Agricultural Salinity Assessment and Management [2nd Revised edition] 0784411697, 978-0-7844-1169-8, 978-0-7844-7648-2

Citation preview

ASCE Manuals and Reports on Engineering Practice No. 71

Agricultural Salinity Assessment and Management, Second Edition Prepared by the Water Quality Technical Committee of the Irrigation and Drainage Council of the Environmental and Water Resources Institute of the American Society of Civil Engineers Edited by Wesley W. Wallender, Ph.D., P.E., and Kenneth K. Tanji, Sc.D.

Library of Congress Cataloging-in-Publication Data Agricultural salinity assessment and management/prepared by the Water Quality Technical Committee of the Irrigation and Drainage Council of the Environmental and Water Resources Institute of the American Society of Civil Engineers ; edited by Kenneth K. Tanji and Wesley W. Wallender. — 2nd ed. p. cm. — (ASCE manual and reports on engineering practice ; no. 71) Includes bibliographical references and index. ISBN 978-0-7844-1169-8 (soft cover : alk. paper) — ISBN 978-0-7844-7648-2 (ebook) 1. Salinization—Control—Handbooks, manuals, etc. 2. Salinization—Environmental aspects—Handbooks, manuals, etc. 3. Irrigation farming—Handbooks, manuals, etc. 4. Agricultural pollution—Handbooks, manuals, etc. 5. Agricultural ecology—Handbooks, manuals, etc. I. Tanji, Kenneth K. II. Wallender, Wesley W. III. Environmental and Water Resources Institute (U.S.). Water Quality Technical Committee. IV. Series: ASCE manuals and reports on engineering practice ; no. 71. S620.A48 2011 628.1'1—dc23 2011030788 Published by American Society of Civil Engineers 1801 Alexander Bell Drive Reston, Virginia 20191 www.asce.org/pubs Any statements expressed in these materials are those of the individual authors and do not necessarily represent the views of ASCE, which takes no responsibility for any statement made herein. No reference made in this publication to any specific method, product, process, or service constitutes or implies an endorsement, recommendation, or warranty thereof by ASCE. The materials are for general information only and do not represent a standard of ASCE, nor are they intended as a reference in purchase specifications, contracts, regulations, statutes, or any other legal document. ASCE makes no representation or warranty of any kind, whether express or implied, concerning the accuracy, completeness, suitability, or utility of any information, apparatus, product, or process discussed in this publication, and assumes no liability therefor. This information should not be used without first securing competent advice with respect to its suitability for any general or specific application. Anyone utilizing this information assumes all liability arising from such use, including but not limited to infringement of any patent or patents. ASCE and American Society of Civil Engineers—Registered in U.S. Patent and Trademark Office. Photocopies and permissions. Permission to photocopy or reproduce material from ASCE publications can be obtained by sending an e-mail to [email protected] or by locating a title in ASCE’s online database (http://cedb.asce.org) and using the “Permission to Reuse” link. Bulk reprints. Information regarding reprints of 100 or more copies is available at http://www.asce.org/reprints. Copyright © 2012 by the American Society of Civil Engineers. All Rights Reserved. ISBN 978-0-7844-1169-8 ISBN 978-0-7844-7648-2 Manufactured in the United States of America. 18 17 16 15 14 13 12 11

12345

MANUALS AND REPORTS ON ENGINEERING PRACTICE (As developed by the ASCE Technical Procedures Committee, July 1930, and revised March 1935, February 1962, and April 1982) A manual or report in this series consists of an orderly presentation of facts on a particular subject, supplemented by an analysis of limitations and applications of these facts. It contains information useful to the average engineer in his or her everyday work, rather than findings that may be useful only occasionally or rarely. It is not in any sense a “standard,” however; nor is it so elementary or so conclusive as to provide a “rule of thumb” for nonengineers. Furthermore, material in this series, in distinction from a paper (which expresses only one person’s observations or opinions), is the work of a committee or group selected to assemble and express information on a specific topic. As often as practicable, the committee is under the direction of one or more of the Technical Divisions and Councils, and the product evolved has been subjected to review by the Executive Committee of the Division or Council. As a step in the process of this review, proposed manuscripts are often brought before the members of the Technical Divisions and Councils for comment, which may serve as the basis for improvement. When published, each work shows the names of the committees by which it was compiled and indicates clearly the several processes through which it has passed in review, in order that its merit may be definitely understood. In February 1962 (and revised in April 1982) the Board of Direction voted to establish a series entitled “Manuals and Reports on Engineering Practice,” to include the Manuals published and authorized to date, future Manuals of Professional Practice, and Reports on Engineering Practice. All such Manual or Report material of the Society would have been refereed in a manner approved by the Board Committee on Publications and would be bound, with applicable discussion, in books similar to past Manuals. Numbering would be consecutive and would be a continuation of present Manual numbers. In some cases of reports of joint committees, bypassing of Journal publications may be authorized.

MANUALS AND REPORTS ON ENGINEERING PRACTICE No.

Title

28 Hydrology Handbook, Second Edition 40 Ground Water Management 45 How to Work Effectively with Consulting Engineers: Getting the Best Project at the Right Price 46 Pipeline Route Selection for Rural and Cross-Country Pipelines 49 Urban Planning Guide 50 Planning and Design Guidelines for Small Craft Harbors, Revised Edition 54 Sedimentation Engineering, Classic Edition 57 Management, Operation and Maintenance of Irrigation and Drainage Systems 60 Gravity Sanitary Sewer Design and Construction, Second Edition 62 Existing Sewer Evaluation and Rehabilitation, Third Edition 66 Structural Plastics Selection Manual 67 Wind Tunnel Studies of Buildings and Structures 71 Agricultural Salinity Assessment and Management, Second Edition 73 Quality in the Constructed Project: A Guide for Owners, Designers, and Constructors, Third Edition 74 Guidelines for Electrical Transmission Line Structural Loading, Third Edition 77 Design and Construction of Urban Stormwater Management Systems 80 Ship Channel Design 81 Guidelines for Cloud Seeding to Augment Precipitation, Second Edition 82 Odor Control in Wastewater Treatment Plants 84 Mechanical Connections in Wood Structures 85 Quality of Ground Water 91 Design of Guyed Electrical Transmission Structures 92 Manhole Inspection and Rehabilitation, Second Edition 93 Crane Safety on Construction Sites 94 Inland Navigation: Locks, Dams, and Channels 95 Urban Subsurface Drainage 96 Guide to Improved Earthquake Performance of Electric Power Systems

No.

Title

97 Hydraulic Modeling Concepts and Practice 98 Conveyance of Residuals from Water and Wastewater Treatment 99 Environmental Site Characterization and Remediation Design Guidance 100 Groundwater Contamination by Organic Pollutants: Analysis and Remediation 101 Underwater Investigations 102 Design Guide for FRP Composite Connections 103 Guide to Hiring and Retaining Great Civil Engineers 104 Recommended Practice for FiberReinforced Polymer Products for Overhead Utility Line Structures 105 Animal Waste Containment in Lagoons 106 Horizontal Auger Boring Projects 107 Ship Channel Design and Operation 108 Pipeline Design for Installation by Horizontal Directional Drilling 109 Biological Nutrient Removal (BNR) Operation in Wastewater Treatment Plants 110 Sedimentation Engineering: Processes, Measurements, Modeling, and Practice 111 Reliability-Based Design of Utility Pole Structures 112 Pipe Bursting Projects 113 Substation Structure Design Guide 114 Performance-Based Design of Structural Steel for Fire Conditions 115 Pipe Ramming Projects 116 Navigation Engineering Practice and Ethical Standards 117 Inspecting Pipeline Installation 118 Belowground Pipeline Networks for Utility Cables 119 Buried Flexible Steel Pipe: Design and Structural Analysis 120 Trenchless Renewal of Culverts and Storm Sewers 121 Safe Operation and Maintenance of Dry Dock Facilities 122 Sediment Dynamics upon Dam Removal

CONTRIBUTORS Manucher Alemi, California Department of Water Resources, Chapter 20 Christopher Amrhein, University of California–Riverside, Chapter 25 R. Aragüés, Agri-food Research and Technology Center of Aragón, Chapter 30 James E. Ayars, Agricultural Research Service, USDA, Chapters 12 and 16 S. E. Benes, California State University–Fresno, Chapter 22 Eduardo Blumwald, University of California–Davis, Chapter 8 A. C. Chang, University of California–Riverside, Chapter 7 W. P. Chen, Chinese Academy of Sciences, Chapter 7 B. Clark, Davids Engineering, Chapter 27 Dennis L. Corwin, U.S. Salinity Laboratory, Chapters 10, 12, and 26 Evan Christen, CSIRO Land and Water, Australia, Chapter 24 Michael Delamore, U.S. Bureau of Reclamation, Chapters 20 and 32 Steven J. Deverel, HydroFocus, Inc., Chapter 4 William Evans, Soil Conservation Service, Chapter 19 Jose I. Faria, California Department of Water Resources, Chapters 20 and 23 Roger Fujii, U.S. Geological Survey, Chapter 4 Suduan Gao, Agricultural Research Service, USDA, Chapter 24 J. R. Gilley, Texas A&M University, Chapter 27 Sabine Goldberg, U.S. Salinity Laboratory, Chapter 4 Stephen R. Grattan, University of California–Davis, Chapters 6, 9, 13, and 22 Catherine M. Grieve, Agricultural Research Service, USDA, Chapter 13 Anil Grover, University of Delhi, Chapter 8 Ardell D. Halvorson, Agricultural Research Service, USDA, Chapter 18 Blaine R. Hanson, University of California–Davis, Chapters 9 and 17 John Hedlund, Soil Conservation Service, Chapter 19 R. W. Hill, Utah State University, Chapter 27 Glenn J. Hoffman, University of Nebraska, Chapter 12 Jan W. Hopmans, University of California–Davis, Chapter 29 D. Isidoro, Agri-food Research and Technology Center of Aragón, Chapter 30 William R. Johnston, Consulting Engineer, Chapter 32 J. J. Jurinak, Utah State University, Chapter 3 S. R. Kaffka, University of California–Davis, Chapter 22 R. Keren, Institute of Soil, Water and Environmental Sciences, Israel, Chapter 21 Keith C. Knapp, University of California–Riverside, Chapter 31 André Läuchli, University of California–Davis, Chapter 6 S. M. Lesch, University of California–Riverside, Chapters 10 and 14 John Letey, University of California–Riverside, Chapter 20 D. B. Lobell, Lawrence Livermore National Laboratory, Chapter 10 v

vi

J. M. Lord, J. M. Lord, Inc., Chapter 27 Eugene V. Maas, U.S. Salinity Laboratory, Chapter 13 Ari M. Michelsen, Texas A&M AgriLife Research, Chapter 33 S. Miyamoto, Texas A&M University, Chapter 21 James D. Oster, University of California–Riverside, Chapters 22 and 25 A. L. Page, University of California–Riverside, Chapter 7 Fred M. Phillips, New Mexico Institute of Mining and Technology, Chapter 33 James Poss, Agricultural Research Service, USDA, Chapter 23 D. Quílez, Agri-food Research and Technology Center of Aragón, Chapter 30 James D. Rhoades, Agricultural Salinity Consulting, Chapters 2 and 26 Jim L. Richardson, National Soil Survey Center, Chapter 18 R. R. Robinson, Coachella Valley Water District, Chapter 27 G. Schoups, Delft University of Technology, Chapter 29 I. Shainberg, Institute of Soil, Water and Environmental Sciences, Israel, Chapter 5 J. Sˇ imu˚ nek, University of California–Riverside, Chapter 26 M. J. Singer, University of California–Davis, Chapter 5 Amanjot Singh, University of Delhi, Chapter 8 E. C. Stegman, North Dakota State University, Chapter 27 Donald L. Suarez, U.S. Salinity Laboratory, Chapters 3, 11, and 28 Kenneth K. Tanji, University of California–Davis, Chapters 1, 15, 27, and 29 Jim Thomas, U.S. Bureau of Reclamation, Chapter 19 Anthony L. Toto, California Regional Water Quality Control Board, Fresno, Chapter 24 Wesley W. Wallender, University of California–Davis, Chapters 1, 15, 27, and 34 Dennis W. Westcot, Consulting Scientist, Chapter 32 Patrick H. Willey, Natural Resources Conservation Service, USDA, Chapter 19 Laosheng Wu, University of California–Riverside, Chapter 25 Charles A. Young, Stockholm Environmental Institute, Chapter 15

BLUE RIBBON REVIEW PANEL Mark E. Grismer, University of California–Davis William F. Ritter, University of Delaware Richard D. Wenberg, Consultant

REVISION TO MANUAL 71 TASK COMMITTEE James E. Ayars Blaine R. Hanson Glen Dale Sanders Donald L. Suarez Wesley W. Wallender Patrick H. Willey

vii

This page intentionally left blank

CONTENTS

FOREWORD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix

PART ONE: INTRODUCTION 1

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salinity and Its Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Extent of Agricultural Salt Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactivity of Salts and Salt Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concerns Over Agricultural Salinity Problems . . . . . . . . . . . . . . . . . . . . The Agricultural Salinity and Drainage Dilemma . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 2 3 6 12 23 23 24

2

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION OF CONTROL PRACTICES: AN OVERVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Potential Adverse Effects of Salts on Soils and Plants: Brief Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diagnosing Salt Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Assessing Reclaimability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Selecting Salinity Control and Management Measures . . . . . . . . . . . . . Using Models to Identify Potential Salinity Problems . . . . . . . . . . . . . .

ix

27 27 31 41 44 52

x

CONTENTS

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

PART TWO: EFFECTS OF SALTS ON SOILS 3

THE CHEMISTRY OF SALT-AFFECTED SOILS AND WATERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Origin of Salt in Soil and Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measurement of Salinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chemistry of Salt-Affected Soil Solutions . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

57 57 62 65 85 85 88

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Processes Affecting Trace Element Concentrations . . . . . . . . . . . . . . . . 90 Biogeochemical Behavior and Distribution of Trace Elements in Soils and Groundwater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Evaluation of Pollution Potential in Soils . . . . . . . . . . . . . . . . . . . . . . . . . 118 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 5

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Solution Composition and Clay Swelling and Dispersion . . . . . . . . . . . 141 Solution Composition and the Hydraulic Conductivity of Sodic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 The Infiltration Rate in Sodic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Depositional Crust: Effect of Salinity and Sodicity . . . . . . . . . . . . . . . . . 152 Effect of Extrinsic Physical Conditions on Susceptibility of Soils to Sodicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Effect of Polyacrylamide and Gypsum on Runoff and Erosion from Sodic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

CONTENTS

xi

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 6

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Salinity and Sodicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Principal Responses of Plants to Salinity . . . . . . . . . . . . . . . . . . . . . . . . . 171 Mechanisms of Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Integration in the Whole Plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 7

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Factors Influencing Trace Element Deficiencies and Toxicities of Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Methods for Diagnosing Trace Element Deficiencies and Toxicities . . . 213 Accumulation of Trace Elements in Vegetation to Levels Potentially Harmful to Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 Assessing the Fates of Trace Elements in Soils . . . . . . . . . . . . . . . . . . . . 223 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228 8

TRANSGENIC STRATEGIES TOWARD THE DEVELOPMENT OF SALT-TOLERANT PLANTS . . . . . . . . . . . 235

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 Conclusions and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261

PART THREE: SAMPLING, MONITORING, AND MEASUREMENT 9

FIELD SAMPLING OF SOIL, WATER, AND PLANTS . . . . . . . 275

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 Sampling Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276

xii

CONTENTS

Considerations in Developing Sampling Strategies . . . . . . . . . . . . . . . . 279 Evaluations of Sampling Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284 Sampling Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293 10 LABORATORY AND FIELD MEASUREMENTS . . . . . . . . . . . . 295 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295 Factors Affecting Soil Salinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 Direct and Indirect Analysis of Soil Salinity . . . . . . . . . . . . . . . . . . . . . . 297 Methods of Laboratory, Lysimeter, and Plot-Scale Soil Salinity Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Soil-Related (Edaphic) Factors Influencing the ECa Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Methods of Field-Scale Soil Salinity Measurement . . . . . . . . . . . . . . . . . 316 Use of Remote Imagery for Measuring Soil Salinity at Field and Landscape Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340

PART FOUR: DIAGNOSIS OF SALT PROBLEMS 11 IRRIGATION WATER QUALITY ASSESSMENTS . . . . . . . . . . 343 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 Salinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Sodicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352 Ionic Balances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362 Boron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 Trace Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 12 LEACHING AND ROOTZONE SALINITY CONTROL . . . . . . . 371 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 Water and Salt Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371

CONTENTS

xiii

Leaching Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 Effect of Shallow Groundwater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380 Soil Salinity Without Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Integration of Soil Salinity by Crops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 13 PLANT SALT TOLERANCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Source and Causes of Soil Salinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 Plant Salt Tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407 Salinity and Nutritional Imbalance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Crop Response to Specific Ions and Elements . . . . . . . . . . . . . . . . . . . . . 429 Parameters Influencing Plant Response to Salt Stress . . . . . . . . . . . . . . 440 Plant Tolerance to Saline Sprinkling Waters . . . . . . . . . . . . . . . . . . . . . . 445 Controlling Soil Salinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 14 STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE AND SPATIAL SALINITY PATTERNS FROM SOIL CONDUCTIVITY SURVEY DATA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482 15 SPATIALLY DISTRIBUTED SOLUTE BALANCE IN A CALIFORNIA WATER DISTRICT . . . . . . . . . . . . . . . . . . . . 483 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483 Local Geology of the Panoche Water District . . . . . . . . . . . . . . . . . . . . . . 484 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510

xiv

CONTENTS

PART FIVE: SALINITY MANAGEMENT OPTIONS 16 ON-FARM IRRIGATION AND DRAINAGE PRACTICES . . . . 511 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 Irrigation and Salinity Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 Drainage and Salinity Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524 Management of Shallow Groundwaters . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 17 DRIP IRRIGATION AND SALINITY . . . . . . . . . . . . . . . . . . . . . . . 539 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559 18 MANAGEMENT OF DRYLAND SALINE SEEPS . . . . . . . . . . . . 561 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561 Factors Contributing to Saline Seep Development . . . . . . . . . . . . . . . . . 562 Types of Saline Seeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564 Water Quality Associated with Saline Seeps . . . . . . . . . . . . . . . . . . . . . . 567 Identification of Recharge and Discharge (Seep) Areas . . . . . . . . . . . . . 569 Methods for Controlling Saline Seeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 Reclamation of Controlled Saline Seep Areas . . . . . . . . . . . . . . . . . . . . . 579 Socioeconomic Concerns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 582 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584 19 PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS . . . . 591 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591 Planning Regional Management Programs . . . . . . . . . . . . . . . . . . . . . . . 591 Salinity Control Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 Project-Level Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598 Legal and Institutional Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602 Salinity Reduction by Improving Water Supplies and Irrigation Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613

CONTENTS

xv

20 SAN JOAQUIN VALLEY, CALIFORNIA, DRAINAGE MANAGEMENT OPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651

PART SIX: LAND RECLAMATION, TREATMENT AND DISPOSAL OF DRAINAGE WATERS 21 RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655 Effects of Salinity and Adsorbed Ions on Soil Properties . . . . . . . . . . . . 656 Reclamation of Salt-Affected Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659 Reclamation of Sodium-Affected Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Reclamation of Boron-Affected Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 676 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 678 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684 22 USE OF SALINE DRAINAGE WATERS FOR IRRIGATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687 Irrigation with Saline Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688 Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 711 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 23 DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721 Treatment of Agricultural Drainage Water . . . . . . . . . . . . . . . . . . . . . . . 722 Trace Element Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729 Disposal Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 733 Reclamation and Reuse Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 751 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755

xvi

CONTENTS

24 AGRICULTURAL EVAPORATION BASINS . . . . . . . . . . . . . . . . 757 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 780 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 25 SALINITY ASSESSMENT OF IRRIGATION WATER USING WATSUIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800 26 LEACHING REQUIREMENT: STEADY-STATE VERSUS TRANSIENT MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . 801 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 801 General Description of Models Used to Estimate Leaching Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805 WATSUIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807 UNSATCHEM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 Model Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 Model Leaching Requirement Estimates . . . . . . . . . . . . . . . . . . . . . . . . . 810 Implications of Leaching Requirement Model Estimates . . . . . . . . . . . . 818 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824 27 CONCEPTUAL WATER FLOW AND SALT TRANSPORT FOR FLUX-LIMITED AND PONDED INFILTRATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825 An Overview of Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826 Water Flow Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827 Solute Transport Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851

CONTENTS

xvii

28 MODELING TRANSIENT ROOTZONE SALINITY (SWS MODEL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 855 Model Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 SWS Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 858 Soil and Water Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 SWS Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 893 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894 29 LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 920 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 922 30 CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923 Description of CIRFLE Hydrosalinity Model . . . . . . . . . . . . . . . . . . . . . . 930 Model Calibration and Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 936 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938 Model Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943 Model Limitations, Model Improvements, and Future Research Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952 31 MICROECONOMICS OF SALINITY AND DRAINAGE MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 953 Dynamic Irrigation Management for a Single Season . . . . . . . . . . . . . . 954 Multiyear Irrigation Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 960

xviii

CONTENTS

Farm and Regional Agricultural Production . . . . . . . . . . . . . . . . . . . . . . 966 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 972 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 974 Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 975 32 SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY . . . . 977 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 977 The Valley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 978 Efforts to Provide Agricultural Drainage for the San Joaquin Valley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 985 Drainage Litigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 995 On-Farm Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1003 Drainage Control Outside the San Luis Unit Service Area . . . . . . . . . 1003 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1024 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1025 33 INSTITUTIONAL AND SALINITY ISSUES ON THE UPPER RIO GRANDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033 Rio Grande Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1033 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1049 34 VIABILITY OF IRRIGATED AGRICULTURE WITH EXPANDING SPACE AND TIME SCALES . . . . . . . . . . . . . . . . . 1053 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053 Farm Control Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1054 Extending the Spatial and Temporal Extent . . . . . . . . . . . . . . . . . . . . . . 1057 Irrigation District Control Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058 Regional Control Volume Management . . . . . . . . . . . . . . . . . . . . . . . . . 1061 Framework for Management Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 1063 Agricultural Viability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1067 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1069 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1071 INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1077

FOREWORD

This updated ASCE manual of practice considers salinity and trace element management in irrigated agriculture and water supplies worldwide. It is comprised of 34 chapters written by academic, agency, and consulting experts whose training ranges from the basic sciences to applied sciences and engineering. While there may appear to be some redundancy in topic coverage among the chapters, this enables each chapter to stand independently, such that if, for example, the reader needs more in-depth information on plant responses to saline/sodic soils after reading a related waterquality chapter, it is readily accessible. Wes Wallender has guided this updated manual to build and expand on the original that was assembled and published under Ken Tanji’s editorship nearly 20 years ago, with the stated goal to be “a reference to help sustain irrigated agriculture.” This manual goes well beyond that need and should accessed by water professionals interested in developing and managing ever-constrained water supplies worldwide. Nature and Extent of Agricultural Salinity. This chapter presents an overview of the nature of salinity in soils and waters, its extent from global to regional scales, the reactivity of salts and salt flows, and the concerns of agriculture and other sectors of society. Note in the opening paragraphs that, in fact, saline drainage water discharge has been challenged and regulations presently exist in California and elsewhere as part of basin plans. Chapter 2: Overview: Diagnosis of Salinity Problems and Selection of Control Practices. This critical chapter summarizes the principal adverse effects of salts on soils and plants, and describes a methodology for diagnosing the nature and cause of salinity problems, while providing guidance on assessment and selection of appropriate reclamation and management practices. This is a difficult area for many practicing engineers; ample guidance is provided here for developing evaluation and monitoring programs for impacted soils. Chapter 3: The Chemistry of Salt-Affected Soils and Waters. When greater in-depth knowledge is needed, this chapter provides essential background information about the weathering process and geochemical xix

xx

FOREWORD

reactions in soils. Don Suarez describes the basic processes, as well as currently available chemical equilibrium-type models. Chapter 4: Chemistry of Trace Elements in Soils and Groundwater. This is a very well written complementary chapter to the previous chapter, with the objective of providing a framework for the factors and processes to consider when evaluating trace element concentrations in soils and shallow groundwater of irrigated agricultural areas. The focus here is on the description of the U.S. Environmental Protection Agency priority trace element pollutants of concern. This information is becoming more crucial to water supplies across the globe as water managers face dilemmas associated with trace element contamination. Chapter 5: Soil Response to Saline and Sodic Conditions. In field soils, accumulation of sodium in the soil solution and the exchange phase generally adversely affect soil physical properties critical to watershed processes affecting drainage and groundwater quality, such as structural stability, hydraulic conductivity, infiltration rate, runoff, and erosion. While extensive reviews of saline-sodic effects on soil properties in laboratory samples exist, Isaac Shainberg notes that “these conditions do not prevail in the field where slower wetting rate and ageing at different AMCs decrease the susceptibility of soils to sodic conditions.” He then seeks “to demonstrate the effect of inherent soil properties and time-dependent physical conditions on the susceptibility of soils to sodic conditions.” This is a critical distinction in the field as lab results to date have had little success in predicting field responses to saline-sodic conditions. At the same time, however, acquiring an understanding of these processes has guided water managers and will continue to do so as better field-applicable knowledge is developed. Chapter 6: Plant Responses to Saline and Sodic Conditions. This is a companion chapter to Chapter 5 in providing a general overview of the principal mechanisms and crop responses to salinity and sodicity stress available in scientific literature. Because many water managers and practitioners may be less knowledgeable in the plant sciences, this is an important chapter to have available, though it may seem a little daunting. Lauchli and Grattan underscore that, since the previous edition of the manual, “our knowledge and understanding of the physiological mechanisms of salt tolerance in plants has greatly increased” with ever-more rapid progress “made in elucidating molecular and genetic aspects of salt tolerance in plants.” However, as shown in the previous chapter, as crops and plants integrate a number of environmental stresses, their responses in the field have been less predictable from the laboratory results considering less than a handful of stress factors in any one set of experiments. Nonetheless, more exciting research opportunities are readily available. Chapter 7: Deficiencies and Toxicities of Trace Elements. In this review of general trace element chemistry in soils, Chen, Chang, and Page

FOREWORD

xxi

consider the factors affecting their availability, mobility, and accumulation by crops, as well as their fate and transport through a generalized mass balance-type model. For many trace elements there is a narrow window of concentrations in soil solutions between deficient and toxic conditions for either the plant or those that consume the plants; understanding something of this window and the processes affecting it is ever-more important. This is essential information because various recycled and treated waters are increasingly reused in irrigated agriculture, as well as from the perspective of water managers considering the quality of these waters. Chapter 8: Transgenic Strategies toward the Development of SaltTolerant Plants. Understanding plant response stresses associated with salinity, sodicity, and trace elements sets the stage for developing alternative, or transgenic, plants that better tolerate some of these stresses brought on by the changing environmental conditions facing irrigated agriculture. This chapter is one of hope; it summarizes some of these opportunities and how they may play out. For example, boron phytotoxicity is a problem for citrus production in many salinity-affected areas, though the general irrigation water quality is acceptable for a wide variety of other uses (including human consumption). Development of a transgenic lemon rootstock would facilitate sustaining the citrus industry in such areas. Chapter 9: Field Sampling of Soil, Water, and Plants. Field sampling and monitoring is critical toward evaluation and assessment of soil-water processes and plant responses to those processes. Unfortunately, this dimension of investigation is often lacking or entirely absent in diagnosing various environmental problems. Hanson and Grattan have the necessary background and provide a straightforward description of what is needed to develop a sampling strategy for different requirements. Integrated simultaneous sampling and monitoring of the soil, waters, and plants are essential to furthering research in these irrigated systems. This chapter provides a basic starting point for the practitioner to design and establish a field monitoring and sampling program. Chapter 10: Laboratory and Field Measurements. Management of the plant rootzone is the focus of irrigated agriculture, and this chapter begins with a description of the “leaching fraction” concept (a more in-depth review of leaching fraction is provided in Chapter 12) and the processes associated with rootzone water and salt balances. The chapter proceeds to describe laboratory and field measurements of soil-water content and salinity with the goal of determining localized or field-wide salinity. The saturation extract, often used to reflect soil-solution chemistry, bears further consideration; this is covered in Chapter 11. Chapter 11: Irrigation Water Quality Assessments. Irrigation water quality is a key component related to sustainability of the soil-water system affecting soils and plants. As available water resources are strained by

xxii

FOREWORD

competing demands, a thorough understanding of the impacts of irrigation water quality is needed. This chapter summarizes the previous work on the effects of irrigation water quality on soils and crop production. The continued weak link in this area is the relationship between irrigation water quality and soil-solution chemistry; guidelines are developed for irrigation water presuming more-or-less parity between irrigation water quality and that of the soil solution. Don Suarez underscores the difference in reviewing the soil-water processes and older work on leaching fractions (also considered from a field perspective in Chapter 12). Chapter 12: Leaching and Rootzone Salinity Control. In regions lacking adequate rainfall leaching of the soil profile salinity and trace elements, irrigation must be carefully managed so as to maintain favorable rootzone water quality without adversely affecting “downstream” (e.g., ground water) water resources. With considerable field experience, Jim Ayars and his colleagues outline the primary rootzone mass balance processes central to determining adequate leaching. They build on the old assumptions of steady-state balances but note that these are effectively conservative in their estimation of leaching requirements. That is, the gap between the actual leaching needed and that suggested by the estimated requirement is narrowing and, together with improved irrigation methods, enabling a closer match between crop water requirements and actual irrigation, so that irrigated agriculture may find sustainability. Chapter 13: Plant Salt Tolerance. Determination of adequate leaching fractions or requirements is driven in part by knowledge of the salt tolerance of the desired crop. Salt and trace element tolerance thresholds of various crops have been studied for decades and are summarized by these same authors in several publications. Together with the information in Chapter 11, water managers must consider the tolerance thresholds to evaluate various water supplies and their applicability, as well as downstream impacts. Unfortunately, threshold information remains based on the saturation extract or soil-solution concentrations (see Chapter 10) from lab or greenhouse studies on generally coarser-textured soils, and their field applicability combined with assessments of irrigation water quality leave some regulators with a gray zone that is difficult to address. Some thresholds are based on research from the 1930s (e.g., citrus) for which updated information is needed related to more recently available rootstocks. Nonetheless, the information of this chapter is critical to the practitioner as a starting point in designing appropriate irrigation strategies. Chapter 14: Statistical Models for the Prediction of Field-Scale and Spatial Salinity Patterns from Soil Conductivity Survey Data. This chapter continues and is complementary to the discussion begun in Chapter 10 about field-wide assessment of soil salinity and the appropriate statistical evaluations necessary to complete that spatial assessment using

FOREWORD

xxiii

the soil conductivity survey information. This chapter provides useful information when soil conductivity assessments are deployed. Chapter 15: Spatially Distributed Solute Balance in a California Water District. This chapter illustrates the benefits of assessments of salinity and mass balance modeling at the larger spatial scales (field to district), and is complementary to Chapters 19 and 20. It combines the information presented in previous chapters into the modeling framework in an effort to determine area-wide salinity and some trace element concentrations, as well as the water quality of the drainage water “leaving” the region, using data available from the water district. Chapter 16: On-Farm Irrigation and Drainage Practices. For the practitioner unfamiliar with commonly available irrigation and drainage systems in use, this chapter provides a brief, general overview and outlines management options that can help prevent or correct salinity problems and minimize water table build-up. Again, Ayars underscores the age-old importance of drainage for salinity management, as well as methods for management of shallow water tables, and an alternative method for disposal of saline drainage water. This and the following chapter are good background reading for those who need this information. Chapter 17: Drip Irrigation and Salinity. Drip irrigation is one method that enables fairly high irrigation distribution uniformity and possible salinity control of the rootzone through more precise application of water and fertilizers than is otherwise possible from surface irrigation methods. This chapter provides a basic description of the factors important to drip irrigation system design for adequate irrigation and possible salinity control. Chapter 18: Management of Dryland Saline Seeps. Saline seeps can result from dryland farming in the Plains states and are a source of downstream degradation and locations of poor, if any, crop production in the landscape. The origin of such seeps is not always readily apparent. This chapter considers the identification, control, and possible reclamation of seep areas. Because seeps occur irrespective of property or government lines and are part of the overall watershed, their identification and reclamation is an important water quality consideration. Chapter 19: Project-Level Salinity Management Options. In order to address water and salinity management across a wide geographical area that transcends water district and various jurisdictional boundaries, both local- and regional-scale management plans must be coordinated across institutional entities—rarely a simple process. This chapter outlines an approach and provides examples of how such management may be achieved, and follows from the modeling example given in Chapter 15. Chapter 20: San Joaquin Valley Drainage Management Options. The San Joaquin Valley Drainage Program is an exceptional example of an interagency entity established to address the contamination associated

xxiv

FOREWORD

with subsurface drainage waters as evidenced in the Kesterson Reservoir in the early 1980s. The program continues to-date and has implemented a range of strategies to address the drainage water disposal issue, including source control through land retirement, disposal via evaporation ponds and salt accumulation (solar evaporator) cells, crop rotations through reuse of progressively greater salinity drain waters, and by-product recycling. Chapter 21: Reclamation of Saline, Sodic, and Boron-Affected Soils. If there is inadequate control of irrigation and soil-water quality such that irrigated soils become salinized and/or otherwise compromised through high boron concentrations, additional reclamation beyond regular leaching is required. In some cases mineral composition of the affected areas limits the options available for reclamation, but the understanding of the reclamation process remains important and may be essential toward development of increased production. Considering various soil processes, Rami Keren reviews the relative success of leaching and addition of amendments to reclaim salt-affected soils, as well as the chemical modeling associated with such processes. Reclamation of boron-affected soils remains problematic, especially in tourmaline-based soils. Chapter 22: Use of Saline Drainage Waters for Irrigation. This chapter is an extension or continuation of Chapters 12 and 13. The discussion considers several practices using saline water for irrigation, including both their benefits and limitations. Trace elements, such as B, Se, and Mo, may also influence the feasibility of using saline-sodic water for irrigation. The same soil-water and plant tolerance principles apply, but special consideration must be given to potential salt or phytotoxic ion accumulations. Clearly, when using saline waters for irrigation there is an emphasis on control of soil salination and adverse effects on soil physical properties when irrigation waters are sodic as well as saline. In addition, as part of water reuse strategies, there is consideration of salt-tolerant plants and crops. Examples are provided of drainage water reuse studies from the San Joaquin Valley Drainage Program. Chapter 23: Drainage Water Treatment and Disposal Options. As considered in the previous chapter, reuse of saline drainage waters for irrigation is one part of the drainage water disposal scheme; others involve possible treatment and alternative disposal methods. Some of these are covered in general terms as part of the San Joaquin Valley Drainage Program report considered in Chapter 20, but the focus here is on the possible technologies available or under development for drain water treatment and final disposal. They underscore the magnitude of the salt disposal problem by noting that “2 to 3 million tons of annually imported salts (in addition to significant amounts of salt mobilized from soils as a result of irrigation) needs to be disposed of to maintain salt balance in the San Joaquin Valley,” and “even an optimistic estimate of the amount that could be commercially marketed would represent a only a small percent-

FOREWORD

xxv

age of the total salts needing to be disposed. Active pursuit of commercial utilization of the salts and selenium is needed and will require a variety of options for separating the salts from productive agricultural fields” before sustainability is achievable. Chapter 24: Agricultural Evaporation Basins. Another disposal option for drain water from irrigated agriculture is through evaporation from open, shallow basins. These basins pose environmental hazards associated with aquatic and bird life and groundwater contamination; understanding the biogeochemical processes occurring in these basins is critical when designing agricultural irrigation systems when drain water disposal elsewhere is not a viable option. This chapter summarizes “their siting, design, operation, chemical/biological characteristics, and environmental and regulatory issues.” While such basins are a temporary disposal method, they are often linked with other strategies, as outlined in Chapter 20. Chapter 25: Salinity Assessment of Irrigation Water Using WATSUIT. Development and application of a rootzone water chemistry model is instrumental when evaluating the suitability of irrigation water and determining the appropriate leaching fractions needed to maintain the soil-water chemistry within an acceptable range for the crop of interest (discussed in Chapter 13). WATSUIT is a steady-state model due to the complexity of chemical reactions but is nonetheless useful in organizing the various processes outlined in Chapter 7 for trace elements and developing a first-approximation assessment. Chapter 26: Leaching Requirement: Steady-State versus Transient Models. This is an interesting chapter outlining steady-state and transient modeling of the leaching process and determination of the leaching fraction of irrigated soils. This chapter is continuation of that material in Chapter 12. The verdict on the operational value of transient leaching fraction evaluation is not yet finalized, but the use of the dynamic soil-water content (unsaturated flow) together with basic chemistry models is an excellent concept depending on how one interprets the meaning of “leaching fraction.” Chapter 27: Conceptual Water Flow and Salt Transport for FluxLimited and Ponded Infiltration. This chapter reviews another model development for rootzone water and salinity dynamics with the added component of bypass flows associated with cracking clay soils. As with the previous two chapters, it outlines the key factors that affect rootzone salinity and that guide possible management strategies. Chapter 28: Modeling Transient Root Zone Salinity. The objective of a management model should be to represent the underlying process without undue burden on the user for collection of site specific characterization or parameter information. This chapter describes development of another rootzone salinity model (SWS), the processes used by the model, as well as applications to management of saline soils or waters. Together

xxvi

FOREWORD

with Chapters 25, 26, and 27, these chapters provide a comprehensive summary of rootzone modeling efforts, their advantages and disadvantages, and their overall applicability to particular problems of interest. With some effort, the average reader or practitioner can ferret through this group of chapters to extract the information needed to determine which model is best applied in the particular situation. Chapter 29: Long-Term Regional-Scale Modeling of Soil Salinity. While each of the modeling chapters considers rootzone water content and chemistry dynamics, they differ in the scale of application or process detail. This chapter describe the development and application of a rootzone model to larger regional areas of the San Joaquin Valley. Its sister chapters are Chapters 15 and 20. In evaluating long-term salinity variations across the region and possible model simplifications that reduce computing time and data requirements, Hopmans, Schoups, and Tanji find that “a simplified modeling approach can be used with annually averaged boundary conditions and a relatively coarse spatial discretization, but that it must include cation exchange and gypsum dissolutionprecipitation reactions.” Perhaps more significant is the role of regional groundwater dynamics and salinity in controlling rootzone salinity. Chapter 30: Conceptual Irrigation Project Hydrosalinity Model. This chapter considers another steady-state hydrosalinity model having similar roots to that developed in Chapter 29 but with an application toward assessing drain-water salinity as part of Total Maximum Daily Load (TMDL) establishment from an irrigated project area. The simplified model requires “only 25” input parameters and is readily available for use on PCs. Chapter 31: Microeconomics of Salinity and Drainage Management. This chapter begins a change in focus regarding the economics of salinity management in irrigated agriculture. This is an important component of the social decisions that balance the needs of food production and environmental preservation. The framework for these salinity and drainage management models follows from economic theory. Process-based models from the physical and biological sciences are incorporated as needed. The models are “used to explore characteristics of the underlying system, develop efficient management practices, and identify shadow values and policy instruments to achieve efficient and equitable solutions to salinity and drainage problems.” The economic models differ from those described in the previous chapters in that they are primarily dynamic optimization models, although some static analyses are included as appropriate. The economic-type models indicate that interseasonal variability for single-crop systems resolve quickly into steady-state conditions, suggesting that steady-state-type hydrosalinity models are adequate at the larger spatial and time scales. Mixed results with future unknown conditional costs are obtained for determinations of whether irrigation uniformity as source

FOREWORD

xxvii

control, drain water reuse, and land retirement options effectively enable sustainable irrigated agriculture. Chapter 32: San Joaquin Valley, California: A Case Study. Case studies provide a living history of the problems, issues, and possible solutions attempted, and the lessons learned there from, at least from the perspective of the storyteller. This chapter considers the story of the San Joaquin Valley of California from the perspective of a former water district engineer/manager, a regulator, and a federal agency engineer. As the authors note, the ultimate solution to the Valley’s drain water disposal problems remains “elusive.” The first effort is to “find a long-term solution to permanently dispose of the salty residue from the irrigated lands, and the second is implementation of measures to relieve the immediate drainage problems” while awaiting resolution of the first. Thus far, many of the methods “to reuse and dispose of drainage water have been driven by regulatory needs for water quality and environmental protection rather than considerations of long-term hydrologic and basin protection that must still be outlined in a long-term strategy.” This shift in emphasis from salinity management has occurred as a result of the selenium contamination issues that arose at the Kesterson Reservoir. While the previous statement may appear contradictory, it reflects the ongoing story of the San Joaquin Valley agriculture and is good reading for all interested. The reader is referred to Chapter 20 for a basic background on the San Joaquin Valley Drainage Program. Chapter 33: Institutional and Salinity Issues on the Upper Rio Grande. The Rio Grande River system is a fascinating complex of issues involving multiple states, Native American communities, and reservoirs, and two countries. With the oldest gaging station in the United States, it also has the longest record of information. This chapter is a case study of the Rio Grande basin from the perspective of hydrologic and institutional management issues, sources and impacts of salinity, and potential management alternatives. In nearly all semiarid region rivers, water salinity increases from the headwaters to the discharge point, with irrigated agriculture exacerbating this effect. An interesting outcome of the Rio Grande studies is that earlier agriculture and waterlogging undoubtedly account for some of the increasing salinity downstream, but that the lower-basin geology is such that salts are leached directly into the river and concentrated in major reservoirs. This is a good history and story that well complements that of the San Joaquin Valley. Chapter 34: Viability of Irrigated Agriculture with Expanding Space and Time Scales. This final chapter is an effort to wrap up all of the discussions in the text, and considers the range of spatial and time scales at play in irrigated agriculture and how they may affect its future viability. This chapter sets a context for the manual and in some ways complements the opening chapter as a conceptual discussion of what is discussed

xxviii

FOREWORD

throughout the manual. As noted, this is good reading for the generalist interested in irrigated agriculture and its future in semiarid and arid regions affected by salinity. Overall, this updated and revised manual continues the Tanji legacy of eliciting worldwide cooperation and integration of scientists and engineers to address the problems of salinity and trace element management in irrigated agriculture. It is never a small feat to bring together such a range of talents and characters. Mark E. Grismer, Chair, Peer Review Committee

PREFACE

Under the editorship of Kenneth K. Tanji, ASCE Manuals and Reports on Engineering Practice No. 71, Agricultural Salinity Assessment and Management, was first published in 1990. In the years following, Ken recognized the significant gains in the knowledge base and advances in technology and initiated the process of revising the manual. In their July 2002 meeting, the Water Quality and Drainage Technical Committee agreed to the recommendation of Editor Tanji to revise the manual. A formal proposal for a Task Committee to revise the manual was first submitted on or about August 2002 to the Irrigation and Drainage Council of ASCE/EWRI (Environmental and Water Research Institute), and at that time the effort was to start on October 1, 2002 and end on September 30, 2005. After approval in 2004, the Task Committee was launched in January 2005 and, by October 2006, it had determined the revisions desired, the authors of existing chapters to be revised, and the authors for new chapters. Drafts of chapters were requested to be completed on or about March 2007, with the goal of completing a complete draft by December 2007. By September 2007, about onethird of the 34 chapters had been submitted. Sadly, Ken passed away suddenly on Friday, September 7, 2007, and this left an enormous void in leadership and vision—not only regarding the completion of Manual 71 but also in expertise in agricultural salinity and management. Wes Wallender volunteered and was approved by the Irrigation and Drainage Council as the editor to complete one of Ken’s many great works. By the fall of 2008, all the chapters had been submitted. They were technically reviewed by the Task Committee (each chapter was reviewed by a member of the Committee) and revised by the authors. The draft document, finalized by the Committee, was given to ASCE for review. The Irrigation and Drainage Council appointed a Review Committee of three members of ASCE, comprising a balance of interests, with expertise in the subject matter. They were independent from the Task Committee and were not authors. Their written reviews were completed by September 2009 and submitted to the Task Committee. By October 2, 2009, the Committee found no differences between reviews to resolve and agreed to make all the suggested changes. All of the reviewers’ suggested changes were made. xxix

xxx

PREFACE

The Kearney Foundation of Soil Science of the University of California provided support for editing. Elizabeth Grieve, Suduan Gao, Sharon Benes, Arnold Bloom, James Poss, Lola Quiles, and Richard Adams provided technical editorial assistance to the Revision to Manual 71 Task Committee. Judson Monroe provided copyediting services.

PART ONE: INTRODUCTION

CHAPTER 1 NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY Kenneth K. Tanji and Wesley W. Wallender

INTRODUCTION Irrigated agriculture has faced the challenge of sustaining its productivity for centuries. Because of natural hydrological and geochemical factors, as well as irrigation-induced activities, soil and water salinity and associated drainage problems continue to plague agriculture. The problems have extended far beyond the farmlands, where saline soils and waters impair crop production. Practices based on the presumption that saline drainage waters will somehow, somewhere, be discharged are now being challenged. New and more restrictive regulations on the discharge of nonpoint source pollutants in agricultural drainage waters are expected in the United States. Issues related to salts in soils include the concentration of salts (salinity) and the composition of sodium relative to calcium and magnesium (sodicity). Salinity concentrations have direct effects on plants independent of other soil conditions. Sodicity can contribute to the deterioration of soil physical properties, which can indirectly affect plants via crusting, reduced infiltration, increased soil strength, and reduced aeration resulting in anoxic or hypoxic conditions for roots. The viability of irrigated agriculture is affected by salinity, sodicity, and, in some cases, concentrations of trace elements. This chapter presents an overview of the nature of salinity in soils and waters, its extent from global to regional scales, the reactivity of salts and salt flows, and the related agricultural and socioeconomic issues.

1

2

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SALINITY AND ITS SOURCES Salinity Constituents Salinity is the concentration of dissolved mineral salts present in waters and soils on a unit volume or weight basis. The major solutes comprising dissolved mineral salts are the cations Na, Ca, Mg, and K and the anions Cl, SO4, HCO3, CO3, and NO3. Other constituents contributing toward salinity in hypersaline waters include B, Sr, Li, SiO2, Rb, F, Mo, Mn, Ba, and Al. Salinity Parameters Salinity is expressed in a number of ways, depending on the method and purpose of the measurements. The salinity constituents listed are frequently reported in terms of mol(c)/L (equivalents per liter) or mg/L (ppm) for major solutes and g/L (ppb) for trace elements. Salinity is often expressed as a lumped parameter, e.g., electrical conductivity (EC), an intensive parameter; total dissolved solids (TDS), an extensive gravimetric measure in mg/L; or total concentration of soluble cations (TSC) and anions (TSA) in mol(c)/L. The EC of saline soils and waters is reported in decisiemens per meter (dS/m, which is equivalent to millimhos per cm), and the EC of lesser saline soils and waters in dS/m  103 (or microsiemens/cm, which is equivalent to micromhos per cm). No exact relationship exists between these measures of lumped salinity parameters, but TDS may be approximated by multiplying EC (dS/m) by a factor of 640 for lesser saline samples, and a factor of 800 for hypersaline samples. To obtain TSC or TSA, multiply EC (dS/m) by a factor of 0.1 for mol(c)/L and a factor of 10.0 for mmol(c)/L. Salinity Measurements The measurement of salinity in waters for EC, TDS, TSC, and TSA is straightforward. In contrast, soil-water contents significantly affect the measurement of salinity as a lumped parameter or dissolved mineral contents in soils. Soil salinity is typically measured (1) in a saturation soil extract, (2) in soil solutions extracted by vacuum in the field, or (3) by electroconductimetric methods. The concentration of salts in the soil solution does not typically change in proportion to change in soil-water contents, because the major solute species participate in sink/source mechanisms, such as mineral precipitation and dissolution, cation exchange, and ion association (Tanji et al. 1967). Moreover, salinity is a dynamic property in soil-water systems, as it is highly mobile and related to water content.

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

3

Salinity Criterion The criterion used for salinity hazard depends on the use of the water or soil. For instance, an irrigation water of EC  0.7 dS/m poses little or no threat to most crops, whereas EC  3.0 dS/m may restrict the growth of most crops (Ayers and Westcot 1985). A soil with an EC  4 dS/m in the soil saturation extract is traditionally characterized as a saline soil (Richards 1954). Soil and water salinity decreases the availability of soilwater and leads to reduced germination, growth, and yields. Furthermore, certain constituents in waters or soils, such as B, may be toxic to plant growth. Others, such as Na, detrimentally affect the soil’s physical properties, such as its infiltration rate. Therefore, characterizing a water or soil as saline is relative and may vary widely, since responses by plants and soils to salinity are highly variable. Sources of Salts The problem of salinity manifests itself in the environment in a number of ways: saline irrigation and drainage waters, saline and sodic soils, saline groundwaters, seawater intrusion, brines from natural salt deposits or geologic formations, and brines from oil and gas fields and mining. The primary source of salts in waters and soils is the chemical weathering of earth materials, i.e., minerals that are constituents of rocks and soils. Evaporative salinization, e.g., surface evaporation of water and transpiration by plants, and dilution, e.g., rainfall, snowmelt waters, and irrigation waters, affect the level of concentration of dissolved mineral salts. Mineral solubility principally regulates the extent to which salts accumulate or dissolve. Natural secondary sources of salts include atmospheric deposits of oceanic salts along coastal areas; seawater intrusion into estuaries due to tidal events; seawater intrusion into groundwater basins in coastal areas due to overdraft; saline water from rising groundwaters, inland saline lakes, and playas; and leaching of saline land forms. Anthropogenic sources of salts include irrigation and drainage waters, soil and water amendments, animals manures and wastes, chemical fertilizers, sewage sludges and effluents, and oil and gas field brines.

EXTENT OF AGRICULTURAL SALT PROBLEMS Global Scale Most of the water in the hydrosphere is salty, and much of the fresh water is frozen. Figure 1-1 shows that the oceans contain about 97.3% of

4

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 1-1. Global water balance and fluxes. From Todd (1970).

the water, continents about 2.8%, and the atmosphere about 0.001% (Todd 1970). About 77.2% of the water associated with land is found in ice caps and glaciers and about 22% in groundwaters, much of which is economically irretrievable. This leaves only a small percentage of readily manageable fresh water as a resource for water supply. The world’s land surface occupies about 13.2  109 ha, no more than 7  109 ha of which are arable and only 1.5  109 ha of which are cultivated (Massoud 1981). Of the cultivated lands, about 0.34  109 ha (23%) are saline and another 0.56  109 ha (37%) are sodic. Figure 1-2 shows that saline and sodic soils cover about 10% of total arable lands and exist in more than 100 countries. Salt-affected soils are not limited to semiarid to arid regions; in several other regions the climate and mobility of salts produce saline waters and soil seasonally.

FIGURE 1-2. Global distribution of salt-affected soils. From Szabolcs (1989) with permission from FAO. 5

6

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

United States Figure 1-3 shows areas with the potential for problems with water and soil salinity in the conterminous United States. The map is based on spatial scales of river basins or watersheds and salt indicators, which include TDS and Cl in waters and salts in surface soils and geologic formations. About 30% of the land in the conterminous United States has a moderate to severe potential for such problems. The western United States, where geologic formations and soils derived from them are salt-affected, has a high potential for problems with salinity. California Table 1-1 shows the extent of salinity and associated problems by land use in California. About 1,720,000 ha (29%) of nonfederal land are saline (EC  4 dS/m) or sodic (exchangeable sodium percentage 15%); 1,100,000 ha (27%) have a water table at a depth of 1.5 m or less; and 1.4 million ha (34%) have a problem with water quality. Table 1-2 indicates that much of the area affected by salinity, sodicity, a high water table, and problems with water quality is in the San Joaquin Valley.

REACTIVITY OF SALTS AND SALT FLOWS World Water Chemistry Figure 1-4 plots TDS of world waters against the cation ratio (Na to Na  Ca) in mg/L. When the scatter of data points is enclosed, a boomerang-shaped figure emerges. In its center are waters with about 50 to 500 mg/L TDS and cation ratios from near zero to about 0.6. Gibbs (1970) indicates that rock dominance or the watershed’s geochemical nature primarily influences these waters. As TDS increases, the cation ratio increases, forming the upper leg of the boomerang with the major oceans at its apex. Gibbs points out that waters become saline as rates of evaporative salinization exceed precipitation (rainfall and snowfall), and the cation ratio increases from the selective precipitation of Ca over Na. As TDS decreases, the cation ratio also increases to form the lower leg. Gibbs attributes this trend to the dilution of the waters by the dominance of rainfall or high runoff over evapoconcentration and the presence of sea salts, mainly NaCl, in high-rainfall coastal areas, which result in high cation ratios. A similar diagram emerges when TDS is plotted against the anion ratio (Cl to Cl  HCO3). It illustrates the three major mechanisms that regulate the chemistry of the world’s water: (1) evapoconcentration; (2) selective mineral precipitation; and (3) rainfall of variable composition.

7

FIGURE 1-3. Soil salinity in the United States. White is 0.0 to 0.4 dSm1 and black is 4.0 dSm1. From USDA (2008).

8

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 1-1. Extent of Salinity and Associated Problems by Land Use in California, in Millions of Hectares

Primary Land Use (1)

Nonfederal Land Area (2)

Salinity/ Sodicity Soils (3)

High Water Table (4)

Water Quality (5)

Irrigated cropland Dry cropland Grazed land Timberland Wildlife land Urban Other

4.1 0.7 7.9 3.6 0.5 2.0 3.6

1.2 0.00 0.32 0.00 0.08 0.04 0.08

1.1 0.04 0.16 0.00 0.04 0.04 0.04

1.4 0.04 0.16 0.04 0.08 0.20 0.12

22.4

1.72

1.42

2.04

Total

From Backlund and Hoppes (1984), © 1984 Regents of the University of California.

Evaporative Salinization Figure 1-5 presents a simplified interpretation of the Hardie-Eugster model for evaporative salinization of waters. The chemical divides lead to several types of hypersaline water. For waters containing the most dominant solute species and subject to evapoconcentration, the first mineral to precipitate in large quantities is calcite (CaCO3). What happens with further evapoconcentration depends on whether the molarity, m, of Ca is greater than or less than carbonate alkalinity (Alk). If 2mCa  Alk, the

TABLE 1-2. Salinity and Drainage Problems by Major Irrigated Areas in California, in Millions of Hectares Location (1)

Irrigated Area (2)

Salinity/ Sodicity (3)

High Water Table (4)

Water Quality (5)

San Joaquin Valley Sacramento Valley Imperial Valley Other areas

2.3 0.85 0.20 0.77

0.89 0.08 0.08 0.12

0.61 0.16 0.20 0.12

0.93 0.12 0.20 0.12

Total

4.12

1.17

1.09

1.37

From Backlund and Hoppes (1984), © 1984 Regents of the University of California.

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

9

FIGURE 1-4. Chemical characteristics of world waters. From Gibbs (1970). Reprinted with permission from AAAS. water would be depleted of HCO3. Further evapoconcentration results in the precipitation of gypsum (CaSO4·2H2O). If mCa  mSO4, the result would be a Cl type of water (e.g., brines in Death Valley, California or the Carson Sink in Nevada). If mCa  mSO4, the brine will also contain appreciable amounts of SO4, such as the world oceans, the Dead Sea, the Salton Sea in California, and the Great Salt Lake in Utah. In contrast, if 2mCa  Alk, the water would be depleted of Ca and the next mineral to precipitate would be sepiolite (MgSi3O6(OH)2). If the water contains 2mMg  Alk, further evapoconcentration results in a brine similar to world oceans. If 2mMg  Alk, the brine would not accumulate Mg, such as in Owens Lake in California and Pyramid Lake in Nevada. There is some question of invoking sepiolite formation as the primary control for Mg. Other possible Mg-controlling precipitates include Mgrich smectite, Mg-rich calcite, or dolomite (CaMg(CO3)2).

10

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 1-5. Hardie-Eugster model for evaporative salinization of waters. From Drever (1988). © 1988, p. 236. Reprinted by permission of Pearson Education, Inc., Upper Saddle River, NJ. Cyclic Salinization and Dilution Natural waters are also subjected to dilution by inputs of fresh water. Figure 1-6 shows classes of precipitates, with evaporites as the most soluble types of minerals, resistates and hydrolyzates as the least soluble. It gives examples of solid phases for each class and the solute species that form or dissolve from these minerals. The precipitation of minerals and the dissolution of minerals are essentially engaged in a tug-of-war that depends on the relative rates of evapoconcentration versus dilution. Hypersaline waters typically would have solute abundances in the order of Na  Mg  Ca and Cl  SO4  HCO 3. Fresh waters would typically have nearly the reverse order of Ca  Mg  Na and HCO3  Cl  SO4. Soil-Water Systems Figures 1-4, 1-5, and 1-6 indicate that mineral solubility principally controls water chemistry and salinity in systems, such as streams and lakes. In a soil-water system, where the ratio of water to soil is considerably smaller than the ratio of water to sediments in streams and lakes, other mecha-

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

11

FIGURE 1-6. Mineral dissolution–mineral precipitation in waters. From Tanji (1972). nisms may also be significant. Figure 1-7 illustrates the complex chemical interactions that take place between the solution, solid, exchanger, and gas phases. A change in soil-water content by irrigation and rainfall or evapotranspiration by plants will cause the equilibrium to shift due to mineral precipitation or dissolution, association or dissociation of ion pairs, adsorption or desorption of cations, and emission or absorption of gases (Paul et al. 1966). Moreover, the free ions and ion pairs are subject to transport by diffusion and convection (e.g., water movement). Salinity and Irrigation Agricultural irrigation changes the soil-water content equilibrium. Figure 1-8 schematically represents water and salt flows in an irrigation project with surface-water supplies. The top left-hand side depicts irrigation water diverted from a river to cropland, the center left depicts irrigated cropland, and the top and bottom left respectively depict the surface and subsurface irrigation return flows. The right-hand side contains water and salt transfers between the interface of irrigation projects and its environs. One way to interpret, synthesize, and simulate water and salt flows in irrigation projects and river basins is hydrosalinity modeling (Tanji 1981), which can help in understanding the dynamic behavior of dissolved mineral salts in waters and soils and the physical, chemical, and biological mechanisms for these changes on micro- and macro-levels.

12

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 1-7. Interactive chemical reactions in soil-water systems. From Tanji (1972).

CONCERNS OVER AGRICULTURAL SALINITY PROBLEMS Historical Perspective The accumulation of salt in soils and the frequently accompanying problem of drainage have plagued irrigated agriculture for centuries. Such accumulation results when plants transpire waters but leave most of the salts in the soil solution. Over time, salts may concentrate to such an extent that they hinder germination, seedling and vegetative growth, and the yield and quality of crops. Historical records for the past 6,000 years reveal that numerous societies based on irrigated agriculture have failed. One of the most highly publicized is ancient Mesopotamia, now Iraq. This once-productive land appears to have suffered salt damage from about 2400 B.C. to 1700 B.C. This problem (Jacobsen and Adams 1958) stemmed from a dispute between two Sumerian cities, Umma and Girsu, over land and water

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

13

FIGURE 1-8. Schematic representation of water and salt flows in an irrigation project with surface water supply. From Hornsby (1973).

rights (a problem still common today). Umma, located upstream from Girsu on the Euphrates River, blocked branch canals that supplied water to Girsu’s agriculture. The latter responded by building other canals off the Euphrates to irrigate a large basin. Flooding, seepage, overirrigation, and

14

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

siltation resulted in a rising water table, which led to excessive soil salinity (Gelburd 1985). The crop records show that production of wheat was phased out over time and replaced by more salt-tolerant barley, but the yields of barley gradually declined to 10 bushels per acre. After 1,000 to 1,500 years of successful irrigated agriculture, the Sumerian civilization declined. Numerous references to canal building in Sumeria from the third millennium BC are available, but no record of drainage canals being built to sustain agriculture (a common present-day occurrence) exists. Similar problems have developed in more recent times in the Indus Plain region, which includes parts of modern-day India and Pakistan, where irrigation began about 2,000 years ago by the Harappa civilization. Serious salinity and drainage problems did not develop until recently, within the last 150 years or so (Taylor 1965). Less well-known are the historical salt and drainage problems in North and South America. The inhabitants of Viru Valley on the coast of Peru developed an irrigation system between 400 B.C. and the time of Jesus Christ (Willey 1953). By 800 A.D., the population of the Viru Valley reached its peak and from 1200 A.D. dramatically decreased. Evidence shows that people relocated from the previously densely settled valley bottoms to the upper narrows of the valley. Historians attribute this relocation to increasing soil salinity and rising water tables from lack of drainage (Armillas 1961). To say that the sole cause of relocation of the people of Viru Valley in the thirteenth century was due to soil salinity may not be entirely accurate, but to ignore this fact is also unacceptable to historians. A second documented case is the Hohokam Indians, who lived in the Salt River region of what is now Arizona. They practiced a form of flood irrigation, similar to that practiced by the farmers in the Viru Valley, beginning about 300 B.C. This civilization flourished through 900 A.D. Historical records are sketchy until about 1275 A.D., when this region, along with much of the southwestern United States, suffered a drought (Willey 1953). After 1450 A.D., no evidence of the Hohokam civilization exists. Records indicate that waterlogging and salt accumulation in the valley floor caused crop failures. Historians surmise that these problems led the Hohokams to either relocate or starve. A Recent California Example The discovery of selenium and other toxic elements in subsurface drainage waters from the west side of California’s San Joaquin Valley, coupled with the age-old threat of salinity and high water tables, has heightened our awareness of the difficulties in sustaining irrigated agriculture (Tanji et al. 1986). Before the 1950s, only salinity and boron were of major concern in agricultural drainage. During the 1960s, more and more attention was paid to nitrates. In the 1970s and the early part of the 1980s,

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

15

pesticides became the major problem. Now, selenium and other toxic elements are of major concern. Eugene Hilgard first identified the growing problem of drainage and alkaline soils in the San Joaquin Valley around 1877. He concluded in 1889 that subsurface drainage was needed for waterlogged, saline, and sodic soils (Kelley 1951). W. W. Mackie in 1905 first investigated land reclamation with an experimental tile-drainage system at Kearney Park, near Fresno, in the San Joaquin Valley (Kelley 1951). Tile-drainage systems were installed beginning in the early 1950s on the west side of the San Joaquin Valley. Since irrigation began in about 1850, surface and subsurface drainage waters have been discharged into the San Joaquin River system. Federal and state agencies envisioned a 450-km master drain to serve the west side of the valley. However, by 1975, the U.S. Bureau of Reclamation had built only the upstream (southern) 137-km reach of the San Luis Drain, terminating at Kesterson Reservoir, a flow-regulating reservoir in the master drain plan. Restrictions in the federal budget due to the Vietnam War delayed completion of the lower reaches of the drain. In 1979, the San Joaquin Valley Drainage Program recommended that the master drain be completed; however, in 1981, uncertainties about the potential environmental and health effects of the discharge of San Luis Drain into the Sacramento-San Joaquin Delta and the San Francisco Bay halted further construction. Meanwhile, the San Luis Drain conveyed saline subsurface drainage waters from a 17,000-ha area affected by drainage to Kesterson Reservoir, which served as a terminal evaporation pond and a wildlife area. From about 1981 to 1986, the average annual flow of tile effluents into the San Luis Drain was about 85,200 ha-cm, collected from about 2,000 ha of tiledrained fields and 1,200 ha drained by the open-joint collector system. In 1982, selenium toxicity of fish and, in 1983, deformed and dead waterfowl were found in the Kesterson Reservoir. The selenium, averaging 300 (g/L (ppb) in the drainwater, came from the Moreno shale, a geologic formation in the Coast Range Mountains that contributed the parent material for some of the soils formed in the west side of the San Joaquin Valley. The discovery of selenium toxicosis of waterfowl at Kesterson Reservoir and the presence of selenium in irrigation return flows in the western United States emphasized the need for agriculture to be concerned about edge-of-field effects on the environment. Discharge of the effluents into the San Luis Drain was halted in June 1986. Massive state and federal investigations have been underway since 1985 on how to contain and dissipate selenium in the Kesterson Reservoir and develop management options to mitigate drainage problems in the San Joaquin Valley (Letey et al. 1987). The San Joaquin Valley Drainage Program and the San Joaquin Valley Drainage Implementation Program

16

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

drafted planning and implementation alternatives to solve agricultural drainage and drainage-related problems in the San Joaquin Valley. Drainage-water disposal through a master drain into the San Francisco Bay Delta system or into the Pacific Ocean is not being considered by the San Joaquin Valley Drainage Program because it is perceived as politically infeasible. This program’s goals are to minimize potential health risks, protect reasonable and beneficial uses of waters, sustain productivity of farmlands, and protect and restore fish and wildlife. After the 6-year, $50 million study, the San Joaquin Valley Drainage Program (SJVDP)—a consortium of five state and federal agencies— produced a final report in 1990 (SJVDP 1990). The SJVDP evaluated drainage problems in the 0.9 million ha of irrigated lands for the entire west side of the San Joaquin Valley and came up with recommended physical and institutional solutions to the drainage problems. The recommended solutions have been monitored by the San Joaquin Valley Drainage Implementation Program (SJVDIP 1998). It became evident that a review of the 1990 recommendations and accomplishments were needed along with consideration of new technologies and developments since the 1990 report. The SJVDIP and the University of California carried out this task, producing eight technical committee reports (Source Reduction, Drainage Reuse, Evaporation Ponds, Drainage Water Treatment, Groundwater Management, Land Retirement, Salt Utilization, and River Discharge; SJVDIP, 1999b–j) and three subarea reports (Grasslands, Westlands, Tulare/Kern; SJVDIP, 1999k–m) in 1999, and a final report in 2000 (SJVDIP 2000). The 2000 final report synthesized and integrated the findings of the twelve 1999 reports and made recommendations. The final report included an illustrative case study with a recommended decision tree created by the authors (Tanji et al. 2002). The case study presents alternative irrigation drainwater management options to meet constraints imposed on the discharge of irrigation return flows into the San Joaquin River in California. The 39,600-ha area of irrigated land is waterlogged and salt-affected but, through subsurface drainage, it is highly productive. The discharge of saline drainage containing naturally occurring boron and selenium is being regulated with waste discharge requirements (WDRs). Irrigation and drainage districts are implementing a combination of drainwater management options but are having difficulties meeting the WDRs for selenium. The case study evaluates alternative management options to meet WDRs, including reduction of deep percolation, drainwater reuse, drainwater treatment, evaporation ponds, groundwater management, limited river discharge, land retirement, and salt utilization. Table 1-3 contains a summary of the eight alternative drainwater management options. The product of this study case is a decision tree recommending drainwater management options sequentially to meet WDRs, which are a moving target.

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

17

TABLE 1-3. Summary of Drainage Management Alternatives for the San Joaquin Valley Westside That Is Constrained by Irrigation Drainage Disposal

Options

Practices

Feasibility/ Adverse Impacts

Estd. Annual Costs, US$ ha1

Source reduction (SR)

Reduce the volume of subsurface drainage by shortening furrow lengths; installing sprinkler, drip, and linear-move systems; modifying irrigation schedules; encouraging use of shallow groundwater to meet crop ET; implementing tiered water pricing, etc.

Some uncertainty exists over farm-level economic benefits of system improvements in contrast to regional benefits. Many of the practices are adopted and in wide use in all subareas.

$60

Drainwater reuse (DR)

Reuse on salt- and borontolerant crops, grasses, trees, and halophytes. Reuse may consist of direct reuse, blending with freshwater, cyclic reuse of fresh and drain waters, or sequential drainwater reuse.

Sustainability of reuse is uncertain for most plants. Potential adverse impacts of sodium, boron, selenium, and molybdenum in soil is not fully ascertained. Residuals of reused waters need to be managed.

$150– $160

Drainwater treatment (DWT)

Remove salts and trace elements by physical, chemical, and biological treatment processes, such as reverse osmosis, and selenium reduction to volatile methylated forms and immobile organic and elemental forms.

Membrane technology is rapidly improving for removal of salts, boron, and selenium. Biological treatment holds the most promise for removal of selenium at economic costs. Operating systems are not in place yet.

$150– $300

Evaporation ponds (EP)

Dispose through evaporation and inadvertent seepage losses in specially designed and managed unlined basins. Impounded waters when desiccated, deposit salts (thenardite, halite, gypsum, calcite, etc) and trace elements.

Evaporation ponds play a major role in sustaining agriculture in the Tulare subarea. Adverse toxic impacts of selenium on birds are being mitigated by compensation and alternative wetland habitats.

$180– $300

18

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 1-3. Summary of Drainage Management Alternatives for the San Joaquin Valley Westside That Is Constrained by Irrigation Drainage Disposal (Continued)

Options

Practices

Feasibility/ Adverse Impacts

Estd. Annual Costs, US$ ha1

Land retirement (LR)

Retire or fallow irrigated lands heavily impacted by selenium and waterlogging, and restore native plant communities.

USBR has implemented a buyout program for lands retired. Large contiguous tracts of retired lands are desired for wildlife. Retired lands could become excessively salinized and seleniferous in waterlogged sites.

Groundwater management (GM)

Control water table by subsurface drains and by pumping regional wellfields. Pumped water of adequate water quality could serve as a substitute for surface water.

Subsurface drains are in $160– place in all subareas. $185 Existing wells are used during drought. Plans for regional wellfields are on hold.

River discharge (RD)

Grassland subarea is the only subarea able to discharge into the river, but is constrained by waste discharge requirements.

This subarea is having some difficulties meeting target concentration and load limits.

Salt utilization (SU)

Harvest salts by desiccatThus far, a market for ing unusable saline drainsalts (mainly waters in solar evaporators. thenardite) produced is not available.

$170

$120

Unknown

ET, evapotranspiration; USBR, U.S. Bureau of Reclamation From SJVDP (1990; 2000).

The following criteria were used when evaluating each management option (SJVDIP 2000): 1. Meet the water quality objectives in the San Joaquin River and comply with WDRs for drain outfalls into the river. 2. Promote efficient use of water in the upslope portion of the area in order to minimize drainage problem downslope.

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

19

3. Utilize water until it is no longer usable to reduce the volume of drainage that must be managed or discharged. 4. In view of the site’s close proximity to state and federal wildlife areas, as well as private duck hunting ponds, protect the environment for fish and waterbirds. 5. Avoid further contamination of groundwater so that it remains available in the future for potential agricultural, industrial, and municipal uses. The primary “driving force” behind the evaluation of management options was the need to meet WDRs. Source reduction (SR) is considered the preferred drainwater management option because it is comparatively easy for many growers to implement, it contributes to managing water more efficiently, and it reduces drainwater that needs to be disposed. The combination of SR and drainwater reuse (DR) is a natural follow-up to reduce irrigation return flows and to meet WDRs for river discharge (RD). Moreover, SR and DR both reduce the volume of drainage water and thus help reduce the need for evaporation ponds (EPs), drainwater treatment (DWT), groundwater management (GM), and land retirement (LR). However, DR could degrade the soil physically and chemically if the loading rate is too large, and deep percolation may eventually further degrade groundwater if the concentrated drainage water is not intercepted and removed. Land retirement will reduce the volume of drainwater that must be managed provided the water used to irrigate the retired land is not applied on other irrigated lands. If some irrigated lands were retired, there would be less need for other management measures. However, there are some costs associated in managing retired lands. Evaporation ponds are an alternative to RD, but high levels of selenium in the collected subsurface waters may pose a hazard to waterbirds and require expensive mitigation measures. If selenium could be removed by DWT to low levels prior to disposal in EPs, mitigation measures will be reduced. EPs and salt utilization (SU) are a natural follow-up to sequential DR. If EPs are feasible, they will reduce the need for not only RD but also LR, DWT and GM. Groundwater management may alleviate waterlogging provided the pumped groundwater is usable. If the groundwater is of marginal quality, GM in combination with DWT, EPs, or SU may be an option. As stated previously, the WDRs for river discharge are the principal driving force for this case study; thus, all other drainwater management options or combinations thereof must be evaluated in terms of meeting WDRs. Table 1-4 ranks the practicality of each management option (SJVDIP 2000). Source reduction is in the top spot because it is highly feasible, both technologically and economically, and has been applied with success at this site. Another, simpler alternative for expanding the use of RD is to

20

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 1-4. Ranking of Drainage Management Options Drainage Option

Rank

SR (source reduction) RD (river discharge) DR (drainwater reuse) DWT (drainwater treatment) LR (land retirement) EP (evaporation pond) GM (ground water management)

Most effective Very effective Effective Somewhat effective Somewhat effective Somewhat effective Somewhat effective

SU (salt utilization)

Somewhat effective

From SJVDIP (2000).

implement a real-time drainage water disposal program. Assuming that some such program can and will be implemented by the Central Valley Regional Water Quality Control Board (CVRWQCB), RD has been assigned a rank of “very effective.” Drainwater reuse has been used in the study area and is viewed by many to be a viable option, although questionable in terms of its long-term effects on soil quality. Drainwater reuse does reduce the volume of drainwater that must be managed and is ranked “effective.” Drainwater treatment is ranked “somewhat effective” because desalting and biological treatment technologies for removal of selenium in drainwaters are still in the pilot scale status. Retiring land will set aside land for wildlife habitat and as resting places for migratory birds. When retiring land, if water formerly used for irrigation of those lands is merely conveyed to nearby lands for irrigation, then little improvement (in terms of reducing drainwater quantity) will be achieved. Land retirement or intermittent fallow is practiced today in the region surrounding the study site, although not yet on the study site itself. The U.S. Bureau of Reclamation now has funds to purchase farmlands for LR, and this option is assigned a rank of “somewhat effective.” Evaporation ponds would certainly decrease the volume of drainwater disposed off-site; however, the mitigation measures and other measures necessary to meet the WDRs will be costly. The case study area has comparatively high selenium in the shallow groundwater, which probably will result in higher-than-normal mitigation costs when subsurface drainwaters are impounded into EPs. For all these reasons, taken together, the EP option has been assigned a rank of “somewhat effective.” Although groundwater management can play a major role in the study site’s (and the Valley’s) drainage problems, the management options available for

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

21

groundwater are all of a long-term nature and may be costly. To ensure that a groundwater program will be effective, modeling will be necessary with associated consulting and field monitoring activities. These will add to the cost. Additionally, state water quality laws forbid the intentional degradation of aquifers, which is a likely side-effect of implementing this option. Thus, GM has been assigned a rank of “somewhat effective.” Salt utilization has good long-term potential for helping to meet the salt balance in the Valley. However, when applying the criteria that say a technology must be doable “today” (i.e., within a time horizon of 2 to 3 years), that a market exists, and that the economics are favorable, SU falls down in the list. The realities of finding a profitable market and arranging economic harvesting and transport (e.g., by rail) are just not met at present. Salt utilization is therefore ranked “somewhat effective.” Table 1-4 identifies the top three management options: source reduction (“most effective”), river discharge (“very effective”), and drainwater reuse (“effective”). The remaining five options are ranked equally in the moderate range (“somewhat effective”), because there remains some question about the feasibility and cost of each. Figure 1-9 presents a decision tree that may be helpful in solving the drainage problem in the study area. To use the decision tree, one begins at “Start” and proceeds through the successive steps, as with a flowchart. As options fail to meet the WDRs, the succeeding options implemented become more drastic, or more costly, or both. In essence, looking at the chart differently, it displays combination of options as the WDRs become more stringent. A combination of options should be selected based on the technical, economic, and institutional issues discussed thus far. Figure 1-9 is an illustrative product for an analytical exercise for assessing potential drainwater management options. Most recently, the Westside Regional Drainage Plan Proposal is a series of six integrated projects designed to eliminate subsurface agricultural production drainage from about 100,000 acres of drainage-affected lands on the west side of the San Joaquin Valley (SJVDIP 2000). The plan also includes water demand reduction, groundwater pumping and management, water transfer elements, and drainage treatment to provide for drainage control and improve water supply reliability for the partners executing the plan. Statewide benefits of the plan include allowing for compliance with state water-quality objectives for salinity, boron, and selenium, as well as increases of available water supply of up to 23,000 acre-feet in dry and critically dry years due to a lessening of need for storage releases for water quality on the San Joaquin River. The plan was developed by the stakeholders and is designed primarily to quick-start identified drainage elements in time to meet water-quality standards. A key element is adaptive management combining investigation, construction of proven drainage components, and operational experience

22

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 1-9. The decision tree, a flowchart outlining a stepwise plan for implementing drainwater management options at the case study site. From SJVDIP (2000). to perfect the final drainage strategy. The chief components include land retirement, groundwater management, source control, regional reuse, treatment, and salt disposal. Contemporary salt problems exist elsewhere in the United States, such as the upper Colorado River basin and the northern Great Plains.

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

23

THE AGRICULTURAL SALINITY AND DRAINAGE DILEMMA As mentioned earlier, the history of agriculture has shown that irrigated agriculture cannot survive in perpetuity without adequate salt balance and drainage. The length of time that irrigated agriculture can survive without adequate drainage depends on hydrogeology and water management. The accumulation of salt in soils depends on the salinity of the applied waters, the salinity of the native soil, and the rate at which salts are leached out of the rootzone. Waterlogging in the rootzone depends on the presence of restricting soil and substrata layers and the vadose region’s capacity for deep percolation. If restricting layers exist close to the surface, waterlogging occurs and the accumulation of salt associated with waterlogged soils develops within a comparatively short time, such as decades. If no restricting layers exist and the vadose region has a large capacity, irrigation may be practiced for a long time (possibly centuries) before problems with surface drainage and groundwater quality arise. The technology to sustain irrigated agriculture exists, but its use is limited by a number of technical and socioeconomic factors, such as lack of economic incentives for irrigators; lack of education on best management practices; the high cost of improving structures, pressurized irrigation systems, and drainage systems; institutional constraints, such as water rights and water transfers; and the effects of irrigation return flows on the environment. Because irrigated agriculture operates in an open system and even the best source-control measures still result in drainage water containing salt, fertilizer, and pesticide residues, measures to dispose of salts in ultimate salt “sinks,” such as the oceans or inland closed basins, are needed.

SUMMARY Changing soil-water conditions can result in substantial changes in the composition of salt in the soil and water, and in total concentrations of salt. These changes can have profound impacts on the long-term suitability of land for agriculture and associated socioeconomic problems. In semiarid to arid climates, irrigated agriculture requires leaching and drainage. In the past, it has often been assumed that residuals from agriculture, especially saline drainage waters, would be disposed of somewhere, somehow. Now other water users have demanded that agriculture minimize the degradation of water quality. Federal, state, and local regulations on nonpoint sources of pollutants are being promulgated. There are a number of management strategies available to address these regulations, although over the long term some means of disposing of accumulated salts in an ultimate sink will need to be found.

24

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Irrigated agriculture needs to strive for a balance between maintaining agricultural productivity and protecting our natural resources. To this end, this manual serves as a reference to help sustain irrigated agriculture.

REFERENCES Armillas, P. (1961). “Land use in pre-Columbian America,” in A history of land use in arid regions, L. D. Stamp, ed., UNESCO Arid Zone Research, 17, Paris, 255–276. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Backlund, V. L., and Hoppes, R. R. (1984). “Status of soil salinity in California,” Calif. Agric., 38(10), 8–9. Drever, J. I. (1988). The geochemistry of natural waters, 2nd ed., Prentice Hall, Upper Saddle River, N.J. Gelburd, D. E. (1985). “Managing salinity lessons from the past,” J. Soil and Water Conserv., 40(4), 329–331. Gibbs, R. J. (1970). “Mechanisms controlling world water chemistry,” Science, 170, 1088–1090. Hornsby, A. G. (1973). Prediction modeling for salinity control in irrigation return flows, EPA-R2-73-168, U.S. Environmental Protection Agency, Washington, D.C. Jacobsen, T., and Adams, R. A. (1958). “Salt and silt in ancient Mesopotamian agriculture.” Science, 128, 1251–1258. Kelley, W. P. (1951). Alkali soils: Their formation, properties and reclamation, Reinhold Publishers, New York. Letey, J., Roberts, C., Penberth, M., and Vasek, C. (1987). An agricultural dilemma: Drainage water and toxic disposal in the San Joaquin Valley, Special Publication 3319, Div. of Agriculture and Natural Resources, University of CaliforniaBerkeley. Massoud, F. I. (1981). “Salt affected soils at a global scale and concepts for control,” Technical Paper, FAO Land and Water Development Div., Food and Agriculture Organization of the United Nations, Rome. Paul, J. L., Tanji, K. K., and Anderson, W. D. (1966). “Estimating soil and saturation extract composition by a computer method,” Proc., Soil Sci. Society of Am., 30, 15–17. Richards, L. A., ed. (1954). Diagnosis and improvement of saline and alkali soils, USDA Handbook No. 60, USDA, Washington, D.C. San Joaquin Valley Drainage Program (SJVDP). (1990). “A management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley,” U.S. Department of the Interior and California Resources Agency, Sacramento, Calif. San Joaquin Valley Drainage Implementation Program (SJVDIP). (2000). Evaluation of the 1990 Drainage Management Plan for the Westside San Joaquin Valley, Cal-

NATURE AND EXTENT OF AGRICULTURAL SALINITY AND SODICITY

25

ifornia, Ad Hoc Coordination Committee (AHCC), Sacramento, Calif, University of California-Berkeley and San Joaquin Valley Drainage Implementation Program. SJVDIP. (1999a). “Drainage technical committee and subarea reports,” Department of Water Resources, Sacramento, Calif., www.dpla.water.ca.gov/agriculture/ drainage/implementation/hq/sjvlib.htm, accessed January 15, 2011. SJVDIP. (1999b). “Executive summary,” Technical Committee Report. SJVDIP. (1999c). “Drainage reuse,” Technical Committee Report. SJVDIP. (1999d). “Drainage water treatment,” Technical Committee Report. SJVDIP. (1999e). “Land retirement,” Technical Committee Report. SJVDIP. (1999f). “Evaporation ponds,” Technical Committee Report. SJVDIP. (1999g). “Source reduction,” Technical Committee Report. SJVDIP. (1999h). “Groundwater management,” Technical Committee Report. SJVDIP. (1999i). “River discharge,” Technical Committee Report. SJVDIP. (1999j), “Salt utilization,” Technical Committee Report. SJVDIP. (1999k). “Grasslands subarea report,” Technical Committee Report. SJVDIP. (1999l). “Westlands subarea report,” Technical Committee Report. SJVDIP. (1999m). “Tulare/Kern subarea report,” Technical Committee Report. SJVDIP. (1998). Drainage management in the San Joaquin Valley, A status report, Department of Water Resources, Sacramento, Calif. Szabolcs, I. (1989). Salt-affected soils, CRC Press, Inc., Boca Raton, Fla. Tanji, K. K. (1981). “River basin hydrosalinity modeling,” Agric. Water Mgmt., 4, 207–225. Tanji, K. K. (1972). Lecture notes from the course, “Chemistry of the Hydrosphere,” University of California, Davis, Calif. Tanji, K. K., Dutt, G. R., Paul, J. L., and Doneen, L. D. (1967). “Quality of percolating waters. II. A computer method for predicting salt concentrations at variable moisture contents,” Hilgardia, 38(9), 307–318. Tanji, K. K., Läuchli, A., and Meyer, J. (1986). “Selenium in the San Joaquin Valley,” Environment, 28(6), 6–11, 34–39. Tanji, K. K., Wallender, W. W., and Rollins, L. T. (2002). “Irrigation drainage water management options: San Joaquin valley case study,” in Scientific Committee (eds.), Proc., Symposium 33, 17th World Congress of Soil Science, International Union of Soil Science, 1986, 1–10 [CD-ROM]. Taylor, A. C. (1965). “Water, history and the Indus plain.” Nat. Hist., 24(5), 40–49. Todd, D. E. (1970). The water encyclopedia, Water Information Center, Port Washington, N.Y. U.S. Department of Agriculture (USDA). (1988). Water quality education and technical assistance plan, USDA-Soil Conservation Service and USDA-Extension Service report. U.S. Department of Agriculture (USDA). (2008). “Digital General Soil Map of U.S.,” http://soildatamart.nrcs.usda.gov, accessed March 18, 2011. Willey, G. R. (1953). Prehistoric settlement patterns in the Viru Valley, Peru, Bulletin 155, Smithsonian Institute, Bureau of American Ethnology, Washington, D.C.

This page intentionally left blank

CHAPTER 2 DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION OF CONTROL PRACTICES: AN OVERVIEW James D. Rhoades

INTRODUCTION This chapter summarizes the principal adverse effects of salts on soils and plants, and describes a methodology for diagnosing the nature and cause of salt problems, assessing reclaimability, and selecting appropriate reclamation and management practices. As an overview, this chapter summarizes and makes reference to the other chapters in this manual that provide greater detail in discussion of salt effects, salinity measurement techniques, diagnostic criteria, and salinity control practices in more detail.

POTENTIAL ADVERSE EFFECTS OF SALTS ON SOILS AND PLANTS: BRIEF SUMMARY The development of effective salinity control practices requires an understanding of cause and effect. Toward that end, this section briefly describes the principal effects of salts on soils and plants (Rhoades et al. 1992). Effects of Salts on Soils The suitability of a soil for cropping depends greatly on the degree to which it conducts water and air (permeability) and on physical properties that control the friability of the seedbed (tilth). Poor permeability and tilth often are major problems in irrigated lands. Saline soils generally have 27

28

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

normal physical properties; however, in sodic soils, physicochemical reactions cause the slaking of aggregates and the swelling and dispersion of clay minerals, leading to reduced permeability and poor structure and quality (tilth). To understand how the poor physical properties of sodic soils are developed, the interaction between the negatively charged colloidal clay particles in the soil and the exchangeable cation envelopes that are associated with those solid clay particles must be understood. The interactions between these exchangeable cations and their counterparts in the external bulk solution are also important. The envelope of exchangeable cations is subject to two opposing processes: (1) cations are strongly attracted to the negatively charged clay surface in proportion to their charge; and (2) they tend to diffuse from the surface of the clay, where their concentration is higher, and into the bulk of the solution, where their concentration is lower. This attraction––diffusion process results in an approximately exponential decrease in cation concentration with distance from the clay surface out to the bulk solution. Divalent cations, such as calcium (Ca) and magnesium (Mg), are attracted to the surface with a force twice as great as sodium (Na) and other monovalent cations. In soils containing high proportions of Ca and Mg relative to Na or high concentrations of total dissolved electrolytes (i.e., saline soils), the exchangeable cation envelope is therefore compressed toward the clay surface. Consequently, the repulsion forces between the like-charged envelopes decrease, and the particles can approach closely enough to permit their cohesion into aggregates. These aggregates have a matrix of greater pore size compared to that of individual clay particles. Hence, the permeability and tilth of aggregated soils are better than those of nonaggregated soils. The repulsion between nonaggregated clay particles also allows more solution to be imbibed between them. This phenomenon is referred to as swelling. Because of the plate-like shape and parallel orientation of clays, such swelling reduces the size of interaggregate pore spaces in the soil, reducing permeability. Swelling is particularly important in soils that contain expandable phyllosilicate minerals, such as smectites, and have high Na levels [exchangeable sodium percentage (ESP) values of greater than about 15]. Dispersion, the release of individual clay platelets from aggregates, and slaking, the breakdown of aggregates into subaggregate assemblages, can occur at ESP values of lower than 15 if the electrolyte concentration is low. Dispersed platelets or slaked subaggregate assemblages can lodge in pore interstices, reducing permeability. Slaking also decreases porosity, which leads to crusting, reduced permeability, and poor tilth. Water entering the soil must pass through the soil surface, so the combined properties of water and soil affect the water-entry rate. Thus, soil permeability and tilth problems must be evaluated in terms of both the salinity of the infiltrating water and the ESP [or its equivalent sodium

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

29

adsorption ratio (SAR) value] of the topsoil. Representative threshold values of SAR (⬃ESP) and the electrical conductivity of infiltrating water for maintenance of soil permeability and tilth are given in Chapter 11. In summary, soil solutions with high salt concentrations and low SAR values are conducive to good soil physical properties. Conversely, low electrolyte concentrations and relatively high proportions of Na salts, that is, high SAR values, adversely affect permeability and tilth. These variables may be manipulated by the grower by, for example, the addition of gypsum to the soil surface or to irrigation water can help to avoid or possibly even alleviate problems with reduced infiltration rate and seedling emergence. Gypsum increases the electrolyte concentration of the infiltrating water and supplies divalent Ca to replace Na. For more specific information on the effects of exchangeable Na, electrolyte concentration, pH, and exchangeable Mg and K on the permeability, infiltration, and crusting of soils, see Chapters 3, 5, and 11. Effects of Salts on Plants Excess salinity within the rootzone reduces plant growth rate. The hypothesis that seems to best fit observations asserts that excess salt reduces plant growth, primarily because it increases the energy that the plant must expend to acquire water from the soil and make the biochemical adjustments necessary to survive. This energy is diverted from the processes that lead to growth and yield, including cell enlargement and the synthesis of metabolites and structural compounds. Typically, growth is suppressed when a threshold value of salinity is exceeded. This threshold value depends on the crop, external environmental factors, such as temperature, relative humidity, or wind speed, and the water-supplying potential of the rootzone. Suppression of growth increases as salinity increases until the plant dies. The salt tolerance of many crops can be expressed as follows (Maas and Hoffman 1977): Yr  100  b (ECe  a)

(2-1)

where Yr  the percentage of the yield of the crop grown under saline conditions relative to that obtained under nonsaline but otherwise comparable conditions; a  the threshold level of soil salinity at which yield decreases begin; and b  the percentage yield loss per increase of salinity in excess of a, and the unit of soil salinity (ECe) is dS/m. Equation 2-1 assumes that crops respond primarily to the osmotic potential of the soil solution. Effects of specific ions or elements must be considered separately. They are generally secondary in importance. Available data on relative crop tolerances to salinity and specific ions and elements are compiled in Chapter 13 (Tables 13-1 through 13-5). Such

30

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salt-tolerance data cannot provide accurate, quantitative values of crop yield losses from salinity for every situation. Actual response to salinity varies with many factors, including climate, soil conditions, agronomic practices, irrigation management, crop variety, stage of growth, and salt composition. For example, most plants are relatively salt-tolerant during germination and are more sensitive during seedling emergence and early stages of seedling growth. Hence, it is imperative to keep salinity in the seedbed low after germination. In addition, differences in salt tolerance often occur between different varieties of a given species. Rootstocks affect the salt tolerances of tree and vine crops because they regulate the uptake and translocation of potentially toxic salts, such as Na and chlorine (Cl), to the shoots. Salt tolerance also depends on the method of irrigation and its frequency. As water becomes limiting, plants experience stresses from low matric potential, as well as low osmotic potential. The available salt tolerance data apply most directly to crops irrigated by furrow and flood irrigation with conventional irrigation management. Salt concentrations within irrigated soils profiles change constantly. They may differ severalfold over the depth of the profile. The plant is most responsive to salinity in that part of the rootzone where maximum water uptake occurs. Sprinkler-irrigated crops are potentially subject to additional damage by foliar salt uptake and burn from spray contact of the foliage. Susceptibility to foliar salt injury depends on leaf characteristics and rate of absorption. It does not correlate with general salt tolerance. The degree of injury caused by spray effects depends on the weather and water stress. Relatively little information is available for predicting yield losses caused by foliar salt uptake and sprinkler spray effects (see Chapter 13). Climate is a major factor affecting salt tolerance. Most crops can tolerate greater salt stress if the weather is cool and humid than if it is hot and dry. Yield is reduced more by salinity when humidity is low. While the primary effect of soil salinity on herbaceous crops is to retard growth, as discussed, certain constituents often present in saline waters are toxic to particular crops (Chapter 13). For example, boron (B), when present in the soil solution at concentrations of only a few parts per million, is highly toxic to many crops. Boron toxicities can be described in terms of a threshold value and yield-decrement slope parameters, as for salinity. In addition, Na and Cl may accumulate in the tissues of woody crops over time to toxic levels that produce foliar burn. In sodic soils, Ca deficiencies may occur, as well as deficiencies of other nutrients such as phosphorus (P), zinc (Zn), and iron (Fe). These conditions can be moderated though the use of certain amendments, such as gypsum and sulfuric acid. For more information on the amelioration of such soils, see Chapter 21. Crops grown on infertile soil may seem more salt-tolerant than those grown with adequate fertility, because fertility is the primary factor limit-

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

31

ing growth. The addition of extra fertilizer will not alleviate growth inhibition by salinity. For more information on how salinity affects the physiology and biochemistry of plants, see Chapters 6, 7, 8, and 13. DIAGNOSING SALT PROBLEMS This section provides a methodology for diagnosing salt-related problems. Some guidelines are given for diagnosing problems caused by salinity, permeability and crusting, toxicity, nutritional imbalances, and drainage. Because salinity is usually the dominant problem in field soils, it should be considered first. If it is found to be dominant, there is no need to proceed with diagnoses of permeability/crusting or toxicity/nutrition problems, especially since the latter two conditions may be altered by the leaching or drainage undertaken to alleviate the excess salinity. If salinity is not a problem, permeability/crusting should be considered next, followed by toxicity/nutrition. Following leaching, toxicity or permeability problems that may have been overridden in the presence of salinity can appear. Subsequent evaluations of these problems may be needed. Visual observations of soils, crops, topography, elevation of water in surface drainages, and plants are rarely sufficient to diagnose a salinity problem conclusively. For example, salinity may reduce the yields of crops by as much as 25% without visible symptoms. Moreover, visual observations may lead to a false diagnosis. For example, the minerals calcite and gypsum are essentially harmless to plants and soil properties, but they can form a white crust on the surface of soils that may be confused with potentially harmful soluble salts. Still, visual information can be useful. It provides clues to the problem and suggests what measurements should be made and where to make them. In situ measurements of soil electrical conductivity and on-the-spot measurements of the salinities of irrigation water and drainage water will help to narrow the possibilities and pinpoint the areas to sample. Final diagnosis must rely on appropriate chemical analyses of plant, soil, and water samples. Before actively correcting problems caused by salt, the cause of a salt problem should be determined and steps taken to eliminate it. Soil Salinity Problems Visual indications Saline fields often can be identified by the presence of spotty white patches of precipitated salt. Such precipitates usually occur on slightly

32

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

elevated positions or in unvegetated areas, where water evaporates and leaves the salts behind. Visible salt crusts obviously indicate surface salt accumulation, but they do not provide reliable evidence of high salinity in the rootzone. Recent excavations and open ditch banks more reliably indicate highly saline soil in the rootzone. If salinity is high, white salt deposits will appear on the newly exposed surfaces of excavations within a few hours, as the water evaporates. During germination and establishment of seedlings, only the salinity of the topsoil affects the plant. Bare spots in a cropped field may indicate only that excessive salt occurs in the shallow topsoil or seedbed rather than in the deeper rootzone. The vigor of the plants adjacent to these bare spots is usually a good indicator of the depth of excessive soil salinity. Crops growing on saline soils typically show uneven growth and may exhibit other symptoms induced by salinity, such as stunted growth or unusually small leaves of a deep bluegreen color. When different crops are grown in a single salt-affected area, a knowledge of which crops show symptoms induced by salinity and which do not can be combined with the crop tolerance values (see Chapter 13) to estimate the magnitude of soil salinity within the rootzone of the area. Some caution should be exercised in using crop distribution and growth to diagnose salinity, because water deficiencies, disease, soil infertility, or misapplication of herbicides all may produce similar visual symptoms. Measurements Soil salinity can be determined in the field by a number of techniques. Soil samples can be extracted and the EC of their paste extracts, ECe measured in the field using field-kit techniques. More conveniently, soil salinity may be determined from measurements of the electrical conductivity of the soil paste itself, ECp, or measurements of the electrical conductivity of the bulk field soil, ECa. The latter measurement is particularly suited for reconnaissance because soil samples do not have to be collected, and information about soil salinity can be quickly derived on the spot. To establish the horizontal and vertical distribution of salinity over a field, ECa, data can be collected from a spatially distributed set of field sampling sites. This information is used to select the areas to be separately evaluated and to determine the number of soil samples necessary to determine the chemical composition of each area. The resulting ECe data from each area of concern are used with appropriate salt-tolerance criteria to determine whether soil salinity is a problem. The principles and practices of measuring ECp and ECa, and determining ECe and chemical composition are described in Chapter 10. Sampling guidelines and statistical procedures are described in Chapters 9 and 14.

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

33

Diagnostic criteria To determine whether salinity is excessive for establishment of seedlings, the ECe of the topsoil is compared with the tolerance of the crop to be grown (see Chapter 13). To determine whether salinity is excessive for the vegetative and reproductive growth of conventionally irrigated, established crops, the mean salinity of the rootzone is compared with the corresponding crop salt tolerances. When evaluating salinity of fields that use high-frequency irrigation or drip irrigation, the water-uptakeweighted salinity, which is the salinity within each soil depth weighted in proportion to the estimated fraction of rootzone water extracted from it, is used instead of the depth-averaged mean. The variation of salinity over the area of concern should be determined and included in the diagnosis. To do this, a salt index is computed and compared with the salt tolerance of the crop to be grown. If a salt problem in any part of the cropped area is unacceptable, the salt index for the entire area is equivalent to the maximum salinity observed in the area. The standard deviation may be computed if at least 10 data values are available. If salt problems are allowable in about 15% of the area, the sum of the mean and standard deviation may be used as the salt index. If the standard deviation of salinity values is small, (that is, less than 15% in coefficient of variability), the sample mean should be used as the salt index. Salinity is diagnosed as a growth-limiting factor (i.e., as the first-order salt-related problem) if the tolerances are exceeded for seedling establishment or vegetative and reproductive growth. The source of this salinity should then be ascertained. If salinity is not limiting, the possibility that either permeability/crusting or toxicity/nutrition problems exist should be evaluated next. Potential causes of soil salinity problems The potential causes or sources of excess salts include saline irrigation water, inadequate leaching, inadequate drainage, indigenous soil salts, and inundation by and evaporation of salt-laden waters. If none of these factors is the cause of excessive salinity, other possibilities should be considered, such as excessively applied fertilizer or manure, or wind-blown saline water spray. Specific methods are not given here for diagnosing such situations, since they are unusual and case-specific. Is irrigation water too saline? To evaluate the possibility that saline irrigation water is the cause of soil salinity problems, compare the upper rootzone (0–30-cm depth) salinity and average rootzone salinity with the salinity levels that would likely result from the use of the irrigation water. The latter values are predicted by multiplying the electrical conductivity

34

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of the irrigation water, ECi, by appropriate concentration factors (Fc) for these depths (see Chapter 11). If rainfall is significant, a rainfall-weighted averaged value of ECi is used. The level of salinity (ECe basis) that should result in the topsoil from irrigation with a water of ECi is estimated as 0.6 ECi. The water is too saline and judged to be the cause of seedling emergence problems if the resulting value of 0.6 ECi (and the observed topsoil ECe) exceed the value that is tolerable for germination and seedling establishment. A spotty stand is typically observed under such circumstances. The level of average rootzone salinity that should result from irrigating with a water of ECi, when leaching is at its upper practical limit, is estimated as 0.8 ECi. The water is judged to be too saline if 0.8 ECi and average observed rootzone ECe exceed the average salinity level tolerable by the crop being grown and the salinity distribution indicates a net downward water flux. In this case, either a crop of greater tolerance must be substituted, or a water of lower salinity should be used for irrigation. This evaluation scheme applies to situations where the soil has been under irrigation for at least 3 to 5 years and salts have not accumulated in the rootzone due to upward flow of groundwater or inundation by saltwater. Upward flow of groundwater from the water table can be ascertained from the distribution of salinity within the rootzone. Is inadequate leaching the problem? To evaluate the possibility that inadequate leaching is the cause of soil salinity problems, the average rootzone salinity or the concentration of a specific solute, such as Cl, and its distribution throughout the rootzone should be determined. These may be excessive for crop growth, even when low-salinity irrigation water is used. The salinity distribution should be used to determine the direction of net flow of water. That, in turn, can be used to diagnose the adequacy of leaching or the presence of shallow saline water tables. Leaching is judged insufficient if the ECe values increase with depth and the ratio of average rootzone ECe to ECi exceeds (ECe⬘/ECi), where ECe⬘ is the maximum average rootzone salinity without yield reduction for the crop in question (see Chapter 13). The adequacy of leaching can also be evaluated by comparing an estimate of the leaching fraction, L, or the theoretically achievable leaching fraction under present management, LA, with the leaching requirement, Lr. The Lr for a given crop can be obtained from Chapter 12. LA is estimated from the following equation: LA  1 

ETc 1 ⋅ Ti

(2-2)

where ETc is the estimated crop evapotranspiration (ET) requirement; and I·Ti is the amount of water infiltrated over the irrigation period. If LA

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

35

(or L) is less than Lr, leaching is judged inadequate. This problem can be alleviated by increasing L, either by increasing the infiltration rate (I) or the irrigation set period (Ti), or by changing to a crop of lower ETe. If excessive surface ponding of water is observed following irrigation, the cause of inadequate leaching is an insufficient infiltration rate. Is inadequate drainage the problem? Inadequate drainage due to a high water table can cause an upward net annual flux of water and salt, which may cause excessive salt accumulation in the soil. To determine whether poor drainage is the cause of a soil salinity problem, the field site is visually inspected for abnormally wet soil or weedy areas dominated by water-loving plants. Plants growing in excessively wet areas often have pale, yellowish leaves, which indicate N or Fe deficiencies. In exposed soil profiles, Mn concretions and red or yellow mottled spots associated with Fe compounds also indicate poor soil aeration. Also, water that fills an excavated hole without appreciably draining for several days indicates a high water table. The adequacy of drainage should be diagnosed from measurements of the magnitude and distribution of ECe above the water table during the cropping season and observations of the water table depth over the same period of time. If these measurements can only be made at one time during the year, they should be made at the end of or late in the irrigation season, when the net effect of irrigation practices and the water table on the rootzone salinity is generally most evident. Problems caused by inadequate drainage often develop progressively through the season, so the late-season condition is reasonable for diagnosis. Drainage is judged inadequate if (1) the ECe pattern with depth indicates that the net annual flux of water is upward into the rootzone and the magnitude of the average rootzone ECe value exceeds that tolerable by the crop in question, or (2) the rootzone is inundated for more than a few days or the soil is waterlogged within a distance of approximately one rootzone depth from the bottom of the crop’s roots. An alternative method for evaluating the adequacy of drainage in the presence of a shallow water table is as follows. The leaching fraction being attained is estimated by using Eq. 2-1, the corresponding value of Fc is obtained from the appropriate table in Chapter 11, and the maximum expected average rootzone salinity, ECi Fc is computed. It may be assumed that drainage is inadequate if the measured value of average rootzone ECe exceeds the value of the quantity ECi · Fc for the crop being grown. For more discussion of drainage, see Chapter 16. Are indigenous salts in the soil the problem? Indigenous soil salts are more common in dryland soils than in irrigated soils. The management and reclamation of soils containing indigenous salts differ from the

36

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

management and reclamation of soils in which salts are derived from external sources. To discern whether indigenous soil salts are the cause of soil salinity problems, the methods described in Chapter 10 can be used to determine the gypsum content, ECe, and the composition of the soluble salts in a soil. Indigenous salts are diagnosed as the likely source of excessive salinity in an irrigated soil when the determined levels of ECe through the rootzone exceed those predicted from the quantity ECi Fc and there is no evidence of a shallow water table. Highly soluble indigenous salts, such as Na and Mg sulfates, require different abatement strategies from the relatively insoluble salts (gypsum, for example). If a soil contains enough gypsum to increase the average rootzone salinity to a level that is greater than the level that could result only from irrigating with a water in the absence of gypsum (i.e., ECi Fc), and if this level exceeds the crop’s tolerance, then a crop of higher salt tolerance should be selected, since it is generally impractical to leach out the gypsum. However, if the salts present in an excessively saline soil are of the highly soluble type, they should be removed by leaching. The total volume of water required to dissolve these soluble salts (the reclamation requirement) can be determined as described in Chapter 21. Is a historic playa the problem? Salinity problems in soils can result from the periodic inundation and subsequent evaporation of salt-laden waters. To determine whether this could be the cause of a soil salinity problem, the local topography and topsoil properties of the site are examined. Evidence of a playa-type topographic feature indicates that this process of salinization has occurred. To the extent that this process is still operative, it should be countered by an appropriate diversion measure (see Chapter 16). Soil permeability and crusting problems A high ratio of exchangeable Na in relation to other exchangeable cations (sodicity) enhances the degradation of the tilth of a surface soil. This is usually manifested in decreased permeability and increased crusting. The loss of permeability may so restrict water infiltration into the rootzone that plants become stressed from lack of water. Crusting may so impede or prevent seedling emergence as to reduce crop stand. Visual indications. The structural condition and color of the surface soil may be used to judge whether permeability or crusting problems caused by sodicity are likely to occur. Sodic soils are poorly aggregated, crusted, and hard when dry, and sticky and plastic when wet. They often are dark brown or black due to surface coatings of dispersed organic

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

37

matter. Following rainfall or irrigation of sodic soils, surface water persists, even though the underlying soil may be dry. Phenolphthalein indicator solution dropped into the ponded water typically turns pink because of the high alkalinity. The pH of such soils and water is usually greater than 8.5. Soils subjected to sodicity conditions for significantly long periods of time and to successive periods of wetting and drying tend to develop distinctive morphological properties. Their clay content is relatively high where minerals, humus, or plant nutrients have been deposited after being washed down from the watershed (the illuvial horizons), and the structure of the upper B horizon is either columnar or prismatic. The appearance of excessive ponding or runoff during irrigations or periods of rainfall is a good visual indicator of permeability problems. The slaked and compacted (parallel orientation) appearance of the topsoil indicates crusting problems. Measurements. Measurements of soil intake rate and soil strength are sometimes made to diagnose a permeability problem. The ultimate criteria for judging permeability are (1) whether the soil water reservoir is being adequately replenished by irrigation, and (2) whether excess salts are being leached from the rootzone. Thus, to diagnose a permeability problem, the depth of wetting and degree of salt build-up in the rootzone resulting from irrigation should be measured. Fortunately, these parameters are much easier to measure than are the intake rate and soil strength. To determine the extent of soil water replenishment (or depth of wetting), changes in soil water content are measured by a suitable means. The depth of water penetration—a useful indirect measure of soil water replenishment—can be measured by inserting a pointed rod into the soil before and after irrigation. Other methods are described in Chapters 9 and 10. To determine the degree of leaching and the direction of net water flux, the magnitude and distribution of ECe within the rootzone is measured. Portable insertion probes are especially useful for this. The principles and methods of such techniques are described in Chapter 10. To diagnose a crusting problem, measurements of the strength of the surface soil could be made in the field, using some appropriate method. However, because the ultimate criterion for judging crust strength is whether seedlings emerge adequately, simple observations of emergence success provide a more appropriate method. Diagnostic criteria. Permeability is diagnosed as a problem if soil water cannot be adequately replenished by irrigation without excessive ponding or runoff. It is also judged to be a problem if leaching is inadequate to maintain a net downward flux of water through the rootzone (as deduced

38

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

from the distribution of salinity with soil depth). Additional evaluations may be necessary to determine whether the apparent permeability problems are simply the result of poor irrigation water management. Crusting is diagnosed as a problem if the strength of the surface soil is determined from observation to significantly hinder seedling emergence. Potential causes of permeability problems and crusting. Inadequate water infiltration rates usually are due to low soil permeability or topsoil crusting. These, in turn, are caused by excessive exchangeable Na, excessively low electrolyte concentrations in the infiltrating water, and poor intrinsic soil properties, including massive structure, high clay content, high bulk density, lack of sufficient macroporosity, low organic matter; or compaction of soil resulting from excessive water application rates or improper tillage and traffic operations. The cause of a permeability problem is important to know because the cause must be abated to achieve permanent solution. The various causes of inadequate permeability include underirrigation, excessive runoff, excessive sodicity, and inherently poor soil physical properties. Underirrigation is diagnosed as the cause of the permeability and crusting problem if the amount of water applied is less than the crop consumptive use. Excessive runoff is diagnosed as the problem if the difference between the amount of water applied and the amount of runoff from the field during an irrigation event is less than the crop water requirement. It may be caused by improper field grade or crusted soils. Measurements of the soil grade and observations of aggregation and crust conditions should be made to distinguish whether the excessive runoff is due to improper field grade, poor soil physical conditions, or improper irrigation system design and management. Such requirements are described in Chapters 10 and 16. To evaluate whether excessive sodicity is the cause of inadequate permeability, surface soil SAR at 0- to 15-cm depth is measured, or the adjusted SAR from the EC, SAR, Ca, and HCO3 concentrations in the irrigation water is calculated. If the SAR of the surface (0- to 15-cm depth) soil, or the adjusted SAR of the irrigation water, exceeds 10, the poor permeability probably is caused by excessive exchangeable Na, especially if the area receives appreciable rainfall or if the electrolyte concentration of the irrigation water is ⬃ 3 meq/L. Other combinations of soil SAR and irrigation water EC may also be harmful (Chapter 11). When the ECi is ⬃ 3 mmolc/L (meq/L), its harmful effect may be reduced by adding electrolyte to the irrigation water. If the adjusted SARi is high (10), its suitability may be improved by adding a Ca amendment, such as gypsum, to the irrigation water. Such measures should be undertaken only if the addition of ⬃3 mmolc/L or less of Ca sufficiently improves the SAR–EC relation. If the bicarbonate concentration of the irri-

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

39

gation water appreciably exceeds the Ca concentration, the suitability of the water for irrigation may be improved by amending it with enough sulfuric acid to neutralize up to 90% (equivalent basis) of the bicarbonate. If the EC, SAR, Ca, and HCO3 of the irrigation water do not indicate high Na, the excessive exchangeable Na might be caused by upward flow of water from a shallow water table. If a shallow groundwater with relatively low salinity and sodic-chemistry is present, and if the SAR distribution in the soil decreases with depth, upward flow of water is indicated. If the excessive exchangeable Na derives from a water table, drainage must be improved before soil reclamation is attempted. If neither the irrigation water nor the groundwater is determined to be the source of excessive exchangeable Na, the Na is indigenous to the soil. Excessive exchangeable Na must be replaced by reclamation. Methods for reclamation are described in Chapter 21. If none of these factors is determined to be the cause of low permeability, excessive compaction or inherently poor soil physical properties are likely. Some soils may be so high in clay content and so poorly structured that their infiltration rate is too low for normal irrigation practices and cropping needs. Such soils can only be irrigated “down-the-cracks,” and irrigations and cultivation must be timed at the proper soil water content. These problems are distinguished from sodic problems (though they may occur simultaneously) by the lack of excessive Na and the presence of high clay content and massive soil structure. For further discussion of soil structure and other physical properties, see Chapter 5. Plant toxicity problems and nutritional imbalance problems Visual indications. Certain crops are more severely affected by Na, Cl, and B, or the lack of Ca, than are others (see Chapter 13). When one of the susceptible crops shows stunted growth or other salt-related effects and other kinds of crops nearby do not, a toxicity problem or nutritional imbalance problem, rather than a salinity problem, may exist. Excessive foliar uptake of Cl and Na may occur with sprinkler irrigation. Therefore, if sprinkled plants show problems and nearby nonsprinkled plants do not, the toxicity is not likely soil-derived. Plant appearance should be compared with typical toxicity and deficiency symptoms to help diagnose toxicity and Ca deficiency problems. These symptoms are described below and in Chapter 10. Ca deficiency generally impairs root growth more than top growth. Affected roots tend to be white and stunted in appearance. The young leaves of new plants are affected before the older leaves, and are often distorted and small with irregular margins and spotted or necrotic areas. There may be dieback at the terminal buds. Alfalfa is a good indicator crop, because it is particularly sensitive to Ca deficiency.

40

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Symptoms of excess Cl include burning and firing of leaf tips or margins, bronzing, premature yellowing, abscission of leaves and, less frequently, chlorosis. Smaller leaves and slower growth also are typical. Symptoms of excess Na include necrotic areas on the tips, margins, or interveinal areas. Incipient injury is indicated by a mottled or chlorotic condition. Boron toxicity occurs in citrus, avocado, persimmon, and many other species. These plants may develop tip burn or marginal burn of mature leaves accompanied by chlorosis of interveinal tissue. Boron injury to walnut leaves is characterized by marginal burn and brown necrotic areas between the veins. Stone fruits, apples, and pears are sensitive to B but do not accumulate high B concentrations in their leaves, nor do they develop leaf symptoms. Cotton, grapes, potatoes, beans, peas, and several other plants show marginal burning and a cupping of the leaf that results from a restriction of the growth of the margin. Measurements. Appropriate plant and soil analysis methods are described in Chapters 9 and 10. The likelihood of a Ca deficiency, Ca/Mg imbalance, or a Na-, Cl-, or B-toxicity problem should be evaluated before chemical analyses are undertaken to diagnose other toxicity- or nutrition-related problems. The tolerance tables given in Chapter 13 should be used to evaluate crop susceptibility. Laboratory analyses of plant tissue samples or soil samples for the suspected toxicant should be made to confirm the field diagnosis. If the foliage is wetted by sprinkler irrigation and either Na or Cl toxicity is suspected, only plant-tissue analyses are necessary. Diagnostic criteria. Specific solute toxicity is diagnosed as a problem if an excessive level of B, Cl, or Na is found in the plant or soil samples. Cl and Na are particularly toxic to fruit crops and woody ornamentals. When the leaves of these plants accumulate more than about 0.5% to 1.0% Cl or 0.25% to 0.5% Na on a dry-weight basis, they develop characteristic leaf-injury symptoms. Boron is toxic to many plants when it occurs in susceptible tissue at concentrations exceeding approximately 100 mg/kg to 300 mg/kg. Tolerances vary among varieties and rootstocks and are influenced by various growth factors. Specific concentrations in plant tissue associated with toxicity vary with sampling method, crop type, and leaf position. More specific diagnostic criteria for assessing B, Na, and Cl toxicities are given in more detail in Chapter 7 and in Reisenauer (1983). Levels of B and Cl in soil saturation paste extracts associated with toxicity problems for various crops are given in Chapter 13. Ca deficiency is an important nutritional imbalance problem associated with some types of salt-affected soils. This problem occurs as a result

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

41

of either excessive soil sodicity, high pH, or a high soluble Mg/Ca ratio in the soil water. Ca deficiency is diagnosed as a problem if the concentration ratio Ca2/Mg2  0.5, or if the absolute concentration of Ca is less than 1 mmolc/L to 2 mmolc/L in the soil solution. Ca deficiency can also be determined from plant tissue tests. Guidelines for various crops are described in Chapters 6 and 13. Potential causes of plant toxicity or plant nutriient imbalances. Plant toxicities are caused by the same factors as salinity problems. Thus, the procedures for their diagnosis are the same as those for diagnosing salinity problems. When Na or Cl plant toxicity occurs and concentrations of Na, Cl, or both, in the irrigation water exceed about 5 mmolc/L, foliar uptake is diagnosed as the likely cause. Ca deficiency can be due to an inadequate concentration of Ca or an excessively low Ca/Mg ratio in the soil solution. If the absolute concentration is too low (⬃2 mmolc/L), the cause is either that the soil is too alkaline (pH 8.5) or the irrigation water is too sodic (adjusted Ca 2 mmolc/L). Inadequate Ca concentration is diagnosed as a problem if the adjusted Ca value is 2 mmolc/L. Ca/Mg imbalance is diagnosed as a problem if the ratio [(adjusted Ca)/(Mgi)·(Fc.)] is 0.5. Methods for calculating adjusted Ca and Fc are given in Chapters 3 and 11, respectively, and their use in water quality assessment is discussed in Chapter 11. Deficiencies of other nutrients, such as P, Fe, and Zn, may be evaluated by analyzing soil samples and plant tissues. Procedures for doing so are described in Chapters 7 and 10. Under sodic soil conditions, permeability and crusting conditions are generally more limiting to plant growth than are nutrient deficiencies, and such deficiencies will usually be eliminated by removing the excess sodicity.

ASSESSING RECLAIMABILITY Once the cause of a salinity problem has been determined, steps should be taken to abate the cause and eliminate the excess salts, toxicants, or sodicity by reclamation. Good drainage is the first requirement for the reclamation of any salt-affected soil. After drainage is provided, excessive salinity and toxic ions can be reduced to acceptable levels by leaching. Some soils, however, have such poor physical properties that leaching is impractical. In sodic soils, for example, amendments may be applied to increase soil permeability and reduce the exchangeable Na content. Before any reclamation is undertaken, it should be determined whether leaching, amendments, or special tillage operations are necessary.

42

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Determine Leachability The feasibility of leaching without amendments may be estimated in two ways. The first method involves plotting the Na-adsorption ratio of the saturation extract or the irrigation water against the ECi, on a figure similar to Fig. 11-3 in Chapter 11. If the plotted point falls into the area of the figure that indicates permeability hazards are unlikely, leaching without amendments is feasible. The second method involves leaching packed soil columns with samples of the irrigation water and with 0.05 N CaCl2 solution. Reclamation without amendments is unlikely to be successful if the permeability of the soil leached by irrigation water is very low relative to the case where the same type of soil is leached by CaCl2. This evaluation may also be made in the field using infiltrometers. These two methods provide estimates of the reclaimability of a soil without amendments but should not be relied on for a final decision. The final evaluation of leachability should be based on the results of field trials. To set up a field trial, salinity-sensing devices and tensiometers are installed at various depths in the field plot. The plot is then leached, using the same water and method of application as will be used for full-scale irrigation. The changes in salinity and hydraulic gradient are measured as a function of time and volume of water applied (or infiltration). Alternatively, measurements of the rate of water infiltration can be used to predict the soil’s reclaimability, as described in Chapter 21. Reclamation by leaching without amendments or special tillage operations is likely to be successful if the rate of leaching or the permeability measured in the field trials is relatively high and sustainable. If soil permeability is found to be inadequate for reclamation by simple leaching, the viability of increasing it sufficiently with the use of amendments should be ascertained. Low soil permeability may result from the low EC of the leaching water, excessive ESP, or the inherently poor physical properties of the soil. To determine whether the electrolyte content of the leaching water can be increased by the addition of amendments enough to make leaching practical, the values of ECi and SAR for the 0- to 15-cm soil depth are plotted on Fig. 11-3 of Chapter 11. If adding electrolyte to the leaching water or using another water of higher salinity, such as drainage water, moves the point from the hazard area into the nonhazard area of the figure, reclamation with such amendments or waters may be feasible. The potential benefit of an amendment or alternative water supply can also be evaluated by leaching a packed soil column with the proposed leachant and determining whether the permeability increases appreciably. This test can also be made in the field using an infiltrometer. A solution of 0.05 N CaCl2 is recommended for this purpose.

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

43

Various soil amendments, such as gypsum, CaCl2, H2SO4, and highelectrolyte water, should be evaluated to determine the most suitable one for increasing permeability. Laboratory column studies, field infiltrometer studies, and the WATSUIT software program (Chapter 25) may be used for this evaluation. The water intake rate of sodic soils is often too low to reclaim the soil at an acceptable rate, especially when gypsum is used as an amendment. Using CaCl2 or H2SO4 instead of gypsum usually eliminates this problem. Another factor affecting the suitability of amendments is their acidity or acid-forming properties. If the soil contains at least 0.5% CaCO3 equivalent, an acid-type amendment can be used. Under some circumstances, the dissolution of CaCO3 can be enhanced by the presence of growing plants (which evolve CO2), supplying the needed Ca and electrolyte concentration to accomplish reclamation. The final decision on amendment requirement and rates should be based on the results of field trials. For more discussion of amendments and reclamation procedures, see Chapter 21. Determine Adequacy of Soil Structure and Need for Tillage If the soil permeability is inadequate when tested using a 0.05 N CaCl2 solution or an acidifying soil amendment, poor physical properties of the soil are the likely cause of the low permeability. The soil profile may often be sufficiently altered by deep plowing, deep chiseling, shattering, or trenching operations to achieve an adequate leaching rate. To help to determine whether these methods will succeed for a given soil profile, note the presence or absence of a topsoil crust or permeable materials deeper in the profile that could be brought up by tilling and mixed with permeabilitylimiting strata. Unfortunately, deep tillage can also bring up saline material into the rootzone. This potential hazard must be carefully evaluated by measuring the salinity of the deeper material. If mixing seems like a possible solution, additional chemical and physical analyses of the deeper soils should be made. The final decision on the reclaimability of the soil by tillage and amendments should be based on the results of field trials. Estimate Amount of Leaching and Time Required for Reclamation Once it is determined that the soil can be reclaimed by leaching with or without amendments and tillage, the amount of leaching required to reclaim the soil and the amount of time needed for reclamation must be estimated. One method involves installing a succession by depth of salinity-sensing devices in the field and then measuring the change in ECe by soil depth, time, and depth of infiltrated water. This directly, simply, and accurately determines the reclamation requirements. Such sensors are described in

44

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Chapter 10. Other methods for estimating the amount of leaching and time necessary for saline or sodic soil reclamation are given in Chapter 21. Other Considerations If the concentrations of Cl, B, or Mg in the soil are excessive, the excessive amounts must be removed by leaching. Assessing the reclaimability of such soils is similar to the assessment of reclaimability of soils containing excessive salinity. If nutrient deficiency poses a problem, the methods for assessing reclaimability are similar to those described for assessing the reclaimability of sodic soils, with particular emphasis on pH reduction.

SELECTING SALINITY CONTROL AND MANAGEMENT MEASURES Soil and water salinization inevitably occur with irrigation. Salinity control measures must therefore be implemented if irrigated agriculture is to be sustained over the long term. Ideally, such measures should be compatible with processes in the natural geohydrologic system. Management practices offer different levels of control. Some are aimed at controlling salinity within the rootzone in a particular field; others control salinity over a much larger scale, such as an irrigation project. The practices should protect off-site environments, such as surface water and ground water. On-farm salinity-management practices consist of agronomic and engineering techniques applied on a field-by-field basis. Management practices applied on a district-wide scale generally consist of structures for water control (both delivery and discharge) and systems for the treatment and disposal of drainage waters. Usually, no single method suffices to control salinity of an irrigated land. Instead, many practices are combined. The appropriate combination of salinity control methods depends on economic, climatic, social, edaphic, and hydrogeologic factors. Thus, diagnostic procedures cannot be given here for selecting a generic set of control practices. Rather, some important goals, principles, and practices that should be considered in selecting salinity-management practices are reviewed. Controlling Plant Toxicity and Rootzone Salinity and Sodicity Management practices for the control of salinity, sodicity, and toxicity in the rootzone include: 1. Selecting crops or varieties (rootstocks) that produce satisfactory yields under the existing conditions of salinity, sodicity, and toxicity.

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

45

2. Using planting procedures that minimize salt accumulation around the seed. 3. Using land preparation methods that facilitate uniform infiltration of irrigation water. 4. Adopting irrigation procedures that maintain sufficient available soil moisture and cause periodic leaching of the soil profile. 5. Installing and maintaining drainage systems that facilitate leaching, rooting, and trafficability. 6. Using treatments, including chemical amendments and organic matter, to maintain soil permeability and tilth. Soil properties, water table depth, and the quality of irrigation water also must be considered when selecting a particular management practice. Growing suitably salt-tolerant crops A crop that produces satisfactory yields under present or anticipated conditions of salinity, sodicity, or specific solute concentrations should be selected using the tolerances given in Chapter 13. Barley (Hordeum vulgare), sugar beet (Beta vulgaris L.), cotton (Gossypium spp.), Bermuda grass (Cynodon spp.), Rhodes grass (Chloris gayana), western wheat grass (Pascopyrum smithii), bird’s-foot trefoil (Lotus corniculatus), table beets (Beta vulgaris var. rubra), kale (Brassica napus and B. oleracea), asparagus (Asparagus officinalis), spinach (Spinacia oleracea), and tomato (Solanum lycopersicum) are all highly salt-tolerant. Radish (Raphanus sativus), celery (Apium graveolens), beans (Family Leguminosae), clovers (Trifolium spp.), and nearly all fruit trees are low-tolerance crops. The maximum expected levels of salinity, sodicity, and specific solute concentration resulting from long-term irrigation can be predicted by using the models described in Chapters 25, 28, 29, and 30. Minimizing salt accumulation in the seed bed Many crops are more sensitive to salinity during early growth. Such crops are particularly susceptible to losses in stand due to high levels of salinity in the seedbed (see Chapter 13). Failure to obtain a satisfactory stand frequently limits the successful cropping of moderately saline soils. Establishment of seedlings in furrow-irrigated row crops is particularly difficult because soluble salts tend to accumulate in raised beds that are wetted by subbing water from the furrow (see Chapter 16). Practices that facilitate establishment of seedlings in saline soils involve modification of seedbed shape, seed number, and placement and irrigation techniques. To speed germination, the seeds may be presoaked before planting. This is effective, yet seldom used. As discussed in Chapter 16, salts tend to accumulate near the seed during furrow irrigation. This tendency is greatest in single-row, round-topped

46

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

beds. Such beds should be avoided under saline conditions. With doublerow beds, most of the salt is carried into the center of the bed, leaving the shoulders relatively free of salt for seedling establishment. Sloping beds are best for saline soils because seedlings can be established on the slope below the zone of salt accumulation. The salt is moved away from the seedlings instead of toward them. Planting in furrows or basins is satisfactory from the standpoint of salinity control but often unfavorable for the emergence of many row crops, because less-favorable crusting and aeration conditions exist in furrows and basins. Pre-emergence irrigation by sprinklers or drip lines placed close to the seed line may help to keep the soluble salt concentration low in the seedbed. Also, temporary small furrows may be used instead of drip lines. After the seedlings are established, these furrows may be abandoned and new furrows made between the rows. The bed shape may also be altered and sprinkling replaced by furrow irrigation. The seeding rate should be increased to help to counter the loss of stand associated with germination and emergence problems. Alternatively, seedling transplants can be used to establish a stand in saline soils. Applying irrigation water uniformly and efficiently Irrigation management is of prime importance in the control of salinity (see Chapters 12 and 16). Precise land leveling is necessary to facilitate the uniform application and infiltration of irrigation water and, hence, to control salinity. Barren or poor areas in otherwise productive fields often occur on high spots that do not infiltrate enough water for crop growth or leaching. Usually, lands that have been irrigated for 1 or 2 years after initial leveling should be replaned to remove unevenness caused by the settling of fill material. Annual crops should be grown following the first leveling so that replaning can be performed before perennial crops are planted. Flood irrigation is suitable for salinity control if the land is sufficiently level, though soil aeration and crusting problems may occur. Furrow irrigation is well adapted to row crops and land that is too steep for flooding. Aeration and crusting problems are minimal with furrow irrigation, but salts tend to accumulate in the beds, as discussed. If excess salt accumulates, a periodic change of irrigation method to flooding or sprinkling, along with a switch to grain crops or other more salt-tolerant crops, is advisable. Alternatively, after the seedlings are established, furrow depths can be made shallower or irrigation depths increased so that the beds will be inundated and leached. Irrigation by sprinkling allows superior control of the amount and distribution of water; it is, therefore, often used on steep land. There is a tendency to apply too little water for leaching requirements with this

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

47

method, and leaching of salts beyond the rootzone often requires special effort. As explained earlier, crusting is more likely to become a problem with sprinkler irrigation. Another potential hazard of sprinkler irrigation is foliar uptake of salt and leaf burn due to contact with water. Sprinkler irrigation should be avoided if the water contains excessive levels of Na and Cl, although sprinkling at night can help in such cases. Subirrigation, in which the water table is maintained close to the soil surface, is not advised when salinity is a problem, unless the water table can be lowered periodically and the accumulated salts leached out by rainfall or by periodic surface applications of water. Drip irrigation, if properly designed, minimizes salinity and matric stresses because the soil water content is maintained at a high level and the salts are leached to the perimeter of the wetted volume, where rooting activity is minimal. Drip irrigation is usually the method of choice when the water is high in salts, though the high build-up of salts in the fringe of the wetted area may eventually become a problem. Solid-set, linear-move, and center-pivot sprinkler systems give good control and distribution of water if managed properly. Gravity systems, if designed and operated properly, can also achieve good control. For tree crops, a low-head bubbler system provides excellent control while minimizing pressure requirements. Laser-controlled, precision land leveling allows better areal water distribution and smaller applications. Combined with automation, it has led to high irrigation efficiencies for dead-level, flooded systems. Such systems improve irrigation efficiency and salinity control. Excessive loss of irrigation water from unlined distribution canals is a major cause of high water tables and saline soils in many irrigation projects (see Chapter 18). Such losses should be reduced by lining the canals with impermeable materials or by compacting the canal wall to achieve low permeability. Closed conduits, rather than open waterways, should be used for laterals wherever possible, because of their more effective off–on control. In addition, they can provide gravitational energy for pressurizing delivery systems or controls. In furrow-irrigated areas, intake uniformity can sometimes be increased and tail water runoff reduced by shortening furrow lengths and using multiset systems. For some soils, surge irrigation can also be used to improve irrigation uniformity in graded furrows. It is important to recognize that inefficient irrigation is the major cause of salinity and shallow water tables in most irrigation projects of the world and that the need for artificial drainage can be substantially reduced through improvements in irrigation management. Many delivery systems encourage overirrigation by supplying water according to fixed schedules or in fixed amounts, regardless of seasonal variations in crop needs. Ways to improve irrigation efficiency should be sought before drainage capacity is increased.

48

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Irrigating frequently to maximize soil water potential The optimum irrigation scheme provides water continuously at a rate sufficient to keep the soil water content in the rootzone within narrow limits, although carefully programmed periods of stress may be desirable to obtain maximum yield with some crops. Cultural practices also may demand periods of dry soil. Thus, the timing and amount of water applications should be carefully controlled to achieve high crop yield and water-use efficiency, especially in saline soils. This calls for water delivery to the field on demand, which, in turn, requires the farmer and the organization that distributes the water to coordinate their efforts closely. The level of salinity that can be tolerated in the soil depends on the distribution of salinity in the soil profile, the frequency and extent to which the soil water is depleted between irrigations, and the water content of the soil. Plants usually tolerate higher levels of salinity under conditions of low matric stress, such as is achieved with high-frequency irrigation methods, such as drip irrigation. High soil-water salinities occurring in deeper regions of the rootzone should be offset by sufficient low-salinity water added to the upper profile. Irrigations should be implemented so as to maximize soil water potential, thereby minimizing the effects of salinity on soil water availability, and to minimize deep percolation, thereby minimizing drainage and pollution problems. Frequently, the efficiency of on-farm water use is limited by lack of knowledge about when to irrigate. Ideal irrigation management would supply enough water to hold the soil water content near field capacity at planting time and reduce water content at least 50% at harvest, and it would maintain water within the rootzone during the major period of vegetative growth as high as possible without incurring aeration or leaching problems. Under saline conditions, this requires that water be applied in excess of consumptive use. Irrigation scheduling requires some method of assessing the water availability to the crop with sufficient lead time to provide for a water application before significant plant stress occurs. In addition, the quantities of water needed for replenishing depleted soil moisture and leaching must be estimated. Both direct and indirect measurements are commonly used to predict the onset of plant water stress and the soil water deficit. Chapter 16 describes some of these methods in detail. Providing leaching and drainage to avoid rootzone salt accumulation Additional water above the amount required by plant transpiration and evaporation must be applied, at least occasionally, to leach out the salt that has accumulated during previous irrigations. This leaching

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

49

requirement depends on rainfall, the salt content of the irrigation water, and the maximum salt concentration permissible in the soil solution. The leaching requirement should be estimated by using the empirical relations given in Chapter 12, the crop production relations given in Chapter 13, or the WATSUIT model described in Chapter 25. Often, much of the necessary leaching is achieved during preirrigations and early-season irrigations, when soil permeability is at its seasonal maximum. If rainfall is significant, especially between cropping periods, then the necessity of irrigative leaching will be reduced or eliminated. The control of salinity by leaching is accomplished most easily in permeable, coarse-textured soils. Medium-textured soils, with their greater water-holding capacity, ordinarily present no major problem from the standpoint of salinity control if they have good structure and overlie subsurface materials that facilitate the removal of drainage water. Preventing salt accumulation is most difficult in fine-textured, slowly permeable soil, especially if drainage is poor. For more discussions on leaching requirements, see Chapter 12. If natural drainage is insufficient, it should be supplemented by drainage installations. For more discussion of drainage and irrigation management for salinity control, see Chapter 16. For a discussion of irrigation requirements, see Chapter 12. For more discussion on project-level management, see Chapter 19. Controlling Salinity of Water Resources Irrigated agriculture cannot be sustained without adequate leaching and drainage to prevent excessive salinization of the soil. Yet these processes are the very ones that contribute to the salt loading of our rivers and groundwaters. Several approaches are available for minimizing salinity pollution of water resources that result from irrigation activities. First, irrigation can be eliminated. This should be undertaken in instances where the detrimental effects of irrigating the land outweigh the benefits. Second, the amount of water lost in seepage and deep percolation can be reduced, lessening the amount of saline water that passes through the soil and substrata. Third, point sources of drainage-return flow into streams or rivers can be intercepted and diverted to other outlets and uses. For example, saline drainage water can be desalted and reused, disposed of by evaporation in ponds or by injection into some isolated deep aquifer, or used as a water supply for applications in which brackish waters are acceptable. The second and third approaches are discussed in greater detail next.

50

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Minimize leaching and deep percolation With irrigation water containing between 0.05 tons and 3.5 tons of salt per 1,000 m3 and crops requiring annual applications of 6,200 m3 to 9,300 m3 water per hectare to meet ET, an average of 0.3 tons to 32 tons of salt/ha are added annually to irrigated soils. Clearly, if the volume of water applied for irrigation can be minimized, then the amount of salt added to the soil and the amount that must be removed by leaching will be minimized. The leaching fraction, L, should also be minimized, to maximize the precipitation of applied Ca, HCO3, and SO4 salts as carbonate and gypsum minerals in the soil and to minimize the amount of weathered and dissolved salts picked up from the soil and substrata. In the case described, by reducing L from 0.3 to 0.1, the amount of salt discharged from the rootzone can be reduced by as much as 2 tons/ha to 12 tons/ha annually. The extent to which leaching can be minimized is limited by the tolerances of plants to increased salinity in the lower part of the rootzone. In most irrigation projects, current leaching fractions could be reduced appreciably without harming crops or soils, especially with improved irrigation management (see Chapter 12). Minimizing leaching may or may not reduce the degradation of the receiving water by salinity. For example, rivers essentially saturated with CaCO3 and gypsum will not benefit from reduced leaching unless salts other than those derived from the diverted water or from soil mineral weathering and dissolution in the rootzone are encountered in the drainage flow path. Each situation must be evaluated according to its specific hydrogeologic conditions. Like surface waters, groundwaters receiving irrigation drainage water may or may not benefit from reduced leaching. If no sources of recharge other than irrigation drainage exist, the chemical composition of the groundwater will eventually approach that of the drainage water. Intercept, isolate, and reuse drainage The ultimate goal for efficient irrigation is to prevent the mixing of highly saline waters and freshwater irrigation supplies and to maximize the utilization of an irrigation water supply, with minimum drainage. If drainage water from one irrigation is acceptable for use by a crop of higher salt tolerance, it should be used again, ideally in the same project. This can be achieved by intercepting the water draining from the rootzone, isolating it from higher-quality water supplies, and using it to irrigate crops of appropriate salt tolerance. Such reuse of drainage water reduces the salinity of the associated river system or groundwater reservoir because the salt load of the drainage return would not be added to it.

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

51

A saline drainage water may be envisioned as consisting of an usable component (which varies with the plants’ salt tolerance) and an unusable component. When the growth-limiting factor is salinity, the ultimate fraction of water in a supply that can be used for crop growth is: 1

ECiw ECm

(2-3)

where ECiw  the electrical conductivity (alternatively, salt concentration can be used) of the water supply, and ECm  the maximum electrical conductivity (or salt concentration) that the plant can tolerate in the rootzone (in terms of the EC of the soil water). Values of ECm vary among the crop species. Approximate typical values are 45 dS/m for highly salt-tolerant crops, such as cotton, sugar beets, and barley; 30 dS/m for crops of intermediate salt tolerance, such as tomatoes, wheat, and alfalfa; and 15 dS/m for sensitive crops, such as beans, clovers, and onions (Bernstein et al. 1975). If the usable component of drainage water is allowed to mix with freshwater and the blend is used to irrigate a crop of the same salt tolerance, then the plant can, at best, only consume through transpiration the freshwater fraction of the mix, and the same “unusable” fraction will pass through the profile again without contributing to transpiration and growth. In the second drainage, more salt may be displaced, or “picked up.” Greater flexibility and opportunity for crop production results if the initial drainage water can be intercepted, isolated, and prevented from returning to the general water supply. The waters can then be blended or left separate for other uses, such as the irrigation of appropriately salt-tolerant plants. Once the waters are mixed, such alternatives are lost. Thus, the best strategy for controlling the salinity of water supplies is to intercept drainage waters before they can flow into low-salinity water supplies, and to substitute the drainage water for conventional irrigation water after seedling establishment of certain crops in the rotation. When the drainage water is too salty for reuse, it should be discharged to evaporation ponds or other appropriate outlets. This strategy will conserve water, sustain crop production, and minimize the salt loading of freshwater supplies that occurs by irrigation return flow. It will also reduce the amount of freshwater diverted for irrigation. The feasibility of this strategy is supported by the following: (1) the maximum possible soil salinity resulting from continuous use of drainage water does not occur during the time such water is only used for a fraction of the time; (2) substantial alleviation of salt build-up resulting from irrigation of salt-tolerant crops with drainage water occurs when salt-sensitive crops are irrigated with fresh water; (3) proper preplant irrigation and careful irrigation management during germination and seedling establishment leaches salts out of the seed area and from shallow soil depths;

52

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and (4) data from field experiments to date support the credibility of this cyclic reuse strategy. For more on the reuse of drainage waters for irrigation, see Chapter 22. Dilution of return flows is often advocated for controlling water salinity and increasing the water supply for irrigation. But, as previously discussed, this concept has serious limitations when its effect on the true volume of usable water for crop transpiration is considered. It is not advocated as a general method of salinity control or water supply enhancement. For more discussion of river basin water and salt flows, salinity control measures, and treatment and disposal options, see Chapters 19, 22, and 23. Summary The natural and artificially induced hydrogeological processes in a given system must be carefully analyzed before management practices for controlling salinity pollution are selected. The relative contributions of rootzone drainage and seepage should be assessed so that priorities for pollution control can be established. For further discussion of this topic, see Chapter 19.

USING MODELS TO IDENTIFY POTENTIAL SALINITY PROBLEMS Modeling Methods Analytical quantitative techniques and numerical computer models are available for predicting soil solution composition, crop response to water applications, and the quantity and quality of subsurface return flows under a variety of water management alternatives. Deterministic, or process-oriented, models can simulate the physical and chemical processes that occur as water moves through the soil profile. The processes include salt precipitation, mineral weathering, and cation exchange reactions. Stochastic models and optimization models may be used to predict problems that are likely to result from the development of new irrigation projects. Management plans can then be developed to prevent or solve these problems in a more effective and timely manner. Economic models can be used to estimate the damages that may occur as a result of salinization of soils and water. Such economic models can also be used to predict the local, state, regional, and national benefits that result from implementation of a salinity control program. Area-wide modeling studies can define the need for and potential benefit of salinity control measures. These studies pinpoint the sources and

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

53

causes of salinity and provide information for selecting the most appropriate control measures. Once the sources of salts are defined, more detailed, local-scale models may help to specify how those sources can be controlled. For more information on the use of models for salinity evaluation and control, see Chapters 14, 19, 25, 28, 29, and 30. Monitoring Methods A viable, permanent, irrigated agriculture system requires periodic information on soil salinity. Only with this information can the salt balance and water-use efficiency of irrigation projects be assessed. The collection and monitoring of soil salinity is complicated by the spatial variability of salinity data. Typically, numerous samples are needed to characterize an area. Monitoring is also complicated by changes in salinity that occur over time due to changing weather patterns, changing management practices, and movement of the water table. Representative areas for collecting samples can be established within irrigation projects to minimize the number of samples required. Also, geophysical instruments often provide efficient, fast measurements of parameters, such as soil EC. A network of representative soil salinity monitoring stations should be established in irrigation projects, especially those projects undergoing changes in operation. When ECe measurements are made at a number of locations, maps of the distribution of ECe by area and depth can be prepared. Iso-lines of ECe on such maps delineate the areas of salinityaffected soil. Shallow water tables often are responsible for such salt accumulations and can be associated with them if contour map overlays of the surface elevations and the groundwater are prepared. Maps of the net vertical soil-water fluxes can help locate areas of inadequate drainage or salt loading. Techniques and equipment for measuring ECa and estimating ECe are described in Chapters 9, 10, and 14.

SUMMARY The diagnosis and treatment of soil salinity problems are complex. There are numerous causes of salinity-related poor soil conditions, and these may vary from site to site at plot, farm, and regional scales. There are a number of visual indications of various salinity problems and attention to these may be an initial step in the diagnosis of problems, but subsequent field measurement of the nature and spatial distribution of each problem may be required before a management response can be developed. Responses to salinity problems may range from source control

54

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

to crop choice to irrigation management (including leaching and drainage). To sustain crop production on saline and sodic soils, it is probably necessary to establish a long-term monitoring program, including periodic routine field measurements to track changes soil conditions in response to (1) changes in natural conditions, such as precipitation and runoff, and (2) changes in response to management. Various management models can enhance understanding of the problems and the response of soils and crops to the salinity management measures being taken on a plot, farm, or regional scale. The chapters that follow provide detailed analyses of the full range of problems facing growers, techniques for diagnosing these problems, and range of management options available.

REFERENCES Bernstein L., Francois L. E., and Clark, R. A. (1975). “Minimal leaching with varying root depths of alfalfa,” Soil Sci. Soc. Am. Proc., 39, 112–115. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment,” J. Irrig. Drain. E., ASCE, 103(2), 115–134. Reisenauer, H. M. (1983). Soil and plant tissue testing in California. Bulletin 1879. Div. of Agricultural Science, University of California, Davis, Calif. Rhoades, J. D., Kandiah, A., and Mashali. A. M. (1992). The use of saline waters for crop production, FAO Irrigation and Drainage Paper 48, Food and Agriculture Organisation of the United Nations, Rome.

NOTATION a  threshold level of soil salinity at which yield decreases begin b  percentage yield loss per increase of salinity in excess of a Em  maximum electrical conductivity that plant can tolerate in the rootzone ECa  electrical conductivity of the bulk field soil ECe  electrical conductivity of paste extracts of soil samples ECi  electrical conductivity of irrigation water ECiw  electrical conductivity of water supply ECp  electrical conductivity of soil paste ET  evapotranspiration ETc  estimated crop evapotranspiration requirement Fc  concentration factor I  infiltration rate I·T1  amount of water infiltrated over irrigation period L  leaching fraction

DIAGNOSIS OF SALINITY PROBLEMS AND SELECTION

55

LA  theoretically achievable leaching fraction Lr  leaching requirement Tt  irrigation set period Yr  percentage of yield of crop grown under saline conditions relative to that obtained under nonsaline, but otherwise comparable, conditions

This page intentionally left blank

PART TWO: EFFECTS OF SALTS ON SOILS

CHAPTER 3 THE CHEMISTRY OF SALT-AFFECTED SOILS AND WATERS Donald L. Suarez and J. J. Jurinak

INTRODUCTION This chapter describes the origins of salts in soil and water and the major pathways by which they accumulate in soils, particularly in soils of arid regions. The basic chemistry of soil-water systems is described, with a focus on the most common salts and the complex chemical interactions within various soil types and combinations of dissolved minerals and salts in the soil solution. Basic methods for calculating salt concentrations are described for a number of different salts, and the complex interactions of salts and variables, such as pH are described.

ORIGIN OF SALT IN SOIL AND WATER The primary source of salts in soil and waters is the continuous geochemical weathering of rocks that form the upper strata of the earth’s continental crust. It represents one step in the geochemical exchange of matter between the land, oceans, crust, and mantle that has been in existence throughout geologic time. Weathering Because most rocks have formed under high temperature and pressure, the constituent minerals usually are thermodynamically unstable when exposed to atmospheric conditions. Weathering is a spontaneous process that transforms primary minerals into other minerals that are 57

58

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

more stable at the earth’s surface. This process is, of course, critical to understanding soil formation and release of salts over geologic time. In most instances the rates of silicate reaction are slow, as there are important kinetic constraints. As a result, silicate weathering influences salt chemical composition and initial salt load but is not important at the management time scale. The reagents involved in geochemical weathering include atmospheric water, oxygen, and carbon dioxide. The biosphere enhances weathering because of its increased levels of CO2, organic matter, and biological activity. Organic matter can serve as a reducing agent and a source of organic acids, which promote weathering and cation migration through complexation. Three types of reactions describe the chemical weathering of rock-forming minerals: congruent dissolution, incongruent dissolution, and reduction-oxidation (redox) reactions. Some of these reactions are reversible, denoted with the ↔ symbol, and some are irreversible, denoted by the → symbol. Congruent dissolution In this reaction, the solution products exist in the same proportion as they occur in the mineral, that is, Mineral → soluble species Examples of congruent weathering processes are CaCO3 (calcite)  H2O  CO2 ↔ Ca2  2HCO 3

(3-1)

SiO2 (quartz)  2H2O ↔ H4SiO4

(3-2)

Most, but not all, minerals that release salt in soils dissolve congruently. These include chloride, sulfate, and most carbonate salts. Incongruent dissolution In this reaction, part of the mineral dissolves and leaves behind a secondary solid phase (secondary aluminosilicate clay minerals) that differs in composition from the original mineral: 2NaAlSi3O8 (albite)  3H2O → Al2Si2O5 (OH)4 (kaolinite)  4SiO2 (quartz)  2Na  2OH

(3-3)

3KA1Si3O8 (orthoclase)  2CO2  14H2O → 2K  2HCO3  6H4SiO4  KA1Si3O10(OH)2 (mica)  4H2O

(3-4)

CHEMISTRY OF SALT-AFFECTED SOILS

59

The silicate minerals, including those found in soils, control the earth surface chemistry under natural conditions over geologic time. However, the silicate dissolution reactions are mostly so slow under natural conditions (Suarez and Wood 1996) that they can be neglected in terms of anthropogenic time scales for salt loading and prediction of soil-solution composition under arid and semiarid lands. Redox reactions Changes in the oxidation states of minerals modifies the weathering process. Redox reactions between ions dissolved in solution and minerals in contact with that solution often influence the pH of the solution and subsequently its composition, for example, FeS2(pyrite)  15/4O2  7/2H2O ↔ Fe(OH)3(s)  4H  2SO2 4

(3-5)

In this reaction, the protons produced have a strong local influence on subsequent weathering and salt release. Oxidation of reduced S materials (including elemental S) has been used in reclamation of calcareous sodic soils. In this instance, the reactions 2S  3O2 ↔ 2SO3

(3-6)

SO3  H2O ↔ 2H  SO 2 4

(3-7)

occur slowly, releasing H that in turn reacts with the calcite in the soil, CaCO3  2H ↔ Ca2  CO2  H2O

(3-8)

resulting in net release of Ca2 and SO 2 4 . Reaction of the acid produced in Eq. 3-5 would result in a similar solution composition. In the absence of carbonates, silicates would slowly dissolve, releasing cations from the minerals, with the protons neutralizing the hydroxols released from weathering. Climate and Landscape Effects Although weathering is continuous and occurs universally, the intensity and extent of the weathering reactions strongly reflect the influence of climate. The presence of water is most important to the weathering process. It serves as a reactant in mineral transformation and is the medium that transports dissolved and suspended matter from the system. The transport of the weathering products depends on sufficient rainfall to

60

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

move the soluble salts through the surface soil into the groundwater, eventually into rivers, and, ultimately, the oceans. The oceans’ chemical composition reflects the constant inflow of salt from the land masses as modified by chemical interaction with the oceanic sediments, evaporation, volcanic outgassing, and aerosol transport. Salt-affected soils are natural components of arid landscapes. Their presence directly correlates with limited rainfall, that is, in arid landscapes the evapotranspiration (ET) greatly exceeds precipitation throughout most of the year. Lack of moisture limits the intensity of the chemical weathering of minerals. Lack of moisture also limits the movement of the product of weathering (salts), and the secondary minerals formed often are constrained to a localized area. In subhumid and humid areas, the properties of the parent rock largely dictate the properties of the soils formed and most solutes are displaced from the watershed in drainage. Most arid-zone soils are classified under two soil orders: entisols, soils that have little or no development of pedogenic horizons; and aridisols, soils that do not have water available to mesophytic plants for long periods and contain only small amounts of organic material. Because water serves as the principal transport vehicle for salts, salinity is closely linked to lowlands or depressions into which water drains and accumulates. Salinization is enhanced when restricted soil drainage promotes a high water table and the balance of mineralized groundwater is regulated by the evaporation of water, transpiration, or both, rather than by surface runoff and drainage. Areas of impeded drainage vary in size from a fraction of a hectare to thousands of square kilometers, such as Utah’s Great Salt Lake basin and Pakistan’s Indus Plain. Fossil or Secondary Deposits Throughout geologic time, saline seas have inundated large areas of present-day continents. These submerged areas have subsequently been uplifted. The resulting geologic formations provide parent material for soils as well as outcrops and underlying saline strata to soils or other formations, all of which are important zones of contact for salt loading of surface and groundwater. The secondary deposits (sedimentary rocks) formed from inland seas and weathering of continental rock during inundation are the major sources of salinity and sodicity. The term “fossil salt” has been used to describe the salinity of these deposits. These deposits, mobilized by irrigation and rainfall, are the major sources of salt loading in western U.S. surface waters. A good example of a secondary saline deposit that markedly affects the salinity of a region is the Mancos shale formation in the upper Colorado River basin. This formation, deposited in inland seas in the late Cretaceous epoch, reaches a maximum thickness of about 2,000 m

CHEMISTRY OF SALT-AFFECTED SOILS

61

(Williams 1975). Mancos shale forms an extensive belt of erodible outcrops and underlying strata in southern Wyoming, eastern Utah, and western Colorado. This formation is a major contributor to the salt load in the Colorado River. Atmospheric Deposition The atmospheric deposition of salt can be of localized importance. Dry and wet aerosol fallout contributes up to 100 kg/y-ha to 200 kg/y-ha along sea coasts and from about 10 kg/y-ha to 20 kg/y-ha in the interior. The composition of atmospheric salt deposition varies with distance from the source. The salt is predominately an NaCl-type at the coast, consistent with the composition of the oceans. It becomes proportionately greater in Ca2 and SO42 ions as the air mass moves inland. Atmospheric contributions to the salt load of arid lands can be from 10% to 25% of the total yearly contribution of weathering (Bresler et al. 1982), but are a much smaller contributor in regions with saline geologic materials. In the overall picture of salinization of soils, the contribution of atmospheric salt is often overlooked, but it is a factor that must be considered in highly weathered landscapes that have poor drainage. A case in point is a large area of saline soils in western Australia. The source of the salts is attributed to long-term inputs from rain in combination with limited drainage and concentration of the salts by surface evaporation and extraction of water by tree roots and subsequent transpiration, leaving most of the salt in the unsaturated zone. Anthropogenic Activities Soils made saline by humans are of major historical and economic importance. Industrialization has increased the atmospheric loading of gaseous nitrogen and sulfur components, both of which result in acidic fallout, which intensifies the soil mineral weathering rate. This impact is of importance primarily in nonsaline areas. Energy-related mining activities have brought to the surface saline and sodic materials that, if left in the ground, would have had little effect on the environment. At the surface in the presence of rain, they are mobilized and contribute to salt loading of surface and shallow groundwaters. Most recently, development of coal bed methane resources in the western United States has resulted in pumping of saline groundwater to dewater the methane-containing materials. Discharge of these waters into natural drainage ways increases salt concentrations in receiving surface waters, while surface application increases surface salinity. Agriculture, both irrigated and nonirrigated, has had a dramatic effect on salt distribution in the terrestrial system. All irrigation water contains

62

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salt in varying amounts and differing types. During ET the plant extracts water with a much lower salt content than in the water source. The majority of the salt (85%–95%) applied in the irrigation water is left behind in the remaining water, inevitably resulting in a drainage water more saline than the applied water. For example, for every 100 mg/L salt in irrigation water, one megagram of salt is added per ha-m of applied water. If one ha-m of water with a salt concentration of 850 mg/L is applied to a crop during the growing season, 8.5 megagrams (8,500 kg or about 9.4 tons) of salts are added. Without salinity management, salts will eventually accumulate in the rhizosphere and affect crop yield. Management of salinity in the rootzone requires application of quantities of water above the actual water use of the crop (for leaching). Inevitably, more water is transported down into the unsaturated and shallow groundwater, thus greatly increasing the mobilization of salts present in those zones. The rate of mineral weathering, primarily gypsum, calcite, and dolomite, is also enhanced by the larger volumes of water passed through the soil under irrigation. The control of salinity or sodicity in the rhizosphere and in surface or ground waters receiving drainage waters is, therefore, closely associated with soil and water management practices.

MEASUREMENT OF SALINITY The chemical and physical properties of a salt-affected soil reflect the amount and type of salt present. Although the use of salinity sensors in the field is increasingly common (see Chapter 10), laboratory analysis of aqueous extracts of soil is still the most common technique for assessing salinity and other potential hazards. One of the earliest methods for determining the amount of dissolved salts is based on evaporating a given volume of water or soil extract and measuring the weight of the residue. The result is called the total dissolved solids (TDS) and the dimensions are mg/L. This method, although still used, has been largely replaced by a more convenient measurement of the electrical resistance of the solution, or its reciprocal, the electrical conductance (EC). The current-carrying capacity of a solution is proportional to the concentration of ions in the solution. The electrical conductance is measured in reciprocal ohms (ohms1), also known as mhos. In SI units, mhos are designated as siemens (S). Measurements of conductivity are typically made in a cell containing two electrodes of defined geometry. An electric potential is imposed across the electrodes and the resistance of the solution between the electrodes is measured. The results are multiplied by a “cell constant” to cor-

CHEMISTRY OF SALT-AFFECTED SOILS

63

rect for variations in cell geometry, allowing all data to be reported as specific EC. The EC is commonly reported per unit volume (1 cm3) of solution in siemens (S) per cm, at 25 °C. Alternatively, EC can be measured using a four-electrode configuration, whereby a known potential is imposed on the two outer electrodes and the potential is measured between the two inner electrodes. This configuration has the advantage of being less sensitive to changes in the surface conductance of the electrodes, and is useful for measurements of soil conductivity in the field (see Chapter 10). The siemen is too large for measuring electrical conductance in most natural systems. Hence, the working unit is often the decisiemen (S  101). Because the basic length in the SI convention is the meter, the preferred dimension for EC is decisiemens per meter (dS/m). The units relationship is as follows: dS/m  mS/cm  mmhos/cm where mmhos/cm (millimhos per cm) are the traditional and now-obsolete dimensions for EC. Saturation Extract A variety of soil/water ratios can be used to obtain an aqueous extract from a soil sample. Therefore, a standard extraction method must be used if saline-soil chemical data are to be compared. Because the amount of water that a soil holds at saturation (saturation percentage, SP) is related to a number of soil parameters, such as texture, surface area, clay content, and cation exchange capacity (Merrill et al. 1987), one widely used technique is to obtain an extract by vacuum filtration of a saturated soil paste made with distilled or deionized water (U.S. Salinity Laboratory 1954). This extract is then analyzed for electrical conductivity (ECe) and soluble constituents of interest. Higher-water-content extracts (soil/water ratios of 1⬊1, 1⬊2, 1⬊5) are easier and faster to prepare, but their solution compositions are less related to those at field moisture condition than those of the saturation pastes and are generally not recommended. As increasing amounts of water are added, there is an increasing importance of mineral dissolution (primarily gypsum and carbonates), cation exchange, and anion desorption on solution composition and EC. However, when only relative salinity changes are of interest, and in the absence of gypsum, the lower soil/water ratios can be used to advantage. Correction of the composition from one water content to another can be made by use of the Extract Chem program (Suarez and Taber 2007). This computer model considers cation exchange, calcite and gypsum dissolution/precipitation, and boron (B) adsorption/desorption, as related to changes in water content. The value of EC is sensitive to the temperature of the solution. The change in EC with temperature depends on the mineral composition of the dissolved salts. However, the correction for natural waters is about a 1.9% increase in EC for each degree increase in solution temperature in

64

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the range of 15 °C to 35 °C. The EC data are usually normalized to a temperature of 25 °C for a meaningful comparison among samples. A temperature correction to 25 °C (EC25) can be approximated by EC25  ECt  0.019(t  25)ECt

(3-9)

where ECt  the value at temperature t. Temperature-compensated EC probes make a similar correction of EC. Predicting Electrical Conductivity from Ionic Composition Theoretical and empirical approaches have been used to predict the EC of a solution from its composition. An example of the theoretical approach is a model based on Kohlrauch’s Law of independent migration of ions (Harned and Owen 1958), in which each ion contributes to the current-carrying ability of an electrolyte solution. The equation used is EC  ∑ ECi  ∑ i

ci (i0   ci ) 1, 000

(3-10)

where EC  the specific conductance (dS/m); ECi  the ionic specific conductance (dS/m) summed over all ith species in solution; ci  the concentration of the ith ion (mmolc/L); 0i  the ionic equivalent conductance (cm2 S/molc) at infinite dilution; and  is an empirical interactive parameter obtained by plotting the ionic equivalent conductance of the ith ion (i 1/2 expressed in cm2 S mol1 . For relatively dilute mixtures, c ) versus (c i)  ranges in value from 2 to 9, with an average of 5.5 (Tanji and Biggar 1972). Values of i0 can be obtained from standard textbooks on electrochemistry and physical chemistry. A more accurate, but mathematically more complex, model based on the modified Onsanger-Fuoss equation and corrected for ion pair formation (which will be discussed later) gives reasonable agreement ( 8%) between calculated and measured EC up to 15 dS/m (Marion and Babcock 1976; Tanji 1969). Marion and Babcock (1976) also developed an empirical approach relating measured specific conductance (EC in dS/m) to total soluble salt concentration (TSS in mmolc/L) and ionic concentration (C in mmolc/L), where C is corrected for ion pairs. In the absence of ion complexation, TSS  C. The derived relationship was determined on a composite database obtained from soil extracts, river waters, and pure salt solutions. The equations, suitable for most purposes, are log C  0.955  1.039 log EC r2  0.997

(3-11)

log TSS  0.990  1.055 log EC r  0.993

(3-12)

CHEMISTRY OF SALT-AFFECTED SOILS

65

Equation 3-11 should be used instead of the empirical relationship (U.S. Salinity Laboratory 1954): TSS (mmolc/L)  10 EC (dS/m)

(3-13)

McNeal et al. (1970) developed a more detailed empirical relationship with improved predictive capability of EC (generally within 5%, listed as Method 3 in their publication), suitable for inclusion in spreadsheets or computer models. This calculation of EC from solution composition is used in the Extract Chem computer model and in the SWS model used for management of salt-affected soils (see Chapter 28).

CHEMISTRY OF SALT-AFFECTED SOIL SOLUTIONS Major cations in salt-affected soils are Na, Ca2, Mg2, and, to a lesser  extent, K. The major anions are Cl, SO42, HCO 3 , NO3 , and, at high pH, CO32. The ions HCO3 and CO2 are usually reported together as carbon3 ate alkalinity. Under high pH conditions (8.5), elevated concentrations of B can result in a significant contribution to alkalinity [dissociation constant for boric acid (pKa) is 9.2]. Other ions that are sometimes present under anaerobic conditions, but usually neglected from a salinity view point, include NH 4 , NO2 , and organic anions (such as acetate). When analyzing high-organic-content waters (such as waste waters), these anions of weak organic acids will also contribute to titratable alkalinity, which needs to be corrected if carbonate alkalinity is to be reported. Dissolution and precipitation of minerals often determine the composition of the soil solution. The degree of saturation of the soil solution with respect to a particular solid phase can be evaluated from the ion activities. The activity concept accounts for the nonideal behavior of ions in solution. Activities can be calculated from solution concentrations and ionic strength as follows: ai  i mi

(3-14)

where i is the activity coefficient and mi is expressed in molal units (moles/kg solvent) concentration of the ith ion. Activities and activity coefficients are dimensionless, since mi is actually the ratio mi/m0 where m0 is the standard state molal concentration (1.0). Except for very saline solutions, the assumption that mi  Mi (moles/L of solution) is reasonable. The activity coefficients are, in turn, related to the ionic strength, I, defined by I  0.5∑ mi zi2 i

(3-15)

66

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

where z  the valence of the ith ion. The summation in Eq. 3-15 is for all charged species in solution. An empirical relationship between EC and ionic strength suitable for most natural waters is given by I  0.0127 EC

(3-16)

where EC is in dS/m at 25 °C (Griffin and Jurinak 1973). The activity coefficients of individual ions are necessary for calculation of the mineral saturation status of a water. The most utilized method of calculating activity coefficients is by applying the Davies equation (Stumm and Morgan 1996): ⎡ ⎤ I log i  0.509 zi2 ⎢  0.2 I ⎥ ⎣ 1 I ⎦

(3-17)

Note that this equation uses the term 0.2I instead of 0.3I as originally proposed by Davies. This equation is reported as valid when I 0.5 and water at 25 °C is the solvent, but substantial errors occur above I  0.1; also, the equation does not consider differences in activity coefficients with ions of the same charge. A better estimate of activity coefficients can be obtained by a modified version of the extended Debye-Huckel equation (Truesdell and Jones 1974), ⎡ ⎤ I log i  A zi2 ⎢ ⎥  bi I ⎣ 1  Bai I ⎦

(3-18)

where A is equal to 0.509 and B is equal to 0.33 at 25 °C and ai and bi are ion-specific adjustable parameters fitted from mean salt calculations (Truesdell and Jones 1974). The coefficients based on Eq. 3-18 can be obtained from Fig. 3-1. Individual ion activities can then be estimated from concentration data using Eq. 3-14. This equation is used in WATEQ (Truesˇ unek 1997), and availdell and Jones 1974), UNSATCHEM (Suarez and Sim˚ able as an option in PHREEQC version 2 (Parkhurst and Apello 1999), among others. The WATEQ and PHREEQC models are available from the U.S. Geological Survey website and UNSATCHEM from the ARS-USDA Salinity Laboratory website. The equation is stated as valid to I  4.0, but this is based on NaCl as the background electrolyte; in mixed electrolyte solutions, the limit is around I  0.3–0.5 depending on the actual solution composition. At higher ionic strength it is recommended to use the Pitzer expressions (Pitzer 1979) with the Harvie et al. (1981) species constants. Equally as important as the specific model chosen for the absolute accuracy of the activity coefficient calculation is internal consistency in

CHEMISTRY OF SALT-AFFECTED SOILS

67

FIGURE 3-1. Calculation of activity coefficients as related to ionic strength using a extended version of the Debye-Huckel equation. From Butler (1964).

the database used for the calculations. Many thermodynamic constants have been derived from solubility experiments; thus, the derived solubility product (Ksp) values depend on the chemical speciation model used. For this reason the activity calculation model needs to use the same activity coefficient calculations and ion association model and constants as used to generate the database (Suarez 1998). Salt-tolerance data for crops usually are expressed as a function of osmotic potential, o, where o  the soil solution’s osmotic pressure. The osmotic pressure, in kPa, can be calculated from  o  2480

∑ mi  ii

(3-19)

i

where i  the osmotic coefficient of the ith salt and   the stochiometric number of ions yielded by the salt. The  value for each salt can be obtained by using the salt’s total concentration (Fig. 3-2). The approximate relationship between o and EC at 25 °C, is o (kpa) 艐  0.40 EC

(3-20)

where EC is in dS/m. The commonly used proportionality constant of 0.36 is for o data obtained at 0 °C (U.S. Salinity Laboratory 1954).

68

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 3-2. Osmotic coefficients of electrolytes at 25 °C. From Robinson and Stokes (1959).

Salt and pH Effect on Chemical Mass Action The general mass action equation for any reaction is aA  bB ↔ cC  dD

(3-21)

where the lowercase letters are the stochiometric coefficients and uppercase letters represent chemical symbols. In Eq. 3-21, the thermodynamic equilibrium constant for the general reaction is K eq 

(C)C (D)d ( A)a (B)b

(3-22)

where ( ) represents activities as defined in Eq. 3-14. This convention is generally used in soil chemical publications. In the chemical and geochemical literature, [ ] is used to denote activities; in this chapter [ ] denotes concentrations.

CHEMISTRY OF SALT-AFFECTED SOILS

69

The dissolution of carbon dioxide in water is represented by the reactions CO2  H2O ↔ H2CO3 ↔ H  HCO3

(3-23)

Biological production and diffusion (water content-dependent) control the CO2 concentration in soil air. The CO2 concentration of the soil solution is thus assumed to be independent of water quality or precipitation-dissolution reactions (i.e., an open system). Equation 3-23 shows that an increase in CO2 results in a production of H and, thus, a reduction in pH. Calcite (CaCO3) is a source of calcium commonly found in most arid soils. This mineral is important because much of the soil-solution’s chemistry in arid regions can be defined in the context of the CO2—CaCO3— H2O system in combination with cation exchange. The dissolution of calcite is represented by CaCO3  CO2  H2O ↔ Ca2  2HCO 3

(3-24)

An increase in aqueous CO2 concentration shifts the reaction to the right. This increases the Ca2 concentration and alkalinity of the solution. The presence of ions other than Ca2 or HCO3 increases the ionic strength, which, in turn, decreases the ionic activity coefficients of Ca2, 2 HCO 3 , and CO3 . This, known as the ionic strength effect, increases the solubility of CaCO3. Dolomite, CaMg(CO3)2, is sometimes present in arid soils, usually derived from carbonate geologic materials. Its solubility is very similar to that of calcite (expressed as a double carbonate mineral), but its dissolution rate is very slow and its precipitation negligible over time frames of agricultural interest. Soils with native dolomite almost always contain calcite. The dissolution reaction is CaMg(CO3)2  2CO2  2H2O → Ca2  Mg2  4HCO 3

(3-25)

We do not consider this reaction to be reversible as it does not form over time frames of interest to agriculture. Gypsum (CaSO4 2H2O) is another mineral that may be found naturally in arid soils. It is frequently added as a soil amendment as a calcium source to reclaim sodic soils and thus maintain or enhance soil permeability. It is moderately soluble and readily precipitated when its solubility is exceeded, such as when irrigating with a high-sulfate water. The dissolution reaction for gypsum is CaSO4 2H2O ↔ Ca2  SO42  2H2O

(3-26)

70

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Addition of salts that contain ions, such as Na, Mg2, Cl, and NO 3, enhances the solubility of gypsum by the ionic strength effect. Adding salts that contain either Ca2 or SO42 decreases gypsum solubility by the common ion effect. For example, gypsum solubility is greatly reduced in the presence of water containing large amounts of Na2SO4 or MgSO4 salt. At constant temperature, the solubility product, Ksp, of gypsum is invariant, that is, assuming the activities of H2O and solid-phase gypsum are unity, the product of the activities (Ca2) (SO42) is constant, even though the concentrations of the constituent ions in solution increase. Gypsum’s solubility is also increased by ion association (the formation of soluble ion pairs and complexes). The concentration of ion pairs can constitute an important percentage of the total ions in a saline solution. For example, in a system of pure gypsum dissolved in water, the concentration of free Ca2 ion is 10 mmol/L, but the total Ca concentration in solution is 15 mmol/L. The difference between the two concentrations is the presence of 5 mmol/L of the CaSO04 ion pair in the solution phase. Because the formation of CaSO04 requires both Ca2 and SO2 4 ions, gypsum solubility increases as ion association occurs. The formation of ion pairs is prevalent between multiple valence ions and, to a lesser extent, mono- and multivalent ions. It is relatively minor for monovalent–monovalent interactions. Ion association can best be considered as a way to fit nonideal behavior among ions rather than as actual physical entities with chemical significance. These always reduce the activity of the free ions and, hence, they enhance the solubility of minerals. These speciation corrections are needed to accurately estimate the soil-solution composition of salt-affected systems when the ion association model is used, but they are not utilized in the Pitzer formulations. Addition of neutral salts can nonetheless affect the soil-solution pH. For example, the addition of gypsum to a solution saturated with respect to calcite increases the Ca ion concentration, causing additional CaCO3 to precipitate, reducing the bulk solution alkalinity, and decreasing the solution pH. However, the solubility of gypsum is not pHdependent. Adding sulfuric acid to a soil suppresses gypsum solubility due to the common ion effect, rather than the effect of the acidity. However, the net result of adding gypsum or gypsum  sulfuric acid is similar in the presence of soil calcite. Conversely, Eq. 3-24 shows that CaCO3 solubility is pH-dependent. Addition of acid reduces the alkalinity of the solution and allows CaCO3 to dissolve until equilibrium is reestablished. Chemical Composition of Surface and Ground Waters The composition of surface and ground water varies greatly in arid regions. Table 3-1 shows data selected from saturation extracts of salt-

CHEMISTRY OF SALT-AFFECTED SOILS

71

TABLE 3-1. Chemical Characteristics of Saturation Extracts of Salt-Affected Soils, Well Waters, and River Waters Analyzed at U.S. Salinity Laboratory Number of Samples (1)

(a) mmolc/L Ca Sat. Ext. Well River Mg Sat. Ext. Well River Na Sat. Ext. Well River K Sat. Ext. Well River SO4 Sat. Ext. Well River Cl Sat. Ext. Well River HCO3 Sat. Ext. Well River (b) (dS/m) EC Sat. Ext. Well River

10% less than (4)

90% less than (5)

Mean (2)

Median (3)

139 115 68 139 115 61 139 115 58 128 101 30

27.8 5.9 4.7 22.2 4.3 3.6 93.2 15.8 7.5 1.6 0.6 0.3

10.6 3.1 3.4 8.0 1.5 2.2 53.5 6.6 3.7 0.5 0.1 0.2

1.4 0.8 1.0 1.1 0.5 0.6 1.8 0.8 0.7 0.1 0.01 0.06

71.8 14.5 10.9 79.1 15.3 9.2 219 44.8 18.9 2.2 0.9 0.5

134 23 58 139 115 58

400 6.7 6.7 95.5 15.0 5.9

29.4 3.6 4.1 34.8 2.5 1.5

3.7 0.4 0.3 1.4 0.2 0.2

94.1 15.4 19.0 281 54.2 20.7

139 115 58

8.3 4.9 3.3

3.0 4.1 3.0

1.2 1.9 1.7

10.3 8.2 5.7

134 115 58

12.9 2.4 1.4

8.8 1.2 0.92

1.1 0.3 0.3

33.7 7.3 3.2

139 115 58

38.5 6.6 3.3

0.9 0.6 0.7

56.1 16.5 7.4

(c) (mmol/L)0.5 SAR

Sat. Ext. Well River

15.1 4.7 2.4

EC, electrical conductance; SAR, sodium adsorption ratio; Sat. Ext., saturation extract

72

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

affected soils, well waters, and surface waters analyzed at the U.S. Salinity Laboratory. The data represent samples obtained throughout the world. The soil extracts were more saline and more sodic than the well waters and surface waters. This reflects the effects of ET and mineral dissolution in the soil. With increasing salinity, Na predominates over Ca because Na salts are more soluble than Ca salts. The importance of Na is also reflected in the increasing sodium adsorption ratio (SAR). These data also show that Cl is the dominant anion in saline waters, whereas SO4 and Cl are prevalent in dilute solutions. This reflects the influence of limited solubility of gypsum and precipitation as the waters are concentrated by ET. The increase in the Mg/Ca ratio reflects the fact that Mg salts are more soluble than Ca salts. In general, the ratio Ca/HCO3 is also 1.0 (when concentrations are expressed in mmolc/L). A Ca/HCO3 ratio of less than 1.0 poses a special sodicity hazard because, when such waters are concentrated by ET, calcite precipitates, the Ca concentration decreases, and the HCO3 increases. The smaller the Ca/HCO3 ratio, the greater the sodicity hazard. Surface Chemistry Clay minerals Incongruent weathering processes in soils result in the production of soluble salts and, more importantly, the formation of secondary aluminosilicate clay minerals. These secondary clays are what give soils their ionexchange and adsorptive properties. Because weathering processes differ in response to changing environmental conditions, and because weathering products from one reaction may simultaneously participate as reactants in other, different reactions, a given soil often contains a wide range of clay minerals. For the same reasons, a given clay type typically exhibits a wide range of chemical composition and physical properties. With their plate-like shape and small particle size, clay minerals exhibit large specific surface areas (m2/kg). This, coupled with their permanent charge and their pH-dependent charge, causes the colloidal clay fraction to be the center of chemical activity in the soil. The permanent charge is the result of structural substitution of cations of lesser charge (primarily Al3 for Si4), resulting in net negative charge at the surface. The pHdependent charge (broken bonds at the surface) is most important on the mineral edges. In addition, the colloidal hydrous oxides of iron and aluminium (the stable end-products of weathering) enhance the adsorptive and exchange capacity of soils. Table 3-2 gives the qualitative unit cell formulae, the range of the cation exchange capacities (CEC), and specific surface areas for three clay minerals commonly found in arid soils. Textbooks on soil mineralogy can be consulted for more information on crystal structure and properties (Dixon and Weed 1989).

CHEMISTRY OF SALT-AFFECTED SOILS

73

TABLE 3-2. Properties of Common Soil Clays Clay (1)

Formula (General) (2)

Surface Area (104m2/kg) (3)

CEC (cmolc/kg) (4)

Smectite (montmorillonite)

Si8(Al,Mg)4O20 (OH)4

60–80

80–120

Mica (illite)

K(Al,Si)8(Al,Fe,Mg)4O20(OH)4

10–20

20–40

Kaolin (kaolinite)

SiAl4O10(OH)8

1–2

3–6

CEC, cation exchange capacity

Although inorganic aspects of the soil matrix are emphasized here, the colloidal (0.2 μm) organic matter fraction of a soil cannot be overlooked. In some soils, the effect of organic matter is more important than the inorganic-fraction processes. The organic matter has exchange properties, with extreme affinity for Ca (affecting the overall Ca–Na selectivity of soils) as well as variable charge, allowing for buffering of solution pH. The diffuse double layer The diffuse double layer (DDL) model developed from basic electrostatic theory describes electrochemical phenomena at the charged solid–liquid interface. The charge associated with a given surface is viewed as a layer, with an adjacent diffuse layer of oppositely charged ions, known as counter ions, in solution. The layer of counter ions maintains a charge in solution that is equal and opposite to the net charge of the mineral-surface layer. Various forces influence the diffuse nature of the counter ions: electrostatic attraction, which draws them toward the charged surface, repulsion among the ions and thermal energy, which tends to equalize the concentration of ions in the system and draw the counter ions back into the bulk solution, as well as close-range van der Waals forces. Applying the DDL model to the clay-mineral surface allows one to predict the distribution of cations (counter ions) and anions (co-ions) at the charged mineral interface. It is an alternate approach to the mass action model (discussed later) for describing the exchange phenomena in soils. Many physical properties of soils can be modeled as an interaction between DDLs of soil clay particles. The degree and nature of interaction is determined by the effective thickness of the DDL, which can be estimated with the K parameter in units of cm1: K

8 e 2 z 2 n0 D ◊k ◊T

(3-27)

74

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

where e  the electron charge (coulomb/ion); z  the valence of the counter ion, n0 is the electrolyte concentration in the bulk solution (ion/cm3); D  the dielectric constant (coulomb/V/cm); k  the Boltzmann constant (V coulomb/K/ion); and T  absolute temperature (K). The effective thickness of the DDL is 1/K, which has units of cm. Compression of the DDL (the desired condition for soil stability) is promoted by (1) increasing the valence of the counter ion, (2) increasing the concentration of the bulk solution, or (3) reducing the dielectric constant of the medium. The first two factors can be readily altered by processes, such as application of a calcium amendment or leaching. Leaching is necessary for salinity control; however, a reduction in salt concentration expands the DDL and thus reduces aggregate stability. Addition of a calcium source (such as gypsum) increases both the concentration of the bulk solution and provides addition of a divalent ion, both of which compress the DDL. Compression of the DDL allows for increased soil stability because, at specified distances, there is a reduction in the repulsive forces between clay particles, thus enabling closer approach of individual clay particles, resulting in aggregation. Aggregation of soil particles results in beneficial soil properties, including development of larger pores, thus enhancing permeability, as well as improving soil tilth. Several excellent references discuss at length the DDL theory and applications (Bolt 1979; Bresler et al. 1982; Singh and Uehara 1998). Anion exclusion The DDL model has successfully predicted anion exclusion, the negative adsorption of ions from the clay–water interface. Anions excluded from the double layer must be associated with cations to maintain solution electroneutrality. This increases the apparent concentration of soluble cations in solution. Figure 3-3 shows how the concentration of counter ions (n) and anions (n) vary with distance in the DDL of a negatively charged particle. Assuming a 1⬊1 symmetrical salt, e.g., NaCl, in the bulk solution (infinite distance from the surface), the concentration of the electrolyte n0 equals n or n. This is represented by the dashed line in the figure. The excess of cations ( ), which neutralizes both the surface charge and the anions in the DDL, is given by area ABCDEF. The amount of anions ( ) in the DDL is given by area AEF. Total surface charge ( tot), equivalent to CEC, is given by CEC  tot    –

(3-28)

CEC  surface cation excess  anion exclusion

(3-29)

CHEMISTRY OF SALT-AFFECTED SOILS

75

FIGURE 3-3. Charge distribution in diffuse double layer (DDL) of a negatively charged particle.

The CEC is equal to the area ABCDEF  area AEF, which equals the area ABCDE, where anion exclusion is measured by the increase of anion concentration in the contacting solution. The CEC is equal to the surface cation excess (total exchangeable cations) only when anion exclusion is negligible. This occurs at low electrolyte concentrations. Increasing electrolyte concentration compresses the DDL and increases anion exclusion. The correction for anion exclusion becomes significant above approximately 10 dS/m at 25 °C. Bower and Hatcher (1962) reported the effect of the anion exclusion correction on exchangeable sodium percentage, ESP, where ESP  NaX 100/CEC. The units of NaX and CEC are molc/kg). Table 3-3 shows selected data from their study. To correct for anion exclusion, they determined the Cl concentration in the saturation extract. The Cl concentration in the whole soil was determined by extensive leaching. The soluble cation content of the saturation extract was then multiplied by the ratio, Clleaching/Clextract.

76

TABLE 3-3. Anion Exclusion Correction on ESP Chloridea

Soluble Sodiuma

Exchangeable Sodiuma

ESP (Percentage)

Soil (1)

SP (2)

ECe (dS/m) (3)

Saturation Extracts (4)

Corrected (5)

Uncorrected (6)

Corrected (7)

Uncorrected (8)

Corrected (9)

Uncorrected (10)

Corrected (11)

1 2 3

101 60 57

9.0 67.4 32.3

0.078 0.406 0.074

0.068 0.368 0.061

0.071 0.464 0.165

0.062 0.420 0.135

0.076 0.184 0.027

0.085 0.228 0.057

21 60 13

23 75 27

4

59

17.7

0.033

0.026

0.076

0.059

0.023

0.040

12

21

a

Chloride, soluble sodium, and exchangeable sodium are given in molc/kg.

ESP, exchangeable sodium percentage From Bower and Hatcher (1962).

CHEMISTRY OF SALT-AFFECTED SOILS

77

Table 3-3 indicates a marked increase in Na ion preference by the exchanger, as seen by the ESP, after correction. This underscores the importance of considering anion exclusion in high-salinity environments. Anion exclusion fostered the concept known as the apparent exclusion volume, the hypothetical volume of water near the surface that must be insoluble to anions to account for the bulk solution’s anion concentration. When calculating anion exclusion, the mathematically simpler concept of apparent exclusion volume is often used. Ion Exchange The ability to predict the distribution of ions between the soil solution and the exchanger phase is vital to the management of salt-affected soils, particularly when dealing with a potential sodic hazard. The complexity of the clay mineral fraction and the multi-ion nature of the soil solution make a rigorous definition of ion distribution in the soil difficult. A common approach to ion exchange is to apply the mass action principle. Mass action The general mass action equation for describing ion exchange in a binary cation system is bAXa  aBAb ↔ aBXb  bAa

(3-30)

where X  one mole of negative charge on the exchanger; a  the valence of cation A; and b  the valence of cation B. Ion concentrations are in mol/L. When a  b, the exchange is homovalent. When a  b, the exchange is heterovalent. Applying the general exchange equation to Ca/Na exchange, where a  2 and b  1, gives CaX2  2Na ↔ 2NaX  Ca2

(3-31)

The thermodynamic equilibrium constant is K eq 

(NaX )2 (Ca2 ) (CaX2 )(Na )2

(3-32)

where ( ) signifies activities in both solution and exchanger phase. The method for calculating activities of ions in solution is well established (Stumm and Morgan 1996), but determining the activities of ions in the exchanger phase is difficult and subject to question. A simple mass action approach, which gives the selectivity coefficient, can be used to calculate the ion distribution between the solution and adsorbed phase.

78

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

For Ca/Na exchange (Eq. 3-31), the selectivity coefficient ks is ks 

[NaX ]2 [Ca2 ] [CaX ][Na+ ]2

(3-33)

where [ ] signifies concentrations units in both the solution and exchanger phases. The ion concentration in the exchanger phase is expressed in units of equivalent fraction, using the Gapon convention, or, less frequently, the mole fraction, using the Vanselow convention. These terms are related, but they produce different values for the selectivity coefficient (Sposito 1977). Although not a true thermodynamic constant, the variability of ks is often small. Using the equivalent fraction, Ei, for the adsorbed ion concentration and mol/L for the solution concentration, Eq. 3-33 becomes ks 

[ENa ]2 [Ca2 ] [ECa ][Na+ ]2

(3-34)

where Ei  iX/CEC and both iX and CEC have the units of cmolc/kg (cmolc/kg is equivalent to the formerly popular units of meq/100 g). Levy and Hillel (1968) found the ks value for this exchange reaction to be constant over a wide range of ENa values (i.e., ENa 艐 0.1 to 0.7) for typical Israeli soils. The magnitude of the ks, however, varied between soils. The Gapon constant The Gapon exchange reaction has been widely used in salinity and sodicity studies. The original expression for Ca/Na exchange is Ca1/2X  Na ↔ NaX  1/2 Ca2

(3-35)

where solution concentration is in mol/L and exchange ion concentration is in cmolc/kg. At equilibrium, we can write kg 

[NaX ][Ca2 ]1/2 [Ca1/2 X ][Na+ ]

(3-36)

where kg  the Gapon selectivity coefficient, often assumed to be constant. The ratio of adsorbed ions is [NaX ] [Na ]  kg [Ca1/2 X ] [Ca2 ]1/2

(3-37)

CHEMISTRY OF SALT-AFFECTED SOILS

79

where the units of kg are (mol/L)1/2. The U.S. Salinity Laboratory (1954) further assumed that Mg behaves similarly to Ca in the adsorbed phase and modified the Gapon equation to the following: [NaX ] [Na ]  k g′  k g′ SAR [CaX  MgX ] [Ca  Mg]1/2

(3-38)

where k⬘g  the modified Gapon selectivity coefficient (mmol/L)1/2; and SAR  the sodium adsorption ratio, defined as SAR 

[Na] [Ca  Mg]1/2

(3-39)

where total analytical concentrations are used (mmol/L) with no account of ion association. Because Ca, Mg, and Na are the most common exchangeable cations in arid soils, Eq. 3-38 may be simplified to [NaX ]  k g′ SAR  ESR CEC  [NaX ]

(3-40)

where ESR is called the exchangeable sodium ratio. In terms of the ESP, Eq. 3-40 becomes [ESP]  k g′ SAR  ESR [100  ESP]

(3-41)

where ESP  NaX 100/CEC. Because obtaining reliable exchangeable ion data is difficult (Amrhein and Suarez 1990), the SAR of the soil solution or extract has become the principle parameter for diagnosing sodic hazards in soils. The value of k⬘g is determined by the regression (slope) of the ESR–SAR relationship, which is linear. A value of k⬘g  0.015 (mmol/L)1/2 is widely used. It is useful in field studies when the ESR 30. Correspondingly, as a rough approximation in this sodicity range, the value of SAR and ESP are assumed equal (U.S. Salinity Laboratory 1954). More recent studies, however, indicate that numerous factors influence the ESR–SAR couple and the value of k⬘g ranges from 0.016 (mmol/L)1/2 to 0.008 (mmol/L)1/2. This suggests that, to be accurate, k⬘g should be estimated using data specific to the site under study (Doering and Willis 1980; Jurinak et al. 1984).

80

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Effect of salinity on ion exchange If the soil solution is concentrated by a factor C, maintaining the same ionic ratios, the initial SAR increased by the factor (C)1/2: SARfinal  (C)1/2 SARinitial

(3-42)

Thus, if the concentration is increased by a factor of 3 and no chemical precipitation occurs, SARfinal  (3)1/2 SARinitial. Increasing soil salinity increases the SAR, and if k⬘g remains constant, then ESR and ESP must increase. There is little evidence for a change in selectivity with increasing salinity (Amrhein and Suarez 1991), hence no practical need to consider solution activity coefficients and ion pairs for cation exchange calculations. As an example, assume a soil is irrigated with water of EC  1.5 dS/m and SAR  8 (mmol/L)1/2. Assuming an ET concentration factor of 3 in the rootzone, and using Eq. 3-42, the SAR of the soil solution will then be 13.9 (mmol/L)1/2. Applying the Gapon convention and assuming that k⬘g is constant and equal to 0.015 (mmol/L)1/2, the steady-state ESR of the soil is 0.21 (ESP  17.4). If ET increases the concentration factor to 6, the SAR will then equal 19.6 (mmol/L)1/2, and the ESR increases to 0.29 (ESP  22.5). Thus, if the selectivity coefficient remains the same, an increase in the concentration of the equilibrium solution increases the affinity of the clay fraction for monovalent cations, and dilution increases the affinity for divalent cations. Ion demixing Early theories of cation exchange assumed exchangeable ions were randomly distributed at the clay surface. More recent evidence suggests, however, that random distribution is rare for expanding lattice clays. Instead, divalent ions are preferentially adsorbed in the interlayer spaces and monovalent ions are adsorbed on the edges and planar surfaces. This phenomenon, referred to as ion demixing, affects a soil’s response to relatively low amounts of exchangeable sodium (see Chapter 5). Rhizosphere Chemistry in Salt-Affected Soils The composition of soil water depends on the composition of irrigation and rain water; chemical reactions, such as dissolution, precipitation, adsorption, and exchange; and the extent to which ET concentrates the water. Predicting the chemistry of a soil solution is difficult due to the dynamic nature of the system. For a more convenient prediction we often assume (though it may be far from true) that steady-state conditions exist.

CHEMISTRY OF SALT-AFFECTED SOILS

81

When assuming steady-state conditions, ion exchange is viewed as a static condition affected only by ET and precipitation-dissolution reactions. It is also assumed that Na, Mg, K, and Cl do not undergo chemical reactions (other than exchange). Concentrations of these ions can be estimated by multiplying their initial concentrations in the applied water, Ciw, by the ratio of the water consumed to water applied. Magnesium carbonates In the rhizosphere, many soil waters are supersaturated with respect to dolomite, CaMg(CO3)2, and possibly magnesite, MgCO3. Dolomite is not considered a sink for Mg because of the kinetic constraints on precipitation existing in rhizosphere environments. Although dolomite has approximately the same molar solubility as calcite, the rate at which it dissolves is about 70 times slower. The result is that dolomite contributes Ca to the soil solution but, in the absence of calcite, is extremely slow to reach equilibrium with a soil solution. Because dolomite does not precipitate readily from the soil solution, whereas calcite does, the Mg/Ca ratio increases as the leaching fraction is reduced. Magnesite precipitation in soils has not been documented. The solubility of magnesite is in dispute, but pure solutions may attain (Mg2) values of about 60 to 180 times greater (at pH 9.5 and 8.5, respectively, depending also on the alkalinity) than the (Ca2) values before another magnesium carbonate, hydromagnesite (Mg5(CO3)4(OH)2 4H2O) precipitates. Before either dolomite or Mg-carbonate minerals form, Mg will most likely form a Mg-silicate mineral, such as sepiolite, MgSi3O6(OH). Co-precipitation with Ca is also another sink for Mg because, depending on the Ca/Mg ratio, the precipitated calcite will often contain 3% to 7% Mg substitution for Ca. Calcium carbonate The dissolution and precipitation of CaCO3 influences rhizosphere chemistry. Indeed, calcite reaction is generally assumed to dominate the chemistry of the soil solution in arid-zone soils. Studies show that calcite is usually supersaturated in waters in and below the rootzone, and the mean value of the ion activity product (IAP) of (Ca2) (CO32) equals 108 (Suarez 1977; Suarez et al. 1992), that is, (Ca2)(CO32)  108.0

(3-43)

This measured IAP does not imply that soil calcite is thermodynamically unstable (Ksp of calcite is 108.47 at 25 °C). Rather, it appears that this supersaturation reflects a kinetic aspect associated with calcite crystal

82

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

growth. At an IAP value of 108.0 calcite is sufficiently supersaturated that heterogeneous nucleation occurs. Once calcite nucleates in the soil, calcite crystal growth is inhibited by dissolved organic carbon in the soil water and the value of 108.0 corresponds to the value at which calcite precipitates via heterogeneous nucleation (Lebron and Suarez 1996; 1998). Soil solutions remain in a metastable condition at this IAP value (threefold supersaturated with respect to calcite), so it can be used as an apparent solubility constant for predictive purposes. Since CO32 is usually a minor species in solution, calcite solubility can more conveniently be represented by expressing it in terms of HCO3 :

K ′′ 

(Ca2 )(HCO3 )2 PCO2

(3-44)

When PCO2 is expressed in kilopascals (1 atm  101 kPa) and a calcite thermodynamic Ksp of 108.47 is used, the theoretical value of K⬙  108.0. If the apparent solubility product of calcite (Eq. 3-43) is used in Eq. 3-44 for predictive purposes, then the value of K⬙ increases to 107.5. In that situation, Eq. 3-45 loses its thermodynamic significance. Computer programs, such as Extract Chem (Chapter 26) are available for calculating equilibrium Ca2 and HCO3 concentrations and activities. A graphical method, based on solution of the following equation, is also available (Suarez 1982):

(Ca2  x)(HCO3  2 x)2 

107.5 PCO2  Ca2  2HCO3

(3-45)

where x  the amount of Ca that is precipitated or dissolved. No correction for ion association is made, so the results are only approximate. In the absence of specific data for the rhizosphere, use PCO2  1 kPa (0.01 atm) for sandy soils and 5 kPa (0.05 atm) for clay soils. The adjusted sodium adsorption ratio As discussed earlier, the SAR relationship is useful for estimating the ESP of the soil. The SAR of soil water, assuming steady-state conditions, is determined from the composition of the irrigation water, after correcting for ET and CaCO3 precipitation or dissolution. An adjustment to the SAR of irrigation water is often made to incorporate the changes in Ca that will occur when the water is equilibrated with

CHEMISTRY OF SALT-AFFECTED SOILS

83

calcite. Various methods are used to adjust the SAR of waters, several of which are in error. The preferred method is to calculate the equilibrium Ca concentration in the water (Caeq in mmol/L) and use in the following equation (Suarez 1981): SAR adj 

Naiw Mg iw  Caeq )

(3-46)

where Na iw and Mg iw are concentrations of Na and Mg (mmol/L), respectively, in the applied water. The Caeq is the Ca value in equilibrium with calcium carbonate, and can best be obtained by use of chemical speciation models, such as Extract Chem. This correction will result in a downward adjustment of the SAR if the initial irrigation water is undersaturated with respect to calcite. More commonly, surface irrigation waters in arid lands are already reacted with carbonates and a (Ca2)(CO32) IAP at or greater than 108.0. In this instance an adjustment to SAR is not needed. The SAR correction is most important when using groundwaters or waste waters for irrigation. In this instance the waters are generally equilibrated at higher concentrations of CO2 and, upon exposure to the air, will degas CO2 and the solution will precipitate calcite, thus increasing the SAR. An alternate equation for calculating SARadj is SAR adj 

Naiw Mg iw  0.215X( PCO2 )1/3

(3-47)

where PCO2 is in kPa and X values are found in Table 3-4 (Suarez 1981) using the molar HCO3/Ca ratio and ionic strength values for the irrigation water. The PCO2 for surface waters can be set to around twice atmospheric (i.e., 103.14 atm or 0.072 kPa). The same concepts can be used to estimate the equilibrium SAR in the rhizosphere. In this instance Eq. 3-47 is modified to SAR adj 

Naiw Fc Mg iw Fc  0.215X( PCO2 )1/3

(3-48)

The concentration factor Fc equals 1/L, in which L  the leaching fraction. Its value depends on the soil depth.

84

TABLE 3-4. X-Values for Various HCO3/Ca Mole Ratios and Ionic Strengths Ionic Strength (I)a HCO3/Ca (1) 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 6.0 7.0 8.0 9.0 10.0 a

0.001 (2)

0.005 (3)

0.01 (4)

0.02 (5)

0.03 (6)

0.04 (7)

73.4 46.2 35.3 29.1 25.1 22.2 20.1 18.3 17.0 15.8 12.1 9.96 8.58 7.60 6.86 6.27 5.80 5.41 4.79 4.31 3.95 3.65 3.41

79.6 50.1 38.2 31.6 27.2 24.1 21.7 19.9 18.4 17.1 13.1 10.8 9.31 8.24 7.44 6.80 6.29 5.86 5.19 4.68 4.29 3.96 3.69

84.1 53.0 40.4 33.4 28.8 25.5 23.0 21.0 19.4 18.1 13.8 11.4 9.84 8.71 7.86 7.19 6.65 6.20 5.49 4.95 4.53 4.19 3.90

90.0 56.7 43.3 35.7 30.8 27.3 24.6 22.5 20.8 19.4 14.8 12.2 10.5 9.33 8.42 7.70 7.12 6.63 5.87 5.30 4.85 4.48 4.18

94.4 59.5 45.4 37.5 32.3 28.6 25.8 23.6 21.8 20.3 15.5 12.8 11.0 9.78 8.82 8.07 7.46 6.96 6.16 5.56 5.09 4.70 4.38

97.9 101.0 106.0 112.0 120.0 125.0 130.0 133.0 139.0 144.0 61.7 63.6 66.8 70.5 75.3 78.8 81.7 84.0 87.9 90.4 47.1 48.5 51.0 53.8 57.5 60.1 62.3 64.1 67.0 69.0 38.9 40.1 42.1 44.4 47.4 49.6 51.5 52.9 55.3 57.0 33.5 34.5 36.3 38.3 40.9 42.8 44.3 45.6 47.6 49.1 29.7 30.6 32.1 33.9 36.2 37.9 39.3 40.4 42.2 43.5 26.8 27.6 29.0 30.6 32.7 34.2 35.4 36.4 38.1 39.2 24.5 25.2 26.5 28.0 29.9 31.3 32.4 33.3 34.8 35.9 22.6 23.3 24.5 25.9 27.6 28.9 30.0 30.8 32.2 33.2 21.1 21.8 22.8 24.1 25.8 26.9 27.9 28.9 30.0 30.9 16.1 16.6 17.4 18.4 19.7 20.6 21.3 21.9 22.9 23.6 13.3 13.7 14.4 15.2 16.2 17.0 17.6 18.1 18.9 19.5 11.5 11.8 12.4 13.1 14.0 14.6 15.2 15.6 16.3 16.8 10.1 10.5 11.0 11.6 12.4 13.0 13.4 13.8 14.4 14.9 9.15 9.44 9.91 10.5 11.2 11.7 21.1 12.5 13.0 13.4 8.37 8.63 9.06 9.57 10.2 10.7 11.1 11.4 11.9 12.3 7.74 7.98 8.38 8.85 9.45 9.88 10.2 10.2 11.0 11.3 7.22 7.44 7.81 8.25 8.81 9.21 9.55 9.83 10.3 10.6 6.39 6.59 6.92 7.30 7.80 8.16 8.46 8.70 9.09 9.37 5.77 5.95 6.24 6.59 7.04 7.36 7.63 7.85 8.20 8.45 5.28 5.44 5.71 6.03 6.44 6.74 6.98 7.18 7.50 7.73 4.88 5.03 5.28 5.57 5.95 6.23 6.46 6.64 6.94 7.15 4.55 4.69 4.92 5.20 5.55 5.80 6.02 6.19 6.46 6.66

I  0.0127 EC (mol/L)

From Suarez (1981).

0.05 (8)

0.07 (9)

0.1 (10)

0.15 (11)

0.2 (12)

0.25 (13)

0.3 (14)

0.4 (15)

0.5 (16)

CHEMISTRY OF SALT-AFFECTED SOILS

85

SUMMARY The salt-mineral composition of the soil solution results from numerous, interdependent, multiphase chemical interactions, including dissolution and precipitation of minerals, formation of inorganic and organic coordination compounds in solution, adsorption of ions and ligands onto the surfaces of minerals and organic compounds, exchange of ions between clay-mineral surfaces and the solution phase, transport of compounds between the gas and liquid phases, and reduction and oxidation of ions and minerals. This list suggests not only the complexity of the system, but also the need for conceptual geochemical computer models to predict the composition of the soil solution as affected by man’s activities. Geochemical models can be linked with soil-water flow models, plant growth models, and other models to predict solute migration in soils. Numerous geochemical models are available. An abbreviated list includes PHREEQC (Parkhurst and Appelo 1999) MINEQL (Westall et al. 1976), EQ3/EQ6 (Wolery 1979; 1983) and MINTEQ (Allison et al. 1990; HydroGeologic Inc. and Allison Geoscience Consultants, Inc. 1999). They represent the current state-of-the-art in computer simulation of soil-solution ˇ unek 1997) incorpochemistry. The UNSATCHEM model (Suarez and Sim˚ rates a variably saturated water flow model with major ion chemistry, B adsorption, plant water uptake, and yield predictions related to salinity. REFERENCES Allison, J. D., Brown, D. S., and Novo-Gradac, K. J. (1990). MINTEQA2/ PRODEFA2: A geochemical assessment model for environmental systems: Version 3.0 user’s manual, Environmental Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Athens, Ga. Amhrein, C., and Suarez, D. L. (1990). “A procedure for determining sodiumcalcium exchange selectivity in calcareous and gypsiferous soils.” Soil Sci. Soc. Am. J., 54, 999–1007. ———. (1991). “Sodium-calcium exchange with anion exclusion and weathering corrections.” Soil Sci. Soc. Am., J. 55, 698–706. Bolt, G. H. (1979). Soil chemistry, B: Physico-chemical models, Elsevier Scientific Publishing Co., New York. Bower, C. A., and Hatcher, J. T. (1962). “Characterization of salt-affected soils with respect to sodium.” Soil Sci., 93, 275–280. Bresler, E., McNeal, B. L., and Carter, D. L. (1982). Saline and sodic soils, SpringerVerlag, New York. Butler, J. N. (1964). Ionic equilibrium, a mathematical approach, Addison-Wesley, Boston. Dixon, J. B., and Weed, S. B. (eds.). (1989). Minerals in soil environments, 2nd ed., Soil Science Society of America Book Series No. 1, Soil Science Society of America, Madison, Wisc.

86

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Doering, E. J., and Willis, W. O. (1980). “Effect of soil solution concentration on cation exchange relationships.” Proc., International Symposium on Salt-Affected Soils, Karnal, India. Griffin, R. A., and Jurinak, J. J. (1973). “Estimation of activity coefficients from the electrical conductivity of natural aquatic systems and soil extracts.” Soil Sci., 116, 26–30. Harned, H. S., and Owen, B. B. (1958). The physical chemistry of electrolyte solutions, Reinhold Publishers, New York. Harvie, C. E., Moller, N., and Weare, J. H. (1981). “The prediction of mineral solubilities in natural waters: The Na-K-Mg-Ca-H-Cl-SO4-OH-HCO3-CO3-H2O system to high ionic strength.” Geochim. Cosmochim. Acta, 48, 723–751. HydroGeoLogic, Inc. and Allison Geoscience Consultants, Inc. (1999). MINTEQA2/ PRODEFA2: A geochemical assessment model for environmental systems, Version 4.0 user’s manual, Environmental Research Laboratory, Office of Research and Development. U.S. Environmental Protection Agency, Athens Ga., www.epa. gov/ceampubl/mmedia/minteq/, accessed January 15, 2011. Jurinak, J. J., Amrhein, C., and Wagenet, R. J. (1984). “Sodic hazard: The effect of SAR and salinity in soils and overburden materials.” Soil Science, 137, 152–158. Lebron, I., and Suarez, D. L. (1996). “Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25 °C and pH 7.5.” Geochim. Cosmochim. Acta, 60, 2765–2776. ———. (1998). “Kinetics and mechanisms of precipitation of calcite as affected by PCO2 and organic ligands at 25 °C. Geochim. Cosmochim. Acta, 62, 405–416. Levy, R., and Hillel, D. (1968). “Thermodynamic equilibrium constants of sodiumcalcium exchange in some Israel soils.” Soil Sci., 106, 393–398. Marion, G. M., and Babcock, K. L. (1976). “Predicting specific conductance and salt concentration of dilute aqueous solution.” Soil Sci., 122, 181–187. McNeal, B. L., Oster, J. D, and Hatcher, J. T. (1970). “Calculation of electrical conductivity from solution composition data as an aid to in-situ estimation of soil salinity.” Soil Sci., 110, 405–414. Merrill, S. D., Deutsch, J. R., and Pole, M. W. (1987). “Saturation percentage,” in Reclaiming mine soils and overburden in the western United States, D. R. Williams and G. E. Schuman, eds., Soil Conservation Society America, Ankeny, Iowa. Parkhurst, D. L. and Appelo, C. A. (1999). User’s guide to PHREEQC (ver 2): A computer program for speciation, batch reaction, one dimensional transport, and inverse geochemical calculations. U.S. Geol. Survey Water Resources Investigations Report 99-4259. U.S. Geological Survey, Denver, Colo. Pitzer, K. S. (1979). Activity coefficients in electrolyte solutions, Chapter 7, CRC Press, Boca Raton, Fla. Robinson, R. A., and Stokes, J. J. (1959). Electrolyte solutions, 2nd ed., Butterworth and Co., London. Singh, U., and Uehara, G. (1998). “Electrochemistry of the double layer,” in Soil physical chemistry, 2nd ed., D. Sparks, ed., Chapter 1, CRC Press, Boca Raton, Fla. Sposito, G. (1977). “The Gapon and Vanselow selectivity coefficients.” Soil Sci. Soc. Am. J,. 41, 1205–1206.

CHEMISTRY OF SALT-AFFECTED SOILS

87

Stumm, W., and Morgan, J. J. (1996). Aquatic chemistry, 3rd ed., John Wiley and Sons, New York. Suarez, D. L. (1977). “Ion activity products of calcium carbonate in waters below the rootzone.” Soil Sci. Soc. Am. J., 41, 310–315. ———. (1981). “Relation between pHc and sodium adsorption ratio (SAR) and an alternative method of estimating SAR of soil or drainage waters.” Soil Sci. Soc. Am. J., 45, 469–475. ———. (1982). “Graphical calculation of ion concentrations in CaCO3 and/or gypsum soil solutions.” J. Environ. Qual., 11, 302–308. ———. (1998). “Thermodynamics of the soil solution,” in Soil physical chemistry, 2nd ed., D. Sparks, ed., Chapter 3, CRC Press, Boca Raton, Fla. ˇ unek. (1997). “UNSATCHEM: Unsaturated water and Suarez, D.L. and J. Sim˚ solute transport model with equilibrium and kinetic chemistry.” Soil Sci. Soc. Am. J., 61, 1633–1646. Suarez, D. L., and Taber, P. (2007). “Extract Chem: Numerical software package for estimating changes in solution composition due to changes in soil water content,” http://ars.usda.gov/Services/docs.htm?docid14567, accessed January 15, 2011. Suarez, D. L., and Wood, J. D. (1996). “Short and long term weathering rates of a feldspar fraction isolated from an arid zone soil.” Chem. Geol., 132, 143–150. Suarez, D. L., Wood, J. D., and Ibrahim, I. (1992). “Reevaluation of calcite supersaturation in soils.” Soil Sci. Soc. Am. J., 56, 1776–1784. Tanji, K. K. (1969). “Predicting specific conductance from electrolytic properties and ion association in some aqueous solutions.” Soil Sci. Soc. Am. Proc., 33, 887–889. Tanji, K. K., and Biggar, J. W. (1972). “Specific conductance models for natural waters and soil solutions of limited salinity levels.” Water Resource Res., 8, 145–153. Truesdell, A. H., and Jones, B. F. (1974). “WATEQ, a computer program for the calculating chemical equilibria of natural waters.” J. Res. U.S. Geol. Surv., 2, 233–248. U.S. Salinity Laboratory. (1954). Diagnosis and improvement of saline and alkali soils, U.S. Department of Agriculture Handbook No. 60, U.S. Government Printing Office, Washington, D.C. Westall, J. C., Zachary, J. L., and Morel, F. M. (1976). MINEQL: A computer program for calculation of chemical equilibrium composition of aqueous systems, Tech Note 18, Dept. Civil Engineering, Massachusetts Institute of Technology, Cambridge, Mass. Williams, J. S. (1975). The natural salinity of the Colorado River, Occasional Paper 7, Utah Water Research Laboratory, Logan, Utah. Wolery, T. J. (1979). Calculation of chemical equilibrium between aqueous solution and minerals: The EQ3/6 software package, UCRL-52658, Lawrence Livermore Lab., University of California, Livermore, Calif. ———. (1983). EQ3NR: A computer program for geochemical aqueous speciation-solubility calculations, Lawrence Livermore Lab., University of California, Livermore, Calif.

88

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

NOTATION A, B, C a a, b, c b ci D e EC ECi ECt ESR ET Fc K k kg k g L Mgiw, mi Naiw n0 SAR T TSS X x z  i oi  () []

 chemical symbols  valence of cation A  stoichiometric coefficients  valence of cation B  concentration of the ith ion  dielectric constant  electron charge  specific conductance  ionic specific conductance  value at temperature t  exchangeable sodium ratio  evapotranspiration  concentration factor  solubility product for calcite  Boltzmann constant  Gapon selectivity coefficient  modified Gapon selectivity coefficient  leaching fraction  concentration of Mg in applied water  molal concentration of the ith ion  concentration of Na in applied water  electrolyte concentration in the bulk solution  sodium adsorption ratio  absolute temperature  total soluble salt concentration  one mole of negative charge on the exchanger  amount of Ca precipitated or dissolved  valence of the counter ion  empirical interactive parameter  activity coefficient  ionic equivalent conductance at infinite dilution  osmotic coefficient of the ith salt  activities  concentrations

CHAPTER 4 CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER Steven J. Deverel, Sabine Goldberg, and Roger Fujii

INTRODUCTION High concentrations of inorganic trace elements in irrigated soils and shallow groundwater pose a threat to agricultural production and the health of humans and animals. They do so in three ways: (1) trace elements can accumulate in plants to levels that cause phytotoxicity; (2) trace elements in plants can adversely affect humans and animals that consume those plants; and (3) trace elements can migrate with seepage through the rootzone and into groundwater, possibly re-emerging with subsurface drainage in surface waters, thereby affecting wildlife, or with groundwater pumped for domestic use, thereby threatening the health of humans. The objective of this chapter is to provide a framework for understanding trace element chemistry in soils and groundwater: the processes that affect trace element concentrations in soils, the biochemical behavior of trace elements in soils, and the methods for evaluating pollution potential in soils in irrigated agricultural areas. Problematically high concentrations of trace elements in soils and groundwater in irrigated areas can occur concomitant with soil and groundwater salinity and can be affected by similar processes that affect soil and groundwater salinity. Problematically high trace element concentrations may also occur independent of salinity. We will describe processes that affect the mobility of trace elements. We have focused on elements designated by the U.S. Environmental Protection Agency (EPA) as priority pollutants (EPA 1986) and on elements that have been documented as pollutants associated with irrigated agriculture. We have divided the trace elements into four categories: (1) alkali and alkaline-earth metals [barium (Ba) and lithium (Li)]; (2) transition metals [chromium (Cr), molybdenum (Mo), and vanadium (V)]; (3) metalloids 89

90

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

[arsenic (As) and boron (B)] and the nonmetal selenium (Se); and (4) informally denoted heavy metals [cadmium (Cd), copper (Cu), lead (Pb), mercury (Hg), nickel (Ni), and zinc (Zn)].

PROCESSES AFFECTING TRACE ELEMENT CONCENTRATIONS Precipitation and Dissolution Mineral precipitation and dissolution reactions often govern the activities of trace elements in soil solutions. These reactions generally are described by a solubility product relation in which the solid dissolves to form soluble constituents. For example, the solubility of solid barite, BaSO4(s), may control the activity of Ba2 in the soil solution of arid-zone soils and groundwaters (Hem 1985). At constant temperature and pressure, a solution in equilibrium with BaSO4(s) is described by the following equation: BaSO4 (s) ⇆ Ba2  SO42

(4-1)

The reduced thermodynamic equilibrium constant expression for this reaction is: Kso  (Ba2) (SO42)

(4-2)

where ( ) denotes activities, the solubility product constant Kso has a value of about 1010 at 25 °C and 1 atmosphere pressure (Hem 1985), and H2O and BaSO4(s) are assumed to be in their standard states with activities equal to unity. Mineral phases affecting the activities of some trace elements in soil solutions and groundwaters can be assessed by calculating mineral saturation indices. The mineral saturation index (SI) is equal to the logarithm value for the quotient of the ion-activity product (IAP) and mineral solubility product constant, or SI  log (IAP/Kso). Negative saturation-index values indicate mineral undersaturation, while positive values indicate supersaturation. Values approaching zero indicate possible thermodynamic mineral equilibrium. Supersaturation often indicates that kinetic constraints are preventing precipitation. The result of this evaluation alone, however, does not constitute proof that a particular mineral phase is present and affecting trace element concentrations, nor that it is absent and not a factor in a water’s trace element chemistry. Mineralogical data can be collected to identify the presence of suspected minerals, thereby providing evidence to support the results of SI calculations.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

91

Evaluation of SIs is limited by the availability and accuracy of mineral thermodynamic data that affect trace element concentrations. Several computer models calculate SIs based on currently available thermodynamic data. Nordstrom et al. (1979b) and Nordstrom and Ball (1984) reviewed a number of these codes. All codes assume that pure solid phases at their standard states are either present or will precipitate. Solid phases with unit activities probably are rare in natural systems (Corey 1981). The evaluation of mineral SIs assumes equilibrium. This may be invalid in some natural systems. In general, the use of equilibrium calculations in mineral aqueous systems is valid only when the residence time of the water greatly exceeds the reaction half-life (Langmuir and Mahoney 1984). For the case of mineral dissolution (mineral-water equilibria) and crystallization, equilibrium conditions may not be present in many natural systems, depending on the specific reactions and residence times. In contrast, the use of equilibrium calculations for solute-solute or solute-water interactions, such as complexation or acid-base reactions, generally will be valid because of the relatively rapid reaction rates, as long as the solution-theory model is adequate to describe the solution considered and the thermodynamic parameters are known with reasonable accuracy. Solution-Phase Speciation of Trace Elements The total (analytical) concentration of any given trace element in the aqueous phase depends on the ionic strength of the solution and the concentrations of other ions with which the trace element forms complexes. For example, the total aqueous concentration of Cd consists of the sum of the concentrations of uncomplexed Cd2 and all of the possible complexes formed in solution. This may be expressed mathematically by the following mass-balance equation: [Cd]total  [Cd2]  [CdSO04]  [CdNO3]  [CdCl]  [CdCl02]  [Cd-Org]  . . .

(4-3)

where [ ] represents molar or molal concentration and Cd-Org represents cadmium complexed by dissolved organic matter. At constant temperature and pressure, the formation of a soluble complex, for example, CdCl, can be described by the following mass action equation: Cd2  Cl ⇆ CdCl

(4-4)

92

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The reduced thermodynamic stability constant expression for this reaction is Ks 

(CdCl ) (Cd 2 )(Cl )

(4-5)

where ( ) denotes activities and H2O is assumed to be at unit activity. Values of stability constants for inorganic complexes involving trace elements or free energy values from which the stability constants can be calculated are published extensively in the literature (see Nordstrom et al. 1979b). For Cd and other metals, pH plays an important role in determining the predominant species in soil solutions and groundwater. For example, at pH values below 7.0, Cd2 is the predominant dissolved Cd species representing 70% to 99%. At pH values above 7.0 in a typical calcareous soil, CdCO3 is the predominant species, representing more than 80% at pH values above 8.5 (Adriano 2001). The activity of a single ion can be expressed as ai  mi yi, where ai  the activity of ion i, mi  the molal (or, in dilute solutions, molar) concentration of species i, and yi is the activity coefficient of species i. For extremely dilute solutions (i.e., approaching infinite dilution), activities are approximated by molar concentrations for all species and yi equals unity. As the concentrations of charged dissolved constituents increase, yi decreases. This decrease represents the degree of departure from ideality of the ionic properties of species i. Individual ion-activity coefficients can be calculated in different ways, depending on the ionic strength of the solution. See Chapter 3 for a complete discussion of calculation of ionic strength and ionic-activity coefficients. Adsorption Activities of trace elements in soil solutions and natural waters frequently are too low to be controlled by precipitation and dissolution of pure solid phases. Additionally, slow precipitation kinetics result in supersaturation with respect to many oxide, hydroxide, and carbonate minerals. Therefore, adsorption reactions occurring at the solid–liquid interface are significant mechanisms controlling trace element activities in many natural waters (Jenne 1977). Adsorption reactions occur mainly in the clay mineral size fraction of the soil and aquifer material at the surfaces of layered aluminosilicate minerals. The reactive surfaces include amorphous oxides and hydroxides of Fe, Mn, and Al; organic matter containing weak-acid functional groups; and metal carbonates, such as calcite, CaCO3(s), and dolomite, CaMg(CO3)2(s) (Jenne 1977). These soil components have the highest surface area and, therefore, the greatest contact with soil water. Even though

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

93

their total concentrations may be relatively low, amorphous oxides, hydroxides, and organic matter occur as coatings on clay-mineral surfaces, thereby exerting considerable control on activities of trace elements in natural waters. In contrast, colloidal-size (1 m) calcite is often found in the clay mineral fraction of arid and semiarid soils. Calcite is an important adsorptive surface for metals. It buffers the soil pH in the alkaline range and decreases metal activities. Jenne (1977) presented an excellent review of these topics. Adsorption of trace elements can be nonspecific, involving simple electrostatic attractions, or specific, involving coordinate covalent bonding. Trace element adsorption occurs at surfaces where the nature of the surface charge affects the extent of the reaction. The permanent negative surface charge of 2⬊1 layered silicates arises from isomorphous substitution of Al3 for silicon (Si4), substitution of a divalent cation for a trivalent cation (e.g., Mg2 for Al3), or substitution of a monovalent cation for a divalent cation (e.g., Li for Mg2) (Bohn et al. 1979). The permanent negative charge of layered silicate minerals is balanced by a diffuse cloud of cations. The cations, or counter ions, are electrostatically attracted to the negative surface, thereby increasing their concentration near the surface. This creates a concentration gradient favoring diffusion away from the surface. The opposite is true for anions. The balancing of these two opposing processes determines the distribution of cations and anions near the permanent charge surface, often referred to as the diffuse or electric double layer (van Olphen 1977). The simplest interpretation of adsorption selectivity of cations is that constant charge surfaces favor higher valence cations (stronger coulombic attraction) and smaller hydrated radii (closer approach to the surface). In contrast to permanent charge sites, charge sites that are pHdependent exist on broken edges of layered silicate minerals, hydroxylated surfaces of metal oxides and hydroxides, and weak-acid functional groups of organic material. Adsorption reactions at pH-dependent charge sites require the formation of surface complexes and some degree of coordinate covalent bonding, which results in the specificity of the adsorption reactions. Figure 4-1 depicts reactions that may take place at the hydroxylated surfaces of metal oxides, as suggested by Schindler (1981). Figure 4-1a shows the weak-acid nature of the surface hydroxyls. At low pH, the surface hydroxyl will be protonated and positively charged, giving rise to some anion exchange capacity. At higher pH, the surface is negatively charged and can contribute to the soil’s cation exchange capacity. Figure 4-1b shows reactions for a dissolved trace metal cation (e.g., Cd2 or Zn2) forming a surface complex by covalent bonding to the surface, displacing protons. Similarly, ligands can exchange with surface hydroxyls (Fig. 4-1c). The adsorption of a dissolved metal ion also can

94

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 4-1. Coordination phenomena at oxide–water interfaces: (a) acid-base reactions at surface hydroxyl groups; (b) deprotonated surface hydroxyls coordinate with dissolved metal ions; (c) surface hydroxyls are replaced by dissolved ligands; (d) dissolved metal ion coordinates with both deprotonated surface hydroxyls and dissolved ligands; (e) dissolved multidentate ligand coordinates with both x and the dissolved metal ion M. From Schindler (1981) with permission. coordinate a dissolved ligand (Fig. 4-1d), and adsorption of a dissolved ligand can coordinate a dissolved metal ion (Fig. 4-1e). Similar reactions occur at edges of layered silicate minerals (e.g., kaolinite edges) and at weak-acid functional groups attached to organic material (e.g., carboxylic acid functional groups).

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

95

Hendrickson and Corey (1981) demonstrated the dependence of Cd selectivity on adsorption site coverage for a variety of solids. A continuum of different types of sites exists in natural systems, ranging from extremely specific sites at low trace element concentrations to nonspecific, cation-exchange-type sites at high concentrations. In natural systems where a small proportion of total site coverage by trace metals or trace element oxyanions is expected, specific adsorption is usually the dominant process controlling the concentrations of many trace metals and trace element oxyanions. Trace metals compete with each other for adsorption on minerals and soils. Kinniburgh et al. (1976) showed that the selectivity sequence for iron oxide gel was Pb2  Cu2  Zn2  Ni2  Cd2  Co2  Sr2  Mg2. For Al oxide gel, the sequence was Cu2  Pb2  Zn2  Ni2  Co2  Cd2  Mg2  Sr2. Forbes et al. (1976) found trace metal affinities for goethite to be Cu2  Pb2  Zn2  Co2  Cd2. Kuo and Mikkelsen (1979) compared Zn2 competition with other trace metals for adsorption on two alkaline soils. They found the adsorption selectivities relative to Zn2 to be: Hg2  Cu2  Fe2  Mn2. Ligand exchange of anions for surface hydroxyls is the predominant process affecting the specific adsorption of oxyanions on goethite (e.g., Goldberg 1985; Hingston 1981) and soils (Fujii et al. 1988; Goldberg et al. 2000, 2002, 2005a,b, 2007, 2008; Neal et al. 1987a,b). Hingston et al. (1968) showed that affinity for anion adsorption on goethite followed the order phosphate  silicate  selenite  fluoride. Describing Adsorption Data Adsorption isotherms describe the relation between the mass of a substance adsorbed (adsorbate) by the solid (adsorbent) and the equilibrium solution concentration (or, more appropriately, activity) supported by the adsorbed phase. These data usually are fit to a model from which adsorption mechanisms may be inferred. Adsorption equations have been incorporated into solute transport models to estimate the retention and release of adsorbing species during transport (e.g., Lewis et al. 1986). One common model used to describe trace element adsorption data is the Freundlich equation: x  KC 1/n m

(4-6)

where x  the mass of trace element adsorbed; m  the mass of solid phase adsorbent; C  the equilibrium concentration of adsorbate in solution; and K and n are empirically determined constants. Many researchers have applied the Freundlich adsorption equation to describe adsorption

96

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of trace metals onto soils and soil minerals (e.g., Bowman et al. 1981; Garcia-Miragaya and Page 1976). Another model commonly used to describe trace element adsorption is the Langmuir equation, which Langmuir (1918) originally developed for adsorption of gases onto solids. When applied to minerals, soils, and sediments, the equation can take the form x kbC  m (1  kC)

(4-7)

where k  a parameter related to the affinity of the adsorbent for the adsorbate; and b  the adsorption maximum. A linearized form of the equation typically is used to describe sorption data as follows: C ⎛ 1 ⎞ ⎛ C⎞ ⎜ ⎟ ⎜ ⎟ x / m ⎝ kb ⎠ ⎝ b ⎠

(4-8)

From a plot of C/x/m versus C, the b and k terms can be estimated from the slope and intercept, respectively. The original derivation of the Langmuir model assumes that only one type of adsorption site exists and that it has constant binding energy (homogeneous surface). However, minerals, soils, and sediments contain surface sites of varying specificity or binding energies. Syers et al. (1973) addressed this problem by using a two-site Langmuir model (high-energy and low-energy sites) to describe the sorption of phosphate by soils. Harter and Baker (1977) and Sposito (1984) addressed other problems associated with using the Langmuir model, such as precipitation of the adsorbate and competition from other adsorbates. Surface complexation models provide molecular descriptions of trace element adsorption using an equilibrium approach that defines surface species, chemical reactions, mass balances, and charge balances. Examples of surface complexation models of the solid–solution interface are: the constant-capacitance model (Stumm et al. 1980), the diffuse-layer model (Dzombak and Morel 1990), the triple-layer model (Davis et al. 1978), and the charge distribution multisite surface complexation (CD-MUSIC) model (Hiemstra and van Riemsdijk 1996). The constant-capacitance model assumes that adsorbing ions form tightly bound inner-sphere complexes. The mechanism of adsorption of anionic trace elements is ligand exchange with surface hydroxyls on oxides and clay minerals. The model has been applied to the adsorption of trace elements including B, Mo, arsenate, selenite, chromate, Cd, Cu, Pb, Ni, Zn, and Hg (Goldberg 1985, 1986; Goldberg and Glaubig 1985; Goldberg et al. 1996; Grossl et al. 1997; Gu and Evans 2007; Gunneriusson

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

97

and Sjöberg 1993). The model has also been used to describe B, Mo, arsenate, and selenite adsorption by soils (Goldberg and Glaubig 1986a; Goldberg et al. 2000, 2002, 2005a,b, 2007, 2008; Sposito et al. 1988). In the diffuse-layer model, adsorbing trace elements also form innersphere surface complexes with anion adsorption proceeding via a ligand exchange mechanism. This model, as its name implies, also includes a diffuse layer formed by the background electrolyte ions. The model has been applied to the adsorption of the trace elements: Cr, V, B, Mo, As, Se, Cd, Cu, Pb, Ni, Zn, and Hg on oxides (Dzombak and Morel 1990; Gustafsson 2003; Peacock and Sherman 2004). The model has not yet been applied to adsorption by soils. In the triple-layer model, adsorption can occur as inner-sphere surface complexation or via an outer-sphere adsorption mechanism that results in the formation of weaker surface complexes containing at least one water molecule between the adsorbing ion and the surface functional group. The model has been applied to the adsorption of the trace elements: Cr, V, B, As, Se, Cd, Cu, Pb, Zn, and Hg on oxides (Balistrieri and Chao 1990; Balistrieri and Murray 1982; Blesa et al. 1984; Davis and Leckie 1978, 1980; Peacock and Sherman 2004; Sarkar et al. 1999) and Hg adsorption on the clay mineral kaolinite (Sarkar et al. 2000). The model has been used to describe chromate (Zachara et al. 1989) and molybdate (Goldberg et al. 1998) adsorption by soils. The CD-MUSIC model considers various types of reactive surface groups: singly, doubly, and triply coordinated hydroxyl groups. The charge of the central ion in the surface complexes is distributed over the coordinating ligands. Inner- and outer-sphere surface complexes are possible. The model has been applied to the adsorption of the trace elements: Cr, Mo, As, Se, Cd, Cu, Pb, and Zn on oxide minerals (Bourikas et al 2001; Hiemstra and van Riemsdijk 1999; Ponthieu et al. 2006; Stachowicz et al. 2006; Venema et al. 1996; Weerasooriya and Tobschall 2000). The model has been used to describe adsorption of arsenate by soil (Gustafsson 2001). Oxidation-Reduction Processes Many trace elements exist in more than one oxidation state (e.g., V, Cr, Se, As), and are therefore affected by electron transfer or oxidation-reduction (redox) reactions. Comprehensive reviews of redox reactions in natural systems are presented by Ponnamperuma (1972), Sposito (1981), and Stumm and Morgan (1981). Most unsaturated soils contain enough oxygen (O2) to maintain oxidizing conditions. In poorly aerated soils where the supply of O2 is limited by the rate of diffusion, anoxic microsites can develop. Under watersaturated conditions, the O2 supply rate may be slower than the O2

98

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

demand, and anaerobic or reducing conditions may develop (Reddy and Patrick 1983). The potential for a substance to accept or donate electrons under equilibrium conditions is represented by standard electrode potentials of reduction half-reactions (Eh) relative to the half-reaction for the hydrogen electrode. These reactions involve the transfer of electrons and depend on the activity of electrons in solution, represented by pE, the negative logarithm of the aqueous electron activity. A major electron-donating process in soils is the microbial oxidation of reduced organic carbon. Oxygen is the primary electron acceptor for this reaction. In the absence of O2, other soil constituents act as the electron acceptor and are reduced. Bohn et al. (1979) lists the principal electron acceptors in soils in the order of their tendency to be reduced, as indicated by the equilibrium potentials of the half-reactions at pH 7: O2  NO 3  MnO2  FeOOH  SO42  H  (CH2O)n. Within the range of redox potentials reported for soils (Baas-Becking et al. 1960), changes in oxidation states for many trace elements are expected. The reduction reactions can increase or decrease the concentration of trace elements. For example, reduction of Mn(III) or Mn(IV) to Mn(II) increases the concentration of Mn because Mn2 is more soluble. In contrast, reduction of SeO42 to the much less soluble elemental Se decreases Se concentration. Selenate reduction may be the major mechanism immobilizing Se in sediments and preventing its transport to the groundwater at the Kesterson National Wildlife Refuge on the west side of the San Joaquin Valley, California. The refuge was formerly an impoundment for agricultural drain water with high concentrations of selenium (e.g., White et al. 1991). Minerals, such as MnO2 and Fe(OH)3, act as sorbents for trace elements; hence, dissolution of these solid phases also can release adsorbed trace elements. During an investigation of As mobility in groundwater, Gulens et al. (1979) found decreased adsorption of As by hydrated ferric oxide under reducing conditions. They attributed this to reduction of Fe(III) to the more soluble Fe(II) and subsequent release of As. Conversely, Reddy and Patrick (1977) reported that water-soluble Pb concentrations decreased (from 45 to 22 g/L at pH 5) with increases in redox potential due to the probable formation of Fe and Mn oxyhydroxides under oxidized conditions, with subsequent adsorption of Pb. Under sufficiently reducing conditions, sulfate is reduced to sulfide and trace metal sulfides precipitate (Lindsay 1979). For example, Bingham et al. (1976) and Reddy and Patrick (1977) reported a decrease in Cd mobility under reducing conditions due to precipitation of cadmium sulfide. Considerable effort has been expended to visually represent stability or predominance fields for minerals and dissolved constituents in natural

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

99

systems, as a function of Eh (or pE) and pH (e.g., Drever 1988; Garrels and Christ 1965; Hem 1985; Lindsay 1979). Their diagrams represent equilibrium systems and provide critical information on possible equilibrium assemblages of minerals and coexisting water compositions. They also help identify disequilibrium assemblages of minerals and aqueous redox couples (Thorstensen 1984). Redox reactions in natural systems generally are considered to be at only partial equilibrium because the kinetics usually are slow; therefore, equilibrium is not achieved, owing to lack of effective coupling of redox reactions (Sposito 1983; Thorstensen 1984). The platinum (Pt) electrode has been used extensively for the measurement of redox potential in field situations. However, Pt-electrode measurements do not necessarily represent the redox potential of the system; they may be the result of a single electroactive redox couple or may represent mixed potentials at the electrode surface. Thus, in natural systems, Pt-electrode measurements should be interpreted with great caution (Lindberg and Runnells 1984; Thorstensen 1984). In natural systems, kinetically limited redox reactions are commonly mediated by microbial organisms. Therefore, the appropriate microbial population must be present to effect a particular reaction. Organisms generally affect only the kinetics of the reaction and do not alter the thermodynamic constraints (Sposito 1983). However, some reactions and species occur that would not be predicted by the overall redox status of the soil due to the presence of microenvironments and microbial populations. Selenium, for example, can be present in oxidized and reduced forms in different microenvironments of the same system. In spite of the problems discussed, environments exist where dominant redox couples are effectively catalyzed and thermodynamic interpretations may be applied. In flooded soils, for example, reduction of Fe and Mn minerals produces high concentrations of Fe2 and Mn2, which can dominate the redox potentials of the system and, thus, allow a quantitative thermodynamic description of redox-active ion activities (Bohn et al. 1979; Ponnamperuma 1972; Sposito 1983). Other investigators have reported Nernstian behavior between measured Eh and various redoxsensitive elements (e.g., Nordstrom et al. 1979a).

BIOGEOCHEMICAL BEHAVIOR AND DISTRIBUTION OF TRACE ELEMENTS IN SOILS AND GROUNDWATER The relative importance of biogeochemical processes discussed in the preceding sections is determined by an individual element’s mode of occurrence. An element’s mode of occurrence is, in turn, determined primarily by the pH and redox status of soil solutions and groundwaters, and by the solid phase controlling solubility and distribution. A review of

100

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the distribution, sources, and behavior of individual trace elements in the near-earth-surface environment is presented here to provide a framework for evaluating potential trace element pollution problems associated with irrigated agriculture. Sedimentary rocks and soils are the solid-earth materials that most affect and determine the trace element chemistry of groundwaters (Hem 1985). Sedimentary rocks can be divided into four groups, based on their resistance to weathering: resistates, hydrolyzates, precipitates, and evaporites. Resistates are rocks composed primarily of residual aluminosilicate minerals that are not easily altered chemically by the weathering of parent rock. Hydrolyzates are composed primarily of insoluble metal oxide and aluminosilicate minerals derived from the weathering of parent rock. Precipitates are rocks that are a result of direct precipitation of mineral matter from aqueous solution. Evaporites are rocks composed of minerals deposited during evaporation. Evaporites affect the composition of some soil solutions and groundwaters associated with irrigated agriculture. However, little is known about their trace element concentrations (Hem 1985). Trace element concentrations generally are higher in hydrolyzates than in resistates and precipitates. This is especially true for those elements that adsorb strongly to metal oxide minerals, such as Zn, Ni, Hg, Pb, Cu, Cr, Ba, and As. Regional assessment of groundwater and drainage-water quality in relation to regional-scale geohydrologic and geochemical processes are common goals of groundwater investigations of trace elements in irrigated areas. Spatial and statistical analysis of data collected over large geographic areas can provide information about the distribution of trace elements and insights about processes affecting their distribution and mobility. Multivariate statistical and spatial analyses can be useful tools for this endeavor. For example, Deverel (1989) used a combination of geostatisitcal analysis (kriging) and principal component analysis integrated using a geographic information system (GIS, Arc/Info) to assess the distribution of principal component scores in the San Joaquin Valley, California. This provided a way of assessing relations among constituents for a large number of samples analyzed for more than 20 chemical constituents. Analysis of the principal component scores in relation to geomorphology elucidated constituent sources and regional processes affecting trace element mobility. Deverel and Galanthine (1989) also successfully utilized spatial analysis using GIS and kriging to assess the relation of soil and groundwater salinity and selenium concentrations in the western San Joaquin Valley. By way of another example, in the western United States during the 1980s and 1990s the U.S. Department of the Interior conducted studies related to Se and other trace elements in soils, drainage waters, surface waters, and biota in major irrigation projects (Engberg et al. 1998; Presser

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

101

et al. 1994). Naftz (1996) attempted to summarize and synthesize the data for the 23 projects where studies had been conducted throughout the western United States. He used a combination of geochemical modeling and simple salt calculations for 1,962 samples collected from the 23 study areas using calculation of the normative salt assemblage with the program SNORM (Bodine and Jones 1986) and pattern recognition modeling to assess selenium–hydrochemical facies associations. The normative salt assemblage can lead to a unique characterization of the sample chemical composition providing information about solute sources and can be visualized as the solid residuum that coexists with the last vestige of water upon evaporation. Naftz’s (1996) analysis demonstrated three distinct hydrochemical facies. Facies 1, which showed an absence of calcium carbonate and elevated concentrations of Na, Ca, and Mg sulfate salts, was positively correlated with hazardous levels of Se. His results are consistent with data collected in the San Joaquin Valley and other areas that point to Se–sulfur associations in primary and secondary minerals as geologic sources for Se in soils, groundwater, and drainage water (Presser and Swain 1990). Alkali and Alkaline Earth Metals Barium Barium is a priority pollutant, and aquatic life and drinking-water criteria have been established (EPA 1986). Mineralogical sources of Ba include aluminosilicate mineral structures, such as feldspars and micas, where Ba substitutes for K and Ca. Barium, which is present in the 2 valence state, also substitutes for Ca in secondary minerals, such as calcite and apatite. In weathering environments, Ba solubility seems to be primarily controlled by reactions with mineral surfaces and by the precipitation and dissolution of barite (BaSO4) and witherite (BaCO3) (Rai and Zachara 1984). The interaction of Ba with mineral surfaces is important in controlling aqueous concentrations. Nonspecific adsorption on constant-charge surfaces exerts substantial control on aqueous Ba concentrations (e.g., Elprince et al. 1980). Barium is also specifically adsorbed onto silicate and Al, Mn, Ti, and Fe oxide minerals (Kinniburgh et al. 1976; Posselt et al. 1978). We found no documentation of high Ba levels in saline environments. The available evidence for marine environments indicates that Ba concentrations are limited by the solubilities of the sulfate and carbonate minerals. Lithium Although typically present as a monovalent cation, lithium’s small ionic and atomic size causes it to behave similarly to the divalent alkaline earth

102

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

metals, especially Mg2. Mineralogically, lithium occurs in Li-halogen and Li-oxygen bonds, primarily in silicate, phosphate, and halide minerals. Released from these primary minerals during weathering, Li commonly is removed from solution by incorporation into clays. Among monovalent cations, Li is the most weakly bonded of all the alkali metals. In soils and groundwater in which Na and K are present at high concentrations, Li may remain in solution and behave conservatively. Deverel and Millard (1988) found that Li concentrations greater than 100 g/L were associated with groundwater salinity in NaSO 4 dominated shallow groundwater beneath the western San Joaquin Valley, California. The environmental toxicity of Li is not well documented; however, it is apparently toxic to plants at concentrations of less than 100 g/L in irrigation water. Bradford (1963) reported that citrus trees may be damaged by irrigation water containing 60 to 100 g/L of Li.

Transition Metals Chromium Chromium exists primarily in the trivalent (Cr(III)) and hexavalent (Cr(VI)) forms. The hexavalent form is substantially more toxic than the trivalent form (EPA 1986). Reduced plant growth (James and Bartlett 1984; Turner and Rust 1971) and soil microbial activity (Ross et al. 1981) were reported due to high concentrations of Cr in soil. Anthropogenic inputs in the form of sewage sludges, waste water, industrial metal processing, and wood preservation and mining activities are the primary sources of high Cr concentrations in soils and groundwaters. Chromium is second to Pb in frequency of occurrence at U.S. Superfund sites (Tiesta 2005). Robertson (1975) reported naturally occurring Cr(VI) concentrations as high as 200 g/L in Arizona, owing to the dissolution and oxidation of Cr(III)-bearing minerals. Oze et al. (2007) summarized data for other locations where high naturally occurring Cr(VI) concentrations have been identified, including Italy, Mexico, and California. Chromite in mafic and serpentinic rocks and sediments occurring at convergent plate margins is the primary source of hexavalent chromium. Oze et al. (2007) demonstrated accelerated chromite oxidation in the presence of birnessite, a common Mn mineral. This natural oxidation can lead to Cr(VI) concentrations that exceed the World Health Organization maximum drinking water contaminant level of 50 g/L. Ball and Izbicki (2004) assessed the processes affecting the natural occurrence of Cr in 200 groundwater samples collected in the Mojave Desert in California. Concentrations ranged from less than 0.1 to 60 mg/L; almost all in the hexavalent form in groundwater samples collected in sediments derived from mafic rock. Close to recharge areas

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

103

where near-oxygen-saturated groundwater pH values were near neutral, Cr concentrations were low. Concentrations increased downgradient as long as oxygen was present. Trivalent Cr was present in groundwater discharge areas where oxygen was below 1 mg/L. Concentrations varied by geologic material. Higher concentrations were associated with mafic rocks, and lower concentrations were associated with alluvial materials derived from mixed volcanic, granitic, and mafic sources. The results of this study indicated that geologic origin of aquifer materials and groundwater redox status are the primary factors affecting Cr species and concentration in groundwater. In Hinkley, California, in the Mojave Desert, where cooling water from an industrial facility containing Cr as a corrosion inhibitor entered the groundwater, there was some evidence for reduction of Cr(VI) to Cr(III) (Andrews and Neville 2003), and pH and Eh data pointed to Cr(III) as the stable species. However, the long-term persistence of Cr(VI) in the upper aquifer indicated that electron donors for reduction of Cr(VI) were limited. In contrast, in the oxic and ferruginous Trinity aquifer in Odessa, Texas, Henderson (1994) presented substantial evidence of adsorption, reduction by iron minerals and immobilization of Cr(VI) at a Superfund site. Trivalent Cr is the thermodynamically stable chemical form in these and most environments (Barnhart 1997). Laboratory studies and modeling by Friedly et al. (1995) and Anderson et al. (1994) demonstrated the importance of reduction of Cr(VI) by ferrous iron in aquifer materials and along groundwater flow paths. However, the time scale for disappearance of Cr(VI) in the field was limited by diffusion resistance. In the model of the field experiments, the reducing agents [Fe(II)-bearing minerals] were heterogeneously distributed in thin strata located between larger nonreducing sand lenses that comprised the bulk of the aquifer solids. The authors identified reducing strata of the order of centimeters thick that were sufficient to contribute enough diffusion resistance to cause the long observed timescales for Cr(VI) reduction in the field relative to laboratory experiments. This reduction process has been identified as a mechanism for Cr(VI) removal along a groundwater flow path underlying and downgradient of a land fill (Loyaux-Lawniczak et al. 2001). Extremely reducing and polluting conditions, such as those identified by Davis et al. (1994) in groundwater underlying and downgradient of a tannery in Woburn, Massachusetts, increased solubility and mobility of Cr(III) due to its association with hydrophilic acids. Manganese also affects the fate and transport of Cr(VI) (Stanin 2005). Oxidation of trivalent chromium by MnO2 is likely to occur in three steps: sorption of trivalent chromium to MnO2 surface sites, subsequent oxidation to hexavalent chromium by Mn(IV), and desorption of Cr(VI). The interaction of Cr and Mn is important in groundwater fate and transport

104

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

as they often show opposing tendencies (Stanin 2005). For example, in the San Joaquin Valley, California, Deverel and Millard (1988) observed high Cr and low Mn concentrations in the upgradient alluvial fan groundwater and the opposite in the more reducing downgradient basin-trough groundwater. Their results indicate that soluble Mn serves as an electron donor for reduction of Cr(VI) to Cr(III). The solubility of Cr(III) in soil solution and groundwater is affected primarily by its interaction with mineral surfaces and co-precipitation with iron-oxide-type minerals. Chromite (FeCr2O4) is stable at low temperatures for most of the water stability region and could be formed as an admixture or co-precipitate with ferrous hydroxide (Hem 1977). Rai and Zachara (1986) identified the solid solution CrxFe1x(OH)3 and Cr(OH)3 (chromium hydroxide) as the solid phase exercising the predominant solubility control on Cr(III) in geologic systems. The amount of Cr(III) adsorbed by a Typic Fragiorthod soil increased with pH increasing to 4 (Bartlett and Kimble 1976). Trivalent Cr was also complexed by organic compounds in soil (Bartlett and Kimble 1976). Trivalent Cr oxidizes to the hexavalent form under conditions prevalent in some field soils (Bartlett and James 1979). Organic material and manganese oxides (Bartlett and James 1979; Rai and Zachara 1986) are important electron acceptors for this reaction. The hexavalent form can be reduced to Cr(III) in soils (James and Bartlett 1983) and in water samples acidified for preservation (Stollenwerk and Grove 1985b). The mobility of Cr(VI) in soil solutions and groundwaters is controlled by adsorption/desorption reactions at mineral surfaces and redox reactions. Amorphous iron oxide has a high capacity for adsorption of Cr(VI) (Leckie et al. 1980). Stollenwerk and Grove (1985a) and Weng et al. (1997) used Langmuir isotherms to describe adsorption of Cr(VI) onto aquifer materials and titanium oxide. Ionic-strength-dependent Cr(VI) adsorption as a function of pH was well described on amorphous iron oxide using the triple-layer model (Davis and Leckie 1980) and on goethite using the diffuse-layer model (Mesuere and Fish 1992). The diffuse-layer model was also used to successfully describe Cr(III) adsorption on silica and aluminum oxide (Csoban and Joo 1999). Adsorption of CrO42 by goethite (Grossl et al. 1997) and titanium oxide (Weng et al. 1997) has been successfully described using the constant-capacitance model. The CDMUSIC model was able to describe Cr(VI) adsorption by goethite as a function of pH and equilibrium CrO42 concentration (Weerasooriya and Tobschall 2000). These approaches underscore the need to consider the supporting electrolyte and competing ions in describing adsorption of Cr(VI). Zachara et al. (1989) described chromate adsorption by soils as a function of pH using the triple-layer model. These authors assumed that the adsorption sites for chromate in these soils were the iron sites of aluminum substituted goethite.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

105

Molybdenum Molybdenum occurs in oxidation states from 3 to 6. Under acidic reducing conditions, the Mo(IV) oxidation state dominates and usually precipitates as a sulfide. Under oxidized conditions, the Mo(VI) oxidation state dominates, primarily as MoO42. Because of its oxyanion form, it commonly behaves conservatively in alkaline, oxidized environments. Consistently, in western San Joaquin Valley shallow groundwater (Deverel and Millard 1988), specific conductance explained 58% to 85% of the variance in Mo concentrations. Molybdenum can substitute for Al, Fe, and Ti in aluminosilicates or metal oxides, and is often associated with organic materials in sedimentary rocks and soils. Little evidence exists of human Mo toxicity from drinking water, and few data are available on chronic Mo toxicity. Ruminants grazing in some parts of the San Joaquin Valley of California were found to be adversely affected by elevated Mo content, especially in legumes and mainly on alkaline soils (Barshad 1948). Unlike legumes, most grasses and grains tend not to accumulate Mo to toxic levels (O’Connor et al. 2001). Mineral surface reactions and interactions with organic matter are the predominant processes controlling the mobility of Mo in soils and groundwaters (Jarrell et al. 1980). Adsorption and incorporation into Mn, Fe, and Al oxides are predominant mechanisms controlling Mo solubility in oxidized environments at low pH (Hem 1977; Reyes and Jurinak 1967). Co-precipitation and adsorption of Mo with and onto manganese and iron oxides decreases with increasing pH greater than 5 (Chan and Riley 1966; Reyes and Jurinak 1967). Molybdenum adsorption has been described on Al and Fe oxides (Goldberg et al. 1996), Ti oxide (Saripalli et al. 2002), and clay minerals (Manning and Goldberg 1996; Motta and Miranda 1989) using the constant-capacitance model. The CD-MUSIC model was used to describe Mo adsorption on titanium oxide as a function of solution Mo concentration and pH (Bourikas et al. 2001). Gustafsson (2003) found the diffuse-layer model to be superior to the CD-MUSIC model in describing molybdate adsorption by amorphous iron oxide. The triple-layer model was able to describe Mo adsorption on Al and Fe oxides, clay minerals, and soils as a function of solution pH and ionic strength (Goldberg et al. 1998, 2008). A new approach for predicting Mo adsorption used a general regression model to predict constant-capacitance model (Goldberg et al. 2002) and triple-layer model (Goldberg et al. 2008) parameters from easily measured soil chemical properties: cation exchange capacity, organic carbon content, inorganic carbon content, and iron oxide content. This approach provides a completely independent model evaluation and was well able to predict Mo adsorption on soils of diverse orders having a wide range of chemical characteristics, as indicated in Fig. 4-2.

106

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 4-2. Prediction of Mo adsorption on Wyo soil: (a) constant-capacitance model; (b) triple-layer model. Circles represent experimental data; model predictions are represented by solid lines. Adapted from Goldberg et al. (2008).

Vanadium Vanadium (V) exists in four different geochemical environments: sulfide deposits, oxidized sulfide ores, aluminosilicates, and iron-oxide deposits (Wedepohl 1972). Vanadium can exist in three oxidation states: 3, 4, and 5 (Hem 1977). The trivalent form of V is incorporated into sulfide minerals, and it often replaces Al in aluminosilicate structures and Fe in iron oxide minerals (Schwertmann and Pfab 1996). In oxidized deposits, V is present in the tri-, tetra- and pentavalent forms. Evans and Garrels (1958) identified a V mineral weathering sequence spanning the valence states from 3 to 5 and associated with oxidized uranium deposits in the Colorado Plateau. Hem (1977) presented evidence that ferric and ferrous vanadates may control the solubility of V in natural waters. The vanadyl ion, VO2, is dominant under reducing environments and is strongly complexed by organic matter (Wehrli and Stumm 1989). A large pool of adsorbed V exists in salt-

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

107

affected Kesterson Reservoir evaporation pond soils that may be mobilized under reducing conditions (Amrhein et al. 1993). Interactions with mineral surfaces substantially influence aqueous V activities in acidic environments. McBride (1979) and Taylor and Giles (1970) described the adsorption of V on montmorillonite and Mn- and Fe3-oxides. Both the vanadyl, VO2 (Wehrli and Stumm 1989), and the vanadate, HVO42 (Peacock and Sherman 2004), species are strongly adsorbed on oxide surfaces. Both the diffuse-layer model and the triplelayer model were successful in describing vanadate adsorption onto goethite (Peacock and Sherman 2004). Because it tends to exist in oxyanion forms, V commonly is mobile in alkaline, oxidized aqueous environments. Consistently, Deverel and Millard (1988) reported V concentrations to be significantly correlated with specific conductance data in shallow groundwater in the San Joaquin Valley, California. Anthropogenic sources of V in soils and groundwater include industrial wastes, sewage sludge, fossil fuel by-products, and mine spoils (Mattigod and Page 1983; National Academy of Sciences 1974). Because of its mobility in oxidized environments, V from these sources may represent a potential pollution hazard. Although tolerance levels for aquatic life have not been established, incidences of industrial V toxicosis, primarily from airborne V, have been documented (National Academy of Sciences 1974).

Metalloids and Nonmetals Arsenic Because of its carcinogenic effects on human health, detectable concentrations (g/L levels) of arsenic (As) in water consumed by humans are undesirable (EPA 1986). In recognition of the hazards arsenic poses to humans and domestic animals, the EPA recently lowered the drinking water standard from 50 to 10 g/L. The predominant factors governing As levels in natural waters are the redox state of the water, aqueous– mineral interactions, and biochemical transformations (Welch et al. 2000). At higher soil redox (500–200 mV), As solubility is low and is predominantly present as As(V) in solution; under moderately reducing conditions (0–100 mV), As solubility is controlled by the dissolution of iron oxides (Masscheleyn et al. 1991). Arsenic exists in several oxidation states. In the tri- and pentavalent forms, it can be incorporated into aluminosilicates and titanium and iron oxide minerals. In the more reduced 3 valence state, As forms sulfides, such as arsenopyrite (FeAsS), realgar (AsS), and orpiment (As2S3). Welch et al. (1988) described geochemical environments in the western United States where As may be present in groundwaters. High As concentrations

108

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

are found in groundwaters in alluvial basins derived from neutral to acidic extrusive and volcanic rocks. Groundwaters in gold mining areas and geothermal areas where arsenosulfide minerals are present generally have high As concentrations. The application of arsenical pesticides is another source of As pollution in groundwaters and soils (Hem 1985). Biogeochemical transformations of arsenic include biomethylation to produce soluble methylated arsenic compounds and mono- or dimethyl arsine gases (Faust et al. 1983; Lemmo et al. 1983; Welch et al. 1988). Volatilization rates of As as arsines are influenced by As form, As concentration, soil moisture, soil temperature, and organic amendment (Gao and Burau 1997). Adsorption and co-precipitation of As by clay minerals and metal oxides are the predominant solubility controls in oxidized environments (Hem 1977; Leckie et al. 1980; Pierce and Moore 1980). Leckie et al. (1980) demonstrated that arsenate adsorption on amorphous iron oxyhydroxide, Fe(OH)3, decreased at pH values greater than 8. Both arsenate, As(V), and arsenite, As(III), are adsorbed on oxide minerals. The adsorption behavior is pH-dependent; at low pH arsenate is adsorbed to a greater extent, while arsenite is adsorbed more at high pH (Raven et al. 1998). Arsenic adsorption has been described on Al and Fe oxides (Gao and Mucci 2001, 2003; Goldberg and Johnston 2001) and clay minerals (Manning and Goldberg 1996) using the constant-capacitance model. The diffuse-layer model (Dixit and Hering 2003; Dzombak and Morel 1990) and the triple-layer model (Benjamin and Bloom 1981; Goldberg and Johnston 2001; Hsia et al. 1992) have been used to describe the adsorption of arsenate and arsenite on iron oxides. Arsenate and arsenite adsorption has been described using the CD-MUSIC model for gibbsite (Weerasooriya et al. 2003, 2004), goethite (Stachowicz et al. 2006), and an allophanic soil (Gustafsson 2001). Similar to the approach for Mo, arsenate adsorption was predicted using a general regression model to predict constant-capacitance-model parameters from easily measured soil chemical properties: cation exchange capacity, inorganic carbon content, organic carbon content, iron oxide content, and surface area (Goldberg et al. 2005b). This approach provided a completely independent model evaluation and was well able to predict As(V) adsorption on many soils from both the southwestern and midwestern United States having a wide range of chemical characteristics, as indicated in Fig. 4-3. Boron Boron generally behaves conservatively in natural waters. It occurs primarily as the undissociated acid H3BO3 at pH  9 and as the B(OH)4 anion at pH  9. Primary and secondary aluminosilicate minerals often

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

109

FIGURE 4-3. Prediction of As adsorption on soils using the constant-capacitance model: (a) Bernow soil; (b) Summit soil. Experimental data are represented by circles for the A horizon and squares for the B horizons. Model predictions are represented by solid lines for the A horizon and dashed lines for the B horizons. Adapted from Goldberg et al. (2005a).

contain B, as it substitutes for silica and Al. In semiarid, and arid environments, B may substantially threaten irrigated agriculture due to its phytotoxic effects and conservative behavior. Soil minerals can attenuate potentially phytotoxic soil-solution B concentrations because plants respond only to B in solution (Keren et al. 1985). Evidence exists that bioaccumulation to levels toxic to waterfowl can occur (Smith and Anders 1989; Hoffman et al. 1990). Deverel and Millard (1988) reported B concentrations that were highly correlated with shallow groundwater salinity in the San Joaquin Valley, California. Boron concentrations can be affected substantially by interaction with mineral surfaces. Most researchers have described the B–soil association as one of adsorption or desorption. Goldberg (1997) and Rai and Zachara (1984) have summarized research in this area.

110

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Elrashidi and O’Connor (1982) used the Langmuir and Freundlich equations to describe B adsorption. The change in adsorptive behavior with changing pH cannot be accurately described using this approach. Boron adsorption onto various minerals and soils is strongly pH-dependent, exhibiting maximum adsorption from pH 7 to 10. Keren et al. (1981) successfully modeled the pH dependence of B adsorption on soils and montmorillonite. Goldberg and Glaubig (1985, 1986a, 1986b) and Goldberg et al. (2000) applied the constant-capacitance model to B adsorption on soils and various minerals from pH 3 to 12. The diffuse-layer model has been applied to B adsorption by amorphous iron oxide (Dzombak and Morel 1990). The triple-layer model was used to describe B adsorption by oxide minerals (Blesa et al. 1984) and kaolinite (Singh and Mattigod 1992). Similar to the approach for Mo and As(V), B adsorption was predicted using a general regression model to predict constant-capacitance-model parameters from easily measured soil chemical properties: surface area, inorganic carbon content, organic carbon content, and Al oxide content (Goldberg et al. 2000, 2004, 2005a). This approach provided a completely independent model evaluation and was well able to predict B adsorption on soils of diverse orders having a wide range of chemical characteristics, as depicted in Fig. 4-4. The prediction equations, developed from B adsorption as a function of pH on a set of soils from the southwestern United States, were able to predict B adsorption as a function of solution B concentration on a set of soils from the midwestern United States, indicating wide predictive capability (Goldberg et al. 2004). The soluble B in irrigated soils may regenerate or increase in concentration after reaching a constant low concentration. Peryea et al. (1985) concluded that slow release of B from adsorption sites and diffusion of B from small pores to large pores are the mechanisms responsible for the regeneration of B in the soil solution. Weathering of soils with a high amount of structural B may release phytotoxic levels of B within a few years of reclamation unless leaching is maintained (Su and Suarez 2004). (See Chapter 19 for a discussion on B reclamation in soils.) Corwin et al. (1999) evaluated five different approaches for predicting B–surface interactions and transport in large lysimeters using the TETrans solute transport model. They concluded that the pH-dependent Keren equation was the best-performing chemical adsorption model. Their results suggest that pH and ionic strength were the most influential chemical factors, followed by temperature and kinetics for B transport. Selenium Selenium generally is similar to sulfur in its chemical behavior (Lakin 1973). Substantial evidence exists that suggests Se is associated primarily with sulfide minerals present in sedimentary environments (Sindeeva

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

111

FIGURE 4-4. Prediction of B adsorption on various depths of a soil profile using the constant-capacitance model: (a) 30–60-cm depth; (b) 60–90-cm depth; (c) 90– 120-cm depth; (d) 120–150-cm depth. Circles represent experimental data; model predictions are represented by solid lines. Adapted from Goldberg et al. (2005b).

1964). The mobility of Se in the aqueous phase depends on aqueous complex formation, redox speciation, and reactions at mineral surfaces. High concentrations of Se in groundwaters and agricultural drain waters in several western U.S. states primarily result from the high mobility of the SeO42 (selenate) ion, the predominant species in alkaline, oxidized environments (Crist 1974; Deverel and Millard 1988; Sylvester et al. 1988). Selenate (SeO42), although it adsorbs on oxide surfaces (Peak and Sparks 2002), generally is nonadsorptive in soils (Neal and Sposito 1989), especially in the presence of high concentrations of sulfate. Selenite (SeO32), in contrast, generally adsorbs strongly onto mineral surfaces (Balistrieri and Chao 1987; Ryden et al. 1987) and soils (Neal et al. 1987a,b). Selenite adsorption has been described on iron oxide minerals using the constant-capacitance model (Goldberg 1985; Duc et al. 2006) and the diffuse-layer model (Dzombak and Morel 1990). The triple-layer model described selenite adsorption on iron and manganese oxide (Balistrieri and Chao 1990). By describing selenite adsorption using inner-sphere surface complexes and Se(VI) adsorption using outer-sphere surface complexes, the triple-layer model (Hayes et al. 1988) and the CD-MUSIC

112

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

model (Hiemstra and van Riemsdijk 1999; Rietra et al. 2001) correctly accounted for the differing strength of adsorption of these ions. Selenite adsorption on four California soils was qualitatively predicted using the surface complexation constants obtained by applying the constant-capacitance model to one other California soil (Sposito et al. 1988). Similar to the approach for Mo, As(V), and B, Se(IV) adsorption was predicted using a general regression model to predict constant-capacitancemodel parameters from easily measured soil chemical properties: inorganic carbon content, organic carbon content, Al oxide content, Fe oxide content, and surface area (Goldberg et al. 2007). This approach provided a completely independent model evaluation and was well able to predict Se(IV) adsorption on many soils from both the southwestern and midwestern United States having a wide range of chemical characteristics, as depicted in Fig. 4-5. Fio et al. (1991) evaluated the mobility of selenite and selenate in San Joaquin Valley irrigated and nonirrigated soils. They conducted sorption studies of selenite and selenate and incorporated the results into a onedimensional solute transport model to simulate changes in soluble and adsorbed Se and to evaluate the potential for leaching of Se to groundwater. Model results showed that selenite can represent a potential longterm source of Se to the groundwater. In contrast, selenate behaves conservatively under the alkaline and oxidized conditions in the western San Joaquin Valley and is easily leached from the soil.

FIGURE 4-5. Prediction of Se adsorption on the B horizon of Dennis soil using the constant-capacitance model. Circles represent experimental data; model predictions are represented by solid lines. Adapted from Goldberg et al. (2007).

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

113

Selenium toxicity in waterfowl at Kesterson Reservoir in the San Joaquin Valley, California, resulted from high concentrations of SeO42 in shallow groundwater (e.g., Presser and Ohlendorf 1987). This Se contamination of the groundwater results from irrigation of soils derived from cretaceous and tertiary geologic materials in the California coast range. The semiarid climate resulted in the formation of naturally saline soils in many areas. Irrigation of these soils resulted in the movement of soluble salts, including Se, into the groundwater. Partial evaporation of a shallow water table in low-lying areas resulted in even higher salinity and Se concentrations (Deverel and Fujii 1988). Farm drainage systems remove this high Se groundwater in some areas. Drain water in some areas is mixed with nonsaline water and reused for irrigation. In other areas, drain water is mixed with surface runoff and irrigation water and discharged to adjacent wetlands and the San Joaquin River. Metals: Cadmium, Copper, Lead, Nickel, and Zinc The biogeochemical behaviors these metals are similar; therefore, they will be discussed as a group. In oxidized sedimentary and secondary mineral environments, the solubility, mobility, and distribution of these metals is primarily affected by adsorption onto and co-precipitation with iron and manganese oxides and oxyhydroxides, pH and interactions with solid- and dissolved-phase organic matter (e.g., Adriano, 2001; Gibbs 1976; Jenne 1968; Kinniburgh et al. 1976; Quirk and Posner 1976). Sewage sludge, animal manures, pesticides, and fertilizers are significant sources of Cd, Cu, Pb, Ni, and Zn in agricultural soils. Mattigod and Page (1983), Alloway (1995) and Adriano (2001) have extensively reviewed the sources and mobility of these metals in soils. Fertilizers and sewage sludge are Cd sources, swine manure and pesticides are Cu sources, and sewage sludge and pesticides are Pb and Zn sources (Adriano, 2001). Sewage sludge is the primary anthropogenic source of Ni. The primary sources of these constituents in agricultural soils are sewage sludge and fertilizers. Pesticides, mining, and smelting activities also contribute these elements to some agricultural soils. Non-anthropogenic sources can also contribute soil concentrations that cause lowered crop productivity and phytotoxicity. For example, Ni concentrations in serpentine-derived soils can be problematically high for some plant growth. Sewage sludge application is a significant source of Cd, Cu, Pb, Ni, and Zn to soils worldwide. There has been some long-term research on the behavior of these elements in sludge-amended agricultural soils as interest in applying sewage sludge to soils has increased substantially during the last two decades. Adriano (2001) summarized results of field experiments that spanned more than 10 years. Data from the United States and Europe indicated Zn as having the highest input rate (usually by an order of magnitude) to sludge-amended soils. The high Zn inputs

114

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

can in turn affect the mobility and bioavailability of other trace metals of concern (Cu, Cd, Ni, and Pb). Incorporation of sewage sludge has led to marked increase in Cd, Cu, Pb, Ni, and Zn concentrations in the upper 30 cm of receiving soils. Field studies generally showed no significant concentration differences between soils with and without sludge application below 30 cm (Adriano 2001; Berti and Jacobs 1998; Dowdy and Volk 1983). However, some studies point to the mobility of these metals to deeper depths (Dowdy et al. 1991; Steehuis et al. 1999). The increased mobility of Cd, Cu, Pb, Ni, and Zn in sludge-amended soils can be attributed to the formation of metal-organic compounds and preferential flow in the unsaturated zone. There are also soil ecological effects of application of sewage sludge as plant growth, soil microbial activity, and fertility can be affected. McGrath et al. (1995) summarized the evidence for effects of metals applied in sewage sludge on soil microbial activity and fertility in longterm for studies conducted in the United States and Europe. Although reduced plant yields were documented in some cases, these were not due to decreased fertility but, rather, to adverse effects on nitrogen-fixing microbes. There was also evidence for reduced total soil microbial mass. Effects on microbes and plants were reduced by increased pH, clay content, and percent organic matter content which enhanced metal sequestration. Notably, effects on the microbial community were observed at soil metal levels below European maximum allowable levels. Application of animal manures (primarily poultry, swine, and bovine) represents a significant source of Cu and Zn in soils. These metals are often added to animal feeds in excessive amounts relative to animal needs and significant amounts are passed through to feces. Sistani and Novak (2006) recently summarized research on trace-metal accumulation in manure amended soils. There is little evidence for movement of these elements past the upper 30 cm of soil. However, there is evidence for Cu and Zn concentrations in surface runoff in the southeastern United States that are toxic to aquatic life. Also, soil concentrations can build up to levels high enough to result in yield reductions and phytotoxicity. Phosphate fertilizers derived from sedimentary phosphates may contain high levels of Cd, although concentrations and amounts of Cd applied to soils are highly variable (Adriano 2001). In contrast, phosphate fertilizers derived from magmatic sources generally contain negligible amounts of Cd. Pesticides, such as copper sulfate, are sources of Cu in many agricultural soils, and there is evidence that application of pesticides and Cu fertilizers has resulted in high soil concentrations that can lead to adverse effects on plants or the microbial community. In tropical areas, long-term continuous application of Cu pesticides and fertilizers has resulted in high crop concentrations and soil residues and occurrence of phytotoxicity symptoms in tropical crops, such as coffee, tea, cocoa, and bananas (Adriano 2001). Application of lead arsenate pesticides prior

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

115

to the 1960s in deciduous fruit tree orchards has led to accumulation of Pb (and As) in orchard surface soils (Peryea and Creger 1994). Atmospheric deposition of Pb from its use in gasoline has resulted in high Pb concentrations in soils in highly industrialized areas. Numerous studies have demonstrated high soil concentrations of Cd, Pb, and Zn and phytotoxicity resulting from atmospheric deposition near Pb, Zn, and battery smelting facilities (Adriano 2001). The phytotoxicity of Cd, Cu, Ni, Pb, and Zn corresponds to their electronegativity (Cd  Cu  Ni  Zn  Pb). Even though there is little evidence for significant movement of Cd, Cu, Pb, Ni, and Zn to groundwater, organic complexation and soil-water and groundwater pH, salinity and redox potential need to be considered when evaluating the mobility, solubility, and bioavailability of these elements in soils and groundwaters. Mobility and bioavailability of metals in soils are generally reduced by increasing pH, organic matter content, cation exchange capacity, and redox potential. Adding lime to soils with high metal concentrations and sludged soils has been shown to reduce plant uptake of Cd, Cu, Pb, Ni, and Zn. These metal species in soil are dependent on pH, as illustrated previously for Cd. The most phytotoxic cationic form predominates at pH values below 7.0, whereas metal hydroxide or carbonate complexed species predominate above pH 7.0. Solubilities of metals are reduced at higher pH values (above 6.5 to 7.0) due to formation of metal phosphates and carbonates. For Cd, Cu, Pb, Ni, and Zn, increasing pH is the predominant factor affecting bioavailability to plants. Low pH (below 4.0) and low redox potential values that favor solubilization of iron minerals facilitate movement in soils and may result in groundwater contamination. This has been observed at industrial hazardous waste sites in which other pollutants have greatly altered subsurface redox and pH. Low redox and pH conditions favor solubilization of metal hydroxides and oxy-hydroxides, thus increasing metal solubility. However, reducing conditions that favor sulfate reduction result in formation of metal sulfides and reduced metal solubility. Cadmium, Cu, Pb, Ni, and Zn strongly adsorb on organic matter in oxic sediments through ion exchange and chelation. Zinc and Ni form insoluble organic complexes and adsorb onto soil organic matter. (Addition of organic matter and/or lime is a common practice for reducing Ni availability in the reclamation of serpentine soils.) For sludge-amended soils, numerous investigators have reported that variable proportions of Cd, Cu, Pb, Ni, and Zn are present in organically bound forms. For example, Hickey and Kittrick (1984) used selective extraction techniques to determine that organically bound heavy metals represented significant proportions of the total metal present in the soil. Other studies with sludgeamended soils also demonstrated that Cu is organically complexed to a greater degree than the other metals (e.g., Emmerich et al. 1982). Paradoxically, metal-organic complexes in soil can enhance aqueous solubility.

116

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Modeling and quantitative estimation of organically complexed metals in natural waters has traditionally been limited by lack of information on the chemical composition of the dissolved organic matter and the appropriate thermodynamic constants. Lerman and Childs (1973), Stumm and Morgan (1981), Davis (1984), Davis and Leckie (1978), and others attempted to quantify the effect of organic ligands on the aquatic solubility and mobility of various metals. Several authors have documented the use of empirical, semi-empirical, and statistical models for estimating metal–organic interactions (e.g., Elzinga et al. 1999; Guy and Chakrabarti 1976; McBride et al. 1997). Several authors have examined the effects of varying salinity on metal mobility and bioavailability. Cadmium mobility and availability for plant uptake generally increased with increasing salinity and Cl concentrations. For example, Amini et al. (2005) assessed the spatial variability of diethylenetriaminepentaacetic acid (DPTA) extractable Cd in Iranian soils and identified a positive and significant correlation with soil electrical conductivity. Similarly, Khoshgoftar et al. (2004) and Norvell et al. (2000) found that NaCl salinity mobilized soil Cd and increased its phytoavailability. In contrast, Zn availability and mobility was generally not affected by moderate salinity levels due to its propensity to adsorb or precipitate independently of soil salinity. Khoshgoftarmanesh et al. (2006) obtained similar results. Their study and other studies (e.g., McLaughlin et al. 1997; Weggler et al. 2004) included the analysis of Zn and Cd species in solution. With increasing salinity, Cd increasingly forms chloride complexes, which increases total solution concentrations and Cd availability to plants. Cadmium-chloride complex formation also results in displacement of Cd from exchange sites. Although we were unable to find studies of effects on groundwater quality, based on these results of these studies, we expect that increasing salinity may increase Cd movement to groundwater. We identified three studies of salinity effects on the bioavailability and mobility of other metals. Preeda et al. (2002) evaluated the effect on DPTA extractable Zn, Cu, Ni, Pb, Cd, and Cr in Thailand sludge-amended soils at different salinity levels ranging from a nonsaline control and 2 to 42dS/m. DPTA-extractable Zn, Cu, and Pb concentrations were higher than the control in the 2 and 4 dS/m treatments, and concentrations were lower in the 8, 19, 31, and 42 dS/m treatments. DPTA-extractable Ni concentrations were unaffected by salinity below 8 dS/m but concentrations were lower at higher salinities. Usman et al. (2005) assessed the effect of metal immobilizing substances (Na-bentonite, Ca-bentonite, zeolite, iron oxides and phosphate fertilizers) and NaCl salinity (1,800 mg/L versus a deionized-water control) on the availability of heavy metals: Zn, Cd, Cu, Ni, and Pb to wheat. The largest reduction in metal bioavailability was found for bentonites.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

117

Irrigation with saline water (1,600 mg L1 NaCl) resulted in a significant increase in metal chloride species (MCl and MCl2). The highest metal complexation with Cl occurred for Cd, which was about 53% of its total soil-solution concentration. The total concentration of Cd in soil solution increased by 1.6- to 2.8-fold in saline water. The NaCl salinity caused a significant increase in uptake and shoot concentration of Cd for two harvests, and small but significant increase in shoot Pb concentration. Saline water increased the availability of Cd and Pb to wheat and decreased the efficiency of bentonites to immobilize soluble Cd. Zinc, Cu, and Ni levels in wheat were not different for the nonsaline and saline treatments, presumably due to the relatively low propensity to form chloride complexes. Keshavarz et al. (2006) assessed the mobility of native-soil Zn under different NaCl salinities in calcareous soils. Their results indicate that the partitioning of soil Zn changed with increasing salinity to 20 dS/m. These authors concluded that increased salinity redistributed nitric-acidextracted Zn to the soluble and exchangeable and organic forms of Zn and became more available to plants with increasing salinity. Mercury Mercury exists in the mercurous (1), mercuric (2), and metallic (0) forms in soils and groundwater (Hem 1970). Mercury readily forms aqueous organic complexes and commonly is present in methylated forms derived from biological metabolism (Jensen and Jernolov 1969). Chronic toxicity of Hg is a dominant environmental issue (EPA 1986), and methyl mercury is of primary concern in surface and subsurface drainage in agricultural soils, especially soils flooded for rice or wetlands. Sources of Hg in soils and groundwater include sewage sludges, mining wastes, atmospheric deposition, soil parent materials, pesticides, fungicides, and fertilizers. Phosphorous fertilizers in particular are reported to have high Hg concentrations—4 g/g to 100 g/g (Anderson 1979). Organic-rich shales, sedimentary clays, and sulfides are the major geologic sources of Hg (Fleischer 1970). The fluxes and forms of Hg in soils and groundwater depend on several factors, including temperature, moisture content, and oxidation and reduction state (e.g., Benes and Havlik 1979; Hem 1985). Soil atmospheric flux can be considerable, with volatilization of inorganic forms and release of methylated forms predominant. Although Hg sometimes can move through the soil profile to shallow groundwater and drain water, it usually is substantially removed from solution by adsorption or incorporation into organic matter (Anderson 1979). In oxidized solutions, the Hg2 ion predominates. The divalent form complexes with Cl– ion, so that aqueous HgCl02 will be the predominant

118

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

dissolved inorganic form in most oxidized waters (Benes and Havlik 1979). Under reducing conditions, Hg exists in the relatively insoluble elemental or sulfide forms, Hg(SH)2 and HgS2 2 . Dismutation of the mercurous form, Hg22, to HgO  Hg2 also increases volatility and solubility (Hem 1970). Organic Hg complexes, including methylated and phenylated forms (Benes and Havlik 1979) comprise a significant portion of soluble Hg in soil solutions and groundwater. These complexes generally are unstable and easily decomposed or volatilized. Adsorption of Hg2 removes Hg from solution. Adsorption of Hg2 onto Mn oxides (Jenne 1970), Fe oxides (Anderson 1979), and soils and sediments (Benes and Havlik 1979) controls the fate of Hg in aqueous environments. Mercury adsorption has been described using various surface complexation models. The constant-capacitance model was used to describe Hg(II) adsorption by goethite (Gunneriusson and Sjöberg 1993; Gunneriusson et al. 1995). Mercury(II) adsorption by Fe oxides and silicas was described using inner-sphere surface complexes in the diffuse-layer model (Bonnissel-Gissinger et al. 1999; Dzombak and Morel 1990; Tiffreau et al. 1995). A combination of inner-sphere and outer-sphere surface complexes were used to describe Hg(II) adsorption on quartz, gibbsite (Sarkar et al. 1999), and kaolinite (Sarkar et al. 2000). Caution must be exercised in the collection and preservation of samples for Hg analysis. Sources of Hg contamination are ubiquitous and the ultraclean sampling methods (see Olsen and deWild 1999; USGS 2011) should be used for groundwater and surface-water sample collection. Adsorption to container walls can be prevented by acidification. Reduction to elemental Hg, which is easily volatilized, can be prevented by the addition of an oxidizing agent, such as potassium dichromate or potassium manganate.

EVALUATION OF POLLUTION POTENTIAL IN SOILS Evaluation of soil quality relative to inorganic trace element concentrations involves general consideration of the soil’s capacity to perform its function within the ecosystem. This historically has been defined as plant productivity and plant uptake and concentration [see Adriano (2001) for more in-depth discussion of assessment of soil quality]. However, other factors requiring consideration include potential leaching to groundwater or drain water and potential movement of trace elements within eroded soils or in surface runoff. Evaluation of uptake by plants requires comparison of trace element concentrations in the soil with levels known to be associated with harmful levels of plant uptake. Evaluation of all considerations requires that the mode of occurrence of the element or elements be

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

119

determined and the effects of the various biogeochemical and physical processes be identified. Soil chemical extractants have generally been used successfully to predict plant uptake, especially under controlled conditions. However, the use of chemical extractants may be ineffective for predicting plant uptake in uncultivated soils and among native species (Gough et al. 1979). Soil chemical extraction procedures theoretically dissolve specific phases and minerals in a soil sample. Chao (1984) wrote a comprehensive review of selective dissolution techniques. The partitioning of elements, such as Se and As, which typically are present as anions and undergo oxidation and reduction reactions in soils, may not be effectively characterized by these procedures (Gruebel et al. 1988). Less destructive techniques, such as scanning electron microscopy, electron-microprobe energy-dispersive x-ray analysis, and x-ray photoelectron spectroscopy, can provide information about the mode of occurrence of trace elements in soils, but their utility is typically limited by high detection limits. Models can be used to evaluate the potential for soils to leach trace elements to the groundwater. The Fio et al. (1991) study cited in the Selenium subsection is one example. Those researchers used a one-dimensional model of Se transport in the unsaturated zone to provide information about future movement to groundwater. Evaluating the spatial distribution of the total concentrations of trace elements in soils in relation to morphological features and hydrologic processes can provide insight about sources, transport processes, and mobility. For example, Tidball et al. (1986a) spatially interpolated the results of the analyses of more than 700 samples collected in the western San Joaquin Valley, California. These samples were analyzed for more than 30 trace elements. Mobile elements, such as Se, were concentrated in areas where there was less leaching of alluvial material or evapoconcentration from a shallow water table. They also found that less-mobile elements, such as Hg and As, tended to be concentrated near the source of the alluvial deposits. Factor analysis provides a way of examining the results from chemical analyses of a large number of samples containing trace elements (Joreskog et al. 1976) by simplifying large quantities of data into associations with master variables. Controversy arises because of the subjectivity involved in identifying master variables and assigning physical and chemical meaning to them. Tidball et al. (1986b) used factor analysis to evaluate the results of the sampling described by plotting the spatially interpolated factor scores. McNeal et al. (1985) determined the mode of occurrence of various trace metals in uncultivated soils of the northern Great Plains by combining partial dissolution techniques (Chao 1984) and factor analysis. Another method of understanding the areal distribution and spatial variability of soil trace element concentrations is the low-density, hierarchical,

120

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

analysis-of-variance approach extensively used by the U.S. Geological Survey Branch of Regional Geochemistry (Tourtelot and Miesch 1975). Geochemical differences among regional subgroups are identified by determining the proportion of the total variability among all samples represented by the subgroups. This division also allows for evaluation of the proportion of the variance attributable to sampling and analytical error. Geographic information systems (GIS) are used extensively for mapping soil trace element concentrations in soils relative to factors that may affect concentrations and transport to groundwater. Of specific interest is the possibility of using GIS coupled with transport modeling to predict potential groundwater contamination. Corwin and colleagues at the USDA-ARS U.S. Salinity Laboratory in Riverside, California, developed GIS for this use (e.g., Corwin 1996; Corwin and Wagenet 1996). Coupling of GIS to transport models for assessing nonpoint source pollution presents numerous challenges resulting primarily from limitations in the mathematical representations of complex transport processes, heterogeneous media, data availability, and difficulties in parameter estimation. To date, applications have included primarily salt, pesticides, and nitrates. GIS linked to deterministic or stochastic transport models offer promise for characterizing the spatially variable nature of groundwater contamination potential due to nonpoint sources of trace elements. Soils are generally considered contaminated if the concentrations exceed background levels unaffected by human activities. However, there are numerous examples of high naturally occurring soil trace element concentrations. For example, Ni concentrations in serpentine-derived soils can be toxic to plants and animals. Sampling methodology and soil clay and organic matter content also affect soil concentrations as trace elements accumulate in these soil fractions. The comparison of soil trace element concentrations with background and/or established acceptable levels needs to be conducted in conjunction with assessment of an element’s mode of occurrence, mobility, and relevant soil physical and chemical characteristics. Several sources of trace element background values are available (Alloway 1995; Bowen 1979; Kabata-Pendias and Pendias 1992). Adriano (2001) provided a comparison of values for several investigators, including Bowen (1979) who compiled the concentration ranges and median values of trace elements shown in Fig. 4-6. Values of total concentrations higher than those listed may indicate that a problem exists. However, because of spatial variability and factors affecting concentrations, it is difficult to assign single values for background concentrations. Also, the mode of occurrence and mobility of an element in the soil needs to be evaluated before drawing conclusions about potential harmful effects to plants or animals or movement to surface or groundwater.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

121

FIGURE 4-6. Ranges of concentrations of trace elements in soils in mg/kg (Bowen 1979). The rectangle represents the middle 50% of the data, and vertical lines represent the median.

122

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

During the last 20 years, various agencies have developed environmental standards for soils by determining potential health effects related to levels and duration of exposure to trace elements. Values vary depending on assumptions about exposure scenarios. Relevant to this chapter, standards for trace elements have been established for application of sewage sludge to agricultural soils. Adriano (2001) described the critical levels used throughout Europe.

SUMMARY The dominant processes that affect trace elements in soils and groundwater are (1) precipitation and dissolution, (2) surface interactions and adsorption, and (3) oxidation-reduction reactions. For each element or group of elements, we provided a review of mode of occurrence, sources, and relevant biogeochemical processes affecting mobility and plant health, and trace element uptake. Last, we presented a discussion of considerations of evaluation for pollution potential in soils. Salient general conclusions follow. • Because of their tendency to form oxyanions, the transition metals, metalloids, and nonmetals (As, B, Cr, Se, V) are often mobile in soils and groundwater. Mobility depends primarily on redox potential and pH and secondarily on surface interactions. Depending on geologic sources, high concentrations of these elements are often observed in arid and semiarid areas because of the typically oxidized and alkaline conditions. • The informally named heavy metals (Cd, Cu, Pb, Ni, and Zn) are generally unlikely to be mobile in soils due to their strong potential to adsorb to soil minerals and soil organic matter, and to precipitate as carbonates and phosphates. There is typically little potential for leaching to groundwater at near-neutral pH and above. Long-term application of sewage sludge, manures, and pesticides can cause high soil concentrations that may affect plant productivity and affect the soil microbial community. Low pH, high salinity, and formation of soluble organic complexes and preferential flow can result in increased mobility and leaching to deeper soil horizons. Cadmium and Pb form soluble chloride complexes in saline soils, which increases their availability to plants and may increase the potential for leaching to groundwater. • Mercury is an important contaminant in many locations throughout the world. It is subject to a complex series of biogeochemical processes that affect its mobility and toxicity.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

123

• Evaluation of the pollution potential of soils requires evaluation of multiple factors and criteria. These include (1) spatial variability of soil trace element concentrations and the processes affecting concentrations, (2) background concentrations unaffected by anthropogenic activity, and (3) the forms of constituents of concern. In evaluating pollution potential, it is important to consider the following: – To effectively compare soil trace element concentrations with background levels requires collection of sufficient numbers of samples to effectively represent measures of central tendency and variance, which necessitates analysis of spatial variability relative to possible sources. – Use of multivariate statistical methods, such as factor and principal component analysis, can provide insight into geochemical associations and processes affecting distribution and mobility and provide a framework for analyzing large amounts of trace element data. – Geographic information systems (GIS) can be an effective tool for regional and subregional assessment of the distribution and mobility of soil and groundwater trace element concentrations and the potential for movement of trace elements to groundwater. – Soil extractants can provide useful information about how trace elements are partitioned in soils, especially relative to plant uptake. However, for some elements, such as As, these selective extraction techniques are ineffective. • Trace element adsorption in soils is a key factor for understanding mobility of many trace elements and can be quantified in different ways. These include surface complexation, constant-capacitance, diffuse-layer, and triple-layer models, which have been applied successfully in a variety of systems and to numerous trace elements.

REFERENCES Adriano, D. (2001). Trace elements in terrestrial environments, Springer, New York. Alloway, B. J. (1995). Heavy metals in soils, Chapman and Hall, London. Amini, M., Khademi, H., Afyuni, M., and Abbaspour, K. C. (2005). “Variability of available cadmium in relation to soil properties and land use in an arid region in central Iran.” Water, Air, and Soil Poll., 162, 205–218. Amrhein, C., Mosher, P. A., and Brown, A. D. (1993). “The effects of redox on Mo, U, B, V, and As solubility in evaporation pond soils.” Soil Sci., 155, 249–255. Anderson, A. (1979). “Mercury in soils,” in The biogeochemistry of mercury in the environment, J. O. Nriagu, ed., Elsevier, Amsterdam, 79–110.

124

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Anderson, L. D., Kent, D., and Davis, J. A. (1994). “Batch experiments characterizing the reduction of Cr(VI) using suboxic material from a mildly reducing sand and gravel aquifer.” Environ. Sci. Technol., 28, 178–185. Andrews, C. B. and Neville, C. J. (2003). “Groundwater flow in a desert basin: Challenges for simulating transport of dissolved chromium.” Groundwater, 41, 219–226. Baas-Becking, L. G. M., Kaplan, I. R., and Moore, D. (1960). “Limits of natural environment in terms of pH and oxidation-reduction potentials.” J. Geol., 68, 243–284. Balistrieri, L. S., and Chao, T. T. (1987). “Selenium adsorption by goethite.” Soil Sci. Soc. Am. J., 51, 1145–1151. ———. (1990). “Adsorption of selenium by amorphous iron oxyhydroxide and manganese oxide.” Geochim. Cosmochim. Acta, 54, 739–751. Balistrieri, L. S., and Murray, J. W. (1982). “The adsorption of Cu, Pb, Zn, and Cd on goethite from major ion seawater.” Geochim. Cosmochim. Acta, 46, 1253–1265. Ball, J. W., and Izbicki, J. A. (2004). “Occurrence of hexavalent chromium in groundwater in the western Mojave Desert, California.” Appl. Geochem., 19, 1123–1135. Barnhart, J. (1997). “Chromium chemistry and implications for environmental fate and toxicity.” J. Soil. Contamination, 6, 561–568. Barshad, I. (1948). “Molybdenum content of pasture plants in relation to toxicity to cattle.” Soil Sci., 66, 187–191. Bartlett, R. J., and James, B. (1979). “Behavior of chromium in soils: III. Oxidation.” J. Environ. Qual., 8, 31–35. Bartlett, R. J., and Kimble, J. M. (1976). “Behavior of chromium in soils: I. Trivalent forms.” J. Environ. Qual., 5, 379–383. Benes, F., and Havlik, B. (1979). “Speciation of mercury in natural waters,” in Biogeochemistry of mercury in the environment, J. O. Nriagu, ed. Elsevier, Amsterdam, 175–202. Benjamin, M. M., and Bloom, N. S. (1981). “Effects of strong binding of anionic adsorbates on adsorption of trace metals on amorphous iron oxyhydroxide,” in Adsorption from aqueous solutions, P. H. Tewari, ed., Plenum Publishing, New York, 41–60. Berti, W. R., and Jacobs, L. W. (1998). “Distribution of trace elements in soil from repeated sewage sludge applications.” J. Environ. Qual., 27, 1280–1286. Biggar, J. W., and Fireman, M. (1960). “Boron adsorption and release by soils.” Soil Sci. Soc. Am. Proc., 24, 115–119. Bingham, F. T., Page, A. L., Mahler, R.J ., and Ganje, T. J. (1976). “Cadmium availability to rice in sludge-amended soil under ’flood’ and ’nonflood’ culture.” Soil Sci. Soc. Am. J., 40, 715–719. Blesa, M. A., Maroto, A. J. G., and Regazzoni, A. E. (1984). “Boric acid adsorption on magnetite and zirconium dioxide.” J. Colloid Interface Sci., 99, 32–40. Bodine, M. W., and Jones, B. F. (1986). The salt norm: A quantitative chemical-mineralogical characterization of natural waters, U.S. Geological Survey Water Resource Investigative Report 86-4086, U.S. Geological Survey, Washington, D.C. Bohn, H. L., McNeal, B. L., and O’Connor, G. A. (1979). Soil chemistry, John Wiley and Sons, New York.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

125

Bonnissel-Gissinger, P., Alnot, M., Lickes, S.-P., Ehrhardt, J.-J., and Behra, P. (1999). “Modeling the adsorption of mercury(II) on (hydr)oxides. II: -FeOOH (goethite) and amorphous silica.” J. Colloid Interface Sci., 215, 313–322. Bourikas, K., Hiemstra, T., and van Riemsdijk, W. H. (2001). “Adsorption of molybdate monomers and polymers on titania with a multisite approach.” J. Phys. Chem. B, 105, 2393–2403. Bowen, H. J. M. (1979). Environmental chemistry of the elements, Academic Press, London, 32–46. Bowman, R. S., Essington, M. E., and O’Connor, G. A. (1981). “Soil sorption of nickel: Influence of solution composition.” Soil Sci. Soc. Am. J., 45, 860–865. Bradford, G. R. (1963). “Lithium survey of California’s water resources.” Soil Sci., 96, 77–81. Chan, K. M., and Riley, J. P. (1966). “The determination of molybdenum in natural waters, silicates and biological materials.” Anal. Chim. Acta, 36, 220–229. Chao, T. T. (1984). “Use of partial dissolution techniques in geochemical exploration.” J. Geochem. Explor., 20, 101–135. Corey, R. B. (1981). “Adsorption vs. precipitation,” in Adsorption of inorganics at solid-liquid interfaces, M. A. Anderson and A. J. Rubin, eds., Ann Arbor Science Publishers, Ann Arbor, Mich., 161–182. Corwin, D. L. (1996). “GIS applications of deterministic solute transport models for regional-scale assessment of non-point source pollutants in the vadose zone,” in Application of GIS to the modeling of non-point source pollutants in the vadose zone, D. L. Corwin and K. Loague, eds., SSSA Special Publication 48, Soil Science Society of America, Madison, Wisc. Corwin, D. L., Goldberg, S., and David, A. (1999). “Evaluation of a functional model for simulating boron transport in soil.” Soil Sci., 164, 697–717. Corwin, D. L., and Wagenet, R. J. (1996). “Application of GIS to the modeling of non-point source pollutants in the vadose zone: A conference summary.” J. Environ. Qual., 25, 403–411. Crist, M. A. (1974). Selenium in waters in and adjacent to the Kendrick project, Natrona County, Wyo, U.S. Geological Survey Water-Supply Paper 2023, U.S. Geological Survey, Washington, D.C. Csoban, K., and Joo, P. (1999). “Sorption of Cr(III) on silica and aluminium oxide: Experiments and modeling.” Colloids Surf. A: Physicochem. Eng. Aspects, 151, 97–112. Davis, J. A. (1984). “Complexation of trace metals by adsorbed natural organic matter.” Geochim. Cosmochim. Acta, 48, 679–691. Davis, J. A., James, R. O., and Leckie, J. O. (1978). “Surface ionization and complexation at the oxide/water interface. II. Surface properties of amorphous iron oxyhydroxide and adsorption of metal ions.” J. Colloid Interface Sci., 67, 90–107. Davis, A., Kempton, H., and Nicholson, A. (1994). “Groundwater transport of arsenic and chromium at a historical tannery, Woburn, Massachusetts, U.S.A.” Appl. Geochem., 9, 569–582. Davis, J. A., and Leckie, J. O. (1978). “Effect of adsorbed complexing ligands on trace metal uptake by hydrous oxides.” Environ. Sci. Technol., 12, 1309–1315. ———. (1980). “Surface ionization and complexation at the oxide/water interface. 3. Adsorption of anions.” J. Colloid Interface Sci., 74, 32–43.

126

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Deverel, S. J. (1989). “Geostatistical and principal component analysis of groundwater chemistry and soil salinity data, San Joaquin Valley, California, in Regional Characterization of Water Quality, Stephen Ragone, ed., Publication No. 182, International Association of Hydrological Sciences, IAHS Press, Centre for Ecology and Hydrology, Wallingford, UK. Deverel, S. J., and Gallanthine, S. K. (1989). “Distribution of salinity and selenium in relation to hydrologic and geochemical processes, San Joaquin Valley, California.” J. Hydrol., 109, 125–149. Deverel, S. J., and Fujii, R. (1988). “Processes affecting the distribution of selenium in shallow ground water of agricultural areas, western San Joaquin Valley, Calif.” Water Resour. Res., 24, 516–524. Deverel, S. J., and Millard, S. P. (1988). “Distribution and mobility of selenium and other trace elements in shallow ground water of the western San Joaquin Valley, Calif.” Environ. Sci. Technol., 22, 697–702. Dixit, S., and Hering, J. G. (2003). “Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: Implications for arsenic mobility.” Environ. Sci. Technol., 37, 4182–4189. Dowdy, R. H., Latterell, J. J., Hinesly, T. D., and Grossman, R. B. (1991). “Trace metal movement in an Aeric Ochraqualf following 14 years of annual sludge application.” J. Environ. Qual. 20, 119–123. Dowdy, R. H., and Volk, V. V. (1983). “Movement of heavy metals in soils,” in Chemical mobility and reactivity in soil systems, D. W. Nelson, ed., SSSA Publication 11, Soil Science Society of America, Madison, Wisc. Drever, J. I. (1988). The geochemistry of natural waters, 2nd ed., Prentice Hall, Upper Saddle River, N.J. Duc, M., Lefevre, G., and Fedoroff, M. (2006). “Sorption of selenite ions on hematite.” J. Colloid Interface Sci., 298, 556–563. Dzombak, D. A., and Morel, F. M. M. (1990). Surface complexation modeling: Hydrous ferric oxide, John Wiley and Sons, New York. Elprince, A. M., Vanselow, A. D., and Sposito, G. (1980). “Heterovalent ternary cation exchange equilibria: NH4–Ba 2–La 3 exchange on montmorillonite.” Soil Sci. Soc. Am. J., 44, 964–969. Elrashidi, M. A., and O’Connor, G. A. (1982). “Boron sorption and desorption in soils.” Soil Sci. Soc. Am. J., 46, 27–31. Elzinga, E. J., Van Grinsven, J. J. M., and Swartjes, F. A. (1999). “General purpose Freundlich isotherms for cadmium, copper and zinc in soils.” Eur. J. Soil Sci., 50, 139–149. Emmerich, W. E., Lund, L. J., Page, A. L., and Chang, A. C. (1982). “Solid phase forms of heavy metals on sewage sludge-treated soils.” J. Environ. Qual., 11, 178–181. Engberg, R. A., Westcott, D. A., Delamore, M., and Hotz, D. D. (1998). “Federal and state perspective on regulation and remediation of irrigation induced selenium problems,” in Environmental chemistry of selenium, W. T. Frankenburger and R. A. Engberg, eds., Marcel Dekker, Inc., New York, 1–26. Evans, H. T. and Garrels, R. M. (1958). “Thermodynamic equilibria of vanadium in aqueous systems as applied to the interpretation of the Colorado Plateau ore deposits.” Geochim. Cosmochim. Acta, 15, 131–149. Faust, S. D., Winka, A., Belton, T., and Tucker, R. (1983). “Assessment of the chemical and biological significance of arsenical compounds in a heavily contami-

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

127

nated watershed. Part II. Analysis and distribution of several arsenical species.” J. Environ. Sci. Health, 18, 389–411. Fio, J. L., Fujii, R., and Deverel, S. J. (1991). “Selenium mobility and distribution in irrigated and nonirrigated alluvial soils.” Soil Sci. Soc. Am. J., 55, 1313–1320. Fleischer, M. (1970). “Summary of the literature on the inorganic geochemistry of mercury,” in Mercury in the environment, U.S. Geological Survey Professional Paper 713, U.S. Geological Survey, Washington, D.C. Forbes, E. A., Posner, A. M., and Quirk, J. P. (1976). “The specific adsorption of divalent Cd, Co, Cu, Pb, and Zn on goethite.” J. Soil Sci., 27, 154–166. Friedly, J. C., Davis, J. A., Kent, D. B. (1995). “Modeling hexavalent chromium reduction in groundwater in field-scale transport and laboratory batch experiments.” Water Resources Res., 31, 2783–2794. Fujii, R., Deverel, S. J., and Hatfield, D. B. (1988). “Distribution of selenium in soils of agricultural fields, western San Joaquin Valley, Calif.” Soil Sci. Soc. Am. J., 52, 1274–1283. Gao, S., and Burau, R. G. (1997). “Environmental factors affecting rates of arsine evolution from and mineralization of arsenicals in soil.” J. Environ. Qual., 26, 753–763. Gao, Y., and Mucci, A. (2001). “Acid base reactions, phosphate and arsenate complexation, and their competitive adsorption at the surface of goethite in 0.7 M NaCl solution.” Geochim. Cosmochim. Acta, 65, 2361–2378. ———. (2003). “Individual and competitive adsorption of phosphate and arsenate on goethite in artificial seawater.” Chem. Geol., 199, 91–109. Garcia-Miragaya, J., and Page, A. L. (1976). “Influence of ionic strength and inorganic complex formation of the sorption of trace amounts of Cd by montmorillonite.” Soil Sci. Soc. Am. J., 40, 658–663. Garrels, R. M., and Christ, C. L. (1965). Solutions, minerals and equilibria, Harper and Row, New York. Gibbs, R. J. (1976). “Transport phases of transition metals in the Amazon and Yukon Rivers.” Geol. Soc. Am. Bull., 88, 829–843. Goldberg, S. (1985). “Chemical modeling of anion competition on goethite using the constant capacitance model.” Soil Sci. Soc. Am. J., 49, 851–856. ———. (1986). “Chemical modeling of arsenate adsorption on aluminum and iron oxide minerals.” Soil Sci. Soc. Amer. J., 50, 1154–1157. ———. (1997). “Reactions of boron with soils.” Plant Soil, 193, 35–48. Goldberg, S., Forster, H. S., and Godfrey, C. L. (1996). “Molybdenum adsorption on oxides, clay minerals, and soils.” Soil Sci. Soc. Am. J., 60, 425–432. Goldberg, S., and Glaubig, R. A. (1985). “Boron adsorption on aluminum and iron oxide minerals.” Soil Sci. Soc. Am. J., 49, 1374–1379. ———. (1986a). “Boron adsorption on California soils.” Soil Sci. Soc. Am. J., 50, 1173–1176. ———. (1986b). “Boron adsorption and silicon release by the clay minerals kaolinite, montmorillonite, and illite.” Soil Sci. Soc. Am. J., 50, 1442–1448. Goldberg, S., Corwin, D. L., Shouse, P. J., and Suarez, D. L. (2005a). “Prediction of boron adsorption by field samples of diverse textures.” Soil Sci. Soc. Am. J., 69, 1379–1388. Goldberg, S., and Johnston, C. T. (2001). “Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectroscopy, and surface complexation modeling.” J. Colloid Interface Sci., 234, 204–216.

128

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Goldberg, S., Johnston, C. T., Suarez, D. L., and Lesch, S. M. (2008). “Mechanism of molybdenum adsorption on soils and soil minerals evaluated using vibrational spectroscopy and surface complexation modeling,” in Adsorption of metals by geomedia II: Variables, mechanisms, and model applications, M. O. Barnett and D. B. Kent, eds., Developments in Earth and Environmental Sciences 7, Elsevier, Amsterdam, 235–266. Goldberg, S., Lesch, S. M., and Suarez, D. L. (2000). “Predicting boron adsorption by soils using soil chemical parameters in the constant capacitance model.” Soil Sci. Soc. Am. J., 64, 1356–1363. ———. (2002). “Predicting molybdenum adsorption by soils using soil chemical parameters in the constant capacitance model.” Soil Sci. Soc. Am. J., 66, 1836–1842. ———. (2007). “Predicting selenite adsorption by soils using soil chemical parameters in the constant capacitance model.” Geochim. Cosmochim. Acta, 71, 5750–5762. Goldberg, S., Lesch, S. M., Suarez, D. L., and Basta, N. T. (2005b). “Predicting arsenate adsorption by soils using soil chemical parameters in the constant capacitance model.” Soil Sci. Soc. Am. J., 69, 1389–1398. Goldberg, S., Su, C., and Forster, H. S. (1998). “Sorption of molybdenum on oxides, clay minerals, and soils: Mechanisms and models,” in Adsorption of metals by geomedia: Variables, mechanisms, and model applications, E. A. Jenne, ed., Proc., American Chemical Society Symposium, Academic Press, San Diego, 401–426. Goldberg, S., Suarez, D. L., Basta, N. T., and Lesch, S. M. (2004). “Predicting boron adsorption isotherms by midwestern soils using the constant capacitance model.” Soil Sci. Soc. Am. J., 68, 795–801. Gough, L. P., Severson, R. C., and McNeal, J. M. (1979). “Extractable and total-soil element concentrations favorable for native plant growth in the northern Great Plains,” in Ecology and coal resource development, vol. 2., M. K. Wali, ed., Pergamon Press, New York, 859–869. Grossl, P. R., Eick, M., Sparks, D. L., Goldberg, S., and Ainsworth, C. C. (1997). “Arsenate and chromate retention mechanisms on goethite. II. Kinetic evaluation using a pressure-jump relaxation technique.” Environ. Sci. Technol., 31, 321–326. Gruebel, K. A., Davis, J. A., and Leckie, J. O. (1988). “The feasibility of using sequential extraction techniques for As and Se in soils and sediments.” Soil Sci. Soc. Am. J., 52, 390–397. Gu, X., and Evans, L. J. (2007). “Modelling the adsorption of Cd(II), Cu(II), Ni(II), Pb(II), and Zn(II) onto Fithian illite.” J. Colloid Interface Sci., 307, 317–325. Gulens, J., Champ, D. R., and Jackson, R. E. (1979). “Influence of redox environments on mobility of arsenic in ground water,” in Chemical modeling in aqueous systems, E. A. Jenne, ed., American Chemical Society, Washington, D.C., 81–95. Gunneriusson, L., Baxter, D., and Emteborg, H. (1995). “Complexation at low concentrations of methyl and inorganic mercury(II) to a hydrous goethite ( -FeOOH) surface.” J. Colloid Interface Sci., 169, 262–266. Gunneriusson, L., and Sjöberg, S. (1993). “Surface complexation in the Hgoethite ( FeOOH)-Hg(II)-chloride system.” J. Colloid Interface Sci., 156, 121–128.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

129

Gustafsson, J. P. (2001). “Modelling competitive anion adsorption on oxide minerals and an allophane-containing soil.” Europ. J. Soil Sci., 52, 639–653. ———. (2003). “Modelling molybdate and tungstate adsorption to ferrihydrite.” Chem. Geol., 200, 105–115. Guy, R. D., and Chakrabarti, C. L. (1976). “Studies of metal–organic interactions in model systems pertaining to natural waters.” Can. J. Chem./Rev. Can. Chim., 54, 2600–2611. Harter, R. D., and Baker, D. E. (1977). “Application and misapplication of the Langmuir equation to soil adsorption phenomena.” Soil Sci. Soc. Am. J., 41, 1077–1080. Hayes, K. F., Papelis, C., and Leckie, J. O. (1988). “Modeling ionic strength effects on anion adsorption at hydrous oxide/solution interfaces.” J. Colloid Interface Sci., 125, 717–726. Hem, J. D. (1970). “Chemical behavior of mercury in aqueous media,” in Mercury in the environment, U.S. Geological Survey Professional Paper 713, U.S. Geological Survey, Washington, D.C. ———. (1977). “Reactions of metal ions at surfaces of hydrous iron oxide.” Geochim. Cosmochim. Acta, 41, 527–538. ———. (1985). Study and interpretation of the chemical characteristics of natural water, U.S. Geological Survey Water Supply Paper 2254, U.S. Geological Survey, Washington, D.C. Henderson, T. (1994). “Geochemical reduction of hexavalent chromium in the Trinity Sand Aquifer.” Groundwater, 32, 477–486. Hendrickson, L. L., and Corey, R. B. (1981). “Effect of equilibrium metal concentrations on apparent selectivity coefficients of soil complexes.” Soil Sci., 131, 163–171. Hickey, M. G., and Kittrick, J. A. (1984). “Chemical partitioning of cadmium, copper, nickel, and zinc on soils and sediments containing high levels of heavy metals.” J. Environ. Qual., 13, 372–376. Hiemstra, T., and van Riemsdijk, W. H. (1996). “A surface structural approach to ion adsorption: The charge distribution (CD) model.” J. Colloid Interface Sci., 179, 488–508. ———. (1999). “Surface structural ion adsorption modeling of competitive binding of oxyanions by metal (hydr)oxides.” J. Colloid Interface Sci., 210, 182–193. Hingston, F. J. (1981). “A review of anion adsorption,” in Adsorption of inorganics at solid–liquid interfaces, M. A. Anderson and A. J. Rubin, eds., Ann Arbor Science Press, Ann Arbor, Mich., 51–90. Hingston, F. J., Posner, A. M., and Quirk, J. P. (1968). “Adsorption of selenite by goethite,” in Adsorption from aqueous solution, R. F. Gould, ed., Advances in Chemistry Series, No. 79, American Chemical Society, Washington, D.C., 82–90. Hoffman, D. J., Camardese, M. B., Lecaptain, L. J., and Pendleton, G. W. (1990). “Effects of boron on growth and physiology of mallard ducklings.” Environ. Toxicol. Chem. 9, 335–346. Hsia, T. H., Lo, S. L., and Lin, C. F. (1992). “As(V) adsorption on amorphous iron oxide: Triple layer modeling.” Chemosphere, 25, 1825–1837. James, B. R., and Bartlett, R. J. (1983). “Behavior of chromium in soils. VI. Interactions between oxidation-reduction and organic complexation.” J. Environ. Qual., 12, 173–176.

130

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1984). “Plant–soil interactions of chromium.” J. Environ Qual., 13, 67–70. Jarrell, W. M., Page, A. L., and Elseewi, A. A. (1980). “Molybdenum in the environment.” Residue Rev., 74, 1–43. Jenne, E. A. 1968. “Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils and water: The significant role of hydrous Mn and Fe oxides.” Adv. Chem., 73, 337–387. ———. (1970). “Atmospheric and fluvial transport of mercury,” in Mercury in the environment, U.S. Geological Survey Professional Paper 713, U.S. Geological Survey, Washington, D.C. ———. (1977). “Trace element sorption by sediments and soils—sites and processes,” in Symposium on molybdenum in the environment, Vol. 2, W. Chappell, and K. Peterson, eds., Marcel-Dekker, Inc., New York, 425–553. Jensen, S., and Jernolov, A. (1969). “Biological methylation of mercury in aquatic organisms.” Nature, 223, 753–754. Joreskog, K. G., Klovan, J. E., and Reyment, R. A. (1976). Geological factor analysis, Elsevier Scientific Publishing, New York. Kabata-Pendias, A. and Pendias, D. C. (1992). Trace elements in soils and plants, CRC Press, Boca Raton, Fla. Keren, R., Bingham, F. T., and Rhoades, J. D. (1985). “Plant uptake of boron as affected by boron distribution between liquid and solid phases in soil.” Soil Sci. Soc. Am. J., 49, 297–302. Keren, R., Gast, R. G., and Bar-Yosef, B. (1981). “pH-dependent boron adsorption by Na-montmorillonite.” Soil Sci. Soc. Am. J., 45, 45–48. Keshavarz, P., Malakouti, M. J., Karimian, N., Fotovat, A. (2006). “The effects of salinity on extractability and chemical fractions of zinc in selected calcareous soils of Iran.” J. Agric. Sci. Technol., 8, 181–190. Khoshgoftar, A. H., Shariatmadari, H., Karimian, N., Kalbasi, M., van der Zee, S. E. A. T. M., and Parker, D. R. (2004). “Salinity and zinc application effects on phytoavailability of cadmium and zinc.” Soil Sci. Soc. Am. J., 68, 1885–1889. Khoshgoftarmanesh, A. H., Shariatmadari, H., Karimian, N., Kalbasi, M., van der Zee, S. E. A. T. M. (2006). “Cadmium and zinc in saline soil solutions and their concentrations in wheat.” Soil Sci. Soc. Am. J., 70, 582–589. Kinniburgh, D. G., Jackson, M. L., and Syers, J. K. (1976). “Adsorption of alkaline earth, transition, and heavy metal cations by hydrous oxide gels of iron and aluminum.” Soil Sci. Soc. Am. J., 40, 796–799. Kuo, S., and Mikkelsen, D. S. (1979). “Zinc adsorption by two alkaline soils.” Soil Sci., 128, 274–279. Lakin, H. W. (1973). “Selenium in our environment,” in Trace elements in the environment, E. L. Kothy, ed., Advances in Chemistry Series 123, American Chemical Society, Washington, D.C., 96–111. Langmuir, I. (1918). “The adsorption of gases on plane surfaces of glass, mica, and platinum.” J. Am. Chem. Soc., 40, 1361–1403. Langmuir, D., and Mahoney, J. (1984). “Chemical equilibrium and kinetics of geochemical processes in ground water studies,” in First Canadian/American conference on hydrogeology, B. Hitchon and E. I. Wallick, eds., National Water Well Association, Worthington, Ohio, 69–95. Leckie, J. O., Benjamin, M. M., Hayes, K., Kaufman, G., and Altmann, S. (1980). Adsorption/coprecipitation of trace elements from water with iron oxy-hydroxide,

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

131

Electric Power Research Institute Report CS-1513, Stanford University, Stanford, Calif. Lemmo, N. V., Faust, S. D., Belton, T., and Tucker, R. (1983). “Assessment of the chemical and biological significance of arsenical compounds in a heavily contaminated watershed. Part I. The fate and speciation of arsenical compounds in aquatic environments: A literature review.” J. Environ. Sci. Health, 18, 335–387. Lerman, A., and Childs, C. W. (1973). “Metal-organic complexes on natural waters,” in Trace metals and metal-organic interactions in natural waters, P. C. Singer, ed., Ann Arbor Science Press, Ann Arbor, Mich., 201–235. Lewis, F. M., Voss, C. I., and Rubin, J. (1986). Numerical simulation of advective-dispersive multisolute transport with sorption, ion-exchange, and equilibrium chemistry, U.S. Geological Survey, Water-Resources Investigations Report 86-4022, U.S. Geological Survey, Washington, D.C. Lindberg, R. D., and Runnells, D. D. (1984). “Ground water redox reactions: An analysis of equilibrium state applied to Eh measurements and geochemical modeling.” Science, 225, 925–927. Lindsay, W. L. (1979). Chemical equilibria in soils, Wiley-Interscience, New York. Loyaux-Lawniczak, S., Lecomte, P., Ehrhardt, J.-J. (2001). “Behavior of hexavalent chromium in a polluted groundwater: Redox processes and immobilization in soils.” Environ. Sci. Technol., 35, 1350–1357. Manning, B. A., and Goldberg, S. (1996). “Modeling arsenate competitive adsorption on kaolinite, montmorillonite, and illite.” Clays Clay Miner., 44, 609–623. Masscheleyn, P. H., Delaune, R. D., and Patrick, W. H. (1991). “Effect of redox potential and pH on arsenic speciation and solubility in a contaminated soil.” Environ. Sci. Technol., 25, 1414–1419. Mattigod S. V., and Page, A. L. (1983). “Assessment of metal pollution in soils,” in Applied environmental geochemistry, I. Thornton, ed., Academic Press, London, 355–391. McBride, M. B. (1979). “Mobility and reactions of VO2 on hydrated smectite surfaces.” Clays Clay Miner., 27, 91–96. McBride, M., Sauvé, S., and Hendershot, W. (1997). “Solubility control of Cu, Zn, Cd and Pb in contaminated soils.” Eur. Soil Sci., 48, 337–346. McGrath, S. P., Chaudri, A. M., and Gille, K .E. (1995). “Long-term effects of metals in sewage sludge on soils, microorganisms and plants.” J. Ind. Microbiol., 14, 94–104. McLaughlin, M. J., Tiller, K. G., Smart, M. K. (1997). “Speciation of cadmium in soil solutions of saline/sodic soils and relationship with cadmium concentrations in potato tubers (Solanum tuberosum L.).” Aust. J. Soil Res., 35, 183–198. McNeal, J. M., Severson, R. C., and Gough, L. P. (1985). “The occurrence of extractable elements in soils from the Northern Great Plains.” Soil Sci. Soc. Am. J., 49, 873–881. Mesuere, K., and Fish, W. (1992). “Chromate and oxalate adsorption on goethite. 1. Calibration of surface complexation models.” Environ. Sci. Technol., 28, 2357–2364. Motta, M. M., and Miranda, C. F. (1989). “Molybdate adsorption on kaolinite, montmorillonite, and illite: Constant capacitance modeling.” Soil Sci. Soc. Am. J., 53, 380–385.

132

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Naftz, D. L. (1996). “Pattern recognition analysis and classification modeling of selenium-producing areas.” J. Chemometrics, 10, 309–324. National Academy of Sciences. (1974). Vanadium, National Research Council, Washington, D.C. Neal, R. H., and Sposito, G. (1989). “Selenate adsorption on alluvial soils.” Soil Sci. Soc. Am. J., 53, 70–74. Neal, R. H., Sposito, G., Holtzclaw, K. M. and Traina, S. J. (1987a). “Selenite adsorption on alluvial soils: I. Soil composition and pH effects.” Soil Sci. Soc. Am. J., 51, 1161–1165. ———. (1987b). “Selenite adsorption on alluvial soils: II. Solution composition effects.” Soil Sci. Soc. Am. J., 51, 1165–1169. Nordstrom, D. K., and Ball, J. W. (1984). “Chemical models, computer programs and metal complexation in natural waters,” in Complexation of trace metals in natural waters, C. J. M. Kramer and J. C. Duinker, eds., Martinus and Nijoff/W. Jurk, The Hague. Nordstrom, D. K., Jenne, E. A., and Ball, J. W. (1979a). “Redox equilibria of iron in acid mine waters,” in Chemical modeling in aqueous systems, E. A. Jenne, ed., American Chemical Society, Washington, D.C., 51–79. Nordstrom, D. K., Plummer, L. N., Wigley, T. M. L., Wolery, T. J., Ball, J. W., Jenne, E. A., Bassett, R. L., Crerar, D. A., Florence, T. M., Fritz, B., Hoffman, M., Holdren, Jr, G. R., Lafon, G. M., Mattigod, S. V., McDuff, R. E., Morel, F., Reddy, M. M., Sposito, G., and Thraikill, J. (1979b). “A comparison of computerized chemical models for equilibrium calculations in aqueous systems,” in Chemical modeling in aqueous systems, E. A. Jenne, ed., American Chemical Society, Washington, D.C., 857–892. Norvell, W. A., Wu, J., Hopkins, D. G., and Welch, R. M. (2000). “Association of cadmium in durum wheat grain with soil chloride and chelate-extractable soil cadmium.” Soil Sci. Soc. Am. J., 64, 2162–2168. O’Connor, G. A., Brobst, R. B., Chaney, R. L., Kincaid, R. L., McDowell, L. R., Pierzynski, G. M., Rubin, A., and Van Riper, G. G. (2001). “A modified risk assessment to establish molybdenum standards for land application of biosolids.” J. Environ. Qual., 30, 1490–1507. Olson, M. L., and DeWild, J. F. (1999). Techniques for the collection and species-specific analysis of low levels of mercury in water, sediment, and biota, U.S. Geological Survey Water Resources Investigations Report 99-4018-B, U.S. Geological Survey, Washington, D.C. Oze, C., Bird, D. K., and Fendorf, S. (2007). “Genesis of hexavalent chromium from natural sources in soils and groundwater.” Proc., Nat. Acad. Sci. USA, 104, 6544–6549. Peacock, C. L., and Sherman, D. M. (2004). “Vanadium(V) adsorption onto goethite ( -FeOOH) at pH 1.5 to 12: A surface complexation model based on ab initio molecular geometries and EXAFS spectroscopy.” Geochim. Cosmochim. Acta, 68, 1723–1733. Peak, D., and Sparks, D. L. (2002). “Mechanisms of selenate adsorption on iron oxides and hydroxides.” Environ. Sci. Technol., 36, 1460–1466. Peryea, F. J., Bingham, F. T., and Rhoades, J. D. (1985). “Mechanisms of boron regeneration.” Soil Sci. Soc. Am. J., 49, 840–843.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

133

Peryea, F. J., and Creger, T. L. (1994). “Vertical distribution of lead and arsenic in soils contaminated with lead arsenate pesticide residues.” Water, Air Soil Pollut., 78, 297–306. Pierce, M. L., and Moore, C. B. (1980). “Adsorption of arsenite on amorphous iron hydroxide from dilute aqueous solution.” Environ. Sci. Technol., 14, 214–216. Ponnamperuma, F. N. (1972). “The chemistry of submerged soils.” Advan. Agron., 24, 29–96. Ponthieu, M., Juillot, F., Hiemstra, T., van Riemsdijk, W. H., and Benedetti, M. F. (2006). “Metal binding to iron oxides.” Geochim. Cosmochim. Acta, 70, 2679– 2698. Posselt, H. S., Anderson, F. J., and Weber, W. J. (1978). “Cation sorption on colloidal hydrous manganese dioxide.” Environ. Sci. Technol., 2, 1087–1093. Preeda, P., Leong, S. T., Laortanakul, P., Torotoro, J. L. (2002). “Influence of salinity and acidity on bioavailability of sludge-borne heavy metals. A case study of Bangkok municipal sludge.” Water, Air, Soil Pollut., 139, 43–60. Presser, T. S., and Ohlendorf, H. M. (1987). “Biogeochemical cycling of selenium in the San Joaquin Valley, California.” Environ. Mgmt., 11, 805–821. Presser, T. S. and Swain, W. C. (1990). “Geochemical evidence for selenium mobilization by weathering of pyretic shale, San Joaquin Valley, California, USA.” Appl. Geochem., 5, 703–717. Presser, T. S., Sylvester, M. A., and Low, W. H. (1994). “Bioaccumulation of selenium from natural geologic sources in western states and its potential consequences.” Environ. Mgmt., 18, 423–436. Quirk, J. P., and Posner, A. M. (1976). “Trace element adsorption by soil minerals,” in Trace elements in soil-plant-animal systems, D. J. D. Nicholas and A. R. Egan, eds., Academic Press, New York. Rai, D., and Zachara, J. M. (1984). Chemical attenuation rates, coefficients, and constants in leachate migration, a critical review, Electric Power Research Institute Report, EA-3356, Volume 1, Electric Power Research Institute, Palo Alto, Calif. ———. (1986). Geochemical behavior of chromium species, Electric Power Research Institute Report EA-4544, Electric Power Research Institute, Palo Alto, Calif. Raven, K. P., Jain, A., and Loeppert, R. H. (1998). “Arsenite and arsenate adsorption on ferrihydrite: Kinetics, equilibrium, and adsorption envelopes.” Environ. Sci. Technol., 32, 344–349. Reddy, C. N., and Patrick, W. H. (1977). “Effect of redox potential on the stability of zinc and copper chelates in flooded soils.” Soil Sci. Soc. Am. J., 41, 729–732. Reddy, K. R., and Patrick, W. H. (1983). “Effect of aeration on reactivity and mobility of soil constituents,” in Chemical mobility and reactivity in soil systems, D. W. Nelson et al., eds., Soil Science Society America, Madison, Wisc., 11–13. Reyes, E. D., and Jurinak, J. J. (1967). “A mechanism of molybdate adsorption in Fe2O3. Soil Sci. Soc. Am. Proc., 31, 637–641. Rietra, R. P. J. J., Hiemstra, T., and van Riemsdijk, W. H. (2001). “Comparison of selenate and sulfate adsorption on goethite.” J. Colloid Interface Sci., 240, 384–390. Robertson, F. N. (1975). “Hexavalent chromium in the ground water in Paradise Valley, Maricopa County, Arizona.” Ground Water, 13, 516–527.

134

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Ross, D. S., Sjogren, R. E., and Bartlett, R. J. (1981). “Behavior of chromium in soils: IV. Toxicity to microorganisms.” J. Environ. Qual., 10, 145–148. Ryden, J. C., Syers, J. K., and Tillman, R. W. (1987). “Inorganic anion sorption and interaction with phosphate sorption by hydrous ferric oxide gel.” J. Soil Sci., 38, 211–217. Saripalli, K. P., McGrail, B. P., and Girvin, D. C. (2002). “Adsorption of molybdenum on to anatase from dilute aqueous solutions.” Appl. Geochem., 17, 649–656. Sarkar, D., Essington, M. E., and Misra, K. C. (1999). “Adsorption of mercury(II) by variable charge surfaces of quartz and gibbsite.” Soil Sci. Soc. Am. J, 63, 1626–1636. ———. (2000). “Adsorption of mercury(II) on kaolinite.” Soil Sci. Soc. Am. J., 64, 1968–1975. Sauvé, S., Hendershot, W., and Allen, H. E. (2000). “Solid-solution partitioning of metals in contaminated soils: Dependence on pH, total metal burden, and organic matter.” Environ. Sci. Technol., 34, 1125–1131. Schindler, P. W. (1981). “Surface complexes at oxide-water interfaces,” in Adsorption of inorganics at solid–liquid interfaces. M. A. Anderson and A. J. Rubin, eds., Ann Arbor Science Press, Ann Arbor, Mich., 1–50. Schwertmann, U., and Pfab, G. (1996). “Structural vanadium and chromium in lateritic iron oxides: Genetic implications.” Geochim. Cosmochim. Acta, 60, 4279–4283. Sindeeva, N. D. (1964). Mineralogy and types of deposits of selenium and tellurium, Wiley-Interscience, New York. Singh, S. P. N., and Mattigod, S. V. (1992). “Modeling boron adsorption on kaolinite.” Clays Clay Miner., 40, 192–205. Sistani, K. R. and Novak, J. M. (2006). “Trace metal accumulation, movement, and remediation in soils receiving animal manure,” in Trace elements in the environment, M. N. V. Prasad, K. S. Sajwan, and R. Naidu, eds., CRC Press, Boca Raton, Fla. Smith, G. J., and Anders, V. P. (1989). “Toxic effects of boron on mallard reproduction.” Environ. Toxicol. Chem., 8, 943–950. Sposito, G. (1981). The thermodynamics of soil solutions, Oxford University Press, New York. ———. (1983). “The chemical forms of trace metals in soils,” in Applied environmental geochemistry, I. Thornton, ed., Academic Press, London, 123–170. ———. (1984). The surface chemistry of soils, Oxford University Press, New York. Sposito G., de Wit, J. C. M., and Neal, R. H. (1988). “Selenite adsorption on alluvial soils: III. Chemical modeling.” Soil Sci. Soc. Am. J., 52, 947–950. Stachowicz, M., Hiemstra, T., and van Riemsdijk, W. H. (2006). “Surface speciation of As(III) and As(V) in relation to charge distribution.” J. Colloid Interface Sci., 302, 62–75. Stanin, F. T. (2005). “The transport and fate of chromium (VI) in the environment,” in Chromium (VI) handbook, J. Guertin, J. A. Jacobs, and C. P. Avakian, eds., CRC Press, Boca Raton, Fla. Steenhuis, T. S., McBride, M. B., Richards, B. K., and Harrison, E. (1999). “Trace metal retention in the incorporation zone of land-applied sludge.” Environ. Sci. Technol., 33, 1171–1174.

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

135

Stollenwerk, K. G., and Grove, D. B. (1985a). “Adsorption and desorption of hexavalent chromium in an alluvial aquifer near Telluride, Colorado.” J. Environ. Qual., 14, 150–155. ———. (1985b). “Reduction of hexavalent chromium in water samples acidified for preservation.” J. Environ. Qual., 14, 396–399. Stumm, W., Kummert, R., and Sigg, L. (1980). “A ligand exchange model for the adsorption of inorganic and organic ligands at hydrous oxide interfaces.” Croatica Chem. Acta, 53, 291–312. Stumm, W., and Morgan, J. J. (1981). Aquatic chemistry, 2nd ed., John Wiley and Sons, Inc., New York. Su, C., and Suarez, D. L. (2004). “Boron release from weathering of illites, serpentine, shales, and illitic/palygorskitic soils.” Soil Sci. Soc. Am. J., 68, 96–105. Syers, J. K., Browman, M. G., Smillie, G. W., and Corey, R. B. (1973). “Phosphate sorption by soils evaluated by the Langmuir adsorption equation.” Soil Sci. Soc. Amer. Proc., 37, 358–363. Sylvester, M. A., Deason, J. P., Feltz, H. R., and Engber, R. A. (1988). “Preliminary results of the Department of the Interior irrigation drainage studies.” Proc. on Planning Now for Irrigation and Drainage, Irrigation and Drainage Div., ASCE, Lincoln, Nebraska, August 1988, ASCE, Reston, Va. Taylor, R. M., and Giles, J. B. (1970). “The association of vanadium and molybdenum with iron oxides in soils.” J. Soil Sci., 21, 203–215. Thorstensen, D. C. (1984). The concept of electron activity and its relation to redox potentials in aqueous geochemical systems, U.S. Geological Survey Open-File Report 84–072, U.S. Geological Survey, Washington, D.C. Tidball, R. R., Grundy, W. D., and Swatzky, D. L. (1986a). “Kriging techniques applied to element distribution in soils of the San Joaquin Valley, California.” Proc., HAZTECH International Conference, Denver, Colorado, August 1986, Haztech International, Englewood, Colo., 992–1009. Tidball, R. R., Severson, R. C., Gent, C. A. and Riddle, G. O. (1986b). Element associations in soils of the San Joaquin Valley, California, U.S. Geological Survey OpenFile Report, 86-583, U.S. Geological Survey, Washington, D.C. Tiesta, S. M. (2005). “Sources of chromium contamination in soil and groundwater,” in Chromium (VI) handbook, J. Guertin, J. A. Jacobs, and C. P. Avakian, eds., CRC Press, Boca Raton, Fla. Tiffreau, C., Lützenkirchen, J., and Behra, P. (1995). “Modeling the adsorption of mercury(II) on (hydr)oxides. I. Amorphous iron oxide and -quartz.” J. Colloid Interface Sci., 172, 82–93. Tourtelot, H. A., and Miesch, A. T. (1975). “Sampling designs in environmental geochemistry.” Geol. Soc. Amer. Spec. Paper, 155, 107–118. Turner, M. A., and Rust, R. H. (1971). “Effect of chromium on growth and mineral nutrition of soybeans.” Soil Sci. Soc. Am. Proc. 35, 755–758. U.S. Environmental Protection Agency (EPA). (1986). Quality criteria for water, 440/5-86-001, U.S. EPA, Washington, D.C. U.S. Geological Survey (USGS). (2011). “USGS Mercury Research Lab,” http:// infotrek.er.usgs.gov/doc/mercury/methods.html, accessed January 15, 2011. Usman, A. R. A., Kuzyakov, Y., Stahr, K. (2005). “Effect of immobilizing substances and salinity on heavy metals availability to wheat grown on sewage sludge-contaminated soil.” Soil and Sed. Contam. Int. J., 14, 329–344.

136

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

van Olphen, H. (1977). An introduction to clay colloid chemistry, John Wiley and Sons, New York. Venema, P., Hiemstra, T., and van Riemsdijk, W. H. (1996). “Multisite adsorption of cadmium on goethite.” J. Colloid Interface Sci., 183, 515–527. Wedepohl, K. H. (1972). “Vanadium,” in Handbook of geochemistry, K. H. Wedepohl, ed., Springer-Verlag, Berlin, 23-A-1 to 23-O-1. Weerasooriya, R., and Tobschall, H. J. (2000). “Mechanistic modeling of chromate adsorption onto goethite.” Colloids Surf. A: Physicochem. Eng. Aspects, 162, 167–175. Weerasooriya, R., Tobschall, H. J., Wijesekara, H. K. D. K., Arachchige, E. K. I. A. U. K., and Pathirathne, K. A. S. (2003). “On the mechanistic modeling of As(III) adsorption on gibbsite.” Chemosphere, 51, 1001–1013. Weerasooriya, R., Tobschall, H. J., Wijesekara, and Bandara, A. (2004). “Macroscopic and vibration spectroscopic evidence for specific bonding of arsenate on gibbsite.” Chemosphere, 55, 1259–1270. Weggler, K., McLaughlin, M. J., Graham, R. D. (2004). “Effect of chloride in soil solution on the plant availability of biosolid-borne cadmium.” J. Environ. Qual., 33, 496–504. Wehrli, B., and Stumm, W. (1989). “Vanadyl in natural waters: Adsorption and hydrolysis promote oxygenation.” Geochim. Cosmochim. Acta, 53, 69–77. Welch, A. H., Lico, M. S., and Hughes, J. L. (1988). “Arsenic in ground water of the western United States.” Ground Water, 26, 333–347. Welch, A. H., Westjohn, D. B., Helsel, D. R., and Wanty, R. B. (2000). “Arsenic in ground water of the United States: Occurrence and geochemistry.” Ground Water, 38, 589–604. Weng, C. H., Wang, J. H., and Huang, C. P. (1997). “Adsorption of Cr(VI) onto TiO2 from dilute aqueous solutions.” Water Sci. Tech., 35, 55–62. White, A. F., Benson, S. M., Yee, A. W., Wallenberg, H. A., Flexser, S. (1991). “Groundwater contamination at the Kesterson Reservoir, California: 2. Geochemical parameters influencing selenium mobility” Water Resour. Res., 27, 1085–1098. Zachara, J. M., Ainsworth, C. C., Cowan, C. E., and Resch, C. T. (1989). “Adsorption of chromate by subsurface soil horizons.” Soil Sci. Soc. Am. J., 53, 418–428.

NOTATION ai  activity of ion i b  adsorption maximum C  equilibrium concentration of adsorbate IAP  ion-activity product k  parameter that reflects affinity of adsorbent for adsorbate K, n  constants (empirically evaluated) Ks  stability constant Kso  solubility product constant

CHEMISTRY OF TRACE ELEMENTS IN SOILS AND GROUNDWATER

m  mass of solid phase adsorbent mi  molal concentration of species i SI  saturation index x  mass of trace element adsorbed i  activity coefficient of species i

137

This page intentionally left blank

CHAPTER 5 SOIL RESPONSE TO SALINE AND SODIC CONDITIONS I. Shainberg and M. J. Singer

INTRODUCTION Saline and sodic conditions reduce the value and productivity of soils (Sumner and Naidu 1998) and affect their physical behavior (Dexter 2004; Oster 1994). Accumulation of sodium in the soil solution and the exchange phase affects soil physical properties, such as structural stability, hydraulic conductivity (HC), infiltration rate (IR), runoff, and erosion (Bronick and Lal 2005; Shainberg and Letey 1984; Suarez et al. 2006). Many reviews have been published on the response of the physical properties of soils to sodicity and salinity (e.g., Levy et al. 1998; Levy 1999; Levy and Shainberg 2004; Shainberg and Levy 2004; Sumner 1993; Sumner and Stewart 1992; Sumner and Naidu 1998). These reviews demonstrate that soil properties, such as soil texture, clay mineralogy, pH, sesquioxides, organic matter (OM), and lime content affect the response of soils to sodic conditions. Soils with high clay content, a high portion of 2⬊1 clay minerals, high pH, and low sesquioxides, lime, and OM content are most susceptible to sodic conditions. However, temporal changes in extrinsic physical conditions, such as wetting rate (WR), antecedent moisture content (AMC), and ageing (time since wetting), also affect the response of soils to sodic conditions and the effect of these conditions have been only partially considered (Levy and Shainberg 2004; Mamedov et al. 2001; Mamedov et al. 2002; Ruiz-Vera and Wu 2006). This chapter describes the various ways in which salinity and sodicity affect soils and the factors that influence these effects. An underlying focus of this chapter is the evaluation of salinity and sodicity and their impacts under actual field conditions. Most evaluations of the effect of soil sodicity on soil’s physical properties were based on 139

140

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

laboratory results on dry, disturbed soil samples, which were packed in columns or trays and then exposed to fast wetting rates. Immediately following wetting (without waiting for some ageing), the soil samples were leached either by flooding or by high-intensity (20 mm/h) simulated rain storms (Bagarello et al. 2006; Levy and Shainberg 2004). Under these conditions, soil physical properties were very susceptible to the effects of sodicity. However, these conditions do not prevail in the field where slower wetting rate and ageing at different AMCs decrease the susceptibility of soils to sodic conditions. One of the objectives of this chapter is to demonstrate the effect of inherent soil properties and time-dependent physical conditions on the susceptibility of soils to sodic conditions. Salinity and sodicity affect soil structure, which must be stable for adequate permeability and water infiltration. High sodium levels combined with low soil-water salinity can lower the soil’s permeability and decrease its infiltration capacity through the swelling and dispersion of clays and the slaking of aggregates. Swelling of clays narrows the conducting pores in soils (McNeal and Coleman 1966; Quirk and Schofield 1955). Dispersion and deposition of clays in the narrow necks of conducting pores also can reduce soil permeability (Frenkel et al. 1978; Shainberg et al. 1981a,b). Aggregate breakdown can reduce both the amount of macropores in a soil and the soil’s HC (Abu-Sharar et al. 1987). Aggregate breakdown by water may result from a variety of physical and physicochemical mechanisms (Falsone and Bonifacio 2006). Four main mechanisms are (1) slaking, that is, breakdown caused by compression of entrapped air during fast wetting (Panabokke and Quirk 1957), (2) breakdown by differential swelling during fast wetting (Panabokke and Quirk 1957), (3) breakdown by raindrop impact (McIntyre 1958; Ramos et al. 2003), and (4) physicochemical dispersion due to osmotic stress upon wetting with low-electrolyte water (Emerson 1977). These mechanisms differ in the type of energy involved in aggregate disruption. For example, swelling can overcome attractive pressures in the magnitude of MPa (Rengasamy and Olsson 1991), while slaking and impact of raindrops can overcome attractive pressures in the range of kPa only (Rengasamy and Sumner 1998). This chapter reviews the effect of sodicity and electrolyte concentration on swelling and dispersion of soil’s clay, the effect of wetting rate and ageing on soil’s aggregates slaking, and the combined effect of clay dispersion and aggregate slaking on consequent changes in soil hydraulic properties. Aggregate stability is the measure of the resistance of aggregates to breakdown (Amezketa et al. 1996). Kemper and Koch (1966) showed this stability to be a function of organic matter, clay, and oxide contents. In subsequent papers, Kemper and Rosenau (1984) and Kemper et al. (1987) showed that aggregate stability and soil permeability to water depend on both soil exchangeable sodium percentage (ESP) and the salt concentration of the percolating solution. Levy and Mamedov (2002) studied aggre-

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

141

gate resistance to slaking of six soils differing in clay content. Results showed that for coarse- and medium-textured soils (25% clay), aggregate stability was relatively low and unaffected by sodicity and salinity of the irrigation water. Conversely, in the fine-textured soils (38% clay), aggregate stability was relatively high because the clay acted as a cementing and binding agent in the soil (Kemper and Koch 1966). Stability of the fine-textured aggregates decreased with the increase in soil ESP (Levy et al. 2005). Evidently, in clay soils where aggregates are a priori stable, conditions favoring dispersive behavior of clay (high sodicity) adversely affect aggregate resistance to slaking, and conditions favoring aggregate slaking (e.g., high wetting rate) increase the susceptibility of soils to sodic conditions. Amendments, such as gypsum (or phosphogypsum-PG) and polyacrylamide (PAM), have been used to prevent seal formation, runoff, and erosion (Flanagan et al. 1997a,b; Sojka et al. 2007; Tang et al. 2006; Yu et al. 2003). Gypsum is effective because, upon dissolution, gypsum releases electrolytes into the rainwater (the electrolyte effect) and because dissolved calcium ions displace Na ions from the exchange complex—the reclamation effect (Keren and Shainberg 1981). Laboratory and field studies with anionic PAM demonstrated that addition of small amounts of PAM (10–20 kg ha1) to the soil surface were effective in maintaining high permeability and decreasing runoff and soil erosion levels, especially when the PAM was applied with gypsum (Flanagan et al. 1997a,b; Shainberg et al. 1990). The response of sodic soils to surface application of PAM and gypsum is also reviewed in this chapter. SOLUTION COMPOSITION AND CLAY SWELLING AND DISPERSION Colloidal clay is the soil fraction that most decisively determines soil physical behavior. It possesses the greatest specific surface area and is, therefore, most active in physicochemical processes, such as swelling and dispersion. Because of its extremely high specific surface area (800 m2/g), smectite, the dominant clay in semiarid and arid regions, is the most active clay in clay-solution interactions. Sodium and Calcium Distribution in Reference Clays Smectitic clay Swelling and dispersion of clays are the primary processes responsible for the degradation of the physical properties of soils in the presence of Na. The effect of sodicity on clay swelling and dispersion in suspension has been reviewed extensively (e.g., Levy et al. 1998; Shainberg and Letey

142

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

1984; van Olphen 1977). The diffuse double layer at the clay surfaces consists of the lattice negative charge and compensating counter ions. Divalent cations are attracted to the surface with a force twice as great as that of monovalent ions. The greater the compression of the ionic atmosphere, the smaller the repulsion forces between clay platelets are. Because of the low repulsion in Ca-smectite, the platelets condense into quasi-crystals consisting of four to nine clay platelets in parallel with interplatelet distance of 0.9 nm (Shainberg and Letey 1984). In a mixture of Na and Ca cations, “demixing” of the cations occurs and Na ions concentrate mainly on the external surfaces of the quasi-crystals (Oster et al. 1980; Shainberg and Letey 1984). This model explains swelling, dispersion, and flocculation in Na-Ca smectite systems (Shainberg and Letey 1984). A small increase in ESP in smectite suspensions has a considerable effect on the electrophoretic mobility of Ca-smectite (curve A in Fig. 5-1) and a negligible effect on the size of the quasi-crystals as determined by the slope of the viscosity equation (curve B in Fig. 5-1). Because clay dispersion and flocculation values are related to the electrophoretic mobility, smectite dis-

FIGURE 5-1. Dependence of electrophoretic mobility (curve A) and relative size (curve B) of smectite particles on the exchangeable sodium percentage (ESP). Relative size is expressed in terms of the slope of Einstein’s equation for the viscosity of the suspension. From Shainberg and Letey (1984).

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

143

persion also increased sharply with a small increase in sodium content (Fig. 5-2). Conversely, swelling of smectite as a function of ESP in Na-Ca systems changed in a way similar to the change in clay size (Fig. 5-1) and was only slightly affected by increases in ESP up to 15 and a steep increase in swelling above ESP 20 (Shainberg and Letey 1984). The “demixing” model explains why in smectite even a small increase in ESP has a considerable effect on clay dispersion and a small effect on clay swelling. In low-ESP systems, Na ions concentrate on the external surfaces of the quasi-crystals. A small increase in total ESP sharply increases Na concentration at the external surfaces of the quasi-crystals and clay dispersion increases significantly. At the same time, in the low-ESP range, the size of the quasi-crystals and the area of osmotic active external surface are not affected by an increase in ESP, and thus swelling is unaffected. The main factor that determines the susceptibility of clays to sodicity is the electrolyte concentration in the soil solution. Whereas in electrolyte concentration 3 mmolc /L, a low sodicity level (ESP 5) may cause clay dispersion (Fig. 5-2), in moderate electrolyte concentration levels

FIGURE 5-2. Flocculation values for Wyoming smectite and Fithian illite as a function of ESP. From Shainberg and Letey (1984).

144

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(e.g., 10 mmolc /L) even high sodicity [e.g., sodium absorption ratio (SAR) and ESP 40] causes only little dispersion (Fig. 5-2). Similarly, Mace and Amrhein (2001) found increasing clay dispersion in columns leached with solutions below 5 mmolc L1 and SAR from 1 to 8. In arid regions, intermittent rain may lower the electrolyte concentration sufficiently to promote clay dispersion even in soils with low sodicity. However, many sodic soils may contain lime, which upon dissolution may maintain electrolyte concentration of 3 to 4 mmolc /L (Shainberg and Letey 1984), and this concentration is sufficient to counter the deleterious effect of sodicity below ESP 5. Effect of smectite mineralogy on its response to sodicity Smectite mineralogy also affects the response of smectites to sodicity (Alperovitch et al. 1985 and Fig. 5-3). Alperovitch et al. (1985) studied the effect of sodicity on the HC of reference clay-sand mixtures (Fig. 5-3). Reference smectite (Otay, California) with high charge (1.2 mmolc/g) and relatively low specific surface area (552 m2 /g) was found to swell less than Wyoming smectite with low charge (0.9 mmolc /g) and high surface area (800 m2 /g). The y-intercepts in Fig.5-3 represent the effect of swelling on relative HC in 10 mmolc/L solution. Much clay swelling in 10 mmol/L solution results in a low intercept, and low swelling in 10 mmol/L solution results in high initial HC. Also, Otay smectite was much less susceptible to sodicity compared with Wyoming smectite (Fig. 5-3). The two other smectites with intermediate charge densities (Belle Fourche and Polkville) also showed intermediate susceptibility to ESP (Fig. 5-3). Illite clay was the clay that swelled the least (the y-intercept) and was least sus-

FIGURE 5-3. Effect of ESP on relative hydraulic conductivity (HC) of clay minerals/sand mixtures equilibrated with 10 mmolc /L solutions and leached with distilled water. From Alperovitch et al. (1985), Fig. 4; reproduced with kind permission of The Clay Minerals Society, publisher of Clays and Clay Minerals.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

145

ceptible to sodic conditions (Fig. 5-3). It can be expected that soils with high-charge smectitic clays will be less susceptible to sodic conditions than soils with smectitic clay similar to Wyoming (low-charge/high surface area) smectite. Swelling and dispersion in illite and kaolinite clays Demixing does not take place in illitic and kaolinitic clays and, as a result, dispersion/flocculation as a function of ESP follows a simple linear relation (dashed line in Fig. 5-2). At low ESP, illite is more dispersed and does not swell as much as smectite. The low osmotic active surface area of illite (120 m2/g) explains the low swelling. Because illite particles have irregular surfaces and the planar surfaces are terraced, edge-to-surface attraction bonds are weak and pronounced dispersion occurs. Thus, illite is more dispersive than smectite. Pure Na-kaolinite is not dispersive at pH 7 because of the attraction forces between the positive charge at the edges and the negative charges on the planar surfaces (Schofield and Samson 1954). At pH 7, Nakaolinite flocculation occurs even in the absence of salt. However, when pH 8, the edges of kaolinite become negatively charged, edge-to-face attraction does not occur, and kaolinite may disperse with an increase in ESP (Lado et al. 2007). Because of the low osmotic active surface area of kaolinite (15 m2/g), the clay does not swell much, and there is a negligible effect of ESP on clay swelling (Lado et al. 2007). Differences of opinion exist in the literature regarding the effect of kaolinite clay on the swelling and HC of soils leached with sodic water. McNeal and Coleman (1966) concluded that the most labile soils were those high in 2⬊1 layer silicates, especially montmorillonite, and the least labile were those high in kaolinite and sesquioxides. However, kaolinitic soils containing small amounts of montmorillonite and mica were found to be dispersive (Frenkel et al. 1978; Schofield and Samson 1954). Frenkel et al. (1978) concluded that, although kaolinitic soils were less sensitive than montmorillonitic soils at low electrolyte concentrations, their HC was reduced markedly, even at ESP values 10, when the kaolinitic was mixed with smectite.

SOLUTION COMPOSITION AND THE HYDRAULIC CONDUCTIVITY OF SODIC SOILS Effect of Electrolyte Concentration on the Hydraulic Conductivity of Sodic Soils Permeability of soils to water depends on the ESP of the soil and the electrolyte concentration of the percolating solution (Quirk and Schofield

146

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

1955). Soil permeability can be maintained, even at high ESP values, provided that the electrolyte concentration of the water is above a critical (threshold) level (Oster 1994). Soils responded differently to the same combination of electrolyte concentration and ESP; thus, a unique threshold concentration exists for each soil (Fig. 5-4). Soil properties significantly affect the response of soil permeability to Na and electrolyte concentration. However, in discussing this response, a separation should be made between percolating solutions with low concentration (e.g., rain or snow water) and irrigation water with electrolyte concentration 3mmolc/L. This separation is justified by the mechanisms responsible for the HC

FIGURE 5-4. Combinations of salt concentration and ESP required to produce a 25% reduction in HC for selected soils. From McNeal and Coleman (1966) with permission from the Soil Science Society of America.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

147

decrease. Clay swelling is the main mechanism responsible for HC deterioration in electrolyte concentration 3mmolc/L, and clay dispersion is the main mechanism in soils leached with electrolyte concentration of soil solutions below 3 mmolc/L (Shainberg and Letey 1984). Soil Properties Affecting the Response of Sodic Soils to Irrigation with Moderate Electrical Conductance Soil texture McNeal et al. (1968) and Frenkel et al. (1978) showed that HC reductions caused by increases in ESP or decreases in electrolyte concentration were greater for soils having higher clay content. This effect was in addition to the decreases in absolute HC associated with increases in clay content. Also, the effect of prewetting rate on the HC of soils was more pronounced in a soil with a clay content of 70% compared to one with 46.5% clay, and the effect of sodicity increased with increase in clay content and prewetting rate (Moutier et al. 1998). Clay mineralogy McNeal et al. (1968) concluded that the most labile soils were those high in 2:1 layer silicates, especially smectites, and the least labile were those high in kaolinite and sesquioxides. McNeal and Colman (1966) found a good correlation between relative soil permeability and the swelling of extracted clays, suggesting that clay swelling was responsible for HC deterioration. Frenkel et al. (1978) found that the ESP and the electrolyte concentration less affected HC values of kaolinitic soils than the HC of smectitic soils. They also found that the effect of ESP increased with increase in soil’s bulk density. Mixing a small percentage of smectite (2%) with kaolinitic soil increased significantly its susceptibility to sodicity (Frenkel et al. 1978). Effect of highly dilute solutions on the hydraulic conductivity of soils Effect of low electrolyte concentration (3 mmolc/L) on the HC of sodic soils is presented in Fig. 5-5 (Shainberg and Letey 1984). Displacing 10 mmolc/L solutions of SAR 10, 15, 20, and 30 with deionized water (DW) reduced the HC of the soil. When leached with DW, even an ESP of 10 was enough to appreciably reduce the HC of the Fallbrook area soil. Electrolyte concentration of 2 mmolc/L in the percolating solution prevented the adverse effect of ESP 10 (Fig. 5-5). The adverse effect of ESP 15 was prevented by a solution of 3 mmolc/L. Clay dispersion and movement were very sensitive to both ESP and electrolyte concentration. For example, in soil equilibrated with SAR 15 solution, leaching with DW, 1

148

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 5-5. Relative HC of Fallbrook soil sand mixture (1:1), equilibrated with 10 mmolc /L solutions of sodium adsorption ratio (SAR) 10, 15, 20, or 30, and leached with deionized water (DW) or salt solutions of 1, 2, or 3 mmolc /L. From Shainberg and Letey (1984).

or 2 mmolc/L resulted in a peak clay concentration of 1.0%, 0.1%, and 0.02% clay, respectively, in the effluent (Shainberg and Letey 1984). Dispersed and mobile clay in the leachate is only observed in sandy soils. When the clay content in the soil is moderate and high, the small size of the conducting pores usually ensures that dispersed clay moves only short distances before it clogs the pores, resulting in reduced HC. Thus, in loams and clays, the dispersion mechanism still operates, but no macroscopic movement of clay particles is observed. Clogging of the pores is responsible for the sharp decrease in HC of noncalcareous sodic loams and clay soils. Suarez et al. (2006), using rainfall simulation and irrigation, found that for a loam soil, SAR 2 produced adverse impacts on infiltration and SAR 4 produced similar affects for a clay soil. These low SARs are not typical of the published research. They suggest that the combination of rain and

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

149

irrigation increases the “hazard” of decreased infiltration compared to irrigation alone. Effect of mineral weathering and lime content Arid land soils can release 3 to 5 mmolc/L of Ca and Mg to the soil solution as a result of the dissolution of calcites and silicates minerals (Rhoades et al. 1968; Suarez and Wood 1996). The solution composition of a calcareous soil at a given ESP, when placed in contact with DW, was calculated and measured (Shainberg et al. 1981b). As CaCO3 dissolves, Ca replaces Na on exchange sites until the solution is in simultaneous equilibrium with the exchange sites and with lime. The electrical conductivity of the soil solution in equilibrium with soils having ESP values of 5, 10, and 20 is 0.4, 0.6, and 1.2 dS/m (Shainberg et al. 1981b). These concentrations reduce the deleterious effect of ESP when the soil is leached with DW. Many sodic soils contain lime (and a few primary minerals) that readily release soluble electrolytes. These soils will not readily disperse when leached with DW at moderate ESPs, because they will maintain a high enough salt concentration in the soil solution to prevent clay dispersion (Shainberg and Gal, 1982). Effect of pH and exchangeable magnesium on hydraulic conductivity of soils The sensitivity of soil HC to pH depends on the quantity of variablecharge minerals and organic matter present in the soil (Suarez et al. 1984). Soils with large amounts of variable charge are the most susceptible to pH effects. Changes in pH affect the edge-charge on clays and the surface charge of variable-charge minerals, such as Fe and Al oxides. At low pH, edge-to-face bonding may occur, as well as bonding of positive iron and aluminium oxides to negative clay surfaces (van Olphen 1977). This type of bonding hinders dispersion and, thus, results in optimum HC. With increasing pH, both edge-to-face clay bonding and Fe and Al bonding to clays decrease and the HC of the soil decreases (Lado et al. 2007). Exchangeable Mg’s effect on the HC of sodic soil is unclear. The U.S. Salinity Laboratory investigators (1954) grouped Ca and Mg together as similar ions that are beneficial for developing and maintaining soil structure. However, McNeal et al. (1968) showed that mixed Na/Mg soils developed lower HCs than did Na/Ca soils under similar conditions. McNeal et al. (1968) also made a distinction between the direct (specific) effect of exchangeable Mg in causing decreases in HC and the lessened ability of Mg in irrigation water (relative to Ca) to prevent the accumulation of exchangeable Na in the soil. They showed that Mg affects soils dominated by illite but not montmorillonitic soils.

150

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Alperovitch et al. (1981) demonstrated that exchangeable Mg does not affect the HC and clay dispersion of calcareous soils. In well-weathered soils that do not contain CaCO3, exchangeable Mg decreased soil’s HC and increased clay dispersion when the Na/Mg soils were leached with DW. Magnesium-saturated soil clays are chemically more stable and do not release electrolytes into the solution (Alperovitch et al. 1981). They thus disperse more easily when leached with DW. Exchangeable Mg’s presence enhances the dissolution of CaCO3 in calcareous soils, and the electrolytes prevent the dispersion of clay and losses of HC in Na/Mg calcareous soils. Alperovitch et al. (1986) reached similar conclusions when they studied clay dispersivity and HC properties of three California soils as a function of ESP and electrolyte concentration for Na/Ca and Na/Mg systems. They also concluded that soils that were relatively chemically stable were the most sensitive to low ESP and to the effects of exchangeable Mg when leached with DW. Such results agreed with the hypothesis that the susceptibility of soils to decreases in HC produced by exchangeable Na and exchangeable Mg depends on their rates of mineral dissolution. Exchangeable Mg reduces the dissolution rates of noncalcareous soils and increases soil susceptibility to ESP under conditions of low electrolyte concentration.

THE INFILTRATION RATE IN SODIC SOILS Infiltration rate (IR) is the volume flux of water flowing into the profile per unit of soil surface area. In general, initial soil infiltration capacity is high, particularly when the soil is initially dry but tends to decrease monotonically until it asymptotically approaches a constant rate—the steady state or final IR. If water delivery rate to the surface is less than the soil infiltration capacity, water infiltrates as fast as it is delivered, and the supply rate determines the IR. When the delivery rate exceeds the soil infiltration capacity, the latter determines the actual IR and the process becomes surface-controlled. In soils with stable surface structures, decreases in infiltration capacity result from the decrease in matric-suction gradient that inevitably occurs as infiltration proceeds. In the initial wetting, the surface saturates, but the soil below remains dry, and the matric suction gradient is steep. As the wetted zone deepens, this gradient is gradually reduced. Decreases in the infiltration capacity of a soil from an initially high rate can also result when the soil’s structure deteriorates and a surface seal forms. When a seal with very low HC forms at the soil surface, the seal determines the IR of the soil (Le Bissonnais and Singer 1992). Crust formation in soils exposed to the beating action of falling water droplets is due to two mechanisms (Agassi et al. 1981; Kazman et al. 1983): (1) breakdown of the soil aggregates by droplets impact and aggre-

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

151

gate wetting, and (2) physicochemical dispersion of soil clays. Dispersed soil clays can then migrate into the soil with infiltrating water and clog the pores immediately beneath the surface (Agassi et al. 1981; Kazman et al. 1983; McIntyre 1958). These factors produce a thin (0.1-mm) skin seal at the soil surface (McIntyre 1958). Effect of Soil Sodicity on Infiltration Rate The IR (and seal formation) in field soils exposed to rain is more susceptible to soil sodicity than the saturated HC of similar soils run in laboratory soil columns under similar conditions (Shainberg and Letey 1984). The high sensitivity of the soil surface to low ESP values was explained by three factors: (1) the mechanical impact of rain drops, which enhances clay dispersion; (2) the absence of surrounding soil matrix which, when present, slows clay dispersion and movement; and (3) the absence of electrolytes released by mineral dissolution. Even ESP 1.0 accelerated the formation of a seal of a noncalcareous loamy soil, and seal formation increased with further increase in ESP to values of 2.2 and 4.6 (Fig. 5-6). Effect of Soil Texture The effect of clay content on the IR of soils with low ESP was studied by Ben-Hur et al. (1985). Soils with 10% to 30% clay were the most susceptible to seal formation and had the lowest IR (Fig. 5-7). With increasing clay content, the soil structure was more stable, and seal formation was diminished. In soils with low clay content, the amount of clay was not enough to clog soil pores, and a poorly developed seal was formed. Le Bissonnais et al. (1995) found a complex relationship among initial moisture content, soil organic matter content, and clay content for high-silt soils of Europe, which suggests that soils need to be considered on a siteby-site basis and that broad generalizations are not sufficient. Effect of Electrolyte Concentration Clay dispersion is very sensitive to electrolyte concentration and sodicity of the applied water (Shainberg and Letey 1984). This is particularly true in the case of seal formation. Agassi et al. (1981) studied the effect of electrolyte concentration on the IR of a sandy loam in rain-simulation studies (Fig. 5-8). These results simulate the IR in soils exposed to sprinkler irrigation, and indicate that the IR is far more sensitive than the HC to the electrolyte concentration of the applied water. These results also explain the beneficial effect of gypsum spread at the soil surface (Fig. 5-6). Dissolution of gypsum increases the electrolyte concentration in rain water and slowed down seal formation.

152

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 5-6. Effect of ESP and phosphogypsum on the IR of sandy-loam exposed to deionized water (DW) rain. From Shainberg and Letey (1984).

Effect of Clay Mineralogy on Seal Formation Most of the studies on seal formation and runoff were conducted on soils in which the dominant clay minerals were smectites (previous sections). These clay minerals are known to be more dispersive than kaolinitic clays (Frenkel et al. 1978). Stern et al. (1991), studying seal formation of kaolinitic soils from South Africa, concluded that pure kaolinitic soils are not susceptible to crusting. However, kaolinitic soils with smectitic impurities were dispersive and susceptible to sealing.

DEPOSITIONAL CRUST: EFFECT OF SALINITY AND SODICITY Slow water penetration may be a problem in crop production in many surface-irrigated soils. In furrow and border irrigation, water penetration at the beginning of the irrigation season might be high and the rate

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

153

FIGURE 5-7. Infiltration rate (IR) of various soils as a function of clay percentage. From Ben-Hur et al. (1985), Fig. 1, p. 285, with kind permission of Springer ScienceBusiness Media. of water penetration might decrease with more irrigations. Overland flow applies shear forces to the soil surface, which causes particle detachment and movement. As the water infiltrates the soil, sediment is deposited at the soil surface, and a depositional crust is formed (Chen et al. 1980; Fox et al. 1998; Lentz et al. 1992; Shainberg and Singer 1985;

154

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 5-8. Effect of electrolyte concentration in rain simulation experiments on the infiltration rate (IR) of a sandy loam. From Shainberg and Letey (1984).

Southard et al. 1988). Fox et al. (1998) and Southard et al. (1988) showed the effects of depositional crust formation on pore size distribution and pore shape. Shainberg and Singer (1985) produced depositional crusts in the laboratory and measured the effects of electrolyte concentration on their per-

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

155

meability. They found that the depositional crust’s permeability was two to three orders of magnitude below that of the underlying soil when the electrolyte concentration of the applied solution was less than 0.3 dS/m. When the electrolyte concentration of the water exceeded 0.5 dS/m, the crust HC was only one order of magnitude below that of the bulk soil. The high susceptibility of the depositional crust to the electrolytic concentration in the irrigation water was similar to that of structural crusts (Agassi et al. 1981). Whereas the soil’s HC was not affected by the electrolyte concentration of the applied suspension, the HC of the depositional crust was very sensitive to the electrolytic concentration. Shainberg and Singer (1985) and Southard et al. (1988) assumed that the dispersion-flocculation status of the sediments in the suspension determined the hydraulic properties of the depositional crust formed. Depositional crusts formed from flocculated particles had an open and permeable structure. Conversely, dispersed soil particles formed a dense crust in which the particles were oriented parallel to the soil surface (Southard et al. 1988). Thus, the HCs of depositional crusts in equilibrium with dilute solutions and rain water are low, while the permeability of the crusts in equilibrium with more concentrated solutions is high (Shainberg and Singer 1985).

EFFECT OF EXTRINSIC PHYSICAL CONDITIONS ON SUSCEPTIBILITY OF SOILS TO SODICITY The extent of sodicity-induced deterioration on soil hydraulic properties depends on aggregate stability (Shainberg et al. 2001). In soils with stable aggregates, higher values of ESP are needed to deteriorate the soil’s physical properties. Similarly, soils with unstable aggregates are expected to be more susceptible to sodicity. When the clay particles are together in a stable aggregate, a higher concentration of exchangeable Na is needed for clay dispersion and clay swelling. Thus, breakdown of sodic soil aggregates, prior to leaching with dilute solutions, should increase the deterioration of soil properties caused by sodicity, and any procedure that stabilizes soil aggregates should decrease soil susceptibility to sodicity. It is expected that fast wetting of dry soil—a procedure that disintegrates soil aggregates—will increase soil susceptibility to sodicity. Conversely, aging of soils—a procedure that stabilizes soil aggregates—will decrease the susceptibility of soils to sodic conditions. Because both processes (disintegration of aggregates by fast wetting and aggregate stabilization by aging) depend on the antecedent moisture content (AMC) of the soil. AMC is also expected to affect susceptibility of soils to sodic conditions.

156

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Effect of Clay Content and Wetting Rate on Aggregate Stability and Hydraulic Conductivity of Soils Aggregate stability increases with increase in clay content (Kay and Angers 1999; Kemper and Koch 1966; Le Bissonnais 1996; Le Bissonnais et al. 2006). This was attributed to clay particles acting as a cementing material holding the primary particles together in the aggregates (Kemper and Rosenau 1984). Structural destabilization during wetting occurs because water stability of aggregates is not strong enough to withstand the stresses produced by differential swelling, entrapped air explosion, rapid release of heat of wetting, and the mechanical action of moving water (Emerson 1977; Kay and Angers 1999; Loch 1994). The extent of slaking during wetting has been shown to increase with increase in the rate of wetting (Kay and Angers 1999; Panabokke and Quirk 1957). Panabokke and Quirk (1957) reported that differential swelling plays a major role in the slaking of clayey soils, whereas for loamy soils the entrapped air is the major cause of slaking. The effect of wetting rate (WR) and clay content on the HC of five smectitic soils of different clay content was studied by Shainberg et al. 2001 (Fig. 5-9). The HC of the loamy sand (9% clay) was high due to the low clay and silt content. Because of the low clay content, this soil contains hardly any aggregates that may disintegrate by fast wetting. Thus, its HC was not sensitive to WR (Fig. 5-9a). The HC of the loam (22% clay) was 15.8 mm/h, which was much lower than that of the loamy sand. The low HC of the loam was due to its high silt content and high silt/clay ratio. Similar to the loamy sand, the HC of the loam was also not affected by the WR. The textural composition of the loam enabled the formation of aggregates but with low stability (Ben-Hur et al. 1985). Thus, even a low WR was enough to disintegrate the aggregates, and further disintegration by the fast WR was negligible (Fig. 5-9a). In the sandy clay and the two clay soils with low ESP (Fig. 5-9a), HC was affected by WR. In the sandy clay, clay content of 39% stabilized the soil aggregates enough to prevent the collapse of the aggregates by slow WR and the HC was high (80.3 mm/h). Moderate and fast WRs disintegrated the aggregates of the sandy clay, and the HC decreased to 67 and 34 mm/h, respectively. The HC values after slow WR for the two clay soils (52% and 62% clay) were lower than that of the sandy clay, and the HC of both soils was more affected by increasing the WR than the sandy clay. Fast WR decreased the relative HC of the soils with 39%, 52%, and 62% clay to 0.42, 0.18, and 0.13 of the slow HC values, respectively. It seems that the two mechanisms responsible for aggregate slaking (explosion by entrapped air and differential swelling) were more active as clay content in the soils increased. These two mechanisms complement one another in disintegrating the aggregates and reducing the HC of clay soils (Shainberg et al. 2001).

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

157

FIGURE 5-9. (a) Effect of wetting rate (1.7, 4.25, and 50 mm/h) on HC of five soils as a function of their clay content. (b,c) The effect of ESP (2, 6, and 10) on the HC of the five soils as a function of clay content and wetting rate. From Shainberg et al. (2001) with permission from CSIRO Publishing.

158

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

No simple relation was found between the HC and clay content in the low-ESP soils (Fig. 5-9a). In the slow WR, HC first decreased with an increase from 9% to 22% clay, then increased upon further increase in clay content to 39%, and then decreased again with a further increase in clay content. The soil with 39% clay maintained high HC values at the slow WR, which decreased as the clay concentration either decreased or increased. Similar behavior was obtained for the five soils with low ESP exposed to moderate or fast WR. This phenomenon was explained by two opposing mechanisms: (1) effect of clay content on aggregate stability, and (2) effect of clay content on aggregate swelling. With an increase in clay content from 22% to 39%, stability of the aggregates increased, and the HC increased. With further increases in clay content, swelling of smectitic clay decreased the size of the transmission pores and the HC. Effect of Exchangeable Sodium Percentage and Wetting Rate on Hydraulic Conductivity of Smectitic Soils The effect of low ESP (10) on HC of the five soils at the two wetting rates is presented in Figs. 5-9b and 5-9c. In the loamy sand and the silt loam, there was a minute effect of WR and ESP on HC. In soils with unstable aggregates, fast wetting did not further disintegrate the aggregates, and the susceptibility of these soils to sodicity was not affected by WR. The effect of ESP on the HC of the sandy clay (39% clay) was affected by WR (Figs. 5-9b and 5-9c). Whereas at fast WR, ESP of 5 and 10 decreased HC to 0.39 and 0.19, respectively, of the HC at ESP 1.6, the corresponding values at the slow WR were 0.98 and 0.46 of the HC at ESP 1.6. The effect of ESP was more pronounced as WR increased. Slaking of the aggregates increased the susceptibility of the sandy clay to sodic conditions. In the soils with 52% and 62% clay, the relative effect of ESP on HC was noticeable only at the slow WR. In these soils, fast WR reduced HC of the nonsodic clay soils to 0.18 and 0.13 of the HC at the slow WR, respectively. The very low HC at the fast WR suggests that most of the aggregates had disintegrated, and clay swelling sealed the soils. Thus, the additional effect of higher ESP in further sealing the soils was not pronounced. The Effect of Aging on the Hydraulic Conductivity of Smectitic Soils The effect of aging duration, water content, and temperature during aging on aggregate stability, IR, and erosion have been studied by Kemper and Rosenau (1984), Kemper et al. (1987), Levy et al. (1997), and Moutier et al. (1998). It has been suggested that (1) the binding of soil particles by clay particles is independent of organic matter content and the activity of a viable microbial population (Kemper et al. 1987); (2) chemical

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

159

processes, such as CaCO3 and silica precipitation (Kemper et al. 1987), and physicochemical binding forces between clay platelets, such as edge-toface forces (Moutier et al. 1998), might be responsible for the development of cohesion forces; and (3) clay movement and reorientation, which increase with an increase in water content, aging, temperature, and clay content, might control the rate of development of cohesive forces. This bonding mechanism is supported by the fact that aggregate stability increases with increase in clay content (Kemper and Koch 1966; Levy and Mamedov 2002). Thus, it is expected that the susceptibility of soils to sodicity decreases with an increase in aging. The Effect of Wetting Rate and Sodicity on Infiltration Rate and Seal Formation Most of the studies on the effects of sodicity and rain kinetic energy on seal formation were done on dry soil samples subjected to rapid wetting— either from below or from above—prior to their exposure to simulated rain. Under these conditions, the soils were highly susceptible to ESP and the affects of soil texture (Figs. 5-6 through 5-8). In this section, the effect of distilled water rain on the measured IR of three of the soils prewetted at rates of 2, 8, and 64 mm/h is presented (Fig. 5-10). These soils represent three groups based on their response to WR and sodicity (Mamedov et al. 2001). The first group contained the loamy sand with IR curves highly sensitive to the ESP of the soil but only slightly affected by the WR (Fig. 5-10a). Note that within a given ESP, the IR curves for the different WRs were close to each other. However, the different ESP groups have completely separate IR curves (Fig. 5-10a). The observed negligible effect of WR on the IR curves in the loamy sand was ascribed to the fact that this soil is without structure (i.e., contains few aggregates) because of its low clay and organic matter content. The second group contained the sandy clay (Fig. 5-10b). In this group the IR curves were moderately affected by both the WR and the ESP of the soils. The third group contained the clay soil with 61.7% clay (Fig. 5-10c). In this group the effect of WR on the IR curves was the dominant factor, and the ESP played a secondary role (Fig. 5-10c). Furthermore, the effect of WR increased with an increase in clay content, between clay contents of 52%, 62%, and 68% clay (Mamedov et al. 2001). Because of the high clay content in these soils, their aggregates were stable and susceptible to WR. The high susceptibility to WR was due to the two mechanisms that were responsible for aggregate slaking: differential swelling and “explosion” of entrapped air (Panabokke and Quirk 1957). Albeit playing a secondary role in the development of seal in the clay soil, the effect of ESP was also evident. In each of the WR treatments, the higher the ESP, the lower the IR curve (Fig. 5-10c).

160

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 5-10. Effects of soil ESP and wetting rate (WR) on the IR of sandy loam (a), sandy clay (b), and clay (c). From Mamedov et al. (2001) with permission from CSIRO Publishing.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

161

The IR curves for the clay soil suggest that the physical disintegration of the surface aggregates by rapid wetting was essential for seal development. Surface sealing is determined by both aggregate disintegration and clay dispersion. Clay dispersion depends to a large extent on the ESP of the soil, whereas aggregate disintegration depends on, among others, the rate of wetting of the aggregates. The results presented show that the effects of ESP and WR on soil IR and runoff are closely associated with soil texture. In soils with low clay content (loamy sand), the effect of WR on seal formation was negligible, and the effect of ESP was important. In the clay soils with stable aggregates (52% clay), aggregate disintegration is essential for seal formation and WR had a predominant effect on the IR and runoff. In these soils, the effect of ESP on seal formation was notable yet secondary to that of the WR. The soils with intermediate clay content (20%–40% clay) were the most susceptible to seal formation, and both WR and ESP had moderate effects on seal formation.

EFFECT OF POLYACRYLAMIDE AND GYPSUM ON RUNOFF AND EROSION FROM SODIC SOILS Use of synthetic polymers as soil additives started as early as the 1950s. Laboratory and field studies with anionic PAM (polyacrylamide with negative charge and high molecular weight) demonstrated that addition of small amounts of PAM (10–20 kg ha1) to the soil surface were effective in maintaining high permeability and decreasing runoff and soil erosion levels, especially when the PAM was applied together with a source of electrolytes (e.g., Flanagan et al. 1997b; Shainberg et al. 1990; Sojka et al. 2007). In nearly all studies on PAM applications for preventing raininduced seal formation, PAM was initially dissolved in water and sprayed onto the soil surface or added to the irrigation water. Neither practice is suitable for rain-fed or irrigated agriculture because it is difficult to dissolve dry PAM in water (Tang et al. 2006; Yu et al. 2003). Studies on erosion control in furrow irrigation have shown that addition of dry granules of PAM to the gated irrigation pipe had favorable effects on preventing erosion and increasing infiltration, comparable to those of adding stock solution of PAM to the furrow inflow (Sojka et al. 2007). This success led Yu et al. (2003) and Tang et al. (2006) to apply dry granular PAM, either mixed or unmixed with phosphogypsum (PG), to the soil surface prior to simulated rain application. In the Yu et al. (2003) studies, on soils with low ESP the application of dry PAM mixed with gypsum was highly effective in decreasing runoff and erosion from the soils exposed to simulated DW rainstorms. The effect of ESP (between ESP 2 and 20) on runoff and erosion from four smectitic soils, ranging in clay

162

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 5-11. Cumulative runoff as a function of exchangeable sodium percent (ESP) and PAM and PG treatments in four soils varying in clay content. From Tang et al. (2006) with permission from Soil Science Society of America. content between 10% and 62%, compared to spreading dry PAM mixed with PG prior to simulated DW rain application, are presented in Figs. 511 and 5-12. Increasing the ESP of the soils increased runoff and erosion in the four untreated soils (control treatments). The soils varied in their response to sodicity, and sodicity affected erosion more than runoff. Spreading PAM mixed with PG or just PG was effective in maintaining low runoff and erosion levels compared with the control. Use of PAM mixed with PG resulted in lower runoff and erosion than just PG alone.

SUMMARY AND CONCLUSIONS The available data suggest that a suite of soil conditions, which vary on a site-by-site basis, largely determine the response of a soil to wetting/irrigation. Factors that may affect the impact of sodicity on a soil’s IR and HC include the soil texture, soil clay content, the type of clay, electrolyte concentration, lime content, pH, exchangeable Mg, depositional crust formation (particularly associated with horizontal flow), aging, and the use of PAM or PG.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

163

FIGURE 5-12. Soil loss as a function of ESP and PAM and PG treatments in four soils varying in clay content. From Tang et al. (2006) with permission from Soil Science Society of America.

Based on currently available research information, only a limited ability exists to predict the impact of sodicity on clay dispersion and soil hydraulic properties in the field. This stems from the fact that the laboratory methods used to obtain data do not closely reflect typical field conditions. The laboratory experiments used disturbed and dry samples, fast wetting, and no aging—conditions that enhanced slaking of aggregates and clay dispersion. Under field conditions, soil aggregation is maintained and the stable aggregates are less susceptible to sodicity.

REFERENCES Abu-Sharar, T. M., Bingham, F. T., and Rhodes, J. D. (1987). “Stability of soil aggregates as affected by electrolyte concentration and composition.” Soil Sci. Soc. Amer. J., 51, 309–314. Agassi, M., Shainberg, I., and Morin, J. (1981). “Effect of electrolyte concentration and soil sodicity on infiltration rate and crust formation.” Soil Sci. Soc. Amer. J., 48, 848–851. Alperovitch, N., Shainberg, I., and Keren, R. (1981). “Specific effect of magnesium on the hydraulic conductivity of sodic soils.” J. Soil Sci., 32, 543–554.

164

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Alperovitch, N., Shainberg, I., and Singer, M. J. (1985). “Effect of clay mineralogy, Al and Fe oxides on HC of clay sand mixtures.” Clays Clay Min., 33, 443–450. Alperovitch, N., Shainberg, I., and Rhoades, J. D. (1986) “Effect of mineral weathering on the response of sodic soils to exchangeable magnesium.” Soil Sci. Soc. Am. J., 50, 901–904. Amezketa, E., Singer, M. J., and Le Bissonnais, Y. (1996). “Testing a new procedure for measuring water stable aggregates.” Soil Sci. Soc. Am. J., 60, 888–894. Bagarello, V., Iovino, M., Palazzolo, E., Panno, M., and Reynolds, W. D. (2006). “Field and laboratory approaches for determining sodicity effects on saturated soil hydraulic conductivity.” Geoderma, 130, 1–13. Ben-Hur, M., Shainberg, I., Baker, D., and Keren, R. (1985). “Effects of soil texture and CaCO3 content on water infiltration in crusted soil as related to water salinity.” Irrig. Sci., 6, 281–294. Bronick, C. J., and Lal, R. (2005). “Soil structure and management: A review.” Geoderma, 124, 3–22. Chen, J., Tarchitzki, J., Morin J., and Banin, A. (1980). “Scanning electron microscope observations on soil crust and their formation.” Soil Sci., 130, 49–55. Dexter, A. R. (2004).”Soil physical quality. Part I. Theory, effects of soil texture, density, and organic matter, and effects on root growth.” Geoderma, 120, 201–214. Emerson, W. W. (1977). “Physical properties and structure,” in Soil factors in crop production in a semiarid environment, J. S. Russell and E. L. Green, eds., Queensland University Press, Brisbane, 78–104. Falsone, G., and Bonifacio, E. (2006). “Destabilization of aggregates in some Typic Fragiudalfs.” Soil Sci., 171, 272–281. Flanagan, D. C., Norton, L. D., and Shainberg, I. (1997a). “Effect of water chemistry and soil amendments on a silt loam. Part 1: Infiltration and runoff.” Trans. ASAE., 40, 1549–1554. ———. (1997b). “Effect of water chemistry and soil amendments on a silt loam. Part 2: Soil erosion.” Trans. ASAE., 40, 1555–1561. Fox, D. M., Le Bissonnais, Y., and Quetin, P. (1998). “The implications of spatial variability in surface seal Rc for infiltration in a mound and depression microtopography.” Catena, 32, 101–114. Frenkel, H., Goertzen, J. Q., and Rhoades, J. D. (1978). “Effects of clay type and content, ESP and electrolyte concentration on clay dispersion and soil hydraulic conductivity.” Soil Sci. Soc. Am. J., 48, 32–39. Kay, B. P., and Angers, D. A. (1999). “Soil structure,” in Handbook of soil science, M. E. Sumner, ed., CRC Press, Boca Raton, Fla., A-229–A-269. Kazman, Z., Shainberg, I., and Gal, M. (1983). “Effect of low levels of exchangeable Na and applied PG on the infiltration rate of various soils.” Soil Sci., 135, 184–192. Kemper, W. D., and Koch, E. J. (1966). “Aggregate stability of soils from western USA and Canada.” USDA Technical Bulletin No. 1355, U.S. Government Printing Office, Washington D.C. Kemper, W. D., and Rosenau, R. C. (1984). “Soil cohesion as affected by time and water content.” Soil Sci. Soc. Am. J., 48, 1001–1006. Kemper, W. D., Rosenau, R. C., and Dexter, A. R. (1987). “Cohesion development in disrupted soils as affected by clay and organic matter content and temperature.” Soil Sci. Soc. Am. J., 51, 860–867.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

165

Keren, R. and Shainberg, I. (1981). “Effect of dissolution rate on the efficiency of industrial and mined gypsum in improving infiltration of a sodic soil.” Soil Sci. Soc. Am. J., 45, 103–107. Lado, M., Ben Hur, M., and Shainberg, I. (2007). “Clay mineralogy, ionic composition and pH effects on hydraulic properties of depositional seals.” Soil Sci. Soc. Am. J., 71, 314–321. Le Bissonnais, Y. (1996). “Aggregates stability and assessment of crustability and erodibility. 1. Theory and methodology.” Eur. J. Soil Sci., 47, 425–437. Le Bissonnais, Y., Blavet, D., De Noni, G., Laurent, J.-Y., Asseline, J., and Chenu, C. (2006). “Erodibility of Mediterranean vineyard soils: relevant aggregate stability methods and significant soil variables.” Eur. J. Soil Sci., 58, 188–195. Le Bissonnais, Y., Renaux, B., and Delouche, H. (1995). “Interactions between soil properties and moisture content in crust formation, runoff and interrill erosion from tilled loess soils.” Catena, 25, 33–46. Le Bissonnais, Y., and Singer, M. J. (1992). “Crusting, runoff and erosion response to soil water and successive rainfalls.” Soil Sci. Soc. Am. J., 56, 1898–1903. Lentz, R. D., Shainberg, I., Sojka, R. E., and Carter, D. L. (1992). “Preventing irrigation furrow erosion with small application of polymers.” Soil Sci. Soc. Am. J., 56, 1926–1932. Levy, G. J. 1999. “Sodicity,” in Handbook of soil science, M. E. Sumner, ed., CRC Press, Boca Raton, Fla., G27–G63. Levy, G. J., Goldstein, D., and Mamedov, A. I. (2005). “Hydraulic conductivity of semiarid soils: Combined effects of salinity, sodicity, and rate of wetting.” Soil Sci. Soc. Amer. J., 69, 653–662. Levy, G. J., Levin, J., and Shainberg, I. (1997). “Prewetting rate and aging effect on seal formation and interrill soil erosion.” Soil Sci., 162, 131–139. Levy, G. J., and Mamedov, A. I. (2002). “Aggregate stability and seal formation.” Soil Sci. Soc. Am. J., 66, 1603–1609. Levy, G. J., and Shainberg, I. (2004). “Sodic soils,” in Encyclopedia of soils in the environment, Vol. 3, D. Hillel, ed., Elsevier, Oxford, UK, 504–512. Levy, G. J., Shainberg, I., and Miller, W. P. (1998). “Physical properties of sodic soils,” in Sodic soils distribution, properties management and environmental consequence, M. E. Sumner and R. Naidu, eds., Oxford University Press, New York, 77–94. Loch, R. J. (1994). “Structure breakdown on wetting,” in Sealing, crusting and hardsetting soils: Productivity and conservation. Proc., 2nd Int. Symp. on Sealing Crusting and Hardsetting Soils: Productivity and Conservation held at the University of Queensland, February 7–11, 1994, H. B. So, ed., Australian Society of Soil Science, 113–132. Mace, J. E., and Amrhein, C. (2001). “Leaching and reclamation of a soil irrigated with moderate SAR waters.” Soil Sci. Soc. Amer. J., 65, 199–204. Mamedov, A. I., Shainberg, I., Levy, G. J., and Letey, J. (2001). “Prewetting rates and sodicity effects on surface sealing.” Aust. J. Soil Res., 39, 1293–1305. Mamedov, A. I., Shainberg, I., and Levy, G. J. (2002). “Wetting rates and sodicity effect on interrill erosion from semi-arid Israeli soils.” Soil Tillage Res., 68, 121–132. McIntyre, D. S. (1958). “Permeability measurements of soil crusts formed by raindrop impact.” Soil Sci., 85, 185–189.

166

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

McNeal, B. L., and Coleman, N. T. (1966). “Effect of solution composition on soil hydraulic conductivity.” Soil Sci. Soc. Am. Proc., 30, 308–312. McNeal, B. L., Layfield, D. A., Norvell, W. A., and Rhoades, J. D. (1968) “Factors influencing hydraulic conductivity of soils in the presence of mixed salt solution.” Soil Sci. Soc. Am. Proc., 32, 187–190. Moutier, M., Shainberg, I., and Levy, G. J. (1998). “Hydraulic gradient, aging, and water quality effects on hydraulic conductivity of a vertisol.” Soil Sci. Soc. Am. J., 62, 1488–1496. Oster, J. D. (1994). “Irrigation with poor quality water.” Ag. Water Mgmt., 25, 271–297. Oster J. D., Shainberg I., and Wood, J. D. (1980). “Flocculation value and gel structure of Na/Ca montmorillonite and illite suspensions.” Soil Sci. Soc. Am. J., 44, 955–959. Panabokke, C. R., and Quirk, J. P. (1957). “Effect of initial water content on stability of soil aggregates in water.” Soil Sci., 83, 185–195. Quirk, J. P. and Schofield, R. K. (1955). “The effect of electrolyte concentration on soil permeability.” J. Soil Sci., 6, 163–178. Ramos, M. C., Nacci, S., and Pla, I. (2003). “Effect of raindrop impact and its relationship with aggregate stability to different disaggregation forces.” Catena, 53, 365–376. Rengasamy, P., and Olsson, K. A. (1991). “Sodicity and soil structure.” Aust. J. Soil Res., 29, 935–952. Rengasamy, P., and Sumner, M. E. (1998). “Processes involved in sodic behavior,” in Sodic soils distribution, properties management and environmental consequence, M. E. Sumner and R. Naidu, eds., Oxford University Press, New York, 35–50. Rhoades, J. D., Krueger, D. B., and Reed, M. J. (1968). “The effect of soil mineral weathering on the sodium hazard of irrigation water.” Soil Sci. Soc. Am. J., 32, 343–347. Ruiz-Vera, V. M., and Wu, L. (2006).” Influence of sodicity, clay mineralogy, prewetting rate, and their interaction on aggregate stability.” Soil Sci. Soc. Amer. J., 70, 1825–1833. Schofield, R. K., and Samson, H. R. (1954). “Flocculation of kaolinite due to the attraction of oppositely charged crystal faces.” Disc. Faraday Soc., 18, 135. Shainberg, I., and Gal, M. (1982). “The effect of lime on the response of soils to sodic conditions.” J. Soil Sci., 33, 489–498. Shainberg, I., and Letey, J. (1984). “Response of soils to sodic and saline conditions.” Hilgardia, 52, 1–57. Shainberg, I., and Levy, G. J. (2004). “Salinization processes,” in Encyclopedia of soils in the environment, D. Hillel, ed., Vol. 3, Elsevier, Oxford, UK, 429–434. Shainberg, I., Levy, G. J., Goldstein, D., Mamedov, A. I., and Letey, J. (2001). “Prewetting rate and sodicity effects on the hydraulic conductivity of soils.” Aust. J. Soil Res., 39, 1279–1291. Shainberg, I., Rhoades, J. D., and Prather, R. J. (1981a). “Effect of low electrolyte concentration on clay dispersion and hydraulic conductivity of a sodic soil.” Soil Sci. Soc. Am. J., 45, 273–277. Shainberg, I., Rhoades, J. D., Suarez, D. L., and Prather, R. J. (1981b). “Effect of mineral weathering on clay dispersion and hydraulic conductivity of sodic soils.” Soil Sci. Soc. Am. J., 45, 287–291.

SOIL RESPONSE TO SALINE AND SODIC CONDITIONS

167

Shainberg, I., and Singer, M. J. (1985). “Effect of electrolyte concentration on the hydraulic properties of depositional crust.” Soil Sci. Soc. Am. J., 49, 1260–1263. Shainberg, I., Warrington, D., and Rengasamy, P. (1990). “Effect of soil conditioner and gypsum application on rain infiltration and erosion.” Soil Sci., 149, 301–307. Sojka, R. E., Bjorneberg, D. L., Entry, J. A., Lentz, R. D., and Orts, W. J. (2007). “Polyacrylamide in agriculture and environmental land management.” Adv. Agron., 92, 75–162. Southard, R. J., Shainberg, I., and Singer, M. J. (1988). “Influence of electrolyte concentration on the micromorphology of artificial depositional crust.” Soil Sci., 45, 278–288. Stern, R., Ben-Hur, M., and Shainberg, I. (1991). “Clay mineralogy effect on rain infiltration, seal formation and soil losses.” Soil Sci., 152, 455–462. Suarez, D. L., Rhoades, J. D., Lavado, R. L., and Grieve, C. M. (1984) “Effect of pH on saturated hydraulic conductivity and soil dispersion.” Soil Sci. Soc. Am. J., 48, 50–55. Suarez, D. L., and Wood, J. D. (1996). “Short and long term weathering rates of a feldspar fraction isolated from an arid zone soil.” Chem. Geol., 132, 143–150. Suarez, D. L., Wood, J. D., and Lesch, S. M. (2006). “Effect of SAR on water infiltration under a sequential rain-irrigation management system.” Ag. Water Mgmt., 86, 150–164. Sumner, M. E. (1993). “Sodic soils: New perspectives.” Aust. J. Soil Res., 31, 683–750. Sumner, M. E., and Naidu, R. (1998). Sodic soils. distribution, properties, management and environmental consequences. Oxford University Press, New York. Sumner, M. E., and Stewart, B. A. (1992). Soil crusting: Chemical and physical processes. Lewis Publishers, Boca Raton, Fla. Tang, Z., Lei, T., Yu, J., Shainberg, I., Mamedov, A. J., and Levy, G. J. (2006). “Runoff and erosion in sodic soils treated with dry PAM and phosphogypsum.” Soil Sci. Soc. Am. J., 70, 679–690. U.S. Salinity Laboratory. (1954). Diagnosis and improvement of saline and alkali soils. USDA Agricultural Handbook No. 60, U.S. Government Printing Office, Washington, D.C. van Olphen, H. (1977). An introduction to clay colloid chemistry, 2nd ed. John Wiley and Sons, New York. Yu, J., Lei, T., Shainberg, I., Mamedov, A. I., and Levy, G. J. (2003). “Infiltration and erosion in soils treated with dry PAM and gypsum.” Soil Sci. Soc. Am. J., 67, 630–636.

NOTATION CEC  cation exchange capacity, cmolc/kg EC  electrical conducitivity, dS/m EPP  exchangeable potassium percentage, % ESP  exchangeable sodium percentage, % FV  flocculation value, molc/m3 IR  infiltration rate, mm/h SAR  sodium absorption ratio

This page intentionally left blank

CHAPTER 6 PLANT RESPONSES TO SALINE AND SODIC CONDITIONS André Läuchli and Stephen R. Grattan

INTRODUCTION The vast majority of crop plants are glycophytes, or nonhalophytes. These plants are unable to tolerate stresses imposed by saline or salinesodic conditions, unlike the halophytes, which thrive under these harsh environments. Plant stress refers to a condition where the plant is unable to express its full genetic potential for growth, development, and reproduction (Läuchli and Epstein 1990). Salinity and sodicity stresses are quantitatively expressed as concentrations, activities, and similar units, but there is no sharp dividing line between saline or sodic stress and lack of stress (Läuchli and Epstein 1990). Rather, there is a continuum from the absence of stress to severe stress. Plants, including genotypes within a species, vary widely in their tolerance to saline and sodic conditions (see Chapter 13). However, there is also no clear distinction between salt tolerance and salt sensitivity. Salt sensitivity of a given plant is indicated by the point or range in the continuum of stress where the plant shows visual or quantitative signs of being adversely affected. This depends not only on the intensity of the saline or sodic stress but also on the chemical composition of the medium and other abiotic and biotic stresses such as temperature, water deficit, flooding, nutritional inadequacies, poor soil physical conditions, pests, and pathogens (Mittler 2006). Thus, plants under field conditions often endure multiple stresses that show potential interactions, which can be both negative and positive (Läuchli and Grattan 2007).This chapter provides a general overview of the principal mechanisms and crop responses to salinity and sodicity stress. The chapter also focuses on how the whole plant integrates these responses and mechanisms. For in-depth reviews 169

170

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

on mechanisms of salt tolerance in halophytes, see papers written by Flowers (1985), Flowers et al. (1977), Wyn Jones (1981), Gorham (1996), Lüttge (1993, 2002), Lovelock and Ball (2002), and Cushman and Bohnert (2002) among others. In addition, our chapter does not focus on biochemical and molecular mechanisms of salt tolerance or on genomics-type technologies such as those discussed in recent reviews by Hasegawa et al. (2000), Zhu (2002), Koiwa et al. (2006), and Bohnert et al. (2006).

SALINITY AND SODICITY Irrigation water supplies contain dissolved mineral salts, but the concentration and composition of the dissolved salts vary from one source to another. The most common cations are calcium (Ca2), magnesium (Mg2), and sodium (Na), while the most abundant anions are chloride (Cl), sulfate (SO42) and bicarbonate (HCO3). Potassium (K), carbonate (CO32), nitrate (NO 3 ), and trace elements also exist in water supplies, but most often the concentrations of these constituents are comparatively low. Conversely, some groundwater sources contain boron (B) at comparatively low concentrations but at levels that may be detrimental to certain crops. There is a clear distinction between salinity and sodicity—the former being related to salt concentration and the latter to salt composition. Salinity refers to the concentration of salts in the irrigation water or soil that is sufficiently high to adversely affect crop yields or crop quality. This is based strictly on the colligative property of the soil solution regardless of its ionic composition. These adverse responses are caused by high concentrations of salts that lower the osmotic potential of the soil solution (i.e., osmotic effects) or by high concentrations of specific ions, such as Cl or Na, that can cause specific injury to the crop (i.e., specificion effects). Sodicity, on the other hand, is related to the proportion of Na in the water, or adsorbed to the soil surface, relative to Ca and Mg. Sodicity can contribute to the deterioration of soil physical properties, which can indirectly affect plants via crusting, reduced infiltration, increased soil strength, and reduced aeration resulting in anoxic or hypoxic conditions for roots. Sodicity has been described in different ways (Jurinak and Suarez 1990). The sodicity of soil is characterized by the exchangeable sodium percentage (ESP). The ESP is the percentage of the cation exchange capacity occupied by Na. The sodicity of the water, however, is a measure of the sodium adsorption ratio (SAR). SAR is defined as SAR 

Na (Ca2  Mg 2 )

(6-1)

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

171

where concentrations of all cations are molarities. The ESP and SAR are related to one another and, for most practical purposes, are numerically equivalent in the range of 3 to 30 (Richards 1954). At the soil surface, infiltration rates and soil tilth can be adversely affected by salinity and sodicity, particularly when irrigation with nonsaline water or rain follows irrigation with saline-sodic water. Water infiltration rates (Oster and Schroer 1979) and the soil hydraulic conductivity (HC) (McNeal and Coleman 1966) decrease with decreasing soil salinity and with increasing exchangeable Na, or sodicity. These processes occur due to the combination of clay swelling and instability of soil aggregates (Quirk 2001). In addition, clay movement and deposition into soil pores reduces the fraction of large pores in the soil. Large-pore-size distribution is important for water movement, adequate drainage, and proper aeration. Other significant factors that influence aggregate stability are soil water contents before wetting occurs, the rate of wetting, and soil texture (Manedov et al. 2001; Shainberg et al. 2001), as well as clay mineralogy, organic matter, CaCO3, sesquioxides, and pH (Levy et al. 1998). Chapter 5 in this manual covers the mechanisms of salinity-sodicity impacts on soil physical conditions in more detail. More than half a century ago, researchers assigned numerical values to soils classifying them as either saline and/or sodic (Richards 1954). Historically soils were considered “saline” when the electrical conductivity (EC) of the saturated soil extract exceeded 4 dS/m at 25 °C, and were classified as “sodic” when the ESP exceeded 15. Because these definitions are different, a soil could be classified as either saline, sodic, saline-sodic, or nonsaline/nonsodic. Since that time, much has been learned about the many factors affecting crop response to salinity and differences in sensitivity to salinity among crop species (Läuchli and Epstein 1990). Similarly, much has been learned about the complexities of soil mineralogy, clay content, organic matter, and ionic strength, and composition of irrigation waters on aggregate stability and soil physical conditions (Jurinak and Suarez 1990). As such, those historical values for classifying soils are no longer valid because many other factors must be considered to assess whether an irrigation water can have negative impacts on crops or soils. Numerical values have to be related to crop type, soil type, climate, and management factors, and their intricate relationships.

PRINCIPAL RESPONSES OF PLANTS TO SALINITY A scheme that illustrates the effects of salinity and sodicity on plants is shown in Fig. 6-1. The effects of salinity on plants are due to one or the other of two properties of saline media, but usually both are implicated. Salinity depresses the external water potential (osmotic effect), and

172

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 6-1. Effects of salinity and sodicity on plants.

particular ions in the solution may have chemical or specific-ion effects. The left and right parts of Fig. 6-1 suggest that some effects of salinity on plants are not deleterious. With few exceptions, however, saline conditions adversely affect crops. Depression of the external osmotic potential by high salt concentrations tends to narrow the gap between the external and internal water potentials. At high salinities, the external osmotic potential may be depressed below that of the cell water potential, resulting in osmotic desiccation. However, even in less extreme situations, the water availability to the plant will at least initially tend to be lessened. The reduction in the osmotic potential of the medium is one of the primary causes of the adverse effects of salinity on plant growth (Maas and Nieman 1978). According to Munns (2002a, 2005; see also Fig. 1 in Läuchli and Grattan 2007), plants show a “two-phase growth response to salinity.” The first phase of growth reduction occurs within minutes after exposure to salinity and is due to an osmotic effect. The second and slower phase is a specific-ion effect which, if it occurs, may take days, weeks, or months and can lead to salt toxicity in the plant, primarily in the older leaves. In addi-

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

173

tion, certain trace elements, such as B, Se, or Mo, may reach high enough concentrations in the soil solution to cause specific toxicities in some plants or accumulate to levels that could pose a health threat to the consumer. As suggested in Fig. 6-1, these specific-ion effects may be differentiated under three categories. First, high concentrations of a given ion may cause mineral nutrition disorders. For example, high Na concentrations may cause deficiencies of other elements, such as K or Ca. Second, certain ions, such as Na or Cl, may have toxic effects that may not always be clearly distinguishable from deficiencies. Third, there may be specific-ion effects that promote the growth or qualitative features of the plant. Sodicity is due to a high activity of Na in the soil solution, relative to activities of Ca2 and Mg2. Sodicity may, therefore, cause two of the effects discussed in connection with salinity: disturbances of mineral nutrition, and toxicity. Sodicity will not elicit strictly osmotic effects. In addition, sodicity affects soil physical properties that lead to crusting, reduced infiltration, and reduced aeration, which can also adversely affect plants. For detailed considerations of salinity and sodicity, refer to Chapter 5. As already mentioned, salinity may also positively affect plant growth and composition. It may promote the growth of halophytes (Flowers et al. 1977) or enhance quality of crops, such as improved freezing tolerance of citrus (Syvertsen and Yelenosky 1988), increased sugar content in carrots, increased soluble solids in tomatoes and melons, and improved grain quality in durum wheat (Maas and Grattan 1999).

MECHANISMS OF RESPONSE Overall, salt tolerance in plants is determined by three distinct features: (1) osmotic tolerance, (2) ability to exclude Na or Cl, and (3) ability of the tissue to tolerate high concentrations of Na or Cl (Munns and Tester 2008). Osmotic Effects If the osmotic potential of the medium were to immediately drop lower than that of the plant’s cells, the plant’s cells would suffer osmotic desiccation. To survive, the plant must adjust osmotically, that is, in a medium of a high salinity it must build up even higher internal solute concentrations. Osmotic tolerance can be achieved by either the absorption of ions from the medium, or synthesis and accumulation of organic solutes. Both processes occur, but the extent to which one process dominates over the other is dependent on the type of plant and level of salinity.

174

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The salt-accumulating halophytes are adapted by absorbing salt from the medium and using it as their major internal osmoticum (Flowers et al. 1977). However, salt in plant cells can be toxic. Much evidence (Greenway and Munns 1980; Munns et al. 1983; Wyn Jones 1981) indicates that high salt concentrations in the cytoplasm damage enzymes and organelles. Salt from the medium apparently serves as an osmoticum in the vacuole, which represents a large fraction of the total cell volume. In the cytoplasm, the function of osmotic adjustment is served mainly by organic solutes synthesized by the plant (Rhodes et al. 2002; Wyn Jones and Gorham 1983, 2002). Thus, organic osmolytes are used to a large extent in only a small fraction of the total cell volume. The tonoplast must transport salt into the vacuole where it can concentrate while, at the same time, prevent any substantial leakage of organic osmolytes from the cytoplasm into the vacuole. A model of intracellular solute compartmentation is shown, for example, in Fig. 1 of Wyn Jones and Gorham (2002), and ranges of ion concentrations in cytoplasm and vacuoles of plants under salinity are presented by those authors in Fig. 2 of the same publication.

FIGURE 6-2. Relationship between leaf Na concentration and the estimated K/Na ratio in the cytoplasm of leaves of barley (cv. Franklin) and durum wheat (cv. Wollaroi) grown in a range of high salinities leading to different leaf Na concentrations. From Munns et al. (2006) with permission from Oxford University Press.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

175

The glycophytes are unable to cause the sharp, asymmetrical, intracellular compartmentation of inorganic and organic solutes that is central in the adaptation of salt-absorbing halophytes to their saline habitats. Exposed to moderate salinities, glycophytes tend to exclude Na or Cl and to sequester what salt they absorb in the roots and stems. Salt exclusion minimizes the exposure of the leaf cells and, hence, the photosynthetic apparatus, to salt. This aspect of salt tolerance is discussed in more detail in the “Integration in the Whole Plant” section. Thus, the glycophytes regulate ion fluxes less effectively at the cellular level than do halophytes, but partition ions more effectively at the level of the organ and tissue. Salt compartmentation in vacuoles is typical for halophytes and is the primary cause for tissue tolerance. Barley, a relatively salt-tolerant crop, can tolerate Na concentrations up to 500 mM in the leaf tissue. In this species (as in halophytes), Na is primarily sequestered in the vacuoles, and in the cytoplasm favorable K:Na ratios at high leaf Na concentrations of 200 to 300 mM are maintained, in contrast to the salt-sensitive durum wheat (Fig. 6-2; Munns et al. 2006). Unable to survive saline conditions by absorbing major quantities of external ions for osmotic adjustment, glycophytes must synthesize organic osmolytes to a larger extent than plants that can use salt itself as a major osmoticum. Many such organic solutes have indeed been identified (Table 6-1). The term “compatible solutes” (or “compatible osmolytes”) is TABLE 6-1. Compatible Solutes (Osmolytes) of Higher Plants Class of Compounds (1)

Amino acid Quaternary ammonium compounds

Polyhydric alcohols Mono- and disaccharides Cyclitols

Tertiary sulfonium compounds

Major Compounds (2)

Proline Glycine betaine Proline betaine B-alaninebetaine Choline-O-Sulfate D-sorbitol D-mannitol Sucrose, glucose, fructose D-pinitol D-ononitol L-quebrachitol Myo-inositol Trehalose 3-dimethylsulphonio propionate

From Bray et al. (2000); Gorham et al. (1985); Hasegawa et al. (2000); Rhodes et al. (2002); Sacher and Staples (1985); Wyn Jones (1984); Yancey et al. (1982).

176

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

often applied to these organic compounds because of their compatibility with cytoplasmic entities and processes. They do not interfere with cellular metabolism even at high concentrations (Bray et al. 2000). There is evidence, however, that not all of these compatible solutes are, in fact, wholly benign (Wyn Jones 1984). Compatible solutes that commonly accumulate to high concentrations in plants are proline, glycine betaine, and sucrose (Munns and Tester 2008). In order to function as compatible osmolytes, these solutes are highly compartmentalized in plant cells. The vacuoles accumulate charged ions and solutes that would otherwise perturb metabolism if accumulated in the cytoplasm. Compatible solutes in the cytoplasm, however, allow the cytosol to be osmotically balanced with the vacuole. As an example, very little glycine betaine (1 mM) is contained in the vacuoles of spinach leaves under salt stress, but its concentration in the cytosol and chloroplasts may reach 300 mM (Bray et al. 2000). At low concentrations, these solutes can function as “osmoprotectants” (McNeil et al. 1999) by, for example, blocking the inhibition of enzyme activity caused by perturbing solutes and elevating the denaturation temperatures of proteins (Rhodes et al. 2002). The synthesis and transport of organic osmolytes, the transport of inorganic ions, and the compartmentation of organic and inorganic solutes as discussed all require the expenditure of metabolic energy (photosynthate) (Epstein 1980). Stavarek and Rains (1985) compared the performance of alfalfa cells selected for salt tolerance in cell culture with that of unselected (nontolerant) cells. At 170 mM NaCl, the unselected cells grew exceedingly slowly. The selected (salt-tolerant) cells at the same salinity differed from those in the control treatment (no salt) only in having a lag period of a few days. Thereafter, their growth (i.e., gain in dry weight) paralleled that of the cells in the control treatment. Measurements of carbohydrate metabolism of the salt-tolerant selected cells showed maintenance costs to be relatively low when the cells grew under saline conditions. The maintenance costs of the unselected cells increased with salinity. Even selected cells, however, must divert energy to the processes of ion transport, synthesis of organic osmotica, and compartmentation. These costs result in diminished growth per unit of substrate (sugars) utilized (Stavarek and Rains 1985; Yeo 1983). Plants vary greatly in the adjustment of their energy economy to the presence of salt (Schwarz and Gale 1981). Maintenance respiration rates usually increase at moderate salinities. The definition of “moderate,” however, depends on the salt tolerance of the plant. Raven (1985) concluded that the energy cost of the synthesis of compatible solutes leads to a potential growth penalty. About seven moles of ATP are needed to accumulate one mole of NaCl as an osmoticum in leaf cells. The ATP requirement for the synthesis or accumulation of solutes has been estimated by Raven (1985) to be about 3.5 moles for either Na or Cl, 41 for proline, 50 for glycine

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

177

betaine, and 52 for sucrose, respectively. Hence, the synthesis of organic, compatible solutes proceeds at the expense of growth, but it may allow the plant to survive (Munns and Tester 2008). Specific Ion Effects: Nutrition A universal feature of salinity is the presence of high concentrations or activities of certain ionic species, such as Na and Cl, in many salt-affected soils (Epstein and Rains 1987; Szabolcs 1989). The ratios of these ions to others may be quite high and may cause deficiencies of nutrient elements present at much lower concentrations. Thus, plant performance may be adversely affected by salinity-induced nutritional disorders, resulting from salinity effects on nutrient availability, competitive uptake, transport, or partitioning within the plant (Grattan and Grieve 1999a). For example, in the saline environments in which Na predominates over K, the plant’s paramount nutritional requirement is K in adequate amounts (Rains and Epstein 1967). A high degree of selectivity of the K transport mechanism was demonstrated in experiments with excised leaf tissue of the mangrove, Avicennia marina. Even so, this species may show evidence of salt-induced K deficiency (Ball et al. 1987). Saline conditions may inhibit the absorption of nitrate. Thus, in shortterm experiments (up to 12 hours) with barley seedlings, Aslam et al. (1984) found that SO42 and, to a greater extent, Cl diminish the rate of NO 3 absorption, the degree of inhibition being 83% at 0.2 M NaCl. The identity of the cation (Na or K) had little effect. Salt did not affect sub sequent NO 3 reduction by the plant. The inhibition of NO3 absorption by salt (0.2 M NaCl) was apparent within 1 minute of imposing the stress; recovery depended on a number of metabolic processes (Klobus et al. 1988). Sodium ions have been shown to cause disturbances in Ca nutrition. Nutritional disorders involving other elements may be linked to the effects of salinity on the transport and metabolism of Ca. High external Ca2 concentrations may mitigate the effects of salinity. High ratios of Na to Ca in the medium tend to be deleterious. Inadequate concentrations of Ca2 may adversely affect membrane function and growth within minutes of salt application (Cramer et al. 1988; Epstein 1961; Läuchli and Epstein 1970). Different genotypes may have widely different responses. Examples of salinity–Ca interactions follow. When LaHaye and Epstein (1969) grew highly salt-sensitive bean plants, Phaseolus vulgaris, in solution cultures at 50 mM NaCl and CaSO4 concentrations of 1 mM or less, they found that NaCl impaired the growth of the plants during the seven days of the experiment. Adding Ca2 at concentrations of 3 or 10 mM completely protected the plants from the adverse effects of NaCl. Similar findings were obtained with bean plants

178

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

grown to maturity (LaHaye and Epstein 1971). Effects of Ca2 on the performance of the plants under saline conditions have since been extensively investigated, especially its role in the integrity and unimpaired function of plant cell membranes and its displacement from membranes by Na ions present at high concentrations (Cramer et al. 1985, 1986; Lynch et al. 1987). Elzam and Epstein (1969) compared two species of wheat grass that were grown in nutrient solutions salinized with NaCl at concentrations ranging up to 500 mM. Salt concentrations that severely affected growth were 5.0 mM for the salt-sensitive species of wheat grass and 100 mM for the salt-tolerant species. In each species, the Ca concentration in the roots dropped precipitously at the same concentrations of salt, suggesting a causal relationship between the salt-induced Ca loss and the failure of the plants. In the experiments just described, the osmolarity of the nutrient solutions increased with increasing salinities, as did the Na/Ca2 ratios. Maas and Grieve (1987) compared the effects of exposing corn, Zea mays, to isosmotic solutions salinized at various Na/Ca2 ratios. At a high ratio (34.6/1 on a molar basis), the plants suffered from Ca deficiency. When that ratio was 5.7/1 or less, no Ca deficiency occurred. In similar experiments with corn, Plaut and Grieve (1988) found that at progressively lower Na/Ca2 ratios in the media, the photosynthetic rate (CO2 fixation) and water-use efficiency declined, the former possibly elicited in part by a Ca-induced Mg deficiency. Salinity–mineral nutrient relations in crops have been reviewed more extensively by Grattan and Grieve (1999a,b). Specific Ion Effects: Toxicity In a number of instances, toxicity of specific ions is supported by evidence from two sources. One is that moderate concentrations of Na, Cl, sulphate, or other ions reduce growth or cause specific injury. The other is that isosmotic solutions of different compositions may elicit significantly different responses. As might be expected, genotypes may differ in these responses, even within a species. Toxicities due to even moderate concentrations of some ions in saltaffected soils are most common in woody plants. Bernstein (1965) shows color photographs of severe leaf injury due to Na or Cl salts in several fruit crops. These crops have little ability to exclude Na or Cl from their leaves, and the plants are long-lived; hence, they often suffer toxicities at even moderate soil salinities. Salinity caused leaf injury identified as phosphate toxicity to soybean plants in experiments conducted by Grattan and Maas (1988). The extent to which salinity caused such leaf injury depended on the concentration of phosphate, the Ca2/Na ratio and the variety. The mechanism by

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

179

which moderate salinity accompanied by high phosphate concentrations elicits phosphate toxicity is as yet poorly understood (Treeby and van Steveninck 1988). An example from experiments with wheat shows the usefulness of isosmotic solutions of different ionic composition, as well as the importance of the genotype. Kingsbury and Epstein (1986) found that the growth of a salt-sensitive line of wheat was adversely affected by nutrient solutions containing high concentrations of Na (100 mM) but not by isosmotic solutions without Na at elevated concentrations. The growth of a line selected for salt tolerance was unimpaired by any of the solutions used. Termaat and Munns (1986) also used isosmotic solutions to investigate osmotic versus specific-ion effects.

INTEGRATION IN THE WHOLE PLANT Response to Salinity Stress during Plant Development For half a century it has been recognized that a crop’s sensitivity to salinity varies from one developmental growth stage to the next (Bernstein and Hayward 1958). The majority of the research indicates that most annual crops are tolerant at germination but are sensitive during emergence and early vegetative development (Läuchli and Epstein 1990; Maas and Grattan 1999). As plants mature, they become progressively more tolerant to salinity, particularly at later stages of development. Salinity affects both vegetative and reproductive developmental processes in the plant, but the effect on one process may be more than on the other. This is particularly important depending on whether the harvested organ of the crop is a stem, leaf, root, shoot, fruit, fiber, or grain (Läuchli and Grattan 2007). For example, if salinity affects the reproductive organs more than the vegetative organs, then it will have a greater negative effect on the plant when fruits and grains are the harvestable product. Conversely, if the vegetative organs, such as roots, tubers, or stems, are the harvestable product, then salinity detriments will be less. Germination and Seedling Emergence Although there are exceptions, such as sugar beet (Läuchli and Epstein 1990), most plants are tolerant during germination and can germinate under high-salinity conditions, including many that are rated as sensitive to salinity, such as corn (Maas et al. 1983), kenaf (Curtis and Läuchli 1985), Limonium (Carter et al. 2005), and tomato (Kurth et al. 1986b). However, salinity stress delays germination, even though the final percentage of germinated seeds will eventually be the same among moderate-salinity

180

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

treatments (Maas and Poss 1989a). However, further increases in salinity will eventually reduce the percentage of germinated seeds as well (Kent and Läuchli 1985; Mauromicale and Licandro 2002). Many crops are less tolerant to salinity during emergence as compared to the germination phase and, as a result, stand establishment is reduced (Maas and Grattan 1999). For example, cotton, a crop classified as salt-tolerant, is particularly sensitive to salinity after germination, and plant densities have been substantially reduced in fields previously irrigated with saline-sodic drainage water (Goyal et al. 1999; Mitchell et al. 2000; Rains et al. 1987). Interestingly, stand establishment of more salt-sensitive crops, including safflower (Goyal et al. 1999) and tomato (Mitchell et al. 2000), was not nearly as affected. Emergence studies have been conducted using different root media under various environmental conditions, making it difficult to interpret results and compare studies. For example, sodic waters (particularly those where NaCl was the sole salinizing salt) can cause a deterioration of soil physical conditions, thus reducing oxygen diffusion rates and increasing soil strength (Grattan and Oster 2003). This condition inadvertently adds additional abiotic stresses to the emerging seedlings as compared with other salinity studies under saline, nonsodic conditions. Vegetative Growth Most of the literature indicates that crops are particularly susceptible to salinity during the seedling and early vegetative growth stage, as compared to germination. Why is early vegetative development so susceptible to salinity? It is well known that salinity, even with an adequate supply of Ca, reduces shoot growth (particularly leaf area) more than root growth (Läuchli and Epstein 1990). However, inadequate Ca supply under saline conditions (i.e., saline-sodic conditions) can adversely affect membrane function and growth of the root within minutes (Cramer et al. 1988; Epstein 1961; Läuchli and Epstein 1970). When the Ca concentration was increased from 0.4 to 10 mM in a medium salinized with NaCl, root length of cotton seedlings was substantially increased, particularly at the higher salinity levels (Cramer et al. 1986). Moreover, root length at this increased level of Ca2 was maintained over a wide range of salinities. Conversely, shoot growth remained salt-sensitive (Kent and Läuchli 1985). These Ca additions favor cell elongation of cotton roots at the expense of radial cell growth, and cell production rates were maintained (Kurth et al. 1986a). Additional studies conclude that supplemental Ca2 alleviates the inhibitory effect of salt on cotton root growth by maintaining plasma membrane selectivity of K over Na (Läuchli 1990, 1999; Zhong and Läuchli 1994).

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

181

The effect of salt stress on shoot growth in many species can also be partly alleviated by supplemental Ca2 (Cramer 2002; Läuchli and Epstein 1990), particularly those exposed to high Na/Ca ratios where Ca deficiency in developing leaves may occur (Maas and Grieve 1987). The Ca status of the growing region of leaves is particularly sensitive to salt stress (Läuchli 1990). The importance of supplemental Ca2 in alleviating salt stress effects in the shoot, as demonstrated originally by LaHaye and Epstein (1971), has been clearly emphasized by Cramer (2002) and Munns (2002b), who recommended adding at least 5 to 10 mM Ca2 to the medium for salinities of 100 to 150 mM NaCl, to counteract the inhibitory effect of high Na concentrations on growth. Reduction in shoot growth by salinity is characterized by stunted shoots with reduced leaf area, but the final leaf size depends on both cell division and cell elongation (Läuchli and Epstein 1990). Although salinity can reduce cell numbers (Munns and Termaat 1986), leaf extension (controlled by cell elongation) has been found to be an extremely salt-sensitive process (Papp et al. 1983). Cell expansion is controlled by processes related to cell-wall extension and cellular water uptake (Cramer and Bowman 1993), the latter depends on the cell osmotic potential, s. Hu and Schmidhalter (2004) concluded that the reduction of leaf elongation by salinity may either be related to decreases in cell-wall extensibility or increases in yield threshold (see, for example, Cramer 1991; Neumann 1993). In most plants, when leaf elongation partly recovered after the initial rapid drop in the elongation rate upon salinization of the medium, osmotic adjustment occurred with the solute content in the leaf cells becoming higher under saline than nonsaline conditions. Although leaf expansion is highly responsive to salt stress by reductions in cell elongation, the leaf initiation rate appears less affected. Kriedemann (1986) found that the rate of leaf expansion in sugar beet decreased linearly with increased salinity while the rate of leaf initiation was unaffected. The relationship between salt sensitivity of leaf growth and the overall salt sensitivity of the plant was hypothesized by Munns and Termaat (1986) using the following model (Fig. 6-3). When NaCl stress is added, leaf expansion continues but at a slower rate (Fig. 6-3a). At the same time, ions begin to accumulate in the leaves in a linear fashion (Fig. 6-3b) until they reach a maximum concentration beyond which concentrations become lethal and the leaf dies. This first occurs in older leaves then sequentially in younger leaves as salt stress continues. Eventually, the rate at which leaves die is faster than the rate of leaf expansion, whereby the carbohydrate reserve in the plant is depleted (Fig. 6-3c). Leaves no longer expand once the carbohydrate reserves drop to the bare minimum required for further growth. As carbohydrate reserves are depleted yet further, the whole plant

182

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 6-3. Changes with time following exposure to NaCl: (a) total leaf area, (b) leaf mean ion concentration, and (c) total reserve carbohydrate content of a plant. The time scale varies with plant species and external NaCl concentration. From Munns and Termaat (1986) with permission from CSIRO. eventually dies. The time scale of this process will vary depending on the plant type or genotype and level of salt stress. The more salt-sensitive the plant and more severe the stress, the shorter the time scale. For the plant to survive salt stress, a salt balance in the leaves must be achieved and some minimum fraction of the total leaf area of the plant must remain healthy to continue photosynthesis and carbohydrate production. Cheeseman (1988) considered the control of Na acquisition and allocation in plants, on the one hand, and carbon allocation and use, on the other, to be critical components of salt tolerance. However, earlier work indicated that photosynthesis per se does not appear to determine plant growth under salt stress for barley (Munns et al. 1982; Rawson et al. 1988), sugar beet (Papp et al. 1983), sunflower (Rawson and Munns 1984), and kenaf (Curtis and Läuchli 1986; Curtis et al. 1988). Although stomatal conductance in response to salinity is commonly reduced, the rates of photosynthesis per leaf area are mostly unaffected (Munns and Tester 2008), as a consequence of salinity-induced higher chloroplast density per unit leaf area. However, photosynthesis expressed on a unit of chlorophyll is usu-

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

183

ally reduced. Overall, photosynthesis measured on a per plant basis is reduced by salinity due to the reduction in total leaf area of the plant (Munns and Tester 2008). Salt accumulation in the chloroplasts of leaf cells could have a direct toxic effect on photosynthesis, or else salt could reach high enough concentrations in the cytoplasm of leaf cells to inhibit enzymes related to carbohydrate metabolism. The influence of salinity on photosynthesis of halophytes and C4-plants has been reviewed in detail by Lovelock and Ball (2002) and by Lüttge (2002). Water relations Could a water deficit in the leaves limit shoot growth following salinization? Termaat et al. (1985) tested whether leaf turgor limits shoot growth in salt-stressed plants by growing plants in a saline soil within a pressure chamber, and applying a pressure to the soil media that raised the shoot water potential by the same amount as the pressure applied to the root. Experiments using barley, wheat, Egyptian clover, white clover, and white lupin indicated that increased leaf-water potential did not improve growth under saline conditions (Munns and Termaat 1986). These results indicate that leaf-water deficit does not limit shoot growth in nonhalophytes grown under saline conditions. The possible role of cell turgor in the response of leaf elongation is unclear. Thiel et al. (1988) determined turgor in leaves of salinized barley seedlings by using the micropressure probe method described by Zimmermann et al. (1980) and correlated the turgor values with leaf elongation rates. Adding NaCl to the medium caused the leaf elongation rate to immediately decline, followed by gradual recovery, the extent of which depended on the NaCl concentration. Salt stress reduced cellular turgor in expanding and expanded leaf tissue. The turgor of expanding tissue recovered quickly, but decreased turgor persisted in expanded tissue. These authors suggested that salinity reduced leaf elongation through osmotic effects on turgor in the short term, but the sustained component of reduction in leaf elongation was not caused by a leaf-water deficit in the long term. The latter part of this conclusion is in agreement with Munns and Termaat (1986). In contrast to the results of Thiel et al. (1988), Yeo et al. (1991) found that leaf elongation in rice grown under salinity declined without a change of turgor in the growing zone. It appears that the long-term reduction in leaf growth due to salinity is not caused by a leaf-water deficit (Fricke and Peters 2002; Munns et al. 2000). This conclusion is supported by recent unpublished work by J. S. Boyer, R. A. James, and R. Munns (summarized in Munns et al. 2006). These researchers used isopiestic psychrometry on wheat and barley and confirmed that tugor was unchanged when plants were exposed to a range of saline solutions. However, when

184

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

they determined the relative water content (RWC; a widely used method of measuring plant water status), they found RWC decreased under salinity. Earlier, James et al. (2002) and Rivelli et al. (2002) showed turgor (calculated from the difference between total water potential and osmotic potential in durum wheat) remained unchanged by salinity, whereas RWC decreased. However, RWC, when measured with the conventional method of detaching leaves and rehydrating in deionized water, cannot be used on salt-exposed plants because of osmotic adjustment as a consequence of higher solute content of the cells in a saline medium than in the absence of salinity. The increased cell solute content in the salt-exposed plants leads to a higher uptake of water than in control leaves, resulting in an apparent low RWC in the salinity treatment. Thus, the conventionally used method of determining RWC is not an indicator of turgor in the plants that are undergoing osmotic adjustment in response to saline exposure (Boyer, James, and Munns, unpublished data). Additional evidence reviewed recently in detail by Läuchli and Grattan (2007) also indicates no direct role of turgor in the response of leaf elongation to salinity. Indirectly, the evidence supports the conclusion of Munns et al. (2006) that hormonal signals are controlling shoot growth and that control of shoot growth under salinity is not mediated by leaf-water deficit or ion toxicity. Reproductive Growth The bulk of the research suggests that after the salt-sensitive early-vegetative growth stage, most crops become progressively more tolerant as the plants grow older (Läuchli and Grattan 2007). In experiments with wheat (Maas and Poss 1989a), sorghum (Maas et al. 1986), and cowpea (Maas and Poss 1989b), the duration of salinity stress was held constant, but the period of salt-stress imposition varied from one developmental stage to the next. These investigators found that these crops were most sensitive during vegetative and early reproductive stages, less sensitive during flowering, and least sensitive during the seed-filling stage. In all these studies, seed weight was the yield component of interest, but similar conclusions regarding growth stage sensitivity were obtained with both determinate crops (the grain crops) and indeterminate (cowpea) crops. Wheat and rice are not only two of the most important grain crops in the world, they also have been the most intensively studied agronomic crops regarding salt sensitivity at different growth stages. Extensive research has also been conducted on tomato [see review articles by Cuartero et al. (2006) and Cuartero and Fernandez-Munoz (1999)], but these leading grain crops are of particular interest not only because they vary so widely in salt tolerance but also because salinity affects their reproductive processes differently (Läuchli and Grattan 2007). Studies on wheat and rice were conducted in a variety of conditions, including the field,

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

185

greenhouse, and laboratory, to better understand detailed changes in developmental processes, as the plants endure various degrees of salt stress at different growth stages. Wheat A mature wheat plant is a consequence of sequential developmental processes that are characterized by changes in shoot apex morphology. The yield components, such as tillers per plant, number of spikelets per spike, and individual grain weights, are developed sequentially as the crop develops. It has long been known that salinity reduces the growth rate of the entire wheat plant and its specific organs, but it also affects plant development (Hu et al. 2005). The duration of plant development is also affected by salinity. Salt stress was shown to retard leaf development and tillering but hasten plant maturity (Maas and Poss 1989a). Salt stress, imposed while the shoot apex is in vegetative stage, can adversely affect spike development and decrease yields of wheat (Maas and Grieve 1990). When wheat was salt-stressed during spike differentiation, reproductive development was stimulated, but the number of spikelets was reduced. Those investigators found that salt stress accelerated the development of the shoot apex on the mainstem and decreased the number of spikelet primordia. Salt stress decreased the yield potential mostly by reducing the number of spikebearing tillers. Therefore, they concluded that salinity stress needs to be avoided prior to and during spikelet development on all tiller spikes if full yield potential is to be achieved. Rice Rice is one of the most important food crops in the world, both economically and nutritionally, and ranks among the most sensitive to salinity (Maas and Grattan 1999). Not only is rice considerably less tolerant to salinity than wheat, but salinity affects its reproductive development quite differently. Rice sensitivity to salinity varies considerably from one growth stage to the next. In terms of grain yield, rice is salt-tolerant during germination (Heenan et al. 1988), sensitive to salinity during emergence and early seedling growth, becomes more tolerant later on in vegetative development, and then becomes sensitive again during reproductive growth (Abdullah et al. 2001; Flowers and Yeo 1981; Khatun and Flowers 1995). The vegetative shoot biomass of rice, however, is often affected much less than reproductive growth (except for young seedlings) (Khatun and Flowers 1995; Munns et al. 2002). Field and greenhouse studies showed that salinity had a negative impact on stand establishment, adversely affected a number of yield components, and even delayed heading (Grattan et al.

186

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

2002). In one study, investigators found linear decreases in several yield components with increased salinity, including the percent of sterile florets, tillers per plant, and spikelets per panicle, which translated into larger reductions in grain weight per plant at a given salinity (Zeng and Shannon 2000, Fig. 6-4). However these investigators suggested that seedling emergence and early seedling growth stages were most sensitive to salinity. It has long been recognized that salinity can cause sterility in rice, particularly if imposed during pollination and fertilization (Pearson and Bernstein 1959). However, some cultivars are more susceptible that others (Akbar and Yabuno 1977), suggesting some genetic control (Khatun et al. 1995). Salinity’s effect on rice has resulted in delayed flowering, a decrease in the number of productive tillers and fertile florets per panicle, and a reduction in individual grain weight (Khatun et. al 1995; Lutts et al. 1995; Zeng and Shannon 2000). The reduction in number of spike-bearing tillers by salt stress during the vegetative and early reproductive development in most cereal crops appears to have a greater negative impact on grain yield than any other yield component. The time from planting to maturity in most cereal crops typically decreases with increased salinity (Grieve et al. 1993), but salinity has just the opposite effect on rice (Khatun et. al 1995; Lutts et al. 1995). When salinity was applied to wheat from seedling emergence, it had a profound influence on reproductive development (Grieve et al 1993). Leaf

FIGURE 6-4. Relationship between salinity and various yield components of rice (Oryza sativa L., cv M-202). Fertility is inversely proportional to sterility. From Grattan et al. (2002); originally adapted from Zeng and Shannon (2000). © 2002 Regents of the University of California.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

187

initiation rate decreased, even though the time of flag leaf initiation was unchanged, indicating that salinity had no influence on the timing of the transition from vegetative to reproductive development but greatly reduced the number of tillers and overall grain yield. Salt stress in rice can reduce seedling emergence and, when imposed at early vegetative stages, reduces tillers and grain-bearing panicles, leading to low yields. However, unlike wheat, certain rice cultivars can develop sterile spikelets by a mechanism that appears to be genetically controlled, leading to further grain yield losses. Regulations of Salt Balance in Leaves Cellular uptake of sodium Sodium levels in roots increase rapidly after plants have been exposed to salt stress. This is the result of several transport processes, including Na uptake, Na efflux to the apoplast and vacuole, and Na transport to the shoot (Maathuis 2007). These Na transport processes require the activity of several membrane transporters that are important outputs for signaling pathways. A recent model of signaling pathways for both cellular Na and Ca2 during salt stress is shown in Fig. 6-5

FIGURE 6-5. The SOS pathway in plant cells in response to salinity stress (see text). From Maathuis (2007) with kind permission from Springer Science  Business Media.

188

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(Maathuis 2007). Some of these transporters may also have sensing functions. Plant roots respond specifically to external Na within seconds (Knight et al. 1997), yet the mechanism of Na sensing is still unknown. One possibility is that a plasma-membrane protein may function as the extracellular Na sensor (Munns and Tester 2008). The first response to an increase in salt stress around the roots is an elevated Ca2 activity in the cytosol of root cells ([Ca2]cyt) (see, for example, Knight 2000; Lynch et al. 1989). This increase in free cytosol Ca2 is brought about by a flux of Ca2 into the cytosol across the plasma membrane and the tonoplast (Munns and Tester 2008). The increase in Ca2cyt in response to salt is a central component of the SOS (salt overly sensitive) pathway (see Fig. 6-5 and Zhu 2002). This Ca2cyt increase leads to Ca binding to SOS3, a Ca2 binding protein (Liu and Zhu 1997). The Ca2-activated SOS3 interacts with and activates the kinase SOS2 (Ward et al. 2003). In turn, the activated SOS2 phosphorylates and activates SOS1, causing an increase in SOS1 activity (Chinnusamy et al. 2004). SOS1 is a key Na transporter in salt stress. It functions as an Na/H antiporter, mediates Na extrusion from the cytoplasm to the outside of the cell, and is located at the plasma membrane. It is expressed in many tissues but particularly in the root epidermis and vascular tissues (Maathuis 2007). The regulation of SOS1 by SOS2 and SOS3 has been addressed in detail by Qiu et al. (2002). The exact role of SOS1 in salt tolerance is uncertain (Munns and Tester 2008), but the SOS pathway is important for some component of salinity tolerance because SOS mutants of Arabidopsis are more sensitive to salinity than wild-type plants (Zhu et al. 1998). The kinase SOS2 can also affect the activity of other transporters, such as NHX1, an Na/H antiporter at the tonoplast that sequesters Na  in the vacuole (Apse et al. 1999), and HKT1, a cation transporter involved in Na uptake at the plasma membrane (Laurie et al. 2002). The latter two transporters contribute to lowering the cytoplasmic Na concentration under conditions of salinity stress. As already mentioned (see Fig. 6-2), the maintenance of an adequate K concentration in the cytoplasm by favorable K:Na ratios under salt stress is an important feature of salt tolerance in plants. This feature is usually called K homeostasis and is important for maintaining cellular metabolism (Zhu 2003). Addition of K to the Arabidopsis SOS mutants alleviated the phenotype, possibly due to an increase in cytoplasmic K concentration in the root (Zhu 2002). A negative correlation between Nainduced net K efflux and salinity tolerance in barley plants, was found by Chen et al. (2005). Chen et al. (2007) concluded that the cytosolic K/Na ratio is an important determinant of salinity tolerance, but its significance is probably located mainly in the root because no strong relationship between leaf K concentrations and salt tolerance has been found thus far (Munns and Tester 2008).

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

189

Calcium signaling in response to salt stress As explained in the previous section (and see Fig. 6-5), the cytosolic Ca2 activity is involved in the regulation of the Na concentration in the cytosol by functioning as a component of the Ca2-signaling network (Fig. 6-6, Hirschi 2004). In the absence of stress, the cytosolic Ca2 activity (“resting Ca2“) is low, remaining at about 0.1 to 0.2 M (Hirschi 2004). A stimulus, such as an increase in external Na concentration, elicits Ca2 mobilizing signals. These mobilizing signals trigger a flow of Ca2 into the cytoplasm, which generates the “on” events. One of these “on” events is the activation of the Na/H antiporters at the plasma membrane and tonoplast, which lower the elevated cytoplasmic Na concentration following exposure to salinity, to low and nontoxic levels. The “high” cytosolic Ca2 activates Ca2 sensitive processes, which are mediated by proteins such as calmodulin. The final step mediated by “off” mechanisms (e.g., transporters, binding proteins) removes Ca2 from the cytosol and restores the Ca2 activity at the resting state. This Ca2 signaling network is the cellular/subcellular basis for the important role that Ca2 plays in mediating a component of salt tolerance in plants. Salt exclusion Regulation of salt balance in the leaves is significant for the salt tolerance of the whole plant. Flowers and Yeo (1986) concluded that for salt-

FIGURE 6-6. Ca2 signaling network in response to a stimulus, such as salinity stress. Adapted from Hirschi (2004).

190

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

sensitive species, such as rice, Na supply to the expanding leaves exceeds the demand for osmotic adjustment at high salt stress, leading to excessive Na concentrations in the leaves. Sodium could accumulate in the cell walls of leaves, as proposed by Oertli (1968). However, recent work showed that Na concentrations in the leaf cell-walls of maize and cotton remained too low to cause a decline in leaf growth under salinity (Mühling and Läuchli 2002a,b). Alternatively, Na reaches excessive intracellular concentrations, leading to ion toxicity. It is also well established that salt concentrations are highest in the oldest leaves of nonhalophytes (Greenway and Munns 1980). Yeo and Flowers (1982) suggested that lower salt concentrations in younger leaves may be partly due to exclusion of specific ions from the xylem vessels that supply them. In rice, the evidence clearly indicates that, upon salinization, Na moves first to the older leaves, which in effect protects the younger leaves. This partial exclusion of Na from the young leaves appears to correlate positively with the salt tolerance of some rice cultivars relative to the germplasm base (Yeo and Flowers 1986). Net Na accumulation in the shoot is the result of several different transport processes in the root and the shoot (Fig. 6-7, Davenport et al. 2007): (1) under conditions of high salinity, Na is transported passively into the root cytosol, probably by nonselective cation channels; (2) antiporters, such as SOS1, pump Na back from the outer region of the root to

FIGURE 6-7. Involvement of different Na transport processes of the root and the shoot in net Na accumulation in the shoot. From Davenport et al. (2007) with permission from Wiley-Blackwell.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

191

the soil or (3) can sequester Na in the root cell vacuoles by NHX transporters; and (4) Na can also leak passively from the vacuole to the cytosol. These transport processes 1 through 4 are located in the root cortex. Transport of Na to the shoot is still poorly understood (Davenport et al. 2007). Other transport processes are in the stele of the root where Na is loaded into the xylem, either passively (5) or possibly actively by SOS1 (6). Transport steps 5 and 6 occur in the mature part of the root or at low transpiration. Na retrieval from the xylem, which can take place in the mature region of the root, base of the stem, and along the veins of the leaves, may also occur via the transport processes 5 and 6. Unloading of Na into the leaf cells (7) probably proceeds passively; and Na can also be recirculated from the leaves to the roots in the phloem (8, 9). The transport steps involved in the net delivery of Na to the xylem have been described in detail by Tester and Davenport (2003). More specifically, the controlling processes in Na transport in durum wheat have been investigated in detail (Davenport et al. 2005). Recently, the thermodynamics of the transport processes and their possible molecular mechanisms have been assessed and summarized by Munns and Tester (2008). What follows from the discussion in the previous section is that each of the transport processes that contribute to the net accumulation of Na in the shoot requires their control by specific cell types in specific locations of the plant, thus facilitating Na transport in a coordinated manner (Munns and Tester 2008; Tester and Davenport 2003). A recent experimental example of elucidating the involvement of specific cell types in governing the major control points that limit Na loading of the xylem is given by Läuchli et al. (2008) (see also Munns et al. 2006; Munns and Tester 2008). This study indicated that the highest concentration of Na in the cells of a relatively salt-tolerant durum wheat genotype was found in the pericycle, and therefore the pericycle may be the transport control point in the radial transport of Na, perhaps limiting xylem loading of Na in the root (Läuchli et al. 2008). The question arises as to how significant salt exclusion from the shoot is as a mechanism of salt tolerance in nonhalophytes. There is a strong correlation between salt exclusion and salt tolerance in many species (see Flowers and Yeo 1986; Munns et al. 2006; Läuchli 1984). In species that retain Na in the woody roots and stems, a strong correlation exists between Cl- exclusion from the leaves and salt tolerance in, for example, citrus (Storey and Walker 1999). In durum wheat genotypes, a negative correlation has been found between leaf Na concentration and salt tolerance measured as shoot dry matter production (Munns and James 2003). In the salt-tolerant wheat genotype, the transport of Na to the leaf blades was maintained at low rates and controlled by xylem loading in the root and sequestration of Na in the leaf sheath. In a recent review, Munns et al. (2006) convincingly argued that all plants must exclude most of the Na

192

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and Cl in the soil solution, or the salt concentration in the leaves becomes so high that the leaves are killed. The salt concentration in the soil solution, percent salt taken up by the roots, and percent water retained in the leaves are related as follows: [NaCl] shoot  [NaCl] soil  (%NaCl taken up/%water retained)

(6-2)

Plants take up about 50 times more water from the soil than they retain in the shoot; thus, they retain only about 2% of the cumulative water absorbed by roots. The vast majority of the water absorbed by roots is transpired by leaves. To prevent the shoot salt concentration exceeding that in the soil, only 2% of the salt should be allowed to move into the shoot; that is, 98% should be excluded (Munns et al. 2006). In fact, most plants exclude about 98% of the salt in the soil solution and only 2% is transported in the xylem to the shoots. Table 6-2 shows some examples of Na and Cl concentrations in the xylem and percent Na and Cl exclusion from the xylem in plants grown in the presence of 50 mM NaCl for two weeks (see Gregory 2006, Table 5.9; data from R. Munns by personal communication). The data in Table 6-2 show that bread wheat excludes Na more effectively than durum wheat, rice, and barley. Although barley is a relatively poor Na excluder, it is quite salt-tolerant because it shows a high degree

Table 6-2. Salt Concentrations in the Xylem and Percent Na and Cl Exclusion from the Xylem of Plants Grown in the Presence of 50 mM NaCl for 2 Weeks Na

Cl

Xylem Exclusion

Xylem Exclusion

Species (1)

Genotype (2)

Na (mM) (3)

(%) (4)

Cl (mM) (5)

(%) (6)

Bread wheat

Janz Chinese Spring Wollaroi Tamaroi IR 36 Clipper Rangpur lime

0.3 1.1 2.0 2.7 2.8 3.2 2.4

99 98 96 95 94 94 95

3.9 — 3.4 2.7 — 4.7 1.5

92 — 93 95 — 91 97

Rough lemon

3.3

93

2.6

95

Durum wheat Rice Barley Citrus

Adapted from Gregory (2006).

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

193

of “tissue tolerance” to Na (see Fig. 6-2). The Cl exclusion data in Table 6-2 do not correlate with salt tolerance as well as those for Na exclusion, but the relatively salt-tolerant citrus genotype Rangpur lime shows the highest percent of Cl exclusion. The use of wild relatives to improve salt tolerance of the corresponding crop species has recently been shown to have potential. For example, several wild relatives of wheat are halophytes and exhibit considerable Na exclusion and moderate Cl exclusion at high salinity (Colmer et al. 2006). Another recent example by Teakle et al. (2007) is the comparison of the two lotus species (perennial forage legumes), Lotus tenuis (relatively salt-tolerant) and L. corniculatus (less salttolerant). L. corniculatus showed much higher Cl concentrations in the xylem and shoot than did L. tenuis; thus, higher salt tolerance in L. tenuis was correlated with a higher degree of Cl exclusion. In irrigated agriculture under field conditions, surface irrigation (particularly on poorly drained land) can lead to an oxygen deficiency in the root zone of the soil (Meek et al. 1983). Since oxygen deficiency alone adversely affects the growth of most dryland species (Jackson and Drew 1984), the combined effects of salinity and oxygen deficiency would be expected to be particularly damaging. In corn, oxygen deficiency in the root medium under saline conditions led to a breakdown of Na exclusion from the shoots and simultaneously inhibited K transport (Drew and Läuchli 1985). Oxygen deficiency in the root zone interacts with salinity relative to salt exclusion from the shoot in solution culture (Drew et al. 1988) and under field conditions (Barrett-Lennard 1986, 2003). Does export of Na and Cl from leaves through the phloem effectively help to maintain a salt balance in the leaves of a salinized plant? Läuchli (1984) and Flowers and Yeo (1986) concluded that little evidence exists to support this. Munns et al. (1986) confirmed this conclusion with an experiment that involved salinized barley. They found phloem export of Na and Cl– from leaves to be only about 10% of the xylem import of those ions. In a more recent study on two genotypes of durum wheat that differ in rates of Na accumulation in the leaves, Na export from the shoot to the root via the phloem was very small (Davenport et al. 2005). Nevertheless, Na recirculation via the phloem may be of some importance in salt tolerance of halophytes (Tester and Davenport 2003). Hormonal Regulation of Salt Tolerance The integration of salt-tolerance processes in the plant and their regulation is clearly dependent on the involvement of signaling pathways and plant hormones. This integrated control is a highly specialized field and is beyond the scope of the present chapter. The interested reader is referred to recent reviews, such as those prepared by Hasegawa et al. (2000), Munns (2002b), Xiong (2007), and Zhu (2002).

194

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

OUTLOOK Since the 1990 edition of this manual, our knowledge and understanding of the physiological mechanisms of salt tolerance in plants has greatly increased. Even faster progress has been made in elucidating molecular and genetic aspects of salt tolerance in plants. What is needed now is a concerted effort in bringing the molecular and genetic plant biologists together with the physiologists and agronomists to integrate the crop responses to salinity at the whole plant and ecosystem levels, and to undertake future studies at realistic and field conditions (the reader is referred to the pertinent review by Munns 2005). In addition, we have argued (Läuchli and Grattan 2007) that crops in the field do not normally encounter single abiotic stresses, such as salinity, but do experience multiple stresses that can vary in type and intensity throughout the growing season. Our knowledge of mechanisms of crop responses to salinity and its interaction with other stresses is limited, and their study should be a future focus (see also Mittler 2006). In the past decade it has become popular to perform salt-tolerance experiments using Arabidopsis thaliana as a convenient and genetically well-understood model plant. However, this model plant is salt-sensitive (Sickler et al. 2007), and there is the rather convincing argument stating that mechanisms of salt tolerance should not be studied in salt-sensitive plants, because the principal features of salt tolerance may only be poorly expressed in salt-sensitive plants, if expressed at all. Fortunately, the recent discovery of a salt-tolerant relative to Arabidopsis, Thellungiella halophila (Munns and Tester 2008; Zhu 2001), a native of the saline soils along the coast of eastern China, opens up promising new opportunities. Although we have a good grasp of many basic salt-tolerance processes, novel and exciting discoveries continue to be made. Two examples are pertinent. A group of Japanese researchers described a novel salt-tolerance mechanism (Kanai et al. 2007). They found in the common reed (Phragmites australis), a salt-tolerant plant, retention of absorbed Na in the shoot base and preferential binding of Na to starch granules that may decrease cytosolic free Na and thus decrease Na toxicity in the cells of the shoot base of this plant. Na binding starch granules could constitute an important component of “tissue tolerance.” Another group of scientists recently discovered in the salt-tolerant bryophyte Physcomitrella patens, the presence of a plasma-membraneassociated Na-ATPase, which is involved in Na extrusion (Lunde et al. 2007). This is an exceptionally interesting and novel Na transporter, which is absent in higher vascular plants that rely on Na/H antiporters for Na extrusion from the cells. It may be conceivable that the expression of this Na-ATPase could be transferred to higher crop plants, making them more salt-tolerant.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

195

SUMMARY Increased salinity concentrations narrow the gap between the external and internal plant water potentials and may have direct toxicity effects. Plants have a variety of responses to increasing levels of salinity, including various methods of salt exclusion, partitioning of salts to older leaves, and other salt isolation/sequestration processes. The mechanisms of salt tolerance vary among crops. For most crops other than halophytes, salinity can (1) affect nutrient uptake, (2) cause direct toxic effects, (3) inhibit germination and seedling emergence, (4) slow growth, (5) affect reproductive growth, and (6) alter the salt balance in leaves. In the field, the adverse effects of salinity can be complicated by other stresses. Continuing research into the mechanisms of plant salt tolerance, combined with the ability to engineer genetic modifications, may suggest ways to increase salt tolerance in food crops.

REFERENCES Abdullah, Z., Khan, M. A., and Flowers, T. J. (2001). “Causes of sterility in seed set of rice under salinity stress.” J. Agron. Crop Sci., 187, 25–32. Akbar, M., and Yabuno, T. (1977). “Breeding for saline-resistant varieties of rice. IV. Inheritance of delayed-type panicle sterility induced by salinity.” Japan J. Breed., 27, 237–240. Apse, M. P., Aharon, G. S., Snedden, W. A., and Blumwald, E. (1999). “Salt tolerance conferred by overexpression of a vacuolar Na/H antiport in Arabidopsis.” Science, 285, 1256–1258. Aslam, M., Huffaker, R. C., and Rains, D. W. (1984). “Early effects of salinity on nitrate assimilation in barley seedlings.” Plant Physiol., 76, 321–325. Ball, M. C., Chow, W. S., and Anderson, J. M. (1987). “Salinity-induced potassium deficiency causes loss of functional photosystem II in leaves of the grey mangrove, Avicennia marina, through depletion of the atrazine-binding polypeptide.” Aust. J. Plant Physiol., 14, 351–361. Barrett-Lennard, E. G. (1986). “Effects of waterlogging on the growth and NaCl uptake by vascular plants under saline conditions.” Reclam. Reveg. Res., 5, 245–261. ———. (2003). “The interaction between waterlogging and salinity in higher plants: Causes, consequences and implications.” Plant Soil, 253, 35–54. Bernstein, L. (1965). Salt tolerance of fruit crops, U.S. Department of Agriculture, Agriculture Information Bulletin No. 292, USDA. Washington, D.C. Bernstein, L., and Hayward, H. E. (1958). “Physiology of salt tolerance.” Ann. Rev. Plant Physiol., 9, 25–46. Bohnert, H. J., Gong, Q., Li, P., and Ma, S. (2006). “Unraveling abiotic stress tolerance mechanisms: Getting genomics going.” Curr. Op. Plant Biol., 9, 180–188. Bray, E. A, Bailey-Serres J., and Weretilnyk, E. (2000). “Responses to abiotic stresses,” in Biochemistry and molecular biology of plants, B. Buchanan, W. Gouissem,

196

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and R. Jones, eds., American Society of Plant Physiologists, Rockville, Md., 1158–1203. Carter, C. T., Grieve, C. M., and Poss, J. P. (2005). “Salinity effects on emergence, survival, and ion accumulation of Limonium perezii.“ J. Plant Nutr., 28, 1243–1257. Chen, Z., Newman, I., Zhou, M., Mendham, N., Zhang, G., and Shabala, S. (2005). “Screening plants for salt tolerance by measuring K flux: A case study for barley.” Plant Cell Environ., 28, 1230–1246. Chen, Z., Pottosin, I. I., Cuin, T. A., Fuglsang, A. T., Tester, M., Jha, D., ZepedaJazo, I., Zhou, M., Palmgren, M. G., Newman, I. A., and Shabala, S. (2007). “Root plasma membrane transporters controlling K/Na homeostasis in salt-stressed barley.” Plant Physiol., 145, 1714–1725. Cheeseman, J. M. (1988). “Mechanisms of salinity tolerance in plants.” Plant Physiol., 87, 547–550. Chinnusamy, V., Schumaker, K., and Zhu, J.-K. (2004). “Molecular genetic perspectives on cross-talk and specificity in abiotic stress signaling in plants.” J. Ext. Bot., 55, 225–236. Colmer, T. D., Flowers, T. J., and Munns, R. (2006). “Use of wild relatives to improve salt tolerance in wheat.” J. Exp. Bot., 57, 1059–1078. Cramer, G. R. (1991). “Kinetics of maize leaf elongation. II. Responses of a sodium excluding cultivar and a Na including cultivar to varying Na/Ca salinity.” J. Exp. Bot., 43, 857–864. ———. (2002). “Sodium-calcium interactions under salinity stress,” in Salinity: environment-plants-molecule, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 205–227. Cramer, G. R., and Bowman, D. C. (1993). “Cell elongation control under stress conditions,” in Handbook of plant and crop stress, M. Pessarakli, ed., Marcel Dekker, Inc., New York, 303–319. Cramer, G. R., Epstein, E., and Läuchli, A. (1988). “Kinetics of root elongation of maize in response to short-term exposure to NaCl and elevated calcium concentration.” J. Expt. Bot., 39, 1513–1522. Cramer, G. R., Läuchli, A., and Epstein, E. (1986). “Effects of NaCl and CaCl2 on ion activities in complex nutrient solutions and root growth of cotton.” Plant Physiol., 81, 792–797. Cramer, G. R., Läuchli, A., and Polito, V. S. (1985). “Displacement of Ca2 from the plasmalemma of root cells. A primary response to salt stress?” Plant Physiol., 79, 207–211. Cuartero, J., Bolarin, M. C., Asins, M. J., and Moreno, V. (2006). “Increasing salt tolerance in the tomato.” J. Exp. Bot., 57, 1045–1058. Cuartero, J., and Fernandez-Munoz, R. (1999). “Tomato and salinity.” Sci. Hort., 78, 83–125. Curtis, P. S., and Läuchli, A. (1985). “Responses of kenaf to salt stress: Germination and vegetative growth.” Crop Sci., 25, 944–949. ———. (1986). “The role of leaf-area development and photosynthetic capacity in determining growth of kenaf under moderate salt stress.” Aust. J. Plant Physiol., 13, 553–565. Curtis, P. S., Zhong, H. L., Läuchli, A., and Pearcy, R. W. (1988). “Carbohydrate availability, respiration and the growth of kenaf (Hibiscus cannabinus) under moderate salt stress.” Amer. J. Bot., 75, 1293–1297.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

197

Cushman, J. C., and Bohnert, H. J. (2002). “Induction of Crassulacean acid metabolism by salinity molecular aspects,” in Salinity: environment-plants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 361–393. Davenport, R., James, R. A., Zakrisson-Plogander, A., Tester, M., and Munns, R. (2005). “Control of sodium transport in durum wheat.” Plant Physiol., 137, 807–818. Davenport, R. J., Muñoz-Mayor, A., Jha, D., Essah, P.A., Rus, A., and Tester, M. (2007). “The Na transporter AtHKT1 controls xylem retrieval of Na in Arabidopsis.” Plant Cell Environ., 30, 497–507. Drew, M. C., Guenther, J., and Läuchli, A. (1988). “The combined effects of salinity and root anoxia on growth and net Na and K accumulation in Zea mays grown in solution culture.” Ann. Bot., 61, 41–53. Drew, M. C., and Läuchli, A. (1985). “Oxygen-dependent exclusion of sodium ions from shoots by roots of Zea mays (cv Pioneer 3906) in relation to salinity damage.” Plant Physiol., 79, 171–176. Elzam, O. E., and Epstein, E. (1969). “Salt relations of two grass species differing in salt tolerance. I. Growth and salt content at different salt concentrations.” Agrochimica, 13, 187–195. Epstein, E. (1961). “The essential role of calcium in selective cation transport by plant cells.” Plant Physiol., 36, 437–444. ———. (1980). “Responses of plants to saline environments,” in Genetic engineering of osmoregulation: Impact on plant productivity for food, chemicals, and energy, D. W. Rains, R. C. Valentine, and A. Hollaender, eds., Plenum Press, New York, 7–21. Epstein, E., and Rains, D. W. (1987). “Advances in salt tolerance.” Plant Soil, 99, 17–29. Flowers, T. J. (1985). “Physiology of halophytes.” Plant Soil, 89, 41–56. Flowers, T. J., Troke, P. F., and Yeo, A. R. (1977). “The mechanism of salt tolerance in halophytes.” Ann. Rev. Plant Physiol., 28, 89–121. Flowers, T. J., and Yeo, A. R. (1981). “Variability in the resistance of sodium chloride within rice (Oryza sativa L.) varieties.” New Phytol., 88, 363–373. ———. (1986). “Ion relations of plants under drought and salinity.” Aust. J. Plant Physiol., 13, 75–91. Fricke, W., and Peters, W. S. (2002). “The biophysics of leaf growth in salt-stressed barley: A study at the cell level.” Plant Physiol., 129, 1–15. Gorham, J. (1996). “Glycinebetaine is a major nitrogen-containing solute in the Malvaceae.” Phytochemistry, 43, 367–369. Goyal, S. S., Sharma S. K., Rains D. W., and Läuchli, A. (1999). “Long-term reuse of drainage waters of varying salinities for crop irrigation in a cotton-safflower rotation system in the San Joaquin Valley of California: A nine year study. I. Cotton (Gossypium hirsutum L.).” J. Crop Prod., 2, 181–213. Grattan, S. R., and Grieve, C. M. (1999a). “Salinity: Mineral nutrient relations in horticultural crops.” Sci. Hort., 78, 127–157. ———. (1999b). “Mineral nutrient acquisition and response by plants grown in saline environments,” in Handbook of plant and crop stress, 2nd ed., Chapter 9, M. Pessarakli, ed., Marcel Dekker, Inc., New York, 203–230. Grattan, S. R., and Maas, E. V. (1988). “Effect of salinity on leaf P accumulation and injury in soybean. II. Role of Cl and Na.” Plant Soil, 109, 65–71.

198

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Grattan, S. R., and Oster, J. D. (2003). “Use and reuse of saline-sodic waters for irrigation of crops,” in Crop production in saline environments: Global and integrative perspectives, S. S. Goyal, S. K. Sharma, and D. W. Rains, eds., Haworth Press, New York, 131–162. Grattan, S. R., Zeng, L., Shannon, M. C., and Roberts, S. R. (2002). “Rice is more sensitive to salinity than previously thought.” Calif. Agric., 56, 189–195. Greenway, H., and Munns, R. (1980). “Mechanisms of salt tolerance in nonhalophytes.” Ann. Rev. Plant Physiol., 31, 149–190. Gregory, P. J. (2006). Plant roots: Growth, activity and interaction with soils, Blackwell Publishing, Oxford, UK. Grieve, C. M., Lesch, S. M., Maas, E. V., and Francois, L. E. (1993). “Leaf and spikelet primordia initiation in salt-stressed wheat.” Crop Sci., 22, 1286–1294. Hasegawa, P. M., Bressan R. A., Zhu J.-K., and Bohnert, H. J. (2000). “Plant cellular and molecular responses to high salinity.” Ann. Rev. Plant Physiol. Plant Mol. Biol., 51, 463–499. Hirschi, K. D. (2004). “The calcium conundrum: Both versatile nutrient and specific signal.” Plant Physiol., 136, 2438–2442. Heenan, D. P., Lewin, L. G., and McCaffery, D. W. (1988). “Salinity tolerance in rice varieties at different growth stages.” Aust. J. Exp. Agric., 28, 343–349. Hu, Y., Fricke, W., and Schmidhalter, U. (2005). “Salinity and the growth of nonhalophytic grass leaves: The role of mineral nutrient distribution.” Funct. Plant Biol., 32, 973–985. Hu, Y., and Schmidhalter, U. (2004). “Limitation of salt stress to plant growth,” in Plant toxicology, B. Hock and E. F. Elstner, eds., Marcel Dekker, Inc., New York, 191–224. Jackson, M. B., and Drew, M. C. (1984). “Effects of flooding on growth and metabolism of herbaceous plants,” in Flooding and plant growth, T. T. Kozlowski, ed., Academic Press, Orlando, Fla., 47–128. James, R. A., Munns, R., and von Caemmerer, S. (2002). “Factors affecting CO2 assimilation, leaf injury and growth in salt-stressed durum wheat.” Funct. Plant Biol., 29, 1393–1403. Jurinak, J. J., and Suarez, D. L. (1990). “The chemistry of salt-affected soils and waters,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 42–63. Kanai, M., Higuchi, K., Hagihara, T., Konishi, T., Ishii, T., Fujita, N., Nakamura, Y., Maeda, Y., Toshiba, M., and Tadano, T. (2007). “Common reed produces starch granules at the shoot base in response to salt stress.” New Phytol., 176, 572–580. Kent, L. M., and Läuchli, A. (1985). “Germination and seedling growth of cotton: Salinity-calcium interactions.” Plant Cell Environ., 8, 155–159. Khatun, S., and Flowers, T. J. (1995). “Effects of salinity on seed set in rice.” Plant Cell Environ., 18, 61–67. Khatun, S., Rizzo, C.A., and Flowers, T. J. (1995). “Genotypic variation in the effect of salinity on fertility in rice.” Plant Soil, 173, 239–250. Kingsbury, R. W., and Epstein, E. (1986). “Salt sensitivity in wheat. A case for specific ion toxicity.” Plant Physiol., 80, 651–654.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

199

Klobus, G., Ward, M. R., and Huffaker, R. C. (1988). “Characteristics of injury and recovery of net No3– transport of barley seedlings from treatment of NaCl.” Plant Physiol., 87, 878–882. Knight, H. (2000). “Calcium signaling during abiotic stress in plants.” Int. Rev. Cytol., 192, 269–324. Knight, H., Trewavas, A. J., and Knight, M. R. (1997). “Calcium signaling in Arabidopsis thaliana responding to drought and salinity.” Plant J., 12, 1067–1078. Koiwa, H., Bressan R. A., and Hasegawa, P. M. (2006). “Identification of plant stress-responsive determinants in Arabidopsis by large-scale forward genetic screens.” J. Exp. Bot., 57, 1119–1128. Kriedemann, P. E. (1986). “Stomatal and photosynthetic limitations to leaf growth.” Aust. J. Plant Physiol., 13, 15–31. Kurth, E., Cramer, G. R., Läuchli, A., and Epstein, E. (1986a). “Effects of NaCl and CaCl2 on cell enlargement and cell production in cotton roots.” Plant Physiol., 82, 1102–1106. Kurth, E., Jensen, A., and Epstein, E. (1986b). “Resistance of fully imbibed tomato seeds to very high salinities.” Plant Cell Environ., 9, 667–676. LaHaye, P. A., and Epstein, E. (1969). “Salt toleration by plants: Enhancement with calcium.” Science, 166, 395–396. ———. (1971). “Calcium and salt toleration by bean plants.” Physiol. Plant., 25, 213–218. Läuchli, A. (1984). “Salt exclusion: An adaptation of legumes for crops and pastures under saline conditions,” in Salinity tolerance in plants: Strategies for crop improvement, R. C. Staples and G. H. Toenniessen, eds., John Wiley and Sons, New York, 171–187. ———. (1990). “Calcium, salinity and the plasma membrane,” in Calcium in plant growth and development, R. T. Leonard and P. K. Hepler, eds., American Society of Plant Physiologists Symposium Series, Vol. 4, American Society of Plant Physiologists, Rockville, Md., 26–35. ———. (1999). “Salinity-potassium interactions in crop plants,” in Frontiers in potassium nutrition, D. M. Oosterhuis and G. A. Berkovitz, eds., Potash and Phosphate Institute, Norcross, Ga., 71–76. Läuchli, A., and Epstein, E. (1970). “Transport of potassium and rubidium in plant roots: The significance of calcium.” Plant Physiol., 45, 639–641. ———. (1990). “Plant responses to saline and sodic conditions,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 113–137. Läuchli, A., and Grattan, S. R. (2007). “Plant growth and development under salinity stress,” in Advances in molecular breeding toward drought and salt tolerant crops, M. A. Jenks, P. A. Hasegawa, and S. M. Jain, eds., Springer-Verlag, Dordrecht, The Netherlands, 1–31. Läuchli A., James, R. A., Huang, C. X., McCully, M., and Munns, R. (2008). “Cellspecific localisation of Na in roots of durum wheat and possible control points for salt exclusion.” Plant Cell Environ., 31, 1565–1574. Laurie, S., Feeney, K. A., Maathuis, F. J. M., Heard, P. J., Brown, S. J., and Leigh, R. A. (2002). “A role for HKT1 in sodium uptake by wheat roots.” Plant J., 32, 139–149.

200

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Levy, G. J., Shainberg, I., and Miller, W. P. (1998). “Physical properties of sodic soils,” in Sodic soils: Distribution, properties, management and environmental consequences, M. E. Sumner and R. Naidu, eds., Oxford University Press, New York, 51–75. Liu, J. P., and Zhu, J.-K. (1997). “An Arabidopsis mutant that requires increased calcium for potassium nutrition and salt tolerance.” Proc. Natl. Acad. Sci. USA, 94, 14960–14971. Lovelock, C. E., and Ball, M. C (2002). “Influence of salinity on photosynthesis of halophytes,” in Salinity: Environment-plants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 315–339. Lunde, C., Drew, D. P., Jacobs, A. K., and Tester, M. (2007). “Exclusion of Na via sodium ATPase (PpENA1) ensures normal growth of Physcomitrella patens under moderate salt stress.” Plant Physiol., 144, 1786–1796. Lüttge, U. (1993). “The role of crassulacean acid metabolism (CAM) in the adaptation of plants to salinity.” New Phytol., 125, 59–71. ———. (2002). “Performance of plants with C4-carboxylation modes of photosynthesis under salinity,” in Salinity: Environment-plants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 341–360. Lutts, S., Kinet, J. M., and Bouharmont, J. (1995). “Changes in plant response to NaCl during development of rice (Oryza sative L.) varieties differing in salinity resistance.” J. Exp Bot., 46, 1843–1852. Lynch, J., Cramer, G. C., and Läuchli, A. (1987). “Salinity reduces membrane-associated calcium in corn root protoplasts.” Plant Physiol., 83, 390–394. Lynch, J., Polito, V. S., and Läuchli, A. (1989). Salinity stress increases cytoplasmic Ca activity in maize root protoplasts.” Plant Physiol., 90, 1271–1274. Maas, E. V., and Grattan, S. R. (1999). “Crop yields as affected by salinity,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Agronomy Monograph 38. American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America (ASA/CSSA/SSA), Madison, Wisc., 55–108. Maas, E. V., and Grieve, C. M. (1987). “Sodium-induced calcium deficiency in saltstressed corn.” Plant Cell Environ., 10, 559–564. ———. (1990). “Spike and leaf development in salt-stressed wheat.” Crop Sci., 30, 1309–1313. Maas, E. V., Hoffman, G. J., Chaba, G. D., Poss, J. A., and Shannon, M. C. (1983). “Salt sensitivity of corn at various growth stages.” Irrig. Sci., 4, 45–57. Maas, E. V., and Nieman, R. H. (1978). “Physiology of plant tolerance to salinity,” in Crop tolerance to suboptimal land conditions, G. A. Jung, ed., American Society of Agronomy, Madison, Wisc., 277–299. Maas, E. V., and Poss, J. A. (1989a). “Salt sensitivity of wheat at different growth stages.” Irrig. Sci., 10, 29–40. ———. (1989b). “Salt sensitivity of cowpea at various growth stages.” Irrig. Sci., 10, 313–320. Maas, E. V., Poss, J. A., and Hoffman, G. J. (1986). “Salinity sensitivity of sorghum at three growth stages.” Irrig. Sci., 7, 1–11. Maathuis, F. J. M. (2007). “Root signaling in response to drought and salinity,” in Advances in molecular breeding toward drought and salt tolerant crops, M. A. Jenks, P. A. Hasegawa, and S. M. Jain, eds., Springer-Verlag, Dordrecht, The Netherlands, 317–331.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

201

Manedov, A. I., Levy, G. J., Shainberg, I., and Letey, J. (2001). “Wetting rate, sodicity, and soil texture effects on infiltration rate and runoff.” Aust. J. Soil Res., 39, 1293–1305. Mauromicale, G., and Licandro, P. (2002). “Salinity and temperature effects on germination, emergence and seedling growth of globe artichoke.” Agronomie, 22, 443–450. McNeal, B. L., and Coleman, N. T. (1966). “Effect of solution composition on soil hydraulic conductivity.” Soil Sci. Soc. Am. J., 30, 313–317. McNeil, S. D., Nuccio, M. L., and Hanson, A. D. (1999). “Betaines and related osmoprotectants. Targets for metabolic engineering of stress resistance.” Plant Physiol., 120, 945–949. Meek, B. D., Ehlig, C. F., Stolzy, L.H., and Graham, L. E. (1983). “Furrow and trickle irrigation: Effects on soil oxygen and ethylene and tomato yield.” Soil Sci. Soc. Am. J., 47, 631–635. Mitchell, J. P., Shennan, C., Singer, M. J., Peters, D. W., Miller, R. O., Prichard, T., Grattan, S. R., Rhoades, J. D., May, D. M., and Munk, D. S. (2000). “Impacts of gypsum and winter cover crops on soil physical properties and crop productivity when irrigated with saline water.” Agric. Water Mgmt., 45, 55–71. Mittler, R. (2006). “Abiotic stress: The field environment and stress combination.” Trends Plant Sci., 11, 15–19. Mühling, K. H., and Läuchli, A. (2002a). “Effect of salt stress on growth and cation compartmentation in leaves of two plant species differing in salt tolerance.” J. Plant Physiol., 159, 137–146. ———. (2002b). “Determination of apoplastic Na in intact leaves of cotton by in vivo fluorescence ratio-imaging.” Funct. Plant Biol., 29, 1491–1499. Munns, R. (2002a). “Comparative physiology of salt and water stress.” Plant Cell Environ., 25, 239–250. ———. (2002b). “Salinity, growth and phytohormones,” in Salinity: Environmentplants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 271–290. ———. (2005). “Genes and salt tolerance: Bringing them together.” New Phytol., 167, 645–663. Munns, R., Fisher, D. B., and Tonnet, M. L. (1986). “Na and Cl– transport in the phloem from leaves of NaCl-treated barley.” Aust. J. Plant Physiol., 13, 757–766. Munns, R., Greenway, H., Delane, R., and Gibbs, J. (1982). “Ion concentrations and carbohydrate status of the elongating leaf tissue of Hordeum vulgare growing at high external NaCl. II. Cause of the growth reduction.” J. Exp. Bot., 33, 574–583. Munns, R., Greenway, H., and Kirst, G. O. (1983). “Halotolerant eukaryotes,” in Encylopedia of plant physiology new series, Vol. 12C, Physiological plant ecology III, O. L. Lange, P. S. Nobel, C. B. Osmond, and H. Ziegler, eds., Springer-Verlag, Berlin, 59–135. Munns, R., Guo, J., Passioura, J. B., and Cramer, G. R. (2000). “Leaf water status controls daytime but not daily rates of leaf expansion in salt-treated barley.” Aust. J. Plant Physiol., 27, 949–957. Munns, R., Husain, S., Rivelli, A. R., James, R. A., Condon, A. G., Lindsay, M. P., Lagudah, E. S., Schachtman D. P., and Hare, R. A. (2002). “Avenues for

202

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

increasing salt tolerance of crops, and the role of physiologically based selection traits.“ Plant Soil, 247, 93–105. Munns, R., and James, R. A. (2003). “Screening methods for salinity tolerance: A case study with tetraploid wheat.” Plant Soil, 253, 201–218. Munns, R., James, R. A., and Läuchli, A. (2006). “Approaches to increasing the salt tolerance of wheat and other cereals.” J. Exp. Bot., 57, 1025–1043. Munns, R., and Termaat, A. (1986). “Whole plant responses to salinity.” Aust. J. Plant Physiol., 13, 143–160. Munns, R., and Tester, M. (2008). “Mechanisms of salinity tolerance.” Ann. Rev. Plant Biol., 59, 651–685. Neumann, P. M. (1993). “Rapid and reversible modifications of extension capacity of cell walls in elongating maize leaf tissues responding to root addition and removal of NaCl.” Plant Cell Environ., 16, 1107–1114. Oster, J. D., and Schroer, F. (1979). “Infiltration as influenced by irrigation water quality.” Soil Sci. Am. J., 43, 444–447. Papp, J. C., Ball, M. C., and Terry, N. (1983). “A comparative study of the effects of NaCl salinity on respiration, photosynthesis and leaf extension growth in Beta vulgaris (sugar beet).” Plant Cell Environ., 6, 675–677. Pearson, G. A., and Bernstein, L. (1959). “Salinity effects at several growth stages of rice.” Agron. J., 51, 654–657. Plaut, Z., and Grieve, C. M. (1988). “Photosynthesis of salt-stressed maize as influenced by Ca:Na ratios in the nutrient solution.” Plant Soil, 105, 283–286. Qiu, Q. S., Guo, Y., Dietrich, M. A., Schumaker, K. S., and Zhu, J.-K. (2002). “Regulation of SOS1, a plasma membrane Na/H exchanger in Arabidopsis thaliana, by SOS2 and SOS3.” Proc. Natl. Acad. Sci. USA, 99, 8436–8441. Quirk, J. P. (2001). “The significance of the threshold and turbidity concentrations in relation to sodicity and microstructure.” Aust. J. Soil Res., 39, 1185–1217. Rains, D. W., and Epstein, E. (1967). “Preferential absorption of potassium by leaf tissue of the mangrove, Avicennia marina: An aspect of halophytic competence in coping with salt.” Aust. J. Bio. Sci., 20, 847–857. Rains, D. W., Goyal, S., Weyrauch, R., and Läuchli, A. (1987). “Saline drainage water reuse in a cotton rotation system.” Calif. Agric., 41, 24–26. Raven, J. A. (1985). “Regulation of pH and generation of osmolarity in vascular plants: A cost-benefit analysis in relation to efficiency of use of energy, nitrogen and water.” New Phytol., 101, 25–77. Rawson, H. M., Long, M. J., and Munns, R. (1988). “Growth and development in NaCl-treated plants I. Leaf Na and Cl– concentrations do not determine gas exchange of leaf blades in barley.” Aust. J. Plant Physiol., 15, 519–527. Rawson, H. M., and Munns, R. (1984). “Leaf expansion in sunflower as influenced by salinity and short-term changes in carbon fixation.” Plant Cell Environ., 7, 207–213. Rhodes, D., Nadolska-Orczyk, A., and Rich, P. J. (2002). “Salinity, osmolytes and compatible solutes,” in Salinity: Environment-plants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 181–204. Richards, L. A., ed. (1954). Diagnosis and improvement of saline and alkali soils, Agriculture Handbook No. 60, U.S. Department of Agriculture, Washington, D.C.

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

203

Rivelli, A. R., James, R. A., Munns, R., and Condon, A. G. (2002). “Effect of salinity on water relations and growth of wheat genotypes with contrasting sodium uptake.” Funct. Plant Biol., 29, 1065–1074. Schwarz, M., and Gale, J. (1981). “Maintenance respiration and carbon balance of plants at low levels of sodium chloride salinity.” J. Exp. Bot., 32, 933–941. Shainberg, I., Levy, G. J., Goldstein, D., Manedov, A. I., and Letey, J. (2001). “Prewetting rate and sodicity effects on the hydraulic conductivity of soils.” Aust. J. Soil Res., 39, 1279–1291. Sickler, C. M., Edwards, G. E., Kiirats, O., Gao, Z., and Loescher, W. (2007). “Response of mannitol-producing Arabidopsis thaliana to abiotic stress.” Funct. Plant Biol., 34, 382–391. Stavarek, S. J., and Rains, D. W. (1985). “Effect of salinity on growth and maintenance costs of plant cells,” in Cellular and molecular biology of plant stress, J. L. Key and T. Kosuge, eds., Alan R. Liss, Inc., New York, 129–143. Storey, R., and Walker, R. R. (1999). “Citrus and salinity.” Sci. Hort., 78, 39–81. Syvertsen, J. P., and Yelenosky, G. (1988). “Salinity can enhance freeze tolerance of citrus rootstock seedlings by modifying growth, water relations, and mineral nutrition.” J. Amer. Soc. Hort. Sci., 113, 889–893. Szabolcs, I. (1989). Salt-affected soils, CRC Press, Boca Raton, Fla. Teakle, N. L., Flowers, T. J., Real, D., and Colmer, T. D. (2007). “Lotus tenuis tolerates the interactive effects of salinity and waterlogging by excluding Na and Cl- from the xylem.” J. Exp. Bot., 58, 2169–2180. Termaat, A., and Munns, R. (1986). “Use of concentrated macronutrient solutions to separate osmotic from NaCl-specific effects on plant growth.” Aust. J. Plant Physiol., 13, 509–522. Termaat, A., Passioura, J. B., and Munns, R. (1985). “Shoot turgor does not limit shoot growth of NaCl affected wheat and barley.” Plant Physiol., 77, 869–872. Tester, M., and Davenport, R. (2003). “Na tolerance and Na transport in higher plants.” Ann. Bot., 91, 503–527. Thiel, G., Lynch, J., and Läuchli, A. (1988). “Short-term effects of salinity stress on the turgor and elongation of growing barley leaves.” J. Plant Physiol., 132, 38–44. Treeby, M. T., and van Steveninck, R. F. M. (1988). “Effects of salinity and phosphate on ion distribution in lupin leaflets.” Physiol. Plant., 73, 317–322. Ward, J. M., Hirschi, K. D., and Sze, H. (2003). “Plants pass the salt.” Trends Plant Sci., 8, 200–201. Wyn Jones, R. G. (1981). “Salt tolerance,” in Physiological processes limiting plant productivity, C. B. Johnson, ed., Butterworth, London, 271–292. ———. (1984). “Phytochemical aspects of osmotic adaptation,” in Phytochemical adaptation to stress, B. N. Timmermann, C. Steelink, and F. A. Loewus, eds., Plenum Press, New York, 55–78. Wyn Jones, R. G., and Gorham, J. (1983). “Osmoregulation,” in Encyclopedia of plant physiology new series, Vol. 12C, Physiological plant ecology III, O. L. Lange, P. S. Nobel, C. B. Osmond, and H. Ziegler, eds., Springer-Verlag, Berlin, 35–58. ———. (2002). “Intra- and inter-cellular compartmentation of ions,” in Salinity: Environment-plants-molecules, A. Läuchli and U. Lüttge, eds., Kluwer Academic Publishers, Dordrecht, The Netherlands, 159–180.

204

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Xiong, L. (2007). “Abscisic acid in plant response and adaptation to drought and salt stress,” in Advances in molecular breeding toward salinity and drought tolerance, M. A. Jenks, P. A. Hasegawa, and S. M. Jain, eds., Springer-Verlag, Berlin, 193–221. Yeo, A. R. (1983). “Salinity resistance: Physiology and prices.” Physiol. Plant., 58, 214–222. Yeo, A. R., and Flowers, T. J. (1982). “Accumulation and localization of sodium ions within the shoots of rice (Oryza sativa) varieties differing in salinity resistance.” Physiol. Plant., 56, 343–348. ———. (1986). “Salinity resistance in rice (Oryza sativa L.) and a pyramiding approach to breeding varieties for saline soils.” Aust. J. Plant Physiol., 13, 161–173. Yeo A. R., Lee K. S., Izard P., Boursier P., and Flowers, T. J. (1991). “Short- and long-term effects of salinity on leaf growth in rice (Oryza sativa L.).” J. Exp. Bot., 42, 881–889. Zeng, L., and Shannon, M. C. (2000). “Salinity effects on seedling growth and yield components of rice.” Crop Sci., 40, 996–1003. Zhong, H., and Läuchli, A. (1994). “Spatial distribution of solutes, K, Na, and Ca and their deposition rates in the growth zone of primary cotton roots: Effects of NaCl and CaCl2.” Planta, 194, 34–41. Zhu, J.-K. (2001). “Plant salt tolerance.” Trends Plant Sci., 6, 66–71. ———. (2002). “Salt and drought stress signal transduction in plants.” Ann. Rev. Plant Biol., 53, 247–273. ———. (2003). “Regulation of ion homeostasis under salt stress.” Curr. Op. Plant Biol., 6, 441–445. Zhu, J.-K., Liu, J., and Xiong, L. (1998). “Genetic analysis of salt tolerance in Arabidopsis thaliana: Evidence of a critical role for potassium nutrition.” Plant Cell, 10, 1181–1192. Zimmermann, U., Hüsken, D., and Schulze, E. D. (1980). “Direct turgor measurement in individual leaf cells of Tradescantia virginiana.” Planta, 149, 445–453.

NOTATION ATP  adenosine triphosphate ATPase  enzyme that catalyzes the breakdown of ATP to ADP and phosphate EC  electrical conductivity ESP  exchangeable sodium percentage HC  hydraulic conductivity HKT1  cation transporter involved in Na and K uptake at the plasma membrane NHX1  Na/H antiporter at the tonoplast RWC relative water content SAR  sodium adsorption ratio SOS  “salt overly sensitive” pathway

PLANT RESPONSES TO SALINE AND SODIC CONDITIONS

205

SOS1 plasma membrane localized Na/H antiporter SOS2 protein kinase SOS3 Ca2 binding protein that activates and recruits SOS2 kinase to the plasma membrane s  solute potential

This page intentionally left blank

CHAPTER 7 DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS W. P. Chen, A. C. Chang, and A. L. Page

INTRODUCTION “Trace element” is used in the geochemical and biochemical literature to refer to a group of otherwise unrelated chemical elements that occur in nature at low concentrations. Interpreted in the broadest sense, “trace elements” encompass more than two-thirds of the 91 naturally occurring elements. The following elements are likely to be significant in agricultural soils and have the potential to affect terrestrial and aquatic biota, beneficially or adversely: • Boron (B), copper (Cu), iron (Fe), manganese (Mn), molybdenum (Mo), nickel (Ni), selenium (Se), and zinc (Zn) are essential for higher plants and may become toxic at higher concentrations. • Chromium (Cr), nickel (Ni), and selenium (Se) in trace amounts are essential for metabolism in mammals and may become toxic at slightly higher concentrations. • Arsenic (As), cadmium (Cd), mercury (Hg), and lead (Pb) have no known vital biological functions and are toxic to virtually all forms of life where they occur in available forms above the critical levels. Human activities, such as irrigation, fertilizer applications, crop harvests, atmospheric deposition from industrial emissions and automobiles, and waste disposal, may alter the distribution and content of trace elements in soils. Plants often experience deficiency levels in the growing medium that hinder plant development and growth or toxicity levels that lead to plant injuries and/or transfer of potentially harmful elements up the ecological echelon. For many trace elements, the margin of safety between beneficial 207

208

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and harmful is narrow (Oliver 1997; Alloway 2005; Shanker 2008). The onset of an environmental calamity is often ambiguous and adverse effects may be hidden for years before they suddenly blow up. Excess or deficient trace elements have caused disease around the world. For example, symptoms of fluoride toxicity in sheep were observed in Iceland about 1,000 years ago (Roholm 1937). Long-term exposure to food grown in selenium-deficient soils has for centuries caused the Kashan and Kaschin-Beck diseases crippling bone disorders throughout China (Atlas 1989). And millions of people who depend on crops from the iodine-deficient soils of eastern Africa are still susceptible to goiter (Jaffiol et al. 1992). The biogeochemical processes governing the fate and transport of trace elements in the terrestrial environment are being adapted to alleviate toxicities and enhance nutritional benefits of harvested crops (Baˇnuelos and Lin 2008). This chapter reviews trace elements in soils and soil factors that influence their availability and accumulation by crops. Their fate and transport in soil are then evaluated by a generalized trace element mass balance model.

FACTORS INFLUENCING TRACE ELEMENT DEFICIENCIES AND TOXICITIES OF PLANTS Nutritional deficiencies and elemental toxicities of crops are related to a number of soil and plant factors (Morrissey and Guerinot 2009). The physical, mineralogical, chemical, and biological properties of soils govern the concentration of an element in the soil solution that, along with the rate at which the element can be replenished, governs the element’s bioavailability. Species respond differently to the same concentration of an element in the soil solution. Hence, soil and plant factors are site-specific and must be evaluated on a case-by-case basis to diagnose soils for trace element deficiencies and toxicities. Soil Factors As the primary medium in which plants obtain nutrients and extract water, the soil significantly influences plant growth. Soil is a porous medium made up of mixtures of weathered mineral fragments, organic debris, and microorganisms. Air and water fill voids between the granular particles. Most important to plant nutrition and soil fertility is the extent to which soil can store, transmit, and retain water. The rate at which applied water enters the soil (infiltration), the rate at which it passes through the soil profile (leaching), and the soil’s water content strongly influence the chemical forms and concentrations of trace elements that are present.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

209

Chemical reactions in the soil can be described as a heterogeneous system with various solid phases existing more or less in equilibrium with chemical species in the soil solution and soil atmosphere. Influx and efflux of water, fluctuation in soil water content, the plant’s absorption of nutrients and excretion of metabolites, and microbial transformations constantly change this equilibrium. Because soil is a porous medium and biological activity is present, true chemical equilibrium is seldom, if ever, attained. Therefore, the chemical forms and concentrations in soil solutions are not easily, if ever, accurately predicted. Chemical characteristics of soils, such as concentrations and species of cations and anions, adsorption, oxidation/reduction potential, kinetics of precipitation and dissolution, and soil pH and buffering capacity influence trace elements in the soil solution. Adsorption of ions and molecules by the solids is a fundamental process governing chemical activities in soils. The extent of adsorption is related to the kind and amount of clay, organic matter, and amorphous oxides of Al, Fe, and Mn. The adsorption removes trace elements from the soil solution. The ability of soils to adsorb trace elements that occur in cationic or molecular form generally increases as their contents of clay and organic matter increase. The adsorption of trace elements occurring in anionic form tends to increase as the soil’s content of amorphous oxides of Fe, Al, and Mn increases. Chemical precipitation in soils results in the formation of sparingly soluble compounds and the removal of ions from the soil solution. Precipitation, a nucleation process, differs mechanistically from adsorption, a surface phenomenon. Because both remove ions from the solution, distinguishing the two is difficult (Sposito 1986). As most soil systems are undersaturated with known inorganic solid compounds of many trace elements, including Cd, Cu, Ni, and Zn, the solubility of trace elements in soils is thought to be governed by either adsorption or the formation of a mixed solid. In saline soils, because concentrations of inorganic ligands, like those of chloride, sulfate, bicarbonate, and carbonate, are orders of magnitude greater than those of trace elements, ion association that results in ion pairing of trace elements with these ligands is the rule, not the exception (Page et al. 1981). Ion association markedly affects the adsorption, precipitation, and concentration of trace elements in the soil solution. The moisture content of the soil governs the oxygen level, which, in turn, governs the reduction/oxidation (redox) potential of the soil. The redox potential of soils ranges from approximately 400 mV for a flooded soil in a strongly reduced environment, to 700 mV for a well-drained soil with oxidized environment. Under reducing conditions, most trace elements, including Cu, Ni, Cd, Pb, Hg, Se, and Zn, have a very low solubility due mainly to the formation of metallic sulfides, such as sulfides of Cu, Ni, Pb, and others, and reduced forms of metalloids, such as selenides.

210

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Because crops, with the exception of rice, cannot tolerate reduced soil conditions for an extended period of time, the redox potential’s effect on the chemistry of trace elements is transitory, depending on the rate at which the products of reduction are re-oxidized after the system returns to oxidative conditions. Minerals of Fe and Mn, however, are exceptions. They are more soluble in their reduced forms than in oxidized forms. The concentration of hydrogen ion (pH) in the soil solution affects all of the previously discussed soil properties, which, along with the hydrogen ion concentration, are interdependent. The surface charge on an inorganic solid in soil consists of a permanent component and a pH-dependent component. The permanent charge is created when an ion in the crystal structure of the soil mineral is substituted by an ion of about equal size and lesser valence, resulting in a charge imbalance. The association of hydrogen ions or dissociation of hydroxyl ions from inorganic and organic surfaces in aqueous solution creates positively charged surface sites capable of adsorbing negatively charged trace elements. Hydrogen and hydroxyl ions are adsorbed on inorganic and organic surfaces. They are also involved in the dissolution and precipitation of minerals and amorphous solids in soils. Through these types of reactions, the pH of the soil solution—more precisely, the activity of hydrogen ions—affects the concentrations of trace elements. The solubility of the cationic trace elements decreases with increasing pH, while the solubility of the anionic trace elements increases with increasing pH. Therefore, deficiencies of Cu, Fe, Mn, and Zn, rather than toxicities of these elements, commonly exist in saline soil and, more commonly, in saline-sodic and sodic soils due to their higher alkalinity. Phytotoxicities caused by excessive amounts of Al, Cd, Mn, and Ni in acid soils essentially do not exist in alkaline soils. Solubilities of the anionic trace elements increase with increasing pH, but seldom reach levels that are toxic to plants. The increased solubility and availability of the anionic Mo and Se species in alkaline soils may pose a problem, though, because certain plants absorb these elements in amounts that are harmful to animals that eat them. Boron is probably the most troublesome trace element in managing saline and alkaline soils in the United States, particularly in California (Nable et al. 1997). While low concentrations of B are essential for the growth of plants, it becomes phytotoxic at concentrations only slightly higher than that required for optimum growth. In soil solutions, B occurs mainly as undissociated boric acid [B(OH)3] and is much more soluble than the other trace elements of concern, such as As, Cd, Ni, and Cu. The water-soluble B pool is readily available for plant uptake (Tsadilas et al. 1994; Xu et al. 2001). Its deficiency in crop plants is widespread (Gupta 1993). However, in arid and semiarid irrigated regions, particularly in humid and temperate climate zones, B toxicity also occurs.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

211

Plant Factors Plants differ in their ability to tolerate trace elements and in their ability to absorb and accumulate trace elements. Table 7-1 summarizes the concentrations of essential and nonessential trace elements in selected crops grown in major producing regions throughout the United States (Wolnick et al. 1983a,b; 1985). This data set is representative of concentrations in crops entering the US marketplace, because plant samples for the study were collected from fields of commercial growers throughout the United States. Worldwide data are available in Kabata-Pendias and Pendias (2001) and Kabata-Pendias and Mukherjee (2007). Concentrations of Cu, Fe, Mn, Mo, and Zn in various species varied, respectively, by factors of about 5 (1.8 mg/kg to 11 mg/kg), 70 (3 mg/kg to 200 mg/kg), 12 (7 mg/kg to 81 mg/kg), 6 (0.098 mg/kg to 6.65 mg/kg), and 3 (15 mg/kg to 46 mg/kg). Similar patterns of variations among crops are observed for the nonessential trace elements Cd, Ni, Pb, and Se. Data presented in Table 7-1 represent crops grown on many different soils. The crops obviously contain different amounts of trace elements and are chemically, physically, and morphologically different. Even among crops grown on

TABLE 7-1. Median Concentrations of Essential and Nonessential Trace Elements in Raw Crops Grown in Major Producing Areas of the United States

Crop Species (1)

Trace Element Concentration (mg/kg, oven-dry weight)

Number of Observations (2)

Cu (3)

Fe (4)

Mn (5)

Mo (6)

Zn (7)

Cd (8)

Lettuce

150

6.1

57

31

0.25

46

Spinach

104

8.5 200

81

0.22

43

Tomato

231

11

48

15

0.30

Wheat

288

4.9

36

43

0.43

Sweet corn

268

1.8

18

7

0.16

Ni (9)

Pb (10)

Se (11)

0.435

—

0.19

0.039

0.80

1.1

0.53

—

22

0.22

0.84 0.027

29

0.036

—

25

0.008

0.26 0.009

0.014 0.082

0.02

Soybean

322

13

71

27

—

45

0.045

4.8

Rice

166

2.1

3

11

0.65

15

0.005

0.26 0.005

Carrot

207

4.7

27

12

0.098 20

0.16

0.41 0.055

Potato

297

4.4

20

7

0.19

15

0.14

—

Onion

228

3.6

13

9

0.14

16

0.09

0.32 0.038

Peanut

320

8.3

20

18

0.28

31

0.068

1.5

Data from Wolnick et al. (1983a,b; 1985).

0.036

0.025 0.040

— 0.19

— — 0.013 — 0.040

212

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the same soil, comparable variations in concentrations of trace elements in plant tissue may be observed. For instance, leafy vegetables, such as lettuce and spinach, are known to take up much more Cd than species, such as corn, from the same soil. For a given species, concentrations of trace elements in various parts of the plants and among different cultivars also vary. For comparison purposes, a plant uptake factor (PUF, the ratio of the concentration of an element in plant tissue to that in soil) is often calculated. A higher PUF indicates that one plant species has a greater tendency to absorb the trace element than another, or one element is more susceptible to plant uptake than another element. Table 7-2 summarizes the PUF of Cd and Pb for different plant species based on the California portion of a data set from a TABLE 7-2. Summary Statistics of the Plant Uptake Factor (PUF) of Cd and Pb for Root, Vegetable, and Grain Crops Grown in California, 1980sa Cadmium Crop (1)

Mean (2)

Min (3)

Potatob

1.051

0.029

Peanutb

2.408

0.130

Onion

0.265

0.23

Tomatoc

1.763

0.167

Cabbage

0.771

0.028

Lettucec

1.549

0.061

Sweet Corn

0.088

0.0042

Soybeand

0.390

0.046

Wheat

0.205

Cornd

0.153

b

c

d

d

d

Lead

Max (4)

n (5)

Mean (6)

Min (7)

6.279

305

0.0072

0.0004

0.367

305

320

0.0021

0.0001

0.033

306

255

0.007

0.047

0.577

255

182

0.003

0.0001

0.064

179

206

0.0076

0.0008

0.122

206

145

0.036

0.0002

0.485

145

1.179

244

0.0025

0.0001

0.0432

252

7.500

339

0.004

0.0002

0.0269

338

0.014

1.375

315

0.004

0.0001

0.04

298

0.0026

4.200

256

0.0025

0.001

0.4

281

0.0027

32.00 2.222 32.00 5.375 12.54

Max (8)

n (9)

Rice

0.128

3.4

142

0.0008

0.0001

0.007

148

Root crops

1.32

32.00

0.029

880

0.0053

0.5770

0.0001

866

Vegetable crops

1.32

32.00

0.028

533

0.0138

0.4850

0.0001

530

Grain crops

0.21

7.50

0.0026 1,296

0.0030

0.4000

0.0001 1,317

a

Based on data from a USDA soil survey during the 1980s (Holmgren et al. 1993)

b

Considered as root crops

c

Considered as vegetable crops

d

Considered as grain crops

PUF  concentration in plant/concentration in soil

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

213

nationwide cropland trace element investigation conducted by a USDA Soil Survey (Holmgren et al. 1993). The uptake potentials of leafy crops generally are greater than those of the root crops and, in turn, those of the grain crops. A single biological mechanism apparently does not cause the differences in nutrient absorption among species and within cultivars. Genetically controlled features of plants, morphological and anatomical differences between plants, and physiology of the ion transport mechanism may all be responsible (Chang et al. 1982). When a plant takes up more cations than anions from the soil solution to satisfy the charge balance, hydrogen ions are exuded by effluxing by the roots. The opposite effect is observed when the plant extracts more anions than cations: the liberation of acids or bases by plant roots changes the pH of the soil immediately adjacent to the root, thereby changing the solubility and availability of trace elements. Plants also exude, through roots, organic ligands that can make trace elements in soils complex and alter the availability of those elements to the crop (Gao et al. 2003; Inskeep and Comfort 1986; Koo et al. 2006; Mench et al. 1988; Mench and Martin 1991; Merckx et al. 1986).

METHODS FOR DIAGNOSING TRACE ELEMENT DEFICIENCIES AND TOXICITIES Plant tissues, or soil extracts, can be analyzed to diagnose whether a plant is experiencing deficiencies or toxicities of trace elements. These data are compared against criteria that define ranges of concentrations related to deficiencies and toxicities. Through this relationship, the concentrations of a trace element in plant tissue or a soil extract may be related to the soil’s ability to supply the plant with adequate amounts for optimum yield. A plant’s response to trace elements is affected by many factors specific to the species and the soil, and sometimes by synergistic or antagonistic chemical and biochemical interactions at the root–soil interface. These factors complicate the diagnosis of trace element deficiencies and toxicities. Soil Methods A soil’s ability to supply the trace elements needed by plants is related to not only the chemical forms and solubility of the element but also a series of reactions. The rate at which the plant extracts trace elements and the rate at which other trace-element pools in the soil system replenish the chemical form of the element determine the total plant uptake. Such uptake depends on the growth rate of the root. Though sufficient amounts may exist in the soil, the rate of release must be fast enough to supply the

214

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

trace element so as to avoid deficiencies and their symptoms. Ideally, the soil extractant used to indicate the availability of the trace element would account for contributions of all applicable physical, chemical, and biological processes in the soil related to the uptake of trace elements. Actually, no static extraction process can simulate these dynamic processes. Thus, the use of trace element extractants to evaluate the soil’s deficiency and toxicity potential is derived from observations on the amounts extracted and the corresponding plant responses (yield or crop quality). Therefore, levels of trace elements associated with deficiency or toxicity for one crop may not indicate a similar response in another crop. Calibration of the criteria for conditions specific to the site and the crop is advisable. Extractants used to diagnose trace element deficiencies and toxicities of crops include water, weak acids, strong acids, buffered and neutral salt solutions, and solutions of organic complexing compounds. They are more frequently used to diagnose deficiencies rather than toxicities. They often are specific to a species or group of species, for example, sensitive species or tolerant species, and a particular type of soil. In the western United States, two similar reagents, DTPA-TEA (diethylene triamine pentaacetic acid-triethanolamine) and DTPA-NH4HCO3, are most often used. In the rest of the nation, HCl and a mixed reagent containing HCl and H2SO4 are commonly used. Other methods involve the use of aqueous solutions of other organic complexing compounds (EDTA and EDDHA), weak acids (HOAc and HOOCCOOH), and buffered and unbuffered salts (NH4OAc, NH4NO3, and MgCl2). DTPA-TEA extraction method The DTPA-TEA extraction method, developed by Lindsay and Norvell (1969) is thought to be the most appropriate method for diagnosing the deficiencies of elements, such as Zn and Cu, in a soil (Brennan et al. 1993; Khan et al. 2005). When amounts of the trace elements, such as Cu, Fe, Zn, Mn, and Ni, in soils are adequate for plant growth, the concentrations extracted with DTPA-TEA from soils significantly correlate with the concentrations at which they occur in various species (Alvarez 2007; Basar 2005). Soltanpour and Schwab (1977) modified the DTPA-TEA method developed by Lindsay and Norvell (1969) by replacing CaCl2 with NH4HCO3 as the background electrolyte to include the extraction of anions of N, P, S, Mo, Se, and As. Soltanpour (1985) used DTPANH4HCO3 extraction to detect potential phytotoxicities due to B, Mo, Pb, Cd, Se, and As and to diagnose the plant nutrient status of NO3-N, P, K, S, Zn, Fe, Mn, and Cu in soils. Table 7-3 shows the concentrations of selected DTPA-NH4HCO3-extractable trace elements in soils that indicate a deficient and toxic condition.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

215

TABLE 7-3. Levels of Trace Elements Extracted from Soils with NH4HCO3-DTPA, Indicative of Possible Deficient, Adequate, or Phytotoxic Levels for Crops Extractable Level Indicative of: Element (1)

Deficiency (mg/kg) (2)

Adequacy (mg/kg) (3)

Phytotoxicity (mg/kg) (4)

Cu Fe Mn

0–0.2 0–3.0 0–0.5

0.3–0.5 3.1–5.0 0.6–1.0

0.5 5.0 1.0

Zn

0–0.2

0.3–0.5

0.5

Data from Soltanpour (1985).

HCl and HCl  H2SO4 Before the mid-1960s, methods were not sensitive enough to detect certain trace elements dissolved in water or neutral electrolytes. Researchers experimented with the use of strong acid extractants to help determine whether trace elements in a soil would be deficient or toxic to plants. Use of inorganic acids as extractants has the advantage of high removal efficiency at low pH, and a strong linear correlation between the amounts of element extracted (such as Zn) and the amount of uptake by plants is often observed. However, the strong acid can damage the soil matrix and sometime it cannot adequately predict the availability. Mehlich 3 extractant Recently, the soil-test method using Mehlich 3 extractant has received considerable attention as a candidate for use as a standard method. This extractant contains 0.2 M CH3COOH, 0.25 M NH4NO3, 0.015 M NH4F, 0.13 M HNO3, and 0.001 M EDTA. Mehlich 3 offers the advantages of simplicity and multi-element extracting ability. The latter is especially attractive since the extractant can simultaneously extract soil P, K, Ca, and Mg, as well as micronutrients Cu, Fe, Mn, and Zn, and improve significantly the efficiency of soil tests. Mehlich 3 has been demonstrated to correlate well with plant uptake of several macro- and micronutrients and agronomic yields (Allen et al. 1994). Other extractants A number of other reagents have been used to indicate levels of soil metals associated with deficiency or toxicity to crops. In all cases, the

216

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

reagent was apparently specific to the crop, the soil, or the element. In the United Kingdom, 0.5 M HOAc, and 0.05 M ETDA are commonly used to extract Cd, Cu, Pb, Ni, and Zn in soils and set the toxicity thresholds (Purves 1985). Solutions of neutral and buffered salts have also been used. Hot-water-soluble B is used most frequently to assess levels of B in soils. Berger and Truog (1939) first developed the method, which Baker and Mortensen (1966) modified to use an aqueous solution of 0.1% CaCl2 2H2O. The CaCl2 electrolyte flocculates the soil particles and facilitates filtration. As with other trace elements, crops differ in their responses to B. Certain crops grow well in soils where the level of B extracted with hot water is less than 0.1 mg/kg, while others suffer from a deficiency of B if the level of hot-water-soluble-B falls below 1 mg/kg. For crops sensitive to B, the margin of safety between levels of B in soils associated with deficiencies and toxicities to crops usually is narrow. Crops grown on saline soils more commonly suffer from B toxicities than from B deficiencies. Both the soil and irrigation water can contribute to excessive B in the growing medium. Maas (1984) developed a classification scheme for B tolerances of crops based on the threshold concentration, which is the maximum concentration of B in the soil solution that does not restrict the yield (Table 7-4). A few of these threshold values have been updated in Maas and Grattan (1999). The NaHCO3 reagent is commonly used on neutral and alkaline soils to assess the availability of phosphorus (P) to crops (Olson and Sommers 1982). Because the chemistry of P is somewhat similar to the chemistry of As, soil tests for available P could reasonably be expected to apply to available As. Woolson et al. (1971) demonstrated that yields of corn correlated with the logarithm of the concentration of As extracted from neutral and alkaline soils by 0.5 M NaHCO3. Cadmium extracted from soils with 1 M NH4OAc and 1 M NH4NO3, respectively, correlated with concentrations of Cd in radish and lettuce (John et al. 1972) and in radish grown in the tested soils (Symeonides and McRae 1977). Molybdenum extracted from soils by Tamm’s Reagent (ammonium oxalate-oxalic acid) appeared to correspond well with levels of Mo in the harvested forage crops (Pierzyneski and Jacobs 1986a,b). Trace elements are present in the soil in solid chemical forms and not readily soluble in the solution phase. As a result, plant uptake of trace elements tends to short range mass transfer processes taking place in the vicinity of rhizosphere. For this reason, organic acids in the root exudates play a significant role in dissolving the soilborne trace elements for plant absorption. Koo et al. (2006) and Chen et al. (2010) proposed that the plant availability of soilborne trace elements be expressed as Mt  C0 x (1  ekt)

(7-1)

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

217

TABLE 7-4. Relative Tolerance of Crops to Boron in Soil Solutions at Saturated Water Content Relative Boron Tolerance (1)

Very sensitive (0.05–0.75 mg/L)

Sensitive (0.75–1.0 mg/L)

Moderately sensitive (1.0–2.0 mg/L) Moderately tolerant (2.0–4.0 mg/L)

Tolerant (4.0–6.0 mg/L) Very tolerant (6.0–15.0 mg/L)

Species (2)

Avocado, grapefruit, orange, apricot, peach, cherry, plum, persimmon, fig (kadota), grape, walnut, pecan, cowpea, onion Garlic, sweet potato, wheat, sunflower, mung bean, sesame, lupine, strawberry, kidney bean, lima bean, peanut Broccoli, red pepper, pea, carrot, radish, potato, cucumber Lettuce, cabbage, celery, turnip, bluegrass, barley, oats, corn, tobacco, mustard, sweet clover, squash, muskmelon, cauliflower Tomato, alfalfa, purple vetch, parsley, beet (red), sugar beet Sorghum (6 mg/L), cotton (10 mg/L), asparagus

Data from Maas (1984).

where Mt (mg/kg)  cumulative trace element removal by plants grown and harvested in t successive years, C0  available trace element pool of the soil (mg/kg); and k  element- and soil-specific available trace element release constant of the soil (1/y). In this manner, the availability of soilborne trace elements may be defined in terms of total amounts potentially available, C0, and the time-dependent rate trace elements absorption, r. They demonstrated that C0 and k of Equation 7-1 could be obtained by a extracting the soils with synthetically formulated root exudates and fitting the results to an exponential decay kinetics model (Chen et al. 2010). Plant Methods The nutritional deficiencies and toxicities of plants can be diagnosed based on visual symptoms and plant tissue analysis. Leaves of plants

218

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

affected by nutrient deficiency or toxicity are frequently characterized by necrosis (tip burn or marginal scorch), chlorosis (turning yellow in color), and abscission (premature dropping). These symptoms may appear most frequently on leaves of a particular age class, depending on whether the toxic trace element is readily translocated (mobile) within the plant. The chlorosis can occur uniformly over an entire leaf surface, as often occurs with Fe and N deficiency, or be restricted to the interveinal regions of the leaf, such as occurs with Mn and Mg deficiency. Bould et al. (1984), Scaife and Turner (1984), and Winsor and Adams (1987) have described symptoms associated with deficiencies and toxicities of trace elements for a wide variety of crops. Unfortunately, visual symptoms caused by the deficiency or toxicity of one element often are similar to those of another element. Therefore, an elemental analysis of the leaf tissue is needed to confirm the cause. The analysis of plant tissue has long been used to diagnose deficiencies and toxicities of trace elements in plants. A number of factors determine the threshold concentrations that indicate deficiencies or toxicities. The trace elements in plants vary according to the species and cultivar, the part of the plant sampled, its stage of development, the position of the tissue on fruiting versus nonfruiting branches, and the substrate. Therefore, to use the trace elements in plant tissue to diagnose nutrient deficiencies and phytotoxicities, the condition under which the threshold levels were derived must be adhered to strictly. Data presented in Table 7-5 show levels of trace elements in leaves of plants that indicate deficient or toxic conditions. The concentrations presented in Table 7-5 are generalized from data in various publications. They should be regarded merely as a guide for screening toxic or deficient conditions. Table 7-6 summarizes concentration ranges of trace elements associated with both normal and phytotoxic levels in forage crops. Among the forage species, trace element concentrations vary by a factor of as much as 10, and concentrations associated with phytotoxicity vary by a factor of 2 or more.

ACCUMULATION OF TRACE ELEMENTS IN VEGETATION TO LEVELS POTENTIALLY HARMFUL TO ANIMALS Plants can grow at normal or near-normal rates but accumulate enough of certain trace elements to cause acute toxicity or chronic metabolic imbalances in consumers of the plants. The trace elements that can be absorbed by plants from soils in amounts high enough to be toxic, or cause metabolic imbalance in consumers, include Cd, Se, Mo, and possibly Be and Co. Arsenic, Cr, Fe, Pb, and Hg are toxic to animals and humans but are not absorbed by terres-

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

219

TABLE 7-5. Threshold Concentrations of Trace Elements in Leaves of Crops Indicative of Possible Deficiency or Phytotoxicity

a

Element (1)

Deficiencya (mg/kg on dry weight) (2)

Phytotoxicityb (mg/kg on dry weight) (3)

As B Cd Co Cu Fe Mn Mo Ni Se

NEc 20 NE 0.2 5 30 15 0.1 NE NE

— 100 10 — 25 — 100 100 50 50

Zn

15

200

Deficiency possible if concentration is less than listed

b

Phytotoxicity possible if concentration is greater than listed

c

NE, nonessential element

trial plants in sufficient concentrations to be harmful to consumers. There could be a soil–plant barrier to limit the translocation of trace elements from soils to plants (Benzarti et al. 2008; Chaney 1980, 1983; Hamon et al. 1999). The barrier protects the food chain from potentially toxic trace elements when one or more of the following occurs: (1) Low solubility in the soil prevents uptake. Bonding of potentially toxic elements to soils can limit transfer to roots. Some metals, such as Cr and Pb, have minimal solubility in soils, which creates a particularly strong barrier to plant uptake. (2) Insolubility (or sequestration) of an element in the fibrous roots prevents translocation to the edible plant tissues. Or (3) phototoxicity of the trace element kills the plant before the concentrations of the element in the edible plant tissues reach concentrations injurious to animals. Although the soil–plant barrier effectively protects consumers of the plants against harmful accumulations of trace elements via the root absorption process, other ways of ingesting contaminated soil, dust particles, or vegetation may cause harmful trace-element intake. For example, infants and toddlers in an inner-city area, where soils and dusts contain large amounts of Pb, may inadvertently ingest high amounts of Pb through the hand-to-mouth route.

220

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 7-6. Concentration of Trace Elements Commonly Observed in Forage Crops Element (1)

Typical (mg/kg on dry weight) (2)

As B Cd Co Cu Mo Ni Se

0.01–1.0 7–75 0.1–1.0 0.01–0.3 3–20 0.1–3.0 0.1–5.0 0.1–2.0

Zn

15–150

Phytotoxic (mg/kg on dry weight) (3)

3–10 75 5–700 25–100 25–40 100 50–100 100 500–1,000

Data from Logan and Chaney (1983).

Cadmium Ample evidence demonstrates that crops grown on Cd-contaminated soils contain elevated concentrations of Cd (Grant and Bailey 1998; Guttormsen et al. 1995; He and Singh 1994; Huang et al. 2004). Loganathan et al. (1995) concluded that if factors, such as topsoil pH and organic matter content, are equal, then the amount of Cd taken up by plants is proportional to the amount accumulated in the topsoil. If a human population regularly consumes Cd-enriched foods over decades, the Cd will accumulate in the people’s bodies and adversely affect the health of susceptible individuals. The body can mistake Cd for essential elements, such as Ca and Zn, so individuals with a deficient diet are at a greater risk for these effects. The most widely publicized example of chronic Cd poisoning occurred in Japan, where wastewater from a Zn mining and smelting complex contaminated the rice grown on fields downstream (Asami 1981; Kobayashi 1978). Most of the cases of Cd poisoning (so-called itaiitai disease) occurred in multiparous farm women of above middle age who also showed symptoms of Ca, Zn, and Fe deficiency. The concentration of Cd in the locally grown rice and the number of years of residence in the area also strongly related to incidence of the disease. No other documented cases exist of chronic Cd poisoning of humans or animals from foods grown on Cd-contaminated soils. The behavior of Cd in plants is closely related to Zn because both elements have affinity for sulphur, particularly sulphydryl groups (Garrett

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

221

1996). The symptoms of Cd toxicity in plants generally are stunning and chlorosic. The chlorosis may be due to the Fe deficiency or direct/indirect interaction of Cd with foliar Fe (Das et al. 1997). Cadmium contents of agricultural soils normally are less than 1 mg/kg. Some soils in irrigated regions may have naturally high levels of Cd, but chronic Cd poisoning is unlikely unless the food grown on these soils makes up a substantial part of a consumer’s regular diet over an extended period of time. The World Health Organization estimates that the adult human intake of Cd from food on a global basis is 70 g/ day, which is already approaching the recommended tolerable intake of Cd (WHO 1992). Molybdenum Molybdenum, essential in the diet of livestock at low concentrations, is toxic at higher concentrations. Ruminant animals are apparently more sensitive to such toxicity, so-called molybdenosis. The amount of Mo that an animal can tolerate depends on the levels of Cu and SO4 in the diet (Spears 2003; Suttle 1991). A negative relationship between concentration of Mo and Cu in the liver from Swedish moose was presented by Frank et al. (2002). If the level of Cu is low, as low as 5 mg/kg of Mo in the forage may adversely affect the health of ruminant animals. Because plants can tolerate considerably larger concentrations of Mo than 5 mg/kg in their plant tissue, forage crops grown on soils high in Mo may become unfit for livestock feed without showing any symptoms of injury. The Mo toxicity to livestock is primarily a nutritional deficiency of Cu, and the ambient sulfate level markedly affects the interaction between Cu and Mo (Underwood, 1977). The dietary ratio of Cu/Mo for ruminants should range from 4 to 5. Any variation in the ratio between Cu/Mo/S in ruminant feed can introduce a secondary deficiency of Cu, and sometimes impaired metabolism of Fe. The Moinduced deficiency has been relatively often observed (MacPherson, 2000; Jones 2005). Soils in arid and semiarid regions, especially ferrasols, usually contain relatively high amounts of Mo. In soils, Mo availability to crops increases as the pH of the soil increases, and Cu availability tends to decrease as pH of the soil increases. Consequently, toxic levels of Mo are more common in forages grown on alkaline soils than in those grown on acid soils. Deficiencies of Mo are likely to occur on acid soils (pH 5.5) and soils with a low Mo content and high Fe oxide levels. The most sensitive to Mo deficiency are some vegetables, such as cauliflower, that exhibit deficiency symptoms at a Mo content of 0.4 mg/kg in leaves. Liming of acid soils is a common practice to increase the Mo availability to plants. Fertilization with Mo was employed in particular environments and on

222

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

particular crops through coating the seed with a dilute solution of Mo salt or foliar spraying. Selenium Selenium is essential to livestock at low concentrations and toxic at slightly higher concentrations (James et al. 1989). Soils from many regions in the world produce forage with amounts of Se that are nutritionally inadequate for animals. Irrigation methods affect the selenium accumulation in forages (Suarez et al. 2003). Selenium-response disease in sheep is observed in areas where soils contain 0.5 mg/kg (Oldfield 2002). Forage Se at a concentration above 0.1 mg/kg satisfies the requirement of most animals (NRC 2001; ATSRD 2002; Fordyce 2005). For beef cattle, the nutritional requirement of Se is 0.1 mg/kg (NRC 1996). For ruminant animals, the requirement is 0.03 to 0.05 mg/kg (McDowell and Valle 2000; Minson 1990). The supplementation of feed by selenomethionine or Se yeast has become increasingly common in animal nutrition in low-Se areas worldwide. Observations have indicated that the supplementation of feed seems to have some prophylactic effect on humans who ingested animal products (Hartikainen 2005). In a few isolated areas, however, concentrations of Se are high enough to produce forage with amounts of Se that would be toxic to animals. The toxic threshold values for Se in fodder has been set at above 3 mg/kg (ATSDR 2002; Fordyce 2005), and some have recommended 5 mg/kg in the diet as the dividing level between toxic and nontoxic feed (Schrauzer 2004b). The maximum tolerable concentration set by NRC (1996) for beef cattle is 2 mg/kg. Se toxicity (i.e., alkali disease) occurs in dairy cattle when concentrations in forage exceed 5 mg/kg (DW) (NRC 2001). Selenium toxicity in humans due to environmental exposure has also been reported (Yang et al. 1983). In the Westside San Joaquin Valley of California, high concentrations of Se in drainage from irrigated fields were linked to deaths and deformities of waterfowl and the elimination of fish from reservoirs receiving the drainage water in the 1980s (Presser and Ohlendorf 1988). The problem was caused by the irrigation of soils derived from Se-laden sediments of marine origin. As the irrigation water leached through the soils, Se was brought into solution and transported through the drainage network to a terminal reservoir (Kesterson Reservoir), where it was further concentrated to ecotoxic concentration. Currently, researchers are examining the potential to use less saline drainage water for the production of forages and canola for biodiesel production (Baˇnuelos 1996; Baˇnuelos et al. 2003). Forages grown on these high-Se soils and irrigated with highSe drainage water for multiple years accumulated up to 10 mg/kg Se (DM) (Suyama et al. 2007), which would require feeding to animals in a mixed diet. Conversely, such high Se levels in these forages could make

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

223

them valuable as processed, organic Se supplements for cattle raised in Se-deficient areas. Cobalt Cattle have been injured when fed diets spiked with levels of soluble salts of Co that exceed 10 mg/kg. Concentrations of Co in forages are normally from 0.01 mg/kg to 0.3 mg/kg. Some crops may grow well at as much as 50 to 100 mg/kg (Logan and Chaney 1983). Although Co is apparently an element that can penetrate the soil–plant barrier, no instances of animals being injured from excessive plant-absorbed Co in their diets have been reported (Logan and Chaney 1983). Cobalt is essential for humans and for most animals as a component of the vitamin B12. The deficiency of Co in humans may affect anemia and anorexia. Co deficiency syndromes in ruminants occur in several continents, and the most severe were observed at the beginning of last century in Australia. According to Schrauzer (2004a), it is likely that Co deficiency will become a problem in the future, since the natural Co content of soils is low and the depletion of this metal occurs through agriculture practices and natural soil leaching processes. It has been observed that grasses grown on soil with a Co content less than 5 mg/kg may be Co-deficient for the normal growth of animals. The deficiency of Co can be controlled by application of Co salts to the soil, or given directly to livestock.

ASSESSING THE FATES OF TRACE ELEMENTS IN SOILS The fate and transport of trace elements in soils are determined by the dynamic equilibrium between those present in various soil components, which, in turn, are governed by interactions including adsorption and desorption, precipitation and dissolution, organic matter formation and decomposition, and oxidation and reduction. Depending on their chemical nature and amounts, portions of trace elements in soil may leach into deeper soil strata and/or be carried away by surface runoff, become airborne, or be absorbed by growing plants. It is difficult to detect the changes of trace elements in soils by the routine field sampling and measurement protocols, since the in-and-out fluxes are generally small. More often, their environmental fates can be evaluated by mass balance models. Two generalized mass balance models (STEM-single layer model, and STEM-profile distribution model) were developed at the University of California, Riverside, to assess the risks associated with trace elements fates in soils. These two models are available for download at Chang (2011).

224

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

In both models, the trace elements present in the rootzone are assumed to be distributed in four conceptually defined phases, namely, the soil solution phase, adsorbed phase, mineral phase, and organic phase. Trace elements in these phases are in kinetic equilibrium and are subjected to three fluxes, including inputs from external sources, leaching out of the rootzone, and plant uptake removal. Using the mass balance model, Chen et al. (2007, 2009) evaluated the fates of two typical elements, As and Cd, under normal California cropping practices. Model simulation results show intensive cultivation does not have a significant long-term effect on the total As content of the cropland soils but will have a significant impact on total Cd content of the receiving soils. Over the 100-year period, the total soil As content of the rootzone shows a slight decreasing trend at an annual input of 31.5 g/ha/yr, while the soil Cd level increases two-fold from 0.22 to 0.54 mg/kg at an annual input of 21.5 g/ha/yr. The fate of As and Cd is greatly affected by soil and plant characteristics, as well as irrigation practice. The solid and solution phase partitioning coefficient (Kd) has direct impacts on the amount of As and Cd accumulation in soils, especially when Kd is less than 1,000 (Fig. 7-1). As Kd increases (i.e., when soils retain incrementally more trace elements), the amount held by the “sinks” (i.e., leaching and plant uptake) diminishes, and the amount accumulated in the solid phase of the soil proportionally increases. When Kd exceeds 1,000, the total As and Cd contents of the receiving soils rapidly approach a plateau, indicating a steady state in the mass balance. Removal by plant uptake (represented by maximum influx rate Jmax) and leaching losses (represented by percolation rate Kh) are interrelated (Fig. 7-2). For heavier-textured soils, the leaching losses decrease, resulting in a higher trace element concentration in the soil solution and more plant uptake. Kh has a minor effect on the total As and Cd contents of the receiving soils when Jmax is high, indicating plant species with high As and

FIGURE 7-1. Effect of partitioning coefficient on accumulation of As and Cd in cropland soils. Data from Chen et al. (2007).

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

225

FIGURE 7-2. Effect of percolation rate, Kh and maximum influx rate, Jmax on accumulation of As and Cd in the receiving soils. Data from Chen et al. (2007).

Cd absorption rates. Leaching loss has a pronounced effect on the accumulation of As and Cd in the receiving soils when Jmax is low, indicating plant species with low As and Cd absorption rates. In the food chain, the transfer of potentially toxic elements is customarily evaluated by a steady-state plant uptake factor (PUF). Plant uptake is a dynamic process that is interrelated with other soil attributes. Judging from the outcomes of the sensibility analyses, PUF may not provide an accurate depiction of the plant uptake of trace elements as the soil texture (Kh), soil chemical properties (Kd), and plant species (Jmax) invariably change from one location to another (Chen et al. 2009). Water input is the driving force of pollutant movement in the soil profile. As the amount of irrigation increases, the leaching loss from the soil profile will increase. Consequently, the amount of plant uptake is reduced because high irrigation results in a lower trace element concentration in the soil solution. The amount of trace element accumulated in the soil may decrease with time (Fig. 7-3). With a given amount of irrigation, the distribution of As and Cd in the soil profile may also be affected by the duration and frequency of irrigation. As the duration and frequency decrease, irrigation intensity will increase, and water flux and leaching loss will also increase. For those elements with high concentrations in the soil solution, the irrigation practice may have a significant effect on their accumulation in the soil and thus their uptake by plants. Hydraulic properties of soils govern water transport in soils and hence the distribution of trace elements in the soil profile. Figure 7-4 illustrates the accumulation of As and Cd over 100 years for three soil textures under typical agricultural practices in California. When the soil texture changes

226

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 7-3. Accumulation of As and Cd in the plow layer with different irrigation practices under typical California agricultural conditions.

from fine to coarse, leaching increases and plant uptake decreases. The effect of soil hydraulic properties on plant uptake is less significant than on leaching. Overall, total soil As and Cd concentrations decrease as soil texture varies from clay soil to sandy soil. Monte Carlo simulation was included in the mass balance model to account for the variance and probability distribution of each model parameter in the simulations. Figure 7-5 summarizes the probability distribution of total As and Cd in soil from 10,000 simulations, in which the model parameters were randomly selected according to their probability distribution function. Based on the outcomes, there is a 60% probability that the As content of soils receiving P fertilizer applications for 100 years will remain equal to or lower than the initial As concentration of the soil. The accumulation of As in soil is not significant, even in extreme cases. Conversely, for Cd, there is only a 10% probability that after 100 years of

FIGURE 7-4. Accumulation of As and Cd in the plow layer at soil texture under typical California agricultural conditions.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

227

FIGURE 7-5. Cumulative probability distribution of total As and Cd content at the end of 100 years of continuous cultivation. Data from Chen et al. (2007).

P fertilizer application, Cd concentrations in soil will remain equal to or lower than the initial concentrations. In the extreme case, soil Cd increased more than four-fold after 100 years of cultivation. The estimated total As and Cd contents of the soils followed log-normal distributions and were skewed toward the high concentration end.

SUMMARY Agronomists have long labored to find ways to diagnose trace element deficiencies and toxicities in plants. Many have focused on determining the thresholds of an element in the growing medium that result in nutritional deficiencies or phytotoxicities. Two approaches are commonly used: (1) relate the concentration of the element in soil to the yield (and therefore the degree of health of the plant); and (2) examine the plant’s foliage for symptoms of nutritional deficiency or phytotoxicity. The amount of a trace element in soil is measured by extracting the element with an appropriate chemical reagent and determining the concentration of the element in the extract, which should, in theory, reflect the plant-available concentration. Since the physical and chemical properties of the soil influence the results, deficiency and phytotoxicity thresholds of an element may vary widely in different soils. A plant’s performance can also be indicated by assaying the amount of an element that has accumulated in its tissue. The elemental content of the tissue usually depends solely on the availability of the element in the soil. To determine the elemental concentrations in plant tissue that result in nutritional deficiency or phytotoxicity, tissue concentrations are usually related to yield or some indicator of growth. Criteria based on the elemental concentration in plant tissue suffer from physiological differences

228

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of different species and cultivars, as well as uncertainties introduced by the climate and cultivation practices. The techniques for diagnosing trace element deficiencies and phytotoxicities in plants grown in saline soils are the same as for plants grown in nonsaline soils. However, the high electrolyte concentration in saline soil solutions, the neutral to alkaline soil chemical conditions, the soil’s susceptibility to desiccation, and the salinity stress experienced by the plants will all influence the diagnostic criteria. In our opinion, 1. Boron is the trace element most likely to be phytotoxic in saline soils. 2. Plants grown on saline soils often experience deficiencies of Cu, Fe, Mn, and Zn because these micronutrients have low solubility in saline soils, which tend to be alkaline. 3. Plants can tolerate high levels of Se and Mo in soils, but some forage crops grown in saline soils can absorb enough Se and Mo to be toxic to consumers. 4. Levels of Cd in soils are generally low and are unlikely to cause problems in crop production, but Cd introduced into the soil by humans may accumulate to toxic amounts in food crops.

REFERENCES Agency for Toxic Substances and Disease Registry (ATSDR). (2002). Draft toxicological profile for selenium, U.S. Department of Health and Human Services, Agency for Toxic Substances and Disease Registry, Atlanta, Ga. Allen, E. R., Johnson, G. V., and Unruh, L. G. (1994). “Current approaches to soil testing methods: Problems and solutions,” in Soil testing: Prospects for improving nutrient recommendations, J. L. Havlin and J. S. Jacobsen, eds., Special Publication No. 40, Soil Science Society of America, Madison, Wisc., 203–220. Alloway, B. J. (2005). “Bioavailability of elements in soils,” In Essentials of medical geology: Impacts of natural environment on public health, O. Selenus et al., eds., Elsevier, London, 347–372. Alvarez, J. M. (2007). “Influence of soil type on the mobility and bioavailability of chelated zinc.” J. Agric. Food Chem., 55, 3568–3576. Asami, T. (1981). “Maximum allowable limits of heavy metals in rice and soil,” in Heavy metal pollution in soils of Japan, K. Kitagishi and I. Yamane, eds. Japan Scientific Societies Press, Tokyo. Atlas. (1989). The atlas of endemic diseases and their environments in People’s Republic of China, Science Press, Beijing. Baˇnuelos, G. S. (1996). “Managing high levels of boron and selenium with trace element accumulator crops.” J. Environ. Sci. Health, A31(5), 1179–1196. Baˇnuelos, G. S., and Lin, Z. (2008). Development and uses of biofortified agricultural products. CRC Press, Boca Raton, Fla.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

229

Baˇnuelos, G. S., Sharmarsakar, S., Cone, D., and Stuhr, G. (2003). “Vegetative approach for improving the quality of water produced from soils in the westside of central California.” Plant Soil, 249, 229–236. Basar, H. (2005). “Methods for estimating soil iron availability to chlorotic peach trees.” Commun. Soil Sci. Plant Anal., 36, 1187–1198. Benzarti, S., Mohri, S., and Ono, Y. (2008). “Plant response to heavy metal toxicity: Comparative study between the hyperaccumulator Thlaspi caerulescens (ecotype Ganges) and nonaccumulator plants: Lettuce, radish, and alfalfa.” Environ. Toxicol., 23, 607–616. Berger, K. C., and Truog, E. (1939). “Boron determination in soils and plants.” Ind. Eng. Chem. Anal. Ed., 11, 540–545. Bould, C., Hewitt, E. J., and Needham, P. (1984). Diagnosis of mineral disorders in plants. Vol. 1: Principles. Chemical Publishing, New York. Brennan, R. F., Armour, J. D., and Reuter, D. J. (1993). “Diagnosis of zinc deficiency,” in Zinc in soils and plants: Developments in plant and soil sciences 55, A. D. Robson, ed., Kluwer Academic Publishers, Dordrecht, The Netherlands, 167–181. Chaney, R. L. (1980). “Health risks associated with toxic metals in municipal sludge,” in Sludge: Health risks of land application, G. Bitton, B. L. Damron, G. T. Edds, and J. M. Davison, eds., Ann Arbor Science Publishers, Inc., Ann Arbor, Mich., 59–83. ———. (1983). “Fate of toxic substances in sludge applied to cropland,” in Proc., International Symposium on Land Application of Sewage Sludge. Oct. 13–15, 1982. Tokyo, 259–324. Chang, A. (2011). “UC-Riverside/faculty/Chang,” www.envisci.ucr.edu/faculty/ chang.html, accessed January 20, 2011. Chang, A. C., Page, A. L., Foster, K. W., and Jones, T. E. (1982). “A comparison of cadmium and zinc accumulation by four cultivars of barley grown in sludgeamended soils.” J. Environ. Qual., 11, 409–412. Chen, W. P., Chang, A. C., and Wu, L. (2007). “Assessing long-term environmental risks of trace elements in phosphate fertilizers.” Ecotoxicol. Environ. Safety, 67, 48–58. Chen, W. P., Chang, A. C., Wu, L., Page, A. L., and Koo, B. (2010). “Trace elements in biosolids-amended soils,” in Trace elements in soils, P. S. Hooda, ed., John Wiley and Sons, London, 113–133. Chen, W. P., Li, L. Q., Chang, A. C., Wu, L. S., Khosrivafard, M., and Chaney, R. L. (2008a). “Characterizing the solid-solution partitioning coefficient and plant uptake factor of As, Cd, and Pb in California croplands.” Agri., Ecosys. Envir., 129(1/3), 212–220. Chen, W. P., Wu, L. S., Sˇ imu ˚ nek, J., Chang, A. C., and Hou, Z. A. (2008b). “Assessing the effect of long-term crop cultivation on distribution of Cd in the root zone.” Ecol. Modeling, 220(15), 1836–1843. Das, P., Samantaray, S., and Rout, R. (1997). “Studies on cadmium toxicity in plants: A review.” Environ. Pollut., 98, 29–36. Fordyce, F. (2005). “Selenium deficiency and toxicity in the environment,” in Essentials of medical geology: Impacts of the natural environment on public health, O. Selinus, B. J. Alloway, J. A. Centeno, R. B. Finkelman, R. Fuge, U. Lindh, and P. Smedley, eds., Elsevier, Amsterdam, 373–415.

230

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Frank, A., Wibomb, R., and Danielsson, R. (2002). “Myocardial cytochrome c oxidase activity in Swedish moose (Alces alces L.) affected by molybdenosis.” Sci. Total Environ., 290, 121–129. Gao, Y., He, J., Ling, W., Hu, H., and Liu, F. (2003). “Effects of organic acids on copper and cadmium desorption from contaminated soils.” Environ. Int., 29, 613–618. Garrett, R. G. (1996). Natural sources of cadmium: Sources of cadmium in the environment, Organisation for Economic Co-operation and Development, Paris, 149–161. Grant, C. A., and Bailey, L. D. (1998). “Nitrogen, phosphorus and zinc fertilizer management effects on grain and cadmium concentration in two cultivars of durum wheat,” in Can. J. Plant Sci., 78, 63–70. Gupta, U. C. (1993). “Deficiency, sufficiency, and toxicity levels of boron in crops,” in Boron and its role in crop production, U. C. Gupta, ed., CRC Press, Boca Raton, Fla., 137–145. Guttormsen, G., Singh, B. R., and Jeng, A. S. (1995). “Cadmium concentration in vegetable crops grown in a sandy soil as affected by Cd levels in fertilizer and soil pH.” Fert. Res., 41, 27–32. Hamon, R. E., Holm, P. E., Lorenz, S. E., McGrath, S. P., and Christensen, T. H. (1999). “Metal uptake by plants from sludge-amended soils: Caution is required in the plateau interpretation.” Plant Soil, 216, 53–64. Hartikainen, H. (2005). “Biogeochemistry of selenium and its impact on food chain quality and human health.” J. Trace Elem. Med. Biol., 19, 125–140. He, Q. B., and Singh, B. R. (1994). “Crop uptake of cadmium from phosphorus fertilizers. II. Relationship with extractable soil cadmium.” Water, Air, Soil Poll., 74, 251–265. Holmgren, G. S., Meyer, M. W., and Chaney, R. L. (1993). “Cadmium, Pb, Zn Cu and Ni in agricultural soils of the United States of America.” J. Environ. Qual., 22, 335–348. Huang, B., Kuo, S., and Bembenek, R. (2004). “Availability of cadmium in some phosphorus fertilizers to field-grown lettuce.” Water, Air, Soil Poll., 158, 37–51. Inskeep, W. P., and Comfort, S. D. (1986). “Thermodynamic predictions for the effects of root exudates on metal speciation in the rhizosphere.” J. Plant Nutr., 9, 567–586. Jaffiol, H. N., Perezi, N., Baylet, R., and Baldet, L. (1992). “A survey on endemic goiter in Togo: Etiological factors analysis.” Bull. Acad. Natl. Med., 176, 557–567. James, L. F., Panter, K. E., Mayland, H. F., Miller, M. R., and Baker, D. C. (1989). Selenium poisoning in livestock: A review and progress, Soil Science Society of America Special Publication No. 23, American Society of Agronomy, Milwaukee, Wisc. John, M. K., VanLaerhoven, C. J., and Chauh, H. H. (1972). “Factors affecting plant uptake and phytotoxicity of cadmium added to soils.” Environ. Sci. Tech., 6, 1005–1009. Jones, B. (2005). “Animals and medical geology,” in Essentials of medical geology: Impacts of the natural environment on public health, O. Selinus, B. J. Alloway, J. A. Centeno, R. B. Finkelman, R. Fuge, U. Lindh, and P. Smedley, eds., Elsevier, Amsterdam, 513–526. Kabata-Pendias, A., and Mukherjee, A. B. (2007). Trace elements from soil to human, Springer-Verlag, Berlin.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

231

Kabata-Pendias, A., and Pendias, H. (2001). Trace elements in soils and plants, 3rd ed., CRC Press, Boca Raton, Fla. Khan, M. A. R., Bolan, N. S., and Mackay, A. D. (2005). “Soil test to predict the copper availability in pasture soils.” Comm. Soil Sci. Plant Anal., 36, 2601–2624. Kobayashi, J. (1978). “Pollution by cadmium and the itai-itai disease in Japan,” in Toxicity of heavy metals in the environment, F. W. Oehme, ed., Marcel Dekker, Inc., New York, 199–260. Koo, B. J., Chang, A. C., Crowley, D. E., and Page, A. L. (2006). “Characterization of organic acids recovered from rhizosphere of corn grown on biosolidstreated medium.” Comm. Soil Sci. Plant Anal., 37(5), 871–887. Lindsay, W. L., and Norvell, W. A. (1969). “Development of a DTPA micronutrient soil test.” Agron. Abstr., 84. Logan, T. J., and Chaney, R. L. (1983). “Utilization of municipal wastewater and sludge on land-metals.” Proc. of the Workshop on Utilization of Municipal Wastewater and Sludge on Land, A. L. Page et al., eds., University of California, Riverside, Calif., 235–326. Loganathan, P., Mackay, A. D., Lee, J., and Hedley, M. J. (1995). “Cadmium distribution in hill pastures as influenced by 20 years of phosphate fertilizer application and sheep grazing.” Aust. J. Soil Res., 33, 859–871. Maas, E. V. (1984). “Salt tolerance of plants,” in Handbook of plant science, B. R. Christie, ed., CRC Press, Cleveland, Ohio. Maas, E. V., and Grattan, S. R. (1999). “Crop yields as affected by salinity.” Agronomy, 38, 55–110. MacPherson, A. (2000). “Trace mineral status of forages,” in Forage evaluation in ruminant nutrition, D. I. Givens, E. Owens, R. F. E. Axford, and H. M. Omed, eds., CABI Publishing, Oxon, UK, 345–371. McDowell, L. R., and Valle, G. (2000). “Major minerals in forages,” in Forage evaluation in ruminant nutrition, D. I. Givens, E. Owens, R. F. E. Axford, and H. M. Omed, eds., CABI Publishing, Oxon, UK, 373–397. Mench, M., and Martin, E. (1991). “Mobilization of cadmium and other metals from two soils by root exudates of Zea mays L., Nicotiana tabacum L. and Nicotiana rustica L.” Plant Soil, 132, 187–196. Mench, M., Morel, J. L., Guckert, A., and Guillet, B. (1988). “Metal binding with root exudates of low molecular weight.” J. Soil Sci., 39, 521–527. Merckx, R., Ginkel, J. H., Sinnaeve, J., and Cremers, A. (1986). “Plant-induced changes in the rhizosphere of maize and wheat. II. Complexation of cobalt, zinc and manganese in the rhizosphere of maize and wheat.” Plant Soil, 96, 95–107. Minson, D. J. (1990). Forage in ruminant nutrition, Academic Press, New York. Morrissey, J., and Guerinot, M. L. (2009). “Trace elements: Too little or too much and how plants cope.” Biol. Rep., 1(14), DOI: 10.3410/B1-14. Nable, R. O. (1990). Forage in ruminant nutrition. Academic Press, New York. National Research Council (NRC). (1996). Nutrient requirements of beef cattle, 7th ed., National Academies Press, Washington, D.C. ———. (2001). Nutrient requirements of dairy cattle, 7th ed., National Academies Press, Washington, D.C. Oldfield, J. E. (2002). Selenium world atlas, updated ed., Selenium-Tellurium Development Association, Grimbergen, Belgium.

232

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Oliver, M. A. (1997). “Soil and human health: A review.” Eur. J. Soil Science, 48, 537–592. Olson, S., and Sommers, L. E. (1982). “Phosphorus,” in Methods of soil analysis. Part 2. Chemical and microbiological properties, 2nd ed., A. L. Page, R. H. Miller, and D. R Keeney, eds., American Society of Agronomy, Madison, Wisc., 403–427. Page, A. L., Chang, A. C., Sposito, G., and Mattigod, S. (1981). “Trace elements in wastewater: Their effects on plant growth and composition and their behavior in soils,” in Modeling wastewater renovation: Land treatment, I. K. Iskandar, ed., John Wiley and Sons, New York, 182–222. Pierzyneski, G. M., and Jacobs, L. W. (1986a). “Extractability and plant availability of Mo from inorganic and sewage sludge sources.” J. Environ. Qual., 15, 323–326. ———. (1986b). “Molybdenum accumulation by corn and soybeans from a Morich sewage sludge.” J. Environ. Qual., 15, 94–98. Purves, D. (1985). Trace-element contamination of the environment, Elsevier, New York. Roholm, K. (1937). Fluorine intoxication, H. K. Lewis and Co., London. Scaife, A., and Turner, M. (1984). Diagnosis of mineral disorders in plants. Vol. 2: Vegetables, Chemical Publishing Co., New York. Schrauzer, G. N. (2004a). “Cobalt,” in Elements and their compounds in the environment, 2nd ed., E. Merian, M. Anke, M. Ihnat, and M. Stoeppler, eds., WileyVCH Verlag, Weinheim, Germany, 1365–1406. ———. (2004b). “Selenium,” in Elements and their compounds in the environment, 2nd ed., E. Merian, M. Anke, M. Ihnat, and M. Stoeppler, eds., Wiley-VCH Verlag, Weinheim, Germany, 1365–1406. Shanker, A. K. (2008). “Mode of action and toxicity of trace elements,” in Trace element nutritional benefits, environmental contamination and health, M. N. V. Prasad, ed., John Wiley and Sons, London, 525–557. Soltanpour, P. N. (1985). “Use of ammonium bicarbonate DTPA soil test to evaluate elemental availability and toxicity.” J. Plant Nutr., 8, 323–338. Soltanpour, P. N., and Schwab, A. P. (1977). “A new soil test for simultaneous extraction of macro- and micro-nutrients in alkaline soils.” Comm. Soil Sci. Plant Anal. 8(3), 195–207. Spears, J. W. (2003). “Trace mineral bioavailability in ruminants.” J. Nutr. 133, 1506–1509. Sposito, G. (1986). “Distinguishing adsorption from surface precipitation,” in Geochemical processes at mineral surfaces, J. A. Davis and K. F. Harges, eds., ACS Symposium Series No. 323, American Chemical Society, Washington, D.C., 217–228. Suarez, D. L., Grieve, C. M., and Pos, J. A. (2003). “Irrigation method affects selenium accumulation in forage Brassica species.” J. Plant Nutr., 26, 191–201. Suttle, N. F. (1991). “The interactions between copper, molybdenum, and sulphur in ruminant nutrition.” Ann. Rev. Nutr., 11, 121–140. Suyama, H., Benes, S. E., Robinson, P. H., Getachew, G., Grattan, S. R., and Grieve, C. M. (2007). “Biomass yield and nutritional quality of forage species under long-term irrigation with saline-sodic drainage water: Field evaluation.” Anim. Feed Sci. Technol. 135, 329–345.

DEFICIENCIES AND TOXICITIES OF TRACE ELEMENTS

233

Symeonides, C., and McRae, S. G. (1977). “The assessment of plant-available Cd in soils.” J. Environ. Qual., 6, 120–123. Tsadilas, C. D., Yassoglou, C. S., Cosmas, C. S., and Kallianou, C. H. (1994). “The availability of soil boron fractions to olive trees and barley and their relationships to soil properties.” Plant Soil, 162, 211–217. Underwood, E. J. (1977). Trace elements in human and animal nutrition, 4th ed. Academic Press, San Francisco. Winsor, G., and Adams, P. (1987). Diagnosis of mineral disorders in plants, Vol. 3: Glasshouse crops, Chemical Publishing Co., New York. Wolnick, K. A., Fricke, F. L., Capar, S. G., Braude, G. C., Meyer, M. W., Satzger, R. D., and Bonnin, E. (1983a). “Elements in major raw agricultural crops in the United States. 1. Cadmium and lead in lettuce, peanuts, potatoes, soybeans, sweet corn, and wheat.” J. Agric. Food Chem., 31, 1240–1244. ———. (1983b). “Elements in major raw agricultural crops in the United States. 2. Other elements in lettuce, peanuts, potatoes, soybeans, sweet corn, and wheat.” J. Agric. Food Chem., 31, 1244–1249. ———. (1985). “Elements in major raw agricultural crops in the United States. 3. Cadmium, lead, and eleven other elements in carrots, field corn, onions, rice, spinach, and tomatoes.” J. Agric. Food Chem., 33, 807–811. Woolson, E. A., Axley, J. H., and Kearney, P. C. (1971). “Correlation between available soil arsenic, estimated by six methods, and response of corn (Zea mays L.).” Soil Sci. Soc. Am. Proc., 35, 101–105. World Health Organization (WHO). (1992). Environmental health criteria, Vol. 134: Cadmium, WHO, Geneva. Xu, J. M., Wang, K., Bell, R. W., Yang, Y. A., and Huang, L. B. (2001). “Soil boron fractions and their relationship to soil properties.” Soil Sci. Soc. Am. J., 65, 133–138. Yang, G. Q., Wang, S., Zhou, R., and Sun, R. (1983). “Endemic selenium intoxication of humans in China.” Amer. J. Clin. Nutr., 37, 872–875.

This page intentionally left blank

CHAPTER 8 TRANSGENIC STRATEGIES TOWARD THE DEVELOPMENT OF SALT-TOLERANT PLANTS Anil Grover, Amanjot Singh, and Eduardo Blumwald

INTRODUCTION As discussed in Chapters 6, 19, and 20, one of the farm management responses to soil salinity is planting salt-tolerant crops, but salt-tolerant crops represent only a small fraction of food crops. Enhancing the salt tolerance of existing crops using transgenic methods may assist growers in maintaining productivity on saline soils. The production of transgenic plants constitutes one of the major developments in plant sciences, but it is a major challenge to genetically engineer crop plants showing improved performance against abiotic stresses, such as suboptimal and supra-optimal temperatures, excess salt, and reduced water availability, among others. Abiotic stresses adversely affect almost all major field-grown plants belonging to varied ecosystems. The severity of abiotic stresses is on the rise due to the intensive cultivation being practiced in the farming areas, as well as environmental deterioration. These stresses cause a great amount of loss, both in terms of biomass as well as economic returns; the extent of this loss depends on the crop species, its location, growth stage, and the intensity of the stress. A considerable proportion of the potential biomass of the crops remains untapped due to such stresses. Although modifying crops to enhance their tolerance of the abiotic stresses of saline soil environments is not a trivial endeavor, there has been substantial progress in this field. In the late 1970s and early 1980s, the transgenic plants produced were mostly those showing resistance to antibiotics (e.g., kanamycin) or expressing reporter genes (e.g., those encoding for chloramphenicol acetyl transferase, -glucuronidase, and green fluorescent protein). In the mid to late 1980s, transgenic tobacco plants showing (1) enhanced resistance to herbicide glyphosate (through 235

236

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

overexpression of the altered enolpyruvyl shikimate 3-phosphate synthase gene), (2) enhanced resistance to tobacco mosaic virus (by overexpressing the coat protein gene of tobacco mosaic virus), (3) enhanced resistance to insect attacks (by overexpression of Bt gene, as well as through gene encoding for protease inhibitor protein), and (4) enhanced resistance to fungal pathogens (through overexpression of chitinase enzyme) were produced (Abel et al. 1986; Broglie et al. 1991; Comai et al. 1985; Hilder et al. 1987; Vaeck et al. 1987). These early attempts paved the way for more elaborate attempts that resulted in the production of transgenic plants showing improved yields, better quality, altered levels of hormones, and/or increased levels of secondary metabolites. Examples of progress in increasing plant tolerance of abiotic stresses include development of (1) transgenic tobacco plants for enhanced cold stress tolerance (Murata et al. 1992), and (2) transgenic tobacco plants for enhanced salt stress tolerance (Tarczynski et al. 1993). Following these two initial reports, transgenic plants showing tolerance to salt stress, water stress, oxidative stress, low-temperature stress, and high-temperature stress have been produced. In this chapter we discuss the deleterious effects of excess salts on crop plants and review the work aimed at the production of salinity-tolerant transgenic plants. Agricultural Productivity and Salt Stress Environmental stress due to salinity is one of the most serious factors limiting the productivity of agricultural crops, which are predominantly sensitive to the presence of high concentrations of salts in the soil. The loss of farmable land due to salinization is directly in conflict with the needs of the world population, which is projected to increase by 1.5 billion (more than 20%) in the next 20 years, and the challenge of maintaining the world food supplies. Although famine in the world currently results from complex problems beyond an insufficient production of food, there is no doubt that the gains in food production provided by the so-called Green Revolution are reaching their ceiling, while the world population continues to rise. Therefore, increasing the yield of crop plants in salinized lands (as well as in other less productive lands) is an absolute requirement for feeding the world. The need to produce stress-tolerant crops was evident even in ancient times (Jacobsen and Adams 1958). However, efforts to improve crop performance under environmental stresses have not had much success because the fundamental mechanisms of stress tolerance in plants remain to be completely understood. Epstein et al. (1980) described technical and biological constraints to the problem of salinity. While there

TRANSGENIC STRATEGIES

237

appears to be more success with the technical solutions to the problem, the biological solutions have been more difficult to develop. A biological approach to the development of salt-tolerant crop varieties requires as a prerequisite the identification of key genetic determinants of stress tolerance and the use of salt stress tolerance-related genes (QTL; quantitative trait loci). The existence of salt-tolerant plants (halophytes) and differences in salt tolerance between genotypes within salt-sensitive plant species (glycophytes) clearly indicates that there is a genetic basis to salt response. While differences in salt tolerance between varieties have been known for a long time (Epstein 1977; 1983) and intraspecific selection for salt tolerance has been reported in rice (Akbar and Yabuno 1977) and barley (Epstein et al. 1980), a large gap in our understanding still exists. Flowers and Yeo (1995) highlighted the paucity of salt-tolerant cultivars and predicted that the number was likely to be fewer than 30. Flowers (2004) pointed out that since 1993, there have been just three registrations of salt-resistant cultivars in Crop Science (Al-Doss and Smith 1998; Dierig et al. 2001; Owen et al. 1994). Two basic genetic approaches currently being utilized to improve stress tolerance are (1) exploitation of natural genetic variations, either through direct selection in stressful environments or through the mapping of QTLs and subsequent marker-assisted selection, and (2) generation of transgenic plants to introduce novel genes or alter expression levels of the existing genes to affect the degree of salt-stress tolerance. We discuss these approaches in some detail, focusing on the recent experimentation with transgenic plants that has led to increased salinity tolerance, with emphasis on the areas of ion homeostasis, osmotic regulation, antioxidant protection and signaling, and transcriptional controls (see also Hasegawa et al. 2000; Sahi et al. 2006; Zhang et al. 2004; Zhu 2001, 2002). Salt Tolerance Using Transgenic Approaches Physiologically, salinity (1) imposes a water deficit that results from the relatively high solute concentrations in the soil, (2) causes ion-specific stresses resulting from altered K/Na ratios, and (3) leads to buildup in Na and Cl concentrations that are detrimental to plants. Plants respond to salinity using different types of mechanisms. Salt-sensitive plants restrict the uptake of salt and adjust their osmotic pressure by the synthesis of compatible solutes like proline, glycinebetaine, and sugars (Greenway and Munns 1980). Salt-tolerant plants sequester and accumulate salt into the cell vacuoles, controlling the salt concentrations in the cytosol and maintaining a high cytosolic K/Na ratio in their cells (Glenn et al. 1999). Therefore, ion exclusion mechanisms could provide a degree of tolerance to relatively low NaCl concentrations but would not work at high

238

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salt concentrations, resulting in the inhibition of key metabolic processes with the concomitant growth reduction. Clearly, three key processes that contribute to salt tolerance at the cellular level include (1) the establishment of cellular ion homeostasis, (2) synthesis of compatible solutes for osmotic adjustment, and (3) increased ability of the cells to neutralize reactive oxygen species generated during the stress response. Also, the regulatory mechanisms that govern these parameters are important. Transgenics against salt stress have largely been produced by manipulating these key processes. We now detail each of these categories of protection mechanisms. Transgenics produced by modifying ion homeostasis mechanisms Although Na is required in some plants, particularly halophytes (Glenn et al. 1999), a high NaCl concentration is toxic to plant growth. At high NaCl concentrations, Na enters through pathways that function in the acquisition of K and alters the ion ratios in plants (Blumwald et al. 2000). The sensitivity of cytosolic enzymes to Na is similar in both glycophytes and halophytes, indicating that the maintenance of a high cytosolic K/Na concentration ratio is a key requirement for plant growth in high salt (Glenn et al. 1999). Strategies that plants could use in order to maintain a high K/Na ratio in the cytosol include (1) extrusion of Na ions out of the cell, and (2) vacuolar compartmentation of Na ions. Given the negative membrane potential difference at the plasma membrane (140 mV) (Higinbotham 1973), a rise in extracellular Na concentration will establish a large electrochemical gradient favoring the passive transport of Na into the cells. Sodium ions can enter the cell through a number of low- and highaffinity K carriers. Three classes of low-affinity K channels have been identified. Inward-rectifying channels (KIRC), such as AKT1 (Sentenac et al. 1992), activate K influx upon plasma-membrane hyperpolarization and display a high K/Na selectivity ratio. A knockout mutant of AKT1 in Arabidopsis (akt1-1) exhibited similar sensitivity to salt as the wild type, suggesting that this channel does not play a role in Na uptake (Spalding et al. 1999). K outward-rectifying channels (KORCs) could play a role in mediating the influx of Na into plant cells. KORC channels showed a high selectivity for K over Na in barley roots (Wegner and Raschke 1994), and a somewhat lower K/Na selectivity ratio in Arabidopsis root cells (Maathuis and Sanders 1995). These channels, which open during the depolarization of the plasma membrane (i.e., upon a shift in the electrical potential difference to more positive values), could mediate the efflux of K and the influx of Na ions (Maathuis and Sanders 1997). Voltage-independent cation (VIC) channels in plant plasma membranes have been reported (de Boer and Wegner 1997; Roberts and Tester 1997).

TRANSGENIC STRATEGIES

239

VIC channels have a relatively high Na/K selectivity, are not gated by voltage, and provide a pathway for the entry of Na into plant cells (Maathuis and Amtmann 1999). Among the carriers mediating Na transport, AtHKT1 from Arabidopsis was shown to function as a selective Na transporter and, to a lesser extent, to mediate K transport (Uozumi et al. 2000). AtHKT1 was identified as a regulator of Na influx in plant roots. This conclusion was based on the capacity of hkt1 mutants to suppress Na accumulation and sodium hypersensitivity in a sos3 (salt-overly-sensitive) mutant background (Rus et al. 2001), suggesting that AtHKT1 is a salt-tolerance determinant that controls the entry of Na into the roots. Na extrusion from the plant cell cytosol is powered by the operation of the plasma membrane H-ATPase generating an electrochemical H gradient that allows plasma membrane Na/H antiporters to couple the passive movement of H inside the cells, along its electrochemical potential, to the active extrusion of Na (Blumwald et al. 2000). AtSOS1 from Arabidopsis thaliana has been shown to encode a plasma membrane Na/H antiporter with significant sequence similarity to plasma membrane Na/H antiporters from bacteria and fungi (Shi et al. 2000). The overexpression of SOS1 improved the salt tolerance of Arabidopsis, demonstrating that improved salt tolerance can be attained by limiting Na accumulation in plant cells (Shi et al. 2003) (Table 8-1). The compartmentation of Na ions into vacuoles also provides an efficient mechanism to avert the toxic effects of Na in the cytosol. The transport of Na into the vacuoles is mediated by an Na/H antiporter that is driven by the electrochemical gradient of protons generated by the vacuolar H-translocating enzymes, the H-ATPase, and the H-PPase (Blumwald 1987). The overexpression of AtNHX1 (a vacuolar Na/H antiporter from Arabidopsis) in Arabidopsis resulted in transgenic plants that were able to grow in high salt concentrations (Apse et al. 1999). The paramount role of Na compartmentation in plant salt tolerance has been further demonstrated in transgenic tomato plants overexpressing AtNHX1 (Zhang and Blumwald 2001). The transgenic tomato plants grown in the presence of 200 mM NaCl were able to grow, flower, and set fruit. Although the leaves accumulated high Na concentrations, tomato fruits displayed minimal amounts of Na (Zhang and Blumwald 2001). Similar results were obtained with transgenic Brassica napus (canola) overexpressing AtNHX1 (Zhang et al. 2001). Leaves of transgenic plants grown in the presence of 200 mM NaCl accumulated Na to up to 6% of their dry weight, but the seed yields and oil quality were not affected, demonstrating the potential use of these technologies for agricultural use in saline soils. Since the initial overexpression experiments, AtNHX1 has been widely employed for generating salt-tolerant transgenics among various species

240

TABLE 8-1. Selective List of Salt-Stress-Tolerant Transgenic Plants Gene

Source Organism

Trans Host

K transporter

AtHKT1 SKC1

A. thaliana O. sativa

A. thaliana O. sativa

Salt tolerance Salt tolerance

Horie et al., 2006 Ren et al., 2005

Vacuolar Na/H antiporter

AeNHX1 AtNHX1 AtNHX1

A. elongatum A. thaliana A. thaliana

A. thaliana A. thaliana B. napus

Qiao et al., 2007 Apse et al., 1999 Zhang et al., 2001

AtNHX1 AtNHX1 AtNHX1 AtNHX1 AtNHX5 OsNHX1 SeNHX1 PgNHX1 TaNHX1 SsNHX1 SOD2 SOD2

A. thaliana A. thaliana A. thaliana A. thaliana A. thaliana O. sativa S. europaea P. glaucum T. aestivum S. salsa S. pombe S. pombe

F. esculentum G. hirsutum F. arundinacea T. aestivum T. fournieri O. sativa N. tabacum O. sativa A. thaliana O. sativa A. thaliana O. sativa

Salt tolerance Salt tolerance, growth, fruit yield Salt tolerance, growth, seed yield, and seed oil quality Salt tolerance Salt tolerance in photosynthesis and yield Salt tolerance Salt tolerance for grain yield in the field Salt tolerance Salt tolerance Increased biomass under salt stress Increased biomass and flowering under salt stress Salt tolerance, enhanced osmotic adjustment Salt tolerance Salt tolerance; higher plant K/Na ratio Salt tolerance

AVP1

A. thaliana

A. thaliana

SsVP-2 TaVP1 TsVP TsVP

S. salsa T. aestivum T. halophila T. halophila

A. thaliana A. thaliana N. tabacum G. hirsutum

OsARP

O. sativa

N. tabacum

Protein Function

Parameters Studied and Effect

Reference

Antiporters/transporters

Vacuolar H pyrophosphatase

Antiporter-regulating protein

Salt tolerance in growth and sustained plant water status Salt tolerance Salt tolerance Salt tolerance Salt tolerance, improved growth, and photosynthetic performance Salt tolerance by Na compartmentation

Chen et al., 2008 He et al., 2005 Zhao et al., 2007 Xue et al., 2004 Shi et al. 2008 Fukuda et al., 2004 Zhou et al., 2008 Verma et al., 2007 Brini et al., 2007 Zhao et al., 2006b Gao et al., 2004 Zhao et al., 2006a Gaxiola et al., 2001 Guo et al., 2006 Brini et al., 2007 Gao et al., 2006 Lv et al., 2008 Uddin et al., 2006

Cation/proton antiporter

GmCAX1

G. max

A. thaliana

Salt tolerance

Luo et al., 2005

Tonoplast intrinsic protein

PgTIP1

P. ginseng

A. thaliana

Salt tolerance; root dependant drought tolerance

Peng et al., 2007

Promote K/Na selectivity

HAL1

S. cerevisiae

L. esculentum

Salt tolerance in growth and fruit production

Rus et al., 2001

HAL1

S. cerevisiae

C. lanatus

Salt tolerance in growth

Ellul et al., 2003

SOS3

A. thaliana

A. thaliana

Salt tolerance

Horie et al., 2006

Sodium accumulation in roots Osmolyte synthesis Apoplastic invertase

Apo-Inv

S. cerevisiae

N. tabacum

Salt tolerance, high osmotic pressure

Fukushima et al., 2001

ABA-, stress- and ripening-induced (ASR) protein

LLA23

L. longiflorum

A. thaliana

Drought and salt resistance

Yang et al., 2005

Betaine aldehyde dehydrogenase

BADH

D. carota

D. carota

Salinity tolerance

Kumar et al., 2004

BADH BADH BADH

S. liaotungensis S. oleracea S. bicolor

Z. mays N. tabacum L. esculentum

Salinity tolerance Salinity tolerance Maintenance of osmotic potential

Wu et al., 2007 Yang et al., 2008 Moghaieb et al. 2008

Choline dehydrogenase

betA

E. coli

N. tabacum

Increased tolerance to salinity stress

Lilius et al., 1996

Choline monooxygenase

CMO

A. hortensis

N. tabacum

Shen et al., 2002

SoCHO

S. oleracea

O. sativa

Better in vitro growth under salinity and osmotic stress Salt tolerance, no obvious phenotype,

codA

E. coli

A. thaliana, B. napus, N. tabacum

Increased stress tolerance

Huang et al., 2000

Choline oxidase

Shirasawa et al., 2006

241

(continued)

242

TABLE 8-1. Selective List of Salt-Stress-Tolerant Transgenic Plants (Continued) Protein Function

Gene

Source Organism

Trans Host

codA codA codA

E. coli E. coli A. thaliana

A. thaliana S. tuberosum

codA codA COX

A. globiformis A. globiformis A. pascens

Ectoin accumulation in chloroplasts

Ect A, Ect C

Chloroplastic glutamine synthetase

Parameters Studied and Effect

Reference

O. sativa O. sativa O. sativa

Salt tolerance in terms of reproduction Higher water content and biomass under stress Seedlings tolerant to salinity stress and increased germination under cold Increased tolerance to salinity and cold Recovery from a week long salt stress Salt tolerance

Sulpice et al., 2003 Ahmad et al., 2007 Hayashi et al., 1997, Alia et al., 1999 Sakamoto et al. 1998 Mohanty et al., 2003 Su et al., 2006

M. halophilus

N. tabacum

Salt and cold tolerance

Rai et al., 2006

GS2

O. sativa

O. sativa

Increased salinity resistance and chilling tolerance

Hoshida et al., 2000

Myo-inositol o-methyltransferase

IMT1

M. crystallinum

N. tabacum

Better CO2 fixation under salinity stress

Sheveleva et al., 1997

Group 3 LEA protein gene

HVA1

H. vulgare

A. sativa

Salinity tolerance in yield/plant

Oraby et al., 2005

HVA1 HVA1 Rab17 ME-leaN4

H. vulgare H. vulgare Z. mays B. napus

O. sativa M. indica A. thaliana L. sativa

Rohila et al., 2002 Lal et al., 2008 Figueras et al., 2004 Park et al. ,2005a

ME-leaN4

B. napus

B. campestris ssp. pekinensis

Drought and salinity tolerance Salinity and drought resistance Resistance to osmotic and salinity stress Enhanced growth and delayed wilting under drought. Salt resistance Drought and salt resistance

PcINO1

P. coarctata

N. tabacum

Salt tolerance

Majee et al., 2004

PcINO1

P. coarctata

O. sativa, B. napus

Salt tolerance with increased amounts of inositol

Das-Chatterjee et al., 2006

L-myo-inositol 1-P synthase

Park et al., 2005b

OsINO1

O. sativa

O. sativa, B. napus

Salt tolerance with increased amounts of inositol

Das-Chatterjee et al., 2006

Mannose-6-phosphate reductase

M6PR

A. graveolens

A. thaliana

Mannitol accumulation under salt stress leading to salt tolerance

Zhifang and Loescher, 2003

Mannitol-1-phosphate dehydrogenase

mt1D

E. coli

A. thaliana

Increased germination under salinity stress

Thomas et al., 1995

mt1D

E. coli

P. tomentosa

Salinity tolerance

Hu et al., 2005

Osmotin

Osm1

N. tabacum

Fragaria x ananassa

Proline accumulation and salt tolerance

Husaini and Abdin, 2008

Pyrroline carboxylate synthase

P5CS

A. thaliana

S. tuberosum

Salinity tolerance

P5CS

V. aconitifolia

O. sativa

Resistance to water and sainity stress

Hmida-Sayari et al., 2005 Su and Wu, 2004

SAMDC

H. sapiens

N. tabacum

Salt tolerance

Waie and Rajam, 2003

S-adenosylmethionine decarboxylase Spermidine synthase

SPE

C. ficifolia

A. thaliana

Chilling, freezing, salinity, drought hyperosmosis

Kasukabe et al., 2004

Trehalose synthesis

TPP1 TPS; TPP TPS1

O. sativa S. cerevisiae S. cerevisiae

O. sativa A. thaliana L. esculentum

Salt and cold tolerance Drought, freezing, salt and heat tolerance Drought, salt and oxidative stress tolerance

Ge et al., 2008 Miranda et al., 2007 Cortina and CulianezMacia, 2005

Oxidative stress responsers

243

Glutathione peroxidase (GPX)-like proteins

ApGPX2 Synechocystis and PCC 6803 ApGPX2

A. thaliana

Oxidative stress, drought and salt resistance

Gaber et al., 2006

Dehydroascrobate reductase

DHAR

A. thaliana

Salt tolerance

Ushimaru et al., 2006

DHAR

A. thaliana

N. tabacum

Tolerance to ozone, drought, salt, and PEG

Elsadig et al., 2006

Glyoxylase1 and 2

Gly1; gly2

O. sativa

N. tabacum

Salt tolerance

Singla-Pareek et al., 2006

O. sativa

(continued)

244

TABLE 8-1. Selective List of Salt-Stress-Tolerant Transgenic Plants (Continued) Protein Function Glutathione S-transferase

Source Organism

Trans Host

GST

S. salsa

O. sativa

Salt and chilling resistance

Zhao and Zhang, 2006

Gene

Parameters Studied and Effect

Reference

Nt107

N. tabacum

N. tabacum

Sustained growth under cold and salinity stress

Roxas et al., 1997

Catalase

katE

E. coli

N. tabacum

Salt tolerance by hydrogen peroxide scavenging

Al-Taweel et al., 2007

Mn superoxide dismutase

SOD

A. thaliana

A. thaliana

Salt tolerance

Wang et al., 2004b

Mn superoxide dismutase  Catalase

SODCAT Z. mays

B. campestris. ssp. pekinensis

Salt tolerance

Tsenga et al., 2007

Cu/Zn superoxide dismutase

SOD1

A. marina

O. sativa

Salt tolerance

Prashanth et al., 2008

Intracellular vesicle trafficking regulating protein

AtRabG3e

A. thaliana

A. thaliana

Salt and osmotic stress tolerance

Mazel et al., 2004

14-3-3 protein

TFT7

L. esculentum

A. thaliana

Salinity tolerance

Xu and Shi, 2008

Calcineurin B like protein

CBL1

A. thaliana

A. thaliana

Salt and drought tolerance and cold sensitivity

Cheong et al., 2003

CBL1

Z. mays

A. thaliana

Salt tolerance

Wang et al., 2007

Ca2dependent protein kinase

OsCDPK7

O. sativa

O. sativa

Increased cold salinity and drought tolerance

Saijo et al., 2000

Calcium binding protein 8

SCABP8

A. thaliana

A. thaliana

Salt tolerance

Quan et al., 2007

Regulatory factors

Ca2- and calmodulindependent serine/ threonine phosphatase

Calcineurin Mouse

O. sativa

Salt stress tolerance via controlled Na accumulation

Ma et al., 2005

Eukaryotic translation initiation factor 4A, eIF4A

PDH45

P. sativum

N. tabacum

Salinity tolerance in yield

Sanan-Mishra et al., 2005

Pyridoxal kinase

SOS4

A. thaliana

A. thaliana

Salt tolerance through Na/K homeostasis

Shi et al., 2002

Protein phosphatase

NtHAL3

N. tabacum

N. tabacum

Yonamine et al., 2004

AtHAL3a

A. thaliana

A. thaliana

Improved salt, osmotic and Lithium tolerance of cell cultures Regulate salinity and osmotic tolerance and plant growth

Serine/threonine protein kinase

TaSTK

T. aestivum

A. thaliana

Salt tolerance

Ge et al., 2007

CCCH-type zinc finger protein

AtSZF1 and AtSZF2

A. thaliana

A. thaliana

Salt tolerance

Sun et al., 2007

C2H2-type zinc finger protein

ZFP252

O. sativa

O. sativa

Higher proline content along with upreulation of stress responsive genes, Salt tolerance

Xu et al., 2008

bZIP Transcription factor

CAbZIP1

C. annum

A. thaliana

Disease, drought and salt tolerance

Lee et al., 2006

APETALA2/EREBP transcription factor

CAP2

C. arietinum

N. tabacum

Drought and salt tolerance

Shukla et al., 2006

JERF1 JERF3 OPBP1

L. esculentum L. esculentum N. tabacum

N. tabacum N. tabacum N. tabacum

Salt tolerance Salt tolerance Salinity and disease tolerance

Zhang et al., 2004 Wang et al., 2004a Guo et al., 2004

Espinosa-Ruiz et al., 1999

(continued)

245

246

TABLE 8-1. Selective List of Salt-Stress-Tolerant Transgenic Plants (Continued) Protein Function Transcription factor

Gene CBF3/ DREB1A DREB1A DREB1A OsDREB1A

Source Organism

Trans Host

A. thaliana

O. sativa

Drought and salt resistance

Oh et al., 2005

H. spontaneum A. thaliana O. sativa

P. notatum N. tabacum A. thaliana

Salinity and dehydration tolerance Salinity tolerance and dwarfing Drought, salt and freezing tolerance

James et al., 2008 Cong et al., 2008 Dubouzet et al., 2003

Parameters Studied and Effect

Reference

MYB homeodomain, and zinc finger proteins

OsMYB3R-2 O. sativa

A. thaliana

Drought, salt, freezing tolerance

Dai et al., 2007

Myb transcription factor interacting protein

STO

A. thaliana

A. thaliana

Salt tolerance

Nagaoka and Takano, 2003

NAC domain containing transcription factor

OsNAC6

O. sativa

O. sativa

Retarded growth and low reproductive yield but highly tolerant to salt stress

Nakashima et al., 2007

SNAC1 SNAC2

O. sativa O. sativa

O. sativa O. sativa

Drought and salt tolerance Cold, salinity, dehydration tolerance

Hu et al., 2006 Hu et al., 2008

RING finger protein

OsCOIN

O. sativa

O. sativa

Cold, salt and drought tolerance and overexpression of P5CS

Liu et al., 2007a

Zinc finger protein

OsiSAP1

O. sativa

N. tabacum

Multiple stress tolerance

OsiSAP8

O. sativa

O. sativa, N. tabacum

Salt drought and cold tolerance

Mukhopadhyay et al., 2004 Kanneganti and Gupta, 2008

Nucleotidase

HAL2

S. cerevisiae

L. esculentum

Salt tolerance in calli and rooting

Arrillaga et al. 1998

9-cis-epoxycarotenoid dioxygenase

NCED

S. guianensis

N. tabacum

Drought and salt tolerance via stomatal closure and enhanced SOD

Zhang et al., 2008

9-cis-epoxycarotenoid dioxygenase

VuNCED1

Z. mays

A. palustris

Salinity and drought resistance

Aswath et al., 2005

Endochitinase synthesis

CHIT33, CHIT42

T. harzianum

N. tabacum

Salt and metal toxicity resistance

Dana et al., 2006

NADP-malic enzyme

NADPME 2

O. sativa

A. thaliana

Salt tolerance

Liu et al., 2007b

Protease inhibitors

WRSI5

T. aestivum

A. thaliana

Salt tolerance

Shan et al., 2008

Role in systematic acquired resistance as well as abiotic stresses

Thaumatin

T. daniellii

N. tabacum

Salt and disease tolerance

Rajam et al., 2007

Serine rich protein

PcSrp

P. coarctata

E. coracana

Salt tolerance

Mahalakshmi et al., 2006

Yeast cadmium factor 1

YCF1

S. cerevisiae

A. thaliana

Heavy metal and salt tolerance

Koh et al., 2006

Miscellaneous

247

248

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(Brini et al. 2007; Chen et al. 2008; Verma et al. 2007; Xue et al. 2004; Zhang et al. 2001; Zhao et al. 2007; Zhou et al. 2008). Subsequently, NHX genes have been isolated from several other species. The introduction of a vacuolar Na/H antiporter from the halophyte Atriplex gmelini conferred salt tolerance in rice (Ohta et al. 2002). The overexpression of GhNHX1 from cotton into tobacco plants (Wu et al. 2004), as well as the overexpression of AtNHX1 in maize (Yin et al. 2004) resulted in enhanced salt tolerance. Brini et al. (2007) isolated TNHX1 and TVP1 (coding for a vacuolar H-pyrophosphatase) from wheat and overexpressed these in Arabidopsis. They noted that the transgenic plants in both cases were able to grow in the presence of 200 mM NaCl. Overexpression of OsNHX1 led to salt tolerance in transgenic rice plants (Fukuda et al. 2004). Tobacco plants overexpressing OsARP, a rice antiporter regulating protein, accumulated more Na in their leaves as compared to wild plants upon salt stress. Isolated tonoplast vesicles from transgenics also showed almost three-fold Na/H exchange rate as compared wild plants (Uddin et al. 2008). Additional evidence supporting the role of vacuolar transport in salt tolerance has been provided by Arabidopsis thaliana plants overexpressing AVP1, coding for the vacuolar H-pyrophosphatase, which displayed enhanced salt tolerance that was correlated with the increased ion content of the plants (Gaxiola et al. 2001). These results suggest that the enhanced vacuolar H-pumping in the transgenic plants provided additional driving force for vacuolar Na accumulation via the vacuolar Na/H antiporter. Recently, Yang et al., (2007) showed that AVP1 overexpressing Arabidopsis, rice, and tomato plants were tolerant to low P levels in soil. It has been postulated that AVP1 may help in alleviation of losses in lowphosphorus tropical/subtropical soils. The overexpression of a Thellungiella halophila vacuolar pyrophosphatase, TsVP, in tobacco caused increased salt-stress tolerance of transgenics (Gao et al. 2006). These transgenics were found to accumulate 20% to 32% more Na under salt-stress conditions as compared to the wild type. Overexpression of the same gene in cotton also led to increased tolerance to salt stress accompanied by an increased accumulation of Na, K, Ca2, Cl and soluble sugars in their roots and leaf tissues as compared to wild type plants (Lv et al. 2008). Several other transporter proteins have also been employed in production of transgenic plants. GmCAX1-overexpressing plants were found to be more tolerant to Na than to K and Li as compared to wild-type Arabidopsis (Luo et al. 2005). Transgenic wheat overexpressing HKT1 showed enhanced growth under stress as well as low Na accumulation in the roots (Laurie et al. 2002). SKC1 is involved in regulating K/Na homeostasis under salt stress, and introduction of this locus in a salt-sensitive rice cultivar rendered it tolerant to salt (Ren et al. 2005). Yeast HAL1 gene facilitates K/Na selectivity and salt tolerance of cells. Ectopic expression of HAL1 in transgenic tomato plants minimized

TRANSGENIC STRATEGIES

249

the reduction in fruit production caused by salt stress (Rus et al. 2001). Maintenance of fruit production by transgenic plants was correlated with enhanced growth under salt stress of calli derived from the plants. HAL1 transgene enhanced water and K contents in both leaf calli and leaves in the presence of salt, which indicates that HAL1 functions in plants using a similar mechanism to that in yeast, namely, by facilitating K/Na selectivity under salt stress. The same gene was used for raising transgenic watermelon, plantlets of which elongated better and produced new roots and leaves in culture media supplemented with NaCl as compared to control (Ellul et al. 2003). Similarly, yeast HAL2 gene was also used for raising transgenic tomato and, under salt stress, callus formation from hypocotyl explants was higher in transgenic-derived progenies than in the control (Arrillaga et al. 1998). Transgenics produced by modifying intracellular osmolyte levels The cellular response of salt-tolerant organisms to both long- and short-term salinity stresses includes the synthesis and accumulation of a class of osmoprotective compounds known as osmolytes. These relatively small, compatible solutes (mainly organic) include amino acids and derivatives, polyols and sugars, methylamines, and so forth. The osmolytes stabilize proteins and cellular structures and can increase the osmotic pressure of the cell (Yancey et al. 1982). This response is homeostatic for cell water status and protein integrity, which is perturbed in the face of soil solutions containing higher amounts of NaCl and the consequent loss of water from the cell. The accumulation of osmotically active compounds in the cytosol increases the osmotic potential to provide a balance between the apoplastic solution, which itself becomes more concentrated with Na and Cl ions, and the vacuolar lumen, which in halophytes can accumulate up to 1 M Na (and Cl). For a short-term stress, this may provide the cells with the ability to prevent water loss. However, for continued growth under salinity stress, an osmotic gradient (toward the cytosol) must be kept in order to maintain turgor and water uptake and facilitate cell expansion. The enhancement of proline and glycinebetaine synthesis in target plants has received much attention (Rontein et al. 2002). Two themes have emerged from the results of these efforts: (1) there are metabolic limitations on the absolute levels of the target osmolyte that can be accumulated, and (2) the degree to which the transformed plants are able to tolerate salinity stress is not necessarily correlative with the amounts of the osmoprotectants accumulated. The metabolic limitations on increasing the concentration of a given osmoprotectant are well illustrated with both proline and glycinebetaine. Initial strategies aimed at engineering higher concentrations of proline began with the overexpression of genes encoding

250

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the biosynthetic enzymes pyrroline-5-carboxylate (P5C) synthase (P5CS) and P5C reductase (P5CR) that catalyze the two steps between the substrate, glutamic acid and the product, proline. P5CS overexpression in tobacco dramatically elevated free proline in transgenic tobacco (Kishor et al. 1995) (Table 8-1). However, the regulation of free proline does not appear to be straightforward. Proline catabolism, via proline dehydrogenase (ProDH), is up-regulated by free proline and there exists a strong evidence that free proline inhibits P5CS (Roosens et al. 1999). Further, a two-fold increase in free proline was achieved in tobacco plants transformed with a P5CS modified by site-directed mutagenesis (Hong et al. 2000). This modification alleviated the feedback inhibition by proline on the P5CS activity and resulted in an improved germination and growth of seedlings under salt stress. Free cellular proline levels are also transcriptionally and translationally controlled. P5CR promoter analysis revealed that P5CR transcripts have reduced translational initiation. A 92-bp segment of the 5’UTR of P5CR was sufficient to provide increased mRNA stability and translational inhibition under salt stress to the GUS reporter gene that was ligated at the 3’ end to this small region (Hua et al. 2001). These results highlighted the complex regulation of P5CR during stress and emphasized the importance of stability and translation of P5CR mRNA during salt stress. An alternative approach to attain significant free proline levels, where antisense cDNA transformation was used to decrease ProDH expression, was utilized (Nanjo et al. 1999). Levels of proline in the transgenic Arabidopsis were twice (100 (g/g fresh weight) that of control plants grown in the absence of stress, and three times higher (600 (g/g fresh weight) than in control plants grown under stress. The high levels of proline were correlated with an improvement in tolerance to salinity, albeit for a short duration exposure to 600 mM NaCl. A mothbean P5CS gene was introduced in rice under the control of a constitutive and an inducible promoter (Su and Wu 2004). Both cases led to an increase in salt tolerance, indicating that the transgenics possessed increased growth of shoots and roots. However, stress-inducible expression caused an increase in the biomass of the transgenics under stress. Hmida-Sayari et al. (2005) introduced Arabidopsis P5CS gene into potato, leading to an increase in the proline content of transgenics, the levels of which were further enhanced on the imposition of salt stress. There has been considerably more experimentation directed at the engineering of glycinebetaine synthesis than for any other compatible solute. Unlike proline, glycinebetaine degradation is not significant in plants (Nuccio et al. 2000), but the problems of metabolic fluxes, compounded with the compartmentation of the substrate and product pools, has made the engineering of appreciable levels of glycinebetaine problematic. In plants that are naturally glycinebetaine accumulators (spinach and sugar beets), synthesis of this compound occurs in the chloroplast,

TRANSGENIC STRATEGIES

251

with two oxidation reactions from choline to glycinebetaine. The first oxidation to betaine aldehyde is catalyzed by choline monooxygenase (CMO), an iron-sulfur enzyme. Betaine aldehyde oxidation to glycinebetaine is catalyzed by betaine aldehyde dehydrogenase (BADH), a nonspecific soluble aldehyde dehydrogenase (Rathinasabapathi 2000). In E. coli, these reactions are cytosolic; in this species the first reaction is catalyzed by the protein choline dehydrogenase (CDH; NAD-dependent enzyme), which is encoded by the betA locus, and the second reaction is catalyzed by BADH, which is encoded by the betB locus. In Arthrobacter globiformis, the two oxidation steps are catalyzed by one enzyme choline oxidase (COD), which is encoded by the codA locus (Sakamoto and Murata 2000). The codA gene of A. globiformis offers an attractive alternative to the engineering of glycinebetaine synthesis as it necessitates only a single gene transformation event. This strategy was employed for engineering glycinebetaine synthesis in Arabidopsis (Hayashi et al. 1997). The 35S promoter-driven construct for transformation included the transit peptide for the small subunit of rubisco enzyme so that the COD protein would be targeted to the chloroplast. Improved salinity tolerance was obtained in transgenic Arabidopsis that accumulated, as a result of the transformation, 1 (mol/g fresh weight glycinebetaine. The same construct was used by for transformation of Brassica juncea (Prasad et al. 2000), and tolerance to salinity during germination and seedling establishment was improved markedly in the transgenic lines. Since these early reports, COD has been widely used for generating transgenic plants in a wide array of species (Mohanty et al. 2002; Sulpice et al. 2003). COX from Arthrobacter panescens, which is homologous to A. globiformis COD, was used to transform Arabidopsis, Brassica napus, and tobacco (Huang et al. 2000). This set of experiments differs from those already mentioned in that the COX protein was directed to the cytoplasm and not to the chloroplast. Improvements in tolerance to salinity and drought and freezing were observed in some transgenics from all three species, but the tolerance was variable. The levels of glycinebetaine in the transgenic plants were not significantly higher than those of wild-type plants, but increased significantly with the exogenous supply of choline to plants, suggesting that the supply of choline is a significant constraint on the synthesis of glycinebetaine (Huang et al. 2000). In another experiment, A. panescens COX was cloned under an ABA inducible promoter and targeted to the chloroplast in rice (Su et al. 2006). In this case, saline growth conditions caused an 89% increase in the amount of glycinebetaine accumulation in the transgenics. Two important issues emerge from the aforementioned discussion. The first is that the concentrations of glycinebetaine in the transgenic plants were much lower than the concentrations noted in natural accumulators. Despite the fact that these levels were not high enough to be osmotically

252

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

significant, a moderate (and significant) increase in tolerance to salinity and other stresses was conferred. This raises the possibility that the protection offered by glycinebetaine is not only osmotic, which is a point raised by several of the above researchers. This explanation was also offered by Bohnert and Shen (1999). Compatible solutes, including mannitol, may also function as scavengers of oxygen radicals, which may be supported by the results of Alia et al. (1999), where the protection to photosystem II in plants expressing codA was observed. An alternative possibility, not necessarily exclusive of the first, is that the increased level of peroxide generated by the COD/COX oxidation of choline causes an up-regulation of ascorbate peroxidase and catalase (Holmstrom et al. 2000), which may also improve the tolerance to salinity stress (Rontein et al. 2002). The second issue is that the level of glycinebetaine production in the transgenics is limited by choline. Because betaine synthesis takes place in the chloroplast, the free choline pool may not reflect its availability to the chloroplast, which may be limited in this compartment by the activity and/or abundance of choline transporters. However, a dramatic increase in glycinebetaine levels to 580 (mol/g dry weight in Arabidopsis) was shown in the transgenic plants when they were supplemented with choline in the growth medium (Huang et al. 2000). This limitation was not explored in the transgenic tobacco expressing E. coli enzymes CDH and BADH in the cytoplasm (Holmstrom et al. 2000). Although these transgenic plants demonstrated an improved tolerance to salinity, glycinebetaine levels were on the order of those mentioned above. BADH has since been used in a variety of crop species to improve tolerance (Kumar et al. 2004; Wu et al. 2007). Kumar et al. (2004) expressed BADH in plastids of carrot cells and reported that the transgenics were able to grow in the presence of even 400 mM of salt. Sakamoto and Murata (2000) also asserted that despite the similarities in tolerance exhibited by transgenic plants engineered to synthesize betaine in either the chloroplast or cytoplasm, the site of synthesis of betaine may play a role in the degree of tolerance shown. Indeed, if the betaine present in these plants is localized primarily in the chloroplast, it may be present at significant concentrations (50 mM) (Hayashi et al. 1997). However, Sakamoto and Murata (2000) downplayed the limitation of the metabolic pool of choline on the levels of glycinebetaine obtained in the engineered plants, by suggesting that the choline oxidizing activity may be the limiting factor. This argument is supported by Huang et al. (2000), who found that the levels of glycinebetaine correlated with the levels of COX activity measured in each plant. The increase in glycinebetaine with exogenous choline argues against this notion. Stronger evidence for the limitations of choline metabolism has been presented by McNeil et al. (2001). By overexpressing spinach phosphoethanolamine N-methyltransferase (PEAMT), which catalyzes the three

TRANSGENIC STRATEGIES

253

methylation reactions required for the conversion of phosphoethanolamine to phosphocholine, up to a 50-fold increase in free choline was obtained. This led to an increase in glycinebetaine levels (60%) in plants that were expressing spinach CMO and BADH in the chloroplast. Further, the addition of ethanolamine to the plant growth medium caused increased choline and glycinebetaine levels, showing that the metabolic flux through this pathway is also limited by the supply of ethanolamine. Because PEAMT is itself inhibited by phosphocholine, further engineering efforts need to include (1) the modification of PEAMT to remove this inhibition (McNeil et al. 2001), (2) increasing the supply of ethanolamine by overexpression of serine decarboxylase, and (3) resolving the compartmentation problem of choline supply and choline oxidation, either by use of choline oxidation in the cytoplasm or by finding the appropriate transporters to improve choline supply to the chloroplast (Rontein et al. 2002). Finally, as the compatible solutes are by definition nontoxic, the interchangeability of these compounds among species has held much interest (Table 8-1). The recent examples include the engineering of (1) ectoine synthesis with enzymes from the halophytic bacterium Halomonas elongata in plants (Ono et al. 1999; Nakayama et al. 2000) and (2) trehalose synthesis, which occurs in bacteria, yeast, and in extremely desiccation-tolerant plants (Goddijn and van Dun 1999) into potato (Yeo et al. 2000) and rice (Garg et al. 2002). An intriguing report on the improved tolerance to salinity in tobacco expressing yeast invertase in the apoplast highlights the potential of manipulating sucrose metabolism (Fukushima et al. 2001). The authors reported improved salt tolerance of transgenic tobacco plants expressing a yeast invertase in their apoplastic space, and concluded that the changes in sucrose metabolism in the transgenic plants protected the photosynthetic apparatus under salt stress. Overexpression of a bifunctional TPSP fusion protein (trehalose-6-phosphate synthase  trehalose-6-phosphate phosphatase) encoded by E. coli otsA and otsB genes in rice led to an enhanced salt tolerance accompanied by vigorous root and shoot growth after a stress of 100 mM NaCl for 4 weeks (Garg et al. 2002). Overexpression of a yeast chimeric bifunctional TPS-TPP enzyme in Arabidopsis caused an increased salt tolerance (Miranda et al. 2007). In a similar approach, yeast TPS has been shown to provide a shield against salt to tomato transgenic plants (Cortina and Cilianez-Macia 2005). Recently, overexpression of a rice trehalose-6-phosphate phosphatase (TPP) gene was shown to confer salt tolerance to transgenic rice plants; although there was no enhanced trehalose accumulation, the transgenic plants had higher constitutive levels of several stress genes (Ge et al. 2008). Majee et al. (2004) isolated L-myo-inositol 1-phosphate synthase gene from Porteresia and transformed it into tobacco, which was found to be capable of growth in 200 to 300 mM NaCl with retention of around

254

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

40% to 80% of photosynthetic competence along with an increase in the inositol production. The overexpression of polyols, such as mannitol (Tarczynski et al. 1993) and D-ononitol (Sheveleva et al. 1997) have been shown to contribute to enhanced drought and salt tolerance in transgenic tobacco plants. Apart from the low-molecular-weight organic compatible solutes, there is great deal of interest in imparting salt stress tolerance employing low-molecular-weight proteins. In plants, a group of hydrophilic proteins, known as “late embryogenesis abundant” (LEA) proteins, accumulate to very high levels during the last stage of seed maturation, as well as during water deficit in vegetative organs (Bray 1997). Most LEA proteins belong to a more widespread group of proteins called hydrophilins. The physicochemical characteristics that define this group of proteins are the presence of glycine in percentage exceeding 6% as well as a hydrophilicity index of greater than 1. Overexpression of some plant hydrophilins in plants has been seen to confer tolerance to water-deficit conditions (Xu et al. 1996). In recent years, overexpression of hydrophilins across different genera has also been seen to account for salt tolerance. Oraby et al. (2005) utilized barley HVA1 gene for raising oat transgenics that showed significant increase in tolerance to salt-stress conditions as well as improved agronomic characteristics. When barley HVA1 was overexpressed in mulberry, transgenic plants showed better cell membrane stability and photosynthetic yield, less photo-oxidative damage, and better water-use efficiency under salt-stress conditions (Lal et al. 2008). Overexpression of maize Rab17 in Arabidopsis has also been shown to provide salt tolerance (Figueras et al. 2004). An LEA protein from rape seed has been shown to provide tolerance to lettuce, as well as to Chinese cabbage (Park et al. 2005a,b). Overexpression of a lily ASR protein in Arabidopsis also caused salt tolerance (Yang et al. 2005). Transgenic plants expressing proteins involved in antioxidant protection An important aspect of salinity stress in plants is the stress-induced production of reactive oxygen species (ROS), including superoxide radicals (O 2), hydrogen peroxide (H 2O 2), and hydroxyl radicals (OH ·). ROS are a product of altered chloroplast and mitochondria metabolism during stress. These ROS cause oxidative damage to different cellular components, including membrane lipids, proteins, and nucleic acids (Halliwell and Gutteridge 1986). The alleviation of this oxidative damage could provide enhanced plant resistance to stress. Plants use lowmolecular-mass antioxidants, such as ascorbic acid and reduced glutathione, and employ a diverse array of enzymes, such as superoxide dismutases (SODs), catalases (CATs), ascorbate peroxidases (APXs),

TRANSGENIC STRATEGIES

255

glutathione S-transferases (GSTs) and glutathione peroxidases (GPXs) to scavenge ROS. Transgenic rice overexpressing yeast mitochondrial Mn-dependent SOD displayed enhanced salt tolerance (Tanaka et al. 1999) (Table 8-1). Prashanth et al. (2008) employed SOD isolated from Avicennia marina for raising transgenic rice plants and showed that the transgenics were able to survive a stress of 150 mM NaCl applied for eight days. The overexpression of a cell wall peroxidase in tobacco plants improved germination under osmotic stress (Amaya et al. 1999). Overexpression of an E. coli katE gene coding for catalase in tobacco showed that katE increased the resistance of chloroplast’s translational machinery to salt stress by scavenging hydrogen peroxide (Al-Taweel et al. 2007). Transgenic tobacco plants overexpressing both GST and GPX displayed improved seed germination and seedling growth under stress (Roxas et al. 1997). Subsequent studies (Roxas et al. 2000) demonstrated that in addition to increased GST/GPX activities, the transgenic seedlings contained higher levels of glutathione and ascorbate than wild-type seedlings, showed higher levels of monodehydroascorbate reductase activity, and the glutathione pools were more oxidized. These results indicated that the increased glutathione-dependent peroxidase scavenging activity and the associated changes in glutathione and ascorbate metabolism led to reduced oxidative damage in the transgenic plants and contributed to their increased salt tolerance. During stress, plants display an increase in the generation of H2O2 and other ROS (Gueta-Dahan et al. 1997; Roxas et al. 2000). GPX-1 and -2 from Synechocystis PCC 6803 were used to generate transgenic Arabidopsis plants, which showed enhanced tolerance to salt tolerance (Gaber et al. 2006). The major substrate for the reductive detoxification of H2O2 is ascorbate, which must be continuously regenerated from its oxidized form. An important function of glutathione in protection against oxidative stress is the reduction of ascorbate via the ascorbate-glutathione cycle, where GSH acts as a recycled intermediate in the reduction of H2O2 (Foyer and Halliwell 1976). Dehydroascorbate reductase (DHAR) is one of the two enzymes important for regeneration of the ascorbate during the process of antioxidation. DHAR is therefore essential for ascorbate recycling. A rice DHAR, when overexpressed in Arabidopsis, led to an increase in resistance to salt tolerance (Ushimaru et al. 2006). Similar overexpression of an Arabidopsis DHAR in tobacco caused increased salt tolerance in terms of higher net photosynthesis. Ruiz and Blumwald (2002) investigated the enzymatic pathways leading to glutathione synthesis during the response to salt stress of wild-type and salt-tolerant Brassica napus (canola) plants overexpressing a vacuolar Na/H antiporter (Zhang et al. 2001). Under salt stress, the wild-type plants showed a marked increased in the activity of enzymes associated

256

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

with cysteine synthesis (the crucial step for assimilation of reduced sulfur into organic compounds such as glutathione), resulting in a significant increase in GSH content. Conversely, these activities were unchanged in the transgenic salt-tolerant plants and their GSH content did not change with salt stress. These results clearly showed that salt stress induced an increase in the assimilation of sulfur, and the biosynthesis of cysteine and GSH aimed to mitigate the salt-induced oxidative stress. The small changes seen in the transgenic plants overexpressing the vacuolar Na/H (Zhang and Blumwald 2001) suggested that the accumulation of excess Na in the vacuoles (and the maintenance of a high cytosolic K/Na ratio) greatly diminished the salt-induced oxidative stress, highlighting the important role of Na homeostasis during salt stress. GlyI and GlyII gene products are required for glutathione-based detoxification of methylglyoxal. Singla-Pareek et al. (2003) showed that overexpression of GlyIGlyII in tobacco led to increased salt tolerance, and there was only 5% loss in total productivity when grown in the presence of 200 mM NaCl. A yeast gene, YCF1, encoding a cadmium factor, which sequesters glutathione-chelates of heavy metals and xenobiotics into vacuoles, has been shown to provide tolerance to Arabidopsis plants (Koh et al. 2006). Overexpression of a yeast SOD2 gene led to improved seed germination and seedling salt tolerance in both Arabidopsis and rice (Gao et al. 2004; Zhao et al. 2006b). Transgenic plants with altered expression of regulatory proteins Several regulatory proteins (transcription factors, kinases, phosphatases, etc.) have been utilized for generating stress tolerant transgenics in various species. The Rab family of monomeric GTPases is conserved from yeast to animals and has been implicated in intracellular vesicle trafficking, as well as in the organization of membranes (Zerial and McBride 2001). Rab proteins continuously cycle between the GTP- and GDP-bound states and between cytosol and membrane compartments. Rab proteins thus regulate several molecular interactions through their effector regions that are specific for individual Rabs. Transgenic Arabidopsis plants overexpressing AtRabG3e were found to be more salt-tolerant and accumulated Na in the vacuoles, and had reduced accumulation of reactive oxygen species during salt stress (Mazel et al. 2004). Transcription factors have been the most widely used proteins for overexpression studies because they can regulate several downstream genes. Kasuga et al. (1999) developed transgenic plants using DREB1A gene under the control of RD29A promoter and showed that the plants were tolerant to multiple stresses, including salt stress. Overexpression of rice DREB1A in Arabidopsis induced overexpression of target stressresponsive genes and contributed to salt tolerance (Dubouzet et al. 2003).

TRANSGENIC STRATEGIES

257

James et al. (2008) isolated a DREB1 homolog from wild barley and introduced it into bahiagrass (Paspalum notatum Flugge), thereby increasing its tolerance to salt several-fold. Overexpression of AtSZF1, a CCCH-type zinc-finger protein, imparted salt tolerance to transgenic Arabidopsis that showed reduced induction of salt stress-responsive genes (Sun et al. 2007). Mukhopadhyay et al. (2004) overexpressed OSISAP1, a rice gene, in tobacco and showed tolerance of transgenic tobacco to salt stress. OsiSAP8, a rice gene coding for a zinc-finger protein, when overexpressed in both rice and tobacco led to salt tolerance via enhancing germination, and gain in fresh weight after stress recovery (Kanneganti and Gupta 2008). Rice TFIIIA-type zinc-finger protein gene has been overexpressed in rice and shown to accumulate more free proline and soluble sugars and also elevate the level of transcripts of stress-responsive genes (Xu et al. 2008). An APETALA2-like protein from chickpeas was overexpressed in tobacco, which led to salt tolerance along with up-regulation of several genes like ERD10B, ERD10C, and IAA4.2 and caused drastic increase in leaf cell size and number of lateral roots (Shukla et al. 2006). Overexpression of Arabidopsis CBF3 in rice caused increased tolerance to salt and, in contrast to Arabidopsis, there was no change in the phenotype of the plants (Oh et al. 2005). Overexpression of JERF1 and JERF3, two ethyleneresponsive transcription factors in tobacco and tomatoes, respectively, led to salt tolerance along with activated expression of GCC box containing genes under normal growth conditions (Wang et al. 2004a; Zhang et al. 2004). Overexpression of OsCOIN, a RING finger protein, in transgenic rice lines significantly enhanced their tolerance to cold, salt, and drought, accompanied by an up-regulation of OsP5CS expression and an increase of cellular proline level (Liu et al. 2007b). Guo et al. (2004) overexpressed the osmotin promoter binding protein 1 (OSBP1), an AP2/EREBP-like transcription factor, in tobacco and showed their salt tolerance; these researchers hypothesized that OSBP1 might be a transcriptional regulator capable of regulating expression in sets of stress-related genes. OsMYB3R2, an R2R3 MYB transcription factor, when overexpressed in Arabidopsis caused enhanced salt tolerance, as well as increase in the steady-state levels of some stress-responsive genes like DREB2A and COR15A (Dai et al. 2007). Overexpression of rice NAC transcription factors, SNAC1 and SNAC2, showed increased salt tolerance with a concomitant increase in the steady-state levels of mRNA of several stressinduced genes (Hu et al. 2006, 2008). Calcium ions represent both an integrative signal and an important convergence point of many disparate signaling pathways. Calcium-binding proteins, like calcineurin B-like (CBL) proteins, have been implicated as important relays in calcium signaling. Overexpression of Arabidopsis CBL1 reduced transpirational water loss and induced the expression of early stress-responsive transcription factors and stress adaptation genes

258

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

in nonstressed plants (Albrecht et al. 2003). Wang et al. (2007) overexpressed maize CBL4 in Arabidopsis and found that the transgenic plants exhibited salt tolerance at germination and seedling stages, along with lesser accumulation of Na and Li. Overexpression of a wheat serine/threonine protein kinase gene in Arabidopsis has been shown to increase the growth of roots under salt stress and improve salt tolerance (Ge et al. 2007). Transgenic plants with altered expression of other classes of proteins Overexpression of several other types of proteins has also been seen to improve the salt tolerance of the transgenic plants. NADP-malic enzyme (NADP-ME) functions in many different pathways in plants and has recently been implicated in plant defense, such as in responses to wounding and UV-B radiation. Transgenic Arabidopsis plants overexpressing a, NADP-ME gene from rice grew well in MS medium with 100 mM NaCl, whereas growth of wild-type Arabidopsis seedlings was strongly inhibited. In addition, under 125 mM NaCl stress, the root lengths of transgenic lines were about twice as long as those of the wild type (Liu et al. 2007a). Transgenic fingermillet expressing a gene encoding a serine-rich protein (Srp) isolated from Porterisia, a wild rice relative, were able to reach maturity under salt stress of 200 mM NaCl and also set seeds. The roots of Srp overexpressing transgenics accumulated more amounts of Na and K as compared to wild-type plants (Mahalakshmi et al. 2006). An aquaporin isolated from Panax ginseng was able to confer salt tolerance to Arabidopsis plants that showed superior growth and seed germination under salt stress (Peng et al. 2007). A pea DNA helicase, PDH45, similar to eukaryotic translation initiation factor eIF4A, when overexpressed in tobacco provided salt tolerance along with accumulation of Na in older leaves and negligible amount in seeds (Sanan-Mishra et al. 2005). Overexpression of a salt tolerance-related gene, STO also caused increased salt tolerance in Arabidopsis. The transgenic plants showed a 33% to 70% increase in root growth as compared to wild-type plants (Nagaoka and Takano 2003).

CONCLUSIONS AND PERSPECTIVES Degradation of agricultural land and water supplies is a result of the intensive agricultural practices employed in developed and developing countries. Ideally, these practices should be changed to a more rational use of land and water resources, but this is unlikely in the foreseeable future. For example, mixed cropping with perennials and trees would alleviate the accumulation of Na and other salts in the upper soil layers. Nevertheless, this kind of change in farming systems and the develop-

TRANSGENIC STRATEGIES

259

ment of new products are likely a long and difficult process, since it will require the use of new land and will not address the problem of growing crops in land that is already compromised. The development and use of crops that can tolerate high levels of salinity in soils is a practical solution, at least for the time being. Conventional breeding programs for raising salt-tolerant genotypes have met with limited success, in part because breeders prefer to evaluate their genetic material in ideal conditions. The issue of raising salt-tolerant transgenic crops is gaining increased support due to the growing interest of commercial seed companies in making agriculture a profitable enterprise. From a business perspective, in order for plant breeding companies to invest in the development of new varieties with enhanced stress tolerance, there will always be the question as to whether the investment in the development of these cultivars is worth the effort. There is no benefit in developing salinity-tolerant plants unless there are economic drivers that will allow the plants to be competitively productive with non-saline-tolerant plants growing on uncompromised soil. These viewpoints might differ from those of basic researchers for whom a significant, although small, increase in salt tolerance is worth the effort. In evaluating the possibility of improving stress tolerance in transgenic plants, we propose a number of considerations that the research community should consider. First, although it has been recognized by many researchers that there are dramatic changes in gene expression associated with all types of stresses, the promoters that are most commonly used for transgene modifications are primarily constitutively expressed, including the CaMV35S promoter, ubiquitin, and actin promoters (Grover et al. 2003). Recent studies have noted that the overexpression of specific stressinduced genes under the control of stress-induced or tissue-specific promoters often displays a better phenotype than the same genes expressed under a constitutive promoter (Kasuga et al. 1999; Zhu et al. 1998). Second, although there have been a number of successes in the production of abiotic stress-tolerant plants using tobacco or Arabidopsis, there is a clear need to begin to introduce these tolerance genes into crop plants. Moreover, even though researchers tend to focus on a few basic plant systems (with Arabidopsis, tobacco, and rice being the major species of choice), there has been no attempt to choose specific genetic backgrounds. Third, it is likely that the effectiveness of a specific transgene will be based on the specific genetic background into which it is transformed. One component of this is the well-known phenomenon of “position effect”; moreover, the ability of a transgene to work may well be determined by the overall genetic background, independent from the chromosomal location of the transgene, referred to as “transgene combining ability” (TCA).

260

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Finally, we also need to establish better comparative systems, as well as look at rational concepts for combining genes, just as the disease resistance researchers are now doing with gene stacking. For example, the overexpression of AtSOS1 in meristems (nonvacuolated cells) and AtNHX1 (for vacuolar Na accumulation), together with the overproduction of compatible solutes, would provide not only the ability of using NaCl as an osmoticum during vegetative growth but also would provide the seedlings with the ability to reduce Na toxicity during early growth and seedling establishment. Wherever applicable, genes for protection against oxidative stress must be combined, particularly in actively photosynthesizing cells that are prone to more chloroplast damage due to ROS. Over the last 50 years many researchers have argued for the development of salt-tolerant crops from true halophytes. Although halophytes are present in a wide diversity of plant forms, to date few halophytic crops have been able to compete effectively with glycophytic crops (Glenn et al. 1999). Moreover, research on the physiology of tolerance suggests that the overall trait is determined by a number of subtraits, any of which might, in turn, be determined by any number of genes. We believe that by comparing different genes and genetic combinations, researchers will be able to advance the field more quickly and develop stress-tolerant germplasms. While progress in improving stress tolerance using glycophytes has been slow, there are a number of opportunities and reasons for optimism. Over the last 10 years a number of functional tools have been developed that can allow us to dissect many of the fundamental questions associated with stress tolerance. These include (1) the development of molecular markers for gene mapping and the construction of associated maps; (2) development of EST libraries; (3) complete sequencing of plant genomes, including Arabidopsis, rice, and maize; (4) production of T-DNA or transposon-tagged mutagenic populations of Arabidopsis; and (5) development of a number of forward genetics tools that can be used in gene function analysis, such as TILLING (Colbert et al. 2001). In addition to these, advancements in tools and techniques in structural and functional genomics have opened unparalleled avenues in gene discovery. We need to focus on looking at the comparative effects and interactions of specific transgenes within a defined genetic background and determine the efficacy of these approaches in the field. SUMMARY Major advances have been made in plant genetic engineering during the past 25 years, and the methodologies for stable genetic transformation, as well as for regulation of the introduced trans-genes, have been highly optimized. While the bulk of the work carried out toward development of transgenic crops has dealt mainly with disease resistance and herbicide-

TRANSGENIC STRATEGIES

261

tolerant phenotypes, the production of abiotic stress-tolerant plants has become a reality, at least in laboratory-based experiments in recent years. Salt-tolerant transgenics have been produced by altering levels of proteins that play roles in ion transport, osmolyte synthesis, and protection against oxidative damage, as well as those that play role in regulatory functions that control stress perception and signaling, transcriptional, posttranscriptional, and translational activities. The transgenic plants made from the aforementioned tools have raised a great deal of enthusiasm; nevertheless, more must be achieved in terms of practical applications at the field level. The current upsurge in genomic research has the potential to accelerate efforts in discovering novel stress-responsive genes. There are also possibilities of engineering the whole cascade of multiple genetic changes through the manipulation of the regulatory genes. More basic research is needed to translate the success of laboratory experiments into field-level product development.

REFERENCES Abel, P. P., Nelson, R. S., De, B., Hoffmann, N., Rogers, S. G., Fraley, R. T., and Beachy, R. N. (1986). “Delay of disease development in transgenic plants that express the tobacco mosaic virus coat protein gene.” Science, 232, 738–743. Akbar, M., and Yabuno, T. (1977). “Breeding saline-resistant varieties of rice. IV. Inheritance of delayed type panicle sterility induced by salinity.” Jap. J. Breed., 27, 237–240. Albrecht, V., Weinl, S., Blazevic, D., D’Angelo, C., Batistic, O., Kolukisaoglu, U., Bock, R., Schulz, B., Harter, K., and Kudla, J. (2003). “The calcium sensor CBL1 integrates plant responses to abiotic stresses.” Plant J., 36, 457–470. Al-Doss, A. A., and Smith, S. E. (1998). “Registration of AZ-97MEC and AZ-97MECST very non-dormant alfalfa germplasm pools with increased shoot weight and differential response to saline irrigation.” Crop Sci., 38, 568. Alia, Kondo Y., Sakamoto, A., Nonaka, H., Hayashi, H., Saradhi., P. P., Chen, T. H. H., and Murata, N. (1999). “Enhanced tolerance to light stress of transgenic Arabidopsis plants that express the codA gene for a bacterial choline oxidase.” Plant Mol. Biol., 40, 279–288. Al-Taweel, K., Iwaki, T., Yabuta, Y., Shigeoka, S., Murata, N., and Wadano, A. (2007). “A bacterial transgene for catalase protects translation of d1 protein during exposure of salt-stressed tobacco leaves to strong light.” Plant Physiol., 145, 258–265. Amaya, I., Botella, M. A., dela Calle, M., Medina, M. I., Heredia, A., Bressan, R. A., Hasegawa, P. M., Quesada, M.A., and Valpuesta, V. (1999). “Improved germination under osmotic stress of tobacco plants overexpressing a cell wall peroxidase.” FEBS Lett., 457, 80–84. Apse, M. P., Aharon, G. S., Snedden, W. A., and Blumwald, E. (1999). “Salt tolerance conferred by overexpression of a vacuolar Na/H antiport in Arabidopsis.“ Science, 285, 1256–1258.

262

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Arrillaga, I., Gil-Mascarell, R., Gisbert, C., Sales, E., Montesinos, C., Serrano, R., and Moreno, V. (1998). “Expression of the yeast HAL2 gene in tomato increases the in vitro salt tolerance of transgenic progenies.” Plant Sci., 136, 219–226. Aswath, C. R., Kim, S., Mo, S., and Kim, D. H. (2005). “Transgenic plants of creeping bent grass harboring the stress inducible gene, 9-cis-epoxycarotenoid dioxygenase, are highly tolerant to drought and NaCl stress.” Plant Growth Regul., 47, 129–139. Blumwald, E. (1987). “Tonoplast vesicles for the study of ion transport in plant vacuoles.” Physiol. Plant., 69, 731–734. Blumwald, E., Aharon, G. S., and Apse, M. P. (2000). “Sodium transport in plant cells.” Biochim. Biophys. Acta, 1465, 140–151. Bohnert, H. J., and Shen, B. (1999). “Transformation and compatible solutes.” Sci. Hort., 78, 237–260. Bray, E. A. (1997). “Plant responses to water deficit.” Trends Plant Sci., 2, 48–54. Brini, F., Hanin, M., Mezghani, I., Berkowitz, G. A., and Masmoudi, K. (2007). “Overexpression of wheat Na/H antiporter TNHX1 and H-pyrophosphatase TVP1 improve salt- and drought-stress tolerance in Arabidopsis thaliana plants.” J. Exp. Bot., 58, 301–308. Broglie, K., Chet, I., Holliday, M., Cressman, R., Biddle, P., Knowlton, S., Mauvais, C. J., and Broglie, R. (1991). “Transgenic plants with enhanced resistance to the fungal pathogen Rhizoctonia solani.” Science, 254, 1194–1197. Chen, L. H., Zhang, B., and Xu, Z. Q. (2008). “Salt tolerance conferred by overexpression of Arabidopsis vacuolar Na()/H() antiporter gene AtNHX1 in common buckwheat (Fagopyrum esculentum).” Transgen. Res., 17, 121–132. Cheong, Y. H., Kim, K. N., Pandey, G. K., Gupta, R., Grant, J. J., and Luan, S. (2003). “CBL1, a calcium sensor that differentially regulates salt, drought, and cold responses in Arabidopsis.” Plant Cell, 15, 1833–1845. Colbert, T., Till, B. J., Tompa, R., Reynolds, S., Steine, M. N., Yeung, A. T., McCallum, C. M., Comai, L., and Henikoff, S. (2001). “High-throughput screening for induced point mutations.” Plant Physiol., 126, 480–484. Comai, L., Facciotti, P. D., Hiatt, W. R., Thompson, G., Rose, R. E., and Stalker, D. M. (1985). “Expression in plants of a mutant aroA gene from Salmonella typhimurium confers tolerance to glyphosate.” Nature, 317, 741–744. Cong, L., Zheng, H., Zhang, Y., and Chai, T. (2008). “Arabidopsis DREB1A confers high salinity tolerance and regulates the expression of GA dioxygenases in tobacco.” Plant Sci., 174, 156–164. Cortina, C., and Culianez-Macia, F. (2005). “Tomato abiotic stress enhanced tolerance by trehalose biosynthesis.” Plant Sci., 169, 75–82. Dai, X., Xu, Y., Ma, Q., Xu, W., Wang, T., Xue, Y., and Chong, K. (2007). “Overexpression of an R1R2R3 MYB gene, OsMYB3R-2, increases tolerance to freezing, drought, and salt stress in transgenic Arabidopsis.” Plant Physiol., 143, 1739–1751. Dana, M. M., Pintor-Toro, J. A., and Cubero, B. (2006). “Transgenic tobacco plants overexpressing chitinases of fungal origin show enhanced resistance to biotic and abiotic stress agents.” Plant Physiol., 142, 722–730. Das-Chatterjee, A., Goswami, L., Maitra, S., Dastidar, K. G., Ray, S., and Majumder, A. L. (2006). “Introgression of a novel salt-tolerant L-myo-inositol 1-phosphate synthase from Porteresia coarctata (Roxb.). Tateoka (PcINO1)

TRANSGENIC STRATEGIES

263

confers salt tolerance to evolutionary diverse organisms.” FEBS Lett., 580, 3980–3988. de Boer, A. H., and Wegner, L. H. (1997). “Regulatory mechanisms of ion channels in xylem parenchyma cells.” J. Exp. Bot., 48, 441–449. Dierig, D. A., Shannon, M. C., and Grieve, C. M. (2001). “Registration of WCL-SL1 salt-tolerant Lesquerella fendleri germplasm.” Crop Sci., 41, 604–605. Dubouzet, J. G., Sakuma, Y., Ito, Y., Kasuga, M., Dubouzet, E. G., Miura, S., Seki, M., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2003). “OsDREB genes in rice, Oryza sativa L., encode transcription activators that function in drought-, highsalt- and cold-responsive gene expression.” Plant J., 33, 751–763. Ellul, P., Rios, G., Atares, A., Roig, L. A., Serrano, R., and Moreno, V. (2003). “The expression of the Saccharomyces cerevisiae HAL1 gene increases salt tolerance in transgenic watermelon [Citrullus lanatus (Thunb.) Matsun., and Nakai.].” Theor. Appl. Genet., 107, 462–469. Epstein, E. (1977). “Genetic potentials for solving problems of soil mineral stress: Adaptation of crops to salinity” in Plant adaptation to mineral stress in problem soils, M. J. Wright, ed., Cornell University Agricultural Experiment Station, Ithaca, New York, 73–123. ———. (1983). “Crops tolerant to salinity and other mineral stresses,” in Better crops for food, Ciba Foundation Symposium, J. Nugent and M. O’Connor, eds., Pitman, London, 61–82. Epstein, E., Norlyn, J. D., Rush, D. W., Kingsbury, R. W., Kelley, D. B., Cunningham, G. A., and Wrona, A. F. (1980). “Saline culture of crops: A genetic approach.” Science, 210, 399–404. Espinosa-Ruiz, A., Belles, J. M., Serrano, R., and Culianez-MacIa, F. A. (1999). “Arabidopsis thaliana AtHAL3: A flavoprotein related to salt and osmotic tolerance and plant growth.” Plant J., 20, 529–539. Fairbairn, D. J., Liu, W., Schachtman, D. P., Gomez-Gallego, S., Day, S. R., and Teasdale, R. D. (2000). “Characterisation of two distinct HKT1-like potassium transporters from Eucalyptus camaldulensis.” Plant Mol. Biol., 43, 515–525. Figueras, M., Pujal, L., Saleh, A., Save, R., Pages, M., and Goday, A. (2004). “Maize Rab17 overexpression in Arabidopsis plants promotes osmotic stress tolerance.” Ann. Appl. Biol., 144, 251–257. Flowers, T. J. (2004). “Improving crop salt tolerance.” J. Exp. Bot., 55, 307–319. Flowers, T. J., and Yeo, A. R. (1995). “Breeding for salinity resistance in crop plants: Where next?” Aust. J. Plant Physiol., 22, 875–884. Foyer, C., and Halliwell, B. (1976). “The presence of glutathione and glutathione reductase in chloroplasts: A proposed role in ascorbic acid metabolism.” Planta, 133, 21–25. Fukuda, A., Nakamura, A., Tagiri, A., Tanaka, H., Miyao, A., Hirochika, H., and Tanaka, Y. (2004). “Function, intracellular localization and the importance in salt tolerance of a vacuolar Na()/H() antiporter from rice.” Plant Cell Physiol., 45, 146–159. Fukushima, E., Arata, Y., Endo, T., Sonnewald, U., and Sato, F. (2001). “Improved salt tolerance of transgenic tobacco expressing apoplastic yeast-derived invertase.” Plant Cell Physiol., 42, 245–249. Gaber, A., Yoshimura, K., Yamamoto, T., Yukinori, Y., Takeda, T., Miyasaka, H., Yoshihisa, N., and Shigeoka, S. (2006). “Glutathione peroxidase-like protein of

264

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Synechocystis PCC 6803 confers tolerance to oxidative and environmental stresses in transgenic Arabidopsis.” Physiol. Plant., 128, 251–262. Gao, F., Gao, Q., Duan, X., Yue, G., Yang, A., and Zhang, J. (2006). “Cloning of an H-PPase gene from Thellungiella halophila and its heterologous expression to improve tobacco salt tolerance.” J. Exp. Bot., 57, 3259–3270. Garg, A. K., Kim, J. K., Owens, T. G., Ranwala, A. P., Choi, Y. D., Kochian, L. V., and Wu, R. J. (2002). “Trehalose accumulation in rice plants confers high tolerance levels to different abiotic stresses.” Proc. Natl. Acad. Sci. USA, 99, 15898–15903. Gaxiola, R. A., Li, J., Undurraga, S., Dang, L. M., Allen, G. J., Alper, S. L., and Fink, G. R. (2001). “Drought- and salt-tolerant plants result from overexpression of the AVP1 H-pump.” Proc. Natl. Acad. Sci. USA, 98, 11444–11449. Ge, L. F., Chao, D. Y., Shi, M., Zhu, M. Z., Gao, J. P., and Lin, H. X. (2008). “Overexpression of the trehalose-6-phosphate phosphatase gene OsTPP1 confers stress tolerance in rice and results in the activation of stress responsive genes.” Planta, 228, 191–201. Ge, R., Chen, G., Zhao, B., Shen, Y., and Huang, Z. (2007). “Cloning and functional characterization of a wheat serine/threonine kinase gene (TaSTK) related to salt-resistance.” Plant Sci., 173, 55–60. Glenn, E. P., Brown, J. J., and Blumwald, E. (1999). “Salt tolerance and crop potential of halophytes.” Crit. Rev. Plant Sci., 18, 227–255. Goddijn, O. J., and van Dun, K. (1999). “Trehalose metabolism in plants.” Trends Plant Sci., 4, 315–319. Greenway, H., and Munns, R. (1980). “Mechanisms of salt tolerance in non-halophytes.” Ann. Rev. Plant Physiol., 31, 149–190. Grover, A., Aggarwal, P. K., Kapoor, A., Katiyar-Agarwal, S., Agarwal, M., and Chandramouli, A. (2003). “Addressing abiotic stresses in agriculture through transgenic technology.” Curr. Sci., 84, 355–367. Gueta-Dahan, Y., Yaniv, Z., Zilinskas, A., and Ben-Hayyinm, G. (1997). “Salt and oxidative stress: Similar and specific responses and their relation to salt tolerance in citrus.” Planta, 203, 460–469. Guo, S., Yin, H., Zhang, X., Zhao, F., Li, P., Chen, S., Zhao, Y., and Zhang, H. (2006). “Molecular cloning and characterization of a vacuolar H-pyrophosphatase gene, SsVP, from the halophyte Suaeda salsa and its overexpression increases salt and drought tolerance of Arabidopsis.” Plant Mol. Biol., 60, 41–50. Guo, Z. J., Chen, X. J., Wu, X. L., Ling, J. Q., and Xu, P. (2004). “Overexpression of the AP2/EREBP transcription factor OPBP1 enhances disease resistance and salt tolerance in tobacco.” Plant Mol. Biol., 55, 607–618. Halliwell, B., and Gutteridge, J. M. C. (1986). “Oxygen free radicals and iron in relation to biology and medicine: Some problems and concepts.” Arch. Biochem. Biophys., 246, 501–514. Hasegawa, P. M., Bressan, R. A., Zhu, J. K., and Bohnert, H. J. (2000). “Plant cellular and molecular responses to high salinity.” Ann. Rev. Plant Physiol. Plant Mol. Biol., 51, 463–499. Hayashi, H., Alia, Mustardy, L., Deshnium, P., Ida, M., and Murata, N. (1997). “Transformation of Arabidopsis thaliana with the codA gene for choline oxidase; Accumulation of glycinebetaine and enhanced tolerance to salt and cold stress.” Plant J., 12, 133–142.

TRANSGENIC STRATEGIES

265

He, C., Yan, J., Shen, G., Fu, L., Holaday, A. S., Auld, D., Blumwald, E., and Zhang, H. (2005). “Expression of an Arabidopsis vacuolar sodium/proton antiporter gene in cotton improves photosynthetic performance under salt conditions and increases fiber yield in the field.” Plant Cell Physiol., 46, 1848–1854. Higinbotham, N. (1973). “Electropotentials of plant cells.” Ann. Rev. Plant Physiol., 24, 25–46. Hilder, V. A., Gatehouse, A. M. R., Sheerman, S. E., Barker, R. F., and Boulter, D. A. (1987). “A novel mechanism of insect resistance engineered in transgenic plants.” Nature, 330, 160–163. Hmida-Sayari, A., Gargouri-Bouzid, R., Bidani, A., Jaoua, L., Savoure, A., and Jaou, S. (2005). “Overexpression of 1-pyrroline-5-carboxylate synthetase increases proline production and confers salt tolerance in transgenic potato plants.” Plant Sci., 169, 746–752. Holmstrom, K. O., Somersalo, S., Mandal, A., Palva, T. E., and Welin, B. (2000). “Improved tolerance to salinity and low temperature in transgenic tobacco producing glycine betaine.” J. Exp. Bot., 51, 177–185. Hong, Z., Lakkineni, K., Zhang, Z., and Verma, D. P. (2000). “Removal of feedback inhibition of (1)-pyrroline-5-carboxylate synthetase results in increased proline accumulation and protection of plants from osmotic stress.” Plant Physiol., 122, 1129–1136. Horie, T., Horie, R., Chan, W.Y., Leung, H. Y., and Schroeder, J. I. (2006). “Calcium regulation of sodium hypersensitivities of sos3 and athkt1 mutants.” Plant Cell Physiol., 47, 622–633. Hoshida, H., Tanaka, Y., Hibino, T., Hayashi, Y., Tanaka, A., Takabe, T., and Takabe, T. (2000). “Enhanced tolerance to salt stress in transgenic rice that overexpresses chloroplast glutamine synthetase.” Plant Mol. Biol., 43, 103–111. Hu, H., Dai, M., Yao, J., Xiao, B., Li, X., Zhang, Q., and Xiong, L. (2006). “Overexpressing a NAM, ATAF, and CUC (NAC) transcription factor enhances drought resistance and salt tolerance in rice.” Proc. Natl. Acad. Sci. USA, 103, 12987–12992. Hu, L., Lu, H., Liu, O., Chen, X., and Jiang, X. (2005). “Overexpression of mtlD gene in transgenic Populus tomentosa improves salt tolerance through accumulation of mannitol.” Tree Physiol., 25, 1273–1281. Hu, H., You, J., Fang, Y., Zhu, X., Qi, Z., and Xiong, L. (2008). “Characterization of transcription factor gene SNAC2 conferring cold and salt tolerance in rice.” Plant Mol. Biol., 67, 169–181. Hua, X. J., VanDe Cotte, B., VanMontagu, M., and Verbruggen, N. (2001). “The 5’ untranslated region of the At-P5R gene is involved in both transcriptional and post-transcriptional regulation.” Plant J., 26, 157–169. Huang, J., Hirji, R., Adam, L., Rozwadowski, K. L., Hammerlindl, J. K., Keller, W. A., and Selvaraj, G. (2000). “Genetic engineering of glycinebetaine production toward enhancing stress tolerance in plants: Metabolic limitations.” Plant Physiol., 122, 747–756. Husaini, A. M., and Abdin, Z. A. (2008). “Development of transgenic strawberry (Fragaria x ananassa Duch.) plants tolerant to salt stress.” Plant Sci., 174, 446–455. Jacobsen, T., and Adams, R. M. (1958). “Salt and silt in ancient Mesopotamian agriculture: Progressive changes in soil salinity and sedimentation contributed to the breakup of past civilizations.” Science, 128, 1251–1258.

266

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

James, V. A., Neibaur, I., and Altpeter, F. (2008). “Stress inducible expression of the DREB1A transcription factor from xeric, Hordeum spontaneum L. in turf and forage grass (Paspalum notatum Flugge) enhances abiotic stress tolerance.” Transgenic Res., 17, 93–104. Kanneganti, V., and Gupta, A. K. (2008). “Overexpression of OsiSAP8, a member of stress associated protein (SAP) gene family of rice confers tolerance to salt, drought and cold stress in transgenic tobacco and rice.” Plant Mol. Biol., 66, 445–462. Kasuga, M., Liu, Q., Miura, S., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1999). “Improving plant drought, salt, and freezing tolerance by gene transfer of a single stress-inducible transcription factor.” Nat. Biotechnol., 17, 287–291. Kasukabe, Y., He, L., Nada, K., Misawa, S., Ihara, I., and Tachibana, S. (2004). “Overexpression of spermidine synthase enhances tolerance to multiple environmental stresses and up-regulates the expression of various stress-regulated genes in transgenic Arabidopsis thaliana.” Plant Cell Physiol., 45, 712–722. Kishor, P., Hong, Z., Miao, G. H., Hu, C., and Verma, D. (1995). “Overexpression of -pyrroline-5-carboxylate synthetase increases proline production and confers osmotolerance in transgenic plants.” Plant Physiol., 108, 1387–1394. Koh, F., Song, W., Lee, Y., Kim, K., Chung, N., Lee, K., Hong, S., and Lee, H. (2005). “Expression of yeast cadmium factor 1 (YCF1) confers salt tolerance to Arabidopsis thaliana.“ Plant Sci., 170, 534–541. Kumar S., Dhingra A., and Daniell, H. (2004). “Plastid-expressed betaine aldehyde dehydrogenase gene in carrot cultured cells, roots, and leaves confers enhanced salt tolerance.” Plant Physiol., 136, 2843–2854. Lal, S., Gulyani, V., and Khurana, P. (2008). “Overexpression of HVA1 gene from barley generates tolerance to salinity and water stress in transgenic mulberry (Morus indica).” Transgenic Res,, 17, 651–663. Laurie, S., Feeney, K. A., Maathuis, F. J., Heard, P. J., Brown, S. J., and Leigh, R. A. (2002). “A role for HKT1 in sodium uptake by wheat roots.” Plant J., 32, 139–149. Lee, S. C., Choi, H. W., Hwang, I. S., Choi, D., and Hwang, B. K. (2006). “Functional roles of the pepper pathogen-induced bZIP transcription factor, CAbZIP1, in enhanced resistance to pathogen infection and environmental stresses.” Planta, 224, 1209–1225. Lilius, G., Holmberg, N., and Bulow, L. (1996). “Enhanced NaCl stress tolerance in transgenic tobacco expressing bacterial choline dehydrogenase.” Biotechnology, 14, 177–180. Liu, K., Wang, L., Xu, Y., Chen, N., Ma, Q., Li, F., and Chong, K. (2007a). “Overexpression of OsCOIN, a putative cold inducible zinc finger protein, increased tolerance to chilling, salt and drought, and enhanced proline level in rice.” Planta, 226, 1007–1016. Liu, S., Cheng, Y., Zhang, X., Guan, Q., Nishiuchi, S., Hase, K., and Takano, T. (2007b). “Expression of an NADP-malic enzyme gene in rice (Oryza sativa L.) is induced by environmental stresses; overexpression of the gene in Arabidopsis confers salt and osmotic stress tolerance.” Plant Mol Biol., 64, 49–58. Luo, G. Z., Wang, H. W., Huang, J., Tian, A. G., Wang, Y. J., Zhang, J. S., and Chen, S. Y. (2005). “A putative plasma membrane cation/proton antiporter from soybean confers salt tolerance in Arabidopsis.” Plant Mol. Biol., 59, 809–820.

TRANSGENIC STRATEGIES

267

Lv, S., Zhang, K., Gao, Q., Lian, L., Song, Y., and Zhang, J. (2008). “Overexpression of an H-PPase gene from Thellungiella halophila in cotton enhances salt tolerance and improves growth and photosynthetic performance.” Plant Cell Physiol., 49, 1150–1164. Ma, X., Qian, Q., and Zhu, D. (2005). “Expression of a calcineurin gene improves salt stress tolerance in transgenic rice.” Plant Mol. Biol., 58, 483–495. Maathuis, F. J. M., and Amtmann, A. (1999). “K nutrition and Na toxicity: The basis of cellular K/Na ratios.” Ann. Bot., 84, 123–133. Maathuis, F. J. M., and Sanders, D. (1995). “Contrasting roles in ion transport of two K -channel types in root cells of Arabidopsis thaliana.“ Planta, 197, 456–464. ———. (1997). “Regulation of K absorption in plant root cells by external K: Interplay of different plasma membrane K transporters.” J. Exp. Bot., 48, 451–458. Mahalakshmi, S., Christopher, G. S., Reddy, T. P., Rao, K. V., and Reddy, V. D. (2006). “Isolation of a cDNA clone (PcSrp) encoding serine-rich-protein from Porteresia coarctata T., and its expression in yeast and finger millet (Eleusine coracana L.) affording salt tolerance.” Planta, 224, 347–359. Majee, M., Maitra, S., Dastidar, K. G., Pattnaik, S., Chatterjee, A., Hait, N. C., Das, K. P., and Majumder, A. L. (2004). “A novel salt-tolerant L-myo-inositol-1phosphate synthase from Porteresia coarctata (Roxb.) Tateoka, a halophytic wild rice: Molecular cloning, bacterial overexpression, characterization, and functional introgression into tobacco-conferring salt tolerance phenotype.” J. Biol. Chem., 279, 28539–28552. Mazel, A., Leshem, Y., Tiwari, B.S., and Levine, A. (2004). “Induction of salt and osmotic stress tolerance by overexpression of an intracellular vesicle trafficking protein AtRab7 (AtRabG3e).” Plant Physiol., 134, 118–128. McNeil, S. D., Nuccio, M. L., Ziemak, M. J., and Hanson, A. D. (2001). “Enhanced synthesis of choline and glycine betaine in transgenic tobacco plants that overexpress phosphoethanolamine N-methyltransferase.” Proc. Natl. Acad. Sci. USA 98, 10001–10005. Miranda, J. A., Avonce, N., Suarez, R., Thevelein, J. M., Van Dijck, P., and Iturriaga, G. (2007). “A bifunctional TPS-TPP enzyme from yeast confers tolerance to multiple and extreme abiotic-stress conditions in transgenic Arabidopsis.” Planta, 226, 1411–1421. Moghaieb, R. E., Tanaka, N., Saneoka, H., Hussein, H. A., Yousef, S., Ewada, M., Aly, M., and Fujita, K. (2008). “Expression of betaine aldehyde dehydrogenase gene in transgenic tomato hairy roots leads to the accumulation of glycine betaine and contributes to the maintenance of the osmotic potential under salt stress.” Soil Sci. Plant Nutr., 46, 873–883. Mohanty, A., Kathuria, H., Ferjani, A., Sakamoto, A., Mohanty, P., Murata, N., and Tyagi, A. K. (2002). “Transgenics of an elite indica rice variety Pusa basmati L. harbouring the codA gene are highly tolerant to salt stress.” Theor. Appl. Genet., 106, 51–57. Mukhopadhyay, A., Vij, S., and Tyagi, A. K. (2004). “Overexpression of a zincfinger protein gene from rice confers tolerance to cold, dehydration, and salt stress in transgenic tobacco.” Proc. Natl. Acad. Sci. USA, 101, 6309–6314.

268

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Murata, N., Ishizaki-Nishizawa, O., Higashi, S., Hayashi, H., Tasaka, Y., and Nishida, I. (1992). “Genetically engineered alteration in the chilling sensitivity of plants.” Nature, 356, 710–713. Nagaoka, S., and Takano, T. (2003). “Salt tolerance-related protein STO binds to a Myb transcription factor homologue and confers salt tolerance in Arabidopsis.” J. Exp. Bot., 54, 2231–2237. Nakashima, K., Tran, L. S., Van Nguyen, D., Fujita, M., Maruyama, K., Todaka, D., Ito, Y., Hayashi, N., Shinozaki, K., and Yamaguchi-Shinozaki, K. (2007). “Functional analysis of a NAC-type transcription factor OsNAC6 involved in abiotic and biotic stress-responsive gene expression in rice.” Plant J., 51, 617–630. Nakayama, H., Yoshida, K., Ono, H., Murooka, Y., and Shinmyo, A. (2000). “Ectoine, the compatible solute of Halomonas elongata, confers hyperosmotic tolerance in cultured tobacco cells.” Plant Physiol., 122, 1239–1247. Nanjo, T., Kobayashi, M., Yoshiba, Y., Kakubari, Y., Yamaguchi-Shinozaki, K., and Shinozaki, K. (1999). “Antisense suppression of proline degradation improves tolerance to freezing and salinity in Arabidopsis thaliana.” FEBS Lett., 461, 205–210. Nuccio, M. L., McNeil, S. D., Ziemak, M. J., Hanson, A. D., Jain, R. K., and Selvaraj, G. (2000). “Choline import into chloroplasts limits glycine betaine synthesis in tobacco: Analysis of plants engineered with a chloroplastic or a cytosolic pathway.” Metabol. Eng., 2, 300–311. Oh, S. J., Song, S. I., Kim, Y. S., Jang, H. J., Kim, S. Y., Kim, M., Kim, Y. K., Nahm, B. H., and Kim, J. K. (2005). “Arabidopsis CBF3/DREB1A and ABF3 in transgenic rice increased tolerance to abiotic stress without stunting growth.” Plant Physiol., 138, 341–351. Ohta, M., Hayashi, Y., Nakashima, A., Hamada, A., Tanaka, A., Nakamura, T., and Hayakawa, T. (2002). “Introduction of a Na/H antiporter gene from Atriplex gmelini confers salt tolerance to rice.” FEBS Lett., 532, 279–282. Ono, H., Sawads, K., Khunajakr, N., Tao, T., Yamamoto, M., Hiramoto, M., Shinmyo, A., Takano, M., and Murooka, Y. (1999). “Characterization of biosynthetic enzymes for ectoine as a compatible solute in a moderately halophilic eubacterium, Halomonas elongata.” J. Bacteriol., 181, 91–99. Oraby, H. F., Ransom, C. B., Kravchenko, A. N., and Sticklen, M. B. (2005). “Barley HVA1 gene confers salt tolerance in R3 transgenic oat.” Crop Sci., 45, 2218–2227. Owen, P. A., Nickell, C. D., Noel, G. R., Thomas, D. J., and Frey, K. (1994). “Registration of ’saline’ soybean.” Crop Sci., 34, 1689. Park, B., Liu, Z., Kanno, A., and Kameya, T. (2005a). “Genetic improvement of Chinese cabbage for salt and drought tolerance by constitutive expression of a B. napus LEA gene.” Plant Sci., 169, 553–558. ———. (2005b). “Increased tolerance to salt- and water-deficit stress in transgenic lettuce (Lactuca sativa L.) by constitutive expression of LEA.” Plant Growth Reg., 45, 165–171. Peng, Y., Lin, W., Cai, W., and Arora, R. (2007). “Overexpression of a Panax ginseng tonoplast aquaporin alters salt tolerance, drought tolerance and cold acclimation ability in transgenic Arabidopsis plants.” Planta, 226, 729–740. Prasad, K. V. S. K., Sharmila, P., Kumar, P. A., and Saradhi, P. P. (2000). “Transformation of Brassica juncea (L.). Czern with bacterial codA gene enhances its tolerance to salt stress.” Mol. Breed., 6, 489–499.

TRANSGENIC STRATEGIES

269

Prashanth, S. R., Sadhasivam, V., and Parida, A. (2008). “Overexpression of cytosolic copper/zinc superoxide dismutase from a mangrove plant Avicennia marina in indica rice var Pusa Basmati-1 confers abiotic stress tolerance.” Transgen. Res., 17, 281–291. Qiao, W. H., Zhao, X. Y., Li, W., Luo, Y., and Zhang, X. S. (2007). “Overexpression of AeNHX1, a root-specific vacuolar Na/H antiporter from Agropyron elongatum, confers salt tolerance to Arabidopsis and Festuca plants.” Plant Cell Rep., 26, 1663–1672. Quan, R., Lin, H., Mendoza, I., Zhang, Y., Cao, W., Yang, Y., Shang, M., Chen, S., Pardo, J. M., and Guo, Y. (2007). “SCABP8/CBL10, a putative calcium sensor, interacts with the protein kinase SOS2 to protect Arabidopsis shoots from salt stress.” Plant Cell, 19, 1415–1431. Rai, M., Pal, M., Sumesh, K. V., Jain, V., and Sankaranarayanan, A. (2006). Engineering for biosynthesis of ectoine (2-methyl 4-carboxy tetrahydro pyrimidine) in tobacco chloroplasts leads to accumulation of ectoine and enhanced salinity tolerance.” Plant Sci., 170, 291–306. Rathinasabapathi, B. (2000). “Metabolic engineering for stress tolerance: Installing osmoprotectant synthesis pathways.” Ann. Bot., 86, 709–716. Ren, Z. H., Gao, J. P., Li, L. G., Cai, X. L., Huang, W., Chao, D. Y., Zhu, M. Z., Wang, Z. Y., Luan, S., and Lin, H. X. (2005). “A rice quantitative trait locus for salt tolerance encodes a sodium transporter.” Nat. Genet., 37, 1141–1146. Roberts, S. K., and Tester, M. (1997). “Permeation of Ca2 and monovalent cations through an outwardly rectifying channel in maize root stelar cells.” J. Exp. Bot., 48, 839–846. Rohila, J. S., Kaur, R., and Wu, R. (2002). “Genetic improvement of Basmati rice for salt and drought tolerance by regulated expression of a barley Hva1 cDNA.” Plant Sci., 163, 525–532. Rontein, D., Basset, G., and Hanson, A. D. (2002). “Metabolic engineering of osmoprotectant accumulation in plants.” Metabol. Eng., 4, 49–56. Roosens, N. H., Willem, R., Li, Y., Verbruggen, I. I., Biesemans, M., and Jacobs, M. (1999). “Proline metabolism in the wild-type and in a salt-tolerant mutant of Nicotiana plumbaginifolia studied by (13)C-nuclear magnetic resonance imaging.” Plant Physiol., 121, 1281–1290. Roxas, V. P., Lodhi, S. A., Garrett, D. K., Mahan, J. R., and Allen, R. D. (2000). “Stress tolerance in transgenic tobacco seedlings that overexpresses glutathione S-transferase/glutathione peroxidase.” Plant Cell Rep. 2000, 41, 1229–1234. Roxas, V. P., Smith, R. K., Jr., Allen, E. R., and Allen, R. D. (1997). “Overexpression of glutathione S-transferase/glutathione peroxidase enhances the growth of transgenic tobacco seedlings during stress.” Nat. Biotechnol., 15, 988–991. Ruiz, J. M., and Blumwald, E. (2002). “Salinty-induced glutathione synthesis in Brassica napus.“ Planta, 214, 965–969. Rus, A., Yokoi, S., Sharkhuu, A., Reddy, M., Lee. B., Matsumoto, T. K., Koiwa, H., Zhu, J. K., Bressan, R. A., and Hasegawa, P. M. (2001). “AtHKT1 is a salt tolerance determinant that controls Na entry into plant roots.” Proc. Natl. Acad. Sci. USA, 98, 14150–14155. Sahi, C., Singh, A., Blumwald, E., and Grover, A. (2006). “Beyond osmolytes and transporters: Novel plant salt-stress tolerance-related genes from transcriptional profiling data.” Physiol. Plant., 127, 1–9.

270

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Saijo, Y., Hata, S., Kyozuka, J., Shimamoto, K., and Izui, K. (2000). “Over-expression of a single Ca-dependent protein kinase confers both cold and salt/ drought tolerance on rice plants.” Plant J., 23, 319–327. Sakamoto, A., and Murata, N. (2000). “Genetic engineering of glycinebetaine synthesis in plants: Current status and implications for enhancement of stress tolerance.” J. Exp. Bot., 51, 81–88. ———. (1998). “Metabolic engineering of rice leading to biosynthesis of glycinebetaine and tolerance to salt and cold.” Plant Mol. Biol., 38, 1011–1019. Sanan-Mishra, N., Pham, X. H., Sopory, S. K., and Tuteja, N. (2005). “Pea DNA helicase 45 overexpression in tobacco confers high salinity tolerance without affecting yield.” Proc. Natl. Acad. Sci. USA, 102, 509–514. Sentenac, H., Bonneaud, N., Minet, M., Lacroute, F., Salmon, J. M., Gaymard, F., and Grignon, C. (1992). “Cloning and expression in yeast of a plant potassium ion transport system.” Science, 256, 663–665. Shan, L., Li, C., Chen, F., Zhao, S., and Xia, G. (2008). “A Bowman-Birk type protease inhibitor is involved in the tolerance to salt stress in wheat.” Plant Cell Environ., 31, 1128–1137. Shen, Y. G., Du, B. X., Zhang, W. K., Zhang, J. S., and Chen, S. Y. (2002). “AhCMO, regulated by stresses in Atriplex hortensis, can improve drought tolerance in transgenic tobacco.” Theor. Appl. Genet., 105, 815–821. Sheveleva, E., Chmara, W., Bohnert, H. J., and Jensen, R. G. (1997). “Increased salt and drought tolerance by D-Ononitol production in transgenic Nicotiana tabacum L.” Plant Physiol., 115, 1211–1219. Shi, H., Ishitani, M., Kim, C., and Zhu, J. K. (2000). “The Arabidopsis thaliana salt tolerance gene SOS1 encodes a putative Na/H antiporter.” Proc. Natl. Acad. Sci. USA, 97, 6896–6901. Shi, H., Lee, B. H., Wu, S. J., and Zhu, J. K. (2003). “Overexpression of a plasma membrane Na/H antiporter gene improves salt tolerance in Arabidopsis thaliana.” Nat. Biotechnol., 21, 81–85. Shi, H., Xiong, L., Stevenson, B., Lu, T., and Zhu, J. K. (2002). “The Arabidopsis salt overly sensitive 4 mutants uncover a critical role for vitamin B6 in plant salt tolerance.” Plant Cell, 14, 575–588. Shukla, R. K., Raha, S., Tripathi, V., and Chattopadhyay, D. (2006). “Expression of CAP2, an APETALA2-family transcription factor from chickpea, enhances growth and tolerance to dehydration and salt stress in transgenic tobacco.” Plant Physiol., 142, 113–123. Singla-Pareek, S. L., Reddy, M. K., and Sopory, S. K. (2003). “Genetic engineering of the glyoxalase pathway in tobacco leads to enhanced salinity tolerance.” Proc. Natl. Acad. Sci. USA, 100, 14672–14677. Spalding, E. P., Hirsch, R. E., Lewis, D. R., Qi, Z., Sussman, M. R., and Lewis, B. D. (1999). “Potassium uptake supporting plant growth in the absence of AKT1 channel activity: Inhibition by ammonium and stimulation by sodium.” J. Gen. Physiol., 113, 909–918. Su, J., and Wu, R. (2004). “Stress-inducible synthesis of proline in transgenic rice confers faster growth under stress conditions than that with constitutive synthesis.” Plant Sci., 166, 941–948.

TRANSGENIC STRATEGIES

271

Su, J., Hirji, R., Zhang, L., He, C., Selvaraj, G., and Wu, R. (2006). “Evaluation of the stress-inducible production of choline oxidase in transgenic rice as a strategy for producing the stress-protectant glycine betaine.” J. Exp. Bot., 57, 1129–1135. Sulpice, R., Tsukaya, H., Nonaka, H., Mustardy, L., Chen, T. H., and Murata, N. (2003). “Enhanced formation of flowers in salt-stressed Arabidopsis after genetic engineering of the synthesis of glycine betaine.” Plant J., 36, 165–176. Sun, J., Jiang, H., Xu, Y., Li, H., Wu, X., Xie, Q., and Li, C. (2007). “The CCCH-type zinc finger proteins AtSZF1 and AtSZF2 regulate salt stress responses in Arabidopsis.” Plant Cell Physiol., 48, 1148–1158. Tanaka, Y., Hibin, T., Hayashi, Y., Tanaka, A., Kishitani, S., Takabe, T., Yokota, S., and Takabe, T. (1999). “Salt tolerance of transgenic rice overexpressing yeast mitochondrial Mn-SOD in chloroplasts.” Plant Sci., 148, 131–138. Tarczynski, M. C., Jensen, R. G., and Bohnert, H. J. (1993). “Stress protection of transgenic tobacco by production of the osmolyte mannitol.” Science, 259, 508–510. Thomas, J. C., Sepahi, M., Arendall, B., and Bohnert, H. J. (1995). “Enhancement of seed germination in high salinity by engineering mannitol expression in Arabidopsis thaliana.“ Plant Cell Environ., 18, 801–806. Tsenga, M. J., Liub, C., and Yiu, J. (2007). “Enhanced tolerance to sulfur dioxide and salt stress of transgenic Chinese cabbage plants expressing both superoxide dismutase and catalase in chloroplasts.” Plant Physiol. Biochem., 45, 822–833. Uddin, M. I., Qi Y., Yamada, S., Shibuya, I., Deng, X. P., Kwak, S. S., Kaminaka, H., and Tanaka, K. (2008). “Overexpression of a new rice vacuolar antiporter regulating protein OsARP improves salt tolerance in tobacco.” Plant Cell Physiol, 49, 880–890. Uozumi, N., Kim, E. J., Rubio, F., Yamaguchi, T., Muto, S., Tsuboi, A., Bakker, E. P., Nakamura, T., and Schroeder, J. I. (2000). “The Arabidopsis HKT1 gene homolog mediates inward Na currents in Xenopus laevis oocytes and Na uptake in Saccaharomyces cerevisiae.” Plant Physiol., 122, 1249–1259. Ushimaru, T., Nakagawa, T., Fujiko, Y., Daicho, K., Naito M., Yamauchi, Y., Nonaka, H., Amako, K., Yamawaki, K., and Murata, N. (2006). “Transgenic Arabidopsis plants expressing the rice dehydroascorbate reductase gene are resistant to salt stress.” J. Plant Physiol., 163, 1179–1184. Vaeck, M., Reynaerts, A., Hofte, H., Jansens, S., Beuckeleer, M. D., Dean, C., Zabeau, M., Montagu, M. V., and Leemans, J. (1987). “Transgenic plants protected from insect attack.” Nature, 328, 33–37. Verma, D., Singla-Pareek, S. L., Rajagopal, D., Reddy, M. K., and Sopory, S. K. (2007). “Functional validation of a novel isoform of Na/H antiporter from Pennisetum glaucum for enhancing salinity tolerance in rice.” J. Biosci., 32, 621–628. Waie, B., and Rajam, M. V. (2003). “Effect of increased polyamine biosynthesis on stress responses in transgenic tobacco by introduction of human S-adenosylmethionine gene.” Plant Sci., 164, 727–734. Wang, M., Gu, D., Liu, T., Wang, Z., Guo, X., Hou, W., Bai, Y., Chen, X., and Wang, G. (2007). “Overexpression of a putative maize calcineurin B-like protein in Arabidopsis confers salt tolerance.” Plant Mol. Biol., 65, 733–746.

272

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Wang, H., Huang, Z., Chen, Q., Zhang, Z., Zhang, H., Wu, Y., Huang, D., and Huang, R. (2004a). “Ectopic overexpression of tomato JERF3 in tobacco activates downstream gene expression and enhances salt tolerance.” Plant Mol. Biol., 55, 183–192. Wang, Y., Ying, Y., Chen, J., and Wang, C. (2004b). “Transgenic Arabidopsis overexpressing Mn-SOD enhanced salt-tolerance.” Plant Sci., 167, 671–677. Wegner, L. H., and Raschke, K. (1994). “Ion channels in the xylem parenchyma of barley roots (a procedure to isolate protoplasts from this tissue and a patchclamp exploration of salt passageways into xylem vessels).” Plant Physiol., 105, 799–813. Wu, C. A., Yang, G. D., Meng, Q. W., and Zheng, C. C. (2004). “The cotton GhNHX1 gene encoding a novel putative tonoplast Na/H antiporter plays an important role in salt stress.” Plant Cell Physiol., 45, 600–607. Wu, W., Su, Q., Xia, X., Wang, Y., and An, L. (2007). “The Suaeda liaotungensis kitag betaine aldehyde dehydrogenase gene improves salt tolerance of transgenic maize mediated with minimum linear length of DNA fragment.” Euphytica, 159, 17–25. Xu, D., Duan, X., Wang, B., Hong, B., Ho, T., and Wu, R. (1996). “Expression of a late embryogenesis abundant protein gene, HVA1, from barley confers tolerance to water deficit and salt stress in transgenic rice.” Plant Physiol., 110, 249–257. Xu, D. Q., Huang, J., Guo, S. Q., Yang, X., Bao, Y. M., Tang, H. J., and Zhang, H. S. (2008). “Overexpression of a TFIIIA-type zinc finger protein gene ZFP252 enhances drought and salt tolerance in rice (Oryza sativa L.).” FEBS Lett., 582, 1037–1043. Xu, W. F., and Shi, W. M. (2008). “Mechanisms of salt tolerance in transgenic Arabidopsis thaliana constitutively overexpressing the tomato 14-3-3 protein TFT7.” Plant Soil, 301, 17–28. Xue, Z., Zhi, D., Xue, G., Zhang, H., Zhao, Y., and Xi, G. (2007). “Enhanced salt tolerance of transgenic progeny of tall fescue (Festuca arundinacea) expressing a vacuolar Na/H antiporter gene from Arabidopsis.” Plant Sci., 167, 849–859. ———. (2004). “Enhanced salt tolerance of transgenic wheat (Tritivum aestivum L.) expressing a vacuolar Na/H antiporter gene with improved grain yields in saline soils in the field and a reduced level of leaf Na.” Plant Sci., 167, 849–859. Yancey, P. H., Clark, M. E., Hand, S. C., Bowlus, R. D., and Somero, G. N. (1982). “Living with water stress: Evolution of osmolyte systems.” Science, 217, 1214–1222. Yang, C. Y., Chen, Y. C., Jauh, G. Y., and Wang, C. S. (2005). “A lily ASR protein involves abscisic acid signaling and confers drought and salt resistance in Arabidopsis.” Plant Physiol., 139, 836–846. Yang, H., Knapp, J., Koirala, P., Rajagopal, D., Peer, W. A., Silbart, L. K., Murphy, A., and Gaxiola, R. A. (2007). “Enhanced phosphorus nutrition in monocots and dicots overexpressing a phosphorus-responsive type I H-pyrophosphatase.” Plant Biotechnol. J., 5, 735–745. Yang, X., Liang, Z., Wen, X., and Lu, C. (2008). “Genetic engineering of the biosynthesis of glycinebetaine leads to increased tolerance of photosynthesis to salt stress in transgenic tobacco plants.” Plant Mol. Biol., 66, 73–86.

TRANSGENIC STRATEGIES

273

Yeo, E. T., Kwon, H. B., Han, S. E., Lee, J. T., Ryu, J. C., and Byu, M. O. (2000). “Genetic engineering of drought resistant potato plants by introduction of the trehalose-6-phosphate synthase (TPS1) gene from Saccharomyces cerevisiae.” Mol. Cells, 10, 263–268. Yin, X. Y., Yang, A. F., Zhang, K. W., and Zhang, J. R. (2004). “Production and analysis of transgenic maize with improved salt tolerance by the introduction of AtNHX1 gene.” Acta Bot. Sin., 46, 854–861. Yonamine, I., Yoshida, K., Kido, K., Nakagawa, A., Nakayama, H., and Shinmyo, A. (2004). “Overexpression of NtHAL3 genes confers increased levels of proline biosynthesis and the enhancement of salt tolerance in cultured tobacco cells.” J. Exp. Bot., 55, 387–395. Zerial, M., and McBride, H. (2001) “Rab proteins as membrane organizers” Nat. Rev, Mol. Cell. Biol. 2, 107–17. Zhang, H., Huang, Z., Xie, B., Chen, Q., Tian, X., Zhang, X., Zhang, H., Lu, X., Huang, D., and Huang, R. (2004). “The ethylene-, jasmonate-, abscisic acidand NaCl-responsive tomato transcription factor JERF1 modulates expression of GCC box-containing genes and salt tolerance in tobacco.” Planta, 220, 262–270. Zhang, H. X., and Blumwald, E. (2001). “Transgenic salt-tolerant tomato plants accumulate salt in foliage but not in fruit.” Nat. Biotechnol., 19, 765–768. Zhang, H. X., Hodson, J. N., Williams, J. P., and Blumwald, E. (2001). “Engineering salt-tolerant brassica plants: Characterization of yield and seed oil quality in transgenic plants with increased vacuolar sodium accumulation.” Proc. Natl. Acad. Sci. USA, 98, 12832–12836. Zhang, J. Z., Creelman, R. A., and Zhu, J. K. (2004). “From laboratory to field: Using information from Arabidopsis to engineer salt, cold and drought tolerance in crops.” Plant Physiol., 135, 615–621. Zhang, Y., Yang, J., Lu, C., Cai, J., and Guo, Z. (2008). “Overexpressing SgNCED1 in tobacco increases ABA level, antioxidant enzyme activities, and stress tolerance.” J. Plant Growth Regul., 27, 151–158. Zhao, F., and Zhang, H. (2006). “Expression of Suaeda salsa glutathione S-transferase in transgenic rice resulted in a different level of abiotic stress resistance.” J. Agric. Sci., 144, 258–265. Zhao, F. Guc, S., Zhang, H., and Zhao, Y. (2006a). “Expression of yeast SOD2 in transgenic rice results in increased salt tolerance.” Plant Sci., 170, 216–224. Zhao, F., Wang, Z., Zhang, Q., Zhao, Y., and Zhang, H. (2006b). “Analysis of the physiological mechanism of salt-tolerant transgenic rice carrying a vacuolar Na/H antiporter gene from Suaeda salsa.” J. Plant Res., 119, 95–104. Zhao, J., Zhi, D., Xue, Z., Liu, H., and Xi, G. (2007). “Enhanced salt tolerance of transgenic progeny of tall fescue (Festuca arundinacea): Expressing a vacuolar Na/H antiporter gene from Arabidopsis.” J. Plant Physiol., 164, 1377–1383. Zhifang, G., and Loescher, W. H. (2003). “Expression of a celery mannose 6-phosphate reductase in Arabidopsis thaliana enhances salt tolerance and induces biosynthesis of both mannitol and a glucosyl-mannitol dimmer.” Plant Cell Env., 26, 275–283. Zhou, S., Chen, X., Zhang, X., and Li, Y. (2008). “Improved salt tolerance in tobacco plants by co-transformation of a betaine synthesis gene BADH and a vacuolar Na/H antiporter gene SeNHX1.” Biotechnol. Lett., 30, 369–376.

274

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Zhu, B., Su, J., Chang, M., Verma, D. P. S., Fan, Y-L., and Wu, R. (1998). “Overexpression of a 1-Pyrroline-5-carboxylate synthetase gene and analysis of tolerance to water- and salt-stress in transgenic rice.” Plant Sci., 139, 41–48. Zhu, J.-K. (2001). “Cell signaling under salt, water and cold stress.” Cur. Op. Plant Biol., 4, 401–406. ———. (2002). “Salt and drought stress signal transduction in plants.” Ann. Rev. Plant Biol., 53, 247–273.

PART THREE: SAMPLING, MONITORING, AND MEASUREMENT

CHAPTER 9 FIELD SAMPLING OF SOIL, WATER, AND PLANTS Blaine R. Hanson and Stephen R. Grattan

INTRODUCTION Diagnostic information is essential for assessing, correcting, or preventing salinization of agricultural land, a worldwide problem in arid climates that affects our food and fiber production. This information can be obtained by sampling soil, water, and plants. Objectives of sampling should include: • • • • •

Selection of parameters Identification of problem areas Description of the severity of the problem Monitoring of time-dependent changes in salinity Assessing effects of different management strategies on salinity and crop yield.

To develop a sampling strategy that will achieve these objectives, the following questions must be answered: • • • • •

What is the region or time period to be sampled? What types of measurements are needed? What sampling technique is appropriate for the particular problem? How accurate must estimates of the population’s parameters be? How many observations are needed in the sample to obtain the desired accuracy? • Should the sample be distributed spatially or temporally?

275

276

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

A sampling strategy involves collecting a set of observations, called a sample, from a population, which is the total of all possible observations. The parameters or characteristics of the population are estimated by statistics—values calculated from the sample. Inferences about characteristics are made from statistics. Examples of statistics are the mean or average value, standard deviation, range, and mode. Two characteristics of an estimator or statistic of a population parameter are its accuracy—a measurement of how close the estimate is to the true value of the population parameter—and its precision—the magnitude of the deviations from a mean obtained from repeated sampling. Due to bias, an estimator can be inaccurate. Such bias can be caused by improperly calibrated instruments, improper analysis procedures, improper sample selection, or improper methods of estimation. Sample selection can cause bias by deliberate selection of a sample (Yates 1981), selecting samples based on a characteristic correlated with the properties of the sampling area of interest, conscious or unconscious bias in the selection of a sample, substitution of one sampling location with another, and failure to cover all of the population. Proper procedures of analysis and estimation and proper methods of sample selection, such as random or systematic methods of selection, must be used to avoid bias. In this chapter we explore sampling techniques, the considerations in selecting a sampling approach, and the basics of sampling for soils, soilwater content, infiltration rates, water quality, and plant condition. SAMPLING TECHNIQUES A number of techniques are used for field sampling (Cochran 1977; Peterson and Calvin 1967; Snedecor and Cochran 1980; Yamane 1967; Yates 1981). Some are described as follows. Judgment Sampling Sampling locations may be selected based on the judgment of the sampler (Peterson and Calvin 1967). They may be selected according to such factors as crop growth, visual appearance of soil or plants, and locations of subsurface drains, ditches, and canals. Judgment sampling inherently involves a preconception about what needs to be sampled and where. For example, samples are frequently taken in a problem area to compare with other samples from a nonaffected area. Samples are often composited to reduce analysis costs. Simple Random Sampling Simple random sampling involves randomly selecting observations on the assumption that each sample has an equal chance of being

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

277

FIGURE 9-1. Examples of different sampling techniques. selected (Fig. 9-1). On a spatial level, a field is divided into various numbered sections and numbers are randomly selected to determine sampling locations. Peterson and Calvin (1967) suggest establishing a grid or transect throughout a field or plot and then using a table of random numbers to select coordinates for the locations of the observations. Random sampling can also be applied at a temporal level, such as sampling irrigation water quality at randomly selected times to detect changes in source water quality. Simple random sampling is appropriate for homogeneous populations. The estimators used for simple random sampling are: sample mean: X

∑ Xi N

(9-1)

standard deviation: ⎡ ∑(X i  X )2 ⎤ s ⎢ ⎥ ⎣ N 1 ⎦

1/2

⎡ N ∑ X i2  (∑ X i )2 ⎤ S ⎢ ⎥ N ( N  1) ⎦ ⎣

1/2

(9-2)

standard error of the mean: sX 

s N

(9-3)

278

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

confidence intervals: X ± ( z)sX

(9-4)

and coefficient of variation: CV 

100 s X

(9-5)

Stratified Random Sampling This technique, appropriate for areas with heterogeneous populations, consists of subdividing the area of interest into relatively uniform subareas or strata (Fig. 9-1). Each stratum is then randomly sampled. Strata need not be of equal size, but they must not overlap. Selection of strata might be based on differences in crop appearance, soil texture, salt accumulation on the soil surface, or management. The number of observations per stratum should be proportional to the percentage of the total area represented by a stratum. Stratified random sampling can be more accurate than simple random sampling for heterogeneous populations. Since stratified sampling eliminates sampling error between strata, the only sampling error comes from within strata. Miyamoto and Cruz (1986) found that a stratified sample based on soil mapping units in a field greatly reduced the sampling error. Their analysis showed the average coefficient of variation for field-wide transects to be 63%, whereas the average coefficient of variation was 36% for the stratified sample. Soil type accounted for 73% of the variability in the soil salinity. However, for homogeneous populations, stratified sampling may be no more accurate than simple random sampling. Estimators used for stratified sampling include: mean: X  ∑ Wh X h

(9-6)

⎛ ∑ Wh2 sh2 ⎞ sX  ⎜ ⎟ ⎝ Nh ⎠

(9-7)

standard error: 1/2

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

279

Systematic Sampling This technique consists of randomly selecting a starting point and then sampling the total area at equal intervals along a grid or a transect (Fig. 9-1). An advantage of this technique is that the area of interest is completely covered, whereas some parts of the field may not be sampled in simple random sampling. However, if a periodicity about equal to the sampling interval exists, a considerable loss of accuracy occurs (Peterson and Calvin 1967). No reliable way to estimate the standard error of the sample mean exists with systematic sampling, since one observation per interval is normally obtained (Snedecor and Cochran 1980). Peterson and Calvin (1967) and Yates (1981) recommend stratifying the systematic sample with at least two samples per block, assuming that the variation within blocks is the only sampling variation, and estimating the sampling error similarly to a stratified random sample. They also suggest obtaining several separate systematic samples from randomly selected starting points, and then calculating the sample mean for each sample. These means are treated as the data from a simple random sample of size equal to the number of separate systematic samples.

CONSIDERATIONS IN DEVELOPING SAMPLING STRATEGIES Selecting a Sampling Technique Selecting a sampling technique depends on the objectives and the sources of variance. For salinity problems in relatively small areas, judgment sampling coupled with compositing, or combining individual samples into one sample, might be used. Any bias or errors in estimates are likely to be inconsequential. If, however, field-wide salinity needs to be assessed, an unbiased and accurate sampling technique may be required, particularly if monitoring salinity over time is an objective. For this objective, sampling errors from spatial variability of salinity must be minimal to detect differences in salinity. Common sources of variance are trend, discontinuities, periodicity, and randomness. Trends in soil salinity may result from varying soil texture or varying amounts of infiltrated water in a field. If trend is a major source of variability, stratified random or systematic sampling may be best. Stratified random sampling should be used in areas with discontinuities or relatively abrupt changes in salinity. If random behavior is the major source of variability and the field-wide population appears homogeneous, then simple random sampling would be satisfactory. Periodicity in soil properties normally is created by human activity, but it can occur naturally (Webster 1977). If periodicity is substantial, the

280

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

sampling strategy must account for this source of variance. Sampling in blocks of observations, with the size of a block equal to the period of the frequency, may be needed for relatively high frequencies or small periods. The blocks of observations could be randomly or systematically distributed over the field, depending on the desired sampling technique. Since any periodicity is most likely created by human activity, the period or size of blocks might be estimated from information on cultural practices. For example, periodicity in the advance of water across a field could be caused by different rates of infiltration for wheel and nonwheel furrows. Estimates of the mean and standard deviation of advance times might be obtained by sampling in blocks equal to the number of furrows cultivated per pass of the cultivator. The number of wheel and nonwheel furrows sampled in a block should be proportional to the number of wheel and nonwheel furrows in a block. Spatial Distributions While population parameters (such as the mean of all data points) are important, equally important is the spatial distribution of salinity. According to Cameron et al. (1971), use of the mean value may be inappropriate for recommendations, since salinity and toxic elements vary spatially. If distributions are known, adequate treatments can be applied in areas with the more severe problems, and the expense of excessive treatment can be avoided in other, less-affected areas. Number of Samples The number of observations needed depends on the variability of the population and the accuracy and confidence level desired in the estimate. A statistical approach for estimating a sample size is (Wadpole and Myers 1978) ⎡ ( z)( s) ⎤ ⎡ ( z)(CV ) ⎤ N ⎢  ⎣ e ⎥⎦ ⎢⎣ k ⎥⎦ 2

2

(9-8)

where s  standard deviation; CV  coefficient of variation; e  absolute allowable error between the estimate of sample mean and the true value; k  error expressed as a decimal fraction; and z  multiplier equal to 1.96, 1.64, or 1.28 for confidence levels of 95%, 90%, and 80%. — The confidence level is the percent of time that the interval of X e contains the true value. For confidence levels of 95%, 90%, and 80%, z  1.96, 1.64, and 1.28, respectively. Cameron et al. (1971) suggest that e  0.1 to 0.2 of the sample mean and an 80% confidence level is sufficient for soil

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

281

recommendations. Thus, the true mean would be expected to be between — — X 0.15 X for 80% of the time for a sample of size N. Estimating the sample size requires an estimate of the variance or the standard deviation, usually unknown before sampling. However, an analysis of data sets on the variability of soil salinity (Biggar and Nielsen 1985; Hanson and Oster 1986; Miyamoto and Cruz 1987) suggests that a CV  35% to 45% might first be used in sampling. Subsequent sample sizes could then be adjusted as better estimates of the standard deviation of a specific field are obtained. For example: —

Calculate the number of samples required so that the interval X

— 0.15 X contains the true mean value for 80% of the time with repeated sampling. Assume the coefficient of variation is 40%. ⎡ (1.28)(0.40) ⎤ N ⎢ mples ⎥⎦  11.6, or collect 12 sam ⎣ 0.15 2

Once the number of observations has been determined, the sampling interval can be determined by systematically spacing the observations across the region of interest, or by randomly locating sampling locations (simple random and stratified random sampling). A minimum interval must be maintained to satisfy independence of observations. Independence of Observations The aforementioned sampling techniques assume independence of samples, that is, the probability of one value occurring does not depend on the probability of another value. This assumption implies that each observation provides completely new information about the population parameter. The assumption may be invalid for a given sampling interval due to spatial variability. Observations made relatively close together are more likely to be similar than those made relatively far apart. Thus, over some distance, a soil property may be autocorrelated, and observations taken within that distance are dependent (i.e., each observation does not provide entirely new information about the population parameter). The distance over which dependence occurs, called the range, is determined with variograms or correlograms (see Chapter 15; Clark 1979; Shumway 1988). Dependence exists if observations are autocorrelated. Autocorrelation generally decreases with distance between observations and the variance between observations increases. At distances where no autocorrelation exists, the observations are independent and the variance is maximal.

282

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The sampling interval must satisfy sample independence if classical statistics will be used, such as in simple random sampling. Sampling at intervals where autocorrelation exists will not improve the estimates of the population parameters but will result in unnecessary sampling. What is the minimum sampling interval for a given area or field? Most likely, a correlogram or variogram for a particular field will not exist. However, studies of field-wide spatial variability of soil salinity (Gajem et al. 1981; Hajrasuliha et al. 1980; Hanson and Oster 1986; Miyamoto and Cruz 1987) revealed that the range was often less than sampling intervals of 4 to 10 m. Thus, the assumption of independence is likely valid for intervals that are practical in terms of cost and time (i.e., intervals of more than 10 m). If a variogram or correlogram is known or can be developed for an area, the standard error can be estimated with fewer observations with a technique called kriging (Clark 1979). McBratney and Webster (1983) reported that estimating the standard error with kriging may require as few as about one-tenth the number of observations as with the classical approach. According to Gutjahr (1984), at least 50 observations are needed for a reliable variogram. Volume–Variance Relationships Also to be considered in sampling is the relationship between the volume sampled per observation and the variance or the coefficient of variation. El-Araby et al. (1987), Hassan et al. (1983), Hawley et al. (1982), and Zobeck et al. (1985) have shown that as the volume sampled per observation increases, the variance of the sample decreases. Hanson and Oster (1986) found a coefficient of variation of about 45% when sampling with a four-electrode salinity probe (volume sampled about 90 cm3) and a CV of about 35% when sampling with an electromagnetic conductivity meter (volume sampled about 1 cm3). These findings imply that the larger the volume sampled per observation (and thus, the smaller the variance), the fewer the number of required observations. For the CV of the four-electrode probe data, 14 observations are necessary. For the electromagnetic conductivity meter, seven are necessary. Variance: Size of Area Sampled The number of observations required might be expected to increase as the area sampled increases. However, as discussed previously, the number of observations depends on the variance or coefficient of variation. The required number of observations increases with the size of the area sampled only if the CV increases. Beckett and Webster (1971) concluded that soil variability, as measured by the CV, increases with the size of the area sampled. However, as much

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

283

as half of the CV present within one ha is already present within a few m2. Their analysis of data sets showed that relationships between CV and size of area sampled vary greatly. This probably reflects changes in soil properties in the size of the area sampled. The effect of the area’s size on variability depends on the spatial distribution of the soil property. The data in Fig. 9-2a show an increase in the CV with distance. The data in Figs. 9-2c and 9-2d show primarily random behavior, or little change in the CV with distance. Local Variability Once field-wide locations have been selected, further care is needed in sampling when substantial small-scale variability in soil salinity exists due to patterns of water flow unique to furrow and drip/trickle irrigation systems (Hoffman et al. 1985; Wadleigh and Fireman 1948). Sampling at a particular location must account for this variability. Sampling in the furrow bed would be appropriate for analysis of salinity for achieving germination and stand establishment but inappropriate for

FIGURE 9-2. Field-wide salinity along transects for: (a) Site DU; (b) Site FR; (c) Site CR; and (d) Site BO.

284

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

analyzing seasonal crop yield–salinity relationships. For drip/trickle systems, sampling midway between emitters would result in different salinity levels than sampling near the emitters. When monitoring timedependent salinity levels, each observation must be made at the same relative position to the pattern of water flow. Composite Sampling Frequently, samples are composited to reduce the cost of sample analysis. Cline (1944) states that composite sampling is valid if: 1. The sampling volume represents a homogeneous population. For heterogeneous populations, knowledge of spatial distributions is important, which a composite sample would fail to provide. 2. Equal amounts of each sampling unit contribute to the composite sample. 3. No interactions occur within the composite sample as a result of the compositing. 4. An unbiased estimate of the mean is the only objective. Composite sampling is unsatisfactory if any other statistic is required.

EVALUATIONS OF SAMPLING STRATEGIES Where little information is known before sampling, which sampling technique is most appropriate? To address this matter, we compared the results of the four sampling strategies, using data from four transects of soil salinity (Fig. 9-2a–d). Each transect data set had a different spatial distribution of salinity. Strategies were developed under the assumption that little information exists on field-wide variance and frequency distributions. Sampling techniques evaluated include simple random sampling, stratified random sampling, and systematic sampling (analyzed with stratified random sampling methods). A sample size of 15 observations was determined, using a confidence level of 80%, an error of 15%, and an assumed CV of 45%. This sample size was used for all strategies. Four strata were arbitrarily used for stratified random sampling. Figures 9-2a–d show the data sets. The primary source of variability at Site DU (Fig. 9-2a) was a gradual trend in salinity across the field. No surface indications of differences in soil texture or salinity were observed. At Site FR (Fig. 9-2b), a discontinuity in soil salinity appeared between 85 m and 90 m along the transect. Visible differences in soil texture were observed between areas of low and high salinity. Soil salinity appeared to be somewhat random at Site CR (Fig. 9-2c), but three large spikes

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

285

occurred at about 100 m, 150 m, and 250 m along the transect. Crop growth was poor at those locations. Figure 9-2d shows a relatively uniform distribution of salinity at Site BO. Random behavior appears to be the major source of variability. Results of this analysis (Table 9-1) revealed that the standard error for Sites BO and CR—where random behavior is the major source of variability—was about the same for all sampling techniques. The standard error where other sources of variability contributed substantially was less for the stratified random and systematic sampling techniques. This suggests that stratified random or systematic sampling should be used when little information is known about the salinity distribution before sampling. Equation 9-8 and an assumed CV were used to estimate the number of observations. For Sites CR and FR, judgment sampling was used. Areas with relatively high salinity levels experienced poor plant growth. Thus, one strategy would be to sample only those areas with poor growth to determine initial salinity levels and monitor changes in salinity with time. At Site CR, the affected areas were small. Hence, a few composited samples might be sufficient. At Site FR, the affected area was about half the total area sampled. A stratified random or systematic sampling method might be more appropriate (Table 9-1).

SAMPLING PARAMETERS Soil Soil chemistry problems are frequently identified by measurements of soil salinity, soluble salts, and toxic materials, such as boron. Other types of measurements that help to develop management strategies for salinity TABLE 9-1. Standard Error of the Mean for Several Sampling Strategies as Applied to Example Data for Four Transects of Soil Salinity Site (1)

Simple Random (2)

Stratified Random (3)

Systematic (4)

DU BO CR FR All data

0.18 0.04 0.09

0.11 0.03 0.10

0.14 0.03 0.09

0.94

0.53

0.49

Saline area

0.69

0.44

0.55

286

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

control include water content, water-holding capacity of the soil, soil texture, soil infiltration rate, hydraulic conductivity, and nutrients and micronutrients. Chapter 10 of this manual and Black (1965), Chapman and Pratt (1982), and Reisenauer (1978) cover methods for measuring the various parameters. Some considerations and recent developments in measuring some of these parameters will be discussed here. Soil Salinity Several devices that allow rapid field-wide sampling of soil salinity have been developed. They include the four-electrode salinity probe (Rhoades and Halvorson 1977; Rhoades and van Schilfgaarde 1976); the Wenner array (Rhoades and Halvorson 1977; Rhoades and Ingvalson 1971), and the electromagnetic conductivity meter (Corwin and Rhoades 1984; Hendrickx et al., 2002; Rhoades and Corwin 1981). All three instruments measure the bulk (or apparent) electrical conductivity (EC) of the soil. The electromagnetic conductivity meter can be attached to a tractor or other motorized device and used in conjunction GPS systems and appropriate computer software to generate field-scale soil salinity maps (see Chapter 10 and Chapter 14 for the prediction of field scale, and spatial salinity patterns from soil conductivity survey data). The four-electrode salinity probe is a probe of small diameter with four electrodes spaced a few centimeters apart along its length of about 150 mm (6 in.). Two of the electrodes create an electric field; the other two measure the electrical resistance of the soil. To use it, the probe is inserted into the soil to the desired depth. The volume sampled is about 90 cm3. The Wenner array consists of placing along the soil surface four equally spaced electrodes, each inserted to a depth of about 40 mm to 50 mm. The depth sampled roughly equals the electrode spacing. The electromagnetic conductivity meter consists of a transmitter and a receiver coil. The instrument is placed on or held above the surface of the soil. Electrical current introduced into the transmitter coil creates a secondary magnetic field in the soil through magnetic induction. The receiver coil detects the secondary field, the strength of which depends on the electrical conductivity of the soil. Theoretically, the bulk EC depends on the salinity of the soil solution, soil-water content, porosity, and type and amount of clay in the soil (McNeill 1980). While laboratory studies have shown the bulk EC to depend strongly on water content (Rhoades et al. 1976), field studies have shown little correlation between water content and bulk EC (Hanson and Oster 1986). Calibration curves for a given soil texture are needed to relate the bulk EC to the EC of a saturated extract (Halvorson et al. 1977). Studies on field-wide sampling (Cameron et al. 1981; de Jong

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

287

et al. 1979; Hanson and Oster 1976; William and Baker 1982) have shown the following results: 1. The four-electrode salinity probe provides good depth resolution, describing the salinity distribution with depth at a particular location. However, it may be difficult to insert this probe into the soil, even with a pilot hole. 2. The Wenner array provides satisfactory depth resolution over a large volume of soil, using the method of Rhoades and Halvorson (1977). However, the instrument is bulky and requires at least two persons for rapid surveying. Another drawback is that the drying of the soil surface can result in poor electrical continuity between the electrodes and the soil. Due to this problem, Hanson and Oster (1986) found the same coefficient of variation for the Wenner array and the four-electrode salinity probe, although the array sampled a much larger volume of soil. 3. The electromagnetic conductivity meter (EM-38) provides a way to survey field-wide salinity patterns rapidly. Its effective depth is about 1 m. Deeper sampling can be conducted with other instruments, such as the EM-31. The depth resolution of the instrument is poor, although some information about changes in salinity with depth can be obtained by operating the instrument in both the vertical mode (coils in vertical position) and horizontal mode (coils in horizontal position). (The EM-38 and EM-31 are manufactured by Geonics, Ltd., Mississauga, Ontario, Canada. Mention of these instruments is not an endorsement.) Gravimetric Sampling For gravimetric sampling of soil-water content, Hawley et al. (1982) found relatively large coefficients of variation for volumes of soil less than about 50 to 60 cm3 per observation, and little change in the CV for volumes from 50 to about 800 cm3. These results suggest that the volume sampled per observation should be about 50 to 100 cm3 for water content estimates. Larger samples would not improve the estimate of water content. For soil salinity, El-Araby et al. (1987) recommend a sample size of about 500 g per observation, based on volume-variance data. Infiltration Rates Traditional methods of measuring soil infiltration rates, such as ring infiltrometers, blocked-furrow infiltrometers, and inflow and outflow measurements, usually sample a particularly small area of a field. They can be highly unreliable due to spatial variability. Elliott and Walker (1980) developed a way to calculate the infiltration rate for furrow or border irrigation systems. Data needed for this method include water advance

288

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

measured at two locations along the length of the field, border or furrow inflow rate, field slope, run length, surface roughness, and furrow geometry. The advantage of this method is that the entire furrow or border is used as the infiltrometer, thus reducing the variance of infiltration measurements. Hanson et al. (1988), however, found infiltration rates calculated from measurements made in a single furrow to be highly unreliable for cracked soils. Their study showed that measurements must be made in blocks of furrows for these soils. Water Measurements made on water samples include salinity, soluble salts, toxic elements, and pH. Other useful measurements for management strategies include flow rates, volumes, and application rates. While strategies for water sampling should be based on the previously discussed techniques, some modifications may be required. Observations could be distributed systematically or randomly. The required number of observations can be calculated from Eq. 9-8 if an estimate of the variance is available. Events responsible for changes in water quality must also be considered in determining sample size and sampling interval. The quality of irrigation water may change slowly during an irrigation season. Thus, infrequent sampling at random or systematic intervals may be sufficient. The quality of subsurface drainage water may vary considerably with irrigation events. Thus, water measurements may be needed just before and just after an irrigation and then at relatively frequent intervals shortly after the irrigation. Plants The most important plant parameters to monitor in a saline environment are growth, visual appearance, element concentrations in tissues, and crop quality. All these parameters vary in time and space. Variability can be at the field-wide scale and the local scale (i.e., within a plant). As indicated in Chapter 6, plants are affected by salinity by both osmotic effects and specific-ion effects. The former results in plant growth reduction, and plants suffering from salinity in the field may show differential growth patterns associated with differential soil salinity patterns. Therefore, plant sampling for growth differences may be assessed by stratified random sampling. Plants affected by salinity may also produce visual injury to the plant. The most notorious elements that produce crop injury are chloride (Cl), boron (B), and sodium (Na) (Maas and Grattan 1999). These elements are usually distributed spatially in predictable patterns within the plant. For example, Cl, and to some degree Na, are often found in highest concentrations in the margins of older leaves. Symptoms occur primarily on tree

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

289

and vine crops and include chlorosis (yellowing of leaves) and necrosis (leaf burn). Symptoms appear first on older tissue and are often absent on younger, developing tissue. Injury symptoms are often correlated with tissue ion concentration, but development of injury may be related to climate (Maas and Grattan 1999). Consequently, segregation of leaves that vary in physiological age could reduce the variance and provide a better indicator of chloride damage. Therefore, dividing the plant into strata based on leaf age may be desirable. This suggests a stratified random sampling approach. For practical purposes, judgment sampling can be used to select leaves for the sample within a stratum. Boron is also an element that can cause injury to plants although its concentration in the tissue is a poor indicator of toxicity (Nable et al. 1997). Boron is an essential element for plants but it has a small concentration range between what is considered optimal and that which is considered toxic (Maas and Grattan 1999). Boron distribution in the plant is very much dependent on its mobility, which is highly crop-specific (Brown and Shelp 1997). In most plants, B is relatively immobile and injury patterns will be very much characterized as those for Na and Cl. Examples of B-immobile plants include pistachio, walnut, tomato, and cotton. However in others, B can be remobilized in the plant to younger tissue and toxic effects can be seen as tip die back and fruit disorders appear (Nable et al. 1997). Examples of B-mobile plants include almond, peach, apple, beans, and onion. Sampling plants as an indicator of B toxicity, therefore, will be dependent upon whether the plant is B-mobile or B-immobile. Sampling older leaves will be a better indicator in B-immobile plants, whereas sampling young developing tissue would be more appropriate in B-mobile plants.

SUMMARY A variety of relatively simple statistical sampling methods are available for evaluating soil-water and other field conditions. Where little is known about the conditions to be studied, stratified random sampling and systematic sampling are recommended. Where there is a basis for doing so, a focused judgment sampling may be appropriate, for example, when there are clear visual indications that one site has poor crop growth and sampling can be focused on the problem area. There are relatively standard data collection protocols for sampling soils, soil-water, infiltration rates, and plants: • Soil electrical conductivity can be sampled with four-electrode salinity probes and/or soil surface arrays (Wenner arrays), or electromagnetic conductivity meters.

290

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

• Gravimetric sampling for soil-water content requires soil samples to be 50 to 100 cm3 each, and sampling for soil salinity analysis samples should be 500 g each. • For furrow or border irrigation in soils that are not cracked, infiltration rates may be measured based on water advance at two locations along the length of the field or border. • Sampling for water quality (salinity, pH, trace minerals, etc.) may involve systematic or random sampling, with temporal changes in irrigation water sampled randomly and subsurface drainage sampled just before and just after irrigation events. • Sampling to diagnose potential salinity effects on plants generally requires a stratified random sampling based on known symptoms of salinity-induced stress (such as chlorosis and leaf burn in older leaves). • Sampling of plants for B toxicity depends on whether the plant is classified as B-immobile or B-mobile.

REFERENCES Beckett, P. H. T., and Webster, R. (1971). “Soil variability: A review.” Soils and Fert., 34, 1–15. Biggar, J. W., and Nielson, D. R. (1985). “Spatial analysis of water table salinity and hydraulic conductivity prior to salinization.” Proc., Int. Symp. on the Reclamation of Salt-Affected Soils, Part 1. May 13–21, 1985, Jinan, China. Black, C. A., ed. (1965). “Methods of soil analysis, Part 1.” American Society of Agronomy Monograph 9, ASA, Madison, Wisc. Brown, P. H., and Shelp, B. J. (1997). “Boron mobility in plants.” Plant Soil, 193, 85–101. Cameron, D. R., de Jong, E., Read, D. W. L., and Oosterveld, M. (1981). “Mapping salinity using resistivity and electromagnetic inductive techniques.” Can. J. Soil Sci., 61, 67–78. Cameron, D. R., Nyborg, M., Toogood, J. A., and Laverty, D. H. (1971). “Accuracy of field sampling for soil tests.” Can. J. Soil. Sci., 51, 165–175. Chapman, H. D., and Pratt, P. F. (1982). Methods of analysis for soils, plants, and waters, Publication 4034, University of California, Riverside, Calif. Clark, I. (1979). Practical geostatistics, Applied Science Publishers Ltd., London. Cline, M. G. (1944). “Principles of soil sampling.” Soil Sci., 58, 275–288. Cochran, W. G. (1977). Sampling techniques, John Wiley and Sons, New York. Corwin, D. L., and Rhoades, J. D. (1984). “Measurement of inverted electrical conductivity profiles using electromagnetic induction.” Soil Sci. Soc. of Am. Proc., 41, 966–971. de Jong, E., Ballantyne, A. K., Cameron, D. R., and Read, D. W. L. (1979). “Measurement of apparent electrical conductivity of soils by an electromagnetic induction probe to aid salinity surveys.” Soil Sci. Soc. Amer. Proc., 43, 810–812.

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

291

El-Araby, A. M., Warrick, A. W., and Tucker, T. C. (1987). “Sample volume effect on determination of soil salinity.” Commun. Soil Sci. Plant Anal., 18, 593–599. Elliott, R. L., and Walker, W. R. (1980). “Furrow irrigation infiltration and advance functions.” ASAE Paper 80-2075, presented at the 1980 American Society of Agricultural Engineers Summer Meeting, San Antonio, Tex., June 15–18, 1980, ASABE, St. Joseph, Mich. Gajem, Y. M., Warrick, A. W., and Myers, D. E. (1981). “Spatial dependence of physical properties of a typic torrifluvent soil.” Soil Sci. Soc. of Amer. Proc., 45, 709–715. Gutjahr, A. (1984). “Spatial variability: Geostatistical methods.” Soil Spatial Var., Proceedings of a workshop of the International Soil Science Society and the Soil Science Society of America, Las Vegas, Nevada, November 30–December 1, 1984, Soil Science Society of America, Madison, Wisc. Hajrasuliha, S., Baniabbassi, N., Metthey, J., and Nielsen, D. R. (1980). “Spatial variability of soil sampling for salinity studies in southwest Iran.” Irrig. Sci., 1, 197–208. Halvorson, A. D., Rhoades, J. D., and Reule, C. A. (1977). “Soil salinity: Four electrode conductivity relationships for soils of the northern Great Plains.” Soil Sci. Soc. of Amer. Proc., 41, 966–971. Hanson, B. R., and Oster, J. D. (1986). “Rapid field-wide assessment of soil salinity.” ASAE Paper 86-2063, presented at the 1986 American Society of Agricultural Engineers Summer Meeting, California Polytechnic Institute, San Luis Obispo, Calif., June 29–July 2, 1986, ASABE, St. Joseph, Mich. Hanson, B. R., Prichard, T. L., Goldhamer, D., and Schulbach, H. (1988). Evaluating and predicting furrow irrigation system performance, draft report on Interagency Agreement B55427 between the California Dept. of Water Resources, Office of Water Conservation, and the University of California Cooperative Extension/Dept. of Land, Air, and Water Resources. Hassan, H. M., Warrick, A. W., Amoozegar-fard, A. (1983). “Sampling volume effects on determining salt in a soil profile.” Soil Sci. Soc. of Amer. Proc., 47, 1265–1267. Hawley, M. E., McCuen, R. H., and Jackson, T. J. (1982). “Volume-accuracy relationship in soil moisture sampling.” Proc., ASCE Irrigation and Drainage Div., 108(IR1), ASCE, Reston, Va., 1–11. Hendrickx, J. M. H., Das, B., Corwin, D. L., Wraith, J. M., Kachanoski, R. G. (2002). “Indirect measurement of solute concentration,” in Methods of soil analysis, Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., SSSA Book Series 5, Soil Science Society of America, Madison, Wisc., 1274–1306. Hoffman, G. J., Shannon, M. C, and Jobes, J. A. (1985). “Influence of rain on soil salinity and lettuce yield,” in Drip/trickle irrigation in action, Proc. Third Drip/Trickle Irrigation Congress, Nov. 18–21, 1985, Fresno, California, American Society of Agricultural Engineers, St. Joseph, Mich. Maas, E. V. and Grattan, S. R. (1999). “Crop yields as affected by salinity,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Agronomy Monograph 38, ASA/CSSA/SSA, Madison, Wisc., 55–108. McBratney, A. B., and Webster, R. (1983). “How many observations are needed for regional estimation of soil properties?” J. Soil Sci., 135, 177–183.

292

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

McNeill, J. D. (1980). Electrical conductivity of soils and rocks, Technical Note TN-5, Geonics Ltd., Mississauga, Ontario, Canada. Miyamoto, S., and Cruz, I. (1986.) “Spatial variability and soil sampling for salinity and sodicity appraisal in surface-irrigated orchards.” Soil Sci. Soc. of Amer. Proc. 50, 1020–1026. ———. (1987). “Spatial variability of soil salinity in furrow-irrigated torrifluvents.” Soil Sci. Soc. Amer. Proc., 51, 1019–1025. Nable, R. O., Bañuelos, G. S., and Paull, J. G. (1997). “Boron toxicity.” Plant Soil, 193, 181–198. Peterson, R. G., and Calvin, L. D. (1967). “Sampling,” in Methods of soil analysis, Part 1. American Society of Agronomy Monograph 9, ASA, Madison, Wisc. Reisenauer, H. M. (1978). Soil and plant-tissue testing in California, Bulletin 1879, University of California, Davis, Calif. Rhoades, J. D., and Corwin, D. L. (1981). “Determining soil electrical conductivitydepth relations using an inductive electromagnetic soil conductivity meter.” Soil Sci. Soc. Amer. Proc., 45, 255–260. Rhoades, J. D., and Halvorsen, A. D. (1977). Electrical conductivity methods for detecting and delineating saline seeps and measuring salinity in northern Great Plains soils, ARS W-2, Agricultural Research Service, U.S. Department of Agriculture, Washington, D.C. Rhoades, J. D., and Ingvalson, R. D. (1971). “Determining salinity in field soils with soil resistance measurements.” Soil Sci. Soc. Amer. Proc., 35, 54–60. Rhoades, J. D., Raats, P. A. C., and Prather, R. J. (1976). “Effects of liquid-phase electrical conductivity, water content, and surface conductivity on bulk soil electrical conductivity.” Soil Sci. Soc. Amer. Proc., 40, 651–655. Rhoades, J. D., and van Schilfgaarde, J. (1976). “An electrical conductivity probe for determining soil salinity.” Soil Sci. Soc. Amer. Proc., 40, 647–651. Shumway, R. H. (1988). Applied statistical time series analysis, Prentice Hall, Upper Saddle River, N.J. Snedecor, G. W., and Cochran, W. G. (1980). Statistical methods, The Iowa State University Press, Ames, Iowa. Wadleigh, C. H., and Fireman, M. (1948). “Salt distribution under furrow and basin irrigation cotton and its effect on water removal.” Soil Sci. Soc. Amer. Proc., 13, 527–530. Wadpole, R. E., and Myers, R. H. (1978). Probability and statistics for engineers and scientists, MacMillan Publishing Co., New York. Webster, R. (1977). “Spectral analysis of gilgai soil.” Aust. J. Soil Res., 15, 191–204. William, B. G., and Baker, G. C. (1982). “An electromagnetic induction technique for reconnaissance survey of soil salinity hazards.” Aust. J.. Soil Res., 20, 107–118. Yamane, T. (1967). Elementary sampling theory, Prentice-Hall, Upper Saddle River, N.J. Yates, F. (1981). Sampling methods for censuses and surveys, Charles Griffin and Co. Ltd., London. Zobeck, T. M., Fausey, N. R., and Al-Hamdan, N. S. (1985). “Effect of sample cross-sectional area on saturated hydraulic conductivity in two structured clay soils.” Trans. ASAE, 28, 791–794.

FIELD SAMPLING OF SOIL, WATER, AND PLANTS

293

NOTATION CV  coefficient of variation, expressed as a percent e  absolute allowable error between the estimate of sample mean and the true value k  error expressed as a decimal fraction of the mean N  total number of observations in sample Nh  number of observations in stratum h s  sample standard deviation sh  standard deviation for stratum h — sX  standard error of the mean s2  sample variance Wh  relative weight of stratum h — X  sample mean or average value Xi  individual observation — Xh  mean for stratum h z  multiplier, which equals 1.96, 1.64, and 1.28 for confidence levels of 95%, 90%, and 80%, respectively

This page intentionally left blank

CHAPTER 10 LABORATORY AND FIELD MEASUREMENTS Dennis L. Corwin, S. M. Lesch, and D. B. Lobell

INTRODUCTION Soil salinity refers to the presence of major dissolved inorganic solutes in the soil solution (i.e., aqueous liquid phase of the soil and its solutes), which consist of soluble and readily dissolvable salts, including charged  2 2 species (e.g., Na, K, Mg2, Ca2, Cl, HCO 3 , NO3 , SO4 and CO3 ), non-ionic solutes, and ions that combine to form ion pairs. The primary source of salts in soil and water is the geochemical weathering of rocks from the earth’s upper strata, with atmospheric deposition and anthropogenic activities serving as secondary sources. The predominant mechanism causing the accumulation of salt in the rootzone of agricultural soils is loss of water through evapotranspiration (ET; the combined processes of evaporation from the soil surface and plant transpiration), which selectively removes water, leaving salts behind. The accumulation of soil salinity can result in reduced plant growth, reduced yields, and in severe cases, crop failure. Salinity limits water uptake by plants by reducing the osmotic potential, making it more difficult for the plant to extract water. Salinity may also cause specific-ion toxicity (e.g., Na ion toxicity) or upset the nutritional balance of plants. In addition, the salt composition of the soil water influences the composition of cations on the exchange complex of soil particles, which influences soil permeability and tilth. Irrigated agriculture, which accounts for 35% to 40% of the world’s total food and fiber, is adversely affected by soil salinity on roughly half of all irrigated soils (totaling about 250 million ha), with more than 20 million ha severely affected by salinity worldwide (Rhoades and Loveday 1990). Because of the potential detrimental impacts of soil salinity accumulation, it is a crucial soil chemical property that is routinely measured 295

296

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and monitored. This chapter describes common field and laboratory techniques for measuring salinity in the soil and water, with discussion of their practicability and reliability.

FACTORS AFFECTING SOIL SALINITY The accumulation of soil salinity is a consequence of a variety of processes, some of which are illustrated in Fig. 10-1. In arid and semiarid areas, for example, where precipitation is less than evaporation, salts can accumulate at the soil surface when the depth to the water table is less than 1 to 1.5 m, depending on the soil texture. The accumulation of salts at the soil surface is the consequence of the upward flow of water and subsequent transport of salts due to capillary rise driven by the evaporative process. However, the most common cause for the accumulation of salts is ET by plants, which results in an increase in salt concentration with depth through the rootzone (see graph in Fig. 10-1) and the accumulation of salts below the rootzone. The level of salt accumulation within and below the rootzone due to ET depends on the fraction of irrigation or pre-

FIGURE 10-1. Various examples of how salts accumulate in soil.

LABORATORY AND FIELD MEASUREMENTS

297

cipitation that flows beyond the rootzone, referred to as the leaching fraction (LF). As the LF increases, the total salts within the rootzone decrease due to their removal from the rootzone by leaching. A third process, which is common in the northern Great Plains of the United States, is the formation of saline seeps. There are several forms of saline seeps differing in their means of development. In general, saline seeps form downslope of recharge areas in locations where discharge is occurring because of the presence of a low conductivity layer and shallow water table (Fig. 10-1). Salts are leached from the upslope recharge area, which tends to be an area of higher conductivity than the downslope discharge area. Once the water and salts from upslope reach the downslope low conductivity layer, they accumulate and are forced to the surface by evaporation.

DIRECT AND INDIRECT ANALYSIS OF SOIL SALINITY The most common technique for the measurement of soil salinity is laboratory analysis of aqueous extracts of soil samples. Soil salinity is quantified in terms of the concentration of total salts in the soil. The measurement of the total salt concentration of the aqueous extracts of soil samples can be done either directly through the chemical analysis of the chemical constitutes that comprise soil salinity, or indirectly through the measurement of electrical conductivity (EC). The chemical species of primary interest in salt-affected soils include four major cations (Na, K, Mg2, 2 Ca2) and four major anions (Cl, HCO3 , SO2 4 and CO3 ) in the soil solution; exchangeable cations (Na, K, Mg2, Ca2); and the precipitated salts calcium carbonate (lime) and calcium sulfate (gypsum). Other soil properties of concern in salt-affected soils include pH, water content of the saturation paste, sodium adsorption ratio (SAR), and exchangeable sodium percentage (ESP). Detailed analytical techniques for measuring all of these salinity-related properties can be found in Methods of Soil Analysis (Part 3, Sparks 1996; Part 4, Dane and Topp 2002). However, a chemical analysis of the salinity-related properties of primary concern is too labor- and cost-intensive to be practical, particularly when large numbers of samples are involved, such as field-scale assessments of salinity; consequently, the salinity of aqueous extracts of soil samples has been most often measured by EC. It is well known that the EC of water is a function of its chemical salt composition and total salt concentration (U.S. Salinity Laboratory 1954). In the laboratory, soil salinity is commonly determined from the measurement of the EC of soil-solution extracts, where the current-carrying capacity of the soil solution is proportional to the concentration of ions in the solution. The total concentration of the soluble salts in soil is measured by EC of the soil solution in dS m1. Over a range of mixed salt

298

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

concentrations commonly found in soils (1 to 50 meq L1), total salt concentration (C) in meq L1 is linearly related to electrical conductance of the solution by Eq. 10-1: C 艐 10 ECw @ 25 °C

(10-1)

where ECw @ 25 °C is the electrical conductivity of the soil solution referenced to 25 °C (dS m1). If C is measured in mg L1 or ppm, then C is related to ECw @ 25 °C by a factor of 640 (i.e., C 艐 640 ECw @ 25 °C). Over a broader range of salt concentrations (1 to 500 meq L1) the relationship between C and ECw @25 °C is no longer linear and is best fit with a thirdorder polynomial or an exponential equation. Another useful relationship is between osmotic potential ( ) and EC, where in bars is related to ECw @ 25 °C by a factor of 0.36 (e.g., 艐 0.36 ECw @ 25 °C; for 3 ECw @ 25 °C 30 dS m1). Theoretical and empirical approaches are available to predict the EC of a solution from its solute composition. Equation 10-2 is an example of a theoretical approach based on Kohlrauch’s Law of independent migration of ions, where each ion contributes to the current-carrying ability of an electrolyte solution: EC  ∑ ECi  ∑ i

i

ci ( 0i   ci ) 1000

(10-2)

where EC is the specific conductance (dS m1), ECi is the ionic specific conductance (dS m1), ci is the concentration of the ith ion (mmolc L1), 0i is the ionic equivalent conductance at infinite dilution (cm2 S mol1 c ), and  is an empirical interactive parameter (Harned and Owen 1958). Equation 10-3 shows an empirical equation developed by Marion and Babcock (1976): log TSS  0.990  1.055 log EC (r2  0.993)

(10-3)

where TSS is the total soluble salt concentration (mmolc L1). Temperature influences EC; consequently, EC must be referenced to a specific temperature to permit comparison. Electrolytic conductivity increases at a rate of approximately 1.9% per degree centigrade increase in temperature. Customarily, EC is expressed at a reference temperature of 25 °C. The EC measured at a particular temperature t (in °C), ECt, can be adjusted to a reference EC at 25 °C, EC25 °C, using Eq. 4 from USDA Handbook 60 (U.S. Salinity Laboratory 1954): EC25 °C  ft ECt

(10-4)

LABORATORY AND FIELD MEASUREMENTS

299

where, ft  0.4470  1.4034 exp(t/26.815) [from Sheets and Hendrickx (1995)].

METHODS OF LABORATORY, LYSIMETER, AND PLOT-SCALE SOIL SALINITY MEASUREMENT Historically, four principal methods have been used for measuring soil salinity in the laboratory, in soil lysimeter columns, and at field-plot scales: (1) the EC of soil solution at or near field capacity, of extracts at higher than normal water contents (i.e., including saturation and soil to water ratios of 1⬊1, 1⬊2, and 1⬊5), or of a saturation paste; (2) in-situ measurement of electrical resistivity (ER); (3) noninvasive measurement of EC with electromagnetic induction (EMI); and, most recently, (4) in-situ measurement of EC with time domain reflectometry (TDR). Electrical Conductivity To determine the EC of a soil solution extract, the solution is placed in a cell containing two electrodes of constant geometry and distance of separation. An electrical potential is imposed across the electrodes, and the resistance of the solution between the electrodes is measured. The measured conductance is a consequence of the solution’s salt concentration and the electrode geometry whose effects are embodied in a cell constant. At constant potential, the current is inversely proportional to the solution’s resistance as shown in Eq. 10-5: ECt  k/Rt

(10-5)

where ECt is the electrical conductivity of the soil solution in dS m1 at temperature t (°C), k is the cell constant, and Rt is the measured resistance in ohms at temperature t. One dS m1 is equivalent to one mS cm1 and one mmho cm1, where mmho cm1 are the obsolete units of EC. Except for the measurement of EC of a saturated soil paste (ECp), the determination of soluble salts in disturbed soil samples consists of two basic steps: (1) preparation of a soil-water extract, and (2) the measurement of the salt concentration of the extract using EC. Customarily, soil salinity has been defined in terms of laboratory measurements of the EC of the extract of a saturated soil paste (ECe). This is because it is impractical for routine purposes to extract soil water from samples at typical field water contents; consequently, soil-solution extracts must be made at saturation or higher water contents. The saturation paste extract is the lowest soil-to-water ratio that can be easily extracted with vacuum, pressure, or centrifugation, while providing a sample of sufficient size to analyze. The

300

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

water content of a saturation paste is roughly twice the field capacity for most soils. Furthermore, ECe has been the standard measure of salinity used in salt-tolerance plant studies. Most data on the salt tolerance of crops have been expressed in terms of the EC of the saturation paste extract (Bresler et al. 1982; Maas 1986). Unfortunately, the partitioning of solutes over the three soil phases (gas, liquid, solid) is influenced by the soil-to-water ratio at which the extract is made, so the ratio needs to be standardized to obtain results that can be applied and interpreted universally. Commonly used extract ratios, other than a saturated soil paste, are 1⬊1, 1⬊2, and 1⬊5 soil-to-water mixtures. These extracts are easier to prepare than saturation paste extracts. With the exception of sandy soils, soils containing gypsum, and organic soil, the concentrations of salt and individual ions are approximately diluted by about the same ratio between field conditions and the extract for all samples, which allows conversions between water contents using dilution factors. The conversion of EC from one extract to another is commonly done using a simple dilution factor. For example, if the gravimetric saturation percentage (SP) is 100%, then ECe  EC1⬊1  5 EC1⬊5 or if SP  50%, then ECe  2 EC1⬊1  10 EC1⬊5. However, this is not recommended because of potential dissolution-precipitation reactions that may occur. At best, the use of a dilution factor to convert from one extract to another is an approximation. Any dilution above field water contents introduces errors in the interpretation of data. The greater the dilution is, the greater the deviation between ionic ratios in the sample and the soil solution under field conditions. These errors are associated with mineral dissolution, ion hydrolysis, and changes in exchangeable cation ratios. In particular, soil samples containing gypsum deviate the most because the calcium (Ca) and sulfate concentration remain nearly constant with sample dilution, while the concentrations of other ions decrease with dilution. The standardized relationship between the extract and the conditions of the soil solution in the field for different soils is not applicable with the use of soil-to-water above saturation. However, the recent development of Extract Chem software by Suarez and Taber (2007) allows for the accurate conversion from one extract ratio to another, provided sufficient chemical information is known (for example, knowledge of the major cations and anions and presence/absence of gypsum). The disadvantage of determining soil salinity using a soil sample is the time and labor required, which translates into high cost. However, there is no more accurate way of measuring soil salinity than with extracts from soil samples. Prior to the 1950s, much of the data on soil salinity were obtained by using a 50-mL cylindrical conductivity cell, referred to as a “Bureau of Soils cup,” filled with a saturated soil paste to estimate soluble-salt concentrations by measuring the ECp. This approach was fast and easy; con-

LABORATORY AND FIELD MEASUREMENTS

301

sequently, it was used to map and diagnose salt-affected soils. When Reitemeier and Wilcox (1946) determined that plant responses to soil salinity correlated more closely with the EC values of the saturation paste extract, the use of the paste was discontinued. A theoretical relationship between ECp and ECe has since been developed to overcome the cell’s shortcomings. This was done by developing a simple method of determining the volumetric water and volumetric solid contents of the saturation paste, the conductance of the sample surface, and the current pathway of the water in the cell (Rhoades et al. 1999b). Even so, the relationship between ECp and ECe is complex; consequently, the measurement of ECp is not recommended except in instances where obtaining an extract of the saturation paste is not possible or is impractical. Figure 10-2 graphically illustrates the theoretical complexity of the relationship between ECp and ECe based on the dual parallel pathway conductance model of Rhoades et al. (1989a,b). Soil salinity can also be determined from the measurement of the EC of a soil solution (ECw), where the water content of the soil is less than saturation, usually at field capacity. Ideally, ECw is the best index of soil salinity because this is the salinity actually experienced by the plant root. Nevertheless, ECw has not been widely used to express soil salinity for various reasons: (1) it varies over the irrigation cycle as the soil water content changes, so it is not single-valued; and (2) the methods for obtaining soil solution samples at typical field water contents are too labor-, time-, and cost-intensive to be practical (Rhoades et al. 1999b). For disturbed soil samples, soil solution can be obtained in the laboratory by displacement, compaction, centrifugation, molecular adsorption, and vacuum- or

FIGURE 10-2. Theoretical relationship between ECe and ECp based on the dual parallel pathway conductance model of Rhoades et al. (1989a,b).

302

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

pressure-extraction methods. For undisturbed soil samples, ECw can be determined with a soil-solution extractor (Fig. 10-3a), often referred to as a porous cup extractor, or using an in situ, imbibing-type porous-matrix salinity sensor (Fig. 10-3b).

FIGURE 10-3. Instruments for obtaining soil-solution extracts at less than saturation, including (a) soil-solution extractor system (from Corwin 2002a), and (b) porous-matrix salinity sensor (from Corwin 2002b). Reprinted with permission from Soil Science Society of America.

LABORATORY AND FIELD MEASUREMENTS

303

Porous cup soil solution extractors include zero-tension and tension (or suction) cups. Historically, suction cups have been more widely used. No single soil-solution sampling device will perfectly sample under all conditions, so it is important to understand the strengths and limitations of a sampler to determine when to apply certain sampling methods in preference over other methods. In structured soils, suction cups do not sample water in preferential flow paths. Zero-tension cups will almost always sample just saturated flow, which is more closely associated with preferential flow channels, and tension samplers will more efficiently sample unsaturated flow within soil aggregates. Zero-tension cups represent the flux concentration, whereas the tension samples are approximations of resident concentrations. The basic design of a suction cup apparatus consists of a suction cup, sample collection bottle, manifold (if there is more than one suction cup), an overflow trap, an applied vacuum, and connective tubing (Fig. 10-3a). The general principle behind the operation of suction cup extractors is simply that suction (preferably the suction at field moisture capacity) is placed against the porous cup. This suction opposes the capillary force of the soil at field capacity, causing soil solution to be drawn across the porous wall of the cup as a result of the induced pressure gradient. The imbibed solution is stored in a sample collection chamber. This approach for extracting soil solution is viable when the soil-water matric potential is greater than about –30 kPa (kilopascals, a standard unit of pressure). The salinity sensor consists of a porous ceramic substrate with an embedded platinum mesh electrode, which is placed in contact with the soil to measure the EC of the soil solution that has been imbibed by the ceramic (Fig. 10-3b). The salinity sensor contains a thermistor designed to temperature-correct the EC readings. Both the electrolytic element and thermistor of a salt sensor (Fig. 10-3b) must be calibrated for proper operation. Calibration is necessary because of (1) the variation in water retention and porosity characteristics of each ceramic, and (2) the variation in electrode spacing, both of which cause the cell constant to vary for each salt sensor. The calibration can change with time, so periodic recalibration is necessary. There are various advantages and disadvantages to measuring EC using soil solution extractors or soil salinity sensors. The obvious advantage is that ECw is being measured, but this is outweighed by the disadvantages. Even though the sample volume of a soil-solution extractor (10 to 100 cm3) is roughly an order of magnitude larger than a salinity sensor (1 to 2 cm3), both measure significantly limited sample volumes; consequently, there are serious doubts about the ability of soil-solution extractors and porousmatrix salinity sensors to provide representative soil-water samples, particularly at field scales (England 1974; Raulund-Rasmussen 1989; Smith et al. 1990). Soil heterogeneity significantly affects chemical concentrations in

304

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the soil solution. Because of their small sphere of measurement, neither solution extractors nor salt sensors adequately integrate spatial variability (Amoozegar-Fard et al. 1982; Haines et al. 1982; Hart and Lowery 1997). Biggar and Nielsen (1976) suggested that soil-solution samples are “point samples” that can provide a good qualitative measurement of soil solutions but are not adequate quantitative measurements unless the fieldscale variability is adequately established. Furthermore, salinity sensors demonstrate a response time lag that is dependent on the diffusion of ions between the soil solution and solution in the porous ceramic, which is affected by (1) the thickness of the ceramic conductivity cell, (2) the diffusion coefficients in soil and ceramic, and (3) the fraction of the ceramic surface in contact with soil (Wesseling and Oster 1973). The salinity sensor is generally considered the least desirable method for measuring ECw because of its low sample volume, unstable calibration over time, and slow response time (Corwin 2002b). Soil-solution extractors have the drawback of requiring considerable maintenance due to cracks in the vacuum lines and clogging of the ceramic cups with algae and fine soil particles. Both solution extractors and salt sensors are considered slow and labor-intensive. The ability to obtain the EC of a soil solution when the water content is at or less than field capacity, which are the water contents most commonly found in the field, is considerably more difficult than extracts for water contents at or above saturation because of the pressure or suction required to remove the soil solution at field capacity and lower water contents. The EC of the saturated paste is the easiest to obtain, followed by the EC of extracts greater than SP, followed by the EC of extracts less than SP. However, ECe is most preferred; consequently, either measuring ECe or being able to relate the EC measurement to ECe is critical. The techniques of ER, EMI, and TDR measure ECa, which is discussed in the next section. Electrical Resistivity Because of the time and cost of obtaining soil-solution extracts and the lag time associated with porous ceramic cups, developments in the measurement of soil EC shifted in the 1970s to the measurement of the soil EC of the bulk soil, referred to as apparent soil electrical conductivity (ECa). Apparent soil electrical conductivity provides an immediate, easy-to-take measurement of conductance with no lag time and no need to obtain a soil extract. However, ECa is a complex measurement that has been misinterpreted and misunderstood by users in the past due to the fact that it is a measure of the EC of the bulk soil, not just a measure of the conductance of the soil solution, which is the desired measurement, since the soil solution is the soil phase that contains the salts affecting plant roots.

LABORATORY AND FIELD MEASUREMENTS

305

The most comprehensive body of research concerning the adaptation and application of geophysical techniques to the measurement of soil salinity within the rootzone (top 1 to 1.5 m of soil) was compiled by scientists at the U.S. Salinity Laboratory. The most recent reviews of this body of research can be found in Corwin (2005), Corwin and Lesch (2005a), and Rhoades et al. (1999b). Electrical resistivity (ER) was originally used by geophysicists to measure the resistivity of the geological subsurface. Electrical resistivity methods involve the measurement of the resistance to current flow across four electrodes inserted in a straight line on the soil surface at a specified distance between the electrodes (Corwin and Hendrickx 2002). The electrodes are connected to a resistance meter that measures the potential gradient between the current and potential electrodes (Fig. 10-4). These methods were developed in the second decade of the 1900s by Conrad Schlumberger in France and Frank Wenner in the United States for the evaluation of near-surface ER (Burger 1992; Rhoades and Halvorson 1977). Although two electrodes (one current and one potential electrode) can be used, the stability of the reading is greatly improved with the use of four electrodes. The resistance is converted to EC using Eq. 10-5, where the cell constant, k, in that equation is determined by the electrode configuration and distance. The depth of penetration of the electrical current and the volume of measurement increase as the interelectrode spacing increases. The fourelectrode configuration is referred to as a Wenner array when the four electrodes are equidistantly spaced (interelectrode spacing  a). For a

FIGURE 10-4. Schematic of four-electrode probe electrical resistivity used to measure apparent soil electrical conductivity. From Corwin and Hendrickx (2002) with permission from Soil Science Society of America.

306

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

homogeneous soil, the depth of penetration of the Wenner array is a and the soil volume measured is roughly a3. Other four-electrode configurations are frequently used, as discussed by Burger (1992), Dobrin (1960), and Telford et al. (1990). The influence of the interelectrode configuration and distance on ECa is reflected in Eq. 10-6: ⎧ ⎪ ⎛ 1000 ⎞ ⎪⎪ ft EC a ,25° C  ⎜ ⎟⎨ 1 ⎝ 2 Rt ⎠ ⎪ ⎪ 11 1  1 ⎪⎩ r1 r2 R1 R2

⎫ ⎪ ⎪⎪ ⎬ ⎪ ⎪ ⎪⎭

(10-6)

where ECa,25 °C is the apparent soil electrical conductivity temperature corrected to a reference of 25 °C (dS m1), and r1, r2, R1, and R2 are the distances in cm between the electrodes as shown in Fig. 10-4. For the Wenner array, where a  r1  r2  R1  R2, Eq. 10-6 reduces to ECa  159.2 ft/aRt and 159.2/a represents the cell constant (k). A variety of four-electrode probes have been commercially developed, reflecting diverse applications. Burial and insertion four-electrode probes are used for continuous monitoring of ECa and to measure soil profile ECa, respectively (Fig. 10-5a,b). These probes have volumes of measurement roughly the size of a football (i.e., about 2,500 cm3). Bedding probes with small volumes of measurement of roughly 25 cm3 were used to monitor ECa in seed beds (Fig. 10-5c), but these probes are no longer commercially available. Only the Eijelkamp conductivity meter and probe are commercially available, which is similar in use and basic design to the insertion probe in Fig. 10-5b. Measuring ER is an invasive technique that requires good contact between the soil and the four electrodes inserted into the soil; consequently, it produces less reliable measurements in dry or stony soils than a noninvasive measurement such as EMI. Nevertheless, ER has a flexibility that has proven advantageous for field application, that is, the depth and volume of measurement can be easily changed by altering the spacing between the electrodes. A distinct advantage of the ER approach is that the volume of measurement is determined by the spacing between the electrodes, which makes a large volume of measurement possible. For example, a 1-m interelectrode spacing for a Wenner array results in a volume of measurement of more than 3 m3. This large volume of measurement integrates the high level of local-scale variability often associated with ECa measurements.

LABORATORY AND FIELD MEASUREMENTS

307

FIGURE 10-5. Examples of various four-electrode probes: (a) burial probe, (b) insertion probe, and (c) bedding probe.

308

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Because ECe is regarded as the standard measure of salinity, a relation between ECa and ECe is needed to relate ECa to salinity. The relationship between ECa and ECe is linear when ECa is above 2 dS m1 and is dependent on soil texture, as shown in Fig. 10-6. Rough approximations of ECe from ECa in dS m1 when ECa 2 dS m1 are: ECe 艐 3.5 ECa for fine-textured soils, ECe 艐 5.5 ECa for medium-textured soils, and ECe 艐 7.5 ECa for coarse-textured soils. For ECa 2 dS m1, the relation between ECa and ECe is more complex. In general, at ECa 2 dS m1 salinity is the dominant conductive constituent; consequently, the relationship between ECa and ECe is linear. However, when ECa 2 dS m1, other conductive properties (e.g., water and clay content) and properties influencing conductance (e.g., bulk density) have greater influence. For this reason, it is recommended that below an ECa of 2 dS m1, the relation between ECa and ECe is established by calibration. The calibration between ECa and ECe is established by measuring the ECe of soil samples taken at a minimum of three to four locations within a study area where associated ECa measurements have been taken. These samples should reflect a range of ECas and should be collected over the volume of measurement for the ECa technology used (i.e., ER or EMI). Electromagnetic Induction Apparent soil electrical conductivity can be measured noninvasively with EMI. A transmitter coil located at one end of the EMI instrument

FIGURE 10-6. Relationships between ECa and ECe for representative soil types found in the northern Great Plains, United States. Modified from Rhoades and Halvorson (1977).

LABORATORY AND FIELD MEASUREMENTS

309

induces circular eddy-current loops in the soil, with the magnitude of these loops directly proportional to the EC in the vicinity of that loop (Fig. 10-7). Each current loop generates a secondary electromagnetic field that is proportional to the value of the current flowing within the loop. A fraction of the secondary induced electromagnetic field from each loop is intercepted by the receiver coil of the instrument and the sum of these signals is amplified and formed into an output voltage, which is related to a depth-weighted ECa. The amplitude and phase of the secondary field will differ from those of the primary field as a result of soil properties (e.g., clay content, water content, salinity), spacing of the coils and their orientation, frequency, and distance from the soil surface (Hendrickx and Kachanoski 2002). The most commonly used EMI conductivity meters in soil science and in vadose zone hydrology are the Geonics EM-31 and EM-38 (Geonics Ltd., Mississauga, Ontario, Canada) and the DUALEM-2 (Dualem Inc., Milton, Ontario, Canada). The EM-38 has had considerably greater application for agricultural purposes because the depth of measurement corresponds roughly to the rootzone (i.e., generally 1 to 1.5 m.). When the instrument is placed in the vertical coil configuration (EMv, with the coils perpendicular to the soil surface), the depth of measurement is about 1.5 m; in the horizontal coil configuration (EMh, with the coils parallel to the soil surface), the depth of the measurement is 0.75 to 1.0 m. The EM-31 has an intercoil spacing of 3.66 m, which corresponds to a penetration depth of 3 and 6 m in the horizontal and vertical dipole orientations, respectively, which extends well beyond the rootzone of agricultural

FIGURE 10-7. Schematic of the operation of electromagnetic induction equipment, using an EM-38.

310

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

crops. However, the EM-38 has one major pitfall—the need for calibration—which the DUALEM-2 does not require. Further details about and operation of the EM-31 and EM-38 equipment are discussed in Hendrickx and Kachanoski (2002). Documents concerning the DUALEM-2 can be found in Dualem (2007). Apparent soil electrical conductivity measured by EMI at ECa 1.0 dS m1 is given by Eq. 10-7, from McNeill (1980):

EC a 

⎛ Hs ⎞ 4 ⎜ ⎟ 2 ⎜ 2  0 fs ⎝ H p ⎟⎠

(10-7)

where ECa is measured in S m1; Hp and Hs are the intensities of the primary and secondary magnetic fields at the receiver coil (A m1), respectively; f is the frequency of the current (Hz); 0 is the magnetic permeability of air (4 107 H m1); and s is the intercoil spacing (m). Both ER and EMI are rapid and reliable technologies for the measurement of ECa, each with its advantages and disadvantages. The primary advantage of EMI over ER is that EMI is noninvasive, so it can be used on dry and stony soils that would not be amenable to invasive ER equipment. The disadvantage is that ECa measured with EMI is a depthweighted value that is nonlinear, whereas ER provides an ECa measurement that is nearly linear with depth. More specifically, EMI concentrates its measurement of conductance over the depth of penetration at shallow depths, while ER is more uniform with depth. Because of the linearity of the response function of ER, the ECa for a discrete depth interval of soil, ECx, can be determined with the Wenner array by measuring the ECa of successive layers by increasing the interelectrode spacing from ai1 to ai and using Eq. 10-8 from Barnes (1952) for resistors in parallel: EC x  EC aiai1 

(EC ai ⋅ ai )  (EC ai1  ai1 ) ( ai  ai1 )

(10-8)

where ai is the interelectrode spacing, which equals the depth of sampling, and ai1 is the previous interelectrode spacing, which equals the depth of previous sampling. Measurements of ECa by ER and EMI at the same location and over the same volume of measurement are not comparable because of the nonlinearity of the response function with depth for EMI and the linearity of the response function for ER. An advantage of ER over EMI is the ease of instrument calibration. Calibrating the EM-31 and EM-38 is more involved then for ER equipment. However, there is no need to calibrate the DUALEM-2.

LABORATORY AND FIELD MEASUREMENTS

311

Time Domain Reflectometry Time domain reflectometry (TDR) was initially adapted for use in measuring water content,  (Topp and Davis 1981; Topp et al. 1980, 1982). The TDR technique is based on the time for a voltage pulse to travel down a soil probe and back, which is a function of the dielectric constant (ε) of the porous media being measured. Later, Dalton et al. (1984) demonstrated the utility of TDR to also measure ECa. The measurement of ECa with TDR is based on the attenuation of the applied signal voltage as it traverses the medium of interest, with the relative magnitude of energy loss related to ECa (Wraith 2002). By measuring ε,  can be determined through calibration (Dalton 1992). The ε is calculated with Eq. 10-9, from Topp et al. (1980): 2 ⎛ ct ⎞ ⎛ l ⎞ ε  ⎜ ⎟  ⎜⎜ a ⎟⎟ ⎝ 2l ⎠ ⎝ lv p ⎠

2

(10-9)

where c is the propagation velocity of an electromagnetic wave in free space (2.997  108 m s1), t is the travel time (s), l is the real length of the soil probe (m), la is the apparent length (m) as measured by a cable tester, and vp is the relative velocity setting of the instrument. The relationship between  and ε is approximately linear and is influenced by soil type, b, clay content, and organic matter (Jacobsen and Schjønning 1993). By measuring the resistive load impedance across the probe (ZL), ECa can be calculated with Eq. 10-10, from Giese and Tiemann (1975), EC a 

ε0 c Z0 ⋅ l ZL

(10-10)

where ε0 is the permittivity of free space (8.854  1012 F m1), Z0 is the probe impedance (), and ZL  Zu[(2V0/Vf)  1]1 where Zu is the characteristic impedance of the cable tester, V0 is the voltage of the pulse generator or zero-reference voltage, and Vf is the final reflected voltage at an exceedingly long time. To reference ECa to 25 °C, Eq. 10-11 is used: ECa  Kc ftZL1

(10-11)

where Kc is the TDR probe cell constant (Kc [m1]  ε0cZ0/l), which is determined empirically. Advantages of TDR for measuring ECa include (1) a relatively noninvasive nature since there is only minor interference with soil processes, (2) an ability to measure both soil water content and ECa, (3) an ability to

312

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

detect small changes in ECa under representative soil conditions, (4) the capability of obtaining continuous unattended measurements, and (5) a lack of a calibration requirement for soil water content measurements in many cases (Wraith 2002). Even so, TDR has not been the instrument of choice for the measurement of salinity, whether in the laboratory or in the field; consequently, it will not be discussed in detail. Soil ECa has become one of the most reliable and frequently used measurements to characterize field variability for application to precision agriculture due to its ease of measurement and reliability (Corwin and Lesch 2003). Although TDR has been demonstrated to compare closely with other accepted methods of ECa measurement (Heimovaara et al. 1995; Mallants et al. 1996; Reece 1998; Spaans and Baker 1993), it is still not sufficiently simple, robust, or fast enough for the general needs of field-scale soil salinity assessment (Rhoades et al. 1999b). Only ER and EMI have been adapted for the georeferenced measurement of ECa at field-scales and larger (Rhoades et al. 1999a,b).

SOIL-RELATED (EDAPHIC) FACTORS INFLUENCING THE ECa MEASUREMENT The earliest field applications of geophysical measurements of ECa in soil science involved the determination of salinity through the soil profile of arid zone soils (Cameron et al. 1981; Corwin and Rhoades 1982, 1984; de Jong et al. 1979; Halvorson and Rhoades 1976; Rhoades and Corwin 1981; Rhoades and Halvorson 1977; Williams and Baker 1982). However, it became apparent that the measurement of ECa in the field to infer soil salinity was more complicated than initially anticipated due to the complexity of current flow pathways arising from the complex interaction of the conductive properties influencing the ECa measurements and from the spatial heterogeneity of those conductive properties. The interpretation of ECa measurements is not trivial because of the complexity of current flow in the bulk soil. Numerous ECa studies have been conducted that have revealed the site specificity and complexity of geospatial ECa measurements with respect to the particular property or properties influencing the ECa measurement at the study site. Table 10-1 (taken from Corwin and Lesch 2005a) is a compilation of ECa studies and the associated dominant soil property or properties measured by ECa for each study. The advantages of the ECa measurement are that it is rapid, reliable, and easy to take, which have made it an ideal field measurement tool. However, because of the multiple pathways of conductance, it is often difficult to interpret. Corwin and Lesch (2003) provided guidelines for the use of ECa in agriculture by identifying the complexities of the ECa measurement

LABORATORY AND FIELD MEASUREMENTS

313

TABLE 10-1. Compilation of Literature Measuring ECa Categorized According to the Physicochemical and Soil-Related Properties either Directly or Indirectly Measured by ECa Soil Property

References

Directly Measured Soil Properties Salinity (and nutrients, e.g., NO 3)

Halvorson and Rhoades (1976); Rhoades et al. (1976); Rhoades and Halvorson (1977); de Jong et al. (1979); Cameron et al. (1981); Rhoades and Corwin (1981, 1990); Corwin and Rhoades (1982, 1984); Williams and Baker (1982); Greenhouse and Slaine (1983); van der Lelij (1983); Wollenhaupt et al. (1986); Williams and Hoey (1987); Corwin and Rhoades (1990); Rhoades et al. (1989b, 1990, 1999a, 1999b); Slavich and Petterson (1990); Diaz and Herrero (1992); Hendrickx et al. (1992); Lesch et al. (1992, 1995a, 1995b, 1998); Rhoades (1992, 1993); Cannon et al. (1994); Nettleton et al. (1994); Bennett and George (1995); Drommerhausen et al. (1995); Ranjan et al. (1995); Hanson and Kaita (1997); Johnston et al. (1997); Mankin et al. (1997); Eigenberg et al. (1998, 2002); Eigenberg and Nienaber (1998, 1999, 2001); Mankin and Karthikeyan (2002); Herrero et al. (2003); Paine (2003) ; Kaffka et al. (2005); Lesch et al. (2005); Sudduth et al. (2005)

Water content

Fitterman and Stewart (1986); Kean et al. (1987); Kachanoski et al. (1988, 1990); Vaughan et al. (1995); Sheets and Hendrickx (1995); Hanson and Kaita (1997); Khakural et al. (1998); Morgan et al. (2000); Freeland et al. (2001); Brevik and Fenton (2002); Wilson et al. (2002); Farahani et al. (2005); Kaffka et al. (2005); Lesch et al. (2005); Sudduth et al. (2005)

Texture-related (e.g., sand, clay, depth to claypans or sand layers)

Williams and Hoey (1987); Brus et al. (1992); Jaynes et al. (1993); Stroh et al. (1993); Sudduth and Kitchen (1993); Doolittle et al. (1994, 2002); Kitchen et al. (1996); Banton et al. (1997); Boettinger et al. (1997); Rhoades et al. (1999b); Scanlon et al. (1999); Inman et al. (2001); Triantafilis et al. (2001); Anderson-Cook et al. (2002); Brevik and Fenton (2002); Lesch et al. (2005); Sudduth et al. (2005); Triantafilis and Lesch (2005)

Bulk density-related (e.g., compaction)

Rhoades et al. (1999b); Gorucu et al. (2001) (continued)

314

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 10-1. Compilation of Literature Measuring ECa Categorized According to the Physicochemical and Soil-Related Properties either Directly or Indirectly Measured by ECa (Continued) Soil Property

References

Indirectly Measured Soil Properties Organic matter-related (including soil organic carbon, and organic chemical plumes)

Greenhouse and Slaine (1983, 1986); Brune and Doolittle (1990); Nyquist and Blair (1991); Jaynes (1996); Benson et al. (1997); Bowling et al. (1997); Brune et al. (1999); Nobes et al. (2000); Farahani et al. (2005); Sudduth et al. (2005)

Cation exchange capacity

McBride et al. (1990); Triantafilis et al. (2002); Farahani et al. (2005); Sudduth et al. (2005)

Leaching

Slavich and Petterson (1990); Corwin et al. (1999); Rhoades et al. (1999b); Lesch et al. (2005)

Groundwater recharge

Cook and Kilty (1992); Cook et al. (1992); Salama et al. (1994)

Herbicide partition coefficients

Jaynes et al. (1995)

Soil map unit boundaries

Fenton and Lauterbach (1999); Stroh et al. (2001)

Corn rootworm distributions

Ellsbury et al. (1999)

Soil drainage classes

Kravchenko et al. (2002)

From Corwin and Lesch (2005a) with permission from Elsevier.

and how to deal with them. As shown in Fig. 10-8, three parallel pathways of current flow contribute to the ECa measurement: (1) a liquid phase pathway (Pathway 1) via salts contained in the soil water occupying the large pores, (2) a solid pathway (Pathway 2) via soil particles that are in direct and continuous contact with one another, and (3) a solid-liquid pathway (Pathway 3) primarily via exchangeable cations associated with clay minerals (Rhoades et al. 1999b). To measure soil salinity, the EC of only the soil solution (Pathway 1) is required; consequently, ECa measures more than just soil salinity. In fact, ECa is a measure of anything conductive within the volume of measurement and is influenced, whether directly or indirectly, by any edaphic properties that affect bulk soil conductance. Because of the pathways of conductance, ECa is influenced by a complex interaction of edaphic properties including salinity, texture (or saturation percentage, SP), water content, bulk density (b), organic matter (OM), cation exchange capacity (CEC), clay mineralogy, and temperature. The SP and b are both directly influenced by clay content (or texture) and OM. Furthermore, the exchange surfaces on clays and OM provide a solid-liquid phase pathway primarily via exchangeable cations; conse-

LABORATORY AND FIELD MEASUREMENTS

315

FIGURE 10-8. Schematic showing the three conductance pathways of apparent soil electrical conductivity (ECa). Pathway 1  liquid phase conductance, Pathway 2  solid phase conductance, and Pathway 3  solid-liquid phase conductance. From Rhoades et al. (1989b). Reprinted with permission from the Soil Science Society of America. quently, clay type and content (or texture), CEC, and OM are recognized as factors influencing ECa measurements. Measurements of ECa must be interpreted with these influencing factors in mind. It is of paramount importance that the concept of parallel pathways of conductance is understood in order to interpret ECa measurements. Interpreting ECa measurements is accomplished best with ground-truth measurements of the soil physical and chemical properties that potentially influence ECa at the point of measurement. An understanding and interpretation of geospatial ECa data can only be obtained from ground-truth measures of soil properties that correlate with ECa from either a direct influence or indirect association. For this reason, geospatial ECa measurements are used as a surrogate of soil spatial variability to direct soil sampling when mapping soil salinity at field scales and larger spatial extents. They are not generally used as a direct measure of soil salinity, particularly at ECa 2 dS m1 where the influence of conductive soil properties other than salinity can have an increased influence on the ECa reading. At high ECa values, salinity is most likely dominating the ECa reading; consequently, geospatial ECa measurements are most likely mapping soil salinity.

316

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

METHODS OF FIELD-SCALE SOIL SALINITY MEASUREMENT Soil salinity is a dynamic soil property that is highly spatially and temporally variable. The dynamic nature of soil salinity makes mapping and monitoring of salinity a difficult challenge. Mapping and monitoring soil salinity at field scale requires a rapid, reliable, easy method of taking geospatial measurements. The use of soil samples to measure salinity (e.g., ECe, EC1⬊1, EC1⬊2, EC1⬊5, or ECp) at field scales is impractical because of the need for hundreds and even thousands of grid samples. The use of soil samples to measure salinity at field scales is only practical when sampling is directed to minimize the number of samples that reflect the range and variability of salinity within the area of study. This can be achieved using easily measured spatial information correlated to soil salinity as a means of directing where to take the fewest samples. Two potential sources of correlated spatial information used to direct where soil samples should be taken to measure ECe are: (1) visual crop observation, and (2) geospatial measurements of ECa with mobile ER or EMI equipment. Associated with visual crop observation but considered a distinct potential approach is the use of multi- and hyperspectral imagery. Even though the use of remote imagery has tremendous potential, at this point it is still restricted to research because the methodology has not been developed for general application to mapping and monitoring salinity. At present, only the use of geospatial measurements of ECa can provide reliable, accurate maps of salinity at field scales. Even so, remote imagery will unquestionably play a future role in mapping salinity, particularly at landscape scales. Visual Crop Observation Visual crop observation is a quick method, but it has the disadvantage that salinity development is detected after crop damage has occurred; consequently, crop yield must be sacrificed to locate areas of salinity development. Furthermore, decreases in crop yield are not necessarily the consequence of only salt accumulation. Crops respond to a variety of anthropogenic (e.g., irrigation uniformity, farm equipment traffic), edaphic (e.g., salinity, water content, texture, OM), biological (e.g., disease, nematodes), meteorological (e.g., precipitation, humidity, temperature), and topographical (e.g., slope, elevation, microrelief) factors, any of which can cause yield reduction. Because of the variety of factors influencing crop yield and quality, the use of visual crop observations to assess soil salinity is not definitive and can be extremely misleading. The least desirable method to measure salinity distributions in the field is visual observation because crop yields are reduced to obtain soil salin-

LABORATORY AND FIELD MEASUREMENTS

317

ity information, and the crop yield decrements may or may not be related to salinity. However, remote imagery is increasingly becoming a part of agriculture and potentially represents a quantitative approach to visual observation. Remote imagery may offer a potential for early detection of the onset of salinity damage to plants. The expectations for the use of multi- and hyperspectral imagery to map and monitor soil salinity as well as the spatial variability of other soil properties (e.g., water content, mineralogy, and others) is high and will no doubt prove fruitful as research continues in this area. Geospatial ECa Measurements Because of the quickness and ease with which geospatial measurements of ECa can be obtained and because ECa measures a variety of properties that potentially influence crop yield and quality (i.e., salinity, water content, texture, OM, bulk density), geospatial ECa measurements can serve as a surrogate to characterize the spatial variability of a variety of properties, particularly soil salinity (Corwin 2005). It has been hypothesized by Corwin and Lesch (2003, 2005b) that spatial ECa information can be used to develop a soil sampling plan that identifies sites reflecting the range and variability of soil salinity and/or other soil properties correlated with ECa. The use of geospatial ECa measurements to direct a soil sampling plan is referred to as ECa-directed soil sampling (Corwin 2005). This approach has been demonstrated for not only mapping salinity at field scale (Corwin et al. 2003a; Corwin and Lesch 2005c) but also for applications in (1) precision agriculture to define site-specific management units (Corwin 2005; Corwin et al. 2003b); (2) monitoring management-induced spatio-temporal changes due to degraded water reuse (Corwin et al. 2006); (3) characterizing soil spatial variability (Corwin 2005); and (4) modeling nonpoint source pollutants in the vadose zone (Corwin 2005; Corwin et al. 1999). Each of these applications uses ECadirected soil sampling to characterize the spatial variability of a soil property or properties of significance to the particular application. Electrical resistivity (e.g., Wenner array) and EMI are both well suited for field-scale applications because their volumes of measurement are large, which reduces the influence of local-scale variability. To obtain geospatial measurements, a mobile means of measuring ECa is essential. Mobile ECa equipment has been developed by a variety of researchers (Cannon et al. 1994; Carter et al. 1993; Freeland et al. 2002; Jaynes et al. 1993; Kitchen et al. 1996; McNeill 1992; Rhoades 1993). The development of mobile ECa measurement equipment has made it possible to produce ECa maps with measurements taken every few meters. Mobile ECa measurement equipment has been developed for both ER and EMI geophysical approaches.

318

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

By mounting the four ER electrodes to “fix” their spacing, considerable time for a measurement is saved. A tractor-mounted version of the “fixedelectrode array” has been developed that georeferences the ECa measurement with a GPS (Rhoades 1993). The mobile, fixed-electrode-array equipment is well suited for collecting detailed maps of the spatial variability of ECa at field scales and larger. Veris Technologies (2011) has developed a commercial mobile system for measuring ECa using the principles of ER, which uses the spacing of 6 coulter electrodes to measure ECa to depths of 0–30 and 0–91 cm (Fig. 10-9a).

FIGURE 10-9. Mobile apparent soil electrical conductivity (ECa) equipment: (a) Veris 3100 electrical resistivity rig, and (b) electromagnetic induction rig developed at the U.S. Salinity Laboratory with a close-up of the sled containing a dual-dipole Geonics EM-38.

LABORATORY AND FIELD MEASUREMENTS

319

Mobile EMI equipment developed at the U.S. Salinity Laboratory (Rhoades 1993) is available for appraisal of soil salinity and other soil properties (e.g., water content and clay content) using an EM-38. Recently, the mobile EMI equipment developed at the U.S. Salinity Laboratory was modified by the addition of a dual-dipole EM-38 unit (Fig. 10-9b) and DUALEM-2. The dual-dipole EM-38 conductivity meter simultaneously records data in both dipole orientations (horizontal and vertical) at time intervals of just a few seconds between readings. The mobile EMI equipment is suited for the detailed mapping of ECa and correlated soil properties at specified depth intervals through the rootzone. The advantage of the mobile dual-dipole EMI equipment over the mobile fixed-array resistivity equipment is that the EMI technique is noninvasive so it can be used in dry, frozen, or stony soils that would not be amenable to the invasive technique of the fixed-array approach due to the need for good electrode–soil contact. The disadvantage of the EMI approach would be that the ECa is a depth-weighted value that is nonlinear with depth McNeill (1980). Scientists at the U.S. Salinity Laboratory have developed an integrated system for the measurement of field-scale salinity consisting of (1) mobile ECa measurement equipment (Rhoades 1993), (2) protocols for ECadirected soil sampling (Corwin and Lesch 2005b), and (3) sample design software (Lesch et al. 2000). The integrated system for mapping soil salinity is schematically illustrated in Figure 10-10. The protocols of an ECa survey for measuring soil salinity at field scale include eight basic elements: (1) ECa survey design, (2) georeferenced ECa data collection, (3) soil sample design based on georeferenced ECa data, (4) soil sample collection, (5) physical and chemical analysis of pertinent soil properties, (6) spatial statistical analysis, (7) determination of the dominant soil properties influencing the ECa measurements at the study site, and (8) GIS development. The basic steps for each element are provided in Table 10-2. A detailed discussion of the protocols can be found in Corwin and Lesch (2005b). Corwin and Lesch (2005c) provide a case study demonstrating the use of the protocols. Arguably, the most significant element of the protocols is the ECa-directed soil sampling design, which warrants discussion. Soil Sample Design Based on Geospatial ECa Data Once a georeferenced ECa survey is conducted, the data are used to establish the locations of the soil core sample sites for (1) calibration of ECa to soil sample ECe and/or (2) delineation of the spatial distribution of soil properties correlated to ECa within the field surveyed. To establish the locations where soil cores are to be taken, either design-based or prediction-based (i.e., model-based) sampling schemes can be used. Design-

320

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 10-10. Schematic of the integrated system for mapping field-scale soil salinity as developed at the U.S. Salinity Laboratory. based sampling schemes have historically been the most commonly used and hence are more familiar to most research scientists. An excellent review of design-based methods can be found in Thompson (1992). Design-based methods include simple random sampling, stratified random sampling, multistage sampling, cluster sampling, and network sampling schemes. The use of unsupervised classification by Fraisse et al. (2001) and Johnson et al. (2001) is an example of design-based sampling. Prediction-based sampling schemes are less common, although significant statistical research has been recently performed in this area (Valliant et al. 2000). Prediction-based sampling approaches have been applied to the optimal collection of spatial data by Müller (2001); the specification of optimal designs for variogram estimation by Müller and Zimmerman (1999); the estimation of spatially referenced linear regression models by Lesch (2005) and Lesch et al. (1995b); and the estimation of geostatistical mixed linear models by Zhu and Stein (2006). Conceptually similar types of nonrandom sampling designs for variogram estimation have been introduced by Bogaert and Russo (1999), Russo (1984), and Warrick and Myers (1987). Both design-based and prediction-based sampling methods can be applied to geospatial ECa data to direct soil sampling as a means of characterizing soil spatial variability (Corwin and Lesch 2005b).

LABORATORY AND FIELD MEASUREMENTS

321

TABLE 10-2. Outline of Steps to Conduct an ECa Field Survey to Map Soil Salinity 1. Site description and ECa survey design a. Record site metadata. b. Define the project’s/survey’s objective. c. Establish site boundaries. d. Select GPS coordinate system. e. Establish ECa measurement intensity. 2. ECa data collection with mobile GPS-based equipment a. Georeference site boundaries and significant physical geographic features with GPS. b. Measure georeferenced ECa data at the predetermined spatial intensity and record associated metadata. 3. Soil sample design based on georeferenced ECa data a. Statistically analyze ECa data using an appropriate statistical sampling design to establish the soil sample site locations. b. Establish site locations, depth of sampling, sample depth increments, and number of cores per site. 4. Soil core sampling at specified sites designated by the sample design a. Obtain measurements of soil temperature through the profile at selected sites. b. At randomly selected locations, obtain duplicate soil cores within a 1-m distance of one another to establish local-scale variation of soil properties. c. Record soil core observations (e.g., mottling, horizonation, textural discontinuities). 5. Laboratory analysis of soil salinity and other ECa-correlated physical and chemical properties defined by project objectives 6. If needed, stochastic and/or deterministic calibration of ECa to ECe or to other soil properties (e.g., water content and texture) 7. Spatial statistical analysis to determine the soil properties influencing ECa a. Perform a basic statistical analysis of physical and chemical data, including soil salinity, by depth increment and by composite depth over the depth of measurement of ECa. b. Determine the correlation between ECa and salinity and between ECa and other soil properties by composite depth over the depth of measurement of ECa. 8. GIS database development and graphic display of spatial distribution of soil properties Modified from Corwin and Lesch (2005b) specifically for mapping soil salinity.

322

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The prediction-based sampling approach was introduced by Lesch et al. (1995b). This sampling approach attempts to optimize the estimation of a regression model, that is, minimize the mean square prediction error produced by the calibration function, while simultaneously ensuring that the independent regression model residual error assumption remains approximately valid. This, in turn, allows an ordinary regression model to be used to predict soil property levels at all remaining (i.e., nonsampled) conductivity survey sites. The basis for this sampling approach stems directly from traditional response-surface sampling methodology (Box and Draper 1987). There are two main advantages to the response-surface approach. First, a substantial reduction in the number of samples required for effectively estimating a calibration function can be achieved, in comparison to more traditional design-based sampling schemes. Second, this approach lends itself naturally to the analysis of ECa data. Indeed, many types of ground-, airborne-, and/or satellite-based remotely sensed data are often collected specifically because one expects these data to correlate strongly with some parameter of interest (e.g., crop stress, soil type, soil salinity), but the exact parameter estimates (associated with the calibration model) may still need to be determined via some type of site-specific sampling design. The response-surface approach explicitly optimizes this site selection process. A user-friendly software package (ESAP) developed by Lesch et al. (2000), which uses a response-surface sampling design, has proven to be particularly effective in delineating spatial distributions of soil properties from ECa survey data (Corwin 2005; Corwin et al. 2003a,b, 2006; Corwin and Lesch 2003, 2005c). The ESAP software package identifies the optimal locations for soil sample sites from the ECa survey data. These sites are selected based on spatial statistics to reflect the observed spatial variability in ECa survey measurements. Generally, 6 to 20 sites are selected depending on the level of variability of the ECa measurements for a site. The optimal locations of a minimal subset of ECa survey sites are identified to obtain soil samples. Once the number and location of the sample sites have been established, the depth of soil core sampling, sample depth increments, and number of sites where duplicate or replicate core samples should be taken are established. The depth of sampling should be the same at each sample site and should extend over the depth of penetration by the ECa-measurement equipment used. For instance, the Geonics EM-38 measures to a depth of roughly 0.75 to 1.0 m in the horizontal coil configuration (EMh), and 1.2 1.5 m in the vertical coil configuration (EMv). Sample depth increments are flexible and depend to a great extent on the study objectives. A depth increment of 0.3 m has been commonly used at the U.S. Salinity Laboratory because it provides sufficient soil profile information over the

LABORATORY AND FIELD MEASUREMENTS

323

rootzone (0–1.2 to 1.5 m) for statistical analysis without an overly burdensome number of samples to conduct physical and chemical analyses. Depth increments should be the same from one sample site to the next. The number of duplicates or replicates taken at each sample site is determined by the desired accuracy for characterizing soil properties and the need for establishing the level of local-scale variability at the site. Duplicates or replicates are not necessarily needed at every sample site to establish local-scale variability. Considerations when Conducting an ECa Survey A number of considerations must be heeded when conducting a geospatial ECa survey to map soil salinity. Each of these considerations can influence the ECa measurement, leading to a potential misinterpretation of the salinity distribution. These considerations account for temporal, moisture, surface roughness, and surface geometry effects. Temporal comparisons of geospatial ECa measurements to determine spatio-temporal changes in salinity patterns of distribution can only be made from ECa survey data that have been obtained under similar watercontent and temperature conditions. Surveys of ECa should be conducted when the water content is at or near field capacity and the soil profile temperatures are similar. For irrigated fields, ECa surveys should be conducted roughly two to four days after an irrigation, or longer if the soil is high in clay content and additional time is needed for the soil to drain to field capacity. For dryland farming, the survey should occur two to four days after a substantial rainfall, or longer, depending on soil texture. The effects of temperature can be addressed by taking soil profile temperatures at the time of the ECa survey and temperature-correcting the ECa measurements, or by conducting the surveys roughly at the same time during the year so that the temperature profiles are the same for each survey. The type of irrigation used can influence the within-field spatial distribution of water content and should be kept in mind as a factor influencing ECa spatial patterns. Sprinkler irrigation has a high level of application uniformity, whereas flood irrigation and drip irrigation are highly nonuniform. In general, flood irrigation results in higher water contents and overleaching at the “head” end of the field, while underleaching and lower water contents can occur at the “tail” end of the field. This general across-the-field trend is observed for both flood irrigation with basins and flood irrigation with beds and furrows, but beds and furrows introduce an added level of localized complexity. Flood irrigation with beds and furrows results in localized variations in water content, with high water contents and greater leaching occurring under the furrows while beds will typically show lower water contents and accumulations of salinity.

324

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The presence or absence of beds and furrows is a significant factor during a geospatial ECa survey. Measurements taken in furrows will differ from measurements taken in beds due to water flow and salt accumulation patterns. In addition, the physical presence of the bed influences the conductivity pathways, particularly when using EMI. These surface geometry effects are in addition to the effects of moisture and salinity distribution patterns that are present in beds and furrows. To assess salinity in a bed-furrow irrigated field, it is probably best to take the ECa measurements in the bed. Above all, the ECa measurements must be consistent, either entirely in the furrow or entirely in the bed. Surveys of drip-irrigated fields are even more complicated than ECa surveys of bed-furrow irrigated fields. Drip irrigation produces complex local- and field-scale three-dimensional patterns of water content and salinity that are particularly difficult to spatially characterize with geospatial ECa measurements (or any salinity measurement technique, for that matter). The easiest approach is to run ECa transects both over and between drip lines to capture the local-scale variation. The roughness of the soil surface can also influence spatial ECa measurements. Geospatial ECa measurements taken on a smooth field surface will be higher than the same field with a rough surface from disking. This is due to the fact that the disturbed disked soil acts as an insulated layer to the conductance pathways, thereby reducing its conductance. When conducting a geospatial ECa survey of a field, the entire field must have the same surface roughness. These factors, if not taken into account when conducting an ECa survey, will likely produce a “banding” effect. For example, if an ECa survey is conducted on a field that has areal differences in water content, soil profile temperature, surface roughness, and surface geometry, then bands of ECa such as those found in Fig. 10-11 will result. These bands reflect the variations in soil moisture, temperature, roughness, and surface geometry, which must be uniform across a field to produce a reliable ECa survey that can be used to direct soil sampling to spatially characterize the distribution of salinity.

USE OF REMOTE IMAGERY FOR MEASURING SOIL SALINITY AT FIELD AND LANDSCAPE SCALES While field-based measurements of soil salinity have progressed greatly over the past decades, they remain limited to mapping soil salinity over a small number of fields in a single day. Assessments of soil salinity across entire landscapes and through time are therefore difficult and expensive to conduct with field-based approaches alone. Remote sensing instruments aboard airplanes or satellites routinely acquire measure-

LABORATORY AND FIELD MEASUREMENTS

325

FIGURE 10-11. A poorly designed apparent soil electrical conductivity (ECa) survey showing the banding that occurs when surveys are conducted at different times under varying water contents, temperatures, surface roughnesses, and surface geometry conditions. ments of energy reflected or emitted from the land surface across wide swaths of land, thus presenting an opportunity for low-cost mapping of salinity at broad scales. Unfortunately, in our estimation, efforts to relate remote sensing data to soil salinity have achieved limited success. Most methods have been highly empirical in nature, and empirical relationships successful in one case have tended to break down when applied to data from different

326

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

scales, years, or locations. This is especially true in the case of mapping salinity at slight to moderate levels (ECe  2 to 8 dS m1). Here we provide a selective review of past work and highlight the approaches we believe are most promising for the future. More exhaustive reviews of remote sensing techniques for soil salinity can be found in Metternicht and Zinck (2003) and Mougenot et al. (1993) As with ECa, remote sensing measurements are influenced by a range of land surface properties, with soil salinity representing only one of these factors. The overall challenge is to find some measure that is sensitive to soil salinity but insensitive to other factors that vary in the landscape. This measure may be reflectance or emittance at a particular wavelength, or a strategic combination of measurements made at different wavelengths, dates, or locations. Importantly, the appropriate measure may depend on the aspect of soil salinity that is of interest. For example, reflectance from a soil surface is affected by salinity only in the upper few centimeters of soil, which may not be representative of average salinity at greater depths. In contrast, reflectance from plant canopies can provide information on soil salinity throughout the rootzone. Much of the work on remote sensing of salinity has been done in the two major agricultural regions affected by salinity: the irrigated systems of India and Pakistan and the rain-fed systems of Australia. Early work relied heavily on visual interpretation of aerial photos or Landsat satellite images. Verma et al. (1994) observed that remote indicators of canopy biomass [Landsat red and near-infrared (NIR) reflectance] during the peak of the cropping season in the Indo-Gangetic Plains successfully distinguished barren saline soils from healthy crop land. A Landsat thermal image was then used to separate saline fields from fallow fields with sandy soils, which had similarly bright reflectance values but lower soil moisture levels. Many similar studies have been conducted throughout India that rely mainly on a lack of vegetation on salt-affected soils (IDNP 2002; Sharma et al. 2000). Other studies have used images acquired prior to the growing season, when white salt crusts on the surface of saline soils are significantly brighter than nonsaline soils (IDNP 2002). Both of these approaches can be quite useful for mapping severely saline soils (ECe  25 dS m1), but are problematic for less severe cases that are not marked by salt crusts and barren land. In general, an important factor in evaluating any study is the range of salinity values sampled. For example, a high model R2 can be driven by a few points above 20 dS m1, even though the model’s predictive power at lower levels is poor. The more challenging problem of mapping slight to moderate salinity has been approached in several ways. A common method has been to use the health of crop condition as an indicator of soil salinity. In aerial photos

LABORATORY AND FIELD MEASUREMENTS

327

taken with color infrared film, dense vegetation appears bright red, saline soils appear bright white, and fields with sparse canopies appear pinkish. Multispectral satellite data can be similarly displayed and interpreted visually. More quantitative measures can also be used, such as the normalized difference vegetation index (NDVI) based on NIR and red reflectance: NDVI  (NIR  Red)/(NIR  Red). For example, Wiegand et al. (1994, 1996) found strong linear relationships between NDVI and rootzone salinity in salt-affected cotton and sugarcane fields. However, such simple relationships are only likely to exist in cases where salinity is the major factor responsible for variability in crop yield. Landscapes with only slight to moderate salinity are likely to have many other factors, such as field management, that affect yields as much or more than salinity. Extrapolation of relationships within a small number of salt-affected fields to an entire landscape can therefore result in large errors. To address this problem, some have proposed using average crop yields over a number of years to filter out noise from nonsoil factors, which tend to vary from year to year. Lobell et al. (2007) found very weak relationships between salinity and yields in individual years in the Colorado River delta region of Mexico, but much stronger correspondence between salinity and maximum yield over a six-year period. In Australia, Furby et al. (1995) reported large commission errors for a classification of saline soils when using a single year of image data, because many areas of poor crop condition were incorrectly labeled as saline. These errors of commission were reduced from 20% to 2% by the addition of a second year of Landsat data. Another common approach is to estimate salinity from soil reflectance measured when the surface is bare. These methods rely on the bright visible reflectance of surface salts, several characteristic absorption features at longer wavelengths, or both (Ben-Dor et al. 2002; Csillag et al. 1993; Dehaan and Taylor 2002). However, because soil reflectance can vary greatly due to spatially and temporally variable moisture or surface roughness conditions, these techniques often result in poor accuracies when applied outside of the calibration dataset. As mentioned, soil salinity at the surface can also correlate poorly with average rootzone salinity. A less-developed but promising approach is to exploit the spatial dimension of remote sensing data. Because salts tend to be spatially heterogeneous, saline fields may be identified by a high standard deviation of NDVI within fields (Metternicht and Zinck 1997). This approach requires relatively high spatial resolution imagery, accurate information on field boundaries, and a relatively low contribution of other factors to within-field heterogeneity.

328

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Overall, no single remote sensing approach has proven particularly effective for mapping salinity at low to moderate levels. Therefore, the most successful approaches are likely to combine information from a variety of sources, including multiple remote sensing measures as well as several nonremote indicators such as landscape position, soil type, and topography (Furby et al. 1995; Metternicht 2001; Tweed et al. 2007). As with any spatial prediction problem, the use of independent validation data will also be critical to evaluating and improving salinity estimates. For example, simply extrapolating local empirical relationships to estimate regional totals (Madrigal et al. 2003) should be avoided. As Furby et al. (1995) demonstrate, reserving a significant fraction (in their case, half) of sites for independent validation can help to identify shortcomings in the original algorithms and suggest improvements. Another important methodological consideration for regional mapping is that sites should be selected at random and not preferentially in saline areas. Table 10-3 presents a summary of elements that are most likely, in our opinion, to result in successful salinity mapping with remote sensing at landscape scales. Recently, Lobell et al. (2010) published a successful regional-scale salinity assessment of 284,000 ha using these recommended elements. TABLE 10-3. Some Elements Key to Successful Remote Sensing of Salinity at Landscape Scales Element

Well-timed image acquisition

Randomly selected training sites

Independent validation data Multiple years of images

Ancillary data

Comment

Images should be selected, if possible, from end of dry season for methods based on soil reflectance, or from peak of growing season for methods based on crop canopy reflectance. A bias of training sites toward highsalinity fields will likely result in an overestimation of regional salinity levels. Prediction errors for test data can be much larger than training errors. Nonsoil factors can heavily influence reflectance in any one year, but will tend to average out over multiple years. Combining remote sensing with other GIS data (soil texture, topography, etc.) can greatly improve model accuracy.

LABORATORY AND FIELD MEASUREMENTS

329

SUMMARY The various methods of measuring/estimating soil salinity have pros and cons: • While precise and reliable, directly measuring the aqueous extract of soil samples in the laboratory is labor-intensive and costly. • The use of soil samples to measure salinity at field scales is only effective if sampling is directed using an easy-to-take surrogate measurement, such as apparent soil electrical conductivity (ECa), to minimize the number of samples. • The sample volumes of soil solution extractors and soil salinity sensors are small, which affects their ability to provide data representative of field conditions. • Measurements of apparent soil conductivity (ECa) can be made based on electrical resistivity (ER), electromagnetic induction (EMI), and time-domain reflectometry (TDR). In general, when measuring ECa, it is important to take into consideration the multiple pathways of electrical conductivity in the bulk soil; consequently, ECa may be affected by salinity, texture, water content, bulk density, organic matter in the soil, cation exchange capacity, clay mineralogy, and soil temperature. • For field-scale salinity measurement, a systematic ECa-directed sampling approach is required that minimizes the primary influences of soil property effects (such as water content, texture, bulk density, and soil temperature) and avoids the confounding secondary influences of soil condition effects (such as surface roughness, presence or absences of beds and furrows, ambient air temperature effects on the instrumentation, and compaction) to reliably measure the target property of soil salinity. • Remote sensing techniques are an experimental approach to mapping soil salinity over regional scales with tremendous potential, but better correlations between energy strength and spectrum and field conditions are needed before the technique is reliable. The technique for measuring and mapping soil salinity at field scale with ECa-directed soil sampling is well understood, and an eight-step protocol is outlined in Table 10-2. REFERENCES Amoozegar-Fard, A., Nielsen, D. R., and Warrick, A. W. (1982). “Soil solute concentration distributions for spatially varying pore water velocities and apparent diffusion coefficients.” Soil Sci. Soc. Am. J., 46, 3–9.

330

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Anderson-Cook, C. M., Alley, M. M., Roygard, J. K. F., Khosla, R., Noble, R. B., and Doolittle, J. A. (2002). “Differentiating soil types using electromagnetic conductivity and crop yield maps.” Soil Sci. Soc. Am. J., 66, 1562–1570. Banton, O., Seguin, M. K., and Cimon, M. A. (1997). “Mapping field-scale physical properties of soil with electrical resistivity.” Soil Sci. Soc. Am. J., 61(4), 1010–1017. Barnes, H. E. (1952). “Soil investigation employing a new method of layer-value determination for earth resistivity interpretation.” Highway Res. Board Bull., 65, 26–36. Ben-Dor, E., Patkin, K., Banin, A., and Karnieli, A. (2002). “Mapping of several soil properties using DAIS-7915 hyperspectral scanner data: A case study over clayey soils in Israel.” Int. J. Remote Sens., 23, 1043–1062. Bennett, D. L., and George, R. J. (1995). “Using the EM38 to measure the effect of soil salinity on Eucalyptus globulus in south-western Australia.” Agric. Water Manage., 27, 69–86. Benson, A. K., Payne, K. L., and Stubben, M. A. (1997). “Mapping groundwater contamination using DC resistivity and VLF geophysical methods: A case study.” Geophysics, 62(1), 80–86. Biggar, J. W., and Nielsen, D. R. (1976). “Spatial variability of the leaching characteristics of a field soil.” Water Resour. Res., 12, 78–84. Boettinger, J. L., Doolittle, J. A., West, N. E., Bork, E. W., and Schupp, E. W. (1997). “Nondestructive assessment of rangeland soil depth to petrocalcic horizon using electromagnetic induction.” Arid Soil Res. Rehabil., 11(4), 372–390. Bogaert, P., and Russo, D. (1999). “Optimal spatial sampling design for the estimation of the variogram based on a least squares approach.” Water Resour. Res., 35, 1275–1289. Bowling, S. D., Schulte, D. D., and Woldt, W. E. (1997). A geophysical and geostatistical methodology for evaluating potential subsurface contamination from feedlot runoff retention ponds, ASAE Paper No. 972087, 1997 ASAE Winter Meetings, December 1997, Chicago, ASAE, St. Joseph, Mich. Box, G. E. P., and Draper, N. R. (1987). Empirical model-building and response surfaces, John Wiley and Sons, New York. Bresler, E., McNeal, B. L., and Carter, D. L. (1982). Saline and sodic soils, SpringerVerlag, New York, 174–181. Brevik, E. C., and Fenton, T. E. (2002). “The relative influence of soil water, clay, temperature, and carbonate minerals on soil electrical conductivity readings taken with an EM-38 along a Mollisol catena in central Iowa.” Soil Surv. Horiz., 43, 9–13. Brune, D. E., and Doolittle, J. (1990). “Locating lagoon seepage with radar and electromagnetic survey.” Environ. Geol. Water Sci., 16, 195–207. Brune, D. E., Drapcho, C. M., Radcliff, D. E., Harter, T., and Zhang, R. (1999). Electromagnetic survey to rapidly assess water quality in agricultural watersheds, ASAE Paper No. 992176, ASAE, St. Joseph, Mich. Brus, D. J., Knotters, M., van Dooremolen, W. A., van Kernebeek, P., and van Seeters, R. J. M. (1992). “The use of electromagnetic measurements of apparent soil electrical conductivity to predict the boulder clay depth.” Geoderma, 55, 79–93. Burger, H. R. (1992). Exploration geophysics of the shallow subsurface, Prentice Hall, Upper Saddle River, N.J.

LABORATORY AND FIELD MEASUREMENTS

331

Cameron, D. R., de Jong, E., Read, D. W. L., and Oosterveld, M. (1981). “Mapping salinity using resistivity and electromagnetic inductive techniques.” Can. J. Soil Sci., 61, 67–78. Cannon, M. E., McKenzie, R. C., and Lachapelle, G. (1994). “Soil-salinity mapping with electromagnetic induction and satellite-based navigation methods.” Can. J. Soil Sci., 74(3), 335–343. Carter, L. M., Rhoades, J. D., and Chesson, J. H. (1993). Mechanization of soil salinity assessment for mapping, ASAE Paper No. 931557, 1993 ASAE Winter Meetings, December 12–17, 1993, Chicago, ASAE, St. Joseph, Mich. Cook, P. G., and Kilty, S. (1992). “A helicopter-borne electromagnetic survey to delineate groundwater recharge rates.” Water Resour. Res., 2 (11), 2953–2961. Cook, P. G., Walker, G. R., Buselli, G., Potts, I., and Dodds, A. R. (1992). “The application of electromagnetic techniques to groundwater recharge investigations.” J. Hydrol., 130, 201–229. Corwin, D. L. (2002a). “Solute content and concentration—Measurement of solute concentration using soil water extraction: Suction cups,” in Methods of soil analysis. Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., SSSA Book Series No. 5. SSSA, Madison, Wisc., 1261–1268. ———. (2002b). “Solute content and concentration—Measurement of solute concentration using soil water extraction: Porous matrix sensors,” in Methods of soil analysis. Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., SSSA Book Series No. 5. SSSA, Madison, Wisc., 1269–1273. ———. (2005). “Geospatial measurements of apparent soil electrical conductivity for characterizing spatial variability,” in Soil-water-solute process characterization: An integrated approach, J. Alvarez-Benedi and R. Munoz-Carpena, eds., CRC Press, Boca Raton, Fla., 639–672. Corwin, D. L., Carrillo, M. L. K., Vaughan, P. J., Rhoades, J. D., and Cone, D. G. (1999). “Evaluation of GIS-linked model of salt loading to groundwater.” J. Environ. Qual., 28, 471–480. Corwin, D. L., and Hendrickx, J. M. H. (2002). “Solute content and concentration— Indirect measurement of solute concentration: Electrical resistivity—Wenner array,” in Methods of soil analysis. Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., Agronomy Monograph No. 9, SSSA, Madison, Wisc., 1282–1287. Corwin D. L., Kaffka, S. R., Hopmans, J. W., Mori, Y., Lesch, S. M., and Oster, J. D. (2003a). “Assessment and field-scale mapping of soil quality properties of a saline-sodic soil.” Geoderma, 114, 231–259. Corwin, D. L., and Lesch, S. M. (2003). “Application of soil electrical conductivity to precision agriculture: Theory, principles, and guidelines.” Agron. J., 95, 455–471. ———. (2005a). “Apparent soil electrical conductivity measurements in agriculture.” Comput. Electron. Agric., 46, 11–43. ———. (2005b). “Characterizing soil spatial variability with apparent soil electrical conductivity I. Survey protocols.” Comput. Electron. Agric., 46, 103–133. ———. (2005c). “Characterizing soil spatial variability with apparent soil electrical conductivity II. Case study.” Comput. Electron. Agric., 46, 135–152. Corwin, D. L., Lesch, S. M., Oster, J. D., and Kaffka, S. R. (2006). “Monitoring management-induced spatio-temporal changes in soil quality through soil sampling directed by apparent electrical conductivity.” Geoderma, 131, 369–387.

332

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Corwin D. L., Lesch, S. M., Shouse, P. J., Soppe, R., and Ayars, J. E. (2003b). “ Identifying soil properties that influence cotton yield using soil sampling directed by apparent soil electrical conductivity.” Agron. J., 95, 352–364. Corwin, D. L., and Rhoades, J. D. (1982). “An improved technique for determining soil electrical conductivity-depth relations from above-ground electromagnetic measurements.” Soil Sci. Soc. Am. J., 46, 517–520. ———. (1984). “Measurement of inverted electrical conductivity profiles using electromagnetic induction.” Soil Sci. Soc. Am. J., 48, 288–291. ———. (1990). “Establishing soil electrical conductivity: Depth relations from electromagnetic induction measurements.” Commun. Soil Sci. Plant Anal., 21(11– 12), 861–901. Csillag, F., Pasztor, L., and Biehl, L. L. (1993). “Spectral band selection for the characterization of salinity status of soils.” Remote Sens. Environ., 43, 231–242. Dalton, F. N. (1992). “Development of time domain reflectometry for measuring soil water content and bulk soil electrical conductivity,” in Advances in measurement of soil physical properties: Bringing theory into practice, G. C. Topp, W. D. Reynolds, and R. E. Green, eds., SSSA Special Publication 30, SSSA, Madison, Wisc., 143–167. Dalton, F. N., Herkelrath, W. N., Rawlins, D. S., and Rhoades, J. D. (1984). “Timedomain reflectometry: Simultaneous measurement of soil water content and electrical conductivity with a single probe.” Science, 224, 989–990. Dane, J. H., and Topp, G. C. (eds.). (2002). Methods of soil analysis: Part 4: Physical methods. SSSA Book Series 5. SSSA, Madison, Wisc. Dehaan, R. L., and Taylor, G. R. (2002). “Field-derived spectra of salinized soils and vegetation as indicators of irrigation-induced soil salinization.” Remote Sens. Environ., 80, 406–417. de Jong, E., Ballantyne, A. K., Caneron, D. R., and Read, D. W. (1979). “Measurement of apparent electrical conductivity of soils by an electromagnetic induction probe to aid salinity surveys.” Soil Sci. Soc. Am. J., 43, 810–812. Diaz, L., and Herrero, J. (1992). “Salinity estimates in irrigated soils using electromagnetic induction.” Soil Sci., 154, 151–157. Dobrin, M. B. (1960). Introduction to geophysical prospecting, McGraw-Hill, New York. Doolittle, J. A., Indorante, S. J., Potter, D. K., Hefner, S. G., and McCauley, W. M. (2002). “Comparing three geophysical tools for locating sand blows in alluvial soils of southeast Missouri.” J. Soil Water Conserv., 57(3), 175–182. Doolittle, J. A., Sudduth, K. A., Kitchen, N. R., and Indorante, S. J. (1994). “Estimating depths to claypans using electromagnetic induction methods.” J. Soil Water Conserv., 49(6), 572–575. Drommerhausen, D. J., Radcliffe, D. E., Brune, D. E., and Gunter, H. D. (1995). “Electromagnetic conductivity surveys of dairies for groundwater nitrate.” J. Environ. Qual., 24, 1083–1091. Dualem. (2007). “DUALEM-2,” Dualem Inc., Milton, Ontario, Canada, www.dualem.com/documents.html, accessed June 2007. Eigenberg, R. A., Doran, J. W., Nienaber, J. A., Ferguson, R. B., and Woodbury, B. L. (2002). “Electrical conductivity monitoring of soil condition and available N with animal manure and a cover crop.” Agric. Ecosyst. Environ., 88, 183–193.

LABORATORY AND FIELD MEASUREMENTS

333

Eigenberg, R. A., Korthals, R. L., and Neinaber, J. A. (1998). “Geophysical electromagnetic survey methods applied to agricultural waste sites.” J. Environ. Qual., 27, 215–219. Eigenberg, R. A., and Nienaber, J. A. (1998). “Electromagnetic survey of cornfield with repeated manure applications.” J. Environ. Qual., 27, 1511–1515. ———. (1999). Soil conductivity map differences for monitoring temporal changes in an agronomic field, ASAE Paper No. 992176, ASAE, St. Joseph, Mich. ———. (2001). Identification of nutrient distribution at abandoned livestock manure handling site using electromagnetic induction, ASAE Paper No. 012193 ASAE Annual International Meeting, July 30–August 1, 2001, Sacramento, California, ASAE, St. Joseph, Mich. Ellsbury, M. M., Woodson, W. D., Malo, D. D., Clay, D. E., Carlson, C. G., and Clay, S. A. (1999). “Spatial variability in corn rootworm distribution in relation to spatially variable soil factors and crop condition,” in Proc., 4th Int. Conf. on Precision Agriculture, St. Paul, Minnesota, July 19–22, 1998, P. C. Robert, R. H. Rust, and W. E. Larson, eds., ASA-CSSA-SSSA, Madison, Wisc., 523–533. England, C. B. (1974). “Comments on ’A technique using porous cups for water sampling at any depth in the unsaturated zone,’ by W. W. Wood.” Water Resour. Res., 10, 1049. Farahani, H. J., Buchleiter, G. W., and Brodahl, M. K. (2005). “Characterization of soil electrical conductivity variability in irrigated sandy and non-saline fields in Colorado.” Trans. ASAE, 48(1), 155–168. Fenton, T. E., and Lauterbach, M. A. (1999). “Soil map unit composition and scale of mapping related to interpretations for precision soil and crop management in Iowa,” in Proc., 4th In. Conf. on Precision Agriculture, St. Paul, Minnesota, July 19–22, 1998, P. C. Robert, R. H. Rust, and W. E. Larson, eds., ASA-CSSA-SSSA, Madison, Wisc., 239–251. Fitterman, D. V., and Stewart, M. T. (1986). “Transient electromagnetic sounding for groundwater.” Geophysics, 51, 995–1005. Fraisse, C. W., Sudduth, K. A., and Kitchen, N. R. (2001). “Delineation of site-specific management zones by unsupervised classification of topographic attributes and soil electrical conductivity.” Trans. ASAE, 44 (1), 155–166. Freeland, R. S., Branson, J. L., Ammons, J. T., and Leonard, L.L. (2001). “Surveying perched water on anthropogenic soils using non-intrusive imagery.” Trans. ASAE, 44, 1955–1963. Freeland, R. S., Yoder, R. E., Ammons, J. T., and Leonard, L. L. (2002). “Mobilizing surveying of soil conductivity using electromagnetic induction.” Appl. Eng. Agric., 18(1), 121–126. Furby, S. L., Wallace, J. F., Caccetta, P., and Wheaton, G. A. (1995). “Detecting and monitoring salt-affected land.” CSIRO Mathematical and Information Sciences, www.cmis.csiro.au/rsm/research/salmapmon/salmapmon.html, accessed July 4, 2007. Giese, K., and Tiemann, R. (1975). “Determination of the complex permittivity from thin-sample time domain reflectometry: Improved analysis of the step response waveform.” Adv. Mol. Relax. Processes, 7, 45–59. Gorucu, S., Khalilian, A., Han, Y. J., Dodd, R. B., Wolak, F. J., and Keskin, M. (2001). Variable depth tillage based on geo-referenced soil compaction data in coastal plain region of South Carolina, ASAE Paper No. 011016, 2001 ASAE Annual

334

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

International Meeting, July 30–August 1, 2001, Sacramento, California, ASAE, St. Joseph, Mich. Greenhouse, J. P., and Slaine, D. D. (1983). The use of reconnaissance electromagnetic methods to map contaminant migration.” Ground Water Monit. Rev., 3(2), 47–59. ———. (1986). “Geophysical modelling and mapping of contaminated groundwater around three waste disposal sites in southern Ontario.” Can. Geotech. J., 23, 372–384. Haines, B. L., Waide, J. B., and Todd, R. L. (1982). “Soil solution nutrient concentrations sampled with tension and zero-tension lysimeters: Report of discrepancies.” Soil Sci. Soc. Am. J., 46, 658–661. Halvorson, A. D., and Rhoades, J. D. (1976). “Field mapping soil conductivity to delineate dryland seeps with four-electrode techniques.” Soil Sci. Soc. Am. J., 44, 571–575. Hanson, B. R., and Kaita, K. (1997). “Response of electromagnetic conductivity meter to soil salinity and soil-water content.” J. Irrig. Drain. Eng., 123, 141–143. Harned, H. S., and Owen, B. B. (1958). The physical chemistry of electrolyte solutions, Reinhold Publishing Corp., New York. Hart, G. L., and Lowery, B. (1997). “Axial-radial influence of porous cup soil solution samplers in a sandy soil.” Soil Sci. Soc. Am. J., 61, 1765–1773. Heimovaara, T. J., Focke, A. G., Bouten, W., and Verstraten, J. M. (1995). “Assessing temporal variations in soil water composition with time domain reflectometry.” Soil Sci. Soc. Am. J., 59, 689–698. Hendrickx, J. M. H., Baerends, B., Raza, Z. I., Sadig, M., and Chaudhry, M. A. (1992). “Soil salinity assessment by electromagnetic induction of irrigated land.” Soil Sci. Soc. Am. J., 56, 1933–1941. Hendrickx, J. M. H., and Kachanoski, R. G. (2002). “Solute content and concentration: Indirect measurement of solute concentration: Nonintrusive electromagnetic induction,” in Methods of soil analysis. Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., Agronomy Monograph No. 9, SSSA, Madison, Wisc., 1297–1306. Herrero, J., Ba, A. A., and Aragues, R. (2003). “Soil salinity and its distribution determined by soil sampling and electromagnetic techniques.” Soil Use Mgmt., 19(2), 119–126. Indo-Dutch Network Project (IDNP). (2002). Indo-Dutch network project. A methodology for identification of waterlogging and soil salinity conditions using remote sensing, CCSRI, Karnal, India and Alterra-ILRI, Wageningen, The Netherlands. Inman, D. J., Freeland, R .S., Yoder, R. E., Ammons, J. T., and Leonard, L. L. (2001). “Evaluating GPR and EMI for morphological studies of loessial soils.” Soil Sci., 166(9), 622–630. Jacobsen, O. H., and Schjønning, P. (1993). “A laboratory calibration of time domain reflectometry for soil water measurements including effects of bulk density and texture.” J. Hydrol. 151, 147–157. Jaynes, D. B. (1996). “Mapping the areal distribution of soil parameters with geophysical techniques,” in Applications of GIS to the modeling of non-point source pollutants in the vadose zone, SSSA Special Publication No. 48, D. L. Corwin and K. Loague, eds., SSSA, Madison, Wisc., 205–216.

LABORATORY AND FIELD MEASUREMENTS

335

Jaynes, D. B., Colvin, T. S., and Ambuel, J. (1993). Soil type and crop yield determinations from ground conductivity surveys, ASAE Paper No. 933552, 1993 ASAE Winter Meetings, December 14–17, 1993, Chicago, ASAE, St. Joseph, Mich. Jaynes, D. B., Novak, J. M., Moorman, T. B., and Cambardella, C. A. (1995). “Estimating herbicide partition coefficients from electromagnetic induction measurements.” J. Environ. Qual., 24, 36–41. Johnson, C. K., Doran, J. W., Duke, H. R., Weinhold, B. J., Eskridge, K. M., and Shanahan, J. F. (2001). “Field-scale electrical conductivity mapping for delineating soil condition.” Soil Sci. Soc. Am. J., 65, 1829–1837. Johnston, M. A., Savage, M. J., Moolman, J. H., and du Pleiss, H. M. (1997). “Evaluation of calibration methods for interpreting soil salinity from electromagnetic induction measurements.” Soil Sci. Soc. Am. J., 61, 1627–1633. Kachanoski, R. G., de Jong, E., and Van-Wesenbeeck, I. J. (1990). “Field scale patterns of soil water storage from non-contacting measurements of bulk electrical conductivity.” Can. J. Soil Sci., 70, 537–541. Kachanoski, R. G., Gregorich, E. G., and Van-Wesenbeeck, I. J. (1988). “Estimating spatial variations of soil water content using noncontacting electromagnetic inductive methods.” Can. J. Soil Sci., 68, 715–722. Kaffka, S. R., Lesch, S. M., Bali, K. M., and Corwin, D. L. (2005). “Site-specific management in salt-affected sugar beet fields using electromagnetic inductive methods.” Comput. Electron. Agric., 46, 329–350. Kean, W. F., Jennings Walker, M., and Layson, H. R. (1987). “Monitoring moisture migration in the vadose zone with resistivity.” Ground Water, 25, 562–571. Khakural, B. R., Robert, P. C., and Hugins, D. R. (1998). “Use of non-contacting electromagnetic inductive method for estimating soil moisture across a landscape.” Commun. Soil Sci. Plant Anal., 29, 2055–2065. Kitchen, N. R., Sudduth, K. A., and Drummond, S. T. (1996). “Mapping of sand deposition from 1993 Midwest floods with electromagnetic induction measurements.” J. Soil Water Conserv., 51(4), 336–340. Kravchenko, A. N., Bollero, G. A., Omonode, R. A., and Bullock, D. G. (2002). “Quantitative mapping of soil drainage classes using topographical data and soil electrical conductivity.” Soil Sci. Soc. Am. J., 66, 235–243. Lesch, S. M. (2005). “Sensor-directed response surface sampling designs for characterizing spatial variation in soil properties.” Comput. Electron. Agric., 46 (1–3), 153–179. Lesch, S. M., Corwin, D. L., and Robinson, D. A. (2005). “Apparent soil electrical conductivity mapping as an agricultural management tool in arid zone soils.” Comput. Electron. Agric., 46 (1–3), 351–378. Lesch, S. M., Herrero, J., and Rhoades, J. D. (1998). “Monitoring for temporal changes in soil salinity using electromagnetic induction techniques.” Soil Sci. Soc. Am. J., 62, 232–242. Lesch, S. M., Rhoades, J. D., and Corwin, D. L. (2000). ESAP-95 version 2.10R: User and tutorial guide, Research Report 146, USDA-ARS, U.S. Salinity Laboratory, Riverside, Calif. Lesch, S. M., Rhoades, J. D., Lund, L. J., and Corwin, D. L. (1992). “Mapping soil salinity using calibrated electromagnetic measurements.” Soil Sci. Soc. Am. J., 56, 540–548.

336

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Lesch, S. M., Strauss, D. J., and Rhoades, J. D. (1995a). “Spatial prediction of soil salinity using electromagnetic induction techniques: 1. Statistical prediction models: A comparison of multiple linear regression and cokriging.” Water Resour. Res., 31, 373–386. ———. (1995b). “Spatial prediction of soil salinity using electromagnetic induction techniques: 2. An efficient spatial sampling algorithm suitable for multiple linear regression model identification and estimation.” Water Resour. Res., 31, 387–398. Lobell, D. B., Lesch, S. M., Corwin, D. L., Ulmer, M. G., Anderson, K. A., Potts, D. J., Doolittle, J. A., Matos, M. R., and Baltes, M. J. (2010). “Regional-scale assessment of soil salinity in the Red River Valley using multi-year MODIS EVI and NDVI,” J. Environ. Qual., 39, 35–41. Lobell, D. B., Ortiz-Monasterio, J. I., Gurrola, F. C., and Valenzuela, L. (2007). “Identification of saline soils with multiyear remote sensing of crop yields.” Soil Sci. Soc. Am. J., 71, 777–783. Madrigal, L. P., Wiegand, C. L., Meraz, J. G., Rubio, B. D. R., Estrada, X. C., and Ramirez, O. L. (2003). “Soil salinity and its effect on crop yield: A study using satellite imagery in three irrigation districts.” Ingenieria Hidraulica En Mexico, 18, 83–97. Mallants, D., Vanclooster, M., Toride, N., Vanderborght, J., van Genuchten, M. T., and Feyen, J. (1996). “Comparison of three methods to calibrate TDR for monitoring solute movement in undisturbed soil.” Soil Sci. Soc. Am. J., 60, 747–754. Mankin, K. R., Ewing, K. L., Schrock, M. D., and Kluitenberg, G. J. (1997). Field measurement and mapping of soil salinity in saline seeps, ASAE Paper No. 973145, 1997 ASAE Winter Meetings, December 1997, Chicago, ASAE, St. Joseph, Mich. Mankin, K. R., and Karthikeyan, R. (2002). “Field assessment of saline seep remediation using electromagnetic induction.” Trans. ASAE, 45(1), 99–107. Maas, E. V. 1986. “Salt tolerance of plants.” Appl. Agric. Res., 1, 12–25. Marion, G. M., and Babcock, K. L. (1976). “Predicting specific conductance and salt concentration of dilute aqueous solution.” Soil Sci., 122, 181–187. McBride, R. A., Gordon, A. M., and Shrive, S. C. (1990). “Estimating forest soil quality from terrain measurements of apparent electrical conductivity.” Soil Sci. Soc. Am. J., 54, 290–293. McNeill, J. D. (1980). Electromagnetic terrain conductivity measurement at low induction numbers, Technical Note TN-6, Geonics Ltd., Mississauga, Ontartio, Canada. ———. (1992). “Rapid, accurate mapping of soil salinity by electromagnetic ground conductivity meters,” in Advances in measurements of soil physical properties: Bringing theory into practice, SSSA Special Publication No. 30, G. C. Topp, W. D. Reynolds, and R. E. Green, eds., ASA-CSSA-SSSA, Madison, Wisc., 201–229. Metternicht, G. (2001). “Assessing temporal and spatial changes of salinity using fuzzy logic, remote sensing and GIS. Foundations of an expert system.” Ecol. Model., 144, 163–179. Metternicht, G., and Zinck, J. A. (1997). “Spatial discrimination of salt- and sodium-affected soil surfaces.” Int. J. Remote Sens., 18, 2571–2586.

LABORATORY AND FIELD MEASUREMENTS

337

———. (2003). “Remote sensing of soil salinity: Potentials and constraints.” Remote Sens. Environ., 85, 1–20. Mougenot, B., Pouget, M., and Epema, G. F. (1993). “Remote sensing of salt affected soils.” Remote Sens. Rev., 7, 241–259. Morgan, C. L .S., Norman, J. M., Wolkowski, R. P., Lowery, B., Morgan, G. D., and Schuler, R. (2000). “Two approaches to mapping plant available water: EM-38 measurements and inverse yield modeling,” in Proc., 5th Int. Conf. on Precision Agriculture, Minneapolis, Minnesota, July 16–19, 2000, P. C. Roberts, R. H. Rust, and W. E. Larson, eds., ASA-CSSA-SSSA, Madison, Wisc., 14 [CD-ROM], Müller, W. G. (2001). Collecting spatial data: Optimum design of experiments for random fields, 2nd ed., Physica-Verlag, Heidelberg. Müller, W. G., and Zimmerman, D. L. (1999). “Optimal designs for variogram estimation.” Environmetrics, 10, 23–37. Nettleton, W. D., Bushue, L., Doolittle, J. A., Wndres, T. J., and Indorante, S. J. (1994). “Sodium affected soil identification in south-central Illinois by electromagnetic induction.” Soil Sci. Soc. Am. J., 58, 1190–1193. Nobes, D. C., Armstrong, M. J., and Close, M. E. (2000). “Delineation of a landfill leachate plume and flow channels in coastal sands near Christchurch, New Zealand, using a shallow electromagnetic survey method.” Hydrogeol. J., 8(3), 328–336. Nyquist, J. E., and Blair, M. S. (1991). “Geophysical tracking and data logging system: Description and case history.” Geophysics, 56(7), 1114–1121. Paine, J. G. (2003). “Determining salinization extent, identifying salinity sources, and estimating chloride mass using surface, borehole, and airborne electromagnetic induction methods.” Water Resour. Res., 39(3), 1059. Ranjan, R. S., Karthigesu, T., and Bulley, N. R. (1995). Evaluation of an electromagnetic method for detecting lateral seepage around manure storage lagoons, ASAE Paper No. 952440, ASAE, St. Joseph, Mich. Raulund-Rasmussen, K. (1989). “Aluminum contamination and other changes of acid soil solution isolated by means of porcelain suction cups.” J. Soil Sci., 40, 95–102. Reece, C. F. (1998). “Simple method for determining cable length resistance in time domain reflectometry systems.” Soil Sci. Soc. Am. J., 62, 314–317. Reitemeier, R. F., and Wilcox, L. V. (1946). “A critique of estimating soil solution concentrations from the electrical conductivity of saturated soils.” Soil Sci., 61, 281–293. Rhoades, J. D. (1992). “Instrumental field methods of salinity appraisal,” in Advances in measurement of soil physical properties: Bringing theory into practice, SSSA Special Publication No. 30, G. C. Topp, W. D. Reynolds, and R. E. Green, eds., SSSA, Madison, Wisc., 231–248. ———. (1993). “Electrical conductivity methods for measuring and mapping soil salinity,” in Advances in agronomy, D. L. Sparks, ed., Vol. 49, Academic Press, San Diego, Calif., 201–251. Rhoades, J. D., Chanduvi, F., and Lesch, S. (1999b). Soil salinity assessment: Methods and interpretation of electrical conductivity measurements, FAO Irrigation and Drainage Paper No. 57, Food and Agriculture Organisation of the United Nations, Rome.

338

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Rhoades, J. D., and Corwin, D. L. (1981). “Determining soil electrical conductivitydepth relations using an inductive electromagnetic soil conductivity meter.” Soil Sci. Soc. Am. J., 45, 255–260. ———. (1990). “Soil electrical conductivity: Effects of soil properties and application to soil salinity appraisal,.” Commun. Soil Sci. Plant Anal., 21, 837–860. Rhoades, J. D., Corwin, D. L., and Lesch, S. M. (1999a). “Geospatial measurements of soil electrical conductivity to assess soil salinity and diffuse salt loading from irrigation,” in Assessment of non-point source pollution in the vadose zone, D. L. Corwin, K. Loague, and T. R. Ellsworth, eds., Geophysical Monograph 108, American Geophysical Union, Washington, D.C., 197–215. Rhoades, J. D., and Halvorson, A. D. (1977). Electrical conductivity methods for detecting and delineating saline seeps and measuring salinity in northern Great Plains soils, ARS W-42, USDA-ARS Western Region, Berkeley, Calif., 1–45. Rhoades, J. D., and Loveday, J. (1990). “Salinity in irrigated agriculture,” in Irrigation of agricultural crops, Agronomy Monograph No. 30, B. A. Stewart and D. R. Nielsen, eds., SSSA, Madison, Wisc., 1089–1142. Rhoades, J. D., Manteghi, N. A., Shouse, P. J., and Alves, W. J. (1989a). “Estimating soil salinity from saturated soil-paste electrical conductivity.” Soil Sci. Soc. Am. J., 53, 428–433. ———. (1989b). “Soil electrical conductivity and soil salinity: New formulations and calibrations. Soil Sci. Soc. Am. J., 53, 433–439. Rhoades, J. D., Raats, P. A. C., and Prather, R. J. (1976). “Effects of liquid-phase electrical conductivity, water content and surface conductivity on bulk soil electrical conductivity.” Soil Sci. Soc. Am. J., 40, 651–655. Rhoades, J. D., Shouse, P. J., Alves, W. J., Manteghi, N. M., and Lesch, S. M. (1990). “Determining soil salinity from soil electrical conductivity using different models and estimates.” Soil Sci. Soc. Am. J., 54, 46–54. Russo, D. (1984). “Design of an optimal sampling network for estimating the variogram.” Soil Sci. Soc. Am. J., 48, 708–716. Salama, R. B., Bartle, G., Farrington, P., and Wilson, V. (1994). “Basin geomorphological controls on the mechanism of recharge and discharge and its effect on salt storage and mobilization: Comparative study using geophysical surveys.” J. Hydrol., 155(1/2), 1–26. Scanlon, B. R., Paine, J. G., and Goldsmith, R. S. (1999). “Evaluation of electromagnetic induction as a reconnaissance technique to characterize unsaturated flow in an arid setting.” Ground Water, 37(2), 296–304. Sharma, R. C., Saxena, R. K., and Verma, K. S. (2000). “Reconnaissance mapping and management of salt-affected soils using satellite images.” Int. J. Remote Sens., 21, 3209–3218. Sheets, K. R., and Hendrickx, J. M. H. (1995). “Non-invasive soil water content measurement using electromagnetic induction.” Water Resour. Res., 31, 2401–2409. Slavich, P. G., and Petterson, G. H. (1990). “Estimating average rootzone salinity from electromagnetic induction (EM-38) measurements.” Aust. J. Soil Res., 28, 453–463. Smith, C. N., Parrish, R. S., and Brown, D. S. (1990). “Conducting field studies for testing pesticide leaching models.” Int. J. Environ. Anal. Chem., 39, 3–21. Spaans, E. J. A., and Baker, J. M. (1993). “Simple baluns in parallel probes for time domain reflectometry.” Soil Sci. Soc. Am. J., 57, 668–673.

LABORATORY AND FIELD MEASUREMENTS

339

Sparks, D. L., ed. (1996). Methods of soil analysis: Part 3—Chemical methods, SSSA Book Series 5, SSSA, Madison, Wisc. Stroh, J. C., Archer, S. R., Doolittle, J. A., and Wilding, L. P. (2001). Detection of edaphic discontinuities with ground-penetrating radar and electromagnetic induction.” Landscape Ecol., 16(5), 377–390. Stroh, J. C., Archer, S. R., Wilding, L. P., and Doolittle, J. A. (1993). “Assessing the influence of subsoil heterogeneity on vegetation in the Rio Grande Plains of south Texas using electromagnetic induction and geographical information system, College Station, Texas.” The Station, March 1993, 39–42. Suarez, D. L., and Taber, P. (2007). ExtractChem software: Version 1.0.18, U.S. Salinity Laboratory, Riverside, Calif. Sudduth, K. A., and Kitchen, N. R. (1993). Electromagnetic induction sensing of claypan depth, ASAE Paper No. 931531, 1993 ASAE Winter Meetings, December 12–17, 1993, Chicago, ASAE, St. Joseph, Mich. Sudduth, K. A., Kitchen, N. R., Wiebold, W. J., Batchelor, W. D., Bollero, G. A., Bullock, D. G., Clay, D. E., Palm, H. L., Pierce, F. J., Schuler, R. T., and Thelen, K. D. (2005). “Relating apparent electrical conductivity to soil properties across the north-central USA.” Comput. Electron. Agric., 46 (1–3), 263–283. Telford, W. M., Gledart, L. P., and Sheriff, R. E. (1990). Applied geophysics, 2nd ed., Cambridge University Press, Cambridge, UK. Thompson, S. K. (1992). Sampling, John Wiley and Sons, Inc., New York. Topp, G. C., and Davis, J. L. 1981. “Detecting infiltration of water through the soil cracks by time-domain reflectometry.” Geoderma 26, 13–23. Topp, G. C., Davis, J. L., and Annan, A. P. (1980). “Electromagnetic determination of soil water content: Measurement in coaxial transmission lines.” Water Resour. Res., 16, 574–582. ———. (1982). “Electromagnetic determination of soil water content using TDR: 1. Applications to wetting fronts and steep gradients.” Soil Sci. Soc. Am. J., 46, 672–678. Triantafilis, J., Ahmed, M. F., and Odeh, I. O. A. (2002). “Application of a Mobile Electromagnetic Sensing System (MESS) to assess cause and management of soil salinization in an irrigated cotton-growing field.” Soil Use Mgmt., 18(4), 330–339. Triantafilis, J., Huckel, A. I., and Odeh, I. O. A. (2001). “Comparison of statistical prediction methods for estimating field-scale clay content using different combinations of ancillary variables.” Soil Sci., 166(6), 415–427. Trianatafilis, J., and Lesch, S. M. (2005). “Mapping clay content variation using electromagnetic induction techniques.” Comput. Electron. Agric., 46, 203–237. Tweed, S . O., Leblanc, M., Webb, J. A., and Lubczynski, M. W. (2007). “Remote sensing and GIS for mapping groundwater recharge and discharge areas in salinity-prone catchments, southeastern Australia.” Hydrogeol. J., 15, 75–96. U.S. Salinity Laboratory. (1954). Diagnosis and improvement of saline and alkali soils, U.S. Department of Agriculture Handbook No. 60, U.S. Government Printing Office, Washington, D.C. Valliant, R., Dorfman, A. H., and Royall, R. M. (2000). Finite population sampling and inference: A prediction approach, John Wiley and Sons, Inc., New York. van der Lelij, A. (1983). Use of an electromagnetic induction instrument (type EM38) for mapping of soil salinity, Internal Report, Research Branch, Water Resources Commission, NSW, Australia.

340

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Vaughan, P. J., Lesch, S. M., Corwin, D. L., and Cone, D. G. (1995). “Water content on soil salinity prediction: A geostatistical study using cokriging.” Soil Sci. Soc. Am. J., 59, 1146–1156. Veris Technologies. (2011) “Products,” Veris Technologies, Salina, Kansas, www.veristech.com, accessed January 21, 2011. Verma, K. S., Saxena, R. K., Barthwal, A. K., and Deshmukh, S. N. (1994). “Remote-sensing technique for mapping salt-affected soils.” Int. J. Remote Sens., 15, 1901–1914. Warrick, A. W., and Myers, D. E. (1987). “Optimization of sampling locations for variogram calculations.” Water Resour. Res., 23, 496–500. Wesseling, J., and Oster, J. D. (1973). “Response of salinity sensors to rapidly changing salinity.” Soil Sci. Soc. Am. Proc., 37, 553–557. Wiegand, C. L., Anderson, G., Lingle, S., and Escobar, D. (1996). “Soil salinity effects on crop growth and yield: Illustration of an analysis and mapping methodology for sugarcane.” J. Plant Physiol., 148, 418–424. Wiegand, C. L., Rhoades, J. D., Escobar, D. E., and Everitt, J. H. (1994). “Photographic and videographic observations for determining and mapping the response of cotton to soil-salinity.” Remote Sens. Environ., 49, 212–223. Williams, B. G., and Baker, G. C. (1982). “An electromagnetic induction technique for reconnaissance surveys of soil salinity hazards.” Aust. J. Soil Res., 20, 107–118. Williams, B. G., and Hoey, D. (1987). “The use of electromagnetic induction to detect the spatial variability of the salt and clay contents of soils.” Aust. J. Soil Res., 25, 21–27. Wilson, R. C., Freeland, R. S., Wilkerson, J. B., and Yoder, R. E. (2002). Imaging the lateral migration of subsurface moisture using electromagnetic induction, ASAE Paper No. 023070, 2002 ASAE Annual International Meeting, July 28–31, 2002, Chicago, ASAE, St. Joseph, Mich. Wollenhaupt, N. C., Richardson, J. L., Foss, J. E., and Doll, E. C. (1986). “A rapid method for estimating weighted soil salinity from apparent soil electrical conductivity measured with an aboveground electromagnetic induction meter.” Can. J. Soil Sci., 66, 315–321. Wraith, J. M. (2002). “Solute content and concentration: Indirect measurement of solute concentration: Time domain reflectometry,” in Methods of soil analysis. Part 4: Physical methods, J. H. Dane and G. C. Topp, eds., Agronomy Monograph No. 9. SSSA, Madison, Wisc., 1289–1297. Zhu, Z., and Stein, M. L. (2006). “Spatial sampling design for prediction with estimated parameters.” J. Agric. Bio. Environ. Statistics, 11, 24–44.

NOTATION CEC  cation exchange capacity EC  electrical conductivity ECa  electrical conductivity of the bulk soil, referred to as apparent soil electrical conductivity ECe  electrical conductivity of an extract of a saturated soil paste

LABORATORY AND FIELD MEASUREMENTS

341

ECp  electrical conductivity of saturated soil paste ECw  electrical conductivity of a soil solution EMh  electromagnetic induction when the instrument coils are parallel to the soil surface EMv  electromagnetic induction when the instrument coils are perpendicular to the soil surface EMI  electromagnetic induction ER  electrical resistivity ET  evapotranspiration NDVI  normalized difference vegetation index SP  saturation percentage TDR  time domain reflectometry

This page intentionally left blank

PART FOUR: DIAGNOSIS OF SALT PROBLEMS

CHAPTER 11 IRRIGATION WATER QUALITY ASSESSMENTS Donald L. Suarez

INTRODUCTION This chapter discusses the effects of inorganic elements in irrigation water on the long-term sustainability of the agricultural soil-water system. It focuses on salinity, sodicity, and the effects of sodicity on soil permeability, major cations and anions, and trace elements. Taking into account interactions of irrigation water, soils, and crops, it describes the net change in salinity resulting from irrigation at various leaching fractions and identifies generally acceptable levels of trace elements in irrigation supplies. A meaningful assessment of the quality of water used for irrigation should consider such local factors as the chemical reactivity of constituents dissolved in the water, the soil’s chemical and physical properties, climate, and irrigation management practices. It should also consider the effects of irrigation on the quality of agricultural drainage, effects on humans and animals of chemicals concentrated in harvested plant products, and economic conditions that determine how much salinity-induced reduction in yield or quality can be tolerated. To avoid the long-term accumulation of toxic amounts of waterborne substances in the rootzone of irrigated lands, the input of those substances to the soil from irrigation and other sources must not exceed the sum of losses from the soil and conversions to unavailable forms. Losses include removal in harvested crops, transport by subsurface drainage, erosion by wind and water, and, for some elements, volatilization of gaseous compounds. Relatively immobile elements, such as arsenic (As) and copper (Cu), often are converted in the rootzone to less available forms (such as adsorbed or precipitated solid phases). These forms are 343

344

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

sometimes considered unavailable; however, availability depends on the subsequent chemical conditions. Conversion processes are primarily due to precipitation or changes in redox status, and these conditions could change again in the future. For example, As in forms that may be considered unavailable under aerobic conditions could be remobilized under anaerobic conditions, or in response to changes in pH. In contrast, selenium (Se) is highly mobile under aerobic conditions and less mobile under anaerobic conditions. Changes in oxidation status, pH, or other chemical conditions could be related to changes in cropping patterns (such as conversion to rice cultivation) or changes in land use. The most mobile ion of importance, chloride (Cl) is relatively nonreactive, as most of its salts are highly soluble, and the ion undergoes little adsorption or exchange. Other mobile ions, such as nitrate (NO3), also undergo little adsorption or exchange but are subject to redox transformations, such as NO3 to ammonium (NH 4 ), which may be retained by the exchange sites, volatilized as ammonia (NH3), or incorporated into organic matter. Other elements, such as sodium (Na) and magnesium (Mg), exist in cationic form (Na and Mg2), are readily exchangeable, and are thus less mobile when going into soil exchange sites. Elements, such as boron (B), are adsorbed and less mobile, followed by elements, such as As, that are highly adsorbed. Most soluble constituents, being relatively mobile, can be removed by leaching. Thus, leaching often can be used to adjust the concentrations of soil chemical constituents to accommodate crop production. If the element of interest is immobile under existing soil conditions and if leaching losses are insignificant, then the elemental inputs not removed by plants or converted in the soil to unavailable forms will accumulate as soluble and labile (adsorbed) forms. These forms are related as follows: Soluble ↔ Labile ↔ Residual The soluble element adsorbs or desorbs into the labile form as the amount in solution increases or decreases. The labile element is transformed to or from the residual (relatively unavailable) form. Only the soluble form is immediately available to the plant. As the soluble element is removed by plant roots, desorption from the labile pool replenishes the soluble pool. Although the residual pool may not impact current agricultural production, it may nonetheless be of environmental concern, both in terms of potential mobility under different chemical conditions and as potential transport as dust to other sensitive environments, such as wetlands. The level of toxicity depends directly on the amount of the toxic constituent in solution, and indirectly on the capacity of the labile pool. The hazard posed by elements that exist in soluble and labile forms in the soil

IRRIGATION WATER QUALITY ASSESSMENTS

345

is that once toxic levels are attained, eliminating or reducing these levels involves removal in harvested crops and conversion to residual forms, both of which are processes that can take decades, even if inputs of the element cease. These situations can be avoided by ensuring that inputs of potentially toxic elements remain below the levels that are tolerable by the most sensitive crop to be grown and by avoiding crops that will bioaccumulate the elements of concern.

SALINITY Salinity in water is defined as the total sum of dissolved inorganic ions and molecules. The major components of salinity are the cations Ca2, Mg2, and Na, and the anions Cl, sulfate (SO2 4 ), and bicarbonate (HCO3 ) and NO3 . The potassium (K) and carbonate (CO2 3 ) ions are usually minor components of the salinity. The effects of these and other minor dissolved constituents, such as B, are generally neglected in assessing the salinity of irrigation waters but nonetheless are important when assessing the suitability of waters for irrigation. Salinity reduces crop growth by reducing the ability of plant roots to absorb water, by accumulation of toxic concentrations of salts in plant tissue, specific ion toxicity, and ion imbalances. The soluble ions and molecules reduce the availability of water to a plant, a phenomenon known as the osmotic pressure effect. The osmotic pressure effect is especially important at high salinity. Water availability in the soil relates to the combined (but not the simple sum) of the matric and osmotic potential stresses. As a first approximation, we can consider that the combined effects of osmotic and matric stress can be represented by multiplying the relative yield response of the individual stresses. For example, if the calculated salinity level is such that we predict a 70% relative yield and the matric stress is such that we predict a 50% relative yield, then the combined effect gives a predicted relative yield of 35%. This calculation must be based on actual measurements or modeling that accounts for the effect of salinity on matric stress and the effect of matric stress on salinity (as both reduce water uptake). The multiplication of yield response from multiple stresses has been utilized by several investigators (Suarez and Sˇ imu˚nek 1997; Shani et al. 2007). Shani et al. (2007) present an extensive review of available data related to plant response to multiple stresses. The resultant user-friendly SWS model (see Chapter 27 of this manual) derived from UNSATCHEM maintains these features. The dynamic models predict water consumption based on the actual stress rather than the evapotranspiration (ET) multiplied by crop coefficient information. In the above example, if the osmotic stress produced a 70% relative yield independent

346

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of matric stress, the water consumption is reduced 30% from the crop optimal ET and the soil salinity and matric stress is reduced; thus, the predicted yield from a dynamic model is greater than the 35% value given. More detail is provided in an example in Chapter 27. As the water content of the soil decreases, the matric and osmotic potential decreases (i.e., it becomes more negative). Evaporation and transpiration by plants remove almost pure water, leaving behind soluble salts in the soil. Depending on the water composition, salinity, plant species, and climatic conditions, about 5% to 10% of the salts are taken up by plants and the remainder is either left in the soil or leached with the drainage water. Electrical Conductivity Specific ion effects on plant yield are most evident in salt-sensitive species, such as rice, lettuce, strawberries, and stone fruits. Toxicity can be related to either the Na cation or Cl- anion, and is related to the ability of the individual plant species and cultivar to restrict uptake and movement of these ions. Salinity is most easily and conveniently measured by determining the electrical conductivity (EC) of the solution (see Chapter 10 of this manual for more detail). The term specific electrical conductance (SpC) is some times used as well. The U.S. Salinity Laboratory (USSL 1954) showed that the EC in soil extracts was highly correlated with total salts when the data were expressed in mmolc/L. The osmotic potential (OP) can be approximately related to EC by the equation OP  36  EC, where OP is expressed in kPa and EC in dS/m at 25 °C. While useful, these approximations should not be used in research experiments where more accurate calculations are warranted. More accurate estimations of OP can be made by consideration of the ion composition of the water, such as presented in the Extract Chem model (Suarez and Taber 2007). Soil-Water Extracts The EC is used as an expression of salinity in the irrigation water (ECiw), salinity in the soil saturation extract (ECe), and salinity in the soil solution (ECSS). The U.S. Salinity Laboratory researchers (1954) developed the saturation paste-saturation extract technique, a way to estimate soil salinity that uses a reference water content. The saturation paste is defined as a mixture of demineralized water added to a soil sample until the mixture (soil paste) glistens and slightly flows when the container is tipped. The soil paste is then typically left overnight to equilibrate and is filtered under suction the next day. The solution obtained is analyzed for

IRRIGATION WATER QUALITY ASSESSMENTS

347

ECe and soluble constituents. This extract, while not ideal, is nonetheless the most recommended for standardized representation of the soil-solution composition. Direct determination of the soil-solution composition is difficult due to the extraction, especially when the soil is not near saturation. Also, direct determination makes spatial and temporal comparisons difficult as the composition depends on water content at time of sampling. Extracts are convenient and rapid, providing data at reference water contents. Other extracts used include 1⬊1, 1⬊2, and 1⬊5 soil/water ratios. Clearly, the larger the dilution, the greater the deviation from the soil-water composition in situ and the more uncertain the interpretation of the data due to dissolution, exchange, and desorption. The saturation extract has the advantage of minimizing salt dissolution, relative to other dilutionextraction methods, since less water is added, but has the disadvantage of being the most time consuming. The water content of the saturated paste is roughly 1.5 to 2 times that of field capacity, but the exact value is quite variable depending on soil texture and mineralogy. The ECe is thus approximately one-half the ECSS at field capacity. These are relatively rough approximations suitable for field evaluation but not for reporting of salt tolerance data, as the errors can be in the range of 10% to 30%. These approximations do not consider the unique water content relation of each soil (saturated paste vs. field capacity), the nonlinearity between EC and salt content, or the reactivity of the soil, especially dissolution of gypsum if present during the addition of water and extraction. Recently Suarez and Taber (2007) developed the Extract Chem program. The program allows for conversion of the inorganic chemical composition of soil water from one water content to another, considering cation exchange, precipitation/dissolution of calcite and gypsum if specified, and adsorption/desorption of B. The model calculates EC using the routines developed by McNeal et al. (1970), based on solution composition. Comparison of the model to analyzed extracts reveals some of the problems associated with extracts, such as incomplete equilibration after reaction overnight (gypsum soils), and variability in CO2 and thus calcium depending on soil biological activity and experimental conditions. The ECe provides a way to assess the salinity of field samples. The relationships among ECiw, ECe, and ECss are critical, as a large amount of data on salinity tolerances of crops is based either on ECe or ECiw, whereas plant response is related to the ECss. The salinity of irrigation waters can be assessed by relating ECiw, the leaching fraction (LF), the ECss at field capacity, and the salt tolerance of crops of interest. Unfortunately, there are various recommendations for calculation of the soil salinity relevant to crop response, and they provide significantly differ-

348

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

ent results. See Chapter 10 for a more detailed discussion of the variability in soil salinity tests. Plant Response to Soil Salinity The most common way to represent the soil-solution EC relevant to plant response has been to use the average soil ECe (Ayers and Westcot 1985). This method simply averages the calculated or measured ECe of several depths. If the ECe data are not available, it has been suggested to calculate average ECe using the ECiw and the concentration factor Fc, which equals 1/LF at the bottom of the rootzone, and an assumed distribution of water uptake (Rhoades 1984; Ayers and Westcot 1985). Using this method, it is assumed that water is removed by ET in proportions of 0.40, 0.30, 0.20, and 0.10, from the rootzone’s first, second, third, and fourth quarters, respectively. Alternatively, an exponential water uptake function can be used; however, the concentration factors (Fc values) would not greatly change. Since the ECe is about one-half of the ECSS, the Fc values to convert from ECiw to ECe are 2.79, 1.88, 1.29, 1.03, 0.87, and 0.77 for LF of 0.05, 0.10, 0.20, 0.30, 0.40, and 0.50, respectively. These Fc values have been used to calculate the ECe values expected in the rootzone as a function of overall LF. These in turn have been used to calculate average rootzone soil ECe as related to LF and ECiw . The use of the average rootzone ECe to predict salinity effects on crop yield is widely accepted but questionable on several grounds. First, plant water uptake is not uniform throughout the rootzone. If we use the same water uptake functions that were used to generate the EC soil profiles, multiply the soil salinity at each depth by these factors, and sum the product for the rootzone, then we generate EC values that correspond to the average EC of the water that the plant has taken up. These uptakecorrected EC values are considerably lower than the average ECe values, and the differences increase with decreasing LF, as shown in Table 11-1. For example, at an LF 0.05, the mean soil EC is 55% greater than the uptake-weighted EC, whereas at an LF of 0.5 it is only 10% greater. It is recommended to use these uptake-weighted factors and not the average salinity to calculate plant response to soil salinity. As long as we use the same function or distribution for water uptake as we used to calculate the soil salinity depth profile from LF and irrigation water EC, then we will have a reasonable estimate of the salinity experienced by the plant. For instance, if the water uptake pattern is different from that assumed here, we still get the same uptake-weighted salinity concentration factors as the water uptake drives the salinity distribution. We need only ensure that we have divided the soil into sufficient compartments (four compartments appears satisfactory in most instances).

IRRIGATION WATER QUALITY ASSESSMENTS

349

TABLE 11-1. Relative Solute Concentrations of Soil Water (Field Capacity Basis, Fc) Compared to That of Irrigation Water Related to Depth in the Rootzone and Leaching Fractiona Fc at Leaching Fraction Values Of: Rootzone in Quarters (1)

0 1 2 3 4 Mean Fcc Uptake-weighted Fcd

b Vcu

(2)

0.05 (3)

0.10 (4)

0.20 (5)

0.30 (6)

0.40 (7)

0.50 (8)

0 40 70 90 100

1.0 1.61 3.03 7.14 20.00

1.0 1.56 2.70 5.26 10.0

1.0 1.47 2.27 3.57 5.00

1.0 1.39 1.96 2.70 3.33

1.0 1.32 1.72 2.17 2.50

1.0 1.25 1.54 1.82 2.00

5.58 3.6

3.76 2.71

2.58 2.07

2.06 1.75

1.74 1.54

1.53 1.40

a

Assuming a water uptake of 0.4, 0.3, 0.2, and 0.1, respectively, from the first through fourth quarters of the root zone b Cumulative percentage of consumptive use above each indicated depth in the rootzone c The average for the rootzone obtained by the sum of quarter of the root zone divided by 4 d The water uptake-weighted mean for the rootzone

The water uptake-weighted salinity, while more realistic than the mean rootzone salinity in representing plant salt stress, is nonetheless still a simplification. It does not consider the following factors: 1. In the short term, plants can compensate for reduced water uptake in some areas of the rootzone by increased uptake in other regions. However, in the longer term, this redistribution of water uptake causes a redistribution of roots and redistribution of the salinity profile, with the water uptake reverting back to the previous concentration factors. For example, if plants consume 90% of the water applied, then over time they must extract water up to the salinity level corresponding to this concentration factor, and the water uptake-weighted salinity goes back to the steady-state concentration factors listed here and in Table 11-1. 2. The concentration factors do not consider the changes in EC due to chemical processes, mostly calcite and gypsum precipitation and dissolution; these can easily change the concentration factors by 10% to 30% or more, depending on the specific conditions. In most instances this results in lower salinity than calculated by the concentration factors. The important exception, where salinity in the soil is greater than

350

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

that calculated by the concentration factors, is when a gypsiferous soil is irrigated with a water containing small concentrations of calcium and sulfate. 3. The steady-state factors do not consider the dynamics of wetting and drying cycles. As the soil dries out, the resultant in situ soil salinity and EC increase. Infrequent irrigation results in increased soil salinity averaged over time, in addition to possible matric stress. This is not an issue in the case of high-frequency irrigation. 4. When the LF is calculated, the actual ET—not the potential ET—must be considered as increased salinity results in decreased plant water uptake. This requires a feedback loop from the salt stress response to the calculation of ET. The UNSATCHEM model (Suarez and Sˇ imu˚nek 1997), and the user-friendly SWS version (see Chapter 27) uses a water uptake response function (separate osmotic and matric functions) at each point in the rootzone. Thus, the LF fraction calculated by the model is not solely defined from ETo, crop coefficients and water inputs. If plant response is to osmotic stress, then osmotic stress needs to be calculated rather than estimated from EC, as there is a significant difference in the relationship of osmotic pressure and EC for chloride salts compared to sulfate salts. The SWS model also calculates osmotic pressure and EC after consideration of chemical processes. The salinity threshold values, meaning the salinity at which plant yields start to decline, are derived from the following relationship between yield and ECe: Yield  100  B (ECe  A)

(11-1)

where A  the salinity concentration at which growth depression (threshold) starts, and B  the percent of yield decrease per unit ECe above the threshold level (Maas and Hoffman 1977). Figure 11-1 shows the relationships between ECss and ECiw for various LF based on calculations as described for Table 11-1. In the previous edition of this manual (1990), Fig. 11-1 was used for high-frequency irrigation systems only and the average rootzone salinity was used for furrow and other nonfrequent irrigation systems. This special consideration has been dropped because, despite theoretical expectations, there is no clear evidence that frequent irrigation reduces salt damage (Shalhavet 1994). Conversion of these ECss data to ECe should consider the specific soil properties and water composition; in the absence of such information, the user would have to use the approximate conversion ECe 艐 0.5 ECss. To use Fig. 11-1 for evaluation of potential yield loss due to salinity damage, determine the ECiw and then estimate the range in LF that can be obtained for the soil with the available irrigation management system. Next, compare the resultant ECss values with the ECss values from the salt

IRRIGATION WATER QUALITY ASSESSMENTS

351

FIGURE 11-1. Relationship between average rootzone salinity (field capacity basis), EC of irrigation water, and LF required to avoid yield loss. Modified from Rhoades (1982).

tolerance tables. This will indicate crops that can be grown successfully without decreases in yield from salinity. For example, if ECiw is 4.0 and an LF of 0.20 is expected, only salt-tolerant plants can be grown without yield loss. If LFs of 0.5 or greater are possible, moderately salt-tolerant plants can be grown. If the nature of the soil hydraulic properties or water availability is such that only very small LFs are possible, then in this instance (where ECiw  4.0) the water will reduce yields in even the most salt-tolerant crops. Thus, assessing the effects of salinity as a parameter of water quality depends on the soil, crops, amount of water available, reference crop ET of the site (ET0), irrigation system, irrigator’s expertise in achieving the needed leaching, and decrease in yield that can be tolerated. In short, from the standpoint of salinity, the suitability of a given irrigation water supply requires an evaluation of how the applied water will interact with the soils, the resultant LF (dependent on ET0 and salt stress), and the net change on soil salinity.

352

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The method of assessing the water salinity as described can be adapted to different sites. The suitability of the water supply can be assessed based on such local conditions as the ease with which the soil can be leached, salt tolerance of the crops, irrigation system, skill of the manager, and climate. Perhaps the weakest link in this system is the estimation of the LF, which is seldom measured directly, but often determined by measuring water application and estimating ET from crop coefficients and ET0. The difficulty is that not all applied water infiltrates (we need to correct for surface runoff, often called tail water), and that actual ET is not an input but a response, depending on crop stress. If there is salinity stress, then for a fixed application of water, as salinity increases, ET decreases and the LF increases, with LF determined by the crop response to salinity as well as by the water application. In irrigation waters that are sprinkled, there is also a potential for direct injury to the plant from absorption of salts in the irrigation water by foliage. The foliar injury from salts on plants depends on the concentrations of the individual ions in the water, sensitivity of the crop, frequency of sprinkling, presence of sunlight, and environmental factors (such as temperature, relative humidity, and water stress of the plants before irrigation). Maas et al. (1982) reported that rates of salt absorption by leaves increased as the frequency of irrigation increased but that a threefold increase in the duration of sprinkling had no measurable effect on salt absorption. Night-time sprinkling reduces foliar absorption and injury. Foliar absorption by Na or Cl ions at concentrations of less than 5 mmol/L damages some fruit trees. Other crops can tolerate Na and Cl ion concentrations of greater than 20 mmol/L. Thus, no concentration limits can be recommended, although an increase in Na or Cl in the water reduces its suitability for sprinkler systems by reducing the types of crops that can be grown without foliar injury. Also, the degree of injury depends on the crop, the irrigation system, and how it is operated. For example, Suarez et al. (2003) observed almost a doubling of the Se shoot concentration of Brassica species under sprinkler rather than flood irrigation, but the relative increase in Se uptake was crop-dependent. Foliar uptake can be expected to be related to shoot morphology, as well as leaf structural characteristics. Consequently, limits or guidelines for sprinkler irrigation at current levels of knowledge are too arbitrary to be useful.

SODICITY Sodium hazards of irrigation and soil waters can negatively affect crop production due to both specific ion toxicity (as discussed) and the adverse effect of Na on soil physical properties, especially water infiltration. The growth of plants is, thus, affected by either an unavailability of

IRRIGATION WATER QUALITY ASSESSMENTS

353

soil water or poor aeration due to reduced water movement and subsequent waterlogging. The reduction in water infiltration caused by Na can usually be attributed to surface crusting, dispersion and migration of clay into the soil pores, and swelling of expandable clays. All of these phenomena relate to the distance of charge neutralization for soil particles, predominantly clay, but also oxides in more weathered landscapes. The hydrated exchangeable cations neutralize the net negative charge on clays. The distance of charge neutralization (the double-layer thickness) depends on the cation valence, hydration energy, and ion concentration in solution. Divalent cations, such as Ca2 and Mg2, neutralize the surface charge in relatively short distances, even at low concentrations. Particles are repulsed when the charge is neutralized too far from the surface and the electrostatic repulsion between particles exceeds the attractive (van der Waals) forces. In contrast to Ca2 and Mg2 ions, the exchangeable Na ion neutralizes the surface charge at a longer distance (much larger, double-layer thickness) and requires high concentrations in solution before particle aggregation and swelling are reduced. Consider Ca2 as a stabilizing ion, Mg2 less so (Dontosova and Norton 2002), and Na as a destabilizing ion in regard to the soil structure. The sodicity of a soil is given by the exchangeable sodium percentage, ESP, which is the percentage of the exchangeable charge neutralized by Na. The ESP of a soil can be estimated from the sodium adsorption ration (SAR) of the water, in other words, ESP  1.475 SAR/(1  0.0147 SAR), based on a set of data from soils in the western United States (U.S. Salinity Laboratory 1954; also see Chapter 3 of this manual). The ESP value alone is insufficient for predicting soil stability. Soil structure depends on many other factors, including soil salinity, tillage, mineralogy, organic matter, and pH. Sodicity Hazard Guidelines The sodic-hazard potential of water is often evaluated from the SAR and salinity. At the same SAR, the dispersion potential of dilute water exceeds that of a more saline water. Various investigators have developed stability lines related to concentration and SAR. Perhaps the most widely used is that presented by Ayers and Westcot (1985). Figure 11-2 shows the guidelines of Rhoades (1982) and Quirk and Schofield (1955) represented as solid and dashed lines, respectively. Rhoades based his guidelines primarily on experience and data from arid soils in California. Quirk and Schofield (1955) based their guidelines on a noncalcareous soil in England. In each instance, the region below the line represents unstable soil structure and permeability loss, and the region above it represents stable permeability.

354

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 11-2. Relationship between SAR and solute concentration (in mmol/L) at which a 25% reduction in soil hydraulic conductivity was observed. The data were obtained from laboratory studies of packed soil columns containing aridland soils. The dashed and solid lines are guideline values recommended by Quirk and Schofield (1955) and Rhoades (1982), respectively.

Figure 11-2 also shows the concentration and SAR values at which a 25% reduction in saturated hydraulic conductivity took place in packed laboratory soil columns, from available published data from arid soils. A general relationship cannot be predicted because soils greatly differ, but a good SAR versus concentration relationship for a set of soils from a region or locality is always possible. For all arid soils examined, decreasing salinity, or increasing sodicity, or both, decreases soil stability.

IRRIGATION WATER QUALITY ASSESSMENTS

355

Differences among soils shown in Fig. 11-2 are at least partly due to experimental procedures used by different researchers, such as different column packing, flow rates, and saturation methods. However, variations in clay mineralogy, clay content, organic matter, and oxide content likely account for most of the variation. Almost all of the soils have been examined under relatively low pH (7.0). For specific groups of soils, several researchers have demonstrated that soil stability correlates well with organic matter, or oxide content, or both. Soils with a very large amount of oxides, such as some tropical soils, show little or no loss of hydraulic conductivity, even when saturated with sodium and equilibrated to minimal levels of salinity. Organic soils may also be highly stable at low salinity, as long as the pH is not elevated. Texture and initial hydraulic characteristics have also not received sufficient attention. Sandy soils with high infiltration rates can remain productive with 25% losses in infiltration rates, but this is not true for clay soils where infiltration may be barely sufficient to supply crop water needs during high ET conditions, even without sodicity effects. Since the effects of variables other than salinity, SAR, and their interactions have not been quantified, EC-SAR suitability figures offer only an approximate guideline. A representation of the stability of arid zone soils as related to irrigation water quality in the absence of rain is shown in Fig. 11-3, based primarily on research at the U.S. Salinity Laboratory. This guideline differs from the other guidelines in that it includes the effect of pH and is based to a considerable extent on longer-term infiltration experiments. The relationships at low SAR and EC are primarily based on the experiments of Suarez et al. (2006, 2008) and D. L. Suarez and A. Gonzalez Rubio (unpublished data). The slopes of the solid lines (EC vs. SAR) are similar to those used by others (Ayers and Westcot 1985; Rhoades 1982). The area between the lines represents a region of little to 25% reduction in infiltration. If we were to select a line where all soils had 25% or less reduction in infiltration, it would be a line almost on the x-axis, as shown by the data in Fig. 11-2. Similarly, the upper left line in Fig. 11-3 is for a typical arid land soil; below that line less stable soils may already have severe reductions in infiltration. Effect of pH on Infiltration Increasing pH is known to increase the salt concentration necessary for flocculation of soil clays (Suarez et al. 1984; Goldberg and Forster 1990). An increase in pH also has an adverse effect on saturated hydraulic conductivity (Suarez et al. 1984). Also shown in Fig. 11-3 are estimates of the impact of irrigation water pH on water infiltration. Increased pH has an adverse effect on infiltration (D. L. Suarez and A. Gonzalez Rubio, unpublished data), as well as saturated hydraulic conductivity. The relative

356

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 11-3. Relationship between SAR and solute concentration (in mmol/L), at which little or no reduction or greater than 25% reduction in infiltration is expected, based on data from arid and semiarid land soils, in the absence of rain. The solid lines represent the effects for waters at pH 6.5, the dashed lines represent the effects for waters at pH 8, and the dotted lines are for waters at pH 9. impact of pH on infiltration may also be expected to depend on texture and clay type. The information on the effect of pH on hydraulic conductivity or infiltration is limited to only a few soils. Figure 11-4 shows a representation of the impact of irrigation water quality in the presence of substantial rain. It represents the results of different experiments conducted in Riverside, California, with various soils of different texture and geographic origin (Suarez et. al. 2006; Suarez et al. 2008; D. L. Suarez and A. Gonzalez Rubio, unpublished data). In this representation, no irrigation waters with an SAR above 5 can be considered safe in the presence of rain. The adverse results of the rain results not only from physical impact of the drops but also from the chemical changes at

IRRIGATION WATER QUALITY ASSESSMENTS

357

FIGURE 11-4. Relationship between SAR and solute concentration (in mmol/L), at which a 25% reduction in infiltration is expected, based on data from arid and semiarid land soils, in the presence of rain.

the soil surface. Rain results in a rapid decrease in EC as the water infiltrates, as shown in Fig. 11-5, for simulations of two calcareous soils of differing texture (Suarez et al. 2006). As shown in Fig. 11-6, with infiltration of rain, there is a much slower change in SAR than EC (Fig. 11-5) and the change depends on the cation exchange content of the soil, with higher cation exchange soils having a greater resistance to changes in SAR. Noncalcareous soils would have a considerably slower change in SAR than shown in Fig. 11-6, thus increased sensitivity to rain on a sodic soil. It is recommended that the effects of an irrigation water be tested directly on the soil of interest with column leaching studies, tests of aggregate stability, or tests of flocculation after the soil has been dispersed in a

358

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

A

B

FIGURE 11-5. Predicted relationship of EC with depth and quantity of rain infiltrated for (a) loam soil, and (b) clay soil. The initial condition was EC  1.0 ds/m and SAR 10. Each curve represents the addition of 1 cm of rain.

IRRIGATION WATER QUALITY ASSESSMENTS

359

A

B

FIGURE 11-6. Predicted relationship of SAR with depth and quantity of rain infiltrated for (a) loam soil, and (b) clay soil. The initial condition was EC  1.0 ds/m and SAR 10. Each curve represents the addition of 1 cm of rain.

360

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

test tube. This need arises because of the variability among soils in their response to Na. The SAR value calculated from analyses of surface waters usually represents the SAR of the irrigation water on the surface of the soil. In this instance there is little justification for making an SAR adjustment. This is not the case for groundwaters that are equilibrated with a much higher partial pressure of carbon dioxide (PCO2) and, thus, are lower in pH. Exposure of the groundwater to atmospheric CO2 conditions, which occurs with sprinkler irrigation or conveyance through open canals, raises the pH and may cause calcite precipitation. The adjusted SAR of the irrigation water is a correction of the SAR to account for the change in Ca concentrations as related to changes in the calcium carbonate solubility (see Chapter 3 of this manual). In such instances, assume a PCO2 of 0.1 kPa at the soil surface and adjust the SAR as described in Chapter 3. The adjusted SAR can also be used to estimate the SAR in or below the rootzone by correcting for mineral precipitation and assuming no ion exchange. The concentration factor (1/LF), the PCO2, and the chemical composition of the irrigation water are needed. If specific PCO2 data in the rootzone are unavailable, the values of 1 kPa and 5 kPa can be used for sandy and clay soils, respectively. Since the publication of the earlier edition of this manual, there has been a dramatic increase in computer availability, user capability, and ease of use of software to calculate chemical equilibria. The Extract Chem software, among many others, can be readily used to calculate a precise, adjusted SAR value. High pH values (i.e., pH 8.5) always indicate waters with an excess of alkalinity (HCO3  CO32) over Ca. These high-pH waters pose an extra sodicity hazard for several reasons. When alkalinity exceeds Ca, the increased concentration of salts in the soil due to ET causes calcite precipitation and a decrease in the Ca concentration (when Ca  alkalinity in mmolc/L), the Ca concentration remains constant or increases slightly during plant water extraction or evaporation of the water). Waters with pH values below 8.5 can also have high alkalinity, depending on PCO2. For groundwaters, samples should be aerated or shaken until the water is equilibrated with ambient CO2 levels and then the pH remeasured. If pH 8.5 after aeration, then the concentration of alkalinity is greater than Ca. The higher the pH, the greater the imbalance. Imbalances in alkalinity and Ca concentrations can also exist in waters with a pH 8.5 if they are dilute waters that are undersaturated with respect to calcite, such as surface waters from snowmelt. The very low EC of rain or snowmelt water (0.1 dS/m) compounds their Na hazard. High pH (pH 9.0) directly and adversely affects infiltration as discussed above, as well as limiting Ca concentrations and increasing the SAR. The sodicity hazard of an irrigation water also depends on the man-

IRRIGATION WATER QUALITY ASSESSMENTS

361

agement system used. Dispersing the soil at the surface requires inputs of energy and a weakening of chemical bonding. Irrigation by sprinkler will increase the infiltration problems associated with irrigation waters. Due to the greater likelihood of surface crusting, sprinkler irrigation may be unsuitable for waters that tend to cause dispersion or swelling. Drip or surface systems will produce less physical disruption with such waters. This is a very important consideration when using waters with a potential dispersion or swelling hazard. High pH values in irrigation waters may cause nutritional and infiltration problems and, thus, need to be amended to reduce the alkalinity. The composition of the divalent ion component slightly affects the stability of soils at a particular ESP value, with Ca slightly more stable than Mg (McNeal et al. 1968). The greater selectivity of most soils for Ca2 as compared to Mg2 means that the Mg-Na system has a higher ESP than does the Ca-Na system at the same SAR values. This, combined with the high pH that frequently occurs in low Ca systems, also accounts for why Mg seems deleterious to infiltration as compared to Ca under field conditions. Weathering of Ca containing minerals, primarily gypsum, calcite, and dolomite, decreases SAR and increases electrolyte concentration. The hydraulic conductivity response of sodic soil in arid areas to rain or waters of minimum salinity seems related to its weathering potential (Shainberg et al. 1981). The more stable soils appear to maintain higher electrolyte levels than do unstable soils. Tropical Hawaiian soils also appear to be much more stable than arid soils at comparable SAR and salinity levels, possibly due to their high oxide content (McNeal et al. 1968). Moderate amounts of organic matter also increase the stability of a soil (Kemper and Koch 1966; Dong et al. 1983). The potential hazard of reduced water infiltration is partly related to the intensity and timing of rainfall in a region. Rainfall, generally 0.06 dS/m, is relatively pure water. When it infiltrates the soil, the salinity of surface soil can decrease rapidly but the soil may remain at almost the same ESP. As a result, the potential for dispersion by rainfall is especially high if the ESP of the soil is high. Rainfall, as with sprinkler irrigation, contributes dispersive energy. Surface (flood, furrow, or drip) irrigation also can cause particles to migrate and result in sealing of the soil surface. However, the inputs of energy are less than with sprinkler irrigation, and lower-quality water can be better tolerated. In areas with little rainfall, such as California’s Imperial Valley, these phenomena are generally neglected. In areas with appreciable seasonal rainfall (200 mm), surface amendments, such as gypsum, can be applied to maintain the electrolyte concentrations above values for dispersion and swelling to allow the water to infiltrate. If rainfall occurs throughout the year, tillage or repeated gypsum applications may be needed.

362

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

IONIC BALANCES Calcium (Ca), Mg, K, S, N, and P are the major elements needed for plant nutrition. Minor requirements include Fe, Mn, Zn, Cu, Mo, and Ni (Marschner 1995). Plants generally tolerate widely varying concentrations of the major cations, including Na, which is not required for plant growth. The Ca requirement of a plant is generally low, i.e., 0.7 mmol/L to 1.5 mmol/L; however it appears to depend on the presence of other ions. The Ca requirement may be related to ion competition and, thus, is better expressed in terms of ion ratios. High Mg/Ca ratios in solution may result in Ca deficiencies in plants, despite high absolute Ca concentrations. Carter et al. (1979) observed reduced growth in barley, starting at Mg/Ca ratios of 1.0, independent of salinity or absolute Ca concentrations. Calcium requirements are also greater at low pH than at high pH (Marschner 1995). Guidelines for specific cation ratios cannot be developed at this time due to insufficient information; cultivars respond in widely varied ways to cation composition, and this has not been sufficiently researched. Concentrations of Mg and K generally are high enough in irrigation waters to prevent deficiency symptoms in plants. The micronutrient cations of Fe, Mn, Cu, and Zn are virtually absent from most irrigation waters, but the soil generally supplies these nutrients. The irrigation water limits the availability of micronutrients if the water causes the soil pH to increase. Specific cation toxicity takes place with excess Na, predominantly in citrus and stone fruits. For anions, specific toxicities occur, rather than ionic imbalances. Although most plants tolerate high Cl concentrations, woody species and some grape rootstocks do not. High levels of nitrate, often associated with saline waters, may narrow the selection of crops suitable for irrigation. Nitrogen is one of the essential elements for the growth of plants. Optimal growth requires 2% to 5% N on a dry weight basis, depending on the species, developmental stage, organ to be optimized, and ultimate use of the plant or parts (Marschner 1995). High levels of nitrate during early growth enhance shoot elongation which, in cereals, increases susceptibility to lodging. High substrate levels of nitrate increase total N in ryegrass but decrease carbohydrates and increase cellulose content. Such high levels of NO3, that is, 1% to 2% by dry weight, can be toxic to grazing animals. High nitrate concentrations can cause excessive vegetative growth and reduce production of fruits and other harvested products. Excess N reduces the production of fruits in some varieties of tomatoes and reduces the sugar content and increases impurities in sugar beets. Due to such problems, Ayers and Westcot (1985) indicate that the restriction on the use of water increases as the NO3-N concentration increases from 5 mg/L to 30 mg/L.

IRRIGATION WATER QUALITY ASSESSMENTS

363

BORON Boron is essential for and potentially toxic to plants. Boron deficiencies take place in the (g/L concentration range in soil solutions. Boron toxicities take place at concentrations above a few mg/L for most plants. In the toxicity range, plants respond to B in the soil solution (Bs) rather than to B adsorbed on soil particles. Hence, solution and sand-culture data are used to evaluate the response of plants to B. Bingham et al. (1985) and Francois (1984) demonstrated that yield decreases related to B toxicity can be fitted to the two-parameter model used to describe salt tolerance (Maas and Hoffman 1977). The expression for this model is Y  100  m (x  A)

(11-2)

where Y  relative yield, m  the decrease in yield per unit increase in B concentration, A  the maximum concentration of B that does not reduce yield (threshold); and x  the B concentration in the nutrient, sand culture, or soil solution. Early recommendations and ranking of B tolerance of plants were largely based on visual symptoms. Francois (1984) showed that visual symptoms of B toxicity do not generally correlate with the yield of marketable product. Decreases in yield from B toxicity depend on the tolerance of the crop to B and on the Bs, which depends on the concentration of B in the irrigation water (Biw), the LF, and the departure from a steady-state relationship between adsorbed B and Bs. At steady-state input and output of B from the rootzone, the mean Bs is related to Biw and the LF, in the same manner as for salinity (Table 11-1), as shown in Fig. 11-7. Since B is adsorbed onto and released from the surfaces of soil particles, soil solutions are buffered against rapid changes in B concentration. If the B in irrigation water is increased, B is adsorbed, resulting in a smaller increase in the solute B concentration than the increase to irrigation water. The time required to reach a steady-state concentration of B depends on the increased B concentration, the amount of water used, the LF, and the sorption capacity of the soil volume of the rootzone. Jame et al. (1982) reported that the time ranged from 3 to 150 years. Three years was adequate for a sandy soil that can adsorb small quantities of B and has been treated with a B solution of 10 mg/L, and 150 years was required for a clay loam soil that could adsorb large quantities of B and was treated with a solution of 0.1 mg/L B. If B in the irrigation water is decreased, the soil releases B and time is needed to reduce the Bs. Also, as a rough approximation, the volume of low-B water needed to reduce the Bs from toxic to nontoxic levels is two to three times

364

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 11-7. Relationship between average rootzone boron (field capacity basis), boron in irrigation water, and LF required to avoid yield loss. Modified from Rhoades (1982).

greater than is needed for a comparable reduction in Cl when the pH is below 7.5. The ratio of concentration in the soil solution at field capacity to the concentration in the saturation extract is approximately 2 for anions not adsorbed or precipitated, such as Cl. However, the ratio is 2 for B because the adsorption of B on the soil surface depends on the concentration. When ET decreases the water content and concentrates the soil solution, B is adsorbed. The concentration factor decreases as the adsorption or buffer capacity of the soil increases. Jame et al. (1982) reported that this ratio ranged from 1.0 to 1.8, depending on the concentration of B and the adsorption capacity of the soil. Consequently, the B concentration in the saturation extract does not adequately represent or indicate B toxicity under field conditions. It is suggested that the Extract Chem model be used to convert B concentrations from one water content to another.

IRRIGATION WATER QUALITY ASSESSMENTS

365

TRACE ELEMENTS Trace elements are those that occur in waters and soil solutions at concentrations of less than a few mg/L, with most concentrations in the g/L range. Some are essential for plants and animals, but all can become toxic to plants and animals at elevated concentrations or doses. Virtually no experiments have yet been conducted to determine the criteria for quality for trace elements in irrigation water. Hence, guidelines have been based on results from sand, solution, and pot cultures, field trials with applications of chemicals, laboratory studies of chemical reactions, and animal feeding and grazing trials. Table 11-2 presents the recommended maximum concentrations of 15 trace elements. Shown for comparison are the U.S. Environmental Protection Agency drinking water standards (US EPA 1985, 2008). In most (but not all) instances, the drinking standards are lower than the recommended maximum concentrations for irrigation. The irrigation standards are designed to protect the most sensitive crops and animals that consume those crops from toxicities when the most vulnerable soils are irrigated. These concentrations should be considered as guidelines but not as criteria for water quality. If sufficient knowledge becomes available to show that these concentrations can be exceeded without adversely affecting soils, crops, and animals, then new guidelines can be established. For example, for the irrigated lands of the west side of California’s San Joaquin Valley, Pratt et al. (1988) recommended that the guideline for Se in the selenate form be increased to 0.10 mg/L and the guideline for Mo be increased to 0.05 mg/L. The conditions included alkaline, fine-textured soils; saline drainage waters, which need high LFs to prevent reduced yields; and drainage waters dominated by SO4 anions, which inhibit the absorption of Se and Mo by plants. These guidelines do not consider the long-term consequences on soil loading or the impact on discharge of drainage water to surface or subsurface water supplies. In most instances these environmental considerations are the limiting factor when using waters elevated in trace elements. Sprinkler irrigation may also result in increased trace element uptake. Other water-quality guidelines list the elements Al, Fe, Sn, Ti, and W (NAS 1973; Ayers and Westcot 1985), but limits for these elements have little meaning. If certain soil conditions develop, such as low pH for Al and highly reduced, waterlogged conditions for Fe, these elements can become toxic to plants due to the dissolution of Al or Fe from soil solids. Aerated soils with pH values above 5.5 will precipitate the Al and Fe in irrigation waters. When pH values are 7, the solubility of most trace metals is greatly reduced. Guidelines for tin (Sn) titanium (Ti), and tungsten (W) cannot be made due to insufficient information.

366

TABLE 11-2. Recommended Maximum Concentrations of 15 Trace Elements in Irrigation Waters for Long-Term Protection of Plants and Animals Recommended Maximum Concentrationa (mg/L) (2)

US EPA MClb Drinking Water (3)

Arsenic

0.10

0.01

This guideline will protect sensitive crops grown on sandy soils. Higher concentrations can be tolerated by some crops for short periods when grown in finetextured soils.

Beryllium

0.10

0.004

Toxicities to plants have been reported at concentrations as low as 0.5 mg/L in nutrient solutions and at levels in the soil greater than 4% of the cationexchange capacity.

Cadmium

0.01

0.005

Concentrations 0.01 mg/L will require 50 years or more to exceed the recommended maximum Cd loading rate. Removal in crops and by leaching will partially compensate and perhaps allow use of the water indefinitely.

Chromium

0.10

0.10

Toxicity in nutrient solutions has been observed at a concentration of 0.50 mg/L and in soil cultures at a rate of 120 kg/ha. Toxicity depends on the form of Cr existing in the water and soil and on soil reactions.

Cobalt

0.05

N.S.e

A concentration of 0.10 mg/L is near the toxic threshold for many plants grown in nutrient solution. Toxicity varies, depending on type of crop and soil chemistry.

Copper

0.20

1.3

Concentrations of 0.1 mg/L to 1.0 mg/l in nutrient solutions have been found to be toxic to plants, but soil reactions usually precipitate or adsorb Cu so that soluble Cu does not readily accumulate.

Fluoride

1.0

4.0

This concentration is designed to protect crops grown in acid soils. Neutral and alkaline soils usually inactivate F, so higher concentrations can be tolerated.

Element (1)

Comments (4)

Lead

5.0

Lithium

2.5c

Manganese

0.20

Molybdenum

0.01

Nickel

0.20

Selenium

0.02

0.05

This guideline will protect livestock from selenosis because of Se in forage. Selenium absorption by plants is greatly inhibited by SO4, so the guideline for this element can be increased for gypsiferous soils and waters.

Vanadium

0.10

N.S.e

Toxicity to some plants has been recorded at V concentrations above 0.5 mg/L.

Zinc

0.50

a

0.015

N.S.e

0.05d

N.S.e

Plants are relatively tolerant to Pb, and soils effectively sorb or precipitate it. Toxicity to animals typically is caused not by Pb absorption from soils but by aerial deposition of lead on foliage of pasture and forage plants. Most crops are tolerant to Li up to 5 mg/L in nutrient solutions. Citrus, however, is highly sensitive to Li. Lithium is a highly mobile cation that will leach from soils over an extended period of time. Some crops show Mn toxicities at a fraction of a mg/L in nutrient solution, but typical soil pH and oxidation-reduction potentials necessary for plant growth control Mn in the soil solution so that the Mn concentration of irrigation water is relatively unimportant. This concentration is below phytoxic level but is recommended to protect animals from molybdosis because of excess Mo in forages. Many plants show toxicity at Ni concentrations of 0.5 mg/L to 1.0 mg/l. Toxicity of this element decreases with increase in pH, so acid soils are the most sensitive.

5d

A number of plants show Zn toxicity at concentration of 1 mg/L in nutrient solution, but soils have a large capacity to precipitate this element. This guideline is designed to provide protection for acid sandy soils. Neutral and alkaline soils can accept much larger concentrations without developing toxicities.

367

Loading rates in kg/ha-yr can be calculated from the relationship that 1 mg/L in the water gives 10 kg/ha-yr when water is used at a rate of 10,000 m3/ha-yr b EPA maximum contaminant level, legal standards for public water systems (US EPA 2008) c For citrus, the maximum recommended concentration is 0.075 mg/L d EPA secondary maximum contaminant levels, voluntary standards for nonhealth-threatening elements e

No EPA standard

368

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SUMMARY To avoid the long-term accumulation of toxic amounts of waterborne substances in the rootzone of irrigated lands, the input of those substances to the soil from irrigation and other sources must not exceed the sum of losses from the soil and conversions to unavailable forms. Losses from the soil include plant uptake (5% to 10%) and leaching. In addition, there is an ongoing and reversible conversion of soluble, labile, and insoluble forms of minerals, which is affected by variables such as the oxygen content and pH of the soil water. Assessing the effects of irrigation water salinity and trace element concentrations on the suitability of a water supply for a given crop thus depends on the soil, crops, amount of water available, reference crop ET of the site (ET0), irrigation system, irrigator’s expertise in achieving the needed leaching, and decrease in yield that can be tolerated. It is recommended that the effects of irrigation water be tested directly on the soil of interest with column leaching studies, tests of aggregate stability, or tests of flocculation after the soil has been dispersed in a test tube.

REFERENCES Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Bingham, F. T., Strong, J. E., Rhoades, J. D., and Keren, R. (1985). “An application of the Maas-Hoffman salinity response model for boron toxicity.” Soil Sci. Soc. Am. J., 49, 672–674. Carter, M. R., Webster, G. R., and Cains, R. R. (1979). “Calcium deficiency in some solonetzic soils of Alberta.” J. of Soil Sci., 30, 161–174. Dong, A., Chesters, G., and Simsiman, G. V. (1983). “Soil dispersibility.” Soil Sci., 136, 208–212. Dontosova, K., and Norton, L. D. (2002). “Clay dispersion, infiltration, and erosion as influenced by exchangeable Ca and Mg.” Soil Sci., 167, 184–193. Francois, L. E. (1984). “Effect of excess boron on tomato yield, fruit size, and vegetative growth.” J. Amer. Soc. Hort. Sci., 109(3), 322–324. Goldberg, S., and Forster, H. S. (1990). “Flocculation of reference clays and aridzone soil clays.” Soil Sci. Soc. Am. J., 54, 714–718. Jame, Y. W., Nicholaichuk, W., Leyshon, A., and Campbell, C. A. (1982). “Boron concentration in the soil solution under irrigation: A theoretical analysis.” Can. J. Soil Sci., 62, 461–471. Kemper, W. D., and Koch, E. J. (1966). Aggregate stability of soils from western United States and Canada, U.S. Department of Agriculture Technical Bulletin 1335, U.S. Government Printing Office, Washington, D.C. Maas, E. V., Clark, R. A., and Francois, L. E. (1982). “Sprinkler-induced foliar injury to pepper plants: Effects of irrigation frequency duration, and water composition.” Irrig. Sci., 3, 101–109.

IRRIGATION WATER QUALITY ASSESSMENTS

369

Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. Drainage Div. ASCE, 103(IR2), 115–134. Marschner, H. (1995). Mineral nutrition of higher plants, 2nd ed., Academic Press, London. McNeal, B. L., Layfield, D. A., Norvell, W. A., and Rhoades, J. D. (1968). “Factors influencing hydraulic conductivity of soils in the presence of mixed-salt solutions.” Soil Sci. Soc. Amer. Proc., 32, 187–190. McNeal, B. L., Oster, J. D., and Hatcher, J. T. (1970). “Calculation of electrical conductivity from solution composition as an aid to estimation of soil salinity.” Soil Sci., 110, 405–414. National Academy of Sciences (NAS). (1973). Water quality criteria 1972, Environmental Studies Board Committee on Water Quality Criteria, U.S. Government Printing Office, Washington, D.C. Pratt, P. F., Albasel, N., Joseph, H., and Resco, C. (1988). Trace element guidelines for irrigation waters in the San Joaquin Valley, Report to the California Water Resources Control Board from the Dept. of Soil and Environmental Science, University of California. Quirk, J. P., and Schofield, S. P. (1955). “The effect of electrolyte concentration on soil permeability.” J. Soil Sci., 6, 163–178. Rhoades, J. D. (1982). “Reclamation and management of salt-affected soils after drainage,” in Proc., 1st Annual Western Provincial Conference on Rationalization of Water and Soil Resources and Management, Lethbridge, Alberta, Canada, November 27–December 2, 123–197. ———. (1984). “Using saline waters for irrigation,” in Scientific reviews on arid zone research, Scientific Publishers, Jodhpur, India, 2, 233–264. Shainberg, I., Rhoades, J. D., Suarez, D. L., and Prather, R. J. (1981). “Effect of mineral weathering on clay dispersion and hydraulic conductivity of sodic soils.” Soil Sci. Soc. Am. J., 45, 287–291. Shalhavet, J. (1994). “Using water of marginal quality for crop production: Major issues.” Agric. Water Mgmt., 26, 233–269. Shani, U., Ben-Gal, A., Tripler, E., and Dudley, L. M. (2007). “Plant response to the soil environment: An analytical model integrating yield, water, soil type and salinity.” Water Resour. Res., 43(8), 1–12. Suarez, D. L., Grieve, C. M., and Poss. J. A. (2003). “Irrigation method affects selenium accumulation in forage Brassica species.” J. Plant Nutr., 26, 191–201. Suarez, D. L., Rhoades, J. D., Lavado, R., and Grieve, C. M. (1984). “Effect of pH on saturated hydraulic conductivity and soil dispersion.” Soil Sci. Soc. Am. J., 48, 50–55. Suarez, D. L., and Sˇ imu˚nek, J. (1997). “UNSATCHEM: Unsaturated Water and Solute Transport with Equilibrium and Kinetic Chemistry.” Soil Sc. Soc. Am. J., 61, 1633–1646. Suarez, D. L., and Taber, P. (2007). “Extract Chem: Numerical software package for estimating changes in solution composition due to changes in soil water content,” http://ars.usda.gov/Services/docs.htm?docid14567, accessed February 3, 2011. Suarez, D. L., Wood, J. D., and Lesch, S. M. (2006). “Effect of SAR on water infiltration under a sequential rain-irrigation management system.” Agric. Water Mgmt., 86, 150–164.

370

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (2008). “Infiltration into cropped soils: Effect of rain and SAR-impacted irrigation water.” J. Environ. Qual., 37, S169–S179. U.S. Environmental Protection Agency (US EPA). (1985). Quality criteria for water, 1986, U.S. EPA, Office of Water Regulations and Standards, Washington, D.C. ———. (2008). “National primary drinking water regulations,” www.epa.gov/ safewater/contaminants/index.html#mcls, accessed February 3, 2011. U.S. Salinity Laboratory (USSL). (1954). Diagnosis and improvement of saline and alkali soils, USDA Handbook 60, USDA, Washington, D.C.

NOTATION A  salinity concentration at which growth depression (threshold) starts, or maximum concentration of boron that does not reduce yield B  percent of yield decrease per unit ECe above the threshold level Biw  concentration of boron in irrigation water Bss  boron in soil solution EC  electrical conductivity ECaw  average ECSS at field capacity ECe  electrical conductivity of soil saturation extract ECiw  electrical conductivity of irrigation water ECss  electrical conductivity of soil solution ESP  exchangeable sodium percentage Fc  concentration factor LF  leaching fraction m  decrease in yield per unit increase in boron concentration OP  osmotic potential x  boron concentration in the nutrient, sand culture, or soil solution Y  relative yield

CHAPTER 12 LEACHING AND ROOTZONE SALINITY CONTROL James E. Ayars, Glenn J. Hoffman, and Dennis L. Corwin

INTRODUCTION Successful water management for salinity control depends on adequate leaching, which takes place whenever irrigation and rainfall exceed the soil’s capacity to store infiltrated water within the crop’s rootzone. In humid regions, rainfall normally results in enough leaching to flush salt from the rootzone. In subhumid and drier regions, irrigation water that exceeds the crop’s water requirements may need to be applied to ensure adequate leaching. Depending on the salinity control needed, leaching may occur continuously or at intervals of a few weeks to a few years. The crop’s water requirement and salinity control must be prime considerations in places where salinity poses a hazard. Proper irrigation restores the soil’s water deficit without a wasteful and potentially harmful excess. Crops need water from irrigation and rainfall to control soil salinity by inducing drainage (leaching). As discussed in this chapter, leaching must remove enough salt to prevent it from accumulating in the rootzone beyond the crop’s salt-tolerance level. Chemical reactions in the soil affect the amount of salt leached, which may be greater than, equal to, or less than the amount of salt added by irrigation water.

WATER AND SALT BALANCE The amount of irrigation water needed to meet the crop’s water requirement can be calculated from a water balance of the crop rootzone. The major flows of water into the crop’s rootzone are irrigation, rainfall, and upward flow from the groundwater. The depths of each are expressed in 371

372

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

equations as Di, Dr, and Dg, respectively. Water flows out due to evaporation, transpiration, and drainage. Their equivalent depths are represented in equations as De, Dt, and Dd, respectively. The difference between the water flowing in and the water flowing out must equal the change in storage, the depth of which is expressed as Ds in equations. The water balance equation for storage change is as follows: Ds  Dr  Dg  Di  De  Dt  Dd

(12-1)

The equation for the change in salt storage, Ss, is the following: Ss  DrCr  DgCg  DiCi  Sm  Sf  DdCd  Sp  Sc

(12-2)

where C  salt concentration; the subscripts r, i, g, and d designate rain, irrigation, upward flow from groundwater, and drainage, respectively; Sm  the salt dissolved from minerals in soil; Sf  the salt added to soil as a fertilizer or amendment; Sp  the salt precipitated; and Sc  the salt removed in the harvested crop. If Dr  Dg  Di is less than De  Dt in Eq. 12-1, the crop water demand is met by extraction from soil storage and reduced drainage in the rootzone. As Ds is depleted, the soil dries, which reduces De, and Dt, and the crop becomes water-stressed. Initially, these processes bring water loss from the rootzone in balance with the water supply at zero drainage. However, without drainage, salt stored in the rootzone concentrates in the remaining stored water, which increases the osmotic stress on the plant, further reducing transpiration. If salts continue to increase in concentration, osmotic stress will reduce plant growth and may result in the plant dying. Alternatively, transpiration may be reduced to the extent that an irrigation results in excess water again being present in the profile and drainage commences (Solomon 1985). This drainage, in turn, carries salt out of the rootzone and the plant survives. The resulting leaching fraction (LF) is the absolute minimum at which the crop can extract water from a saline rootzone. This LF, however, is far less than that needed to prevent a reduced yield. When a shallow water table exists, deficiencies in Di  Dr may be offset by Dg. When flow is upward from the groundwater, drainage is zero and salt will not be exported from the entire rootzone. This situation cannot continue indefinitely. In the field, upward flow and drainage may take place alternately during the year. Typically, drainage takes place in the winter and early in the irrigation season, when the water requirements of the crop are low and rainfall or irrigation water applications are high. Upward flow takes place late in the irrigation season, when water requirements are high and rainfall and applications of irrigation water are insufficient to meet crop demand. If upward flow continues and suf-

LEACHING AND ROOTZONE SALINITY CONTROL

373

ficient leaching does not take place, soil salinity will ultimately reduce the crop’s water consumption to the point that the crop dies. Temporary use of water from soil storage beyond that normally removed between irrigations or from shallow groundwater is a useful strategy for managing water. However, over the long term and where salinity is a hazard, a net downward flow of water through the rootzone is needed to sustain crop productivity. One management strategy in shallow groundwater areas is to utilize preplant irrigation to leach the rootzone prior to planting (Ayars 2003). Rarely will conditions controlling the water that flows into and out of the rootzone last long enough for a true steady state to exist. As a result, the amount of salt stored in the rootzone fluctuates continually. The goal of water management is to maintain this fluctuation within limits that neither allow excess drainage nor reduce crop growth. Rainfall The concentration of salt in rainfall (Cr) varies according to distance from the ocean, topography, direction of the wind, intensity of rainfall, and geographical distribution of the storms. The annual deposit of salt from rainfall has been estimated at 100 to 200 kg/ha near the sea coast and 10 to 20 kg/ha in the continental interiors (Downes 1961; Cope 1958; Yaalon 1963). Although small, these deposits can add up to sizable amounts of salts in areas with low rainfall after several decades. In irrigated areas, the salt applied annually in the irrigation water normally far exceeds the salt contribution from rainfall. Thus, DrCr is normally assumed to be zero. Mineral Weathering Soils in arid and semiarid regions, except for ancient land masses, such as in parts of Australia, are relatively unweathered. Unweathered minerals provide plant nutrients but are also a source of soil salinity (Sm). Rhoades et al. (1968) have shown that increases in salt content of 200 to 300 mg/L are common when arid-land soil solutions remain in contact with relatively unweathered soil minerals for substantial periods of time. The amount of salt dissolved under such conditions depends on the level of carbon dioxide in the soil profile. The partial pressure of carbon dioxide can reach 10% or more when oxygen is consumed and carbon dioxide is released during soil respiration (Bohn et al. 1979). Studies using various simulated irrigation waters from the western United States (Rhoades et al. 1973, 1974) showed that the dissolution of primary minerals is most important when the irrigation water’s salt content is low—less than 100 to 200 mg/L—or when the LF is at least 0.25.

374

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

For example, irrigation with water from California’s Feather River, which has a salt content of 60 mg/L, results in more salt in the drain water due to weathering than due to the salt content of the irrigation water (Rhoades et al. 1974). The major concern about mineral weathering is the sodicity hazard of relatively low-salinity irrigation water. For salt-affected soils, mineral weathering is seldom a significant part of salt balance computations and Sm is generally assumed to be zero. Salt Precipitation As indicated in Eq. 12-2, the salt balance is affected by precipitation reactions (Sp) involving slightly soluble salts, such as gypsum, carbonates, and silicate minerals. Consequently, the amount of salt leached below the rootzone may be less than that applied, as was demonstrated in a three-year lysimeter study (Rhoades et al. 1974). When irrigation waters have a concentration of salt greater than 100 to 200 mg/L and if LFs are less than 0.25, some salts precipitate in the rootzone and are stored in the soil profile. When irrigation waters have a moderate amount of salt, such as the 800 mg/L that occurs in the Colorado River’s lower reaches, and LFs are below 0.25, salts precipitated in the soil profile exceed the amount weathered. Figure 12-1 shows the relative amounts of salt that may chemically precipitate or become soluble in water due to weathering for various types of irrigation water as a function of LF. All irrigation waters illustrated have concentrations of salt of above 500 mg/L. Thus, mineral weathering does not exceed chemical precipitation, except for some waters at an LF of 0.2 and above. At low LFs (LF  0.1), 20% or more of the salt in irrigation water precipitates and is not contained in the drainage water. Consequently, salt precipitation may be a significant part of calculating the salt balance when the LF is low for some water. Salt Removal by Crops Salt removed by agronomic crops (Sc) is insufficient to maintain salt balance. The average amount of salt contained in mature crops of alfalfa, barley, corn silage, Sudan grass, and sweet clover grown in Texas’s Rio Grande area was 3.6% (Lyerly and Longenecker 1962). At intermediate levels of salinity, Chapman (1966) reported that the salt content of alfalfa, corn, and sorghum was about 3% of the dry tissue weight. In one study, water from the Pecos River with an electrical conductivity (ECi) of 3.3 dS/m was applied to alfalfa grown on sandy loam soil. The amount of salt removed in the harvests was 3% of the dry forage for LFs varying from 0.1 to 0.3 (Rhoades et al. 1974). Assuming a depth of application of 2 m and a total of 2,112 ppm salt for the irrigation water, the applied salt load is approximately 40 Mg/ha; assuming a yield of 17 Mg/ha, the

LEACHING AND ROOTZONE SALINITY CONTROL

375

FIGURE 12-1. Net contribution of mineral weathering and salt precipitation to salt content of drainage water, expressed as a percent of the salt applied from various river waters. Each line represents an average of the percentages for calcareous and noncalcareous Pachappa sandy loam soil. From Rhoades et al. (1974) with permission from the American Society of Agronomy. removed salt is approximately 0.5 Mg/ha (about 1% of the applied salt). Francois (1981) reported a salt content of 3% to 4% for alfalfa grown under saline conditions. Plants that are very efficient in removing salt from saline soils, such as sea-blithe (Suaeda fruticosa), remove less than 3 Mg/ha with each harvest (Chaudri et al. 1964). Under most agricultural

376

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

conditions where salinity is a concern, salt removal by crops can be ignored in the salt balance equation. Fertilizer Salts The upper limit of recommended fertilizer applications for crops, such as corn, is about 250 kg/ha of nitrogen (Kearney et al. 1980). If the nitrogen is applied as ammonium sulphate (21% N, 73% SO4), the amount of fertilizer applied may be as high as 1.2 Mg/ha. When only the sulphate contributes to the salt load, 0.9 Mg/ha is the upper limit. If this amount of fertilizer were added to corn irrigated by 750 mm of water with 800 mg/L of salt, the fertilizer’s contribution of salt would be 15% of the amount added by irrigation. The amount of fertilizer in this example is considered excessive for many crops. While fertilizer salt may not be inconsequential, it is not routinely included in the salt balance calculation.

LEACHING REQUIREMENT Salts in irrigation water accumulate in the rootzone as a consequence of the extraction of nearly pure water by plant roots leaving residual salts behind. The resulting salinity profile typically increases in salt concentration with depth. The salts residing in the rootzone can detrimentally affect plant productivity due to (1) osmotic effects that limit plant water uptake, (2) specific-ion toxicity, (3) plant nutrient imbalance, and (4) influences on soil physical properties such as permeability and tilth. The concept of a leaching requirement (Lr) grew out of the need to control salinity in the rootzone. The U.S. Salinity Laboratory investigators (George E. Brown Jr. Salinity Laboratory) developed the concept of Lr in the early 1950s as an irrigation management tool to control salinity affecting plant growth. Leaching requirement is based on the concept of leaching fraction (LF), which is defined as the fraction of infiltrating water that moves beyond the rootzone and is a measure of the level of leaching of salts. As the LF increases, the concentration of salts in the rootzone and concomitantly the electrical conductivity (EC) decreases. Quantitatively, LF is defined by Eq. 12-3: LF 

Dd Ca ECa   Da Cd ECd

(12-3)

where Dd (mm) and Da (mm) are the depths of drainage water and infiltrating applied water, respectively; Ca (mg/L) and Cd (mg/L) are the salt contents of the applied and drainage water, respectively; and ECa (dS/m) and ECd (dS/m) are the electrical conductivities of the applied and

LEACHING AND ROOTZONE SALINITY CONTROL

377

drainage water, respectively. Leaching requirement was originally defined by the U.S. Salinity Laboratory researchers (1954) as the fraction of water infiltrating the soil that must move beyond the rootzone to prevent soil salinity from exceeding a specified value. The Lr represents the minimum LF that will adequately leach salts in the rootzone to a level that does not measurably reduce crop yield; consequently, the rootzone salin* ity level is the maximum permissible salinity level of ECdw (i.e., ECdw ) that will still result in optimum plant growth. Quantitatively, the original Lr model as defined by the U.S. Salinity Laboratory (1954), which assumes steady-state conditions, is represented by Eq. 12-4: Lr 

Ca ECa  Cd* ECd*

(12-4)

where Cd* is the maximum permissible salt content of the drainage water. Equation 12-4 must still include a relationship between plant response and EC of the bottom of the rootzone to be useful in determining the required leaching level. It has generally been assumed that the plant responds to the linearly averaged rootzone EC of the saturation extract (ECe) (Shalhevet and Bernstein 1968; Shalhevet et al. 1969), which is an assumption derived from early salt-tolerance experiments that were conducted at extremely high LFs, resulting in fairly uniform salt concentrations throughout the rootzone. Rhoades (1974) introduced an estimate of ECd* with Eq. 12-5: ECd*  5 ECe*  ECi

(12-5)

where ECe* (dS/m) is the linearly averaged rootzone EC of the saturation extract for a given crop appropriate to the tolerable degree of yield depression (usually 10% or less) and equivalent to the plant salt tolerance threshold EC values as defined by Maas (1990) and Maas and Hoffman (1977), and ECi is the EC of the irrigation water. Substitution of Eq. 12-5 into Eq. 12-4 yields Eq. 12-6, which ties Lr to irrigation water salinity and crop salt tolerance and is referred to as the Rhoades Lr model: Lr 

ECi 5ECe*  ECi

(12-6)

Hoffman and van Genuchten (1983) developed a steady-state model that determined the linearly averaged, mean rootzone salinity by solving the continuity equation for one-dimensional vertical flow of water through soil, assuming an exponential plant water uptake function. The

378

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT —

linearly averaged salt concentration of the rootzone (C ) as a ratio of the salt concentration of the irrigation of water (Ci) is C 1    ln[LF  (1  LF )e z/ ] Ci LF Z(LF )

(12-7)

where LF is the leaching fraction; Z is the depth of the rootzone, and  is an empirical constant set to 0.2 Z. Figure 12-2 shows the Lr as a function of salinity of the applied irrigation water and salt tolerance based on the Hoffman–van Genuchten model. Other steady-state models of Lr have been developed by Ayers and Westcot (1976) and Rhoades (1982). Hoffman (1985) compared calculated leaching requirements from the

FIGURE 12-2. Leaching requirement (Lr) as a function of the salinity of the applied irrigation water (ECi) and plant salt-tolerance threshold (EC*e ). From Hoffman and van Genuchten (1983) with permission from the American Society of Agronomy.

LEACHING AND ROOTZONE SALINITY CONTROL

379

Rhoades, Hoffman–van Genuchten, and two other steady-state models. Of the four models tested, the Hoffman–van Genuchten model agreed well with the measured values throughout the range of Lr of agricultural interest. These models can be used not only to determine the Lr but also to determine the maximum irrigation water salinity that can be applied to a crop for a specific LF. Table 12-1 provides a comparison of the estimated maximum irrigation water salinity (i.e., ECi) that could be used to grow tomatoes with an EC*e of 2.5 dS/m and LFs of 0.05, 0.10, 0.15, and 0.20 using the Rhoades and Hoffman–van Genuchten Lr models. The data for the Hoffman–van Genuchten model are shown in Fig. 12-2. At all LFs the Hoffman–van Genuchten model indicated that higher levels of salinity could be used for irrigation without loss compared to the Rhoades model. The Hoffman–van Genuchten model is in closer agreement with transient models (see Chapter 26 of this manual) than other steady-state models, which are too conservative in the quality of irrigation water that can be used without reducing yields. The aforementioned Lr models, including the Rhoades and Hoffman– van Genuchten models, only consider salt tolerance of the crop grown and salinity of the irrigation water while assuming steady-state conditions. However, steady-state conditions do not exist under most field situations. This is because there are commonly occurring factors that cause perturbations to steady state, including rainfall, crop rotations, alteration of the irrigation management strategy, variation in irrigation water quality, and variations in soil profile water content and salinity resulting from variations in plant root water uptake. In addition, Lr is influenced by numerous factors, including salinity of applied water, crop salt tolerance, precipitation-dissolution reactions, TABLE 12-1. Estimated Maximum Irrigation Water Salinity That Could Be Used to Grow Tomatoes with a Salt Tolerance Threshold of 2.5 dS/m for Various Leaching Requirements (LRs) as Calculated from the Rhoades and Hoffman–van Genuchten LR Models Maximum Irrigation Water Salinity (dS/m) Leaching Fraction (1)

Rhoades Model (2)

Hoffman and van Genuchten Model (3)

0.05 0.10 0.15

0.6 1.1 1.6

0.7 1.3 2.0

0.20

2.1

3.0

380

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

transient root water uptake distributions, preferential flow, climate, runoff, extraction of shallow groundwater, and leaching from effective precipitation, as well as the questionable appropriateness of the assumption of steady-state conditions. Based on the exclusion of these factors from consideration, recent publications by Corwin et al. (2007) and Letey and Feng (2007) have shown that the steady-state Lr models are conservative, suggesting that a new paradigm may be needed, particularly for research applications. Chapter 26 provides a detailed discussion of the appropriateness of transient Lr models over steady-state Lr models and demonstrates that models that can account for additional processes influencing Lr will provide less conservative estimates (i.e., Lr estimates are lower). For general applications, the two existing models presented here will be adequate for water management. Accounting for nonuniformity of irrigations to estimate Lr has not been addressed to date. If the Lr is not met everywhere in the field, salinity will increase wherever ET plus the Lr is not met. Whether to apply enough water to ensure that the Lr is met throughout the field or to accept some reduction in yield in parts of the field, rather than overirrigate most of the field, must be determined. Adopting advanced irrigation technologies and implementing advanced management alternatives are needed to approach the goal of achieving the Lrs. Inefficient irrigation inadvertently provides excessive leaching, which is costly and leads to a loss of water, energy, and nutrients; deteriorates the quality of groundwater; and increases the need for drainage facilities. Consequently, knowing the Lrs of crops and striving to attain them is vital.

EFFECT OF SHALLOW GROUNDWATER The upward movement of shallow saline groundwater and its subsequent evaporation at the surface of the soil adds to the salination of soils. Drainage systems are generally used to manage the water table depth to minimize the rate at which salt accumulates and, thus, reduce the salinity hazard (USBR 1993). The effects of the water table depth and the soil properties on the rate of upward movement must be known to determine the depth at which to maintain the water table. This information is also needed to estimate the amount of groundwater available to plants from upward movement (Ayars et al. 2006). Starting from saturation, the drying rate of the surface of the soils will at first be limited by the atmospheric evaporative conditions. When the surface becomes dry enough, the evaporation rate will be limited by the rate of water movement to the surface in the liquid phase. As the soil dries further, vapor movement is possible but relatively unimportant, particularly since diurnal fluctuations in temperature may cause the vapor

LEACHING AND ROOTZONE SALINITY CONTROL

381

movement to reverse directions. The length of the period from rapid drying to the vapor phase depends on depth of application of water, soil type, and the presence of vegetation. When a shallow water table exists, upward flow becomes important in the salination process. Gardner and Fireman (1958) studied how the rate of upward flow relates to the water table depth in a fallow area. This study verified the steady-state solutions proposed by Gardner (1957), who based his solutions on the relation between hydraulic conductivity (k) and soil matric potential (suction, S) of the form k

a S b n

(12-8)

where a, n, and b are constants. For many soils, values of n equal to 2 or 3 fit best with experimental data. For Chino clay, k  1,100/(S2  565) cm/d, where S is in mbars. For Pachappa sandy loam, k  32/(S3  104  2.6) cm/d. Figure 12-3 gives the theoretical maximum rate of upward

FIGURE 12-3. Maximum theoretical rate of upward water flow for chino clay and Pachappa sandy loam as a function of the depth of the water table.

382

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

flow from the water table for these two soils as a function of the depth to the water table. Two maximum rates of flow must be considered: the potential rate of evaporation from the soil surface dictated by the atmospheric conditions, and the maximum rate at which water can be transmitted upward from the water table based on soil hydraulic properties. Obviously, the lesser of these limits upward movement. Excluding a shallow water table and humid conditions, the water-transmitting properties of the soil most often limit upward flow. This type of analysis can be used to select the depth at which a water table should be maintained to keep a desired upward flow. In the past the goal was to maintain a minimum upward flow. Using the data in Fig. 12-3, lowering the water table from the surface to a depth of about 1 m would be of little benefit in most soils. Upward flow at these shallow depths could exceed 2.5 mm/d for clay soils (Fig. 12-3) and be even greater for coarser-textured soils, depending on the atmospheric evaporative demand. As the water table is lowered below 1 m, the soil’s hydraulic properties and depth limit the rate of upward flow (Fig. 12-3). Lowering the water table from 1.2 to 3.0 m in Pachappa sandy loam decreased upward flow by a factor of 10. When the water table is at 2.5 m, further lowering reduces upward flow only slightly. Upward movement and evaporation of water from the surface of the soil is possible even with a water table that has a depth of 10 m. Harmful amounts of soluble salts could slowly accumulate in the upper part of the soil profile if the groundwater is sufficiently saline and rainfall and irrigation amounts are inadequate. These results, verified by field observations, have led to the installation of most subsurface drainage systems at depths of 1.5 to 2.5 m wherever salinity poses a hazard. This is reflected in the recommendation provided in the drainage design manual developed by the U.S. Department of the Interior’s Bureau of Reclamation (USBR 1993). In the past, the recommendations for drain placement were made at a time when drainage systems ran continuously with little concern for the environment. This practice is no longer environmentally practical and has resulted in modifications of the criteria used to design drainage systems (Grismer 1990; Guitjens et al. 1997; Ayars et al. 1997). The current thinking with regard to design of drainage systems includes a water quality criterion such that the drain placement is shallower with resulting narrower lateral placement than in the past (Ayars et al. 1997). The analysis of drain flow lines indicates that deep placement of laterals results in deep flow lines that mine salt from deep within the soil profile (Jury et al. 2003). This results in excess salt being discharged into the environment with minimal effect on the salinity in the rootzone. Shallow placement of the drain lines results in shallower flow lines and reduced salt loading. An alternative to modifying the drain spacing and depth is to provide drainage controls, which also reduces the depth of the flow lines

LEACHING AND ROOTZONE SALINITY CONTROL

383

to the laterals. This has resulted in a need to have an active management of a drainage system, where in the past that management has been passive and the flow has been continuous (Ayars 2003). The inclusion of drainage system controls has also dictated a change in the orientation of the drainage system laterals to be perpendicular to the surface grade rather than parallel to the surface grade (Ayars 1999). The modifications in the drainage system design have also prompted a change in thinking with regard to salinity management. In the past the objective has been the nearly complete removal of salinity from the crop rootzone to levels that would be adequate for most crops. With the need to minimize the environmental impact of salinity and trace elements and other pollutants in the drainage water, the goal now is to only remove the absolute minimum level of salt needed to sustain production of the selected crop. The objective is one of salt management, as well as water management, within the rootzone. Water supplied to a crop by capillary rise from shallow groundwater can be an important resource. Benefits of using this water include reduced irrigation, lower production costs, movement of a more moderate amount of groundwater to deeper aquifers, and a decreased amount of groundwater that needs to be disposed through subsurface drainage systems (Ayars et al. 2006). The distribution of salts in the soil profile above the water table depends on the groundwater’s depth and salt content, the amount of applied water and its salt content, the water uptake pattern of the crop’s root system, and whether the water table is controlled. The flux to the rootzone will be determined by the unsaturated soil hydraulic conductivity, which is determined by the soil type, and the soil matric potential gradient established in the soil profile as a result of both crop water use and evaporation from the soil surface. Soil water flux is often computed in one dimension using Darcy’s law, as shown in Eq. 12-9: hz

z  ∫0

dh 1  q  k( h)

(12-9)

where z is the distance between the water table and a position in the soil profile with a constant flux of q. The hydraulic conductivity (k) is given as function of the matric potential (h). Since the unsaturated hydraulic conductivity is a function of the soil type, it is apparent that the soil type is a dominant factor affecting the flux from the water table to a plant. The closer the rootzone is to the water table, the higher will be the potential crop water use, since it is possible to maintain the flux at a higher rate over a shorter distance. There is still the problem of creating the gradient needed to move water up in the profile. It has been demonstrated that plants will take water from the areas of the soil profile with the highest potential energy, so the higher the soil water content in the rootzone, the

384

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

lower is the potential for use from shallow groundwater. This means that the soil in the rootzone has to be dried out sufficiently to create an upward gradient. The gradient is also affected by the osmotic potential in the soil water and groundwater. Several formulas have been derived for estimating flow from a water table to fallow and crop land. Equation 12-9 was simplified and solved analytically by using an exponential form for the hydraulic conductivity function for the soil being studied. The maximum steady state flux then becomes qm  Aebz

(12-10)

where qm is the flux (cm/d), A and b are regression coefficients related to the soil properties, and z is the depth (cm) to the water table (Ragab and Amer 1986). Use of this expression gives an indication of the potential crop water use for the given conditions. Other research (Grismer and Gates 1988) has indicated that upflux (qu) from the water table may be adequately represented by qu  a  bD

(12-11)

where a and b are empirical coefficients that depend on the soil hydraulic parameters, and D is the depth to the watertable. The values for a are highly variable, while the values for b depend only on the soil type. Grismer and Gates (1988) demonstrated the application of this equation for cotton water use from shallow groundwater on three different soil types. The regression equations for water use by cotton from shallow groundwater in different soils are shown in Fig. 12-4. The data demonstrate that for a given depth to the water table, the percentage of water extracted from the water table is reduced as the soil clay content increases. This is a consequence of a reduction of the unsaturated hydraulic conductivity in finer textured soil. The data also show that for a given soil type an increase in the depth to the water table results in a reduction of crop water use from the shallow groundwater, as predicted in Eq. 12-11. Wu et al. (1999) modeled crop water use from shallow groundwater with an empirical model developed by W. S. Meyer that captures the interaction of soil water content, root development, crop water requirement, and soil type. The equation is ⎛ ⎞ ⎜ ⎟ ⎜ ⎟ a qu  ⎜ ⎛ z R ⎞ ⎟ * ET c ⎞⎟ ⎜ b ⎜⎜⎝ zmax ⎟⎟⎠ ⎛ x0.01 ⎜1  e ⎟⎟ ⎜e ⎝ ⎠⎠ ⎝

(12-12)

LEACHING AND ROOTZONE SALINITY CONTROL

385

FIGURE 12-4. Contribution of shallow, saline ground water to evapotranspiration (ET) of cotton as a function of soil type and depth to water table. From Grismer and Gates (1988). © 1988 The Regents of the University of California. where qu is upflux (mm/d); a, b, and c are regression coefficients; ZR is the depth from one-third of the depth of the rootzone to the groundwater level (m); Zmax is the threshold water table depth below which upflow would be less than 1 mm/d as defined by Talsma (1963) (m); and x is the relative water content described by the relation x

 s  avg  s  l

(12-13)

where s is saturated water content (cm3/cm3), l is lower limit of plant available water (cm3/cm3), and avg is average water content of the unsaturated layer. The values suggested by Wu et al. (1999) for the

386

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

regression coefficients are a  3.9, b  3.8, and c  0.5. The suggested values for Zmax are soil-dependent and vary from 1.5 m for coarse sand, to 6 m for sandy clay loam, and to 1.5 m as the clay content increases beyond the sandy loam texture. The Zmax indicates the upflux potential for the soil type and should be related to hydraulic conductivity, air entry value, and soil water retention curve for a certain soil. Wu et al. (1999) provided a graph of the proposed values for Zmax. The bracketed coefficient in Eq. 12-12 represents the percentage of shallow groundwater that is used to meet crop ET. The ultimate salinity distribution in the soil profile will depend on whether the water table was static, as in a lysimeter study, or was dynamic, as would be found in field studies. In a lysimeter study in Texas, researchers studied soil salinity profiles in a Willacy fine sandy loam above a shallow water table (Namken et al. 1969). The study consisted of two water treatments and three depths of water tables. Because differences in soil salinity between the water treatments were small, Fig. 12-5

FIGURE 12-5. Soil salinity distribution for different ground water salinities and depths of groundwater. From Namken et al. (1969) with permission from the American Society of Agronomy.

LEACHING AND ROOTZONE SALINITY CONTROL

387

illustrates only the influence of the groundwater’s depth and salt content. During the study’s first year, the groundwater had a level of salinity (ECGW) of 6 dS/m to 8 dS/m. During its last three years, the ECGW ranged from 0.9 dS/m to 1.6 dS/m. The cotton crop took 57%, 38%, and 28% of the water used when the water table was at depths of 0.9 m, 1.8 m, and 2.8 m, respectively. When the water table was 1.8 m deep or lower, the upper half of the profile remained nonsaline, while the lower half became salinized. When the depth was 0.9 m, the groundwater’s level of salinity influenced the entire profile. Cotton grown on a loam soil in the San Joaquin Valley of California with a water table located 2.0 m to 2.5 m below the surface received at least 60% of its ET from shallow groundwater with an EC of 6 dS/m (Wallender et al. 1979). The fewer the irrigations, the more the groundwater contributed to ET. However, the yields of lint were reduced. Figure 12-6 illustrates how cotton’s use of groundwater affected soil salinity. Concentrations of soil Cl from early in the irrigation season (July 5) are compared with concentrations after harvest (November 28). The equivalent depth of water used in ET from the groundwater equaled 362 mm. This was based on concentrations of soil Cl and the concentration of Cl in the groundwater (17.4 mol/m3) and soil bulk density. The amount agreed with the contribution of groundwater based on the soil profile’s water budget.

FIGURE 12-6. Seasonal change in soil chloride levels as a function of soil depth and ET (in percent) from groundwater. From Wallender et al. (1979) with permission from the American Society of Agronomy.

388

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Recent field studies have demonstrated that it is possible to manage salinity in the rootzone both with and without the presence of a drainage system. In a long-term saline water reuse study, Ayars et al. (1993) demonstrated that preplant irrigation and rainfall were adequate to restore salinity levels to the spring conditions (Fig. 12-7). In a separate study Ayars (2003) demonstrated that it was possible to manage the soil salinity in the rootzone within limits to permit production of tomato and cotton provided the regional groundwater flow and vertical drainage were adequate to reduce the groundwater to a depth of 1.5 m during the fallow period (Fig. 12-8). Use of groundwater by alfalfa and corn varies from 15% to 60% of the total seasonal use, but the data are too inconsistent to establish a relationship. Use of groundwater by alfalfa from a water table with a depth of 0.6 m in the Grand Valley of Colorado (Kruse et al. 1985) varied from 46% to 94% of the total seasonal use in two different years, when ECGW equaled 0.7 dS/m. It varied from 23% to 91% of the total seasonal use between years, when ECGW equaled 6 dS/m. In the same study, Kruse et al. (1985) reported that corn obtained 52% to 68% of its seasonal water requirement when the water table was 0.6 m deep and obtained 25% to 32% of its seasonal water requirement when the water table was 1 m

FIGURE 12-7. Distribution of soil electrical conductivity (EC) under drip plots (D1, D2) irrigated with saline (6 dS/m) water and furrow-irrigated plots (F1, F2) irrigated with low-salinity (0.4 dS/m) water. From Ayars et al. (1993).

LEACHING AND ROOTZONE SALINITY CONTROL

389

FIGURE 12-8. Distribution of soil electrical conductivity (EC) under drip (A) and furrow (B) irrigated plots in a field with shallow ( 2 m) saline (6 dS/m) groundwater. From Ayars (2003). deep. The proportion of use remained unaffected when ECGW varied from 0.7 dS/m to 6 dS/m. Soils with a shallow water table frequently depress yields due to reduced soil aeration and inhibited root extension. If the shallow groundwater is saline, yields may be further reduced. Hanson et al. (2006) demonstrated that it is possible to grow tomatoes in areas with shallow saline groundwater without a loss in yield using a subsurface drip irrigation (SDI) system with a high irrigation frequency. The drip system operation provided adequate leaching around the drip line, which enabled growth. Using SDI or surface drip provided improved control of irrigation application and reduced deep percolation losses, which enables production.

390

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Subsurface drainage benefits crop production in salt-affected soils, but few long-term drainage experiments have been conducted that quantify increased yields and reduced salinity. El-Mowelhi et al. (1988) undertook one such experiment in the Nile Delta of Egypt from 1976 to 1986. Soil salinity to a depth of 1.5 m was reduced from an average of 5.3 dS/m to 2.2 dS/m after 1 year of drainage (Fig. 12-9) without additional water being applied beyond the normal irrigation amounts and rainfall. For three crop rotations, subsurface drains spaced 20 m apart and placed 1.5 m deep in clay soil increased the yield of cotton and rice by 100%, and the yield of wheat and Berseem clover by 50%.

SOIL SALINITY WITHOUT LEACHING Leaching in the context of this chapter implies that salt is removed from the rootzone and then is eventually removed from the soil profile. The following examples demonstrate the results when this process is not completed.

FIGURE 12-9. Influence of subsurface drainage on average soil salinity profile over 10 years. From El-Mowelhi et al. (1988) with permission from Elsevier.

LEACHING AND ROOTZONE SALINITY CONTROL

391

An example of the effect of virtually no drainage on soil salinity is found in southwest Australia (Peck et al. 1981). Figure 12-10 illustrates the Cl concentration and the downward velocity of the soil solution for one location in a Mediterranean climate with an annual rainfall of 800 mm on a native Eucalypt forest. Chloride concentration increased most at a soil depth of 7 m, where the downward velocity of the soil solution equaled 0.04% of the annual rainfall, or 0.3 mm/yr. Below 7 m, Cl decreased linearly to less than 2,000 mg/L just above the water table at a depth of 17 m. In this case the salt is moving slowly to the groundwater. The deeper the soil, the greater the capacity to store salt with minimal yield reduction. One of the first studies of the effect of no leaching involved alfalfa grown in a greenhouse and irrigated by water with an electrical conductivity (ECi) of 1 dS/m. The plants were grown without leaching in sandy loam soil profiles with depths of 0.6 m, 1.2 m, and 1.8 m for periods of 9, 14, and 20 months, respectively (Francois 1981). Yield was reduced less than 25%, yet 14 Mg/ha, 30 Mg/ha, and 45 Mg/ha of salt, respectively were stored in the lower portions of the three different soilprofile depths. Drastic reductions in yield took place when the salt began to build up in the upper portion of the rootzone. This study demonstrated

FIGURE 12-10. Soil chloride concentration and the downward rate of soil-water movement as a function of soil depth in poorly drained soils of southwest Australia. From Peck et al. (1981) with permission from Elsevier.

392

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

that, regardless of soil depth, alfalfa can be grown for a considerable period of time without removal of salt from the rootzone if the upper part of the rootzone is maintained at a low level of salinity. In this case, the upper part of the rootzone was being leached with each irrigation. However, the salt accumulation in the bottom of the rootzone resulted in transport of salt to the surface and the ultimate salination of the entire rootzone. The Broadview Irrigation District, which was located on the west side of the California’s San Joaquin Valley, is a well-documented example of the effect of accumulating soil salinity on a large scale (Wichelns et al. 1988). The district was made up of 4,000 ha of field crops that were irrigated with water containing approximately 300 mg/L of salt (ECi of 0.5 dS/m) starting in 1957. To facilitate leaching, subsurface drains were installed on more than 80% of the irrigated land. The district had no drainage outlet until 1983, so it blended its surface runoff and subsurface drainage water with irrigation supply water. The ratio of drain water to fresh water increased from near zero in the early 1960s to about half in the early 1980s, when the mean salt content of the drainage water was about 2,800 mg/L. Although the fields were leached, the salts were reapplied to the fields. Thus, no disposal of salts took place. Crop selection switched to salt-tolerant crops, such as cotton, to maintain yield, while the amount and yield of more salt-sensitive crops, such as tomatoes, dropped drastically as soil salinity increased over time. Eventually, a drainage outlet was established and the disposal of excess salt resulted in a change back to more salt-sensitive crops. The presence of Se in the drainage water resulted in the loss of drainage water disposal alternatives and saline water was again recycled within the district (Wichelns et al. 2002). The lack of a drainage water disposal site resulted in the closure of the district and the fallowing of all the land.

INTEGRATION OF SOIL SALINITY BY CROPS In the field, the distribution of salts is neither uniform nor constant. Water and salt management strategies will require an understanding of the plant responses to salinity, which varies according to time and the soil depth. These responses must be known to apply the results from experiments on the salt tolerance of crops. The following sections will illustrate plant response to salinity variation with depth and time. Integration with Soil Depth Hoffman et al. (1983) conducted a field experiment to establish the salt tolerance of corn in the Sacramento-San Joaquin delta of California using two irrigation methods. One consisted of mini-sprinklers, each with a wetted diameter of about 4 m, spaced 1.5 m apart along laterals in every

LEACHING AND ROOTZONE SALINITY CONTROL

393

other row of corn. Water was applied uniformly to achieve about 50% leaching. Figure 12-11 illustrates the resulting soil salinity profile for the nonsaline treatment (ECi equaled 0.2 dS/m) and a saline treatment (ECi equaled 6 dS/m). Figure 12-11 gives the values of soil salinity for measurements from soil samples, soil water samples extracted by vacuum through suction cups, and the monitoring of direct-burial, four-electrode salinity probes (Rhoades 1979). Figure 12-11 also gives the composite values from these three techniques. The linear averages of the composite values through the rootzone are 1.9 dS/m for the nonsaline treatment and 7.3 dS/m for the saline treatment.

FIGURE 12-11. Time-weighted averages of EC of soil water from suction cups, salinity probes, and soil samples and composite averages for 0.2 dS/m and 6 dS/m saline water applied to corn, Sacramento-San Joaquin delta, 1981. From Hoffman et al. (1983) with kind permission of Springer ScienceBusiness Media.

394

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The second method was subirrigation, which is the one most commonly used method in the delta. It consisted of ditches, spaced every 16 rows of corn, dug approximately 15 cm wide and 60 cm deep by a trencher each year in mid June. Irrigation water applied in the ditches moved horizontally and vertically through the soil profile, raising the shallow water table to about 15 cm from the surface of the soil. Figure 12-11 gives the salinity profiles for the same treatments as for the sprinkled plots. These profiles are representative of those expected in situations with no irrigation, low rainfall, and shallow, saline groundwater. The linearly averaged values for the composite salinity profiles were 3.0 dS/m for the 0.2-dS/m treatment and 8.6 dS/m for the 6-dS/m treatment. When the linearly averaged values for these treatments and other levels of salinity tested during the 3-year experiment are plotted, the salt tolerance response curves for the sprinkled and subirrigated treatments do not differ statistically (Fig. 12-12). This suggests that plants respond to a linear average of

FIGURE 12-12. Relative grain yield of corn grown in the Sacramento-San Joaquin delta as a function of soil salinity for sprinkler irrigation and subirrigation water application methods. From Hoffman et al. (1983) with kind permission of Springer ScienceBusiness Media.

LEACHING AND ROOTZONE SALINITY CONTROL

395

the salinity values through the rootzone and that the salt-tolerance coefficients apply where the salinity distribution is not uniform with the depth of the soil. The response of corn to salinity when grown on organic soil agrees with the response of peanuts (Shalhevet et al. 1969) and tomatoes (Shalhevet and Yaron 1973) grown on mineral soils to salinity. Integration over Time Soil salinity is typically monitored at the beginning and the end of the crop’s growing season, and the mean soil salinity is determined by averaging the values. In experiments, soil salinity is normally monitored more often and by a combination of soil sampling, vacuum extraction of soil water, and various devices that measure salinity. The integration of soil salinity over time is difficult because sensitivity varies from one stage of growth to the next for some crops. Cereal crops seem particularly variable. Results indicate that corn, for example, is most sensitive during the vegetative stage (Maas et al. 1983). Although soil salinity delayed the emergence of seedlings of corn, salinity of up to ECSW of 9 dS/m did not reduce the emergence of seedlings after six days of germination. Increasing the salinity of the irrigation water to 9 dS/m at the tassel or grain-filling stages did not decrease the yield of corn ear or grain significantly below that achieved where soil salinity was constant throughout the growing season. Bernstein and Pearson (1954) compared the influence of a constant level of soil salinity with cycles of slowly increasing and then abruptly decreasing levels of soil salinity. They reported that peppers responded to the seasonal mean soil salinity, whereas tomatoes were more affected by periods of high soil salinity. Meiri and Poljakoff-Mayber (1970) noted from different salinity experiments that the relationship between salinity and relative leaf area was linear. Plant response to mean seasonal soil salinity is probably a reasonable estimate unless soil salinity during the season ranges both lower and higher than the salt-tolerance threshold for the crop or unless salinity occasionally exceeds the range over which linear salt-tolerance response is observed, as probably was the case for tomatoes. Evaluating the response of perennial crops to salinity over time is more complex than evaluating the response of annuals. This is primarily due to the extended length of time during which the yield of a perennial crop may be affected by soil salinity. With this increased time period come problems of how to compensate for dormant periods, drastic weather changes, such as monsoons and winter rains, and large changes in atmospheric evaporative demand. Deciduous fruit trees exemplify a perennial crop whose response to salinity over time is difficult to assess. Hoffman et al. (1989) assessed the response of 20-year-old Santa Rosa plum trees in California’s San Joaquin Valley to soil salinity. The experiment involved the use of irrigation water with six levels of salinity (ECi of

396

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

0.3 dS/m to 8 dS/m). The water was applied through two mini-sprinklers for each tree to apply published measurements of ET (seasonal ET of 1,030 mm) and the desired LF (0.3). Figure 12-13 presents soil salinity profiles before the irrigation season (February or March) and during its

FIGURE 12-13. Soil salinity profiles during a salt-tolerance experiment on plum trees [during 1986; nonsaline water (0 dS/m) was applied to the 8 dS/m treatment.] From Hoffman et al. (1989) with kind permission of Springer ScienceBusiness Media.

LEACHING AND ROOTZONE SALINITY CONTROL

397

second half (June to October) for three treatments during the study’s first three years. When the experiment began in 1984, all of the treatments had the same low level of salinity before the irrigation season. Winter rainfall that followed the 1984 irrigation season leached soil profiles to below 75 cm before the 1985 irrigation season. The same leaching took place before the 1986 season. The 8-dS/m treatment resulted in such severe salinity damage by the end of 1985 that nonsaline water was applied to that treatment in 1986. This accounts for the low salt content during the second half of the 1986 irrigation season. Soil salinity was relatively uniform with the depth of the soil (Fig. 12-13). Thus, regardless of the integration process used to account for variability with depth, the resulting average soil salinity would be close to a simple average in the increments of depth sampled. Soil salinity over time, however, changed significantly, as Fig. 12-14 illustrates. The salinity level rises quickly after irrigation begins and drops rapidly due to leaching induced by winter rainfall. Time-integrated values of soil salinity were determined from data similar to that presented in Fig. 12-13 to develop a salt-tolerance curve, as proposed by Maas and Hoffman (1977). To account for salinity’s influence on shoot growth, which contributes to bud formation the year before harvest, soil salinity measurements were integrated over the two years before each harvest. Excluded were the months from November to March, when the trees

FIGURE 12-14. Mean root zone salinity over time for plum trees irrigated with water of electrical conductivities 0, 4, and 8 dS/m. From Hoffman et al. (1989) with kind permission of Springer ScienceBusiness Media.

398

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 12-15. Salt tolerance of mature plum trees based on integrating soil salinity over a 2-year period. From Hoffman et al. (1989) with kind permission of Springer ScienceBusiness Media. were dormant. Data on flower formation, fruit set, and budwood development can be analyzed to establish a more accurate time frame for integration. The yield response of mature plum trees to soil salinity is based on the results of the first three years of the field trial (Fig. 12-15). According to these results, soil salinity can apparently be integrated over two years for plum trees. The proper period of time undoubtedly depends on the crop and its environment.

SUMMARY Salinity always threatens agriculture in arid or coastal environments. However, management strategies for using saline soil and water to produce crops have improved immeasurably by knowledge and experience gained over the past century. The basic premise that leaching is essential remains true. The gap between the leaching requirement and the leaching achieved on most irrigated land is being narrowed. As our ability to match crop water requirements with water applications improves throughout each field, our ability to minimize excess drainage will improve proportionately. The ultimate goal is to acquire the skills and knowledge necessary to use as efficiently as possible all available irrigation waters. Achieving this objective will minimize the amount of drain water requiring disposal or treatment, thus ensuring the sustainability of irrigated agriculture.

LEACHING AND ROOTZONE SALINITY CONTROL

399

REFERENCES Ayars, J. E. (1999). “Integrated management of irrigation and drainage systems,” in Water management, purification and conservation in arid climates, Vol. 1, M. F. A. Goosen and W. H. Shayya, eds., Technomic, Lancaster, Pa., 139–164 ———. (2003). “Field crop production in areas with saline soils and shallow saline groundwater in the San Joaquin Valley of California.” J. Crop Prod., 7, 353–386. Ayars, J. E., Christen, E. W., Soppe, R. W. O., and Meyer, W. S. (2006). “Resource potential of shallow groundwater for crop water use: A review.” Irrig. Sci., 24, 147–160. Ayars, J. E., Grismer, M. E., and Guitjens, J. C. (1997). “Water quality as design criterion in drainage water management system.” J. Irrig. Drain. Eng., 123, 148–153. Ayars, J. E., Hutmacher, R. B., Schoneman, R. A., Vail, S. S., and Pflaum, T. (1993). “Long term use of saline water for irrigation.” Irrig. Sci., 14, 27–34. Ayers, R. S., and Westcot, D. W. (1976). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Food and Agriculture Organisation of the United Nations, Rome. Bernstein, L., and Pearson, G. A. (1954). “Influence of integrated moisture stress achieved by varying the osmotic pressure of culture solutions on growth of tomato and pepper plants.” Soil Sci., 77, 355–368. Bohn, H. L., McNeal, B. L., and O’Connor, G. A. (1979). Soil chemistry, WileyInterscience, New York. Chapman, H. D. (1966). Diagnostic criteria for plants and soils, Div. Agricultural Science, University of California, Oakland, Calif. Chaudri, I. I., Shak, B. H., Nagro, N., and Mallick, I. A. (1964). “Investigations on the role of Suaeda fruiticosa Forsk in the reclamation of saline and alkaline soils in West Pakistan plains.” Plant Soil, 21, 1–7. Cope, F. (1958). “Catchment salting in Victoria,” Soil Cons. Auth. Victoria Bull., 1, 1–88. Corwin, D. L., Rhoades, J. D., and Sˇ imu ˚ nek, J. (2007). “Leaching requirement for soil salinity control: Steady-state versus transient models.” Agric. Water Mgmt., 90, 165–180. Downes, R. G. (1961). “Soil salinity in non-irrigated arable and pastoral land as the result of unbalance of the hydrologic cycle.” Teheran Symposium Salinity Problems in the Arid Zones, UNESCO, Paris, 105–110. El-Mowelhi, N., El-Bershamgy, A., Hoffman, G. J., and Chang, A. C. (1988). “Enhancement of crop yields from subsurface drains with various envelopes.” Agric. Water Mgmt., 15, 131–140. Francois, L. E. (1981). “Alfalfa management under saline conditions with zero leaching.” Agron. J., 73, 1042–1046. Gardner, W. R. (1957). “Some steady-state solutions of the unsaturated moisture flow equation with application to evaporation from a water table.” Soil Sci., 85(4), 228–232. Gardner, W. R., and Fireman, M. (1958). “Laboratory studies of evaporation from soil columns in the presence of a water table.” Soil Sci., 85(5), 244–249. Grismer, M. E. (1990). “Subsurface drainage system design and drain water quality.” J. Irrig. Drain. Eng., 119, 537–543.

400

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Grismer, M. E., and Gates, T. K. (1988). “Estimating saline water table contributions to crop water use.” Calif. Agri., March-April, 23–24. Guitjens, J. C., Ayars, J. E., Grismer, M. E., and Willardson, L. S. (1997). “Drainage design for water quality management: Overview.” J. Irrig. Drain. Eng., 123, 148–153. Hanson, B. R., Grattan, S. R., and Fulton, A. (1993). Agricultural salinity and drainage, Water Management Handbook Series, Publication No. 93-01, University of California, Davis, Calif. Hanson, B. R., Hutmacher, R. B., and May, D. M. (2006). “Drip irrigation of tomato and cotton under shallow saline groundwater conditions.” Irrig. and Drain. Sys., 20, 155–175. Hoffman, G. J. (1985) “Drainage required to manage salinity.” J. Irrig. and Drain. Div., ASCE, 111, 199–206. Hoffman, G. J., Catlin, P. B., Mead, R. M., Johnson, R. S., Francois, L. E., and Goldhamer, D. (1989). “Yield and foliar injury responses of mature plum trees to salinity.” Irrig. Sci., 10, 215–229. Hoffman, G. J., Maas, E. V., Prichard, T. L., and Meyer, J. L. (1983). “Salt tolerance of corn in the Sacramento-San Joaquin Delta of California.” Irrig. Sci., 4, 31–44. Hoffman, G. J., and van Genuchten, M. T. (1983). “Water management for salinity control,” in Limitations to efficient water use in crop production, H. Taylor, W. Jordan, and T. Sinclair, eds., American Society of Agronomy Monograph, ASA, Madison, Wisc., 73–85. Jury, W. A., Tuli, A., and Letey, J. (2003). “Effect of travel time on management of a sequential reuse drainage operation.” Soil Sci. Soc. Amer. J., 67, 1122–1126. Kearney, T. E., Kegel, F. R., Ou, J. P., Smith, J. D., and Ingelbretsen, K. H. (1980). Field corn production in California, Div. Agricultural Science Leaflet 21163, University of California, Oakland, Calif. Kruse, E. G., Yoder, R. E., Cuevas, D. L., and Champion, D. F. (1986). Alfalfa water use from high, saline water tables, American Society of Agricultural Engineers Paper No. 86-2597, December, Chicago, Ill., ASABE, St. Joseph, Mich. Kruse, E. G., Young, D. A., and Champion, D. F. (1985). “Effects of saline water tables on corn irrigation,” in Proc., ASCE Specialty Conference, San Antonio, Tex., American Society of Civil Engineers, Reston, Va., 444–453. Letey, J., and Feng, G. L. (2007). “Dynamic versus steady-state approaches to evaluate irrigation management of saline waters.” Agric. Water Mgmt., 91, 1–10. Lyerly, P. J., and Longenecker, D. E. (1962). Salinity control in irrigation agriculture, Texas Agricultural Experiment Station Bulletin 876, Texas AgriLife Research, Texas A&M System, College Station, Tex. Maas, E. V. (1990). “Crop salt tolerance,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 262–304. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. and Drain. Div. ASCE, 103(IR2), 115–134. Maas, E. V., Hoffman, G. J., Chaba, G. D., Poss, J. A., and Shannon, M. C. (1983). “Salt sensitivity of corn at various growth stages.” Irrig. Sci., 4, 45–57. Meiri, A., and Poljakoff-Mayber, A. (1970). “Effect of various salinity regimes on growth, leaf expansion, and transpiration rate of bean plants.” Soil Sci., 109, 26–34.

LEACHING AND ROOTZONE SALINITY CONTROL

401

Namken, L. N., Wiegand, C. L., and Brown, R. G. (1969). “Water use by cotton from low and moderately saline static water tables.” Agron. J., 61, 305–310. Peck, A. J., Johnston, C. D., and Williamson, D. R. (1981). “Analyses of solute distributions in deeply weathered soils.” Agric. Water Mgmt., 4, 83–102. Ragab, R. A., and Amer, F. (1986). “Estimating water table contribution to the water supply of maize.” Ag. Water Mgmt., 11, 221–230. Rhoades, J. D. (1974). “Drainage for salinity control,” in Drainage for agriculture, J. van Schilfgaarde, ed., Agronomy Monograph No. 17, SSSA, Madison, Wisc., 433–461. ———. (1979). “Inexpensive four-electrode probe for monitoring soil salinity.” Soil Sci. Soc. Amer. J., 43, 817–818. ———. (1982). “Reclamation and management of salt-affected soils after drainage,” in B. L. Colgan, ed., Proc., 1st Western Provincial Conference for Rationalization of Water and Soil Research and Management, Lethbridge, Alberta, Canada, November 29–December 2, 23–198. Rhoades, J. D., Ingvalson, R. D., Tucker, J. M., and Clark, M. (1973). “Salts in irrigation drainage waters: I. Effects of irrigation water composition, leaching fraction, and time of year on the salt compositions of irrigation drainage waters.” Soil Sci. Soc. Amer. Proc., 37, 770–774. Rhoades, J. D., Krueger, D. B., and Reed, M. J. (1968). “The effect of soil-mineral weathering on the sodium hazard of irrigation waters.” Soil Sci. Soc. Amer. Proc., 32, 643–647. Rhoades, J. D., Oster, J. D., Ingvalson, R. D., Tucker, J. M., and Clark, M. (1974). “Minimizing the salt burdens of irrigation drainage water.” J. Environ. Qual., 3, 311–316. Shalhevet, J., and Bernstein, L. (1968). “Effects of vertically heterogeneous soil salinity on plant growth and water uptake.” Soil Sci., 106, 85–93. Shalhevet, J., Reiniger, P., and Shimshi, D. (1969). “Peanut response to uniform and nonuniform soil salinity.” Agron. J., 61, 384–387. Shalhevet, J., and Yaron, B. (1973). “Effect of soil and water salinity on tomato growth.” Plant Soil, 39, 285–292. Solomon, K. H. (1985). “Water-salinity-production functions.” Trans. ASAE, 28, 1975–1980. Talsma, T. (1963). “The control of saline groundwater.” Meded Landbouwhogeschool (Wageningen), 63(10), 1–68. U.S. Department of the Interior, Bureau of Reclamation (USBR). (1993). Drainage manual, 3rd ed., U.S. Department of the Interior, Bureau of Reclamation, Denver, Colo. U.S. Salinity Laboratory. (1954). “Diagnosis and improvement of saline and alkali soils,” in L. A. Richards, ed., USDA Agricultural Handbook 60, USDA, Washington, D.C. Wallender, W. W., Grimes, D. W., Henderson, D. W., and Stromberg, L. K. (1979). “Estimating the contribution of a perched water table to the seasonal evapotranspiration of cotton.” Agron. J., 71, 1056–1060. Wichelns, D. W., Cone, D., and Stuhr, G. (2002). “Evaluating the impact of irrigation and drainage policies on agricultural sustainability.” Irrig. and Drain. Sys., 16, 1–14. Wichelns, D. W., Nelson, D., and Weaver, T. March (1988). Farm-level analyses of irrigated crop production in areas with salinity and drainage problems, San Joaquin

402

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Valley Drainage Program, U.S. Dept. of the Interior, Bureau of Reclamation, Sacramento, Calif. Wu, Q., Christen, E. W., and Enever, D. (1999). Basinman: A water balance model for farms with pipe drainage and on-farm evaporation basins, CSIRO Land and Water Technical Report 1/99, Commonwealth Scientific and Industrial Research Organisation, Clayton South, Victoria, Australia. Yaalon, D. H. (1963). “The origin and accumulation of salts in groundwater and in soils of Israel.” Bull. Res. Council Israel, 11G, 105–131.

NOTATION A, a, b, n C Ca Cd Cg Cr Da Dd De Dg Di Dr Ds Dt d ECa ECd ECe ECGW ECi ECSW ET g i k LF LR

 constants determined by experimental data  salt concentration  salt content of applied water  salt content of drain water  salt concentration of groundwater  salt concentration of the irrigation water  depth of applied water (irrigation plus rainfall)  depth of flow of water out of the crop’s rootzone due to drainage  depth of flow of water out of the crop’s rootzone due to evaporation  depth of flow of water from groundwater into the crop’s rootzone  depth of flow of water from irrigation into the crop’s rootzone  depth of flow of water from rainfall into the crop’s rootzone  depth of stored soil water  depth of flow of water out of the crop’s rootzone due to transpiration  drainage  electrical conductivity of applied water  electrical conductivity of drainage water  electrical conductivity of soil saturation extract  electrical conductivity of groundwater  electrical conductivity of irrigation water  electrical conductivity of soil water  evapotranspiration  upward flow from groundwater  irrigation  hydraulic conductivity  leaching fraction  leaching requirement

LEACHING AND ROOTZONE SALINITY CONTROL

q qu r S Sc Sm Sp Ss St Z *

 soil water flux  soil water flux as a percentage of ET  rain  soil matric potential  salt removed in the harvested crop  salt dissolved from minerals in soil  salt precipitated  change in salt storage  salt added to soil as fertilizer or amendment  distances  required values

403

This page intentionally left blank

CHAPTER 13 PLANT SALT TOLERANCE Catherine M. Grieve, Stephen R. Grattan, and Eugene V. Maas

INTRODUCTION As noted in previous chapters of this manual, the quality of irrigation water is an important factor in determining sustainability of agriculture on salt-impaired lands. For a number of reasons, the availability of lowsalinity irrigation supplies has led to (1) an interest in using alternative supplies, such as recycled wastewaters, and (2) innovative plant and water management strategies to mitigate the adverse effects of salt and specificion stresses these poor-quality waters may impose on plant growth, yield, and quality. A second motivating factor is the lack of suitable drainage outlets in many agricultural areas of the world. Drainage of irrigated lands is one of the requisites for sustaining agricultural productivity in a given region over the long term. Adequate drainage not only allows for better aeration in the crop rootzone but provides a means by which salinity and toxic elements can be managed and controlled. Reuse of drainage water for irrigation is one way of expanding the useable water supply while at the same time reducing drainage volume. This chapter provides a management perspective on (1) how plants respond to salinity and toxic elements (e.g., Na, Cl, and B); (2) crop salt tolerance and the various factors that influence plant response to salinity; (3) the extent to which salinity affects crop yields and quality; and (4) management strategies to optimize yields by controlling soil salinity. It is not our intent to provide a comprehensive review of physiological effects of salinity crops. That topic is covered in detail in Chapter 6.

405

406

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SOURCE AND CAUSES OF SOIL SALINITY Salts found in soils and waters originate from parent rock material that has undergone geochemical weathering. Over geologic time, primary minerals have reacted with water, oxygen, and carbon dioxide to form secondary minerals and salts that were transported by water to oceans or depressions in the landscape. Inundation of large land masses by saline seas deposited sedimentary materials that have become the major source of salt in arid regions. In coastal or delta regions, salination of soils may occur predominantly by salts that contaminate freshwater supplies by seawater intrusion. Seawater intrusion can impair groundwater quality when wells are pumped to the extent that they overdraft aquifers near coastal areas. Salinity in freshwater channels near deltas is affected by the tide and can increase dramatically during high tides when stream flows are low. Coastal agriculture may also be subjected to cyclic salts where saline aerosols are produced by violent wave activity during storms or high winds on the sea. These salts can move inland considerable distances, but the most harmful effects occur on vegetation or crops grown close to the shore. Salts, however, can be found in groundwater at relatively high concentrations without originating from the sea. The concentration and composition of groundwater is largely dependent on the hydrological and geochemical environment that the infiltrating water encounters en route to the groundwater. This is particularly true in irrigated soils formed from marine sediments. Salts contained in irrigation water, regardless of their source, can salinize agricultural land if the mass of salts that moves out of the rootzone is less than the mass of salts entering the rootzone for an extended period of time. A favorable salt balance within the rootzone must be maintained by adequate leaching. In closed hydrologic basins, salts may have been present in the soil long before irrigation was introduced to a region. Upon irrigation, saline water tables can develop in poorly drained areas in relatively short time periods (i.e., years). Even if good-quality water is used for irrigation, salination may occur from capillary movement of salts and water to the surface from rising saline water tables. Rising water tables are a result of excessive deep percolation and are often associated with inefficient water-management practices, such as overirrigation or inadequate drainage systems. The two processes described, (1) salination from irrigation with saline water, and (2) salination from shallow saline water tables, are the most common cause of large-scale soil salination in irrigated agriculture. They are, of course, not mutually exclusive, and highly saline water tables can often occur from or in association with saline irrigation water. Other processes of soil salination described in Chapter 1 occur on a smaller scale and will not be discussed in this section.

PLANT SALT TOLERANCE

407

PLANT SALT TOLERANCE Plant salt tolerance is defined differently depending on the intended use and value of the plant. For agricultural and horticultural crops, growers are most concerned with achieving economic yield and quality under saline conditions. For landscape designers and managers, the ability of the plant to maintain an aesthetic quality without excessive growth is of primary concern. And for the ecologist, the interest is most often on plant survival and species dominance in an environmentally sensitive area affected by salinity. Therefore, no one definition is appropriate that covers all interest groups. Relative Yield–Response Curves for Agronomic and Horticultural Crops The salt tolerance of a crop can be described as a complex function of yield decline across a range of salt concentrations (Maas and Hoffman 1977; Maas and Grattan 1999; van Genuchten and Hoffman 1984). Salt tolerance can be adequately described on the basis of two parameters: threshold, the electrical conductivity (ECt) that is expected to cause the initial significant reduction in the maximum expected yield, and slope, the percentage of expected yield reduction per unit increase in salinity above the threshold value. There is considerable uncertainty regarding the yield-threshold soilsalinity values. The salinity coefficients (yield threshold and slope values) for the piece-wise linear slope-threshold model introduced by Maas and Hoffman (1977) are now determined by nonlinear least-squares statistical fitting that determines the slope and threshold values from a particular experimental dataset. Despite intense control of salinity and all other important variables related to plant yield in salt tolerance trials, for many crops the standard errors associated with the threshold values can be 50% to 100% percent of the best-fit threshold value. Salinity studies on rice grown in northern California, for example, resulted in a threshold value of 1.9 dS/m of the field water (Grattan et al. 2002) with a 95% confidence limit ranging between 0.6 and 3.2 dS/m (J. Poss, U.S. Salinity Laboratory, personal communication, 2004). Obviously, very large ranges of uncertainty exist and additional studies to resolve this theoretical maximum are needed to refine water quality standards to a greater degree of confidence. One approach recently described by Steppuhn et al. (2005a,b) substitutes a nonlinear relationship between relative yield and soil salinity similar to the nonlinear model introduced by van Genuchten and Hoffman (1984) for the linear “yield threshold” model. A curvilinear relationship better describes relative crop yield data than does the yield-threshold

408

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

expression. In this curvilinear relationship, there is no longer a “yield threshold” but, rather, a continuous decline in yield with increased soil salinity. Using nonlinear models, the numerically most reliable curve-fitting parameter seems to be the value at which yield is reduced by 50% (C50). The C50 value can still be estimated when too few data points exist to provide reliable information on the threshold and slope. The set of equations developed by van Genuchten and Hoffman (1984) takes advantage of the stability of C50. The C50 value, together with a response curve steepness constant (p), may be obtained by fitting the appropriate function (van Genuchten 1983) to observed salt tolerance response data (van Genuchten and Gupta 1993). This approach has been used to develop a salt tolerance index (ST-index) as a revised indicator of the inherent salinity tolerance or resistance of agricultural crops to rootzone salinity (Steppuhn et al. 2005a,b). Both nonlinear models as well as the piece-wise linear fit (Maas and Hoffman 1977) describe the data extremely well (r2  0.9) (Steppuhn et al. 2005a,b). Water quality regulators prefer the latter since the concept of a “threshold” value provides them with a regulatory limit to impose on wastewater dischargers. The former, however, may best describe plant response from a physiological perspective, but there remains some uncertainty regarding the method that best describes the data in the relative yield range of 100% to 80%, the range of interest to most users and regulators. Since both curve-fitting methods describe the data well, we choose to report the most comprehensive and historically familiar list of salinity coefficients for the Maas-Hoffman model since this chapter is written as a user’s manual. Herbaceous crops Table 13-1 lists threshold and slope values generated by the MaasHoffman model for 81 crops in terms of seasonal average ECe in the crop rootzone. Most of the data were obtained where crops were grown under conditions simulating recommended cultural and management practices for commercial production in the location tested. Consequently, the data indicate relative tolerances of different crops grown under different conditions and not under some standardized set of conditions. Furthermore, the data apply only where crops are exposed to fairly uniform salinities from the late seedling stage to maturity. Plants are likely to be more sensitive to salinity than the tables indicate should crops be planted in initially high-salinity conditions. Where crops have particularly sensitive stages, the tolerance limits are given in the footnotes. The data in Table 13-1 apply to soils where chloride (Cl) is the predominant anion. Because of the dissolution of CaSO4 when preparing saturated-soil extracts, the ECe of gypsiferous, (nonsodic, low Mg2) soils will

TABLE 13-1. Salt Tolerance of Herbaceous Cropsa Crop Common Name (1) Artichoke, Jerusalem Barleye Canola or rapeseed

409

Canola or rapeseed Chick pea Cornh Cotton Crambe Flax Guar Kenaf Lesquerella Millet, channel Oats Peanut Rice, paddy Roselle Rye Safflower Sesamej Sorghum Soybean Sugarbeetk

Salt-Tolerance Parameters Botanical Nameb (2)

Tolerance Based On: (3) Fiber, grain, and special crops

Helianthus tuberosus L. Hordeum vulgare L. Brassica campestris L. [syn. B. rapa L.] B. napus L. Cicer arietinum L. Zea mays L. Gossypium hirsutum L. Crambe abyssinica Hochst. ex R.E. Fries Linum usitatissimum L. Cyamopsis tetragonoloba (L). Taub. Hibiscus cannabinus L. Lesquerella fenderli (Gray) S. Wats. Echinochloa turnerana (Domin) J.M. Black Avena sativa L. Arachis hypogaea L. Oryza sativa L. Hibiscus sabdariffa L. Secale cereale L. Carthamus tinctorius L. Sesamum indicum L. Sorghum bicolor (L.) Moench Glycine max (L.) Merrrill Beta vulgaris L.

Tuber yield Grain yield Seed yield Seed yield Seed yield Ear FW Seed yield Seed yield Seed yield Seed yield Stem DW Seed yield Grain yield Grain yield Seed yield Grain yield Stem DW Grain yield Seed yield Pod DW Grain yield Seed yield Storage root

Thresholdc (ECe) (dS/m) (4)

Slope (% per dS/m) (5)

Ratingd (6)

0.4 8.0 9.7

9.6 5.0 14

MS T T

11.0 — 1.7 7.7 2.0 1.7 8.8 8.1 6.1 — — 3.2 3.0i — 11.4 — — 6.8 5.0 7.0

13 — 12 5.2 6.5 12 17 11.6 19 — — 29 12i — 10.8 — — 16 20 5.9

T MS MS T MS MS T T MT T T MS S MT T MT S MT MT T (continued)

410

TABLE 13-1. Salt Tolerance of Herbaceous Cropsa (Continued) Crop Common Name (1) Sugarcane Sunflower Triticale Wheat Wheat (semidwarf)f,a Wheat, Durum

Salt-Tolerance Parameters

Botanical Nameb (2) Saccharum officinarum L. Helianthus annuus L. X Triticosecale Wittmack Triticum aestivum L. T. aestivum L. T. turgidum L. var. durum Desf.

Thresholdc (ECe) (dS/m) (4) 1.7 4.8 6.1 6.0 8.6 5.9

Slope (% per dS/m) (5) 5.9 5.0 2.5 7.1 3.0 3.8

Ratingd (6) MS MT T MT T T

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

2.0 — — 6.0 — 6.9 —

7.3 — — 7.1 — 6.4 —

MS T* T* MT MS T MS*

Shoot DW Shoot DW Shoot DW

1.6 — — —

9.6 — — —

MS MT* MT MS*

— — 1.5 1.5

— — 12 5.7

MS* MT MS MS

Tolerance Based On: (3) Shoot DW Seed yield Grain yield Grain yield Grain yield Grain yield Grasses and forage crops

Alfalfa Alkaligrass, Nuttall Alkali sacaton Barley (forage) Bentgrass, creeping Bermudagrassm Bluestem, Angleton Broadbean Brome, mountain Brome, smooth Buffelgrass Burnet Canarygrass, reed Clover, alsike Clover, Berseem

Medicago sativa L. Puccinellia airoides (Nutt.) Wats. & Coult. Sporobolus airoides Torr. Hordeum vulgare L. Agrostis stolonifera L. Cynodon dactylon (L.) Pers. Dichanthium aristatum (Poir.) C.E. Hubb. [syn. Andropogon nodosus (Willem.) Nash] Vicia faba L. Bromus marginatus Nees ex Steud. B. inermis Leyss Pennisetum ciliare (L). Link. [syn. Cenchrus ciliaris] Poterium sanguisorba L. Phalaris arundinacea L. Trifolium hybridum L. T. alexandrinum L.

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

Clover, Hubam Clover, ladino Clover, Persian Clover, red Clover, strawberry Clover, sweet Clover, white Dutch Corn (forage)f Cowpea (forage) Dallisgrass Dhaincha Fescue, tall Fescue, meadow Foxtail, meadow Glycine Gram, black or Urd bean Grama, blue Guinea grass Hardinggrass Kallargrass Kikuyugrass Lablab bean

411

Lovegrassn Milkvetch, Cicer Millet, Foxtail

Melilotus alba Dest. var. annua H.S.Coe Trifolium repens L. T. resupinatum L. T. pratense L. T. fragiferum L. Melilotus sp. Mill. Trifolium repens L. Zea mays L. Vigna unguiculata (L.) Walp. Paspalum dilatatum Poir. Sesbania bispinosa (Linn.) W.F. Wight [syn. Sesbania aculeata (Willd.) Poir] Festuca elatior L. Festuca pratensis Huds. Alopecurus pratensis L. Neonotonia wightii [syn. Glycine wightii or javanica] Vigna mungo (L.) Hepper [syn. Phaseolus mungo L.] Bouteloua gracilis (HBK) Lag. ex Steud. Panicum maximum Jacq. Phalaris tuberosa L. var. stenoptera (Hack) A. S. Hitchc. Leptochloa fusca (L.) Kunth [syn. Diplachne fusca Beauv.] Pennisetum clandestinum L. Lablab purpureus (L.) Sweet [syn. Dolichos lablab L.] Eragrostis sp. N. M. Wolf Astragalus cicer L. Setaria italica (L.) Beauvois

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

— 1.5 — 1.5 1.5 — — 1.8 2.5 — —

— 12 — 12 12 — — 7.4 11 — —

MT* MS MS* MS MS MT* MS* MS MS MS* MT

Shoot DW Shoot DW Shoot DW Shoot DW

3.9 — 1.5 —

5.3 — 9.6 —

MT MT* MS MS

Shoot DW

—

—

S

Shoot DW Shoot DW Shoot DW

— — 4.6

— — 7.6

MS* MT MT

Shoot DW

—

—

T

Shoot DW Shoot DW

8.0 —

—

T MS

Shoot DW Shoot DW Dry matter

2.0 — —

8.4 — —

MS MS* MS (continued)

412

TABLE 13-1. Salt Tolerance of Herbaceous Cropsa (Continued) Crop Common Name (1) Oatgrass, tall Oats (forage) Orchardgrass Panicum, Blue Pea, Pigeon Rape (forage) Rescuegrass Rhodesgrass Rye (forage) Ryegrass, Italian Ryegrass, perennial Ryegrass, Wimmera Salt grass, desert Sesbania Sirato Sphaerophysa Sudangrass Timothy Trefoil, big Trefoil, narrowleaf birdsfoot Trefoil, broadleaf birdsfoot

Botanical Nameb (2) Arrhenatherum elatius (L.) Beauvois ex J. Presl & K. Presl Avena sativa L. Dactylis glomerata L. Panicum antidotale Retz. Cajanus cajan (L.) Huth [syn. C. indicus (K.) Spreng.] Brassica napus L. Bromus unioloides HBK Chloris Gayana Kunth. Secale cereale L. Lolium multiflorum Lam. Lolium perenne L. L. rigidum Gaud. Distichlis spicta L. var. stricta (Torr.) Bettle Sesbania exaltata (Raf.) V.L. Cory Macroptilium atropurpureum (DC.) Urb. Sphaerophysa salsula (Pall.) DC Sorghum sudanense (Piper) Stapf Phleum pratense L. Lotus pedunculatus Cav. L. corniculatus var tenuifolium L. L. corniculatus L. var arvenis (Schkuhr) Ser. ex DC

Salt-Tolerance Parameters Thresholdc (ECe) (dS/m) (4) —

Slope (% per dS/m) (5) —

Ratingd (6) MS*

— 1.5 — —

— 6.2 — —

T MS MS* S

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

— — — 7.6 — 5.6 — — 2.3 — 2.2 2.8 — 2.3 5.0

— — — 4.9 — 7.6 — — 7.0 — 7.0 4.3 — 19 10

MT* MT* MT T MT* MT MT* T* MS MS MS MT MS* MS MT

Shoot DW

—

—

MS

Tolerance Based On: (3) Shoot DW Straw DW Shoot DW Shoot DW Shoot DW

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

Vetch, common Wheat (forage)l Wheat, Durum (forage) Wheatgrass, standard crested Wheatgrass, fairway crested Wheatgrass, intermediate Wheatgrass, slender Wheatgrass, tall Wheatgrass, western Wildrye, Altai Wildrye, beardless Wildrye, Canadian Wildrye, Russian

Vicia angustifolia L. Triticum aestivum L. T. turgidum L. var durum Desf. Agropyron sibiricum (Willd.) Beauvois

Shoot DW Shoot DW Shoot DW Shoot DW

3.0 4.5 2.1 3.5

11 2.6 2.5 4.0

MS MT MT MT

A. cristatum (L.) Gaertn.

Shoot DW

7.5

6.9

T

A. intermedium (Host) Beauvois A. trachycaulum (Link) Malte A. elongatum (Hort) Beauvois A. smithii Rydb. Elymus angustus Trin. E. triticoides Buckl. E. canadensis L. E. junceus Fisch.

Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW Shoot DW

— — 7.5 — — 2.7 — —

— — 4.2 — — 6.0 — —

MT* MT T MT* T MT MT* T

6.1 4.1 1.0 — 1.8 — 4.0 1.3 — 1.8 1.0 1.5 1.8 1.7 4.9

11.5 2.0 19 — 20.7 — 9.0 15.8 — 9.7 14 14.4 6.2 12 12

MT T S MT* S MS MT MT MS* MS S MS* MT MS MT

Vegetable and fruit crops

413

Artichoke Asparagus Bean, common Bean, lima Bean, mung Cassava Beet, redk Broccoli Brussels Sprouts Cabbage Carrot Cauliflower Celery Corn, sweet Cowpea

Cynara scolymus L. Asparagus officinalis L. Phaseolus vulgaris L. P. lunatus L. Vigna radiata (L.) R. Wilcz. Manihot esculenta Crantz Beta vulgaris L. Brassica oleracea L. (Botrytis Group) B. oleracea L. (Gemmifera Group) B. oleracea L. (Capitata Group) Daucus carota L. Brassica oleracea L. (Botrytis Group) Apium graveolens L. var dulce (Mill.) Pers. Zea mays L. Vigna unguiculata (L.) Walp.

Bud yield Spear yield Seed yield Seed yield Seed yield Tuber yield Storage root Head FW Head FW Storage root Petiole FW Ear FW Seed yield

(continued)

414

TABLE 13-1. Salt Tolerance of Herbaceous Cropsa (Continued) Crop Common Name (1) Cucumber Eggplant Fennel Garlic Gram, black or Urd bean Kale Kohlrabi Lettuce Muskmelon Okra Onion (bulb) Onion (seed) Parsnip Pea Pepper Pigeon pea Potato Pumpkin Purslane Radish Spinach Squash, scallop Squash, zucchini

Botanical Nameb (2) Cucumis sativus L. Solanum melongena L. var esculentum Nees. Foeniculum vulgare Mill. Allium sativum L. Vigna mungo (L.) Hepper [syn. Phaseolus mungo L.] Brassica oleracea L. (Acephala Group) Brassica oleracea L. (Gongylodes Group) Lactuca sativa L. Cucumis melo L. (Reticulatus Group) Abelmoschus esculentus (L.) Moench Allium cepa L. Pastinaca sativa L. Pisum sativum L. Capsicum annuum L. Cajanus cajan (L.) Huth [syn. C. indicus (K.) Spreng.] Solanum tuberosum L. Cucurbita pepo L. var Pepo Portulaca oleracea L. Raphanus sativus L. Spinacia oleracea L. Cucurbita pepo L. var melopepo (L.) Alef. C. pepo L. var melopepo (L.) Alef.

Salt-Tolerance Parameters Tolerance Based On: (3) Fruit yield Fruit yield Bulb yield Bulb yield Shoot DW

Top FW Fruit yield Pod yield Bulb yield Seed yield Seed FW Fruit yield Shoot DW Tuber yield Shoot FW Storage root Top FW Fruit yield Fruit yield

Thresholdc (ECe) (dS/m) (4) 2.5 1.1 1.4 3.9 —

Slope (% per dS/m) (5) 13 6.9 16 14.3 —

Ratingd (6) MS MS S MS S

— — 1.3 1.0 — 1.2 1.0 — 3.4 1.5 —

— — 13 8.4 — 16 8.0 — 10.6 14 —

MS* MS* MS MS MS S MS S* MS MS S

1.7 — 6.3 1.2 2.0 3.2 4.9

12 — 9.6 13 7.6 16 10.5

MS MS* MT MS MS MS MT

Strawberry Sweet potato Swiss chard Tepary bean Tomato Tomato, cherry Turnip Turnip (greens)

Fragaria x Ananassa Duch. Ipomoea batatas (L.) Lam. Beta vulgaris L. Phaseolus acutifolius Gray Lycopersicon lycopersicum (L.) Karst. ex Farw. [syn. Lycopersicon esculentum Mill.] L. lycopersicum var. Cerasiforme (Dunal) Alef. Brassica rapa L. (Rapifera Group)

Watermelon Winged bean

Citrullus lanatus (Thunb.) Matsum. & Nakai Psophocarpus tetragonolobus L. DC

Fruit yield Fleshy root Top FW Fruit yield

1.0 1.5 7.0 — 2.5

33 11 5.7 — 9.9

S MS T MS* MS

Fruit yield Storage root

1.7 0.9

9.1 9.0

MS MS

Top FW Fruit yield Shoot DW

3.3 — —

4.3 — —

MT MS* MT

a

These data serve only as a guideline to relative tolerances among crops. Absolute tolerances vary, depending on climate, soil conditions, and cultural practices. b Botanical and common names follow the convention of Hortus Third (Liberty Hyde Bailey Hortorium Staff, 1976) where possible. c In gypsiferous soils, plants will tolerate ECe of about 2 dS/m higher than indicated. d Ratings with an * are estimates. e Less tolerant during seedling stage; ECe at this stage should not exceed 4 or 5 dS/m. f Unpublished U. S. Salinity Laboratory data h Grain and forage yields of DeKalb XL-75 grown on an organic muck soil decreased about 26% per dS/m above a threshold of 1.9 dS/m. i Because paddy rice is grown under flooded conditions, values refer to the electrical conductivity of the soil-water while the plants are submerged. Less tolerant during seedling stage. j Sesame cultivars Sesaco 7 and 8 may be more tolerant than indicated by the S rating. k Sensitive during germination and emergence; ECe should not exceed 3 dS/m. l Data from one cultivar, Probred. m Average of several varieties. Suwannee and Coastal are about 20% more tolerant, and common and Greenfield are about 20% less tolerant than the average. n Average for Boer, Wilman, Sand, and Weeping cultivars. Lehmann seems about 50% more tolerant. *Estimated. DW, dry weight; FW, fresh weight; S, sensitive; MS, moderately sensitive; MT, moderately tolerant; T, tolerant.

415

Adapted from Maas and Grattan (1999).

416

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

range from 1 to 3 dS m1 higher than that of nongypsiferous soils having a similar soil water conductivity value at field capacity (Bernstein 1962). The extent of this dissolution depends on the exchangeable ion composition, cation exchange capacity, and solution composition. Therefore, plants grown on gypsiferous soils will tolerate ECes approximately 2 dS m1 higher than those listed in Table 13-1. The last column provides a qualitative salt tolerance rating that is useful in categorizing crops in general terms. The limits of these categories are illustrated in Fig. 13-1. Some crops are listed with only a qualitative rating, because experimental data are inadequate to calculate the threshold and slope. The salt tolerance parameters shown in Table 13-1 are given in terms of the relative, rather than absolute, yield response under salinity, and for that reason may be somewhat misleading for growers in selecting crops for maximum yield and profitability given specific saline field conditions. Comparison of the relative and absolute yields of alfalfa, a high-value, high-quality leguminous forage, and tall wheatgrass, a forage of moderate value and quality, provides a good illustration. The salt tolerance threshold and slope, expressed on a relative yield basis, for alfalfa are 2 dS m1 and 7.3%, respectively; the crop is rated as moderately salt-sensitive. Tall wheatgrass, conversely, is considerably more tolerant to salinity; threshold-slope values are 7.5 dS m1 and 4.2%, respectively (Table 13-1).

FIGURE 13-1. Divisions for classifying crop tolerance to salinity. From Maas and Hoffman (1977).

PLANT SALT TOLERANCE

417

A year-long greenhouse sand culture experiment was conducted at the U.S. Salinity Laboratory in which both crops were irrigated with waters at two salinity levels: 15 and 25 dS m1 (Grattan et al. 2004a). In this system, irrigation water salinity (ECi) is equivalent to that of the sand water (ECsw) which, in turn, is approximately 2.2 times the EC of the saturatedsoil extract (ECe). The soil-water dynamics of the sand are similar to those found in field soils (Wang 2002). The salinity treatments were, therefore, estimated at 7.0 and 11.7 dS m1 expressed as ECe. The threshold-slope model predicts that at the lower salinity (7 dS m1), the relative yield of tall wheatgrass would not be reduced at all, but that alfalfa yield would be reduced more than 40%. However, annual absolute biomass production for alfalfa irrigated with 15-dS m1 waters was 28 t ha1, whereas tall wheatgrass produced 20 t ha1 in the same treatment. As irrigation water salinity increased to 25 dS m1, the relative salt sensitivity of alfalfa became obvious inasmuch as absolute biomass was reduced by nearly 50% (15 t ha1). In contrast, biomass of tall wheatgrass was reduced only 15% in response to the higher salinity. It is likely that if salinity increased even further (e.g., to ECi 30 dS m1) the survivability of alfalfa would be in question, although tall wheatgrass would, in all probability, maintain reasonable biomass production. Therefore, both salt tolerance and absolute biomass production must be considered in crop selection. Clearly the grower must have a priori knowledge of the overall crop production potential in order to make an appropriate crop selection for anticipated saline field conditions. Woody crops The salt tolerance of trees, vines, and other woody crops is complicated because of additional detrimental effects caused by specific-ion toxicities. Many perennial woody species are susceptible to foliar injury caused by the toxic accumulation of Cl and/or Na in the leaves. Because different cultivars and rootstocks absorb and transport Cl and Na at different rates, considerable variation in tolerance may occur within an individual species. Tolerances to these specific ions are discussed in the following. In the absence of specific-ion effects, the salt tolerance of woody crops, like that of herbaceous crops, can be expressed as a function of the concentration of total soluble salts or osmotic potential of the soil solution. One could expect this response to be obtained for those cultivars and rootstocks that restrict the uptake of Cl and Na. The salt tolerance data given in Table 13-2 for woody crops are believed to be reasonably accurate in the absence of specific-ion toxicities. Because of the cost and time required to obtain fruit yields for extended periods of time (i.e., multiple years), particularly for alternate-bearing trees, tolerances of woody crops have been determined for vegetative growth only. In contrast to other

418

TABLE 13-2. Salt Tolerance of Woody Cropsa Crop Common Name (1) Almond Apple Apricot Avocado Banana Blackberry Boysenberry Castorbean Cherimoya Cherry, sweet Cherry, sand Coconut Currant Date palm Fig Gooseberry Grape Grapefruit Guava Guayule

Botanical Nameb (2) Prunus duclis (Mill.) D.A. Webb Malus sylvestris Mill. Prunus armeniaca L. Persea americana Mill. Musa acuminata Colla Rubus macropetalus Doug. ex Hook Rubus ursinus Cham. and Schlechtend Ricinus communis L. Annona cherimola Mill. Prunus avium L. Prunus besseyi L., H. Baley Cocos nucifera L. Ribes sp. L. Phoenix dactylifera L. Ficus carica L. Ribes sp. L. Vitis vinifera L. Citrus x paradisi Macfady Psidium guajava L. Parthenium argentatum A. Gray

Jambolan plum Jojoba Jujube, Indian Lemon Lime Loquat

Syzygium cumini L. Simmondsia chinensis (Link) C. K. Schneid Ziziphus mauritiana Lam. Citrus limon (L.) Burm. f. Citrus aurantiifolia (Christm.) Swingle Eriobotrya japonica (Thunb). Lindl.

Salt-Tolerance Parameters Tolerance Based On: (3) Shoot growth Shoot growth Shoot growth Fruit yield Fruit yield Fruit yield Foliar injury Foliar injury Foliar injury, stem growth Foliar injury, stem growth Fruit yield Plant DW Shoot growth Fruit yield Shoot and root growth Shoot DW Rubber yield Shoot growth Shoot growth Fruit yield Fruit yield Foliar injury

Thresholdc (ECe) (dS/m) (4) 1.5 — 1.6 — — 1.5 1.5 — — — — — — 4.0 — — 1.5 1.2 4.7 8.7 7.8 — — — 1.5 — —

Slope (% per dS/m) (5) 19 — 24 — — 22 22 — — — — — — 3.6 — — 9.6 13.5 9.8 11.6 10.8 — — — 12.8 — —

Ratingd (6) S S S S S S S MS* S S* S* MT* S* T MT* S* MS S MT T T MT T MT S S* S*

Macadamia Mandarin orange; tangerine Mango Natal plum Olive Orange Papaya Passion fruit Peach Pear Pecan Persimmon Pineapple Pistachio Plum; Prune Pomegranate Popinac, white Pummelo Raspberry Rose apple Sapote, white Scarlet wisteria Tamarugo Walnut

Macadamia integrifolia Maiden & Betche Citrus reticulata Blanco Mangifera indica L. Carissa grandiflora (E.H. Mey.) A. DC. Olea europaea L. Citrus sinensis (L.) Osbeck Carica papaya L. Passiflora edulis Sims Prunus persica (L.) Batsch Pyrus communis L. Carya illinoinensis (Wangenh.) C. Koch Diospyros virginiana L. Ananas comosus (L.) Merrill Pistacia vera L. Prunus domestica L. Punica granatum L. Leucaena leucocephala (Lam.) de Wit [syn. Leucaena glauca Benth.] Citrus maxima (Burm.) Rubus idaeus L. Syzygium jambos (L.) Alston Casimiroa edulis Llave Sesbania grandiflora Prosopis tamarugo Phil Juglans spp.

Shoot DW Shoot growth Fruit yield Shoot growth Shoot DW

— — — — — 1.3 — — 1.7 — — — — — 2.6 — —

— — — — — 13.1 — — 21 — — — — — 31 — —

MS* S* S T MT S MS S* S S* MS S* MT MS MS MS MS

Foliar injury Fruit yield Foliar injury Foliar injury Shoot DW Observation Foliar injury

— — — — — — —

— — — — — — —

S* S S* S* MT T S*

Seedling growth Shoot growth Foliar injury Shoot growth Seedling growth, Fruit yield Fruit yield Seedling growth, foliar injury Shoot growth, Fruit yield Nut yield, trunk growth

a

419

These data serve only as a guideline to relative tolerances among crops. Absolute tolerances vary, depending on climate, soil conditions, and cultural practices. The data are applicable when rootstocks are used that do not accumulate Na or Cl rapidly or when these ions do not predominate in the soil. b Botanical and common names follow the convention of Hortus Third (Liberty Hyde Bailey Hortorium Staff, 1976) where possible. c In gypsiferous soils, plants will tolerate ECe about 2 dS/m higher than indicated. d Ratings with an * are estimates. *Estimated. DW, dry weight; FW, fresh weight; S, sensitive; MS, moderately sensitive; MT, moderately tolerant; T, tolerant. Adapted from Maas and Grattan (1999).

420

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

crop groups, most woody fruit and nut crops tend to be salt-sensitive, even in the absence of specific-ion effects. Only the date palm is rated as relatively salt-tolerant. Olives, pistachios, and a few others are believed to be tolerant to moderately high salinity, at least during the first few years of growth. As time under exposure to salts increased, however, tolerance may decline due to progressive toxic levels of salts accumulated in leaves or woody tissues. The long-term effects of olives grown under field conditions provide a striking example. Two years after planting and imposition of salt stress, the olive cultivar Arbequina was rated in 1999 as salttolerant with a threshold (ECe) of 6.7 dS m1. In 2000, the threshold decreased to 4.7 dS m1. By 2001, Arbequina was rated as moderately saltsensitive as the threshold declined to 3.0 dS m1 (Aragüés et al. 2005). However, this decline in salt tolerance over the years was not observed in plums. The more salt-sensitive Santa Rosa plum (Prunus salicina Lindl) on Marianna 2624 rootstock (P. cerasifera Ehrh.  P munsoniana Wight and Hedr.) showed little change in tolerance due to age. At the end of a 6-year field trial, the salt tolerance parameters (i.e., threshold  2.6 dS m1; slope  31%) based on fruit yield determined after the first 3 years of the trial (1984–1985) were not significantly different than those obtained for 1987–1989 (Catlin et al. 1993). Quality of salt-stressed agronomic and horticultural crops While crop salt tolerance is based solely on yield, salinity adversely affects the quality of some crops. By decreasing the size and/or quality of fruits, tubers, or other edible organs, salinity reduces the market value of many vegetables, such as carrots, celery, cucumbers, peppers, potatoes, head cabbage and lettuce, artichoke, and yams (Bernstein et al. 1951; Bernstein 1964; Francois and West 1982; Francois 1991, 1995). Rye grown on saline soils produces grain with poorer bread-baking quality (Francois et al. 1989). Salinity appears to have only limited effect on the quality of citrus fruit (Maas 1993). Not all the effects of salinity on crop quality are negative, however (Grieve 2010). Salinity often confers beneficial effects on crops, which may translate into economic advantages (Pasternak and De Malach 1994). Salinity can increase yields in crops that show a strong competition for photosynthates between vegetative and reproductive structures. In certain crops, salt stress can slow growth of the vegetative parts, allowing the excess photosynthates to flow to the generative organs. Cotton is a good example of such a crop. Saline water (ECe  4.4 dS/m) irrigation resulted in 15% increases in fruit dry matter (g/plant) and number of bolls on fruiting branches as well as a 20% increase in boll number per plant (Pasternak et al. 1979). Although final internode number was reduced by 11%, reduction in total dry matter yield and plant height was not significant.

PLANT SALT TOLERANCE

421

In addition to the conservation of high-quality waters, the controlled use of degraded waters offers a second benefit by providing a unique opportunity for the production of value-added crops with healthpromoting constituents (Grieve 2010). Many plants adapt to salt stress by enhanced biosynthesis of secondary metabolites, such as soluble solids, sugars, organic acids, proteins, and amino acids (Ashraf and Harris 2004), which may act as osmolytes or osmoregulators to maintain plant turgor under salt stress. The presence of these metabolites often greatly increases the nutritive quality and marketability of fruits and vegetables (Mizrahi and Pasternak 1985). Beneficial effects include increased sugar concentration of carrots (Bernstein and Ayers 1953) and asparagus (Francois 1987), increased total soluble solids in tomatoes (Adams and Ho 1989; Krauss et al. 2006; Campos et al. 2006), muskmelon (Shannon and Francois 1978; Botia et al. 2005; Colla et al. 2006), cucumber (Chartzoulakis 1992; Trajkova et al. 2006), mandarin orange (Garcia-Sanchez et al. 2006), and improved grain quality and protein content of durum (Francois et al. 1986) and bread wheat (Rhoades et al. 1988). Salt-stress may increase firmness and improve postharvest handling characteristics in eggplants (Sifola et al. 1995), strawberries (Sarooshi and Cresswell 1994), tomatoes (Krauss et al. 2006), and melons (Navarro et al. 1999). Onion bulb pungency may be reduced by salt-stress, although the content of flavor precursors often increases (Chang and Randle 2005). Salinity may also cause oxidative stress and induce production of reactive oxygen species, which are damaging to all classes of biomolecules. The primary defensive plant response to oxidative stress is the biosynthesis of antioxidants (Bartosz 1997). As a result, salt-stressed plants often contain enhanced concentrations of antioxidants, such as flavonoids, ascorbate, tocopherols, carotenoids, and lycopene. With proper management practices, it is likely that economic losses associated with yield reductions due to salinity may be offset by production of high-quality food crops that can be marketed at a premium to meet the changing demands of the market and healthconscious consumers (Cuarto and Fernández-Muñoz 1999; De Pascale et al. 2001). Ornamental and landscape species Research on the salt tolerance of floriculture species continues to be largely devoted to providing information that would help commercial growers maintain crop productivity, quality, and profitability if recycled waters are used for irrigation. Quality standards for landscape use are far less stringent than those required by the floriculture industry. For example, a major quality determinant for important cut flowers is stem length, a growth parameter that is generally reduced when the plant is challenged by salinity. In their drive for high-quality products suitable for

422

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

premium markets, commercial growers would likely use the highestquality water available to maximize inflorescence length, flower diameter, and plant height. However, a flowering stalk of stock (Matthiola incana), a moderately salt-sensitive crop, would still be aesthetically acceptable for landscape purposes if, compared to a premium-grade stalk, the flowering stem was 5 cm shorter and the inflorescence contained one or two fewer florets. Minimal reduction in growth and flowering capacity should be permissible, provided that the overall health of the plant is not compromised, the stems are robust, the colors of the leaves and flowers remain true, and no visible leaf or flower damage due to salt stress is evident. For landscape purposes, stock is rated as very salt-tolerant. Applying salt-tolerance criteria derived from the ecophysiological literature to landscape plants sometimes results in completely misleading tolerance ratings. The performance of statice (Limonium spp.) under saline conditions provides a good example. In HALOPH, a database of salt-tolerant plants of the world, Aronson (1989) lists more than 50 species of Limonium. The commercially important species, L. perezii and L. sinuatum, are listed among those that will complete their life cycles in waters more salty than seawater (e.g., EC  50 dS/m). That these species grow to maturity under highly saline conditions is clearly a halophytic characteristic. Although one would not expect either species to produce a high-quality crop under irrigation with hypersaline waters, the question arose: Could flowers suitable for the commercial market or for landscape purposes be produced at lower salinities, for example, in the range of 20 to 30 dS m1? To answer this question, both statice species were grown under irrigation with waters ranging from 2 to 30 dS m1 (Grieve et al. 2005; Carter et al. 2005). These trials confirmed that both species were halophytic; both flowered and set seed in all treatments. However, neither species possessed a high degree of salt tolerance as understood by horticulturists and agronomists whose research focuses on crop yield and quality. Growth response of statice more closely resembled that of glycophytic plants. Height of the flowering stalks decreased consistently as salinity increased. Those plants receiving the 30-dS m1 treatment were only one-quarter as tall as those irrigated with nonsaline waters. The salt tolerance of both species, rated for commercial production on the basis of stem length, is correctly rated as low (Farnham et al. 1985). Reduction in stem length should not necessarily be the limiting factor in species selection for landscape plants, however. Even under severe salt stress, both ‘American Beauty’ and ‘Blue Seas’ produced acceptable, healthy plants with attractive foliage and colorful inflorescences on sturdy, albeit short, stems. For landscape purposes, the species fall in the “very tolerant” category. Many examples are available illustrating that the effects of salinity on landscape plants are not always adverse. Salt-related stress can benefi-

PLANT SALT TOLERANCE

423

cially affect quality and disease resistance of plants. If plant aesthetics are not compromised, salt-stressed landscape plants will be slower growing, requiring less trimming and maintenance. In some instances, the uptake and accumulation of salinizing ions stimulates growth. Cabrera (2000) and Cabrera and Perdomo (2003) observed a positive correlation between relatively high leaf-Cl concentrations (0.45%) and dry weight for containergrown rose (‘Bridal Pink’ on Rosa manetti rootstock). Yield and quality were unaffected. Salinity imposed early in the life cycle of some cut-flower species tends to limit vegetative growth, with favorable results. Salinityinduced reduction of stem length may be beneficial in species such as chrysanthemum, where tall, rangy cultivars are treated with growth regulators to keep the plants compact and short. While plant height is often reduced by moderate salinity, the length of time to maturity and the size of developing floral buds generally remain unaffected by stress (Lieth and Burger 1989). Salt tolerance ratings of selected landscape species (Table 13-3) are based on aesthetic value and survivability. In some cases, two contrasting ratings are given. Differences may be due to variety, climatic, or nutritional conditions under which the trials were conducted. In addition, some of the ratings are derived from data collected from closely related varieties of horticultural or agronomic value. There are no data, for example, on the salt tolerance of ornamental brassicas, such a kale and cabbage, but it would be reasonable to assume that their salt tolerance would not differ very sharply from that of the same leafy vegetable crop grown under field conditions in agricultural settings. Excellent resources for additional information regarding the salt tolerance of landscape plants are the Salt Management Guide (Tanji 2007) and Abiotic Disorders of Landscape Plants: A Diagnostic Guide (Costello et al. 2003). Potential uses of halophytes A promising approach for the practical use of heavily salinized soils and waters that are otherwise unsuitable for conventional agriculture is the use of highly salt-tolerant plant species, (halophytes). Many halophytes are valuable for economic reasons (human food, fodder, oil, fuel) or for ecological reasons, such as dune stabilization, erosion control, CO2 sequestration, reclamation, and desalinization (Koyro 2003). True halophytes are defined as those plants that are able to survive and complete their life cycles in hypersaline environments and whose maximum growth occurs at a soil water salinity of ⬃20 dS/m (Salisbury 1995). Halophytes have developed a number of morphological adaptations and physiological mechanisms to avoid and resist salt stress: salt hairs and salt glands, waxy cuticles, selective ion uptake, salt exclusion from different

424

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 13-3. Salt Tolerance of Selected Landscape Plants Botanical Name (1)

Common Name (2)

Agapanthus orientalis Ageratum houstonianum Alstroemeria hybrids Amaranthus hypochondriacus Amaranthus tricolor Anthurium andreanum Antirrhinum majus Artemesia stelleran Begonia bunchii Begonia Rex-cultorum Begonia ricinifolia Bouvardia longiflora Brassica oleracea Brassica oleracea Calendula officinalis Callistephus chinensis

Lily of the Nile Ageratum Inca lily, Peruvian lily Pygmy Torch Love-Lies-Bleeding Anthurium Snapdragon Dusty Miller Begonia Rex Begonia Begonia Bouvardia Ornamental Cabbage Ornamental Kale Pot Marigold China Aster

Calocephalus brownii Camellia japonica Carathamus tinctorius Catharanthus roseus Celosia argenta cristata Celosia argenta cristata Cereus peruviana Chlorophytum comosum Chrysanthemum morifolium Clematis orientalis Coleus blumei Codiaeum punctatus Consolida ambigua Cosmos bipinnatus Coreopsis grandiflora Crassula ovata Cyclamen persicum Cymbidium spp. Dianthus barbatus Dianthus caryophyllus Dianthus chinensis Eschscholzia californica Euphorbia pulcherrima Euphorbia pulcherrima Euryops pectinatus Eustoma grandiforum

Cushion Bush Camellia Safflower Vinca Crested Coxcomb Chief Celosia Apple Cactus St. Bernard’s Lily Mum Clematis Coleus Croton Larkspur Cosmos Coreopsis Jade Plant Cyclamen Orchid Pinks Carnation Carnation California Poppy Poinsettia ‘Red Sails’ Poinsettia ‘Barbara Ecke’ Golden Marguerite Lisianthus

Salt Tolerancea (3) sensitive moderately sensitive very sensitive tolerant tolerant very sensitive tolerant moderately sensitive sensitive very sensitive sensitive moderately sensitive sensitive sensitive moderately tolerant moderately sensitive moderately tolerant moderately sensitive sensitive moderately tolerant sensitive moderately sensitive tolerant moderately sensitive tolerant moderately tolerant very tolerant tolerant moderately tolerant sensitive very sensitive moderately sensitive moderately sensitive sensitive very sensitive moderately sensitive moderately tolerant moderately tolerant moderately tolerant sensitive very sensitive sensitive moderately sensitive

PLANT SALT TOLERANCE

425

TABLE 13-3. Salt Tolerance of Selected Landscape Plants (Continued) Botanical Name (1) Felicia amelloides Fuchsia hybrida Gardenia augusta Gazania aurantiacum Gerbera jamesonii Gazania spp. Gladiolus spp. Gomphrena globosa Gyposphila paniculata Helianthus annuus Helianthus debilis Hibiscus rosa-sinensis Hippeastrum hybridum Hymenocallis keyensis Impatiens x hawkeri Kalanchoe spp. Kochia childsii Lathyrus japonica Lilium spp. Lilium spp. Limonium spp. Limonium latifolium Limonium perezii Limonium sinuatum Lobularia maritima Matthiola incana Narcissus tazetta Oenthera speciosa Ophiopogon jaburan Ornithogalum arabicum Pelargonium x hortorum Pelargonium domesticum Pelargonium peltatum Petunia hybrida Portulaca grandiflora Phalaenopsis hybrid Protea obtusifolia Rhododendron hybrids Rhododendron obtusum Rosa x hybrida Stapelia gigantea

Common Name (2) Felicia Fuchsia Gardenia Gazania Gerbera Daisy Treasure Flower Gladiola Globe Amaranth Baby’s Breath Sunflower Cucumber Leaf Hibiscus Amaryllis Spiderlily Impatiens Kalanchoe Kochia Sweet Pea Asiatic Hybrid Lily Oriental Hybrid Lily Japanese Limonium Sea Lavender Statice Statice Sweet Alyssum Stock Paperwhite Narcissus Mexican Evening Primrose Giant Turf Lily Arabian Star Flower Geranium Geranium Ivy Geranium Petunia Moss Rose Orchid Protea Azalea Azalea Rose Starfish Flower

Salt Tolerancea (3) sensitive very sensitive sensitive moderately tolerant moderately sensitive very tolerant sensitive moderately sensitive moderately tolerant moderately tolerant very tolerant sensitive very sensitive moderately tolerant sensitive moderately tolerant tolerant moderately tolerant sensitive sensitive very tolerant very tolerant Sensitive; very tolerant Sensitive; very tolerant moderately tolerant very tolerant sensitive moderately tolerant moderately sensitive very sensitive sensitive tolerant moderately tolerant tolerant very tolerant very sensitive moderately tolerant moderately sensitive sensitive sensitive moderately tolerant (continued)

426

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 13-3. Salt Tolerance of Selected Landscape Plants (Continued) Botanical Name (1) Strelitzia reginae Tagetes erecta Tagetes patula Trachelium caeruleum Tropaeolum majus Vinca major Vinca minor Viola x wittrockiana Zinnia elegans a

Common Name (2) Bird of Paradise Marigold Marigold Blue Throatwort Nasturtium Periwinkle Myrtle Pansy Zinnia

Salt Tolerancea (3) very sensitive moderately tolerant moderately tolerant sensitive moderately sensitive moderately tolerant sensitive sensitive moderately sensitive

Criteria for assigning salt tolerance: not more than 50% reduction in growth, no visually observable foliar burn, and maximum permissible ECe (dS m1) as follows: 2 2–3 3–4 4–5 5–6 6

very sensitive sensitive moderately sensitive moderately tolerant tolerant very tolerant

Adapted from Grieve et al. (2007) and Tanji (2007).

plant organs (root, stem, leaf or fruit), salt sequestration in vacuoles or in senescent leaves, succulence, dilution of plant salt concentration by increased growth, osmotic adjustment, compatible osmotic solutes, root excretion of salts, and root molecular sieves (Ungar 1998). Halophytes may also be of value in water treatment. Improvement of water quality through the use of natural or constructed wetlands is a relatively new concept for treating effluents from agricultural operations, such as dairies, livestock feedlots (Ibekwe et al. 2003; Ibekwe et al. 2007; Ray and Inouye 2007), and nurseries (Arnold et al. 2003). Wastewaters from agricultural operations are generally brackish and typically contain high levels of nutrients and other pollutants. Vegetation plays a significant role in wastewater purification by reducing nitrogen and the biochemical oxygen demand and removal of suspended solids (Gersberg et al. 1986). Certain aquatic plant species possess unique anatomical and morphological features that together with their pollutant uptake capacity and survivability, make them of prime importance in wetlands ecosystems. Wetland species improve water quality by direct uptake of nutrients and also by reducing water velocity, which allows suspended particles to settle (Ray and Inouye 2007). Ecologically valuable species for these purposes include bulrush (Scirpus validus), common tule (S. acutus), rush (Juncus balticus), spike rush (Eleocharis palustris),

PLANT SALT TOLERANCE

427

common reed (Phragmites communis), cattail (Typha latifolia), and carex (Carex nebrascensis). Additional information concerning ecologically important species may be found in HALOPH (Aronson 1989). This compilation lists halophytic crops by plant family and gives maximum reported salinity tolerance, geographical distribution, and potential economic uses for many species. Another valuable resource is Cash Crop Halophytes (Lieth and Mochtchenko 2003), which addresses topics, such as ecophysiological research on salt tolerance of plants, halophyte utilization for reclamation of soils, and sustainable systems under irrigation with seawater. A CD-ROM accompanies the text and gives taxonomic classification and highest reported salinity tolerance for more than 2,000 species.

SALINITY AND NUTRITIONAL IMBALANCE Salinity can induce elemental nutrient deficiencies or imbalances in plants depending on the ionic composition of the external solution. These specific effects vary among species and even among varieties of a given crop. The optimal concentration range for a particular nutrient element in the soil solution depends on many factors, including salt concentration and composition (Grattan and Grieve 1994). This is not surprising since salinity affects nutrient ion activities and produces extreme ion ratios (e.g., Na/Ca2, Na/K, Cl/NO 3 ) in the soil solution. Nutrient imbalances in the plant may result from the effect of salinity on (1) nutrient availability, (2) the uptake and/or distribution of a nutrient within the plant, and/or (3) increasing the internal plant requirement for a nutrient element resulting from physiological inactivation (Grattan and Grieve 1999). A substantial body of information in the literature indicates that nutrient element acquisition by crops is reduced in saline environments, depending, of course, on the nutrient element in question and the composition of the salinizing solution. The activity of a nutrient element in the soil solution decreases as salinity increases, unless the nutrient in question is part of the salinizing salts (e.g., Ca2, Mg2, or SO42). For example, phosphate (P) availability is reduced in saline soils not only because the ionic-strength effect reduces the activity of phosphate, but also because its concentration is controlled by sorption processes and by the precipitation of Ca-P minerals. Therefore, P concentrations in many full-grown agronomic crops decrease as salinity increases (Sharpley et al. 1992). Soil salinity can affect nutrient acquisition by severely reducing root growth. Reductions of 40% to 50% have been reported in root weight and lengths of citrus and tomato (Zekri and Parsons 1990; Snapp and Shennan 1992).

428

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Other evidence indicates that salinity may cause some physiological inactivation of P, thereby increasing the plant’s internal requirement for this element (Awad et al. 1990). These investigators found that when NaCl concentrations were increased from 10 to 100 mM, P concentration in the youngest mature tomato leaf necessary to achieve 50% yield nearly doubled. Moreover, at any given leaf P concentration, foliar symptoms of P deficiency increased with increased NaCl salinity. It would not be surprising to find similar relationships involving other crops or even other nutrients. Nutrient uptake and accumulation by plants is often reduced under saline conditions as a result of competitive processes between the nutrient and a major salt species. Although plants selectively absorb K over Na, Na-induced K deficiencies can develop on crops under salinity stress by Na salts (Janzen and Chang 1987). On the other hand, Cl salts can reduce NO 3 uptake and accumulation in crops even though this effect may not be growth-limiting (Munns and Termaat 1986). Even under nonsaline conditions, significant economic losses of horticultural crops have been linked to inadequate calcium (Ca2) nutrition (Shear 1975). Many factors can influence the amount of plant-available Ca. These include the total supply of Ca2, the nature of the counter-ions, the pH of the substrate, and the ratio of Ca2 to other cations in the irrigation water (Grattan and Grieve 1999). Calcium-related disorders may even occur in plants grown on substrates where the Ca2 concentration appears to be adequate (Pearson 1959; Bernstein 1975). Deficiency symptoms are generally caused by differences in Ca2 partitioning to the growing regions of the plant. All plant parts—leaves, stems, flowers, fruits— actively compete for the pool of available Ca2 and each part influences Ca2 movement independently. Organs that are most actively transpiring are those most apt to have the highest Ca2 concentrations. In horticultural plants whose marketable product consists primarily of large heads enveloped by outer (“wrapper”) leaves [e.g., cabbage, lettuce, escarole, or endive], excessive transpiration by the outer leaves diverts Ca2 from the rapidly growing meristematic tissue (Bangerth 1979). Calcium deficiency appears as physiological disorders of the younger tissues: internal browning of cabbage and lettuce, blackheart of celery. Calcium deficiency disorders may also occur in reproductive tissues and may reduce market quality: blossom-end rot of tomatoes, melons, and peppers; “soft-nose” of mangoes and avocados; and cracking and “bitter pit” of apples. Artichokes grown under arid but nonsaline conditions also exhibits Ca-deficiency injury as necrosis of inner bracts. The incidence of the disorder increased when salt-stress was imposed (Francois 1995). Any hazard to horticultural crops that are susceptible to Ca-related disorders in the absence of salinity becomes even greater under saline conditions. As the salt concentration in the rootzone increases, the plant’s

PLANT SALT TOLERANCE

429

requirement for Ca also increases (Bernstein 1975). At the same time, Ca uptake from the substrate may be depressed because of ion interactions, precipitation, and increases in ionic strength (Grattan and Grieve 1999). Significant reduction in market quality and associated economic losses occur when these susceptible crops are also challenged by salinity. Sodium-induced Ca2 deficiencies have been observed by many researchers when the Na/Ca2 ratio in the solution, at a given salinity level, increases above a threshold level. This is particularly true for many crops in the grass family (e.g., corn, sorghum, rice, wheat, and barley) and striking differences have been observed among species and cultivars. Calcium deficiency may be related, at least in part, to the effect of Na on Ca2 distribution within the plant. Some researchers found that Na inhibits the radial movement of Ca2 from the root epidermis to the root xylem vessels (Lynch and Läuchli 1985), while others found that Ca2 transport to meristematic regions and developing leaves was inhibited (Maas and Grieve 1987; Grieve and Maas 1988). Salinity-induced Ca2 deficiency has also been observed on crops from different families, such as blossom-end rot in tomatoes and bell peppers and black heart in celery (Geraldson 1957).

CROP RESPONSE TO SPECIFIC IONS AND ELEMENTS In addition to osmotic effects that reduce plant biomass and yields and salinity’s effect on mineral nutrition, specific ions (i.e., Na, Cl, and B) can cause additional injury to the crops, causing further crop damage. These specific ions will be discussed separately. Sodium Sodium is not considered an essential element for most crop plants, but it does beneficially affect growth of some plants at concentrations below the salt-tolerance threshold. At concentrations above the threshold, Na can have both direct and indirect detrimental effects on plants. Direct effects are caused by the accumulation of toxic levels of Na and are generally limited to woody species. The ability of a plant to tolerate excessive amounts of Na varies widely among species and rootstocks. Na injury on avocados, citrus, and stone fruit is rather widespread and can occur at Na concentrations as low as 5 mol m3 in soil water. The symptoms may not appear immediately after exposure to saline water, however. Initially, Na is retained in the roots and lower trunk, but after 3 or 4 years the conversion of sapwood to heartwood apparently releases the accumulated Na, which is transported to the leaves and causes leaf burn (Bernstein et al. 1956).

430

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Indirect effects include both nutritional imbalance and impairment of soil physical conditions. The nutritional effects of Na are not simply related to the exchangeable Na percentage of soils but depend on the concentrations of Na, Ca2, and Mg2 in the soil solution. In sodic, nonsaline soils, total soluble salt concentrations are low and, consequently, Ca2 and/or Mg2 concentrations are often nutritionally inadequate. These deficiencies, rather than Na toxicity per se, are usually the primary cause of poor plant growth among nonwoody species and, in many cases, woody species as well. Furthermore, since Na uptake by plants is strongly regulated by Ca2 in the soil solution, the presence of sufficient Ca2 is essential to prevent the accumulation of toxic levels of Na. This is particularly important with Na-sensitive woody crops. As a general guide, Ca2 and Mg2 concentrations in the soil solution above 1 mol m3 each are nutritionally adequate in nonsaline, sodic soils (Carter et al. 1979; Hanson 1983). As the total salt concentration increases into the saline-sodic range, Ca2 concentrations become adequate for most plants and osmotic effects begin to predominate. However, some species are susceptible to salinityinduced Ca2 deficiencies as previously indicated. Therefore, for most crops species, rather than having tolerance limits for Na per se, it would be more valuable to list a favorable Na/Ca ratio or sodium adsorption ratio (SAR), an approach used by Ayers and Westcot (1985). Sodic soil conditions affect almost all crops because of the deterioration of soil physical conditions. Dispersion of soil aggregates in sodic soils decreases soil permeability to water and air, thereby reducing plant growth. Poorly structured soils also result in prolonged saturated environments, encouraging root disease. Therefore, yield reductions in crops that are not specifically sensitive to Na generally reflect the combined effects of nutritional problems and all problems associated with impaired soil physical conditions. Chloride Chlorine is an essential micronutrient for plants but, unlike most micronutrients, it is relatively nontoxic when supplied at low concentrations sufficient only to meet plant requirements (Maas 1986). In fact, most nonwoody crops are not specifically sensitive to Cl even at higher concentrations. One exception to this generalization involves certain cultivars of soybeans that tend to accumulate excessive and toxic amounts of Cl (Abel and McKenzie 1964; Parker et al. 1983). Tolerant cultivars restrict Cl transport to the shoots. Many woody species are also susceptible to Cl toxicity, which varies among varieties and rootstocks within species. As in soybeans, these differences usually reflect the plant’s ability to prevent or retard Cl translocation to the shoots or scions. Cooper (1951, 1961) found that the salt tolerance of avocados, grapefruits, and oranges is

PLANT SALT TOLERANCE

431

closely related to the Cl accumulation properties of the rootstocks. Similar effects of rootstocks on salt accumulation and tolerance have been reported for stonefruit (Bernstein et al. 1956) and nut trees (Ferguson et al. 2002). Large differences in the salt tolerance of grape varieties have been linked with the Cl-accumulating characteristics of different rootstocks (Ehlig 1960; Sauer 1968; Bernstein et al. 1969; Groot Obbink and Alexander 1973). By selecting rootstocks that exclude Cl from the scions, this problem can be avoided. Table 13-4 lists the maximum Cl concentrations permissible in the soil water that do not cause leaf injury in selected fruitcrop cultivars and rootstocks. In some cases, however, the osmotic threshold may be exceeded so that yield is decreased without obvious injury. The list is by no means complete, and most popular rootstocks are not listed because quantitative data are not available. The major detrimental effect of Cl results from its contribution to the overall osmotic stress. No comprehensive testing has been done to specifically determine crop tolerances to Cl salinity but, since most of the salttolerance data were obtained in field plots salinized with Cl salts of Na and Ca2, the data can be converted to express tolerances in terms of Cl concentration. If Cl is the predominant anion in the soil solution, then Cl concentration [Cl], expressed in meq/L (mmolc/L) is approximately 10 times the EC expressed in dS/m (USSL 1954). Therefore, multiplying the threshold values given in Tables 13-1 and 13-2 by 10 gives the maximum allowable Cl concentration in mol m3 in the saturated-soil extract without a loss in yield. Dividing the slope by 10 estimates the percent yield-potential decrease expected per each 1 mol m3 increase in Cl concentration above the threshold. Boron Boron (B) is an essential micronutrient for plants. The optimum concentration range of plant-available B, however, is very narrow for most crops. Various criteria have been proposed to define levels that are necessary for adequate B nutrition and yet low enough to avoid B toxicity symptoms, plant injury, and subsequent yield reduction (Ayers and Westcot 1985; Gupta et al. 1985; Keren and Bingham 1985). Boron deficiency is more widespread than B toxicity, particularly in humid climates, whereas excess B toxicity tends to be more of a concern in arid environments. Like salt tolerance, B tolerance fluctuates with climate, soil conditions, and plant variety. Much of the existing B tolerance data were obtained from experiments conducted from 1930 through 1934 by Eaton (1944). These data provided threshold tolerance limits for more than 40 different crops. While very useful, Eaton’s experimental data cannot be fitted to any reliable

432

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 13-4. Chloride Tolerance Limits of Some Fruit-Crop Rootstocks and Cultivars

Crop (1) Rootstocks Avocado (Persea americana)

Rootstock or Cultivar (2)

Maximum Permissible Cl in Soil Water without Leaf Injurya (mol m3) (3)

West Indian Guatemalan Mexican

15 12 10

Citrus (Citrus sp.)

Sunki mandarin, grapefruit Cleopatra mandarin, Rangpur lime Sampson tangelo, rough lemon Sour orange, Ponkan mandarin Citrumelo 4475, trifoliate orange Cuban shaddock, Calamondin Sweet orange, Savage citrange Rusk citrange, Troyer citrange

50 50 30 30 20 20 20 20

Grape (Vitis sp.)

Salt Creek, 1613-3 Dog Ridge

80 60

Stone fruit (Prunus sp.)

Marianna Lovell, Shalil Yunnan

50 20 15

Boysenberry Olallie blackberry Indian Summer raspberry

20 20 10

Grape (Vitis sp.)

Thompson seedless, Perlette Cardinal, Black Rose

40 20

Strawberry (Fragaria sp.)

Lassen Shasta

15 10

Cultivars Berriesb (Rubus sp.)

a

For some crops, these concentrations may exceed the osmotic threshold and cause some yield reduction. b Data available for one variety of each species only Adapted from Maas and Grattan (1999).

PLANT SALT TOLERANCE

433

growth–response function for most crops. Nevertheless, his results are the source of most of the threshold tolerance limits presented in Table 13-5. Plant response to excess B can be described by the two-piece linear response model (Bingham et al. 1985; Francois 1984, 1986, 1988, 1989, 1991, 1992). These data have provided threshold and slope parameters for a limited number of crops and are included in Table 13-5. With few exceptions, the B tolerance data are based on crop responses to different B levels in sand cultures. The thresholds indicate maximum permissible concentrations in the soil water that do not cause yield reductions. Some crops, however, may exhibit leaf injury at low to moderate concentrations without decreasing yield. For example, heavy leaf damage due to longterm B accumulation in grapes had no effect on the commercial fruit yield (Yermiyahu et al. 2006). Based on response to B, crops have been classified in six groups, ranging from very sensitive to very tolerant. Like salt tolerance, B tolerance varies with climate, soil conditions, and crop cultivars; therefore the data may not apply to all cultural conditions. Because different rootstocks of citrus and stone fruits absorb B at different rates, that tolerance will likely be improved by using rootstocks that restrict B uptake. A number of these rootstocks are listed in order of increasing B accumulation in Table 13-6. Francois and Clark (1979) examined the response of 25 ornamental shrub species to irrigation with waters containing either high (7.5 mg/L) or low (2.5 mg/L) B concentrations. Boron tolerance was based on growth reduction and overall plant appearance. The salt tolerance of these species had been established in an earlier study (Bernstein et al. 1972) and no correlation was found between B tolerance and salinity tolerance of the species tested. Symptoms of boron toxicity. At the early stages, symptoms of salinity and specific ion toxicities in plants are often difficult to distinguish from each other. Foliage may be off-color green with yellowing of the leaf tips or margins. This observation, however, is of little diagnostic value unless accompanied by chemical analysis for specific ions in the tissue. As B in the root environment increases, however, characteristic visual symptoms are evident. Sharp boundaries often distinctly separate the affected and the green unaffected tissues. Leaf margins become scorched and necrotic, and finally the leaf drops prematurely. Boron toxicity patterns are generally correlated with the venation of the leaf in that chlorosis followed by necrosis appears first at the end of the veins. Parallel-veined leaves (e.g., grasses, lilies) generally show necrosis in leaf tips where the veins terminate. A similar pattern is found in lanceolet leaves (e.g., stock, carnations) where the principal vein terminates in the tip. In species of geranium or broccoli, for example, where veins are of more radial distribution, B toxicity appears as an injured zone

434

TABLE 13-5. Boron Tolerance Limits for Agricultural Crops Crop Common Name (1) Alfalfa Apricot Artichoke, globe Artichoke, Jerusalem Asparagus Avocado Barley Bean, kidney Bean, lima Bean, mung Bean, snap Beet, red Blackberry Bluegrass, Kentucky Broccoli Cabbage Carrot Cauliflower Celery Cherry Clover, sweet Corn

Botanical Name (2) Medicago sativa L. Prunus armeniaca L. Cynara scolymus L. Helianthus tuberosus L. Asparagus officinalis L. Persea americana Mill. Hordeum vulgare L. Phaseolus vulgaris L. Phaseolus lunatus L. Vigna radiata (L.) R. Wilcz. Phaseolus vulgaris L. Beta vulgaris L. Rubus sp. L. Poa pratensis L. Brassica oleracea L. (Botrytis group) Brassica oleracea L. (Capitata group) Daucus carota L. Brassica oleracea L. (Botrytis group) Apium graveolens L. var. dulce (Mill.) Pers. Prunus avium L. Melilotus indica All. Zea mays L.

Boron-Tolerance Parameters Tolerance Based On: (3) Shoot DW Leaf and stem injury Laminae DW Whole plant DW Shoot DW Foliar injury Grain yield Whole plant DW Whole plant DW Shoot length Pod yield Root DW Whole plant DW Leaf DW Head FW Whole plant DW Root DW Curd FW Petiole FW Whole plant DW Whole plant DW Shoot DW

Thresholda (g m3) (4) 4.0–6.0 0.5–0.75 2.0–4.0 0.75–1.0 10.0–15.0 0.5–0.75 3.4 0.75–1.0 0.75–1.0 0.75–1.0 1.0 4.0–6.0 < 0.5 2.0–4.0 1.0 2.0–4.0 1.0–2.0 4.0 9.8 0.5–0.75 2.0–4.0 2.0–4.0

Slope (% per g m3) (5)

4.4

12

1.8

1.9 3.2

Ratingb (6) T S MT S VT S MT S S S S T VS MT MS MT MS MT VT S MT MT

435

Cotton Cowpea Cucumber Fig, Kadota Garlic Grape Grapefruit Lemon Lettuce Lupine Muskmelon Mustard Oats Onion Orange Parsley Pea Peach Peanut Pecan Pepper, red Persimmon Plum Potato Radish Sesame Sorghum Squash, scallop Squash, winter

Gossypium hirsutum L. Vigna unguiculata (L.) Walp. Cucumis sativus L. Ficus carica L. Allium sativum L. Vitis vinifera L. Citrus x paradisi Macfady Citrus limon (L.) Burm. f. Lactuca sativa L. Lupinus hartwegii Lindl. Cucumis melo L. (Reticulatus group) Brassica juncea Coss. Avena sativa L. Allium cepa L. Citrus sinensis (L.) Osbeck Petroselinum crispum Nym. Pisum sativa L. Prunus persica (L.) Batsch. Arachis hypogaea L. Carya illinoinensis (Wangenh.) C. Koch Capsicum annuum L. Diospyros kaki L. f. Prunus domestica L. Solanum tuberosum L. Raphanus sativus L. Sesamum indicum L. Sorghum bicolor (L.) Moench Cucurbita pepo L. var melopepo (L.) Alef. Cucurbita moschata Poir

Boll DW Seed yield Shoot DW Whole plant DW Bulb yield Whole plant DW Foliar injury Foliar injury, Plant DW Head FW Whole plant DW Shoot DW Whole plant DW Grain (immature) DW Bulb yield Foliar injury Whole plant DW Whole plant DW Whole plant DW Seed yield Foliar injury Fruit yield Whole plant DW Leaf and stem injury Tuber DW Root FW Foliar injury Grain yield Fruit yield Fruit yield

6.0–10.0 2.5 1.0–2.0 0.5–0.75 4.3 0.5–0.75 0.5–0.75 < 0.5 1.3 0.75–1.0 2.0–4.0 2.0–4.0 2.0–4.0 8.9 0.5–0.75 4.0–6.0 1.0–2.0 0.5–0.75 0.75–1.0 0.5–0.75 1.0–2.0 0.5–0.75 0.5–0.75 1.0–2.0 1.0 0.75–1.0 7.4 4.9 1.0

12

2.7

1.7

1.9

1.4 4.7 9.8 4.3

VT MT MS S T S S VS MS S MT MT MT VT S T MS S S S MS S S MS MS S VT T MS (continued)

436

TABLE 13-5. Boron Tolerance Limits for Agricultural Crops (Continued) Crop Common Name (1) Squash, zucchini Strawberry Sugar beet Sunflower Sweet potato Tobacco Tomato Turnip Vetch, purple Walnut Wheat

Botanical Name (2) Cucurbita pepo L. var melopepo (L.) Alef. Fragaria sp. L. Beta vulgaris L. Helianthus annuus L. Ipomoea batatas (L.) Lam. Nicotiana tabacum L. Lycopersicon lycopersicum (L.) Karst. ex Farw. Brassica rapa L. (Rapifera group) Vicia benghalensis L. Juglans regia L. Triticum aestivum L.

a

Boron-Tolerance Parameters Tolerance Based On: (3) Fruit yield Whole plant DW Storage root FW Seed yield Root DW Laminae DW Fruit yield

Thresholda (g m3) (4) 2.7 0.75–1.0 4.9 0.75–1.0 0.75–1.0 2.0–4.0 5.7

Root DW Whole plant DW Foliar injury Grain yield

2.0–4.0 4.0–6.0 0.5–0.75 0.75–1.0

Slope (% per g m3) (5) 5.2

3.4

Ratingb (6) MT S T S S MT T

3.3

MT T S S

4.1

Maximum permissible concentration in soil water without yield reduction. Boron tolerances may vary, depending on climate, soil conditions, and crop varieties. b The B tolerance ratings are based on the following threshold concentration ranges: 0.5 g m; s3 very sensitive (VS); 0.5–1.0 sensitive (S); 1.0–2.0 moderately sensitive (MS); 2.0–4.0 moderately tolerant (MT); 4.0–6.0 tolerant (T); and 6.0 very tolerant (VT). Adapted from Maas and Grattan (1999).

PLANT SALT TOLERANCE

437

TABLE 13-6. Citrus and Stone-Fruit Rootstocks Ranked in Order of Increasing Boron Accumulation and Transport to Scions Common Name

Botanical Name

Citrus Alemow Gajanimma Chinese box orange Sour orange Calamondin Sweet orange Yuzu Rough lemon Grapefruit Rangpur lime Troyer citrange Savage citrange Cleopatra mandarin Rusk citrange Sunki mandarin Sweet lemon Trifoliate orange Citrumelo 4475 Ponkan mandarin Sampson tangelo Cuban shaddock Sweet lime

Citrus macrophylla C. pennivesiculata or C. moi Severina buxifolia C. aurantium x Citrofortunella mitis C. sinensis C. junos C. limon C. x paradisi C. x limonia x Citroncirus webberi x Citroncirus webberi C. areticulata x Citroncirus webberi C. reticulata C. limon Poncirus trifoliata P. trifoliata x C. paradisi C. reticulata C. x Tangelo C. maxima C. aurantiifolia

Stonefruit Almond Myrobalan plum Apricot Marianna plum Shalil peach

Prunus duclis P. cerasifera P. armeniaca P. domestica P. persica

Adapted from Maas and Grattan (1999).

all around the margin. In leaves with a well-developed network of veins, and with many veins ending in areas between principal side veins (gerberas, asters, eucalyptus, most citrus species), symptoms first develop as interveinal yellow or red spots. As injury progresses, chlorosis spreads to the margins (Oertli and Kohl 1961). Other B toxicity symptoms commonly observed in landscape plants include terminal twig dieback, necrotic leaf spots, abnormal leaf forms

438

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and texture, and bark cracking. Necrosis associated with B is often black and sometimes red (e.g., eucalyptus) and is most severe on the older foliage (Chapman 1966). Characteristic symptoms of B toxicity in stonefruit trees are reduction of flower bud formation, poor fruit set, and malformed fruit exceptionally poor flavor (Johnson 1996). In citrus trees, symptoms often progress from tip chlorosis and mottling to the formation of tan-colored, resinous blisters on the underside of the leaves (Wutscher and Smith 1996). Boron toxicity and how it is expressed by plants is related to some extent on the mobility of B in the plant. Although in most plant species B is thought to be immobile, accumulating in the margins and tips of the oldest leaves, B can be remobilized by some species (Brown and Shelp 1997). These B-mobile plants have high concentrations of polyols (sugar alcohols) that bind with the B and allow it to be mobilized in the shoot. Examples include almonds, apples, grapes, and most stonefruits. For these crops, B concentrations are higher in younger tissue than in older tissue, and injury is expressed in the young, developing tissue. This likely explains symptoms, such as reduced bud formation and twig die-back. Boron-immobile plants, such as pistachios, tomatoes, walnuts, and figs, do not have high concentrations of polyols and the B concentrates in the margins of older leaf tissue. Injury in these crops is expressed as the classical necrosis on leaf tips and margins. Salinity–boron interactions. Because excess B often occurs in areas with saline soils and waters, it is relevant to consider B uptake by plants under saline conditions inasmuch as B toxicity may be confounded with the associated problems of salt accumulation (Nicholaichuk et al. 1988). Although plant response to high concentrations of B in the root media has been extensively reviewed (Nable et al. 1997), the interactive effects of salinity and B on plant performance have received less attention (Grieve and Poss 2000; Alpaslan and Gunes 2001; Ben-Gal and Shani 2002; Diaz and Grattan 2009; Edelstein et al. 2005; Tripler et al. 2007; Yermiyahu et al. 2007, 2008). Moreover, studies addressing the interaction of the dual stresses on crop response reach widely different conclusions. Bingham et al. (1987) reported that wheat shoot growth was influenced by each stress independently but not by their interaction. Several studies have shown that salt stress may increase B toxicity symptoms and reduce crop yield (Aspaslan and Gunes 2001; Grieve and Poss 2000; Supanjani 2006; Wimmer et al. 2003). Conversely, results of other studies suggest that increased salinity may reduce B uptake and mitigate its toxic effects in wheat (Holloway and Alston 1992), chickpeas (Yadav et al. 1989), melons (Edelstein et al. 2005), and eucalyptus (Grattan et al. 1997; Marcar et al. 1999). Recent research at the USDA-ARS U.S. Salinity Laboratory shows that there are complex interactions among salinity, B, and pH (Grieve et al.

PLANT SALT TOLERANCE

439

2010; Smith et al. 2010a,b). These new findings suggest that more research is needed to better understand the mechanisms of these interactions. Selenium and other trace elements Since the publication of ASA Monograph 38 Agricultural Drainage (Skaggs and van Schilfgaarde 1999), environmental concerns related to trace elements have added a new dimension to drainage management and disposal. Certain irrigated soils derived from sedimentary rock materials contain, or at least originally contained, high concentrations of trace elements that dissolve in the soil-water and move to the shallow groundwater. Pratt and Suarez (1990) provide a table that lists the recommended maximum concentration of 15 trace elements in irrigation waters that provide long-term protection of plants and animals. However, the recent concern that these elements pose for irrigated agriculture is not so much their effects on limiting production but rather the toxicological effects they can cause when drainage effluents that contain them are used for irrigation or are discharged into bodies of water. If such effluents are used to supplement irrigation water supplies, certain trace elements may accumulate in the soil and/or crop to levels that pose a health hazard to consumers. Molybdenum (Mo) and selenium (Se) are readily absorbed by plants and can be toxic to animals and humans (Page et al. 1990). If trace-element-tainted drainage effluents are discharged into channels, lakes, ponds, estuaries, or other bodies of water, there are ecological concerns that they may concentrate as they move up the food chain, a process called biomagnification. The composition of salts in the drainage effluent can influence the uptake of certain trace elements by plants. Selenium, for example, is found in soil solutions in California’s San Joaquin Valley, where it exists together with high concentrations of sulfate. Uptake of both SeO42 and SO42 by plants is mediated by the same high-affinity enzyme, and the anions compete for binding sites on this cell membrane carrier (Läuchli 1993). Plant accession of Se from a substrate high in sulfate will be significantly lower than from a Cl system. Irrigation with drainage water dominated by sulfate salts reduced selenate accumulation in vegetables (Burau et al. 1991), wheat (Grieve et al. 1999), soybeans (Wang et al. 2005), and the seed oil crop lesquerella (Grieve et al. 2001). In some areas where total soil Se is high (5 mg/kg dry wt) and sulfate concentrations are much lower than those in the San Joaquin Valley, plants can accumulate Se to phytotoxic levels, such as the case with wheat grown in an isolated area in the Punjab state of India (G. S. Dhillon, personal communication, 2007). Selenium accumulation by plants is also influenced by the irrigation method. Although root uptake of Se is inhibited by the presence of sulfate in the external media, a similar interaction apparently does not occur in leaf tissue. Therefore, Se is readily taken up by leaves of forage forage Brassica

440

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

species (Suarez et al. 2003), Swiss chard, spinach, and soybeans (D. L. Suarez, personal communication) sprinkler-irrigated with Se-containing, sulfate-dominated saline waters. Studies conducted in sand tank cultures have shown that sulfate salinity can also reduce Mo accumulation in alfalfa shoots (Grattan et al. 2004b) but had the opposite effect on tall wheatgrass (cv ‘Jose’) (Diaz and Grattan 2009).

PARAMETERS INFLUENCING PLANT RESPONSE TO SALT STRESS Although crop yields are a function of salt concentrations within the rootzone, it must be recognized that this relationship is influenced by interactions between salinity and various soil, water, and climatic conditions. While other environmental stresses may limit crop yield, they may increase, decrease, or have no effect on the apparent salt tolerance of the crop. It is important, therefore, that the effects of any interacting factor be compared on the basis of relative crop yield. Even though expressing the yields on a relative basis minimizes large differences in absolute yield from experiments conducted in different sites and conditions, these factors can still affect the apparent salt tolerance expressed on a relative basis. Soil water content Salt-affected crops often must contend with water deficits or excess as well. Therefore, actual crop performance during the growing season is related to how the plant responds to both salinity and water stress. In flooded or poorly drained soils, the overall diffusion of oxygen to roots is reduced, thereby limiting root respiration and plant growth (Sharpley et al. 1992). When the rootzone is saturated with saline water, the combined effects of salinity and oxygen deficiency can adversely affect seed germination (Aceves-N et al. 1975), selective ion transport processes in the plant (Drew et al. 1988; Barrett-Lennard 2003), and shoot growth (Aubertin et al. 1968; Aragüés et al. 2004; Isidoro and Aragüés 2006). Water deficit, at least to some degree, is practically unavoidable under field conditions, since the soil-water content varies temporally and spatially throughout the season. Exactly how the plant responds to the combination of stresses from salinity and water deficit remains unresolved (Meiri 1984). Obviously the combination of stresses is more damaging than either one alone, but are they additive or antagonistic? Quantifying the growth-limiting contribution of each is difficult, since both change over time and space. Water-deficit stress may predominate in the upper rootzone, while salt stress may predominate in the lower rootzone.

PLANT SALT TOLERANCE

441

Wadleigh and Ayers (1945) first demonstrated that bean plants responded to the additive combination of water deficit and salt stress. However, Meiri (1984) concluded from data collected by Parra and Romero (1980) that matric potential affected bean shoot growth more than did osmotic potential. Thermodynamically, matric and osmotic components are additive, but resistance to soil water flow must be considered. For example, plant response to these stresses under conditions of low evaporative demand is likely to be different than that observed under high evaporative demand, since matric rather than osmotic potential dominates control of water flow from soil to roots. The magnitude of the difference may be related to differences in evaporative demand and rootlength density. Regardless of how plants respond to integrated stress, they presumably do better when grown on saline soils if water-deficit stress is minimized. However, increasing irrigation frequency does not necessarily improve yields of salt-stressed crops (Bresler and Hoffman 1986; Shalhevet et al. 1982, 1986). Salt-stressed plants are smaller, grow slower than non-salt-stressed plants, and require less water over a given time. Consequently, salt-stressed plants deplete a smaller percentage of available soil water than do nonsaline plants, so they are less responsive to frequent irrigations. Therefore, increased irrigation frequency benefits salt-stressed plants only when it reduces water stress; maintains the salt concentration in the soil solution below growth-limiting levels; and does not contribute to additional stresses, such as O2 deficit or root disease (e.g., phytophthora). As Wadleigh and Ayers (1945) concluded more than a half a century ago, it is not that salt-stressed plants should necessarily be irrigated more frequently, but rather that they should be irrigated at lower soilwater depletion. Salt composition. The composition of salts in water varies widely across the globe. In most waters, the dominant cations are Na, Ca2, and Mg2, while the dominant anions are Cl, SO42, and HCO 3 (Grattan and Grieve 1999). Most horticultural crops are subjected to irrigation water or soil solutions with Na/(Na  Ca2) in the range of 0.1 to 0.7, suggesting that the composition of saline water employed in experimental studies should reflect this ratio. Despite recommendations by early investigators of plant salt tolerance that plants under salt stress require higher concentrations of Ca2 than under nonsaline conditions (Hayward and Wadleigh 1949; Pearson 1959; Hayward and Bernstein 1958; Bernstein 1975), a high percentage of salinity studies of agronomic and horticultural crops continue to be conducted with NaCl as the sole salinizing agent. The use of this unrealistic salinizing composition may induce ion imbalances that contribute to Na-induced Ca2 deficiencies and Ca-related physiological disorders in certain susceptible crops (Shear 1975; Maas and Grieve 1987;

442

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Sonneveld 1988; Suarez and Grieve 1988). Furthermore, the use of singlesalt solutions in salt-tolerance experiments may result in misleading and erroneous interpretations about plant response to salinity. A similar argument can be offered for the anions. Although the majority of salinity studies use Cl as the sole salinizing anion, most soil solutions contain a high proportion of SO42and HCO3 . Plants can perform equally well with moderate variations in the Cl/SO42 ratio, but at measurably low ratios at the same salinity level, some plants perform better in the sulfate-dominated solutions. Bicarbonate is somewhat different from Cl or SO42 because it can be damaging under even mildly saline conditions when it is the dominant anion. It is likely that more can be learned if future salinity-nutrition studies, regardless of experimental scale or objectives, are conducted with more realistic ion ratios. Much of the salt-tolerance information has been derived from studies of plant responses to Cl-dominated saline irrigation waters that typically contain both NaCl and CaCl2.. A few research teams have evaluated plant salt tolerance by using irrigation waters prepared to simulate recycled or saline waters typical of a specific location or site. Dutch growers frequently employ solutions with compositions adjusted to the average salt composition of surface waters in the western region of the Netherlands (Bik 1980; Sonneveld 1988). Saline waters (EC  2.5 to 4.5 dS m1) from local wells in Israel continue to be used successfully for cut-flower production on more than 700 ha throughout the Negev Desert (Shillo et al. 2002). Arnold and coworkers (2003) demonstrated that recycled runoff effluents from a nursery operation and water from a constructed wetland were suitable for irrigating certain bedding and cut-flower crops. Irrigation waters used in recent research at the U.S. Salinity Laboratory were prepared to mimic waters available at three locations within California: (1) Na- and SO42-dominated drainage effluents present in the San Joaquin Valley (Grattan et al. 2004a,b; Grieve et al. 2005; Skaggs et al. 2006a,b); (2) compositions of increasing salinity that would result from concentration of Colorado River waters (Grieve et al. 2006); and (3) waters affected by seawater intrusion along the California coastal areas (Carter et al. 2005; Carter and Grieve 2008). Soil biota Full coverage of the interactions of salinity and soil flora and fauna is clearly beyond the scope of this chapter. However, the importance of soil organisms cannot be ignored. The use of controlled mycorrhization has been shown to alleviate deleterious effects of salt stress and improve yields of tomatoes (Al-Karaki 2006), lettuce (Ruiz-Lozano et al. 1996), sorghum (Cho et al. 2006), and bananas (Yano-Melo et al. 2003). Rhizobium spp., which are integral to legume production, seem more salt-tolerant than

PLANT SALT TOLERANCE

443

their host plants, but evidence indicates that nodulation and N2 fixation by some crops are impaired by salinity (Läuchli 1984). Growth of several legumes was reduced more when grown symbiotically than with N fertilization. Some investigators have suggested that mycorrhizal symbioses improve the ability of some crops to tolerate salt by improving P nutrition (Hirrel and Gerdemann 1980; Ojala et al. 1983; Poss et al. 1985) or by enhancing K⬊Na status (Sannazzaro et al. 2006). Although salinity does not specifically cause plant diseases, saltstressed plants may be predisposed to infection by soil pathogens. Salinity has been reported to increase the incidence of phytophthora root rot in chrysanthemums (MacDonald 1982), citrus (Blaker and MacDonald 1986), chili peppers (Sanogo 2004), and tomatoes (Snapp et al. 1991); the colonization of pistachios (Mohammadi et al. 2007) and olives (Levin et al. 2007) rootstocks by Verticillium dahlia; and the incidence and severity of crown and root rot of tomatoes by Fusarium oxysporum (Triky-Dotan et al. 2005). The combined effects significantly reduced fruit size and yield of tomatoes (Snapp et al. 1991). Wetter soil under salt-stunted plants may contribute to increased susceptibility to fungal diseases. Inadequate drainage could exacerbate this condition. Soil fertility In irrigated agriculture, fields are usually fertilized to achieve maximum productivity. Sometimes fertilizer applications are inadequate or even omitted because of cost or availability. If crops are grown on lowfertility soils, they may seem more salt-tolerant than those grown with adequate fertility. The reason is that fertility, not salinity, is the primary factor limiting plant growth. Proper fertilizer applications would increase yields whether or not the soil was saline but proportionately more if it were nonsaline. The results of Bernstein et al. (1974) indicate that the effects of salinity and nutritional stresses tend to be additive, provided that neither of these stresses are extreme. When yields are limited similarly by salinity and infertility, the effects of decreasing salinity or increasing fertility will give similar benefits. However, if yields are reduced much more by one factor than the other, alleviating the most severe condition will increase yield more than alleviating the less restrictive condition. Therefore, one must be careful in interpreting salinity  fertility studies in terms of whether fertilizer additions increase or decrease crop salt tolerance. Response functions are based on relative crop yield as salinity increases from non-growth-limiting to severely growthlimiting levels (Maas and Grattan 1999). Although suboptimal soil fertility may be the most growth-limiting factor at low salinity, salt stress may be the most growth-limiting at higher salinity levels with the same level of fertility (Grattan and Grieve 1994). Therefore, depending on the

444

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

severity of salt stress, fertilizer additions may increase or decrease crop salt tolerance. Although crop salt tolerance is expressed on a relative basis, actual yields must be considered in evaluating the benefits of fertilizer. For example, fertilizer additions may decrease crop salt tolerance, but it still may be economically advantageous to fertilize if absolute yields are increased. However, unless salinity causes specific nutritional imbalances, fertilizer applications exceeding that required under nonsaline conditions have rarely been beneficial in alleviating growth inhibition by salinity. Most studies indicate that excess N, P, and K applications have little effect or that they reduce salt tolerance (Grattan and Grieve 1994); however, Ravikovitch and Yoles (1971) found that N, P, or both seemed to increase the salt tolerance of millet and clover. Reliable data on the salt tolerance of crops during emergence and seedling growth are extremely limited (Maas and Grieve 1994). Although salt stress may delay emergence, the final emergence percentage for most crops is not affected if salt concentrations remain at or below the tolerance threshold for mature yields. No systematic evaluation of the tolerance of crop seedlings grown under actual or simulated field conditions has ever been undertaken. Clearly, more research is needed to better understand how crops respond to integrated stresses they encounter between germination and emergence. Irrigation methods The method of irrigation can affect the crop’s response to salinity. The irrigation method (1) influences the salt distribution in the soil, (2) determines whether leaves will be subjected to wetting, and (3) determines the ease at which high soil-water potentials can be achieved (Bernstein and Francois 1973; Shalhevet 1984). Since irrigation methods that maintain a higher soil-water potential reduce the time-averaged salt concentration in the soil-water, they allow for optimal plant performance. With pressurized systems, such as drip and sprinkler, small applications of water can be applied to fields uniformly, unlike surface irrigation methods, such as furrow, basin, or flood. Surface irrigation systems require some minimum quantity of water to enable uniform applications over the field. This minimum quantity may be in excess of the yield-threshold soilwater depletion, thereby resulting in unnecessary drainage losses. Therefore, pressurized systems (sprinkler, drip, etc.) are more conducive for light, uniform irrigations. Although irrigating at lower soil-water depletion (i.e., higher matric potential) may be desirable to maintain a favorable soil-water environment, use of sprinkler irrigation to achieve this creates an additional problem. Salts in the irrigation water can be readily absorbed by wetted foliage

PLANT SALT TOLERANCE

445

and cause foliar injury. This subject will be addressed in more detail later in this section. In light of this discussion, it is not surprising that crop salt tolerance has been found to vary under different irrigation methods (Bernstein and Francois 1973; Bernstein and Francois 1975; Meiri et al. 1982), where crop performance was best under drip irrigation and worst under sprinkler irrigation. Salt distribution patterns, such as those described, are related to the combined effects of root water extraction patterns and the net direction of water flow in the soil. Salt accumulation is lowest at the point in the soil where irrigation water contacts the soil but increases in the direction of soil water flow. Water moves in the direction where it is transpired or evaporated, thereby concentrating salts in areas where it occurs (Kruse et al. 1990). Salt accumulation patterns under furrow, sprinkler, drip, and subsurface drip irrigation methods have been described by Oster et al. (1984) and Wang et al. (2002). Subsurface drip irrigation practices can create unique salt accumulation patterns where salts accumulate in the soil above the drip line (Hanson et al. 2009). Plant roots encounter unexpected salination when rain moves salts accumulated at the soil surface back into the rootzone.

PLANT TOLERANCE TO SALINE SPRINKLING WATERS Sprinkler-irrigated crops are subject to additional salt damage when the foliage is wetted by saline water. Salts are directly accumulated by the leaves and, as a result, some species become severely injured and lose their leaves. Of course, sprinkler-irrigated crops are subject to injury from both soil salinity and salt spray. Any genetically controlled mechanisms that may have evolved in plants to restrict Na and Cl from the shoot may become irrelevant under sprinkler irrigation. The degree of injury is related to the salt concentration in the leaves, but weather conditions and water stress can influence the onset of injury. For instance, leaves may contain toxic levels of Na or Cl for several weeks without exhibiting any injury symptoms, but the first hot, dry weather will cause severe leaf necrosis. Consequently, there are no practical guidelines for correlating foliar injury to salt concentrations in the leaves. Obviously, saline irrigation water is best distributed through surface distribution systems. However, if sprinkling with marginally saline water cannot be avoided, several precautions should be considered (Maas 1986). If possible, susceptible crops should be irrigated below the plant canopy to eliminate or reduce wetting of the foliage. Since injury is related more to the number of sprinklings than to their duration, infrequent, heavy irrigations would be preferable to frequent, light irrigations. Intermittent

446

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

wetting by slowly rotating sprinklers that allow drying between cycles should be avoided. Lateral sprinkler systems might be moved downwind, when possible, so that salts accumulated on the leaves from salt drift would be washed off as the sprinkler moves past. Perhaps the best strategy for minimizing foliar injury to plants is to irrigate at night when both evaporation and salt absorption are reduced. Daytime sprinkling should be avoided on hot, dry, windy days. Sprinkling with low-salinity water for 3 to 5 minutes either prior to or after sprinkler irrigations with saline water effectively reduced foliar salt accumulation and injury in barley and corn (Aragüés et al. 1994; Benes et al. 1996). These investigators concluded that much of the salt accumulated by wetted leaves is absorbed during the first few minutes of irrigation and also after sprinkling when the saline water evaporates and concentrates on the leaf surface. Sprinkling barley with 9.6 dS m1 water, for example, reduced grain yields by 58% compared to nonsprinkled plants, but when saline-sprinkled plants received both pre- and post-washing with nonsaline water, yields were reduced only 17% (Benes et al. 1996). The soil surface was covered to shed the sprinkling waters in all cases. Post-rinsing of soybean plants with nonsaline water prevented leaf injury due to potentially toxic levels of Cl (Wang et al. 2002; Grieve et al. 2003). In this field trial, the soil surface was not covered; therefore, Cl and other ions were accumulated via both the root pathway and foliar absorption. This information may be useful to growers who have access to and can readily switch between sources of irrigation waters of different quality.

CONTROLLING SOIL SALINITY Most of the crop salt-tolerance data provided in Tables 13-1 and 13-2 reflect how the plant responds to a relatively uniform soil-salinity profile from the established seedling stage to harvest. Although useful, particularly for crop comparison purposes, field-grown crops respond to salinity profiles that change over time, making relative yield predictions understandably difficult. There are advantages, however, in imposing water management practices that allow salinity profiles to change over time, as opposed to maintaining relatively constant soil salinity profiles. With controlled changes in soil salinity, crops with different tolerances to salinity can be included within a crop rotation (Rhoades et al. 1988, 1989). Increases in soil salinity are also acceptable when the tolerance of a crop increases within a season (Shennan et al. 1995; Steppuhn et al. 2009). Adequate control of soil salinity changes requires that the farmer has access to multiple and dependable supplies of irrigation water where at least one supply is of good quality. Within limits, farmers who have irrigation water supplies of different qualities can use them alter-

PLANT SALT TOLERANCE

447

nately (cyclically) in different years or at different times of the year, or they can blend supplies to achieve a suitable quality water (Grattan and Rhoades 1990). Regardless of the irrigation water supplies and quality available to the grower, irrigation practices must be managed to control soil salinity within an acceptable level. This requires that a favorable salt balance be attained. This does not suggest, however, that a calculated leaching requirement must be achieved each irrigation. Leaching fractions (LFs) often decrease as the season progresses. In fact, a reduced LF is a consequence of a mature, deep-rooted crop actively growing in a soil with low permeability during months of high evaporative demand. Prolonged periods of saturation required to achieve leaching could produce anoxic conditions and encourage root disease. Nevertheless, a favorable salt balance must be maintained, even if intermittent leaching (e.g., during the winter, alternate years) is the only means to remove excess salts from the soil. A long-term salt balance can only be achieved at the farm scale if there is adequate drainage beyond the rootzone. Crops grown in areas affected by rising saline water tables are subjected to salination. Crop production in these situations cannot be sustained indefinitely, since a long-term salt balance cannot be achieved. Use of saline water to irrigate crops grown in soils with high water tables accelerates the problem. Moreover, the required leaching further raises the saline water table, thereby salinizing the rootzone even more. This paradox can only be overcome by adequate drainage and disposal, thereby ensuring that crop yields can be sustained over the long term (van Schilfgaarde 1990).

SUMMARY In making decisions about salinity management and the use of lowsalinity irrigation water, there are a number of variables that a grower may consider. First is that the published threshold and slope values for various crops represent statistical means, not absolute values, and actual crop tolerance falls within a range around these means. Crop selection should therefore include consideration of the relative total production of a crop, since a high-production crop may have a net economic yield high enough to offset the effects of salinity stress. In addition, there are some potential crop-specific benefits of high-salinity environments, such as increases in sugars, total soluble solids, postharvest handling characteristics, and the concentration of various flavonoids, ascorbates, tocopherols, carotinoids, and lycopene. For ornamental species, the visible effects of salt stress may also not affect the aesthetics of the plant, and salt stress effects may have benefits, such as low growth and low water uptake. Salt-tolerant plants

448

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(halophytes) may also be planted for a combination of their crop value and their value in treating drainage and other wastewater. In addition, there are plant-specific effects of high-salinity environments on nutrient uptake; for example, high sodium levels may inhibit plant uptake of Ca2. The grower may wish to consider the plant-specific effects of the specific salt composition of the field on the proposed crop and adjust cropping accordingly. Finally, salt stress may be affected by soil-water content, salt composition, soil biota, soil fertility, irrigation method, and timing of irrigation. Management of these variables may reduce the cumulative stresses on a crop and thus minimize the net impact of salt stress. REFERENCES Abel, G. H., and MacKenzie, A. J. (1964). “Salt tolerance of soybean varieties (Glycine max L. Merrill) during germination and later growth.” Crop Sci., 4, 157–161. Aceves-N., E., Stolzy, L. H and Mehuys, G. R. (1975). “Combined effects of low oxygen and salinity on germination of semi-dwarf Mexican wheat.” Agron. J., 67, 530–532. Adams, P., and Ho, L. C. (1989). “Effects of constant and fluctuating salinity on the yield, quality and calcium status of tomatoes.” J. Hort. Sci., 64, 725–732. Al-Karaki, G. N. (2006). “Nursery inoculation of tomato with arbuscular mycorrhizal fungi and subsequent performance under irrigation with saline water.” Sci. Hort., 109, 1–7. Alpaslan, M., and Gunes, A. (2001). “Interactive effects of boron and salinity stress on the growth, membrane permeability and mineral composition of tomato and cucumber plants.” Plant Soil, 236, 123–128. Aragüés, R., Puy, J., and Isidoro, D. (2004). “Vegetative growth response of young olive trees (Olea europaea L., cv. Arbequina) to soil salinity and waterlogging.” Plant Soil, 258, 69–80. Aragüés, R., Puy, J., Royo, A., and Espada J. L. (2005). “Three-year field response of young olive trees (Olea europaea L., cv. Arbequina) to soil salinity: Trunk growth and leaf ion accumulation.” Plant Soil, 271, 265–271. Aragüés, R., Royo, A., and Grattan, S. R. (1994). “Foliar uptake of sodium and Cl in barley sprinkler-irrigated with saline water. Effect of pre-irrigation with fresh water.” Eur. J. Agron., 3, 9–16. Arnold, M. A., Lesikar, B. J., McDonald, G. V., Bryan, D. L., and Gross, A. (2003). “Irrigating landscape bedding plants and cut flowers with recycled nursery runoff and constructed wetland treated water.” J. Environ. Hort., 21, 89–95. Aronson, J. A. (1989). HALOPH: A database of salt tolerant plants of the world, Office of Arid Lands Studies, University of Arizona Press, Tucson, Ariz. Ashraf, M., and Harris, P. J. C. (2004). “Potential biochemical indicators of salinity tolerance in plants.” Plant Sci., 166, 3–16. Aubertin, G. M., Rickman, R.W., and Letey, J. (1968). “Differential salt-oxygen levels influence plant growth.” Agron. J., 60, 345–349.

PLANT SALT TOLERANCE

449

Awad, A. S., Edwards, D. G., and. Campbell, L. C. (1990). “Phosphorus enhancement of salt tolerance of tomato.” Crop Sci., 30, 123–128. Ayers, A. D., and Eberhard, D. L. (1960). “Response of edible broadbean to several levels of salinity.” Agron. J., 52, 110–111. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Bangerth, F. (1979). “Calcium-related physiological disorders of plants.” Ann. Rev. Plant Physiol., 97, 1136–1145. Barrett-Lennard, E. G. (2003). “The interaction between waterlogging and salinity in higher plants: Causes, consequences and implication.” Plant Soil, 253, 35–54. Bartosz, G. (1997). “Oxidative stress in plants.” Acta Physiol. Plant., 19, 47–64. Benes, S. E., Aragüés, R., Austin, R. B, and. Grattan, S. R. (1996). “Brief pre- and post-irrigation sprinkling with freshwater reduces foliar salt uptake in maize and barley sprinkler irrigated with saline water.” Plant Soil, 180, 87–95. Ben-Gal, A., and Shani, U. (2002). “Yield, transpiration and growth of tomatoes under combined excess boron and salinity stress.” Plant Soil, 247, 211–221. Bernstein, L. (1962). “Salt-affected soils and plants,” in Problems in arid zones, Proc., 18th UNESCO Symp., Paris, 139–174. ———. (1964). Salt tolerance of plants, USDA Information Bulletin 283, U.S. Government Printing Office, Washington, D.C. ———. (1975). “Effects of salinity and sodicity on plant growth.” Ann. Rev. Phytopath., 13, 295–312. Bernstein, L., and Ayers, A. D. (1953). “Salt tolerance of five varieties of carrots.” Proc. Am. Soc. Hort. Sci., 67, 367–370. Bernstein, L., Ayers, A. D., and Wadleigh, C. H. (1951). “The salt tolerance of White Rose potatoes.” Am. Soc.. Hort. Sci., 57, 231–236. Bernstein, L., Brown, J. W., and Hayward, H. E. (1956). “Influence of rootstock on growth and salt accumulation in stone-fruit trees and almonds.” Proc. Am. Soc. Hort. Sci., 68, 86–95. Bernstein, L., Ehling, C. F., and Clark, R. A. (1969). “Effect of grape rootstocks on Cl accumulation in leaves.” J. Amer. Soc. Hort. Sci., 84, 584–590. Bernstein, L., and Francois, L. E. (1973). “Comparison of drip, furrow, and sprinkler irrigation.” Soil Sci., 115, 73–78 ———. (1975). “Effects of frequency of sprinkling with saline waters compared with daily drip irrigation.” Agron. J., 67, 185–190. Bernstein, L., Francois, L. E., and Clark, R. A. (1972). “Salt tolerance of ornamental shrubs and ground covers.” J. Am. Soc. Hort. Sci., 97, 550–556. ———. (1974). “Interactive effects of salinity and fertility on yields of grains and vegetables.” Agron. J., 66, 412–421. Bik, R. A. (1980). “Effect of irrigation water salinity on postharvest performance of cyclamen grown on rockwool and potting compost.” Act Hort., 99, 189–196. Bingham, F. T., Strong, J. E., Rhoades J. D., and Keren, R. (1985). “An application of the Maas-Hoffman salinity response model for boron toxicity.” Soil Sci. Soc. Am. J., 49, 672–674. ———. (1987). “Effects of salinity and varying boron concentrations on boron uptake and growth of wheat.” Plant Soil, 97, 345–351.

450

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Blaker, N. S., and MacDonald, J. D. (1986.) “The role of salinity in the development of Phytophthora root rot of citrus.” Phytopathology, 76, 970–975. Botia, P., Navarro, J. M., Cerdá, A., and Martinez, V. (2005). “Yield and fruit quality of two melon cultivars irrigated with saline water at different stages of development.” Europ. J. Agron., 23, 243–253. Bresler, E., and Hoffman, G. J. (1986.) “Irrigation management for soil salinity control: Theories and tests.” Soil Sci. Soc. Am., J. 50, 1552–1560. Brown, P. H., and Shelp, B. J. (1997). “Boron mobility in plants.” Plant Soil, 193, 85–101. Burau, R. G., Shennan, C., May, D., and Grattan, S. (1991). Uptake of boron and selenium by vegetables grown in the field as affected by boric acid, selenate and sulfate levels in irrigation water and the impact on the diets of consumers, 1990–1991, University of California Salinity/Drainage Task Force Technical Progress Report, University of California Division of Agricultural and Natural Resources. Cabrera, R. I. (2000). “Evaluating yield and quality of roses with respect to nitrogen fertilization.” Acta Hort., 511, 133–142. Cabrera, R. I., and Perdomo, P. (2003). “Reassessing the salinity tolerance of greenhouse roses under soilless production conditions.” Hort. Sci., 38, 533–536. Campos, C. A. B., Fernandes, P. D., Gheyi, H. R., Blanco, F. F., and. Campos, S. A. F. (2006). “Yield and fruit quality of industrial tomato under saline irrigation.” Sci. Agric., 63, 146–152. Carter, C. T., and Grieve. C. M. (2008). “Mineral nutrition, growth, and germination of Antirrhinum majus L. (snapdragon) when produced under increasingly saline conditions.” Hort. Sci., 43, 710–718. Carter, C. T., Grieve, C. M., Poss, J. A., and Suarez, D. L. (2005). “Production and ion uptake of Celosia argentea irrigated with saline wastewaters.” Sci. Hort., 106, 381–394. Carter, M. R., Webster, G. R., and Cairns, R. R. (1979). “Calcium deficiency in some solonetzic soils of Alberta.” J. Soil Sci., 30, 161–174. Catlin, P. B., Hoffman, G. J., Mead, R. M., and Johnson, R. S. (1993). “Long-term response of mature plum trees to salinity.” Irrig. Sci., 13, 171–176. Chang, P. T., and Randle, W. M. (2005). “Sodium Cl timing and length of exposure affect onion growth and flavor.” J. Plant Nutr., 28, 1755–1766. Chapman, H. D. (1966). Diagnostic criteria for plants and soils, University of California Division of Agricultural and Natural Resources, Oakland, Calif. Chartzoulakis, K. S. (1992). “Effects of NaCl salinity on germination, growth and yield of greenhouse cucumber.” J. Hort. Sci., 67, 115–119. Cho, K. M., Toler, H., Lee, J., Ownley, B., Stutz, J. C., Moore, J. L., and Auge, R. M. (2006). “Mycorrhizal symbiosis and response of sorghum plants to combined drought and salinity stresses.” J. Plant Physiol., 163, 517–528. Colla, G., Rouphael, Y., Cardarelli, M., and Rea, E. (2006). “Effect of salinity on yield, fruit quality, leaf gas exchange and mineral composition of grafted watermelon plants.” Hort. Sci., 41, 622–627. Cooper, W. C. (1951). “Salt tolerance of avocado on various rootstocks.“ Texas Avocado Soc. Yearbook, 4, 24–28. ———. (1961). “Toxicity and accumulation of salts in citrus trees on various rootstocks in Texas.” Florida State Hort. Soc. Proc., 74, 95–104.

PLANT SALT TOLERANCE

451

Costello, L. R., Perry, E. J., Matheny, N. P., Henry, J. M., and Geisel, P. M. (2003). Abiotic disorders of landscape plants: A diagnostic guide, Publication 3420, University of California Division of Agricultural and Natural Resources. Cuartero, J., and Fernández-Muñoz, R. (1999). “Tomato and salinity.” Sci. Hort., 78, 83–125. De Pascale, Maggio, A., Fofliano, V., Ambrosino, P., and Ritieni, A. (2001). “Irrigation with saline water improves carotenoids content and antioxidant activity of tomato.” J. Hort. Sci. Biotech., 76, 447–453. Diaz, F. J., and Grattan, S. R. (2009). “Performance of tall wheatgrass (Thinopyrum ponticum, cv. ‘Jose’) irrigated with saline-high boron drainage water: Implications on ruminant mineral nutrition.” Agri. Ecosystems Environ., 131, 128–136. Drew, M. C., Guenther, J., and Läuchli. A. (1988). “The combined effects of salinity and root anoxia on growth and net Na and K accumulation in Zea mays grown in solution culture.” Ann. Bot., 61, 41–43. Eaton, F. M. (1944). “Deficiency, toxicity, and accumulation of boron in plants.” J. Agric. Res., 69, 237–277. Edelstein, M., Ben-Hur, M., Cohen, R., Burger, Y., and Ravina, I. (2005). “Boron and salinity effects on grafted and non-grafted melon plants.” Plant Soil, 269, 73–284. Ehlig, C. F. (1960). “Effects of salinity on four varieties of table grapes grown in sand culture.” Proc. Am. Soc. Hort. Sci., 76, 323–331. Farnham, D., Ayers, R. S., and Hasek, R. F. (1985). Water quality: Its effect on ornamental plants, Leaflet 2995, University of California Division of Agricultural and Natural Resources. Ferguson, L., Poss, J. A., Grattan, S. R., Grieve, C. M., Wang, D., Wilson, C., Donovan, T. J., and Chao, C.-T. (2002). “Pistachio rootstocks influence scion growth and ion relations under salinity and boron stress.” J. Am. Soc. Hort. Sci., 127, 194–199. Francois, L. E. (1984). “Effect of excess boron on tomato yield, fruit size, and vegetative growth.” J. Am. Soc. Hort. Sci., 109, 322–324. ———. (1986). “Effect of excess boron on broccoli, cauliflower and radish.” J. Am. Soc. Hort. Sci., 111, 484–498. ———. (1987). “Salinity effects on asparagus yield and vegetative growth.” J. Am. Soc. Hort. Sci., 112, 432–436. ———. (1988). “Yield and quality responses of celery and crisphead lettuce to excess boron.” J. Am. Soc. Hort. Sci., 113, 538–542. ———. (1989). “Boron tolerance of snapbean and cowpea.” J. Am. Soc. Hort. Sci., 114, 615–619. ———. (1991). “Yield and quality responses of garlic and onion to excess boron.” Hort. Sci., 26, 547–549. ———. (1992). “Effect of excess boron on summer and winter squash.” Plant Soil, 147, 163–170. ———. (1995). “Salinity effects on bud yield and vegetative growth of artichoke (Cynara scolymus L.).” Hort. Sci., 30, 69–71. Francois, L. E., and Clark, R. A. (1979). “Boron tolerance of twenty-five ornamental shrub species.” J. Am. Soc. Hort. Sci., 104, 319–322. Francois, L. E., Donovan, T. J., Lorenz, K., and Maas. E. V. (1989). “Salinity effects on rye grain yield, quality, vegetative growth, and emergence.” Agron. J., 81, 707–712.

452

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Francois, L. E., Maas, E. V., Donovan, T. J., and Youngs, V. L. (1986). “Effect of salinity on grain yield and quality, vegetative growth, and germination of semi-dwarf and durum wheat.” Agron. J., 78, 1053–1058. Francois, L. E., and West, D. W. (1982). “Reduction in yield and market quality of celery caused by soil salinity.” J. Am. Soc. Hort. Sci., 107, 952–954. Garcia-Sanchez, F., Perez-Perez, F. G., Botia, P., and Martinez, V. (2006). “The response of young mandarin trees grown under saline conditions depends on the rootstock.” Eur. J. Agron., 24, 129–139. Geraldson, C. M. (1957). “Factors affecting calcium nutrition of celery, tomato, and pepper.” Soil Sci. Soc. Proc., 21, 621–625. Gersberg, R. M., Elkins, B. V., Lyon, S. R., and Goldman, C. R. (1986). “Role of aquatic plants in wastewater treatment by artificial wetlands.” Water Res., 20, 363–368. Grattan, S. R., and Grieve, C. M. (1994). “Mineral nutrient acquisition and response by plants grown in saline environments,” in Handbook of plant and crop stress, M. Pessarakli, ed., Marcel Dekker Inc., New York, 203–226. ———. (1999). “Salinity-mineral nutrient relations in horticultural crops.” Sci. Hort., 78, 127–157. Grattan, S. R., Grieve, C. M., Poss, J. A., Robinson, P. H., Suarez, D. L., and Benes, S. E. (2004a). “Evaluation of salt-tolerant forages for sequential water reuse systems. I. Biomass production.” Agric. Water Mgmt., 70, 109–120. ———. (2004b). “Evaluation of salt-tolerant forages for sequential water reuse systems. III. Potential implications for ruminant nutrition.” Agric. Water Mgmt., 70, 137–150. Grattan, S. R., and Rhoades, J. D. (1990). “Irrigation with saline groundwater and drainage water,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 432–449. Grattan, S. R., Shannon, M. C., Grieve, C. M., Poss, J. A., Suarez, D. L., and Francois L. E. (1997). “Interactive effects of salinity and boron on the performance and water use of eucalyptus.” Acta Hort., 449, 607–613. Grattan, S. R., Zeng, L., Shannon, M. C., and Roberts, S. R. (2002). “Rice is more sensitive to salinity than previously thought.” Calif. Agric., 56, 189–195. Grieve, C. M. (2010). “Salinity-induced enhancement of horticultural crop quality,” in Handbook of plant and crop stress, M. Pessarakli, ed., CRC Press, Boca Raton, Fla., 1175–1196. Grieve, C. M., and Maas, E. V. (1988). “Differential effects of sodium:calcium ratio on sorghum genotypes.” Crop Sci., 29, 659–665. Grieve, C. M., and Poss, J. A. (2000). “Wheat response to interactive effects of boron and salinity.” J. Plant Nutr., 23, 1217–1226. Grieve, C. M., Poss, J. A., and Amrhein, C. (2006). “Response of Matthiola incana to irrigation with saline wastewaters.” Hort. Sci., 41, 119–123. Grieve, C. M., Poss, J. A., Grattan, S. R., Shouse, P. J., Lieth, J. H., and Zeng, L. (2005). “Productivity and mineral nutrition of Limonium species irrigated with saline wastewaters.” Hort. Sci., 40, 654–658. Grieve, C. M., Poss, J. A., Grattan, S. R., Suarez, D. L., and Smith, T. T. (2010). “The combined effects of salinity and excess boron on mineral ion relations in broccoli.” Sci. Hort, 125, 178–187.

PLANT SALT TOLERANCE

453

Grieve, C. M., Poss, J. A., Suarez, D. L., and Dierig, D. A. (2001). “Lesquerella growth and selenium uptake affected by saline irrigation water composition.” Ind. Crops Prod., 13, 57–65. Grieve, C. M., Suarez, D. L., and Shannon, M. C. (1999). “Effect of saline irrigation water composition on selenium accumulation by wheat.” J. Plant Nutr., 22, 1443–1450. Grieve, C. M., Wang, D., and Shannon, M. C. (2003). “Salinity and irrigation method affect mineral ion relations in soybean.” J. Plant Nutr., 26, 901–913. Groot Obbink, J., and Alexander, D. M. (1973). “Response of six grapevine cultivars to a range of Cl concentrations.” Am. J. Enol. Viticult., 24, 65–68. Gupta, U. C., Jaime, Y. W., Campbell, C. A., Leyshon, A. J., and Nicholaichuk, W. (1985). “Boron toxicity and deficiency: A review.” Can. J. Soil Sci., 65, 381–409. ˇ ˚ nek, J., Hopmans, J. W., and Hutmacher, R. B. Hanson, B. R., May, D. E., Simu (2009). “Drip irrigation provides the salinity control needed for profitable irrigation of tomatoes in the San Joaquin Valley.” Calif. Agric., 63(3), 131–136. Hanson, J. B. (1983). “The roles of calcium in plant growth.” Curr. Topics Plant Biochem. Physiol., 1, 1–24. Hayward, H. E., and Bernstein, L. (1958). “Plant-growth relationships on saltaffected soils.” Bot. Rev., 24, 584–635. Hayward, H. E., and Wadleigh, C. H. (1949). “Plant growth on saline and alkali soils.” Adv. Agron. 1, 1–38. Hirrel, M. C., and Gerdemann, J. W. (1980). “Improved growth of onion and bell pepper in saline soils by two vesicular-arbuscular mycorrhizal fungi.” Soil Sci. Soc. Am. J., 44, 654–655. Holloway, R. E., and Alston, A. M. (1992). “The effects of salt and boron on growth of wheat.” Aust. J. Agric. Res., 43, 987–1001. Ibekwe, A. M., Grieve, C. M., and Lyon, S. R. (2003). “Characterization of microbial communities and composition in constructed dairy wetland wastewater effluent.“ App. Environ. Microbiol., 69, 5060–5069. Ibekwe, A. M., Lyon, S. R., Leddy, M., and Jacobson-Meyers, M. (2007). “Impact of plant density and microbial composition on water quality from a free water surface constructed wetland.” J. Appl. Microbiol., 102, 921–936. Isidoro, D., and Aragüés, R. (2006). “Modeling survival of young olive trees (Olea europaea L. cv. Argequino) to saline and waterlogging field conditions.“ Agron. J., 98, 795–799. Janzen, H. H., and Chang, C. (1987). “Cation nutrition of barley as influenced by soil solution composition in a saline soil.” Can. J. Soil Sci., 67, 619–629. Johnson, R. S. (1996). “Stone fruit: Peaches and nectarines,” in Nutrient deficiencies and toxicities in crop plants, W. F. Bennett, ed., The American Phytopathology Society, St. Paul, Minn., 171–175. Keren, R., and Bingham, F. T. (1985). “Boron in water, soils, and plants.” Adv. Soil Sci., Vol. I, R. Stuart, ed., Springer-Verlag, Inc., New York, 229–276. Koyro, H.-W. (2003). “Study of potential cash crop halophytes by a quick check system: Determination of the threshold of salinity tolerance and the ecophysiological demands,” in Cash crop halophytes: Recent studies, H. Lieth and M. Mochtchenko, eds., Kluwer Academic Publishers, Norwell, Mass., 5–17. Krauss, S., Schnitzer, W. H., Grassman, J., and Woitke, M. (2006). “The influence of different electrical conductivity values in simplified recirculating soilless

454

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

system on inner and outer fruit quality characteristics of tomato.” J. Agric. Food Chem., 54, 441–448. Läuchli, A. (1984). “Salt exclusion: An adaptation of legumes for crops and pastures under saline conditions,” in Salinity tolerance in plants: Strategies for crop improvement, R. C. Staples and G. H. Toenniessen, eds., John Wiley and Sons, Inc., New York, 171–187. ———. (1993). “Selenium in plants: Uptake, functions, and environmental toxicity.” Bot. Acta, 106, 455–468. Levin, A. G., Lavee, S., and Tsror (Lahkim), L. (2007). “The influence of salinity on Verticillium dahliae in stem cuttings of five olive cultivars.” J. Phytopath., 155, 587–592. Lieth, J. H., and Burger. D. W. (1989). “Growth of chrysanthemum using an irrigation system controlled by soil moisture tension.” J. Am. Soc. Hort. Sci., 114, 387–392. Lieth, H., and Mochtchenko, M. (2003). Cash crop halophytes: Recent studies, Kluwer Academic Publishers, Boston. Lynch, J., and Läuchli, A. (1985). “Salt stress disturbs the calcium nutrition in barley (Hordeum vulgare L.).” New Phytol., 99, 345–354. Maas, E. V. (1986). “Salt tolerance of plants.” Appl. Agric. Res., 1, 12–36. ———. (1993). “Salinity and citriculture.” Tree Physiol., 12, 195–219. Maas, E. V., and Grattan, S. R. (1999). Crop yields as affected by salinity, agricultural drainage, Agronomy Monograph No. 38, R. W. Skaggs and J. van Schilfgaarde, eds., ASA/CSSA/SSSA, Madison, Wisc., 55–108. Maas, E. V., and Grieve, C. M. (1987). “Sodium-induced calcium deficiency in saltstressed corn.” Plant Cell Environ., 10, 559–564. ———. (1994). “Salt tolerance of plants at different growth stages,” in Proc., Int. Conf. Current Developments in Salinity and Drought Tolerance of Plants. January 7–11, 1990, Tando Jam, Pakistan, 181–197. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. Drain. Div. ASCE, 103, 115–134. MacDonald, J. D. (1982). “Effect of salinity stress on the development of phytophthora root rot of chrysanthemum.” Phytopathology, 72, 214–219. Marcar, N. E., Guo, J., and Crawford, D. F. (1999). “Response of Eucalyptus camaldulensis Dehnh., E. globulus Lavill. ssp globulus and E. grandis W. Hill to excess boron and sodium chloride.” Plant Soil, 208, 251–257. Meiri, A. (1984). “Plant response to salinity. Experimental methodology and application to the field,” in Soil salinity under irrigation, I. Shainberg and J. Shalhevet, eds., Springer-Verlag Inc., New York, 284–297. Meiri, A., Shalhevet, J., Shimshi, D., and Tibor, M. (1982). “Irrigation of spring potatoes with saline water,” in Annual report, Agricultural Research Organization, Institute of Soils and Water, Volcani Center, Bet Dagan, Israel. Mizrahi, Y., and Pasternak, D. (1985). “Effect of salinity on quality of various agricultural crops.” Plant Soil, 89, 301–307. Mohammadi, A. H., Banihashemi, Z., and Maftoun, M. (2007). “Interaction between salinity stress and verticillium wilt disease in three pistachio rootstocks in a calcareous soil.” J. Plant Nutr., 30, 241–252. Munns, R., and Termaat, A. (1986). “Whole-plant responses to salinity.” Aust. J. Plant Physiol., 13,143–160.

PLANT SALT TOLERANCE

455

Nable, R. O., Bañuelos, G. S., and Paul, J. G., (1997). “Boron toxicity.” Plant Soil, 193, 181–198. Navarro, J. M., Botella, M. A., and Martinez, V. (1999). “Yield and fruit quality of melon plants grown under saline conditions in relation to phosphate and calcium nutrition.” J. Hort. Sci. Biotech., 74, 573–578. Nicholaichuk, W., Leyshon, A. J., Jaime, Y. W., and. Campbell, C. A. (1988). “Boron and salinity survey of irrigation projects and the boron adsorption characteristics of some Saskatchewan soils.” Can. J. Soil Sci., 68, 77–90. Oertli, J. J., and Kohl. H. C. (1961). “Some considerations about the tolerance of various plant species to excessive supplies of boron.” Soil Sci., 92, 243–247. Ojala, J. C., Jarrell, W. M., Menge, J. A., and Johnson, E. L. V. (1983). “Influence of mycorrhizal fungi on the mineral nutrition and yield of onion in saline soil.” Agron. J., 75, 255–259. Oster, J. D., Hoffman, G. J., and Robinson F. E. (1984). “Management alternatives: Crop, water, and soil.” Calif. Agric., 38, 29–32. Page, A. L., Chang, A. C., and Adriano, D. C. (1990). “Deficiencies and toxicities of trace elements,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 138–160. Parker, M. B., Gascho, G. J., and Gaines, T. P. (1983). “Chloride toxicity of soybeans grown on Atlantic coast flatwoods soils.” Agron. J., 75, 439–442. Parra, M. A., and Romero, G. C. (1980). “On the dependence of salt tolerance of beans (Phaseolus vulgaris L.) on soil water matric potential.” Plant Soil, 56, 3–16. Pasternak, D., and De Malach, Y. (1994). “Crop irrigation with saline water,” in Handbook of plant and crop stress, M. Pessarakli, ed., Marcel Dekker, Inc., New York, 599–622. Pasternak, D., Twersky, M., and De Malach, Y. (1979). “Salt resistance in agricultural crops,” in Stress physiology in crop plants, H. Mussel and R. S. Staples, eds., John Wiley and Sons, New York, 128–152. Pearson, G. A. (1959). “Factors influencing salinity of submerged soils and growth of caloro rice.” Soil Sci., 87, 198–206. Poss, J. A., Pond, E., Menge, J. A., and Jarrell, W. M. (1985). “Effect of salinity on mycorrhizal onion and tomato in soil with and without additional phosphate.” Plant Soil, 88, 307–319. Pratt, P. F., and Suarez, D. L. (1990). “Irrigation water quality assessments,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 220–236. Ravikovitch, S., and Yoles, D. (1971). “The influence of phosphorus and nitrogen on millet and clover growing in soils affected by salinity. I. Plant development.” Plant Soil, 35, 555–567. Ray, A. M., and Inouye, R. S. (2007). “Development of vegetation in a constructed wetland receiving irrigation return flows.” Agric. Ecosys. Environ., 121, 401–406. Rhoades, J. D., Bingham, F. T., Letey, J., Dedrick, A. R. (1988). “Reuse of drainage water for irrigation: Results of Imperial Valley Study. I. Hypothesis, experimental procedures and cropping results.” Hilgardia, 56, 1–16. Rhoades, J D., Bingham, F. T., Letey, J., Hoffman, G. J., Dedrick, A. R., Pinter, P. J., and Replogle, J. A. (1989). “Use of saline drainage water for irrigation: Imperial Valley study.” Agric. Water Mgmt., 16, 25–36.

456

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Ruiz-Lozano, J. M., Azcon, R., and Gomez, M. (1996). “Alleviation of salt stress by arbuscular-mycorrhizal Glomus species in Lactuca sativa plants.” Physiol. Plant, 98, 767–772. Salisbury. F. B. (1995). Units, symbols, and terminology for plant physiology, Utah State University, Logan, Utah. Sannazzaro, A. I., Ruiz, O. A., Alberto, E. O., and Menendez, A. B. (2006). “Alleviation of salt stress in Lotus glaber by Glomus intraradices.” Plant Soil, 285, 279–287. Sanogo, S. (2004). “Response of chili pepper to Phytophthora capsici in relation to soil salinity.” Plant Dis., 88, 205–209. Sarooshi, R. A., and Cresswell, G. C. (1994). “Effects of hydroponic solution composition, electrical conductivity and plant spacing on yield and quality of strawberries.” Aust. J. Exp. Agric., 34, 529–535. Sauer, M. R. (1968). “Effects of vine rootstocks on Cl concentration in Sultana scions.” Vitis, 7, 223–226. Shalhevet, J. (1984). “Management of irrigation with brackish water,” in Soil salinity under irrigation: Processes and management, I. Shainberg and J. Shalhevet, eds., Springer-Verlag, New York, 298–318. Shalhevet, J., Heuer, B., and Meiri, A. (1982). “Irrigation interval as a factor in the salt tolerance of eggplant.” Irrig. Sci., 4, 83–93. Shalhevet, J., Vinten, A., and Meiri, A. (1986). “Irrigation interval as a factor in sweet corn response to salinity.” Agron. J., 78, 539–545. Shannon, M. C., and Francois, L. E. (1978). “Salt tolerance of three muskmelon cultivars.” J. Am. Soc. Hort. Sci., 103, 127–130. Sharpley, A. N., Meisinger, J. J., Power, J. F., and Suarez, D. L. (1992). “Root extraction of nutrients associated with long-term soil management,” in Advances in soil science. Limitations to plant root growth, Vol. 19, J. I. Hatfield and B. A. Stewart, eds., Springer-Verlag, New York, 151–217. Shear, C. B. (1975). “Calcium-related disorders in fruits and vegetables.” Hort. Sci., 10, 361–365. Shennan, C., Grattan, S. R., May, D. N., Hillhouse, C. J.,. Schachtman, D. P., Wander, M., Roberts, B., Tafoya, S., Burau, R. G., McNeish, C., and Zelinski, L. (1995). “Feasibility of cyclic reuse of saline drainage in a tomato-cotton rotation.” J. Environ. Qual., 24, 476–486. Shillo, R., Ding, M., Pasternak, D., and Zaccai, M. (2002). “Cultivation of cut flower and bulb species with saline water.” Sci. Hort., 92, 41–54. Sifola, M. I., De Pascale, S., and Romano, R. (1995). “Analysis of quality parameters in eggplant grown under saline water irrigation.” Acta Hort., 412, 176–184. Skaggs, R. W., and van Schilfgaarde, J. (1999). Agricultural drainage, ASA Monograph 38, ASA-CSA-SSSA, Madison, Wisc. Skaggs, T. H., Poss, J. A., Shouse, P. J., and Grieve, C. M. (2006a). “Irrigating forage crops with saline waters. I. Volumetric lysimeter studies.” Vadose Zone J., 5, 815–823. Skaggs, T. H., Shouse, P. J., and Poss, J. A. (2006b). “Irrigating forage crops with saline drainage waters. II. Modeling drainage and root water uptake.” Vadose Zone J., 5, 824–837.

PLANT SALT TOLERANCE

457

Smith, T. E, Grattan, S. R., Grieve, C. M., Poss, J. A., and Suarez, D. L. (2010a). “Salinity’s influence on boron toxicity in broccoli. I: Impacts on yield, biomass distribution, and water use.” Agric. Water Mgmt., 97, 777–782. ———. (2010b). “Salinity’s influence on boron toxicity in broccoli. II: Impacts on boron uptake, uptake mechanisms and tissue ion relations.” Agric. Water Mgmt., 97, 783–791. Snapp, S. S., and Shennan, C. (1992). “Effects of salinity on root growth and death dynamics of tomato, Lycopersicon esculentum Mill.” New Phytol., 121, 71–79. Snapp, S. S., Shennan, C., and van Bruggen, A. H. C. (1991). “Effects of salinity on severity of infection by Phytophthora parasitica Dast. ion concentrations and growth of tomato, Lycopersicon esculentum Mill.” New Phytol., 119, 275–284. Sonneveld, C. (1988). “The salt tolerance of greenhouse crops.” Neth. J. Agric. Sci., 36, 63–73. Steppuhn, H., van Genuchten, M. T., and Grieve, C. M. (2005a). “Root-zone salinity: I. Selecting a product yield index and response function for crop salt tolerance.” Crop Sci., 45, 209–220. ———. (2005b). “Root-zone salinity: II. Indices for tolerance in agricultural crops.” Crop. Sci., 45, 221–232. Steppuhn, D., Wall, K. G., and P. G. Jefferson. (2009). “Seed size effects on emergence, height growth, and shoot biomass of first-year alfalfa grown in saline media.” Trans. Agri. Biol. Eng., 52, 121–132. Suarez, D. L., and Grieve, C. M. (1988). “Predicting cation ratios in corn from saline solution composition.” J. Exp. Bot., 39, 605–612. Suarez, D. L., Grieve, C. M., and Poss, J. A. (2003). “Irrigation method affects selenium accumulation in forage Brassica species.” J. Plant Nutr., 26(1), 191–201. Supanjani, K. D. L. (2006). “Hot pepper response to interactive effects of salinity and boron.” Plant Soil Environ., 52, 227–233. Tanji, K. K. (2007). “Salt management guide. Guide for landscape irrigation with recycled water in coastal southern California,” CD database available from California Department of Water Resources, www.water.ca.gov, accessed February 12, 2011. Trajkova, F., Papadantonakis, N., and Savvas, D. (2006). “Comparative effects of NaCl and CaCl2 on cucumber grown in a closed hydroponic system.” Hort. Sci., 41, 437–441. Triky-Dotan, S., Yermiyahu, U., Katan, J, .and Gamliel, A. (2005). “Development of crown and root rot disease of tomato under irrigation with saline water.” Phytopathology, 95, 1438–1444. Tripler, E., Ben-Gal, A., and Shani, U. (2007). “Consequence of salinity and excess boron on growth, evapotranspiration and ion uptake in date palm (Phoenix dactylifera L., cv. Medjool).” Plant Soil, 297, 147–155. Ungar, I. A. (1998). “Are biotic factors significant in influencing the distribution of halophytes in saline habitats?” Bot. Rev., 64, 176–199. U.S. Salinity Laboratory Staff (USSL). (1954). Diagnosis and improvement of saline and alkaline soils, USDA Agricultural Handbook 60, USDA-ARS, Washington, D.C.

458

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

van Genuchten, M. T. (1983). Analyzing crop salt tolerance data. Model description and user’s manual, USDA-ARS/U.S. Salinity Laboratory Research Report No. 120, U.S. Government Printing Office, Washington, D.C. van Genuchten, M. T., and Gupta, S. K. (1993). “A reassessment of the crop tolerance response function.” J. Indian Soc. Soil Sci., 41, 730–736. van Genuchten, M. T., and Hoffman, G. J. (1984). “Analysis of crop salt tolerance data,” in Soil salinity under irrigation: Processes and management: Ecology studies 51, I. Shainberg and J. Shalhevet, eds., Springer-Verlag, New York, 258–271. van Schilfgaarde, J. (1990). “Irrigated agriculture: Is it sustainable?” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports of Engineering Practice No. 71, ASCE, Reston, Va., 584–595. Wadleigh, C. H., and Ayers, A. D. (1945). “Growth and biochemical composition of bean plants as conditioned by soil moisture tension and salt concentration.” Plant Physiol., 20, 106–132. Wang, D. (2002). “Dynamics of soil water and temperatures in above ground sand cultures used for screening plant salt tolerance.” SSSA J., 66, 1484–1491. Wang, D., Grieve, C. M., and Suarez, D. L. (2005). “Composition of irrigation water salinity affects growth characteristics and uptake of selenium and salt ions by soybean.” J. Plant Nutr., 28, 1073–1088. Wang, D., Shannon, M. C., Grieve, C. M., Shouse, P. J., and Suarez, D. L. (2002). “Ion partitioning among soil and plant components under drip, furrow, and sprinkler irrigation regimes: Field and modeling assessments.” J. Environ. Qual., 31, 1684–1693. Wimmer, M. A., Mühling, K. H., Läuchli, A., Brown, P. H., and Goldbach, H. E. (2003). “The interaction between salinity and boron toxicity affects the subcellular distribution of ions and proteins in wheat leaves.” Plant Cell Environ., 26, 1267–1274. Wutscher, H. K., and Smith, P. F. (1996). “Citrus,” in Nutrient deficiencies and toxicities in crop plants, W. F. Bennett, ed., American Phytopathology Society, St. Paul, Minn., 165–170. Yadav, H. D., Yadav, O. P., Dhankar, O. P., and Oswald, M. C. (1989). “Effect of Cl salinity and boron on germination, growth, and mineral composition of chickpea (Cicer arietinum L.).” Ann. Arid Zone, 28, 63–67. Yano-Melo, A. M., Saggin, O. J., and Mala, L. C. (2003). “Tolerance of mycorrhizal banana (Musa sp. cv. Pacovan) plantlets to saline stress.” Agric. Ecosys. Environ., 95, 343–348. Yermiyahu, U., Ben-Gal, A., Keren, R., and Reid, R. J. (2008). “Combined effect of salinity and excess boron on plant growth and yield.” Plant Soil, 304, 73–87. Yermiyahu, U., Ben-Gal, A., and Sarig, P. (2006). “Boron toxicity in grapevine.” Hort. Sci., 41, 1698–1703. Yermiyahu, U., Ben-Gal, A., Sarig, P., and Zipilevitch, E. (2007). “Boron toxicity in grapevine (Vitis vinifera L.) in conjunction with salinity and rootstock effects.” J. Hort. Sci. Biotech., 82, 547–554. Zekri, M., and Parsons, L. R. (1990). “Comparative effects of NaCl and polyethylene glycol on root distribution, growth, and stomatal conductance of sour orange seedlings.” Plant Soil, 129, 137–144.

PLANT SALT TOLERANCE

459

NOTATION C50  soil salinity where crop yields are reduced 50% EC  electrical conductivity of the irrigation water (ECi), the soil water (ECsw), and the saturated soil extract (ECe) ECt  yield threshold soil salinity (ECe) value (also referred to as the yield threshold “A” coefficient in other chapters) above which yields decline p  coefficient that reflects the curve steepness and ST-index is salttolerance index

This page intentionally left blank

CHAPTER 14 STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE AND SPATIAL SALINITY PATTERNS FROM SOIL CONDUCTIVITY SURVEY DATA S. M. Lesch

INTRODUCTION The collection of apparent soil electrical conductivity (ECa) survey data for the purpose of characterizing various spatially referenced soil properties has received considerable attention in the soils literature over the last two decades (Corwin and Lesch 2005a,b). Although now commonly used in many precision agriculture survey applications, most of the original interest in ECa survey data was motivated by the need to characterize and map soil salinity in a cost-effective manner (Rhoades et al. 1999; Hendrickx et al. 2002). The need for such surveying work is expected to increase over time, as more agricultural land becomes degraded due to salinization. Apparent soil conductivity survey data often correlate reasonably well with various soil properties (salinity, soil texture, soil water content, etc.) under different field conditions (Corwin and Lesch 2005b; Lesch and Corwin 2003). However, as a general rule, ECa survey readings tend to be strongly correlated with soil salinity levels. Thus, accurate salinity predictions can normally be constructed from ECa survey data in semi-saline and saline fields using fairly simple statistical calibration techniques. Additionally, accurate maps of the field-scale salinity pattern can sometimes also be produced from ECa survey data in marginally saline fields, provided that other important soil properties (such as soil texture and soil water content) exhibit fairly minimal spatial variation.

461

462

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

After ECa survey data have been acquired in a field, calibration soil samples are normally collected at a certain number of ECa survey locations. The measured salinity levels associated with these soil samples are then used (in conjunction with the co-located survey data) to estimate some type of spatial-statistical or geostatistical model. This statistical model is in turn used to predict the detailed spatial soil-salinity pattern from the full set of acquired survey data. This chapter discusses the simplest and most frequently used statistical modeling approach for calibrating ECa survey information with measured salinity data, such as ordinary regression. Ordinary linear regression models represent a special case of a much more general class of models commonly known as linear regression models with spatially correlated errors (Schabenberger and Gotway 2005), hierarchical spatial models (Banerjee et al. 2004), or geostatistical mixed linear models (Haskard et al. 2007). This broader class of models includes many of the geostatistical techniques familiar to soil scientists, such as universal kriging and kriging with external drift, as well as standard regression techniques—ordinary linear regression (LR) models and analysis of covariance models. The remainder of this chapter is organized as follows. A technical review of the basic linear regression model estimation and validation techniques is presented in “Regression Models” and “Regression Model Validation Tests.” Some suitable sampling strategies for calibrating linear regression equations are discussed in “Sampling Strategies,” while the subsequent section presents a brief overview of the ESAP software package. Two salinity assessment examples are then presented in “ Data Analysis Examples”; these data analyses demonstrate the statistical calibration and prediction techniques discussed in this chapter, along with some of the types of analysis output that the ESAP software can produce. Regression Models: Estimation and Prediction Formulas Site-specific prediction of diverse soil properties from EM survey data can be achieved using regression model estimation and prediction techniques. In the regression modeling approach advocated by Lesch and Corwin (2008), Lesch (2005), Rhoades et al. (1999), and Lesch et al. (1995), a suitable linear equation is specified that relates the target soil property of interest to a linear combination of conductivity signal data readings and (possibly) trend surface coordinates. One example of such an equation would be y1  0  1[EMV,i]  2[EMH,i]  3[cx,i]  4[cy,i]  εi

(14-1)

where the response variable (y) represents the soil property of interest (e.g., salinity, texture, water content) at the ith survey location; the predic-

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

463

tor variables (EMV, EMH, cx, cy) represent the corresponding EM38 vertical and horizontal signal readings and associated ith survey site coordinate locations, respectively; the  parameters represent empirical regression model coefficients; and ε represents the random error component associated with the model. Equation 14-1 relates the response variable (e.g., soil property) to both EM signal and trend surface components, and thus can be viewed as a “signal  trend” model. The trend surface components specified in Eq. 14-1 are optional and should only be included if they are found to be necessary (i.e., if the associated parameter estimates are statistically significant or if the inclusion of such components is needed to address an obvious spatial trend in a residual plot). The optimal estimation of the aforementioned (or similar) regression model depends on the assumptions placed on the random error component. If the errors are assumed to be normally distributed and exhibit spatial correlation, then Eq. 14-1 is commonly called a spatial linear regression model in the statistical literature, or a kriging with external drift model in the geostatistical literature (Cressie 1991; Schabenberger and Gotway 2005). Such models can be efficiently estimated using maximum likelihood or restricted maximum likelihood fitting techniques (Littell et al. 1996). In contrast, if the errors can be assumed to be approximately uncorrelated, then ordinary least squares (OLS) fitting techniques can be used. In this latter case the model becomes identical to an ordinary linear regression equation, the only difference being that the predictions are spatially referenced. The likelihood of the residual errors being approximately uncorrelated (as opposed to spatially correlated) depends primarily on (1) the method used to select the calibration sample sites, and (2) the degree to which the conductivity signal data correlates with the response variable of interest. When the signal data is strongly correlated with the target soil property and specialized sampling strategies are employed, the assumption of approximate residual independence is often satisfied. For detailed discussions concerning these issues, see Lesch and Corwin (2008) and Lesch (2005). Although appropriate prediction statistics can be derived for either case, only the OLS results are presented here. Additionally, all of the following results are presented in matrix notation; a good review of matrix notation from a regression modeling viewpoint is given in Myers (1986). Following standard matrix notation, note that we can express Eq. 14-1 as y  X  e, where y represents a (n  1) vector of soil property measurements (collected across n sites), X represents the corresponding (n  p) regression model design matrix and e represents the (n  1) vector of residual errors. Then, under the uncorrelated residual error assumption, the best linear unbiased estimate (BLUE) for  is ˆ  (X T X )1X T y

(14-2)

464

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

with a corresponding variance of Var(ˆ )   2 (X T X)1

(14-3)

where 2 represents the regression model mean square error (MSE) component. Likewise, the residuals (i.e., empirical model errors) for Eq. 14-1 are defined to be r  y  Xˆ

(14-4)

and these residuals provide an unbiased estimate of the MSE component, that is ˆ 2  (r T r )  (n  p). Now, let yz represent the (unknown) vector of soil property values at all of the remaining survey locations and define Xz to be the corresponding design matrix associates with these sites. Then, again under the uncorrelated residual error assumption, the best linear unbiased prediction (BLUP) of these soil property values can be shown to be yˆ z  X zˆ

(14-5)

with a corresponding variance estimate of Var( y z  yˆ z )   2 (I  X z (X T X )1X Tz ).

(14-6)

Corresponding formulas for both individual and field average prediction estimates can also be immediately derived from standard linear modeling theory. For example, individual survey site predictions (and their corresponding variance estimates) become yˆ 0  x zˆ Var { y 0  yˆ 0 }   2 (1  x z (X T X )1x Tz )

(14-7)

where xz represents the (1  p) design vector associated with a specific prediction site. Likewise, the average prediction associated with the entire nonsampled survey grid can be computed as

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

yˆ ave  x aveˆ Var { y ave  yˆ ave }   2 (1  ( N  n)  x ave (X T X )

465

(14-8)

where xave represents the average of the N  n design vectors associated with the nonsampled survey positions. Note that all of these results are exactly identical to ordinary linear regression model parameter estimation and prediction formulas presented in standard regression model textbooks (Myers 1986). In many practical survey applications, determining the probability that a new prediction exceeds some specific threshold value is also of interest. Although not commonly discussed in most classical linear modeling textbooks, regression models can also be used to produce such probability estimates. More specifically, upon adopting a Bayesian perspective, the probability that an unobserved y0 lies within the interval (a, b) can be computed as h

i [ a , b]  Prob( a ≤ y 0 ≤ b)  ∫g t( np ) dt

(14-9)

where t(np) represents a central t-distribution having n  p degrees of freedom (i.e., the regression model residual degrees of freedom), g  ( a  yˆ 0 )  Var { yˆ 0 } , and h  (b  yˆ 0 )  Var { yˆ 0 } (Press 1989: assuming vague prior distributions on the model parameters). These latter probability predictions can in turn be used to calculate a range interval estimate (RIE) defined as RIE[ a , b] 

100 Nn ∑ i [ a , b] N  n i1

(14-10)

which represents a prediction of the percentage of nonsampled sites (on the survey grid) that exhibit soil property values falling within the interval (a, b). For example, one might be interested in predicting the percentage of survey sites in a field with salinity levels in excess of 4 dS/m. Equations 14-9 and 14-10 can be used to calculate this value, while simultaneously adjusting out the “shrinkage-effect” inherent in the associated regression model predictions. Lesch et al. (2005) discuss the above estimates in more detail and show multiple examples of their application.

466

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Regression Model Validation Tests If an ordinary linear regression model is to be successfully used in place of the geostatistical or spatial linear model, then more-restrictive modeling assumptions need to be met. In addition to the assumption of a normally distributed error process, the critical assumption in the linear regression model is the uncorrelated residual assumption. A formal test for spatial correlation in the residual pattern can be carried out using either a nested likelihood ratio test or via the Moran residual test statistic (Upton and Fingleton 1985; Haining 1990; Tiefelsdorf 2000; Schabenberger and Gotway 2005). The likelihood ratio test can only be performed after first estimating a suitable geostatistical or spatial linear model [see pages 343–344 of Schabenberger and Gotway (2005) for more discussion of this topic]. In contrast, the Moran test can be carried out directly on the ordinary regression model residuals. As originally introduced by Brandsma and Ketellapper (1979), the Moran test statistic was designed to detect spatially correlated residuals in conditionally and/or simultaneously specified spatial autoregressive models (Schabenberger and Gotway 2005). However, it can also be used to assess the uncorrelated residual assumption in a general linear modeling framework. The Moran residual test statistic (M) is defined as M 

r T Wr rT r

(14-11)

where r is defined in Eq. 14-4, W represents a suitably specified proximity matrix, and ˆ is calculated using Eq. 14-2. While the specification of W can be application-specific, in most soil survey applications it is generally reasonable to specify W as a scaled inverse distance squared matrix. Under such a specification, where dij represents the computed distance between the ith and jth sample locations, the {wij} elements associated with the ith row of the W matrix are defined as wii  0 and wij  dij2

n

∑ dij2,

(14-12)

j1

respectively. Brandsma and Ketellapper (1979) describe how to calculate the first two moments of M, i.e., E(M) and Var(M) [see also Lesch and Corwin (2008) and Lesch (2005)]. The corresponding Moran test score can then be computed as zM  ( M  E( M ))  Var( M )

(14-13)

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

467

and compared to the upper (one-sided) cumulative standard normal probability density function. A test score in excess of 1.65 ( ⬇ 0.05) is normally interpreted as being statistically significant. Provided that the regression model has been correctly specified, such a test score implies that the model residuals exhibit significant spatial correlation. In this situation, the parameter estimates and survey predictions may be highly inefficient and the mean square error estimate and parameter test statistics may be substantially biased. If sufficient data are available (or additional data can be collected), then a suitable spatial or geostatistical linear modeling approach should instead be employed. In addition to the uncorrelated residual assumption, one must also verify that the model residuals satisfy the usual standard normal error assumption and that the hypothesized model is correctly specified. Fortunately, most well-known residual analysis techniques used in an ordinary regression analysis are just as useful when applied to a spatially referenced linear regression model. These include assessing the assumption of residual normality using quantile (QQ) plots and the Shapiro-Wilk test (Shapiro and Wilk 1965), detecting outliers and/or high leverage points (plots of internally or externally studentized residuals), and detecting model specification bias (residual versus prediction plots, partial regression leverage plots, influence plots, etc.). The standard jack-knifing techniques commonly used to assess the predictive capability of an ordinary regression model are also directly applicable. Most standard statistical software packages can readily produce jack-knifed residual and/or prediction estimates in a computationally efficient manner. Cook and Weisberg (1999) and Myers (1986) offer good reviews of many relevant regression model diagnostic and assessment techniques. Sampling Strategies for Spatially Referenced Linear Regression Models Space limitations preclude a detailed discussion of the numerous sampling strategies one can employ to estimate spatially referenced regression models. Broadly speaking, the most common strategies currently employed can be classified as either (1) probability-based (design-based) sampling, (2) prediction-based (model-based) sampling, and (3) grid sampling. Brief descriptions of each of these approaches are given here. In general, probability-based sampling strategies tend to be commonly employed in most spatial research problems. Probability-based sampling strategies include techniques, such as simple random sampling, stratified random sampling, cluster sampling, capture-recapture techniques, and line transect sampling. Thompson (1992) provides a good review of multiple types of probability-based sampling strategies.

468

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Probability-based sampling strategies have a well-developed underlying theory and are clearly useful in many spatial applications (Thompson 1992; Brus and de Gruijter 1993). However, they are not designed specifically for estimating models per se. Indeed, most probability-based sampling strategies explicitly avoid incorporating any parametric modeling assumptions, relying instead upon randomization principles (which are built into the design) for drawing statistical inference. Prediction-based sampling strategies represent an alternative approach for developing sampling designs that are explicitly focused toward model estimation. The underlying theory behind this approach for finite population sampling and inference is discussed in detail in Valliant et al. (2000). More generally, response surface design theory and optimal experimental design theory represent two closely related statistical research areas that also study sampling designs specifically from the model estimation viewpoint (Atkinson and Donev 1992; Myers and Montgomery 2002). Techniques from these latter two subject areas have been applied to the optimal collection of spatial data by Müller (2001), the specification of optimal designs for variogram estimation by Müller and Zimmerman (1999), the estimation of spatially referenced regression models by Lesch (2005) and Lesch et al. (1995), and the estimation of geostatistical linear models by Brus and Heuvelink (2007), Minasny et al. (2007), and Zhu and Stein (2006). Conceptually similar types of nonrandom sampling designs for variogram estimation have been introduced by Bogaert and Russo (1999), Warrick and Myers (1987), and Russo (1984). Grid sampling represents another form of nonrandom sampling that has been used for many years in the soil sciences. Grid sampling has historically been recommended for accurately mapping soil boundaries and/or as a precursor to an ordinary kriging analysis (Burgess et al. 1981; Burgess and Webster 1984). Theoretically, any of these sampling approaches can be used for the purposes of estimating a regression model, although each approach exhibits various strengths and weaknesses. Lesch (2005) compares and contrasts probability-based and prediction-based sampling strategies in more detail, and highlights some of the strengths of the prediction-based sampling approach. Overview of the ESAP Software Package Many types of diverse software programs can be utilized for the assessment and quantification of soil salinity inventory information via soil conductivity survey data. The more common types of software applications include spatial mapping software, GIS software, statistical software, and geophysical software (when appropriate). Nonetheless, the need for a stand-alone, comprehensive salinity assessment software package was recognized some years ago by the technical staff at the U.S. Salinity Labo-

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

469

ratory. The ESAP software package was specifically developed to handle field scale salinity inventorying and assessment work, primarily in response to this need (Lesch et al. 2000). The ESAP software package contains a series of integrated, comprehensive software programs, designed for the Windows XP (or equivalent) operating system. This software can be used for the prediction of fieldscale, spatial soil salinity information from conductivity survey data and has specifically been designed to facilitate the use of cost-effective, technically sound soil salinity assessment and data interpretation techniques. The current publicly available shareware version of ESAP (version 2.35) contains three data processing programs designed to guide the analyst through the entire survey process: ESAP-RSSD, ESAP-SaltMapper, and ESAP-Calibrate. The ESAP-RSSD program can be used to generate optimal model-based soil sampling designs from conductivity survey data. The ESAP-SaltMapper program may be used to generate 1-D transect plots and/or 2-D raster maps of either raw soil conductivity or predicted soil salinity data. Additionally, the SaltMapper software can be used to identify and locate tile line positions in fields suffering drainage-related salinity problems. The ESAP-Calibrate program is normally used to convert raw conductivity data into estimated soil salinity data, via either statistical or deterministic calibration modeling techniques. However, this latter program can also be used to estimate other soil properties from conductivity survey data and/or analyze various soil property/conductivity relationships. ESAP version 2.35 contains two additional utility programs: ESAPSigDPA and the DPPC-Calculator. The SigDPA program can be used to perform various conductivity data preprocessing chores, such as screening out negative conductivity readings and/or assigning row numbers to transect conductivity survey data. The DPPC-Calculator can be used to convert insertion four-probe readings into calculated soil salinity levels (in conjunction with measured or estimated soil temperature, texture, and water content information). The ESAP-RSSD and ESAP-Calibrate programs contain the bulk of the model-based sampling and statistical modeling algorithms within the ESAP software package. As discussed, the ESAP software package represents an integrated, self-contained salinity assessment software system. All of the data analysis examples presented in the next section were performed using the version 2.35 ESAP software components (i.e., RSSD, Calibrate, and SaltMapper). Data Analysis Examples Example 1: A survey of a marginally saline lettuce field in Indio, California An electromagnetic induction (EMI) survey was performed by the Coachella Valley Resource Conservation District in June 2003 within a

470

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

14-ha lettuce field located in Indio, California. The primary goals of this survey were threefold: (1) to construct an accurate soil salinity inventory for the field, (2) to determine if this field should be leached before the fall cropping season, and (3) to construct relevant yield-loss projections, based on the predicted field soil salinity conditions. A total of 2,040 Geonics EM38 vertical (EMV, mS/m) and horizontal (EMH, mS/m) signal readings were collected across 29 north-south survey transects within this field and then processed through the USDA-ARS ESAP software package. This software selected 12 survey locations for soil sampling, using a prediction-based ESAP sample design (Lesch et al. 2000). Soil samples were collected from 0 to 0.6 m and 0.6 m to 1.2 m depths and analyzed for soil salinity (ECe, dS/m), soil saturation percentage (SP, %), and gravimetric water content (g, %). Table 14-1 lists the univariate summary statistics (mean, standard deviation, minimum, and maximum) for the EM38 survey and soil sample data, respectively. Figure 14-1 shows the interpolated EMV signal map for this field, along with the spatial positions of the 12 sampling locations. Note also that some advanced statistical aspects concerning this specific data analysis are discussed in Lesch and Corwin (2008). The results from an exploratory regression modeling analysis performed in ESAP suggested that the following natural log(ECe)/log(EM)

TABLE 14-1. Basic EM38 and Soil Sample Summary Statistics: Indio Lettuce Field Variable

EMV EMH

Units

N

Mean

Std. Dev.

Min

Max

mS/m mS/m

2040 2040

63.67 38.02

13.87 10.28

36.25 17.63

119.25 81.75

Variable

Units

ECe

dS/m

SP

%

g

%

Depth

N

Mean

Std. Dev.

Min

Max

0–0.6 m 0.6 m–1.2 m 0–0.6 m 0.6 m–1.2 m

20 20 20 20

1.86 1.93 36.95 32.92

1.18 1.28 4.09 5.14

0.72 0.26 32.20 26.35

4.22 3.92 45.55 44.10

0–0.6 m 0.6 m–1.2 m

20 20

16.76 16.64

3.41 5.33

9.85 10.60

21.10 25.20

ECe  soil salinity EMH  EM38 horizontal signal EMV  EM38 vertical signal SP  soil saturation percentage g  gravimetric water content

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

471

FIGURE 14-1. Survey of a marginally saline lettuce field in Indio, California, showing the interpolated EMV signal map for this field, along with the spatial positions of the 12 sampling locations.

regression equation should be used to describe the soil salinity/signal conductivity relationship in this field: ln(ECij)  0j  1j (z1i)  2j (z2i)  3j(z21i)  εij

(14-14)

where z1i  ln(EMV,i)  ln(EMH,i), and z2i  ln(EMV,i)  ln(EMH,i)

(14-15)

In Eq. 14-14, the subscript j  1, 2 corresponds to the two sampling depths, i  1, 2, . . . 2,040 corresponds to the EM38 sampling locations, 0j through 4j represent the two sets of regression model parameters (which define the two depth-specific prediction functions), and the residual errors for each sampling depth are assumed to be spatially uncorrelated. Table 14-2 presents the key summary statistics for each estimated regression function; these statistics include the R2, root mean square error (RMSE) estimate, overall model F-score and associated p-value, and the

472

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 14-2. Summary Statistics for Depth-Specific ln(ECe) Linear Regression Models: Indio Lettuce Field R2

RMSE

F-score

Pr  F

0–0.6 m

0.922

0.196

31.37

0.001

0.652

0.6–1.2 m

0.798

0.490

10.54

0.004

0.067

Depth

Moran Score

Pr  ZM

0.257 0.5

corresponding Moran test score and p-value. These latter Moran scores suggest that the uncorrelated residual assumption is valid. Likewise, residual QQ plots (not shown) confirm that the regression model errors follow a normal distribution and, hence, the ordinary linear regression modeling approach can be adopted. Additionally, the R2 values suggest that these regression models can be used to describe 92% and 80% of the 0 to 0.6 m and 0.6 to 1.2 m observed spatial log(ECe) patterns in this field, respectively. The spatial salinity pattern in the 0 to 0.6 m depth was of primary interest in this survey. More specifically, the field was scheduled to be leached if (1) 50% of the field was predicted to exhibit 0 to 0.6 m depth salinity levels 2 dS/m and/or the field average ln(ECe) level exceeded ln(2)  0.693, or (2) 25% of the field was predicted to exhibit 0 to 0.6 m depth salinity levels 3 dS/m. Table 14-3 presents the predicted field average ln(ECe) levels (and corresponding 95% confidence intervals), as well as the range interval estimates for both sampling depths. These predictions can be automatically calculated in the ESAP software package (using Eqs. 14-8 through 14-10, respectively). Figure 14-2 shows the corre-

TABLE 14-3. Regression Model Predicted Field Average ln(ECe) Levels and Range Interval Estimates: Indio Lettuce Field

Field average ln(ECe) 95% confidence interval

0–0.6 m Depth

0.6–1.2 m Depth

0.494 (0.35, 0.64)

0.548 (0.19, 0.91)

Range Interval Estimates (% Area of Field Classified into RIEs)

2.0 dS/m 2.0–3.0 dS/m 3.0–6.0 dS/m 6.0 dS/m RIE  range interval estimate

66.5 22.9 10.1 0.5

54.8 19.9 20.0 5.3

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

473

FIGURE 14-2. The corresponding predicted spatial salinity map for the field in Fig. 14-1. This map was produced within the ESAP SaltMapper program by interpolating the back-transformed, individual ln(ECe) predictions onto a finemesh grid using an adjustable smoothing kernel.

sponding predicted spatial salinity map for this field. This map was produced (within the ESAP SaltMapper program) by interpolating the backtransformed, individual ln(ECe) predictions onto a fine-mesh grid using an adjustable smoothing kernel. The results shown in Table 14-3 and Fig. 14-2 suggest that this field does not need to be leached. The 0 to 0.6 m field average ln(ECe) estimate is 0.494, and 66.5% of the individual 0 to 0.6 m depth predictions are calculated to be below 2 dS/m. Additionally, only 10.6% of these predictions are calculated to exceed 3 dS/m. Thus, none of the specified criteria for implementing a leaching process are met in this field. Within the preceding 5 years, the landowner had grown alternating winter vegetable crops of romaine lettuce and broccoli in this field. The ESAP software can be used to convert the Fig. 14-2 salinity map into projected relative yield loss maps for these crops, using standard salt-tolerance equations published for these vegetables (Mass and Hoffman 1977). Figure 14-3 shows the projected relative yield loss map for romaine lettuce, based on a threshold of 1.3 dS/m, a slope estimate of 13% yield loss

474

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 14-3. For the field shown in Figs. 14-1 and 14-2, the projected relative yield loss map for romaine lettuce, based on a threshold of 1.3 dS/m, a slope estimate of 13% yield loss per one unit increase in ECe (beyond the threshold), and a 80:20 root-weighting distribution (for the 0-m to 0.6-m and 0.6-m to 1.2-m depths, respectively). per one unit increase in ECe (beyond the threshold), and a 80% to 20% root-weighting distribution (for the 0 to 0.6-m and 0.6- to 1.2-m depths, respectively). The calculated field average romaine lettuce yield loss in this field is 8.7%. The corresponding calculated field average yield loss for broccoli is 1% (using a threshold of 2.8, a slope of 9.2%, and a 70% to 30% root-weighting distribution). These additional yield loss estimates also suggest that a full-scale leaching of this field is currently unwarranted, particularly if broccoli is the next scheduled crop in the rotation. Example 2: Pre- and postleaching surveys of a Coachella Valley vegetable field Pre- and postleaching EM surveys were performed by U.S. Salinity Laboratory personnel in July and October 2003 within a 13-ha vegetable field located in Thermal, California. The main goal of this survey was to spatially quantify the leaching process and determine the percent reduction in the post- versus preleaching median salinity levels in the field. A

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

475

total of 1,243 and 1,288 Geonics EM38 vertical (EMV, mS/m) and horizontal (EMH, mS/m) signal readings were collected within this field during the pre- and postleaching survey processes, respectively, and processed through the USDA-ARS ESAP software package. This software was again used to select 12 locations for soil sampling in each survey, using a prediction-based ESAP sample design (Lesch et al. 2000). Soil samples were collected from the 0 to 0.6 m sample depth and analyzed for soil salinity (ECe, dS/m), soil saturation percentage (SP, %), and gravimetric water content (g, %). Table 14-4 lists the univariate summary statistics for the EM38 survey and 0 to 0.6 m sample data associated with each survey event. Note that one soil sample in the preleaching survey event had to be discarded due to contamination during the laboratory analysis procedures. Figures 14-4 and 14-5 show the interpolated July (preleaching) and October (postleaching) EMH signal maps for this field, along with the spatial positions of the sampling locations. The results from an exploratory regression modeling analysis performed in ESAP confirmed that the following simple log(ECe)/log(EM) regression equation could be used to describe the soil salinity/signal conductivity relationship for each survey event in this field: ln(ECij)  0j  1j (z1ij)  εij

(14-16)

TABLE 14-4. Basic EM38 and Soil Sample Summary Statistics: Coachella Valley Vegetable Fielda Variable

Units

Date

N

Mean

Std.Dev.

Min

Max

EMH

mS/m

July

1243

23.25

9.12

10.63

79.75

EMV

mS/m

July

1243

44.35

13.29

27.25

124.63

EMH

mS/m

October

1288

30.99

13.10

15.25

121.88

EMV

mS/m

October

1288

48.26

18.69

27.75

175.38

ECe

dS/m

July

11

1.83

0.99

0.75

3.69

SP

%

July

11

32.53

2.36

29.44

37.33

g

%

July

11

0.12

0.03

0.06

0.16

ECe

dS/m

October

12

0.98

0.39

0.63

1.94

SP

%

October

12

34.07

5.88

28.63

46.33

g

%

October

12

0.24

0.10

0.11

0.44

a

All soil samples acquired from 0–0.6 m sampling depth ECe  soil salinity EMH  EM38 horizontal signal EMV  EM38 vertical signal SP  soil saturation percentage g  gravimetric water content

476

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 14-4. The interpolated July (preleaching) EMH signal maps of a Coachella Valley, California, vegetable field, along with the spatial positions of the sampling locations.

where z1ij  ln(EMV,i,j)  ln(EMH,i,j). In Eq. 14-16, the subscript j  1, 2 now corresponds to the two sampling dates, the i subscript correspond to the EM38 sampling locations acquired during each survey process, {01, 11} and {02, 12}represent the two sets of regression model parameters (which define the two time-dependent prediction functions), and the residual errors for each sampling depth are again assumed to be spatially uncorrelated. Table 14-5 presents the key summary statistics for each estimated regression function; these statistics again include the R2, root mean square error (RMSE) estimate, overall model F-score and associated p-value, and the corresponding Moran test score and p-value. The Moran scores and residual QQ plots (not shown) suggest that the normally distributed, uncorrelated residual assumption is valid. The RMSE and R2 values suggest that the postleaching LR model is more accurate; this increase in prediction accuracy is most likely due to the presence of higher and more uniform soil moisture conditions during the post-leaching survey process. In September 2003, a total of 64 cm of Colorado River water was applied to this field over a seven-day leaching cycle. The leaching was performed using 25 m-wide ponding basins laid out across the field,

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

477

FIGURE 14-5. The same field as in Fig. 14-4, in October (postleaching). after the soil had been deep-chiseled, plowed, and land-planed. The basins were laser-leveled and the water was released from a standpipe located within the northwest corner of the field (a head channel along the north edge of the field was used to deliver the water to each basin). Calculations from the flow and volume measurements performed during the leaching process suggested that approximately 55 cm of water infiltrated the soil and that the distribution uniformity of the basin system was 93%. The temporal change in the spatial salinity pattern in the 0 to 0.6 m depth was of primary interest in this survey. Table 14-6 shows the ESAPpredicted pre- and postleaching salinity summary statistics for this field. The postleaching median salinity level is estimated to be 0.91 dS/m, TABLE 14-5. Summary Statistics for Time-Specific ln(ECe) Linear Regression Models: Coachella Valley Vegetable Field Date

R2

RMSE

F-Score

July

0.600

0.340

13.51

October

0.837

0.148

51.37

Pr  F

Moran Score

Pr  ZM

0.005

1.37

0.5

0.001

0.55

0.5

478

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 14-6. Regression Model Predicted Field Average ln(ECe) Levels and Range Interval Estimates: Coachella Valley Vegetable Field

Field average ln(ECe) 95% confidence interval

July

October

0.513 (0.26, 0.76)

0.098 (0.19, 0.00)

Range Interval Estimates (% Area of Field Classified into RIEs)

1.0 dS/m 1.0–1.5 dS/m 1.5–2.0 dS/m 2.0–3.0 dS/m

16.6 27.2 21.4 20.9

68.0 25.5 4.6 1.7

3.0 dS/m

13.9

0.2

RIE  range interval estimate

which represents about a 46% decrease over the pre-leaching level (1.67). The ESAP-Calibrate software can perform a t-test on the difference between two field median (log mean) estimates; the corresponding t-score is this example is –5.14 (p  0.0001). Additionally, 68% of the field is estimated to exhibit postleaching salinity levels below 1 dS/m, and less than 2% of the field exceeds 2 dS/m. These estimates imply a substantial leaching effect, given that the corresponding preleaching estimates were 16.7% (1 dS/m) and 34.8% (2 dS/m), respectively. The predicted pre- and postleaching 0 to 0.6 m salinity maps for this field are shown in Figs. 14-6 and 14-7. A pronounced leaching effect can be clearly seen in the postleaching salinity map, and the near-surface salinity levels across the entire field appear to be significantly reduced. These results are perhaps not that surprising, given the large volume of water used during the leaching process (⬇ 8.3 ha-m). Finally, it is worthwhile to observe that the raw October (postleaching) EM38 signal data exhibited a higher average level than the July (preleaching) data (see Table 14-4 and Figs. 14-4 and 14-5). The general increase in the EM signal response was again most likely due to the elevated nearsurface soil moisture conditions. The top 30 cm of the soil profile was particularly dry during the July survey; these dry surface conditions undoubtedly depressed the EM38 signal response. These results demonstrate why a direct interpretation of EM38 signal data is often misleading. Note that the median near-surface soil salinity level in this field decreased by nearly 46%, even though the average horizontal EM signal reading increased from 23.3 mS/m to 31.0 mS/m.

FIGURE 14-6. The predicted preleaching 0- to 0.6-m salinity map for the field shown in Figs. 14-4 and 14-5.

FIGURE 14-7. The predicted postleaching 0- to 0.6-m salinity map for the field shown in Fig. 14-6. A pronounced leaching effect can be clearly seen here and the near-surface salinity levels across the entire field appear to be significantly reduced. 479

480

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SUMMARY This chapter demonstrates that a practical, regression-based methodology for the prediction of field-scale, spatial salinity patterns from soil conductivity survey data has substantial advantages in programs for the management of soil salinity. The basic parameter estimate and salinity prediction formulas for the ordinary linear regression model have been reviewed, along with the necessary modeling assumptions that have been built into the ESAP model, which also provides guidance for soil salinity sampling. The two case studies presented highlight the model estimation and salinity prediction capabilities of the ESAP software and demonstrate how bulk soil electrical conductivity survey data can be efficiently interpreted and used to quantify field-scale soil salinity information. It is worthwhile to note that although the focus of this chapter has been on predicting soil salinity from survey conductivity data, the associated statistical prediction methods discussed here are actually quite general. Indeed, these methods can be used to effectively model many different soil property/sensor data relationships, provided that the underlying modeling assumptions are satisfied. For a review of these more general calibration techniques, see Lesch and Corwin (2003) and/or the references contained in Table 1 of Corwin and Lesch (2005a). REFERENCES Atkinson, A. C., and Donev, A. N. (1992). Optimum experimental designs, Oxford University Press, Oxford, UK. Banerjee, S., Carlin, B. P., and Gelfand, A. E. (2004). Hierarchical modeling and analysis for spatial data, CRC Press, Boca Raton, Fla. Bogaert, P., and Russo, D. (1999). “Optimal spatial sampling design for the estimation of the variogram based on a least squares approach.” Water Resour. Res., 35, 1275–1289. Brandsma, A. S., and Ketellapper, R. H. (1979). “Further evidence on alternative procedures for testing of spatial autocorrelation amongst regression disturbances,” in Exploratory and explanatory statistical analysis of spatial data, C. P. A. Bartels and R. H. Ketellapper, eds., Martinus Nijhoff, Boston, 113–136. Brus, D. J., and de Gruijter, J. J. (1993). “Design-based versus model-based estimates of spatial means: Theory and application in environmental soil science.” Environmetrics, 4, 123–152. Brus, D. J., and Heuvelink, G. B. M. (2007). “Optimization of sample patterns for universal kriging of environmental variables.” Geoderma, 138, 86–95. Burgess, T. M., and Webster, R. (1984). Optimal sampling strategy for mapping soil types. I. Distribution of boundary spacings.” J. Soil Sci., 35, 641–654. Burgess, T. M., Webster, R., and McBratney, A. B. (1981). “Optimal interpolation and isarithmic mapping of soil properties. IV. Sampling strategy.” J. Soil Sci., 32, 643–654.

STATISTICAL MODELS FOR THE PREDICTION OF FIELD-SCALE

481

Cook, R. D., and Weisberg, S. (1999). Applied regression including computing and graphics, John Wiley and Sons, New York. Corwin, D. L., and Lesch, S. M. (2005a). “Apparent soil electrical conductivity measurements in agriculture.” Comp. Electron. Ag., 46, 11–43. ———. (2005b). “Characterizing soil spatial variability with apparent soil electrical conductivity. I. Survey protocols.” Comp. Electron. Ag., 46, 103–134. Cressie, N. A. C. (1991). Statistics for spatial data, John Wiley and Sons, New York. Haining, R. (1990). Spatial data analysis in the social and environmental sciences, Cambridge University Press, Cambridge, UK. Haskard, K. A., Cullis, B. R., and Verbyla, A. P. (2007). “Anisotropic Matèrn correlation and spatial prediction using REML.” J. Ag. Bio. Environ. Stats., 12, 147–160. Hendrickx, J. M. H., Das, B., Corwin, D. L., Wraith, J. M., and Kachanoski, R. G. (2002). “Indirect measurement of solute concentration,” in Methods of soil analysis, Part 4: Physical methods. Soil Science Society of America Book Series, J. H. Dane and G. C. Topp, eds., Soil Science Society of America, Madison, Wisc., 1274–1306. Lesch, S. M. (2005). “Sensor-directed response surface sampling designs for characterizing spatial variation in soil properties.” Comp. Electron. Ag., 46, 153–179. Lesch, S. M., and Corwin, D. L. (2008). “Prediction of spatial soil property information from ancillary sensor data using ordinary linear regression: Model derivations, residual assumptions and model validation tests.” Geoderma 148, 130–140. Lesch, S. M., and Corwin, D. L. (2003). “Using the dual-pathway parallel conductance model to determine how different soil properties influence conductivity survey data.” Agron. J., 95, 365–379. Lesch, S. M., Corwin, D. L., and Robinson, D. A. (2005). “Apparent soil electrical conductivity mapping as an agricultural management tool in arid zone soils.” Comp. Electron. Ag., 46, 351–378. Lesch, S. M., Rhoades, J. D., and Corwin, D. L. (2000). ESAP-95 version 2.10R: User manual and tutorial guide, Research Report 146, USDA-ARS, George E. Brown, Jr., ed., U.S. Salinity Laboratory, Riverside, Calif. Lesch, S. M., Strauss, D. J., and Rhoades, J. D. (1995). “Spatial prediction of soil salinity using electromagnetic induction techniques: 2. An efficient spatial sampling algorithm suitable for multiple linear regression model identification and estimation.” Water Resour. Res., 31, 387–398. Littell, R. C., Milliken, G. A., Stroup, W. W., and Wolfinger, R. D. (1996). SAS system for mixed models, SAS Institute Inc., Cary, N.C. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. and Drainage Div., ASCE, (103 IR2), 115–134. Minasny, B., McBratney, A. B., Walvoort, D. J. J. (2007). “The variance quadtree algorithm: Use for spatial sampling design.” Comput. Geosci. 33, 383–392. Müller, W. G. (2001). Collecting spatial data: Optimum design of experiments for random fields, 2nd ed., Physica-Verlag, Heidelberg, Germany. Müller, W. G, and Zimmerman, D. L. (1999). “Optimal designs for variogram estimation.” Environmetrics, 10, 23–37. Myers, R. H. (1986). Classical and modern regression with applications, Duxbury Press, Boston. Myers, R. H., and Montgomery, D. C. (2002). Response surface methodology: Process and product optimization using designed experiments, 2nd ed., John Wiley and Sons, New York.

482

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Press, S. J. (1989). Bayesian statistics: Principles, models, and applications, John Wiley and Sons, New York. Rhoades, J. D., Chanduvi, F., and Lesch, S. M. (1999). Soil salinity assessment: Methods and interpretation of electrical conductivity measurements, FAO Irrigation and Drainage Paper No. 57, Food and Agriculture Organisation of the United Nations, Rome. Russo, D. (1984). “Design of an optimal sampling network for estimating the variogram.” Soil Sci. Soc. Am. J., 48, 708–716. Schabenberger, O., and Gotway, C. A. (2005). Statistical methods for spatial data analysis, CRC Press, Boca Raton, Fla. Shapiro, S. S., and Wilk, M. B. (1965). “An analysis of variance tests for normality (complete samples).” Biometrika, 52, 591–611. Thompson, S. K. (1992). Sampling, John Wiley and Sons, New York. Tiefelsdorf, M. (2000). Modeling spatial processes: The identification and analysis of spatial relationships in regression residuals by means of Moran’s I, Springer-Verlag, New York. Upton, G., and Fingleton, B. (1985). Spatial data analysis by example, John Wiley and Sons, New York. Valliant, R., Dorfman, A. H., and Royall, R .M. (2000). Finite population sampling and inference: A prediction approach, John Wiley and Sons, New York. Warrick, A. W., and Myers, D. E. (1987). “Optimization of sampling locations for variogram calculations.” Water Resour. Res., 23, 496–500. Zhu, Z., and Stein, M. L. (2006). “Spatial sampling design for prediction with estimated parameters.” J. Ag. Bio. Environ. Stats., 11, 24–44.

NOTATION BLUE  best linear unbiased estimate ECe  soil salinity EMH  EM38 horizontal signal EMI  electromagnetic induction EMV  EM38 vertical signal e  (n  1) vector of residual errors RIE  range interval estimate SP  soil saturation percentage X  (n  p) regression model design matrix Y  (n  1) vector of soil property measurements   (p  1) parameter vector M  Moran residual test statistic g  gravimetric water content

CHAPTER 15 SPATIALLY DISTRIBUTED SOLUTE BALANCE IN A CALIFORNIA WATER DISTRICT Charles A. Young, Wesley W. Wallender, and Kenneth K. Tanji

INTRODUCTION In arid irrigated agricultural regions, a primary concern is the concentration of salts in the rootzone. Excessive soil-water salt concentration can retard plant growth and degrade soil structure if sodium is the dominant cation. The salts may be naturally present in the soil matrix, as well as introduced via applied irrigation water. In regions with a shallow water table, salts can move upward into the rootzone through capillary and evaporative processes. The west side of the San Joaquin Valley, California suffers from these problems, and they threaten the productivity of the area (SJVDP 1990). Subsurface drain systems have been installed to control water table elevations and rootzone salinity. However, soils in this region, derived from marine sediments, contain selenium that is mobilized by irrigation water and collected in drain systems. The disposal of seleniumtainted drain water has become a serious problem due to the potentially toxic impacts on wildlife from bioaccumulation of selenium in the food chain. Regulatory limits on selenium discharge may threaten the ability of the region’s growers to control salt accumulation in the rootzone. The Panoche Water District and six neighboring water and irrigation districts, which make up the Grasslands Basin, have historically discharged their drainage water to the San Joaquin River using unlined canals. Starting in October 1996, the districts ceased transporting drain water through these canals due to environmental concerns within the region through which the canals pass. Instead, drain water is now transported to the San Joaquin River using the concrete-lined San Luis Drain. The agreement allowing the use of the San Luis Drain stipulates that selenium load targets must be met at the point of discharge into the river. If 483

484

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

these load limits are exceeded, the water districts risk monetary penalties and the loss of access to the San Luis Drain. In the past, the Panoche Water District has been able to maintain acceptable salt levels in the rootzone for agricultural production because off-site drainage was allowable. It is unclear, however, if constraints imposed by current or future regulations will reduce drainage to the point where salt accumulation occurs in crop rootzones. Past efforts at calculating salt balances have been at a scale too coarse to identify areas where salt accumulation may first occur. This requires knowledge of the spatial variability in those factors that affect the occurrence of salts. These factors include hydrogeology, water table depth, location and performance of subsurface drains, and irrigation management. This chapter describes how a spatially distributed total dissolved solids (TDS) and selenium balance was developed for the 15,000-ha Panoche Water District, using data from 98 subregions within the district, and the management complexities that were identified as a result.

LOCAL GEOLOGY OF THE PANOCHE WATER DISTRICT The single most important factor affecting soil salinity levels in the Panoche Water District is the local geology. Figure 15-1 shows that the district lies on two alluvial fans and the interfan deposits between them (Fio 1994). The western side of the district lies on the upper and middle reaches of the Little Panoche Creek alluvial fan, and the eastern side lies on the northern boundary of the much larger Panoche Creek alluvial fan. A relatively small area in the south end of the district lies on interfan deposits. Differences in the chemical composition of groundwater between the two fans are related to the types of rock in their respective drainage basins. The Little Panoche Creek alluvial fan contains groundwater that is a sodium chloride type, relatively low in salinity, with selenium concentrations ranging from 1 to 27 g/L. In the Panoche Creek alluvial fan the groundwater is a sodium sulfate type, relatively high in salinity, with selenium concentrations ranging from 20 to 400 g/L (Fio and Leighton 1994). Salinity varies within, between, and at the interface of alluvial fans. Deverel and Gallanthine (1989) showed that groundwater salinity is correlated with soil salinity, which increases at alluvial fan boundaries. Fio and Leighton (1994), using oxygen isotope analysis, concluded that regions with a shallow water table have experienced evapoconcentration resulting in water with much higher selenium and salt concentrations. At the northern edge of the district on the Panoche Creek alluvial fan, evapoconcentrated water from a depth of 8.3 m had a selenium concentration of 1,700 g/L.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

485

FIGURE 15-1. Panoche Water District and alluvial fan boundaries.

The water table is within 3 m of the ground surface for more than half of the district, and drainage systems control the water table elevation and remove accumulating salts (Fig. 15-2). According to numerical modeling and water quality analyses in a similar region to the south of the district, drain flow is a combination of deep, evapoconcentrated groundwater and less-saline, shallow groundwater (Deverel and Fio 1991a,b). Each of the installed drain systems delivers water to one of 50 sumps where flow volumes and water quality are monitored. Drain water from the sumps flow

486

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-2. Location of drainage systems and extent of shallow water table.

in open collection drains to the northeast, eventually meeting the San Luis Drain for transport to the San Joaquin River. Irrigation water management also plays a role in soil and water salinity levels. Salts added to the rootzone by irrigation must be removed by applied water when the soil salinity is in excess of the threshold salinity of crops grown in this area; this is typically called the leaching requirement. The common management strategy is the application of a large irrigation prior to planting (pre-irrigation) that flushes salts from the rootzone that have built up from the previous year’s irrigation. However, in regions with

SPATIALLY DISTRIBUTED SOLUTE BALANCE

487

gypsiferous saline soils, irrigation in excess of the leaching requirement also increases the downward salt flux through dissolution of gypsum. To evaluate the rate of salt accumulation or leaching, Orlob (1991) calculated a salt balance for the entire San Joaquin Basin. Prior to development of the basin’s water resources, salts were excreted and accreted at the same rate. In the early 1950s, with the completion of the Friant Dam and the Delta-Mendota Canal (which imports water from northern California), salts began to accumulate in the basin (Fig. 15-3). The accumulation of salts within the basin was due to the reduction in outflow from the basin, importation of salts with water from northern California, and irrigation of marginally productive, saline soils. However, salt has not accumulated uniformly throughout this region. Tanji (1977) calculated a district-scale salt balance for the rootzone of the Panoche Water District in calendar year 1975. On average, the storage of salt was decreasing in the rootzone and naturally occurring gypsum was considered the likely source of the leaching salts.

METHODS The availability of spatially distributed water flow and salinity data for the Panoche Water District now allows for the calculation of a salt balance at a finer scale. Young and Wallender (2002) describe a water balance for 98 subregions within the Panoche Water District. Using the results of that water balance study, the objectives of this study were to: 1. Develop a methodology for calculating the spatial distribution of salt and selenium balance components using data collected by the Panoche Water District.

FIGURE 15-3. Net salt accumulation in the San Joaquin basin, 1930–1989. Data from Orlob (1991).

488

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

2. Develop and evaluate spatially distributed drain water disposal costing strategies according to drainage volume, selenium load, deep percolation volumes, or selenium discharge to the environment. Data Data layers containing the boundaries of water balance control volumes and drainage areas were obtained from Young and Wallender (2002) and Fio (1994), respectively. Monthly drain flow and TDS concentration data were collected by district staff during the study period. Selenium concentration data at individual sumps were collected by the district for varying time intervals from June 1991 to September 1994. Total Dissolved Solids Balance Using the control volumes (area times the depth of rootzone or deeper in the profile to include part or all of the groundwater systems) in the water balance calculations (Young and Wallender 2002), a salt balance was used to estimate change in salt storage as the closure term (Fig. 15-4). No data were available on application of fertilizers, animal manures, and other soil amendments. Control volumes with drains extend from the ground surface to the bottom of the drain trench (Fig. 15-4). In undrained control volumes, the bottom boundary is the bottom of the rootzone. The salt balance equations, assuming fluxes into the control volume are positive, are (I  CI)  (R  CR)  (D  CD)  (GW  CGW)  STDS (drained)

FIGURE 15-4. Salt balance schematic used in model.

(15-1)

SPATIALLY DISTRIBUTED SOLUTE BALANCE

(I  CI)  (R  CR)  (WT  CWT)  (Dp  CDp)  STDS (undrained)

489

(15-2)

where I  infiltrating applied water (m3/y) CI  total dissolved solids (TDS) concentration of applied water (kg/m3) R  infiltrating rainfall (m3/y) CR  TDS concentration of rainfall (kg/m3) D  drainflow (m3/y) CD  TDS concentration of drainage water (kg/m3) GW  recharge to the groundwater system (m3/y) CGW  TDS concentration of recharge to the groundwater system (kg/m3) Dp  deep percolation (m3/y) CDp  TDS concentration of deep percolation (kg/m3) WT  water from water table (m3/y) CWT  TDS concentration of groundwater at water table (kg/m3) STDS  change in TDS storage within control volume (kg/y) The magnitudes of I, R, D, Dp, WT, and GW were measured or calculated in the water balance study (Young and Wallender 2002). TDS concentrations of the balance components were measured or estimated using data from previous studies. The electrical conductivity of delivered irrigation water was reported as 0.7 dS/m (Burt and Styles 1994). An acceptable conversion factor for this water is 640 mg/L per dS/m (Tanji 1990). This resulted in an equivalent of 448 mg/L TDS. Rainfall was assumed to have a TDS concentration of 10 mg/L (Tanji 1977). Drained control volumes District personnel measured the TDS concentration of drain flows at 50 sumps on a monthly basis. The quality of drainage water is a reflection of the contributions of deep evapoconcentrated groundwater and shallow, less-saline groundwater (Fio and Deverel, 1991a,b). TDS concentration varies in time as the relative contributions of these components vary. In this study it was assumed that the monthly measurements represented the average TDS concentration of drain flow for a given month. Calculated monthly drainage volumes for each delivery-drainage control volume (Young and Wallender 2002) were multiplied by the monthly TDS concentration measured at the corresponding sump. Monthly TDS loads removed by drainage were then summed for the year. The annual totals from each delivery-drainage polygon were then summed by delivery control volume to give the total drain load of TDS from each delivery control volume. To estimate the quality of the shallow and deep groundwater, it was assumed that the minimum and maximum TDS concentration measured

490

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

at each sump during the study period represented the quality of shallow (within the control volume) and deep groundwater (below the control volume), respectively. Recharge to the deep groundwater system (positive GW) is this water leaving the control volume and was assumed to have the salinity of shallow groundwater, and deep groundwater discharge entering the control volume (negative GW) was assumed to have the salinity of deep groundwater. This assumption is supported by numerical modeling and water chemistry analyses done in a similar region located near the study area (Deveral and Fio 1991a,b). Again, in these analyses drain water was found to be a combination of deep, evapoconcentrated groundwater and shallow, less-saline water. Therefore, it is assumed that water leaving the control volume (positive GW) is made up of shallow, less-saline water and can be represented by the lowest TDS concentration measured in the drains. Negative GW represents an influx of deeper, more-saline water and is estimated by the largest value of TDS concentration measured. Figures 15-5 and 15-6 show the installed drain system boundaries and the maximum and minimum measured drainage TDS concentrations. Maximum and minimum TDS concentration values for each control volume were calculated using volume-weighted averaging of the contributions from each drainage area that contributed drainage to the control volume. This was done because drain system boundaries did not correspond with the delivery control volume boundaries. Groundwater TDS concentration was lower and less variable on the Little Panoche Creek alluvial fan than on the Panoche Creek alluvial fan (Figs. 15-5 and 15-6). This supports past research showing that older drains on the Little Panoche Creek alluvial fan have removed more of the nearby evapoconcentrated water in comparison to the Panoche Creek alluvial fan (Fio and Leighton 1994). The highest TDS concentrations on each fan are found at the margins of the Little Panoche Creek alluvial fan and in the northern areas of the Panoche Creek alluvial fan, in agreement with the findings of Deverel and Gallanthine (1989). Undrained control volumes The quality of water leaving the rootzone (Dp) in the undrained areas was not measured and, without an estimate of this value, it is impossible to calculate a change in TDS storage. To overcome this, a linear function through the origin was fitted to data relating the amount of TDS removed from drained control volumes having groundwater recharge and the amount of water traveling through it, the sum of D and GW (Fig. 15-7). This relationship was developed using drained control volumes that recharged the groundwater system, because this is the case most similar to the undrained regions where the control volumes end at the rootzone

SPATIALLY DISTRIBUTED SOLUTE BALANCE

491

FIGURE 15-5. Maximum measured TDS concentration at sumps (10/1/95– 9/30/96). (Fig. 15-4). Data for the Panoche Creek alluvial fan and interfan areas were grouped according to similarities in groundwater quality (Figs. 15-5 and 15-6). The linear functions were used to predict the load of TDS removed from the undrained control volumes as a function of deep percolation (Dp). An overestimation of TDS removal may have occurred using this method because soil salinity levels decrease in upper alluvial fan areas (Harradine 1950). This effect would be more pronounced on the Little Panoche Creek

492

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-6. Minimum measured TDS concentration at sumps (10/1/95– 9/30/96). alluvial fan because a relatively large percentage of it is within the study area. In addition, this approach ignored differences in soil salinity and mineral weathering between the saturated and unsaturated zones and differences in soil leaching caused by preferential flow paths. However, salt loading differences between the alluvial fans were captured. For the case where Dp was negative, either a net movement of water upward from a shallow water table occurred or actual evapotranspiration was less than potential due to underirrigation. The change in TDS storage

SPATIALLY DISTRIBUTED SOLUTE BALANCE

493

FIGURE 15-7. TDS pickup rates for Panoche and Little Panoche alluvial fans. was calculated as the sum of TDS from applied water and rainfall. Additional TDS load was added if the water table was shallow enough to contribute water to transpiration or evaporation. Selenium Balance Irrigation and rainfall did not contribute selenium to the control volumes. Removing these components from Eqs. 15-1 and 15-2 and assuming selenium is nonreactive, (GW  CsGW)  (D  CsD)  Ss (drained)

(15-3)

(WT  CSWT)  (Dp  CsDp)  Ss (undrained)

(15-4)

where CsD  selenium concentration of drainage water (kg/m3) CsGW  selenium concentration of recharge to the groundwater system (kg/m3) CsDp  selenium concentration of deep percolation (kg/m3) CSWT  selenium concentration of groundwater at water table (kg/m3) Ss  change in selenium storage within control volume (kg/y) Drained control volumes The same methodology described for TDS was used for selenium. Selenium removal by the drainage systems was calculated on a sump-bysump basis using an average of concentration measurements made by the

494

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

district during 1991–1994. No other data were available; however, a primary goal of this study was the development of a procedure for calculating mass balance. Future efforts will be enhanced by additional data collection. Selenium concentration was multiplied by the total annual drain flow from each delivery-drainage polygon (Young and Wallender 2002). The annual totals from each delivery-drainage polygon were grouped by delivery control volume and the total drain load of selenium was calculated. Selenium concentrations for groundwater recharge and discharge in the drained regions were the maximum (Fig. 15-8) and minimum (Fig. 15-9) of measurements taken at drainage sumps during 1991–1994. Maximum

FIGURE 15-8. Maximum measured selenium concentration at sumps (1991–1994).

SPATIALLY DISTRIBUTED SOLUTE BALANCE

495

FIGURE 15-9. Minimum measured selenium concentrations at sumps (1991–1994). and minimum selenium concentrations were calculated for each control volume using volume-weighted averaging. Consistent with the findings of Fio and Leighton (1994), drain water on the Panoche Creek alluvial fan has higher selenium concentrations than the Little Panoche Creek alluvial fan (Figs. 15-8 and 15-9). On the Little Panoche Creek alluvial fan, the highest selenium concentrations are at the fan boundaries. The selenium concentrations show no clear pattern on the Panoche Creek alluvial fan, but only a small portion of this relatively large fan is considered.

496

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Undrained control volumes The relationship between selenium load and water flow for drained control volumes having recharge to the groundwater system (Fig. 15-10) was used to estimate the change in selenium storage in undrained control volumes. Selenium load was linearly related, with the line through the origin, to the volume of water passing through the control volume (Dp). When Dp was negative, the change in storage was equivalent to the load of selenium from shallow water table evaporation. Drainage Disposal Cost Distribution The spatial distribution of drain water disposal cost was calculated assuming the district had exceeded its selenium discharge limits to the San Luis Drain and was required to pay a hypothetical $100,000 penalty. The simplest method for distributing this cost is a fee charged per acre of irrigated land within the district. However, this method would not account for the spatial variation in selenium discharge. This cost could be spatially distributed using alternative measurements, such as drainage volume, drainage selenium load, deep percolation, or control volume selenium discharge load. The drainage volume charge per control volume was the drainage produced in each control volume divided by the district total drainage (fraction) multiplied by the total $100,000 penalty. For drainage selenium load, the cost was distributed by the fraction of the total selenium load per

FIGURE 15-10. Selenium pickup rates for Panoche and Little Panoche alluvial fans.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

497

control volume. For deep percolation, it was by the fraction of the total volume of deep percolation in excess of the leaching requirement (calculated assuming a maximum rootzone soil water salinity of 7.5 dS/m). Last, the control volume selenium discharge load charge per control volume was calculated by multiplying the total penalty by the fraction of the total selenium discharge. The control volume discharge load was the amount of selenium that left the control volume. Because the imported irrigation water is nearly selenium-free, in undrained control volumes and drained control volumes that recharged the groundwater, the discharge load was equivalent to the change in storage. In drained control volumes discharging groundwater (negative GW), the discharge load was the drainage load.

RESULTS Total Dissolved Solids Balance The TDS loading by rain and irrigation water (Fig. 15-11) follows the pattern of applied water found in the water balance study (Young and Wallender 2002). The highest drainage loads of TDS were in the center and northwestern parts of the district (Fig. 15-12). This corresponds with the location of greatest drainage (Fig. 9, Young and Wallender 2002). Salts are contributed to the deep groundwater system in the drained regions (Fig. 15-13), with the maximum occurring at the alluvial fan boundaries. In the northern region salts move upward from the deep groundwater to the control volume (negative sign). Control volumes showing zero TDS to deep groundwater in the south and west areas of the district neither received nor contributed TDS to the deep groundwater system because of zero deep percolation and a deep water table. The change in TDS storage in tons/ha, as the closure term in Eq. 15-1, is shown in Fig. 15-14. The change was not uniform, with some regions on both alluvial fans and the interfan accumulating TDS (positive change in storage), while in other regions TDS was depleted (negative change in storage). Accumulation largely occurred in the drained regions, with the maximums occurring roughly in regions of maximum groundwater discharge (Young and Wallender 2002). In undrained regions, control volumes with crop consumption in excess of applied water show salt accumulation in the rootzone because water lost from the control volumes as evapotranspiration left the salt behind. On both alluvial fans and the interfan, more TDS were removed from storage in the undrained regions compared to the drained regions (Table 15-1). Greater deep percolation in the undrained regions (Young and Wallender 2002) coupled with the salt dissolution were causes for higher TDS removal. If the salt pickup rate is

498

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-11. TDS load from applied water and rain (10/1/95–9/30/96). lower than assumed on upper fan reaches, this difference decreases. The control volumes in the drained areas extend below the rootzone, so it is possible that TDS accumulated below, not within, the rootzone. However, the accumulation of salts in the shallow groundwater and soils below the rootzone can pose a problem over time if these salts are carried upward in the water moving into the rootzone. The difference in average change in TDS storage between the alluvial fans is attributable to geological differences. The high soil and water salinity observed on the Panoche Creek alluvial fan and interfan make salt accumulation more likely.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

499

FIGURE 15-12. TDS load removed by drainage systems (10/1/95–9/30/96).

District-wide, the average change in TDS storage was 3.0 tons/ha, signifying a loss of TDS from the control volumes. This value is considerably less than 16.8 tons/ha calculated in an earlier district-scale rootzone salt balance (Tanji 1977). The difference may be due to one or more of the following reasons: 1. The rate of gypsum dissolution may have decreased in the 20 years since the district-scale study performed by Tanji (1977). 2. In drained areas, the control volumes extend to just below the plane of the drains. The 1977 district-scale model included only the rootzone.

500

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-13. TDS load contributed to the deep groundwater system below the control volume (10/1/95–9/30/96).

Salt accumulation below the rootzone increases the average change in storage of salts. 3. The assumptions regarding groundwater salinity may be incorrect. Selenium balance The amount of selenium removed by the drains (Fig. 15-15) was greatest on the Panoche Creek alluvial fan and the interfan. Selenium removed by the drainage system was added to selenium leaving the control vol-

SPATIALLY DISTRIBUTED SOLUTE BALANCE

501

FIGURE 15-14. Net TDS accumulation in control volumes (10/1/95–9/30/96).

umes via the groundwater (Fig. 15-16) to calculate change in selenium storage (Fig. 15-17). Average decrease in selenium storage on the Panoche Creek alluvial fan and the interfan occurred at a much higher rate than on the Little Panoche Creek alluvial fan (Table 15-1). Selenium storage decreased in proportion to the volume of deep percolation in the undrained regions. In undrained control volumes, selenium did not accumulate because rainfall and irrigation water contained no analytically detectable selenium, and the water table was too deep in most areas to contribute selenium to the rootzone. In drained regions, selenium accumulated in

502

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-15. Selenium removed by drainage systems (10/1/95–9/30/96). areas similar to that seen in the TDS balance, with the maximum occurring on the Panoche Creek alluvial fan and the interfan. Drainage charges The spatial distribution of the drainage penalty in $/ha, assuming a charge on drainage volume, is shown in Fig. 15-18. However, the wide range in drain water selenium concentrations (Figs. 15-8 and 15-9) suggest that a charge levied uniformly on a volume basis across the district would unfairly impact growers on the Little Panoche alluvial fan where relatively little selenium originated.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

503

FIGURE 15-16. Selenium contributed to the groundwater system (10/1/95– 9/30/96).

The per hectare costs associated with a charge levied per kilogram of selenium discharged from the drainage systems is shown in Fig. 15-19. In this scenario, the growers on the Panoche Creek alluvial fan and the interfan would pay more for drainage disposal in accord with higher selenium loads. Neither of the aforementioned methods penalizes poor management in the upslope, undrained regions that may be contributing to the drainage problem. Upslope regions recharged the water table at more than three times the rate of the downslope, undrained regions. Particle-tracking

504

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-17. Net selenium accumulation in control volumes (10/1/95– 9/30/96).

experiments using groundwater models show that deep percolation in upslope regions enters downslope drainage systems (Fio 1994; Purkey, personal communication, 1997). What is less clear is how reducing deep percolation in upslope areas will affect downslope water table elevations. Assuming a hydraulic connection between upslope deep percolation and downslope water table elevation, a map was developed showing the drainage charge spatially distributed based on the volume of deep percolation in excess of leaching requirements (Fig. 15-20). Some regions of the

SPATIALLY DISTRIBUTED SOLUTE BALANCE

505

TABLE 15-1. TDS and Selenium Change in Storage (10/1/95–9/30/96) ha

Panoche Creek alluvial fan and interfan Drained Undrained Total Little Panoche Creek alluvial fan Drained Undrained Total District Total

Tons TDS

Tons/ha TDS

kg Se

kg/ha Se

6,719 2,237

16,229 34,308

2.4 15.3

855 1,158

0.1 0.5

8,956

50,537

5.6

2,013

0.2

4,185 2,645

8,433 4,890

2.0 1.8

126 39

3.0  102 1.0  102

6,830

3,543

0.5

87

1.3  102

15,786

46,994

3.0

1,926

0.1

FIGURE 15-18. Per hectare cost of drain volume charge (10/1/95–9/30/96).

506

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-19. Per hectare cost of selenium load charge (10/1/95–9/30/96). district would have no disposal cost because irrigation water was insufficient to leach salts that accumulated in the rootzone. The distribution of the drainage penalty shown in Fig. 15-20 penalizes upslope areas for excess deep percolation; however, differences in selenium loading caused by geological variation are not considered. A charge on the amount of selenium discharged into the environment from a control volume accounts for both factors (Fig. 15-21). This charge might stimulate better water management in undrained regions and in areas that are more likely to discharge selenium due to local geology.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

507

FIGURE 15-20. Per hectare cost of excess recharge to water table (10/1/95– 9/30/96).

SUMMARY AND CONCLUSIONS The results from the calculation of the salt balance demonstrate the difficulty faced by growers in reducing selenium discharge in drainage while maintaining acceptable soil salinity levels. Specific conclusions include: 1. Calculation of a spatially distributed TDS and selenium balance is possible with currently collected data sets. However, verification of

508

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 15-21. Per hectare cost of selenium load discharged to environment (10/1/95–9/30/96). assumptions regarding the concentration of deep percolating waters is required. 2. The accumulation and discharge of salts and selenium is highly correlated to the regional geology. The Panoche Creek alluvial fan and interfan areas had a calculated reduction in TDS storage of 5.6 tons/ha, while the Little Panoche Creek alluvial fan reduced storage by 0.5 tons/ha. Selenium storage was reduced by 0.2 kg/ha on the Panoche Creek alluvial fan and interfan areas and by 1.3 3 1022 kg/ha on the Little Panoche Creek alluvial fan.

SPATIALLY DISTRIBUTED SOLUTE BALANCE

509

4. The goal of selenium load reduction at the district drainage discharge point is in conflict with the need of individual growers to control rootzone salinity. The results of this study show that reductions in drainage volume/selenium load will likely increase TDS and selenium accumulation in drained areas. 5. Financial charges designed to reduce selenium drainage load should be a function of geological variation and water management. • A charge on drainage volume will unfairly penalize growers on the Little Panoche Creek alluvial fan. • A charge on drainage selenium load will be more equitable, yet it will not factor in adverse effects caused by poor management upslope of the drained areas. • A charge on deep percolation in excess of leaching requirements will motivate better management through reduced deep percolation but with no accounting for geological differences. • A charge on the load of selenium discharged from a control volume will stimulate a reduction in deep percolation and will account for differences in geology. 6. The methods outlined in this and Young and Wallender (2002) can provide information regarding the source and fate of salts to district and ranch management without the use of complex numerical groundwater and solute models. REFERENCES Burt, C. M., and Styles, S. (1994). Grassland basin irrigation and drainage study, final report, Irrigation Training and Research Center, California Polytechnic Institute, San Luis Obispo, Calif. Deverel, S. J., and Fio, J. L. (1991a). “Groundwater flow and solute movement to drain laterals: Western San Joaquin Valley, California, 1. Geochemical assessment.” Water Resour. Res., 27(9), 2233–2246. ———. (1991b). “Groundwater flow and solute movement to drain laterals: Western San Joaquin Valley, California, 2. Quantitative hydrologic assessment.” Water Resour. Res., 27(9), 2247–2257. Deverel, S. J., and Gallanthine, S. K. (1989). “Relation of salinity and selenium in shallow groundwater to hydrologic and geochemical processes, western San Joaquin Valley, California.” J. Hydrol., 109, 125–149. Fio, J. L. (1994). Calculation of a water budget and delineation of contributing sources to drainflows in the western San Joaquin Valley, California, U.S. Geological Survey Open File Report 94-45, USGS, Washington, D.C. Fio, J. L., and Leighton, D. A. (1994). Effects of ground-water chemistry and flow on quality of drainflow in the western San Joaquin Valley, California, U.S. Geological Survey Open File Report 94-72, USGS, Washington, D.C. Harradine, F. (1950). Soil survey of western Fresno County, University of California Press, Berkeley, Calif.

510

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Orlob, G. T. (1991). “San Joaquin salt balance: Future prospects and possible solutions,” in The economics and management of water and drainage in agriculture, Kluwer Academic Publishers, Boston, 143–167. San Joaquin Valley Drainage Program (SJVDP). (1990). “A management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley,” California Natural Resources Agency, http://ceic.resources. ca.gov/catalog/USBureauOfReclamationMidPacificRegionMPGIS/LibraryOfTheSanJoaquinValleyDrainageProg.html, accessed February 5, 2011. Tanji, K. K. (1977). “Managing saline water for irrigation.” Paper presented at International Conference on Managing Saline Water for Irrigation: Planning for the Future, International Society of Soil Science, Wageningen, The Netherlands. Tanji, K. K., ed. (1990). Agricultural salinity assessment and management, ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va. Young, C. A., and Wallender, W. W. (2002). “Spatially distributed irrigation hydrology: Water balance.” Trans. ASAE, 45(3), 609–618.

NOTATION CD  TDS concentration of drainage water, kg/m3 CDp  TDS concentration of deep percolation, kg/m3 CGW  TDS concentration of recharge to groundwater system, kg/m3 CI  TDS concentration of applied water, kg/m3 CR  TDS concentration of rainfall, kg/m3 CsD  selenium concentration of drainage water, kg/m3 CsDp  selenium concentration of deep percolation, kg/m3 CsGW  selenium concentration of recharge to groundwater system, kg/m3 CsWT  selenium concentration of groundwater at water table, kg/m3 D  drainflow, m3/y Dp  deep percolation, m3/y GW  recharge to the groundwater system, m3/y I  infiltrating applied water, m3/y R  infiltrating rainfall, m3/y Ss  change in selenium storage within control volume, kg/y STDS  change in TDS storage within control volume, kg/y

PART FIVE: SALINITY MANAGEMENT OPTIONS

CHAPTER 16 ON-FARM IRRIGATION AND DRAINAGE PRACTICES James E. Ayars

INTRODUCTION Irrigators have many options for managing their irrigation and drainage systems. The options that they choose significantly affect the salinization of the soil and the depth of the water table, which, in turn, influence the productivity and profitability of their farms. On-farm water management decisions can also change the quantity and quality of surface and subsurface return flows, affecting both neighboring farms and the general environment. This chapter contains brief descriptions of the most common types of irrigation and drainage systems. It outlines associated management options that help prevent or correct salinity problems and minimize buildup of water tables, demonstrates the importance of drainage for salinity management, describes methods for management of shallow water tables, and describes an alternative method for disposal of saline drainage water.

IRRIGATION AND SALINITY CONTROL Over the long term, irrigation must be adequate but not excessive to prevent harmful accumulation of salt in the rootzone and to prevent a high water table that may contribute to salt accumulation at the soil surface. Infiltrated depths of water must be relatively uniform to meet the crop’s needs and leach salts adequately, without excessive surface runoff or deep percolation. To meet such depth and uniformity requirements, irrigation systems must be suited to the site, well-designed, and well-managed. 511

512

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Surface Irrigation In surface irrigation systems, such as the furrow, border dike, and basin systems, water flows over the field’s surface and part of the water infiltrates as the stream advances. The length of the advance time will determine the uniformity as a result of the differences in infiltration opportunity time between the head and tail end of the field. When the water has advanced to the end of the field, runoff may occur or the water may pond on the soil surface along the lower end of the field. This ponding will continue until inflow is stopped and all surface water has infiltrated. For uniform infiltration, the variability of soil characteristics across the field must be small, and the infiltration opportunity times must be uniform in all parts of the field. A major limitation of surface irrigation results from the depth of infiltration being controlled by the soil surface characteristics, such as surface cracking or sealing, the antecedent water content, roughness, and compaction.

Furrow and Corrugation Irrigation Furrow irrigation is commonly used for trees, vines, and annual row crops. Corrugations, similar to furrows but with smaller cross sections, are used for alfalfa, small grains, and other close-growing crops. Furrow and corrugation irrigations can be reasonably uniform if advance times to the end of the field are small, on the order of one-fourth to one-fifth of the time needed to infiltrate the necessary water. Advance times can be decreased by using the maximum flow rate of the non-erosive irrigation stream and by shortening run lengths, such as 800 m to 400 m. Excessive runoff can sometimes be prevented by reducing the size of each furrow stream (cutback) after advance is complete. An alternative to cutting back the stream size is to collect tailwater and return it to the irrigation stream or apply it to an adjacent field. On some soils, especially those with coarser textures, surge flow (cycling the furrow inflow stream on and off) causes advance to be completed with less water than with continuous furrow inflow (Stringham and Keller 1979) and results in more uniform infiltration. Chemical amendments (e.g., polyacrylimide, PAM) have been applied to both the soil surface and the irrigation stream to modify the infiltration. There have been mixed results with PAM both increasing and decreasing the infiltration rates depending on the concentrations (Trout et al. 1995; Lentz 2003). PAM has been effective in controlling erosion due to surface irrigation and provides significant water quality benefits (Trout et al. 1995). Mathematical techniques for simulating the advance and recession of the stream and depth of infiltration have been under development since the mid 1970s. Such techniques can be used to determine the optimal

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

513

stream size and water application time for irrigation of furrows. However, the infiltration conditions must be known accurately and expressed mathematically, which is difficult since infiltration varies from irrigation to irrigation, year to year and furrow to furrow. Choosing measurement sites that represent an entire field is difficult, and direct measurements of infiltration are time-consuming. Therefore, supplying representative intake information for simulation techniques can be difficult. Walker and Skogerboe (1987) and Strelkoff and Clemmens (2007) summarize simulation procedures and list references. Walker (2003) developed the model SIRMOD which has found wide application in the design and operation of surface irrigation systems (Hornbuckle et al. 2003). Border Irrigation Border dike irrigation is used primarily for close-growing crops. The land is prepared by constructing parallel earthen dikes to contain the irrigation streams. The dikes are generally parallel to the greatest land slope. When this slope exceeds 4% for sod-forming crops, and 2% for other crops, the dikes may be angled to reduce slope in the direction of water flow. Slight cross slopes (up to 0.1%) are allowable if streams are sufficiently large and slopes in the direction of flow are sufficiently gradual. The irrigation stream is usually introduced from a single turnout at the upslope end of each strip and needs to be relatively large to cover the width of the border and to advance to the end in reasonable time. Stream size and application time are set so that recession rates will parallel advance rates to permit uniform depth of infiltrated water. For detailed information on border irrigation design, refer to the Natural Resources Conservation Service (NRCS) National Engineering Handbook (USDA 1997c). The WinSRFR model has been used extensively by NRCS in the design of border irrigation systems. (Bautista et al. 2006) Basin Irrigation Basins are used to irrigate close-growing crops, orchards, and vinyards. Each basin surface is leveled to eliminate cross slope and leave little or no slope in the direction of irrigation. Each basin is surrounded with dikes and may be any shape. To irrigate, a predetermined volume of water is ponded in the basin until infiltration is complete. This requires a high flow rate to be discharged from one or more inlets along one end of the basin (Dedrick et al. 1982). This system is capable of applying a relatively small depth of water because of the rapid advance. A variation, basin-furrow irrigation, can be used to irrigate row crops. The furrows are similar to those for sloping furrows, but they have no slope and are larger to accommodate the larger stream sizes. For detailed information on level

514

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

basin design, see the aforementioned NRCS National Engineering Handbook and Clemmens et al. (2007). The WinSRFR model (Bautista et al. 2006) is also available to model surface irrigation systems.

Land Leveling Earth-moving on irrigated fields to improve surface slope and smoothness is often referred to as land leveling. It establishes uniform field topography even though the entire slope may not be removed. Land leveling will produce uniform slopes in the direction of irrigationstream advance and may totally eliminate slope in level basins. Land planing, including laser leveling, removes minor irregularities in the surface just before planting. Land smoothing removes high and low areas but leaves the field with variable slope. Land planing and land smoothing improve conditions for salinity control by improving uniformity of surface irrigation. Thus, water for salt leaching can be applied more effectively and deep percolation from nonleaching irrigations can be minimized, restricting the water table rise. When soil is moved during land leveling to create a field of constant or zero slope, the depth of soil removed from some portions of the field may be a significant fraction of the rootzone depth, and subsoils that are highly saline or low in plant nutrients may be exposed. Earth-moving cuts may also lower the soil surface close to slowly permeable soil lenses or perched water tables. If the lowered soil surface is within the capillary fringe of a water table, preventing salt accumulation in the soil surface will be difficult. Thus, soil profiles and groundwater conditions should be carefully investigated before significant land leveling is begun. In situations that may lead to exposure of saline subsoils, surface soils may be removed and stockpiled before reconfiguring, then replaced on the field after leveling is completed.

Distribution of Applied Irrigation Water in Soil With surface irrigation, intake opportunity time and, thus, the infiltrated depth of water are normally greatest at the upstream end of each furrow, border, or basin. If applications merely replace the soil-water deficit at the upstream end of a furrow, the downstream end will be underirrigated. Refilling the entire profile at the downstream end causes excessive deep percolation at the upstream end. Furrow irrigation uniformity can be improved by using larger furrow streams for faster stream advance, or by reducing furrow length. Infiltrated water from furrows may not be distributed uniformly perpendicular to the furrows because of nonuniform soil conditions and vari-

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

515

able furrow geometry. On some soils, movement of infiltrated water from the furrow to the adjacent ridges is limited. This poses no problem for plants with well-developed root systems but may hamper seed germination or cause stress in young plants. In such cases, shallower furrows or longer water application times can be used to increase lateral water movement to the center of the seed bed. However, increased infiltration time may increase deep percolation losses, which can exacerbate problems in areas with shallow groundwater. If irrigation water is applied to all furrows equally, less infiltration will occur in furrows compacted by tractor traffic. The lateral distribution of infiltrated water after irrigation will vary from site to site, being less uniform for coarser textured soils and every-other-furrow irrigation. Distribution of infiltrated water is most uniform for border dike systems if the borders have low cross-slope and evenly sloping or slightly convex profiles in the direction of flow. Irrigation stream size and duration may need to be adjusted for the infiltration and flow resistance conditions that exist at the time of each water application. Water application can be highly efficient and uniform for level-basin and basin-furrow systems if leveling is precise and the application rate and volume are appropriate for the basin’s size and soil conditions. Irrigation streams must be large enough to inundate the basin’s surface in a fraction of the time needed for the desired net irrigation depth to infiltrate (Dedrick et al. 1982). The irrigation stream must be turned off when the applied volume of water is equal to the design irrigation application. Since no runoff can take place, the net infiltration will nearly equal the gross water application. Salt Accumulation Patterns In arid and semiarid irrigated agriculture, salts are present in both the irrigation water and in the soil where they are dissolved and carried in solution by infiltrated water. When water is adsorbed by roots or evaporated from the soil surface, the salts are left behind either as precipitates or concentrates in the soil water. When deep percolation occurs, the concentrated saline water leaches below the rootzone to the groundwater. Soil salinity may vary widely from the bottoms of furrows to the tops of beds. Capillary flow carries infiltrated water into the beds, where water use by plants increases the salt concentration and evaporation leaves salt deposits at the soil surface. The distribution will depend on the bed size and shape. In permanent bed systems, this accumulation of salt may be great enough by the end of an irrigation season to hamper seedling development in the following year unless leaching is provided. Figure 16-1 shows typical patterns of salt accumulation in beds between furrows. It

516

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 16-1. Typical salt accumulation patterns in ridge and bed cross sections under furrow irrigation. From Bernstein et al. (1955) and Bernstein and Fireman (1957).

illustrates seed locations where young plants can avoid the highest salt concentrations. Plowing or other tillage can redistribute this salt to allow continued crop production. Variation of salinity in the direction of furrows is generally unimportant if the infiltration is uniform enough to meet crop water needs. For border and basin irrigation in the absence of high water tables, water and salt flow downward and significant surface salt accumulations are unlikely. Salt distribution in basin furrows is much like that with graded furrows. Sprinkler Irrigation Types of systems Sprinkler systems can be categorized as set or mobile. With set systems, the sprinklers remain in a fixed location while applying water, whereas mobile systems move continuously, in a straight line (linear move) or a circle (center pivot) while irrigating. Mobile systems generally have a higher cost than set systems but require less labor to operate.

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

517

Distribution of applied water Pressure losses in sprinkler lateral lines cause different operating pressures at individual sprinklers and, thus, different discharge rates. Sprinkler heads at different elevations on fields of irregular topography will operate at different pressures. Pressure regulators, flow controlled, or appropriately sized nozzles for individual sprinkler heads can minimize differences in discharge rates due to elevation differences and pressure losses and provide uniform irrigation. Uniformity of sprinkler application, especially for set systems, is affected by both the water application pattern of each individual sprinkler head, the degree of overlap, and the wind conditions. Computer simulation programs are available to determine optimal sprinkler spacing for any given sprinkler distribution pattern. Windy conditions during sprinkler application distort the water distribution pattern. Set systems can be designed to maximize uniformity for any predominant wind condition, principally by selecting the proper type of sprinkler head and by correct spacing of sprinkler heads and laterals [refer to NRCS National Engineering Handbook USDA (1997b)]. Water applied to the soil surface uniformly may not infiltrate uniformly. If application rates exceed the soil intake capacity, some water will pond or flow across the soil surface to finally accumulate in low areas or run off the field. If application rates everywhere are less than intake capacity and total application amounts are less than the available soil water holding capacity, uniformity of stored soil water will be good even for nonuniform soils. However, this may lead to a deficit soil water condition that affects crop yield. Tillage or residue management practices may be used to maintain adequate infiltration conditions. Salt accumulation patterns Since well-designed, well-operated sprinkler systems apply water reasonably uniformly over the complete soil surface, they tend to leach salts evenly and surface salt accumulations are unlikely to occur. In general, as the depth of infiltrated water increases, the greater the resulting leaching. The variability in the salt distribution will be a result of the variability of applied water. Microirrigation System description Microirrigation refers to irrigation methods, such as surface drip, subsurface drip, bubbler, and microsprinkler (Lamm et al. 2007). A common feature of microirrigation systems is water delivery near plants through

518

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

a network of tubing with closely spaced, low-flow-rate emitters, and water typically infiltrating at the point of application. The soil volume wetted is therefore controlled by the number of application points, lateral movement of water in the soil, and the depth of application. Microirrigation systems are easily automated, making frequent, light water applications possible without high labor costs (Ayars and Phene 2007). A high level of management is needed because a large soil water reservoir is typically not maintained, and plants can be stressed if irrigations are improperly timed. Distribution of applied water Microirrigation systems do not cover the soil surface uniformly with applied water; instead, they apply water to small areas relative to the rooting pattern of the plants. On many fields the limited wetting of the surface soil is an advantage, as it inhibits weed growth and evaporation from the soil surface. Subsurface drip irrigation has the advantage of not wetting the soil surface and applying fertilizers and water close to the rootzone (Lamm and Camp 2007). This results in reduced weed growth and better use of fertilizers with a proportionate increase in water use efficiency (Ayars et al. 1999). The water distribution with closely spaced in-line emitters results in a reasonably constant wetted soil pattern in the direction of the irrigation line. These systems are especially well adapted for use with row crops, where the roots occupy the wetted areas and dry zones occur between rows. Microirrigation of perennial crops such as vineyards and orchards may require multi-lines and variable emitter spacing to apply water to the rootzone. The number of emitters will vary depending on the plant spacing and size and there may not be an overlap of wetted areas (Boman 2007). Salt accumulation patterns The movement of water in the soil wetted by a microirrigation system dictates the movement and relocation of salts. A gradient of water and salt develops between the emitter and the boundary of the wetted soil volume. Salts remain when the water evaporates or is absorbed by roots, and salt concentrations become highest at the soil surface and at the boundaries of the wetted soil volumes. As the irrigation frequency increases this becomes less of a problem (see Chapter 17). West and Merrigan (1979) concluded that when tomatoes are drip-irrigated with saline water, irrigation must be managed to provide enough wetted soil volume with low salt concentration to minimize contact between roots and zones of high salinity.

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

519

Irrigation Management Practices for Salinity Control Table 16-1 summarizes the most common irrigation methods, their applications, and their effects on salinity control. The following sections cover additional aspects of salt management. Crops Different plant cultivars have different salt tolerances (Maas 1990). The profitability of crop production can be increased by selecting the proper crop for the site’s salinity conditions (see Chapter 13). Meiri and Plant (1985) review practices that maximize yield under saline conditions. These include providing leaching to maintain rootzone salinity, proper placement of seeds relative to salt accumulation zones, changing planting density, and alternating water sources. These are discussed in the following sections. Management of different irrigation methods often depends on the crop’s characteristics. Temporary sprinklers are sometimes used to germinate and establish salt-sensitive crops. Sprinklers can apply small depths of water uniformly, keeping the seed bed adequately moist and salt-free (Robinson and Mayberry 1976). Surface irrigation is then used to grow the established crop. If the sprinkler irrigation water is saline (see Chapters 11 and 13), some crops may be adversely affected by salt deposits on leaves (Collier 1984; Maas 1985). Deciduous fruit trees are especially susceptible (Hoffman et al. 1980). The salts that accumulate on leaves are largely those that remain in solution in the intercepted water at the end of irrigation. As this water evaporates, the salts are concentrated and deposited on the foliage. Larger, less frequent irrigations may reduce the salt deposits relative to the volume of irrigation applied. Irrigation at night or another low-evaporation period also minimizes salt concentration and absorption. Completing irrigation with low-salinity water to wash salt off vegetation is also a possibility. Planting practices If microirrigation emitters are located near individual plants of perennial crops, salts tend to move away from the roots and concentrate in intermediate soil areas. To avoid problems with germination or salt stress on seedlings of annual crops, it is important to plant precisely where previous microirrigation has left low concentrations of salt. Planting seeds of furrow-irrigated crops on the sides of furrow ridges may keep seedlings out of the most saline soil zone (Fig. 16-1). An alternative is to apply a preplant irrigation during a fallow period that will move the salt below the germination zone (Ayars 2003).

TABLE 16-1. Factors Affecting Selection of Irrigation Method Under Saline Conditions 520

Water Application Method (1)

Application (2)

Pattern of Salt Accumulation (3)

Leaching Effectiveness (4)

Special Considerations (5)

Furrow

Row crops, low to medium infiltrationrate soils.

High in ridges between furrows, may increase in direction of slope if irrigations are nonuniform.

Effective leaching beneath furrow channels, salt left in ridges. Leaching requires more water than for methods with lighter, intermittent applications.

None

Corrugation

Close-growing crops.

Leaves saltier strips between corrugation channels unless entire field surface is inundated.

Similar to furrows, above.

None

Border dike

Close-growing crops.

Leaves salt in dikes that separate borders.

Areas between dikes are leached uniformly, but more water is required than for light, intermittent applications.

None

Sprinkler: Set

Most crops, all but very fine-textured soils.

No salt concentrations in root zone if system is designed and managed properly.

Uniform leaching. Can be used to leach salt accumulations left by other irrigation methods.

May encourage disease in sensitive crops, e.g., beans. Salty irrigation water may leave harmful deposits on leaves.

Sprinkler: Mobile

Most crops, except trees, vines. Can be used to irrigate fields on rolling topography.

No salt concentrations in root zone if system is designed and managed properly.

Uniform leaching. Same as for set sprinklers.

None

Microirrigation (surface drip, subsurface drip, micro-sprinklers)

Because of high initial costs, used mostly for high-value crops or crops with high irrigation labor costs.

Salt concentrates at outer fringes of the soil mass wetted by each emitter.

Soil mass wetted by each emitter is well-leached. Difficult to leach all soil to depth of root zone.

When automated for light, frequent irrigations, saline water can be used because low matric stress compensates for osmotic stress.

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

521

Tillage Deep plowing can redistribute salt in the soil profile. The practice should be evaluated on a small land area before plowing entire fields. Soil layers that prevent leaching water from flowing downward can be chiseled or ripped to facilitate leaching. Some minimum tillage techniques allow furrow ridges formed for one crop to remain undisturbed for the next. Planting on such ridges may place seed in the most highly saline zone of the soil profile. Seeds should be planted on sides of ridges to avoid the most saline zones. Also, ridge splitting (a minimum-tillage method of planting deep in the ridge) may be used to provide a suitable seed bed for row crops. Organic residue left on the soil surface by minimum-tillage practices presents a problem for furrow irrigation. It slows furrow stream advance, increasing longitudinal nonuniformity. Surge flow may be useful in furrows with large amounts of residue. Surge-caused reductions in intake rates tend to speed stream advance, compensating for the slowing caused by the residue (Evans et al. 1987). Irrigation scheduling Irrigation scheduling is the process of determining when and how much to irrigate. It is important when management includes salinity considerations. Proper timing will avoid low levels of soil water that concentrate salts in the soil solution to detrimental levels. Frequent water applications maintain low matric water stress, which may compensate for the osmotic stresses caused by saline water. Frequent irrigations also keep the salts moving through and away from the root mass. High-frequency irrigations are generally small and can seldom be applied as uniformly with surface methods as with sprinklers or microirrigation. Shainberg and Shalhevet (1984) reviewed the effect of the frequency of saline water irrigation on yield and concluded that higher frequencies result in higher yields. Scheduling each irrigation application amount allows the depth of leaching water to be accurately determined and prevents excessive deep percolation. Jensen et al. (1990) and Allen et al. (1998) presented methods for estimating rates of evapotranspiration (ET) of irrigated crops and scheduling individual irrigations. Limiting irrigation applications to amounts necessary for replacement of rootzone water and leaching also helps to minimize deep percolation and build-up of the water table in areas with shallow groundwater. As water management improves, irrigation schedules will have to account for in-situ use of shallow groundwater by crops (Ayars and Hutmacher, 1994; Ayars et al. 2000). Microirrigation systems have helped to maintain suitable water potentials in the plant rootzone, even with saline irrigation water. Irrigation intervals of several days to allow for internal drainage are unnecessary because large soil volumes are not saturated. While frequent irrigations

522

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

may allow the use of saline irrigation water, irrigators should be aware of the hazards of omitting irrigations. Matric stress may occur quickly and, added to the osmotic stress, cause immediate crop damage. If microirrigation causes high salt concentrations to accumulate near the soil surface, unexpected rainfall can move the salt down into the root mass. Irrigations should be scheduled during or after rains to leach the salts out of the rootzone before damage to the crop occurs (Hanson and Bendixen 1995). Water Measurement To follow irrigation scheduling recommendations and irrigate efficiently, irrigation water applications should be measured. Sprinkler, level basin, and microirrigation systems can usually be designed so that none of the applied water runs off the irrigated area. This simplifies water measurement because the total water volume applied per set, divided by the surface area, gives the average net depth of application. For surface irrigation on sloping fields, the net depth of application can be calculated by measuring the volumes applied to and running off each set. A less accurate alternative is to measure the water applied and multiply this by a water application efficiency determined from past measurements or experience. Numerous types of weirs, flumes, and meters exist for measuring openchannel or pipeline flows with sufficient accuracy for irrigation management. See Bureau of Reclamation (2001) and Replogle and Kruse (2007). Leaching A wide range of waters containing dissolved salts can be used for leaching. With waters of higher salinity, larger applications are needed to maintain a salt balance in the crop rootzone. The use of sodic water [sodium adsorption ratio (SAR) 15] is not recommended, except on highly sandy soils, since such water may damage soil structure and reduce permeability, making continued leaching difficult. Because the electrical conductivity (EC) of the water increases, it is possible to use sodic water on a wide range of soil types without adverse effects. Chapter 12 discusses in detail the control of salinity in the rootzone. Chapter 21 reviews chemical amendments that aid in correcting salinity and sodicity problems. The volume of water needed for a given degree of leaching may be greater for furrow and other surface irrigation methods than for sprinkler irrigation. The minimum depth of water that can be applied uniformly by most surface methods is several times greater than the minimum for sprinkler applications. Hoffman (1980) suggests amounts for different degrees of leaching (Fig. 16-2). Small, intermittent applications are more effective for leaching than continuous flooding applications. The unsaturated flow of soil water achieved by the former more effectively removes

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

523

FIGURE 16-2. Depth of water per unit depth of soil required to leach saline soil by continuous or intermittent ponding. From Hoffman (1980).

salt from the smaller soil pores. Much of the flow achieved by the latter occurs in large pores and root channels. With level basins, water can be applied uniformly over the entire surface of the soil. If the soil-water deficit in the rootzone at the time of irrigation is known accurately and is uniform, the portion of the application that will cause leaching can be precisely controlled. Border irrigations require greater care to adjust flows to varying infiltration rates and application depths. Sprinkler irrigation applies water reasonably uniformly over the soil surface and application depths are easily controlled. However, application rates must not exceed the soil intake capacity or surface runoff will occur. Sprinklers can be operated long enough to replace the moisture deficit of the rootzone and apply the desired additional depth for leaching.

524

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Some types of microirrigation systems, especially those using minisprinklers, wet most of the soil surface and move salts downward. For most types of microirrigation, however, salt moves and accumulates in a more complex manner. The type and spacing of emitters, soil types, and crop rooting patterns vary, making general recommendations for leaching difficult. Salts should be carefully monitored by direct measurements (soil sampling), indirect methods, such as electromagnetic induction (such as with the Geonics EM-38, a small portable conductivity meter) (Corwin and Lesch 2005; Kaffka et al. 2005), or by frequent, careful observations of crop conditions. Long-term use of microirrigation may create large, irregularly distributed salt accumulations. If a crop with a different row spacing or rooting pattern is to be produced on such an area, special leaching methods may be needed. Similarly, nonuniform surface irrigation applications may cause zones of salt concentrations that are inadequately leached. If salt accumulation is observed, leaching maybe accomplished by an annual preplant irrigation using either sprinklers or surface methods (Ayars 1999).

DRAINAGE AND SALINITY CONTROL Drainage of Irrigated Fields Applying irrigation water to soil disturbs the natural hydrologic balance of the soil profile. Since irrigation water cannot be applied with complete uniformity, some water will percolate below the rootzone. If the deep percolation rate is less than the natural drainage capacity of the soil, a water table will not develop and the net movement of salt in the profile will be downward. If the deep percolation rate exceeds the natural drainage capacity of the soil, shallow groundwater will develop and the water will rise in response to the volume of deep percolation. Soil layers with low permeability may also restrict percolating water from flowing downward and cause perched water tables. When the water table is too close to the soil surface in an arid region, water and salt will be carried upward by capillary action and the upper soil profile and surface may become salinized as the water evaporates. If enough irrigation water is applied annually for net downward movement of water through the profile, a favorable salt balance in the rootzone can exist even in the presence of a high water table. The shallower the water table, the more care needs to be taken with water applications to ensure a net downward movement. If the natural drainage capacity is so limited that normal deep percolation of irrigation water causes the water table to rise close to the soil surface, a subsurface drainage system must be installed. Poor irrigation uniformity

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

525

and excessive deep percolation increase the amount of water that drainage systems must remove. Interception of Lateral Flow If the subsurface geology contains shallow impermeable layers and embedded geologic streams, underground water may be forced to the surface or transmitted to adjacent fields. Interceptor drains are installed perpendicular to the subsurface flow, upstream of the land to be protected, to collect excess lateral seepage water before the water table rises too close to the surface. By intercepting subsurface soil water that exceeds the soil’s natural drainage capacity, the soil may more easily carry the remaining water, helping to maintain the water table at an acceptable depth. The design and location of interceptor drains requires local topographic and groundwater surveys to identify sources of water and suitable outlets (Bureau of Reclamation 1993). Drains Excess water from irrigation and rainfall is removed by both surface and subsurface drains. In irrigated areas the most common type of subsurface drains consists of perforated pipes installed in a regularly spaced pattern. They are placed at a depth of 1.5 to 2 times the desired midpoint water table depth, with the spacing determined by the soil characteristics and the deep percolation schedule. The amount of water to be removed is the excess between the total infiltrated and the amount stored in the rootzone for use by the crop. Water provided for leaching, nonuniform applications, and overirrigation will increase the drainage requirement. Steady or non-steady-state (transient) methods can be used to calculate the spacing of relief drains in arid areas. Two techniques of designing relief drain depth and spacing, the Donnan steady-state method and the U.S. Bureau of Reclamation transient method (Bureau of Reclamation 1993), must be used with corrections for equivalent depth to barrier. Generally, the transient method better represents irrigated conditions. As environmental concerns develop with regard to the discharge of saline water, it is necessary to consider water quality in the design of subsurface drainage systems (Ayars et al. 1997). The design criteria for the midpoint water table depth and the drain depth should be modified to increase crop water use from shallow groundwater, thus reducing drainage discharge and salt loading of surface water. Controls should also be installed to control water table position (Christen and Ayars 2001). Pumped drainage wells, closely spaced so that their areas of influence overlap, can be used as relief drains to lower water tables if subsurface geologic conditions are suitable (Luthin 1978).

526

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Water Table Depths The water table should be deep enough to allow adequate aeration in the active plant rootzone, with deep-rooted crops requiring a deeper water table than shallow-rooted crops. The water table should be deeper in clay soils than in sandy soils, as the capillary rise of water is greater in the clay soils. Salt deposits on the soil surface or crops that appear to be suffering from poor soil aeration may indicate a water table that is too shallow. The desired water table depth varies with crop, soil, and spacing of drains (Bureau of Reclamation 1993). In arid regions, drains have traditionally been placed at approximately 1.8 m deep and designed to remove 2.0 mm to 5.0 mm of water per day to prevent waterlogging and salinization of the rootzone. Water quality concerns and improved salt management practices have required revised design criteria that are resulting in recommendations for shallower drain placement and the addition of controls to regulate the water table position (Christen and Ayars 2001). Drainage of Nonirrigated, Saline Fields Nonirrigated cropped areas may need subsurface drains, depending on the annual amount and distribution of rainfall, soils, and plant water use. Seeps or saline areas may develop when the amount of water present exceeds the natural drainage capacity of the soil. For example, if an area of natural vegetation is converted to cropland and then fallowed, excess water from precipitation that is not used by the crops may cause drainage and salinity problems. Saline seeps sometimes appear on nonirrigated, arid lands when precipitation exceeds ET during the process of soil fallowing. Excess water percolates down to the water table and then moves laterally to lower soil surfaces, dissolving salts in the material through which it flows. Subsurface conditions eventually force the flow to the surface, where it evaporates and results in a saline seep. Chapter 18 covers saline seep management. Control of Upward Flow from Confined Aquifers In stratified alluvial soils, confined aquifers may contain water under artesian pressure. When a well punctures the confining layer, the water may flow without pumping. Unbroken confining layers may allow slow, vertically upward seepage over large areas, creating high water tables and saline areas. Some form of relief drainage must be used to reverse the net annual flow direction in the rootzone from upward to downward. In confined aquifers it may be possible to drill wells into the aquifer and pump them to relieve artesian pressure. Such action will reverse the

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

527

direction of water flow in the overlying soil if the pressure head is lowered well below the soil surface. Gravity relief wells have been proposed to reduce artesian pressure, but they must be closely spaced and discharge into the bottom of deep drains. If the confining layer is impermeable, the artesian pressure may have no effect other than to prevent water from moving downward, and pumped or gravity relief wells will be ineffective. In this case, relief drains above the confining layer may solve high-water-table problems. Drainage Wells Wells in unconfined aquifers can be pumped to lower the water table. Pumped wells have continuing operation and maintenance costs. If the aquifer water is of good quality, it may be used for irrigation or domestic purposes and the costs of lowering the water table will be offset by other benefits. The deep water table that results from pumping will eliminate salinity and drainage problems if the overlying soil has no layers of low permeability. Water Extraction by Plants Healthy plants rapidly remove excess water from the soil. Normal rates of ET are two to three times greater than subsurface drainage rates in arid areas. Plants are not normally used to remediate waterlogging, because crops may suffer physiological root damage before they can remove enough water to aerate the rootzone adequately. However, integrated water management of irrigation and drainage systems should include in-situ use of shallow groundwater whenever possible (Ayars et al. 2000). In-situ use will reduce the total irrigation requirement and reduce the total drainage volume and salt load from an irrigated area (Christen and Ayars 2001). Significant quantities of water may be extracted from the shallow groundwater over a wide range of groundwater salinity (Ayars et al. 2006b). Disposal of Drainage Effluent Concerns and regulations about water quality may prevent the discharge and disposal of drainage effluents in natural or man-made channels. Tanji et al. (1985) and Wu et al. (1999) describe ways to design and manage on-farm evaporation ponds. Hanson (1984) discusses general solutions to drainage water quality and disposal, especially in those areas where off-farm disposal is impossible. A suitable outlet must be located before undertaking any drainage work. Traditionally, drainage water was discharged to a gravity outlet, such as an open drain, river, or evaporation

528

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

basin. Environmental concerns have significantly limited the options for on-farm drainage water disposal and, as a consequence, methods were developed to provide on-farm disposal options. One option is sequential biological concentration of drainage water (Heuperman et al. 1996; Cervinka et al. 1999; Blackwell et al. 2005). Sequential Biological Concentration Drainage water disposal is a significant problem for irrigated areas throughout the world. If irrigation is to be sustained, methods will have to be developed to dispose of saline drainage water in an environmentally friendly matter (Oster and Wichelns 2003). One proposed solution is sequential biological concentration of the drainage water coming from irrigated agriculture. This is a process whereby drainage water is collected and applied to successively more salt-tolerant crops until the concentration reaches a level that is no longer suitable for irrigation. At this point the drainage water then is disposed of in the evaporation facility. This concept is called integrated on-farm drainage management (IFDM) in the United States (Cervinka et al. 1999; Ayars and Basinal 2005) and serial biological concentration in Australia (Heuperman et al. 1996; Blackwell et al. 2005). One limitation of the theory is that increased salt concentration in the deep percolate with each reuse has not been observed in some experimental situations (Bethune et al. 2004). This occurs when the subsurface drainage system captures significant quantities of regional saline groundwater that mask the concentrating effect of crop water use (Jury et al. 2003). The successful operation of these systems incorporates knowledge of improved irrigation management, redesign of drainage systems, management of salinity, selection of crops, and use of nontraditional crops. The major components for successful operation will be minimizing the throughput of drainage water and maximizing the area of economic crops being grown in the system (Wichelns and Oster 2006). These systems represent an interim solution until other alternatives can be developed to sustain irrigated agriculture.

MANAGEMENT OF SHALLOW GROUNDWATERS Subirrigation is practiced as an active form of water table management in humid areas (Fouss et al. 2007). In arid and semiarid areas, groundwater management historically has been passive and an almost incidental part of irrigation schemes. Groundwater management is now considered a means to use water more effectively in arid and semiarid regions (Ayars et al. 2006a). Research has shown that crop production can be maintained with water of higher salinity than previously thought

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

529

possible (van Schilfgaarde et al. 1974; Rhoades 1984; Ayars 2003). Problems with water disposal have intensified the need to minimize the volume of drainage water, and other demands for water have increased the need for agriculture to maximize its use of all available water. The salinities of the soil, groundwater, and irrigation water influence the proper management of shallow groundwater. Operation of Shallow Groundwater Systems Shallow groundwater management requires a systems approach that includes the design and operation of irrigation systems and consideration of irrigation and groundwater qualities, soil salinity limits, crop tolerances, and cropping patterns. In arid areas, the objectives of drainage are to maintain favorable oxygen availability and salinity levels in the crop rootzone. Traditionally, drains in arid and semiarid areas were placed as deep as possible, often in excess of 2 m, and no attempt was made to control the rate of decline or position of the water table except to maintain some minimum depth to the water table at the midpoint between drains (Bureau of Reclamation 1993). Doering et al. (1982) proposed a shallow water table control concept for drainage design that allows plants to use some water from the water table. The water table control is passive, a function of drain placement and irrigation management. Based on studies of the use of water by corn and sugar beets, Benz et al. (1981) found that drain depths of 1.5 m adequately protect the crop from high water tables and provide a substantial part of the crop water requirement. Maintaining the water table at a depth of 1 m resulted in maximum yields of corn, sugar beets, and alfalfa. Ayars (1996) demonstrated water table management using a drainage system that had drain lines laid perpendicularly to the direction of the field’s maximum surface slope. Each lateral line was controlled by a valve, as was the outlet to the sump. The drain orientation prevented excessive buildup of the water table at the field’s low end. Salt Accumulation and Distribution In many arid areas with shallow groundwater, salinity increases with depth to a value equal to the groundwater salinity, as a result of leaching. The salinity increases in the rootzone over the growing season (Fig. 16-3). The total salt increase is a function of the water extracted from groundwater, effective leaching during irrigation, and salt added by the irrigation water. A preseason salinity profile can be restored by leaching during a fallow period. If not leached, salt will accumulate in the profile until plants can no longer grow. Rainfall and preplant irrigation can leach areas where drains are installed.

530

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 16-3. Salinity changes in soil profile over a 2-year period. From Ayars (2003).

Management Practices Crop use of saline subsurface water The depth of the water table, quality of the groundwater, and salt tolerance of the plants determine the amount of controlled, shallow groundwater available to a crop (Ayars et al. 2006b). For instance, a saltsensitive crop grown in the presence of groundwater at a depth of 1 m and with an EC () of 8 dS/m probably would extract little groundwater and might suffer from the salt carried up through the profile by water that evaporates from the soil surface. The more closely the crop’s rooting depth and salt tolerance match the groundwater‘s depth and salinity, the more likely plants are to extract groundwater. Maas (1990) tabulated salt tolerances for a wide range of crops (see Chapter 13). Ayars and Hutmacher (1994) found that the groundwater’s maximum direct contribution to crop water requirement was a function of both water quality and water table depth. Management of soil salinity The change in profile salinity caused by the crop’s water use from groundwater can be approximated for a particular situation if the initial average soil salinity, irrigation water salinity, and groundwater salinity are known. Knowing the crop water requirement and proportioning

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

531

water use between irrigation and groundwater, the final average soil profile salinity may be estimated by: (16-1) where F and 1 represent the final and initial average ECs of the soil solution in the profile, D1 and DF are the initial and final depths of water stored in the profile at saturation, i and gw represent the conductivities of the irrigation and groundwater, and Di and Dgw are the depths of irrigation and groundwater used. It is assumed that no leaching occurs, no salt precipitation or dissolution takes place, and the relation between salt concentration and EC is approximately linear (Hoffman and van Genuchten 1983). A sample calculation for a 1.2-m profile with a 55% volumetric water content at saturation, 1  7 dS/m, i  0.2 dS/m, gw  8 dS/m, Di  397 mm, and Dgw  174 mm yields a final salinity, F  9.6 dS/m (although the average over the growing season would be 8.3 dS/m). Using this type of calculation, a manager can select crops and develop a strategy to maximize use of groundwater without a loss in productivity due to salinity. The average soil profile salinity should remain below the salttolerance level of the crop. If groundwater supplies part of the crop water requirement, the concentration of the applied water is determined on a volume-weighted basis. It is assumed that the leaching requirement, Lr, can be estimated by Lr  a/D*, where a  the weighted EC of the water available to the crop, including irrigation water supplied and groundwater used, and D* is the desired soil salinity value (Hoffman 1985). In the previous example, a would equal (i Di  gw Dgw)/(Di  Dgw)  2.6 dS/m, not simply the 0.2 dS/m in the irrigation water. If the required D* was 7.0, Lr  2.6/7  0.37. As the groundwater becomes more saline, the potential for soil salinization appears to increase rapidly if significant quantities of groundwater are used. If rainfall and preplant irrigation water are included, then the volume-weighted concentration of applied water decreases and the Lr decreases. To avoid raising the water table to the extent that the crop is damaged, leaching should take place during a fallow period or early in the growing season when the crop’s root system is shallow and the water demand is small (Rhoades 1984). Chapter 12 provides an in-depth discussion of leaching. Irrigation scheduling The depth of the water table in uncontrolled areas is a function of lateral inflows and deep percolation from irrigation and rainfall. Good

532

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

management of irrigation will allow the crop to use some shallow groundwater (Ayars et al. 2000). Scheduling irrigation by use of traditional water balance methods is difficult if shallow groundwater exists because quantifying the groundwater contribution to crop ET is a problem. Most scheduling programs assume no contribution from the groundwater and assume irrigation and precipitation to be the only sources of water in calculating soil water depletion. When groundwater contributes to the crop water requirement, the rootzone water content decrease is slower than estimated by scheduling programs. Scheduling programs that specify the allowable depth of water to be depleted before irrigation often do not consider the osmotic stress that takes place in a saline environment. Excessive stress levels to the crop can result as the soil dries and salts in the soil water become more concentrated. Howell et al. (1984a) developed ways to use plant stress, either matric, osmotic, or the sum of both, to indicate the time to irrigate. Leaf-water potential and crop-water stress index (CWSI) provide the bases for two such methods (Idso et al. 1982). The CWSI method uses plant canopy temperature along with the vapor pressure deficit to indicate stress. Howell et al. (1984b) developed values of CWSI to use in scheduling irrigations for cotton. The needed depth of application at the time of irrigation can be estimated from soil water content as determined by neutron probe, gravimetric analysis, or another method. Irrigation scheduling and water tables can be managed jointly for either the maximum contribution from the shallow groundwater, little or no groundwater contribution, or some intermediate value of groundwater use as defined by the user. In the case of maximum contribution from the groundwater, the salt tolerance of the crop and the availability of shallow groundwater limit total water use. To maximize availability, the water table position is controlled. Its elevation is carefully monitored to ensure an adequately aerated zone. Water table depths of 1 m to 1.5 m have provided the maximum contribution from groundwater for a wide range of soil types. The shallower water table depth applies to sandy, nonsaline soils (Ayars et al. 2006). With maximum groundwater contributions, irrigations can best be scheduled using plant-based measures. The optimum time to irrigate is at the highest stress level that does not reduce yield. Irrigation at lower stress levels would result in more frequent irrigations, more deep percolation, and less contribution to ET from the water table. The depth of irrigation water to apply is estimated from soil-water measurements. The time of the first irrigation is critical for unrestricted plant growth and root development (Ayars and Schoneman 1986). Soil salinity measurements at the end of the irrigation season can be used to calculate the leaching required for re-establishing a favorable soil salinity profile for the next

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

533

growing season. Rainfall plus preplant irrigation can provide necessary leaching. Most types of irrigation systems allow this management. To obtain little or no groundwater contribution, management is somewhat simpler. Only irrigation and rainfall supply the crop water requirement. Deep percolation is minimized. This can be done most easily by using high-frequency irrigation to supply water used by the crop, thus preventing significant contribution from the water table. Periodic leaching during the season will prevent the build-up of salts. An automated irrigation system that can provide highly uniform applications will be required, with solid set or mobile sprinklers and drip-irrigation systems being preferred. Lack of data on the plant’s temporal extraction of groundwater presents an obstacle to obtaining an intermediate amount of groundwater use, which can achieved on a seasonal basis by eliminating the final irrigation of the season. Hutmacher et al. (1986) found that cotton and wheat grown in the presence of a shallow saline water table did not suffer a reduction in yield when the last irrigation of the season was eliminated.

SUMMARY There are a number of commonly available irrigation and drainage systems in use, and for each there are management options that help prevent or correct salinity problems and minimize water table build-up. Management options involve variations in the method, timing, depth, and magnitude of irrigation, as well as crop selection. There are interactions among these variables, and thus an optimal irrigation regime requires a systematic and comprehensive approach. For example, the use of drainage wells and plants that intercept and take up saline drainage water can be comanaged to accomplish the objective of minimizing groundwater level rise. Similar coordinated decisions can be made for management decisions, such as the reuse of drainage water and various salt tolerant crops. Regardless of the irrigation and drainage method, drainage and management of shallow water tables are critical elements in salinity management.

REFERENCES Allen, R. A., Pereira, L. S., Raes, D., and Smith, M. (1998). Crop evapotranspiration: Guidelines for computing crop water requirements, FAO Irrigation and Drainage Paper 56, Food and Agriculture Organisation of the United Nations, Rome. Ayars, J. E. (1996). “Managing irrigation and drainage systems in arid areas in the presence of shallow groundwater: Case studies.” Irrig. and Drain. Sys., 10, 227–244.

534

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1999). “Integrated management of irrigation and drainage systems,” in Water management, purification and conservation in arid climates, Vol. 1, M. F. A. Goosen and W. H. Shayya, eds., Technomic, Lancaster, Pa., 139–164. ———. (2003). “Field crop production in areas with saline soils and shallow saline groundwater in the San Joaquin Valley of California.” J. Crop Prod., 7, 353–386. Ayars, J. E., and Basinal, L. (2005). A technical advisor’s manual: Managing agricultural irrigation drainage water: A guide for developing integrated on-farm drainage management systems, Center for Irrigation Technology, California State University, Fresno, Calif. Ayars, J. E., Christen E. W., and Hornbuckle, J. W. (2006a). “Controlled drainage for improved water management in arid regions irrigated agriculture.” Agric. Water Mgmt., 86(1–20), 128–139. Ayars, J. E., Christen E. W., Soppe, R. W. O., and Meyer, W. S. (2006b). “Resource potential of shallow groundwater for crop water use: A review.” Irrig. Sci., 24, 147–160. Ayars, J. E., Grismer M. E., and Guitjens, J. C. (1997). “Water quality as design criterion in drainage water management system.” J. Irrig. and Drain. Eng., 123, 148–153. Ayars, J. E., and Hutmacher, R. B. (1994). “Crop coefficients for irrigating cotton in the presence of groundwater.” Irrig. Sci., 15(1), 45–52. Ayars, J. E., Hutmacher, R. B., Schoneman, R. A., Soppe, R. W., Vail, S. S., and Dale, R. (2000). “Realizing the potential of integrated irrigation and drainage water management for meeting crop water requirements in arid and semi-arid areas.” Irrig. and Drain. Sys., 13, 321–347. Ayars, J. E., and Phene, C. J. (2007). “Automation,” in Microirrigation for crop production: Design, operation, and management, F. R. Lamm et al., eds., Elsevier, Amsterdam, 259–284. Ayars, J. E., Phene, C. J., Hutmacher, R. B., Davis, K. R., Schoneman, R. A., Vail, S. S., and Mead, R. M. (1999). “Subsurface drip irrigation of row crops: A review of 15 years of research at the Water Management Research Laboratory.” Agric. Water Mgmt., 42, 1–27. Ayars, J. E., and Schoneman, R. A. (1986). “Use of saline water from a shallow water table by cotton.” Trans. ASAE, 29(6), 1674–1678. Baustista, E., Schlegel, J. L., Strelkoff, T., Clemmens, A. J., and Strand, R. J. (2006). “An integrated software package for simulation, design, and evaluation of surface irrigation systems.” Proc., World Water and Environmental Resources Congress, EWRI/ASCE, Reston, Va. Benz, L. C., Doering, E. J., and Reichman, G. A. (1981). “Water table management saves water and energy.” Trans. ASAE, 24(4), 995–1001. Bernstein, L., and Fireman, M. (1957). “Laboratory studies on salt distribution in furrow-irrigated soil with special reference to the pre-emergence period.” Soil Sci., 83, 249–263. Bernstein, L., Fireman, M., and Reeve, R. C. (1955). Control of salinity in the Imperial Valley, California, USDA-ARS-41-4, USDA, Washington, D.C. Bethune, M. G., Gyles, O. A., and Wang, Q. J. (2004). “Options for management of saline groundwater in an irrigated farming system.” Aust. J. Exp. Agric., 44, 181–188.

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

535

Blackwell, J., Jayawardane, N., Biswas, T., and Christen, E. W. (2005). “Evaluation of a sequential biological concentration system in natural resource management of a saline irrigated area.” Aust. J. Water Resour., 9(2), 169–176. Boman, B. J. (2007). “Microsprinkler irrigation” in Microirrigation for crop production: Design, operation, and management, F. R. Lamm et al., eds., Elsevier, Amsterdam, 575–608. Bureau of Reclamation. (1993). Drainage manual, 3rd ed., U.S. Department of the Interior, Bureau of Reclamation, Washington, D.C. ———. (2001). Water measurement manual, U.S. Department of the Interior, Bureau of Reclamation, Washington, D.C. Cervinka, V., Diener, J., Erickson, J., Finch, C., Martin, M., Menezes, F., Peters, D., and Shelton, F. (1999). Integrated system for agricultural drainage management on irrigated farm land, Report 4-FG-20–11920, U.S. Bureau of Reclamation, U.S. Department of the Interior, Washington, D.C. Christen, E. W., and Ayars, J. E. (2001). Subsurface drainage system design and management in irrigated agriculture: Best management practices for reducing drainage volume and salt load, CSIRO Land and Water, Technical Report 38-01, CSIRO, Clayton South, Victoria, Australia. Clemmens, A. J., Walker, W. R., Fangmeier, D. D., and Hardy L. A. (2007). “Design of surface systems,” in Monograph on irrigation systems design and management, G. J. Hoffman and R. E. Evans, eds., ASABE, St. Joseph, Mich., 499–531. Collier, F. W. (1984). “The use of saline water in irrigation,” in Irrigation, drainage and flood control, K. K. Tanji, ed., State-of-the-Art Publication No. 3, International Commission on Irrigation and Drainage, New Delhi, India, 203–229. Corwin, D. L., and Lesch, S. M. (2005). “Apparent soil electrical conductivity measurements in agriculture.” Comput. Elect. Agric., 46, 11–43. Dedrick, A. R., Erie, L. J., and Clemmens, A. J. (1982). “Level-basin irrigation,” in Advances in irrigation, D. Hillel, ed., Academic Press, London, 105–145. Doering, E. J., Benz, L. C., and Reichman G. A. (1982). “Shallow water-table concept for drainage design in semiarid and sub-humid regions,” in Advances in drainage, Proc., ASAE 4th National Drainage Symposium, ASAE, St. Joseph, Mich., 34–41. Evans, R. G., Aarstad, J. S., Miller, D. E., and Kroeger, M. W. (1987). “Crop residue effects on surge furrow irrigation hydraulics.” Trans. ASAE, 30(2), 424–429. Fouss, J. L, Evans, R. W., Ayars, J. E., and Christen, E. W. (2007). “Water table control and shallow groundwater utilization,” in Management of farm irrigation systems, G. J. Hoffman et al., eds., ASABE, St. Joseph, Mich., 684–724. Hanson, B. R. (1984). “Effects of increasing drainage in the San Joaquin Valley.” Calif. Agric., 38(10), 40–41. Hanson, B. R., and Bendixen, W. E. (1995). “Drip irrigation controls soil salinity under row crops.” Calif. Agric., 49, 19–23. Heuperman, A. F., Heath, J., and Greenslade, R. (1996). “Serial biological concentration of salts: A management system for saline drainage effluent,” in Managing environmental changes due to irrigation and drainage, 16th ICID Congress, ICID-CIID, Cairo, Egypt. Hoffman, G. J. (1980). “Irrigation management: Salinity control,” in Irrigation challenges of the 80’s, Proc., ASAE 2nd National Irrigation Symposium, ASAE, St. Joseph, Mich., 166–175.

536

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1985). “Drainage required to manage salinity.” J. Irrig and Drainage Div., ASCE, 111(IR3), 199–206. Hoffman, G. J., Ayers, R. S., Doering, E. J., and McNeal, B. L. (1980). “Salinity in irrigated agriculture,” in Design and operation of farm irrigation systems, M. E. Jensen, ed. ASAE, St. Joseph, Mich., 145–188. Hoffman, G. J., and van Genuchten, M. T. (1983). “Soil properties and efficient water use: Water management for salinity control,” in Limitations to efficient water use in crop production, H. M. Taylor, W. R. Jordan, and T. R. Sinclair, eds., ASA, Madison, Wisc., 73–85. Hornbuckle, J. W., Christen, E. W., and Faulkner, R. D. (2003). “Improving the efficiency and performance of furrow irrigation using modeling in South-Eastern Australia. Workshop on improved irrigation technologies and methods: R&D and testing,” in Proc., 54th Executive Council of ICID and 20th European Regional Conference, Montpellier, France, September 14–19. Howell, T. A., Hatfield, J. L., Rhoades, J. D., and Meron, M. (1984a). “Response of cotton water stress indicators to soil salinity.” Irrig. Sci., 5, 25–36. Howell, T. A., Hatfield, J. L., Yamada, H., and Davis, K. R. (1984b). “Evaluation of cotton canopy temperature to detect crop water stress.” Trans. ASAE, 27, 84–88. Hutmacher, R. B., Ayars, J. E., Schoneman, R. A., and Vail, S. S. (1986). Furrow irrigation management of cotton, wheat and sugar beets in the presence of a shallow, saline water table, Annual Report, Water Management Research Laboratory, Fresno, Calif., 3–4. Idso, S. B., Reginato, R. J., and Farah S. M. (1982). “Soil and atmosphere-induced plant water stress in cotton as inferred from foliage temperature.” Water Resour. Res., 18, 1143–1148. Jensen, M. E., Robb, D. C. N., and Franzoy, C. E. (1970). “Scheduling irrigations using climate-crop-soil data.” J. Irrig. and Drain. Div., ASCE, 96, 25–38. Jury, W. A., Tuli, A., and Letey, J. (2003). “Effect of travel time on management of a sequential reuse drainage operation.” Soil Sci. Soc. Amer. J., 67, 1122–1126. Kaffka, S. R., Lesch, S. M., Bali, K. M., and Corwin, D. L. (2005). “Site-specific management in salt-affected sugar beet fields using electromagnetic induction.” Comput. Elect. Agric., 46, 329–350. Lamm, F. R., Ayars, J. E., and Nakayama, F. S., eds. (2007). Microirrigation for crop production: Design, operation, and management, Elsevier, Amsterdam. Lamm, F. R., and Camp, C. R. (2007). “Subsurface drip irrigation,” in Microirrigation for crop production, F. R. Lamm, J. E. Ayars, and F. S. Nakayama, eds., Elsevier, Amsterdam, 473–552. Lentz, R. D. (2003). “Inhibiting water infiltration with polyacrylamide and surfactants: Applications for irrigated agriculture.” J. Soil and Water Cons., 58(5), 290–300. Luthin, J. N. (1978). Drainage engineering, Robert E. Krieger Publishing Co., Huntington, N.Y. Maas, E. V. (1985). “Crop tolerance to saline sprinkling water.” Plant Soil, 89, 273–284. ———. (1990). “Crop salt tolerance,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 262–304. Meiri, A., and Plant, Z. (1985). “Crop production and management under saline conditions.” Plant Soil, 89(1/3), 253–271.

ON-FARM IRRIGATION AND DRAINAGE PRACTICES

537

Oster, J. D., and Wichelns, D. (2003). “Economic and agronomic strategies to achieve sustainable irrigation.” Irrig. Sci., 22, 107–120. Replogle, J. A., and Kruse, E. G. (2007). “Delivery and distribution systems,” in Design and operation of farm irrigation systems, G. J. Hoffman and R. E. Evans, eds., ASABE, St. Joseph, Mich., 347–391. Rhoades, J. D. (1984). “New strategy for using saline waters for irrigation,” in Water today and tomorrow, Proc., Specialty Conference of Irrigation and Drainage Division, ASCE, Flagstaff, Arizona, July 24–26, ASCE, Reston, Va., 231–236. Robinson, F. E., and Mayberry, K. S. (1976). “Seed coating, precision planting, and sprinkler irrigation for optimum stand establishment.” Agron. J., 68, 674–675. Shainberg, I., and Shalhevet, J., eds. (1984). Soil salinity under irrigation processes and management, Springer-Verlag, New York. Strelkoff, T., and Clemmens, A. J. (2007). “Hydraulics of surface systems,” in Design and operation of farm irrigation systems, G. J. Hoffman and R. E. Evans, eds., ASABE, St. Joseph, Mich., 436–498. Stringham, G. E., and Keller, J. (1979). “Surge flow for automatic irrigation,” in Irrigation and drainage in the nineteen-eighties, Proc., Specialty Conference of Irrigation and Drainage Division, ASCE, Albuquerque, NM, ASCE, Reston, Va., 132–142. Tanji, K. K., Grismer, M. E., and Hanson, B. R. (1985). “Subsurface drainage evaporation ponds.” Calif. Agric., 39(9/10), 10–12. Trout, T. J., Sojka, R. E., and Lentz, R. D. (1995). “Polyacrylamide effect on furrow erosion and infiltration.” Trans. ASAE, 38(3), 761–765. U.S. Department of Agriculture (USDA). (1997a). “Border irrigation,” in National Engineering Handbook, Chapter 4, Section 15 (Irrigation), USDA/NRCS, Washington, D.C. ———. (1997b). “Sprinkler irrigation,” in National Engineering Handbook, 2nd ed., Chapter 11, Section 15 (Irrigation) USDA/NRCS, Washington, D.C. ———. (1997c). National Engineering Handbook, 2nd ed., Chapter 6, Section 15 (Irrigation) USDA/NRCS, Washington, D.C. van Schilfgaarde, J., Bernstein, L., Rhoades, J. D., and Rawlins S. L. (1974). “Irrigation management for salt control.” J. Irrig. Drainage Div., ASCE, 100(IR3), 321–338. Walker, W. R. (2003.) “SIRMODII Surface irrigation simulation, evaluation and design, guide and technical documentation.” Biological and Irrig. Eng., Utah State University, Logan, Utah. Walker, W. R., and Skogerboe, G. V. (1987). Surface irrigation theory and practice, Prentice-Hall, Inc., Englewood Cliffs, N.J. West, D. W., and Merrigan, I. F. (1979). “Soil salinity gradients and growth of tomato plants under drip irrigation in saline soils.” Soil Sci., 127(5), 281–291. Wichelns, D., and Oster, J. D. (2006). “Sustainable irrigation is necessary and achievable, but direct costs and environmental impacts can be substantial.” Agric. Water Mgmt., 86, 114–127. Wu, Q., Christen, E. W., and Enever, D. (1999). On-farm and community scale salt disposal basins on the Riverine Plain: BASINMAN—A water balance mode for farms with subsurface pipe drainage and on-farm evaporation basins, CSIRO Land and Water Technical Report 01/99, CRC for Catchment Hydrology Technical Report 00/06, CSIRO Land and Water, Clayton South, Victoria, Australia.

538

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

NOTATION Di  depth of irrigation water D1  initial depth of water stored in profile at saturation DF  final depth of water stored in profile at saturation Dgw  depth of groundwater Lr  leaching requirement a  weighted electrical conductivity of water available to crop D  desired soil salinity value, expressed as electrical conductivity F  final average electrical conductivity of soil solution in profile gw  electrical conductivity of groundwater i  electrical conductivity of irrigation water 1  initial average electrical conductivity of soil solution in profile

CHAPTER 17 DRIP IRRIGATION AND SALINITY Blaine R. Hanson

INTRODUCTION Drip irrigation systems can apply water and chemicals more uniformly throughout a field (field-wide uniformity) at a higher irrigation frequency than other irrigation methods, and the water applications tend to be more uniform over time. This ability is not only governed by the technology but also by the design, installation, operation, and maintenance of the systems. While the field-wide uniformity of applied water can be high under drip irrigation, the distribution of the soil water content and salt is highly variable around emitters, with the highest water contents occurring near the emitter and decreasing with distance and depth from the emitter, whereas the lowest salinity levels are usually near the drip line and increase with distance and depth from the emitter. Drip irrigation has some specific advantages under saline conditions (Shalhevet 1994). First, no foliar absorption of salts occurs during irrigation, as would occur under sprinkle irrigation. Another advantage is that the wetting pattern around emitters results in highly leached soil near the drip line, a zone where root density frequently is the highest, particularly for row crops. Thus, lower levels of soil salinity are more likely to occur in the rootzone for a given amount of leaching than under other irrigation methods. A third advantage is high-frequency irrigation, which maintains relatively constant soil water content and soil salinity over time near drip lines. Common irrigation frequencies used for drip irrigation of trees and vines may be twice to once per week; drip systems used for row crops may be irrigated multiple times per day or multiple times per week. A main disadvantage of drip irrigation under saline soil conditions is salt accumulation near the periphery of the wetted pattern. Thus, the 539

540

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

placement of emitters relative to the plant row is critical for crops that are sensitive or moderately sensitive to soil salinity. Salt accumulation above buried drip lines also is a concern. This chapter reviews the general approaches to managing salts in drip irrigation systems and provides a case study conducted in California’s San Joaquin Valley that addresses several important management questions regarding drip irrigation in saline soils: (1) Is drip irrigation economically feasible? and (2) Does drip irrigation provide for or enhance salt management? Wetting and Salt Patterns under Drip Irrigation Wetting patterns The salt distribution around drip lines depends on wetting patterns during irrigation and subsequent redistribution of soil water content. Wetting patterns around a surface drip line (Fig. 17-1) and a subsurface drip line (Fig. 17-2) just after an irrigation show spatially varying soilwater content with radial distance from the drip line. The highest soilwater content values occur near the drip lines and the smallest values at the periphery of the wetted pattern. Root distributions around drip lines reflect wetting patterns. Roots were highly concentrated near drip lines where drip line placement coincided with the plant row (Hanson and

FIGURE 17-1. Soil water pattern under surface drip irrigation.

DRIP IRRIGATION AND SALINITY

541

FIGURE 17-2. Soil water pattern under subsurface drip irrigation. Depth of drip line is 200 mm. May 2007). Offsetting drip lines from plant rows results in a root distribution offset from the zone of highest water content. Salt patterns Bresler (1975) showed that near the emitter, soil salinity approached the salinity of the irrigation water. Soil salinity increased slightly with radial distance from the emitter until near the periphery of the wetted pattern, where large increases in salinity occurred over small increases in radial distance. Salt leaching occurred near the drip line, whereas salt accumulated near the periphery of the wetted pattern. Bresler concluded that the general pattern of soil salinity distribution depended on the initial soil salinity at the start of drip irrigation, salinity of the irrigation water, rate of irrigation, and soil hydraulic characteristics. Under conditions of continued high-frequency irrigation (multiple irrigations per week) in commercial fields, considerable leaching can occur over time beneath drip lines. Low salinity levels occurred near the drip lines and extended downward directly below the drip lines in a sandy loam and clay loam soil, with salt accumulating between drip lines and near the edge of the bed (Fig. 17-3A and B, respectively) (Hanson and Bendixen 2004). Under subsurface drip irrigation, salt accumulated above the drip line with the highest levels near the soil surface indicating no leaching above the drip line, but leaching occurred below the drip line as indicated by the relatively low salinity levels near and extending below the drip line (Fig. 17-4) (Hanson and Bendixen 1995).

542

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 17-3. Soil salinity (ECe) pattern under surface drip irrigation.

Soil salinity around subsurface drip lines under saline shallow groundwater conditions was found to depend on the depth to the groundwater, salinity of the shallow groundwater, salinity of the irrigation water, and amount of applied water in clay loam soil (Hanson and May 2004). Soil salinity throughout the profile was relatively uniform for a water table depth of about 2 m (Fig. 17-5A), but for water table depths less than 1 m,

FIGURE 17-4. Soil salinity (ECe) pattern under subsurface drip irrigation.

DRIP IRRIGATION AND SALINITY

543

FIGURE 17-5. Soil salinity (ECe) around the drip line for (A) a water depth of about 2 m, EC of irrigation water  0.3 dS m1, and EC ground water  8 to 11 dS m1; (B) a water depth of about 0.6 to 1 m, EC of irrigation water  0.3 dS m1, and EC ground water  5 to 7 dS m1; and (C) a water table depth of 0.6 to 1 m, EC of irrigation water  1.1 dS m1, and EC ground water  9 to 16 dS m1. soil salinity varied spatially around the drip lines (Fig. 17-5B,C). Higher soil salinity near the drip line occurred for higher irrigation water salinity (Fig. 17-5C). Salt accumulated directly above the drip line. Crop Yield Response to Salinity under Drip Irrigation The effect of soil salinity on crop yield historically has been described by crop tolerance curves that show no yield effect until a threshold soil salinity is reached, and thereafter a linear decline in yield as soil salinity

544

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

increased (Maas and Hoffman 1977). Soil salinity is described by the salinity of the saturated extract (ECe). The threshold soil salinity is the maximum salinity at which no yield reduction occurs. The relationship between relative yield (Yr) and ECe is described by Yr  100  b  (ECe  a), ECe  a

(17-1)

where a is the threshold soil salinity and b is the slope of the line describing the relationship between Yr and ECe. Tables containing threshold values and slopes (Yr/ECe) are used to assess the salinity hazard (Maas and Grattan 1999). The Maas and Grattan tables reflect values obtained with sprinkle or furrow irrigation. Using data from experiments to assess the effect of irrigation water salinity or soil salinity on crop yield under drip irrigation that included rootzone soil salinity data, the yield–soil salinity data were compared with the historical relationships to determine whether drip irrigation reduced the effect of soil salinity on yield compared to furrow/ sprinkle irrigation. In one experiment, daily drip irrigation using irrigation water salinities of 1.2, 3.5, 8.2, and 10.5 dS m1 showed a significant yield reduction for seven cultivars of iceberg lettuce for water salinities of 8.2 and 10.5 dS m1, but only a slight reduction occurred for three cultivars of romaine lettuce (Pasternak et al. 1986). The 10.5-dS m1 irrigation water resulted in an average ECe of 7.7 dS m1, which caused a 14% yield reduction for the romaine lettuce and 40% yield reduction for the iceberg lettuce. The yield reduction based on the historical crop tolerance values is 83%. The experiment a and b (Eq. 17-1) were 1.98 dS m1 and 5.6%/ECe, respectively, compared to the historical values of 1.3 dS m1 and 13%/ECe (Mass and Grattan 1999). Based on these results, these lettuce cultivars were classified as salt-tolerant under drip irrigation, whereas under sprinkle or furrow irrigation the historical classification is moderately saltsensitive. In another experiment, two cultivars of lettuce irrigated with drip irrigation on clay loam were found to be moderately salt-tolerant compared to the historical classification of moderately salt-sensitive (De Pascale and Barbieri 1995). Experiment values of a and b (Eq. 17-1) were 2.7 dS m1 and 5.8%/ECe, respectively, compared to the respective historical values of 1.3 dS m1 and 13.0%/ECe (Mass and Grattan 1999). Drip irrigation of lettuce on a sandy soil using irrigation water salinities of 1.7, 3.2, and 4.7 dS m1 and varying water applications showed decreasing yield with increasing irrigation water salinity (Russo 1987). At the higher irrigation water applications, the yield–soil salinity relationship approached that described in Maas and Grattan (1999). Yields of daily drip irrigated processing tomatoes decreased as irrigation water salinity increased for water salinities of 1.2, 4.5, and 7.5 dS m1

DRIP IRRIGATION AND SALINITY

545

(Pasternak et al. 1986). Applying saline water immediately after sowing reduced the yield by 60% for the 7.5 dS m1 water, whereas applying the saline water at the 4-leaf stage and the 11-leaf stage reduced the yield by 28% and 30%, respectively. The average yield reduction of the 4.5 dS m1 water was 9.5%, regardless of the stage at which saline water was applied. Using the measured ECe values of the experiment, these yield reductions were smaller than that predicted by the historical crop tolerance values except for the 7.5 dS m1 water applied immediately after sowing. Pasternak and De Malach (1995) found that processing tomato yields decreased as drip irrigation frequency decreased (1 day1, 2 day1, every 2 days, every 3 days) on a fine sand. Relative yields (both total and marketable) of a saline water treatment (6.2 dS m1) ranged from 0.40 to 0.46 of the fresh-water treatment (1.2 dS m1) yield for the first three respective irrigation frequencies and was 0.31 for irrigation every third day. They also found that higher yields generally occurred for daytime irrigations compared to night irrigations. Fresh-water yields of pulsed irrigation (five times per day) were 0.66 (total yield) and 0.72 (marketable yield) of the yield of one irrigation per day. However, saline water yields under pulse irrigation were 1.75 (total) and 1.92 (marketable) of the yields of the one irrigation per day with nonsaline water. It was believed that the pulse irrigation better controlled the soil salinity of the rhizosphere under saline soil conditions. Slightly higher total and marketable tomato yields occurred under drip irrigation than under furrow irrigation on a clay loam soil for irrigation water salinities ranging from 0.55 to 4.5 dS m1 (Malash et al. 2005). However, the rate of the relative yield decrease with increasing irrigation water salinity was similar for both irrigation methods. Higher yields also occurred using continuous applications of saline water compared to alternate applications of fresh and saline water. Gawad et al. (2005) showed significantly higher tomato yields with drip irrigation compared with furrow irrigation irrigated with saline water on a clay soil. Irrigation water salinity ranged from 0.6 to 9 dS m1. The leaching fraction was estimated to be 15%. However, similar relationships occurred between relative yield and irrigation water salinity for both irrigation methods (relative yields were based on the maximum yield of each irrigation method). In addition, an experiment using 10 tomato varieties showed a significant variety effect on the response of tomato yield to irrigation water salinity. Yields of the varieties ranged from 8.8 to 58.4 Mg ha1 for the fresh water application. Yields of all varieties decreased as irrigation water salinity increased, with relative yields ranging from 0.35 to 0.73 for 6.0-dS m1 irrigation water and from 0.22 to 0.50 for 9.0-dS m1 water. The relative yields were based on the maximum yield of each variety. Also, a comparison of continuous application of saline water versus cyclic applications of fresh and saline water generally showed higher yields for the continuous application approach.

546

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Smaller yield reductions of onions irrigated with saline water were found compared to the historical methods (Pasternak et al. 1984). Both fresh water (1.2 dS m1) and saline water (4.4 dS m1) were used for stand establishment (sprinkle irrigation). The applied water was at least 30% higher than the crop evapotranspiration (ET) between irrigations. Irrigation with the saline water reduced the yield by 6.2% for the fresh-water stand establishment and by 28.6% for saline-water stand establishment. No differences in yield were found among the four cultivars used in this experiment. De Malach et al. (1989) also found that the later the application of the saline water (sowing, two-leaf stage, and five-leaf stage), the higher the onion yield. Drip irrigation of maize (field and sweet) on a sandy loam using continuous and alternate applications of saline water, with irrigation water salinities of 1.2, 4.5, 7.0, and 10.5 dS m1, showed that continuous saline water irrigation reduced the yield by 11.2%/ECe (field) and 11.9%/ECe (sweet) (Pasternak et al. 1985), similar to the historical yield reduction (Maas and Grattan 1999). Yields of the alternate treatment were reduced only for the 10.5-dS m1 salinity. The applied water was at least 30% higher than the crop ET between irrigations. Pasternak et al. (1995) found relative yields of 14 sweet corn cultivars drip-irrigated with 6.2 dS m1 water to range from 0.54 to 0.82 of the maximum yield of each variety (obtained with 1.2-dS m1 irrigation water). Soil salinity over the 0- to 30-cm depth interval was about 6.5 dS m1 for the saline water and 1.4 to 2.9 dS m1 for the fresh water. Based on these soil salinities, a yield reduction of 0.54 would be expected for the saline water using the historical approach. Daily drip irrigation (3 pulses per day) was used on potatoes with irrigation water salinities of 1.2, 4.5, and 6.2 dS m1 (Bustan et al. 2004). Soil salinity levels ranging from 1.6 to 6.6 dS m1 had no effect on the yield of the drip-irrigated potatoes grown in a deep sandy soil as long as extreme heat conditions did not occur. These soil salinity values exceeded the historical threshold value of 1.2 dS m1, indicating that the drip irrigation and its management reduced the effect of soil salinity on crop yield under these conditions. Assouline et al. (2006) found that pulsed drip irrigation (10 pulses per day) on bell peppers had no effect on yield compared with one irrigation per day for irrigation water salinities of 1.8 and 4.2 dS m1 and for water applications equal to 100% and 125% of the predicted seasonal crop water requirement. Smaller yields occurred for the saline water, with relative yield–ECe relationship similar to the historical relationship. The results of these studies suggest that drip irrigation has a potential for higher crop yields under saline conditions compared to other irrigation methods. This potential depends on site-specific conditions, such as crop type, soil type, and climate, among others. Relative yield–ECe data for

DRIP IRRIGATION AND SALINITY

547

lettuce (De Pascale and Barbieri 1995; Pasternak et al. 1986), onion (Pasternak et al. 1984), and potato (Bustan et al. 2004) showed a smaller soil salinity effect under drip irrigation on relative yield than predicted by the historical relationships for sprinkle or furrow irrigation. However, studies on tomatoes (Gawad et al. 2005; Malash et al. 2005; Pasternak et al. 1986), maize (Pasternak et al. 1985), and bell peppers (Assouline et al. 2006) showed a relative yield response similar to the historical crop tolerance relationships (Maas and Grattan 1999). The historical crop tolerance relationships of these crops were determined from studies generally conducted decades ago using crop varieties of those time periods. Gawad et al. (2005) and Pasternak et al. (1995) found a significant variety effect of tomatoes and maize, respectively, on the yield response to irrigation water salinity. Thus, the crop varieties used in these previously discussed experiments may be a factor in the relative yield response to ECe under drip irrigation, making it difficult to compare the effect of drip irrigation only under saline soil conditions with the historical relationships. Other factors that could contribute to the yield–salinity relationship include climate, soil texture, root distribution as related to the salt distribution, and time and duration of exposure to saline water (Shalhevet 1994). One advantage of drip irrigation is no foliar absorption, as would occur under sprinkle irrigation. Shalhevet (1994) reported on a comparison of sprinkle and drip irrigation of potatoes using saline water that found similar threshold values for sprinkle and drip irrigation. However, the decrease in yield with increasing irrigation water salinity for sprinkle irrigation was twice that of drip irrigation, attributed to foliar absorption of salt under sprinkle irrigation. The highest pepper yield occurred under drip irrigation compared to furrow and sprinkle irrigation, regardless of water quality (450 and 2,450 ppm) (Bernstein and Francois 1973). The saline water greatly reduced the yield of sprinkle irrigation to statistically small values, believed to be due to foliar absorption. However, similar relationships between yield and applied water were found in a sandy loam soil for sprinkle and drip-irrigated cotton irrigated with water qualities of 3.7 dS m1 and 7.1 dS m1 (Meiri et al. 1992). These different yield responses reflect the crop’s sensitivity to foliar injury from saline water. Cotton is not highly susceptible to foliar injury, whereas peppers are very susceptible (Grattan 2006). While potatoes were not rated, their response suggests a high level of susceptibility. Salinity Control under Drip Irrigation The key to profitable drip irrigation under saline conditions is adequate salinity control in the rootzone. Salinity control involves leaching salts from the rootzone by applying irrigation water in excess of the soil

548

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

moisture depletion. The leaching fraction (LF), used as a measure of adequacy of leaching, is the ratio of the amount of water draining below the rootzone to the amount of infiltrated water. Leaching under drip irrigation depends on the EC of the irrigation water, the amount applied, and the wetting pattern, which is also dependent on the applied amount. Lack of leaching around the drip line due to insufficient applied water can cause a zone of high soil salinity near drip lines (Fig. 17-6A), whereas a low-salt zone can occur under leaching conditions (Fig. 17-6B). Applying sufficient water for a leaching fraction of 17% resulted in the low-salt zone extending horizontally to about 300 mm from a surface drip line, whereas the horizontal distance of the low-salt zone extended only about 30 mm for a leaching fraction of 2%

FIGURE 17-6. Soil salinity (ECe) pattern under conditions of (A) no leaching around the drip line, and (B) leaching around the drip line.

DRIP IRRIGATION AND SALINITY

549

(Hoffman et al. 1985). A higher water application also has been shown to produce a larger low-salt zone around the drip line than occurred for a smaller application for subsurface drip irrigation (200 mm deep) under saline, shallow groundwater conditions (Fig. 17-7) (Hanson et al. 2006). Experiments by Hoffman et al. (1979), Hoffman and Jobes (1983), and Jobes et al. (1981) showed yield under drip irrigation to increase to a maximum as LF increased. The salinity of the irrigation water was 2.3 dS m1. Maximum yields occurred for wheat, sorghum, barley, oats, celery, and cauliflower at LFs of 14%, 16%, 10%, 13%, 14%, and 17%, respectively. For lettuce, cowpeas, and tomatoes, maximum yields were not reached even though maximum experimental LFs of 18%, 16%, and 19%, respectively, were obtained. The yield increase with LF reflected an increase in crop ET due to a combination of lower soil salinity near the drip line, a larger zone of low salt soil around drip lines, and higher soil-water contents as the LF increased. Several methods have been historically used to determine LFs in commercial fields. One method requires measurements of the irrigation water EC and the average rootzone soil salinity (ECe). Chloride concentrations

FIGURE 17-7. Effect of amount of applied water on zone of low salt soil around drip line for subsurface drip irrigation.

550

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

are sometimes used instead of the EC. These data are then related to the LF using leaching curves or appropriate equations (Ayers and Westcot 1985; Rhoades 1982). However, both salt patterns and root patterns vary spatially around drip lines, which may result in uncertainty in estimating LFs using soil salinity data. The water balance is commonly used to estimate field-wide leaching fractions using cumulative amounts of applied water and ET during a given time period. The leaching amount is the difference between applied water and ET. The ET is commonly calculated as the product of crop coefficients and a reference crop ET, usually the ET of well-water grass calculated from climatic data and appropriate equations or from pan evaporation data. Actual LFs were estimated for drip-irrigated almonds in silt loam to clay loam soil using soil chloride concentrations (Nightingale et al. 1991). Amounts of applied irrigation water were 50%, 100%, and 150% of the ET of the 100% water application. The water balance method indicated that no leaching occurred for the 50% and 100% applied-water treatments and a 50% LF for the 150% treatment, but actual LFs, calculated from chloride concentrations, were 4% to 6% for the 50% water treatment, 10% to 22% for the 100% treatment, and 31% to 36% for the 150% treatment. These values were based on the soil chloride levels of the first 1- to 1.6-m distance from the tree. This behavior suggests that the water balance approach to calculating field-wide LFs is inappropriate for drip irrigation because spatially varying soil water patterns under drip irrigation cause leaching below the drip lines, even for conditions considered to be deficit irrigation (i.e., applied water amounts are smaller than the 100% ET). Under subsurface drip irrigation, no leaching by drip irrigation occurs above the drip lines. Thus, periodic leaching either by rainfall or sprinkle irrigation is necessary to control rootzone soil salinity above the drip lines. Sufficient leaching water should be applied to move the salts below the drip line, where they will eventually be leached by subsequent drip irrigation. Figure 17-8 shows the redistribution of salts in the spring due to rainfall and sprinkle irrigation for the salinity distribution shown in Fig. 17-4, which occurred at the end of the previous crop season. The placement of drip lines relative to plant rows is critical in salinity control with drip irrigation. Salinity control will be best where drip lines and plant rows coincide. Under this condition, root density generally will be the highest near the drip lines, where leaching will be the greatest (Hoffman et al. 1983; Moshrefi and Reese 1985; Hanson and May 2007). Offsetting drip lines from plant rows can shift the zone of high root density away from the zone of highly leached soil, and, in some cases, into the zone of salt accumulation (Mantell et al. 1985).

DRIP IRRIGATION AND SALINITY

551

FIGURE 17-8. Effect of winter rainfall (356 mm) on soil salinity (ECe) for the site used in Fig. 17-4.

Drip Irrigation under Shallow, Saline Groundwater Conditions: A Case Study in the San Joaquin Valley, California Background Many areas in the world are affected by adverse levels of soil salinity due to salt accumulation in the rootzone from the upward flow of saline, shallow groundwater. Leaching of salts from the rootzone may be difficult with surface irrigation methods under these conditions because of the water table rise caused by irrigation. In some areas, the lack of drainage water disposal facilities limits the use of subsurface drainage systems for salinity and water table control. Drip irrigation offers a potential for better water table control and better rootzone salinity control. Along the west side of the San Joaquin Valley, California, about 400,000 ha of irrigated land are affected by excessive levels of soil salinity, the result of saline, shallow groundwater conditions. After more than 30 years of research on drainage water disposal, the lack of widespread economically, technically, and environmentally feasible subsurface drainage water disposal facilities continues to plague agriculture along the west side of the valley. In some areas, land retirement is being implemented as a solution. The only options available to growers to address the salinity/drainage problem without retiring land are better management of irrigation water to reduce drainage, increasing crop water use of the shallow groundwater without any yield reductions, and drainage water reuse. All of these methods, however, require adequate salinity control in the rootzone. A University of California study (Shoups et al. 2005) concluded that a soil

552

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salt balance must be maintained that allows for productive cropping systems, and that continued irrigation without changing management practices is not sustainable. One option is to convert from furrow or sprinkle irrigation to drip irrigation. The main disadvantage of this option is the installation cost of drip irrigation systems, which, based on grower experience, could range from $1,480 ha1 to $2,470 ha1. Profitability is the key for viable drip irrigation in these salt-affected soils. In the late 1980s two large-scale comparisons of subsurface drip irrigation and furrow irrigation of cotton showed that higher cotton yields and smaller water applications consistently occurred for subsurface drip irrigation compared to furrow irrigation under saline, shallow groundwater conditions. However, the profit of furrow-irrigated cotton was much higher than that of drip irrigation at one location, whereas a slightly higher profit was found for drip irrigation at the other site (Fulton et al. 1991; Styles et al. 1997). Thus, growers converting to drip irrigation of cotton assume an economic risk. Subsurface drip irrigation of high-cash-value crops, such as processing tomatoes, offers a better potential for increased profitability compared to cotton. However, these higher-valued crops generally are more salt-sensitive, which could reduce crop yields in salt-affected soil. For example, tomatoes are classified as moderately sensitive to soil salinity with a threshold ECe of 2.5 dS m1. The threshold value for cotton is 7.7 dS m1. Field experiments A 3-year project (1991–1993) conducted in a commercial field under saline, shallow groundwater conditions found higher cotton and processing tomato yields for subsurface drip irrigation than for furrow irrigation (Ayars et al. 2001). The shallow groundwater supplied about 40% of the cotton water requirement and 10% of the processing tomato water requirement. No long-term salt accumulation was found under drip irrigation. Between 1998 and 2003, experiments in three commercial fields on the west side of the San Joaquin Valley showed that subsurface drip irrigation of processing tomatoes was highly profitable compared to sprinkle irrigation under saline, shallow groundwater conditions (Hanson and May 2003, 2004). The average yield of the subsurface drip-irrigated fields was 90.7 Mg ha1 versus 75.9 Mg ha1 for sprinkle irrigation—statistically significant at a level of significance of 0.05. The average difference in soluble solids between the two irrigation methods was not statistically significant. Subsurface drip irrigation increased the profit by $1,195 ha1 compared to sprinkle irrigation. Water table depths at these fields ranged

DRIP IRRIGATION AND SALINITY

553

from 0.45 m to 2 m. The EC of the irrigation water ranged from 0.3 dS m1 (irrigation district water) to 1.1 dS m1 (groundwater). EC of the shallow groundwater ranged from 4.0 to 16.4 dS m1. Soil type was clay loam at all fields. Drip irrigations occurred two to three times per week. At a fourth commercial field, an experiment evaluating the yield response of tomatoes and cotton to amount of applied water under very shallow groundwater conditions (0.46 m to 0.61 m deep) showed tomato yield to decrease with decreasing applied water, but little response of cotton to applied water was found (Hanson et al. 2006). The maximum tomato yield was similar to those found in the previously discussed commercial fields. EC of the irrigation water was 0.5 dS m1 and the EC of the groundwater was 8 to 10 dS m1. Daily irrigations occurred. At all sites, little response of water table depth to drip irrigation was found except when overirrigation occurred in one year at one site. A subsequent reduction in applied water at that site caused the water table to decline. The field-wide leaching amounts calculated using the water balance approach showed little or no field-wide leaching at most of these commercial sites (Table 17-1), which suggests inadequate salinity control and TABLE 17-1. Seasonal Applied Water and Evapotranspiration, and Field-Wide Leaching Fractions Calculated from a Water Balance for Four Commercial Sites, Designated as BR, DI, DE, and BR2

Year

Seasonal Applied Water (mm)

Seasonal Evapotranspiration (mm)

Leaching Fraction (%)

BR 1999 2000 2001

406 427 521

1999 2000 2001

564 737 582

516 544 582

0 0 0

638 640 676

0 13 0

615 587

14 0

617

0

DI

DE 2000 2001

732 561 BR2

2002

589

554

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

raises questions about the long-term viability of drip irrigation under saline conditions. The soil salinity data (Fig. 17-5), however, clearly showed that substantial localized leaching was occurring around the drip lines and that the leaching was concentrated near the drip line. Computer simulations Because the soil salinity data indicated leaching around drip lines, but the water balance method generally indicated no leaching in the commercial fields, the HYDRUS-2D computer simulation model was used to evaluate leaching with subsurface drip irrigation and to estimate LFs under the saline, shallow groundwater conditions found in the commercial fields (Hanson et al. 2008). This model simulated the movement of water and salt in soil under drip irrigation for a 42-day period and determined the amount of drainage below the rootzone for water table depths of 0.5 m and 1.0 m; irrigation water salinities of 0.3, 1.0, and 2.0 dS m1; and applied water amounts of 60%, 80%, 100%, and 115% of the 100% ET. The drip line was 200 mm deep. The EC of the shallow groundwater was 10.0 dS m1 and 8.0 dS m1 for the 0.5- and 1.0-m water table depths, respectively, based on measured levels. These simulations showed that larger water applications applied less frequently reclaimed the soil faster than smaller applications applied more frequently. Salinity patterns were consistent with those measured in the commercial fields. As time progressed, the volume of reclaimed soil increased, with most of the reclamation occurring below the drip line. Salts accumulated near the soil surface. The larger the amount of applied water, the larger the volume of reclaimed soil below the drip line (consistent with the field measurements), but the amount of applied water had little effect on the volume of reclaimed soil above the drip line. The salinity near the drip line increased as the irrigation water salinity increased. Actual or localized LFs ranged from 7.7% (60% water application) to 30.5% (115% water application) and was 24.5% for the 100% water application. Even for applications considered to be severe deficit irrigation, drainage below the rootzone occurred, caused by the wetting patterns under drip irrigation. Is subsurface drip irrigation viable under saline conditions? The field experiments and computer simulations show that subsurface drip irrigation of processing tomatoes has a high potential for profitability, the key to which is salinity control. Salinity control involves irrigating with relatively low-salt irrigation water, applying sufficient irrigation water for localized leaching, leaching salts that accumulate

DRIP IRRIGATION AND SALINITY

555

above the drip line, and preventing shallow groundwater intrusion into the rootzone. Recommendations for long-term viability of subsurface drip irrigation under saline, shallow groundwater conditions are as follows: • Seasonal water applications should be about equal to the seasonal ET. This amount of water provides sufficient localized leaching, minimizes crop water use of the saline, shallow groundwater, and prevents groundwater intrusion into the rootzone. Higher applications could raise the water table; smaller applications could decrease tomato yield. • The EC of the irrigation water should be about 1.0 dS m1 or less for the varieties grown in these salt-affected soils. Higher EC levels may reduce yield. • Periodic leaching of salt accumulated above the buried drip lines will be necessary with sprinkle irrigation for stand establishment if winter and spring rainfall is insufficient. • Drip-irrigation systems should be designed for a high uniformity of applied water. • Drip-irrigation systems should be properly maintained to prevent emitter clogging. Are subsurface drainage systems and drainage water disposal methods needed under drip irrigation? No subsurface drainage systems were used at these sites. Water table depths ranged from 0.45 m to about 2 m. At all sites, groundwater salinity was high, but no trend in yield with water table depth or soil salinity was found. Substantial localized leaching occurred around the drip line at all sites, regardless of the water table depth and its salinity. Subsurface drip irrigation continues to be used at this time at these sites. Little response of the water table to drip irrigation occurred except at one site, where overirrigation occurred part of the time. Although drainage below the rootzone occurs under subsurface drip irrigation (based on the HYDRUS-2D modeling), the amount of drainage per irrigation is small because of the small water applications per irrigation, and its distribution over time is relatively uniform because of the high irrigation frequency. Thus, the natural subsurface drainage in these fields appeared to be sufficient to prevent groundwater intrusion into the rootzone. This behavior suggests that, for the conditions found in these fields, subsurface drainage systems and drainage water disposal methods are not needed for properly managed and designed drip-irrigation systems.

556

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SUMMARY Subsurface drip irrigation involves more frequent but smaller water applications and results in a different distribution of low-salinity irrigation water and accumulated salts than do sprinkler or furrow irrigation. Under drip irrigation, salts tend to accumulate in the soil above the drip line and along the perimeter of the area wetted by the irrigation. Crop response to drip irrigation includes concentration of roots in the zone near and below the drip lines. Depending on the timing of irrigation and the placement of plants in relation to the drip line, the more localized distributions of water and salt provided by drip irrigation may result in a different yield–ECe relationship than is predicted in the literature and may effectively reduce adverse effects of irrigation with saline waters on salt-sensitive plants. For high-value crops, such as tomatoes, subsurface drip irrigation may be cost-beneficial provided that (1) the application rate is maintained close to the ET rate, (2) the electrical conductivity of the irrigation water is not greater than about 1.0 dS m1, (3) the soil above the drip line is leached periodically, and (4) the drip-irrigation system is properly designed and maintained. With these conditions met, subsurface drip irrigation does not appear to affect water table elevation and will reduce the potential for irrigation to raise groundwater levels into the rootzone. REFERENCES Assouline, S., Moller, M., Cohen, S., Ben-Hur, M., Grava, A., Narkis, K., and Silber, A. (2006). “Soil-plant system response to pulsed drip irrigation and salinity: Bell pepper case study.” Soil Sci. Soc. Am. J., 70, 1556–1568. Ayars, J. E., Schoneman, R. A., Dale, F., Meso, B., and Shouse, P. (2001). “Managing subsurface drip irrigation in the presence of shallow groundwater.” Agric. Water Mgmt., 47, 243–264. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper (Rev. 1), Food and Agriculture Organisation of the United Nations, Rome. Bernstein, L., and Francois, L. E. (1973). “Comparisons of drip, furrow, and sprinkler irrigation.” Soil Sci., 115, 73–86. Bresler, E. (1975). “Two-dimensional transport of solutes during nonsteady infiltration from a trickle source.” Proc., Soil Science Society of America, 39, 604–613. Bustan, A., Sagi, M., De Malach, Y., and Pasternak, D. (2004). “Effects of saline irrigation water and heat waves on potato production in an arid environment.” Agric. Water Mgmt., 90, 275–285. De Malach, Y., Pasternak, D., Mendlinger, S., Borovic, I., and Abdel, N. (1989). “Irrigation with brackish water under desert conditions. VIII. Further studies on onion (Allium cepa L.) production with brackish water.” Agric. Water Mgmt., 16, 204–215.

DRIP IRRIGATION AND SALINITY

557

De Pascale, S., and Barbieri, G. (1995). “Effects of soil salinity from long-term irrigation with saline-sodic water on yield and quality of winter vegetable crops.” Scientia Hort., 64, 145–157. Fulton, A. E., Oster, J. D., Hanson, B. R., Phene, C. J., and Goldhamer, D. A. (1991). “Reducing drainwater: Furrow vs. subsurface drip irrigation.” Calif. Agric., 45(2), 4–8. Gawad, G. A., Arslan, A., Gaihbe, A., and Kadouri, F. (2005). “The effects of saline irrigation water management and salt tolerant tomato varieties on sustainable production of tomato in Syria (1999–2002).” Agric. Water Mgmt., 78, 39–53. Grattan, S. R. (2006). “Salt accumulation in leaves under sprinkler irrigation,” in Agricultural salinity and drainage, Division of Agriculture and Natural Resources Publication 3375, University of California. Hanson, B. R., and Bendixen, W. E. (1995).” Drip irrigation controls soil salinity under row crops.” Calif. Agric., 49(4), 19–23. ———. (2004). “Drip irrigation evaluated in Santa Maria Valley strawberries.” Calif. Agric., 58(1), 48–53. ˇ ˚ nek, J. (2008). “Leaching with subsurface Hanson, B., Hopmans, J. W., and Simu drip irrigation under saline, shallow groundwater conditions.” J. Vadose Zone Hydrol., 7(2), 810–818. Hanson, B. R., Hutmacher, R. B., and May, D. M. (2006). “Drip irrigation of tomato and cotton under shallow saline groundwater conditions.” Irrig. and Drainage Sys., 20, 155–175. Hanson, B. R., and May, D. M. (2003). “Drip irrigation increases tomato yields in salt-affected soil of San Joaquin Valley.” Calif. Agric. 57(4), 132–137. ———. (2004). “Effect of subsurface drip irrigation on processing tomato yield, water table depth, soil salinity, and profitability.” Agric. Water Mgmt., 68, 1–17. ———. (2007). The effect of drip line placement on yield and quality of drip-irrigated processing tomatoes. Irrig. and Drainage Sys., 21(2), 109–118. Hoffman, G. J., and Jobes, J. A. (1983). “Leaching requirement for salinity control. III. Barley, cowpea, and celery.” Agric. Water Mgmt., 6, 1–14. Hoffman, G. J., Jobes, J. A., and Alves, W. J. (1983). “Response of tall fescue to irrigation water salinity, leaching fraction, and irrigation frequency.” Agric. Water Mgmt., 7, 439–456. Hoffman, G. J., Rawlins, S. L., Oster, J. D., Jobes, J. A., and Merrill, S. D. (1979). “Leaching requirement for salinity control. I. Wheat, sorghum, and lettuce.” Agric. Water Mgmt., 2, 177–192. Hoffman, G. J., Shannon, M. C., and Jobes, J. A. (1985). “Influence of rain on soil salinity and lettuce yield,” in Drip/Trickle Irrigation in Action, Proc., 3rd International Drip/Trickle Irrigation Congress, Fresno, California, November 18–21, ASABE, St. Joseph, Mich., 659–665. Jobes, J. A., Hoffman, G. J., and Wood, J. D. (1981). “Leaching requirement for salinity control. II. Oat, tomato, and cauliflower.” Agric. Water Mgmt., 4, 393–407. Maas, E. V., and Grattan, S. R. (1999). “Crop yields as affected by salinity,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Agronomy Monograph 38, ASA/CSSA/SSA, Madison, Wisc., 55–108. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. and Drainage Div. ASCE, 103(IR2), 115–134.

558

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Malash, N., Flowers, T. J., and Ragab, R. (2005). “Effect of irrigation systems and water management practices using saline and non-saline water on tomato production.” Agric. Water Mgmt., 78, 25–38. Mantell, A., Frenkell, H., and Meiri, A. (1985). “Drip irrigation of cotton with saline-sodic water.” Irrig. Sci., 6, 95–106. Meiri, A., Frankel, H., and Mantell, A. (1992). “Cotton response to water and salinity under sprinkler and drip irrigation.” Agron. J., 84, 44–50. Mosehrefi, N., and Reese, F. (1985). “Effect of irrigation system on salt and root distribution,” in Drip/Trickle Irrigation in Action, Proc., 3rd International Drip/Trickle Irrigation Congress, Fresno, California, November 18–21, ASABE, St. Joseph, Mich., 706–711. Nightingale, H. I., Hoffman, G. J., Rolston, D. E., and Biggar, J. W. (1991). “Trickle irrigation rates and soil salinity distribution in an almond (Prunus amhgdalus) orchard.” Agric. Water Mgmt., 19, 271–283. Pasternak, D., and De Malach, Y. (1995). “Irrigation with brackish water under desert conditions X. Irrigation management of tomatoes (Lycopersicon esculentum Mills) on desert sand dunes.” Agric. Water Mgmt., 28, 121–132. Pasternak, D., DeMalach, Y., and Borovic, I. (1984). “Irrigation with brackish water under desert conditions. I: Problems and solutions in production of onions.” Agric. Water Mgmt., 9, 225–235. Pasternak, D., De Malach, D., and Borovic, I. (1985). “Irrigation with brackish water under desert conditions II. Physiological and yield response of maize (Zea mays) to continuous irrigation with brackish water and to alternating brackish-fresh-brackish water irrigation. Agric. Water Mgmt., 10, 47–69. ———. (1986). “Irrigation with brackish water under desert conditions VII. Effect of time of application of brackish water on production of processing tomatoes (Lycopersicon esculentum Mill.).” Agric. Water Mgmt., 12, 149–158. Pasternak, D., De Malach, D., Borovic, I., Shram, M., and Aviram, C. (1986). “Irrigation with brackish water under desert conditions IV. Salt tolerance studies with lettuce (Lactuca sative L.).” Agric. Water Mgmt., 11, 303–311. Pasternak, D., Sagih, M., De Malach, Y., Keren, Y., and Shaffer, A. (1995). “Irrigation with brackish water under desert conditions XI. Salt tolerance in sweetcorn cultivars.” Agric. Water Mgmt., 28, 325–334. Rhoades, J. D. (1982). “Reclamation and management of salt-affected soils after drainage.” in Proc., 1st Annual Western Provincial Conference on Rationalization of Water and Soil Resources and Management, Lethbridge, Alberta, Canada, November 27–December 2, 123–197. Russo, D. (1987). “Lettuce yield-irrigation water quality and quantity relationships in a gypsiferous desert soil.” Agron. J., 79, 8–14. Schoups, G., Hopmans, J. W., Young, C. A., Vrugt, J. A., Wallender, W. W., Tanji, K. K., and Panday, S. (2005). “Sustainability of irrigated agriculture in the San Joaquin Valley, California.” Proc. Natl. Acad. Sci. USA, 102(43), 15352–15356. Shalhevet, J. (1994). “Using water of marginal quality for crop production: Major issues.” Agric. Water Mgmt., 25, 233–269. Styles, S., Oster, J. D., Bernaxconi, P., Fulton, A., and Phene, C. (1997). “Demonstration of emerging technologies,” in Agroecosystems: Sources, control and remediation, J. Guitjens and L. Dudley, eds., Pacific Division, American Association for the Advancement of Science, San Francisco, 183–206.

DRIP IRRIGATION AND SALINITY

559

NOTATION a  threshold soil salinity b  slope of the line describing the relationship between Yr and ECe ECe  electrical conductivity of an extract of a saturated paste Yt  relative yield

This page intentionally left blank

CHAPTER 18 MANAGEMENT OF DRYLAND SALINE SEEPS A. D. Halvorson and J. L. Richardson

INTRODUCTION This chapter discusses the identification, diagnosis, control, and reclamation of dryland saline seeps. A saline seep results from a salinization process, often accelerated by dryland farming that enhances water flow through salt-laden substrata below the rootzone. This results in intermittent or continuous saline water discharge at or near the soil surface, commonly downslope from recharge areas under dryland conditions. Saline water reduces or eliminates the crop growth in the discharge area due to increased soluble salt concentrations in the rootzone. Saline seeps can be differentiated from other saline soil conditions by their recent and localized origin, saturated rootzone profile, shallow water table, and sensitivity (short-term response) to precipitation and cropping systems (Brown et al. 1983). Although this chapter is focused on seeps in the Great Plains of the United States, similar processes of localized seep formation may occur in areas with similar characteristics, and this chapter has application to seep formation processes in other regions. Saline seeps occur frequently in dryland farming areas throughout the Great Plains (Ballantyne 1963; Berg et al. 1986; 1991; Brown et al. 1987; Colburn 1983; Doering and Sandoval 1976b; Halvorson and Black 1974; Neffendorf 1978; Vander Pluym 1978). Miller et al. (1981) estimated that nearly 1 million ha of productive cropland have been salinized in the northern Great Plains. Saline seep problems also exist in Australia (Malcolm 1982; Matheson 1968), India, Iran, Turkey, and Latin America (Olson 1978). Saline seeps result from a combination of geologic, climatic, hydrologic, and cultural (land-use) conditions. The primary cause is a change from grassland or forest to a cropping system, such as crop–summerfallow rotation, that

561

562

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

allows rainfall in watershed recharge areas to move below the rootzone and provide seepage water.

FACTORS CONTRIBUTING TO SALINE SEEP DEVELOPMENT Saline seeps in the Great Plains possess similar characteristics and causes (Berg et al. 1986, 1991; Brown et al. 1983; Doering and Sandoval 1976a,b; Halvorson and Black 1974; Vander Pluym 1978). Typically, native grasses or naturally occurring vegetation has been replaced with agricultural fields and cropping systems with lower evapotranspiration (ET) potential. Precipitation that exceeds ET use and the rootzone’s soil storage capacity (which takes place, according to these authorities, primarily during summerfallow periods) is the source of seepage water that drives the salinization of seeps. The crop-summerfallow system of dryland farming has contributed significantly to the development of saline seep problems in the northern Great Plains but is not the sole cause (Brown et al. 1983; Christie et al. 1985; Halvorson and Black 1974). Also contributing to the seep development are periods of above-normal precipitation, restricted surface and subsurface drainage due to the building of roads and pipelines, large snowdrifts, gravelly and sandy soils, obstructions across natural drainageways (e.g., roads), uncapped or poorly cased artesian water wells, leaky ponds and dugouts, and crop failures. These factors, combined with certain geologic conditions, can result in saline seeps years after vegetation has changed. Water conservation practices in the southern Great Plains, such as level bench-terraces and unlined irrigation or drainage canals, have contributed to the development of saline seeps by increasing soil water concentrated at specific points on the landscape (Berg et al. 1986, 1991; Naney et al. 1986). During the 1930s, drought fostered use of moisture-conserving management systems. Using an area with a substantial amount of saline seeps, the Palmer Drought Index (PDI) illustrates (Fig. 18-1A) the drought severity during that period (NOAA 2007). Data in the PDI are normalized by region and represent a water supply and demand based on temperature, precipitation, and Thorntwaite potential ET with antecedent moisture, and some soil factors (NOAA 2005). Land owners who lived through the drought of the 1930s became conservative regarding soil-water management and did not change readily from a crop-summerfallow management system. The PDI also can be used to show periods of excess moisture. In Fig. 18-1B, an illustration of the pluvial (wet climatic) conditions is given in which saline seeps have developed. Saline seeps developed during the 1960s and 1970s, even though the pluvial conditions were not as intense as the drought was dry, as can be seen by contrasting the two PDI illus-

MANAGEMENT OF DRYLAND SALINE SEEPS

563

FIGURE 18-1. (A) Palmer drought index (PDI) illustrating the severity of the 1934–1939 conditions in the northern Great Plains. (B) PDI illustrating the pluvial conditions in the 1970s when saline seeps were occurring. From NOAA (2007).

564

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

trations. Many people still consider the crop-summerfallow system to be a lower risk than other management systems.

TYPES OF SALINE SEEPS Miller et al. (1981) discussed geologic conditions in the northern Great Plains that produced saline seeps. The soil parent materials vary from glacial till deposited over shales to highly stratified, water-lain geologic deposits. Brown et al. (1983) diagrammed several geologic conditions that can result in seeps in the northern Great Plains (Fig. 18-2). Seeps generally develop on side slopes and/or toe slopes of rolling to undulating topography, where permeable material lies above less permeable strata (e.g., an aquifer above an aquitard). Such anisotropic conditions are conducive to development of perched water tables and lateral flow to discharge areas. Characteristics of different seep types are: 1. Geologic outcrop seep. The recharge area lies above material of low hydraulic conductivity (HC), such as shale, dense till, or clay. Soil above the low-HC layer may vary in texture and thickness. Most outcrop seeps expand laterally and downslope. Only limited seep expansions occur upslope. The seeps develop on planar or convergent slopes and are frequently crescent-shaped, illustrating the convergence of water. 2. Coal seam seep. The recharge area lies above coal, which, in turn, overlies clay of low HC. This lower clay unit, called underclay, typically is present in coal deposits. Soil above the coal seam varies in texture. Water rapidly moves laterally through the coal material. Seepage occurs where truncated coal beds crop out. Coal seeps typically expand laterally, forming sharp discharge areas that are wettest at the discharge points and are drier downslope. 3. Glaciated Fort Union seep (Brown et al. 1983). In this variant type of geologic outcrop seep, the recharge area of glacial till lies above sandstone, siltstone, lignite, and clay strata of the Fort Union Formation of Wyoming, Montana, and parts of adjacent states. As in other formations, water in the recharge area enters permeable strata to form a water table above a low-HC zone at a discontinuity, such as a change in four strata in the Fort Union Formation. Seepage water moves downslope to glacial till. The till has low HC and truncates the permeable zone, causing the water table to rise with some artesian pressure. Seep expansion is diffuse, moving upslope, downslope, and laterally. 4. Textural change seep (soil restrictive layers). The recharge area is in a material with low HC at depth; often, the same texture has a high HC in the surface (soil structure aquifer) but has a low HC below. In the

MANAGEMENT OF DRYLAND SALINE SEEPS

565

FIGURE 18-2. Schematic diagrams illustrating seven geologic conditions for saline seep development. From Brown et al. (1983).

566

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

soil above the zone of low HC, water moves through the rootzone to the zone of low HC. The water then moves laterally downslope until the water movement slows, which causes the water table to rise to or near the soil surface. Often downslope areas have a finer texture than upslope areas. Seep expansion is lateral and downslope. The upper boundary of this seep type is more diffuse than most seeps. 5. Slope change seep. The recharge area lies above geologic material of low HC. Soil above the zone of low HC varies in texture. Water moves through the rootzone to the zone of low HC, then laterally downslope to where the slope decreases. The reduced slope gradient at this point causes water to move slower, resulting in a water table at or near the surface. Seep expansion is mostly lateral and downslope with some upslope expansion. 6. Hydrostatic pressure seep. The recharge area lies above geologic material of low HC. Soil above the dense layer varies in texture. Water moves through the rootzone to the zone of low HC and then laterally downslope to a zone of low HC above the saturated zone. At this point, the confined water is under hydrostatic pressure. This often forces water through fractures to the surface of the soil and causes a seep. For example, the eastern edge of Glacial Lake Agassiz has recently been found to have several miles of such saline seeps. Such seeps expand mostly downslope and out onto the toe slope where the seep can move laterally. The recharge area may be at a greater distance and at a higher elevation than for other types of seeps. 7. Pothole seep. The recharge area has potholes or poorly drained areas that lie above material of low HC. Water moves through slowly permeable material in a pothole to a zone of even lower HC, then moves downslope and laterally as throughflow to other potholes or a stream valley where it discharges, creating a seep. The soils reflect the movement of the water and are often Calciaquolls or soils with calcareous B-horizons. The seep primarily expands laterally. Tillage aids in moving salinity upslope by capillary action. Richardson et al. (2001) describes the groundwater flow around potholes in more detail. Their discharge wetlands have edge-focused salinity development that are consistently saline seeps. An additional effect is the freezing impact of sodium sulfates as mirabilite (hydrated sodium sulfate). Mirabilite has 10 waters of hydration and freezes even before ice forms in the pond or the groundwater has enough salt content. Magnesium sulfates maintain their mobility and, by not freezing, are separated from the sodium sulfates (Richardson et al. 1990). Seeps are also created by three additional field conditions by land management that may locally affect agriculture: (1) prolonged sewage lagoon leakage (Griffin et al. 1985) has been found in numerous locations; (2) irri-

MANAGEMENT OF DRYLAND SALINE SEEPS

567

gation and drainage ditches that have low gradients create a local recharge to the adjacent land (Skarie et al. 1986), and (3) oilfield brine pits that leak NaCl into adjacent landscapes (Foss et al. 1985). These latter types of seeps can vary widely in expression in the field. Understanding the geology and circumstances that cause a saline seep to form helps in designing ways to control or prevent them. Agronomic practices generally work well to decrease some seep types (e.g., geologic outcrop, coal seam, glaciated Fort Union, texture change, and slope change seeps). Agronomic practices need to be combined with drainage and landleveling for mitigation of hydrostatic pressure and pothole seep control.

WATER QUALITY ASSOCIATED WITH SALINE SEEPS As water passes through the soil profile toward the perched or permanent water table, salts dissolve and move downward or laterally with the water. Hydrologic studies show that seeps are generally sustained by local recharge areas (Doering and Sandoval 1976a; Halvorson and Black 1974; Halvorson and Reule 1980; Hendry and Schwartz 1982; Naney et al. 1986). Numerous studies document movement of soluble salts and NO3-N toward and into shallow water tables. Ferguson and Batteridge (1982) have estimated that as much as 90 Mg/ha of salt has migrated toward the groundwater table in glacial till soils of north-central Montana after several decades of grain production. Christie et al. (1985) reported that the soil-profile salinity of cultivated land decreased more than that of an adjacent native noncultivated area, indicating salt movement to deeper depths. Doering and Sandoval (1981) reported that a drained seep area had lost salt and 50 kg NO3-N/ha. The data in Table 18-1 show the chemical composition of waters associated with several saline seeps in the northern and southern Great Plains. The shallow groundwater often associated with saline seeps is unsuitable for consumption by humans and livestock due to high salt and NO 3 (0.7 mmol/L) level, and unsuitable for irrigation due to total salt concentration. Calcium, magnesium, and sodium are the dominant cations, and sulfate is the dominant anion in most of the shallow groundwater associated with saline seeps. Compared to sulfates, chlorides exist in water and soil at relatively low concentrations in the northern Great Plains. They occur at slightly higher concentrations in the southern Great Plains. Soils in seep areas are generally in equilibrium with gypsum, lime, and other Ca-Mg sulfate minerals (Brun and Deutch 1979; Doering and Sandoval 1981; Oster and Halvorson 1978; Timpson et al. 1986; Timpson and Richardson 1986). Freezing of the water in saline seeps that are dominantly sulfates concentrates the Na in the ice and the Mg in the groundwater. Therefore, the Na is found near discharge areas, and the Mg is

568

TABLE 18-1. Chemical Composition of Waters Associated with Saline Seeps in the U.S. Great Plains Location (1)

pH (2)

EC (dS/m) (3)

Caa (4)

Mga (5)

Na (6)

HCO3a (7)

NOa3 (8)

Cla (9)

SO4a (10)

MT recharge MT seep MT seep MT seep MT recharge ND seep ND seep OK seep

8.4 8.2 7.9 8.4 8.2 3.7 4.6 8.1

5 9 14 26 7 10 8 5

7 8 10 1 3 9 9 15

11 21 37 108 21 36 30 16

18 66 109 211 39 59 40 26

3.8 9.8 8.1 4.0 2.4 — — —

4.3 0.4 29.5 5.4 6.2 5.7 4.7 0.6

0.7 0.8 2.6 7.6 11.2 2.1 2.5 12.3

21 52 80 225 44 70 55 27

OK seep

8.2

3

3

17

13

—

—

16.0

15

a

Chemical elements in mmol/L.

EC, electrical conductivity; MT, Montana; ND, North Dakota; OK, Oklahoma.

Reference (11)

Halvorson and Black (1974) Halvorson and Black (1974) Halvorson and Black (1974) Miller (1971) Miller (1971) Doering and Sandoval (1981) Doering and Sandoval (1981) Berg et al. (1986) Naney et al. (1986)

MANAGEMENT OF DRYLAND SALINE SEEPS

569

mobilized some distance downslope or gradient because of its continual travel in the winter (Richardson et al. 1990). Researchers in the northern Great Plains have concluded that the NO3 in groundwater originates mainly from exchangeable NH4 of geologic origin located deep in the profile and from NO3 that comes from mineralized organic matter that is leached from the rootzone during periods of summerfallow (Doering and Sandoval 1981; Hendry et al. 1984; Power et al. 1974). The NH4 is oxidized to NO3. Little, if any, of the NO3 had its origin as fertilizer N because little fertilizer N was applied by dryland farmers in the northern Great Plains before the early 1970s, when saline seeps were recognized as a problem.

IDENTIFICATION OF RECHARGE AND DISCHARGE (SEEP) AREAS Early detection and diagnosis of a saline-seep problem may allow a farmer to change current cropping systems to minimize the damage. Postponing the use of control practices obviously leads to a problem that is more difficult to control. Visual Assessment Brown (1976) described multiple visual symptoms of impending development of saline seeps: vigorous growth of kochia (Kochia scoparia L.) or other weeds after grain harvest in areas where the soil normally would be too dry to support weeds; the presence of salt crystals on the soil surface; prolonged wetness of the surface in localized areas after rainfall; the slipping of tractor wheels or the bogging down of equipment in areas of a field, or the seepage of water into wheel tracks, with salt crystals becoming visible as the soil dries; crop growth accompanied by lodging in areas of the field that previously produced normal growth; increased infestations of foxtail barley (Hordeum jubatum L.); stunted or dying trees in a shelterbelt or windbreak; and poor seed germination. Field Assessment of Soil Salinity Electrical conductivity (EC) methods for measuring soil salinity have been developed to identify areas where saline seeps may develop or extend the boundaries of existing saline seeps. Halvorson and Rhoades (1974, 1976) and Halvorson et al. (1977) used a four-electrode resistivity technique to characterize salinity levels of the soil profile and to identify recharge areas and incipient and existing saline seep areas (Fig. 18-3).

570

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 18-3. Typical four-probe electrical conductivity (ECa) readings as a function of soil depth in saline seep recharge area, encroaching saline seep area, and saline seep itself. From Halvorson and Rhoades (1974).

Rhoades and Halvorson (1977) also used EC methods to delineate saline seep areas and map soil salinity (see also Chapters 9–11 of this manual). In recent years, the use of Geonics EM-38 or similar electromagnetic induction (EMI) devices has replaced almost all other EC measurement instruments because of the speed and ease of use (Cameron et al. 1981, Wollenhaupt et al. 1986). Currently, EMI instruments coupled with GPS and GIS instruments can locate and map saline seeps accurately at walking or slow driving speeds, provided one does not get stuck. Additionally, the EMI method can also be used to verify areas of high and low salinity in the field without laboratory analyses. Saline seeps generally have high levels of salinity at the soil surface that decreases with depth. The surface concentration is created by matric or capillary movement of water from the wetter soil to the dry soil surface that is undergoing evaporation or freezing, which creates a tension that moves more water to the surface. The salts are concentrated when the water evaporates (Richardson 2005). Developing or potential seep areas, however, generally have low to medium levels of salinity at the soil surface, with higher salinity at 30- to 90-cm depth and lower salinity below. Salinity generally increases gradually with increasing depth in recharge areas.

MANAGEMENT OF DRYLAND SALINE SEEPS

571

Locating Recharge Areas The size of recharge areas must be identified and delineated before effective treatments for controlling saline seeps can be designed. Generally, recharge areas are a short distance (180 m to 600 m) upslope from the discharge or seep area. If gravel beds and sandy soils are involved, recharge areas may be within 30 m. The recharge area usually is directly upslope or at an angle across the slope from the discharge area (Brown et al. 1983). The following methods may be useful for determining the location and size of recharge areas. 1. Topographic maps, soil maps, and geologic information. Topographic maps present a clear picture of slopes. Saline seeps are created by water moving via gravity that preferentially concentrates in certain slope positions that can be deduced from the topography. Soil surveys can be used to locate sandy or gravelly soils upslope from the discharge area, as well as poorly drained areas, such as potholes. A recent tool that currently provides soil information for almost any site in the United States is the Web Soil Survey program (NRCS 2007). Because the program uses recent orthophotographs, the actual color and tonal shades on the image and the soils themselves can be used to identify the recharge area and determine some characteristics about the discharge (saline seep) site. The Web Soil Survey program will be coupled with a topographic map selection in the near future. Geologic maps can provide information on subsurface stratification, the type of saline seep, and depths to permanent groundwater tables. 2. Soil moisture probes and test holes. If a seep is surrounded by elevated topography on several sides, a soil moisture probe (Brown 1958) can be used to identify the recharge area relative to the seep by locating abnormally wet soil in one general direction. Augering or coring machines can be used to examine soil profiles to greater depths. Each drilled hole should be carefully logged during drilling. The depths at which dense or impermeable materials (aquitards), such as clay and shale, or highly permeable materials (aquifers), such as sand, gravel, silt, and lignite, are encountered should be recorded. The depth to the water table should be noted and the hole cased with perforated pipe so that depths of the water table can be periodically monitored and water samples can be collected. Information collected from the test holes, including well log data, water depth, and salinity measurements, can be combined with visual site observations and topography to delineate the recharge area. Often, soil moisture probes and a few well-placed test holes will provide the most economical way to locate a recharge area. 3. Visual inspection. When soil survey maps, drill rigs, and equipment for measuring soil salinity in the field are unavailable, visually locate the

572

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

upslope area, direction of seep expansion, and upslope factors that may contribute water to the discharge area. Bear in mind that (1) the recharge area is higher in elevation than the seep area; (2) the recharge area is generally within 600 m of the seep area; (3) saline seeps in glacial till areas generally expand laterally and upslope toward the recharge area; (4) saline seeps in nonglaciated areas tend to expand laterally and downslope away from the recharge area; and (5) if the seep does not begin to dry up within 2 to 3 years of implementing control measures, such as planting alfalfa or grasses or annually cropping the suspected recharge area, the boundary of the recharge area was incorrectly identified, the recharge area was larger than estimated, or the seepage water may be coming from an artesian source.

METHODS FOR CONTROLLING SALINE SEEPS Since seeps are caused by water moving below the rootzone in the recharge area, the saline seep problem will not be permanently solved unless control measures are applied to the recharge area to interrupt the throughflow water source. Two procedures for managing seeps are (1) mechanically drain ponded surface water before it infiltrates and intercept lateral flow of subsurface water with drains before the water reaches the discharge area; and (2) use plants, such as phreatophytes, to consume the water before it percolates below the rootzone or escapes from the recharge area. A phreatophyte is a deep-rooted plant (such as desert willow, mesquite, or alfalfa) that can draw water from the water table. In general, these plants use copious amounts of water if it is available. Drainage Undulating, near-level land with poor surface drainage (potholes) can create localized recharge areas for saline seeps; these “hummocky” areas create local groundwater flow areas (Richardson et al. 2001). Runoff takes place after rainfall and snowmelt, causing these areas to fill with water temporarily. Where possible, surface drains can be installed to prevent the temporary ponding of surface water. In flatter topography as in the Red River Valley of the northern Great Plains, land-leveling works by removing the depression entirely. Drainageways, such as culverts under roadbeds, should be kept clear of debris and sediment so they do not cause temporary ponding of surface water. In the central Great Plains, level bench-terraces serve as temporary water impoundments that may be contributing water to saline seeps (Berg et al. 1986, 1991; Naney et al. 1986). Their use may need to be evaluated if saline seepage is a problem.

MANAGEMENT OF DRYLAND SALINE SEEPS

573

Drainage studies have shown that hydraulic control can be accomplished quickly with subsurface interceptor drains on the upslope side of the seep area (Doering and Sandoval 1976a; Sommerfeldt et al. 1978). However, a suitable outlet for disposal of saline water must be available. Outlet considerations include easement for drainage water transport across intervening lands and the effect of drainage waters on the quality of receiving streams or reservoirs. Because seep waters are saline and typically high in nitrate, disposal into downstream surface waters or groundwaters is difficult due to physical and legal constraints and costs. Therefore, subsurface drainage is generally not a satisfactory solution to the problem. The best approach is to use the soil-water for crop growth when the water is in the rootzone of the recharge area and is relatively nonsaline. Mole drains have been used in Alberta, Canada, to maintain water tables at a sufficient depth to prevent the salt accumulation on the soil surface (Sommerfeldt et al. 1978). Procedures for using mole-type drains are specific to the site. With moist, cohesive, fine-textured soils and shallow water tables (100 cm), the drains work well when installed on proper grade. Such drains may not work in noncohesive soils or with a water table that is 100 cm below the surface of the soil (Sommerfeldt 1976). Oosterveld (1978) used seep discharge water to irrigate the recharge area, thus recycling the salts. Limited water supplies for irrigation, the cost of an irrigation system to deliver the water, and the build-up of soil salinity in the recharge area may reduce the usefulness of this technique. Agronomic Practices Hydraulic control of saline seep areas can be achieved by growing plants that use available soil water supplies in the rootzone of the recharge area or can draw water from the water table. First, the recharge area must be delineated, and plants that maximize the use of soil-water and minimize deep percolation must be adopted (notably phreatophytes). In Australia, plantations of eucalypts are capable of intercepting groundwater and reclaiming sand plain seeps within 5 years (Eastham et al. 1993; George 1991). Evidence from one site suggests that 200 eucalypts were sufficient to intercept and transpire approximately 1,000 m3 of brackish to saline groundwater. This is a strategy that may work in the Great Plains of the United States and is being explored by the Agroforestry Center of the U.S. Forest Service in Lincoln, Nebraska. Seeded in recharge areas, alfalfa (Medicago sativa L.), a pheatophyte, is one crop that helps to control seep discharge areas hydraulically (Brown et al. 1983; Brun and Worcester 1975; Halvorson and Reule 1980). In Montana, Halvorson and Reule (1980) found that alfalfa extracted more water from the soil profile than did native grass sod or small grain crops (Fig. 18-4). Alfalfa utilized more soil water to a depth of 3.0 m than did

574

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 18-4. September 1976 soil-water profiles of a saline seep recharge area in native range sod, alfalfa (seeded in 1973), spring wheat stubble, and summerfallow. From Halvorson and Reule (1980).

other crops. This effectively tapped a larger reservoir of the soil’s stored precipitation and thus reduced the potential loss of water via deep percolation. With the upslope recharge area in alfalfa, the saline seep area dried sufficiently to once again obtain normal yields (Halvorson 1984). In Colorado, alfalfa established in a terraced recharge area in 1984 hydrologically controlled an active saline seep (Halvorson 1988). By the autumn of 1985, the seep had dried sufficiently to allow the seep area to once again be worked with farm machinery. In 1987, three cuttings of alfalfa were harvested from the discharge area where only salt-tolerant weeds had grown in 1984. Brown and Miller (1978) and Miller et al. (1981) showed that alfalfa controlled saline seeps effectively, while Brown (1983) further showed that it took 7 to 8 years to recharge the dried soil profile to field-capacity water content, when a summerfallow-winter wheat-barley rotation followed 3 years of alfalfa (Table 18-2). Halvorson and Reule (1980) reported a rise in the level of the water table after a farmer reverted to a crop-summerfallow system of farming in the recharge area following several years of alfalfa production, during which hydraulic control of the seep area had been achieved and the seep area supported near-normal crop production.

MANAGEMENT OF DRYLAND SALINE SEEPS

575

TABLE 18-2. Total Soil-Water Content (0.0–4.6 m) at the End of Each Growing Season Following Three Years of Alfalfa Year (1)

Crop/Summerfallow (2)

Fall Soil-Water (mm H2O/4.6 m) (3)

Annual Precipitation (mm) (4)

1973 1974 1975 1976 1977 1978 1979 1980

Alfalfa (3rd yr) Summerfallow (no crop) Winter wheat Barley Summerfallow (no crop) Winter wheat Barley Summerfallow (no crop)

217 342 330 393 448 461 461 524

— 278 563 371 363 418 208 380

1980

Estimated field capacity

573

—

From Brown (1983).

These studies indicate that even though a saline seep area has been controlled, reclaimed, and returned to normal crop production, a farmer cannot permanently return to a conventional crop-summerfallow system of farming in the recharge area. The soil water needs to be managed continually to prevent the recurrence of saline seep. Other work has shown that small grain crops, such as rye, wheat, and triticale can be used to control saline seep areas (Alberta Agriculture 1986; Bramlette 1971; Halvorson and Reule 1976; Holm 1983; Steppuhn and Jenson 1984). Using annual small-grain cropping systems to control seep discharge areas hydraulically is slower than using alfalfa because less soil water is extracted and rooting depths are shallower. Oil-seed crops that are deeper-rooted than small grains, such as safflower and sunflower (Table 18-3), can help to deplete the stored soil water to greater depths, thereby increasing the soil capacity to store precipitation between crops or during summerfallow periods. Black et al. (1981) describe several dryland cropping strategies for controlling saline seeps in the northern Great Plains. They suggest using intensive, flexible cropping systems with adapted crops and soil, water, and crop management practices to improve the crop–production-water-use relationship enough to eliminate or reduce the need for summerfallow. Flexible cropping involves planting a crop only in years when soil water and precipitation are expected to be sufficient to produce an economic crop yield. Each year, based on data regarding soil-water and expected precipitation during the growing season, the farmer decides to

576

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 18-3. Rooting Depth and Soil-Water Use by 11 Dryland-Grown Crops Fort Benton, Montana

Culbertson, Montana

Rooting Depth (m) (2)

Soil-Water Use (mm) (3)

Rooting Depth (m) (4)

Soil-Water Use (mm) (5)

Alfalfa (1st yr) Alfalfa (4th yr) Sanfoin (1st yr) Sanfoin (4th yr) Russian wild rye (1st yr) Russian wild rye (4th yr) Sweet clover (1st yr) Sweet clover (2nd yr) Safflower Sunflower Winter wheat Rapeseed Spring wheat Barley

2.1 5.5 1.5 4.0 2.1 3.0 1.8 2.7 2.2 2.0 1.8 1.5 — 1.4

178 666 150 561 318 475 276 403 249 206 200 170 — 190

— — — — — — — — 2.1 — 1.6 — 1.2 1.1

— — — — — — — — 229 — 190 — 152 135

Corn

1.2

94

—

—

Crop (1)

From Black et al. (1981).

crop or summerfallow the area where the seep occurs (Alberta Agriculture 1986; Brown et al. 1981; Naney et al. 1986). Recropping or annual cropping is ill-advised when less than about 76 mm of soil-water is available at planting time (Alberta Agriculture 1986; Black and Ford 1976). Farmers can use a moisture probe to determine soil moisture profiles (Brown 1958), or they may determine soil water content in another way. Halvorson and Kresge (1982) developed a computer model, FLEXCROP, to help farmers select the best cropping and soil management strategies for wheat (Triticum aestivum L.), barley (Hordeum sativum, Jess.), oats (Avena sativa L.), and safflower based on stored soil-water and expected precipitation. Weed control and soil fertility are also critical factors in developing flexible dryland cropping systems. Black et al. (1981) reported that crops grown under annual cropping systems used an average of 75% to 81% of the precipitation received

MANAGEMENT OF DRYLAND SALINE SEEPS

577

between crop harvests within a grass barrier system. Conventional spring wheat-summerfallow systems used only 40% (Table 18-4). The amount of unused available water between crops averaged 473 mm for spring wheatsummerfallow systems and only 72 mm to 98 mm for annual cropping systems. These data show that more water, nitrates, and dissolved salts can be moved below the rootzone with a spring-wheat-summerfallow system than with an annual cropping system. Adequate fertility is essential for optimizing yields with annual cropping systems (Black et al. 1982; de Jong and Halstead 1986; Halvorson et al. 1976; Schneider et al. 1980). If intensive, flexible cropping systems are to succeed, more efficient methods for storing soil-water during fallow periods must be found. In the northern and central Great Plains, supplies of soil-water can be increased by controlling the growth of weeds and volunteer grain after harvest, leaving standing stubble to trap snow, using annual or perennial barriers or windbreaks for snow trapping, and using reduced- or no-tillage cropping systems (Black and Siddoway 1976; Nicholaichuk and Gray 1986; Smika and Whitfield 1966). All of these practices enhance the efficiency of soilwater storage. However, more intensive cropping systems than the conventional crop-summerfallow system must be used. Otherwise, the development of saline seeps will intensify. Cropping Strategies Crops need different amounts of water to produce an economical yield because they have different rooting depths and water extraction patterns. Black et al. (1981) reported that safflower in Montana used more soil water and withdrew water from greater depths in 1 year than any other annual dryland crop (Table 18-3). Alfalfa used only slightly less water the first year than safflower and sweet clover (Melilotus officinalis L.), but its ability to use precipitation plus soil-water from progressively deeper depths in successive years makes it the best crop to use first to hydrologically control seep recharge areas. Crops listed in order of decreasing rooting depth and soil water use are alfalfa, sweet clover, safflower, sunflower, winter wheat, rapeseed (Brassica napus L.), spring wheat, barley, and corn (Zea mays L.). To decide the following crop, knowledge of the amount and depth of soil-water depleted by the previous crop is needed. Crops should be grown in sequential order with increasing rooting depths until the depth and amount of soil water removed exceeds soil water recharge during fallow periods (Black et al. 1982). Summerfallow should be used only when needed, such as after planting alfalfa or safflower, or when less than 76 mm of soil water exists at planting. Crops must be rotated in a sequence that avoids weeds, diseases, and insect infestations. Rotating oilseed crops and small grain crops allows

578

TABLE 18-4. Average Precipitation Use Efficiency (PUE) per Cropping Sequence, as Influenced by Cropping System Within a Tall Wheat Grass Barrier System over a 12-Year Period Annual Grain Yieldc

Cropping System (1)

WUEc,d

a

Number of Crops per Year (2)

Total Precipitation per Crop (mm) (3)

Total Water Use per Crop (mm) (4)

PUEb (%) (5)

Without Nitrogen (kg/ha) (6)

With Nitrogen (kg/ha) (7)

Without Nitrogen (kg/ha-mm) (8)

With Nitrogen (kg/ha-mm) (9)

1.00

396

322

81

1328

1794

3.4

4.5

Annual cropping 6WW-B-S-B-WW-S-B 5SW-S-B-WW-B-WW-B-WW

1.00

394

296

75

993

1822

2.5

4.6

4SW-S-B-WW-S-SW-B-WW-B

1.00

390

318

82

969

1590

2.5

41

0.66

569

333

59

997

1416

2.6

3.7

WW-F

0.50

788

404

51

1019

1247

2.6

3.1

SW-F

0.50

786

313

40

853

1065

2.2

2.7

Three-year rotation SW-WW-F Crop-summerfallow

WW  winter wheat, SW  spring wheat, B  spring barley, S  safflower, F  Summerfallow a

Water use per crop is based on soil-water use to 120-cm depth plus precipitation received from seeding to harvest. PUE  [(total water use per crop)/(total precipitation received per crop)]  100 c Applied nitrogen of 34 kg N per ha each crop year d WUE  water use efficiency  [(grain yield/ha)/(total precipitation/crop rotation)]  100 b

From Black et al. (1981).

MANAGEMENT OF DRYLAND SALINE SEEPS

579

grass herbicides to be used, helping to control the build-up of grassy weeds in the small grain crops (Berg et al. 1986; 1991; Naney et al. 1986). Soil fertility is almost as important as water in an annual cropping system. As cropping frequency increases, the need for N increases and responses to P fertilizer depend on the level of soil P and crop N needs (Halvorson and Black 1985). Nitrogen needs should be balanced carefully with expected water supplies and the potential yield of the crop. One should also remember that N fertilizers are labile salts. If overapplied, they add to the saline seep problem. Strict adherence to a crop-summerfallow rotation restricts farmers to a fixed cropping system with limited flexibility to adjust cropping patterns to fit available water supplies. Selection of alternate cropping strategies to use available water supplies effectively requires knowledge of the amount of water available at any given time, potential ET requirements and rooting depths of adapted crops, and expected growing-season precipitation. Knowledge of the depth to some restricting or impermeable geologic strata and water table is essential if a cropping strategy to control or prevent saline seeps is to be developed. Soil surveys help identify and evaluate these conditions.

RECLAMATION OF CONTROLLED SALINE SEEP AREAS Before reclaiming a saline seep area, water flow from the recharge area must be reduced so that the water table in the seep area is deep enough to prevent salts from moving up by capillary action into the rootzone. If a saline water table is less than approximately 90 cm below the surface, salts can move to the surface by capillary action. Water table depth often varies during the year and is shallower in spring and early summer than during the rest of the year. Figure 18-5 illustrates the general relationship between the water table depth and soil salinity in the upper 30 cm of soil (Halvorson and Rhoades 1976). Observation wells should be installed at strategic locations in recharge and seep areas to monitor water tables. A drill rig is needed to install deep wells in recharge areas, but a tractor-mounted post-hole auger or bucket auger can be used to install wells that are less than 180 cm deep. The water table level should be monitored monthly. A rising water table that persists into the summer months indicates that cropping practices should be intensified to increase use of soil-water. The results of research and the experiences of farmers indicate that reclamation occurs quite rapidly (Brown and Miller 1978; Halvorson 1984, 1988; Halvorson and Reule 1976, 1980). If the water table depth in the seep area exceeds 150 cm, reclamation procedures to remove salts from the rootzone can proceed. The rate of reclamation depends on the amount of

580

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 18-5. Four-probe soil electrical conductivity (ECa) readings as a function of water table depth in a saline seep area. From Halvorson and Rhoades (1974).

precipitation available to leach salts. Therefore, practices that enhance the downward movement of water in the salt-affected area, such as snow trapping or summerfallowing, accelerate reclamation. Summerfallowing can be used during reclamation to help increase downward leaching of salts from the soil profile. None of these practices will be effective, however, until hydrologic control is achieved in the recharge area and the water table is significantly lowered in the seepage area. Halvorson (1984) reported that soil salinity at the 0- to 30-cm depth was markedly reduced 2 years after a seep was ameliorated, and various crops could be grown. Adding straw mulch to reduce loss through evaporation from the fallow area helped to accelerate the removal of salt from the 0- to 90-cm depth. The application of gypsum did not accelerate reclamation, probably because sufficient naturally occurring gypsum had been precipitated in the soil profile during the saline seep formation. Because the salts present were Ca, Mg, and Na sulfates, neither the permeability nor the soil structure deteriorated during reclamation. Seven years after hydrologic control was achieved, soil salinity was still higher in the arrested saline seep than in adjacent areas (Fig. 18-6). Miller et al. (1981) reported the control of a serious saline seep problem (4 ha in size) in an 32-ha field near Fort Benton, Montana. “Ladak 65” alfalfa was seeded over the entire field in 1971, when the water table was 0.3 m below the surface in the seep area and 5.8 m below the surface in the recharge area. Six years later the water table had dropped to 3 m below the

MANAGEMENT OF DRYLAND SALINE SEEPS

581

FIGURE 18-6. Electrical conductivity (EC) of saturated soil extracts as a function of soil depth and time, for a saline seep area that was brought under hydrologic control and an adjacent non-seep-affected soil. From Halvorson (1984). surface in the seep area and to 8.5 m below the surface in the recharge area. Alfalfa roots had penetrated to a depth of 4.6 m and had depleted 48 cm of water from the soil profile in the recharge area. The receding water table in the seep area was caused by the reduced flow from the recharge area. Soil salinity in the seep area was 21.3 dS/m and 13.9 dS/m for the 0- to 30-cm and 30- to 60-cm depths, respectively, in 1971, and 4.3 dS/m and 6.3 dS/m in 1977. With the drop in the level of the water table in the seep, salts had been leached below 60 cm and the land once again supported economical crop production. In 1977, winter wheat yield in the area of the arrested saline seep was 70% of the surrounding area; in 1978, it was 100%. Halvorson and Reule (1980) controlled a saline seep that developed in 1971 near Sidney, Montana, where the land was farmed in a crop-summerfallow system. ‘Ladak 65’ alfalfa was seeded on about 80% of the recharge area in 1973. Figure 18-7 shows changes in the water table level before, during, and after alfalfa establishment. The water table in the recharge and seepage areas began to recede shortly after the alfalfa was seeded. By 1975, the surface of the seepage area was dry enough to cross with farm machinery. By 1977, the water table had receded to about 2.4 m below the soil surface in the seepage area. In the recharge area, one of two observation wells was dry by 1977. The depth of water in the other well had receded from 1.8 m to 3.3 m. Salinity in the top 30 cm of soil in the seep area decreased from 20 dS/m in 1972 to about 5 dS/m in 1978. Crop

582

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 18-7. Effect of alfalfa, seeded in summer of 1973, on water table depths in recharge and seep areas on a farm in northeast Montana. From Halvorson and Reule (1980). yields in the area of the arrested saline seep equaled average county yields after 3 to 4 years of hydrologic control (Table 18-5). Saline seeps would reappear if the phreatophytes were removed and the crop-summerfallow system was resumed. Halvorson (1984) reports that when a saline seep recharge area that had been hydraulically controlled was converted from alfalfa production back to a crop-summerfallow cropping system in 1979, soil salinity had within 3 years begun to increase at the 30- to 60-cm depth in the saline seep area. Data from the Fort Benton seep site indicated that 6 years of crop-summerfallow rotation had recharged the 4.6-m soil profile to field capacity and that continued crop-summerfallowing in the recharge area would reactivate the former saline seep area.

SOCIOECONOMIC CONCERNS Saline seeps do not respect property lines. A recharge area on one farmer’s property can supply water to a discharge area on a neighboring farm, or the seep discharge can contaminate a stream, natural drainageway, or farm pond. Except for small, uncomplicated seeps, such as geologic outcrop and coal seam seeps, most farmers need help in diagnosing their saline seep problem and in developing cropping systems or other control measures. When a recharge area is on an adjacent farm, landown-

MANAGEMENT OF DRYLAND SALINE SEEPS

583

TABLE 18-5. Yields of Several Crops Grown in Two Reclaimed Saline Seeps in 1978 and 1979, Compared to County Yields in Northeastern Montana Average County Yield (kg/ha)

Yield (kg/ha) Crop (1)

1978 (2)

1979 (3)

Average (4)

2,462 4,547 3,385 5,708

Seep A 1,586 2,135 1,577 9,834

2,024 3,341 2,481 7,771

Richland County 2,184 1,398 2,382 1,333 1,971 1,247 4,346 3,360

Spring wheat Barley Oats

2,426 3,861 5,273

Seep B 1,781 3,279 2,175

2,104 3,570 3,724

Roosevelt County 1,848 1,270 2,091 1,409 1,756 1,247

Corn (silage)

16,948

3,474

10,211

17,920

Spring wheat Barley Oats Alfalfa

1978 (5)

1979 (6)

11,200

From Halvorson (1984).

ers need to cooperate. Knowledgeable individuals or agencies can help by characterizing the problem and recommending control measures. Legislation could provide ways for farmers to form salinity control districts and achieve collectively what cannot be done individually. A saline seep is not just one farmer’s problem. Any loss of farmland decreases the nation’s food and tax base. Unless saline seeps are controlled, salty water from seeps can pollute fresh surface waters and add to groundwater salinity. The saline seep problem has political implications, involving such issues as subsidies, crop-acreage allotments, and landowner rights. Federal farm programs sometimes have inadvertently adversely affected the progress of saline seep control programs by restricting the acreage that can be planted with small grains or other crops to provide economic control of a saline seep problem. Hectares of summerfallow are often increased, magnifying the saline seep problem.

SUMMARY Saline seeps result from a combination of geologic, climatic, hydrologic, and cultural (land-use) conditions. The primary cause is a change

584

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

from grassland or forest to a cropping system, such as crop-summerfallow rotation, that allows rainfall in watershed recharge areas to move below the rootzone and provide seepage water. Seepage water then moves laterally downslope in permeable soil until it either encounters an impermeable soil and rises up to the soil surface or is discharged where the permeable soil stratum is discontinuous. Thus, saline seeps generally develop on side slopes and/or toe slopes of rolling to undulating topography, where permeable material lies above less permeable strata (e.g., an aquifer above an aquitard). A number of different geologic formations are associated with saline seeps in the Great Plains. Seeps can also be created by structures that concentrate water, such as canals and ponds. Identification and characterization of saline seeps requires consideration of topography, the underlying geology, and the ongoing land use (agricultural practices). Once a saline seep condition has been identified, there are several basic management strategies. Subsurface drainage can be intercepted with surface drains (if the drainage can be discharged to a surface water) or the land can be leveled so that saline water remains below the rootzone. There are practical constraints to these approaches and the best general approach is to use the soil-water for crop growth (when the water is in the rootzone of the recharge area and is relatively nonsaline). Management of crop rotations using phreatophytes and small grain crops is intended to balance crop water use and precipitation to minimize the seepage of precipitation past the rootzone, thus reducing the volume of potential seepage. This type of management can both reduce the potential for saline seeps and remediate saline seep conditions, but a return to management techniques, such as summerfallowing, will result a revival of the saline seepage problem.

REFERENCES Alberta Dept. of Agriculture and Rural Development (Alberta Agriculture). (1986). Dryland saline seep control, Alberta Agriculture AGDEX 518-11, Alberta Dept. of Agriculture and Rural Development, Edmonton, Alberta, Canada. Ballantyne, A. K. (1963). “Recent accumulation of salts in the soils of southeastern Saskatchewan.” Can. J. Soil Sci., 43, 52–58. Berg, W. A., Cail, C. R., Hungerford, D. M., Naney, J. W., and Sample, G. A. (1986). “Saline seep on wheatland in northwest Oklahoma,” in Proc. National Conference on Ground Water Quality and Agricultural Practice, Lewis Publishing Co., Chelsea, Mich., 265–271. Berg, W. A., Naney, J. W., and Smith, S. J. (1991). “Salinity, nitrate and water in rangeland and terraced wheatland above saline seeps.” J. Environ. Qual., 20, 8–11.

MANAGEMENT OF DRYLAND SALINE SEEPS

585

Black, A. L., Brown, P. L., Halvorson, A. D., and Siddoway, F. H. (1981). “Dryland cropping strategies for efficient water use to control saline seeps in the northern Great Plains, U.S.A.” Agric. Water Mgmt., 4, 295–311. Black, A. L., and Ford, R. H. (1976). “Available water and soil fertility relationships for annual cropping systems,” in Proc. Regional Saline Seep Control Symposium, Cooperative Extension Service Bulletin 1132, Montana State University, Bozeman, Mont. Black, A. L., and Siddoway, F. H. (1976). “Dryland cropping sequences within a tall wheatgrass barrier system.” J. Soil Water Conserv., 31, 101–105. Black, A. L., Siddoway, F. H., and Aase, J. K. (1982). “Soil moisture use and crop management (DRYLAND),” in Proc., 1st Annual Western Provincial Conference on Rationalization of Water and Soil Resources and Management, Lethbridge, Alberta, Canada, November 27–December 2. Bramlette, G. (1971). “Control of saline seeps by continuous cropping,” in Proc. Saline Seep-Fallow Workshop, Feb. 22–23, 1971, Great Falls, Mont., Highwood Alkali Control Association, Highwood, Mont. Brown, P. L. (1958). Soil moisture probe. U.S. Patent No. 2,860,515. ———. (1976). “Saline seep detection by visual observation,“ in Proc. Regional Saline Seep Control Symposium, Cooperative Extension Service Bulletin 1132, Montana State University, Bozeman, Mont., 59–61. ———. (1983). “Saline seep control and soil water recharge under three rotations following alfalfa.” Paper presented at Montana Chapter Soil Conservation Society of America meeting, Feb. 4–5, 1983, Bozeman, Mont. Brown, P. L., Black, A. L., Smith, C. M., Enz, J. W., and Caprio, J. M. (1981). Soilwater guidelines and precipitation probabilities for growing barley, spring wheat, and winter wheat in flexible cropping systems in Montana and North Dakota, Montana Cooperative Extension Service Bulletin 356, Montana State University, Bozeman, Mont. Brown, P. L., Ferguson, H., and Holzer J. (1987). “Saline seep development and control in Montana,” in A century of action, natural resource development and conservation in Montana, J. W. Bauder, ed., Montana Chapter of Soil Conservation Society of America, Bozeman, Mont. Brown, P. L., Halvorson, A. D., Siddoway, F. H., Mayland, H. F., and Miller, M. R. (1983). Saline-seep diagnosis, control and reclamation, USDA/ARS, Conservation Research Report No. 30, USDA, Washington, D.C. Brown, P. L., and Miller, M. R. (1978). “Soil and crop management practices to control saline seeps in the U.S. northern plains,” in Proc. of Meeting of Sub-commission on Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24. Brun, L. J., and Deutch, R. L. (1979). “Chemical composition of salts associated with saline seeps in Stark and Hettinger Counties, North Dakota.“ No. Dak. Agric. Exp. Sta. Farm Res., 37(1), 3–6. Brun, L. J., and Worcester, B. K. (1975). “Soil-water extraction by alfalfa.” Agron. J., 67, 586–589. Cameron, D. R., de Jong, E., Read, D. W. L., and Oosterveld, M. (1981). “Mapping salinity using resistivity and electromagnetic inductive techniques.” Can. J. Soil Sci., 61, 67–78.

586

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Christie, H. W., Graveland, D. N., and Palmer, C. J. (1985). “Soil and subsoil moisture accumulation due to dryland agriculture in southern Alberta.” Can. J. Soil Sci., 65, 805–810. Colburn, E. (1983). “Salt buildup in soil, slicing Texas yields.” Crops and Soil Mag., 35(4), 26. de Jong, E., and Halstead, E. H. (1986). “Field crop management and innovative acres,” in Proc. Moisture Management in Crop Production Conference, Nov. 18–20, 1986, Calgary, Alberta, Alberta Agriculture, Edmonton, Alberta. Doering, E. J., and Sandoval, F. M. (1976a). “Hydrology of saline seeps in the northern Great Plains.” Trans. ASAE, 19, 856–861, 865. ———. (1976b). Saline-seep development on upland sites in the northern Great Plains, USDA/ARS-NC-32, USDA, Washington, D.C. ———. (1981). “Chemistry of seep drainage in southwestern North Dakota.” Soil Sci., 132, 142–149. Eastham, J., Scott, P. R., Steckis, R., Barton, A. M., and Hunter, J. (1993). “Survival, growth and productivity of tree species under evaluation for agroforestry to control salinity in the wheatbelt of Western Australia.” Agroforestry Sys., 21, 223–237. Ferguson, H., and Batteridge, T. (1982). “Salt status of glacial till soils of northcentral Montana as affected by the crop-fallow system of dryland farming.” Soil Sci. Soc. Am. J., 46, 807–810. Foss, J. E., Richardson J. L., Prunty, L., Sweeney, M. D., Cudworth, D. K., and Hoag, B. K. (1985). “Task 1: Identification of salt-seepage areas from oilfield brine pits,” in Characterization of detrimental effects of salts and other chemical constituents carried in surface and subsurface waters from brine and drilling fluid disposal pits buried during oil development, E. C. Doll, J. E. Foss, G. J. McCarthy, and E. C. Murphy, eds., Water Resource Research Institute, OWRT Project Research Technology Completion Report, North Dakota State University, Fargo, N.D., 15–45. George, R. J. (1991). “Management of sandplain seeps in the wheatbelt of western Australia.” Agri. Water Mgmt., 19, 85–104. Griffin, D. M., Jr., Skarie, R. L., Maianu, A., and Richardson, J. L. (1985). “Effects of prolonged lagoon leakage on agricultural land.” J. Civil Engr., 4, 797–806. Halvorson, A. D. (1984). “Saline-seep reclamation in the northern Great Plains.” Trans. ASAE, 27, 773–778. ———. (1988). “Role of cropping systems in environmental quality: Saline seep control,” in Cropping strategies for efficient use of water and nitrogen, Special Publication No. 51, ASA/CSSA/SSSA, Madison, Wisc. Halvorson, A. D., and Black, A. L. (1974). “Saline seep development in dryland soils of northeastern Montana.” J. Soil and Water Cons., 29, 77–81. ———. (1985). “Long-term dryland crop responses to residual phosphorus fertilizer.” Soil Sci. Soc. Am. J., 49, 928–933. Halvorson, A. D., Black, A. L., Sobolik, F., and Riveland, N. (1976). “Proper management: Key to successful winter wheat recropping in northern Great Plains.” No. Dak. Agric. Exp. Sta. Farm Res., 33(4), 3–9. Halvorson, A. D., and Kresge, P. O. (1982). FLEXCROP: A dryland cropping systems model, USDA Production Research Report No. 180, USDA, Washington, D.C. Halvorson, A. D., and Reule, C. A. (1976). “Controlling saline seeps by intensive cropping of recharge areas,” in Proc. Regional Saline Seep Control Symposium,

MANAGEMENT OF DRYLAND SALINE SEEPS

587

Cooperative Extension Service Bulletin 1132, Montana State University, Bozeman, Mont. ———. (1980). “Alfalfa for hydrologic control of saline seep.” Soil Sci. Soc. Am. J., 44, 370–373. Halvorson, A. D., and Rhoades, J. D. (1974). “Assessing soil salinity in identifying potential saline-seep areas with field soil resistance measurements.” Soil Sci. Soc. Am. Proc., 38, 576–581. ———. (1976). “Field mapping soil conductivity to delineate dryland saline seeps with four electrode technique.” Soil Sci. Soc. Am. J., 40, 571–575. Halvorson, A. D., Rhoades, J. D., and Reule, C. A. (1977). “Soil salinity: Four-electrode conductivity relationships for soils of the northern Great Plains.” Soil Sci. Soc. Am. J., 41, 966–971. Hendry, M. J., McCready, R. G. L., and Gould, W. D. (1984). “Distribution, source and evolution of nitrate in a glacial till of southern Alberta, Canada.” J. Hydrol., 70, 177–198. Hendry, M. J., and Schwartz, F. (1982). “Hydrogeology of saline seeps,” in Proc., 1st Annual Western Provincial Conference on Rationalization of Water and Soil Resources and Management, Lethbridge, Alberta, Canada, November 27–December 2, 1982. Holm, H. M. (1983). Soil salinity: A study in crop tolerances and cropping practice, Saskatchewan Agriculture, Plant Industry Branch, Regina, Saskatchewan, Canada. Malcolm, C. V. (1982). Wheat belt salinity: A review of the salt land problem in southwestern Australia, Western Australian Dept. of Agriculture Technical Bulletin No. 52, Department of Agriculture and Food, Western Australia, Perth, Western Australia. Matheson, W. E. (1968). “When salt takes over.” J. Agric., South Aust., 71, 266–272. Miller, M. R. (1971). “Hydrology of saline-seep spots in dryland farm areas, a preliminary evaluation,” in Proc. Saline Seep-Fallow Workshop, Feb. 22–23, 1971, Great Falls, Montana, Highwood Alkali Control Association, Highwood, Mont. Miller, M. R., Brown, P. L., Donovan, J. J., Bergatino, R. N., Sonderegger, J. L., and Schmidt, F. A. (1981). “Saline seep development and control in the North American Great Plains: Hydrologic aspects.” Agric. Water Mgmt., 4, 115–141. Naney, J. W., Berg, W. A., Smith, S. J., and Sample G. A. (1986). “Assessment of ground-water quality in saline seeps,” in Proc. Agricultural Impacts on Ground water, Aug. 11–13, 1986, Omaha, Nebraska, National Water Well Association, Dublin, Ohio. Neffendorf, D. W. (1978). “Statewide saline seep survey of Texas.” M.S. thesis, Texas A&M University, College Station, Tex. Nicholaichuk, W., and Gray, D. M. (1986). “Snow trapping and moisture infiltration enhancement,” in Proc. Moisture Management in Crop Production Conference, Nov. 18–20, 1986, Calgary, Alberta, Alberta Agriculture, Edmonton, Alberta. National Oceanic and Atmospheric Administration (NOAA). (2005). “The Palmer Drought Severity Index,” www.drought.noaa.gov/palmer.html, accessed February 5, 2011. ———. (2007). “Climate data online,” www.ncdc.noaa.gov/oa/climate/online prod/drought/xmgr.html, accessed February 5, 2011. Natural Resources Conservation Service (NRCS). (2007). “Web soil survey,” http://websoilsurvey.nrcs.usda.gov/app/, accessed February 5, 2011.

588

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Olson, G. W. (1978). “Some observations on dryland saline seepage in several countries and some comments on causes, effects, and cures,” in Proc. Sub-Commission on Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24, Sec. 2, 7–29. Oosterveld, M. (1978). “Disposal of saline drain water by crop irrigation,” in Proc. of Meeting of Sub-commission of Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24. Oster, J. D., and Halvorson, A. D. (1978). “Saline seep chemistry,” in Proc. of Meeting of Sub-commission of Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24. Power, J. F., Bond, J. J., Sandoval, F. M., and Willis, W. O. (1974). “Nitrification in Paleocene shale.” Science, 183, 1077–1079. Rhoades, J. D., and Halvorson, A. D. (1977). Electrical conductivity methods for detecting and delineating saline seeps and measuring salinity in northern Great Plains soils, USDA-ARS Western Region ARS W-42, USDA, Washington, D.C. Richardson, J. L. (2005). “Soluble salts: Translocation and accumulation,” in Encyclopedia of soil science, 2nd ed., Vol. 2, R. Lal, ed., Taylor and Francis, New York, 1664–1665. Richardson, J. L., Arndt, J. L., and Enz, J. W. (1990). “Effects of freezing on sulfate salts in North Dakota soils and wetlands,” in Proc. International Symposium Frozen Soil Impacts on Agricultural, Range, and Forest Lands, Spokane Wash., March 21–22, 1990, K. R. Cooley, ed., CRREL Special Report 90-1, U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, N.H., 212–215. Richardson, J. L., Arndt, J. L., and Montgomery, J. A. (2001). “Hydrology of wetland and related soils,” in Wetland soils: Their genesis, morphology, hydrology, landscapes, and classification, J. L. Richardson and M. J. Vepraskas, eds., CRC Press, Boca Raton, Fla., 35–84. Schneider, R. P., Johnson B. E., and Sobolik, F. (1980). “Saline seep management: Is continuous cropping an alternative?” No. Dak. Agric. Exp. Sta. Farm Res., 37(5), 29–31. Skarie, R. L., Richardson, J. L., Maianu, A., and Clambey, G. K. (1986). “Soil and groundwater salinity along drainage ditches in eastern North Dakota.” J. Environ. Qual., 15, 334–340. Smika, D. E., and Whitfield, C. J. (1966). “Effect of standing wheat stubble on storage of winter precipitation.” J. Soil Water Conserv., 21, 138–141. Sommerfeldt, T. G. (1976). “Mole drains for saline seep control,” in Proc. Regional Saline Seep Control Symposium, Cooperative Extension Service Bulletin 1132, Montana State University, Bozeman, Mont., 296–302. Sommerfeldt, T. G., Vander Pluym, H., and Christie, H. (1978). “Drainage of dryland saline seeps in Alberta,” in Proc. of Meeting of Sub-commission of Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24. Steppuhn, H., and Jenson, D. (1984). “Barley can help control dryland salinity.” Crops and Soils Mag., 36(8), 22–23. Timpson, M. E., and Richardson, J. L. (1986). “Ionic composition and distribution in saline seeps of southwestern North Dakota, USA.” Geoderma, 25, 295–305.

MANAGEMENT OF DRYLAND SALINE SEEPS

589

Timpson, M. E., Richardson, J. L., Keller, L. P., and McCarthy, G. J. (1986). “Evaporite mineralogy associated with saline seeps in southwestern North Dakota.” Soil Sci. Soc. Am. J., 50, 490–493. Vander Pluym, H. S. A. (1978). “Extent, causes and control of dryland saline seepage in the northern Great Plains of North America,” in Proc. Sub-commission on Salt-Affected Soils, 11th International Soil Science Society Congress, Edmonton, Alberta, Canada, June 21–24. Wollenhaupt, N. C., Richardson, J. L., Foss, J. E., and Doll, E. C. (1986). “A rapid method for estimating soil salinity from apparent soil electrical conductivity measured with an above-ground electromagnetic induction meter.” Can. J. Soil Sci., 66, 315–321.

This page intentionally left blank

CHAPTER 19 PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS John Hedlund, William Evans, Jim Thomas, and Patrick H. Willey

INTRODUCTION Addressing salinity and water management over a wide geographical area requires a regional or project-level program that incorporates numerous local-scale projects. Such projects might include augmentation of water supplies, restructuring of water use, enactment of water-related laws, development of interstate and international agreements on water use, improvement of the technology used to apply irrigation water, and water management techniques (Fig. 19-1). The Economic and Environmental Principles and Guidelines for Water and Related Land Resource Implementation Studies (Water Resources Council 1983), informally known as the U.S. National Principles and Guidelines, provides concepts that can be used to identify, analyze, plan, and evaluate soil and water resources projects. Obtaining funding for water resource salinity management projects requires the consideration of economic conditions and current political objectives and policies. For a project-level salinity management program to succeed, federal, state, and local agencies, farmers, and the public must all cooperate. Sample objectives and actions are provided in Table 19-1. PLANNING REGIONAL MANAGEMENT PROGRAMS To plan and implement a regional soil and water salinity management program, the following steps must be taken: 1. Define the problem. 2. Analyze irrigation and drainage systems and management practices. 591

592

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 9-1. Project-level management options. 3. Develop alternative plans. 4. Compare alternative plans. 5. Select a plan for implementation. Defining the Problem One must go beyond saying that the main problem is downstream salinity or soil salinization and then quantify the problem. At a project

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

593

TABLE 19-1. Sample Objectives and Actions of a Regional Salinity Management Program Objective (1)

1. Improve surface water quality and reduce downstream salinity damages.

2. Reduce salinity damages to ground water.

3. Limit construction of additional water supply structures that adversely affect fish and wildlife. 4. Increase crop production.

5. Improve overall economic efficiency of agricultural production.

Action (2)

• Reduce salinity loading from irrigation return flows by improvements to irrigation systems and management that results in reduced diversion from surface water sources. • Provide for on-farm disposal of saline water. • Utilize drainage water on more salt tolerant crops. • Improve irrigation systems to reduce deep percolation and salt loading to aquifers. • Install pipelines in place of irrigation ditches prone to seepage. • Install impervious linings for irrigation ditches. • Reduce irrigation demands by improving irrigation efficiencies.

• Implement irrigation water management practices that minimize the accumulation of soil salinity. • Implement salinity management program that includes salinity monitoring capability for agricultural producers.

level, salinity and drainage problems are generally multifaceted and involve physical, economic, and environmental factors within a social/ political context. Physical factors To analyze how salt loading affects streams, the seepage and deep percolation reductions that will result from improved delivery and onfarm irrigation systems must be determined. For example, many saline

594

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

formations underlying soils have an unlimited salt supply, and saline subsurface return flows will continue to the river. If irrigation systems and management are improved so that seepage from canals and ditches and percolation from fields into the saline aquifer are reduced, then salt loading will be reduced proportionately. In other cases, deep percolation or seepage may displace residual saline ground water. Projects that control the saline water’s rate of displacement could be used. The analysis is based on the technical ability to estimate accurately the reduction in seepage and deep percolation occurring with various approaches to irrigation improvement. This translates into salinity changes. Basin-wide models are maintained to balance water and salt contributions from each study unit. Salt and water budgets form the nucleus for evaluating plan alternatives. Physical budget analysis is never straightforward. Concurrent flow and water quality records for all desired locations in the basin over a representative time period are seldom available. Further, more than one mechanism could be at work to increase salinity between measuring points. Such mechanisms include irrigation, natural runoff, mining activity, and point discharges. Moreover, the number of these mechanisms varies over the years in response to hydrologic and economic activity. In summary, data-intensive interdisciplinary effort is needed to identify and quantify the salinity-producing mechanisms in an irrigated river basin. Many diverse pieces of data must be integrated to understand overall the movement of water and salt through the system. Sufficient evidence should be obtained from water and salt budgets to conclude whether salinity will be reduced by implementing the proposed salinity control program. Economic factors A salinity control program is evaluated similarly to other kinds of water-related land resource developments (Hedlund et al. 1978). Since a program affects on-farm water users and downstream water users, the extent to which both groups are benefited should be analyzed. Also, several levels of resource development should be analyzed so that the optimal level can be determined. This permits the selection of the best combination of control measures within an area and the best combination of areas to treat within a river system (Evans et al. 1981). In formulating a program, the physical effects of each alternative are measured and translated into economic terms so that each alternative’s benefits can be compared with its cost (Fig. 19-2). To identify the optimal level of economic development, the point of maximum net benefit (point 2 shown in Fig. 19-2), which is when the incremental benefit equals the incremental cost, is determined.

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

595

FIGURE 9-2. Example of benefits and costs for alternate plans.

Environmental factors Improving the use and management of water for conservation makes full use of limited water supplies, increases in-stream flows for fish and wildlife, and reduces the salinity of downstream water. Irrigation improvements may adversely affect wildlife by eliminating wetlands that rely on irrigation water to satisfy the hydrology requirement of a wetland determination (irrigation-induced wetlands). The processes that link the area’s physical, chemical, and biological elements—including fish and wildlife, threatened and endangered species, cultural resources, and the value of prime farmland—are analyzed. The effects of proposed actions, as well as the options and tradeoffs, are identified and measured. Water users, the concerned public, and state and

596

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

federal agencies can then evaluate the choices for that area between water quality (salinity) and fish and wildlife habitat (Skogerboe et al. 1979). Analyzing Irrigation Practices Alternative irrigation systems used within the project area are listed, and how those systems and associated irrigation practices are adapted to specific site conditions are described. Developing alternative plans Available saline water and soil treatment opportunities (Fig. 19-1) are analyzed. Several promising technologies are selected and then the contribution that each alternative makes toward solving the salinity problem is evaluated. Acceptable alternative plans are evaluated in terms of physical, environmental, and economic factors. Comparing alternative plans A display that summarizes significant physical, economic, environmental, and social effects helps to readily identify the tradeoffs between alternatives. Alternative plans are analyzed to identify the plan that maximizes net on-farm monetary benefits, net off-farm or downstream benefits, total (on-farm and downstream) net monetary benefits, water conservation, salinity reduction, and environmental protection for fish and wildlife. Table 19-2 summarizes the results of four typical alternatives for improving irrigation management and systems. Although the analysis is based on economic conditions in the 1970s, the process of evaluation costs and benefits is valid for any time frame. Salt load and salinity concentration reduction and off-farm monetary benefits are used to illustrate salinity improvement. Summaries such as this can fully reveal the benefits and costs of each and provide a basis by which all interested groups can decide on salinity control. Plan A achieves a nominal amount of salinity reduction at the least cost. Plan B maximizes net benefits to the farmer. Plan C maximizes total net benefits. Plan D maximizes the reduction of salinity downstream. Selecting a recommended plan Local farmers select an acceptable plan that meets state, interstate, national, and/or international water conservation and water quality goals as appropriate. The plan may include canal and lateral improvements, structural on-farm improvement measures, improved irrigation water management, water reuse, selective land retirement, technical assistance,

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

597

TABLE 19-2. Summary of Alternative Plans for Improving Irrigation Systems and Management Alternatives

Item (1)

Annual benefits On-farm Off-farm Total Annual cost Total net benefits Net benefits on-farma Net benefits off-farmb Salt load reduction (metric tons) Sediment load reduction (metric tons) Loss of wetland (ha)

Plan A: Least-cost Salinity Reduction (2)

Plan B: Maximize Farmer’s Net Benefits (3)

Plan C: Maximize Total Net Benefits (4)

Plan D: Minimize Downstream Salinity (5)

$115,000 100,000 100,000

$160,000 150,000 140,000

$175,000 180,000 160,000

$180,000 190,000 180,000

115,000

170,000

195,000

190,000

15,000 0

20,000 10,000

15,000 20,000

0 10,000

3,600

5,400

6,500

6,900

10,900

16,300

11,000

11,600

40

57

65

73

121

170

194

218

Decrease in electrical energy use (MWh)

11

14

16

18

On-farm irrigation efficiency (%)

58

71

78

81

Conversion of wetland habitat to upland habitat (ha)

a

Net benefits accruing on-farm when all costs are borne by on-farm interests Net benefits off-farm when all costs are borne by off-farm interests

b

From Hedlund et al. (1978).

and cost-share support. For example, Plan C in Table 19-2, which maximizes net benefits, would be recommended for federal implementation. It also provides close to the maximum amount of salinity reduction while providing on-farm benefits to farmers (Hedlund et al. 1978).

SALINITY CONTROL OPTIONS A good salinity control program thus involves a strategy for total management of a river system. A broad systems viewpoint and long-term

598

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

perspective are essential to the national planning of solutions to salinity problems. To illustrate project-level actions, three case studies will be examined.

PROJECT-LEVEL PROGRAMS San Joaquin Valley Drainage: An Ongoing Dilemma The total gross cash receipts from California agriculture in 2005 were $31.7 billion, which represented approximately 13.3% of the total U.S. value. Irrigation on 3.2 million ha of irrigated land throughout the state sustains this level of production. The value was greater than the combined value of the No. 2 and 3 states, Texas and Iowa, respectively. In 1984 nearly 40% of the approximately 2.3 million ha of irrigated land in the San Joaquin Valley (SJV) was threatened by salinity and high water table problems (Backlund and Hoppes 1984). The presence of selenium and other toxic trace elements in the subsurface drainage water from irrigated agriculture in the San Joaquin Valley were responsible for the environmental crisis at the Kesterson Reservoir, which led to the closure of the reservoir (SJVDP 1990) and the elimination of any drainage disposal capacity. There are 240,000 ha of irrigated land in the San Joaquin Valley that lack an outlet for the disposal of saline drainage water, which has resulted in the loss of production and the retirement of 40,000 ha of irrigated land as of 2007. A multi-agency study conducted by the U.S. Bureau of Reclamation, California Department of Water Resources, California Fish and Game, the U.S. Fish and Wildlife Service, and the U.S. Geological Survey identified options for managing the subsurface drainage, which were restricted to in-valley solutions (SJVDP 1990). These included source control, reuse for supplemental irrigation, evaporation ponds, and land retirement. Source control is the implementation of improved irrigation practices to reduce deep percolation losses from inefficient irrigation. This has resulted in the gradual shift from surface irrigation to drip irrigation, subsurface drip irrigation, sprinkler irrigation, and center pivot irrigation. Reuse of saline water for supplemental irrigation has increased with time, particularly in water-short years. Evaporation ponds have not been used and existing ponds have been closed because the concentration of selenium in the evaporation ponds creates an environmental hazard. Land retirement is proceeding with the prospect of an additional 40,000 ha of land being considered for retirement. Lacking an out-of-valley drainage solution, there has been research to develop on-farm solutions that has resulted in a system called Integrated On-Farm Drainage Management (IFDM) (Ayars and Basinal 2005). This is

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

599

a process of sequential reuse of drainage water to reduce the volume and concentrate the salt load prior to discharge in an evaporation facility (Blackwell et al. 2005). This system requires improved irrigation management, drainage system management, and reuse of saline drainage water. It is critical to minimize the drainage water reuse component to achieve an economic system (Wichelns and Oster 2006). This type of system will provide an interim solution while long-term solutions can be developed that are environmentally sustainable and politically acceptable. The Tulare Lake Drainage District (TLDD) serves an area of 95,000 ha in the southern San Joaquin Valley that was formerly the lakebed of Tulare Lake. About 18.5 million m3 of drain water is collected each year from the approximate 10,000 ha of subsurface drained land. The district contains more than 1,250 ha of evaporation ponds as disposal sites for drainage water that is not used for irrigation. Evaporation ponds are subject to many regulatory controls and, as a result, are very expensive to operate because of the need for continuous monitoring and reporting on the environmental aspects of the operation. To eliminate the environmental concerns of open evaporation ponds and to maximize the consumptive use of drain water, TLDD has investigated many alternatives for disposal, including the use of constructed wetlands, irrigation of salt-tolerant grasses, and subsurface drip irrigation systems (Zoldoske et al. 2005) (see also Chapter 17 in this manual). Solving the disposal of saline drainage water in the SJV is California’s biggest challenge to sustaining a viable agricultural economy. One proposed solution involves building a pipeline over the mountains of the Coast Range for ocean disposal, but the cost is prohibitive. Another less expensive option is to route return flows to the ocean via the San Francisco Bay. This has met with continued and vehement political opposition on environmental grounds. Colorado River Basin Salinity Control Program With a drainage area of approximately 64 million ha (244,000 mi2), the Colorado River Basin extends across parts of seven states: Arizona, California, Colorado, Nevada, New Mexico, Utah, and Wyoming. The river and its tributaries supply water to nearly 27 million people and to approximately 1.6 million ha (4 million ac) of irrigated land in the United States. In Mexico, it supplies water to 2.3 million people and to about 202,000 ha (500,000 ac) of irrigated land (USDI 2003). It was the recognition of salt damage to irrigated crops in Mexico that led to the signing of Minute No. 242 (1973) of the International Boundary and Water Commission (an amendment to the Mexican Water Treaty of 1944). Minute No. 242 placed a limit on the salinity of Colorado River water flowing into Mexico. In order to address salinity requirements of both the federal Clean Water Act

600

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and Minute No. 242, the U.S. Congress passed the Colorado River Basin Salinity Control Act (CRBSCA) (Public Law 93-320) in 1974. Title I of CRBSCA provides for the construction, operation, and maintenance of a desalting plant near Yuma, Arizona, to reduce the salinity of drainage water from the Wellton-Mohawk Irrigation District. Title II of CRBSCA establishes specific salinity control units within the Colorado River basin. Salinity control planning has been oriented toward reducing the salt load to the river (Hedlund 1984). Salinity reduction through practices, such as lining of irrigation canals and reservoirs to reduce seepage losses, improvements to irrigation systems that reduce deep percolation losses, and management of farm irrigation systems to reduce deep percolation losses, have been successful in salinity reduction to the point that only test operation of the Yuma desalting plant has been necessary. If future demands on Colorado River water lead to increases in salinity concentrations, it may be necessary to bring the desalting plant on-line. Public Law 98-569, an amendment to CRBSCA signed into law on October 30, 1984, authorizes establishment of additional salinity control units not listed in the original Act. The amendment directs the Secretary of Agriculture to establish a major voluntary on-farm cooperative salinity control program (USBR/USDA 1986). The Salinity Control Act requires full coordination and cooperation between the U.S. Department of the Interior, the U.S. Department of Agriculture, and the U.S. Environmental Protection Agency. The salinity control problem has many technically complex and institutionally challenging aspects. The cooperative state and federal program is based on maintenance of lower mainstem salinity concentrations established by the Clean Water Act (1972), while the upper basin develops its compact-apportioned waters. Expanding water use in the upper basin reduces the river’s flow rate and causes salt loading to increase. Both these effects are counterproductive to the maintenance of salinity levels in the lower reaches of the river. Salt reduction at minimal cost is what determines which salinity control units or practices should be initiated (Walker et al. 1977). However, each on-farm alternative undergoes a complete evaluation to ensure that on-farm benefits offset the farmers’ costs. One of nine active salinity control units in the Colorado River basin, the Grand Valley Unit, located along the Colorado River near Grand Junction, Colorado, illustrates some of the salinity management approaches being implemented. In this unit, seepage from the conveyance and application of irrigation water is the driving force for salinity loading to the Colorado River. Before implementation, the contribution of salt from irrigation-related practices within the Grand Valley Unit was estimated at

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

601

52,600 metric tons per year. Seepage water in the Grand Valley dissolves natural salts from the soil and from the underlying Mancos shale formation, and transports those salts to the Colorado River. The Mancos shale is a marine formation in western Colorado and Utah that contains a nearly limitless supply of crystalline salt that easily dissolves in percolating water and contributes substantially to the salt load of the Colorado River. The goal of the Grand Valley Unit is to reduce salt loading by reducing the volume of seepage reaching the Colorado River. The USDA Natural Resources Conservation Service (NRCS) is working with land owners to reduce the on-farm contributions to salt loading by reducing the quantity of water lost to deep percolation (seepage below the rootzone) and by reducing the quantity of water diverted for irrigation. Practices being applied are lining on-farm irrigation ditches, replacing ditches with pipelines, improving irrigation systems, and improving irrigation water management. By the end of 2006, 83% of the targeted land in the Grand Valley Project was under contract for salinity reduction measures. The U.S. Bureau of Reclamation is providing improvements to the irrigation delivery infrastructure to reduce off-farm salt loading. Sources of off-farm salt loading are seepage from unlined (or permeable earth-lined) irrigation delivery canals and laterals, and seepage from ditches that return flows to the Colorado River. On-farm salt loading results when irrigation water drains below the crop rootzone and seepage water comes from irrigation canals and ditches. Reducing seepage from water conveyances has been the primary focus of the Grand Valley Unit (USDA/NRCS 2006). Upper Arkansas River: Irrigation with Saline Water The Arkansas River in southeastern Colorado is one of the most saline rivers in the United States. The total dissolved solids (TDS) concentration of irrigation water used by farmers there is at least 2,000 mg/L and frequently exceeds 4,000 mg/L in the lower reaches of the valley. Water containing more than 2,000 mg/L salt would normally be assumed to be unsuitable for irrigation, but such water is used successfully on more than 80,000 ha of land in the Arkansas Valley, Colorado. The salinity in the Arkansas River basin is more a problem of reuse and full development of a limited water resource than of salt loading. About 85% of the total surface water supply is consumed before the river leaves Colorado. Although controlling salt loading as much as practically possible is desirable, it is expected that salinity levels will remain high as long as water is used. Therefore, the greatest potential for reducing salinity damages is apparently in learning how to use and reuse highly saline water.

602

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

LEGAL AND INSTITUTIONAL FACTORS Legal and institutional constraints often make implementing salinity control projects more difficult. Some progress has been made in modifying these constraints to make implementation of projects more feasible. Water Quality Standards for Salinity In 1965 the U.S. Congress passed the Water Quality Act, which requires each state to establish water quality standards for its portions of any interstate waters or waterways. In 1972, Congress passed the Federal Water Pollution Control Act Amendments, Public Law 92-500. The U.S. Environmental Protection Agency, which was designated as the agency to enforce these amendments, interpreted them as a mandate to establish numeric salinity standards for problem river systems, such as the Colorado River (Colorado River Basin Salinity Control Forum 1984). This created specific standards that facilitate the implementation of salinity control projects. Basin-Wide Jurisdiction For an effective salinity control program, a single entity with legislative authority to plan and implement basin-wide projects is necessary. That same entity also would be responsible for pushing for appropriate legislation and funding. For example, the seven Colorado River basin states introduced legislation in Congress in 1972 and 1973 to authorize a basin-wide salinity control program. To ensure cooperation, these states formed the Colorado River Basin Salinity Control Forum, comprised of up to three governorappointed representatives from each state. With this basin-wide approach to controlling salinity, the basin states have been able to adopt numeric water quality standards for salinity, develop a plan for salinity control, achieve passage of federal legislation, and obtain funding. Effect of Water Rights State water rights constrain, to some degree, virtually all salinity control options. In some states, water rights prevent the retirement of highsalt-yielding farmland, the transfer of “saved” water to nonpolluting use, the collection and reuse of saline water, the collection and disposal of saline flows, or reducing the exportation of good-quality water from a river system. Lack of Urgency Much of the control problem arises from how the salinity issue is perceived locally and nationally. Irrigation return flow is an off-site or down-

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

603

stream problem; the upstream farmer may be unaffected and, therefore, unaware of it. At the national level, the salinity problem is perceived as a regional problem to be solved by the states. The lack of urgency that sometimes prevails may be due to confusion or disbelief about hydrologic predictions for the future. Federal agencies, regional interest, and environmental groups often perceive predictions as self-serving. International Treaties Treaties between the United States and the Republic of Mexico on the problem of salinity have fostered the development of project-level salinity management programs in the Colorado River basin (Hobart 1984). Approximately 90% of the water in the Colorado River is consumed within the United States before the river flows across the border into Mexico. In 1964 the Colorado River was completely dry from the Mexicali Valley, Mexico, to the Gulf of California. The river was, and still is, fully consumed before reaching the ocean. On August 30, 1973, the United States reached an agreement with Mexico in Minute No. 242 of the International Boundary and Water Commission, titled “Permanent and Definitive Solution to the International Problem of the Salinity of the Colorado River.” That agreement states that the average annual TDS content of all waters delivered to Mexico upstream from Morelos Dam may be no more than 115 mg/L above the TDS content of Colorado River water arriving at Imperial Dam. The United States has been in compliance since the agreement was made. Project planning and implementation under Public Law 93-320 has involved cooperative actions by federal, state, and local entities, and individuals. Title I of the project involves several water management measures below the Imperial Dam, which include lining the Coachella Canal, construction of a regulatory pumping station, and construction of a desalting complex on the Wellton-Mohawk drain. The USDA-NRCS’s involvement specifically relates to on-farm treatment and irrigation water management improvement in the WelltonMohawk Irrigation and Drainage District (WMIDD) near Yuma, Arizona (Advisory Committee on Irrigation Efficiency 1974). From 1975 through 1986, the USDA-NRCS (then the Soil Conservation Service) implemented a program to improve irrigation systems and management in the Wellton-Mohawk Valley. As a result, drainage flows decreased from an initial rate of more than 250 million m3/year to less than 125 million m3/yr. A decrease in the diversion of irrigation water to the WMIDD has resulted from the reduction of irrigated area, and improvements in irrigation systems and management have reduced the consumption of irrigation water within the WMIDD (Clemmens and Allen 2005). Since 1990, drainage flow from the WMIDD has varied due to an increase in land devoted to vegetable crops. WMIDD has committed to the pursuit

604

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of a goal to permanently reduce irrigation drainage from the district (USDI 2003).

SALINITY REDUCTION BY IMPROVING WATER SUPPLIES AND IRRIGATION MANAGEMENT Augmentation of Water Supplies The expansion of water supplies to achieve greater dilution may in some cases be an important salinity control option. For example, in southeastern Australia the River Murray Commission has implemented a plan to contain and reduce salinity in the River Murray by augmenting water supplies and increasing water storage capacity. The plan also calls for establishing numeric water quality standards for various points along the river, monitoring and modeling of spatial and temporal water quality trends, and diverting saline drainage waters away from the river. The salinity control options for the Colorado and Rio Grande Rivers in the United States and Mexico are similar to those for the River Murray. As water is used and then discharged back to the river, it becomes more saline to downstream users. Earlier in this century, the United States vigorously augmented its water supplies in the Colorado River basin and throughout the rest of the country by construction of numerous dams, reservoirs, canals, and pipelines. However, the potential for expansion of the nation’s water supply system has become significantly limited. Few rivers remain undammed, and the construction of additional canals, reservoirs, and other water resources system infrastructures typically involve considerable environmental and legal problems and extremely high construction costs. Inter-basin transfers of fresh water In the arid and semiarid portions of the western United States with serious salinity problems, water is scarce, and nearly all existing water supplies are committed to beneficial uses. The opportunities for importation and transfers for augmenting water supplies are minimal. Little public support exists at either the local or national level for water resource development and interregional water transfers. In addition, construction costs have escalated, and a strong, broad-based, largely antidevelopment environmental movement has emerged. Climate change and weather modification Climate change may affect the availability of water supplies due to changes in timing, intensity, amount, and type (rain or snow) of precipita-

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

605

tion. Where water is supplied by snowmelt, weather changes may affect the timing of when that water is available. Current weather modification research may develop a reliable system for increasing precipitation in some local areas. The net benefits, the flexibility of use, and opportunities to obtain even greater water yields with new techniques point to cloud seeding as a desirable tool for increasing water supplies to dilute salinity. The technical, legal, and institutional problems of measuring and allocating increased flows will present major obstacles to a basin-wide weather modification program. Even if additional high-quality water could be produced, no assurances that it would be used for salinity control exist. On the contrary, junior appropriators could claim the water and use it to irrigate more land, thereby worsening the salinity problem. Desalting processes Current technology has made several commercial desalting processes, such as reverse osmosis, electrodialysis, and ion exchange, available for use in the treatment of saline water. Process selection depends on the volume and quality of the saline water and major economic factors, such as energy costs and the costs for disposing concentrated salt brine (Bureau of Reclamation 2003). Saline water may be desalted at the point of diversion or prior to use. For irrigation purposes, this would involve large desalting plants. An alternative involves desalting the irrigation return flows before they are discharged to a stream. In some areas, desalting seawater and groundwater to augment fresh-water supplies may prove economical. Desalting is technically feasible, but the costs are high and the disposal of brines presents a problem. Vegetation and watershed management In most of the arid and semiarid western United States, the opportunity for water salvage by watershed management or phreatophyte control remain limited (see also Chapter 18 regarding phreatophyte use in managing saline seeps). According to the U.S. Forest Service, removal of riparian vegetation on perennial or intermittent streams seldom saves water. Shade removal increases evaporation and transmission losses may be high. Ecologically sound watershed-management techniques could reduce the amount of sediment and associated salt from some areas. Reuse of saline drainage water for irrigation One strategy for controlling the salinity of river systems involves intercepting drainage returns before they are mixed back into the river and using them to irrigate certain crops in the rotation (Rhoades 1984). When

606

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the drainage water quality is such that its potential for reuse is exhausted, then this drainage is discharged into evaporation ponds, injected into deep wells, or desalted. This topic is covered extensively in other chapters of this manual. Improvement of Water Delivery Systems An important part of salinity management is the reduction of canal and lateral seepage into underlying saline aquifers. If a water delivery system is unlined and soils are permeable, considerable water may be lost through seepage. Not only is this seepage unavailable to the farmer, it may pick up other pollutants and salt before returning to a stream or entering a groundwater aquifer. Locating and quantifying canal leakage In the United States about one-quarter of all the water withdrawn from streams and reservoirs for irrigation does not reach the farm. On a delivery system, determining the rate at which seepage occurs and locating sections of high seepage is difficult. Leaky reaches of canals are typically identified through geologic maps, construction records, aerial photographs, soil surveys, ground reconnaissance, hydrologic measurements, groundwater data, and interviews with local farmers as well as canal company and cooperative extension service personnel. A walking survey of the canal is recommended. The presence of wetland vegetation below a canal can help to identify seepage. The rate of seepage from a given canal depends on such factors as the physical characteristics of the material in the channel and subsoil, depth of water in the canal, depth to water table, sediment concentration in the water, temperature of canal water, groundwater, biological growth in the water and along the canal, and the time of year. To obtain reliable estimates of seepage, field measurements are needed. Where canal and lateral lengths are extensive, seepage measurements on all canals and laterals would be time-consuming and might be cost-prohibitive. An alternative to a complete inventory of seepage measurements is to characterize canals by soil type, conveyance properties, and geohydrologic setting, with seepage measurements taken at a limited number of locations that represent these conditions. Seepage estimates can then be estimated for intermediate locations based on soil characteristics. Several methods for estimating seepage exist, but none is entirely satisfactory for all situations. They include ponding tests, inflow-outflow measurements, seepage meters, barrel tests, and mathematical determinations (Dhillon 1967).

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

607

Reducing salt loading Efforts to reduce seepage alone do not necessarily lead to reduced salinity. Unless the seepage water dissolves minerals or displaces saline groundwater on its journey to the underlying groundwater system or receiving stream, reducing seepage may negligibly affect salinity. Thus, a basin-wide or irrigation-project-wide investigation is needed to identify salinity-producing mechanisms and develop a control strategy. Reduction of canal losses Canals and laterals with large seepage losses formerly were lined to conserve water, with no thought given to salinity control. It is now recognized that reducing seepage in areas where soluble salts are present saves water and reduces salt loading to streams. Seepage has been successfully controlled by lining canals with compacted earth, bentonite, concrete, various geomembranes, or combinations of these materials. For smaller canals and laterals, closed conduits of concrete or polyvinyl chloride are practical alternatives. Pipelines reduce evaporation, produce a pressure head, and can be buried below the land surface. They are often useful for linking or otherwise combining canals. Water control structures Water measurement devices and control structures facilitate better management and reduce salt pickup independent of a seepage reduction program. In open channels, various culverts, checks, drops, and dividers are commonly required to regulate water levels, control velocities, prevent erosion, and direct water deliveries. Improved management A modern irrigation delivery system, with appurtenant water control structures and flow measurement devices, must be operated efficiently to achieve crop production and salinity control goals. Lining canals or installing pipelines will control seepage and the associated salinity contributions. It is equally important to manage water deliveries to match crop requirements (Clemmens 1987). Good district records of water deliveries are needed. Accurate measurement and accounting for most of the water is essential for good irrigation-water management. It is vitally important to match physical improvements to the delivery system with on-farm water conservation programs. Improving On-Farm Irrigation Systems and Management Surface and subsurface irrigation return flows carry agricultural pollutants. These include dissolved salts, sediment, nitrogen, phosphorus

608

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

compounds, pesticides, and oxygen-consuming organic compounds. The most persistent pollution problem in the western United States is salt loading of streams and groundwater aquifers by ditch seepage and deep percolation of irrigation water. Improving on-farm irrigation with practices that can reduce ditch seepage, deep percolation, and the associated salt load picked up in irrigation return flow is an important project-level strategy for salinity control (Table 19-3). Soil salinization poses an increasing problem on irrigated land, especially where on-farm drainage systems have not been installed and where no permanent disposal site for highly saline drainage effluent exists. Proper leaching of salts, management of irrigation water, and the disposal of saline drainage flow all must be considered in planning a program to prevent salinization of irrigated soils (Willardson and Hanks 1976). Water and salt budgets are needed to evaluate the project’s effects on soil salinity and downstream water quality.

TABLE 19-3. Sources of On-Farm Salt Loading and Practices for Attaining Salt-Load Reduction Salt Sources (1)

Application Management (2)

Structural Practices (3)

On-farm ditches

Utilize water measuring devices. Practice good maintenance.

Install ditch lining or pipelines and appurtenant water structures.

Tailwater ditches

Follow irrigation water management (IWM) plan to reduce tailwater.

Install lined collection ditches or lined ponds.

Excessive and nonuniform application of irrigation water

Adjust the number and frequency of irrigations, time of set, and flow rate. Follow IWM plan. Change method of irrigation.

Install measuring devices. Land-level to improve length of run. Automate the system.

Inefficient layout of fields and irrigation systems

Combine fields and reorganize or change to more efficient method of irrigation.

Convert existing open drains to buried drains. Relocate ditches and pipelines. Relevel and select proper length of run.

Seepage

Deep percolation

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

609

On-farm irrigation water management One part of a salinity control plan is the improvement of on-farm salinity water management. This may only involve providing a farmer with an irrigation water management plan that includes the use of devices to measure the rate and volume of water delivered to the field, the monitoring of soil moisture, the scheduling of irrigations, training the irrigator on how and when to apply the water and otherwise follow an irrigationwater management plan, maintenance of the irrigation system, and following good agronomic practices to maintain water infiltration and promote full root development. Ditch lining and pipelines Another part of a salinity control plan is to line ditches or install pipelines to reduce seepage. Seepage from ditches can contribute significantly to salt load pickup. On-farm improvements should include concrete ditches with headgates and ports or notches, or pipelines and gated pipes with control valves. Delivery systems can be improved by including water control structures with built-in measuring devices. Gated pipes may be semi-automated by including timing devices on control valves, surge controls, and cablegation. Improving a significant reach of the off-farm water delivery system may be needed to operate the on-farm improvements properly. A pipeline could be installed to develop gravity pressure head to operate a sprinkler system. Installing new or improved systems Yet another part of a salinity-control plan—sometimes inseparable from the other two—is an improvement of the on-farm irrigation system to reduce deep percolation and associated salt loading. Obtaining precise land-leveling and proper length of run on surface-irrigated fields will help to achieve uniform application. Combining fields or changing to a more efficient method of irrigation may also help reduce deep percolation. On-farm drainage, tailwater ditches, collector ponds, reuse systems To combine fields and reorganize irrigation systems, existing open drains must be converted to subsurface drains. Installing on-farm drainage systems to handle the leaching fraction of irrigation water will prevent soil salinization. Lining tailwater collection ditches and collection ponds can help to prevent seepage into underlying saline aquifers and reduce salt pickup. Irrigation drain water can sometimes be reused. Pump-back systems permit such reuse on the same field.

610

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Selective land retirement and irrigation water acquisition For optimal salinity-control strategies in project-level planning, selective land retirement must be considered. In Arizona’s Wellton-Mohawk Irrigation and Drainage District, the salinity control program involved the purchase of 2,520 ha of land with poor irrigation efficiency. This land was then retired from irrigation to reduce the volume and cost of desalting return flows for use in Mexico. Monitoring and evaluation Another part of a salinity control plan is to coordinate with the farmer and adjust his irrigation water management plan to fit the system and changing farm practices. Monitoring and evaluation provides valuable economic, environmental, and irrigation information (USDA/SCS 1982).

Managing Wastewater To alleviate drainage water disposal problems, point sources of drainage water can be intercepted and diverted to other outlets. Several alternatives for reusing or disposing of drain water are described. Export to the ocean or another basin Irrigation drainage water may be disposed of in the ocean or in terminal lakes. The Imperial Valley in California, situated in the once-desolate Colorado Desert, has become one of the world’s most productive farming regions. However, the Salton Sea, a terminal lake in the basin, has been the receptor of saline water collected from 32,000 km of drains in the 186,000-ha irrigation project. The Salton Sea varies from 55 to 65 km long and 25 to 30 km wide. Since 1994, reduced agricultural inflows to the Salton Sea have resulted in a decline in the average water surface elevation (see Fig. 19-3). The reduced inflows result from the adoption of more efficient irrigation water management practices and conversion from surface irrigation methods to trickle irrigation, which contributes very little drainage to the Salton Sea. The concentrations of salts, nutrients, and pesticides have increased as a result of the declining water volume. Periodic algal blooms, promoted by high nitrogen and phosphorus levels, lead to hypoxic conditions that kill fish. However, a resident population of tilapia has managed to survive. Tilapia are able to survive salinity concentrations that approach twice that of seawater. Environmental issues in the Salton Sea are expected to increase if the water surface level continues to decline.

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

611

FIGURE 9-3. Salton Sea average monthly water surface elevations, 1994–2004. From USGS (2011).

Evaporation ponds The situation in California’s San Joaquin Valley illustrates some of the conflicting issues of using evaporation ponds. Although the ponds provide temporary relief, a permanent solution is necessary because evaporation results in the deposition of salts that contain hazardous levels of trace elements (Hall 1986). Desalting The Wellton-Mohawk Salinity Control Project in Arizona was created to help treat irrigation drainage water from the Colorado River basin and improve the quality of water sent downstream to Mexico. The keystone of the Wellton-Mohawk Salinity Control Project is the Yuma desalting plant. The design capacity of this reverse osmosis (RO) desalting plant is 273,000 m3 of drainage water per day (Trompoter 1987). To dispose of RO reject water, an 85-km, concrete-lined drain extends from the desalting plant across the Mexico border, to the upper end of the Santa Clara Slough, which empties into the Gulf of California. The operating plan calls for the 83 million m3 per year of product water, at a salt concentration of 295 mg/L, to be mixed with the flow of the Colorado River and delivered to Mexico during periods when salinity levels in the river exceed limits established by the Colorado River Basin Salinity Control Act. Since its completion in 1992, the United States has been able to meet its obligation without operating the plant. Operation of the plant has been limited to a 6-month test run in 1992–1993 and a 3-month test run in 2007.

612

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Deep well injection Injection of saline water into an isolated deep aquifer may be a feasible disposal method in many areas of the American West. Deep well injection currently is a key part of the Paradox Valley Unit of the Colorado River Basin Salinity Control Program (Chafin 2003). SUMMARY In project-level or regional-level efforts to manage soil salinity and agricultural drainage, planning, implementation, and management are substantially more complicated than at the field or farm level. In formulating a regional program, the physical parameters and effects of each alternative are quantified and compiled, often using computer models, and then translated into economic terms so that each alternative’s benefits can be compared with its costs. In this analysis, benefits and costs may be distributed to different entities in each alternative; the alternatives may thus reflect (1) different sources and allocations of funds, (2) different total levels of benefit, and (3) different allocations of benefit among growers, urban water users, and fish and wildlife. A good salinity control program thus involves a strategy for total management of a river system. A broad systems viewpoint and long-term perspective are essential to the national planning of solutions to salinity problems. This level of planning involves evaluation and eventual integration of on-farm, local, and regional solutions, based on consideration of the problem and the suite of management options that are available and economically and socially acceptable. Project-level solutions may be phased in over decades, as in California’s San Joaquin Valley, where the focus of the last 30 years has been on on-farm and local management, pending future solutions at a larger spatial and temporal scale. In developing these programs, there are legal and institutional constraints, such as constraints related to water rights and the beneficial uses of water. Finally, there are a number of technical solutions for reducing salinity and drainage problems on farms and at a regional level: • • • • • •

Increasing water supplies Water transfers Desalting Vegetation management Reuse of saline water in drainage Improving water supply systems to minimize loss of supply in transit, including repair of canals • Reducing salt loading • Better management of supplies to match demands

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

613

• Use of evaporation ponds to concentrate and sequester salts in drainage • A suite of on-farm irrigation management practices that control mobilization and leaching of salts as a result of irrigation. Project-level planning involves integrating these various technical approaches to solving the physical problems that cause a salinity/drainage problem, in a systematic program that meets economic and environmental objectives. REFERENCES Advisory Committee on Irrigation Efficiency. (1974). Special report on measures for reducing return flows from the Wellton-Mohawk Irrigation and Drainage District, U.S. Department of the Interior, Bureau of Reclamation, Yuma, Ariz. Ayars, J. E., and Basinal, L. (2005). A technical advisor’s manual: Managing agricultural irrigation drainage water: A guide for developing integrated on-farm drainage management systems, Center for Irrigation Technology, California State University, Fresno, Calif. Backlund, V. L., and Hoppes, R. R. (1984). “Status of soil salinity in California.” Calif. Agric., 38(10), 8–9. Blackwell, J., Jayawardane, N., Biswas, T., and Christen, E. W. (2005). “Evaluation of a sequential biological concentration system in natural resource management of a saline irrigated area.” Aust. J. Water Resourc., 9, 169–176. Bureau of Reclamation. (2003). Desalting handbook for planners, 3rd ed., U.S. Department of the Interior, Bureau of Reclamation, Denver, Colo. Bureau of Reclamation and U.S. Department of Agriculture (USBR/USDA). (1986). Joint evaluation of salinity control programs in the Colorado River basin, U.S. Department of the Interior, Bureau of Reclamation, Denver, Colo. Chafin, D. T. (2003). Effect of the paradox valley unit on the dissolved-solids load of the Dolores River near Bedrock, Colorado, 1988–2001, U.S. Geological Survey, Washington, D.C. Clemmens, A. J. (1987). “Delivery system schedules and required capacities, in planning operation, rehabilitation and automation of irrigation water delivery systems,” in Proc. of ASCE Conference, Irrigation Systems for the 21st Century, Portland, Oregon, July 1987, ASCE, Reston, Va. Clemmens, A. J., and Allen, R .G. (2005). “Impact of agricultural water conservation on water availability,” in Proc. EWRI 2005: Impacts of Global Climate Change 2005, ASCE, Reston, Va. Colorado River Basin Salinity Control Forum. (1984). Water quality standards for salinity, Colorado River system, U.S. EPA, Denver, Colo. Dhillon, G. S. (1967). “Measurement of seepage losses from irrigation canals.” Ind. J. Power and River Valley Dev., 17(1). Evans, R. G., Walker, W. R., and Skogerboe, G. V. (1981). Optimizing salinity control strategies for the upper Colorado River basin, AER80-81RGE-WRW-GVSI, Colorado State University, Ft. Collins, Colo.

614

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Hall, S. K. (1986). “Evaporation basins: Environmental disaster, or economic necessity?” in J. B. Summers and S. S. Anderson, eds., Toxic substances in agricultural water supply and drainage—Proc. Regional meeting of U.S. Committee on Irrigation and Drainage, Phoenix, Ariz., October 22–24, 1986, USCID, Denver, Colo., 27–35. Hedlund, J. D. (1984). “USDA planning process for Colorado River basin salinity control,” in Proc. 1983 International Symposium on State-of-the-art Control of Salinity, ASCE, Reston, Va., 63–77. Hedlund, J. D., Russell, C., and Collins, H. (1978). “Techniques for planning and evaluating water conservation and water quality measures in irrigated agriculture,” in Proc. 33rd Annual SCSA Meeting, Soil Conservation Society of America, Ankeny, Iowa. Hobart, M. S. (1984). “The lower Colorado, a salty river.” Calif. Agric., 38(10), 6–8. Rhoades, J. D. (1984). “Reusing saline drainage waters for irrigation: A strategy to reduce salt-loading of rivers,” in R. H. French, ed., Salinity in water courses and canals, Chapter 43, Ann Arbor/Butterworth Publishing Co., Boston, 455–464. San Joaquin Valley Drainage Program (SJVDP). (1990). Management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley, San Joaquin Valley Drainage Program, Sacramento, Calif. Skogerboe, G. V., Walker, W. R., and Evans, R. G. (1979) Environmental planning manual for salinity management in irrigated agriculture, Report EPA-600/2-79-062, U.S. EPA, Ada, Okla. Trompoter, K. M. (1987). The Yuma desalting plant: A water quality solution, draft paper, U.S. Dept. of the Interior, Bureau of Reclamation, Yuma, Ariz. U.S. Department of Agriculture, Natural Resources Conservation Service (USDA/NRCS) (2006). Monitoring and evaluation report, Grand Valley Unit, Colorado River Salinity Control Project, USDA Natural Resources Conservation Service, Grand Junction, Colo. U.S. Department of Agriculture, Soil Conservation Service (USDA/SCS). (1982). Monitoring and evaluation plan for Grand Valley Unit, Colorado and Uintah Basin Unit, Utah, USDA Soil Conservation Service, Denver, Colo. U.S. Department of the Interior (USDI). (2003). Quality of Colorado River water, progress report 21, U.S. Department of the Interior, Bureau of Reclamation, Upper Colorado Region, Craig, Colo. U.S. Geological Survey (USGS). (2011). “Water resources data – California, water years 1994–2004,” http://pubs.usgs.gov/wdr/2004/wdr-ca-04-2/WDR.CA. 04.vol2.pdf, accessed February 5, 2011. Walker, W. R., Skogerboe, G. V., and Evans, R. G. (1977). “Development of best management practices for salinity control in Grand Valley,” in Proc. National Conference on Irrigation Return Flow Quality Management, Colorado State University, Fort Collins, Colo., 385–395. Water Resources Council. (1983). Economic and environmental principles and guidelines for water and related land resource implementation studies, U.S. Government Printing Office, Washington, D.C. Wichelns, D., and Oster, J. D. (2006). “Sustainable irrigation is necessary and achievable, but direct costs and environmental impacts can be substantial.” Agric. Water Mgmt., 86, 114–127.

PROJECT-LEVEL SALINITY MANAGEMENT OPTIONS

615

Willardson, L. S., and Hanks, R. J. (1976). Irrigation management affecting quality and quantity return flow, Report EPA-600/2-76-226, U.S. EPA, Ada, Okla. Zoldoske, D., Adhikari, D. D., Goorahoo, D., Kimmell, T., Anderson, S., and Hitt, D. (2005). Evaluation of alternative saline drainage water disposal methods, Project Report, California State University, Agricultural Research Initiative, Fresno, Calif.

This page intentionally left blank

CHAPTER 20 SAN JOAQUIN VALLEY, CALIFORNIA, DRAINAGE MANAGEMENT OPTIONS Michael Delamore, Manucher Alemi, John Letey, and Jose I. Faria

INTRODUCTION This chapter is a case study in salinity and drainage management at a regional scale, where management solutions were constrained by (1) the need to address a trace element ecotoxicity issue, and (2) at least temporary restrictions on drainage of agricultural return water to the ocean. In many ways, this case study represents a worst-case scenario and illustrates the principles outlined in Chapter 19 for the development of project-level/regional-scale management programs. In 1984, in response to findings of impacts to waterfowl from selenium in agricultural drainage water at Kesterson Reservoir, the San Joaquin Valley Drainage Program (SJVDP) was established to investigate drainage and drainage-related problems and to develop possible solutions. The SJVDP was a joint federal/state effort involving the U.S. Fish and Wildlife Service (USFWS), the U.S. Geological Survey (USGS), the U.S. Department of the Interior, Bureau of Reclamation (USBR), the California Department of Fish and Game (DFG), the [California] State Water Resources Control Board (SWRCB), and the California Department of Water Resources (DWR). Figure 20-1 shows the five SJVDP study subareas. SJVDP initially conducted a preliminary investigation of all drainage management options, including out-of-valley drainage water disposal. In 1987, an SJVDP-commissioned report (Brown and Caldwell 1987) presented possible areas for drainage water disposal, such as the Pacific Ocean. In reaction to that report, the Citizen’s Advisory Committee recommended to the SJVDP Policy Group that program efforts be focused on in-valley solutions. SJVDP thereafter adopted the approach that agriculture should strive to correct, to the extent feasible, the drainage problem in-valley before resorting to out-of-valley disposal options. 617

618

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 20-1. San Joaquin Valley Drainage Program study subareas.

The SJVDP developed A Management Plan for Agricultural Subsurface Drainage and Related Problems on the Westside San Joaquin Valley, also known as the Rainbow Report, to manage drainage problems from 1990 to 2040 (SJVDP 1990), hereafter called the 1990 Plan. Although the 1990 Plan was based on managing the problems in-valley for several decades without exporting drainage water and salts to the ocean, it also stated that, “ultimately, it may become necessary to remove salt from the Valley” (p. 1 of the 1990 Plan).

DRAINAGE MANAGEMENT OPTIONS

619

The SJVDP investigated treatment technology but did not recommend it as a drainage management component because of technology and cost uncertainties. The SJVDP did recommend, however, continuing treatment technology research, demonstration projects, and continued monitoring to assess the progress and efficacy of various management measures. Key components of the SJVDP recommended plan were source reduction, drainage reuse, land retirement, evaporation ponds, groundwater management, discharge to the San Joaquin River, water for protection of fish and wildlife, and institutional changes. Federal and state agencies initiated the San Joaquin Valley Drainage Implementation Program (SJVDIP) in 1991 to pick up where SJVDP left off, following through on program recommendations (SJVDIP 1991). Four federal agencies (USBR, USFWS, USGS, and the Natural Resources Conservation Service) and four state agencies (DWR, DFG, SWRCB, and the Department of Food and Agriculture) signed a Memorandum of Understanding (MOU) in December 1991. The agencies agreed to use SJVDP’s 1990 Plan as the guide to correct the SJV’s subsurface drainage problems. They agreed to work together to identify specific tasks and associated responsible parties, seek needed funding and authority, and set schedules to implement all components of the SJVDP 1990 Plan. Those signing the MOU recognized that the success of the program depended on local districts and irrigators carrying out effective drainage management measures. Because drainage is a regional problem, however, federal and state agencies needed to remain involved to coordinate efforts. Beginning in 1997, in cooperation with the University of California and representatives of the drainage-impacted subareas, the SJVDIP undertook an updated evaluation of the SJVDP drainage management options (SJVDIP 2000, 1999a–k). This chapter presents a summary and update of that work. Problem Statement History is replete with examples of irrigation projects that ultimately failed because of salt accumulation and the inability to remove salt from soils and shallow groundwater. The classic historic example is Mesopotamia. Beginning about five millennia ago, irrigation projects in the area that is presently southern Iraq created a highly productive agricultural system. Copious quantities of water were available, contributing to seepage and a consequent rise in the water table. Initially, only a few fields were affected, but increases in salt-affected fields were recorded between 1200 and 1800 BC. Declining yields and a shift to cultivating more salttolerant crops paralleled increasing salinity. The southern part of the alluvial plane never recovered from the decline resulting from salinization. The story of Mesopotamia is ancient, yet the story of Mesopotamia (with

620

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

minor variations) has been repeated for millennia in other lands and could be repeated in the western SJV. Climatic conditions in the western SJV require importation of irrigation water for economic agricultural production. Dissolved salts in the irrigation water are simultaneously imported. Irrigation to achieve high yields is virtually impossible without some water percolating below the crop rootzone. Indeed, some deep percolation of irrigation water is required to leach salt that accumulates in the rootzone soil. Water percolating below the rootzone moves vertically downward through the unsaturated zone until it reaches the water table. After reaching the water table, water flows in the direction of hydraulic gradients, which might eventually carry it a considerable horizontal distance to a point where the hydraulic head is lower, and thereby potentially contribute to shallow groundwater conditions on other lands. Deep percolation may also cause the water table to rise directly underneath the field where there is an underlying confining layer and hydraulic conductivities are low. The fine-textured alluvium in the western SJV is derived from sedimentary coastal range deposits containing significant quantities of soluble mineral salts and trace elements such as selenium, chromium, arsenic, boron, lead, mercury, cadmium, copper, and zinc. Most of the undesirable characteristics associated with western SJV soils are directly due to their origin from marine sedimentary parent materials of the coastal ranges. Water that percolates below the rootzone moves through these sedimentary materials and dissolves salts and other chemicals. As the water table approaches the land surface, drainage systems may be installed to keep the water table from encroaching into the crop rootzone. Drainage waters typically have high concentrations of salts and various trace elements, with concentrations varying with location. The challenge is to properly reduce, reuse, or dispose of these drainage waters. The salt in the drainage water affects its reuse for irrigating agricultural crops, and the selenium in the water can negatively affect wildlife if discharged into wetlands or other water bodies, greatly decreasing its utility for creating wetland habitat. Agricultural lands with a shallow water table ultimately must have a drainage system to lower the water table, remove salt, and maintain productivity. If it were not for the presence of certain constituents, such as selenium in the drainage water and their associated wildlife impacts, an out-of-valley drain such as the partially constructed San Luis Drain likely would have been completed. Such a drain would have enabled drainage waters to be conveyed from the SJV and discharged directly into the Sacramento-San Joaquin delta or Suisun Bay and flow to the Pacific Ocean. The presence of selenium in drainage water not only curtailed completion of the San Luis Drain but also is the major obstacle toward finding alternative solutions.

DRAINAGE MANAGEMENT OPTIONS

621

Solution Approach Eight technical committee reports were prepared by the SJVDIP (SJVDIP 1999a–h) summarizing the scientific information on individual management options that may address the problems associated with invalley management of drainage, salinity, and toxic trace elements in the western SJV. No single management option is adequate to solve all drainage-related problems, and many options require interaction with other options for maximum benefits. As described in Chapter 19, the challenge was to identify the optimal mix of options. Because of variable conditions throughout the SJV, the optimal mix of management options differs based on location. Construction of an out-of-valley discharge facility was not part of the SJVDP solution mix. A schematic presentation of the in-valley options is presented in Fig. 20-2. The figure does not show the surface and subsurface hydrologic system that provides the spatial connection between management options. Surface hydrology is visible and can be quantified. For example, the endpoint of drainage water pumped to the surface can be traced by following its flow to an evaporation pond, a field for irrigation reuse, a ditch where it is diluted, and so on. Subsurface hydrology is much more complex and has not been completely characterized at most locations. Nevertheless, subsurface hydrology provides the spatial connection between management options that affect subsurface flows. For example, deep percolation on one farm may travel some distance and result in drainage water on another farm. Therefore, the consequent effects of management options can be completely evaluated only from a spatial

FIGURE 20-2. Schematic presentation of in-valley drainage management options.

622

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

viewpoint encompassing the interconnection of surface and subsurface hydrology. Because groundwater flow is relatively slow, the temporal effects of management must also be considered. For example, a change in management affecting the amount of deep percolation from an upslope grower may not be reflected in the amount of drainage for a hydrologically connected downslope grower for several years.

Reducing Drainage Water Volume Irrigation and drainage management, groundwater management, and land retirement are three options to reduce the volume of drainage water requiring surface disposal. Each was the topic of a technical committee report. Reduction in volume of applied irrigation water and discharged drainage water will be referred to as source reduction. Source reduction and groundwater management permit continued agricultural production, whereas land retirement converts land from agricultural production to other uses. Irrigation and drainage management Reduction of drainage volume by irrigation and drainage management is largely driven by three factors: (1) the ability to precisely control the rate of water infiltration into the soil, (2) the ability to apply water uniformly across a field, and (3) the degree of drainage water reuse. The latter is discussed in “Drainage Reuse,” following. The ability to control the rate of infiltration is necessary to achieve the goal of replenishing the soil water supply to maintain crop yield without applying excess water and increasing drainage volumes. Source reduction can be achieved by improving irrigation scheduling. One method involves using reference evapotranspiration (ET) data, provided by the California Irrigation Management Information System (CIMIS) weather stations, combined with appropriate crop coefficients (Kc). A plant’s water usage varies during the growing season; therefore, the most accurate estimates of crop ET will be obtained by employing improved seasonal crop coefficients. Irrigation systems can be broadly classified into pressurized irrigation systems and surface-applied gravity-flow systems. Pressurized systems are those in which water is delivered through a pipe and then discharged through various orifices such as a sprinkler head or drip emitter. In surface-applied systems, irrigation water is applied at one end of a field and flows across to the other side of the field, such as in a furrow irrigation system. Pressurized systems, such as sprinkler and drip, allow precise valve control on the rate and volume of water applied during irrigation.

DRAINAGE MANAGEMENT OPTIONS

623

Nonuniform distribution of irrigation means that some parts of a field receive more water than other parts of a field. If irrigation is applied to achieve high yields on those portions of the field receiving the least water, the result would be an excess of irrigation water applied to other parts of the field with resultant large drainage volumes. Conversely, if irrigation is designed to reduce drainage from the sections of the field receiving the most water, yields would be severely reduced on the sections of the field receiving the least water. Nonuniform irrigation requires a significant trade-off between achieving high yields and minimizing drainage volumes. Uniform distribution of irrigation water achieves both high yields and low drainage volumes. The uniformity of water distribution is dependent on the design and maintenance of the irrigation system. Surface-applied systems provide the least control over both irrigation uniformity and the amount of irrigation. Two factors contribute to nonuniformity of surface irrigation systems. First, water must flow across the field and, therefore, is in contact with the soil longer at the head of a field than at the tail of a field. This is referred to as opportunity time nonuniformity. Second, the infiltration of the water depends on soil properties. Fields may have high soil variability with nonuniform infiltration rates. The amount of water that infiltrates the soil is dependent not only on how long the water is applied to the field but also the soil infiltration rate. Further, the uniformity of sprinkler systems is decreased by factors including wind. Thus, the irrigator has limited control on the amount of water that infiltrates a given field. Deep percolation resulting from furrow irrigation can be reduced by properly designing furrows with shorter lengths. Deep percolation can also be reduced by improving water delivery management, such as switching to surge-flow. Both of these overall techniques—better furrow design and better water delivery management—help to make the distribution of applied irrigation water more uniform. Deep percolation can be excessive during germination and growth of seedlings in furrow-irrigated fields. Water must be applied in sufficient quantity to wet the full length of the furrows, yet the young plants are not large enough to take up much water. Switching to sprinkler irrigation can be helpful in reducing deep percolation in the first stage of crop growth. By using sprinklers, the irrigator can wet the soil sufficiently and uniformly, with lower application rates than with furrow irrigation. Surface and subsurface drip irrigation, when well-designed and managed, will substantially reduce deep percolation losses. However, drip systems typically are economical only when the crop is of relatively high value. With drip irrigation, a wetted zone forms around each emitter and salt tends to accumulate at the perimeter of this wetted area. Eventually, this salt must be leached from the soil.

624

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Potentially, the relative opportunity for increased uniformity and control on the amount of applied irrigation water is in the order of furrow sprinkler microirrigation systems. The costs associated with each irrigation system are also in the same order. Constraints in converting to potentially higher-performance irrigation systems are economics and crop suitability for pressurized irrigation systems. The major economic question is whether the increased costs involved in upgrading irrigation systems are offset by increased benefits. Many cost/benefit analyses only consider the effects of the irrigation system conversion on crop yield without consideration of the reduced costs associated with managing drainage water. If the benefits associated with reduced drainage water are not considered in the cost/benefit analysis, the results are biased toward the cheaper and less efficient irrigation systems. Crop use of shallow groundwater The management option known as rootzone water extraction involves allowing deeper-rooted plants to satisfy part of their ET requirement by extracting water from a shallow water table (Ayers et al. 1996; URS 2002; see also Chapter 18). Promoting root extraction of groundwater may result in reduction of applied irrigation water and the volume of drainage water that must be collected and managed. A limitation is there are only a few crops (such as cotton) with roots deep enough to reach the groundwater table and are tolerant of salinity present in the groundwater. Trees, such as eucalyptus, have been used to lower shallow water tables or intercept flow of shallow groundwater. Trees have been used for this purpose in a demonstration project at Red Rock Ranch in Fresno County, California (WRCD 1999). Most drainage systems in the SJV are designed with lateral drain lines discharging into a main line leading to a sump that is pumped, or into an open drain. When the water table is higher than the drain line, water flows into the drain line until the water table has been reduced to the drain line depth. Any irrigation that exceeds the water-holding capacity of the rootzone soil increases deep percolation and groundwater recharge, thereby causing the water table to rise and initiate drain flows. The rate of water table drop is related to the spacing of the drainage lines and the hydraulic conductivity of the soil. A control valve can be placed on the drain outlet, regulating drainage flows and retaining more water in the soil profile for later use by the crop. Drainage outlet control has the dual advantage of reducing drainage outflow and also reducing the need for irrigation water. Water normally removed through the drainage system can be retained for crop use when of suitable water quality. The control valve could be temporarily opened

DRAINAGE MANAGEMENT OPTIONS

625

to release drainage if the water table rises too high in the crop rootzone and/or to discharge water for salt reduction within the rootzone. A major impediment to implementing the drainage control option is the present design of drainage systems. Placing the control on the main line discharging into the sump, or controlling the pumping of water from the sump, is not adequate. Drainage laterals are progressively at higher elevations moving up the field. Controlling the water table elevation at the sump would have minimal effect on water table control on laterals at higher elevations. Therefore, a system must be designed that can control each individual lateral. Comparatively little research has been devoted to drainage outlet control management. Both the engineering and management aspects require additional research before guidelines could be established for this practice. Drainage outlet control is important to increasing the utility of solar evaporators and discharge to the San Joaquin River, as will be discussed later. Groundwater management High water tables and substantial drainage volumes are the direct result of an imbalance in regional groundwater budgets. During surface irrigation, water is typically applied to the soil surface at a volume that exceeds the soil water-holding capacity and the carrying capacity of the groundwater system. When recharge rates exceed the groundwater system capacity to discharge via subsurface lateral flow and flow to wells, water tables rise until they intersect with drains or the topographically lower portions of the basin. Regional groundwater budgets must be altered to reduce drainage volumes. Modeling studies show that this can be accomplished through a combination of reductions in groundwater recharge and increases in groundwater pumpage. Reductions in recharge can be accomplished by reducing deep percolation through source reduction, crop use of shallow groundwater, and land retirement. The notion in recent years that the increases in pumpage would have to come from the semiconfined aquifer or from new wells drilled specifically for water table management is incorrect. Regional groundwater models and basic hydrogeologic principles demonstrate that increased pumpage can occur in existing wells. Further, allocating a significant portion of that pumpage to wells tapping the sub-Corcoran confined aquifer can be quite effective for lowering the water table regionally. This occurs by inducing increased rates of downward leakage regionally across the Corcoran clay. Increased pumpage would have the benefit of providing increased water for irrigation and decreased demand for existing surface water supplies.

626

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

In view of regional modeling studies that elucidate groundwater system processes, the notion in recent years that the groundwater management option should be implemented locally or only as a short-term solution is no longer appropriate. It is now clear that if local or regional drainage volumes are to be significantly reduced, long-term regional groundwater management is necessary. This strategy alone would significantly alter the regional groundwater budget that ultimately controls water table elevations. Some local implementation of groundwater management can perhaps affect water table elevations locally, but the net impact of such a strategy would be negligible regionally. In general, concentrations of both dissolved solids and trace elements decrease with depth in the semiconfined aquifer overlying the Corcoran clay layer. Better water quality is found in the confined aquifer system under the Corcoran clay layer. Pumping the better-quality water from deeper wells causes a downward movement of the poorer-quality water found at shallower depths. Plants extract the water, resulting in a high salt concentration in the water leaving the rootzone. In practice, good-quality water is extracted by pumping and replaced by poor-quality water percolating below the rootzone, causing a gradual depletion of the good-quality groundwater supply. Presently, groundwater pumping is increased during drought years when surface water supplies are limited. Exploiting the supply of good-quality groundwater decreases the opportunity to reduce the impact of drought by increased pumping in future years. Increasing the groundwater pumping rates would accelerate the ongoing, downward movement of poor-quality groundwater. Because this process will occur even without increases in pumpage, it is not clear whether the relative water quality impacts would be significant, given that the groundwater quality is already being degraded at an unknown rate. Groundwater quality at some local supply wells would probably experience much more rapid quality degradation than predicted on the regional scale (SJVDIP 1999f). Several state laws prohibit degradation of groundwater, with exceptions being made in rare instances where the degradation is deemed beneficial to the people of California. Proactively managing groundwater in a manner that results in accelerated groundwater quality degradation would require such an exception but would be consistent with the fact that groundwater-quality degradation is already occurring under present pumping practices in the SJV. Regional groundwater analyses indicate that increases in pumpage can significantly lower the water table without creating excessive risk of inducing land subsidence (i.e., without dropping confined aquifer water levels below historical lows) (Belitz and Phillips 1995). Significant improvements in monitoring of groundwater quality, groundwater levels, pumpage, and subsidence are needed to support

DRAINAGE MANAGEMENT OPTIONS

627

implementation of the groundwater management option in an adaptive framework. Even if a groundwater management option is not adopted, such information is necessary for basic stewardship of water resources in the basin. In summary, opportunities in groundwater management include (1) maintenance of agricultural productivity, (2) decreased demand for surface water supplies resulting from increased pumpage, and (3) decreased amount of drainage water requiring disposal. Constraints in groundwater management include (1) reduction in the supply of good-quality groundwater limiting future opportunities for conjunctive use, (2) imposition of the requirement of universal participation and regulatory compliance in regional management instead of voluntary action, (3) increased potential for accelerated degradation of groundwater, and (4) unsuitability of groundwater quality pumped from above the Corcoran clay layer for some uses. Land retirement Land retirement eliminates most irrigation and therefore implicitly ends the need for drainage on retired lands. The original purpose in including land retirement in the 1990 Plan was a means to isolate lands with relatively high concentrations of selenium in the soil and shallow groundwater. Other benefits of land retirement have since gained in importance. Water that would have been applied for irrigation becomes available for irrigation use elsewhere or potentially for other uses. Retired land could become suitable as wildlife habitat for upland endangered species. The nature of the restored habitat is partially dependent on land management. A whole range of scenarios could be considered based on the type and level of adaptive land management and management costs. As a voluntary program, lands most likely to be retired have very low agricultural economic return because of existing high water tables and salinized soil and water resources. The lands are typically located at the lower elevations near the trough of the SJV. Water tables would be expected to drop under lands not irrigated. However, depending on precipitation and lateral flows into the area, water tables could be maintained at a depth close enough to the surface that water would move by capillary action to the surface and evaporate. Upward water flow would carry salts and toxic elements, such as selenium, to the surface and deposit them through evaporation. This would lead to land with sparse vegetation, wind erosion, and poor-quality and possibly toxic habitat. Therefore, one of the major questions related to the land retirement option has been whether the water table would be deep enough to prevent salinization and selenification of the soil surface. Some retired lands could require ongoing management, such as pumping of groundwater, to prevent soil

628

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salinization caused by saline groundwater entering the site from below, adjacent, or upslope areas. Otherwise, lands taken out of agricultural production could lose environmental quality and future value, including for wildlife habitat. The socioeconomic impact on local communities of the value of crops not grown must be counted as part of the cost of retiring a parcel of agricultural land. Substantial direct costs may be involved in the purchase of the land for retirement, and monitoring and management of the land after retirement. Restoration for wildlife habitat will incur additional costs. A number of land management measures and alternative strategies to permanent land retirement and complete cessation of irrigation could achieve the same objectives of source reduction and reduced drainage volume, while minimizing or avoiding soil salinization and reduced plant growth. Alternative measures and strategies include: • Systematic implementation of rotational-, distributed-, or periodicfallowing programs • Pumping of groundwater for reuse as limited irrigation of winter crops to counter the upward transport of salt from shallow groundwater to the soil surface, while providing plant growth opportunities for both agricultural and upland wildlife habitat uses. Summary Each option to reduce the volume of drainage water has advantages and disadvantages. Groundwater management allows continued agricultural operation but requires regional management and compliance in order to maintain a lowered water table. Groundwater resources would be used in lieu of surface water supplies, reducing the option for future conjunctive use. Reducing drainage volume by irrigation and/or drainage outlet control has the benefit of maintaining crop production while being technically feasible but has economic considerations. Control of drainage outlets would require additional research to demonstrate utility, but the control would complement discharge to solar evaporators or the San Joaquin River. Overall, irrigation systems that allow the greatest control are also the most expensive but may be cost-effective when considering the overall costs of drainage water management and disposal. Land retirement does not allow continued agricultural productivity on the lands retired, but it does free surface water supplies for use elsewhere and eliminates the need to dispose of drainage water from the retired land. The long-term consequences of land retirement depend on what type of adaptive land management is adopted. One of the most critical factors affecting land retirement is whether the water table will be sufficiently deep to prevent the transport and accumulation of salts and trace

DRAINAGE MANAGEMENT OPTIONS

629

elements to the surface, affecting soil quality and creating an environmental hazard. Any lateral water flow from upslope growers to downslope growers affects the implementation of irrigation water source reduction or drainage outlet control. Upslope growers who do not have a high water table have no financial incentive to reduce deep percolation by upgrading to more expensive irrigation systems. Drainage outlet control is not an option if they do not have a drainage system. Conversely, downslope growers incur costs associated with the additional drainage contributed by upslope farmers. Drainage Water Treatment Although source reduction, groundwater management, and land retirement provide opportunities to reduce the amount of drainage water, some drainage water will still require reuse or disposal. One option is water treatment to improve the quality and thus the utility of drainage water. Reverse osmosis (RO) is a promising technology for complete treatment of drainage water, that is, removal of dissolved salts including selenium. Continuing advances in membrane technology since the late 1990s have increased the efficacy of RO treatment. The technology is available; implementation of RO treatment is driven by economic considerations. The total amortized production cost of RO is $250 to $1,200/acre-ft. An increase in energy cost of $0.01/kWh would increase the total cost by $50/acre-ft (California Department of Water Resources 2003). The higher investment would be required if extensive pretreatment of the water prior to RO were necessary. RO is an energy-intensive operation and the costs are greatly affected by energy costs. The stated capital and operating costs do not include the costs of collecting the drainage water, delivering the treated water, or disposing of the waste brine. A number of significant benefits could be associated with implementation of membrane treatment technologies, such as RO treatment systems alone or in combination with other drainage management options. In the case of the imposition of more stringent regulations and prohibition of drainage discharge from the Grasslands Subarea to Mud Slough and the San Joaquin River, RO treatment offers an alternative to the discharge of drainage and could allow for the continuance of the present level of agricultural production. RO results in one useful product now in short supply in the SJV—pure water. The purified water could be sold to a municipality, possibly at a profit to the RO operator. The resultant brine, the salinity of which is dependent both on the quality of the original feed water and the recovery (i.e., percent of feed water “recovered” as clean product water), could be used on halophytic crops appropriate for the salt concentration. The concentrated drainage collected from the halophytic crops could then be discharged into a solar evaporator, resulting in salt desiccation and

630

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

recovery. A commercial market for the salt is not available at the present time; however, if RO coupled with salt separation and disposal or utilization could be accomplished economically, the cycle would be closed and drainage would have a beneficial use. In the absence of a market for salt products, the brine or salts can be discharged into lined disposal facilities. However, potential environmental impacts of selenium and possibly other constituents in the brine is a significant consideration in brine management options. If the RO process could be made sufficiently feasible, it could be used to treat shallow groundwater that is currently too saline for general agricultural use. The pumping and treatment of shallow groundwater could help to reduce existing shallow groundwater levels, as well as create a new freshwater supply. Specifically, the RO process could be used to treat shallow groundwater under retired lands in the eventuality that rising shallow groundwater could be affecting soil quality and therefore wildlife habitat quality. The two major obstacles to extensive RO technology implementation are the costs of operation and the current limitations on brine disposal. Purified water would have to be sold at a price greater than most agricultural operations could afford to offset the operational costs. Therefore, treatment of drainage water through RO becomes more feasible if water transfer through an open market is developed between the agricultural and the urban communities. Growers could use treated drainage water in lieu of surface water supplies, which could then be transferred to the urban sector. Disposal of the brine could pose an environmental impact. The brine could be extremely concentrated in salts as well as selenium, depending on the initial concentration of selenium in the drainage water and the degree of concentration achieved in the RO process. Solar ponds, solar evaporators, or lined disposal facilities are potential brine disposal options. Salinity-gradient solar ponds A salinity-gradient solar pond is a body of saline water that combines solar energy collection with long-term thermal energy storage. It is a constructed basin where highly concentrated saline water is placed on the bottom of the basin, with less-saline water at the surface. A density gradient is created with the densest water at the bottom and the least dense water at the top of the water column. This arrangement provides an opportunity to capture solar energy and convert it into electricity. Solar rays pass through the stratified, ponded water, heating and raising the temperature of the lower saline water. In ordinary ponds, warmer and lighter bottomwater rises to the surface, displacing the heavier, colder water above and causing convection currents. These currents rapidly disperse the heat

DRAINAGE MANAGEMENT OPTIONS

631

throughout the pond, preventing any portion of it from reaching a high temperature. The dense saline water at the bottom of a solar pond can stabilize under solar heating, with cessation of convection currents and pond circulation. The bottom layer of hot brine, called the storage zone, is the system’s energy-accumulating component. The stored heat must then be extracted from the lower layer of the pond for utilization. The water from the bottom layer can be removed from one side of the pond, passed through a heat exchanger, and the cooled water returned to the pond. With special care, the water from the lower level can be cycled without disturbing the established density gradient. Potentially, solar ponds allow the opportunity to produce energy as well as dispose of brine. Treatment of drainage water would provide a continual stream of brine. That would require continual expansion of solar ponds to accommodate the continual brine stream. The use of solar ponds for electric power generation and other uses has been extensively researched by the University of Texas–El Paso. A demonstration unit has been successfully operated at El Paso for several years. However, the climatic conditions at El Paso are slightly different from those in the SJV. The economic benefit of solar ponds needs to be determined. Since selenium may be present in the pond brine, there is also a question about hazards to birds that might inhabit the pond. Solar evaporators A solar evaporator is defined as an evaporation system where drainage water is not allowed to pond within the system. The flow of drainage water into the solar evaporator is regulated to equal or be less than the rate of evaporation. Two benefits associated with using solar evaporators are reduced wildlife impacts and facilitated salt harvesting. The major disadvantage is that the rate of discharge must match the rate of evaporation. Since both drainage flows and rates of evaporation vary with time, matching the two is difficult. However, with an efficient RO unit, the salt brine could be made highly concentrated and therefore the volume of drainage water to be evaporated would be minimized. A control system on the drainage outlet, which was discussed earlier in “Crop Use of Shallow Groundwater,” would be complementary to the use of solar evaporators. Drainage discharge could be timed to more closely accommodate the evaporation potential. Treatment for selenium removal Treatment of drainage water to remove only selenium would still leave highly saline water requiring reuse or disposal. Nevertheless, the removal of selenium would increase the options for reusing or disposing of the drainage water.

632

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 20-3. Salt-gradient solar pond near El Paso, Texas.

Chemical reduction to treat drainage water for selenium removal has been investigated. Zero-valent iron filings can be used to reduce selenium. However, tests have indicated that the beds tend to become cemented with precipitate (SJVDIP 1999b). Ferrous hydroxides are also possible but generally have a slow rate of reaction. In both cases, nitrate concentration in the water interferes with the removal process. Also, each system would require a reactor and, while the economics have not been evaluated, it is probable that the systems would be overly expensive. Several laboratory investigations have demonstrated that bacteria can effectively reduce selenium. The selenium concentration in water can be reduced in open systems. For example, an algal-bacterial selenium removal system consisting of a series of specially designed ponds has been tested (Lundquist et al. 1994). The concept of this process is to grow micro-algae to use nitrate and then utilize the naturally settled algal biomass as a carbon source for native bacteria. The bacteria in the absence of oxygen reduce the remaining nitrate to nitrogen gas and reduce selenate to insoluble selenium. The insoluble selenium is then removed from the water by sedimentation in deep ponds and, as needed, by dissolved air floatation and sand filtration. Flowing water through wetlands has been demonstrated to reduce selenium concentrations in water. This system consists of substrate con-

DRAINAGE MANAGEMENT OPTIONS

633

taining organic detrital matter and actively growing plants, all in a flowthrough system of ponds. Removal of selenium occurs by several mechanisms, including reduction of inorganic selenium to elemental selenium, adsorption of selenite to the charged surfaces of minerals and organic matter, plant uptake, and microbial volatilization. The extent of selenium removal by a flow-through wetland system varies with hydraulic residence time and with seasonal changes in temperature and growth rates of plants in the wetland. A positive feature of the wetland flow-through system is that it may provide a relatively inexpensive means to reduce the selenium load in drainage water. A field experiment was conducted to investigate the effectiveness of a wetland flow through system for selenium removal (Tanji et al. 2001). The research evaluated the effectiveness of various types of vegetation and water retention times on selenium removal. The research also identified and quantified the fate of the selenium. Mass balance showed that about 41% of the total inflow selenium left the wetlands cells through outflow, seepage, and volatilization. About 53% of the inflow selenium was retained within the wetland cells, primarily in the surface sediment and organic detritus layer, and the remainder was unaccounted for (Tanji et al. 2001). From an economic point of view, treatment to reduce or remove selenium from drainage water allows it to be disposed of with less impact to wildlife. However, in contrast to the purified water resulting from the RO process, the market value of biologically treated water is not greatly enhanced because of the remaining salt content. Therefore, the added value of the water after treatment cannot be used to offset the cost for the treatment. Drainage Water Disposal in Evaporation Ponds One means of disposing of drainage water is to set aside a portion of land to create a basin for ponding water for evaporation. Except for the limited opportunity to discharge drainage into the San Joaquin River, evaporation ponds in the Tulare Lake basin and a few solar evaporators elsewhere are the only means of isolating salt from productive agricultural lands in use in the SJV as of 2008. Use of evaporation ponds can be severely hindered by the presence of selenium, which can affect wildlife exposed to the systems. Waterborne and sediment selenium within evaporation ponds bioaccumulates into the aquatic food chain by bioconcentration and biomagnification mechanisms. The extent of bioaccumulation depends on the route of exposure (e.g., diet, water, or sediment) and the chemical form of the selenium. Some previously operational evaporation ponds have shut down and are subject to closure and postclosure maintenance plans because of regulatory criteria and costs associated with selenium management.

634

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The future utilization of evaporation ponds for drainage water disposal is dependent on practices to eliminate or minimize bird impacts. In the Tulare Lake basin, a variety of waterfowl and shorebirds seasonally inhabit or utilize evaporation ponds for resting, foraging, and nesting. Waterfowl may be adversely affected from exposure to and bioaccumulation of selenium through the food chain. Adverse impacts may range from impaired health and condition of adult birds, reduced hatchability of eggs, and embryonic deformities. Although species-specific differences exist among waterfowl, the focus has been mainly on the American avocet and black-necked stilt. A number of complex interacting environmental and biological factors need to be taken into account to assess the potential adverse effects of selenium to wildlife. Ultimately, the controlling factor in evaporation pond management will be the nature of regulatory requirements. Presently, waste discharge requirements (WDRs) for drain-water disposal in evaporation ponds are based on the design and management of the ponds, as well as on sitespecific mitigation. WDRs may also specify that compensation habitat or alternative habitat is provided. Compensation habitat is a waterfowl resting, feeding, and nesting area built outside the functional landscape of the evaporation pond to provide breeding habitat in the presence of lowselenium water. Such habitat has been constructed at the Tulare Lake Drainage District (TLDD). Alternative habitat is a waterfowl area built within the functional landscape of an evaporation pond to provide yearround habitat to dilute the diet of birds with respect to selenium. An experimental habitat was constructed at Westlake Farms in Kings County, California. WDRs usually specify one or more protocols for assessing the effects of the pond on waterfowl. Such protocols typically are based on egg selenium content and/or waterborne concentrations (the current WDR for evaporation ponds having elevated selenium require that bird eggs are sampled each year, for measurement of selenium content). The WDR would identify a number of facility designs and operational parameters that are intended to reduce and avoid adverse impacts of the evaporation basin on wildlife. In addition, the WDR may specify wetland habitat to compensate for unavoidable impacts, thereby reducing the overall effect of the proposed basin operations to less than significant levels, as defined by the California Environmental Quality Act (CEQA). Redesign and maintenance of evaporation ponds to reduce impacts to wildlife may include a minimum water depth of 2 ft, steepening levee slopes, reducing vegetative cover, and removing windbreaks. Also included are disease surveillance and control programs, invertebrate sampling, and bird hazing. All of these measures contribute to decreased use of evaporation ponds by birds. Methods that cause disruption of the selenium food chain, such as the commercial production and harvesting brine shrimp

DRAINAGE MANAGEMENT OPTIONS

635

within evaporation ponds, are being developed and implemented. Reduction in selenium concentration in drainage water before discharge into ponds can reduce the hazard. However, none of these practices provides an absolutely safe bird habitat without some potential impact. Results of biological monitoring at evaporation ponds in the Tulare Lake basin have shown substantial reductions in the numbers of nesting waterfowl, particularly American avocets and black-necked stilts, after modifications have been implemented (SJVDP 1999k). The results of monitoring have also shown that the numbers of stilts and avocets successfully nesting at compensation habitats is substantially higher than originally expected. Monitoring is essential to refine the performance of compensation habitats and to address questions such as: • The use of saline water with low-selenium concentrations as a water supply for wetlands • Performance under drought conditions • Alternative wetland habitat design and operations • The relationship between waterfowl production on compensation wetlands relative to the mitigation requirements to reduce unavoidable evaporation basin impacts to less than significant levels • The function of alternative habitats for reducing selenium dietary loads • The contribution of compensation habitat production to the adult waterfowl population and the associated assessment of net environmental benefits. Compliance with WDRs requires monitoring of waterfowl nesting, abundance, nest fate, egg selenium, and embryonic conditions within operating evaporation basins and compensation and/or alternative habitats. Policies that allow for compensation habitats to offset associated impacts of ponds will enhance the future utility of evaporation ponds. Studies (e.g., SJVDIP 1999k) on compensation habitat have shown that: • Compensation wetland habitat can be designed and operated successfully to support high densities of nesting wild birds. • Nesting success can be high at several compensation habitats where predatory exclusion is effective. • A carefully designed vegetation control program can contribute to the long-term success of the mitigation site. • Relatively larger numbers of young waterfowl are produced at compensation wetland habitats when compared to current estimates of waterfowl nesting at several of the evaporation basins. Compensation habitat does require setting aside land and good-quality water that might otherwise be used for agricultural production. However,

636

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

this may be a modest cost if it allows productive agriculture to proceed on acreage that requires drainage. Protocols (Hanson 1995; Hanson Environmental, Inc. 1993; USFWS 1995a) are available to estimate unavoidable adverse impacts on American avocets and black-necked stilts, and the acreage of uncontaminated compensation wetland to mitigate unavoidable losses as required by CEQA. A second protocol (USFWS 1995b) has also been proposed for alternative wetland habitats to provide foraging for targeted waterfowl so that selenium dosing from contaminated basins could be reduced. These protocols calculate compensation and alternative habitats utilizing sitespecific information on waterborne selenium concentrations, abundance of nesting stilts and avocets at the evaporation basin, and the anticipated density (number per acre) of stilts and avocets at a managed wetland site. Such analyses require site- and species-specific appraisals. Although selenium is the principal constituent of concern, others, such as salt and boron, are of concern, too. Other factors include predation, flooding of nests, entrapment in phosphate foams along shorelines, diseases (such as avian botulism), levee maintenance, and other disturbances. These must also be evaluated by separate risk analysis and risk management. During the operation of an evaporation pond, salts are concentrated as a result of evaporation. Those salts eventually must be harvested and utilized or isolated from the environment. One farm in Kern County, California, has converted an evaporation pond to a solar evaporator to minimize wildlife impacts by concentrating the salts into crystals through evaporation. Drainage Reuse Drainage waters can be reused for irrigation of salt-tolerant crops. Integrated On-Farm Drainage Management (IFDM) systems have been designed and put into operation in the SJV. IFDM systems sequentially reuse drainage water on increasingly salt-tolerant or halophytic crops to concentrate and decrease the volume of drainage water. Ultimately, the remaining drainage water is discharged to solar evaporators where the water is evaporated and the salt deposited. Opportunities could be developed for beneficial use of the harvested salt. Implicit in this system is the expectation that water percolating below the rootzone is captured in the drainage system to be passed on to the next, more salttolerant crop. Even for closely spaced drain lines, some of the deep percolate may not be captured in the lines immediately below the field. Depending on the local hydrology and the location of the fields, some of the drainage water may migrate to other areas, and/or the drainage systems may capture considerable water originating from more distant areas.

DRAINAGE MANAGEMENT OPTIONS

637

High salt concentration, and in some cases boron concentration, limits the utility of drainage water for crop production. Some crops are tolerant to both salinity and boron; other crops are tolerant to salinity but not tolerant to boron, or vice versa. Thus, crop tolerance to both salinity and boron relative to the concentrations in the drainage water is important in choosing appropriate crops. Because reusing drainage waters for irrigation is putting the salt back into the system, this practice creates a continual accumulation of salts with long-term limitations. Ultimately, salts must be removed from agricultural lands to maintain productivity. Although some crops are tolerant to salinity, all plants have a limit to salt tolerance, and yields are decreased if the salinity level in the soil and water exceeds the tolerance level. Therefore, leaching accumulated salt from the rootzone soil by the annual application of excess irrigation water is required for all crops if there is not enough rainfall. Indeed, the required leaching fraction to maintain high crop yields may be large if the salinity of the irrigation water is high, even if the crop is salt-tolerant. Destruction of soil physical properties creates crusting, which restricts germination, decreases infiltration rates, and is a hazard associated with using saline irrigation waters. The use of amendments, such as gypsum, can mitigate the negative impact of saline waters on soil physical properties. This is a management practice that imposes an additional cost but must be implemented to prevent loss of soil quality. There may be a temptation to use poorer-quality land for reuse of drainage water. This may create problems, particularly if a fairly high leaching fraction is required. If the soil-water transmission properties are not high, large leaching fractions cannot be achieved and the soil profile may remain excessively wet, leading to oxygen depletion and a negative impact on plant growth. The combination of high salinity and low oxygen supply could greatly reduce yields. Summary In summary, drainage waters can be reused for irrigating salt-tolerant crops. After evaluation of soil and water quality on a given farm, the decision to implement a reuse system is largely one of economic considerations. If a farm has no outlet for drainage water, assigning part of the property to utilize the drainage water, or implementing an IFDM system to ultimately discharge salt into a solar evaporator, may be economically feasible. The economically optimal solution would be derived from comparing the costs associated with source reduction, such as crop use of shallow groundwater, drain-water treatment, and drainage reuse. Included in the cost evaluation should be the economic return from the production of higher-value, less-salt-tolerant crops on soils leached of salt with newly installed drainage systems.

638

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Discharge to the San Joaquin River The Grasslands area in the SJV has the opportunity to discharge some drainage water into the San Joaquin River. The amount of discharge is constrained by the requirement to meet water-quality objectives for the river. The opportunity to discharge salts and selenium on an annual basis, without violating the water-quality objectives, would be increased if the discharge could be timed to match the assimilative capacity of the river. The assimilative capacity varies seasonally because of precipitation and water release to the river. Matching release to the assimilative capacity has commonly been referred to as real-time management. Real-time management is facilitated by the opportunity to control the time and amount of drainage water released. This condition operationally requires the storage of drainage water from times when the drainage water exceeds the assimilative capacity to times when the assimilative capacity of the river is higher. Some storage might be accommodated within the soil profile if a drainage outlet control was in practice. This storage capacity is constrained by the requirement that the water table cannot be maintained too high in the rootzone for an extended period of time. The construction of holding ponds provides another storage option. The main disadvantage of holding ponds is the potential hazard to wildlife using the ponds and being exposed to selenium toxicity. In evaluating the consequences of discharging drainage water into the San Joaquin River, ecotoxicity of selenium compounds probably constitutes the most complex issue. The large gaps in knowledge have their roots in the extensive biogeochemical transformation and bioaccumulation of selenium. These research gaps were addressed in the 1998 “Peer Consultation Workshop on Selenium Aquatic Toxicity and Biocumulation” held by the U.S. EPA (EPA 1998). The consensus from the ninemember panel was that waterborne selenium concentration is not always a reliable indicator of selenium adverse effects on the aquatic top predators. This is because selenium exposure and effects in top predators (the major concern for selenium contamination) is mainly mediated through diet, that is, the food chain organisms in which biotransformation and bioaccumulation occur. The consensus emphasizes that the sediment and its resident food-chain organisms are major sinks for selenium bioaccumulation and biotransformation. Since these biogeochemical processes are highly complex, they may be highly variable from site to site, resulting in the need to address selenium impact on a site-by-site basis. Our present knowledge of these processes is inadequate to allow an extrapolation from waterborne selenium concentrations to selenium impact on top predator on a site-specific basis. Nevertheless, such extrapolation is needed for setting appropriate water-quality criteria for different site conditions. For sustainable protection of water quality, research is

DRAINAGE MANAGEMENT OPTIONS

639

also needed to assess the biogeochemical assimilatory capacity of a given system with respect to biological or ecological impacts. Such impacts are the sole reason of concern over the trace elements such as selenium.

Salt Utilization Major constraints in managing drainage water in the SJV are the salinity effects on plant growth and the selenium toxicity to wildlife. High concentrations of salt are harmful to most plants and, therefore, salt levels in soil and water must be maintained within a certain range in order for productive agriculture to continue. Selenium can be toxic to wildlife, and wildlife exposure to selenium in agricultural drainage must be avoided or minimized. However, both salt and selenium have essential and established beneficial uses in industry, and selenium is an essential nutrient in animal nutrition. Many areas of the world, including parts of California, suffer from a deficiency of selenium. Problems associated with salt and selenium then become ones of separation and distribution, not disposal. An evaluation of these elements as resources rather than pollutants is therefore justified. The salt composition of drain water differs from that of seawater. Seawater contains primarily sodium-chloride salt, whereas drain water from the west side of the SJV typically contains sodium-sulfate salt. When drain water is concentrated by evaporation, the dominant minerals that precipitate are thenardite (sodium sulfate), halite (sodium chloride), gypsum (hydrated calcium sulfate), and calcite (calcium carbonate). The drain water also contains several trace elements of concern: selenium, arsenic, boron, and molybdenum. During the evaporation process, those elements will associate with or become incorporated into the precipitated mineral salts. Such contamination of the salt minerals may have positive or negative implications, depending on the intended use of the salt. The commercial utilization of sodium sulfate includes dying of textiles, glass making, glazing, and other industrial uses. All of these utilization options and others have been evaluated in the Salt Utilization Technical Committee Report (1999h). For certain commercial and industrial uses, salt must first be purified. For example, in the sodium-sulfate industry, purity exceeding 90% may be required. The U.S. market for sodium sulfate is about 1.5 million tons per year, compared to a 1989 estimate for the combined annual deposition of salt in evaporation ponds in the SJV of an estimated 0.8 million tons per year. The harvesting and marketing of that much sodium sulfate could drive down the price, possibly to levels so low that it would become uneconomical to harvest the salt. Transportation must also be considered in planning to utilize SJV salt. The cost of

640

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

freighting the harvested material to a salt refinery or other market must be low enough to provide a profit. Collection of water in solar evaporators facilitates the harvesting of salt. Thus, the feasibility of using solar evaporators, as previously discussed, is relevant to the goal of salt utilization. Indeed, if significant commercial markets were established for the utilization of the salts, it would provide an economic incentive to work toward the utility of solar evaporators. Considering that 2 to 3 million tons of salt influx per year by irrigation water (in addition to significant amounts of salt mobilized from soils as a result of irrigation) need to be disposed of to maintain salt balance in the SJV, even an optimistic estimate of the amount that could be commercially marketed would represent a small percentage of the total salts requiring disposal. Active pursuit of commercial utilization of the salts and selenium is justified, and will require all the other options for separating the salts from productive agricultural fields. However, the salt utilization approach should not negate pursuit of other salt disposal options, such as disposal in lined storage facilities or ocean disposal.

Status of Implementation of Drainage Management Options Drainage reuse Although progress in development of sequential reuse systems has been made, it has not been implemented as predicted by the 1990 Plan. More emphasis is now placed on developing forage and halophytic crops with reused drainage water, as opposed to tree crops that have not generally proven to be effective. The IFDM system was developed as an alternative for farmers in the west side of the SJV who have no drainage disposal outlet. IFDM is a zero-liquid-discharge farming system that manages its drainage water by sequentially reusing it to grow increasingly salt-tolerant crops and evaporating the final effluent in a solar evaporator. Implementation of IFDM technology has demonstrated the cultivation of higher-value crops and increased yields through soil improvement of salt-laden lands. IFDM systems are defined and regulated under Article 9.7 of Health and Safety Code, Section 25209.11, (c), (1 4) of the state of California. An education and outreach program was developed to make information available to individuals and district areas interested in implementing IFDM technology. Landowner and technical advisor manuals—each providing information on IFDM system design, construction, operation, monitoring, and maintenance requirements—have been published (WRCD/CIT 2005a,b). The concepts of drainage reuse have been applied to plans for regional management of drainage water. Regional drainage reuse is a key component currently being used by the Grassland Area Farmers (a regional

FIGURE 20-4. The concept of integrated on-farm drainage management. From California Department of Water Resources. 641

642

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

drainage entity under the San Luis and Delta-Mendota Water Authority) on the west side of the SJV to manage discharges to the San Joaquin River, and is a key component of their long-term drainage management strategy. Approximately 4,000 acres of land within the Grassland drainage area were purchased as part of the San Joaquin River Water Quality Improvement Project (SJRIP) for the purpose of drain-water disposal. The first phase of the SJRIP was implemented in 2001 with the planting of salttolerant crops and construction of distribution facilities. More than 1,800 acres were irrigated with drainage water or blended water. In 2003, 2,420 acres were irrigated with more than 5,300 acre-feet of subsurface drain water. Subsurface drains were being installed within the SJRIP area. This project ultimately allows for planting and irrigation of the entire 4,000 acres with drainage water, along with purchase of additional acreage, installation of subsurface drainage systems, and, ultimately, implementation of treatment and salt-disposal components (San Joaquin River Exchange Contractors Water Authority 2003). In evaluating drainage service alternatives for the federal San Luis unit of the Central Valley Project, the Bureau of Reclamation (USBR) found that in all alternatives evaluated it was most cost-effective to first reduce the volume of water requiring disposal through the use of regional drainage reuse areas (USBR 2006). The USBR plan that was recommended for implementation (USBR 2007) includes up to 16 regional drainage reuse areas totaling approximately 12,500 acres. Reuse-area crops would include salt-tolerant perennial pasture grasses such as Bermuda grass and Jose tall wheatgrass, salt-tolerant alfalfa, barley, canola, and other salt-tolerant grains or forage mixes. Salt-tolerant grasses and grains would be harvested for hay, silage, and/or greenchoppped for local livestock producers. Grazing would also be used to harvest the grasses. The cropping mixes would consume approximately 70% of the incoming drainage water. Subsurface drains would be installed in the reuse areas. The reuse areas would serve to some extent as an underground regulating reservoir to control the flow to downstream treatment and/or disposal facilities (USBR 2006). Drainage treatment Treatment systems were not sufficiently advanced or economical to allow for recommendation at the time of the 1990 Plan. RO filtration systems for the removal of both salt and selenium have now advanced and appear viable for demonstration and implementation. Several pilot or fully operating RO systems have been constructed or proposed for construction in the Grasslands and Westlands subareas of the SJV. However, the brine disposal from a large RO system still remains problematic. Research and development of biological selenium treatment systems, including wetlands treatment systems not envisioned in the 1990 Plan,

DRAINAGE MANAGEMENT OPTIONS

643

have been conducted (Tanji et al. 2001; USBR 2008). Water from filtration or distillation treatment systems may be reused directly on-farm. Biologic selenium treatment systems may be integrated with other drainage management systems, such as evaporation ponds and river discharge, in order to improve wildlife safety. The USBR, in partnership with DWR and others, conducted pilot studies of RO treatment at the Panoche Drainage District and Red Rock Ranch (RRR) in the Westlands Water District between July 2003 and December 2005 for the purpose of collecting data to develop designs and cost estimates for full-scale treatment plants (USBR 2008). The system performed in a stable fashion at 50% recovery1 and 99% rejection2 of total dissolved solids (TDS) and selenium (40%–60% for boron). New, recently commercialized antiscalent chemicals offer the possibility of increasing recovery in the RO units, and tests were conducted to evaluate the performance of RO membranes on drainage water at recoveries higher than 50%. The increase in operation and maintenance costs due to running the RO at higher recoveries is considered modest compared to the significant decrease in costs associated with concentrate disposal that may be realized. However, the antiscalent mixtures tested were unsuccessful and serious calcium sulfate scaling of the RO pilot system and of downstream equipment occurred at 64% recovery. The pilot system was subsequently cleaned and the recovery was reduced to 55%. Scaling continued to occur, even at 55% recovery. Consequently, USBR concluded that 50% recovery should be the assumption for the design of full-scale treatment plants (USBR 2008). WaterTech Partners (Moraga, California) in collaboration with PCI Membranes (Milford, Ohio) developed and tested a system that they termed “double pass preferential precipitation reverse osmosis” (DP3RO™). This process uses a seeded first-pass nanofliltration step in which calcium sulfate is removed as a slurry from the water, followed by a secondpass RO system in which more-soluble salts, principally sodium chloride, are removed. A pilot test performed at the Panoche site was partially successful. The researchers were able to demonstrate significant removal of calcium sulfate and, therefore, an overall system recovery of product water at 88%; however, the solids in the slurry abraded the surface of the nanofiltration membranes. Follow-up tests with a more durable membrane were not conducted long enough to demonstrate whether this problem had been resolved.

1

Recovery refers to percentage of influent water that is recovered as clean product water. 2 Rejection refers to percentage of constituent that is excluded from passing through the RO membrane, i.e., from the product water.

644

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Development and testing of drainage water desalting technologies is continuing. DWR has recently initiated a pilot ion exchange treatment project at Red Rock Ranch. The Panoche Drainage District, on behalf of the Grassland Area Farmers group, is planning further pilot/demonstration projects to investigate treatment options for final salt disposal. Selenium risks to wildlife remain a major constraint on drainage disposal options, particularly to those involving discharge to an open water body such as evaporation ponds or the San Joaquin River. Investigations of selenium treatment in the SJV have focused largely on biological treatment processes. In 2003, USBR became aware of a new biotreatment technology (ABMet®) that was patented and commercialized by Applied Biosciences, Inc. of Salt Lake City, Utah (subsequently Xenon Environmental, now owned by GE Power and Water). In June 2003, USBR in collaboration with DWR and others, initiated a series of pilot projects at the Panoche Drainage District and Red Rock Ranch to investigate the technology and ultimately develop data for design of full-scale selenium treatment plants. The process consists of multiple trains of two bioreactor tanks in series filled with granulated activated carbon (GAC). The tanks are inoculated with cultivated bacteria found to be effective in reducing nitrate and selenium. The GAC media provides a surface area for development of a biological film that reduces the dissolved selenium to solid form that is captured in the biomass. The reducing bacteria are sustained by dosing with a molasses-based nutrient. Initial pilot tests in 2003 at Panoche demonstrated that the technology could reduce selenium in agricultural drain water to below 10 g/L, but various design and operational problems were encountered. Subsequent pilots were conducted using both raw drain water and concentrate reject from an RO system as feed water. Following pilot system retrofits, stable operation of the biotreatment systems was achieved with selenium effluent concentrations at or below 10 g/L at both sites (USBR 2008). Land retirement In the 1990 SJVDP Plan, land retirement was seen as a last-resort measure to manage agricultural lands with high concentrations of selenium in the soil and groundwater. Significantly more land retirement has occurred since 2000 in the Westlands and Grassland subareas than was contemplated in the SJVDP plan. This has been driven as much by litigation and water supply issues, as well as wildlife habitat restoration objectives, as by the drainage considerations originally intended. Approximately 37,000 acres of drainage-impacted lands in Westlands were permanently retired as part of settlement of litigation among landowners, the Westlands Water District, and the United States. In addition, settlement of internal litigation within Westlands concerning water allocation issues resulted in

DRAINAGE MANAGEMENT OPTIONS

645

a program where Westlands has purchased and fallowed approximately 50,000 acres in the district. The entire Broadview Water District in the Grassland subarea (approximately 9,000 acres) was purchased by Westlands in 2005 for the water supply and has been fallowed. Under the Central Valley Project Improvement Act (CVPIA), the USBR along with FWS and the Bureau of Land Management (BLM) has retired approximately 9,000 acres in the Westlands and the Tulare subareas for the purpose of evaluating soil and groundwater response to land retirement and developing and documenting wildlife habitat restoration techniques (U.S. Department of the Interior Interagency Land Retirement Team 2005). A drainage plan put together by water districts in the Westlands and Grassland subareas (San Joaquin River Exchange Contractors Water Authority 2003) proposes permanently retiring up to 200,000 acres (inclusive of the acres described). The USBR drainage service plan (USBR 2007) for the same area proposes retirement of a total of 194,000 acres (again inclusive of the acres described). The Westside Regional Drainage Plan proposes potential uses of retired land to include drainage reuse and treatment facilities, commercial and industrial uses, flood control and water storage, dryland farming, and hunting and wildlife uses. The USBR plan assumed for cost-estimating purposes that retired lands not used for drainage project purposes would be managed by a combination of fallowing, dryland farming, and grazing. The CVPIA land retirement demonstration program has shown a steady decline in groundwater tables beneath retired lands, and that transport and accumulation of salts and trace elements to the surface soils has not occurred (USBR 2005). The program has also demonstrated successful habitat restoration at the Tulare basin site, although initial investment costs can be significantly high, and seed stocks for large-scale native habitat restoration effort do not exist. The long-term management of large tracts of retired land is a significant issue not addressed in the 1990 Plan. Evaporation ponds The efficacy of structural and operational modifications to traditional evaporation ponds in reducing wildlife use was not foreseen by the 1990 Plan; neither was the high bird productivity of compensatory mitigation wetlands. The 1990 Plan recommended an evaporation pond:mitigation habitat ratio of 1⬊1 (one acre of pond with more than 2-ppb selenium required one acre of selenium-free mitigation habitat). Based on more recent findings from studies at mitigation habitats, the WDRs issued by the Central Valley Regional Water Quality Control Board (CVRWQCB) for evaporation ponds have required fewer habitats to mitigate the unavoidable impacts than recommended by the 1990 Plan. Traditional evaporation ponds can now be managed to reduce wildlife impacts and

646

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

meet regulatory requirements, although at a high economic cost. The USBR plan for drainage service for the federal San Luis unit of the Central Valley Project (USBR 2007) uses evaporation ponds as the only currently proven technology for the final in-valley disposal of drainage water and salt after reuse and treatment. However, despite minimizing the volume of drainage water through reuse and RO treatment, and including selenium treatment to below 10 g/L prior to discharge into ponds, design features to protect both fish and wildlife resources as well as groundwater resources add significantly to the cost of evaporation ponds. Including the cost of selenium treatment, the cost of brine disposal remains the chief impediment to drainage service. Nonstandard evaporation ponds proposed in the 1990 Plan have not yet been developed for use in the SJV, although enhanced evaporation and alternatives to evaporation ponds continue to be aggressively explored. Solar evaporators (not proposed in the 1990 Plan) have proven efficacious in evaporating high-salt- and -selenium drainage water while minimizing the hazard to wildlife, acting as the final component of sequential reuse systems. DWR and its cooperators performed a 3-year study of a 10,000-ft2 solar evaporator pilot project for storing and concentrating subsurface drainage water effluent from an SJV farming operation. The goal of this project was to collect information for the development of a farmscale solar evaporator for the 640-acre IFDM system at Red Rock Ranch. Concentrated agricultural subsurface drainage water collected from the last drainage water reuse cycle of the IFDM system is discharged and evaporated in the solar evaporator, using timed spray sprinklers. The remaining salts are stored on the surface of the evaporator to be later recycled or disposed. Enclosed storage is included for water in excess of the evaporation rate so that no surface water is exposed to the atmosphere. The studies included evaluation of salt deposition rates in relation to nozzle height and studies of air quality and particle emissions associated with the solar evaporator. The results are promising and the Westlands Water District is proposing the use of the sprinkler solar evaporator system in lieu of the evaporation ponds that are included in the USBR drainage service plan. Other evaporation technologies have been pursued, including solar distillation for water and salt recovery, and salt-gradient solar ponds for production of electricity. DWR engineers investigated the use of a closedsystem solar still at Red Rock Ranch. Although the technology did not appear to be economically competitive as an alternative to the sprinkler solar evaporator system, the effort recovered distilled water and provided data on the chemical process of the salt crystallization stage, which may be useful for reclamation of salts. DWR also led the principal investiga-

DRAINAGE MANAGEMENT OPTIONS

647

FIGURE 20-5. Pilot solar evaporator at Red Rock Ranch in Fresno County, California. tion into salt-gradient solar ponds with the Bureau of Reclamation, the University of Texas–El Paso, and the California Department of Food and Agriculture as project collaborators. Five potential solar pond locations were identified in an earlier report prepared by USBR (Lu et al. 2001). Preliminary investigations were conducted to collect information on the following: regulatory requirements, local markets for potential application of the heat energy, geotechnical analysis, drainage-water pretreatment, and an environmental impact assessment. Thus far, there has not been sufficient further interest or funding for implementation of a salt-gradient solar pond project in the SJV. Source reduction Improved methods of water distribution uniformity and water-use efficiency allow source reduction to nearly meet or exceed the projections of the 1990 Plan. Water supply limitations that have occurred since 1990, as well as drainage management considerations and other economic considerations, have all contributed to increasing acreage being converted to drip irrigation and the adoption by farmers and districts of other irrigationefficiency measures. Although measures have been widely implemented

648

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and considerable source reduction achieved, further reductions can be made in some areas. In particular, techniques for increased direct use of shallow groundwater by deeper-rooted and salt-tolerant crops may be refined and more widely implemented. Groundwater management Unlike all the other recommendations of the 1990 Plan, no coordinated regional effort has been made to implement groundwater management as a drainage management measure due to its perceived infeasibility. However, lowering of the shallow groundwater table has been an indirect effect of ongoing groundwater pumping, particularly during drought years, and the SJVDIP Technical Committee concluded that coordinated, monitored, and managed groundwater extraction could still be an effective means to manage drainage through lowering the regional shallow groundwater table. The committee further outlined the technical and regulatory process that would allow for implementation. However, an exception to the regulatory prohibition of degradation of a water supply would have to be made in order for groundwater management to be implementable. River discharge The focus of management of drainage discharge to the San Joaquin River is on meeting selenium load limits, which were based on water year types, and became progressively more restrictive until fully effective in the year 2010. The 1990 Plan recommendation to extend the San Luis Drain (a part of which now functions as the Grassland Bypass Channel) to the San Joaquin River below the Merced River confluence is now viewed as requiring additional analysis to determine its need. The 1990 Plan projected increased discharge of drainage water to the San Joaquin River subject to water-quality objectives. However, the adoption by the CVRWQCB of a selenium total maximum daily load (TMDL) for the river and, more recently, the adoption of TMDLs for salt and boron, severely constrain opportunities for discharge of drainage water to the river. Drainage plans developed by both the Grassland subarea districts (San Joaquin River Exchange Contractors Water Authority 2003; USBR 2007) now anticipate zero discharge of subsurface drainage to the river, relying instead on the reuse, treatment, and disposal technologies previously discussed. Extension of the Grassland Bypass Channel/San Luis Drain as recommended in the 1990 plan could still have utility in the future as part of a plan to discharge high flows resulting from extreme hydrologic events and/or discharge salts from the basin in conjunction with a real-time water quality management program.

DRAINAGE MANAGEMENT OPTIONS

649

Salt and selenium utilization A major aspect of the 1990 Plan was the postponement of addressing the issue of finding an appropriate endpoint for salt. The 1990 Plan left the salt to gradually build up in the SJV soils, become stored in the semiconfined aquifer by groundwater management, and then discharge to evaporation ponds with a relatively small amount being discharged to the San Joaquin River. The Ad Hoc Coordination Committee Report (SJVDIP 2000), a synthesis report prepared by the SJVDIP evaluating the status of implementation of the 1990 plan, recognized the separation of salt from SJV soils and groundwater and disposal or utilization of that salt to be the fundamental issue of drainage management and agricultural and environmental sustainability. Commercial utilization of salts separated from agricultural drainage water was not a component of the 1990 Plan. Opportunities for the commercial marketing and utilization of some salt products may exist if economical separation, purification to commercial standards, and marketing of agricultural salts can be developed (Table 20.1). The same is true for selenium, with the addition of the now-recognized important health and nutritional benefits of selenium in the diets of both humans and animals. As of 2008, salt harvesting continued to be a focus of the solar evaporator test program at Red Rock Ranch. The solar evaporation concentration ratios indicate that calcium and bicarbonate begin precipitating at the TABLE 20-1. Value of Extractable TDS of Red Rock Ranch Subsurface Drainage Water Constituents Constituent (1)

Sodium Sulfate Sodium Chloride Calcium Sulfate Magnesium Chloride Sodium Nitrate Calcium Carbonate Boron Potassium Chloride Selenium Other

Average Dry Weight (%) (2)

Average Market Price, 2006 (3)

37 33 16 7 3 2 1 0.8 0.1

$120/ton $25/ton $30/ton $270/ton $300/ton N/A $425/ton $270/ton $50/lb

0.19

Source: U.S. Geological Survey Mineral Commodities and Chemical Market Reporter.

650

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

onset of evaporation and continue throughout the process. A sequence of process steps to separate sodium chloride, sodium sulfate, boric acid, magnesium oxide, and potassium nitrate as salable products is proposed. Refined selenium is a commodity used in electrical devices, pigments, glass production, and metallurgy. The market for selenium has increased significantly in recent years and remains strong at approximately $60 per pound in 2007. Recovery of selenium from the waste biomass produced in the bioreactors is being explored. If this is feasible, recycling the selenium byproduct would result in significant savings in waste disposal costs associated with the selenium treatment system tested and proposed by USBR.

SUMMARY AND CONCLUSIONS The management options reviewed here can contribute to in-valley drainage management strategies. However, except for what can be discharged into the San Joaquin River or discharged to the ocean in some other manner, evaporation ponds and solar evaporators—in possible combination with RO treatment systems—are the only options for longterm separation of salts from the soils of SJV agricultural lands. Various practices can be implemented to decrease the volumes of water requiring ultimate disposal into evaporation systems. However, all options (other than discharge and separation in evaporation systems) maintain salinity in agricultural lands, with long-term consequences. For example, blending drainage waters with good-quality surface waters can be used for irrigation. However, in the absence of adequate leaching of salt from the rootzone and salt removal, this practice will contribute to the continual salinization process, with its resultant long-term negative impacts on agricultural productivity. Since evaporation systems serve as the only repository for salts that isolates them from productive agricultural fields, regulations on the operation of evaporation ponds will serve a pivotal role in determining the longrange agricultural productivity in the western SJV. The primary problem with evaporation ponds is potential selenium toxicity to waterfowl. Several steps can be taken to reduce the hazard to birds. Examples include design and management of ponds to reduce their attractiveness to birds, bird hazing, various approaches to disrupting the selenium food chain, and reduction of selenium concentration in the water by treatment prior to discharge into evaporation ponds. A combination of these steps can be taken to greatly reduce negative impacts on birds, but it would be virtually impossible to design a system to be completely bird-safe. To reduce wildlife impacts when the selenium concentration is above 50 ppb, the 1990 Plan recommended accelerated-rate evaporation ponds, where water would be pumped and sprayed at an elevation aboveground for enhanced

DRAINAGE MANAGEMENT OPTIONS

651

evaporation of drainage water. This conceptual project has not been tested. However, a solar evaporator that is a modified form of an accelerated-rate evaporation pond has been tested in the SJV (California Department of Water Resources 2006). In a solar evaporator, drainage water is discharged to the pond at a rate equal to daily evaporation, and thus no ponding occurs. Conversion of evaporation ponds to solar evaporators may present an option to reduce wildlife impacts. A cost/benefit analysis of operation of an evaporation pond and its mitigation and compensation habitat requirements in comparison to operation of a solar evaporator and associated drainage reuse system should be done prior to selection. Studies on compensation habitat during recent years have shown that they can be designed and operated successfully to support high densities of nesting waterfowl. A policy that permits mitigation and compensation of potential impacts increases the opportunity to use a combination of management options to sustain high agricultural productivity in the SJV. There has been substantial progress in many aspects of salinity, drainage, and selenium management in the SJV since 1990, which has allowed profitable agricultural production to continue in the region. There has been a substantial reduction in the volume of drainage from agricultural lands, and progress has been made in drainage-water treatment and disposal. Ongoing programs evaluate the feasibility and economics of desalination, solar ponds, solar evaporators, and the use of the salts in irrigation drainage. These have extended the life of irrigated agriculture, but groundwater management has not been addressed regionally and a permanent drainage solution has not yet been designed or implemented.

REFERENCES Ayers, J. E., Schoneman, R., Mead, R., Soppe, R., Dale, F., and Mesa, B. (1996). Shallow groundwater management project, Final report to the California Department of Water Resources, Contract B-58166, California Department of Water Resources, Sacramento, Calif. Belitz, K., and Phillips, S. P. (1995). “Alternative to agricultural drains in California’s San Joaquin Valley: Results of a regional-scale hydrogeologic approach.” Water Resour. Res., 31(8), 1845–1862. Brown and Caldwell Consulting Engineers. (1987). Screening potential alternative geographic disposal areas, prepared for the San Joaquin Valley Drainage Program by Brown and Caldwell, Walnut Creek, Calif. California Department of Water Resources. (2003). Desalination Demonstration Report for Buena Vista Water Storage District, Bakersfield, Calif. ———. (2006). White paper: Solar evaporator for integrated on-farm drainage management system at Red Rock Ranch, San Joaquin Valley, Calif.

652

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Hanson, C. (1995). “Testimony in the matter of the petitions of Lloyd Carter and Patrick Porgans, Bay Institute of San Francisco, et al., and U.S. Fish and Wildlife Service for review of waste discharge requirements adopted by and environmental impact reports certified by the California Regional Water Quality Control Board, Central Valley Region, for 14 Tulare Lake Basin Agricultural Drainage Dischargers.” SWRCB/OCC FiZes A-858, A-858(a) and A-858(b). Hanson Environmental, Inc. (1993). Tulare Lake Drainage District north, Hacienda, and south evaporation basins: Kings County site-specific biological impact analysis and response to comments, prepared for the California Regional Water Quality Control Board, Central Valley Region, June 1993, Hanson Environmental, Inc, Walnut Creek, Calif. Lu, H., and Walton, J. C., Irvine, S., and Remmers, H. (2001). Conceptual application and feasibility of salinity gradient solar pond technology in San Joaquin Valley, California, Memorandum Report, California Department of Water Resources, December 2001, Sacramento, Calif. Lundquist, T. J., Green, F. B., Tresan, R. B., Newman, R. D., Oswald, W. J., and Gerhardt, M. B. (1994). “The algal bacterial selenium removal system: Mechanisms and field study,” in Selenium in the environment, W. T. Frankenberger Jr. and S. Benson, ed., Marcel Dekker, Inc., New York, 251–278. San Joaquin River Exchange Contractors Water Authority. (2003). Westside Regional Drainage Plan, prepared by San Joaquin River Exchange Contractors Water Authority, Broadview Water District, Panoche Water District, Westlands Water District, available at www.waterboards.ca.gov/centralvalley/ water_issues/salinity/programs_policies_reports/westsd_regnl_drng_plan_ may2003.pdf, accessed February 8, 2011. San Joaquin Valley Drainage Program (SJVDP). (1990). A management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley, California: Final report of the San Joaquin Valley Drainage Program, U. S. Department of Interior and California Resources Agency, Sacramento, Calif. San Joaquin Valley Drainage Implementation Program (SJVDIP). (1991). A strategy for implementation of the management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley, U. S. Department of the Interior and California Resources Agency, Sacramento, Calif. ———. (1999a). Drainage reuse technical committee report. ———. (1999b). Drainage treatment technical committee report. ———. (1999c). Land retirement technical committee report. ———. (1999d). Evaporation pond technical committee report. ———. (1999e). Source reduction technical committee report. ———. (1999f). Groundwater management technical committee report. ———. (1999g). River discharge technical committee report. ———. (1999h). Salt utilization technical committee report. ———. (1999i). Grasslands subarea report. ———. (1999j). Westlands subarea report. ———. (1999k). Tulare/Kern subarea report. ———. (2000). Ad hoc coordination committee report, evaluation of the 1990 drainage management plan for the westside of the San Joaquin Valley, California Department of Water Resources, Sacramento, Calif.

DRAINAGE MANAGEMENT OPTIONS

653

Tanji, K., Gao, S., Terry, N., and Lin, Z. Q. (2001). Selenium removal and mass balance in a constructed flow-through wetland system, UC Center for Water Resources, Salinity/Drainage Program Annual Report 2000–2001, University of California at Riverside, Riverside, Calif. URS Corp. (URS). (2002). Source control technical memorandum, prepared for the Bureau of Reclamation, San Luis Drainage feature re-evaluation, June 17, 2002, URS Corp., San Francisco, Calif. U.S. Department of the Interior, Bureau of Reclamation (USBR). (2006.) San Luis drainage feature re-evaluation, final environmental impact statement, U.S. Department of the Interior, Bureau of Reclamation, Fresno, Calif. ———. (2007). San Luis drainage feature re-evaluation, record of decision, U.S. Department of the Interior, Bureau of Reclamation, Fresno, Calif. ———. (2008). San Luis drainage feature re-evaluation, feasibility report, U.S. Department of the Interior, Bureau of Reclamation, Fresno, Calif. U.S. Department of the Interior Interagency Land Retirement Team. (2005). “Land Retirement Demonstration Project, five-year report,” http://esrp.csustan.edu/ projects/lrdp/, accessed February 8, 2011. U.S. Environmental Protection Agency (EPA). (1998). Report on the peer consultation workshop on selenium aquatic toxicity and bioaccumulation, EPA Report No. 822R98007, U.S. EPA, Washington, D.C. U.S. Fish and Wildlife Service (USFWS). (1995a). Alternative habitat protocol for drainwater evaporation basins, Sacramento Fish and Wildlife Office, Sacramento, Calif. ———. (1995b). Compensation habitat protocol for drainwater evaporation basins, Sacramento Fish and Wildlife Office, Sacramento, Calif. Westside Resource Conservation District (WRCD). (1999). Integrated system for agricultural drainage management on irrigated farmland, Final Report, Grant No. 4-FG-20-11920, U.S. Department of the Interior, Bureau of Reclamation, Washington, D.C. Westside Resource Conservation District and Center for Irrigation Technology (WRCD/CIT). (2005a). A technical advisor’s manual, managing agricultural irrigation drainage water: A guide for developing integrated on-farm management systems, Center for Irrigation Technology, California State University, Fresno, Calif. ———. (2005b). A landowner’s manual, managing agricultural irrigation drainage water: A guide for developing integrated on-farm management systems, Center for Irrigation Technology, California State University, Fresno, Calif.

This page intentionally left blank

PART SIX: LAND RECLAMATION, TREATMENT AND DISPOSAL OF DRAINAGE WATERS

CHAPTER 21 RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS R. Keren and S. Miyamoto

INTRODUCTION Saline-sodic soils exist in arid and semiarid climates and in regions with poor drainage. Irrigation with poor-quality water, inadequacy of leaching, seepage from canals high-lying adjacent areas, together with the presence of a high water table and high evaporation rate, all account for the secondary salinization of irrigated soils. The composition and concentration of salts in the soil solution influence the growth of plants by osmotic and specific-ion toxicity effects and by changing the physical properties of the soils. The accumulation of dispersive cations, such as sodium (Na), on the exchange phase affects the physical properties of the soil. This, in turn, affects the production of crops. A low infiltration rate (IR) is associated with sodic soils. Irrigating with water of low electrolyte concentration (such as rainwater) results in ponding and runoff of water. Plants, which respond only to water in the rootzone, consequently have less available water. The reclamation of sodic soils ameliorates adverse conditions, such as surface crusting, and increases the hydraulic conductivity (HC), which affects the infiltration and storage of water, the emergence of seedlings, and the development of roots. Whereas adsorbed Na and Ca ions affect plant growth through changes in soil physical properties, boron (B) may cause injury to plants even when present at very low concentrations in the soil solution. Because plants obtain B from the soil solution, and since there is a relatively small range between B levels in the soil solution that cause deficiency and toxicity symptoms in plants, B distribution between solid and liquid phases 655

656

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of soil under various conditions is important. The suitability of irrigation waters is evaluated on the basis of B concentration, which determines the potential of the irrigation water to injure plants and reduce yield. It is important, therefore, to clarify the conditions that allow the soil to act as a buffer with respect to B concentration in the soil solution. This knowledge can improve the efficiency of using waters of varying B concentrations. Irrigated agriculture in semiarid and arid regions has faced the challenge of sustaining its productivity for generations. Because of natural hydrological and geochemical factors, as well as irrigated-induced activities, soil salinity, sodicity, and associated drainage and soil erosion continue to plague agriculture. Considerable progress has been made in managing and controlling salinity and sodicity in irrigated lands. However, coping with salinity and sodicity problems has become much more complex because of changes in soil physical properties and increasing environmental constraints. A successful irrigated agriculture requires permanent control of salinity, sodicity, and B levels in soils and irrigation water. The reclamation of sodic soils ameliorates adverse conditions, such as surface crusting, and increases the HC, which affects the infiltration and storage of water, the drainage, the emergence of seedlings and the development of roots. Reclamation of affected soils can reduce natural hazards and make a stable background for intensive cropping. Such reclamation will be the focus of this chapter.

EFFECTS OF SALINITY AND ADSORBED IONS ON SOIL PROPERTIES Smectites are the dominant clay mineral in many semiarid and arid regions, and they determine much of the physical properties of soils, such as swelling and dispersion, due to the large specific surface area and the electrostatic charge. These two processes alter the geometry of soil pores and thus affect soil properties such soil structure, permeability, and soil water retention. Montmorillonite platelets in aqueous suspension may flocculate in three possible modes of particle association (Van Olphen 1977): (1) association between siloxane planes of two parallel platelets (FF), (2) association between edge surfaces of neighboring particles (EE), and (3) association between an edge surface and a siloxane planar surface (EF). In the presence of electrolyte (e.g., NaCl) at low concentration, the double layers on the Na-montmorillonite particles are well developed and osmotic repulsion prevents particle association. Thus, a stable suspension of individual platelets is obtained (Van Olphen 1977). As the concentration of NaCl in the suspension increases, double layers at both the planar and the edge

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

657

surfaces are compressed, and both EF and EE association can occur (Heller and Keren 2001). As the electrolyte concentration increases further, the FF mode of particle association occurs, and “oriented aggregates” are formed (Keren et al. 1988). The mechanism of swelling is analogous to the osmotic characteristics of a semi-permeable membrane. The ions between the clay platelets in the exchange phase are constrained there by the electrical attraction of the negatively charged clay surfaces. Thus, under a nonequilibrium condition, the gradient for water movement is from the bulk solution to the inner platelet region, since the concentration of ions is greater there. The movement of the water into this region creates a hydrostatic pressure that forces the clay to expand and swell. A Na soil with a larger number of ions in the inner platelet region, coupled with their weak interaction with the clay, is therefore more susceptible to swelling and dispersion than a Ca soil. Therefore, excess adsorbed Na most adversely affects the permeability of the soil (Alperovitch et al. 1981; Frenkel et al. 1978; Keren and Singer 1988; McNeal et al. 1968), which decreases with the square of the pore radius. Swelling, or the movement of clay, or both, significantly affects permeability. The extent of swelling, or dispersion of clays, or both, depends on the mineralogy of the clays, composition of the adsorbed ions, concentration of salt in the solution, and content of hydroxy-aluminum and -iron (Goldberg and Glaubig 1987; Keren and Singer 1988, 1989, 1991; Oster et al. 1980; Shainberg et al. 1971). The swelling of clay increases as the electrolyte concentration of the solution decreases. The dispersion of clay occurs only at electrolyte concentrations below the flocculation value (FV) of the clay. The degree to which the clay swells before the solution is replaced with one that has an electrolyte concentration below the clay’s FV determines whether clay particles will move into the conducting pores and leave the rootzone, or remain trapped in the narrow pores and decrease the HC of the soil (Keren and Singer 1988). When leached with water of low salt concentrations, the soil’s potential for mineral weathering (i.e., for releasing electrolytes) influences the HC. Calcareous soils and soils with feldspar minerals release electrolytes into the solution in the presence of exchangeable Na. If the electrolyte concentration of the soil solution is maintained above the FV, the clay does not disperse and the HC changes minimally. The electrolyte concentration of the applied water controls the HC of soils without weathered minerals. When these soils are leached with low-salt waters (below FV), clay disperses and the HC decreases. Clay swelling and dispersion depend, in part, on the type of exchangeable cations and the electrolyte concentration of the soil solution. Because the effect of exchangeable Na on swelling and dispersion of clays is countered by a high electrolyte concentration, soil sodicity hazard cannot be assessed independently of information on the accompanying level of

658

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

electrolyte concentration. Thus, an important parameter in assessing sodicity hazard is the total electrolyte concentration of the applied water and the change in the soil solution concentration and electrolyte composition during the growing season. It is obvious, therefore, that in any attempt to assess sodicity hazard in soils, the soil texture, clay mineralogy, exchangeable sodium percentage and total electrolyte concentration of percolating solution should be considered. In coarse-textured soils that contain a limited amount of swelling clay, dispersion and particle transport is the main process in the reduction of HC. Usually there is a physical plugging of pores. In general, however, coarse-textured soils are more easily managed when the exchangeable sodium percentage (ESP) could be a problem. In soils having stable surface structures, decreases in IR result from the inevitable decrease in the matric suction gradient that occurs as infiltration proceeds. Decreases in soil IR from an initially high rate can also result from gradual deterioration of soil structure and the formation of a surface crust. When a crust of low concentrated HC is formed, its reduced permeability determines the IR of the soil and the wetting front depth has only a slight effect on the infiltration rate (Morin et al. 1989). Thus, in cultivated soils from semiarid regions, the organic matter content is low, soil structure is unstable, and sealing is a major determinant affecting the steadystate IR (Morin et al. 1989). Seal formation at the soil surface is in turn due to two processes: (1) physical disintegration of soil aggregates and their compaction caused by the impact of water, especially water drops; and (2) chemical dispersion and movement of clay particles and the resultant plugging of conducting pores. Both of these processes act simultaneously, with the first enhancing the second. Similar to HC, the IR depends also on the salt concentration in the irrigation water and the ESP of the soil (Agassi et al. 1981). Water infiltration, however, is more sensitive to ESP than the HC of the soil profile for the following reasons (Oster and Schroer 1979): (1) the mechanical impact of the water drops, (2) the absence of the soil matrix, which slows clay movement, and (3) concentration of electrolytes in the surface soil solution is determined solely by the composition of the applied water, because dissolution of CaCO3 and primary minerals is too slow to affect the surface solution concentration. Thus, when water with low electrolyte concentration is applied (rainwater or snow water), salt concentration in the soil surface solution remains low even for calcareous soils, and clay dispersion is possible. Conversely, soil solution concentration deeper in the soil profile is affected by CaCO3 and primary mineral dissolution and might be maintained at a concentration above 3 molc m3, which is enough to prevent clay dispersion at low ESP values. It is generally found that a soil containing nonexpanding clays as illite (mica), kaolinite, and sesquioxides are less sensitive to the ESP value. In

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

659

these soils, a total clay analysis could be misleading for interpreting the sodic hazard unless a mineralogical identification is also included in the analysis. Saline-sodic soil reclamation requires a certain flow of water through the soil profile and, to be effective, an appropriate profile HC must be achieved. The end result of reclamation must be a decrease in erodibility to reduce erosion, and a sufficient and stable porosity that provides a favorable physical environment for plants.

RECLAMATION OF SALT-AFFECTED SOILS Concepts and Principles Leaching is the usual way to reclaim salt-affected soils, since plant uptake removes insignificant amounts of salt. It is suited to areas where leaching water is available. Nonirrigated areas must rely on natural precipitation for leaching. Salt leaching involves the dissolution of soluble salts in the soil, the passage of water through soil profiles, and the removal of salt from the rootzone. Thus, soils to be reclaimed must be permeable and have outlets for drainage. Drainage systems may need to be established before soils with high water tables are leached. The extent of leaching required depends largely on initial soil salinity, the salt tolerance of the crops, and the depth of the water table. The prevailing idea in earlier reclamation trials was to leach all excess salts from the entire depth of the rootzone. However, under good drainage conditions, the aim is to reduce salinity in the top 45 to 60 cm of the soil to below the threshold values of the crop (see Chapter 13). Land-use objectives dictate the amount of time to allot for leaching. To reclaim salt-affected virgin land, leaching may be conducted for several months or more. To reclaim cropped fields, such as permanent pastures and orchards, leaching must take place in a short time period to avoid lack of aeration. Leaching generates highly saline drainage water, which can contaminate water and contribute to rising water tables. Evaluating the environmental and legal implications must be a part of any reclamation program. Efficiency of Salt Leaching and Water Requirements The efficiency of salt leaching can be defined as the quantity of soluble salts leached per unit volume of water applied. When leaching must take place within a short time or under high evaporative conditions, also consider the time factor. Factors that determine efficiency as defined include

660

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the initial salinity of the soil, the amount of leaching water applied, the content of the soil water during leaching, hydrodynamic dispersion, spatial variation, and, when applicable, layout of drainage systems. One-dimensional leaching The efficiency of salt transport—the salt leaching efficiency normalized by the initial concentration of salt—increases as the contents of the soil water decrease during leaching. The range of soil pore size distribution also affects the efficiency of salt transport. Salt-transport efficiency is especially low in deeply plowed, wet, clay-rich fields because the leaching water percolates between large clods that are too wet to slake. Drying and disking the soil reduces this problem, yet pore sizes may vary enough to cause substantial hydrodynamic dispersion, especially during ponded leaching. Such hydrodynamic dispersion decreases under unsaturated flow, resulting in the more efficient transport of salt (Nielsen et al. 1966). Hoffman (1980) proposed the following empirical formula for the salttransport efficiency under one-dimensional leaching based on field data from various parts of the world: c k  co ⎛ D ⎞ ⎜ ⎟ ⎝ Ds ⎠

(21-1)

where C  salt concentration in the soil; Co  initial salt concentration in the soil; D  depth of leaching water applied; Ds  depth of the soil to be leached; and k an empirical coefficient that ranges from 0.1 (for sandy loam) to 0.3 (for clay) under ponded leaching as shown in Fig. 21-1 part A. Hoffman found that the coefficient of 0.3 also applies to silty clay loam, silty clay, and clay loam. He also found that, under intermittent ponding with 5 to 15 cm per application (Fig. 21-1 part B), the empirical coefficient is about 0.1, irrespective of the type of soil. Oster et al. (1972) found that sprinkler irrigation of silty clay produced a similar coefficient. The equation here applies after the actual drainage has begun or in the range of D/Ds  k. Equation 21-1 can readily be used to calculate the amount of water needed to leach salts from a desired depth of soil for a specified initial and desired salinity. The salt transport efficiency decreases sharply when D/Ds ratios exceed 0.5 for sandy loam and 0.75 for clay loam to clay. Even after adjusting for lesser water requirements, leaching by sprinkling can take more time than continuous ponding. This is because leaching water is applied at the rate below the saturated HC. It increases the leaching time and water evaporation, which, in turn, increases the salinity

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

661

FIGURE 21-1. Relationship between fraction of initial salt concentration remaining in soil, C/Co, and depth of leaching water applied per unit depth of soil, D/Ds, (A) for continuous ponding; (B) for intermittent ponding. Data from Hoffman (1980).

of leaching water. Carter and Fanning (1964) and Minhas and Khosla (1986) demonstrated that leaching by sprinkling or intermittent ponding works better at low evaporation rates. Drainage conditions affect redistribution of soil water and salts. In well-drained, uniform soils, water after infiltration penetrates deeper into the profile than in poorly drained soils with high water tables or abrupt soil stratification. Consequently, salts will be leached deeper (Terkeltoub and Babcock 1971). Two-dimensional leaching Leaching into tile or open-ditch drains rapidly removes salt directly above and near the drains, and removes little midway between the drains. Leaching efficiency decreases as soon as the rapid flow reaches the drains. This uneven leaching is more evident in water-saturated fields (Luthin et al. 1969) and is less evident in unsaturated fields, where leaching above the water table is a one-dimensional process (Talsma 1967). To leach more efficiently, avoid ponding water near the drains (Miyamoto and Warrick 1974), or intermittently apply water to keep the water table as low as possible (Talsma 1967).

662

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Leaching of spatially variable fields When fields consist of several types of soils, leaching takes place first in sandy portions and then in clay-rich portions. If leaching is continued until the clay-rich portion is sufficiently leached, an excessive amount of water percolates into the sandy area. This problem is pronounced under ponded leaching, during which ponded water percolates predominantly through the sandy portion. In fact, the distribution of the salinity of the soil in basin-, border-, or furrow-irrigated fields coincides closely with the distribution of soil type (Miyamoto and Cruz 1986, 1987). To leach much more efficiently, leach based on the distribution of type of soil (Miyamoto and Cruz 1986), or use the stochastic analysis of solute transport in spatially variable fields (Russo 1984) to divide a field into small sections that need different amounts of water. Leaching Methods Continuous ponding This method is used extensively in surface-irrigated areas, and usually accomplishes leaching in the shortest period of time with the least cost in slowly permeable fields. The leaching efficiency is less than with other methods, especially when used in water-saturated fields with tile or open drains (Talsma 1967). The land must be leveled and, when permeability is low, must also consist of subsoils or be deeply plowed (Longenecker and Lyerly 1974). Ideally, the soils should be dry before leaching. The permeability and initial salinity largely control the rate of leaching. For more details, refer to Reeve et al. (1948), Reeve et al. (1955), Rasmussen et al. (1972), and Hulsbos and Boumans (1960). In clay-rich soils with minimal permeability, ponded leaching may be conducted along with the temporary cultivation of rice or stock fish. Intermittent ponding This method is especially suitable in fields with tile drains, as it allows the water tables to draw down, which greatly increases leaching efficiency (Talsma 1967). When soils develop a surface seal, intermittent ponding may also help water to infiltrate by forming cracks. However, the longer period of wet soil exposure increases evaporation. Carter and Fanning (1964) demonstrated that combining intermittent ponding with mulching greatly improved the performance of intermittent ponding over a 5-month period. Intermittent ponding with mulch is especially suitable when low rates of drainage and high water tables exist. Most soils in the middle Rio Grande project have been reclaimed by deep plowing and then intermittent ponding.

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

663

Sprinkling This method can be used advantageously when the field is unprepared for ponded leaching. It has high leaching efficiency (e.g., Nielsen et al. 1966), except under high evaporative conditions, such as windy days, but is energy- and capital-intensive. It is compatible with unleveled sandy soils, especially when water for leaching is scarce and costly. It can also be used advantageously when a large drainage load is prohibited. Alternate row or border leaching Alternate-row irrigation is used in saline areas to minimize the accumulation of salt in crop beds. Stewart et al. (1976) found that a similar system removed salts in areas with a high water table. High ridge rows are made and every other furrow is then flooded. This moves salts laterally into the unirrigated furrows. The seeped water, because of high water table, can then be surface-drained away. Another method used successfully in salt-affected orchards with high water tables is alternate-border leaching (Miyamoto 1988). Two borders are placed along each row of trees. Leaching water is applied to the border strip but not to the strip between the rows. This provides rapid leaching of salt near the trees and pushes some of the salts into the dry strip between the rows. Salts accumulated between the rows would be leached after lowering the water table. In that study, this method kept trees alive until drainage was completed. Surface flushing High concentrations of soluble salts are often found at and near the surface of the soil, especially in soils with shallow water tables. It is desirable to remove surface-accumulated salts instead of leaching into the soil profile. However, Reeve et al. (1955) found that passing water over the surface by sheet flow failed to remove the salts. When plowed fields with a clay substratum are bedded and the aforementioned alternate-row leaching technique used, the efficiency may improve. The physical removal of salt crusts by mechanical means is an alternative. Post-Leaching Operations The last phase of reclamation is drying the field and, when necessary, taking steps to enhance soil structure, such as applying organic matter or gypsum and cropping with grains or legumes. Prolonged ponding destroys the structure of the soil and depletes nitrogen in the soil. Sprinklers or intermittent ponding generally cause less destruction and may not need elaborate post-leaching measures.

664

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

RECLAMATION OF SODIUM-AFFECTED SOILS Concepts and Principles Reclamation of sodium-affected soils usually involves replacing exchangeable Na with Ca2. This Ca may originate from the dissolution of Ca-containing minerals in the soil, such amendments as gypsum and calcium chloride, or irrigation water with Ca ions. A significant factor in reclaiming sodic soils is the maintenance of HC by providing a sufficiently high electrolyte concentration in the soil solution to counter the influence of exchangeable Na. Generally, the higher the electrolyte concentration, the higher the exchangeable Na fraction (ENa) at which a relatively high permeability can be maintained (Quirk and Schofield 1955). McNeal and Coleman (1966) showed that soils responded differently to the same combination of electrolyte concentration and ENa. Each soil has a unique threshold of salinity concentration. The electrolyte concentration affects the HC less when the contents of the soil-water is low (Russo and Bresler 1977). If the electrolyte concentration of the percolating solution is adequate to reduce clay swelling, the permeability of the soil remains high. When low-salinity water (or rainwater) follows the saline water, permeability can be maintained by applying an electrolyte source on the surface of the sodium-affected soil. Applying a slow-dissolving salt adds sufficient electrolytes to the rainwater to prevent clay dispersion (Keren and Shainberg 1981), even if the Na-Ca exchange process in the adsorbed phase is limited (Keren et al. 1983). The effectiveness of amendments in reclaiming sodic soils depends on their dissolution properties (Kemper et al. 1975; Keren and O’Connor 1982b). In spite of the documented effect of the electrolyte concentration in the soil solution on the permeability of the soil, the main criteria for determining the sodic hazard are the soil ENa and the sodium adsorption ratio (SAR) of the saturation extract or irrigation water (Chapter 3). However, soil sodicity cannot be assessed without information on the total electrolyte concentration of the soil solution.

Reclamation of the Soil Profile by Adding Soil Amendments Amendments commonly used to provide soluble Ca include gypsum (CaSO4 2H2O) and calcium chloride dihydrate (CaCl2 2H2O). Amendments that produce Ca in calcareous soils by enhancing the conversion of soil CaCO3 to gypsum include sulfuric acid, sulfur, and iron and aluminum sulfates. Gypsum, sulfuric acid, and sulfur are the amendments most commonly used for reclamation due to their effectiveness and relatively low cost. Chemical amendments may improve water penetration

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

665

caused by excessive Na if the texture of the soil, compaction, or waterrestricting layers is not the cause of low permeability. Gypsum Due to its solubility, low cost, and availability, gypsum is the most commonly used amendment for reclaiming Na-affected soil and reducing the harmful effects of high-Na irrigation waters. Gypsum comes from both mines and as a byproduct of the phosphate fertilizer industry. Under the same conditions, the rate of dissolution of industrial gypsum is much higher than that of mined gypsum (Fig. 21-2). Gypsum added to a sodic soil increases permeability by increasing electrolyte concentration and by reducing exchangeable Na (Keren and Shainberg 1981; Loveday 1976). Shainberg et al. (1982) evaluated the relative significance of the electrolyte effect and the cation exchange on the HC by comparing the effects of gypsum and CaCl2 in chemically equivalent

FIGURE 21-2. Rate of dissolution of industrial and mined gypsum particles of two fragment sizes. Data from Keren and Shainberg (1981).

666

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

amounts. Infiltration of distilled water into a noncalcerous soil initially treated with CaCl2 decreased and eventually stopped. The gypsum treatment maintained high HC. They observed no difference between the two amendments in a calcareous soil. Since both amendments had similar Na replacement, they attributed the differences to an electrolyte effect. Thus, the release of electrolytes by gypsum particles must be sustained to maintain the high HC of soils irrigated by waters with a low electrolyte concentration. The relative significance of the two effects should be noted. If the electrolyte effect is sufficient to prevent soil clays from dispersing and swelling, the surface application of gypsum may be worthwhile. In this case, the amount of gypsum required depends on the amount of low-salt water applied and the rate of gypsum dissolution. In soils where the electrolyte concentration effect is less significant and the main effect results from cation exchange, the amount of gypsum required depends on the amount of exchangeable Na in the depth of soil. The amount of exchangeable Na to be replaced during reclamation depends on the initial exchangeable Na fraction (ENai), the soil cation exchange capacity (CEC, mmol/Mg), soil bulk density (Mg m3), the desired final exchangeable sodium fraction (ENaf), and the depth of soil to be reclaimed (Dr, m). After the parameters are determined, the amount of exchangeable Na to be replaced per unit of land area (Na, molc ha1) can be calculated from Na  104 (Dr) (b) (CEC))ENai  ENaf

(21-2)

The value of ENaf depends on the response of the physical conditions of the soil and the Na tolerance of the crop. The gypsum requirement (GR, the amount of gypsum needed to reclaim a sodium-affected soil, (metric ton ha1) can be calculated from GR  86.1  106 Na

(21-3)

As the soil depth increases, the concentration of Ca and Na in the soil solution decreases and increases, respectively, as Ca-Na exchange occurs. Eventually, the SAR of the soil solution reaches a value in equilibrium with the existing ENa. The remaining soluble Ca will not react with exchangeable Na and is removed by drainage. To correct for this loss, Doering and Willis (1975) suggest including a correction factor in Eq. 21-3. The efficiency of applied Ca to remove adsorbed Na varies with ENa. It is much greater at high ESP values (Chaudhry and Warkentin 1968). The efficiency of Na exchange at ENa levels below about 10 is low (about 30%)

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

667

because a greater fraction of applied Ca displaces exchangeable Mg (Greene and Ford 1985; Loveday 1976). Efficiency may also be low (20% to 40%) in fine-textured soils due to the slow kinetics of Na-Ca exchange inside the structural elements (Manin et al. 1982). A factor seldom considered is that applying large amounts of gypsum may temporarily reduce HC, since small gypsum particles lodge in the soil pores (Keren et al. 1980). The benefit of gypsum for reclaiming sodic soils depends not only on the infiltration characteristics of the soil but also on the gypsum dissolution properties. Some of the factors that influence the dissolution rate are the surface area of gypsum fragments, soil-water velocity during leaching, and the electrolyte composition of the soil solution (Kemper et al. l975; Keren and O’Connor l982b). The time-dependent dissolution of gypsum is dC/dt  K (Cs  C)

(21-4)

where dC/dt is the net rate of dissolution, K is the dissolution coefficient, and Cs and C are the solution concentration at saturation and at a particular time, respectively. Equation 21-4 is a first-order kinetic model with the term (Cs  C) being the chemical concentration gradient. In accordance with Kemper et al. (1975), Keren and O’Connor (1982b) concluded that the rate of dissolution of gypsum is strongly influenced by both gypsum content and soil-water velocity. Increasing solution velocity increased the dissolution rate coefficient but decreased the contact time between gypsum and the flowing solution; the net effect was a decreasing dissolution rate with increasing soil-water velocity. Integrating Eq. 21-4 yields ⎛ C⎞ ln ⎜1  ⎟  kt C ⎝ s ⎠

(21-5)

and a plot of ln (l  C/Cs) versus t (time) should yield a straight line. Since the thickness of the film around the gypsum fragments is changing with the soil-water velocity, the dissolution rate coefficient for a given surface of gypsum fragment is also changing. Thus, the lines obtained from Eq. 21-5 are not linear under such conditions. Keren and O’Connor (1982b) reported that the left-hand side of Eq. 21-5 is empirically related linearly to the square root of time (tc): ⎛ C⎞ ln ⎜1  ⎟  t 1/2  c Cs ⎠ ⎝

(21-6)

668

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

where and  are the slope and intercept, respectively, and tc is the time for an increment of solution to leave a given soil depth: tc 

L V

(21-7)

where L is the length of the layer of soil-gypsum mixture and V is soilwater velocity (V  v/, where v is the soil water flux and  is the soil porosity). Combining Eqs. 21-6 and 21-5 gives k

1 ( t 1/2  ) c tc

(21-8)

Introducing Eq. 21-7 in 21-8 yields ⎛V⎞ k ⎜ ⎟ ⎝L⎠

1/2

1/2 ⎤ ⎡ ⎛V⎞ ⎢  ⎜ ⎟ ⎥ ⎝L⎠ ⎢⎣ ⎥⎦

(21-9)

Increasing velocity increases the dissolution rate coefficient but decreases the time during which gypsum and the flowing solution are in contact. The net result is a decreasing dissolution rate with increasing velocity. Thus, the chemical equilibrium in reclamation models probably can be assumed only when the velocity of the soil-water is lower than the rate of diffusion or when sufficient surface area of gypsum is available to reach equilibrium during leaching. Surface area of gypsum fragments has also been invoked to account for differences in infiltration rates of soils resulting from mined and phosphogypsum (byproduct of the phosphorus industry; Keren and Shainberg l98l). The latter, with a bulk density half of the former, dissolved 10 times faster (Fig. 21-2). Chemical equilibrium for gypsum dissolution may not be reached for soils with comparatively large HC, soils with low gypsum content, or soils with large particles of gypsum (Keren and Shainberg 1981) or particles of gypsum coated by CaCO3 (Keren and Kauschansky 1981). Thus, lowering the rate of water application whenever possible, such as with sprinkler irrigation, is preferable. It increases the rate of gypsum dissolution and enhances the efficiency of exchange (Fig. 21-3). Generally, soil-water penetrates too slowly to reclaim sodium-affected soils in a single leaching. For example, a 50-cm depth of water applied for leaching can only dissolve about 12 Mg/ha of gypsum. Larger gypsum treatments should not be made unless sufficient soil-water penetrates to allow larger applications of water. Thus, Na-affected soil normally can

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

669

FIGURE 21-3. Distribution of exchangeable Na fraction with depth as a function of soil water velocity for a given amount of dissolved calcium. Data from Keren and O’Connor (1982).

only be reclaimed to a limited depth in the first year. This will often permit a shallow-rooted crop to be established after leaching. Subsequently, amendments and leaching volumes can be applied annually to reclaim the entire profile. In estimating the amount of gypsum to be applied, little attention has been directed to the possibility that incorporating a large amount of gypsum powder into the soil may cause a temporary reduction in HC. The influence of gypsum fragments and content on the HC of soils before the gypsum has dissolved was studied for loamy soil and sandy loam soil (Keren et al. l980). Because of the limited solubility of gypsum, excess small particles of gypsum may block the conducting pores of the soil and reduce its HC in the early stage of the dissolution process. It was concluded that the effect of small particles of gypsum in diminishing the HC of soils should be considered when recommendations are made based on the sodicity of the soil, using large amounts of gypsum as an amendment. Acids and sulfur Sulfuric acid is an amendment that treats Na-affected soils. In some areas, acid waste products from mining and industrial activities are available. Using these waste products as amendments may provide a safe way

670

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

to dispose of them (Miyamoto et al. 1975b). The acid reacts with soil calcium carbonate to form gypsum (using H2SO4) or calcium chloride (using HCl). Sulfur requires an initial phase of microbiological oxidation to produce H2SO4. In soil column studies, Yahia et al. (1975) and Prather et al. (1978) found that H2SO4 increases the penetration of water into calcareous Na-affected soils more effectively than does gypsum. Field experiments on yield response have generally shown results in favor of H2SO4 (Overstreet et al. 1951). Equivalent amounts of gypsum reduced soluble and exchangeable Na in the surface soil, similarly to H2SO4, but produced smaller yield responses. The concentrated acid, applied directly on the surface of the soil, results in better distribution, less destruction of soil aggregates, and more efficient leaching of salts (Miyamoto et al. 1975a). Another technique used commercially (Tisdale 1970) is direct application by chiseling into the soil in bands about 45 cm apart. H2SO4 and SO2, being highly corrosive, should not be added to water that will pass through metal or concrete irrigation systems. The amount of acid required can be calculated from equations using CEC and concentrations of Na, Ca, and Mg in the saturation extract made with the leaching water (Miyamoto et al. 1975a). The major problem in using H2SO4 is the hazard associated with handling and application. Acidulants that first must be oxidized, such as sulfur, pyrites, and polysulfides, act more slowly than H2SO4. Their effectiveness in field experiments has varied. It may relate to the presence or absence of microbial populations. For elemental S application, dust poses a problem. This can be overcome by using conventional fluid fertilizer equipment to apply water suspensions containing 55% to 60% S (Thorup 1972) or granulated S. Acids applied at high rates lower the soil’s pH in a portion of the rootzone and consequently increase the availability of P, Zn, Mn, and Fe. This is important in obtaining improved crop responses (Miyamoto et al. 1975b). Calcium chloride Although CaCl2 2H2O generally is too expensive to compete with other amendments, it is available as an industrial waste product and could, therefore, be considered for reclamation. Similarly to GR (Eq. 21-3), the CaCl2 2H2O requirement, CCR, (metric ton ha1) can be calculated from CCR  75.5  106 Na

(21-10)

Being highly soluble, calcium chloride initially yields high electrolyte levels and high rates of water intake. This makes it a more efficient amendment than gypsum for soils with high ENa (Alperovitch and Shain-

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

671

berg 1973; Prather et al. 1978). Since CaCl2 is highly soluble, it is quickly leached from the soil profile. Despite the lower ENa value after treatment, noncalcareous soils may seal when the electrolyte concentration of the percolating solution falls below the FV (Shainberg et al. 1982). Therefore, a combination of CaCl2 and gypsum may more quickly and effectively reclaim soil with high ENa than gypsum alone (Prather et al. 1978).

Reclamation without Adding Soil Amendments Calcareous soils CaCO3 particles isolated from soils have ion activity product (IAP) values expected for calcite (Suarez and Rhoades 1982), but a calcite supersaturation is usually found in soils (Levy 1981). The supersaturation appears to be due to the presence of Ca-silicates in the soil more soluble than calcite and is the result of unstable CaCO3 phases (Suarez and Rhoades 1982). The rate at which CaCO3 dissolution in water approaches equilibrium depends on a number of factors, including the surface area solution volume ratio, the ionic composition of the solution, the ion composition of the adsorbed phase, the affinity of the clay minerals to cations, the temperature, and the local partial pressure of CO2. Amrhein et al. (1985) concluded that the kinetics of CaCO3 dissolution in soils is not a simple diffusion-controlled or first-order reaction. Soil CaCO3 may be dissolved to contribute Ca, especially in reclaiming saline and sodium-affected soils in which its solubility is enhanced (Oster 1982). Calcite supersaturation may occur in alkaline soils due to decomposition of organic matter coupled with inhibition of calcite precipitation due to poisoning of the calcite crystal surfaces by dissolved organic carbon (Amrheim and Suarez 1987; Lebron and Suarez 1996). The only hazard associated with this management practice is that the electrical conductance (EC) must be sufficiently high and pH sufficiently low during reclamation to maintain soil structural stability until the SAR decreases to a safe level. Suarez (2001) predicted [using the model UNSATCHEM of Sˇimu˚nek and Suarez (1996, 1997)] that the HC of a sodic soil (very fine sandy loam, hyperthermic Typic Torrifluvent, Riverside, California) was reduced significantly during reclamation in the absence of gypsum. This reduction was due to the initial decrease in EC while the SAR was still relatively high. However, the dissolution of CaCO3 may maintain the concentration levels of the soil solution above the FV of the soil clays and prevent HC reduction in the profile of soils at low ENa levels (Alperovitch et al. 1981). The extent to which reclamation can be achieved in calcareous soils by dissolution of calcium carbonate in the profile was simulated by Suarez

672

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(2001) and appeared feasible. The ideal practice would be to incorporate fresh organic matter into the soil when the soil is warm, thereby having high CO2 production as a result of decomposition. If the soil is kept at or near saturation, CO2 diffusion out of the soil is greatly restricted, CO2 concentrations are elevated, and substantial calcite can be dissolved and pH maintained within desirable levels. Successive dilutions of high-salt waters containing divalent cations The high-salt dilution method is particularly effective for soils with expanding clay-type minerals that have an extremely low HC at low electrolyte concentration in the soil solution and need an excessive amount of reclamation time or amendment. It involves successively diluting a high-salt water containing divalent cations (Reeve and Bower 1960). It tested successfully in experiments. A monogram was developed to predict the reclamation at various stages of leaching (Reeve and Doering 1966). In the early phase, high salinity of the water flocculates soil colloids and provides a source of Ca for exchange with Na. When diluted with water with a low concentration of salt, the SAR of the irrigation water is reduced by the square root of the dilution factor. Equations were derived for calculating the depth of water need to reclaim a sodic soil by leaching with such water. Misopolinos (1985) theoretically analyzed the use of high-salt water to add a constant quantity of Ca2 and thus reclaim sodic soils. Jury et al. (1979) demonstrated that leaching reclamation without amendments can take place. Good drainage through the soil profile, adequate leaching water, and a soil source of Ca that can be mobilized are all needed. Saline sodic soils frequently contain precipitated CaCO3 and gypsum that, upon leaching, dissolve to provide sufficient Ca to exchange with adsorbed Na. Intermittent ponding requires less water than continuous ponding to achieve the same degree of leaching (Miller et al. 1965). Sprinkle irrigation is more efficient than other methods at removing salt from small pores in the soil profile (Nielsen et al. 1966). This method has also reclaimed a slowly permeable, Na-affected soil in a humid environment, where HC and IR increased from 30% to more than 100% (Rahman et al. 1974). Soil profile modification Deep tillage may benefit stratified, Na-affected soils with impermeable layers between permeable layers, which are often found along the Snake River and its tributaries in the northwestern United States, in much of Arizona’s Salt River Valley, in parts of the Rio Grande Valley in New Mexico and Texas, and in California’s San Joaquin Valley. In loca-

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

673

tions where soil with significant quantities of gypsum or CaCO3 lies beneath Na-affected soils or subsoils, deep plowing has effectively broken up and mixed the layers while supplying soluble Ca to aid reclamation. The depth of plowing required varies from 0.5 m to more than 1.0 m, depending on the concentration and depth of the sodic- and Ca-rich layers. A procedure to predict the optimum depth of plowing to maintain adequate permeability during reclamation is available (Rasmussen and McNeal 1973). Soil Reclamation Models Two currently available types of models that describe the reclamation of sodic soils are chromatographic (plate) models and miscible displacement models. The former are generally simpler to use. The latter demand extensive input data. For both types, spatial variability needs to be taken into account when applied in the field. Chromatographic models have been used to describe sodic soil reclamation with gypsum (Dutt et al. 1972; Tanji et al. 1972). The models, which include such reactions as the dissolution of CaCO3 and gypsum and cation exchange, are based on the assumption that these reactions reach equilibrium in each soil segment of plates. When these models were field-tested, differences between predicted and measured values were found to be less than horizontal variations typically found in sodic soils. More recently, Tanji and Deverel (1984) expanded this approach to include the kinetic rate of several chemical reactions, including sulfur oxidation. Though they can apparently predict the chemistry of reclamation, chromatographic models are less able to predict soil-water status or the rate of reclamation. Determining the appropriate number of soil segments for simulation also poses a problem. The solution of solute-transport equations about convective transfer and hydrodynamic dispersion is the core of miscible displacement models (see Chapter 23). It rigorously describes solute-transfer processes. Miscible displacement models, when coupled with chemical reaction models (Robbins et al. 1980), provide insight into the processes of reclamation. However, they have not been used extensively in the field, probably because solving them requires extensive data specific to the site. Also, existing models neither fully account for solute effects on HC nor consider spatial variability, both of which are among the most important aspects of sodic soil reclamation. Determinations of gypsum requirements can be based on the quantitative calculation of exchange efficiency, calcite dissolution, and the Ca contribution of the irrigation water using the numerical model UNSATCHEM (Sˇ imu˚nek and Suarez 1996; Suarez and Sˇ imu˚nek 1997). The model pre-

674

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

dictions of EC and SAR after reclamation gave a satisfactory fit to the measured values of a field study (Suarez 2001). This model considers the effect of EC, SAR, and pH on the HC of soil. Preventing Crust Formation Gradual deterioration of soil structure and the formation of a surface crust can decrease an initially high rate of IR. The IR is also sensitive to the electrolyte concentration in the applied solution (Agassi et al. 1981), which is important in controlling infiltration of Na-affected soils. Applying gypsum reduced surface runoff (Table 21-1). Although the gypsum minimally affected the ENa of the surface soil, it significantly affected the IR (Keren et al. 1983). The IR of a Na-affected soil exposed to rain depends on the source, amount, and size of the gypsum fragments (Fig. 21-4). The efficiency of gypsum in maintaining a high IR is a function of its rate of dissolution. Industrial gypsum, because of its finely sized particles, is more effective than mined gypsum in maintaining a high IR (Keren and Shainberg 1981). The dissolution rate of gypsum is important

TABLE 21-1. Effect of Industrial Gypsum Treatments on Surface Runoff in Rainstorms from Loess Soil at Two Levels of ENa Surface Runoff (Percent of Rainfall) ENa 0.046

Storm No. (1)

Amount of Rainfall (mm) (2)

Time Between Storms (3)

1

16

2

20

40

3

19

4

59

5 6 7

ENa 0.193

Gypsuma Spread Control Over Control (4) (5) (6)

Gypsuma Gypsuma Mixed Spread In Over (7) (8)

0

0

23.1

2.6

0

6.5

0

13.3

3.0

0.7

20

1.3

0.5

27.3

10.5

3.8

6

21.5

3.8

41.0

21.2

17.2

44

9

13.4

2.1

45.0

22.5

12.4

12

1

12.5

2.5

40.0

25.0

12.5

12

15

10.1

2.5

31.7

13.8

6.7

22.8

3.8

64.2

30.1

18.7

Percent of annual rainfall (182 mm) 12.5

2.1

35.3

16.5

10.3

Total rainfall (mm) a

The rates of industrial gypsum application were 5 Mg/ha and 10 Mg/ha for the soils having ENa values of 0.046 and 0.193, respectively.

ENa, exchangeable sodium fraction. Data from Keren et al. (1983).

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

675

FIGURE 21-4. Effect of industrial and mined gypsum on infiltration rate of loess soil as a function of cumulative rainfall. Data from Keren and Shainberg (1981). because rainwater and gypsum particles at the surface of the soil are in contact for a short time. Surface spreading of gypsum was more beneficial than mixing it in the upper layer because the former concentrates the amendment in the zone of crusting (Agassi et al. 1982; Keren et al. 1983). The amount of gypsum needed depends on the amount of low-salt-concentration water applied (or rain) and the rate of gypsum dissolution. Typical application rates in semiarid regions range from 5 Mg ha1 to 10 Mg ha1.

676

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

RECLAMATION OF BORON-AFFECTED SOILS Boron Hazard Boron is one of the essential micronutrients required for the growth of plants. However, a relatively narrow margin exists between levels of B that cause deficiency and levels that cause toxicity. Boron toxicity occurs most frequently in soils of arid and semiarid regions, where soils are irrigated with water high in B, while B deficiency is found primarily in humid regions. Concepts and Principles Boron can be specifically adsorbed by different clay minerals, hydroxy oxides of Al, Fe, and Mg, and organic matter. Keren and Bingham (1985) review the factors that affect the adsorption and desorption of B by soil constituents and the mechanisms of adsorption. The Langmuir equation describes the adsorption of B by soils (Hatcher and Bower 1958) and clays (Hingston 1964). Deviations occur at higher solution concentrations at high pH (Hingston 1964). Hingston (1964) related the variation in B adsorption with pH to changes in the surface of the clays, which induced more adsorption sites to form at high pH but decreased adsorption due to the decrease of the concentration of boric acid relative to borate ions (Hingston 1964). This adsorption model has two disadvantages: It does not consider that two B species are involved and that their relative presence in solution varies with pH; and different values of the Langmuir coefficients, namely, maximum adsorption and binding energy constant, are needed to predict adsorption at any given pH. Another approach for describing B adsorption by clays, hydroxy-Al, and soils has been derived by Keren et al. (1981). It is assumed that B(OH)3, B(OH)4 , and OH–S compete for the same adsorption sites. Calculated and experimental results correlated well (Keren and Gast 1983; Keren and Mezuman 1981; Keren and O’Connor 1982a; Mezuman and Keren 1981). Boron concentrations in the soil solution after drying or wetting tend to remain constant due to buffering by sorption and desorption reactions (Fig. 21-5). This buffering capacity depends on the soil affinity for B and the maximum B adsorption value. Thus, plants grown in the coarse-textured soil had the greatest uptake of B and plants grown in the fine-textured soil had the least (Keren et al. 1984, 1985). Reclamation Methods and Models Before high-B soils can be successfully used for agriculture, their soluble B contents must be reduced to nonphytotoxic levels. Such reclamation

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

677

FIGURE 21-5. Boron concentration in soil solution as a function of solution/soil ratio for a given total amount of boron. (A) No interaction between boron and soil; (B) adsorption accounted for. Data from Mezuman and Keren (1981).

is typically accomplished by leaching the soil. The relative decrease of soluble B in field soils during reclamation is described as ⎛ C ⎞⎛ D ⎞ ⎟ ⎜ ⎟  0.6 ⎜ ⎝ C0b ⎠ ⎝ Ds ⎠

(21-11)

where C and C0b are the final and initial soluble B concentrations and D/Ds is the depth of leaching water per unit depth of soil (Hoffman 1980). Equation 21-11 is independent of the method of water application—be it sprinkling or ponding. Native soil B appears to be more difficult to leach than B accumulated from previous irrigations.

678

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Hatcher and Bower (1958), Tanji (1970), Suarez (2001), and Communar and Keren (2005, 2006, 2007) have used mathematical models to predict the movement of B in soil profiles, considering adsorption isotherms and a chromatographic displacement theory. They found that theoretical and experimental results correlated well. The computer models predict profile distribution of in-situ soil water contents. These models may be used to estimate the amount of water required to reclaim B-affected soils only when B is associated with the soil adsorption sites. When irrigation cycles consist of water infiltration, evaporation, and internal drainage, B transport in unsaturated soils occurs under nonequilibrium conditions (Communar and Keren 2007). The model developed by Communar and Keren (2006, 2007) was based on the convection-dispersion equation with a rate-limited reaction term coupled with the Richards’ equation and a B adsorption equation. This model successfully simulated B transport in unsaturated soils under irrigation cycles consisting of water infiltration, internal drainage, and evaporation. The numerical model UNSATCHEM mentioned (Suarez 2001) also has routines suitable for B adsorption-desorption (using the constant capacitance model of Goldberg et al. 2000). This model is also capable to simulate B transport in soils. A functional model TETrans (Trace Element Transport) was tested by Corwin et al. (1999) for evaluating B transport in soil. They concluded that this model was capable of simulating the movement of B through the rootzone with reasonable accuracy. However, the pH-dependent Keren equation (Keren et al. 1981) was the best-performing chemical adsorption model when several other adsorption models were compared (Corwin et al. 1999). All of the models discussed may apply to soils without native high-B minerals (tourmaline, for example). The reduced B concentrations that follow the leaching of soils with native high-B minerals may be temporary. Longer storage times of water in the soil profile resulted in larger increases in effluent B concentrations (Rhoades et al. 1970; Bingham et al. 1972). Peryea et al. (1985) inversely related this phenomenon, termed B regeneration, to the amount of water used for the initial leaching. Boron regeneration is therefore of primary concern during the early stages of soil reclamation, when appreciable residual sources of regenerable B are present. The potential for reestablishing phytotoxic concentrations in a soil is keyed to the relative completeness of the reclamation processes and the rate of B dissolution after reclamation.

SUMMARY A successful irrigated agriculture requires permanent control of salinity, sodicity, and B levels in soils and irrigation water. The reclamation of

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

679

sodic soils ameliorates adverse conditions such as surface crusting, increases the HC, which affects the infiltration and storage of water, the drainage, the emergence of seedlings, and the development of roots. There are a number of methods for leaching salts from soils, the selection of which depends on soil type, the availability of subsurface drains, and the variability of salt distribution in the area to be leached. The longterm success of leaching depends on a means of conveying drainage water to the ocean or another suitable sink. The reclamation of Na-affected soils may be aided by management of irrigation to maintain HC, but also generally requires replacement of the Na with Ca2. Amendments that produce Ca in calcareous soils by enhancing the conversion of soil CaCO3 to gypsum include sulfuric acid, sulfur, and iron and aluminium sulfates. Gypsum added to a sodic soil increases permeability by increasing electrolyte concentration and by reducing exchangeable Na. To be effective, the release of electrolytes from the gypsum must be maintained. Fine-grained particles of gypsum are preferred because they dissolve more rapidly and thus help maintain infiltration rate in the upper soil layers. Acids (such as H2SO4) are also effective in reclamation of sodic soils, with some advantages over gypsum but also including some handling hazards. Acidulants that first must be oxidized, such as sulfur, pyrites, and polysulfides, act more slowly than H2SO4. Calcium chloride (CaCl2 2H2O) is generally is too expensive to compete with other amendments, but as an industrial waste product it may be considered for reclamation. The needed Ca2 may also be provided in calcareous soils by adding organic matter under warm conditions to enhance the solubility of CaCO3. An alternative is leaching with successive dilutions of high-salt waters containing divalent cations, provided there is good drainage through the soil profile, adequate leaching water, and a soil source of Ca that can be mobilized. Boron is a particular concern because its phytotoxicity concentration is only marginally higher than the level needed for plant growth. Before high-B soils can be successfully used for agriculture, their soluble B contents must be reduced to nonphytotoxic levels. Such reclamation is typically accomplished by leaching the soil. Leaching is affected by the presence of B in the soil, as well as in the irrigation water, and in soils with residual sources of regenerable B. The potential for reestablishing phytotoxic B concentrations in a soil is keyed to the relative completeness of the reclamation processes and the rate of B dissolution after reclamation.

REFERENCES Agassi, M., Morin, J., and Shainberg, I. (1982). “Infiltration and runoff control in the semi-arid region of Israel.” Geoderma, 28, 345–356.

680

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Agassi, M., Shainberg, I., and Morin, I. (1981). “Effect of electrolyte concentration and soil sodicity on the infiltration rate and crust formation.” Soil Sci. Soc. Am. J., 45, 848–851. Alperovitch, N., and Shainberg, I. (1973). “Reclamation of alkali soils with CaCl2 solutions,” in Physical aspects of soil, water and salts in ecosystems, A. Hadas et al., eds., Springer-Verlag, Berlin, 431–440. Alperovitch, N., Shainberg, I., and Keren, R. (1981). “Specific effect of magnesium and hydraulic conductivity of sodic soils.” Clays Clay Min., 33, 443–450. Amrhein, C. Jurinak, J. J., and Moore, W. M. (1985). “Kinetics of calcite dissolution as affected by carbon dioxide partial pressure.” Soil Sci. Soc. Am. J., 49, 1393–1398. Amrhein, C., and Suarez, D. L. (1987). “Calcite supersaturation as a result of organic matter mineralization.” Soil Sci. Soc. Am. J., 51, 932–937. Bingham, F. T., Marsh, A. W., Branson, R., Mahler, R., and Ferry, G. (1972). “Reclamation of salt-affected high boron soils in western Kern County.” Hilgardia, 41, 195–211. Carter, D. L., and Fanning, C. D. (1964). “Combining surface mulches and periodic water applications for reclaiming saline soils.” Soil Sci. Soc. Am. Proc., 28, 564–567. Chaudhry, G. H., and Warkentin, B. P. (1968). “Studies on exchange of sodium from soils by leaching with calcium sulfate.” Soil Sci., 105, 190–197. Communar, G., and Keren, R. (2005). “Equilibrium and nonequilibrium transport of boron in soil.” Soil Sci. Soc. Am. J., 69, 311–317. ———. (2006). “Rate-limited boron transport in soils: The effect of soil texture and solution pH.” Soil Sci. Soc. Am. J., 70, 882–892. ———. (2007). “Effect of transient irrigation on boron transport in soils.” Soil Sci. Soc. Am. J., 71, 306–313. Corwin, D. L., Goldberg, S., and David, A. (1999). “Evaluation of a functional model for simulating boron transport in soil.” Soil Sci., 164, 697–717. Doering, E. J., and Willis, W. O. (1975). Chemical reclamation for sodic strip-mine spoils, USDA/Agricultural Research Service, ARS-NC-20, U.S. Department of Agriculture, Washington, D.C. Dutt, G. R., Terkeltoub, R. W., and Rauschkolb, R. S. (1972). “Prediction of gypsum and leaching requirements for sodium-affected soils.” Soil Sci., 114, 93–103. Frenkel, H., Goertzen, J. O., and Rhoades, J. D. (1978). “Effect of clay type and content, exchangeable sodium percentage, and electrolyte concentration on clay dispersion and soil hydraulic conductivity.” Soil Sci. Soc. Am. J., 42, 32–39. Goldberg, S., and Glaubig, R. A. (1987). “Effect of saturating cation, pH and aluminum and iron oxide on the flocculation of kaolinite and montmorillonite.” Clays and Clay Min., 35, 220–227. Goldberg, S., Lesch, S. M., and Suarez, D. L. (2000). “Predicting boron adsorption by soils using chemical parameters in the constant capacitance model.” Soil Sci. Soc. Am. J., 64, 1356–1363. Greene, R. S. B., and Ford, G. W. (1985). “The effect of gypsum on cation exchange and leaching in two red duplex wheat soils.” Aust. J. Soil Res., 23, 61–74. Hatcher, J. T., and Bower, C. A. (1958). “Equilibria and dynamics of boron adsorption by soils.” Soil Sci., 85, 319–328.

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

681

Heller, H., and Keren, R. (2001). “Rheology of Na-montmorillonite suspension as affected by electrolyte concentration and shear rate.” Clays Clay Min., 49, 286–291. Hingston, F. J. (1964). “Reaction between boron and clays.” Aust. J. Soil Res., 2, 83–95. Hoffman, G. J. (1980). “Guidelines for the reclamation of salt-affected soils,” in G. A. O’Connor, ed., Proc. 2nd Inter-American Conf. on Salinity and Water Management Technology, Juarez, Mexico, December 11–12, 1980, 49–64. Hulsbos, W. C., and Boumans, J. H. (1960). “Leaching of saline soils in Iraq.” Netherlands J. Agr. Sci., 8, 1–10. Jury, W. A., Jarrell, W. M., and Devitt, D. (1979). “Reclamation of saline-sodic soils by leaching.” Soil Sci. Soc. Am. J., 43, 1100–1106. Errata (1987), Soil Sci. Soc. Am. J., 51, 1092. Kemper, W. D., Olsen, J., and DeMooy, C. (1975). “Dissolution rate of gypsum in flowing water.” Soil Sci. Soc. Am. J., 39, 458–463. Keren, R., and Bingham, F. T. (1985). “Boron in water, soils, and plants,” in Advances in soil science, Vol. 1, B. A. Stewart, ed., Springer-Verlag, New York, 229–276. Keren, R., Bingham, F. T., and Rhoades, J. D. (1984). “Plant uptake of boron as affected by boron distribution between liquid and solid phases in soils.” Soil Sci. Soc. Am. J., 48, 297–302. ———. (1985). “Effect of clay content in soil on boron uptake and yield of wheat.” Soil Sci. Soc. Am. J., 49, 1466–1470. Keren, R., and Gast, R. G. (1983). “pH-dependent boron adsorption by montmorillonite hydroxy-aluminum complexes.” Soil Sci. Soc. Am. J., 47, 1116–1121. Keren, R., Gast, R. G., and Bar-Yosef, B. (1981). “pH-dependent boron adsorption by Na-montmorillonite.” Soil Sci. Soc. Am. J., 45, 45–48. Keren, R., and Kauschansky, P. (1981). “Coating of calcium carbonate on gypsum particle surfaces.” Soil Sci. Soc. Am. J., 45, 1242–1244. Keren, R., Kreite, J. F., and Shainberg, I. (1980). “Influence of size of gypsum particles on the hydraulic conductivity of soils.” Soil Sci., 130, 113–117. Keren, R., and Mezuman, U. (1981). “Boron adsorption by clay minerals using a phenomenological equation.” Clays Clay Min., 29, 198–204. Keren, R., and O’Connor, G. A. (1982a). “Effect of exchangeable ions and ionic strength on boron adsorption by montmorillonite and illite.” Clays Clay Min., 30, 341–346. ———. (1982b). “Gypsum dissolution and sodic soil reclamation as affected by water flow velocity.” Soil Sci. Soc. Am. J., 46, 726–732. Keren, R., and Shainberg, I. (1981). “The efficiency of industrial and mined gypsum in reclamation of a sodic soil-rate of dissolution.” Soil Sci. Soc. Am. J., 45, 103–107. Keren, R., Shainberg, I., Frenkel, H., and Kalo, Y. (1983). “The effect of exchangeable sodium and gypsum on surface runoff from loess soil.” Soil Sci. Soc. Am. J., 47, 1001–1004. Keren, R., Shainberg, I., and Klein, E. (1988). “Settling and flocculation value of sodium-montmorillonite particles in aqueous media.” Soil Sci. Soc. Am. J., 52, 76–80. Keren, R., and Singer, M. J. (1988). “Effect of low electrolyte concentration on hydraulic conductivity of Na/Ca-montmorillonite sand systems.” Soil Sci. Soc. Am. J., 52, 368–373.

682

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1989). “Effect of low electrolyte concentration on hydraulic conductivity of clay-sand hydroxy polymers.” Soil Sci. Soc. Am. J., 53, 349–355. ———. (1991). “Hydroxy-aluminum’s effect on permeability of clay-sand mixtures.” Soil Sci. Soc. Am. J., 55, 61–65. Lebron, I., and Suarez, D. L. (1996). “Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25 °C and pH 7.5.” Geochim. Cosmochim. Acta, 60, 2765–2776. Levy, R. (1981). “Effect of dissolution of alumosilicates and carbonates on ionic activity products of calcium carbonate in soil extracts.” Soil Sci. Soc. Am. J., 45, 250–255. Longenecker, D. E., and Lyerly, P. J. (1974). Control of soluble salts in farming and gardening, Texas Agricultural Experiment Station Bulletin 876, Texas A&M University, College Station, Tex. Loveday, J. (1976). “Relative significance of electrolyte and cation exchange effects when gypsum is applied to a sodic clay soil.” Aust. J. Soil Res., 14, 361–371. Luthin, J. N., Fernandez, P., Woerner, J., and Robinson, F. (1969). “Displacement front under ponded leaching.” J. Irrig. Drain. Div. ASCE, 95(IR1), 117–125. Manin, M., Pissarra, A., and Van Hoorn, J. W. (1982). “Drainage and desalinization of heavy clay soil in Portugal.” Agric. Water Mgmt., 5, 227–240. McNeal, B. L., and Coleman, N. T. (1966). “Effect of solution composition on soil hydraulic conductivity.” Soil Sci. Soc. Am. Proc., 30, 308–312. McNeal, B. L., Layfield, D. A., Norvell, W. A., and Rhoades, J. D. (1968). “Factors influencing hydraulic conductivity of soils in the presence of mixed salt solutions.” Soil Sci. Soc. Am. J., 32, 187–190. Mezuman, U., and Keren, R. (1981). “Boron adsorption by soils using a phenomenological adsorption equation.” Soil Sci. Soc. Am. J., 45, 722–726. Miller, R. J., Biggar, J. W., and Nielsen, D. R. (1965). “Chloride displacement in Panoche clay loam in relation to water movement and distribution.” Water Resour. Res., 1, 63–73. Minhas, P. S., and Khosla, B. K. (1986). “Solute displacement in a silt loam soil as affected by dry method of water application under different evaporation rates.” Agr. Water Mgmt., 12, 64–74. Misopolinos, N. D. (1985). “A new concept for reclaiming sodic soils with highsalt water.” Soil Sci., 140, 69–74. Miyamoto, S. (1988) “Reclamation of salt-affected orchards,” in Proc. Texas Pecan Conf., Texas A&M University, College Station, Tex., 25–37. Miyamoto, S., and Cruz, I. (1986). “Spatial variability and soil sampling for salinity and sodicity appraisal in surface-irrigated orchards.” Soil Sci. Soc. Am. J., 50, 1020–1025. ———. (1987). “Spatial variability of soil salinity in furrow-irrigated torrifluvents.” Soil Sci. Soc. Am. J., 51, 1019–1025. Miyamoto, S., Prather, R. J., and Stroehlein, J. L. (1975a). “Sulfuric acid and leaching requirements for reclaiming sodium-affected calcareous soils.” Plant Soil, 43, 573–585. Miyamoto, S., Ryan, J., and Stroehlein, J. L. (1975b). “Potentially beneficial uses of sulfuric acid in south-western agriculture.” J. Environ. Qual., 4, 431–437. Miyamoto, S., and Warrick, A. W. (1974). “Salt displacement into drain tiles under ponded leaching.” Water Resour. Res., 10, 275–278.

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

683

Morin, J., Keren, R., Benjamini, Y., Ben-Hur, M., and Shainberg, I. (1989). “Water infiltration as affected by soil crust and water moisture profile.” Soil Sci., 148, 53–59. Nielsen, D. R., Biggar, J. W., and Luthin, J. N. (1966). “Desalinization of soils under controlled unsaturated flow conditions,” in Proc. 6th Congr. Int. Comm. on Irrigation and Drainage, New Delhi, 19.15–19.24. Oster, J. D. (1982). “Gypsum usage in irrigated agriculture: A review.” Fertil. Res., 3, 73–89. Oster, J. D., and Schroer, F. W. (1979). “Infiltration as influenced by irrigation water quality.” Soil Sci. Soc. Am. J., 43, 444–447. Oster, J. D., Shainberg, I., and Wood, J. D. (1980). “Flocculation value and gel structure of Na/Ca montmorillonite and illite suspension.” Soil Sci. Soc. Am. J., 44, 955–959. Oster, J. D., Willardson, L.S., and Hoffman, G. J. (1972). “Sprinkling and ponding techniques for reclaiming saline soils.” Trans. ASAE, 15, 115–117. Overstreet, R., Martin, J. C., and King, H. M. (1951). “Gypsum, sulfur and sulfuric acid for reclaiming an alkali soil of the Fresno series.” Hilgardia, 21, 113–127. Peryea, F. J., Bingham, F. T., and Rhoades, J. D. (1985). “Regeneration of soluble boron by reclaiming high boron soils.” Soil Sci. Soc. Am. J., 49, 313–316. Prather, R. J., Goertzen, J. O., Rhoades, J. D., and Frenkel, H. (1978). “Efficient amendment use in sodic soil reclamation.” Soil Sci. Soc. Am. J., 42, 782–786. Quirk, J. P., and Schofield, R. K. (1955). “The effect of electrolyte concentration on soil permeability.” J. Soil Sci., 6, 163–178. Rahman, M. A., Hiler, E. A., and Runkles, J. R. (1974). “High electrolyte water for reclaiming slowly permeable soils.” Trans. ASAE, 17, 129–133. Rasmussen, W. W., and McNeal, B. L. (1973). “Predicting optimum depth of profile modification by deep plowing for improving saline-sodic soils.” Soil Sci. Soc. Am. Proc., 37, 432–437. Rasmussen, W. W., Moore, D. P., and Alban, L. A. (1972). “Improvement of a solonetizic (slick spot) soil by deep plowing, subsoiling, and amendments.” Soil Sci. Soc. Am. Proc., 36, 137–142. Reeve, R. C., Allison, L. E., and Peterson, D. F. (1948). Reclamation of saline-alkali soils by leaching, Utah Agricultural Experiment Station Bulletin 335, Utah State University, Logan, Utah. Reeve, R. C., and Bower, C. A. (1960). “Use of high-salt waters as a flocculant source of divalent cations for reclaiming sodic soils.” Soil Sci., 90, 139–144. Reeve, R. C., and Doering, E. J. (1966). “The high salt-water dilution method for reclaiming sodic soils.” Soil Sci. Soc. Am. Proc., 30, 498–504. Reeve, R. C., Pillsbury, A.F., and Wilcox, L. V. (1955). “Reclamation of a saline and high boron soil in the Coachella Valley of California.” Hilgardia, 24(4), 69–91. Rhoades, J. D., Ingualson, R. D., and Hatcher, J. T. (1970). “Laboratory determination of leachable soil boron.” Soil Sci. Soc. Am. J., 34, 938–941. Robbins, C. W., Wagenet, R. J., and Jurinak, J. J. (1980). “A combined salt transport-chemical equilibrium model for calcareous and gypsiferous soils.” Soil Sci. Soc. Am. J., 44, 1191–1200. Russo, D. (1984). “Spatial variability considerations in salinity management,” in Soil salinity under irrigation, I. Shainberg and J. Shalhevet, eds., Springer-Verlag, New York, 198–219.

684

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Russo, D., and Bresler, E. (1977). “Effect of mixed Na/Ca solutions on the hydraulic properties of unsaturated soils.” Soil Sci Soc. Am. J., 41, 713–717. Shainberg, I., Bresler, E., and Klausner, Y. (1971). “Studies on Na/Ca montmorillonite systems. I. The swelling pressure.” Soil Sci., 111, 214–219. Shainberg, I., Keren, R., and Frenkel, H. (1982). “Response of sodic soils to gypsum and calcium chloride application.” Soil Sci. Soc. Am. J., 46, 113–117. Sˇ imu˚nek, J., and Suarez, D. L. (1996) UNSATCHEM Code for simulating the onedimensional variably saturated water flow, heat transport, carbon dioxide production and transport, and multicomponent solute transport with major ion equilibrium and kinetic chemistry, Part A, U.S. Salinity Laboratory Research Report No. 128, U.S. Salinity Laboratory, Riverside, Calif. Stewart, E. H., Alberts, R. R., and Orth, P. G. (1976). “Water and salinity relationships in Perrine marl soils of south Florida.” Soil and Crop Soc. Sci. Florida, 36, 89–93. Suarez, D. L. (2001). “Sodic soil reclamation: Modeling and field study.” Aust. J. Soil Res., 39, 1225–1246. Suarez, D. L., and Rhoades, J. D. (1982). “The apparent solubility of calcium carbonate in soils.” Soil Sci. Soc. Am. J., 46, 716–722. Suarez, D. L, Sˇ imu˚nek, J. (1997). “UNSATCHEM: Unsaturated water and solute transport model with equilibrium and kinetic chemistry.” Soil Sci. Soc. Am. J., 61, 1633–1646. Talsma, T. (1967). “Leaching of tile-drained saline soils.” Aust. J. Soil Res., 5, 37–46. Tanji, K. K. (1970). “A computer analysis on the leaching of boron from stratified soil columns.” Soil Sci., 110, 44–51. Tanji, K. K., and Deverel, S. J. (1984). “Simulation modeling for reclamation of sodic soils,” in Soil salinity under irrigation, I. Shainberg and J. Shalhevet, eds., Springer-Verlag, New York, 238–251. Tanji, K. K., Doneen, L. D., Ferry, G. V., and Ayers, R. S. (1972). “Computer simulation analysis on reclamation of salt-affected soil in San Joaquin Valley, California.” Soil Sci. Soc. Am. J., 36, 127–133. Terkeltoub, R. W., and Babcock, K. L. (1971). “Calculation of the leaching required to reduce the salinity of a particular soil depth beneath a specific value.” Soil Sci. Soc. Am. Proc., 35, 411–413. Thorup, J. T. (1972). “Soil sulphur application.” Sulphur Inst. J., 8, 16–17. Tisdale, S. L. (1970). “The use of sulphur compounds in irrigated arid land agriculture.” Sulphur Inst. J., 6, 2–7. Van Olphen, H. (1977). An introduction to clay colloid chemistry, 2nd ed., John Wiley and Sons, New York. Yahia, T. A., Miyamoto, S., and Stroehlein, J. L. (1975). “Effect of surface applied sulfuric acid on water penetration into dry calcareous and sodic soils.” Soil Sci. Soc. Am. J., 39, 1201–1204.

NOTATION C  salt concentration in soil C  final soluble B concentrations Co  initial salt concentration in soil

RECLAMATION OF SALINE, SODIC, AND BORON-AFFECTED SOILS

C0b CCR CEC D D/Ds Ds Dr EC ENaf ENai FV GR k SAR

 initial soluble B concentrations  CaCl2 2 H2O requirement  soil cation exchange capacity  depth of leaching water  depth of leaching water per unit depth of soil  depth of soil to be leached  depth of soil to be reclaimed  electrical conductance  final exchangeable sodium fraction  initial exchangeable sodium fraction  flocculation value  gypsum requirement  empirical constant  sodium adsorption ratio

685

This page intentionally left blank

CHAPTER 22 USE OF SALINE DRAINAGE WATERS FOR IRRIGATION Stephen R. Grattan, James D. Oster, S. E Benes, and S. R. Kaffka

INTRODUCTION In many arid parts of the world such as Australia, the Middle East, the former Soviet Union, and the United States, large quantities of saline drainage water and other saline groundwater sources exist. These waters are often considered to have little value for irrigation since they contain large quantities of dissolved salts and sometimes other undesirable constituents. Depending on quality and availability, however, saline water is suitable for irrigation based on worldwide experiences with its use. The following are examples where water typically classified as unsuitable for irrigation is used successfully in different parts of the world. In the Pecos Valley of Texas, water averaging 2,500 mg/L total dissolved solids (TDS) has been used as irrigation water for decades (Moore and Hefner 1976). In Colorado’s Arkansas Valley, alfalfa, sorghum, and wheat have been grown with water containing 1,500 to 5,000 mg/L TDS (Miles 1977). In Iraq, pear trees have been grown successfully with water up to 4,000 mg/L TDS (Hardan 1976). In Israel, cotton has been grown commercially with 4.6 dS/m EC (electrical conductivity) water (about 3,000 mg/L TDS) for many years (Frenkel and Shainberg 1975; Keren and Shainberg 1978). In India, saline water with an EC 6.0 dS/m has been successfully used to irrigate various agronomic crops in a cyclic manner with nonsaline water (Minhas and Sharma 2003). And in the western San Joaquin Valley (SJV) of California, a number of grass forages have been successfully grown using saline-sodic drainage water as the sole irrigation source, including tall wheatgrass (Thinopyrum ponticum, var. ‘Jose’) and creeping wildrye (Leymus triticoides, var. ‘Rio’) irrigated with drainage water of 8 to 12 dS/m for more than 7 years and growing in soils of 19 and 13 dS/m ECe, 687

688

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

respectively (Suyama et al. 2007a). ECe is the average electrical conductivity (dS m1) of the saturated soil extract within the crop rootzone. At another location in the SJV, Bermuda grass (Cynodon dactylon, vars. ‘Common’ and ‘Giant’) is currently growing well under irrigation with salinesodic drainage water (2 to 8 dS/m EC) and supports beef cattle production in soils averaging 13 dS/m ECe to a depth of 30 cm (Kaffka et al. 2004). This chapter discusses several practices that use saline water and drainage water for irrigation, including both their benefits and limitations. Particular emphasis is directed toward the control of soil salination, adverse effects on soil physical properties when irrigation waters are sodic as well as saline, and the use of crops that are suitable for saline irrigation systems. Trace elements, such as boron (B), selenium (Se), and molybdenum (Mo) may also influence the feasibility of using salinesodic water for irrigation. We address these elements and discuss how these constituents may or may not affect the overall feasibility of using saline water for irrigation. This review places a large emphasis on studies conducted in the SJV, where trace element management is a significant challenge.

IRRIGATION WITH SALINE WATER Use of saline or saline-sodic water for irrigation requires several changes from standard management practices, including selection of appropriate crops and crop rotations, appropriate water and soil management, and in some cases, the adoption of advanced irrigation technology. Particular care needs to be directed toward achieving salinity control within the rootzone, avoiding deterioration of soil physical conditions, and minimizing the accumulation of certain trace elements (e.g., B, Se, Mo) that may be problematic to crop production or wildlife, should they be present. Management practices that optimize production depend on whether good-quality water is also available for irrigation at critical times during the season or for salt-sensitive crops in a rotation. Rainfall sufficient to leach salts from the upper portion of the rootzone and provide stored soil-water for the crop also facilitates the use of saline water for irrigation. In the absence of fresh-water augmentation, management opportunities using saline water are more limited, and crop rotations may be limited to planting crops that are more salt-tolerant. Irrigation Strategies with Saline Water Several methods for utilizing saline water have been tested experimentally or demonstrated under field conditions. The methods differ

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

689

regarding where, when, or how the saline water is applied and whether nonsaline water is available, and depend on whether rain is a significant source of water for crop needs, which include both evaportranspiration (ET) and leaching. Blending water supplies One common way of improving the quality of saline water is by blending it with water of lower salinity. The goal of the “blending strategy” is to combine two sources of irrigation water to produce irrigation water of suitable quality while increasing the overall irrigation water supply. The suitability of the blended water depends on the salt tolerance of the crop being irrigated and the long-term management plan for irrigation and crop production at that site. To use a blending strategy, a controlled way of mixing the water supplies must exist. Shalhevet (1984) discussed two blending processes: network dilution and soil dilution. With network dilution, water supplies are blended in the irrigation conveyance system to achieve the targeted blended ratio. To do this, a facility for blending must be built into the system. With soil dilution, the soil acts as the media for mixing water of different qualities. Different water qualities are alternated, according to availability, between or within an irrigation event. Dinar et al. (1986) described a method to calculate optimal ratios of saline and nonsaline irrigation water for crop production. Caution is advised because blending saline drainage water with goodquality water does not unconditionally increase the usable water supply (Grattan and Oster 2003), nor is it always economically feasible (Dinar et al. 1986). There is an upper salinity limit beyond which the saline water is unsuitable for blending. Good-quality waters that are blended with water that is more saline than this hypothetical maximum (i.e., the maximum salinity of the irrigation water above which is lethal to the plant) will actually reduce the overall “available water” supply. As an illustration, 1 ha-m of pristine snowmelt water blended with 1 ha-m of seawater will produce 2 ha-m of water with a quality equivalent to about halfstrength seawater. After the blend, 1 ha-m of plant-available water would be lost if one were planning on irrigating most horticultural crops, since half-strength seawater is lethal to the crops. Blending is not useful if the saline water cannot supply at least 25% of the total irrigation water requirement (Grattan and Oster 2003). The costs of increased management and risks of potential crop damage associated with adding salts to the irrigation supply will likely outweigh the benefits of a modest increase in the total water supply if the salinity of the total supply is significantly increased by blending.

690

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Cyclic use of saline and nonsaline water The cyclic strategy was first introduced and tested by Rhoades (1984). In this method, saline water is used only for selected crops during certain less sensitive stages of their growing season, while nonsaline water is used at more sensitive phases of growth. The objective of the cyclic strategy is to minimize plant salinity stress during salt-sensitive growth stages, or when salt-sensitive crops are grown in a rotation of crops. This does not necessarily imply that saline drainage water is only applied to salttolerant crops after they reach a salt-tolerant growth stage. In a soil that is initially nonsaline, soil salination lags behind saline water application, allowing a more salt-sensitive crop to be irrigated with saline water for a given duration (Bradford and Letey 1992; Shennan et al. 1995). Likewise, without preplant leaching by irrigation or rainfall, it may be difficult to quickly return to a salt-sensitive crop in the rotation once the seedbed or rootzone is salinized. The yield-threshold levels of salinity, often used as standards for irrigation water quality, reflect the response of crops to the average rootzone salinity after the establishment of seedlings. However, salt tolerance of many crops increases as the plant matures (Maas and Grattan 1999). For this reason, crops can tolerate relatively higher salinity levels late in the growing season without suffering a loss in yield. Furthermore, applying saline water later in the season reduces crop exposure to salinity, allowing waters of higher salinity to be used. With a cyclic strategy, the soil salinity is purposefully reduced by irrigation with good-quality water, thereby facilitating germination and permitting crops with lesser tolerances to be included in the rotation. The cyclic strategy keeps the average rootzone salinity lower than that under the blending method, especially in the upper portion of the profile—a zone critical for emergence and plant establishment. A major limitation is that saline water needs to be available at the appropriate time and in sufficient amounts. If a reservoir of saline water is available, such as saline groundwater, then timely availability will not be an issue. The remaining issues would then be whether the amount is sufficient and whether this amount can be delivered to the location where it is to be used. Sequential use In this practice, part of the farm or subregion is designated as the reuse area. It consists of a field or sequence of fields within the boundaries of a farm, or an irrigation district, that are irrigated with saline water that is collected as subsurface drainage or runoff from a larger area of the farm, or from several farms (SJVDP 1990). Subsurface drainage water collected by a tile drainage system is usually more saline than the original

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

691

irrigation water. This subsurface drainage water is then used to irrigate the next field in the sequence. Subsurface drainage from this field may be applied to a third, and so on. The crops receiving saline drainage water have greater tolerance to salinity than those grown on the initial set of drained fields (Fig. 22-1). The main purposes are to (1) reduce the soil salinity in fields that are tile drained, thereby increasing their productivity; (2) obtain an additional economic benefit by using drainage water as a resource for crop production; (3) minimize the area affected by shallow water tables, thereby increasing the area available to grow high-value salt-sensitive crops; and (4) reduce the volume of drainage water that requires disposal. Such sequential reuse systems are being tested in California’s San Joaquin Valley. One approach is the integrated on-farm drainage management (IFDM) system, which explores many options for managing salinity, including a wide range of species in the rotation with various tolerances to salinity (including forages and halophytes), and potential uses of salt harvested from the terminal solar evaporator (Ayars and Basinal 2005). Although sequential reuse is conceptually attractive, there is a substantial lag time for salts at the beginning of the reuse sequence to reach the final stage. Jury et al. (2003) conducted a water and solute flow study using a transfer function model. Assuming typical drain-line spacing and

FIGURE 22-1. Schematic representation of a sequential drainage water reuse system.

692

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

water management practices, they found that such a reuse system would never reach steady state. Time periods of decades or longer are required for water and salts to move though a sequential system. Because sites with low-lying, fine-textured soils are often where such systems are considered, salt transport times may be even longer. The influence other of local groundwater dynamics can also be significant in estimating timing and direction of salt and water movement. Therefore, caution is advised when designing sequential reuse systems and estimating the rate of salt movement through the sequential system. Steady-state assumptions will result in poor designs. Another example of sequential use has been adopted using Bermuda grass (Cyanadon dactylon, [L.] Pers) grown on salinized land where it is irrigated with water of variable quality, including saline drainage, municipal waste waters from a nearby city, and high-quality water from the Kings River (Corwin et al. 2003, 2006; Kaffka et al. 2002, 2004). These ongoing studies have shown that beef cattle can successfully graze Bermuda grass as a sole source of feed during much of the year. These irrigation strategies are not mutually exclusive but, rather, a combination may be most practical in many cases. For example, within a sequential reuse scheme, blending and/or cyclic methods may be used occasionally for salinity management depending on water availability or to germinate and establish crops, and the types and amounts of water will vary. Fundamental Principles Related to Irrigation with Saline and Saline-Sodic Water The salinity and sodicity of the saline water are the main parameters that determine the feasibility of its use, but other constituents (e.g., B, Se, Mo, NO3) must also be addressed should they be present. Use of this water will require an integrated approach, including new on-farm skills related to irrigation, crop, and soil management, all within the context of being economically feasible and environmentally sound. The entire cropping system must be planned to account for the most important water quality aspects of the low-quality water and its effects on different crops and soils over time. These include crop types with varying tolerance to salinity, nutrient and irrigation management, accounting for trace element accumulation in soils and plants, and maintaining tolerable levels of soil salinity and good soil physical conditions. Salt tolerance Plants vary widely in their tolerance to salinity (see Chapter 13). Most traditional crops are glycophytes that have evolved under nonsaline

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

693

conditions and are unable to cope with the stresses of saline and sodic conditions. Conversely, “salt-loving” plants, the halophytes, thrive under these conditions (see Chapter 6). Crop tolerance to salinity is a major factor affecting irrigation management when using saline waters. Maas and Hoffman (1977) proposed that the salt tolerance of a crop is best be described by plotting relative yield as a continuous function of average rootzone salinity (ECe). They proposed that this response curve could be represented by two line segments, one, a tolerance plateau with a zero slope, and the second, a concentrationdependent line the slope of which indicates the yield reduction per unit increase in ECe. Relative yield (RY) or yield potential can be estimated using the following expression (see Chapter 13): RY (%)  100  B(ECe  A)

(22-1)

where ECe is the average electrical conductivity (dS m1) of the saturated soil extract within the crop rootzone, A is the threshold salinity (i.e., the maximum average rootzone salinity the crop can tolerate before a reduction in yield occurs), and B is the slope of the yield–salinity curve when ECe is greater than A. Salt-tolerant crops usually have a high value of A and a low value of B, although a few have both a low A and B value. In this case, the low B (yield loss in response to increasing salinity) can compensate for the low A (threshold value). Values for these coefficients for various crops can be found in tables presented here and in Chapter 13. Adequate leaching to control rootzone salinity Soil salinity is controlled by avoiding excessive salt accumulation in the crop rootzone. The sustained, long-term use of saline water for irrigation requires a net downward movement of water and salt past the rootzone. This downward movement is commonly referred to as leaching and is necessary regardless of plant type to maintain plant productivity. The leaching fraction (LF), or leaching requirement, is the ratio of the amount of water percolating below the rootzone to the amount of water that infiltrated the soil. The required LF is dependent on plant tolerance to salinity, salinity of the irrigation water, and site-specific conditions. The greater the salt tolerance, the lower the leaching requirement (LR); for a given salt tolerance, the higher the irrigation water salinity, the greater the required leaching. The leaching requirement is an attractive concept but has serious limitations. First, the ET of the crop is assumed to be independent of the average rootzone salinity. As a result, calculated crop water requirements will be overestimated when the average rootzone salinity exceeds the threshold

694

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salinity of the crop, which corresponds to a yield potential less than 100% (Letey and Dinar 1986; Shani et al. 2005; Letey and Feng 2007). Consequently, crop ET will be reduced and leaching will be increased. Other issues also affect the proper calculation of crop water requirements: (1) initial levels of salinity in the rootzone, (2) spatial variation in amount of water applied and the amount that infiltrates into the soil, and (3) difficulty of the application and infiltration of sufficient water to a field to achieve the desired LF. Despite these limitations, in order to control salinity, leaching must occur whether it is achieved at the beginning, through the season, or at the end of the crop season (Ayers and Westcot 1985; Shalhevet 1994). To allow this, soil physical conditions must not be allowed to deteriorate to the point where adequate water for both the crop and for leaching will not infiltrate and move through the soil. This is primarily a problem when the water used for irrigation is sodic, as well as saline, or the normal declines in infiltration rates result in an inability to infiltrate sufficient water to refill the soil within the rootzone. Saline-sodic waters Irrigation with saline-sodic water requires a higher level of management to avoid salinization to sustain production over the long term than is required for irrigation with nonsaline water. Soil physical properties can be altered by irrigation with saline-sodic water, made apparent when good-quality water is used or rainfall occurs after saline-sodic water application (Oster and Jayawardane 1998; Oster et al. 1999a; Shainberg and Letey 1984). Potential adverse effects include reduced infiltration and redistribution within the soil, poor soil tilth, and inadequate aeration resulting in anoxic conditions for roots (Oster et al. 1999a). These negative impacts can be reduced with appropriate soil and water amendments, such as gypsum, sulfur, and sulfuric acid (Oster et al. 1992). In addition, if high levels of boron (B) are present in the water, its accumulation in the soil could adversely affect crop production (Grattan and Oster 2003). There may be an interaction between salinity and B that negates, to some extent, B’s toxic affect on the crop (Ferreyra et al. 1997), but studies have shown that salinity can aggravate B injury (Läuchli and Grattan 2007). The need to leach salts and B from the rootzone will also leach NO 3. Nitrate losses can be mitigated by additional fertilizer application, but such losses are uneconomic and environmentally damaging. On the other hand, if saline drainage water contains NO3 and is used for irrigation, some crops can be adversely affected while other crops can benefit (Kaffka et al. 1999). Trace elements, such as Se or Mo, if present in the salinesodic water, could accumulate in the crop and pose a health risk to either the animal or human that consumes them. Both negative and positive

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

695

aspects of using saline-sodic waters for irrigation will be discussed in following sections. Water quality impacts on soil physical properties Deterioration of soil physical conditions can occur in drainage water reuse systems. Soil tilth and permeability to water and air can be altered by irrigation with saline-sodic water, particularly when irrigation with nonsaline water or rainfall follows. Soil hydraulic conductivities (HCs) (McNeal and Coleman 1966) and infiltration rates (IRs) (Oster and Schroer 1979; Suarez et al. 2006) decrease with decreasing soil salinity and with increasing exchangeable Na. This occurs because both clay swelling and the stability of microstructures in soils depend on an interconnection between salinity and sodicity (Quirk and Schofield 1955). Aggregate deterioration, clay swelling into the water-conducting pores, and clay movement and deposition within the macropores are the mechanisms responsible for loss in permeability. An important aspect of soil management is the recognition of the different responses of surface and subsurface soils to sodicity and salinity (Oster and Shainberg 2001). Because surface soils have a soil–atmosphere interface and are subjected to tillage, they are affected more than subsoils by water drop impact (by rainfall or sprinklers), rapid wetting, irrigation water quality, animal and vehicular traffic, tillage, and surface mulches. The bonding mechanisms associated with organic matter (Nelson et al. 1997) and aging are continually changing in surface soils compared with subsoils. Subsoils have lower wetting rates, water contents prior to wetting are usually higher, organic matter content is usually lower, and the chemical state of organic matter is more stable than in surface soils. Because of these differences, the criteria of acceptable combinations of sodicity and salinity are different for surface and subsurface soils, as are the methods of soil management. Criteria used in California (Table 22-1) are biased toward conditions required to maintain adequate infiltration rates. Farming practices to promote satisfactory soil physical properties Application of gypsum and sulfuric acid to the soil surface are common practices in many arid parts of the world for improvement of infiltration rates and for reclamation of sodic soils (Oster and Jayawardane 1998). Appropriate management practices following gypsum application can prolong its benefits, leading to a reduced need for repeated application. For example, Greene and Wilson (1989) demonstrated that over a period of 3.5 years, the beneficial effects of gypsum on clay dispersion were lost as a result of leaching, but because the establishment of pasture protected the surface from impacts of raindrops, no loss in HC was evident.

696

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 22-1. Combined Effects of Sodium Adsorption Ratio (SAR) and Salinity of Either a Saturation Extract (ECe) or an Irrigation Water (ECiw) on the Likelihood of Problems with Low Infiltration Rates (IRs) Potential Water Infiltration Problem When SAR of the Irrigation Water or Soil Water Is: (1)

Unlikely if ECe or ECw Is: (2)

Likely if ECe or ECw Is: (3)

0–3 3.1–12 12.1–20

0.7 2.0 3.0

0.3 0.5 1.0

20.1–40

5.0

2.0

SAR is the sodium adsorption ratio. SAR  Na/(Ca  Mg) are expressed in mmol L1.

1/2

where the ion concentrations

ECe is the electrical conductivity of a water extract obtained from a saturated soil paste (dS/m). ECw is the electrical conductivity of the soil water (dS/m). From Ayers and Westcot (1985).

Applying gypsum directly to irrigation water to foster improved infiltration rate is another common practice in California, in use since at least the 1950s (R. S. Ayers, personal communication). Injection of gypsum directly into the irrigation water at rates ranging from 0.17 to 0.34 kg/m3 is a common practice for low-salinity irrigation waters (Oster et al. 1992). These rates correspond to the addition of 2 to 4 mmolc/L of Ca and an increase of 0.15 to 0.30 dS/m in EC of the irrigation water. Both changes represent a considerable improvement in the quality of the irrigation water by slightly increasing the EC of the water while reducing the sodium adsorption ratio (SAR) at the same time. Injection of gypsum into the lowsalinity water used in a cyclic strategy with saline-sodic water could be an appropriate practice. Incorporation of organic matter to the soil can also affect soil physical conditions. Taylor and Olsson (1987) and Quirk (1978) demonstrated that increased levels of organic matter arising from pasture root systems stabilize soil structure after gypsum is no longer present at the soil surface in sufficient amounts. Under pasture, aggregate stability increases with time, partly because of increased levels of fresh organic matter leading to reduced susceptibility of soils to an unfavorable combination of sodicity and salinity decreases. Other significant factors that influence aggregate stability are the antecedent water contents before wetting occurs, the rate of wetting, and soil texture (Mamedov et al. 2001; Shainberg et al. 2001),

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

697

as well as clay mineralogy, organic matter, CaCO3 sesquioxides, and pH (Levy et al. 1998). Permanent crop cover also protects the soil surface from dispersion by water droplet impact. The adoption of farming practices, such as minimum tillage and direct drilling without tillage, leads to increased retention of crop residues in the form of surface mulches. This encourages soil microbial activity that increases and maintains the continuity of large biopores, which are effective at conducting water and air to subsoils (Jayawardane and Chan 1994). However, in some instances the opposite effect can occur. Cover crops that were incorporated into saline-sodic soils actually increased crust strength and substantially reduced cotton emergence (Mitchell et al. 2000). These researchers demonstrated that the incorporated stubble acted to reinforce and strengthen soil crusts. Direct sodic effects on plants In addition to the effects of salinity/sodicity on soil physical properties that indirectly affect crops, sodic conditions may also create direct harm to crops such as Na-induced Ca deficiency. Calcium deficiency in the crop may be diagnosed by obvious symptoms, such as a whiplike appearance in young emerging leaves in cereals, blackheart in celery, and blossom end rot in tomatoes and peppers. More commonly, Ca deficiency symptoms occur without visual symptoms (Grattan and Grieve 1999). Other specific-ion interactions have been described in detail elsewhere in this manual (Chapter 6). Boron accumulation and potential toxicity to plants Under some circumstances, B may be a concern where saline-sodic water contains high amounts of this potentially toxic element. Such is the case in the shallow groundwater on the west side of the SJV in California, where B concentrations typically range from 2 to 10 mg/L (SVJDP 1990). Soils in parts of Australia and Chile also have high B under saline conditions. Saline byproduct water from oil drilling can also have high concentrations of B, which represents an additional constraint for the use of vegetated systems to dispose of this water (R. Soppe, personal communication). Despite these levels of B in the drainage water in the SJV, B toxicity has not yet been reported in annual crops. In the SJV, saline-sodic drainage water containing 7 to 10 mg/L of B has been used to irrigate cotton, melons, sugar beets, tomatoes, safflower, and wheat (Ayars et al. 1990; Ayars et al. 1993; Bassil and Kaffka 2002a,b; Grattan et al. 1987; Kaffka et al. 1999; Mitchell et al. 2000; Rhoades et al. 1988; Shennan et al. 1995). At least part of the success may be attributed to the fact that cotton, sugar beets, and tomatoes are tolerant of B (Maas and Grattan 1999), but wheat is reported as sensitive to B (Maas and Grattan 1999). An additional factor is that rainfall

698

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

reduces the B hazard, a factor that normally is not taken into account when assessing B hazards. Moreover, B toxicity is most commonly diagnosed by visible leaf injury or injury to growing tips. Crop injury and yield are not invariably related, in part because injury that develops on leaves usually occurs late in the growing cycle. Indeed, leaf B concentrations have been related to B injury in eucalyptus, where the interactive effects of salinity and B on foliar injury depended on the physiological age of the leaf (Poss et al. 1998). Interactions between salinity and B on crop growth and yield are not well understood (Läuchli and Grattan 2007). Some crops are able to remobilize B within the plant such that younger tissue has higher B concentrations than older tissue (Brown and Shelp 1997). In these instances, B injury is expressed as injury to the meristematic tissue such as developing leaves and buds. Others have described such injury in deciduous trees as “twig dieback” (Ayers and Westcot 1985). Selenium and molybdenum accumulation in crops Naturally occurring trace elements in soils and in the underlying shallow groundwater adds another dimension to the management of saline drainage waters (van Schilfgaarde 1990). Trace elements, particularly Se and Mo, occur on the western SJV of California and pose a threat to the sustainability of irrigated agriculture in this area. Selenium and Mo are important because they are found in relatively high concentrations at many locations in their geochemically mobile forms (i.e. SeO2 and 4 MoO42, respectively). These chemical species are also in forms that are most biologically available. The uptake of Se and Mo by plants is the primary process where essential trace elements for humans and animals are introduced naturally into their diet from the terrestrial environment. Selenium and Mo concentrations in plants vary widely among plant types, plant organs, and location. A number of greenhouse and field studies have measured the accumulation of Se in edible tissue of crops (Tanji et al. 1988; Valoppi and Tanji 1988). Field studies included experiments where crops where grown in high-Se soils (Bañuelos et al. 2003) and/or where crops were irrigated with saline drainage water containing high levels of Se. The results indi2 cated that the presence of SO2 uptake, 4 dramatically reduced SeO 4 resulting in lower concentrations of Se in its tissue and a nonuniform distribution of Se within the plants. The general trend was that in the shoot, concentrations were higher in the leaves and stems than in the fruit or grain. The Se concentration range in edible portions of plants in all these field experiments and surveys reviewed by Valoppi and Tanji (1988) was 0.002 to 5.0 mg/kg, dry weight (DW). A survey health assessment of these studies indicated that crops consumed at these levels contribute minor amounts to the daily Se intake of humans.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

699

As important as the potential dietary concerns for humans are, potential toxicological problems exist for livestock whose diet may rely almost entirely on forage grown in high-Se and -Mo areas within the SJV. Selenium and Mo are essential elements to animals and the concentration range in forage between deficiency and toxicity to livestock is rather narrow (James et al. 1968; Ohlendorf 1989; Osweiler et al. 1985). Selenium toxicity, such as alkali disease, can occur in livestock that graze on forage containing high levels of Se (Rosenfeld and Beath 1964). The National Research Council set the maximum tolerable concentration (MTC) for Se in beef cattle at 2 mg/kg DM (NRC 1996), but for dairy cattle 5 to 40 mg/kg was listed as causing chronic toxicity (alkali disease) in several weeks to months (NRC 2001). Molybdenosis is a nutritional disorder that ruminant animals, particularly sheep and cattle, may develop if the animals feed on forage that contains high levels of Mo (Kubota and Allaway 1972; Ward 1978). Molybdenosis results from an Mo-induced Cu deficiency and is often called molybdenum-induced hypocuprosis (Mason 1990). It can also delay first oestrus in cattle and reduce the pregnancy rate when Mo concentrations in forage are as low as 5 mg/L (Phillippo et al. 1987). Although Cu deficiency symptoms can be easily treated by Cu supplements, fertility symptoms appear to be related directly to the Mo concentration in the forage and are not readily correctable. Two beef cattle grazing studies are currently underway in the SJV that will examine Se and/or Mo accumulation in tissue when irrigated with saline drainage water and will hopefully shed more light on this issue. Effect of saline drainage water on crop quality Saline drainage water can have positive effects on crop quality. Examples include increased protein content and total digestible nutrients in alfalfa (Rhoades et al. 1988); improved netting, flesh color, and taste in cantaloupes (Rhoades et al. 1988); increased soluble solids, fruit acidity, and improved firmness and color in tomatoes (Grattan et al. 1987; Mitchell et al. 1991; Pasternak et al. 1986); and increased flour protein and loaf volume in wheat (Rhoades et al. 1988). However, improvements in these quality characteristics are not always consistent and do not justify the application of saline drainage water to crops. Nevertheless, improved quality characteristics are potentially beneficial secondary effects. Irrigation with saline water does not necessarily improve quality in all crops. For example, Rains et al. (1987) and Bassil and Kaffka (2002b) did not observe any difference in the percentage or quality of oil in safflower seed irrigated with saline drainage water as compared with nonsaline water. Moreover, many of the studies in California did not include, for example, leafy vegetables that are known to be sensitive to

700

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salinity and where resulting size reductions can significantly affect marketability. Grieve et al. (2005) found that cut flowers, such as statice (Limonium spp.), can be irrigated with saline waters, but stem length may be reduced, which adversely affects market quality. When using drainage water, correlated water quality characteristics may also be important. Sugar beet requires low nitrates in the crop rootzone late in the season to increase beet sugar content. Since saline drainage water may contain high concentrations of NO3 , irrigation with saline drainage water late in the growing season can reduce the sugar content in the beet (Kaffka et al. 1999). Conversely, moderate salinity reduced the solution NO3 nitrogen requirement of the commercially grown Matthiola to where production of marketable stem-length flowers was maintained in a closed irrigation system (Grieve et al. 2008), thereby reducing off-site NO3 pollution. Field Research Studies A number of field studies have been conducted over the past 20 years that tested various strategies for irrigating crops with saline or salinesodic water. Rhoades (1987) tested the cyclic strategy of using waters of different salinities and found it to be sustainable in maintaining crop rotations that included both moderately salt-sensitive and salt-tolerant crops. In a successful test of the cyclic strategy conducted in the SJV of California, a 0.5-dS/m water was used to irrigate cotton during germination and seedling establishment, and 7.9-dS/m, SAR 11 water was used thereafter (Rhoades 1987). Wheat was subsequently irrigated with the 0.5-dS/m water, followed by 2 years of sugar beets with the cyclic strategy used again for irrigation. Rhoades et al. (1988; 1989) reported the results from a second study conducted in the Imperial Valley of California. In a rotation of wheat, sugar beets, and melons, Colorado River water (1.5 dS/m, SAR 4.9) was used to irrigate cantaloupes, a moderately salt-sensitive crop, and for the preplant and early irrigations of wheat and sugar beets. Alamo River drainage water (4.6 dS/m, SAR 9.9) was used for all other irrigations. Sugar beet and wheat yields were not reduced and crop qualities were often improved from the use of saline drainage water. Ayars et al. (1990, 1993) used drip irrigation for 3 consecutive years to apply a 7- to 8-dS/m, SAR 9 water to cotton after it was established with 0.4- to 0.5-dS/m water. The saline water supplied 50% to 59% of irrigation water requirement. A wheat crop irrigated with the 0.5-dS/m water followed cotton; sugar beets followed wheat and were irrigated with the 8.0-dS/m water after stand establishment. Yields under these conditions were the same as from continuous irrigation with the good-quality water. The investigators did note, however, an increase in soil B levels

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

701

over time from drainage water application (5 to 7 mg/L of B) despite annual rainfall and preseason applications of nonsaline water in excess of 150 mm. Others have expanded on the cyclic reuse approach. Shennan et al. (1995) tested two cyclic, drainage-water reuse practices on processing tomatoes in a 3-year rotation with cotton over a 6-year period. In both practices, drainage water was applied to processing tomatoes after first flower to take advantage of salinities enhancement of fruit quality, and continued as the sole source of irrigation water until the end of the season. In one practice, nonsaline water was used at all other times (i.e., saline water applied in 1 out of 3 years). In the other practice, drainage water (EC  7.4 dS/m, SAR 12) was also applied to the following cotton crop after thinning, while nonsaline water was applied at all other times (i.e., saline water was applied in 2 out of 3 years). Nonsaline water (EC  0.4 dS/m, SAR 1.6) was used as the source of irrigation water at other times, and both reuse practices were compared to rotations where only nonsaline water was used. In the practice where drainage water was applied in 1 out 3 years, yields of tomatoes were sustained even though drainage water supplied up to nearly 60% of the irrigation water requirements. In the cyclic reuse treatment where drainage water was applied in 2 out of 3 years, tomato yields were reduced in 2 of the 6 years. Soluble solids in tomato fruit, conversely, were increased in 5 of the 6 years in drainage-water treatments. Similar results were found in field studies with tomatoes conducted in different locations in the SJV but for only 1 year, where drainage water supplied more than 65% of the irrigation water requirement (Grattan et al. 1987). Pasternak et al. (1986) also reported an increase in soluble solids in tomatoes when irrigated with a 7.5-dS/m water after the fourth or eleventh leaf stage, but yields were reduced by 30%. Differences between these studies may be due, in part, to differences in the anion composition of the saline water. In Israel, where Pasternak and colleagues conducted their work, the saline water is normally Cl-dominated, whereas in the SJV of California the saline drainage water is sulfate-dominated. Chloride salts are likely to be more damaging than sulfate salts. A study by Shennan et al. (1995) also examined the behavior of salts and B over time at different soil-depth increments. On a relative basis, salts were more readily leached than B. At the 60- to 140-cm depth interval, ECe in plots irrigated with saline water increased the first year (1986). Concentrations of B at this depth, on the other hand, were not found to increase until 1988 in plots that received drainage water in 2 out of 3 years, and 1989 in plots that received drainage water in 1 out of 3 years. In the upper 15 cm of the soil profile, the ECe increased after saline-water irrigation, but then after 2 years of irrigation with nonsaline water, the ECe returned to the level found in the control. Boron, conversely, was not

702

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

as responsive and concentrations in drainage-treated plots remained high after 2 years of application with aqueduct water. As indicated earlier, the feasibility of cyclic reuse practices depends, in part, on changes in soil chemical and physical properties over the long term. In the long-term cyclic study conducted on a clay loam soil by Shennan et al. (1995), they found no difference in water IRs, measured with a steady-state infiltrometer, between plots that received saline water and those that did not. Nevertheless, they did find a significant reduction in cotton stands in 1988 in plots that were salinized the previous year. Use of saline drainage water (9,000 mg/L TDS and 16 to 30 SAR) on a clay soil also caused a reduction in stands of both cotton and safflower (Rains et al. 1987; Rolston et al. 1988) within 2 years. The inability to prepare a seedbed with the tilth necessary for optimal emergence for cotton seed resulted in poor stand establishment (Oster 1994). Mitchell et al. (2000) also found a reduction in cotton seedling establishment related to irrigation with saline-sodic water. Surprisingly, the inclusion of cover crops to improve soil physical conditions actually aggravated the situation. As indicated earlier, they suggested that stand establishment was reduced by the formation of “stubble-reinforced” surface crusts. Kaffka and Hembree (2004) studied the emergence of sugar beet seedlings in salt-affected soils in the San Joaquin and Imperial Valleys (IV). Different seed treatments were evaluated for their effect on emergence. At the IV site, salinity accumulated at and near the soil surface due to surface irrigation and upward movement of salt from a shallow saline water table. Primed seeds treated with imidicloprid, a neonicotinid insecticide, emerged at significantly higher rates in this location than where the other treatments were used, resulting in superior emergence under transient saline conditions. At the SJV site, salinity in the upper 30 cm of soil ranged from 2 to 10 dS/m due to prior saline irrigation applications (Shennan et al. 1995; Kaffka et al. 1999; Bassil and Kaffka 2002a,b). Therefore, seedling dry matter accumulation and rates of emergence declined at ECe levels greater than 6.0 dS/m. However, final plant populations were not affected over the salinity range. The rate of emergence of primed seeds in saline plots equaled that of nonprimed seeds in plots with low levels of salinity. In the same field used by Shennan et al. (1995), Kaffka et al. (1999) conducted a study with sugar beets. Results were consistent with most of the reports already cited. Transient salinity levels in saline-irrigated plots exceeded threshold levels for yield reduction (see Chapter 13). The capacity of a deep-rooted crop to recover water from a larger volume of soil allowed the beets to adjust without effects on root yield. Alternating saline with nonsaline irrigation water (using saline water either before or after nonsaline in midseason) did not affect root yields but, because of

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

703

NO 3 present in the saline well-water source, late-season irrigation with saline water reduced sugar yields more than early-season irrigation. High concentrations of nitrate in saline drainage water may be a resource or a nuisance. Nitrogen is often a growth-limiting nutrient, so N-fertilizer applications can be adjusted downward to take advantage of the N applied in the drainage water. At the same time, the crop is absorbing nitrate from the drainage water, which can adversely affect crop quality by reducing root sugar content (Kaffka et al. 1999). Therefore, accounting for N and making sure that soil N levels are sufficiently low near the end of the season are necessary steps to assure sugar beet root quality. In general, accounting for NO 3 and other nutrients in drainage water is essential for a prudent drainage water management program. Following sugar beets, Bassil and Kaffka (2002a,b) grew safflower in the same long-term drainage reuse plots. Because safflower is susceptible to Phytophthora when irrigated late in the growing season, all irrigations were completed by late April. Safflower used less water in saline plots, produced less shoot biomass, and matured earlier, but seed yields were similar in saline and nonsaline plots. Safflower was able to adjust its harvest index under salt stress and overall oil yield was not affected. Testing Crops for Saline Reuse Systems in California A number of conventional crops are currently grown in the SJV (e.g., cotton, melons, safflower, sugar beets, tomatoes, and wheat) that were tested and found to be successful in both short- and long-term drainage water reuse studies (Ayars et al. 1990; Ayars et al. 1993; Bassil and Kaffka 2002a,b; Grattan et al. 1987; Kaffka et al. 1999; Mitchell et al. 2000; Rhoades et al. 1988, 1989; Shennan et al. 1995). In many reuse systems, however, a wide range of saline conditions can occur. For example, in fields at the end of the sequential reuse systems, soil salinities (ECe) can be in excess of 40 dS/m. This provides an opportunity to introduce a number of crops with a wide range in salt tolerance in these systems, including highly salttolerant crops that would otherwise not exist in the valley. Different crops may be more appropriate than others, depending on the field conditions. Desirable and undesirable characteristics for crop selection under irrigation with saline water are listed in Table 22-2. Vegetable crops At the U.S. Salinity Laboratory, Shannon et al. (1998) tested a number of novel leafy vegetable species to determine whether they could fill a niche within a drainage-water reuse sequence. Consistent with results from tests with most commercial vegetable cultivars, these were moderately sensitive to salinity, suggesting that they will perform best in the

704

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 22-2. Criteria for Selecting Crops for Saline Water Reuse Selection Criterion (1)

Desirable (2)

Undesirable (3)

Economic value/ marketability

High/marketable

Low/unmarketable

Crop salt tolerance

Tolerant

Sensitive

Biomass production

High biomass production

Low biomass production

Crop boron/chloride tolerance

Tolerant

Sensitive

Crop potential to accumulate toxic element

Excludes toxic element

Accumulates toxic element

Crop quality

Improved by saline water irrigation

Unaffected or adversely affected by saline water

nonsaline portion of a sequential reuse system. Purslane (Portulaca oleracea), on the other hand, grew well in sand tank experiments irrigated with simulated drainage effluent, suggesting that this crop may be suitable in saline portions of a sequence (Grieve and Suarez 1997). Tree crops Eucalyptus plantations have been established throughout California’s SJV for the purpose of reducing the volume of drainage water that needs ultimate disposal (Cervinka 1994). Eucalyptus trees were proposed as an important component of the sequential reuse system (SJVDP 1990), but were not found to be as tolerant to frost or to waterlogged soils as desired (Benes et al. 2004). Currently, they are now more commonly used in interceptor strips to reduce subsurface flows between the different stages of the sequential reuse system (Cervinka et al. 1999). Eucalyptus trees have been found to reduce dryland salinization in areas cleared of native vegetation in western Australia (Bari and Schofield 1992). Revegetation with eucalyptus trees reduces groundwater recharge and lowers shallow saline-water tables, thereby reducing salinization of the upper portion of the soil profile (Morris and Thomson 1983). A study was conducted at the U.S. Salinity lab using large sand tanks to determine potential ET rates. Eucalyptus camaldulensis clone 4544 was irrigated with simulated drainage water that varied in salinity (EC 2 to 28 dS/m) and B concentration (1 to 30 mg/L) (Shannon et al. 1998). Sand tanks were used because they drain well and the concentrations of salts and B in the soil water are close to that in the irrigation-treatment water. When the average rootzone salinity of the soil water was 15 dS/m, tree ET

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

705

was reduced to 53% of those in the nonsaline treatment when evaluated over the entire period where trees were subjected to saline water. The reduced ET is attributed largely to salinity’s affect on tree biomass. Eucalyptus tree biomass decreased as the salinity of the soil water increased above 6 dS/m (Shannon et al. 1998). In addition, these investigators found a significant interaction between B and salinity (Grattan et al. 1996). Tree biomass was affected by increased B at low salinity but not at higher salinity (i.e., 22 dS/m). Based on this sand tank study, Eucalyptus camaldulensis is more appropriately classified as moderately salt-tolerant rather than salt-tolerant. A field study testing the effectiveness of eucalyptus (Eucalyptus camaldulensis, clones 4543, 4544, and 4573) was conduced in a hydrologically closed basin within the SJV of California (Oster et al. 1999a). This district had fine-textured soils with many in the textural class of clay to clay loam. The trees were planted and irrigated with nonsaline water to facilitate survival of the young trees. Trees were then irrigated with saline-sodic water (EC 8.5 dS/m, SAR 33) several months later after they were established. The average ECe and SAR in the 0- to 60-cm depth from 1996 through 1998 were 15 dS/m and 36, respectively. Tree biomass was greatest in those plots that received gypsum applications. Fall-applied gypsum improved soil aeration, infiltration, and drainage during the winter when rain occurred. After substantial rainfall occurred in 1998 between day-ofyear (DOY) 31 and 125, resulting in ponding in all treatment plots, oxygen diffusion rates remained at 0 g(O2)/(cm2 min) in the untreated plot to DOY 80 to 190, whereas rates increased to 0.3 g (O2)/(cm2 min) in the gypsum-treated plot after DOY 140. Therefore, gypsum application substantially increased the oxygen diffusion rates in winter months and improved tree performance. Pistachios are a highly salt-tolerant nut crop. In a 9-year study, Sanden et al. (2004) evaluated the Pistacia vera ‘Kerman’ scion on four rootstocks under irrigation with saline drainage water at salinities from 0.5 to 12 dS/m ECw and found no impact on yield with irrigation waters up to 8 dS/m. At 12 dS/m ECw, the nut yield was 81% of the low-salinity control yield, and the cumulative water use (ET) was 64% of the control treatment. A salinity tolerance threshold for the pistachio rootstocks tested ranged between 9 to 10 dS/m ECe. The authors caution the reader that some deep roots of trees in high-salinity plots may have extracted nonsaline water from adjacent low-salinity plots even though plastic barriers were installed to a 1.5-m depth between treatments. Athel (Tamarisk aphylla) and mesquite (Prosopis alba) trees can grow at considerably high salinities (15 dS/m ECe). In the sequential reuse system (integrated on-farm drainage management, IFDM) at Red Rock Ranch in the SJV, P. alba has been established in soils reaching 27 dS/m ECe (S. Benes, personal communication). Although athels have little

706

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

economic value, the wood of P. alba has high value for furniture construction and charcoal production. In addition, the pods it produces contain 35% sugar and 9% protein and are highly palatable for livestock (Velarde et al. 2003). Salt-tolerant forages A number of salt-tolerant forages may play an important role in reducing drainage water volumes in the SJV while at the same time producing a food source for dairy cattle and possibly sheep. Masters et al. (2007) point out that in spite of relatively little effort to improve the feed value of salt-tolerant plants through breeding or selection programs, there is a range of plants capable of growing under saline conditions that could provide a feed source for livestock. At lower salinity levels, this includes both grass and legume forages, particularly when the availability of water is high. At high salinity (25 dS/m), production is substantially reduced and forage choices are fewer, but there is a range of halophytic grasses and shrubs that can produce 0.5 to 5 Mg/ha of edible dry matter per year. Fortunately, most research has shown that salinity does not lower organic forage quality (Suyama et al. 2007b) and, in some cases, it may slightly improve certain quality characteristics (Robinson et al. 2004; Masters et al. 2007). The exception may be the accumulation of certain minerals, particularly Se, Mo, and S, to excessive levels, as discussed earlier, or the stimulation of secondary metabolites specific to plants growing in saline environments, which could be inhibitory or toxic (e.g., oxalate, coumarate, and nitrate) (Masters et al. 2007). A greenhouse study was conducted in sand tanks at the U.S. Salinity laboratory to evaluate the performance of 10 promising forages that included both grass and legume species (Grattan et al. 2004a,b; Robinson et al. 2004). The legumes (alfalfa and trefoil) were sensitive to salinity despite having good forage quality; however, the overall biomass production of the alfalfa species (‘Salado’ and ‘SW 9720’) was high under saline conditions relative to other more salt-tolerant species. Tall wheatgrass (cv ‘Jose’), Bermuda grass (cv ‘Tifton’), and paspalum (cv ‘PI 299042’) emerged as the top choices based on combined factors including biomass production, salt tolerance, and overall forage quality (both organic and inorganic). Investigators cautioned that elevated tissue sulfur concentrations could be problematic to ruminants that would rely on these forages as a sole source of feed (Grattan et al. 2004b). Several field studies were conducted to test the feasibility of a number of salt-tolerant forages and forage cropping strategies (Corwin et al. 2003, 2006; Oster et al. 1999a; Kaffka et al. 2002, 2004; Suyama et al. 2007a), as well as other crops, such as wheat, forage brassicas, and safflower, as suitable crops for irrigation with saline-sodic water.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

707

One such field study has been reported by Kaffka et al. (2002, 2004) and Corwin et al. (2003, 2006): a 24-ha site with variable salinity in the western SJV was developed in 1999 to study the use of saline drainage and other waste waters (ECw average 3.6 dS/m with range from 1.2 to 12 dS/m) for the production of forages and cattle. Tile drains were installed at a depth of 1.3 m and spaced 37 m apart. Baseline soil physical and chemical properties were assessed before the project began in summer 1999 (Corwin et al. 2003). Soil maps describing the chemical and physical properties were developed with GPS mapping and soil sample site locations determined using ESAP software (Lesch et al. 2000). A second survey was carried out in March 2003 and investigators found that the salinity and B in the top 60 cm of soil declined, while subsoil concentrations (60–120 cm) were unchanged (Corwin et al. 2006). Bermuda grass (Cynodon dactylon, vars. ‘Common’ and ‘Giant’), planted to eight paddocks each, remained productive (1.5 to 2.5 tons/ha DW, depending on cultivar) after 5 years of irrigation with saline drainage and other waste waters (soil salinity averaging 13 dS/m ECe, 0–30 cm depth) (Kaffka et al. 2004). In places where ECe exceeded 20 dS/m, stands failed. Livestock trials were carried out for 3 years (2001–2003) (S. Kaffka, personal communication). Cattle were estimated to remove 40% to 60% of the standing forage biomass while grazing during the summer, with a larger percentage by late fall (October) when grass growth rates declined. These fields supported beef cattle with weight gains of 0.75 kg/day once copper supplementation was administered to offset a deficiency due in part to high sulfur and Mo in the drainage water. In areas of the pasture where the soil salinities exceeded 20 dS/m ECe, the Bermuda grass did not grow well. On a hay basis, forage crude protein (CP) contents averaged 9.0 %, (range: 4.2% to 22.1%), acid detergent fiber (ADF): 29.6% (range: 20.7% to 42.3%), B: 245.4 mg kg1 DW (range: 73 to 1,004 mg kg1); Mo: 1.44 mg kg1 DW (range: 0.3 to 5.3 mg kg1) (Kaffka et al. 2004). CP and trace element concentrations were greater in the upper portion of the canopy selected by cattle; Na content was greatest in the lower portion of the canopy. At this location, Mo rather than Se was found in large amounts in some areas of the field. Generally, Cu⬊Mo ratios in ruminant diets below 3 or 4 are thought to affect cattle health, but there is little formal research in this area (Suttle 1991). The Cu⬊Mo ratios at this site averaged 5.2 and no Mo toxicity was observed during the study period. In field studies conducted by Suyama et al. (2007a), ‘Jose’ tall wheatgrass emerged as a top candidate for a particular drainage water reuse system, among the forages tested, due its ability to maintain adequate dry matter yield (7.0 Mg/(ha-yr) and high forage quality (metabolizable energy of 9.3 MJ/kg DW), even when growing in soils having ECe of 19 dS/m, SAR of 37, and B  24 mg/kg DW. Creeping wildrye had higher dry

708

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

matter yield (11.5 mTons/ha) when grown in a less saline field (13 dS/m ECe), but its forage quality was considerably lower (metabolizable energy  8.1 MJ/kg DW). At this SJV location, Se, not Mo, concentrations are exceptionally high in the soils and drainage water and, after multiple years of drainage-water irrigation, these forages accumulated 6 to 11 mg/kg Se, well above the MTCs) of 2 to 5 mg/kg DW (NRC 1996, 2001). Such levels of Se could presumably cause toxicity in ruminants if used as a sole forage source, but the forage could be used as an Se supplement in Se-deficient areas of the SJV. A beef cattle study is currently under way at this ranch to examine the potential for adverse effects to animal health due to this enrichment in Se and sulfur in these forages. In other studies conducted in California’s SJV, important work was carried out to evaluate the effects of high soil Se concentrations on grazing cattle. These cattle grazed on forage grown in what used to be Pond 12 at the Kesterson wildlife refuge in California’s SJV for approximately 6 months in 1999 (J. Maas, personal communication). During this period, forage Se concentrations exceeded 1 mg/kg DW. Blood and fecal Se concentrations were elevated, but none of the cattle tested exhibited any signs of Se toxicity, and their general health was not adversely affected. Additional research must be continued in this area as well as applying research experiences from around the world to the conditions in the SJV (Sharma and Tyagi 2004; Qadir and Oster 2004). Among the halophytes, Atriplex spp. and, to a lesser extent, salt grass (Distichlis spicata) have potential as forages, while others have potential for shade and for water table control (salt cedar) (Oster et al. 1999b). In addition to the ‘Nypa Forage’ salt grass (Nypa Inc., Tucson, Arizona), a native distichlis variety has performed well in a saline drainage-water reuse system at a ranch in the southern SJV (Benes et al. 2004). Extreme halophytes Salicornia (Salicornia bigelovii, commonly known as dwarf saltwort) and iodine bush (Allenrolfea occidentalis) have limited economic value at present, but they can thrive in soils with 50 to 60 dS/m ECe in the top 30 cm of soil (Benes et al. 2004). Therefore, they both have a potential indirect economic benefit as crops that can further decrease drainage volume. Under irrigation with high-Se drainage water (800–1,200 g/L), they can accumulate up to 14 and 7.5 g/kg DM of selenium, respectively (Benes et al. 2005), which gives them potential as processed, organic Se supplements that are commonly administered to livestock in Se-deficient areas. Salicornia can maintain ET rates in excess of reference evapotranspiration (ETo) when irrigated with hypersaline drainage water (29 dS/m and 25 mg/L B) (Grattan et al. 1999). More remarkable is that this leafless plant loses the vast majority of its water as transpiration, not evaporation from a wet soil surface (Grattan et al. 2008). Salicornia, which is one

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

709

of the most salt-tolerant of the vascular plants, is sold as a salad supplement in Europe and produces oil in its seed that is equal in quantity and quality to soybean oil (Glenn et al. 1999). It has considerable promise as a halophyte for seawater irrigation (Glenn et al. 1998). Although this plant grows vegetatively well in the SJV, its seed production is considerably less than when grown along the coast. Worldwide efforts to improve crop selection for saline agriculture The search for new and improved plants that are appropriate for saline agriculture is an ongoing effort in many parts of the world (e.g., Chabbra 1996; Epstein et al. 1980; Glenn et al. 1999; Greig 1996; NRC 1990). Some groups have attempted to improve the salt tolerance of existing agronomic and horticultural crops, but the success, at least over the past two decades, has been rather dismal (Flowers and Yeo 1995) largely due the complexity and multiple genes affecting salt tolerance. Others have studied or are currently studying halophytes as potential new crops (Boyko and Boyko 1959; Glenn et al. 1999). The International Center for Biosaline Agriculture was established in 1996 in Dubai in the United Arab Emirates to further develop possibilities for the use of saline waters for irrigation, including screening of the most suitable species and varieties, with particular emphasis on forage crops (ICBA 2005). Uncertainty about the lower salt-tolerance limit at which a plant is considered a halophyte makes it difficult to compile a list of potential halophytes that have crop potential (Glenn et al. 1999). Nevertheless, Mudie (1974) began to compile a list that was later expanded by Aronson (1989) and the National Research Council (1990) to more than 1,500 species, ranging from existing salt-tolerant crops (e.g., sugar beets, barley, date palms) on the low end to a handful of crops, such as Salicornia bigelovii, that can continue to grow well when the soil-water salinity exceeds twice that of seawater (Glenn et al. 1997). The species mentioned here do not constitute a comprehensive list of the potential crops that could be adopted when using saline and salinesodic waters for irrigation. Numerous other salt-tolerant crops, shrubs, and trees have potential for the production of food, fuel, fodder, and fiber when irrigated with saline or saline-sodic water. More comprehensive crop lists are contained in the NRC’s Saline Agriculture: Salt-Tolerant Plants for Developing Countries (NRC 1990), along with other articles cited in this section. SUMMARY AND CONCLUSION Irrigation with saline-sodic water is necessary in parts of the world, particularly those areas with limited supplies of good-quality water or those affected by shallow saline groundwater. Careful management of

710

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

saline-sodic irrigation water, and monitoring changes in soil chemical and physical characteristics, are essential for success. Also required are proper crop selection, a means of providing sufficient leaching to control soil salinity, accounting for rainfall, and the use of amendments to control the effects of sodicity on soil structure and infiltration. If trace elements are present, close attention should be given to accumulation in the soil and bioaccumulation within the crop, particularly forages. Without careful management, in the long term the area of degraded, salinized lands will increase. Various methods for reusing saline or saline-sodic water for irrigation include blending, cyclic use of two different water supplies that vary in salinity, and sequential reuse of water in a particular region. In practice, farmers may use available water opportunistically, obscuring differences in reuse techniques. The long-term success of irrigation with saline-sodic water, regardless of the strategy used, will depend on the evolution of practical management strategies applying new research and practical farming experiences as they emerge in the future. The major factors affecting the sustained use and reuse of saline-sodic water are magnitude and type of salinity, soil sodicity hazard, and crop salt tolerance. Studies have shown that soil quality can be maintained if sound irrigation and soil management is used, coupled with sufficient leaching for the crop grown. Slow infiltration of water and subsequent slow redistribution within soils is characteristic of many soils, particularly those with high clay contents. Adverse effects on soil structure and infiltration occur not so much by the direct use of saline-sodic water for irrigation but more by subsequent rainfall or irrigation with low-salinity water in soils where a sodium hazard has developed. Irrigation with saline-sodic water followed by rain or irrigation with nonsaline water can enhance soil crusting, adversely affect the tilth of the seedbed, reduce seedling emergence, reduce infiltration rates, and aggravate waterlogged conditions that can reduce soil aeration, thereby affecting crop growth. Researchers and farmers have established that these effects can be largely mitigated by incorporation of gypsum or other amendments that liberate free Ca2 in the upper portion of the soil profile. Trace elements, such as B, Se, and Mo, if present in the water, also affect the feasibility or extent to which this water can be used to irrigate certain crops. Particularly important is crop tolerance of B, or the accumulation of Se and/or Mo to levels hazardous for human or livestock nutrition. Boron toxicity has not been observed on many commercial (e.g., tomato, cotton, sugar beet, melon) or nonconventional crops (e.g., salicornia) in drainage water reuse studies in California. Boron toxicity has been observed, however, on fruit trees, pistachios, and eucalyptus trees. On eucalyptus and pistachios, some of this injury has been transient, and it is not clear to what extent the observed injury reduces tree growth.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

711

Some studies indicate that there is an interaction between salinity and B in many crops whereby salinity reduces the toxic effects of B. Despite these encouraging observations, long-term effects from B accumulation in soils still remain somewhat uncertain. Boron tends to be more resistant to leaching than are salts. Consequently, it is possible that B concentrations in soils may continue to increase to levels that reduce yields of sensitive crops. Although Se concentration was found to increase in crops grown in the SJV of California that were irrigated with saline drainage water, in no instance was it reported that the Se concentrations in crops pose a potential health risk to humans, even when irrigated with drainage water containing more than 300 g/L Se. It remains unknown whether or not Se and/or Mo will pose a potential threat to ruminants if they are fed with forages irrigated with saline-sodic water containing these constituents. Much depends on the type of livestock used, their stage of development, and the amount and type of forage in their diets, including how long they are fed with forages containing high levels of trace elements. This warrants further study, and some work to address this concern is currently under way. In addition, much more research is needed to identify appropriate crops for use in water reuse systems. In particular, research is needed to identify appropriate salt-tolerant forages. In California’s SJV there is a shortage of forages, and this trend will likely continue in the near future as dairies continue to increase and pressure to remove beef cattle from foothill and mountain ecosystems escalates. Potentially suitable crops for saline reuse systems have been identified by researchers around the world, but by no means is this list complete. Much has been learned about use and reuse of saline-sodic water over the past three decades, but much more can be learned due to the wide range of conditions affecting salt-affected lands around the world. Although the same set of scientific principles applies in all cases, there is no one management practice that uses saline-sodic water that will be appropriate in all areas and appropriate to every farmer or farming operation. Rather, use and reuse of saline-sodic water will have to be customized to site-specific conditions to be sustainable. REFERENCES Aronson, J. A. (1989). HALOP: A data base of salt tolerant plants of the world, Arid Land Studies, University of Arizona, Tucson, Ariz. Ayars J. E., and Basinal, L. (2005). A technical advisor’s manual. Managing agricultural irrigation drainage water: A guide for developing integrated on-farm drainage management systems, Westside Resources Conservation District and Center for Irrigation Technology, California State University, Fresno, Calif.

712

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Ayars, J. E., Hutmacher, R. B., Hoffman, G. J., Ben-Asher, J., and Pflaum, T. (1990). “Response of sugar beet to nonuniform irrigation.” Irrig. Sci., 11, 101–109. Ayars, J. E., Hutmacher, R. B., Schoneman, R.A., Vail, S. S., and Pflaum, T. (1993). “Long-term use of saline water for irrigation.” Irrig. Sci., 14, 27–34. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Bañuelos, G. S., Sharmarsakar, S., Cone, D., and G. Stuhr. (2003). “Vegetative approach for improving the quality of water produced from soils in the westside of central California.” Plant Soil, 249, 229–236. Bari, M. A., and Schofield, N. J. (1992). “Lowering of a shallow, saline water table by extensive eucalypt reforestation.” J. Hydrol., 133, 273–291. Bassil, E., and Kaffka, S. R. (2002a). “Response of safflower (Carthamus tinctorius L.) to saline soils and irrigation. I. Consumptive water use.” Agric. Water Mgmt., 54, 67–80. ———. (2002b). “Response of safflower (Carthamus tinctorius L.) to saline soils and irrigation. II. Crop growth and yield.” Agric. Water Mgmt., 54, 81–92. Benes S., Grattan S., Finch C., and Basinal, L. (2004). “Plant selection for IFDM,” in A landowner’s manual: Managing agricultural irrigation drainage water. A guide for developing integrated on-farm drainage management systems, T. Jacobsen and L. Basinal, eds., Chapter 6, California State Water Resources Control Board, Hudson Orth Communications, Fresno, Calif., 1–21. Benes, S. E., Grattan, S. R., and Robinson, P. H. (2005). “Cultivation of halophytes to reduce drainage volumes on the westside San Joaquin Valley of California,” in Final report to the California State University Agricultural Research Institute (ARI), Project No. 00-1-003, Oct. 18, 2005, California State University, Fresno, Calif. Boyko, H., and Boyko, E. (1959). “Seawater irrigation, a new line of research on a bioclimatic plant-soil complex.” Int. J. Bioclim., 3, 1–17. Bradford, S., and Letey, J. (1992). “Cyclic and blending strategies for using nonsaline and saline waters for irrigation.” Irrig. Sci., 13, 123–128. Brown, P. H., and Shelp, B. J. (1997). “Boron mobility in plants.” Plant Soil, 193, 85–101. Cervinka, V. (1994). “Agroforestry farming system for the management of selenium and salt on irrigated farmland,” in Selenium in the environment, W. T. Frankenberger, Jr. and S. Benson, eds., Marcel Dekker, Inc., New York, 237–250. Cervinka, V., Diener, J., Erickson, J., Finch, C., Martin, M., Menezes, F., Peters, D., and Shelton, J. (1999). Integrated system for agricultural drainage management on irrigated farmland, Final Report (Oct. 99), Grant 4-FG-20-11920, U.S. Dept. of the Interior, Bureau of Reclamation, Washington, D.C. Chabbra, R. (1996). Soil salinity and water quality, A. A. Balkema Publishers, Brookfield, Vt. Corwin, D. L., Kaffka, S. R., Oster, J. D., Hopmans, J., Mori, Y., van Groenigen, J. W., and van Kessel, C. (2003). “Assessment and field-scale mapping of soil quality properties of a saline-sodic soil.” Geoderma, 114, 231–259. Corwin D. L., Lesch, S. M., Oster J. D., and Kaffka, S. R. (2006). “Monitoring management-induced spatio-temporal changes in soil quality through soil sampling directed by apparent electrical conductivity.” Geoderma, 131(3/4), 369–387.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

713

Dinar, A., Letey, J., and Vaux, H. J., Jr. (1986). “Optimal ratios of saline and nonsaline irrigation waters for crop production.” Soil Sci. Soc. Am,. J., 50(2), 440–443. Epstein, E., Norlyn, J., Rush, D., Kingsbury, R., Kelly, D., Cunningham, G., and Wrona, A. (1980). “Saline culture of crops: A genetic approach.” Science, 210, 399–404. Ferreyra, R. E., Alijaro, A. U., Ruiz, R. S., Rojas, L. P., and Oster, J. D. (1997). “Behavior of 42 crop species grown in saline soils with high boron concentrations.” Agric. Water Mgmt., 32, 111–124. Flowers, T., and Yeo, A. (1995). “Breeding for salinity resistance in crop plants: Where next?” Aust. J. Plant Physiol., 22, 875–884. Frenkel, H., and Shainberg, I. (1975). “Irrigation with brackish water: Chemical and hydraulic changes in soils irrigated with brackish water under cotton production,” in Proc. Irrigation with Brackish Water, International Symposium, BeerShevos, Israel, Negev University Press, Jerusalem, 175–183. Glenn, E. P., Brown, J. J., and Blumwald, E. (1999). “Salt tolerance and crop potential as halophytes.” Crit. Rev. Plants Sci., 227–255. Glenn, E. P., Brown, J. J., and O’Leary, J. W. (1998). “Irrigating crops with seawater.” Scientif. Am., August, 76–81. Glenn, E., Miyamoto, M., Moore D., Brown, J. J., Thompson, T. L., and Brown, P. (1997). “Water requirements for cultivating Salicornia bigelovii Torr. with seawater on sand in a coastal desert environment.” J. Arid Environ., 36, 711–730. Grattan, S. R., Benes, S. E., Diaz, F., and Peters, D. W. (2008). “Feasibility of irrigating Salicornia bigelovii Torr. with hyper-saline drainage water.” J. Environ. Qual., 37, S149–S156. Grattan, S. R., Benes, S. E., Peters, D. W., and Mitchell, J. P. (1999). “Potential suitability of the halophyte Salicornia bigelovii as the final crop in a drainage water reuse sequence,” in Proc., 17th International Congress on Irrigation and Drainage, September 11–19, Granada, Spain, Special Session R-8, 107–120. Grattan, S. R., and Grieve, C. M. (1999). “Salinity–mineral nutrient relations in horticultural crops.” Sci. Hort., 78, 127–157. Grattan, S. R., Grieve, C. M., Poss, J. A., Robinson, P. H., Suarez, D. L., and Benes, S. E. (2004a). “Evaluation of salt-tolerant forages for sequential water reuse systems. I. Biomass production.” Agric. Water Mgmt., 70, 109–120. ———. (2004b). “Evaluation of salt-tolerant forages for sequential water reuse systems. III. Potential implications for ruminant mineral nutrition.” Agric. Water Mgmt., 70, 137–150. Grattan, S. R., and Oster, J. D. (2003). “Use and reuse of saline-sodic waters for irrigation of crops,” in Crop production in saline environments: Global and integrative perspectives, S. S. Goyal, S. K. Sharma, and D. W. Rains, eds., Haworth Press, New York, 131–162. Grattan, S. R., Shannon, M. C., Grieve, C. M., Poss, J. A., Suarez, D. L., and Francois, L. E. (1996). “Interactive effects of salinity and boron on the performance and water use of eucalyptus.” Acta Hort., 449, 607–613. Grattan, S. R., Shennan, C., May, D. M., Mitchell, J. P., and Burau, R. G. (1987). “Use of drainage water for irrigation of melons and tomatoes.” Calif. Agric., 41, 27–28. Greene, R. S. B., and Wilson, I. B. (1989). “Amelioration of some physical properties and nutrient availability of an exposed B horizon of a red-brown earth.” Soil Use and Mgmt., 5, 66–71.

714

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Greig, J. (1996). “Saltland’s Cinderellas,” in Proc. 4th National Conference and Workshop on the Productive Use and Rehabilitation of Saline Lands, Albany, Western Australia, March 25–30, 1996, Promaco Conventions Pty Ltd., Canning Bridge, Western Australia, 293–296. Grieve, C. M., Poss, J. A., Carter, C. T., and Shouse, P. J. (2008). Modeling growth of Matthiola incana in response to saline wastewaters differing in nitrogen level. Hort. Sci., 43, 1787–1793. Grieve, C. M., Poss, J. A., Grattan, S. R., Shouse, P. J., Lieth, J. H., and Zeng, L. (2005). “Productivity and mineral nutrition of Limonium species irrigated with saline wastewaters.” Hort. Sci., 40, 654–658. Grieve, C. M., and Suarez, D. L. (1997). “Purslane (Portulaca oleraceae L): A halophytic crop for drainage water reuse systems.” Plant Soil, 192, 277–283. Hardan, A. (1976). “Irrigation with saline water under desert conditions,” in Proc. International Salinity Conference, Managing Saline Water for Irrigation, Texas Tech University, Lubbock, Texas, August 1976, Texas Tech University, Lubbock, Tex., 165–169. International Center for Biosaline Agriculture (ICBA). (2005). Annual report 2005 (1425-26H), ICBA, Dubai, UAE. James, D. W., Jackson, T. L., and Harvard, M. E. (1968). “Effect of molybdenum and lime on the growth and molybdenum content of alfalfa grown on acid soils.” Soil Sci., 105(6), 397–402. Jayawardane, N. S., and Chan, K. Y. (1994). “The management of soil physical properties limiting crop production in Australian sodic soils: A review.” Aust. J. Soil Res., 32, 13–44. Jury, W. A., Tuli, A. and Letey, J. (2003). “Effect of travel time on management of a sequential reuse drainage operation.” Soil Sci. Soc. Am. J., 67, 1122–1126. Kaffka, S. R., Daxue, D., and Peterson, G. (1999). “Saline water can be used to irrigate sugarbeets, but sugar may be low.” Calif. Agric., 53(1), 11–15. Kaffka, S. R. and Hembree, K. (2004). “The effects of saline soil and irrigation on sugarbeet stand establishment.” J. Sugar Beet Res., 41(3), 61–72. Kaffka, S. R., Oster, J. D., and Corwin, D. L. (2002). “Using forages and livestock to manage drainage water in the San Joaquin Valley,” in Proc., 17th World Congress of Soil Science, August 14–21, 2002, Bangkok, Thailand, I. Kheoruenrome, ed., Symposium No. 34, 2059-1–2059-12. Kaffka, S., Oster, J., Maas, J., and Corwin, D. (2004). “Forage production and soil reclamation using saline drainage water,” in Proc., 2004 National Alfalfa Symposium and the 34th CA Alfalfa Symposium, December 13–15, 2004, San Diego, California, Dept. of Agricultural and Range Science Cooperative Extension, University of California-Davis, Davis, Calif., 247–253. Keren, R., and Shainberg, I. (1978). “Irrigation with sodic and brackish water and its effect on the soil and on cotton fields.” Harrade, 58, 963–976. Kubota, J., and Allaway, W. H. (1972). “Geographic distribution of trace element problems,” in Micronutrients in agriculture, J. J. Mordtvedt, P. M. Giordano, and W. L. Lindsay, eds., SSSA, Madison, Wisc., 525. Läuchli, A., and Grattan, S. R. (2007). “Plant growth and development under salinity stress,” in Advances in molecular breeding toward drought and salt tolerant crops, M. A. Jenks, P. M. Hasegawa, and S. M. Jain, eds., Springer-Verlag, Dordrecht, The Netherlands, 1–32.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

715

Lesch, S. M., Rhoades, J. D., and Corwin, D. L. (2000). ESAP-95 version 2.01R: User manual and tutorial guide, Research Report No. 146, G. E. Brown Jr., ed., U.S. Department of Agriculture, George E. Brown Jr. Salinity Laboratory, Riverside, Calif. Letey, J., and Dinar, A. (1986). “Simulated crop-production functions for several crops when irrigated with saline waters.” Hilgardia, 54, 1–32. Letey, J., and Feng, G. L. (2007). “Dynamic versus steady-state approaches to evaluate irrigation management of saline waters.” Agric. Water Mgmt., 98(4), 502–506. Levy G. J., Shainberg, I. and Miller, W. P. (1998). “Physical properties of sodic soils,” in Sodic soils: Distribution, properties, management and environmental consequences, M. E. Sumner and R. Naidu, eds., Oxford University Press, New York, 51–75. Maas, E. V., and Grattan, S. R. (1999). “Crop yields as affected by salinity,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Monograph 38, ASA, Madison, Wisc., 55–108. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. Drain. Div., ASCE, 103(IR2), 115–134. Mamedov, A. I., Levy, G. J., Shainberg, I., and Letey, J. (2001). “Wetting rate, sodicity, and soil texture effects on infiltration rate and runoff.” Aust. J. Soil Res., 39, 1293–1305. Mason, J. (1990). “The biochemical pathogenesis of molybdenum-induced copper deficiency syndromes in ruminants: Towards the final chapter.” Irish Vet. J., 43(1), 18–20. Masters, D. G., Benes, S. E., and Norman, H. (2007). “Biosaline agriculture for forage and livestock production.” Ag. Ecosyst. Environ., 119, 234–248. McNeal, B. L., and Coleman, N. T. (1966). “Effect of solution composition on soil hydraulic conductivity.” Soil Sci. Soc. Am. J., 30, 313–317. Miles, D. L. (1977). Salinity in the Arkansas Valley of Colorado, Interagency Agreement Report EPA-Iowa G-D4-0544, U.S. Environmental Protection Agency, Denver, Colo. Minhas, P. S., and Sharma, O. P. (2003). “Management of soil salinity and alkalinity problems in India,” in Crop production in saline environments: Global and integrative perspectives, S. S. Goyal, S. K. Sharma, and D. W. Rains. eds., Haworth Press, New York, 181–230. Mitchell, J. P., Shennan, C., Grattan, S. R., and May, D. M. (1991). “Tomato fruit yields and quality under water deficit and salinity.” J. Amer. Soc. Hort. Sci., 116(2), 215–221. Mitchell, J. P., Shennan, C., Singer, M. J., Peters, D., Miller, R. O., Prichard, T., Grattan, S. R., Rhoades, J. D., May, D. M., and Munk, D. (2000). “Impacts of gypsum and winter cover crops on soil physical properties and crop productivity when irrigated with saline water.” Agric. Water Mgmt., 45(1), 55–71. Moore, J., and Hefner, J. V. (1976). “Irrigation with saline water in the Pecos Valley of west Texas,” in Proc. International Salinity Conference, Managing Saline Water for Irrigation, Texas Tech University, Lubbock, Texas, August, 1976, Texas Tech University, Lubbock, Tex., 339–344. Morris, J. D., and Thomson, L. A. J. (1983). “The role of trees in dryland salinity control.” Proc. R. Soc. Vict., 95(3), 123–131.

716

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Mudie, P. (1974). “The potential economic uses of halophytes,” in Ecology of halyphytes, R. Reimold and W. Queen, eds., Academic Press, New York, 565–597. National Research Council (NRC). (1990). Saline agriculture: Salt-tolerant plants for developing countries, National Academies Press, Washington, D.C. ———. (1996). Nutrient requirements of beef cattle, 7th ed., National Academies Press, Washington, D.C. ———. (2001). Nutrient requirements of dairy cattle, 7th ed., National Academies Press, Washington, D.C. Nelson, P. N., Barzegar, A. R., and Oases, J. M. (1997). “Sodicity and clay type: Influence on decomposition of added organic matter.” Soil Sci. Soc. Am J., 61, 1052–1057. Ohlendorf, H. M. (1989). “Bioaccumulation and effects of selenium in wildlife,” in Selenium in agriculture and the environment, L. W. Jacobs, ed., SSA Special Publication No. 23, SSA, Madison, Wisc., 133–177. Oster, J. D. (1994). “Irrigation with poor quality water.” Agric. Water Mgmt., 25, 271–297. Oster, J. D., and Jayawardane, N. S. (1998). “Agricultural management of sodic soils,” in Sodic soils: Distribution, processes, management and environmental consequences, M. E. Summer and R. Naidu, eds., Chapter 8, Oxford University Press, New York, 125–147. Oster, J. D., Kaffka, S.R., Shannon, M. C., and Grattan, S. R. (1999b). “Saline-sodic drainage water: A resource for forage production?” Proc. of 17th International Congress on Irrigation and Drainage, Vol. 1F, Granada, Spain, 67–79. Oster, J. D., Macedo, T. F. Davis, D., and Fulton, A. (1999a). Developing sustainable reuse and disposal of saline drainwater on eucalyptus (E. camaldulensis), Final Report, Dept. Environmental Science, University of California, Riverside, Calif., and Tulare Lake Drainage District, Corcoran, Calif., U.S. Department of the Interior, Bureau of Reclamation, and Tulare Lake Drainage District. Oster, J. D., and Schroer, F. W. (1979). “Infiltration as influenced by irrigation water quality.” Soil Sci. Am. J., 43, 444–447. Oster, J. D., and Shainberg, I. (2001). “Soil responses to sodicity and salinity: Challenges and opportunities.” Aust. J. Soil Res., 39(6), 1219–1224. Oster, J. D., Shainberg, I., and Abrol, J. P. (1999b). “Reclamation of salt-affected soil,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Agronomy Monograph 38, ASA/CSSA/SSSA, Madison, Wisc., 659–691. Oster, J. D., Singer, M. J., Fulton, A., Richardson, W., and Prichard, T. (1992). Water penetration problems in California soils: Prevention, diagnoses and solutions (revised 1993), Kearney Foundation of Soil Science, DANR, University of California. Osweiler, G. D., Carson, T. L., Buck, W. B., and van Gelder, G. A. (1985). Clinical and diagnostic veterinary toxicology, 3rd ed., Kendall/Hunt Publishers, Dubuque, Iowa, 87–103, 132–142. Pasternak, D., DeMhalach, Y., and Borovic, I. (1986). “Irrigation with brackish water under desert conditions VII. Effect of time of application of brackish water on production of processing tomatoes (Lycopersicon esculentum Mill.).” Agric. Water Mgmt., 12, 149–158.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

717

Phillippo, M., Humphries, W. R., Atkinson, T., Henderson, G. D., and Garhwaite, P. H. (1987). “The effect of dietary molybdenum and iron on copper status, puberty, fertility and oestrus cycles in cattle.” J. Agric. Sci., 109, 321–336. Poss, J. A., Grattan, S. R., Grieve, C. M., and Shannon, M. C. (1998). “Characterization of leaf boron injury in salt-stressed Eucalyptus by image analysis.” Plant Soil, 206, 237–245. Quirk, J. P. (1978). “Some physico-chemical aspects of soil structural stability: A review,” in Modification of soil structure, W. W. Emerson, R. D. Bond, and A. R. Dexter, eds., John Wiley and Sons, New York. Quirk, J. P., and Schofield, R. K. (1955). “The effect of electrolyte concentration on soil permeability.” J. Soil Sci., 6, 163–178. Rains, D. W., Goyal, S., Weyrauch, R., and Läuchli, A. (1987). “Saline drainage water reuse in a cotton rotation system.” Calif. Agric., 41, 24–26. Rhoades, J. D. (1984). “Use of saline water for irrigation.” Calif. Agric., 38(10), 42–43. ———. (1987). “Use of saline water for irrigation.” Water Qual. Bull., 12, 14–20. Rhoades, J. D., Bingham, F. T., Letey, J., Dedrick, A. R., Bean, M., Hoffman, G. J., Alves, W. J., Swain, R. V. Pacheco, P. G., and Lemert, R. D. (1988). “Reuse of drainage water for irrigation: Results of Imperial Valley study.” Hilgardia, 56, 1–45. Rhoades, J. D., Bingham, F. T., Letey, J., Hoffman, D., Pinter, A. R., Alves, W., Swain, R., Pacheco, P., Lemert, R., and Replogle, J. A. (1989). “Use of saline drainage water for irrigation: Imperial Valley study.” Agric. Water Mgmt., 16, 25–36. Robinson, P. R., Grattan, S. R., Getachew, G., Grieve, C. M., Poss, J. A., Suarez, D. S., and Benes, S. E. (2004). “Biomass accumulation and potential nutritive value of some forages irrigated with saline-sodic drainage water.” Animal Feed Sci. Tech., 111, 175–189. Rolston, D. E., Rains, D. W., Biggar, J. W., and Läuchli, A. (1988). “Reuse of saline drain water for irrigation.” Paper presented at UCD/INIFAP conference, Guadalajara, Mexico, March 1988. Rosenfeld, I., and Beath, O. A. (1964). Selenium: Geobotany, biochemistry, toxicity and nutrition, Academic Press, New York. San Joaquin Valley Drainage Program (SJVDP). (1990). A management plan for agricultural subsurface drainage and related problems on the westside San Joaquin Valley: Final report of the San Joaquin Valley Drainage Program, U.S. Dept. of the Interior and California Resources Agency, Sacramento, Calif. Sanden, B. L., Ferguson, L., Reyes, H. C., and Grattan, S. R. (2004). “Effect of salinity on evapotranspiration and yield of San Joaquin Valley pistachios.” Acta Hort., 664, 583–589. Shainberg, I., and Letey, J. (1984). “Response of soils to sodic and saline conditions.” Hilgardia, 52, 1–57. Shainberg I., Levy G. J., Goldstein, D., Manedov, A. I., and Letey, J. (2001). “Prewetting rate and sodicity effects on the hydraulic conductivity of soils.” Aust. J. Soil Res., 39, 1279–1291. Shalhevet, J. (1984). “Management of irrigation with brackish water,” in Soil salinity under irrigation: Processes and management, I. Shainberg and J. Shalhevet, eds., Springer-Verlag, New York, 298–318.

718

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1994). “Using water of marginal quality for crop production: Major issues.” Agric. Water Mgmt., 24, 233–269. Shani, U., Ben-Gal, A., and Dudley, L. (2005). “Environmental implications of adopting a dominant factor approach to salinity management.” J. Environ. Quality, 34(5), 1455–1460. Shannon, M. C., Grieve, C. M., Wilson, C., Poss, J., Suarez, D. L., Lesch, S., and Rhoades, J. D. (1998). Growth and water relations of plant species suitable for saline drainage water reuse systems, final report to California Dept. of Water Resources, Project DWR B-59922, California Dept. of Water Resources, Sacramento, Calif. Sharma, D. P., and Tyagi, N. K. (2004). “On-farm management of saline drainage water in arid and semi-arid regions.” Irrig. and Drain., 53, 87–103. Shennan, C., Grattan, S. R., May, D. M., Hillhouse, C. J., Schactman, D. P., Wander, M., Roberts, B., Burau, R. G., McNeish, C., and Zelinski, L. (1995). “Feasibility of cyclic reuse of saline drainage in a tomato-cotton rotation.” J. Environ. Qual., 24, 476–486. Suarez, D. L., Wood, J. D., and Lesch, S. M. (2006). “Effect of SAR on water infiltration under a sequential rain-irrigation management system.” Agric. Water Mgmt., 86, 150–164. Suttle, N. F. (1991). “The interactions between copper, molybdenum and sulphur in ruminant nutrition.” Ann. Rev. Nutr., 11, 121–140. Suyama H., Benes, S. E., Robinson, P. H., Getachew, G., Grattan S. R., and Grieve, C. M. (2007a). “Biomass yield and nutritional quality of forage species under long-term irrigation with saline-sodic drainage water: Field evaluation.” Animal Feed Sci. Technol., 135, 329–345. Suyama H., Benes, S. E., Robinson, P. H., Grattan, S. R., Grieve, C. M., and Getachew, G. (2007b). “Forage productivity, and quality under irrigation with saline-sodic drainage water: Greenhouse evaluation.” Ag. Water Mgmt., 88, 159–172. Tanji, K. K., Valoppi, L., and Woodring, R. C., eds. (1988). Selenium contents in animal and human food crops grown in California, ANR Publication 3330, University of California, Oakland, Calif. Taylor, A. J., and Olsson, K. A. (1987). “Effect of gypsum and deep ripping on lucerne (Medicago sativa L.) yields on a red-brown earth under flood and spray irrigation.” Aust. J. Exp. Agr., 27, 841–849. Valoppi, L., and Tanji, K. K. (1988). “Are the selenium levels in food crops and waters of concern?” in Selenium contents in animal and human food crops grown in California, K. K. Tanji, L. Valoppi, and R. C. Woodring, eds., DANR Publication 3330, University of California, Oakland, Calif. van Schilfgaarde, J. (1990). “Irrigated agriculture: Is it sustainable?” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 584–594. Velarde, M., Felker, P., and Degano. C. (2003). “Evaluation of Argentine and Peruvian Prosopis germplasm for growth at seawater salinities.” J. Arid Envi., 55, 515–531. Ward, G. M. L. (1978). “Molybdenum toxicity and hypocuprosis in ruminants: A review.” J. Animal Sci., 46, 1078–1085.

USE OF SALINE DRAINAGE WATERS FOR IRRIGATION

NOTATION A  yield threshold soil salinity above which yields decline B  percentage yield loss per increase in salinity in excess of A EC  electrical conductivity ET  evapotranspiration HC  hydraulic conductivity IR  infiltration rate LF  leaching fraction LR  leaching requirement RY  relative yield SAR  sodium adsorption ratio

719

This page intentionally left blank

CHAPTER 23 DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS Jose I. Faria and James Poss

INTRODUCTION Treating and disposing of subsurface drainage water from irrigated agricultural lands presents unique technical challenges. Complex chemical characteristics of drainage water further complicate treatment options. Treatment of drainage water has long been considered one of the lastresort drainage management options due to its high costs. However, this perspective is changing due to the increasing demand for fresh water caused, in part, by increases in population, industrialization, and agricultural activities in addition to environmental restrictions imposed on state and federal water supply and distribution systems. Additionally, treatment system costs and energy requirements for the most promising treatment technologies have decreased over the last two decades. For subsurface drainage water containing extremely high levels of salinity, selenium (Se), molybdenum (Mo), and other trace elements, the treatment objectives are as follows: 1. Meet agricultural water management goals. 2. Reduce salt and toxic constituents below hazardous levels. 3. Meet water quality objectives in surface waters. 4. Reduce trace elements below hazardous concentrations for wildlife. This chapter will cover the treatment and disposal of subsurface drainage from irrigated lands. The current status of the technology of drainage water treatment and disposal options will be reviewed, and current research on treatment technology will be discussed.

721

722

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TREATMENT OF AGRICULTURAL DRAINAGE WATER Treatment Technology Needs Substantial technological advancements in treatment of agricultural drainage water have been made since the first edition of this manual was published in 1990. Long before the Se problem emerged, drainage water reclamation was being seriously considered at the tubular RO plant in Firebaugh in the early 1970s. The motivation for construction of this experimental facility arose from two fundamental issues relevant to agricultural drainage. Of primary concern was augmentation of irrigation water supplies by drainage water desalinization. This was indeed a challenging application of the emerging technology under development at the time. A second goal was directed toward reduction of drainage water volume. Management of salt accumulation could then be enhanced by such waste minimization technology. The management of drainage from irrigated lands is an important part of any agricultural development plan. History has recorded many instances of fertile lands subsequently made barren by salt (FAO 1973). More recently, traditional approaches to sustaining and optimizing agricultural productivity by salinity control have been complicated by the need to protect public health and the environment from any potential effects of residual fertilizers, pesticides, herbicides, fungicides, and trace toxic substances in drainage from irrigated lands. The environmental impacts of Se and other trace elements in agricultural drainage water was not fully recognized until about 1985, when high levels of Se were identified in biota in Kesterson Reservoir in the San Joaquin River basin, California. The alarm was raised when dead and deformed birds were found at the reservoir (Ohlendorf 1984). Consequently, the need to develop a technology that will adequately control and manage drainage from irrigated lands was recognized. The treatment of agricultural drainage water effluent presents a challenge due to the complex chemical characteristics of most drainage waters (Lee 1994). One of the major challenges for treating drainage water in the San Joaquin Valley is that typical drainage water is saturated with calcium sulfate (CaSO4). Earlier attempts to use membrane technology to desalinate agricultural drainage water have failed entirely or have been limited to low recovery (50%) because of CaSO4 fouling problems. Table 23-1 presents a summary of analyses for drainage water taken from three locations within the western San Joaquin Valley. These waters represent water sources utilized by the various treatment demonstration projects described in this chapter. A variety of processes can be used to treat seawater, brackish, or waste waters to meet industrial, urban, and drinking water standards. Many of

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

723

TABLE 23-1. Drainage Water Quality Analysis/Various Agricultural Drainage Sumps Locations, Western San Joaquin Valley, California Parameter

Buena Vista mg/l

Westlands mg/l

Northerly Area mg/l

Sodium Potassium Calcium Magnesium Total Hardness Chloride Carbonate Bicarbonate Sulfate Nitrate (as N) Boron Selenium Silica

1400 4 630 100 2000 2400 1 320 1350 19 5 0.04 36

2200 7 560 270 2500 1600 4 200 4700 48 15 0.23 37

600 9 290 93 1100 550 4 170 1500 14 9 0.07 22

TDS

6000

9900

3300

Source: California Department of Water Resources.

these processes could potentially be applied to the treatment of agricultural drainage water. Treatment processes for drainage water can be categorized into those that reduce the total salinity of the drainage water and those that remove specific trace elements. Most desalinization processes also remove trace elements, but their costs are often prohibitive. Less costly methods for the removal of trace elements are being developed. Methods for the removal of trace elements can be biological, physical, or chemical. In an earlier review, Lee (1994) described available drainage water treatment and disposal technologies. The San Joaquin Valley Drainage Implementation Plan (SJVDIP 1999b) also reviewed treatment technologies for removing Se from agricultural drainage water. The next section summarizes their findings. Desalinization The numerous desalinization processes include ion exchange, thermal distillation, electrodialysis, and reverse osmosis (RO). Of these processes,

724

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

RO is considered to be the most promising for the treatment of agricultural drainage water, mainly due to its comparatively low cost (Tanji and Neeltje 2002). Thermal distillation may be an attractive option if a source of low-cost heat is available, such as waste heat from a power plant facility. Reverse osmosis is a process capable of removing different contaminants, including dissolved salts and organics. In RO, a semipermeable membrane separates water from dissolved salts and other suspended solids. Pressure is applied to the feed-water, forcing the water through the membrane and leaving behind salts and suspended materials in a brine stream. The energy consumption of the process depends on the salt concentration of the feed-water and the salt concentration of the effluent. Depending on the quality of the water to be treated, pretreatment might be crucial to preventing fouling of the membrane. Pretreatment steps could include multiple filtration, addition of antiscalants, pH corrections, and lime treatment, along with ion exchange. Following is a brief description of the most important desalination efforts performed with subsurface agricultural drainage water.

Firebaugh Water District The first attempt at drainage water reclamation began in 1971 in Firebaugh, California (McCutchan et al. 1976). A small membrane desalinization pilot plant utilizing hand-cast cellulose acetate tubular membranes was designed and built at the UCLA School of Engineering and Applied Science. The plant remained on-line for approximately 3 years and was operated jointly by UCLA and the California Department of Water Resources (CDWR). Water quality at this site varied in TDS levels between 2,000 and 7,000 mg/L, and calcium and sulfate ion concentrations were near saturation with respect to gypsum. A limiting issue in processing this water was the potential deposition of scale-forming calcium sulfate (gypsum) on membrane surfaces. Scale control was investigated first by treatment with sodium hexametaphosphate, followed by installation of a cation exchange system for calcium removal. Product water recovery based on chemical and ion-exchange treatment was reported at 60% and 90%, respectively.

Los Baños (California) Demonstration Desalting Facility As a continuation of the reclamation program started in Firebaugh, between 1982 and 1985 the CDWR conducted a pilot-plant-scale demonstration of RO of saline drainage water using cellulose acetate membranes. Throughout the studies, bacterial and chemical fouling of the

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

725

membrane were major problems. As a pretreatment, the drainage water was treated with potash alum (KAl(SO4)2 12H2O) and passed through a solids-reactor clarifier; it was then chlorinated, filtered, and processed through an ion-exchange (IX) system using a strong-acid resin for calcium removal before being desalted by RO. The highly concentrated Na RO reject was used to regenerate the IX resin. The spent IX regenerant was concentrated further for use in the solar pond operations at the facility. In spite of this level of pretreatment, the membranes tended to foul due to the precipitation of gypsum and calcite. The permeate is the product (desalted) water and the concentrate is the brine water. The results show that TDS can be desalted from 9,800 to 640 ppm, boron (B) from 14.5 to 7.6 ppm, and Se from 325 to 3 ppb in a three-stage RO system. The efficiency of removal declines with each stage. The drainage water was saturated with respect to calcite and gypsum. Other desalting processes tested at the facility included electrodialysis reversal and vapor-compression evaporation. These processes also experienced scaling issues. In addition, a vertical fluted-tube foamy evaporator (VTFE) was tested in solar pond operations that used the pond’s hot brine heat as the driving force for evaporation. Heat transfer rates ranged from 500 to 800 Btu/hr-ft2-F during the course of testing in the VTFE mode (CDWR 1986). Buena Vista Water Storage District From 2000 through 2002, state, local, and private entities collaborated in a project to investigate treatment costs, identify and resolve drainage water treatment issues, and demonstrate the ability of commercially available RO membranes to treat agricultural drainage at the Buena Vista Water Storage District in Kern County, California. In 2000 a 20-gpm RO unit was operated to treat tile-drain water. However, the drainage feedwater was switched to shallow groundwater because of the lack of a sufficient volume of drainage water due to the reduction of irrigation allocations during 2001 and 2002. Raw tile-drainage TDS concentrations averaged 7,010 mg/L, while the shallow groundwater, pumped from two wells, 60 and 80 ft deep, averaged 3,980 mg/L. The overall TDS removal was 97% throughout the operation of the project. Treatment costs were estimated to range from $651/acre-ft for a 1 million gallon per day (MGD) plant to $459/acre-ft for a 10 MGD plant, irrespective of costs to collect and transport saline water to the plant or the cost to dispose of the concentrated reject brine. A final report prepared by Boyle Engineering Corp. of Bakersfield, California was published in December 2003 (Boyle Engineering et al. 2003). Figure 23-1 outlines the process flow diagram used for this desalination project.

726

FIGURE 23-1. Process flow diagram for Buena Vista Water Storage District demonstration desalting. From Boyle Engineering Corp. (2003).

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

727

Panoche Drainage District Through the Energy Innovation Small Grant (EISG) program, the California Energy Commission sponsored a pilot demonstration project to demonstrate technical feasibility of a conceptual RO process. The patented double-pass preferential-precipitation reverse osmosis (DP3RO™) process is a two-pass membrane process that induces preferential precipitation of calcium sulfate in the first pass. The first pass employs tubular membranes. Treated water costs, as stated in the 2003 final report (Enzweiler and Strasser 2003), are estimated to range between $564/acre-ft and $801/acre-ft for drainage that ranged from 3,625 mg/L to 10,500 mg/L TDS concentration. The DP3RO™ process was periodically operated at Panoche Drainage District’s (PDD) DP25 test site from April 2005 through April 2006 (USBR 2008a). Continuous testing did not extend more than a week at a time due to equipment and programming constraints. In addition, due to the location of Panoche’s test site, remotely monitoring and operating the system was difficult. Despite many problems, the system was operable for a total of 30 days, of which 23 were trouble-free. During this trouble-free period, the system operated with a greater than 80% recovery rate.

Westlands Water District From 2003 through 2005, state, federal, and local entities investigated RO drainage water treatment at Red Rock Ranch (RRR) near Five Points, California, and the Panoche Drainage District (PDD) near Firebaugh, California. The 2003 study at RRR consisted of the operation of a pilot-scale membrane unit that tested RO and nanofiltration membranes to evaluate cost and performance in the treatment of agricultural drainage water. In 2004 the pilot membrane test unit was moved to the DP25 test site in PDD. To further develop technical feasibility and costs, testing at PDD took place during two phases; Phase I occurred from August 2004 to December 2004, and Phase II occurred from August 2005 to December 2005. After the initial treatment process, the concentrate reject brine stream was further treated to remove Se and nitrates by a new biotreatment technology using bioreactors. Testing indicated that operating the RO unit at a recovery greater than 50% was not practical when treating the concentrate reject brine for Se and other constituents due to the propensity of the reject to precipitate calcium sulfate. This condition was shown by operating the RO unit at a recovery of 64% using a single antiscalant at the beginning of Phase II from August 17 to October 13, 2005. Reducing the recovery to 55% along with the changing to antiscalant to a mixture of two antiscalants from October 26 to December 12, 2005, did not alleviate the condition. These projects provided data to

728

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

develop full-scale reverse osmosis treatment plant design parameters and cost estimates (Table 23-2) in the evaluation of drainage solution alternatives for the San Luis Unit of the Central Valley Project in California (USBR 2008b).

Desalination Studies at the University of California–Los Angeles Investigations at the Polymer and Separations Research Laboratory at the University of California–Los Angeles continued to develop the understanding of the principles that lead to scale formation resulting in membrane fouling. A series of rigorous laboratory tests were performed in problematic areas associated with membrane desalting, using simulated or actual drainage water. The work provided a fundamental understanding of the process of surface mineral scale formation and developed diagnostic tools and protocols for assessing the effectiveness of antiscalants and the propensity for membrane scaling (Rahardianto et al. 2006). Tests were performed to compare and rank antiscalant effectiveness and showed that the method was useful for assessing the impact of particle matter in the induction of mineral salt crystallization. The work prompted a detailed study on drainage water collected from five sumps of varying water quality located throughout the San Joaquin Valley. The drainage water was first analyzed for constituents; then potential biofouling assays were performed on field samples using two reference membranes. Prefiltration needs based on turbidity and silt-density index analyses were evaluated, mineral salt-scaling thresholds were determined, and mineral

TABLE 23-2 Full-scale Reverse Osmosis Treatment Plant Design Parameters and Cost Estimates for the San Luis Drainage Feature Preferred Alternative Parameter

Nominal feed flow, gpm

North Westlands

Central Westlands

South Westlands

Panache (Northerly Area)

570

1,690

1,070

11,000

15,000

11,000

14,000

6,100

Product recovery

50%

50%

50%

generally 50%

Number of vessels

19

38

24

Influent TDS (mg/L)

308

Membrane elements

114

228

144

1,848

Power (Kw-hr/year)

959,000

3,230,000

2,110,000

11,000,000

8,383

7,575

Building area (sq. ft.) Construction cost (2006) Source: USBR (2008a).

5,865 $8,000,000

$12,500,000 $10,000,000

18,560 $40,000,000

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

729

salt-scaling propensities of different water samples based on scaling experiments were compared with a low-scaling potential reference membrane (Cohen et al. 2008).

TRACE ELEMENT TREATMENT Biological Processes for Selenium Removal The concept of the algal-bacterial Se removal process is to grow microalgae in the drainage water at the expense of nitrate, and then to utilize the naturally settled algal biomass as a carbon source for native bacteria. In the absence of oxygen, the bacteria reduce the remaining nitrate to nitrogen gas and further reduce selenate to insoluble Se. The insoluble Se is then removed from the water by sedimentation in deep ponds and, as needed, by dissolved air flotation and sand filtration. Supplemental carbon sources, such as molasses, can be employed as reductant in addition to algal biomass (Lundquist et al. 1994) Anaerobic Treatment at Adams Avenue Agricultural Drainage Research Center As a result of the initial testing that occurred in the mid-1980s, development of anaerobic processes to reduce and remove drainage-water Se continued from 1990 through 1995 at the Adams Avenue Agricultural Drainage Research Center located near Tranquility, California. At one time, seven processes were operated consisting of an upflow anaerobic sludge blanket reactor (UASBR), two fluidized bed reactors (FBRs), two slow sand filters (SSFs), a packed bed reactor (PBR), and a pilot UASBR. The processes were arranged in two treatment trains and three standalone processes. The four first-stage processes achieved soluble selenium (SSe) reduction between 47% and 55% and the two second-stage processes reduced SSe to between 41% and 87%. The lowest SSe concentration obtained was 68 ug/L for the UASBR  FBR  SSF train, which was greater than the target water quality objective concentration of 5 mg/L as established by the State of California Regional Water Quality Control Board, Central Valley Region for the San Joaquin Basin. The Adams Avenue investigation identified additional specific work needed in the removal of drainage-water Se (CDWR 2004). Tulare Lake Drainage District’s Flow-Through Wetlands Project Selenium removal from drainage water using wetlands was investigated by Tulare Lake Drainage District (TLDD) from 1996 through 2001.

730

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The Flow-Through Wetlands Project investigated Se uptake and volatilization by plants and monitored the Se concentration in sediment and water. The project site was located on five acres of TLDD property near Corcoran, California, and consisted of 10 unlined cells (ponds), each 50 ft by 250 ft. Testing parameters included type of vegetation, flow rate, and pond water depth (Gao et al. 2003; Tanji and Gao 1999). Algae-Bacterial Selenium and Nitrate Removal (ABSR Process) The School of Civil and Environmental Engineering at the University of California–Berkeley tested an algal-bacterial Se removal (ABSR) process to remove nitrate and Se from drainage water. A 75 m3/day pilotscale ABSR facility was used to study the mechanisms and rates of Se and nitrate removal. Subsurface drainage was dosed with a carbon and energy source for bacteria (usually animal food-grade molasses) and then injected into a baffled and covered anoxic reduction pond. In the reduction pond, bacteria denitrify and reduce selenate to selenite, elemental Se, and bacterial-associated organic Se. Much of the reduced Se settles in the pond. Settled bacterial biomass in the reduction pond undergoes anaerobic decomposition, so the volume of solid residues increases quite slowly. Removal of the Se-containing solids should not be required for many years, possibly not even decades. Over two years, the ABSR facility at Panoche, California removed 95% of the influent nitrogen load and 80% of the influent SSe load. The addition of physical-chemical flotation and filtration processes to remove particulate Se increased total Se removal to 87%. Dozens of bacterial species have been isolated from the ABSR facility and were identified by 16S rRNA sequencing, including the prevalent Acinetobacter Johnson II/genospecies 7, Pseudomonas mendocina, and Xanthomonas maltophilia. Pure cultures of several of these bacteria have been proven to reduce selenite in the laboratory (Green et al. 2003). Planned “zero discharge” drainage management in the San Joaquin Valley will create brines that require treatment. The high salt concentration of brines may inhibit bacterial Se reduction. It was found that denitrification and selenate reduction are unaffected by NaCl in concentrations augmented up to 22 g/L or approximately the salinity of seawater. Inland seas, such as the Salton Sea, California, can average salinity upward of 46 g/L. Selenium and Nitrate Removal at Red Rock Ranch The USBR contracted with Applied Biosciences Inc. to conduct a threephased pilot testing of their patented biotreatment process named ABMet® for reduction of nitrate and Se. The testing occurred at RRR and PDD

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

731

between June 2003 and September 2006. The pilot units treated both raw agricultural drainage water and RO concentrate reject. The process developed into two stages, where the first stage primarily reduced nitrate and the second stage treated the water for Se removal. Each stage consisted of a tank filled with granular activated carbon (the bioreactor) that was coated with a biofilm to provide surface area for the ABMet® microbes (cultured bacteria). The microbes were sustained with a molasses-based nutrient (USBR 2008b). The microbes first removed nitrogen and then reduced the dissolved Se to a particulate that was captured in the biomass that accumulated at the bottom of the tank. In November 2005 the PDD feed was switched to raw drainage water from RO concentrate reject due to scaling problems in bioreactor 1 (first stage). Testing at PDD was completed on June 5, 2006, while RRR operations continued until September 20, 2006. After treating raw drainage water laden with high concentrations of Se (500 ppb to 1,000 ppb), the target effluent Se concentration of 10 ug/L was consistently met at the end of the Phase 3 at both sites. Chemical Processes Chemical treatment processes refer to the use of chemicals to remove trace elements from contaminated water. Chemicals are frequently used for industrial wastewater treatment but are not effective in agricultural drainage water due to their often-complex chemical characteristics (Lee 1994). Chemical processes have been developed for the reduction of selenate to elemental Se by means of ferrous hydroxide (Murphy 1988). Under laboratory conditions, ferrous hydroxide was able to reduce and precipitate Se by 99% in 30 minutes (Moody et al. 1988). In field studies, although 90% of the selenate was reduced, the reactor time required up to 6 hours. It appeared that dissolved bicarbonate, oxygen, and nitrate influenced the reduction process (CDWR 2004). Physical Processes Physical processes involve the adsorption of ions on natural and synthetic surfaces of active materials, including ion exchange resins. At PDD, Harza Engineering Company tested a pilot-scale treatment plant in 1985 for the removal of heavy metals using a process patented by Mayenkar (1986). The process used iron filings in flow-through beds. The principle was based on the idea that oxygen could activate the surface of the iron, which could then adsorb Se. Testing was discontinued as the beds quickly cemented with precipitates. The advantage of zero-valent iron is that it can significantly reduce the concentration of Se. This method could be used as a polishing step following microbial treatments. Where the waste

732

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

is anaerobic after microbial treatment, the formation of secondary precipitates is minimized (Harza Engineering Co. 1986). In a study combining both chemical and physical processes, Southern Illinois University evaluated Se removal from agricultural drainage water and synthetic solutions. Batch and kinetic studies were conducted on the removal of Se, and the effectiveness of various remediation materials was determined. Agricultural drainage water samples were obtained by CDWR from locations in the San Joaquin Valley and were provided to the researchers. The study showed that nano-sized zero-valent NiFe and Fe particles rapidly reduced and immobilized selenate from aqueous solutions. Nearly 100% selenate removal was obtained in 5 hours under most conditions. The data show that, at identical solids loading, the use of NiFe particles (as compared to Fe and Ni particles) accomplished 42% and 56% removal, respectively. Reduction of Se using bimetallic nano-sized NiFe particles resulted in nearly complete Se removal from agricultural drainage water samples. The presence of sulfates in the aqueous solutions decreased the degree of removal. However, sufficient removal is possible using these particles and can be used to achieve the 10 ppb U.S. Environmental Protection Agency (EPA)-mandated levels. Immobilization of selenate with barium chloride also appears to be an effective method, with the final cleanup of Se with NiFe bimetallic particles. The adsorption studies on both selenite and selenate removal showed that the commercially available sorbents, such as  alumina, alumina, and activated carbon, showed some promising results for selenite removal. However, they were found to be completely ineffective for selenate removal, which is one of the predominant Se species in the agricultural drainage water. The data also showed that  alumina provided higher selenite removal percentages (99%) as compared to alumina (94%), activated carbon (87%), and chitin (49%). The selenite removal was found to decrease with increasing initial Se (IV) concentration in the solution. Adsorption capacities of the adsorbents are reported in terms of their Langmuir adsorption isotherms. The adsorption capacity (on a unit mass basis) of the adsorbents for selenite is in the order: chitin  activated carbon  alumina   alumina. Generally, low pH of the solution resulted in favorable Se removal. Adsorption experiments at controlled pH conditions confirmed that surface charge density can have significant influence in equilibrium uptakes of these oxyanions. Modification of the carbon surface by copper cations significantly enhanced the equilibrium uptakes of both selenite and selenate. The surface modification of activated carbon resulted in up to 68% and 217% enhancement in uptakes from aqueous solutions containing 1 mg/L selenite and selenate, respectively. Similarly, the increase in selenite and selenate uptakes with the use of modified Southern Illinois University fly ash-derived char carbon (SIUF_C) was evaluated to be

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

733

240% and 80%, respectively, while those employing modified Lake of Egypt fly ash-derived char carbon (LOEF_C) showed an increase of 155% and 128%, respectively, over the as-extracted carbons. Sulfate and nitrate were observed to hinder the uptake of selenite, while chloride did not affect selenite uptake (Lalvani 2005).

DISPOSAL OPTIONS Considerations for Safe Disposal Drainage water management is normally concerned with reducing the amount of drainage water and with managing its disposal. However, this aim is more complex than it appears. Agricultural drainage of soils is practiced to maintain aeration in an otherwise waterlogged rootzone, and/or to leach excess soil salinity beyond the rootzone to sustain agricultural production. The drainage water generated must then be managed for reuse purposes for as long as it is of suitable water quality before it is finally discharged or disposed of. The discharge of drainage waters in watercourses may have impacts ranging from beneficial to deleterious, and the discharge of drainage water into wetlands, lakes, rivers, and coastal waters requires consideration of the quantity and quality that can be allowed while still maintaining appropriate environmental conditions and functions of the given water body. To sustain irrigated agriculture, maintenance of suitable soil salinity levels is the major concern when assessing the minimum drainage disposal requirements. In order to maximize source reduction, minimum leaching is required to maintain favorable salt balances in the rootzone. The minimum leaching fraction depends on the salinity of the irrigation water and the salt tolerance of the crops grown. In addressing this issue, it is important to note that maximizing source reduction is inconsistent with maximizing reuse of drainage water. Minimum disposal requirements also depend on the requirements of downstream water uses. In addition, many inland fisheries are currently threatened because of increasing water pollution, degradation of aquatic habitats, and excessive water abstraction (Barg et al. 1997; FAO 1999). With the increase in environmental awareness, there is increasing pressure to preserve sufficient water for aquatic environments, habitats, and biodiversity. Wetlands, for example, are considered to be some of the world’s most valuable ecological resources. Not only do they provide habitat for many species of plants and animals, they also perform other vital functions, such as water storage, flood prevention and mitigation, and water purification. Wetlands additionally provide economic benefits, of which fisheries, recreation, and tourism are among the most important.

734

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

These functions, values, and attributes can only be maintained if the ecological processes of wetlands continue functioning (Ramsar Convention Bureau 2000). Consequently, if the ecological processes of wetlands depend on drainage water inflow, then in order to sustain the wetlands’ functions the minimum volume of drainage discharge contributing to the wetlands will depend on the quality and quantity of the drainage water. Disposal Conditions Depending on the location, hydrology, and topography of the drainage basin (and the ecology and environmental conditions of receiving water bodies), drainage water might be suitable for discharge to open surface water bodies, such as rivers, lakes, outfall drains, and oceans. Ocean disposal is often regarded as the safest and most appropriate final disposal site for agricultural drainage water unless the drainage water contains sediments, nutrients, and other pollutants and the disposal site is in the vicinity of fragile coastal ecosystems, such as mangroves and coral reefs (FAO 2002). Under these circumstances, at the point of discharge, pollution may be of greater concern than dissolved salts. In general, oceans have significant dilution or assimilative capacity. However, this can be limited in many cases, especially in enclosed and semi-enclosed seas. Inland drainage water disposal to freshwater bodies, such as lakes and rivers, requires careful consideration. The multiple uses of rivers often require maintenance of specific water quality, and accumulation of salts and other pollutants in freshwater lakes could threaten ecosystem functions and aquatic life (FAO 2002). Several agricultural drainage-water disposal options are available, including discharge to surface waters, discharge to constructed wetlands, land application, irrigation, evaporation in ponds, evaporation in enhanced solar evaporators, concentration and evaporation in salinity-gradient solar ponds, integration into drainage management systems, development of aquaculture, and deep-well injection. Disposal to Surface Waters The primary aim of safe disposal of agricultural drainage water into freshwater bodies, such as lakes and rivers, is to protect beneficial downstream water uses. Rivers and lakes are normally multifunctional, providing water supplies for urban and industrial uses in addition to sustaining fisheries, recreation, and agriculture. Surface water resources like rivers and streams associated with delicate riparian habitats also have intrinsic ecological value. Furthermore, rivers feed lakes, floodplains, wetlands, estuaries, and bays. To protect these functions, it is necessary to determine the assimilative capacity of the river, including its tributaries and

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

735

ecosystems, to absorb salinity, sediment, and/or trace elements. It is also necessary to identify the constituents of concern in the drainage water in order to determine the overall discharge requirements. Moreover, the potential myriad of environmental impacts of the discharged water on sediments, riparian habitats, and floodplains needs to be considered. The discharge requirements should specify the maximum allowable concentration of each constituent of concern [limiting factor(s)] and the volume of drainage water (load limits) that would provide acceptable mitigation. Constituent concentrations are the water quality parameters normally used in establishing drinking water standards or for protecting the health of aquatic ecosystems. In California, the state and regional Water Quality Boards have regulatory oversight of these issues, which are outlined in Basin Plans (CVRWQCB 2008). In the San Joaquin Valley, for example, discharge limitations into the San Joaquin River include limits on the concentration of specific constituents and their loads. The concentration limitations include limits on salinity, boron (B), Se, and Mo in order to protect downstream water quality for domestic and agricultural uses and to safeguard the environment. To discourage dilution or inefficient water usage, load limits for these constituents are enforced for the discharge of irrigation return flows into the river. The assimilative capacity of the receiving water body varies by place and time and is influenced by numerous local conditions. Conditions can include upstream uses, climate, and other physical considerations. Until the mid-1990s, irrigated agriculture discharged its drainage water into the Grassland Water District (GWD). The district’s primary function is delivery of water to the landowners within its boundaries, which encompasses approximately 51,537 acres, the majority of which is in wetland habitat. The board of directors of GWD is proud of what the district has achieved, particularly with its success in securing and managing a long-term water supply to preserve and enhance one of the nation’s most valuable wildlife resource areas. The private landowners and sport hunters and fishers within the Grasslands, working with the district and other public agencies, have been responsible for the preservation and maintenance of the largest remaining freshwater marsh habitat on the Pacific flyway (Grassland Water District 2008). Since the 1980s discovery of Se avian poisoning at Kesterson Reservoir from subsurface drainage water conveyed from users of the Bureau of Reclamation’s Central Valley Project’s San Luis Unit and San Joaquin River Exchange Contractors, drainage from these users is no longer discharged into the GWD. Instead, it is discharged directly into the San Joaquin River at Mud Slough via the San Luis Drain and a bypass called the Grasslands Bypass (Fig. 23-2). Salt Slough and Mud Slough have become major contributors of TDS and Se in the lower reaches of the San Joaquin River. As a consequence,

736

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 23-2. Schematic of the Grasslands Bypass Project. From FAO (2002).

the Central Valley Regional Water Quality Control Board (CVRWQCB) has established monthly waste discharge requirements for Salt Slough and Mud Sloughs (CVRWQCB 2008). The water quality objectives were established to protect downstream water uses, as well as to export water to southern California. The salinity water quality objective for Vernalis, a benchmark station on the SJR, is 1 dS/m for a 30-day running average for the period September 1 to March 30, and 0.7 dS/m for a 30-day running average for the period April 1 to August 31. The boron objective was established from Sack Dam to the Merced River (the reach above and below Salt Mud Sloughs) at a 2.0 mg/L monthly mean for the year, with a maximum of 5.8 mg/L for a given month. The Se water quality objective was established for Salt and Mud Sloughs at 5 g/liter, 4-day average, to protect aquatic biota, with a maximum annual load of 3,632 kg to control mass emission. Exceeding the specified allowable loads of salt and Se as measured in the Sacramento River-San Francisco Bay delta results in monetary fines to the dischargers. The Se load maximum is the most difficult for irrigated agriculture to meet. Generally, rivers with adequate flows continually cleanse themselves. In most cases, lakes do not have this capacity because they may not have an outlet or because flow volume from the lake is limited. The disposal of drainage water into lakes could cause substantial long-term problems; thus, doing so involves special consideration. Numerous worldwide exam-

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

737

ples indicate that disposal of agricultural drainage water into freshwater lakes, often in combination with reduced freshwater inflow, has contributed to catastrophic environmental crises. Examples of significant effort being required to halt environmental degradation include Lake Chapala in Mexico, Lake Manchhar in Pakistan, and Lake Biwa in Japan. Evaporation Ponds In inland drainage basins without an appropriate surface water outlet, one of the few options is disposal of drainage effluent into constructed evaporation ponds. The impounded water is dissipated by evaporation, transpiration, and seepage losses. Disposal into constructed ponds is practiced worldwide, and the planning and design of evaporation ponds must consider numerous potential environmental problems including: • Waterlogging and salinization problems in adjacent areas resulting from excessive seepage losses • Salt-dust and spray to areas downwind of the pond during dry and windy periods, potentially damaging vegetation and affecting the health of humans and animals • Concentration of trace elements that might become toxic to fish and waterbirds because of bioaccumulation in the aquatic food chain. The following example highlights the opportunities and constraints of evaporation ponds in California. Evaporation Ponds in California Between 1972 and 1985, 28 evaporation ponds were constructed covering an area of about 2,800 ha and receiving about 39 million m3 annually of subsurface drainage from 22,700 ha of tile-drained fields (SJVDIP 1999a). Basin-wide, the pond area was about 12% of the total area drained. Most of the ponds were located in the Tulare subarea, a closed basin in the San Joaquin Valley. The salt concentration in the waters discharged into these ponds ranged from 6 to 70 dS/m with an annual salt load of 0.88 million tons, about 25% of the annual salt load accumulating in the more than 0.9 million ha of cropland. The concentration range of Se, the principal constituent of concern, was from 1 to more than 600 ppb. An Se concentration of 2 ppb is considered the upper limit for aquatic life in ponds as Se tends to bioaccumulate 1,000- to 2,500-fold in the aquatic food chain. Selenium is toxic to water birds because it substitutes for sulphur in essential amino acids, resulting in embryo deformities and reduced reproduction rates. Ten ponds (with a surface area of about 4,900 ha) are still active and are managed by seven

738

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

operators. The other ponds have been voluntarily deactivated due to the high cost of meeting the waste discharge requirements and mitigation measures for bird toxicity. The waste discharge requirements for drainage water discharge to evaporation ponds define the requirements and compliance schedules designed to discourage wildlife use of evaporation ponds and/or provide mitigation and compensation measures to offset adverse effects on birds. The evaporation ponds were be constructed by excavating the soil to form berms with side-slopes of at least 3⬊1 (h⬊v). The pond bottoms are unlined but compacted and the pond environs are kept free of vegetation. A minimum pond water depth is maintained at 60 cm. Migratory water birds attracted to the pond site are scared (hazed) off, and bird diseases must be kept under control and immediately reported to the California Department of Fish and Game. Compliance monitoring includes seasonal drainage water and sediment Se concentrations, and biological monitoring of birds for abundance and symptoms of toxicity. When adverse Se impacts are noted, off-site mitigation measures must be implemented with either compensation or alternative habitats, guided by approved protocols and risk analysis methods developed by the U.S. Fish and Wildlife Service. Compensation habitats are constructed to mitigate unavoidable migratory bird losses by providing a wetland habitat safe from Se and predators. Alternative habitats are year-round freshwater habitats immediately adjacent to contaminated ponds in order to provide dietary dilution to Se exposure. In spite of the stringent waste discharge requirements and associated high costs, farmers and districts in the Tulare subarea use evaporation ponds for drainage water disposal basins because there are no opportunities for off-site discharge into the San Joaquin River or other designated sinks for drainage water (SJVDIP 1999a). Integrated On-Farm Drainage Management Integrated On-Farm Drainage Management (IFDM) is water management system designed to manage irrigation, surface, and subsurface drainage flows within a farming unit or group of farms and to provide the ultimate disposal of all drainage water, including saline water, in an environmentally sound manner. Implicit in this definition is the consideration that source control is a significant component of an IFDM system. The goal of the integrated system is to minimize the drainage water volume by implementing source control. Reuse of drainage water is also a component to provide additional reduction of the drainage volume in an economic manner. The components needed for the ultimate disposal must be completely developed. The ultimate disposal may be either on a salttolerant crop or in a solar evaporator. There is no single design for a solar evaporator. Various configurations of solar evaporators are possible

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

739

(CIT 2004). The state of California has written regulations covering the solar evaporator design, construction, operation, and closure requirements (California Code of Regulations 2008). It must be emphasized that there is no standard design for an IFDM system. The designer develops an irrigation and drainage water management system that disposes of saline drainage water on the farm. The manner and components of the system that will be developed typically are based on the existing equipment, crops grown, and the farmer’s long-term goals. An IFDM system currently serves as a viable alternative for on-farm subsurface drainage water disposal system throughout the San Joaquin Valley. Evaporation ponds are being used in some locations, but the requirements for mitigation and compensation wetlands, and the environmental monitoring, limit the viability of ponds to those areas with verifiably low levels of Se (CIT 2005). Figure 23-3 illustrates the concept of an IFDM system. Land Application and Irrigation Under certain conditions, these disposal options can be a viable, beneficial use of drainage water. Benefits include volume reduction through evapotranspiration, increased availability of high-quality water for irrigation, additional revenue from sale of irrigated crops, and aesthetic value of created landscapes. Factors associated with land application include the salinity tolerance of target vegetation to salinity, the ability to meet groundwater quality standards, the availability and cost of land, percolation rates, and irrigation needs. Before adopting land application, an assessment of the compatibility of target vegetation with the proposed drainage water quality is conducted, including an assessment of the sodium adsorption ratio (SAR), trace metals uptake, and other vegetative and percolation factors. Regulations governing groundwater quality and protection of drinking water aquifers should be investigated to confirm the acceptability of this alternative. Where salinity levels are excessive, special salt-tolerant plant species (halophytes) could be considered for irrigation. Maintaining ample groundcover to eliminate dust problems is also an important consideration for land applications. In general, drainage from land application disposal sites would still require either (1) subsequent treatment and/or volume reduction through further irrigation on even more salt-tolerant crops, (2) the use of evaporation ponds, or (3) the use of solar evaporators, such as those implemented in an IFDM system to ultimately prepare the salts for disposal. In the southern San Joaquin Valley, as part of its IFDM system, the commercial fruit and vegetable producer AndrewsAg Inc. (Kern County, California) uses the halophytic iodine bush (Allenrolfea occidentalis) and native salt grass (Distichlis spicata L.) to further concentrate drainage water reused on the ranch before placing the final effluent into a solar

740

FIGURE 23-3. Concept of integrated on-farm drainage management. Courtesy of the California Dept. of Water Resources, 2008.

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

741

evaporator (CIT 2005). Irrigation of halophytic plants increases the range of drainage concentrate salinities that can be discharged into solar evaporators; however, markets for halophytic commodities are not yet well established. Solar Evaporators A solar evaporator is an enhanced evaporation system that increases evaporation by increasing water temperature, water surface area, and the vapor pressure difference between the atmosphere and the surface. Solar evaporators reduce surface tension, as well as the bond between water molecules, and take advantage of wind energy and increased surface roughness. In the San Joaquin Valley, solar evaporators are used to collect highly concentrated agricultural subsurface drainage water from the final drainage water reuse cycle of an IFDM (Fig. 23-3). The drainage water brine is discharged and evaporated into the solar evaporator by timed sprinklers or spray nozzles. The surface of the evaporator is lined and covered with a gravel pack to absorb solar heat, enhancing evaporation. The surface of the evaporator is sloped to drain any unevaporated runoff into an underground cistern that circulates back into the spray nozzle system. Before being pumped back into the spray system, the brine in the cistern is mixed with other farm drainage water. Leftover dried salts from the evaporator process are stored on the surface of the evaporator for recycling or for disposal. The solar evaporator does not contain standing water. When climatic conditions are not favorable for evaporation, drainage water is stored in underground or covered reservoirs. A covered fence is placed around the perimeter of the solar evaporator to prevent salt drift. The lower portion of Fig. 23-3 depicts a solar evaporator constructed at Red Rock Ranch, near Five Points, California, as a part of an IFDM system. In 2003, CDWR constructed and evaluated a 100-ft  100-ft solar evaporator for the purpose of final disposal of agricultural drainage water collected from farming operations at Red Rock Ranch. Figure 23-4 shows evaporation rates achieved during one season. The solar evaporator achieved enhanced evaporation up to 3.3 times the normal pan evaporation rates using 1.5-ft-high fan sprinklers (CDWR 2006). Salts stored on the surface of the evaporator can be recovered for future use, with calcium sulfate and sodium sulfate being the most feasible salts to separate. Selenium and B can also be extracted from the salt crust by implementing various biological and chemical processes. Solar evaporators can be designed to separate salts by taking advantage of the saturation-point properties of the different salts present in drainage water at different concentrations and temperatures.

742

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 23-4. Evaporation rates achieved by the pilot solar evaporator at Red Rock Ranch, central California, during the 2003 season. Courtesy of the California Dept. of Water Resources. In the San Joaquin Valley, two viable farming operations have successfully implemented and operated IFDM systems for more than a decade, and the Westlands Water District is proposing use of solar evaporators in lieu of the evaporation basins proposed by the Bureau of Reclamation in their San Luis Unit Drainage Feature Re-evaluation EIS preferred alternative (USBR 2006). Several other farmers working on drainage-impaired land on the west side of the San Joaquin Valley are either in the process of implementing IFDM systems or have expressed interest in doing so (see Fig. 23-5). Deep-Well Injection Disposal by injection into deep aquifers involves injecting drainage water into a well for placement into a porous geologic formation below

743

FIGURE 23-5. Schematic of Westlands Water District’s preferred alternative to provide drainage service to its district. Courtesy of the U.S. Bureau of Reclamation.

744

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the soil surface (Lee 1994). The oil and gas industries use deep-well injection to dispose of waste brine from their processes. In California, deepwell injection technology has been used for more than 60 years for the disposal of oil-field brines. This technology could potentially be applied to the disposal of agricultural drainage water (URS Corp. 1986) provided the conditions of the receiving geologic formation are adequate and the costs reasonable. The main environmental concern is leakage from pipes and confining aquifers. Drainage leaks could contaminate freshwater zones. The drainage water and the geochemistry of the geologic formation should be evaluated to ensure that the injecting fluids do not cause the formation to clog biologically and prematurely fail (Lee 1994) Furthermore, the receiving geologic formation should have sufficient porosity and thickness to receive the injected water. If water containing nitrate is injected into aquifers containing organic matter and ferrous iron, the growth of nitrate-reducing bacteria might clog the pores of the receiving formation as they accumulate (Westcot 1997). A prototype deep-well injection system to dispose of up to 4 000 m3/d of drainage water was built by the Westlands Water District. The drainage water was injected into shale and sand formations 1,554 and 2,164 m, respectively, beneath the ground surface. The well was drilled to a total depth of 2,469 m at a cost of about $1 million. The casing was perforated at 13 perforations per meter from depths of 2,245 to 2,344 m and from 2,411 to 2,414 m in the Martinez geologic formation, for a total length of perforations of 102 m. Following the recovery of some natural formation water samples, an injection test was conducted. The fluid injected was irrigation water filtered through a 0.5-micron filter and treated with 2% potassium chloride and a chlorine biocide. The filtered and treated water was injected through 4.75-cm tubing at a rate of 12 L/s for a total of 175,000 L of water injected. Then a 48-hour pressure fall-off test was conducted, revealing that the permeability of the geologic formation was 12 mm/d. This permeability value was too low to achieve the desired injection rate of 44 L/s. In spite of the filtration and chlorination of the injected water, plugging of the conducting pores occurred in the shale (Johnston et al. 1997). The U.S. EPA rejected a request for permission to conduct a second injection test in the sandy, more permeable formation above the shale. The Westlands Water District decided to abandon this method of disposal.

RECLAMATION AND REUSE OPTIONS There are many ways to reclaim and reuse agricultural drainage water and good management practices exist. A few of the more promising ones are described as follows:

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

745

Salt-Tolerant Crops Chapter 20 of this manual covers in detail the use of saline drainage water for irrigating salt-tolerant crops. There are many options available for salt-tolerant cropping patterns. The selected ones are typically the those that provide the most economic benefit. Soil type and quality and the potential irrigation-water quality will guide plant selection. For freshwater-irrigated acreage, the crops grown will be fairly typical, ranging in salt tolerance from sensitive to tolerant. In reuse areas, forages and salttolerant agronomic or industrial crops, such as canola or cotton, can be grown. Currently, ‘Jose’ tall wheatgrass (Agropyron elongatum) and canola are examples of crops being grown economically and successfully in drainage-impaired areas for the production of animal feed and biofuels. If needed or desired, a subsequent reuse area of halophytes (highly salttolerant, undomesticated plants) can be grown. In Kern County, California, Andrews Ag utilizes native halophytes (iodine bush and salt grasses) in their IFDM system to reduce the volume of drainage water disposed of in the solar evaporator (Benes et al. 2005). The halophyte plants consume an average of 5 acre-ft per year of drainage water with EC values exceeding 15 dS/m. Figure 23-6 depicts the iodine bush plantation at Andrews Ag.

FIGURE 23-6. Native iodine bush plantation, part of the Andrews Ag IFDM system in Kern County, California. Courtesy of the California State University– Fresno (2005).

746

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Salt-tolerant trees also have the potential to lower water tables and reduce the volume of drainage through evapotranspiration and consumptive use. An agroforestry program for the management of saline drainage water was demonstrated in the San Joaquin Valley (Jorgensen et al. 1994). In 1985, eucalyptus, casuarina, poplar, mesquite, and Eldarica pine trees were planted on approximately 75 ha. The seedlings were irrigated using fresh water during the first year. After the trees were established they were irrigated with saline drainage water from the groundwater system. Findings, to date, indicate that drainage water can be used to irrigate salttolerant trees. Once established, trees can draw water from the groundwater system for growth and evapotranspiration. An added incentive for the use of trees is the uptake, and later disposal, of trace elements. Harvesting trees provides biomass that can fuel power plants that generate electricity and produce pulp. Currently, growth of the halophytic trees Prosopis alba (Argentine mesquite) and a Pauwlonia tree hybrid are being investigated by CDWR for their economic potential to produce high-quality wood for the furniture market and biomass, respectively, while helping to manage regional shallow groundwater in salinity-impaired lands. Aquaculture Aquaculture offers an opportunity for disposal and beneficial utilization of highly saline agricultural drainage water. Currently, markets for shrimp and fish species, such as tilapia, exist. Since 2001, brine shrimp (Artemia franciscana) has been successfully raised, harvested, and marketed as a food source for aquaculture hatcheries, aquariums, and aquatic pet markets in Tulare Lake Drainage District evaporation pond cells (D. Davis, general manager of Tulare Lake Drainage District, personal communication, 2007). The brine shrimp thrive at salinities between 70,000 and 80,000 mg/L and at a pH between 8 and 9. However, aquaculture needs careful management to meet standards of ecological sustainability and to maintain a net benefit to the environment. Power Plant Cooling Drainage water can be used to cool power plants but, due to its chemical characteristics, it needs to be treated beforehand for corrosion and scale control. A test program in California used water from a tiled drainage system for power plant cooling (CDWR/UC 1978). The study involved a plant designed to treat 43.5  103 m3/d (11.5 MGD) for a 1,000-MW power plant. Treatment processes softened water by ion-exchange resins and then regenerated the resin using concentrated cooling tower blowdown without adding new chemicals. The estimated cost of treating the drainage water was about $132 per 1,000 m3 ($163 per acre-ft), adjusted to 1984 costs but not

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

747

including the disposal of any wastewater. Additional costs for brine disposal in evaporation ponds added $88 per 1,000 m3 ($108/acre-ft). Therefore, the sum of treatment and evaporation pond disposal costs totaled about $220 per 1,000 m3 ($271 per acre-ft). If Se or other contaminants in the waste brine were to reach hazardous levels, the disposal costs could be substantially higher because special holding ponds might be required. Use of drainage water from the Palo Verde Irrigation Outfall Drain in the Colorado River basin was studied for a hypothetical power generating station near Las Vegas, Nevada. Laughlin (1986) prepared a report as part of the Bureau of Reclamation’s Colorado River Water Quality Improvement Program. The report addresses several ways to treat drainage water and concluded that drainage is a viable source of cooling water, which would help lower salt loads in the basin. Reclamation for Reuse Treated water can be used for municipal, industrial and agricultural purposes if it can meet reuse standards. Limitations on reuse depend on the desalination process and the drainage water’s chemical composition. Not only must the TDS be reduced, but substances, such as B and toxic trace elements, must also meet standards for beneficial uses. For example, in reuse for agricultural irrigation, levels of TDS and B are particularly important, but for environmental or human consumption uses the levels of toxic trace elements are important to evaluate. The desalination process is considered an expensive treatment for agricultural reuse purposes; however, desalination for municipal or industrial purposes may be considered affordable if other supplies are too expensive to obtain. To date, the desalination of drainage water for reuse has not been widely practiced. Salinity-Gradient Solar Pond Salinity-gradient solar pond (SGSP) technology could have an important role in the utilization and disposal of agricultural subsurface drainage water. Basically, a SGSP is a body of water or brine that captures solar energy for either evaporation, such as in a saltworks operation, or to provide useful heat (University of Texas 1995). The SGSP can store large amounts of salt, evaporate drainage water, and capture a portion of the solar energy that radiates through the surface of the pond. Solar ponds also have several advantages over other solar technologies; they have a lower cost per unit area of collector, inherent storage capacity, and are easily constructed over large areas (University of Texas 1995). In addition, SGSPs can generate heat energy for a variety of applications such as industrial process heating, space heating, refrigeration, desalinization, and electricity. These

748

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

applications can provide significant benefits to agriculture and could offset the cost of constructing the pond. Despite its potential, a commercial SGSP has not been constructed to date because the economic benefits of SGSPs remain uncertain. Use of evaporation ponds and desalination technologies results in endproducts that must somehow be disposed, and the SGSP technology presents a possibility for using the resulting waste brine from evaporation and desalination to produce electrical energy. Several types of solar ponds for converting solar energy to electrical energy have been described in the literature, but the discussion is limited to nonconvective, salt-gradient solar ponds (USBR 1982). The nonconvective ponds are about 2 to 5 m deep, with three distinct zones. The top layer, or the upper convecting zone, is about 0.2 to 0.5 m deep and consists of uniform, relatively low-salinity water of up to about 60,000 mg/L (MIT 1985). The intermediate zone is the nonconvective gradient zone, in which both temperature and salinity increase with depth. This zone, about 0.75 to 1.5 m thick, helps to insulate the underlying zone of heat storage. The third and lowest zone, usually about 1.0 to 3 m thick, consists of very saline brine of about 150,000 mg/L to 250,000 mg/L. Ordinarily, heated bottom-water in water bodies expands to become lighter and rises to the surface to lose its absorbed heat by convection and radiation. However, in SGSPs, the salt content of the lower zones causes them to remain denser than the zones above, even when heated. Shortwave solar radiation penetrates the upper zones into the heat storage zone and raises its temperature. The stored heat can be used as a low-temperature energy source. A study of solar ponds in Israel indicates that about 20% to 25% of the incident irradiation can be collected and extracted at about 82 °C to 93 °C (Ormat Turbines, Ltd. 1981). Using specially designed turbines and generators, Ormat showed that the low-temperature energy can be effectively converted into electrical energy. The Bureau of Reclamation demonstrated the Ormat technology at El Paso, Texas (Hightower and Bronicki 1987). The energy budget for solar ponds depends on the following factors: net incident solar radiation, penetration of solar radiation to the zone of heat storage (as affected by solar angle and water clarity), diffusion of heat from the zone of heat storage to the intermediate and nonconvecting gradient zone, loss of ground heat from the zone of heat storage, and thermal energy extraction (MIT 1985). Energy extraction is accomplished by one of two methods. One is to place the heat exchangers in the zone of heat storage. The other way is to pass the heated brine through an external heat exchanger and return the cooled brine to the pond. Solar ponds may need to meet stringent construction standards due to the possibility of concentrating trace element contaminants to hazardous

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

749

levels. If toxic containments were required, the cost would increase substantially. The SGSP technology is not currently considered a viable part of the solution to drainage-water treatment, but technology may advance to make them feasible. Such advancement will be governed largely by the supply and demand for renewable energy (Lee 1994). Salt Recovery The recovery of salts or minerals from agricultural drainage water after concentration in evaporation ponds and/or solar evaporators may present the ultimate solution to the problem of salt disposal. The salts produced could be sold to cover some, if not all, of the disposal costs. Mineral extraction companies have expressed interest in extracting various constituents from San Joaquin Valley agricultural drain water. However, the trends in worldwide markets and the proximity of the markets to the salt sources are important factors in determining the viability of salt recovery. The commercial marketing of recovered salt could be limited unless the source was favorably located near the market. The major constraints in managing drainage water are the effects of salinity on plant growth and Se toxicity to wildlife. High concentrations of salt are harmful to most plants and necessitate maintaining soil and water salt concentrations within a specific range in order to safeguard productive agriculture. Selenium can be toxic to wildlife, and wildlife exposure to Se in agricultural drainage water must be prevented. However, both salt and Se have essential and established beneficial uses in industry, and in the case of Se, as an essential nutrient in mammalian nutrition. Many areas of the world, including parts of California, lack sufficient sources of naturally occurring Se. Consequently, the difficulties associated with salt and Se are related to separation and distribution, and not merely disposal. An evaluation of the concentrated drainage-water elements as resources rather than pollutants is therefore justified. The salt composition of agricultural drainage water differs from that of seawater. Seawater primarily contains sodium chloride, whereas drain water from the west side of the San Joaquin Valley typically contains sodium sulfate. When drain water is concentrated by evaporation, the dominant minerals that precipitate are thenardite (sodium sulfate), halite (sodium chloride), gypsum (hydrated calcium sulfate), and calcite (calcium carbonate). The drain water also contains several trace elements of concern, including Se, arsenic, B, and Mo (Jenkins et al. 2003). During the evaporation process, those elements will associate with, or become incorporated into, the precipitated mineral salts. Such contamination of the salt minerals may have positive or negative implications depending on the intended use of the salt.

750

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Gypsum is the salt that has the most potential for recovery and reuse. According to the California Department of Food and Agriculture, in 2006, agriculture in the San Joaquin Valley used 826,000 tons of gypsum (CDFA 2007). In addition, commercial uses of sodium sulfate include textile dyeing, glass making, and glazing, as well as other industrial uses. Options for salt recovery and reuse have been evaluated in the SJVDIP’s Salt Utilization Technical Committee report (1999b), For certain commercial and industrial uses, salt must first be purified. The sodium sulfate industry, for example, may require a purity standard exceeding 90%. The U.S. market for sodium sulfate is approximately 1.5 million tons per year. However, as of 1989, the combined annual deposition of salt in evaporation ponds in the San Joaquin Valley was an estimated 0.9 million tons per year, which could mean that harvesting and marketing of that volume of sodium sulfate could drive down the price to levels so low that salt harvesting from drainage water could become uneconomical. Transportation costs must also be considered in salt-harvest plans as the cost of freighting the harvested material to a salt refinery or other market must be low enough to still ensure a profit. Collection of concentrated drainage water in large-scale solar evaporators would facilitate salt harvesting because salts are left over on the surface of the solar evaporator, thus facilitating its recovery for reuse. It is estimated that establishment of a large-scale (1 million tons per year) plant to process agricultural drainage salts would require approximately 10 years of research, development, testing, design, and construction. While this is being accomplished, it would be necessary to store the drainage salts. Considering that 2 to 3 million tons of annually imported salts (in addition to significant amounts of salt mobilized from soils as a result of irrigation) need to be disposed of to maintain salt balance in the SJV, even an optimistic estimate of the amount that could be commercially marketed would represent a only a small percentage of the total salts needing to be disposed. Active pursuit of commercial utilization of the salts and Se is needed, and will require a variety of options for separating the salts from productive agricultural fields. However, adopted salt harvesting and utilization options should not negate pursuit of other salt disposal options, such as in lined storage facilities or ocean disposal. Table 23-3 presents the 2007 market value of salts and certain trace elements present in agricultural drainage water. Other Reuse Options The reuse of reclaimed drainage water has been suggested for many other purposes. A novel way to integrate several reclamation and reuse options is cogeneration-desalination, which was explored for the Westlands Water District, California (URS Corp. 1987). It involves using gas

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

751

TABLE 23-3 Salt Utilization and Separation Extractable TDS from Subsurface Drainage Water, Red Rock Ranch Average Dry Weight (%) (1)

Average Market Price, 2007 (2)

Sodium sulfate Sodium chloride Calcium sulfate Magnesium chloride Sodium nitrate Calcium carbonate Boron Potassium chloride Selenium

37 33 16 7 3 2 1 0.8 0.01

$134/ton $57/ton $18/ton $270/ton $300/ton $125/ton $425/ton $270/ton $33/lb

Other

0.19

Price Source: USGS Mineral Commodities and Chemical Market Reporter.

turbines to generate electrical power, with the bypass heat used to desalinate drainage water. Salt from the brine stream crystallizes, the desalinated water is made available for reuse, and surplus electrical power from plant operation is sold. Although the technology of each process is well developed, the processes have yet to be integrated into an economical system. Economic considerations include whether the sale of surplus electrical energy and salt products can cover the costs of processing (Lee 1994).

SUMMARY In this chapter we have examined the potential drainage-water treatment and disposal strategies and technologies in light of recent field experience, with an emphasis on the ability of technologies to address the magnitude of the drainage problems. To date, there has been progress in increasing the effectiveness of desalinization of agricultural drainage water and in bringing unit costs down (in the $500/acre-ft to $800/acre-ft range, not including conveyance to a facility). There has also been significant work on treatment of Se, using both physical and biological processes. On-farm management to reduce the volume of highly saline drainage water has been successful. Recent field methods for minimizing the attractiveness of evaporation ponds for wildlife and providing alternative habitats for waterfowl have been successful, and the TLDD’s use of

752

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

drainage ponds to raise a commercial product (brine shrimp) have reduced the net costs of evaporation pond management. The long-term technological challenges remain how to concentrate and locally sequester any salts that are the byproduct of desalination or evaporation technologies, or how to remove these salts from the system. Reuse of drainage water for solar power or cooling of power plants appears feasible but costly. Recent efforts to sequester salts in deep groundwater have raised issues related to the ability of this approach to provide for rapid sequestration of large volumes of drainage water. In addition, there are still cost, ecological, and water-quality issues associated with disposal of salts to the ocean. Based on recent work, it is apparent that there are well developed technological approaches to solving the problems, but the processes have yet to be integrated into an economical system at an adequate scale.

REFERENCES Barg, U., Dunn, I. G., Petr, T., and Welcomme, R. L. (1997). “Inland fisheries,” in Water resources: Environmental planning, management and development, A. K. Biswas, ed., McGraw-Hill, New York. Benes, S. Grattan, S. Cervinka, V., and Diener, J. (2005). Cultivation of halophytes to reduce drainage volumes on the westside San Joaquin Valley of California, California State University Agricultural Research Initiative Report, Fresno, Calif. Boyle Engineering Corp., California Dept. of Water Resources, Kern County Water Agency, and University of California at Los Angeles. (2003). Desalination demonstration report for Buena Vista Water Storage District, Boyle Engineering Corp., Bakersfield, Calif. California Code of Regulations. (2008). Title 27. “Environmental Protection Division 2. Solid Waste Subdivision 1. Consolidated regulations for treatment, storage, processing or disposal of solid waste. Chapter 7. Special treatment, storage, and disposal units. Subchapter 6. Solar evaporators,” www.directory.westlaw. com, accessed February 8, 2011. California Department of Food and Agriculture (CDFA). (2007). Fertilizing materials use report (July 2005, June 2006), California Department of Food and Agriculture, Sacramento, Calif. California Department of Water Resources (CDWR). (1986). Technical information record on the physical/chemical pretreatment system at the Los Banos demonstration desalting facility: Sacramento, California. ———. (2004). Selenium removal at Adams Avenue Agricultural Drainage Research Center, California Department of Water Resources, Sacramento, Calif. ———. (2006). White paper: Solar evaporator for integrated on-farm drainage management system at Red Rock Ranch, San Joaquin Valley, California. California Department of Water Resources and University of California (CDWR/ UC). (1978). Agricultural wastewater for powerplant cooling development and testing of treatment processes, Vol. II, California Department of Water Resources, Sacramento, Calif.

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

753

Center for Irrigation Technology (CIT). (2004). A landowner’s manual managing agricultural irrigation drainage water: A guide for developing integrated on-farm drainage management systems, Center for Irrigation Technology, California State University, Fresno, Calif. ———. (2005). A technical advisor’s manual managing agricultural irrigation drainage water: A guide for developing integrated on-farm drainage management systems, Center for Irrigation Technology, California State University, Fresno, Calif. Central Valley Regional Water Quality Control Board (CVRWQCB). (2008). “Water quality control plan for the Sacramento and San Joaquin River basins,” www.swrcb.ca.gov/centralvalley/water_issues/basin_plans/, accessed February 8, 2011. Cohen, Y., Glater, J., McCool, B., Rahardianto, A., and Kim, M. (2008). “Membrane desalination of agricultural drainage water,” in Salinity drainage, A. C. Wang and D. Brikle, eds., University of California Press, Berkeley, Calif. Enzweiler, R. J., and Strasser, J. (2003). Energy-efficient process for using membrane technology to treat and recycle agricultural drainage water, January 29, 2003, Final report submitted to the Energy Innovation Small Grant Program of the California Energy Commission, Grant No. 52735A/01-23, California Energy Commission, Sacramento, Calif. Food and Agriculture Organisation of the United Nations (FAO). (1973). Irrigation drainage and salinity, Food and Agriculture Organisation of the United Nations, Rome. ———. (1999). Review of the state of world fishery resources: Inland fisheries. FAO Fisheries Circular No. 942, Rome. ———. (2002). Agricultural drainage water management in arid and semi-arid areas, Food and Agriculture Organisation of the United Nations, Rome. Gao, S., Tanji, K. K., Linb, Z. Q., Terry, N., and Peters, D. W. (2003). “Selenium removal and mass balance in a constructed flow-through wetland system.” J. Env. Qual., 32, 1557–1570. Grassland Water District. (2008). http://www.grasslandwetlands.org/about/ Green, F. B., Lundquist, T. J., Quinn, N. W. T., Zarate, M. A., Zubieta, I. X., and Oswald W. J. (2003). “Selenium and nitrate removal from agricultural drainage using the AIWPS® technology.” Water Sci. Technol., 48(2), 299–305. Harza Engineering Co. (1986). Selenium removal study, prepared for Panoche Drainage District, Harza Engineering Co., Firebaugh, Calif. Hightower, S., and Bronicki, L. (1987). Installation and operation of the first 100 kW solar pond power plant in the United States, U.S. Dept. of the Interior, Bureau of Reclamation, Denver, Colo. Jenkins, B. M., Sun, G., Cervinka, V., Faria, J., Thy, P., Kim, D. H., Rumsey, T. R., and Yore, M. W. (2003). Salt separation and purification concepts in integrated farm drainage management systems, Paper No. 032236, ASAE Annual International Meeting, Las Vegas, Nev. Johnston, W. R., Tanji, K. K., and Burns, R. T. (1997). “Drainage water disposal,” in Management of agricultural drainage water quality, C. A. Madramootoo, W. R. Johnston, and L. S. Willardson, eds., FAO Water Reports No. 13, Food and Agriculture Organisation of the United Nations, Rome. Jorgensen, G. S., Solomon, K. H., and Cervinka, V. (1994). Agroforestry farming system for the management of selenium and salt on irrigated farmland, Project

754

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Report, Center for Irrigation Technology, California State University, Fresno, Calif. Lalvani, S. B. (2005). Selenium removal from agricultural drainage water: Lab scale studies, Final Report to the California Department of Water Resources, Fresno, Calif. Laughlin, J. K. (1986). Study of saline water use at the Etiwanda generating station, Prepared for U.S. Dept. of the Interior, Bureau of Reclamation, Denver, Colo. Lee, E. W. (1994). “Drainage water treatment and disposal,” in Management of water use in agriculture, K. K. Tanji and B. Yaron, eds., Springer-Verlag, New York. Lundquist, T. J., Green, F. B., Tresan, R. B., Newman, R. D., Oswald, W. J., and Gerhardt, M. B. (1994). “The algal bacterial selenium removal system: Mechanisms and field study,” in Selenium in the environment, W. T. Frankenberger, Jr., and S. Benson, eds., Marcel Dekker, Inc., New York, 251–278. Massachusetts Institute of Technology (MIT). (1985). A state-of-the-art study of nonconvective solar ponds for power generation, Report prepared for the Electrical Power Research Institute (EPRI), Palo Alto, Calif. Mayenkar, A. (1986). Removal of Dissolved Heavy Metals from Aqueous Waste Effluents. U.S. Patent No. 4,565,633. McCutchan, J. W., Goel, V., Bryce, D. B., and Yamamoto, K. (1976). “Reclamation of field drainage water.” Desalination, 19, 153–160. Moody, C. D., Murohrt, A. P., Ralston, B. G., Hulsey, R. A., and Hyde, G. M. (1988). Experimental results for removing selenate from agricultural drainage waters, U.S. Dept. of the Interior, Bureau of Reclamation, Denver, Colo. Murphy, A. P. (1988). “Removal of selenate from water by chemical reduction.” Ind. Eng. Chem. Res., 27, 181–191. Ohlendorf, H. M. (1984). “The biological system,” in Conference on Toxicity Problems at Kesterson Reservoir, California, Proceeding of a Research Meeting, Sacramento, December 5–7, 1983. Ormat Turbines, Ltd. (1981). A study of the feasibility of a solar pond generating facility in the state of California, USA, Final Report, Volume I of a report prepared for the Southern California Edison Co., Rosemead, Calif. Rahardianto, A., Shih, W. Y., Lee, R-W., and Cohen, Y. (2006). “Diagnostic characterization of gypsum scale formation and control in RO membrane desalination of brackish water.” J. Membrane Sci., 279(2006), 655–668. Ramsar Convention Bureau. (2000). “Ramsar information paper No. 1,” www. ramsar.org/, accessed February 8, 2011. San Joaquin Valley Drainage Implementation Plan (SJVDIP). (1999a). Evaporation pond technical committee report, San Joaquin Valley Drainage Implementation Program, Department of Water Resources, Sacramento, Calif. ———. (1999b). Salt utilization technical committee report, San Joaquin Valley Drainage Implementation Program, Department of Water Resources, Sacramento, Calif. Tanji, K. K., and Gao, S. (1999). TLDD flow-through wetland system: inflows and outflows of water and total selenium as well as water Se speciation and sediment Se speciation, Annual Report, UC Salinity Drainage Program, University of California, Riverside, Calif.

DRAINAGE WATER TREATMENT AND DISPOSAL OPTIONS

755

Tanji, K. K., and Neeltje, K. C. (2002). Agricultural drainage water management in arid and semi arid areas, Food and Agriculture Organisation of the United Nations, Rome. University of Texas.(1995). Salinity gradient solar ponds, The El Paso Solar Pond Project, Mechanical and Industrial Engineering Department, University of Texas, El Paso, Tex. U.S. Dept. of the Interior, Bureau of Reclamation (USBR). (2008a). San Luis drainage feature re-evaluation, feasibility report, Appendix D: Reverse osmosis analyses reports, U.S. Dept. of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (2008b). San Luis drainage feature re-evaluation, feasibility report, Appendix E: Selenium biotreatment pilot reports. ———. (2006). San Luis drainage feature re-evaluation, final environmental impact statement. ———. (1982). Preliminary study of solar ponds for salinity control in the Colorado River basin, Appendix A: Solar pond brine properties and pretreatment options, U.S. Dept. of the Interior, Bureau of Reclamation, Denver, Colo. URS Corp. (1986). Deep-well injection of agricultural drain waters, prepared for the San Joaquin Valley Drainage Program, Sacramento, California, URS Corporation, San Francisco, Calif. ———. (1987). Agricultural drainage salt disposal, report prepared for Westlands Water District, URS Corporation, San Francisco, Calif. Westcot, D. W. (1997). “Drainage water quality,” in Management of agricultural drainage water quality, C. A. Madramootoo, W. R. Johnston, and L. S. Willardson, eds., FAO Water Report No. 13, Food and Agriculture Organisation of the United Nations, Rome.

NOTATION TDS  total dissolved solids

This page intentionally left blank

CHAPTER 24 AGRICULTURAL EVAPORATION BASINS Suduan Gao, Evan Christen, and Anthony L. Toto

INTRODUCTION Agricultural evaporation basins are storage areas used to impound and dissipate agricultural drainage water. Evaporation basins are used worldwide in areas where there are constraints on disposal of saline drainage waters to natural water bodies such as oceans, rivers, lakes, or natural depressions. Evaporation basins are mostly used in arid and semiarid regions such as Australia, central Asia, Pakistan, and the southwestern United States (Tanji et al. 1993; Tanji and Kielen 2002). These areas are commonly associated with irrigated agriculture, high levels of salts naturally occurring in soils (primary salinity), high evapotranspiration rates, and waterlogging problems. For productive irrigated farming to continue, adequate leaching and drainage to remove excess salt in the rootzone are necessary (Hoffman 1985). The natural drainage capacity of the soil and the groundwater system in these areas is usually insufficient to remove water that has infiltrated in excess of crop requirement. Therefore, engineered drains are necessary to prevent waterlogging and salinization of the crop rootzone (Chapter 16 in this manual). Surface or subsurface drains act to remove water from the soil profile and to allow leaching of salts from the crop rootzone. Evaporation basins are used to store and evaporate water as a disposal mechanism for saline drainage water where other disposal options, such as discharge into rivers, are deemed unacceptable. Drainage waters impounded in such basins are desiccated by evaporation, resulting in elevated levels of dissolved mineral salts and trace elements as well as precipitation of evaporite minerals. Trace elements present in drainage water, such as selenium (Se), may pose a hazard to wildlife attracted to the ponds. Seepage losses may degrade surface and groundwater quality in adjacent areas. However, in the absence of toxic trace elements, evaporation ponds 757

758

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

can be developed into valuable ecosystems and wildlife refuges. The overall concerns that evaporation ponds will concentrate salts in a small area, that the salts may contain toxic trace elements, and that the ponds will in all likelihood leak to the groundwater system have led to the development of policies and regulations to minimize the negative environmental impacts associated with their use. This chapter reviews and summarizes information about agricultural evaporation basin use, their siting, design, operation, chemical/biological characteristics, and environmental and regulatory issues. Most detailed information on evaporation basin use comes from the San Joaquin Valley (SJV) of California, USA, and the Murray-Darling Basin (MDB) in Australia. Thus, this chapter concentrates on information from these regions, which broadly represent the conditions in many irrigated semiarid regions of the world. Evaporation Basins in the San Joaquin Valley of California (Background, Status, and Operation) The San Joaquin Valley is in the southern part of the Central Valley of California in the United States (Fig. 24-1). The climate is Mediterranean or semiarid to arid, that is, summers are hot and dry and winters are cool and moist (Arroues and Anderson 1986). Total precipitation averages 100 to 200 mm per year, occurring mostly between October and April, and evapotranspiration (ET) ranges from 1,400 to 1,600 mm per year (CIMIS 1999). The highest ET0 (reference ET) in summer months can be up to 240 mm. Currently, irrigation water comes mostly from surface water supplies imported from northern California (Tanji et al. 1986). Drainage has been the greatest challenge for sustainable agricultural production in the SJV. The valley is surrounded by the Sierra Nevada Mountains (igneous and metamorphic rocks) to the east, the coastal ranges (marine and nonmarine sedimentary rocks) to the west, and the San Emigdio and Tehachapi ranges (igneous and metasedimentary) to the south. The Tulare Lake Basin, located in the southern part of the SJV, is a hydrologically closed basin. Soils on the west side of the valley (the “Westside”), derived from sedimentary parent materials from the coastal ranges, contain high levels of salts and trace elements (chiefly Se, As, B, Mo, V, and U). Soils in the Tulare Basin are derived from flood-basin, lacustrine, and marsh deposits that are also high in salts and trace elements. Much of the west side of the SJV (about 750,000 acres) is affected by rising or shallow groundwater tables (less than 5 ft) due to irrigation and the presence of an impermeable layer of Corcoran clay (Tanji et al. 1992). The poor subsurface drainage conditions on the Westside and the hydrologically closed Tulare Basin require supplemental drainage for sustainable crop production.

AGRICULTURAL EVAPORATION BASINS

759

FIGURE 24-1. Locations of agricultural evaporation pond facilities in the San Joaquin Valley constructed between 1972 and 1985, and status in 1987. By 2007, only eight facilities (shaded) were operating (Nos. 9, 11, 12, 21, and 22 in the Tulare Lake bed; Nos. 8 and 23 in the alluvial fan, and No. 6 in the basin trough were active). Modified from Ford (1988).

A federal water project, the San Luis Project, was initiated in 1960 with the purpose of constructing a 280-mile drainage channel from the valley to the Pacific Ocean through the bay delta (Tanji et al. 1993). An upper 85-mile San Luis Drain was constructed during 1960 to 1975, terminating at Kesterson Reservoir (a wildlife refuge). The remainder of the drain was never completed. High Se levels in the drainage water disposed into the Kesterson Reservoir were discovered to be toxic to fish and water birds, causing the shutdown of the reservoir by 1986 (Ohlendorf et al. 1993; Tanji et al. 1986). Since then, drainage water disposal in the region continues to face great challenges and relies on temporary resolutions within

760

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the valley (SJVDP 1990; EPTC 1999). Most proposed actions, including source reduction and remediation, are either expensive or subject to taking some land out of production. Some reuse of drainage water through integrated farm drainage management systems is practiced on a small scale with final disposal of salts collected in the terminal solar evaporator, but this still presents problems (Chapter 17). Currently, evaporation basins used at a regional scale are considered the only economically feasible solution to disposal of agricultural drainage water in the Tulare Basin and Westside of the SJV. The evaporation basins in the SJV were constructed mostly between 1981 and 1986 and encompass a total surface area of 2,800 ha (Fig. 24-1) (Ford 1988; Tanji et al. 1993). Most of these basins are operated by the local drainage or water district. The high cost of meeting regulatory requirements in monitoring and reporting has resulted in closure of some basins (Tanji et al. 1992). Some basins were converted to a component of integrated on-farm drainage management (IFDM) systems (Cervinka et al. 1999; Blackwell et al. 2005). By 2007, there were only eight basins still operational, covering a total surface area of ⬃1,940 ha (Fig. 24-1). In the late 1980s, the drainage water volume impounded into the SJV evaporation ponds was about 39 million m3 coming from 22,700 ha of drained fields (Ford 1988). This drainage volume has decreased in recent years. Evaporation Basins in the Murray-Darling Basin, Australia (Background, Status, and Operation) The Murray-Darling Basin (MDB) is located in the interior of southeastern Australia and its name is derived from its two major rivers, the Murray River and the Darling River. It is Australia’s most important agriculture region, producing one-third of Australia’s food supply. Evaporation basins associated with irrigated agriculture in Australia have been exclusively used in the MDB (Fig. 24-2) and they have been extensively studied, resulting in detailed guidelines as to their design, management, and operation (Jolly et al. 2000). About 1.8 million ha are irrigated in the MDB. The majority of this irrigation occurs in the south-central part of the Basin widely known as the Riverine Plain. The climate in the Riverine Plain is described as Mediterranean or semiarid, with hot dry summers and cool winters. Annual average evaporation ranges from 1,600 to 2,000 mm, and annual rainfall from 200 to 600 mm. Even in its pre-European state, the MDB contained a vast amount of salt, which was stored in its soils and groundwater. This salt is of aerial origin, having been deposited over millennia from rainfall. The use of irrigation, the leakage of water from the associated network of water distribution and drainage channels, and, following clearance of deep-rooted perennial plants, the replacement with shallow-rooted annual crops have

AGRICULTURAL EVAPORATION BASINS

761

FIGURE 24-2. Map showing location of the Riverine Plain of the Murray Basin, Australia. altered the water balance and caused the water tables to rise throughout the basin. This has resulted in mobilization of the stored salt and soil salinization and waterlogging from the rising water tables. It was estimated in 1987 that 96,000 ha of irrigated land in the MDB were visibly affected by soil salinization and that 560,000 ha had water tables within 2 m of the surface (MDBMC 1987). Currently, the areas of high water tables have been reduced due to drought conditions, which have limited the rainfall and irrigation water availability. Duncan et al. (2008) analyzed drainage volumes and qualities from irrigated areas in the MDB and showed that about 50% of the salt load from irrigated regions comes from subsurface drainage installed to combat salinization and waterlogging, despite the fact that the area of subsurface drainage is only a small fraction of the total area. The Salinity and Drainage Strategy produced and published by the Murray-Darling Basin Commission (MDBC 1999) imposed constraints on

762

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the amount of river disposal possible. Moreover, in the past two decades there has been reduced political and community tolerance of continued disposal to river systems. The option of export of saline drainage water to the sea via a pipeline has been considered a number of times in the past (SRWSC 1978; Earl 1982; GHD 1990). However, studies have indicated that this option was relatively uneconomic and that the impacts of this option on the marine environment remain unclear. Thus, to meet salinity targets set for the Murray River, land disposal of drainage to disposal basins has become more prevalent. The use of evaporation basins historically has been as regional-scale basins that accept drainage water from multiple farms and irrigation districts and may be situated outside the districts themselves (hence, salt is exported from the area in which it is produced). Regional basins were often developed on sites most convenient from an engineering standpoint, sometimes with detrimental environmental, socioeconomic, and aesthetic impacts. The viewpoint that the beneficiaries of irrigation should be responsible for their own drainage management assumes that this would encourage more efficient irrigation and drainage management and, hence, minimize the environmental and other impacts of disposal basins and irrigation on downstream users. Local-scale basins can be privately owned on-farm basins that occupy parts of individual properties, and community basins within a drainage area that are shared by a small group of properties and are either privately or authority-owned (such as the Girgarre Basin near Shepparton). Local-scale disposal basins represent an expensive long-term commitment and are also terminal storage areas for salt. For these reasons, the desired basin is one that meets community and environmental standards, is economically viable, and results in the least amount of salt export. Hostetler and Radke (1995) collated all available hydrogeological, engineering, and operational data on more than 150 existing basins in the MDB. While the data for many basins are incomplete, the study provides a summary of available information: • 107 basins were reported as being active, with a total area of 15,900 ha, a total storage capacity of 113,342 ML, and an annual disposal volume of 210,044 ML/year. • Of the 107 active basins, 90 were reported as being used for drainage disposal (i.e., not for groundwater interception schemes or groundwater discharge), with a total area of 14,531 ha, a total storage capacity of 113,074 ML, and an annual disposal volume of 181,495 ML/year. • Of the 90 active drainage disposal basins, 9 (representing 3,338 ha) were located on the Riverine Plain, the rest being concentrated mostly in the Riverland (South Australia) and Sunraysia (Victoria) regions.

AGRICULTURAL EVAPORATION BASINS

763

Since the publication of the Hostetler and Radke (1995) report, at least another 10 on-farm basins have been constructed on the Riverine Plain in the Murrumbidgee Irrigation Area (MIA).

Basin Design and Hydrology Designs Evaporation basins are constructed by excavating soils from the interior of the basins to build up embankments, and then drainage water is discharged into the ponds by gravity flow and pumping (Tanji et al. 1985; Melville et al. 1993). Evaporation basins are usually constructed on clay loam or clay soils without lining (due to high costs) in relatively lowelevation areas. Pond designs and operation are very much dependent on the volume of drainage water disposal required, the hydrology of the basins, and the cost factor. Tanji et al. (1985) analyzed the feasibility of using evaporation basins for drainage water disposal and the requirements for basin designs in the SJV. On-farm evaporation basins provide opportunities for more reuse of drainage water than do regional basins. However, they remove land from production on the farm. Individuals are responsible for construction costs, maintenance, and monitoring, and their available funds might be limited. Seepage could degrade adjacent productive soil quality on the farms. Regional basins offer opportunities to be built on marginal land with possibly better design and construction, as well as better maintenance and monitoring. The disadvantages are limited opportunities for water reuse and the need for an extensive drainage collection system in the region. One estimate of pond surface area required for disposal of drainage on a 320-acre farm was 50 acres (15.6% of the total area that included 240 acres of irrigated productive land producing 168 acreft of drainage water per year). The total surface area of basin facility in the SJV is usually 10% to 15% of the total acreage of land drained but could be as low as 3% if the drainage were reused or as high as 33% without reuse. Most regional basin facilities are comprised of multicells (SJV) or bays (MDB) that vary in number and surface area. In the SJV, basins contain varying numbers of cells depending upon their overall size, with surface areas ranging from a few to 750 ha per cell and ranging in depth from 0.5 to 2 m (Tanji et al. 1993; Johnson et al. 1997). In the MDB, evaporation basins usually are comprised of “bays,” mainly to reduce wind effects. Some locations, such as the Wakool-Tullarkool evaporation basins, contain 40 bays—a layout that includes holding, concentrating, and crystallization ponds (Roberts 1995). Ponds are separated by earthen banks and water typically flows by gravity. The slopes of the banks are deeper in

764

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the SJV (3 : 1, h/v) (Tanji and Kielen 2002) than in the MDB (6 : 1) (Jolly et al. 2000) to reduce the likelihood of birds nesting along the banks. Salts accumulated in the crystallization bays in the MDB are harvested, but those in the SJV have no commercial value currently due to the presence of trace elements. In the SJV, in addition to the primary concern about Se and wildlife impacts due to evaporation pond use, other concerns include seepage causing possible contamination of groundwater and spillage associated with flooding (EPTC 1999). Perimeter drains are used to cut off the seepage and, therefore, seepage effects are considered insignificant. Ponds are designed, operated, and maintained to have a 100-year return period in regard to spillage or flooding. The cost of evaporation basins is an important factor in their use. In the MDB, a review of the cost of basins between 2 and 770 ha was found to be between A$3,000 and A$25,000/ha in 2001 based on the estimates by Singh and Christen (2001). A detailed analysis of size and design found that large basins cost less to construct on a per-unit-area basis. A welldesigned and -sited 2-ha basin cost about A$19,000/ha, whereas a 200-ha basin under the same conditions would cost about A$5,000/ha (Singh and Christen 2001). This is due to economies of scale in construction, especially with regard to bank length, and also because small basins require compaction to control leakage. Thus, there may be a significant advantage to using a smaller number of large basins compared to many smaller basins, but this will depend on the transportation costs of the drainage water. This scenario of few regional versus many on-farm basins has been analyzed by Singh and Christen (2000). They showed that the total cost of a regional basin with a piped drainage collector system could be either 12% less than or 11% more than on-farm basins, depending on the scenario. Leakage control is an important factor in cost minimization. Basin designs that require no additional leakage control measures are cheapest. Therefore, it is important to find sites that do not need additional leakage control measures, such as compaction. Basin disposal capacity The total amount of drainage water that can be disposed of in a disposal basin is referred to as its potential disposal capacity. This capacity results from the effects of evaporation, rainfall, leakage, and interception on the amount of drainage water that can be disposed into a given basin but does not consider recycling of shallow lateral or vertical flow (Leaney and Christen 2000a). Design disposal capacity refers to the amount of drainage water that a disposal basin can dispose of if the recycled water intercepted by the drainage system is equal to that leaked from the basin.

AGRICULTURAL EVAPORATION BASINS

765

The design disposal capacity, when matched to the required drainage for an irrigated area, determines the percentage of area that needs to be reserved for disposal basins. In the Riverine Plain of the MDB, probable potential disposal capacities for basins were estimated based on field evaluation of the individual effects of a range of factors, including direct water balance terms (rainfall and evaporation) and conditions within the basin itself. Leakage amounts, climate, soil properties, salinity of the disposal water, and the area of basin are all important factors. The complex interaction of these factors determines the potential and design disposal capacity of a basin. Increased rainfall lowers the disposal capacity, while increased evaporation raises it. Increased leakage below a basin leads to a direct increase in the potential disposal capacity. Salinity increase reduces the evaporation rate and hence the basin disposal capacity. Basin leakage, even at low rates (1 mm/day), will moderate the effect of evapoconcentration and allow the basin to function at near the maximum evaporative capacity. Without leakage, the size of basin required to dispose of a unit volume of drainage will be governed by the rate of evaporation from a saturated brine solution, which will be about 70% that of fresh water. Thus, basins without leakage will need to be at least 40% larger than those with substantial leakage, depending on the relative magnitudes of the rainfall and potential evaporation of the site and the scope of the oasis effect (Leaney and Christen 2000a). An example of this effect on the design disposal capacity for a hypothetical 1-ha basin at Hillston is shown in Fig. 24-3. Depending on the input water salinity, the design disposal capacity may be reduced by 60% to 70% for a lined (zero leakage) basin as opposed to a basin with a leakage of between 0.5 and 1 mm/day. In situations with low input salinities (10,000 mg/L), high design disposal capacities can be maintained with leakage rates less than 0.5 mm/day; however, it is generally not practical or beneficial to reduce leakage rates below this level. Similarly, the increase in design disposal capacity is less marked for leakage in excess of 1 mm/day. Therefore, there is little benefit to be gained by allowing leakage rates much higher than this. Basin area or size affects basin disposal capacity. Smaller basins can dispose of more water per unit area than larger basins due to higher leakage rates (Leaney and Christen 2000b) that maintain lower basin salinity and higher net evaporation rate (i.e., the potential disposal capacity is higher). The design disposal capacity, however, is only slightly higher for smaller basins. The higher leakage rates become especially significant when basins are to be sited above relatively fresh groundwater systems where contamination is to be avoided. Large basins, when placed on a similar site, will leak less and therefore present a lesser threat of groundwater salinization per unit of water disposed (although they may cause many more significant problems locally because of their size).

766

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 24-3. Design disposal capacity for hypothetical 1-ha basin at Hillston with varying leakage rate. There are two main ways that basin area influences the disposal capacity of a basin (Leaney and Christen 2000a): 1. The oasis effect, where larger basins tend to develop their own microclimate, resulting in increased humidity above the basin and hence less evaporation. The magnitude of this effect is also controlled by the humidity of the surrounding area. The effect will be greatest where there is no (or very little) irrigation occurring in the surrounding areas. 2. Larger basins tend to have less leakage than smaller basins for similar site conditions. Figure 24-4 illustrates the relationship between observed leakage and perimeter/area (P/A) ratio for existing basins on the Riverine Plain of the MDB in shallow water table areas where much of the leakage is shallow lateral flow away from the basin. Basins that have a larger perimeter compared to their area can have higher leakage rates—larger basins leak less than smaller basins. From observations of existing basins, together with the modeled behavior of hypothetical basins, it is considered that leakage rates of 0.5 to 1 mm/day should be considered as desirable and achievable for basins located at suitable sites in the Riverine Plain. To achieve such rates, basins less than 100 ha in area will need to have their floors compacted and be maintained with a year-round cover of water. In general, we recommend

AGRICULTURAL EVAPORATION BASINS

767

FIGURE 24-4. Effect of basin size (perimeter/area ratio) on the basin leakage rate, where P is the perimeter of the basin (m), and A is the surface area (m2). the use of design disposal capacity (all leakage recycled to the basin) when determining the area of the basin. However, it is important to make allowance for the additional operating costs incurred by complete recycling of leakage. Hydrology The hydrology of an evaporation basin is relatively simple (Fig. 24-5). The main inputs are drainage water from crop land and small amounts of rainfall, and, when applicable, pumped drainage water from perimeter drainage installed to intercept pond seepage. The outputs are mainly evaporation or evapotranspiration with some seepage (leakage). In some basins in the MDB, vascular plants grow in ponds with low-EC waters but not in ponds with high EC water (Roberts 1995). Therefore, evapotranspiration could be significant in some ponds. In the SJV, however, transpiration from aquatic vegetation is minor because no vascular plants are allowed to grow to minimize the risk of poisoning water birds due to Se intake. In the SJV and the MDB, pond evaporation rates are roughly 1.4 to 1.8 m/year, much exceeding the rainfall of about 0.2 to 0.4 m/year. In new ponds, seepage rates can be as great as 10 mm/day. This decreases dramatically to a few mm per day within the first year or two due to the progressive “bottom sealing” from continuous inundation and materials from microbial activity in bed materials (Grismer et al. 1993; Leaney and Christen 2000a,b).

768

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 24-5. Water balance of an agricultural evaporation basin.

Evaporation/evapotranspiration rates are affected by air temperature, wind speed, humidity, solar radiation, and water characteristics, such as temperature, color, and salinity (Tanji et al. 1992; Grismer et al. 1993; Johnson et al. 1997). Direct measurements of evaporation from the pond water using floating evaporation pans showed that evaporation rates from pond water declined as salinity increased (Grismer et al. 1993). A ratio of 0.73 to 0.77 for evaporation rates from pond waters of EC 14 dS/m to dryland pan evaporation rates was determined, but this may also be affected by the oasis effect of basins. The potential evaporation rate from pond waters can also be estimated from reference ET rates, ETo, provided by a California Irrigation Management Information System (CIMIS) station: E  Y(ETo)

(24-1)

where Y  1.3234  0.0066 EC (dS/m) for water of EC up to 60 dS/m. A correction factor of 1.2, 1.07, and 0.92 for waters of EC  20, 47, and 59 dS/m, respectively, can be used (Tanji et al. 1992). Precipitation of salts occurs when temperatures are cool, and dissolution occurs when temperatures rise at the water surface, which may also affect pond evaporation rates. Seepage losses from evaporation basins depend on soil, hydrogeological, and topographic conditions (Tanji and Kielen 2002). For fine-textured soils with some compaction of the basin floor and year-round coverage of water, leakage rates at many sites in the Riverine Plain of the MDB can be reduced to ⬃0.5 to 1 mm/day. Leakage is often limited by the capacity for shallow lateral flow in underlying aquifers. For similar site conditions, smaller basins will leak at a much higher rate than large basins (Leaney and Christen 2000a). In the SJV, three key factors have been identified that affect pond seepage rates: (1) spatial variability of soil texture; (2) clay content of the soil in the bed materials; and (3) microbial activity in the

AGRICULTURAL EVAPORATION BASINS

769

pore structure (Grismer et al. 1993). Microbial activity was considered the most important factor limiting pond seepage because average seepage rates from ponds for more than 2 years were similar and on the order of a few mm/day, regardless of any differences in initial seepage rates and soil texture (Grismer et al. 1993).

Monitoring Requirements for Water Quality and Sediments In California, monitoring of evaporation ponds (water quality and sediments) is required by the Regional Water Quality Control Board (Central Valley Region) under waste discharge requirements (WDRs) in compliance with the California Water Code. The soluble threshold limit concentration (STLC) and total threshold limit concentration (TTLC) are used to characterize hazardous wastes. The STLC for Se, As, B, and Mo and the TTLC for sediments specified in Title 22 and Title 23 of the California Administration Code are given in Table 24-1 (Tanji et al. 1992). If pond waters or sediments of a basin exceed the hazardous waste criteria, the basin may be subject to closure. Accumulations of salts (e.g., halite, thenardite, and mirabilite) and trace elements are commonly found in pond waters (Tanji 1990; Chilcott

TABLE 24-1. Trace Element Concentrations in Evaporation Basins in the San Joaquin Valley, and Hazardous Waste Criteria Element (1)

Basin Concentration Rangea (2)

Waste Criteriab (3)

Solid phase: Se A B Mo Solution phase: Se As B Mo

(mg/kg) ND–33 ND–8.6 ND–980 ND–94 (mg/L) 0.008–8 ND–12 3–200 1–282

TTLC (mg/kg) 100 500 7,000 3,500 STLC (mg/L) 1.0 5.0 70 350

a

Concentration range found in San Joaquin Valley basins. For Se, As, and Mo, hazardous waste criteria for soluble threshold limit concentration (STLC) and total thresh limit concentration (TTLC) (solid phase) were according to Title 22, California Administration Code. For B only, the criteria are according to Title 23, California Administration Code.

b

ND, none detected. From Tanji et al. (1992), © 1992 Regents of the University of California.

770

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and McVay 1993; Chilcott et al. 1993; Westcot et al. 1993; Ong et al. 1997). Trace element concentrations (chiefly B, As, Se, Mo, U, and V) were much higher than natural background concentrations in seawater and natural saline-sink lakes in the western United States after taking into consideration the average pond age and physiographic areas (Westcot et al. 1993). Generally higher levels of Se and B are found in the west alluvial fan region, and higher As levels are found in the Tulare Lake bed (Westcot et al. 1993). In one of the largest pond facilities in the Tulare Basin, an investigation in 1985–1986 showed that inlet water had median total As and Se concentrations of 97 and 2 g/L, respectively, and pond waters concentrated As to 110 to 420 g/L and Se to 21 to 29 g/L (Fujii 1988). Data for 2004–2005 indicated that As concentrations in inlet water accumulated from 120 up to 1,000 g/L, while Se concentrations were reduced from inlet water of 15 g/L to 10 g/L (Gao et al. 2007a,b). Data reported in the early 1990s indicated that certain elements (B, Se, and As) already approached or exceeded the STLC in some ponds (Ong et al. 1995). Although the sink mechanism resulted in the accumulation of trace elements in sediments, no case exceeding the TTLC has been reported, including the most recent sediment data for As and Se (Ong et al. 1995, 1997; Westcot et al. 1993; Gao et al. 2007a,b). Molybdenum, U, and V were also considered to have some sink mechanisms, while B was considered to be a conservative element. In the MDB, a reconnaissance survey included the analysis of water and sediment samples from five basins in the Murrumbidgee Irrigation Area (MIA) and one in the Shepparton Irrigation Region (SIR) (Christen et al. 2000). The results from this study showed that the concentrations of B, Cu, Cd, Pb, and Mn were, in some instances, above guideline levels for water and sediments (Table 24-2). All of the pesticides included in the study (atrazine, diuron, metalochlor, endosulfan, and chlorpyrifos) were detected above guideline levels for protection of aquatic environments. These pesticides were also found at elevated levels in the sediments. However, the total hardness of such waters is very high (100–300 mg/L). This will greatly reduce the toxicity of heavy metals and possibly the other contaminants. These results indicate that the waters and sediments in the basins should be regularly monitored and the sites treated as potentially contaminated. Further investigation is required to determine the appropriate guideline levels for these waters and hence the appropriate management of these sites. Basin Salinity and General Water Chemistry In the MDB, basin water chemistry is simply related to salt accumulation from evapoconcentration, although there are potential concerns for nutrients and pesticides. The conductivity of basin waters is in the range

TABLE 24-2. Toxicant Level Guidelines and Levels in Basin Water AWQGFMW Guidelinesa

Toxicant (1)

Raw Drinking Water (2)

Use in Aquaculture (3)

Protection of Aquatic Ecosystems (4)

Use for Irrigation (5)

1 0.1

0.5–6 1 0.2–2

0.05–0.1 0.005–0.1 0.002–0.005 0.005–0.05 0.5

5 0.2 2 100b

Boron (mg/L) Iron (mg/L) Manganese (mg/L) Sodium (mg/L) Phosphorus (mg/L) Aluminum (mg/L) Copper (mg/L) Zinc (mg/L) Atrazine (g/L)

1 0.3 0.1 300

Diuron (g/L) Metalochlor (g/L) Endosulfan (g/L)

40 800 40

0.01

0.03 0.02 0.01 (0.001)

Chlorpyrifos (g/L)

2

0.001

0.001

a

0.2 1 5

0.1 300 0.2 0.005 0.05

2

AWQGFMW, Australian Water Quality Guidelines for Fresh and Marine Waters (ANZECC 1992). Not specific; 100 g/L is the guideline for all herbicides in New South Wales Department of Agriculture, Orange, Australia.

b

771

Data from Christen et al. (2000).

Maximum Value Found in Existing Pond Waters (6)

9.6 0.1 0.65 52,000 2,810 0.1 0.058 0.1 2.82 0.25 0.27 5.9 0.57

772

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

of 37 to 162 dS/m in the Wakool-Tullakool evaporation basins (Roberts 1995) and 6 to 80 dS/m for on-farm basins in the MIA, from drainage waters of 2 to 25 dS/m, indicating evapoconcentration of three- to fourfold (Leaney and Christen 2000a). The water is predominantly Na-Cl with associated cations and sulfates. However, a detailed study of basin water chemistries has not been undertaken. In the SJV, drainage waters impounded in the evaporation ponds have high levels of dissolved salts and also trace elements. Water salinity in evaporation ponds increases drastically due to evaporation, as reflected in the increase in total dissolved solids (TDS), EC, and soluble ion concentrations. Pond water ECs ranged from 10 to 120 dS/m, but reached up to 178 dS/m (Ong et al. 1995; Gao et al. 2007a). Pond water can potentially evapoconcentrate up to about 388,000 mg/L, which is more than 10 times seawater salinity of 35,000 mg/L (Johnson et al. 1997). As evaporation continues, certain mineral salts exceed their solubility and precipitate out. This is not reflected in TDS or EC measurements. An evapoconcentration factor (ECF) can be calculated considering the conservative nature of Cl (Tanji 1990): ECF  [Cl]pond water/[Cl]inlet

(24-2)

The ECF can be used to reflect evaporation until halite precipitates form. In multicell drainage evaporation ponds, the ECF increases along the water flow path and can reach 20 at the terminal cell, corresponding to a EC of 120 dS/m (Tanji 1990; Gao et al. 2007a). Most pond-water pHs range from about 8 to above 9, reflecting carbonate-dominated systems. As an example, Fig. 24-6 shows the correlation between chemical constituents in pond waters and Cl concentrations from evapoconcentration in a large basin facility in the Tulare Basin. In the SJV, major elements in evaporation pond waters are Na, Cl, and SO4, and pond-water types are typically identified as either Na-Cl, NaSO4, or Na-Cl-SO4 mixtures (Westcot et al. 1993; Ong et al. 1995). Other constituents present in low levels include Mg, Ca, bicarbonate, K, N, and P. Among the trace elements, B was found to be highest (on the order of mg/L) and others (Se, As, U, V) were on the order of g/L. Some elements (Mg, K, B, and As) also show conservative properties as demonstrated by a positive relationship with EC or Cl concentration. Others, such as Ca, bicarbonate, and sulfate, are not linearly related to increasing salinity due to precipitation. Trace elements B and As sometimes exhibit linear increases with salinity, while Se does not increase and even decreases from the inlet drainage concentrations in a Tulare Basin pond facility (Gao et al. 2007a,b). Precipitation of mineral salts occurs from evapoconcentration when their solubility is exceeded (Chapter 1 in this manual). Minerals with the

AGRICULTURAL EVAPORATION BASINS

773

FIGURE 24-6. Correlation between soluble chemical constituents. (a) Na; (b) SO4-S and Mg; (c) Ca and bicarbonate alkalinity; (d) dissolved organic carbon; (e) B; (f) As and Se with Cl concentration, which is assumed to be a conservative parameter in an evaporation pond facility in Tulare Basin, California. From Gao et al. (2007a) with permission from Elsevier.

774

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

lowest solubility product constant, Ksp (e.g., calcite), precipitate out first. Using a brine chemical equilibrium model, Smith et al. (1995) predicted that the most common evaporites formed early in the evapoconcentration process are calcite and gypsum, followed by thenardite and halite However, the formation of calcite and gypsum is limited due to low levels of Ca. The most common minerals observed in SJV ponds are halite (NaCl), thenardite (Na2SO4), and mirabilite (Na2SO4 10H2O) (Tanji et al. 1992). Other minerals present in minor amounts include gypsum (CaSO4 2H2O), calcite (CaCO3), bloedite [Na2Mg(SO4)2 4H2O], glauberite [(Na2Ca(SO4)2] and nesquehonite (MgCO33H2O) (Westcot et al. 1993; Smith et al. 1995). Trace elements in evaporite minerals are considered minor based on their concentrations (⬃100 times depleted relative to the solution phase) (Ong et al. 1997). However, the presence of these trace elements provides one of the greatest challenges for consideration of harvesting profitable salts from evaporation ponds. Trace Elements and Concerns in San Joaquin Valley Evaporation Basins Drainage disposal in the SJV of California is complex and challenging due to the presence of a number of toxic or potentially toxic trace elements at elevated concentrations. Selenium in evaporation basins continues to be a major concern due to its toxicity to water birds. Intensive research or investigations on pond chemistry in the SJV have provided more detailed chemical information than anywhere else in the world. This information is summarized in the following section. Redox transformations are active in evaporation ponds because of algal and microbial activity. In a 10-cell pond facility, the reducing environment develops along the water flow path as EC of the water increases (Tanji et al. 2006; Gao et al. 2007a). Along this flow path, electron acceptors, such as dissolved oxygen (DO), NO3, and Fe(III) decrease, while reduced products of ammonium (NH4 ), Mn(II), Fe(II), and sulfide increase. Dissolved organic carbon (DOC) also shows a positive correlation with EC, whereas organic matter in the surface sediments does not (average 32–36 g/kg). The initial water-receiving cell had high chlorophyll-a (463 g/L) and DO content (14 mg/L) compared to a cell toward the terminal end (chlorophyll-a of 6 g/L and DO of 4 mg/L) in surface water (Tanji et al. 2006). Although water depth was generally less than 1 m, water collected near the sediment was found to be more reducing than surface water (Tanji et al. 2006). In the initial drainage-receiving cell, DO averaged 3.7 mg/L in the bottom layer water and 14.2 mg/L in the surface layer. The development of reducing conditions in evaporation ponds results in active redox transformations of trace elements, such as Se, As, Mo, U, and V, with the exception of B.

AGRICULTURAL EVAPORATION BASINS

775

Selenium Selenium is the element of most concern in drainage water disposal in the SJV because of its toxicity to wildlife. All oxidation states (VI, IV, 0, and II) of Se have been detected in pond waters or sediments. Greater than 90% of Se in drainage water entering evaporation ponds is in the most oxidized selenate form [Se(VI) or SeO42]. As reducing conditions develop, Se(VI) is reduced to selenite [Se(IV) or SeO2 3 ], elemental Se [Se(0)], selenide, or organic Se [Se(II)]. Selenite is much less soluble because it is more strongly adsorbed onto soil and mineral surfaces than Se(VI). Selenide [Se(II)] can be incorporated into metal selenides (insoluble) and organic compounds that can be volatile, soluble, and solid materials. All of these species were identified in pond waters or sediments, confirming that all major reduction reactions occurred in evaporation ponds (Zhang et al. 1999; Gao et al. 2007b). Formation of reduced species resulted in removal of Se and partitioning into sediments. In a 10-cell pond facility in the Tulare Basin, most pond waters had Se concentrations of 10 g/L, lower than the inlet value of 16 g/L (Gao et al. 2007b). Greater than 95% of the Se in inlet water was Se(VI). Selenite increased to 40% and org-Se up to 27% in some ponds. Elemental Se was determined to be about 39% to 53% of the sediment Se. Selenium incorporated into the organic phase and extracted with NaOH was on average 33% to 49% of the sediment Se where 50% or more was identified as Se(IV). The data indicate the importance of Se(VI) reduction in immobilizing Se into sediments. This removal may help alleviate potential Se risks to water birds to some extent but appears insignificant in places where drainage water contains much higher Se concentrations, such as on Westside of the SJV (Tanji 1990; Ong and Tanji 1993; Ong et al. 1995). Volatilization of Se from pond waters from indigenous algae and bacteria in evaporation ponds was also identified, but volatilization loss accounted for only about 2% of the total Se loss measurement (Frankenberger and Karlson 1989; Fan et al. 1997). Arsenic Arsenic levels are much higher in Tulare Basin soils and drainage waters than in those on Westside of the SJV. Arsenic was found to be as high as 1,000 g/L in pond water of terminal cells of a 10-cell facility in the Tulare Basin (Tanji et al. 2006; Gao et al. 2007a). Total As concentrations increased almost linearly with Cl concentration or ECF (Gao et al. 2007a), illustrating its conservative nature, although some sink mechanisms were identified (Tanji 1990; Ong et al. 1995; Tanji et al. 2006). Arsenic has four oxidation states (V, III, 0, and III). Elemental As is unstable and rare in nature. All other species of As were identified in the ponds. Fresh drainage water and water in the initial drainage-receiving

776

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

cell had 95% arsenate [As(V)], 5% arsenite [As(III)], and negligible org-As. Arsenite and org-As increased to 35% and 14% of the total dissolved As, respectively. Unlike Se, the oxidized form As(V) is more strongly adsorbed onto soil or mineral surfaces than its reduced and more toxic form As(III). A group of organic As species (org-As) was produced from biomethylation (e.g., monomethylarsenic acid and dimethylarsenic acid as well as volatile arsines) (Cullen and Reimer 1989). There have been no direct measurements of As volatilization from pond waters. Reduction of As was not observed in the initial cell (where reduced Se species were detected), indicating that a more reducing condition is required for As reduction compared to Se reduction. Sink mechanisms for As removal from pond waters were identified (Tanji 1990a; Ong et al. 1995). Sediments near the outlet of an initial drainage-receiving cell and sediments toward terminal cells contained up to 80 mg/kg As. Surface sediments accumulated the highest As concentrations. Arsenic concentrations decreased with depth (Ong et al. 1995). Tanji et al. (2006) determined soluble As, adsorbed As, and As associated with carbonates, organic matter (OM), and oxides, using sequential extraction methods with KCl, K2HPO4, NaOAc, NaOCl, and NH2OHHNO3 respectively. Reducing sediments near a terminal cell contained more soluble (19%) and exchangeable As (27%) compared to 13% and 12%, respectively, under oxidizing conditions from the initial drainagereceiving cell. Association with oxides was found to be important under oxidizing conditions (27% vs. 9% under reducing conditions). Incorporation into solid organic phase was relatively small (6%–10%) for As compared to Se, while carbonate minerals played a more important role in immobilizing As into the sediments (26%–35%) under alkaline conditions and a broad range of redox conditions. Other trace elements Boron is not subject to redox transformations but can be adsorbed to metal oxides under oxidizing conditions. Boron generally shows conservative behavior and is positively correlated with EC or Cl concentrations in pond waters. Direct information on transformation of other trace elements (U, V, and Mo) in evaporation pond waters is limited. Some incubation studies from pond waters or sediments indicate that these elements can be subject to reduction. Uranium accumulation was observed in evaporation pond waters to above 1,000 g/L (Westcot et al. 1993). Oxidized U(VI) ion (UO2 2 , uranyl) is unstable at high pH, undergoing hydrolysis and forming complexes with carbonate (CO32), such as (UO2)2CO3(OH)3 , UO2COo3 , UO2(CO3)22 and UO2(CO3)34 (Duff et al. 1999). Ponds with high salinities and high alkalinities contain the highest aqueous U concentration. At highly reduc-

AGRICULTURAL EVAPORATION BASINS

777

ing conditions, U(VI) can be reduced to U(IV), which is sparingly soluble. However, U(IV) can also form soluble complexes with carbonates as U(CO3)44, and U(CO3)56 and OM (Duff et al. 1997). Molybdate [MoO42, Mo(VI)], the most oxidized form, is highly soluble and dominant in drainage waters. Adsorption of Mo to metal oxides, clay minerals, or CaCO3 occurs mostly at low pH and, therefore, may not be important in evaporation ponds because little adsorption occurs at high pH (e.g., above 8) (Goldberg et al. 1996). Incubated sediments under reducing conditions caused substantial loss of Mo from solution, possibly due to reduction and precipitation of molybdenite (MoS2) (Amrhein et al. 1993). Vanadium concentrations in evaporation pond waters were elevated compared to seawater but were in the same range as the western salt-sink lakes (Westcot et al. 1993). Vanadium in nature can exist in oxidation states of 3, 4, and 5; there is not enough direct information on V transformation in pond systems. Vanadium concentrations in bottom-sediment decreased from 40 to 21 mg kg1 in cells following a flow path, indicating more accumulation in initial drainage-receiving cells (Fujii 1988). The surface sediment is organically rich and has been identified as the major sink for all trace elements. The surface 6-cm sediments were found to have elevated levels of B, As, Se, Mo, and U under reducing conditions (Chilcott and McVay 1993). At depths greater than 10 to 12 cm, concentrations for all elements except As returned to background levels. No clear trend of Se and As accumulation with sediment depth was observed from the analysis of 25-cm sediment cores at 5-cm increment (Gao et al. 2007a,b). It should be noted that immobilized reduced forms of trace elements in sediments can be released upon redox status changes (e.g., oxidation). Selenium oxidation and remobilization was reported in surface soils from the closure of Kesterson Reservoir, which was used to store agricultural drainage waters for a number of years (Wahl et al. 1994; Martens and Suarez 1997). Oxidation and remobilization of accumulated trace elements in the soil and sediments become a concern when a pond is allowed to dry out. Pond biology and selenium toxicity Evaporation ponds are biologically active and attract many migratory and resident birds because they are permanent water bodies in otherwise semiarid zones where often other wetland habitats have been lost. Selenium toxicity to water birds has been the primary concern in the SJV (Ohlendorf et al. 1993; Tanji et al. 2002). Cases of Se toxicity to water birds in evaporation ponds were observed (e.g., Ohlendorf et al. 1993; Gordus 1999). Nesting birds at the ponds are mostly shore birds. American avocets and black-necked stilts account for 50% and others include grebes, plovers, and ducks (Ohlendorf et al. 1993). The national chronic criterion

778

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

for Se concentration in freshwater aquatic habitat is 5 g/L (US EPA 1987). An even lower (2 g/L) criterion was recommended based on observed toxic effects (Hamilton and Lemly 1999). Algae and bacteria, along with organic detritus, serve as principal foods for aquatic invertebrates, which in turn provide food for other invertebrates, fish, and birds. Selenium can bioconcentrate from water to phytoplankton and aquatic vegetation by about 500 to 1,000 times, and be further biomagnified at 1.5 to 2.5 times to zooplankton, zooplankton to fish, and rooted plants to birds. The overall bioaccumulation in the food chain to birds and fish is about 1,500 to 2,000 times (Skorupa 1998). Selenium is considered to substitute for sulfur in essential amino acids such as methionine and cystine, leading to reduced reproduction and deformities in embryos. Other trace elements at elevated levels, such as B, As, and Mo, were not identified to cause toxicity symptoms to water birds (Ohlendorf et al. 1993). Wetland habitats were developed in evaporation basins in Australia after 10 years of basin construction (Roberts 1995). A study showed that the Wakool-Tullakool evaporation basins (2,100 ha) acquired a rich aquatic plant flora and became a wetland of local, regional, and possibly international significance for water birds and waders. Vascular plant coverage was extensive (up to 100% in low-EC waters), although the coverage dramatically decreased with increasing EC (zero coverage in waters with EC around 100 dS/m). The presence of aquatic plants and, in particular, of five macrophytes including four Ruppia L. species, is highly significant in terms of species richness, species distribution, and dispersal mechanisms. Sixty species of water birds were recorded in the basins, including 12 ducks and 27 waders (18 migratory). It was suggested that the size of these basins may have contributed to the large and diverse water bird populations. However, some water bird species, such as Australia shelduck (Tadoma tadomoides) and black swan (Cygnus atratus), are a threat to the rice farms near the basins and so present a management problem. Other basins also developed nuisance populations of birds. Regulatory and Environmental Issues and Mitigation Measures In the SJV, discharge of drainage waters and operation of evaporation ponds are highly regulated by the enforcement of state and federal laws regarding water quality and wildlife protection (Campos et al. 1993). WDRs issued by the California Regional Water Quality Control Board, Central Valley Region require “a program of management actions to reduce, avoid, and mitigate for adverse environmental impacts to wildlife” (EPTC 1999; Tanji et al. 2002). Compliance monitoring includes seasonal water and sediment Se concentration, and biological monitoring of birds for abundance and symptoms of toxicity. Continuous flow, monthly

AGRICULTURAL EVAPORATION BASINS

779

salinity, and annual trace element (As, B, Mo, Se, U) characterization are required for inflow. Within the ponds, weekly water depth, monthly salinity, and annual trace elements in waters and sediments must be recorded. Wildlife monitoring includes semimonthly bird breeding and nesting, all included in an annual report. These stringent monitoring requirements and high associated costs have precipitated the closure of many small evaporation pond facilities. When evaporation ponds are closed, a series of actions must be followed (EPTC 1999; Tanji et al. 2002). After removing all free liquids, there are two options: clean-close or close-in-place. Clean-close requires that all residual wastes are completely removed and discharged to an appropriate waste management unit. If this attempt is infeasible, the basin is closed as a landfill. Close-in-place requires that all residual wastes be compacted and the basin closed as a landfill. Closure methods considered are compaction of sediments, coverage with nonhazardous soil materials, and vegetation cover. Methods of closure and postclosure monitoring must minimize environmental impact by preventing wind blowing of sediments off-site, limiting access of wildlife to the sediment, and preventing any substantial impact to groundwater quality. To alleviate Se toxic risk to water birds, a number of mitigation measures have been used (EPTC 1999; Tanji et al. 2002). On-site mitigation includes pond configuration with steep side-slopes of at least 3⬊1, removal of windbreak islands within the ponds, tires for stabilizing banks, removal of vegetation, and a minimum water depth of 1 m to discourage feeding and nesting of shore birds. When unavoidable Se impacts are observed, off-site mitigation measures using alternative or compensation habitats must be implemented. An 84-ha compensation wetland habitat was constructed in the Tulare Basin in an effort to mitigate adverse environmental impacts to wildlife (Tanji et al. 2002). The compensation habitat was to provide a safe foraging and nesting environment for water birds. The habitat has good-quality water with low Se (2 g/L) and low salinity, and nesting islands. The habitat is a flow-through system and also is designed to deter predators (perimeter electric fence). Although the mitigation efforts did not reduce Se intake by stilts that continued to nest at the evaporation basin, the freshwater wetland provided a “clean” foraging area for stilts and successfully attracted stilts away from the evaporation basin (Gordus 1999). The establishment of the compensation habitat resulted in a significantly higher percentage of nests hatched or presumed hatched in adjacent evaporation ponds. After a few years the number of nests at the evaporation ponds declined to nil due to the presence of the compensation habitat (Tanji et al. 2002). In Australia, any development of disposal basins is carried out within the framework of catchment land and water management plans (LWMPs).

780

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

When investigating or implementing the use of disposal basins, local agencies need to be consulted and management plans should be developed in conjunction with the local environmental protection authority. At the time of this writing there was no legislation at the federal, state, or local government level that specifically dealt with the use of local-scale saline disposal basins. However, various aspects of disposal basin siting and use may fall under a range of legislation, regulations, and by-laws (e.g., Victorian EPA 1994; New South Wales EPA 1997). It is, therefore, important that during the planning stage of a basin a thorough investigation of the statutory responsibilities is carried out to ensure legal compliance.

SUMMARY Evaporation basins have been used as a drainage disposal option for irrigated agriculture in many arid and semiarid areas around the world. Various types and sizes of evaporation basins are used at local or regional scales depending on drainage water volumes to be disposed of, hydrology, and management, as well as cost factors. On-farm basins promote more water reuse but in this method some lands are taken out of production and farmers bear the costs of associated construction and maintenance. Regional basins collecting water from multiple farms can be built on marginal agricultural lands but require a master drainage collection system and more management. Most basins are developed on fine-textured soils and are compacted to minimize percolation. Actual seepage loss depends on soil and groundwater conditions. Over time, basin seepage generally has been found to decrease to rates of about 0.1 mm/day. Lateral seepage loss varies with basin size and groundwater conditions, with shallow groundwaters leading to large lateral seepage losses. To prevent seepage from waterlogging and salinizing, adjacent ground-perimeter drains are used around basins to intercept and recycle pond leakage. Various public or environmental concerns over evaporation basin use have been presented. Elevated Se concentration in drainage water and its toxicity to wildlife is a unique case in the SJV of California. Policies and regulations are developed to minimize Se toxic risks to water birds, including closure of ponds presenting high risks and establishment of compensation ponds nearby to divert wildlife. Conversely, where there are no toxic elements, the ponds can be valuable environmental assets that lead to diverse populations of water birds, as experienced in Australia. Other issues including odor, windblown salt, and contamination of groundwater continue to present challenges in the management of evaporation basins. Various monitoring programs have been put in place.

AGRICULTURAL EVAPORATION BASINS

781

REFERENCES Amrhein, C., Mosher, P. A., and Brown, A. D. (1993). “The effects of redox on Mo, U, B, V, and As solubility in evaporation pond soils.” Soil Sci., 155, 249–255. Arroues, K. D., and Anderson, C. H., Jr. (1986). Soil survey of Kings County, California. USDA-Soil Conservation Service, Washington, D.C. Australian and New Zealand Environment and Conservation Council (ANZECC). (1992). Australian water quality guidelines for fresh and marine waters, ANZECC, Canberra, Australia. Blackwell, J. J., Biswas, N. T., and Christen, E. W. (2005). “Evaluation of a sequential biological concentration system in natural resource management of a saline irrigated area.” Aust. J. Water Resour., 9(2), 169–176. California Irrigation Management Information System (CIMIS). (1999). Reference evapotranspiration map, California Department of Water Resources, Sacramento, Calif. Campos, M. A., Mass, L., Puckett, L., and Carlson, R. L. (1993). “Legal and regulatory considerations in agricultural drainage water management,” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen, and C. M. U. Neale, eds., ASCE, Reston, Va., 612–620. Cervinka, V., Dienet, J. Erickson, J., Finch, C., Martin, M., Menezes, F., Peters, D., and Shelton, F. (1999). Integrated system for agricultural drainage management on irrigated farm land, Report 4-FG-20-11920, U.S. Department of the Interior, Bureau of Reclamation, Washington, D.C., 41. Chilcott, J. E., and McVay, L. A. (1993). Sediment quality in evaporation basins used for the disposal of agricultural subsurface drainage water in the San Joaquin Valley, California: 1990–1991. California Central Valley Regional Water Quality Control Board Report, Sacramento, Calif. Chilcott, J. E., Westcot, D. W., Johnson, S. D., and Toto, A. L. (1993). Water quality in evaporation basins used for the disposal of agricultural subsurface drainage water in the San Joaquin Valley, California: 1990–1991, California Central Valley Regional Water Quality Control Board Report, Sacramento, Calif. Christen, E. W., Gray, L., and Spark, K. (2000). A reconnaissance survey of trace metals and pesticides in saline disposal basins in the Riverine Plain, CSIRO Land and Water Technical Report 23/00, CSIRO Land and Water, Griffith, New South Wales, Australia. Cullen, W. R., and Reimer, K. J. (1989). “Arsenic speciation in the environment.” Chem. Rev., 89, 713–764. Duff, M. C., Amrhein, C., and Bradford. G. (1997). “Nature of uranium contamination in the agricultural drainage water evaporation ponds of the San Joaquin Valley, California, USA.” Can. J. Soil Sci., 77, 459–467. Duff, M. C., Hunter, D. B., Bertsch, P. M., and Amrhein, C. (1999). “Factors influencing uranium reduction and solubility in evaporation pond sediments.” Biogeochemistry, 45, 95–114. Duncan, R. A., Bethune, M. G., Thayalakumaran, T., Christen, E. W., and McMahon, T. A. (2008). “Management of salt mobilisation in the irrigated landscape: A review of selected regions.” J. Hydrol., 351, 238–252.

782

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Earl, G. C. (1982). “An engineering solution to dryland salting in the Mallee.” Paper presented to the Focus on Mallee Salinity Seminar, Manangatang [Australia], March 1982, State Rivers and Water Supply Commission Report (unpublished). Evaporation Ponds Technical Committee (EPTC). (1999). Evaporation ponds, final report, The San Joaquin Valley Drainage Implementation Program and the University of California Salinity/Drainage Program, Sacramento, Calif. Fan, T. W. M., Lane, A. N., and Higashi, R. M. (1997). “Selenium biotransformations by a euryhaline microalga isolated from a saline evaporation pond.” Environ. Sci. Technol., 31, 569–576. Ford, S. A. (1988). Agricultural drainage evaporation ponds in the San Joaquin Valley – Progress of the investigation, Memorandum Report, California Dept. of Water Resources, Sacramento, Calif. Frankenberger, W. T. Jr., and Karlson, U. (1989). “Environmental factors affecting microbial production of dimethylselenide in a selenium-contaminated sediment.” Soil Sci. Soc. Am. J., 53, 1435–1442. Fujii, R. (1988). Water-quality and sediment-chemistry data of drain water and evaporation ponds from Tulare Lake Drainage District, Kings County, California, March 1985 to March 1995, U.S. Geological Survey Open-File Report 87-700, U.S. Geological Survey, Sacramento, Calif. Gao, S., Ryu, J., Tanji, K. K., and Herbel. M. J. (2007a). “Arsenic accumulation and speciation in evaporating waters of agricultural evaporation basins.” Chemosphere, 67, 862–871. Gao, S., Tanji, K. K., Dahlgren, R. A., Ryu, J., Herbel, M. J., and Higashi. R. (2007b). “Chemical status of selenium in evaporation basins for disposal of agricultural drainage.” Chemosphere, 69, 585–594. Goldberg, S., Forster, H. S., and Godfrey, C. L. (1996). “Molybdenum adsorption on oxides, clay minerals, and soils.” Soil Sci. Soc. Am. J., 60, 425–432. Gordus, A. W. (1999). “Selenium concentrations in eggs of American avocets and black-necked stilts at an evaporation basin and freshwater wetland in California.” J. Wildlife Mgmt., 63(2), 497–501. Grismer, M. E., Karajeh, F., and Bouwer, H. (1993). “Evaporation pond hydrology,” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen and C. M. U. Neale, eds., ASCE, Reston, Va., 580–586. Gutteridge, Haskins, and Davey (GHD). (1990). Pipeline to the sea: Pre-feasibility study, Report to the Murray-Darling Basin Commission, ACIL Australia, and Australian Groundwater Consultants. Hamilton, S. J., and Lemly, A. D. (1999). “Water-sediment controversy in setting environmental standards for selenium.” Ecotox. Environ. Safety, 44, 227–235. Hoffman, G. J. (1985). “Drainage required to manage salinity.” J. Irrig. Drain. Eng, ASCE, 111(3), 199–206. Hostetler, S., and Radke, B. (1995). An inventory of saline disposal basins, Murray Basin, AGSO Record 94/4, Australian Geological Survey Organisation, Canberra, Australia. Johnson, W. R., Tanji, K. K., and Burns, R. T. (1997). “Drainage water disposal,” in Management of agricultural drainage water quality, C. A. Madramootoo, W. R. Johnston, and L. S. Willardson, eds., Chapter 5, FAO Corporate Document

AGRICULTURAL EVAPORATION BASINS

783

Repository, Water Report 13, International Commission on Irrigation and Drainage, Food and Agriculture Organisation of the United Nations, Rome. Jolly, I., Christen, E. W., Gilfedder, M., Leaney, F., and Walker, G. (2000). On-farm and community-scale salt disposal basins on the Riverine Plain: Guidelines for basin use, CSIRO Land and Water Technical Report 12/00, CRC for Catchment Hydrology Technical report 00/07, CSIRO Land and Water, Canberra, Australia. Leaney, F. W., and Christen, E. W. (2000a). On-farm and community-scale salt disposal basins on the Riverine Plain: Evaluating basin leakage rate, disposal capacity and plume development, CSIRO Land and Water Technical Report 17/00, CRC for Catchment Hydrology Technical Report 00/11, CSIRO Land and Water, Adelaide, Australia. ———. (2000b). On-farm and community-scale salt disposal basins on the Riverine Plain: Basin leakage: Site studies at Girgarree, Victoria and Griffith, New South Wales, CSIRO Land and Water Technical Report 16/00, CRC for Catchment Hydrology Technical Report 00/10, CSIRO Land and Water, Adelaide, Australia. Martens, A., and Suarez, D. L. (1997). “Changes in the distribution of selenium oxidation states with sample storage.” Soil Sci. Soc. Am. J., 26, 1711–1714. Melville, D. K., Reynolds, R. L., Bradford, D. F., Marsh, R. E., and Salmon, T. P. (1993). “Design and management of evaporation ponds,” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen, and C. M. U. Neale, eds., ASCE, Reston, Va., 604–611. Murray-Darling Basin Commission (MDBC). (1999). Salinity and drainage strategy– Ten years on, 1999, Murray-Darling Basin Commission, Canberra, Australia. Murray-Darling Basin Ministerial Council (MDBMC). (1987). Salinity and drainage strategy background paper 87/1, Murray-Darling Basin Commission, Canberra, Australia. New South Wales EPA. (1997). The protection of the environment operations act, 1997 (draft), NSW EPA, Chatswood, Australia. Ohlendorf, H. M., Skorupa, J. P., Saiki, M. K., and Barnum, D. A. (1993). “Foodchain transfer of trace elements to wildlife,” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen, and C. M. U. Neale, eds., ASCE, Reston, Va., 596–603. Ong, C. G., Dahlgren, R. A., Herbel, M. J., and Tanji, K. K. (1997). “Trace element (Se, As, Mo, B) contamination of evaporites in hypersaline agricultural evaporation ponds.” Environ. Sci. Technol., 31, 831–836. Ong, C. G., Tanji, K. K. (1993). “Evaporative concentration of trace elements in a multi-cell agricultural evaporation pond.” J. Agric. Food Chem., 41, 1507–1510. Ong, C. G., Tanji, K. K., Dahlgren, R. A., Smith, G. R., and Quek, A. F. (1995). “Water quality and trace element evapoconcentration in evaporation ponds for agricultural waste water disposal.” J. Agric. Food Chem., 43, 1941–1947. Roberts, J. (1995). “Evaporation basins are wetlands.” Aust. J. Environ. Mgmt., 2(1), 7–18. San Joaquin Valley Drainage Program (SJVDP). (1990). A management plan for agricultural subsurface drainage and related problems on the Westside San Joaquin Valley, Final report, September 1990, San Joaquin Valley Drainage Program, Sacramento, Calif.

784

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Singh, J., and Christen, E. W. (2000). On-farm and community-scale salt disposal basins on the Riverine Plain: Cost comparisons of on-farm and community disposal basins: Case studies of the MIA and Wakool, CSIRO Land and Water Technical Report 14/00, CRC for Catchment Hydrology Technical report 00/08, CSIRO Land and Water, Griffith, Australia. ———. (2001). “Evaporation basins: opportunities for cost minimisation in siting design and construction.” Irrig. and Drain., 50, 19–29. Skorupa, J. P. (1988). “Selenium poisoning of fish and wildlife in nature: Lesson from twelve real-world examples,” in Environmental chemistry of selenium, W. T. Frankenberger, Jr., and R. A. Engberg, eds., Marcel Dekker, New York, 315–354. Smith, G. R., Tanji, K. K., Burau, R. G., and Jurinak J. J. (1995). “C Salt: A chemical equilibrium model for multicomponent solutions,” in Chemical equilibrium and reaction model, R. H. Loeppert, A. P. Schwab, and S. Goldberg, eds., Chapter 15, SSSA Special Publication No. 42, Soil Science Society of America, Madison, Wisc., 289–324. State Rivers and Water Supply Commission (SRWSC). (1978). Shepparton region drainage and the Lake Tyrrell scheme. Part 1, State Rivers and Water Supply Commission, Melbourne, Australia. Tanji, K. K. (1990). “Accumulation of salts and trace elements in agricultural evaporation ponds.” Trans. 14th International Congress of Soil Science, Vol. VII, 80–185. Tanji, K. K., Davis, D., Hanson, C., Toto, A., Higashi, R., and Amrhein, C. (2002). “Evaporation ponds as a drainwater disposal management plan.” Irrig. Drain. Syst., 16(6), 279–296. Tanji, K. K., Ford, S., Toto, A., Summers, J., and Willardson, L. (1993). “Evaporation ponds: What are they, why some concerns?” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen, and C. M. U. Neale, eds., ASCE, Reston, Va., 573–579. Tanji, K. K., Gao, S., and Ryu, J. (2006). Selenium mass balance and modeling in agricultural evaporation basins, California Department of Water Resources Contract No. 460000567, Final Report, October 10, 2006, California Department of Water Resources, Sacramento, Calif. Tanji, K. K., Grismer, M. E., and Hanson, B. R. (1985). “Subsurface drainage evaporation ponds.” Calif. Agr., Sept-Oct., 1985. Tanji, K. K., and Kielen, N. C. (2002). “Drainage water disposal,” in Agricultural drainage water management in arid and semi-arid areas, FAO Irrigation and Drainage Paper 61, Food and Agriculture Organisation of the United Nations, Rome, 91–99. Tanji, K. K., Läuchli, A., and Meyer, J., (1986). “Selenium in the San Joaquin Valley.” Environment, 28(6), 6–11, 34–39. Tanji, K. K., Ong, C. G., Dahlgren, R. A., and Herbel M. J. (1992). “Salt deposits in evaporation ponds: An environmental hazard?” Calif. Agric., 46(6), 18–21. U.S. Environmental Protection Agency (US EPA). (1987). Ambient water quality for selenium-1987, Publication EPA-440 5-87-006, USEPA, Office of Water Regulations and Standards, Washington, D.C. Victorian EPA. (1994). Draft state environment protection policy (groundwaters of Victoria), Publication No. 288, Victorian EPA, Melbourne, Australia.

AGRICULTURAL EVAPORATION BASINS

785

Wahl, C., Benson, S., and Santolo, G. (1994). “Temporal and spatial monitoring of soil selenium at Kesterson Reservoir, California.” Water Air Soil Pollu., 74, 345–361. Westcot, D. W., Chilcott, J. E., and Smith, G. (1993). “Pond water sediment and crystal chemistry,” in Management of irrigation and drainage systems: Integrated perspectives, Proc. ASCE 1993 Natl. Conf. of Irrigation and Drainage Engineers, Park City, Utah, 21–23 July 1993, R. G. Allen, and C. M. U. Neale, eds., ASCE, Reston, Va., 587–594. Zhang, Y. Q., Moore, J. N., and Frankenberger, W. T. Jr. (1999). “Speciation of soluble selenium in agricultural drainage waters and aqueous soil-sediment extracts using hydride generation atomic absorption spectrometry.” Environ. Sci. Technol., 33, 1652–1999.

NOTATION [Cl]inlet  chloride concentration in inlet water to a pond facility [Cl ]pond water  chloride concentration in pond water E  evapotranspiration rate EC  electrical conductivity ECF  evapoconcentration factor ET  evapotranspiration ET0  reference ET Y  empirical correction factor 

This page intentionally left blank

CHAPTER 25 SALINITY ASSESSMENT OF IRRIGATION WATER USING WATSUIT Laosheng Wu, Christopher Amrhein, and James D. Oster

INTRODUCTION Managing salinity in irrigated lands is a major challenge for sustainable agricultural production. Chemical reactions in the rootzone affect salinity, sodicity, and the salt load of the drainage water. These reactions include dissolution and precipitation of carbonate-, sulfate-, and silicateminerals; cation exchange between the major cations in solution (Na, Ca, Mg, and K) and clay mineral surfaces; adsorption and desorption of carbon dioxide gas, and ion pair formation within the soil solution. Due to mineral precipitation and dissolution reactions in the soil, the amount of salt leached may be greater than, equal to, or less than that applied in the irrigation water. Early attempts to account for these reactions in soils resulted in empirical equations to predict rootzone salinity, drainage water salinity, leaching requirements, and the sodium hazard of the irrigation water (Rhoades 1968). In this chapter we describe how a model can be used to keep track of the various chemical reactions in the rootzone and project soil solution chemical composition. Evolution of the WATSUIT Model The sodium adsorption ratio (SAR) is the classic approach for assessing the sodium hazard of irrigation water (U.S. Salinity Lab Staff 1954). Soil scientists realized early on that the SAR was sensitive to dissolution and precipitation of calcium carbonate, and one of the first attempts to “adjust” the SAR was proposed by Bower et al. (1968). This correction was based on the Langelier Saturation Index but did not properly consider the effects of calcium activity, alkalinity, and CO2 on calcite equilibrium, prompting 787

788

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Suarez (1981) to publish another correction. This new version of the “adjusted SAR” was popularized following inclusion in Ayers and Westcot (1985). Keeping track of chemical reactions in the rootzone was one of the first applications of computers in soil science. Dutt (1962), Dutt and Tanji (1962), and Tanji (1969) were the first to demonstrate the use of computer programs to predict soil solution chemical composition. Rhoades and Merrill (1976) developed WATSUIT based on the chemistry model developed by Oster and McNeal (1971) and modified by Oster and Rhoades (1975). A version of the model that runs in DOS is available from the George E. Brown Salinity Laboratory (USDA/ARS 1991), and an updated version of the model with a Windows-based interface is available at the University of California, Riverside (UCR 2003).

WATSUIT Basics WATSUIT calculates the chemical composition and electrical conductivity (EC) of soil water and drainage water based on the composition of the applied irrigation water and various management practices, including leaching fraction and amendment additions. The program calculates an average rootzone salinity, which can help select crops based on salttolerance data. The WATSUIT model is a “steady-state model” and therefore assumes that a particular leaching fraction remains constant over time. Consequently, cation exchange and adsorption are assumed to be at equilibrium. This would occur in a soil that had been used the same irrigation management practice for many years. Realistically, this rarely occurs and leaching fractions change from year to year and season to season. However, over long time periods (decades) an average leaching fraction (LF) can be assumed based on soil properties, irrigation methods, and average rainfall. This model gives a first-approximation of the long-term average chemistry of the rootzone and drainage water.

Rootzone Water Uptake In this model, the rootzone is divided into quarters and a relative water uptake of 40%, 30%, 20%, and 10% is assumed (Rhoades and Merrill 1976). That is, 40% of the total evapotranspiration (ET) is extracted from the upper quarter of the rootzone, 30% from the second quarter, 20% from the third quarter, and 10% from the lowest portion of the rootzone. The depth of the rootzone is assumed to be constant but it is not defined. Water that is not transpired from the rootzone goes on to become

SALINITY ASSESSMENT USING WATSUIT

789

drainage. The drainage volume can be calculated from the LF, which is predetermined by the operator. Several LFs can be chosen initially in order to compare changes in soil water chemistry as LF varies. The available LF values in the model are 0.05, 0.1, 0.2, 0.3, and 0.4 (5%, 10%, 20%, 30%, and 40% of the applied water goes to drainage below the rootzone). The model reports the LF at the surface of the soil (1.0) and at the bottom of each quarter of the rootzone based on the 40%, 30%, 20%, 10% water uptake distribution. The model also reports 1/LF, or the concentration factor, for these same depths, and it is used to initialize the ion concentrations for each depth. User-Controlled Variables WATSUIT calculates the concentrations of the major cations (Na, 2 2 Ca2, Mg2, and K) and anions (Cl, HCO 3 , CO3 , and SO4 ) in the soil water at each depth based on the composition of irrigation water, the LF for each soil depth, and whether gypsum or sulfuric acid was added to the irrigation water. The model allows for distinguishing between calcareous and noncalcareous soils by checking a box labeled “Saturate with CaCO3.” If this box is checked, the soil is considered calcareous and the applied water is brought to calcium carbonate equilibrium at each soil depth. The partial pressure of CO2 increases with depth, which affects the amount of calcium carbonate that will dissolve or precipitate throughout the soil profile. If the box labeled “Saturate with CaCO3“ is not checked, calcite—also referred to as “lime” in the model—is allowed to precipitate if saturation is reached in the soil solution, but dissolution does not occur. The solubility product constant for calcite in the model is set at 108.0 (Ca2 activity  CO32 activity), which is substantially higher than the constant for pure calcite (108.48) or aragonite (108.19). This higher solubility constant is consistent with field and laboratory observations of calcite solubility in soils (Suarez 1977; Suarez and Rhoades 1982; Suarez et al. 1992; Lebron and Suarez 1996). Gypsum (CaSO4 2H2O) is allowed to precipitate in the rootzone if the solubility product (2.45  105) is exceeded in the soil solution. Ion activities are calculated at each depth, taking into account ion pair formation and ionic strength effects on the activity coefficients. Magnesium carbonate precipitation does not occur in WATSUIT, although an early version of the model had it as a possible solid phase. The model calculates soil solution pH using the chemical composition of the soil water and a partial pressure of CO2(PCO2) that increases with depth through the rootzone. At the surface, the PCO2 is set at 0.07 kPa and 0.5, 1.5, 2.3, and 3.0 kPa going from the upper quarter of the rootzone to the lowest quarter. The PCO2 cannot be adjusted by the user.

790

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Chemical Amendments To facilitate the infiltration and leaching, soil and water amendments are often added for reclaiming salt-affected soils. A common amendment for sodic soils is gypsum, which provides soluble Ca2 that reduces SAR and at the same time increases the salinity of the soil water, both of which help to maintain percolation rate. Gypsum can be added to the irrigation water at rates equivalent to 1 mmolc L1 or 20 mmolc L1 (in the model, the check-boxes are labeled “1 Eq. CaSO4” and “20 Eq. CaSO4”). These amendment rates are equivalent to adding 86 g gypsum per m3 water and 1.72 kg gypsum per m3 of irrigation water, respectively (230 and 4,600 lbs gypsum per acre-foot). The 1 mmolc L1 can be considered a low dose of calcium that will maintain soil structure at the soil surface and maintain infiltration rates. This might be similar to the calcium addition from a gypsum injector or a light top-dressing of gypsum on the soil. The 20 mmolc L1 rate will facilitate reclamation of a sodic soil with high exchangeable sodium. The user has another amendment option, which is adding sulfuric acid to remove 90% of the irrigation water alkalinity (HCO3  CO3). In this reaction, the sulfate from the acid replaces an equivalent amount of bicarbonate in the water. Adding sulfuric acid to water will increase the dissolution of soil calcium carbonate and aid in removing exchangeable sodium from the soil. The user can select any or all of these amendment options and several LF values, resulting in a large matrix of output. The “no amendment” option is the default condition and is always run.

Chemical Composition The chemical composition of the applied water (irrigation water) has to be in milliequivalent (meq L1) or millimole of charge (meq L1, or mmolc L1), which are equal. The concentrations of the following ions are required to run the model: calcium, magnesium, sodium, potassium, chloride, sulfate, and bicarbonate. If the chemical composition data are given in units of mg L1 (ppm), then the conversion to mmolc L1 (meq L1) requires division of the mg L1 by factors given in Table 25-1. If the analyTABLE 25-1. Conversion Factors to Obtain mmolc L1 from mg L1 Concentrations Ions (1)

Ca2 (2)

Mg2 (3)

Na (4)

K (5)

Cl (6)

SO42 CO2 HCO NO 3 3 3 (7) (8) (9) (10)

Conversion Factors

20.0

12.2

23.0

39.1

35.5

48.0

30

61.0

62.0

SALINITY ASSESSMENT USING WATSUIT

791

sis includes carbonate, add its concentration (in mmolc L1) to bicarbonate before entering the bicarbonate concentration. If the analysis includes nitrate, add its concentration to chloride before entering the chloride concentration. If the chemical analysis is not balanced (i.e., the sum of the cations does not equal the sum of the anions), the program will balance the solution. WATSUIT adjusts the concentrations to achieve charge balance as follows: if the sum of cations exceeds the sum of anions, the difference is added to the Cl. If the sum of anions exceeds the sum of cations, the program will subtract the difference from the Cl, then SO4, then HCO3. For the soil-water composition at the soil surface, the ion concentrations of the irrigation water are used. For each rootzone interval, calculation begins by dividing the ion concentrations by the LF appropriate to the LF at the bottom boundary of the interval. WATSUIT also calculates the EC and SAR of the soil water at the soil surface, the bottom of each quarter of the rootzone, an average for the whole rootzone, and an average of the upper half of the rootzone. The EC of the soil water is calculated from the calculated ion composition of the soil water using method 3 of McNeal et al. (1970). The EC values are expressed in terms of decisiemens per meter (dS m1), which is equal to millimhos per cm (mmho cm1). To convert the EC of the soil solution or drainage water to “total dissolved solids” (TDS), the following rules-of-thumb can be used: TDS (mg L1) ⬇ EC (dS m1)  640

when 0.1  EC  5.0 dS m1

TDS (mg L1) ⬇ EC (dS m1)  800

when EC  5.0 dS m1

Sodium Adsorption Ratio The SAR has been used to predict potential problems with a soil’s physical properties, including permeability, saturated hydraulic conductivity, aeration, tilth, and tendency to “hard-set” (Sumner 1993). The SAR is defined as the ratio Na/兹苶 ([Ca 苶 苶苶 Mg 苶/2 ] 苶 ) where Na, Ca, and Mg concentrations are in mmolc L1 (U.S. Salinity Lab Staff 1954). Dissolution and precipitation reactions within the soil can have a dramatic effect on the SAR of the soil water. Early attempts to correct for calcite dissolution and precipitation resulted in less-than-correct equations to calculate an “adjusted SAR.” Clearly, a computer model that calculates the actual amount of calcium minerals dissolved or precipitated will give a more accurate estimate of the SAR of the soil water. WATSUIT calculates the SAR of the soil water at the soil surface, the bottom of each quarter of the rootzone, an average for the whole rootzone, and an average of the upper half of the rootzone.

792

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

WATSUIT Outputs The chemical analyses for five irrigation waters are shown in Table 25-2 and will be used as examples of how the WATSUIT model works. Figure 25-1 shows the calculated EC of drainage water (ECdw) for three of the rivers in Table 25-2 when used for irrigation at various LFs. If no mineral precipitation reactions are considered, the ECdw is substantially higher than the WATSUIT predictions, except for the low-salinity Feather River water applied at high LFs (low values of 1/LF). Under these conditions, oversaturation with respect to calcium carbonate does not occur in the rootzone, and the simple equation ECdw  ECiw/LF holds true. The Colorado and Pecos Rivers have high salinities and the evapoconcentration in rootzone leads to extensive calcite and gypsum precipitation, thereby reducing the drainage water EC and the average rootzone salinity (Fig. 25-1). Table 25-3 shows the output from a simulation of the Feather River water applied to a calcareous soil with an LF of 0.05. Comparing the chemical composition of the “0 DEPTH” with the irrigation water shows that calcite is dissolved at the surface of the soil, adding 1.3 mmolc L1 Ca and HCO3CO3 to the water. The laboratory analysis was slightly imbalanced and the program subtracted 0.05 mmolc L1 Cl from the input concentration. Magnesium, Cl, NaK, and SO4 were all conservative and increased in concentration in direct proportion to the concentration factor (1/LF). The last entry (DEPTH 4) gives the ion concentrations of soil water leaving the rootzone, which is the water that becomes deep percolation or drainage water. The LF for each soil layer is given in Table 25-3, with the soil surface getting a 100% leaching (all of the irrigation water passes the surface) and the bottom of the fourth layer having an LF of 0.05. The second half of Table 25-3 gives the calculated pH at each depth, the Ca/Mg ratio, the sum of cations (SUM CAT) in mmolc L1, the calculated TABLE 25-2. Chemical Composition of Irrigation Waters from the Western United States River (1)

Caa (2)

Mga (3)

Naa (4)

Ka (5)

Cla (6)

SO4a (7)

HCO3a (8)

ECb SARc (9) (10)

Feather Porterville Colorado Gila

0.45 0.10 6.95 7.22

0.36 0.05 3.63 5.88

0.20 4.90 3.4 18.55

0.04 0.05 0.22 0.09

0.08 1.70 1.03 20.17

0.16 1.25 9.31 8.48

0.86 2.10 3.73 3.17

0.1 0.3 0.6 17.8 1.3 1.5 3.1 7.3

16.98

9.07

11.38

0.08

12.13

22.39

3.11

3.3

Pecos

in mmolc L1. in dS m1. c in (mmolc L1)1/2. a

b

3.2

SALINITY ASSESSMENT USING WATSUIT

793

FIGURE 25-1. Calculated drainage-water electrical conductivity (ECdw) with and without calcite and gypsum precipitation (ppt) in the soil. EC (dS m1), SAR, and the amount of “lime” and gypsum dissolved (negative values) or precipitated (positive values or less negative values than the previous depth). The soil solution pH is approximately a whole pH unit lower than is typically reported for calcareous soils, with the exception of the surface sample. This is due to the increasing PCO2 as a function of soil depth. Changes in the alkalinity (HCO3CO3) with depth due to calcite dissolution and evapoconcentration, and the increasing PCO2 with depth result in a nonsimple change in the pH with depth. Soil solutions with Ca/Mg ratios less than 1 can be a problem for some plants and can be used to recommend gypsum amendment. In the Feather River water example (Table 25-3), soil calcium carbonate (lime) is dissolving in the first two soil depths but starts precipitating in the third and fourth depths (explained in more detail below). The soil solution never reaches saturation with respect to gypsum in this example. Interpretation of CaCO3 and Gypsum Precipitated or Dissolved Precipitation is indicated by positive values, given in units of concentration, in the columns marked “LIME” and “GYP” in the output file (Table 25-3). Negative values indicate minerals have dissolved. WATSUIT deals with ion and calcite and gypsum concentrations, not the mass of

794

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 25-3. WATSUIT Output from a Simulation of Feather River Water Applied to a Calcareous Soil with a Leaching Fraction of 0.05 (5% of the Applied Water Goes to Drainage) Case ID: Feather River water, applied to a calcareous soil Amendments: None Leaching fraction treatment: 0.05 All concentrations are in mmolc L1 Depth

0 1 2 3 4

LF

1/LF

Ca

Mg

NaK

CL

CO3

HCO3

SO4

1.00 0.62 0.33 0.14 0.05

1.00 1.61 3.03 7.14 20.0

1.79 3.29 4.61 4.63 3.32

0.36 0.58 1.09 2.57 7.20

0.24 0.39 0.73 1.71 4.80

0.03 0.05 0.09 0.21 0.60

0.43 0.42 0.41 0.43 0.50

1.76 3.54 5.44 7.14 11.02

0.16 0.26 0.48 1.14 3.20

Depth

pH

Ca/Mg

SUM CAT.

EC

SAR

LIME

GYP

0 1 2 3

8.20 7.63 7.33 7.26

4.96 5.67 4.23 1.80

2.39 4.26 6.43 8.92

0.28 0.42 0.58 0.77

0.19 0.23 0.36 0.75

–1.34 –2.57 –3.25 –1.42

0.00 0.00 0.00 0.00

4

7.31

0.46

15.32

1.24

1.74

5.68

0.00

SUM CAT.  sum of cations (mmolc L1) EC  electrical conductivity (dS m1) SAR  sodium adsorption ratio GYP  gypsum

calcite and gypsum precipitated. Also, conversion of calcite and gypsum concentrations for each depth to mass of each mineral is not a simple matter. This would require knowing the volume of applied water. Also, the calcite and gypsum concentrations for each depth do not take into account the corresponding concentrations at other depths. They are the concentrations that result from dividing the ion concentrations of the irrigation water by the LF for each depth and correcting them for the solubility of calcite and gypsum. WATSUIT Outputs: Summary Table The second output table from WATSUIT is a summary table that contains the most important information—the average rootzone salinity and

SALINITY ASSESSMENT USING WATSUIT

795

average chloride concentration as a function of LF. The average rootzone salinity (AVG. EC) can be used to estimate yield reductions based on salinity and the average Cl concentration (AVG. CL) used to predict chloride toxicity. Table 25-4 shows the summary table for the Feather River water applied to a calcareous soil at five different LFs. It is important to remember that all of the output concentrations and EC values are for the actual soil water. Historically, salinity and toxicity assessment have been based on the EC and Cl concentration of a saturation paste extract. A handy rule is the actual soil solution is two times more concentrated than a saturation paste extract. In order to compare the AVG. EC values from WATSUIT with the salinity tolerance threshold EC values (often called the Mass-Hoffman salt tolerance coefficients), divide the AVG. EC by 2 (Grieve et al. 2011). If the 1⁄2 AVG. EC exceeds the salt tolerance threshold value for a particular crop, a yield reduction due to salinity stress can be expected at that LF. Higher LFs reduce the AVG. EC values and allow for more salt-sensitive crops to be grown without a loss in yield.

TABLE 25-4. Summary Data from WATSUIT Output for Feather River Water Applied to a Calcareous Soil All leaching fractions were run without amendments applied. LF

0.05 0.10 0.20 0.30 0.40

AVG. EC

UP. EC

AVG. SAR

UP. SAR

AVG. CL

UP. CL

0.63 0.57 0.54 0.52 0.51

0.43 0.42 0.42 0.41 0.41

0.58 0.43 0.31 0.25 0.22

0.25 0.24 0.22 0.21 0.20

0.17 0.11 0.08 0.06 0.05

0.05 0.05 0.05 0.04 0.04

SUR. EC  0.28 SUR. SAR  0.19 Program Options Used: CaCO3 forced to saturation LF  leaching fraction AVG. EC  Average electrical conductivity of the whole rootzone. Units are dS m1. UP. EC  Average electrical conductivity of the upper half of the rootzone (layers 1 and 2). AVG. SAR  Average sodium adsorption ratio (SAR) of the soil water in the rootzone. Units are (mmol L1)1/2. UP. SAR  Average SAR of the upper half of the rootzone (layers 1 and 2). AVG. CL  Average chloride concentration in the rootzone in mmolc L1. UP. CL  Average chloride concentration in the upper half of the rootzone (layers 1 and 2). SUR. EC  Electrical conductivity of the surface soil in dS m1. SUR. SAR  Sodium adsorption ratio of the surface soil in (mmolc L1)1/2.

796

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Likewise, divide the AVG. CL (average chloride concentrations) by 2 to compare with the Cl toxicity threshold values published by Grieve et al. (2011). For the Feather River water example (Table 25-4), the AVG. EC and AVG. CL values indicate that this water can be used with an LF of 0.05 on a calcareous soil without exceeding the EC or Cl threshold values for even the most sensitive crops. The EC of the surface layer of the soil (SUR. EC) should be used in conjunction with surface SAR (SUR. SAR) to estimate problems associated with slaking, dispersion, crusting, and reductions in infiltration; see, for example, Fig. 21 in Ayers and Westcot (1985). If the SUF. EC is less than 0.5 dS m1, poor infiltration and crusting problems can occur even when SUR. SAR equals zero. When the surface SAR equals 10, the surface EC should be greater than 1 dS m1 to prevent problems with infiltration and crusting. At higher SARs, SUR. EC needs to be greater than 0.1 times the surface SAR to prevent problems with infiltration and crusting. These guidelines should be used with care, since infiltration characteristics vary considerably among soils. Irrigation Water Amendments The “Porterville water” (Table 25-2) is a difficult water to manage for irrigation because the low salinity (EC  0.6 dS m1) is coupled with a high SAR (17.8 (mmol L)1/2), which leads to dispersion and slaking at the soil surface, crusting, and low infiltration rates. Adding gypsum is particularly effective in improve the suitability of this water for irrigation. If no amendments are added to the Porterville water, the average rootzone SAR ranges from 25 to 40, increasing with decreasing LF (Table 25-5). Adding 1 mmolc L1 gypsum to the irrigation water decreases the surface SAR to 6.5 (mmolc L1)1/2 and the rootzone SAR ranges from 9 to 21, depending on LF. Addition of 20 mmolc L1 gypsum to the irrigation water decreases the surface SAR to 1.6 (mmolc L1)1/2 and the rootzone SAR ranges from 3 to 7 (mmolc L1)1/2, depending on the LF. The addition of sulfuric acid to remove 90% of HCO3 reduces the bicarbonate from 2.10 to 0.21 mmolc L1. Table 25-5 shows the results of running WATSUIT with and without checking the box marked “Saturate with CaCO3,” which simulates adding the acid-treated water to noncalcareous and calcareous soil. When applied to a noncalcareous soil, the acid-treated water had no effect on the average SAR of the rootzone (Table 25-5). In order for sulfuric acid to be an effective amendment, the soil must contain calcium carbonate. Rerunning the program with the variable “Saturate with CaCO3” checked lowers the surface SAR to 4.7 (compared to 17.9 in the noncalcareous soil), and the average rootzone SAR ranged from 6 to 14, making sulfuric acid an effective treatment if the water is applied to a calcareous soil (Table 25-5).

SALINITY ASSESSMENT USING WATSUIT

797

TABLE 25-5. WATSUIT Summary Table for the Porterville Water Comparing the Effects of Gypsum and Sulfuric Acid Amendments Amendment: None, noncalcareous soil. LF

0.05 0.10 0.20 0.30

AVG. EC

UP. EC

AVG. SAR

UP. SAR

AVG. CL

UP. CL

2.70 1.87 1.30 1.05

0.94 0.88 0.81 0.75

39.9 32.5 27.9 25.2

23.6 23.0 22.1 21.3

9.75 6.57 4.51 3.60

3.17 2.99 2.72 2.51

UP. SAR

AVG. CL

UP. CL

9.75 6.57 4.51 3.60

3.17 2.99 2.72 2.51

AVG. CL

UP. CL

9.75 6.57 4.51 3.60

3.17 2.99 2.72 2.51

SUR. EC  0.55

SUR. SAR  17.9

Amendment: 1 meq/L CaSO4, noncalcareous soil. LF

0.05 0.10 0.20 0.30

AVG. EC

UP. EC

2.92 2.07 1.48 1.20

1.07 1.01 0.93 0.86

SUR. EC  0.63

AVG. SAR

20.61 13.63 10.08 9.10

8.53 8.31 7.97 7.69

SUR. SAR  6.46

Amendment: 20 meq/L CaSO4, noncalcareous soil. LF

0.05 0.10 0.20 0.30

AVG. EC

UP. EC

4.57 3.78 3.20 2.94

2.78 2.73 2.60 2.50

SUR. EC  1.86

AVG. SAR

7.33 4.84 3.29 2.62

UP. SAR

2.36 2.22 2.05 1.92

SUR. SAR  1.58

Amendment: H2SO4, noncalcareous soil. LF

0.05 0.10 0.20 0.30

AVG. EC

UP. EC

AVG. SAR

UP. SAR

AVG. CL

UP. CL

2.85 2.00 1.40 1.13

1.01 0.95 0.87 0.81

37.7 32.5 27.9 25.2

23.6 23.0 22.1 21.3

9.75 6.57 4.51 3.60

3.17 2.99 2.72 2.51

UP. SAR

AVG. CL

UP. CL

9.75 6.57 4.51 3.60

3.17 2.99 2.72 2.51

SUR. EC  0.59

SUR. SAR  17.9

Amendment: H2SO4, calcareous soil. LF

0.05 0.10 0.20 0.30

AVG. EC

UP. EC

AVG. SAR

3.12 2.27 1.67 1.40

1.21 1.16 1.08 1.02

13.7 9.9 7.2 5.9

SUR. EC  0.71

SUR. SAR  4.65

Same abbreviations as in Table 25-4.

6.00 5.70 5.27 4.93

798

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SUMMARY AND CONCLUSION Management that reduces leaching will increase the salt concentration in the rootzone and reduce the volume and salt load of the drainage water. Computer programs of varying complexity exist that can help scientists and irrigation managers evaluate the potential effects of irrigation management, soil and water amendments, and soil type on salt accumulation and leaching. However, no one model will serve all needs perfectly and the availability of many types of soil/water/plant models is necessary depending on intended use. Steady-state models, like WATSUIT, require the least number of inputs and can be used as a first-approximation to assess water suitability and management needs to prevent salinization of the soil and yield loss. Transient-state models are available (see Chapter 28) and can be used to assess the salinity and sodicity hazards of irrigation water. Transient-state models can account the following effects: consequences of changing salinity with time on crop growth and yield (Cardon and Letey 1992; Bradford and Letey 1992); salinity-induced reduction in plant growth and transpiration; and extra water uptake from portions of the rootzone where water and salinity stresses are low to compensate for reduced water uptake from zones where these stresses are higher (Pang and Letey 1998; Shani et al. 2007). When these factors are taken into account, plant response to salinity is more dependant on the zone where most of the water uptake occurs (Letey and Feng 2007) than on the linear average soil-water salinities for the entire rootzone resulting from the 40%, 30%, 20%, and 10% water uptake distribution used in WATSUIT. This finding would favor the use of the upper-layer EC (UP. EC) instead of AVE. EC when using the output of WATSUIT to determine the leaching requirement.

REFERENCES Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Bower, C. A., Ogata, G., and Tucker, J. M. (1968). “Sodium hazard of irrigation waters as influenced by leaching fraction and by precipitation or solution of calcium carbonate.” Soil Sci., 106, 29–34. Bradford, S., and Letey, J. (1992). “Cyclic and blending strategies for using nonsaline and saline waters for irrigation.” Irrig. Sci., 13, 123–128. Cardon, G. E., and Letey, J. (1992). “Plant water uptake terms evaluated for soil water and solute movement models.” Soil Sci. Soc. Am. Proc., 32, 1876–1880. Dutt, G. R. (1962). “Prediction of the concentration of solutes in soil solutions for soil systems containing gypsum and exchangeable Ca and Mg.” Soil Sci. Soc. Am. Proc., 26, 341–343.

SALINITY ASSESSMENT USING WATSUIT

799

Dutt, G. R., and Tanji, K. K. (1962). “Predicting concentrations of solutes in effluents.” J. Geophys. Res., 67, 3437–3439. Grieve, C. M., Gratten, S. R., and Mass, E. V. (2011). “Plant salt tolerance,” Chapter 13, this volume. Lebron, I., and Suarez, D. L. (1996). “Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25°C and pH 7.5.” Geochim. Cosmochim. Acta, 60, 2765–2776. Letey, J., and Feng, G. L. (2007). “Dynamic versus steady-state approaches to evaluate irrigation management of saline waters.” Agric. Water Mgmt., 91, 1–10. McNeal, B. L., Oster, J. D., and Hatcher, J. (1970). “Calculation of electrical conductivity from solution composition data as an aid to in-situ estimation of soil salinity.” Soil Sci., 110, 405–414. Oster, J. D., and McNeal, B. L. (1971). “Computation of soil solution composition variation with water content for desaturated soils.” Soil Sci. Soc. Amer. Proc., 35, 436–442. Oster, J. D., and Rhoades, J. D. (1975). “Calculated drainage water compositions and salt burdens resulting from irrigation with river waters in the western United States.” J. Environ. Qual., 4, 73–79. Pang, X. P., and Letey, J. (1998). “Development and evaluation of ENVIRO-GRO, an integrated water, salinity, and nitrogen model.” Soil Sci. Soc. Amer. J., 62, 1418–1427. Rhoades, J. D. (1968). “Mineral-weathering correction for estimating the sodium hazard of irrigation waters.” Soil Sci. Soc. Amer. Proc., 32, 648–652. Rhoades, J. D., and Merrill, S. D. (1976). Assessing the suitability of water for irrigation: Theoretical and empirical approaches, FAO Soils Bulletin 31, Food and Agriculture Organisation of the United Nations, Rome, 69–109. Shani, U., Ben-Gal, A., Tripler, E., Dudley, L. M. (2007). “Plant response to the soil environment: An analytical model integrating yield, water, soil type, and salinity, 2007.” Water Resour. Res., 43. Suarez, D. L. (1977). “Ion activity products of calcium carbonate in waters below the rootzone.” Soil Sci. Soc. Amer. J., 41, 310–315. ———. (1981). “Relation between pHc and sodium adsorption ratio (SAR) and an alternative method of estimating SAR of soil or drainage water.” Soil Sci. Soc. Am. J., 45, 469–475. Suarez, D. L., and Rhoades, J. D. (1982). “The apparent solubility of calcium carbonate in soils.” Soil Sci. Soc. Amer. J., 46, 716–722. Suarez, D. L., Wood, J. D., and Ibrahim, I. (1992). “Reevaluation of calcite supersaturation in soils.” Soil Sci. Soc. Amer. J., 56, 1176–1784. Sumner, M. E. (1993). “Sodic soils: New perspectives.” Aust. J. Soil Res., 31, 683–750. Tanji, K. K. (1969). “Solubility of gypsum in aqueous electrolytes as affected by ion association and ionic strengths up to 0.15M and at 25°C.” J. Environ. Sci. Tech., 3, 656–661. University of California–Riverside (UCR). (2003). “Laosheng Wu, WATSUIT for Windows (new–revised 2/03),” www.envisci.ucr.edu/faculty/wu.html, accessed February 20, 2011. U.S. Dept. of Agriculture, Agricultural Research Service (USDA/ARS). (1991). “WATSUIT model, year 1991, version 1.0,” http://ars.usda.gov/Services/ docs.htm?docid8968, accessed February 20, 2011.

800

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

U.S. Salinity Laboratory Staff. (1954). Diagnosis and improvement of saline and alkali soils, Agricultural Handbook 60, L. A. Richards, ed., USDA-ARS, Riverside, Calif. Available online at www.ars.usda.gov/Services/docs.htm?docid10158 &page2, accessed February 9, 2011.

NOTATION AVG. CL  average chloride concentration AVG. EC  average electrical conductivity EC  electrical conductivity GYP  gypsum LF  leaching fraction SAR  sodium adsorption ratio SUM CAT  sum of cations TDS  total dissolved solids

CHAPTER 26 LEACHING REQUIREMENT: STEADY-STATE VERSUS TRANSIENT MODELS Dennis L. Corwin, James D. Rhoades, and J. Sˇimu ˚ nek

INTRODUCTION In the southwestern United States, irrigated agriculture is responsible for roughly 80% of the demand on surface and groundwater resources. Similar demand levels can be found in irrigated arid and semiarid regions throughout the world. The sizable consumption of water to support irrigated agriculture is a growing concern, particularly in arid zone regions of the world. Greater scrutiny of irrigated agriculture’s sizable demand on water resources grows as a consequence of water scarcity due to increased demand on finite water resources and increased frequency of drought conditions resulting from erratic weather attributable to climate change or alterations in historical weather patterns. Finite water resources that are stretched to their limits must be used judiciously. One means of diminishing demand on finite water resources is to decrease the volumes of irrigation water necessary to remove salts from the rootzone to maintain crop productivity. Excess salts accumulate in the rootzone of arid and semiarid irrigated soils largely as a result of the process of evapotranspiration (ET). In the ET process, plant roots remove pure water, thus concentrating any salts present in the irrigation water, resulting in salinity profiles that typically increase with depth, as shown in Fig. 26-1. The accumulated salts can cause a reduction in crop yields and even crop failure due to (1) osmotic effects that limit plant water uptake, (2) specific-ion toxicity effects (e.g., excess Na), (3) upsetting the plant nutrient balance (e.g., Ca in the presence of excess Na), and (4) salt composition effects [e.g., high sodium adsorption ratio (SAR) and low electrical conductivity (EC)] that influence soil physical properties such as soil permeability and tilth. 801

802

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 26-1. Typical salinity profiles resulting from the process of evapotranspiration (ET) for various leaching fractions (LFs). The accumulation of excessive soluble salts in the rootzone, which threaten crop productivity on irrigated soils, can be prevented by applying water in excess of what is required to meet ET needs to leach excessive soluble salts. The water needed to remove excessive salts that cause a crop yield decrement is referred to as the leaching requirement (LR). Leaching requirement was originally defined as the fraction of infiltrated water that must pass through the rootzone to keep average rootzone soil salinity from exceeding a level that would significantly reduce crop yield, assuming steady-state conditions with associated good management and uniformity of leaching (U.S. Salinity Laboratory Staff 1954). The original concept of LR developed by the U.S. Salinity Laboratory was based on the concept of leaching fraction (LF), where LF is defined as the fraction of the applied irrigation water that moves beyond the plant rootzone and represents the level of drainage and leaching of salts. As the LF increases, the level of leaching of salts increases and the salts accumulating in the rootzone decrease, which is graphically illustrated in Fig. 26-1. The LF is quantitatively defined by Eq. 26-1: LF 

Ddw ECiw  Diw ECdw

(26-1)

where Ddw (mm3 mm2) and Diw (mm3 mm2) are the unit depths of drainage water and infiltrating irrigation water, respectively, and ECiw (dS/m) and ECdw (dS/m) are the electrical conductivities of the irrigation and drainage water, respectively. The LR represents the lowest value of

STEADY-STATE VERSUS TRANSIENT MODELS

803

LF that could be allowed without ECdw (and thus, inferentially, soil salinity) becoming excessive for optimal plant growth. Thus, the minimum value of LF (i.e., LR) would be given when the maximum permissible * ) was inserted into Eq. 26-1, resulting in salinity level of ECdw (i.e., ECdw Eq. 26-2, which is considered the original LR model: LR 

ECiw * ECdw

(26-2)

The LR is an estimate of what the LF must be to keep soil water salinity within tolerable limits for crop production. Equation 26-2 must still include a relationship between plant response and EC of the bottom of the rootzone. Equation 26-2 only considers salt tolerance of the crop grown and salinity of the irrigation water while assuming steady-state conditions. Steady-state conditions do not exist under most field situations. In addition, LR is influenced by numerous factors, including irrigation nonuniformity, mineral precipitation-dissolution reactions, transient root water uptake distributions, preferential flow, climate, runoff, extraction of shallow groundwater, and leaching from effective precipitation, as well as the questionable appropriateness of the assumption of steady-state conditions. Based on the exclusion of these factors from consideration, recent publications by Corwin et al. (2007) and Letey and Feng (2007) have brought into question the appropriateness of Eq. 26-2 as a reasonable means of calculating LR, suggesting that a new paradigm may be needed, particularly for research applications. The questionable ability of Eq. 26-2 to accurately calculate LR stems from (1) the assumption of steady state, and (2) influencing factors that are not taken into account. Steady state occurs when water content and salt concentration remain constant over time at a given soil depth. The assumption of steady state is probably not reasonable in most situations, particularly over short time periods of a few years or less, because both water content and salinity continuously change over time within the rootzone due to the extraction of water by roots and replenishment by irrigation and precipitation. In addition, several factors can cause perturbations to steady-state conditions, including a change in the crop, variation in irrigation water quality, alteration in irrigation management, and transient water uptake by plant roots. Furthermore, osmotic and matric effects on roots will cause plants to uptake water from that area of the rootzone where the least energy is expended to extract water. The dynamic uptake of water by roots enables a greater ability to tolerate average rootzone salinities higher than the plant’s salt-tolerance values, which are experimentally derived from the linearly averaged rootzone EC of the saturation

804

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

extract (ECe) of high-LF experiments producing nearly uniform rootzone salinity. By accounting for the transient uptake of water by roots, the LR will be lower than that calculated by Eq. 26-2. Aside from not considering transient conditions, other factors that influence LR are often not taken into account. Depending on the chemical composition of the irrigation water and the minerals present in the soil, salts in the soil water can precipitate or minerals in the soil can dissolve, resulting in changes in the salinity in the soil water. A low LF will increase the salt concentration in the soil water, increasing the likelihood of salt precipitation. The original LR method (i.e., Eq. 26-2) ignores the chemical process of salt precipitation which can, in some cases, significantly reduce levels of soil salinity within the rootzone. The failure to account for precipitation can lead to an overestimation of the LR, whereas the failure to account for dissolution reactions will have the opposite effect. Climatic factors such as humidity can, in some cases, increase a plant’s salt tolerance, which will lower the LR. The original LR method does not account for preferential flow, which influences water flow and the efficiency of salt leaching, resulting in an increase in LR. Runoff reduces the volume of infiltrating water, which reduces the leaching of salts raising the LR. If the plant can extract water from the groundwater, then salts accumulating in the rootzone have less of an effect, thereby lowering the LR. Leaching from effective precipitation will lower the volume of water necessary to remove salts from the rootzone, thereby lowering the LR. Furthermore, the estimation of LR does not include (1) the manner in which spatiotemporal variation in salinity within the rootzone affects crop response and water uptake, (2) scale issues, (3) horizontal leaching and subsequent redistribution of salts for cracking soils when flood irrigation is used, (4) basing the LR on the most salt-sensitive crop in a crop rotation, and (5) uncertainties in salt-tolerance data developed from experimental plots when applied to field situations. Some of these have been discussed in Rhoades (1999). It is also noteworthy that LR does not provide sufficient information concerning optimal irrigation because optimal irrigation is the amount of water that maximizes profit, and maximum profit may not coincide at all times with maximum yield (Letey et al. 1985). The relationship between crop yield and seasonal amount of water required is essential to determine the optimal irrigation management (Letey et al. 1985). For this reason, cropwater production functions have been advocated as a means of determining the economically optimal amount of water that is needed to prevent excessive accumulation of salts. Nevertheless, LR is still widely used by growers and irrigation management districts in the southwestern United States and many other irrigated arid and semiarid regions of the world. Transient models enable the simulation of complex processes with time-dependent variables. The development of transient models has been

STEADY-STATE VERSUS TRANSIENT MODELS

805

primarily facilitated by the development of high-speed computers. To evaluate the appropriateness of a steady-state approach for estimating LR, Corwin et al. (2007) compared a variety of steady-state and transient LR models to determine whether differences existed, the extent of the differences, and the reasons for the differences, and to analyze the implications of the differences with respect to irrigation management and salinity control. A compilation of the most significant results of Corwin et al. (2007) follows.

GENERAL DESCRIPTION OF MODELS USED TO ESTIMATE LEACHING REQUIREMENT Four models are compared to evaluate the appropriateness of steadystate versus transient conditions and to evaluate the significance of precipitation-dissolution reactions, transient water uptake by roots, and preferential flow to the estimation of LR. Each is considered to have potentially significant effects on LR for the fine-textured soils of the arid southwestern United States. The four models selected to compare and contrast their estimation of LR are (1) the traditional LR model, which is an LR model by Rhoades (1974) based on the original LR developed by the U.S. Salinity Laboratory Staff (1954), (2) WATSUIT (Rhoades and Merrill 1976), (3) TETrans (Corwin and Waggoner 1990a,b; Corwin et al. 1990), and (4) UNSATCHEM (Sˇimu˚nek and Suarez 1994). These models reflect a spectrum of categories of models ranging from steady-state to transient models and from functional to mechanistic, which provide potential insight into the influence of physical and chemical processes on the estimation of LR. The traditional LR and WATSUIT models are steady-state models, whereas TETrans and UNSATCHEM are transient models. The WATSUIT and UNSATCHEM models account for precipitation and dissolution reactions, but the traditional LR and TETrans models do not. The UNSATCHEM model determines ET and plant yield as a function of matric and osmotic stresses, while the traditional LR model, WATSUIT, and TETrans do not. Finally, TETrans is the only model within the group that accounts for preferential flow. Table 26-1 provides a summary of the four models, which includes the type of model (steady-state or transient) and the processes included in the model (salt effects on plant growth, osmotic and matric effects on root water uptake, precipitationdissolution reactions, and preferential flow). Steady-State Leaching Requirement Models Steady-state LR models are based on simple salt-balance concepts and an assumption of long-time average conditions that will result in steady state.

806

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 26-1. Summary of Leaching Requirement Model Type and Processes Included in Each Model Type of Model

Leaching Requirement Model (1)

Processes Included in Model

Salt Effects on Plant SteadyGrowth State Transient and EvapoModel Model transpiration (2) (3) (4)

Traditional

X

WATSUIT

X

Osmotic and Matric Effects on Plant PrecipitationWater Dissolution Preferential Uptake Reactions Flow (5) (6) (7)

X

TETrans

X

UNSATCHEM

X

X X

X

X

Traditional model The determination of LR, as originally formulated in Eq. 26-2, required * the selection of the appropriate value of ECdw for the crop in question. Such crop-related values were not known. However, data obtained from salt tolerance studies conducted in test plots utilizing relatively uniform soil conditions and optimal irrigation and crop management were available at that time (Bernstein 1974; Maas and Hoffman 1977). These studies related the response of many crops to average rootzone soil salinity in terms of the ECe (dS/m), which is approximately half that of the soil-water salinity at field capacity (U.S. Salinity Laboratory Staff 1954). The nearly uniform rootzone ECe values that resulted in 50% yield decreases in forage, field, and vegetable crops and 10% yield decreases in fruit crops were originally * substituted for ECdw in Eq. 26-2 to estimate LR. No direct evidence supports the appropriateness of this substitution or the corresponding LR values, nor is there any direct evidence to support the assumption that plants respond primarily to average rootzone soil salinity. Based on empirical distribution of soil salinity by depth, Rhoades * (1974) introduced a procedure for approximating values of ECdw for use in Eq. 26-2 using Eq. 26-3: EC*dw  5 ECe*  ECiw

(26-3)

where EC*e (dS/m) is the average EC of the saturation extract for a given crop appropriate to the tolerable degree of yield depression, usually 10% or less and equivalent to the threshold EC values as defined by Maas

STEADY-STATE VERSUS TRANSIENT MODELS

807

(1990). Substitution of Eq. 26-3 into Eq. 26-2 yields Eq. 26-4, which has become referred to as the traditional LR model: LR 

ECiw 5ECe*  ECiw

(26-4)

Equation 26-4 ties LR to irrigation water salinity and crop tolerance. The traditional LR model assumes uniform water applications and does not adjust for salt precipitation or dissolution, nor does it account for irrigation frequency effects, upward water flow, water chemical composition, and salt removal in surface runoff.

WATSUIT In contrast to the traditional steady-state model previously described, WATSUIT considers the chemical composition of the irrigation water [i.e., major cations and anions and presence or absence of soil lime (CaCO3) and gypsum (CaSO4 2H2O)] and includes the processes of mineral precipitation (salt deposition) and mineral weathering (salt pickup). The assumption is made that plant water uptake occurs from successively deeper quartile-fractions of its rootzone in the ratios of 40/30/20/10. The concentrations of the major cations and anions in the soil water within an irrigated rootzone are predicted at equilibrium by WATSUIT as a function of the irrigation water composition, quartile LF, presence or absence of soil CaCO3, and several alternative amendment treatments, such as gypsum. The WATSUIT model accounts for the precipitation and dissolution of the two most relevant soil minerals, calcite and gypsum (Rhoades and Merrill 1976). With WATSUIT, the LR is determined by accounting for the chemistry of the irrigation water and soil mineralogy to estimate the LF for which the level of average rootzone salinity equals the threshold value for the crop in question (i.e., the maximum salinity that can be tolerated without excessive loss in yield). The WATSUIT model also considers irrigation management in the determination of LR, distinguishing between conventional irrigation and high-frequency forms of irrigation. The effect of salinity on ET (mm) is not taken into account. It is assumed that there will be no loss in yield due to salinity and, concomitantly, no loss in ET, provided the average rootzone salinity does not exceed the threshold value of salinity (EC*e ; dS/m). The same assumption is also made in the TETrans model, which is discussed later. The WATSUIT model also assumes uniform water application and does not account for the effects of irrigation frequency and upward water flow from a shallow water

808

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

table. Further details regarding WATSUIT can be found in Rhoades and Merrill (1976). Transient Leaching Requirement Models Steady-state conditions are the exception rather than the rule. Perturbations to the system result in transient conditions that can reduce the general applicability of the traditional LR model approach, rendering a temporal tracking of the system with transient approaches more appropriate. TETrans The TETrans model is a functional, “tipping-bucket,” layer-equilibrium model that predicts incremental changes over time in amounts of solute and water content occurring within the crop rootzone (Corwin et al. 1990; Corwin and Waggoner 1990a,b). In TETrans, transport through the rootzone is modeled as a series of events or processes within a finite collection of discrete depth intervals. These sequential events or processes include infiltration of water, drainage to field capacity, plant water uptake resulting from transpiration, and/or evaporative losses from the soil surface. Each process is assumed to occur in sequence within a given depth interval, as opposed to reality where transport is an integration of simultaneous processes. Other assumptions include (1) the soil is composed of a finite series of discrete depth intervals with each depth interval having homogeneous properties, (2) drainage occurs through the profile to a depth-variable field capacity water content, (3) the depletion of stored water by ET within each depth increment does not go below a minimum water content that will stress the plant, (4) dispersion is either negligible or part of the phenomenon of bypass, and (5) upward or lateral water flow does not occur. Included within TETrans is a simple mechanism to account for preferential flow or bypass. The phenomenon in which all or part of the infiltrating water passes through a portion or all of the soil profile via large pores or cracks without contacting or displacing water present within finer pores or soil aggregates is referred to as bypass. This process is typical of cracking clay soils (such as those in the Imperial Valley of California). The net effect of bypass is that some resident salt is not miscibly displaced by incoming water; this reduces the leaching efficiency and increases the amount of salt retained within successive soil-depth intervals, which requires additional water to leach the salts, thereby increasing the LR. In TETrans, bypass is approximated using a simple mass-balance approach; it is simulated by ascribing a spatial variation in the fractional quantity (or % water bypass) of the resident pore-water present in the soil

STEADY-STATE VERSUS TRANSIENT MODELS

809

at the time an infiltration event occurs that is not involved in piston-type displacement following the event. The means of estimating bypass is by assuming that any deviation from piston flow for the transport of a conservative solute is due to bypass (Corwin et al. 1990). Additional details regarding TETrans can be found in Corwin et al. (1990). UNSATCHEM The UNSATCHEM model is a sophisticated mechanistic, numerical model that simulates the flow of water in unsaturated soils, along with transport and chemical reactions of solutes, and crop response to salinity (Sˇimu ˚ nek and Suarez 1994; Sˇimu ˚ nek et al. 1996). The model has submodels accounting for major ion chemistry, crop response to salinity, CO2 production and transport, time-varying concentration in irrigated rootzones, and the presence of shallow groundwater. While variably-saturated water flow is assumed to be described using the Richards’ equation, the transport of solutes and CO2 is described using the convection-dispersion equation. Root growth is described using the logistic growth function and root distribution can be made user-specific. Precipitation, ET, and irrigation fluxes can be specified at any user-defined time interval. While UNSATCHEM has not been used to determine LR, it is suited to do so by determining the minimum LF that can be used under a specified set of soil, crop, and management conditions while preventing undue losses in crop yields. The UNSATCHEM model does not account for the phenomenon of bypass. The complex transient chemical processes included are precipitation and/or dissolution of solid (mineral) phases, cation exchange, and complexation reactions as influenced by the CO2 composition of the soil air, which largely controls the soil pH, as well as sulfate ion association, which affects the solubility of gypsum. Additional details regarding UNSATCHEM can be found in Sˇ imu˚ nek and Suarez (1994) and Sˇimu˚nek et al. (1996).

MODEL INPUTS In order to estimate LR using the previously described steady-state and transient models, a database is needed for the following: climate, crops grown, crop rotations, soil physical and chemical properties related to solute transport (e.g., soil salinity initial conditions, field capacity, wilting point, bulk density, infiltration rate, texture, and hydraulic conductivity properties), irrigation management practices, drainage conditions, irrigation scheduling and amounts, ET, root water uptake, irrigation water composition, crop salt-tolerance parameters, and a schedule of events

810

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(e.g., planting and harvesting dates, dates and amounts of irrigation and rainfall, root development, mature root penetration depths, root water extraction patterns, and stages of plant growth). For comparative analysis a set of realistic conditions representative of California’s Imperial Valley was developed and used as input for the LR models. Details describing the development of the dataset from available data sources can be found in Corwin et al. (2007). To estimate the LR for the entire Imperial Valley, a primary consideration is the crop sequence grown. A single rotation was sought that would be representative of the valley-wide cropping pattern. From available records, it was found that the dominant crops grown in the Imperial Valley during the period 1989–1996 were field crops, with alfalfa as the most dominant field crop, followed by wheat. Next, the garden crops were dominant, with lettuce as the most-grown garden crop. Consequently, a representative crop rotation for the Imperial Valley is a 6-year crop rotation consisting of 4 years of alfalfa, followed by 1 year of wheat and 1 year of lettuce in sequence (i.e., alfalfa/alfalfa/alfalfa/alfalfa/wheat/lettuce). This rotation was selected as a basis for evaluating the various models for estimating LR for the Imperial Valley.

MODEL LEACHING REQUIREMENT ESTIMATES As shown in Table 26-2, the LR values determined by the traditional method from Eq. 26-4 for the individual alfalfa, wheat, and lettuce crops are 0.14, 0.04, and 0.23, respectively, assuming the EC of the irrigation water (the Colorado River) is 1.23 dS/m, and the tolerable levels of average rootzone soil salinity are 2.0, 6.0, and 1.3 dS/m, respectively. The weighted-average LR for the 6-year rotation during crop growth only and the 6-year rotation during growth and fallow periods (referred to as the overall rotation period) were 0.14 and 0.13, respectively, assuming the ETc [estimated crop evapotranspiration  ET0Kcb, where ET0 is the potential reference evapotranspiration (mm) and Kcb is the crop coefficient] values for alfalfa, wheat, and lettuce are 5,273, 668, and 233 mm, respectively (Table 26-2). The overall rotation period refers to the growth period of all the crops plus all fallow periods between crops. Additional irrigation water must be added to compensate for the amount of ETc (actually, for evaporation only) that occurs during unplanted periods and for the depletion (with reference to field capacity) of soil water that occurred during cropping. The estimated LR values from WATSUIT are 0.09, 0.03, and 0.13 for the individual alfalfa, wheat, and lettuce crops, respectively (Table 26-3). The corresponding weighted LR values for the crop growth period and overall rotation period are estimated to be 0.09 and 0.08, respectively (Table 26-3).

STEADY-STATE VERSUS TRANSIENT MODELS

811

TABLE 26-2. Leaching Requirements (LR) as Determined by the Traditional Method Leaching Requirement Estimates (Traditional Method)

Period (1)

ETca (mm) (2)

Alfalfa (Year 1) Alfalfa (Year 2) Alfalfa (Year 3) Alfalfa (Year 4) Wheat Lettuce Crop growth

1,642 1,740 1,740 1,511 668 233 7,534

Overall

7,731

LRb (3)

0.14 0.14 0.14 0.14 0.04 0.23

Diwc (mm) (4)

Ddwc (mm) (5)

Weighted LR (6)

1,909 2,023 2,023 1,757 699 304 8,715

267 283 283 246 31 71 1,181

0.14e

8,912

1,181

0.13f

a

Crop evapotranspiration (ETc) from Table A-2 in UCCE (1996). Leaching requirement (LR) calculated from LR  ECiw/(5ECe*  ECiw). c Required irrigation, Diw ETc/(1  LR). d Required drainage, Ddw Diw  ETc. e (Required drainage/Required irrigation) during crop growth period. f (Required drainage/Required irrigation) during overall rotation period. b

Crop growth  6-year rotation during crop growth period. Overall  6-year rotation during crop growth and fallow periods. Diw  unit depth of irrigation water (mm3 mm2) Ddw  unit depth of drainage water (mm3 mm2) From Corwin et al. (2007) with permission from Elsevier.

Figure 26-2a,b shows quite clearly from WATSUIT simulations that salt precipitation under steady-state conditions is a significant factor in reducing average soil salinity for this water composition. At steady state, the soil water salinity will be predicted with WATSUIT to be reduced by salt precipitation by about 25% at an LF of 0.03, 20% at LF 0.05, 13% at LF 0.10, 9% at LF 0.15, and 5% at LF 0.20 (Fig. 26-2b). These depositions of salt reduce the need for leaching. As shown in Tables 26-2 and 26-3 and Fig. 26-2b, the LR for alfalfa (ECe* of 2.0 dS/m) is reduced from 0.14 to 0.09 by salt precipitation; the LR for lettuce (EC*e of 1.3 dS/m) is reduced from 0.23 to 0.13, and the LR for wheat is reduced from 0.04 to about 0.03. The process of salt precipitation, in which the salts are made innocuous to plants and removed from the soil and drainage waters, significantly reduces the LR. To illustrate, the LR value for the crop rotation period obtained using the WATSUIT model is estimated to be about 0.08 to 0.09,

812

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 26-3. Leaching Requirements as Determined by the WATSUIT Model Leaching Requirement Estimates (WATSUIT Model)

Period (1)

ETca (mm) (2)

Alfalfa (Year 1) Alfalfa (Year 2) Alfalfa (Year 3) Alfalfa (Year 4) Wheat Lettuce Crop growth

1,642 1,740 1,740 1,511 668 233 7,534

Overall

7,731

LRb (3)

0.09 0.09 0.09 0.09 0.03 0.13

Diwc (mm) (4)

Ddwd (mm) (5)

1,804 1,912 1,912 1,660 685 266 8,239

162 172 172 149 17 34 706

0.09e

8,436

706

0.08f

Weighted LR (6)

a

Crop evapotranspiration (ETc) from Table A-2 in UCCE (1996). Leaching requirement (LR) obtained from Fig. 26-1b. c Required irrigation, Diw ETc/(1  LR). d Required drainage, Ddw Diw  ETc. e (Required drainage/Required irrigation) during crop growth period. f (Required drainage/Required irrigation) during overall rotation period. b

Crop growth  6-year rotation during crop growth period. Overall  6-year rotation during crop growth and fallow periods. Diw  unit depth of irrigation water (mm3 mm2) Ddw  unit depth of drainage water (mm3 mm2) From Corwin et al. (2007) with permission from Elsevier.

compared to LR values of about 0.13 to 0.14 obtained using the traditional LR model. The TETrans model was used to test whether the steady-state LRs determined from the traditional method would result in lower, comparable, or higher levels of soil salinity under transient conditions; consequently, irrigation timings and amounts for TETrans were adjusted to match those of the steady-state LRs determined using the traditional method. Preseason irrigations were given only in amounts sufficient to return the soil to field-capacity water content; no special irrigations, such as reclamation leaching, were included in the simulations. The cumulative LFs that actually were obtained in the simulations were 0.14, 0.04, and 0.17 for alfalfa, wheat, and lettuce, respectively, and an overall rotation LR of less than 0.13. These results and their time trends are shown in Fig. 26-3. The simulations reveal that, when bypass is 40% or less, soil

STEADY-STATE VERSUS TRANSIENT MODELS

813

FIGURE 26-2. WATSUIT-simulated results when irrigated with Colorado River water. (A) average soil salinity (0–120 cm) with and without salt removal by precipitation as related to LF, and (B) percent reduction in salt concentration in soil water due to salt precipitation as a function of LF. From Corwin et al. (2007) with permission from Elsevier.

salinity is less than the threshold ECe levels of each crop grown in the rotation, even though the LFs were based on the steady-state traditional LR model. At most, the yield of alfalfa would be reduced by 1.5% during the first season. Even under the extreme conditions of 80% bypass, alfalfa yield would be reduced by only 3% during the first year of production; no loss would occur in the next 3 years of production. Wheat yield would not

814

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 26-3. Time trend in average salinity (ECe, dS/m) for the soil profile (0–120 cm) over the period of a 10-year cycle of crop rotation (A–C), and for the crop rootzone (alfalfa: 0–120 cm, wheat: 0–90 cm, lettuce: 0–60 cm) over the period of a 6-year cycle of crop rotation (D–F) as predicted by TETrans for various levels of bypass: (A and D) 0% bypass, (B and E) 40% bypass, and (C and F) 80% bypass. From Corwin et al. (2007) with permission from Elsevier.

be reduced under such extreme conditions of bypass; lettuce yield would be reduced by no more than 5%. The results show that the LRs estimated from the steady-state traditional model are not too low, but they are probably too high. The results presented in Figs. 26-3d–f and 26-4 show that the relatively high levels of salinity that develop over time in the lower portion of the rootzone are subsequently displaced to deeper depths and eventually out of the rootzone as the subsequent crop is irrigated. The effect of bypass is also illustrated in these figures. The levels and distributions of soil salinity are not much affected by bypass up to at least 40%. This level of bypass slightly increases salinity levels in the relatively shallow soil profile depths in the early period of the crop season, but not enough to reduce yield. The predicted salinity levels when the bypass is very high (⬃80%) are higher, especially during the periods of wheat and lettuce production (see Fig. 26-2f). These levels are not high enough to reduce wheat yield,

STEADY-STATE VERSUS TRANSIENT MODELS

815

FIGURE 26-4. Soil salinity levels (ECe, dS/m) by depth at selected times for alfalfa (A–C), wheat (D–F), and lettuce (G–I) as predicted by TETrans for various levels of bypass: (A, D, and G) 0% bypass, (B, E, and H) 40% bypass, and (C, F, and I) 80% bypass. From Corwin et al. (2007) with permission from Elsevier.

but they could slightly reduce lettuce growth during the early part of its growing season. While the extent of bypass occurring in the Imperial Valley soils has not been established, it is doubtful that it reaches the level of 80%. Thus, it is doubtful that crop yields would be reduced by the levels of soil salinity resulting under the conditions of simulated crop rotation, even considering the bypass phenomenon. Simulations using TETrans show that the LRs of the crops in rotation are not greater than those estimated using the traditional model. This is because the estimate of LR by the traditional model is slightly more conservative than by TETrans, that is, the maximum levels of salinity predicted to occur at steady-state do not result under transient conditions. Because TETrans does not account for salt precipitation, predictions of salinity distributions in the rootzone are still higher than would be expected.

816

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The amounts of irrigation, precipitation, crop ET (i.e., ETc), and the levels of resulting leaching and deep percolation predicted from UNSATCHEM for each crop and for the entire rotation are summarized in Table 26-4. The LRs estimated from UNSATCHEM are 0.10 for alfalfa, 0 for wheat, 0.13 for lettuce, and an overall rotation LR of less than 0.08. The estimates of LR obtained with the steady-state WATSUIT model (i.e., 0.09, 0.03, and 0.13 for alfalfa, wheat, and lettuce, respectively, and an overall rotation LR of 0.08) appear to provide estimates of LR for salinity control as reasonable as those of the transient model UNSATCHEM. The LR values of 0.09 for alfalfa and of 0.13 for lettuce appear to be close to the minimum. The LR value of 0.03 for wheat is about as low as feasible, though the salinity level as determined by UNSATCHEM is still much below tolerable by this crop. It may be concluded that the LR may be as low (or possibly lower) as 0.08 for the overall crop rotation and about 0.10 for alfalfa, 0 for wheat, and 0.13 for lettuce. The manner in which the distribution of salinity within the soil profile (0–120 cm) changes during the crop rotation is shown in Figs. 26-5 and 26-6. The relatively low levels of salinity maintained within the rootzones of these crops during most of their cropping seasons, especially in the upper half of the rootzones, illustrates the adequacy of the simulated irrigation/leaching management for salinity control.

TABLE 26-4. Estimates of Deep Percolation and Leaching Fraction (LF) Obtained with the UNSATCHEM Model Time Period (Day Numbers) (1)

Crop (2)

No. of Days (3)

Adjusted ETc (cm) (4)

Precipitation (cm) (5)

3491814

alf

1,465

672.5

27.2

721.0

1814–2038

wh

224

72.8

3.7

55.1

0 14.0

0

2038–2170

let

132

29.8

1.2

33.2

0

0.14

349–2170

rot

1,821

775.1

32.1

809.4

0

Irrigation SW (cm) (cm) (6) (7)

0

DP (cm) (8)

LF (9)

75.7

0.10

4.7

66.4 0.08

alf  alfalfa wh  wheat let  lettuce rot  alfalfa/alfalfa/alfalfa/alfalfa/wheat/lettuce rotation bare  fallow ETc  crop evapotranspiration SW  change in soil water content DP  deep percolation  precipitation  irrigation  ETc  change in soil water content From Corwin et al. (2007) with permission from Elsevier.

STEADY-STATE VERSUS TRANSIENT MODELS

817

FIGURE 26-5. Soil salinity levels (ECe, dS/m) by depth at selected times as predicted by UNSATCHEM for (A) alfalfa, (B) wheat, and (C) lettuce. From Corwin et al. (2007) with permission from Elsevier.

FIGURE 26-6. Soil salinity levels (ECe, dS/m) by depth for alfalfa, wheat, and lettuce as predicted by UNSATCHEM for (A) single selected days late in each crop season (i.e., alfalfa: Day 1752, wheat: Day 1951, and lettuce: Day 2170); and (B) an average over the entire crop season (i.e., 4 years for alfalfa, 1 year for wheat, and 1 year for lettuce). From Corwin et al. (2007) with permission from Elsevier.

818

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

IMPLICATIONS OF LEACHING REQUIREMENT MODEL ESTIMATES A summary of the values of LR obtained by the various methods is given in Table 26-5. Comparing steady-state models to transient models supports the notion that steady-state models overestimate LR, but only to a minor extent. Estimates of LR by steady-state models were found to be slightly conservative. The steady-state traditional model and transient TETrans model are directly comparable because they are based on the same water-salt balance relations and exclude the effects of salt precipitation. Similarly, the steady-state model WATSUIT is directly comparable to the transient model UNSATCHEM since both take mineral precipitationdissolution reactions into account. In both comparisons, there is only a slight difference in estimated LRs (see Table 26-5). The actual levels of rootzone salinity will be slightly less than the predicted steady-state levels for the cases of annual crops and time-varying cropping since there is insufficient time to develop the maximum levels found under steadystate conditions, which result only after longer periods of continuous cropping, such as with perennial crops. The estimates of LR were significantly reduced when the effect of salt precipitation was included in the salt-balance calculations, regardless of whether the model was steady state or transient. For example, the LR for TABLE 26-5. Summary Table of Leaching Requirements as Estimated by Various Methods Leaching Requirement (LR) Crop or Cropping Period Model (1)

Traditional WATSUIT TETrans

UNSATCHEM

Tablea (2)

26-2 26-3 Corwin et al. (2007) 26-4

Lettuce (5)

Crop Growthb (6)

0.14 0.04 0.09 0.03 0.14 0.04

0.23 0.13 0.17

0.14 0.09 0.13

0.1

0.13

0.08

Alfalfa (3)

Wheat (4)

0

a

Table number in this chapter where data were obtained. Crop growth refers to period included in crop simulation. c Overall rotation includes entire rotation with fallow periods. b

From Corwin et al. (2007) with permission from Elsevier.

Overall Rotationc (7)

0.13 0.08

STEADY-STATE VERSUS TRANSIENT MODELS

819

the overall crop rotation was reduced from 0.13 for the traditional model to 0.08 for the WATSUIT method by accounting for salt precipitation (Table 26-5). Simulated data from WATSUIT show that the concentration of soil-water salinity is reduced by about 13% and 25% at LFs of 0.1 and 0.04, respectively, as a result of salt precipitation (Fig. 26-2b). The average soil-salinity levels predicted with the transient UNSATCHEM model were essentially the same as those obtained with the steady-state WATSUIT model (Table 26-5). Both models clearly show that with salt precipitation lower LR would be expected. The predicted levels of salinity simulated by UNSATCHEM within the rootzones of alfalfa, wheat, and lettuce never exceeded levels that would cause crop-yield losses at any time during the transient conditions of crop rotation. These and other results obtained with UNSATCHEM indicate that (1) reclamation and the use of less water than that estimated by the traditional LR method could control soil salinity in the alfalfa/ wheat/lettuce crop rotation selected as representative of Imperial Valley conditions, and (2) the LR is lower than that determined using the traditional method. The two transient models, TETrans and UNSATCHEM, estimated the LR to be lower than the traditional steady-state approach. The weakness of the traditional LR approach is that steady-state conditions seldom exist except over long time periods, and processes, such as preferential flow and precipitation-dissolution reactions, are not taken into account. The difference between the traditional steady-state and transient approaches is expected and adds credence to the recommendation that any estimation of LR first consider the use of a transient model, particularly for research applications. The same general conclusion recommending the use of a transient over a steady-state approach for estimating LR was also found by Letey and Feng (2007) when focusing on the influence of plant water uptake using the transient ENVIRO-GRO model compared to two steadystate models. The small difference in the estimated LR between WATSUIT and UNSATCHEM shows that accounting for salt precipitation under conditions representative of the Imperial Valley was more important than whether the model was a steady-state or transient model. This suggests that in some instances accounting for all the dominant mechanisms influencing the leaching of salts may be nearly as important as capturing the temporal dynamics of the leaching process. This fact suggests that there may be certain instances where steady-state models can be used as long as the models account for all the dominant mechanisms (e.g., bypass flow, mineral precipitation-dissolution reactions, plant water uptake) that are affecting the leaching of salts and that few or no perturbations that have occurred over a long time period would prevent steady-state conditions, or nearly so. For instance, in situations where precipitation-dissolution reactions

820

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

are dominant and temporal dynamic effects are minimal, LR could be adequately estimated using WATSUIT. Or, in situations where the irrigation water of quality and amount minimized the temporal dynamic effects of plant water uptake, LR could be adequately estimated using the exponential-water-uptake, steady-state model by Hoffman and van Genuchten (1983). Using the area of every crop and an estimate of the LR for each crop with the traditional model to obtain a valley-wide LR based on the weighted average of the crop areas and LRs, Jensen and Walter (1998) obtained an LR value of 0.14 for the entire Imperial Valley. In addition, field studies by Oster et al. (1986) showed a similar steady-state estimate of LR of 0.12. The LR value obtained from Corwin et al. (2007), as discussed herein for the representative Imperial Valley crop rotation using the traditional method of estimating LR was 0.13. The three results are essentially the same. However, the valley-wide LR is more accurately estimated using the selected representative crop rotation and either the WATSUIT or UNSATCHEM model. Based on the results obtained with these models, an LR value of 0.08 is concluded to be reasonable for the entire Imperial Valley. This conclusion is based on the fact that both models predict that soil salinity will not accumulate to levels that would cause losses to any crop grown in rotation at the ascribed level of leaching. Furthermore, the 6-year crop rotation is made up of the dominant crops grown in the Imperial Valley and of crops that are dominantly salt-sensitive (alfalfa and lettuce). The LR would be proportionately lower if the assessment was based on more salt-tolerant crops. The validity of a valley-wide LR of 0.08 is supported by the results of a field experiment carried out in the Imperial Valley in which a succession of crops were successfully grown in two different rotations (cotton/wheat/alfalfa and wheat/sugar beets/ cantaloupes) with an LF of about 0.1, even while substituting water that was four times as saline as Colorado River water (i.e., Alamo River water) in place of Colorado River for 30% to 50% of the total irrigation supply (Rhoades et al. 1989). The field studies by Bali and Grismer (2001) and Grismer and Bali (2001) also support the notion that a valley-wide LR for the Imperial Valley of 0.08 is reasonable from results that showed no decrease in the yield of alfalfa and Sudan grass hay at an LF of 0.10 or less. The salient points to be derived from the LR model simulations that are specific to the conditions representative of the Imperial Valley include: (1) for cracking soils representative of the Imperial Valley, preferential flow does not appear to be a significant factor influencing LR; and (2) salt precipitation is a primary factor for reducing LR for the Imperial Valley. The implication is that reducing the estimated LR from 0.13 to 0.08 will reduce irrigation water needs that deplete scarce surface-water supplies and will reduce drainage volumes that affect the environment when disposed. Each year an estimated 2.46  109 m3 (2 million ac-ft) of water infil-

STEADY-STATE VERSUS TRANSIENT MODELS

821

trates into the cropped soil of Imperial Valley; consequently, reducing the LR from 0.14 to 0.08 would reduce the drainage volume by approximately 1.23  108 m3 (100,000 ac-ft). However, cautionary notes should be weighed when considering the practicality and validity of a valley-wide LR of 0.08 for the Imperial Valley. First, the effect of irrigation uniformity has not been addressed in this study, nor has runoff been considered. The lack of irrigation uniformity caused by uneven application of irrigation water and/or within-field spatial variability results in a greater application of irrigation water to attain maximum yield. The issues of nonuniformity effects on LR are discussed in detail by Rhoades (1999). The inability to more precisely control the spatial distribution of irrigation application also causes runoff. However, site-specific irrigation technology may eventually overcome the problems of application distribution associated with flood irrigation and withinfield spatial variability through site-specific sprinkler irrigation. The use of level basins may also ameliorate, to some extent, the nonuniformity of infiltration seen with flood irrigation. A second cautionary note pertains to the small effect of bypass on EC values, especially at deeper depths, suggesting that bypass will not significantly influence LR estimates. This is the consequence of the observation in soil lysimeters containing Imperial Valley silty-clay soil that bypass primarily occurred from the soil surface to 30 to 45 cm below the surface, which may not be a realistic assumption in the field. The effect of bypass is small because it occurs only in the top 30 cm of the soil profile, where concentrations are relatively small and downward fluxes are large. Had the bypass been active in deeper layers, where concentrations are large and fluxes small, the effect would be significantly larger. Third, the ability to control a 0.05 reduction in LF will require a change from current irrigation management that results in significant runoff, and will not be realized until more efficient site-specific irrigation management is adopted. An inherent limitation in the LR model comparison by Corwin et al. (2007), using representative data, is that it is an indicator but not a confirmation that transient models results are better. Confirmation that transient models provide a more robust estimation of LR than steady-state models can only be shown through more controlled experimental conditions. Our lower estimates of LR by transient models suggest the need for a reevaluation of the traditional means of estimating LR, but caution must be taken in considering the transient model approach as the new paradigm until experimental data can provide direct evidence of its enhanced accuracy for determining LR. Many issues still remain that confound our knowledge of applying models, such as issues related to temporal and spatial scales, the complexities of uniformity of irrigation water application, and spatial variability, just to mention a few. However, this cautionary note should not preclude the use of transient models in place of

822

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

steady-state models as tools to help develop irrigation management guidelines and recommendations, as long as the transient models are not misused, which is an essential caveat.

SUMMARY Calculations of the LR and LF can be made using a number of either steady-state or transient models, each of which describes one or more of the following processes: • • • •

Salt effects on plant growth and ET Osmotic and matric effects on plant water uptake Precipitation-dissolution reactions Preferential flow.

Each of the four models described produces a different LR. The original or traditional model describes the LR in terms of the relative EC of water at irrigation infiltration depth and drainage-water depth, and generally provides the highest LR value. WATSUIT accounts for irrigation water quality and effects on two important soil chemicals (calcite and gypsum) as irrigation water flows through the soil. This provides additional insight into the conditions in the soil being evaluated. Both the traditional model and the WATSUIT model are steady-state models, that is, they do not account for incremental changes in soil conditions over time. The two transient models (TETrans and UNSATCHEM) use different methods to capture incremental changes in soil processes and conditions over time. LR simulations using the four models (Table 26-5) vary by as much as ⬃40%. The implications of these simulations are that (1) estimates of LR are subject to substantial variation, depending on the method, (2) the more recent transient models can capture and evaluate more of the variables that may affect LR, and (3) the traditional models may overestimate the LR. Given the complexities of irrigation and drainage, and the economic and ecological consequences of excessive drainage, it is probably appropriate to develop more accurate tools for estimating LR.

REFERENCES Bali, K. M., and Grismer, M. E. (2001). “Reduced-runoff irrigation of alfalfa in Imperial Valley, California.” J. Irrig. Drain. Eng., 127(3), 123–130. Bernstein, L. (1974). “Crop growth and salinity,” in Drainage for agriculture, J. van Schilfgaarde, ed., Agronomy Monograph No. 17, SSSA, Madison, Wisc., 39–54.

STEADY-STATE VERSUS TRANSIENT MODELS

823

Corwin, D. L., and Waggoner, B. L. (1990a). TETrans: Solute transport modeling software user’s guide (IBM Compatible Version 1.5), U. S. Salinity Laboratory Report 123, USDA-ARS, U.S. Salinity Laboratory, Riverside, Calif. ———. (1990b). TETrans: Solute transport modeling software user’s guide (Macintosh Version 1.6). U.S. Salinity Laboratory Report 121, USDA-ARS, U.S. Salinity Laboratory, Riverside, Calif. Corwin, D. L., Rhoades, J. D., and Sˇimu˚nek, J. (2007). “Leaching requirement for soil salinity control: Steady-state versus transient models.” Agric. Water Mgmt., 90, 165–180. Corwin, D. L., Waggoner, B. L., and Rhoades, J. D. (1990). “A functional model of solute transport that accounts for bypass.” J. Environ. Qual., 20, 647–658. Grismer, M. E., and Bali, K. M. (2001). “Reduced-runoff irrigation of Sudan grass hay in Imperial Valley, California.” J. Irrig. Drain. Eng., 127(5), 319–323. Hoffman, G. J., and van Genuchten, M. T. (1983). “Soil properties and efficient water use: Water management for salinity control,” in Limitations to efficient water use in crop production, H. Taylor, W. Jordan, and T. Sinclair, eds., ASA/CSSA/SSSA, Madison, Wisc., 73–85. Jensen, M. E., and Walter, I. A. (1998). Review of the report: Imperial Irrigation District water use assessment for the years 1987–1996, March 1998, Report to Bureau of Reclamation, November 14, 1998, U.S. Dept. of the Interior, Bureau of Reclamation, Washington, D.C. Letey, J., and Feng, G. L. (2007). “Dynamic versus steady-state approaches to evaluate irrigation management of saline waters.” Agric. Water Mgmt., 91, 1–10. Letey, J., Dinar, A., and Knapp, K. C. (1985). “Crop-water production function model for saline irrigation waters.” Soil Sci. Soc. Amer. J., 49, 1005–1009. Maas, E. V. (1990). “Crop salt tolerance,” in Agricultural salinity assessment and management, K. K. Tanji ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 262–304. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. Drainage Eng. ASCE, 103(IR2), 115–134. Oster, J. D., Meyer, J. L., Hermsmeier, L., and Kaddah, M. (1986). “Field studies of irrigation efficiency in the Imperial Valley.” Hilgardia, 54(7), 1–15. Rhoades, J. D. (1974). “Drainage for salinity control,” in Drainage for agriculture, J. van Schilfgaarde, ed., Agronomy Monograph No. 17, SSSA, Madison, Wisc., 433–461. ———. (1999). Assessment of Imperial Valley leaching requirement: 1990-1996, A special report for the U. S. Bureau of Reclamation, U. S. Bureau of Reclamation, Boulder City, Nev. (published Oct. 20, 1999 and revised Dec. 3, 2002). Rhoades, J. D., Bingham, F. T., Hoffman, G. J., Dedrick, A. R., Pinter, A. R., and Replogle, J. A. (1989). “Use of saline drainage water for irrigation: Imperial Valley study.” Agric. Water Mgmt., 16, 25–36. Rhoades. J. D., and Merrill, S. D. (1976). “Assessing the suitability of water for irrigation: Theoretical and empirical approaches,” in Prognosis of salinity and alkalinity. FAO Soils Bulletin 31, Food and Agriculture Organisation of the United Nations, Rome, 69–109. Sˇimu˚nek , J., and Suarez, D. L. (1994). “Major ion chemistry model for variably saturated porous media.” Water Resour. Res., 30(4), 1115–1133.

824

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Sˇimu˚nek , J., Suarez, D. L., and Sejna, M. (1996). The UNSATCHEM software package for simulating one-dimensional variably saturated water flow, heat transport, carbon dioxide production and transport, and solute transport with major ion equilibrium and kinetic chemistry, Version 2, Research Report No. 141, USDA-ARS, U.S. Salinity Laboratory, Riverside, Calif. University of California Cooperative Extension (UCCE). (1996). Guidelines to production costs and practices for Imperial County, field crops 1996–1997, K. S. Mayberry and G. J. Holmes, eds., Circular 104-F, UC Cooperative Extension, Holtville, Calif., 39. U.S. Salinity Laboratory Staff. (1954). Diagnosis and improvement of saline and alkali soils, Agricultural Handbook 60, USDA-ARS, Riverside, Calif.

NOTATION Ddw  unit depth of drainage water (mm3 mm2) Diw  unit depth of infiltrating water (mm3 mm2) ECdw  electrical conductivity of the drainage water (dS m1) ECe  electrical conductivity of the saturation extract (dS m1) ECiw  electrical conductivity of the irrigation water (dS m1) 1 *  maximum permissible salinity level of EC ECdw dw (dS m ) * ECe  average electrical conductivity of the saturation extract (dS m1) for a given crop appropriate to the tolerable degree of yield depression, usually 10% or less and equivalent to the threshold electrical conductivity values defined by Maas (1990) ET  evapotranspiration (mm) ETc  estimated crop evapotranspiration (mm)  ET0Kcb where ET0 is the potential reference evapotranspiration (mm) and Kcb is the crop coefficient LF  leaching fraction LR  leaching requirement

CHAPTER 27 CONCEPTUAL WATER FLOW AND SALT TRANSPORT FOR FLUX-LIMITED AND PONDED INFILTRATION* W. W. Wallender, K. K. Tanji, B. Clark, R. W. Hill, E. C. Stegman, J. R. Gilley, J. M. Lord, and R. R. Robinson

INTRODUCTION Water and salt flow are particularly complex for drip-irrigated soils, as well as for surface-irrigated, cracking clayey soils, and thus predicting the effect of water and land management scenarios is also complex. Sophisticated analytical and numerical models are leading the way to understand these systems and connections. As one might expect, the data requirements in terms of the number of input variables and their initial conditions and boundary conditions—not to mention their spatial variation—may become overwhelming, particularly for irrigation system designers and managers. There is an urgent need to develop a management-type model to calculate the disposition of water and salt for various water and cropping system scenarios. Before describing the *All text, figures and tables in the chapter are taken either directly or with modification from the following two articles with kind permission from Springer ScienceBusiness Media. Wallender, W. W., Tanji, K. K., Gilley, J. G., Hill, R. W., Lord, J. M., Moore, C. V. Robinson, R. R., and Stegman, E. C. (2006). “Water flow and salt transport in cracking clay soils of the Imperial Valley, California.” Irrig. and Drainage Sys., 20, 361–387. Wallender, W. W., Tanji, K. K., Clark, B., Hill, R. W., Stegman, E. C., Gilley, J. R., Lord, J. M., and Robinson, R. R. (2007). “Drip irrigation water and salt flow model for table grapes in Coachella Valley, California.” Irrig. and Drainage Sys., 21, 79–95. 825

826

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

conceptual model proposed herein, a brief review of research-type models is fitting.

AN OVERVIEW OF RESEARCH A number of efforts to model the wetting pattern of drip-irrigation systems have been advanced. One line of modeling is based on solving Richards’ equation for water flow in two and three dimensions (e.g., Warrick 1986; Sˇ imu ˚ nek et al. 1999; Annandale et al. 2003; Cote et al. 2003; Skaggs et al. 2004; Gardenas et al. 2005). Richards’ equation requires a robust input data typically obtained for research plots. A second line of modeling effort is a semi-empirical approach for determining the geometry of wetting patterns for line surface and subsurface drip systems (e.g., Schwartzman and Zur 1986; Cook et al. 2003; Singh et al. 2006; Warrick and Orr 2007). This approach is less data-intensive, assuming an elliptical geometry for wetting. A few drip irrigation models consider solute transport utilizing advective dispersion equation for salts (e.g., Sˇ imu ˚ nek et al. 1999) and nitrate (e.g., Gardenas et al. 2005). In the case of cracking clay soils, a particularly difficult modeling challenge is simulating preferential flow of water and solutes. Numerous research articles in the scientific literature document measured preferential flow (Singh and Kanwar 1991; Li and Ghodrati 1997; Andreini and Steenhuis 1990; Casey et al. 1997; Clothier et al. 1992; Anderson and Bouma 1977a,b; Blake et al. 1973; Seyfried and Rao 1987; Ritchie et al. 1972; Kissel et al. 1973). Field experiments showed that water flows quickly through macropores and arrives below the rootzone. Laboratory breakthrough curves (solute concentration of water flowing out of a soil column with time or with the number of pore volumes) indicated that tracers, such as the chloride ion move rapidly through large connected pores. Bouma and Dekker (1978) documented short-circuiting (preferential flow) and measured infiltration into dry clay soil. Water containing methylene blue tracer short-circuited the upper soil layers and moved rapidly through initially large air-filled vertical pores and bypassed dry or moist soil inside peds. This was evidenced by small vertical bands of stain to a depth of 1 m. Grass roots were concentrated along the vertical prism faces. Ponded infiltration rates measured in laboratory studies on undisturbed samples ranged from 11 to 75 m/day. Anderson and Bouma (1977 a,b) demonstrated the use of surface crusts and reduced sprinkling rates to eliminate short-circuiting. Under Bouma’s leadership, a quasi three-dimensional model to simulate infiltration was developed. Vertical infiltration of water into the top surface of the soils was simulated using a one-dimensional vertical infiltration model (van Keulen and van Beek 1971) and horizontal absorption

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

827

of water into the side walls of the peds was simulated using a model developed by van der Ploeg and Benecke (1974). Flow into the cracks from the surface began when a preset ponding depth (0.2 cm) was reached. The volume of water infiltrating horizontally from the cracks into the vertical surface of the ped depended on the vertical area wetted and the absorption rate. In contrast to Imperial Valley (California) soils in which the vertical area was wetted nearly simultaneously as the cracks fill during a surface irrigation event, these soils were sprinkled and the contact area grew with time as the number of bands increased and enlarged. Although this early attempt simulated quasi three-dimensional water flow, it did not attempt to calculate the movement of solutes. Van Genuchten and Dalton (1986) reviewed more complex “tworegion-type” (Coats and Smith 1964) models for simulating flow as well as salt movement in aggregated soils. Solutes move by convection and dispersion through connected macropores and by diffusion inside micropores. A first-order rate model is used for solute exchange between the mobile (macropore) and immobile (micropore) liquid zones. More recent simulation models of water flow and salt transport (Wallach and Steinhuis 1998; Casey et al. 1997; Ma et al. 1995) were also based on this mobile and immobile model (MIM). Improvements include multiple soil layers or several regions for flow, but these models rely on the data-intensive characterization of soil hydraulic and chemical properties that is beyond the scope of most management modeling approaches. There is an additional need to develop conceptual models requiring input data normally taken or utilized to manage soil water and salinity in drip-irrigated soils (such as ETo, Kc, water application rates and duration, irrigation water quality, soil water contents, and soil salinity), as well as for irrigated cracking clay soils (crack volume, etc.). We further suggest that such conceptual simulator models could be fine-tuned with ongoing field-monitored data (such as soil water content, soil salinity, real-time irrigation scheduling, etc.). We shall refer to such modeling efforts as adaptive simulation. The model should be user-friendly and coded on Microsoft Excel spreadsheets so that the manager can easily modify the program as needed, as well as evaluate water and salinity management on a laptop computer in the field, and make appropriate changes to management. This chapter is the compilation as well as additional synthesis of two papers published by the authors in the journal Irrigation and Drainage Systems (Wallender et al. 2006; Wallender et al. 2007).

WATER FLOW MODEL The conceptual flow and transport model described in this chapter is a mass balance, perfect mixing, chemical equilibrium model. The soil

828

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

simulation zone is spatially discretized into imaginary containers. For a given container, water and salt flow in, they mix perfectly with the existing soil water, and the mixture reacts chemically. Some solution is retained while excess flows out horizontally or vertically, depending on soil physical properties. This process cascades through the soil domain to give the water and salt amount and concentration for each infiltration event. Cascading is spatially sequential and predetermined and is not time-dependent. Because the sequencing is predetermined and explicit, the simulation is programmed in a spreadsheet using relative references. Two example applications are summarized in what follows in order to illustrate the adaptability and utility of the conceptual model for water and salt management. The first is for drip irrigation of grapes in the Coachella Valley of California (Wallender et al. 2007) and the second is for surface irrigation of cracking clay soils in the Imperial Valley of California (Wallender et al. 2006). Drip Irrigation The water flow model for drip irrigation is shown in Fig. 27-1. For a detailed description of the model, refer to Wallender et al. (2007). 1. Soil, plant, irrigation, and rainfall data are input to the model. These include surface geometry (Fig. 27-2) (including wedge angle , row angle , and radius of the wedge r) as well as profile geometry (including soil depth and radius intervals) (Fig. 27-3). 2. The predetermined flow route is from the cell (container) with the emitter to its contiguous horizontal and vertical cells according to a routing factor , which is the ratio of horizontal to vertical outflow volume. This cascades through the simulation domain. Volumetric soil water content at field capacity determines the maximum water retained before outflow, and a minimum water content for water extraction sets the lower limit for water removal by the crop. 3. The evapotranspiration (ET) partition coefficient is the fraction of total water extraction between irrigation events and it varies with depth, radius, and cell volume. The ET between irrigations is required, as are drip irrigation volume and rain and or sprinkler irrigation depth. At the beginning of the simulation, the soil water content is initialized. Water is extracted from the soil according to the ET partition coefficient prior to irrigation or rain. Volume of extraction from each cell is the volume of ET for the interval multiplied by the partition coefficient. 4. Volume of drip-irrigation water for the irrigation interval is applied to the top cell containing the emitter; the volume of sprinkler irrigation water and/or rainfall is added to the top cell or cells. Steps 6 and 7 in Fig. 27-1 may be skipped in the case of rainfall, hydrocooling, or recla-

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

829

FIGURE 27-1. Schematic of drip-irrigation water flow model.

mation leaching using sprinkler irrigation, for example, but otherwise the process repeats until simulation is complete. 5. Simulation output includes water content and salinity distribution within the wedge as well as flow and salinity out the bottom layer of the wedge. Cracking Soil Water flow routing for surface irrigation of cracking clay soils involves initial crack filling and bypass flow directly to the groundwater, and,

830

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-2. Horizontal wetting pattern from emitters along laterals for two rows. thereafter, horizontal flow from cracks to the soil mass, downward flow in the soil mass during internal drainage, water extraction by the crop, and finally upward flow from groundwater if soil water is insufficient to meet ET demand (Fig. 27-4). In contrast to the drip-irrigation application in which there is horizontal as well as vertical flow, the simulation domain assumed in cracking soil is only in the vertical direction using a single stack of containers. Bypass flow is routed past the soil containers directly to the groundwater container. During reclamation leaching, after the cracks fill, additional water is applied at the surface to enhance flow through the soil mass and thus leaching. Internal drainage continues and drainage flow out of the groundwater is produced. The schematic of water flow is presented in Fig. 27-5 and the steps are summarized as follows. For a more detailed description of the model, refer to Wallender et al. (2006). 1. Soil, plant, and groundwater data are input to the model. Input variables include soil depth intervals (seven layers of 0.3-m thickness), water content immediately after irrigation (saturation), water content

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

831

FIGURE 27-3. Soil wedge control volume with lines representing the edges of the outside bounding surfaces of each cell.

2.

3.

4. 5.

after drainage (field capacity), minimum water content for water extraction by the crop, ET partition profile, initial soil salinity profile, groundwater thickness and porosity, fraction of irrigation bypassing the rootzone to groundwater, and ET for the irrigation interval. At the beginning of the simulation, the soil water content profile is initialized to the condition following the irrigation at the end of the previous season. Water is extracted from the soil (e), not the groundwater, because the root system develops prior to the first seasonal irrigation. Extraction from each layer is the depth interval multiplied by the difference between the water content after irrigation and the water content before a seasonal irrigation (calculated in Step 7). Seasonal irrigation begins after the root system is developed. Surface irrigation water moves into the cracks, infiltrates horizontally (hi), and a fraction bypasses (bp) below the rootzone. The bypass flow arrives at the groundwater system and is retained for later crop extraction via upward flow from the groundwater. Water infiltrates horizontally and wets the soil profile. The bottom layer remains near

832

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-4. Drainage as bypass plus internal drainage minus upflow for cracking soil. saturation (the groundwater is in contact with the bottom of layer 7) and does not infiltrate water horizontally. 6. Water drains vertically through the soil profile step by step after horizontal infiltration. The amount of internal drainage (id) from each layer is the difference between the wet water content and the water content after irrigation multiplied by the thickness of the layer. During the first step of this internal drainage phase, the drainage from layer 1 cascades through the layers below. For the second step, the drainage contributed by layer 2 cascades down through the underlying layers. This process repeats four more times for a total of six steps, not five, because the seventh layer remains wet and does not gain or release water. Groundwater is in contact with the bottom of layer 7 and water passes downward and upward, but water content is unchanged. Less water drains through the upper layers than the lower layers because irrigation water short-circuits the upper layers during crack filling. As with the bypass flow, the internal drainage is retained for potential crop water use via upflow from the water table.

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

833

FIGURE 27-5. Schematic of water flow model for cracking soil. 7. Water extracted from each layer is the product of the ET for the period, the partition factor, and the depth interval. ET for the layer is subtracted from available water and the water content is adjusted if available water is sufficient. Available water is the water stored in the soil between the water content after drainage (after) and the allowable

834

8.

9.

10.

11. 12. 13.

14.

15.

16.

17.

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

depletion water content (allowable). If available water storage is insufficient, soil water content is reduced to the allowable water content. If ET exceeds the available water, as just mentioned in Step 7, the groundwater contribution needed to meet the demand is calculated as ET minus available soil water storage. Water flow from the groundwater through the bottom of layer 7 is the sum of that required for all layers, while water movement past the bottom of layer 1 is the groundwater requirement for layer 1 only. To complete a seasonal irrigation period, retained bypass and retained internal drainage in excess of upflow are released as drainage. Thus, drainage is bypass plus internal drainage minus upflow and must be greater than zero. If additional irrigations are required, Steps 4 through 9 are repeated. If reclamation is not required, the next crop is simulated. Reclamation irrigation is applied after the final seasonal irrigation period, if required to lower soil salinity. Reclamation leaching is a water application to remove and/or redistribute salt in the profile in preparation for the next crop. Irrigation water moves into the cracks, infiltrates horizontally, and a fraction bypasses below the rootzone, the same as in Steps 3 and 4. Water infiltrates horizontally and wets the soil profile. The bypass flow arrives at the groundwater system and is released from the groundwater by drain flow. Additional irrigation water is applied at the surface and, because the cracks have swelled shut, all the applied water flows through the top layer [vertical infiltration (vi)] and does not short-circuit. This water cascades down through the profile as a pulse. Internal drainage continues to produce drain flow via the groundwater system during vertical infiltration, and this is released as drainage water. Vertical infiltration ceases when ponding stops and the internal vertical drainage is drained from the groundwater. Water in the soil profile moves downward according to the internal drainage process presented above (for a seasonal irrigation). Drainage is released from the groundwater. If another crop is called for, Steps 1 through 16 are repeated. The final water content is the initial water content in Step 2. If the final crop is reached, simulation ends and the results are presented.

SOLUTE TRANSPORT MODEL Because chemical reactivity is central to the solute transport model, the chemical model will be presented first as it applies to evapoconcentration

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

835

of soil water from ET and mineral precipitation (salt deposition) and then to mixing of infiltrating irrigation water and resident soil water and mineral dissolution (salt pickup). Following the description of the chemical model, the transport process models will be explained and, finally, the solute transport model will be summarized. Evapoconcentration/Precipitation Algorithm for Drip-Irrigated Grape The Colorado River has been the source water for irrigation in Imperial Valley for the past 100 years or so, and in Coachella Valley for about 50 years or so. The dissolved mineral salts in this water sampled at the drip emitter in the vineyard is made up of 4.22 meq/L Ca, 3.07 meq/L Mg, 6.51 meq/L Na, 3.65 meq/L Cl, 2.33 meq/L HCO3, and 7.41 meq/L SO4 ions. The chemistry of this river water was evaluated using an equilibrium chemistry model known as WATSUIT (Oster and Rhoades 1975). Colorado River water was found to be saturated with respect to calcite (CaCO3), meaning it would have a tendency to deposit calcite upon exposure to the atmosphere. When the 898 mg/L Colorado River is evapoconcentrated one-fold in WATSUIT, the equilibrium salt concentration decreases to 871 mg/L because of calcite precipitation. By 10-fold evapoconcentration and above, both calcite and gypsum (CaSO4 2H2O) are deposited and the concentration after salt (mineral) precipitation is increasingly less than the salt concentration after ET, as shown in Fig. 27-6. Data points from WATSUIT for 1-, 1.3-, 2-, 3.33-, 5-, 10- and 20-fold “salt concentration before precipitation” (i.e., before mineral precipitation but after evapoconcentration) are plotted along with “salt concentration after precipitation” (i.e., after mineral precipitation), and a linear salinity function (Wallender et al. 2007) quantifies the reactivity of the Colorado River water. This salt reactivity function is also used in the inverse manner. The initial “concentration after precipitation” is first transformed into “concentration before precipitation” using the inverse of the function shown in Fig. 27-6. Then this preprecipitated solution is evapoconcentrated. Finally, the function is used to calculate the “concentration after precipitation.” Mineral precipitation is calculated as the difference in initial and final salt mass, knowing the initial and final concentrations, water contents, and cell volume. The following example illustrates the evapoconcentration procedure. If, by ET, the water content decreases from 0.125 to 0.0125, the after-ET concentration increases from 899 to 8,990 mg/L [(0.125/0.0125)  899)]. The “concentration after salt precipitation” can be found by drawing a vertical line from 8,990 mg/L on the x-axis to the curve, drawing a horizontal line from the intersection of the curve to the y-axis and reading off (or by using the regression equation in Fig. 27-6), the “concentration

836

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-6. Reactivity of Colorado River water. “Before precipitation” is the salt concentration after ET but before mineral precipitation. “After precipitation” is the salt concentration after mineral precipitation. A second-order polynomial describes this salt reactivity function. It can be found by drawing a vertical line from 8,990 mg/L on the x-axis to the curve, drawing a horizontal line from the intersection of the curve to the y-axis, and reading off (or by using the regression equation in this figure) the “concentration after salt precipitation” (7,857 mg/L). Salt precipitation is the decrease in salt mass caused by evapoconcentration. The final mass of dissolved salt is the product of final “concentration after salt precipitation,” final water volumetric water content, and the cell volume or 174.8 mg (7,857 mg/L  0.0125  0.00178 m3  1,000 L/m3). Initial mass of dissolved salt is the product of initial “concentration after salt precipitation,” initial water content, and depth interval or 200.0 mg (899 mg/L  0.125  0.00178 m3  1,000 L/m3). The mass of salts (minerals) precipitated is initial minus final salt mass or 25.2 mg (200.0 mg  174.8 mg).

after salt precipitation” (7,857 mg/L). Salt precipitation is the decrease in salt mass caused by evapoconcentration. The final mass of dissolved salt is the product of final “concentration after salt precipitation,” final water volumetric water content, and the cell volume or 174.8 mg (7,857 mg/L  0.0125  0.00178 m3  1,000 L/m3). Initial mass of dissolved salt is the product of initial “concentration after salt precipitation,” initial water content, and depth interval or 200.0 mg (899 mg/L  0.125  0.00178 m3  1,000 L/m3). The mass of salts (minerals) precipitated is initial minus final salt mass or 25.2 mg (200.0 mg  174.8 mg).

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

837

Mixing and Dissolution Algorithm The process of mixing solutions of different concentrations and volumes will cause dissolution (negative precipitation or salt pickup) in the presence of excess soil calcite and gypsum derived previously from evapoconcentration of this irrigation water. Figure 27-6 is again used to calculate the final “concentration after precipitation.” First, the “concentration after precipitation” for each solution to be mixed is converted to “concentration before precipitation.” Second, the solutions are mixed according to concentration and volume. Finally, the volume-weighted salt “concentration before precipitation” of the mixture is used along with Fig. 27-6 to estimate the “concentration after precipitation” for Colorado River water. The initial-minus-final concentration after precipitation is multiplied by the mixed volume to give dissolution (negative precipitation).

Transport Algorithm for Drip-Irrigated Grape For salt transport, water enters the cell from above and/or the left and it is mixed with all water initially stored in the cell. After mixing, it moves out the bottom or to the right of the cell (Fig. 27-3). The concentration of water leaving the cell is the mixed concentration. Mass of salt transported across each boundary is the product of the volume of water flowing across the boundary and “concentration before precipitation.” The volume of solution remaining in the cell before mixing is the initial soil solution volume, and its concentration is the initial soil solution “concentration before precipitation.” After mixing, the remaining solution concentration is converted back to “concentration after precipitation” for the final solution concentration. The final salt mass after precipitation is the product of “concentration after precipitation” and the final volume of water. “Change in storage” is the difference between the salt mass before and after flow and mixing. The chemical and mixing models are applied in conjunction with the water flow model in the solute transport model. For the drip model, input data include the initial soil salinity distribution with depth and distance from the emitter; irrigation water salinity; rain salinity; and the chemical reactivity function. For the surface irrigation model, initial groundwater is needed. In both cases, prior to the first irrigation the soilwater concentration increases via evapoconcentration and mineral precipitation occurs using the evapoconcentration/precipitation algorithm. There is no transport between cells without irrigation or rain. For drip irrigation, the initial soil solution is mixed with the solution entering the top or left boundary of the cell during irrigation or rainfall. The mixed solution moves down as well as away from the emitter in the horizontal plane and the process propagates through all cells. For surface irrigation,

838

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

salt moves in the vertical direction only following the algorithm. The transport process reverses for upflow from the groundwater. Under drip irrigation during horizontal redistribution, salt is once again moved using the flows and the concentrations. At the first cell, the final concentration is simply the initial concentration after drip irrigation. At the second cell, the concentration is the weighted concentration of the flow from the first cell and the initial concentration. The process repeats until the last cell is updated. Evapoconcentration and mineral precipitation occur between irrigations or if ET is zero; then the next irrigation or rainfall is added to the top layer and the process repeats. For surface irrigation, after the last season irrigation, reclamation leaching moves salt downward in the profile. After all the irrigations and rainfalls are complete, soil salinity profiles as well as salt flow across the bottom are displayed as output. Transport Algorithm for Cracking Clay Soil For salt transport, water enters the layer from above (or below), and it is mixed with all water initially stored in the layer before it moves out the bottom (top) of the layer (Figs. 27-4 and 27-5). This process begins in the top (bottom) layer and cascades through the layers to the bottom (top) layer for each step. The concentration of water leaving the cell depends on all the cells above (below), which are contributing internal drainage (upflow) water. The mixing algorithm is used to calculate the final concentration and the dissolution (negative precipitation) of the mixture of water from the layer above, and the initial soil water stored in the layer before a fraction of this water is released to the layer below. Concentration of water entering through the top of the layer is the initial “concentration after precipitation” of the layer above it. Mass of salt transported across each boundary is the product of the volume of water flowing across the boundary and “concentration after precipitation.” The volume of solution remaining in the cell before mixing is the initial soil solution volume and its concentration is the initial soil solution “concentration after precipitation.” In preparation for mixing, the “concentration after precipitation” is converted to “concentration after ET” using the reactivity function for Colorado River water for the solution remaining in the cell and for the solution entering the cell. These solutions are mixed to give a volumeweighted “concentration after ET,” which is converted back to “concentration after precipitation” for a final solution concentration. The final salt mass after precipitation is the product of “concentration after precipitation” and the final volume of water is the sum of the flow from the layer above and the initial soil water volume. Dissolution (negative precipitation) is the difference between the salt mass before and after mixing. After this mixing and dissolution process, water at the final “concentration after

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

839

precipitation” flows to the layer below and the process continues through the remaining cells unit the last layer is drained. For Steps 2 through 6, the process repeats, starting at the top-most draining cell and again cascades down through the rootzone. The process reverses for upflow from the groundwater. Summary of Solute Transport Model The chemical and mixing models are applied in conjunction with the water flow model to give the solute transport model. What follows is a description of the model as well as a corresponding schematic (Fig. 27-7). 1. Initial groundwater salinity and irrigation water salinity are input. 2. At the beginning of the simulation, the salinity profile is initialized by specifying the initial mass of salt in each layer. 3. Prior to the first seasonal irrigation, the soil-water concentration increases and the precipitation occurs using the evapoconcentration/ precipitation algorithm. There is no transport prior to the first seasonal irrigation. 4. Salt is carried in the irrigation water during the first seasonal irrigation. 5. During each seasonal irrigation, irrigation water moves horizontally into the soil and mixes with the soil water. In addition, the bypass flow transports irrigation-water salt into the groundwater systems where it is retained for later mixing and transport up into the soil or released as drainage. 6. Salt in the water internally drains vertically through the soil profile via the transport algorithms following horizontal infiltration. Salt moves through the top and bottom boundary of the bottom layer and, because the concentrations of these flows are different, the salt mass changes during mixing, but the final and initial water contents are the same. 7. As with the bypass transport, salty water from internal drainage is retained and blended with the initial groundwater. These solutions are volume- and concentration-blended using the reactivity function (Fig. 27-6). The blended water moves upward (upflow) to meet the ET demand of the crop (Step 9) or drains away (Step 11). 8. If soil water storage exceeds ET for the layer, the water content in the layer is reduced according to the ET for the layer. If soil-water storage is insufficient, soil-water content is reduced to the allowable depletion water content. The evapoconcentration/precipitation model is used to transform the soil water using the Colorado River water reactivity function (Fig. 27-6). 9. If groundwater flows upward to meet the ET demand, salts are transported using the transport algorithm and the reactivity function for Colorado River water (Fig. 27-6) is used to calculate dissolution during

840

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-7. Schematic of solute transport model for cracking soil. mixing. The initial water content is temporarily elevated by the amount to be extracted by ET. 10. The mixed solution is finally evapoconcentrated/precipitated using the Colorado River water reactivity function to give the final concentration and salt mass. 11. Blended groundwater (Step 7), not used for ET, is released as drainage. The mass of salt discharged is the product of the blended groundwater

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

12. 13. 14. 15.

16. 17. 18.

19.

841

concentration (using the groundwater reactivity function) and the volume of drainage water, which is the volume of bypass plus internal drainage minus upflow. If required for salinity control, reclamation irrigation is applied after the final seasonal irrigation period. Irrigation water transports salt into the cracks where it moves horizontally and mixes with the soil water. Irrigation water that bypasses the rootzone is mixed with the groundwater and drained. Salts move and mix with the vertically infiltrating water after the cracks swell shut. The transport algorithms are used to move the salt along with the water, and the reactivity functions are used in conjunction with mixing. After vertical infiltration, the drain water is mixed with the groundwater and drainage occurs. Ponding stops after the vertical infiltration stops and the salts drain from the profile as in Step 6. Internally drained water is mixed with groundwater and drained from the groundwater. Steps 2 through 17 repeat for multiple-crop simulations, with the year-end salinity profile carrying into the initial condition for the next year. Finally, simulation results are output.

RESULTS Drip Irrigation of Grapes Soil, plant, and irrigation/rainfall data are required inputs to the model. For grape production, the lateral and row spacing was 3.65 m and the emitter spacing was 0.91 m, weekly ET was 43 mm between drip irrigations, and weekly drip irrigation was 49 mm. A 10-degree wedge, oriented at right angles to the lateral, had a radius of 1.82 m and was chosen because the surface between the rows was wetted and soil data were collected along a transect centered on the wedge. The wedge was equally divided into seven layers over the 2.1-m root depth and seven radial intervals, also of equal length for a total of 49 cells of increasing volume with distance from the emitter. Initial water content for this sandy soil profile was assumed uniform throughout the profile at 12.5% by volume; the minimum soil water content for plant water extraction was 6.25% by volume; and the soil had a saturated water content of 25% by volume. The ET partitioning with depth and with distance from the emitter was from Nagarajah et al.’s (1987) study of red flame grapes. Their data were normalized to a function of Y  (1  D), as adapted by Smart et al. (2006) for

842

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

grapes from Gale and Grigal (1987) for forest trees. Y is the proportion of roots (cumulative fraction) from the surface to soil depth D in cm, and  is a fitting function. Nagarajah et al.’s data yielded a  value of 0.975 for flame grapes, in contrast to 0.9278 for n  240 grapes of varying varieties by Smart et al. (2006). We have assumed that this root distribution function (Y) in the vine row serves as the root water extraction pattern to meet ET needs and is the same radially halfway into the space between vine rows. Y, the cumulative fraction of extraction, is converted to extraction by depth (Fig. 27-8). As mentioned previously, the adapted chemical reactivity function (Fig. 27-6) for this Colorado River water was obtained from outputs from WATSUIT. Initial soil salinity profiles at 1/7, 4/7, and 7/7 the distance along the radius are shown in Fig. 27-9a–c. ECe plotted in Fig. 27-9a is the EC of the extract from a saturated soil paste, a traditional unit used to describe soil salinity. ECe in dS/m is obtained from salinity in mg/L using approximate conversion factors (for EC  5 dS/m, EC  mg/L salinity  reciprocal of 735 mg/L per dS/m; for EC  5 dS/m, EC  mg/L salinity 

FIGURE 27-8. ET partition with depth (same used for horizontal distance from emitter for Red Flame grapes). Adapted from Nagarajah et al. (1987).

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

843

FIGURE 27-9. Initial and final soil salinity profiles at (a) 1/7 of r, (b) 4/7 of r, and (c) at 7/7 (r), respectively, for grapes on sandy soil in the Coachella Valley of California.

reciprocal of 800 mg/L per dS/m). This input data came from a particular field irrigated with Colorado River water on a sandy soil. In contrast, the final measured salinity value is the mean of 42 drip-irrigated grape fields in Coachella Valley since data on salinity in the particular field was incomplete. The calibration goal was to reproduce the initial salinity profiles after repeated irrigation cycles by adjusting only the routing factor . After eight cycles the profiles stabilized and the calibrated horizontal-to-vertical flow routing ratio was 0.6. There is remarkable agreement between measured and simulated salinity (Fig. 27-9). The corresponding soil moisture profiles (Fig. 27-10) show the expected high water contents with depth at the emitter, the decrease in surface water content with radial distance, and the increase with depth, at the distal end of the wedge. Deep percolation is greatest at the emitter and declines with radial distance (Fig. 27-11), and 64% of irrigation water passes below the rootzone. Multiplying deep percolation depth unit area and by salt concentration, and adjusting units, shows that the mass of salt passing below the grape rootzone decreases with distance from the emitter (Fig. 27-12), following the trend in deep percolation (Fig. 27-10). Here it must be emphasized that the model is location-specific. It can, however, be applied when knowing the soil type, initial and boundary conditions, and root water extraction pattern, as well as irrigation application quantity and quality. With such readily available information, the model can be applied location-by-location in order to assess flow and quality of deep percolation recharging the groundwater system. In addition,

844

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-10. Volumetric water contents at (a) 1/7 of r, (b) 4/7 of r, and (c) 7/7 of r. the model can be easily applied at the field scale to guide farm irrigation and salinity management. This derives from the capacity to predict the soil water quantity and quality outcomes for possible land and water management scenarios. Surface Irrigation of Cracking Soil Soil, plant, and groundwater data for the model application to surface irrigation on cracking clay soils in the Imperial Valley of California are

FIGURE 27-11. Deep percolation per irrigation with radial distance from the emitter.

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

845

FIGURE 27-12. Mass of salt per unit area in deep percolation per irrigation. required inputs to the model. Water content profiles after irrigation before internal drainage (wet), after internal drainage (after), before irrigation (before), and allowable (allowable), as well as ET partitioning, are shown in Fig. 27-13. The immediately-after-irrigation (wet) and afterdrainage (after) water content profiles were adapted from Waller and Wallender’s (1991) Fig. 5 for cracking clay soil in the Imperial Valley. The ET partitioning was adapted from Wallender et al.’s (1979) study of water table contribution to ET. ET for alfalfa of 1.7 m is from Jensen and Walters (1997) as well as Bali et al. (2001), and the ET for wheat (0.463 m) and lettuce (0.433 m) were from Jensen and Walters (1997). Porosity of the groundwater system was set equal to the saturated water content of layer 7 (0.42) and the groundwater mixing zone (thickness) of 6 m was assumed. The fraction of irrigation water bypassing the rootzone to groundwater of 25% was that reported by Waller and Wallender (1991). The adapted chemical reactivity function is from Tanji (2000). Allowable water content profile and ET partitioning profile were used to calibrate and validate. The goal was to match the salinity profiles following three consecutive years of alfalfa reported by Bali et al. (2001). The model was calibrated to fit the salinity profile after the first year of alfalfa (Fig. 27-14) and was validated using the salinity profiles after the second and third year (Fig. 27-14). The model overpredicts the measured salinity build-up between 0.9 and 1.5 m, but does not predict the slight salinity increase in the top 0.6 m (Fig. 27-15). During reclamation leaching,

846

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-13. Water content profiles after irrigation before internal drainage (wet), after internal drainage (after), before irrigation (before), and allowable (allowable), as well as ET partition for cracking clay soils in the Imperial Valley of California.

FIGURE 27-14. Measured salinity profiles and predicted salinity profiles by the cracking soil model. From Bali et al. (2001).

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

847

FIGURE 27-15. Initial, irrigation 1, irrigation 5, irrigation 10, and postseason soil salinity profiles.

salinity in the top layers should decrease, and the salinity peak should attenuate and move downward if the surface is ponded and there is no bypass flow, as the model predicts (Fig. 27-7). Soil salinity increases during the first 2 years of alfalfa and then does not change appreciably for the next 2 years of alfalfa (Fig. 27-16). In preparation for the wheat crop, a 0.15-m reclamation leaching is applied. As mentioned, the top 0.6 m is reclaimed and the salt bulge between 0.9 and 1.2 m is displaced downward. During wheat production, salinity in the near surface increased but fell deeper in the profile. Salinity decreased at all depths for lettuce and these changes can be explained by the ET and its partitioning. The partitioning per 0.3-m layer from the top down for wheat is 60%, 30%, 5%, 4%, 2%, and for lettuce it is 70%, 30%, 0%, 0%, 0%. Greater extraction in the top 0.6 cm increases salt accumulation in the upper profile. Furthermore, a larger fraction of ET is met from soil moisture storage and less water is required to move upward from the water table. Hence, a greater fraction of applied water passes through the lower layers during internal drainage and the lower layers are leached. Similarly, frequent irrigations of lettuce (Table 27-1) leach the salt out of the profile to levels below initial concentrations (Fig. 27-16) below 0.6 m. For the wheat and lettuce, bypass was reduced from the 25% of irrigation water applied for alfalfa to 10% and 0%, respectively, for wheat and lettuce. Water extraction between irrigations is particularly limited for lettuce and bypass is reduced.

848

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-16. Salinity profiles for a crop sequence of 4 years of alfalfa, a 0.15-m reclamation leaching, and wheat and lettuce for cracking reactive soil. Soil salinity is much higher without the moderating effect of precipitation of salts (Fig. 27-17). The nonlinearity of the reactivity function (Fig. 27-6) limits the concentration after precipitation to those measured by Bali et al. (2001). Ignoring the chemical reactivity effect gives artificially high values for salinity. To simulate the effect of salt pickup as surface flow advances down the field, irrigation water salinity is increased to 3.125 dS/m. The salinity

TABLE 27-1. Summary of Irrigation Performance for Cracking Reactive Soil Alfalfa Alfalfa Alfalfa Alfalfa Year 1 Year 2 Year 3 Year 4 (1) (2) (3) (4)

Reclamation (5)

Wheat (6)

Lettuce Total (7) (8)

ET

1.70

1.77

1.77

1.77

0.00

0.46

0.43

7.91

Irrigation

2.00

2.01

2.01

2.01

0.20

0.56

1.03

9.81

Drainage

0.30

0.24

0.24

0.24

0.20

0.09

0.59

1.90

Leaching Fraction

0.15

0.12

0.12

0.12

1.00

0.17

0.58

0.19

ET/Irrigation

0.85

0.88

0.88

0.88

0.00

0.83

0.42

0.81

ET, evapotranspiration

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

849

FIGURE 27-17. Salinity profiles for a crop sequence of 4 years of alfalfa, a 0.15-m reclamation leaching, and wheat and lettuce for cracking nonreactive soil. profiles shift to the right and exceed threshold salinity of 2.0 and 1.3 dS/m for alfalfa and lettuce, respectively, but not the 8.6-dS/m threshold for wheat (Fig. 27-18). The profiles are consistent with those reported by Bali et al. (2001) in their Fig. 2d, shown as follows (Fig. 27-19).

FIGURE Figure 27-18. Salinity profiles for a crop sequence of 4 years of alfalfa, a 0.15-m reclamation leaching, and wheat and lettuce on cracking nonreactive soil with an irrigation water quality of 3.125 dS/m rather than 1.25 dS/m.

850

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 27-19. Salinity along border reported by Bali et al. (2001) for alfalfa.

SUMMARY AND CONCLUSIONS The conceptual model described in this chapter is location-specific and is adaptable for (1) local soil; (2) initial and boundary conditions to compensate for model structure errors; and (3) variation in infiltrating water, soil, and groundwater. The spreadsheet model is sufficiently robust and flexible to adjust according to measurements commonly taken by irrigation managers. This approach is similar to an irrigation scheduling program in which the model parameters are adjusted according to periodic field measurements. The calibrated model reproduced field salinity profiles remarkably well for drip-irrigated grapes and surface-irrigated cracking clay soils— two complex irrigation systems. With this capacity, the model adjusts to ongoing measurements taken by irrigation managers and thereby predicts soil-water quantity and quality outcomes for possible land and water management scenarios. See also Chapter 34 for a discussion of how this type of modeling can be adapted to link management decision-making at various spatial and temporal scales.

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

851

REFERENCES Anderson, J. L., and Bouma, J. (1977a). “Water movement through pedal soils: I. Saturated flow.” Soil Sci. Soc. Am. J., 41, 413–418. ———. (1977b). “Water movement through pedal soils: II. Unsaturated flow.” Soil Sci. Soc. Am. J., 41, 419–423. Andreini, M. S., and Steenhuis, T. S. (1990). “Preferential paths of flow under conventional and conservation tillage.” Geoderma, 46, 85–102. Annandale, J. G., Jovanovic, N. Z., Campbell, G.S., Du Sautoy, N., and Benade, N. (2003). “A two-dimensional water balance model for micro-irrigated hedgerow tree crops.” Irrig. Sci., 22(3–4), 157–170. Bali, K. M., Grismer, M. E., and Tod, I. C. (2001). “Alfalfa water use and reducedrunoff irrigation management in the Imperial Valley, California.“ ASCE J. Irrig. Drainage Eng., 127(3), 123–130. Blake, G., Schlichting, E., and Zimmermann, U. (1973). “Water recharge in a soil with shrinkage cracks.” Soil Sci. Soc. Am. J., 37, 669–672. Bouma, J., and Dekker, L. W. (1978). “A case study on infiltration into dry soil.” Geoderma, 20, 27–40. Casey, F. X. M., Logsdon, S. D., Horton, R., and Jaynes, D. B. (1997). “Immobile water content and mass exchange coefficient of a field soil.” Soil Sci. Soc. Am. J., 61, 1030–1036. Clothier, B. E., Kirkham, M. B., and McLean, J. E. (1992). “In situ measurement of effective transport volume for solute moving through soil.” Soil Sci. Soc. Am. J., 56, 733–736. Coats, K. H., and Smith, B. D. (1964). “Dead-end pore volume and dispersion in porous media.” Soc. Pet. Eng. J., 4, 73–84. Cook, F. J., Thorburn, P. J., Fitch, P., and Bristow, K. L. (2003). “WetUp: A software tool to display approximate wetting patterns from drippers.” Irrig. Sci., 22(3–4), 129–134. Cote, C. M., Bristow, K. L., Charlesworth, P. R., Cook, F. J., and Thornurn, P. J. (2003). “Analysis of soil wetting and solute transport in subsurface trickle irrigation.” Irrig. Sci., 22(3–4), 143–156. Gale, M. R., and Grigal, D. F. (1987). “Vertical distribution of northern tree species in relation to successional status.” Can. J. Forestry Res., 17, 829–834. Gardenas, A. I., Hopmans, J. W., Hanson, B. R., and Simunik, J. (2005). “Twodimensional modeling of nitrate leaching for various fertigation scenarios under micro-irrigation.” Agric. Water Mgmt., 74, 219–242. Jensen, M. E., and Walters, I. A. (1997). Assessment of 1987–1999 water use by the Imperial Irrigation District using water balance and cropping data, Special Report to U.S. Bureau of Reclamation, U.S. Dept. of the Interior, Bureau of Reclamation, Washington, D.C. Kissel, D. E., Ritchie, J. T., and Burnett, E. (1973). “Chloride movement in undisturbed swelling clay soil.” Soil Sci. Soc. Amer. Proc., 37, 21–24. Li, Y., and Ghodrati, M. (1997). “Preferential transport of solute through soil columns containing constructed macropores.” Soil Sci. Soc. Am. J., 61, 1308–1317. Ma, Q. L., Ahuja, L. R., and Rojas, K. W. (1995). “Measured and RZWQM predicted atrazine dissipation and movement in field soil.” Trans. ASAE, 38, 471–479.

852

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Nagarajah, S., Kinnard, C., Salter, D., McDonald, D., Nesbitt, A., and Kenna, G. (1987). “Root distribution patterns in drip irrigated Red Globe, Flame and Menindee grapevines,” in Horticulture, Section 4.2.4., NSW Dept. of Primary Industries, Alstonville, New South Wales, Australia, 45–55. Oster, J. D., and Rhoades, J. D. (1975). “Calculated drainage water compositions and salt burdens resulting from irrigation with river waters in the western United States.” J. Environ. Qual., 4, 73–79. Ritchie, J. T., Kissel, D. E., and Burnett, E. (1972). “Water movement in undisturbed swelling clay soil.” Soil Sci. Soc. Am. Proc., 36, 874–879. Schwartzman, M., and Zur, B. (1986). “Emitter spacing and geometry of wetted soil volume.” ASCE J. Irrig. Drainage Eng., 112, 242–253. Seyfried, M.S., and Rao, P. S. C. (1987). “Solute transport in undisturbed columns of an aggregated tropical soil: Preferential flow effects.” Soil Sci. Soc. Am. J., 51, 1434–1444. Sˇ imu˚nek, J., Sejna, M., and van Genuchtan, M. Th. (1999). The Hydrus-2D software package for simulating two-dimensional movement of water, heat and multiple solutes in variably saturated media. Version 2.0, IGWMC-TPS-53, International Ground Water Modeling Center, Colorado School of Mines, Golden, Colo., 1–251. Singh, D. K., Rajput, T. B. S., Singh, D. K., Sikarwar, H. S., Sahoo, R. H., and Ahmad, T. (2006). “Simulation of soil wetting pattern with subsurface drip irrigation from line source.” Agric. Water Mgmt., 83, 130–134. Singh, P., and Kanwar, R. (1991). “Preferential solute transport through macropores in large undisturbed saturated soil columns.” J. Environ. Qual., 20, 295–300. Skaggs, T. H., Trout, T. J., Sˇ imu˚nek, J., and Shouse, P. J. (2004). “Comparison of HYDRUS-2D simulations of drip irrigation with experimental observations.” J. Irrig. Drainage Eng. ASCE, 130(4), 304–310. Smart, D. R., Schwass, E., Lakso, A., and Morano, L. (2006). “Grapevine rooting patterns: A comprehensive analysis and a review.” Am. J. Enology and Viticul., 57, 89–104. Tanji, K. K. (2000). Salt balance in Imperial Irrigation District, A technical paper for Water Advisory Committee, Coachella Valley Water District, Coachella, Calif. van der Ploeg, R. R., and Benecke, P. (1974). “Unsteady, unsaturated n-dimensional moisture flow in soil: A computer simulation program.” Soil Sci. Soc. Am. Proc., 38, 881–885. van Genuchten, M. T., and Dalton, F. N. (1986). “Models for simulating salt movement in aggregated field soils.” Geoderma, 38, 165–183. van Keulen, H., and van Beek, C. G. M. (1971). “Water movement in layered soils: A simulation model.” Neth. J. Agric. Sci., 19, 138–153. Wallach, R., and Steinhuis, T. S. (1998). “Model for nonreactive solute transport in structured soils with continuous preferential flow paths.” Soil Sci. Soc. Am. J., 62, 881–886. Wallender, W. W., Grimes, D. W., Henderson D. W., and Stromberg L. K. (1979). “Estimating the contribution of a perched water table to the seasonal evapotranspiration of cotton.” Agron. J., 71, 1056–1060. Wallender, W. W., Tanji, K. K., Clark, B., Hill, R. W., Stegman, E. C., Gilley, J. R., Lord, J. M., and Robinson, R. R. (2007). “Drip irrigation water and salt flow

CONCEPTUAL WATER FLOW AND SALT TRANSPORT

853

model for table grapes in Coachella Valley, California.” Irrig. and Drain. Sys., 21(2), 79–95. Wallender, W. W., Tanji, K. K., Gilley, J. G., Hill, R. W., Lord, J. M., Moore, C. V., Robinson, R. R., and Stegman, E. C. (2006). “Water flow and salt transport in cracking clay soils of the Imperial Valley, California.” Irrig. and Drain. Sys., 20(4), 361–387. Waller, P. M., and Wallender, W. W. (1991). “Infiltration in surface irrigated swelling soils.” Irrig. and Drain. Sys., 5, 249–266. Warrick, A. W. (1986). “Soil water distribution,” in Trickle irrigation for crop production, 1st ed., F. S. Nakayama and D. A. Bucks, eds., Chapter 2.3, Elsevier, New York. Warrick, A. W., and Orr, D. (2007). “Soil water concepts,” in Trickle irrigation for crop production, 2nd ed., F. S. Nakayama and D. A. Bucks, eds., Chapter 2, Elsevier, New York.

This page intentionally left blank

CHAPTER 28 MODELING TRANSIENT ROOTZONE SALINITY (SWS MODEL) Donald L. Suarez

INTRODUCTION Modeling processes in soils serves multiple purposes. Most models are developed for either regulatory or management purposes, in which case they should be easy to use and have readily available input requirements. The management models usually contain a set of simplified and generalized scientific relationships but may sometimes be exclusively statistical and thus without explicit description of processes. Alternatively, models developed for research purposes consider a set of known or hypothesized processes serving as a tool for data analysis and thus furthering the scientific understanding of processes. Some take the position that all models, but especially management models, should be as simple as needed to represent the data. The difficulty with this approach is that unless the data set is extremely robust, only a limited set of conditions are examined and thus represented in the model. In this instance, locations with different specific conditions may result in unsatisfactory predictions. Most modeling efforts result in simulation of existing data, with the modeler adjusting input parameters, enabling a satisfactory match between the model and the data. Such an approach is of very little use as a management tool, where predictive capability is required and the collected data may not be sufficient to allow for validation. Adding more complexity to a model, even if the science is correctly represented in the mathematical relationships, it may not necessarily improve the absolute predictive capability of the model. However, the advantage of a model based on tested algorithms of processes is that a user can evaluate the impact of changes on those processes. If the model has incorporated the major processes controlling the system, then the 855

856

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

model will be able to provide predictions of the system response to changes and thus serve as a valuable management tool. The objective of a management model should thus be to represent the underlying process, without undue burden on the user for collection of site-specific characterization or parameter information. In this chapter we describe the development of rootzone salinity models, the SWS (Soil-Water-Salinity) salinity model and processes used by the model, as well as applications to management of saline soils or waters.

MODEL DEVELOPMENT Both steady-state and transient models have a place in management of saline soils and waters. Steady-state models are easier to use and provide information about the long-term effect of a given change, such as the effect of a different irrigation water on soil properties. These models will input data, such as the average irrigation water volumes and concentrations, and average evapotranspiration (ET), and will calculate average yearly leaching fraction (LF) and soil salinity. Transient models provide detailed information on temporal changes; this can be either unneeded detail (if the time scale of the change is exceedingly short) or critical (if the change is at a time frame that impacts other parts of the system, such as plant response to salt stress). As will be demonstrated, transient modeling is not just consideration of the transition from one steady state to another but, rather, analysis of the continual dynamic change experienced by natural systems. Depending on the scale of the changes, steady state can be considered for systems with seasonal fluctuations, but some processes, for instance, plant response to water salinity or toxic ions, may occur on a much shorter time scale. The initial salinity models considered either mass balances of either water or chemicals. These models are still used, primarily for analysis of large scale systems, such as irrigation districts or hydrologic basins. Among these models are the ASTRAN model, (Labadie and Khan 1979) and the Hydrosalinity model, which also considers gypsum dissolution-precipitation (Quilez et al. 2011; Chapter 30 of this manual). Both of these models have been applied for salinity management at the basin scale. Subsequently, rootzone water and chemical steady-state models were developed and then transient flow models. Hanks and Bowers (1962) developed a numerical solution for description of variably saturated water flow. Chemical models also evolved from mass balance models to thermodynamic equilibrium models, to models with kinetic considerations. Dutt (1962) developed a computer program to predict gypsum sol-

MODELING TRANSIENT ROOTZONE SALINITY

857

ubility coupled with cation exchange. Truesdell and Jones (1974) developed the WATEQ mineral equilibrium model, enabling determination of the mineral saturation status of waters. Bresler (1973) described modeling of water and nonreactive solutes (coupled water and solute flow) under transient unsaturated flow conditions. Robbins et al. (1980) developed a combined chemical model considering cation exchange and calcite and gypsum equilibria coupled with variable water flow. This approach was further developed by Wagenet and Hutson (1987), into LEACHM, a model that is still widely utilized. Suarez and Sˇ imu˚ nek (1992, 1997) described the UNSATCHEM model, which has similar objectives to LEACHM (with combined water flow, cation exchange, and equilibrium expressions for calcite and gypsum) but with added processes and interactions. Among the unique features, the model includes descriptions of kinetic expressions for calcite, a CO2 production and transport routine for prediction of CO2 concentrations needed for pH prediction, a boron (B) transport routine that considers adsorption-desorption as a function of pH, application of Pitzer expressions (Pitzer 1973; Felmy 1990) for ion activity calculations at high ionic strength and calculation of osmotic pressure, as well as a routine describing the impact of chemical properties [electrical conductivity (EC), sodium adsorption ratio (SAR), and pH] on the soil hydraulic properties. The outputs from the various models are not always in agreement (Suarez and Dudley 1998) due to assumptions made regarding system response. For example, work by Robbins et al. (1980) and the initial LEACHM model (Wagenet and Hutson 1987) assume a fixed input CO2 but do not properly predict soil pH and alkalinity changes, apparently due to the numerics of the calcite routine, which is only clearly evident with irrigation of high-alkalinity waters (Suarez and Dudley 1998). Current versions of LEACHM have corrected this problem (J. L. Hutson, personal communication, April 2008). The Dutt et al. (1972) model assumed that the soil is a closed system (no transfer of material in or out of the system) with respect to CO2. The model predicts that the pH increases and the CO2 concentration decreases as the water content increases and calcite is dissolved. This is correct for a closed system (such as soil and water in a closed flask over a short time interval), but the prediction is contrary to field observations (Buyanovsky and Wagner 1983) and considerations that there is gas exchange between the soil and the atmosphere. Most models consider a fixed soil pH or fixed CO2 (specified by the user). The fixed pH assumption is reasonable only for extremely shortterm simulations or where the water composition remains constant. The fixed-CO2 assumption is preferable to the closed-system assumption and to the fixed-pH assumption because it allows prediction of the impact of water composition and mineral reactions on soil pH. It is reasonable to

858

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

assume that the chemical reactions have a minimal impact on the gas composition in soil; thus, the gas composition can be defined outside of the chemical system. However, as discussed in the following sections, the fixed-CO2 assumption may be adequate for a steady-state model, but the CO2 concentration is not constant because the soil is a dynamic system, where the gas concentration is defined by the processes of production and transport. The UNSATCHEM model predicts that as the water content increases, the CO2 transport out of the rootzone is decreased and the CO2 concentration increases. These predictions are consistent with observations of O2 depletion under wet conditions. Following are descriptions of some of the relevant processes.

SWS MODEL The SWS model was developed as a user-friendly transient-water-flow chemistry model for salinity management. The modeling approach is deterministic in that it is based on a set of mathematically defined process, such that with each set of data input a unique and reproducible prediction is obtained (Addiscott and Wagenet 1985). The base of the program is that of UNSATCHEM (Suarez and Sˇ imu˚ nek 1992, 1997) with addition of calculations for ET, a new B adsorption routine, and with a userfriendly interface that makes extensive use of default parameters to minimize the need for user expertise in soil physics and chemistry. Water Flow Hydraulic functions The SWS model uses a modified version of the one-dimensional Richards’ equation: ⎞⎤ ∂w ∂ ⎡ ⎛ ∂h  ⎢k ⎜  1⎟⎥  S ⎠⎦ ∂t ∂z ⎣ ⎝ ∂z

(28-1)

where h is the water pressure head, w is the volumetric water content, K is the hydraulic conductivity, t is time, z is the depth coordinate, and S is the sink term, representing extraction of water from the soil by plant roots. The effects of thermal and density gradients on water flow are neglected, and it is further assumed that the gas phase dynamics do not affect water flow. These simplifications are not justified in all instances. For example, density gradients can be significant when saline waters are present, but consideration of these processes increases the complexity beyond the scope of this already complex model.

MODELING TRANSIENT ROOTZONE SALINITY

859

The unsaturated soil hydraulic properties are described by a modified version of those proposed by van Genuchten (1980). The water retention and hydraulic conductivity (HC) functions are given by ( h)  r 

 s  r (1  hn )m

(28-2)

and )m]2 K(h)  Ks Krr  KsrSe1/2[1  (1  S1/m e

(28-3)

respectively, where m  1  1/n

n1

(28-4)

and Se 

  r  s  r

(28-5)

where r and s are the residual and saturated water content (expressed as cm3cm3), respectively, Ks is the saturated conductivity [cm d1], Kr is the relative HC (scaled from 0 to 1), Se is relative saturation, and m, n, and

[cm1] are the empirical parameters of the hydraulic characteristics. In order to increase numerical stability in the range of h from 0 to 2 cm, we specify a constant (s) for that interval. Hydraulic characteristics are determined by the set of six parameters, r, s, , n, Ks, and the unique variable r, representing a reduction function (scaled from 0 to 1) describing the effect of soil chemistry on hydraulic properties and is discussed in more detail later. Use of the model requires optimizing the first five parameters from the experimental water retention, pressure head, and saturated conductivity data. It is not realistic to expect users of a management model to conduct detailed studies on the water retention curve and unsaturated HC of each soil considered. For use in a crop-irrigation management model, the water retention versus pressure head curve can be approximated by the functions obtained from soil texture (Carsel and Parrish 1988), and thus are included in the interface of the present model. The estimates of saturated HC given in this data set are likely the major error for our applications to irrigated agriculture. The saturated HC is important because water or rain applied at a rate in excess of infiltration may result in surface runoff and thus infiltration below the applied amount. In some instances the values presented by Carsel and Parrish (1988) appear greater than what we

860

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

observe locally, for example, Ks for a loam soil. In this instance, it is suggested that the user maintain the water retention versus pressure head curve of the default parameters based on the soil texture (Carsel and Parrish 1988) and then input their own or local estimates of Ks for their soil. Alternatively, users with hydraulic information can use the advanced option and input their own hydraulic parameters. Chemical effects on hydraulic conductivity It is well known that soil hydraulic properties are affected by chemical properties of the soil, but to date this is not accounted for in other models. Equation 28-3 differs uniquely from previous hydraulic expressions in that it includes a reduction term, r, which scales the HC in relation to the EC, pH, and SAR conditions in the soil. Optimal soil chemical conditions for infiltration are represented by values of r  1. Elevated levels of exchangeable Na result in swelling of smectitic clays, detachment of clay particles, dispersion, and subsequent clay migration and redeposition. All of these processes result in blocking of pores at low salinity and in the presence of exchangeable sodium (McNeal 1968; Shainberg and Levy 1992). These processes are observed in the natural development of clay pan layers in soils and, more dramatically, in sodic, nonsaline soils. In addition, elevated levels of pH adversely affect saturated HC, separate from the sodicity and salinity interactions (Suarez et al. 1984). Suarez and Sˇ imu˚ nek (1997) represented the chemical effects on hydraulic properties by the use of a reduction function, r, given by r  r 1 r2

(28-6)

where r1 is the reduction due to the combined adverse effects of low salinity and high exchangeable sodium fractions on the clay, and r2 is the adverse effect of pH. The r1 term is given by McNeal (1968) as r1  1 

cx n 1  cx n

(28-7)

where c and n are empirical factors, and x is defined by x  fm 3.6  104 ESP* d*

(28-8)

where fm is the mass fraction of smectite (defined as montmorillonite and beidellite) in the soil, d* is an adjusted interlayer mineral spacing, and ESP* is an adjusted exchangeable sodium percentage (percentage of

MODELING TRANSIENT ROOTZONE SALINITY

861

the total cation exchange charge neutralized by Na). The term d* is defined by d*  0

C0  300 mmolcL1

d*  356.4(C0)0.5  1.2

C0 300 mmolc L1

(28-9)

where C0 is the total salt concentration of the solution, and the term ESP* is given by ESP*  ESPsoil  (1.24  11.63 logC0)

(28-10)

The reduction factor r2, representing the effect of pH on HC, was calculated from the SAR-pH saturated HC experimental data given in Suarez et al. (1984). The data were first corrected for the effects of salinity and exchangeable sodium using the r1 values calculated from the aforementioned. Based on this limited data, r2  1

for pH  6.83

r2  3.46  0.36 pH

for 6.86 pH 9.3

r2  0.1

for ph  9.3

(28-11)

In view of the differences among soils, these specific values may not be generalized predictors of soil HC, but they do represent conditions of arid land soils examined at the U.S. Salinity Laboratory. The response of soil hydraulic properties to pH has not been extensively studied, but it is reasonable to assume that soils differ in their reaction to these factors. This option in the model should not be considered as a quantitative prediction of what will occur at a specific site but is useful to evaluate the relative importance of the chemical effects under different soil and water conditions. Many other factors in addition to sodicity and pH affect soil aggregate stability, such as organic matter, soil texture, clay mineralogy, or tillage, and it is reasonable to assume that there are interactions between these factors and the chemical factors considered here. By use of the reduction function it is implied that the relative response to sodic conditions, obtained under saturated conditions, is applicable to unsaturated conditions. This is likely not entirely correct, but for irrigated agriculture the most important water flow occurs at and near saturation (as water is applied) and the available data are entirely based on saturated flow experiments. Another important simplifying assumption is that the processes are reversible. This assumption is likely valid when the reduction in HC is due to swelling, but it is likely not valid when clay

862

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

dispersion occurs. The extent to which sodic reclamation restores HC (with or without tillage) is unknown. Extensive research on reclamation has focused primarily on chemical changes and secondarily on improvements in hydraulic properties relative to the initial degraded condition. Plant Modeling Water uptake by plant roots Water uptake by plants is related to the rooting distribution. There are two options in the SWS model related to root distribution: a user-specified fixed root distribution and an initial user-specified distribution coupled to root growth. The fixed rooting distribution option is used when simulating perennial crops such as alfalfa and pasture grasses, but it can also be used for simplified input for annual crops. Water uptake by plants is related to the rooting distribution, input ET, and water and salt stress simulated by the model. The model predicts relative yield based on the ratio of predicted ETa, (actual ET calculated by the model after consideration of stress) to optimum ETc, (optimum ET of the crop). The root growth option can be used for simulation of annual crops. In this case the user inputs an initial root distribution from which the roots will develop. This option requires additional inputs, such as initial rooting depth, maximum rooting depth, and growing degree days for the crop. The sink term in Eq. 28-1 is the volume of water removed from a unit volume of soil per unit of time as a result of plant water uptake. The root water uptake in response to water and salinity stress is expressed as S  Sp s(h) (h)

(28-12)

where Sp is the potential water uptake [cm3cm3d1], s(h) is the water (matric) stress function, h is the matric head [m], (h) is the osmotic stress function, and h is the osmotic head [m]. As shown by Eq. 28-12, water uptake is obtained by multiplying the water stress reduction function, matric stress reduction function, and potential water uptake. The model calculates the stress functions and water uptake at each time step. There is uncertainty as to how to best represent the response to combined stresses, but the multiplicative approach appears preferable to the alternative addition of the osmotic and matric pressure and representation of a single stress function or the assumption that only the most limiting stress need be considered (see Grieve et al., Chapter 13 of this manual, for more discussion).

MODELING TRANSIENT ROOTZONE SALINITY

863

The water stress response function, s(h), is a dimensionless function of the soil water pressure head (0 s(h) 1) described by van Genuchten (1987) as

s ( h) 

1 p ⎛ h ⎞ 1⎜ ⎟ ⎝ h50 ⎠

(28-13)

where h50 [m] and p are empirical constants. The default parameter of the model is set at h50 equal to 50 m and p  3. The parameter h50 represents the pressure head at which the water extraction rate is reduced by 50%. Specific crop values of h50 are not available but a default value can be estimated from the wilting point. This water stress response function, s(h), does not consider transpiration reduction near saturation. The decrease in water uptake that is sometimes observed at saturation is related to oxygen stress and is more properly treated based on prediction of the gas phase composition (for models such as SWS that include CO2 production and transport). An expression similar to Eq. 28-13, only with (h) and h50 for osmotic stress, is used for salinity:

 ( h ) 

1 ⎛ h ⎞ 1⎜  ⎟ ⎝ h 50 ⎠

p

(28-14)

Specific values of the h50 and p parameters for salinity are presented in the model for selected crops. If other crops are selected it is suggested that in the absence of detailed information, the h50 value be calculated from the more traditional Maas-Hoffman relationship (Grieve et al., Chapter 13 of this manual; Maas and Hoffman 1977). The Maas-Hoffman relationship represents data in terms of a threshold electrical conductivity (EC) or osmotic pressure above which there is a yield decline and a slope that describes the yield decline with increasing salinity (expressed in terms of EC or osmotic pressure). From the salt response threshold and slope, the h50 can be easily calculated and p can be set to 3.0. The potential water uptake rate in the rootzone is expressed as the product of the potential transpiration rate, Tc [cm d1], and the normalized water uptake distribution function, (z) [cm1], where z is depth (cm). The normalized water uptake distribution function describes the variation with depth of the potential water uptake rate, Sp, over the rootzone, as follows: Sp  (z)Tc

(28-15)

864

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

For a fixed root distribution, the function (z) is specified by the user. The actual (calculated) transpiration Ta is given by L

L

Ta 

∫

LLr

S( h, h , z)dz  Tc

∫

s ( h)  ( h )( z)dz

(28-16)

LLr

The terms L and Lr are, respectively, the depth at the soil surface (0) and the depth of the deepest root. The total actual transpiration for each time interval is calculated by summation of the actual transpiration amounts for each of the rootzone depth intervals. The transpiration in each of the depth intervals is based on the root distribution function, the potential crop transpiration, and the stress calculated in that depth interval. There is currently no compensation at other depths for reduced water uptake within any depth interval. The total transpiration for the simulation is the sum of the actual transpiration time intervals. The ratio of actual transpiration to optimal transpiration is used to calculate the relative yield. This calculation does not currently consider the change in water use efficiency under salt stress. Changes in water use efficiency (unit fresh weight/unit of water consumed) related to salt stress are currently available for only a few crops and cannot be generalized because some increase and some decrease with increasing salinity. The fixed root option is always selected for a perennial crop. It is also possible to use the fixed root option for predicting the water uptake and relative crop yield for an annual crop. In this instance, the input ETc values are ET0 multiplied by the crop coefficient. Values for these coefficients are crop- and locality-specific, as well as varying with time during growth, thus are ideally provided by the model user. Use of this option requires more detailed information but may provide more accurate prediction of water requirements and use if the crop factors are known for the crop and locality to be simulated. The user manual presents coefficient data on selected crops and at different stages of growth. Water uptake by plant roots: root growth option A specification of the root growth option enables use of a simplified crop-root growth model. In this instance, the input is still ETc and this value must be input or calculated by the model. Additional plant-specific information is required, including planting date, growing degree days (GDD) to maturity, and harvest date. The plant is divided into various stages of development and the initial rooting depth must be specified. If the shallow initial rootzone dries out, there will be water stress. It is suggested that the user ensure that the initial conditions are reasonable with regard to the initial root distribution and the total amount of water extracted.

MODELING TRANSIENT ROOTZONE SALINITY

865

Root growth. The root depth, Lr, can be either constant or variable during the simulation. For annual vegetation, the plant submodel is required to simulate the change in rooting depth with time. In UNSATCHEM (Suarez and Sˇimu˚nek 1996, 1997) the root depth is the product of the maximum rooting depth, Lm [cm], and the root growth coefficient, fr(t): Lr(t)  Lm fr(t)

(28-17)

To calculate the root growth coefficient, fr(t), Sˇimu˚nek and Suarez (1993) combined the Verhulst-Pearl logistic growth function with the GDD function. The logistic growth function is used to describe biological growth at constant temperature, and the GDD model is utilized for determining the time between planting and plant maturity. The model uses a modified version of the GDD relation developed by Logan and Boyland (1983), who assumed that this function is fully defined by the temperature, T [K], expressed by a sine function approximating the temperature variation during the day, and by the three temperature limits, T1, T2, and T3 [K]. When the actual temperature is below the base value T1, plants do not grow. Plant growth is at a maximum rate at temperature T2, with growth constant up to a maximum temperature T3, above which increased temperature has an adverse effect on growth. Based on this information, Sˇ imu˚ nek and Suarez (1993) developed the following dimensionless growth function:

(28-18)

where TBas are the heat units [KT] necessary for the plant to mature and the roots to reach the maximum rooting depth; tp, tm, and th represent time of planting, time at which the maximum rooting depth is reached and time of harvesting, respectively; and parameter  is the reduction in optimal growth due to the water and osmotic stress. The expression inside the brackets of Eq. 28-18 equals TBas at time tm when roots reach the maximum rooting depth. The individual integrals in Eq. 28-18 are evaluated only when the arguments are positive. Parameter  is defined as the ratio of the actual to potential transpiration rates: 

Ta Tp

(28-19)

866

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Biomass and/or root growth is represented with the Verhulst-Pearl logistic growth function: fr (t) 

L0 L0  (Lm  L0 )ert

(28-20)

where L0 is the initial value of the rooting depth at the beginning of the growth period [cm] and r is the growth rate [cm d1]. Combining the concepts for GDD (Eq. 28-18) and logistic growth (Eq. 28-20), the time to maturity is expressed as (Sˇ imu˚ nek and Suarez 1993) t  tmg(t)

(28-21)

where tm is the time when GDD reaches the specified plant species heat units for maturity (TBas). Calculations of Crop Evapotranspiration Water consumption at any time step is calculated based on ETc and the stress reduction factor. In the absence of input ETc information, the model will predict ETc using the FAO version of the Penman-Monteith equation (Allen et al. 1998), given as ET0 

0.408 Δ(Rn  G)  ( t900 273 )U 2 ( e s  e a ) 10( Δ  ((1  0.34U 2 ))

(28-22)

where ET0 is expressed in cm/day,  is the slope of the saturation vapor pressure curve (kPa °C1), Rn is the net radiation at the crop surface (MJ m2d1), G is the soil heat flux density (MJ m2d1),  is the psychrometric constant (kPa °C1), t is the mean daily air temperature (°C), es is the saturation vapor pressure (kPa) at the specified temperature, ea is the measured or calculated vapor pressure, and U2 is the wind speed at a height of 2 m above the surface (m s1). The terms on the right side of Eq. 28-22 are all calculated using the expressions given in FAO Irrigation and Drainage Paper 56 (Allen et al. 1998). The input variables needed to calculate ET0 using this approach can be reduced to wind speed, latitude, elevation, calendar date, mean daily temperature, daily temperature fluctuation, fraction of the day that is clear, and maximum relative humidity—all factors that should be readily available on a daily basis. Crop coefficients and calculation of ETc Crop coefficients (kc) serve to convert the ET0 values into ETc for the crops of interest. The reference ET0 is for a hypothetical crop with an

MODELING TRANSIENT ROOTZONE SALINITY

867

assumed height of 0.12 m having a surface resistance of 70 s m1 and an albedo of 0.23, resembling a grass crop of uniform height, well-watered, and growing actively. For annual crops, the stage of growth, as well as crop characteristics, affect the coefficients; thus, the kc values must be growth stage-dependent. In the absence of a coupled crop-specific growth model, the transition to various stages depends on climatic factors; thus, the crop coefficients vary according to location as well as time. The SWS user manual presents length of crop stages for use in calculation of crop coefficients, crop coefficients for major crops, and selected locations and planting dates, all taken from FAO Irrigation and Drainage Paper 56 (Allen et al. 1998). Concentration/Production of Carbon Dioxide Factors controlling soil carbon dioxide concentrations The carbon dioxide concentration in the soil air is always elevated relative to the concentration in the earth’s atmosphere. Carbon dioxide is produced in the soil primarily as a result of two processes: microbial respiration and root respiration from plants. The soil CO2 concentration is dynamic with both seasonal changes and short-term changes. Changes in concentration are due to changes in production of CO2 as well as changes in the transport rate of CO2, which is mostly related to changes in the airfilled porosity of the soil but can also be related to the flow of water. In the rootzone and at some distance below it, the quantity of CO2 added or removed by mineral dissolution/precipitation reactions is usually relatively small compared to the gas production and flux values and thus is neglected for simplification. A few meters below the rootzone, microbial respiration is greatly reduced, and mineral reactions may need to be considered in order to predict CO2 gas concentration. Carbon dioxide production. Sˇ imu˚ nek and Suarez (1993) described a general model for CO2 production and transport that is included in the SWS model. The CO2 production is represented as the sum of the production rate by soil microorganisms, s [cm3 cm3T1], and the production rate by plant roots, p [cm3cm3T1]: P   s   p   s 0 ∏ fsi   p 0 ∏ f pi i

(28-23)

i

where the subscript s refers to soil microorganisms, and the subscript p refers to plant roots;  fsi is the overall reduction coefficient for microbial CO2 production and is the product of reduction coefficients dependent on depth, temperature, pressure head (the soil water content), CO2

868

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

concentration, and osmotic head. The term  fpi is the overall reduction coefficient for plant root CO2 production and is the product of reduction coefficients dependent on depth, temperature, pressure head (the soil water content), CO2 concentration, osmotic head, and time. The parameters s0 and p0 represent, respectively, the optimal CO2 production by the soil microorganisms or plant roots for the whole soil profile at 20 °C under optimal water, solute, and soil CO2 concentration conditions [cm3cm2 T1]. The individual reduction functions are given in Sˇ imu˚ nek and Suarez (1993), and a discussion of selection of the values for optimal production, as well as coefficients for the reduction functions, is given in Suarez and Sˇ imu˚ nek (1993). Carbon dioxide transport. The SWS model uses the one-dimensional CO2 transport model presented by Sˇ imu˚ nek and Suarez (1993). The model considers CO2 transport in the soil by both the liquid and gas phases. Thus, the CO2 transport is described by convective transport in the aqueous phase and diffusive transport in both gas and aqueous phases, and by CO2 production and/or removal. The one-dimensional CO2 transport is described by ∂CT ∂  ( J da  J dw  J ca  J cw )  Scw  P ∂t ∂z

(28-24)

where Jda is the CO2 flux resulting from gas phase diffusion [cm d1], Jdw is the CO2 flux resulting from dispersion in the dissolved phase [cm d1], Jca the CO2 flux caused by convection in the gas phase [cm d1], and Jcw the CO2 flux caused by convection in the dissolved phase [cm d1]. The term cT is the total volumetric concentration of CO2 [cm3cm3] and P is the CO2 production/sink term [cm3cm3 d1]. The term Scw represents the dissolved CO2 removed from the soil by root water uptake. This assumes that when plants take up water, the dissolved CO2 is also removed from the soil-water system. Details of the production and transport routines are given in the user manual. SOIL AND WATER CHEMISTRY Transport The governing equation for one-dimensional advective-dispersive chemical transport under transient flow conditions in partially saturated porous media is taken as (Suarez and Sˇ imu˚ nek 1992): ⎤ ∂ˆ ∂cTi ∂ ∂⎡ ∂   cTi   cTi  ⎢D cTi  qcTi ⎥ i  1, ns ⎦ ∂t ∂t ∂t ∂z ⎣ ∂z

(28-25)

MODELING TRANSIENT ROOTZONE SALINITY

869

where cTi is the total dissolved concentration of the aqueous component i [ML3], c–Ti is the total adsorbed or exchangeable concentration of the aqueous component i [Mkg1], cˆTi is the nonadsorbed solid-phase concentration of aqueous component i [Mkg1],  is the bulk density of the soil [ML3], D is the dispersion coefficient [cm2 d1], q is the volumetric flux [cm d1], and ns is the number of aqueous components. The second and third terms on the left side of Eq. 28-25 are zero for components that do not undergo ion exchange, adsorption, or precipitation/ dissolution. The coefficient D is the sum of the diffusion and dispersion components: D  Dm  

q 

(28-26)

where  is the tortuosity factor, Dm is the coefficient of molecular diffusion [cm2 d1], and  is the dispersivity [cm]. This equation is a simplified treatment of the diffusion process. A more detailed description of the diffusion process requires calculation of the diffusion rates of individual species requiring coupling of individual ion fluxes to the concentration gradients of all individual species. This simplification appears justified since, in soils, errors generated by uncertainty in determination of the tortuosity factor and velocity vectors are more significant for solute transport than errors associated with this treatment of diffusion. Chemical Model The chemical model includes consideration of nine major aqueous components, consisting of Ca, Mg, Na, K, SO4, Cl, alkalinity, NO3 and B. Alkalinity is defined as Alkalinity  [HCO3]  2[CO32]  2[CaCO03]  [CaHCO3]  2[MgCO03 ]  [MgHCO3]  2[NaCO 3] [NaHCO03 ]  [B(OH)4 ]  [H]  [OH]

(28-27)

where brackets represent concentrations (mol kg1). From these components we obtain 11 primary species: Ca2, Mg2, Na, K, SO42, Cl, 2   HCO 3 , CO3 , NO3 , B(OH)4 and H3BO3. In addition, we include the ion 0 0  0 0 pair/complexes CaHCO , 3 CaCO3 , CaSO4 , MgHCO3 , MgCO3 , MgSO4 ,   NaHCO30, NaCO , NaSO , and KSO . Alkalinity as defined in Eq. 28-27 is 3 4 4 a conservative quantity, affected only by dissolution or precipitation of a carbonate phase (such as calcite). After calculation of the soil air phase CO2 partial pressure (from the production and transport routines or from the user specified input),

870

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

H2CO3* (sum of aqueous CO2 and H2CO3) is calculated using a Henry’s Law expression: kCO2 

(H 2 CO *3 ) PCO2 (H 2 O)

(28-28)

where PCO2 is the partial pressure of CO2 (atm.), and parentheses denote activities. From the H2CO*3 value and utilizing the equations for the first and second dissociation constants of carbonic acid, conservation of mass, and the equations for dissociation of the complexes, we solve the equations using an iterative approach. The soil solution pH is determined as a dependent variable from solution of Eq. 28-27 for [H] and the activity coefficient. All equilibrium constants are calculated from available temperature-dependent expressions. Soil temperature is calculated from a heat flow submodel, with input of air temperature data and the initial soil temperature profile. Osmotic pressure The osmotic pressure is used to calculate the impact of salinity on water uptake and plant yield. Osmotic pressure is calculated from P  RT

Ms vm Vs m0

(28-29)

where P (Pa) is the osmotic pressure of the solution, R is the gas constant, T is absolute temperature, Vs is the partial molar volume of the water, m0 is unit molality, m is molality of the solution,  is the osmotic coefficient of the solution, and M is the molar weight (Stokes 1979) . The osmotic pressure in Pa is converted to osmotic pressure in m by the expression h 

P g

(28-30)

The osmotic coefficient is calculated from Pitzer (1973). Detail is provided in the user manual. Activity coefficients Activity coefficients in the SWS model are determined by default using an extended version of the Debye-Huckel equation (Truesdell and Jones 1974): ln  

Az 2 I  bI 1  Ba I

(28-31)

MODELING TRANSIENT ROOTZONE SALINITY

871

where A and B are constants depending on the dielectric constant of water, density, and temperature; z is the ionic charge; a and b are adjustable ion parameters; and I is the ionic strength. At higher ionic strength (0.3 M), the solution is sufficiently concentrated that all ion interactions must be considered for calculation of activity coefficients. In this instance, the Pitzer expressions are utilized. The activity coefficients are expressed in a virial-type expansion having the form (Pitzer 1979) ln i  ln iDH  ∑ Bij ( I )m j  ∑ ∑ Cijk m j mk  ... j

j

(28-32)

k

where iDH is a modified Debye-Huckel activity coefficient and Bij and Cij are coefficients specific to each ion interaction. The Pitzer approach considers ion–ion interactions for every species in solution; thus, it does not consider the individual ion pairs and complexes, such as NaSO 4 described above as a species when using the extended Debye-Huckel equation and ion association model. The Pitzer model is considered suitable for prediction of species activity in solutions up to 20 mol kg1, a concentration well above the intended use of the SWS model. Solid phases Because the model considers a restricted set of solid phases, it cannot be used to predict the composition of a brine undergoing evaporation. The minerals considered include calcite, gypsum, hydromagnesite, nesquehonite, and sepiolite. Since the model attempts to predict water composition, it cannot be based only on thermodynamic considerations. Dolomite precipitation is not considered by the present model because ordered dolomite has not been observed at near-earth surface conditions. The kinetics of dissolution are also sufficiently slow that it is not reasonable to assume that a solution is dolomite-saturated merely because dolomite is present in the soil profile. It is beyond the scope of this model to consider the detail necessary for a kinetic description of dolomite dissolution. This omission is significant only if dolomite is present and calcite is not. The model tracks all changes in the quantities of the solid phases due to precipitation or dissolution. Calcite precipitation The equilibrium condition of a solution saturated with calcite in the presence of CO2 can be described by the expression (Ca2 )(HCO3 )2 

C KSP KCO2 K a1 C KT PCO2 (H 2 O)  KSP K a2

(28-33)

872

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

where parentheses denote activities, KCO2 is the Henry’s Law constant for the solubility of CO2 in water, Ka1 and Ka2 are the first and second dissociC ation constants of carbonic acid in water, and KSP is the solubility product for calcite. To obtain equilibrium [i.e., when the ion activity product (IAP) is equal to the solubility product Ksp], a quantity x of Ca2 and HCO 3 must be added or removed from the solution (right-hand side of Eq. 28-33) to satisfy the equilibrium condition. It has been shown that waters below irrigated regions are supersaturated with respect to calcite, and the average IAP can be represented by 108.0 (Suarez 1977; Suarez et al. 1992)—a value about three times greater than that predicted by calcite equilibrium. Thus, the equilibrium condition significantly underestimates the Ca solubility in soil solutions. The cause of supersaturation has been shown to be due to poisoning of crystal surfaces by dissolved organic matter (Inskeep and Bloom 1986; Lebron and Suarez 1996). For the purposes of the SWS model, the nucleation rate is sufficiently fast, and the crystal growth rate sufficiently slow, that the calcite solubility can be taken at the point of supersaturation at which there is no further nucleation. This level of supersaturation is significantly close to the supersaturation with respect to calcite observed in field measurements. The model thus uses the apparent solubility of 1.0  108, with the temperature dependence determined for calcite. This is not an equilibrium value, but it is the suitable value to simulate calcium carbonate solubility in the soil zone. Gypsum The model allows the user to specify the initial presence of gypsum, requiring input of the quantity present. If gypsum is present in any soil layer at the given time step, the model forces the solution to gypsum equilibrium. The program tracks changes in the amount of gypsum present if all gypsum is dissolved in a soil layer, such as during reclamation of a sodic soil, in which case gypsum equilibrium is no longer forced. In all cases, gypsum precipitates wherever supersaturation is indicated by solution calculations. For the objectives of this model, it is reasonable to assume that the kinetics of gypsum dissolution/precipitation are sufficiently fast that the equilibrium condition can be used. Magnesium precipitation The model considers that Mg precipitation can occur as a carbonate (either nesquehonite or hydromagnesite), or as a silicate (sepiolite). Since this is a predictive model, it considers only phases that either precipitate under earth surface conditions or occur frequently and are reactive under earth surface conditions; these need not necessarily be the thermodynam-

MODELING TRANSIENT ROOTZONE SALINITY

873

ically most stable. With this consideration, magnesite is neglected because it apparently does not form under earth surface temperatures, is relatively rare, and its dissolution rate is exceedingly small, such that its solubility has not yet been satisfactorily determined from dissolution studies at or near 25 °C. Similarly, dolomite precipitation is not considered, because true dolomite appears to form very rarely in soil environments. If nesquehonite or hydromagnesite saturation is reached, the model will precipitate the predicted Mg carbonate. The Mg carbonate precipitated, combined with calcite precipitation, will likely represent the mixed Ca-Mg precipitate that is observed in hypersaline environments, often called protodolomite (sometimes incorrectly called dolomite). However, the resulting solution composition is much different from that produced by simply forcing equilibrium with respect to dolomite, as the model forms this mixed precipitate (calcite  hydrated magnesium carbonate) under conditions of high supersaturation with respect to dolomite. This result is consistent with the high levels of dolomite supersaturation maintained in high-Mg waters. Precipitation (or dissolution, if present in the soil) of sepiolite is also considered by the model. Sepiolite will readily preS cipitate into a solid with a KSP greater than that of well-crystallized sepiolite. Formation of this mineral requires high pH, high Mg concentrations, and low CO2 partial pressure. Cation Exchange Cation exchange is generally the dominant chemical process for the major cations in solution in the unsaturated zone. Generally, cation exchange is treated with a Gapon-type expression of the form (White and Zelazny 1986) K ij 

x 1 x ciy (c j ) c jx (ciy )1 y

(28-34)

where y and x are the valences of species i and j, respectively, and the overscored concentrations are those of the exchange phase (expressed in molc mass1). It is assumed that the cation exchange capacity cT is constant, and for nonacid soils 2

cT  Ca  Mg

2



 Na  K



(28-35)

Experimentally determined selectivity values for a given cation pair are not constant over the full range of composition. In addition, the cation exchange capacity (CEC) varies as a function of pH due to variable charge materials such as organic matter. It has been observed that soils have an

874

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

increased preference for Ca2 over Na, and Ca2 over Mg, at low levels of exchange phase Ca2. Suarez and Wood (1993) developed a mixing model used in UNSATCHEM (Suarez and Sˇ imu˚ nek 1997) that is able to approximate the nonconstant values of the soil selectivity coefficient by taking into account the organic matter content of the soil and using the published constant selectivity values for clay and organic matter. Calcium preference decreases as the organic matter exchange sites (which have higher Ca preference than clays) become Ca saturated. This approach is useful to predict the input exchange constants if the selectivity values of the soil are not known and if substantial organic matter is present. This option is not directly available in the SWS model interface but can be applied by the user via the input file using a standard file editor. The model has default exchange selectivity values, but the user should specify soil-specific values if available.

Boron The major chemical process affecting B concentration and transport in soils is adsorption. Various adsorption models are available, but the constant capacitance model (CCM; Herbelin and Westall 1996) has been demonstrated to well-represent B adsorption with soils (Goldberg and Glaubig 1986; Goldberg et al. 2000). Application of chemical complexation models into transport models has been hampered primarily by the need to have the soil-specific characterization and model parameters, requiring time-consuming laboratory studies. When fitting the CCM model to experimental data, Goldberg et al. (2000) found that a good fit to the CCM model was obtained by selecting the surface species as SH3BO 4 ; the surface reaction was written as SOH  H 3 BO3 ↔ SH 3 BO4  H 

(28-36)

The intrinsic equilibrium constants are given as

K 

[SOH 2 ] exp( F  RT ) [SOH ][ H  ]

(28-37)

K 

[SO ][ H  ] exp(F  RT ) [SOH ]

(28-38)

K B 

[SH 3 BO4 ][ H  ] exp(F  RT ) [SOH ][ H 3 BO3 ]

(28-39)

MODELING TRANSIENT ROOTZONE SALINITY

875

where F is the Faraday constant (C molcL1), is the surface potential (V), R is the molar gas constant, T is the absolute temperature (K), and brackets indicate concentrations in mol L1 (Goldberg et al. 2000). Goldberg et al. (2000) developed regression equations for the prediction of the three CCM surface complexation constants based on generally available soil properties. The following equations were developed: LogKB  9.14  0.375ln(SA)  0.167ln(OC)  0.11ln(IOC)  0.466ln(Al)

(28-40)

LogK  7.85  0.102ln(OC)  0.198ln(IOC)  0.622ln(Al)

(28-41)

LogK  11.97  0.302ln(OC)  0.0584ln(IOC)  0.302ln(Al)

(28-42)

where SA is the surface area, measured by ethylene glycol monoethyl ether (EGME); OC is the organic carbon; IOC is the inorganic carbon; and Al is the extractable Al (including absorbed and reactive hydroxides and oxides). Using these relationships, Goldberg et al. (2000) predicted the absorption envelopes (adsorption as a function of pH) for a series of arid land soils. They concluded that the fits using the CCM with the constants determined from the stated predictive equations were acceptable for use in modeling adsorption. These constants have been added to the SWS model along with a routine to solve the CCM equations, enabling soilspecific prediction of B adsorption and transport as related to soil properties and solution pH.

SWS APPLICATIONS This section describes several published applications and utilizations of the model or incorporated routines. The capabilities to predict changes in water content, CO2 concentration, and leaching of salts and sodium during reclamation are demonstrated. Also presented are several examples of model simulations useful for water managers and engineers. Prediction of Variable Water Content and CO2 Suarez and Sˇimu˚nek (1993) utilized field data published by Buyanovsky and Wagner (1983). This field data set is relatively unusual in that it presents detailed information about the extent and timing of rain events, average air temperature, and the changes in water content and CO2 throughout a period of almost 1 year. The simulation used the described soil

876

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

properties, including texture, organic matter, and air porosity at field capacity, but the available data set did not contain detailed hydraulic information. The data set input to the model was thus limited to the soil properties, rooting depth, rain events and quantities, average air temperature, and ET0 estimates. These inputs represent a level of information that would be realistic for management decisions. As shown in Fig. 28-1, there was a good correspondence between the measured water content at 0.2 m and the model-simulated water content. Similarly, Fig. 28-2 shows that the model was well able to predict the field CO2 concentrations (Suarez and Sˇ imu˚ nek 1993), over a range of conditions, including those where transport (under winter conditions and low water content) and CO2 production (warm conditions) predictions could be evaluated. Saline Sodic Soil Reclamation: Model Versus Field Data Suarez (2001) examined the reclamation of a saline sodic field and compared the field results to predictions using UNSATCHEM 3.1 (Suarez and Sˇ imu˚ nek 1997). The 40-ha field was mapped for salinity using an electromagnetic (Geonics EM-38) unit. Soil samples (24 cores sampled in 30-cm intervals) were also collected before and after reclamation based on the EM map to capture the field variability. Shown in Figs. 28-3 and 28-4 [after Suarez (2001)] are the initial and final median EC and SAR values as a function of depth. A total of 114 cm of water was applied to the field

FIGURE 28-1. Measured (Buyanovsky and Wagner 1983) and calculated water contents at a depth of 0.20 m for a Missouri wheat experiment, 1982. From Suarez and Sˇ imu˚ nek (1993).

MODELING TRANSIENT ROOTZONE SALINITY

877

FIGURE 28-2. Measured (Buyanovsky and Wagner 1983) and calculated CO2 concentrations at a depth of 0.20 m for a Missouri wheat experiment, 1982. Vertical bars show standard deviations. From Suarez and Sˇ imu˚ nek (1993).

FIGURE 28-3. Median EC values with depth for both initial and final (after leaching) conditions. Reclamation consisted of application of 24 Mg/ha of gypsum and application of 114 cm of water. The dashed lines indicate the 95% confidence limits of the median for the field. From Suarez (2001).

878

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 28-4. Median SAR values with depth for both initial and final (after leaching) conditions. Reclamation consisted of application of 24 Mg/ha of gypsum and application of 114 cm of water. The dashed lines indicate the 95% confidence limits of the median for the field. From Suarez (2001).

(ponded) along with incorporation of 24 Mg/ha of gypsum to a depth of 15 cm. Based on climatic data for the time interval during the ponding event, it was calculated that only 74 cm infiltrated, with 40 cm of surface evaporation. The model simulation used only the initial soil profiles, water applied (volume and EC), initial soil saturation extract EC and SAR, soil texture (used to estimate hydraulic parameters), estimated soil CEC (based on mineralogy and texture), and quantity and depth of gypsum applied. As shown in Figs. 28-5 and 28-6, both the EC profile and the change in SAR were satisfactorily predicted with the correct amount of water. Note that the model was not “calibrated” by adjusting parameters, nor was the input modified, demonstrating that a deterministic transient model can give useful results for a field application without excessive input data. Optimizing Reclamation Using SWS When reclaiming a sodic field, many options are available. Although a model prediction cannot be used alone to select the best option, it can be used as part of the decision-making process. The following examples are from Suarez (2001). Deeper placement of gypsum increases costs of sodic

MODELING TRANSIENT ROOTZONE SALINITY

879

FIGURE 28-5. Comparison of measured and model predicted changes in EC with depth after mixing 24 Mg/ha of gypsum into the top 15 cm and then infiltration of 70 and 80 cm of water. From Suarez (2001).

FIGURE 28-6. Comparison of measured and model predicted changes in SAR with depth after mixing 24 Mg/ha of gypsum into the top 15 cm and then infiltration of 70 and 80 cm of water. From Suarez (2001).

880

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

soil reclamation but may result in use of less water. Shown in Fig. 28-7 is a comparison of the relative effectiveness of reclamation as related to depth of gypsum incorporation (equal quantities of gypsum). As shown, shallow placement of gypsum (top 2 cm) will not reclaim this field adequately with 80 cm of water. Additional water (and thus time) is required to reclaim to the same extent as predicted for deeper placement. This is due to the fact that gypsum solubility is enhanced in the presence of high exchangeable Na, so that the gypsum dissolves with less water if it is incorporated deeper in the profile. The optimal placement depth of gypsum depends on several factors. Deep placement of gypsum is more expensive and will enable reclamation to a greater depth, but, depending on the quantity used, it may not adequately reclaim the important top 15 cm. Based on Fig. 28-7, it can be concluded that for this site and the amount of gypsum and water used, 8- to 15-cm placement of the gypsum is sufficiently deep. The optimal depth of placement and quantity of gypsum to apply will thus depend on the depth needed to be reclaimed, initial exchangeable sodium percentage, CEC, cost and availability of water, and cost of gypsum incorporation The SWS model is suited to predict various scenarios and enable the user to decide the optimal practice for the specific site. For example, gypsum placement on the surface or in the water may be less expensive than

FIGURE 28-7. Comparison of model-predicted changes in SAR with depth after mixing 24 Mg/ha gypsum into the top 2, 8, 15, and 30 cm of soil and leaching with 80 cm of water. From Suarez (2001).

MODELING TRANSIENT ROOTZONE SALINITY

881

incorporating into the soil, but it will require more time and larger quantities of water. An alternative or complement to gypsum reclamation for calcareous soils is enhancement of the CO2 concentration in the soil air and reclamation by dissolution of calcite. The concept of green manuring as a sodic reclamation practice has been discussed and applied with mixed results. Shown in Fig. 28-8 are simulations with a CO2 partial pressure of 5 kPa, comparable to what could be achieved by incorporating organic matter and flooding the soil under warm soil conditions. Although less effective than gypsum, use of the calcite in the soil can nonetheless reduce the quantity of gypsum and can sometimes avoid the entire cost of gypsum and its application. In a calcareous soil, calcite dissolution and its reclamation effect should be considered when determining gypsum requirements. The disadvantage of using the soil calcite alone for sodic soil reclamation, as compared with gypsum, is that it requires more water to achieve the same final SAR in the soil. This is shown in Fig. 28-8 compared to Fig. 28-7. As shown in Fig. 28-9 (Suarez 2001), up to 32 mmolc/L of alkalinity may be released when reclaiming a sodic soil using calcite and green manuring. This indicates that calcite, in combination with cation exchange, can release an amount of Ca comparable to that released from gypsum

FIGURE 28-8. Model-predicted changes in SAR with depth after elevating the CO2 to 5 kPa in the presence of calcite, then leaching with 20, 50, 80, and 114 cm of water. From Suarez (2001).

882

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 28-9. Model-predicted alkalinity concentrations (mmolc/L) with depth after elevating the CO2 to 5 kPa in the presence of calcite, and then leaching with 20, 50, 80, and 114 cm of water. From Suarez (2001). placement on the surface. Under cold soil conditions, the model predicts that the CO2 production is reduced and very large quantities of water are required. The model can also be used to examine the interaction of EC, SAR, and hydraulic relationships to ensure that the EC does not drop rapidly and cause soil dispersion before the SAR is reduced to a safe level. Based on the simulation, green manuring will fail when insufficient calcite is present, the initial EC is not sufficiently high for the initial SAR to prevent dispersion or swelling, or CO2 production is inadequate (cold conditions). The hazard to water supplies receiving drainage water from a sodic soil reclamation project must also be considered. As discussed previously, up to 32 mmolc/L of alkalinity may be released when reclaiming a calcareous sodic soil using green manuring. In this instance the drainage water would be of very high pH (9.0) once it degasses, with high alkalinity and low Ca concentration. Reclamation with gypsum will also increase the salt load of discharging waters, in this instance primarily sodium sulfate. Effect of Rain on Sodium Adsorption Ratio Shown in Fig. 28-10 are the soil EC profiles after 1 to 5 cm of rain infiltrated into a loam soil (Suarez et al. 2006). These simulations are for cal-

MODELING TRANSIENT ROOTZONE SALINITY

883

FIGURE 28-10. Predicted relationship of EC with depth and quantity of rain infiltrated for Glendive loam soil. The initial condition was EC  1.0 dS/m and SAR  10. Each curve represents addition of 1 cm of rain. From Suarez et al. (2006). careous soils first irrigated with water of EC  1.0 dS/m and SAR  10. The initial soil EC is higher than the input irrigation water EC and it increases with depth due to predicted calcite dissolution in the soil. The simulation input included the measured soil CEC. As shown, the predicted EC at the surface decreased during the rain event, decreasing to 0.42 dS m1 at the surface after infiltration of 5 cm of rain. Calcite dissolution during the rain event is enhanced by the exchange of solution Ca for Na on the clay exchange sites, thus causing a reduction in the SAR with time, as shown in Fig. 28-11. The SAR was still equal to 5.5 at the surface after 5 cm of rain. The decrease in SAR was not sufficient to compensate for the decrease in EC; thus, the simulation shows that the sodium hazard is increased during the rain event. A surface treatment (such as gypsum)

884

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 28-11. Predicted relationship of SAR with depth and quantity of rain infiltrated. The initial condition was EC  1.0 dS/m and SAR  10. Each curve represents addition of 1 cm of rain. From Suarez et al. (2006).

is likely needed if there is substantial rain on this soil when it is also irrigated with water of SAR  10 and EC  1 dS m1. The predicted change in EC and SAR for rain on a clay soil irrigated with the same water is shown in Figs. 28-12 and 28-13, respectively (Suarez et al. 2006). The decrease in EC at the surface is similar to but slightly less than that observed for the loam soil (Fig. 28-10). This difference is caused by the increased dissolution of calcite due, in turn, to the increased cation exchange of the clay soil. As shown in Fig. 28-13, the SAR of the clay soil was buffered, and there was only a small SAR reduction after infiltration of 5 cm of rain. The high CEC of the clay soil allows the soil exchange sites to buffer the solution SAR. The soil surface of the clay soil at the end of the

MODELING TRANSIENT ROOTZONE SALINITY

885

FIGURE 28-12. Predicted relationship of EC with depth and quantity of rain infiltrated into the clay soil. The initial condition was EC  1.0 dS/m and SAR  10. Each curve represents addition of 1 cm of rain. From Suarez et al. (2006).

rain event is thus at low EC with almost no decrease in SAR relative to the pre-rain condition. These simulations suggest that the chemical effects related to the infiltration hazard of rain or irrigation waters of low salinity on a sodic soil would be greater for soils of greater CEC. The model provides the temporal changes during the irrigation season and allows for simulation of different applications and timing of surface amendments. Management of High-Boron Waters Used for Irrigation Waters with B concentrations above 1 mg L1 can be potentially toxic to B-sensitive crops, and almost all crops are adversely affected when

886

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 28-13. Relationship of SAR with depth and quantity of rain infiltrated into clay soil. The initial condition was SAR  10. Each curve represents addition of 1 cm of rain. From Suarez et al. (2006).

concentrations in the soil water exceed 10 mg L1. The SWS simulations below demonstrate how the model can be used as a management tool when using low-quality waters for irrigation. Irrigation drainage water from Westside of the Central Valley in California typically has a B concentration of 4 to 8 mg L1 and an EC of 8 to 14 dS m1. These waters are typically considered unusable for irrigation, or, if usable, then only with salt- and B-tolerant crops. With the traditional steady-state approach (in this case no adsorption-desorption), the B concentration is increased in the soil proportionally with the Cl concentration. Therefore, the traditional

MODELING TRANSIENT ROOTZONE SALINITY

887

recommendation when irrigating with high-B water is to increase the leaching fraction to maintain a lower B concentration in the rootzone. The subsequent model simulations examine two leaching regimes when applying a high-B concentration irrigation water on an initially low-Bcontaining soil. The irrigation water used in the simulations contained a B concentration of 0.8 mmol L1 (10 mg L1), considered to be unusable by all criteria. Although low-quality water is often not usable for sustained agricultural production, it can be utilized either by blending or cyclic use of low- and high-quality water. A steady-state model can adequately consider this by using the average B concentration. However, in a drought condition no high-quality water may be available, or it may be available only during part of the season. The soil profile was initially free of B. In the simulation shown in Fig. 28-14 (Goldberg and Suarez 2006), the ET was 1 cm/day with irrigation applications corresponding to an average input of 2 cm/day (leaching fraction of 0.5). A total of 200 cm of water was applied during the 100-day growing season. The surface area of 100 m2 g1 soil corresponded to a soil with relatively low B adsorption capacity, while that at 1,000 m2 g1 was a soil with high adsorption capacity. The irrigation of the low-adsorption-capacity soil caused the B concentration to rapidly increase to toxic levels (with concentrations approaching the steadystate values), while the higher-adsorbing soil was able to maintain the B in solution below that of the irrigation water. The results of irrigating the same soils with the same waters at a leaching fraction of 0.1 are shown in Fig. 28-15 (Goldberg and Suarez 2006). The low-adsorbing soil had B concentrations increasing to higher levels with low leaching as compared to high leaching. The high-adsorbing soil saw relatively low concentrations of B, even after 100 days; at this time the B front is just reaching 25 cm. The mean rootzone B concentrations as a function of time are shown in Fig. 28-16 for the four simulations (Goldberg and Suarez 2006). The conclusions are that for a soil with high B adsorption capacity, there is little B hazard during the first year of cropping, and it is best to use low leaching fractions to minimize the B concentration and accumulation. At steady state, the lower leaching would eventually result in proportionally higher B concentrations than the more leached soil. For the low-B-adsorbing soil, the recommendation would be to use low leaching for 70 days and then switch to high leaching to prevent further B accumulation. These recommendations are opposite to recommendations based on the steady-state analysis. The mean B concentration in the rootzone is sufficiently low that many crops could be grown without yield loss. Sustained management would clearly require a better water source, either winter rains or leaching with higher-quality water, in the subsequent crop cycle.

888

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 28-14. Change in boron concentration with depth and time (0, 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 days) for leaching fraction of 0.5 and soil surface area of (a) 100 m2g1, and (b) 1,000 m2g1. From Goldberg and Suarez (2006).

MODELING TRANSIENT ROOTZONE SALINITY

889

Figure 28-15. Change in boron concentration with depth and time (0, 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 days) for leaching fraction of 0.1 and soil surface area of (a) 100 m2g1, and (b) 1,000 m2g1. From Goldberg and Suarez (2006).

890

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Figure 28-16. Mean root zone soil solution B concentration with time as related to leaching fraction (LF) and soil surface area (SA), expressed in 103  m2g1. From Goldberg and Suarez (2006). Plant Response to Salinity and Water Stress As discussed, the SWS model predicts plant response to water and salt stress under dynamic conditions. The model uses the predicted decreases in plant water uptake to predict the decrease in biomass production. This calculation assumes that yield is directly proportional to water consumption (constant WUE, or water use efficiency). Improved predictions of yield loss can be obtained if the user has crop-specific information on the change in WUE as related to crop water consumption. Prediction of the yield of individual plant parts (such as seed or fruit) can be obtained by consideration of the relation of reduction in plant water uptake and yield response of the plant part of interest. The following example, taken in part from Suarez (2011), provides model predictions based on water stress, salt stress, and combined water and salt stress compared to steady-state predictions. Scenario 1: No stress In an initial simulation we examined crop production in the absence of matric or osmotic stress. A perennial crop with a 100-cm rootzone depth on a loam soil (ks  25 cm/day) was irrigated for 200 days. The first irrigation of 11 cm was applied after 10 days. After another 10 days, 22 cm of

MODELING TRANSIENT ROOTZONE SALINITY

891

water was applied over 2 days, followed by irrigations of 22 cm every 20 days thereafter, for a total of 209 cm of applied irrigation water. The potential ET of the crop for full yield was 200 cm and, for simplicity, we assumed a constant potential ET value of 1 cm/day. The crop was irrigated with a water of the following composition: Ca  2 mmolc L1, Mg  1 mmolc L1, Na  10 mmolc L1, Cl  10.4 mmolc L1, SO4  1.4 mmolc L1, and alkalinity  0.2 mmolc L1. The soil was initially at a moderate water content, with an input matric pressure of 100 cm at the surface and 85 cm at the 100-cm soil depth (the volumetric water content was calculated to be 0.244 at the surface and 0.257 at 100-cm depth). The h50 value for Eq. 28-14 was set at 150 m and p  3. The model predicted 100% yield, with a water consumption of 200 cm, producing a calculated LF  0.043. The steady-state calculation of salt stress would also predict no yield loss at this salinity level. When the salt stress value, h50 value was set to 50 m, we predicted a 99% relative yield. Scenario 2: Matric stress In this case we predict yield response to water stress. We irrigated a loamy sand soil (ks  356 cm/day) with the same root distribution, total water application, frequency daily ET, and water composition as in Scenario 1. Here we used an h50 value of 50 cm and p  3 for the water stress response function (Eq. 28-13). The initial matric pressure was again 100 cm at the surface and 85 cm at the 100-cm depth (the volumetric water content was calculated to be 0.072 at the surface and 0.075 at the 100-cm depth). At the end of 200 days the model predicted a relative yield of 57%. The reduced yield is attributed to matric stress between irrigations on this loamy sand. The SWS model also predicted an LF value of 0.56, as compared to 0.043 with more frequent irrigation into the loam soil (Scenario 1). A steady-state model cannot predict the matric stress resulting from inadequate irrigation frequency and depending on soil properties, ET, and quantities of water applied. Thus, steady-state calculations cannot predict salt distribution, or the increased LF and drainage volume that results in these instances. Scenario 3: Salt stress We used the SWS model to predict plant yield reduction from salt stress. We utilized all the same conditions as in Scenario 1, with the exception of the water composition. We used the loam soil properties of Scenario 1. The initial soil water and irrigation water composition was as follows: Ca  5 mmolc L1, Mg  5 mmolc L1, Na  50 mmolc L1, Cl  52 mmolc L1, SO4  7.0 mmolc L1, and alkalinity  1.0 mmolc L1. The h50 for osmotic stress was set at 50 m. We set the h50 for matric stress to 50 cm to ensure that there was no matric stress. The predicted relative

892

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

yield was 62% and the predicted LF was 0.42. The increased leaching and decreased water uptake was due entirely to salt (osmotic stress). The same scenario can be evaluated using the steady-state calculation as recommended by Ayers and Westcot (1985). In this calculation we consider the crop requirement of 200 cm of water and the applied water quantity of 209 cm. The EC of the irrigation water is estimated from the concentration and Eq. 3-11 (see Chapter 3 of this manual), yielding EC  6.20 dS/m. The average rootzone salinity is calculated from the average salinity of the rootzone, using the irrigation water salinity and the salinity at the bottom of each of the four quarters of the rootzone. The salinity is calculated for each quarter as EC  ECiw/LFq, where LFq is the leaching fraction at the bottom of that quarter of the rootzone, using the assumption that water uptake is 40% in the first quarter, 30% in the second quarter, 20% in the third quarter, and 10% in the fourth quarter. In this case the salinity at the top is 6.2 dS/m and the salinity of the first through fourth quarters is 10.05, 18.8, 44.7, and 144.2 dS/m, respectively. The average rootzone salinity is calculated as 44.8 dS/m. Using Eq. 3-20 (see Chapter 3), this corresponds to calculated mean rootzone osmotic pressure of 179 m. Using the salt response of the crop utilized for these scenarios (h50  50 m and p  3), the mean osmotic pressure of 179 m, and applying Eq. 28-14, we calculate an (h) value of 0.02. The predicted relative yield is thus 2% using the Ayers and Westcot (1985) calculation method. A similar result would be obtained using the steady-state WATSUIT calculation. The major discrepancy between these calculations and the SWS predictions is the failure of these “traditional” calculations to predict the reduction in water consumption by the crop and, thus, the rootzone salinity and LF. Note that the LF was assumed to be 0.043 based on applied water and crop water demands (ET); however, the SWS model predicts reduced water uptake and an LF  0.42. The differences between the model predictions (less stress) and the simple calculation method are even greater when we consider waters that precipitate gypsum in the soil, thus reducing the salt concentrations in the soil. Scenario 4: Water and salt stress In this scenario we again use the SWS model with the same conditions for water quantities, irrigation, potential ET, etc. The irrigation water composition was the same as used for Scenario 3 (salt stress only), with the difference being that this time we irrigated the loamy sand soil from Scenario 2 (matric stress only). If we were to simply combine the stresses by multiplying the independently calculated stress response functions, we would predict a relative yield of 35% (57%  62%). The model prediction is 46% relative yield, accounting for the interaction of the stresses. In this case, the reduced water uptake by the salt stress reduced the soil

MODELING TRANSIENT ROOTZONE SALINITY

893

pressure head, thus reducing the matric stress. The reduced matric stress also allowed the salt concentrations and thus the salinity stress to be lower than would be expected by considering them separately. The hazard with a static analysis and overly simplifying assumptions is also illustrated by the following calculations. A prediction of the outcome of Scenario 4 could be obtained by combining experimental observations under matric stress and the steady-state analysis of salt response. Adding the responses to both osmotic stress and matric stress together, without consideration of their interaction, would result in a predicted yield of only 19%. Alternatively, if we were to add the water uptake averaged water matric and osmotic stress for Scenarios 2 and 3, we obtain an overall pressure of 138.5 m (70.0 m from matric alone and 68.5 m from osmotic alone). Using Eq. 28-14 and combining the mean salt and matric pressure with an h50 of 50 m would give a predicted relative yield of only 4.5%. As discussed in Scenario 2, the static calculation without correction for reduced water uptake would predict a 2% yield based on salt stress alone. Although the preceding example is somewhat extreme in terms of the close correspondence between water application and crop water demand (209 cm vs. 200 cm), such irrigation efficiency is not unusual for new irrigation technologies, such as drip irrigation. It appears that dynamic modeling is necessary for irrigation management when low target LFs are the objective under conditions of potential yield loss due to salinity.

SUMMARY The SWS model is a variation on the UNSATCHEM model (Suarez and Sˇ imu˚ nek 1992, 1997) with addition of calculations for ET, a new B adsorption routine, and with a user-friendly interface that makes extensive use of default parameters to minimize the need for user expertise in soil physics and chemistry. The SWS model accounts for a number of functions and processes known to influence practical aspects of irrigation and drainage management, including: • Hydraulic functions • Chemical effects on HC • Water uptake by plant roots, including an optional root growth function • Calculations of crop ET • Factors controlling soil CO2 concentrations (production and transport) • Soil-water chemistry (transport, osmotic pressure, chemical activity, calcite precipitation, gypsum content, magnesium precipitation, cation exchange, boron).

894

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Based on comparisons of the model simulations and reliable, readily available field data, the SWS model predicts actual field conditions with reasonable accuracy: • Soil water content and soil CO2 • EC and SAR • Reclamation effects of various alternatives amendments (such as gypsum, organic matter, and/or acid dissolution of calcite, alone or in combination), including secondary effects such as increases in pH or salt loading • Effects of rain on soil SAR • Effects management protocols for boron in soils with variable leaching fractions • Plant response to various salinity and water stress combinations. The comparison of field results to results of SWS simulations suggests that transient models like the SWS model may provide more accurate and precise predictions of actual field soil and water conditions as a result of irrigation and other soil management inputs than are feasible with more traditional steady-state prediction models. This greater predictive ability of such models is obtained with input data that are often readily available from routine on-farm measurements. This type of dynamic modeling may be necessary for irrigation management when low target LFs are the objective under conditions of potential yield loss due to salinity.

REFERENCES Addiscott, T. M., and Wagenet, R. J. (1985). “Concepts of solute leaching in soils: A review of modeling approaches.” J. Soil Sci., 85, 411–424. Allen, R. G., Peieira, L. S., Raes, D., and Smith, M. (1998). Crop evapotranspiration, FAO Irrigation and Drainage Paper 56, Food and Agriculture Organisation of the United Nations, Rome. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1 Food and Agriculture Organisation of the United Nations, Rome. Bresler, E. (1973). “Simultaneous transport of solutes and water under transient unsaturated flow conditions.” Water Resour. Res., 9, 975–986. Buyanovsky, G. A., and Wagner, G. H. (1983). “Annual cycles of carbon dioxide level in soil air.” Soil Sci. Soc. Am. J., 47, 1139–1143. Carsel, R. F., and Parrish, R. S. (1988). “Developing joint probability distributions of soil water retention characteristics.” Water Resour. Res., 24, 755–769. Dutt, G. R. (1962). “Prediction of the concentration of solutes in soil solutions for systems containing gypsum and exchangeable Ca and Mg.” Soil Sci. Soc. Am. Proc., 26, 341–343.

MODELING TRANSIENT ROOTZONE SALINITY

895

Dutt, G. R., Shaffer, M. J., and Moore, W. J. (1972). Computer simulation model for dynamic bio-physical processes in soils, University of Arizona Agricultural Experiment Station Bulletin, Tucson, Ariz. Felmy, A. R. (1990). GMIN: A computerized chemical equilibrium model using a constrained minimization of the Gibbs free energy, Pacific Northwest Lab, Richland, Wash. Goldberg, S. R., and Glaubig, R. A. (1986). “Boron adsorption on California soils.” Soil Sci. Soc. Am. J., 50, 1173–1176. Goldberg, S. R., Lesch, S. M., and Suarez, D. L. (2000). “Predicting boron adsorption by soils using the soil chemical parameters in the constant capacitance model.” Soil Sci. Soc. Am. J., 64, 1356–1363. Goldberg, S., and Suarez, D. L. (2006). “Prediction of anion adsorption and transport in soil systems using the constant capacitance model,” in Surface complexation modeling, J. Luetzenkirchen, ed., Interface Science and Technology Series, Elsevier, Amsterdam, 11, 491–517. Grieve, C. M., Maas, G., and Grattan, S. (2011). “Plant salt tolerance,” in Agricultural salinity assessment and management, Chapter 13, this volume. Hanks, R. J., and Bowers, S. A. (1962). “Numerical solution of the moisture flow equation for infiltration into layered soils.” Soil Sci. Soc. Am. Proc., 26, 530–535. Herbelin, A. L., and Westall, J. C. (1996). FITEQL: A computer program for determination of the chemical equilibrium constants from experimental data, Report 96-01, Ver. 3.2, Dept. of Chemistry, Oregon State University, Corvallis, Ore. Inskeep, W. P., and Bloom, P. R. (1986). “Kinetics of calcite precipitation in the presence of water soluble organic ligands.” Soil Sci. Soc. Am. J., 50, 1167–1172. Labadie, J. W., and Khan, I. A. (1979). “River basin salinity management via the ASTRAN method, I: Model development.” J. Hydrol., 42, 301–321. Lebron, I., and Suarez, D. L. (1996). “Calcite nucleation and precipitation kinetics as affected by dissolved organic matter at 25 °C and pH 7.5.” Geochem. Cosmochim Acta, 60, 2767–2776. Logan, S. H., and Boyland, P. B. (1983). “Calculating heat units via a sine function.” J. Am. Soc. Hort. Sci., 108, 977–980. Maas, E. V., and Hoffman, G. J. (1977). “Crop salt tolerance: Current assessment.” J. Irrig. Drainage Div. ASCE, 103(IR2), 115–134. McNeal, B. L. (1968). “Prediction of the effect of mixed-salt solutions on soil hydraulic conductivity.” Soil Sci. Soc. Am. Proc., 32, 190–193. Pitzer, K. S. (1973). “Thermodynamics of electrolytes I: Theoretical basis and general equations.” J. Phys. Chem., 77, 268–277. ———. (1979). Activity coefficients in electrolyte solutions, Chapter 7, CRC Press, Boca Raton, Fla. Quilez, D., Isidoro, D., and Aragüés, R. (2011). “Conceptual irrigation project hydrosalinity model,” in Agricultural salinity assessment and management, Chapter 30, this volume. Robbins, C. W., Wagenet R. J., and Jurinak J. J. (1980). “A combined salt transportchemical equilibrium model for calcareous and gypsiferous soils.” Soil Sci. Soc. Am. J., 44, 1191–1194. Shainberg, I., and Levy, G. J. (1992). “Physico-chemical effects of salts upon infiltration and water movement in soils,” in R. J. Wagenet, P. Baveye, and B. A. Stewart, eds., Interacting processes in soil science, Lewis Publishers, CRC Press, Boca Raton, Fla.

896

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Sˇ imu˚ nek, J., and Suarez, D. L. (1993). “Modeling of carbon dioxide transport and production in soil: 1. Model development.” Water Resour. Res., 29, 487–497. Stokes, R. H. (1979). “Thermodynamics of solutions,” in Activity coefficients in electrolyte solutions, R. M. Pitkowicz, ed., CRC Press, Inc., Boca Raton, Fla. Suarez, D. L. (1977). “Ion activity products of calcium carbonate in waters below the rootzone.” Soil Sci. Soc. Am. J., 41, 310–315. ———. (2001). “Sodic soil reclamation: Model and field study.” Aust. J. Soil Res., 39, 1225–1246. ———. (2011). “Soil salinization and management options for sustainable crop production,” in Handbook of crop and plant stress, 3rd ed., M. Pessarakli, ed., CRC Press, Boca Raton, Fla. Suarez, D. L., and Dudley, L. (1998). “Hydro chemical considerations in modeling water quality within the vadose zone,” in Agroecosystems and the environment: Sources, control, and remediation of potentially toxic trace element oxyanions, L. Dudley and J. Guitjens, eds., American Association for the Advancement of Science–Pacific Division, San Francisco University, San Francisco, Calif. Suarez, D. L., Rhoades, J. R., Lavado, R., and Grieve, C. M. (1984). “Effect of pH on saturated hydraulic conductivity and soil dispersion.” Soil Sci. Soc. Am. J., 48, 50–55. Suarez, D. L., and Sˇ imu˚ nek, J. (1992). The UNSATCHEM code for simulating onedimensional variably saturated water flow, heat transport, carbon dioxide production and transport, and multicomponent solute transport with major ion equilibrium and kinetic chemistry, Ver 1.1, Research Report No. 129, U.S. Salinity Laboratory, USDA-ARS, Riverside, Calif. ———. (1993). “Modeling of carbon dioxide transport and production in soil: 2. Parameter selection, sensitivity analysis and comparison of model predictions to field data.” Water Resour. Res., 29, 499–513. ———. (1996). “Solute transport modeling under variably saturated water flow conditions,” in Reactive transport in porous media, reviews in mineralogy, Vol. 34, P.C. Lichtner, C.I. Steefel, and E. H. Oelkers, eds., Mineralogical Society of America, Washington, DC. ———. (1997). “UNSATCHEM: Unsaturated water and solute transport model with equilibrium and kinetic chemistry.” Soil Sci. Soc. Am. J., 61, 1633–1646. Suarez, D. L., and Wood, J. W. (1993). “Predicting Ca-Mg exchange selectivity of smetitic soils,” in Agronomy Abstracts, American Society Agronomy, Madison, Wisc., 236. Suarez, D. L., Wood, J. W., and Ibrabim, I. (1992). “Reevaluation of calcite supersaturation in soils.” Soil Sci. Soc. Am. J., 56, 1776–1784. Suarez, D. L., Wood, J. W., and Lesch, S. M. (2006). “Effect of SAR on water infiltration under a sequential rain-irrigation management system.” Agric. Water Mgmt., 86, 150–164. Truesdell, A. H., and Jones, B. F. (1974). “WATEQ, a computer program for calculating chemical equilibria of natural waters.” J. Res. U.S. Geol. Surv., 2, 233–248. van Genuchten, M. T. (1980). “A closed-form equation for predicting the hydraulic conductivity of unsaturated soils.” Soil Sci. Soc. Am. J., 44, 892–898.

MODELING TRANSIENT ROOTZONE SALINITY

897

———. (1987). A numerical model for water and solute movement in and below the rootzone, unpublished research report, U.S. Salinity Laboratory, USDA-ARS, Riverside, Calif. Wagenet, R. J., and Hutson, J. L. (1987). LEACHM: Leaching estimation and chemistry model, Continuum 2, Water Resources Institute, Cornell University, Ithaca, New York. White, N., and Zelazny, L. W. (1986). “Charge properties in soil colloids,” in Soil physical chemistry, D. L. Sparks, ed., CRC Press, Boca Raton, Fla., 39–81.

This page intentionally left blank

CHAPTER 29 LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY Jan W. Hopmans, G. Schoups, and K. K. Tanji

INTRODUCTION When predicting rootzone salinity and salt load in drainage and groundwater of irrigated agricultural systems, we must consider various interacting factors. The main mechanism of soil salt build-up is by evapoconcentration of the soil solution from evaporation, and transpiration of water (Tanji 1990). In the presence of shallow saline water tables, capillary rise can result in periodic transport of salts upward into the rootzone. The resulting rate of soil salinization will depend on water table depth, shallow groundwater salinity, and soil type. The degree of soil salinization may be further affected by the precipitation and dissolution of salts, primarily gypsum and calcite (Oster and Rhoades 1975). Cation exchange reactions between the soil solution and the soil exchange complex can further complicate salinity dynamics by altering the composition of cations in the solution that might lead to precipitation or dissolution of soil minerals and compositional changes in soil solution salinity (Robbins et al. 1980). Moreover, Tanji (1969) demonstrated the effects of ion pairing, common ions, and ionic strength on total solubility of gypsum in aqueous systems. An extensive review of salt chemistry in soil-water systems was presented by Oster and Tanji (1985). Because of the great complexity and interdependence of the soil processes affecting soil salinization, it is useful to identify the main processes and seek simplified process descriptions, thereby reducing the computational burdens and associated data requirements when considering long-term and regional scales. A key question, then, relates to the upscaling of point-scale physical and chemical processes to the larger time and spatial scales of interest (Hopmans et al. 2002), in the general presence of large model input uncertainties. 899

900

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

A common approach to temporal process scaling is the time-averaging of boundary conditions. This usually simplifies the calculations significantly, making it especially attractive when large-scale hydrologic systems need to be simulated. Previous studies on the effects of timeaveraging of boundary conditions on vadose zone solute transport (Wierenga 1977; Beese and Wierenga 1980; Destouni 1991; and Vanderborght et al. 2000) show that solute transport under transient variably saturated conditions may be approximated by a transport model with steady-state water flux and time-indifferent soil moisture content, and using a larger effective soil dispersivity to compensate for neglecting short-term moisture content variations. Nevertheless, time-averaging may provide satisfactory predictions of the spatial-ensemble distribution or statistical moments of the variables of interest, as noted by van der Zee and Boesten (1991) in their detailed transient numerical flow and reactive transport simulations. Schoups and Hopmans (2002) presented an efficient analytical solute transport model, designed for large-scale applications in the presence of large uncertainty. Other studies have shown that simple transport models can perform equally well as more complex models, when large time and space scales are involved (e.g., Jothityangkoon et al. 2001; Schoups and Hopmans 2002, 2006; van der Linden and Woo 2003). After a brief introduction of the UNSATCHEM model, we describe benchmark simulations that are representative for the western San Joaquin Valley, California, and discuss various model simplifications relative to (1) the time-scale of the boundary conditions, (2) the level of vertical discretization, and (3) the complexity of the soil chemistry system, as detailed in Schoups et al. (2006). Then we show the results of a long-term (60 years), regional-scale (1,400 km2) hydrosalinity modeling study (Schoups et al. 2005a) that illustrates the need to adequately define the controlling chemical and hydrological processes, irrespective of scale. Finally, we summarize the key findings of an ongoing sensitivity modeling study that evaluates the impact of climate change on irrigated agriculture in the region.

METHODS Description of the Numerical Model UNSATCHEM is a one-dimensional numerical soil water flow and transport model, simulating variably saturated flow, heat transport, CO2 production and transport, and solute transport of individual ions, coupled to equilibrium and kinetic chemistry routines (Sˇ imu˚nek and Suarez

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

901

1994; Sˇ imu˚nek et al. 1996). Variably saturated flow is simulated with the Richards’ equation: ⎞ ∂ ∂ ⎛ ∂h  ⎜ K  K ⎟  rw ⎠ ∂t ∂z ⎝ ∂z

(29-1)

where h is the soil water pressure head (L),  is volumetric water content (L3/L3), K is the unsaturated hydraulic conductivity (L/T), t is time (T), z is the spatial coordinate (positive upward) (L), and rw defines the root water uptake term (1/T). Root water uptake is simulated as a function of depth and time, and is a function of the pressure head and the osmotic head to account for water and salt stress, respectively. The dependence of K and h on  is represented by van Genuchten-Mualem–type models. Solute transport of each aqueous species considered is simulated with the advection-dispersion equation: ⎞ ∂c k ∂ ⎛ ∂c ∂c ∂c  ⎜ D k  qc k ⎟  b k  b k ⎠ ∂t ∂z ⎝ ∂z ∂t ∂t

∀k  1,, K , 7

(29-2)

where ck is total dissolved concentration of aqueous species k (M/L3), –c k is – total sorbed phase concentration of aqueous species k (M/M), –c k is total solid phase concentration of aqueous species k (M/M), b is soil bulk density (M/L3), D is the dispersion coefficient (L2/T), and q is the Darcy water flux (L/T). For ions that do not adsorb or react (conservative species, such as Cl), the second and third terms on the right side are zero. For all the other ions, the second and third terms are determined by solving the reaction system. This study considers transport of seven major ions, namely,  2 Ca2, Mg2, Na, K, HCO 3 , SO4 , and CI . The chemical reactions include ion complexation, cation exchange, and mineral precipitation-dissolution of both calcite and gypsum, as summarized in Table 29-1. All reactions are assumed to be at equilibrium, and both the modified Debye-Huckel and Pitzer expressions are incorporated to calculate single ion activities. All cations in solution are assumed to be in instantaneous equilibrium with their sorbed counterparts. These sorbed components balance the total net negative charge of the clay minerals and organic matter, defined by the soil’s cation exchange capacity (CEC). The complete chemical model consists of a nonlinear algebraic system, which is solved iteratively using several convergence criteria as described by Sˇ imu˚nek et al. (1996). Values for several chemical equilibrium constants are listed in Table 29-2. All other chemical constants were adapted from Sˇ imu˚nek et al. (1996).

902

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 29-1. Chemical Reactions Considered by UNSATCHEMa Reaction (1)

Equilibrium Constant (2)

CO2-H2O system H2O 3 H  OH CO2 (g)  H2O 3 H2CO*3 H2CO*3 3 H  HCO 3  2 HCO 3 H  CO 3 3

Kw KCO2 Ka1 Ka2

Ion pairing (complexation) CaSO04 3 Ca2  SO42 CaCO03 3 Ca2  CO32 CaHCO3 3 Ca2  HCO 3 MgSO04 3 Mg2  SO42 MgCO30 3 Mg2  CO32 MgHCO3 3 Mg2  HCO 3  2 NaSO 3 Na  SO 4 4  2 NaCO 3 3 Na  CO3 0  NaHCO 3 3 Na  HCO 3 KSO4 3 K  SO42

K1 K2 K3 K4 K5 K6 K7 K8 K9 K10

Cation exchange reactions 1 2 1 2 1 2

1 2 1 2

2 Ca  Mg ⇔ Ca 

1 2

Mg

2

2  Ca  Na ⇔ Ca  Na

Ca  K ⇔ Ca  K 2



KMg–Ca KCa–Na KCa–K

Mg

2

 Na ⇔ Mg  Na

KMg–Na

Mg

2

 K ⇔ Mg  K 

KMg–K

Na  K ⇔ Na  K 

KNa–K

Precipitation-dissolution CaSO4 2H2O(s) 3 Ca2  SO2 4  2H2O CaCO3(s)  CO2(g)  H2O 3 Ca2  2HCO 3

KGsp KCsp

a

All reactions are in instantaneous equilibrium. Chemical species with an overscore denote sorbed species. The “*” in H2CO*3 indicates a species of low stability.

From Sˇ imu˚nek et al. (1996).

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

903

Benchmark Numerical Simulations The sensitivity analysis considers a hypothetical situation of an irrigated cotton crop, grown under conditions that are typical for the western San Joaquin Valley (SJV) in California. The benchmark model domain consists of a one-dimensional vertical 210-cm-deep soil profile, representing the crop rootzone. The domain is discretized by 234 nodes to yield a uniform nodal spacing of a little less than 1 cm. Each simulation extends for a period of 10 years, with each year consisting of a fallow winter period followed by the irrigated cotton crop. The upper boundary conditions of rainfall, irrigation, and potential evapotranspiration (ET) are specified on a daily basis. Data correspond to the October 1987 to September 1988 hydrologic year (Hanford, California, weather station) with an annual average rainfall amount of 213 mm. These same boundary conditions were repeated every year. Six irrigations were applied during the growing season, with amounts and timing representative for a furrowirrigated cotton crop. For the lower boundary condition we specified either gravity drainage (deep water table) or a constant zero pressure head to represent a shallow water table condition. Selected soil physical and chemical properties are representative for the clayey soils that are typical for the study area (Table 29-2). Annual average ion concentrations were specified for rain and irrigation and the shallow water table (Table 29-3). Initial soil solution concentrations were set equal to irrigation water concentration, whereas the initial soil calcite and gypsum values were assumed constant with depth (Table 29-2). Initially, the soil exchange complex was assumed to be saturated with Ca, with chemical equilibrium between all ions in solution and the sorbed and solid phases. Salinity was simulated with (R) and without (NR) chemical reactions for a range of leaching fractions (LF’s), water table depths, groundwater salinities, and initial gypsum contents, for a total of seven cases (Table 29-4). The first three cases (cases 1, 2, and 4) correspond to free drainage (i.e., a deep water table and net leaching conditions), whereas cases 5 through 8 represent a saline shallow water table with deficit irrigation, causing net capillary rise upward and limited salt leaching. We intentionally omitted case 3 (58 mmolc/kg gypsum and NR) in Table 29-4, because initial gypsum content is irrelevant for the NR cases. The shallow groundwater cases were repeated for two different groundwater salinities and composition (samples 1 and 2 in Table 29-3), representative of shallow groundwater qualities in the western SJV (Deverel and Gallanthine 1989). Salinity of groundwater sample 1 is moderate (3,000 mg/L), while salinity of sample 2 is high (8,000 mg/L) and saturated with respect to gypsum. Simulations with samples 1 and 2 were conducted with initial soil gypsum content values of 10 (0.1% w/w) and 58 mmolc/kg (0.5% w/w), respectively. Although these various scenarios do not provide all possible combinations of soil gypsum content, ground-

TABLE 29-2. Overview of Parameter Values Used in the UNSATCHEM Simulations 904

Parameter (1) Soil bulk density Hydraulic parameters for clay Ks

n r s Molecular diffusion coefficient Dispersivity Cation exchange coefficient Gapon selectivity coefficients KMg–Ca KCa–Na KCa–K Gypsum solubility, KGsp (20 °C) Calcite solubility, KCsp (20 °C) Soil temperature CO2 content Initial dissolved concentrations Initial sorbed concentrations —– —– —– — (Ca /Mg/Na /K ) Initial calcite Initial gypsum Annual rainfall Annual potential ET Number of irrigation events Irrigation amount (by event) ET, evapotranspiration

Value (2) 1.4 14.8 0.015 1.25 0.10 0.46 105 8.3 350 0.63 6.3 0.36 2.5 105 3.5 109 20 0.00033 (top) 0.02 (bottom) Irrigation water 350/0/0/0 400 10 or 58 213 927 6 160/160/100

Units (3) 3

g/cm

Reference or Comment (4) Jury et al. (1991)

Schaap et al. (1998) cm/day 1/cm — — — cm2/s cm mmolc/kg soil

Fetter (1999) Biggar and Nielsen (1976) Jury et al. (1991)

— — — — — °C vol%

Robbins et al. (1980) Robbins et al. (1980) Robbins et al. (1980) Simunek et al. (1996) Simunek et al. (1996) Constant in time Constant in time Table 29-3

mmolc/kg soil mmolc/kg soil mmolc/kg soil mm/yr mm/yr mm

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

905

TABLE 29-3. Salinity and Composition of Rain, Irrigation, and Shallow Groundwater Shallow Groundwater Variable (1)

Unit (2)

Ca Mg Na

mmolc/L mmolc/L mmolc/L

K HCO3 SO4 Cl

mmolc/L mmolc/L mmolc/L mmolc/L

SAR TDS

a

SAR 

Sample 1 (5)

Sample 2 (6)

0.002 0.0032

0.998 1.23

14.5 9.0

28.4 15.6

0.0183 0.0005 0.005 0.009

2.54 0.0 0.822 1.10

31.0 0.0 2.82 30.0

87.0 0.18 3.5 109

0.01 0.36 1.61

2.85 2.40 308

21.7 9.04 3500

18.9 18.5 8880

Rain (3)

0.5

(mmolc/L) mg/L

Reference NADP (2003) a

Irrigation Water (4)

CA-DWR (1990)

Deverel and Gallanthine (1989)

Na Ca  Mg 2

water salinity, and irrigation amount, they collectively represent a wide range of conditions for the western SJV. For each case, the two main variables of interest were average rootzone salinity, expressed as total dissolved solids (TDS in mg/L or ppm), and annual drainage salt load, S (g/m2/year), at the conclusion of the 10-year simulation period. We note that the salt load is negative if the salt flux is downward and out of the rootzone.

Model Simplifications The benchmark simulations were used as a reference (M0 in Table 29-5) to quantify errors resulting from subsequent model simplifications. From this information we will identify the optimal level of model complexity that accounts for the most important processes. Simplifications (models M1 through M5 in Table 29-5) were introduced incrementally and are related to (1) the time-scale of the boundary conditions, (2) the level of vertical discretization, and (3) the complexity of the reaction system in Table 29-1. Time-averaged boundary conditions used annual-averaged

906

TABLE 29-4. Parameter Values for the Different Cases Water Table Depth (cm) (2)

Net Leaching (3)

Irrigation Amount per Event (mm) (4)

Initial Gypsum (mmolc/kg)a (5)

Groundwater Salinity (6)

Reactions (7)

1 2 4 5 6 7

Free drainage Free drainage Free drainage 210 210 210

Yes Yes Yes No No No

160 160 160 100 100 100

— 10 58 — 10 —

— — — Sample 1 Sample 1 Sample 2

NR R R NR R NR

8

210

No

100

58

Sample 2

R

Case (1)

a

For NR cases, soil gypsum content is irrelevant.

R, reaction; NR, no reaction.

TABLE 29-5. Summary of Different Models with Increasing Levels of Simplification Level of UNSATCHEM Model Simplification (1)

Boundary Conditions (2)

Nodal Spacing (cm) (3)

Cation Exchange (4)

Calcite DissolutionPrecipitation (5)

Ion Complexation (6)

Gypsum DissolutionPrecipitation (7)

M0 M1 M2 M3 M4

Daily Annual Annual Annual Annual

1 1 15 15 15

Yes Yes Yes No No

Yes Yes Yes Yes No

Yes Yes Yes Yes No

Yes Yes Yes Yes Yes

M5

Annual

15

No

No

No

No

907

908

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

values for evaporation, transpiration, and infiltration rates, as computed from the benchmark simulations. The time-averaged model uses a fixed root water uptake distribution, equal to that of the fully developed crop of the benchmark model. As an additional simplification, we decreased the number of spatial nodes from 234 to 15. This coarser discretization with a nodal spacing of about 15 cm (M2 through M5 in Table 29-5) allowed much larger time steps, thereby reducing computing time. We expect that additional errors are introduced with increasing time steps, since transport and reactions are solved in sequence rather than simultaneously. A third simplification was considered by simplifying the salt chemistry of Table 29-1 incrementally in three steps. First, in simulation M3 the cation exchange reactions were turned off by setting the CEC equal to 0. Second, in simulation M4 no reactions were included except for gypsum dissolution-precipitation using a constant solubility product. Finally, the third step (M5) does not include any chemical reactions. The approach of using a constant solubility product neglects the effects of cation exchange, ion complexation, and ionic strength on total solubility of gypsum and calcite precipitationdissolution. The various stepwise model simplifications implemented are summarized in Table 29-5, realizing that the various salt chemistry levels (M3–M5) are only relevant for the R cases of Table 29-4. The errors due to simplification are quantified by the relative absolute difference (%) of average rootzone TDS and annual drainage salt load, S, at the end of the 10-year simulation period, or

ε

X X X

(29-3)

— where X is the prediction with the simplified model, and X is the daily prediction from the benchmark simulation. The comparison of model performance after 10 years diminishes the influence of the assumed initial conditions.

RESULTS AND DISCUSSION Benchmark Simulations for Different Cases (M0) Figure 29-1A presents the 1-year daily dynamics of the water balance components for the deep groundwater scenario (cases 1, 2, and 4). The occasional rain events in the winter when no crop is grown cause large fluctuations in soil evaporation (E). The pre-irrigation (I) in February

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

909

FIGURE 29-1. Daily water balance for one year for (A) deep water table scenario (cases 1–4, Table 29-4), and (B) shallow groundwater table with no net leaching (cases 5–8, Table 29-4). Definitions of symbols: E, evaporation; T, transpiration; D, drainage; R, rainfall; and I, irrigation.

(Table 29-4) is intended to leach salts by drainage (D) that accumulate during the previous growing season and to provide favorable soil moisture conditions before planting. The cotton crop is planted in April, resulting in subsequent transpiration (T) and decreasing E, except after the first two irrigations when the soil is only partly covered. During the growing season, five additional 160-mm irrigations were applied to satisfy crop water requirements, resulting in significant drainage fluxes (negative). We note that the bottom-water flux (D) is never upward (positive), because of the gravity drainage boundary condition. Figures 29-2A and B compare the simulated dynamics of daily average rootzone soil salinity (TDS) and cumulative annual drainage salt load (S) during the 10-year simulation periods for cases 1 (NR) and 4 (R). The corresponding sodium adsorption ratio (SAR) and soil gypsum values for the R case are presented in Fig. 29-2C and D. Considering the 10-year simulation period, for the case with reactions (R), the model predicts a decrease in soil salinity over time. This slow decrease follows the initial

910

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 29-2. Simulated variables with the daily UNSATCHEM model (M0) for deep groundwater scenarios, case 4 (free drainage, initial gypsum content  58 mmolc/kg, with chemical reactions “R”) and case 1 (free drainage, without chemical reactions “NR”). (A) daily rootzone average TDS, (B) annual drainage salt load S, (C) daily rootzone average SAR for case 4, (D) daily rootzone average gypsum content for case 4. instantaneous dissolution of gypsum at time zero, resulting in an initial soil salinity level of about 2,300 mg/L that is much larger than the irrigation water TDS (308 mg/L, Table 29-3). As irrigation proceeds, irrigation with gypsum-undersaturated water gradually leaches the gypsum (Fig. 29-2D) and decreases soil salinity (Fig. 29-2A). Continuing the simulations beyond the 10-year period is expected to cause additional leaching of gypsum and a soil salinity level similar to the nonreactive (NR) case. Simulation results show that for the gypsum-undersaturated irrigation water,

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

911

gypsum dissolution is the main contributor to soil salinity for deep water table conditions, and is much more important than calcite because of its much higher solubility. Because of the significant gypsum dissolution, the drainage salt load, S, (Fig. 29-2B) is also larger (more negative) than the corresponding nonreactive (NR) case. Finally, since the simulations start with an initially Ca-saturated sorption complex, the subsequent exchange of sorbed Ca by Na of the rainfall and irrigation water gradually increases the SAR (Fig. 29-2C) during the simulation period. For the shallow groundwater scenario with no net leaching (cases 5 through 8), all irrigation amounts were reduced to 100 mm, thereby simulating deficit irrigation conditions leading to negative leaching (annual net upward flow) in Fig. 29-1B. The bottom boundary consisted of a fixed water table. The reduced irrigations caused significant upward fluxes (positive value for D), both between irrigation events and the nongrowing season. The resulting effects on soil salinization are shown in Fig. 29-3.

FIGURE 29-3. Simulated variables with the daily UNSATCHEM model (M0) for shallow groundwater scenarios with no net leaching, case 8 (210-cm water table, initial gypsum content  58 mmolc/kg, with chemical reactions “R”) and case 7 (210-cm water table, without chemical reactions “NR”). (A) daily rootzone average TDS, (B) annual drainage salt load S, (C) daily rootzone average SAR for case 8, (D) daily rootzone average gypsum content for case 8.

912

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The total net water flux at the bottom of the rootzone is upward, resulting in zero leaching of salts, leading to very high rootzone TDS levels (Fig. 29-3A). This clearly illustrates the need for maintaining a salt balance in the rootzone by periodic leaching of salts. The salt load is consistently positive, directed upward into the rootzone, resulting in an annual salt influx of about 8,000 g/m2/yr after a few years (Fig. 29-3B). Comparing annual average advective flux from the saline (sample 2) shallow water table with computed diffusive transport shows that the upward salt flux is dominated by diffusion and dispersion accounting for about 6,000 g/m2/yr. This large salt mass is controlled by continuous precipitation of the incoming salts. For the R case, significant buffering of rootzone salinity occurs by precipitation of gypsum (Fig. 29-3D), thereby decreasing TDS as in Fig. 29-3A. The predicted SAR values are large and increasing with time due to the inflow of Na-rich groundwater and the precipitation of Ca as gypsum. The combined results for the no-leaching cases show that predicted rootzone salinity can be largely overestimated if there is no accounting for gypsum precipitation. For each of the seven presented cases, the first column in Tables 29-6 and 29-7 lists the benchmark simulation results of TDS and S, respectively. The other five columns present the TDS (Table 29-6) and S (Table 29-7) simulation results for the model simplifications M1 through M5, with the relative error in percent between parentheses. Effects of Annually Averaging the Boundary Conditions (M1) and Vertical Discretization (M2) The third column in Tables 29-6 and 29-7 compares TDS and S, using annual-averaged boundary conditions for the corresponding benchmark cases in the second column. The magnitude of the relative errors varies between near zero and 42%. A comparison of the TDS values for cases 1 through 4 (Table 29-4) indicates that there is a consistent underprediction or bias by the time-averaged model. In contrast, for cases 5 through 8 with no net leaching, the time-averaged model slightly overpredicts rootzone TDS. The underprediction for the first three cases is caused by the omission of the water flux direction changes when applying the annual timeaveraged model that uses the simplified averaged net downward flux at all times, thereby underestimating the mixing of soil solution by dispersion as caused by changes in soil water flow direction during and between irrigation events (e.g., capillary rise, root water uptake, and redistribution). The largest effects on annual salt drainage load, S, in Table 297 (case 8) using an annual-average boundary condition were caused by the elimination of periodic flow direction changes as compared to the benchmark simulations, affecting salt transport and gypsum precipitation, especially for the cases with high soil gypsum and groundwater salinity. In general, we found that in comparisons between the reactive (R) and

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

913

TABLE 29-6. Predicted Rootzone Average TDS (mg/L) and Relative Errors (%)a Level of Model Complexity Case (1)

M0 (2)

M1 (3)

1

1,130

944 (17)

2

1,090

4

2,050

1,860 (9)

1,860 (9)

5

11,100

12,040 (9)

6

5,970

6,510 (9)

7

26,200

26,100 (0)

27,600 (5)

8

8,280

8,690 (5)

9,030 (9)

978 (11)

M2 (4)

M3 (5)

943 (17)

M4 (6)

943 (17)

973 (11)

1,040 (5)

943 (17) 943 (14)

M5 (7)

943 (17) 943 (14)

1,980 (3)

1,910 (7)

12,500 (13)

12,500 (13)

12,500 (13)

12,500 (13)

6,620 (11)

10,100 (69)

9,760 (63)

12,500 (109)

943 (54)

27,600 (5)

27,600 (5)

27,600 (5)

22,800 (175)

22,000 (165)

27,600 (234)

a

After 10 years between daily model (M0) and different models using increasing levels of simplification, as defined in Table 29-5. The different cases are summarized in Table 29-4.

Errors larger than 50% are highlighted in bold face.

nonreactive (NR) cases, the prediction errors were lower when reactions are considered. Differences in relative errors between the time-averaged models with fine (M1) and coarse discretization (M2) reflect errors in vertical resolution. Results in Tables 29-6 and 29-7 show little effect on simulation results when applying the coarser discretization. TABLE 29-7. Predicted Annual Drainage Salt Load S (g/m2/yr) and Relative Errors (%)a Level of Model Complexity Case (1)

M0 (2)

M1 (3)

M2 (4)

M3 (5)

M4 (6)

M5 (7)

1

278

288 (4)

288 (4)

288 (4)

288 (4)

288 (4)

2

275

279 (1)

283 (3)

307 (12)

289 (5)

288 (5)

4

566

564 (0)

563 (0)

653 (15)

607 (7)

288 (49)

5

692

765 (11) 658 (7)

767 (11)

767 (11)

767 (11)

767 (11)

6

615

743 (21)

743 (21)

767 (25)

767 (25)

7

1,770

1,940 (10)

1,950 (10)

1,950 (10)

1,950 (10)

1,950 (10)

8

7,960

2,210 (72)

2,030 (75)

2,010 (75)

2,200 (72)

1,950 (76)

a

After 10 years between daily model (M0) and different models using increasing levels of simplification, as defined in Table 29-5. The different cases are summarized in Table 29-4.

Errors larger than 50% are highlighted in bold face.

914

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Effects of Simplifying Salt Chemistry (M3, M4, and M5) In addition to model simplifications M1 and M2, the benchmark simulations were compared with those that simplified the complex salt chemistry of Table 29-1. The simplification included exclusion of cation exchange reactions (M3), consideration of gypsum dissolution-precipitation only (M4), using a constant total solubility of gypsum in distilled water of 30.7 mmolc/L, and exclusion of all chemical reactions (M5). Comparing the relative errors of M3 with the M2 simulations, the differences for cases 1 through 4 were small; however, we computed a large overestimation (69% and 175%, respectively) for cases 6 and 8. For the corresponding high-saline, Na-rich shallow groundwater cases, the neglect of cation exchange of Ca by Na resulted in an underestimation of gypsum precipitation, as would occur in the absence of cation exchange. In summary, cation exchange may significantly influence rootzone TDS by affecting the precipitation-dissolution of gypsum. The error magnitude will depend on cationic composition differences between the sorption complex and the infiltrating soil solution. Further inspection of Tables 29-6 and 29-7 illustrates the negligible effects of simplification (M4), ignoring calcite precipitation-dissolution, ionic strength, and ion complexation reactions. Therefore, we conclude that inclusion of cation exchange reactions is more important when predicting rootzone salinity than considering calcite and ion complexation. Because the solubility of calcite is much lower than the solubility of gypsum (Ksp of 109 versus 105), calcite plays a secondary role in soil salinity. Finally, the M5 (NR) simulation results in the last column of Table 29-6 demonstrate the high importance of accounting for gypsum dissolutionprecipitation, because errors are as large as 50% (case 4) by ignoring dissolution, and larger than 200% (case 8) by eliminating gypsum precipitation.

Regional-Scale Application For this purpose, the regional-scale hydrology model MODHMS (Panday and Huyakorn 2004) was coupled with the soil chemistry module of UNSATCHEM (Sˇ imu˚nek et al. 1996) to provide for an integrated threedimensional, variably saturated subsurface flow and reactive salt transport (Schoups 2004), solving for the flow and transport equations [1] and [2], extended in three spatial dimensions. The horizontal boundaries of the model domain coincided with the hydrologic boundaries of an earlier regional groundwater flow model (Belitz and Philips 1995), defined by the trough of the SJV on the east, the Coast Range foothills in the west, and no-flow boundaries in the north and south of the regional flow domain (Fig. 29-4A). The model domain was discretized into a regular finite difference grid of 2,960 square cells of 805 m (0.5 mi) side length and 64-ha area, corresponding to a typical field size. In the vertical direction,

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

915

FIGURE 29-4. Overview of the study area. (A) Location of the study area in the western San Joaquin Valley, California, that includes 13 water districts, (B) soil texture map, and (C) soil gypsum contents. the model domain extended from the land surface to the top of the Corcoran clay, using 17 layers of increasing thickness from the surface downward. The total number of active model grid cells was 36,040. Hydrologic flows and salt transport were simulated for a 57-year period, from 1940 to 1997, using annual-average boundary conditions and grid cell-specific soil parameters. The salinity module included reactions, such as cation exchange and precipitation and dissolution of gypsum and calcite (Tanji et al. 1967; Schoups et al. 2006a), using spatially distributed information as

916

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

presented in Figs. 29-4A and B. Using historical crop acreage and water delivery records for each water district, crops and irrigation amounts were randomly distributed, leading to the annual assignment of a single crop to each grid cell. Other required boundary conditions were needed to quantify vertical (across Corcoran clay and into deep groundwater) and lateral (toward the San Joaquin River) water flow and salt fluxes and exchange between the simulated domain and its surroundings, so that an annual salt balance could be estimated. Spatially distributed water flow and salinity reaction and transport parameters were obtained from soil survey data and 242 well logs [more information is available in the online supplement in Schoups et al. (2005a)]. Hydrological parameter values were either optimized (Schoups et al. 2005b) or obtained from existing information. Results of this comprehensive modeling study were presented in Schoups et al. (2005a), within the context of evaluating the sustainability of irrigated agriculture in the SJV. Most of the simulation results during the 57-year historical period were intuitively clear and could be explained by the historical development of irrigation projects in the region, shifting from groundwater to mostly surface water allocations, multiple multiyear droughts, and presence of relatively large quantities of soil gypsum. The hydrologic component of the hydrosalinity model simulated the dynamics of the regional variation in water table depths, reconstructing the gradual increase in shallow water table area from the 1950s to the 1990s due to increased recharge from irrigated agriculture compared to predevelopment conditions, and the shift in irrigation water supply from locally pumped groundwater to imported surface water beginning in the early 1970s. Deep percolation of water through the Corcoran clay was highest during the 1950–1970 period, when pumping rates from the confined aquifer were the highest. As surface water was increasingly used, the hydraulic head gradient across the clay layer decreased, thus reducing deep percolation flows. Drainage flows were relatively small, starting in the late 1950s and reaching a maximum when the subsurface drainage systems in the Westlands Water District were operated from 1980 to 1985. Much of the spatial and temporal dynamics in rootzone and groundwater salinity was adequately described with the hydrosalinity model. The salinity dynamics in the shallow groundwater generally followed that of the rootzone, indicating that the two systems were closely connected. The relatively slow movement of salts to larger depths indicated the long travel time needed for salts to move into the deeper groundwater. Our model simulations demonstrated that a significant portion of the soil salinity dynamics was controlled by the cycling of soil gypsum through dissolution and precipitation, as caused by changes in salt leaching with time and soil depth, and soil cation exchange between Ca and Na. A summary of the 57-year soil salinity dynamics is represented in Fig. 29-5 by a time-series of the number of model grid cells with a rootzone

FIGURE 29-5. Simulated salinity changes of regional example of Fig. 29-4. (A) Time-series of number of model grid cells with a simulated average rootzone ECe greater than 4 dS/m (solid line) and greater than 8 dS/m (dashed line). Symbols correspond with measured data. (B) Simulated changes in total salt storage and dissolved salts (in million tons) since 1940.

917

918

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

average salt concentration (ECe) greater than 4 dS/m, which identifies the salt-affected soils, and 8 dS/m. The few measured data points in Fig. 29-5A were derived from aggregating measured soil salinity data reported in 1969 and 1992 soil surveys. Initially (1940), soil salinity was high and decreased gradually until about 1975 due to salt leaching when water tables were relatively deep. Model simulations were able to describe temporal changes in leaching as controlled by irrigation water source (surface water versus groundwater) and water table depth. For example, the reversal in the general pattern of soil salinity decrease during the 1970s was caused by water table rise in the eastern portion of the simulation domain, resulting in capillary rise of relatively high-salinity groundwater into the rooting zones. The hydrosalinity model was also able to reconstruct the effects of droughts in 1976–1977 and 1991–1992, resulting in peaks in soil salinity. The simulated cumulative change in total salt storage over the 60-year simulation period (Fig. 29-5B) across the whole depth domain (to the Corcoran clay) shows that a pseudo-equilibrium developed after 1970. Although the salt balance results indicate that crop productivity can be maintained, sustainability is threatened in two ways. First, the storage of dissolved salts has increased continuously since 1945 at an average rate of about 0.5 million tons/year (Fig. 29-5B) due to gypsum dissolution. Second, the simulations also showed that the deeper aquifers below the Corcoran clay accumulate salt, thereby degrading deep groundwater quality. This process of salinization of the deeper groundwater bodies may take many decades or longer, putting the sustainability of current irrigation practices into question. In addition, Hopmans and Maurer (2009) examined the potential regional-scale impacts of climate change on sustainability of irrigated agriculture for the same hydrologic domain, considering potential changes in irrigation water demand and supply, and they quantified impacts on the hydrologic system, soil, and groundwater salinity with associated crop yield reductions. This analysis was based on archived output from General Circulation Model (GCM) climate projections through 2100, which were downscaled to the 1,400 km2 study area. We accounted for uncertainty in GCM climate projections by considering two different GCMs, each using three greenhouse gas emission scenarios. Significant uncertainty in projected precipitation created large uncertainty in surface water supply. The largest demand reductions were computed for the dry and warm scenarios, because of increased land fallowing with corresponding decreased total crop water requirements. A decrease in seasonal crop ET by climate warming, despite an increase in evaporative demand, was attributed to faster crop development with increasing temperatures. Simulations of hydrologic response to climate-induced changes suggest that the salt-affected area will be slightly expanded. However, irrespective of climate change, the sensitivity analysis showed that salinity is expected to

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

919

increase in downslope areas, thereby limiting crop production to mostly upslope areas of the simulation domain. Results show that increasing irrigation efficiency may be effective in controlling salinization by reducing groundwater recharge and improving soil drainage, and in mitigating climate-warming effects, by reducing the need for groundwater pumping to satisfy crop water requirements.

SUMMARY AND CONCLUSIONS The numerical simulation of long-term, regional, variably saturated subsurface flow and transport remains a computational challenge, even with today’s computing power. Therefore, it is appropriate to develop and use simplified models that focus on the main hydrosalinity processes operating at the pertinent time and space scales. This study investigated the effects of various model simplifications on the prediction of long-term salt transport and soil salinity in irrigated soils. We presented an analysis of various model simplifications on the prediction of long-term soil salinity and salt transport in irrigated soils, using UNSATCHEM. Model simulations consist of benchmark scenarios, with and without soil chemical reactions. These hypothetical benchmark simulations are compared with the results of various model simplifications that considered (1) annual average boundary conditions, (2) coarser spatial discretization, and (3) reduced complexity of the salt-soil reaction system. We conclude that for longterm salinity modeling purposes, a simplified modeling approach can be used with annually averaged boundary conditions and a relatively coarse spatial discretization, but that it must include cation exchange and gypsum dissolution-precipitation reactions. The main advantage of these simplifications is that it leads to much smaller simulation times and reduced data input requirements. We illustrate this approach to regional-scale modeling, clarifying the importance of precipitation-dissolution reactions for the long-term salt balance. The regional long-term simulation of water and salt flows showed that salinization issues are critical to the sustainability of irrigated agriculture in the San Joaquin Valley and likely for many other areas of the world with relatively closed groundwater systems. We also concluded that many of the simulated adverse effects, such as soil salinization, are caused by regional groundwater dynamics of the hydrologic system in the study area, irrespective of climate change. Acknowledgments The authors acknowledge financial support by the USDA Fund for Rural America, Project No. 97-36200-5263, and the scientific contributions by the principal investigators W. W. Wallender, T. C. Hsiao, S. L.

920

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Ustin, R. E. Howitt, T. H. Harter, and G. E. Fogg. We also thank D. L. Suarez and J. Sˇ imu˚nek for providing us with the UNSATCHEM software. The research was made possible through funding by the UC Water Resources Center Project SD011.

REFERENCES Beese, F., and Wierenga, P. J. (1980). “Solute transport through soil with adsorption and root water uptake computed with a transient and a constant-flux model.” Soil Sci., 129, 245–252. Belitz, K., and Phillips, S. P. (1995). “Alternative to agricultural drains in California’s San Joaquin Valley: Results of a regional-scale hydrogeologic approach.” Water Resour. Res., 31(8), 1845–1862. Biggar, J., and Nielsen, D. R. (1976). “Spatial variability of the leaching characteristics of a field soil.” Water Resour. Res., 12, 78–84. California Department of Water Resources (CA-DWR). (1990). Water quality data of the California Aqueduct, California Department of Water Resources, Sacramento, Calif. Destouni, G. (1991). “Applicability of the steady-state flow assumption for solute advection in field soils.” Water Resour. Res., 27, 2129–2140. Deverel, S. J., and Gallanthine, S. K. (1989). “Relation of salinity and selenium in shallow groundwater to hydrologic and geochemical processes, western San Joaquin Valley, California.” J. Hydrol., 109, 125–149. Fetter, C. W. (1999). Contaminant hydrogeology, Macmillan, New York. Hopmans, J. W., Nielsen, D. R., and Bristow, K. L. (2002). “How useful are smallscale soil hydraulic property measurements for large-scale vadose zone modeling,” in Heat and mass transfer in the natural environment, the Philip volume, D. Smiles, P. A. C. Raats, and A. Warrick, eds., AGU Geophysical Monograph Series No. 129, American Geophysical Union, Washington, D.C., 247–258. Hopmans, J. W., and Maurer, E. P. (2008). Impact of climate change on irrigation water availability, crop water requirements and soil salinity in the SJV, CA, University of California Water Resources Center Technical Completion Report Project SD011, Berkeley, Calif., http://escholarship.org/uc/item/0g21p5hs. Jothityangkoon, C., Sivapalan, M., and Farmer, D. L. (2001). “Process controls of water balance variability in a large semi-arid catchment: Downward approach to hydrological model development.” J. Hydrol., 254, 174–198. Jury, W. A., Gardner, W. R., and Gardner, W. H. (1991). Soil physics, John Wiley and Sons, New York. Oster, J. D., and Rhoades, J. D. (1975). “Calculated drainage water compositions and salt burdens resulting from irrigation with river waters in the western United States.” J. Environ. Qual., 4, 73–79. Oster, J. D., and Tanji, K. K. (1985). “Chemical reactions within the rootzone of arid zone soils.” J. Irr. Drain. Eng., 111, 207–217. Panday, S., and Huyakorn, P. S. (2004). “A fully coupled physically-based spatially-distributed model for evaluating surface-subsurface flow.” Adv. Water Resour., 27, 361–382.

LONG-TERM REGIONAL-SCALE MODELING OF SOIL SALINITY

921

Robbins, C. W., Wagenet, R. J., and Jurinak, J. J. (1980). “A combined salt transport-chemical equilibrium model for calcareous and gypsiferous soils.” Soil Sci. Soc. Am. J. 44, 1191–1194. Schaap, M. G., Leij, F. J., and van Genuchten, M. T. (1998). “Neural network analysis for hierarchical prediction of soil water retention and saturated hydraulic conductivity.” Soil Sci. Soc. Am. J., 62, 847–855. Schoups, G. (2004). “Regional scale hydrologic modeling of subsurface water flow and reactive salt transport in the western San Joaquin Valley, Ca.” Ph.D. dissertation. University of California, Davis, Calif. Schoups, G., and Hopmans, J. W. (2002). “Analytical model for vadose zone solute transport with root water and solute uptake.” Vadose Zone J., 1, 158–171. ———. (2006). “Evaluation of model complexity and input uncertainty of fieldscale water flow and salt transport.” Vadose Zone J., 5, 951–962. Schoups, G., Hopmans, J. W., and Tanji, K. K. (2006). “Evaluation of model complexity and space-time resolution on the prediction of long-term soil salinity dynamics, western San Joaquin Valley, CA.” Hydrol. Proc., 20, 2647–2668. Schoups, G., Hopmans, J. W., Young, C. A., Vrugt, J. A., Wallender, W. W., Tanji, K. K., and Panday, S. (2005a). “Sustainability of irrigated agriculture in the San Joaquin Valley, California.” Proc. Nat. Acad. Sci. USA, 102, 15352–15356. Schoups, G., Hopmans, J. W., Young, C. A., Vrugt J. A., and Wallender, W. W. (2005b). “Multi-criteria optimization of a regional spatially-distributed subsurface water flow model.” J. Hydrol., 311, 20–48. Sˇ imu˚nek J., and Suarez, D. L. (1994). “Two-dimensional transport model for variably saturated porous media with major ion chemistry.” Water Resour. Res., 30, 1115–1133. Sˇimu˚nek J., Suarez D. L., and Sejna, M. (1996). The UNSATCHEM software package for simulating one-dimensional variably saturated water flow, heat transport, carbon dioxide production and transport, and multi-component solute transport with major ion equilibrium and kinetic chemistry, U.S. Salinity Laboratory, Research Report No. 141, U.S. Salinity Laboratory, Riverside, Calif. Tanji, K. K. (1969). “Solubility of gypsum in aqueous electrolytes as affected by ion association and ionic strengths up to 0.15M and at 25°C.” Environ. Sci. Tech., 3, 656–661. ———. (1990). “Nature and extent of agricultural salinity,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va. Tanji, K. K., Doneen, L. D., and Paul, J. L. (1967). “Quality of percolating waters. III. The quality of waters percolating through stratified substrata, as predicted by computer analyses.” Hilgardia, 38, 319–353. Vanderborght, J., Jacques, D., and Feyen, J. (2000). “Deriving transport parameters from transient flow leaching experiments by approximate steady-state flow convection-dispersion models.” Soil Sci. Soc. Am. J., 64, 1317–1327. Van der Linden, S., and Woo, M. (2003). “Application of hydrological models with increasing complexity to subarctic catchments.” J. Hydrol., 270, 145–157. Van der Zee, S. E. A. T. M., and Boesten, J. J. T. I. (1991). “Effects of soil heterogeneity on pesticide leaching to groundwater.” Water Resour. Res., 27, 3051–3063. Wierenga, P. J. (1977). “Solute distribution profiles computed with steady-state and transient water movement models.” Soil Sci. Soc. Am. J., 41, 1050–1055.

922

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

NOTATION CEC EC ET GCM LF SAR SJV TDS

 cation exchange capacity  electrical conductivity  evapotranspiration  general circulation model  leaching fraction  sodium adsorption ratio  San Joaquin Valley  total dissolved solids

CHAPTER 30 CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL D. Quílez, D. Isidoro, and R. Aragüés

INTRODUCTION Salt Balance and Irrigation Return Flows (IRF) Agricultural irrigation return flows (IRFs) are receiving increasing attention due to proliferating water-quality regulations and the public’s mounting concern about water conservation. The overuse and misuse of water in irrigated agriculture causes undesirable on-site effects (e.g., waterlogging and soil salinization and/or sodification) as well as off-site effects (e.g., degradation of water quality that affects potential downstream users) (Tanji and Kielen 2002). In arid and semiarid areas where the pressure on water resources is increasing, the management of irrigated areas to prevent salinization and degradation of water quality is becoming a critical issue (Prendergast et al. 1994; Minhas 1996; Heaven et al. 2002; Burkhalter and Gates 2006). Control of surface water quality in the United States is enforced by the federal Clean Water Act through the Total Maximum Daily Load (TMDL) program (US EPA 1991). Several TMDLs have been established for various pollutants such as nitrogen, pesticides, sediments, phosphorus, mercury, and salinity (Cl, SO4, or TDS) (US EPA Region 6 2006; TCEQ 2007). Under a salt TMDL, an irrigation area is assigned with a load allocation that shall not be exceeded, stressing the importance of determining the actual salt contribution of the area in relation to its salt load allocation. The program recommends having modeling tools that allow for identifying this load under different management scenarios that will comply with the load allocation, or to determine which practices would lead to noncompliance.

923

924

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The European Water Framework Directive (EU 2000) demands the member states to ensure by 2010 “an adequate contribution of the different water uses, disaggregated into at least industry, households and agriculture, to the recovery of the costs of water services.“ Thus, the salt load contribution of irrigated agriculture needs to be established accurately so as to assign to agricultural uses their share in the recovery of environmental costs of the degradation in water quality induced by (irrigated) agriculture. One of the most effective ways to reduce salt loading in IRF is to limit subsurface drainage from irrigated lands. This emphasizes the need to determine IRF quantity and quality from irrigated areas. Hornsby (1973) classified IRF into two categories: surface and subsurface return flows. Surface return flows, if irrigation is properly managed, retain approximately the same salinity level as the applied irrigation water. In contrast, subsurface return flows, which percolate through the rootzone, typically contain levels of salinity that have increased due to evapotranspiration (ET) and soil chemical weathering processes (Aragüés and Tanji 2003). The latter include cation exchange reactions, mineral precipitation-dissolution reactions, dissolution and leaching of fertilizers and soil amendments, ion association, temperature changes that affect the reaction rates, oxidation-reduction reactions, and interaction with groundwater. Figure 30-1 shows an idealized drawing of the IRF system. Water is diverted from a river for irrigation, applied to agricultural fields, and returned as drainage water to the river. IRF includes the water delivery subsystem’s canal seepage and operational spills, and the farm subsystem’s surface and subsurface drainage. IRF is one component of the water balance equation when applied to an irrigation district. Scofield (1940) introduced the concept of salt balance (account of salt inputs minus salt outputs) for an irrigation system to assess salt leaching from irrigation areas—a pioneer step to account for in-situ sustainability. Since then, salt balances have been valuable tools for (1) establishing the long-term sustainability of irrigated areas (Sharma 1999), (2) establishing salt loads in IRF and identifying unaccounted sources and sinks of water and salts (Kaddah and Rhoades 1976; Tedeschi et al. 2001), and (3) assessing soil salinization processes (van Rensburg et al. 2008). An uptake-weighted salt balance in the rootzone is the key to establishing rootzone salinity and determining the traditional leaching requirement of crops (Ayers and Westcot 1985). Thayalakumaran et al. (2007) reviewed the application of the salt balance concept at various spatial scales (rootzone, farm and subsurface drainage, irrigation area, and hydrologic basin) and concluded that maintaining the salt balance is necessary to keep acceptable soil salinity levels, whereas keeping the water balance at larger scales is the key strategy for minimizing offsite impacts.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

925

FIGURE 30-1. Idealized drawing of irrigation return flow system. From Skogerboe and Walker (1981) with permission from Marcel Dekker. Salt balance equations provide the basis for the most simple salinity models for irrigation areas (Eq. 30-1). A general salt balance equation is given by CiVi  CdVd  Mhc  (Msp  Msd)  Mis  Mfs  Mpre  Mafm  Mpr  Miwr  Mdp  0

(30-1)

where Vi and Vd  volumes of irrigation and drainage water, and Ci and Cd their concentrations; Mhc  salt removal with the harvested crops; (Msp  Msd)  salt pickup  salt deposition; Mis  initial amount of salts in the soil solution; Mfs  final amount of salts in the soil solution; Mpre  salt mass in precipitation or snow; Mafm  salts contributed by soil amendments and fertilizers; Mpr and Miwr  salts picked up by irrigation or precipitation runoff; and Mdp  salt mass in deep percolation waters. Ion exchange is not considered explicitly, since it produces a change in the ionic composition rather than a change in concentration. When natural mineral deposits are present or chemical amendments are applied, the most important sinks and sources of salt generally are Mdp, Mis, and Msp  Msd. Despite its simplicity, the terms of the salt balance are often difficult to quantify, thus requiring the help of modeling tools. The use of models is also needed to assess the influence of changes in climate or management practices on IRF. Hydrosalinity models range from a simple, mass balance approach that often aggregates inputs and outputs over a wide time span (conceptual, steady-state models), to much more complicated routines that consider

926

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the individual processes that may contribute to the salinity of IRF, generally with small input time steps (transient soil water and reactive transport models). In recent decades there has been an increase in the number of soil water and reactive transport models. Tanji and Kielen (2002) analyzed the three main groups of modeling approaches (i.e., mechanistic, empirical, and conceptual) and the various rootzone hydrosalinity models ranging from simple conceptual to complex scientific. They also described the basic concepts behind various hydrosalinity models such as LEACHM, UNSATCHEM, HYDRUS, SWAP, CIRF, SALTMOD, and SAHYSMOD. Some of these models are presented or discussed in other chapters of this manual. At the core of any physically based salinity model for IRF, the salt balance of the rootzone must be considered, since evapoconcentration due to ET and mineral weathering are the leading processes contributing to salinity in drainage and in IRF (El-Ashry et al. 1985). However, IRFs are also affected by other flows such as operational spills from distribution conveyances, canal seepage, and plot tailwaters (i.e., surface runoff from irrigated fields) (Fig. 30-1). Also, drainage waters may mobilize salts from below the rootzone (Christen et al. 2001; Smedema and Shiati 2002), affecting salt loading in IRF. Therefore, soil water modeling is the main but not the only process involved in modeling the off-site effects of irrigated agriculture. Complex (Transient Soil Water and Reactive Transport) Models Bastiaansen et al. (2007) reviewed soil water models and highlighted the relevant and recent developments of these kinds of models, reflected both in the number of models available and in the growing number and complexity of the modeled processes. They also emphasized their underutilization in real-world irrigation and drainage systems and the fact that usually a given model is focused in a particular process and goal (crop production, drainage design, pesticide or nitrogen movement, reactive transport, ET calculation, etc.). The main weaknesses and threats for the application of these complex models to irrigated areas are: 1. Aggregation of point results to the whole irrigated area—a difficult task due to the inherent spatial heterogeneity in soil properties and input parameters and the presence of lateral flows. This weakness may be overcome by (a) coupling non-saturated, unidimensional flow and transport models to regional groundwater models (Schoups et al. 2005; Burkhalter and Gates 2006; Guganeshharajah et al. 2007), and (b) applying the soil water and salt models to several hydrologic units and integrating the results over a wider area by means of distributed basin models and/or GIS applications. Generally, available basin scale models

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

927

(AGNPS, ANSWERS, SWAT, SWRRB, APEX) have been developed for or applied to modeling of nonpoint source pollution from agriculture (erosion, nutrient, and pesticide transport), and not directly to salt exports, although there have been some uses of basin-scale models and GIS to watershed mass balance (Richards and Kump 1997) and IRF (Burkhalter and Gates 2006) studies. 2. Collection of data required for model application (Bastiaansen et al. 2007). Generally, irrigation district managers and practicing engineers have neither the time nor the resources to gather the data needed for sophisticated models. Often, soil sampling programs are not developed to obtain the soil properties needed by this type of model. Simpler models may make use of more readily available soil information. 3. Training of the potential users (irrigation district managers, engineers) in the use of complex models (Bastiaansen et al. 2007). In this regard, there is also room for the use of simple, salt balance-type models. 4. Difficulties in the determination of the adequate lower boundary conditions for the application, which may result in an overparameterization (Bastiaansen et al. 2007). Generally, salt balance models require fewer and easier-to-determine parameters and account for the shallow groundwater–unsaturated zone interaction in a semiempirical manner, or can skip this interaction. Simple, Conceptual (Steady-State, Mass Balance) Models Conceptual hydrosalinity models are based on the principle of continuity of mass (Eq. 30-1). They typically model salinity as a reactive parameter through chemical equilibrium, which considers only the difference between initial and final states (steady-state models). Most chemical reactions occur rapidly enough that attaining chemical equilibrium is relatively easy, especially in systems with small water-flow rates. Chemical kinetic equations are used as necessary, such as to describe the precipitation of lime and dissolution of gypsum. The advantages of using these conceptual models are that they (1) take under consideration management factors that are not easily accounted for in more complex models, (2) use sparse or less intensive data and obtain simple answers from very limited inputs, (3) compare in an easy manner (usable by not only highly trained professionals) the different outputs or environmental behavior of different irrigation systems, (4) get quick, rough results of how the outputs will change after variations in the irrigation system (sensitivity analysis), and (5) perform long-term analyses and scenarios based on the changes of bulk variables (such as irrigation efficiency) without regard to the actual processes leading to those changes.

928

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The outputs of these conceptual models, though based on mass balance equations and physical processes, are less rigorous than those obtained with complex models; to obtain reliable results, the models should be carefully calibrated and validated for each area of application (van Rensburg et al. 2008). For example, salt loads in IRF, or average soil salinity estimates, may not be as accurate as those obtained with scientific models, but the data demand is much lower, and complex processes, such as a shift in crop patterns or distribution systems or the reuse of drainage waters for irrigation, may be accounted for easily. Applications of Hydrosalinity Models Some recent applications of regional salt balance studies have focused on the long-term sustainability of irrigated systems, establishing best management practices, or simply accounting for irrigation contribution to salt load. SALTMOD (Oosterbaan 2000) is one of the most-used irrigation area models. This model performs water and salt balances (not accounting for mineral weathering or salt dissolution) in up to four seasons in a year, thus allowing for a better definition of IRF along time. It works upon user-specified areas of similar cropping, drainage, or irrigation patterns. This leads to results better-adjusted to actual agricultural practices and local conditions. The final step of this model is to aggregate the results for the whole area. SALTMOD has been applied in several irrigated areas, not generally as a predictor of IRF quality but rather to help in the design of drainage systems in Turkey (Bahçeci et al. 2006) and India (Singh et al. 2002), or in the management of irrigated areas (Bahçeci and Nacar 2007). In Australia, Hornbuckle et al. (2005) developed the Tiddalik model to specifically address the contribution of irrigation return flows to river systems. The model is applied to several cropping units (user-defined spatial discretization) and soil layers, performs a bucket-type water balance, and calculates drainage with the Hooghoudt equation. Tiddalik does not track salts as they move in the system (evapoconcentration or dilution effects), and the concentrations of each flow path are assigned based on actual knowledge. This salt transport component is yet to be developed. Burkhalter and Gates (2006) evaluated several management scenarios (reduced recharge, reduced seepage, increased pumping of groundwater, varying drain spacing, and its combinations) to the salinity problem in the lower Arkansas River. They used MODFLOW to model groundwater flow and M3TDMS to model the unsaturated zone. The aforementioned scenarios were evaluated in terms of changes in soil salinity and water table depth (closely linked in the system), groundwater salinity, relative crop yields, consumptive water use, and salt loads to the river, thus modeling implicitly the subsurface IRF salt load. It is noteworthy that the

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

929

models were calibrated and validated with extensive field data in a process that allowed estimating the main fluxes in the system and establishing the nature of the problem (Burkhalter and Gates 2005). Schoups et al. (2005) developed a hydrosalinity model to integrate salt transport with subsurface hydrology to assess the long-term sustainability of irrigation in the San Joaquin Valley (SJV) of California. The model permitted the reconstruction of salinity records for the last 60 years, covering the initial period of groundwater irrigation, the import of surface irrigation water from the northern Central Valley, and the beginning of the salinity and selenium toxicity problems in the SJV. They found two main problems questioning the sustainability of the system: on-site, or the steady accumulation of about 5  105 Mg salt/year in the irrigated soils due to gypsum dissolution, and off-site, or the salinization of the deeper aquifers due to recharge from subsurface IRF. Guganesharajah et al. (2007) developed the coupled surface and groundwater model HYDRO-GW to assess the regional and temporal variation in groundwater and soil salinity levels under selected irrigation and drainage scenarios, and applied it to the Kashkadarya region in Uzbekistan. Among other variables, the model considers seepage losses from distribution and field irrigation canals that are substantial in this area (i.e., overall conveyance efficiency for the distribution systems is estimated as 72%). Along with HYDRO-GW, Guganesharajah et al. (2007) applied a simplified mass balance model for salts. The results of the simplified model for rootzone and groundwater salinity for different scenarios followed patterns similar to the complete model but with quite higher values. Salt balance-type models are usually steady-state models, performing a season- or year-aggregated salt balance. These steady-state models yield time-averaged results for the quality of IRF regardless of the distribution in time of these flows and their concentrations. Letey and Feng (2007) emphasized the advantages of transient models for soil water and salt dynamics over steady-state approaches in establishing actual soil salinity levels for crop production and leaching requirements (LRs), as transient models give lower LRs than those calculated by the traditional FAO method (Ayers and Westcot 1985). In a similar study, Corwin et al. (2007) concluded that steady-state (WATSUIT) or transient (UNSATCHEM) models that account for salt dissolution and mineral precipitation in the soil predict quite similar LRs and that these estimates are lower than those predicted by other transient models that do not incorporate this chemistry routine. Their results suggest that under certain conditions, accounting for salt precipitation may be more important than accounting for temporal variability. Schoups et al. (2006) utilized the transient model UNSATCHEM to determine soil salinity and drainage salt load. They undertake several

930

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

simplifications in the inputs and model operation (annual average boundary conditions, only cation exchange processes and gypsum dynamics in the soil chemistry module, and coarser spatial discretization) over a 10-year simulation period. They established that time-averaging the boundary conditions yielded differences with the full model lower than 20% (except if the time-averages eliminated short-term water table fluctuations). The approach is relevant because it suggests the use of annually averaged inputs when seeking long-term results (reducing the work load and the data requirements). The testing of a model should incorporate various steps: (1) calibration, which involves fine-tuning the model’s coefficients and parameters so that the computed data fit more closely to the observed data; (2) validation, which involves testing the calibrated model with observed data other than the set previously used for calibration; and (3) sensitivity analysis, which determines the relative importance of the various input parameters on the model’s output. Without this step process for the area of application, the model may give unsatisfactory results.

DESCRIPTION OF CIRFLE HYDROSALINITY MODEL This section describes a conceptual irrigation return flow hydrosalinity model developed by Tanji (1977), revised by Aragüés et al. (1985, 1990) with the name CIRF, and updated by Quílez (1998) with the name CIRFLE (Conceptual Irrigation Return Flow hydrosalinity model with consideration for the Leaching Efficiency of salts). The description is conceptual and is focused on the main inputs and outputs of the model. The mathematical equations and details of the internal variables and parameters can be found in the tutorial accompanying the computer program (CITA 2011). CIRFLE focuses on the crop’s rootzone and considers only the main flow pathways of water and salts. The model assumes that the masses of water and salt are conservative and that steady-state conditions can model long-term transient conditions approximately. All of the model’s spatial and temporal variables and parameters are taken as average values. Therefore, the model can be applied to large systems, including irrigation districts, and for long periods of time, such as an irrigation season, a hydrologic year, or a series of consecutive years. CIRFLE consists of a hydrologic submodel coupled to a salinity submodel. Hydrologic Submodel Figure 30-2 shows an idealized representation of the hydrology and flow pathways in a typical irrigation project, with particular focus on the

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

931

FIGURE 30-2. Freebody diagram of hydrologic submodel, focusing on crop rootzone. The symbol Q denotes quantity of water. rootzone. With the continuity equation, the rate of change of rootzone water storage can be described as n m dQs  ∑ Qi  ∑ Qo dt

(30-2)

where Q  quantity of water and t  time (hydrological year, irrigation season); subscripts s, i, and o respectively denote storage, inputs, and outputs from the crop’s rootzone; and n and m the number of inputs and outputs. The hydrologic inputs are defined by n

∑ Qi  Qdiw  Qp  Qrim

(30-3)

where the subscripts diw, p, and rim are diverted irrigation water, precipitation (rain or snow), and rim inflows from adjacent systems, respectively.

932

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The hydrologic outputs are defined by m

∑ Qo  Qet  Qevdiw  Qevp  Qdp  Qpro  Qiwro  Qsdw  Qrim

(30-4)

where Qet, Qevdiw, Qevp, Qdp, Qpro, Qiwro, and Qsdw are ET of soil water to the atmosphere, evaporation of irrigation water, evaporation of precipitation, deep percolation from rootzone, precipitation runoff, irrigation runoff, and collected subsurface drainage water (including tile drainage effluent and seepage into natural and manmade channels), respectively. If tile drainage effluents and surface runoff are collected in a single drain, the last four terms in the equation constitute the surface IRF system. The net change in water storage in the rootzone is given by dQs  Q fsw  Qisw dt

(30-5)

where isw and fsw denote initial and final stored soil water, respectively. A portion of the initial stored soil water and the effectively infiltrated irrigation and precipitation (Qsw) is evapotranspired, and the rest (Qpsw) goes into final stored soil water (Qfsw) and water available for subsurface drainage and deep percolation (Qppsw) (Fig. 30-2). From a management standpoint, several variables linked to the hydrologic subsystem are defined: • Evapotranspiration concentration factor (ETCF), which gives the proportion of soil water before ET (Qsw) to soil water after ET (Qpsw): ETCF 

Qsw Qpsw

(30-6)

• Leaching fraction (LF), the portion of the infiltrated irrigation and precipitation that percolates through the rootzone: LF 

Qppsw Qeaiw  Qep

(30-7)

• Water use efficiency (WUE), the portion of the infiltrated irrigation and precipitation that undergoes ET: WUE 

Qet Qeaiw  Qep

(30-8)

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

933

• Irrigation application efficiency (Eiae, a required input to the model), the proportion of diverted irrigation water that effectively infiltrates the soil: Eiae 

Qeaiw Qdiw

(30-9)

Salinity Submodel Figure 30-3 is a freebody diagram that describes salinity and its flow pathways. The symbol C represents total dissolved solids (TDS) and M the mass of salts obtained from the product of salt concentration (TDS) and water volume. Based on conservation of salt mass, salinity is described with j k dMs  ∑ Mi  ∑ Mo dt

(30-10)

where dMs/dt is the rate of change in mass of salts stored in the crop’s rootzone; and l and k are the number of inputs and outputs, respectively.

FIGURE 30-3. Freebody diagram of salinity submodel, focusing on crop rootzone. The symbols C and M denote total dissolved solids and mass of salts, respectively.

934

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The salt inputs are defined by j

∑ Mi  Mdiw  Mp  Mrim

(30-11)

where all subscripts have been defined in the hydrologic submodel section. Salt outputs are defined by k

∑ Mo  Mdp  Mpro  Miwro  Msdw  Mrim

(30-12)

where all subscripts have been previously defined. The rate of change in mass of salt stored in the crop’s rootzone, including sources and sinks, is also defined by dMs  M fsw  Misw  Msd  Msp  M gsp dt

(30-13)

where Msd denotes lime (CaCO3) and gypsum (CaSO4 2H2O) precipitation as the soil solution is concentrated by ET, and Msp and Mgsp denote, respectively, lime and gypsum dissolution from the soil. Misw represents the initial soluble salts other than gypsum, and Mfsw the final soluble salts present in the soil solution. It is assumed that waters flowing over the soil’s surface may pick up salts, as is reflected by adding Ciwrosp and Cprosp to Cdiw and Cp, respectively, in calculating Miwro and Mpro (Fig. 30-3). The salt concentration (TDS) of the initial stored soil water (Cisw) is corrected for the solubility of gypsum when it is present in the soil, because the TDS due to gypsum cannot be significantly concentrated during ET due to its limited solubility (2.63 kg/m3 or 2.2 dS/m in deionized water). Although this approach has limitations because gypsum solubility depends on other soil variables (such as soil sodicity or the composition of the soil solution), it is preferred over that where it will be allowed to freely evapoconcentrate due to ET. The volume-weighted average of effective applied irrigation water (Ceaiw), effective precipitation (Cep), and initial soil water (Cisw) corrected for the solubility of gypsum, gives the TDS of soil water (Csw), which is then concentrated by an ET factor (ETCF) to account for water lost in ET.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

935

CIRFLE computes the term Msp  Msd (salt pickup  salt deposition, or mineral dissolution  mineral precipitation) with WATSUIT (described in Chapter 25 of this manual) that calculates the extent to which applied irrigation water, as it becomes concentrated in the soil solution due to ET, effectively dissolves CaCO3 from the soil or precipitates out CaSO4 2H2O and CaCO3. The required input data by WATSUIT include the chemical composition of the irrigation water and the LF. A linear regression equation that relates Csp  Csd and LF is introduced in the model to account for this effect. For simplicity, values of the intercept and slope of this equation are given for representative irrigation water compositions and electrical conductivity (EC) values, although users may introduce their own empirical values if desired (see CIRFLE computer program for details). Salt pickup due to gypsum dissolution (Mgsp) is treated separately in the model. It is assumed, for this purpose, that the solubility of gypsum is 2.63 kg/m3 and that gypsum will dissolve by the amount of water in the soil after ET (Qpsw, Fig. 30-2) until saturation or until its supply is exhausted. The concentration and mass of soil water after evapotranspiration losses (Cpsw and Mpsw), adjusted by gypsum dissolution (Mgsp) and salt pickup minus salt deposition (Msp  Msd) gives the final salt concentration and mass in the soil solution (C psw and M psw, Fig. 30-3). CIRFLE estimates the TDS of final stored soil water (Cfsw) from the TDS of initial stored soil water, a leaching efficiency coefficient (k), and the amount of percolating water per unit rooting depth, using the empirical approach developed by Hoffman (1986). This is a major modification over the CIRF model given in Aragüés et al. (1990). The empirical coefficient k takes into account the inefficiency of salt leaching that depends on soil physical and chemical characteristics such as pore size distribution, bypass flows, and soil water content, among other factors. Typical k values for continuous ponding irrigation are around 0.1 for sandy loam soils (i.e., high leaching efficiency), 0.3 for clay loam soils, and 0.45 for peat soils (i.e., low leaching efficiency), whereas k is around 0.1 and is independent of soil texture in intermittent ponding or sprinkler irrigation. After the model calculates the concentration and mass of salts in final stored soil water (Cfsw and Mfsw, Fig. 30-3), the soil solution is subdivided into two components of equal salt concentration Cppsw: deep percolation (Cdp) and collected subsurface drainage (Csdw). The mass of salts in deep percolation (Mdp) and subsurface drainage (Msdw) is then calculated from these concentrations and the corresponding amounts of water (Qdp and Qsdw). The concentration of surface irrigation return flow (Csirf) is the volume-weighted average of the concentrations of subsurface drainage (Csdw), runoff components (Cpro and Ciwro), and lateral contributions (Crim).

936

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

From a management standpoint, one variable linked to the salinity subsystem is defined, the mass of salts in IRF per unit irrigated area (Mua, Mg/ha): Mua 

Msirf A

(30-14)

MODEL CALIBRATION AND VALIDATION The original model by Tanji (1977) was first calibrated with data from the Glenn-Colusa Irrigation District (California) for the 1973 irrigation season. The water and soils in this district are low in dissolved mineral salts. After calibration, the modeled flow, TDS, and salt mass in the irrigation return flows were within 2% of the observed data (Aragüés et al. 1990). The calibrated model was then validated with data from the 1970 irrigation season. The model’s estimates were within 33% and 23% of the observed values (Table 30-1).

TABLE 30-1. Results of CIRFLE Validationa Irrigation Return Flow (IRF) (1)

Volume IRF (ha-m) Measured Estimated % Differencec TDS IRF (mg/L) Measured Estimated % Differencec Salt mass IRF (Mg) Measured Estimated % Differencec a

GC (1970)b (2)

PA (1975) (3)

VID (1984) (4)

BID (1992) (5)

VID (2006) (6)

34,512 23,127 33

5,585 6,552 17

4,307 4,282 1

18,749 18,418 2

1,715 1,455 15

192 236 23

1,955 1,818 7

1,843 2,071 12

675 762 13

1,982 2,280 15

66,305 54,614 18

109,267 119,191 9

79,391 88,656 12

126,519 140,407 11

33,994 33,121 3

Irrigation Districts: GC, Glenn-Colusa; PA, Panoche; VID, Violada; BID, Bardenas I. The hydrological year of validation is given in parentheses. b For GC, the values are for the irrigation season. c 100  (Estimated  Measured)/Measured TDS, total dissolved solids

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

937

The model was then applied to a second irrigation project, the Panoche Drainage District, which contains salt-affected soils. The estimated volume of surface IRF was within 17% of the measured value, but the TDS and mass of salts in surface irrigation return flows were greatly underestimated (Aragüés et al. 1990) because the original model did not consider the presence of native soil salts, particularly gypsum. When salt contributions from gypsum were considered, the calculated results for TDS and mass of salts in the irrigation return flows were, respectively, within 7% and 9% of measured values (Table 30-1). Aragüés et al. (1985) modified Tanji’s 1977 model and applied it to Violada Irrigation District (VID) in the middle Ebro River basin of Spain. The soils in VID are high in gypsum and low in soluble salts. The modified model (CIRF) added routines that describe chemical weathering of soil gypsum, lateral contributions from adjacent areas, and initial and final stored soil water. CIRF was calibrated and validated in VID. Faci et al. (1985) describe in detail the study area and the methods used to collect data and estimate parameters. The model was calibrated for the 1982 hydrological year. After calibration, the predicted values for volume, concentration, and salt load in IRFs were within 2% of measured values (Aragüés et al. 1990). The calibrated model was validated with data from hydrological years 1983 and 1984. The results of model validation show that model predictions were in both years within 3% and 13% of the observed values. Table 30-1 gives the results for the 1984 hydrological year. The application of CIRFLE for the 2006 hydrological year, a dry season where irrigation water was only 544 mm as compared to values of about 1.000 mm in the 1980s, produced error estimates within 15% of measured values (Table 30-1). These satisfactory results for both normal and dry years give reliability to CIRFLE for simulating IRF in VID. Quílez (1988) updated CIRF to account for the inefficiency of the displacement of salts in the soil. The modified model (CIRFLE) was applied to the 46.5-ha Bardenas I Irrigation District (BID) in the middle Ebro River basin of Spain. The first step in model application was to divide the district into eight homogeneous areas based on irrigation (irrigated vs. nonirrigated land) and soil salinity (ECe in the ranges 2 dS/m, 2–4 dS/m, 4–8 dS/m and 8 dS/m). Soil salinity was mapped using a Geonics EM38 electromagnetic sensor that was calibrated against ECe. Based on the performed soil survey, it was considered that gypsum was present in areas with ECe 4 dS/m. Different values of the leaching efficiency coefficient (k) were assigned to each soil salinity interval to take into account that salt leaching becomes less effective as salts are leached out (Hoffman 1986). Thus, the k values were 0.3 for ECe 8 dS/m, 0.4 for ECe  4–8 dS/m, 0.6 for ECe  2–4 dS/m, and 0.8 for ECe 2 dS/m. The model estimates for BID were within 2% and 13% of observed values (Table 30-1), a

938

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

satisfactory result taking into account the simplifications performed in delineating the homogeneous areas. CIRFLE was further validated by comparing the TDSmean calculated for each ECe homogeneous area (Table 30-2) with the mean annual TDS measured in waters draining areas within BID with similar ECe ranges. The calculated TDSmean for ECe 2 dS/m was 446 mg/L, as compared to a measured average TDS of 388 mg/L (standard deviation  64 mg/L) in drainage waters from five areas with ECe 2 dS/m (error estimate  15%). Corresponding calculated TDSmean and measured TDS were 866 and 755 mg/L (one area with ECe  2–4 dS/m) (error estimate  15%), and 2.241 and 2.301 mg/L (standard deviation  184 mg/L) (three areas with ECe  4–8 dS/m) (error estimate  3%). The modeling results demonstrate that CIRFLE could properly simulate the average water and salinity IRFs in the aforementioned irrigated areas.

SENSITIVITY ANALYSIS For sensitivity analysis, one of the variables is varied while simultaneously keeping all other variables constant. Model results are compared. If outputs differ significantly, the variable is said to be sensitive. Sensitivity analysis provides valuable information on the behavior of the study system and how the variables interact with each other. Table 30-3 summarizes results of the sensitivity analysis for VID using data for hydrological year 1982. The order of sensitivity of the hydrological inputs for Qsirf (volume of IRFs) is Qdiw  Qet  Qp  Qrim, and all the variables behave linearly. Analyzing for Csirf (concentration of IRFs), Fig. 30-4 demonstrates that the sensitivity behavior of some variables is not linear. An extreme example is gypsum content, which has no effect on Csirf when present in amounts that result in a soil solution saturated with gypsum, but exerts a sharp influence when present in amounts small enough to cause the solution to become undersaturated. Irrigation application efficiency (Eiae) strongly affects Csirf (Fig. 30-4), as it establishes the partitioning between runoff water of low salt concentration and subsurface drainage water of higher salt concentration. Csirf decreases with decreases in diverted irrigation water (Qdiw) and precipitation (Qp), and increases in ET (Qet), because the amount of water available for gypsum dissolution decreases. However, in irrigation districts without gypsum or other types of soluble salts, the concentrations of subsurface drainage and return flows will tend to increase with decreases in Qdiw or Qp and increases in Qet due to the ET concentration effect. The combined effects of the inputs soil salinity (ECe), leaching efficiency coefficient (k), presence or absence of gypsum, and volume of irri-

TABLE 30-2. Results of CIRFLE Application for Bardenas I Irrigation District (BID) Using Data from Hydrological Year 1992a

Model Output for IRF (1)

Volume (ha-m) Salt mass (Mg) TDS (mg/L) TDSmean (mg/L)

2 dS/m  ECe  4 dS/m

4 dS/m  ECe  8 dS/m

ECe  2 dS/m

NI (3)

I (4)

N (5)

I (6)

NI (7)

I (8)

NI (9)

Total (10)

345 3,191 931

7,396 60,020 811

284 6,453 2,266

1,253 26,248 2,096

48 2,898 6,033

102 2,819 2,762

4 361 9,220

18,418 140,407 762

ECe  2 dS/m I (2)

8,987 38,415 428 446

a

866

2,241

3,001

760

The district was divided into eight homogeneous areas based on four ECe intervals and the irrigated (I) and nonirrigated land (NI) within each ECe interval. TDSmean is the volume-weighted average TDS for each ECe interval.

IRF, irrigation return flow; TDS, total dissolved solids.

939

940

TABLE 30-3. CIRFLE Sensitivity Analysis for Violada Irrigation District (VID) with Qualitative Classification of Input Parameters According to Their Influence on Outputsa Input Parameter Model Output for IRF (1)

Qdiw (2)

Eiae (3)

Qp (4)

Qisw (5)

Qet (6)

Qrim (7)

Crim (8)

Cdiw (9)

Cp (10)

ECe (11)

SP (12)

Gyp (13)

Volume (Qsirf) TDS (Csirf)

3 2

0 3

3 2

0 0

3 2

1 1

0 0

0 1

0 0

0 3

0 0

0 2

Salt mass (Msirf)

3

3

3

3

3

1

0

1

0

3

0

2

a

Plus signs indicate that the output increases as the input increases. Minus signs indicate that the output decreases as the input increases.

0: Linear response; input variation of 100% gives output variation less than 5%. 1: Linear response; input variation of 100% gives 5% to 20% output variation. 2: Nonlinear response; for small input variations, output variation as 1; for large input variations, output variation as 3. 3: Linear response; input variation of 100% gives output variation greater than 20%. IRF, irrigation return flow; TDS, total dissolved solids. Q, volume of water; C, salt concentration; and M, salt mass in sirf, surface irrigation return flows; diw, diverted irrigation water; p, precipitation; isw initial soil water; et, evapotranspiration; and rim, rim inflows-outflows. Eiae, irrigation application efficiency; ECe, electrical conductivity of soil saturation extract; SP, saturation percentage; Gyp, soil gypsum content.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

941

FIGURE 30-4. Sensitivity analysis: effect of different variations of input variables (%) on TDS of irrigation return flows (Csirf). gation water (Qdiw) on the outputs salt concentration (Csirf) and salt mass (Msirf) in irrigation return flows are explored for a generic semiarid irrigation district. As expected, soil salinity is the input variable with the larger effect on the salinity of irrigation return flows (Csirf and Msirf) (Fig. 30-5a). Thus, mapping of soil salinity (and soil gypsum, if present) is critical to properly estimate salinity in IRFs. The salt concentration and mass in subsurface drainage waters and in IRFs decrease with increasing k values because of the lower leaching efficiency of salts, especially in soils high in soluble salts and without gypsum (Fig. 30-5b). Decreasing the diverted irrigation water (Qdiw) from 1,000 to 800 mm increases Csirf, especially for low k values and soils without gypsum. Since the volume of subsurface drainage water (Qsdw) decreases with decreasing Qdiw, the mass of salts in irrigation return flows (Msirf) from areas with gypsum also decrease, whereas it remains practically unchanged in areas without gypsum (Fig. 30-5b). Minimizing subsurface drainage through an efficient water management is therefore a key strategy for controlling the export of salts from salt-affected and/or gypsum irrigated areas. Thus, as indicated in a recent review by Thayalakumaran et al. (2007), managing the water balance is likely to be more effective in controlling off-site impacts of irrigation than managing the salt balance.

942

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 30-5. Combined effects on TDS (Csirf) and salt mass (Msirf) of irrigation return flows of (a) soil salinity (ECe) and leaching efficiency coefficient (k), and (b) volume of irrigation (Qdiw), leaching efficiency coefficient (k) in the presence or absence of gypsum.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

943

The aforementioned simulations emphasize the importance of performing a sensitivity analysis for each particular study area, since it provides a way to rank the more important measurements and estimates, and gives insights for a better understanding of the system.

MODEL APPLICATION An advantage of the CIRFLE model is that it only requires 25 input data and model coefficients and yet apparently has the characteristics needed to predict salt loading of IRFs. The sensitivity analysis of CIRFLE proved that the key to minimizing salt loading in IRFs from Violada (VID) and Bardenas I (BID) Irrigation Districts was to decrease the applied irrigation water and, thus, reduce subsurface drainage water. This option is clearly viable, as is shown by the irrigation districts’ low to medium water use efficiencies (WUEs), an average of 0.45 for 1982–1984 VID years and 0.79 for BID. Concurrently, calculated LFs were high in VID (0.51) and moderate in BID (0.21). Considering the low salinity of irrigation water in these districts (EC 0.4 dS/m) and the crops grown in the area, their theoretical leaching requirements are one order of magnitude lower than these LFs. CIRFLE was applied in VID and BID to evaluate three water management alternatives for salinity control: (1) reduction in applied irrigation water, (2) change from surface to pressure irrigation systems, and (3) reuse of drainage waters for irrigation. Table 30-4 gives the model estimates for VID in 1983 and the corresponding model predictions assuming that salt control practices resulted in hypothetical reductions of 10% and 20% in diverted irrigation water (Qdiw). Even though the ET concentration factor (ETCF) increased by 5% to 12% when Qdiw decreased by 10% 20%, the final soil water TDS remained unchanged because it was gypsum-saturated. Hence, the TDS of percolating waters in VID is basically independent of its volume, and salt loading (Msirf) can be approximated from knowledge of the water balance only. Obviously, if soil gypsum had been absent or more readily soluble salts had been present in the soil profile, the TDS would have increased due to ETCF. Thus, a similar analysis in the Bardenas I Irrigation District predicted that the TDS of subsurface drainage waters (Csdw) will increase by 11% and 41% for 10% and 20% decreases in Qdiw, respectively. The mass of salts in irrigation return flows (Msirf) was substantially reduced when Qdiw decreased because the corresponding decrease in the amount of water percolating through the rootzone reduced the mass of dissolved gypsum. Thus, the calculated Msirf was 21.2 Mg/ha in 1983 and decreased by 14% and 27% when irrigation water decreased by 10% and

944

TABLE 30-4. CIRFLE Estimates in Violada Irrigation District (Year 1983) and CIRFLE Predictions for Hypothetical 10% and 20% Reductions in Diverted Irrigation Water (Qdiw). Qdiw  10% % Differencea (4)

Year 1983 (2)

Diverted irrigation water (m) Final soil-water TDS (mg/L) Salt pickup-salt deposition (mg/L) Salt mass due to gypsum (Mg/ha) Volume IRF (Qsirf) (m)

1.13 2,824 38.5 25.0

1.02 2,829 43.1 22.3

10 0 12 11

0.91 2,836 48.6 19.6

20 0 26 22

0.91

0.80

12

0.69

24

2,330 21.2 0.52 1.72

2,291 18.3 0.48 1.81

2 14 7 5

2,244 15.4 0.43 1.92

4 27 17 12

0.48

0.52

8

0.57

19

Salt mass IRF (Msirf) (Mg/ha) Leaching fraction (LF) ET concentration factor (ETCF) Water use efficiency (WUE) a

100  (Predicted  Year 1983)/Year 1983

ET, evapotranspiration; IRF, irrigation return flow; TDS, total dissolved solids

Predicted (5)

% Differencea (6)

Variable (1)

TDS IRF (Csirf) (mg/L)

Predicted (3)

Qdiw  20%

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

945

20%, respectively. A similar analysis for the irrigated land in BID shows that, even though Csdw increased as indicated previously, Msirf decreased by 20% and 32% for 10% and 20% decreases in Qdiw, respectively. Table 30-4 shows that another consequence of reducing Qdiw is the precipitation of salts in the soil profile due to a reduced leaching fraction. Researchers at the U.S. Salinity Laboratory (Bernstein and Francois 1973) postulated the concept of “minimum leaching fraction” as a way to reduce salt loading in IRFs. Its applicability to irrigation waters similar to those of VID is expressed by the equation Csp  Csd (mg/L)  102  122.9 LF (at PCO2  3.5  104atm)

(30-15)

which shows that the amount of salts precipitating at 10% LF almost doubles that at 40% LF. These LF reductions can be achieved with proper water management in VID without affecting crop yields or soil’s structural stability because of the gypsum-saturated, low soil water SAR. A second CIRFLE application carried out in BID simulated a change from actual surface irrigation systems to future pressure irrigation systems. This simulation provided information on the advantages and limitations of this projected irrigation modernization in BID. It was anticipated that this change from surface to pressure irrigation systems would increase the irrigation application efficiency (Eiae) from 0.65 to about 0.8. In addition, since the leaching requirements of the crops grown in BID are negligible, a target LF of 0.1 was imposed with a concomitant decrease in Qdiw from 1.11 to 0.70 m (Table 30-5). This low LF was not attainable in the old surface-irrigated schemes of BID but is feasible in modern sprinkler systems, as shown by Tedeschi et al. (2001) for the new sprinkler-irrigated schemes of Monegros II (Ebro River Basin, Spain). CIRFLE predicts that the TDS of BID IRFs (Csirf) will increase by 35% (from actual 806 mg/L to predicted 1,092 mg/L). In contrast, the IRF volume (Qsirf) and salt mass (Msirf) will decrease, respectively, by 60% and 46% (Table 30-5). Although these figures could vary depending on Eiae, LF, and other variables, they show that the modernization in BID will substantially reduce salt loading in IRFs without inducing detrimental salt concentrations in these returns. Finally, another management strategy analyzed with CIRFLE was the reuse of return flows for irrigation as a way to reduce salt loading in IRFs. Table 30-6 shows model predictions for 1983 VID and 1993 BID (2  ECe (dS/m)  4 homogeneous area) assuming that 25% of the applied water comes from reused IRFs. The mixing of the canal irrigation water with these IRFs in this proportion produced irrigation waters low to moderate in TDS (965 mg/L in VID and 333 mg/L in BID) and relatively high in

946

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

TABLE 30-5. CIRFLE Estimates for Actual Surface Irrigation Systems in Bardenas I Irrigation District (Year 1993) and CIRFLE Predictions for a Hypothetical Change to Pressure Irrigation Systemsa

Variable (1)

Irrigation application efficiency (Eiae) Leaching fraction (LF) Diverted irrigation water (Qdiw) (m) Volume IRF (Qsirf) (m) TDS IRF (Csirf) (mg/L) Salt mass IRF (Msirf) (Mg/ha) ET concentration factor (ETCF) Water use efficiency (WUE)

Actual Surface Irrigation (2)

Future Pressure Irrigation (3)

% Differenceb (4)

0.65 0.21

0.80 0.10

23 52

1.11 0.62

0.70 0.25

37 60

806 5.0 3.3 0.79

1,092 2.7 4.6 0.90

35 46 39 14

This change was characterized by new values of irrigation application efficiency (Eiae  0.80) and leaching fraction (LF  0.10). The diverted irrigation water (Qdiw  0.70) was accordingly modified. b 100  (Future  Actual)/Actual a

ET, evapotranspiration; IRF, irrigation return flow; TDS, total dissolved solids.

Ca, so they can be used safely without affecting crop yields and soil’s structural stability. Model predictions for VID show that, after the first reuse, subsurface drainage waters and irrigation return flows become nearly saturated with gypsum, so that subsequent reuses negligibly affect the concentration of these waters. Since TDS of the IRF (Csirf) remained unchanged with this reuse strategy, the IRF salt mass decrease (30%) was almost parallel to the IRF volume decrease (31%). In contrast, the selected area in BID is moderate in soil salinity (2 dS/m  ECe  4 dS/m) and without gypsum. As a consequence, the first reuse will increase the TDS of the IRF by 24% and the salt mass (Msirf) will decrease only by 22%, for a 45% decrease in Qsirf (Table 30-6). CIRFLE also predicts that subsequent reuses of IRF for irrigation will drastically increase its TDS, making the long-term sustainability of this strategy unacceptable in BID. The three strategies simulated with CIRFLE should be analyzed in each area of interest, because results will change depending on their characteristics. Even so, these strategies show that TMDLs could be attained by decreasing the volume of applied irrigation water, through the modernization of irrigation districts, and by the reuse of drainage waters for irrigation.

TABLE 30-6. CIRFLE Estimates for 1983 Violada and 1993 Bardenas I (2–4 dS/m ECe Homogeneous Area) Irrigation Districts, and CIRFLE Predictions Assuming 25% of the Applied Irrigation Water Comes from Reused Irrigation Return Flows Violada Irrigation District Variable (1)

Canal irrigation water (m) TDS diverted irrigation water (Cdiw) (mg/L) Reused IRF for irrigation (m) TDS reused IRF (mg/L) Volume IRF (Qsirf) (m) TDS IRF (Csirf) (mg/L) Salt mass IRF (Msirf) (Mg/ha) a

VID: 100  (Reused IRF  Year 1983)/Year 1983 BID: 100  (Reused IRF  Year 1993)/Year 1993

b

IRF, irrigation return flow; TDS, total dissolved solids

Bardenas I Irrigation District 2  ECe  4 (dS/m)

Year 1983 (2)

Reused IRF (3)

% Differencea (4)

Year 1993 (5)

Reused IRF (6)

% Differenceb (7)

1.13 190 — —

0.85 965 0.28 2,330

25 408 — —

1.11 197 — —

0.84 333 0.28 927

25 70 — —

0.91 2,330

0.63 2,389

31 3

0.62 927

0.34 1,146

45 24

21.2

14.9

30

5.8

4.5

22

947

948

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

MODEL LIMITATIONS, MODEL IMPROVEMENTS, AND FUTURE RESEARCH NEEDS CIRFLE, a management-oriented model, is inherently limited by its simplifying assumptions. One of the model’s limitations is its assumption of steady-state conditions. Whether or not this holds true for the case to be simulated needs to be ascertained. CIRFLE may therefore simulate longterm conditions but not short-term, transient conditions. Another limitation is that the model’s stationary characteristics are coupled with some nonstationary variables, such as the volume and concentration of initial and final soil water, gypsum content, soil salinity, and the leaching efficiency of salts. For example, gypsum content changes with time due to its dissolution, or its precipitation, or both. If gypsum is present in excess, the soil solution remains saturated at all times and this variable can be considered stationary. However, soluble salts may be leached by percolating waters so that the initial and final salt contents differ according to leached pore volumes and the empirical leaching efficiency parameter. Problems may arise from the fact that input variables are averaged over time and space. This simplification could prove limiting in cases with highly heterogeneous soil properties or nonuniform application of irrigation water. In such cases, the irrigation district must be subdivided into homogeneous subareas, the model has to be applied to each area independently, and global irrigation return flows must be calculated by adding up the individual simulations (i.e., volume-weighted averaging). Future research is needed to establish possible interactions between subsurface drainage waters and groundwaters as well as the nature of the salt-mixing processes at the interface between percolating waters and resident shallow groundwaters. This is an important consideration because many salt-affected irrigation districts have shallow water tables with salinity and ion composition that may differ from that of the percolating waters. Assuming that no water is used from the water table, this situation could be modeled by considering the presence of a noninteracting volume and a mixing volume. Although the total volume of subsurface drainage water would remain unaffected by the mixing mechanism, its concentration will change according to the equation Cdsw  g Cgw  (1  g) Csw

(30-16)

where g is a global interactive coefficient between groundwater and subsurface drainage. However, other models presented in this manual are best suited to conform to this scenario.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

949

SUMMARY AND CONCLUSIONS In recent decades, significant advances have been made in the modeling of IRFs. Models range from simple, mass balance routines, to state-ofthe-art, modular digital codes. Simple, mass balance, steady-state models have inherent limitations and users should have a proper knowledge of the study areas, and should be aware of scenarios where their use is inappropriate. In contrast, more sophisticated, transient models may provide more accurate estimates but the extensive requirements in model inputs and parameters may prevent or limit their use in areas with insufficient information. CIRFLE is a mass balance, steady-state hydrosalinity model, simple in concept and formulation that requires only 25 input data and coefficients. Yet CIRFLE has reliably predicted the volume and salt concentration and mass of IRFs in areas with nonsaline to saline soils and with native gypsum, and has the inherent characteristics needed to evaluate water management practices, irrigation water quality, and soil properties that affect salt loading of IRFs. Of particular relevance is the sensitivity analysis of calibrated and validated models because this provides a way to rank the more important measurements and estimates, and gives insights for a better understanding of the system. CIRFLE has proven to be useful for evaluating alternative salinity control practices such as decreasing the volume of applied water, the modernization of irrigation districts (i.e., a change from surface-irrigation to pressure-irrigation systems), and the reuse of drainage waters for irrigation. Model simulations of these and other strategies should help in devising best management alternatives for a proper control of salt loading in IRFs in order to comply with the mandatory TMDL program that is being developed around the world. The CIRFLE model is written in C# and runs interactively, with numerous built-in comments and explanations. It also includes a module for sensitivity analysis. The model is currently available in a Windows-based format, available free from the Agrifood Research and Technology Center of Aragón (Spain) (CITA 2011). REFERENCES Aragüés, R., and Tanji, K. K. (2003). “Water quality of irrigation return flows,” in Encyclopedia of water science, B. A. Stewart and T. A. Howell, eds., Marcel Dekker, Inc., New York, 502–506. Aragüés, R., Tanji, K. K., Quílez, D., Alberto, F., Faci, J., Machin, J., and Arrue, J. L. (1985). “Calibration and verification of an irrigation return flow hydrosalinity model.” Irrig. Sci., 6, 85–94.

950

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Aragüés, R., Tanji, K. K., Quílez, D., and Faci, J. (1990). “Conceptual irrigation return flow hydrosalinity model,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va., 504–529. Ayers, R. S., and Westcot, D. W. (1985). Water quality for agriculture, FAO Irrigation and Drainage Paper 29, Rev. 1, Food and Agriculture Organisation of the United Nations, Rome. Bahçeci, I., Dinç, N., Fuat Tari, A., Agar, A. I., and Sönmez B. (2006). “Water and salt balance studies, using SaltMod, to improve subsurface drainage design in the Konya-Çumra plain, Turkey.” Agric. Water. Mgmt., 85, 261–271. Bahçeci, I., and Nacar, S. (2007). “Estimation of rootzone salinity using SaltMod, in the arid region of Turkey.” Irrig. Drain., 56, 601–614. Bastiaansen, W. G. M., Allen, R. G., Droogers, P., D’Urso, G., and Steduto, P. (2007). “Twenty-five years modeling irrigated and drained soils: State of the art.” Agric. Water Mgmt., 92, 111–125. Bernstein, L., and Francois, L. E. (1973). “Leaching requirement studies: Sensitivity of alfalfa to salinity of irrigation and drainage waters.” Soil Sc. Soc. Am. Proc., 37, 931–943. Burkhalter, J. P., and Gates, T. K. (2006). “Evaluating regional solutions to salinization and waterlogging in an irrigated river valley.” J. Irrig. Drain. Eng., 132(1), 21–30. ———. (2005). “Agroecological impacts from salinization and waterlogging in an irrigated river valley.” J. Irrig. Drain. Eng., 131(2), 197–209. Centro de Investigación y Tecnología Aroalimentaria de Aragón (CITA). (2011). “Software Download, Soils and Irrigation Department,” www.cita-aragon.es, accessed February 24, 2011. Christen, E. W., Ayars, J. E., and Hornbuckle, J. W. (2001). “Subsurface drainage design and management in irrigated areas of Australia.” Irrig. Sci., 21, 35–43. Corwin, D. L., Rhoades, J. D., and Sˇ imu˚ nek, J. (2007). “Leaching requirements for soil salinity control: Steady-state versus transient models.” Agric. Water Mgmt., 90, 165–180. El-Ashry, M. T., van Schilfgaarde, J., and Schiffman, S. (1985). “Salinity pollution from irrigated agriculture.” J. Soil Water Conserv., 40, 48–52. European Union (EU). (2000). “Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for community action in the field of water policy.” Official J. Euro. Communities, L 327, December 12, 1–73. Faci, J., Aragüés, R., Alberto, F., Quílez, D., Machin, J., and Arrue, J. L. (1985). “Water and salt balance in an irrigated area of the Ebro River basin (Spain).“ Irrig. Sci., 6, 29–37. Guganesharajah K., Pavey J. F., van Wonderen J., Khasankhanova G. M., Lyons D. J., and Lloyd, B. J. (2007). “Simulation of processes involved in soil salinization to guide soil remediation.” J. Irrig. Drain. Eng., 133, 131–139. Heaven, S., Kolosov, G. B., Lock, A. C., and Tanton, T. W. (2002). “Water resources management in the Aral basin: A river basin management model for the Syr Darya.” Irrig. Drain., 51,109–118. Hoffman, G. (1986). “Guidelines for reclamation of salt-affected soils.” Applied Agric. Res., 1(2), 65–72.

CONCEPTUAL IRRIGATION PROJECT HYDROSALINITY MODEL

951

Hornbuckle, J., Christen, E., Podger, G., White, R., Seaton, S., Perraud, J. M., and Rahman, J. (2005). “Predicting irrigation return flows to river systems: Conceptualisation and model development of an irrigation area return flow model,” in Proc. MODSIM 2005 Congress, 12–15 December 2005, Melbourne, Modelling and Simulation Society of Australia and New Zealand Inc. (MSSANZ), The Australian National University, Canberra, Australia, 2700–2706. Hornsby, A. G. (1973). Predicting modeling for irrigation return flows, U.S. Environmental Protection Agency Publication R2-73-168, U.S. EPA, Washington D.C. Kaddah, M. T., and Rhoades, J. D. (1976). “Salt and water balance in Imperial Valley, California.” Soil Sci. Soc. Am. J., 40, 93–100. Letey, J., and Feng, G. L. (2007). “Dynamic versus steady-state approaches to evaluate irrigation management of saline waters.” Ag. Water Mgmt., 91, 1–10. Minhas, P. S. (1996). “Saline water management for irrigation in India.” Ag. Water Mgmt., 30(1), 1–24. Oosterbaan, R. J. (2000). SALTMOD: Description of principles, user manual and examples of application, International Institute for Land Reclamation and Improvement (ILRI), Wageningen, The Netherlands. Prendergast, J. B., Rose, C. W., and Hogarth, W. L. (1994). “Sustainability of conjunctive water use for salinity control in irrigation areas: Theory and application to the Shepparton Region, Australia.“ Irrig. Sci., 14(4), 177–187. Quílez, D. (1988). “La salinidad de las aguas superficiales de la Cuenca del Ebro, Análisis del impacto potencial del regadío de Monegros II.” Ph.D. dissertation, University of Lérida, Spain. Richards, P. L., and Kump, L. R. (1997). “Application of the geographical information systems approach to watershed mass balance studies.” Hydrol. Proc., 11, 671–694. Schoups G., Hopmans J. W., and Tanji K. K. (2006). “Evaluation of model complexity and space-time resolution on the prediction of long-term soil salinity dynamics, western San Joaquin Valley, California.” Hydrol. Proc., 20, 2647–2668. Schoups G., Hopmans J. W., Young C. A., Vrugt, J. A., Wallender, W. W., and Tanji K. K. (2005). “Sustainability of irrigated agriculture in the San Joaquin Valley, California.” Proc. Nat. Acad. Sci. USA, 102(43), 15352–15356. Scofield, C. S. (1940). “Salt balance in irrigated areas.” Agric. Res., 61, 17–39. Sharma, B. R. (1999). “Regional salt- and water-balance modelling for sustainable irrigated agriculture.” Ag. Water Mgmt., 40, 129–134. Singh, M., Bhattacharya, A. K., Singh A. K and Singh A. (2002). “Application of SALTMOD in coastal clay soil in India.” Irrig. Drain. Syst., 16 (3), 213–231. Skogerboe, G. V., and Walker, W. R. (1981). “Impact of irrigation on the quality of groundwater and river flows,” in D. Yaron, ed., Salinity in irrigation and water resources, Marcel Dekker, Inc., New York. Smedema L. K., and Shiati, K. (2002). “Irrigation and salinity: A perspective review of the salinity hazards of irrigation development in the arid zone.” Irrig. Drain. Syst., 16(2), 161–174. Tanji, K. K. (1977). “A conceptual hydrosalinity model for predicting salt load in irrigation return flow,” in Managing saline water for irrigation, Texas Tech University, Lubbock, Tex., 49–70. Tanji K. K., and Kielen, N. C. (2002). Agricultural drainage water management in arid and semi-arid areas, FAO Irrigation and Drainage Paper 61, Food and Agriculture Organisation of the United Nations, Rome.

952

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Tedeschi, A., Beltrán, A., and Aragüés, R. (2001). “Irrigation management and hydrosalinity balance in a semi-arid area of the middle Ebro river basin (Spain).” Agric. Water Mgmt., 49, 31–50. Texas Commission on Environmental Quality (TCEQ). (2007). “Two total maximum daily loads for chloride and total dissolved solids in the Colorado River below E. V. Spence Reservoir for Segment Number 1426,” www.tceq.state.tx.us/ assets/public/implementation/water/tmdl/32colorado/32-uppercoloradotmdl adopted.pdf, accessed February 12, 2011. Thayalakumaran, T., Bethune, M. G., and McMahon, T. A. (2007). “Achieving a salt balance: Should it be a management objective?” Agric. Water Mgmt., 92, 1–12. U.S. Environmental Protection Agency (US EPA). (1991). “Guidance for waterquality based decisions: The TMDL process,” EPA 440/4-91-001, www.epa. gov/owow/tmdl/decisions/, accessed February 12, 2011. U.S. Environmental Protection Agency, Region 6 (US EPA Region 6). (2006). TMDLs for turbidity, sediment, TSS, chloride, sulfate, and TDS for subsegments 100309, 100602, and 100603 in the Red River basin, Louisiana, U.S. Environmental Protection Agency Region 6 Water Quality Protection Division Oversight and TMDL Team, Dallas, Tex. van Rensburg, L. D., Strydom, M. G. du Preez, C. C., Bennie, A. T. P., le Roux, P. A. L., and Pretorius, J. P. (2008). “Prediction of salt balances in irrigated soils along the lower Vaal River, South Africa.” Water SA, 34(1), 11–18.

NOTATION C  salt concentration EC  electrical conductivity ECe  electrical conductivity of soil saturation extract ET  evapotranspiration ETCF  evapotranspiration concentration factor IRF  irrigation return flows LF  leaching fraction M  salt mass Q  volume of water TDS  total dissolved solids

CHAPTER 31 MICROECONOMICS OF SALINITY AND DRAINAGE MANAGEMENT Keith C. Knapp

INTRODUCTION Numerous practices exist for salinity and drainage management in irrigated agriculture. These include crop rotations, volume and timing of irrigation water, investment in improved irrigation systems, installing subsurface drainage systems, reusing drainage water, and treating or disposing of water collected in subsurface drains. From an economic point of view, these actions are evaluated by effects on crop output and revenue, water and production costs, and environmental consequences arising from deep percolation flows and other emissions. For economic efficiency, actions are selected to maximize the social net benefits where all inputs and outputs from the system are valued from the perspective of the overall economy. Grower profit-maximization will be economically efficient if all inputs and outputs are correctly priced. Incorrect pricing of inputs and outputs from the system provides an opening for public policy. One approach to selecting best management practices is to simulate alternate management policies using crop-water production functions, and then choose the best according to some criterion. Another approach is to formulate an optimization problem and solve it with appropriate algorithms. Simulation allows for detailed physical, chemical, and biological process models but does not optimize beyond simple enumeration or trial and error. Optimization finds the best management practice under specific conditions, but computational considerations usually limit model complexity. The two approaches may be combined for some applications and are usefully considered as complements. Various options can be screened with an optimization model, and then one or more simulation models used to evaluate the selected options. 953

954

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

This chapter focuses on computational economics models for salinity and drainage management. The framework for these models follows from economic theory. Process-based models from the physical and biological sciences are incorporated as needed, and parameter values generally follow from direct observation or estimation from experimental data. The models are solved on the computer and used to explore characteristics of the underlying system, develop efficient management practices, and identify shadow values and policy instruments to achieve efficient and equitable solutions to salinity and drainage problems. This chapter primarily emphasizes dynamic optimization models, although some static analyses are included as appropriate. It describes the available optimization models and possible extensions of those models for a variety of applications at the field and farm scales. Regional production economic models coupled to hydrologic models for surface and groundwater systems are not considered here due to space limitations [Knapp (1999) provides a review].

DYNAMIC IRRIGATION MANAGEMENT FOR A SINGLE SEASON Irrigation scheduling selects irrigation timing, volume, and quality to maximize net benefits over the course of an irrigation season at the field level. Social net benefits equal crop revenue minus irrigation and production costs, and reuse/disposal costs of drainage water. Social net benefits will equal farm profits if all inputs and outputs are correctly priced. This means, for example, that the farmer is charged for the off-farm environmental costs that result from disposal of drain water. In this event, private and social optimums will be identical; if not, then policy intervention is necessary for efficiency. Irrigation scheduling models can be applied to a wide variety of issues, including water management, water demand and conservation, pollution control costs, drainwater reuse, and environmental damages from water salinization and climate change. Bioeconomic Model for Field-Scale Crop Growth and Management The spatial dynamic model for optimal irrigation scheduling is for water, salinity, and drainage management at the field scale over the course of a single irrigation season. The crop and irrigation system are given, and there are two irrigation sources differing in quality and price. Water is distributed nonuniformly over the field in response to soil heterogeneity and/or nonuniform irrigation systems; in particular, the field is spatially variable with n subareas characterized by different water infiltration characteristics. Plant growth, soil moisture, and soil salinity evolve

MICROECONOMICS OF SALINITY AND DRAINAGE

955

on a daily time scale, where T(1) and T(2) denote the Julian date of plant and harvest, respectively. Deep percolation flows below the field carry salts and other substances and may damage receiving water, thus incurring societal costs. The object is to select irrigation dates and quantity and quality to maximize net benefits, defined as crop revenue net of production and deep percolation costs. Annual net benefits to land and management per-unit area are n

T (2)

i1

tT ( 1)

 p∑ y i ai   

∑

n

( p1 w 1 t  p 2 w 2 t )  p d ∑

T (2)

∑

dit ai

(31-1)

i1 tT ( 1)

where yi  crop yield; ai  the fractional area of the field such that, n

∑ ai  1, w1t and w2t

are field-average applied water depth from sources 1

i1

and 2, respectively; and dit  deep percolation emissions/leaching, for subarea i  1,...n, and Julian day t. Parameters are p, p1, p2, and pd as the price of crop output, water, and deep percolation, respectively, and  is nonwater production costs associated with the cropping system. – w – w – with salt Field-level applied water depth on day t is w t 1t 2t c s 1 w1 t  c s 2 w 2 t concentration cwt  , where cs1 and cs2 are salt concentrations wt of the two irrigation sources. A typical interpretation is that the first source is fresh water from surface sources with a high price and low salt concentration (high quality), whereas the second source is a low-quality/ high-salinity water potentially representing drainwater reuse. While there may be nominal costs associated with drainage reuse, such as pumping costs or gypsum treatments, p2 may be negative, meaning that reuse has beneficial value in reducing drainage problems. A static, seasonal model of spatially variable water infiltration was proposed by Seginer (1978) and used subsequently in Feinerman et al. (1983), Dinar et al. (1985), and Berck and Helfand (1990). The dynamic model presented here follows that literature by assuming that water infiltration in each subarea of the field is characterized by a water infiltration coefficient i , i  1,...n, and  苸[0, ]. These coefficients give the fraction of field-average water depth infiltrating in each subarea of the field. Thus, water infiltration in the ith subarea of the field on day t is – with salt concentration c . The infiltration coefficients  are wit  iw t wt i distributed over the field according to a spatial density function, f(), with n

mean   E[]  ∑i ai  1, and a standard deviation SD[] that depends i1

on the system type.

956

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Daily evapotranspiration (ET) rates for the ith subarea, eit, are given by

(31-2)

where c it is soil water salt concentration, h it is matric potential, and –e (b )  maximum daily ET under nonstressed conditions. Parameters it are e1, which converts soil salinity to osmotic potential, e2 is an exponent with no physical interpretation, and h 50 is soil-water potential causing a 50% reduction in ET. Matric potential is calculated according to where mit is soil moisture, mrz is rootzone depth, and the other symbols denote parameters [van Genuchten (1978)]. Deep percolation flows follow the relation dit  Max[0, wit  (mfc  (mit  eit))] under the assumption that deep percolation only occurs for moisture contents greater than field capacity, mfc. Soil moisture dynamics are given by the accounting relation mi,t1  mit  wit  eit  dit

(31-3)

following mass balance. Here soil moisture is calculated as water depth in the rootzone, and increases with applied water and decreases with ET and deep percolation flows. Soil salinity is calculated as cit  sit/mit and rootzone salt mass dynamics are si,t1  sit  cwtwit  cdtdit

(31-4)

where cdt is salt concentration of deep percolation. This relation is also a mass balance identity as salt is added in the irrigation water and exits the rootzone in deep percolation flows. Plant growth follows a logistic growth function with growth rates influenced by relative ET. In particular,

is nonstressed – growth where b is the intrinsic growth rate and b is the maximum possible biomass (carrying capacity of the environment). Biomass dynamics are then (31-5)

MICROECONOMICS OF SALINITY AND DRAINAGE

957

eit is relative ET, which depends on soil moisture and soil salinity. e (bit ) This relation implies maximal growth under nonstressed conditions; however, biomass growth is reduced in proportion to crop stress as measured by relative ET. The plant-level production function for yield is in turn yit  y1  y2bi,T(2)  y3b2i,T(2), where bi,T(2) is plant biomass at harvest time. For crops such as alfalfa where crop yield equals biomass, then y1  0,y2  1,y3  0. For other crops, y2 converts biomass to crop yield and the other coefficients allow for threshold effects and excessive vegetative growth that reduces marketable yield. where

Model solution procedures – – Control variables are daily applied water depths w 1t and w2t from each irrigation source, while the state variables are soil moisture, soil salinity, and crop biomass for each of the subareas. The optimization problem is to – and w – to maximize net benefits, subject to the equations of select w 1t 2t motion and other definitions and constraints defined previously, nonnegativity conditions and possibly other constraints, and given initial values for the state variables. This formulation implicitly assumes that rainfall and other climatic variables are deterministic. In the stochastic case, decision rules for the control variables are chosen to maximize the expected value of net benefits. An ideal solution procedure for this problem is dynamic programming (Bertsekas 1976) as this can incorporate extensions of the aforementioned framework to stochastic problems and to fixed irrigation setup costs, which are a key element in irrigation system investment along with spatial variability and capital costs of the system. The primary difficulty with dynamic programming is that computational feasibility generally limits problems to a few state variables, implying that high-dimension spatial systems cannot be solved. Under deterministic conditions and without irrigation setup costs, nonlinear programming techniques can be used. However, most nonlinear programming algorithms are guaranteed to find global optima only for convex problems. Empirical specifications for the ET function (Eq. 31-2) are generally nonlinear and may exhibit significant nonconvexities. The evidence suggests that irrigation setup costs significantly affect the number of irrigations, and these setup costs introduce nonconvexities and discontinuities into the irrigation cost function that conventional optimization algorithms cannot handle. Bras and Seo (1987) propose the use of extended linear quadratic control in the irrigation scheduling problem. The first step is to choose an initial path of control and state variables. The system is linearized around this path, and a new set of values for the control variables is chosen by using linear quadratic control techniques. Expectations are calculated

958

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

where needed, and modifications are introduced to handle inequality constraints on the control and state variables. Iterations are repeated until convergence to an optimum is achieved. The method shows promise for handling problems with many state variables. The drawbacks include the use of a suboptimal controller for stochastic problems and possible convergence to a local, but not global, optimum for nonconvex problems. Empirical Studies Several studies analyze optimal irrigation scheduling under saline conditions. Yaron et al. (1980) focus on sorghum with crop ET assumed to be a linear function of soil moisture. They do not consider rainfall and water movement between the rootzone and the water table. Soil salinity follows a model developed by Bresler (1967) in which soil salinity after an irrigation is calculated by mass balance from soil salinity before the irrigation, salts added in the irrigation water, and salts leached through deep percolation. The volume of deep percolation is the excess of irrigation over the moisture deficit (field capacity minus current moisture content), and the salt concentration is the average of soil salinity before and after irrigation. If the volume of deep percolation water exceeds two pore volumes, then soil salinity after the irrigation is assumed to equal the irrigation water’s salt concentration. Yaron et al. (1980) calculate yield as a function of the number of critical days in several subperiods, where a critical day is any day in which soil water potential is less than a specified level. Irrigation can take place every 6 days, and irrigation costs include only the cost of water. Knapp and Dinar (1987) consider optimal irrigation scheduling for cotton. Crop ET is a function of matric potential plus osmotic potential. They use a piston-flow model to compute soil salinity and assume yield to be a function of seasonal ET. They also do not consider rainfall and water movement between the rootzone and the water table. Deep percolation is calculated as the excess, if any, of irrigation over the moisture deficit. Irrigation can occur on any day of the season, and irrigation costs include a fixed setup cost plus the cost of water. Dynamic programming is used to solve the optimization problem. Bras and Seo (1987) consider corn. Evaporation is a function of time, while transpiration depends on soil moisture and salinity. Rainfall is stochastic; upward flow from a water table is calculated exogenously; and deep percolation is a function of the soil moisture content. Crop yield equals a weighted average of transpiration in each stage. Irrigation expenses include water and fixed costs of irrigation, although the latter are not treated as an integer variable. They use the extended linear quadratic control algorithm discussed earlier to solve the problem.

MICROECONOMICS OF SALINITY AND DRAINAGE

959

These empirical studies provide a number of insights into irrigation scheduling under saline conditions. In practice, salt leaching typically occurs at the beginning of the season. However, Yaron et al. (1980) and Knapp and Dinar (1987) find that some leaching during the season is also optimal. Yaron et al. (1980) and Knapp and Dinar (1987) also find that initial soil salinity has a major effect on seasonal applied water quantities. In Knapp and Dinar (1987), for example, seasonal water applications double when initial ECe—the electrical conductivity of a saturated extract— increases from 2 dS/m to 12 dS/m. Yaron et al. (1980) find that an increase in irrigation water salinity significantly increases irrigation volumes and decreases net income. Changes in the price of irrigation water scarcely affect irrigation volumes for cotton (Knapp and Dinar 1987), but more significantly affect corn (Bras and Seo 1987). This is likely due to the crop biological characteristics and their relative values. Knapp and Dinar (1987) also evaluate the effect of drainage water disposal costs on irrigation applications, yields, and profits. They find that increased drainage water disposal costs can reduce irrigation volumes by up to 10%, depending on initial salinities. Knapp and Dinar (1987) find that irrigation setup costs are a major determinant of the number of irrigation events during the season. Eliminating these costs increases irrigation events by a factor of 4 to 6, depending on initial salinity levels. This corresponds to the observation that irrigations are more frequent under drip and sprinkler systems than under surface systems, where the costs of initiating irrigations are higher. To some extent, this result depends on the model. In the Knapp and Dinar (1987) model, crop ET is reduced continuously as moisture falls below field capacity. Yields will likely decline as a consequence, depending on total seasonal ET. However, if maximum ET can be maintained over a range of soil moistures and soil salinity is low, then setup costs will not affect the frequency and depths of irrigation as much. Knapp and Dinar (1987) also consider the use of low-quality water for irrigation. This augments fresh-water supplies and reduces the disposal and environmental costs of collected drain water. Depending on initial salinities, they find that low-quality water contributes from 33% to 100% of the total water applied, depending on the particular situation being considered. Extensions Extensions must be made to the work as described for economic irrigation scheduling models under saline conditions to be fully operational. The existing studies implicitly assume uniform irrigation conditions; however, irrigation water is generally applied nonuniformly over the

960

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

field, and soil parameters are highly variable over space. These variabilities largely determine optimal water applications and are known to be quantitatively significant from static analysis (Dinar et al. 1985) in that it is not possible to understand observed water applications and deep percolation values without accounting for those variabilities. Parameter variations by depth are also ignored here. Some evidence suggests that crops can minimize stress by extracting water from layers where water is relatively abundant or nonsaline. Depth variations also influence the quantity and quality of deep percolation. In addition, the existing literature with soil salinity does not include explicit biomass dynamics as stated. Instead, it assumes exogenous maximum ET for a given day; however, this is not realistic as earlier plant stress means that biomass and, hence, maximum ET for a given day are actually variables in the analysis. As discussed in the next section, salinity and moisture levels at the end of the season influence production and profitability in future years and thus need to be accounted for in the current year. This can be accomplished by including a terminal value function in the objective function (Eq. 31-1). The terminal value function gives discounted net benefits in future years as a function of the soil salinity and soil moisture at the end of the current year. Inclusion in Eq. 31-1 implies that both current and future profits will be taken into account when irrigation decisions are made in the current year. The interseasonal models discussed in the next section can be used to specify terminal value functions. High water tables, common in areas with saline soils, exacerbate the salinity problem. Irrigation decisions influence and are influenced by the presence of an underlying water table. Irrigation scheduling models should include the water table height, salt concentration, and upward flows to the rootzone as endogenous variables. Bras and Seo (1987) consider uncertain rainfall. However, crop prices, production costs, pest problems, and meteorological conditions during the growing season are also likely to be uncertain when irrigation decisions are made. Ideally, the probability distributions of these variables would be characterized and the problem solved as a stochastic control problem, although this is likely to be computationally burdensome. An alternative is Bras’s and Seo’s (1987) strategy in which the optimization problem is solved repeatedly throughout the season as new observations on the state variables become available. Finally, many of the parameters underlying the various relations are not known with certainty. Parameter uncertainty with learning is an adaptive control problem.

MULTIYEAR IRRIGATION MANAGEMENT Interseasonal field-level dynamics arise from several factors, including moisture carryover, salt accumulation, crop rotations, tree and perennial

MICROECONOMICS OF SALINITY AND DRAINAGE

961

crops, and investment in irrigation and drainage systems. The implication is that current period management must consider both current net returns and impacts on future growing seasons. As an example, if a salt-tolerant crop such as cotton is followed by a salt-sensitive crop such as tomatoes, then optimal irrigation of cotton may involve additional salt leaching beyond what would be needed by cotton alone in order to provide acceptable soil salinity for tomatoes. While the conceptual model in the previous section can be extended in a number of directions to account for these effects, this likely results in a computationally intensive model. Accordingly, the economic literature to date on this topic typically works at a higher level of aggregation.

Salt Accumulation Salt accumulation in the rootzone depends on irrigation practices, cropping patterns, rainfall, height of the water table, and irrigation salt concentration. A model for interseasonal salt accumulation over time under uniform conditions can be formulated as follows. The present-value of annual net benefits is defined as T

∑ t [ pyt  (wt , ct )]

(31-6)

t1

where is the discount factor, and wt and ct are seasonal applied water depth and salt concentration, respectively, for year t  1,...T. Crop yield yt depends on applied water and soil salinity st, and can be estimated from experimental data, while (wt,ct) gives irrigation costs allowing for the possibility of multiple sources. The equation of motion for soil salinity is st1  g(st,wt,ct)

(31-7)

which increases with applied water up to a point, then decreases as salts are leached from the soils. This relation may be specified from theoretical models or statistical analysis of experimental data. The optimization problem is to select seasonal applied water to maximize the present value of profits, subject to the soil salinity equation of motion and other constraints. Dinar and Knapp (1986) solved this problem for alfalfa and cotton. They use data from field experiments to statistically estimate a salt balance and yield relation. Dynamic programming computes optimal decision rules for an infinite time horizon. These give the profit-maximizing volume of seasonal water that must be applied, expressed as a function of the salinity of the soil at the beginning of the irrigation season. In general,

962

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

water applications increase as initial soil salinity increases. Optimal time paths for soil salinity can be calculated from the optimal decision rules for various initial conditions. For cotton, soil salinity (measured as the EC of a saturated extract) converges to a steady-state level of 6 dS/m, regardless of the initial conditions. More generally, annual fluctuations in salinity can be expected from crop rotations and changes in input costs and price received. Dinar and Knapp (1986) also consider the optimal mix of highquality irrigation water and low-cost, low-quality (saline) irrigation water. The results show that no saline water is used until the price of the goodquality water exceeds $8.00/ha-cm (in 1985 dollars). Thus, a single-season decision rule, calculated by maximizing net returns in each year without consideration for future years, results in discounted net returns as much as 26% lower than those from an optimal interseasonal decision rule. Yaron and Olian (1973) examine the interseasonal dynamics of salinity management for citrus with winter leaching and fixed amount of water applied in the summer as determined by traditional farming practices. Mass balance relations describe physical-system dynamics in a modified version of the leaching model developed by Bresler (1967), with coefficients describing salinity-induced yield losses based on information from soil scientists and data from local sources. They calculate a decision rule for an optimal leaching policy over a range of observed levels of state variables. The decision rule suggests the optimal leaching fraction (LF) to use in the approaching summer as a function of soil salinity at the end of winter and the quality of available leaching and irrigation water. Economic losses from soil salinity for various water qualities and initial soil salinities are used to estimate the farm-level value of improvements in water quality. These results, as noted by the authors, depend heavily on assumptions about salt leaching, accumulation functions, and the relationship between yield and salinity. Matanga and Mariño (1979) extend Yaron and Olian (1973) to include seasonal irrigation water as a decision variable. They use stochastic dynamic programming and simulation to choose the leaching depths and seasonal water applications that maximize discounted net benefits over finite and infinite planning horizons. Seasonal rainfall is treated as a stochastic variable, and probabilities conditional on rainfall during the previous period are used to compute its expected value in any season. The model is applied to corn, grain sorghum, and pinto beans in a single-crop framework for each. For each crop, results from the interseasonal model provide decision rules describing the optimal amount of seasonal irrigation water and leaching water to apply as a function of soil salinity and expected rainfall at the beginning of any stage. Empirical results are given for 30-year finite and infinite planning horizons. As with the intraseasonal irrigation scheduling problem, a major consideration is incorporating field-level spatial variability in applied water

MICROECONOMICS OF SALINITY AND DRAINAGE

963

or soil conditions. This is necessary for accurately representing field-level response to crop and water management, as well as including irrigation system investment as different systems vary in terms of infiltration uniformity. A straightforward approach is to consider the field as a collection of grid cells with soil salinity, soil moisture, and crop biomass dynamics specified for each cell, and with applied water in each cell a function of the field-level applied water depth. Conceptually, the algorithms discussed earlier could be used to solve this optimization problem; however, this approach may quickly become computationally infeasible depending on the number of grid cells and horizon length (ideally infinite). An alternate approach is discussed in a later section.

Tree and Perennial Crops The previous models consider either annual crops or, in the case of Yaron and Olian (1973), perennial crops where crop response to water and soil salinity in the current period does not depend on management in the previous year. In general, however, growth of tree crops and perennial crops does depend on water and salt management in previous periods, which in turn affects yield, ET, water management, and profits in the current period as well as soil salinity evolution over time. For example, if previous water management restricts tree growth, then smaller trees in the current period will have lower biomass, yield, and ET, implying lower profits (although a given volume of irrigation water will now leach additional salts from the profile). The interseasonal model can be extended for tree and perennial crops by including a biomass variable bt and an associated growth equation. Annual net benefits and soil salinity dynamics are as before, except that now they potentially depend on biomass as well as the other variables. The main new component is the biomass equation of motion: bt1  bt 

et g(bt ) et (bt )

(31-8)

where et is actual ET, –e (bt) is maximum possible ET as a function of biomass, and g is the growth function under nonstressed conditions. This formulation implies that maximum ET in a given year is endogenous instead of exogenous, as typically assumed. Maximum crop growth occurs if the crop is not stressed by soil moisture and soil salinity. Knapp and Sadorsky (2000) consider agroforestry production (eucalyptus) with saline water. The trees are cut periodically and can be used for several purposes, including firewood and energy production. Use of saline drainwater for irrigating the trees provides a disposal mechanism

964

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

in areas adversely affected by saline, high water tables. The objective function includes a harvest variable with harvest and establishment costs, and the biomass equation of motion is modified to include periodic harvests. Decision variables in this analysis include when to cut the trees and replant, and irrigation volumes from low and high saline water sources. This problem is solved using dynamic programming methods for conditions typical of Westside of the San Joaquin Valley in California. Management with freshwater irrigation is considered first. Generally, water applications are quite responsive to soil salinity but are not highly dependent on tree biomass except for very young trees, and the harvest decision is relatively constant with respect to soil salinity. For system dynamics, after a transition period, the system converges to a limit cycle, with periodic harvests and water management varying with the tree stage of growth. Profits are positive although not very large, implying that agroforestry production is not likely to compete with other crops in the area absent consideration of drainwater reuse. Management with saline drainwater reuse depends on salt concentration of the low-quality source and the shadow value of drainwater. The shadow value of drainwater reuse reflects the damages and disposal costs associated with saline, high water tables, implying a benefit from net reuse defined as saline irrigation net of deep percolation from reuse activities. This is a benefit since net reuse is effectively disposing drainage water (although not the salts in the drainage water) and thus contributing to maintaining water tables at acceptable levels. At typical values for salt concentration and drainage shadow values in the study area, all irrigation comes from the saline source, and net reuse and social net benefits are substantial. Both saline irrigation volumes and net reuse decline substantially as irrigation salt concentration increases. These entities are somewhat sensitive to the drainage shadow value at intermediate irrigation salinity values but less sensistive at either high or low salinity levels. Overall, these results are favorable for agroforestry production in closed drainage basins. At typical values for the study area, both net reuse volumes and annual net benefits of production are substantial. Farm profits and incentive to practice agroforestry for drainwater reuse will be appropriate provided growers are compensated for drainwater reuse at an appropriate level. Spatial Variability and Irrigation System Investment Knapp (1992a) develops an interseasonal analysis to address a variety of issues. Field-level management over an infinite horizon is considered. Two irrigation sources varying in cost and salt concentration are available, and there is an exogenous crop rotation. Annual net benefits include crop revenue, water and production costs, irrigation system costs, and

MICROECONOMICS OF SALINITY AND DRAINAGE

965

deep percolation costs. Spatial variability is modeled with a different approach than the grid cell approach considered previously. Here soil salinity and water infiltration for the field are characterized by a spatial density function with a finite number of moments. The distribution moments are then treated as state variables and evolve over time as a function of the moments and control variables in the previous period. This problem is likely to be more computationally tractable if the number of moments characterizing the spatial distribution is small. Irrigation system investment is also included. Suppose there are n(k) types of irrigation systems available, kt and t are integer variables denoting the type and age of irrigation system in place at the beginning of year t, and let kt be an integer variable denoting investment in a new irrigation system at the beginning of year t. Setting kt  0 means no new investment takes place and, hence, the irrigation system remains the same, while kt 苸{1,...n(k)} implies investment in a new irrigation system. Then kt1  kt

t1  t  1

(31-9a)

when there is no new investment (kt  0), and kt1  kt

t  1

(31-9b)

when there is investment (kt  0). This is the equation of motion for the irrigation system state variables. With this formulation, the dynamic optimization problem is selection of applied water depths and irrigation system investment to maximize the present value of net benefits, subject to the equations of motion and other constraints and definitions. This problem is solved using dynamic programming. This model was explored in Knapp (1992b) for cotton and tomatoes production typical of the San Joaquin Valley in California. Irrigation systems considered are furrow with half-mile and quarter-mile runs, sprinkler, low energy precise application (LEPA), linear move, and drip. Decision rules are characterized for both continuous cotton and a cotton-cotton-tomatoes rotation. These demonstrate that water applications are generally increasing in field-level average salinity but decrease after some point. The latter effect is due to decreased ET and, hence, enhanced leaching from a given applied water depth when initial soil salinity is sufficiently high. The results also demonstrate different decision rules for the same crop, depending on place within the rotation. In particular, additional leaching water is applied for second-year cotton for a given soil salinity level to maintain an improved soil status for the subsequent salt-sensitive tomatoes

966

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

crop. Irrigation investment was investigated for a range of water and deep percolation prices. The results indicate that traditional furrow systems are selected for low irrigation and deep percolation prices, but that moreuniform, capital-intensive systems are selected as either or both of these prices increase. Knapp (1992c) develops several policy implications with this framework. Short-run and long-run water demand are investigated by varying water prices. Water demand is demonstrated to be quite elastic for low water prices, inelastic for higher prices, and finally reaching a point where water prices are too high to maintain positive profits. The analysis suggests considerable water savings starting from a situation of low initial prices, as is typically the case. Likewise, the analysis finds that starting from a zero deep percolation shadow value, considerable reduction in deep percolation is achievable to a point, after which profits would be adversely affected by higher prices with consequent little-reduced deep percolation. Damages from irrigation water salinization are also evaluated and found to be somewhat smaller than previous estimates, likely due to the enhanced range of management variables included in this analysis. This work also analyses drainwater reuse economics. For the cottoncotton-tomatoes rotation and typical water prices, reuse on cotton is optimal for irrigation salinities less than 10 dS/m. Shadow values for the low-quality source can be negative, representing reuse benefits consequent to reduced drainage disposal requirements elsewhere in the system. These can substantially increase use of a low-quality source while enhancing net returns.

FARM AND REGIONAL AGRICULTURAL PRODUCTION Previous sections considered irrigation and drainage management at the field level. However, most farms consist of several fields and, in some instances, simultaneous operation of all fields must be considered. This can occur if: 1. Resource constraints exist for the farm as a whole. For example, if the total quantity of water available to the farm is constrained, then the amount applied to one field affects the amount available for other fields. 2. Marketing quotas or contracts for specified crops affect the choice of crops that can be grown in a given year. 3. Technical externalities exist between fields. For example, deep percolation from one field may influence water table heights under other fields, or pest management practices in one field may affect pest populations in other fields.

MICROECONOMICS OF SALINITY AND DRAINAGE

967

4. The grower is risk-averse and the problem is stochastic. Then valuation of farm profits is the expected value of a nonlinear function of farm profits and not equal to the sum of the expected value of profits from individual fields. Each of these circumstances implies that decisions on all fields must be considered simultaneously.

Programming Model Suppose that there are n(c) possible crops that can be grown and n(k) possible irrigation systems. Then social net benefits from agricultural production are given by n( c ) n( k )

n( c ) n( k )

i1 j1

i1 j1

∑ ∑ ( pi yij  ij  w wij )xij  psw qsw  pgw qgw  pdw ∑ ∑ dij xij

(31-10)

where xij  land area, wij  applied water depth, yij  yield, and dij  deep percolation flows for crop i and irrigation system j. The variable qsw denotes surface water purchases by the farm, and qgw is groundwater withdrawal. Parameters are pi  crop price, w  pressurization cost, ij  nonwater production costs, psw and pgw are surface and groundwater pumping costs, respectively, and pdw is the deep percolation shadow value. Crop-water production functions are defined by {yij, dij}  f(wij) to give crop yield and deep percolation as a function of applied water depth by crop and irrigation system type. Crop areas are constrained by n( c ) n( k )

∑ ∑ xij ≤ x

(31-11)

i1 j1

where x– is the total available land area on the farm, while water applications must satisfy n( c ) n( k )

∑ ∑ wij xij ≤ qsw  qgw

(31-12)

i1 j1

implying sufficient water supply to cover applied water. For economic efficiency, the land and water variables are chosen to maximize social net benefits, subject to the various constraints. This shows the economically efficient cropping patterns, choice of irrigation systems, and water applications given prices, costs, shadow values, and resource

968

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

constraints. This framework can be used to generate a source control cost function showing the cost of achieving various specified levels of drainage reductions, where cost is defined as the reduction in net benefits for a given level of drainage emission reductions. Finally, attaching charges to and/or constraints on the variables (e.g., water use, drainage emissions, type of irrigation system) allows evaluation of grower policy response. This framework is a generic model intended to illustrate an overall approach to estimating the appropriate mix of strategies and deep percolation reduction costs; actual empirical studies may considerably modify or add to this framework. For example, additional constraints might include upper bounds on surface water usage or groundwater withdrawals and constraints on cropping patterns. Calibration terms representing land quality and other factors may be added to the objective function (Howitt 1995a,b), and some researchers model irrigation systems as continuous functions of uniformity. Spatial variation over a region and/or multiple farms can be modeled by specifying a set of equations, such as Eqs. 31-11 and 31-12 for each subarea or farm in the region; net returns for the region are then as in Eq. 31-10 except summed over all the subareas and/or farms. Demand equations or marketing constraints for output may also need to be included, depending on the fractional share of individual crops grown in the region relative to the market size. Further developments might include regional dynamics, salinity, and uncertainty in prices, and other parameters. Empirical Studies Posnikoff and Knapp (1997) apply this framework to a representative farm on Westside of the San Joaquin Valley, California. Four crops are considered, there are five possible irrigation systems, and water is distributed lognormally over the field with the variance a function of the irrigation system. They compute farm management variables and source control costs for various levels of drainage reduction. There is some reduction in cropped areas as drainage is reduced; however, the bulk of reduction in deep percolation flows comes from improved irrigation systems and reduced water applications. For source control, the first units of drainage reduction (pollution control) can be achieved at relatively low cost, but additional levels of drainage reduction incur increasingly greater costs. Wichelns and Weinberg (1990) consider drainage reduction in the Broadview Water District (Firebaugh, California). They estimate a statistical relation for drainage flows as a function of applied water and other variables, and a production function giving cotton yields as a function of applied water and other variables. They found that reducing drainage flows by an average of 0.21 ft/year at the district level (30% reduction

MICROECONOMICS OF SALINITY AND DRAINAGE

969

from unregulated levels) reduced net returns by $32/acre-year (a 12% reduction), for the field groups considered under 1988 production conditions. They also point out that an equiproportionate reduction for all farms can result in substantial inequities because some observed drainage flows on a given farm are likely the result of lateral subsurface flows. Dinar et al. (1991) also consider a multifarm framework on Westside of the SJV. They consider source control strategies of crop areas, including land fallowing, irrigation uniformity, and applied water levels, plus they allow for switching between water sources (surface and groundwater). In the least-cost solution, they found that drainage water could be reduced by 0.31 ft/year at a cost of $34.20/acre-year, and reduced by 0.43 ft/year at a cost of $77.58/acre-year in comparison to the original unconstrained level of drainage flows. Note that these costs are a reduction in net benefits from agricultural production and do not include policy costs to growers, such as a drainage emissions fee. Weinberg (1991) and Weinberg et al. (1993a,b) report an extensive analysis of source control costs at the regional level. This study considers 68,000 acres in the drainage problem area, divided into subareas that account for spatial variability in climatic conditions and water availability, among other factors. Their results indicate a steady decline in applied water and a steady increase in irrigation uniformity as the allowable level of drainwater flows decreases but suggest relatively little change in overall cropping areas, as well as the proportion of acreage devoted to specific crops. Furthermore, their results suggest that significant levels of drainage reduction can be achieved at relatively small costs; in particular, they found that a 30% reduction in collected drainwater levels could be achieved via a 4% decrease in crop returns. Dinar et al. (1993) consider a dynamic version of this model, with salt accumulation dynamics in which soil salinity is characterized by a farmlevel index, thus implicitly assuming uniformity across the farm with respect to soil salinity. Production functions are estimated from lysimeter data for several crops, and slope coefficients are adjusted to account for two different irrigation systems. Results are generated for different water and drainage prices. Spatial Dynamics Previous sections in this chapter have stressed within-field variability; however, variability also exists across fields and farms in a region. Natural causes include land quality and microclimate variability, but different crops and irrigation systems also cause variation. At the same time, different dynamic forces exist, as emphasized before. Here we stress studies that have considered salt accumulation at farm and regional scales while still allowing for spatial variability.

970

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

One approach to farm-level modeling is to treat the farm as a collection of separate fields. For each field, a set of equations may be written relating state variables to control variables and, possibly, to variables in other fields. The object is to choose the levels of the control variables in each time period that maximize the present value of profits, subject to constraints on the controls and the equations of motion for the state variables in each of the fields. Gradient methods may be used to solve this problem provided the problem is deterministic, the functions are well-behaved, no discrete variables exist, and the number of time periods and variables is limited. Dynamic programming offers a theoretical approach to quite general problems, but computational considerations limit its usefulness to fairly small problems. Yaron and Voet (1982) provide an empirical example of this discretearea approach. They consider a farm with two orchards and a field that can be used for growing various crops, and a fixed amount of water is available for irrigation. The orchards require a fixed amount of water for irrigation and the remainder is allocated to leaching in each orchard and for growing field crops. Equations of motion describe soil salinity dynamics in each of the orchards as a function of the variable leaching quantities and the fixed irrigation quantities. The optimization problem is solved via dynamic programming. An iterative approach can be used where a limited resource forces simultaneous consideration of all fields. For example, suppose that the farm’s total water supply is limited. Decisions on each field can be optimized separately for a given water price, and then the water price adjusted until supply constraints are satisfied. Yaron and Dinar (1982) give an algorithm based on this approach. Knapp (1992c) explores a shadow-value approach to the same problem. Individual field optimization problems are solved over a set of alternate time-series for watershadow values, and this is used to construct a farm-level profit function and annual water demand as functions of the shadow values. Shadow values are then selected to maximize farm profits, subject to annual constraints on water availability to the farm. A second approach to farm-level modeling is to consider the farm as one area with possibly different soil qualities, and the areas devoted to various crops are continuous variables. Lansford et al. (1986) give an example of this continuous-area approach for the intraseasonal salinity and drainage problem. They consider a farm with fresh and saline water sources, a fixed amount of land, and several crops that can be grown. An irrigation scheduling simulation model is used to determine the yield from a number of alternate irrigation scheduling policies. These policies vary according to irrigation timing and the respective amounts of fresh and saline water used during the season. A linear programming model is then formulated, where the variables include the amount of land devoted

MICROECONOMICS OF SALINITY AND DRAINAGE

971

to a particular crop irrigated by a particular scheduling policy. Farm profits are maximized subject to constraints on land and the amount of fresh and saline water available to the farm. For the interseasonal problem, each variable or activity in the linear/nonlinear programming model could represent the acreage devoted to a specified time path of decisions on crops and irrigation activities over the horizon. For example, x1 could represent the acreage of the 3-year rotation cotton/cotton/tomatoes over the horizon, x2 the acreage of continuous alfalfa, and both with a specified irrigation volume in each year. Implicit in each sequence, xi is a time path for soil salinity, drainage water volumes, crop yields, and so on. The object is to choose the values for xi that maximize the present value of profits over the horizon, subject to various constraints, such as land, water supplies, and marketing quotas. The author is not aware of any studies of farm-level salinity and drainage that use this approach for the interseasonal problem. Feinerman and Yaron (1983) describe another variant of the continuous-area approach. Suppose that the problem being considered is interseasonal soil salinity management, that a specified range of soil salinities is being considered (e.g., ECe from 0 dS/m to 20 dS/m), and let this range of salinities be subdivided into n intervals. Let ai(t) be the area of the farm with soil salinity in the ith salinity interval in year t and xij(t) be the area of the jth cropping activity planted on soil in the ith salinity interval in year t. Cropping activity here refers to a specific crop grown with a specific irrigation volume, specific fertilizer quantity, and so on. Evolution of the ai(t) is then governed by ak (t  1)  ∑ ∑ ijk xij (t) i

(31-13)

j

where ijk  the fraction of land in salinity interval i planted to cropping activity j, which has soil salinity in interval k at the beginning of the next year. In a deterministic problem, ijk is a binary, 0-1 variable with a summation over k for given i, j equal to 1. In a problem with stochastic rainfall, as considered by Feinerman and Yaron (1983), these coefficients range between 0 and 1, but the same summation applies. The xij(t) must also satisfy

∑ xij (t) ≤ ai (t)

(31-14)

j

for all i and all t. This guarantees that the area of cropping activities planted on soil with salinity in interval i does not exceed the total area in that interval. The optimization problem is to choose values of xij(t) in each

972

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

year to maximize the present value of profits over the horizon, subject to Eqs. 31-13 and 31-14 and other constraints, such as water availability and marketing quotas.

SUMMARY AND CONCLUSIONS This chapter reviews economic optimization models for salinity and drainage management at the field and farm level. The irrigation scheduling studies suggest that initial salinity levels significantly affect optimal irrigation volumes during the season, and that the setup costs of irrigation are a major determinant of the number of irrigation events during the season. The interseasonal analysis described here finds convergence to an optimal steady state that is independent of initial conditions for deterministic problems when a single crop is grown continuously, and this convergence happens relatively quickly. This implies that policy analysis at larger spatial and temporal scales can utilize steady-state assumptions, thereby providing computational tractability for largescale models. Early salinity economics research focused on losses due to irrigation water salinization; however, the more recent literature studies salinity and drainage management in limited drainage conditions. Several management and policy conclusions can be drawn from this latter analysis: • Irrigation management may have significant economic net benefits. Some studies indicate that improved irrigation management, including deficit irrigation, can achieve significant savings in deep percolation. This is especially true for traditional systems where spatial variability is high. High crop yields as well as acceptable net returns can be maintained; however, this only holds to a certain level, after which other strategies will be necessary or more significant losses will be incurred. • Systems to ensure irrigation uniformity may have economic net benefits, depending on capital costs. Mixed evidence is provided by economic analyses that irrigation system uniformity should be a prime candidate for source control strategies. Some studies find that relatively high irrigation prices or deep percolation shadow values are needed for investment in capital-intensive systems, while other studies find investment optimal for more moderate prices. • Drain water reuse has economic net benefits, whether or not disposal and environmental costs are considered. The economic studies provide considerable evidence that drain water reuse can exist at current prices when the only benefit is the replacement of fresh water. However, reuse will likely become a much more attractive

MICROECONOMICS OF SALINITY AND DRAINAGE

973

option when the disposal costs and environmental costs of drain water are taken into account. • Crop switching and land retirement are not necessarily least-cost strategies. The various analyses appear to find relatively little cropswitching in least-cost solutions, nor do the analyses suggest that land retirement/fallowing is a particularly appealing strategy. Although land retirement can reduce drainage provided that surface water is transferred out of the basin, it is generally not found to be economically efficient because other strategies can efficiently supplant it, at least for conditions indicative of California. Additionally, land retirement can lead to environmental problems if upslope areas contribute to downslope drainage and if other methods are not implemented to maintain water tables at appropriate depths. • Significant levels of water conservation and/or drainage reduction can be achieved at fairly low costs in terms of lost net benefits from agricultural production. However, beyond some point, these costs can go up dramatically. The significance is that water quality standards that are likely to impose smaller costs on agricultural production could perhaps be adapted with a lower burden of proof, whereas tighter water quality standards imposing larger costs require stronger justification. • A number of approaches to regulation of salinity drainage may help growers remain productive and profitable. Grower profits will be affected by required levels of land and water conservation as well as pollution reduction, and they will also be determined by the type of policy instrument used to induce various biophysical control strategies. However, the evidence suggests that for at least some problems and regions, high levels of grower profits can be maintained while achieving resource conservation and pollution reduction by using either quantity-based instruments or financial instruments that allow for return of the revenue to the agricultural sector. The latter might include tiered pricing, emission charges with rebates, or markets for water or transferable discharge permits. While insights into optimal salinity and drainage management can be gained with existing models, much remains to be done before these models can be used in actual situations. Further improvement of crop-water production functions for saline conditions is needed, especially for calibration, endogenous plant growth, and spatial variability. The irrigation scheduling models require incorporation of field-level spatial variability, and improved handling of crop rotations in intertemporal models is needed. Finally, this chapter has concentrated on agricultural production economic models for salinity and drainage management. A complete analysis requires coupling of the agricultural production models developed

974

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

here to explicit hydrologic models at regional spatial scales and decadal or longer temporal scales. REFERENCES Berck, P., and Helfand, G. (1990). “Reconciling the von Liebig and differentiable crop production functions.” Am. J. Agric. Econ., 72, 985–996. Bertsekas, D. P. (1976). Dynamic programming and stochastic control, Academic Press, New York. Bras, R. L., and Seo, D. J. (1987). “Irrigation control in the presence of salinity: Extended linear quadratic approach.” Water Resour. Res., 23(7), 1153–1161. Bresler, E. (1967). “A model for tracing salt distribution in the soil profile and estimating the efficient combination of water quality and quantity under varying field conditions.” Soil Sci., 104, 227–233. Dinar, A., Aillery, M. P., and Moore, M. R. (1993). “A dynamic model of soil salinity and drainage generation in irrigated agriculture: A framework for policy analysis.” Water Resour. Res., 29(6), 1527–1537. Dinar, A., Hatchett, S. A., and Loehman, E. T. (1991). “Modeling regional irrigation decisions and drainage pollution control.” Nat. Resour. Model., 5(2), 191–212. Dinar, A., and Knapp, K. C. (1986). “A dynamic analysis of optimal water use under saline conditions.” Western J. Agric. Econ., 11(1), 58–66. Dinar, A., Letey, J., and Knapp, K. C. (1985). “Economic evaluation of salinity, drainage and nonuniformity of infiltrated irrigation water.” Agric. Water Mgmt., 10, 221–233. Feinerman, E., Letey, J., and Vaux, H. J. (1983). “The economics of irrigation with nonuniform infiltration.” Water Resourc. Res., 19(6), 1410–1414. Feinerman, E., and Yaron, D. (1983). “Economics of irrigation water mixing within a farm framework.” Water Resourc. Res., 19(2), 337–345. Howitt, R. E. (1995a). “Positive mathematical programming.” Am. J. Agric. Econ., 77, 329–342. ———. (1995b). “A calibration method for agricultural economic production models.” J. Agric. Econ., 46(2), 147–159. Knapp, K. C. (1992a). “Irrigation management and investment under saline, limited drainage conditions. 1. Model formulation.” Water Resourc. Res., 28(12), 3085–3090. ———. (1992b). “Irrigation management and investment under saline, limited drainage conditions. 2. Characterization of optimal decision rules.” Water Resourc. Res., 28(12), 3091–3097. ———. (1992c). “Irrigation management and investment under saline, limited drainage conditions. 3. Policy analysis and extensions.” Water Resourc. Res., 28(12), 3099–3109. ———. (1999). “Economics of salinity and drainage management in irrigated agriculture,” in Agricultural drainage, R. W. Skaggs and J. van Schilfgaarde, eds., Chapter 40, Agronomy Society of America, Madison, Wisc. Knapp, K. C., and Dinar, A. (1987). “Optimum irrigation under saline conditions,” in Irrigation systems for the 21st century, L. G. James and M. J. English, eds., ASCE Press, Reston, Va.

MICROECONOMICS OF SALINITY AND DRAINAGE

975

Knapp, K. C., and Sadorsky, P. (2000). “Economics of agroforestry production in irrigated agriculture.” J. Agric. Resour. Econ., 25(1), 286–306. Lansford, R. R., McGuckin, T., Mapel, C. L., and Sammis, T. W. (1986). “Optimization of irrigation scheduling with alternative saline water supplies.” Presented at the annual meeting of the American Agricultural Economics Association, in Reno, Nevada, July 1986, Agricultural and Applied Economics Association, Milwaukee, Wisc. Matanga, G. B., and Mariño, M. A. (1979). “Irrigation planning II: Water allocation for leaching and irrigation purposes.” Water Resourc. Res., 15(3), 679–683. Posnikoff, J. F., and Knapp, K. C. (1997). “Farm-level management of deep percolation emissions in irrigated agriculture.” J. AWRA, 33(2), 375–386. Seginer, I. (1978). “A note on the economic significance of uniform water application.” Irrig. Sci., 1, 19–25. van Genuchten, M. T. (1978). Calculating the unsaturated hydraulic conductivity with a new closed-form analytical model, Report 78-WR-08, Water Resources Program, Department of Civil Engineering, Princeton University, Princeton, N.J. Weinberg, M. (1991). “Economic incentives for the control of agricultural nonpoint source water pollution,” unpublished Ph.D. dissertation, University of California, Davis, Calif. Weinberg, M., Kling, C. L., and Wilen, J. E. (1993a). “Water markets and water quality.” Am. J. Agric. Econ., 75, 278–291. ———. (1993b). “Analysis of policy options for the control of agricultural pollution in California’s San Joaquin River basin,” in Theory, modeling, and experience in the management of nonpoint source pollution, C. Russell and J. Shogren, eds., Kluwer Academic Press, Boston. Wichelns, D., and Weinberg, M. (1990). “Economics of agricultural drainage policies.” Calif. Agric., 44(4), 8–10. Yaron, D., Bresler, E., Bielorai, H., and Harpinist, B. (1980). “A model for optimal irrigation scheduling with saline water.” Water Resour. Res., 16(2), 257–262. Yaron, D., and Dinar, A. (1982). “Optimal allocation of farm irrigation water during peak seasons.” Am. J. Agric. Econ., 64(4), 681–689. Yaron, D., and Olian, A. (1973). “Application of dynamic programming in Markov chains to the evaluation of water quality in irrigation.” Am. J. Agric. Econ., 55(3), 467–471. Yaron, D., and Voet, H. (1982). “Application of an integrated dynamic and linear programming model to the analysis of optimal irrigation on a farm with dual quality (salinity) water supply,” in Planning and decisions in agribusiness: Principles and experience, C. F. Hanf and G. W. Schierer, eds., Elsevier, Amsterdam.

NOTATION ai, ak(t)  area by land type – bit, bt, b  actual and maximum biomass cs1, cs2, cwt, ct, cit, cdt  salt concentration of irrigation, soil and deep percolation water dit, dij  deep percolation depth

976

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

eit, e–(bit)  daily and maximum evapotranspiration f(wij)  crop-water production function g(bt)  biomass growth function hit, h50  matric potential: actual and 50% reduction in ET i  1,...n cell i  1,...n(c) crops j  1,...n(k) irrigation systems kit, kt  irrigation system capital and investment mfc, mrz  field capacity and rootzone depth mit  soil moisture n(c), n(k)  crop and irrigation system numbers p, pi, p1, p2, and pd  prices: crop output, water for source 1 and 2, and deep percolation psw, pgw, pdw  surface, groundwater, and deep percolation price qsw, qgw  surface and groundwater quantity sit, st  rootzone salt mass and soil salinity t  1,...T time (Julian day or year) wit, wij  water depth w– 1t, w– 2t, w– t  field-average applied water depth by source and totals xij, xij(t), x–  crop area and total land area yit, yij, yt  crop yield

 discount factor i  water infiltration coefficient ijk  transfer coefficient: crop i and salinity interval j to salinity interval k w, , (wt,ct), ij  pressurization and non-water production costs  profits h1, h2, h3, e1, e2  matric potential and ET function parameters b, y1, y2, y3  biomass and yield parameters res, sat  residual and saturated moisture content t  irrigation system age

CHAPTER 32 SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY William R. Johnston, Dennis W. Westcot, and Michael Delamore

INTRODUCTION The emphasis in water supply and water supply development worldwide is taking a major turn. Good-quality supplies, which previously were plentiful and readily available, are becoming fully developed. There is increased attention toward protecting these limited supplies from any type of degradation that might reduce their usability or limit their development. Water-quality problems from irrigated agriculture are at the forefront of this increased concern. Of the water-quality concerns, salinity management is likely a top priority in any intensively managed agricultural water project in order to maintain the productivity of the agricultural land. Salts from these intensively farmed areas must be managed in a manner that protects the river basin water quality. Irrigated agriculture, being the largest user of water worldwide, can have a significant impact on river basin salinity. A few examples are the Rio Grande, Pecos, Colorado, and San Joaquin River basins in the United States, the River Murray-Darling basin in Australia, the Indus River basin in Pakistan, the Tigris and Euphrates River basins of Syria and Iraq, and the Nile River basin of Sudan and Egypt. The solutions to salinity management in these basins will be as varied as the basins themselves (French 1984; FAO 1997). Data show that in many areas a significant portion of the salinity increase in surface and ground water comes from irrigated agriculture and, especially, areas that have used subsurface drainage to maintain the productivity of the land. In many of these areas, the solutions being developed are often based on new regulatory requirements for water quality and fisheries protection. Drainage water reuse and disposal to protect 977

978

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

water quality are relatively new areas of resource management. The objective of this chapter is to review the experiences and approaches used in the San Joaquin Valley of California to meet these challenges. In the San Joaquin Valley, two separate efforts have moved forward simultaneously. The first is to find a long-term solution to the drainage problem and the methods to permanently dispose of the salty residue from the irrigated lands. The second is the implementation of measures to relieve the immediate drainage problem while awaiting the long-term strategy. Many of the methods studied and implemented to reuse and dispose of drainage water have been driven by regulatory requirements for water quality and environmental protection rather than considerations of long-term hydrologic and basin protection that must still be outlined in a long-term strategy. The measures the various sections of the San Joaquin Valley are pursuing best fit their immediate drainage needs, while an overall salt management strategy for the valley remains elusive. This chapter will review the steps the various sections of the valley have taken, both successful and unsuccessful, to solve their immediate drainage needs.

THE VALLEY The California Central Valley is a 650-km (400-mi)-long bowl that is bounded over its entire length by the Sierra Nevada Mountains (reaching more than 4,400 m (14,490 ft) in elevation) on the east side and the coastal mountain range (up to 1,830 m (6,000 ft) in elevation) on the west. The Central Valley covers 157,000 km2 (60,480 mi2) and is made up of two distinct and separate regions. The first is the Sacramento Valley [70,530 km2 (27,210 mi2)], which drains to the south and where the largest portion of the state’s developed water supply originates. The second is the drier San Joaquin Valley [86,240 km2 (33,270 mi2)], which drains to the north. River flows from the two valleys converge and form the Sacramento-San Joaquin Delta, which ultimately drains to the San Francisco Bay. The delta is one of the largest estuaries in the United States and is a maze of river channels, wetlands, and diked islands covering more than 2,590 km2 (1,000 mi2). Figure 32-1 provides an overall picture of the Central Valley of California. The delta also acts as a diversion and transfer point for the flows from the Sacramento Valley to the San Joaquin Valley, southern California, and other coastal cities. The diversion of water south from the delta into the San Joaquin Valley, as well as the extensive development of local surface and ground water, make the San Joaquin Valley one of the largest intensively farmed regions in the world, with more than 3.7  106 ha (9.1  106 ac) of irrigated cropland. The annual water use for irrigated agriculture in the San Joaquin Valley is estimated to be about 2.07  106 ha-m (16.8  106 ac-ft).

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

979

FIGURE 32-1. Location map for the Central Valley of California. The water supply and irrigation development in the San Joaquin Valley and its hydrologic characteristics are the principal reasons why the valley struggles with salinity management. The San Joaquin Valley is actually made up of two distinct and different hydrologic basins: the San

980

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Joaquin River basin and the Tulare Lake basin. Figure 32-2 provides a view of the San Joaquin Valley. San Joaquin River Basin The San Joaquin River basin includes the entire area drained by the San Joaquin River. The San Joaquin River flows from the Sierra Nevada Mountains near Fresno in an east to west direction. Near the valley

FIGURE 32-2. Location map for the San Joaquin River and Tulare Lake basins showing the major hydrologic features of each basin.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

981

trough, the river makes an abrupt turn north and flows 160 km (100 mi) to the Sacramento-San Joaquin Delta. The San Joaquin River, when it turns north, essentially divides the valley floor into east and west sides. The majority of the flow of the San Joaquin River is diverted to another hydrologic basin just as the river exits the Sierra Nevada Mountains near Fresno. A portion of the river below this point has little or no flow but, as the river moves on its route to the delta, it receives flows from several large eastside tributaries draining the Sierra Nevada Mountains, including the Fresno, Chowchilla, Merced, Tuolumne, Stanislaus, Calaveras, Mokelumne, and Cosumnes Rivers. All but the Cosumnes are highly regulated. Little tributary flow comes from the drier coastal range drainages on the west side of the valley. Tulare Lake Basin The second distinct hydrologic basin is the Tulare Lake basin, which comprises the 45,100-km2 (17,390-mi2) drainage area of the San Joaquin Valley south of the San Joaquin River drainage area. Major surface flows from east-side streams draining the west face of the Sierra Nevada Mountains occur in the Kings, Kaweah, Tule, and Kern Rivers. These flows are highly regulated and are almost completely utilized for irrigated agriculture and other uses. Surface flow from only a small portion of the Tulare Lake basin drains north into the San Joaquin River basin, and this only occurs during years of extreme rainfall or flooding. Imported water supplies are also brought into the Tulare Lake basin through a series of canals and aqueducts. These imported water supplies make up more than onehalf of the water supply used for irrigated agriculture in the basin. The Tulare Lake basin is essentially a closed hydrologic basin where all water supplies and the salts they contain—whether from natural runoff or imported supplies—remain in the basin and must be considered in all water management decisions. San Joaquin Valley Geology and Salinity The San Joaquin Valley is a large, relatively flat alluvial basin surrounded by mountains that are the primary source of both the valley’s water supply and alluvial material. The San Joaquin Valley is basically a horseshoe-shaped basin with three key geologic zones. The Sierra Nevada Mountains ring the eastern side of the valley and rise in elevation to 4,400 m (14,490 ft). The coastal ranges that line the western edge of the valley floor are much lower in elevation, up to 1,830 m (6,000 ft). The San Joaquin Valley floor makes up the third geologic zone and has received erosional debris almost continually from the mountains flanking the valley floor, with most of these deposits originating from the Sierra

982

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Nevada Mountains. These deposits are more than 305 m (1,000 ft) thick in most places. The geology of each of these mountain ranges has had a marked influence on the valley floor sediments and salinity. The deposits from the drainage of the western slope of the Sierra Nevada Mountains have created large alluvial fans of low-salinity, well-sorted gravels and sands on the eastern side of the San Joaquin Valley. This has resulted in coarsetextured alluvial material on the east side of the valley floor that is low in natural salinity. This coarse-textured material becomes finer as these alluvial deposits move toward the valley trough. In contrast, the coastal ranges are made up of Jurassic and Cretaceous sandstones and shales of marine origin. These are known to be high in salt. The lower rainfall on the western side of the San Joaquin Valley has resulted in poorly sorted sediments that, as a general rule, are of lower permeability and higher salinity when compared to those on the east side. Near the trough of the valley, the fine-sediment deposits from both the east and west sides of the valley grade into, and intermingle with, extensive masses of fine-grained sediments that were deposited in lakes and swamps. The area of lake and swamp deposition has shifted widely in the past in response to climatic and geologic changes, but has stayed primarily near the valley trough. The present valley floor is not homogeneous. Build-up has been through a succession of alluvial deposition from the surrounding mountains and from fluctuating lakes and swamplands. This stratified area and the higher soil salinity on the western side of the valley have led to salinity and drainage problems as irrigation development has progressed. Irrigation Development and Drainage Needs Prior to 1850 the San Joaquin Valley was devoted largely to rain-fed grain and cattle production. Irrigation development began sporadically in the decade following the 1850s, when individual farmers made diversions to lands lying immediately adjacent to the perennial eastside streams. Many of these early irrigation areas were already natural overflow lands that had been used for pasture prior to that time (CDWR 1965). Construction of the railroad through the San Joaquin Valley from 1869 to 1875 increased the demand for more intensive cultivation, as markets in the larger coastal cities were accessible to valley farmers. Large-scale irrigation began in the valley around 1870, and by 1880 almost 81,000 ha (200,000 ac) were planted to cereals and alfalfa (CDWR 1931). Development proceeded generally from east to west across the valley, though some lands along the valley trough were irrigated during this period. The most significant water diversions to the western side of the valley occurred in 1872 when the San Joaquin River was diverted through the

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

983

Miller and Lux canal system west of Fresno, near where the San Joaquin River turns north when it reaches the valley trough (CDWR 1965). By the 1890s and early 1900s sizeable areas in the trough of the San Joaquin Valley were being forced out of production by salt accumulation and shallow water tables. Much of this land laid idle until the 1920s, when development of reliable electric pumps and the energy to power them accelerated the expansion of irrigated agriculture by making available vast groundwater resources under the valley floor. This groundwater pumping lowered the water table in many areas (CSWRCB 1977; Ogden 1988). Large-scale irrigation diversions by cooperative ventures of individual landowners and by local water agencies extended water deliveries to additional land. By 1943, the San Joaquin Valley had more than 1.42  106 ha (3.5  106 ac) under irrigation, largely using groundwater. Declining groundwater elevations and the desire of landowners to bring new land into production, and the desire of the U.S. government to expand population settlement in the western part of the nation led to the formulation of several large-scale water resource development plans for importing water from outside the San Joaquin Valley. However, salt management was not included in any of these early plans. As late as 1949, the U.S. Department of the Interior, Bureau of Reclamation (USBR)’s Comprehensive Report on Planned Water Resources Development in the Central Valley (USBR 1949) made no mention of salt management (CSWRCB 1977). The only official reference to the problem was contained in these 1946 comments by the U.S. Department of Agriculture on the draft report: The plan does not discuss drainage or include costs relative to constructing or operating drainage systems. In the light of experience with lands that have been irrigated, we feel that properly integrated plans for drainage should be made a part of any proposed new irrigation development plans. (CSWRCB 1977)

The first consideration of salt management occurred by the 1950s and 1960s, as large-scale water development plans began to take shape and salt management plans began to be incorporated into the process. The USBR’s 1955 Feasibility Report for the San Luis Unit of the Federal Central Valley Project, in the west-central part of the valley, recognized the need for drainage and proposed an interceptor drain for the unit (USBR 1955). In addition, a California Department of Water Resources (CDWR) report to the California legislature recommended that the state study a “comprehensive master drainage works system” for the valley. The California State Water Plan prepared by CDWR included the concept of a “valleywide master drain” and, in looking at the southern portion of the valley, assessed the problem in 1957 by stating, “Drainage must be considered an

984

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

integral and indispensable part of the development and utilization of water resources. Adequate provision must be made, therefore, in the total program” (CDWR 1957). The San Luis Unit of the Federal Central Valley Project (CVP) and the State Water Project, each authorized in 1960, began delivering northern California water to agricultural lands in the southern San Joaquin Valley in 1968. Together, they were to provide water to irrigate about 405,000 ha (1.0  106 ac). The voter-approved State Water Project authorized state participation in a drain if drainage repayment contracts were signed by users. Public Law 86-488, authorizing construction of the federal San Luis Unit, also mandated either participation with the state in a master drain or construction of an interceptor drain to collect irrigation drainage water from the federal San Luis Unit service area and carry it to the delta for disposal. Unfortunately, no final proposal had been prepared and there was a recognized need to jointly study drainage needs with the state. This began a long process of developing a proposal for consideration. A second major change occurred during the same time period with the increased recognition that water pollution nationwide was limiting the beneficial use and development potential of many streams and rivers. Protection of the quality of existing water supplies was recognized as critical to the success of these and other projects that attempted to increase the intensity of water resource utilization. For the federal CVP, questions were being raised about the potential effects of the discharge of untreated agricultural drainage on the water quality in the delta and San Francisco Bay (SJVDP 1990). This concern was reflected in a rider to the CVP appropriation act by Congress in 1965, which stated, [. . .] the final point of discharge for the interceptor drain for the San Luis Unit shall not be determined until development by the Secretary of Interior and the State of California of a plan which shall conform with the water quality standards of the State of California as approved by the Administrator of the Environmental Protection Agency. (SJVDP 1990)

A similar rider has been included in every annual CVP appropriation act since that time (SJVDP 1990). In California, the Porter Cologne Water Quality Control Act was passed in 1970, followed by the federal Water Pollution Control Act in 1972 (Clean Water Act). Both of these landmark pieces of legislation required regulatory oversight of any salty drainage water discharges. This regulation, however, must be implemented under the policy direction of a basin water-quality control strategy and not through a piecemeal effort. Planning for this basinwide strategy began in 1972 with the completion of a Water Quality Control Plan (hereafter, the Basin Plan) in 1975 for each of the California hydrologic basins (CSWRCB 1975a,b) and continues to be updated (CSURCB 1995,

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

985

1998). These plans were approved by the U.S. Environmental Protection Agency (EPA) as part of the Clean Water Act requirements.

EFFORTS TO PROVIDE AGRICULTURAL DRAINAGE FOR THE SAN JOAQUIN VALLEY Over the past 60 years a number of studies have been conducted to determine the magnitude and the appropriate methods to manage the extensive saline subsurface agricultural drainage problem facing the San Joaquin Valley of California. Land drainage is well understood, and onfarm implementation proceeded smoothly early on. The principal issue is how to manage and dispose of the collected saline water in an economically and environmentally sound manner. Some of the studies on drainwater management and disposal have been localized, but many have attempted to look at the problem from a valley-wide perspective. Recent studies have looked at how to deal in the short term with continuing drainage problems, with a focus on the quality of the drainage water and how to minimize the impact of the various disposal methods being considered. The following description of the efforts to develop a drainage water disposal solution for the entire San Joaquin Valley illustrates the difficulty of preparing such a plan when there are various groups in the community both supporting and opposing such a project. It will become obvious to the reader that many of the decisions regarding the disposal of saline drainage water were based not just on technical considerations but also on policy, political, institutional, economic, social, and other factors—all of which complicate the effort. Early Attempts The earliest known attempts to deal with the build-up of a high water table and the associated drainage and salinity problems was the development of deep, open drains along the lowest portions of the western side of the San Joaquin River in the early 1900s. The increased flows from these open drains also required that ephemeral streams and sloughs leading to the San Joaquin River be enlarged to accommodate the additional drainage flows. As irrigation with imported surface supplies expanded upslope, the drainage problem also moved upslope (Johnston and Hall 1990). Just prior to 1950, operation of the federal CVP in the San Joaquin Valley was initiated with the completion of Friant Dam on the San Joaquin River. This facility transformed the hydrology of the San Joaquin River basin when the better-quality river water was diverted out of the basin at Friant Dam and sent south for use on the east side of the Tulare Lake basin. The river supply that was originally being used to irrigate lands on

986

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

the western side of the San Joaquin River basin was replaced by the lesserquality (but more reliable) supply from the Sacramento-San Joaquin Delta through what is known as the Delta-Mendota canal. This increased delta water supply to the western side of the valley expanded the intensively irrigated area. The drainage problem in the irrigated area immediately began to increase, and farmers on the west side of the San Joaquin River began to install on-farm drains to remove the shallow saline water table from the crop rootzone. The collected drainage water was discharged into open drains or canals and then either reapplied as irrigation water on farms or wetlands, or it continued down-gradient to eventually end up in the San Joaquin River. During the 1950s, as detailed planning for the San Luis Unit of the CVP commenced, the USBR recognized that there was a significant shallow subsurface drainage problem on the west side of the San Joaquin Valley. The USBR conducted a drainage study to determine the magnitude of the drainage problem in the proposed San Luis Unit, which was being designed to bring supplemental irrigation water to 160,000 ha (400,000 ac) on the valley’s west side, mostly located in the Tulare Lake basin just south of the San Joaquin River basin. The 1955 USBR feasibility report for the San Luis Unit stated that a drain would be necessary to export saline subsurface drainage water from the area. The feasibility report described the drain as “an earthen lined ditch with a discharge in the SacramentoSan Joaquin Delta” (USBR 1955). An interceptor drain was authorized to be constructed for the San Luis Unit when Congress authorized the San Luis Unit in 1960. The plans, however, did not remain static. Planning for a Joint State/Federal Drainage Facility The state of California also recognized the need to provide drainage to its service area in the lower San Joaquin Valley (Tulare Lake basin), and began studies on constructing a valley-wide drain. The state was also looking at a Sacramento-San Joaquin Delta discharge point. Because of the need for other areas of the San Joaquin Valley to be provided with drainage service, in 1962 the USBR changed its plans for the interceptor drain to a concrete-lined channel with a capacity of 30,600 m3 hr1 (300 ft3 sec1), almost three times the original proposal. The additional capacity was for service areas outside the San Luis Unit (USBR 1962). As specific plans for a drainage outlet began to take shape, opposition began to develop in regard to any drain-water discharge to the San Francisco Bay or the Sacramento-San Joaquin Delta and estuary. The U.S. Fish and Wildlife Service (USFWS), which was then known as the Bureau of Sport Fisheries and Wildlife, also expressed concern about potential toxic effects on the Sacramento-San Joaquin Delta and estuary from agricultural drainage water discharge, and requested the USBR to conduct studies on

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

987

agricultural drainage water and to consider evapotranspiration for disposal of the drainage water. In 1963, the USBR advised the Secretary of the Interior that it was evaluating potential environmental impacts of a drain discharge in the delta and suggested the possible use of evaporation ponds and/or an ocean discharge, either of which might require additional Congressional authorization (USBR 1978). In 1964, the CDWR indicated that it wanted to provide a master drain for the entire San Joaquin Valley. With the request by the state to work on a joint state/federal facility, planning for a separate drain for the federal San Luis Unit was suspended and a cooperative study for the Valleywide Master Drain started. The CDWR issued a preliminary edition of Bulletin 127 entitled San Joaquin Master Drain (CDWR 1965), which advocated the formation of a valley-wide drainage district to finance the construction and operation of the drain. This recommendation was met with much resistance within the valley. In spite of many efforts by the state to gain support for the formation of such a district, the resistance was too great and the district was never formed. An additional key change was made to the alternative plans for the San Luis Unit in 1964. A regulating reservoir was added along the drain alignment to temporarily retain drainage water (USBR 1964). The purpose of the regulating reservoir was to ensure control of the drainage effluent that would be discharged into the delta estuary. Another reason for adding a reservoir was litigation that had been filed by local irrigation districts lying downslope from the San Luis Unit. These districts insisted that drainage service be provided to the San Luis Unit before any irrigation water could be delivered, and the USBR needed the reservoir to prove to a federal court that drainage service could be provided, even though no outlet to a drain had yet been selected. After USBR officials assured the court that the San Luis interceptor drain would be constructed, and provided a construction schedule for the drain and the drainage collector system, the federal judge dismissed the drainage litigation without prejudice, meaning that if the USBR did not perform, the downslope districts could go back to court and ask for relief. Simultaneously, the USBR was proceeding with the construction of the irrigation and drainage systems for the Westlands Water District (WWD), the district that served the majority of 160,000 ha (400,000 ac) in the San Luis Unit. Potential Environmental Impacts Rise to the Forefront During the mid 1960s, considerable environmental controversy developed regarding the planned discharge of drainage effluent into the Sacramento-San Joaquin Delta and estuary. The principal areas of concern were the high nutrient levels of the drainage water and, in particular, the potential problem of excessive nitrates in that water. The former Federal

988

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Water Pollution Control Administration (FWPCA) (now the EPA), issued a report warning of the potential problem with excess nitrates in the drainage water (FWPCA 1967). To determine the potential impacts, the CDWR, the USBR, and the FWPCA initiated an interagency nitrate removal research program at Firebaugh, California, commonly known as the Firebaugh studies. The FWPCA issued the report on the Firebaugh studies, which concluded that nitrates could be sufficiently removed from the drainage water, so there would not be a nitrate problem in the delta if the drainage water was discharged there (CDWR 1968; U.S. EPA 1972). Even though the report showed no potential adverse effect, the controversy over the location of the discharge outlet and the potential waterquality impacts continued to escalate. Because of the environmental controversy, Congress required that the USBR and the state of California agree on a final discharge point for the drain and the water-quality standards that should be met before any construction of the drain north of the proposed regulating (evaporation) reservoir would be allowed. A rider to the CVP appropriations was added that prohibited selection of a final point of discharge for the drain until an agreement was reached with the state. A rider with similar, but not identical, language has been included in every annual CVP Congressional appropriation since that time. Initial Efforts to Complete a Valley-Wide Drain In the late 1960s the CDWR advised USBR that it would not participate in the construction or operation of a San Joaquin Valley master drain and that the USBR should proceed with completing a drain for only the federal San Luis Unit. This decision was partly based on the fact that the initial irrigation deliveries were being made to the San Luis Unit and drainage facilities would be soon needed. The USBR began to move forward with construction of a drainage facility in 1968 and, by 1975, had completed 192 km (120 mi) of collector drains in WWD—the first 136 km (85 mi) of the main drain from its southern-most point to a regulating reservoir within the San Joaquin River basin, and the first phase of the regulating reservoir called Kesterson Reservoir (SJVDP 1990). However, federal budget constraints and growing environmental concern about releasing the drainage water into the Sacramento-San Joaquin Delta halted work on the drain beyond the new Kesterson regulating reservoir. Kesterson Reservoir now became the terminal reservoir for any drainage water until an outlet for the drain could be identified. Unfortunately, the decision whether to begin construction of the San Luis Drain at the outlet, or complete the drain to the outlet, would begin a series of problems that still have not been overcome. With the stoppage of construction of the drain at Kesterson Reservoir, there was an urgent need for an agreement between the USBR and the

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

989

state of California with regard to a discharge point for a drain and the water-quality standards that were to be achieved with any drainagewater discharge. To set the agreement format, the USBR, the CDWR, and the California State Water Resources Control Board (CSWRCB) formed the San Joaquin Valley Interagency Drainage Program (SJVIDP) in 1975. The goal of the SJVIDP was to develop a plan for the conveyance and disposal of saline agricultural subsurface drainage effluent, with recommendations for implementing and financing a plan that was economically, environmentally, and politically acceptable. The objectives for the plan were to • Protect agricultural productivity in the San Joaquin Valley by providing the facilities necessary for disposal of the saline effluent from on-farm drainage systems, • Protect good-quality surface and ground waters of the valley from degradation by saline waters, and • Promote the beneficial use of saline drainage effluent. The final report of the SJVIDP recommended completion of the drain to a discharge point in the bay-delta-estuary area near Chipps Island, at the confluence of the Sacramento and San Joaquin Rivers (SJVIDP 1979). The CSWRCB issued Interim Guidance on Possible Waste Discharge Requirements for the Proposed San Luis Drain (CSWRCB 1981) in an effort to achieve an agreement with the USBR on the discharge point and the water-quality protection standards. The USBR started a technical “waste discharge study” that would be designed to allow it to obtain a discharge permit from the CSWRCB for a drain discharge as recommended by the SJVIDP (Hydroscience Inc. 1978; USBR 1982a,b; USBR 1983). Public meetings were held on the study plans and, in mid-1983, the CSWRCB notified the USBR that it concurred with the USBR study plans for completing the report of waste discharge for the San Luis Drain with certain recommended modifications. The CDWR, however, notified the USBR that the state of California would not participate in the development and financing of any San Joaquin Valley drainage facilities. The USBR continued to prepare for the discharge from the San Luis Unit without the participation of the CDWR in constructing and operating the facilities. Decisions on Drainage Issues Cause Concern A series of events occurred in the 1980s that set the tone for how drainage is still handled today. By 1980, the USBR had been providing irrigation supply water to the San Luis Unit for more than a decade without providing a drainage outlet. The high water tables and soil salinity were beginning to accelerate and worsen. The WWD drainage collector

990

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

system had been completed on 16,800 ha (42,000 ac) in the northern portion of the district, along with the San Luis Drain from WWD to the Kesterson Reservoir, but no on-farm drainage water had been allowed into these facilities. Under pressure from growers faced with high water tables, the USBR began taking water into the completed portions of the San Luis Drain and conveying the water to Kesterson Reservoir, which was now part of the newly formed Kesterson National Wildlife Refuge and was operated jointly by the USBR and the USFWS. Within 18 months of the decision to accept drainage, on-farm drains for 2,800 ha (7,000 ac) had been installed and/or connected to the main collector drains. Because of the drainage flows generated by these on-farm drains and the open-joint collector drains, the USBR advised that it would not allow any further on-farm drain hookups and would not approve any new connections to the drainage system because Kesterson Reservoir had reached capacity. In just a few short years the concept of using Kesterson Reservoir as a temporary terminal point for drainage water had hit its first snag. The reservoir was not capable of evaporating drainage water at a high enough rate to provide drainage relief to the full drainage service area. This was the first warning bell to be sounded. Soon after the decision to stop any further connections to the drainage collection system, the USFWS notified the USBR of high levels of selenium (Se) being found in mosquitofish (Gambusia affinis) within the San Luis Drain and in Kesterson Reservoir. The Se appeared to be in the drainage water going into Kesterson Reservoir. Within a few months, the USFWS began field studies on potential Se effects on waterfowl that were using Kesterson Reservoir or were in the vicinity of the reservoir. By June 1983, the USFWS notified the USBR about development of abnormalities in birds, including deformities and deaths of aquatic birds, at Kesterson Reservoir. This finding sparked a number of efforts by numerous state and federal agencies to assess the impact of these findings. Many meetings, discussions, and conferences were held regarding the findings at Kesterson Reservoir. In January 1984, the USBR notified the water-quality regulatory agency that excessive precipitation could overtop Kesterson Reservoir. In September 1984, a bird hazing program was also initiated at the reservoir and 1.85  103 ha-m (15  103 ac-ft) of clean water was provided for alternative habitat to minimize the impacts of Se on waterfowl. In October 1984, the CSWRCB assumed regulatory responsibility for the Kesterson Reservoir and all issues pertaining to the discharge of drainage water into the reservoir, and began a series of evidentiary hearings. At about the same time, the U.S Geological Service (USGS) issued a report (Deveral et al. 1984) that concluded that soils within the San Luis service area contained high concentrations of native Se that were leached by the application of irrigation water. In February 1985, the CSWRCB

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

991

adopted an order requiring the USBR to abate nuisance conditions at Kesterson Reservoir and develop and submit a clean-up plan for Kesterson Reservoir within 5 months (CSWRCB 1985). This action by the CSWRCB made the continued discharge of drainage water from the San Luis Drain questionable. On March 15, 1985, the U.S. Department of the Interior (DOI), the parent agency of USBR, announced the closure of Kesterson Reservoir and the discontinuance of the delivery of irrigation water to the 16,800 ha (42,000 ac) in WWD served by the drainage collector system because of concerns over violations of the Federal Migratory Bird Treaty Act that were occurring at the reservoir. Concurrent with the proceedings before the CSWRCB, WWD began looking at other options for the disposal of the water being transported to Kesterson Reservoir from the WWD drainage system (CH2M Hill 1985a). Once the CSWRCB adopted the Final Cleanup and Abatement Order for Kesterson Reservoir, the WWD filed a plan with the USBR to allow continued use of WWD on-farm drains (CH2M Hill 1985b). It was imperative to the farmers that the flow of irrigation water continue because substantial investments had been made in land preparation, fertilization, irrigation, and planting of fields in the drainage service area. Three weeks after the DOI announcement, on April 3, 1985, a negotiated settlement was reached. The WWD and the DOI agreed that WWD would stop the flow of drainage water into the San Luis Drain under a strict time schedule; in exchange, the USBR would continue the delivery of irrigation water to the area served by the WWD drainage collector system. The WWD was now faced with the immense task of developing an alternate plan to cope with the continuing flow of drainage water and no ability to put it into the San Luis Drain. The WWD announced that it would use diluted drainage water to irrigate salt-tolerant forage grasses as an interim method for the disposal of drainage water previously flowing into the San Luis Drain. WWD also would develop a plan for longterm disposal of the drainage water within the district. In June, the WWD announced a revised plan to store drainage water in storage basins, from September 30, 1985, through March 31, 1986. The plan was to have these basins serve as a pilot project for a proposed long-term evaporation basin project. A request for a waste discharge permit was submitted to the Central Valley Regional Water Quality Control Board (CVRWQCB) in support of the WWD proposal. The WWD announced that the following immediate steps would be taken: (1) irrigation of salt-tolerant plants would not be a part of the WWD long-term plan; (2) WWD authorized a contract to develop a longterm drainage plan; (3) WWD would seek to use the portion of the San Luis Drain within the WWD boundaries for drainage water transport; and (4) WWD would authorize up to $2 million to develop a drainage water management program.

992

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

After months of intensive work and detailed reporting (CH2M Hill 1985c; Jones and Stokes Assoc. 1985, 1986), and negotiations between the WWD, its consultants, the regulatory agencies, water suppliers, landowners, and political leaders, the district finally announced on September 25, 1985, that, in order to be able to comply with the April 3 agreement with the DOI, it would not construct evaporation basins. Instead, it would plug the drainage collector system and discontinue drainage service to land that was currently draining into the district’s drainage collector system. Even though the CVRWQCB issued a discharge permit for the evaporation of drainage water, the cost of building the evaporation basins exceeded the value of the land to be drained, and WWD was unable to obtain a discharge permit that would guarantee the basins could operate permanently. The WWD board of directors concluded that there was no alternative but to plug the collector drains to the degree necessary to achieve a zero flow in the San Luis Drain outside of the district boundaries. On January 21, 1986, the WWD board of directors adopted the final environmental impact report approving a drainage project, including water conservation, voluntary recycling, and the plugging of the collector drainage system. WWD landowners and the cities of Mendota and Firebaugh protested that the WWD environmental documents were inadequate, and they formally requested that WWD not plug the drainage system. Once again the lower-lying districts filed suit, this time in Fresno County Superior Court, challenging the drain plugging. After working with the protestors and trying to resolve the litigation, WWD commenced shutting off drainage pumps to reduce flow into the San Luis Drain, and announced a timetable for plugging of the collector drainage system. The district staff designed a plugging technique for the district’s open-joint drains that would essentially restore the shallow water table to the conditions that existed prior to the installation of the drainage system. The district decided that the drainage water collected by the on-farm drains would be the responsibility of the landowners. Over the next 3 months, a total of 113 earthen and steel plugs were installed to decommission the system. In addition, more than 500 observation wells were installed in order to monitor the status of the water table after the drains were plugged. To help some of the affected landowners, the WWD board of directors adopted a plan that allowed recycling of drainage water by individual farmers as an alternative to plugging some of the district drains. Five water users, farming about 5,520 ha (13,800 ac) of land, participated in the voluntary program. These water users received a financial incentive of $25 for each acre-foot of drainage water recycled. This program was only for a limited period of time, and eventually the landowners who participated in this program either plugged the on-farm drains to stop the production of drainage water or they discontinued farming the land as the soils became too saline

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

993

for agriculture. Drainage-water flow from the WWD system into the San Luis Drain was terminated by June 9, 1986, in accordance with the April 3 agreement with the DOI. In order to do everything possible to resolve the drainage dilemma, the WWD authorized a $6.6 million drainage treatment and disposal project to study a prototype Se removal plant and a prototype deep injection well. This program also failed to produce any positive results to allow the opening of any of the drains that had been plugged. Concurrently, the USBR prepared a series of regulatory documents pertaining to the closure and clean-up of Kesterson Reservoir, while conducting a comprehensive characterization and research program in partnership with Lawrence Berkeley National Laboratory aimed at finding an effective strategy for remedial action at Kesterson (Benson et al. 1990). Ultimately, the CSWRCB issued an order (CSWRCB 1988) requiring that the reservoir be filled with clean material to a height of 15 cm above the rising groundwater to eliminate aquatic food chain exposure to toxins at the reservoir. Approximately 8.0  105 m3 (1.0  106 yd3) of material was imported and used to fill 290 ha (713 ac) of the reservoir to an average depth of 25 cm (10 in.) (Benson et al. 1990). The site remains in this condition today. Reconsideration of Drainage Alternatives The 1983 discovery of deformities in and deaths of aquatic birds at Kesterson Reservoir altered the perception of drainage problems on the western side of the San Joaquin Valley. Previously, the emphasis for drainage management was on salt management and dealing with the immense volumes of salty water being generated by irrigated land drainage. Salt management continues to be the primary long-term issue in the valley for water-quality protection, but the findings at Kesterson Reservoir shifted the entire perception of this drainage water to one of toxin control. Selenium poisoning was determined to be the culprit; it was coming from the soils on the western side of the valley and was being mobilized by the irrigation practices. The USBR, in providing a briefing to federal and Congressional representatives on agricultural drainage and salt disposal in the San Joaquin Valley, said that in light of the Se problem in Kesterson Reservoir, the SJVIDP’s 1979 report recommendations would need to be reformulated. In 1984 the governor of California and the Secretary of the Interior announced the establishment of the interagency San Joaquin Valley Drainage Program (SJVDP) to conduct “comprehensive studies to identify the magnitude and sources of the (drainage) problem, the toxic effects of selenium on wildlife, and what actions need to be taken to resolve these issues.”

994

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

In February 1987 the SJVDP issued a document describing activities to date and planned activities for the next 4 years, describing the purpose of the program as: To investigate the problems associated with the drainage of irrigated agricultural lands on the west side of the San Joaquin Valley and to formulate, evaluate, and recommend alternatives for the immediate and long-term management of those problems. (SJVDP 1987a)

In April 1987 the SJVDP issued a draft report discussing potential ocean disposal of San Joaquin Valley drainage water (Brown and Caldwell 1987). In response to adverse public reaction to the Brown and Caldwell report, the SJVDP’s Policy and Management Committee narrowed the focus of the SJVDP by issuing the following directive: Recognizing and agreeing with the advisory action of July 27, 1987, by the Citizens Advisory Committee, the Policy and Management Committee requests the Program managers to focus investigative and planning efforts on in-valley solutions to the San Joaquin Valley drainage problem . . . No studies of out-of-valley disposal of agricultural drainwater, or concentrated brine, will be undertaken by the Program. (SJVDP 1987b,c)

This decision narrowed the focus of this study for a drainage solution to in-valley management of agricultural drainage problems and away from the achievement of a long-term, sustainable salt balance on the irrigated lands on the west side of the San Joaquin Valley. The SJVDP’s final report (SJVDP 1990) recommended a drainage management plan that provided for the management of subsurface drainage on the west side of the San Joaquin Valley through the use of • • • • • • •

Source control; Drainage reuse; Evaporation; Land retirement; Groundwater management; Discharge to the San Joaquin River; Protection, restoration, and provision of substitute water supplies for fish and wildlife habitat; and • Institutional change. The plan was presented as a framework through which drainage problems could be managed “for several decades,” and represented actions that would be necessary before any eventual export of salt from the valley. Farmers, districts, and state and federal agencies have been pursuing these drainage management components to varying degrees of success since that time.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

995

Coping with Drainage Needs in the San Luis Unit Service Area To date, all drainage service for the WWD (which comprises most of the federal San Luis Unit Service area) has been discontinued. A significant portion of the drainage problem area never did receive even temporary service. The farmers have been coping with the shallow saline water table problems either with on-farm water management through improved irrigation practices, continued use of on-farm drains where some evaporation techniques can be utilized, or, when necessary, by retiring land that becomes too saline to produce even the most salt-tolerant commercial crops. The use of evaporation basins is complicated because the CVRWQCB does not allow any standing water that will attract wildlife where the wildlife will accumulate sufficient Se to adversely affect reproduction.

DRAINAGE LITIGATION In December 1986, soon after the closure of Kesterson Reservoir and plugging of the collector system in WWD, a compromise settlement of long-standing litigation involving a number of issues between the federal government and WWD was entered and filed as a District Court stipulated judgment in Barcellos & Wolfsen, Inc. v. Westlands Water District (E.D.Cal.) (No. CV 79-106-EDP). The so-called Barcellos Judgment, among other things, required the federal government to develop a plan for “drainage service facilities” by December 31, 1991. Section 6.1.1 of the judgment stipulated that the facilities should have: [. . .] sufficient capacity and capability to transport, treat as necessary, and dispose of, the annual quantity of subsurface agricultural drainage water from the District (not less than 7.4  103 ha-m (60,000 acre-feet) and not more than 12.3  103 ha-m (100,000 acre feet)) required to be disposed of by December 31, 2007.

To help finance construction of the drainage service facilities, the district was required to make annual contributions into a trust fund established under the judgment. The SJVDP was scheduled to conclude its work in September 1990, but the scope of that program included the entire west side. It became clear that, as described, the SJVDP plan would be a broad framework strategy and not a plan upon which structures could be built. In 1989 the USBR initiated an effort to develop a drainage service plan specific to the federal San Luis Unit in parallel with and based on the ongoing work of the SJVDP. In December 1991, the USBR published the San Luis Unit drainage program plan formulation report and draft environment impact statement. The proposed action included site-specific actions for the federal San Luis Unit based on the recommendations in the SJVDP final report. It

996

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

included source control activities, controlled discharge to the San Joaquin River (utilizing small regulating reservoirs and an extension of the San Luis Drain) from the northerly San Luis Unit districts in the San Joaquin River watershed, and an adaptive technology development and demonstration program directed toward concentrating tile-drainage water, isolating and removing contaminants, and disposal of salts (USBR 1991). Following publication of the SJVDP’s final report in September 1990, a group of landowners in WWD filed claims for damages, based on the lack of drainage, against the WWD and the DOI in federal District Court on January 31, 1991. The suit was found to be premature since the Barcellos Judgment provided that a plan was to be provided in December 1991. Following publication of the USBR plan in December, the landowners again filed suit. On January 27, 1992, the WWD petitioned the court to reject the USBR proposed drainage plan on the grounds that it did not meet the terms of the previous court order. WWD also petitioned the court for the release of the drainage trust funds.

Initial Court Ruling The U.S. District Court for the Eastern District of California in Fresno, ruled on June 2, 1992, that the USBR had failed to develop, adopt, and submit a drainage plan that satisfied the requirements of the approved settlement agreement. The court held that “drainage service facilities” referred specifically to physical structures to be constructed to allow the drainage of subsurface water from the district, that is, an interceptor drain. The USBR plan was rejected and the WWD motion for release of the drainage trust funds was granted. Since 1992, litigation on drainage issues has continued. Prior to the June 1992 decision, in 1991 a group of landowners within WWD sued the USBR and other federal agencies for failing to provide drainage to the San Luis Unit of the CVP. This lawsuit became known as the “Sumner-Peck litigation” (No. CIV-F-91-048), named after one of the landowner groups. The Sumner-Peck litigation was later partially consolidated with a similar lawsuit brought in 1988 by neighboring Firebaugh Canal Water District and Central California Irrigation District (No. CIV-F-88-634) for the purpose of determining the federal government’s obligation with respect to drainage in the San Luis Unit. In May 1993, the District Court issued an order that held the following: The San Luis Act requires the Secretary (of Interior) to make provision for drainage for the San Luis Unit as specified in the Act. The failure to do so violates the Act. It remains to be determined, whether the Secretary’s duty has been excused, and if not, how the duty can be performed. (May 17, 1993, Memorandum Opinion and Order Re: Plaintiffs’ Motions for Partial Summary Judgment in Firebaugh Canal Water District, et al.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

997

v. United States, et al. Case No. CIV-F-88-0634 OWW DLB; Consolidated with CIV-F-91-048 OWW DLB)

In August and September 1994, a nonjury trial was held, addressing the following issues: (1) whether the obligation of the Secretary of the Interior to provide drainage for the San Luis Unit had been excused by factual or legal impossibility; (2) the extent of the court authority to order compliance with that obligation; and (3) what nonmonetary relief, if any, should be ordered if the court found the Secretary’s obligation had not been excused. The federal District Court judge issued a lengthy Findings of Fact and Conclusions of Law in December 1994, including a finding that: The Secretary of Interior through the Bureau of Reclamation has made the policy decision not to complete the San Luis Drain, in violation of Section 1 of the San Luis Act. This action constitutes agency action unlawfully withheld. (Case No. CIV-F-88-0634)

In March 1995 the judge issued an order that the USBR must take reasonable and necessary actions to promptly prepare, file, and pursue an application for a discharge permit for the San Luis Drain. Appeals Court Ruling In April 1995 the USBR initiated a process with the Central Valley Regional Water Quality Control Board to determine the permitting requirements, and, concurrently, the federal government appealed the District Court decision to the U.S. Court of Appeals for the Ninth Circuit. The appeal was heard on October 6, 1998, and, on February 4, 2000, the court issued its opinion. The Appeals Court affirmed the District Court’s conclusion that the federal government must act promptly to provide drainage service but reversed the part of the order directing the USBR to apply for a discharge permit, holding instead that the federal government has the discretion to provide drainage service by means other than an interceptor drain. The case was sent back to the District Court for further proceedings consistent with the Appeals Court ruling. On December 18, 2000, the District Court judge issued an amended order directing the Secretary of the Interior, through the USBR, to “without delay, provide drainage to the San Luis Unit” and to submit to the court a plan describing the actions it would take to promptly provide drainage to the San Luis Unit (203 F3d 568, 9th Cir. 2000). Settlement Negotiations Apart from the Declaratory and Injunctive Orders, the damage and other claims by the Sumner Peck plaintiffs and the adjoining districts continued to be pursued, and a series of settlement negotiations were held through the 1990s and onward. In late 2002, settlement of the Sumner

998

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Peck plaintiffs’ claims was concluded. Under the settlement, the plaintiffs received in excess of $100 million from the federal government and WWD, and approximately 37,100 acres of land were retired and nonirrigation covenants and drainage easements were placed on the lands. In April 2001, pursuant to the December 2000 court order, the USBR submitted a Plan of Action to the Court describing the steps it must take in order to provide drainage service, including completion of an Environmental Impact Statement (EIS) and Record of Decision under the National Environmental Policy Act. The USBR then undertook an evaluation of all reasonable alternatives for providing drainage service to the San Luis Unit. In identifying and formulating alternatives to meet the court order, the USBR identified certain project objectives that the alternatives should meet, including (1) the drainage service alternative should consist of measures and facilities to provide a complete drainage solution, from production through disposal, and avoid a partial solution or a solution with undefined components; and (2) the drainage service alternative must be based on technically proven components and be provided in a timely manner. Development of the Bureau of Reclamation Drainage Plan In formulating alternatives, the USBR determined the acreage of land that would require drainage service and determined a reasonable future drainage output from the federal San Luis Unit. The USBR determined that 1.21  105 ha (2.98  105 acres) in WWD, or almost half of the district, and about two-thirds of the northern San Luis Unit and adjacent lands (which is often referred to as the Grasslands Drainage Area) [3.28  104 ha (8.1  104 ac)], will require service. The estimated average annual output of drainage from the total 1.54  105 ha (3.8  105 ac) is 1.2  104 ha-m (9.7 104 ac-ft) per year. Although the area is, in general, already highly efficient in its water use, all alternatives include an estimate of additional reasonable, costeffective measures that could and are expected to be taken at the farm and district level to reduce the drainage output. It was estimated that these measures would reduce drainage output from the 1.54  105 ha (3.8  105 ac) to 8.64  103 ha-m (7.0  104 ac-ft) per year. Seven action alternatives were evaluated in the EIS. The alternatives were grouped by their final discharge location: delta, ocean, and in-valley evaporation. Four alternatives—delta discharge at one of two potential locations, ocean discharge, and in-valley evaporation—provide drainage service to all lands that require it, that is, 1.54  105 ha (3.8  105 ac). Three additional alternatives combine in-valley evaporation with varying levels of land retirement. Land retirement is defined as the removal of lands from irrigated agricultural production. This action would reduce drain-

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

999

water production and thus reduce the size of the in-valley treatment and disposal facilities. The three alternatives would cease irrigation on • 3.75  104 ha (9.3  104 ac), • 7.86  104 ha (1.95  105 ac), and • 1.39  105 ha (3.08  105 ac). and would reduce drainage production from 8.64  103 ha-m (7.0  104 ac-ft) per year for these alternatives to • 7.53  103 ha-m (6.1  104 ac-ft), • 5.55  103 ha-m (4.5  104 ac-ft), and • 3.33  103 ha-m (2.7  104 ac-ft). The USBR found it cost-effective in all alternatives to further reduce the volume of water requiring disposal through regional drain-water reuse areas. The collected drainage water would be transported to up to 16 regional reuse areas where the water would be applied to salt-tolerant crops and forages. Drainage water from the reuse areas would then be treated as necessary and disposed according to the alternative. For the ocean disposal alternative, water from the reuse areas would be transported and discharged approximately 2.25 km (1.4 mi) off the California coast near Point Estero at a depth of about 61 m (200 ft). For the delta disposal alternatives, water from the reuse areas would be processed through a biological Se treatment plant prior to discharge at one of two locations, near Chipps Island or in the Carquinez Straits. For the in-valley alternatives, water from the reuse areas would undergo reverse osmosis treatment producing about 50% clean/reusable product water. The remaining 50% more-concentrated water would undergo Se treatment prior to disposal in evaporation ponds. The USBR published a final EIS (USBR 2006) and in March 2007 issued a Record of Decision (ROD) (USBR 2007). The alternative recommended for implementation includes land retirement, other source control measures, a collector system, drainage-water reuse facilities, treatment systems, evaporation ponds, and mitigation habitat. About 1.54  105 ha (3.8  105 ac) require drainage service. Of these, a total of 7.86  104 ha (1.95  105 ac) of land in WWD and the former Broadview Water District would be retired (including lands already retired under litigation settlements). Lands identified for retirement would not receive drainage service but, instead, owners would be offered compensation in exchange for a nonirrigation covenant on their land. On lands remaining in production, farmers and districts are expected to continue to implement source control measures such as irrigation system improvements, seepage control, recycling, and shallow groundwater management.

1000

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The federal drainage system would be constructed in four sub areas: the Northerly (Grasslands) Area, WWD-North, WWD-Central, and WWD-South. A drainage collector system would be constructed in WWD; the Northerly Area collection system already exists. Drainage would be transported to the regional reuse areas (totaling about 5.06  103 ha (1.25  104 ac) where about 75% of the drainage water would be consumed by salt-tolerant plants. Drainage from the reuse areas would go to one of four reverse osmosis (desalting) treatment plants. About 50% of the inflow would come out as clean, reusable water and 50% as a concentrated brine. The brine would go to one of four Se treatment plants where Se is reduced to less than 10 ppb. The low-Se water would then go to evaporation ponds (totaling about 8.7  102 ha (2.15  103 ac). The USBR has completed feasibility-level cost estimates that confirmed the need for Congressional reauthorization in order to implement the plan. The San Luis Drainage Feature Re-evaluation Feasibility Report (USBR 2008) was completed in March 2008 and submitted to Congress in July 2008. This report concluded that the alternative selected in the ROD was technically and environmentally feasible but financially and economically infeasible according to federal planning guidelines, largely because the $2.7 billion cost was beyond the capability of San Luis Unit contractors (hereafter, the Contractors) to repay. Continuing Negotiations As of August 2008, the drainage litigation remained under the jurisdiction of the U.S. District Court in Fresno, California. Implementation of the drainage service plan developed by the USBR will require Congressional authorization. Concurrent with the planning process, and as the true cost of the recommended plan became apparent, in late 2006 the USBR initiated a collaborative process with water users and interested stakeholders. The process has been guided by seven basic principles: 1. Provide a timely solution to the drainage problem. 2. Sustain agriculture in the San Joaquin Valley. 3. Eliminate the drainage obligation and any potential liability to the federal government. 4. Provide benefits to the environment. 5. Eliminate the need for federal funding. 6. Avoid redirected impacts to third parties. 7. Ensure compatibility with the state water project operations. The current proposal (hereafter, the Proposal) developed through the collaborative process continues to evolve over time in response to key stakeholder interests. However, as of August 2008, the Proposal includes the following main components:

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1001

• Drainage obligation. The Contractors will assume full obligation for providing drainage service to the San Luis Unit service area and the Secretary of the Interior will be relieved of the statutory obligation to provide drainage service to the San Luis Unit. Further, all existing or future claims by the Contractors and landowners within the Contractor districts based upon the lack of drainage service in the past will be waived. • Drainage management plan. The Contractors will assume the same drainage obligation currently required of the United States for their respective service areas. In that regard, WWD, the other San Luis Unit contractors, the Firebaugh Canal Water District (FCWD), and the Central California Irrigation District (CCID) will, at their own expense, be responsible for managing drain water within their service areas in a manner that is generally consistent with the ROD. WWD will permanently retire a minimum of 100,000 acres of drainageimpaired lands in conjunction with the drainage management plan. WWD will retain discretion to retire more land from irrigated agricultural production than the current minimum 100,000-acre objective. In all cases, drainage service must comply with state and federal laws and regulations and applicable permit conditions. • Financial provisions. The USBR will be relieved of the obligation to pursue the selected alternative in the ROD, estimated at $2.69 billion in capital costs. Within 5 years of enactment of enabling legislation, the Contractors will pay the present-worth equivalent of their respective outstanding CVP capital repayment obligations, adjusted to reflect their assumption of the drainage obligation. Upon repayment of their adjusted capital obligations, the Contractors would no longer be subject to the full-cost pricing provisions of the Reclamation Reform Act (RRA) and the tiered pricing requirements of the Central Valley Project Improvement Act (CVPIA). The Contractors will continue to pay their allocated share of future capital costs, annual operations and maintenance costs, and all required CVPIA restoration fund charges. • Contracts. Existing contracts between the USBR and the Contractors will be converted from water service (“9e”) contracts to repayment (“9d”) contracts, which would continue without expiration so long as the Contractors continue to pay all applicable rates and charges and comply with all provisions of the contract. The amount of water delivered in any given year will continue to be adjusted under the terms and conditions of their contracts to reflect legal requirements, such as Endangered Species Act (ESA) requirements and orders of the CSWRCB . • Environmental water. WWD will reduce its contract quantity by 100,000 acre-feet and assign that water on a permanent basis to the USFWS to be managed for the benefit of fish, wildlife, and habitat restoration. Contingent upon enabling legislation, the USFWS will have broad authority to sell, purchase, exchange, and store water, like any other CVP “south-of-the-delta” water contractor.

1002

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The Contractors proposed a drainage management plan that is substantially similar to the federal alternative with the following exceptions: • Evaporation technology. The ROD identifies the use of evaporation ponds as the method for evaporating drain water. The WWD proposes instead to use a sprinkler evaporation system. • Land retirement. The WWD would retire a minimum of 100,000 acres rather than the 194,000 acres prescribed in the ROD. The land retirement element of the WWD proposal is adaptive. The WWD may ultimately take more acres out of irrigated agricultural than 100,000 acres, based upon progress in achieving drainage service objectives. • The Northerly Area Contractors, together with the FCWD and CCID, will implement the drainage plan in their area. They intend to implement the plan adaptively in order to take advantage of advancing treatment technologies and provide flexibility to adapt to changes over the proposed 50-year planning and construction period. The Contractors anticipate implementing the drainage management plan at substantially less cost than estimated under the federal alternative. The largest cost factors of the federal alternative are the land acquisition component (estimated at $660 million); evaporation ponds (estimated at $842 million); and mitigation requirement costs (estimated at $76 million), for a total of $1.6 billion. First, under the Proposal, there are virtually no land acquisition costs because the Contractors already own much of the land to be retired. Second, the sprinkler evaporation system is anticipated to be substantially less expensive than the evaporation ponds and associated mitigation requirements. Third, because of adaptive implementation and advances in technology, the contractors anticipate substantial savings on many of the elements proposed under the federal alternative. As with the federal alternative identified by the USBR in the ROD, implementation of the Proposal developed through the collaborative process will require that enabling legislation be passed by the Congress. In September 2010 the Commissioner of Reclamation wrote to California Senator Dianne Feinstein requesting assistance in developing “longterm legislative drainage strategy” that would accomplish the goals of the current administration (Micheal L. Connor, unpublished letter, September 1, 2010). Basically, the USBR is requesting legislation that will implement portions of the ROD outlined above but transfer all drainage responsibility to local control and require the USBR to stop delivery of CVP water to all parcels of land for which the districts fail to provide acceptable drainage service within a specified time frame (something the USBR has been unable to do for more than 40 years). The drainage issue for the San Luis Unit remains unsettled until Congressional action and/or final resolution in federal court.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1003

ON-FARM EFFORTS Throughout the course of the litigation-driven events described above, farmers and water districts in the federal San Luis Unit, as well as elsewhere on the west side of the San Joaquin Valley, have been coping with their drainage problems in varying ways depending on the specific circumstances of the farm or district. Following publication of the final report of the SJVDP, federal and state agencies initiated the San Joaquin Valley Drainage Implementation Program (SJVDIP) in 1991 to pick up where SJVDP left off, following-through on program recommendations (SJVDIP 1991). Through various partnerships between state and federal agencies, the University of California, water and drainage districts, and individual farmers, a number of efforts have been undertaken to develop and implement drainage management measures, including source control, drainage reuse, treatment, enhanced evaporation, and other approaches that were recommended in the 1990 SJVDP plan. These efforts are discussed in Chapter 20 and elsewhere in this manual.

DRAINAGE CONTROL OUTSIDE THE SAN LUIS UNIT SERVICE AREA With the early decision by the state to not participate in a Valleywide Master Drain, and the federal government having indicated that it was moving forward with construction of a drain to serve only the federal San Luis Unit service area in the central portion of the valley, many farming operations and water districts in both the northern San Joaquin River basin and the southern Tulare Lake basin found themselves either without a drainage service outlet or not being able to plan for participation in a joint facility. The immediate drainage needs were growing and interim solutions needed to be explored both at the farm level and at the drainage district level. Because each area had different hydrologic conditions, the northern and southern valley areas developed their own methods to cope with the immediate drainage dilemma. Southern Valley Area (Tulare Lake Basin) The state originally planned to develop a master drain extending from near the Buena Vista lakebed in the southern portion of the Tulare Lake basin to the delta, jointly with the federal service area. With the decision by the state in the 1960s to not participate in the Valleywide Master Drain, the federal government began moving forward with construction of a drain to serve only the federal service area in the central portion of the San Joaquin Valley. Many of the farming operations and water districts in the southern

1004

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

end of the Tulare Lake basin found themselves without a drainage service outlet and drainage problems continuing to mount. Because there was no drainage outlet available, large water and drainage districts and individual farmers began plans to develop and implement evaporation basins for their drainage water disposal. The first of these facilities, at the Tulare Lake Drainage District (TLDD), was ready to be built on 730 ha (1,800 ac) prior to the completion of the Basin Plan, the water-quality control planning effort being done by the state under the federal Clean Water Act. To not delay the implementation, the state set regulatory requirements on the facility prior to completion of the Basin Plan. The state regulatory requirements included: • Total containment of the drainage water in the basins • Provision of flood protection for a 100-year flood event • Use of a multicelled salt routing design that provides for final salt disposal in a single cell • Provision for final salt disposal within the basin area • Provision of access for fish and wildlife officials to prevent the basins from being a wildlife impairment • Provision of groundwater protection by interceptor drains or other means • Construction to be near or on the alignment of the proposed Valleywide Master Drain. When the water-quality control planning efforts were completed in 1975, the plan recognized the importance of drainage control and disposal for the southern end of the San Joaquin Valley. The plan states, “Salts in the drainage water could amount to approximately 55 percent of the total estimated agricultural salt impact on the Basin. Disposal of the drainage water must be in a manner that isolates the salts from the main groundwater body.” The Basin Plan recognizes that isolation may be accomplished by either an export drain or by a properly constructed evaporation basin. The policy in the plan that regulatory efforts “will promote the development of any system which yields a permanent solution to the disposal of salts.” It recognizes, however, that at the time the plan was developed the only permanent solution consisted of constructing an export drain because the capability of evaporation basins to protect water quality on a long-term basis was not known. The primary unknowns were associated with the ability to protect groundwater quality as salinity concentration in the basins increase, and the ability to permanently handle and store the continuing accumulation of evaporated salts. The policy stated, “Evaporation reservoirs will not be considered permanent solutions until documentation is sufficient to show adequate water quality and environmental protection” (CSWRCB 1975b; Hall 1986).

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1005

Under this policy, by the early 1980s there were 27 separate evaporation basins constructed and operating for the disposal of agricultural subsurface drainage water in the Tulare Lake basin. These basins covered 2,900 ha (7,160 ac) and ranged in size from 5 to 730 ha (10 to 1,800 ac). The total storage capacity was 8,500 ha-m (69,000 ac-ft) of drainage water. An additional 20,000 ha (50,000 ac) of evaporation basins were in the planning stages for the lower San Joaquin Valley. At that time, the only other known experience with evaporation basins outside of the Tulare Lake basin were the 2,070 ha (5,100 ac) of basins in Australia (van der Lelij and Flint 1984; Evans 1989) and a 1,200-ha (3,000-ac) basin with a storage capacity of 14,000 ha-m (113,500 ac-ft) in Texas. The latter was built to isolate and trap brine water seeping into the Red River (Rought 1984). Recent information shows that constructed evaporation basins are currently being utilized in the Indus River basin of Pakistan and India (Trewhella and Badruddin 1991) and the Syr Dar’ya and Amu Dar’ya River basins in central Asia (Micklin 1991; Tanji et al. 2002). Based on the negative experience and findings at Kesterson Reservoir in the mid-1980s, related to the problem of Se toxicity to wildlife, a reevaluation of all the constructed evaporation basins in the Tulare Lake basin was conducted by the regulatory agencies in the Central Valley in cooperation with the pond owners and fish and wildlife officials. Initially, a full water-quality survey of the evaporation basins was conducted with an emphasis on the trace element Se that was indicated as the main source of impacts at Kesterson Reservoir. Evaporation Pond Chemistry Selenium was found in the inlet flows to the evaporation basins, with a geometric mean Se concentration of 16 g/L (ppb), a level only 4% to 6% of that found at Kesterson Reservoir, which averaged about 300 g/L (ppb) (Skorupa 1998). The inflow salinity levels to the evaporation basins in the Tulare Lake basin were equal to or greater than those found at Kesterson Reservoir. In addition, several other trace elements were found in the inflows to the various evaporation basins (Chilcott et al. 1990; Westcot et al. 1993). The salinity concentrations in the inflow to the evaporation basins ranged from 6 to 70 dS m1, almost twice that of seawater (Chilcott et al. 1990; Westcot et al. 1993). Salinity in the evaporation basins consisted mainly of sodium, sulfate, and chloride salts. Salinity in the evaporated water in the individual cells in the basins ranged up to 388,000 mg/L, more than 10 times the 35,000-mg/L seawater salinity (Westcot 1988; Chilcott et al. 1990). Some of the basins were multiple cells and were operated in series, with the final cells producing hypersaline waters in which a series of evaporite deposits form as their solubility products are exceeded

1006

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(Smith et al. 1995). It is these evaporite deposits that will need long-term management or final disposal in a landfill. Trace elements found in the drainage water entering the basins included arsenic (As), boron (B), molybdenum (Mo), selenium (Se), vanadium (V), and uranium (U), with Se being of highest concern because of its potential toxicity to waterfowl (Westcot 1988; Chilcott et al. 1990; Westcot et al. 1993). The actual suite and concentration of trace elements varied by evaporation basin. The occurrence of trace elements in the inflow water and in the basin cells was strongly associated with the physiographic locations on the valley floor. For example, basins containing subsurface drain water from the alluvial fan areas of the western side of the San Joaquin Valley contained high concentrations of B and Se, whereas those in the lakebed areas of the valley floor trough showed strongly elevated levels of As but little Se. None of the trace element levels in the influent flows exceeded any known hazardous waste levels. By contrast, however, due to evapoconcentration, the concentrations of these trace elements in the hypersaline waters of the individual basin cells approached or exceeded hazardous liquid waste criteria. The exception was Se, which did not show any signs of evapoconcentration. The lack of evapoconcentration was seen even in the highest Se-concentration ponds. There are various ways which Se can be immobilized or converted in such a pond environment. The exact mechanisms are highly complicated and will require research to define (Tanji et al. 2002). None of the evaporate deposits contained Se, As, B, or Mo at concentrations that exceeded California’s hazardous solid waste criteria (Tanji et al. 1992). Researchers surmise that these trace elements in waters are somehow excluded, immobilized, or dissipated during the crystallization of salts and only accumulate in the liquid phase. It is this liquid phase of the evaporation ponds that needs special attention. Wildlife Concerns The San Joaquin Valley is an important component of the Pacific Flyway, the migratory bird route from Canada to Mexico and areas farther south. Because the lower San Joaquin Valley is highly arid, any open bodies of water (such as evaporation ponds) strongly attract waterfowl and become a magnet for resident species. Within the one or two decades after the evaporation basins were first constructed, they became a strong attractant to species such as American avocets, black-necked stilts, grebes, plovers, and numerous species of migrating geese and ducks. Many of these birds are attracted to the saline waters of the evaporation basins since they mirror their native habitat in other areas. The basins are also a strong attractant to migrating waterfowl because the ponds often contain

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1007

large populations of aquatic insects and other benthic macroinvertebrates, which are one of the primary food sources for migrating waterfowl. Since Se was found in almost all of the basin inflows, a second phase of the reevaluation of the constructed evaporation basins in the Tulare Lake basin was initiated to investigate whether any waterfowl or wildlife impacts that were seen at Kesterson Reservoir were also occurring at these ponds. The focus of this investigation was on Se because of the impacts it had caused at Kesterson Reservoir, even though the Se levels measured in the evaporation basins within the Tulare Lake basin were well below those measured at Kesterson Reservoir. Other constituents, including B, Mo, and As were not initially considered as they did not appear to cause reproductive failure or mortality as Se had (Ohlendorf et al. 1993; Ohlendorf 2002). However, researchers felt that exposure to this mix of trace elements could result in (1) increased mortality, (2) reduced growth or impaired condition, (3) reproductive impairment, (4) reductions in species abundance, or (5) cumulative effects when viewed with past, current, and future exposure, and that future research may be needed to identify which is or are most important (Hothem and Welsh 1994; Paveglio et al. 1992, 1997; Skorupa and Ohlendorf 1991; and Davis et al. 2008). It was initially theorized that the deleterious impact of drain-water Se on fish and water birds observed at Kesterson Reservoir resulted partly from the inflow Se concentrations that were about 300 g/L (ppb). For the majority of the basins in the southern San Joaquin Valley, the inflow concentrations were typically below 20 g/L (ppb). The pond surveys continued, however, as wildlife scientists felt that the threshold toxic level of Se in freshwater systems was something below 5 g/L (ppb). Wildlife surveys conducted on several of the evaporation basins in 1988 showed adverse impacts on water birds (reduced reproduction and embryonic deformity) (Skorupa 1998). An example is the Tulare Lake Drainage District (TLDD) facility, which is composed of three separate evaporation basins, each of which is fed by a distinct drainage water inflow. In the North TLDD evaporation basin, inflow water had a geometric mean of 1.8 g/L (ppb), while the two southern basins, the South TLDD evaporation basin and the Hacienda Ranch TLDD evaporation basin, had an inflow with a geometric mean of about 10 g/L (ppb). There were no significant adverse impacts on waterfowl (e.g., mallard, cinnamon teal, northern shoveler, ruddy duck, coot, eared grebe) and shorebirds (e.g., black-necked stilts, American avocet, snowy plover, killdeer) seen at the North TLDD basin. In contrast, both of the basins in the southern part of TLDD that received the higher Se concentration in the inflow water showed elevated levels of Se in bird eggs, liver, and muscle tissues. These levels were determined by wildlife officials to cause reproductive effects, especially to the black-necked stilts and American avocets (Skorupa 1998).

1008

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Due to the findings of wildlife harm at the evaporation basins, the regulatory agencies placed new requirements and compliance schedules on the pond owners and operators. These new requirements were to consider any factors that would affect wildlife, especially waterfowl. In addition to Se effects, the pond operators needed to consider impacts from nest flooding, entrapment of young birds in tires and other materials used for levee stabilization, and mortality resulting from vegetation control and maintenance activities. Impacts were to be considered for both individual evaporation basins and on a cumulative basis. The effects of evaporation basin operations and exposure of wildlife to elevated concentrations of waterquality constituents, with particular emphasis on Se exposure, were considered to be significant if they individually or collectively: • • • •

Increased mortality Reduced growth or condition Resulted in reproductive impairment Caused or contributed to substantial short- or long-term reductions in species abundance • Contributed directly or indirectly to substantial cumulative effects when viewed with past, current, and reasonably foreseeable future projects. The new regulatory requirements focused on discouraging wildlife use of the ponds and/or providing mitigation and compensation habitat for any unavoidable losses. Specifically, the actions asked for were “a program of management actions to reduce, avoid, and mitigate for adverse environmental impacts to wildlife.” To comply, the pond operators took various steps to modify the design and operation of the ponds to reduce or avoid adverse environmental impacts. Some of the most successful were: • Steepening the inside slopes of the individual basins to at least 3:1 to discourage shorebird feeding • Keeping the entire pond area and banks free of vegetation • Maintaining a minimum water depth of at least 60 cm (24 in.) • Removal of windbreak islands that attract nesting birds due to their isolated location • Removal of synthetic bank stabilization material, such as automobile tires and concrete chunks, that are used for nesting and shelter • Immediate removal of any sick or dying birds to keep avian diseases under control • Hazing of migratory waterfowl attracted to the ponds • Compliance monitoring of drain-water inflows and pond water and sediment for Se and salinity

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1009

• Biological monitoring of birds for abundance and any signs of toxicity • Development and management of compensation habitats for unavoidable losses. The severity of the impacts seen at the evaporation basins in the southern San Joaquin Valley (Tulare Lake basin) were not expected when looking at the Se water concentrations alone. The harm in many cases was as severe as or more severe than that seen at Kesterson Reservoir. Part of the explanation can be found in the environment of the Tulare Lake basin as compared to the San Joaquin River basin where Kesterson Reservoir was located. The evaporation basins in the Tulare Lake basin are unique in that they are aquatic islands in the middle of a semiarid to arid zone. As such, they are highly attractive to waterfowl that are resident in the area or are on the migratory Pacific Flyway. This effect can be seen when comparing the Kesterson Reservoir experience to the evaporation basins in the Tulare Lake basin. Kesterson Reservoir was built as a wildlife refuge in the middle of several thousand acres of other wetlands in the San Joaquin River basin. The evaporation basins in the southern portion of the Tulare Lake basin, conversely, were isolated facilities surrounded by thousands of hectares of irrigated agriculture and semiarid and arid grasslands. Highly contrasting environments, yet both had similar impacts on wildlife from Se and showed similar Se levels in the bird eggs, livers, and muscle tissues, even though the Se inflow concentrations at Kesterson Reservoir were up to 30 times the concentrations seen in the lower San Joaquin Valley. One theory for this similarity is that the birds in the Kesterson Reservoir environment had alternative feeding sources in the lower-Se adjacent wetlands. The evaporation basins, in contrast, did not have this dilution effect and many of the birds became resident of only that site. This is, in part, shown by the fact that the greatest impact occurred to semi-resident shorebird species, such as black-necked stilts and American avocets. Any development results in some type of unavoidable impact. Because of the unavoidable bird losses in some of the basins due to Se, the regulatory agencies required that compensation habitat become a part of the management of these basins. The goal of the compensation habitat was twofold. First, it was to provide equivalent habitat to make up for any unavoidable losses, and second, it was to provide a potential dilution effect for any waterfowl that used the evaporation basins for a temporary food source. An example of how this compensation habitat can be effective is the one constructed by the TLDD. TLDD was required to construct and operate a compensation wetland habitat, so it constructed a 124-ha (306-ac) wetland habitat to specifically provide safe foraging and nesting habitat for the shorebird species most affected by the evaporation basins, the black-necked stilts and American avocets (Tanji et al.

1010

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

2002; TLDD 2003). Some of the unique design characteristics included the following (TLDD 2003): • • • • • • • •

A gently sloped shore of 12:1 A water depth of 10 to 15 cm (4 to 6 in.) Deeper pond areas for macroinvertebrate production Isolated islands with both foraging and nesting habitat Control of predator access to the habitat area Ability to adjust the water levels in the habitat A blend of fresh and saline waters as source waters A flow-through system that allows for depth control, controlled evapoconcentration, flushing, and draining as needed

The performance of the constructed compensation habitat shows that proper design and operation of such habitat can promote shorebird safety and reproduction. When comparing the number of nest starts by avocets and stilts in the TLDD basins with those of the compensation habitat constructed to comply with regulatory requirements, the TLDD experience demonstrates that within a few years after commencing operation of the compensation habitat how the number of nesting starts of avocets and stilts at the evaporation basins declined to zero, due primarily to hazing and design modifications and more suitable safe habitat offered by the compensation wetland (Tanji et al. 2002; Davis et al. 2008; TLDD 2003). Another key factor in the success is the high percentage of nests classified as hatched or presumed hatched in the compensation habitat, 82% for avocets and 75% for stilts (Davis et al. 2008). These levels are not seen in natural systems due primarily to predation, which is excluded from the compensation habitat. Northern Valley Area (San Joaquin River Basin) The irrigation areas in the west side of the northern San Joaquin Valley were faced with the same challenge as the southern portion of the valley: how to move ahead in the absence of a Valleywide Master Drain and with the federal government intending to construct a drain to serve only the federal San Luis Unit service area in the central portion of the San Joaquin Valley. Although drainage problems had been present in this portion of the valley for years, overdraft pumping and limited surface water supply had made them manageable. The importation of water from northern California, however, changed the hydrology, and significant salinity and drainage problems appeared almost immediately. Using the River for Salt Management The difference between this northern area and the southern end of the valley was that the northern drainage area was in the San Joaquin River

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1011

basin, and the river provided an outlet for the drainage water and salts. The implementation of on-farm drainage began in the 1950s, prior to the passage of the federal Clean Water Act, and no initial regulatory controls were established. Figure 32-3 shows the rate of development of subsurface drainage for a 40,000-ha (97,000-ac) drainage problem area of the 168,000-ha (370,000-ac) Grassland watershed on the western side of the San Joaquin River basin. Development of the on-farm systems and collector systems was essentially complete by the late 1970s. During implementation of the on-farm drainage systems, the drainage flow rate to the San Joaquin River increased just as the flow in the river upstream was decreasing, because diversions were increasing at Friant Dam for irrigation on the eastern side of the lower San Joaquin Valley. These two actions, occurring simultaneously, resulted in a significant degradation of the river and prompted a declaration by the California legislature in 1961 that the river was impaired and that no further impairment shall occur; this declaration was made part of the California Water Code. The California Department of Water Resources conducted a full investigation of water quality in the San Joaquin River and concurred that water quality had deteriorated significantly (CDWR 1969). Salinity is affecting many river basins that are under intense water development or that receive return flows from irrigated agriculture, especially high-salinity subsurface drainage water discharges. For example, the

FIGURE 32-3. Development of on-farm subsurface drainage in the drainage problem area flowing through the Grassland watershed into the San Joaquin River.

1012

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

River Murray-Darling in Australia shows a typical progressive increase in salinity as it flows downstream. This pattern is seen in many river basins, such as the Nile, Euphrates, and Indus River basins where water-supply development is increasing the intensity of water use in the basin and where irrigated agricultural return flows enter the river system (O’Brien 1984; Westcot 1988). Management of drainage water inflows in this scenario is a matter of matching discharge rate to stream dilution flows. While the River Murray-Darling, the Nile, the Euphrates, and the Indus River basins and the San Joaquin River are all affected by salty subsurface drainage water discharges, the San Joaquin River shows a different pattern. Due to heavy diversions for irrigation supply upstream, the water quality of the river decreases immediately below the Grassland watershed subsurface drainage water discharge points at Salt and Mud Sloughs. The quality, however, improves downstream as better-quality water enters from several eastside tributaries (Fig. 32-4). It is estimated that subsurface drainage entering the San Joaquin River on an annual basis accounts for about 1% of the total river flow entering the delta estuary but accounts for approximately 17% of the total salt load (Oppenheimer and Grober 2004). Management of drainage-water inflows in this scenario is more difficult because discharge must be matched to potential flows from east-side streams. Often the two do not match and water-quality problems result. Under this scenario, some type of real-time management or release of the drainage flows to correspond with downstream dilution flows would be needed to avoid water-quality problems (Grober 1996). This is not easily done. This degradation in water quality is significant, and the Basin Plan for the San Joaquin River watershed recommends the construction of a separate collection and discharge drain for the subsurface drainage water to isolate it from the river system. To encourage the construction, the Basin Plan also recommended that it be done jointly with the federal service area for possible joint use of all facilities. Because the federal government was moving ahead with construction of a separate facility, the water users in the San Joaquin River Basin were faced with having to continue to use the river as its outlet for their drainage water and the salt it contains, since there was no planned capacity for these drainers in the federal facility for the San Luis Unit. Beginning in the early 1950s (Fig. 32-3) and continuing for almost three decades, farmers in the Grassland watershed discharged both their surface and subsurface drainage water into channels that led to the San Joaquin River. A large percentage of this water was captured by downslope wildlife refuge facilities, prior to the drainage water entering the river. The captured drainage water was combined with other water supplies and used to maintain portions of a 40,500-ha (100,000-ac) wetland area within the watershed. This wetland area represented one of the

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1013

FIGURE 32-4. Median salinity for June through September 1985 at various points along the San Joaquin River in the United States. largest remaining contiguous wintering waterfowl habitats on the Pacific Flyway. Focus Shifts from Salt Management The entire focus of water management in the Grassland watershed and in the San Joaquin River used to be on salt and salinity management, but the discovery of Se impacts at Kesterson Reservoir in the early 1980s

1014

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

changed this focus entirely. The experience at Kesterson Reservoir showed that subsurface drainage water containing elevated levels of Se likely caused sensitive waterfowl to show high levels of Se in their blood and tissues. Because the Grassland watershed drainage problem area was located directly adjacent to the lands that drained to Kesterson Reservoir, on soil types similar to those known to contain strongly elevated Se levels, a reevaluation of drainage discharges into the San Joaquin River was conducted by the regulatory agencies in cooperation with the drainage water managers and fish and wildlife officials. Initially, a full water-quality survey of the drainage water discharges was conducted with an emphasis on the trace element Se. Selenium concentrations in the on-farm subsurface drainage water ranged from less than 1 g/L (ppb) to 1,800 g/L (ppb) (Chilcott et al. 1988). The combined surface and tile-drainage flows from the drainage problem area that were being used by wetland managers had Se concentrations at levels approaching those found in the inflow to Kesterson Reservoir (Deveral et al. 1984; Gilliom 1986). Although not measured directly, the Grassland waterfowl area was likely experiencing a similar impact since drainage water being used to supplement the water supply for this valuable wildlife refuge area contained elevated levels of Se. Based on the new water-quality information, the Grassland refuge managers discontinued the use of blended Se-laden drainage water and diverted the subsurface and surface drainage flows directly to the San Joaquin River. As a result of these diversions, the Se loads to the San Joaquin River between 1984 and 1985 likely increased from about 1,180 kg (2,600 lbs) to 4,180 kg (9,200 lbs) per year, an increase of more than 300%. Selenium concentrations increased from near background levels (1.0 g/L) to levels consistently above 5 g/L, which was the US EPA water quality criteria at the time for protection of aquatic life. Figure 32-5 shows the estimated Se levels in the San Joaquin River for a normal water year based on differences in management of discharges of agricultural subsurface drainage water. This dramatic increase in river Se concentration prompted the regulatory agencies to begin the process of establishing water-quality objectives for the river to protect downstream beneficial uses and establishing a regulatory program to limit the Se loads in the drainage water discharges (CSWRCB 1986). Era of Regulatory Control Both the federal and state governments established a water-quality objective for Se of 5 g/L (ppb) for protection of all beneficial uses, including wildlife and aquatic life (CVRWQCB 1988, 1996). The state, however, set a 2 g/L (ppb) Se monthly mean water-quality criterion for all uses that involve wetlands and waterfowl habitat in the San Joaquin River Basin (CVRWQCB 1988). In 1988, the regulatory agencies in California

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1015

FIGURE 32-5. Estimated selenium levels in the San Joaquin River immediately downstream of the Merced River inflow for a normal water year and differences.

developed a program to begin to control the agricultural subsurface drainage discharges (CVRWQCB 1988) from the Grassland watershed. Several policy actions were taken, including: • The control of Se in the drainage water was set as the first priority. • The San Joaquin River could continue to be used to remove salts from the basin provided that water-quality objectives for Se were met. • Any further increase in drainage water discharges to the San Joaquin River from the Grassland watershed were prohibited until load levels were shown to be within the water-quality objectives. • Regulation of Se discharges would be pursued on a regional basis rather than at individual farms. • Reuse of drainage water would be encouraged. • A separate and isolated valley-wide facility to take drainage water out of the basin would continue to be promoted as the best longterm alternative. These policy actions were supplemented with a program of implementation that relied on voluntary actions by individual farmers improving their irrigation efficiency, with the goal of reducing the load of Se that each farm was discharging. The program had mixed success, for various reasons. The most significant were that individual farmers did not know the consequences of their actions on improving downstream water quality,

1016

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

and there was no overall responsible party who could direct the actions of the individual farmers or verify results (Westcot et al. 1996; Engberg et al. 1998). Initial Efforts at Regulation of Selenium Regulatory agencies normally use a permitting system to regulate the discharge of wastewater from individual dischargers. The focus of a permit is on pollutant reduction by source control and treatment at discrete points of discharge. These actions succeed because dischargers are successful at manipulating both source control and treatment to achieve the desired results at the point of discharge. At all times they know the results of their actions. Conversely, control of Se from an irrigated area is not as straightforward. These discharges are the result of an activity that is a legal practice: the irrigation of cropland. The pollutant Se originates from a natural source; it is either brought in with the water or is mobilized from the soil by the practice of irrigation. This is in contrast to a standard permitted discharge, which controls a pollutant that has been added to the water system. In this case, Se is not being added to the irrigation system or water. The source of Se is normally diffuse over a large area and is mobilized by the practice of irrigation. Control of this diffuse source of Se may or may not depend on the changes made to the irrigation practice. Regulators are also faced with the fact that the actions of one farmer alone may or may not affect water quality; the total effect often comes from the cumulative impact of many farm operations doing the same thing but none of which individually affects water quality. The manipulation of Se discharges in the Grassland watershed has not been easy. The source is not well understood; the mechanisms for control or reduction are modified agricultural practices that take time to implement and the responses from control actions are not well understood; and measuring success may take several years. A key element in this process is an overall responsible entity that can direct these efforts. Regional entities can then develop control mechanisms for nonpoint sources of Se by developing long-term solutions based on local and regional variability (Young and Congdon 1994). Two examples of Se contamination that originate from entirely different sources are the Grassland watershed of the San Joaquin River basin in the Central Valley of California and the Imperial Valley of the Salton Sea basin of southern California. The program to control these Se sources was developed for the local conditions but, in these two cases, the results were similar. In the Imperial Valley the Se is being brought in with the irrigation water supply from the Colorado River. It is concentrated by the practice

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1017

of irrigation and appears in the subsurface drainage water being discharged to the Salton Sea. The obvious control mechanism for this Se source is controlling the amount of Colorado River water brought into the area. Increasing irrigation efficiency would greatly reduce the total load brought into the irrigated area. The Se concentration of the Colorado River water supply is about 2 g/L (ppb) (Setmire et al. 1990). Increasing irrigation efficiency reduced the quantity of both the subsurface drainage flows and surface return flows, but increased the Se concentrations in the drainage canals to levels that often exceed the federal and state waterquality criteria for the protection of aquatic life, and this caused concern among regulators (CRRWQCB 1993). A different scenario exists in the Grassland watershed of the San Joaquin River Basin, where Se is not in the water supply but is in the soil that is being irrigated. In this case, the practice of irrigation mobilizes the Se. The options and costs to manage Se were outlined by the University of California through a series of reports focused on irrigation and drainage water management and improvement (UC Committee of Consultants 1988a,b) and the San Joaquin Valley Drainage Program (SJVDP 1987a), and were verified by field studies (Burt and Katen 1988). The regulatory action in 1988 to reduce Se loads entering the San Joaquin River focused on initial implementation of improved on-farm irrigation efficiency to reduce deep percolation, thus reducing Se mobilization into the drainage water. Immediately following this regulatory action, extended drought conditions led to water supply shortages, which, in turn, led to significant improvements in irrigation efficiencies within the San Joaquin River basin. The result was similar to the Imperial Valley experience. Discharges of Se to the San Joaquin River were cut in half and the amount of Se in the river water improved significantly (Karkoski 1994; CVRWQCB 1996). The Se concentration of the internal channels in the Grassland watershed that were dominated by these drainage flows, however, did not improve. In fact, water quality in the channels deteriorated as a result of water conservation and improved irrigation efficiency. The cause, like in the Imperial Valley, was that water conservation significantly reduced drainage flows but the concentration of constituents in the drainage increased even though the total load of Se discharged was less. With the return of the wetter cycles after 1993 and a full water supply to each farm, the total load of Se discharged went back up to the 1988 levels, indicating that the irrigation efficiency improvements alone would not provide a long-term solution. One of the major problems with the approach taken by the regulatory agencies was that individual farmers had no knowledge of or control over how their actions affected water quality downstream of their farming operations. It was found that, like in the Imperial Valley, as irrigation efficiency improved there was a significant reduction in the subsurface drainage

1018

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

flows but also a significant reduction in the high-quality surface runoff (irrigation tailwater) and operational spills from individual farms that had previously diluted the agricultural subsurface drainage flows. The result was that the discharge from the individual farms was smaller and the total load of Se was reduced, but flow was now dominated by the poor-quality subsurface drainage. At the district or broader level, improvements in how water was distributed and managed were also being made. However, the result was that operation spills and on-farm losses that previously had been available for dilution of the subsurface drainage water discharges were also no longer available. There did not appear to be any connection between the two operations—one at the farm level and one at the district level—nor was there any connection between these operations, flows in the collection channels, and river flows available for dilution. Present Regulatory and Compliance Efforts The lack of compliance with water-quality objectives prompted the regulatory agencies to reconsider the direction established in 1988. In 1990, the San Joaquin Valley Drainage Program (SJVDP 1990) outlined a three-step process to manage Se in the basin. The first step was to minimize Se mobilization from the irrigated area; the second was to capture and reuse the Se-laden drainage water to the maximum extent possible; and the final step was to isolate, treat, and/or dispose of the remaining Se-laden water. The regulatory agencies established a new approach for these nonpoint-source Se discharges in 1996. The new focus continued on source control efforts, including improved distribution and delivery efficiency, improved on-farm efficiency, and continued off-farm reuse of drainage water, but expanded efforts to control the final collected discharge runoff. The new focus would be implemented under a formal waste discharge requirement (WDR, the state equivalent of a permit) with monthly and annual load limits for specific water bodies. This was the first time a permit-type approach had been used to control or regulate a nonpoint source discharge. Because the mechanisms for controlling Se were not well understood, the use of such a permit was not intended to be at each individual farm but, rather, at the final discharge point from the Grassland watershed. The permit would be issued to a responsible entity that would have the administrative power to implement the load limitations. The permit would also prohibit the introduction of subsurface drainage water into channels that supply the wetland areas with water. It was also recognized in the 1996 decision that meeting final water-quality objectives in some water bodies would be difficult or impossible, especially in constructed canals, drains, or natural water bodies that were strongly drainage-water-dominated. The regulatory control policy also recognized

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1019

that not all actions could be accomplished at the same time. Wetlands and wetland water-supply channels were given the highest priority for protection, followed by protection of in-stream aquatic life in the San Joaquin River, and finally protection of in-stream aquatic life in the effluent-dominated tributaries (CVRWQCB 1996; Westcot et al. 1996). Wetlands and wetland channel protection in the Grassland watershed was accomplished by the Grassland water users through a series of reconstructed channels in the wetland area and diversion of the drainage flows into specified channels, including a 47-km (29-mi) portion of the then-idle San Luis Drain. Diversion into the drain, now called the Grassland Bypass Channel, allowed the drainage flows to be routed around the wetland area. This consolidation and rerouting removed the Se-laden flows from the 40,500-ha (100,000-acre) wetland area within the watershed and from 145 km (90 mi) of the channels serving this waterfowl habitat area (McGahan and Falaschi 2002). The result was a higher beneficial use of the wetland areas and wetland water-supply channels, because alternative freshwater supplies could now be brought in without mixing with the subsurface drainage flows, which contained elevated levels of Se and salts (Quinn et al. 2006). Figure 32-6 shows that Se concentrations in Salt Slough, a major wetland channel, were high before the implementation of the Grassland Bypass Project in September 1996 and dropped quickly after implementation. This pattern was echoed at other wetland water-supply monitoring points (CVRWQVB 2005). The rerouting has resulted in an overall higher benefit on a watershed basis, even though 10 km (6 mi) of the remaining channels carry almost all the subsurface drainage water and it will be difficult, if not impossible, to bring them to the same level of beneficial use. Wetland and wetland channel protection was then followed by efforts to achieve aquatic life protection in the San Joaquin River. Selenium load reductions were determined to be the method to achieve this goal (Karkoski et al. 1993; Karkoski 1994; CVRWQCB 1996). The regulatory approach for implementing load reduction is through establishing effluents limits for the entire watershed. The Grassland water users can then design their own methods of compliance. This approach allowed greater participation by the on-farm drainers and their local water and drainage agencies in deciding how to apportion loads among themselves to achieve the most cost-effective methods for compliance (Young and Congdon 1994). To ensure that aquatic life in the San Joaquin River would be protected, the use of the then-idle San Luis Drain as the Grassland Bypass was made contingent on compliance with monthly and annual Se load targets and the formation of a regional drainage management authority. Monthly Se loads were set based on the average monthly Se load for the drainage problem

1020

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 32-6. Monthly average selenium concentration at Station “F” (Salt Slough) for water years 1985 to 2006. The Grassland Bypass Project was implemented to reduce the selenium levels in Salt Slough and the surrounding wetlands and wetland channels. area during the 9-year (1985–1994) period prior to the project. The annual Se load was capped at 3,000 kg (6,600 lbs) per year, which was less than the sum of the monthly Se load targets. The loads were established with a high probability that water-quality objectives in the San Joaquin River would be met (Karkoski 1994). The Se loads were reduced in each year of the project with implementation over a period of time, beginning in 1997. The Se load reductions were nearly met in 1997, the initial year of the program. The following year, however, was extremely wet and the extreme rainfall mobilized Se in the drainage problem area. The irrigators had almost no control over this mobilization, and the agencies controlling the discharge rates to the San Joaquin River were overwhelmed by the ongoing flood flows and could do little to control the Se discharges. This extreme weather pattern was not seen during the 9-year period used to establish the Se loads. Nevertheless, after that extreme year the Grassland Bypass Project has been an unqualified success in reducing Se loads to the San Joaquin River. The load limitations have been met on an annual basis since that extreme weather pattern in the second year of the project (Fig. 32-7) (McGahan and Falaschi 2002; Quinn et al. 2006). In instances where Se load limits are used by regulatory agencies, a daily load limitation is set. For the Grassland watershed, however,

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1021

FIGURE 32-7. Grassland drainage area annual loads of selenium discharged to the San Joaquin River for the period 1986–2010 in comparison to the annual load limitations adopted by the Central Valley Regional Water Quality Control Board in 1996. because of the uncertainty in being able to control the discharges from a nonpoint source on a daily basis and the fact that Se is a bioaccumulative constituent whose impact occurs over a longer period, the regulatory agency established a monthly Se load limitation (Karkoski et al. 1993; Karkoski 1994). Figure 32-8 shows the monthly Se load limitations for Water Year 2005 and the actual discharges that occurred during each month of that water year. In this particular water year, the total annual Se load discharged from the watershed was below the limitation set by the regulatory agencies (Fig. 32-7), but the monthly Se load did exceed the load limitations in certain months (Fig. 32-8). While the annual Se load continues to exhibit an overall downward trend annually, there are instances where monthly Se load limits have been exceeded, as seen in Water Year 2005. These exceedances normally occur in the January–April period when several events occur, often simultaneously. Events such as spring pre-irrigations and periodic late, heavy spring rainfall can mobilize significant Se. Pre-irrigation is the application of irrigation water to cropland prior to planting a crop, and is used by farmers in the Grassland watershed to ensure that there is sufficient soil moisture available to establish a crop early in the growing season. These applications are made following winter rains and at a time when there are few crops available to

1022

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 32-8. Monthly selenium load discharged from the Grassland watershed in Water Year 2005.

reuse the drainage water generated. These conditions historically translate into an annual peak in Se loads discharged from the basin (Fig. 32-8) (McGahan and Falaschi 2002). On-Farm and District Efforts to Control Selenium To meet the Se load limitations, the farmers and water district managers in the Grassland drainage problem area have implemented the most aggressive source control and drainage management program ever conceived (Quinn et al. 2006). In the initial years, several physical changes were made to the drainage and irrigation systems to allow for more direct management of the subsurface drainage water. Continuous flowmeters have been installed at each main discharge point within the water and drainage districts. Telemetric water-quality sensors were installed, allowing realtime access to each district’s contribution to the overall drainage flow. Water meters were retrofitted on drainage sumps and discharge points within each district to determine the flow and mass of Se discharged. Water and drainage districts began to limit or prohibit the discharge of good-quality tailwater into the drainage system. In many instances, new plumbing systems were installed to allow a district to directly recirculate drainage water into the irrigation supply canals. Individual farmers also constructed tailwater return systems. In-line control weirs were

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1023

also installed on tile-drainage systems to regulate the flow. Where the tiledrainage systems discharged directly to a sump, sump control sensors were raised to allow discharge only when water tables in the field rose above a certain level (McGahan and Falaschi 2002). To allow farmers and others to monitor shallow groundwater levels, shallow monitoring wells were retrofitted with color-coded floating risers that protruded from the observation wells and were observable from the roadside. When the red-colored band on the riser showed, this indicated that water levels in the field were sufficiently high to affect crop yields. This device also provided indirect peer pressure on landowners whose water management practices contributed to excess deep percolation after irrigation (Quinn et al. 2006). This single practice was highly effective at improving on-farm practices. In the case where the tile drains discharged directly to open collection ditches, weir control structures were installed at the outlet to help store more drainage water beneath each field prior to discharge to the main drainage collection system (McGahan and Falaschi 2002). District policies also promoted improved on-farm management for Se source control. The techniques focused on improved irrigation practices and efficiency of water use; recycling of drainage water into the supply water at the farm and district levels during periods of the year when crop tolerance will withstand the additional salt; and tiered water pricing for both cropping-season supply and for pre-irrigations. The on-farm activities included an extensive program of education and outreach on improved water timing, especially during the pre-irrigation period of the year. Other subsurface drainage water management techniques include the use of drainage water for production of salt-tolerant forage crops, such as Bermuda grass, tall wheat grass, alfalfa, and other grasses used for grazing; temporary fallowing of land; and permanent land retirement. Most of these techniques are coordinated by the regional drainage water management agency. The use of each varies across the basin, but with Se discharges from the watershed being managed by an overall responsible authority, the Se load limitations are now being met (McGahan and Falaschi 2002). Future Efforts to Control the Remaining Selenium The most difficult step in the future will be to achieve aquatic life protection in the remaining 10 km (6 mi) of effluent-dominated constructed channels and natural waterways. Data show that Se load reductions alone will not be enough to achieve final water-quality objectives in these effluent-dominated systems (CVRWQCB 1996). The achievement of aquatic life protection in effluent-dominated channels will depend on the ability to develop technology to treat and manage Se in the drainage flows while

1024

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

maintaining the economic viability of the watershed. At present, some of the technology is not available, and the regulatory agencies have been forced to allow a reasonable time period for technology development and compliance. When compliance is difficult or technology not available, consideration must be given to the development of site-specific waterquality objectives for these water bodies in the Grassland watershed (CVRWQCB 1996). To move toward compliance in the effluent-dominated channels, the water users in the Grassland watershed are looking at expanding the area devoted to water reuse on forage crops. They recognize, as do others, that this is not a final solution; this method only concentrates the drainage water into a smaller volume. This smaller volume has a negative sideeffect—a higher concentration of salts and other constituents. While this may reduce loads in the long term, it makes handling and disposal of this water more difficult because of the increased concentration of both Se and salts. A number of experimental efforts to treat water for Se removal and evaporation of the final effluent are ongoing, but none has proved costeffective or reliable for Se removal. Nevertheless, the salt still needs to be dealt with on a basin-wide or more long-term basis. All of the efforts in the last two decades in the Grassland watershed, like other areas in the San Joaquin Valley, have been directed at reducing the impacts from Se in the drainage water and not at the long-term management of salt and salinity levels. As a result, the San Joaquin River still remains impaired due to salt discharges.

SUMMARY In the San Joaquin Valley, two separate efforts to address salinity and drainage issues have moved forward simultaneously. First, there has been an effort to find a long-term solution to the drainage problem and methods to permanently dispose of the salty residue from the irrigated lands. The second effort has been the implementation of measures to relieve the immediate drainage problem while awaiting the long-term strategy. Although there are some differences in the problems facing the southern/western portion and the northern/eastern portion of the San Joaquin Valley, these two issues are addressed in each region. The initial focus was on a classic drainage strategy involving leaching of salts from the soil and disposing of these salts via a canal that would discharge to San Francisco Bay and ultimately to the Pacific Ocean. This option was partially accomplished but remains unfinished due primarily to cost and environmental issues, and a lack of political consensus regarding the discharge of drainage water to the San Joaquin River delta. A 2008 agreement under which costs of any solution would be borne by growers

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1025

and water agencies addresses some of the institutional concerns. The 2008 agreement is a complex mix of financial and management commitments, including commitments for land retirement and reallocation of water saved by land retirement, to environmental uses. The 2008 agreement does not provide final direction on the nature and siting of a drainage solution. Pending implementation of a drainage solution, the viability of agriculture in the San Joaquin Valley has been largely maintained using the water management techniques described in this manual. Irrigation management, crop selection, enhanced drainage, land retirement, and better design of drainage evaporation ponds have, collectively, resulted in sustained profitable production and lower environmental impacts. There is a general recognition in the San Joaquin Valley that some means of disposing of accumulated salts, including toxic trace elements, will be required. All of the efforts in the last two decades in the San Joaquin Valley have been directed at reducing the impacts but not at the long-term management of salt and salinity levels. This fundamental issue must be addressed to ensure the productivity of San Joaquin Valley agriculture.

REFERENCES Benson, S. M., Delamore, M., and Hoffman, S. (1990). “Kesterson crisis: Sorting out the facts,” in Proc. National Conference on Irrigation and Drainage Engineering, July 11–13, 1990, Durango, Colorado; ASCE Irrig. and Drain. Eng., 119(3), 471–483; Report No. LBL-30587, Lawrence Berkeley National Laboratory, Berkeley, Calif. Brown and Caldwell Consulting Engineers. (1987). Screening potential alternative geographic disposal areas, prepared for the San Joaquin Valley Drainage Program by Brown and Caldwell, Sacramento, Calif. Burt, C. M., and Katen, K. (1988). Technical report to the Office of Water Conservation on the Westside Resource Conservation District, 1986/87 Water Conservation and Drainage Reduction Program, California Polytechnic University, San Luis Obispo, Calif. California Department of Water Resources (CDWR). (1931). San Joaquin River basin, Division of Water Resources Bulletin No. 29, prepared as part of the Reports on State Water Plan, California Department of Water Resources, Sacramento, Calif. ———. (1957). California water plan, Department of Water Resources Bulletin No. 3-35, California Department of Water Resources, Sacramento, Calif. ———. (1965). San Joaquin Valley drainage investigation—San Joaquin Master Drain, Department of Water Resources Bulletin No. 127, California Department of Water Resources, Sacramento, Calif. ———. (1968). Progress report on San Joaquin Valley drainage investigation—Quality and treatment studies through December 31, 1967, Department of Water Resources Bulletin No. 174-2, California Department of Water Resources, Sacramento, Calif. _____. (1969). Lower San Joaquin River water quality investigation, Department of Water Resources Bulletin No. 143-5, California Department of Water Resources, Sacramento, Calif.

1026

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

California State Water Resources Control Board (CSWRCB). (1975a). Water quality control plan (basin plan) for the Sacramento River and San Joaquin River basins and the Sacramento-San Joaquin delta basin, California State Water Resources Control Board, Sacramento, Calif. ———. (1975b). Water quality control plan (basin plan) for the Tulare Lake basin, California State Water Resources Control Board, Sacramento, Calif. ———. (1977). San Joaquin Valley Interagency Drainage Program environmental assessment—Phase I, prepared for the California State Water Resources Control Board by Environmental Impact Planning Corporation, San Francisco, Calif. ———. (1981). Interim guidance on possible waste discharge requirements for the proposed San Luis Drain, California State Water Resources Control Board, Sacramento, Calif. ———. (1985). Order No. WQ 85-1, and Clean and Abatement Order No. 85-1, California State Water Resources Control Board, Sacramento, Calif. ———. (1986). Regulation of agricultural drainage to the San Joaquin River, Final Report by the CSWRCB Order WQ 85-1 Technical Committee, California State Water Resources Control Board, Sacramento, Calif. ———. (1988). Order No. WQ 85-1, and Clean and Abatement Order No. 88-7, California State Water Resources Control Board, Sacramento, Calif. ———. (1995). Water quality control plan (basin plan) for the Tulare Lake basin, 2nd ed., California State Water Resources Control Board, Sacramento, Calif. ———. (1998). Water quality control plan (basin plan) for the Sacramento River and San Joaquin River basins, 4th ed., California State Water Resources Control Board, Sacramento, Calif. Central Valley Regional Water Quality Control Board (CVRWQCB). (1988). Amendments to the water quality control plan for the Sacramento River and San Joaquin River basins for the control of agricultural subsurface drainage discharges, Staff report, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. ———. (1996). Amendments to the water quality control plan for the Sacramento River and San Joaquin River basins for the control of agricultural subsurface drainage discharges, Staff report, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. ———. (2005). The Grassland Bypass Project status report, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. CH2M Hill. (1985a). Evaluation of alternatives to dispose of subsurface agricultural drainage water, prepared for Westlands Water District, Fresno, Calif. ———. (1985b). Preliminary plan report on alternative drainage disposal facilities, prepared for Westlands Water District, Fresno, Calif. ———. (1985c). Proposed plan of works for long-term drainage water storage and evaporation project, prepared for Westlands Water District, Sacramento, Calif. Chilcott, J. E., Westcot, D. W., Toto, A. L., and Enos, C. A. (1990). Water quality in evaporation basins used for the disposal of agricultural subsurface drainage water in the San Joaquin Valley of California, 1988 and 1989, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. Chilcott, J. E., Westcot, D. W., Werner, K., and Belden, K. K. (1988). Water quality survey of tile drainage discharges in the San Joaquin River basin, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1027

Colorado River Regional Water Quality Control Board (CRRWQCB). (1993). Salinity control in the Imperial Valley—An overview: Amendments to the water quality control plan for the control of agricultural drainage discharges, Staff report, CRRWQCB, Palm Desert, Calif. Davis, D. E., Hanson, C. H., and Hansen, R. B. (2008). “Constructed wetland habitat for American avocet and black-necked stilt foraging and nesting.” J. Wildlife Mgmt., 72(1), 143–151. Deveral, S. J., Gilliom, R. J., Fujii, R., Izbicki, J. A., and Fields, J. C. (1984). Areal distribution of selenium and other inorganic constituents in shallow ground water of the San Luis Drain service area, San Joaquin Valley, California: A preliminary report, U.S. Geological Survey Water Resources Investigation Report 84-4319, U. S. Geological Survey, Washington, D.C. Engberg, R. A., Westcot, D. W., Delamore, M., and Holz, D. D. (1998). “Regulation of irrigation-induced problems,” in Environmental chemistry of selenium, W. T. Frankenberger, Jr. and R. A. Engberg, eds., Marcel Dekker, Inc., New York, 1–25. Evans, R. S. (1989). “Saline water disposal.” BMR J. Aust. Geol. and Geophys., 11, 167–185. Federal Water Pollution Control Administration (FWPCA). (1967). San Joaquin Master Drain: Effects on water quality of San Francisco Bay and delta, U.S. EPA, Washington, D.C. Food and Agricultural Organisation of the United Nations (FAO). (1997). “Management of agricultural drainage water quality,” in C. A. Madramootoo, W. R. Johnston, and L. S. Willardson, eds., FAO Water Report No. 13, Food and Agricultural Organisation of the United Nations, Rome. French, R H., ed. (1984). “Salinity in watercourses and reservoirs,” in Proc. Conference on the State-of-the-Art Control of Salinity, July 1983, Salt Lake City, Utah, Ann Arbor Science, Boston. Gilliom, R. J. (1986). Selected water quality data for the San Joaquin River and its tributaries, California: June to September 1985, U.S. Geological Survey Open File Report 86-74, U.S. Geological Survey, Washington, D.C. Grober, L. F. (1996). “Sources and circulation of salt in the San Joaquin River basin,” in Proc., ASCE North American Water and Environment Congress, ASCE, Reston, Va., 649–654. Hall, S. K. (1986). “Evaporation basins: Environmental disaster or economic necessity?” Proc., USCID Conference on Toxic Substances in Agricultural Water Supply and Drainage: Defining the Problems, USCID, Denver, Colo., 27–35. Hothem, R. L., and Welsh, D. (1994). “Contaminants in eggs of aquatic birds from the grasslands of central California.” Arch. Environ. Contam. Toxicol., 27, 180–185. Hydroscience, Inc. (1978). Assessment of water quality impacts: Alternatives for discharge of San Joaquin Valley agricultural drainage to the western delta-Suisun Bay region— Final report, prepared for the State Water Resources Control Board, Hydroscience, Inc., Toms River, N.J. Johnston, W. R., and Hall, S. K. (1990). “Institutional and legal constraints,” in Agricultural salinity assessment and management, K. K. Tanji, ed., ASCE Manuals and Reports on Engineering Practice No. 71, ASCE, Reston, Va. Jones & Stokes Associates, Inc. (1985). Expanded initial study/ negative declaration Westlands Water District interim drainage water storage project, prepared for Westlands Water District, Sacramento, Calif.

1028

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

———. (1986). Final environmental impact report elimination of drainage water flow into the San Luis Drain, prepared for Westlands Water District, Sacramento, Calif. Karkoski, J. (1994). A total monthly load model for the San Joaquin River, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. Karkoski, J., Young, T. F., Congdon, C. H., and Haith, D. A. (1993). “Development of a selenium TMDL for the San Joaquin River,” in Proc. Management of Irrigation and Drainage Systems, ASCE Irrigation and Drainage Systems Division, Park City, Utah, July 21–23, 1993, ASCE, Reston, Va. McGahan, J. C., and Falaschi, D. (2002). “Innovative drainage reduction in the San Joaquin Valley, California, USA,” in Proc. International Committee on Irrigation and Drainage, Montreal, ICID, New Dehli. Micklin, P. P. (1991). The water management crisis in Soviet central Asia, The Carl Beck Papers No. 905, Center for Russian and East European Studies, University of Pittsburgh, Pittsburgh, Pa. O’Brien, T. A. (1984). “The problem of salinity and its control, River Murray, Australia,” in Salinity in water courses and reservoirs, Proc. 1983 International Symposium on State-of-the-Art Control of Salinity, R. H. French, ed., Butterworth, Boston, 33–42. Ogden, G. R. (1988). Agricultural land use and wildlife in the San Joaquin Valley, 1769–1930: An overview, SOLO Heritage Research, San Joaquin Valley Drainage Program, U.S. Department of the Interior, Sacramento, Calif. Ohlendorf, H. M. (2002). “Ecotoxicology of selenium,” in Handbook of ecotoxicology, 2nd ed., D. J. Hoffman, A. Barnett, G. Rattner, A. Burton, Jr., and J. Cairns, Jr., eds., Lewis Publishers, Boca Raton, Fla. Ohlendorf, H. M., Skorupa, J.P., Saiki, M.K., and Barnum, D.A. (1993). “Food chain transfer of trace elements to wildlife,” in Management of irrigation and drainage systems: Integrated perspectives, R. G. Allen and C. M. U. Neale, eds., Proc., ASCE National Conference on Irrigation and Drainage Engineering, ASCE, Reston, Va. Oppenheimer, E. I., and Grober, L. F. (2004). Technical total maximum daily load report for the amendments to the water quality control plan for the Sacramento and San Joaquin River basins for the control of salt and boron discharges into the lower San Joaquin River: Staff Report, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. Paveglio, F. L., Bunck, C. M., and Heinz, G. H. (1992). “Selenium and boron in aquatic birds from central California.” J. Wildlife Mgmt., 56, 31–42. Paveglio, F. L., Kilbride, K. M., and Bunck, C. M. (1997). “Selenium in aquatic birds from central California.” J. Wildlife Mgmt., 61, 832–839. Quinn, N. W. T., Linneman, J. C., and Tanji, K. K. (2006). “The San Joaquin Valley westside perspective,” presented at the World Environmental and Water Resources Congress, ASCE/EWRI Symposium, Omaha, Nebraska, May 2006, Lawrence Berkeley National Laboratory Paper LBNL-60613, ASCE, Reston, Va. Rought, B. G. (1984). “The southwest salinity situation: The Rockies to the Mississippi River,” in Salinity in water courses and reservoirs, R. H. French, ed., Butterworth, Boston, 115–124. San Joaquin Valley Drainage Program (SJVDP). (1987a). Farm water management: Options for drainage reduction, report by the Agricultural Water Management Subcommittee for the San Joaquin Valley Drainage Program, U.S. Department of the Interior and California Resources Agency, Sacramento, Calif.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1029

———. (1987b). Prospectus, U.S. Department of Interior and California Resources Agency, Sacramento, Calif. ———. (1987c). Policy and Management Committee meeting minutes for August, U.S. Department of the Interior and California Resources Agency, Sacramento, Calif. ———. (1990). A management plan for agricultural subsurface drainage and related problems on the westside San Joaquin Valley, California: Final report of the San Joaquin Valley Drainage Program, U.S. Department of the Interior and California Resources Agency, Sacramento, Calif. San Joaquin Valley Drainage Implementation Program (SJVDIP). (1991). A strategy of implementation of the management plan for agricultural subsurface drainage and related problems on the westside San Joaquin Valley, December 1991, U.S. Department of the Interior and California Resources Agency, Sacramento, Calif. San Joaquin Valley Interagency Drainage Program (SJVIDP). (1979). Agricultural drainage and salt management in the San Joaquin Valley—Final report, U.S. Dept. of the Interior, Bureau of Reclamation, California Department of Water Resources, and California State Water Resources Control Board. Setmire, J. G., Wolfe, J. C., and Stroud, R. K. (1990). Reconnaissance investigation of water quality, bottom sediment and biota associated with irrigation drainage in the Salton Sea area, California, 1986-87, U.S. Geological Survey Water Resources Investigations Report No. 89-4102, U.S. Geological Survey, Washington, D.C. Skorupa, J. P. (1998). “Selenium poisoning of fish and wildlife in nature: Lessons from twelve real-world examples,” in Management of irrigation and drainage systems: An integrated approach, R. Allen and C. M. U. Neale, eds., Proc. ASCE Specialty Conference, Park City, Utah, ASCE, Reston, Va., 315–354. Skorupa, J. P., and Ohlendorf, H. M. (1991). “Contaminants in drainage water and avian risk thresholds,” in The economics and management of water and drainage in agriculture, A. Dinar and D. Zilberman, eds., Kluwer Academic Publishers, Boston, 345–368. Smith, G. R., Tanji, K. K., Jurinak, J. J., and Burau, R. G. (1995). “Applications of the Pitzer equations-based model to hypersaline solutions,” in Chemical equilibria and reaction models, R. H. Loeppert, A. D. Schwab, and S. Goldberg, eds., Soil Science Society of America Special Publication 42, SSSA, Madison, Wisc., 113–141. Tanji, K. K., Ong, C. G. H., Dalgren, R. A., and Herbel, M. J. (1992). “Salt deposits in evaporation ponds: An environmental hazard?” Calif. Agric., 46(6), 18–21. Tanji, K. K., Davis, D. E., Hanson, C. E., Toto, A. L., Higashi R., and Amrhein, C. (2002). “Evaporation ponds as a drainwater disposal management option.” Irrig. Drain. Sys., 16, 279–295. Trewhella N. W., and Badruddin, M. (1991). Evaporation ponds for the disposal of saline drainage effluents in Pakistan, Netherlands Research Assistance Project, Lahore, Pakistan. Tulare Lake Drainage District (TLDD). (2003). Performance assessment of mitigation actions implemented at the Tulare Lake Drainage District evaporation basins: 1993–2001, Technical report prepared for Tulare Lake Drainage District for submittal to California Regional Water Quality Control Board Central Valley Region in Compliance with WDR 93-136, California Regional Water Quality Control Board, Central Valley Region, Fresno, Calif. U.S. Department of the Interior, Bureau of Reclamation (USBR). (1949). Central Valley basin: Comprehensive report on planned water resources development in the

1030

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Central Valley, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1955). Feasibility report for the San Luis Unit of the Federal Central Valley Project, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1962). Drainage appendix for San Luis Unit, Central Valley Project, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1964). Alternative solutions for drainage, San Luis Unit, Central Valley Project, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1978). Special task force report on San Luis Unit, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1982a). San Luis Drain plan of study, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1982b). San Luis Drain, Central Valley Project—California status of plans for completion of report of waste discharge, prepared for the California State Water Resources Control Board, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1983). San Luis Drain, Central Valley Project—California public input and USBR responses on toxicity and monitoring programs presented in status of study plans for completion of report of waste discharge, prepared for the California State Water Resources Control Board, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (1991). San Luis Unit drainage program, Central Valley Project, California: Draft environmental impact statement, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (2006). San Luis drainage feature re-evaluation, final environmental impact statement, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (2007). San Luis drainage feature re-evaluation, record of decision, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. ———. (2008). San Luis drainage feature re-evaluation feasibility report, U.S. Department of the Interior, Bureau of Reclamation, Sacramento, Calif. U.S. Environmental Protection Agency (U.S. EPA). (1972). Possibility of reducing nitrogen in drainage water by on-farm practices: Part of the collected papers regarding nitrates in agricultural drainage water, Report 13030ELY 5-72-11, U.S. EPA, Washington, D.C. (also published by the California Department of Water Resources as Report CDWR No. 174-14 and by the U.S. Department of the Interior, Bureau of Reclamation (USBR) as Report No. REC-R2-72-11). University of California Committee of Consultants on Drainage Water Reduction (UC Committee of Consultants). (1988a). Opportunities for drainage water reduction, Salinity and Drainage Task Force and the Water Resources Center of the University of California. ———. (1988b). Associated costs of drainage water reduction, Salinity and Drainage Task Force and the Water Resources Center of the University of California. van der Lelij, A., and Flint, S. E. (1984). “Salinity control problems associated with irrigation in south-west New South Wales, Australia,” in Salinity in water courses and reservoirs. R. H. French, ed., Butterworth, Boston, 265–274. Westcot, D. W. (1988). “Reuse and disposal of higher salinity subsurface drainage water: A review.” Agric. Water Mgmt., 14, 483–511.

SAN JOAQUIN VALLEY, CALIFORNIA: A CASE STUDY

1031

Westcot, D. W., Chilcott, J. E., and Smith, G. R. (1993). “Pond water sediment and crystal chemistry,” in Management of irrigation and drainage systems, an integrated approach, R. Allen and C. M. U. Neale, eds., Proc. ASCE Specialty Conference, Park City, Utah, ASCE, Reston, Va., 587–594. Westcot, D. W., Karkoski, J., and Schnagl, R. J. (1996). “Non-point source policies for agricultural drainage,” in Proc. North American Water and Environment Congress, Orange, Calif., E. Bathala, ed., ASCE Environmental Engineering Division, ASCE, Reston, Va., 83–88. Westcot, D. W., Rosenbaum, S. E., Grewell, B. J., and Belden, K. K. (1988). Water and sediment quality in evaporation basins used for the disposal of agricultural subsurface drainage water in the San Joaquin Valley, California, California Regional Water Quality Control Board, Central Valley Region, Sacramento, Calif. Young T. F., and Congdon, C. H. (1994). Plowing new ground—Using economic incentives to control water pollution from agriculture, Environmental Defense Fund Nonpoint Source Pollution Report, Environmental Defense Fund, Washington, D.C.

This page intentionally left blank

CHAPTER 33 INSTITUTIONAL AND SALINITY ISSUES ON THE UPPER RIO GRANDE Fred M. Phillips and Ari M. Michelsen

INTRODUCTION This is a case study of the upper Rio Grande basin from the perspective of hydrologic and institutional management issues, sources and impacts of salinity, and potential management alternatives. The chapter begins with an overview of Rio Grande physical, climate, and geographic conditions; water sources, flows and use; and water allocation, management, and institutions. This is followed by a discussion of growing salinity problems, Rio Grande salinity trends and causes, and potential management alternatives.

RIO GRANDE OVERVIEW The Rio Grande, known as the Rio Bravo in Mexico, is 1,900 miles (3,000 km) long, making it the fifth longest river in North America. Originating in the southern Colorado Rocky Mountains with peaks of more than 14,000 ft, the river flows south through central New Mexico and Albuquerque to the border cities of El Paso, Texas and Ciudad Juarez, Mexico. Downstream of El Paso, the river forms the international border between the United States and Mexico on its way to the Gulf of Mexico. The Rio Grande is hydrologically and jurisdictionally divided and managed as two river systems. The upper Rio Grande, the focus of this chapter, extends 700 miles (1,120 km) from its headwaters in Colorado to Fort Quitman, Texas, 100 miles southeast of El Paso. The reasons for the division of the Rio Grande into upper and lower basins are both physical and institutional. In the upper basin, spring runoff 1033

1034

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

from mountain snowpack is the dominant source of river water, accounting for approximately 70% of annual flow. The average annual surface supply produced in the headwaters is 1.57 million acre-ft but is highly variable and subject to prolonged periods of drought (Booker et al. 2005). Reservoirs were constructed to capture spring flows, and water is released primarily for agricultural use during the irrigation season. When the Rio Grande reaches the U.S.–Mexico border at El Paso, all of the water is diverted out of the river channel, except during flood events or canal maintenance, by the American Dam for U.S. water users and the International Dam for water users in Mexico (Mexican Irrigation District No. 009), resulting in reduced to little-or-no river flow except for minimal agricultural and municipal return flows and infrequent storm event flows below Fort Quitman, Texas to Presidio, Texas. At Presidio the Rio Conchos from Mexico joins the Rio Grande and, along with the Pecos River that joins below Big Bend National Park, they become the largest contributors to flows in the lower Rio Grande. Lower Rio Grande inflows are dominated by summer monsoonal precipitation, including inland remnants of Pacific and Gulf of Mexico tropical storms. These physical differences, combined with historical use and allocation agreements, contribute to the institutional and management division of the Rio Grande into the upper and lower basins. Topography, Climate, and Water Use The upper basin from Albuquerque south is characterized by basin-andrange topography. The climate is semiarid upstream of Albuquerque and arid downstream. Average annual precipitation exceeds 25 in. only in small areas and at the highest elevations. In more than two-thirds of the basin, precipitation is between 7 and 15 in. (Niemi and McGuckin 1997). In the Chihuahuan Desert of south-central New Mexico and far west Texas, average precipitation is less than 9 in. per year. Irrigated agriculture accounts for about 90% of Rio Grande water withdrawals, and return flows from irrigation are an important source of water supply. As discussed later in this chapter, irrigation return flows may not represent as large a contribution to salinity concentrations as has been previously hypothesized. The water supply in the upper Rio Grande is highly variable. Over the last 100 years there have been several prolonged periods of drought, most notably in the 1950s, late 1970s, and early to mid 2000s. During these periods runoff fell to less than 10% of average long-term river flows, and in 2003 the quantity of water stored in reservoirs fell below 5% of storage capacity (Michelsen and Cortez 2007). Salinity concentrations also become elevated during low flows. These water supply and salinity conditions affect all water users and have significantly influenced agricultural, urban, and environmental water-use patterns, planning, and management strategies.

SALINITY ISSUES ON THE UPPER RIO GRANDE

1035

Urban water demand in the upper Rio Grande basin, similar to much of the western United States, has been growing rapidly. Most of the population and urban water demand is concentrated in two areas, Albuquerque (with a population of 615,000 and elevation about 5,300 ft) and the border cities of El Paso (with a population of 700,000) and Ciudad Juarez (with a population of 1,400,000 with an elevation around 3,800 ft). Historically, most urban water use in the region has been supplied from deep aquifers, not the river. This is due to several factors, including the seasonal and uncertain Rio Grande flows, increased salinity and sulfate concentrations in the southern region of the upper basin, predominant agricultural ownership of surface water rights, and increased costs of surface versus groundwater treatment for potable use. However, the rate of recharge to major aquifers relied on by Albuquerque, El Paso, and Juarez is much lower than the rate of withdrawal, and/or they have levels of dissolved solids above drinking water standards, such as arsenic, limiting their use and supply sustainability (Niemi and McGuckin 1997). For this reason the cities have turned to the Rio Grande as an alternative source. Beginning in the 1940s, El Paso acquired river-water rights and constructed treatment facilities that can now provide more than 50% of annual demand in years when the river yields a full water supply. To provide water during drought periods and to meet increasing demands, in 2007 El Paso Water Utilities began operation of a brackish groundwater desalination plant capable of supplying 27.5 million gallon per day. This plant can provide up to 25% of annual urban demand, currently around 120,000 acre-ft. The source of the brackish water is the Hueco Bolson aquifer, shared by El Paso and Ciudad Juarez. Juarez is developing plans to convert its surface-water allocation from the upper Rio Grande from agricultural use to urban use, but there are political, hydrologic, and institutional issues that must be resolved. Albuquerque, which has relied completely on groundwater for urban use, has also acquired surface-water rights, both from the Rio Grande and from San Juan-Chama Project water imported from the Colorado River into the Rio Grande, and is constructing a diversion system from the Rio Grande and facilities to treat this surface water. These transfers of urban water sources from groundwater to surface water, and the associated effluent return flows, are likely to affect future Rio Grande salinity concentrations. Rio Grande Allocation and Management: Dividing the River Allocation and management of water resources in the Rio Grande involves two countries and numerous federal and state agencies and other organizations, highlighting the complexity of water supply, use, jurisdictions, and management of the river. Agencies and organizations involved in water resources of the region include: U.S. Bureau of Reclamation, U.S.

1036

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

International Boundary and Water Commission, U.S. Geological Survey, U.S. Environmental Protection Agency, U.S. Army Corps of Engineers, U.S. Department of Agriculture, North American Development Bank, Border Environment Cooperation Commission, Texas Commission on Environmental Quality, Texas Water Development Board, Texas State Soil and Water Conservation Board, El Paso County Water Improvement District No. 1, Elephant Butte Irrigation District, Hudspeth County Conservation and Reclamation District, El Paso Water Utilities, Las Cruces Water Utilities, Juarez Municipal Water Utility, National Water Commission, and Irrigation District No. 009 (Hanks et al. 2005). In the United States, allocation of Rio Grande water in the upper basin is governed by the Rio Grande Compact, a legal agreement signed by the states of Colorado, New Mexico, and Texas in 1938 and ratified by the U.S. Congress in 1939 (Hill 1974). Compact provisions specify how much water is obligated to be delivered to each state. Delivery obligations are not fixed amounts; they vary with flow conditions. Colorado deliveries to New Mexico under the Compact are measured at Lobatos on the Colorado– New Mexico state line. The delivery schedule defines Colorado’s water rights to be senior to those of New Mexico based on historical water use; thus, the Compact allows Colorado to use from 40% to 80% of its total runoff. New Mexico’s delivery requirement to Texas is on a sliding schedule based on New Mexico’s annual supply, defined as flows at the Otowi stream gauge north of Santa Fe, New Mexico (Ward et al. 2006). Article V of the Compact and a February 1948 resolution of the Compact oblige New Mexico to deliver water to Texas measured above Elephant Butte Reservoir in New Mexico, more than 120 miles upstream from the Texas border. The Texas water allocation is used by farmers in southern New Mexico and farmers and the city of El Paso in Texas. Rio Grande Compact requirements affect both the quantity of water available and salinity concentration. For example, water delivered to Texas is stored in Elephant Butte and Caballo Reservoirs, which have very high evaporation rates, reducing the quantity of water while correspondingly increasing in salinity. While the Rio Grande Compact does not include requirements on the quality of water delivered, the increased salinity of water delivered from New Mexico to farmers in El Paso County has been a major issue related to U.S. Bureau of Reclamation operation of releases and deliveries to the Elephant Butte Irrigation District (EBID) in New Mexico and El Paso County Water Improvement District No. 1 (EPCWID1) in Texas. Rio Grande Project Above El Paso, flow in the river is largely controlled by releases from Elephant Butte and Caballo Reservoirs. These two reservoirs, with a com-

SALINITY ISSUES ON THE UPPER RIO GRANDE

1037

bined capacity of 2.2 million acre-ft and associated infrastructure of 6 diversion dams, 139 miles of canals, 457 miles of laterals, and 465 miles of drains, are called the Rio Grande Project. The Project was authorized in 1905 and was one of the first constructed by the Bureau of Reclamation. Farmers in the Elephant Butte Irrigation District with 90,640 water-right acres, and the El Paso County Water Improvement District No. 1 with 60,100 acres, agreed to repay the cost of the Project through assessments on their land. After the debt obligation to the Bureau was repaid, ownership of the diversion dams, canals, laterals, drains and rights of way were turned over to the irrigation districts. The Bureau retained ownership and operation of the reservoirs and releases. There has been substantial litigation over water-right ownership between the Bureau and irrigation districts, between the states claiming excess use of river water from alluvial groundwater pumping, and disagreements over accounting and operation of the Project and over delivery of lower-quality water to Texas consisting of return flows rather than higher-quality Project water releases. In February 2008, after many years of litigation, many of these issues were resolved with the signing by USBR, EBID, and EPCWID1 of a Rio Grande Project Operations Agreement. In a full-water-supply year, 790,000 acre-ft of water are released from Project reservoirs. When combined with additional groundwater pumped by EBID and the city of Las Cruces and irrigation and urban return flows, total deliveries from the Project are 930,000 acre-ft. This includes 60,000 acre-ft delivered to Mexico under the Convention between the United States and Mexico for the Equitable Distribution of the Waters of the Rio Grande, often referred to as the 1906 Treaty (U.S. Department of State 1906). The U.S. Section of the International Boundary and Water Commission (USIBWC) is the federal government agency responsible for implementation of land and water treaties between the United States and Mexico. Mexican flows are delivered at the International Dam that crosses the Rio Grande between Juarez and downtown El Paso. The 1906 Treaty does not stipulate the quality of water to be delivered. Salinity Concern and Impacts Water quality, particularly salinity, is a large and growing concern in the upper Rio Grande basin, especially in the Rio Grande Project area. On average, each year approximately 400,000 tons of salt dissolved in the water are released from Elephant Butte Reservoir to users (Miyamoto et al. 1995). Geologic sources and agricultural and municipal use in the Project area contribute additional salts to the river (Phillips et al. 2003; Hogan et al. 2007). During periods of low flows and in the winter months, the salinity of the Rio Grande at El Paso at times exceeds drinking water standards of 1,000 ppm (mg/L) total dissolved solids (TDS) and/or 300 ppm of sulfate.

1038

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

When this occurs the river cannot be used as a municipal supply source and water treatment plant diversions for the city of El Paso cease until the water quality improves. High salinity adversely affects agricultural production and urban landscaping in this part of the basin. The high salinity levels reduce crop yields and restrict the selection of crops that can be grown. The surface and alluvial water quality deteriorate in the downstream direction from Elephant Butte Reservoir to Fort Quitman. The major crops in the region are pecans, pima, and acala cotton, and alfalfa. While lettuce, chile, and onions are important in the upper portion of the Rio Grande Project, these salinitysensitive crops gradually drop out of the crop mix as the river flows downstream through El Paso County and Hudspeth County. Return flows and drainage water from Project lands also provide a supplemental supply to agriculture below El Paso County for about 18,000 acres in Hudspeth County, Texas. Soil salinity in riverbanks and levees along the Rio Grande in the project area has been measured at more than 30,000 ppm, exceeding seawater concentrations (S. Miyamoto, unpublished data). One concern over upper basin water quality is that as salts from the upper Rio Grande are flushed downstream and join with Pecos River and Rio Conchos flows, the salinity in Amistad Reservoir below Big Bend has increased and is approaching drinking-water quality limits (Miyamoto et al. 2005, 2006). In a study conducted for the Rio Grande Project Salinity Coalition, Michelsen et al. (2009) calculated that preliminary estimates of the economic damages (cost) of salinity in the Rio Grande Project area, from San Acacia, New Mexico, to Fort Quitman, Texas, was $10.2 million per year. Of this amount, urban economic impacts account for 76% of total damages and agricultural damages account for the remaining 24% of total damages. The highest single category of damages is residential, 42% of the total, followed by agricultural, commercial, and urban landscape. However, this preliminary study did not account for several important factors such as salinity-induced changes to production of lower value crops, seasonal variability in salinity concentrations, environmental impacts and future damages from increasing growth in urban populations, and use of Rio Grande water in the region.

Salinity of the Rio Grande: Trends and Causes Salinity variation with flow distance The headwaters of the Rio Grande lie in the San Juan Mountains of southern Colorado. The runoff that sustains flow in the river originates largely from snowmelt off of peaks that extend up to 14,000 ft in elevation. This runoff is quite dilute, averaging ⬃40 mg/L TDS at Rio Grande Reservoir. However, by the time the river has reached El Paso, Texas, 1,000 km

SALINITY ISSUES ON THE UPPER RIO GRANDE

1039

to the south, the TDS averages 700 mg/L during the irrigation season. One-hundred and fifty kilometers farther south at Fort Quitman, the TDS varies between 2,000 and 5,000 mg/L (Fig. 33-1). In the northern portion of the drainage basin the quality of the water can be characterized as very good, and the water is suitable for all but the most demanding uses. At the southern end of the upper basin the water quality is poor, and uses are limited to water for livestock and irrigating a few salt-tolerant crops. This large transformation raises several practical questions: What causes the increase of salinity? What role does agriculture play? Are salinity problems increasing? Can anything be done to mitigate the salinization? Salinity problems on the Rio Grande have undoubtedly been exacerbated by human activities. Significant diversion of water for irrigation probably began in the thirteenth century when Pueblo Indian groups moved from the highlands to the floodplain of the river (Clark 1987). These diversions, although small relative to the discharge of the river, were substantially increased by population expansion in the centuries following the Spanish conquest in 1598. However, the amount of water that could be diverted was still severely limited by the difficulty of maintaining diversion structures on a river subject to uncontrolled annual floods. This difficulty was at least partly overcome by industrial-era technology for

FIGURE 33-1. Total dissolved solids (TDS) concentration of the Rio Grande during 2000 and 2001. 2001 was the beginning of a major drought and the drought was well advanced in 2002. Distances are downstream of the Rio Grande Reservoir in Colorado. Data from Mills (2004).

1040

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

water diversion structures that could be imported after the penetration of railroads in the late nineteenth century. The subsequent expansion in irrigated acreage substantially increased the consumptive use of Rio Grande water and also highlighted the need for water storage to alleviate shortages during the later part of the irrigation season. In response, Elephant Butte Dam was completed in 1916, with smaller reservoirs being constructed both upstream and downstream in later years. Evaporation from these reservoirs and from the additional acreage that was put in cultivation when a reliable water supply was provided, substantially increased the consumptive loss of Rio Grande discharge. Most recently, expansion of urban populations and industry during the late twentieth century have resulted in greatly increased pumping of groundwater and discharge of wastewater to the river. All of these activities have the potential to progressively increase the salinity of the Rio Grande. Recognition of salinity problems on the Rio Grande came early in the application of agronomic science in the region. Major solute concentrations in Rio Grande irrigation water were among the first water-quality analyses to be reported for New Mexico (Goss 1900), shown in Fig. 33-2.

FIGURE 33-2. Monthly average TDS concentrations at El Paso, for 1893–1894, 1905–1907, 1930–1939, and 1990–1999. Note that the period of record for the first two intervals is only 1 to 3 years. Data for 1893–1894 are from Goss (1900), for 1905–1907 from Stabler (1911), for 1930–1939 from Wilcox (1968), and 1990–1999 from USGS (2008).

SALINITY ISSUES ON THE UPPER RIO GRANDE

1041

At the time when the construction of Elephant Butte Dam and the irrigation project associated with it were being considered by the federal government, the U.S. Geological Survey performed extensive water-quality analyses of the river above and below the proposed dam site (Stabler 1911). In spite of the precautions taken during the development of the Rio Grande Project (Elephant Butte Dam), increasing salinity problems were noted during the subsequent decades, with significant portions of the Rio Grande valley being rendered unfit for cultivation due to salinization (Clark 1987; Wozniak 1987). As a result, evaluation of the water quality throughout most of the Rio Grande basin above El Paso was an important part of a massive Department of the Interior investigation (Natural Resources Committee 1938; Scofield 1938). This study was immediately followed up by a U.S. Department of Agriculture Salinity Laboratory research project that lasted 30 years (1934–1963) (Wilcox 1957, 1968). Continuing problems led the U.S. Environmental Protection Agency to investigate salinity sources in irrigation return flows during the 1970s (Trock et al. 1978). Finally, in the twenty-first century, more geochemically oriented approaches were employed by the U.S. Geological Survey and the Center for Sustainability of Semiarid Hydrology and Riparian Areas (Anning et al. 2006; Hogan et al. 2007; Phillips et al. 2003). These various studies have produced a range of conclusions regarding the cause of the progressive salinization of the Rio Grande along its course. J. B. Lippincott (1939), a prominent agricultural engineer of the early twentieth century, evaluated the data then available and concluded that “The increase in salinity of the waters of the Rio Grande [is] due to their use and re-use [for irrigation] in its long drainage basin . . .”; in other words, that evapotranspirative loss accompanying irrigation concentrated the salts naturally present in the river water. L. V. Wilcox, who conducted the 30-year salinity study for the U.S. Agricultural Research Service, in 1957 wrote, “There is a relatively large increase in the tonnage of both sodium and chloride from the upper to the lower stations . . . [that can be] attributed to the displacement of salty groundwater in the course of irrigation and drainage operations.” He did not describe the source of the salty groundwater. Trock et al. (1978) ascribed the salinization to leaching of soil salts during irrigation and evapotranspirative consumption of the river water. Salinity dynamics Figure 33-3 illustrates changes in the TDS concentration, load, and discharge of the Rio Grande upper basin with both position downstream and time. For all of the decades illustrated, the data show an increase in TDS concentration with distance downstream. However, the variation with time at each station reveals a rather surprising trend: TDS concentrations have generally decreased from the 1930s through the 2000s. Also surpris-

1042

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 33-3. (a) Average discharge, (b) average TDS load, (c) dischargeweighted average TDS concentration of the Rio Grande as a function of time interval and station. Stations are arranged from upstream forward to downstream rear. Data for 1905–1907 are from Stabler (1911), for 1930–1939 from Wilcox (1968), and 1940–2004 from USGS (2008). ingly, the variation in TDS concentration shows little correlation with variations in the river discharge. In contrast to the trend of increasing TDS concentrations with flow distance, Figure 33-3 shows that the TDS load is relatively invariant south of San Marcial (731 km). Other sources (Anning et al. 2006) demonstrate that

SALINITY ISSUES ON THE UPPER RIO GRANDE

1043

FIGURE 33-3. (Continued).

the load does not increase south of San Acacia (655 km). The flat trend of the TDS load is due to reduction in discharge with distance south of San Acacia (Fig. 33-3) counterbalancing the increase in concentration. Another marked difference from the TDS concentration trends is that TDS load does show a strong correlation with discharge, as previously observed by Yuan and Miyamoto (2004). The observation that, at an individual sampling location, load varies with discharge whereas concentration does not, suggests that the concentration may be controlled by the equilibrium solubility of moderately soluble minerals, such as calcite and gypsum. However, the trend of steadily increasing TDS concentration with distance contradicts this inference, inasmuch as the equilibrium solubilities should be constant with distance as well as time. In order to better understand how to control the increasing TDS concentration with flow, Phillips et al. (2003) and Hogan et al. (2007) examined the trend of chloride concentration and the chloride/bromide ratio with distance downstream (Fig. 33-4). The Cl/Br ratio (mass/mass) in natural waters is useful for fingerprinting the source of the Cl because the ratio is relatively low in natural precipitation (30 to 200), whereas it is usually very high (1,000 to 5,000) in subsurface brines (Davis et al. 1998; Moysey et al. 2003). Figure 33-4 illustrates that the Cl/Br ratio increases from ⬃50 in the headwaters area to 1,000 south of El Paso, as the Cl concentration increases from ⬃0.5 mg/L to 500 mg/L. This covariation indicates that influx of brines originating from deep subsurface sources

1044

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

FIGURE 33-4. Cl/Br ratio as a function of flow distance down the Rio Grande. Data from Mills (2004).

(e.g., connate or geothermal fluids) is contributing significantly to the increase of the Cl concentration. Simple evapotranspiration, as proposed by Lippincott (1939) and Trock et al. (1978), would result in an increase of Cl while the Cl/Br ratio remained low. Based on a quantitative analysis of the Cl budget, together with a more qualitative consideration of changes in the Cl/Br ratio, Lacey (2006) constructed the graph of Cl sources as a function of flow distance that is shown in Fig. 33-5. The combination of deep saline discharges and tributary inputs (which are also largely derived from saline discharges) accounts for about two-thirds of the Cl concentration increase of the Rio Grande below Otowi. Detailed examination of the river geochemistry along its course, such as was conducted by Wilcox (1968), shows that the loads of Ca2, Mg2, and SO42 remain approximately constant between San Acacia and El Paso, while HCO 3 decreases slightly. The concentrations of all these solutes except SO42 increase by 30% to 50% over the same reach (SO42 increases by ⬃60%). The large increase in TDS concentration shown in Figs. 33-1 and 33-3 (⬃75%) is thus accomplished by large increases in Na and Cl (200%) and a smaller one in SO42. These solutes are the common major constituents of sedimentary brines. This observation helps to explain the patterns observed in Figs. 33-3 and 33-4. The TDS concentration at individual stations is poorly correlated with discharge because above El Paso most constituents (Ca2, Mg2, some

SALINITY ISSUES ON THE UPPER RIO GRANDE

1045

FIGURE 33-5. Cumulative sources of Cl to the Rio Grande, as a function of flow distance. Data from Lacey (2006).

SO42, and HCO 3 ) are controlled by equilibrium with minerals in the soil under irrigated fields, probably principally calcite, dolomite, and gypsum. As discharge rises, more minerals are dissolved to achieve equilibrium, and as it falls, minerals are precipitated out into the soil. Relatively constant concentration over periods of varying discharge then produces good correlation of TDS load with discharge at individual stations. In contrast, Na, Cl, and some SO42 are principally derived from deep groundwater discharges. These solutes progressively accumulate with flow distance, producing the increases in both TDS concentration and load with downstream distance that are seen in Fig. 33-3, and also the increase of the Cl/Br ratio seen in Fig. 33-4. The increase in ionic strength from the addition of these solutes probably also increases the equilibrium solubility of the soil minerals, thus permitting the modest downstream increases observed for the remaining solutes. The systematics as described also help to explain the progressive decrease in TDS concentration over the twentieth century (Fig. 33-3). During this period, the Cl and Na concentrations also decreased markedly, whereas the Ca2 concentration decreased only slightly. This indicates that the TDS concentration decrease was largely due to progressive flushing of highly soluble salts that were stored in the system, rather than increased dissolution of soil minerals. It is possible that the higher concentrations of the most soluble salts were caused by increased deep

1046

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

groundwater discharge, but a more likely explanation is changes in agricultural practices. Due to an expanding population, the mid to late nineteenth and early twentieth centuries saw substantial expansion of the traditional agricultural system. These acequias (community-operated waterways for irrigation) generally did not provide wasteways for excess irrigation water, which was simply dumped on the floodplain below the last farm on the ditch (Wozniak 1987). During the same time period the bed of the Rio Grande began to aggrade (Clark 1987). This was probably partly due to channelization of the river between flood-control levies but probably also partly due to decreased peak discharges resulting from large water diversions for irrigation in the San Luis Valley, the portion of the Rio Grande drainage basin just downstream of the river headwaters in the San Juan Mountains in Colorado. In addition, during this period clear-cutting of timber and overgrazing by livestock in the headwaters of the Rio Grande greatly accelerated the rate of soil erosion, dumping large amounts of sediment into the river. The combined result of these modifications to the river system was that large areas of the Rio Grande Valley in the upper basin the floodplain, including the agricultural areas, became a hydrological sink from which water could escape only by evaporation (Phillips et al. 2011). A substantial proportion of the Rio Grande Valley in the upper basin was severely salinized (Wozniak 1987). The less-soluble salts were probably precipitated, but Na, Cl, and some SO42 were stored in the soils and groundwater of the floodplain. The recognition of the regional character of these problems helped to spur the creation of two regional irrigation districts, the Elephant Butte Irrigation District in 1918 and the Middle Rio Grande Conservancy District in 1928. One of the first orders of business for both districts was the construction of a regional drainage network. The improved drainage alleviated soil waterlogging and salinization and eventually substantially lowered water tables. It seems likely that the progressive decrease in TDS concentration of the Rio Grande over the period 1930–2000, accomplished mostly by decrease in Na, Cl, and SO42, was a result of the progressive flushing of soluble salts accumulated over centuries of traditional agriculture. Mitigation of River Salinity An improved understanding of the salinity dynamics of the Rio Grande may result in a better capability to control river salinity. First, it is clear that there is little to be gained from attempting to reduce the component of the TDS that is contributed by Ca2, Mg2, HCO3, and at least part of the SO42. These solutes are maintained at relatively constant levels by equilibrium with minerals in agricultural soils. The median annual TDS concen-

SALINITY ISSUES ON THE UPPER RIO GRANDE

1047

tration of ⬃350 mg/L at Bernardo (Anning et al. 2006), which is upstream of most of the points of deep groundwater discharge, thus probably represents an absolute minimum for possible dissolved solids in the downstream reaches. One strategy currently being pursued that might help to improve salinity levels is to reduce evapotranspiration, which will in turn reduce evaporative concentration of the more soluble salts. Eradication of introduced phreatophytic riparian vegetation, mostly saltcedar (Tamarix) and Russian olive, is being attempted over large areas. Whether this effort will actually result in significant water savings and salinity reduction remains to be seen. Saltcedar tends to accumulate salts in the leaves and this salt accumulates on the soil surface from the deposition of leaf litter. The mobilization of this salt after eradication may produce a pulse of salinity in the river. Furthermore, the evapotranspiration rate in riparian settings may be more a function of water-table depth than of vegetation type (Moayyad et al. 2003). Simply replacing invasive species by native ones may not result in significant water savings. The single largest evaporative sink on the Rio Grande is Elephant Butte Reservoir. The reservoir might be managed so as to make evaporation reduction a priority, but this type of flexibility in management is limited under the provisions of the Rio Grande Compact. Proposals have been suggested to switch from surface-water storage to groundwater storage, which would greatly reduce evaporation. However, issues of cost and practicality have not been thoroughly assessed. Furthermore, although evaporative concentration of solutes would certainly be reduced, reactions with aquifer minerals and mixing with ambient groundwater might also add significant solutes to the stored water being pumped out. Wastewater discharges are a large source of salts (Fig. 33-5), especially from the city of Albuquerque. A substantial proportion of those solutes arise from the groundwater that is pumped for the municipal water supply, which is higher in TDS concentration than the Rio Grande water that flows through the city. Albuquerque is currently in the process of shifting from a predominantly groundwater source for its supply to diversions from the Rio Grande as the predominant source. This will have two effects: increasing the TDS of the river by reduction of discharge (due to consumptive use between diversion and wastewater discharge) and reducing the TDS by reducing the amount of highTDS groundwater that is pumped and discharged. The net effect of these changes remains to be seen, but it may well be a small decrease in the TDS concentration. Another possible strategy is to attempt to reduce the inflows of saline deep groundwater. Certain portions of the regional agricultural drainage networks have been identified as locations where these brines enter the system, presumably due to favorable geologic structures beneath them

1048

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(Mills 2004). These portions might be backfilled or otherwise blocked or intercepted. More speculatively, it might be possible to drill wells to intercept subsurface saline sources and to then reinject the brines in deep formations. Similar efforts on the Colorado River have apparently reduced river salinity (Anning et al. 2006). The technology for this approach exists, but it is relatively expensive and should not be attempted until its costeffectiveness and feasibility under the specific hydrogeological characteristics has been carefully evaluated.

SUMMARY The salinity of the Rio Grande progressively increases by a factor of 50 over a flow distance of ⬃745 miles (1,200 km). All of the major solutes increase down the course of the river, but most of the salinization, in the lower part of the river at least, is due to increasing content of Na and Cl, and to a lesser extent SO42. At specific points on the river the TDS concentration correlates poorly with river discharge, but the TDS load is well correlated. Surprisingly, the TDS concentration of the river above El Paso has steadily decreased throughout the twentieth century. Contrary to the findings of earlier investigators, recent studies using geochemical techniques have concluded that irrigated agriculture plays only a secondary role in the salinization of the river. The primary agent is small discharges of high-concentration geological brines along deepseated faults and other favorable hydrogeological pathways. Relatively 2 reactive solutes (Ca2, Mg2, HCO 3 , and some SO 4 ) are largely controlled by equilibrium with soil minerals in the agricultural system. The more soluble constituents (Na and Cl, and part of the SO42), however, progressively accumulate from deep groundwater discharges. The decrease in the TDS concentration of the river over the twentieth century can probably be attributed to the slow flushing of soluble salts accumulated during centuries of traditional agriculture, and especially during late nineteenth and early parts of the twentieth centuries when a combination of circumstances produced widespread waterlogging and salinization. These salts have been slowly flushed following the construction of a modern drainage network between 1920 and 1940. Reduction of the river salinization may be possible through reduction of riparian evapotranspiration, reduction of pumping of relatively high-TDS groundwater, managing water storage so as to minimize evaporation, and blocking or intercepting brine discharges. However, all of these approaches have considerable practical uncertainties regarding their long-term effectiveness and should be carefully evaluated before implementation. The Rio Grande is not unique. Most of the large streams of the semiarid Southwest exhibit similar increases of TDS concentration with flow dis-

SALINITY ISSUES ON THE UPPER RIO GRANDE

1049

tance (Anning et al. 2006). Although the importance of different mechanisms of salinization vary from place to place, it seems likely that ones similar to the Rio Grande are common, especially discharge of deep saline groundwater. The findings of recent studies on Rio Grande salinization may thus help to guide investigations of salinization in semiarid rivers worldwide.

REFERENCES Anning, D. W., Bauch, N. J., Gerner, S. J., Flynn, M. E., Hamlin, S. N., Moore, S. J., Schaefer, D. H., Anderholm, S. K., and Spangler, L. E. (2006). “Dissolved solids in basin-fill aquifers and streams in the southwestern United States,” in Scientific Investigations Report 2006-5315, U.S. Geological Survey, Washington, D.C. Booker, J. F., Michelsen, A. M., and Ward, F. A. (2005). “Economic impact of alternative policy responses to prolonged and severe drought in the Rio Grande basin.” Water Resour. Res., 41(WO2626), 1–15. Clark, I. G. (1987). Water in New Mexico: A history of its management and use, University of New Mexico Press, Albuquerque. Davis, S. N., Whittemore, D. O., and Fabryka-Martin, J. (1998). “Uses of chloride/ bromide ratios in studies of potable water.” Ground Water, 36, 338–350. Goss, A. (1900). Principles of water analysis as applied to New Mexico waters, Bulletin No. 34, New Mexico College of Agriculture and Mechanic Arts, Agricultural Experiment Station, Mesilla Park, Santa Fe, N.M. Hanks, N., Morrison, W., and Michelsen, A. M. (2005). Water related organizations in the far west Texas region, Far West Texas Water Planning Group, El Paso, Tex. Hill, R. A. (1974). “Development of the Rio Grande Compact of 1938.” Nat. Resour. J., 14, 163–200. Hogan, J. F., Phillips, F. M., Mills, S. K., Hendrickx, J. M. H., Ruiz, J., Chesley, J. T., and Asmerom, Y. (2007). “Geological origins of salinization in a semiarid river: The role of sedimentary brines.” Geology, 35, 1063–1066. Lacey, H. (2006). “Quantification and characterization of chloride sources in the Rio Grande,” M.S. thesis, New Mexico Institute of Mining and Technology, Socorro, N.M. Lippincott, J. B. (1939). “Southwestern border water problems.” J. AWWA, 31, 1–29. Michelsen, A. M., and Cortez, F. (2007). Drought watch on the Rio Grande, Texas Agricultural Experiment Station, Texas A&M University System, El Paso Agricultural Research and Extension Center, Lubbock, Tex. Michelsen, A. M., McGuckin, T., Sheng, Z., Lacewell, R. L., and Creel, B. (2009). Rio Grande Salinity Management Program: Preliminary economic impact assessment, prepared for the Rio Grande Salinity Management Coalition with partial support from the U.S. Army Corps of Engineers, Albuquerque Office, Texas AgriLife Research Center, El Paso, Tex. Mills, S. K. (2004). “Quantifying salinization of the Rio Grande using environmental tracers,” M.S. thesis, New Mexico Institute of Mining and Technology, Socorro, N.M.

1050

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Miyamoto, S., Fenn, L. B., and Swietlik, D. (1995). Flow, salts, and trace elements in the Rio Grande: A review, TR-169, Texas Agricultural Experiment Station El Paso, Texas Water Resources Institute, El Paso, Tex. Miyamoto, S., Yuan, F., and Anand, S. (2005). Reconnaissance survey of salt sources and loading into the Pecos River, TR-291, Texas Agricultural Experiment Station El Paso, Texas Water Resources Institute, El Paso, Tex. ———. (2006). Influence of tributaries on salinity of Amistad International Reservoir, TR-292, Texas Agricultural Experiment Station El Paso, Texas Water Resources Institute, El Paso, Tex. Moayyad, B., Bawazir, S. A., King, J. P., Hong, S., and Hendrickx, J. M. H. (2003). “Groundwater depth and arid zone riparian evapotranspiration,” in Understanding water in a dry environment: Hydrological processes in arid and semi-arid zones, I. Simmers, ed., A. A. Balkema Publishers, Lisse, The Netherlands, 188–195. Moysey, S., Davis, S. N., Zreda, M., and Cecil, L. D. (2003). “The distribution of meteoric 36Cl/Cl in the United States: A comparison of models.” Hydrogeol. J., 11, 615–627. Natural Resources Committee. (1938). Regional Planning: Part VI, The Rio Grande joint investigation in the upper Rio Grande basin, Colorado, New Mexico, and Texas, 1936–1937, U.S. Government Publishing Office, Washington, D.C. Niemi, E., and McGuckin, T. (1997). Water management study: Upper Rio Grande basin, Report to the Western Water Policy Review Advisory Commission, National Technical Information Service, Springfield, Va. Phillips, F. M., Hall, E. H., and Black, M. E. (2011). Reining in the Rio Grande: People, land, and water, University of New Mexico Press, Albuquerque, N.M., 67–80. Phillips, F. M., Hogan, J. F., Mills, S., and Hendrickx, J. M. H. (2003). “Environmental tracers applied to quantifying causes of salinity in arid-region rivers: Preliminary results from the Rio Grande, southwestern USA,” in Water resource perspectives: Evaluation, management, and policy, A. S. Alarshan and W. W. Wood, eds., Elsevier, Amsterdam, 327–334. Scofield, C. S. (1938). Quality of water of the Rio Grande Basin above Fort Quitman, Texas: Analytical data, Water-Supply Paper WSP 839, U.S. Geological Survey, Washington, D.C. Stabler, H. (1911). Some stream waters of the western United States, with chapters on sediment carried by the Rio Grande and the industrial application of water analyses, Water-Supply Paper 274, U.S. Geological Survey, Washington, D.C. Trock, W. L., Huszar, P. C., Radosevich, G. E., Skogerboe, G. V., and Vlachos, E. C. (1978). Socio-economic and institutional factors in irrigation return flow quality control, EPA-600/2-78-174c, Robert S. Kerr Environmental Research Laboratory, U.S. Environmental Protection Agency, Ada, Okla. U.S. Department of State. (1906). Convention between the United States and Mexico equitable distribution of the waters of the Rio Grande, U.S. Department of State, Washington, D.C. U.S. Geological Survey (USGS). (2008). “USGS Water Data for the Nation.” National Water Information System, U.S. Geological Survey, Washington, D.C., http://waterdata.usgs.gov/nwis. Ward, F. A., Booker, J. F., and Michelsen, A. M. (2006). “Integrated economic, hydrologic and institutional analysis of policy responses to mitigate drought impacts in the Rio Grande basin.” J. Water Resour. Plan. Mgmt., 132(6), 488–502.

SALINITY ISSUES ON THE UPPER RIO GRANDE

1051

Wilcox, L. V. (1957). “Analysis of salt balance and salt-burden data on the Rio Grande,” in Problems of the upper Rio Grande: An arid zone river, P. C. Duisberg, ed., U.S. Commission for Arid Resource Improvement and Development, Socorro, N.M., 39–44. ———. (1968). Discharge and salt burden of the Rio Grande above Fort Quitman, Texas, and salt-balance conditions on the Rio Grande Project: Summary report for the 30-year period 1934–1963, Research Report No. 113, U.S. Salinity Laboratory, Soil and Water Conservation Research Division, Agricultural Research Service, U.S. Department of Agriculture, Riverside, Calif. Wozniak, F. E. (1987). Irrigation in the Rio Grande Valley, New Mexico: A study of the development of irrigation systems before 1945, Report on Contract BOR-87-1, U.S. Dept. of Agriculture and Bureau of Reclamation, Southwest Regional Office, Amarillo, Tex. Yuan, F., and Miyamoto, S. (2004). “Influence of the Pacific Decadal Oscillation on hydrochemistry of the Rio Grande, USA, and Mexico.” Geochem. Geophys. Geosys., 5(12), 1–10.

This page intentionally left blank

CHAPTER 34 VIABILITY OF IRRIGATED AGRICULTURE WITH EXPANDING SPACE AND TIME SCALES Wesley W. Wallender

INTRODUCTION Viability of an irrigated agroecosystem through time depends on farmer profitability in the context of competing municipal, industrial, and environmental demands for fresh water. Experiences on the west side of the San Joaquin Valley as well as in the Imperial and Coachella Valleys of California are used to guide the presentation, but the hope is that these ideas are applicable elsewhere and to no small degree to precipitationdriven agricultural landscapes. In the first section of this final chapter, the control volume (land area and depth of the system analyzed) provides a conceptual framework to assess water and salinity management for viability in the context of an evolving irrigation project. Increasing control-volume size from the farm to the district and finally to the region were chosen in accordance with management decisions made at those scales. Decisions are taken according to the costs and benefits of inputs and outputs that depend on area and depth (e.g., from the surface through the groundwater system) of the hydrology system. It is shown that because the cost of deep percolation historically is ignored early in a project, the control-volume depth changes with time and, thus, the decisions are not optimal. The analysis is segmented into stages during which the depth of the control volume may or may not be too shallow. An argument is made for extending the spatial and temporal extent of the analysis toward socially optimal management of water and salinity. The remainder of the chapter applies water and solute transport concepts in conjunction with the control-volume framework to demonstrate that the systems are not in equilibrium or steady state, but viability 1053

1054

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

(profitability while meeting reasonable-beneficial water use criteria) is predictable. The audience is primarily researchers, but managers hopefully will find the discussion useful, particularly the section on viability.

FARM CONTROL VOLUME The evolution through time (stages) of an irrigated agroecosystem or project is the context for what follows. The disposition of groundwater and salinity, depth of farm control volume, returns to water, and excess deep percolation and runoff losses are discussed for each stage of a project. Initially, fresh groundwater is extracted for irrigation and then the supply shifts to imported surface water as the fresh groundwater is exhausted and salinized. The groundwater system is recharged and saline groundwater encroaches on the rootzone with the corresponding salinity and drainage challenges. At the farm (field) scale, management decisions are made for a control volume created by the area of the farm and by the depth of the rootzone, but may also include the groundwater system depending on its quality, quantity, and depth through time. Water and salt management focuses on maximizing returns to water. “Returns to water” is the benefit (salinitydependent yield times price) minus the costs of water, which are the product of volume of irrigation, runoff, drainage, and deep percolation water and their quality-dependent unit costs, as well as the required labor and annualized irrigation equipment costs. Yield is related to total applied water and average soil salinity using a production function. Stage 1 At the outset of a project the groundwater for irrigation may be of good quality and shallow such that pumping costs are low, yet not too shallow to cause drainage problems. The control volume on which management decisions are made extends from the surface though the affected groundwater system, because the depth to groundwater affects pumping cost. Initially, the rootzone is reclaimed to reduce soil salinity, and typically the initial decline in environmental services associated with the mobilization of salty water to the deep vadose zone and to the groundwater is ignored despite being part of the control volume. Because the unit costs are low, yield is near maximum and water losses to excess deep percolation and runoff are not minimized in the process, maximizing returns to water. Hence, the traditional engineering design goal of minimizing losses is violated because deep percolation in excess of that required to control soil salinity and runoff losses are relatively high. The

VIABILITY OF IRRIGATED AGRICULTURE

1055

time scale for this stage may be less than a decade, depending on the rate of groundwater overdraft. Stage 2 If, through time, groundwater extraction exceeds recharge, the groundwater level on the local farm, as well as on those nearby, may fall. Not only will the pumping cost increase for the farm, but also the cost of lowering the water level in adjacent farms may rise through litigation. Farm costs associated with the impairment of local, district, and regional municipal, industrial, and environmental services, which are also dependent on the same fresh groundwater resource, may also increase. Decline in groundwater quality caused by salts in deep percolation from the rootzone is generally ignored. Discharge of surface runoff, via surface washoff, is commonly recirculated on the farm, but it may escape and impair downstream uses if the assimilative capacity of the receiving waters is exceeded. At this stage, however, the groundwater and surface water systems likely assimilate the salts without obvious impact on biological and physical systems down- or upstream of the control volume, and these costs are deferred until later stages. The water-quantity-related increases in farm, district, and regional costs affect farm decisions as they appear in the returns to water relationship, via the cost of irrigation water and the added cost for water-related litigation. As the unit costs are higher (e.g., the groundwater lift is higher), yield is likely further from maximum, while the losses decrease and are nearer to the traditional engineering design criteria of minimizing runoff and excess deep percolation. The time scale for this stage may be several years, depending on the rate of groundwater overdraft. Stage 3 If overpumping continues, the groundwater level falls farther, its quantity declines, land may subside, and the upper salty layer of the groundwater may mix with the fresh groundwater during pumping. Furthermore, the quantity and quality of groundwater and its depth may fall in control volumes beyond the farm, at the district and regional scales (control volumes). Costs associated with salty deep percolation, increased cost of pumpage and subsidence, and the costs associated with impaired services outside the farm control volume, which were previously ignored, are included in the farm cost. Reduction in yield benefit is caused by applying the saltier irrigation water. Returns to water decline and a shift to lowervalued crops may occur. In addition, more irrigation water may be applied to maintain favorable rootzone salinity. This causes an increase in deep percolation, but the losses to deep percolation in excess of leaching and to

1056

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

runoff decrease to near the design goal of minimizing their sum. Eventually, however, irrigation may stop (land retirement) because groundwater quantity and quality may become limiting and because salinity in the rootzone may elevate beyond the limits of economic crop productivity. However, the previously ignored costs associated with the impairment of soil, groundwater, and land subsidence, and of surface water services remain for capture in later stages. The time scale may be years to decades. Stage 4 To avoid land retirement, arrest land subsidence, and escape litigation costs associated with extracting groundwater from control volumes beyond the farm, inexpensive surface water may be imported for irrigation. By replacing groundwater with surface water, the farm control volume may shrink to that of the rootzone if the cost of leaving behind the depleted salty groundwater system and the related effects are ignored. This control volume includes only the rootzone, which is in contrast to Stage 1 in which the groundwater system was part of the control volume because pumping cost was affected by lift and quantity. If the imported surface water unit cost is less than pumped groundwater, the optimal decision at the farm scale may be to apply more water than is necessary to maintain favorable soil water and salinity. Overirrigation is greater than in Stage 1 where the irrigation design is far short of minimizing runoff and excess deep percolation losses. Recall that the impairment of environmental services caused by the flow of salty water that passes the bottom of the rootzone or is discharged in surface runoff, via surface washoff, is ignored and the control volume is the rootzone only. The time scale for this stage may be decades, depending on the groundwater system and the rate of overirrigation. Stage 5 Through time, however, in the case of imported irrigation water mentioned, the groundwater levels may rise (more rapidly than if overirrigation did not occur) and encroach on the rootzone because the assimilation capacity (lateral or vertical discharge from the regional groundwater control volume) of the groundwater is exceeded. The groundwater receives and releases water and salt from lateral flows from the rootzone via deep percolation during irrigation events; it sends water and salt back to the rootzone between irrigations (upward flow driven by evapotranspiration, ET); and it discharges water and salt to the drains. The groundwater is a mixing zone of lateral flows and deep percolation from the rootzone. This mixture leaves the groundwater as lateral flow, as upward flow into the rootzone between irrigations, via the drains. The

VIABILITY OF IRRIGATED AGRICULTURE

1057

groundwater at the water table is more saline than irrigation water, because it is a blend of deep percolation and deeper groundwater. Compared to higher-quality irrigation water, water flowing upward is of lower quality and hence of diminished value. The drain-water blend has more saline deep groundwater (salts from previous stages are stored in the groundwater) and is thus saltier than water moving upward to meet ET. Being more saline, it is costly to reuse or dispose. Obviously, dissolution and precipitation of salts affect their concentration and composition. These impaired waters affect farm water and salinity management decisions, and thus the control volume includes the rootzone and the groundwater. With the groundwater system discharging previously ignored salt to the rootzone and to the drains, the related reductions in benefits and increases in costs, respectively, are now borne at the farm level. Returns to water decline because the rootzone salinity may be elevated. In some cases deep percolation is less than upward movement to meet ET and the rootzone salinizes at times during the season. Returns also decline because drainage costs are included. If the drain water is continually recycled on the same land, the rootzone salinizes, and the land may have to be retired.

EXTENDING THE SPATIAL AND TEMPORAL EXTENT Imagine a groundwater system that connects the upslope farm to the downslope farm as previously described. Upstream, the groundwater is deep, the cost of deep percolation is ignored, and the tendency is to overirrigate because that cost is ignored. This overirrigation recharges the groundwater system and causes the groundwater level to rise more than it would otherwise in the downslope farm control volume. Added cost associated with the rise disadvantages the downstream farm. In contrast, if the same owner operated the upstream and downstream farms, the returns to water relations for both farms would be combined, and it is likely that the upstream farm would be managed to reduce deep percolation if that cost is less than the added cost of drainage at the downstream farm. This shows that length scale affects optimal decision making. In a similar way, time scale affects optimal decision making. To illustrate this, consider the case in which the control volume extends from the surface through the entire groundwater systems over all five stages. The deferred groundwater depletion and salinization costs during the early stages would be included in the cost terms. These costs might motivate the farm water user to apply less water in both the early and the last stages of the project. The temporal and spatial extent of the analysis affects decision making. From a temporal point of view, if ignored costs were included in the early stages of the returns to water relation, the control volume would extend

1058

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

from the soil surface through the groundwater system and less water may be applied. Similarly, from a spatial extent point of view, if upstream and downstream farms were managed by the same farmer, less water may be applied upstream. By ignoring these costs, upslope farms may have advantages over downslope farms as the project matures. Land value in downslope highwater-table areas falls, land may be retired, and the irrigation water may be shifted to upslope farms. The time scale for this stage may be years, depending on the groundwater system and the rate of overirrigation. This progress may at least be slowed if the time and space scales were extended to the entire district and through the entire duration of the project.

IRRIGATION DISTRICT CONTROL VOLUME Once again, the evolution of the project is the context for analyzing the water and salinity management at the district scale. The distribution of the water and salinity, the depth of the control volume, and cost are discussed for each stage. At this scale, decisions are made for a control volume that includes the area of the district and from the soil surface through the rootzone to some depth in the groundwater system. The expanded control volume also includes surface features, such as the water delivery and drainage conveyance structures, as well as natural channels. The distribution of water and salt within, as well as the flows of water and salt across the external surfaces of the control volume may affect district management decisions. A goal is to minimize cost to farms and to manage surface and groundwater flows into and out of the boundaries of the control volume, not to maximize profit or returns to water as in the case of farm control volumes. A related goal, shared by the district and the farms, is to use water reasonably and beneficially. Reasonable-beneficial use may be defined as the use of water in such a quantity as is necessary for economic and efficient utilization for a purpose and in a manner that are both reasonable and consistent with the state and county land use plans and the public interest. The latter criteria often evolve through court and regulatory actions and do not rely solely on economic measures; rather, they default to efficiency as well as to return flow quantity and quality measures. These measures and targets are documented in water management plans that may or may not include salinity metrics. In the recent past there has been interest in privatizing water markets in a way similar to that done in the energy sector. The objective would be to maximize profit while satisfying the reasonable-beneficial use criteria.

VIABILITY OF IRRIGATED AGRICULTURE

1059

For the case herein, the district cost is balanced by the water charges to farms for these nonprofit entities. The connection between the farm and the district is through the cost to farms. They work together to meet the reasonable-beneficial use criteria found in district water management plans. Once again, the stages are used here to discuss the project evolution through time. The time scale for these stages is the same as for the farm scale. Stage 1 Parallel to the development stated for the farm control volume, assume that at the outset of the project the groundwater is of good quality, pumping costs are low, and it is used for irrigation. As mentioned, the groundwater system is generally managed by individual farmers, but if the groundwater is managed by the district it may maintain a well network along with a water delivery system that distributes the groundwater to farms. In addition, the district manages surface flows generated by irrigation and precipitation, if necessary. The control volume extends from the surface through the groundwater system because the district assumes responsibility for allocating the common groundwater and surface water resources within the district, as well as for managing the groundwater and surface water flow across the control volume boundaries shared by contiguous entities that depend on the same connected surface and water resources. From a water-quality point of view, costs associated with salinizing the fresh groundwater with salt leaving the rootzone and with salty water leaving the control volume and impairing downstream environmental services are ignored and not passed on to the farms. The time scale for district decisions may vary from the time for an irrigation water delivery to years, depending on the quantity of groundwater and surface water. The cost to farms increases with time as the groundwater storage is depleted and lift increases. Stage 2 Groundwater overdraft continues and the control volume remains the same as in Stage 1, from the surface through the groundwater system. Surface water leaving the control volume may not impair downstream uses because the assimilative capacity of the surface waters is not exceeded. Once again, the salt added to the groundwater system is also ignored but, because groundwater is overdrafted, access to ground water outside the district may be impaired and a new cost may emerge from litigation. This cost, as well as the increased cost of pumping as the lift increases, is passed on to the farms.

1060

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Stage 3 In this last stage of groundwater overdraft, the salty deep percolation water mixes with the fresh groundwater and the land may subside further. The decline in water quality may require blending of water from different wells. Added costs associated with the water-blending infrastructure are passed on to farms. Land subsidence may add further costs due to damage of farm infrastructure, increased pumping costs caused by reductions in the specific capacity of the aquifer, and related litigation. As the amount and quality of groundwater available declines and land is retired, the unit cost of water increases. Recall that the soil, groundwater, and surface water degradation costs are ignored. Stage 4 To avoid land retirement caused by overpumping and salination of the groundwater, surface water is imported for irrigation. In the short term, the depth of the district control volume shrinks to the bottom of the rootzone. The thick, unsaturated zone above the water table and the salinized groundwater system are ignored from a decision-making point of view. The district manages the irrigation- and precipitation-driven surface water flows for minimum cost, which might be lower than when groundwater was pumped. This lower cost is reflected in lower water cost to farms. As mentioned, the reduction in unit water cost may lead to excess deep percolation and runoff at the farm scale. Additional costs may be incurred if the runoff is collected and redistributed by the district. Stage 5 In the final temporal phase, the groundwater levels may encroach on the rootzone in regions within the district, and it may become necessary to install farm drainage systems (cost to the farmers) to control water and salt levels in the rootzone. If the district collects the drain water, it can be discharged outside the district control volume or be reused within the control volume. Within-district reuse might include blending drain water with imported water or sequentially using the drain water and eventually discharging to evaporation basins. Without salt discharge from the district, the groundwater system may salinize further, and the downslope land is eventually retired. Land retirement will likely continue until the net groundwater flows out of the district equal or exceed deep percolation. In the case of out-of-district drain water and salt discharge, this process may be slowed, but the threat of land retirement remains. For both within- and outof-district drain water and salt disposal there is an associated cost that translates into an increased cost passed on to the farms.

VIABILITY OF IRRIGATED AGRICULTURE

1061

As at the farm scale, the district control volume changes with depth through time as the project evolves. Initially, deep percolation and runoff costs are ignored and then later captured and hence influence management at the district scale. Increased district cost is transferred to the farm and, generally, deep percolation is reduced, particularly in downslope areas. As mentioned, downslope regions may be retired and the irrigation water shifted to upslope areas with deeper water tables as the project matures. If the majority of the district has a shallow water table, all land in the district may be retired and the water transferred to another district in a deep-water-table region. To meet its goal of reasonable-beneficial use at minimum cost, the district can use tiered water pricing to decrease excess water application, particularly in upslope regions. The lowest-cost tier might be set according to water duty to remain revenue neutral. The duty of water includes water required to mature a particular type of crop, including consumptive use, required leaching, dust control, and frost control, as well as evaporation and seepage from ditches and canals, and the water eventually returned by percolation, via groundwater and by surface runoff. The amount of water justifiably returned may be limited by quality criteria. Unit cost of water is higher for water applied in excess of the water duty. Using district tiered water pricing, the reasonable-beneficial use criteria at the farm scale are more likely to be met because farmers will likely apply less water than if a low flat rate is charged. Applying less water might dissolve and mobilize less salt in the return flows.

Extending the Spatial and Temporal Extent As at the farm scale, the temporal and spatial extent of the district can be extended. All five stages are included, and the spatial extent is increased to adjacent districts. Losses would decrease because ignored costs in time and in space would be internal to the districts. Furthermore, less salt may be dissolved and mobilized.

REGIONAL CONTROL VOLUME MANAGEMENT The areal extent of the control volume increases to that of the enclosing political boundaries, which may or may not coincide with watershed boundaries and may include municipal and industrial uses in addition to environmental and agricultural uses. The depth of the control volume extends from the surface and may include part or all of the entire groundwater system. Once again, the depth of the control volume is determined by inputs and outputs of the system.

1062

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The time scale varies from days to decades. For example, water quality in surface receiving waters might be regulated daily, whereas the groundwater quality might be regulated on a decade basis. The goal is to maximize social benefits, including environmental services. Agricultural water use is linked to other uses through reasonable-beneficial use criteria. If agriculture does not meet that criteria, water quality is often impaired for other uses. Water markets may guide allocation decisions between and within competing uses of water of various qualities. They remain, however, within the context of institutional policy and legal considerations. Stage 1 Following the project scenario, start with fresh groundwater pumping only. The control volume may include a wide range of water-related municipal, industrial, agricultural, and environmental structures (including rivers) on the surface, as well as the subsurface from soil surface through the unsaturated zone and the affected groundwater system. Depending on the hydrologic connection between the groundwater and river systems within the control volume, there may be little impairment of environmental services to surface waters in terms of quantity or water quality. Similarly, there is little impairment of groundwater services. With abundant fresh groundwater, little policy intervention is invoked. Stage 2 With increased water use and groundwater overdraft through time, the groundwater quantity declines. This may lead to litigation between users within the region competing for the limited groundwater. Water used for irrigation may decline, while for municipal and industry uses it may increase because those users can afford to pay more for water. Environmental costs associated with the impaired groundwater and surface water systems may be ignored. Stage 3 With the groundwater system exhausted, land is retired and the supporting municipalities and industries decline if they depend on agriculture. Maximum social benefit falls and the soil, groundwater, and surfacewater degradation costs are ignored. Stage 4 To avoid the decline in groundwater-based irrigation and industrial, municipal, and environmental services, fresh surface water is imported.

VIABILITY OF IRRIGATED AGRICULTURE

1063

The control volume shrinks to the near surface, and allocation and regulation decisions are made with little, if any, regard to the impairment of groundwater resources, depending on the connection to impaired surface waters. Maximum social benefit rebounds to possibly higher levels than in Stage 1 if the subsidized surface water cost is less than for pumped groundwater. Stage 5 With imported surface water, the groundwater levels recover and once again may encroach on the rootzone in the irrigation regions within the control volume. Drainage water reuse and disposal may impair regional water quality and quantity, and hence potentially all services within as well as outside the control volume. Drainage-water penalties, levied by the government, associated with discharge outside the control volume are meant to represent society’s lost downstream opportunities to reasonably and beneficially use the water. These added costs translate into an increase in cost within the region. These costs obviously affect management decisions made by industrial, municipal, and agricultural users depending on how the costs are distributed across users. Unless cropping patterns shift to higher-value crops, which they often do, maximum social benefit falls as land is retired. Extending the Spatial and Temporal Extent The temporal and spatial extent of the region to the watershed can be extended. All five stages are included, and the spatial extent is increased to adjacent regions within the watershed. Internal losses within agriculture, which violate reasonable-beneficial use criteria, would decrease because ignored costs in time and in space would be internalize. Furthermore, less salt may be dissolved and mobilized within the watershed. Water markets are a mechanism to achieve the goal of optimal social use of water.

FRAMEWORK FOR MANAGEMENT MODELS Because the hydrology and economy of these systems are complex and interconnected, they have been difficult to model and measure, and therefore forecasting outcomes for different policies across time and space scales has been a challenge. The disposition of water and salinity in the groundwater systems is especially difficult to measure and model at high temporal and spatial resolution and over large space and time scales. As a consequence, management decisions have largely been reactive rather

1064

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

than proactive. The system is inherently changing in space (nonuniform) and time (unsteady). Our modeling and measuring tools continue to improve and the hope is to close the gap such that management can shift to the proactive mode underpinned by improved accuracy of predicting outcomes. Toward closing this gap, the next section lays out a quantitative framework for model development and measurement to interpret the viability of irrigation- and precipitation-driven agroecosystems. Modeling to Interpret Viability of Agroecosystems Across the range of length scales, chemicals react, dissolve, precipitate, and diffuse while being convected in the flowing water. The time scale of these complex processes influences when steady state will be achieved. Generally, the water flow (convection), water content, and chemical concentration are nonuniform in space and unsteady in time within and at the boundaries of a control volume. If at some time conditions at the control volume boundary are assumed steady, the time to steady state within a control volume is achieved when the flow and chemical concentration within the control volume do not change. Even if the water velocity field becomes steady, the rates of diffusion, reaction, dissolution, and precipitation may delay the time to steady state of the chemical concentration field. When both the water velocity field and the concentration fields are steady, travel time is the travel distance divided by the length. Travel times for a convected chemicals are generally less in the partially saturated rootzone than in the saturated groundwater system, because the length scale is less even though hydraulic conductivity is lower in the rootzone. In addition, heterogeneity in the subsurface porous medium commonly increases travel time as low-permeability regions limit speed through the control volume. For surface flows, the travel times are generally much less than for subsurface porous media flows because the flow velocity is far greater. Irrigation control volumes are rarely at steady state in terms of flow or chemical concentration because the conditions at the boundaries are inherently unsteady. Furthermore, the boundary position of the control volume, being defined by the user according to the analysis of interest, changes with time, as previously mentioned. In quantitative terms, the unsteady mass balance equation (Whitaker 1982) for the water, including chemicals, is d dV  ∫ (v  w ) ◊ndA dt Va∫(t ) Aa ( t )

(34-1)

in which the first term is the rate of change of mass in the temporally deforming control volume Va(t), and the second term is the net flux of the

VIABILITY OF IRRIGATED AGRICULTURE

1065

fluid entering the temporally deforming surface Aa(t) of the control volume. Fluid density is , the fluid velocity is v, the boundary velocity is w, and n is the unit vector normal to the surface of the control volume. The more familiar unsteady form (Bird et al. 2002) is dmtot  Δw dt

(34-2)

in which mtot is the total mass in the control volume, and w  w2  w1, which is the exit value minus the entrance value where w is the mass rate of flow. If the total mass within the control volume does not change with time and flow is steady, 0  w

(34-3)

In the case of chemical species in the fluid, the law of conservation of mass for unsteady conditions is d   dV  ∫  (v  w ) ◊ndA  ∫ r dV  ∫ r s dA dt Va∫(t ) Aa ( t ) Va ( t ) Aa ( t )

(34-4)

in which the first term is the rate of change of mass of species , the second term is the net flux of the species entering the surface Aa(t) of the deforming control volume, the third term is the rate of production of the species in the solution (homogeneous reactions), and the last term is the rate of production of the species at solid surfaces, such as those at the fluid and soil particle interface. The mass of the species per unit volume of water only is  , the species velocity is v , the boundary velocity is w, and n is the unit vector normal to the surface of the control volume. If convection dominates over diffusive transport, the fluid velocity can be substituted for species velocity. The more familiar unsteady form (Bird et al. 2002) is dm ,tot  Δw  r ,tot dt

(34-5)

in which m ,tot is the total mass of the species in the control volume, w

is the net mass rate of addition of the species carried in the flowing water as well as at the fluid–soil particle interface, and the last term is the net rate of production of species within the fluid volume (homogeneous reactions) and production at the soil particle surfaces (heterogeneous reactions). Bird and his colleagues separate the second term into net mass flow of the species at entrances and exits, and net mass flow at “boundary surfaces,” which, for this case, might be the fluid–soil particle interface.

1066

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

The latter should not be confused with heterogeneous reactions at the fluid–soil particle interface. If the total mass of species within the control volume does not change with time, dm ,tot  0  Δw  r ,tot dt

(34-6)

The net mass flow across fluid and solid control-volume boundaries is equal to the net of the homogeneous and heterogeneous reactions in the fluid and on the solid surfaces, respectively. If net reactions are zero, mass inflow equals the mass outflow at steady state and 0  w . Occasionally the terms “steady state” and “equilibrium” are mistakenly used interchangeably. Steady state occurs when the velocity, temperature, and chemical concentration do not change with time but, in general, can be nonuniform in the control volume. Equilibrium occurs when the fluid is at rest, and the temperature, solution chemical concentration, and the electric field are uniform. Fluid momentum transfer, heat transfer, and chemical charge transport are nonequilibrium processes in irrigationcontrol volumes because they are driven by spatial variation (gradients) in velocity, temperature, concentration, and charge. Respectively, gradients of these fields are used in conjunction with transport coefficients in the constitutive equations by Newton to calculate fluid shear stress, by Fourier for heat transfer, by Fick for diffusion, and by Ohm’s Law for current. These transport processes assume that the characteristic time to reach local equilibrium is very small (molecule-to-molecule interactions) compared to the characteristic time at large spatial scales for the dynamic system. If local equilibrium is achieved, then the equations of state are applicable at the small scale. Although thermodynamics, for example, can be used to calculate heat transfer resulting from nonequilibrium conditions, it does not quantify the rate of heat transfer given by Fourier’s constitutive equation. Furthermore, local equilibrium may not be achieved for finite-rate chemical reactions and phase changes, and the equations of state do not apply to the dynamic system. In most cases the processes are dissipative, that is, they are irreversible and global entropy increases. The empirical constitutive equations remain useful for dissipative conditions. As mentioned, the velocities, temperatures, or concentrations are unsteady and nonuniform from the small scale up to the length scale of the control volume of the rootzone and beyond. Having made the case that the control volume is not in equilibrium or in steady state, it is unlikely that control volume is in salt balance. Even if the control is in salt balance, it simply means that w  r ,tot

(34-7)

VIABILITY OF IRRIGATED AGRICULTURE

1067

Though net mass flow across fluid and solid control-volume boundaries is equal to the net of the homogeneous and heterogeneous reactions in the fluid and on the solid surfaces, respectively, the control volume may or may not be profitable for agricultural production (viable) or for environmental services. If, in addition, net reactions within the control volume balance to zero and hence mass inflow equals the mass outflow over the control-volume surfaces at steady state 0  w , movement of salt within the control volume may also preclude agricultural viability if salt accumulates in the rootzone.

AGRICULTURAL VIABILITY From the introductory section it is clear that, with the possible exception of the plant rootzone, salt balance of the entire system historically has not been a management objective. Rather, agricultural profitability— while attempting to meet reasonable-beneficial use criteria—is the goal for agriculture viability. It was shown that meeting the criteria of reasonable-beneficial use may extend the life of the project if the cost of salty deep percolation is included in the decision-making process. In so doing, the control volume extends through the groundwater system and over the region. Furthermore, the temporal extent used for decision making extends through the life of the project. The evolution and life of the project viability is predictable but uncertain. State-of-the-art unsteady, spatially distributed hydrology, chemistry, and economic models can be linked at appropriate spatial and temporal resolution, as well as spatial and temporal extents toward accurately predicting socially optimal water and salinity management. For a given initial condition of the system, the economic model predicts water allocation. Next, the hydrology and chemistry models calculate the outcome, and the process repeats through time. Thus, there is a predictable time series of water, salinity, costs, and benefits, spatially distributed over the region. However, uncertainty in policy, markets, surface-water imports, chemistry of the applied water, groundwater, soil, geology, and climate, among other factors, influence the evolution and life of the unsteady, nonuniform agroecosystem. With prediction and its accompanying uncertainty, the role of the scientist and engineer advances from reactive to proactive participation in decision making. For example, one might simulate a land and water use scenario in which imported salt as well as salt dissolution in the system matches the socially acceptable and biophysically limited ability of the surface and subsurface receiving waters to accept those salt flows. That landscape might support an irrigated area of lesser extent than was historically viable.

1068

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

SUMMARY Decision making at various spatial and temporal scales has cascading effects on irrigation and salinity management. Decisions at the farm level affect the farm and adjacent growers. The combined decisions of many growers over time result in accumulated impacts, which may have effects at the water district scale. Water district policies, in turn, affect conditions at a regional scale. At each scale, physical conditions and responses to them change over time as impacts accumulate, and there are both (1) a lag time between management actions and degradation of soilwater conditions, and (2) an inconsistent spatial and temporal distribution of the accumulated effects. The sequence of the processes that lead to irrigation and drainage problems is similar at the various spatial and temporal scales, and thus, there is a potential to develop linked models. The general trends of management at each spatial and temporal scale can be modeled and the models can be linked, although there is uncertainty in policy, markets, surface water imports, chemistry of the applied water, groundwater, soil, geology, and climate, among other factors. In addition, irrigation management measures may not be uniformly distributed over space and time. Because the hydrology and economy of these systems are complex and interconnected, they have been difficult to model and measure, and forecasting outcomes for different policies across time and space scales has been a challenge. But the sequence of effects can be reasonably characterized at the farm, multifarm, water district, and regional scales, taking into account variables such as water velocity and concentration; how chemicals react, dissolve, precipitate, and diffuse while being convected in the flowing water; and the time scale of these complex processes when steady state is achieved. In short, we know a good deal about the sequence and general effects of irrigation in areas with saline soils; the challenge is developing models that predict the effects of various management regimes at the on-farm, regional, and water district levels. How, for example, does tiered pricing affect the distribution of irrigation water or the method of irrigation? Working from the perspective of a defined control volume, which will be different at each spatial and temporal scale, state-of-the-art unsteady, spatially distributed hydrology, chemistry, and economic models can be linked at appropriate spatial and temporal resolution as well as spatial and temporal extents toward accurately predicting socially optimal water and salinity management.

VIABILITY OF IRRIGATED AGRICULTURE

1069

REFERENCES Bird, R. B., Stewart, W. E., and Lightfoot, E. N. (2002). Transport phenomena, 2nd ed., John Wiley & Sons, Inc., New York. Whitaker, S. (1982). Introduction to fluid mechanics, R. E. Kreiger Publishing Co., Malabar, Fla.

This page intentionally left blank

APPENDIX

Table A-1. Conversion Factors for SI and Non-SI Units To convert Column 1 into Column 2, multiply by

Column 1 SI Unit

Column 2 Non-SI unit

To convert Column 2 into Column 1, multiply by

Length 0.621 3.28 0.00394 0.394

kilometer, km meter, m millimeter, mm centimeter, cm

mile, mi foot, ft inch, in. inch, in.

1.609 0.305 25.4 2.54

acre, ac acre, ac acre, ac sq. foot, ft2

0.405 0.00405 4,047 0.0929

acre-inch, ac-in. acre-foot, ac-ft million acre-feet, maf gallon, gal

102.8 1234 1.234 3.785

pound, lb ton (US), ton pound, lb ton (US), ton

454 907 0.454 0.907

pounds per acre, lb/ac pounds per acre, lb/ac US ton per acre, ton/ac

1.12 0.00112

Area 2.47 247 0.000247 10.76

hectare, ha sq.km, km2 sq. meter, m2 sq. meter, m2

0.00973 0.00081 0.81 0.265

cubic meter, m3 cubic meter, m3 cubic kilometers, km3 liter, L

0.0022 0.0011 2.205 1.102

gram, g kilogram, kg kilogram, kg tonne, t

Volume

Mass

Yield 0.893 893

kg/ha tonne/ha tonne/ha

(continued)

1071

1072

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Table A-1. Conversion Factors for SI and Non-SI Units (Continued) To convert Column 1 into Column 2, multiply by

Column 1 SI Unit

Column 2 Non-SI unit

To convert Column 2 into Column 1, multiply by

Rate 264106 0.107 2.24

m3/day L/ha m/sec

million gallons per day, mgd gallons per acre, gal/ac miles per hour, mi/hr

3,785 9.35 0.447

atmosphere, atm bar pound per square foot, lb/ft2 pound per square inch, psi atmosphere, atm

0.101 0.1 47.9 6,900 101.3

millimho/cm, mmho/cm millimho/cm, mmho/cm micromho/cm, umho/cm micromho/cm, umho/cm

0.1 1 1,000 1

ac-in. cfs cfs gal/min ac-ft ac-ft ac-in. ac-ft ac-ft/ac ac-ft/ac gpm

102.8 101.9 0.028 0.227 0.1233 1234 0.0103 12.33 3,047 0.305 3.788

meq/100g % ppm ounce/gal lb/gal

1 10 1 7.489 119.8

Pressure 9.9 10 0.0209 0.000145 0.00987

megapascal, MPa megapascal, MPa pascal, Pa pascal, Pa kilopascal, kPa

10 1 0.001 1

Siemen/m, S/m decisiemen/m, dS/m decisiemen/m, dS/m millisiemen/cm,mS/cm

Electrical Conductivity

Water Measurement 0.00973 0.00981 35.59 4.4 8.11 0.00081 97.28 0.0821 0.000328 3.279 0.264

m3 m3/hr m3/sec m3/hr ha-m m3 ha-m ha-cm m3/ha m3/m2 L/min Concentration

1 0.1 1 0.1335 0.00835

centimole/kg g/kg mg/kg g/L g/L

APPENDIX

1073

Table A-2. Chemical Conversion Units To convert Column 1 into Column 2, multiply by

Chemical Conversions for Ions

To convert Column 2 into Column 1, multiply by

Column 1 milligram/Liter

Column 2 milliequivalent/Liter

0.0499

mg/L Ca

meq/L Ca

20.04

0.0823

mg/L Mg

meq/L Mg

12.15

0.0435

mg/L Na

meq/L Na

22.99

0.0256

mg/L K

meq/L K

39.1

0.0164

mg/L HCO3

meq/L HCO3

61.02

0.033

mg/L CO3

meq/L CO3

30

0.0282

mg/L Cl

meq/L Cl

35.45

0.0208

mg/L SO4

meq/L SO4

48.03

0.0161

mg/L NO3

meq/L NO3

62

0.0554

mg/L NH4

meq/L NH4

18.04

To convert Column 1 into Column 2, multiply by

Chemical Conversions for Ions Column 1 milligram/Liter

Column 2 millimole/Liter

To convert Column 2 into Column 1, multiply by

0.025

mg/L Ca

mM/L Ca

40.08

0.0411

mg/L Mg

mM/L Mg

24.31

0.0435

mg/L Na

mM/L Na

22.99

0.0256

mg/L K

mM/L K

39.1

0.0164

mg/L HCO3

mM/L HCO3

61.02

0.0167

mg/L CO3

mM/L CO3

60

0.0282

mg/L Cl

mM/L Cl

35.45

0.0104

mg/L SO4

mM/L SO4

97.06

0.0161

mg/L NO3

mM/L NO3

62.01

0.0554

mg/L NH4

mM/L NH4

18.04

1074

AGRICULTURAL SALINITY ASSESSMENT AND MANAGEMENT

Table A-3. Other Useful Conversions mg/L TDS  EC dS/m  640 mg/L TDS  EC dS/m  735 (preferred for Colorado River water) lbs/ac-ft TDS  mg/L TDS  2.72 tons/ac-ft TDS  mg/L TDS  0.00136 atm osmotic pressure  EC dS/m  0.36 1 ac  43,560 sq ft 1 mi  5,280 ft 1 ac-ft soil  4 million lbs (approx.) 1 ton/ac  20.8 g/sq ft 1 g/sq ft  96 lb/ac 1 lb/ac  0.0104 g/sq ft 1 cu ft  7.48 gals 1 gal  8.345 lb cfs  448.8 gpm 1 cfs/24 hr  1.98 ac-ft 1 mgd  3.07 ac-ft/24 hr 1 mgd  1.547 cu ft/sec 1 mgd  694.4 gpm 1 ac-ft  325,851 gal 1 atm  14.7 psi 1 psi  14.22 kg/sq cm 1 bar  14.5 psi 1 bar  1,023 cm water

APPENDIX

1075

Table A-4. Soil Water

Soil texture

Field capacity or water holding capacity (inches water per ft soil)

Available soil moisture (inches water per ft soil)

Sand Loamy sand Sandy loam Loam Silt loam Sandy clay loam Sandy clay Clay loam Silty clay loam Silty clay

1.2 1.9 2.5 3.2 3.6 3.5 3.4 3.8 4.3 4.8

0.7 1.1 1.4 1.8 1.8 1.3 1.6 1.7 1.9 2.4

Clay

4.8

2.2

This page intentionally left blank

INDEX

The letter t following a page number denotes a table; the letter f following a page number denotes a figure. amorphous iron oxyhydroxide, 108 anaerobic water treatment, 729 analytical solute transport model, 900 animal manure, 113, 114 anion exclusion, 74–75, 76t, 77 anion ratio, 6 anions, 73, 93, 95, 170, 297 anoxic microsites, 97 anthropogenic salt sources, 3, 61–62 antioxidant protection, 254–256 apparent soil electrical conductivity (ECa) measurement, 304–315, 329; edaphic factors influencing, 312–315; electrical resistivity, 304–308; electromagnetic induction, 308–310; geospatial measurements, 317–319; time domain reflectometry, 311–312 aquaculture, 746 Arabidopsis, 188, 238 Arabidopsis plants, 239, 255, 260 Arabidopsis thaliana, 194, 248 Argentine mesquite tree, 746 aridisols, 60 arid landscapes, 60, 69; chemical composition of surface/ground waters, 70, 71t, 72; evaporation basins and, 757 Arkansas Valley, 601, 687 arsenic, 107–108, 109f; in San Joaquin Valley evaporation basins, 775–776 Arthrobacter globiformis, 251 Arthrobacter panescens, 251 asparagus, 421 atmospheric deposition, 61, 115 Atriplex spp., 708

A Abiotic Disorders of Landscape Plants, 423 acidic fallout, 61 acidulants, 670, 679 activity coefficient calculation, 65–67, 67f Adams Avenue Agricultural Drainage Research Center, 729 adsorption: data description, 95–97; mercury, 118; trace elements and, 91–92 aggregate stability, influences on, 156, 157f, 158, 171 agricultural evaporation basins. See evaporation basins agricultural salinity problems: California’s San Joaquin Valley, 14–22, 22f; drainage dilemma, 23; effects of salts on plants, 29–31; effects of salts on soils, 27–29; global scale, 3–4, 4f, 5f; historical perspective, 12–14, 23; modeling methods for identification of, 52–53; monitoring methods, 53. See also salinity control and management measures; soil salinity; specific topic Albuquerque, New Mexico, 1035, 1047 alfalfa, 39, 176, 388, 416–417, 573–574, 577, 810, 811t, 812, 812t, 847 algae-bacterial selenium removal (ABSR) process, 730–731 alkali and alkaline earth metals, 101–102; barium, 101; lithium, 101–102 alternate row leaching, 663 aluminosilicates, 106 amorphous iron oxide, 104, 105

1077

1078

INDEX

Australia, 61, 326, 327, 928; MurrayDarling Basin, 760–763, 761f; regulatory and environmental issues, 779–780 Avicennia marina, 177 Ayers and Westcot model, 378

B Babcock, K. L., 64 Ball, J. W., 91 barium, 101 barley, 175, 177, 182, 183, 192, 193, 549, 576 basin irrigation, 513–514 battery smelting facilities, 115 bedding probe, 306, 307f bell peppers, 546 bentonites, 116 bicarbonate, 170 bioeconomic model, 954–957 biomethylation, 108 birnessite, 102 bloedite, 774 border irrigation, 513 border leaching, 663 boron, 14, 30, 40, 44, 108–110, 111f, 217t, 288, 776, 777; accumulation and plant toxicity, 697–698; constant capacitance model and, 874; crop response to, 431, 433, 434t–437t; irrigation water quality assessment of, 363–364, 364f; reclamation of boron-affected soils, 655, 676–678; salinity-boron interactions, 438–439; toxicity, 289, 433, 437–438 Bower, C. A., 75 brine shrimp, 746 Broadview Irrigation District, 392 Buena Vista Water Storage District, 725–727, 726f “Bureau of Soils cup,” 300–301 burial four-electrode probe, 306, 307f

C Caballo Reservoir, 1036–1037 cadmium, 115, 116, 216, 220–221 cadmium-chloride complex formation, 116 calcareous soils, 671–672

calcite, 62, 69, 774; as adsorptive surface for metals, 93; evaporative salinization and, 8; WATSUIT model and, 807 calcite precipitation, 871–872 calcium, 30, 170, 177–178, 567; deficiency of, 39, 40–41, 428–429; as soil amendment, 74 calcium carbonate (CaCO3), 81–82, 297 calcium chloride, as soil amendment, 664, 670–671, 679 calcium sulfate (gypsum), 297 California: evaporation ponds in, 737–738; extent of salinity problems, 6, 8t; lettuce field survey, 469–474, 470t, 471f, 472t, 473f; Mojave Desert, 103; toxic elements in San Joaquin Valley, 14–22, 22f. See also San Joaquin Valley canal leakage, locating and quantifying, 606 canal losses reduction, 607 carbonate, 170 carbonate alkalinity, 8, 65 carboxylic acid functional groups, 94 carrots, 421 cation exchange, 873–874 cation exchange capacities, 72, 314, 315, 901 cation exchange reactions, 899 cation ratio, 6, 362 cations, 93, 170, 297; adsorption or desorption of, 11; sodic soils and, 28 cauliflower, 549 CD-MUSIC model, 104, 105, 108, 111–112 celery, 549 cell expansion, 181 cell turgor, 183–184 cellulose acetate membranes, 724–725 cereal crops, 186. See also specific crop charge distribution multisite surface complexation model, 96, 97 chelation, 115 chemical conversion units, 1073 chemical speciation model, 63, 65, 67 chemistry, of world water, 6, 9f chloride, 170, 288, 567; crop response to, 430–431, 432t; osmotic stress and, 431; toxicity, 178–179 chlorine (Cl) toxicity, 30, 40, 44 chlorosis, 289 chromatographic (plate) models, 673 chromium, 102–104 chromium hydroxide, 104 CIRFLE, 930–936, 936t, 939t, 940t, 944t–947t, 948, 949

INDEX Ciudad Juarez, 1035 clay mineralogy, 147, 152 clay minerals, 72–73, 73t clay swelling and dispersion, 657–658 Clean Water Act, 923 climate: salinity and, 59–60; salt tolerance and, 30 Coachella Valley: irrigation and, 835; preand postleaching surveys, 474–478, 475t–476f, 477f–479f coal bed methane resources, 61 coal seam seep, 564 cobalt, 223 cogeneration-desalination, 750 co-ions, 73 collector ponds, 609 Colorado Plateau, 106 Colorado River Basin Salinity Control Program, 599–601, 602 Colorado River Water Quality Improvement Program, 747 compatible solutes (osmolytes), 175–176, 175t compensation habitats, 634, 635, 651 composite sampling, 284 computational economics models. See microeconomics of salinity/drainage management conceptual Irrigation Return Flow hydrosalinity model. See CIRFLE confidence level, 280 congruent dissolution, 58 constant-capacitance model, 96–97, 104, 105, 108, 110, 111–112, 118, 874–875 continuous ponding, 662 conversion factors for SI and non-SI units, 1071t–1072t conversion units, 1073–1074 copper, 115, 116 corn, 178, 179, 193, 216, 388, 546 correlograms, 281, 282 corrugation irrigation, 512–513 Corwin, Dennis, 120 cotton, 180, 190, 327, 387, 420, 553 counter ions, 73, 93 cowpea, 184, 549 cracking soil, 829–834, 833f; modeling for, 826; surface irrigation of, 844–845, 846f, 847–849, 848f, 848t, 849f, 850f crops: drip irrigation and, 543–547; effects of salts on, 29–31, 39–40; herbaceous, 407, 409t–415t, 416–417, 416f; on-farm salinity control practices, 519; response

1079

to specific ions/elements, 429–440; salt removal by, 374–376; salt tolerance of, 29–30, 745–746; trees and perennial crops, 963–964; woody, 417, 418t–419t, 420. See also agricultural salinity problems; plants; plant salt tolerance; specific crop crop-summerfallow rotation, 561, 562, 579, 581. See also saline seeps crop switching, 973 crop-water production functions, 953 crusting, 31, 33, 36–39 crystallization, 91 cucumbers, 421

D date palm, 420 Davies equation, 66 Debye-Huckel equation, 66, 67f, 901 decisiemen, 63 deep percolation, 843 deep-well injection, 612, 742–744 depositional crust, 152–155 desalinization, 723–724 desalinization studies, 728–729 desalting processes, 605 diagnosing salt problems, 31–41 diffuse double layer, 73–74, 75f, 93 diffuse-layer model, 96, 97, 104, 105, 107, 108, 110, 111 dispersion, 28 dissolution, 122; minerals, 91; trace elements and, 90–91 dissolution reaction: dolomite, 69; gypsum, 69; precipitation, 69, 81; silicate, 59 ditch lining, 609 dolomite, 62, 69; rhizosphere chemistry, 81 drainage, 405; salinity and inadequacy of, 35. See also microeconomics of salinity/drainage management; onfarm irrigation and drainage practices; San Joaquin Valley Drainage Program drainage reuse, 636–637, 640, 642 drainage water disposal, 733–744; conditions, 734; deep-well injection, 742–744; evaporation ponds, 737–738; Integrated On-Farm Drainage Management, 738–739, 740f; land application and irrigation, 739, 741;

1080 safe disposal considerations, 733–734; solar evaporators, 741–742; to surface waters, 734–737, 736f drainage water reclamation and reuse, 17t, 19, 20, 20t, 744–751, 955, 964, 966, 972, 984, 999; aquaculture, 746; cogeneration-desalination, 750; power plant cooling, 746–747; reuse standards, 747; salinity-gradient solar pond, 747–749; salt recovery, 749–750, 751t; salt-tolerant crops, 745–746, 745f drainage water treatment, 722–733; Buena Vista Water Storage District, 725–727, 726f; desalinization, 723–724; Firebaugh Water District, 724; Los Baños Demonstration Desalting Facility, 724–725; Panoche Drainage District, 727; technology needs, 722–723, 723t; trace element treatment, 729–733 drainage wells, 527 drainwater treatment, 17t, 19, 20t drip irrigation, 539–556; advantages of, 539; cracking soil, 829–834, 833f; crop yield response to salinity under, 543–547; disadvantages of, 539–540; evapoconcentration/precipitation algorithm (grape crop), 835–837; fieldwide uniformity and, 539; grapes, 841–844, 842f, 843f; notation, 559; research overview, 826–827; salinity control under, 547–550, 548f, 549f, 551f; salt patterns, 541–543, 541f–543f; San Joaquin Valley case study, 551–555, 553t; water and salt flow, 825–826; water flow model, 827–834, 829f, 830f; wetting patterns, 540–541, 540f, 826 dryland saline seeps. See saline seeps DTPA-TEA extraction method, 214 DUALEM-2, 309–310 durum wheat, 421 dwarf saltwort, 708

E EC. See electrical conductivity (EC) Economic and Environmental Principles and Guidelines for Water and Related Land Resource Implementation Studies, 591 Egypt, 390 Eijelkamp conductivity meter, 306

INDEX electrical conductivity (EC), 62–63, 299–304, 301f, 302f; as expression of soil salinity, 2; factors influencing ECa measurement, 312, 313t–314t, 315, 315f; predicting, from ionic composition, 64–65; salinity of irrigation water, 34; temperature and, 298 electrical resistance, 62 electrical resistivity (ER), 304–308, 305f, 307f, 308f electric double layer, 93 electrodialysis reversal, 725 electrolyte concentration, and sodicity, 151–152, 154f electromagnetic conductivity meter, 286, 287, 289 electromagnetic induction (EMI), 308–310, 309f electron-microprobe energy-dispersive x-ray analysis, 119 Elephant Butte Reservoir, 1036–1037, 1040, 1047 El Paso, Texas, 1035 Energy Innovation Small Grant program, 727 entisols, 60 environmental issues, 779–780 Environmental Protection Agency priority pollutants, 89 ESAP software package, 468–469 European Water Framework Directive, 924 evapoconcentration, 6, 8–9, 838 evapoconcentration factor, 772 evaporation basins, 757–780; closure of, 779; cost of, 764; design, 763–764; disposal capacity, 764–767, 766f, 767f; hydrology, 767–769, 768f; monitoring water quality and sediments, 769–770, 769t; in Murray-Darling Basin (Australia), 760–763, 761f, 765; notation, 785; pond biology and selenium toxicity, 777–778; pond seepage rates, 767–769; redox transformations, 774; salinity and water chemistry, 770, 772–774, 773f; in San Joaquin Valley, 758–760, 769t, 771t, 774–778; wetland habitats in, 778 evaporation ponds, 17t, 19, 20, 20t, 611, 633–636, 645–647, 647f, 650, 737–738, 751; chemistry of, in Tulare Lake Basin, 1005–1006 evaporative salinization, 3, 8–9, 10f

INDEX evaporites, 100 evapotranspiration (ET), 60, 295, 296, 296f, 757, 801 exchangeable cations, 297, 300 exchangeable sodium percentage (ESP), 28, 170, 658–659; hydraulic conductivity of smectitic soils and, 158, 160f Extract Chem program, 63, 65, 67, 82, 300

F farm and regional agricultural production, 966–972; empirical studies, 968–969; programming model, 967–968; spatial dynamics, 969–972 farm control volume, 1054–1057 ferrous hydroxide, 731 fertilizers, 33, 113, 114; fertilizer salts, 376; mercury concentrations in, 117 field sampling: bias, 276, 279; B toxicity, 289; composite sampling, 284; evaluation of strategies, 284–285, 285t; gravimetric sampling, 287, 290; independence of observations, 281–282; infiltration rates and, 287–288; judgment sampling, 276, 279, 289; local variability, 283–284; number of samples, 280–281; objectives, 275; plant parameters, 288–289; questions to be answered, 275; selecting a technique, 279–280; simple random sampling, 276–277, 284; size of area sampled (variance), 282–283, 283f; soil parameters, 285–286; soil salinity parameters, 286–287, 290; spatial distributions, 280; strategies involved in, 276; stratified random sampling, 278, 284, 289; systematic sampling, 279, 284; techniques, 276–279, 277f; volume-variance relationships, 282; water parameters, 288, 290 field-scale crop growth and management model, 954–957; empirical studies, 958–959; extensions, 959–960; model solution procedures, 957–958 field-scale soil salinity measurements, 316–324, 329; geospatial ECa, 317–319; geospatial ECa-based design, 319–323; remote imagery, 324–328; survey considerations, 323–324; visual crop observation, 316–317

1081

Firebaugh Water District, 724 Flow-Through Wetlands Project (Tulare Lake), 729–730 fossil deposits, 60–61 four-electrode salinity probe, 286, 287, 289 foxtail barley, 569 Freundlich equation, 95–96, 110 fruit orchards, 115, 395 furrow irrigation, 512

G Gapon exchange reaction, 78–79 gases, emission or absorption of, 11 General Circulation Model climate projections, 918 geochemical models, 85 geochemical weathering, 57–59 geographic information system, 100, 120 geologic outcrop seep, 564 Geonics EM-31 and EM-38, 309–310, 309f geospatial ECa: measurement, 317–319, 318f, 320f, 321t; soil sample design based on data of, 319–323; survey considerations, 323–324, 325f geostatistical analysis, 100 geostatistical mixed linear models, 462 germination, 179–180 Gibbs, R. J., 6 Girsu, 12–14 glaciated Fort Union seep, 564 glauberite, 774 global scale, of agricultural salt problems, 3–6 glycophytes, 169, 175 glycophytic crops, 260 goethite, 95, 118 Grand Valley Unit, 600–601 grapes, 835–837, 841–844, 842f, 843f Grassland Water District, 735, 736f gravimetric sampling, 287 Great Plains, 562, 563f, 567, 569, 573, 575 grid sampling, 468 groundwater: chemical composition of, 70, 71t, 72; effect of shallow groundwater, 380–390, 381f, 385f–390f; farm control volume, 1054–1057; irrigation district control volume, 1058–1061; management, 18t, 19, 20–21, 20t; management model framework, 1063–1067; regional control volume management, 1061–1063; San Joaquin

1082 Valley Drainage Program and, 625–629; sedimentary rocks/soils and, 100; shallow groundwaters management, 528–533; trace elements in, 89, 99–118 (see also specific trace element); upward flow, 34, 39, 296, 371, 372–373, 381–383 gypsum, 29, 30, 36, 62, 297, 774; evaporative salinization and, 9; as irrigation water amendment, 696, 790, 796, 797t; recovery and reuse, 750; runoff and erosion from sodic soils and, 161–162, 162f, 163f; salinity measurement of, 300; as soil amendment, 69–70, 74, 665–669, 665f, 669f, 674–675, 679; solubility of, 70; SWS model and, 872; WATSUIT model and, 807

INDEX hydrolyzates, 100 hydromagnesite, 81 hydrosalinity modeling, 11 hydrosalinity models, 923–949; applications of, 928–930; CIRFLE, 930–936, 936t, 939t, 940t, 944t–947t, 948, 949; complex models, 926–927; hydrologic submodel, 930–933, 931f; limitations, improvements, and future research needs, 948; model application, 943–946, 944t, 947t; model calibration and validation, 936–938, 936t, 939t; notation, 952; range of, 925–926, 949; salinity submodel, 933–936, 933f; salt balance and irrigation return flows, 923–926, 925f; sensitivity analysis, 938–943, 940t, 941f, 942f; simple, conceptual models, 927–928 hydrostatic pressure seep, 566 HYDRUS-2D, 554, 555

H halite, 774 HALOPH, 422, 427 halophytes, 169, 175, 237, 260, 708–709; potential uses of, 423, 426–427 Hardie-Eugster model, 8 Harvie et al. species constants, 66 Hatcher, J. T., 75 HCl and HCl = H2SO4 extractant, 215 heavy metals, 90, 115, 116, 122; removal of trace elements, 731 herbaceous crops, 407, 409t–415t, 416–417 heterogeneous nucleation, 82 Hickey, M. G., 115 hierarchical spatial models, 462 high-salt dilution method, 672 Hilgard, Eugene, 15 Hoffman-van Genuchten model, 377–378, 379 Hohokam Indians, 14 hot-water-soluble B extractant, 216 Hueco Bolson aquifer, 1035 hydraulic conductivity (HC) of soil, 171; aging and, 158–159; clay content and wetting rate and, 156, 157f, 158; exchangeable sodium percentate/wetting rate and, 158; pH and exchangeable magnesium and, 149–150; very dilute solutions and, 147–149, 148f HYDRO-GW model, 929 hydrologic submodel, 930–933, 931f

I illite clays, swelling and dispersion in, 145 Imperial Valley, 835 incongruent dissolution, 58–59 India, 326 indigenous salts, 35–36 Indio, California, lettuce field survey, 469–474, 470t, 471f, 472t, 473f Indus Plain region, 14 industrialization, 61 infiltration rates, 287–288; effect of wetting rate and sodicity on, 159, 161; ESP of soil and, 658; in sodic soils, 150–152, 153f; sodic soils and, 655; in soils with stable surface structures, 658 insertion four-electrode probe, 306, 307f Integrated On-Farm Drainage Management, 738–739, 740f intermittent ponding, 662 international treaties, 603–604 intracellular osmolyte levels, 249–254 iodine bush, 708, 739 ion-activity product, 90 ion adsorption, 731 ion complexation, 901 ion effects: nutrition, 177–178; specific toxicities, 295; toxicity, 178–179 ion exchange, 77–80, 115; effect of salinity on, 80; Gapon constant, 78–79; ion demixing, 80; mass action, 77–78

INDEX ion homeostasis mechanisms, 249–254 ion hydrolysis, 300 ionic balances, 362 ionic composition, predicting EC from, 64 ionic strength effect, 69 ion pairs, 11, 70 iron (Fe), 30, 41 iron-oxide deposits, 106 irrigated agriculture viability through time, 1053–1068; agricultural viability, 1067; agroecosystem viability, 1064–1067; extending spatial and temporal extent, 1057–1058; farm control volume, 1054–1057; irrigation district control volume, 1058–1061; management model framework, 1063–1067; regional control volume management, 1061–1063 irrigation: analyzing practices of, 596–597, 597t; avoiding rootzone salt accumulation, 48–49; ditch lining and pipelines, 609; drainage water disposal and, 739; evaporation basins and, 757; improvement of on-farm management, 607–610; indigenous soil salts and, 36; installing new or improved systems, 609; local variability and, 283–284; low infiltration rate and, 655; maximizing soil water potential, 48; multiyear management, 960–966; on-farm drainage, 609; on-farm water management, 609; optimal management of, 804; oxygen deficiency and, 193; plant salt tolerance and methods of, 444–445; reuse of saline drainage waters for, 605–606, 687–711; salinity and, 11, 13f, 33–34, 39, 61–62, 170, 295 (see also WATSUIT model); salinity control measures, 44; salt tolerance and, 30; scheduling, 521–522, 531–533; singleseason management, 954–960; trace elements in areas of, 100–101, 102, 113, 117, 170; underirrigation, 38; uniform and efficient application of, 46–47; Upper Arkansas River, 601. See also drip irrigation; leaching requirements; on-farm irrigation and drainage practices irrigation district control volume, 1058–1061 irrigation return flows (IRFs), 923–926, 925f

1083

irrigation water quality assessments, 343–368; boron, 363–364, 364f; electrical conductivity, 346; ionic balances, 362; plant response to soil salinity, 348–352; salinity, 345–352, 368; sodicity, 352–361; soil-water extracts, 346–348; trace elements, 365, 368

J judgment sampling, 276, 289

K kaolinite clays, swelling and dispersion in, 145 kaolinite edges, 94 kenaf, 182 Keren equation, 110 Kesterson National Wildlife Refuge, 98, 759 Kesterson Reservoir, 15, 107 Kittrick, J. A., 115 Kohlrauch’s Law, 64, 298 K+ outward-rectifying channels, 238 kriging, 282

L laboratory salinity measurements: direct and indirect analysis, 297–299; electrical conductivity, 299–304, 301f, 302f; electrical resistivity, 304–308, 305f, 307f; electromagnetic induction, 308–310, 309f; time domain reflectometry, 311–312 land leveling, 514 land retirement, 18t, 19, 20, 20t, 627–628, 644–645, 973 landscape, 59–60; salt-tolerant plant species for, 421–423, 424t–426t Langelier Saturation Index, 787 Langmuir equation, 96, 110, 676 Langmuir isotherms, 104 Las Cruces, New Mexico, 1037 leaching, 74, 406, 522–524, 523f, 659–662, 679; alternate row or border leaching, 663; continuous ponding, 662; under drip irrigation, 547–548; effect of shallow groundwater, 380–390, 381f, 385f– 390f; intermittent ponding, 662;

1084 notation, 402–403; one-dimensional, 660–661, 661f; post-leaching operations, 663; rootzone salinity control and, 371, 693–694; salinity and inadequacy of, 34–35; soil salinity without, 390–392, 391f; of spatially variable fields, 662; sprinkling, 663; surface flushing, 663; twodimensional, 661; water and salt balance, 371–376. See also leaching fraction (LF); leaching requirements leaching fraction (LF), 297, 372, 548–550, 802–804, 802f, 822 leaching requirements, 376–380, 378f, 379t, 801–822; background information, 801–805, 802f; implications of model estimates, 818–822, 818t; model inputs, 809–810; model requirement estimates, 810–818, 811t–812t, 813f–815f, 816t–817t; notation, 824; steady-state models, 803, 805–807; TETrans model, 805, 808–809, 812, 815, 818; transient models, 804–805, 808–809; UNSATCHEM model, 805, 806t, 809, 816, 816t, 818, 819, 820; WATSUIT model, 807–809, 810–811, 818–820 lead, 115, 116 leaf expansion, 181 leaf extension, 181 lettuce, 216, 544, 548, 810, 811t, 812, 812t, 847 likelihood ratio test, 466 Limonium, 179 linear regression models, 462, 476, 477 liquid phase pathway, 314 lithium, 101–102 livestock, 699, 708 Los Baños Demonstration Desalting Facility, 724–725 lotus, 193 low-affinity K+ channels, 238 lysimeter columns, 299 lysimeter study, 386–387

M Maas and Gratan threshold tables, 544 Maas-Hoffman model, 408 Mackie, W. W., 15 magnesium, 567 magnesium carbonates, 81

INDEX magnesium (Mg), 44, 170; sodicity and, 1, 28 magnesium precipitation, 872–873 maize, 190, 260, 546, 547 Mancos shale formation, 60–61 mandarin oranges, 421 manganese, 103–104 mangrove, 177 manure, 33, 113, 114 Marion, G. M., 64 Massachusetts, 103 mass action principle, 73, 77–78 Mayenkar process, 731 McNeal, B. L., 65 measuring salinity, 2, 62–65; predicting EC from ionic composition, 64–65; saturation extract, 63–64 Mehlich 3 extractant, 215 melons, 421 membrane scaling, 728 mercury, 117–118, 122 mesopotamia, 12–14 metalloids and nonmetals, 107–113; arsenic, 107–108, 109f; boron, 108–110, 111f; selenium, 110–113, 112f Methods of Soil Analysis, 297 methyl mercury, 117 Mexico, 327; -U.S. treaties, 603–604 mhos, 62 microbial oxidation of reduced organic carbon, 98 microeconomics of salinity/drainage management, 953–974; drain water reuse, 972; farm and regional agricultural production, 966–972; multiyear irrigation management, 960–966; notation, 975–976; single season irrigation management, 954–960 microirrigation, 517–518 mineral dissolution, 300 mineral precipitation, 838 mineral precipitation-dissolution, 901 mineral saturation index, 90 mineral solubility, 3 mineral solubility product constant, 90 mineral-water equilibria, 91 mineral weathering, 149, 373–374, 375f, 657 mining activities, 61, 113 mini-sprinklers, 392–393 mirabilite, 774 miscible displacement models, 673 mixing and dissolution algorithm, 837 mobile and immobile model, 827 MODFLOW model, 928

INDEX MODHMS model, 914 Mojave Desert, 103 mole drains, 573 molybdate, 777 molybdenosis, 699 molybdenum, 105, 106f, 221–222; accumulation of, with saline drainage waters, 698–699 montmorillonite platelets, 656 Moran residual test statistic, 466 M2TDMS model, 928 Mud Slough, 735–736 multispectral satellite data, 327 multiyear irrigation management, 960–966; salt accumulation, 961–963; spatial variability and irrigation system investment, 965–966; tree and perennial crops, 963–964 Murray-Darling Basin (Australia), 765; evaporation basins in, 760–763; Riverine Plain, 760, 761f, 762–763, 765–766, 768; water chemistry, 770, 772 Murrumbidgee Irrigation Area, 763, 770 muskmelons, 421

N NaHCO3 reagent, 216 Natural Resources Conservation Service, 601 near-surface electrical resistivity, 305 necrosis, 289 nesquehonite, 774 nickel, 115, 116, 120 Nile Delta, 390 nitrates, 14–15, 170, 177; algae-bacterial removal, 730 nonhalophytes, 169, 190 nonmetals. See metalloids and nonmetals Nordstrom, D. K., 91 normalized difference vegetation index, 327 nutrition, 33; diagnosing, 31; as ion effect, 177–178; salt tolerance and nutritional imbalance, 427–429

O oasis effect, 765, 766 oats, 549, 576 ocean disposal, 734

1085

one-dimensional leaching, 660–661, 661f on-farm irrigation and drainage practices, 511–533; basin irrigation, 513–514; border irrigation, 513; confined aquifers and, 526–527; disposal of drainage effluent, 527–528; distribution of applied water in soil, 514–515; drainage of irrigated fields, 524–525; drainage wells, 527; drains, 525; furrow and corrugation irrigation, 512–513; interception of lateral flow, 525; irrigation and salinity control, 511–524; irrigation scheduling, 521–522; land leveling, 514; leaching, 522–524, 523f; microirrigation, 517; nonirrigated saline fields, drainage of, 526; notation, 538; salinity control, 519, 520t, 521–522; salt accumulation patterns, 515–516, 516f; sequential biological concentration, 528; shallow groundwaters management, 528–533; sprinkler irrigation, 516–517; surface irrigation, 512; tillage, 521; water measurement, 522; water table depths, 526 onions, 421, 546 Onsanger-Fuoss equation, 64 orchards, 115 ornamental species, salt-tolerant, 421–423 osmoprotectants, 176 osmotic desication, 173 osmotic effects, 173–177 osmotic pressure calculation, 67, 68f, 870 overparameterization, 927 oxidized sulfide ores, 106 oxyanions, 95, 105, 107, 122

P Pakistan, 326 Palmer Drought Index, 562, 563f Palo Verde Irrigation Outfall Drain, 747 Panoche Drainage District (California), 483–509, 727; calculation of spatial distribution of salt balance, 487–497, 488f; data, 488; drainage charges, 502–507, 505f–508f, 508; drainage disposal cost distribution, 496–497; drained control volumes, 489–490; local geology of, 484–487, 485f, 486f; notation, 510; selenium balance, 493–495, 494f–496f, 500–502, 502f, 503f, 504f, 505t; total dissolved solids

1086 balance, 488–493, 497–507, 498f–501f; undrained control volumes, 490–493, 491f, 492f, 493f parallel pathways of conductance, 315 Pauwlonia tree hybrid, 746 Pecos Valley, 687 peppers, 395 periodicity, 279–280 permeability, 27–28, 31, 33; crusting problems and, 36–39; potential causes of, 38–39; soil amendments for increasing, 43 pesticides, 113, 114 Phaseolus vulgaris, 177 phosphate fertilizers, 114 phosphorus (P), 30, 41, 216 photosynthate, 176 photosynthesis, 182–183 Phragmites australis, 194 PHREEQC model, 66 Physcomitrella patens, 194 phytotoxicity, 89, 114, 115 pipelines, 609 Pitzer expressions, 66, 901 Pitzer formulations, 70 planting practices, 519 plants: calcium signaling and salt stress, 189, 189f; cellular uptake of sodium, 187–188, 187f; development, and response to salinity stress, 179; effects of salts on, 29–31; field sampling, 288–289; germination/seedling emergence, 179–180; mechanisms of response, 173–179; notation, 204–205; nutrition, 177–178; osmotic effects, 173–177, 174f, 175t; principal responses to salinity, 171–173, 172f; principal responses to sodicity, 172f; reproductive growth, 184–187; response to salinity/water stress, 890–893; salinity stress during development of, 179; salt exclusion, 189–193, 190f, 192t; salt tolerance, 193–194 (see also salt tolerant plant development); toxicity and nutritional imbalance problems, 39–41, 177–179; vegetative growth, 180–184, 182f; water relations, 183–184. See also crops; plant salt tolerance; specific plant plant salt tolerance: boron, 431–439, 434t–437t; chloride, 430–431, 432t; controlling soil salinity, 446–447; crop response to specific ions/elements,

INDEX 429–440; halophytes, potential uses of, 423, 426–427; herbaceous crops, 407, 409t–415t, 416–417; irrigation methods and, 444–445; ornamental and landscape species, 421–423; parameters influencing, 440–445; quality of crops, 420–421; relative yield-response curves, 407–427; saline sprinkling waters and, 445–446; salinity and nutritional imbalance, 427–429; salt composition and, 441–442; selenium and other trace elements, 439–440; sodium, 429–430; soil biota, 443–443; soil fertility and, 443–444; soil water content and, 440–442; woody crops, 417, 418t–419t, 420 playa-type topographic features, 36 plot-scale soil salinity measurement, 299 plums, 395, 420 pollution: mercury, 117; priority pollutants, 89 polyacrylamide, 161–162, 162f, 163f porous cup extractor, 302, 302f, 303 porous-matrix salinity sensor, 302, 302f potassium, 170 potatoes, 546 pothole seep, 566 power plant cooling, 746–747 precipitated salts, 297 precipitates, 100 precipitation, 122; trace elements and, 90–91 prediction-based sampling strategies, 468 preferential flow, 826 principal component analysis, 100 priority pollutants, 89 probability-based sampling strategies, 467–468 project-level salinity management programs, 591–597, 592f, 593t, 595f; augmentation of water supplies, 604–606; basin-wide jurisdiction, 602; climate change and weather modification, 604–605; Colorado River Basin Salinity Control Program, 599–601; defining the problem, 592–596; desalting processes, 605; economic factors, 594; effect of water rights, 602; efficiency of, 607; environmental factors, 595–596; inter-basin fresh water transfers, 604; international treaties, 603–604; irrigation practices, analysis of, 596–597, 597t; lack of urgency,

INDEX 602–603; legal and institutional factors, 602–604; on-farm irrigation systems/ management, improvement of, 607–610; physical factors, 593–594; reuse of drainage water for irrigation, 605–606; salinity control options, 597–598; San Joaquin Valley Drainage, 598–599; summarized, 612–613; Upper Arkansas River, 601; vegetation and watershed management, 605; wastewater management, 610–612; water delivery systems improvement, 606; water quality standards for salinity, 602 Prosopis alba, 746

Q quantile (QQ) plots, 467 quantitative trait loci, 237

R radishes, 216 rainfall, 373; of variable composition, 6 range, determining, 281 reactivity of salts and salt flows, 6, 9–12, 9f; cyclic salinization and dilution, 10, 11; evaporative salinization, 8–9, 10f; soil-water systems, 10–11, 12f; world water chemistry, 6, 9f recharge areas, 571–572 reciprocal ohms, 62 reclaimability, assessing, 41–44; adequacy of soil structure and need for tillage, 43; estimating time required for, 43–44; leachability, 42–43; other considerations, 44 reclamation of boron-affected soils: boron hazard, 676–678, 679; concepts and principles, 676; methods and models, 676–678, 677f reclamation of sodium-affected soils, 664–675, 679; concepts and principles, 659, 664; high-salt dilution method, 672; notation, 684–685; preventing crust formation, 674–675, 674t, 675f; salt leaching and water requirements, 659–662 (see also leaching); with soil amendments, 664–671, 679; soil profile modification, 672–673; soil reclamation

1087

models, 673–674; without soil amendments, 671–673 redox reactions. See reduction-oxidation (redox) reactions Red Rock Ranch, 647f, 649t; selenium and nitrate removal, 730–731; solar evaporator, 741, 742f reduction-oxidation (redox) reactions, 122; in evaporation ponds, 774; trace elements and, 97–99; weathering, 59 regional salinity management programs. See project-level salinity management programs regional-scale modeling of soil salinity, 899–919; application, 914–919, 914f, 917f; averaging boundary conditions and vertical discretization, 912–913, 913t; benchmark numerical simulations, 903, 904t, 905, 905t–907t, 908–912, 909f–911f, 919; description of numerical model, 900–901, 902t; methods, 900–908; model simplifications, 905, 908; notation, 922; results and discussion, 908–919; salt chemistry simplification, 913t, 914 remote imagery soil salinity measurements, 324–328, 328t, 329 residual analysis techniques, 467 resistates, 100 retiring land, 18t, 19, 20, 20t reuse systems, 609 reverse osmosis, 723–724, 727, 728t rhizosphere chemistry, 80–84; adjusted sodium adsorption ratio, 82–83, 84t; calcium carbonate, 81–82; magnesium carbonates, 81 Rhoades equation, 806 Rhoades model, 378, 379 rice, 183, 184–187, 186f, 190, 192, 260 Richards’ equation, 809, 826, 901 Rio Grande basin: allocation and management, 1035–1036; evapotranspiration and, 1047; mitigation of river salinity, 1046–1048; overview, 1033–1048; Rio Grande Project, 1036–1037; salinity concern and impacts, 1037–1038; salinity dynamics, 1041–1046, 1042f–1045f; salinity variation with flow distance, 1038–1041, 1039f, 1040f; topography, climate, and water use, 1034–1035; total dissolved solids (TDS), 1037, 1039f, 1040f, 1041, 1042f–1043f, 1044–1047

1088 Rio Grande Compact, 1036 river discharge, 18t, 19–20, 20t, 21 Riverine Plain, 760, 761f, 762–763, 765–766, 768 root water uptake, 901 rootzone: excess salinity within, 29, 30, 32, 33; managing salinity in, 62; modeling transient salinity of, 855–858 (see also SWS model); salinity control with drip irrigation, 547–550; WATSUIT model for tracking chemical reactions in, 787–798. See also leaching; leaching requirements roses, 423 runoff, excessive, 38 Russian olive, 1047 rye, 420

S safflower, 180, 576 Salicornia, 708 Saline Agriculture: Salt-Tolerant Plants for Developing Countries, 709 saline drainage waters for irrigation, 687–711; blending water supplies, 689; boron accumulation and, 697–698; crop quality and, 699–700; cyclic use of saline/nonsaline water, 690; direct sodic effects on plants, 697; extreme halophytes, 708–709; farming practices to promote, 695–697; field research studies, 700–703; fundamental principles related to, 692–700; leaching to control rootzone salinity, 693–694; notation, 719; saline-sodic waters, 694–695; salt tolerance, 692–693; salttolerant forages, 706–708; selenium and molybdenum accumulation, 698–699; sequential use, 690–692, 691f; strategies, 688–692; testing California crops for, 703–709, 704t; tree crops, 704–706; vegetable crops, 703–704; water quality impacts, 695, 696t; worldwide efforts, 709 saline seeps, 297, 561–584; agronomic practices, 573–577, 574f, 575t, 576t, 578t; control methods, 572–579; cropping strategies, 577; development factors, 562, 563f, 564, 565f; drainage, 572–573; identification of recharge/discharge areas, 569–572,

INDEX 570f; recharge area location, 571–572; reclamation of controlled areas, 579–582, 580f, 581f, 582f, 583t; socioeconomic concerns, 582–583; types of, 564; water quality associated with, 567, 568t, 569 saline water spray, 33 salinity: constituents, 2; criterion, 3; cyclic salinization and dilution, 10, 11; defined, 1, 2; diagnosing problems of (see diagnosing salt problems); measurements, 2; microeconomics of management (see microeconomics of salinity/drainage management); parameters, 2; and sodicity compared, 170–171. See also agricultural salinity problems; salinity sources; soil salinity salinity coefficients, 407 salinity control and management measures, 44–52, 597–598; growing salt-tolerant crops, 45; irrigation and, 46–47; minimizing accumulation in seed bed, 45–46; plant toxicity and rootzone salinity/sodicity, 44–45. See also project-level salinity management programs salinity-gradient solar pond, 630–631, 747–749 salinity management options, 591–613; defining the problem, 592–596; projectlevel management programs, 591–597, 592f, 593t salinity patterns, based on soil electrical conductivity survey data, 461; data analysis examples, 469–478; ESAP software package, 468–469; estimation and prediction formulas, 462–465; notation, 482; regression models, 462–465; regression model validation tests, 466–467; spatially referenced linear regression models, 467–468 salinity sources, 3; anthropogenic activities, 3, 61–62; atmospheric deposition, 61; climate and landscape effects, 59–60; fossil or secondary deposits, 60–61; irrigation and, 11, 13f; weathering and, 57–59 salinity submodel, 933–936, 933f salt accumulation patterns, 515–516, 516f, 517, 518 saltcedar, 1047 salt exclusion, 189–193 salt grass, 708, 729

INDEX salt index, 33 salt loading reduction, 607, 608t Salt Management Guide (Tanji), 423 SALTMOD, 928 Salton Sea, 610, 611f salt overly sensitive (SOS) pathway, 188, 239 salt precipitation, 374 Salt Slough, 735–736 salt tolerance, 29–31; drainage water reuse option, 745–746; irrigation with saline drainage waters and, 692–693. See also plant salt tolerance; salt-tolerant plant development salt tolerance index, 408 salt-tolerant plant development, 235–261; agricultural productivity and salt stress, 236–237; conclusions and perspectives, 258–260; using transgenic approaches, 237–239, 240t–247t, 248–258 salt utilization, 18t, 19, 20t, 21 Salt Utilization Technical Committee Report, 639 San Joaquin Valley, 98, 104; agricultural drainage efforts, 985–995; arsenic levels, 775–776; boron, uranium, molybdate, and vanadium, 777; Broadview Irrigation District, 392; cotton in, 387; desalination studies, 728; drainage litigation, 995–1002; drainage management alternatives, 16, 17t–18t, 18–22, 22f, 735; drainage plan development, 998–1000; drainage water treatment, 722; drip irrigation case study, 551–555, 553t; evaporation basins in, 758–760, 759f; evaporation pond chemistry, 1005–1006; geographics of, 978–985, 979f, 981–982; grass forages, 687; irrigation development/drainage needs, 982–985; modeling study, 916; northern valley area, 1010; on-farm drainage efforts, 1003; regulatory/environmental issues, 778–780, 1014–1016, 1018–1022, 1020f–1022f; River basin, 980–981, 980f; saline drainage waters and, 687; salinity and, 981–982; San Joaquin River and salt management, 1010–1013, 1011f, 1013f; selenium levels, 764, 775, 1013–1018, 1015f, 1022–1024; solar evaporators, 741–742,

1089

743f; toxic elements in subsurface drainage waters, 14–22, 22f; trace element concentrations, 769t, 771t, 774–778; trace elements in, 105, 107, 112, 119; Tulare Lake basin, 981, 1003–1005; water chemistry, 772; Westside Regional Drainage Plan Proposal, 21; wildlife concerns, 1006–1010 San Joaquin Valley Drainage Implementation Plan, 598–599, 723 San Joaquin Valley Drainage Program, 15–16, 617–651; crop use of shallow groundwater, 624–625; discharge to San Joaquin River, 638–639; drainage reuse, 636–637, 640, 642; drainage treatment, 642–644; drainage water treatment, 629–633; evaporation ponds, 633–636, 645–647, 647f, 650; groundwater management, 625–627, 648; irrigation and drainage management, 622–624; land retirement, 627–628, 644–645; problem statement, 619–620; reducing drainage water volume, 622–629; river discharge, 648; salinity-gradient solar ponds, 630–631; salt utilization, 639–640, 649–650, 649t; selenium removal treatment, 631–633; selenium utilization, 649–650, 649t; solar evaporators, 631; solution approach, 621–622, 621f; source reduction, 647 San Luis Drain, 15, 728, 728t, 759 San Luis Valley, 1046 saturated soil paste, 299–300, 301, 304 saturation extract, 63–64 saturation index, 90–91 saturation percentage, 63, 314 scanning electron microscopy, 119 Schlumberger, Conrad, 305 sea-blithe, 375 seal formation, 152 secondary deposits, 60–61 sedimentary rocks, and groundwater chemistry, 100 seedling emergence, 179–180 selective land retirement, 610 selective mineral precipitation, 6 selenate, 111–112 selenite, 111–112 selenium, 110–113, 112f, 222–223, 779; accumulation of, with saline drainage waters, 698–699; algae-bacterial

1090 removal, 730; biological processes for removal of, 729; in evaporation basins, 774; in Panoche Drainage District, 493–495, 494f–496f, 500–502, 502f, 503f, 504f, 505t; plant tolerance for, 439–440; in San Joaquin Valley, 14–15, 16, 113, 775, 1013–1018, 1015f; San Joaquin Valley Drainage Program and, 631–633; toxicity to wildlife, 749; as trace element, 100–101, 777–779; trace element removal methods, 731–733; utilization, 649–650, 649t sensitivity analysis, 938–943, 940t, 941f, 942f sepiolite, 9, 81 sewage sludge, 113–114 shallow groundwaters management, 528–533, 530f; irrigation scheduling, 531–533; operation, 529; saline subsurface water use, 530; salt accumulation and distribution, 529; soil salinity management, 530–531 Shapiro-Wilk test, 467 Shepparton Irrigation Region, 770 short-circuiting flow, 826 siemens (S), 62–63 simple random sampling, 276–277, 284 slaking, 28 slope change seep, 566 slope-threshold model, 407 sludge-amended soils, 115 smectite mineralogy, 144–145 smectites, 28, 656 smectitic clay, 141–144, 142f, 143f smelting, 113 SNORM program, 101 sodicity: crusting problems and, 36–37; defined, 1; development of, 28; effect of pH on infiltration, 355–357, 356f–359f, 360–361; exchangeable sodium percentage (ESP), 28, 170; hazard guidelines, 353–355, 354f; irrigation water quality assessment, 352–361; lime content and, 149; nutrition/toxicity and, 173; and salinity compared, 170–171; sodium adsorption ratio (SAR), 28–29, 170–171; solution composition and hydraulic conductivity, 145–147 sodium, 30, 170–171, 288; crop response to, 429–440; saline seeps and, 567; toxicity, 178–179. See also salinity; sodicity

INDEX sodium adsorption ratio (SAR), 28–29, 170–171; effect of rain on, 882–885, 883f–886f; WATSUIT model, 787, 791–792 sodium sulfate recovery and reuse, 749–750 soil: amendments, 43, 664–671, 679; effects of salinity/adsorbed ions on properties of, 656–659; effects of salts on, 27–29; evaluating pollution potential in, 118–122, 121f, 123; extractants, 119, 123; hydraulic conductivity, 171; sodicity of, 170; source/causes of soil salinity, 406; texture, 147, 151 soil chemical extractants, 119 soil electrical conductivity survey data. See salinity patterns, based on soil electrical conductivity survey data soil permeability and crusting, 36–39; diagnostic criteria, 37–38; measurements, 37; potential causes of, 38–39; visual indications, 36–37 soil reflectance, 327 soil response to saline/sodic conditions, 139–163; depositional crust, 152–155; infiltration rate in sodic soils, 150–152; notation, 167; polyacrylamide and gypsum and, 161–162; properties affecting, 147–150; solution composition and clay swelling/ dispersion, 141–145; solution composition and hydraulic conductivity, 145–150, 146f; susceptibility of soils, 155–161 soil salinity: controlling, 446–447; defined, 295; diagnostic criteria, 33; direct and indirect analysis of, 297–299; field assessment of, 569–571, 570f; historic playa and, 36; inadequate drainage and, 35; inadequate leaching and, 34–35; indigenous salts and, 35; integration of, by crops, 392–398; integration over time, 395–398, 396f–398f; integration with soil depth, 392–395, 393f, 394f; irrigation water and, 33–34; measurements, 32; notation, 402–403; plant response to, 348–352; potential causes of, 33; salt sources, 295; visual indications, 31–32, 569. See also plant salt tolerance; regional-scale modeling of soil salinity; salinity sources

INDEX soil salinity measurement: electrical conductivity (EC), 299–304, 301f, 302f; electrical resistivity (ER), 304–308, 305f, 307f, 308f; electromagnetic induction (EMI), 308–310; factors influencing ECa measurement, 312, 313t–314t, 315, 315f; geospatial ECa, 317–324, 320f, 321t; notation, 340–341; remote imagery, 324–328, 328t; sample design based on geospatial ECa data, 319–323; time domain reflectometry (TDR), 311–312; visual crop observation, 316–317 soil-solution extractor, 302–304, 302f soil-solution pH, 70 soil solutions, chemistry of salt-affected, 65–84, 67f; anion exclusion, 74–75, 76t, 77; clay minerals, 72–73, 73t; diffuse double layer, 73–74, 75f; ion exchange, 77–80; notations, 88; rhizosphere chemistry, 80–84; salt and pH effect on chemical mass action, 68–70; surface and ground waters, 70, 71t, 72; surface chemistry, 72–77 soil solutions, measuring EC of, 301–304 soil-water extract, 299–300, 346–348 soil-water systems, 10–11, 12f soil water table, 1075 solar evaporators, 18t, 625, 628, 631, 647f, 651, 741–742, 742f, 743f solid-liquid pathway, 314 solid pathway, 314 solubility product relation, 90 soluble threshold limit concentration, 769–770 solute transport model, 834–841, 836f; mixing and dissolution algorithm, 837; summary, 839–841; transport algorithm for cracking clay soil, 838–839, 840f; transport algorithm for drip-irrigated grape, 873–838 sorghum, 184, 549 source reduction, 17t, 19, 20t, 21 Southern Illinois University, 732 soybeans, 178 spatial analysis, 100 spatial-ensemble distribution, 900 spatially referenced linear regression models, 467–468 spatially variable fields, leaching of, 662 spinach, 176

1091

sprinkler irrigation, 516–517, 547; distribution of applied water, 517; salt accumulation patterns, 517 sprinkling, 663; plant tolerance to saline sprinkling waters, 445–446 standard deviation, 281 statice (Limonium spp.), 422 statistical moments of the variables of interest, 900 steady-state leaching requirement models, 805–807; traditional, 805, 806–807, 806t; WATSUIT, 805, 806t, 807–809, 813f stock (Matthiola incana), 422 stratified random sampling, 278, 284, 289 strawberries, 421 subirrigation, 394 subsurface drip irrigation. See drip irrigation suction cup measurements, 302f, 303 sugar beets, 182 sugarcane, 327 sulfate, 170, 567 sulfide deposits, 106 sulfuric acid, 30; as irrigation water amendment, 790, 796, 797t; as soil amendment, 669–670 Sumerian civilization, 12–14 sunflowers, 182 Superfund sites, 102, 103 supersaturation, 90 surface complexation models, 96 surface flushing, 663 surface hydroxls, 93, 94f surface irrigation, 512 surface mineral scale formation, 728 surface waters, 70, 71t, 72 swelling, 28, 657 SWS model, 858–868; activity coefficients, 870–871; applications of, 875–893; boron, 874–875; calcite precipitation, 871–872; carbon dioxide concentration/production, 867–868; carbon dioxide transport, 868; cation exchange, 873–874; chemical effects on hydraulic conductivity, 860–862; chemical model, 869–873; crop coefficients and ETc calculations, 866–867; crop evapotranspiration calculations, 866; effect of rain on sodium adsorption ratio, 882–885, 883f–886f; gypsum, 872; high-boron irrigation waters, management of, 885–887, 888f–890f; hydraulic

1092 functions, 858–860; magnesium precipitation, 872–873; matric stress scenario, 891; no stress scenario, 890–891; optimizing reclamation using, 878, 880–882, 880f, 881f, 882f; osmotic pressure, 870; plant modeling, 862–867; plant response to salinity/water stress, 890–893; root growth, 865–866; saline sodic soil reclamation, 876, 877f, 878, 878f, 879f; salt stress scenario, 891–892; soil and water chemistry, 868–875; solid phases, 871; transport, 868–869; variable water content and CO2 prediction, 875–876, 876f, 877f; water and salt stress scenario, 892–893; water flow, 858–862; water uptake by plant roots, 862–866 systematic sampling, 279, 284

T tailwater ditches, 609 Tamm’s Reagent, 216 TDS. See total dissolved solids (TDS) temporal process scaling, 900 tension (suction) cups, 302f, 303 TETrans model, 110, 678, 805, 806t, 808–809, 812, 815, 818 Texas, 103 textural change seep, 564, 566 Thellungiella halophila, 194 thenardite, 774 thermodynamic mineral equilibrium, 90 Tiddalik model, 928 tile-drainage systems, 15 tillage, 43 TILLING, 260 tilth, 27–29 time-averaging of boundary conditions, 900, 905, 908 time domain reflectometry (TDR), 311–312 titratable alkalinity, 65 tobacco, transgenic, 235–236 tomatoes, 179, 180, 184, 389, 395, 421, 544–545, 547, 549, 553 total dissolved solids (TDS), 2, 62; extractable, on Red Rock Ranch, 649, 649t; international treaty agreements on, 603; in saline drainage waters, 687, 702, 723t, 725. See also under Panoche Drainage District (California); Rio Grande basin

INDEX Total Maximum Daily Load program, 923 total threshold limit concentration, 769 toxicity, 33; diagnosing, 31; as ion effect, 178–179. See also specific element trace elements, 1, 89–123; adsorption, 92–95; adsorption data description, 95–97; alkali and alkaline earth metals, 89, 101–102; biogeochemical behavior and distribution of, 99–118; categories of, 89–90; in drainage waters, treatment of, 729–733; heavy metals, 90, 115, 116, 122; irrigation water and, 170, 365, 366t–367t, 368; mercury, 117–118; metalloids and nonmetals, 90, 107–113; metals, 113–118; notation, 136–137; oxidation-reduction processes, 97–99; pollution potential in soils, 118–122, 121f; precipitation and dissolution, 90–91; processes affecting, 90–99; solution-phase speciation of, 91–92; transition metals, 89, 102–107 trace elements, deficiencies/toxicities of, 207–228; accumulation in plants, and potential harm to animals, 218–223; assessing trace elements in soils, 223–227, 224f–227f; plant factors, 211–213, 211t, 212t; plant methods, 217–218, 219t, 220t; soil factors, 208–210; soil methods of diagnosing, 213–217, 215t, 217t transgenic approaches to salt tolerance development, 237–239, 240t–247t, 248–258; altered expression of other protein classes, 258; altered expression of regulatory proteins, 256–258; expressing proteins in antioxidant protection, 254–256; modification of intracellular osmolyte levels, 249–254; modification of ion homeostasis mechanisms, 238–239, 248–249 transient leaching requirement models, 804–805, 808–809 transient soil water and reactive transport models, 926–927 transition metals, 102–107; chromium, 102–104; molybdenum, 105–106; vanadium, 106–107 Trinity aquifer (Texas), 103 triple-layer model, 96, 97, 104, 105, 107, 108, 110, 111 trivalent chromium, 104 Tulare Lake Basin, 758, 770, 775, 980f, 981, 1003–1005

INDEX Tulare Lake Drainage District, 729–730, 746 turgor, 183–184 two-dimensional leaching, 661 “two-region-type” models, 827

U Umma, 12–14 underirrigation, 38 undersaturation, 90 United States, extent of salinity problems, 6, 7f. See also specific location University of California–Berkeley ABSR process, 730 University of California–Los Angeles desalination studies, 728–729 UNSATCHEM model, 66, 85, 678, 805, 806t, 809, 816, 816t, 817f, 818, 819, 820, 900–901, 902t, 914, 919, 929–930 Upper Arkansas River irrigation, 601 uranium, 776, 777

V vacuum filtration, 63 vadose zone solute transport, 900 vanadium, 106–107, 777 van der Waals forces, 73, 353 van Genuchten-Mualem-type models, 408, 901 vapor-compression evaporation, 725 variograms, 281, 282 vegetative growth, 180–184 vertical discretization, 900, 905, 942 vertical infiltration model, 826 Viru Valley (Peru), 14 visual crop observation, 316–317 voltage-independent cation channels, 238–239 volume-variance relationships, 282

W waste discharge requirements (WDRs), 16, 19, 769 wastewater management, 610–612; deep well injection, 612; desalting, 611; evaporation ponds, 611; export to ocean or another basin, 610, 611f; in Rio Grande basin, 1047

1093

WATEQ model, 66 water: chemistry, 6, 9f; field sampling, 288; infiltration rates, 171; sodicity of, 170–171 water control structures, 607 water flow model, 827–834, 829f, 830f; summary of steps in, 830–834, 831f–833f waterfowl: boron toxicity in, 109; selenium toxicity in, 15, 113, 222, 617, 634–636, 650, 990, 1006–1009, 1014 waterlogging, 14, 19, 23, 757 Water Quality Act, 602 water resources, controlling salinity of: intercept, isolate, and reuse drainage, 50–52; minimizing leaching/deep percolation, 50 watershed management, 605 WATSUIT model, 807–809, 805, 806t, 810–811, 813f, 818, 819–820, 842, 929; basics, 788; chemical amendments, 790; chemical composition, 790–791, 791t; evolution of, 787–788; irrigation water amendments, 796–798; notation, 800; outputs, 792–793, 792t, 793f, 794t; rootzone water uptake, 788–789; sodium adsorption ratio, 787, 791–792; summary table, 794–796, 795t, 797t; user-controlled variables, 789–790 weathering, 57–59, 373–374, 375f, 406, 657; congruent dissolution, 58; incongruent dissolution, 58–59; redox reactions, 59 Wellton-Mohawk Irrigation and Drainage District, 603–604, 611 Wenner, Frank, 305 Wenner array, 286, 287, 289, 305–306, 310, 317 Westlands Water District (California), 727–728; solar evaporators, 742, 743f Westside Regional Drainage Plan Proposal, 21 wetland habitats, 426, 733–734, 778; irrigation and, 595 wetting rate, 156, 158, 159, 160f, 161 wheat, 179, 184–185, 191–192, 549, 576, 577, 810, 811t, 812, 812t, 847 wheatgrass, 178, 416–417 wildlife: compensation habitats, 634, 635, 651; protection of, 15–16, 19, 749, 757–758, 778, 779–780; in San Joaquin Valley, 1006–1010; trace elements as hazard to, 757, 775, 777–778 woody crops, 417, 418t–419t, 420 world water chemistry, 6, 9f

1094 X x-ray photoelectron spectroscopy, 119

Y yield threshold model, 407–408 yield-threshold soil salinity values, 407

INDEX Z Zea mays, 178 zero discharge drainage management, 730 zero-tension cups, 303 zero-valent iron, 731 zinc (Zn), 30, 41, 115, 116