Misreading the African landscape: society and ecology in a forest-savanna mosaic 9780511089718, 9780521564991, 9780521563536

Table of contents :
Frontmatter
List of plates (page x)
List of figures (page xi)
List of tables (page xiii)
Note on authorship and research collaboration (page xiv)
Acknowledgements (page xvi)
Linguistic conventions (page xviii)
Introduction (page 1)
1 Convictions of forest loss in policy and ecological science (page 24)
2 Forest gain: historical evidence of vegetation change (page 55)
3 Settling a landscape: forest islands in regional social and political history (page 86)
4 Ecology and society in a Kuranko village (page 115)
5 Ecology and society in a Kissi village (page 149)
6 Enriching a landscape: working with ecology and deflecting successions (page 176)
7 Accounting for forest gain: local land use, regional political economy and demography (page 210)
8 Reading forest history backwards: a century of environmental policy (page 237)
9 Sustaining reversed histories: the continual production of views of forest loss (page 261)
10 Towards a new forest-savanna ecology and history (page 279)
Appendix I Glossary of plant names (page 296)
Appendix II Cassette recordings of oral accounts and discussions (page 310)
Cassette recordings of oral accounts and discussions (page 310)
Notes (page 314)
List of references (page 327)
Index (page 348)

Citation preview

Islands of dense forest in the savanna of ‘forest? Guinée have long been regarded by both scientists and policy-makers as the last relics of a once more extensive forest cover, degraded and degrading fast due to its inhabitants’

land use. Through meticulous use of historical sources, and close investigation of inhabitants’ technical knowledge and practices, James Fairhead and Melissa Leach question these entrenched assumptions. They show, on the contrary, that people have created forest islands around their villages, and have turned fallow vegetation more woody, so that population growth has

implied more forest, not less. They also consider the origins, persistence and consequences of a century of erroneous policy. Interweaving historical, social anthropological and ecological data, this unique study advances a novel theoretical framework for ecological anthropology, forcing a radical re-examination of some central tenets in each of these disciplines.

BLANK PAGE

Misreading the African landscape

African Studies Series 90 Editorial board Professor Naomi Chazan, The Harry S. Truman Research Institute for the Advancement of Peace, The Hebrew University of Jerusalem

Professor Christopher Clapham, Department of Politics and International Relations, Lancaster University Professor Peter Ekeh, Department of African American Studies, State University of New York, Buffalo Dr John Lonsdale, Trinity College, Cambridge Professor Patrick Manning, Department of History, Northeastern University, Boston

Published in collaboration with THE AFRICAN STUDIES CENTRE, CAMBRIDGE A list of books in this series will be found at the end of this volume

Misreading the African landscape Society and ecology in a forest—savanna mosaic James Fairhead and Melissa Leach with the research collaboration of Dominique Millimouno and Marie Kamano

CAMBRIDGE &:) UNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE

The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS

The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York, NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcon 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org

© Cambridge University Press 1996 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1996 Reprinted 1999, 2002

Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication data Fairhead, James, 1962—

Misreading the African landscape: society and ecology in a forest — savanna mosaic / James Fairhead and Melissa Leach; with the research collaboration of Dominique Millimouno and Marie Kamano. p. cm. — (African studies series; 90) Includes bibliographical references. ISBN 0 521 56353 4 (hc). — ISBN 0 521 56499 9 (pb) 1. Human ecology — Guinea — Kissidougou (Region) 2. Landscape assessment — Guinea — Kissidougou (Region) 3. Forest ecology — Guinea — Kissidougou (Region) 4. Savanna ecology — Guinea — Kissidougou (Region) 5. Environmental policy — Guinea — Kissidougou (Region)

5. Kissidougou (Guinea: Region) — Environmental conditions. I. Leach, Melissa. II. Title. III. Series. GF746.2.F35 1996 304.2°096652 —dce20 95-52318 CIP

ISBN 0 521 56353 4 hardback ISBN 0 521 56499 9 paperback

WD

This book is dedicated to the memories of Mr Oury Bah and Professor Rowland Moss, and to their confidence in the ability of African farmers in the transition zone to manage their own environments.

BLANK PAGE

Contents

List of plates page Xx List of figures x]

List of tables xlil

Note on authorship and research collaboration XIV Acknowledgements XV1 Linguistic conventions XViil

Introduction 1

1 Convictions of forest loss in policy and ecological science 24

2 Forest gain: historical evidence of vegetation change 55

political history 86 4 Ecology and society in a Kuranko village 115 3 Settling a landscape: forest islands in regional social and

5 Ecology and society in a Kissi village 149 6 Enriching a landscape: working with ecology and

deflecting successions 176

7 Accounting for forest gain: local land use, regional

political economy and demography 210

environmental policy 237 of views of forest loss 261

8 Reading forest history backwards: a century of

9 Sustaining reversed histories: the continual production

10 Towards a new forest-savanna ecology and history 279

Notes 314 Index 348

Appendix! Glossary of plant names 296 Appendix II Cassette recordings of oral accounts and discussions 310

List of references 327 1X

Plates

Plate section between pages 76 and 77 2.1 Comparison of 1952 and 1991 aerial photographs and 1992 SPOT satellite images, Kiss 2 project area, north-east Kissidougou 2.2 Comparison of 1952 aerial photographs and 1992 SPOT satellite images around Sandaya, north-west Kissidougou 2.3 Comparison of 1952 aerial photographs and 1992 SPOT satellite images around Moria, northern Kissidougou 2.) Comparison of 1952 aerial photographs and 1989 SPOT satellite images around Kofodou, eastern Kissidougou 2.6 Comparison of 1952 aerial photographs and 1989 SPOT satellite images in Koundiadou district, southern Kissidougou

0.1 The forest-savanna mosaic of Kissidougou prefecture page 3

0.2 Forest island in the savanna 4

0.3 Road entering a forest island from the savanna 5 1.1 Savanna after the passage of dry season fire 27 2.4 Comparison of 1952 and 1991 aerial photographs to show formation

of a forest island in Bambassiria district, north-east Kissidougou 61 3.1 Silk cotton tree (Ceiba pentandra) planted and emergent

in a growing village forest island 90

4.1 Young men’s work group cultivates an inland valley swamp 125 4.2 Female farmers working the land of an abandoned village site 128 4.3 A woman farmer planting peanuts on the mounds of a regularly

gardened swampside site 129 4.4 Aman weaves a sleeping mat for sale in the village 134

4.5 Arecently cleared farm-site in gallery forest 141 5.1 View of Toly’s territory from the top of Sumafaa hill 151

association 152 coming-out 153

5.2 Coming-out ceremony following initiation into the men’s power 5.3 “Toly Condition’: Painted clay sculptures overseeing the men’s

island 219

5.4 Women’s weeding group in an upland rice farm 159 7.1 An urban-based logger sawing timber felled in a village forest

xX

Figures

of West Africa page 2

0.1 Kissidougou prefecture in the forest-savanna mosaic zone

0.2 Location of research sites in Kissidougou prefecture 19

1.1 Rainfall in Kissidougou Prefecture, 1922-1993 28

1.2 Changes in season length, 1934-1993 31

language groups 33

1.3 Map to show approximate distribution of Kissidougou’s main 1.4 Present-day vegetation forms seen as anthropogenically

degraded derivates of previous vegetation 38

1.5 Rainfall variation within Kissidougou prefecture AO

Aubréville 1949 4]

1.6 Limits of assumed past forest cover in West Africa, after

2.1 Districts in which large expanses of savanna have ceded to

forest thicket vegetation since 1952 57

2.2 Vegetation change in the localities of Toly, Fondambadou and Bamba from 1952 aerial photographs and 1989/92 SPOT

satellite images 62

2.3 Map to show routes taken by observers who described

Kissidougou’s landscape in the early twentieth century 65 2.4 Comparison of archival maps and modern remotely sensed

imagery to show changes in forest island area 71

3.1 Spatial distribution of termite mounds in humid savannas 91 3.2 Population density in Kissidougou prefecture by district, 1993 96

region, 1600—1850 101 4.1 Sketch map to show village territory, Sandaya 117 3.3. Map to show main population movements in the Kissidougou

organisation, Sandaya 119

4,2 Schematic representation of clan, compound and farm-household

4.3 Indigenous swamp improvement techniques 139 5.1 Sketch map to show village territory, Toly 150 6.1 Species used as roofing materials in five villages 178 6.2 Species used to make selected items of household and agricultural

equipment in five villages 182 6.3. Fuelwood sources in different seasons 185 XI

Xll List of figures

6.4 Schematic transect of an inland valley swamp 192

1917-1993 213

6.5 Vegetation cover on tombondu and ‘new land’, Sandaya 199 7.1 Population change in Kissidougou prefecture by district,

and use 223

7.2 Relationship between population density and forest island cover 214 7.3. Diagram to show three organisational patterns of land tenure

7.4 Relationship between population density and all forest cover 232

Tables

1.1 Extent of humid forest in forest Guinée in different periods page 36 2.1 Characteristics and descriptive terms for major vegetation

forms in Kissidougou’s forest-savanna mosaic 58

2.2 Areas of forest and secondary forest thicket vegetation in

1952, 1982, 1989/92 in five localities 60 2.3 Comparison of archival landscape descriptions at the turn of the century with present-day vegetation 66 2.4 Comparison of forest islands as described by Brossart, 1910-11

with present existence and 1982 area from aerial photographs 68

2.5 Village forest island origins 80 3.1 Early population size estimates of Kuranko areas in Kissidougou 98

3.2 Natal origins of current married village residents 109 4.1 Changes in the use of forest and savanna farm-sites since the

early twentieth century in Sandaya 123

Sandaya, 1993 126

4.2 Peanut, cassava and fonio fields and personal rice swamps in 5.1 Changes in the use of forest and savanna farm-sites since the

early twentieth century in Toly 157

6.1 Plant treatments for common ailments 186 6.2 Tree species recognised by villagers as apt for different

propagation techniques 190

6.3 Site characteristics in survey of tombandu and ‘new land’ 197 6.4 Tree species occurrence on tombondu and ‘new land’, Sandaya 200 6.5 Examples of dominant species composition and structure of

different forest island areas 208

7.1 Population change in Kissidougou by sub-prefecture, 1917-1993 211 7.2 Official statistics of cattle numbers in Kissidougou from 1900 227

xiil

Note on authorship and research collaboration

This research is the product of several forms of collaboration. The first is between us, Melissa Leach and James Fairhead, as the joint and equal co-authors of this book. James Fairhead is a social anthropologist who began this study while a researcher at the Natural Resources Institute (NRIJ), and completed it as a Research Fellow of the Department of Sociology and Anthro-

pology, School of Oriental and African Studies, University of London. A graduate in agriculture and forestry sciences, his social anthropology PhD and subsequent work had focused on African farmers’ knowledge and agrarian change. Melissa Leach, also a social anthropologist, is a Fellow of the Institute of Development Studies at the University of Sussex, and co-director of its Environment Group. A graduate in geography, her social anthropology PhD

and subsequent work had focused on gender and broader social issues in natural resource management, particularly around West African forests. While the scope of the present study grew out of a combination of our interests and analytical perpectives, it was broadened by our shared interest in historical approaches and deepened by interaction during fieldwork, analysis and writing so close that it is now impossible to distinguish our respective roles. Equally, our own perspectives have been enriched through a second form of collaboration, with the Guinean researchers Dominique Millimouno and Marie Kamano with whom we formed a team in Guinée called COLA (Connaissance

et Organisation Locales Agro-écologiques). Dominique Millimouno is a graduate in history and philosophy of the University of Kankan, who had earlier studied Kissi myths and oral traditions, and who focused principally on the recording of oral histories during this study. Marie Kamano, a graduate in botany of the University of Kankan, carried out many of the resource surveys during this research. While we did most of the in-depth fieldwork in two main Study villages, our Guinean colleagues were more mobile, comparing these localities with others which we would visit for shorter periods. The third form of collaboration was between COLA and several Guinean institutions: the Direction National de la Recherche Scientifique et Technique

(DNRS) of the Republic of Guinée’s Ministry of Higher Education, whose XIV

Note on authorship XV support to this research was fundamental; the Direction National des Foréts et Faune (DNFF) of the Republic of Guinée’s Ministry of Agriculture and

Animal Resources, and thirdly the Projet de Dévéloppement Rural de Kissidougou (DERIK).

Acknowledgements

As co-authors, our first acknowledgements must be due to our colleagues in

this research, Dominique Millimouno and Marie Kamano. Without their combination of investigatory spirit and patience this study would have come to nothing. Special thanks are due to the villagers of Sandaya, especially the du ti

Saragbe Mansare and the so ti Denwulen Mansare, and our landlords and compound hosts Mamadi and Kali Mansare, Manty, Mundu, Kanko Kourouma

and her daughter Sayo; and to Toly’s inhabitants, especially the du ti Faya Yaradouno, elders Kissi Kondiano and Bisi Tolno, and our compound hosts, the family Wamouno. We especially thank Dauda Yaradouno, our indefatigable companion in Toly. We would also like to thank the inhabitants of Moria, Fondambadou, Bamba and Massamaya for their assistance to the team, and the many elders of the region who spoke to us of their own and their ancestral past, many of whose names are cited in Appendix II or in the text. In Kissidougou, we are also grateful for the assistance of the district level representatives, Kaba Traore of Foria and Dominique Leno of Bongoro, for supporting our stay. We

gratefully acknowledge the support and help of Kissidougou’s Prefet, the Sécrétaire Général de la Décentralisation, Fadama Kourouma, and the SousPrefets and CRD authorities of Sangardo, Koundiadou, Albadaria, Firawa and Yende Millimou. This research was made possible by a grant from the Economic and Social Committee on Research of the British Overseas Development Administration (ODA). We are extremely grateful for this support.

At national level in Guinée our gratitude is due to the Ministére de l’Enseignement Superieur, with special thanks to the then Minister, Charles Pascal Tolno; the Directeur des Affaires Etrangéres Ibrahim Solo Conde, and Dr Kabine Kante of the Direction National de la Recherche Scientifique et

Technique. Equally we are grateful to the rectors of the Universities of Conakry, Kankan and Faranah for their support, and to the staff and students of

these universities for their useful comments and questions during COLA’s seminars presented at these universities. We are in debt to the staff of the Direction National des Foréts et Faune (DNFF), Conakry, especially their late director, Mr Oury Bah, whose intellectual and administrative support was XV1

Acknowledgements XVli fundamental for this research; staff at the Bureau Technique; Paul Anspach and Dietrich Suhlrie of the GTZ projet Conseiller Forestier, and the Programme d’Aménagement des Bassins-Versants, especially Tahirou Barry and Pietro Marino of the Cellule de Co-ordination, Conakry, and projects Kiss 1 and 2. Many thanks are due to Mr Diaby, the Director of Projet DERIK and his staff, as well as the German GTZ team and its leader, Otto Honke. The GTZ Service d’ Administration provided us with invaluable administrative help, and the staff of both DNFF and DERIK provided us with useful comments and clarifications during our assorted seminars, for which we are most grateful. We would also like to thank Projet SNAPE and the Institut de Topographie et de Cartographie, Conakry for their assistance with aerial photographs, and

the archivists of the Archives Préfectorales, Kissidougou, the Archives Nationales of the Republique de Guinée — especially the director, Mr Cissé — the Archives Nationales du Senegal (Dakar) and the Archives Nationales de France, section Outre Mer (Aix-en-Provence). We are in debt to the Directeur of the Centre de Recherche Agronomique de Seredou, and to their researcher,

Mr Jean Louis Hellié, for his incomparable skills in the identification and

use of the region’s flora. We are equally in debt to our team of tape transcribers/translators who performed this arduous task with admirable skill. We would also like to thank Robert Harms who gave us invaluable guidance

in the practicalities of historical research, especially in Guinée, and made available for our reflections the most spectacular flat in Conakry. Gerald and Penelope Leach, and Thomas and Jane Fairhead made valuable comments and editorial suggestions on earlier drafts; many thanks for these,

and for so much encouragement throughout this study. The manuscript’s anonymous reviewers made useful comments which have helped to improve this book, and it has also benefitted from the comments of participants in numerous seminars and conferences in England, France and North America. While we are grateful for the financial, administrative and intellectual support

of all the above people and institutions, we must note that the opinions represented here, and any errors, are the co-authors’ own, not those of any of our collaborators or of the ODA.

Linguistic conventions

This book deals with three non-English languages: French, Kissi and Kuranko.

French titles and citations from archives and modern literature have been translated into English by the authors, as have longer citations in the two national languages. Kissi language speakers are more properly called Kissia in the plural, and the language which Greenberg (1966) classifies within the Mel sub-branch of the West Atlantic language group is more properly called Kissieé. Kuranko (sometimes spelt Kouranko or Koranko in the literature) is a dialect of Maninka, a language of the Manden/Manding branch of the Northern Mande language group. Both Kissi and Kuranko words and phrases are italicised in the text, but we have also differentiated Kissi and Kuranko words and phrases by typeface. Kissi words thus appear as Kissi and Kuranko as Kuranko. Several

dialects of each language are spoken in Kissidougou prefecture, and much spoken Kuranko directly uses Maninka forms; we have made no attempt to differentiate between these. Such Kissi and Kuranko words as we use are translated in the text, and transcribed broadly following the orthographic conventions of the new alphabet of the Republic of Guinée designated in 1989 by the Service National d’Alphabétisation and the Institut de Recherche Linguistique Appliqué in Conakry. Kissi has four functional tones, but we have not marked these in this work. The symbols ¢, 9 and n in these written languages are pronounced as follows: ¢ like the ‘e’ in English ‘pet’ 9 like the ‘o’ in English ‘pot’ Sometimes as a consonant similar to english ‘ng’ as in ‘sing’, and sometimes to add nasality to the preceding vowel.

XVIll

Introduction

Kissidougou’s landscape is striking. Over open expanses of grassy savanna

tower patches of dense, verdant, semi-deciduous rainforest. These forest islands, scattered over the gently rolling hills, are generally circular, perhaps a

kilometre or two in diameter, and most conceal at their centre one of the prefecture’s 800 or so villages. Apart from these islands, dense forest vegetation is found only in narrow strips along streamsides or swampy valley bottoms. This landscape resembles that in many parts of the West African forest-savanna mosaic or ‘transition’ zone, which stretches along the northern fringe of the forest zone from Sierra Leone eastwards to Nigeria and beyond (figure 0.1). Since the first French occupation in 1893, Guinée’s administrators have been convinced that these forest patches are the last relics of an original dense

humid forest which once fully covered the landscape. They suppose that inhabitants have progressively converted this forest into ‘derived’ savanna through their shifting cultivation and fire-setting practices, preserving only

the narrow belts of forest around their villages. From the outset, this ‘savannisation’ has been a major policy concern because of the threat it posed to the local agricultural and tree crop economy and to regional climate and hydrology. Today, and with the addition of global environmental concerns, the degraded and degrading landscape of Kissidougou attracts major international funding for environmental rehabilitation. These concerns are not unique to Kissidougou: they have been expressed thoughout much of the transition zone, and elaborated during a century of scientific investigation. Our research was originally conceived to engage with natural science and policy debates concerning savannisation, in a landscape the broad vegetation

history of which seemed well known. We were broadly interested in the social dynamics of deforestation; in how different inhabitants understood this process and responded to it technically and socially. Put crudely, we sought to elucidate its social causes, consequences and ramifications, and especially to identify how wider issues in the region’s political economy articulated with these.

Early in the research, however, it became clear that the very pattern and 1

2 Misreading the African landscape

Yy Y/Y | NIGER Ly “iy MAURITANIA | MALI

et )t| /J! NUN BURKINA antUy )~ ;;a / FASO S-“N

GUINEA NS

SIERRA “no | —

SO : " a . J ae a / 3 fl . AN f CEES aC LIBERIA PN ree ete tae ra Susiiunisenceseekieees ( D’IVOIRE es

0 km 400 REE COTE G A © a eraretarararererere esas

| Faranah | 4 , x . Kankan ~e “ ° areeeeataes®, ’

; Kouroussa ' { Lowland rainforest (wetter types) ‘N. , a cant, oN ~ Lowland rainforest (drier and mixed types)

° :;\ ,Mees “tT. op ae Kissidougou-

‘

.* _ . “,Se Kerouane ? _* t Forest—savanna mosaic __2 Beyla L.

y bone . Gueckedous Maecenas A L Sudan woodland and savanna (southern)

: Wi po New‘ ckore Rol a Sudan woodland and savanna (northern)

fe SS ast COTE Sahel grassland and thicket

fF “LIBERIA » pes A OIRE Li °

0 km 100° en Ment 5 |

Figure 0.1: Kissidougou prefecture in the forest-savanna mosaic zone of West Africa (based on the UNESCO/AEFTAT/UNSO vegetation map of Africa).

direction of vegetation change in Kissidougou’s forest-savanna mosaic was open to dramatic reappraisal. As this book explores, elders and others living behind the forest walls provide quite different readings of their landscape and its making. At their most contrasting, they bluntly reverse policy orthodoxy, representing their landscape as half-filled and filling with forest, not halfemptied and emptying of it. Forest islands, some villagers suggested, are not relics of destruction, but were formed by themselves or their ancestors in savanna. And rather than disappearing under human pressure, forests, we were

Introduction 3

CNN oe lg LE: Nee Me A Ie | es LOT ee eae 1 OS Ie smRaCCN EL yA. |

Plate 0.1: The forest-savanna mosaic of Kissidougou prefecture: patches of

dense, semi-deciduous forest and strips of streamside gallery forest lie scattered in relatively open savanna uplands.

shown, are associated with settlement, and come and go with it. In short, it became clear that by treating forest islands as relics and savannas as derived, policy-makers may have been misreading Kissidougou’s landscape, by reading forest history backwards. That a landscape could become subject to such conflicting readings invites — even demands — consideration of the worlds which have generated them. In

Kissidougou, our research agenda thus broadened not only to address what could be said about the making of this landscape and what its inhabitants made of it, but also to consider critically what scientists and policy-makers had made of it too, and the social contexts which had generated these different views. Considering all landscape interpretations as in part socially constructed does not, however, negate the fact that certain readings can be demonstrated as false, and that historical evidence might support some more than others. Indeed if, as

we shall argue, the views driving policy are demonstrably false but have brutal instrumental effects, it is the clarification of ‘real? history which renders

apparent the power relations of which these views are part. This issue is more than academic. The view of Kissidougou’s landscape as degraded and

degrading has justified state action to take resource control from local

i

Fe mn ee FF e-. i ii =F =. ad * .

4 Misreading the African landscape

aee.a E . if--_ FF. , ,. — = h”hrlrrrr—..—=Ci“(< iéeCOC;CCC;:CC =

SE ESSE SER Ree RMR ee ABS RS 0 Bic Soe = URES ER SiS [OBES SS RE Se ..©=6Ch6Me.lmrwrmCm™Cm™C#P’-'[_fm.

oe oe ge I ee a eee ne rer oo : ee ee ee ,—=“ E—E .—rt— e **Ld bd * Ld

inhabitants, and repressive policies to reorientate what has been seen as destructive land management. Many aspects of local land use have been

€ . * * Ly ¢ t Ld «

criminalised. Setting bush fires — whose timely application has been a central feature of villagers’ land management — carried the death penalty in the 1970s. ' Formal fines and informal extortion by environmental services further tax an

a Ld e °

already cash-poor population. A landscape filling, not emptying, of forests under local land management would have very different implications.

The making of a landscape had

A fundamental element of this study has therefore been to establish with as

«*2e.°

much precision as possible the course, pattern and causes of vegetation change in Kissidougou’s forest-savanna mosaic. Clearly, this first involves attention to evidence of vegetation change, which we have compiled from documentary,

photographic and oral accounts during the last century and before. It also involves attention to specific land management techniques, grounded in farmers ecologica owledge, and to the social ical knowled d to thand laleconomic and icrelations relati whic hich

Introduction 5

9— 7 : caf errr t—‘ai‘_ ‘_ONNSOO*iOtisCOrsOéCd«s«sr— (wr eee

Plate 0.3: Road entering a forest island from the savanna (Dembayara, western Kissidougou).

have conditioned land use. It is important at the outset to clarify the particular perspectives which we have taken in analysing these issues. Land management and socialised ecology

This study adds to a now large literature which attempts to document the apro-ecological knowledge of African farmers. In general, this literature demonstrates the sophistication of local land, soil and vegetation management techniques, and the wisdom and creativity of farmers; in effect, the operation of a dynamic indigenous science (e.g. Richards 1985, Chambers et al. 1989, de Boef et al. 1994, Scoones and Thompson 1994, Warren et al. 1995). When considering the ecological concepts and explanations employed by Kissidougou’s farmers, however, it has not always been straightforward to understand these in

terms familiar to the western scientific literature on forest-savanna ecology. This problem stems, at a fundamental level, from the framing of scientists’ and inhabitants’ explanations within very different root assumptions concerning the relationship between social and ecological processes.

In short, since the Enlightenment, western science has conceptualised natural and social phenomena as being of a different order; as a priori separate. It is assumed that ‘natural’ phenomena can be investigated as separate from

6 Misreading the African landscape human society, except in as much as people and their social world are subject to ‘nature’ and act on ‘it’. Boundary problems are interesting, but do not undermine the conceptual scheme. It is partly this perspective which circumscribes

the concerns of ecological science in the forest-savanna transition zone, encouraging ecologists to break down their consideration of vegetation into ‘natural’ forms and processes (involving plants, animals, soils, water and so on) as if uninfluenced by people and society — so-called anthropic factors. Ecologists who seek out untouched nature — pristine forest — against which to assess human impact are drawing on and reaffirming this divide; they may be disappointed, but not conceptually challenged, when they find pottery sherds in their soil pits beneath ‘natural’ vegetation (Van Rompaey 1993, Jones 1956). Nature is an essential reference point against which to discuss human impact.

Kissidougou’s forest islands have proved a prime subject for this view, considered by ecologists as islands of more original, albeit threatened nature:

the legacy of a natural ‘climax’ vegetation remaining within an otherwise

abused landscape. This ideal of nature has, of course, been central to conservation policies which have commonly deemed the exclusion of people as necessary for the preservation, or reestablishment, of nature (Anderson and Grove 1987, Adams 1990). Treating nature and people as opposed, and people’s society as phenomenologically different, does not of course foreclose on analysis which theorises the

relationship between ‘nature’ and ‘society’, culture and environment, or people and habitat (human ecology and cultural ecology). It does, however, structure the ways that such relationships are theorised.2 Whatever the merits or otherwise of this perspective, one must recognise, first, that not all people conceive of social and ecological relations in this way: of nature and society/culture as separate objects of analysis which can be examined on their own, and in interrelationship. In many African societies, such a distinction is alien; categories of thought are structured in very different ways which cross-cut a nature—culture divide (cf. Croll and Parkin 1992).3 Indeed unversed in western views about the nature of nature, Kissidougou’s rural inhabitants, we shall suggest, broadly associate the presence of forest not with a near ‘natural’ state, but with a ‘settled’ state. And rather than seeing savanna as a Cultural artefact (derived savanna, abused nature), savanna is associated more with vagrant, mobile and impermanent lifestyles; with the human world perhaps of hunters, the mad, pastoralists and refugees. One might be tempted to equate settlement with ‘social’, and vagrant with natural, but when one sees that it is not just human life which has a settled existence and forest association, but also many land spirits and termites, which also live in ‘villages’ associated with denser vegetation, such an equation becomes more problematic. Furthermore, even if these conceptual categories were treated as analogous,

Introduction 7 one would find a second order of difference. While in western thought nature and society are, in their ideal form, inherently distinct, albeit acting on each

other, for Kissia and Kuranko these potentially analogous concepts are inherently indistinct, moral and fallible. Cycles of birth, growth, maturity and

death common to people, land spirits and termites, and to animals, crops and plants, can directly interfere with each other and must be scrupulously and actively maintained as separate. Space for human settlements and reproduction

must be established as very separate both from the settled life of spirits and termites, and from the vagrancy of animals and crops. The problem is not how to theorise the interaction of these (the subject of western human ecology) but how to prevent it. That inhabitants’ categories cannot be fully understood in terms of western

ones is noted in works throughout the West African region, and elucidating such differences has been the stuff of much social anthropological inquiry (e.g. Gottleib 1992). Such studies underscore how attempts to consider local

representation in terms of these modern western categories obscure inhabitants’ own perspectives on social and ecological relations. The relevance of this critique has, however, often been overlooked in the study of farmers’

agro-ecological knowledge. Indeed, the extent of attempts to overcome western Scientific categories can be used to distinguish two broad genres in the analysis of what has come to be termed ‘indigenous technical knowledge’ or

ethnoscience,* at least when focusing on soil and vegetation phenomena (Fairhead 1992). In such studies, the first and predominant analytical strategy

has been to consider the wisdom of ‘folk’ representations in terms of the conceptual apparatus of western natural sciences. This has evaluative aims, seeking to assess the validity of indigenous knowledge, and how it can be brought into a productive partnership with western science in environmental management and development processes (e.g. Chambers et al. 1989, Farrington and Martin 1987). A second genre, more firmly rooted in social anthropology, takes the more

comparative perspective, asking how agro-ecological concepts (in natural science as much as ethnoscience) develop and make sense within their

particular socio-cultural and economic, as well as ecological, milieux (Fairhead 1992, Hobart 1993, Thrupp 1989, Scoones and Thompson 1994). From this perspective, for example, instead of searching for, and failing to find, the local equivalent of a western concept of plant ‘disease’ to explain poor yields, it would be appropriate to investigate the broad ways in which farmers consider plant health. If farmers do not recognise disease, analysts would not deduce ignorance and hence the need to teach about it, but instead seek out other explanatory frameworks in which farmers consider ill-health, with which scientists might be able to establish dialogue (e.g. Trutmann et al. 1993). To take a second example, instead of seeking out local knowledge of soil nutrient

8 Misreading the African landscape cycling, texture and morphogenesis in exploring ‘indigenous knowledge’ of soil quality, this perspective would hope to comprehend the operation of the field of local concepts pertaining to site productivity; concepts which may have no equivalents in western agricultural science. In Kuranko-speaking areas, as we will suggest, ideas of past habitation and of soil ‘maturity’ and ‘oiliness’ which have little obvious relevance to agronomic science are fundamental for comprehending farmers’ rationales. In attempting to evade the centrisms of the evaluative perspective, this genre explores the references or allusions which local ecological explanations make

to other phenomena and everyday experiences, including those perhaps concerning human health, wellbeing and social life. Identifying these metaphors may be essential to comprehending what is being said. Attention to such ‘metaphors’, moreover, can show up relationships which are considered to be stronger than mere likeness; where social and ecological processes are linked together intercausally (Overing 1985, Croll and Parkin 1992). Thus when fishing fails, Kuranko might say that the water is ‘tied’, when hunting fails that the bush is ‘tied’, and when a woman ‘fails’ to become pregnant, that she is ‘tied’. Certain activities can provoke a state of tying across all of these domains; if, for example, a menstruating woman enters a fishing pool, both she

and the water may simultaneously become tied. Treating tying merely as metaphor would obscure this, as well as hiding common aspects of fishing and

hunting ecology in local representation: for example, when a senior hunter dies, both fishing and hunting immediately become tied. Importantly, attention to such links throws into relief particular ways that representations of ecology might be socialised. In contrast, the first, evaluative genre obscures how agro-

ecological reasoning helps to constitute and reproduce the socio-political conditions within which it gains its sense, as well as the social implications of differing ‘ecological’ opinions and assertions. While much of western science examines natural and social phenomena separately, it is conventionally the disciplines of human and cultural ecology which have examined their interaction. Often locked into the nature—culture impasse, works in this genre have rarely sought to reflect on alternative ways

in which ‘ecological phenomena’ are ‘socialised’, and social phenomena ecologised. The dominant perspective in these sciences, which became popular from the late 1950s, ‘suffered from a naive organismic view of society and a functionalism that saw culture as having adaptive value with respect to the general goals of living systems’ (Peet and Watts 1993: 239). Myth, ritual and symbolic systems were thus examined for the regulatory role they played in Shaping land use and understanding, such that ecological and social systems remained in equilibrium; equilibrium which was, in the most extreme versions, seen to operate through cyclical and homeostatic mechanisms (e.g. Rappaport 1968). In Kissidougou, it is this kind of theoretical position which has enabled

Introduction 9 some studies to postulate the past existence of an equilibrium between a functional social order and ‘nature’ — in its iconic form of a full forest cover.°

Supposed forest degradation is thus explained in terms of emerging social dysfunction, and a breakdown of the culture—nature equilibrium. Such views of change in terms of ‘disruption’ or ‘breakdown’ of the systemic order, and

a subsequent dialectic in which ‘culture’ and ‘environment’ are out of synchrony, are inherent to this approach.

As Amanor argues, however, this perspective on ecology in effect ‘robs society of its history’ (1994: 19), replacing it with another ‘history’ which

obscures the parameters of local struggles (cf. Croll and Parkin 1992). Furthermore in that the functional equilibrium of society and environment is seen as worked out in unconscious ways, lying above human cognition, these

cultural ecology perspectives tend either to silence local people’s own analyses, or to co-opt them by reinterpreting them. In reducing ‘human Organisation and consciousness to a regulative mechanism for preserving an equilibrium’, functional equilibrium models rob people of their action and consciousness (Amanor 1994: 19). As will become terribly apparent in the Kissidougou case, this is not merely academic debate. Social science analyses which draw on such functionalist models have become the complement to natural science in sustaining policy-makers’ views of degradation. In contrast, local peoples’ own theories concerning social and ecological issues carry very different implications for issues of agency, cause and responsibility in explaining vegetation change. At the same time as theories of equilibrium have made a major impact on cultural ecology, ecologists themselves have been questioning their applicability to the natural world, refashioning their view of nature in the process, albeit within the society—nature distinction. This rethinking has involved a shift

in fundamental assumptions about the orderly functioning of bio-physical systems, which had been enshrined in the dominant perspectives of systems ecology (e.g. Odum 1952). New attention has been given to issues of inherent instability, non- and disequilibria, chaotic fluctuations and the role of specific contingent events in ecosystems (e.g. Worster 1990, Noy-Meir 1982, Sprugel 1991). This undermines classical representations of ecosystems in terms of tendency to equilibria, and of universal laws — such as of plant succession to climax vegetation (Clements 1916) — conditioning their structure and function. Sometimes heralded as a ‘new ecology’, giving the semblance of providing a coherent alternative, this rethinking arguably amounts less to the provision of

a single alternative than to freeing ecological reasoning from the strictures of unified theory, enabling theoretical pluralism (McIntosh 1987). While they still examine the world within the conceptual nature—culture dichotomy, these

views of nature nevertheless deny visions of society and ecology as in functional equilibrium (cf. Zimmerer 1994). Correspondingly, they compromise

10 Misreading the African landscape managerial perspectives, whether in cultural ecology or policy, which assume

that society can necessarily predict and manage ecological relations for stability (Botkin 1990, McIntosh 1987).

We will argue that a more pluralistic ecology, and appreciation of nonequilibrium issues, provide conceptual space for reevaluating vegetation dynamics in the forest-savanna transition zone. When freed from its nature— culture straitjacket, this perspective enables better comprehension of local ecological reasoning, and reappraisal of people’s roles in achieving environmental stability or directing change. This book therefore moves with others in this trend towards pluralism within ecological science, but simultaneously

makes the case for an incorporation of land users’ own perspectives and conceptual frameworks.

A pluralistic ecology also allows appreciation of actual environmental histories, rather than marginalising these to a narrow interest in intra-system oscillations as equilibrium perspectives do (McIntosh 1987). As Worster points out ‘Abstracted from time . . . ecosystems may have a reassuring look of permanence; but out there in the real, the historical, world, they are more perturbed than imperturbable, more changing than not’ (1988: 297). In this respect, newer approaches in ecology actually share much in common with an older tradition of studying natural history, which pre-dated Clementsian succession-and-climax theory (e.g. Thoreau 1860). Indeed historical perspectives in ecology have never been entirely suppressed, and existed in some tension with the systems ecology perspectives even when these dominated in the 1960s and 1970s (McIntosh 1987). Even at this time one proponent then

argued that ‘more history and less philosophy are needed for progress in ecological study in Africa’ (Thomas 1965).

Social and economic conditions of land management

To understand historical patterns of landscape change, it is necessary to pay

attention not only to the technical aspects of land management and its interaction with ecology, but also to the broader conditioning of land use. We

therefore examine the changing local social and political relations which influence how people deal with land and vegetation, and, more broadly, how

regional economic, political and demographic changes shape these local dynamics.

Here, we draw both directly and indirectly on a large and useful literature

on agrarian change in Africa which, while not necessarily focusing on environmental relationships per se, investigates how land use is shaped by the

dynamics of resource access, use and control, setting these firmly within questions of social form and process (e.g. Berry 1989, Guyer 1986, 1984, 1991). Works in this genre identify how social conditions influence people’s

Introduction 11 interests, opportunities and choices, in relation perhaps to their gender — the

focus of much work — or to wealth, age, stage in social cycles or other influences. Rather than assume the locus of production or resource management to lie in units such as ‘the household’ or ‘the community’, they ask what these — among other unities — mean for the resource management of different members, and what resource management means for the constitution of such locally specific institutional forms (e.g. Whitehead 1984, Leach 1991, Linares 1992). Some of these authors explicitly locate their analysis within broader theoretical approaches (notably of a neo-marxist kind; e.g. Mackintosh 1991). Others, aware of the importance of broader political and economic structures, nevertheless find such structural theory over-deterministic, marginalising

people’s own determination and influence, and so take more eclectic or pluralistic perspectives. But whatever their perspective, these studies clearly show how social arrangements condition resource use and management: for example through tenurial arrangements over land and trees (e.g. Bruce and Fortmann 1992, Berry 1988), or through capital and labour relations, including social divisions and routines of work, relationships between personal and ‘household’ production, and people’s control over their own and access to others’ labour (e.g. Guyer 1991). Equally they trace how people’s roles and responsibilities — for example in providing food or equipment — shape their concerns with particular trees or sites, crops or products. Examining land management from this perspective shows that resource positions and responsibilities do not depend on ‘conventions’, although they might be conditioned by them, but are the subject of ongoing ‘struggles’ (e.g. Carney and Watts 1991). In such struggles, competing claims to resources may well be ‘articulated through cultural idioms in the charged contexts of local politics’ (Moore 1993), but are not determined by them. As Berry (1989) shows, competing interests over resource use and management are generally resolved in ways which reflect broader relations of power and influence. This renders the diverse social institutions which condition such relations as central to land use and resource management; an idea which Berry herself addresses through the metaphor of people’s ‘investment in institutions’. Marriage or initiation societies, for example, may have little overtly to do with ‘environment’ but nevertheless be important in understanding environmental use and

management. In structuring the relations of power between people, they become relevant to issues of tenure, labour, crop and product control. Tenure, labour, and resource control patterns can shift over time, along with broader social institutional arrangements, with major effects on material land and vegetation use. Shifts can relate to demographic conditions, for instance, or to changing local and regional political and economic relationships. The

effects of such shifts on the details of socially differentiated farming and resource management have been revealed especially effectively in historically

12 Misreading the African landscape focused studies by anthropologists and historians examining gender issues (€.g. Guyer 1984, Leach 1994, Moore and Vaughan 1994, Berry 1993). Considering how regional, national or international processes articulate with local ones over

a historical time-frame is also a shared feature of approaches in what has come to be termed political ecology (e.g. Bryant 1992, Peet and Watts 1993,

Zimmerer 1994). Blaikie and Brookfield (1987), in particular, analysed environmental change by linking proximate household or farm-level issues up to regional and international levels through what they termed ‘chains of causation’ (cf. also Blaikie 1985). Like these seminal studies, much political ecology has concentrated on analysis of regional political economy issues at the expense of a rather scanty, or mechanistic, treatment of the ways that local social relations interact with them. Other studies have, however, valuably

focused on gender and other localised struggles around environmental resources in the context of these broader changes (e.g. Rocheleau 1992; Mackenzie 1990, 1991; Carney 1993; Leach 1992). In the case of Kissidougou, the past century has seen some major changes in the wider political economy. These include the effects of three severe political regimes: the rise and decline of Samori Touré’s Dioula state (1860s—1893 in Kissidougou), the French colonial administration (1893-1958), and the state socialist ‘revolutionary’ First Republic under Sékou Touré (1958-84). They

include some major economic changes: for instance the early use of Kissidougou as a labour reserve during nineteenth-century slavery and early colonial rule, and the rise and fall of rubber, oil palm product and coffee export economies. In some periods populations have increased, but in others they have declined due to war and economic or political outmigration, and seasonal migration patterns have altered. Our historical analysis will trace how these changes have articulated with local social relations in influencing land use, and in the making of the forest-savanna landscape. The effects of environmental policy are themselves an important influence, since the environmental service has been one of the more penetrative arms of state intervention in Guinée’s rural areas during the past century. Studies in environmental history have examined the history of colonial and post-colonial environmental controls and regulations (e.g. Millington 1987, Beinart 1939, Diallo 1989), and some have examined these in interaction with local land-use

practices (e.g. Tiffen et al. 1993, Amanor 1994). But in following this analytical tradition, caution is needed in ascribing too great or direct a role to policy in the making of the African landscape. The effects of environmental policies must be contextualised in relation not only to the many other processes involved in shaping the environment, but also to the ways that villagers have dealt with, resisted, subverted or lived with them. Our focus on the actual historical relations of socially differentiated land use practices leads us to criticise a number of generalised arguments concerning

Introduction 13 the way people impact on the environment. These include first, institutionaleconomic arguments based on the ‘tragedy of the commons’ (Hardin 1968); that unless resources are held within private or state property regimes, it is in every individual’s interest to use resources to the maximum, even though it may be in the ‘general interest’ not to, so that they will tend to be degraded. Second, they also include the general counter-arguments to this: that local institutions can effectively regulate, manage and sustain “common property’ resources, but that environmental degradation is likely to result if political change or emergent social differentiation undermine such common property resource management institutions (e.g. Bromley and Cernea 1989, Berkes 1989). A third argument holds that poverty leads to environmental degradation, by rendering people less able to make long-term ‘investments’ in resource conservation. Instead, preoccupied by the needs of immediate survival, they

engage in short-term, resource-degrading practices (e.g. Durning 1989, Leonard 1989). A fourth set of general arguments concerning the ‘people— environment nexus’ is demographic, with one (Malthusian) position holding that population increase outstrips the capacity of the resource base sustainably to meet requirements, and hence leads to degradation. This is countered by Boserupian positions (Boserup 1965) which suggest that innovation and social transformations might sustain the resource base in the face of population expansion. While evidence can be supplied from numerous studies to support each of these arguments, we find them unhelpful in describing the course of historical change in people’s land relations. This is first because the very multiplicity of resources which Kissidougou’s villagers use, and the complex and changing social configurations within which they have used them, mean that while elements of each hypothesis could be found valid in relation to particular resources and particular social groups, none serves to describe or explain the course of change as a whole. Second, these arguments — even in their opposition — all assume that the use of land and vegetation is degrading, in the absence of specific regulation, long-

term investment, or technology to make it otherwise. This is, indeed, a root assumption in western cosmology (Collingwood 1940); one related again, arguably, to Enlightenment notions of human society as separate from nature, as acting on it and, in doing so, degrading it. But while this may be the case, it cannot be assumed. If forest is considered to be ‘closer to nature’, for example,

the assumption that use degrades contradicts the Kuranko experience that untouched land tends to be savanna, and that where one cultivates, trees multiply. Different ecologies — interacting soils, vegetation and so on — respond

differently to use, and the same uses which do lead to ‘degradation’ in one place may lead to improvement in another. Thus ecology is a player; one which in local discourse is even personified at times, adopting different characters. It

14 Misreading the African landscape is necessary to examine precisely the sorts of practices that resource constraints or ‘individualistic’ uses lead to, and their impacts under particular ecological conditions. We will argue that in the ecological conditions of Kissidougou at

certain times, the use practices of some of the most resource-constrained, ‘short-termist’ farmers are leading, almost ‘inevitably’, to increases in land productivity and biomass.

Power and representation in landscape and its history As we came to comprehend villagers’ readings of the forest-savanna landscape, and the sense these made of the historical evidence of vegetation change, their contrast with the readings prevalent in policy circles forced us to consider the latter more critically. While we had originally treated scientific and policy knowledge broadly as ‘empirical reality’, grounded in a large body of scientific literature, we now had to see it as emanating from a particular social context. As will become clear, the ways that views of degradation in Kissidougou have

been formulated are intimately linked not only to the intellectual structures of science, but also to particular institutional and financial structures, and particular social relations. The ‘truth’ about savannisation, produced and diffused through science and its methods, deduction, and language needed to be considered in relation to its political economy; originally within the colonial regime, but not now restricted to it. As such, the conviction of degradation in Kissidougou, and its power in excluding alternative landscape readings, may be considered as “a discourse’ in the Foucauldian sense (Foucault 1972, 1980). Furthermore given the stable and absolute conviction in degradation among powerful state and global institutions, and given its proven capacity to silence

and subjugate inhabitants’ views within policy and scientific circles, this landscape reading and its scientific elaboration might be taken to exemplify what Peet and Watts term a ‘hegemonic’ or ‘totalising’ discourse (Peet and Watts 1993). A central aim of this book is to recover the landscape readings of Kissidougou’s inhabitants, and to put them into wider circulation.

In establishing competing claims to truth about forest-savanna dynamics and landscape history, and in carrying very different implications for the contro] of rural resources in their links to particular institutions, inhabitants’ landscape readings might also be considered as part of (a very different) discourse. Yet during investigation we heard less a single truth than contested opinion. Among the prefecture’s Kissi- and Kuranko-speaking farmers, there is a plethora of ways of representing landscape history and dynamics which can

vary in their implications for different people: for example for the resource control and claims of different women and men, and of longstanding citizens

or recently arrived strangers. It would thus be wrong for us to portray a homogeneous ‘local’ perspective on landscape and its making. Rather, we

Introduction 15 attempt to give a sense of the terms, operation and discursive structuring of local debate.

In addressing landscape representation in this way, this book joins a theoretical alignment across a number of areas of social science. Cultural geographers have recently retheorised ‘landscapes’ as not only the material outcomes of historical interactions between society and environment — a longstanding focus of geographical work (e.g. Sauer 1925) — but also as open to — in a sense constituted by — diverse interpretations. Often treating landscape metaphorically as ‘text’ or ‘spectacle’, this geographical work examines how its reading or viewing is embedded within discursive fields linked to particular institutions, and how these, in turn help to shape socio-cultural processes (e.g.

Duncan and Ley 1993, Cosgrove 1984). In social anthropological inquiry, elucidating particular local representations of landscape has long been an issue, although often tangentially when considering other issues (e.g. Richards 1939). More recent studies have, however, focused directly on ecology and landscape, and dovetail with those of geographers in considering how representations are the effects of discursive formations and are deployed in political relations (e.g. Bender 1993, Croll and Parkin 1992, Milton 1993). Detailing the material aspects of environmental history provides evidence against which to assess diverse representations of landscape. Nevertheless, it is also fundamental for comprehending inhabitants’ landscape readings, since it is in part these changing material conditions and responses of landscape to use

which villagers reflect on. Put another way, the relations of production of ‘knowledge’ and local discourse concerning Kissidougou’s landscape are rooted in lived history and its experience. Such experiential and historical issues are all too often overlooked by ‘discourse’ analyses in their focus on how things are represented. This basis in lived historical experience is, furthermore, a key feature differentiating villagers’ representations of landscape history from the discourse of scientists and policy-makers. For in constituting the problematic of forest-savanna transition, natural scientists have formulated historical scenarios on the basis of deductive methods, without using historical methods or data sets. The essence of this study is the challenge which ecological history and local experience and comprehension

of it make to these ‘scientific’ meanings. As a landscape, Kissidougou’s forest-savanna mosaic has been read backwards: reversing this reading will simultaneously suggest reversals of power.

Describing and representing landscape history It will not have escaped the reader’s notice that the way we have formulated our problematic raises major methodological problems. On the one hand we are dealing with landscape and its history as representation, but on the other hand

16 Misreading the African landscape we are attempting to reveal its empirical ‘reality’, facts or events. Very often it

is the same sources and data sets which must be used for both. Clearly, in asserting or adjudicating between historical accounts, we inevitably enter into the relations of power surrounding these alternative histories; a positionality which we are well aware of, cannot avoid and make no apology for. A second

and related impasse is that on the one hand we are seeking to consider ecological processes from multiple and comparative perspectives, attempting to escape the centrisms and straitjackets of western scientific viewpoints, and arguing that all concepts — whether in western science or as used by local people — are socially and discursively grounded. Yet on the other hand, this leaves us no ‘neutral language’ to describe ‘real events’; for considering how the landscape ‘actually’ responds to use. Furthermore in addressing a discourse

and its problems one is inevitably forced to enter its assumptions to some extent. Thus the concepts and categories of ecological science, for example, inevitably become part of our language even as we recognise and attempt to show their inadequacy. It is hard to avoid becoming trapped in a language of response, especially if we want to address readers in a language with which they are familiar.

Similar problems have been recognised in other recent studies which examine social and environmental history as both event and representation. In particular, Moore and Vaughan (1994) faced this issue head-on in their analysis of the Zambian citemene system, recognising, for example, that: We have . .. had to contend with the problem which besets any researcher influenced by social constructionism in its ‘weaker’ form, which is that we must simultaneously treat accounts as though they were both factual and simultaneously constructed ... We have to recognise that we simultaneously use accounts of all kinds, whether from the

past or the present, both as representations and as data. There is no escape from the unease which this dualism produces. (Moore and Vaughan 1994: xxiii—xxiv)

One response to this inescapable unease is to accumulate multiple sources, and

multiple perspectives, on any given issue, comparing them for their mutual support or inconsistencies, and interrogating them iteratively for the clues to ‘facts’ or representation which can be discerned. The sources we have used in this study are very diverse. They include archival documents, the writings of colonial anthropologists and Guinean scholars, policy documents and reports, old and new aerial photographs and maps, oral accounts, interview data in villages and policy circles, participant observation, and village resource and vegetation surveys. They also include scientific works in ecology, forestry, botany, soil science, hydrology and climatology. Such an assortment is rarely treated together in the same field of critical inquiry, where each reflects on the other, often very uneasily. From whatever source, we read accounts both ‘in’ and ‘out’ of their contexts,

Introduction 17 in a way similar to Moore and Vaughan who describe how ‘To explore the degree to which these representations are representations, we have often had to displace them from their context, from the naturalised parameters of meaning within which they make sense, and to read against the grain of their intentions’ (1994: xxii). This is what we do for written colonial accounts, for example, sometimes reading them for what they say about representations of landscape and people at the time — for their place in discourse — and sometimes seeking

in them the data about past landscape which they provide. Reflect, for example, on the following account of a Kissidougou administrator in 1914, which provides information both on his interpretation of forest degradation, and on the actual state of contemporary vegetation: Never, I believe, has a year so dry occurred in Kissidougou. I can only say that from year to year, rain becomes rarer and rarer. And this I do not find extraordinary — even the contrary would astonish me — given the considerable and even total deforestation in certain parts of this region. From Kissidougou to Gueckedou, all has been cut. . . the effects of this de-wooding are disastrous; one will soon see nothing more than entirely naked blocks of granite. A region so fertile become a complete desert. Now there rests no more than a little belt of trees around each village and that is all.®

We attempt to treat modern scientific accounts from various disciplines in a similar way, interrogating them at times for what they reveal about scientific categories and assumptions in representing forest-savanna dynamics, and at

others, for what they might imply about the detail of soil and vegetation processes. In the latter respect, to understand the transition landscape, we have had to consider the specific effects of soil-water relations, termites, fire, and animal grazing, for example, on vegetation associations. In general we have tended to prioritise those scientific analyses which prove to match both local representations and observed vegetation dynamics.

Both multiple sources, and reading them ‘in and out of context’, were important in facing this study’s most basic methodological dilemma: in having

confidence in the surprising landscape event history it revealed, which so sharply contradicted ‘conventional’ views. Surprise itself was, however, an advantage. That we so little expected to find such a history of forest gain gave

it, in a sense, a greater plausibility (it could not be a figment of our own prejudged expectations). And we became increasingly convinced of it as initial

premises from one set of sources proved, apparently against the odds, to be confirmed by others, and then others again. A central element of this study has been the classical social anthropological method of participant observation, residing in a village and learning subtleties of local language and practice. In this case both of us, working with our two

Guinean colleagues, lived for a total of 12 months over a 21-month period in two villages where different languages, Kissi and Kuranko, were spoken,

18 Misreading the African landscape oscillating between these. We find such immersion essential for comprehending local concepts, their discursive production, and the social relations of land use. Living and working with villagers was especially necessary given the extent to which their representations grapple with experience and practice as well as verbal discourse (cf. Moore and Vaughan 1994, Dupré 1991).

Of course, as outsiders, there are large areas of life of which we remain ignorant or have only passing comprehension. We can make no claim whatsoever to a fully authoritative and comprehensive analysis of social life in this

region. We have tried to gain sufficient familiarity with issues relevant to ecology, but even in this, we inevitably confront the limits of ‘knowing’ imposed on us by villagers themselves, not least through the region’s gender-

specific power associations which are structured around hierarchies of withheld knowledge and comprehension (cf. Bledsoe 1984, Bellman 1984). As is sung at Kissi girls’ initiation (Ifono 1989): Some know, the others don’t. It’s necessary to suffer to know — Our parents know, But the non-initiates believe in simple folklore.

Folklore, in this case, can be taken to refer to the tales and myths related to non-initiates, describing particular prohibitions and accounting for them in mythological terms without revealing other significances. As researchers, the way villagers located us in their politics of knowledge has certainly circumscribed the information and understandings accessible to us. Furthermore we cannot know the extent of these limits, since withheld knowledge of the extent of ignorance is integral to the politics of secrecy. Two detailed village studies were, nevertheless, insufficient. Kissidougou prefecture is extremely diverse, ecologically, socially and historically, and a

consideration of both landscape ‘event’ history and local discourses representing it required this to be taken into account. Highly aware of the danger both of unwarranted generalisations about ‘Kissi’ or ‘Kuranko’ land use and understandings from a single village, and of the reverse danger of compiling a generalised (and ethnicised) picture from diverse localities, we have attempted to examine a range of localities ‘on their own terms’. To this end, we visited six other localities, variously Kissi-, Kuranko- or Lele-speaking, which are situated in different parts of the forest-savanna transition around the prefecture (figure 0.2). Our colleagues or ourselves stayed in one village of each locality for several weeks, working there and in neighbouring settlements to discuss land use and farming, and to record resource use patterns and oral accounts of landscape history. Many of these discussions were explicitly comparative. For

instance, as we learned about particular soil or vegetation concepts in one locality, we raised them again in another. Such intra-locality comparison was

Introduction 19 Yomadou Merna Rn

|

Albadaria Kamerendou SEE SEO Bambass; Niand== ambassiria

Sangardo

Yardo

Wossokoroma

Yalakoro Manfran

Kenema — Gbangbadou

Bomba

Re poss Wy “4 : Beonoe Soni. eeyDembayara Dandou Yomadou _ satatat atte Massa- re omara

“= ©

koundou Gbang. Ouéndekere

Beind badou 8 Yombiro > wine"

4S Bagbe fo if Gban=|Trakore /Gben-F ssadot Bardou / Banama A ikorogadou Fermess¢ Wondeto {_adou Bah Mantrania fos

Wondeto

Koundiadou Bongoro

E53 Bamba) AEE

yh Balansama

. EY Soulakolo

—— Major road

“District boundary

Ld

@ = Kissidougou urban centre 0 10 kilometres Research sites:

1 In-depth study village (Sandaya); air photograph/oral history analysis; Kuranko 2 In-depth study village (Toly); air photograph/oral history analysis; Kissi 3 Comparative study and air photograph analysis; Kuranko 4 Comparative study and air photograph analysis; Kissi 5 Comparative study and air photograph analysis; Kissi 6 Comparative study; Kuranko 7 Comparative study; Kissi 8 Comparative study; Lele

Figure 0.2: Location of research sites in Kissidougou prefecture

indeed quite normal to villagers who have frequently travelled and moved residence several times in their lives. Indeed in many contexts people’s own representations of landscape take an explicitly locally comparative form.

20 Misreading the African landscape The emphasis on critical anthropological perspectives and methods in this study has not been confined to village-level issues. Anthropology has been as important in the analysis of science and policy as constituted within discourse, as it has been in the village-level analysis. As others have recently emphasised,

the ‘worlds’ of environmentalists and development policy-makers and agencies can be legitimately and usefully treated in anthropological inquiry (e.g. Milton 1993, Ferguson 1990); where this study perhaps goes further is in taking this approach to ecological science itself. In this respect, we treat the diverse ‘texts’ — written, verbal and visual — of scientists and policy-makers as part of the ethnographic data set. To a certain extent, we were able to be

participant-observers in policy, development and scientific circles. The ‘applied’ aspects of the research, aimed at rendering the results more useful, led

us into meetings, conferences and field visits from which we learned a great deal about the views of today’s environment and development personnel in Guinée, and the social and authority relations within which they live and work. Understandably, our relationships with these institutions were at times uneasy, as we exposed ‘alternative’ landscape histories which raised uncomfortable political issues and were sometimes found threatening. Guinée is presently very welcoming to foreign researchers such as ourselves, but this has not always been the case. A colonial anthropology did develop in Kissidougou, most notably through the work of Denise Paulme whose Les gens du riz (1954) addressed many issues of concern to this study. While appreciating the detail of Paulme’s work, we have, like many subsequent Guinean scholars, found aspects of her account problematic, and hence we treat it here as one among many other representations of Kissi life and landscape. Guinée is unusual for the rarity of post-colonial foreign scholarship, a result of the political and ideological isolationist conditions of Sékou Touré’s First Republic, which permitted only a few notable exceptions (e.g. Riviere 1971, Person 1975, Suret-Canale 1970). The 1958—84 period nevertheless saw much academic work in Guinée’s national universities, in particular in the student theses which we draw on here both for their specific accounts of localised historical events, and to see the ways these have been conditioned by colonial legacies and then by revolutionary ideologies. Since 1984, amid an influx of foreign investment, aid and agencies, the relationship between national and foreign research has shifted once again. While we like to see this work as a social anthropological study, then, other social anthropologists and ecologists may disagree. It is, however, also a study in environmental history. At least at one level, it engages directly with the substantive and methodological programme of those historians now focusing on environmental issues (e.g. Worster 1988). This is neither a recent nor a disciplinary prerogative. Travellers, administrators, foresters, geographers and others have been ‘doing environmental history’ in Africa for decades, if

Introduction 21 not centuries, and it has remained a strong strand of concern in other disciplines.

The structure of this book This book builds up through ten successive chapters an understanding of Kissidougou’s landscape and what people have made of it. Chapters deal to different degrees with policy and local discourses; with landscape history in terms of representation and of events; and with the layered village, locality and regional analyses which are required to address the landscape’s responses to historically changing use. Chapter 1 introduces the landscape in the terms of policy discourse: of derived savanna and ongoing deforestation. It investigates how administrators, policy-makers and development agencies presently express their conviction concerning this environmental change, consider its social and demographic

causes, and translate it into imperatives for intervention in modern Kissidougou. These convictions are conveyed in scientific texts, policy documents and everyday discourse. We examine concepts and terms of debate

concerning forest-savanna dynamics in the wider West African region, to explore the contribution of discourse in ecological science to the discourse of degradation in Kissidougou. Historical data presented in chapter 2 fundamentally challenge this reading of the landscape. We show how air photographic, archival and oral sources from across the prefecture build up into alternative — and sometimes directly reversed — landscape histories. We present much of the evidence which, as we gathered it, threw into sharp focus the contradiction between policy narratives and demonstrable pasts in this part of the forest-savanna mosaic. Alternative landscape readings inform the next five chapters, which focus on villagers’ own experiences and discourses. Here, several temporal and spatial scales overlap and build on each other. Initially, we consider forest islands in relation to long-term, regional patterns of settlement, population movement and political culture. Then we enter the two main study villages to examine how farmers’ concepts and explanations of vegetation, soil, water and plant health issues are embedded in everyday experiences of landscape in farming, collecting, hunting, and making a living, as well as in local political and social struggles. We explore how landscape change during villagers’ lifetimes has related to changes in social and economic relations within households, families and the village. While these two parallel chapters address similar issues, they do so in very different contexts, both social and environmental. Sandaya, a

Kuranko-speaking village in north-west Kissidougou, has experienced a persistent mosaic of forest islands in savanna, with some increases in upland savanna tree cover since earlier this century. Toly, a Kissi-speaking village

22 Misreading the African landscape in the south-east, has experienced a generalised transformation of upland

savannas to forest fallow since the 1950s. Ways of working with and manipulating ecological processes, and enriching the landscape with vegetation species and forms of perceived value, are basic to local experiences and discourses. In chapter 6, we compare all eight study localities to consider these everyday resource priorities, and detail the numerous ways that villagers work with ecological processes to influence forest-savanna dynamics. While scientists

and policy-makers have failed to recognise or appreciate many of these possibilities, they make use of processes for which ecological scientific support can, in fact, be found. These practices are important in accounting for demon-

strable patterns of vegetation change, but trajectories of vegetation history depend on changes in their application under shifting economic, political and demographic circumstances. Chapter 7 therefore addresses vegetation trajectories during the present century and some possible ways of accounting for them, focusing on land use domains of particular importance in altering upland vegetation: forest island management, farming, the cattle economy and fire control. It examines changes in these domains in terms of the articulation

of local social and economic relations with wider political, economic and demographic changes at regional, national and international levels, addressing the varied outcomes in different localities.

Chapters 8 and 9 switch back to focus on Scientists’ and policy-makers’ landscape readings, addressing more fully their discursive formation and instrumental effects. Chronologically, chapter 8 considers the elaboration of deforestation visions over a century in colonial science, and through their institutionalisation within Guinée’s environmental services, their persistence post-Independence. It documents the repertoire of policies and interventions in

agriculture, forestry and rural development, and examines their material impacts ~— and the resistances provoked — among Kissidougou’s rural populations. We examine how these planned changes based on a backwards reading of forest history influenced Kissidougou’s landscape, but consider how their material effects were circumscribed by the ways they interacted with other social and economic changes, and with villagers’ own practices. Chapter 9 addresses the production of policy-makers’ landscape readings,

and why they have remained so persistent and resistant to challenge. It examines the intellectual, social, political and financial conditions which have mutually supported each other in upholding this powerful backwards reading

of history, and the ways that aspects of both global development concern and local political process have been incorporated into it. It is these political relations, harnessed to the structures of ecological science, we suggest, which

give the savannisation discourse its totalism and hegemonic capacity. In chapter 10, then, we conclude by suggesting how and why a more pluralistic

Introduction 23 approach to ecology is a necessary — although not sufficient — condition for a recovery of local, more experience-grounded landscape histories. Putting them into wider circulation could, we suggest, have major implications for the way forest-savanna dynamics are understood throughout the West African zone, for understandings of history in other African environments, and for the science of ecology in general.

1 Convictions of forest loss in policy and ecological science

Visions of forest loss Guinée’s present national forestry plan suggests that Kissidougou’s landscape is degrading fast: The opinion, quasi-general, is that ... the areas .. . north of Macenta, Gueckedou, Kissidougou will soon be no more than a vast poor savanna, the [forest] islands and gallery forests still present at risk of being rapidly destroyed. (République de Guinée 1988: 31)

A French forestry advisory team describes the forest which policy-makers consider to have been lost:

The region of Kissidougou was covered by a deciduous forest of Khaya sp., Chlorophora sp., Antiaris africana, Afzelia africana, Ceiba pentandra, Triplochiton scleroxylion . . . (Estéve et al. 1989: 181)

Many modern studies which have informed environmental and rural development projects in the region think that this extensive forest cover has been lost within the past 50 years; within the lifetime of the region’s present inhabitants. Thus we read that: Around 1945, the forest, according to the elders, reached a limit 30 km north of Kissidougou town. Today, its northern limit is found at the level of GueckedouMacenta, thus having retreated about 100 km . . . This deforestation is essentially the result of human action. (Ponsart-Dureau 1986: 9-10) Or: In the green belts which surround the villages, one finds the relics of original primary forests. The value of these biotypes in the heart of a nearly 100% degraded environment is inestimable. One finds no individual of [characteristic savanna tree species] more than 35 years old . . . supporting the thesis that the site has burned systematically only since then. (Green 1991: 10—11)

Fortunately, one can consult the air photographs and archives left by the colonial administration. As the next chapter will show in more detail, 24

Forest loss in policy and ecological science 25 the vegetation in the 1940s and 1950s seems not to have been the forest that these modern observers think existed, but consisted of: ‘oases of equatorial vegetation in the middle of savannas burned by the sun and fire . . . all in

regression’ (Adam 1948: 22). Adam, the botanist then responsible for Kissidougou’s forest policy, later remarked that: The information obtained from the oldest inhabitants confirms what we supposed. The whole region was covered with forests around 75 years ago. (Adam 1968: 926)

Seventy-five years from 1968 takes us back to 1893, the date of the first French occupation and its first reports. Kissidougou’s military conquerors did not find

the extensive forests which Adam supposed to have existed then. During Combes’ invasion, for example, he described his troops’ marches through tall grassy savannas, the bush fires that threatened him, and a landscape: ... dotted with clumps of forest of great vegetational strength, rendered impenetrable by a tight confusion of trees and interlaced creepers . . . It is at the centre of these clumps, in the middle of large clearings . . . that one finds the villages.!

Kissidougou’s first military resident in 1893 again presumed that extensive forest cover had once existed, and set its date back further: The soil of the valleys has a more or less thick humus bed. This humus derives, as one understands it, from the immense forests which covered a large part of the soil, and which covered it entirely at an epoque relatively little distant from our own.?

At a first approximation, it would seem that all administrators and environmental policy-makers for a century have been looking at a similar landscape,

but thinking it to be the relic of a quite different one, suffering past and ongoing deforestation. Most, like these authors, consider its forest loss to have occurred within the lifetimes and through the activities of present inhabitants.

Others are more cautious about the timescale, considering, like the botanist Schnell did, that Kissidougou may have been ‘deforested early’ (Schnell 1952), but are equally convinced of past and ongoing forest loss and its anthropogenic causes.

Deforestation and the intervention imperative Agencies concerned with environment and development within Guinea and Kissidougou define and justify their policies and activities according to this landscape reading. Indeed as a part of West Africa’s forest-savanna transition zone where deforestation and savannisation problems are thought to be particularly pressing, Kissidougou attracts considerable international and national attention and funding. A strong consensus exists as to what has happened to the landscape, why this

26 Misreading the African landscape landscape change matters, the role of local inhabitants in it, what this says about the nature and evolution of rural society, and what needs to be done about

it. In the first part of this chapter we examine these views, not only in policy documents but also in ‘informed’ opinion; within national environmental institutions and the international donor community on the one hand and, on the other, among Kissidougou’s general and environmental administrators, project staff and urban-based intelligentsia. After flying south from Kissidougou to Nzerekore in 1993, Kissidougou’s senior administrator, the préfet, was appalled at the severity of deforestation that he saw. Addressing a meeting of projects and traders in preparation for some upcoming ‘Environment Awareness Days’ he asserted that ‘as part of Guinée’s “forest region”, Kissidougou and Gueckedou are no longer worthy

of the name’. During these days, intended to reflect on the Prefecture’s environmental problems and potential solutions, one of the sub-préfets suggested that: ‘if one travels towards Gueckedou one now sees just bare hill tops — all has been felled’. The recently formed national environmental non-governmental organisation

(NGO), ‘Friends of Nature’, based in Kissidougou and run by schoolteachers, helped organise these environmental days. The organisation’s statute claims that ‘our environment continues to degrade because of over-exploitation of

forests and bush fire setting .. . This natural capital, which we continue to exhaust through scandalous management, needs to be protected for the well-being of humanity.’ During the meeting, its representatives referred to Kissidougou’s present environment as ‘this desert of today’. This view is unquestioned throughout the education service. It is taught in

primary school classes, and is invariably the one which appears in undergraduate dissertations. In one, for example, which focused specifically on forests in Kissidougou, the author forces the reader ‘to note that the forest of Kissidougou, once extensive, has been strongly degraded by peasant practices ... Today the largest part of the region is only savanna covered in high grasses’ (Sangare 1983: 14). Multiplied in the ‘geographical’ section which appears in all prefecture-focused dissertations of whatever discipline, this reading seems as true and undeniable as Kissidougou’s latitude and longitude. Before starting the research, one of our team members had himself written in his philosophy dissertation that: “The country is characterised by dense forest ... All the same,

today, under the effect of bush fires, loggers and shifting cultivation, this forest has a greater and greater tendency to disappear, ceding place to new vegetation forms of a wooded savanna type’ (Millimouno 1991: 7). The same view has been held by the prefectoral forestry administration, and by donor funded projects working with forestry staff in and around the prefecture. As a report informing the environmental components of the prefectoral integrated rural development project put it:

+ s a LZ

‘i

Forest loss in policy and ecological science 27

“oss % ©. . 3 Bho OR eS

&Seeai—‘ ernea “ oeeeee rrr LL rrr CC ee AOE 8 eS ee ee ee ee ee Pee ee ue a ee ee ee Ce Ce er ee ees ee ee feeoe SR ee eh ae ee ce i ee : eC OSaea.ee ee

eo neEE oar: oe,Cehh oeRee ¥ IS eSan Sage IRMm ae oe we a F Piety TEE rrr i(kCOE gs a et ERPee ee .isa:BE SE SS RE CeSeeeeeeee ete ee ee eeeEr ce fe EER ee oe ae pee r—rr—“ —r—SCa“i Gee ee Rar OOS so SEE SS ri—“i—~w”~—~—~—~—~—~——~—C”—COCOCOCOdCtéwsC™Cisée es r—“‘i—i—S—S—SC—S—O—sSOSesSSséseCC a — —Crtri“‘ (Ore ;

wh

é le - to

\ . 40 ten of cA, aes SS 7 ww

ya “fF - f ®&,— . ae .aN KR ° OS oe Se" VA aa.*;.it sf 1 ws = °: ~ a 8, " = id 8 $ are =: ca: oem z* |>oh os r , q | ~ = ~ j ° Se ~ = as, a4 oe SY , = = > > a8) ee


aspo} ;vas Ss ageeerre . = aa > wae tye i ' + aa _ Ses SG 82 aor —_— ,

: -eTer) : 3 Pree a m4 aek ° . '¥E ‘eae & opt E woStC

wi IN p ‘ o7 nal Na = ' =4 =: =5 :—%2~ -« “er ——f-0% < —— = >=—:: ; = im =| cS =| a. y= raed =< iais os “soe ~ Wy APRS es :ne 223 oy“™TN ~i rean fH(== ~ ja@° 5poOb oOA>) ae = PATE NS MOR Cr oman ne i S S os aePSN ate a_4 Pa £P>7 abe £ =“. © ore ay NI ~ ie is 7 TAT io Ooiuw v rakewe wali Sar a4= :=om S38 aon a C3 7 8 “1 eeread & Fant "S @ i % ae : F ve a ee a aaa Per? ; XA a & i, Dy ze + : = id x_ SE Ory Fr. v3 : Ce is tinea volS the .ne Par323 2= 3¢ NA Se er ' Ne j 2 i S42 Fe ~ re : ~ d BS —— = : 3 s > “ae . eeeesy) Ln feyzoar: yi qoeee cee ror me a 7a Me yyISWare oe 4z We yuad eee tra ™ A + Sa ° ia — ay “r etinare iL, to 3e Ee — 2:: sy wat

Z& = | : BS sé a . 8 —E| | ag bb : 7 S-"

ae)

& os

=

: | 1 £3 _ — | | Ze 9 © : )2PR. |2£833 | 2gma Bo is 2) “ See Rol =

Rat uymn es 3s48gas Bee 8/ Qy eisSaS aS §83 x Ss © 5sa ee R ES REELS 7 Ht 'S) a= ESS 3 33 Sy =~ 3 3 S ns see 8& Sgge eas 8 Ss SEReS = S > aw) :::Sear 7 Re 2 5 e288 = i aHI = a, See U ae eth Saat s : , sees 8 § S32 8 80 — ~— wz igs ss Prose § Ss : |S=38Sy ~~e8gSE ae.

O one i |ao g32 % SS ;oO)23% Ze S |,fn 23% 2% . ;7os) :i}>,ES3|:& 2 rE

a +38 | InL HEae : co os 5 83 ”Y

Say 223

»g=2

E §385

d » = . Rion 5ann 83 5 2 s£ Bp 9 ons * 5.

5 &S : 2& a 82 oO SsSS 6= > On & 28a HmH & x5 nBas 2 &5g— csfe 2“ 22895» 88 257 ESenors a a 55 Ses 5e2e 4GO Se S &opSs 6 & aeo >ao. 2& 3evoF3c a=) 2 22O'y KeOo oN& w os oa > Y2 = Ee VUES ore 2 in) so A (SQ: . .EU oeeoOE: a., —5=

3 32io3a=3 = ra) oD

oma oO fa)

5 := 3E cE. E 5 r¢ wy

~ Sti 28 S¥oS se SSy Sess Ss8 E3=::=:: S Sy 83 SES FSSES

se ERQSPY SEEPS SE& SSSEL sss SS ose BS S 8Ss, on= =S5

eSeeh8SIaS S8ke8 SPRL S SE S, §0&F S Soo So Se=Sse 8S SR SESE SS3: > s§ 38 8 & 8 8 § FS §

SSS esYXsgs 2 SES si 28 FSSER SRSEER S 8 eee ceessz, SS = ss 5 < 8g8Bt S % 5 8 Sa Sas .wan & s § S335 sé S BD SS§+«=Se28 3.8 SF 8 oe§3 7] & wo

eet

827=) =&38 2=a 3SO O§& =a525 ns ESE a)

60 Misreading the African landscape Table 2.2. Areas of forest and secondary forest thicket vegetation in 1952, 1982, 1989/92 in five localities

Area Percentage analysed Forest area Forestarea Forest area increase

Locality (hectares) 1952 (ha) 1981 (ha) 1989/92 (ha) 1952-92

Moria (locality 3) 9,500 448 562 633 41*

Sandaya (locality 1) 7,000 512 778 900 76*

Fondambadou (locality 4) 3,525 698 654 1,064 52 Bamba (locality 5) 5,400 634 no data 1,411 123

Toly (locality 2) 4,550 708 no data 4,288 506 *ambiguous; possibly an artefact of differences in SPOT and air photograph resolution.

Kissidougou has certainly not diminished in an alarming fashion, but on the contrary, one can believe that it has tended to increase over the last 40 years. (Scheepmans et al. 1993: 71)

While it was impossible to subject the entire prefecture to full stereoscopic and

quantitative analysis, we carried this out in five localities, which we also visited for fieldwork and ground-truthing. These localities, indicated in the map of study zones in the introduction (figure 0.2), were chosen to capture differences in the prefecture’s sub-ecologies: principally the broad north-west to south-east spatial gradient of increasing forest density. Figure 2.2 and plates 2.2 and 2.3 illustrate changes in forest and savanna cover between 1952 and the present in these localities. Overall changes in forest area are summarised in table 2.2. Further selected photographic comparisons from different parts of the prefecture are shown in plates 2.4, 2.5 and 2.6. Various sorts of longer-term change are evident from these comparisons of photographic and satellite images, taking into account the cyclical vegetation changes related to the dynamics of farm—fallow cycles. In all these localities,

while there is a certain stability in the forest islands around villages, it is evident from the photographs that this is dynamic. In many cases forest islands have expanded. This is common, for example, for the villages around Bamba in the south-west and Fondambadou in the east. For the 17 villages covered by 1991 air photographs in the Niger protection programme Kiss 2 project zone, the area covered by forest islands has expanded by 30% since 1952. In certain

instances, entirely new forest islands have appeared: for example in the Sandaya locality in the north-west a nascent forest island is developing around a new village site. In the example from the district of Bambassiria (plate 2.4), a Strip of gallery forest in 1952 has become, by 1991, a forest island completely encircling a village.

Nevertheless, there is also some evidence of forest island diminution or

O i |idence SEs SS ORR ORC

F t gain: historical ev LJ

.

: Pea mree Soe meee Bernences , +Sar ESEnyeee are eens ERE See aues eaeeae eeeOREN soci eee Beg Saga . a 2 SA SESE SEneDESF 2 ky Sty =coe REreer a aprereeas * Pepe nnees: ea 3See One 5 sR

OSs See LE BAe oeee yeGEO “sees BeSee es CO ae esete Jeeste area 6wee Cee ee lL| rr LLCC —— Gar eSSS so cat eaeoe eS aioe eni oh giceam SER ea , ns oaaSs —lrcttr”r—sS— Cr *«wo Sy, “PORES eee eeRas SORE ERR Ste ES Nees a ee Seog Bae ee eeSee SR OBER, mc . : 7 ERS PERS Le fhe el SoHE Eee eee Sin Saasae oN a ee oa rears RE re aes OR ate Mgt eES ope SO Se ORRieee ARR RS ue a pase!ese eee ee ae ane Seae ea feee 2 7 > eeoh. ees SER ea RRRSR Garena. sSemeeree | a, | ve ieon er S ec ae SaSomes ETS aa Been ae eSee Se ae a Reo sok Taat RE a PRR Sees Sia kocamo eR. Bo So eSSee: oo_

es: re : BEE “A ES ee— EC eae SECs een Se eee eeree Re RES So RE prot eens awey (yA EE RE rrr—“_“_™OCOCOCO | Pou oe Re Sar ER ee EE TOES i Rss cSees eeCOE iBone ee Soc alSUE Ss een | voo RE es ee eee eaSean Rep hee ciSR Se aRS SIRE SNeee aeRee RR SN Dee aROR gates etSEE conn oeOE aDIE SRRE RLBeas poACS Ree odo OREae SERS sgaicgn eA eee SRO i RR aise eeee seeeSe Serer Saeteees Reet | Bere Si yee RSSea eR Ske SRE See Sa eso BeBa OStc eae SEae eeSee

re eySe Cee ee eeOr essreSee asaS Sree a arenRSE, es Ree EGBS POR Le ORtsEe BePee EEE geRSE ee psig aneerie BO es BeSES or ee Deena RES a “BSB BOesSee UBeRE ereperi SS Meee Se ees.gaRS eeeSeaes Nemeas Re Rue

fas Boe geiePUN EAL SLE aee,ae SpoOER Geaeern aeeee Meee tea oe SaSRS “CERES Ss Sree REE SER CRE Beasen See: ne aia —xoe --Pos iE Se pe nae Se etieMesicn 3 mae ss Soper. RIED ROE SS agsGR Sraeeenne e: BOER SES Scene Wetec ss eS SR ee ee Es eo ;° :azsne :. ‘DSS .;we eS SE Pee By, PRR RES SSIES: eeeeRe Spee aeEe REE Siete Yoon Pahoe SR Pee CRI os seeaatitenne Bees SOAS {See etepee teeR aera 28 ne GOSS. SSeS oeny ReOER IES fear RaeRO SS Eage eRote: SteeSES So AR a ee.. PS oo EERS BRO ESae ee soca sur. Stk PRG Re SeBatten SeSee aaoo See o GUS Ee ROS SOA Cy CR eee A OES ES 2 Rea SRE SOR -siaeee 2has [seus pices ocean aie eS Se neesSecs ‘ SOREN RRO AR ser3 ene eer ee BESS Pe Og an ey bs 3 teeTEESE TER Ee aatier eeSere ees ee RSceeeeBipire aeeeedaeSee Shin SOaeaN etre nee Geert aneeeSSRN wag See aEEeRE Oaacon ee ee

oedS 2s.— o .ee _. Os : - ee ae TB ~ pasvae sates eS = Pai LOPE ESese REPO Pee GEE weeEN re ES 2 eeooeeSeeeoeSee

ee EO ee ee pete BORE aeeeeeee ee es amet oe| — vee seesis ea RES ce Sees RaaASE GRReSeeeeaeRS . oe @ 2Pane, EyeUSE feBuea Sh LS SPSS SEee Repaieu ES ee ee ee etn ake pees es aeBee Se EI anne es es PE “re” ere PS oeEe SEES eSeu 8‘ Sees renee Pekan Saeeseat aeeaee] ee ac=pak ig See geri ee ant era feet st SEER ESRee eee ee Lo eee SOYee ee Soe As Sees ES SES Rte eaeaSe SOR Us RE eeos ae mene ae hee SURES “Se SySR es ee eeeSERRE ate ee: SPR SESE, eRe eno a rs ¥ Bak eee esePE SESeee ead eSIS eSAEE ae eraece “pe ES ee Seaes eeSRern seegSftSES Sone eeoo ae eseoae oe Re ee BER TSS: gOS

MES Pee ee Bieee eae cag Eos ria sy Sea eraser ee 2 i eee ee Se a eee oe ee ok. aOER _ eseeee eee eae+ Se tg ae Pan : eee os eeCUE — ee cee Seeae ae eae 2 FF Se ieeeeo PESfeOE ee ees a a. Se epee ee BO oe SO See ee 2 ee ce ae oo Se ereSRR Be Mie ene Sao ages omeperre oe RST OS OS eee ieee SRSISS Soha eeSR gee EE o¢ BSE EER eee Be aN noe See RN GRAS Reeeeee oeeecre aaage PEE esCierra SORE ee ag PULSES Bg eee RRC NeROE eens goer aaess aS Sa eC eS cohert aE ASS eneCCV Se SEY OS HSRe SEC Sn eyaes eee gees SSNS Reeey Fete eeeSE eg Shag eeeVES 7 heee TEELIELS TEs eeeaee Sscaters oaRAR Rese ees SNe eee SPE aoeBR SoA eeSaran Rabanne hi nea03 ee

Wagh ELSLE waSS IERE DeeaBEF a ee So Seeeaene eeeae SRE caren oaeoeae eae See eae EE pape sae er hors eS eee, ReOE rs Eee SER ae ea Pepe 5sae Se es SESE ood eeaaiene ae Bees See, Scabaeas sor Sertans EDNhee NAS eeeSe heeaea SO N eee Ee eR eed eT Se: Soe Se | UES, EE SES Remain aseSe eemiei Reece cuaES n pa enenee mame eae cree SeenSe eeeISSEES eo Wipe eet emis ESAOS SERN ecemrre RR ae ee

Se leeeSer cE Ee Speer ett SEE aren reer SAS) See akeRNR eR BP ae SE OOee BOhie asee eS eee SSDS Nee ee Rk enoneSES. tareBIN SP Se ase Si eee ReeeT a Rose, TaeSe akeeaRE ie as Ree

POOR ee OC Srnn Sep aeesad Se RR a ag TR ot neyee eee Be SiShae OFS Sees Ree See 7Seas oeBRE Sie Sa. Eee ers Re? SEE RSARS BEog sae Bo WY ERPe SSCS STE SES geEs ae ieee aF Hoes ae Sa SCR a Sn =ASE Sess eeSee aBE See “Oe GERR SERRE So aati Coe Re enate *aS ESE ceOe ees LEP ren gh ene aeSOS ee TA eae enna eset Re SSE Eee ES: SON Secret PSS ee oe ee eee fA ES Bee EROS, SN Seeae,SES Saree sean ee Sse ReNe eter nae een ase

Png? 2 Gan EE Se ee ee Sees BSoe BER a Seo aes REE . © ee Boree ROSe es eoee BOE Eos Po -a SLE é oe “3renSPEER ™ Bee ck .ee Ss es ee, peestge scores|2See ae “Sees Reais, SEE ee Soom oe eee wp fs SEE Ss spies See BSS SEE SEgs a OREN Cine.ee eesSee. 3 ee o Ss. :iP giSe — ae SEES ae asBae Se ORS TES EES RS Sees bee Ge re RE Eeewe ee Se & eee Seinen See I SOREN Seo PES Rie SP Re See 8 NAO a¥ L Sat See oe Se ae Se ee eiemeenamne senate ad oot, ean Sat ea see “E BEESESS ie ¢ at TROEEE" SeesBtBENE beg eit oSocean Bo TREK os. saree het Nghe Sage Boeee Paui SSaY Saha UES EEE ee Brisa cwerabentens necnns oe Ear po geste, Ses DeeAS ocean S g ESSee ReFERS LLG Rah TCS SRE Bea tesa SEER ee eee cesSee eaeeee Taree

& RS SE . 23|SEE eeBEE Soa he Fy See mae ee ShOES EEG eegeBERS BRE ae ee eeoo ey VO UyEEE REaSE Ab Seee Seoe ares Races Pagano ee neee Es, Eee ett. SAPic! Re Se Sea 2keer a=nSeee ee eeesSs * gh REESE aeseee one SR ees Re Rte ESSeee eeeeBoe ae Beaters ere ty cSae oecape peeee eee : : JOR SAS ALES a eer eae Seae Snare eee. Te AParSe. seont ss pti Seer nes enn ea Ee aogasa

eet IEEE SON See Oia ea Ee eesene Ri FSR ie SeRoe COO See ERS Bee ects SSoeeee 5 Fete Beets: eraneSSSR eeSe SReas2.Be YEE 3 eo ooSees - oe ee a _Se ee SERRE oT ES SUESRS ESSee SEES eeeaCRE Se Re ae SS ee oeerence Ses ater the Peane TEES ee ee a eae aa iS

ereBREE ESS NESS eee eganMs aeosasPR OEFae aso, PEE PE SRE AE atin eee SheosBs Oe pene Soe ELL Oe, SRE ee pieee regeee Sica ERRem Bee ESS eRSEO Be ee ye, FES— ee ooae — 6 2eee “sede Se : 0ES eeaSEES Bho aes oeeeenema eee ere BERTON Beeine ee — es foe 7eeeoo |Sresccsn oe es eee ee San Saplesta Se ee Sree ee Se EES ged ceoeieee — .eteesBoe ear Ses eae See esBe "eeeee ee eeeee eeecon ee : otARS ees2eeSee vane eeeEOE Risers Riccio NER re,ieee iSee: Sree seeeee tern aaa Beeches wont, ees L SoD SELES Tee ee KS santana SSM fo| eee raeSE 3 3.BeLoBERS Ee. Saf ai See st Seeee Ft Or ONS RENE S Bene Cees oeSe eee Se oe ..

a Se ee Ee ae - : eB. See a a eee es Lag AEE Bee AS ea meen ae Rae en ease aes Shor. eae Re Ne eee we ROS 2. See eos OE Sree eR OR es cg Ra SE reece te Sa Be Es

ae eSSen RG Sag Se ay he Se eee B cheyenne 3 Fee pects tenes cere begeSy Se ES es OeSSeS eeeaeSE he agi yeoe SaOES 2S TE) Wroe Pl SS eeeee ee ee hot nons Sean poet ee gees Bs ioeBRN ae FR So most Sses oe. oad Por TL fe SREY Bear ae Po aecee Soe ee ees Saha eaiene ca se SERRE eee AR ES ie BiaE eeeeatin ees= Se

UN RS gtVE cm Be es ee ORE Res Sera eae eeLE Seema See aORO ee5Seen oe! SodEe gy nfs 7 22 RReIRoR ©Soeehe ese ea5 PsSieire Baer Re ae Se Se eeeSRE Boe Saad oeSiig, ee BREE eeR56, ee gee SeSRR Ra eae ATyseeF coe i oe . ST SG ER Se ee te Sees eee rae oe seesaes Sais. eeaiOE COO Que She Ee |beh Se US mae ee oe ear Ripe eee eee epee trres See Se eS Se ie gees es peda ace eran . ~~. Sees Oo UF 22 ES CSRS iy AS ES 2 eR OR era PEE pee See aay SOON Oe ee Sa Recenter eats ies re oe: ae oe oe : ee es SE SRR os Se oT eee ee ee ia ae oe . _ oe ors se. ere eae pees eae ees a ee SES ie ae eae setae > ERE BESS Sh eae oe EE Se

REESE SUE US Se Sea a ane Be meacana ene Rees RUSE Oes Cphase RES on IaeEERE nee See eters PRE ESE SEO 8ES yseS :aon 2 RF SEE nar Sots SE eee2, 2eames EES AeTE eC os Soeeeet woke. Soeens SSN sae pene SRS POSE EE Fees FScee SeSe a en ae SEEsais SRaes 2eae pes :“ORT BOLLE. (SOS Se TE Se Rp tsSR ag cn eae aR oe BR eoiyCRSP ATOR TVA SRE aees pre ee = cane

Seer ee a ie ee Oey ge See rs Pe ces

Voyl ru J ESE Sa WEE Ee Spo eee eS Penns Ee aed te Re DEE eon pea oe Beecher ee Be ee (SP Ca See EE ee ee SE ESS ee es See T "SOS eR SR Ee eee

By ETNl) BAERS 2s UP Soe AE a ee Nein ue Vee geoe hcRiceE eae ee ee : OF Bee aa a i Ne SEES RS eee = aegaye ae he See Benes Re See Shee SEESBeat a ea Sug Be arOES erste Nan seeLO a aSec OOM ES a. BPP PSE SsoO 2SEES eee ee ose eeeee Se og ree yee See Se oreee 5eee aRee EE act SeER tas ee ee Daeata nese Re eeoe te RE es eeVARS ee SESE aoareBee 3mrFoe Sa SS: SaeySena, Shee Be RES ee Rares" Reese a Pa eeBee Ss BoGSa Sah SAS ee tee SRO Re aeBe See eeeghSRS MESES Ghee ES ae ee Ss ee eePaeriesest faeSRonan Sy fhOSE VT ISSneS Eee Reig Bie Ra ee BSS SS-eaSapiens Brae Pe ReareRR Sere: cetera a SEE Rone ee eS sak coms SIRE ae ee oe aoe hae -—rrr———“‘“C ‘CQ So See aa ee SS raha Saas aS GX. eee aees Sees pce ees UR ene SR ace ee aaa Ee ae ee Pera ne ae ne eeend Se TEER SEB See eeejRees bee Rea scieneat sreSa Rteeee eo fae ae Pees Pee Soe CORE, SRna EeEcono Bae RE3 Sg SE Singepeee Boe: Sa Si Se cs aes Bo eh esoeeeSg eer RE ReRP eRe. FER See ee ge REN eS eeeee eRe Reese Pees es Renee Tor ER So USEESSE ae ee Soe re ana Sieben SRoe RES Ra nmap ae Tes ORE gs oe oo So Ts TER AF.3ShTee SpeeSich Ae nee = eres pierce Sewanee poe SOR RIprSOO OR ES, Sees at ReNneA oe cena BeyteOeceaeeee Scns ee Recor Leger eeay ee es 7_: a .

oS Soe |. ... oe ee ee eae pert ol ae ee “a aes Sra Sa

SS a CC ae 8 re eee 3 ae 4 Ce

i aoe Ee ee oe SC eeee:aeSeee ee* | or ; | cae aeie 7oo - aEL Se ES a fo% Bee ae Sa ee oe oo eee Me Py © Pe re. SESS ge EES. SSR

apea.ORE Oi | SER Be Bio ecen eeeneSe ee ‘ PeSeta ge REE os ens eee. eeeae oneaeSees SEGRE RS SeeeEE ee a Sees BOERS Pee ee eee ee RS, eegg PESOS Eee. ee cn ae BES ga ae eee =Seas pene names Beis a eS 2 ESESS i cae ees Se oe es4feBee ee Ze Pe SNe eee ere Se aeBe (ES SeHoe Sefo cae a7 Wg RSE a ae eee Oe ee ge eres Beer a eee BSR SIE Te PES anes oeSN ee & |- |hu. ee ee“eS ae ESEE Se Fe eee ere bh ees ReSee ee BeBasta” EES SE, ES Subd eoe We ee ee Pu a. one SF ee ee EE | RES xSe «Spe, Soon a Ra Boe Sarna Uae Bee Bees et ae een eee ae ee eee ees SOA STE SIE OLE SEE Lae EEE: SESS SO oe Ree hee eee ES Saachicn pian Songer eres “Segara ee SNe oe | : ee ae : Baga ti SERRE aa EN ote nan mn a See. Rap woe ee RI. 2 Se ee Soa oon Sse Sou SLReee oe a oN aR CS ARE vs Je Bao BE Me ES SESE o PASE TSE CAE os ee ODER BE ae ee = See gine tan eee a RSS a renee SEES ee BA eae JETS E02 E Sai Ug GENEL » § SRL Semis Se ee SERIRDII' SEcoa SEESEES Bie oe as soos sene te ae See oo SREB ener Uno wet lt .SESE Et [RSS Se ¥cee ee esSP EES ae =Sir- cn— . . TL .:PASLEEE oe Wet SEITE, eee Pi aRus aaoe

natePomparison oO ial photographs to show 2.4: C f 1952 and 1991 ae p 1S in distir 3d O - vation : tion ost. iriof istirict, north-east Kissi aBambassiria forest island inougou *

) dou late vicin 1 ) | ig V . In the Tol ' r icc ; eis forest fa , . hile entifie tly rec . ; ‘ } hile to : ' V Luiustr c c| in SO |

l u xX tracted in ce 3 / Nn 0 - te 3 2 " Vy t island around the abandoned co; D man“Nn r haletely Care . Islan ) nia . ,tne com

even Aroun old, d village site, for pcomple d Sandaya’s destruction. «hee examp e te he orest ha di rtain parts, although it h Pp oe

anage

IISSI riathe locali y>5 :_| “a town, n ; 2 .ia, late vicinity of Kissidougou , Cc illages in 1m ne ; andsan Cessane, hav ome;Case who O Gbangbadou, Dounikono Orland Ve; enin for buil Ing p rtiall y CONV erted into elard ‘ ne. LJ

mics

y t y locality in outn-e

On the lop and plateaux between t and Vege . ith-cast

a betwe he different localities. I ; . eet ,

veg on Ww Cc D reas of suc 10 ege Wh er! rance of a ; wine Vv conn a fallow ; otographs, it isvegetatio getatiCc |

exp anses O Pp t in 1952 have wholly ceded p 0 fo . + h oe OW

ISntly V1S1IDIC : * this sab equ ile today theONT vegetation of this; locality d

sessmen ith ; arge toeach district. ortions of sic . Onepr vegetation chan mo

in |inée’ Gu n . na t as 2.5 ‘degraded fore oas "her i Nn tactional ’. Plates‘forest and 2. Ll ‘ex-Savanna . 6OV i" is ssidou mm Nn OCcCu it evidently, } 2 naS ane transfo uthern iSSI ou n) where this sam Kissidoug d. . é

e

herly localities, inter- n €

ii

= \8 a|RY: ’— xx4 Pe= 50 |ca wae” z DIS ‘ : = Peal a a a a £5 ) ‘ isha 4 Sore \ 5 = oF ;, ae = E5D oS oy4 ;‘ , .Ric ‘ fe© 2Nepa abe. af ow ;~ :an = thor sé “ ~8 | avg wn— 2 Zs) > oy xy = r3 * vO Bet (ee = , key a” E wiped . ey = ~.:ates Hae ge ¢ | = + x= ey ‘* 2 —= =QO. a ax: 7axX

Fh OAS : wy ok : ‘Gate KR ry) ia) Ss g eur £ b= ~ y cob ze a > ' Be ae ; E ; ” 8 PN! a a os \, tee wr £ 32Rs us) Se. Fy \ Sons geBoss 2s Po UX EpsOIE 3 IN3@ ot

» Ue od ._ ya BS ~ ° x Pees S nN 7dhAUN ete gS 5 4 NN . 3 a > 4 > oe ‘ es m™ >~3\.;:f

LEN Ors + ata | m~ {Sear eee >..lCU CU Oo

:cae SEEKS ' } ; c Oareahy r g Se O eat, ee «Yl SEER L a 9 2 & =| ANY SSapeat es md awlNEES fe!2, 5sg6.8©oS3 / PENS se . SS Sos. Sar ue Mab ba 824 33 5 oaAb. Da ae SI 828 z> = 0tw ieOe |] ee »6oO oO OF \.=Ss 2ha—€ & pene ~@ °SSNad 25o28

Pe 39 85 | 2f ee? 3 WSS : e Si ° 8 8 83 2 i wf , © OO ae oS sRS5 os 5 ee r.tr, oO” nn2 am) $ Sorgen Sa dol Pea 2& SF O Ye ™ 3 \ & weRe he OD °OS onPier ap Is Se = ‘“4 ‘ey ‘t n sf Oe ; ‘¢ a SESE an S ape ty oth 2 Bae jTet ee a « , Y Pp ait E S r ¢ ghee or a 50 be \ Md : o 0 Past . me | le > 2 t ; % COR ed 5. 4S Ai &

REE SS '2—_ oe LADY o) N N ie & y> ty, rs) Cc ear aSOE ~~ @ Rae ]22 an; 5Fp) += Q ga.oF ws ‘Ae! *LAY} . a “” ©23 %at 3GQE 4 3lece

t E : o, ©

& wt ohfor 7fe :alig »Final “Py *eob s=8N , oO oa AALS -< 4f0Ce ‘.ty ai?anty . .ve fs ae . “hate oO os Ke, Sf a ti P —BSFeo) Oo wv &§ SF £5Oo &F wee ww aGB 9 co ¢a

RQ .& a Ss OD QD eo S&S 8 DO VRS [Sa 422 eo F6k zee ae aus .8 N& o9© 2= os9A,Ss5wile et o 6 > 4 o oO & 8 oO Ww a) a =| 72) cy 5 £& © onan =_Dy aad ee es t~ Ww)

Ne! 7 moe 3 rc on HA a>ro) NAN ma NN gD an DO CO & ono oe ew Ou E=*ES meAD oomttoenNnunwMND NHN a OS st

+-~rRwAaAN we A Nn Fei nn tr LH mH & +

=| 2 oS is) :=-O 2

= o 2 2B >. o z

% Oo 5° ~o 2FA5 E | | 232. bard

Aw 3sSs© &s Oo = Awe6 &0 oOmw wofgy oBYM OS oDYY m ay od Som! +20 SSODO &CL 9S 2 oO ron FPF yy BS eYF — =o EF=€a™ SB Pega oa PE — %@ bo a®toD %--FR E Awl & Cre tamdt Funda tf a Ee aa ee Be ie LS ee es Ge PEE Bet. an ee “Ee Le. ro BE emer + os

aa ee ee ee phi 3. e e ia gs ee ree a ee eei2ee ee SeeA ee “yg ae ee 2, We a ee ee oe rg a ee es Been 8 ee ee g a er ee Fe = eS ee aeeaa=~ aoY_

Se to we _. VERBS BS eS ee SP ee bi eet Neageer ete. NES Cte aie me a ES

PES wie Set Pee “Bee Sees | a, SE be Se Pe Lee RY i, ORI Pe a oe wing ve EO Te eae

a ea a cr rn ee as ee MM Cease i, Ce EE, RRR ak Bass. Bf tT Og ae Pi as MR RE Se ERLE ng | aie, pit, A pen eo an se eG Re Spee oe ae Aaa a © sheet eg ae ee 6D

Ce a Ce ee PS UE esc DN, 6, ee ee a ES a apa Phe ic, gy ee

ee LeECoG ae eR eeeeeLegh ma FE aga AO ar eeeaaoa a ae Mess 0 | = = Se, erayae 6RL og innRMR pmms Ceae eeSige ee ee eee eegeswe |PRE pcreREE ee eee eeeeee aSige eeeEE TE a ON ay Me Pig OEE igs gO Re eet ty. # jd Fa) SEEM OSE, BEE IRS MCN 5PA reme tac a 2%, bs ee MN a, 9 hs ee Ss gh Egg Pt ES EOE oe SY (@) B 2 eyee la SR SARE eon Pee. ie iS >"> (st SE Ree PRM OR RBPe eR ges TB sreg eaeFg Ee ae EE OF we Ys. TSE SP * Eee ee aes a ee ee we IE ee ES Ge Ee OE Sede Dieaeen Nag EiES RIES ee Sigg Se See BE GEE SrSR RaONE” ER 3Be RegeEi4 Rp lo eae ag ayrar hae a eeoe RO EROaRRM RE,SS LENS SRAgs 98s A oe CRig eRe OR RE TCT.

aegee i Ase aeegee esa #ee “Pee —_ "Ei oe eeOR gs SeBe ea ee

PoE EO eee eeagoe, Gio eeaEee eri orre. sae PS, aie rirecee — Fee aera eemas Rie adee AER eet AR ato en > SR be 3— eRere 2 oe es

Seeneae eae & See, RUNRMRS Sn oRA ee eee eta a 51 Raars, MMNe Ree ae nt Ree ee nk BEeen Cop RD SeTe eae 3ead ee te ae CH PF aee neSh es$n” Riceiis.” MS reaBae ie a ee fna eae ee eae rs aGare ped Oya etBracke alcI ee es Se aOS ae aett 5SRa LR aS Pe agages poe ee Bese Pape“Boy Tae eeae: aneSete ae ee"Sea Ee Be Ts aa Bo SEER EE.ERE are? |Mag ie)

be ye a ees 0 ee ee ee NM i i. fy tee eee ee I se I rag 8 8 ES Cg ee Sang

et Meee EL eSog Beeg eg Ee MRR a ray} eee | eee gee~ES arag tee ee| ee 2, be ee ee fr ee ee fePt EngARR fe AP aee po bOI EEa ee oo “ee vie ee genet re aPe eeage oa Bethe ee § i) fxs en a| yepe Ss eg aeMOS Osgees Mo aCah PEE|w ORES ae gee % Cs eg Pe Ne a ae Be SEE RR PPR EF Be AR Sp, ERR PE o.

en a ee A ea C) CI

eli ice =en ea fees? BRR we SreSee en, Fe“eeaeeTUES ae 8BR oegt| EB ae, OS Te a 28 Cc “SS. Sp ee, zSIS OOS EIus SSR Se eiA STE ee aeEE RR SPE: rrRi, ge ersee ingaay te noeMk fa ceeAe ERE hbge es Ee“OSes VE ge eeSLF Ee ee, (ee pero. EO a aera ae 2,eeSc Sian ESaee TE eg eee tg SURE Pe ARR Ree ns a SR Jd :a... pec aie Sea NI a Bs Vin Ey MN EL«= apeeAOE ee ae ee ees ESEee SE Ee BER | aae Beg genAue ne EP ae eam ne SOR ee MM AM BE Ss aonES YSSoe Paes eee eeekS Bete EE OY ONY. Sea © ceSea ae mn Rsss cae tec SSe 5%". pene Bee IE Sees: EEener Daeel BeaR egaeae Ee1LE Re,

| Mee mn 6 4S SF ean Oe ets. gee "hig, Se eee OR ee ea ne ee a ee en Oo

- eae ee ee e Peale OMee OM SG oe re ng EE ae Ego A vad t4Gs ,. age ee Be nF eeNe Se MR es Bee aeeM aeiy eeBBE RIE he RE BES GREYS FSS Re es ee ee aaa eae Pe 3 OL, ye nn PM cag6ce Bg Sa? aeEi os eee Bor, ee2 ee Eg oe Oe eegtae gD ERREES BAR ee lon EE eSea ee :" eg oe ae eee ee nn er ere eS ee ee ee aq bcs SRdin. ye ge ag2 Wage SBR eee a Fee eeRESE BEee ee ie, TEBE 2edieee ee ee ae * fe EE FOE FE Sb2is,BG a im © pO ihe ae a gee a igeen gn Re EE eaaSkSe#L “a oF ¢g “-Hee NE: eg CS 2? ~aeeePEERS Be te eeaaake eae a Cy ev=ORE WS Re Se Ee rsgDAe ee eeete See eeaye eeeae Ee aa ew Es pod ~EF -SEE 2an eee es MOE aeae aes meat, as IE a7: aes ee ee — dn Pom 7 ae pe BT, Pee We ea MR Sm RR ACpetA+—_ cP) f BEES. : . fin. RO % = Sei oe ee gen pec ~ he eae Wie ee ae fe Big * wet Se RR ee ee hm liens a Pie ee ee igs nae fi %. Sues, ig ° (tan ny Fe Digg he EE B ed Re a CO ge OU ee ge Raa a ie ee $r Bed =aeka iFERC et EES POWEES eeeaeee aeHR, aeMEeeee pera = - 7 et ee EMSow MB ec 8. RR se RA OFT

7 fs - ek Ber - eM a a a ee ea, LL - ee Es _ Beng Re ee a ee es aes per ee eee ee:

~ we. :37, SS :Paes moESS eS ER aT a aS — “=. ee Bee ee eeBe i IReed RAC eM SLM EEE meee. io See See a" “ms A © Bes _ Foie Eee WL" Pee Ee eget » me “ at EEE es Ree 4 Se ee ee cared pS . Ware” "ety, Ee eee : 7-«: *a +8 FUESGHE Td2BaoaedSEE aDEB eee ge a +, Bo & St. nage 7~ ne .ca FS - Poa e|oes Aoie SEA Bee ge # ey . &, s ey ,a Bpte. on ee Bee oe * "4 PE se“UREA, ae ae faseEee CSB - Me JB . @Se ages 4,an’ES A eee riteaan =we & Lee eng “wr Bae SMa ty Be, ibe

£ -gePE .:HR 7bg iFg -s ges “Be Pe whey oe es .77a eke -= aah vey # ae rd : ik oe | ane wee % , oo eee a woe we att ” Po a @ & See es Sine $ 26S ee 2e Es Re RE BES = OES Ga PGE, MR : : = ; a 7 ” gary ae ging Bek nS ee ~ 4 = Sah # en ee BO Pe Sy . _* 7 BERS : a ke aae t _ 7 el: age OR moe ee = oe , z a Heng SI OE ag PLE Shien. USHA 2 . BPE 5, ee MR Ae 4 -2 _oe. Begreeee - #5 BP"BEERS gee - Seg) 2 Ree eesae 3 q -*Sn BE Se a SeRE GEEee DERN OE. - tt aalis Fa - a = Bh SES See Se ie cae aye” ES Re Pe

a -og agin ge Ce Aage OeiEi okiOn yeMER 2 ih eae ee na “ * ee em OR ee ee we 'S at = we gia pee re ie ae ee ee ie a oe eee ee Enon EIS ge ay enTER a ee OM MET eeTiEES me os eonad EEE eee eel Le A eee ees. Bue ee ieee BPO SR 0" TPoR oy oe esee Rt Rae ee eeBak! Bee ge Ee he PSY ee

:tefee Roof isk. Fa RRR ets oegE eeeee ee er Kee ee ed er Bes an. ; -ca PRES ees eIERS 0 eaee, gsEs Aleee aBBS eRe Re Ra:Pe eee | ee Pals Bh:ne .: abate eeeeeRR OL ORC eee are Sie ie. Ea Ea BE,eee EEE si es 7-5Soe: Be,eeee Te a aeOE RMR eeeCs aeS, gegen x!” t WEES. seriar ReeERAS aan ie, cE a Moree Sage a :> a BE i.Brsee Oe. a8 eSiae SEDO IRS LEY | ee Bee eiacree RER EGET, (ORG ES Lee Ripe eee ne ea arr =See ee RN Sea, 4 SE OM Ee aeraee aeaenees eS BE

2HF aoa ce ee aee ee "EO owe es oeekOe eg en aeFR eeOE Ps :Jo, . “feb .eeHS . obs eehgee I iEs ern es ee . & i= 3 SE, PSOE ae SA Oe ED ee ESERIES. apts ge eae Sane Ee PSE eee EEE ae ie Pens EPO BF TE, eas 7 Bes sa far, OR EZ a CR CRE BARGER. cee cee ee Hey in RE, gr EE Bauer ABE ae

. - :Ta aa ae eed=Sh fee at ye aRts SEteTSS a ee EEae Se oe eG, Sa eaiaaBe oe :# :-- si mn sche BS faire eons Se Teil” Es oe ON aPits, de oe g.ee -#€ *farea ee EMae eaee Boris esaee ee. fa Wea afEE eeeEe ee eRRe LN. aeOm. Sao ° ; me : ee eT he ae oa a ae “Spt me Os Ee aSORA eee eet I foes Bye fee 2 a 1B RR. AE. ee RU es is .* B = Be ic Me te. i ee ae oe Set ii SRS EE eat, . Ze... rhs sel. we : . : ee eg Oe ee oR 5 car pee ee ee :.Ei xk . Pd, : Si Ta eae Os, a a a % : oe 2 EEE SS BS wy, SP ee oe ee eae zg.a =: Se eee yseng 0 aEs Ka* ‘,esce gw EE Ee : . ot ES: Pee ee ee Boa ELITES: iy a: eae

a eeESE as: “=. : zo 8‘ 7eeTE at oe*&*a«By, .Epes RCM 5 ee etn. Se, ee - sth ee ee : wea:ndke 3-2th oe * “Sng ae BER eeeBie ee fo 2 ee fs SS eS. z8eee Hy, DORE

ge PRR AA TET ion pt, 3x.eR SE Be eeaeeEee Mi Re A z® Re ae i.ORE Hee EES, Ree ees al oe wv OF ware yey ots, “gee ER og“Cae 7a ES ‘a ..>etre we eel Sia .iPSE EL Laces eee SS ES > 2, FfaS aESee ee ce ABR Sp TRB ag eg brng? PE , ee” 8 Tian, ees >Ed f.P 0 2S BIT ERR pee eS PRES ie a Be * . “hy. EE PP 7S ~ BE a RY Furi s “Sef see ae a et # . % SE ae RAY 2A ee Tele ie & . Shs ote oF BMRB LES Ba hie ge ESR fog ok 2 ay a EES Ot EE : a-~-sot Eee een OO” caPN*Ps PE‘ oo RS ea Be. ast(oy - ~. ge 3, TE Sse SYga ER eeBe ie, “SES TaBoe Bre 3oezOE FS geet ES ES SSR Se BD PHL SAR OE Fede, tees eee eete- % SRT, ~be oe He BF Bee

-a =€ak = Fo A ap ee er = tg oO” BPE Bee “MARES. ata TB Sta Fy +ASEE a ag ae en eSVEE edecon PAB, BO pe oe RgSE a anne . :2ae * “Gog - Bey "gs| in as 0". RR Hee PET Lag SeOE 7 oe zg F oe 3 a | bee ~ Behe + Bo meg wes, : BRE TISES a Ree fe“ -" SEat-EE2kPSEPR nh : oo .: OF Fags Bare eT sigesie ty Seep s Zon. - Bese ee ae 4* A - Eo my7 Te “PARE eaPe4 SEE a Spee “& = ne %,—~sia EE OE Bd : . +PCSS gageARS peFe ees Be Sono es OR Seth : mo. EPPS ate 2 Ee * aoiait 22. oa2 Sa * aegeoon "ueniee, 7 eg, 7 -~. Os TES BS :7‘ #_. 4a- ro gelBRD STILE Te Ee Le vee Titoe EE- “ie eee4, = ae,PPoh, aa = Boe. =ee.En oo cei bor St Pee et E BE ere BS,aeat Beow. geet we “Benes, Be= FF x. ie 44 wa REE id Se eeeee Leer eee : Seen EESan Saree Soomeat™ -Oe °“Retest Pow Steg Bard weet

:a.an ot Be . SOs : (Bog PEE aeAEeee ers eee + Eee es eeof seer -ee ee ee ~ae SE's WED DURSLEY Lhe eee LB Beg a .Date ee 2Be Be, .a Bee ain Se. SE eeBESO ey 7Me & ow ieaeka Bh |ee pad oe :es Weg gaPADRE edie Bat eB nT aLees Bsoe ee ;UBS, gz Pk SF oem .so : .-wee eat EE giEF ed ; eee |5eet Mbeg eee -ae%“ih, Bee RS HB Ee, dace Oe Bevhate 1LAR: ase -EOP 2SORTS Sh atRIE ESD ell aera tr) aBe eo oe - nrCES ee, feaPoe EEE AS, EESeee arte

fa:: UE rd ‘a a-:2woER “ESSE ae ee eeAAR ee nec Or ke ee pats 7 Sanit gS eT iad on EO ERR MO ECIE™ SF eae ONeS BeEs ° ORE 5ee it = Be ee ift ee ps Be areee ee 2= BE, _ re,Ee Es Me ane oe Ee ee LR PE wo ae |syea ae tin iReaesaeee aocreae : Le 4 EE BES 85 eae Reet Oe‘Pe RC PSERS 8 Ae & caeSES, rr Rr Eye a Oe Memeee ge ~ PIE. . pier eo~ Sr celirins aea aRni>Ge OTE. sedee, eaDetae eg egBatis aera aeaa

: : Se ie eee ic Nc rns a Fa ee ae es ee

OSOBE RE ER EEESEEE SSeee eAeee apeg th,oma Beewae RI.nee Es BE mEeyReae . -o Pe oF DEaCe te Ipic" BNO a tOoe . yeriaBeSpies SF BBA Sea mlENE | EEF EE Sacre, SERIES. * EOS BE BeREA ees OEE Gee EET eM METAR gO as PEM vad - - —_ 7 OE pl ESAS OLR i IRR eR ROC OO RONDE SLE NR EE POR SLIME ELLIO HSE ELEN EY: STS FTES a

TE I5is, Det ae ee eee eee ; am wet eae ge I. Siagliee = ,: abs a in ggCy EUS ee re esete a ara Sing ee ees ee

ee See le OMIT Te te el) RT RE oe, RO ee ‘

ge og RE ea serertnen earoneioweparee ons asentoo nee

ee SieGS erry kekePoop ameee cl i eae teasaa SER NSS kaa sacaeee 4 aZe .BES - Ce PEE hgSe Es eee Bi a-RRO, eeSE BO FyCarr aE eis Pe eos ESE ”es”BES 8h Rey ee eeOSI aaSec a laHES IES 2 BEES Ti awe ea Bouts %% tae EERSROPEEP"g PR CO ES PEREGO ES. EEE TESS SI ee OR Lan eaOe ae© tae See hinge oa iar BO, 4 Bae Sikes, Oe &- BYR aoe centre aRae LLei: SRSI Beha eerie straps. | Eere 1 .ERE PRATT aeDESIST ER REE porn ee ers ae ee oe oeee ae i.Lo Bes FSS ML i eee RI die Sexfang Bur gables * eofa GEER 3NTR E Ee NFBp ir Oe ee Po. Ri eee YEE ES ty GEE SSrreeeee eae ee een EIA SO AD ES eae es Sep eet, Ere HEL Soe eee ;Br£208 ;SRR PRBS BS LE Bee Be OE cere See ine meESSE eeSoteSO aes BIR RP eSLOR ee aeat ae ee oF aeCF sdog

oe oe r ee an ee oe ve. Se ges Bo 7 38 ce ee Ce ae 2. Co POS mee ahE eTEe EEScie BRS,Bays Bete Be ob PS BPEBOS of A AOR... te SUNRUISER RASA opeares Bie Seen ae SSR naCRaa S 2f PUL 2 Pe ee eye arecaSAS aes aotere Se SSges es ee RRACREA GR ERS SEP poe aayisee has pa ee Beis iss Paes Hi aeEASA See er acral heeeeAU Sa A S ene a8

OR RISES REIN seaREE recom a&Eee 2Bagge ee weit Uesei RSE REE ORS Sak Be eeeSeETE. ee Re ee Cas ne eR “go ees, aes FIO SSeC ee ey AENRs eee ee -. EE enandl HR Oh ©Ra Bg ee iABR ERE, BBE isis ET EOSE Dsee SE EES OR Beas eae Re ois ise S20. J“init, SEP ihMOTs hacia EEE, eae UAE. Se $CBa RS EE, SEE Ree es ee geoS ee ooeepatie’ 2ye oo i:LR IO ee fe SOS x, 5tk, Bere 2at2S Gog A ee SEeo ReSol RE pa cae EO De ORM ENO eRe, Easeae as (ese FeINES ORBEES REes ee as eeSie Ao GABOR Raees eateo LED ERRSa EtseR Ree te OSE Se pee ees Bsa SEU oie 8 of .

AOS ES Aa Be APOE CR aye Bee B eee ere aaeSea eae Bn aSaari ea ee eeleow aoe 2eeeoeeeeME oo aERS oe yl LBRO oeaOS a8ee he PGR eg go OES 84tas SRNR aCOREE eCogeg SRA LEO CTS OR RE aace. RR OM SIE eRe aug BSL OT LES Sei on ee nd oe gee eS Eber ol SR EM AUIS ear eeBERG PSE Mia SORES COE BN os aN agates reae See Ree eee Re Be eee Bae Rt reigy a ee.EES FESS RaeCAUSES ee Eee bas Meee Meee Siete, vl EP OES AIRS TORE Recap re ES Merten ath ts ROE et Sanya e Rags ee ne RR: Rite aSy ee ee aCaeeeTE PIS ED REEG A> ee ee ere am Bia ee2eee eas BS Say NAR RR BE Ci eeTED 1 Bs EE SB SSO eG gs AEE Be eee he ae ee a not Bae OSS, ee De SeenON PREIS SLE IOS, OSE Se eT eR ee nO se SIGE: eae fea See act See Ro roe ROSTERS Goaeee pes a ¢ See Rn feyaPRLS BEEBig, Bo ETA|ten ee Reo CR eSee EEFEI RT Cacb $fer SENG 8SEGA Dee eather Bee BREA SR RRSBI eae a COE pene ory Sea © PE aa sy ‘ate EEE SES Ro RR ae aTee EES SS Se EB NE eeSS a ee ee ee Se Spee “A Ses.GC ee aaa ee ee ee. SRR PERSO, SAT: oggaC TSS CEES OOS |Oe SESS ee ie ORE DURE Oi ea OAL guSNA SS56 IE NGS oaSLES jeaST ORE EES pes ees es Bae Se Oe eee 5 "is ee er AE See aagit OIE RE ene BO ook, geBet ORGIES ORO OA SRE HES Cose ASE ma OE Eas Se Sic BRE oR eaeee Gea een Soh hPERE ES: z bn aa ae REE RS reEeG Cnn ee©Bs SSRee SRS. tT ESr ENE Boag iGhee eR ERE eeES ae ee BOS A OS Se SEO SeSL CO a a aSerine mers nara: ge Pe ae Cee CUR EB tgdit B Bekis ee LON eens Sage ee Op pRB ge SEE EE RACES Le oe oe Sa ree OE, SER EE Ss axis 8wikee BREE eres. LASERS, 2858 eeehoeno EE IEEE Gy, PES ee TMS EERIE BP Eig EE aSOe Seee oo. .88a .a LDS Gale as Be bs ee 5 ne ene er ee LES BROS SB SCE Ee SSSI er REPRISE Od DEL: eS ae ae, ee a eee eG ee PBR Meo eA a Se. Sie ee gt Ea See 2 f n Be ge S23 WEES She STE Se a ene RRR. SO LOE RO ea a 2 NE at. TERS HERR ROS AEE es ONE Sie ie ee ee oo. oe She eet ees PIERS ee BG a ES BEG ei Ot RE ak OE eile Side teat en ennai SCR ars pea sees es * Byte ee ney DE RI Le ee iii,a)eege eee nemo mae pee BeeeeRee ceHEaNi yk, SE EER Rirapets Bucs eee oeSse eISPbanoint: Soke EESE EP ISSR ee seen pseees — Fe ae Ee UME Lesa wal! REi Sarees SO EP 4 eee MnEee RE cea.nmr egOE eee N pealES etot ee aee nO eee bs 2 thee |.oeBs oe Oe I aI ee SaSeo YS Beets it nAET EN ae PUES RY SP i» on & ee." earn Ss AE RRM OYTee Ca ealae TERR agee PORE ESRR sin ere aeie ESN aeSeep 2) SER RRELE aECO CLLRS an ale ISaeEOCENE rear kaaStaats ee oe DEES SUC ame eeMe mn Bees Ee Be CRS Ee Baan ER 7 EU eeasSRL IE OI Bi SEES Rg ae Se 83 :ae5M a7 ‘Ss a i,ORR SEES Ge Se ieee reTS Lae Soe Rea anNee ee ee (GE & PR iPaaah RR eeaeHea Pee aePa Snes ein saeeke BOE, Ss Seta a_oo ESSN OE RLS asie BEERS UE UR orate Teanga PsneESN EN titan aaa OES LORIE OR Reee eePage ageeee eS 7ee zSNe 7Ee oSOe - eed ieo Boi eee eRRRO Rayer BOISE: Set Sane EL RRs hetS, ES, Gap Sane estoy GEE Ncpene Sa ae ite RE: eae egy eae ete ROSES COLI escq| ae pai mmncime &.Geng Se tS. PRS Fee&EOS ees BEE Se EEE EEeeay ISIC EARRING PET. op Tien sashMES eaeNee te en 2 Pe ERED RAS GOERS RsRee RMN eer §ec PEER Pee a aar a am poetor, BE, itiona Weieee fag ERASERS ARC ROD sep OT SONOS RS sriiseke. eeey nenEe EE OORT SR ee ase Bey ere Ce a_ Sohey ieuerage bee ieeES asFePee NS eaepakMeg OE SEE AeBoA Be"eee BeRES Apis muon cee EB eae fe eRenai «DB OS RELL oe ee ee ee

ee nni eocars cewots hr ee TO ae? Bigeae Ce oeee2EC os| is, ee ae ee ee ey ena te ee ee i ee ee : . SC flrceaes ee pt gEeeeae Cl—. eeto-7. LD ree ee pOPe OOeee ee es a eee ae rr EES ERS, a es Epea Mn RCOA eeeSBOE a eee PsSENG en ROM PeNeSEES BES EEBAe ERO 8 OE Be.| ReSee ee pee: so Be EBs 2Ie ae i ae mR enerOe ae ae Ec EERE SE pa erISS SS a MRRIES, ced GRIER |. Tks aS RO IR TO 2 et oh ORI CEar B.Sena SSiecrt EE Se) IOANC Ee PRE SSHi SUID EES, SSUES EG DRIES as Se ee ae PPIEE iGOBEE iSO ISEEO, ck UIBOR ee: Hepat a REE PES et De eee gikeiee EE ROS Re Sie lt SSEEE, TE ERS Rae gy ERA ee CEO STC RSE EeCo REaES. _ eeoe. ee mean oe ee Ses DUI CEE” EC GSH ee aR ne re SAL Oe ON gees SE MESS OBEN Bios: eee DERORAL oon BY ROIS SSD oS SIO FUROR OURO ne Be Seminars Sine) BED 8 SS Sek SeORR oo SOI BEBor RRR Se peerage SE.r aSEAO es DER RING nF tn ESSE aieSE eeeSILe CSC PLEee REE eReAs Sea nee iesAe PRGA | aCERES ae 6SSeea Sep ere: ee1k eeBOOS pean aoeRE eS Bao epuntuamere, mac erSR el ea Bega: othe a: NOES, Ee IIIS EEIOSEIS RO geoODIO Si Sal RR ast eae cyLAS a eeIE EG Picco S OEPNaes BEOR ERE «TI LY EP OS SE i ae esNa aaSrOE nora —_— 8rnce=eyAage i

Se BeUE ees: eeIEC ES es Nias reap aioNae agci Spee i See ergTenet PEs rece, Serene CCSSarria tectsSear fee Reise LEER EOE ARTES oe ARMS: Sie. REE ER GS tke. peesSERN COS ACC enn Page Srpeatis EO ree Sh Sete Somat ee FBRC TeteOS tayeee Sai aire Pages oe LS Bs Ra ER Ns"Se St aT ISa: EE OL Cue Sos ER SOON A Behe aes Spee Pos ERE A Eeeeen ie mmaSD y Cem in (ROR BARR TesiEset na apees ip cee Ce Pa eS Be Eee es ceosSes BF en Che edseeBE Tne eee ESN MOEaeLaICS he, Ae hee eRe ieee SRDS Shs AE CRG CLBUT RGNEM SE tebe PBEESE ROSS EOC eau uncecisagamasiee ee ee eee meaeepeed

mn Se I BERN Sa eae cis er ee Sane tne re Bee: ee ier ae SER ALN CE. aes Danone * SS 7. - oe EE SE ee BeTs, PRLeee2 ae enSERRE eee ia eee Se Rape poe:EeGiassien 8 PE ‘ | .7| 3a _a ee Mab SSR SeeEk ceae eeeES Bde SEPESOT cell BEEEE BSUS a aE PN Se eeeoa a ae BigP ZEN iS nie ee Rete mee ola ee ES EES * ae ee Br EE DEES nS ER IEP as oo oo 4 a

_ rr Le pee 4 eke ee te re ee Ap 2. LO rl Resp: 2 PEPE od

eras Be 8 eS Ga aSeSes PS ON NG SS EMER EES ey pee ee Bae eaSepik iegIcs PS — |Sie - _*.. 4a Lae eePoe LES SSR eae OE ane, SU ego SRE eshot oe ES. ThE RR Seeae Lee he ee LL ee Bi eae OpSE Sra artaBONE TEES SISA eae: Seen See Bye Se. Ae ENO Teg I Gest Naa AES eRe USERS, Bay Ree eeeUae 28 SeUe GENES 200 eaor SA geeee ROE OBE, tL ESan Raener LEE STG I eeRan pee oe a a= Pe ee |BO fo Ue we eee. eeeee REE LBS ;Bye eee ESS RT Ne RI Be COREE: iEetSE EP eae GOR. BO SES oe oo ee Pt NgRee SEROUURL, ce IESRR lcoe pee ADEE a 38 PEE Bae oneee Bn ieaged RS EeES Se RE 6Seee SPee bora. ae Re Ls sPO es ROSES DisisiE Bie RE eee pocheSS Aneel wg 1CE ake, Bg eee, STE MM BEtae ie ieng aeEM Yeeee ddinate ee FRR. BAR, GRA OBOREAS Someee a Be Bia Oh i aEee, ceEER cee ee an bE SSUES Ge RRR ot Pha ES BAS | 8Oey Le aESTE REED se Sb:Se eeoh eee PES GIA SeeeaBee SEO EeaeNS, RE °

Do. i 7) at ag Rego 2 Ca ae ee, ARS el. Pes ee Se Mes pe oe a oS MMe reee i aeOE en ea eae ETE I Ee ae PEER OE BB ee ee oo eeoe ee oe Vo oo ; ae i rr”—OC ee . RRO 8 ot as aroe al Oe (sa ore Cee ee 8 ee a ,—rt—~—~——... Ci ew Ce — eo il—TEES eee ee re ee ate ee pee MO ee ee ; RARER go PSR eS ae re Bee oth gee ee oesee Pee CEE .Ree Meeeee ee ee. we Epes iei ee Benteas BNEA ee _— eisee a Ey ee Lg EOE” ee eeeRES eean, OTR eae ares Ne elosS : ee ©aeoe Eos aTge. EE

aa eo oO Sore SSA BEE eeeRAE pee EES, Gee ieee BOSER «si EES ee ey2 Fe Oeoo ES aBg eRSeek STeS ECR aN ane PBSLo, RS re REE EO, BE BSS Bb, BR ORES EPRiRS neLESS IEEE te OEi, ShPRT. RAS gso 2S eeeeee ee Leees Oeas Be oe OIG aeBENE oeeae i ne meee Se EEG -LE Ge SEES, ARTES. BEERS OsAe, BLEEDS eee op Bap EN ateIPRES, yl OS LEas ge ae Ge ee Co Bee LeRee PGR BO en oe aSee ESE eeKine aete, oe aaa ee eae ae AS SE ISSUES SORE BR Ee ee oe Ca —] .eo [Sco ee ce ae EG ER BES Gy, AB ERP CES SSalee, nee esBED aan BOE ASE Ea OE ee spa ee oe 8gS7.ee Be 5. eas eeSESE oa ema ee oa ee a ee ea Sk Rr Pee, See Ea PRU aoe ae OO ESOS Th, SERRA ee RE a So eee “i ISeer pe a COE etPe Aes S ee eePigs BRE EGER eePee ee! a ee eePEE. eaaSeal. adhe reSt ieee PROUT RS Meee. 1 58 OES AIwares eg LES aE aoneee S or Re aeShe eeead ee BO Rein wie ge, Pe % —fihe, T USORANES sBEE. aa SOLOS Ie 7uP aeeamn aaONON PRON OER ROU ERR EERE Ce eee ee eeegn gee SE eee EOE BeOS age EE Se RE. oa ome gra ts ace POR ie feeONS TR ES EISELE ARES LOnes LSse ee NER EERE OS Ea OS eeeRae, aE BEL Oa agewen ee SS 3 PRE EOE es erehe ES MORES ROM ari. «4SR AesRe ees Eimer aia esBe a -i7

. oe acea cpr ee mae ee PIRES Ss oe ROR NEN ia Rae OSGvee es aapie, BEE. OEaeBEES ese AUN a RPaeSERS iG,SOME Cee ees TO hae Se— : BRO Sr eee Se BO SESS eee oS, eeeBy Bee go TyFUE Ee ¢ ONES OPE eeen ereanee ais Foes DAANOE tok aeEIS PESESS es Roe Fae BO oORS LGaeBORE OStape ee ee ae PRESSE SOR Sa apc sited er ee By ORE Poe | Ee ME ly RES FRE oe SOM Se su gett sp maecea erence eo gh Es Ga SOB eee STS IES See TRIE Sas Ore: es aie ee oo asa pe BOE att ee Be EEGEee, yoy aa SES RS RAR DO UES SCL esaE FEeeRE SE Becca OR os I cage. CHES aBey ome oe as eee BRS” BREE. Jee arose 1 GSU EEOES eSaDiigo SEES FCore ésBin aRe EEE PROPremera 1 TA EES See aAeree ene REg RO ase, EEagORC eshme RoR Renae eager EPSis aREN Gs eae ee —Ce oe— Fes ere Se re SRR oa Pat % NLEEES, ©US SRO RR aE aEPES a RASS het BON SeeBee BaeOT rear OLBg ROB SESE Be — = ee. aesi aoe oo Be eee SOG RUDE 088s : reer ESE aa REE PE Rea oe cso pea Bre SEE Raoe OEE SEE Bieae IRE ingROR. toy) eepeek ss Ree RSLS AeREE eaeBe BERS aSON Ge aE ae RRPDR See VRE MORAL ER SSI, ig Seta eeiS ee EIN IS SRN Mee, ea essai eee eeEe aeoe oe |_. BEAT Sent ep eSwhCS as Sgn getole a eRe Baier eet. keES OE BRP U SR PESeEROS aSDBO SLOSS SoEE ae era TrPASE age Bee te eam eee eee nn ee ea ERpaid EELah Netiie GE eae PSUS ER CEOS oe Bee eeaie eee ee Ss aRease SRE: CRB iFcasSE PEE ESSE SS SIR ShPEM HR ene ECE Be og eee Cen ce eee ad "TR ORE ea gn ACR ORC SSPE a aeenarte tee ee FEL Sy ee SRO eh to GE OP San EEL oie Pa 8, ae ee SSIS SEP EE IRIE OE PE eae EERE oe é ee Sapiens ee BER ere Se” aie Pe EROS Bad SE EE ESE EES is Seed Sha sie eee es SMS ee OER E RT F5 OOK RIE. eee Peedi PR OE ES ee ee Site oe Lo eo . |

PROS, 4 og eR ; - Sr Se. Ay Be @ ee EEE EE ES: pee oe aes cc ae en ae Lo BE IE eee. TEES Se ee ie a ee S

GRC iseaes ies, iea ohio SER see Sek SES : Es REE OO ESS eraroSUE aa er eeeSCS BL Ee oe . e| poe BO, Se eeYORE PES Sg” |ee Bee ane Bna oe Map We ere eeae LRG ae eg oa — Be pce as ieSOE une teeS" Pe a ce ROE SSO TG anee” oepee mae ue aLON ES eeee oo ee ee PO ee eee | we « FEES ee ec wee Eee EGER eee PERG EEEE OSE as TIES BRE. 3 PE an, SEES Vey ARGS is CEERI RIS USED Aaa OER a EEGs Gas ay ar NSE SER ss LS 2B Re gee Weer ee IB ARN Geena ese ay i ot RE 2 AeA See Rae Ge ee eee Re tae IU SCRE SE by EE Sees EBS ER Ge atid RIEOEU RE ais shel eee SO Met CRETE a Bee LEE SORES OEE ate eae See eae es _ Ee COR ease amarine f Sag HE aeee ERasSbanaimieoats TEE otee heen ESR eeSCO eam Bae Gt NR Die Cae ny nea enBELGE Se LyeS ee Re eecal aes a oe CORES Aiea NEES YS Ee Ee. a.eyBetbie” ETRE Loe giES, SHG Eee aepene HEISE. Se eeTREES pciee RY RS Ne aOS ee ae. aCo oe ee |dha be oe ne oe Ra ea ee eeBet arGieieee eRgee a ikas aes SS Bon SS eras Tas Br SOTe Beeches Re Sofueat aero© eeescwaty * zi,SR hog oA BeLPecan, So. SE OES 9Sirs Rr eRe ap sis Roe BME +RORISNE. EOS, EES.Nee RCER SG SSE Ie Siar aeIEE ae2 oeee os aot egee eS te ease me eScca JS Beye yess ee Pe eeRRO EGS LST EEE Gs ES CREE eiSeae ae a |iema aeae ere aera Eee Ea ES yee ee ee SNcee eke SIE ARG EA NS ama es EEGs Maen” OS AEC SIO as:Ee oe Se ee ae. ae Pies : ro oeEee aeSSE Bee emia Tee aaig eee erog EEE SERS PIRES IES” Pl Rae ate Ghee ESSEC ES ORs Sa ST "iam Sheng DEY EtI oe PES SRE sg BeaFon G Mage eSEES oeSaas eee ce ee Sie eee eh HR eB pees Naeee See ae tse E a ERS, EORRR SRST EEE, AOE 2 oF oe esa :

gOS Pea RS SS Ogsete aSD soe i ate eee re lee hgNes SESeeSe Ce RE teSsEE a eeReaes, are CC _ oe eG ae fe ee is SE eB ee ee emer EEO ye 2s GE oe eae nts eeeeaoe es "aes aa EER. Eg Mee ee ee Ceeeaecae Meo Say Mee De Se th, a2) HER coe ee BEL ORES 4SS 2eg CR eee BEES Gand 8 -oea9 ee eri Ta nMa EeSISOS py eer SIG i 8eee a= ByNeh 7% eae mae DR es ee eeee OR, BESiy, “seh GREE TROLS Ss PCRSSe1Fe SB Baeee eeCl _ee -2-_ee 8 ee apes

ee ee ee oe ere i es, ee ee rn a a a De 2 See eee Se a ee a We I cE eh tin ee ee SERS BL ENE BO See Te aie, Me Reh gaa aie Renn Bo Bans eR a Pp rees Zk THERE PARES UE Rk ESE SS BOLERO CENA CAE BES SRC RAO, AE RAT LaDy ee BS, Ee toys bea

Ben raieS nh Seater BBG od RE AEGES eaGE geaR ee “6 cin Eeeae apee ED oi ae pa hes BUS Os os Re Cee EEReaccee BES ENS Le SIE Si CRU OR syOise ee pas eePoStee ae ee PE Seni |= SS ©- ae ESPe SES Bee Oe eerE ee SOR hseSPs aeFoR SCG SD EEE eer roae eee Pa Ea nee eesRa eeEGS " PY| oe Bee EE Eege eeBae uncer ORR” PP EE EERE. ER eeoeerne SOT SEC OS nha eae ESD oReDSO tee ERO esAOE Bae.Lr oe -Be &eM .UE oe eeEt STC ne aeSeva EEE EDS. EEeeRR GSAT RNY REE LI Ts EOIN Site ae ss SOP gi Taso a SE Seoe als aaa es ae eso fia Lee SUE

28 oe - . as teh ee Seg US Sess es Boe Mie, ee Seed, GEOR Ti ang SEER Ee ORS RUEE ies dae EE NRE ea NEO ee RR Se aa sSouamvirne 38 od BiB eee FS £

Pokeee cc ES ee a 1TOROS, kBea ISSae Eigdiate Somereoaeerenes” CIENTS S ER“ SESeeeee ah ieEE aio TRIE tiene, nrBe Page 8a a daeee ON SaRee BeeeS ES: TUES IRIE EEE ee Mada‘ EE Eee BR naa ARAAi EE EES eR: 1By Sgr SE eae SOA ge SO MRR BS SeSees Neycae oeSES a Dae ee ooSGM EE Bee EU Bee Bs, I ISESS: Es SESE Rea EE OOM eg eeenSSEEE, TEA Oe ©nee UE A ee aEe2

ee i ee 2 Oy Fe ees. OE ee = —r—“‘_C"R to poe ee Poe ee ee oo oe Se ie eee ee a a eae 2 ie ae ll nr Pe ee BSE i eee es DS SS cs RN Da NG) Se as IRIE EE RS Cares Ce ees 5 ee a eres ace SEE, ee ge ee ae Bee Se hee ieee ic eae Pc EI SR Ep tac eae na Seuss PO DE, ENCE ke A SRR EIS riot oie eR Ree a MOREL ener Sepia ance cae a eae 2 ee eo | oe - — es gy ee RR aICRLN Oearea Sis SE ROIE RES reePaes et ae eeeia IUD Cee SU ey SI ars ESS ee USE eCeetia PRSSoe TSE oaReRe Gem esAore peepee ‘ee peas —. oe eaOEE OER crea eer Se. Se ee .ESee Hersey aHee Sees ee es oan ean aR BS WRB eee Lees oo. — LOR S-o . oe < SS oeoe pareees a Ree eREa PENI areas DORE Pees RO CDA Vee es, Ee ARE, RaPer SS aca SSAA Sees Ceaac at reentioe RG eeS ESOC mR Sse SpeeS: BER i Ae RSgee aaea wm

BoE was Be Wee Ge BBR MS i URC Uany orate cine Cai 1, ete ERS RCRD DE aaa are a io ca ae Be Ea, See ee ee po ee oe ee _s = Ba ome SE NOR eer ee eee Syke HSE RSE eeeeeee IEE nase esBEES eePra SOLS eae ee eee Baer ee oeEe eg — IIE RRS Doe eees Pang eySee SRO Mag ete se rary ps pentARe LE ae ReSaEe OS SeeSEEN ae aBeBe ei NOLS eeORES em Reco gene Spe:eo sees OE AIRE oy OE DE RE Cee le: RUE Ae eee ah nes Bee eee SOR UE Oe aaeo ae Ayee COSA aa eesBier AGRA Ee Gia © ee SenSte ee er Bish:Bn SER Sake acs. OR Aap ee eres ae PERC EPR, ee ore eae SOAR Le ESR 2nas, See aeER ili PURr ahereee oe neStores aeSONG eng a9. oo geeUa Ba GSE en BE OS oe =Senee i.. aa /

Fr EE cee ses ee) ee ee Seon Big RAS SEO SA EIene ea nee Se aseCs oCISe EUnet UeOE INO eis CR, Sere SSE Jae a ETog Boag3tore SaSE ee oe.a ne —tes2tse =2 eae eeeBa Lo seSOR CAL ee ee tice ae ee ee em aaa oe seg et eer oe ree LEU Tie ie SBR Gai etpen og EE Rs, TERE CRESS Me SE. Tepe ees SSS RUE oo TD Bee gre” 3AIEEE Sy eee eeae eeBE ee See Bee comma Se EE erOG ee EGS RD IEE Been eee Aa eB oe BF esPSS SSIS. ane Sea ane. SuSy ane miharitceneeas Ree esee Sess re ged ERG mee gees aerasot eR aeIES ce TES LOREAL sais ieee knae Sieh RpMa ae pee a 2.: : Ep SABRES ERG ere er See ee ee LSID SerSees naege iit, era Ce ers Ee ae ECR ea Ce| aERAS, aee eRSEAS eeBas eeEnea Sana aan Bekisc caeSSES septie. ‘A _oe SS Eaage AO ER LO I AI CBR Soe toetsoem eeoR aERe LORSEC Signet ens FEO Re AL peeRR tee? See Rimmer Seca hy eae WM keoo esoe 2Cas ee Ce ooIRE a5

IES eI BASE EY Foe ee aan es As SE ERE TERR pea tan Oe Siace eee cee be eed . Pig osSS ee re ee oo sepee ..-|ae-|_ ee ARRON SSE at LEN rar PienREG ise paar peaceOS UR eeeaeees Sg ee Nae BSA CtA SCIRUS IER EEO iaSe a As wee . ee reo is Bs ee Seeman 2,ean, cases Ba! emt ae gy ape ias eee- sy REE OeS OO OE eae eg EEE Es RRS UE ae! 1fg Vs ee Be ae Sete Me ae EOE OE aaa ep BONER SaaS LSPee TEI SERGE SARA ete tae anee POR IOfy ES: nes oe Sens SDBy , ota: Epsteec SEE RgAEE ssSIE Searle Fi SUR ONES A SRR loee aess Seana RS OR eas Riser eee, PEED SRP IN case aaaEN i BeSS OOS OPT Spe aesGa fg atneeETE aeBeBe DNR. Ge oscpes, Bae 7 ata S ee ESA Rea mRRea SSee ie eeaees aoets 2 ie ESEePee DES PEO Ra ESA Nae OE SRE CE ESSER aytd EOE EO ROE Re ge SEE ReER ee OER OEf | eee ha :eae 3 SoD SSeS an —

i ce i, ge ee 0 Plate 2.2a: omparison of 1952 9 I elite J

% BRBoge Gases osee SBSH pO ES870 SPSS OEE SG sate aRBE OAR IERIE SOROS ier gees BEERS GIS 2 EGS See mee eee ey ie eeeeRe BR IES aesCn ME es ey IESE ek ge OLESEN BE BORG SRN Lah Ce (By,USER ANTae. Rae REDRG Ta ee ERO Fag EE ES kee aBkee aSeeae eea ame 5 Bee agkSISA SS LSS Meo Ea hess BR oeBOR OYON AcsAS race GeED RRO 8 Ss OEae Seine tagesURES 8 OBE IR. IROCTRDESES fo. aeBOs Ree BEOE ORR Eaton ccas aeen Sta 2k cee heerrr ae

CARRIE Shia ei Oa Le asEE Be Sein tar Ra tte Ree ae |Lo ERSENG EE SOOO 3 IE ESIGe Ge NER e AY Pisses SERB te,EE SERS TERRES, Bptanhig RORBe seesSs ta er ca EO SR Re SF Re me UREN DEIN fe 8 ee ee SO BNE BE ROR heAESENS GSES ER Fs SsHELIO ae aati SRBe: ONE a ABee eeRe eeESEe ee ees aeen PEER STE Eine dhSa pe aS a;7;eeoeRSA IER, ne aa Ce ae BER Ur eS SSE SESE OS ae EEE ene OC OE: TBE gS cen ee en ae ee ee ces ee Co ee, De ERE RH RIO RRO Se EE OREe LenS ON See aera tech Enna ISIE PERUSE TEEEERE ELSE oki EES Ras notges RRRua et unre ener ON eet Ts ee BEERI CE Bey iatPT SSOe SRR ec aaa. Re ee ROIS TR BE SRE Sa ote GREED Se eee SONS MEGS 5Rae” ie ehae i LEO NODE Sie Se ESE ICES Loe igre menfsSs Se 7e cn aah IE TSS Ne SOE BREE SIS Seeee Seen ane te PS RSIS UID SRI IeM oR TOL Oy aaa aoe ac eire Siem ace SADIE SOTO RS USE Ee eeROE ae BRO Ce ee BE Bee eeee eeLe Cad Rat eeIEEE SURO SERS ee YG OS EERE ESRI EG OOE oy Bin, Sew LE Rae ORDOSER ie SE, eees cnn ene Ss ao ee canara ees (PER ES EO ARS ENR MOLES CN eOE SSE aSA IRS aBe ee IS eS TIEN, oe a aaa eRe OE PSU Se IE seg aba, oo $1: RES BE CR NESE SAE SS ESRI REE ee Pe ORE REGS SS See eR oo Lo ee en ee ee ae POS ae oe EERE. igs ROE ties Si. a ae Bie eeeM ae HESS ORES ES as ST aS NE ees Rees ES aa aba eee RS, OAS RR, be Ee Bae aig em EE CPOE ‘ee Beet Re ee SLES oh Re Bin, BEE ges Pee ee ER ESE Oe ONS eo Fe Reise, ga. Pees phar some ear ._ 2. = COC PE OR ag i ae eee pee _ —. —. i. 2 Fug of - :ok 2 oe ee hed EEE RE aes TEE sig ee OSU RT SS geERG mE UII OR SC Tole BES CRE tegen Mee EEO CG2. Seen oe oe RU Bsee Saas icAus + tnciy, SERRE Paice Bee ako GAIT SR ae ESS pipe STE. Tans cartoon ered PARAS SNe aDSe Na sag re, BRE eM poms OES SEE ee eeRe, Shae EOE EE, aee RS ae SESE ee ee, SROs ipps heMMe RE. ;eS ees SEN: TiS SES EES? BSE EMR oa fae Ee eS RIS EE SEN DR eenBa neae” RePegs gee espean ees - as ._4 oe, ee|Oo Bo eae an Regia se Svea Ein. axene Ophir eae fee iy Senne hes SIAL Rene hy Deen 4 $i ae Bee Reerenn aeMEIER ein oeeg SOR OL ERIET, Seine SES, iSaag oe eaR8em ae So aBig Bee Bis aOAPee ee as ee iecco eee eee eae PR ee «aes ad BAY de Seen eee tyPe aaa aANe ght Roa Pia ree BEG Re eee Tay ERE ta —_ Ke oF ES eee ee FOOSE EMS re eat ERS! 8aeeaRca oh = ‘aicork ico ae ee bs Pups, BBR ee ee OR Ey: BT ae ee oe ee ES art—“‘“‘“i‘is™sS—~O~CSCC eeEsBU Ay aie OR RNR SEIS RE REN eG DI RO UO ESNINg iireRE ee hr oeSO peee ee a.ae so RRS 7 GigE ee uM SS EP ATE pia Riees eae perp aea Sa ES OR SS aoa ES BS Sia eee

Lees es Eee Le nr a RE a ee ee Bee ee a g ee eee oe

;mages ; an 2 SSPOT 2g andaya north iSSI u " around 5 -west ssido satelli aerial photogra hs

°a

ane - : :- ifi} fe! Eee! mas wr a . Ledges “Sage? Bs, : er Bibi te EE EERE “EP 2. BREE OS SS >Ree ST a’

are Ih

igoiars, Fey cgetSacuee eee US Tesee Sipe BecA ae é BREE. 2B. ¢. .. 3ooSEE ES th . : . lesa ~ : weet sey sips Fires Weis .- aye ne cg |.Skgeme ys ESSUE TSEOD : : PTET Sa . By ERte oO . are meWEED 2 SSTM ee = CRE f@ a-Ay F q a, ms nee a. 2“Le : :SU ciawres, “UIE 2} .:iUR :Poon : Banged

. Wee Ea. Ben aeZep: Bea: .a -“hee

=+ . ini oe_ .Ee : : = . . - - ee. ‘2. - 2 TOR RE = 2 .

oan i FP . : : . a SM Be : “Sais AR : ; as ; 5- -..;~SRS 7 z peectlh. -F : .ae me a Lo on - 2 eS ty - ae RES

oF EBs as BRE 4 Spek spent GES sees a - ts3Do, . . a TE : ey q- “ee =a- =.: :Ss . - Pea SUD: we Avge Ra eaTEESERE Be | ooSE osee: an ES : Fae 5 wd ME

: Bete : eat Pa Sige aa Eee EO! fuk .SPE .. 2a Rend and “Soe SN PBB geet ee 7Boat Boa |ee oon Fe .ot bete :5 = . : TEs. ee ee ee Bee a& % ae eo [ie “eg a ae re aa ; . SS . JF a =.>: ae) eS i cot Te: 4 we a2rneT re Paced ate “2 :: wo :me Bee aSess *i :we *Eine, F i: Rage oy 34 ky/ tages ue, wo.iasiass Se be “4 :ey = RES og =. iA SSS ES IoUg ak wel Fygat vas neree . coe : _ rn ood BS % SEP SPF ligaeaeRES

Se al . . ar Se . i So eee ; Be: wn. . . . : - ole wea i ee i :mT, Fee AE be Be PR yang i ae 2 F 4B . CE os { * PBT -.oy .we : 2~EU: tg ae 7 oy . 4 tee a 7 : :

aera en . oe va 2 . .. wt + wi LEE BR Loe - Boa. Ploes ae 4 Plate 2.2b

;OS - Caen Ve hUhUmhUmUmUmULrc COCs eeaeeee- ee yeee eeee page eM cece. ey Sa SE Bul 28 US a Re rhc eee cr mes ee, ee ee ee

I Oe Ce et ER EL NT EE HN See a LLY FOIE SEMSG ON, LISDB ETESS SOOO MCE NIS, BLAE ELR ON, REN eeee ues reeves orem res EAS TSE) OA 2Do SAU ERS Ea eA gs CoE toe een a OR MSpate ee ee eesaeae ee 8 eg ae bp * * ee ee ee a Bee Ca a mee ae eae Pee Rae ene Sue ea ae EI ae pS iictce SR el eeRG eee Se 2B. 8 “cgay” oh Ug OMIA MOM aes Eat: Ueeee eee CSE ee Oe Be Se ERE aaa, taaRERE PT BOO SeSe Se LE ce BEE PABRITS Be one eCRgGUN ESS BAAD SREB i ae eg er aewe TORE MEC Bi oe Se 2 eee eA Ee Bg ee BeBe _ BR 2 RRO Gig UCAS EN ii: YOR G BEES Sas. eS one Tae Co en ee ee ee eee ee Pe GO ale. FRIAS Re ESN REE. ERS OBE oe fo HBR. oe PM So) Ter. Ca ne a eS a Ee ESR PEGE ROR Bee PP PEGE FS oe eg trie Ge on oe iis. Eee eee id LE BRO eerer os aage ee So llee Bee 8rp beee ote, 6tecaenc SeeLe ee ee eo eae Be esa PORE BR, SPRBON EE. ee EC, BAe aeaaron eee ESBOS RS RS Re Sacre cannon peer a7 aeaa aS Sg Se .eeaGee we ar ee ee ee eee eae ER ee gee OE EG IS SEINE TEE SE RS ES: oo RR aaa GR rey eaar Ssee a eS, cc Eee mM MR eA 1EAGER nck EES BBA ES EIR eo Te Hs aon ee 287Begs eee ee ee Eee REGRESS. OES. Ee Oe Cs Rae aSe rs i ee She, Ne Be aeeFEE EG ee i FE Er en aiees oe Eile Be enaty ge BeBe ems afer Ge Saee Bog peea Sire eee eee eeeR 2S SO, ees Bae oo 3 Eee SE!REE Rs OEE aseI ee ee Bee:hg peeRe ee aEe pe a ee. eee ty Fa Beg aye ee ee ee Le i ee Se ie te ee TERRES HERE ee: See GEE i EAS ee ae a Bee > me ee me Oe, SR Gg BER ES D0 Ss NR Ea Ft aoe en ee ae ae ae ee SESURON Gautier, ee cae Be, cage < RRSLOT wegaER as SE EgeRee, BEEBB ReaiiesTEES Be eeEs os aRivera. Mies acetate BR Bicutteges Hee Se cn ee we Bia eee eee Sesee Grete Ee eecree eae ee ees HES searae smeeaaapee wre FM BERERCRE R25 ek ae ee y SE ar Ca RS LSS, Bh ABR SUROEERISSS SERS So Re aOR aarnnet nn Se i fe eer ee aie rn Buses po EEE Sas ay Yen Be AIS na ae OL, SR MRR CL EE G/B OR ets elec 2 aoe Ogee Ge EEA pe RAMS oe en I ne a a $e BtMc a a ERS CG GRY eSeee Be thIBo Bek PACE SEES, AEE SE Ee Ee EES SS 5 ECEEE pies Re 2.2 MHESES eee Se a EE, Oe eR Eage BEee amen anee os gtk. deena Ono Meme. ee ott, Te ae Ge cee ee EEE. oo eae EIR 2GES Oe ee aeeeeeBRU SE OR aie aghe aeeee3Be es a BORE 0seSes eeBEER, CTS oa eS eat oe § ea, etewe bgpaci RS RR =. ON. Tapper Ui GE ORT SOLE et SEWR EEE SY, WEE I RyeaE CEL SEED GesRM SA Beep E Wed Saas Bees ge BR Ee, ESS ei Ry, SR an adeee ene B kets CE HR~ Bm UENO RRR REE ORE Solo ty, | Wh ORAS Menace aan ae mee ame re en fae ames... RR 3,RePES Ses isSeseeys RS RM RR gee ° ig OSS OTR LePOR eaeG Eras IEE ree BS RE ie BA Ro esUE AMIS He OO ay Deere eMESES ES aeMmmm 0) aBet ae a OER: ME ge on Bs & eSee, aeeeERS a1et ee ee BER “Et aes es PEER OR toe, Nama ateae Se eeSSE eea ase ee ee eee ee Eekar eg SSE Ee Es Panes Ye Be Su” EERSTE ie ake =eREE eee 2Bas, LR eae eee Goce ee ee ee ee ee3See GG see ‘4 Geni eR ekCee E. REESE Se EER” Bette ot SEPP BEeee ae MRCE ry RR ae inEP neigh ACRE BENEC, Biiscnn EES BAS eT Geng OESpre ca se EEE A emu RN ESiA a ESO bee By Ey af es ala af°ep a ee SO So EP" pee TEE ERRa, aESEN tai By ees SgBok, ciai GES PR ANray 8Sti PUREED USE eS ae eeee eee? EE S| aea ee

Vo ee eS OO ee ee ee ee a ee ee LL ee pe yoa aoe oe i Se eeCree eeaeSe oe iee Vere byeee Gh eeEee ed REPasoo Pe : ; ; Eiger . Oh aoabee BRE to aSiG roSeee eee eewen neRSG eee geeeeaeee| = woeSg re f°Bigs ice eae goo Raat ES o *7Peres Lys PR eat gaee EEE RR TEgE, th EAS fo RR eeSUE Se ee sae7 omTsepeg vee Bak. eo ce Ee ER EO a eae ee ee Ce ee es es ee ee pS Ee eign SE see ale PR g SOE ie PR Ge ee EE hg EL ME RE eae EDEL Pe ee ee aS om ee ee ee Be i ae ie Be OR BE 4 EEE SEE EE # BE Ege Pa 2 pCa ye pone FO PR Re Ee OE nS Re Bg a ee ee So mee a ee ee eee eeoaGSee ae phERS BO OS A ag a ee! BpEe GEL OkSO Eeaeykcea BRB i SERB aReerie a i RE ae Li a 5i ERR SecEit ER pocket Bee cePan Ae ae kBE Ba GeRS gasSeGa Mie ee stiePee A:RD «Bre as SE Se PS Ss rad iad Se ee EE ode RSpps SeesedokTE LPR: Geeta a tee SN i= a ae BREE ES nen FO EB a ee EA EC hg gee a ae ae Foe pep

E at OP Ge a ee 2 ee ee a ee

Eee BE ges BE ee a Sees ee ete pee oe ae ig Lag mie Or ee ee ee eee ee ee a ee ae

Bes Bee Re eae ee eeeeBes ea eea OPES on oeOSee eepee on tiewee Oe eo ee 2. Aig eseeaehe BM PES ee ee o OP eeSe ae i yy a. ee eae ee i peoe ee HES, ee ee ee ieRD ee Bo Egy OEE ee ge: PI RGgt Ee ee a PE ae Boe OTEee a ee. Cae

rare ele RR as Re a ee Ee Gig BuO paige BN RP ES Peg ei Ee, ¢ RO Se ea eR BLED REE GEER GRUES ient « gmimiMao og SERRE SERB BE ores 8 Bags es BS Ge ait EERE.” SS ea EPRI 5 BBS mn DS i a Ls SO CR BEERS Se a OR EAS eet ee 6 eae, aie Bs OS PEGA LE SPO ANG oe a ee Le ae rw rage oy SELES 9 aE Baste Ee Dae 8 re ee SEY Bee De eS See en ee eee eee oePR a eee &Cee oe TOS gEOE GRRBs Te. San Ss,Seg US PELE SEE, Catpea ee Bie8(PARE PP yee SiS07 ae oo Bee ee ates eee. tO See ofEN ae Se EES GEE Eee eo COS SEO” TORS atin 2. Pe 24 tS ERE *ae Tey OS Bg: EE cae eg chy. Se a ae ijik, ame Rae RARE CLRRR eeesee asO Bs 8I URE BEES A ngs eeEe aLe ceaOe gg oe EAM ORB EKe 22Ls, PERE Samat * 8d MOREA SRE, TE IAS OIE aSBe emi, le kee Uiky GI BOR Pag SER iseee he Bey Le ean aeeGE Eee Te SST. eeeeeR we SE ES jieae GOT a edTEE PE Aase ee ESSERE? FA Bp aeCE ORS aeCORL ae eee reaeAeae ee ae ics, BE eee, aeRC Ee aEIiBU Ra Fe eeHee Da LE oe ee ee EEa gee Ea ee GO RSit EO REA eeRtEe ISeae-AR SEgeERS RI CEL RTS ET! AOL tips CG RGR REIER! ELIZ? SRa LE 2Be RA oe eS ie pee Be a OE Sid aa me arnEBRE aieee PaoFE FeSEE SAMMR ERY OE og OF 0ODA ge SOLE Bee i 28SR RE Ea caer a roaSee ie, SRTTSTR RSIS ER ty gE LEE BESe SLOBC aHage ies Ee eee aeee ee Pe Sie en ge Gee ee neUR FEE BS PIII,oe 4 6S Le”, REere CAE Ba ie TE BE ss Ej TERRE AIRES 2 ETRE oes eee ceaRe omas ee ee Ra Bb ie: Bigs GP ed EeEM: Beea EGE DBE LeSeLeek ee GOGR OE cae CRU aoeIR CneBY CSEE DePee ega!a Bee goTTEe ag GL IOS a3Ute ccsaeLees aea Pee ee Gey ae eee eo

EE ee Bs ae ae eeee a ee a ae en igs ae ere eg ee, a ee ae Ue, ie ge ae Pe ee ee aa ee ee ep eeOL ee EE ee 2) ee ee,

ee ee) i ee ee er eS Ba ogee Con na eg 2g ee ey ee Oe a ee Be ee ee ae eon ne Oh 9 eee ce BE LUCER ites. it, Soe Aig eS ee eee Ge Oe OEE: ye LT OEE ILO GE |Reios LaLae, ee«RR oe Ce hae aeRRahie: Be eeeeeeBRR yyILL oes oo ee eg ieeae Se: meIS ERS ge DSP Se OUeee, Ric BN RG EN ROERR Bi Geta eaeoe BP open: ee eon apeeae Soren mee PEER nae ia, aCae eR2eyeae Gee ee eRMRE Ge LEST ane as preg TORR ne ey ee conker Be, IB ee Merrie Pp Bee SAREE tiegee ey emp aee ees ee geMae, a ee REE teEDU Basis SOE eC SNE Ae SER ee Be geenna ee ee eePe Se ea ELE LEean. EEE TE CLE TEL ROP? ED EREE SO, AURA 8 ENON a ae inte De RRS SCT Re SEE hiAES a OL RC egUs ET IIE EER PEE GSeae Snea®, Se ASs Rego eeme DES ee ee ee) Ee Le OeeM bee eee oeBeEE Aag RR: «ias ee enexRASA aERROR ae ee fee age: esLER SE aR8sRe gs Sekine: Bei, ie BOSieae ey LOE naMutae een rae aa aS aN. aes Bae iso" Sai ae ee ee. oe Fe eR AES ee ae BI GM aSees gee OP ee SEE me RE SE as aes On ae te MR eg CSSee DS RCL CSYO Re a LE TESS AI SOT CSay CB a ARM ReaRae 6 Te OE sig TR egPER Le CGos on aee FO OBEOS ESSERE SozaFERS |ee. 8RE, EEN tee ae ea BSED TEA RIIBE Pits: Satay eeAEE a eR ee BP ae RO Sige ESacaORE EE PO REE OES2 SAY IE iES eePek ee cee RRS SSR eee eee EERE EE Se EE wn ee REE rn oe iRe AE OES RE Mn Sele, OE eS Bie ee oe eee We lagi Bee 2 UR RRS “age Be Ss es RG SOR OREN Pitear smear, Be a ieee Ee ase ee oepee He pee pee ee Seeaee MN OsSee ie ANN ae shes SE” SE 3"Des ge aesae os en wnEAS ee MM ea ee Pee escarasaccent 7 Sie aea es i: eae Oyeeeee BASES os Ea Bee ee he reeeeeR UeER EAL DE a TE SS BE oe: a: en aa?a ee te .okSed

Leeee geod ases aa Bre eeFe uae CO nee 20 ee eepe eeLORS eOE ee ee aeeam Ee ae eeeee eepe La ee oeeee eAOe ee RES 5Ce Ee a(A er ec ea oeee ee RS ; Lo ee oe een ee 8 ee a ee “| g L ee ee Pe OO ae IE Site, EC ae “Megas oe a Pe oe a a eee i 2 rn 8 UR A is ee ae nih Ee ye Ee i, OG oe OS SHEE Bee ye ee. ge are ey ee nn a nt a By Seg i ee eg ne ence RS eee Bais ie Bae ee eign Ree BGSOE Nia SB Sg. 8 ARE REP eed eee ee ae ee ee Es ye ee ‘ . en ded ee oo EEE Sehoe. HBS tage dee, RT a GEER CRMee Naren: feteee ne we 4 I SO?Samm eS a ee a ogee, ¥ te. oe 48 Bie, ceBae eyode See OR| B.ee aeSESS eeeScape sf ree SLcoca LYSESee, ae ee eR Bien a eR eee”: : i an Rise? Yai “Ug EE EP CORE BRE Fd SE BU | Bey ny Pen See Tose aR ee, . 2 ie, cee mr eo ES RT oe ee Re a SE: ROSIE 7a 1M ses 4 Fee CREE SS Ee REE BE wey Sees se i a EB gs 0 mee ea a a RE RB OS, Bodh er Ue es ie = ee ee Be... ; on . 3, RS.Se BB ERY VAR cate A ee JESSE ASS Se eee Ses PRR of Dorlas sy i at ae aed: cao 8 icumnie in: & SEMEAIS She, Vets SREY Tes UE ee A ee aa a 2 en, ie ey ee © a7 i: Be “Sees. age eee ee Bus oe Beg foe a He a ee ae ee eee ae SS SF alt ee oe z 4 . z _ Bede a .

6g ee A eeee Be 8 eeeee epee ee ee AR pe a oe, Bg ee enaci Be Boe, F Pe.BeLi ageeeae. OP a oe Peaece ee ene 0heeBo afyi, aa eo 2 2 aa sean HS Ce 7. et es oe ae ee FERS EE LG ee ee me PEER ES OOS REE OC RRA Pe ee os e F a

PRG a Ne. ae oe ORR, ESEee Sobseoeeee 2 ee CE Se Prem ae: eepio SOREEe. 2 7ST Re IaeMERE pas Bo iit Ie, CEB SEES.ee b ReOe 1S ASE Seer eee: rior er dp bos .Rouge 2 . § "ROE8# e . Ee a ceRIre eae eee te Bee 2 ee ee i LkSESE a i EE” BBR ERS. the AE PBR ee eee ierceeeBETIS ae 2Eepipe +

OR GEES peer oDO “PEE BES.. ., SEES Rape oS REBF RPBA aa ee aeRea a FoaS 2 age pa Pik ets Sorcerer Se SES teBaek i Se gad eaccae ad Pepin Mess, “tae: ES s aiken WU Be ee cmmmmariee. ee tae a RO, Ca E, 4iFE aage ‘Wey Beg Ss SSESea Be egaeBek PMR S57 teSea aBot RI Ser PRR RCC enCRE See, aU.toiSine Recent Pere? SE, Sree SEae“5 REA aERENCES Shee cite, Tas :ewe hg ts. .Pg 2 ee UMBREE *ae SRR 3 Bice: age ce RUECRE SN a eameee seiien isnet tioim ee hace aes ae coe EESRR EE PORN SE SO Lene aa CM RE. Eee wt &&

StS CRM MORE keeee Ste ee otek eR SSRs 2 BER, EEene ae RSS cencin aesan SeaPP Rs RRR Seager te Reke NBae eRe RNa RN ae CeADRS i TERE ae a Bae Prunny oan anpE mmncemane kage FS ing 3 & ae oes aa ia ee eeineee ee Ban yaa Can secu aan rTee Se Jee NON SES SOE SR AUN BR pe eae RR ey geaU Behe tefits adhe # By, S,Bs Lae Oe A a4ge ST Se ae aise ee aOnee ee se SS ES ier aOS Be ese aeeos hop Eis aes WR ee eee eee eee hw aEEee Ne aEO eT ea ae aRgae ERE Ss, BESS yee ee gees,tee ee com tseed ye hee aeeae Poa 5ee RSW Rea eeMRE ee ee ENDER eeepc saesBeene ES Ed Ea Soa EMBO, 2ohee ae MER eeee aeSSS aecits asEe ER eeNERD ee ee aapi Me ec osBG co aA aie igBo Re al geRE aes a aoa. Seed netES Hee eee Se By tae MR ts Bee eteeee Ae ek, SER er aeee eenO ne em aed PE GRY RRR GEG, te ee) a a .4 :; Re a ee re a a Se Sem Ce ee ee ee eee Be ee ee Lhe ae ann 6 6 6 a pe ee ee ERR OE Ok oe Mn, eee ee ee py geste wt, S Ram ee pene TER ER : Re LESTE, gee Mee nnll=CO 6 ee Cr ee Leen Cae Bee ee Pon ee a ee sss VeCOU ee eneae ee _eeee Lak 8 fee es:lhlchc t— ieae rDCLUCUCO*COti“‘ ¥ Z . = ? e 4: . F aaiF. >i;.ah :.:2_5.t. ;:~:‘:;:£i Ls . : hee 3 3 :°“4a . Uh ce aa i : hao Uk fee : — oh5 Linge : : .sets oeS i:

= ; | LET UES F 4x#:

=2 “2 Ulee B.a UR , BOOMER Fy googe) ESE x aece OS wget So Uiph tns Sor .- : Bes re A.2ce gee OY : wg Aooice BPt Pog geass FOSS AER . 47° . , Bes Say. ge ABRSR™e

ea ay . ae . . : . ge Bg et Ee at f

Pee ie an - es ; . : c 8.2L) Va hia ae garde PE : : Pet iacd : Pa : Re Bes fos a SA ap! oo", BAP GR GREE gro eek SPR ee SSO :

fee va a =f BF Liege Shp Seofo STE aaS ceIRS egal ave sy TOS Sass TUR Ar ope eee ==r£a.=:.-.ao : +rgb: . =‘ Posh 7 wget 3asODE ose cin, Sag ens ain J OEY eRa5=o Eee SSeS Bg Bt

pe F_BO vo :: an , ae aot Bo ea UU Brga.Sepo eeeae SE Laise : are otee ER ee SeoOT, :SE ayers i aaa ES Se Rie oe &i ke aae -_ 7gl oo EAE dee eetg2Ake SERS oogS RR AR UR eke eefpae - EBT oeae-,: cr Plate 2.3b

ee ee eee Pa eo en re ee: ee . ERS fo ee a a ee oy. ae Ge Lee a Zo E SIERRA TSISMIS SRE LAAT bas auabatilacCokaaaet aA RA Nl a REESE NEAR ME LSA BU A AE Seb ute hes aes 0S 5am aaa ai roecermmenty nad rine = ba i psoas ei: yas . * eee Rie ree pee mone sous ;

BU ena, Rabe ie, OUR Gen natSeBak} By a CE on BES ON, ES aa na einer ame Ryee, Cyne SBaee gmcmene wn SYR SE “Boy Felk eee eeiSee SB pO Ree aSN SERN tinn Biliige ae: Pi, mOB yhFey BE REPS BB hahah SERA i Ree Sseae aetn, hePeeeae, See Cae HBOS SS eee SS 5 Rs gSRee EM Co ee NE NES a Hg ROGIER, GREee SiH, Semmes Heeeege BR Bibi geek Be ae ee ee yee ag te aaa ais Ree See Be aytaes Fle BEERS, cect ig x sgh OUR Oswee Ok Be Sia Diane eeI|e rag” oases RRoP E RC abienta “ise

GS PERG coe, SeBe2S adBOURe 8 ie eeSeee ee eeAe ee ek.eee ee BL en ee ae ee ee ED aI EI a ELE Rags EO e ad, £Buoe eds CEE CERES ET a2SPFGES oe LIE Dee Po SONS De ArI,Soge ee pe eee ee Be Oe MRC BERD aeseaeae *BEEBE a ee aPetes ae Te Gee ea EOD Ge IUN eR ae eee RS eg SES RSRa ESR PaB.eon GE EF Sah ees EIS ec BRS eaeerSe aRRE asaELE 2 mine 1BSY ee tan OE REED REED wR.COR We bw mien See BE Se SCR TAO Ee aE ORE ee Pocccbsera Seonee sa BSsabe ee Ee eg aN Se Eg Exe * Fon BEG SESE ts ooSS EES RE PareeeRR eeRE LOE Shitesto Fo eer 9RS ORR RS AO eSBOeSRE nyu*ASR REE a COS Se eaaeeen Re OREN Carerren Since oS See re RORY ae co

Sn eee eee FY aRARE eM eICameaBee,aOP IIE SE pn Ty aRRRce eR Bee eeeeee a bee Peemee oGURU BEES oe, Le eeRe eeUeee eeeee Fe Poiottciarenni, SES Ty dh Me Med pl, tEAMMe Rema RS ee aRCe nee a aoF CaaS, Soa nee ere ee a reSe ee ke, _ ce .e ee PVE eee eo ae ea on ea ma re ce eee GMMR Tee Ly uf ea PG. Ee AS a RP es BE as Ie gre ea Ge ane oe, Pog ean ee eae fo Bae aU ee RES a Re Becca a ae ee eee GSEs, BP SR I Ae ee ath Bs ERS, MERU econ eee a a Sete ae cae Sie LO i cs EE Ree LEED pate ME age een gee ee ee eae es ; me ee | :eoPo 2 wn dy BE EER EES Paes Boe. EP Be Pee De ie es ee ee oo oe oe AES OEE Book ee Bae ZL oo hr SME Be Pees Ee Sees, LEE weak SS oth BaP bee ee Sees gees ey ns EEE eee Oe et 5, DEIN SPEER LEB ES TS PSI OD ET, RRR RELIC MONS tins SHER ep e. 0 P=

BR macnn TRL | Sees eg Tee ee placing SOUR7IOE FANT eC RESe RESa REeee sh Saab nCanue ce, A ieriea BR SRS ER Sc ee Oe BET ad BERR Reeeee Ee Sek Seema re ig es BRaE RE peeSePe ee FEO a reeueRs ho EES SERof SRS I cea ng aS peoi are 5 EES eyenaenn ene.Ammon Co, RAMEE RN BOSSeSete ag ain NC eeLip Cneermee nese ER BRAR. iee.cuMera” Ieee eRe,

%

Bohai: Re! EINER, AG ts, RAR ENG eaeiotiuatie Mei cths 7 TREE Sati ER ae Sa a eae ee aR OM LE SP gO LUC eS aaa PY Oa Sg a USE RNR RE BGs Ree SO RAE SUNS ROE Mn BRC U UG ie UR BEN ON OnE Le Deas coy Canc Rieter na ot

Ses Soe 7G, EE Re eee aa Murer iinee ued, caine, Amina enna cee eing Gos OE EE Ee Bi ce 6 gp. EO LUO ORL A OIE POT SOS ROS RRS ERR AMEE BO epee mano Bey teeta hae pan eel, ire LE TE ON RL COTE REA OS RRO, 9 Eee ROA Bi et 4 By pies sR Be eee ee LEE HL as OL Ree OG eas le ne OO a RS ETE Nr OU SP AS gs UMM MRIS OE 7 COINS BSc MB OSES MR SEC EI OU GE

A OME eiORC ee ee me oa ae Le eea 7EES Beas ST & cha eNOS *ey,BoLR Ge yeeere ap ee a aayger sn RRRey 7eeGSE ESR MR ORR NBG ROO FsOe Be BRC IT a RAB CORR ns 9TR BeRTE Se Ve MRR Ne OEOO Oeaan ARE PEOS OR OEELIS MMR ROSE RR CESS Be es LR ER ieSe heSel ESAS oeraan meneas Gath. ER oe Eee ed SERRE i an,IRS Saheap iSasea eae eae ne Pe, ea AR i RCE Oe aaa SS, EERE CURE “ie AIR a alegre ree ae gas rag iar mein ONG enanaa galees pupae, ErieEOS eae Gi otBob Senta enOR oeea Re. RERASS RRS Se RA eRe Seca a ne cr cor RaSRI eeeSee See gahi nent te cane weap man UU i pone BER BO ieee ep ene SeeaCe pare Sac cae ge EESSURE i Bass er ifeeBeaAEDS RE CERES SUNT OTE Li ae ea ee snl SE SSE Ag a aeerSe LOSI eared eeeShit, SEA GL, Pan Gs Me egSacco Ce Hai ME air Sa Guin 3iaa EE 3 ae ROE eeeSEeeeeaane re Oe ae me te ele ie Lg SE i oeBE OP ERT MT ESR Goce CMeee ME ee NAR ge RISD ae EES Cais RE LORE: eR SR EOS Lae GREee RRO iia ea NE Tse SEER ein oa BAM oe nae PEai epRS er EBee are eaia ECS SiteeseeSOR ee PB pre eee lea I EE I DS Ss ReOEBh ae A ITLL OsORES aa y oS eeeREa RM es GAGES RMN RateORSER SUE SE COALeyCEG RAN nUrann aa a ERA AREER Cae Seeee PLO R E Riis, Sanghi BERR i nit Smee Rice nig eg rennoede IEE: eae oes epee LE Ss Se OE NRE Dos oe ater Teme, 1h Or oy Bir cb oa epee CSE ERTL fat pies iON A See E OR iia heim ter angie AOS OOO eRe? SS. POR aia, Regen AS RUS amu Sn Ca eau name eae

ER Ca Ue OE OE ENS ORAS SSEEO EO a OREIS SORE SSEBM SENS st a SN cis eS ay GBR RONEN GEO MOL ali ON ie oe eee Se Oe As SRR AR Bet it GT SIE neEE Ser RAO. RRAEE aS PTET NO iaeae ORG ROS RESIS SER CMR OE Aeagi tit, | Re ie men ee a oa EB Aig ROSE | Uae cee Ga EGG: 1 RE A RS ee Uiae mere LOSER ONen hy SESE REISS ES Ee Sei oeas UN Ba aban agseiontag ae Ba or me ee IED EESaS IESREIL Bis: PD RIS 5 AEE ee Ss ieLE RE gti Re ROS ee Sg RN ere es Fis i POSURE:psiats RR OLAS IOI SEFCR RI ARGS Se OT aoperas ROTTERee eg a eRere =e

Tea ER Pe eeCaINI ee PEE aSesoe AE er NS ERE Ooi ee Me eeeeEOSeSEE EE ee aeREMIee ES OR ieeaMR aGEaSyA TE YEE ae se ee ee ee ee ee Eeoe ON Oe EEE Rg iRO 2OMR 2 RENth ee eS See Ren yd OIE OO aR Les Hee 2 ee aRa | 3TGS OS a teal, eee peas eee TAGS oe PEGIee Se BRE cia oyINA AERC HES Oo area Renee eiere Tania ee RR FE Lita eet eae GE!TC FR oe EDEL Ge Ee,acme tat hESN RRND SEM re Oe TS Gia i Tia te REE arrSE SOY pee Bee | pe LEE ERE SStaper BIRR pe hes OE Ope a Ret hr oe od ail oe eRag oes RLDR ROR reee Me api ae SERIE Siee eae PRA Aee ic ane Ree nee Oe a ag! A nS, csgessht tsee SY SOROS CERI os ee ee Pe aR ERO OY RE 0ITe RR G1 TEs SUPRISE ES IE ngseaee Or ETA OT SIE Re aoeMe TS Re RCO Re ne ame eR IE ELE ae el ee Rn ea Oe pe 2SRBRS ht eee Lupe Bee 275 BRR Ee RE RR aSsee ee Ce BS Oe ge Se Reta ee 5 ee ee eeRae ee ee ie LO REE agRR EE ee ag Oe ae RT SO ees IaaereTe EO EIS EERE ESSE ca ipRRR, Me OL NL AIS SIRES aege DP 7 BaRR a= ECL Og a RE ERE eee eeetee BE on a Lo Eeaeeee ae NEEBOS Saas gM aGh a CeJe BE CGE. een

MRR SE Re OR Bia A aRG SRL RR eed RUB UgEL sp geae, ap Bi 2UO RO EERE n RR ayayogc PReR tans aan SEE eeSirs DE oe Rap otBaeTE IE cee Gh I od aeCeO OLaBeen aySaies Bini SRC ROS AEPOE eeeOO ieope ees Eea eg IEE a ea Nee LEE Se PAE EILER LBD TOTS =” eae PeaS bee aeRe anBEES OE ELE Ceanie, Eg Soe enates: eee ye a hee Go | Sage Speer RESET OD aS CLEC SRAR 1Segre teeDe ROR RE a aR 2 SET es PMG ones BN ne n eee enSEE engae. SERRE HES ReSS MiMirae aMRS Een Te pnd emer teGoa Re BE eget TeBe ST oPRC RS Bing nSaeRh a5Sty Se Bets eesLE figs pean pane ggemee SaaS eoaus coe eaGe ER as s. Gin gleams Se EE I Bt CLFS 5,ROT A hee a SNS ea aa LET es Ue RE Eee, Hee Pe BAReOe RIESE RE ME Dis ae ee, Se LEY id RG LOS pp SE BIE 5, IR Be OLE aeeee SECRETE gig teBe aeCE REear OOLae EN SUL ROU OLE OI ES NS aaa OS CIEE atOe Reges RsPega IE ARERR eR, ist CU eeeee eaoyeeGLE aLEseeci RE iyRA ae a8aeDEG Bs Raa ea oO Sg a |

Wee Le ee ae A Lo i an ee ee A - 3 Zee age Dea eee ee eee ee aee eeNeg aes eee TD, SB Be ee ee CLL ee Be ee ae me ee aee ee eegta Ch eeet a he a ae ee yg erPO : EE PBS, Re ee ee ee Pe ee Puen a Be Ee ee ee Be gas ee eee SGT EERIE. IGE ae ees ee ee ae ee ee” a ee ee er eS i a i ee ae fe ae ee ae be Peeio Fo) Ye, me ae ERLE ee LL ie ee LERRL ERRR is Ef OT RSBieR CN OR ETE RaORE LLAeREE RR OE BS a ER ERES LC map ane ee | SEE OES RIOR. Sead. oePN RSCp BRON GE Oy te aEE Le aE LOyee LER RO LT EE ae te eeegS eeEs AGEL sserSB SL es on Rep ISD ee RE waaay Pgky EIR, EBT OTe BE AAT RM OS RES gBEET loee i Oe Le Ray A eIaA ae UIE ROE RO og LN EOARRORE Ee EN SO REIS OE Sh ag A USL BON SECIS ELEe St cE GO EO geGEL Git EE Beat PeSee eT RO CL OE SIL oe, dORSLED RE asHOE, SR ER a eo nis? 3S ae Lp GaLERene RaOgi Ee 1 A aBPO ae En SRE a,ILLL aOEa LE OE AR a aEe eR 7 EO SeUR BE Maa OhRR LORE CMCC PISA BEE Hatt: RRO AL OS SOE IES. EeNg 2iSEEG SSE ae ER aaa RE eR Ee Bee ILESSEE) EG 3BhesASeTIDUS ON ea Aeee:oeesLR oO aL ETE TT OR RR OOS ae ees geLAR SESE Eng SRR BEARCe Sa 3 aAS, ee raeee pee: cee Ds BIect OI Ey a OY as CaM Once, Paes SRaRet en BS oi anER ney eae ae BR peneLe ane RIS i SRE aE BI BRHee eae ARC. aggee as aeCe ee ea aeSR Bais pee aii SO Siac AES Sie es aa a RE eg aR ERE AARON a NNEC RE eR ESLER” RE RCO 0D Oe nage eee een Ug aac RR as a Lappe SRE eg Se Rg AO I OR RN eR SO ee CR

age IS SS ERIM LEE SE SRE RRC GRA IEEE EE GP Me Ae OI ee as IN eae ee oe ee ee ame Se ee ee eee

EEE SEP RES ang Sy EE RGU ee ae ee ae ECR os Rae eens ES. BEE PEis Bee amegeaeae eerie Stree Sah Sie aren ee cee REan, aE Lo Aa ERESE EL RIal RR28 RRR REaera REC Fe peBE see sy20g I aapp PhFEBip Ei CEE, IEeeee ge Rr a PE TOE BEE PPaera aIaeee” ee Baha Bk BREESE Sag SedSEE NE EERO Ohi eee EpGR See

POE geOEY en ee Gt eee a Ree I La ae reeLE ae i me EE aus Bip Bilas ne ge ge Liey i aege a SB a tien1EO SOREN eG ee Re RSIE MGR Te Pa aE So rb:Se es ia RR HESS | St RR TOE RI ee LE— HE EA aaAEE cS LS ee ep IeeG aeRS Epea ei Se INN aa Fe5EeRR SES BobaleGaggia its ae SAR Eigaeee eee ee ME sea iE OS BeeESE eee BES Fin BESS SAR gee ore RRR RR ame areeg RUeS ee RE ATU Sic EDEL CERN eeiy) ee Tene neeea aioBN onc nce Ca aoe Pe anACES noe

Le Be ee aes Bi Be ee es ey ig Pike... Og rg wee ee ee en OU ae ee ee eee ee OO 3

PRE Ae i pee ee a ae ee, eg Sie TOI Es, REE SR NEN Mit BOS ERE 8 ve BERG rigid Pb Re RA Re MLM es ea ee ecco ROO eae = RR MR

EY ME RyrageTE ge cei CERES fo SE San RIS CeeeMS oyoraLe Be PR GE Pathe DEALT cite ERA RR TIEy Ae aim Beiteea RTA : eirues eeeeeS: Bip Ge ee",PES Vine Ripa ST on es,Ke Sey pana eee es a SSS Le RRS Socata meason er omeara gi: PGT “apmareSRI e é 2 SERA Bee Ss agOO acl cess eg W ay Pte Meee pa gn DRA : Bla aan opty RUE Me Ae ST Ree Pac Geo aes RIB nS 8S TEN A Seg ahi prep a iain eg i BR RE RR eRe er OEE: geen NS Se ae eR Ee Bet ne Be eit eR agp Su aaRi DUD LRM tOP TA ig athe oe TS Fy oh ERE ER 5 FS

Se era RR eS MR Oe ae ee Do Sn Oe a LEG ae LY “DE BBP 2 PRGA TN Seite Boney, Pe eR ER Ae ee ee ai aa oe TIF ie OS IRE ESS >So ae. *goes Paeee es oe yee PRB er EEF eS Y ane oe eae ee 2° REDE Ben . to” appre Meee ee8 aIE e “spd 54DP a een. Sore EP ahh . .: .cag ge ne, . 1EE EE BF#0ORE FoOe nn f; 7Beg oeeeBe be ORES BG BEEP Gage a iia. oeete Bee ByEE. BERS ae

Bel Fy a i 7an Ss RES eg, ROR: [er ae ee | Seutube: ~ -OO int pgm agae ay .i,3 ee Awg, ee a hgh" Se ee Seer Be SE 4 gee aSa RES a, RS She rsEB sere bag: eae eeeeOb! ree f3 igo. ae nee « Shder oe EO #BtpRB =BER EEE ee ee asota eeYesee: Mn ee reeeae cae aera aes ae on ER Se rgd I Cee Boggs EEge neetyees BE a Sy elesge CEE ee eet, state PERS ee

Bo. Bit a Ce eea aie ek a ee ei Se ea RRM REaereRS eM er eeSOR LER dig. TERESA MAGE 3E E.R can . a i Bis aeect AG oeaA eR A aaoege IPED SRS Ea SS aaah ee ne gees ma tie eeEe NRE ek i O jan pe ry w wy aa 5 3 $§ SS © NS nN

o)o om dh me) — —_—

iF 2)

8 ee § ¢ ZS s & ae OD : os o & & os

Oo © —) 5So) eo3qq006~«

im n Tm To > : " “A SS = 5

cE 3 oo g¢ ¢ = jo) SS OO =| 1) © ~MwBAS SUne amas AH = oa0 aes D ray NOaN wo yam N wseagne a — ome vVn~_—

3 ‘ast: . wv Se ho Zo 3 oogd Ens Oy Oo

VY T M65 5 > aaVn, DSW tono Vvio Q S224 ZsSsgMa

oma = WS x ‘wwe’ Faiai,£—- £2 plo“6ofOoELS wn

ro) vy, os cd = Faw} oa fae) os | = B = oS = 25 cS S 5 3 S op 3 ng g 8 aasSete - g

Asses TBS Ses SHES A 3 ae O go M oo & eSsovy o = OO 8 oOo oO oO os o) Oo S top ob os Ob 00 OH 8 s of = Sosa Gees & Ss) =2)awiGo = Sos x ESBB=5 =

) a mS 3S SM oe -— —O 7) MM > a> > sPtfan} ih Om He bY eSo—_ >>

2. a a n ZVUD'S «4 w Y a cS van cect eon VU UO 25 Sf we ab 4 e%eo Se< a LR && x -3~ BEpE aonSFSEs £EF5SES SEs & sa25°95 se O egos ododesda EESSEEEEESS a) wah a Bh Yo qs a3 90s 8945 =I ~ 8x Fs > co 3 & > > oe >> os © S) &bd SAS GF obs bb 5 Dee PSHE GSS O'S oD oD so oO o fot} >

3so FSooggo 5 VMao, 7)

ws O me MO Mm a Pas

— —Oo —8 fFSO wo St ~~ G CeS—~~ OFVw GGVS aegeeeeei qe _ ae wf) a sot GCmw Se Ss Ors) w i oor Oo o o oo HS Oo o o + 2 Ly, OOM OO OwO p> VU0UnMTP POOP OAD O S

>—e33

oO ~~ Lae oO Or as — °. © Pend — Ww ‘A 3 aos? =BS ao 8s £E aS} ooo 23 a2 |—No 75a)o.5 eo cs £2 om ess < s~0% %p a 3&Sosa 3 CHe se2eo5 #8 me O S Besa wcEGS OC Favs SSSR SC —, a8)

S|] 2;ODS5EEE RSZSES EE Qo «3 ©SQSEE Qo YO OD OoBEER 8 Oo RHBROCONMG oO BPERSS S Om OO O 0ER 6 Oo oa BeIlPlsSmsrev QGeLRZ ZA

ae) oO —_

Z ral oO

w

>

os

3

< vVvr——

NN

On

™ 00

ao ie os . v EM MAAN HN°iv, a75aeog AAAAD Agaaa ce oy.s Ss

LON Oo

km c = =

a NoMsomi tela olnte & ea os c ee ee ee ne) o 8 — SB BBB 5 o =%2.8 8@~ Oo¢O¢€ § 5 Pee ee 8 oOoOD OD rot =i 3 & Tes eeee Ewa KX Loa w= -&.S .& .E v2 Oo AN ON 5So aMeEeecs 00000 >aEooy v7 x Vy © gas pe 656 6S a 6 ° O oa=mb RY BR & n 2 e) eI 8 os rn! oO wo=

moa “™2osgegs OoOr= s3% E WN Se eemM ee ee mmm ee | 2m |© hoe) a}2 v. Ze 2 Sm 25 so 8 Zaaadad a © E se

aes | ce) oo Os ¥ ™m™ OOO 0 90 = i =

— 3 > AnAnANYHA NHYN nAOs&s © A SESCANAAHA om Cp 338385 3[eal 8oD so '8 65-O =

- =awo ABBSBEB sé EC m= 7s) >m We pe 4 So co3¢os=A} S

es = c sf § § oo os oe

nw nA wn~~Nn Aw = 3Wn a 0070 Oo + —_ _=

2s ae eels e8&

To UC -— TO ~~ aTW oOo woc eUc ©¢¢— ¢ Ww oo nN @

Ss8SoS Sass PHa ®oOx x ao 8s— } x oO os oO 3 eeerFeE ESESSE SSEREREE —_— oo se OO ao—_ Roggsdu m~ ww OR—_ SG EO See ee 6BMUBMUEEUSE SE AwRA GOS A DG DO MAUSS SSoDSERRE 5 6D SaaS 6D oD SA

&.E.S ac E.G @ . .& .& a — ~wuuouweees vouetuvus BR FRG BED ooecg oUy 0My oo on Oo oO OD & wo Oo ©} ee -9eF CHK weee HY eS e@¢ S&vy 2 Ee eS soo aeS oOo vwoco oto NX

PePEE Pec tee veo ee

UUUP Dp UVUPO0 0 OmMWomOUmp oO

—~ = oO a “ a5 oo _— = ae S eo 5 a ESases o

o os eA 2 bs eRFOUNM SD = ao = —~M 652 ¢ 7 TD esg8, 2 Litt i § eg BSeg >~A STS Ty, sa2aaug M5 6 6he ~ SBOGCSZ we FHC H HH%xz= ee Poo

o & SE Sc 3 sap AS =e oD SSZPEREE RS HDEEEEE SSGesPprg es

aa ts OS gZG AMMMMMNM te aaeaeteae ses fF OO HM BS OD eS Og FAARDDAA HABEOAOZEAD

82 Misreading the African landscape case an unfamiliar European audience on a first and unfortunately unprepared visit) the damage done by French environmental policy and administrators during the colonial period.

Evidence of earlier vegetation As the Fondambadou case suggests, Kissidougou’s vegetation in the more distant past cannot be assumed to be one of more extensive forest cover with uni-directional loss. Assessing vegetation prior to the late nineteenth century, without the support of rich colonial archives, is necessarily more conjectural. Nevertheless, such documentary evidence as there is supports oral accounts in suggesting a less, not more, woody landscape in earlier centuries. Thus Harrison, who spent some time in Kissi country during an inland visit from the Liberian coast in c. 1780, generalised about the landscape of the interior in the following terms: These countries were in general flat; and the soil sandy, and covered with grass. Where

rising grounds and woodlands occasionally intervened, the soil was better. In such places, especially on the summits of the hills, the people built their towns . . . Timber is everywhere extremely scarce: in some places nothing deserving the name of a tree can be found. (Ludlum 1808: 6, cf. Hair 1962)

The possibly greater extent of grasslands in Kissi country is in keeping with landscapes described earlier in the nineteenth century in neighbouring parts of the forest—savanna transition zone. Thus both Seymour visiting Toma country to the south-east in 1858, and Laing visiting Kuranko country in Sierra Leone to the west in 1822, describe the existence of relatively open savannas in areas which today are either dense humid forest or heavily wooded (Seymour 1860, Laing 1825; cf. Fairhead and Leach 1994). Several sets of earlier oral history data support the picture of less, not more forest cover in the nineteenth century; and indeed of rather more sparsely wooded savannas. First, all the villages above claiming foundation in savanna were eStablished during or before the nineteenth century. Second, several village foundation stories from the south refer to a scarcity of construction wood. In the Toly locality, it was conflicts between brothers for the timber of a swamp tree (Mitragyna stipulosa) which is said to have provoked the

rivalries within a patrilineal descent group which led them to split and establish separate villages (Ki 54). While this story may be apocryphal, making symbolic use of a tree motif linking tenure and construction, it would

have little credibility if timber had not been relatively scarce, and seems incredible given the area’s present forest and thicket vegetation. Third, in the northernmost part of the prefecture, village foundation stories

Forest gain: historical evidence 83 dating from the early immigration of Kuranko families at least as far back as the mid-seventeenth century describe the existence of the particular type of

short grass savanna or steppe known as fua. Independently, elders in the villages of Moria and Baldou in the Moria locality, and in Fourdoukoro in the prefecture’s extreme north-east, told of past expanses of this vegetation in their territories, in which small trees of somo (Uapaca guineensis) were scattered. It could also be referred to as ‘senba gbers’ (‘elephant compactedground’): The limits of somo began from Massamaya and Baldou, continuing up to Kouroussa. One saw nothing but grass and small somo trees. It was there that elephants went to feed.

All along the way elephants passed, breaking branches with their tusks and their trunks ... there were only a few rare forests, but the great savanna there, all was somo and fua ... the vast fua.

Today, fua is not found in Kissidougou prefecture, where even in the north the savannas are dominated by the tall grass Andropogon gayanus, Hyparrhenia sp., and a diversity of trees. Fua now occurs only much further north, in the area around Kouroussa, and even here it 1s associated mainly with the lateritic soil forms often called bowal.

Conclusion

Available historical evidence tells a very different vegetation-change story from the one which has been assumed for Kissidougou, and which is governing policy. Far from being relics of the destruction of an extensive forest cover,

most of Kissidougou’s forest islands prove to have been created by local populations or extended from much smaller forest patches. And far from undergoing progressive diminution, woody cover on the upland slopes and plateaux between the forest islands has generally increased during this century. In many parts of the north and east of the prefecture, grass savannas have become more densely wooded with relatively fire-resistant savanna tree species and greater

numbers of oil palms. In the south and south-east, grass and sparse shrub savanna over large areas have ceded entirely to the forest thicket of bush fallows. At least in recent times, it would seem, scientists and policy-makers have been reading Kissidougou’s forest history backwards. Indeed during the very century of ‘modernity’ when forest cover has been supposed to be in most precipitous decline, it has been increasing unnoticed. Although each of the data sets we have considered here presents interpret-

ative problems, their combination provides, we would argue, a relatively secure picture of vegetation change in the forest-savanna mosaic during this century. A careful reading of oral evidence, within an awareness of vegetation

84 Misreading the African landscape in everyday resource use and political discourse, has proved consistent with photographic and archival data sets, suggesting the utility of oral accounts in understanding this region’s vegetation history. Indeed, it shows a greater potential than has often been appreciated for their more independent use in regions less amenable to photographic and documentary analysis. Two further factors have reinforced our own confidence in the validity of this vegetation history. First, it was unpredicted and indeed surprising to us, and therefore not at all what we looked for. Second, we hypothesised this history first on the basis of only archival and oral evidence (Fairhead and Leach 1992), and found it confirmed in our eventual consultation of the relatively more incontrovertible photographic sources. These results, which are so commonplace for the forest islands’ inhabitants, proved shocking to the development agencies in Kissidougou engaged in environmental rehabilitation. Sceptical at first, these agencies were prompted to check our results by making more systematic inquiries about vegetation change during studies in their own localities of operation. Their findings came to support those documented in this chapter, contrasting strongly with their previous convictions of degradation, and representing a significant turn-around stimulated by our research. For example staff of the Kiss 2 project, north-east Kissidougou, came to recognise that forest islands around the villages they were working with had been created by their inhabitants (Courtieux pers. comm.). The study conducted by DERIK has already been cited (Scheepmans et al. 1993). To cite others: The oral testimony of Deya’s villagers, ground observation, 1952 and 1990 air photo-

graphs appear to indicate that there has not been a significant degradation of the environment in Deya’s territory since at least 40 years. (Niger protection programme, Project Kiss 1: Duwiquet et al. 1993) All the villagers [of Mara] testify that their forest has grown, which can be asserted by the aerial photographs and the information of the French archives. (National forestry service: Maatje 1993)

In suggesting forest advance, the vegetation history outlined here clearly contradicts interpretations within ecological science of the forest-savanna mosaic not only as shaped by ongoing derivation of savanna from forest, but also as a stable pattern determined largely by physical factors. As in other

West African regions where forest advance has been observed, climatic rehumidifaction is a plausible part of the explanation. The earlier oral and documentary evidence which we have examined, albeit speculatively, would be consistent with the idea of drier climatic conditions during an extended period prior to the mid-nineteenth century. But climate alone is insufficient; inhabitants have been engaged in afforesting their landscape. For historical evidence also questions the underlying assumption throughout the debate

Forest gain: historical evidence 85 reviewed in chapter 1 that forest islands are ‘natural’ and can be studied as such. In suggesting an association between the establishment of forest patches and settlement, the vegetation history here opens up new questions about the relationship between vegetational and social histories, and the ways that people’s use and management of vegetation can influence its form.

3 Settling a landscape: forest islands in regional social and political history

Perhaps the most striking landscape reinterpretation opened up by examining historical data concerns Kissidougou’s forest islands. Forest islands have been represented as natural formations, the epitome of ‘nature’, whether as the relics

of a past forest cover, or as more stably associated with particular soil conditions. That they are often found encircling settlements is, from these viewpoints, either due to their selective preservation there amid surrounding deforestation, or because inhabitants have chosen to site their villages where soils already enabled such patches to exist. But while it is true that some forest patches have shrunk or disappeared, and that certain villages were established in pre-existing forest islands, historical evidence suggests that in most cases villagers have formed and extended the patches around their settlements. Considering forest patches in relationship with settlement, as Kissidougou’s villagers do, accounts for their existence, distribution and forms in a way very

different from dominant ecological explanations. It is social processes of settlement foundation and habitation which are central to local experiences and representations of forest island ecology. And people’s particular positions in these processes — their social identities — can lead them to represent particular aspects of forest islands in particular ways. Initially examining these issues at the scale of particular settlements, the chapter goes on to locate forest island

patterns and distribution within aspects of the longer-term dynamics of regional social history, population movements and political culture. Within this landscape-focused history, the chapter comes to consider key aspects of social and political life, for example concerning the region’s kinship

and power associations. In this, it shows links with vegetational issues often ignored in historical and anthropological studies, which thus overlook important ecological dimensions of sociality.

Settlement and forest island formation Living in a forest island offers a number of important advantages. Peri-village forest belts stop the path of the savanna fires which can so rapidly devastate thatched houses and grass fences in the dry season: fires which throw far ahead 86

Forest islands in regional history 87 burning material such as nests and epiphytes which can carry flame over several hundred metres. The forest belt also protects the village from high winds and excessive heat, and provides convenient sources of forest products

to meet everyday needs, whether for food, medicine, fuel, construction or equipment manufacture. Furthermore, the forests shelter tree crops such as kola (Cola nitida), coffee and fruit trees which, for reasons of microclimate or fire, cannot be grown in savanna. Forest islands provide suitable concealed sites for social and ritual activities, and have, in times of warfare, been central to settlement protection and fortification. It is this very breadth of vital roles which is implied when Kuranko say that “haraye ye tu le ro’ (prosperity / luck / success is in the forest); implying, too, that settlement success and survival cannot be assured without a forest island. Villagers encourage forest islands to develop more or less deliberately in the course of everyday life. This is done occasionally by planting trees, but principally by creating fire and soil conditions which favour forest regeneration in savanna. The processes involved can be observed today around recently established settlements.

New settlements are often preferentially sited where gallery forests or swamps give partial protection from fires. A fire- and wind-break which completes the protection is created largely through everyday activities which reduce the quantity of grasses around the village, thus reducing the fuel for such fires. Men, for example, collect thatch and fencing grasses on the village margins: a close and convenient source. Families also frequently tether their cattle in savannas at the forest edge when they must be tied during the farming season, and their grazing and trampling diminish grass quantities and sub-

sequent fire strength. Here, and on the assorted paths out of the village, villagers also slash the grass down as they pass. When these everyday activities are judged insufficient to form an adequate fire-break, young men cut one purposefully. Early in the dry season, elders may also choose to arrange a controlled burn which then eliminates the fuel for more threatening late-season fires. Thus protected, village-edge areas gradually begin to develop dense, semideciduous moist forest vegetation. Over the years, as the belt of forest expands,

grass collection and grazing are gradually moved further out, facilitating further island enlargement in their wake. Meanwhile, the various daily and domestic activities of many people living in a village tend to fertilise and enrich

the village-edge soils, encouraging the vigorous development of vegetation. Household, cooking and crop processing wastes are deposited there, as is the ash of cooking fires. People also defecate there, in the privacy the developing island offers, as do domestic animals. In a similar vein, the gardens which many women maintain behind their kitchens become sites of concentrated soil fertility, achieved through repeated careful burying of wastes, weeds and

88 Misreading the African landscape residues. If a garden site in a village is abandoned, woody vegetation readily develops on these deep, soft soils in which tree seedlings can easily establish, to become part of the forest island. Indeed, in many cases, new settlements have been established on old garden sites where inhabitants have found their forest island formation especially rapid. In local tradition, this is the case for the village of Waldou in eastern Kissidougou, as its name, translating as ‘taro

land’, reflects. Villagers sometimes garden for a limited period around an inhabited settlement specifically to remove grasses and improve the soils, and

encourage forest establishment. This was the case in the northern village of Yomadou, where forest cover was accelerated in this way to surround the village as a fortress (Ku61).

While certain activities are deliberately forest-promoting, many of the individual activities which contribute to forest establishment are nevertheless undertaken without this outcome in mind. Forest island development in this sense depends on the diverse activities of villagers, rather than on deliberate management by village authorities or community institutions. Indeed, for most village men and women, the origins of forest islands are the logical extension into the past of processes experienced in the present. It is unremarkable to them

that the gradual, cumulative effect of diverse activities on savanna village margins should be the establishment and expansion of a belt of woodland. All residents do not necessarily ‘know’ a forest island’s particular history, especially if they have come from other localities, whether as male immigrants Or aS wives moving to their husband’s home at marriage. They may suppose that the forest has ‘always’ existed, while remaining highly aware that every-

day activities cause the gradual extension of forest island area. From this perspective, the rings of large silk-cotton trees (Ceiba pentandra) and Bombax buonopozense which are a distinctive feature of many forest islands tend to be interpreted as the overgrown relics of living fences, around those past garden sites now abandoned and overgrown. People note how effectively fence poles of many species ‘take’ and grow in fertile garden soils, and these tree species are among those used to make garden enclosures today. Forest patches are also associated with once-inhabited village and hamlet

sites, referred to in Kuranko as tambon and in Kissi as ce pomdo. Most villagers consider such forest patches to exist today because of the legacy of past inhabitants’ everyday lives. With the abandonment of gardens, forest has grown over them — if not over the actual house ruins themselves. All their fence

poles have been able to grow up as forest trees. The land around these sites has been super-fertilised. Since it is normally everyday life activities which promote the growth of a forest island, and as such activities can be important to maintaining it, village abandonment can also precipitate forest island loss. Villages may be abandoned for various reasons, whether settlement consolidation, shortage of space or water, fear of falling trees, or social and medical

Forest islands in regional history 89 problems. If inhabitants remain in the locality, an uninhabited forest island may be maintained for its economic trees and ancestral significance. But management and protection becomes difficult at a distance from the new settlement site. As a new settlement develops its own forest, therefore, interest in the trees and tree crops of the old forest sometimes declines. Especially in cases where old coffee plantations housed in forest islands are not maintained, flammable materials build up. Fire encroachment because of this, or more generally in dry years and droughty places, may little by little diminish the island’s size. The

forest may eventually be converted wholesale for agricultural use to take advantage of the highly fertile soils beneath it.

Certain elderly men and women describe the formation of forest patches as more intentional. It is most commonly those of patrilineal descent groups who claim settlement-founding roles for their ancestors, and whose political authority derives from this, who emphasise how their ancestors arrived in empty, relatively inhospitable savanna and initiated the beginnings of a forest island and a settlement there by planting ‘starter’ trees. As one Kissi elder said of his village, Yiffo: “The firstcomers planted cotton trees. There is still one which carries the name of the planter’ (Ki39). Such foundational trees, within the forest island or the village itself, can be attributed origins which emphasise the founder’s extraordinary capabilities.! But more normally, founding trees are remembered more prosaically as individuals of fast-growing species which were transplanted as wildlings to suppress fire and accelerate rapid forest succession within their protection. From this foundational perspective, the rings of Ceiba pentandra and Bombax buonopozense trees in forest islands are represented as rings of starter trees.

Frequently, one of these founding trees recalls the establishment of a relationship — almost a ‘contract’ — with the area’s land spirits: a relationship

maintained ritually by a founder’s descendants to ensure a place both for human settlement and reproduction, and for productive farming, hunting and fishing in the territory to sustain it. Relationships between descent groups are implicated in establishing these productivity relations, and the representation

of forest island origins in terms of initial tree-planting is significant in upholding these. For ruling families, the trees remain markers invoking historical planting events which legitimise their current social and political status. This significance can persist even in abandoned forest islands, the ‘initial’ trees of which invoke the political authority of ancestral settlementfounders. Those claiming descent from them commonly emphasise these reasons, rather than more economic ones, for preserving islands no longer inhabited.

That settled social life promotes forest establishment seems to be a basic local precept, albeit one with socially differentiated dimensions. It is the needs

and activities associated with habitation which create reasons and enabling

| 3. 2 ,

90 Misreading the African landscape

eS Wee Se? ile

2 ee Oe

oS ea we ao 4 dea 2 pees SS __-_— ees =... YF ee ee cl ce ° NS tC i hhh—hr—. Ls SES TR, og Sie SEER OER yaaaees tae Tf, eee cet SEOUL inte teas ENLES Sa ee PeeeOE :teony SPE Yate eeSe Eger, BeOR a ae =Ree RRR Beit Gees eee ESRSRR geek: OP a Ne aed gOS ge USES Ue oe Pa 2S Ber Bo pe me RESP Se ee RR Re aes a SRO gg ae Aas EE Oe Sage ac oa BkBoSk Se Rca ee eee oeTas ee wot woGeo ee >yey een.gleee RSeae heaeee es Se a aR SERIES ER ceeUE ESR. 2 BORE RE aigBeh 2 aeRRA SEES os RR OEMe Seah Sa Se FE RE ge Ee oS ne a Re eResa it LES TAMA ee cnet eae IR SY

Pei Ao. Sr7 lS SSE io BESS Y -smeraeetanein ree «agingvouSNE eeoaeeae SSR E | cy 8 eee ER ete Ss REEa ee “US A Pn US GRR pre te os Ess srs sa Se tc ati:ee MRC Seas ages Satesoie SRRRE a oa,hae See ERR Spe CARES CSEEEE des |

Ee ES Ss BE Becta ane SefeLe beecg a aoe gos Shap #:eRRoSee eees, . oe TheooSe SEAR cn SS oe,eeoe GPS S-SEE SEER SS Saag BRSRR ee fae ESReg SS Bae “Rees SRR, TR Seay See SeSRS Sage ie as SME ONES pete,AREER BePears BaarsoS a ee Sh EE Se Sea ee seek We ee Beers 2eoPO ee Eeaes - ial eeeED eg aSEER pie een ca ee Soncbigieel neaTES Pate SEES SRR tS eo Etot Se Sane ARe n> aaSe es oer oOaaae TOD URES URE Lis CERES ee ee aa iePile a teFs oeSEER PER aed SOR EEESE ee oS ISotdens geSeage Sahay te Sa RigSeo aroaee. LEER TE etreSE aR RR oecseni RSS Sesi SERRE URE oI Bh ge es Seca Bie ote a :aGES ceeees eeree Eraot, a aa JSR. SORES, EERE oewes SSR RE EP ceReBe PE, PRP Poaeca s Seince ee SE ee Cet Re eae (Ae Eee oye eR SE Soeapes eteBe LEP ane SOS SR ok Sok OTT. ‘ age Wo RRS SOfot Gg ce PR Mn Se ER Oeaii SOT. tet ee RCRRR oe ERE ee CORRE ceeSaas ey SMe ST Cette Sen A See an StkCERRSRESS SRE EEE Sather sega ae Ty re See SRRee eRee eS aae a HERES “SabiRe Sees eg TOR wach, USESSEES ote i pepe oahagye ne ge SRERRE Ree aC SEN SRE SRSEISE Reofeet heaan EEscaee : SES i: Aah So AaeeneSch. ceewie Cea EES ES

re Hogs, § Prairie 3 A Pe a errno ees

pe er SRR 2 ee eRe EE a See Re Dies USES a a a S sa "as SEL BREESE gee SR EE

S.A ees Sees BF Se ee ROS Se ee ghee MSE. Sie Cc SRE ED PE See cue oe RBS ee See CAS See —— De aaa ge EE SERRE ye tue Se -aeBes ae oeRe. SSieeePEE ke ae. ee elRee CREB OOnet Tkbsee osBag MES = BP RSET SEGRE gee EEnee reRR Se eeneee Reae gseerc.FEES Soe ORE RE RN ER co, CREASE re a bet 2 lg EES SREURELS SS SIRES IESEAR RAR oe Soares “S eeSO RreR SEE. a ws 2bo URME BEER Re (SUR acoe: PPO RISE ES RELy

= Secor oe ea Stod SPE CURSE ag ESO oy SRR SeBea SI orate Re rt 7 Loads SS ISSR GERRY SRA Se cee aeee AS:OS 2 etsae SaBe TASTIER in OEE ct atc perccay 7Se Se REE ByeEihacne” BS nN lad a sg i | Vol,Nee Rea PET 058 Sc aeRE ee hes” 2 aglee: wn ER aeBye eer.SRR aesRR Sg,oS“SOSBE ty ME eg eRe Si re Ben

ee bp ae, ow ® , ee ee ee : oor re Bebe RE bg

s&OEE OOFBe SOBER Se,Bc Re.RS oes eetae “ -gp _ SRSA RRR rg SET sR. eeeS REBS RRRte ah oe eaechee a TAME SE hase: Bc Te oe as co.fosERR cones aR: Boni LeaseCNSR CERO LSE SeeOe GIES NR Senne OES

eT ERE SEE Soa. oR agi SPR ROR RR aoo. SPS : SSE REE USE, OnE) SUISSE rg PE . ge OU : “ER ee ee "San onTOT “SSe: SESLSR TEaeERR SER gens SeeTEUSE

cei .- ::yg se: e, ole! PRES i BES CORRES sie fs OSES sheUSS mSRRO GSR ES igo ET = Be ys SOURED slPera Ga Bon :Hee , :RR a EE RRR Bs OSE ONSESSREIERES SM ESUE age BO ks : BM SRRETESR Euee, ac .PERS a Rc, ee. Daegett, oeneg gee eS Cneewip inteSe ietle OgRE esr Tes a OS .a. os Tone Ea me CU EERARE PooS REL TP CEE 5SES ES EeTeRat LO geetREIS TeOe IN 2ceea are ° .os es ee ed $e SSSEREESS> CORES eeSI Seggheb israettect I aii Ba CERES gE eee . ee Boas “AF Bay deeETE : SESSChagas “SeE hogsSOS BER Ya ORE Cae SS RE ONES OS ie iia Pet SURI BSE Seoo neee a ' . wet ERE SUSE . SL UU see Ras coe ee aes7een. aoe RON CORRES RES Sie yc REETR seats SEES : cares Bd mGa 7ARI ee REE Lee msMRS ae edt SOORM aaaREaes Or oe ASR JRE : Shag - SESE ee aOEE csSeer Sk”tace ieee ne iene Tage SeeSEE 2 ERE ae LR Sg : DRE EeOR eectER ec ee

.=oe: fi * co Ope MP 2 OE wee eee 4eee gicigeetrehccccay eh Y_EE SSR Sel ae eeegRE EETRE Pa lide ode SERRE TTLEST Sse eefee ASE Sn Eh ees ch ey teh Ss Lag CAVE oe RES EESEE ee emer J SESE Panis Reet gh Sth wo Copies qa *:Pod ee e— se re, : sere *- ee See xSas Be wfae :mee ee st 7c oe Sa ea. Sit ES “x ae er me es a oe _ee : a8 ;eet. -Pag é i : ! hee. ee E, eck tong Se =o eos ig sn ot Be Se oa Pees poet ~ Be oa — ye mS ce wn ee. - ~~ ae

ae wesaog lous Aeaia ans eS ee z se ae nes 3bg er sis heee ae esfee i *oe oeoe — a.a

=: : : ee Fi> Site arty wta eeod fos: ae : : 3 ;:: f= { "se : f a:DP SS ‘ a agc pee v1,¥ 2% S Be ee on Sag Pe eee ee bs — see fe ie see. e a sepee Sean aa Seos ae a om — ee ‘= Seoe eet 7a. See aoS-aeco .Sey“hee neoeaoe ee « eee oe Stee " .aiewe s, Z Soe —ee% :espe oad 7 ee-— ee . aae —Bee so7 ee aeBee ee ee ih. oe oe iaeeE 3 - Be Geysoe Se wes ea esscyoee™ ae Be ee ee aoe of

ag a FX alee a sie ue. e a & « ase =,== oo Bee caiatAe RE oes ee % ae Se ‘ a 2S ooOy a ssae “ Le? SSi ee oStephagc aed we “ar— nD gee — a ne a, 2 , | | . . x 3 .

ee eo o 208 eo es a pa oe eK * Loy, ag en: Oe ee a ..FoEee egioeoe 8 -ge ge orc ‘pe ae 2See go% Ae : BS Se. — eS :ee a: -eis aeae aSao Poe ee oe‘aeS ee 3 hey Je La:Pee sat ~~ Oh om Ae est B a .Pe co ae sya ; ° Oe ¥poe Se,osSe. eeeeeeae tages sees: = Be a oo S ee a Sa fae Sores hak ae se Sea So ee Say Sanit pe Fe BS See Sens zs “3 aS 2 “8 .; re ux e Ld e 3 °. e 2

* Ld * . ° e e b) . 3

is common in Kissi areas, it is the lineage of the political chiefs, here the Tolla, which dominates leadership positions within the women’s society. These roles are often described as if set in stone — and they are represented almost literally as such in the clay sculptures prepared by the men’s society for their comingout (plates 5.2 and 5.3) — perhaps concealing historical shifts of power and allegiance. This enshrining of relations of authority over village territory and people is

a « *. s

2

«°*Li * »e °tdLJ. 6e .

repeated in land division into the large blocks (tandaran) representing each

*. * La s e in a Kissi village 153 Ecology and society

Se Mg eT BM TN ge .

ee ESUTE lrrrt—“(‘CO;™*™SCOC*SCOCisi‘ESSSsC hl Po eR ets 8 i hLhmrmrt—~—

E

So SS a5

8 88 Nom Sak 2 o EE) 3s oO C4 % 5 > ss B 2 Ehe 34a Oo €

~ § = q 0 + SX

aOR S o32> bs COR FAM]

Enriching a landscape 193 communities, and as targets for elimination. In contrast, women find the weeds associated with swamp communities easier to control. They tend to compete less strongly with rice and be easier to extract. In humid, inundated conditions,

sule actually “self-weeds’ (it falls over and rots) and while Cyperus haspan needs to be weeded, this is not hard, and farmers note that its decay rapidly

supplies the growing crop with nutrients. Inundated swamp vegetation communities also tend to be relatively inhospitable to some key crop pests, especially frogs and rats. Farmers work with assorted ecological processes to achieve the conversion from plain to swamp vegetation. As we saw in the Sandaya case they direct water flow to the margins using canals, and reduce water flow with barrages and hedges to achieve land shaping and levelling. Along with assiduous annual cultivation over periods of many years, this can progressively level the land on more gently sloped swamp edges, and suppress the plain weed community. Grazing cattle in and around swamps during the dry season not only fertilises them, but also helps to spread seed from swamp centres to margins. Operating incrementally, these year-to-year processes progressively deflect margin land to swamp vegetation, expanding the swamp. Farmers sometimes accelerate these gradual vegetation changes by directly

enriching swamp vegetation with species thought to enhance and preserve humidity. Thalia geniculata and Mitragyna stipulosa are sometimes introduced, for example, for their capacity to raise surface water levels.4 Kuranko farmers say of Mitragyna that a di dyi yele sanma; dyi di bon, ‘it will bring water upwards; water will emerge’. And as a Kissi man in Fondambadou described: If a swamp has a watercourse in the middle, you deviate it to the edge with a channel. You plant bananas at the edge of this channel to maintain the water ... Bananas attract

water. Then again, one can plant poo (Mitragyna stipulosa). With that, the place becomes a swamp. Each year you move the channel a bit and the swamp just increases in size .. . If you transplant p99 it normally succeeds well. Then again you can sow seeds; they can grow. I have a swamp here, it was I who planted the poo and they did well. When it became very humid, I began to kill a few by putting grass at the foot of the tree and burning, so that the roots would not prevent cultivation. Even if some trees die the water rests. (Ki 40)

Farmers also recognise the dangers of swamps reverting to plains. This can be provoked by drought, the degeneration of water control structures or indeed

by cultivation practices which permit dry community plants to recolonise swampland. Thus cultivating peanuts on mounds in marginal swamps can ‘bring /mperata’. Reversion can also result from eliminating all Mitragyna trees, unless canalisation is managed extremely well. This problem has sometimes been provoked by agricultural development agencies who advise full swamp de-stumping to facilitate cultivation with ox-ploughs. Much of the

194 Misreading the African landscape art of successful swamp management and incremental swamp improvement lies in balancing these multiple influences on vegetation. Upland savanna—forest transitions

Certain patterns of farming and vegetation use can also deflect fallow vegetation cycles in an enduring way on the slopes and plateaux. To date, it has been taken as axiomatic for the Kissidougou region that such enduring vegetation conversions as occur through farming are in the direction forest to savanna. Indeed, chapter 1 showed how dominant scientific views throughout West Africa’s forest-savanna transition zone hold farming to have assisted the conversion of forest and forest fallow vegetation to savannas. To recap briefly, this is seen to involve the invasion of savanna grasses into the fallow, enabling repeated dry season fires to penetrate, which prevents the regeneration of

secondary forest thicket. Subsequent soil degradation, it is said, only intensifies this effect. It is argued that in subsequent fallow cycles savanna successions dominate, where forest successions had been before. Axiomatic to local ecological thought and experience, however, is the reverse: that farming and other forms of vegetation use assist transitions from savanna to different forms of woody vegetation, whether to savanna woodland

or to forest. Such deflection of succession brings numerous agricultural advantages, recognised by farmers: notably an increase in the amount of vegetation and hence fertility made available to crops when the fallow is cleared; improved soil qualities and water relations, and the suppression of noxious grass weeds. The possibility of deflecting successions pivots on the effects of local land use on fire. From villagers’ perspective, the control of fire need not depend on avoiding setting it, but can rather be achieved through land management practices which limit its spread and intensity. Such practices bring together diverse ecological processes involving soils, water, animals and plant propagation and competition, which combine in their effect on vegetation and fire. There are many ways in which this savanna—forest deflection can come about, of varying significance on ecologically different sites and in the prefecture’s different localities.

1. Edaphic improvement through extended cultivation. A first example of the deflection of savanna to forest vegetation succession 1s provided by the tombandu sites discussed in Sandaya, where gardening-like,

extended cropping sequences using mounding and organic matter incorporation have enduring effects on subsequent soil and vegetation. When tombondu soils are left fallow, both the soil structure and the nature of the grass

cover heavily influence the subsequent vegetational succession, deflecting it towards forest rather than savanna. In explaining why this occurs, farmers

Enriching a landscape 195 recognise a number of individual effects of ‘gardening’, but it is in their combination — and their combined effect on fire — that they describe the relationship between their farming and vegetation. In many cases, local explanations are supported by ecological and agronomic studies, although these studies have not necessarily been previously applied to the question of forest-savanna dynamics. Gardening, farmers say, ‘opens’ the soils and enables water to enter them. Despite conventional arguments that working the soil considerably reduces infiltration, encourages water runoff and is to be avoided (e.g. Lal 1976), recent research now indicates the wisdom of farmers’ view that cultivation improves infiltration. This is for several reasons. First, the surfaces of uncultivated soils

tend to form crusts which prevent water entering rapidly, especially on the more clayey textured soils which farmers tend to prefer. Even superficial cultivation — or indeed uprooting weeds — immediately enables better infiltration and reduced runoff, relative to uncultivated land (e.g. Casta et al. 1989). Other works show how deeper cultivation improves infiltration relative to superficial or non-cultivation, and noticeably improves the water supply to

crops and, hence, the vegetation cover (e.g. Kalms 1977, Chopart and Kone 1985, Chopart 1987, Hoogmoed et al. 1991). Mounded soils have especially high infiltration benefits over unmounded soils, especially during periods of lower rainfall (Lamarcheére 1991). The extent to which infiltration benefits are enduring is uncertain, but no research, to our knowledge, indicates that infiltration is damaged by mounding. Experimentation on the relationship between soils, land use and cultivation is, however, unlikely to confirm or falsify farmers’ wisdom in an unequivocal way. As Hoogmoed et al. argue, the number of very significant variables is high, and ‘one piece of information missing may render an entire experiment virtually useless’ (1991: 96). This severely limits the validity of generalisations which can be made from such experiments. While these results support elements of local reasoning, there has never, to our knowledge, been direct investigation of the hydrological effects of the repeated mounding so important to many farmers in Kissidougou and the region.

Villagers also suggest that heavily gardened tambondu soils retain more water, and for longer into the dry season, which assists vegetation development, retards desiccation and limits fire damage. While soil scientists would suggest that water relations have much to do with underlying geomorphology and soil formation processes strongly influenced, among other things, by

position on the slope, certain scientists would also consider aspects of water—soil—plant relationships to be alterable within the limits set by such structural conditions. First, water movement down and up the soil profile,> and root development

in it, can be influenced by sharp changes in soil properties in the soil

196 Misreading the African landscape profile. Uncultivated soils are characterised by a humic topsoil very sharply

distinguished from the earth beneath it.6 Such sudden changes in soil properties, including similar ‘horizons’ further down, can impede tree root development and hence vegetation potential (especially if the lower soil has restricted porosity). This effect has been described in the forest—-savanna mosaic in Céte d’Ivoire (Avenard et al. 1974) and Nigeria (Moss and Morgan 1977, Lal 1987). Deeper mounding can contribute to horizon mixing, both directly and by stimulating termite activity. The latter is not lost on farmers who appreciate the role of termite channels in bringing water and waterretaining clay to the soil surface, enabling roots to go deeper, and homogenising the soil. Second, villagers emphasise the ‘ripe’, ‘oily’, workable quality of matured tombondu soils, and how tree seedlings establish more easily in them. This is also in keeping with ecological research on the conditions for tree establish-

ment. Small changes in the soil surface can have a major impact on tree germination and seedling survival. Moss and Morgan (1977) suggest that soil moisture variation in the uppermost layers — possibly even the top inch (2.5 cm)

— may well be critical, and can indeed have a determinant influence on vegetational form in the forest-savanna mosaic.’ Farmers also emphasise how extended cultivation eliminates the tall, dense grass associated with open savanna and new soils (Andropogon gayanus), and

how it does not immediately reinvade. Instead, the post-cultivation site is associated with shorter grasses such as Pennisetum violaceum and Rottboellia exaltata, the latter associated with tombondu in its name (tombe), and these offer less fuel for dry season fire.8 These are also the preferred palatable upland

grazing species for cattle and many wild animals. Cattle left to roam and forage for themselves in the dry season often graze such sites, especially as their more humid soil conditions mean the grasses here remain green and fresh

for longer, and their grazing and trampling therefore reduce fire fuel still further. Scientists’ and policy-makers’ analysis of ‘fire control’ fails to appreciate these ways that villagers use vegetation itself — and alterations to soil which influence moisture and vegetation — to control fire. It is likely to be a combination of these interacting soil, water, vegetation, animal and fire factors that means that bushes and trees can rapidly establish on ‘gardened’ sites, despite their reduced fertility. Many of the initially invading

trees are leguminous savanna species, such as Albizia zygia (tongbenin or tombon-gbenin). With time, as villagers describe, these themselves cede to trees characteristic of forest formations. Such succession varies according to locality in ways which depend on the particular seed sources and ‘paths of fire’ in the area. To gain a better indication of the outcomes of these processes, we compared the vegetation and soil conditions on ‘improved land’ (tambondu) and ‘new

Enriching a landscape 197 Table 6.3. Site characteristics in survey of tombaondu and ‘new land’ Tombondu yrs. since last

Site Location Tombondu history cultivation §Dukura history Yelfagberedu Across slope Old farm camp site, c. 8 years Farmed with 1—2-year

between two abandoned c. 1940s, since cassava/fonio sequence

swamps cultivated at intervals with more than 15 years ago;

45-year peanut/cassava otherwise no known

sequences cultivation

Soradou 1 Flat plateau Old village site, abandoned 12 years Never farmed c. 1880, since cultivated at intervals with 4—5 year

peanut/cassava sequences

Soradou 2 Flat plateau Old village site, abandoned > 20 years Never farmed c. 1880, since cultivated at intervals with 4-5 year peanut/cassava sequences

land’ (dukura) in Sandaya. Villagers identified for us three sites where there was a known boundary between land subjected to extended cropping in the past and land which had ‘never been cultivated’ (table 6.3). In each case, topographic and other conditions either side of the boundary were similar. The survey methods were designed to give only indicative, not statistically significant results. At each site, a transect line was laid across the boundary and

5 X 5 m quadrats marked off at 10 m intervals along it to a distance giving 8 quadrats in each of the dukura and tombondu areas (total sample area of 200 m2 of each land type at each site). In each quadrat, tree and bush species present were counted within four height categories (< 1 m, 1-3 m, 3-7 m, > 7m), dominant grasses were recorded, and grass cover estimated to ac. 5% margin.

For soil comparison, we used the simple terminology and standard terms

outlined by Smyth and Montgomery (1962) to describe the soil profile, examining it using a soil auger at c. 20 m intervals along the transect. The aim was internally comparative: to reveal major differences between profiles at

different points on the transect. The characteristics of each horizon were described in turn, dealing with the various physical characteristics in order (viz.: depth, colour, consistency and structure, organic matter content, texture, larger particle content, root presence, nature of horizon boundaries). The matured tomboandu soils were found to have a deeper, darker, and in places more clayey humic A horizon. This is usually twice as deep, penetrating to

198 Misreading the African landscape 30 cm (slightly deeper than the hoe reaches when mounding) and less defined in relation to the B horizon. Results of the vegetation analysis are shown in figure 6.5 and table 6.4. At each site, the difference in grass cover between land types is striking, with new land showing a high (often 100%) cover of Andropogon gayanus in the field layer, and tombondu a much lower percentage herbaceous cover dominated by grasses such as Pennisetum violaceum and forbs such as Aspilia latifolia, offering less fuel for fire. Tree numbers are much greater on ripened than on new land: the latter carries scattered mature trees while ripened land not only has more of these but also has a vigorous woody undergrowth layer containing young trees regenerating since the last cultivation period, as well as numerous small ligneous plants and creepers not individually counted in the survey. Table 6.4 compares the tree species which occur on each land type, and indicates how they were classified by the colonial botanist Adam during his Studies of Kissidougou’s vegetation. There is certainly a large overlap, with many trees found on both new and ripened land. These overlapping species are largely, according to Adam’s (1968) classification, those typical of woody savanna with fire-resistant characteristics.? Many species occur on tambondu which are not found on new land, including species identified by villagers as

particularly characteristic of tombondu: Ficus spp., Albizia zygia, Entada africana and Newbouldia laevis. These species are almost all characterised by Adam as typical either of ‘secondary forest’, and/or of the intermediate vegetation form which he identified as ‘savannised forest’. At our Yelfagberedou transect, these species also occurred where the new land transect passed over a termite mound whose particular soil characteristics had allowed a small island

of woody vegetation to develop in the savanna. Newbouldia laevis and Allophylus africanus are species commonly associated with such environments.

Thus the vegetation structure of new land and its species composition represents a sparsely wooded savanna community. Tombondu, in contrast, contains a mixture of savanna and forest species, and of species typical of such mixed formations. Although earlier scientists such as Adam linked such species to ‘savannised forest’, recent ecological work now recognises that

such transition woodland communities generally represent the advance of forest into savanna (Adejuwon and Adesina 1992, Moss 1982). Appreciation of the vegetation-upgrading effects of settlement and intensive gardening-type cultivation is not limited to Sandaya. It is equally recognised in the more northerly Kuranko savanna areas around Moria. In the Kourouman

Kissi area south-west of Sandaya, women seem to have incorporated the Kuranko noun tombondu into the local dialect of Kissi as a verb, tombolo, to

describe the process of ‘ripening’ their garden soils. We did not hear this expression used in other Kissi areas, but the soils of old villages and farm camps are distinguished there as pomdo. In areas of the east where uplands are

5 marae gt ae ee ie °3 seca > ae pao a a asee | iE

Enriching a landscape 199 sratatal

2= +ee BESoe cense Resets Seesae eects fone ehh ma eee oes 52.5 2) oa ees ee mecreate

= 1.5 +peBe aspeace _ aefees So mel esReed oq Ae sone me S|, soo cory ee o 17| eee Bes Sean meee) Ee peeee Boas2)BSBS

= 0.5 | Ree aS oo) Ra od Bee Sl ee oq Bees Se 0 EES Ea SSeS] BR Sy Bees Secs | Rees =o] Bes Bo dukura tombondu dukura tombondu dukura tombondu

Yelfagberedou Soradou 1 Soradou 2 Tree height

7 metres (a) Mean number of trees by height class

100 90 SRO

5 OT Ge henna io) tetetetetatatetetetetatan anatatetaverererereteren SERRE cretcnrnsreenatenaese 50 reletelelecereseneneseses nrereatanreatetetecatees Dabney pease tentettatataneSes Rebeietacaieietececess aretelecetmeeeaereneets 3A~~otetetetetessserecenereet 40 7 Eeeescsee:sieleeresteterstesese wstiteeearertanrtets Soest tetenteretrerenee 30 eects ————--—— reielepeceanierneerens Datrtecenateteteterreese misteeceseesereenetats

oD sisiesesseeeeseetonerees ——— scoceleteretesseeeees partetetetetoresneneey weleresetereonnerereree

o 207 ESEEESEEE: seeseicreisetsehces ssrestentereet BORO EHOD AOS ssatieeenteteetees S40 soiepepenecerssepeneneets cseetciirntet sreleteneieesenestonses tatetedatetenecetenesese

dukura tombondu dukura tombondu dukura tombondu

Yelfagberedou Soradou 1 Soradou 2 Grass species

Andropogon gayanus Pennisetum violaceum Aspilia latifolia [| Other (b) Grass cover and dominant herbaceous species

Figure 6.5: Vegetation cover on tombandu and ‘new land’, Sandaya

partly savanna, these are also associated with a dense transition woodland type of vegetation (Ki50). As was described in the village of Waldou: Pomdo is not the same as the other soils. The things which grow well here don’t grow well on the savanna slopes. Even taro is not planted on normal soils — only on pomdo and soils of foyo [transition woodland]; if there is foyo the soil is not hard. When we cultivated taro here (on the new village site) the soil was well worked. Like behind the huts, ppmdo is a very cultivated place where one threw ash and ‘did anything’ [defecation].

Li ~ 2 3 " | 6 _ “|8 ) >, 7 _ s bot “ ; £8 " , | t:a «> :reptOM pe pepe °," 3; A tt) Som |

bs ~

bas

_ 1 mo 7 ps sm Mm OM rw.

:m >< iw re 7 : mM oe mM Oe 23 ~ rm OP 1 : OPS bas_ | | “nA 2 A : me * A2A

Ps

»

—

mm OR

:3;:4iA re ye am Os re ;fo rm PS OP _ ha ha AN | oe m , E rea OG :V pe a & ° | i : rs * ra gael s mr Az , : mS " | : ~ OM ye pe

> ba a |

S$ 5 e

. rma PG

E “ re Mm OR ms ras

wv)

% x | >::

a) a 2 Ps mm OF : Pe:

> ps

::re) . pe pe 5=| ; 23 .V wm OM iH 22] a ~~ OM 5 | :oO8 opa. 58‘A8; Zo e”~ Za «a 3 8;zB :BL2 EE : sEREEREB _273DG NB) wh 2: 5a fae}

rat

nr

pe

isi

:=~ man ew fe , ae: I oy fy 2 S&S E fess x2sbeeeee sk 8 6 : ;ce) Sy SS 3 ::3Sse £38 SS 2S eS & wy ks 33 Sas §3 : ; : : S bo QO ~ Ss feec bled as ere : S 3 Q, Es=. 20MEH THU %

NAN ee~& ecb2% bee 2 o

2 F228 S585 g&§ é o & SSE ES 28 ye hY sg 3 Es Beiso 2SRESS eo SYS Ss) =ebpeisigisiagaes 3 esSs888% 9SSes 8£S28 2s8iS 8 $ses8 SELss sses 22g S555 Rsei RS & xy yx ESA 5 se&siaes a, &

> > > ~ x > ro re « x

>>>* >»

ro > x > >» > > >

>

ro x be

i.

= OF}

rae]

=

ie

om 00

Oo

= nee”

fon

E = : E a2 In the early 1950s the same reasoning and policy response would again recur, and the better financed administration then did promote oil palms

as the basis for reforestation.33 In the 1950s too, the observable increases in peanut production were then also taken as evidence for palm loss: thus it was argued that farmers stepped up peanut production to meet the need for a Substitute oil as palms disappeared and as soils degraded.*+ Administrators seem to have been blind to the increases in palm density and distribution which were already taking place in many regions due to positive management by local populations, and the very different reasons for increases in peanut production which earlier chapters have outlined.

The war economy: 1939-50 During wartime the imperative for food production took precedence over any environmental policy objectives. The forestry service was cut back. Fire policy had virtually no impact at this time, it was claimed, because of the indifference of the public authorities who had to put it into effect (Rouanet 1951a). Cercle administrators passed on to canton chiefs responsibility for providing wartime rice and palm oil quotas at artificially low prices, and today’s elders remember this as a period of heavy taxation in kind of compound heads, and of poverty and hunger. Local responses included significant land use changes, as earlier chapters have shown. Coffee production plummeted from 956 tonnes to 40 tonnes between 1940 and 1941; a result of local disinterest in coffee production given its low price, and the problematic labour trade-off with rice and secondary crop production.*> To reinstate coffee production and to meet war effort food quotas, the Cercle

and national administrations made a major attempt to upgrade once again the agricultural extension services which had operated so weakly during the previous decade. In 1942 the Cercle administrator made funds available to a French representative of the agriculture service. He imposed coffee quotas and, critical of local rice and coffee management and yields, called for an agricultural organisation capable of educating 60,000 indigenous farmers. To this end, he established a ‘farm school’ in 1943 to train extension workers and run model coffee plantations, and a network of canton farms to relaunch the ox-plough through demonstration and instruction.*° But again, higher invest-

ment in Kissidougou’s agricultural service proved shortlived. Its French representative left in 1945, leaving only one deputy monitor in Kissidougou and the staff of the by now debt-ridden SIP. The farm school folded, and all

agricultural activity except SJP-funded coffee nurseries, seed loans and

248 Misreading the African landscape ox-plough sales concentrated on the regional agricultural administrative centre at Macenta.3’

Post-war prosperity: the 1950s After the immediate post-war period, concerted attempts at agriculture and forestry policy implementation were again made, with this time a sustained growth in personnel and financial resources. In 1950 a French representative of

the agricultural service returned to Kissidougou, and by 1956 was working

with five trained Guinean monitors and a European export crop quality controller. The forest service also expanded: by 1953, there were eight forest guards and one auxiliary operating in Kissidougou,*® and in 1956 a forestry ‘Inspectorate’ was opened there. While environmental concerns had been important in the 1914—50 period,

they became central to both the agricultural and forestry policies from the 1950s, integrating them in a mutually reinforcing way. The prevailing environmental problematic was thus inscribed into agricultural as well as forestry institutions. Indeed Kissidougou’s 1956 Cercle master plan for rural development stated up-front that ‘all must be subordinated to the conservation of soils and forests’ .°?

In agricultural policy, efforts were focused on shifting rice production to swamps and river plains.4° The top priority for rural development was said to be ‘action in favour of wetland rice cultivation. The future will be preserved

by the progressive reduction of upland rice.’*! Extension activities focused both on large-scale water control structures to irrigate large swamps and river plains for mechanised and plough cultivation,*” and on encouraging all farmers to cultivate their own smaller swamps intensively. The administration in the

early 1950s felt vindicated in observing that ‘over the last 20 years, swamp improvement has been proceeding in Kissidougou’,*? a process which, as earlier chapters have suggested, greatly accelerated in the north of the Cercle during the 1950s. They assumed that in making this change, farmers were responding to declining soil fertility on ill-maintained and eroded uplands.*4 But while the trend was in keeping with agricultural and environmental policy

since the 1930s, and benefited from the availability of new wetland rice varieties introduced over this period, the shift and its timing have been more attributable to a combination of migration-related social and economic changes at this time, as indicated in earlier chapters. Equally, social and economic circumstances at this time deterred southern Kissi farmers from making such a shift despite policy-makers’ efforts. Soil degradation began to dominate environmental agendas in the 1950s.

While soil issues had long been of concern in Guinée, a new heightened awareness reflected Africa-wide publicity for soil degradation around the first

Reading forest history backwards 249 inter-African soil conference in Goma in 1948, and the publication of Aubréville’s ‘Climates, Soils and Desertification’ in 1949. Soil conservation brought new funding envelopes. Existing environmental policies were now Strongly justified in terms of soil conservation,*> as the impacts of upland farming and fire were imaged in terms of the loss of vegetation for the soil which it protected (Aubréville 1947, Rouanet 1951b). In 1952 the director of the forest region’s Circonscription Agricole described the relationship between deforestation, fire, soil erosion and laterisation,4® as he saw it, as the most pressing agricultural development problem. As he put it: ‘Forest Guinée is characterised by an originally rich soil on the way to sterilisation following the irrational agricultural procedures of the indigenous inhabitants.’4’ Building on the recommendations of the Goma soil conference, the idea to assist the regularisation of the major West African rivers flowing north from Guinea, first proposed in 1934, was now linked to soil conservation priorities, and funded in 1950 (Diallo 1989). The grandiose scheme envisaged placing 15% of the entire upper Niger watershed including northern Kissidougou into large reserves, and establishing 30% reserve cover in the more densely wooded southern areas, including southern Kissidougou. The programme was initially launched in two pilot sectors, not touching Kissidougou (Bonnet and Vidal 1959, Vidal 1954).*° The work was to be done by force until the inhabitants realised the benefits to be gained. The effects of bush fire had been strongly debated, with polarised opinions about whether or not fire was useful and under what circumstances.*? In 1945 a more clearly defined fire policy was imposed uniformly throughout the country, which banned running fire at any time except during December, imposed precise controls on agricultural land clearance, and demanded running fire to be immediately extinguished, once observed (Rouanet 1951a). Some unease was expressed about the universal applicability of this policy at the time, devised, it was admitted, mainly with reference to ecological conditions pertaining in the Fouta Dyallon. Early burning-based strategies might, it was felt, be more suitable for Kissidougou and Haute Guinée. Indeed early burning had, since 1935, formed part of national fire policy, and in 1942 instructions had been given to set early fires throughout the territory. A more regionally sensitive policy was again reinstated by 1950, including early-burning in the north of Kissidougou Cercle. For villagers, these policy swings made little difference to their experience of fire policy, since in all cases implementation involved the removal of local control over fire management. Specific policies remained indistinct, within a general perception that forest guards banned

villagers’ fire-setting, and then sometimes (in early burning) set it themSelves.

While earlier fire policy had been implemented too weakly to affect local practices, from the early 1950s it was implemented with greater administrative

250 Misreading the African landscape force. The strategy involved not only rules and sanctions, the tools of earlier policy, but also direct demonstration and action by teams of guards coordinated between the forestry, agriculture and Cercle administrations.5° Forest guards

were urged to act as a repressive police force (Rouanet 1951a), and to use draconian military methods. These came easily to the ex-military guards recruited into the service who until the establishment of a forestry school at Mamou in 1956 received no other form of training. Combined with the increase

in forest guard numbers, fire policing in Kissidougou now increased sufficiently to have an impact on local fire management practices. The impact of fire policy was felt very differently in northern and southern

areas, reflecting the differences in local fire management strategies as discussed in the previous chapter. In the north, the forestry service failed to appreciate how inhabitants understood and used early-burning, often assuming them ignorant of it, and indiscriminate in their fire use. In villagers’ eyes, forest guards were attempting to ban fire, and then bizarrely setting it themselves, in ways which villagers felt were inappropriate (and hence destructive) to local ecological conditions. Fear of fines nevertheless led to a loss of public activity around fire setting and control, and villagers lost much control over and flexibility in their sequencing of fire-using activities, so necessary given the high variability of dry season lengths and intensity. Elders in the north who remember earlier fire management consider the late colonial period as the beginning of a period of greater fire threat and damage to their property and plantations — an effect which would deepen during the First Republic. In the higher rainfall and more populous southern parts of the Cercle, the situation was somewhat different. Here early-burning was forbidden, and policy focused on fire-breaks and the banning of all running fire. As we have seen, in

certain areas this policy coincided with local fire management techniques already in place; but it would be the 1960s before the changing herding— farming balance increased villagers’ incentives to practise them more strictly, when fire policy and local priorities and practices came finally to dovetail. Forest reservation was also pursued with new vigour in the post-war period, and several forest islands near Kissidougou town were reserved.°! This was justified by the studies of the forest officer Rouvain>? and botanist Adam. The latter, in 1948 reemphasised the annual retreat of woody vegetation in the forest region (Adam 1948), and again attributed it to the southwards movement of Maninka people. He considered Kissidougou’s forest islands to be destined to disappear within a few decades. Rouanet, the head of the forestry service,

fully endorsed Adam’s analysis, considering that dense forest must have covered Kissidougou 200 years previously. In 1955 (and despite much higher than average rainfall) alarm was again expressed that the region was drying out and turning from a forest region to a sudanian zone no longer apt for coffee; urgent forest protection and reconstitution were needed to restore its forest

Reading forest history backwards 251 character. These perceptions gave new vigour to forest reservation (Rouanet 1951a). However in Kissidougou, reservation confronted a particular problem since Kissidougou’s ‘relic’ forest islands were now filled with coffee. In the context of the high coffee prices of the late 1940s and 1950s, farmers had increased their production so that between 1949 and 1955 Cercle production increased tenfold to 10,000 tonnes per year. It was impossible, therefore, to turn many more forest islands into strict forest reserves. The forest conservation strategy which was proposed instead had two main

elements. The first was to exploit the valuable timber from within these ‘otherwise doomed’ forest islands. To this end, a sawmill was established in 1947, paving the way for the establishment of many similar privately owned commercial timber exploitation outfits throughout Kissidougou, which served growing urban markets. By 1956, Kissidougou was recognised as Guinée’s

foremost timber producer.°3 A second element was to encourage forest reconstitution in the savanna abutting forest islands. This involved the planting

of quick-growing, fire tolerant exotic species, such as Cassia siamea and Gmelina sp. which were introduced around many village forest islands at this time, intended to extend forest islands in association with coffee, kola and other

tree crops.°4 By this time, however, coffee was sufficiently profitable for villagers to be already doing this, using their long-established fire-exclusion and soil improvement techniques. While content to try out the introduced

exotics, villagers eventually found that these ‘artificial trees’ suppressed further natural regeneration. In the 1950s, permits for tree-felling thus became a major source of revenue for the forestry service. Permits were imposed for felling any tree from a list of protected species, effectively giving felling control to the state and permit

holders. The species were principally those most valuable as commercial timber, and which were also perceived as most representative of the Cercle’s diminishing natural forest cover. In combination with a policy which gave villagers only a 10% royalty (paid in kind) when timber was felled by private, generally French, contractors, villagers lost substantial aspects of their tenurial control over trees which, in many cases, they recognised to exist because of their or their ancestors’ own work and encouragement. Villagers also faced problems when they wished to fell small individuals of protected tree species for use as poles, or in the farming of gallery and upland forest fallows, as this had become illegal.

In the immediate pre-Independence period, the repressive tactics of the policy-empowered ‘forest police’ reached such proportions that they became a target of political movements in Guinea (cf. Riviere 1974). Politicians in the run-up to elections could easily rally rural political support through slogans such as ‘we promise to give you back your lands and forests’ (Diallo 1989). This led to a position in which the Forest Department could no longer count on

252 Misreading the African landscape the direct support of other colonial administrators, and was forced to reform. The forestry service shrank, funds were reduced, and in an attempt to reduce its unpopularity there was a partial declassification of forest reserves, converting them into ‘sylvo-agricultural reserves’ with, in effect, no restrictions over their use. In Kissidougou, the forest islands which had been classified in 1945 were now declassified. It was, as Rouanet (1951a) suggested, necessary to rethink the country’s forest policy. Such changes, however, were not to occur at more than a cosmetic level during the First Republic.

The First Republic: 1958-84 In 1958, Guinée voted ‘No’ to further collaborative policies with the French in the Post Independence era. It installed a ‘revolutionary’ African state socialist

regime, the Parti Démocratique de Guinée (PDG) under the Presidency of Sékou Touré. Among initial actions, the regime abolished the unpopular ‘traditional’ cantons, chiefs and SIPs through which the colonial admin-

istration had ruled. New forms of rule were based on elected village revolutionary councils, with villages grouped into localised assemblies and these in turn into arondissements. In 1959 all land was declared the property of

the state. Public sector investment was initially dominated by the needs of organisational restructuring with little investment in the rural sector, whether for agriculture or for the environment (Mullen 1989). Indeed agricultural production was sacrificed to satisfy urban demand for low food

prices to boost industrialisation. Such rural investment as there was, was concentrated both on modernisation/mechanisation, and on collectivisation. Despite these major political changes, the colonial vision of a degraded and degrading environment was carried over wholesale into the post-Independence period, with environmental policy and the analysis informing it best characterised in terms of continuity. This continuity in policy was partly institutional. The renowned French botanists who had created the degrading vision had, by the end of the colonial period, become the most senior figures in the French colonial environmental administration. Their legacy was not merely academic, but was inscribed within the very organisation and operation of the Eaux et Foréts institution, not least in the permits, fines, and revenues from state forest reserves which formed the basis of its resources. Certain reforms were made, including the abolition of the position of Forest Guard in favour of forest agents

who were required to have forestry training and could not be ex-soldiers. Nevertheless, forestry agents became much more intrusive in village life, and cosmetic reforms made little difference to the service’s practice as experienced by villagers, who since the colonial period have referred to forest agents under the same term, ‘gadifore’ (gardes de forét). Initially during the 1960s, and following the dismissal of the forest guards,

Reading forest history backwards 253 the forestry service was not large, but it gradually increased in size. The number of trained foresters remained small (about 200 in Guinée in 1984) but the service was reinforced by recruitment from the vast pool of agronomists trained in the university agronomy faculties which were established in each Prefecture. By 1984 there were 45 agronomists and foresters working in the Kissidougou section of Eaux et Foréts.

The lack of immediate post-independence administrative priority for environmental issues, and the capacity of staff to finance themselves through fines and permit-granting assisted the maintenance of the status quo. In as

much as environment was made the subject of policy review, it was the ex-colonial service head Rouanet who was called in to advise, and who carried over colonial analysis of forestry objectives (Rouanet 1961). The new service combined the colonially inherited environmental vision and policies with the slogans and state socialist momentum of the First Republic’s regime. New and more draconian forestry laws were created: there was, for instance, a total ban on all bush fire from 1972, when forest burning was criminalised and carried

the death penalty (Jean 1990). These, combined with a real decline in state functionaries’ remuneration and official benefits, and with dysfunctional bureaucratic management, created the conditions in which the large numbers of forest agents could operate in villages in an extremely repressive, but often freelance, way. Repression was thus doubly felt by villagers in both its official and unofficial capacities, intensifying their already severe experiences of the late colonial period. In the north, villagers describe how: During the period of taxes and famine, which weighed heavily on us during the First Republic, forest agents prevented us from felling trees, from setting fire even in the field, from cutting chewing sticks and from hunting even on our own territory. We were concerned and disorientated in land management as we felt ourselves to be strangers and robbers from the State. (Moria locality, cited in Millimouno 1993)> Whether the fire originates here or not, forest agents pass from village to village to the chiefs, saying that they have a bereavement. Consequently, each village must present its condolences, usually a lump sum which they fix themselves. We gather this from each compound, (Massamaya locality, in Millimouno 1993)

Policy in the north of the region was thus perceived as interfering with existing fire management strategies. Villagers increasingly felt highly constrained in their upland farming since in this relatively dry and sparsely populated area, the

labour costs of preventing the spread of fire outside field boundaries were prohibitive. Gallery forest farming required either a permit or extreme secrecy.

This may have been a further factor encouraging the shift of rice farming from uplands to swamps. Nevertheless, needing to continue with some upland farming and with setting early fires to protect property, villagers developed

254 Misreading the African landscape various coping and resistance strategies. These ranged from offering hospitality and largesse to forest agents to prevent them visiting their territory, to strategies for the secret setting of fires.5® As one villager in the east expressed it: ‘We came to judge it best to continue managing our land as before, giving the forest agents money each year, because they clearly need only money and have no concern with forest protection.’ Such coping strategies seem to have mitigated the greater potential for late fires otherwise implied by these policies. Certainly, fire seems not to have diminished, as policy intended, but whether

the problems villagers experienced in fire control generally led to more devastating fires seems unlikely, given evidence of increasing woody vegetation over the period. The effects of late fire on vegetation may have been compensated for by other contemporary factors, such as those linked to cultivation and cattle. Citations from the south-east suggest, as earlier argued, that the effects of external fire intervention were now to reinforce existing strategies: Our penalties and fines against bush fire were not strong before. But following the arrival of the forest agents, they became ten times more severe. From that time, firebreaks around fields became essential. Thanks to this reinforcement from outside, the savanna has disappeared here. Fire no longer passes through here. But oddly, despite our efforts, the forest agents do not cease to menace us and to ask for money when we burn a field, when a tree dies naturally in the forest, or when one cuts a tree for carpentry. (Toly, in Millimouno 1993)

In this region, therefore, the policy seems to have assisted the cession of herbaceous savanna to forest thicket vegetation during this period. However, with the growing trees, farmers’ problem of ‘fire’ developed into the problem of authorisation of tree felling, whether in their forest islands, gallery forests or on their fallows. We ask ourselves if the State really pays the forest agents. If they really are paid, then they are abusing our status as farmers. Because it has been more than twenty years now that fire no longer passes outside the fields here. With all the time that we take to protect and enlarge our forests, one is still fearful to build a house or make a bench. They come immediately to ask you the source of their wood. And because it is wood, you have thus cut the forest. They then make you pay a large sum. (Kofodou, in Millimouno 1993)

In agriculture, policy pursued two linked strategies: modernisation through mechanisation, which intensified strands of colonial policy, and encouragement to collective forms of production, in line with state socialist goals. The Centres Nationaux de Production Agricole (CNPA) which took over the role of colonial agricultural research stations placed emphasis on highly mechanised monocultures and chemical fertilisers, although funding constraints meant very little research progress was made during the period (Leunda 1973-4). Centres

Reading forest history backwards 255 de Modernisation Rurale (CMRs) were established to transmit national research support to farmers through model farms, but by 1962 these had collapsed due to lack of funds, problems with spare parts, and failures in extension (Leunda 1973-4). From 1960, in a major propaganda campaign, farmers were encouraged to

join cooperatives (CAP, Coopératives Agricoles de Production). These grouped the farmers or herders of one or several villages, who were encouraged, by stages, initially to create some collective fields on which to use

ploughs or tractors purchased with credit, and eventually to make all fields except kitchen gardens collective. Produce was sold directly to the state. There

were annual production norms for each ‘cooperator’ which they met from either individual or collective fields. Between 1960 and 1962, 481 CAPs were created in Guinée, with an average of 123 members per CAP. In 1965, as part of a strategy to encourage greater membership, cooperative production was linked with the provision of consumption goods such as cloth, pots and pans, and zinc pan roofing.>’ The cooperative movement declined from the mid-1960s, when it became

evident that those who formed coops did so largely to gain benefits and equipment which they then used on their own land, rather than in collective farming. Furthermore, many were concentrated around major towns and on the Niger plains, and became the family enterprises of traders and salaried

workers. Indeed the policy came to encourage the emergence of a large farmer/trader class (the “bourgeoisie du tracteur’) (cf. Riviere 1971b, 1973).

Villagers were reluctant to engage in such externally imposed collective farming which had little basis in local agricultural organisation: even where village work was communal, as in labour circles, fields and products had been more individually controlled (cf. Leunda 1973-4). In 1967-8 the regime imposed a more wholesale rural reorganisation, based on ‘brigades’ organised at the level of the grouped village assemblies, now termed Pouvoirs Révolutionnaires Locales (PRL). Among brigades dedicated to various activities were agricultural brigades, which in 1973 were mechanised with tractors or ox-ploughs. These were expected to farm some land collectively; encourage other forms of collectivisation such as fields farmed by youth groups, and promote ‘rational’, mechanised production on members’

individual fields by making equipment available (Leunda 1973-4). They also stored and marketed the entire village’s individual production quotas>® at official prices, and organised the distribution of consumer goods in the village.

Between 1975 and 1978 most government agricultural expenditure — now taking a greater (27%) share of public sector investment (Mullen 1989) — was

concentrated in various types of brigade. By 1976, Kissidougou had 47 brigades mechanised with tractors, and 32 with ox-ploughs (Touré 1987). Brigade organisation proved ineffective in achieving state mechanisation

256 Misreading the African landscape and collectivisation goals, however, both because of shortages of spare parts, fertilisers and other inputs, and because of farmers’ ongoing preference to concentrate on their private farmland. Furthermore traders were seen to shortcircuit brigade marketing functions with the blackmarket. In response, the

regime sent more state agricultural technicians and party militia to exert further control over marketing and to encourage collective work forms. But eventually, in 1979, it abandoned brigades in favour of more directly controlled state farms (Fermes Agro-Pastorales d’Arondissement, FAPA). By 1984, fourteen FAPA had been established in Kissidougou, more or less one per Arondissement (Touré 1987). These were equipped with tractors and staffed by former employees of the government’s by then largely disbanded agricultural service.°? Each FAPA farmed a quota of land to different crops, and was intended to serve as a centre of research and demonstration (e.g. of swamp irrigation techniques) as well as of production. Their outputs were often poor. Villagers received little from FAPAs, although they were expected to provided 60% of their labour, contribute to their maintenance through their taxes, and give or sell their cattle to them, at official prices (World Bank 1983). Like the earlier brigade-organised collective fields, FAPA placed demands on farmers’ time, and imposed new management techniques on parts of their landscape. In Sandaya, for example, a large area of the Niandan river plain was subjected to tractor-based FAPA rice cultivation, with bulldozed irrigation channels. Tractors were also used for repeated peanut cultivation on an area of tombondu soil, which left an enduring legacy of reduced woodland regeneration and sparser fallow on that portion. Villagers from nine neighbouring villages were obliged to contribute a day’s labour in each phase of the farming cycle. But villagers’ contributions to such activities seem to have had little enduring influence on their work patterns, and on their management of parts of village territory not farmed by the state, which they continued to use as their own. Thus apart from required periodic contributions to collective fields, most

villagers carried on with their accustomed agricultural organisation and techniques. The greatest impact on land-use patterns seems to have come from the more

indirect effects of high taxation, economic repression and consequent emigration of people and livestock. Villagers found their product commercialisation and trading activities highly constrained by the regime’s progressive, and increasingly severe, attempts at state control of marketing. State marketing boards were established to cover all regional and national commercialisation, while the repression of private traders culminated in 1975 in a major offensive against them which became known as ‘Cheytan 75’. Commerce was paralysed,

and was only liberated from the ‘economic police’ in 1977. Consequences of these policies included a fall in coffee production in Kissidougou from an estimated 10,000 tonnes in 1955 to 1,500 tonnes in 1976 (Touré 1987), a

Reading forest history backwards 257 consequence of the drastic fall in producer prices. By 1983 an estimated 10,000 livestock a year were exported clandestinely from Guinée to avoid low prices

and requisitioning (World Bank 1983). Many young people sought urban work, and others sought refuge across international borders to escape taxation demands and the surveillance of the ‘economic police’, and rural populations diminished. Commonly, only one son remained in the village to care for his parents as they aged. The shifts in agricultural and land-use patterns resulting

from such pressures, and their impacts on vegetation, have already been discussed.

The Second Republic: 1984—present Towards the end of Sékou Touré’s regime, his administration weakened under

economic pressure and increasing international disapproval. Soon after his death in 1984, a military takeover ushered in Lansana Conté as president. In late 1993, his PUP party won the reinstated multi-party presidential elections. His regime initially alleviated the taxation and restrictions which had been burdening villagers, and in doing so managed to gain considerable rural sympathy. In accordance with World Bank/IMF imposed Structural Adjustment policy, agriculture was made a priority sector of investment, with the accent on small-scale producers and liberalisation of markets to increase producer prices and encourage individual and family production (Mullen 1989). Producer prices had risen three- to fourfold by 1986 (Clapp 1993). State

farms were abandoned, often converted into integrated rural development projects. As a new cornerstone of development activity, such projects took on many ex-FAPA staff as extension workers, and much infrastructure. Several received foreign donor support, including the project Dévéloppement Intégré Rurale de Kissidougou (DERIK). In Kissidougou’s agriculture, such projects

focused on the twin national policy emphases of swamp rice and coffee. New seeds and fertiliser and improved water control practices were promoted.

The emphasis on mechanisation was reduced in favour of encouraging agricultural intensification through manual methods and ox-ploughs where appropriate.

But in contrast with agricultural liberalisation, environmental repression continued. The forestry administration reinstated the colonial ‘forest police’, the forest guards, whether because it considered forest protection and fire control policy during the First Republic ineffective, or because of the need to absorb unwanted military personnel. In Kissidougou, 30 ex-soldiers joined the forestry service in 1984, bringing the total number of forest agents to 75; about five times the number present at the end of the colonial era. Indeed, between 1958 and 1986 the number of forestry personnel in Guinée as a whole increased

from 138 to 3,194 (République de Guinée 1988), although public service

258 Misreading the African landscape reduction under Structural Adjustment reduced staff in 1987 — by thirteen in Kissidougou. Environmental policy and intervention in the late 1980s became more than ever influenced by the agendas of foreign aid donors, in the context of global

concerns with the environment and ‘sustainable development’. Numerous foreign advisory missions came to play an important role in formulating national policy documents, including Guinée’s forestry action plan in 1988 (République de Guinée 1988) and a proposed national environmental action plan. Expatriate advisers have been posted at the highest levels within the agriculture and forestry ministries to influence future strategies, and to arrange

the disbursement of financial assitance. As we have seen, in 1990 the programme to improve and regularise the flow of the major Soudano—Sahelian rivers by restoring the vegetation cover of their upstream watersheds in Guinée, first outlined in 1934 and first implemented in the 1950s, was relaunched, with

major USAID and European Union funding, and two sub-projects in the prefecture of Kissidougou. International donor pressure has been instrumental in developing an orientation towards local ‘participation’ in environmental and forestry strategies. Guinée’s 1988 forestry action plan, for example, emphasises local community involvement in the implementation of forest and fire policy as does the proposed national environmental action plan (Ministre du Plan 1993). The latest round of the Niger River protection projects are implemented in the framework

of devising with villagers joint management plans for village territory: an approach internationally known under the French term “Gestion/Aménagement des Terroirs Villageoises’ (GTV) (cf. Painter 1991). Such plans incorporate village-level planning of bush fire, upland and forest use, tree-planting, and reservation of forest islands in favour of ‘the community’. Some consideration

has been given to the reform of the forestry service and its laws, and to changing its relationship with the rural population to foster such an approach.

Thus in 1993 forest guards were again abolished. A more ‘hands-off’ approach, implemented through the prefecture and a new restructured and decentralised rural administration framework, is envisaged.®° The new forestry code of 1989 incorporated a number of these goals.

Despite these administrative changes, the environmental intervention repertoire has hardly changed, and shows few signs of doing so. Projects still emphasise reduced upland farming in favour of swamps, and the rationalisation and intensification of residual upland farming, through ‘model’ agroforestry systems and crop rotation systems. Forest reconstitution is promoted through tree-planting in village territories, largely in the form of community woodlots

of exotic and indigenous species. Policies continue to focus on bush-fire control, with prefecture-level regulations (e.g. concerning early-burning and its timing) providing a limiting framework on village-level plans. Lastly, there

Reading forest history backwards 259 have been attempts to control the perceived ongoing destruction of forest islands by the urban-based, chainsaw operators which proliferated from the mid-1980s. In this context, forestry department control over timber felling has been tightened. Common to these recent policies and project interventions

is their recourse to technology ‘packages’ such as inland valley swamp development and tree planting from nurseries which, given their long history,

are now well established in the region. Common, too, is their attempt to establish or to reestablish ‘control’ and ‘organisation’ in villagers’ environmental activities, perceived generally as disorganised and unsustainable. Whether the state or “the community’ is best placed to manage such control is a matter of ongoing debate among Kissidougou’s various funders, policymakers and administrators. The continuity in environmental interventions is unsurprising given that policy is still working with the idea of recent and ongoing degradation of extensive forest cover in the Kissidougou region. Indeed the major causes of deforestation postulated today are those which evolved during the colonial period. As chapter 1 showed, modern national policy and project documents

frequently emphasise the irrationality of upland shifting cultivation, the problem of uncontrolled bush fire, the negative effects of increased cattle numbers, the ignorance and indiscriminate activities of the population, and anthropogenically induced negative climatic change.®!

Notably, recent calls for a more ‘participatory’ approach have not been driven by any new recognition of indigenous technical skills. Nor, it seems, does local participation extend into the analyses of environmental change and the setting of environmental values which guide policy and project objectives. Villagers who are expected to ‘participate’ in the environmental projects, initiated by state agencies with considerable foreign support and presence,

understandably have many anxieties. These include loss of land to ‘the project’s’ trees; the loss of control over management of the environment to outsiders ignorant of its local specificities, and the unknown future demands that apparently generous projects of unknown origin and intent, huge financial

resources and foreign interests may later exact. Everyday strategies of resistance are one response: letting project tree nurseries and plantations burn in the dry season, for instance.

Conclusions

The reading of Kissidougou’s landscape as degraded and degrading has informed forestry and agricultural policy from its inception. It has been developed, institutionalised and reinforced during a century which has seen a series of very different regimes. Ostensibly concerned with the environment and the sustainability of resource use, this landscape reading in policy has had

260 Misreading the African landscape

the instrumental effect of appropriating resource control and revenue from villagers, and of extending state bureaucracy into rural areas (cf. Ferguson 1990). Indeed, for a long time, forest agents have been the most invasive and extractive state presence in the prefecture’s rural villages.

Understandably, Kissidougou’s rural population bears considerable antipathy towards state environmental services. This is not least because interaction with them has often proved problematic for their own land use and enrichment practices. By removing rights to set fire, clear fallows and fell timber, state policy has reduced villagers’ capacity to control fire in the north, reduced their flexibility in farming, and reduced their control and profits from timber production. The capacity for commercial timber operators to draw on State authority has reduced villagers’ control over tree felling in their forest islands. The capacity of pastoralists to draw on state support for rights to settle and set late fires has reduced villagers’ control in balancing savanna and forest vegetation forms. That these problematic interactions have not had a more degrading effect on vegetation is owed at least partly to the effective resistance strategies which villagers developed. Remarkably, Kissidougou’s broad vegetation history during the past century — of increasing forest cover in savanna — has in fact coincided closely with policy objectives, although these changes were largely unperceived by policy-

makers. Evidently, local interests in managing the environment for agroecological productivity and local needs sometimes produce the results which

policy, driven by other internal and external agendas, aims for. Given the complex influences on vegetation change examined in the previous chapter, it is nevertheless hard to attribute a clear causal role to policy; especially policy whose implementation has been patchy and at times weak and ineffective. Where environmental policy did have an impact on land use changes, it was not because people adopted external suggestions wholesale, but because they used possibilities made available as inputs into their own resource management strategies; for example in the case of swamp rice and peanut varieties, and of

the greater sanctions on failures of local fire control in the south. Thus at particular moments in the conjunctural history of land use, external interventions dovetailed with emergent needs. Moreover, it can be argued that local

resource management patterns have responded less to environmental and agricultural policy than to changing social, economic and demographic conditions. These have of course themselves been influenced by policies, for example concerning taxes, prices, road-building and employment, but policies

whose environmental effects were unforeseen consequences, and indeed remained unacknowledged within the dominant conviction that Kissidougou’s vegetation was degrading fast.

9 Sustaining reversed histories: the continual production of views of forest loss

The persistence of a reversed reading of Kissidougou’s landscape over a century seems hard to credit. In this chapter we explore some explanations for its endurance, examining how degradation visions have succeeded in excluding other environmental perspectives, including those of local inhabitants, and thus why the vegetation changes which have occurred have been rendered invisible. The degradation position, we will argue, has been sustained not on the basis of ignorance, but through the continual production of supportive knowledge.

Those who are convinced of deforestation and savannisation do not lack ‘evidence’ to draw on in support of their convictions. At stake here is partly the inheritance and transfer of scientific ideas and supportive theories. But while

tracing the genealogy of landscape readings through individuals and institutions is useful, it is also inadequate. It overlooks how successive generations of observers repeatedly ‘rediscover’ readings for themselves, within common

sets of intellectual structures and social relations. The persistent conviction of savannisation, we suggest, owes a great deal to this process of continual derivation. In this production of knowledge, the concepts and methodological structures

of ecological sciences have been centrally important, but the relations of production extend beyond ‘science’ itself. This chapter therefore explores the social, political and financial conditions which have mutually supported each other in upholding this powerful backwards reading of history, and the ways that local political processes have been incorporated into it. It is these political relations, harnessed to the structures of ecological science, we suggest, which give the savannisation discourse its hegemonic capacity, silencing challenge.

Forest loss perceived Outsiders who consider that Kissidougou is subject to progressive forest loss deduce this from observations made in the prefecture, conditioned both by particular methods of observation and by broader visions of West Africa’s environment. Scientists very commonly use ‘snapshot’ or short-term landscape 261

262 Misreading the African landscape observations to deduce vegetation change and people’s impact on it. While the approach of inferring process from form has been elaborated and has acquired credibility within certain sciences, administrators and other interested outsiders also do this, equally convinced that what they see today reveals a particular

history. Chapter 1 has already considered some of these ‘indicators’ of deforestation and savannisation, and the challenges to which they are open on ecological grounds alone. Yet as subsequent chapters have shown, the perspective of villagers’ own vegetation use and management practices opens up further ways of interpreting these indicators, with very different implications. It is worth briefly bringing these together in summary. The most basic of the indicators is the presence of forest islands, taken as natural relics of an extensive forest cover, indicating a historical process of forest loss and forest destruction by inhabitants. This deduction is commonly made not only from on-the-ground botanical, forestry and vegetation survey observations, but also when forest islands appear in remotely sensed imagery. Yet to the extent that forest islands are encouraged to form around settlements, as shown in chapters 2 and 3, they testify instead to inhabitants’ capability in forest establishment.

The very species diversity of forest islands is taken to indicate that ‘they must be natural relics’, having once conjoined in larger forests to have acquired that diversity. But this view overlooks how people import species, and actively

create diverse forest island composition to meet a range of local needs. It overlooks how diversity builds up over long time periods, and a much more complicated history of forest islands (and inter-island vegetation) coming and going with settlement in different parts of the landscape, opening and closing

species distribution routes at different times. That silk cotton trees (Ceiba pentandra) often disproportionately dominate forest islands has been taken to indicate the felling of other species, and their loss of diversity. But while other

species may indeed have been thinned out for timber, this overlooks how people have used Ceiba in forest establishment.

Botanists have surveyed the vegetation of Kissidougou’s forest patch boundaries, and where they showed a mixture of forest and savanna species,

they have generally deduced this ‘transition woodland’ to represent the advance of savanna into forest (e.g. Adam 1968, Thies 1993). They have not considered how transition woodland might form in savanna as a complex outcome of farming or grazing strategies, as chapter 6 has outlined. Other aspects of vegetation form are also used to indicate forest retreat. In individual trees, a tall, straight unbranched bole is often taken to indicate that it grew up in forest. Where such ‘forest forms’ are found on open land, they thus indicate forest recession. Yet not only can most pioneer forest species grow into this form on open land (being ‘self-pruning’) (Hawthorne pers. comm.), but these observations overlook how trees might be ‘trained’ —

Sustaining reversed histories 263 encouraged to take different forms for different purposes (e.g. the look-out posts considered in chapter 3) — or indeed, take particular forms because of the ways they were planted, as stakes or seeds. Observers have repeatedly considered the presence of oil palms to indicate that original forest has retreated from the area, on the grounds that palms hardly regenerate in regularly burned savanna (e.g. Valentin 1893,! Keay 1959, Green

1991). Yet this overlooks how people can and do establish palms outside forest, often in association with farming cycles which provide necessary fire protection during their vulnerable early years of growth. And it overlooks how oil palms feature in people’s establishment and enrichment of forest islands, in the dynamics of their creation and destruction.

Similarly, observations of processes seen in the short term are taken to indicate long-term trends. Observations of the felling and clearing of forest islands for farming have been taken to indicate the ongoing and unilineal destruction of forest islands by their inhabitants. Equally, the arrival of species associated with drier vegetation in forest islands, sometimes following their thinning for timber or plantations, has been taken to indicate a general process

of savannisation (e.g. Thies 1993). The error of reasoning is clear when the landscape is considered on a larger and longer-term scale. While one particular forest may be being felled or thinned, or may be acquiring savanna species, another nearby may be in the course of establishment and accumulation of forest species. In a similar vein, the felling of gallery forests for farming, and the establishment of grass cover in the first years following cultivation, has been taken to indicate villagers’ progressive savannisation of forest through farming. Such deductions overlook forest reestablishment later in the fallow cycle, and the steps farmers often take to encourage this.

Foresters have taken observations of ‘more fire this year than last’ to represent an extension of an historically worsening fire problem. Observers have also taken low rainfall and the drying of water sources to infer a gradual climatic desiccation, frequently considering it to result from forest loss. Once again this overlooks timescale issues, as dry phases seem, in many parts of West Africa, to be statistically explicable as part of longer term cycles (van

Rompaey 1993). Second, as expressed in chapter 1, it overlooks the more global causation of rainfall changes. Third, there can be many other reasons for

the drying of sources and swamps, related to the precise balance between infiltration and evapotranspiration, and to changing swamp farming practices.” These alternative interpretations have been excluded as indicators have, today and over the past century, been read in conformity with a ubiquitous structure

of reasoning concerning people’s role in forest-savanna dynamics. In this, people are considered as external to ecological systematics, and as external agents, they impact only in converting natural forest to derived savanna — 1.e. in degrading ‘nature’. Arguments of this kind by early colonial botanists and

264 Misreading the African landscape

foresters are echoed in the most modern of scientific literature. Thus in the most recent work on the nature and dynamics of forest-savanna boundaries (Furley et al. 1992) there is no reference to positive human impact on any forest vegetation. While forest advance is frequently noted in this literature, it is attributed to non-human influences such as climatic change, or the influence

of wild and domesticated animals or termites (Hopkins 1992). Or it is attributed to reduced ‘anthropic factors’ or reduced ‘human interference’; as people move out (Eden and McGregor 1992), or as people set less fire (Kershaw 1992). Where forest is avancing in the presence of people, this is found to be despite their presence: as Ratter argues: ‘No doubt [forest advance] would have been observed much more widely, were it not for man’s [sic] effect

in preventing forest extension by agricultural activities involving the use of fire, grazing etc.’ (1992: 426). While recent work does highlight how anthropic degradation has sometimes been exaggerated (e.g. Eden and McGregor 1992, Ellery et al. 1992), people are never considered to have any positive influence on forest vegetation on the boundary.

Within Guinée, environmental services have been so convinced of the degradation they are combating that they have not thought it necessary to compare the air photographs and satellite images they have commissioned with

those available from 1952. Furthermore, comparative interpretation, even when carried out, is frequently not independent of preconceived ideas of vegetation change. In Kissidougou, the incredulous reactions of forestry staff

when presented with 1952-1990 air photograph comparisons showing increased woody vegetation led them to a sceptical search for ways to render

the comparison invalid (the photographs were taken in atypical years, or incomparable seasons . . . ). In other parts of Africa, similarly surprising results have simply been disbelieved and dismissed: Apparent increases in bio-mass from pre-disturbance [vegetation] to present were labelled ‘discrepancy’ and such discrepancies were omitted from further analysis . . .

there seems little possibility that biomass has increased as a result of land use. (Houghton et al. 1993: 310, 312)

In contrast, such scepticism was cast aside when a comparison of eastern Guinée satellite images taken ten years apart seemed to show significant vegetation degradation, and on which basis major donor funds for a regional environmental rehabilitation programme were secured (Grégoire et al. 1988). The major conclusions were accepted, despite Grégoire’s own (rather hidden)

warnings concerning the incomparability of years, and that the results constituted little more than a methodological experiment. These locality-specific observations conform not only with broad assump-

tions, but also with the iconography of West African vegetation maps. Showing the forest zone, forest-savanna transition zone, guinean, sudanian

Sustaining reversed histories 265 and sahelian savanna zones and desert in more or less horizontal bands, these

easily lend themselves to interpretation as temporal, as well as spatial transitions, whether from desertification, sahelianisation or savannisation of forest. In Guinée, policy-makers have been convinced of these southward shifts. Indeed in 1993, it was the conflation of spatial with temporal zones which provided the logic for a major donor-funded environmental rehabilitation project to take 40 Kissidougou farmers on a journey to northern Mali, to see the future of their own landscape if protective measures were not undertaken.

Within each vegetation zone (e.g. the forest-savanna transition zone) the picture of spatio-temporal shifts on the vegetation map is complemented by the iconography of ‘divergence from a climax vegetation type’: the notional

maximum vegetation which could exist given climatic conditions. This contains the idea of the previous existence of a ‘bigger’ and ‘better’ vegetation

type ‘prior to human disturbance’ — closer to the ‘Eden’ which Africa’s environment so often represented in colonial imaginations — which cannot regenerate because of human land use which maintains it as an anthropogenic

sub-climax. In this way, the present state of each vegetation zone may be envisaged as the anthropogenically degraded derivate of its predecessor. Interpretations of vegetation degradation are reinforced not only by local observation, but also by the global and regional level analyses with which they are in keeping, and which carry the weight of international authority. Given

FAO figures concerning rapid forest loss in West Africa (FAO 1993), for example, it appears inconceivable that Kissidougou should be experiencing anything else. In the context of the growing internationalisation of information flows, such data, so frequently publicised in the glossier development literature and on the radio, are far more accessible to the environmental administrations and urban public concerned with Kissidougou than are analyses of the locality itself. Equally the rhetoric of shared environmental crisis, made so apparent in the 1992 UNCED conference in Rio, appeals far more powerfully to local officials than the statements of the villagers supposedly experiencing these problems. Thus in the 1993 ‘Environment Days’ designed to raise awareness of Kissidougou prefecture’s environmental problems, both the prefect and Kissidougou’s urban-based environmental NGO framed their speeches in terms of global concern with biodiversity loss and the common West African struggle against desertification. While the projection of global and regional

concerns onto Kissidougou’s environmental management has _ recently heightened, they have periodically had a major influence on earlier administrative perceptions of environmental change: whether in the early regional climate and hydrology concerns, or in the Africa-wide soil conservation concerns of the 1940s.

The ways that the relationship between scientists and state institutions

266 Misreading the African landscape

evolved assisted the incorporation of these visions into policy-making processes. The assumption of anthropogenic degradation of a prior natural

forest formation was integral to the early delineation of West African vegetation zones by the botanists Chevalier (e.g. 1911) and Aubréville (e.g. 1938). This analysis was transferred directly into contemporary policy since Chevalier was, at the time, the most senior advisor to the French West African

colonial administrations responsible for environmental concerns. Subsequently, deductions made from landscape observations by botanists such as

Aubréville, Adam and Schnell reinforced the view that the forest-savanna mosaic was in temporal transition. As the chronology considered in the previous chapter suggests, Aubréville inherited ideas from Chevalier, and Adam from Aubréville. They also incorporated ideas from prevailing environ-

mental debates in the West African anglophone world, which in turn had scientific links with the Indian sub-continent. In their characteristic approach, these botanists directly observed landscape features and deduced landscape history and the impact of local practices from them. Their disciplinary position and the social conditions of their fieldwork reinforced their pejorative vision of the environmental impact of local farming and fire management practices, rendering it both difficult and seemingly unnecessary properly to verify change with inhabitants themselves. As Aubréville, Adam and Schnell, in turn, became senior figures in French West Africa’s forestry administrations, so their methods and analysis became

institutionalised: written into the developing forest administration both intellectually and financially. Intellectually, their publications became key texts in comprehending West African environmental change more generally (e.g. Aubréville 1938, 1949; Schnell 1971). They became key references not only in the training given to foresters and botanists, but also in establishing the

basic environmental parameters which informed the studies of historians and other social scientists. In Guinée, by the end of the colonial period, the deforestation analysis had entered formal sector education and the popular consciousness of state functionaries. It was this view of environmental change and of farmers’ role in it which was incorporated and reinforced through the training institutions of the Sékou Touré period. These included the ‘Centres d’ Education Revolutionnaires’ (CER) of the 1960s, and the ‘Cités Socialistes’ of the 1970s, which saw youth training, of an ideological as well as technical kind, as the cornerstone of rural modernisation (Leunda 1973-4, World Bank 1983). Today, as we have outlined, deforestation infuses geography and history curricula, from primary school to university level. This analysis of environmental change for Guinée and Kissidougou also came to be inseparable from the economic context in which environmental institutions operate. The concern of early colonial administrations with the

perceived destructiveness of African environmental management arose

Sustaining reversed histories 267 because the colonial economy was heavily dependent on ‘threatened’ natural resources; first wild rubber and then, in Kissidougou, on palm products and tree crops grown in forest patches. In the later colonial and post-colonial periods, regional and global economic concerns with environmental degradation built on existing national ones. In the 1950s new administrative funding possibilities were made available for regional soil, climate and hydrological conservation following the heightened Africa-wide environmental concern at the time. More recently, administrative solvency and development activities have come to rely even more heavily on foreign aid, and have thus become subject to various forms of ‘green conditionality’ (Davies 1992, Davies and Leach 1991). This greening of aid, and the specific forms it takes, reflects donors’ needs to satisfy

home political constituencies heavily influenced by media images and northern environmental pressure groups, as well as donors’ assessments of African environmental problems.

In Guinée, a large proportion of foreign assistance is now allocated, sectorally and by region, directly to environmental rehabilitation. Between 1990 and 1993 the agricultural and forestry sector accounted each year for about 16% of assistance, and this has been increasingly oriented towards environmental and sustainability concerns. Over the same period, foreign assistance to the natural resources sector, of which direct environmental concerns are about 10%, increased from 3 to 12% (PNUD 1993). This trend is

in keeping with those elsewhere in Africa, where in 1993 10% of donor assistance focused on environmental concerns, and where this is projected to

rise to 25% by 1998. In response in Guinée, there is now not only a new generation of heavily funded environmental projects, but also overt environ-

mental sustainability components in agricultural and other development activities, important for attracting future funds. Kissidougou’s prefecture administration, agriculture and forestry services are well aware of the packages

which satisfy the donors in this respect: agroforestry programmes, forest conservation and improvement, bush fire control, and rationalisation and reduction of shifting cultivation in favour of intensive wetland rice. During Kissidougou’s 1993 Environment Days the prefecture’s second administrator

Stated explicitly that: ‘Donors are interested principally in environmental projects, so we must solicit their aid to ensure the development of the prefecture’. He suggested that other localities learn from the example of the Niger

protection project zones, where schools, water and other infrastructural developments were provided in exchange for local participation in environ-

mental protection. The proliferation of environmental NGOs in Guinée, including Kissidougou’s “Friends of Nature’ society, also responds to donor interest in environmental issues, and to donor appreciation of apparent NGO capacity to achieve more ‘participatory’ development. In short, presenting a degrading or threatened environment has become an imperative to gain access

268 Misreading the African landscape to donors’ funds. In this respect, our own findings were often considered as subversive of the prefecture’s future financial and development interests, and to threaten the continued employment and material privileges of environmental project administrators and extension workers. Considering the environment as degrading is crucial to the solvency of state environmental institutions even when they do not receive donor support. Making the case for pressing environmental degradation helps to justify state budgetary allocation. And as we saw in the previous chapter, francophone West African forestry services since their inception have derived revenues from the environmental controls they impose: the sale of permits and licences for timber and wildlife exploitation, and fines for breaking state environmental laws. Acquiring control over the management of natural resources (e.g. fire and trees) from inhabitants has been achieved by imaging the latter as inadequate resource custodians whose destructive activities are in need of regulation and

repression. Revenues are thus ensured by a reading of the landscape as degraded and degrading. The importance to forestry staff of informal receipts

derived from application of policies of repression only accentuates the imperative for this environmental reading, while the antagonistic relationship

between forestry agents and villagers which is thus engendered bars communication about local people’s own experiences of environmental change and management. Thus at local and national, as well as international levels, the

economic structures within which environmental agencies operate frame the ways that information is derived. The attitudes of forestry staff can be traced not only to their financial and educational status as members of the forestry service, but also to their sociocultural positions. They share with many other formally educated urban-based Guineans a particular vision of villagers’ resource management capabilities. The image of the rural farmer as environmental destroyer, and of the need for modernisation of resource management and farming techniques, conforms to and helps to justify the self-distinction of urban intellectuals as modern. Such distinctions were reinforced under Sékou Touré’s 1958-84 regime when the urbanised were politically and economically privileged, and their vision of a highly mechanised, capital-intensive technical future dominated approaches to

rural development (Riviere 1971a). The environmental component of this degrading view of village capabilities has become dominant with the greater attention now being paid to environmental rather than agricultural issues. Generalised notions about ‘people’s destructive impact on the environment’, projected locally and coupled with images of rural backwardness, have entered the numerous processes through which many urban educated people under-

stand themselves as relatively more ‘civilised’ or ‘globalised’ (cf. Bledsoe 1990). Just as urban circles benefited from the agricultural modernisation

which wrested control from villagers, so they have become the main

Sustaining reversed histories 269 beneficiaries of environmental control, which once again gives them the moral high ground while allowing them to gain from policies which remove timber cutting rights from ‘irresponsible’ villagers. Reading the landscape in terms of degradation is implicit in these intellectual, social and institutional conditions. It is from within them that landscape

forms — even when casually observed — come so obviously to indicate degradation; interpretative tendencies which are accentuated by roadside and seasonal biases in observation. Thus the conversion to farmland of a few forest islands near the town can easily suggest forest island diminution everywhere. Casual landscape readings are often made in the dry season, during which visits of external consultants, forestry agents and urban nationals to villages are concentrated. This is the apparently ‘destructive’ part of villagers’ normal seasonal cycle, when bush is cleared for farming, fires sweep the savanna and trees are cut for construction or sale. Regeneration during the rainy season, anyway more subtle to observe, escapes attention within this seasonal bias (cf.

Chambers 1983). The failure of policies (e.g. in control of fire and timber felling) is also taken as further evidence of local ignorance and wanton destructiveness. This suggests the need to implement policy with greater force,

without questioning whether the policies themselves (and their underlying analysis) may have been inappropriate. Images of forest loss in Kissidougou are also reinforced as part of processes

of ethnic distinction, as a result of both colonial portrayals, and the ways they have been incorporated into local political discourse. As we have seen, colonial constructions of ethnic difference among Kissidougou’s populations rested partly on stereotypes concerning their environmental behaviour. As a

‘forest people’, Kissia were contrasted ethnically with the more northerly ‘savanna people’ of Kuranko or Maninka origin. In the context of historical and ongoing southwards Maninka migration, their fire-setting in savanna farming, honey collecting, and hunting were considered responsible for southwards savannisation (Adam 1948). Where Kuranko lived within forest islands, this was perceived as learning from the Kissia, as it is by modern environmental projects. Such stereotypes overlook evident similarities in everyday ecological knowledge and resource management, as well as the complicated nature of Kissidougou’s settlement history in which many supposedly ‘forest’ Kissi families can trace descent from a Maninka family of savanna origin. Yet, ethnic environmental stereotypes were reinforced and acquired greater

importance in Kissidougou’s local political discourse during the First Republic, when the regime encouraged villages to move out of the ‘mystified

obscurity’ of their forest islands into the ‘open’; into the ‘clarity’ and ‘modernity’ of the roadside savanna world (Riviére 1969). Many Maninka

self-representations draw on the ideal of social clarity, of openness and

270 Misreading the African landscape

simplicity in language and expression, and contrast explicitly their clear ‘savanna language’ (kan gbe) with the secrecy and obscurity of the forest culture and languages which they find difficult to learn. Many Kissia perceived Sékou Touré’s regime as Maninka-biased, and the attempts it made to evict Kissia from their forests and suppress forét sacrée schools were considered as

attempts to disempower the Kissi institution which had until then defended Kissia from Maninka domination, both culturally and militarily. In this context, both the present privileging of the forest, and the view that it

is threatened as portrayed by the forestry service, coincide with the broader politico-ethnic interests of urban Kissia; interests heightened in the run-up to multi-party elections in December 1993. Sharing one forest — where the forest islands of neighbouring villages have come to touch each other — is one of the

strongest metaphors of Kissi political solidarity, linked as it used to be to alliance in warfare and forest initiation. Accepting the idea that the Kissi region could (until even recently) have been united in one forest provides a politically

appealing vision of unity, as does the attribution of forest loss to Maninka immigration. Such views are most often voiced within the politically influential urban Kissi community, but can also be heard in villages when rural Kissia use environmental issues to make politico-ethnic points. Distinctions between urban-institutional, and rural villagers’, perceptions of

environmental change also derive from different valuations of vegetation quality. For urban observers and the forestry service alike, the large trees of forest islands tend to be the central focus of attention and appreciation, whether for nostalgic or commercial reasons. Villagers do not share this valuation, not

least because the forestry laws designed to regulate timber exploitation (preserve the environment) deny them all but an insignificant royalty from trees cut by outsiders in their forest islands. As we have seen, their values are conditioned, instead, by the importance of different vegetation types and species in agriculture, gathering, settlement and tree crop protection and cultural practices, and in which lower bush fallow vegetation is frequently more useful than high forest. Frequently, the large trees of forest islands are more the ‘fortuitous’ consequence of villagers’ environmental management for other reasons, than a deliberately encouraged feature. While the felling of these

trees may be of little consequence to villagers (or to forest area in the long term), to urban and official observers it epitomises, and thus reinforces their conviction of, environmental destruction.

Forest loss explained

The image of environmental degradation in Kissidougou is supported by apparently successful explanations for it, and for its acceleration. Just as intellectual and institutional structures encourage certain readings of how

Sustaining reversed histories 271 the environment has changed and obscure others, so these same structures encourage particular reasoning as to why the environment has changed. The view that local land use encourages the conversion of forest to savanna and reduces savanna tree cover and soil quality has long dominated policymakers’ thinking. This external image of local land use as inevitably degrading is combined with particular theories about the impact of demographic, social and economic change to account for the long-term degradation which policymakers believe has taken place. Discussions in development circles of the links between population and environment, poverty and environment, and social organisation and environmental management have set terms of debate which guide causal interpretations by development personnel, consultants, and national institutions. Given that it is explanations of supposed environmental degradation which are being sought, and given the prevailing intellectual, social and fiscal structures which condition causal analysis, all but the dominant strands of thinking within these debates tend to be suppressed at the project level. Thus it is Malthusian views of the relationship between population and environment, the deduction that impoverishment forces villagers to draw down their natural resources, and the notion of a ‘tragedy of the commons’, which are used to explain increasing environmental degradation in Kissidougou.

Environmental degradation is, as we have seen, attributed to assumed demographic trends by policy-makers who believe that, since local land use is

degrading, more people must mean more degradation. An image of low pre-colonial population densities is commonly linked to the supposed previous

existence of extensive forest cover in the prefecture, and rapid population growth during this century is held to account for forest decline. Short fallows and long cultivation periods on savanna uplands are often taken as evidence

of modern population pressure, and fuelwood consumption by growing populations is assumed to add to the problem. That local farmers use intensive cultivation practices for positive ecological and economic reasons, unrelated to population pressure, is not considered. That fuelwood is obtained largely as a by-product of agricultural land use, and is thus not necessarily an independent source of deforestation (cf. G. Leach and Mearns 1988) is often overlooked, as is the possibility that population growth could enable environmental ‘improvement’, however defined.

Socio-economic theories to explain supposed recent environmental degradation were considered in chapter 1, whether concerning modern poverty and ‘short-termism’, or disruption to social and tenurial structures (cf. Fairhead and Leach 1995). Casual observation of hunting groups along the roadside, of

urban entrepreneurs engaged in rural charcoal making, and of people transporting fuelwood and bushmeat to urban markets seems to provide tangible evidence of such trends. The idea that resource exploitation is disorganised and

272 Misreading the African landscape individualistic — subject to a tragedy of the commons — justifies arguments both

for more state control (e.g. over timber cutting and fire), and for the privatisation and registration of land tenure. Social scientists have also invoked more

recent strands of the commons debate (cf. Ostrom 1990, Berkes 1989), considering that local institutions can and once did control environmental management effectively, and that recent degradation is due to their breakdown. The logical policy implication is that resource use can be rendered sustainable by improving forms of ‘regulation’, ‘authority’ and ‘organisation’, although

from within this perspective, a role is seen for new or reinforced local institutions as well as state ones.

These dominant social and demographic causal explanations for degradation, and the phenomenon of degradation itself, are mutually sustaining. They combine in a compelling story in which each supports the other: a ‘narrative’ in Roe’s (1991) terms, where stories of apparently incontrovertible logic provide scripts and justifications for development action. Such narratives are the stuff of the overviews and synthesis documents written within develop-

ment agencies and policy research institutes, and of analytical alignment in development approaches. Since as orthodoxies these dominate academic

debate in development circles, reports phrased in their terms gain easy acceptance and credibility. Narratives thus help decision-makers to fill confidently the gap between ignorance and expediency. The specific narratives concerning vegetation change and its social causes used to support policy in Kissidougou are examples of a broader-level narrative written into national, regional and international policy documents. Thus it was lamented by a French advisor to the Ministry of Agriculture that: This degradation of the natural environment . . . is the result of an evolution of rural societies little adapted to the rapid structural, demographic and economic changes this

century, and above all, these last years .. . The problem today is the recession of traditional control of the orderly exploitation of space and its resources, which has not managed to follow or adapt to the recent and very rapid change in the rural world. This

management becomes insufficient given a brutal increase in population [and] a progressive loss in the power of traditional control, due to the destructuration of rural society, the new amplitude of migration and the push towards agrarian individualism and monetisation of the local economy. (République de Guinée 1989: 8)

This narrative is the script of international donors, and one could fill shelves with versions of it across Africa and beyond. Focusing on the population component, for example, a recent World Bank policy review argues that: traditional farming and livestock husbandry practices, traditional dependency on wood

for energy and for building material, traditional land tenure arrangements and traditional burdens on rural women worked well when population densities were low and population grew slowly. With the shock of extremely rapid population growth... these practices could not evolve fast enough. Thus they became the major source of forest destruction and degradation of the rural environment. (Cleaver 1992: 67)

Sustaining reversed histories 273 More than empirical evidence, it can be argued, such narratives depend on — and expose — the field of western imagination and stereotyped images concerning African society. Within this field, African society is seen, at origin, in terms of a traditional ‘functional order’ once harmoniously integrated with

‘natural’ vegetation, and African farming, land and resource-use practices degrade or are at best benign to the original vegetation. Degradation is thus prevented only by functional social organisation (regulation and authority), so that from environmental degradation one can diagnose the social ills of organisational dysfunction. Linked to this is the assumption that African

society iS essentially sedentary and subsistence-oriented with an anticommercial sentiment; money, mobility and trade are modern and lead to socio-environmental dysfunction. Thus African history consisted of the continuous reproduction of tradition until it began to become ‘modern’, whether with markets and mobility, colonial intervention, or (in some work) the arrival of Islam. A related stereotype is that African rural populations only increase, and do so fast. Population increase is as such environmentally and socially damaging. The romantic links contained within these assumptions mean that assumed

vegetation change carries very profound moral messages. ‘Original forest vegetation’ and ‘traditional functional society’ provide fundamental baselines, so that whether the concern is about society or the environment, it is possible to judge that something is wrong and assess the extent of damage. From such a vantage point, the imperative is to intervene. The institutional and financial structures in which social science is applied to environmental problems in Guinée strongly encourage the re-elaboration

of such narratives. Donor agencies and projects have commissioned socioeconomic studies to help them tackle more appropriately and participatorily the environmental problems on which their institutional survival depends. Very often, the environmental reading is built into the terms of reference of consultants who have neither the time nor the social position to investigate village natural resource management and its changes on any other terms.? This problem is not necessarily solved when consultants are Guinean, even when working in their home areas; indeed it can be compounded by the urban bias and politico-intellectual images which such local consultants bring to bear. In this context, it has been surprising to us how little the personal lifetime experiences of local consultants and development workers influence the way

that Kissidougou’s environment has come to be perceived. This may be because personal environmental histories have too limited a spatial coverage to challenge a generality, or because unbroken personal histories are themselves rare: state officials are frequently transferred and are in preference posted to areas with which they are unfamiliar, so they have frequently been away from

their childhood village environments for long periods. Such people almost

274 Misreading the African landscape

invariably justify their perceptions of historical deforestation and its social causes with examples drawn from major roadsides and urban peripheries, with which they have more continual familiarity but which in Kissidougou are the proverbial exceptions to the rule.

Forest loss unchallenged

Perhaps the most surprising feature of the external narratives about Kissidougou’s environment is their century-long persistence despite the varied, but highly contrasting, opinions of the prefecture’s rural population. Problems

of effective communication at the interface between development organisations and those they engage with have been examined in many contexts (e.g. Long 1989, Long and Long 1992). In the Kissidougou case, the interface rooted in the colonial encounter and developed over a long period and in antagonistic circumstances, has rendered the expression of local environmental experiences highly problematic.

Faced by direct questions about deforestation and environmental change couched in the environmental services’ terms, villagers tend to confirm what they know the questioners expect to hear. Villagers have, over decades, become familiar with the principal elements of outsiders’ analyses, whether during extension meetings or in primary or later education. Agreeing with (or not denying) visitors’ views, just like offering largesse, is a way of coping with

what are generally extractive or repressive encounters. Furthermore, agreement can reflect not only fear or politeness, but also the desire to maintain good relations with authoritative outsiders who may bring as yet unknown benefits; a school, road or advantageous recognition to the village, for example. Like the prefecture administration, many village authorities realise the benefits which can accompany community participation in environmental rehabilitation, and in this context may publicly agree to the ‘urgent need’ to plant trees, establish

village environmental management committees and so on. Nevertheless, acceptance is not without anxiety, and as chapter 8 showed, everyday forms of resistance frequently underlie overt participation. The social and political contexts of such discussions thus inflect villagers’ representations of past

ecology, just as they do in the different political context of oral histories concerning settlement foundation, where images of initial vacancy (high forest, empty savanna, or abundant wild animals) may justify the firstcomer Status of current residents (cf. Dupré 1991, Hill 1984). Encounters with state officials, development agencies and other visitors silence much local ecological expression through not only their social context,

but also their conceptual context. Local ecological expression can be incomprehensible to outsiders, phrased as it is in concepts relatively incommensurable with those of western environmental science. It is the concepts of

Sustaining reversed histories 275 the forestry service which dominate development encounters. The importance

to village ecology and expression of old village sites, soil opening and ripening, or of gina spirits in influencing water flows, for example, may be entirely missed by outsiders unfamiliar with these idioms, or written off as

religious, not ecological thought (cf. Fairhead 1992). Such conceptual incommensurability is itself couched within what one might call root incommensurability: between western ideas of human degradation or conservation of nature, and local ones of living and working with ecological processes and their inseparability from social ones. In this respect, it is worth noting that it took at least six months of village living before we — and our Guinean colleagues — began to apprehend villagers’ conceptual frameworks and their significance in explaining their experiences of vegetation change. While these factors condition communication at the interface, the lack of challenge should not be reduced to this. Even where forest agents on the ground might acquire some clarity in considering local enviromental perspectives, and

come to have doubts about their messages and the fines which they impose,

they have to date had almost no capacity to influence their institutions’ environmental agendas. How can they be confident enough in what villagers say to question both their institutions’ logic and the scientific canon as they understand it? Beyond what villagers say and what they have seen, what is the evidence for challenge? Who would listen to them? How would expressing this square with their training documents and job instructions? What effect would

it have anyway at a higher level? How would adopting this perspective institutionally alter their future employment and status? Prior to the 1950s, and within the context of a racialist colonial ministry, questioning French seniors’ orthodoxy was certainly not the job of local-level

Guinean administrators. When in the 1950s, the environmental services became a para-military operation, its structures became even less tolerant of critique from the ranks. Post independence, science was the tool of the revolution, and ‘mystification’ its enemy. Again, this was no context for any individual’s populism, least of all given the surveillance and repression of potentially critical individual expression during the Sékou Touré regime (e.g. Bah 1990). Despite an appeal to valorise ‘popular’ African medicine and veterinary medicine, simultaneously valorising the imaged ‘authentic’ African culture underlying the revolution (Touré 1976), this was limited to aspects such as leaf chemicals which could be easily assimilated within the concepts and

therapeutic repertoire of dominant science (cf. Thrupp 1989). The less assimilable aspects carried dangerous overtones of ‘mystification’. Strikingly, embracing even these aspects of African ‘popular culture’ in medical terms did not extend to agriculture and forestry, which were deemed almost wholly backward and in need of radical modernisation. Since 1984, and especially in the 1990s, attempts are being made to render

276 Misreading the African landscape the environmental services more responsive to local knowledge and priorities in acknowledgement of some of these problems (cf. Bah 1989). But the shifts in organisational structures, approach and culture argued elsewhere as central for agencies to engage effectively with farmers’ perspectives (cf. Chambers et al. 1989, Scoones and Thompson 1994), while in some respects outlined and planned in Guinée, have yet to be effected. Meanwhile, exceptional individuals remain frustrated. One dedicated Kissidougou forestry worker who did become interested in farmers’ opinions, and who sought to work and experiment with

farmers’ manipulation of soils and fallows, subsequently found himself blocked by higher levels of his organisation which gave him no space or encouragement to pursue this interest. In this context, there are also questions to be asked about the relevance to environment and development services of a more accurate understanding of environmental change. Given present institutional culture and the terms and conditions of their work, it is not surprising that there is a certain despondency

and confusion of objectives among state service employees. Perhaps, as villagers frequently suggest, forest agents are not really interested in the environment: ‘they only need money, and have no concern for the protection

of forests’, as one said (cited in Millimouno 1993). And perhaps not Surprisingly, it is the material benefits of service or project employment which dominate employees’ work objectives, and there is fear of jeopardising these by questioning institutional aims and strategy.

Fundamentally, then, the precepts basic to local science which challenge conventional savannisation wisdom are not easily apprehended by agencies, visitors and researchers ill-disposed either to listen or to understand. But even within the conceptual framework of western science, challenge to the dominant analysis in Guinée is rare. This is partly because the scientific information and ecological theory which question the derived savanna sub-climax model — the debates addressed in chapter 1, for instance — are dispersed among different

disciplines and their specialist academic journals. So too are the detailed scientific studies of soil, water, plant and animal relations which, while couched in different conceptual terms, do often support the farmers’ explanations that we have heard, as shown in chapter 6. These sources have remained largely inaccessible to Guinée’s academics and policy-makers. Furthermore, it would seem that information from any discipline alone (e.g. botany, hydrology, soil science, demography or climate history) is inadequate to shift thinking in a sufficiently fundamental way. Lack of interdisciplinary criticism, in this respect, is a support to the consistency of views. In any case,

few of the findings or discussions of modern ecological science enter the information bulletins of multinational organisations, NGOs, development journals and the media; the sources on which most development personnel rely for environmental science information.

Sustaining reversed histories 277 While we are not the best placed to consider how this research has been received, it may nevertheless be instructive briefly to consider it in this context.

Carried out under the auspices of Guinée’s national scientific research directorate,* our research programme was nevertheless elaborated in collaboration with both the national forestry service and Kissidougou’s prefectoral integrated development project, DERIK. It was intended that results would also be relevant to the Niger River protection projects. Given the unexpected nature of the research findings, and the challenges they pose to present policy and institutional ‘raison d’étre’, responses have been mixed. At the national level, when our results have been presented in

development fora and formal seminars, they have been appreciated for exemplifying the importance of proper consultation of local populations — pedagogically useful in the current reorientation towards more participatory approaches. In this respect, subsequent phases of the river protection projects, elaborated since this research, have incorporated a more critical historical analysis of environmental problems, providing a better context for realistic ‘participatory’ approaches.

Our collaborators in Kissidougou were, however, highly sceptical of our preliminary results concerning vegetation change. It took the strongly visual evidence of air photographs and their own verificatory studies to begin their reconsideration of environmental problematics. Eventually, this did bring about a basic shift in understanding of environmental change: few would now uphold the idea of recent and rapid loss of forest area in the prefecture, and the assorted reasons supposed for it. This has provided a basis for more realistic consideration of the social and technical aspects of local agro-ecological management; discussions we initiated during a number of seminars and round tables, and the two-day prefectoral environment conference. Nevertheless, there are many institutional, individual and intellectual reasons why people in the prefecture will be unwilling to reconsider their views. In this respect the

research has perhaps served more to open up local debate than to settle the question; a debate which is itself valuable, but which clearly concerns much more than empirical realities about environmental change and people’s management of it.

Conclusions Through wide ranging overview, this chapter has tried to convey how intellectual, social, political and financial structures have all played a part in creating

and sustaining the vision of environmental degradation in Kissidougou. Knowledge about and convictions of deforestation have been produced, not only through particular methodologies but also within diffuse institutions of political and economic power. While technicists might see the problem lying

278 Misreading the African landscape in “bad science’, and its solution in ‘good science’ and training, we have argued that there are much broader and more intransigent reasons why the degradation

view has made sense, which impinge on — or condition — any scientific

endeavour. In this sense, one can speak of a Foucauldian discourse of degradation.

In as much as the degradation conviction can be traced to convergences in the views expressed by different actors, albeit for different reasons, in the contexts that matter to it, it has had an impregnable, totalising capacity. Images

of environmental change invoked in ethnic discourse, for example, have converged with those in discussions of agricultural modernisation, social change and problems of modernity, and of financing the prefecture’s development. Thus the degradation vision has evolved over time and in ways which mean that today it cannot be attributed only to scientists, donor agencies and their narratives. It is partly the product of a long history of interaction with and incorporation into local social and political processes, and is thus today partly

sustained by them. This is not to say that villagers’ everyday ecological

practice and experiential readings of landscape are influenced by the deforestation reasoning, but merely that their ecological reasoning is subjugated in much political interaction.

10 Towards a new forest—savanna ecology and history

Not all ecologists, foresters and botanists would interpret West Africa’s forest-savanna mosaic in terms of past and ongoing forest loss. But over the past century, all those who have actually examined the mosaic in Kissidougou have interpreted it in this way. Close examination of the historical record and of local management practices shows not only the error in this perspective, but also what it has obscured: the creation of forest islands, their dynamics, and the enriching of open savannas with more woody vegetation forms. It is within this dynamic that Kissi- and Kuranko-speaking villagers conceive of their relationship with their landscape; a relationship with deep historical roots. Considering the landscape in terms of degradation has obscured how people live and work with ecological processes on a day to day basis, often improving land productivity and value — improvements recognised in tenurial claims. For

Kissia and Kuranko, it is in part through using land and bringing it into productivity that both common and differentiated social identities and relations are realised. And in reflecting on ecology people also reflect on their relations

not only with the world around them, but also with each other. While their conceptions of ecological dynamics often contrast strongly with scientific orthodoxy, they do nevertheless postulate relationships which isolated and recent scientific studies would support. It is largely these local land use and enrichment practices, and their articulation with major historical shifts in economy and polity which account for the particular course of vegetation change during this century.

None of this has been in the least influential in the elaboration of Kissidougou and Guinée’s environmental policy which has for a hundred years been based on a backwards reading of landscape history. Not only has this had

the instrumental effect of alienating resource control and income from villagers, but it has also provoked environmental problems, often mitigated only by the resistance it has met. The origins and endurance of such landscape misreading have depended, in part, on the relations of production of scientific knowledge within the powerful economic and institutional structures which apply it. But the long history of interaction of these institutions with local discourse has at times also lent support to policy-makers’ convictions. 279

280 Misreading the African landscape In this last chapter, we address three broad issues emerging from what has, for us, been a rather surprising study. First, we consider whether there are other ways within ecological science of conceptualising forest-savanna dynamics which better account for the changes revealed here, and which open up different interpretative possibilities concerning the role of local land use in vegetation change. Second, we consider why the case in Kissidougou cannot be dismissed as a unique and isolated exception either in African forest-savanna dynamics, Or in mismatches between received wisdom and local experience concerning African environmental change. Third, and in this light, we reflect on broader relationships between landscape readings, policy and power. Re-placing people in forest—-savanna ecology

The discursive processes sustaining visions of degradation in Kissidougou, while not reducible to them, have nevertheless rested heavily on the conceptual frameworks of ecological science. As we noted in the introduction, however,

ecological theory has recently seen some major reorientations, notably in theorising issues of disequilibrium in ecological systems, associated more generally with a greater appreciation of plural approaches. Here we consider how non-equilibrium positions might be applied to West Africa’s forest-savanna dynamics, arguing that they provide an alternative

model for considering the region’s ecology, and thus for considering vegetation change in Kissidougou. While in itself this attempt is revisionist, remaining within the discursive terms of ecological science and thus locked into cartesian reasoning about nature and society, it does provide a first step in opening up conceptual space for a more radical incorporation of inhabitants’ own perspectives on people—environment relations.

Revising forest-savanna ecology

It can be argued that the most fundamental props to scientists’ and policymakers’ readings of the forest-savanna mosaic derive from the imperative to explain the disappearance of a ‘natural’ vegetation, imaged as a climatic climax

at equilibrium. The image of a climax contained the image of ‘bigger’ and ‘better’ vegetation — in this case, dense semi-deciduous forest. It allowed all vegetation change to be imaged as a lineal divergence from an undisturbed, original form. And it enabled people’s impact to be logically imaged as external disturbance which modifies or destroys ‘natural’ vegetation. The landscape thus seen appeared as a set of anthropogenic sub-climaxes, containing within them a vision of destruction which hid local enrichment practices and their significance.

Certain recent strands of ecological thinking, however, strongly criticise

Towards a new forest—savanna ecology and history 281 such notions of original vegetation, and in so doing, we would argue, enable a re-placement of people’s role in vegetation change. These strands abandon notions of a single, natural climax. This has become necessary, at a first level, with the increasing recognition that natural disturbances are a feature of most environments. Whether originating in animal, wind or fire effects, they keep the ecosystem from ever reaching a stable equilibrium in any place. Instead, ecosystems come to be imaged as a patchwork or mosaic of ages of recovery from localised disturbances; in forest ecology, a mosaic often portrayed as a set of different stages in cycles of regeneration (cf. Whitmore 1990).

While this revised vision rejects the notion of equilibrium at any single place, it can be seen to displace it from place to the broader level of a landscape, which could reach a dynamic equilibrium through a balance of disturbed and

regenerating patches (cf. Sprugel 1991). However, ecologists now consider even such arguments for stability at aggregate landscape level as potentially flawed, for several reasons. First, climatic changes can interrupt the cycle of ‘return’ and hence the movement to equilibrium. When climate historians suggest that West Africa has experienced both long-period, deep climatic fluctuations and changes in climatic variability, the history of vegetation form begins to have to be seen as a history of continual transition, rather than as one of divergence from a single, once extant climax. Second, one-off events and very large disturbances may have long-lasting impacts, whose effects on the

land or species composition are felt even through subsequent disturbances. Third, there may be long-term trends (such as reduced fire) which lead to major variations in the proportion of the landscape at different developmental stages, perhaps of the order of a decade or a century. If that occurs, then populations of species that depend on different developmental stages — or processes that are accelerated or diminished there — will not be even approximately constant on

a landscape scale (Sprugel 1991: 4); in other words, proportionality has qualitative effects.

Within these potential instabilities, and given that the legacy of one vegetation state can enduringly influence subsequent ones, pathways of vegetation development can therefore be unique and chancy. The cession of savanna to forest thicket vegetation in southern Kissidougou, for example, may partly reflect lag effects of the climatic rehumidification between the mid nineteenth and twentieth centuries (mediated perhaps by lags in the diffusion of particular termites or fungal mycorrhiza). Equally it may reflect the increase in cattle numbers in the 1950s, and their sudden decline at a time when fire control became more effective; the greater availability of forest species as a result of reduced fire further south, or a unique sequential blend of these. Because vegetation change might need to be understood in terms of such historically contingent transitions, it is inappropriate to evaluate vegetation in terms of the ‘climactic potential’ at any particular moment, and certainly

282 Misreading the African landscape

incorrect to assume that, because a particular vegetation form (e.g. forest) could exist under present environmental conditions, that it did once exist. Recent ecological theory suggests how pathways of vegetation development might best be considered as transitions between particular stabilised vegetation states, each determined by a multi-factor complex. In this, the impact of change in any particular factor will be to induce less a smooth trend than a shift from one state to another. Furthermore if the transition-causing factor reverts to its pre-transition level, the vegetation need not return to its initial state and may even move to a third (Sprugel 1991, Behnke and Scoones 1991, Dublin et al.

1990, Holling 1973, Noy-Meir, 1982, Westoby et al. 1989). Given the multiplicity of interacting factors influencing each state, as well as the importance of each state’s legacy on future possibilities, shifts between them can be ‘chaotic’ in nature, and especially likely to respond to particular, possibly unique, historical conjunctures of ecological factors. While highly diversified in themselves, forest and savanna forms can — from an ecological viewpoint — be seen as such multiply determined and strongly differentiated ‘states’, especially in terms of grass exclusion and hence the

exclusion of fire. Where, as in the transition zone, climatic conditions are relatively marginal for forest, the presence of forest forms might depend on a constellation of factors including fertility cycling, soil structure, water holding and mycorrhiza, micro-climate, the absence of fire, and germination potential interacting with the water made available by climate. Furthermore in any place, a shift from a forest to savanna form or vice versa is especially likely to be relatively enduring both because of internal stability effects encouraging the persistence of each form (Moss 1982) — for example the tendency for savanna to perpetuate itself by fuelling fire, or forest to suppress fire — and because the legacy of any transition tends to be enduring, inscribed in soil structure, fauna and texture, and hence in edaphic qualities.

Forest and savanna vegetation forms evidently offer very different production and gathering possibilities, as earlier chapters have shown. While people use and value a diversity of vegetation states, there are frequently major agricultural productivity gains to be had by deflecting savanna regeneration to forest vegetation regeneration in fallows, as well as protective, plant product and other advantages in having forest around one’s settlement. Thus where conditions are marginal for forest, leaving a precarious balance between forest

regeneration and pyrogenic savanna, the Kissidougou case suggests that people’s manipulation of these processes can tip the balance. It suggests that villagers may actively direct such transition ecological processes to promote a greater proportion of forest vegetation in the landscape when they have reason to do so. By altering the balance of interacting factors, people can initiate a shift between vegetation states, perhaps even deflecting one state to another which

Towards a new forest—savanna ecology and history 283 might otherwise be unattainable (or much less likely) through sequential transitions involving only ‘natural’ ecological processes. For example, by bringing about the simultaneous need for improved soils and soil fauna, fertility boost, seed source, and fire exclusion, people may facilitate the establishment of forest vegetation in savannas which (for lack of the necessary rare conjuncture) might otherwise never make this transition. Theoretically at

least, people can therefore increase the range at which any particular stable state (e.g. semi-deciduous forest) might be found with respect to any particular variable (e.g. rainfall) by providing the ‘transition bridge’ for its establishment, effectively increasing the plasticity of that state with respect to that factor. And

one transition having been precipitated, it provides (in time perhaps, and as other ecological factors respond to it) the springboard for others. It seems to be precisely this that we saw happening in chapter 6, as people worked with ecological processes to ratchet (channel) transitions to their advantage. A major implication of this perspective is that the limits of any particular vegetation state (e.g. dense semi-deciduous forest) with respect to any variable (in particular climate) are more likely to be set by human management, rather than that variable per se. Put another way, in managed environments, people’s

activities may be more important than climate in determining vegetation extremes. From this perspective, then, forest patches in savanna reflect the results of particular transition pathways for which people’s mediation has been essential, rather than existing either as natural relics or as responses to stable ecological conditions.

Socialised ecology and social history

To date, ecological science has overlooked these human influences, considering the role of inhabitants only as destructive of forest. The processes which do assist forest advance it treats as ‘natural’, imaging them universally

as ‘despite’ human activities. This is a problem of recent non-equilibrium theories as much as of older climax reasoning, and is partly rooted in the fundamental dualism in western thought which conceptually separates people from ‘nature’. This tendency to conceive of society and nature as separate, the former ‘acting on’ the latter or the two related in an alternation of domination

and subordination, can be located as a product of the particular history of western thought. It is a product which carries particular moral implications, not least in allowing all aspects of nature to be considered autonomously of human

relations, and thus be freely commoditised (e.g. Marx 1867, Taussig 1980, Taylor 1992). Cartesian ecology is, in this way, in keeping with commodity markets. What goes on on the land has only very indirect bearing on social and human health. A fundamental prerequisite for comprehending Kissi and Kuranko debates

284 Misreading the African landscape about vegetation change and management has been to consider how ecology is

socialised in very different ways. As we have seen, local analysis seems to draw distinctions along very different lines which confound the society—nature

divide. It encompasses many analogies, for instance in likening the effect of people to the effect of termites on the landscape, not drawing a fundamental distinction between these when examining their ecological effects. Villagers consider many of the key ecological processes in their landscape as settlement—

social ones, whether in the concentration of woody vegetation where they live and work, or where the settled societies of termites or gina do. Indeed distinctions are less between human and natural processes than, in many cases,

between settled and vagrant ways; and emphasis less on distinguishing between the realms of human health and prosperity and those of the land, and more on cycles of birth, growth, maturity and death common to plants, animals, soils and people. In this, there is no a priori distinction between phenomena

which western thought treats as social and natural; in as much as such a distinction is made (e.g. between village and bush) it is a moral and fallible one, in which land and people’s cycles can interfere with each other and must therefore be actively maintained as separate. From within this conceptual focus On processes common to people and aspects of the world around them, there is scope for their intercausality. Wrongly placed or inappropriately timed sexual activity can disrupt plant, animal or climatic cycles as well as the human ones within which they are integrated in a socialised ecological order. This socialised ecological order should not be confounded with a ‘naive

holism’ which may see people’s thoughts and actions as harmoniously integrated with ‘the environment’. Rather, power and struggle are part of it: as the analyses of Sandaya and Toly showed, ecological processes are considered within a structured conceptual ordering which is also integral to the creation and maintenance of power relations within rural communities. Thus a further

problem of considering Kissi and Kuranko ecology within the western society—-ecology dualism is its obfuscation both of the ways that ecology is locally politicised, and of how people’s experiences and work with ecology

shape sociality itself. It is not just people’s livelihoods but also their identities which are linked with agro-ecological practices. Social and political relationships can be constructed through ecological relationships: whether in sharing a forest island, or maturing soils for a descent group. Social and political identities and boundaries can be constructed in ecological terms, as in the modern ethnicisation of forest or savanna practices. Thus there are elements of agro-ecology in local constructions of social boundaries and transitions, and elements of sociality in local interpretations of agro-ecological boundaries. In this respect the social and historical meanings of landscape features can be

invoked in struggles over social and political status (e.g. descent-group origins), or over resources and their control; for example over timber trees, or

Towards a new forest—savanna ecology and history 285 ‘improved’ soils. It is the extent to which representations of landscape are embedded in — the product of — social and political relations that makes it necessary to speak of local discourses and terms of debate, rather than of a homogeneous local ‘ecocosmology’, when considering socialised ecology and representations of landscape. Earlier chapters have considered how the event histories of people’s land use respond to changing economic and political conditions. We have tried to show how the social history of vegetation management has articulated with ecologyas-history to shape vegetation trajectories. People’s vegetation management cannot be described as ‘traditional’ in the sense that their preferences are conditioned by political and economic history. In as much as the balance between (and quality of) vegetation forms within the mosaic reflects the local values attributed to them, its evolution and hence that of the landscape must be understood in relation to these historical changes, and the cultural processes and social negotiations which they engender. In this respect, gender issues have been central in understanding the making of the forest-savanna landscape. This book has not been about gender; nor does it explicitly address the implications of ecological change in Kissidougou for gendered power and resource control — this we do elsewhere (Leach and Fairhead

et al. 1992; Leach and Fairhead 1994b). However gender has been a central axis of social difference in the village-level analyses of changing farming and vegetation use patterns, with this study thus joining others which show that gender relations shape patterns of environmental management with tangible ecological effects (cf. Joekes et al. 1995, Leach 1994, Leach and Green 1995). Furthermore we have shown how ecology itself is socialised in gendered ways — for instance through people’s understandings of human reproduction. The major changes in vegetation management in Kissidougou have often responded to historical conjunctures of factors. Shifts in the balance of farming practices with different vegetational effects, for example, have reflected how changes in migration, farm-household structure, marketing conditions and gender relations have come together at different times and in different places. Once such a conjunctural history is appreciated, unilineal arguments about people’s ‘effects on the environment’ — including arguments concerning social

breakdown and demographic change — appear theoretically inadequate and misleading. Concerning ‘social breakdown’ arguments, for example, this study suggests that environmental management in this region depends — and has always depended — less on community-level authorities and socio-cultural organisations (which might be ‘threatened’ by social change), than on the sum of a much more diffuse set of relations; a constellation more than a structure. The maintenance of long-term productivity is in many cases built into short-term production patterns; whether carried out for oneself, one’s household or one’s

286 Misreading the African landscape compound. And these improvements frequently interact with others — spatially or temporally — so that the combined effect on resource enrichment is greater than the sum of their parts. Thus, the fires set in the early and mid-dry season by hunters to clear small hunting grounds, and by others to protect property and fallows, create barriers to more devastating later fires; and the small tree crop plantations which people make and protect behind their kitchen gardens add to the creation of the village forest island. For much resource management there

is no need for village or higher-level management structures to ‘regulate degrading pressures’; the root assumption in both ‘tragedy of the commons’

and ‘common property resource management’ debates. When village authorities do intervene in certain vegetation-influencing activities, it is generally to orchestrate activities which are already on people’s agenda, for

example in early-burning around the village, in protecting palm trees, in imposing cattle-tethering dates, in coordinating the fallow rotations of farmers’ contiguous plots in some Kissi areas, or negotiating with external representatives of the forest service. This is not to deny social and economic change or emergent social and economic problems, but these changes are rendered visible in the landscape largely through changing land-use and management priorities, not through organisational ‘breakdown’ of ‘good management’, and vegetation degradation. Re-placing people in a less unilineal view of ecological change as has been necessary in Kissidougou certainly contradicts the simplistic “population growth—deforestation’ relationships upheld by the policy narrative. Indeed, as chapter 7 argued, from an earlier situation of probable greater savanna extent, there seems to have been a broadly positive relationship between the peopling of this region and its forest cover. But this cannot be taken as an endlessly continuing or indeed necessary relationship. On one hand, population levels and distribution themselves need to be seen in relation to more conjunctural events, as well as trends: to warfare for instance, and to migration related to economic and political conditions. On the other hand, in Kissidougou it is the way that particular demographic conditions have articulated with particular economic and social ones that has made possible and desirable particular land management options. Historical perspectives have strongly challenged assumptions made about

Kissidougou’s forest-savanna mosaic. Indeed this study has argued for a perspective which treats ecology more as history; for understanding present

vegetation in terms of unique pathways conditioned by social as well as ecological change. Such a perspective has recently been emphasised in analyses of forest history in areas as diverse as temperate North America (Sprugel 1991), South-east Asia (McNeely 1994) and Latin America (e.g. Posey and Balée 1989, Rival 1995). While studies of African ecology in the drier zones have moved towards this perspective (e.g. Dublin et al. 1990,

Towards a new forest-savanna ecology and history 287 Scoones 1995) it has so far been little applied to Africa’s humid forests or forest—savanna transition zone.

Old ideas of environmental optima once dovetailed neatly with ideas of static social optima: of tradition and functional structure typical of, but persisting beyond, colonial anthropology and earlier theories in cultural ecology. But notions of society with a given social structure and order, maintained by functional adaptation and/or by rules and regulation, are challenged by more recent social theories giving weight to social action and processes, and their capacity to shape and determine rules. Such continual structuration, over time and through social change, challenges the notion of a baseline societal state. A view of African social forms as in continual transition better dovetails

with a view of ecology as constantly changing, and of landscapes as continually made and re-made.

To recap, as an alternative to climax reasoning, non-equilibrium ecology offers better possibilities for interpreting forest-savanna ecology within its real historical specificity. By removing the strictures of a concept of natural climax vegetation, it also opens up scope for better considering people’s impact on vegetation. We have tried to suggest here how an understanding of people’s landscape relations can be fruitfully integrated with non-equilibrium ecology

in ways which better account for the actual, linked vegetational and social dynamics of forest—-savanna transition ecology and history. While this research has not ‘tested’ this theory in ecological terms — and indeed in keeping with arguments concerning ecological pluralism (McIntosh 1987), we would not seek to establish it as any new orthodoxy — we can argue that this perspective provides a valuable counterpoint to climax views. In Kissidougou, moreover, it provides a framework for re-reading the landscape which potentially rescues local landscape readings from the reversed histories that have silenced them.

Beyond Kissidougou This study has concentrated on a small portion of West Africa’s forest-savanna transition zone. Two pressing questions thus need to be considered. First, are

these findings a unique and isolated exception in African forest-savanna dynamics? Second, is Kissidougou (or Guinée) special in this astonishing mismatch between received wisdom and local experience concerning environmental change, or might it exemplify an issue of wider significance in Africa and beyond?

Within the wider forest-savanna transition zone, there is certainly no shortage of modern studies which suggest past and ongoing savannisation of forest. These are generally accepted unproblematically, as indeed they were in Kissidougou prior to the present research. There are, however, no reasons for Supposing that the processes which have produced Kissidougou’s landscape

288 Misreading the African landscape misreadings are not equally operative elsewhere in West Africa’s transition zone. Indeed, it would be surprising if this were not the case. Certainly, very few of these studies are founded on historical data, and those which are have been uncritical in their methods; an issue taken up by Blanc-Pamard and Peltre

(1987), for example, in disputing Gayibor’s (1986) historical analysis of derived savanna in Benin. Certainly, other countries have a similar legacy of colonial science in shaping how environments came to be perceived — indeed the same scientists were involved. Furthermore, landscape readings elsewhere

respond to institutional and global perceptions of deforestation and environmental degradation in a comparable way to those in Guinée. The Kissidougou case underlines the urgent need to interrogate the validity of studies which suggest recent savannisation. Both historical data and a critical examination of the context in which past studies had been conducted will be essential in this. In short, degradation visions may have concealed forest-savanna histories comparable to Kissidougou’s elsewhere. This materially affects many millions of people. The imperative for a more critical examination of forest-savanna dynamics

and people’s role in them is further underlined by a number of studies in neighbouring countries, all of which exemplify how people have encouraged forest and palm forest development in savanna. Notably, none of these studies

has emphasised their wider implications for forest-savanna dynamics and history, and all have remained marginal to broader theoretical debates in this area.

In farming, moving east along the transition into Cote d’Ivoire, several authors have shown processes of forest advance into savanna despite the presence of fire (e.g. Avenard et al. 1974), and highlight the importance of cultivation in this. Blanc-Pamard and others show how ‘In certain forest— savanna borders in the south of the Baoulé V, people are an inducing factor of reforestation who accelerate by their cultivation the process of shifting the equilibrium, already unstable, in favour of forest’ (1979: 68). Here, farmers explicitly state that ‘where one cultivates, the forest advances’ (Spichiger and Blanc-Pamard 1973: 199; cf. also Adjanohoun 1964: 131). While the authors conclude that ‘a population increase in the Baoulé V would increase cultivated areas in forest, forest margin and savanna, strongly accelerating the process of reforestation by secondary forest thicket’ (1973: 207 — cf. Adjanohoun 1964),

their caution led them to delimit these conclusions strictly to their study area. But are such limitations valid? Oral history in the forest zone east and south-

east of the Baoulé V suggests that the region was savanna in historical times (Ekanza 1981). So perplexed was the historian in finding this, that he felt the need to deny the veracity of his own evidence. West of the Baoulé V, again in

what is now the forest zone, several Gouro villages carry names such as

Towards a new forest-savanna ecology and history 289 Koumodji, meaning ‘forest island in savanna’, or Deragon, meaning ‘open savanna’ (Deluz 1970).

East of Céte d’Ivoire, in Ghana, Amanor analyses historical conditions which have led to reduced forest cover in the forest-savanna boundary zone. Yet he also shows how Krobo-speaking farmers there have developed innovative regenerative technologies, albeit ignored, overlooked and rendered invisible in policy. Certain of these halt the spread of savanna, and conserve tree stocks on farms, but more significantly, others reconvert savanna lands

back into secondary forest thicket by excluding fire during extended cultivation periods. As Amanor puts it, farmers do this by a ‘harnessing of the regenerative energies contained in forest succession and adaptation of the agroecosystem to harmonise with forest regenerative principles’ (Amanor 1994a:

63; cf. Amanor 1994b). Frequently farmers introduce a preferred tree, Newbouldia laevis. The wider implications of Amanor’s findings are, however,

suppressed by two features of the study. First, these farming practices are imaged as ‘innovatory’, hence recent and not of long-term significance to the forest-savanna transition. Second, as the study is couched within consider-

ations of rehabilitation of recently-derived savanna, it implies that these phenomena might be specific only to such ecological situations. Thus the interpretation closes off inquiry about the possible use of similar practices in the more distant past, for example in establishing the palm forests known to have existed in these regions in the mid-nineteenth century, sometimes derived from savanna.! That Ghana’s forests have been extending into savannas was noted as early as 1937 (Vigne 1937). Intriguingly, oral accounts from Cote d’Ivoire suggest

that the region in the heart of Ghana’s forest zone (the Denkyira town of Apibweon) once had ‘no trees; there was nothing except short grass’ (Perrot 1974: 199).

In parts of Togo’s forest-savanna mosaic, where farming has classically been carried out in forest fallow fields, recent demographic pressures have resulted in the expansion of shifting cultivation into the savanna area. In some cases trees are encouraged in these fallows, and forest crops are then established. Successional stages in the establishment of forest depend on disturbance or ‘accidental intervention by farmers’, who, by initial cultivation,

encourage the germination of pioneer, light-demanding forest species (e.g.

Harungana madagascariensis), and then protect the developing forest vegetation (Guelly et al. 1993).

In the course of our own research we briefly visited parts of north-west Sierra Leone,2 where Susu farmers outlined the techniques they invest in to upgrade savannas to forest thicket fallows. They use a combination of inten-

sive grazing and organic matter incorporation, stressing the role of termite activity in soil maturation processes, suggesting that it 1s in part their soil

290 Misreading the African landscape improvements which enable secondary forest thicket to form. In Sierra Leone’s

Kono region, oral accounts collected at the turn of this century by Willans asserted that at the time of settlement the country was ‘all grass, not bush at all’ insisting that it was the people that brought the bush (1906: 141). We also visited Sankaran farmers in Guinée’s Faranah prefecture, a slightly

drier part of the transition zone just north-west of Kissidougou and north of Kono. They use similar methods to direct succession to establish a more dense dry forest fallow dominated by Uapaca somon (a tree hugely multiplied by root suckers on cultivation); a formation locally termed samo tu (somo forest). In forest-savanna transition zones elsewhere in the world — for example in Latin America — similar examples of farming practices encouraging forest

advance into savanna can be found. In parts of Amazonia, for instance, farmers ‘plant’ termite mounds and otherwise manipulate soils to initiate forest patches (Anderson and Posey 1989). The importance of settlement sites, past and present, to the development of forest vegetation is also suggested elsewhere, albeit as offhand remarks in a

dispersed and largely obscure literature. For example, Aubréville himself describes the transformation of savanna to forest in association with settlement, but obscures its significance by considering it an isolated exception: We observed a very curious case, incontestably the installation of forest by people in poorly wooded savanna of Hymenocardia acida. Batéké country (Middle Congo), near Okoyo and Evo is scattered with islands of dense forest; at a glance one can see 10, 20 in the landscape. Each one marks the site of an old village. The Batéké always install their villages in savanna and cultivate their manioc only in savanna; in the rare vestiges of ancient forest which remain, they only plant a few yams, and some sugar cane; only the women cultivate, and they do not want to enter the forest. In the villages, to make their houses and fences the inhabitants use poles of iron-wood, Milletia Laurentii . . . and Ficus sp. These trees propagate by cuttings very easily, the poles take root, oil palms spontaneously establish around houses, other species disseminated by seed subsequently establish. After a few years this forest vegetation becomes too dense for Batéké taste, who do not like forest, and the village is moved a little further, into clean savanna. The old site transforms in a dozen years into a thick wood, mixed with palms. Whereas in general the populations of the forest zone seek forest for hiding in, living in and for defence, and clearing it for cultivation, the Batéké form an original exception, fleeing the forest, and recreating it. (Aubréville 1949: 318)

In Nigeria, the establishment of forest islands (Kurmi) around inhabited savanna settlements is alluded to by Lamb (1942, cf. also Jones 1963). In Nigeria’s Oyo division, Abimbola (1964) observes that there are many forest patches covering ruined villages in the savanna belt — some up to 2-3 miles in diameter — attributing their existence to soil enrichment following settlement.

Abimbola writes that “Cocoa is being planted in the forest surrounding these ruins. I am, however, convinced that forest is not native to this area. Geographers have a task here to explain whether forest, like savanna cannot be

Towards a new forest—savanna ecology and history 291 “derived”. But derived from what?’ ‘One clever way of detecting ruins in this area’ the author suggests “is by the dense forest that tends to grow on them’ (1964: 18). Keay (1947) reinforces this interpretation: there is no reason to suppose that the vegetation before these villages were built was other than ordinary savanna woodland. The better growth on these sites and the fact that several of the trees are usually found in moist situations, appear to be due solely to human occupation. (1947: 44)

These findings are similar to those of the archaeologist Davies who used clumps of dense forest in otherwise ‘orchard-bush’ to identify old village sites

in the Afram Plains, Mo Plains and Dayi valley of Ghana’s northern forest-savanna transition zone. He suggests that sites more than 300 years old might lose their forests, but retain their soil specificity. On one site, perhaps more than 1,000 years old, ‘there was up to twelve feet of deep rich soil; the area having been farmed with some surface erosion, and there was today no indication of the site from the vegetation’ (Davies 1964: 29). In this same zone, the archaeologist Effah-Gyamfi (1979) describes the higher nutrient content of ruined site soils, and their vegetational specificity.

In Latin America, Balée (1989) shows how the enriched soils of past habitation sites support a distinctive forest vegetation and are preferred for

cultivation. Apart from these isolated instances, the extent to which uninhabited forest outliers in savanna are associated with old habitation, seems — to our knowledge to date — hardly to have been considered.

Such evidence of people’s establishment of forest vegetation in savanna through farming, coupled with the links between settlement, anthropogenic soils and forest establishment, strongly suggests that the formation of forest patches by people may not be unique to Kissidougou. Given that the presence

of forest islands has been used as the most powerful indicator of past deforestation throughout West Africa, if this is the case, it strongly suggests the need to rethink the forest-savanna mosaic landscape throughout the zone.

Kissidougou provides an extraordinary example of the production of erroneous knowledge concerning landscape history and the degradation problematic. While perhaps extreme, it does show that even the apparently most secure analysis and received wisdom can be challengeable, and exemplifies relationships between power and knowledge surely relevant not only in other

parts of the transition zone, but elsewhere in Africa. Powerful received wisdoms concerning environmental change and people’s impact on it operate in many ecological zones: whether concerning desertification and overgrazing by pastoralists; soil erosion induced by growing highland population pressure,

or the recent degradation of ‘pristine’ tropical rainforests. While these processes must surely have their place, several studies have subjected these orthodoxies to more critical historical, social and ecological analysis for

292 Misreading the African landscape particular localities, and have found them wanting (Leach and Mearns 1996). Whether in the drylands of Zimbabwe or the Sahel (e.g. Scoones 1996, Swift 1996), the highlands of Kenya and Ethiopia (Tiffen et al. 1993, Hoben 1995)

or the forest zone of southern Guinea (Fairhead and Leach 1994a, 1995), historical data fail to match the supposed trajectory of landscape change and the images of local practices contained within them. In reappraising history, each of these studies has allowed better appreciation of the logic of local practices, and certain of them have suggested a similar re-examination of the ecological theory which has underpinned the degradation conviction. And in varied ways, these studies reveal the political processes involved in the production of demonstrably false environmental readings.

Policy and power Where the views of landscape which are driving policy are demonstrably false, greater historical precision renders clearly apparent their relations of power and sometimes their brutal material effects. In the Kissidougou case, these have been strikingly apparent. Considering vegetation only in terms of degradation

has, as we have shown, obscured and marginalised many methods which farmers themselves have been using to enrich their landscapes. Ironically, policy-makers and environmental scientists do, of course, consider people able to improve the wooding of savannas, but only through the agency of external

programmes and projects. Seemingly, it is only outsiders’ technologies and organisational impetus — in ‘community woodlots’, agroforestry systems, planting seedlings from nurseries and state-assisted village-level planning — which are adequate to this task. The degradation discourse thus constructs inhabitants as incapable resource stewards, while simultaneously instilling the imperative to intervene and improve the situation on their behalf. Such intervention has had the instrumental effect of robbing valuable resources from

local control, and placing their stewardship in the hands of a technical and managerial elite (cf. Roe 1995). Policies elaborated within the degradation discourse have sometimes under-

mined or compromised existing vegetation use practices, while introducing new social and economic problems. In removing local control over resources, they have sometimes interfered with local management of them. Removal of resource tenure has also reduced villagers’ abilities to profit from past enrichment activities and their incentives for further landscape enrichment. The implementation of repressive environmental policies has in effect taxed rural populations for supposedly harmful activities which were, in fact, benign

or beneficial. Attempts from within the discourse to ‘decentralise’ resource control by establishing village-level organisation and management plans risk

undermining the existing flexible and diverse constellation of resource

Towards a new forest-savanna ecology and history 293 management relations. And the investment in ‘redressing’ Kissidougou’s Supposed environmental degradation, an investment reaching unprecedented levels amid current aid donor concerns, carries heavy opportunity costs in

terms of other more pressing rural development problems now left unaddressed, or half addressed, for want of funds. In this, Kissidougou provides an extreme example, but of processes which

would appear to be much more general. The political economy of ‘truth’ concerning environmental change is intimately linked with a very material political economy concerning who manages whose environment and in whose ‘interests’. Views of change which construct rural inhabitants as incapable resource managers have, as has been argued elsewhere, served to maintain the financial status and institutional continuity of environment and development agencies (Roe 1995); to promote the commoditisation of rural resources in the interests of global capital (Amanor 1994a); to assist governments in the political marginalisation of certain groups and areas (Swift 1996); or to effect the expansion of state bureaucracy into rural areas (Ferguson 1990). The importance of historical perspectives and of non-equilibrium ecological dynamics for understanding Kissidougou’s vegetation raises further difficult questions about the management models underlying many environment and ‘sustainable development’ policies. Ecological managerialist perspectives are founded on views of nature — or of ecological systems — as stable, predictable and manageable. But examples such as this indicate that ecologies may not show equilibrium characteristics, and may be subject to contingent, unpredictable changes, rendering such management objectives an illusory goal. As Amanor has made clear, if the relationship between social and environmental history shows ‘nature’ to be less amenable to domination or integration with stable social forms, many of the assumptions about the society—environment

interface in managerialist models begin to look questionable (cf. Amanor 1994a). Other authors have raised the possibility that rejecting managerialist models may open up a pandora’s box of anarchic degradation: as Worster (1977) warns, for example, seeing environment in terms of disequilibria could

function as a cover for legitimating environmental destruction. But while this possibility cannot be ruled out, the assumption that unregulated use is

destructive, and that regulated use is not, may be unwarranted. The Kissidougou case suggests that under particular ecological conditions, use itself can lead to ‘improvement’; people here have created forest vegetation in conditions where otherwise they would be unlikely to occur. It would also support the argument that environmental ‘abuse’, not sustainability, can be the

result of ill-conceived managerialist control. As Amanor demonstrates in Ghana’s forest-savanna transition zone, ‘if global systems of environmental management are an illusion, present managerial frameworks may exacerbate degradation by adding layers of unwanted political control, bureaucracy and

294 Misreading the African landscape

ideology masquerading as science’ (1994a: 220). Visit Kissidougou to see this. Where history reveals policy to be driven by misreadings of environmental

change, it may also reveal ways of revising policy. The striking reversal of landscape history in the Kissidougou case, indeed, suggests many revisions through which existing technical and organisational policy could be made more appropriate to the region’s ecology and sociality.> The many techniques and land uses that have served Kissidougou’s farmers to enrich their landscapes and increase their forest cover are surely an effective basis for external support — should this issue be demonstrated as a priority. In working with the local

ecology of fire, soils, vegetation successions and animal dynamics, the ‘integrated vegetation management’ practices described in chapter 6 are more

locally appropriate, integrated with the social matrix and thus more costeffective in terms of labour than are the forestry packages generally proposed by outside agencies; packages whose inappropriateness the landscape event history reveals. Given that farming in the region is not inevitably degrading, environmental policy may look to support as well as to rationalise and regulate it, and specifically to support those upland farming practices which improve

soils and fallow vegetation rather than concentrating technical effort exclusively on swamps. Rather than increase external intervention in the organisation of resource management within villages, the more important priority would be to create the enabling policy and socio-economic conditions in which local resource management constellations can act effectively. This

implies a shift on the part of environmental agencies away from direction (through repression or organisational restructuring as in assisted ‘community control’) towards recognising and supporting the diverse institutions which are actually engaged in resource management, and towards a more responsive role in providing requested services at the village level. The history of people’s

‘opportunistic’ responses and uses of non-equilibrium ecology suggests grounds for respecting and supporting these through a policy framework which

enhances, rather than reduces, people’s resource management control and flexibility.

These suggestions are revisionist in as much as they can operate within existing institutional structures; the structures of science, policy, aid, administration and state—village relations. Yet even such revisonist shifts derive from a reversed landscape reading, and this creates a contradiction which makes such a revisionist programme somewhat naive. For as we showed in chapter 9, it is within these institutional structures that the knowledge of degradation is produced; the institutions and their financing are thus part of the discourse and, indeed, this is what makes it a discourse in the Foucauldian sense. In as much as they are mutually constructive, it is hard to conceive of either knowledge or institutional relations ceding without the other. Can more radical alternatives

Towards a new forest-savanna ecology and history 295 be envisaged which rework the power relations between institutions as well as simply what those institutions are doing? Meanwhile, it is hard to underestimate the importance of the degradation discourse’s instrumental effects on many aspects of Kissidougou’s life. These have impoverished people through taxes and fines, reduced people’s ability to benefit from their resources, and diverted funds from more pressing needs. They have accused people of wanton destruction, criminalised many of their everyday activities, denied the technical validity of their ecological knowledge and research into developing it, denied value and credibility to their cultural forms, expressions, and basis of morality, and at times even denied people’s

consciousness and intelligence. The discourse has been instrumental in accentuating a gulf in perpectives between urban and rural; in undermining the credibility of outside experts in villagers’ eyes; in provoking mutual disdain

between villagers and authority, and in imposing on the former images of social malaise and incapacity to respond to modernity. In rewriting people’s history, the degradation discourse has constructed alien ideas of tradition and

community, reinforced ethnic stereotypes and differentiation, and denied people their own history with all its significance for their social and political relations and capacity to live on their own terms.

Appendix | Glossary of plant names

Botanical names of principal trees, shrubs, bushes, palms and herbaceous plant

species found in Kissidougou prefecture, together with usual names in Maninka, Kuranko, Kissi and Lele.

Sources: Field identification with the assistance of Jean-Louis Hellié and villagers; Scheepmans et al. 1993.

296

« & @ 2 s 2 3 Bs > Hm. 868 a gs ws , 2 © — SS 3> a3as. q0f0 a8 §& 2 00 fs Bosse & QaGo°o = S Sas oS O spa x QO SEB xs

50

3/8230 _ 5 § OH&S SS wetelwMesgstd FBoE S\5Bs£se 8 PEES2eesSEh EB~sFEES

+ >? 42) 2 2a S 5Sbl£5ate wmQD nesses D 8mm FFB HD

gs 5 os 22oc: ©

oe)

O

= GS S& = ‘og q 2s o %. e) o Ss os PS w oft g a & % Q & = S a Com %) G6 =a} oo

n= sss S83 VigSf SEU & SOR AOR GS ekSETE spes 2 O2x § ee

o}/PRe AO ASPs ASLESSRSESSRFSSESE — Oo 2S 8 Sa. 8 V5 F/SSRERPSTSSRPE_

si/ er BFR SSPFSFSESPeskR Ass Res EVs Sloe SERS Soe Bay Sx st 2 Ss 38 § 88 S/S ARSSEESEPESSES SASSI SSS ST HHEM SS Ole. s Bs 8s 3 SFaSaSSES SERRA s es Bee

S/ESSESSSSSSRRRSSSTISSSSSEES ss C1 HS Svuseswe_A A HSX_SCSCCseeseSES Ss sSsSBESERS

0 i On a i OO Oi i Cs Ca Os OH 0D 0-0 0 Ci oi Cn Oo oo oi oc oo

oO Ww es So) oO & > 2b O af z o a oS SSY= og SSE = OU oj—U = = = =| ° _ . od

f=

wo oOASo8sonar oc —=aWS &OW —~ =° wnt w% oSayoS rae =a"=> =Q2| woo56 PA OHO m4ose) ©?)

QO

o= oO °ne! gg S So on & 600 ag oO 3 = =| ga 8 S& Sa od a oD ° so SO 5 JU ae Sos ee

wo

as E oa)

3 oo ae Oo = Lao) 5BEYoes =| =

al w Eo 8 oe Mw 55 AG OUUVAEBSMAOBRF Bow

ee

Oo #9) aWe a> QO WH ran ad Og a ad i oo = Ee io) Oav as » = Oo D oO = = Om & x ae 35 > 2 = oOd ey” = = o Sas gS §S &Ey,sm LMS =) 4 ow = Sas §g Som & dem gy Se&U & = i. == = 82 wo £&

So8s 8B SEBD&A BESE mw MSS88EG MR obo) O & Ss B Bes »weo So ase Ff & 8 is

fon = a Ss i$ S25 = S'SSS ~x

i= 5 fon Fan] re oO — os ne a an sd o LY Sg~os! =fo)=!‘==“4 Go Oo o = r= Eg ow

Sa. = USE =) eas aeg@eaa &O 8 gy ©

SGe2e6 SSezE& FEas && aes wt he MO 8a 0aafFos 2 mame *Y*ea& w So x Zo =

= S Sas n& 8 Qo 8 Aes ™ | gS vant sot. cs 8s = * "m= Wise = S So 8b SSSses ZEB FSVUA5 ne Veli _ FAs g sSBeos SS eTSsSSprsesSBpss gous Qggsse&F 8% =i 222

a0} 2 wo

a=

wa O w Oo 8 4 Se Sen = Deo 2VY 28E R= ee]

3°}

fam he | i2 no! 28= ~rand

© s 8 sss Ss = 4» seks % S Ss} @ssags Sa SSBac gs ee hed ,|fa) Z8S Bs $858 8 s = BSESsS SESS SSS 8 Ss oH a) wo as § Sess te I~ — ~~ 2°83 >= ae SSSSESIESS 2s S35 2 2s aS FISss S88 § ShSe Ses ESec ese&SEES S gSSS SE SSEE g SSL PSsss =§& = Rn a) aa = Sixkses Prati tietire gs

= Pas ea APU Eee

Ss | 8 SSSS Vee Ss re pPaAR 0 = OS Re O S06 A SSEREREESE RAKES LesEE 8 Raa aA AS PEEPPEREESE TES EES SS OOSSSS222 8SSS SS8S8 SESSatatatatatatorors SSS ER ZSRS SCOSOS

om = Z* 4 .s v S ay,

a=f & og gs3 0

> S Zot a ww @2ha 5 z 2 re SSE BES BeStBssoze FESES = a)

co 2 own & a “S300 ¢ vo >» Ss oD QO

aes3 SEM &SsSEBS E2SStasess BESS 26 as SELES FES HSE GM

S oO Sa oz a = > 28of ©

o Oo wo3soos =Oo 0.0 3 80ew a=i.—s r) =~ Of ~~ 0 op a Oo mm SG Gos YW On > S cS aa oO 8 Ome oS x i a faa > Oo < ono Ww

& =“BD ) f= S se SS +t

> a's 1G o § 8s oa §& 2.6 oa oS oS = SadaG a)

awo

o

as ‘a & S GS S sik

O'AZO Ss -—asao © o So 3 a6 6 a) — oo =na FN os} om 3 8°30 G6 aa any 9 8s. & 5S oO Sad © ty ap c « Gs S "8 Soe oO Came’ Peas > Ww mamim oo.

vo) 89 S'h «0 S06 Ss: oO Far SaBASgM SG908.8509 Oe oD HO m4 oo -~ BEESEaR AHH AH — % oy ~ o

= Se S a > SRG ss 688 SS = ees x 8 6 3.8 sc Ss

Sy}

5=

29

8

=

= IEL Vw Y Se CS og 3 g 8 8

LS 4 n 2 SA PRB ase SPS SSsau 8 Ss Ss§? 8 SESQGSRRES SESH Sg ws g SPR ESGDR IS SBESSGSXSE HERB VEGE = 5.38 = OS EVE oO Oo oem oO s S Sow FES esggdege ses& SRPISSSALLSSSSSTAFESESESSESS STIS ss SSPES sS Sess LSS ssssee HS eS SPAS SESE MPR S ‘sesO FAA ss MESEC SPE

SELLS KPKPYSSISSERSSSSSSSSTWSSSCCESS SSFSSLSLSSSERSESSSSSSS S HSSIISSES SFSRAARRPSES SSS SSCS SSS SC ASe2gNsgSsssys

Ss SSS ST YV VV VSS SSS See SXMESBR PAH VOVOARAARAAAARAAARAAAARAAHRRAQHAA

::z¢3 o had 5) & fe) Z Ss. S. Sf 3= =5a ss]SaOw S gg =2 oy 5 ac

v° eC os

SS wn ob oD QW w &===&oD ay aa) Saag = =~ A oO a Sj 0902s =° & a Ssh 96 OD oS 5a SnAaa zo MM eB)

a°

|

= oO

3of= MWS o4 a OO §& se Sakae a 88 Ra Bio YVePooezZLovsgsgegesoos 3/5 £88558 YRRsSsskoeas sees s 4/9 BESBSs BSHEPEBEZESESEFARERS2

ro) 63 £2 Ss) © |Nw oS ¢ Oo 8 a, © o S Sa SX oO a al cot | ae) —_ wo Ss 4S. oO 5 ge OU & om so 3/ 2M, 8S 8segBega5wwB2 ao 52488bd5GB(8S soo c

a=

Y Ww

#2)

os = & 2 ; oc 8 © a is Od Be 5 cd x S CD omy om oe f= cS ne as o #~ a"D> 0 s 0= 3onS&5 a6 Ss

S He&8 §S oo = 8 = s& ¢ 2 2

= S2E&a #28 a iM cos 4 SD Q ‘oo

Ss S 3 = o 8 38 sis S nm es Sm oy Sss ss .,~s ssa se s§& »|/ 2 8 8 xe SR SWE Fae g Bers SU a ae) wy

=

aS ~~

seS.eB £&FsCESRPITARERSERS , S 8=

S BS see 8 © ~

2 SSSSSESEPSSSSLIESSSERSELMEPEERS i ‘> WD om TRB wA SESE CSS Vs VeSS FI SFAEEESESCLSSLSESVRSSESC SS POSH s§ 8S 8 — |(SET TESSSERSESSSFRESESQSIESSSE oS ve GxrxrrgeseBPBiuanaSR GSP ES AEC x &Ss a &Ye SC Sees PFSBRRBSBVRESLK VSoudDLRaASSaESE {as te eeRSPAS Oe SRP8§ Oo OH LS =US ~ SIAR RB ERESSLSSSSSESSssseassaRSES RS ae eS SSF ESER SES AAAs ssk_RUVS je) Sspgg FS h§ FoyeseOeoeerRrrRHARS SSAAA SC sglAAA ergs MIARAAAARAAAA AAA AARAAA

r4 =3s 8& 5= faa, os = ow as of co gO cs S = tm "TS pe S cas) no6s °Es om sok 2s~~ $3 EE oS Q So & “«EOo =m~ 20 oo ond OOay Oo 2)_a© S & Ob a)One =

So oa

x aS

20 20

©

ne

a=

5op &

-— ©

i)3 BS fa} ss)ou oOgs65a og “oS a aa & i ow § 5& eect eo FY eo =< = ZeEEs ~3Ss3 mse MS FPSSESh SHotsse S5BZ Sepa Sa Yess Bo 83 OEE ~F 8 ete Samuo¥zsgyns 5xems 28 moO a2 000& 08DAounnnoeoan 08 OD OY WwW

= ~ = WwMenn ns OU O

a)

ad

a§

55 a S A= io ™> 20 a mW 2 OS o m = aa 2&& 2 2aai os Sess 8 22823saOoEX 6d“ad ‘a

Eg a, CS Oc

> i=4_my B os cs i ae} a = 2 o cg ‘a3 ee 5S 2 EPa=- BRS > g6 mi84528oO

a & "op Ts) LA oa 2&3 ayers, eM oO i=

5 OA OD AA oa 23 ae, oe MO

“8 Ss = iS ~ s&s SsfoS_ o BREE $ 3 cS = § 8

S§SSEESS LS SS sess SSSESSESCRESSESSES WASPSES_s SESESS.LESS FEE

oe QR = >is 8 sf

SER SEATS SSSSESSESSSRIETSSS ass sSsrPseVSsesrasvssggs SPs ®EsVvRP sss Fg K eee Soa 2 8 DO 0 = SegaSSEROES ~ Ssegs