Effective ecological monitoring [Second ed.] 9781486308927, 1486308929

Table of contents :
Cover
Title
Copyright
Contents
Acknowledgments
Preface to Second Edition
Chapter 1 Introduction
Some of the ecological values and uses of long-term datasets
Time until expression
Informing policies and legislation in environmental management
Use in simulation modelling
Tests of ecological theory
Development of co-located, collaborative and multidisciplinary work
Detection of surprises
Poor record of long-term ecological monitoring
Why we wrote this book
1. Societal need
2. Correcting the record – countering the perception that long-term studies in ecology are poor quality science
3. Making sense of the vast monitoring literature
4. Providing an overview of success and failure
5. New perspectives
References
Chapter 2 Why monitoring fails
Characteristics of ineffective monitoring programs
Failure to ask the preliminary and fundamental question – Is monitoring needed at all?
Passive, mindless and lacking questions
Lack of trigger points for action
Poor experimental design
Snowed by a blizzard of ecological details
Squabbles about what to monitor – 'It's not monitoring without the mayflies'
Assumption that 'one size fits all'
Big machines that go ' bing'
Disengagement
Rush to get 'real work' happening on the ground and accusations of program over- engineering
Poor data management
Breaches of data integrity
Other factors contributing to ineffective monitoring programs
Lack of funding – grant myopia
The loss of a champion
Out of nowhere
Excessive bureaucracy
Summary
References
Chapter 3 What makes long-term monitoring effective?
Characteristics of effective monitoring programs
Good questions and evolving questions
The use of a conceptual model
Selection of appropriate entities to measure
Good design
Well- developed partnerships
Strong and dedicated leadership
Potential to identify key emerging issues
Ongoing funding
Frequent use of data
Scientific productivity
Maintenance of data integrity and calibration of field techniques
Little things matter a lot! Some ' tricks of the trade'
Field transport
Field staff
Access to field sites
Time in the field
The adaptive monitoring framework
Examples of the adaptive monitoring framework
Adaptive monitoring is a general and not a prescriptive framework
Increased future role for adaptive monitoring
Summary
References
Chapter 4 The problematic, the effective and the ugly – some case studies
The problematic
PPBio Australasia
The Alberta Biodiversity Monitoring Program (ABMP)
EMAP
The effective
Rothamsted
Ecosystem Health Monitoring Program (EHMP) for Moreton Bay in South East Queensland, Australia
The Hubbard Brook Ecosystem Study
The Central Highlands of Victoria, south- eastern Australia
Need to wait and see
NEON/ TERN
The ugly
Summary
References
Chapter 5 The upshot – our general conclusions
Changes in culture needed to facilitate monitoring
The academic culture and rewards systems
Structure of organisations
Funding
Societal culture
Good things that can come from non- question based monitoring
The role of citizen science in long-term monitoring
The challenge of intellectual property and data sharing
The challenges in effective monitoring of rare, threatened and endangered species
The major challenge of keeping monitoring and long-term studies going
The big issue of integrating different kinds of monitoring
Approaches to integrate data from different kinds of monitoring
The challenges posed by differences in the kinds of entities that are monitored in different ecosystems
Using environmental and economic accounts as a way to demonstrate the value of monitoring and cement support for monitoring in place
Concluding remarks
References
Index
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
Y
Z

Citation preview

EFFECTIVE ECOLOGICAL MONITORING

We are most pleased to dedicate the Second Edition of this book to Phyllis C Likens, now deceased. Her indefatigable spirit, good humour and exceptional support and encouragement for this book project are enthusiastically acknowledged and greatly missed.

SECOND EDITION

EFFECTIVE ECOLOGICAL MONITORING SECOND E DI TI O N

David B. Lindenmayer and Gene E. Likens

© David Lindenmayer and Gene Likens 2018 All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO Publishing for all permission requests. A catalogue record for this book is available from the National Library of Australia. Published by CSIRO Publishing Locked Bag 10 Clayton South VIC 3169 Australia Telephone: +61 3 9545 8400 Email: [email protected] Website: www.publish.csiro.au Front cover: (clockwise from top left) Dan Florance in the field (photo by Dave Blair); Litoria fallax (photo by Damian Michael); Fungus in the Victorian Central Highlands (photo by Elle Bowd); Bull moose at Isle Royale National Park (photo by Shawn Malone) Back cover: Eastern yellow robins (photo by Dave Blair) Set in 11/13.5 Minion & Helvetica Neue LT Std Edited by Karen Pearce Cover design by James Kelly Typeset by Desktop Concepts Pty Ltd, Melbourne Index by Max McMaster Printed in China by Toppan Leefung Printing Limited CSIRO Publishing publishes and distributes scientific, technical and health science books, magazines and journals from Australia to a worldwide audience and conducts these activities autonomously from the research activities of the Commonwealth Scientific and Industrial Research Organisation (CSIRO). The views expressed in this publication are those of the author(s) and do not necessarily represent those of, and should not be attributed to, the publisher or CSIRO. The copyright owner shall not be liable for technical or other errors or omissions contained herein. The reader/user accepts all risks and responsibility for losses, damages, costs and other consequences resulting directly or indirectly from using this information. Original print edition: The paper this book is printed on is in accordance with the rules of the Forest Stewardship Council ®. The FSC® promotes environmentally responsible, socially beneficial and economically viable management of the world’s forests.

Contents

Acknowledgments ix Preface to Second Edition

Chapter 1

xi

Introduction 1 Some of the ecological values and uses of long-term datasets

4

Time until expression

5

Informing policies and legislation in environmental management

7

Use in simulation modelling

8

Tests of ecological theory

9

Development of co-located, collaborative and multidisciplinary work

12

Detection of surprises

12

Poor record of long-term ecological monitoring

13

Why we wrote this book

15

1. Societal need

15

2. Correcting the record – countering the perception that long-term studies in ecology are poor quality science

15

3. Making sense of the vast monitoring literature

16

4. Providing an overview of success and failure

17

5. New perspectives

17

References 18

Chapter 2

Why monitoring fails

27

Characteristics of ineffective monitoring programs

28

Failure to ask the preliminary and fundamental question – Is monitoring needed at all?

28

Passive, mindless and lacking questions

29

Lack of trigger points for action

30

Poor experimental design

31

Snowed by a blizzard of ecological details

32

Squabbles about what to monitor – ‘It’s not monitoring without the mayflies’ 32 v

vi

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Assumption that ‘one size fits all’

35

Big machines that go ‘bing’

37

Disengagement 40 Rush to get ‘real work’ happening on the ground and accusations of program over-engineering 41 Poor data management

42

Breaches of data integrity

43

Other factors contributing to ineffective monitoring programs

45

Lack of funding – grant myopia

46

The loss of a champion

46

Out of nowhere

47

Excessive bureaucracy

48

Summary 49 References 50

Chapter 3

What makes long-term monitoring effective? Characteristics of effective monitoring programs Good questions and evolving questions

59 59 60

The use of a conceptual model

62

Selection of appropriate entities to measure

63

Good design

65

Well-developed partnerships

67

Strong and dedicated leadership

70

Potential to identify key emerging issues

75

Ongoing funding

76

Frequent use of data

77

Scientific productivity

77

Maintenance of data integrity and calibration of field techniques

78

Little things matter a lot! Some ‘tricks of the trade’

78

Field transport

78

Field staff

80

Access to field sites

80

Time in the field The adaptive monitoring framework

81 81

Examples of the adaptive monitoring framework

83

Adaptive monitoring is a general and not a prescriptive framework

85

Increased future role for adaptive monitoring

86

Summary 86 References 89

Contents

Chapter 4

vii

The problematic, the effective and the ugly – some case studies

99

The problematic

101

PPBio Australasia

101

The Alberta Biodiversity Monitoring Program (ABMP)

104

EMAP 109 The effective

114

Rothamsted 114 Ecosystem Health Monitoring Program (EHMP) for Moreton Bay in South East Queensland, Australia

118

The Hubbard Brook Ecosystem Study

121

The Central Highlands of Victoria, south-eastern Australia

129

Need to wait and see

139

NEON/TERN 139 The ugly

150

Summary 150 References 151

Chapter 5

The upshot – our general conclusions Changes in culture needed to facilitate monitoring

163 165

The academic culture and rewards systems

165

Structure of organisations

171

Funding 172 Societal culture

174

Good things that can come from non-question based monitoring

175

The role of citizen science in long-term monitoring

175

The challenge of intellectual property and data sharing

177

The challenges in effective monitoring of rare, threatened and endangered species

178

The major challenge of keeping monitoring and long-term studies going

180

The big issue of integrating different kinds of monitoring

181

Approaches to integrate data from different kinds of monitoring

185

The challenges posed by differences in the kinds of entities that are monitored in different ecosystems

185

Using environmental and economic accounts as a way to demonstrate the value of monitoring and cement support for monitoring in place

186

Concluding remarks

187

References 188 Index 199

This page intentionally left blank

Acknowledgments

Since the completion of the first edition of this book, DBL would like to acknowledge the support of many organisations who have begun to or continued to support long-term work in south-eastern Australia including the Australian Research Council, the Commonwealth Department of Environment and Energy (and its various previous incarnations), Murray Local Land Services, Riverina Local Land Services, Parks Australia, Victorian Department of Environment, Water and Land Planning (and its many earlier incarnations), Parks Victoria, the now sadly defunct Land and Water Australia, the Ian Potter Foundation, the Thomas Foundation, and the Graeme Wood Foundation. Data from long-term studies have been guided in their collection and then expertly analysed by truly outstanding professional statisticians, especially Ross Cunningham, Jeff Wood, Wade Blanchard and Alan Welsh. The field research officers who expertly gathered high-quality data over many years have included David Blair, Mason Crane, Daniel Florance, Ryan Incoll, Chris MacGregor, Lachlan McBurney, Damian Michael, Rebecca Montague-Drake, Sachiko Okada, Thea O’Loughlin and Matthew Pope. Finally, many of the key insights in this book have been shaped through collaborations with the many outstanding scientists in both the Australian Government’s Threatened Species Recovery Hub and the Australian Long-Term Ecological Research Network (LTERN) that is part of the Terrestrial Ecosystem Research Network (TERN) (especially Emma Burns). DBL would particularly like to thank Karen Viggers, Ryan Lindenmayer and Nina Lindenmayer for their understanding of a husband/father spending too much time parked in front of a computer. GEL initiated the Hubbard Brook Ecosystem Study in 1963 with colleagues, F Herbert Bormann, Robert S Pierce and Noye M Johnson, and has continued to work within this large and complicated ecosystem study since then. Throughout the years, numerous other colleagues, students and technicians have contributed to this study. John S Eaton, Donald C Buso and the late Phyllis C Likens, three long-term support staff, and the current field and data management staff, Tammy ix

x

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Wooster, Brenda Minicucci and Geoff Wilson, contributed in major ways to my efforts and learning within the Hubbard Brook Ecosystem Study. Long-term financial support for these studies was provided by the National Science Foundation, including the Long-Term Research in Environmental Biology and Long-Term Ecological Research programs and The Andrew W Mellon Foundation. Additional support to GEL in the writing of this second edition was provided by the Australian Research Council (via a grant to David Lindenmayer) and The Australian National University, the Cary Institute of Ecosystem Studies and the National Science Foundation through the Hubbard Brook Long-Term Ecological Research and Long-Term Research in Environmental Biology programs. GEL also would like to thank the following people for their contributions to this book: Tammy Wooster, Brenda Minicucci, Donald C Buso, Jonathan Cole, Daniel Schindler, John Pastor, Michele Burford, Matthew Gillespie, Sylvia Lee, Stan Temple, Emily Saeck, Jerry Franklin, Adam Wilson, Amy Schuler, Deborah Fargione, Kathleen Weathers and Jim Nichols. Tabitha Boyer completed a range of key editing and other tasks associated with the preparation of this manuscript. We thank her for these efforts. Boxes for the text were written by Charles Krebs and John Pastor. In addition, Penny Olsen and Stephen Garnett provided an update on heroic efforts to recover the Norfolk Island boobook owl. We would like to thank all of the photographers who contributed images, and Andrew Rankine from Atypica for drawing the figures for this book. John Manger from CSIRO Publishing has been a strong advocate for this book project and made it clear that a second edition of this book was needed.

Preface to Second Edition

It has been more than eight years since we wrote the first edition of this book, on which we received many comments. Many colleagues asked if we would update the book, particularly given the amount of new science associated with long-term research and monitoring that had been published in the past few years. The task of recrafting this book was far from straightforward. A lot has happened in eight years. How to do justice to the rapidly expanding body of new science on monitoring, yet ensure that the book was relatively short but also accessible, to the point, and a reasonable representation of our own philosophies and attitudes towards long-term ecological monitoring? We have added some new sections, including those on the uses of new technology in monitoring, the role of citizen science in monitoring, the challenges created by big data, making decisions when to (and not to) monitor, the special case of threatened species monitoring, and some ways to increase the management and policy relevance of monitoring programs, such as through linking the data gathered from them to initiatives like environmental accounting, ecosystem assessments and mandated environmental reporting. We acknowledge that despite these additions, there will still nevertheless be key topics that we have not addressed or only addressed superficially. Omissions were inevitable in trading off manuscript length and readability. Perhaps we might visit missing topics in yet a third revision of this book, but the reality is that the expansion of the scientific literature is now so rapid and substantial that is quickly becoming beyond the ability of most mortals to write texts about even the most highly constrained subject areas.

The structure of this book This book is about a framework and underlying philosophy for effective ecological monitoring. It is not a manual of field techniques – whether to do point interval counts or transects for monitoring birds or which equipment is needed to track changes in airshed quality. xi

xii

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Our book comprises five short chapters. Readers may wish to go to particular chapters or even read the chapters in random order. The summary and set of references at the end of each chapter are designed to give the reader both a quick overview and an opportunity to pursue particular topics in greater detail. In Chapter 1 we define three broad kinds of monitoring – mindless or passive monitoring, mandated monitoring and question-driven monitoring. We then outline some of the ecological values that can be generated from long-term datasets. We conclude Chapter 1 by discussing the poor record of long-term ecological monitoring programs as one of the primary motivations for writing this book. In Chapter 2 we discuss some of the reasons why many monitoring programs are of poor quality and often fail. We then discuss some of the features of good monitoring programs in Chapter 3. Based on that discussion, we present our adaptive monitoring framework, which incorporates many of the key features and ingredients that should characterise good and effective long-term monitoring programs. In Chapter 4, we present case studies of some good and problematic long-term ecological research and monitoring programs to illustrate the salient points in the previous two chapters. We are acutely aware that the choices for these examples represent our considered opinion and that some readers will be offended by the content of Chapter 4, and we have ventured into this territory with trepidation. Nevertheless, we have written this book in the spirit of constructive criticism in the hope that the problems we identify might be avoided in new monitoring programs or rectified in existing ones. It seems to us that constructive criticism lies at the heart of good science, our emphasis being on ‘constructive’. The final chapter, Chapter 5, not only summarises the main points of our book, but it offers some thoughts about why and how many more long-term monitoring programs might be instigated and how existing ones can be maintained and improved. We also offer some thoughts about how some issues in the future (e.g. the challenges of big data) might be resolved or used to make progress in understanding.

What is new in this version of the book? This updated version of the book has been informed by much that has happened in the past eight years, including several new books on monitoring and long-term ecological studies. We endeavoured to maintain our focus on what we considered to be important attributes of successful and effective monitoring programs. To this end, we have included new sections on such topics as: (1) how major observation networks like the National Earth Observation Network (NEON) and similar major national and international initiatives link with effective ecological monitoring; (2) the role of citizen science in ecological monitoring; (3) when to (and not to)

Preface to Second Edition

xiii

monitor; (4) the special case of, and particular challenges associated with, monitoring of threatened species; and (5) ways in which the place of monitoring might be better cemented into the psyche (and budget lines) of managers and decision makers. We also have tried to update particular case studies as well as modify our opinions on some topics in the light of new information and extensive discussions with colleagues.

Caveats This book is not intended to be an exhaustive treatment of the vast amount of literature on monitoring, but adequate references are given to guide the reader to pertinent sources. The book is relatively short, drawing extensively on our personal experiences and perspectives as to what does and does not work in long-term ecological research and monitoring. Our discussions are based on our respective expertise in long-term population studies and biodiversity monitoring (David B Lindenmayer, DBL) and long-term ecosystem and biogeochemical studies and environmental monitoring (Gene E Likens, GEL). We also have targeted terrestrial and aquatic ecosystems. However, we believe that the general principles we outline are also relevant to marine ecosystems. We have selected case studies to provide a range of long-term research and monitoring programs. However, we readily acknowledge that this book contains a bias to our own work. We also admit to a bias towards the science of long-term monitoring rather than strictly a policy perspective on this topic. This bias arose because we are both scientists and not policy makers, policy implementers or resource managers. Nevertheless, we readily acknowledge (in Chapter 3) that partnerships among people with different roles and sets of skills are one of the fundamental ingredients of successful monitoring programs. We have sought to focus more philosophically and conceptually on monitoring programs, and we do not make specific recommendations or give precise ‘recipes’ about ‘how to do monitoring’. We do not present a detailed review of field methods to guide monitoring programs – there is a truly enormous amount of literature on methods that is readily accessible to those initiating new work or revising existing programs. We believe that a monitoring program generally needs to be tailored to a particular question or site, although we acknowledge there sometimes can be a role for surveillance monitoring. Also, we do not examine the equally extensive literature on statistical and analytical methods and experimental designs for monitoring programs. Statistical challenges will often need to be resolved, often in consultation with a professional statistical scientist. Finally, we aimed to present our opinions without religious-like ‘fundagelical’ fervour. We believe that too much of science is characterised by the attitude that

xiv

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

‘my way is right and yours is wrong’. We certainly recognise that we have made mistakes in our own research and monitoring. As we point out in Chapter 5, there will be occasions where the main ‘rules’ we propose to guide monitoring programs will be breached and yet valuable outcomes are still produced. David B Lindenmayer and Gene E Likens May 2017

1 Introduction

Organisms, including humans, depend upon the integrity of ecosystems for their wellbeing and survival. High-quality ecological information collected over long periods provides valuable insights into changes in ecosystem structure, ecological processes, and the services ecosystems provide. Without this information, we would have no knowledge about the changing status of the life support system of the planet (Muir 2010; Lindenmayer et al. 2015b). Nor will we know how effective management has been for actions like weed control, restoration programs, recovery efforts for endangered species, or abatement of nutrient enrichment. For example, a lack of effective monitoring means that Australia is unable to determine the effectiveness of the millions of dollars it invests in biodiversity management (State of Environment 2017). Similarly, there is no credible monitoring program for bees anywhere in the world, despite the serious global declines in this group and the critical role they play in pollination (Inoye et al. 2017). Indeed, the public has every right to ask what scientists and resource managers have been doing in spending of public money over many decades! For the purposes of this book, we consider such information, collected conscientiously and continuously for at least 10 years, and then analysed, to be long-term monitoring (see also Box 1.1). We see the broad area of monitoring falling into three types: ●●

Curiosity-driven or passive monitoring. This type is monitoring, sometimes also called surveillance monitoring (e.g. see Wintle et al. 2010) is typically devoid of specified questions, with little or no purpose other than curiosity and no

2

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 1.1. What is not long-term monitoring? Although one can easily do runs from a simulation model and make projections for thousands of years into the future, we do not consider simulation modelling per se to be long-term work. Studies that substitute space for time (e.g. snapshot investigations (Diamond 1986) or retrospective investigations (Cambridge et al. 2007)) can provide long-term perspectives but we do not consider them to be long-term work per se. We also do not consider haphazard revisits to a site after a prolonged absence to be programs of long-term research or monitoring (e.g. Currie and Parry 1999; Smith et al. 2007). Contrary to widely held beliefs in some organisations, we also do not consider that simply measuring something in the environment constitutes monitoring. Rather, as we outline in the following chapters, monitoring needs to be driven by questions, an experimental design, a conceptual framework, and data integrity through repeatable application of appropriate field protocols (usually the same ones employed through time) (Lindenmayer and Likens 2010). We acknowledge there are many nuances to discussions about what constitutes ‘long term’ in ecological research and monitoring. The review by Stankey et al. (2003) is instructive. They note that some workers consider long-term studies to be those that continue beyond the generation time of dominant organisms in an ecosystem or sufficiently long to quantify the key processes that structure the ecosystem under investigation. This definition would mean that studies of bacterial assemblages with very rapid generation times would be long-term investigations if they were to persist for a year, a month or even a week. Conversely, a 300-year study of stands of giant sequoia (Sequoiadendron giganteum) in which the dominant trees may live for over 1000 years would not qualify as a long-term study. These considerations are important because they emphasise the variable lifespan of different organisms, but they are not feasible to use for many ecosystem analyses. For the purposes of this book, we use a practical, operational definition and consider long-term as monitoring efforts that continue beyond 10 years.

●●

underlying experimental design. It may be done out of inquisitiveness, but often has limited usefulness in addressing environmental problems or in discovering how the world works (but see Olsen et al. 1993). Passive monitoring has these limitations because: (1) it is not hypothesis-driven, and (2) it lacks management interventions or different experimental treatments that facilitate scientific understanding about ecosystem responses to natural or human disturbance. Curiosity-driven, passive or surveillance monitoring could be purely mindless or it could be based on other motivations (see Chapter 4). Mandated monitoring. This is monitoring for which environmental data must be gathered as a stipulated requirement of government legislation or a political directive, such as monitoring of weather, air quality or river flow funded by governments and conducted by government agencies. For example, populations

1 – Introduction

●●

3

of many threatened species listed under the US Endangered Species Act must be subject to monitoring, making such species unusual among those rare and/or threatened globally that are typically not subject to rigorous monitoring. The challenges associated with monitoring of rare and endangered species are discussed in Chapter 5. Similarly, Mexico has developed a mandated monitoring and reporting framework for assessing land condition and degradation in that country (Garcia-Alaniz et al. 2017). Rigid quality assurance/quality control protocols are usually strictly stipulated in this type of monitoring. Mandated monitoring is driven by the potential to answer broad and often practical questions. These questions almost always are posed post-hoc in regard to some environmental problem and are not derived from a conceptual model of a particular system or process. Mandated monitoring does not attempt to identify or understand the mechanism influencing a change in an ecosystem or an entity. Rather, the focus is usually to identify trends in a given entity (e.g. whether environmental conditions are getting ‘better or worse’). Question-driven monitoring. This type of monitoring is guided by a conceptual model of an ecosystem or some other entity (e.g. a population of organisms) and guided by a rigorous experimental design. The use of a conceptual model will typically result in a priori predictions that are then tested as part of the monitoring program. This kind of monitoring might be undertaken by a single investigator or it might fall under a program like the US Long-Term Research in Environmental Biology (LTREB) program (https://www.nsf.gov/funding/ pgm_summ.jsp?pims_id=13544). In question-driven monitoring, mechanisms can be discovered whereby prospective scenarios of trends can be calculated and modelled (see Chapters 3 and 4). Often such learning is informed by strongly contrasting management interventions (Carpenter et al. 1995) and in statistical parlance such studies might be best termed ‘longitudinal studies with interventions’ (Cunningham and Lindenmayer 2017). This approach can lead to robust predictive capacity and enable an investigator to pose new questions – an advantageous part of the adaptive monitoring paradigm (sensu Lindenmayer and Likens 2009) that we outline in Chapter 3. This predictive capacity can be of immense value for ecologists, resource managers and decision makers (Lindenmayer et al. 2015b). In contrast, in curiosity-driven or passive monitoring, and in mandated monitoring to a lesser extent, the predictive and prospective capability of monitoring can be limited such as simply extending trend lines without supporting data.

Obviously, there can be overlap between broad categories of monitoring. For example, a rigorous statistical framework can characterise both mandated and question-driven monitoring (Cunningham and Lindenmayer 2017), although many long-term studies were never initially designed to be long-term studies (see Box 1.2).

4

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

There can be advantages and disadvantages of mandated monitoring programs and question-driven research and monitoring. As we discuss later in this chapter, mandated monitoring is often coarse scale, leading to assessments of resource condition, but providing limited understanding of ecological mechanisms that have given rise to that condition. Question-driven monitoring is often the converse. It is finer scaled and often process based (see examples across Australian ecosystems in Lindenmayer et al. 2014c), but it is very difficult to make valid spatial extrapolations to larger scales (e.g. at the state, province or national level). Thus, in Chapter 5, we discuss the challenges posed by the integration of data, insights and management recommendations from these broad kinds of monitoring programs.

Some of the ecological values and uses of long-term datasets Countless scientific articles, books, management plans and other documents have been written about the need to do long-term monitoring (Strayer et al. 1986; Likens 1989, 1992; Goldsmith 1991; Spellerberg 1994; Thompson et al. 1998; Lovett et al. 2007; Krebs 2008; Muller et al. 2011; Lindenmayer et al. 2012; Rowland and Vojta 2013; Hamilton 2015; Kupschus et al. 2016; Reynolds et al. 2016). As part of writing this book, a search of the ecological literature published between 1985 and early 2017 produced more than 131 300 articles with the term ‘monitoring’ in the title or abstract. There are even entire journals directly or indirectly focused on long-term research and monitoring programs. Environmental Monitoring and Assessment and Journal of Environmental Monitoring are just two of many examples. In this vast literature, many ecologists and managers of natural resources have readily acknowledged the importance of long-term research, which normally incorporates monitoring, for improved understanding and management of complex systems (Likens 1989; Muller et al. 2011). Indeed, long-term research and monitoring contribute disproportionately to ecological knowledge and policy relative to other kinds of studies (Hughes et al. 2017). Long-term data are valuable for many reasons. For example, they are fundamentally important for (after Lindenmayer et al. 2012): ●●

●●

●●

●●

●●

documenting and providing baselines against which change or extremes can be evaluated (e.g. climate change) evaluating ecological responses to disturbance, such as from experimental manipulations detecting and evaluating changes in ecosystem structure and function, such as change in forest biomass accumulation or change in nutrient limitation of aquatic productivity generating new and important questions about population, community and ecosystem dynamics providing empirical data for testing ecological theory and models

1 – Introduction

Box 1.2. A key caveat: many long-term studies were never initially intended to become long-term studies Much of this book discusses the design, implementation and maintenance of long-term research and monitoring. However, we fully acknowledge that while some studies were designed to be long-term investigations from inception (e.g. RussellSmith et al. 2014), many were not; most of the champions for these studies never envisaged that many years or even decades later they would still be working in the same area or in the same ecosystems. The initial study was established to answer a given question, but other questions arose or more time was needed to quantify temporal trends or responses to interventions. This may demand changing the design of a monitoring program over time – the adaptive monitoring paradigm (Lindenmayer and Likens 2009) that we describe in Chapter 3. Good design is nevertheless a key part of all robust ecological studies (Hayward et al. 2015; Cunningham and Lindenmayer 2017), both short-term (often cross-sectional or snapshot) studies as well as long-term investigations. Indeed, a robust design for a short-term or crosssectional study may mean it has a stronger chance of continuing to be robust should it morph into a long-term investigation. However, we readily acknowledge that short-term investigations may, at the outset, have different design features than one that aims to be long-term from the outset (Cunningham and Lindenmayer 2017).

●● ●● ●●

●●

providing a platform for different kinds of research located in the same area data mining when exploring new questions engaging the community in the practice of science and contributing to data collection informing and evaluating policies and legislation on environmental management.

In the remainder of this section, we illustrate some of these values with a few brief examples.

Time until expression Long-term monitoring is important because many ecological phenomena and ecological processes take a long time to manifest (see Box 1.3), often on much longer temporal scales than typical of most ecological or environmental measurements (Krebs et al. 2001; Smith et al. 2007; Owens 2013). For example, in marine ecosystems, oceanographic cycles occur over decades and even centuries, making long-term monitoring programs critical, although such programs are extremely uncommon in places like deep-sea environments (which also happen to be the largest environments on Earth). Long-term studies can allow separation of short-term fluctuations like intraand interannual variability from longer-term trajectories (Keeling et al. 1995, 1996;

5

6

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 1.3. Time to expression – long-term monitoring at Lake Washington Increasing inputs of nutrients from sewage and other sources contributed to Lake Washington (Seattle, Washington, United States) becoming increasingly eutrophic, especially during the 1940s and 1950s (Edmondson 1991). Noxious, blue-green algae (the cyanobacterium, Oscillatoria rubescens) was first noticed in the lake in 1955. Although Seattle began to divert raw sewage to treatment plants in 1926, increased numbers of people, increased urban development, and forest clearing in the catchment by the 1950s and 1960s had increased nutrient loading to the lake. A local newspaper, the Seattle Post-Intelligencer, even referred to Lake Washington as ‘Lake Stinko’ (Edmondson 1991) because of the increased eutrophication. In response, local municipalities passed a bond issue (‘Metro’) in 1958 to fund the construction of a ‘diverter system’ ringing the lake to direct sewage away from the lake and into nearby Puget Sound. The diversion began in 1964; the phosphorus content promptly decreased and the lake began to recover. Long-term research and monitoring commenced in 1955 by Professor WT (Tommy) Edmondson and colleagues (Fig. 1.1). Their work included retrospective studies, natural history studies and monitoring (Edmondson 1974, 1991; Edmondson and Litt 1982) and was instrumental in not only establishing the basis for this management intervention, but also for chronicling its success. In 1965, Edmondson presented a scientific paper at the International Association of Theoretical and Applied Limnology (SIL) Congress in Warsaw, Poland where he made the prediction that the clarity of the lake in 1971 would be better than 1950 summer values had been. Scientific colleagues were sceptical of such a rapid recovery, and Edmondson and Professor K Wuhrmann (a prominent Swiss scientist) entered into a public wager as to when the lake would recover. Edmondson won this bet, presenting his data in a paper at the SIL Congress in the Soviet Union in 1971 to great applause from the audience (Edmondson 1972).

Fig. 1.1.  Professor Edmondson at Lake Washington, Seattle (Photo courtesy of the Seattle Times)

1 – Introduction

7

A bottle of fine liquor was awarded to Edmondson by Wuhrmann at a subsequent meeting in Kiel, to acknowledge and highlight publicly this successful interaction between long-term research (science), monitoring and management. This is an excellent example of the value of long-term research and monitoring for the public good in protecting and managing natural ecosystems. Long-term studies and monitoring are maintained on Lake Washington (David E Schindler, pers. comm.) which continues to change. For example, in 1976 the lake surprisingly became more transparent than ever before observed, due to the appearance of significant numbers of Daphnia spp., a zooplantonic grazer of small algae, requiring changes (adaptive) in the monitoring regime (i.e. adaptive monitoring; see Chapter 3). More recently, the effects of warming climatic conditions have changed the phenology of thermal stratification, and the associated timing of the spring bloom of diatoms and the zooplankton herbivores that consume them (Winder and Schindler 2004a,b). Many of the climate-driven processes that have been discovered only recently were actually occurring when Edmondson was pursuing his research on eutrophication. It was only after decades of long-term monitoring that the signal of these climateinduced changes emerged from the inherent variation in ecosystem processes.

Franklin and Whelan 2009). This understanding is particularly important where there is substantial temporal variability in ecosystem conditions and populations of organisms. Long-term studies can be important in quantifying the effectiveness of conservation management activities like ecosystem restoration and reconstruction where prolonged periods might be required for systems to recover following major human disturbance. For example, Buckney and Morrison (1992) implemented a spatially and temporally replicated sampling design to evaluate the success of revegetation treatments on mined Australian coastal sand plains. They showed that revegetated areas were on a trajectory towards development of a new ecological community that differed significantly in species composition from pre-mining vegetation and adjacent un-mined vegetation. Whether the restoration trajectories reach a new stable state or converge over time to resemble the un-mined community can be resolved only with continuation of the sampling (Wilkins et al. 2003).

Informing policies and legislation in environmental management Long-term datasets such as those gathered through continuous monitoring can be pivotal to evidence-based environmental legislation (e.g. laws to control air and water pollutants). Legislation regulating the release of pollutants like carbon dioxide or sulphur dioxide is typically implemented without a clear view of the time frame of the response. In the United States, the Clean Air Act of 1970 and Amendments of 1990 mandated a reduction in particulate and sulphur dioxide emissions from industrial combustion of coal and oil to reduce the amount of air

8

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

The 50-year Record of Acid Precipitation at Hubbard Brook 100

Hydrogen Ion (µeq/L)

80

60

40

1970 Clean Air Act

20

1990 Clean Air Act Amendments

0

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

Water-year Fig. 1.2.  Changes in the acidity of precipitation from 1963–1964 to 2013–14 (50 years) at Hubbard Brook Experimental Forest, New Hampshire (modified from Holmes and Likens 2016)

pollution and acid rain falling on eastern North America. Emissions of sulphur dioxide to the atmosphere are a major precursor in the formation of acid rain. No-one had expected that prolonged data collection would be required until the effectiveness of this legislation could be tested. It required 18 years of continuous monitoring of precipitation chemistry in the Hubbard Brook Experimental Forest before ecologists could determine that the acidity of precipitation had actually decreased (Fig. 1.2) (Likens 1989, 1992, 2010) in response to these federally mandated reductions in emissions. This result was badly needed by policy makers dealing with the problem of acid rain, but it took a long time to obtain this ostensibly ‘simple’ result (Likens 1992, 2010). Importantly, the results of long-term monitoring and compliance with the Clean Air Act of 1970 and Amendments of 1990 have had major positive benefits for human health; by 2020 the number of deaths prevented as a result of a reduction of air pollution is estimated to be ~230 000.

Use in simulation modelling Simulation modelling is an important part of ecology and environmental management. Indeed, some things like projection of future changes in climate cannot be done without simulation modelling. However, the best simulation

1 – Introduction

9

modelling will typically be based on an understanding of ecosystem structure and function, ecological process, the biotic assemblage or species targeted for study, and a detailed understanding of interactions, such as biogeochemical processes (Burgman 2005; Holmes and Likens 2016; Honrado et al. 2016). Long-term empirical data are often essential for simulation modelling studies (Lindenmayer et al. 2012). As an example, long-term ecological data were used in extensive population modelling studies of arboreal marsupials in the Central Highlands of Victoria in south-eastern Australia (Lindenmayer and McCarthy 2006; Todd et al. 2016). This work was motivated by a need to estimate the risk of extinction of iconic species such as the nationally endangered Leadbeater’s possum (Gymnobelidues leadbeateri) (Lindenmayer and McCarthy 2006; Lindenmayer et al. 2014b). Data used in the modelling included life history information (Smith 1984), information on long-term habitat dynamics based on known temporal relationships between vegetation structure and animal occurrence (Lindenmayer et al. 1994, 2014a), long-term population trend information (Lindenmayer et al. 2003, 2011), and post-disturbance recovery processes in animal populations and forest composition and structure (Lindenmayer et al. 2013a). The results of modelling that was built upon these various kinds of data guided the development of midscale protected areas for the Leadbeater’s possum within forests broadly designated for timber production (Lindenmayer and Possingham 1995) and more recently large-scale (5000+ ha) protected areas (Todd et al. 2016; Taylor et al. 2017). Longterm data underpinned not only an improvement to the reserve system, but also the modelling per se, motivating a cycle of ongoing field data collection, testing of simulation models, and continuous upgrading of simulation models (Lindenmayer and McCarthy 2006; Todd et al. 2016).

Tests of ecological theory Empirical field data provide the only true test of ecological theory (ShraderFrechette and McCoy 1993; Franklin and MacMahon 2000; Driscoll and Lindenmayer 2012). This assertion has been demonstrated in virtually all ecosystems that have been studied worldwide. A classic example is the Kluane Boreal Forest Monitoring Project in northern Canada (Krebs et al. 2017) (see Box 1.4). Another outstanding example is the long-term and large-scale study conducted in the tropical moist forests of Panama since 1923, which provided data to test models in community ecology about the maintenance of diversity (Condit 1995). Research expanded to plant physiology, canopy biology and animal ecology, and there is now a network of over 40 large, permanent plots worldwide (http:// www.ctfs.si.edu/group/About/). The complete enumeration and spatial mapping of all trees in 50-ha plots stimulated the development of the unified neutral theory of biodiversity and biogeography that seeks to explain the relative abundance of species in ecological communities (Hubbell 2001, 2006).

10

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 1.4. Kluane Boreal Forest Monitoring Project Charles Krebs University of British Columbia Vancouver, Canada http://www.zoology.ubc.ca/~krebs/kluane/html In 1973, my students and I began studies in the boreal forest of the south-western Yukon, based out of the Arctic Institute’s Kluane Lake Research Station (Fig. 1.3), one of the few research stations in northern Canada. The original studies were centred on the small rodents, but it quickly became clear that the dominant herbivore in the ecosystem was the snowshoe hare. The research program was refocused on an attempt to understand the ecological mechanisms driving the nine to 10-year cycle of snowshoe hares, a cycle that occurs across the boreal zone of Alaska and Canada (Krebs et al. 2017). Tony Sinclair and Jamie Smith joined the Kluane Project in 1976 and we set up the first of a series of large-scale experiments to test hypotheses about the factors driving the hare cycle. From 1976 to 1984 we tested the food hypothesis by providing extra food to hares on four areas. The results were disappointing because adding food made no impact on the cyclic decline. By the early 1980s, we had become convinced that predators were a major factor in this ecosystem, and we decided to expand the study to encompass the entire terrestrial food web from plant dynamics to predator changes. The research team expanded to eight faculty members from four Canadian universities, and from 1986 to 1996 we conducted a larger set of experiments manipulating the ecosystem bottom up by adding fertiliser and hare food and top down by predator exclusion

Fig. 1.3.  The Kluane study in Yukon, northern Canada (Photo by Charles Krebs)

1 – Introduction

11

with electric fences. The results have been summarised in a book (Krebs et al. 2001) and in over 265 publications. By 1996 when the major funding ceased, we had to decide whether or not to stop the study. We elected to continue monitoring key elements in the food web; fungal and plant production, hare, squirrel, and the abundance of rodents and major predators. By 2016, we had gathered 44 years of monitoring data on rodents and hares in this ecosystem, 31 years of predator monitoring, and 19 to 23 years of monitoring the less-studied components (fungi, ground berries). One result of this monitoring has led to unexpected insights since climate change became a major focus of research interest in northern Canada. For example, we have developed a predictive model of how regional climate variables affect white spruce cone crops (Wexler et al. 2014). Snowshoe hare cycles have shown a disturbing trend to lower peak population densities sequentially in the five cycles that we have monitored since 1973, a trend we are following with great interest. The amplitude of the hare cycle to the present can be predicted by the density of predators like lynx during the previous low phase of the cycle. Red squirrels have adjusted their breeding season earlier in the spring as a response to climate change (Réale et al. 2003). The monitoring has led to a range of hypotheses about ecosystem function in the plants and animals of the Yukon boreal forest. While the major finding has been that for the terrestrial mammals this is largely a top-down system driven by the predators, rather than a bottom-up system driven by soil nutrients and plant production, as with most ecological systems, the devil is in the details of how the ecosystem components interact. The major question currently is to anticipate how this ecosystem will respond to climate change, and this question can now be addressed with a background set of quantitative data on the dominant species in this ecosystem. This study could not have been done without a dedicated group of ecologists interested in understanding community dynamics in this ecosystem, an enthusiastic group of graduate students and technicians, and continued funding from the Natural Sciences and Engineering Research Council of Canada and two private foundations.

Vitousek and Reiners (1975) and Gorham et al. (1979) proposed a theory to explain the biogeochemical response of nitrogen to forest disturbance. They hypothesised that disturbed forest ecosystems will be ‘leaky’ (i.e. have limited retention) for nitrogen with large losses of nitrate in stream water. However, such systems would strongly retain nitrogen during aggradation stages of forest maturation. They also postulated that mature forests would be less retentive of nitrogen and become leaky again (e.g. increased nitrate lost in stream water). However, long-term monitoring, combined with catchment-scale experimental manipulation of the Hubbard Brook Experimental Forest showed that the current, ‘mature’ forest ecosystem is strongly retentive of nitrogen because nitrate concentrations in stream water are at the lowest concentration during the 53-year

12

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

record (Likens 2013; Holmes and Likens 2016). Either the theory is inadequate or something else is happening, and this topic is currently the subject of intense monitoring (Bernal et al. 2012; Wexler et al. 2014). Long-term studies are also critical for informing ecological genetic theory and models aimed at understanding the ability of species to evolve and/or adapt to environmental changes (Hoffmann and Sgro 2011). Much of our understanding of evolutionary potential comes from long-term studies like those involving the Galapagos finches (Grant and Grant 2002) and Soay sheep (Ovis aries) (Ozgul et al. 2009). These datasets have been essential for testing models of adaptation and for developing an understanding of how trait interactions and patterns of natural selection operate and determine the long-term viability of populations and species.

Development of co-located, collaborative and multidisciplinary work Long-term studies can often be the stimulus for other investigations either at the same site or at other sites. Such kinds of co-located long-term ecological studies can become particularly powerful when a multidisciplinary approach is taken (Johnston et al. 2011). An example is the fascinating discovery of birds comprised of four sexes at Cranberry Lake in upstate New York. This area has been the site of more than 25 years of focused research on the white-throated sparrow (Zonotrichia albicollis) and it has entailed answering a series of questions around differences in the colour morphs of birds, breeding and nesting behavioural studies and genetic analyses (Tuttle et al. 2016). The breadth of co-located long-term work at Cranberry Lake has produced many extraordinary insights, not only that birds of the same colour morph do not interbreed, but also that the different morphs are characterised by markedly different patterns of breeding and other behaviour and that key phonological and behavioural differences can be traced to genetic mutations and ‘super-genes’ (Tuttle et al. 2016) There are numerous other examples from around the world that could not have been derived without co-located, multifaceted and long-term research. The longterm studies of mammal population cycles in the boreal forests of Canada is one of many such cases (see Box 1.4). An important aspect of co-located investigations is the instigation of mechanistic studies that account for the long-term patterns that have been identified (e.g. Persson et al. 2009; Tuttle et al. 2016).

Detection of surprises Long-term monitoring can reveal significant ecological surprises of environmental and conservation consequence (Likens 2007; Contamin and Ellison 2009; Lindenmayer et al. 2010; Dodds et al. 2012; Majer et al. 2013). An example is the population monitoring of the giant panda (Ailuropoda melanoleuca), a Chinese and global conservation icon. Populations of the species had been carefully monitored in Wanglang reserve for many years and numbers were estimated to be just 27 individuals in 1998 (Zhan et al. 2006). However, in the mid-2000s, new

1 – Introduction

13

Box 1.5. Why does long-term research often lead to ecological surprises? Ecological surprises can be broadly defined as highly unexpected findings about the natural environment. They are critically important in ecology because they are catalysts for questioning and reformulating views of the natural world. Ecological surprises are most likely to be uncovered through long-term research and monitoring (Lindenmayer et al. 2010; Majer et al. 2013). There are several reasons why long-term studies often lead to ecological surprises. First, time is a major driver of change (e.g. time following major natural disturbances) and some key phenomena do not become apparent unless targeted entities like populations, ecosystems and/or ecological processes are studied for a prolonged period. Second, spending time in an ecosystem targeted for study leads to key new observations that are not seen by desk-bound scientists and policy makers. Third, increasing research experience with a target ecosystem, coupled with an understanding of the natural history of that ecosystem, increases the chance that appropriate ecological questions can be posed and important new insights subsequently found, including ones that are unexpected. Fourth, over time, many research programs expand to encompass different strands of research, including adding additional groups of taxa or exploring the impacts of additional ecological processes. Ecological surprises are often found by interrelating the results of sets of parallel and/or concurrent studies. This increased value occurs because it becomes easier to quantify interactions between and among multiple factors including the unintended consequences of management actions in already modified ecosystems, which are a common cause of ecological surprises (Lindenmayer et al. 2010; Majer et al. 2013).

survey methods were employed using molecular markers and, to the great surprise of biologists, the population of the giant panda was estimated to be ~66 individuals or at least twice that thought in 1998 (Zhan et al. 2006). More recent studies have suggested that the total population of the giant panda China-wide (excluding cubs less than 1.5 years old) appears to be approaching 2000 individuals and a case has been made for downlisting the animal from its current endangered status (Swaisgood et al. 2017). We further discuss the monitoring program for the giant panda in Chapter 2, particularly in regard to the use of standardised field protocols and relationships to the long-term maintenance of data integrity. There are several reasons why long-term research and monitoring can lead to ecological surprises and we briefly outline these in Box 1.5.

Poor record of long-term ecological monitoring Some in the scientific community have traditionally viewed monitoring as a management activity unrelated to scientific research (e.g. Hellawell 1991).

14

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 1.6. Not all ecological research needs to be long-term and not everything needs to be monitored Not all ecological studies need to be long-term. As an illustration, many kinds of studies in behavioural ecology can be completed relatively rapidly and will not necessarily benefit from repeated measurements over a prolonged period. For example, studies of the calling behaviour of birds will often be very informative when conducted as a snapshot investigation at particular times of the day when all members of the bird assemblage are vocalising (Catchpole and Slater 1995). Similarly, not everything needs to be monitored, and sometimes monitoring will distract from (or divert resources from) other key actions or management interventions without tangible return on investment or demonstrable capacity for learning (McDonald-Madden et al. 2010).

Moreover, monitoring has sometimes been considered a routine activity where data collected in this way do not merit significant consideration by managers (‘mindless monitoring’, i.e. an activity that is neither ‘science’ nor contributes to achieving goals, or have an expected outcome) (Yoccoz et al. 2001; Nichols and Williams 2006). In contrast, we strongly consider that well-conceived and well-executed monitoring is a fundamentally important component of scientific research, although not all research needs to be long term (see Box 1.6). Good scientists lead with good questions and they develop good questions by: (1) using critical thinking; (2) building robust conceptual models of how ecosystems work to guide research; (3) testing ‘true’ policy questions (Walters 1986); (4) promoting open dialogue between scientists and managers (Likens 1989; Lawton 2007; Likens et al. 2009; Lindenmayer et al. 2013b); and (5) critically evaluating experimental manipulation, both designed and opportunistic (Cunningham and Lindenmayer 2017; Rosi-Marshall et al. 2016). Although there have been some highly successful long-term monitoring programs (e.g. Lund 1978; Goldman 1981; Lawes Agricultural Trust 1984; Lindenmayer et al. 2014c; Holmes and Likens 2016; Timofeyev 2016), there is a prolonged history of poorly planned and unfocused monitoring programs that are either ineffective or fail completely (see Orians 1986; Krebs 1991; Allen 1993; Norton 1996; Legg and Nagy 2006; Sachs et al. 2010; Brewitt and Holl 2014). In this book, we discuss some of the deficiencies in monitoring programs. Then, based upon our collective experience spanning 80 years in establishing a range of natural resource and environmental monitoring programs, we discuss ways to resolve some of the problems underlying poorly planned and unfocused monitoring programs. As part of our discussion, we outline an adaptive monitoring framework. This framework is driven by tractable questions, a rigorous statistical design at the outset of the study, a human need to know about ecosystem change, and a plan to

1 – Introduction

15

monitor a small number of things well rather than many entities badly. The framework allows a monitoring program to evolve and adapt iteratively as new information is obtained or new questions emerge, while maintaining the integrity of the core objectives of the initial program (Lindenmayer and Likens 2009). We then provide some examples of adaptive monitoring.

Why we wrote this book Why did we feel compelled to write yet something else on monitoring and add to an already enormous scientific and management literature? There were five key reasons for our decision.

1. Societal need The first reason we wrote this book was to foster a greater interest in, and an improvement of, ecological monitoring. It is now increasingly critical to do highquality, question-driven, statistically designed monitoring, given the rapid increase in effects of climate change and the need to reverse the current environmental crisis (see Box 1.7). It could be argued (and we do) that given the increasing, widespread environmental degradation throughout the world, there never has been a more important time to establish good and effective monitoring networks.

2. Correcting the record – countering the perception that long-term studies in ecology are poor quality science The second reason we wrote this book is that in the broader scientific and policy communities some highly ill-informed individuals consider long-term ecological research and monitoring to be second-rate and overpriced science, and believe the

Box 1.7. Societal need for long-term monitoring There are numerous examples of the critical value of long-term monitoring. One of these comes from South Africa where water is scarce and demand will soon outstrip supply. Work from Europe in the early 1900s suggested that water yields are greater in areas with tree cover. This, in part, triggered the establishment of large areas of exotic plantations of trees, including in areas of native grassland. Long-term monitoring in the Drakensburg Mountains, a key water production area in South Africa, actually showed that plantation establishment significantly reduced water yields (Bosch and Hewlett 1982). These results could not have been obtained without long-term monitoring (South African Environmental Observation Network 2015) and they have major implications for water, forest and land clearing policies in South Africa – a country under increasing environmental stress as a result of rapid increases in human population, climate change and many other factors.

16

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

evidence it accumulates does not matter (see Birkhead 2014). We repudiate such myopic perspectives and, where possible throughout this volume, outline the many critical insights that can be gained only through long-term work. A related reason is that budgets for environmental management, including ecological monitoring, are dwindling rapidly. For example, between 2013 and 2019, funding from the Australian Government for the environment will decline by almost 40% even though overall government spending will have increased by more than 20%. By the end of this decade, spending on the environment will be roughly three cents of every A$100 expended. Hence, it is critical that the quality and effectiveness of ecological monitoring be far higher than it has ever been.

3. Making sense of the vast monitoring literature There is an immense and rapidly increasing body of literature on monitoring and monitoring-related topics now spanning well over 100 000 peer-reviewed scientific articles and numerous books. Much of it can be very confusing for scientists, policy makers and resource managers alike. Although this book is short, we invested a great effort in reading a very large amount of the literature on ecological monitoring and in doing so, attempted to alert readers to some distinct biases that often may not be particularly useful for many people proposing to initiate new monitoring programs. First, much has been written about specific methods of monitoring a particular entity, but these will often be relevant only to that entity (e.g. a species or group of species; point interval counts for birds Pyke and Recher 1983; Sutherland 1996) or to a given place; they will not be readily transferable to other entities or landscapes. Phrases like ‘this should be useful for monitoring’ or ‘X should be monitored’ appear repeatedly in the literature but without much discussion of why and in what circumstances. In many cases, the response to the issue of what to monitor is relatively straight-forward – it will depend on what questions need to be answered (Lindenmayer and Likens 2011a). A second bias in the monitoring literature is on generic lists of entities that are advocated as ‘mandatory’ to monitor and generic frameworks to guide the measurement of these entities. These lists and frameworks are often illustrated with a case study from a particular ecosystem, but the transferability of these approaches to other ecosystems can be problematic (Lindenmayer et al. 2015c). Third, much of the monitoring literature focuses on statistical methods and experimental design. While this information is valuable, much of it may not be relevant to address specific questions or problems that a particular monitoring program aims to address. Fourth, much of the monitoring literature has discussed ‘indicators’ and contains claims that entity X or species Y is an indicator. Often these claims are

1 – Introduction

17

unsubstantiated (Caro 2010; Westgate et al. 2014; Lindenmayer et al. 2015a). Where they are substantiated, the generality of the indicator function can be limited, either spatially, taxonomically or both, and this approach may not be helpful for those proposing to establish new monitoring programs (Lindenmayer et al. 2015c). Fifth, there has been an increase in recommendations to establish or further expand what are called Earth Observation Networks, but these often lack key questions or well-developed conceptual frameworks, and the vast amounts of data they can generate (often in real time) has the potential for science to be done backwards and hence become inefficient, ineffective or fail completely (Lindenmayer and Likens 2013) (see Chapter 4). Sixth is the push to do national-level or even global monitoring without careful consideration of the problems associated with integrating data from different studies that have been gathered in different ways for different purposes (e.g. see Sachs et al. (2010) with a corresponding critique by Lindenmayer and Likens (2011b)).

4. Providing an overview of success and failure Another reason why we wrote this book was that the numbers of monitoring studies worldwide is now sufficient to allow us to identify what we believe makes an effective monitoring program. It is important to identify factors that contribute to success and failure so that the public perspectives of ecological monitoring are improved and bureaucrats do not lose patience with the scientific community and abandon funding investments in monitoring programs.

5. New perspectives We have identified some important perspectives on monitoring programs that, to the best of our knowledge, have not previously been canvassed by others. One of these is the adaptive monitoring paradigm and the framework that accompanies it (Lindenmayer and Likens 2009). Similarly, we have not found other treatments of the advantages and disadvantages of mandated monitoring programs and question-driven research and monitoring. Finally, we believe we have identified some confusion of scale and function among monitoring programs. As we outlined earlier (and discuss further in Chapter 5), mandated monitoring is often coarsescale and focused on resource condition whereas question-driven monitoring is often sites-based and/or finer scaled, often with a focus on ecological processes. We believe these differences are often overlooked in the literature in an effort to develop an all-encompassing, one-size-fits-all framework. This generality can lead to arguments at cross-purposes, particularly about the relative merits of different approaches, frameworks and methods. A major and often unresolved issue is how to combine data and approaches from different kinds of monitoring to guide critical management decisions.

18

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

References Allen RB (1993) An appraisal of monitoring studies in South Island tussock grasslands, New Zealand. New Zealand Journal of Ecology 17, 61–63. Bernal S, Hedin LO, Likens GE, Gerber S, Buso DC (2012) Complex response of the forest nitrogen cycle to climate change. Proceedings of the National Academy of Sciences of the United States of America 109, 3406–3411. doi:10.1073/ pnas.1121448109 Birkhead T (2014) Stormy outlook for long-term ecology studies. Nature 514, 405. doi:10.1038/514405a Bosch JM, Hewlett JD (1982) A review of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55, 3–23. doi:10.1016/0022-1694(82)90117-2 Brewitt P, Holl KD (2014) Better monitoring of fish in dam projects. Nature 513, 33. doi:10.1038/513033d Buckney RT, Morrison DA (1992) Temporal trends in plant species composition on mined sand dunes in Myall Lakes National Park, Australia. Australian Journal of Ecology 17, 241–254. doi:10.1111/j.1442-9993.1992.tb00806.x Burgman MA (2005) Risks and Decisions for Conservation and Environmental Management. Cambridge University Press, New York. Cambridge ML, How JR, Lavery PS, Vanderklift MA (2007) Retrospective analysis of epiphyte assemblages in relation to seagrass loss in a eutrophic coastal embayment. Marine Ecology Progress Series 346, 97–107. doi:10.3354/meps06993 Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington, DC. Carpenter SR, Chisholm SW, Krebs CJ, Schindler DW, Wright RF (1995) Ecosystem experiments. Science 269, 324–327. doi:10.1126/science.269.5222.324 Catchpole CK, Slater PJ (1995) Bird Song: Biological Themes and Variations. Cambridge University Press, Cambridge, England. Condit R (1995) Research in large, long-term tropical forest plots. Trends in Ecology & Evolution 10, 18–22. doi:10.1016/S0169-5347(00)88955-7 Contamin R, Ellison AM (2009) Indicators of regime shifts in ecological systems: what do we need to know and when do we need to know it? Ecological Applications 19, 799–816. doi:10.1890/08-0109.1 Cunningham R, Lindenmayer DB (2017) Approaches to landscape scale inference and design issues. Current Landscape Ecology Reports 2, 42–50. Currie DR, Parry GD (1999) Changes to benthic communities over 20 years in Port Phillip Bay, Victoria. Marine Pollution Bulletin 38, 36–43. Diamond J (1986) Overview: laboratory experiments, field experiments and natural experiments. In Community Ecology. (Eds J Diamond, TJ Case) pp. 3–22. Harper and Row, New York. Dodds WK, Robinson CT, Gaiser EE, Hansen GJ, Powell H, Smith JM, Morse NB, Johnson SL, Gregory SV, Bell T, Kratz TK, McDowell WH (2012) Surprises and

1 – Introduction

19

insights from long-term aquatic data sets and experiments. Bioscience 62, 709–721. doi:10.1525/bio.2012.62.8.4 Driscoll D, Lindenmayer DB (2012) Framework to improve the application of theory in ecology and conservation. Ecological Monographs 82, 129–147. doi:10.1890/11-0916.1 Edmondson WT (1972) The present condition of Lake Washington. Internationalen Vereinigung für Theoretische und Angewandte Limnologie. 18, 284–291. Edmondson WT (1974) Secondary production. Mitteilungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie 20, 229–274. Edmondson WT (1991) The Uses of Ecology: Lake Washington and Beyond. University of Washington Press, Seattle. Edmondson WT, Litt AH (1982) Daphnia in Lake Washington. Limnology and Oceanography 27, 272–293. doi:10.4319/lo.1982.27.2.0272 Franklin JF, MacMahon JA (2000) Messages from a mountain. Science 288, 1183–1184. doi:10.1126/science.288.5469.1183 Franklin DC, Whelan PI (2009) Tropical mosquito assemblages demonstrate “textbook” annual cycles. PLoS One 4, e8296. doi:10.1371/journal.pone.0008296 Garcia-Alaniz N, Equihua M, Pérez O, Benítez J, Maeda P, Urrutia F, Martínez JJ, Villela Gaytán S, Schmidt M (2017) The Mexican National Biodiversity and Ecosystem Degradation Monitoring System. Current Opinion in Environmental Sustainability 26–27, 62–68. doi:10.1016/j.cosust.2017.01.001 Goldman C (1981) Lake Tahoe: two decades of change in a nitrogen deficient oligotrophic lake. Verhandlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie 21, 45–70. Goldsmith B (1991) Monitoring for Conservation and Ecology. Chapman and Hall, London. Gorham EP, Vitousek M, Reiners WA (1979) The regulation of chemical budgets over the course of terrestrial ecosystem succession. Annual Review of Ecology and Systematics 10, 53–84. doi:10.1146/annurev.es.10.110179.000413 Grant PR, Grant BR (2002) Unpredictable evolution in a 30-year study of Darwin’s finches. Science 296, 707–711. doi:10.1126/science.1070315 Hamilton S (2015) The Value of Water Monitoring. Aquatic Informatics: Vancouver, British Columbia. Hayward MW, Boitani L, Burrows ND, Funston PJ, Karanth KU, MacKenzie DI, Pollock KH, Yarnell RW (2015) Ecologists need robust survey designs, sampling and analytical methods. Journal of Applied Ecology 52, 286–290. doi:10.1111/1365-2664.12408 Hellawell JM (1991) Development of a rationale for monitoring. In Monitoring for Conservation and Ecology. (Ed. FB Goldsmith) pp. 1–14. Chapman and Hall, London. Hoffmann AA, Sgro CM (2011) Climate change and evolutionary adaptation. Nature 470, 479–485. doi:10.1038/nature09670

20

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Holmes RT, Likens GE (2016) Hubbard Brook: The Story of a Forest Ecosystem. Yale University Press, New Haven, Connecticut. Honrado JP, Pereira HM, Guisan A (2016) Fostering integration between biodiversity monitoring and modelling. Journal of Applied Ecology 53, 1299–1304. doi:10.1111/1365-2664.12777 Hubbell SP (2001) The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, Princeton. Hubbell SP (2006) Neutral theory in ecology and the evolution of ecological equivalence. Ecology 87, 1387–1398. doi:10.1890/0012-9658(2006)87[1387:NTATEO] 2.0.CO;2 Hughes BB, Beas-Luna R, Barner AK, Brewitt K, Brumbaugh DR, Cerny-Chipman EB, Close SL, Coblentz KE, de Nesnear KL, Drobnitch ST, Figurski JD, Focht B, Friedman M, Freiwald J, Heady KK, Heady WN, Hettinger A, Johnson A, Karr KA, Mahony B, Moritsch M, Osterback AK, Reimer J, Obinson J, Rohrer T, Rose JM, Sabal M, Segui LM, Shen C, Sullivan J, Zuercher R, Raimondi PT, Menge BA, Grorud-Colvert K, Novakk M, Carr MH (2017) Long-term studies contribute disproportionately to ecology and policy. Bioscience 67, 271–281. doi:10.1093/biosci/ biw185 Inoye D, Droege S, Mawdsley J (2017) Word alone will not protect pollinators. Science 355, 357. Johnston FH, Hanigan IC, Henderson SB, Morgan GG, Bowman DMJS (2011) Extreme air pollution events from bushfires and dust storms and their association with mortality in Sydney, Australia. Environmental Research 111, 811–816. doi:10.1016/j. envres.2011.05.007 Keeling CD, Whorf TP, Wahlen M, van der Plicht J (1995) Interannual extremes in the rate of rise of atmospheric carbon dioxide since 1980. Nature 375, 666–670. doi:10.1038/375666a0 Keeling CD, Chin JF, Whorf TP (1996) Increased activity of northern vegetation inferred from atmospheric CO2 measurements. Nature 382, 146–149. doi:10.1038/382146a0 Krebs CJ (1991) The experimental paradigm and long-term population studies. The Ibis 133, 2–8. Krebs C (2008) The Ecological World View. CSIRO Publishing, Melbourne. Krebs CJ, Boutin S, Boonstra R (Eds) (2001) Ecosystem Dynamics of the Boreal Forest: The Kluane Project. Oxford University Press, New York. Krebs CJ, Boonstra R, Boutin S (2017) Using experimentation to understand the 10-year snowshoe hare cycle in the boreal forest of North America. Journal of Animal Ecology doi:10.1111/1365-2656.12720 Kupschus S, Schratzberger M, Righton D (2016) Practical implementation of ecosystem monitoring for the ecosystem approach to management. Journal of Applied Ecology 53, 1236–1247. doi:10.1111/1365-2664.12648 Lawes Agricultural Trust (1984) Rothamsted: The Classical Experiments. Rothamsted Agricultural Experiment Station, Rothamsted, England.

1 – Introduction

21

Lawton JH (2007) Ecology, politics and policy. Journal of Applied Ecology 44, 465–474. doi:10.1111/j.1365-2664.2007.01315.x Legg CJ, Nagy L (2006) Why most conservation monitoring is, but need not be, a waste of time. Journal of Environmental Management 78, 194–199. doi:10.1016/j. jenvman.2005.04.016 Likens GE (Ed.) (1989) Long-term Studies in Ecology: Approaches and Alternatives. Springer-Verlag, New York. Likens GE (1992) The Ecosystem Approach: Its Use and Abuse. Ecology Institute, Oldendorf/Luhe, Germany. Likens GE (2007) Surprises from Long-term Studies at the Hubbard Brook Experimental Forest, USA. Asahi Glass Foundation, Tokyo, Japan. Likens GE (2010) The role of science in decision-making: does evidence-based science drive environmental policy? Frontiers in Ecology and the Environment 8, e1–e9. doi:10.1890/090132 Likens GE (2013) Biogeochemistry of a Forested Ecosystem. Springer-Verlag, New York. Likens GE, Walker KF, Davies PE, Brookes J, Olley J, Young WJ, Thoms MC, Lake PS, Gawne B, Davis J, Arthington AH, Thompson R, Oliver RL (2009) Ecosystem science: toward a new paradigm for managing Australia’s inland aquatic ecosystems. Marine and Freshwater Research 60, 271–279. doi:10.1071/MF08188 Lindenmayer DB, Likens GE (2009) Adaptive monitoring – a new paradigm for long-term research and monitoring. Trends in Ecology & Evolution 24, 482–486. doi:10.1016/j.tree.2009.03.005 Lindenmayer DB, Likens GE (2010) Improving ecological monitoring. Trends in Ecology & Evolution 25, 200–201. doi:10.1016/j.tree.2009.11.006 Lindenmayer DB, Likens GE (2011a) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/s10021-010-9394-6 Lindenmayer DB, Likens GE (2011b) Effective monitoring of agriculture. Journal of Environmental Monitoring 13, 1559–1563. doi:10.1039/c0em00691b Lindenmayer DB, Likens GE (2013) Don’t do big-data science backwards. Nature 499, 284. doi:10.1038/499284d Lindenmayer DB, McCarthy MA (2006) Evaluation of PVA models of arboreal marsupials. Biodiversity and Conservation 15, 4079–4096. doi:10.1007/ s10531-005-3367-7 Lindenmayer DB, Possingham HP (1995) The Risk of Extinction: Ranking Management Options for Leadbeater’s Possum. Centre for Resource and Environmental Studies, Canberra. Lindenmayer DB, Cunningham RB, Donnelly CF (1994) The conservation of arboreal marsupials in the montane ash forests of the Central Highlands of Victoria, south-eastern Australia. VI. The performance of statistical-models of the nest tree and habitat requirements of arboreal marsupials applied to new survey data. Biological Conservation 70, 143–147. doi:10.1016/0006-3207(94)90282-8

22

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Lindenmayer DB, Cunningham RB, MacGregor C, Incoll RD (2003) A long-term monitoring study of the population dynamics of arboreal marsupials in the Central Highlands of Victoria. Biological Conservation 110, 161–167. doi:10.1016/ S0006-3207(02)00171-4 Lindenmayer DB, Likens GE, Krebs CJ, Hobbs RJ (2010) Improved probability of detection of ecological “surprises”. Proceedings of the National Academy of Sciences of the United States of America 107, 21957–21962. doi:10.1073/pnas.1015696107 Lindenmayer DB, Wood J, McBurney L, Michael D, Crane M, MacGregor C, Montague-Drake R, Banks S, Gibbons P (2011) Cross-sectional versus longitudinal research: a case study of trees with hollows and marsupials in Australian forests. Ecological Monographs 81, 557–580. doi:10.1890/11-0279.1 Lindenmayer DB, Likens GE, Andersen A, Bowman D, Bull CM, Burns E, Dickman CR, Hoffmann AA, Keith DA, Liddell MJ, Lowe AJ, Metcalfe DJ, Phinn SR, RussellSmith J, Thurgate N, Wardle GM (2012) Value of long-term ecological studies. Austral Ecology 37, 745–757. doi:10.1111/j.1442-9993.2011.02351.x Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks S, Driscoll D, Smith A, Gill AM (2013a) Fire severity and landscape context effects on arboreal marsupials. Biological Conservation 167, 137–148. doi:10.1016/j.biocon.2013.07.028 Lindenmayer DB, MacGregor C, Dexter N, Fortescue M, Cochrane P (2013b) Booderee National Park management: connecting science and management. Ecological Management & Restoration 14, 2–10. doi:10.1111/emr.12027 Lindenmayer DB, Barton PS, Lane PW, Westgate MJ, McBurney L, Blair D, Gibbons P, Likens GE (2014a) An empirical assessment and comparison of species-based and habitat-based surrogates: a case study of forest vertebrates and large old trees. PLoS One 9, e89807. doi:10.1371/journal.pone.0089807 Lindenmayer DB, Blair D, McBurney L, Banks S (2014b) Preventing the extinction of a globally endangered species – Leadbeater’s Possum (Gymnobelideus leadbeateri). Journal of Biodiversity & Endangered Species 2, 140. doi:10.4172/2332-2543. 1000140. Lindenmayer DB, Burns E, Thurgate N, Lowe A (Eds) (2014c) Biodiversity and Environmental Change: Monitoring, Challenges and Direction. CSIRO Publishing, Melbourne. Lindenmayer DB, Barton P, Pierson J (Eds) (2015a) Indicators and Surrogates of Biodiversity and Environmental Change. CSIRO Publishing, Melbourne. Lindenmayer DB, Burns EL, Tennant P, Dickman CR, Green PT, Keith DA, Metcalfe DJ, Russell-Smith J, Wardle GM, Williams D, Bossard K, deLacey C, Hanigan I, Bull CM, Gillespie G, Hobbs RJ, Krebs CJ, Likens GE, Porter J, Vardon M (2015b) Contemplating the future: acting now on long-term monitoring to answer 2050’s questions. Austral Ecology 40, 213–224. doi:10.1111/aec.12207 Lindenmayer DB, Pierson J, Barton P, Beger M, Branquinho C, Calhoun A, Caro T, Greig H, Gross J, Heino J, Hunter M, Lane P, Longo C, Martin K, McDowell WH, Mellin C, Salo H, Tulloch A, Westgate M (2015c) A new framework for selecting

1 – Introduction

23

environmental surrogates. The Science of the Total Environment 538, 1029–1038. doi:10.1016/j.scitotenv.2015.08.056 Lovett GM, Burns DA, Driscoll CT, Jemkins JC, Mitchell MJ, Rustad L, Shanley JB, Likens GE, Haeuber R (2007) Who needs environmental monitoring? Frontiers in Ecology and the Environment 5, 253–260. doi:10.1890/1540-9295(2007)5[253:WNEM]2.0.CO;2 Lund J (1978) Changes in the phytoplankton of an English lake, 1945–1977. Hydrology Journal 14, 10–27. Majer JD, Heterick B, Gohr T, Hughes E, Mounsher L, Grigg A (2013) Is thirty-seven years sufficient for full return of the ant biota following restoration? Ecological Processes 2, 19. McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, Possingham HP (2010) Monitoring does not always count. Trends in Ecology & Evolution 25, 547–550. Muir MJ (2010) ‘Are we measuring conservation effectiveness?’ Report to Conservation Measures Partnership, www.conservationmeasures.org. Muller F, Baessler C, Schubert H, Klotz S (Eds) (2011) Long-term Ecological Research: Between Theory and Application. Springer, Dordrecht. Nichols JD, Williams BK (2006) Monitoring for conservation. Trends in Ecology & Evolution 21, 668–673. doi:10.1016/j.tree.2006.08.007 Norton DA (1996) Monitoring biodiversity in New Zealand’s terrestrial ecosystems. In Papers from a seminar series on biodiversity. (Eds B McFadgen, S Simpson) pp. 19–41. Department of Conservation, Wellington, New Zealand. Olsen P, Fuller P, Marples TG (1993) Pesticide-related eggshell thinning in Australian raptors. Emu 93, 1–11. doi:10.1071/MU9930001 Orians GH (1986) The place of science in environmental problem-solving. Environment 28, 12–41. doi:10.1080/00139157.1986.9928824 Owens B (2013) Slow science. Nature 495, 300–303. doi:10.1038/495300a Ozgul A, Tuljapurkar S, Benton TG, Pemberton JM, Clutton-Brock TH, Coulson T (2009) The dynamics of phenotypic change and the shrinking sheep of St. Kilda. Science 325, 464–467. doi:10.1126/science.1173668 Persson I-L, Nilsson MB, Pastor J, Eriksson T, Bergström R, Danell K (2009) Depression of belowground respiration rates at simulated high moose population densities in boreal forests. Ecology 90, 2724–2733. doi:10.1890/08-1662.1 Pyke GH, Recher HF (1983) Censusing Australian birds: a summary of procedures and a scheme for standardisation of data presentation and storage. In Methods of Censusing Birds in Australia. (Ed. SJ Davies) pp. 55–63. Proceedings of a symposium organised by the Zoology section of the ANZAAS and the Western Australian Group of the Royal Australasian Ornithologists Union. Department of Conservation and Environment, Perth. Réale D, McAdam AG, Boutin S, Berteaux D (2003) Genetic and plastic responses of a northern mammal to climate change. Proceedings. Biological Sciences 270, 591–596. doi:10.1098/rspb.2002.2224

24

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Reynolds JH, Knutson MG, Newman KB, Silverman ED, Thompson WL (2016) A roadmap for designing and implementing a biological monitoring program. Environmental Monitoring and Assessment 188, 399–424. doi:10.1007/s10661016-5397-x Rosi-Marshall EJ, Bernhardt ES, Buso DC, Driscoll CT, Likens GE (2016) Acid rain mitigation experiment shifts a forested watershed from a net sink to a net source of nitrogen. Proceedings of the National Academy of Sciences of the United States of America 113, 7580–7583. doi:10.1073/pnas.1607287113 Rowland MM, Vojta CD (2013) A technical guide for monitoring wildlife habitat. USDA, Forest Service. Russell-Smith J, Edwards A, Woinarski JC, Fisher A, Miurphy B, Lawes M, Crase B, Thurgate N (2014) Northern Australia tropical savannas: the Three Parks Savanna Fire-effects Plot Network. In Biodiversity and Environmental Change: Monitoring, Challenges and Direction. (Eds DB Lindenmayer, E Burns, N Thurgate, A Lowe) pp. 335–378. CSIRO Publishing, Melbourne. Sachs J, Remans R, Smukler S, Winowiecki L, Andelman SJ, Cassman KG, Castle D, DeFries R, Denning G, Fanzo J, Jackson LE, Leemans R, Leemans J, Milder JC, Naeem S, Nziguheba G, Palm CA, Reganold JP, Richter DD, Scherr SJ, Sircely J, Sullivan C, Tomich TP, Sanchez PA (2010) Monitoring the world’s agriculture. Nature 466, 558–560. doi:10.1038/466558a Shrader-Frechette KS, McCoy ED (1993) Method in Ecology: Strategies for Conservation. Cambridge University Press, Cambridge. Smith AP (1984) Demographic consequences of reproduction, dispersal and social interaction in a population of Leadbeater’s Possum Gymnobelideus leadbeateri. In Possums and Gliders. (Eds AP Smith, ID Hume) pp. 359–373. Surrey Beatty & Sons, Sydney. Smith TB, Purcell J, Barino JF (2007) The rocky intertidal biota of the Florida Keys: fifty-two years of change after Stephenson and Stephenson (1950). Bulletin of Marine Science 80, 1–19. South African Environmental Observation Network (2015) SAEON Highlights 2011–2015. South African Environmental Observation Network, South Africa. Spellerberg IF (1994) Monitoring Ecological Change. 2nd edn. Cambridge University Press, Cambridge. Stankey GH, Bormann BT, Ryan C, Shindler B, Sturtevant V, Clark RN, Philpot C (2003) Adaptive management and the Northwest Forest Plan: rhetoric and reality. Journal of Forestry 101, 40–46. State of Environment (2017) State of Environment Report. Department of Environment and Energy, Canberra, Australia. Strayer DL, Glitzenstein JS, Jones C, Kolasa J, Likens GE, McDonnell M, Parker GG, Pickett STA (1986) Long-term ecological studies: An illustrated account of their design, operation, and importance to ecology. Institute of Ecosystem Studies, Millbrook, New York. Sutherland W (1996) Ecological Census Techniques. Cambridge University Press, Cambridge.

1 – Introduction

25

Swaisgood RR, Wang D, Wei F (2017) Panda downlisted but not out of the woods. Conservation Letters doi:10.1111/conl.12355. Taylor C, Cadenhead N, Lindenmayer DB, Wintle BA (2017) Improving the design of a conservation reserve for a critically endangered species. PLOS One 12(1), e0169629. Thompson WL, White GC, Gowan C (1998) Monitoring Vertebrate Populations. Academic Press, London. Timofeyev MA (2016) Safegurarding the world’s largest lake. Nature 538, 41. doi:10.1038/538041a Todd CR, Lindenmayer DB, Stamation K, Acevedo-Catteneo S, Smih S, Lumsden LF (2016) Assessing reserve effectiveness: application to a threatened species in a dynamic fire prone forest landscape. Ecological Modelling 338, 90–100. doi:10.1016/j.ecolmodel.2016.07.021 Tuttle EM, Bergland AO, Korody ML, Brewer MS, Newhouse DJ, Minx P, Stager M, Betuel A, Cheviron ZA, Warren WC, Gonser RA, Balakrishnan CN (2016) Divergence and functional degradation of a sex chromosome-like supergene. Current Biology 26, 344–350. doi:10.1016/j.cub.2015.11.069 Vitousek PM, Reiners WA (1975) Ecosystem succession and nutrient retention: a hypothesis. Bioscience 25, 376–381. doi:10.2307/1297148 Walters CJ (1986) Adaptive Management of Renewable Resources. Macmillan Publishing Company, New York. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899 Wexler SK, Goodale CL, McGuire KJ, Bailey SW, Groffman PM (2014) Isotopic signals of summer denitrification in a northern hardwood forested catchment. Proceedings of the National Academy of Sciences of the United States of America 111, 16413– 16418. doi:10.1073/pnas.1404321111 Wilkins S, Keith DA, Adam P (2003) Measuring success: evaluating the restoration of a grassy eucalypt woodland on the Cumberland Plain, Sydney, Australia. Restoration Ecology 11, 489–503. doi:10.1046/j.1526-100X.2003.rec0244.x Winder M, Schindler DE (2004a) Climate change uncouples trophic interactions in a lake ecosystem. Ecology 85, 2100–2106. doi:10.1890/04-0151 Winder M, Schindler DE (2004b) Climatic effects on the phenology of lake processes. Global Change Biology 10, 1844–1856. doi:10.1111/j.1365-2486.2004.00849.x Wintle BA, Runge MC, Bekessy SA (2010) Allocating monitoring effort in the face of unknown unknowns. Ecology Letters 13, 1325–1337. doi:10.1111/j.1461-0248.2010. 01514.x Yoccoz NG, Nichols JD, Boulinier T (2001) Monitoring of biological diversity in space and time. Trends in Ecology & Evolution 16, 446–453. doi:10.1016/ S0169-5347(01)02205-4 Zhan X, Li MY, Zhang Z, Goossens B, Chen Y, Wang H, Bruford M, Wei F (2006) Molecular censusing doubles giant panda population estimate in a key nature reserve. Current Biology 16, R451–R452. doi:10.1016/j.cub.2006.05.042

This page intentionally left blank

2 Why monitoring fails

‘… like a long proof in mathematics in which the first calculation is wrong, following which all other calculations move you further away from how things were when they made sense.’ (Richard Ford, Canada, 2012, p. 7) These lines from the novel by Richard Ford relate to a failed marriage, but they are equally apt to failed monitoring programs in which the initial steps are flawed and subsequent steps are also fraught with miss-steps.

There have been some highly successful long-term ecological monitoring programs. Perhaps one of the most famous and enduring long-term ecological studies is the Park Grass Experiment at Rothamsted Experimental Station in Hertfordshire, England. The study commenced in 1856 to test the effect of fertilisers on hay yields, but has since been modified to include other treatments. Results over this long period have demonstrated that: (1) conventional field trials probably underestimate threats to plant biodiversity from long-term changes, such as soil acidification; (2) plant species richness, biomass and soil pH are strongly interrelated; (3) competition between plants can magnify the effects of climatic variation on communities; and (4) local evolutionary changes can occur under

28

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

different selection pressures (Silvertown et al. 2006). Examples of excellent longterm studies in aquatic ecosystems include those by Goldman (1981), Schindler et al. (1985), Kingsford et al. (2008), Holmes and Likens (2016) and Timofeyev et al. (2016) (see Chapters 3 and 4). Alongside the cases of effective monitoring, there is an unfortunate history of poorly planned and unfocused monitoring programs that are either ineffective or fail completely (Orians 1986; Norton 1996; Nichols and Williams 2006). For example, Allen (1993) and Norton (1996) describe how nearly half of the more than 55 monitoring programs on tussock grasslands in New Zealand were unreported, indicating a failure rate that is extremely high. A large number of long-term studies in Australia also remain unreported or have been subject to only limited reporting and scientific publication (Youngentob et al. 2013). Similarly, Ward et al. (1986) lament about the ‘data-rich but information-poor’ syndrome in water quality monitoring. There are also several cases in threatened species monitoring where species were monitored without appropriate management intervention until they went extinct (Lindenmayer et al. 2013b) – a clear case of not only scientific but also policy, political, moral and ethical failure (Woinarski et al. 2017). Another issue, but of lesser importance in this regard, is the failure of investigators to update their information and PowerPoint slides when making public presentations. This sloppiness gives a poor impression of the true value of long-term data and may provide misleading information about important current trends. Environmental changes can occur rapidly, particularly in these times of increased environmental damage, making it imperative to not only keep records/trends up-to-date, but to inspect them frequently. In this chapter, we briefly outline why monitoring programs often fail. The reasons, summarised in Table 2.1, include a lack of questions, poor experimental design, unresolved arguments about what to monitor, over-emphasis on built infrastructure, a rush to be seen to be doing ‘practical’ on-the-ground work, disengagement of the scientific community, a lack of trigger points in the monitoring program for management action, poor data management and breaches of long-term data integrity. We also discuss other factors that contribute to the failure of monitoring programs: loss and/or lack of funding, the loss of a pivotal person, unexpected problems (such as highly obstructive individuals) and excessive bureaucracy. The content of this chapter is a prelude to the following chapter where we propose antidotes to the problems we discuss here.

Characteristics of ineffective monitoring programs Failure to ask the preliminary and fundamental question – Is monitoring needed at all? A fundamental question before the commencement of any monitoring program must be: Is a monitoring program actually needed? There may be cases where

2 – W h y m o n i t o r i n g f a il s

29

Table 2.1.  Reasons why monitoring programs and long-term studies can fail or be ineffective Problem

Reference

Mindless, lacking questions

(Lindenmayer and Likens 2009)

Poor experimental design

(Bernhardt et al. 2005)

Monitoring too many things poorly rather than fewer things well

(Zeide 1994)

Failure to agree on what entities to monitor

(Lindenmayer et al. 2015c)

Flawed assumption that all monitoring programs can be the same

(Lindenmayer and Likens 2011b)

Scientific disengagement from monitoring programs

(Franklin et al. 1999)

Lack of trigger points for management action

(Lindenmayer et al. 2013b))

Poor data management

(Caughlan and Oakly 2001)

Loss of integrity of the long-term data record

(Strayer et al. 1986)

Lack of funding

(Caughlan and Oakly 2001; Timofeyev et al. 2016)

Loss of key personnel

(Kendeigh 1982)

Unexpected major event

(Laurance and Luizao 2007)

monitoring is not necessary (McDonald-Madden et al. 2010), such as where a given management agency has no intention of changing management practices (irrespective of monitoring results) or where there is limited capacity to gain new knowledge. We note how rarely that such a fundamental question is actually asked before the commencement of a monitoring program.

Passive, mindless and lacking questions Some monitoring programs have been driven by short-term funding or a political directive rather than being underpinned by carefully posed questions and objectives. Roberts (1991) and Nichols and Williams (2006) argued that too often monitoring has been ‘planned backwards on the collect now (data), think-later (of a useful question) principle’. A paucity of questions is a serious problem because it often results in monitoring programs being poorly focused and incapable of delivering effective outcomes (Legg and Nagy 2006; Martin et al. 2007). In other cases, this deficiency means that it is not possible to diagnose the cause of a change which, in turn, limits predictive capability either through time or spatially to other landscapes or environments. An example of the problems created by passive monitoring is that of the Norfolk Island boobook owl (Ninox novaeseelandiae undulata). Norfolk Island is a small oceanic island over 350 km east of the city of Brisbane in eastern Australia. It supported a rare island-endemic subspecies of owl. Populations of the species declined over a prolonged period, as revealed by a passive and highly haphazard monitoring program. That work was not guided by any meaningful or tractable

30

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 2.1.  Nestlings of hybrids of the Norfolk Island boobook owl and the New Zealand morepork owl (Photo by Penny Olsen)

scientific questions, nor were there any contrasts in management treatments or interventions to determine the causes of decline, although many possible causes for it were proposed (Olsen 1996). Hence, all that could be surmised from the passive monitoring work was that the population was declining. The result was that the species became functionally extinct when a single individual (a female) remained (Garnett et al. 2011) (Olsen 1996; Norman et al. 1998). In 1987, two male New Zealand morepork owls (Ninox novaeseelandiae novaeseelandiae) were released on Norfolk Island and one bred with the single remaining Norfolk Island boobook owl and their offspring (Fig. 2.1) have assisted a recovery of the population (Garnett et al. 2011) to an estimated 40 birds (P Olsen, pers. comm.).

Lack of trigger points for action The demise of the Norfolk Island boobook owl is a far from isolated case in which a species has been monitored until it went extinct or became functionally extinct. Lindenmayer et al. (2013b) provide a list of more than 20 species, from insects such as butterflies to mega-charismatic vertebrates like the West African black rhinoceros (Diceros bicornis ssp longipes) that have been monitored until they went regionally, nationally or globally extinct. These monitoring programs lacked trigger points for action once monitoring work had identified a substantial decline in key populations. In these cases, there were no changes in management despite the clear demise of the species. This lack of action may have occurred because the drivers of decline were unknown, the political will to address the problem was

2 – W h y m o n i t o r i n g f a il s

31

lacking, populations had surpassed a size beyond which is was possible to recover, or a range of other factors (Lindenmayer et al. 2013b). Irrespective of the reasons, each of these monitoring programs could only be described as failures because ongoing decline and eventual extinction was not averted. Notably, some authors have argued that infamous extinctions like that of the passenger pigeon (Ectopistes migratorius) in eastern North America might have been averted if monitoring were to have been in place, alerting managers to its decline and the need for appropriate interventions, particularly controls on overhunting (Stanton 2014). More recently, some authors have called for a formal coronial inquest approach that is applied to policy makers and resource managers who fail to act appropriately to prevent a species from going extinct (Woinarski et al. 2017).

Poor experimental design A third problem has been that monitoring programs have often been very poorly designed at the beginning of a study (Krebs 1991). The quote at the beginning of this chapter highlights the problems that can besiege a monitoring program from the outset when time is not taken to get the experimental design right. Good design is an inherently statistical process (Cunningham and Lindenmayer 2017), but professional statisticians are often left out of the experimental design phases of monitoring programs (Hayward et al. 2015). Issues are then overlooked, including: (1) calculations of statistical power to detect trends (Reed and Blaustein 1995; Strayer 1999; Foster 2001; Field et al. 2007) such as levels of replication of different treatments (Cunningham and Lindenmayer 2017); (2) detectability of particular individual biotic species or chemical elements (Yoccoz et al. 2001; Pellet and Schmidt 2005; Martin et al. 2007; Banks-Leite et al. 2014); (3) optimisation of field methods and statistical design (Mac Nally 1997; Joseph et al. 2006); (4) the importance of contrasts between treatments (e.g. where there is a human intervention and where there is not) (Cunningham and Lindenmayer 2017); (5) the value of innovative rotating of sampling methods for increasing inference (Welsh et al. 2000; Lindenmayer et al. 2012); and (6) matching the timing and periodicity of repeat measures to the phenomena of interest (e.g. rates of tree mortality Acker et al. 2015). Poor design has numerous knock-on effects that can result in the failure of a monitoring program (Legg and Nagy 2006). For example, it can lead to the results of work not being written up, or when it is, making it difficult for findings to be published in reputable outlets. Poor design also means that it is difficult to assess the effectiveness of a management intervention, such as application of a prescribed burning regime in forest ecosystems (Whelan 1995) or a major environmental initiative including national or even continental agri-environmental schemes (Kleijn and Sutherland 2003; Kay et al. 2016). For example, it has not been possible to assess the effectiveness of ~US$15 billion worth of projects on river restoration in the United States because of poor experimental design and a lack of rigorous

32

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

monitoring of interventions designed to improve river and stream environments (Bernhardt et al. 2005). Indeed, in ~90% of projects there was no form of assessment or monitoring of project effectiveness and limited data to determine which activities had been successful and which had not (Bernhardt et al. 2005). Hence, key opportunities for management learning were lost. In other cases, poorly designed monitoring programs could lead to an incorrect decision being made, such as the downlisting of an endangered species when it should not have been (Martin et al. 2007).

Snowed by a blizzard of ecological details A fourth issue is that the design of monitoring programs is often prefaced by protracted (and frequently unresolved) arguments about what to monitor. One response has been to monitor a large number of things (the so-called ‘laundry list’). As we outline in the case studies in Chapter 4, some monitoring programs are indeed based on very extensive lists (Zeide 1994). However, the laundry list approach can have a range of problems. First, it can divert those responsible for establishing a monitoring program from posing well-crafted and tractable questions. Second, resource and time constraints frequently mean that a poorly focused laundry list will result in many things being monitored badly. It is simply not possible to properly monitor a vast number of entities (Zeide 1994). Third, a laundry list may make a monitoring program too expensive to be sustained financially beyond the short-term and may ultimately lead to its collapse. In cases where the objective of a monitoring program is to assess the impacts of resource management practices (e.g. prescriptions for logging operations), demands to measure a long list of attributes may mean that the costs of a monitoring program are mismatched with the level of economic return from that management practice (Franklin et al. 1999). This conflict also can quickly undermine a monitoring program. Finally, a laundry list approach can create problems with the statistical design of a monitoring program. Monitoring a long list of entities can only ever realistically be done on a few sites, but often a robust statistical design for a monitoring program will sometimes call for many sites to be subject to repeated survey (Cunningham and Lindenmayer 2017), such as to provide sufficient statistical power to detect an effect or identify a trend. We suggest that laundry lists should be regarded only as starting points in planning from which a subset of entities appropriate to answer well-defined questions are selected. This approach is important because laundry lists do not reflect the realities of operating or financing a credible monitoring program.

Squabbles about what to monitor – ‘It’s not monitoring without the mayflies’ An alternative response by some workers to the laundry list approach has been to argue that ‘indicator species’ or ‘indicator groups’ should be the targets of

2 – W h y m o n i t o r i n g f a il s

33

monitoring programs. Many would argue that the group of organisms they study is special and any valid monitoring program cannot proceed without including them. One of us (DBL) recalled this problem vividly in discussions about establishing monitoring programs in the forests of the Australian state of New South Wales. One scientist in the meeting repeatedly stated that the proposed monitoring program (which actually did not eventuate) was invalid because one group – mayflies – were not included. Mayflies are undoubtedly important animals but monitoring populations of mayflies would not have answered the questions that were being posed in the proposed monitoring program. We have found that over 55 major taxonomic groups have been proposed as indicators, ranging from viruses, fungi and bryophytes to invertebrates and virtually all major vertebrate groups (Lindenmayer and Likens 2011a). Only very rarely was it explicitly stated: (1) what these species or groups were actually indicative of, particularly at the ecosystem level; and (2) the circumstances in which these species or groups were or were not appropriate indicators (Lindenmayer and Likens 2011a). There is a vast literature on indicators and indicator species concepts (Westgate et al. 2014) (including the journal Ecological Indicators, which focuses extensively on the topic) and it is far beyond the scope of this book to review them here. However, we briefly outline five major, problematic issues. First, almost all applications of the indicator species concept are highly idiosyncratic; that is, they are organism-specific, landscape-specific or ecosystemspecific. Hence, transferability to other groups, landscapes, ecosystems, environmental circumstances or over time is problematic and when circumstances change, the difficult and often time-consuming process of identifying new indicator species must be instigated (Lindenmayer et al. 2015c). For example, there are many cases where a given group is recommended as an indicator species in one environment, but is then found to perform poorly in that surrogate role in another environment or landscape (Sergio et al. 2008). Examples include raptors (Thiollay 1996; Sergio et al. 2008; Rodriguez-Estrella et al. 1998; Roth and Weber 2008), small mammals (Pearce and Venier 2005; Lehmkuhl et al. 2008), arboreal marsupials (Milledge et al. 1991; Lindenmayer and Cunningham 1997; Fig. 2.2) and plants (Gegout and Krizova 2003). Even the same species may respond in markedly different ways in different landscapes (Dooley and Bowers 1998), aquatic ecosystems (Potapova and Charles 2007) or seascapes (Kirby et al. 2006). Second, different species respond in often quite markedly different ways to environmental change. This difference can be true even for closely related species such as those from the same genus or guild (Thiollay 1996). Third, there is a general lack of understanding of causal relationships between indicator species and the entities they are assumed to indicate (Lindenmayer et al. 2015c). Indeed, in several cases where causal relationships

34

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 2.2.  The yellow-bellied glider. This spectacular wide-ranging gliding arboreal marsupial from Australian forests was proposed as an indicator species of forest suitability for arboreal marsupials and other vertebrates (Milledge et al. 1991), but subsequent analyses showed that an average of only two (of a possible seven) other species in the arboreal marsupial assemblage co-occurred with it (Lindenmayer and Cunningham 1997). Moreover, the responses of arboreal marsupials to landscape conditions and disturbance regimes (Lindenmayer et al. 2013a) proved to be markedly different from those of other groups such as small mammals (Banks et al. 2011), birds (Lindenmayer et al. 2014a) and plants (Blair et al. 2016). (Photo by Esther Beaton)

have been carefully examined, a purported indicator species subsequently has proven to be a misleading surrogate of environmental conditions or the occurrence of other elements of biota. For example, early research on the bivalve mollusc Velesunio ambiguus in Australian rivers suggested that the species was an indicator of heavy metal pollution (Walker 1981). Subsequent work found that the uptake of heavy metals by V. ambiguus did not reflect the extent of pollution in the surrounding river system, making the mollusc an unreliable and unsuitable indicator species (Millington and Walker 1983; Maher and Norris 1990). There are other examples of this problem. Overstorey vegetation cover is a useful surrogate for the occurrence of some bird species in the temperate woodlands of inland Australia, but there is a marked lack of consistency in the relationship between change in overstorey cover and population trends among bird species, both within and between geographic regions. This measure of vegetation structure also provides only limited information about individual species population trends (Pierson et al. 2016).

2 – W h y m o n i t o r i n g f a il s

35

A fourth problem with the indicator species approach is that there is often poor or vague specification of the environmental conditions for which particular indicators are purported to be indicative of (Lindenmayer and Likens 2011a), such as ‘ecosystem health’ (Bouyer and Sana 2007), ‘disturbed health’ (Mariogómez et al. 2006) and ‘ecosystem vital signs’ (Dixit et al. 1992). Finally, in many cases there is over-simplified, somewhat misleading and often confusing usage of the term ‘indicator’ as simply the entity which is being measured (e.g. The Heinz Center 2008). Caution when employing the indicator species and other surrogate approaches is important. It is critical for resource managers, policy makers and scientists to be made aware that a particular species may not be a good indicator of the presence or abundance of other species, of the suitability of the ecosystem to support species assemblages, or of ecosystem processes. Otherwise they may believe that by conserving a so-called indicator species they have effectively conserved all other biota when in fact they have been deceived by a false sense of management security (Lindenmayer et al. 2015c). There is no doubt that particular species and sets of species should be the focus of long-term research and monitoring programs. It would be nonsensical to argue otherwise. But it should not necessarily be assumed that these species or groups of species are indicators. Sometimes it will be better to quantify the direct measure and not assume that a particular species or set of species is a surrogate for something else (Lindenmayer and Likens 2011a). However, we also readily acknowledge that in some circumstances, the real-time direct measurement of some entities is not possible. As an example, it is obviously not possible to measure paleoclimates or environmental conditions directly; the presence of species such as diatoms has been used as a surrogate in efforts to reconstruct past climates, ocean conditions and other environmental conditions (e.g. Charles et al. 1990). We reiterate that a key part of selecting surrogates and indicators is that appropriate entities to monitor should emerge from the questions being asked. In Box 2.1, we describe the application of an adaptive surrogacy framework to further assist in the selection of surrogates and indicators.

Assumption that ‘one size fits all’ Many scientific articles make recommendations about generic frameworks for monitoring programs (e.g. Sachs et al. 2010). Recommendations are made about lists of entities to be measured and the way they should be measured. These kinds of recommendations are understandable given a desire to standardise data recording and increase coordination and comparison across studies. However, this approach is problematic for several reasons. It is often not appropriate to measure the same entities (e.g. the same species groups or ecosystem function) in different places, which is frequently mandated in

36

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 2.1. An adaptive surrogacy framework to tackle problems associated with deciding what to monitor The issue of which surrogates or proxies are the best ones to measure is a vexed problem that has been resolved in very few cases. A new framework – termed an adaptive surrogacy framework – has recently been proposed in an attempt to assist scientists, resource managers and policy makers to better identify and then rigorously assess surrogates and indicators (Lindenmayer et al. 2015c). The scientific

Fig. 2.3.  The adaptive surrogacy framework showing links and feedback loops between the seven components, and sub-components, that might be considered in the development, application, and learning about surrogates (modified from Lindenmayer et al. 2015c)

2 – W h y m o n i t o r i n g f a il s

37

underpinnings of the adaptive surrogacy framework include hypothesis testing and risk assessment, where proposed surrogates are treated as working hypotheses to be subjected to rigorous testing. This approach provides a formal framework to improve the application of surrogates continuously over time. The sequence of steps is shown in Fig. 2.3. Selection of appropriate surrogates is guided by community input regarding which ones are suitable for monitoring. They are identified against a background of a conceptual model of how an ecosystem functions in response to changes in the proposed surrogate.

‘big science’ efforts such as was done in the International Biological Program (IBP) as well as in initiatives like the Long-Term Ecological Research (LTER) network and National Earth Observation Network (NEON) in the United States (see Chapter 4). Measuring diverse assemblages of butterflies may well be valid in tropical rainforests where this group can be species rich, but is likely to be of little merit in desert environments where this group may have relatively few representatives (Lindenmayer et al. 2015b). Similarly, the kinds of threatening processes in a tropical savanna and an alpine meadow may be so different that environmental monitoring programs would most likely need to be very different. Even within the same broad vegetation type (e.g. tropical rainforest), the application of identical monitoring protocols may not be particularly informative because of differences in biota, key ecological processes or other factors (see also Martins et al. 2007). As we have outlined, the choice of what entities to measure and how they should be measured is best guided not by a generic framework but rather by well-defined and scientifically tractable questions.

Big machines that go ‘bing’ ‘An exciting mix of nano-engineered materials, miniaturised computers and rapid wireless communications is giving rise to a new generation of environmental sensors. These new sensors will be able to process information they collect and adapt to changing conditions. They’ll be able to measure an amazingly wide range of environmental variables, and they will be able to share this information with a network of sensors. This may be the way of the future for all forms of environmental management, be it catchment management, bushfire risk management, agriculture or monitoring invading pests.’ The above quote is a part of a media release from the Australian Academy of Sciences in late 2008. It suggests that new equipment is the smart science, not the questions being asked to solve a problem. We suggest that appropriate equipment,

38

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

electronic sensors and aspects of social media are indeed very important for long-term research and monitoring programs, but an over-emphasis on these approaches can be problematic especially when human observers are not involved with these approaches. An increasingly common issue in long-term research and monitoring programs is the focus on building expensive infrastructure (e.g. fences, flux towers, and electronic detection and transmission systems) (Walters and Scholes 2017). This infrastructure may create a range of problems and cause monitoring programs to fail. The first is experimental design. When a piece of equipment is expensive there will rarely be many of them, and sometimes only one per catchment, lake, region, state or country. This scarcity, in turn, creates problems for statistical design because it precludes replication, prevents quantification of variability, and limits the spatial extent of inferences that can be made from findings. A second issue is that some kinds of expensive scientific instruments can generate vast quantities of data very quickly. This potential overload can lead to the temptation to do science backwards (Lindenmayer and Likens 2013). That is, data are mined for answers, rather than questions being posed and then data gathered to answer them. This backward approach can create a disconnect between patterns that are measured and identifying the causal and ecological factors that give rise to those patterns. An example of this problem is the discovery of the ozone hole over Antarctica; extensive data-generating methods were employed by American researchers but without their being underpinned by well-focused questions, they missed detection of the ozone hole. By contrast, British researchers developed well-focused questions and using relatively simple methods that did not generate enormous volumes of data, discovered the ozone hole (Shanklin 2010). It also can be difficult to know how to analyse veritable mountains of data that are generated from sophisticated field equipment and then, in turn, use it in real-world environmental management. Indeed, the editor of Nature magazine has argued that large datasets can be prone to errors or the creation of artefacts that are misinterpreted as ecologically meaningful results from which careless mistakes are published (Anon. 2012). We acknowledge that there are enormous opportunities for new discoveries to be made from analysing ‘big data’ (e.g. Reichman et al. 2011; McNutt et al. 2016; Franklin et al. 2017). However, we suggest that such opportunities should not replace the critical need for clear thinking, good questions and robust experimental design in the inception of monitoring programs (see Chapters 4 and 5). Third, a focus on expensive equipment often fails to recognise that people are the most important infrastructure for maintaining long-term research and monitoring programs. For example, the environment sections in the roadmap document for the Australian National Collaborative Research Infrastructure Scheme (https://www.education.gov.au/2016-national-research-infrastructure-

2 – W h y m o n i t o r i n g f a il s

39

roadmap) has a distinct focus on remote detection devices and other equipment; people are somehow regarded as a negative component of monitoring programs and observation networks as it increasingly seems that it is easier to obtain funds to purchase equipment than to employ people to oversee and operate this equipment. Yet, there are many sobering cases where remotely-sensed information lacks the accuracy required for some kinds of monitoring work. People are essential to ground-truth the data streamed by remote sensors. An example is remotely-sensed vegetation cover data in north Queensland, Australia which purports to show the location of habitat for the southern cassowary (Casuarius casuarius). Even a very cursory on-the-ground reconnaissance quickly reveals that this is anything but the case, leading to inappropriate management actions of limited conservation significance. It seems absurd to us that literally billions of dollars can be spent on very expensive equipment yet the employment of people for on-the-ground work is increasingly difficult. Fourth, the purchase of expensive new equipment may change the way particular entities (e.g. water flows, nutrient levels and animal populations) are measured. This approach can breach the integrity of past data collection and invalidate the continuity of some measures such as time series information (Shapiro and Swain 1983). At the very least, the data being gathered with addition of new technology and built infrastructure must be carefully calibrated against pre-existing data measurements under highly varied conditions, although in some cases these calibrations may indicate that the data streams are simply not compatible (Cunningham et al. 2004). Fifth, sophisticated equipment can be expensive to install and maintain. Funding or interest can wane or the equipment can become outdated or superseded and the operators walk away after a relatively short time. This problem may result in little more than littering of the landscape in the long term. Directing large amounts of funding around the need to maintain and operate built infrastructure often occurs at the expense of some other source of scientific funding. Therefore, the opportunity costs of large infrastructure grants may be the reduction of funding for question-based and curiosity-driven science. Large, sophisticated equipment are powerful data-gathering tools and can make important contributions to long-term research and monitoring. However, as we discuss further in the case studies in Chapter 4, major problems can arise in programs with an excessive focus on built infrastructure and an associated failure of funding bodies to recognise that people are a critical part of the infrastructure needed to facilitate and maintain high-quality monitoring programs. We therefore strongly argue that expenditure on infrastructure should not unjustifiably dominate budgets for long-term research and monitoring programs. We recognise that some people will criticise us for being old-fashioned about seeming to be against the adoption of new ways to conduct ecological monitoring.

40

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

We are in no way opposed to the application of new methods in research and monitoring or the use of sophisticated, expensive field equipment. But the purchase and deployment of such equipment should be justified by the questions asked, and should not jeopardise the funding needed for critical personnel at a field site. We further discuss these challenges in Chapter 4, particularly in the context of major national and international initiatives that shape modern ecological research and monitoring like NEON in the United States and the Terrestrial Ecosystem Research Network (TERN) in Australia.

Disengagement Although scientists have stressed the importance of monitoring for many years in innumerable scientific articles and books, they have often refused to involve themselves directly in monitoring, which they have viewed as the routine collection of data for non-scientific purposes (see Franklin et al. 1999). In other cases, scientists have maintained a ‘superiority’ complex and considered monitoring to be too ‘trivial’ to warrant their involvement. This prejudice of the scientific community is based on a variety of factors, particularly the failure of academic reward systems to recognise and credit involvement in monitoring programs (Taylor et al. 1998; Nisbet 2007; McDowell 2015). Finally, scientists may not feel qualified or physically engaged to participate in new big-science efforts, which are focused on continuously measuring ‘everything’ on huge, spatial scales (McDowell 2015; see also Chapter 4 in this book). The fact that many resource managers do not consider science or scientists to be essential participants in development and operation of their monitoring programs is yet another problem. In other cases, scientists do not treat policy makers with appropriate respect and often regard them as ‘cash cows’ for providing funding to support their ‘pet’ research projects. However, while scientists may design elegant experiments and other kinds of studies that go on to produce fascinating papers in high-quality scientific journals, their work may spectacularly fail the test of management relevance. Russell-Smith et al. (2003) provide a sobering example of this disconnect in a fire experiment in the savannas of northern Australia, which was published in the prestigious journal Ecological Monographs. Although that study encompassed elegantly designed and replicated experimental plots, key questions about the effects of multiple fires at several spatial scales could not be addressed, even though answers to them were of primary interest and relevance to resource managers (Russell-Smith et al. 2003). A key part of the problem was that the experiment was designed and implemented without the engagement of landowners and resource managers (Russell-Smith et al. 2003). Conversely, resource managers and policy makers will often have demands to monitor the effectiveness and outcomes of important management interventions and environmental programs, but lack the scientific and statistical skills to do it. Partial failure to demonstrate effectiveness

2 – W h y m o n i t o r i n g f a il s

41

in multi-billion dollar agri-environmental schemes in Europe and North America has been attributed, in part, to a paucity of scientific engagement with resource managers and the subsequent development of robust and credible monitoring programs (Kleijn and Sutherland 2003; Kay et al. 2016). In summary, disengagement is a truly substantial issue, particularly from the scientific community. It is also clear that scientists cannot expect engagement from policy makers and extract significant funds to support monitoring unless they can make a compelling case for it. This case can only be advocated when scientists are able to convince policy makers that they can overcome problems with experimental design, reach an agreement on what to monitor, and provide policyrelevant information.

Rush to get ‘real work’ happening on the ground and accusations of program over-engineering Some policy managers perceive the only ‘true’ monitoring to be that which takes place in the field, with people doing hard physical work. This perception has sometimes meant there has been a rush to ‘get people to work’, without giving the necessary time to think through and then carefully plan a monitoring design and subsequent field program. Indeed, time spent planning may be regarded as idle time wasting. This view can occur despite workers in quite different disciplines recognising that difficult problems and complex problems (both of which monitoring programs can sometimes create) require one to take the time to think deeply (Stirzaker 2010). Our collective experience in establishing long-term studies is that it can take many months to develop a tractable and practical design for a program that may then run for many years, if not several decades; time spent planning is critical to get the design and implementation right. A related issue is that some managers and policy makers consider that scientists over-engineer long-term studies to deliver a ‘Rolls-Royce’ program that scientists want rather than a basic model Ford that some policy makers think they need. As an example, one of us (DBL) was responsible for developing the monitoring program for the Australian Government’s Environmental Stewardship Program in which private farmers were paid under contract to manage threatened woodlands for conservation outcomes (Lindenmayer et al. 2012). A key part of the program was to monitor the change in woodland condition and in biodiversity resulting from stewardship management activities like control of grazing by livestock, weed control and replanting efforts. Over 150 stewardship patches had to be monitored. However, to tell whether stewardship management was effective, we needed also to monitor patches where stewardship management had not occurred. These matched ‘reference’ patches needed to be on the same farms as the stewardship patches. Initially, policy makers for the Environmental Stewardship Program did not want to invest in monitoring non-stewardship patches and claimed that the monitoring

42

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

was over-engineered. Yet, when it rains in woodland environments, all areas improve in condition and during droughts the condition of all areas deteriorates. Reference sites are critical to account for these major interannual changes driven by rainfall variability and to determine whether stewardship management is truly effective. The need for reference sites was finally accepted by policy makers for the Environmental Stewardship Program with the recognition of the fact that the monitoring was anything but over-engineered. Importantly, the monitoring program has been able to demonstrate that the Environmental Stewardship Program is leading to improvements in woodland condition and a range of measures of biodiversity (Kay et al. 2016; Sato et al. 2016). Sadly, in late 2017 the Australian Government radically altered the monitoring program and it will no longer be possible to tell if the Environmental Stewardship Program is effective (see Box 2.3).

Poor data management Many monitoring programs have failed to employ adequate data management procedures, such as appropriate standards for data documentation, and few have provided adequate financial support for data management (Caughlan and Oakly 2001). A general rule of thumb for data management costs in the US LTER program is ~15% of project costs, but some sites are higher and some lower. Some non-LTER recommendations for information management, including data management, in long-term research and monitoring programs may be 20% or even more of total project costs (National Research Council 1995; Caughlan and Oakly 2001; Berman and Cerf 2013). More recent calls for data to be open access create even greater challenges for data management, particularly given the very substantial time and costs of curation such as the need to check data quality carefully and to format data and create appropriate metadata specifications (Leonelli 2016). Indeed, many ‘big datasets’ may contain considerable errors that lead to significant problems such as geographic inaccuracies in diversity patterns or inflated estimates of species richness (Maldonado et al. 2015). Data management costs are rarely supported in research funding and other grants (Berman and Cerf 2013). Adequate management of long-term environmental datasets is very challenging and requires substantial technical expertise and significant financial support. Managers and scientists have often failed to recognise the technical challenges, costs and critical role of data management (see Box 2.2), preferring to use financial resources for other priorities. The reality is that the curation of data is often an afterthought in most of ecological projects. As a consequence, immense amounts of important environmental data, including long-term records and/or their metadata, have disappeared or effectively been lost due to inadequate documentation and quality control (Norton 1996). Indeed, one of us (DBL) has personally witnessed several cases where government agencies have

2 – W h y m o n i t o r i n g f a il s

43

Box 2.2. The challenges of long-term data management Some of the challenges of data management from long-term studies are illustrated by the experience in the 34-year forest ecosystem studies in the wet forests of Victoria (Lindenmayer 2009a; Lindenmayer et al. 2015a). Over the duration of the project, the evolution of data storage methods has included magnetic tapes, 5½-inch floppy disks, 3½-inch floppy disks, CDs, DVDs, memory sticks and, most recently, online cloud-based storage. The time to redundancy of these storage methods has continued to shorten. The development of new data storage methods has required regular updating, and sometimes complete re-coding of datasets, a process that can lead to errors being introduced to datasets or data loss. Moreover, it is very time consuming. For example, the re-coding of old data has required very careful checking as error rates otherwise average 7–10% of the total imported information. A revelation in the project has been the importance of primary data, often in the form of pencil entries on paper copies of field datasheets, although data collection is now switching to tablets with an inbuilt program to signal input errors. The sobering aspect of the digital age (and its many and varied forms) is that many past errors were rectified only by cross-reference with original field datasheets. Field data records are among the most valuable of records kept in the long-term biogeochemical studies at the Hubbard Brook Experimental Forest. These records are referred to regularly to confirm conditions when samples were collected to determine whether unusual environmental conditions may have influenced the results obtained. Moreover, the orderly and safe storage of vast quantities of paper records is important, as is its duplication and storage offsite in case of an unforeseen event such as an office fire, a fate not unknown in government agencies and universities. We suspect that these same problems and issues are common in most long-term research and monitoring programs.

discarded high-quality datasets that soon after were recognised to have been extremely important.

Breaches of data integrity Monitoring programs can fail when the integrity of the data is breached by changes in measurement protocols. There are numerous such cases including this famous example, which has been published in prestigious journals, cited widely, used in textbooks and used in litigation (Harris 1986). Schelske and Stoermer (1971) collected extensive data from 1926 to 1962 showing evidence of significant temporal declines in the amount of dissolved silica in Lake Michigan, which they linked to increased populations of diatoms that they suggested were responding to increased loadings of total phosphorus (eutrophication) in the lake. Dissolved silica is an important nutrient for diatoms. However, Shapiro and Swain (1983) argued that instead, the decline in dissolved silica was due to a change in the analytical method for silica determination and in a change in the laboratories doing the analyses (Fig. 2.4). As

44

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

6 5

Si02 (ppm)

4 3 2 1

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

Year

Fig. 2.4.  Changes in time series data on dissolved silica in Lake Michigan following changes in analytical methods. Vertical line indicates when analytical laboratory was changed (modified from Shapiro and Swain 1983)

such, they considered the quality assurance/quality control to be inadequate and therefore they argued that these long-term data were almost worthless. Schelske (1988) countered with a more detailed analysis of the analytical methods and concluded that there had indeed been a significant decrease in dissolved silica during 1954–69 because of the uptake and sedimentation of silica by diatoms, this process being tightly coupled to phosphorus in the lake. Schelske et al. (1983) further argued that the sediment record in Lake Michigan supported the presence of increased diatoms in the lake. This controversy raged at scientific conferences, in prominent scientific journals, and even in the courts for almost two decades before ending at the US Supreme Court (Mortimer 1981). The answer to the correct interpretation of these long-term data had high relevance for the management of eutrophication in this highly visible and anthropogenically important Laurentian Great Lake. Reviewers of the large-scale study of landscape change at Tumut in southeastern Australia suggested volunteer observer bird surveys be replaced by automatic bird call recorders. However, calibration of the two methods revealed that they gathered markedly different kinds of data (Cunningham et al. 2004; Lindenmayer 2009b). Thus, replacement of one method (human counts) by another (automatic bird call recorders) would have breached the integrity of the long-term data record, making a valid interpretation of the temporal changes in bird populations in the study region impossible (Lindenmayer 2009b). Another recent example is included in Box 2.3.

2 – W h y m o n i t o r i n g f a il s

45

Box 2.3. Altering the monitoring program, breaching data integrity and the loss of effective monitoring of one of Australia’s agri-environmental schemes Large parts of Australia are used for agricultural production and this has had major negative impacts on biodiversity and environmental conditions. One of the most extensively altered ecosystems is the nationally endangered box-gum grassy eucalypt woodland community, which is up to 95–99% cleared in some regions of temperate southern Australia. As part of attempts to recover this ecosystem, the Australian Government entered into contract payments over a 15-year period with 158 farmers to undertake conservation work on their properties. A major monitoring program was instigated to quantify the effectiveness of management interventions on these farms. Box-gum grassy woodlands are relatively slow growing ecosystems and it takes considerable time for these environments to begin recovering (Lindenmayer et al. 2014b). Some positive signs of recovery were apparent from early monitoring work between 2010 and 2016 (Sato et al. 2016). However, more definitive evidence was needed over a longer period to be confident the ecosystem was truly recovering. Unfortunately, in late 2017, the Australian Government elected to change both who was doing the monitoring and the way monitoring was done. This change has breached the data integrity and will mean that two data collection periods, each of approximately seven years in duration, will be incompatible and will have both run for insufficient time to be able to tell if, in fact, the management interventions have been successful. The value proposition for those who have invested time and money in this program is therefore questionable.

There are numerous other cases where new methods can dramatically improve field protocols (e.g. DNA-profiling for detecting individual animals) and enhanced measures like counts of animal abundance. This was the case for the giant panda where molecular methods doubled the known population of the species (Zhan et al. 2006). However, this result also meant that it was not possible to compare new counts with past ones and ascertain what changes in abundance (if any) had taken place in response to reservation and management strategies (Zhan et al. 2006). Other emerging methods such as camera trapping can improve detections of animals (Meek et al. 2014; Burton et al. 2015), but these too create data records that are not readily comparable with previous survey protocols.

Other factors contributing to ineffective monitoring programs Some monitoring programs may be ineffective for reasons other than the ones already summarised. We touch on four of these in the remainder of this chapter, but are mindful that there are still others which we have not discussed here.

46

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Lack of funding – grant myopia Loss or lack of adequate funding can lead to the rapid termination of monitoring programs, including those that are both important and have been highly successful. The scientific literature is littered with innumerable cases of funding problems that have led to the failure of monitoring programs, such as the 40-year study of seabirds of the coast of Wales in the United Kingdom (Birkhead 2014; see also Box 2.4). Short-term funding cycles often emphasise new work rather than long-term studies and at the same time fail to comprehend the profound influence of longterm factors such as site histories and baselines for understanding ecological phenomena (see Silvertown et al. 2006; Gustavsson 2007; Likens and Buso 2010a; Lindenmayer et al. 2014b). Therefore, adequate and sustained funding for monitoring programs is rarely available and valuable insights are lost (Caughlan and Oakly 2001). Since monitoring is often directed towards assessing rather than conducting management activities, funds for monitoring typically have low priority. In many cases scientists and policy makers have failed to calculate the costs of a long-term monitoring program relative to the benefits that it will generate (Box 2.4). Even if adequate funds were provided to plan and initiate a monitoring program (a relatively rare occurrence), most programs have subsequently been starved for financial and logistical resources. As a result, severely under-funded monitoring programs cannot be conducted properly and ultimately fail (see Box 2.3). It is very difficult to assure that adequate financial resources will be continuously available. Most government agencies and corporations have annual budget cycles that make monitoring programs, which are necessarily long-term, vulnerable to budget cuts as short-term shifts occur in organisational priorities and objectives. In our experience, monitoring programs are among the first to lose funding during economic downturns. Many organisations are naïve to the longterm financial and logistical implications of establishing a monitoring program (Walsh and White 1999) and unaware that: ‘Monitoring sustainability will only be possible if the monitoring system is itself sustainable.’ (Watson and Novelly 2004)

The loss of a champion Many monitoring programs fail when the leader or champion for the work leaves, retires or dies (Strayer et al. 1986; Norton 1996). For example, the 27-year studies of birds in the forests of central Illinois by Kendeigh (1982) stopped in 1976 when he retired (Strayer et al. 1986). Garcia and Lescuyer (2008) describe how the handing over of large institutional monitoring programs to local communities of people nearly always results in these programs failing quickly. This problem results from

2 – W h y m o n i t o r i n g f a il s

47

Box 2.4. The potential collapse of a critical monitoring program in the world’s deepest lake Lake Baikal is one of the largest bodies of fresh water, and one of the oldest and most biodiverse lakes, in the world (Fig. 2.5). Water temperature and transparency, plankton abundance and species composition has been monitored at weekly intervals at ~70 stations throughout Lake Baikal since 1945. Currently, coastal zone waters are under threat of eutrophication in this massive, iconic lake because of excessive nutrient and toxic material inputs from industrial and household pollutant runoff from catchments, including from gold mining and logging. This makes monitoring extremely important, but a funding cut of 30% has been proposed from a budget of approximately US$70 000 per year which is trivial relative to the extraordinary ecological and economic value of Lake Baikal (Timofeyev et al. 2016).

Fig. 2.5.  Lake Baikal where a loss of funding threatened the collapse of a critical long-term monitoring program (Photo by Marco Fieber)

the departure of the project champion and loss of the ongoing impetus of the work, as well as a failure to engage with local stakeholders. These kinds of failures reflect the failure of organisations to institutionalise monitoring programs.

Out of nowhere An unforeseen natural disaster or a major human disturbance can threaten long-term research and monitoring programs. For example, Likens (1985)

48

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

described how the development of an interstate highway through the White Mountains of New Hampshire in the United States led to major changes in the integrity of Mirror Lake and its catchment, which had been studied for seven years previously. Some of the massive amounts of salt added to the highway during winter to prevent icing found its way into the lake and greatly increased the salinity of the lake water (Likens 1985; Rosenberry et al. 2007; Likens and Buso 2010b). In response, the New Hampshire Department of Transportation spent more than US$500 000 on structural modifications to divert salt from the lake. At the same time, application of salt to a small, town road adjacent to the lake and crossing two sub-catchments for the lake has become the major source of salt to the lake. Stream input from these two sub-catchments currently provides more than three times as much salt to the lake as does the interstate highway. Meanwhile, the salt concentration in Mirror Lake continues to rise and the funding to monitor this extraordinary saga has ended (Likens and Buso 2010b). There are other cases where a major disturbance can trigger exciting opportunities for ecological studies that lead to important new scientific discoveries and insights. The 1980 volcanic eruption at Mount St Helens and the wildfires at Yellowstone National Park in 1988 in the United States are classic examples. Post-disturbance work in these places resulted in new perspectives on disturbance theory, ecological succession and ecosystem recovery (Franklin and MacMahon 2000; Turner et al. 2003). Very important opportunities can arise where major disturbances damage existing long-term sites, providing a before and after set of contrasts to quantify changes in biodiversity and/or ecosystem condition and where there has and has not been perturbation. This was the case in the wet forests of Victoria, Australia. Fires in 2009 burned a subset of more than 160 long-term sites where 25 years of pre-fire data on vegetation structure and animal occurrence had been gathered (Lindenmayer et al. 2015a). These fires made it possible to rigorously quantify changes pre- and post-fire as well as contrast burned and unburned sites. A critically important part of that work was to commence monitoring as soon as possible after the fire to enable data on key ecosystem variables such as the severity of the fire (Lindenmayer et al. 2010).

Excessive bureaucracy As successful long-term monitoring programs age, they can get bigger and involve more projects and more people. More of the science is then run by committees and more meetings are required to maintain communication across a large group of researchers and administrators. More projects mean more and larger datasets that require data management and data sharing. Intellectual property issues assume greater importance. Increased competition, rather than cooperation (especially among staff from agencies and organisations), can become very counterproductive for the results of long-term monitoring. More people and more projects mean that more grants are required to generate the funds to continue the work. Each of these

2 – W h y m o n i t o r i n g f a il s

49

Fig. 2.6.  The Darwin Cabin in 2009 that was part of the infrastructure underpinning the Trophic Cascade Project in northern Michigan in the United States, but which has now been destroyed (see text) (Photo by J. Hodgson)

factors indicates that the bureaucracy involved in maintaining long-term monitoring programs can increase over time. When this bureaucracy is not carefully managed, monitoring programs that have been successful can ultimately fail because of an increasing focus on ‘busy work’ rather than science. Another less obvious, but no less real, impediment to long-term monitoring is what might be called the loss of cultural infrastructure. A field site that might appear to administrators or bureaucrats to have ‘spartan’ or even unsafe living quarters and/or laboratory facilities, may in fact be the ‘heart’ of innovative and productive science for the project, allowing scientists to work and live together while doing research, adjacent to the research site. This certainly was the case in the early days of the Hubbard Brook Ecosystem Study where scientists lived, worked and thought together in the old Pleasant View Farmhouse at the base of the Hubbard Brook Valley. It was true in the famous, long-term studies at the Experimental Lakes Area in Ontario as well, where scientists ‘worked full-time, brainstorming late into the night’ at a remote facility in Canada (Stokstad 2008). The so-called ‘Darwin Cabin’ (Fig. 2.6), which had been used so effectively to develop and brainstorm ideas by the Trophic Cascade Project at a remote University of Notre Dame Environmental Research Center in northern Michigan in the United States, has been torn down to ‘upgrade facilities’. It will remain to be seen what impact this change will have on the very productive science stemming from this outstanding project.

Summary Any one or a combination of factors may make monitoring fail. Ineffective monitoring programs can be characterised by one or more of the following features:

50

●● ●● ●● ●● ●● ●● ●● ●● ●● ●●

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

failure to set achievable objectives lack of tractable scientific questions deficiencies in experimental design lack of agreement on which ecological entities to monitor disengagement of key parties – policy makers, resource managers and scientists inattention to data management and data quality failure to maintain data integrity over long periods excessive bureaucracy and squabbles about intellectual property a failure to recognise that people are critical infrastructure loss of the champion for a project.

These characteristics mean that it can be difficult to determine when it is appropriate for a monitoring program to continue, to cease, or when the program can be modified and made more effective during the study. Well-designed and implemented monitoring programs that do not have these characteristics may still fail for other reasons, such as a lack of funding, or an unforeseen event like a major human or natural disturbance (although the latter can sometimes provide research and monitoring opportunities). As a result of these numerous problems and the frequent lack of agreement among scientists about what to monitor, there is little wonder that decision makers may be hesitant to fund monitoring programs. With so many opportunities for failure, some readers may wonder whether any monitoring programs have been successful, a topic we turn to in the next chapter.

References Acker SA, Boetsch JR, Bivin M, Whiteaker L, Cole C, Philippi T (2015) Recent tree mortality and recruitment in mature and old-growth forests in western Washington. Forest Ecology and Management 336, 109–118. doi:10.1016/j. foreco.2014.10.008 Allen RB (1993) An appraisal of monitoring studies in South Island tussock grasslands, New Zealand. New Zealand Journal of Ecology 17, 61–63. Anon. (2012) Editorial: Error prone. Nature 487, 406. doi:10.1038/487406a Banks SC, Dujardin M, McBurney L, Blair D, Barker M, Lindenmayer DB (2011) Starting points for small mammal population recovery after wildfire: recolonisation or residual populations? Oikos 120, 26–37. doi:10.1111/j.1600-0706.2010.18765.x Banks-Leite C, Pardini R, Boscolo D, Cassano CR, Püttker T, Barros CS, Barlow J (2014) Assessing the utility of statistical adjustments for imperfect detection in tropical conservation science. Journal of Applied Ecology 51, 849–859. doi:10.1111/1365-2664.12272 Berman F, Cerf V (2013) Who will pay for public access to research data? Science 341, 616–617. doi:10.1126/science.1241625

2 – W h y m o n i t o r i n g f a il s

51

Bernhardt ES, Palmer MA, Allan JD, Alexander G, Barnas K, Brooks S, Carr J, Clayton S, Dahm C, Follstad-Shah J, Galat D, Gloss S, Goodwin P, Hart D, Hassett B, Jenkinson R, Katz S, Kondolf GM, Lake PS, Lave R, Meyer JL, O’Donnell TK, Pagano L, Powell B, Sudduth E (2005) Synthesizing US river restoration projects. Science 308, 636–637. doi:10.1126/science.1109769 Birkhead T (2014) Stormy outlook for long-term ecology studies. Nature 514, 405. doi:10.1038/514405a Blair D, McBurney LWB, Banks S, Lindenmayer DB (2016) Disturbance gradient shows logging affects plant functional groups more than fire. Ecological Applications 26, 2280–2301. doi:10.1002/eap.1369 Bouyer J, Sana Y (2007) Identification of ecological indicators for monitoring ecosystem health in the Trans-boundary W Regional Park: a pilot study. Biological Conservation 138, 73–88. doi:10.1016/j.biocon.2007.04.001 Burton AC, Neilson E, Moreira D, Ladle A, Steenweg R, Fisher JT, Bayne E, Boutin S (2015) Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. Journal of Applied Ecology 52, 675–685. doi:10.1111/1365-2664.12432 Caughlan L, Oakly KL (2001) Cost considerations for long-term ecological monitoring. Ecological Indicators 1, 123–134. doi:10.1016/S1470-160X(01)00015-2 Charles DF, Binford MW, Furlong ET, Hites RA, Mitchell MJ, Norton SA, Oldfield F, Paterson MJ, Smol JP, Uutala AJ, White JR, Whitehead DR, Wise RJ (1990) Paleoecological investigation of recent lake acidification in the Adirondack Mountains, N.Y. Journal of Paleolimnology 3, 195–241. doi:10.1007/BF00219459 Cunningham R, Lindenmayer DB (2017) Approaches to landscape scale inference and design issues. Current Landscape Ecology Reports 2, 42–50. doi:10.1007/ s40823-016-0019-4 Cunningham RB, Lindenmayer DB, Lindenmayer BD (2004) Sound recording of bird vocalisations in forests. I. Relationships between bird vocalisations and point interval counts of bird numbers - a case study in statistical modeling. Wildlife Research 31, 195–207. doi:10.1071/WR02062 Dixit SS, Smol JP, Kingston JC, Charles DF (1992) Diatoms- powerful indicators of environmental change. Environmental Science & Technology 26, 22–33. doi:10.1021/ es00025a002 Dooley JL, Bowers MA (1998) Demographic responses to habitat fragmentation: experimental tests at the landscape and patch scale. Ecology 79, 969–980. doi:10.1890/0012-9658(1998)079[0969:DRTHFE]2.0.CO;2 Field SA, O’Connor PJ, Tyre AJ, Possingham HP (2007) Making monitoring meaningful. Austral Ecology 32, 485–491. doi:10.1111/j.1442-9993.2007.01715.x Foster JR (2001) Statistical power in forest monitoring. Forest Ecology and Management 151, 211–222. doi:10.1016/S0378-1127(01)00591-6 Franklin JF, MacMahon JA (2000) Messages from a mountain. Science 288, 1183–1184. doi:10.1126/science.288.5469.1183 Franklin JF, Harmon ME, Swanson FJ (1999) Complementary roles of research and monitoring: lessons from the U.S. LTER Program and Tierra del Fuego. Paper

52

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

presented to the Symposium, ‘Toward a unified framework for inventorying and monitoring forest ecosystem resources.’ Guadalajara, Mexico, November 1998. Franklin JF, Serra-Diaz JM, Syphard AD, Regan HM (2017) Big data for forecasting the impact of global change on plant communities. Global Ecology and Biogeography Letters 26, 6–17. doi:10.1111/geb.12501 Garcia CA, Lescuyer G (2008) Monitoring, indicators and community based forest monitoring in the tropics: pretexts or red herrings? Biodiversity and Conservation 17, 1303–1317. doi:10.1007/s10531-008-9347-y Garnett S, Butchart S, Olsen P, Hoffman A (2011) Did hybridization save the Norfolk Island boobook owl Ninox novaeseelandiae undulata? Oryx 45, 500–504. doi:10.1017/S0030605311000871 Gegout J, Krizova E (2003) Comparison of indicator values of forest understorey plant species in Western Carpathians (Slovakia) and Vosges Mountains (France). Forest Ecology and Management 182, 1–11. doi:10.1016/S0378-1127(03)00068-9 Goldman C (1981) Lake Tahoe: two decades of change in a nitrogen deficient ­oligotrophic lake. Internationalen Vereinigung für Theoretische und Angewandte Limnologie 21, 45–70. Gustavsson E (2007) Land use more than 200 years ago explains current grassland plant diversity in a Swedish agricultural landscape. Biological Conservation 138, 47–59. doi:10.1016/j.biocon.2007.04.004 Harris GP (1986) Phytoplankton Ecology, Structure, Function and Fluctuation. Chapman and Hall, London. Hayward MW, Boitani L, Burrows ND, Funston PJ, Karanth KU, MacKenzie DI, Pollock KH, Yarnell RW (2015) Ecologists need robust survey designs, sampling and analytical methods. Journal of Applied Ecology 52, 286–290. doi:10.1111/1365-2664.12408 Holmes RT, Likens GE (2016) Hubbard Brook: The Story of a Forest Ecosystem. Yale University Press, New Haven, Connecticut. Joseph L, Field SA, Wilcox C, Possingham HP (2006) Presence-absence versus abundance data for monitoring threatened species. Conservation Biology 20, 1679–1687. doi:10.1111/j.1523-1739.2006.00529.x Kay G, Barton PS, Driscoll D, Cunningham S, Blanchard W, McIntyre S, Lindenmayer D (2016) Incorporating regional-scale ecological knowledge to improve the effectiveness of large-scale conservation programmes. Animal Conservation 19, 515–525. doi:10.1111/acv.12267 Kendeigh SC (1982) Bird populations in east central Illinois: fluctuations, variations, and development over a half-century. Illinois Biological Monographs 52, 1–152. Kingsford RT, Halse SA, Porter JL (2008) Aerial surveys of waterbirds – assessing wetland condition. Final report to the National Land & Water Resources Audit, University of New South Wales, Sydney. Kirby MF, Smith AJ, Barry J, Katsiadaki I, Lyons B, Scott AP (2006) Differential sensitivity of Flounder (Platichthys flesus) in response to oestrogenic chemical exposure: an issue for the design and interpretation of monitoring and research

2 – W h y m o n i t o r i n g f a il s

53

programmes. Marine Environmental Research 62, 315–325. doi:10.1016/j. marenvres.2006.04.068 Kleijn D, Sutherland WJ (2003) How effective are European agri-environment schemes in conserving and promoting biodiversity? Journal of Applied Ecology 40, 947–969. doi:10.1111/j.1365-2664.2003.00868.x Krebs CJ (1991) The experimental paradigm and long-term population studies. The Ibis 133, 2–8. Laurance W, Luizao R (2007) Driving a wedge into the Amazon. Nature 448, 409–410. doi:10.1038/448409a Legg CJ, Nagy L (2006) Why most conservation monitoring is, but need not be, a waste of time. Journal of Environmental Management 78, 194–199. doi:10.1016/j. jenvman.2005.04.016 Lehmkuhl JF, Peffer RD, O’Connell MA (2008) Riparain and upland small mammals on the east slope of the Cascade Range, Washington. Northwest Science 82, 94–107. doi:10.3955/0029-344X-82.2.94 Leonelli S (2016) Data-centric Biology. University of Chicago Press, Chicago. Likens GE (Ed.) (1985) An Ecosystem Approach to Aquatic Ecology: Mirror Lake and its Environment. Springer-Verlag, New York. Likens GE, Buso DC (2010a) Long-term changes in streamwater chemistry following disturbance in the Hubbard Brook Experimental Forest, USA. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie 30, 1577–1581. Likens GE, Buso DC (2010b) Salinization of Mirror Lake by road salt. Water, Air, and Soil Pollution 205, 205–214. doi:10.1007/s11270-009-0066-0 Lindenmayer DB (2009a) Forest Pattern and Ecological Process: A Synthesis of 25 Years of Research. CSIRO Publishing, Melbourne. Lindenmayer DB (2009b) Large-Scale Landscape Experiments. Lessons from Tumut. Cambridge University Press, Cambridge. Lindenmayer DB, Cunningham RB (1997) Patterns of co-occurrence among arboreal marsupials in the forests of central Victoria, southeastern Australia. Australian Journal of Ecology 22, 340–346. doi:10.1111/j.1442-9993.1997.tb00680.x Lindenmayer DB, Likens GE (2009) Adaptive monitoring – a new paradigm for long-term research and monitoring. Trends in Ecology & Evolution 24, 482–486. doi:10.1016/j.tree.2009.03.005 Lindenmayer DB, Likens GE (2011a) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/s10021-010-9394-6 Lindenmayer DB, Likens GE (2011b) Effective monitoring of agriculture. Journal of Environmental Monitoring 13, 1559–1563. doi:10.1039/c0em00691b Lindenmayer DB, Likens GE (2013) Don’t do big-data science backwards. Nature 499, 284. doi:10.1038/499284d Lindenmayer DB, Likens GE, Franklin JF (2010) Rapid responses to facilitate ecological discoveries from major disturbances. Frontiers in Ecology and the Environment 8, 527–532. doi:10.1890/090184

54

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Lindenmayer DB, Zammit C, Attwood SA, Burns E, Shepherd CL, Kay GE, Wood J (2012) A novel and cost-effective monitoring approach for outcomes in an Australian biodiversity conservation incentive program. PLoS One 7, e50872. doi:10.1371/journal.pone.0050872 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks S, Driscoll D, Smith A, Gill AM (2013a) Fire severity and landscape context effects on arboreal marsupials. Biological Conservation 167, 137–148. doi:10.1016/j.biocon.2013.07.028 Lindenmayer DB, Piggott M, Wintle B (2013b) Counting the books while the library burns: why conservation monitoring programs need a plan for action. Frontiers in Ecology and the Environment 11, 549–555. doi:10.1890/120220 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks SC, Driscoll DA, Smith A, Gill AM (2014a) Complex responses of birds to landscape-level fire extent, fire severity and environmental drivers. Diversity & Distributions 20, 467–477. doi:10.1111/ddi.12172 Lindenmayer DB, Burns E, Thurgate N, Lowe A (Eds) (2014b) Biodiversity and Environmental Change: Monitoring, Challenges and Direction. CSIRO Publishing, Melbourne. Lindenmayer DB, Blair D, McBurney L, Banks S (2015a) Mountain Ash: Fire, Logging and the Future of Victoria’s Giant Forests. CSIRO Publishing, Melbourne. Lindenmayer DB, Burns EL, Tennant P, Dickman CR, Green PT, Keith DA, Metcalfe DJ, Russell-Smith J, Wardle GM, Williams D, Bossard K, deLacey C, Hanigan I, Bull CM, Gillespie G, Hobbs RJ, Krebs CJ, Likens GE, Porter J, Vardon M (2015b) Contemplating the future: acting now on long-term monitoring to answer 2050’s questions. Austral Ecology 40, 213–224. doi:10.1111/aec.12207 Lindenmayer DB, Pierson J, Barton P, Beger M, Branquinho C, Calhoun A, Caro T, Greig H, Gross J, Heino J, Hunter M, Lane P, Longo C, Martin K, McDowell WH, Mellin C, Salo H, Tulloch A, Westgate M (2015c) A new framework for selecting environmental surrogates. The Science of the Total Environment 538, 1029–1038. doi:10.1016/j.scitotenv.2015.08.056 Mac Nally R (1997) Monitoring forest bird communities for impact assessment – the influence of sampling intensity and spatial scale. Biological Conservation 82, 355–367. doi:10.1016/S0006-3207(97)00050-5 Maher W, Norris RH (1990) Water quality assessment programs in Australia: deciding what to measure and how and where to use bioindicators. Environmental Monitoring and Assessment 14, 115–130. doi:10.1007/BF00677912 Maldonado C, Molina CI, Zizka A, Persson C, Taylor CM, Alban J, Chilquillo E, Ronsted N, Antonelli A (2015) Estimating species diversity and distribution in the ear of Big Data: to what extent can we trust public databases? Global Ecology and Biogeography 24, 973–984. doi:10.1111/geb.12326 Mariogómez I, Soto M, Cancio I, Orbea A, Garmedndia L, Cajaraville MP (2006) Cell and tissue markers in mussel, histopathology in hake and anchovy from Bay of Biscay after Prestige Oil spill (monitoring campaign 2003). Marine Pollution Bulletin 53, 287–304. doi:10.1016/j.marpolbul.2005.09.026

2 – W h y m o n i t o r i n g f a il s

55

Martin J, Kitchens WM, Hines JE (2007) Importance of well-designed monitoring programs for the conservation of endangered species: case study of the Snail Kite. Conservation Biology 21, 472–481. doi:10.1111/j.1523-1739.2006.00613.x Martins S, Sanderson JG, Silva-Junior J (2007) Monitoring mammals in the Caxiuana National Forest, Brazil – first results from the Tropical Ecology, Assessment and Monitoring (TEAM) program. Biodiversity and Conservation 16, 857–870. doi:10.1007/s10531-006-9094-x McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, Possingham HP (2010) Monitoring does not always count. Trends in Ecology & Evolution 25, 547–550. doi:10.1016/j.tree.2010.07.002 McDowell WH (2015) NEON and STREON: opportunities and challenges for the aquatic sciences. Freshwater Science 34, 386–391. doi:10.1086/679489 McNutt M, Lehnert K, Hanson B, Nosek BA, Ellison AM, King JL (2016) Liberating field science samples and data. Science 351, 1024–1026. doi:10.1126/science. aad7048 Meek P, Fleming P, Ballard G, Banks P, Claridge A, Sanderson J, Swann D (Eds) (2014) Camera Trapping: Wildlife Management and Research. CSIRO Publishing, Melbourne. Milledge DR, Palmer CL, Nelson JL (1991) ‘Barometers of change’: the distribution of large owls and gliders in Mountain Ash forests of the Victorian Central Highlands and their potential as management indicators. In Conservation of Australia’s Forest Fauna. (Ed. D Lunney) pp. 53–65. Royal Zoological Society of NSW, Sydney. Millington PJ, Walker KF (1983) Australian freshwater mussel Velesunio ambiguus (Phillipi) as a biological indicator for zinc, iron and manganese. Australian Journal of Marine and Freshwater Research 34, 873–892. doi:10.1071/MF9830873 Mortimer CH (1981) The Lake Michigan Pollution Case. Sea Grant Institute and Center for Great Lakes Studies. University of Wisconsin, Wisconsin, USA. National Research Council (1995) Finding the Forest in the Trees: The Challenge of Combining Diverse Environmental Data. National Research Council, Washington, DC. Nichols JD, Williams BK (2006) Monitoring for conservation. Trends in Ecology & Evolution 21, 668–673. doi:10.1016/j.tree.2006.08.007 Nisbet E (2007) Cinderalla science. Nature 450, 789–790. doi:10.1038/450789a Norman J, Olsen P, Christidis L (1998) Molecular genetics confirms taxonomic affinities of the endangered Norfolk Island Boobook owl Ninox novaeseelandiae undulata. Biological Conservation 86, 33–36. Norton DA (1996) Monitoring biodiversity in New Zealand’s terrestrial ecosystems. In Papers from a seminar series on biodiversity. (Eds B McFadgen, S Simpson) pp. 19–41. Department of Conservation, Wellington, New Zealand. Olsen P (1996) Re-establishment of an endangered subspecies: the Norfolk Island Boobook Owl Ninox novaeseelandiae undulata. Bird Conservation International 6, 63–80. Orians GH (1986) The place of science in environmental problem-solving. Environment 28, 12–41. doi:10.1080/00139157.1986.9928824

56

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Pearce J, Venier L (2005) Small mammals as bioindicators of sustainable forest management. Forest Ecology and Management 208, 153–175. doi:10.1016/j. foreco.2004.11.024 Pellet J, Schmidt BR (2005) Monitoring distribution using call surveys: estimating occupancy, detection probabilities and inferring absence. Biological Conservation 123, 27–35. doi:10.1016/j.biocon.2004.10.005 Pierson JC, Mortelliti A, Barton PS, Lane PW, Lindenmayer DB (2016) Evaluating the effectiveness of overstory cover as a surrogate for bird community diversity and population trends. Ecological Indicators 61, 790–798. doi:10.1016/j.ecolind. 2015.10.031 Potapova M, Charles DF (2007) Diatom metrics for monitoring eutrophication in rivers of the United States. Ecological Indicators 7, 48–70. doi:10.1016/j. ecolind.2005.10.001 Reed JM, Blaustein AR (1995) Assessment of ‘nondeclining’ amphibian populations using power analysis. Conservation Biology 9, 1299–1300. doi:10.1046/j.1523-1739.1995.9051295.x-i1 Reichman OJ, Jones MB, Schildhauer MP (2011) Challenges and opportunities of open data in ecology. Science 331, 703–705. doi:10.1126/science.1197962 Roberts KA (1991) Field monitoring: confessions of an addict. In Monitoring for Conservation and Ecology. (Ed. FB Goldsmith) pp. 179–212. Chapman and Hall, London. Rodriguez-Estrella R, Donazar JD, Hiraldo F (1998) Raptors as indicators of environmental change in the scrub habitat of Baja California Sur, Mexico. Conservation Biology 12, 921–925. doi:10.1046/j.1523-1739.1998.97044.x Rosenberry DO, Winter TC, Buso DC, Likens GE (2007) Comparison of 15 evaporation methods applied to a small mountain lake in the northeastern USA. Journal of Hydrology 340, 149–166. doi:10.1016/j.jhydrol.2007.03.018 Roth T, Weber D (2008) Top predators as indicators for species richness? Prey species are just as useful. Journal of Applied Ecology 45, 987–991. doi:10.1111/j.1365-2664.2007.01435.x Russell-Smith J, Whitehead PJ, Cook GD, Hoare JL (2003) Response of Eucalyptusdominated savanna to frequent fires: lessons from Munmarlary 1973–1996. Ecological Monographs 73, 349–375. doi:10.1890/01-4021 Sachs J, Remans R, Smukler S, Winowiecki L, Andelman SJ, Cassman KG, Castle D, DeFries R, Denning G, Fanzo J, Jackson LE, Leemans R, Leemans J, Milder JC, Naeem S, Nziguheba G, Palm CA, Reganold JP, Richter DD, Scherr SJ, Sircely J, Sullivan C, Tomich TP, Sanchez PA (2010) Monitoring the world’s agriculture. Nature 466, 558–560. doi:10.1038/466558a Sato C, Wood J, Stein JA, Crane M, Okada S, Michael D, Kay G, Florance D, Seddon J, Gibbons P, Lindenmayer DB (2016) Natural tree regeneration in agricultural landscapes: the implications of intensification. Agriculture, Ecosystems & Environment 230, 98–104. doi:10.1016/j.agee.2016.05.036

2 – W h y m o n i t o r i n g f a il s

57

Schelske CL (1988) Historic trends in Lake Michigan silica concentrations. International Review of Hydrobiology 73, 559–591. doi:10.1002/iroh.19880730506 Schelske CL, Stoermer EF (1971) Eutrophication, silica depletion, and predicted changes in algal quality in Lake Michigan. Science 173, 423–424. doi:10.1126/ science.173.3995.423 Schelske CL, Stoermer EF, Conley DJ, Robbins JA, Glover RM (1983) Early eutrophication in the lower Great Lakes; new evidence from biogenic silica in sediments. Science 222, 320–322. doi:10.1126/science.222.4621.320 Schindler DW, Mills KH, Malley DF, Findlay DL, Shearer JA, Davies IJ, Turner MA, Linsey GA, Cruikshank DR (1985) Long-term ecosystem stress: the effects of years of experimental acidification on a small lake. Science 228, 1395–1401. doi:10.1126/ science.228.4706.1395 Sergio F, Caro T, Brown D, Clucas B, Hunter J, Ketchum J, McHugh K, Hiraldo F (2008) Top predators as conservation tools: ecological rationale, assumptions and efficacy. Annual Review of Ecology Evolution and Systematics 39, 1–19. doi:10.1146/ annurev.ecolsys.39.110707.173545 Shanklin J (2010) Reflections on the ozone hole. Nature 465, 34–35. doi:10.1038/465034a Shapiro J, Swain EB (1983) Lessons from the silica “decline” in Lake Michigan. Science 221, 457–459. doi:10.1126/science.221.4609.457 Silvertown JPP, Johnston E, Edwards G, Heard M, Biss PM (2006) The Park Grass Experiment 1856–2006: its contribution to ecology. Journal of Ecology 94, 801–814. doi:10.1111/j.1365-2745.2006.01145.x Stanton J (2014) Present-day risk assessment would have predicted the extinction of the Passenger Pigeon (Ectopistes migratorius). Biological Conservation 180, 11–20. doi:10.1016/j.biocon.2014.09.023 Stirzaker R (2010) Out of the Scientist’s Garden: A Story of Water and Food. CSIRO Publishing, Melbourne. Stokstad E (2008) Canada’s experimental lakes. Science 322, 1316–1319. doi:10.1126/ science.322.5906.1316 Strayer DL (1999) Statistical power of presence-absence data to detect population declines. Conservation Biology 13, 1034–1038. doi:10.1046/j.1523-1739.1999.98143.x Strayer DL, Glitzenstein JS, Jones C, Kolasa J, Likens GE, McDonnell M, Parker GG, Pickett STA (1986) Long-term ecological studies: An illustrated account of their design, operation, and importance to ecology. Institute of Ecosystem Studies, Millbrook, New York. Taylor B, Kremsater L, Ellis R (1998) Adaptive management of forests in British Columbia. British Columbia Ministry of Forests – Forest Practices Branch, Victoria, British Columbia. The Heinz Center (2008) The State of the Nation’s Ecosystems 2008. The H. John Heinz III Centre for Science, Economics and the Environment and Island Press, Washington, DC.

58

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Thiollay JM (1996) Rain forest communities in Sumatra: the conservation value of traditional agroforests. In Raptors in Human Landscapes. (Eds DM Bird, D Varland, JJ Negro) pp. 245–261. Academic Press, London. Timofeyev MA, Silow EA, Mackay AW, Moore MV, Likens GE (2016) Safeguarding the world’s largest lake. Nature 538, 41. doi:10.1038/538041a Turner MG, Romme WH, Tinker DB (2003) Surprises and lessons from the 1988 Yellowstone fires. Frontiers in Ecology and the Environment 1, 351–358. doi:10.1890/1540-9295(2003)001[0351:SALFTY]2.0.CO;2 Walker KF (1981) Ecology of freshwater mussels in the River Murray. Australian Government Publishing Service, Canberra. Walsh PD, White LJ (1999) What it will take to monitor forest elephant populations. Conservation Biology 13, 1194–1202. doi:10.1046/j.1523-1739.1999.98148.x Walters M, Scholes RJ (Eds) (2017) The Geo Handbook on Biodiversity Observation. Springer Open, Switzerland. Ward RC, Loftis JC, McBride GB (1986) The “data-rich but information-poor” syndrome in water quality monitoring. Environmental Management 10, 291–297. doi:10.1007/BF01867251 Watson I, Novelly P (2004) Making the biodiversity monitoring system sustainable: design issues for large-scale monitoring systems. Austral Ecology 29, 16–30. doi:10.1111/j.1442-9993.2004.01350.x Welsh AH, Cunningham RB, Chambers RL (2000) Methodology for estimating the abundance of rare animals: seabird nesting on North East Herald Cay. Biometrics 56, 22–30. doi:10.1111/j.0006-341X.2000.00022.x Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899 Whelan RJ (1995) The Ecology of Fire. Cambridge University Press, Cambridge. Woinarski JCZ, Garnett ST, Legge SM, Lindenmayer DB (2017) The contribution of policy, law, management, research and advocacy failings to the recent extinctions of three Australian vertebrate species. Conservation Biology 31, 13–23. doi:10.1111/ cobi.12852 Yoccoz NG, Nichols JD, Boulinier T (2001) Monitoring of biological diversity in space and time. Trends in Ecology & Evolution 16, 446–453. doi:10.1016/ S0169-5347(01)02205-4 Youngentob K, Likens GE, Williams JE, Lindenmayer DB (2013) A survey of long-term terrestrial ecology studies in Australia. Austral Ecology 38, 365–373. doi:10.1111/j.1442-9993.2012.02421.x Zeide B (1994) Big projects, big problems. Environmental Monitoring and Assessment 33, 115–133. doi:10.1007/BF00548593 Zhan X, Li MY, Zhang Z, Goossens B, Chen Y, Wang H, Bruford M, Wei F (2006) Molecular censusing doubles giant panda population estimate in a key nature reserve. Current Biology 16, R451–R452. doi:10.1016/j.cub.2006.05.042

3 What makes long-term monitoring effective?

In Chapter 2 we outlined some of the characteristics of ineffective monitoring programs and highlighted why some fail. As an antidote to that rather depressing discussion, in this chapter we describe the characteristics of effective and successful monitoring programs. We then discuss (and provide examples of) the adaptive monitoring paradigm, which collects together many of these characteristics in a broad framework to guide successful monitoring programs (Lindenmayer and Likens 2009). Many of the attributes of successful monitoring programs will be apparent in the examples of successful case studies that appear in Chapter 4. Ecologists, like all scientists, learn by doing (Walters and Holling 1990; Hall and Fleischman 2010) and that is the theme of the next two chapters.

Characteristics of effective monitoring programs If we ask, ‘what are the key characteristics of effective monitoring programs?’, we can isolate what appear to be key points, but we should be mindful that the ‘secrets’ to success are all strongly interrelated. There are several ways of gaining long-term information, including empirical or natural history studies, retrospective studies, modelling studies, space-for-time

60

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

studies, experimental manipulation studies and direct long-term monitoring (Likens 1989; Cunningham and Lindenmayer 2017). Each of these approaches can provide important insights into how systems work, but a combination of these approaches is even more powerful, giving an integrated view of long-term phenomena. Cost, expertise and time often limit the use of multiple approaches, but multiple approaches should remain a goal, even when it is not always possible to achieve the integration they could provide. Long-term records of ecological phenomena are relatively rare (Youngentob et al. 2013) and difficult to obtain, but they provide unique insights into how an ecosystem works (Lindenmayer et al. 2012a; see Chapter 1). As such, these records are a critical component of overall ecological inquiry (Muller et al. 2011). Longterm records are usually developed through monitoring of ecosystem parameters (optimally guided by questions, not mindless collection of data). The integrity and application of quality assurance/quality control protocols are pivotal to the success of long-term research (see Buso et al. 2000; Lovett et al. 2007). Without high-quality assurance/quality control, long-term records can be seriously compromised. Exemplary examples of monitoring include the pioneering efforts at Rothamsted, England (see Chapter 4), the US Weather Service, the US Geological Survey, and British Breeding Bird Survey. In the past three decades in the United States, there have been major attempts to initiate and sustain long-term research and monitoring. Examples include Long-Term Research in Environmental Biology (LTREB), Long-Term Ecological Research (LTER, although it is not a monitoring network per se), Atmospheric Integrated Research Monitoring Network (AIRMoN), Clean Air Status and Trends Network (CASTNeT), Canadian Air and Precipitation Monitoring Network (CAPMoN), US Forest Inventory Analysis (FIA), and the Temporally Integrated Monitoring of Ecosystems (TIME) programs. In spite of the documented value of such efforts (Likens 1989), they often are supported by fickle finances, and thus are difficult to maintain over the truly long term (Lindenmayer et al. 2017). As a recent and most unfortunate example, funding for the Australian Long-Term Ecological Research Network (LTERN) was terminated and the array of plots in this initiative will be decommissioned by the end of 2017.

Good questions and evolving questions Posing good questions lies at the heart of good science and hence is essential to effective monitoring (Nichols and Williams 2006). This is far from being a trivial task – 90% of a problem is often defining the problem. Some authors argue that ecologists and resource managers have been poor at problem definition and objective setting (Peters 1991). Good question setting must result in quantifiable objectives that offer unambiguous signposts for measuring progress. That is, posing good questions

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

61

Transformed monitoring Altered (adaptive) management

Existing monitoring program

New New policy policy conceptual questions model Common ground New New research research conceptual questions model

Transformed (adaptive) monitoring program

New information about management

Altered (adaptive) management Transformed monitoring TRENDS in Ecology & Evolution

Fig. 3.1.  Interactions between management-relevant adaptive monitoring and policy development

means that it should be possible a priori to write down how data that are proposed to be gathered in a monitoring program can be used to test well-formulated hypotheses (e.g. Efford and Dawson 2012; Majer et al. 2013; Dickman et al. 2014; Hayward et al. 2015). As outlined later in this chapter, setting good questions requires a well-developed partnership among scientists, statisticians, resource managers and policy makers (Lindenmayer et al. 2013b) (Fig. 3.1). Thus, good questions must be scientifically tractable and test real policy and resource management options (Walters 1986). For a wide range of reasons, the key questions being addressed in a project may change or evolve during a long-term study or monitoring program. Such changes are challenging because they require a nimble approach to allow the changes to be made when required (Ringold et al. 1996) but without breaching the integrity of the overall monitoring program. As an example, the long-term monitoring program for arboreal marsupials in the wet forests of Victoria (south-eastern Australia) was initially focused on the critically endangered species, Leadbeater’s possum (Lindenmayer et al. 2003). Although this is a high profile species of conservation concern, it is also relatively rare and it can be difficult to gather large quantities of data on it. After the monitoring program for the species had commenced, officials from the Victorian Government requested that the focus of work be altered to include the abundance of all species of arboreal marsupials. A rotating, site-selection process was sufficiently flexible to enable this suggested change to be accommodated (Lindenmayer et al. 2003). Subsequent demands from

62

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

the Victorian Government changed the focus of the monitoring program back to Leadbeater’s possum, in part because of the demands of the State of Victoria’s Recovery Plan for the species. Again, the statistical methodology employed to guide site selection and survey allowed this change in focus to take place without breaching the integrity of the overall monitoring program (see Chapter 4). Importantly, the monitoring program was carefully focused on Leadbeater’s possum and other species of arboreal marsupials and sought to answer key questions about the population dynamics of these species. At no stage was it ever assumed that either Leadbeater’s possum or any other species was valid as an indicator species (Lindenmayer and Cunningham 1997; Lindenmayer et al. 2014a) (see Chapter 2). Finally, in 2009, major wildfires in the area demanded yet further changes to the long-term monitoring program in which new questions were posed as part of quantifying post-fire ecological recovery. Accordingly, the rotating-site sampling strategy had to accommodate comparisons of burned and unburned sites (Lindenmayer et al. 2013a, 2015b).

The use of a conceptual model A conceptual model of the ecosystem or particular entity to be targeted for monitoring is one way to help guide the development of the question(s) to be addressed (Woodward et al. 1999). Such a model can guide the development of a priori predictions about how an ecosystem or other kind of entity (e.g. a population of a species) might behave, for example, in response to a human intervention (e.g. clearcutting of a catchment or prescribed burning of a patch). A conceptual model, developed at the beginning of a study, forces the collection of ideas to formulate theory about how an ecosystem or target entity works, and helps to ensure that all the relevant components are captured in the project design. A conceptual model needs to capture the inputs and outputs for an ecosystem and thereby show clearly the connections of that particular ecosystem with the remainder of the biosphere. Those abiotic and biotic inputs and outputs then highlight the critical attributes of the ecosystem that need to be managed. These inputs and outputs identify the clear targets for management of environmental problems (see case studies in Chapter 4). By understanding the inputs and outputs within a conceptual model framework, it is possible to understand the mechanisms for change in an ecosystem and what responses might occur as a consequence of management interventions. This approach is a powerful way for making informed predictions and subsequently testing them as part of ecosystem management. A conceptual model needs to be able to guide continued and ongoing research and thinking. It becomes a focal point for discussions among partners – scientists, managers and policy makers – about how an ecosystem might be managed or protected. Such discussions are particularly important when dealing with complex

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

Growing vegetation

Benthic & suspended detritus

Standing living & dead trees

Litter & soil organic matter

Riparian vegetation

Coarse woody debris

Soil inorganic matter & bedrock

Inorganic substrate (bedrock, sediment)

GROWTH

AQUATIC

63

Soil nutrients and water

DECAY

SOILS

Fig. 3.2.  The conceptual model for guiding research at Warra Long-Term Ecological Research site in southern Tasmania (redrawn from Grove 2007)

natural systems and bringing people together to generate a common view about an appropriate way forward. We suggest that such discussions are critical for successful long-term monitoring programs. Conceptual models can fail to guide long-term research and monitoring when they are either too detailed (‘the devil is in the detail’), too abstract or vague, or unsuitable for answering specific or new questions (like having the wrong vessel in a laboratory experiment), or not being particularly pertinent to the research site. Many such conceptual models fail to guide research or be used, particularly regarding adaptive monitoring issues, in long-term studies. We show in Fig. 3.2 a conceptual model of the dynamics of coarse woody debris in the wet forests of Tasmania. It was developed to guide a series of major studies that are taking place at the Warra Long-Term Ecological Research site (Grove 2007). The model is quite complex. As we discuss in Chapter 4, conceptual models are often most useful for developing tractable questions and subsequent long-term studies when they are relatively simple. An example of a more simple and tractable conceptual model is the one that has been used successfully for decades to guide thinking and research for the Hubbard Brook Ecosystem Study in the White Mountains of New Hampshire in the United States (Bormann and Likens 1967) (Fig. 3.3).

Selection of appropriate entities to measure Progress in many monitoring programs is hamstrung by wrangles over what to monitor. In Chapter 2, we highlighted two approaches that are commonly but often unsuccessfully employed to tackle this problem. These were attempts to monitor (almost) everything or to monitor so-called indicator species thought to be

Soil & bedrock

Colonisation by mycorrhizae

Geology & landform

Photosynthesis

Insolation

Climate & weather

Microclimate

Flower visitors

Flowering & gamete production

AQUATIC MODULE

Carbon assimilation rate

Transpiration rate

Nutrient & water uptake

Plant growth

Growth rate, NPP

Maintenance & alteration of microclimate

AQUATIC MODULE

SOILS MODULE

DECAY MODULE

Leaf & litterfall

Vegetation succession & stand development

Growing vegetation

Fire, wind, drought & other disturbance

Herbivory

Herbivory rate

Successional pathways & dynamics

Disturbance type, frequency & intensity

Climate & weather

Herbivores

DECAY MODULE

Ageing, competition, climatic stress, pests & diseases, root & butt rot

Dependent biodiversity

Survivorship

Microhabitat type & range

Infection rate

Tree & stand structure

GROWTH MODULE

Fig. 3.3.  The conceptual model for guiding biogeochemical research and monitoring for the Hubbard Brook Ecosystem Study (redrawn from Bormann and Likens 1967). Notably, a new conceptual was developed for the Hubbard Brook Ecosystem Study, but it has been little used (see Chapter 4).

Soil type

Provision of physical structure for colonisation

Mortality & stocking rate

Germination & establishment

Herbivory rate

Dispersal agent

Dispersal ability

Pollinators

Pollination & fertilisation

Seasonality, phenology

Pollination rate

Seed & spore dispersal

Preditors

Fruiting & spore production

Microhabitat & microclimate

SOILS MODULE

Seed & spore rain

Predition rate

Fruit & seed predation

Fecundity

64 E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

65

surrogates for other entities (e.g. effects of a pollutant; see Spellerberg 1994) or for other groups of biota or ecosystem processes (reviewed by Caro 2010; Lindenmayer et al. 2015a). We suggest the problems of laundry lists and indicator species can be avoided by carefully crafting questions at the onset of a monitoring program, using a well-conceived conceptual model to help with thinking about a particular ecosystem and make predictions about ecosystem behaviour and response (see previous section). These key steps will help identify those entities most appropriate for monitoring. Such a focused scientific approach obviates the need to measure a vast array of things and ensures that a subset of entities can be monitored well, rather than vice-versa. In addition, making direct measures of targeted entities avoids the need to assume that those entities are surrogates or indicators of other entities (Lindenmayer and Likens 2011). The selection of appropriate entities to measure in a monitoring program may include not only particular outcomes but also factors that might influence those outcomes. For example, a monitoring program might be designed to assess the effectiveness of restoration efforts for elements of farmland biodiversity such as native bird species. Appropriate entities to measure would certainly include the number and array of different bird species and how it changes over time as a result of a management intervention like the establishment of planted and restored vegetation. However, to explain such temporal changes, the monitoring program also may need to measure attributes that characterise the management intervention (such as the size and shape of the restored plantings) as well as the inputs to the intervention like the cost of establishing fences and controlling the intensity of grazing by domestic livestock (Lindenmayer and Gibbons 2012). In essence, this approach to design seeks not only to document patterns of temporal change but also to uncover the factors influencing the patterns that have been documented, including aspects of cost-effectiveness related to management effectiveness.

Good design A monitoring program can be based on excellent questions that are developed using a conceptual model and well-focused on an appropriate set of attributes to measure, but still fail when it is poorly designed. Good statistical design is a critical component of any successful monitoring program and the methods employed must be sensitive enough to measure and detect change that is both statistically and biologically meaningful (Reynolds et al. 2011). Good design is also efficient (see Box 3.1). Martin et al. (2007) and Hayward et al. (2015) correctly believe that proponents of monitoring programs should spend more time getting their study design right. For example, some of the strongest designs to guide a monitoring program will often be those where there are contrasting management interventions that enable strong inferences to be made about both how and why an ecosystem or

66

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 3.1. Rolls Royce or Ford? Determining monitoring project design Some people have argued that scientists are prone to over-engineering monitoring projects and including expensive design features that are not necessary and too costly. A common question is: Can you design something that is less expensive – a Ford rather than a Rolls Royce? The answer to this question lies in the questions that need to be addressed and the objectives of the monitoring. Good statistical design relates to efficiency. That is, sufficient replication and controls to answer the problem being tackled. It is inefficient to have too many replicates of a given treatment, either because it adds unnecessary cost or leads to an opportunity cost (e.g. the chance to do two experiments rather than one over-replicated experiment) (Cunningham and Lindenmayer 2017). At issue here then is not whether a Rolls Royce or a Ford is needed. Rather, it is what is an efficient design to answer the questions that are being posed? This highlights the importance of statistical science in guiding the design of robust monitoring programs.

other entity like a population has changed either spatially, temporally, or spatiotemporally (Cunningham and Lindenmayer 2017). We contend that professional statistical advice should be at the core of most of the key phases of a monitoring program. This recommendation is made because good study design is an inherently statistical process (Cunningham and Lindenmayer 2017; see Box 3.2) and many key approaches to experimental design and the analysis of environmental data are not those which many scientists, resource managers and policy makers may know how to do. Good study design is invariably linked with expert statistical analyses of data once data are gathered. In fact, the analysis and interpretation of results from monitoring programs can be extremely challenging (Nichols and Williams 2006). For example, separating long-term population declines from annual or cyclical fluctuations in population size is a non-trivial task (Lindenmayer et al. 2014b; see Fig. 3.4). It is particularly problematic for some rare species for which it may be difficult to determine if they were declining (Taylor and Gerrodette 1993; Martin et al. 2007) or for populations of species which can exhibit rapid short-term fluctuations in population size, such as after disturbances like logging (Fedrowitz et al. 2014) or fire (Swanson et al. 2011; Pulsford et al. 2016). The challenges of good design, coupled with the rigour of subsequent statistical analyses of high-quality environmental data, emphasise the fact that monitoring programs can (and should) be good science (Nichols and Williams 2006; Lindenmayer et al. 2012a). Conquering these challenges means that the distinction between research and monitoring also should be blurred (Franklin et al. 1999). Following the suggestion of Walters (1997), we support the idea that researchers

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

67

Fig. 3.4.  Surveys of the red-backed salamander (Plethodon cinereus) provide a good example of a non-standard monitoring program that leads to changes in inference. Traditionally, the size of terrestrial salamander populations in forest ecosystems has been determined by destructive sampling of downed, rotting logs, raking apart deep accumulations of forest litter, examining under downed logs, under rocks or under artificial cover objects, such as pieces of plywood or cardboard placed on the surface of the forest floor. Burton and Likens (1975) found that much higher and more accurate population sizes could be obtained by assaying salamander abundances directly on dark, rainy nights when the litter and herbaceous layers were soaked with water. The red-backed salamander is the most abundant terrestrial salamander in the Hubbard Brook Experimental Forest. The species does not have lungs and breathes through its skin. It comes to the surface of the forest floor on rainy nights to forage. In early 1970, a population size of ~520 salamanders/ha was obtained using traditional methods, but ~2600 salamanders/ha were found along transects on rainy nights using powerful flashlights/headlamps. The biomass of the red-backed salamander was about twice that of birds and about equal to small mammals in the Hubbard Brook forest. Currently, the number of animals at Hubbard Brook is ~30% of what it was in 1970 (Holmes and Likens 2016). Photograph by F Kynd (CC BY 2.0).

should be attracted to the prospect of conducting large-scale, well-designed and replicated monitoring studies of direct relevance to natural resource management.

Well-developed partnerships The previous section emphasised the role of statistical science in the experimental design, data analysis and data interpretation phases of monitoring programs. Most successful monitoring programs are built on partnerships between people with different but complementary sets of skills (Hong and Page 2004; Sergeant et al. 2012; Lindenmayer et al. 2013b). These include scientists, statisticians, policy makers and resource managers who may be from government and nongovernment organisations, universities, research institutions and other

68

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 3.2. A novel and non-standard overlapping and rotating statistical design for increasing inference from monitoring programs The majority of monitoring programs are based on traditional visit–revisit protocols. That is, a specified number of sites is visited at a given frequency (e.g. annually). However, this design can be limiting when it is difficult to measure more than a small number of sites during any given time interval (e.g. a field season). This limitation is the case for a major long-term monitoring program of arboreal marsupials in the wet forests of Victoria in Australia. For that work, all large trees on a given one-hectare site must be watched simultaneously at dusk by well-trained volunteers to generate an accurate count of the numbers of each species of arboreal marsupial on that site (Lindenmayer et al. 1991). This is a time-consuming and labour-intensive method, but no others provide counts with an appropriate level of accuracy (Smith et al. 1989; Lindenmayer et al. 2011b). Weather conditions, logistical factors and financial constraints mean that no more than 40 sites can typically be surveyed in any given year. A monitoring study of 40 sites would be limited to a subset of the environmental conditions relevant to arboreal marsupials in the study area. Hence, levels of inference from the study would be restrictive. These problems were solved by adapting a novel and non-standard overlapping and rotating statistical design from a long-term monitoring program of seabirds in the Coral Sea off the coast of northern Queensland. The monitoring program was based on a retrospective approach to re-survey populations of arboreal marsupials on sites first surveyed in 1983. A pool of 161 sites was established and each site has a history of repeated field surveys over the past two decades. All 161 sites were re-surveyed in 1997–98 and a randomly selected subset of between 30 and 40 of the 161 sites has been re-sampled for animals between 1999 and 2016 (Lindenmayer et al. 2003, 2012b, 2013a). The probability of site selection was based on the count of the number of animals recorded on those sites in the previous surveys, including the census years in 1997 and 1998. Overall, the monitoring design called for ~80% of all sampled sites to remain the same from one year to the following year with an emphasis on sites with high numbers of animals. Sites with few animals typically were surveyed every three to four years (Lindenmayer et al. 2003). A further complication in the design of the monitoring program has been the impact of a major wildfire in 2009 which burned ~50% of the long-term sites at varying severity. The design of the monitoring program, including the rotating sampling design, was further adapted to incorporate contrasts between burned and unburned sites and to quantify the response to (and speed of recovery after) the 2009 wildfires (e.g. Lindenmayer et al. 2013a). The monitoring program in the Victorian forests has been specifically designed to provide strong statistical inferences about population trends in arboreal marsupials. Year-to-year partial sampling has allowed short-term temporal fluctuations in population dynamics to be separated from long-term trends such as population declines (Welsh et al. 2000) as well as long-term responses to natural disturbance. A key advantage of the non-standard overlapping and rotating

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

69

sampling approach was that it enabled a far greater number and range of sites to be included in the monitoring program than if a traditional annual sample and re-sample method were used. This approach greatly broadened the kinds of inferences that could be made from the monitoring program. The statistical advantages of non-standard, overlapping, rotating designs for monitoring programs are further discussed by (Welsh et al. 2000).

organisations. Well-developed partnerships between these groups of people are needed to make sure that investigator-initiated projects are also policy-relevant and management-relevant projects with the scientific and statistical rigor required to ensure that results are robust and conclusions are defensible (Gibbons et al. 2008). Ludwig et al. (2001) even contend that although scientists are good at identifying problems, they are often not good at solving the policy-implementation component of these problems. In contrast, policy professionals are frequently better equipped to find tractable solutions. Thus, the need for these diverse groups to work together to find workable solutions to environmental problems is clear (Lawton 2007; Pielke 2007; Likens et al. 2009; Colglazier 2016). Partnerships are important for other reasons. They can facilitate the flow of information between parties in ways that people from different backgrounds and with different expertise can readily understand (Gibbons et al. 2008). In other cases, particular agencies may no longer maintain the expertise and capability to do long-term monitoring (Lindenmayer et al. 2012b) (see Chapter 5). True collaborative partnerships are also essential because policy makers and resource managers will often not know how to frame questions in ways that can be resolved by well-executed monitoring, or may initially pose too many questions without prioritising them. They also may have unreasonable expectations about what questions can be answered or what problems can and cannot be solved by scientific projects and how much effective monitoring costs. Thus, policy makers need to understand the scientific approach and the importance of posing the right questions and in the correct way (see Chapter 2). Conversely, scientists need to articulate what kinds of questions they can and cannot answer and how long it will take. They also need to understand the complexity of the policy process (Clark 2002; Colglazier 2016). Scientists will often not fully comprehend the kinds of key problems faced by policy makers and resource managers that need to be addressed by long-term research and monitoring (Russell-Smith et al. 2003). Nor will scientists necessarily be aware of the policy options and the range of on-ground management interventions available for testing and monitoring in a particular ecosystem (Walters 1986; Lindenmayer et al. 2011a). Thus, without guidance, scientists may implement long-term research and monitoring programs that

70

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

answer questions with limited value for informing specific management actions. As we outlined earlier, professional statistical advice is essential to ensure that an appropriate experimental design can be developed to answer rigorously the questions conjointly conceived by scientists, policy makers and resource managers (Cunningham and Lindenmayer 2017). Developing partnerships is often difficult because representatives of different groups speak different ‘languages’, have different work cultures, different reward systems and different skill sets (Gibbons et al. 2008). Trust and mutual respect is essential, in part because various kinds of knowledge – scientific, policy and political knowledge – together influence decision making in natural resource management (Pielke 2007; Lindenmayer et al. 2013c). Hence, for example, good science alone is insufficient as it often will need to be translated into a form that politicians, policy makers and resource managers can understand and use (Colglazier 2016; Ellison 2016). Indeed, scientists will often need to strongly advocate for the considerable benefits that can be generated from their long-term monitoring (Box 3.3), sometimes engaging strongly in the political process to make their case.

Strong and dedicated leadership Strong, dedicated and focused leadership is essential in almost all effective monitoring programs – someone with ‘fire in the belly’ to keep the work going (Strayer et al. 1986; Lovett et al. 2007; Lindenmayer et al. 2015b). In many cases, long-term projects and team leaders even become synonymous (Strayer et al. 1986); the long-term work on Australian waterbirds by Richard Kingsford (Kingsford et al. 2008; Porter et al. 2016) is but one of many examples. Effective leadership is pivotal to all of the fundamental characteristics of successful monitoring programs described earlier in this chapter – setting appropriate questions, identifying new questions, developing a workable conceptual model, resolving suitable entities to measure, guiding study design, analysing data, communicating results to management agencies, policy makers and the public, and establishing and maintaining partnerships (see Box 3.9). Leadership is also critical to many other factors that are discussed in this chapter, such as securing funding and ensuring good project management, both of which contribute to effective long-term monitoring programs. Leadership embodies both scientific and management leadership, as most successful long-term programs have a champion within a management agency. Indeed, partnerships appear to work best when there is a problem with a shared responsibility to find a solution (Pielke 2007; Gibbons et al. 2008). Major environmental problems provide a good example of where these partnerships are critical. Strength of leadership, paradoxically, also can be the Achilles heel of monitoring programs. This problem occurs because many programs are prone to collapse when a leader leaves (Strayer et al. 1986; Norton 1996) (see Chapter 2).

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

Box 3.3. Tests of management relevance – the impacts of acid rain on terrestrial and aquatic ecosystems in the Hubbard Brook Valley Effective monitoring programs are those that pass the test of management relevance (Russell-Smith et al. 2003). Discovery, monitoring, understanding, and then taking action to mitigate the effects of acid rain illustrates this interaction well. The very first sample of rain collected as part of the Hubbard Brook Ecosystem Study in the White Mountains of New Hampshire (www.hubbardbrook.org), collected in July 1963, was surprisingly acid (pH 3.7). But it took decades of monitoring and study to understand the numerous connections between atmospheric emissions, deposition and ecological response in recipient terrestrial and aquatic ecosystems in the Hubbard Brook Valley and elsewhere (Likens 2010; Holmes and Likens 2016). Long-term data on precipitation chemistry were critical to the debate and ultimate passage of the 1990 Amendments to the US Clean Air Act. These long-term data also raised many new questions requiring new scientific studies and increased awareness about the effects of human activities on natural ecosystems. An important finding from these long-term studies was that concentrations and input/output of SO42– in precipitation and stream water at the Hubbard Brook Experimental Forest were significantly correlated with emissions of SO2 from the emission source area hundreds to thousands of kilometres distant (Holmes and Likens 2016). Like SO42–, atmospheric deposition of NO3– at the Hubbard Brook Experimental Forest is correlated with NOx emissions (Butler et al. 2001, 2003). Long-term monitoring has been crucial for determining these critical relationships. Moreover, changes in the chemistry of precipitation need to be monitored in the future as political decisions affecting the Clean Air Act Amendments play out. The causes, distribution and ecological effects of acid rain have been studied and debated in North America since 1972 (Likens et al. 1972, 1979, 2001; Likens and Bormann 1974; Schindler et al. 1985; Charles 1991; Driscoll et al. 2001; Weathers et al. 2009; Likens 2010). The Hubbard Brook Experimental Forest is an important site for monitoring atmospheric pollutants in the north-eastern United States because of the long record, location ‘downwind’ of major emission sources and lack of large, local pollution sources. Acid rain has had major environmental impact on terrestrial and aquatic ecosystems of the Hubbard Brook Experimental Forest despite decreases in emissions of sulphur dioxide and nitrogen oxides from the emission source area (Likens et al. 2002; Likens 2004, 2010; Holmes and Likens 2016). For example, in response to the decline in emissions of sulphur and nitrogen oxides, the acidity of precipitation at Hubbard Brook has declined by ~80% since 1965 (Holmes and Likens 2016). Nevertheless, given the cost and angst involved with federal legislation to reduce emissions of the major acid rain precursors (sulphur dioxide and nitrogen oxides), it is important to continue long-term measurement of acid rain and its impact on recipient forest and aquatic ecosystems, which now have become increasingly sensitive to these acidic inputs because of the depletion of the acid neutralizing components in the soil (Likens et al. 1996, 2001, 2004; Likens 2010; Likens and Buso 2010; Holmes and Likens 2016).

71

72

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 3.4. Large mammalian herbivore–ecosystem interactions: a case study from the boreal forests of Isle Royale John Pastor Department of Biology University of Minnesota Duluth, Minnesota Large mammalian herbivores have long generation times and range over large portions of the landscape. Their foraging can greatly change species composition, net primary productivity, litterfall and soil nitrogen availability over long periods and large areas of the landscape (Danell et al. 2006). Elucidating these long-term and large-scale processes requires comparable long-term, systematic experiments and monitoring designed to test hypotheses specifically focused at these large and long scales. Moose are one of the major herbivores of boreal forests worldwide. A single adult moose requires ~10 kg of dry matter per day. Moose therefore have considerable potential to control boreal ecosystems through their foraging decisions. When we began our research in 1984, our hypothesis was that by foraging on small diameter aspen (Populus tremuloides) and birch (Betula papyrifera), moose convert boreal forests more quickly to dominance by unpreferred spruce (Picea glauca) and balsam fir (Abies balsamea). Aspen and birch leaves and twigs have high nitrogen and low lignin contents but needles and twigs of conifers have low nitrogen, high lignin and high phenolic contents. These same chemical properties determine decay and nitrogen release rates of litter. We therefore predicted that productivity and nitrogen cycling would decrease as moose foraging increases dominance by conifers. Clearly, these effects of moose may require decades to take effect. Fortunately, some long-term experiments can be found ready-made. Yosef Cohen and I, and our graduate students Ron Moen and Pam McInnes, began our work by taking advantage of several long-term moose exclosures on Isle Royale established by Laurits Krefting of the University of Minnesota St Paul in the late 1940s. Krefting had already established that aspen, birch and other preferred species remained dominant inside the exclosures while the vegetation outside the exclosures became increasingly dominated by the less preferred spruce and fir. If our hypothesis was correct, litterfall should be more dominated by conifer litter and net primary productivity, microbial biomass and nitrogen mineralisation rates should be lower in control plots than in the exclosures. These predictions were confirmed by McInnes et al. (1992) and Pastor et al. (1988, 1993). Newer long-term exclosure experiments established in northern Sweden extend these results beyond Isle Royale to Scandinavia (Persson et al. 2005, 2009). As clear as these experimental results were, they raised more questions. How common are these results over the entire landscape? Are the effects of moose overridden by topography, bedrock, disturbances and other factors? Are these effects randomly distributed across the landscape or is there some spatial pattern to them? To answer these questions, we began a second phase to our long-term

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

Fig. 3.5.  Isle Royale – the location of long and detailed studies of moose and herbivore impacts on ecosystem dynamics (see text) (Photo by Shawn Malone)

program by monitoring the spatial distribution and dynamics of moose browsing intensity, plant species composition and soil nitrogen availability in two, 6-ha grids of 100 points apiece at an average spacing of 24 m in two valleys on Isle Royale for the past 18 years (Fig. 3.5). We initially found that the spatial distribution of these properties is not random but instead distributed in waveform patterns (Pastor et al. 1998). These waveform patterns do not correspond to topography, bedrock geology, fire history or any other conceivable cause other than the effects of the moose population. They also resemble similar waveform patterns in reaction-diffusion models of a predator (in this case, the moose) and its prey (the two plant species). During 18 years of monitoring, the spatial distributions of moose browsing and its effects on plants and soil nitrogen availability have changed in response to moose population cycles. The waveform distribution was most clearly expressed when moose population densities were at their peak, but when moose population densities crashed to historically low levels, the distribution of browsing and its effects became more random (De Jager and Pastor 2009; DeJager et al. 2009). Long-term monitoring has demonstrated that moose exert a great deal of control of plant species composition, net primary productivity, litterfall and soil nitrogen availability across boreal landscapes. The nature of this control varies with the phase of the moose population cycle. Because these cycles happen on decadal timescales, it was imperative for us to maintain a continuous monitoring program across almost two decades and that exclosures four decades old were available to us. Measurements made on shorter timescales may have been contingent on the phase of the moose population cycle in which they occurred. A full understanding of how large mammals control ecosystem processes can only emerge from longterm studies.

73

74

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 3.5. The importance of doing it right In the early 1950s, Dr Charles David Keeling, a postdoctoral atmospheric scientist at the California Institute of Technology, was measuring carbon in river water and carbon dioxide in the air. He found, surprisingly, that the air had a constant value of 310 ppm if samples were collected remotely, away from obvious sinks and sources of carbon dioxide. He then moved to the Scripps Institution of Oceanography in La Jolla, California, to work with Roger Revelle on carbon dioxide in the atmosphere. Because of prior connections with Harry Wexler of the US Weather Service, he was able to establish a monitoring station on Mauna Loa on the island of Hawaii in the Hawaiian Island archipelago in 1958. There were several potential problems with this location, including that the site was close to an active volcano. On the positive side, the site was above the tree line and situated to sample the easterly trade winds with no major air pollution sources upwind for thousands of kilometres. Being a thoughtful and careful scientist, Keeling set aside a large cylinder of carbon dioxide at the beginning of his study to use in calibrating his gas-measuring instruments as the study progressed. He wasn’t particularly concerned that the stated concentration of carbon dioxide in the cylinder was accurate, just that it would not change over time and serve as a standard or reference to calibrate his instruments in the future to this initial value. Such standards are key to successful long-term monitoring and they can be of many types, both biotic and abiotic. After some 10 years of careful measurements of carbon dioxide at the Mauna Loa monitoring site, carefully removing any samples from upslope winds, which may have been ‘contaminated’ by carbon dioxide produced by vegetation at lower elevations, Keeling made two remarkable discoveries: (1) the concentration of carbon dioxide in the atmosphere was increasing, and (2) there was a consistent, seasonal pattern of increase and decrease of carbon dioxide concentration, which Keeling attributed to global biotic processes of photosynthesis and respiration (Fig. 3.6). These two key conclusions are important in that they allow short-term trends to be distinguished from long-term change. Keeling’s funding was under threat at various stages. His major funding sources, the National Science Foundation and the National Oceanographic and Atmospheric Administration (NOAA), questioned whether this monitoring needed to be continued – ‘Haven’t we learned all we need to know from this study?’ ‘Aren’t 20 years long enough?’ One of us (GEL) was asked to join a Department of Energy Technical Advisory Group on carbon dioxide to address this issue. The review team of 12 met at Scripps Institution of Oceanography in LaJolla, California in 1977–78. Keeling was adamant that the study needed to continue and the review committee agreed, giving him and the monitoring program a strong endorsement to continue. At that point, the Department of Energy came to his rescue with funding for continuation of the program. The rest is history, so to speak. The study did continue, as did the long-term record of carbon dioxide concentrations from Mauna Loa, arguably one of the most important environmental records in the world (Fig. 3.6). These continuous and long-term data have provided the basic information about

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

75

410 400 NSF & NOAA coordinate to terminate Scripps program

parts per million

390 380

NSF warning on “routine warming”

370 360 350

Congressionally mandated budget cuts force shutdown

“Mandated convergence” with NOAA

340

DOE pulls out of global carbon research

330 DOE rescues program

320 310 1960

1970

1980

1990

2000

2010

Year Fig. 3.6.  The CO2 (Keeling) curve from Mauna Loa, Hawaii, United States (redrawn from http://scrippsco2.ucsd.edu)

carbon dioxide increase in the world’s atmosphere, which in turn has led to understandings about global warming and climate change. Other monitoring sites throughout the world have confirmed Keeling’s original findings about carbon dioxide, and the continuing long-term record (now funded and maintained at NOAA’s Mauna Loa Observatory) establishes the baseline for evaluating ongoing changes in atmospheric chemistry, as well as providing discoveries and new scientific questions to be explored (e.g. Keeling et al. 1995, 1996). Keeling died in 2005 at the age of 77, having received numerous prestigious awards for his work. The Keeling curve is arguably one of the most important long-term environmental trends globally and the data that underpin it continue to be gathered. Without Keeling’s leadership, we would not have our current understanding of how carbon dioxide is increasing in the atmosphere.

This problem illustrates the need for succession planning, often at an earlier stage than many senior scientists and project managers recognise. We return to this key topic of succession planning in Chapter 5.

Potential to identify key emerging issues The value of collecting long-term datasets on conditions in the atmosphere demonstrated by the Keeling curve (Box 3.5) can be particularly powerful when they identify emerging problems, as has been occurring in parallel work on the

76

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

condition of the atmosphere at Cape Grim in north-west Tasmania. As there are no nearby cities or industry that would contaminate the atmosphere, the airshed coming off the South Pacific/Antarctic Oceans at Cape Grim is thought to be representative of a large area of the Southern Hemisphere, unaffected by regional pollution sources, and often is referred to as the cleanest air on Earth. Importantly, long-term monitoring at Cape Grim has produced evidence of major changes in concentrations of two relatively unfamiliar greenhouse gases since 1978. Although their impact on incoming solar energy, expressed as the radiative forcing effect (the consequences on the irradiance entering the troposphere), is presently far less than the combined effects of better known greenhouse gases such as carbon dioxide and methane, they could have the potential to increase global temperatures. One of these greenhouse gases, nitrogen trifluoride (NF3) has an estimated global warming potential 17 000 times that of carbon dioxide over 100 years, and remains in the atmosphere for ~550 years (Weiss et al. 2008). It is a by-product of manufacturing plasma screen televisions and other goods. Measurements at Cape Grim showed that concentrations of this gas increased at a rate of 11% per year to 0.454 parts per trillion between 1978 and 2008. The second greenhouse gas, sulfuryl fluoride (SO2F2), is a crop fumigant used to preserve fresh produce. It has a warming potential 4780 times greater than carbon dioxide over 100 years, and remains in the atmosphere for up to 40 years (Muhle et al. 2009). Its concentration at Cape Grim increased between 1978 and 2008 at a rate of 5% per year.

Ongoing funding No project can proceed or be maintained without access to funding. Generating the funding to maintain long-term monitoring programs is a truly major challenge and very few individuals and organisations have ever managed to do it successfully (Strayer et al. 1986). We discuss the many reasons for this challenge in Chapter 5, but suggest that many of the underlying problems appear to be associated with the culture of ‘short-termism’ that pervades the culture of science and the culture of western societies; cultures which are fundamentally mal-adapted to long-term work. Successful programs where funding has been maintained are characterised by many of the factors discussed in this chapter, such as strong leadership, evolving questions that are highly relevant to resource management and society, high levels of scientific productivity (i.e. published articles, books and media reports) and management and policy relevance. In some cases, successful ongoing funding is linked with appropriate matching of the size of a monitoring program relative to the scale of a human activity (e.g. a logging operation) (Lindenmayer and Franklin 2002) or the overall budget for the management of a National Park or nature reserve (Edwards et al. 2003). Some have argued that ~10% of the budget for a management activity (e.g. a program for

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

77

widespread restoration across many landscapes or riparian areas) should be dedicated to monitoring the effectiveness of that program (Lindenmayer and Gibbons 2012). In some cases, the ‘fire in the belly’ is so strong that considerable personal funds have been used for years to support long-term monitoring (also see Chapters 4 and 5).

Frequent use of data Another key ingredient for maintaining long records of high quality is the frequent examination and use of these data. Such examinations result in important discoveries and stimulate new research and management questions. They are also the primary way errors, artefacts or other problems are uncovered, and enthusiasm about the research is sustained. Moreover, it is much easier to resolve a problem in long-term data when it is identified in a timely fashion and while observers and methods are still available for examination and discussion. A further issue with the frequent use of data is that some workers use large amounts of public funds to gather datasets but then never examine and/or publish the results of their work. We are only too familiar with many such cases. We argue that it is unethical not to analyse and then report on efforts expended in gathering scientific information. The public (as well as donors to private foundations that support research) rightly have an expectation to see a return on their investment in science and monitoring.

Scientific productivity One criterion often used to gauge success is scientific productivity. We strongly believe that the results of monitoring programs must be published in peer-reviewed literature. This outreach is essential to inform the public, funders and resource managers about valuable findings, and it helps to establish the credibility, quality and visibility of a project. This outreach is, in turn, essential to convince funders that investments are appropriate and should be maintained. It is notable that high numbers of scientific publications have characterised all of the effective monitoring programs featured in this chapter and the following one (see Chapter 4). A potential problem is that it takes a long time to generate long-term trends and patterns with empirical data. Hence, it can take a long time to generate substantial scientific publications from such kinds of work. This delay can be perceived as a lack of productivity and threaten the continuity of a project. As a counter to this problem, we suggest there can be considerable value in exploring avenues to generate rapid returns on research and monitoring investments and hence highlight scientific productivity. A long-term monitoring program can be used as a framework around which to co-locate shorter-term projects (Lindenmayer et al. 2012a). For example, retrospective or cross-sectional studies can serve as a prelude to longer-term projects and can provide key initial insights

78

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

or questions. These other projects, often built around the major research program (including student projects), add value to overall research effort and enrich the overall body of scientific knowledge. Various kinds of collective evidence derived from several kinds of studies can strengthen an overall research effort, make a body of scientific knowledge more compelling, and allow for an examination of consistency of research outcomes. Working with the news media and social media to generate high-quality and timely reports about long-term data, particularly about trends and extremes discovered, can be another way to enhance outreach from long-term research and monitoring programs. Such reports can be very effective in informing the public, policy makers and funders about environmental issues and the research being done (e.g. Likens 1992; Burns and Lindenmayer 2014).

Maintenance of data integrity and calibration of field techniques New sensors, modified or new analytical procedures and real-time data can add significantly to long-term data, but offsets and glitches generated by new methodology are a common problem in long-term datasets and must be addressed carefully (see the case of population monitoring of the giant panda in Chapter 2). We return to topics of big data, observation networks and related themes in Chapters 4 and 5. In the long-term Hubbard Brook Ecosystem Study, an analytical method or procedure is not replaced with a new one without first overlapping the two. This overlapping period may be for many months or for a year or more to compare results and avoid offsets in the record due to different methodologies (Buso et al. 2000). Also, many samples are stored for later analysis to help reconcile problems and to enable new questions to be pursued when new technology becomes available (e.g. Alewell et al. 1999; see also Box 3.6).

Little things matter a lot! Some ‘tricks of the trade’ Several seemingly small factors contribute enormously to the success of long-term monitoring, sometimes out of proportion to expectations. These include: (1) safe and adequate field vehicles (trucks, snowmobiles, boats, etc.); (2) continuity of employment of field staff; (3) access to field site(s); and (4) spending time in residence at field sites. We base all of these things on the assumption that longterm monitoring occurs primarily at a field site, often remote, and sometimes relatively inaccessible.

Field transport The availability of reliable transportation to a field site is critical to maintain routine sampling protocols. It is especially important for sampling during unusual events and on short notice, such as following a storm or a fire, where samples must

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

79

Box 3.6. The importance of archiving samples Archiving field samples can make a significant difference to the success of a long-term monitoring program (Whitlock 2011; see Fig. 3.7). An important example is the use of archived precipitation samples, collected during the long-term Hubbard Brook Ecosystem Study, for new and originally unplanned purposes. Isotopic measurement and analysis of ∂18O and ∂2H in precipitation samples can provide a mechanistic understanding of the changing hydrologic cycle, such as more frequent heavy rains or snowy winters. Weekly precipitation samples (rain and snow) archived from 1968 to 2010 (43 years) provided a treasure trove for these new isotopic analyses after determining that these stored samples could be used in this way. Unexpected results from this study provided evidence that changes in the hydrology of the north-eastern United States are governed by the interactions among the Atlantic Multi-decadal Oscillation, loss of Arctic sea ice, the fluctuating trajectory of the jet stream and regular incursions of polar air into the north-eastern United States (Puntsag et al. 2016). Over the duration of the long-term Hubbard Brook Ecosystem Study, use of archived samples has provided the basis for some 30 peer-reviewed publications, and more importantly has allowed pursuit of questions not thought of at the outset of the study.

Fig. 3.7.  A series of archived photographic images that highlight the size of populations of Guillemot nesting on the coast of Stora Karlsö, a Swedish island in the Baltic sea (Photos from Jonas Hentati-Sundberg)

80

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

be collected quickly to characterise the initial conditions following such a disturbance (Lindenmayer et al. 2010). Waiting to obtain access to a vehicle through ‘normal channels’ (e.g. commercial rental or university fleet service late at night or over a weekend) could seriously compromise the integrity of a long-term record, and indeed, the scientific validity of a study. Administrators and funding agencies frequently argue that the expense for purchase and maintenance of such field vehicles is an unnecessary luxury. They do not understand the critical importance of this basic infrastructural support. Moreover, specialised field vehicles can be as important to the success of monitoring as an electron microscope or mass spectrometer are to a laboratoryorientated scientist. It is very important to be able to access field sites quickly, safely and on short notice whenever the scientifically driven need arises. Even hours of delay in collecting such samples could jeopardise the integrity, value and novelty of a long-term record. Manual sampling of storm-generated hydrographs in streams or rivers during flooding conditions is a good example.

Field staff Maintaining motivated staff for long periods is critical to the success of long-term monitoring and pivotal to the collection of high-quality, long-term data. Dedicated, intelligent and trained staff that work for a long time on a project bring many vital intangibles to the effort. Such staff: ●● ●● ●● ●●

●●

●● ●● ●● ●●

know what to do and how to do it run, maintain and interpret data generated by machines such as field sensors ground-truth remote sensing are dedicated to the collection of samples or data, often under arduous environmental conditions ensure the consistency of analysis (e.g. taxonomic knowledge of diverse organisms, such as phytoplankton or zooplankton) and are able to notice in a timely manner when there is a temporal change in such organisms maintain enthusiasm for the overall success of a project contribute to scientific publications play a key role in communication efforts generate new questions that could or should be pursued as a result of spending a great deal of time in the field on a project.

Access to field sites Access to remote field sites can be problematic because of inclement weather, poor road or boating conditions, security barriers or potential harm to the organisms or structure of the field site (e.g. excessive disturbance by trampling during sample collection). Care and planning can help anticipate and avoid these problems and

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

81

sustain the integrity of long-term records. Maintaining the infrastructure that is necessary for site access should be a high priority. Security of tenure is part of this requirement, and this includes not only the immediate field site but the surrounding landscape which could affect processes that spill over into the immediate study area (e.g. Laurance and Luizao 2007).

Time in the field Based on our collective experience, we are convinced that it is extremely important for research personnel, including principal investigators, to spend significant amounts of time working together at long-term monitoring sites. Numerous examples exist where increased idea generation and novel and profound research initiatives have been generated from the melting pot of senior and junior scientists spending significant time working together at field sites (e.g. see also Chapter 2 of Lawton 2007). We further discuss problems associated with a lack of time in the field in Chapter 5. Conversely, a lack of time in the field can result in some very ordinary science. As an example, a close colleague of ours recounted reading a paper in a leading ecological journal outlining niche separation of closely related grass species in the western United States. The authors of the paper had clearly never visited the study area, nor understood the ecology of the species in the paper; the two ‘species’ being examined were in fact the same taxon with the one grass species being allocated a different name following taxonomic revision!

The adaptive monitoring framework Based on some of the salient features that should accompany effective monitoring programs, we propose the use of the adaptive monitoring paradigm and show its key steps in Fig. 3.8 (Lindenmayer and Likens 2009). The adaptive monitoring framework is motivated by questions carefully posed at the outset. In fact, we argue that effective monitoring can occur only when it is based on questions. Without them, many important discoveries cannot be made because most discoveries usually represent answers to questions. This suggestion might seem almost banal but as we show in Chapter 4, it is extraordinary how often this initial question-setting step is done poorly or overlooked entirely, even in very large government-funded programs. Close attention to questions may quickly indicate that long-term research and monitoring are not needed (McDonald-Madden et al. 2010) (see Box 1.4 in Chapter 1). Setting clear objectives and framing questions will help resolve squabbles about what to monitor because that will be based on the questions asked. It also will circumvent arguments about whether particular entities are ‘indicators’; again because the entities selected for monitoring will be those that are appropriate for

82

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

ADAPTIVE MONITORING

TIME

Design a monitoring approach for the ecosystem to answer questions

Collect data

Analysis and interpretation

New or restated questions or initial questions addressed

Maintain the integrity of long-series repeated, core measures

Questions

Understanding

Question changes or evolves into a new question

Analytical approach changes or evolves

New technology

Fig. 3.8.  The adaptive monitoring framework (redrawn from Lindenmayer and Likens 2009)

answering the questions being posed (Lindenmayer and Likens 2011) (see Chapter 2). A fundamental part of the adaptive monitoring paradigm is that the question setting, experimental design, data collection, data analysis and data interpretation are iterative steps (Fig. 3.8). A monitoring program then can evolve and develop in response to new information or new questions. For example, it may be appropriate to alter the frequency of data collection when key entities are changing at rates different than those initially anticipated. An adaptive monitoring approach also enables questions to change, new questions to be posed and new protocols to be embraced when, for example, new technology arises to enhance field or laboratory measurements within the overall monitoring framework. Several authors have highlighted the importance of evolving questions as part of a robust approach to monitoring programs (Hicks and Brydes 1994; Ringold et al. 1996). An adaptive monitoring approach represents an iterative and linked framework that allows this evolution to take place in a logical way. An important caveat here is that the adoption of new sampling or analytical methods must ensure that the integrity of the long-term data record is neither breached nor distorted. Another caveat is that sometimes particular questions cannot be addressed with a pre-existing monitoring design or a given long-term dataset (that it was not originally designed to address) and an entirely new investigation may need to be established.

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

83

An ability to pose new questions through the adaptive monitoring framework is critical for several reasons: ●●

●●

●● ●●

It is impossible to know all the questions to answer at the outset of a long-term research project or monitoring program (Lovett et al. 2007). New insights from the results of a study stimulate new questions and new problems that can be addressed. New questions help maintain enthusiasm for a long-term project. The ability to pose new questions ensures that a project can be adapted to solve new or emerging problems, thereby highlighting its relevance to management and policy.

Examples of the adaptive monitoring framework The adaptive monitoring paradigm was first developed in 2009 (Lindenmayer and Likens 2009). The idea has been widely welcomed but to date there are relatively few documented examples of its application (see Box 3.7). This should not be surprising given that the related concepts such as adaptive management have been in vogue for many decades (e.g. Holling 1978; Walters 1986) and discussed in tens of thousands of scientific and other articles, yet there are almost no formally documented examples of true active adaptive management experiments (Westgate et al. 2013). However, we suspect there are actually many unreported examples of adaptive monitoring such as occur when a new analytical procedure or new instrument becomes available. Nevertheless, there are some useful examples of adaptive monitoring that have been documented. For example, the tropical savannas in the Northern Territory of Australia have been targeted in a long-term monitoring program that spans several iconic national parks including Kakadu National Park (Russell-Smith et al. 2014). The original design of the monitoring program was based on a compromise between access to sites for recording changes in animal and plant occurrence and training of Indigenous rangers (Russell-Smith et al. 2014). Major declines in mammal fauna in Kakadu National Park and some other protected areas in northern Australia have demanded that the design of the monitoring program be altered to determine the drivers of decline and the effectiveness of management interventions that attempt to halt and eventually reverse the decline. The planning of this change in the monitoring program is ongoing but it aims to ensure that the value of the datasets gathered in the past will be maintained and used as a platform for new work (Brendan Wintle and Graeme Gillespie, pers. comm.). Other examples of adaptive monitoring may be less extreme than the case outlined for the tropical savannas of northern Australia. Adaptive monitoring may entail a relatively simple shift in, or addition of, the entities being measured in response to emerging problems. An example is the use of lichens as indicators of

84

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

pollutants and airshed quality, which has a long and often highly successful history spanning more than a century (Branquinho et al. 2015). However, the factors affecting lichen diversity have changed over time. From the industrial revolution until the 1980s, sulphur dioxide was a primary driver of lichen diversity in urban environments. Since the 1980s, and following management to reduce sulphur emissions as atmospheric pollutant, nitrogen pollution has become the key driver of lichen diversity (Holmes and Likens 2016). If attempts to control nitrogen pollution are successful, effects of climate change and industrial heat islands in cities will be the primary drivers of lichen diversity in urban environments. However, measures of diversity may be ineffective as measures of climate change because of redundancy between, and turnover among, species. This response suggests that other measures, such as functional diversity based on traits, also might be useful additional measures to be recorded. Another example of adaptive monitoring comes from monitoring the Environmental Stewardship Program that was established to assess the effectiveness of management interventions associated with a major agrienvironment scheme in the temperate woodlands of eastern Australia. Under this program, farmers are paid to undertake particular management activities to improve the condition of patches of endangered box-gum grassy woodlands on their land. In total, the program comprises 158 farms stretching over 2000 km (south to north). Patches of box-gum grassy woodland targeted for stewardship management on each of the 158 farms are matched with a control patch (where no stewardship management occurs). Both patches are monitored (Lindenmayer et al. 2012b). The budget for the monitoring was initially substantial as the government agency responsible for the Environmental Stewardship Program demanded that every patch on every farm be monitored every 1–2 years. However, the inevitable round of funding cuts occurred four years into the 15-year program. Adaptive monitoring had to be adopted to prevent the entire monitoring effort collapsing. Unfortunately, a change in the way monitoring is to be done from 2018, and who will be doing that monitoring, will breach the integrity of the data collection in the program and negate the earlier value of the monitoring (see Box 2.3 in Chapter 2). A rotating sampling monitoring approach (Welsh et al. 2000) was adopted in which 65% of farms were surveyed in any given year with a complete ‘census’ of all farms undertaken twice in four years. The emphasis of monitoring switched from an assessment of compliance with contractual demands for implementing particular kinds of conservation management to estimation of changes in condition across a large ‘population’ of sites. Importantly, recent analyses of the data gathered in the monitoring program show that stewardship management is leading to significantly improved woodland condition and also increases in biodiversity (Kay et al. 2016; Sato et al. 2016).

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

85

Box 3.7. A hypothetical example of adaptive monitoring We provide a further illustration of the potential value of adaptive monitoring with a hypothetical example in an aquatic setting. Preceding and following a management intervention to construct a wastewater treatment plant to lower phosphorus loading to a lake where eutrophication was thought to be driven by excess loading of phosphorus, long-term monitoring of water chemistry and phytoplankton diversity and productivity had been established. This before-after framework had been deliberately set up to guide robust monitoring of the actual impact of this management action on eutrophication of the lake. After a decade of postconstruction monitoring, it became apparent that because of changing climate, it would also be important to monitor the changes in thermal stratification and duration of ice cover for the lake, in addition to water quality, in order to evaluate more precisely the rate of recovery from eutrophication. To adapt the original monitoring scheme to include climate change effects, several issues would need to be resolved. For example: (1) How do these new parameters fit into the conceptual model for eutrophication of this lake? (2) What new question(s) should be posed and new measurements made to address the climate-change problem? For example, a possible new question might be: ‘Is there an effect of increasing temperature on water circulation patterns and algal productivity in the lake?’ (3) What statistically based experimental design would be needed to answer this question? (4) How can the integrity of the overall monitoring program be maintained given that new questions have been posed and new or additional monitoring protocols would be required possibly at different times of the year?

Adaptive monitoring is a general and not a prescriptive framework The core principles of the adaptive monitoring framework, with its focus on questions, a conceptual model and an experimental design, mean that the approach is relevant to a wide range of circumstances and to all kinds of monitoring: from very simple to very complex programs, and from mandated monitoring programs usually conducted at a coarse scale to site- or landscape-level curiosity-driven monitoring programs (as defined at the beginning of Chapter 1). However, as illustrated in examples already described, the adaptive monitoring paradigm does not lead to a set of highly specific prescriptions that can be applied uncritically to any given monitoring program. Rather, its specific application will be context-dependent, and will vary in response to the particular problem to be resolved, the questions being posed, and the composition and ecological processes of particular ecosystems. Indeed, adaptive monitoring can be applied in a range of circumstances, such as when: (1) there is a major cut in funding and a new design is required; (2) new equipment for measuring the environment becomes available; or (3) new ecological or other problems emerge and a new approach is needed to identify the factors creating the problem. Our approach also emphasises that

86

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

adaptive monitoring is not mindless data collection, but instead pivots on the legitimate scientific practice of posing rigorous questions and carefully designing and implementing appropriate studies to answer them. The adaptive monitoring paradigm shares many common elements with the adaptive management paradigm, which is much discussed but rarely implemented (Westgate et al. 2013). That is: (1) the question-setting step will often be best motivated and implemented by testing management interventions, which are relevant to policy options for the management of ecosystems and natural resources (Walters and Holling 1990); (2) there is an explicit acknowledgment of ignorance, that is, there are things that are not known that the monitoring work specifically sets out to answer (Walters 1986); (3) the questions developed for testing are based on a priori predictions from a conceptual model suggesting how an ecosystem might function and what the response of monitored entities might be to competing management interventions in that ecosystem; and (4) there are strong links between the iterative steps in the adaptive monitoring framework, which connect question setting with management actions (e.g. an intervention) and an improved understanding of ecosystem or population responses.

Increased future role for adaptive monitoring We trust that the adaptive monitoring framework will become increasingly important in the future to help make long-term ecological research and monitoring programs more effective. Novel opportunities for adaptive monitoring will arise because of largely anthropogenic perturbations such as climate change, acid rain, elevated levels of ozone and increased occurrences of large-scale, natural disturbances. These human-accelerated environmental changes represent novel, but unreplicated, ‘experimental’ opportunities for long-term monitoring. Because such environmental problems are now at the forefront of society’s concerns, adaptive monitoring will be important in finding solutions to them or designing ways to mitigate their effects. For example, the application of the adaptive monitoring framework should help in the development of new questions within existing long-term monitoring programs about rapid climate change effects. In addition, planned experimental manipulation of environmental parameters, particularly at the catchment scale, also will provide vital information for the development of scientific understanding and protocols for managing of natural resources. The scientific value of these manipulated disturbances is evident from the results of long-term monitoring at places like the Hubbard Brook Experimental Forest (see Box 3.8).

Summary This chapter outlines some of the key features that should characterise effective monitoring programs. They should: (1) address well-defined and tractable

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

Box 3.8. Long-term research and monitoring in the Hubbard Brook Experimental Forest The long-term monitoring of hydrologic and chemical fluxes to and from the various watershed ecosystems of the Hubbard Brook Valley of New Hampshire (air, land and water interactions) is an instructive example of adaptive monitoring. Over time, long-term monitoring was refined as necessary to address each new question better, but without altering the fundamental data collection format. These long-term studies have revealed new understanding about several major environmental problems, including acid rain and its effects on forest growth and soil neutralising capacity (Likens et al. 1972, 1996, 2004; Cogbill and Likens 1974; Likens and Bormann 1974; Driscoll et al. 2001; Likens 2010; Holmes and Likens 2016), hydrological and biogeochemical response to clearcutting of forests (e.g. Likens et al. 1970, 1998, 2002; Bormann and Likens 1979; Holmes and Likens 2016), major declines in bird and salamander populations (Holmes and Likens 2016), salt contamination of surface waters (e.g. Bormann and Likens 1985; Rosenberry et al. 1999; Kaushal et al. 2005), dilution of precipitation and stream water, documenting the effects of climate change (Holmes and Likens 2016), and long-term legacies of forest disturbance (e.g. Likens and Buso 2010; Holmes and Likens 2016). The studies have also been fundamentally important in the generation of new approaches to management and policy (e.g. Likens et al. 1978, 2004; Driscoll et al. 2001; Likens 2010; Holmes and Likens 2016). Change is a predominant feature of all ecosystems and has been the impetus of long-term studies of watershed ecosystems and associated aquatic ecosystems within the Hubbard Brook Valley (Bormann and Likens 1979; Likens and Buso 2010; Holmes and Likens 2016). The characteristics of change in the Hubbard Brook Valley include those brought about by natural biological succession of forest and aquatic ecosystems (e.g. development of soil profiles and organic debris dams) and those caused by disturbance such as acid rain, ice- and wind-storm damage and climate change. Some of the major changes during the past five decades include decreases in concentration and flux of sulphate in precipitation, stream and lake water (Likens et al. 2002; Winter and Likens 2009; Likens 2010; Holmes and Likens 2016) related to decreases in sulphur dioxide emissions from the source area of the Hubbard Brook Experimental Forest (Butler et al. 2001, 2003; Likens et al. 2001, 2005; Likens 2013; Holmes and Likens 2016), decreases in concentration and flux of calcium and magnesium and increases in pH of precipitation and stream water (Likens et al. 1996, 1998, 2004, 2005; Likens 2013; Holmes and Likens 2016), decreases in maximum snow depth and water content of the snow, while mean annual air temperature, particularly during winter, has increased (Campbell et al. 2007; Holmes and Likens 2016), and decreases in bird and salamander populations (Holmes and Likens 2016). Long-term records have been indispensable for identifying, documenting and characterising important changes affecting the Hubbard Brook Experimental Forest. As a result of these long-term data, trends have become clearer and more meaningful to managers and policy makers. Nevertheless, after more than 53 years of continuous research and monitoring, there continues to be many exciting and compelling questions to pursue with adaptive monitoring.

87

88

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 3.9. Some critical components for maintaining effective monitoring programs •• Permanently mark and identify plots and study sites. Detailed descriptions of study areas and field protocols should be filed in more than one location, and provide sufficient detail so that other investigators can find sites, and reproduce calculations and methods at some later date. •• Establish appropriate and adequate references and/or control sites at the beginning of the study. •• Ensure availability of appropriate field equipment. •• Establish long-term security of research sites and field equipment. •• Ensure reliable access to field sites, including availability of safe and reliable vehicles, such as trucks, boats, snowmobiles. •• Pay careful attention to field and laboratory protocols. Standardise methods and procedures to the extent possible, and inter-calibrate them with other organisations or individuals doing similar studies. Calibrate analytical results by comparison against standardised samples. Analytical methods or collection procedures should not be changed without fully testing the effect of the new procedure on the long-term record. •• Determine sampling frequency from questions addressed and from analyses of results. Ensure the duration of measurements is at least as long as the phenomenon being evaluated, or scaled to the frequency of the event or the life history of the organism being studied. •• Do not adopt methods or procedures developed for one location or study for another area or study without careful testing and justification. •• Employ strict database management and data storage, including the agility to change with changes in technology. Datasets should be stored in at least two separate locations to avoid accidental loss. Long-term storage of samples is highly desired. •• Respect and support ‘fire in the belly’ of the project’s leadership (a champion for the project). •• Encourage and support staff to maintain stability and competence. •• Resolve intellectual property issues at the outset of a project. •• Match of the scale of monitoring to the spatial and temporal dimensions of the question being addressed. •• Ensure senior and junior scientists spend significant time in the field working together. •• Constantly update, scrutinise for errors and rigorously review continuous datasets. •• Use long-term datasets to answer questions. Regularly examine, analyse and publish the long-term data. •• Maintain a stream of publications to develop project credibility, credentials and outreach. •• Maintain scientific independence and integrity of the project by avoiding vested interests.

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

89

•• Establish partnerships among scientists, policy makers and staff from resource management agencies to ensure that long-term work passes the test of management relevance. •• Ensure availability of adequate, sustained and reliable funding. •• Encourage ongoing development and evolution of questions that can use the information from the monitoring program as a frame or operate parallel to it.

questions specified before the commencement of a monitoring program; (2) be underpinned by rigorous statistical design; and (3) be driven by a human need to know about an ecosystem (e.g. the effects of a pollutant or changes in climate) so that they ‘pass the test of management relevance’ (sensu Russell-Smith et al. 2003). Effective monitoring programs will often have other typically interrelated features and we have summarised them in a type of checklist in Box 3.9. There are many examples of how long-term data can change perspectives or understanding about how the world works, including ‘surprises’ or unexpected responses to stresses or disturbances (Lindenmayer et al. 2012a) (Chapter 1). Long-term monitoring can invalidate or change pre-conceived ideas about how ecosystems change with time by showing how different time steps of observation or analysis (weekly, monthly, annual or decadal) can give different results (see examples in Box 3.4 and Box 3.5). The adaptive monitoring paradigm we have outlined in this chapter is characterised by many of the features in Box 3.9. We argue that adaptive monitoring has the potential to improve significantly the record of high-quality, long-term ecological research and monitoring programs worldwide. The adaptive monitoring paradigm should increase the credibility of monitoring programs within the scientific community by demonstrating the pivotal roles of the traditional scientific method of posing and then answering questions. This paradigm should elicit greater engagement of resource managers by ensuring that the questions posed pass the test of management relevance. It should also encourage and provide confidence to policy makers and funders in their attempts to be responsible in the use of public funds. Finally, the adaptive monitoring paradigm should assist scientists and resource managers in moving beyond protracted debates about how to monitor and what to monitor – debates that have greatly impeded progress in ecological research over the past few decades.

References Alewell C, Mitchell MJ, Likens GE, Krouse HR (1999) Sources of stream sulfate at the Hubbard Brook Experimental Forest: long-term analyses using stable isotopes. Biogeochemistry 44, 281–299. doi:10.1007/BF00996994

90

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Bormann FH, Likens GE (1967) Nutrient cycling. Science 155, 424–429. doi:10.1126/ science.155.3761.424 Bormann FH, Likens GE (1979) Pattern and Process in a Forested Ecosystem. SpringerVerlag, New York. Bormann FH, Likens GE (1985) Air and watershed management and the aquatic ecosystem. In An Ecosystem Approach to Aquatic Ecology: Mirror Lake and its Environment. (Ed. GE Likens) pp. 436–444. Springer-Verlag, New York. Branquinho C, Matos P, Pinho P (2015) Lichens as ecological indicators to track atmospheric changes: future challenges. In Indicators and Surrogates of Biodiversity and Environmental Change. (Eds D Lindenmayer, P Barton, J Pierson) pp. 77–90. CSIRO Publishing, Melbourne. Burns E, Lindenmayer DB (2014) Policy Handbook: Learning from Long-term Research to Better Manage Biodiversity in Australia. CSIRO Publishing, Melbourne. Burton TM, Likens GE (1975) Salamander populations and biomass in the Hubbard Brook Experimental Forest, New Hampshire. Copeia 3, 541–546. doi:10.2307/1443655 Buso DC, Likens GE, Eaton JS (2000) Chemistry of precipitation, stream water and lake water from the Hubbard Brook Ecosystem Study: A record of sampling protocols and analytical procedures. USDA Forest Service, Northeastern Research Station, Newtown Square, Pennsylvania. Butler TJ, Likens GE, Stunder BJ (2001) Regional-scale impacts of Phase I of the Clean Air Act Amendments in the USA: the relation between emissions and concentrations, both wet and dry. Atmospheric Environment 35, 1015–1028. doi:10.1016/S1352-2310(00)00386-1 Butler TJ, Likens GE, Vermeylen FM, Stunder BJ (2003) The relation between NOX emissions and precipitation NO3- in the eastern USA. Atmospheric Environment 37, 2093–2104. doi:10.1016/S1352-2310(03)00103-1 Campbell JL, Driscoll CT, Eagar C, Likens GE, Siccama TG, Johnson CE, Fahey TJ, Hamburg SP, Holmes RT, Bailey AS, Buso DC (2007) Long-term trends from ecosystem research at the Hubbard Brook Experimental Forest. USDA Forest Service, Northern Research Station, Newtown Square, Pennsylvania. Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington, DC. Charles DF (Ed.) (1991) Acidic Deposition and Aquatic Ecosystems: Regional Case Studies. Springer-Verlag, New York. Clark TW (2002) The Policy Process: A Practical Guide for Natural Resource Professionals. Yale University Press, New Haven, Connecticut. Cogbill CV, Likens GE (1974) Acid precipitation in the northeastern United States. Water Resources Research 10, 1133–1137. doi:10.1029/WR010i006p01133 Colglazier W (2016) Encourage governments to heed scientific advice. Nature 537, 587. doi:10.1038/537587a Cunningham R, Lindenmayer DB (2017) Approaches to landscape scale inference and design issues. Current Landscape Ecology Reports 2, 42–50. doi:10.1007/ s40823-016-0019-4

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

91

Danell K, Bergström R, Duncan P, Pastor J (Eds) (2006) Large Mammalian Herbivores, Ecosystem Dynamics, and Conservation. Cambridge University Press, England. De Jager N, Pastor J (2009) Declines in moose population density on Isle Royale and accompanying changes in landscape patterns. Landscape Ecology 24, 1389–1403. doi:10.1007/s10980-009-9390-4 DeJager N, Pastor J, Hodgson A (2009) The effects of moose browsing on plant geometry regulates subsequent food intake rates across the landscape. Ecological Monographs 79, 281–297. Dickman C, Wardle G, Foulkes J, de Preu N (2014) Desert complex environments. In Biodiversity and Environmental Change: Monitoring, Challenges and Direction. (Eds DB Lindenmayer, E Burns, N Thurgate, A Lowe) pp. 379–438. CSIRO Publishing, Melbourne. Driscoll CT, Lawrence GB, Bulger AJ, Butler TJ, Cronan CS, Eagar C, Fallon Lambert K, Likens GE, Stoddard JL, Weathers KC (2001) Acidic deposition in the northeastern United States: sources and inputs, ecosystem effects, and management strategies. Bioscience 51, 180–198. doi:10.1641/0006-3568(2001)051[0180:ADITNU]2.0.CO;2 Edwards A, Kennett R, Price O, Russell-Smith J, Spiers G, Woinarski J (2003) Monitoring the impacts of fire regimes on vegetation in northern Australia: an example from Kakadu National Park. International Journal of Wildland Fire 12, 427–440. doi:10.1071/WF03031 Efford MG, Dawson DK (2012) Occupancy in continuous habitat. Ecosphere 3, 1–15. Ellison AM (2016) It’s time to get real about conservation. Nature 538, 141. Fedrowitz KF, Koricheva J, Baker SC, Lindenmayer DB, Palik B, Rosenvald R, Beese W, Franklin JF, Kouki J, Macdonald E, Messier C, Sverdrup-Thygeson A, Gustafsson L (2014) Can retention forestry help conserve biodiversity? A metaanalysis. Journal of Applied Ecology 51, 1669–1679. doi:10.1111/1365-2664.12289 Franklin JF, Harmon ME, Swanson FJ (1999) Complementary roles of research and monitoring: lessons from the U.S. LTER Program and Tierra del Fuego. Paper presented to the Symposium, ‘Toward a unified framework for inventorying and monitoring forest ecosystem resources.’ Guadalajara, Mexico, November 1998. Gibbons P, Zammit C, Youngentob K, Possingham HP, Lindenmayer DB, Bekessy S, Burgman M, Colyvan M, Considine M, Felton A, Hobbs R, Hurley C, McAlpine C, McCarthy MA, Moore J, Robinson D, Salt D, Wintle B (2008) Some practical suggestions for improving engagement between policy makers and researchers in natural resource management. Ecological Management & Restoration 9, 182–186. doi:10.1111/j.1442-8903.2008.00416.x Grove SJ (2007) Ecological research coverage at the Warra LTER Site, Tasmania: a gap analysis based on a conceptual ecological model. Tasforests 15, 43–53. Hall JA, Fleischman E (2010) Demonstration as a means to translate conservation science into practice. Conservation Biology 24, 120–127. doi:10.1111/j.1523-1739.2009.01297.x Hayward MW, Boitani L, Burrows ND, Funston PJ, Karanth KU, MacKenzie DI, Pollock KH, Yarnell RW (2015) Ecologists need robust survey designs, sampling and

92

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

analytical methods. Journal of Applied Ecology 52, 286–290. doi:10.1111/13652664.12408 Hicks BB, Brydes TG (1994) A strategy for integrated monitoring. Environmental Management 18, 1–12. doi:10.1007/BF02393745 Holling CS (Ed.) (1978) Adaptive Environmental Assessment and Management. John Wiley and Sons, Toronto. Holmes RT, Likens GE (2016) Hubbard Brook: The Story of a Forest Ecosystem. Yale University Press, New Haven, Connecticut. Hong L, Page SE (2004) Groups of diverse problem solvers can outperform groups of high-ability problem solvers. Proceedings of the National Academy of Sciences of the United States of America 101, 16385–16389. doi:10.1073/pnas.0403723101 Kaushal SS, Groffman PM, Likens GE, Belt KT, Stack WP, Kelly VR, Band LE, Fisher GT (2005) Increased salinization of fresh water in the northeastern United States. Proceedings of the National Academy of Sciences of the United States of America 102, 13517–13520. doi:10.1073/pnas.0506414102 Kay G, Barton PS, Driscoll D, Cunningham S, Blanchard W, McIntyre S, Lindenmayer D (2016) Incorporating regional-scale ecological knowledge to improve the effectiveness of large-scale conservation programmes. Animal Conservation 19, 515–525. doi:10.1111/acv.12267 Keeling CD, Whorf TP, Wahlen M, van der Plicht J (1995) Interannual extremes in the rate of rise of atmospheric carbon dioxide since 1980. Nature 375, 666–670. doi:10.1038/375666a0 Keeling CD, Chin JF, Whorf TP (1996) Increased activity of northern vegetation inferred from atmospheric CO2 measurements. Nature 382, 146–149. doi:10.1038/382146a0 Kingsford RT, Halse SA, Porter JL (2008) Aerial surveys of waterbirds – assessing wetland condition. Final report to the National Land & Water Resources Audit, University of New South Wales, Sydney. Laurance W, Luizao R (2007) Driving a wedge into the Amazon. Nature 448, 409–410. doi:10.1038/448409a Lawton JH (2007) Ecology, politics and policy. Journal of Applied Ecology 44, 465–474. doi:10.1111/j.1365-2664.2007.01315.x Likens GE (Ed.) (1989) Long-term Studies in Ecology: Approaches and Alternatives. Springer-Verlag, New York. Likens GE (1992) The Ecosystem Approach: Its Use and Abuse. Ecology Institute, Oldendorf/Luhe, Germany. Likens GE (2004) Some perspectives on long-term biogeochemical research from the Hubbard Brook Ecosystem Study. Ecology 85, 2355–2362. doi:10.1890/03-0243 Likens GE (2010) The role of science in decision making: does evidence-based science drive environmental policy? Frontiers in Ecology and the Environment 8, e1–e9. doi:10.1890/090132 Likens GE (2013) Biogeochemistry of a Forested Ecosystem. Springer, New York.

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

93

Likens GE, Bormann FH (1974) Acid rain: a serious regional environmental problem. Science 184, 1176–1179. doi:10.1126/science.184.4142.1176 Likens GE, Buso DC (2010) Long-term changes in streamwater chemistry following disturbance in the Hubbard Brook Experimental Forest, USA. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie 30, 1577–1581. Likens GE, Bormann FH, Johnson NM, Fisher DW, Pierce RS (1970) Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecological Monographs 40, 23–47. doi:10.2307/1942440 Likens GE, Bormann FH, Johnson NM (1972) Acid rain. Environment 14, 33–40. doi:1 0.1080/00139157.1972.9933001 Likens GE, Bormann FS, Pierce RS, Reiners WA (1978) Recovery of a deforested ecosystem. Science 199, 492–496. doi:10.1126/science.199.4328.492 Likens GE, Wright RF, Galloway JN, Butler TJ (1979) Acid rain. Scientific American 241, 43–51. doi:10.1038/scientificamerican1079-43 Likens GE, Driscoll CT, Buso DC (1996) Long-term effects of acid rain: response and recovery of a forest ecosystem. Science 272, 244–246. doi:10.1126/ science.272.5259.244 Likens GE, Driscoll CT, Buso DC, Siccama TG, Johnson CE, Lovett GM, Fahey TJ, Reiners WA, Ryan DF, Martin CW, Bailey SW (1998) The biogeochemistry of calcium at Hubbard Brook. Biogeochemistry 41(2), 89–173. doi:10.1023/A:1005984620681 Likens GE, Butler TJ, Buso DC (2001) Long- and short-term changes in sulfate deposition: effects of The 1990 Clean Air Act Amendments. Biogeochemistry 52, 1–11. doi:10.1023/A:1026563400336 Likens GE, Driscoll CT, Buso DC, Mitchell MJ, Lovett GM, Bailey SW, Siccama TG, Reiners WA, Alewell C (2002) The biogeochemistry of sulfur at Hubbard Brook. Biogeochemistry 60, 235–316. doi:10.1023/A:1020972100496 Likens GE, Dresser BK, Buso DC (2004) Short-term, temperature response in forest floor and soil to ice storm disturbance in a northern hardwood forest. Northern Journal of Applied Forestry 21, 209–219. Likens GE, Buso DC, Butler TJ (2005) Long-term relationships between SO2 and NOX emissions and SO42- and NO3- concentration in bulk deposition at the Hubbard Brook Experimental Forest, New Hampshire. Journal of Environmental Monitoring 7, 964–968. doi:10.1039/b506370a Likens GE, Walker KF, Davies PE, Brookes J, Olley J, Young WJ, Thoms MC, Lake PS, Gawne B, Davis J, Arthington AH, Thompson R, Oliver RL (2009) Ecosystem science: toward a new paradigm for managing Australia’s inland aquatic ecosystems. Marine and Freshwater Research 60, 271–279. doi:10.1071/MF08188 Lindenmayer DB, Cunningham RB (1997) Patterns of co-occurrence among arboreal marsupials in the forests of central Victoria, southeastern Australia. Australian Journal of Ecology 22, 340–346. doi:10.1111/j.1442-9993.1997.tb00680.x

94

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Lindenmayer DB, Franklin JF (2002) Conserving Forest Biodiversity: A Comprehensive Multiscaled Approach. Island Press, Washington, DC. Lindenmayer DB, Gibbons P (Eds) (2012) Biodiversity Monitoring in Australia. CSIRO Publishing, Melbourne. Lindenmayer DB, Likens GE (2009) Adaptive monitoring – a new paradigm for long-term research and monitoring. Trends in Ecology & Evolution 24, 482–486. doi:10.1016/j.tree.2009.03.005 Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/s10021-010-9394-6 Lindenmayer DB, Craig SA, Linga T, Tanton MT (1991) Public participation in stagwatching surveys for a rare mammal – applications for environmental education. Australian Journal of Environmental Education 7, 63–70. doi:10.1017/ S0814062600001865 Lindenmayer DB, Cunningham RB, MacGregor C, Incoll RD (2003) A long-term monitoring study of the population dynamics of arboreal marsupials in the Central Highlands of Victoria. Biological Conservation 110, 161–167. doi:10.1016/ S0006-3207(02)00171-4 Lindenmayer DB, Likens GE, Franklin JF (2010) Rapid responses to facilitate ecological discoveries from major disturbances. Frontiers in Ecology and the Environment 8, 527–532. doi:10.1890/090184 Lindenmayer DB, Likens GE, Haywood A, Meizis L (2011a) Adaptive monitoring in the real world: proof-of-concept. Trends in Ecology & Evolution 26, 641–646. doi:10.1016/j.tree.2011.08.002 Lindenmayer DB, Wood J, McBurney L, Michael D, Crane M, MacGregor C, Montague-Drake R, Banks S, Gibbons P (2011b) Cross-sectional versus longitudinal research: a case study of trees with hollows and marsupials in Australian forests. Ecological Monographs 81, 557–580. doi:10.1890/11-0279.1 Lindenmayer DB, Likens GE, Andersen A, Bowman D, Bull CM, Burns E, Dickman CR, Hoffmann AA, Keith DA, Liddell MJ, Lowe AJ, Metcalfe DJ, Phinn SR, RussellSmith J, Thurgate N, Wardle GM (2012a) Value of long-term ecological studies. Austral Ecology 37, 745–757. doi:10.1111/j.1442-9993.2011.02351.x Lindenmayer DB, Zammit C, Attwood SA, Burns E, Shepherd CL, Kay GE, Wood J (2012b) A novel and cost-effective monitoring approach for outcomes in an Australian biodiversity conservation incentive program. PLoS One 7, e50872. doi:10.1371/journal.pone.0050872 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks S, Driscoll D, Smith A, Gill AM (2013a) Fire severity and landscape context effects on arboreal marsupials. Biological Conservation 167, 137–148. doi:10.1016/j.biocon.2013.07.028 Lindenmayer DB, MacGregor C, Dexter N, Fortescue M, Cochrane P (2013b) Booderee National Park management: connecting science and management. Ecological Management & Restoration 14, 2–10. doi:10.1111/emr.12027

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

95

Lindenmayer DB, Willinck E, Crane M, Michael D, Okada S, Cumming C, Durant K, Frankenberg J (2013c) Murray Catchment habitat restoration: lessons from landscape-level research and monitoring. Ecological Management & Restoration 14, 80–92. doi:10.1111/emr.12051 Lindenmayer DB, Barton PS, Lane PW, Westgate MJ, McBurney L, Blair D, Gibbons P, Likens GE (2014a) An empirical assessment and comparison of species-based and habitat-based surrogates: a case study of forest vertebrates and large old trees. PLoS One 9, e89807. doi:10.1371/journal.pone.0089807 Lindenmayer DB, Burns E, Thurgate N, Lowe A (Eds) (2014b) Biodiversity and Environmental Change: Monitoring, Challenges and Direction. CSIRO Publishing, Melbourne. Lindenmayer DB, Barton PS, Pierson J (Eds) (2015a) Indicators and Surrogates of Biodiversity and Environmental Change. CSIRO Publishing, Melbourne. Lindenmayer DB, Blair D, McBurney L, Banks S (2015b) Mountain Ash: Fire, Logging and the Future of Victoria’s Giant Forests. CSIRO Publishing, Melbourne. Lindenmayer DB, Burns EL, Dickman C, Green PT, Hoffmann AA, Keith DA et al. (2017) Save Australia’s ecological research. Science 357, 557. Lovett GM, Burns DA, Driscoll CT, Jemkins JC, Mitchell MJ, Rustad L, Shanley JB, Likens GE, Haeuber R (2007) Who needs environmental monitoring? Frontiers in Ecology and the Environment 5, 253–260. doi:10.1890/1540-9295(2007)5[253:WNEM]2.0.CO;2 Ludwig D, Mangel M, Haddad B (2001) Ecology, conservation and public policy. Annual Review of Ecology and Systematics 32, 481–517. doi:10.1146/annurev. ecolsys.32.081501.114116 Majer JD, Heterick B, Gohr T, Hughes E, Mounsher L, Grigg A (2013) Is thirty-seven years sufficient for full return of the ant biota following restoration? Ecological Processes 2, 19. doi:10.1186/2192-1709-2-19 Martin J, Kitchens WM, Hines JE (2007) Importance of well-designed monitoring programs for the conservation of endangered species: case study of the Snail Kite. Conservation Biology 21, 472–481. McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, Possingham HP (2010) Monitoring does not always count. Trends in Ecology & Evolution 25, 547–550. doi:10.1016/j.tree.2010.07.002 McInnes PF, Naiman RJ, Pastor J, Cohen Y (1992) Effects of moose browsing on vegetation and litterfall of the boreal forest, Isle Royale, Michigan, USA. Ecology 73, 2059–2075. doi:10.2307/1941455 Muhle J, Huang J, Weiss RF, Prinn RG, Miller BR, Salameh PK, Harth CM, Fraser PJ, Porter LW, Greally BR, O’Doherty S, Simmonds PG (2009) Sulfuryl fluoride in the global atmosphere. Journal of Geophysical Research, D, Atmospheres 114, D05306. Muller F, Baessler C, Schubert H, Klotz S (Eds) (2011) Long-term Ecological Research: Between Theory and Application. Springer, Dordrecht.

96

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Nichols JD, Williams BK (2006) Monitoring for conservation. Trends in Ecology & Evolution 21, 668–673. doi:10.1016/j.tree.2006.08.007 Norton DA (1996) Monitoring biodiversity in New Zealand’s terrestrial ecosystems. In Papers from a seminar series on biodiversity. (Eds B McFadgen, S Simpson) pp. 19–41. Department of Conservation, Wellington, New Zealand. Pastor J, Naiman RJ, Dewey B, McInnes P (1988) Moose, microbes, and the boreal forest. Bioscience 38, 770–777. doi:10.2307/1310786 Pastor J, Dewey B, Naiman RJ, McInnes PF, Cohen Y (1993) Moose browsing and soil fertility in the boreal forests of Isle Royale National Park. Ecology 74, 467–480. doi:10.2307/1939308 Pastor J, Dewey B, Moen R, White M, Mladenoff D, Cohen Y (1998) Spatial patterns in the moose-forest-soil ecosystem on Isle Royale, Michigan, USA. Ecological Applications 8, 411–424. Persson I-L, Pastor J, Danell K, Bergström R (2005) Impact of moose population density and forest productivity on the production and composition of litter in boreal forests. Oikos 108, 297–306. doi:10.1111/j.0030-1299.2005.13844.x Persson I-L, Nilsson MB, Pastor J, Eriksson T, Bergström R, Danell K (2009) Depression of belowground respiration rates at simulated high moose population densities in boreal forests. Ecology 90, 2724–2733. doi:10.1890/08-1662.1 Peters R (1991) A Critique for Ecology. Cambridge University Press, Cambridge. Pielke RA (2007) The Honest Broker: Making Sense of Science in Policy and Politics. Cambridge University Press, Cambridge, England. Porter JL, Kingsford RT, Brandis K (2016) Aerial survey of wetland birds in Eastern Australia – October 2016 annual summary report. Centre for Ecosystem Science, University of New South Wales. Available at: https://www.ecosystem.unsw.edu.au/ files/2016-Eastern-Australia-Waterbird-Survey-Summary-Report.pdf Pulsford S, Lindenmayer DB, Driscoll D (2016) A succession of theories: a framework to purge redundancy in post-disturbance theory. Biological Reviews of the Cambridge Philosophical Society 91, 148–167. doi:10.1111/brv.12163 Puntsag T, Mitchell MJ, Campbell JL, Klein ES, Likens GE, Welker JM (2016) Arctic Vortex changes alter the sources and isotopic values of precipitation in northeastern US. Scientific Reports 6, 22647. doi:10.1038/srep22647 Reynolds JH, Thompson WL, Russell B (2011) Planning for success: identifying effective and efficient survey designs for monitoring. Biological Conservation 144, 1278–1284. doi:10.1016/j.biocon.2010.12.002 Ringold PL, Alegria J, Czaplwski RL, Mulder BS, Tolle T, Burnett K (1996) Adaptive monitoring design for ecosystem management. Ecological Applications 6, 745–747. doi:10.2307/2269479 Rosenberry DO, Bukaveckas PA, Buso DC, Likens GE, Shapiro AM, Winter TC (1999) Movement of road salt to a small New Hampshire lake. Water, Air, and Soil Pollution 109, 179–206. doi:10.1023/A:1005041632056

3 – W h a t m a k e s l o n g -t e r m m o n i t o r i n g   e f f e c t i v e?

97

Russell-Smith J, Whitehead PJ, Cook GD, Hoare JL (2003) Response of Eucalyptusdominated savanna to frequent fires: lessons from Munmarlary 1973 - 1996. Ecological Monographs 73, 349–375. doi:10.1890/01-4021 Russell-Smith J, Edwards A, Woinarski J, Fisher A, Murphy B, Lawes M, Crase B, Thurgate N (2014) North Australian tropical savannas: the Three Parks Savanna Fire-Effects Plot Network. In Biodiversity and Environmental Change: Monitoring, Challenges and Direction. (Eds D Lindenmayer, E Burns, N Thurgate, A Lowe) pp. 335–378. CSIRO Publishing, Melbourne. Sato C, Wood J, Stein JA, Crane M, Okada S, Michael D, Kay G, Florance D, Seddon J, Gibbons P, Lindenmayer DB (2016) Natural tree regeneration in agricultural landscapes: the implications of intensification. Agriculture, Ecosystems & Environment 230, 98–104. doi:10.1016/j.agee.2016.05.036 Schindler DW, Mills KH, Malley DF, Findlay DL, Shearer JA, Davies IJ, Turner MA, Linsey GA, Cruikshank DR (1985) Long-term ecosystem stress: the effects of years of experimental acidification on a small lake. Science 228, 1395–1401. doi:10.1126/ science.228.4706.1395 Sergeant CJ, Moynahan BJ, Johnson WF (2012) Practical advice for implementing long-term ecosystem monitoring. Journal of Applied Ecology 49, 969–973. doi:10.1111/j.1365-2664.2012.02149.x Smith AP, Lindenmayer D, Begg RJ, Macfarlane MA, Seebeck JH, Suckling GC (1989) Evaluation of the stag-watching technique for census of possums and gliders in tall open forest. Australian Wildlife Research 16, 575–580. doi:10.1071/WR9890575 Spellerberg IF (1994) Monitoring Ecological Change. 2nd edn. Cambridge University Press, Cambridge. Strayer DL, Glitzenstein JS, Jones C, Kolasa J, Likens GE, McDonnell M, Parker GG, Pickett STA (1986) Long-term ecological studies: An illustrated account of their design, operation, and importance to ecology. Institute of Ecosystem Studies, Millbrook, New York. Swanson ME, Franklin JF, Beschta RL, Crisafulli CM, DellaSala DA, Hutto RL, Lindenmayer DB, Swanson FJ (2011) The forgotten stage of forest succession: early-successional ecosystems on forest sites. Frontiers in Ecology and the Environment 9, 117–125. doi:10.1890/090157 Taylor BL, Gerrodette T (1993) The uses of statistical power in conservation biology: the Vaquita and Northern Spotted Owl. Conservation Biology 7, 489–500. doi:10.1046/j.1523-1739.1993.07030489.x Walters CJ (1986) Adaptive Management of Renewable Resources. Macmillan Publishing Company, New York. Walters C (1997) Adaptive policy design: thinking at large spatial scales. In Wildlife and Landscape Ecology. (Ed. J Bissonette) pp. 386–394. Springer-Verlag, New York. Walters CJ, Holling CS (1990) Large scale management experiments and learning by doing. Ecology 71, 2060–2068. doi:10.2307/1938620

98

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Weathers KC, Strayer DL, Likens GE (Eds) (2009) Fundamentals of Ecosystem Science. Elsevier Academic Press, New York. Weiss RF, Muhle J, Salameh PK, Harth CM (2008) Nitrogen trifluoride in the global atmosphere. Geophysical Research Letters 35, L20821. doi:10.1029/2008GL035913 Welsh AH, Cunningham RB, Chambers RL (2000) Methodology for estimating the abundance of rare animals: seabird nesting on North East Herald Cay. Biometrics 56, 22–30. doi:10.1111/j.0006-341X.2000.00022.x Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive management of biological systems: a review. Biological Conservation 158, 128–139. doi:10.1016/j. biocon.2012.08.016 Whitlock MC (2011) Data archiving in ecology and evolution: best practices. Trends in Ecology & Evolution 26, 61–65. doi:10.1016/j.tree.2010.11.006 Winter TC, Likens GE (Eds) (2009) Mirror Lake: Interactions among Air, Land and Water. University of California Press, Los Angeles. Woodward A, Jenkins K, Schreiner EG (1999) The role of ecological theory in longterm monitoring: report on a workshop. Natural Areas Journal 19, 223–233. Youngentob K, Likens GE, Williams JE, Lindenmayer DB (2013) A survey of long-term terrestrial ecology studies in Australia. Austral Ecology 38, 365–373. doi:10.1111/j.1442-9993.2012.02421.x

4 The problematic, the effective and the ugly – some case studies

The previous two chapters outlined some of the characteristics of what we considered to be poor or failed monitoring programs (Chapter 2) and good or effective monitoring programs (Chapter 3). In this chapter, we reinforce some of these points through a series of case studies. The case studies cover terrestrial, marine, and inland aquatic landscapes; they come from a range of places (three continents); and they represent both mandated monitoring and monitoring programs driven by investigator curiosity. They include environmental monitoring (e.g. nutrients and water flows) as well as monitoring of populations of particular organisms. In each case study, we briefly describe the location of the work, the broad aims of the study, some of its important negative and positive features, and an occasional anecdote. These case studies, selected from a large number that could have been highlighted, illustrate the range of features we deemed important within the scope of this small book. We wrote this chapter with great trepidation because of the potential to deeply offend some people, including many friends and colleagues who are truly outstanding scientists. (We expand on this in Box 4.1.) It would have been far easier to leave this chapter out of the book entirely! However, we felt it was necessary to include it for three primary reasons. First, we wish to illustrate with clear and

100

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 4.1. Chapter 4 trepidation Writing this chapter was quite difficult and certainly far from ‘career-enhancing’ because we have been critical of a major province-level program (in Alberta, Canada) and several national level programs (the NEON and TERN initiatives in the United States and Australia respectively, and EMAP, now morphed into the National Aquatic Resource Surveys (NARS) program in the United States (S Lee, pers. comm.). We had concerns about these programs for several reasons. First, they were characterised at the outset by either a lack of focused questions or only very general questions that may not readily be tractable from a scientific perspective. Second, in the case of NEON and TERN, they are ecological and environmental infrastructure initiatives with questions ‘retro-fitted’ after the infrastructure had been decided upon, leading to a risk that the wrong infrastructure may be targeted for investment. Third, they generally lacked a conceptual model directly related to ecosystem function, a robust underpinning of experimental design, or both. Each of these programs may, in fact, prove to be successful and exciting discoveries may well emerge from them. However, we do suggest that the lack of the key ingredients of successful monitoring programs – motivating questions, a conceptual model and a robust experimental design – are nagging legacies that may lead to considerable risks of mistakes being made and large sums of money being used inefficiently and ineffectively. These deficiencies may not make ecologists look good in the eyes of the public and could endanger future funding for monitoring. Few long-term monitoring programs or funding initiatives can afford such inefficiencies. This was the basis for our concerns and we argue throughout this book that such key characteristics as well-developed questions and an appropriate experimental design are fundamental to effectiveness and success of monitoring programs.

tangible examples what features we think underpin good long-term research and monitoring programs and those that are poor or that fail. Second, we wanted to highlight some aspects of studies for which we ourselves have been personally responsible by showing that they are far from perfect. The third reason we felt compelled to write this chapter was to demonstrate ways that long-term research and monitoring programs could be improved. We suggest this latter goal is critical if the scientific and management communities were to address the mind-boggling number of major environmental problems facing humanity. In fact, this goal underlies the reason why we wrote this book (see Chapter 1). We readily acknowledge that our assessments of the case studies arise from our value judgements, but then this book is based on our personal perspectives. We have attempted to make our criticisms carefully because we know that the proponents of these activities are well intentioned. Our definition of an ineffective or problematic case study was that it was not particularly relevant on a cost-benefit

4 – The problematic, the ef fective and the ugly – some case studies

101

analysis to management, provided inadequate or misleading answers to key resource management problems, or was not scientifically productive. We argue that these problems mean the work breaches many of the principles that we outline as being important in the adaptive monitoring framework. We note, however, that all the case studies of ineffective monitoring have some good points and we suggest there are important opportunities to improve them. Indeed, if the commentary we have provided on any of the case studies were to lead to improvements in any of the monitoring programs we have outlined, then we would feel that such an outcome was an important achievement and worth the angst.

The problematic PPBio Australasia www.griffith.edu.au/ppbio, http://ppbio.inpa.gov.br Program for Planned Biodiversity and Ecosystem Research (PPBio) is a standardised research and monitoring platform designed to promote the collection, storing and sharing of biological information (Hero et al. 2010, 2014). The documentation accompanying the PPBio sites in Queensland indicates that an array of measurements can be made using this standardised platform, including vegetation composition and biomass, soils, hydrology, amphibians, birds, vegetation structure, reptiles, leaf litter, mammals, spiders and invertebrates. It was suggested that vascular plants, birds, and beetles would be ‘indicator taxa’ of value for monitoring (Hero et al. 2014). The PPBio approach was implemented in Brazil (Magnusson et al. 2005) and then adapted for use in Queensland, Australia (Hero et al. 2010; Lollback et al. 2015). There is also an extension to Chitwan National Park in Nepal (Hero et al. 2010). The PPBio method is based on a 5 km × 5 km grid of survey trails within which is located a grid of 30 permanently marked 1-ha plots. A PPBio grid has been established at Karawatha Forest Park in the city of Brisbane, Queensland, comprising 33 plots spaced at 1-km intervals (Hero et al. 2013). Other PPBio grids with the plots set out at 1-km intervals have been established at Lake Broadwater (Hero et al. 2010) and Currawinya National Park is south-western Queensland (Lollback et al. 2015). One of the stated aims of PPBio is to establish PPBio grids in all the major eco-regions of Australia, make these areas akin to Long-Term Ecological Research (LTER) sites such as those established in the United States, allowing for ‘comparable assessment of biodiversity indicators across a range of spatio-temporal scales’ (Hero et al. 2010). Such assessments include monitoring of the effects of climate change on biodiversity. This is an important potential advantage of the PPBio effort because of the great difficulties in harmonising data and questions across different long-term studies that have been designed and implemented for different reasons.

102

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

E1

D2

B3

C3

E2

D4 C4

B4

C5

G3

E3

D3

F3

F4

E4

F5

E5

D5

G5

B5 A6

D6

C6

B7

C7

F6

E6

D7

G6

E7

A7

A8 0

300

600

Metres

N

Plots 10m contours Karawatha Forest Park

Fig. 4.1.  The layout of a PPBio grid at Karawatha Forest Park with survey points spaced at a set interval of 1-km grid squares (redrawn from Hero et al. 2010)

The aims of the PPBio approach are noble and the champions of the method are very well intentioned and enthusiastic. However, we have concerns that the approach suffers from some of the deficiencies we have outlined for the Alberta Biodiversity Monitoring Program (ABMP), later in this chapter. Paucity of questions From our reading of documents available to us, the PPBio approach is not guided by well-defined and testable questions about ecosystem function or the dynamics of populations. The design of the work is not characterised by key contrasts in environmental conditions, which would allow inferences to be made about the effects of an intervention in one place versus no intervention in another. Thus, the ability to quantify the management relevance is questionable.

4 – The problematic, the ef fective and the ugly – some case studies

103

Lack of conceptual model PPBio appears to lack a conceptual model that lays out how a targeted ecosystem might function or how it might respond to intervention. Data gathered from the approach may therefore not assist in resolving the reasons or mechanisms for any changes that might occur in biodiversity or ecosystem processes. One size fits all PPBio recommends a highly-standardised survey framework for all sites and plots within sites (Fig. 4.1). This framework assumes that the same general survey protocol will be appropriate for the array of entities proposed to be measured in PPBio. However, this assumption may be readily violated. The spatial scale might be the correct one for small reptiles, but inappropriate for more mobile species that range over larger areas, like birds. A related issue is that a highlystandardised protocol assumes that it is appropriate to measure precisely the same list of entities at all places. We have outlined in Chapter 2 why this is problematic and suggested that the entities selected for measurement should be guided by the questions being asked. For example, surveys of reptiles may be useful in a highly reptile-rich region like south-eastern Queensland where responses to fire might be a critical aspect of the environment, but they may be of limited value in other jurisdictions like tropical Amazonia where other ecological features, like forest structure and plant species composition, have a far greater impact on biodiversity and biotic responses. Statistical design A lack of questions means that it is difficult to assess the efficacy of the experimental design. However, we suggest there may be some problems with the design. First, the spatial clustering of plots in the standardised grid risks pseudoreplication (Hurlbert 1984) for some groups (e.g. birds). Second, in highly variable ecosystems such as the ones being studied in south-eastern Queensland, a survey comprised of ~30 plots may have limited statistical power to detect trends or effects, particularly when the survey design is not underpinned by clear contrasts such as places (plots) where a management intervention has been implemented versus places where they have not (see Cunningham and Lindenmayer 2017). Third, there appears to be limited stratification to guide plot selection other than space – the plots are distributed on a 1 km × 1 km grid spacing. The PPBio approach is undoubtedly an interesting one and is strongly motivated by frustrations that biodiversity monitoring worldwide has been poor and uncoordinated (Hero et al. 2010). Despite what we consider to be serious problems resulting from a paucity of questions and the overall survey design, PPBio has some positive aspects including limited bureaucracy, well-documented protocols for field sampling (Hero et al. 2014), many well-developed national and

104

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

international partnerships including those with the lay community, and an enthusiastic and highly dedicated advocate who has been instrumental in publishing some high-quality scientific articles from the PPBio initiative (Hero et al. 2010, 2013; Lollback et al. 2015). The approach is also relatively inexpensive (at this stage of its development), particularly in comparison with other similar kinds of one-size-fits-all methods for long-term ecological research and monitoring programs such as ABMP.

The Alberta Biodiversity Monitoring Program (ABMP) The Alberta Biodiversity Monitoring Program (ABMP) is run by the Alberta Biodiversity Monitoring Institute (ABMI) in Alberta, Canada (Alberta Biodiversity Monitoring Institute 2016). Alberta is a large, resource-rich province in western Canada that supports globally significant forestry and energy industries (Boutin et al. 2009) and is home to more than 70 000 species of native vertebrates and invertebrates. Early documents about the program suggest that ABMP aims to track changes in ~2000 of these species or ~3% of the vertebrate and invertebrate taxa estimated to live in the province. Changes in biodiversity will be tracked at two spatial scales; at individual field sites and at large, spatial scales (presumably at landscape scales) using aerial photography and satellite imagery (Alberta Biodiversity Monitoring Institute 2009; Alberta Biodiversity Monitoring Program 2009). The ABMP aims to not only quantify biodiversity responses to land use but also undertake predictive modelling of changes in the relative abundance of species (Sólymos et al. 2015). Background information for ABMP indicates that it aims to monitor 1656 sites at points evenly spaced in a 20 × 20-km grid across the entire province (Fig. 4.2). The list of entities to be monitored is large, as shown in Table 4.1. The approach aims to survey the groups and attributes in Table 4.1 at 350 sites annually and hence visit all 1656 sites every five years (Alberta Biodiversity Monitoring Institute 2009; Alberta Biodiversity Monitoring Program 2009; Sólymos et al. 2015). Examination of relatively recently published articles (e.g. Sólymos et al. 2015) suggests that approximately half of the 1656 sites had been surveyed between 2003 and 2013 (see Fig. 4.2). Background information indicates that the initiative has a well-developed governance structure (Alberta Biodiversity Monitoring Institute 2016) and data management protocol (Sólymos et al. 2015). A project as large and ambitious as this is costly and in 2009 the annual operating cost was estimated to be approximately C$10 million. Some strong claims were made about the ABMP at its inception, such as that it ‘is perhaps the most advanced biodiversity monitoring program in the world’ (Alberta Biodiversity Monitoring Program 2009). We remain unsure about these claims because we consider ABMP to have some problems, which we outline below.

4 – The problematic, the ef fective and the ugly – some case studies

105

Fig. 4.2.  The survey design for the Alberta Biodiversity Monitoring Institute (redrawn from Sólymos et al. 2015)

Paucity of questions The initial documentation we read for ABMP did not contain any information on the questions that were being posed. Thus, it appears that ABMP, at least at its inception, was not driven by well-defined and testable questions. As outlined elsewhere in this book, it is difficult to gauge progress on management

Invertebrates

Bryophytes

Vascular plants

Macro-invertebrates

Mites

Springtails

Zooplankton

Vascular plants

Vascular plants

Benthic algae

Phytoplankton

Birds

Birds

Fish

Fungi

Mammals

Winter mammals

Species in lakes and rivers

Lichens

Wetland species

Terrestrial species

Table 4.1.  Entities targeted for monitoring in the ABMI project

Habitat structure

Trees, snags, coarse woody debris

Physical and ecological characteristics

Terrestrial habitat

Water physiochemistry

Wetland size and depth

Wetland habitat

Patch size distribution

Diversity of habitat types

% composition of habitat types

Landscape habitat

Linear features

Energy footprint

Forestry footprint

Agricultural footprint

Urbanisation

Human footprint

106 E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

4 – The problematic, the ef fective and the ugly – some case studies

107

interventions when there are no clearly defined or tractable scientifically-based questions being asked to guide a project. In addition, the statistical design of the project (see below) may not be an appropriate one for answering particular questions after the spatial array of monitoring sites had been implemented. Statistical design The design of ABMP is an interesting one that is somewhat unusual for a monitoring program. It is not possible for us to assess the design critically because there are no clearly defined questions against which to determine its suitability. It appears that the ‘stratification’ is based on space (Fig. 4.2); that is, the sites are distributed evenly across the province (Fig. 4.2). In some cases, the distribution of key organisms and attributes of interest (see Table 4.1) may be strongly influenced by other factors, which seem to us to be perhaps better ones on which to stratify. For example, the age of the forest and level of logging disturbance might be among the factors more likely to influence terrestrial boreal forest taxa (Burton et al. 2003). Hence, some direct assessment of trend patterns and logging impacts on those species might be better determined from contrasts between sites before and after they are logged and matched with reference sites that remain unlogged (a so-called BACI – before-after-control-impact – design). In addition, the spatial design of the sites makes it likely that rare environments will be represented by few replicates. However, these rare environments may well support rare species of conservation and/or management concern for which it may be critically important to assess their status over time (Burton et al. 2003). Notably, to sample these rare habitats ABMP contains off-grid sites in addition to the 1656 evenly spaced on-grid sites. The frequency of surveys (every five years at a given [on-grid] site) also may be a problem as it might not be well matched to quantify some of the key ecological phenomena of interest in the ecosystems of Alberta, such as outbreaks of beetles that kill large areas of forest (Raffa et al. 2008; Government of British Columbia 2011) or the population cycles of animals (Krebs 2012; see also Box 1.4 in Chapter 1). Passivity ABMP is probably best defined as a passive monitoring program (see Chapter 1); that is, it is largely aimed at tracking change over time in the province of Alberta. Hence, the project is not characterised by management interventions – which we consider are important for the effectiveness of monitoring programs (see Chapter 3), particularly for documenting why changes have occurred. In fact, early documentation for ABMP notes that ‘the intent of monitoring is not to understand why the change has occurred’ (Shank et al. 2002). This approach is puzzling and make it difficult to identify the mechanisms underlying the trend patterns that may emerge from the monitoring program. It also may curtail predictive ability of any findings and hence possibly limit the potential for proactive

108

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

environmental management. Indeed, this approach is a particular concern given that one of the objectives of ABMP is that the datasets from the initiative should be ‘an integral part of predictive modelling of species’ relative abundances’ (Sólymos et al. 2015, p. 473). Laundry list ABMP is a very ambitious initiative and aims to survey a large number of things (Table 4.1). Our concern here is the risk of a large number of things being measured poorly rather than a few things measured and analysed well. There is also an assumption that the statistical design and the spatial scale of sites will be appropriate for all of the target groups, ranging from zooplankton to wide-ranging winter mammals. We worry that this assumption may not hold. Moreover, although the governance structure for ABMP is seeking to secure a long-term endowment, which is laudable, the very large number of sites (many of them remote; see Fig. 4.2), coupled with the very large number of entities being measured, could eventually be financially untenable and ultimately lead to the collapse of the monitoring program. From examining the ABMP website, we note that between 2003 and 2009 in the Lower Athabasca Region in Alberta (the hotspot for oil sands operations), 68 of the proposed 235 permanent survey sites were visited. Thus, it appears to us that ABMP may currently lack the resources to undertake the proposed monitoring protocols for one of the most significant ecological regions in the province of Alberta. This potential deficiency is a particular concern given the recent changes in federal legislation to go ahead with the construction of the Keystone Pipeline to move oil from the tar sands in Alberta to the United States. Lack of conceptual model We have argued that a conceptual model of how an ecosystem works is a key part of our effective monitoring and a pivotal part of adaptive monitoring (see Chapter 3). However, we could find no conceptual model for ABMP (e.g. Boutin et al. 2009), although this oversight is common to many monitoring programs, including some of our own (see the case study on Victorian forests later in this chapter). The lack of a conceptual model, coupled with a paucity of a priori questions and management interventions (or other kinds of interventions), further suggests to us that ABMP has many of the characteristics of both mandated and passive monitoring programs (see Chapter 1). We therefore reiterate our concern that the status of particular entities will be documented at particular points in time, but there may be limited opportunity to determine why those entities have changed and the ecological mechanism influencing the emergent pattern may not be discovered. Complex aggregation issues One of the aims of ABMP is to produce indices that reflect the state of biodiversity in Alberta (Alberta Biodiversity Monitoring Program 2009). For

4 – The problematic, the ef fective and the ugly – some case studies

109

example, an ‘intactness’ index was created to reflect the status of biodiversity for particular regions in Alberta (Lamb et al. 2009). These kinds of indices can be useful for generating reports and communicating to politicians and government officials, especially those who want ‘one number’. However, such indices need to be constructed very carefully and the assumptions underlying them need to be well understood and clearly articulated (Lamb et al. 2009). Indeed, there are vexed issues with creating a composite, scientifically robust index from an amalgam of sub-indices (e.g. see Lindenmayer et al. 2015b,e) and the results can sometimes be misleading (Wright-Stow and Winterbourn 2003). Moreover, it is sometimes not clear how coarsely-scaled, summary data can be used to identify trends for biodiversity or how these rather crude trends would be useful to assist on-the-ground resource management, such as in guiding targeted management interventions to solve a particular environmental problem or protect a rare species. We acknowledge that our assessment of ABMP is somewhat critical and some of it may not be justified or appropriate as the project has continued to develop and mature (see Box 4.2) (Alberta Biodiversity Monitoring Institute 2016). Moreover, we do see some positive aspects to the initiative. First, the proponents of ABMP have clearly given considerable thought to ways of generating sufficient long-term funding to maintain the initiative (Boutin et al. 2009). Few monitoring programs do that at the beginning. Second, although the governance structure seems large and possibly unwieldy (Alberta Biodiversity Monitoring Institute 2016), the approach to bring together many stakeholders and forge partnerships is a valuable one. Third, while ABMP has the potential to create an extremely valuable set of baseline datasets (Sólymos et al. 2015), we do suggest there might be a distinct opportunity for adding carefully selected sites to allow for rigorous quantification of the mechanisms giving rise to key temporal trends in biodiversity. Fourth, the ABMI responsible for the ABMP has worked extremely hard to curate and manage the very large amounts of data that are generated from the work (Sólymos et al. 2015). Careful management of this very large dataset is a difficult and timeconsuming task, but these efforts should pay off in the future. Our hope is that the perspectives we have set out above might stimulate discussion about many topics such as the questions being posed, the experimental design underpinning ABMP and the large array of species and attributes being measured. Lively debate is critical to improving monitoring programs, as it is for advancing any form of credible science.

EMAP In 1990 the US Environmental Protection Agency (EPA) created EMAP, the Environmental Monitoring and Assessment Program, in its Office of Research and Development (ORD) (Messer et al. 1991; McDonald et al. 2004). This initiative followed discussions started in 1988 in response to increasing pressure by the US

110

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 4.2. Rejoinder: an alternative set of perspectives of ABMP Writing our assessment of ABMP made us distinctly more nervous than any other part of this book, with the possible exception of NEON and TERN. It led to much trepidation, not the least because we have had collaborative working relationships with several of the advocates of ABMP and there are many excellent scientists associated with the program. Correspondence with one of the scientists in the program has indicated there are some important alternative perspectives to those we have set out in this chapter. Indeed, ABMP is a truly novel and different approach to monitoring that is quite unlike any other we have seen to date. We touch on some of these below because in no way do we wish to see our criticisms lead to ABMP being irreparably damaged and then terminated. Rather, we believe that it is critical for Alberta to have an effective monitoring program. First, a lack of stratification in ABMP (other than spatially) was deemed important to provide an unbiased sample of the province and also provide sufficient sites in any given region to allow suitable regional reporting of trends (Sólymos et al. 2015). The approach is also considered appropriate for detecting shifts in the distributions of species with climate change, although a recent review suggests that more sampling of ecotonal areas such as between parkland and boreal forest regions would be required to resolve climate-change driven changes in species distributions (see Bayne et al. 2015). In addition, it was believed that the approach would not foreclose future options for monitoring environmental changes that are presently not foreseen. However, we reiterate our concerns that current spatial stratification for ABMP is statistically unusual and may not be the most statistically and ecologically efficient study design. Indeed, such inefficiency would be a concern when one of the aims of ABMP is to ‘reduce costs and redundancy in provincial environmental monitoring’ (Sólymos et al. 2015, p. 473). Second, an essentially passive approach was selected as a potential antidote to designed monitoring programs that fail the test of management relevance, including being conducted at a scale that is not meaningful for detecting cumulative effects (such as those across entire landscapes). Furthermore, a passive approach was advocated because it was not clear what future problems might arise that would need to be monitored. Nevertheless, we still have concerns that the apparent passive approach to monitoring might not produce the desired results, especially in terms of diagnosing the reasons for observed changes. Third, the full suite of entities initially proposed for monitoring since the first edition of this book (see Table 4.1) has been pared down to make the monitoring effort more tractable (Alberta Biodiversity Monitoring Institute 2016), given the practical realities of surveying numerous remote sites. This reduction is important because of an appreciation for not trying to monitor too many things with the potential of doing them badly. We view this reduction as a very positive step in ABMP.

4 – The problematic, the ef fective and the ugly – some case studies

111

Government for ‘measuring for environmental results’ (Reilly 1989). This monitoring program, designed to cover the entire United States, was formulated largely because of the need for large-scale environmental information on trends in the condition of ecological resources ranging from forests and agroecosystems to lakes, rivers and streams (Stevens 1994). The consequences of the absence of such data on the number and percentage of acidified lakes and streams in the United States became evident during the ‘acid rain wars’ period (Likens 1992; Oreskes and Conway 2010) in the 1980s. Although this large and complex program has now been replaced with the National Aquatics Resource Surveys (NARS) program, some lessons from EMAP are worth reviewing here. EMAP was the brainchild of Dr Rick Linthurst, an enthusiastic and forwardthinking employee of the US EPA, who had been heavily involved in trying to resolve the acid rain issue. The emphasis on measuring for results that could be applied (Jackson and Paulsen 2009), the experience with acid rain policy, and the broad scope of the program meant that EMAP (e.g. see Whittier et al. 2002) was born already carrying a heavy mantle of bureaucracy. EMAP was described as a series of ‘research programs to develop tools to monitor and assess the status and trends of the nation’s ecological resources’ (www. epa.gov/emap). Although these are laudable objectives, we found there were some substantial problems with the original configuration and operation of EMAP. Indeed, EMAP was scaled back by the EPA during the 1990s (Jackson and Paulsen 2009) and then replaced with the NARS program (e.g. Hobbs et al. 2015; United States Environmental Protection Agency 2015). Some of the monitoring concepts, such as the sampling design, and many of the methods were carried over from EMAP to NARS (S Lee, pers. comm.). The NARS program is a collaborative effort among the EPA, states and native tribes, and has four major components: ●● ●● ●● ●●

National Coastal Condition Assessment (NCCA) National Lakes Assessment (NLA) National Rivers and Streams Assessment (NRSA) National Wetland Condition Assessment (NWCA).

Lack of focused questions and a conceptual model The general objectives that drove the EMAP monitoring networks from the beginning were: ●●

●●

‘To estimate current status, extent, changes, and trends in indicators of the Nation’s ecological resources on a regional basis with known confidence’ ‘To monitor indicators of pollutant exposure and habitat condition, and to seek correlative relationships between human-induced stresses and ecological condition that identify possible causes of adverse effects’

112

●●

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

‘To provide periodic statistical summaries and interpretive reports on ecological status and trends to the EPA Administrator and the public.’ (Messer et al. 1991).

Although these questions were quite vague, Messer et al. (1991) stated that EMAP was designed ‘to answer critical questions for policy and decision makers and the public’. Examples of the more detailed questions included: ●●

●●

●● ●●

●●

What is the current extent and location of our ecological resources (e.g. estuaries, lakes, forests and wetlands) and how are they distributed geographically? What percentages of the resources appear to be adversely affected by pollutants and other man-induced environmental stress, and in which regions are the problems most severe or widespread? Which resources are degrading, where and at what rate? What are the relative patterns and magnitudes and the most likely causes of adverse effects? Are the proportions of adversely affected ecosystems improving as expected in response to control and mitigation programs?

These still rather vague questions became more specific for each category as the program evolved. EMAP and its successors were able to answer some of these questions for coastal waters, lakes and wadeable streams (Whittier et al. 2002) as well as for some large rivers and wetlands (www.epa.gov/owow/monitoring/ nationalsurveys.html). NARS was designed to answer questions such as: ●● ●● ●● ●●

What percent of waters support healthy ecosystems and recreation? What are the most common water quality problems? Is water quality improving of getting worse? Are investments in improving water quality focused appropriately?

These questions remain somewhat vague with embedded vague concepts, such as ‘healthy ecosystems’, whose undefined meaning is unclear. EMAP was not initiated around a formal conceptual model, although many scientists and consultants had been employed to help think through the development of the initial program, particularly the statistical design. Because of the absence of a conceptual model, EMAP attempted to monitor a very large number of environmental entities, a laundry list, of ‘sentinel’ species, ‘keystone’ species and a ‘suite of indicators’ plus numerous abiotic variables in a large number of natural resources, both terrestrial and aquatic, all with an elaborate statistical design for detecting spatial and temporal trends in these variables (e.g. McDonald et al. 2004). EMAP developed an Indicator Development Strategy, which focused on physical stressors, chemical stressors and biological stressors (Messer et al.

4 – The problematic, the ef fective and the ugly – some case studies

113

1991). The EMAP strategy was to identify and measure suites of indicators in all categories for each ecosystem type. The suite of response indicators was planned to reflect any actual adverse effects of both anticipated and unanticipated stressors (e.g. new pollutants entering the environment). The focus on biotic indicators to assess condition, rather than environmental conditions, was rather unusual at the time, but reflected the goals of EMAP. Statistical design One of the strengths of, yet major problems with, EMAP was its elaborate and rigid statistical design, which we suggest was based on an unrealistically designed set of polygon-shaped spatial units. EMAP planned to focus on ‘the development of unbiased statistical sampling frameworks’. Moreover, EMAP stated that, ‘Probability-based sampling within a statistical survey design … provides the only unbiased estimate of the condition on an aquatic resource over a large geographic area from a small number of samples’ (McDonald et al. 2004). The return interval for sampling was every four years. Nevertheless, a main justification for these polygon-shaped spatial units was that they would evenly cover the curved surface of the Earth. A systematic grid of sampling points was established across the United States and continental shelf forming the basis for regional sampling at ~12 500 points in the contiguous 48 states; each point was ~27 km apart in a triangular grid (Messer et al. 1991). After around 1994, EMAP focused on the value of the probability-based survey designs rather than the value of the hexagonal-grid design. Management relevance Although EMAP operated for many years, relatively few State of the Environment reports were issued (www.epa.gov/roe/), and the EMAP surveys often did not contribute to these reports. Thus, its relevance for advising the nation about environmental issues appears to have been significant in some areas (Whittier et al. 2002), but overall rather limited given the size and expectations of the program. The components of the NARS program have issued the following reports as part of their assessments: ●● ●● ●●

●●

National Coastal Condition Assessment (results and report in 2010) National Lakes Assessment (results in 2007 and 2012) National Rivers and Streams Assessment (results in 2008–09; wadeable streams assessment in 2004) National Wetland Condition Assessment (results and report in 2011).

The website (www.epa.gov/national-aquatic-resource-surveys) gives a listing of publications in scientific journals (NCCA and Great Lakes = 6; NLA = 13; NRSA = 26 and NWCA = 3), but there is no indication as to whether this is a complete listing.

114

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Loss of leadership Although Linthurst continued to be a strong champion for EMAP as Associate Director for Ecology at one of the EPA’s National Research Laboratories after he left his post as head of EMAP in 1996, in reality the passionate leader and champion for funding was lost and the program became immersed in the federal bureaucracy of the US EPA. EMAP suffered from a string of directors with relatively short tenures and loss of funding, both potentially lethal conditions to any long-term monitoring program. Additionally, with the change in federal administrations in the mid-1990s, EMAP’s US$43 million/year budget was reduced to US$23 million/year with over half of that directed to a new competitive grants program. Partnerships Clearly, EMAP was trying to do the right thing (see Box 4.3) with many viable partnerships, including a variety of other federal agencies, state agencies and native tribes, which greatly expanded its already very substantial source of funding from the US Government. The goal of EMAP from the beginning was to integrate data from all of these networks (Messer et al. 1991), which was laudable, but in practice was not very successful. According to EPA’s 2009 Annual Performance Plan and Congressional Justification, EMAP was to be ‘decreased’ in 2009 as the national aquatic surveys were to be transferred to other program offices, states, and native tribes, but with EMAP oversight and support, named the NARS Program (see above and also Jackson and Paulsen 2009). In reality, there was no more research into the design of monitoring, and ORD tended to view monitoring per se as outside of its mission.

The effective Rothamsted https://www.rothamsted.ac.uk/ It is not possible to write a book about long-term ecological research and monitoring without including arguably the most famous, and certainly the most long-running project of all: Rothamsted in England (Fig. 4.3, Fig. 4.4). Research at Rothamsted is conducted at two sites, one at Harpenden in Hertfordshire and a second near Bury St Edmunds in Suffolk. Rothamsted was established in 1834 and is the oldest active agricultural research station in the world (Rothamsted Research 2009). The two founders of Rothamsted, John Lawes and Henry Gilbert, established nine field experiments between 1843 and 1856 and only one of these was abandoned (in 1878). The remaining eight experiments continue to this day, the longest running field experiments in the world (Rothamsted Research 2006). Some of the classic experiments have been modified over time to address new

4 – The problematic, the ef fective and the ugly – some case studies

115

Box 4.3. Rejoinder: EMAP EMAP was a major monitoring program in the United States that produced more than 1000 research articles (Jackson and Paulsen 2009), but it also was a monitoring research and development program as part of the EPA’s Office of Research and Development. As such, it attempted to bring new approaches to monitoring at large, spatial scales. The statistical approach developed by EMAP has been adopted and endorsed by EPA’s Office of Water Coastal Condition Reports (United States Environmental Protection Agency 2015); Wadeable Streams Assessments and National Lakes Assessment (Kaufmann et al. 2014); (see www.epa. gov/owow/monitoring/nationalsurveys.html). Despite what we consider to be shortcomings in a long-term monitoring program based on our criteria for success, we acknowledge EMAP’s contributions to the development of environmental monitoring and particularly to assessment protocols (Kaufmann et al. 2014). It may not be obvious to readers from what we have written, but there were questions generated at the beginning, although we think these questions were not as focused and scientifically rigorous as we would have liked to guide a massive program such as EMAP. The questions were general and directed at only management assessments. The EPA established performance measures in its strategic plan under the Government Performance Results Act for streams and coastal waters based on EMAP-designed surveys. It was contended that a conceptual model had been proposed for the EMAP monitoring program (see Fig. 1 in Messer et al. 1991), but we have reservations about this model for several reasons. The primary one is because it is based on three categories of what we consider are indirect environmental indicators: response indicators, exposure indicators and stress indicators. Using indicators instead of direct measures is not only problematic (Lindenmayer and Likens 2011), but this approach falls short of using a broad, conceptual framework to guide understanding of ecosystem function that we propose as being critical in driving successful monitoring efforts (see Chapter 1).

questions or to add further treatments, as would be appropriate under the adaptive monitoring paradigm we describe in Chapter 3. Considerable long-term monitoring has been conducted at Rothamsted and some of the world’s truly classical experiments have been done there. They span studies of the effects of fertiliser application on crop yields, methods of crop improvement, pest resistance to agricultural chemicals, long-term changes in soil fertility associated with different cropping systems, and GM crop resistance to insect pests (Cressey 2015). One of the most extraordinary studies resulted from analyses of levels of radioactivity from immaculately archived samples gathered over many years. These provided values for ‘pre-atomic’ baselines or background estimates of radioactivity for comparison with samples from the post-atomic era.

116

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 4.3.  Aerial view of agricultural treatments at Rothamsted, England (Photo courtesy of the Lawes Agricultural Trust)

There are many reasons for the successes of the work at Rothamsted, and we briefly summarise some of them here. Well-defined questions The classical experiments at Rothamsted (Rothamsted Research 2006), as well as the many subsequent studies conducted there are motivated by well-articulated and testable research questions about sustainable agriculture. Statistical design The work at Rothamsted is characterised by outstanding experimental design that reflects the value of scientists and statisticians working collaboratively. In many ways, the experiments at Rothamsted were the birth of modern statistics with luminaries such as RA Fisher being responsible for analysing major datasets gathered there and subsequently articulating key principles of experimental design and statistical analyses (Fisher 1925, 1935; Salsburg 2001). Notably, Fischer identified several substantial design issues from the first decades of experimentation at Rothamsted. These were addressed in subsequent studies such as through the application of what were then novel approaches like replication and

4 – The problematic, the ef fective and the ugly – some case studies

117

Fig. 4.4.  Archiving datasets at Rothamsted, England (Photo courtesy of the Lawes Agricultural Trust)

randomisation (Fisher 1925, 1935). The Lawes Agricultural Trust, which provides some of the financial support for the work at Rothamsted, also maintains support for the sophisticated and widely used statistical package GENSTAT. Long-term financial stability Access to ongoing funding has been part of the success of the Rothamsted research. In 1889, John Lawes established a trust with £100 000 to fund the ongoing agronomic experiments. This endowment is administered by trustees and a trust committee. Leadership Long-term stable leadership lies at the core of the work at Rothamsted. John Lawes worked at Rothamsted for 50 years. Since then, subsequent directors have typically had long tenures. Adaptive monitoring Work at Rothamsted has been long-term, but nevertheless highly responsive to new issues of management relevance. New questions have been posed in response to the

118

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

insights gained from earlier findings, but these have not breached the integrity of the initial experiments. Therefore, existing projects have evolved and new ones have been instigated. This approach is highly congruent with the adaptive monitoring framework we describe in Chapter 3. For example, current work at Rothamsted includes studies of genetically modified crops and their ability to repel insect pests (Cressey 2015) and carbon sequestration and climate change (Rothamsted Research 2009) – topics that could not have even been comprehended by the founders of the work 175 years ago. Obviously, the evolution of existing projects, coupled with new ones, add considerably to the overall breadth and importance of work at Rothamsted and the resulting body of knowledge.

Ecosystem Health Monitoring Program (EHMP) for Moreton Bay in South East Queensland, Australia http://hlw.org.au/ Moreton Bay (Fig. 4.5) is a large embayment in south-eastern Queensland, bordering the city of Brisbane and other urban areas, and receiving drainage through the Brisbane, Logan, Pine, Caboolture and other smaller rivers from a catchment of some 19 430 km2 (Dennison and Abal 1999). The human population of the region is 3.4 million (as of 2016) and has grown very quickly over the past 30 years (on average 65 000 people per year for last 10 years), and is expected to increase by 2 million by 2041 (Skinner et al. 1998; The State of Queensland Department of Infrastructure 2016). Moreton Bay is an important marine environment, supporting many significant ecosystems (including extensive mangrove forests and seagrass meadows), important marine mammal and other vertebrate populations (e.g. Chilvers et al. 2005), and an active seafood, recreational and tourist economy. The catchment of Moreton Bay and associated freshwater and estuarine waterways is also highly valued by the local community, significantly contributing to the quality of life of residents (Johnston and Beatson 2015, 2016). The management of Moreton Bay represents a substantial challenge, not only because of its environmental values and large and rapidly increasing human population, but also because: (1) the Bay supports commercial fisheries and recreational fishing; (2) it is a major shipping lane to national and international destinations and there are many associated industrial developments; (3) there is a high catchment to estuary ratio (14:1), with substantial discharges from numerous sub-catchments, particularly during large flood events, which are common in the area; and (4) there can be long residence times for pollutants in some parts of the bay (~200 days) (Chilvers et al. 2005). In an effort to deal with the myriad environmental challenges associated with the management of Moreton Bay, such as nitrogen and phosphorus loading from sewage treatment discharge and stormwater runoff, sediment from growing urban centres and agricultural areas, and installation of upstream dams and

4 – The problematic, the ef fective and the ugly – some case studies

119

Fig. 4.5.  Moreton Bay, Queensland (Photo by Bert Knottenbeld, CC BY 2.0)

dredging downstream (see Dennison and Abal 1999), a regional partnership was established between local government, state government, industry and utilities. (This has been a long-term partnership, although the name has changed several times: 2000–04 Moreton Bay Waterways and Catchments Partnership, 2005–10 South East Queensland Healthy Waterways Partnership, 2011–16 Healthy Waterways, 2016–17 Healthy Waterways and Catchments, 2017–present Healthy Land and Water.) Under the auspices of the partnership, a monitoring program commenced in 1999 called the Ecosystem Health Monitoring Program (EHMP). The original program monitored the environmental condition of more than 250 estuarine and marine sites monthly, and some 130 freshwater sites were sampled biannually (Ecosystem Health Monitoring Program 2007). The total cost for this monitoring and reporting program is (and remains) approximately A$3 million per year. Consistent with our adaptive management paradigm outlined in Chapter 3, after 15 active years, the program underwent a significant review and redesign in 2014. The review revisited and updated the objectives of the program, and found that the focus needed to shift from ambient environmental sampling, to understanding the spatial extent and intensity of pressures on waterways. The program wanted to inform how and where management actions should be applied and provide a measure of their impact. The revised program also wanted to link waterway condition with the social and economic benefits to the wider community. As such, a revised suite of indicators (existing and new) was selected to address the revised program objectives. The number of monitoring sites visited and frequency of visitation was also re-evaluated and consequently reduced, while remaining comprehensive for the area covered (E Saeck, pers. comm.). Importantly, this

120

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

redesign effected cost savings and incorporated major advances in automated water quality monitoring and predictive modelling. Working with, and identifying the needs of, the partner network, including state and local governments, industry and utilities, was central to this redesign. In the redesigned program, the following are monitored and/or modelled on a yearly to three-yearly basis: estuarine and marine water quality, freshwater communities and processes, habitats (seagrass, riparian and wetlands), catchment pollutant loads, and social and economic community benefits. Estuarine samples are collected from 182 sites during eight months of each year, freshwater samples are collected from 75 sites once a year during autumn (48 ‘fixed’ sites are sampled every year and one of three sets of 27 ‘rotating’ sites is sampled once every three years, totalling 129 sites assessed every three years), and seagrass depth range is assessed at 17 sites, twice a year. (Citizen scientists also collect data on seagrass cover at more than 4000 sites throughout Moreton Bay.) In addition, monitoring of the fish community, catchment loads (six sampling stations, monthly during base flow and throughout major flow events) and riparian vegetation is done, as is environmental modelling (http://hlw.org.au/report-card/monitoring-program). A community benefits survey of South East Queensland residents is undertaken annually (commenced in 2015) to collect data on residents’ attitudes and behaviours towards using and valuing their local waterways. The successful ‘report card’ format (see ‘High visibility’) was retained in the redesigned program. The partnership for Moreton Bay is an excellent model that demonstrates ways of combining targeted monitoring programs with long-term ecological and hydrological research. The research and monitoring are based on management questions focused on improving management practices and environmental conditions, and incorporate the complexity of environmental and management issues to be tackled (O’Neill 2008). We suggest there are several features which have been pivotal to the success of this program. Questions with management relevance The research, and especially the monitoring, is driven by diverse questions with management relevance, such as how much (and where) the loading of nutrients and sediment need to be reduced to improve the water quality and general environmental condition of Moreton Bay, and how this could be measured. This relevance occurs, in part, because the program deals with pressing environmental problems that are highly pertinent to large numbers of people who, at the same time, are aggressive stakeholders in Moreton Bay and its catchment. The work has led to a demonstrable improvement in environmental conditions in Moreton Bay since commencing in 2000 (e.g. Ecosystem Health Monitoring Program 2007, 2008; Pitt et al. 2009; Saeck et al. 2013) and hence it has a proven track record of success and management relevance.

4 – The problematic, the ef fective and the ugly – some case studies

121

Partnerships and governance The work in the Moreton Bay area has been built on strong partnerships forged between key stakeholders. The project is truly cross-disciplinary and links some of the best hydrologists and limnologists in the world to engineers, resource managers, policy makers, politicians and the general public. Important parts of the program’s success are: ●●

●●

●●

independence – it is coordinated by a non-government organisation, independent of all levels of government and industry robust science – the whole program is under the review of an independent scientific advisory panel, made up of research leaders with extensive knowledge of the region ongoing engagement of partners through functional committees – the strategic direction is led by steering committees with representatives from all partners.

High visibility The Moreton Bay partnership is highly visible to politicians, the general public and resource managers. It has some strong and highly articulate advocates and produces a readily accessible and interpretable annual report card on changes in environmental conditions in Moreton Bay (Ecosystem Health Monitoring Program 2008; 2016 report card launch at http://hlw.org.au/report-card). This report card has proven to be a politically powerful management tool. Extensive briefings occur before the issue of the report card to stakeholders at all levels – from catchment officers to executives – of government (local and state), industry and utilities, including mayors and councillors. Credibility and design Initially, this program was designed by an independent group of scientists, engineers and managers before it was implemented. Similarly, the redesign was developed over several years before implementation, led by the stated needs of the members of the partnership and overseen by independent expert panels. As a result, the program was developed with a high degree of rigor and scientific integrity.

The Hubbard Brook Ecosystem Study The Hubbard Brook Ecosystem Study was initiated on 1 June 1963. It would become the long-term research and monitoring program of ecological and biogeochemical structure and function of the Hubbard Brook Experimental Forest in the White Mountains of New Hampshire in the United States (Holmes and Likens 2016). Using a medical metaphor, it was proposed that the chemistry of the stream water, draining a complex watershed ecosystem, could be used to diagnose

122

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

ecosystem function like the chemistry of blood and urine is used to diagnose human health. The following is taken from the first proposal to the National Science Foundation (Bormann and Likens 1962). Given the basic abiotic and biotic complexity of land, the phenomena of succession and retrogression, a multiplicity of managerial goals, and a desire for more efficient use of the land, it is obvious that some theoretical framework upon which we can assemble and interrelate these diverse components is a necessity. The ecosystem concept provides this framework … (and) may be visualised as a series of components such as species populations, organic debris, available nutrients, primary and secondary minerals, and atmospheric gases, linked together by food webs, nutrient flow, and energy flow. Knowledge of these components and the links involved, leads to an understanding of the interrelationships within the systems and of the ramifications of any manipulation applied at any point in the system. ... the ecosystem may be visualised as being connected to the surrounding biosphere by a system of inputs and outputs … Nutrient cycles are strongly geared, at many points, to the hydrologic cycle. As a consequence, measurement of nutrient input and output requires simultaneous measurement of hydrologic input and output. An investigation is proposed into the absolute rates of mineral generation, withdrawal and circulation within a small ecosystem. The experimental area selected is a small watershed in the Hubbard Brook Experimental Forest (HBEF) in West Thornton, New Hampshire. Some basic approaches and objectives identified in this first proposal included: (1) the need for a conceptual model to guide research (see Fig. 4.9); (2) the value of the ecosystem concept; (3) the need for watershed-scale experimental studies with a reference (‘control’); (4) the evaluation of the critical relation between ‘mineral cycles’ and hydrology; (5) the need for the perspective of long-term data; (6) the realisation that budgets (mass balances) can provide diagnostic tools for analysing complicated ecosystems; (7) the value of archiving samples for future use; and (8) the need to relate scientific research to land and water management issues. The original founders of the Hubbard Brook Ecosystem Study included F Herbert Bormann, plant ecologist; Gene E Likens, aquatic ecologist; Noye M Johnson, geologist; and Robert S Pierce, soil scientist. They, and their respective expertise, blended well in pursuing this complicated ecosystem study (Holmes and Likens 2016). Serendipity played a critical role in the success of the early long-term research and monitoring program at the Hubbard Brook Experimental Forest. Many peers

4 – The problematic, the ef fective and the ugly – some case studies

123

were sceptical that this study would be a viable way to understand large, complicated ecosystems, because no-one had done that kind of study on the scale of an entire watershed. To do the extensive chemical analyses (none of the founders were trained as chemists), the Hubbard Brook team had to develop and test their own procedures, and there was much to be done. Johnson argued that the focus should be on calcium, magnesium, sodium and potassium because, as a geologist trained in the mid-western United States, his geochemical view of the world depended upon carbon dioxide and carbonate. Likens and Bormann considered that phosphorus and nitrogen would be extremely important ecologically, but the team went along with Johnson’s approach as a way to start. Even so, how to do it? It was planned to use an EDTA titration of water samples, which is a laborious, wet-chemistry method for measuring base cations (calcium, magnesium, sodium, potassium) in solution. Bormann and Likens visited Clarence Grant, an excellent soil chemist who had just joined the faculty of the University of New Hampshire in nearby Durham, for advice. They spent about an hour telling him what they planned to do, including all the chemistry. Grant was highly sceptical, if not discouraging, of the possibility of success! Just as Bormann and Likens were leaving Grant’s office, Grant said, ‘I remember something that came in the mail the other day, a flyer about a new instrument that might be helpful to you.’ He looked through a pile of papers and pulled out the flyer. ‘Here it is, a new instrument called the ... atomic absorption spectrophotometer.’ This instrument had been developed by a scientist in the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia largely to avoid interferences in the measurement of base cations, and was primarily for use in analysing the chemistry of blood and urine. Returning home to Dartmouth College, Likens called the Perkin-Elmer Corporation in Norwalk, Connecticut, and promptly bought one of these new instruments (Model 303) on the grant from the National Science Foundation. It was stamped ‘No. 15’ on the back, and was the first one purchased or used outside a medical facility in the United States. As it turned out, this instrument made the Hubbard Brook Ecosystem Study possible. It is probably fair to state that there would not have been a long-term Hubbard Brook Ecosystem Study without that tool at just that time. This was a serendipitous event: a phone call, a meeting that caught the fancy of someone with additional knowledge, all making the research project so much more enriched and productive. And this early development of the Hubbard Brook Ecosystem Study was not dependent on buying people’s time or services or some tool per se; instead it developed through a question, cooperation, enthusiasm and, especially, trust among the interested investigators. The Hubbard Brook Ecosystem Study has been in continuous operation for 54 years (Holmes and Likens 2016). The US Forest Service had started hydrometeorological measurements some eight years earlier. A large number of entities

124

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 4.6.  Precipitation (rain and snow) measurement at Hubbard Brook Experimental Forest

Fig. 4.7.  Gauging weir at Hubbard Brook Experimental Forest (Photo by A.S. Bailey, Northern Research Station, U.S. Forest Service, reproduced with permission from Holmes and Likens 2016)

4 – The problematic, the ef fective and the ugly – some case studies

125

Fig. 4.8.  Mirror Lake, Hubbard Brook Experimental Forest (Photo by D.C. Buso)

are now monitored on a continuing basis in the Hubbard Brook Valley including solar radiation; air, soil, stream and lake temperature; precipitation amount and chemical content (Fig. 4.6); air chemistry; stream water, including flow and chemical content (Fig. 4.7); soil-water chemistry; forest population dynamics and biomass accumulation; litterfall; bird, salamander and snail populations; and a variety of limnological parameters in stream ecosystems and Mirror Lake (Fig. 4.8), including ice in and ice out dates (Likens 2011; Holmes and Likens 2016). Experimental watershed-ecosystem manipulations in response to questions at the ecosystem scale have been a hallmark of the long-term research of the Hubbard Brook Ecosystem Study and several are ongoing including those addressing questions about forest harvest (clearcutting) and nutrient depletion from the effects of acid rain. Paired watershed ecosystems provide a reference (Likens 1985) for evaluating the long-term effects of the experimental manipulations (Holmes and Likens 2016). Well-defined and evolving questions The Hubbard Brook Ecosystem Study has been driven by carefully formulated and scientifically tractable questions. These have continued to evolve as new insights trigger new lines of inquiry. Indeed, over the years the steady stream of proposals submitted to funding agencies such as the National Science Foundation needed to be driven by compelling questions to obtain ongoing funding in a very competitive, peer-review environment.

126

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 4.9.  A conceptual model for biogeochemical relationships and input and output hydrologic and chemical fluxes in a terrestrial ecosystem (redrawn from Bormann and Likens 1967)

Conceptual model The early development and continued use of a conceptual biogeochemical model (Fig. 3.2, Fig. 4.10; Bormann and Likens 1967) to guide ecosystem and biogeochemical research and analysis, helped to formulate new questions and identify gaps in research directions. With time, this conceptual model became more complicated but still retained its usefulness in guiding thinking and research for the Hubbard Brook Ecosystem Study. This long-term usefulness was related in large part because of the biogeochemical focus of the Hubbard Brook Ecosystem Study. In 2004, a new model was developed and published (Groffman et al. 2004b). This model is more complicated than the one in Fig. 4.9 in terms of entities and concepts included, but to-date its usefulness for guiding the thinking and activity of the Hubbard Brook Ecosystem Study has been limited. Focus on scientific productivity and synthesis materials The research at the Hubbard Brook Experimental Forest has been highly productive (see Holmes and Likens 2016; www.hubbardbrook.org). There are many important facets of this work and these are often intimately interlinked in the ecosystem approach used. Cross-fertilisation often leads to greater insights than would be apparent from consideration of the individual parts in isolation. Given this, the integration of results has led to many synthesis books (Likens et al. 1977; Bormann and Likens 1979; Likens 1985, 1992; Likens and Bormann 1995; Winter

4 – The problematic, the ef fective and the ugly – some case studies

127

Fig. 4.10.  Changes in forest biomass over time at Hubbard Brook Experimental Forest, New Hampshire (modified from data of RH Whittaker, TG Siccama, JJ Battles and others – see Battles et al. 2014)

and Likens 2009; Holmes and Likens 2016) and monographs or major synthesis efforts (e.g. Bormann et al. 1970, 1974; Likens et al. 1970, 1994, 1998, 2002; Whittaker et al. 1974; Fahey et al. 2005; Lovett et al. 2005; Groffman et al. 2006). Demonstrated importance of long-term research The Hubbard Brook Ecosystem Study has continued to highlight and nurture the concept of long-term studies (e.g. Likens et al. 1977; Bormann and Likens 1979; Likens 1992, 2004; Likens and Bormann 1995; Holmes 2007; Holmes and Likens 2016) even though it was often very difficult to maintain uninterrupted funding. For example, although the forest in south-facing catchments of the Hubbard Brook Experimental Forest appears green and robust, research conducted since 1965 is now indicating, surprisingly, that this forest is no longer accumulating biomass. It is now losing biomass as a result of increased mortality of dominant tree species, primarily sugar maple (Acer saccharum), because of loss of soil nutrients and increased toxic aluminium in soil water caused by acid rain, and introduced pests and pathogens (Fig. 4.10; Likens et al. 1998; Siccama et al. 2007; Likens and Franklin 2009; Holmes and Likens 2016). From a carbon sequestration perspective, it might be argued that this long-term record became interesting only after ~20 years of monitoring.

128

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Collaboration Long-term research and monitoring at the Hubbard Brook Experimental Forest has been characterised by strong research partnerships among scientists of the US Forest Service, the US Geological Survey, and numerous universities and research institutes to work cooperatively as part of a scientific team. Dozens of scientists, including a large number of senior scientists, students and support staff from many institutions and agencies have worked at the Hubbard Brook Experimental Forest as part of the Hubbard Brook Ecosystem Study over the past 50+ years. Field presence Initially, a small, focused and dedicated team of senior researchers spent much time in residence, interacting together at the Hubbard Brook Experimental Forest (see Chapter 5). Moreover, long-term, competent, dedicated and enthusiastic support personnel have been central to the success of the Hubbard Brook Ecosystem Study (see also Chapter 2). Careful use and calibration of field equipment A feature of the biogeochemical monitoring at the Hubbard Brook Experimental Forest has been the use of analytical procedures that were neither changed nor replaced without first overlapping and comparing results from the long-term method with those from a proposed new method. This approach avoided artefacts in the long term chemical record (Buso et al. 2000). Deficiencies In spite of the long-term successes of the Hubbard Brook Ecosystem Study, there are several deficiencies and problems. This long-term study is place based, a ‘one of a kind’ study, in that it is not, and cannot be, replicated, which is important in scientific research. Moreover, watershed ecosystem experimental manipulations can be subject to pseudo-replication (Hurlbert 1984), but this problem is alleviated by the paired-watershed approach and the long-term nature of the study. Nevertheless, sometimes results are so obvious that statistics are not necessary and may be a diversion. A related issue is the decreasing availability of undisturbed watershed ecosystems for new experimental manipulations as more watershed ecosystems are used for long-term experimental manipulation (Likens and Buso 2010). Another issue is the increasing bureaucracy associated with running the research and monitoring program with increasing numbers of scientists from increasing numbers of institutions. Such bureaucracy potentially adds to complicated, interpersonal dynamics of interacting scientists, the need for more structured communication venues, and fragmentation in the focus of the study. Over the years, sustaining the funding necessary to continue the long-term

4 – The problematic, the ef fective and the ugly – some case studies

129

monitoring has been very difficult, and currently is especially so given the tight funding environment for ecological research in the United States. Other problems include: extrapolation of detailed site results to larger areas and different ecoregions, and issues of leadership succession; three of the four founding leaders are deceased (Johnson in 1987; Pierce in 1993; and Bormann in 2012). Likens stepped down as co-director of the Hubbard Brook Ecosystem Study in 1990, although he continues to be active in this study. Charles Driscoll and Tim Fahey were the principal investigators of the NSF-funded, Long-Term Ecological Research (LTER) program from 1988 through to 2015 at the Hubbard Brook Experimental Forest. Gary Lovett and Peter Groffman have now assumed this role. The LTER program is a major funder of the overall Hubbard Brook Ecosystem Study. In response to growth and complexity of this study, a revised governance structure was initiated in 2004 (Groffman et al. 2004a). Pamela Templer currently chairs the Scientific Co-ordinating Committee in this governance structure.

The Central Highlands of Victoria, south-eastern Australia The wet montane ash eucalypt forests of the Central Highlands of Victoria are the focus of this case study (Fig. 4.11). The study area comprises about one degree of latitude and longitude and lies ~100 km north-east of Melbourne, the second largest city in Australia. These are spectacular forests and support the tallest

Fig. 4.11.  A stand of old growth mountain ash. The person in the mid-ground highlights both the size and height of the trees in this stand. (Photo by Esther Beaton)

130

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

flowering plants in the world, with trees in old growth stands sometimes exceeding 100 m in height (Beale 2007). They are also critical forests for: (1) timber and pulpwood production; (2) biodiversity conservation – the distribution of the endangered Leadbeater’s possum (Gymnobelideus leadbeateri) is virtually confined to stands of ash-type eucalypts (Lindenmayer et al. 2015c; Todd et al. 2016); (3) the production of water to supply Melbourne (Viggers et al. 2013); (4) human recreation (Keith et al. 2016); and (5) carbon storage (Keith et al. 2009, 2014b). The work in the Central Highlands was initially established to quantify the factors influencing the distribution and abundance of arboreal marsupials throughout the ash-type forests (Smith and Lindenmayer 1988). A particular focus was on the critically endangered Leadbeater’s possum because: (1) it was important to know where the species was most likely to occur to provide a spatial context for managing logging operations; (2) there were concerns about the potential impacts of logging operations on the species; (3) arboreal marsupials are cavity-dependent animals and typically occupy trees that are 150 years or older. Logging operations alter the structure of forests and, in particular, reduce the abundance of cavity trees and significantly delay the time until new ones are recruited (Lindenmayer et al. 2012a, 2016); and (4) there were community demands to develop approaches to mitigate logging effects on biodiversity (including Leadbeater’s possum) and introduce new strategies to integrate wood production and conservation better in ash-type forests (Lindenmayer et al. 2015c). The work commenced in 1983 and continues today. It has been led by a single investigator (Lindenmayer) since that time. Since then, the research program has grown to span several broad themes: (1) from single species work to species assemblages research; (2) spatial patterns (e.g. forest cover and composition) to work on ecological processes (e.g. fire and tree fall); (3) empirical work to simulation modelling; (4) observational studies to manipulative experiments; (5) ecological theory to applied ecological research (Lindenmayer et al. 2015c); and (6) the application of economic and environmental accounting to quantify trade-offs in the use of different natural resources and their relative contribution to economic activity (Keith et al. 2016, 2017). The study which formed the backbone of the work was a carefully designed and executed set of repeated animal and vegetation surveys at over 160 sites distributed widely in the study area. The selection of the 160+ sites was guided by strong contrasts in a range of key environmental conditions. These included: (1) logging and disturbance history – places that have recently been logged and regenerated versus others with no history of cutting; (2) topographic location and elevation – sites on steep slopes versus those on flat terrain; and (3) proximity to human disturbance – sites close to human habitation and those remote from it. There was spatial dispersion of sites to avoid geographical bias and extensive replication of sites within cells representing combinations of stratifying factors to

4 – The problematic, the ef fective and the ugly – some case studies

131

quantify variability and ensure that results of subsequent statistical modelling of field data were robust. That initial study continues to provide the framework for a long-term monitoring study of animal population dynamics, the dynamics of vegetation structure and composition and the inter-linkages between the two (Lindenmayer et al. 2003, 2013b, 2016). Key questions that have motivated the monitoring program have been: ●● ●●

●●

●●

●●

What are temporal patterns of animal abundance? What are the temporal patterns of abundance of large trees with hollows? These trees are primary nesting and denning sites for cavity-dependent animals, including Leadbeater’s possum. What are the relationships between the abundance of trees with hollows and the occurrence of different species of arboreal marsupials? Do these relationships change as the number of trees with hollows declines over time? Are actual changes in populations of animals congruent with those forecast to occur as a consequence of declining populations of trees with hollows?

An early (and obvious) insight was that Leadbeater’s possum did not exist in isolation from other animals, including other species of arboreal marsupials. Information on other species of arboreal marsupials was therefore needed to help better identify where Leadbeater’s possum did and did not occur and the underlying reasons for those patterns of occurrence and co-occurrence (Lindenmayer et al. 2014a). Hence, studies of a single species expanded to encompass other species and ultimately the entire assemblage of possums and gliders, as well as many taxa from other groups such as birds (Lindenmayer et al. 2014b) and terrestrial mammals (Banks et al. 2011). The composition and structure of the forests strongly influence where these animals live (Lindenmayer et al. 2015c, 2017) and the research program expanded to include work on this topic. Habitat suitability is strongly shaped by the ecological process of disturbance and this made it essential to quantify the impacts of fires and logging on the stand structure and landscape composition (Lindenmayer et al. 2013b, 2016). Learning how to mitigate the impacts of human disturbance on forest biodiversity was therefore informed by: (1) quantifying where species occurred; (2) why they occurred where they did; and (3) how disturbance dynamics influenced the key resources required by the target species of management concern. In particular, how logging changed key attributes of stand structure and landscape composition and, in turn, how logging practices might be altered to reduce such impacts (Lindenmayer et al. 2013a). Research on arboreal marsupials at the long-term monitoring sites demonstrated highly significant relationships between the presence and abundance of arboreal marsupials and the abundance of trees with hollows on those sites

132

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Dataset 2 Dataset 3 Dataset 4

Number of large old trees per hectare

Dataset 1

Greater Glider

Mountain Brushtail Possum

7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0

11

1

2

4

8

16

32

1

2

4

8

16

32

Abundance of hollow-bearing trees

Fig. 4.12.  Relationships between the probability of arboreal marsupials and the abundance of trees with hollows in different datasets that have been gathered over the past three decades (thereby demonstrating consistency in results) (redrawn from Lindenmayer et al. 2014a)

(Lindenmayer et al. 2014a). Other studies identified a series of interrelated ecological mechanisms influencing these relationships: (1) different species partition the hollow-tree resource and select trees with different characteristics (Lindenmayer et al. 1991c); (2) different species rarely co-occur in the same tree (Lindenmayer et al. 1990); and (3) each species of arboreal marsupial swaps regularly between den trees and typically uses between six and 20 different trees in a given year (Lindenmayer et al. 1996). These factors mean that only those sites with many trees with hollows are likely to fulfil the requirements of most species of arboreal marsupials. This pattern, in turn, explains the highly significant relationships identified for animal presence and abundance and the abundance of trees with hollows in montane ash forests (Fig. 4.12; Lindenmayer et al. 2014a).

4 – The problematic, the ef fective and the ugly – some case studies

(a)

133

(b)

Fig. 4.13  (a) Marked standing tree and (b) collapsed tree (Photos by Dave Blair)

A 30+-year study of over 1100 marked large trees with hollows (Fig. 4.13) revealed that the death and collapse of these trees is far outstripping the rate of recruitment of new ones (Lindenmayer et al. 2012a, 2016). Predictions from these data indicate that there will be a chronic shortage of these trees throughout large parts of the montane ash forests for 50 or more years beyond 2050 (Fig. 4.14) (Burns et al. 2015). Given known relationships between the abundance of trees with hollows and the abundance of arboreal marsupials, an array of species of cavity-dependent arboreal marsupials are forecast to undergo substantial declines in the coming 40 years. These projections have, in turn, been appropriate for making comparisons with actual data on animal population dynamics gathered from the long-term monitoring sites between 1990 and 2016. That is, projected changes have been compared with actual changes (Lindenmayer et al. 2011). The results of those comparisons have been instructive, although also somewhat sobering. These data indicate that arboreal marsupials have actually declined in all age cohorts of forest (Fig. 4.15; Keith et al. 2016, 2017). The work in the Central Highlands of Victoria has led to some valuable insights. It was some of the first research in Australia to rigorously quantify the impacts of conventional clearcut logging methods on forest vegetation structure, plant species composition (Lindenmayer et al. 2015c; Blair et al. 2016) and the response of animals such as Leadbeater’s possum, as well as other species of arboreal marsupials (Lindenmayer et al. 2015c). For many years, forest management in Victoria was based on a simple assumption that wet eucalypt forests are almost exclusively even-aged as a result of stand-replacing wildfires. However, the work in this study showed that almost all old growth stands actually comprised dominant trees of multiple ages, changing the comprehension of how these forests function

134

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Number of large old trees per hectare

8

4

No future fire or logging

2

10% fire in unburnt forest 20% fire in unburnt forest

1 Current TRP x3 TRP x3 + 20% fire

0.5 2000

2010

2020

2030

2040

2050

2060

Year

Fig. 4.14.  Projections of the future abundance of trees with hollows, where TRP corresponds to the five-year timber release plan for 1–3 five-year cutting periods (redrawn from Burns et al. 2015)

(Lindenmayer et al. 2000). Using empirical analysis based on extensive field vegetation data, it was then demonstrated that random and deterministic influences of climate, terrain and other factors significantly influenced the development of multi-aged stands in particular parts of forest landscapes (Mackey et al. 2002). Such multi-aged forests are important because they support the highest diversity of

Number of animals per 1 ha site

Age Class Old Growth

6

Fire 1939

5

Mixed Logged

4

Fire 2009

3 2 1 0 0

20

40

60

80

100

120

Forest age (years)

Fig. 4.15.  Decline in the abundance of cavity-dependent arboreal marsupials (redrawn from Keith et al. 2016)

4 – The problematic, the ef fective and the ugly – some case studies

135

arboreal marsupials (Lindenmayer et al. 1991b). Discoveries on old growth, multiaged forest and patterns of natural disturbance resulted in the development of new perspectives on the impacts of post-disturbance salvage logging (Lindenmayer and Ough 2006; Blair et al. 2016). It was shown that a combination of natural disturbance (fire) and human disturbance (post-fire salvage logging) could lead to cumulative effects more significant and more prolonged in their impacts than either kind of disturbance in isolation (Lindenmayer and Ough 2006; Blair et al. 2016). These new perspectives have led to new conservation management strategies within forests broadly designated for wood production, including greater regulation of where logging can and cannot take place, and limits on where post-fire salvage logging is permitted (Lindenmayer et al. 2015c). It also led to the instigation of a replicated logging experiment with greater levels of tree retention on cut sites (Lindenmayer et al. 2015a). A key additional area of work in the Central Highlands has been a major focus on the effects of wildfire and post-fire recovery. On 7 February 2009, major wildfires began burning throughout large parts of the Central Highlands of Victoria, as well as other parts of that state. They eventually became the worst fires in Australia’s history with 173 lives lost and over 3000 homes destroyed. The fires had an enormous impact on the long-term studies in the ash forests of the region and many sites were damaged as a result of the conflagration (Fig. 4.16). However, it has been important to treat the fire as a new opportunity to learn about post-fire ecological recovery. In mid-April 2009, as soon as access to the forest was permitted, work commenced remeasuring the vegetation and the vertebrate fauna at all 161 long-term monitoring sites (i.e. both burned and unburned). The fires also have meant that the work in the Victorian forests has become an example of adaptive monitoring (sensu Chapter 3). Adaptive monitoring was needed because the wildfire demanded that new questions be answered and some changes in the protocols for monitoring such as the way in which sites were selected for surveying in any given year. Some examples of the new research questions include: ●●

●●

●●

What are the relationships between pre-fire plant and animal populations and post-fire population trajectories? How does burn severity influence these trajectories, for example, in terms of species richness, community composition, and populations of individual species? What are these relationships for nationally threatened species such as Leadbeater’s possum? How are recovery trajectories influenced by initial conditions (stand age at the time of the 2009 fire), the severity of the 2009 fire, and the interaction of initial conditions and burn severity? Does burn severity and past fire history differentially influence certain kinds of species (e.g. functional groups of plants or particular guilds of arboreal marsupials and birds)? Are there particular sets of life history attributes common to species which increase v. decrease over time?

136

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Fig. 4.16.  A stand of burned old growth forest following the February 2009 wildfires. The site is the same forest as shown before the fires in Fig. 4.11. (Photo by Dave Blair) ●●

●●

Is there a well-defined successional replacement of species within particular groups over time, as predicted by the habitat accommodation model (Fox 1982)? Are there landscape context effects? That is, is the recovery process at a site influenced by the extent of unburned and burned forest surrounding a site?

There have been many important discoveries as a result of the post-fire research in the montane ash forests, including a range of unexpected ones. For example, many species of birds and almost all species of arboreal marsupials are all highly sensitive to the effects of fire, including fire at both the site and surrounding landscape scales. As just three (of many) examples: (1) Species such as Leadbeater’s possum are almost totally absent from burned sites and also largely absent from intact sites where there has been fire in the surrounding landscape (Lindenmayer et al. 2013b); (2) Post-fire recovery by populations of small mammals appears to be driven by residual animals that remain on burned sites (rather than from colonisation by animals from intact forest beyond the fire boundary (Banks et al. 2011); and (3) Fire, even very high severity fire, results in far less depletion of carbon stocks than predicted, with severely burned old forests supporting up to 90% of pre-fire carbon stocks (Keith et al. 2014a).

4 – The problematic, the ef fective and the ugly – some case studies

137

Work in the montane ash forests is ongoing and new work is compiling an array of time series datasets (Lindenmayer et al. 2015c) such as those being used to examine longer-term responses of species to disturbance gradients spanning fire, logging and the combination of the two (e.g. Blair et al. 2016). The long-term studies and monitoring program in the Central Highlands of Victoria are characterised by some important features. Stability The same person (Lindenmayer) has been the leader of the project over the past 34 years and has maintained the work even though there have been five major restructures of government agencies responsible for forest and conservation management in that time. Expert statistical design Four highly qualified professional statisticians have played pivotal roles in the research. They have been responsible for assisting in the framing of tractable scientific questions, the design of observational studies and experiments (see Cunningham and Lindenmayer 2017), conducting cutting-edge statistical analyses of complex datasets, the development of new statistical methods, and contributing to the ecological auditing of projects to determine whether field data collection should continue or if detailed analyses of data were appropriate. Tractable approach to field research and monitoring Although many studies have been completed over the 34-year history of work, each investigation has been characterised by careful posing of questions and the collection of a tractable set of covariates for subsequent analyses. There has never been a desire to ‘monitor everything’ by the laundry list method, nor has there been an assumption that Leadbeater’s possum (nor any other species or group) was an indicator species for some other entity or species or group (Lindenmayer and Cunningham 1997; Lindenmayer et al. 2014a). Ongoing development of questions and research themes The scope and breadth of work undertaken has grown over the period of the project. This directed growth has occurred because new insights and discoveries have prompted new questions and triggered new studies, such as those of ecological recovery following the 2009 wildfire. These studies have been built around a framework of monitoring sites without compromising the integrity of that framework or the integrity of the long-term datasets being gathered. This approach has added greatly to the body of knowledge about the study region and created valuable cross-fertilisation between research themes and topics, thereby creating a better understanding of the forest patterns and ecological processes and

138

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

the mechanisms that link them in ash-type eucalypt forests (Lindenmayer 2009; Lindenmayer et al. 2015c). High levels of scientific productivity The research in this study has led to the publication of more than 220 scientific articles as well as eight books. This has helped establish the scientific credentials of the work and emphasised its value to policy makers, resource managers and politicians. Partnerships Much of the work in this study has involved establishing working relationships between researchers and staff from government agencies responsible for forest management and wildlife conservation. This development of relationships has been done deliberately to ensure that the results of the work can be adopted in on-theground management practices. It also has been critical to the establishment of logging experiments designed to identify more environmentally sensitive silvicultural systems than traditional clearcutting. Large numbers of volunteers have been involved in some of the field-counting protocols and partnerships with community conservation groups have greatly assisted field data collection (Lindenmayer et al. 1991a). Quality of on-the-ground staff Part of the success of this project is due to very high-quality field staff, working full-time in the forest to gather field data. These field staff also have been adept at data curation and database management as well as providing high-quality input to many scientific articles (on which they are either co-authors or lead authors). Deficiencies The work in the Central Highlands of Victoria has several weaknesses and problems. First, much of the initial work was based on quantifying patterns and far less emphasis was placed on the ecological processes or mechanisms giving rise to those patterns. Additional effort has been focused on rectifying this deficiency in the past decade, but a lot more work is required to address this problem. Second, the use of a tractable and explicit conceptual model that is at the core of the adaptive monitoring framework (see Chapter 3) and fundamental to the success of the Hubbard Brook Ecosystem Study (discussed earlier) was adopted in the past eight years within the Central Highlands of Victoria. Third, and typical of virtually all long-term monitoring, is funding. For many years, the project was run using book royalties and other small and highly insecure sources of money. The current work program has three additional years of funding (to the end of 2020) and the future of project following that date remains uncertain. A fourth

4 – The problematic, the ef fective and the ugly – some case studies

139

deficiency has been that the initial study, which now forms the basis for the 34-year monitoring program of populations of arboreal marsupials, was an observational, retrospective investigation (Cunningham and Lindenmayer 2017). Although it was carefully designed and implemented, these kinds of studies are less powerful for making inferences than traditional experiments and natural experiments. More recent work has been based around the implementation of true experiments (Lindenmayer et al. 2015a) and innovative statistical approaches are being developed to link past observational work with more recently established experiments. Fifth, the leader of the long-term study has fallen victim to the modern problem of fragmentation of effort, ideas and time, and is responsible for five other large-scale projects in south-eastern Australia. Hence, the focus on the long-term study in the Central Highlands of Victoria is not as acute as it should be. Sixth, there have been periods over the past 34 years where communication between researchers and resource managers and policy makers has not been nearly as frequent as it should have been and this has led to such problems as logging of monitoring sites and other kinds of damage to experimental sites. Finally, although the Central Highlands of Victoria was formerly part of an informal alliance of (now sadly defunct) LTER sites in Australia, neither the data nor insights from the work has been drawn into the national State of the Environment Report series (State of the Environment Advisory Council 1996; Commonwealth of Australia 2011). This integration between environmental reporting and monitoring programs is an unresolved and vexed problem that we discuss further in Chapter 5.

Need to wait and see NEON/TERN This case study outlines two very large programs designed to help establish infrastructure for long-term ecological research and monitoring. The research programs are the US National Ecological Observatory Network (NEON) and the Terrestrial Ecosystem Research Network (TERN) in Australia. They were initiated by the US National Science Foundation (NSF) and the Australian Government’s (then) Department of Innovation, Industry, Science and Research (now Department of Education and Training), respectively. Both programs are funded from portfolios that have traditionally been used to support the construction of expensive facilities for the physical sciences, such as large telescopes for astronomy. NEON and TERN are, in effect, leaders in a large-scale (almost global) push towards Earth Observation Networks in which the primary focus is on data gathering and the subsequent safe, reliable storage and eventual sharing of those data. The focus is clearly on data collection and sharing and not research; although NEON has a US$400 million budget over 30 years (raised to US$469 million), it

140

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

does not have funds to support research to analyse data that are gathered (although workshops are funded to educate users on how to obtain data). TERN is similar, with its specific mandate from the National Collaborative Research Infrastructure Strategy (NCRIS) being to build and maintain infrastructure to enable research but not to fund any research. There are many other initiatives globally following the broad template set by NEON and TERN that are underpinned by the ‘EON observatory model’. The South African Environmental Observation Network (SAEON; e.g. SAEON 2016) and the Global Lake Ecological Observatory Network (GLEON) in the aquatic realm (e.g. McDowell 2015), are examples. The Integrated Marine Observing System (IMOS) in Australia is equivalent to TERN but in the marine domain (and also far better funded; see Box 4.6). NEON and TERN have similar recent histories even though they are in different countries and have vastly different budgets (TERN has a far smaller budget than NEON, of approximately A$80 million over the past 10 years). They have been characterised by protracted discussions about what infrastructure should be funded, extensive and repeated consultations among scientists about an appropriate way for the programs to proceed, considerable complication of the funding process (including funding cuts, some of which have been severe), multiple changes in leadership, and various issues with governance (see Mervis (2015) and Cesare (2016) in the case of NEON). At the time of writing, the future of TERN funding is unclear, and a major review of its programs has resulted in the closure of the Long-Term Ecological Research Network (LTERN) within TERN at the end of 2017 (Lindenmayer et al. 2017). TERN comprises several facilities that have different key roles. These include (among several others): AusCover, which documents changes in land cover on the Australian continent based on remotely-sensed information; OzFlux, a set of flux towers located in different parts of Australia which are used to monitor atmospheric conditions such as carbon fluxes; AusPlots, a large-scale surveillance monitoring program comprising a large number of intensively but once-off sampled plots, primarily located in rangelands with a much smaller set in forests; SuperSites, which are small-scaled and highly instrumented sites distributed across a range of environments that are co-located with the flux towers in OzFlux; and LTERN, a set of networks of long-term monitoring plots in different ecosystems across Australia (TERN 2013). One of us (DBL) had a close association with the TERN program as the Science Director of LTERN, making a critical evaluation of this case study more challenging to write. The procurement of extensive infrastructure to develop observation networks designed to support environmental research and monitoring is useful and some important discoveries have already been made (Luo et al. 2011; Groffman et al. 2014; Wilson et al. 2015). However, whether this approach leads to ecologicallyeffective and cost-effective ecological monitoring remains unclear – the reason

4 – The problematic, the ef fective and the ugly – some case studies

141

why we have this case study located under the heading ‘wait and see’. As outlined in the following sections, there are some key additional reasons why this case study has been assigned to the ‘wait and see’ category. Lack of tractable overarching questions, conceptual model or a priori predictions Neither NEON nor TERN were driven at the outset by testable scientifically-based questions. There was no conceptual model associated with any of the funding, nor any a priori predictions about how the ecosystems where the proposed infrastructure would be built might respond to environmental change or management interventions, other than in the most general terms. For example, the three overarching questions for NEON were: (1) How are ecosystems across the United States affected over time by changes in climate, land use and invasive species? (2) How do biogeochemistry, biodiversity, hydroecology and biotic structure and function interact with changes in climate, land use and invasive species across the nation? (3) How do these feedbacks vary with ecological context and scale over time? The problem of developing tractable questions triggered considerable deep thinking but unfortunately no resolution on how to address it. Such an outcome at the beginning of large ecological and environmental programs is not new. For example, it characterised the US$57 million US participation in the International Biological Program from the late 1960s to the mid-1970s (Boffey 1976). The words ‘It all started out very badly’ are prominent in an appraisal of the International Biological Program (Boffey 1976; Mitchell et al. 1976). In response to concerns and various controversies within the scientific community, especially regarding the source of the large amount of operational funding required by NEON, a committee was formed by the US National Academy of Sciences/National Research Council (NAS/NRC) to study the problems. This committee (see Tilman 2003) endorsed the concept of NEON, but called for a ‘refined focus and more detailed plan for implementation’, and identified six ‘Grand Challenges’ in environmental biology to guide the overall effort. These related to: biodiversity; biogeochemical cycles; climate change; ecology and evolution of infectious diseases; invasive species; and land and habitat use. Like NEON, TERN was initially underpinned by high-level questions that are either very difficult to test or the investment required to address them adequately falls far short of what is required. These questions included: (1) How does natural and human-induced change, both short- and long-term interact to alter ecosystem structure and function including remedial and mitigation measures? (2) How does biotic structure vary and act as both a cause and consequence of ecological fluxes and energy? (3) What is the outcome for the net terrestrial CO2 sink of the competing effects of CO2 fertilisation, declining water availability, global warming, changing nutrient dynamics and disturbance by impacts? (4) How do altered

142

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

community structure and ecosystem dynamics affect ecosystem services? (5) How can decision support systems, based on models of complex processes, lead to mitigation and/or remedial measures based on human behaviour? (6) Which human actions influence the frequency and magnitude of change, or form of disturbance regimes across ecosystems, and conversely, which human actions are required to mitigate or remediate degradation? Since that time, the scientists collaborating within the various facilities within TERN have worked very hard to rework their questions to articulate the key problems they want to address. To some extent, issues of tractable questions have been partially resolved within the different programs that comprise TERN. However, questions relating to the overarching network of facilities across TERN per se do not appear to be well resolved, and possibly may not be resolvable without a major overhaul in the design of the network. In essence, it appears that many of the ‘real’ scientific questions that will be answered from the NEON and TERN initiatives will be post-hoc questions posed after (and during) dataset collection. Retrofitting questions is common in ecological science and can undoubtedly produce some interesting and important outcomes. However, it is often not the most cost-effective and ecologically-effective approach for doing science; in many respects, we argue that this is a backwards way of doing science (Lindenmayer and Likens 2013). Post-hoc observational studies are generally the weakest of the various kinds of studies, relative to others such as quasi-experiments, natural experiments and true experiments (Manly 1992; Cunningham and Lindenmayer 2017; Fig. 4.17). This problem occurs because observational studies are mensurative and therefore the inferential status of the results is weak in that they cannot answer causal questions.

Fig. 4.17.  Key differences in the broad types of studies in landscape ecology; viz: true experiments, quasi experiments, natural experiments and observational studies; strength of inference increases from bottom to top (i.e. from observational studies to true experiments). The primary difference between strong versus weak quasi-experiments is whether interventions are exogenous or endogenous. Endogenous variables are those internal to an ecosystem such as natural regeneration of trees and other vegetation) rather than being assigned to them (i.e. exogenous or originating outside the ecosystem such as a pre-determined intervention involving planting of vegetation). (Redrawn from Remler and Van Ryzin 2014)

4 – The problematic, the ef fective and the ugly – some case studies

143

This deficiency relates to the difficulty of separating the effects attributed to different variables and, therefore, the association of a variable with a particular outcome. The design of observational studies typically does not include welldefined ecological contrasts and there are few constraints on site selection, although stratification of sites can be employed. No method of analysis can overcome difficulties associated with confounding effects of unknown or unmeasured variables (Cunningham and Lindenmayer 2017). Narrow focus on infrastructure The primary focus of NEON and, to a lesser extent some parts of TERN, is on the construction of infrastructure or equipment for which the aim is to provide data streams (often very large ones), which are then to be archived for potential later use by scientists other than those who collected or generated the data. This approach is, in turn, based on the concept of ‘big data’ in which key discoveries can be made through analyses of very large datasets (Soranno and Schimel 2014), although there is not yet extensive evidence available to support the concept (but see Luo et al. 2011). The NEON and TERN programs have arisen, in part, because some advocates believe that the large equipment grants typical in the physical sciences should be extended into the ecological and environmental sciences. There is no doubt that many kinds of observations will be done best (most accurately with high frequency and repeatability) by a sensor (e.g. water levels in a stream or cloud cover). However, while equipment is undoubtedly important in facilitating long-term research and monitoring, the importance of trained, dedicated people to maintain and groundtruth equipment cannot be overlooked (see Box 4.4). A fixation on equipment has been sustained in assessments of TERN with the NCRIS roadmap (https://www.

Box 4.4. Mirror Lake ice cover observations Beginning in the mid-1960s, the occurrence and loss of a full ice cover on Mirror Lake in the White Mountains of New Hampshire, USA has been determined annually by a field observer. This approach allowed for a relatively precise time frame for verifying each of these two, major events for the lake (ice in = +/– 2 days and ice out = +/– 1 day) (Likens 2011), which is very good and somewhat unusual quality control for records of this type (Magnuson 2000). Since 1968, the length of ice cover on the lake has decreased at about –0.475 days per year. In an effort to ‘modernise’ this long-term monitoring, a remote, continuously transmitting camera was positioned on a hill overlooking the lake. Unfortunately, on more than one occasion, what appeared to be complete ice cover in the transmission of the image was in fact light reflection from the surface that had the same appearance. A field observer was required to verify the misleading, remotely-done observation.

144

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

education.gov.au/2016-national-research-infrastructure-roadmap) having a distinctly anti-person flavour, advocating the widespread use of technology and various kinds of instrumentation to tackle environmental problems. We argue that it is vital to recognise the critical importance of on-the-ground measurements by people, especially for the measurement of many elements of the biota. We have no doubt that exciting new insights can be gained from large and sophisticated equipment such as flux towers or remote sensing science (e.g. Lidar; see Thomas et al. 2013; Nolke et al. 2015). Such equipment can represent a very powerful tool for ecological studies. However, as we outline in Chapter 2, there can be serious statistical and experiment design issues (e.g. pseudo-replication) associated with expensive equipment that is a one-off, can cost a lot of money to maintain, and/or for which only a small number can be purchased. There are limitations in the spatial and temporal inferences that can be made from work at a single site or from a small number of sites in countries the size of the United States and Australia. There also can be limitations to conclusions drawn about environmental impacts from studies with few sites (Cunningham and Lindenmayer 2017). Finally, with a strong focus on infrastructure and questions posed post-hoc (after the infrastructure has been established), there is a danger that the two will be mismatched. That is, the infrastructure cannot be used to address the key questions that need to be answered. On the positive side, there is a strong commitment for broad geographic (continental) coverage of the infrastructure, which would be necessary to address large-scale problems such as climate change. In the case of NEON, mobile or ‘relocatable’ units are being installed for each of the 20 ecoclimatic ‘domains’ selected for study. We also appreciate that the NEON process identified 20 domains based on a sophisticated statistical analysis of climatic state variables and wind vectors. This approach was a valuable design effort, but it was not proceeded by a process of setting questions at the beginning or of integrating questions with the design. Governance structure NEON and TERN have been characterised by complex and often radically changing governance structures. For example, TERN has a CEO, a governing advisory board, and other architecture such as an Executive Group and a Scientific Advisory Committee. Governance has been reviewed within TERN (KPMG 2014) and these issues are planned to be tackled in the future. Initially, NEON was supervised by a CEO and a board of 15 members. More recently, the decision was made to take the program away from NEON Inc. and partner with a contractor to construct, operate and maintain the infrastructure (under contract with and supervision by the National Science Foundation). Both NEON and TERN have had serious issues with turnover in leadership. This turnover occurs, in part perhaps, because of the enormous challenges of providing robust leadership for such complex, interdisciplinary (yet also highly

4 – The problematic, the ef fective and the ugly – some case studies

145

Box 4.5. Synthesis versus integration – a major challenge within TERN The issue of how initiatives like TERN can be a network and deliver more than the sum of their collective parts has exercised the minds of many of the scientists within TERN for years. This is also the case within the Long-Term Ecological Research Network (LTERN) within TERN – the facility comprising long-term research and monitoring plots in an array of ecosystems across Australia. A key challenge is that the different sets of plots within LTERN have been established for different reasons to answer different questions in different ecosystems (Lindenmayer et al. 2015d). Gathering data to answer these different questions has almost always involved employing different field methods that are appropriately suited to the particular ecosystem in question. Integration of datasets across the different studies has therefore proven to be extremely difficult. To some extent, a lack of cross-plot network synthesis might be seen as a reason to support approaches that measure the same thing everywhere (e.g. the PPBio model discussed earlier in this chapter). However, this is counter-balanced by the need to match the entities being measured with the questions that need to be addressed, and this will often vary according to the species, communities, ecosystems or other entities being monitored (see Chapter 2; Lindenmayer et al. 2015d). In contrast to the difficulties of studies based on data integration, synthesis work based on the collective experience of leaders of different studies has been very successful. Articles outlining the values of long-term studies and a book on the long-term trends in biodiversity in different ecosystems across the Australian continent have been published (Lindenmayer and Gibbons 2012; Lindenmayer et al. 2012b, 2014c, 2015d). Yet other work on applying methods for formal IUCN protocols for ecosystem assessment was undertaken for a range of ecosystems (e.g. see Burns et al. 2015). Further synthesis studies are planned, but the issues around how best to complete data integration work remain unresolved, not only within LTERN, but across TERN more broadly (in part because it will be questiondependent). Continued efforts by many outstanding scientists are trying to determine whether such problems have a resolution. Notably, similar kinds of integration problems to those facing LTERN and TERN face the US-LTER network in which, again, the challenges of conducting cross-site comparisons across very different systems have yet to be resolved. Thus, achieving the aim of creating a true network of LTER sites that is more than the sum of its parts has remained elusive. The issues associated with the challenges of integration versus synthesis within TERN are no longer apt. The LTERN facility within TERN will be terminated in late 2017 and Australia will lose much of its capacity to do long-term ecological research and monitoring (Lindenmayer et al. 2017).

bureaucratic) initiatives spanning many partners and large numbers of scientists, with often markedly different views on how things should be done. There are other key issues that add to the complexity of leadership. For example, TERN has experienced significant uncertainty in funding, changes in the political party

146

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

heading the government, and changes in the Australian Government portfolio to which it reports. In the case of NEON, there also has been a difficult relationship with the major funder, the National Science Foundation, with suggestions by some of excessive micromanagement by the latter. Indeed, the challenges are enormous as NEON and TERN can be seen as non-traditional science initiatives in which the emphasis is on data collection and curation rather than scientific discovery or answering key scientific questions. Such a remit is quite different from the background of most science leaders. Because of their extensive bureaucratic structures and relatively large size, it will be interesting to determine what metrics are used to gauge the success of NEON and TERN over time. TERN is gathering information on how frequently the datasets that it has gathered will be re-used as one of its performance metrics, with an added focus on ensuring that there are mechanisms for citing dataset use in journal articles. A further key issue with the governance structures of both NEON and TERN is that there are very high transaction costs in being part of a large group of collaborating scientists, such as travel to meetings, time spent at meetings (and investigator salaries), and extensive and frequent reporting. These transaction costs deflect from the important business of gathering data to answer key questions as part of effective ecological monitoring. High transaction costs can act as a major disincentive for prolonged engagement by already heavily time-pressed scientists. Finally, there are other non-trivial governance issues with NEON, such as major increases in operations cost. The price tag for it was recently (July 2016) increased by US$69 million, although the estimated cost to run the network of US$60–80 million per annum was not considered reliable and is likely to be more. This potentially creates problems for the rest of the budget allocated to biological sciences within the National Science Foundation (Science Magazine, 8 July 2016, p. 104). Indeed, governance issues such as missed deadlines, cost overruns, frequently changing directions, micromanagement by the National Science Foundation and internal (as well as external) conflict has long characterised the development of NEON (Mervis 2015; Cesare 2016). Key issues for resolution in the future of NEON and TERN Like all of our case studies in Chapter 4, we argue that the NEON and TERN programs have been well-intentioned and genuinely aimed at attempting to improve long-term ecological research and monitoring (Scholes et al. 2008; Pennisi 2010). There are many valuable aspects of both programs and much has been achieved and learned through both initiatives. For example, NEON and TERN have provided new sources of funding and promoted new collaborative efforts among scientists in often quite different disciplines who may never have previously worked together (TERN 2013). Attempts to bring together many other scientists is

4 – The problematic, the ef fective and the ugly – some case studies

147

an inherently good thing, although there are risks that other scientists who are not involved may be alienated (as was the case in the US International Biological Program) (Boffey 1976). In addition, the new infrastructure may prove to be valuable for generating new scientific insights. Innovations associated with data sharing and re-use of datasets such as developing citation metrics for citing datasets are also valuable steps forward (see Parsons and Minster 2010; Michener and Jones 2012). However, we suggest there are some key questions that need to be answered before it is possible to determine whether initiatives like NEON and TERN (as well as numerous others like SAEON in South Africa) have been successful, and the general approach and advantages of developing large-scale observation networks (see Hernandez et al. 2012; Rüegg et al. 2014) can be considered to be a costefficient way of conducting effective ecological monitoring at large scales. 1. Can NEON, TERN and other observation networks effectively convert huge data streams into useful scientific knowledge and understanding (and in turn enhance the management of natural resources and lead to improved public policy)? As outlined above, these large-scale initiatives have a focus on data gathering and storage but limited or no funding for research. Analyses of very large datasets, particularly those from multiple sources can be extremely challenging and it will be important to assess the effectiveness of efforts to develop robust insights from the vast amounts of new information generated (McDowell 2015). We do note that there are new generations of scientists raised in a data-rich world that are excited by the very large data streams generated by observation networks, and these data streams are undoubtedly very useful for many purposes not imagined a priori. Notably, initiatives like NEON have developed tutorials to help guide researchers in the use of the data that they are generating and/or collecting (http://neondataskills.org/teaching-modules/ disturb-events-co13/). 2. How do global or national observation networks interface with the scales of, and solutions to, environmental problems and the scales needed for ecological monitoring? For example, some workers might argue that global networks are appropriately scaled for tackling global problems but they may not be appropriate for tackling smaller-scaled problems where site, landscape, catchment or regional solutions are required. 3. Can NEON, TERN and other observation networks function as true networks where the sum of the data gathering and other activities within them function more than the sum of their parts? Synergy between components of these largescale programs is important because it underscores the value of a network. This is a major challenge in the TERN program where some facilities have worked well together but few projects successfully integrate data or ideas from

148

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

all of the facilities as might be expected to occur in a true network. (See also Box 4.5.) 4. Can large-scale observation networks that are essentially based on non-question-driven kinds of monitoring and which generate very large data streams actually detect major changes in ecological conditions that are needed to understand environmental change and how that change is influencing ecosystems and biodiversity? Similarly, can observation networks help enhance effective ecological monitoring and, for example, help determine the effectiveness of key management interventions that lie at the heart of improving environmental outcomes? As the Editor of Nature pointed out in 2012 (volume 487, p. 406), massive data streams can readily produce major errors in data interpretation and conclusions and this is occurring on an increasingly frequent basis (see also Maldonado et al. 2015). Conversely, some patterns generated from large observation networks may not be visible from smaller data streams like those gathered from area-focused, question-driven monitoring (e.g. Groffman et al. 2014). 5. Are major funding initiatives like NEON a more efficient and effective way of investing limited resources in environmental monitoring and management than more focused, question-based empirical investigations? Indeed, given competition for limited (and often rapidly shrinking resources), we would not (and nor would most practising scientists) want to see the funding for NEON and TERN infrastructure lead to the demise of other important ecological programs. We suggest that careful consideration needs to be given to the balance between question-driven monitoring and large-scale non-questiondriven monitoring initiatives like NEON, TERN and other observation networks. Indeed, the balance between funding for infrastructure versus research has been a prominent point of contention in the marine sector in the United States (Kintisch 2015; Witze 2015). 6. Can important place-based, long-term site and landscape studies be integrated with much larger scaled observational networks? For example, NEON has a strong focus on climate change. Big datasets gathered at large scales by sensing arrays and other technology may well be very important, but it remains unclear how this approach may or may not fit with finer scaled, place-based responses to climate change, such as movements of particular species or colonisation of new areas by invasive species. Some work has been done on how such integration might take place (e.g. Turner 2014) and initiatives like the Group on Earth Observations Biodiversity Observation Network (GEO BON) are attempting to coordinate across governments to connect datasets gathered from in-situ observations, airborne sensors and satellites at scales from genes to species and even ecosystems (Scholes et al. 2012) NEON also is trying to address this issue, for example, through workshops on examining

4 – The problematic, the ef fective and the ugly – some case studies

149

remotely sensed data (e.g. http://www.neonscience.org/updates-events/update/ apply-now-two-nsfneon-remote-sensing-data-workshops). Recent advances in hierarchical statistical theory in combination with increasing computational resources have enabled a new generation of methods for integrating data collected across spatial and temporal scales (Banerjee et al. 2014). For example, Wilson et al. (2011) developed statistical methods to scale up from field observations of above-ground biomass at centimetre resolution with satellite-derived observations of vegetation dynamics at 30-m resolution to enable use of multiple decades of satellite observations to monitor post-fire recovery and ecosystem resilience (Wilson et al. 2015). However, there remains a clear need for on-the-ground work to calibrate the data that are generated from satellites and other kinds of remote sensors, including for promoting improved environmental management (Cord et al. 2015). A related question is: How can we effectively integrate different datasets that have been collected for different reasons to answer different questions of management significance in different places? 7. Can the leadership of large-scaled observational networks meet the challenge of generating (and then helping to answer) valuable practical questions of environmental management and policy significance? Can this leadership truly lead such large and complicated projects to new levels of scientific understanding?

Box 4.6. The marine people can do it – why can’t terrestrial people do it too? Some large research infrastructure programs have been very successful in the marine domain. A sister program to TERN in Australia – the Integrated Marine Observing System (IMOS) – has not been characterised by the very difficult gestational issues faced in the terrestrial domain. Notably, IMOS also has received considerably more funding than TERN. Some people have asked: Why can’t the terrestrial people get their act together? One of the key reasons is that the dominant driver in marine and aquatic ecosystems is water – a physical entity that can be readily measured by sensors and related kinds of infrastructure. By contrast, the biota is the dominant driver in the majority of terrestrial ecosystems, and the way biota is best measured and monitored (and indeed which elements of the biota are best to measure and monitor) varies on an ecosystem-by-ecosystem basis (Odum 1962). Some attributes of terrestrial ecosystems can, of course, be readily measured remotely (e.g. via satellites), whereas others need measurements at much finer-scale resolution (Turner 2014). There are other advantages of marine ecosystems, which are dominated by physical entities in that this allows ‘an easier fit’ for infrastructure such as ships to deploy infrastructure to gather data streams (such as buoys).

150

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

We currently do not have the answers to the seven questions listed above. To some extent they cannot be resolved until the ‘experiment’, that is big science and big data gathered through large-scale observation networks, is subject to rigorous assessment of its performance and the utility of the data gathered has been demonstrated. A potential risk here is that the trend towards large-scale observation networks is another ecological fad (sensu Redford et al. 2013) which is later found to have been far less effective than the more traditional kinds of question-based science that it is has often replaced. A likely positive outcome of attempts to answer some of the questions we have posed in this case study on NEON and TERN will be to develop some kind of balance between passive observation networks focused on collecting data and investigator-driven research aimed at answering questions. In fact, there have been relatively recent efforts to change marine science funding formulas in the United States in which science infrastructure spending has been reduced with a greater funding emphasis on research (Kintisch 2015).

The ugly As with our previous edition of this text, we have chosen not to devote any significant space to ‘The Ugly’ category of monitoring in this book. Hopefully, there are very few examples! Let the one example we give in Box 4.7 suffice.

Summary In this chapter, we have presented a series of case studies of long-term ecological research and monitoring programs that demonstrate what we consider to be problematic and effective monitoring. We suggest there are sets of common problems with problematic programs and, conversely, a suite of features that characterise effective monitoring programs. To be fair, none of the effective programs are perfect nor do the problematic programs lack redeeming features. In our view, the problematic programs generally have two or more of the following deficiencies: (1) a lack of well-defined scientifically tractable questions; (2) the absence of a useful conceptual model; (3) no apparent experimental design or a poor experimental design; and (4) laundry lists of attributes to be measured. In contrast, effective long-term studies usually have two or more of the following features: (1) well-defined and scientifically tractable questions, with new questions often posed as new insights are gained that do not breach the integrity of long-term research and monitoring; (2) a simple but useful conceptual model that is applied to help frame questions and identify the ecological processes underlying observed changes or patterns in measured phenomena; (3) a rigorously developed and applied experimental design; (4) a carefully selected set of variables or entities

4 – The problematic, the ef fective and the ugly – some case studies

151

Box 4.7. The Crawford Caper The Woods Hole Oceanographic Institution (WHOI) research vessel Crawford completed regular scientific voyages between Massachusetts and Bermuda. On the afternoon of 19 March 1960, with Bermuda as the last port of call, the Crawford arrived at the WHOI pier and quickly cleared customs and immigration with the usual New Bedford (Massachusetts) agents. When unloading was nearly finished, four Boston customs agents materialised to confiscate the booze that, by longstanding practice, voyage participants had been bringing home to the United States. More than 80 cases were confiscated, two personal vehicles were seized, and Crawford captain David Casiles was charged with allowing illegal importation. Another WHOI ship, Chain, due in three days later, was cryptically notified of these events and Captain Hiller put all the liquor aboard the ship in bond (for later seizure, as it turned out). Crawford was allowed to continue operating, although it was for a time officially the property of US Customs (which assigned custody to Captain Casiles!). A three-person internal investigating committee determined that the liquor was not intended for commercial purposes, and payment of a US$4500 fine (contributed by those who had benefited from the contraband over the years) cleared Captain Casiles. New policies were put in place by WHOI, but life with US Customs was made difficult for some time (modified from Cullen 2005). Ostensibly, a major justification for this ‘rum run’ route, was to monitor temperature and salinity profiles (T-S diagrams) in the ocean between Massachusetts to Bermuda. Although WHOI has done some of the most exemplary research and monitoring of the world’s oceans, it seems to us that the results leading to this ‘caper’ qualifies as mindless monitoring of the ‘ugly’ type.

to be measured; and (5) they take advantage of serendipity (defined as keeping one’s eyes, ears and mind open to new questions and then moving expeditiously to address them) in combination with adaptive monitoring to update and energise long-term research (as with analyses of radioactivity at Rothamsted and the discovery of acid rain at the Hubbard Brook Experimental Forest). As well as these four key design and implementation characteristics, successful programs also appear to have one or more of the following additional features (Strayer et al. 1986; Lovett et al. 2007): (1) management relevance; (2) strong and stable leadership; (3) strong advocates; (4) ongoing base funding; (5) welldeveloped partnerships between important stakeholders; and (6) careful attention to database management and archiving of samples.

References Alberta Biodiversity Monitoring Institute (2009) Program overview. Alberta Biodiversity Monitoring Institute, Edmonton.

152

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Alberta Biodiversity Monitoring Institute (2016) Quality Management Plan. Alberta Biodiversity Monitoring Institute, Edmonton. Alberta Biodiversity Monitoring Program (2009) Program overview and consultation backgrounder. Alberta Biodiversity Monitoring Program, Edmonton, Canada. Banerjee S, Carlin BP, Gelfand AE (2014) Hierarchical Modeling and Analysis for Spatial Data. CRC Press, Boca Raton, Florida. Banks SC, Dujardin M, McBurney L, Blair D, Barker M, Lindenmayer DB (2011) Starting points for small mammal population recovery after wildfire: recolonisation or residual populations? Oikos 120, 26–37. doi:10.1111/j.1600-0706.2010.18765.x Battles JJ, Fahey TJ, Driscoll C, Blum J, Johnson CE (2014) Restoring soil calcium reverses forest decline. Environmental Science and Technology Letters 1, 15–19. Bayne EM, Stralberg D, Nixon A (2015) Adapting monitoring to more effectively assess the impacts of climate change on Alberta’s biodiversity. Alberta Biodiversity Monitoring Institute, Edmonton, Canada. Beale R (2007) If Trees Could Speak: Stories of Australia’s Greatest Trees. Allen and Unwin, Sydney. Blair D, McBurney L, Banks SWB, Lindenmayer DB (2016) Disturbance gradient shows logging affects plant functional groups more than fire. Ecological Applications 26, 2280–2301. doi:10.1002/eap.1369 Boffey PM (1976) International Biological Program – was it worth the cost and effort? Science 193, 866–868. doi:10.1126/science.193.4256.866 Bormann FH, Likens GE (1962) Hydrologic-mineral cycle interaction in a small watershed. Proposal to the National Science Foundation. Bormann FH, Likens GE (1967) Nutrient cycling. Science 155, 424–429. doi:10.1126/ science.155.3761.424 Bormann FH, Likens GE (1979) Pattern and Process in a Forested Ecosystem. SpringerVerlag, New York. Bormann FH, Siccama TG, Likens GE, Whittaker RH (1970) The Hubbard Brook Ecosystem Study: composition and dynamics of the tree stratum. Ecological Monographs 40, 373–388. doi:10.2307/1942336 Bormann FH, Likens GE, Siccama TG, Pierce RS, Eaton JS (1974) The export of nutrients and recovery of stable conditions following deforestation at Hubbard Brook. Ecological Monographs 44, 255–277. doi:10.2307/2937031 Boutin S, Haughland DL, Schieck J, Herbers J, Bayne E (2009) A new approach to forest biodiversity monitoring in Canada. Forest Ecology and Management 2585, 5168–5175. Burns EL, Lindenmayer DB, Stein J, Blanchard W, McBurney L, Blair D, Banks SC (2015) Ecosystem assessment of mountain ash forest in the Central Highlands of Victoria, south-eastern Australia. Austral Ecology 40, 386–399. doi:10.1111/ aec.12200 Burton PJ, Messier C, Smith DW, Adamowicz WL (2003) Towards sustainable manage­ ment of the boreal forest. National Research Council of Canada, Ottawa, Canada.

4 – The problematic, the ef fective and the ugly – some case studies

153

Buso DC, Likens GE, Eaton JS (2000) Chemistry of precipitation, stream water and lake water from the Hubbard Brook Ecosystem Study: A record of sampling protocols and analytical procedures. USDA Forest Service, Northeastern Research Station, Newtown Square, Pennsylvania. Cesare C (2016) Ecology aims high. Nature 529, 274–276. doi:10.1038/529274a Chilvers BL, Lawler IR, Macknight F, Marsh H, Noad M, Paterson R (2005) Moreton Bay, Queensland, Australia: an example of the co-existence of significant marine mammal populations and large-scale coastal development. Biological Conservation 122, 559–571. doi:10.1016/j.biocon.2004.08.013 Commonwealth of Australia (2011) Australia State of the Environment. Commonwealth of Australia, Canberra. Cord E, Seppelt R, Turner W (2015) Monitor ecosystem services from space. Nature 525, 33. doi:10.1038/525033a Cressey D (2015) GM wheat that emits pest alarm signals fails in field trials. Wheat modified to send out insect pheromones does not repel aphids. Nature News 25 June 2015. doi:10.1038/nature.2015.17854 Cullen V (2005) Down to the Sea for Science: 75 Years of Ocean Research, Education, and Exploration at the Woods Hole Oceanographic Institution. Woods Hole Oceanographic Institution, Woods Hole, Massachusetts. Cunningham RB, Lindenmayer DB (2017) Approaches to landscape scale inference and study design. Current Landscape Ecology Reports 2, 42–50. doi:10.1007/ s40823-016-0019-4 Dennison WC, Abal EG (1999) Moreton Bay Study. South East Queensland Regional Water Quality Management Strategy, Brisbane. Ecosystem Health Monitoring Program (2007) Ecosystem Health Monitoring Program, 2006–07 Annual Technical Report. South East Queensland Healthy Waterways Partnership, Brisbane. Ecosystem Health Monitoring Program (2008) Report Card 2008 for the waterways and catchments of south-east Queensland. Ecosystem Heath Monitoring Program, South East Queensland Healthy Waterways Partnership, Brisbane, Queensland. Fahey TJ, Siccama TG, Driscoll CT, Likens GE, Campbell J, Johnson CE, Aber JD, Cole JJ, Fisk MC, Groffman PM, Holmes RT, Schwarz PA, Yanai RD (2005) The biogeochemistry of carbon at Hubbard Brook, New Hampshire, USA. Biogeochemistry 75, 109–176. doi:10.1007/s10533-004-6321-y Fisher RA (1925) Statistical Methods for Research Workers. Oliver and Boyd, Edinburgh, Scotland. Fisher RA (1935) The Design of Experiments. Oliver and Boyd, Edinburgh, Scotland. Fox BJ (1982) Fire and mammalian secondary succession in an Australian coastal heath. Ecology 63, 1332–1341. doi:10.2307/1938861 Government of British Columbia (2011) Mountain Pine Beetles in British Columbia. Available on-line at http://www.for.gov.bc.ca/hfp/mountain_pine_beetle/, www.for. gov.bc.ca/hfp/mountain_pine_beetle/.

154

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Groffman PM, Driscoll CT, Eagar C, Fisk MC, Fahey TJ, Holmes RT, Likens GE, Pardo L (2004a) A new governance structure for the Hubbard Brook Ecosystem Study. Bulletin of the Ecological Society of America 85, 5–6. doi:10.1890/00129623(2004)85[5b:ANGSFT]2.0.CO;2 Groffman PM, Driscoll CT, Likens GE, Fahey TJ, Holmes RT, Eagar C, Aber JD (2004b) Nor gloom of night: a new conceptual model for the Hubbard Brook Ecosystem Study. Bioscience 54, 139–148. doi:10.1641/00063568(2004)054[0139:NGONAN]2.0.CO;2 Groffman PM, Hardy JP, Driscoll CT, Fahey TJ (2006) Snow depth, soil freezing, and fluxes of carbon dioxide, nitrous oxide and methane in a northern hardwood forest. Global Change Biology 12, 1748–1760. doi:10.1111/j.1365-2486.2006.01194.x Groffman PM, Cavender-Bares J, Bettez ND, Grove J, Hall SJ, Heffernan JB, Hobbie SE, Larson KL, Morse JL, Neill C, Nelson K, O’Neil-Dunne J, Ogden L, Pataki DE, Polsky C, Chowdhury RR, Steele MK (2014) Ecological homogenization of urban USA. Frontiers in Ecology and the Environment 12, 74–81. doi:10.1890/120374 Hernandez RR, Mayernik MS, Murphy-Mariscal ML, Allen M (2012) Advanced technologies and data management practices in environmental science: lessons from academia. Bioscience 62, 1067–1076. doi:10.1525/bio.2012.62.12.8 Hero J-M, Castley JG, Malone M, Lawson B, Magnusson WE (2010) Long-term ecological research in Australia: innovative approaches for future research. Australian Zoologist 35, 216–228. doi:10.7882/AZ.2010.010 Hero J-M, Castley JG, Butler SA, Lollback G (2013) Biomass estimation within an Australian eucalypt forest: meso-scale spatial arrangement and the influence of sampling intensity. Forest Ecology and Management 310, 547–554. doi:10.1016/j. foreco.2013.08.062 Hero J-M, Castley G, Lollback G (2014) PPbio Karawatha Field Manual. Environmental Futures Centre, Griffith University, Brisbane. Hobbs WO, Lafrancois BM, DiDonato E (2015) Nearshore conditions in the Great Lakes national parks: a baseline water quality and toxicological assessment. Park Science 32, 36–44. Holmes RT (2007) Understanding population change in migratory songbirds: longterm and experimental studies of Neotropical migrants in breeding and wintering areas. The Ibis 149(Suppl.2), 2–13. doi:10.1111/j.1474-919X.2007.00685.x Holmes RT, Likens GE (2016) Hubbard Brook: The Story of a Forest Ecosystem. Yale University Press, New Haven, Connecticut. Hurlbert SH (1984) Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54, 187–211. doi:10.2307/1942661 Jackson LE, Paulsen SG (2009) Special issue: The Eighth Symposium of the Environmental Monitoring and Assessment Program (EMAP) – research and partnerships for accountability. Environmental Monitoring and Assessment 150, 1–2. doi:10.1007/s10661-008-0667-x Johnston K, Beatson A (2015) Healthy Waterways Social Science Research Report 2015. QUT Business School Available at https://eprints.qut.edu.au/93606.

4 – The problematic, the ef fective and the ugly – some case studies

155

Johnston K, Beatson A (2016) Healthy Waterways Social Science Research Report 2016. QUT Business School. Kaufmann PR, Peck DV, Paulsen S, Seelinger CW, Hughes RM, Whittier TR, Kamman NC (2014) Lakeshore and littoral physical habitat structure in a national lakes assessment. Lake and Reservoir Management 30, 192–215. doi:10.1080/10402381.201 4.906524 Keith H, Mackey B, Lindenmayer DB (2009) Re-evaluation of forest biomass carbon stocks and lessons from the world’s most carbon-dense forests. Proceedings of the National Academy of Sciences of the United States of America 106, 11635–11640. doi:10.1073/pnas.0901970106 Keith H, Lindenmayer DB, Mackey BG, Blair D, Carter L, McBurney L, Okada S, Konishi-Nagano T (2014a) Accounting for biomass carbon stock change due to wildfire in temperate forest landscapes in Australia. PLoS One 9, e107126. doi:10.1371/journal.pone.0107126 Keith H, Lindenmayer DB, Mackey BG, Blair D, Carter L, McBurney L, Okada S, Konishi-Nagano T (2014b) Managing temperate forests for carbon storage: impacts of logging versus forest protection on carbon stocks. Ecosphere 5(6), 1–34. doi:10.1890/ES14-00051.1 Keith H, Vardon M, Stein J, Stein J, Lindenmayer DB (2016) Experimental Ecosystem Accounts for the Central Highlands of Victoria. The Australian National University, Canberra. Available at http://fennerschool-associated.anu.edu.au/ documents/CLE/VCH_Accounts_Summary_FINAL_for_pdf_distribution.pdf. Keith H, Vardon M, Stein JA, Stein JS, Lindenmayer DB (2017) Ecosystem accounts define explicit and spatial trade-offs for managing natural resources. Nature Ecology and Evolution doi:10.1038/s41559-017-0309-1 Kintisch E (2015) A moment of truth arrives for marine science. Science 347, 463. doi:10.1126/science.347.6221.463 KPMG (2014) National Collaborative Research Infrastructure Strategy Project Reviews – Overarching Report. Available at: https://docs.education.gov.au/system/files/doc/ other/ncris_project_reviews_final_report_web.pdf Krebs CJ (2012) Biodiversity monitoring in Canada’s Yukon: The Community Ecological Monitoring Program. In Biodiversity Monitoring in Australia. (Eds DB Lindenmayer, P Gibbons) pp. 151–155. CSIRO Publishing, Melbourne. Lamb EG, Bayne E, Holloway G, Schieck J, Soutin S, Herbers J, Haughland DL (2009) Indices for monitoring biodiversity change: are some more effective than others. Ecological Indicators 9, 432–444. doi:10.1016/j.ecolind.2008.06.001 Likens GE (Ed.) (1985) An Ecosystem Approach to Aquatic Ecology: Mirror Lake and its Environment. Springer-Verlag, New York. Likens GE (1992) The Ecosystem Approach: Its Use and Abuse. Ecology Institute, Oldendorf/Luhe, Germany. Likens GE (2004) Some perspectives on long-term biogeochemical research from the Hubbard Brook Ecosystem Study. Ecology 85, 2355–2362. doi:10.1890/03-0243

156

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Likens GE (2011) Limnological measures related to climate change in the Hubbard Brook Valley, USA. Inland Waters 1, 93–99. doi:10.5268/IW-1.2.378 Likens GE, Bormann FH (1995) Biogeochemistry of a Forested System. Springer-Verlag, London. Likens G, Buso D (2010) Long-term changes in streamwater chemistry following disturbance in the Hubbard Brook Experimental Forest, USA. Verhandlungen der Internationalen Vereinigung fur Theoretische und Angewandte Limnologie 30, 1577–1581. Likens GE, Franklin JF (2009) Ecosystem thinking in the Northern Forest – and beyond. Bioscience 59, 511–513. doi:10.1525/bio.2009.59.6.9 Likens GE, Bormann FH, Johnson NM, Fisher DW, Pierce RS (1970) Effects of forest cutting and herbicide treatment on nutrient budgets in the Hubbard Brook watershed-ecosystem. Ecological Monographs 40, 23–47. doi:10.2307/1942440 Likens GE, Bormann FH, Pierce RS, Eaton JS, Johnson NM (1977) Biogeochemistry of a Forested Ecosystem. Springer-Verlag, New York. Likens GE, Driscoll CT, Buso DC, Siccama TG, Johnson CE, Ryan DF, Lovett GM, Fahey T, Reiners WA (1994) The biogeochemistry of potassium at Hubbard Brook. Biogeochemistry 25, 61–125. doi:10.1007/BF00000881 Likens GE, Driscoll CT, Buso DC, Siccama TG, Johnson CE, Lovett GM, Fahey TJ, Reiners WA, Ryan DF, Martin CW, Bailey SW (1998) The biogeochemistry of calcium at Hubbard Brook. Biogeochemistry 41, 89–173. doi:10.1023/A:1005984620681 Likens GE, Driscoll CT, Buso DC, Mitchell MJ, Lovett GM, Bailey SW, Siccama TG, Reiners WA, Alewell C (2002) The biogeochemistry of sulfur at Hubbard Brook. Biogeochemistry 60, 235–316. doi:10.1023/A:1020972100496 Lindenmayer DB (2009) Forest Pattern and Ecological Process: A Synthesis of 25 Years of Research. CSIRO Publishing, Melbourne. Lindenmayer DB, Cunningham RB (1997) Patterns of co-occurrence among arboreal marsupials in the forests of central Victoria, southeastern Australia. Australian Journal of Ecology 22, 340–346. doi:10.1111/j.1442-9993.1997.tb00680.x Lindenmayer DB, Gibbons P (Eds) (2012) Biodiversity Monitoring in Australia. CSIRO Publishing, Melbourne. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/s10021-010-9394-6 Lindenmayer DB, Likens GE (2013) Don’t do big-data science backwards. Nature 499, 284. doi:10.1038/499284d Lindenmayer DB, Ough K (2006) Salvage harvesting in the montane ash forests of the Central Highlands of Victoria, south-eastern Australia. Conservation Biology 20, 1005–1015. doi:10.1111/j.1523-1739.2006.00501.x Lindenmayer DB, Cunningham RB, Tanton MT, Smith AP (1990) The conservation of arboreal marsupials in the montane ash forests of the Central Highlands of Victoria, south-east Australia. 2. The loss of trees with hollows and its implications

4 – The problematic, the ef fective and the ugly – some case studies

157

for the conservation of Leadbeater’s Possum Gymnobelideus leadbeateri McCoy (Marsupialia, Petauridae). Biological Conservation 54, 133–145. doi:10.1016/0006-3207(90)90138-F Lindenmayer DB, Craig SA, Linga T, Tanton MT (1991a) Public participation in stagwatching surveys for a rare mammal – applications for environmental education. Australian Journal of Environmental Education 7, 63–70. doi:10.1017/ S0814062600001865 Lindenmayer DB, Cunningham RB, Tanton MT, Nix HA, Smith AP (1991b) The conservation of arboreal marsupials in the montane ash forests of the Central Highlands of Victoria, South-East Australia. 3. The habitat requirements of Leadbeater’s Possum Gymnobelideus leadbeateri and models of the diversity and abundance of arboreal marsupials. Biological Conservation 56, 295–315. doi:10.1016/0006-3207(91)90063-F Lindenmayer DB, Cunningham RB, Tanton MT, Smith AP, Nix HA (1991c) Characteristics of hollow-bearing trees occupied by arboreal marsupials in the montane ash forests of the Central Highlands of Victoria, south-east Australia. Forest Ecology and Management 40, 289–308. doi:10.1016/0378-1127(91)90047-Y Lindenmayer DB, Welsh A, Donnelly CF, Meggs RA (1996) Use of nest trees by the Mountain Brushtail Possum (Trichosurus Caninus) (Phalangeridae, Marsupialia). 1. Number of occupied trees and frequency of tree use. Wildlife Research 23, 343–361. doi:10.1071/WR9960343 Lindenmayer DB, Cunningham RB, Donnelly CF, Franklin JF (2000) Structural features of old growth Australian montane ash forests. Forest Ecology and Management 134, 189–204. doi:10.1016/S0378-1127(99)00257-1 Lindenmayer DB, Cunningham RB, MacGregor C, Incoll RD (2003) A long-term monitoring study of the population dynamics of arboreal marsupials in the Central Highlands of Victoria. Biological Conservation 110, 161–167. doi:10.1016/ S0006-3207(02)00171-4 Lindenmayer DB, Wood J, McBurney L, Michael D, Crane M, MacGregor C, Montague-Drake R, Banks S, Gibbons P (2011) Cross-sectional versus longitudinal research: a case study of trees with hollows and marsupials in Australian forests. Ecological Monographs 81, 557–580. doi:10.1890/11-0279.1 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks S, Likens GE, Franklin JF, Stein J, Gibbons P (2012a) Interacting factors driving a major loss of large trees with cavities in an iconic forest ecosystem. PLoS One 7, e41864. doi:10.1371/journal. pone.0041864 Lindenmayer DB, Likens GE, Andersen A, Bowman D, Bull CM, Burns E, Dickman CR, Hoffmann AA, Keith DA, Liddell MJ, Lowe AJ, Metcalfe DJ, Phinn SR, RussellSmith J, Thurgate N, Wardle GM (2012b) Value of long-term ecological studies. Austral Ecology 37, 745–757. doi:10.1111/j.1442-9993.2011.02351.x Lindenmayer DB, Blair D, McBurney L, Banks SC, Stein JAR, Hobbs RJ, Likens GE, Franklin JF (2013a) Principles and practices for biodiversity conservation and restoration forestry: a 30 year case study on the Victorian montane ash forests and

158

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

the critically endangered Leadbeater’s Possum. Australian Zoologist 36, 441–460. doi:10.7882/AZ.2013.007 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks S, Driscoll D, Smith A, Gill AM (2013b) Fire severity and landscape context effects on arboreal marsupials. Biological Conservation 167, 137–148. doi:10.1016/j.biocon.2013.07.028 Lindenmayer DB, Barton PS, Lane PW, Westgate MJ, McBurney L, Blair D, Gibbons P, Likens GE (2014a) An empirical assessment and comparison of species-based and habitat-based surrogates: a case study of forest vertebrates and large old trees. PLoS One 9, e89807. doi:10.1371/journal.pone.0089807 Lindenmayer DB, Blanchard W, McBurney L, Blair D, Banks SC, Driscoll DA, Smith A, Gill AM (2014b) Complex responses of birds to landscape-level fire extent, fire severity and environmental drivers. Diversity & Distributions 20, 467–477. doi:10.1111/ddi.12172 Lindenmayer DB, Burns E, Thurgate N, Lowe A (Eds) (2014c) Biodiversity and Environmental Change: Monitoring, Challenges and Direction. CSIRO Publishing, Melbourne. Lindenmayer D, Wood J, McBurney L, Blair D, Banks SC (2015a) Single large versus several small: the SLOSS debate in the context of bird responses to a variable retention logging experiment. Forest Ecology and Management 339, 1–10. doi:10.1016/j.foreco.2014.11.027 Lindenmayer DB, Barton P, Pierson JC (Eds) (2015b) Indicators and Surrogates of Biodiversity and Environmental Change. CSIRO Publishing, Melbourne. Lindenmayer DB, Blair D, McBurney L, Banks S (2015c) Mountain Ash: Fire, Logging and the Future of Victoria’s Giant Forests. CSIRO Publishing, Melbourne. Lindenmayer DB, Burns EL, Tennant P, Dickman CR, Green PT, Keith DA, Metcalfe DJ, Russell-Smith J, Wardle GW, Williams R, Bossard K, deLacey C, Hanigan IC, Bull CM, Gillespie G, Hobbs RJ, Krebs CJ, Likens GE, Porter J, Vardon M (2015d) Contemplating the future: acting now on long-term monitoring to answer 2050’s questions. Austral Ecology 40, 213–224. doi:10.1111/aec.12207 Lindenmayer DB, Pierson J, Barton P, Beger M, Branquinho C, Calhoun A, Caro T, Greig H, Gross J, Heino J, Hunter M, Lane P, Longo C, Martin K, McDowell WH, Mellin C, Salo H, Tulloch A, Westgate M (2015e) A new framework for selecting environmental surrogates. The Science of the Total Environment 538, 1029–1038. doi:10.1016/j.scitotenv.2015.08.056 Lindenmayer DB, Blanchard W, Blair D, McBurney L, Banks SC (2016) Environmental and human drivers of large old tree abundance in Australian wet forests. Forest Ecology and Management 372, 226–235. doi:10.1016/j.foreco.2016.04.017 Lindenmayer DB, Blanchard W, Blair D, McBurney L, Banks SC (2017a) Relationships between tree size and occupancy by cavity-dependent arboreal marsupials. Forest Ecology and Management 391, 221–229. doi:10.1016/j.foreco.2017.02.014 Lindenmayer DB, Burns EL, Dickman C, Green PT, Hoffmann AA, Keith DA et al. (2017b) Save Australia’s ecological research. Science 357, 557.

4 – The problematic, the ef fective and the ugly – some case studies

159

Lollback GW, Mebberson R, Evans N, Shuker JD, Hero J-M (2015) Estimating the abundance of the Bilby (Macrotis lagotis): a vulnerable, uncommon, noctural marsupial. Australian Mammalogy 37, 75–85. doi:10.1071/AM14024 Lovett GM, Likens GE, Buso DC, Driscoll CT, Bailey SW (2005) The biogeochemistry of chlorine at Hubbard Brook, New Hampshire, USA. Biogeochemistry 72, 191–232. doi:10.1007/s10533-004-0357-x Lovett GM, Burns DA, Driscoll CT, Jemkins JC, Mitchell MJ, Rustad L, Shanley JB, Likens GE, Haeuber R (2007) Who needs environmental monitoring? Frontiers in Ecology and the Environment 5, 253–260. doi:10.1890/1540-9295(2007)5[253:WNEM]2.0.CO;2 Luo Y, Ogle K, Tucker C, Fei S, Gao C, Ladeau S, Clark JS, Schimel DS (2011) Ecological forecasting and data assimilation in a data-rich era. Ecological Applications 21, 1429–1442. doi:10.1890/09-1275.1 Mackey B, Lindenmayer DB, Gill AM, McCarthy MA, Lindesay JA (2002) Wildlife, Fire and Future Climate: A Forest Ecosystem Analysis. CSIRO Publishing, Melbourne. Magnuson JJ, Robertson DM, Benson BJ, Wynne RH, Livingstone DM, Arai T, Assel RA, Barry RG, Card V, Kuusisto E et al. (2000) Historical trends in lake and river ice cover in the Northern Hemisphere. Science 289, 1743–1746. Magnusson WE, Lima AP, Luizao R, Costa FR, de Castillo CV, Kinupp VF (2005) RAPELD: a modification of the Gentry method for biodiversity surveys in longterm ecological research sites. Biota Neotropica 5, 19–24. doi:10.1590/S167606032005000300002 Maldonado C, Molina CI, Zizka A, Persson C, Taylor CM, Alban J, Chilquillo E, Ronsted N, Antonelli A (2015) Estimating species diversity and distribution in the era of Big Data: to what extent can we trust public databases? Global Ecology and Biogeography 24, 973–984. doi:10.1111/geb.12326 Manly BFJ (1992) The Design and Analysis of Research Studies. Cambridge University Press, Cambridge, England. McDonald M, Blair R, Bolgrien D, Brown B, Dlugosz J, Hale S, Hedtke S, Heggem D, Jackson L, Jones K, Levinson B, Linthurst R, Messer J, Olsen A, Paul J, Paulsen S, Stoddard J, Summers K, Veith G (2004) The U.S. Environmental Protection Agency’s Environmental Monitoring and Assessment Program. In Environmental Monitoring. (Ed. GB Wiersma) pp. 649–668. Taylor & Francis Group, CRC Press. McDowell WH (2015) NEON and STREON: opportunities and challenges for the aquatic sciences. Freshwater Science 34, 386–391. doi:10.1086/679489 Mervis J (2015) NSF looks for new contractor to finish troubled observatory. Science 350, 1454. doi:10.1126/science.350.6267.1454 Messer JJ, Linthurst RA, Overton WS (1991) An EPA program for monitoring ecological status and trends. Environmental Monitoring and Assessment 17, 67–78. doi:10.1007/BF00402462

160

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Michener WK, Jones MB (2012) Ecoinformatics: supporting ecology as data-intensive science. Trends in Ecology & Evolution 27, 85–93. doi:10.1016/j.tree.2011.11.016 Mitchell R, Mayer RA, Downhower J (1976) An evaluation of three biome programs. Science 192, 859–865. doi:10.1126/science.192.4242.859 Moreton Bay Waterways and Catchments Partnership (2003) Annual Report 2002– 2003. Moreton Bay Waterways and Catchments Partnership, Brisbane. Moreton Bay Waterways and Catchments Partnership (2004) Report Card 2004 for the waterways of south east Queensland. Moreton Bay Waterways and Catchments Partnership, Brisbane. Nolke N, Fehrmann L, Jaya IN, Tiryana T, Seidel D, Kleinn C (2015) On the geometry and allometry of big-buttressed trees – a challenge for forest monitoring: new insights from 3D-modeling with terrestrial laser scanning. iForest 8, 574–581. O’Neill G (2008) Jump-starting environmental monitoring. Ecos 143, 14–17. Odum EP (1962) Relationships between structure and function of the ecosystem. Japanese Journal of Ecology 12, 108–118. Oreskes N, Conway EM (2010) Merchants of Doubt. Bloomsbury Press, New York. Parsons MA, Minster J-B (2010) Data citation and peer review. Eos (Washington, D.C.) 91, 297–298. doi:10.1029/2010EO340001 Pennisi E (2010) A ground-breaking observatory to monitor the environment. Science 328, 418–420. doi:10.1126/science.328.5977.418 Pitt KA, Connolly RM, Maxwell P (2009) Redistribution of sewage-nitrogen in estuarine food webs following sewage treatment upgrades. Marine Pollution Bulletin 58, 573–580. doi:10.1016/j.marpolbul.2008.11.016 Raffa KF, Aukema BH, Bentz BJ, Carroll AL, Hicke JA, Turner MG, Romme WH (2008) Cross-scale drivers of natural disturbances prone to anthropogenic amplification: the dynamics of bark beetle eruptions. Bioscience 58, 501–517. doi:10.1641/B580607 Redford K, Padoch C, Sunderland T (2013) Fads, funding, and forgetting in three decades of conservation. Conservation Biology 27, 437–438. doi:10.1111/cobi.12071 Reilly WK (1989) Measuring for environmental results. EPA Journal 25, 2–4. Remler DK, Van Ryzin GG (2014) Research Methods in Practice: Strategies for Description and Causation. Sage, Los Angeles. Rothamsted Research (2006) Guide to classical and other long-term experiments, datasets and sample archive. Lawes Agricultural Trust, Bury St. Edmunds, England. Rothamsted Research (2009) Making a Difference: the past and future economic and societal impact of Rothamsted Research. Rothamsted Research, Harpenden, England. Rüegg J, Gries C, Bond-Lanberty G, Bowen GI, Felzer BS, McIntyre NE, Soranno PA, Vanderbilt KL, Weathers KC (2014) Completing the data life cycle: using information management in macrosystems ecology research. Frontiers in Ecology and the Environment 12, 24–30. doi:10.1890/120375 Saeck EA, O’Brien KR, Weber TR, Burford MA (2013) Changes to chronic nitrogen loading from sewage discharges modify standing stocks of coastal phytoplankton. Marine Pollution Bulletin 71, 159–167. doi:10.1016/j.marpolbul.2013.03.020

4 – The problematic, the ef fective and the ugly – some case studies

161

SAEON (2016) SAEON. Highlights 2011–2015. Department pf Science and Technology, National Research Foundation and South African Environmental Observation Network, South Africa. Salsburg D (2001) The Lady Tasting Tea: How Statistics Revolutionized Science in the Twentieth Century. W.H. Freeman, New York. Scholes RJ, Mace GM, Turner W, Geller GN, Jurgens N, Larigauderie A, Muchoney D, Walther BA, Mooney HA (2008) Toward a global biodiversity observing system. Science 321, 1044–1045. doi:10.1126/science.1162055 Scholes RJ, Walters M, Turak E, Saarenmaa H, Heip CHR, Tuama ÉÓ, Faith DP, Mooney HA, Ferrier S, Jongman RHG, Harrison IJ, Yahara T, Pereira HM, Larigauderie A, Geller G (2012) Building a global observing system for biodiversity. Current Opinion in Environmental Sustainability 4, 139–146. doi:10.1016/j. cosust.2011.12.005 Shank C, Schieck J, Farr D (2002) The Alberta Biodiversity Monitoring Program: Updated Technical Summary. Alberta Biodiversity Monitoring Program. Siccama TG, Fahey TJ, Johnson CE, Sherry TW, Denny EG, Girdler EB, Likens GE, Schwartz PA (2007) Population and biomass dynamics of trees in a northern hardwood forest at Hubbard Brook. Canadian Journal of Forest Research 37, 737–749. doi:10.1139/X06-261 Skinner JL, Gilliam E, Rohlin C-J (1998) The demographic future of the Moreton region. In Moreton Bay and Catchment. (Eds R Tibbetts, N Hall, WC Dennison) pp. 245–265. School of Marine Science, University of Queensland. Smith AP, Lindenmayer D (1988) Tree hollow requirements of Leadbeater’s Possum and other possums and gliders in timber production ash forests of the Victorian Central Highlands. Australian Wildlife Research 15, 347–362. Sólymos P, Morrison SF, Kariyeva J, Schieck J, Haughland DL, Azeria ET, Cobb T, Hinchliffe R, Kittson J, McIntosh AC, Narwani T, Pierossi P, Roy M-C, Sandybayev T, Boutin S, Bayne E (2015) Data and information management for the monitoring of biodiversity in Alberta. Wildlife Society Bulletin 39, 472–479. doi:10.1002/wsb.564 Soranno PA, Schimel DS (2014) Macrosystems ecology: big data, big ecology. Frontiers in Ecology and the Environment 12, 3. doi:10.1890/1540-9295-12.1.3 State of the Environment Advisory Council (1996) State of the Environment Australia 1996. Report for Commonwealth Minister for the Environment. CSIRO, Melbourne. Stevens DL (1994) Implementation of a national monitoring program. Journal of Environmental Management 42, 1–29. doi:10.1006/jema.1994.1057 Strayer DL, Glitzenstein JS, Jones C, Kolasa J, Likens GE, McDonnell M, Parker GG, Pickett STA (1986) Long-term ecological studies: An illustrated account of their design, operation, and importance to ecology. Institute of Ecosystem Studies, Millbrook, New York. TERN (2013) Transforming Australian Ecosystem Science. National Collaborative Research Infrastructure Strategy, Canberra. The State of Queensland Department of Infrastructure (2016) Shaping SEQ: Draft South East Queensland Regional Plan October 2016. Available at http://www.dilgp.

162

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

qld.gov.au/planning/regional-planning/review-of-the-south-east-queenslandregional-plan.html. Thomas RQ, Kellner JR, Clark DB, Peart DR (2013) Low mortality in tall tropical trees. Ecology 94, 920–929. doi:10.1890/12-0939.1 Tilman GD (2003) NEON: Addressing the Nation’s Environmental Challenges. National Academy Press, Washington, DC. Todd CR, Lindenmayer DB, Stamation K, Acevedo-Catteneo S, Smih S, Lumsden LF (2016) Assessing reserve effectiveness: application to a threatened species in a dynamic fire prone forest landscape. Ecological Modelling 338, 90–100. doi:10.1016/j.ecolmodel.2016.07.021 Turner W (2014) Sensing biodiversity. Science 346, 301–302. doi:10.1126/ science.1256014 United States Environmental Protection Agency (2015) National Coastal Condition Assessment. Field Operations Manual. United States Environmental Protection Agency, Office of Water, Washington, DC. Viggers JI, Weaver HJ, Lindenmayer DB (2013) Melbourne’s Water Catchments: Perspectives on a World-Class Water Supply. CSIRO Publishing, Melbourne. Whittaker RH, Bormann FH, Likens GE, Siccama TG (1974) The Hubbard Brook Ecosystem Study: forest biomass and production. Ecological Monographs 44, 233–254. doi:10.2307/1942313 Whittier TR, Paulsen SG, Larsen DP, Peterson SA, Herlihy AT, Kaufmann PR (2002) Indicators of ecological stress and their extent in the population of northeastern lakes: a regional-scale assessment. Bioscience 52, 235–247. doi:10.1641/00063568(2002)052[0235:IOESAT]2.0.CO;2 Wilson A, Silander JA, Gelfand A, Glenn JH (2011) Scaling up: linking field data and remote sensing with a hierarchical model. International Journal of Geographical Information Science 25, 509–521. doi:10.1080/13658816.2010.522779 Wilson AM, Latimer AM, Silander JA (2015) Climatic controls on ecosystem resilience: postfire regeneration in the Cape Floristic Region of South Africa. Proceedings of the National Academy of Sciences of the United States of America 112, 9058–9063. doi:10.1073/pnas.1416710112 Winter TC, Likens GE (Eds) (2009) Mirror Lake: Interactions among Air, Land and Water. University of California Press, Los Angeles. Witze A (2015) US ocean sciences told to plot fresh course. Nature 517, 538–539. doi:10.1038/nature.2015.16780 Wright-Stow AE, Winterbourn MJ (2003) How well do New Zealand’s streammonitoring indicators, the Macroinvertebrate Community Index and its quantitative variant, correspond? New Zealand Journal of Marine and Freshwater Research 37, 461–470. doi:10.1080/00288330.2003.9517180

5 The upshot – our general conclusions

Long-term ecological monitoring programs are fundamental to evidence-based environmental decision making. They are essential to gauge the effectiveness of management interventions and policy decisions (Box 5.1). They will become increasingly critical as part of efforts to mitigate, or better adapt to, the effects of rapid climate change and human-accelerated environmental change (Steffen et al. 2015; Hughes et al. 2017). It is now perhaps more important than ever to ensure that the currently limited (and declining) resources expended on environmental management in general and long-term research and monitoring in particular are spent as effectively as possible. Indeed, the risks of not doing monitoring properly are enormous. These risks include misuse of public and other funds, failure to detect major environmental problems before it is too late or extremely expensive to tackle them, and failure to recognise success and capitalise on it with management interventions. As we argued at the start of this book, we suggest that many of the attributes of good and effective monitoring programs are generic and apply equally to biodiversity monitoring as they do to environmental monitoring, and are as relevant to marine ecosystems as they are to terrestrial and aquatic ecosystems. This book has outlined some of the attributes that we consider to be pivotal for effective long-term ecological research and monitoring programs, where we define

164

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 5.1. Monitoring and management interventions How effective is environmental management? Are efforts to remove weeds, restore degraded areas, improve water and air quality or recover populations of endangered species actually working? Are protected areas effective for conserving biodiversity? Monitoring is critical to determine whether these and other kinds of management interventions are achieving their aims. It is, however, extraordinary how often management actions are not monitored and entire environmental programs go unassessed. These deficiencies characterise multi-million to even billion-dollar programs such as agri-environment schemes (Kleijn et al. 2006), river restoration projects (Bernhardt et al. 2005), efforts to tackle secondary salinity (Pannell and Roberts 2010), native vegetation restoration efforts in farming landscapes (Hajkowicz 2009), and the effects of removing dams on aquatic ecosystem integrity (Stanley and Doyle 2002; Orr et al. 2006; Brewitt and Holl 2014). Australia, for example, cannot determine whether its millions of dollars of biodiversity conservation programs are effective (Lindenmayer and Gibbons 2012; State of Environment 2017). An extensive review of conservation projects showed that only 5% incorporate question-driven monitoring (Muir 2010). In contrast, many of the monitoring programs with which we have been involved have been designed to quantify if management actions have been effective and the mechanisms that have (or have not) led to management success. These include whether weed control is effective (Lindenmayer et al. 2012b, 2015d; Sato et al. 2016), whether incentive schemes are restoring biodiversity in endangered woodland environments (Lindenmayer et al. 2012b; Sato et al. 2016) including in replanted areas (Lindenmayer et al. 2016), and if small mammal and bird species respond to the variable retention harvesting system as an alternative silvicultural system to clearcutting in wet forests (Lindenmayer et al. 2015a). The majority of these (and many other) monitoring programs have been underpinned by the general principles and thinking that we have outlined in this book.

long-term to be programs that run for at least 10 years. We are fully cognisant of the fact there is no such thing as a perfect monitoring program. The deficiencies in our own work (see Chapter 4) bear testament to this. However, it is clear that we must continue to improve on the poor record of ecological monitoring to date. We must make it more efficient, more effective and more successful. Only then will governments, private foundations and the general public be willing to support the significant increases in funding urgently needed for environmental monitoring. The adaptive monitoring paradigm we present is a logical framework for drawing together many of the attributes needed to improve all kinds of monitoring, but particularly mandated monitoring and question-driven monitoring programs (see Chapter 3).

5 – The upshot – our general conclusions

165

Changes in culture needed to facilitate monitoring As important as the adaptive monitoring framework is, many other factors outside this framework conspire to prevent monitoring programs from being successful and new ones from being instigated. They also conspire to make many monitoring programs ineffective. We outline these in the following sections, and argue that the culture of science, and indeed many aspects of society per se, prevent monitoring programs being established and maintained. These problems must be tackled if improvements in the understanding of the array of environmental problems facing humanity were to be achieved.

The academic culture and rewards systems We are acutely aware that some researchers do not have the right psyche to conduct long-term monitoring programs, for instance, because they get bored or because long-term studies have opportunity costs that preclude researchers from doing other things. Also, there is a deeply ingrained scientific culture and associated reward system that creates substantial disincentives to undertake long-term ecological monitoring. Bias against longer papers that are often required to present long-term studies ‘A researcher who reveals something exciting is more likely to get a highprofile paper (and a permanent position) than someone who spends years providing solid evidence for something that everyone in the field expected to be true.’ (Editor of Nature, 2012, Volume 467, p. 406) There is a widespread and increasing emphasis on ‘newsy’, short or provocative articles in scientific journals (e.g. through page-length, citation constraints and page charges). This emphasis favours multiple, brief publications over longer works of ecological synthesis. It penalises long-term studies, which are often best reported as major bodies of work drawing together many parallel themes of research. Long-term monitoring programs are also difficult for scientists in the early stages of their careers because they may not generate publications rapidly enough for them to build a track record of published work for promotion or other rewards. A second problem, related to the one outlined in the preceding paragraph, is the current emphasis from editors to reduce the number of citations in scholarly papers, presumably to save journal space. This limitation, combined with some journals now requiring citations from the journal where the paper is being published to increase the journal’s impact factor (Silver 2006), produces both a real and an ethical dilemma for scientists attempting to synthesise a large body of information based on long-term monitoring. It is interesting to note that

166

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Conservation Letters, which publishes generally short papers, has an impact factor approaching almost double that of Conservation Biology which often contains longer, more data-rich articles. Bibliometric emphasis on more (and shorter) papers The bibliometric emphasis in academic reward systems directs scientists to produce many short articles rather than fewer (longer) synthesis publications (the LPU or Least Publishable Unit issue, Likens 1998). This emphasis has led to an information ‘superglut’ which no one can read fully or carefully. It also has meant that people are not reading the literature or are overlooking important earlier scientific articles (Belovsky et al. 2004), ‘re-inventing wheels’ and purporting to be making discoveries but actually making re-discoveries (although not claimed as such). Even powerful searchable literature databases such as the Web of Science are limited in how far back in time they can go and they do not allow for scientific articles to be located when relevant results are not apparent from the title, abstract or keywords. In an interesting example of the potential limitations of keyword searches, one of us recently completed a major review of work on large old trees (Lindenmayer and Laurance 2017). Only 25% of the citations in that paper were identified from a systematic search of the literature, with the remainder coming from careful rereading of the large store of archived articles collected over the previous 30 years. Emphasis on scientific discoveries Current scientific culture favours discoveries and new work rather than maintaining ongoing work (Arnqvist 2013). As an example, the funding criterion for the National Science Foundation and the Australian Research Council makes this abundantly clear, although the Long-Term Research in Environmental Biology (LTREB) program is an exception in the United States. This requirement creates a distinct bias against maintaining existing and productive long-term studies. It is also a major disincentive for young scientists to take over pre-existing long-term projects. Emphasis on short-term funding There is an emphasis on short-term funding; grants and projects are typically for three years or less. However, it is surprising to us how many papers in the literature regard two to three years of study as long-term monitoring (e.g. Pavlik and Barbour 1988; Buck et al. 1996; Goldstein and Beyer 1999). Conversely, flat-funding on multiple-year grants like the LTREB is a prognosis for future failure, as most of the cost in long-term projects is for salaries. The problem of fads Ecology has long been plagued by fads and ill-conceived or poorly defined concepts (Redford et al. 2013). For example, in Australia there was a push to focus on threatened species, then on landscape management, and to some extent this has

5 – The upshot – our general conclusions

167

now reverted to a focus on threatened species. Of course, this fad-like approach is absurd given the need to have a combined species-based and landscape or ecosystem-based focus, especially as the two approaches are often complementary and much cross-fertilisation and enhanced understanding and management can be gained from doing both (Likens and Lindenmayer 2012). Embracing fads doesn’t help promote a culture of careful, question setting or foster a psyche of making prolonged sets of repeated measurements. Moreover, a culture has developed in recent years to teach and research (and publish) what is profitable in terms of academic advancement, not necessarily what is important (Gendron 2008; Shanley and López 2009). Lack of focus on temporal trends Many aspects of ecology also have been dominated by a spatial bias, with much less emphasis on temporal effects. When temporal factors have been explored, they often have been examined using space-for-time substitution, rather than true time series data. Bias against place-based research Long-term ecological research and monitoring is, by definition, focused on temporal changes in given entities at a particular location. Yet, in leading journals, there is a distinct trend away from place-based research. Ecological papers in magazines like Nature and Science are almost invariably decorated with a global map of some pattern or other and/or accompanied by a meta-analysis of some trend that underpins the supposed earth-wide patterns. The results of place-based long-term ecological research only very rarely find a place among the highly summarised short articles in such ‘tabloid’ magazines. Indeed, Ferreira et al. (2016) claim that this bias also leads to the de-funding and eventual demise of field-based investigations as well as the natural history skills that underpin such empirical studies. Some people argue that place-based, long-term monitoring is not relevant to issues of human concern. We strongly disagree. Place-based studies can act as sentinels for problems elsewhere, including national and global environmental problems. We suggest that place-based work can serve as models for application elsewhere to determine if particular ecological or environmental problems are indeed more widespread. The global collapse of frog and bee populations are good examples, in which declines detected in focused place-based work was later found to be replicated in many other parts of the world. The rise of the modeller There often has been a focus on modelling per se rather than empirical data collection and analysis, even though models need to be based on empirical data, and are tools and not an end in themselves (Canham et al. 2003; Todd et al. 2016). There also has been a rapid increase in the amount of ‘content-free’ simulation

168

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

modelling almost devoid of any empirical data. While this kind of modelling is an understandable approach because modelling is comparatively cheap, can generate results quickly, and can lead to numerous papers, we caution that ‘virtual reality can quickly descend into real stupidity’ (or fantasy). The tyranny of multi-tasking Scientists are increasingly pushed to work simultaneously across many different projects and many different disciplines, that is, multi-tasking (Forgasz and Leder 2006). We are chronic multi-taskers and don’t deny that many multi-taskers can be very productive, but many people currently are badly overcommitted (including ourselves). Multi-tasking can be quite distracting to most scientists, by prohibiting them from thinking deeply. Relative to the topic of this book, we believe that the multi-tasking culture is placing an emphasis on quantity rather than quality. The multi-tasking approach to science can lead to fragmentation of ideas, of funding, and of time (with a focus on deadlines). We understand that multi-tasking is a part of modern life, but we hope that the modern academic culture can also find a way to provide the opportunity for young scientists to focus on and think deeply about some long-term projects. The demise of the naturalist The passion, drive and ecological insights needed to maintain long-term ecological research programs are often derived from spending time in the field. However, there has been a distinct move away from the development of people with fieldbased skills, including natural history (Lu 2015). The demise of natural history research is leading to major knowledge gaps that will eventually undermine conservation and management efforts. For example, Lu (2015) laments the fact that while the genome is planned to be sequenced for the world’s 10 500 bird species, basic information such as clutch size remains unknown for half of these species. Many senior scientists do not spend time in the field to view first-hand and over time, the environmental changes that they are interested in and/or are working on. Similarly, many universities have moved away from field courses and field stations (e.g. Tydecks et al. 2016). They are expensive to run and can be hindered by occupational health and safety issues as well as liability risks. These changes contribute to a loss of natural history skills as well as natural history information (Noss 1996; Dijkstra 2016) and a lack of training in how to do the empirical component of long-term monitoring. Moreover, many field and research stations have been closed, are close to closing or dramatically altered in recent decades (Hildrew 2004; Tydecks et al. 2016). For example, Monks Wood Experimental Station in Cambridgeshire (England) was closed in 2009 with an ‘obituary’ published in Nature magazine and maintained online to this day (http://www. monks-wood.org.uk/).

5 – The upshot – our general conclusions

169

Finally, it is not at all clear to us if, in an era of remote data acquisition (sensors, cameras, satellites etc.), big data, data mining, and an almost anti-field work approach to environmental management and monitoring (Walters and Scholes 2017), these new ways of doing science will encourage new cohorts of young people to enter ecology and environmental science (McDowell 2015). To effectively integrate these new technologies into ecology, staff skillsets will have to include the ability to support large-scale, remote data generating exercises and data mining analyses of those remotely gathered data. It is somewhat perverse to us that at a time when members of the medical community are deeply concerned about many people’s disconnect with the natural environment (viz: the so-called nature deficit disorder (Louv 2005)), ecological and environmental sciences are swinging sharply away from an emphasis on people in the field! The problem of mobility and career enhancement Career development in science often promotes moving between institutions. There is some value in such moves, such as preventing ‘inbreeding depression’ resulting from a person spending their entire career at one place and with one group of people. It is also often the only way for scientists to gain promotion and increases in salary. However, we consider that frequent shifts between institutions make it more difficult to maintain continuity at a given site/location over a prolonged period. We also have great concerns about the significant (and often overwhelmingly negative) effects of the current science culture on young academics. The enormous competitive pressures to publish, attract funding, and win a secure position are driving young scientists into other occupations and create disincentives for others to enter science (Powell 2016). A related problem associated with mobility and career development in young scientists is that it makes it extremely difficult for senior scientists to plan for leadership succession and thereby find new people to effectively maintain longterm projects (Box 5.2). The rise of the science bureaucrat Concurrent with a rise in the modeller and a decline in the naturalist, there has been a meteoric rise in the science bureaucrat; so much of modern science appears to be hamstrung by a mountain of administrative over-burden. Scientists have always needed administrative support and structure (often/usually unfunded), but science is increasingly hard to do because of seemingly endless forms and associated checks and balances on every cent that is spent or justification of every minute of work time and allocation of that time to a particular grant. (It does seem that thinking is often not considered to be a valid use of time.) In summary, we strongly argue that there is a need to push back against some of the perverse aspects of existing science culture that impede the maintenance of

170

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box. 5.2. Leadership succession Planning for leadership succession is very important in the maintenance of longterm monitoring programs (see Chapter 2). However, it is very difficult to do (see Scudellari 2015) (and neither of us have done it well!). There are many reasons, including: (1) difficulty in putting out the ‘fire in the belly’ to keep a successful project running; (2) ego investment on the part of the chief investigator; (3) concern that a successor will do things differently and alter the trajectory of the work; and (4) not admitting that succession is needed (at least not now). Senior scientists responsible for often very valuable, long-term studies need to contemplate more carefully the need for succession planning and think about how to best achieve it – including better mentoring of young scientists in conducting long-term studies. As an example, the instigators of the 30-year Desert Ecology program in Central Australia (Dickman et al. 2014) have a well-developed plan to transfer management of the research to a non-government conservation body when they eventually retire (Chris Dickman, pers. comm.). Issues of leadership succession need far closer attention to ensure that the value of monitoring programs does not die along with their instigators. To the best of our knowledge, few if any universities or research institutes would make an academic appointment to maintain an already established long-term ecological research or monitoring program. For instance, the University of British Columbia did not take the opportunity to replace Charles Krebs with a person to maintain the exceptional work at Kluane in the Yukon (see Box 1.4 in Chapter 1). Thus, the problems of a lack of culture in succession planning in universities and research institutes may well warrant re-examination. We suggest that universities and other kinds of research institutions should be pushed to make academic appointments that maintain an already established long-term ecological research or monitoring programs. One of us (DBL) has been working for nearly a decade, with varying degrees of success, to cultivate the development of young post-doctoral researchers so that they can assume increasing responsibility for existing large-scale, long-term projects. However, the issues of finding permanent employment for these young researchers is a never-ending source of concern that has no straight-forward solution (and is indeed getting harder in academia per se). In other cases, succession planning has been well considered long in advance. In 2014, after 50 years, one of us (GEL) passed the long-term chemical analysis of precipitation and streamwater samples, which has been the core of the Hubbard Brook Ecosystem Study, to the US Forest Service, and in 2017 passed the supervision and maintenance of the collection and analysis of precipitation and streamwater samples to two younger scientists. The leaders of the networks of long-term plots within the Long-Term Ecological Research Network within the TERN initiative in Australia (see Chapter 4) suggest there are at least four key considerations in enhancing successful succession. These are (after Burns et al. 2014):

5 – The upshot – our general conclusions

171

•• Ensure that the successor knows the methodology of actually performing the field work, and managing and maintaining the database. •• Choose a successor who is well placed to inherit the responsibility for maintaining the research. That is, someone with a permanent position, or good prospects for securing such a position. •• Publish co-authored papers with the successor to increase the chances of future grant success. •• Ensure that the successor feels they are the new custodian of the monitoring site(s), with freedom to make strategic decisions about funding, management and strategy.

long-term studies. For example, there is a need to mount a concerted attack on editor obsession with novelty (Stephan et al. 2017) as well as more stridently make the case that time is a critical part of ecological dynamics (in place-based locations). There is also a need to recapture parts of the science publishing agenda and refocus it away from the current, and sadly increasing, emphasis on short, ‘newsy’ or provocative articles in scientific journals. There is far more to good ecological science than global syntheses and meta-analyses and a greater focus is needed on long-term studies that are often best reported as major bodies of work drawing together (synthesising) many parallel themes of research. An additional key part of pushing back against current science culture must be to halt the demise of naturalists and field-based ecologists; indeed, without people gathering data, modellers will eventually run out of content to simulate. Moreover, as part of making the case for naturalists and natural history, it will be crucial to continue to demonstrate that people are ‘critical infrastructure’ and essential for on-the-ground work; that is, effective long-term ecological science is much more than just equipment. Many critical discoveries could not have been made without people spending time in the field, which triggered key questions based on careful measurements and observations that are not possible with machines (even very sophisticated ones). For example, none of the discoveries about the evolution of beak morphology in finches in the Galapagos Islands (Grant and Grant 2002) nor the amazing revelation that the white-throated sparrow in northeastern America comprises four sexes (Tuttle et al. 2016; see Chapter 1) would have been made using only field equipment.

Structure of organisations The structure of many organisations is far from conducive for instigating and then maintaining long-term monitoring programs. Many organisations have high staff turnover and are characterised by a rapid loss of corporate culture. There is limited or no psyche of long-term studies or things allied with them, like fastidious data

172

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

and sample curation or long-term data management. In addition, there is usually an emphasis on reducing staff numbers in these organisations during budget crises, leading to a severe drop in morale. Yet, people on the ground are essential for successful monitoring of critical natural systems, which provide the fundamental life-support for humanity. Many organisations involved in environmental management have focused on the procurement of equipment rather than on retaining staff. However, as we discuss in Chapter 2, a focus on built infrastructure can create substantial problems, such as breaching the integrity of time series data, precluding studies that are guided by a robust experimental design, and creating problems for data analysis by generating vast datasets (see Berkelmans and Hendee 2002). Part of the focus on built infrastructure in some organisations (e.g. NEON in the United States) concerns the development of equipment to gather and transmit real-time data. An example is remotely stationed flux towers or temperature sensors that collect and transmit information at rapid intervals. We suggest these approaches can have inherent value but that field observations and measurements are vital to calibrate these instruments and to identify when these instruments produce incorrect or misleading information. Moreover, such kinds of instrumentation may create a disincentive for people to physically go out into the field and better understand how an ecosystem is functioning. Finally, a robust design built on questions needs to drive this type of data collection to avoid being overwhelmed by masses of unwanted or unneeded data. Many organisations contract out major projects (outsourcing), for example, to consulting companies. Other organisations, like the CSIRO in Australia, are continuing to move away from on-the-ground work. The business plan of these organisations means they are no longer capable of doing truly long-term monitoring because they are required to do other kinds of work such as consulting or short-term contracts. Moreover, many mandated monitoring programs change with the politics of the day, which can lead to termination of ongoing programs and false starts in others. Many scientists and resource managers are on short-term, non-continuing contracts. Such a lack of certainty makes it unlikely that they will want to set up long-term monitoring programs. The points we make above suggest to us that current organisational mandates, structures and cultures are often far from conducive to establishing and maintaining long-term monitoring programs. We believe that this culture must be changed in at least some organisations if high-quality, long-term monitoring were to succeed.

Funding Access to funding is an obvious factor influencing the success of monitoring programs. A major reduction in funding (or an abrupt cut to funding) can result in the collapse of long-term monitoring programs (e.g. Birkhead 2014). Such

5 – The upshot – our general conclusions

173

Box 5.3. ‘Close shaves’ Close shaves in maintaining long-term research and monitoring programs are common. Charles David Keeling had several where his CO2 monitoring program at Mauna Loa (see Box 3.5 in Chapter 3) came very close to being terminated. Apparently, Keeling had a reputation for being ‘grumpy’. We can particularly relate to that as it is very difficult and stressful to keep monitoring funded and operational, especially when there is a strong belief that data gaps are a major nemesis. Similar, close shaves have occurred on more than one occasion in the Hubbard Brook Ecosystem Study and in the Victorian Central Highlands Study (both outlined in Chapter 4), including as recently as mid-2015. Again, this problem is not uncommon and the challenge is to be nimble enough and innovative enough to find ways to meet these threats to maintaining the integrity of the long-term records. With hindsight, it is now clear how important it was that the Keeling curve not be terminated.

reductions were the basis for the axing of the Long-Term Ecological Research (LTER) network in Australia in 2017 (Lindenmayer et al. 2017), and many projects have encountered ‘close shaves’ where funding has almost been lost (Box 5.3). Many aspects of funding are not well suited to the establishment and maintenance of long-term monitoring programs. Indeed, monitoring programs are often seen as a luxury and not a core part of the remit of many resource management organisations. They are therefore usually the last initiatives to be funded (after all other, higher priority ones have been supported) and the first ones to be cut during budget shortfalls. In addition, budget cycles emphasise short-term projects with rapid achievement of milestones. Funding initiatives of one to three years are rarely congruent with the time frames appropriate for effective monitoring. Thus, there often is a fundamental mismatch between long-term environmental management aspirations and short-term financial realities. A limited number of monitoring programs now operate through the support of long-term endowments. Our case study in Chapter 4 on the 180+-year research program at Rothamsted in England provides one of the best known examples worldwide (Silvertown et al. 2006). Although such long-term funding is a welcome relief from the constant tribulation of maintaining the funding for a long-term project, we acknowledge that this approach needs to be applied carefully to ensure that guaranteed long-term funding does not create disincentives to be productive, innovative, and to add exciting new projects that contribute additional insights to an overall body of work – a little ‘hunger’ can be a good thing! Too much hunger is not – it leads to starvation. A trend away from evidence-based management and policy Best practice management and policy should be based on the best available evidence (Sutherland et al. 2004). Long-term research and monitoring are crucial to generating the body of evidence needed to underpin evidence-based management

174

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

and evidence-based policy (Lindenmayer et al. 2012a). Yet several recent policy decisions in the environment have blatantly ignored scientific evidence (Gleick et al. 2010; Lindenmayer 2013; Russell-Smith et al. 2015; Schiermeier 2016). Indeed, the Oxford Dictionaries named the term ‘post-truth’ as their word of the year in 2016 (see Higgins 2016). An unwillingness to engage in evidence-based management and evidence-based policy, including the current ‘war on science’ (Otto 2016) and the general loss of trust in science and scientists (Banks et al. 2012; Lubchenco 2017) erodes the appetite of governments and other organisations to maintain long-term research and monitoring to gather scientific evidence.

Societal culture Many of the traits needed to guide monitoring programs – endurance, tenacity, persistence and thoroughness – are not particularly common ones in western society. Rather, society has developed characteristics which are the antithesis of these: ●●

●●

●● ●●

An epidemic of busyness (Likens 1998) created by cell phones, email, Facebook, Twitter and other rapid communication technology and ease of travel to meetings worldwide. While these innovations can assist the scientific enterprise in many ways, this epidemic can restrict time for clear thinking, thoroughness and creativity, which are essential to frame incisive ecological questions and conduct effective monitoring. Nature deficit disorder or a disconnect between people in modern society and their natural environment (Louv 2005; Zaradic and Pergams 2007; Pergams and Zaradic 2008). This disconnect is allied with a loss of the natural history skills in the general community (Noss 1996) that are essential for good monitoring and long-term research (see Box 5.4). Politicisation of administrative decisions. Rapid turnover in employment.

In summary, it appears to us that a perverse outcome of the culture of western society, and parts of it like science culture and its reward system, is that this culture has inadvertently created disincentives to develop the monitoring systems to improve the management of ecosystems on which human kind depends. We are unsure how to best tackle major deep-seated issues like western and scientific cultures. Some corporations and other organisations have attempted to tackle problems associated with issues such as the epidemic of busyness through small steps like imposing email-free days and regular exercise programs. Far more substantial changes are required. Those with quite different skills to ours (e.g. Louv 2005), such as in psychology, sociology, public and science policy, and institutional arrangements, need to examine these problems and suggest ways to try to rectify them.

5 – The upshot – our general conclusions

175

Good things that can come from non-question based monitoring We have spent much of this book highlighting what we believe constitutes ineffective monitoring. We also have outlined an adaptive monitoring approach (Lindenmayer and Likens 2009; Lindenmayer et al. 2011) to gather many of the features of effective monitoring programs within a single coherent and logical framework. However, there are examples where long-term studies have produced valuable information, even when they have not been guided by carefully crafted scientific questions, where no formal conceptual model was used to identify those questions, and there was a noticeable absence of a statistical-based experimental design (Wintle et al. 2010). We highlight in Box 5.4 two outstanding examples of studies of this kind and we include them here to illustrate that we are not zealots in pushing our adaptive monitoring framework to the total exclusion of all other approaches. We readily acknowledge that many different approaches and perspectives are important in the evolution of improved long-term ecological research and monitoring programs. We do note that both case studies summarised in Box 5.4 are characterised by three features which are also common to effective monitoring programs (see Chapter 3): (1) they were maintained by a competent, dedicated and highly motivated individual; (2) record keeping was accurate and immaculate; and (3) they were driven by curiosity.

The role of citizen science in long-term monitoring Much has been written in recent times about the role that citizen science can play in assisting data collection in long-term studies, including in monitoring programs (Gardiner et al. 2012; McGowan et al. 2014; Jackson et al. 2015; see also Volume 10, issue 6 of Frontiers in Ecology and Evolution for a set of excellent case studies). We agree that citizen scientists have much to offer and there are many examples where data collected by them have led to profound new insights (e.g. Petrovan and Schmidt 2016). As one of numerous examples, citizen scientists have helped create inventories of the numbers and locations of large old trees, including exceptional individuals that require special protection. In the Romanian province of Transylvania, a citizen education project enlisted the assistance of school teachers and children in more than 20 villages to ‘find the oldest trees’ in wood pastures in southern Transylvania. The location, circumference and species of trees were recorded and a photograph taken. Information was then field checked by scientists (Moga et al. 2016). This example and others illustrate the power of citizen science to engage the broader community in conservation and pollution-protection programs, in generating datasets, and in the subsequent outcomes of considerable management and policy significance (McGowan et al. 2014; Kosmala et al. 2016). A key issue in the use of citizen science is that data collection needs to be guided by a well-designed and carefully framed scientific framework and

176

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 5.4. When good things come from a lack of formal, testable, scientifically driven questions Because of an intense desire to know more about the natural world, coupled with a strong discipline of recording natural history information in a personal journal, Aldo Leopold (Fig. 5.1a) accumulated long-term (13 years) data on weather patterns and phenological records of migratory bird arrivals and departures, plant flowering times, etc. for central Wisconsin in the United States. Leopold died in 1948, but these long-term observations were restarted by his daughter, Nina, and her husband, Charles Bradley, in 1976 and have been continued to the present (Ellwood et al. 2013). This rich record of natural history observations has provided significant insights into the effects of climate change on the timing of biological phenomena for this area (Bradley et al. 1999; Ellwood et al. 2013), and is available to researchers at the Aldo Leopold Legacy Center in Baraboo, Wisconsin and worldwide (www. aldoleopold.org). Similarly, Daniel Smiley (Fig. 5.1b) religiously recorded weather and a broad range of natural history events for more than 60 years at his family’s property in the New Paltz area of the Shawangunk Mountains of New York State in the United States. These data were collected carefully, but stored mostly on index cards in cardboard boxes. Because this information was potentially so important, one of us (GEL) strongly urged Mr Smiley in the late 1970s to organise these cards and collections and store them in a fire-proof facility. This suggestion was supported by others and this collection of data has now been computerised, and is secure and available to researchers as part of the Daniel Smiley Research Center at the Mohonk Preserve and worldwide (www.mohonkpreserve.org). Several peer-reviewed publications have used these daily observations of phenology, rainfall amount, air

(a)

(b)

Fig. 5.1.  (a) Aldo Leopold (Photo courtesy of the Aldo Leopold Foundation). (b) Daniel Smiley (right) with Paul Huth (Photo by Gerd Ludwig (1986) courtesy of Mohonk Preserve Daniel Smiley Research Center)

5 – The upshot – our general conclusions

177

temperature, ice in/ice out data for a small lake on the property, and acidity of surface waters (Cook et al. 2008). While both of these extraordinary men have died, their legacy lives on, and the long-term records they maintained are being continued and used by others. These two didn’t have a formal conceptual model (although they undoubtedly had a mental one), a statistical design for the activity, or formal question(s), but they had ample curiosity to guide these data collections. Yet it is clear that the rigor and tenacity of doing careful, long-term measurements paid off in providing important insights about how nature works and can change over time.

preferably overseen by professional scientists, otherwise data quality may be compromised (Kosmala et al. 2016). As an example, a small number of volunteer ornithologists were used in long-term studies of birds inhabiting reserves and military training areas on coastal south-eastern Australia. The statistical design of these large-scale comparative studies encompassed numerous sites surveyed at multiple times by different observers on different days with the field protocols implemented to reduce day and observer effects on the data. Differences in the ability of observers to detect particular species were quantified in a replicated experiment on a subset of sites. Importantly, observer heterogeneity was found not to have significant effects on broader inferences drawn from the work (Lindenmayer et al. 2009).

The challenge of intellectual property and data sharing Successful long-term research and monitoring programs result in the collection of high-quality empirical data and are characterised by high-quality data management. However, an important, but often unresolved issue is intellectual property and data sharing in the wider ecological and resource management communities. Only rarely have reasonable and ethical rules of engagement been developed for data sharing between those who gather long-term field data and others who desire access to those data for modelling, data mining and other analyses. Without appropriate attention to intellectual property issues at the beginning of a study, the development of better ways of data sharing will be impaired and the full potential of valuable empirical data for enhanced environmental management will not be realised. A policy regarding public data dissemination should be thought through and stated early in the process of setting up a study. Likewise, within teams of scientists doing long-term research and monitoring, there needs to be a clear, articulated understanding, especially at the beginning of the effort, about the role of team members and the use and attribution of data collected by individuals in the team. The substantial costs of data sharing are a related issue that also requires careful attention (Berman and Cerf 2013). At present, there is an expectation that those who

178

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

gather monitoring data also will meet the costs of curating those data to a level and standard that enables others to readily use them. However, in an era of rapidly shrinking budgets for field programs, many funding agencies simply do not make provision for the substantial logistical and financial costs that can be associated with data curation and data management, and the often considerable time required to service demands from others for data. This is a non-trivial issue as some ‘big datasets’ are actually riddled with errors that can lead to significant artefacts or even errors in analysis and interpretation (Maldonado et al. 2015). As outlined elsewhere in this chapter, it is actually far easier to be a modeller or data miner than a field-based empirical scientist and to avoid most of the associated demands and non-trivial costs of data curation and data sharing. Some argue that this disproportionate burden on field-based scientists has contributed to the decline in funding able to be won and maintained by non-modellers, with corresponding impacts on with the demise of natural history skills (Ferreira et al. 2016). Despite the intellectual property issues associated with making high-quality data readily accessible, it is clear that open access for datasets is increasingly required (including as part of journal paper publication and funding agencies). Many articles highlight the rise in open access publications and highlight the considerable benefits that can accrue from data sharing (Pullin and Salafsky 2010; Joppa et al. 2016; McNutt et al. 2016); however, open access can have its drawbacks (Box 5.5). Legal, funding, cultural and other issues must continue to be overcome (Huang et al. 2012; Litton 2017) and the benefits of data sharing must be made more transparent, such as the fact that sharing of data can substantially increase science publication impact (Parsons and Minster 2010; Reichman et al. 2011) and limit losses of research data (Gonzalez and Peres-Neto 2015). One approach to doing this has been to develop metrics for recording citations of datasets that result in academic and other credit to those whom have expended the often considerable effort in gathering the datasets in the first place (McNutt et al. 2016). Another way is to strongly encourage authors of papers to publish their data in readily accessible and citable repositories (e.g. Ecological Archives, see Box 5.6).

The challenges in effective monitoring of rare, threatened and endangered species The primary focus of this book has been on gathering high-quality ecological data to answer key questions of management significance. However, there can be major challenges in gathering sufficient data for effective ecological monitoring and management of threatened species. Often threatened or endangered species are so uncommon that gathering high-quality data on them is extremely difficult; they also can become extremely difficult (and increasingly costly) to detect as they decline (Chades et al. 2008). Contrasts in management interventions between areas

5 – The upshot – our general conclusions

179

Box 5.5. Some things are better off not being open access Open access is not a panacea and can have a significant negative impact on some aspects of conservation. For example, publication of species locations and descriptions can promote poaching of wildlife populations. Poaching is a significant global issue and has substantial detrimental effects on a wide range of species (Phelps et al. 2016). As an example, there are an increasing number of cases of newly described species being targeted by the illegal wildlife trade (Auliya et al. 2016). Location data contained in scientific publications can lead to entire populations of some species being collected not long after taxonomic descriptions of them are published (Hughes 2017). Indeed, Hughes (2017) suggested that more than 20 species of newly described reptiles have been targeted in this way through the illegal wildlife and pet trade, potentially leading to extinction in the wild not long after taxonomic descriptions have been published. The Chinese tiger gecko (Fig. 5.2) is a case in point. We suggest that it is critical that some data are not made publicly available in open access and other digital platforms; locations of endangered species are a sobering example of such data (Lindenmayer and Scheele 2017).

Fig. 5.2.  Chinese tiger gecko, a species collected so heavily in the wild soon after its taxonomic description that it is now close to extinction (Lindenmayer and Scheele 2017). Notably, typing in the name of this species into a search engine autopopulates the search with offers to sell animals on the illegal pet trade market. (Photo by Carola Jucknies)

supporting these species also may not be possible, thereby potentially limiting the underlying design on long-term monitoring studies, and precluding traditionally recommended approaches like adaptive management. Indeed, extensive surveys of the literature suggest that there are almost no adaptive management studies of endangered species globally (Westgate et al. 2013). The difficulties in monitoring

180

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Box 5.6. Journals for publishing datasets Many new journals are emerging that publish datasets. One example is Ecological Archives, a new journal from the Ecological Society of America. The journal ‘publishes materials that are supplemental to articles that appear in the [various] ESA journals … as well as peer-reviewed data papers with abstracts published in the printed journals.’ Ecological Archives is published in digital, internet-accessible form and enables the publication of data papers. For instance, the long-term (50 year) biogeochemical results for precipitation and streamwater chemistry for the Hubbard Brook Ecosystem Study are published in Ecology (Ecological Archives) in a citable format (Likens 2017).

threatened species are highlighted by the paucity of such studies for some groups in particular jurisdictions (Petrovan and Schmidt 2016). For example, in Australia, many species of amphibians are rare and populations are declining rapidly, yet ~40% of the frog species on the continent are not monitored. In many cases, monitoring may need to be targeted specifically to detect a given endangered species as more general biodiversity surveys will not produce sufficient records of the species of conservation concern to facilitate robust statistical analyses and data interpretation. New technologies, such a remote cameras (Meek et al. 2014; Burton et al. 2015; Steenweg et al. 2017) and eDNA (Hänfling et al. 2016), may prove useful in the monitoring of endangered species, although their use will be most effective if their application is preceded by the steps reiterated many times throughout this book (i.e. development of good questions, robust experiment design, etc.). A critical addition to all monitoring programs for threatened and endangered species must be to have trigger points for action in which major changes in management are instigated when there are signs that populations of a target species of conservation concern are undergoing rapid decline (Lindenmayer et al. 2013). Other entities in addition to rare and endangered species can be profoundly difficult to monitor in the long term. Large old trees are an example, given that they are very long-lived and can be very slow growing, with lifespans well exceeding even by far the longest existing monitoring programs (such as the 175+-year program at Rothamstead, see Chapter 4). Similarly, efforts to monitor important phenomena, like ecosystem collapse, are a major challenge because there are currently no indicators and surrogates that can reliably predict such catastrophic changes, making it impossible to determine which entities are appropriate to target for monitoring (Sato and Lindenmayer 2017).

The major challenge of keeping monitoring and long-term studies going A question that both authors have often been asked is: What can be done to keep a long-term study going? The features that characterise good and effective

5 – The upshot – our general conclusions

181

monitoring programs such as good questions, good design, high-quality data management, management relevance, and high levels of scientific productivity, can be important for meeting the challenge of maintaining such programs. However, there are other factors which, based on our collective experience, can help with the continuity of long-term studies. These include continuing to pose new and important questions of management relevance, reacting in a rapid and timely manner to new opportunities such as those associated with major natural disturbances, fostering partnerships with key stakeholders, involving the community in the research (e.g. through scientifically-directed citizen science), engaging in the political process to generate support for long-term work, planning funding bids well in advance (including to non-traditional sources such as philanthropists), and using a range of forms of communication to demonstrate the value of the work to a wide audience (and therefore well beyond the scientific community). Another way to improve the chances of maintaining a monitoring program is to demonstrate that the work leads to increased capacity for prediction and hence to provide policy makers and resource managers with an improved capacity to better and more proactively do their jobs. This is particularly powerful when monitoring can demonstrate a ‘good news story’ – a relative rarity in environmental management but something that can be a welcome respite to an otherwise seemingly endless diet of bad news (Garnett and Lindenmayer 2011; Garnett et al. 2017). Despite embracing some or all of these activities, funding to maintain long-term studies can nevertheless be erratic and sometimes punctuated, and it is important for the leaders of projects to have plans that enable projects to be mothballed and then readily restarted when funds become available once again. Successful mothballing strategies will necessarily include careful and readily accessible documentation of field site locations, field protocols, locations of archived material, locations of datasets and descriptions of the variables they contain. Sometimes it will be appropriate to completely terminate a monitoring program (see Box 5.7).

The big issue of integrating different kinds of monitoring The material we have presented throughout this book, together with the series of case studies in Chapter 4, clearly show that there are several kinds of long-term monitoring programs. For the general purposes of this book, we have crudely assigned them into three broad categories: question-driven monitoring, mandated monitoring, and curiosity-driven or passive monitoring. Different kinds of studies within these broad categories are conducted in different ways and at different scales. This diversity is fundamental to our adaptive monitoring framework, which is clearly not a one-sizefits-all approach (see Fig. 5.3), but rather emphasises a need to focus on well-crafted questions resulting in a study design, set of attributes and an implementation approach that will be different in each monitoring program (see Chapter 3).

182

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Conceptual model of how system works (assumed) Key questions to be posed Experimental survey or survey design Infrastructure or methods needed to gather data Data collection Data analysis and interpretation Answer questions, complete synthesis and publish/report results

Fig. 5.3.  Sequence of steps underpinning a successful monitoring program

Yet, politicians and high-level policy makers want some kind of world-, stateor national-level reflection of environmental performance. A large and growing number of regional, state/provincial, national and international organisations are producing State of the Environment reports (see www.cnie.org for a good, but partial listing of the sources for these reports). Examples include the highly visible Worldwatch Institute’s annual State of the World reports; The State of The Nation’s Ecosystems for the United States (The Heinz Center 2008); and the five-yearly Australian State of Environment reports (State of Environment 2017). Countless

Box 5.7. When to stop a long-term monitoring program When should a long-term monitoring program be terminated? Some long-term monitoring programs need to be stopped because they are ineffective. Others that are currently ineffective might be made more effective as a result of key changes – in questions, design, protocols or other factors – thereby becoming examples of adaptive monitoring (see Chapter 3). In yet other cases, a previously effective monitoring program might be terminated when there are no further important questions that need to be answered or when the value of the information gathered from a monitoring program is of limited value (Maxwell et al. 2015). Other issues to consider are the opportunity costs of maintaining a given monitoring program relative to other important environmental management activities that could be undertaken (McDonald-Madden et al. 2010). We emphasise that it is not always possible to predict when the data from a pre-existing monitoring program might become important again in the future or if there may be a need to recommence a previously terminated project. On this basis, it is important that field data (and the protocols used to gather those data) are stored in secure places where they can be readily accessed by others.

5 – The upshot – our general conclusions

183

United Nations documents on the state of the planet’s natural resources and its environment are also produced. These mandated monitoring programs will often produce coarse summaries of temporal changes in resource condition (e.g. status reports), but provide limited understanding about the site-specific mechanisms that have given rise to those changes. The large spatial scale of this information also may not be particularly useful for guiding targeted on-the-ground management interventions to improve environmental conditions in a given location. Conversely, question-driven, long-term monitoring programs will often be operating at the level of sites, landscapes or regions. When they are based on well-defined and scientifically tested questions (as well as many other features) they can provide important long-term environmental data on emerging environmental problems, such as the extreme dilution of precipitation and stream water occurring at the Hubbard Brook Experimental Forest (Likens and Buso 2012), as well as insights about the mechanisms or ecological processes giving rise to these emergent patterns. Such programs, in turn, can be valuable for informing resource management. However, spatial generalisation from question-driven, site-based, long-term monitoring programs is difficult because the results from such studies may not extrapolate well to other regions, states or to the national level (Scholes 2017) (but see Steenweg et al. 2017). That is, these are challenging spatial scale issues making it difficult to produce national perspectives on environmental conditions by integrating across such kinds of different site-level, landscape-level, or region-level studies (Lindenmayer et al. 2015c). Thus, these kinds of detailed programs are usually one-of-a-kind projects that do not produce data at the scale many governments want or need. For example, even if it were possible to roll out the excellent ‘report card’ system for Moreton Bay in South East Queensland, Australia (see Chapter 4) to the rest of Australia’s many estuaries, we are not aware of a scientifically defensible method that would enable a single, nationally aggregated estuary assessment. Therefore, there is an inherent tension between state- and national-level mandated monitoring and site- and region-based monitoring, as well as between academic- and organisation-based monitoring programs. This tension occurs because these are often quite disparate programs. For example, we were stunned to note that no data from the some two dozen LTER sites, costing US taxpayers ~US$23 million per year, were explicitly included in the US State of the Nation’s Ecosystems report done in 2008 (The Heinz Center 2008). The tension between mandated monitoring programs and site- and regionbased monitoring programs is reminiscent of tensions between top-down and bottom-up approaches in ecological thinking and research. As in the case of taking advantage of the different methods of study in ecology to facilitate progress, there also must be ways to capitalise on synergies between national- and state-based mandated monitoring and question-driven monitoring.

184

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Many authors have argued that ecology is a case-study discipline (e.g. ShraderFrechette and McCoy 1993) and it is clear that as outstanding as some particular long-term programs have been, such as those at Rothamsted (England) or the Experimental Lakes Area in Ontario (Canada), they are one-offs that are virtually impossible to replicate elsewhere. However, we believe that these kinds of projects serve as ‘in depth, reference models’ often identifying major environmental problems (e.g. atmospheric CO2 increase at Mauna Loa, acid rain at the Hubbard Brook Experimental Forest) and showing what key environmental changes or important ecological processes are occurring in other places. These kinds of long-term research and monitoring programs have led to major discoveries and resulted in very large problems in ecology and environmental management being identified and/or addressed. For example, while it is very difficult to extend the specific results of the Hubbard Brook Ecosystem Study beyond the Hubbard Brook Valley, the findings from the long-term research and monitoring at the Hubbard Brook Experimental Forest clearly suggested that acid rain and its effects were far more extensive throughout North America (Likens and Bormann 1974; Likens 2010; Holmes and Likens 2016). The long-term data from the Hubbard Brook Ecosystem Study underpinned the passage of the 1990 Clean Air Act Amendments in the United States, primarily because the results were so long and of such unassailable quality (see Chapter 4). Moreover, the research of the Hubbard Brook Ecosystem Study in New Hampshire and the Experimental Lakes Area in Ontario helped to catalyse the commencement of national networks for precipitation chemistry in the United States and Canada. The important lesson then is that leaders of site-based monitoring programs should think about the broader (regional, state, province or national) implications of their findings and discoveries, including the implications of site-based work as models for larger-scaled, mandated monitoring programs (Steenweg et al. 2017). We also suggest that the fundamental characteristics of some of the best examples of question-driven monitoring programs (well-defined questions, well-articulated conceptual models, rigorous experimental designs) are features that should be much more widely embraced as part of efforts to improve mandated monitoring programs. Hence, the adaptive monitoring framework we describe in Chapter 3 and the iterative steps that comprise the framework (Fig. 3.7), are just as relevant to mandated monitoring programs as they are to question-driven monitoring programs. We argue that any kind of monitoring program will be effective only when it is based on well-defined questions, makes use of a wellformulated conceptual model, and is guided by a carefully crafted experimental design. This formula may sound trite, but in the process of writing this book, we have been shocked at how often these seemingly simple ingredients are missing from a very large number of monitoring programs. The rapidly increasing areas of disturbed ecosystems (such as forests worldwide; see Seidl et al. 2014) highlight the need for more and better monitoring to guide critical and urgent management decisions to protect vegetation cover,

5 – The upshot – our general conclusions

185

water supplies, soils, and biodiversity. The US National Science Foundation had a small program for immediate response and funding following such major, natural disturbances (Special Grants for Ecological Research). Such programs are extremely important because of the opportunistic nature of quick response following major disasters. Sadly, this program has not been continued. We would argue that all countries should be prepared in this way to gather key ecological information via research and monitoring, to guide managers following major natural disturbances (Lindenmayer et al. 2009).

Approaches to integrate data from different kinds of monitoring There are advantages and disadvantages of question-led monitoring, large-scale mandated monitoring and passive (or surveillance) monitoring (see Chapter 1 and Chapter 4). Two of the key (and strongly interrelated) challenges are: (1) What is the optimal balance between the two broad kinds of monitoring? We suggest the most effective approach to spending limited resources will be for surveillance monitoring programs to form a small component of the overall expenditure on long-term monitoring, with more cost-effective, questiondriven programs being the primary kind of monitoring being undertaken. This tension has to some extent been apparent in attempts to determine levels of investment in research infrastructure versus investments in research, as seen in ongoing debates around observation networks such as NEON and related initiatives (Mervis 2015; Witze 2015; Cesare 2016). (2) Are there approaches to integrate data and insights from different kinds of monitoring in ways that are more seamless and more useful for environmental management (e.g. see Box 5.8)? In the remainder of this section, we describe some of the ways the challenge of integrating different kinds of monitoring might be met. These are key questions given the trends towards, and extensive increase in funding and infrastructure in, dedicated earth observation networks (Walters and Scholes 2017) and the impacts that such initiatives have on funding for questiondriven research and monitoring.

The challenges posed by differences in the kinds of entities that are monitored in different ecosystems One way to coordinate across different monitoring programs would be to adopt a uniform set of protocols for data collection and subsequently to facilitate data compatibility. This is a ‘one-size-fits-all’ approach and we outlined the suite of problems generic monitoring frameworks can create in Chapter 2. Enforcing the measurement of a common set of variables across sites could result in a ‘race to the bottom’ (Lindenmayer et al. 2015c). That is, it may lead to a very crude set of common measurements of limited value for environmental management. Many

186

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

measurements in, for example, long-leaf pine ecosystems in the south-eastern United States, which are strongly fire-dependent, may be irrelevant to the deserts of Antarctica or tropical rainforests of Puerto Rico. An additional problem is that it remains unclear how to combine sensibly the metrics for environmental conditions that have been estimated or calculated in different systems (Lindenmayer et al. 2015b). There is a long, but largely unsuccessful, history of this kind of problem in other topics in ecology. Examples include the difficulties of combining sub-indices for habitat attributes in constructing Habitat Suitability Index models (reviewed by Van Horne and Wiens 1991) and combining landscape indices into meaningful measures of landscape cover (Cale and Hobbs 1994; Stirnemann et al. 2015). It seems to us that combining outcomes from quite different monitoring programs is a broadly similar kind of problem, albeit at large spatial scales and over longer time frames. As an example, poor environmental performance in one place added to a good performance in another could be summed to provide an average overall result and to suggest that the environmental conditions are reasonable, but this result would obscure the fact that environmental conditions are degrading in some places. The problem of data incompatibility from different studies has confronted the Long-Term Ecological Research Network (LTERN) in Australia which is part of the Terrestrial Ecosystem Research Network (TERN). One solution to the problem is not to try to solve the problem directly by seeking to find a subset of common measurements nor a set of homogenised metrics, but rather to recognise that different critical entities need to be measured in different ecosystems, allowing robust reporting on temporal change in those important entities in any given ecosystem. Indeed, this type of monitoring and reporting is precisely what has happened in all of the (currently 12) networks of plots within the various long-term studies that comprise the now sadly defunct LTER network in Australia (see Chapter 4). In 2014, a very challenging synthesis project was completed in which a report of the temporal changes in major ecosystems around Australia, from tall eucalypt forests to deserts and from the wettest to the most arid parts of the continent was made (Lindenmayer et al. 2014). The report for each ecosystem was done in a way that, if possible, the report could be repeated at regular intervals to not only document temporal changes derived directly from the field measurements but also to identify and then describe the mechanisms giving rise to the documented patterns (Lindenmayer et al. 2014).

Using environmental and economic accounts as a way to demonstrate the value of monitoring and cement support for monitoring in place One way to demonstrate the value of monitoring is to ensure that the data generated by monitoring programs are needed for high-level decision making by policy makers and politicians. An example is the development of economic and environmental accounts which are underpinned by high-quality environmental

5 – The upshot – our general conclusions

187

Box 5.8. An example of integrated kinds of monitoring – the State of the Great Barrier Reef Reports The Great Barrier Reef is the largest living structure on earth. It suffers many environmental pressures including coral bleaching, fishing and nutrient run-off from agriculture in adjacent agricultural areas (http://www.aims.gov.au/ documents/30301/2107350/State+of+the+GBR.pdf/cf4b474b-32c4-4733-bf8e27457cd40835). A mandated State of the Great Barrier Reef monitoring program has been maintained for more than two decades by the Australian Institute of Marine Science as an independent data gatherer (http://www.gbrmpa.gov.au/managing-thereef/great-barrier-reef-outlook-report). In 2006, there was a demand to commence monitoring of no-take fishing zones to answer questions about the effectiveness of management interventions. On this basis, the monitoring program switched to a two-year alternating structure with State of the Reef monitoring in a given year and then monitoring of no-take fishing zones in the subsequent year. The monitoring work on the Great Barrier Reef is very large in scale and provides, as one of its outputs, major outlook reports such as those published in 2009 and 2014. Outlook reports are required under Great Barrier Reef Marine Park Act 1975. Monitoring has documented patterns of decline but also recovery of coral cover on many reefs throughout the Great Barrier Reef. It also has quantified the causal drivers of change including damage by predators such as the crown-of-thorns starfish, as well as the impacts of disturbances such as cyclones and coral bleaching (De’ath et al. 2012). The work on the Great Barrier Reef is an excellent example of the integration of mandated and question-driven monitoring that is not only documenting the major changes currently occurring in this globally iconic environment but also answering key questions associated with the causal mechanisms underpinning those changes.

datasets. They combine information from a range of sources to facilitate the calculation of credible metrics of environmental condition as well as enable the contribution of various natural assets such as water, carbon, timber and biodiversity to economic activity (Keith et al. 2016, 2017). Economic and environmental accounts is a well-developed and standardised framework for economic and environmental accounting that has been created by the United Nations (United Nations 2012) and is being trialled in more than 50 countries globally. There are proposals to create a set of economic and environmental accounts that are broadly similar to economic accounts, which are used to gauge economic trends (e.g. gross domestic product, employment numbers, budget surpluses and budget deficits) (Vardon et al. 2016).

Concluding remarks We suggest that many of the arguments in the scientific and resource management literatures about monitoring frameworks and frustrations about failed monitoring

188

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

programs stem from a failure to recognise the inherent values of, and differences between, large-scale mandated monitoring programs and smaller-scaled questiondriven monitoring programs. The next major challenge for long-term monitoring is to work out how to combine the datasets, results and outcomes that are generated from different kinds of monitoring at different scales to produce integrated assessments useful to decision-makers. There may well be methods that have been used to develop broadly based state-, province- or national-level economic indicators as well as track economic performance in different (often more localised) sectors of economies, which may be usefully applied in solving problems of aggregating datasets from environmental monitoring programs. The development of such approaches could help reflect environmental conditions at local scales as well as highlight overall environmental performance at larger (aggregated scales) (e.g. at a state, province or national level). However, we readily admit that we currently do not know how to do this accounting for environmental monitoring programs and associated datasets in ways that are scientifically defensible, of value for resource managers and useful to policy makers. We know that many readers will find our admission of ignorance to be unsatisfactory and frustrating. However, unless this challenge is resolved, our ability to deal with pressing problems such as rapid climate change and human-accelerated environmental change will be severely limited.

References Arnqvist G (2013) Editorial rejects? Novelty, schnovelty! Trends in Ecology & Evolution 28, 448–449. doi:10.1016/j.tree.2013.05.007 Auliya M, Altherr S, Ariano-Sanchez D, Baard EH, Brown C, Brown RM, Cantu J-C, Gentile G, Gildenhuys P, Henningheim E, Hintzmann J, Kanari K, Krvavac M, Lettink M, Lippert J, Luiselli L, Nilson G, Nguyen TQ, Nijam V, Parham JF, Pasachnik SA, Pedrono M, Rauhaus A, Cordova DR, Sanchez M-E, Schepp U, van Schingen M, Schneeweiss N, Segniagbeto GH, Somaweera R, Sy EY, Turkozoan O, Vinke S, Vinke T, Vyas R, Williamson S, Zeigler T (2016) Trade in live reptiles, its impact on wild populations, and the role of the European market. Biological Conservation 204, 103–119. doi:10.1016/j.biocon.2016.05.017 Banks P, Lunney D, Dickman C (Eds) (2012) Science Under Siege: Zoology Under Threat. Royal Zoological Society of New South Wales, Sydney. Belovsky GE, Botkin DB, Crowl TA, Cummins KW, Franklin JF, Hunter ML, Joern A, Lindenmayer DB, MacMahon JA, Margules CR, Scott JM (2004) Ten suggestions to strengthen the science of ecology. Bioscience 54, 345–351. doi:10.1641/00063568(2004)054[0345:TSTSTS]2.0.CO;2 Berkelmans R, Hendee JC (2002) Automatic weather stations: tools for managing and monitoring potential impacts to coral reefs. Marine Technology Society Journal 36, 29–38. doi:10.4031/002533202787914250

5 – The upshot – our general conclusions

189

Berman F, Cerf V (2013) Who will pay for public access to research data? Science 341, 616–617. doi:10.1126/science.1241625 Bernhardt ES, Likens GE, Hall J, Buso RO, Fisher DC, Burton SG, Meyer TM, Mc JL, Dowell WH, Mayer MS, Bowden WB, Findlay SE, Macneale KH, Stelzer RS, Lowe WH (2005) Can’t see the forest for the stream? In-stream processing and terrestrial nitrogen exports. Bioscience 55, 219–230. doi:10.1641/00063568(2005)055[0219:ACSTFF]2.0.CO;2 Birkhead T (2014) Stormy outlook for long-term ecology studies. Nature 514, 405. doi:10.1038/514405a Bradley N, Leopold AC, Ross J, Huffaker W (1999) Phenological changes reflect climate change in Wisconsin. Proceedings of the National Academy of Sciences of the United States of America 96, 9701–9704. doi:10.1073/pnas.96.17.9701 Brewitt P, Holl KD (2014) Better monitoring of fish in dam projects. Nature 513, 33. doi:10.1038/513033d Buck JA, Brewer LW, Hooper MJ, Cobb GP, Kendall RJ (1996) Monitoring Great Horned Owls for pesticide exposure in southcentral Iowa. The Journal of Wildlife Management 60, 321–331. doi:10.2307/3802231 Burns E, Lindenmayer D, Tennant P, Dickman C, Hanigan I, Hoffmann A, Keith D, Metcalfe D, Nolan K, Russell-Smith J, Wardle G, Welsh A, Williams R, Yates C, Green P (2014) Making monitoring successful: insights and lessons from the Long-term Ecological Research Network. Long-term Ecological Research Network, Canberra, Australia. Burton AC, Neilson E, Moreira D, Ladle A, Steenweg R, Fisher JT, Bayne E, Boutin S (2015) Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. Journal of Applied Ecology 52, 675–685. doi:10.1111/1365-2664.12432 Cale PG, Hobbs RJ (1994) Landscape heterogeneity indices: problems of scale and applicability, with particular reference to animal habitat description. Pacific Conservation Biology 1, 183–193. doi:10.1071/PC940183 Canham CD, Cole JJ, Lauenroth WK (Eds) (2003) Models in Ecosystem Science. Princeton University Press, Princeton, New Jersey. Cesare C (2016) Ecology aims high. Nature 529, 274–276. doi:10.1038/529274a Chades I, McDonald-Madden E, McCarthy MA, Wintle B, Linkie M, Possingham HP (2008) When to stop managing or surveying cryptic threatened species. Proceedings of the National Academy of Sciences of the United States of America 105, 13936– 13940. doi:10.1073/pnas.0805265105 Cook BI, Cook ER, Huth PC, Thompson JE, Forster A, Smiley D (2008) A cross-taxa phenological dataset from Mohonk Lake, NY and its relationship to climate. International Journal of Climatology 28, 1369–1383. doi:10.1002/joc.1629 De’ath G, Fabricius KE, Sweatman H, Puotinen M (2012) The 27-year decline of coral cover on the Great Barrier Reef and its causes. Proceedings of the National Academy

190

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

of Sciences of the United States of America 109, 17995–17999. doi:10.1073/ pnas.1208909109 Dickman C, Wardle G, Foulkes J, de Preu N (2014) Desert complex environments. In Biodiversity and Environmental Change: Monitoring, Challenges and Direction. (Eds DB Lindenmayer, E Burns, N Thurgate, A Lowe) pp. 379–438. CSIRO Publishing, Melbourne. Dijkstra K-D (2016) Restore our sense of species. Nature 533, 172–174. doi:10.1038/533172a Ellwood ER, Temple SA, Primack RB, Bradley NL, Davis CC (2013) Record-breaking early flowering in the eastern United States. PLoS One 8, e53788. doi:10.1371/ journal.pone.0053788 Ferreira C, Rios-Saldana CA, Delibes-Mateos M (2016) Hail global fieldwork, not just global models. Nature 534, 326. doi:10.1038/534326b Forgasz HJ, Leder GC (2006) Academic life: monitoring work patterns and daily activities. Australian Educational Researcher 33, 1–22. doi:10.1007/BF03246278 Gardiner MM, Allee LL, Brown PM, Losey JE, Roy HE, Smyth RR (2012) Lessons from lady beetles: accuracy of monitoring data from US and UK citizen-science programs. Frontiers in Ecology and the Environment 10, 471–476. doi:10.1890/110185 Garnett S, Lindenmayer DB (2011) Conservation science must engender hope to succeed. Trends in Ecology & Evolution 26, 59–60. doi:10.1016/j.tree.2010.11.009 Garnett S, Woinarski J, Lindenmayer DB, Latch P (Eds) (2017) Success in Conservation. CSIRO Publishing, Melbourne. Gendron Y (2008) Constituting the academic performer: the spectre of superficiality and stagnation in academia. European Accounting Review 17, 97–127. doi:10.1080/09638180701705973 Gleick PH, Adams RM, Amasino RM, Anders E, Anderson DJ, Anderson WW, Anselin LE, Arroyo MK, Asfaw B, Ayala FJ, Bax A, Bebbington AJ, Bell G, Bennett MVL, Bennetzen JL, Berenbaum MR, Berlin OB, Bjorkman PJ, Blackburn E, Blamont JE, Botchan MR, Boyer JS, Boyle EA, Branton D, Briggs SP, Briggs WR, Brill WJ, Britten RJ, Broecker WS, Brown JH, Brown PO, Brunger AT, Cairns J, Canfield DE, Carpenter SR, Carrington JC, Cashmore AR, Castilla JC, Cazenave A, Chapin FS, Ciechanover AJ, Clapham DE, Clark WC, Clayton RN, Coe MD, Conwell EM, Cowling EB, Cowling RM, Cox CS, Croteau RB, Crothers DM, Crutzen PJ, Daily GC, Dalrymple GB, Dangl JL, Darst SA, Davies DR, Davis MB, de Camilli PV, Dean C, Defries RS, Deisenhofer J, Delmer DP, Delong EF, Derosier DJ, Diener TO, Dirzo R, Dixon JE, Donoghue MJ, Doolittle RF, Dunne T, Ehrlich PR, Eisenstadt SN, Eisner T, Emanuel KA, Englander SW, Ernst WG, Falkowski PG, Feher G, Ferejohn JA, Fersht A, Fischer EH, Fischer R, Flannery KV, Frank J, Frey PA, Fridovich I, Frieden C, Futuyma DJ, Gardner WR, Garrett CJR, Gilbert W, Goldberg RB, Goodenough WH, Goodman CS, Goodman M, Greengard P, Hake S, Hammel G, Hanson S, et al. (2010) Climate change and the integrity of science. Science 328, 689–690. doi:10.1126/science.328.5979.689

5 – The upshot – our general conclusions

191

Goldstein A, Beyer J (1999) Monitoring and assessment of Swainson’s hawks in Argentina following restrictions on monocrotophos use. Ecotoxicology (London, England) 8, 215–224. doi:10.1023/A:1026448415467 Gonzalez A, Peres-Neto PR (2015) Act to staunch loss of research data. Nature 520, 436. doi:10.1038/520436c Grant PR, Grant BR (2002) Unpredictable evolution in a 30-year study of Darwin’s finches. Science 296, 707–711. doi:10.1126/science.1070315 Hajkowicz S (2009) The evolution of Australia’s natural resource management programs: towards improved targeting and evaluation of investments. Land Use Policy 26, 471–478. doi:10.1016/j.landusepol.2008.06.004 Hänfling B, Handley LL, Read DS, Hahn C, Li J, Nichols P, Blackman RC, Oliver A, Winfield IJ (2016) Environmental DNA metabarcoding of lake fish communities reflects long-term data from established survey methods. Molecular Ecology 25, 3101–3119. doi:10.1111/mec.13660 Higgins K (2016) Post-truth: a guide for the perplexed. Nature 540, 9. doi:10.1038/540009a Hildrew A (2004) FBA News 25. Holmes RT, Likens GE (2016) Hubbard Brook: The Story of a Forest Ecosystem. Yale University Press, New Haven, Connecticut. Huang X, Hawkins BA, Lei F, Miller G, Favret C, Zhang R, Qiao G (2012) Willing or unwilling to share primary biodiversity data: results and implications of an international survey. Conservation Letters 5, 399–406. doi:10.1111/j.1755-263X.2012.00259.x Hughes AC (2017) Trading in extinction: how the pet trade is killing off many animal species. The Conversation 7 February. Hughes BB, Beas-Luna R, Barner AK, Brewitt K, Brumbaugh DR, Cerny-Chipman EB, Close SL, Coblentz KE, de Nesnear KL, Drobnitch ST, Figurski JD, Focht B, Friedman M, Freiwald J, Heady KK, Heady WN, Hettinger A, Johnson A, Karr KA, Mahony B, Moritsch M, Osterback AK, Reimer J, Obinson J, Rohrer T, Rose JM, Sabal M, Segui LM, Shen C, Sullivan J, Zuercher R, Raimondi PT, Menge BA, Grorud-Colvert K, Novakk M, Carr MH (2017) Long-term studies contribute disproportionately to ecology and policy. Bioscience 67, 271–281. doi:10.1093/biosci/biw185 Jackson MM, Gergel SE, Martin K (2015) Citizen science and field survey observations provide comparable results for mapping Vancouver Island White-tailed Ptarmigan (Lagopus leucura saxatilis) distributions. Biological Conservation 181, 162–172. doi:10.1016/j.biocon.2014.11.010 Joppa LN, O’Connoer B, Visconti P, Smith C, Geldman J, Hoffmann M, Watson JE, Butchart SH, Virah-Sawmy M, Halpern BS, Ahmed SE, Balmford A, Sutherland WJ, Harfoot M, Hilton-Taylor C, Foden W, di Minim E, Pagad S, Genovesi P, Hutton J, Burgess ND (2016) Filling in biodiversity threat gaps. Science 352, 416–418. doi:10.1126/science.aaf3565 Keith H, Vardon M, Stein J, Stein J, Lindenmayer DB (2016) Experimental Ecosystem Accounts for the Central Highlands of Victoria. The Australian National

192

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

University, Canberra. Available at http://fennerschool-associated.anu.edu.au/ documents/CLE/VCH_Accounts_Summary_FINAL_for_pdf_distribution.pdf. Keith H, Vardon M, Stein JA, Stein JS, Lindenmayer DB (2017) Ecosystem accounts define explicit and spatial trade-offs for managing natural resources. Nature Ecology and Evolution doi:10.1038/s41559-017-0309-1 Kleijn D, Baquero RA, Clough Y, Diaz M, De Estaban J, Fernandez F, Gabriel D, Herzog F, Holzchuh A, Johl R, Knop E, Kruess A, Marshall EJ, Steffan-Dewenter I, Tscharntke T, Verhulst J, West TM, Yela JL (2006) Mixed biodiversity benefits of agri-environment schemes in five European countries. Ecology Letters 9, 243–254. doi:10.1111/j.1461-0248.2005.00869.x Kosmala M, Wiggins A, Swanson A, Simmons B (2016) Assessing data quality in citizen science. Frontiers in Ecology and the Environment 14, 551–560. doi:10.1002/ fee.1436 Likens GE (1998) Limitations to intellectual progress in ecosystem science. In Successes, Limitations and Frontiers in Ecosystem Science. (Ed. M Pace) pp. 247–271. Springer-Verlag, New York. Likens GE (2010) The role of science in decision-making: does evidence-based science drive environmental policy? Frontiers in Ecology and the Environment 8, e1–e9. doi:10.1890/090132 Likens GE (2011) Limnological measures related to climate change in the Hubbard Brook Valley, USA. Inland Waters 1(2), 93–99. Likens GE (2017) Fifty years of continuous precipitation and streamwater chemistry data from the Hubbard Brook Ecosystem Study (1963–2013). Ecology 98, 2224. doi:10.1002/ecy.1894 Likens GE, Bormann FH (1974) Acid rain: a serious regional environmental problem. Science 184, 1176–1179. doi:10.1126/science.184.4142.1176 Likens GE, Buso DC (2012) Dilution and the elusive baseline. Environmental Science & Technology 46, 4382–4387. doi:10.1021/es3000189 Likens GE, Lindenmayer DB (2012) Integrating approaches leads to more effective conservation of biodiversity. Biodiversity and Conservation 21, 3323–3341. doi:10.1007/s10531-012-0364-5 Lindenmayer DB (2013) From biodiversity to bioperversity: from good environmental science to poor policy. Pacific Conservation Biology 19, 250–255. doi:10.1071/ PC130250 Lindenmayer DB, Gibbons P (Eds) (2012) Biodiversity Monitoring in Australia. CSIRO Publishing, Melbourne. Lindenmayer DB, Laurance W (2017) The ecology, distribution, conservation and management of large old trees. Biological Reviews of the Cambridge Philosophical Society 92, 1434–1458. doi:10.1111/brv.12290 Lindenmayer DB, Likens GE (2009) Adaptive monitoring: a new paradigm for longterm research and monitoring. Trends in Ecology & Evolution 24, 482–486. doi:10.1016/j.tree.2009.03.005

5 – The upshot – our general conclusions

193

Lindenmayer DB, Scheele B (2017) Do not publish: limit open access information on rare and endangered species. Science 356, 800–801. doi:10.1126/science.aan1362 Lindenmayer DB, Wood JT, MacGregor C (2009) Do observer differences in bird detection affect inferences from large-scale ecological studies? Emu 109, 100–106. doi:10.1071/MU08029 Lindenmayer DB, Likens GE, Haywood A, Meizis L (2011) Adaptive monitoring in the real world: proof-of-concept. Trends in Ecology & Evolution 26, 641–646. doi:10.1016/j.tree.2011.08.002 Lindenmayer DB, Likens GE, Andersen A, Bowman D, Bull CM, Burns E, Dickman CR, Hoffmann AA, Keith DA, Liddell MJ, Lowe AJ, Metcalfe DJ, Phinn SR, RussellSmith J, Thurgate N, Wardle GM (2012a) Value of long-term ecological studies. Austral Ecology 37, 745–757. doi:10.1111/j.1442-9993.2011.02351.x Lindenmayer DB, Wood J, Montague-Drake R, Michael D, Crane M, Okada S, MacGregor C, Gibbons P (2012b) Is biodiversity management effective? Crosssectional relationships between management, bird response and vegetation attributes in an Australian agri-environment scheme. Biological Conservation 152, 62–73. doi:10.1016/j.biocon.2012.02.026 Lindenmayer DB, Piggott MP, Wintle BA (2013) Counting the books while the library burns: why conservation monitoring programs need a plan for action. Frontiers in Ecology and the Environment 11, 549–555. doi:10.1890/120220 Lindenmayer DB, Burns E, Thurgate N, Lowe A (Eds) (2014) Biodiversity and Environmental Change: Monitoring, Challenges and Direction. CSIRO Publishing, Melbourne. Lindenmayer D, Wood J, McBurney L, Blair D, Banks SC (2015a) Single large versus several small: the SLOSS debate in the context of bird responses to a variable retention logging experiment. Forest Ecology and Management 339, 1–10. doi:10.1016/j.foreco.2014.11.027 Lindenmayer DB, Barton PS, Pierson J (Eds) (2015b) Indicators and Surrogates of Biodiversity and Environmental Change. CSIRO Publishing, Melbourne. Lindenmayer DB, Burns EL, Tennant P, Dickman CR, Green PT, Keith DA, Metcalfe DJ, Russell-Smith J, Wardle GM, Williams D, Bossard K, deLacey C, Hanigan I, Bull CM, Gillespie G, Hobbs RJ, Krebs CJ, Likens GE, Porter J, Vardon M (2015c) Contemplating the future: acting now on long-term monitoring to answer 2050’s questions. Austral Ecology 40, 213–224. doi:10.1111/aec.12207 Lindenmayer DB, Wood J, MacGregor C, Buckley Y, Dexter N, Fortescue M, Hobbs RJ, Catford J (2015d) A long-term experimental case study of the ecological effectiveness and cost effectiveness of invasive plant management in achieving conservation goals; Bitou Bush control in Booderee National Park in eastern Australia. PLoS One 10, e0128482. doi:10.1371/journal.pone.0128482 Lindenmayer DB, Lane PW, Barton PS, Crane M, Ikin K, Michael DR, Okada S (2016) Long-term bird colonization and turnover in restored woodlands. Biodiversity and Conservation 25, 1587–1603. doi:10.1007/s10531-016-1140-8

194

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Litton E (2017) We must urgently clarify data-sharing rules. Nature 541, 437. doi:10.1038/541437a Louv R (2005) Last Child in the Woods: Saving Our Children from Nature-Deficit Disorder. Algonquin Books of Chapel Hill, Chapel Hill, North Carolina. Lu X (2015) Hot genome leaves natural histories cold. Science 349, 1064. doi:10.1126/ science.349.6252.1064-a Lubchenco J (2017) Environmental science in a post-truth world. Frontiers in Ecology and the Environment 15, 3. doi:10.1002/fee.1454 Lindenmayer DB, Burns EL, Dickman C, Green PT, Hoffmann AA, Keith DA et al. (2017) Save Australia’s ecological research. Science 357, 557. Maldonado C, Molina CI, Zizka A, Persson C, Taylor CM, Alban J, Chilquillo E, Ronsted N, Antonelli A (2015) Estimating species diversity and distribution in the ear of Big Data: to what extent can we trust public databases? Global Ecology and Biogeography 24, 973–984. doi:10.1111/geb.12326 Maxwell SL, Rhodes JR, Runge MC, Possingham HP, Ng C, McDonald-Madden E (2015) How much is new information worth? Evaluating the financial benefit of resolving management uncertainty. Journal of Applied Ecology 52, 12–20. doi:10.1111/1365-2664.12373 McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, Possingham HP (2010) Monitoring does not always count. Trends in Ecology & Evolution 25, 547–550. doi:10.1016/j.tree.2010.07.002 McDowell WH (2015) NEON and STREON: opportunities and challenges for the aquatic sciences. Freshwater Science 34, 386–391. doi:10.1086/679489 McGowan KA, Westley F, Fraser EDG, Loring PA, Weathers KC, Avelino F, Sendzimir J, Chowdhury RR, Moore ML (2014) The research journey: travels across the idiomatic and axiomatic toward a better understanding of complexity. Ecology and Society 19, 37–49. doi:10.5751/ES-06518-190337 McNutt M, Lehnert K, Hanson B, Nosek BA, Ellison AM, King JL (2016) Liberating field science samples and data. Science 351, 1024–1026. doi:10.1126/science. aad7048 Meek P, Fleming P, Ballard G, Banks P, Claridge A, Sanderson J, Swann D (Eds) (2014) Camera Trapping: Wildlife Management and Research. CSIRO Publishing, Melbourne. Mervis J (2015) NSF looks for new contractor to finish troubled observatory. Science 350, 1454. doi:10.1126/science.350.6267.1454 Moga CI, Samoila C, Ollerer K, Bancila RI, Reti K-O, Craiveanu C, Poszet S, Rakosy L, Hartel T (2016) Environmental determinants of the old oaks in wood-pastures from a changing traditional social-ecological system of Romania. Ambio 45, 480–489. doi:10.1007/s13280-015-0758-1 Muir MJ (2010) Are we measuring conservation effectiveness? Report to Conservation Measures Partnership, www.conservationmeasures.org. Noss R (1996) The naturalists are dying off. Conservation Biology 10, 1–3. doi:10.1046/j.1523-1739.1996.10010001.x

5 – The upshot – our general conclusions

195

Orr H, Rogers KL, Stanley EH (2006) Channel morphology and P uptake following removal of a small dam. Journal of the North American Benthological Society 25, 556–568. doi:10.1899/0887-3593(2006)25[556:CMAPUF]2.0.CO;2 Otto S (2016) The War on Science: Who’s Waging It, Why It Matters, What We Can Do About It. Milkweed Editions, Minneapolis, Minnesota. Pannell DJ, Roberts AM (2010) Australia’s National Action Plan for Salinity and Water Quality: a retrospective assessment. The Australian Journal of Agricultural and Resource Economics 54, 437–456. doi:10.1111/j.1467-8489.2010.00504.x Parsons MA, Minster J-B (2010) Data citation and peer review. Eos (Washington, D.C.) 91, 297–298. doi:10.1029/2010EO340001 Pavlik BM, Barbour MG (1988) Demographic monitoring of endemic sand dune plants, Eureka Valley, California. Biological Conservation 46, 217–242. doi:10.1016/0006-3207(88)90069-9 Pergams OR, Zaradic PA (2008) Evidence for a fundamental and pervasive shift away from nature-based recreation. Proceedings of the National Academy of Sciences of the United States of America 105, 2295–2300. doi:10.1073/pnas.0709893105 Petrovan SO, Schmidt BR (2016) Volunteer conservation action data reveals large-scale and long-term negative population trends of a widespread amphibian, the Common Toad (Bufo bufo). PLoS One 11, e0161943. doi:10.1371/journal.pone.0161943 Phelps J, Biggs D, Webb EL (2016) Tools and terms for understanding illegal wildlife trade. Frontiers in Ecology and the Environment 14, 479–489. doi:10.1002/fee.1325 Powell K (2016) Young, talented and fed-up. Nature 538, 446–449. doi:10.1038/538446a Pullin AS, Salafsky N (2010) Save the whales? Save the rainforest? Save the data! Conservation Biology 24, 915–917. doi:10.1111/j.1523-1739.2010.01537.x Redford K, Padoch C, Sunderland T (2013) Fads, funding, and forgetting in three decades of conservation. Conservation Biology 27, 437–438. doi:10.1111/cobi.12071 Reichman OJ, Jones MB, Schildhauer MP (2011) Challenges and opportunities of open data in ecology. Science 331, 703–705. doi:10.1126/science.1197962 Russell-Smith J, Lindenmayer DB, Kubiszewski I, Green P, Costanza R, Campbell A (2015) Moving beyond evidence-free environmental policy. Frontiers in Ecology and the Environment 13, 441–448. doi:10.1890/150019 Sato C, Lindenmayer DB (2017) Meeting the global ecosystem collapse challenge. Conservation Letters doi:10.1111/conl.12348. Sato C, Wood J, Stein JA, Crane M, Okada S, Michael D, Kay G, Florance D, Seddon J, Gibbons P, Lindenmayer DB (2016) Natural tree regeneration in agricultural landscapes: the implications of intensification. Agriculture, Ecosystems & Environment 230, 98–104. doi:10.1016/j.agee.2016.05.036 Schiermeier Q (2016) Pristine forests at risk. Nature 530, 393. doi:10.1038/530394a Scholes RJ (2017) Taking the mumbo out of the jumbo: progress towards a robust basis for ecological scaling. Ecosystems 20, 4–13. doi:10.1007/s10021-016-0047-2 Scudellari M (2015) A grand exit. Nature 521, 21–24.

196

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

Seidl R, Schelhaas M-J, Rammer W, Verkerk PJ (2014) Increasing forest disturbances in Europe and their impact on carbon storage. Nature Climate Change 4, 806–810. doi:10.1038/nclimate2318 Shanley P, López C (2009) Out of the loop: why research rarely reaches policy makers and the public and what can be done. Biotropica 41, 535–544. doi:10.1111/j.1744-7429.2009.00561.x Shrader-Frechette KS, McCoy ED (1993) Method in Ecology: Strategies for Conservation. Cambridge University Press, Cambridge. Silver S (2006) Gaming the impact factor. Frontiers in Ecology and the Environment 4, 283. doi:10.1890/1540-9295(2006)4[283:GTIF]2.0.CO;2 Silvertown J, Poulton P, Johnston E, Edwards G, Heard M, Biss PM (2006) The Park Grass Experiment 1856–2006: its contribution to ecology. Journal of Ecology 94, 801–814. doi:10.1111/j.1365-2745.2006.01145.x Stanley EH, Doyle MW (2002) A geomorphic perspective on nutrient retention following dam removal. Bioscience 52, 693–701. doi:10.1641/00063568(2002)052[0693:AGPONR]2.0.CO;2 State of Environment (2017) State of Environment Report. Department of Environment and Energy, Canberra, Australia. Steenweg R, Hebblewhite M, Kays R, Ahumnda J, Fisher JT, Burton C, Townsend SE, Carbone C, Rowcliffe JM, Whittington J, Brodie J, Royle JA, Switalski A, Clevenger AP, Heim N, Rich LN (2017) Scaling-up camera traps: monitoring the planet’s biodiversity with networks of remote sensors. Frontiers in Ecology and the Environment 15, 26–34. doi:10.1002/fee.1448 Steffen W, Richardson K, Rockström J, Cornell S, Fetzer I, Bennett EM, Biggs R, Carpenter S, de Vries W, De Wit CA, Folke C, Gerten D, Heinke J, Mace GM, Persson LM, Ramanathan V, Reyers B, Sorlin S (2015) Planetary boundaries: guiding human development on a changing planet. Science 347, 1259855. doi:10.1126/science.1259855 Stephan P, Veugelers R, Wang J (2017) Blinkered by bibliometrics. Nature 544, 411– 412. doi:10.1038/544411a Stirnemann I, Mortelliti A, Gibbons P, Lindenmayer DB (2015) Fine-scale habitat heterogeneity influences occupancy in terrestrial mammals in a temperate region of Australia. PLoS One 10, e0138681. doi:10.1371/journal.pone.0138681 Sutherland WJ, Pullin AS, Dolman PM, Knight TM (2004) The need for evidencebased conservation. Trends in Ecology & Evolution 19, 305–308. doi:10.1016/j. tree.2004.03.018 The Heinz Center (2008) The State of the Nation’s Ecosystems 2008. The H. John Heinz III Centre for Science, Economics and the Environment and Island Press, Washington, DC. Todd CR, Lindenmayer DB, Stamation K, Acevedo-Catteneo S, Smih S, Lumsden LF (2016) Assessing reserve effectiveness: application to a threatened species in a

5 – The upshot – our general conclusions

197

dynamic fire prone forest landscape. Ecological Modelling 338, 90–100. doi:10.1016/j.ecolmodel.2016.07.021 Tuttle EM, Bergland AO, Korody ML, Brewer MS, Newhouse DJ, Minx P, Stager M, Betuel A, Cheviron ZA, Warren WC, Gonser RA, Balakrishnan CN (2016) Divergence and functional degradation of a sex chromosome-like supergene. Current Biology 26, 344–350. doi:10.1016/j.cub.2015.11.069 Tydecks L, Bremerich V, Jentschke I, Likens GE, Tockner K (2016) Biological field stations – a global infrastructure for research, education, and public engagement. Bioscience 66, 164–171. United Nations (2012) System of Environmental-Economic Accounting Central Framework. United Nations, New York. Available at http://unstats.un.org/unsd/ envaccounting/pubs.asp Van Horne B, Wiens JA (1991) Forest bird habitat suitability models and the development of general habitat models. U.S. Fish and Wildlife Service, Washington, DC. Vardon M, Burnett P, Dovers S (2016) The accounting push and the policy pull: balancing environment and economic decisions. Ecological Economics 124, 145– 152. doi:10.1016/j.ecolecon.2016.01.021 Walters M, Scholes RJ (Eds) (2017) The GEO Handbook on Biodiversity Observation Networks. Springer Open, Switzerland. Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive management of biological systems: a review. Biological Conservation 158, 128–139. doi:10.1016/j. biocon.2012.08.016 Wintle BA, Runge MC, Bekessy SA (2010) Allocating monitoring effort in the face of unknown unknowns. Ecology Letters 13, 1325–1337. doi:10.1111/j.1461-0248.2010.01514.x Witze A (2015) US ocean sciences told to plot fresh course. Nature 517, 538–539. doi:10.1038/nature.2015.16780 Zaradic PA, Pergams OR (2007) Videophilia: implications for childhood development and conservation. The Journal of Developmental Processes 2, 130–144.

This page intentionally left blank

Index

ABMP see Alberta Biodiversity Monitoring Program academic culture and rewards system  165–71 bias against longer papers that are often required to present long-term studies 165–6 bias against place-based research  167 bibliometric emphasis on more (and shorter) papers  166 demise of the naturalist  168–9, 171 emphasis on scientific discoveries  166 emphasis on short-term funding  166 lack of focus on temporal trends  167 problem of fads  166–7 problem of mobility and career enhancement 169 rise of the modeller  167–8 rise of the science bureaucrat  169 tyranny of multi-tasking  168 acid rain changes, Hubbard Brook Experimental Forest  8, 184 extension of Hubbard Brook Experimental Forest results to other parts of the US  184 impacts on terrestrial and aquatic ecosystems, Hubbard Brook Valley 71 US studies  111 adaptive management paradigm  86, 119, 164 adaptive monitoring examples 83–4

as a general and not a prescriptive framework 85–6 hypothetical example  85 Rothamsted 117–18 adaptive monitoring framework  14–15, 17, 81–3, 175, 184 examples 83–4 importance of posing new questions  83 increasingly future role for  86 question setting  81–3 adaptive monitoring paradigm  17, 81, 83, 86, 89 adaptive surrogacy framework to tackle problems associated with deciding what to monitor  35, 36–7 Ailuropoda melanoleuca see giant panda air pollution, reduction, and improved human health  8 air quality indicators, lichen diversity, urban environments 84 Alberta Biodiversity Monitoring Program (ABMP)  104–9, 110 aims 104 complex aggregation issues  108–9 entities targeted for monitoring  106 lack of conceptual model  108 lack of stratification  107, 110 laundry list approach  108, 110 passivity  107–8, 110 paucity of questions  105–7 positive aspects  109 rejoinder: an alternative set of perspectives 110 statistical design  107

199

200

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

survey design  105 Aldo Leopold Legacy Center, Baraboo, Wisconsin 176 analytical methods, changes in, and data integrity  43–5, 78, 82, 128 arboreal marsupials, wet forests, Victoria distribution and abundance  130, 131–4 initial observational, retrospective investigation 139 monitoring program changes using a novel and non-standard overlapping and rotating statistical design  61–2, 68–9 population number relationship with abundance of trees with hollows 131–4 archiving field samples  79, 115, 117, 122, 143, 151 assumption that all monitoring programs can be the same  35–7, 185 atmospheric carbon dioxide, monitoring, Mauna Loa site  74–5, 184 Atmospheric Integrated Research Monitoring Network (AIRMoN)  60 atmospheric monitoring, Cape Grim, Tasmania 75–6 AusCover 140 AusPlots 140 Australian Research Council, funding criterion 166 backward approach to data collection  29, 38 ‘big data’ concept of  143 conversion into useful scientific knowledge and understanding  147 errors with  38, 42, 148 biogeochemical response of nitrogen to forest disturbance 11–12 biomass changes over time, Hubbard Brook Experimental Forest  127 bird records, migratory, central Wisconsin 176 bird species, effectiveness of restoration efforts for elements of farmland biodiversity 65 bird studies, long-term, and observer heterogeneity, coastal south-eastern Australia 177

bird surveys, Tumut, change in recording methods 44 boreal forests Kluane Boreal Forest Monitoring Project, Canada  9, 10–11, 170 moose impacts on ecosystem dynamics, Isle Royale  72–3 Bormann, F. Herbert  122, 123, 129 box-gum grassy woodlands, Victoria, alteration of monitoring program  45, 84 breaches of data integrity  43–5, 78 bureaucracy impact of long-term monitoring programs  48–9, 128–9, 171–2 NEON/TERN 144–6 calibration of field techniques  78 camera trapping  45 Canadian Air and Precipitation Monitoring Network (CAPMoN)  60 Cape Grim, Tasmania, atmospheric monitoring 76 carbon dioxide, monitoring program  74–5, 173 carbon sequestration  118, 127 career development  169 case studies authors’ trepidation about including  99– 101 effective  114–39, 150–1 impossibility to replicate elsewhere  184 problematic  101–14, 150 ugly  150, 151 ‘wait and see’  139–50 Central Highlands of Victoria, south-eastern Australia, wet montane ash eucalypt forest monitoring projects  9, 48, 61–2, 68–9, 129–39 Chinese tiger gecko  179 citizen science, role in long-term monitoring  175, 177 Clean Air Act and Amendments (US)  7–8, 71, 184 Clean Air Status and Trends Network (CASTNeT) 60 climate change insight into through natural history observations, Wisconsin  176 monitoring  85, 86, 101, 110, 118, 144, 148

I n d ex

co-located long-term ecological studies  12 conceptual models  3 Hubbard Brook Ecosystem Study  62–3, 121–2, 126 lack of  103, 108, 112–13, 115, 141 Warra Long-Term Ecological Research site, Tasmania  63 consulting companies, outsourcing of projects to  172 Cranberry Lake, New York, white-throated sparrow studies  12, 171 Crawford caper  151 cross-sectional studies  77 CSIRO  123, 172 cultural changes needed to facilitate monitoring 165 academic culture and rewards system 165–71 funding 172–3 societal culture  174 structure of organisations  171–2 curiosity-driven monitoring  1–2 Daniel Smiley Research Center, Mohonk Preserve 176 data approaches to integration of data from different kinds of monitoring  185–7 frequent examination and use of  77 data collection backward approach to  29, 39 uniform set of protocols for  185 data incompatibility from different studies 186 data integrity breaches of  43–5 maintenance of  78, 82, 128 data management, poor  42–3 data sharing costs of  177–8 and intellectual property  177, 178 data storage methods, evolution of  43 datasets combining from different types of monitoring programs  188 integration at spatial and temporal scales 148–9 journals for publishing  178, 180 metrics for recording citations of  178 open access for  178

201

see also long-term datasets Department of Innovation, Industry, Science and Research (Australia)  139 Desert Ecology Program, Central Australia 170 disengagement of key parties  40–1 dissolved silica, Lake Michigan, changes in time series data  43–4 disturbed ecosystems, need for more and better monitoring  184–5 Earth Observation Networks  139, 140 Ecological Archives  178, 180 ecological communities, relative abundance of species in  9 ecological genetic theory  12 ecological research, not all studies need to be long-term 14 ecological surprises, long-term studies leading to  12–13 ecological theory, long-term datasets to test 9–12 ecological values, and uses of long-term datasets 4–13 ecology, as case-study discipline  184 economic and environmental accounts, as a way to demonstrate the value of monitoring and cement support for monitoring 186–7 ecosystem collapse, monitoring  180 Ecosystem Health Monitoring Program (EHMP) for Moreton Bay, South East Queensland 118–21 ecosystem monitoring, challenges posed by differences in the kind of entities that are monitored in different ecosystems  185– 6 eDNA 180 effective case studies  114–39 features 150–1 effective leadership  70, 75 effective monitoring programs characteristics  59–78, 86, 89, 150–1 critical components  88–9 EMAP (Environmental Monitoring and Assessment Program)  109–14 aims 111 lack of conceptual model  112–13, 115 lack of focused questions  111–12, 115 leadership changes  111, 114

202

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

limited reporting  113 management relevance  113 partnerships 114 rejoinder 115 statistical design  113, 115 endangered species challenges in effective monitoring of 178–80 need for trigger points for action  180 suppression of locations of  179 environmental conditions, difficulty of combining metrics from different systems 186 environmental legislation, long-term datasets pivotal for  7–9 environmental management disturbed ecosystems  184–5 effectiveness of  164 integration of different kinds of monitoring data  185 reporting of ‘good news stories’  181 trend to blatantly ignore scientific evidence 174 Environmental Protection Agency (US), EMAP  109–14, 115 Environmental Stewardship Program (private farmers’ management of threatened woodlands for conservation outcomes)  41, 84 adaptive monitoring  84 need for matched ‘reference patches’  41– 2, 84 rotating sampling monitoring approach 84 see also box-gum grassy woodlands epidemic of busyness  174 evidence-based management and policy, trend away from  173–4 evolving questions  61–2, 82–3, 137, 181 expensive scientific equipment, focus on  37–9, 143–4, 172 experimental design good design  65–7 knock-on effects of poor design  31–2 poor design at beginning of the study  31 see also statistical design Experimental Lakes Area, Ontario  184 fad-like approach to research  166–7

failure to agree on what entities to monitor  32–5, 63–5 failure to ask the preliminary and fundamental question – is monitoring needed at all?  28–9 field, time in the  81, 168–9, 171 field courses, demise of  168 field presence  171, 172 Hubbard Brook Experimental Forest 128 field samples, archiving  79 field sites access to  80–1 infrastructure at  49, 81 field staff continuity of  80 quality of  138 field stations, demise of  168 field techniques, calibration of  78 field transport  78–80 fire experiments, in savannahs of northern Australia, lack of relevance to resource managers 40 fires, wet montane ash eucalypt forests, Central Highlands, Victoria, pre- and post-fire ecological studies  48, 135–6 Fisher, R.A.  116 flawed assumption that all monitoring programs can be the same  35–7 forest disturbance, biogeochemical response of nitrogen to  11–12 Forest Inventory Analysis (FIA)  60 frequent examination and use of data  77 frog species, Australia, lack of monitoring 180 funding ‘close shaves’  173 of expensive scientific equipment  37–9, 143–4, 172 favouring new work rather than maintaining ongoing work  166 Hubbard Brook Ecosystem Study  129, 173 Integrated Marine Observing System (IMOS) 149 lack of/loss of  46, 47 large-scale observation networks versus more focused, question-based empirical investigations  148

I n d ex

for long-term monitoring programs  16, 46, 47, 76–7, 172–3, 181 Rothamsted  117, 173 short-term  46, 166 wet montane ash eucalypt forest projects, Central Highlands, Victoria  138, 173 Galapagos finches  12, 171 giant panda, molecular methods enhanced population numbers  13, 45 Global Lake Ecological Observatory Network (GLEON) 140 global observation networks, interfacing with scales needed for ecological monitoring 147 good design  31, 65–7 challenges of  66–7 Rolls Royce or Ford?  41, 66 good question setting  60–1 Ecosystem Health Monitoring Program, Moreton Bay  120 Hubbard Brook Ecosystem Study  125 Rothamsted 116 wet montane ash eucalypt forests, Central Highlands, Victoria  131 Great Barrier Reef, integration of mandated and question-driven monitoring  187 Group on Earth Observations Biodiversity Observation Network (GEO BON)  148 Guillemot nesting, Stora Karlsö, Sweden, archived photographic images  79 Gymnobelideus leadbeateri see Leadbeater’s possum Habitat Suitability Index models  186 Hubbard Brook Ecosystem Study  121–9 atomic absorption spectrophotometer 123 basic approaches and objectives  122 conceptual model guiding biogeochemical research and monitoring  63, 64, 121–2, 126 deficiencies 128–9 demonstrated importance of long-term research 127 entities monitored  123–5 experimental watershed-ecosystem manipulations 125

203

founding members and leadership issues  122–3, 129 funding  129, 173 impacts of acid rain on terrestrial and aquatic ecosystems, tests of management relevance  71 importance of archiving samples  79 loss of cultural infrastructure  49 overlapping of analytical methods or procedures  78, 128 results as catalyst for other precipitation chemistry studies  184 succession planning  170 well-defined and evolving questions  125 Hubbard Brook Experimental Forest  87, 121, 183 acid rain  8, 184 biogeochemical response of nitrogen to forest disturbance  11–12 careful use and calibration of field equipment  78, 128 changes in biomass over time  127 field data records  43 field presence  128 focus on scientific productivity and synthesis materials  126–7 leaders, succession in leadership  129 partnerships 128 red-backed salamander, surveys  67 serendipity in success of long-term research and monitoring program 122–3 illegal wildlife trade  179 indicator species concept  32–3, 63–5, 137 problematic issues  33–5 selection of indicators  35, 36–7 ineffective monitoring programs  29, 49–50 characteristics  28–45, 150 other factors contributing to  45–9 infrastructure emphasis on large and sophisticated equipment  38, 144, 172 expenditure, and long-term research and monitoring programs  38, 39 at field sites  49, 81 marine ecosystem programs  149 narrow focus on, NEON/TERN  143–4, 172

204

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

people as critical  38–9, 144, 171 Integrated Marine Observing System (IMOS) (Australia)  140, 149 integrating data from different kinds of monitoring 185 challenges posed by differences in the kinds of entities that are monitored in different ecosystems  185–6 State of the Great Barrier Reef Reports 187 using environmental and economic accounts as a way to demonstrate the value of monitoring and cement support for monitoring  186–7 intellectual property, and data sharing 177–8 interactions between management-relevant adaptive monitoring and policy development 61 International Biological Program (IBP)  37 Johnson, Noye M.  122, 123, 129 Karawatha Forest Park, Queensland, PPBio grid layout  101, 102 Keeling (CO2) curve  75, 173 key personnel, loss of  46–7 Kluane Boreal Forest Monitoring Project, Canada  9, 10–11, 170 Krebs, Charles  10, 170 lack of conceptual model Alberta Biodiversity Monitoring Program 108 EMAP  112–13, 115 NEON 141 PPBio 103 TERN 141 lack of focused questions, EMAP  111–12, 115 lack of funding  46, 47 lack of tractable overarching questions, NEON/TERN 141–2 lack of trigger points for management action 30–1 Lake Baikal, loss of funding threatens long-term monitoring program  47 Lake Washington, long-term research and monitoring 6–7

lakes, eutrophication monitoring, addition of climate change effects to the monitoring scheme 85 landscape ecology studies, key differences in types of  142 large mammalian herbivore–ecosystem interactions, boreal forests, Isle Royale 72–3 large old trees, monitoring challenge  180 large-scale observation networks conversion of huge data streams into useful scientific knowledge  147 detection of major changes in ecological conditions 148 funding of, versus funding of more focused, question-based empirical investigations 148 integration with place-based, long-term site and landscape studies  148–9 interface with the scales needed for ecological monitoring  147 key questions for  146–50 leadership challenges  149 synergy between components of  147–8 see also NEON; TERN laundry list approach  32, 65, 137 Alberta Biodiversity Monitoring Program  108, 110 Lawes Agricultural Trust  117 Leadbeater’s possum, population modelling, Victoria  9, 61–2, 130, 131, 137 leadership challenges, large-scale observational networks 149 effective  70, 75 EMAP  111, 114 Hubbard Brook Ecosystem Study  122–3, 129 NEON/TERN 144–5 programs prone to collapse when leader leaves  46–7, 70, 75 Rothamsted 117 strong and dedicated  70 succession planning  75, 169, 170–1 wet montane ash eucalypt forest projects, Central Highlands, Victoria  131, 137, 139 lichen diversity, urban environments, as indicators of air quality  84

I n d ex

Likens, Gene E.  122, 123, 129 Lindenmayer, David B.  130, 140 Linthurst, Rick, as EMAP leader  111, 114 long-term data frequent examination and use of  77 sloppiness in not reporting trends in  28 long-term data management, challenges of  42, 43 long-term datasets adequate management of  42–3 detection of surprises  12–13 development of co-located, collaborative and multidisciplinary work  12 informing policies and legislation in environmental management  7–8 potential to identify emerging problems 75–6 tests of ecological theory  9–12 time until expression  5–7 use in simulation modelling  8–9 values of  4–13 long-term ecological monitoring funding  16, 46, 47 highly successful  27–8 importance of  15–16 poor record of  13–15 Long-Term Ecological Research (LTER) program (US)  37, 60, 101 data from not included in State of the Nation’s Ecosystems report  183 funding of Hubbard Brook Ecosystem Study 129 rule of thumb for data management costs 42 Long-Term Ecological Research Network (LTERN) (Australia)  140 key considerations in succession planning 170–1 problem of data incompatibility from different studies  186 synthesis versus data integration  145 termination  60, 140, 145, 173 long-term monitoring citizen science role  175, 177 definition 2 to overcome temporal variability in ecosystem conditions  7 small factors contributing to the success of 78–81

205

societal needs for  15, 174 what is not?  2 long-term monitoring programs challenge of keeping them going  180–1 funding  16, 46, 47, 76–7, 172–3, 181 time taken to generate substantial scientific publications  77–8 when to stop?  181, 182 Long-Term Research in Environmental Biology (LTREB) program  3, 60, 166 long-term studies challenge of keeping them going  180–1 never initially intended to become long-term studies  3, 5 reasons they fail or are ineffective  29 longitudinal studies  3 loss of the champion for a project  46–7, 70, 114 loss of cultural infrastructure, impact of  49 loss of integrity of the long-term data record  43–5, 78 Lovett, Gary  129 LTER see Long-Term Ecological Research program (US) LTERN see Long-Term Ecological Research Network (Australia) maintenance of data integrity  43–5, 78, 82, 128 major human disturbance, effect of  47–8 management interventions and monitoring  164 need for trigger points for action  30–1, 180 mandated monitoring  2–3, 164 at a coarse scale  4, 17, 85, 183, 188 features to improve  184 tension between state- and national-level monitoring and site- and region-based monitoring 183 marine ecosystems, success of large research infrastructure programs  149 Mauna Loa, Hawaii, atmospheric carbon dioxide monitoring  74–5, 184 mindless monitoring, lacking questions  29 Mirror Lake, New Hampshire ice cover observations  143 impact of interstate highway on salt levels in 48

206

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

mobility and career development of young scientists 169 modelling focus on, rather than empirical data collection 167–8 see also simulation modelling monitoring cultural changes needed to facilitate 165–74 integration of data from different kinds of 185–7 and management interventions  164 mandated  2–3, 4, 17, 85, 164, 183, 188 non-question driven  148 not everything needs to be monitored  14 optimal balance between two broad kinds of 185 question-driven  3, 4, 15, 17, 148, 183, 184, 188 types of  1–3 monitoring literature, making sense of 16–17 monitoring programs assumption that ‘one size fits all’  35–6, 185–6 challenge of keeping them going  180–1 deficiencies  14, 150 effective characteristics  59–78, 86, 89, 150–1 critical components  88–9 ineffective, characteristics  28–45, 150 integrating different types of  181–5 is a monitoring program needed at all? 28–9 new perspectives  17 overview of success and failure  17 poorly planned and unfocused  28 reasons they fail or are ineffective  29 sequence of steps underpinning successful 182 see also long-term monitoring programs monitoring too many things poorly rather than fewer things well  32–5, 65 montane ash forests see wet montane ash eucalypt forests, Central Highlands, Victoria moose–ecosystem interactions, boreal forests, Isle Royale  72–3 Moreton Bay, Queensland  118

environmental management challenges 118–19 Moreton Bay, Queensland, Ecosystem Health Monitoring Program  118–20 aims and cost  119 credibility and design  121 estuarine, freshwater and marine sampling 120 questions with management relevance 120 regional partnerships  119, 120, 121 report card system  121, 183 review and redesigned program  119–20 Mount St Helens volcanic eruption, postdisturbance studies  48 multi-tasking, tyranny of  168 National Aquatic Resource Surveys (NARS) program  100, 111, 114 aims 112 assessment reports  113 components 111 National Collaborative Research Infrastructure Strategy (NCRIS)  140 National Ecological Observation Network see NEON national observation networks, interfacing with scales needed for ecological monitoring 147 National Science Foundation (NSF) (US)  129, 139, 146 funding criterion  166 program for funding of major, natural disturbances 185 natural experiments  142 natural history observations, central Wisconsin 176 naturalists, demise of the  168–9, 171 nature deficit disorder  169, 174 NEON (National Ecological Observation Network) (US)  37, 139 focus on climate change  148 focus on data collection and sharing  139 funding  139–40, 146, 148 governance structure  144–6 high transaction costs  146 key questions for resolution in the future 146–50

I n d ex

lack of tractable overarching questions 141 leadership turnover  144–5 narrow focus on infrastructure  143–4 no a priori predictions  141 no conceptual model  141 operations costs  146 problems of integrating data from different facilities  146 newly described species, targeted by illegal wildlife trade  179 nitrogen oxide emissions, acidity of precipitation changes, Hubbard Brook 71 nitrogen trifluoride, monitoring, Cape Grim 76 no-take zones, monitoring, Great Barrier Reef 187 non-question driven monitoring  148 success stories  175, 176–7 non-standard monitoring programs arboreal marsupials, wet forests, Victoria  61–2, 68–9 red-backed salamander, Hubbard Brook Experimental Forest  67 Norfolk Island boobook owl, passive monitoring 29–30 novel and non-standard overlapping and rotating statistical design for increasing inference from monitoring programs 68–9 observational studies  142, 143 old data, recoding of  43 one-size-fits-all methodology  36–7, 185–6 PPBio 103 ongoing funding, long-term monitoring programs 76–7 open access to datasets  178 significant impact on some aspects of conservation 179 organisational structure, and long-term monitoring programs  48–9, 128–9, 171–2 outsourcing of projects to consulting companies 172 over-engineering of long-term studies  41

207

overstorey vegetation cover, and occurrence of bird species  34 OzFlux 140 ozone hole detection, American vs British approaches 38 partnerships among scientists, statisticians, resource managers and policy makers  61, 67–70, 71 Ecosystem Health Monitoring Program, Moreton Bay  119, 120, 121 EMAP 114 Hubbard Brook Experimental Forest 128 wet montane ash eucalypt forest projects, Central Highlands, Victoria  138 passenger pigeon  31 passive monitoring  1–2, 29–30 Alberta Biodiversity Monitoring Program  107–8, 110 limitations 2 paucity of questions  29 Alberta Biodiversity Monitoring Program 105–7 PPBio 102 people as critical infrastructure for long-term research and monitoring programs  38–9, 144, 171 field staff  80, 138 partnerships among scientists, statisticians, resource managers and policy makers  61, 67–70, 71 rush to get them to do ‘real’ on-ground work 41 time in the field  81, 168–9 turnover  144, 171, 174 tyranny of multi-tasking  168 see also academic culture and rewards system; leadership; succession planning Pierce, Robert S.  122, 129 place-based research, bias against  167 poaching of wildlife populations  179 politicisation of administrative decisions  29, 30, 71, 145, 172, 174 poor data management  42–3 poor experimental design  31–2

208

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

post-hoc observational studies  142 PPBio approach  101 aims 101–2 lack of conceptual model  103 one-size-fits-all methodology  103 overseas sites  101 paucity of questions  102 positive aspects  103–4 statistical design  103 PPBio Australasia  101–4 Queensland sites  101, 102, 103 predictive capacity of question-driven monitoring 3 probability-based sampling  113 problematic case studies  101–14, 150 program over-engineering  41 Program for Planned Biodiversity and Ecosystem Research see PPBio approach public-funded research responsibility for public to see a return on its investment  77 results published in peer-reviewed journals 77–8 quasi-experiments 142 question-driven monitoring  3, 4, 15, 17, 148, 183, 184, 188 as finer scaled  4 predictive capacity  3 questions evolving  61–2, 82–3, 137, 181 focused, lack of  111–12, 115 of management relevance  181 paucity of  29, 102, 105–7 setting, adaptive policy framework  81–3 setting good  60–1, 116, 120, 125, 131 tractable  137, 138 lack of  141 rare species, challenges in effective monitoring of  178–80 red-backed salamander, non-standard monitoring program  67 remote cameras  180 remote detection/sensing devices  39, 143, 172 report card system  121, 183 retrospective investigations  2, 77

revegetation treatments on mined Australian coastal sand plains, restoration trajectories 7 river restoration projects, poor experimental design 31–2 rotating sampling monitoring approach  68–9, 84 Rothamsted  114–18, 184 adaptive monitoring  117–18 field experiments  114–16 founders 114 leadership 117 long-term financial stability  117, 173 park grass experiment, Hertfordshire 27–8 statistical design  116–17 well-defined questions  116 rush to get ‘real work’ happening on the ground 41–2 savannahs northern Australia, fire experiments  40 Northern Territory, adaptive monitoring to determine drivers of mammalian fauna decline  83 science bureaucrats, rise of  169 scientific discoveries, emphasis on  166 scientific disengagement from monitoring programs 40–1 scientific journals bias against longer papers that are often required to present long-term studies 165 emphasis from editors to reduce the number of citations  165 emphasis on ‘newsy’, short or provocative articles  165, 171 impact factor  165–6 need to push back against current science culture 171 for publishing datasets  178, 180 trend away from place-based research 167 scientific productivity emphasis on more (and shorter) papers 166 Hubbard Brook Experimental Forest 126–7 time taken to produce papers  77–8

I n d ex

wet montane ash eucalypt forest projects, Central Highlands, Victoria  138 selection of appropriate entities to measure 63–5 short-term funding  46, 166 simulation modelling  2, 167–8 long-term datasets use in  8–9 site- and region-based monitoring programs implications of findings from  184 tension with mandated monitoring programs 183 sloppy reporting by investigators  28, 77 snapshot investigations  2 snowshoe hare cycles, ecological mechanisms driving 10–11 Soay sheep  12 societal culture, and long-term monitoring  15, 174 South African Environmental Observation Network (SAEON)  140, 147 spatial scale studies, limitations on generalisation to broader scales  144, 183 staff see people State of the Environment Reports  139, 182 State of the Great Barrier Reef Reports  187 statistical design  3 Alberta Biodiversity Monitoring Program 107 EMAP  113, 115 importance of good  31, 65–7 PPBio 103 Rothamsted 116–17 wet montane ash eucalypt forest projects, Central Highlands, Victoria  137 see also experimental design structure of organisations  48–9, 128–9, 171–2 succession planning  75, 129, 169, 170 key considerations  170–1 sulfuryl fluoride, monitoring, Cape Grim 76 sulphur dioxide emissions, acidity of precipitation changes, Hubbard Brook 71 SuperSites 140 surrogates, selection of  35, 36–7, 65 surveillance monitoring  1 survey methods, changes in, affecting data integrity  43–5, 78, 82, 128

209

temporal changes in major Australian ecosystems, synthesis project  186 temporal effects, lack of focus on  167 Temporally Integrated Monitoring of Ecosystems (TIME) programs  60 TERN (Terrestrial Ecosystem Research Network) (Australia)  100, 139, 186 focus on data collection and sharing  139, 140 funding  139, 140, 148 governance structure  144–6 high transaction costs  146 key facilities and roles  140 key questions for resolution in the future 146–50 lack of tractable overarching questions 141–2 leadership turnover  144–5 narrow focus on infrastructure  143–4 no a priori predictions  141 no conceptual model  141 problem of integrating data from all facilities 147–8 synthesis versus data integration  145 terrestrial ecosystem infrastructure programs, versus marine ecosystem infrastructure programs  149 threatened species challenges in effective monitoring of 178–80 fad-like focus on  166–7 monitoring without management intervention 28 need for trigger points for action  30–1, 180 time in the field  81, 168–9 time until expression  5, 7 long-term monitoring at Lake Washington 6–7 tractable overarching questions lack of, NEON/TERN  141 wet montane ash eucalypt forests projects, Central Highlands, Victoria  137, 138 transportation to field sites  78–80 trigger points for action, endangered and threatened species  30–1, 180 Trophic Cascade Project, Michigan, loss of cultural infrastructure  49

210

E f f e c t i ve E c o l o g i c a l M o n i t o r i n g

tropical savannahs, Northern Territory, adaptive monitoring to determine drivers of mammal fauna decline  83 true experiments  142 tussock grasslands, New Zealand, monitoring programs unreported  28 ugly case studies  150, 151 unexpected major events, impact of  47–8 unforeseen natural disasters, effect of  47–8 United Nations, economic and environmental accounting  18 University of British Columbia, Kluane Boreal Forest Monitoring Project  9, 10–11, 170 Velesunio ambiguus, as indicator of heavy metal pollution, Australian rivers  34 visit–revisit protocols, limitations  68 ‘wait and see’ case studies  139–50 Warra Long-Term Ecological Research site, Tasmania, conceptual model for guiding research 63 water-quality monitoring, limited reporting of 28 weather and natural history events, Shawangunk Mountains, New York State 176 weather patterns and phenological records of migratory bird arrivals and departures, central Wisconsin  176 well-developed partnerships  61, 67–9 benefits of  69–70 West African black rhinoceros, lack of trigger points for action  30–1 wet forests, Tasmania, woody debris, conceptual model of dynamics  63 wet montane ash eucalypt forests, Central Highlands, Victoria  129–39 arboreal marsupials monitoring  61–2, 68–9, 130, 131–3, 139

conservation management strategies 135 as critical forests  130 deficiencies 138–9 funding  138, 173 impacts of clearcut logging methods 133–4 importance of multi-aged stands  134–5 initial research objectives  130 insufficient communication between researchers and resource managers 139 key questions that have motivated the monitoring program  131 Leadbeater’s possum, population modelling  9, 61–2, 130, 131 leadership  130, 137, 139 ongoing development of questions and research themes  137–8 partnerships 138 quality of field staff  138 repeated animal and vegetation surveys 130–1 research themes  130, 137–8 scientific productivity  138 statistical design  137 tractable approach to field research and monitoring  137, 138 wildfires, pre- and post-fire ecological studies  48, 135–6 white-throated sparrow, four sexes of  12, 171 Woods Hole Oceanographic Institution  151 yellow-bellied glider, as proposed indicator species of forest suitability for arboreal marsupials 34 Yellowstone National Park, post-disturbance studies 48 Zonotrichia albicollis see white-throated sparrow