Tropical Rainforest

C O N S E R V A T I O N B I O L O G Y SERIES
Series Editors
D r F. B . Goldsmith
Ecology and Conservation Unit, Department of Biology, University College
London, Gower Street, London W O E 6BT, U K
Tel: +44(0)171-387-7050 x2671. Fax: +44(0)171-380-7096.
email: ucbtl96@ucl.ac.uk
D r E. Duffey O B E
Chez Gouillard, 87320 Bussiere Poitevine, France
The aim of this Series is to provide major summaries of important topics in

conservation. The books have the following features:
• original material
• readable and attractive format
• authoritative, comprehensive, thorough and well-referenced
• based on ecological science
• designed for specialists, students and naturalists
In the last twenty years conservation has been recognized as one of the
most important of all human goals and activities. Since the United Nations
Conference on Environment and Development in Rio in June 1992, bio-
diversity has been recognized as a major topic within nature conservation,
and each participating country is to prepare its biodiversity strategy. Those
scientists preparing these strategies recognize monitoring as an essential
part of any such strategy. Chapman & H a l l has been prominent in publish-
ing key works on monitoring and biodiversity, and with this new Series
aims to cover subjects such as conservation management, conservation
issues, evaluation of wildlife and biodiversity.
The Series contains texts that are scientific and authoritative and present
the reader with precise, reliable and succinct information. Each volume is
scientifically based, fully referenced and attractively illustrated. They are
readable and appealing to both advanced students and active members of
conservation organizations.
Further books for the Series are currently being commissioned and those
wishing to contribute, or who wish to know more about the Series, are
invited to contact one of the Editors or:
D r R. C . J. Carling
Senior Editor, Life Sciences, Chapman & H a l l , 2-6 Boundary Row,
London SEI 8 H N , U K .
Tel: +44(0)171-865-0066. Fax: +44(0)171-410-6907.
email: bob.carling@chall.co.uk
Books already published and others being planned or written are listed overleaf...
Already Published
1. Monitoring Butterflies for Ecology and Conservation

E. Pollard and T . J . Yates (eds) (Hb 1993 o/p, Pb 1995) xiv+274pp.
ISBN 0-412-63460-0
2. Insect Conservation Biology

M . J. Samways (Hb 1994 o/p, Pb 1994) xv+358pp. ISBN 0-412-63450-3
3. Monitoring for Conservation and Ecology

E B. Goldsmith (ed.) (Hb/Pb 1991, Pb reprinted four times) 275pp.
ISBN 0-412-35600-7
4. Evaluation and Assessment for Conservation: Ecological Guidelines

for Determining Priorities for Nature Conservation
I. F. Speilerberg (Hb 1992 o/p, Pb 1994 reprinted three times)
xvi+260pp. I S B N 0-412-44280-9
5. Marine Protected Areas: Principles and Techniques for Management

S. Gubbay (ed.) (Hb 1995) xii+232pp. I S B N 0-412-59450-1
6. Conservation of Faunal Diversity i n Forested Landscapes

R. M . DeGraaf and R. I. Miller (eds) (Hb 1995) xxi+633pp, with
colour plate section and colour foldout. ISBN 0-412-61890-7
7. Ecology and Conservation of Amphibians

T. J . Beebee (Hb 1996) viii+214pp. ISBN 0-412-62410-9
8. Conservation and the Use of Wildlife Resources

M . Bolton (Hb 1997) xviii + 278pp, with colour plate section. ISBN
0-412-71350-0
9. Conservation Management of Freshwater Habitats

N . C M o r g a n and P. S. Maitland (Hb 1997) ISBN 0-412-59412-0
10. Tropical Rain Forests: A Wider Perspective

E B . Goldsmith (ed.) (Hb/Pb 1998) I S B N H b 0-412-81510-9, Pb
0-412-81520-6
Forthcoming
11. Valuation of Costs and Benefits of Wildlife i n Africa

H . H . T. Prins and J . G . Grootenhuis (eds) (Hb 1998) ISBN 0-412¬
79730-5
12. Primate Conservation Biology

G . Cowlishaw and R. I. M . Dunbar (Hb 1998) ISBN 0-412-71340-3
Coastal Management for Nature Conservation
P. Doody ISBN 0-412-59470-6
Wetland Ecology and Management

B. D . Wheeler ISBN 0-412-59430-7
Grazing as a Tool i n Conservation Management

S. E. van Wieren, J . P. Bakker and J . Bokdam ISBN 0-412-47520-0
Full details of these and all Chapman & H a l l books are available online
http://www.chaphall.com
Tropical Rain Forest:
A Wider Perspective
Edited by
F.B. Goldsmith
m
Springer-Science+Business Media, B.V.
First edition 1998
© 1998 Springer Science+Business M e d i a Dordrecht

Originally published by Chapman & Hall Ltd in 1998
Thomson Science is a division of International Thomson Publishing I(T)P
Typeset in 10/12pt Sabon by Columns Design Ltd, Reading
ISBN 978-0-412-81520-1 I S B N 978-94-011-4912-9 (eBook)

D O I 10.1007/978-94-011-4912-9
All rights reserved. No part of this publication may be reproduced, stored in a retrieval sys-
tem or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording or otherwise, without the prior written permission of the publishers. Applications
for permission should be addressed to the rights manager at the London address of the pub-
lisher.
The publisher makes no representation, express or implied, with regard to the accuracy of
the information contained in this book and cannot accept any legal responsibility or liability
for any errors or omissions that may be made.
A catalogue record for this book is available from the British Library
Library of Congress Catalog Card Number: 97-69618

Contents
1 Tropical rain forests - what are they really like? 1

F. B. Goldsmith
1.1 Where are rain forests? 1
1.2 How fast is rain forest disappearing? 3
1.3 Definitions 4
1.4 Boundaries 5
1.5 So what are they really like? 5
1.6 Why so much diversity? 11
1.7 How serious is fragmentation? 11
1.8 Quantitative studies in rain forest 13
1.9 Case study: Tiwai Island 13
1.10 Sketch from the Amazon 14
1.11 Other regional studies 15
1.12 Conclusions 15
Acknowledgements 15
References 16
2 Indigenous non-timber benefits from tropical rain forest 21

G. T. Prance
2.1 Introduction 21
2.2 Indigenous use of the forest 21
2.3 Indigenous management of plant resources 23
2.4 Extractive reserves 24
2.5 Sustainability 32
2.6 The Convention on Biological Diversity 36
2.7 Oligarchic forests 38
2.8 Timber and non-timber forest products 39
2.9 Conclusions 39
References 40
3 Degeneration versus regeneration - logging in tropical rain forests 43

N. Brown
3.1 Regeneration of tropical rain forests 43
3.2 Silvicultural systems 48
3.3 The effects of logging on forest structure and composition 56
Vlll Contents
3.4 The hydrological cycle in tropical rain forests 64
3.5 The effects of logging on forest hydrology and erosion 65
3.6 Impact on animal populations 67
3.7 Conclusions 67
References 68
4 Forest people and conservation initiatives: the cultural

context of rain forest conservation in West Africa 75
B. Sharpe
4.1 Who are the indigenes? 78
4.2 First the forest: villages and the socialization of space 80
4.3 Forest destruction: who are the villains? 84
4.4 Myths revisited: environmental management plans and
local realities 87
4.5 Not all doom and gloom - engagement of conservation with
local aspirations 91
4.6 The UCUODA community forest management project 92
4.7 Conclusion: forest futures, forest cultures 93
Acknowledgement 95
References 95
5 Forest and environmental degradation 99

M.]. Eden
5.1 Forest degradation 100
5.2 Biophysical implications of forest degradation 105
5.3 Strategies for forest conservation 110
References 112
6 Soil and vegetation effects of tropical deforestation 119

S. M. Ross
6.1 Introduction 119
6.2 Effects of different deforestation practices on soil and
vegetation 121
6.3 Deforestation effects on microclimate 121
6.4 Deforestation effects on soil physical properties 124
6.5 Deforestation effects on soil erosion 126
6.6 Deforestation effects on soil chemical properties 128
6.7 Changes in soil conditions with time after deforestation 135
6.8 Deforestation effects on soil biology 146
6.9 Deforestation effects on vegetation 153
6.10 Conclusions 163
References 165
Contents IX
7 Causes of tropical deforestation and institutional constraints

to conservation 175
K. S. Bawa and S. Dayanandan
7.1 Deforestation rates 176
7.2 Causes of deforestation 182
7.3 Institutional constraints 189
704 Conclusions 194
Acknowledgements 194
References 195
8 The economics of the tropical timber trade and sustainable

forest management 199
E. B. Barbier
8.2 Trends in forest products trade and deforestation 200
8.3 Market access and trade barriers 207
804 Certification and labelling 221
8.5 The costs of implementing sustainable forest management 229
8.6 Financing sustainable forest management 239
8.7 Summary and conclusions 245
References 250
9 Can non-market values save the tropical forests? 255

D. Pearce
9.1 Introduction: the forest crisis 255
9.2 What do we know about non-market values? 257
9.3 Will non-timber values save the forests? 265
Acknowledgement 265
References 265
10 The role of policy and institutions 269

James Mayers and Stephen Bass
10.1 Why policy matters: dealing with multiple interests and
change 269
10.2 Policy influences on forests 273
10.3 Policy challenges for sustainability - national and
international 279
lOA Institutions and stakeholders 282
10.5 Emerging lessons on 'policy that works' 289
10.6 Signs of change 299
References 300
x Contents
11 Modelling tropical land use change and deforestation 303
A. Grainger
11.2 Background 303
11.3 A theoretical model of the causes of deforestation 305
11.4 Regression models of the causes of deforestation 311
11.5 Theoretical models of trends in forest cover 319
11.6 Modelling trends in forest cover 322
11.7 Modelling long-term spatial deforestation trends 335
References 341
12 Communicating the message: a case study from the Royal

Botanic Gardens, Kew 345
Colin Clubbe
12.2 What is the message and why should we bother? 346
12.3 Types of audience 347
12.4 Resources available 355
12.5 Is the message getting across? 361
References 365
13 Effective campaigning 367

A. Juniper
13.1 The campaigners 367
13.2 The campaign - origins and impacts 368
13.3 The timber trade - importing deforestation 371
13.4 The private sector in the rain forests 377
13.5 Aid, development, the international financial institutions
and the rain forests 380
13.6 International agreements 383
13.7 Rain forests - campaign priorities for the future 385
Acknowledgements 389
References 389
14 Synthesis 391
F. B. Goldsmith
14.1 Production 391
14.2 Protection 392
14.3 Conservation 392
14.4 Services for people 392
14.5 Policies and institutions 392
Contents Xl
14.6 Knowledge base 393

References 394
List of acronyms 395
Index 397
List of contributors
Edward B. Barbier
Environmental Economics and Environmental Management Department,
University of York, Heslington, York Y01 5DD, UK
Stephen Bass
Forestry and Land Use Programme, International Institute for
Environment and Development (lIED), 3 Endsleigh Street, London
WC1HODD, UK
Kamal S. Bawa
Department of Biology, University of Massachusetts, 100 Morrissey
Boulevard, Boston, MA 02125, USA
Nick Brown
Department of Plant Sciences, University of Oxford, South Parks Road,
Oxford OX1 3RB, UK
Colin Club be
Royal Botanic Gardens, Kew, Surrey TW9 3AB, UK
S. Dayanandan
Michael J. Eden
Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK
Barrie Goldsmith
Department of Biology, University College London, Gower Street, London
WC1E 6BT, UK
Alan Grainger
School of Geography, University of Leeds, Leeds LS2 9JT, UK
Tony Juniper
Friends of the Earth, 26-28 Underwood Street, London N1 7JQ, UK
XIV List of contributors
James Mayers
Forestry and Land Use Programme, International Institute for Environment
and Development (lIED), 3 Endsleigh Street, London WC1H ODD,
UK
David Pearce
Centre for Social and Economic Research on the Global Environment
(CSERGE), University College London, Gower Street, London WC1E
6BT, UK
Ghillean T. Prance
Sheila M. Ross
SGS Environment, Yorkshire House, Chapel Street, Liverpool, Merseyside
L3 9AG, UK
Barrie Sharpe
GECIESRC Rainforest Research Group, Department of Anthropology,
University College London, Gower Street, London WC1E 6BT, UK
Foreword
The international perspective for this book is the unprecedented level of

concern over deforestation, recognized by the meeting of world leaders at
the 1992 Earth Summit, in Rio do Janeiro, and culminating in the appoint-
ment of the Intergovernmental Panel on Forests (IPF), under the auspices of
the UN Commission on Sustainable Development. The wide range of issues
covered by the authors in this volume reflects the breadth of the interna-
tional debate, from national policies and activist campaigning, through eco-
nomic and social objectives, to the sustainable management of forest and
soil resources.
Since the conservation campaigns of the 1980s, the focus of international
concern has widened from tropical rain forests to all forest formations, in all
regions, with increased recognition of global values and common responsibil-
ities. However, while forest cover in some temperate countries is increasing,
irrational deforestation, at historically unprecedented levels of damage to
biodiversity and to other environmental values, remains most acute in tropi-
cal countries, where the need to use the natural resources for sustainable
development is greatest, and the capability weakest. While accepting the
urgency of the situation, and the need for greater coherence of action at a
global level, the 1997 report of the IPF to the UN Commission emphasized
the powers and responsibilities of national governments, and the importance
of National Forest Programmes, but with the fuller participation of local
communities, and with enhanced access to international assistance.
Seen from a historical perspective, concern over the impact of tropical
deforestation on environmental and socio-economic values, at both local
and national levels is not new. In 1847 Dr Alexander Gibson, appointed
the first Conservator of Forests in the Bombay Presidency in India, under-
took an inventory - not of the trees, but of the rivers and streams of the
Malabar Coast which had silted up within living memory to the extent that
ships could no longer ride at anchor, apparently as a consequence of defor-
estation in the hills. In response to this and similar reports, in 1850 the
British Association for the Advancement of Science appointed a Committee
to 'consider the probable effects in an Economical and Physical point of
view of the destruction of tropical forests'. The key member, Dr Hugh
Cleghorn, was like Dr Gibson a medical officer, selected for his scientific
knowledge. The report strongly emphasised the environmental importance
of forests for the 'fertility of the land and the prosperity of the people',
XVI Foreword
including their contributions to fuel supplies, animal fodder and soil fertil-
ity. It also drew attention to the commercial value of the rain forests, not
only for their timbers, but for the wide variety of gums, resins, dyes, oils
and drugs, both those already known and others awaiting discovery -
recognition of the option values of biodiversity, a century before that term
had been coined. The recommendations made in the report led, in 1855, to
the creation of the Indian Forest Service, and the assumption by govern-
ment of the legal powers and responsibilities for forest conservation.
The Indian Forest Service had a distinguished record of scientific study
and documentation of the forests, and the model for the legal establish-
ment and management of a national forest estate was adopted under
British colonial administration in other tropical countries. Despite the con-
cern over tropical deforestation expressed by scientists in 1850, most tropi-
cal countries then still possessed very extensive natural forests and, with
low levels of human population, the reservation of selected areas for envi-
ronmental protection, resource conservation, and management for timber
production, could be made in parallel with the conversion of other forested
areas to farmland, or their continued use by local communities for grazing,
fuel and fodder production. However, government forest policies tended to
maximize industrial and revenue objectives, with inadequate provision for
reinvestment in the sector, and insufficient attention to the forest related
needs of local communities. With rising demands from population growth
and economic growth, and increased access by the rich and powerful to the
technology for large-scale timber exploitation, and to profitable markets,
the environmental, social and economic values of tropical rain forests,
identified and documented a century and a half ago as the basis for conser-
vation, came increasingly into conflict.
The need to restore a sustainable balance between social, environmental
and economic uses is particularly acute and difficult to achieve for tropical
rain forests. The unique quality of these forests, of global significance, is
their rich biodiversity. However, despite the supposition of high potential
value in the future commercialization of biodiversity, the prospects for sig-
nificant financial flows to meet the costs of conservation are distant and
uncertain. There are substantial 'trade-offs' between the need of present
generations of poor countries, and poor rural people, to exploit the forest
and land resources, and future benefits from their conservation, which may
incur in part to other relatively affluent communities. This brings the need
for a better balance also in the sharing of costs between global, national
and local interest groups, taking into account the interests of future genera-
tions.
Governments, both in tropical countries and in the more affluent indus-
trialized economies, which should protect the interests of the poorer ele-
ments in society, and represent those of 'voiceless' future generations, have
many immediate demands on scarce financial resources, and are commonly
Foreword XVll
influenced by more powerful short-term sectional interests. There is

increasing recognition that the expropriation of local peoples' rights of
access to forest lands and forest resources, both by private industry and by
governments, has been a frequent cause of conflict and of related forest
degradation. The narrow perspective of many government forestry services
and entrenched official attitudes, to issues of legal access, property rights
and a more equitable share of benefits to local communities from publicly
owned forests, are slow to change. Similarly, at the international level, the
hard-won acceptance under the Biodiversity Convention of the need for
new and additional sources of finance to assist developing countries to
meet their costs for biodiversity conservation, and the related provisions
under the Global Environment Facility, have so far failed to deliver the nec-
essary assistance on the scale and with the urgency needed. The report of
the IPF and its recommendations for actions by national governments, with
international assistance, are welcome, but the level of political will to
implement the necessary radical changes to current practices with the
urgency required is questionable. The effectiveness of the proposed
Intergovernmental Forum on forests, intended to oversee the implementa-
tion of the IPF recommendations and to find ways to strengthen interna-
tional mechanisms for forest conservation, is also doubtful, unless
accompanied by strong contributions to the debate from those most
directly concerned, and best informed, at national and local levels.
It was the mobilization of public and media attention to tropical defor-
estation by environmental campaigns that first raised the issues to a high
political level. That stimulus is still needed, but with a wider perspective,
which seeks to maximize the value of the tropical rain forests to the coun-
tries and rural communities, through their sustainable use, including the
harvesting of timber and other products, in harmony with conservation of
biodiversity and other environmental values. The World Commission on
Forests and Sustainable Development (WCFSD), in informal liaison with
the IPF, but without the formal authorization of governments that was
invested in the Panel, has sought to involve civil society generally to resolve
current conflicts that contribute to deforestation, and to achieve a more
positive relationship between local communities, private industry and gov-
ernment, in the sustainable development of forest resources. This approach
has significant policy and institutional implications, as increasing public
concern over environmental and social values of forests in many countries,
and the rapid development of international communication and informa-
tion systems, combine to strengthen the global influence of civil society on
the actions of governments in respect of both their national and interna-
tional responsibilities. It is significant that India, largely as the result of the
influence of its academic and environmental campaigners, is now in the
forefront of new approaches to involve local communities in participatory
management of tropical forests.
XVlll Foreword
The chapter by James Mayers and Stephen Bass suggests that, in the con-
text of the national acceptance of each country's responsibility for forest
conservation, civil society should provide a wider perspective, through the
development of a national vision for forests, with the creation of some kind
of national forum, representing all the interest groups most directly related
to the forest. Such a national vision requires improved understanding of
many complex and contentious issues, relating for example to the policy
and institutional framework; the social and ecological impacts of forest
degradation; the safe limits to forest use, in terms of soil resources, regener-
ation processes and ecosystem resilience; the potential of non-timber forest
products in multiple use management; the importance of social and cul-
tural values to conservation strategies; the possible mechanisms for
increased financial support for sustainable forest management linked to the
international timber trade andlor compensatory payments for global ser-
vices of tropical rain forests, etc. These and other important issues form the
main content of this compilation, by authors exceptionally well qualified to
provide this wider perspective.
My own involvement with tropical forests, initiated nearly 50 years ago
by concern for wildlife, and spent almost entirely in government service
and in research into forest conservation and management, led to my
involvement in the 1980s in international policy development, in the con-
ception of the Tropical Forestry Action Plan, and the early years of the
International Tropical Timber Organization. Those initiatives failed to
meet the high expectations at the time, partly through too narrow focus
within the forest sector, and the related failures to address underlying
causes of deforestation. Since my formal retirement from government ser-
vice in 1990, my involvement as Chairman of the UK Tropical Forest
Forum has impressed on me the importance of the wider perspective, link-
ing conservation with sustainable use of the forests, being shared as widely
as possible within civil society as a whole. Initiated by NGOs concerned
with both environmental and social issues, the UKTFF has support from
government and from the scientific and academic community, and involves
the private sector also, in the attempt to improve understanding and to
make Britain's total contribution to the conservation and sustainable man-
agement of tropical forests, through all these international relationships,
more coherent, and more effective.
The wider perspective provided in this book will be of value to students
of ecology, forestry, environmental economics and related fields of anthro-
pology and conservation, at graduate and postgraduate levels, and also to
those directly involved, and the wider public deeply interested, in the use
and conservation of tropical rain forests.
Ron Kemp
Preface
I have valued tropical rain forest since 1969 when, at the suggestion of
Professor Paul Richards, I went on the Royal Society/Royal Geographical
Society Expedition to Mato Grosso, Brazil. This spell of 3 months was fol-
lowed by visits to Nigeria and Ghana in the 1970s, to the Caribbean and
Sierra Leone in the 1980s and to India, Bolivia and Australia more
recently. During this time I have been concerned about the losses and cur-
rent rate of loss, and the fact that so many well-meaning agencies have
been unable to stem the losses.
This book has grown from a module designed for the MSc courses in
Conservation and Environmental Economics at University College London.
The former has run since 1960 and the module was for nearly a decade
under the tutelage of Professor Andrew Warren. I became responsible for it
in 1996 and I decided to bring the range of material offered to a wider
audience. Each year Professor Sir Ghillian Prance has kindly invited us to
Kew Gardens and Dr Colin Clubbe has organized the day for us. It is a
pleasure to thank them too. I am also indebted to three research students,
Jo Pires, Chris Birkinshaw and Joe Comiskey, who have worked on various
tropical rain forest topics under my supervision and who have expanded
my horizons and deepened my knowledge of tropical rain forest.
We need to understand the resource, the needs of local people, the insti-
tutions involved, and the kinds of incentives that might work: indeed, to
take a truly comprehensive and interdisciplinary approach. This has always
been the approach of our MSc course and we have pleasure in now sharing
with others the material that we have collected. The only real test of its
usefulness is if more people become involved in an appropriate and effec-
tive way, and the rates of forest loss are at least slowed.
I am privileged to be able to select and edit this volume and hope it rep-
resents a useful partner to Paul Richards' second edition of The Tropical
Rain Forest, which is exclusively ecological. This book, we believe, pre-
sents a wider perspective and one that looks forward to the future.
Barrie Goldsmith
-1------
Tropical rain forests - what are they
really like?
F. B. Goldsmith
Hardly a week passes without a new report of threats to another area of

tropical rain forest. As I write, this week's article is about Guyana, where
two international companies, one Canadian and the other Malaysian, are
trying to acquire 1.5 and 0.5 million acres, respectively (Guardian, 26 June
1996). Guyana, on the northern edge of the Amazon Basin, is still 85% lush
tropical forest, and rich in species unique to the Guyana Shield. The forests
are home to Amerindians, including such little-known tribes as the Wia-Wai,
Makushi and Wapishani. All depend on the forest for hunting and farming.
For future drugs, genes for new agricultural innovations, and global climatic
stability, we in the developed world also depend on these forests.
Countries such as Guyana are faced with high unemployment, low
wages and a sagging economy, and multinational logging companies offer
governments a tempting solution: they could sell off vast areas of rain
forest and the foreign investment would briefly boost the economy, at least
until the next general election.
1.1 WHERE ARE RAIN FORESTS?

They occur in 30 different countries scattered across South and Central
America, Africa, the Indian subcontinent, and South-East Asia. They occur
between 23° Nand 23° S, that is between the Tropics of Cancer and
Capricorn, which is the band of sub-tropical high pressure between the
Tropics (Figure 1.1).
These areas have high rainfall (1700-3000 mm/year), more-or-Iess
equally distributed month by month, high humidities and high mean
annual temperatures (20-28° C) (Walsh, 1966). Temperatures can vary
2 Tropical rain forests - what are they really like?
Figure 1.1 The main areas of tropical rain forest (after Golley, 1983).
considerably due to altitude and height in the canopy as well as the daily
range and season.
The total area of tropical rain forest is not precisely known and esti-
mates range from around 600 million ha to over 1000 million, depending
on definition and assumed losses in historical times. Estimates of timber
volumes also vary from 5-35 m3/ha in Africa to much higher estimates in
Asia, of the order of 50-120 m3/ha. The actual biomass above ground can
be around 100 tlha (including about 10 tlha litter) with about 100 tlha
below ground. The growth rate is around 2 m3/halyear.
But these bald statistics do not reveal anything of the richness of rain
forest: its immense extent; enormous trees; large number of species (297
lowland rain-forest birds in Africa alone); spectacular taxa such as pri-
mates (chimpanzees, orang-utans, gibbons, gorillas, monkeys and pro simi-
ans such as lemurs - Smuts, 1987; Davies and Oates, 1994), elephants,
okapi, pangolin, genet, duiker, fruit bats, parrots, macaws, plantain-eaters,
toucans and horn bills; a wide range of niches; specialized pollinators; and
a wide range of seed dispersers (including birds, lemurs in Madagascar, and
even elephants). Prance estimates that there are one million different
species in rain-forest environments (Prance, 1982, 1989; Prance and
Campbell, 1988).
This 7% of the world's land area has 50% of its species. The richest
areas, such as Para in Brazil, may have as many as 230 tree species to the
hectare (Prance, 1989) or 306 trees/ha in Ecuadorian Amazon (Whitmore,
1990). At this scale of diversity every second tree on a 1-ha plot is likely to
be a different species.
How fast is rain forest disappearing? 3
However, there are several paradoxes here. This species-rich biome is not
always very varied: there are areas that are quite uniform and even monot-
onous to walk through. Even in such areas, the old saying holds true that
in rain forest you look for one thing and usually find another. A further
paradox is that not all areas have rich soils: many of the nutrients are often
tied up in the living parts of the ecosystem and rather little is found in the
underlying mineral soil. There is a dearth of good research on soil
processes (Tanner, 1985; Ross, this volume). But why should we expect any
generalization to hold true for something that is both so diverse and covers
such a huge area. Also, the rainfall does not occur regularly throughout the
year; neither does it normally rain all day as some people imagine.
Useful and interesting general accounts of rain forests are given by
Odum, 1970; Meggers et aI., 1973; UNESCO, 1978; Golley, 1983; Sutton
et ai., 1983; Whitmore, 1984, 1990; Lieth and Werger, 1989; Kricher,
1989; Mabberley, 1992, Archibold, 1995; Richards, 1996.
1.2 HOW FAST IS RAIN FOREST DISAPPEARING?

The answer varies, ranging from countries such as Sierra Leone which has
lost 85% of its forest (and 97% of its primary forest) to Guyana which
retains 85%, and Brazil which may retain 90% of its total rain forest but
has lost 88% of its Atlantic coastal forest (Brown and Brown, 1992).
Unfortunately nobody can be sure, as the statistics for cutting are unreli-
able and satellite imagery cannot always differentiate between virgin tropi-
cal rain forest, secondary forest and tropical plantations (Sayer and
Whitmore, 1990). Frequent cloud cover also makes viewing and interpreta-
tion difficult. Current island biogeographic theory considers the relation-
ship between remaining area and species number, and suggests that if we
can retain 20% of the area, we will keep most of the species (Myers, 1979,
1989). Others would dispute this figure, or emphasize that it would depend
how the protected areas were distributed and whether climatic change
occurred. (For a more detailed discussion about species-area relationships,
see Simberloff, 1992; Whitmore and Sayer, 1992; Peres and Terborgh,
1995.) M. Holdgate, in the Foreword to Whitmore and Sayer (1992), sug-
gests that between the years 1600 and 1992, 83 species of mammal, 113
species of bird and 384 species of vascular plant have been lost.
The threats to rain forest are many and varied. The most important are
shifting cultivation and cutting for timber. Shifting cultivation is covered by
Sharpe (this volume). Cutting for timber is often carried out in a selective
manner in tropical forests, taking the high-value species and leaving the rest,
known as high-grading or 'cookie-jar management'. In some ways this is
less damaging than clear-cutting, but it results in genetic deterioration of the
harvested species. (Useful general books about forest management include
Waring and Schlesinger, 1985; Perry, 1995; more specifically about tropical
forests or the shade tolerance of tropical trees are Wyatt-Smith, 1963;
Champion and Seth, 1968; Symington, 1974; Evans, 1984; Anderson et al.,
1995.) The effect of high-grading is sometimes referred to as 'dysgenic', i.e.
it damages the genetic range of the target tree species. Other timber is taken
for fuel wood or charcoal production, and this often exceeds 50% of the
extracted volume. For a useful summary of the damage caused by logging
and the differences in fauna between disturbed and undisturbed areas, see
Johns, 1992. However, the roads constructed for logging may have more
serious effects than the logging itself, because they give access to other users
such as migrant farmers, who burn the remaining biomass and plant crops.
Areas of poor soil, which are widespread, support arable cropping only for
a few rotations, and are then abandoned. The probability of return to the
original complex rain forest is slim (Sharpe, this volume, argues that the
complex rain forest of West Africa is a humanly managed system). The criti-
cal issue is that shifting cultivation is sustainable at low population density,
but the growth of populations has increased the area of impact and reduced
the length of fallow periods. Other developments include slash-and-burn on
an extensive scale, planned government-supported agriculture, and mining.
All result in a loss of the nutrients in the biome, a loss of organic matter,
water-retaining capacity and nutrients in the litter (which is usually largely
destroyed by fire or oxidation by the sun), loss of soil structure, and com-
paction. At the same time sources of nitrogen such as soil mycorrhizae and
nitrogen-fixing epiphylls are also lost.
Extensive rain-forest clearance, especially of areas such as the Amazon,
might result in local changes in climate. It has also been suggested that this
could result in global climatic changes via changes in global carbon dioxide
levels and the albedo of the planet surface, resulting in global temperature
nse.
1.3 DEFINITIONS
There are myriad different terms relating to tropical rain forests, such as:
• tropical rain forest
• tropical moist forest
• tropical evergreen forest
• humid tropical forest
• monsoon forest
• tropical deciduous rain forest
These are all used more or less interchangeably in this chapter; generally
the term used is that in the original publication. The English colloquial
term for rain forest is 'bush' or 'jungle'; fortunately the Latin 'pluviisylva'
and Greek 'Hylaea' seem to have fallen out of use.
The term tropical forest has a more general meaning, whereas terms such
So what are they really like? 5
as tropical deciduous forest, tropical dry forest, savanna(h), montane rain
forest, and subtropical rain forest have different meanings, which should
be clear. While recognized as being important, they have been omitted from
this chapter.
1.4 BOUNDARIES
With the exception of coastlines, tropical rain forests do not have clear
boundaries (Walter, 1971; Swaine et al., 1976: Bullock et al., 1995). Where
boundaries do occur, e.g. in Nigeria, they may be sharpened by human
activities such as burning (Hopkins, 1974). Rain forest grades into dry
tropical forest such as the cerrado of South America, the dry seasonal sal
(Shorea robusta) forest in India, and moist temperate forest as found in
Australia. These other forest types are also important, with high biodiver-
sity levels and rare and distinctive species, and are similarly under threat.
In some ways these transitional zones are as interesting and important as
the core areas of rain forest.
1.5 SO WHAT ARE THEY REALLY LIKE?

1.5.1 Structure
Primary rain forest is not usually dense and is not difficult to penetrate,
although this statement is not true of river banks or the edges of clearings,
farms and roads. The ground is often more or less clear of vegetation. This
is probably because only 1-3% of the light actually reaches the forest floor
(Hemming et al., 1988). Much is made of the complex structure of tropical
rain forest. Trees are dominant and evergreen, but the number of layers in
the canopy is the subject of much debate and may not be very important.
What is important, however, is the variety of structures including many dif-
ferent woody climbers or lianas, trees with large canopies, emergents, a
wide variety of epiphytes, trees that flower on their branches and trunks
(cauliflory), buttress roots to prop up such giants, stilt roots on trees such
as figs, and roots that drop from above or 'sinkers'. Leaves often have
swollen petiolar joints and an exaggerated acumen or 'drip' tips, up to
4 cm long, to aid the shedding of water and concomitant fungal spores.
(Publishers may inadvertently reveal their ignorance by placing pho-
tographs or figures with such leaves upside down.) Leaves and leaflets tend
to be of uniform size, length between 5 and 10 cm long, entire-margined
and mesophyllous.
Figure 1.2 shows a profile of rain forest (Golley, 1983); Figure 1.3 shows
diagrammatically a range of features characteristic of rain forests. The first
impression is likely to be that the majority of the plants are woody, includ-
ing many of the climbers, and there may be many epiphytes (Richards,
Figure 1.2 Profile of rain forest in French Guiana (after Halle et aI., 1978, repro-
duced by Golley, 1983).
1996). Secondly, attention is drawn to the size of the trees: 46-55 m

(150-180 feet). Thirdly, to a temperate botanist at least, familiar plants or
their close relatives in rain forest are large and woody, such as vervains
(Verbenaceae), milkworts (Polygalaceae), periwinkles (Apocynaceae), vio-
lets (Violaceae) and daisies (Asteraceae). Even grasses are large and wide-
leaved. The canopy is shown in Figure 1.3 as having leaves with drip tips.
The leaves and their epiphyllae enhance cation absorption from precipita-
tion and, along with twigs or wood, may yield exudates such as mucilages,
varnishes, resins and latex. Many are useful to humans but their real value,
and the reason they are there, is to protect the plant from invertebrate or
vertebrate attack. Some of the exudates may be allelopathic and prevent
the germination of the same or other species (Mabberley, 1992). The trees
also support lianas, which are able to reach the light without having to
epiphyte
drip-tips
allelopathic
exudates
lianas
x t r - - - cauliflory
termitaria
;; ~oo.~
~ ectotrophic
mycorrhizae
Figure 1.3 A range of features characteristic of rain forest.
invest in so much wood (capital - or is it essential infrastructure?). The

trees show cauliflory as flowers may be more readily pollinated if they are
more conspicuous away from the mass of foliage in the canopy. Trees show
great buttresses to help support such large structures. These are like the fins
of a rocket and extend up to 3 m up the trunk of the tree. This does not
make harvesting easier and may necessitate the construction of platforms
by timber fellers, often resulting in rather inefficient extraction of wood.
Tropical rain forests contain a large number of epiphytic plants. Many of
these are orchids and a large number of species are found. In the Malay
Peninsula the Orchidaceae is the major family, with 800 species (Richards,
1996). In the New World the Bromeliaceae are particularly well represented.
1.5.2 Regeneration
Tropical rain forests are obviously dynamic systems whose future depends on
their natural regeneration. Studies have shown that it is difficult to generalize
about the regeneration requirements of different tree species, and such gen-
eralizations can lead to serious consequences in relation to appropriate
harvesting and management directed at promoting regeneration (Janzen,
1976; Hall and Swaine, 1980; Whitmore, 1983; Gomez-Pompa et al.,
1991). Some trees are shade-tolerant whilst others are light-demanding.
Each has its role in the regeneration cycle. The abundance, distribution and
geometry of canopy gaps and the way they affect regeneration have been
widely studied recently, especially in areas such as Barro Colorado Island,
Panama (Hubbell and Foster, 1986). There are different theories concern-
ing what controls the replacement of species, but the main parameters are
the species involved, their reproductive strategy and the size of gap. There
is much discussion of 'gap ecology' and whilst some people find it useful to
identify 'regeneration guilds', each species is really a law unto itself, having
slightly different regeneration requirements. The identification of guilds
may be a pigeon-holing system simply for our convenience, lacking any
real biological meaning.
1.5.3 Seasonal rhythms

A popular myth is that there are no seasonal rhythms in tropical rain for-
est. There may be some areas with no rhythm but the vast majority of
areas have marked, if somewhat variable, annual cycles. Trees may be like
the evergreen trees of the temperate zone, or leaf fall may be seasonal, but
leaf fall can also be irregular and complete. Some trees show continuous
growth whilst others show intermittent growth. There is often synchro-
nization of leaf production and flowering. Such annual phenological
rhythms can include two leaf flushes but only one flowering. There are
obvious advantages in such synchrony but the mechanisms are not always
clear. They may be controlled by the water status of trees, as has been
detected by the use of dendrometer bands. Tropical forest phenology has
been modelled, for example by Pires O'Brien and O'Brien (1995) who
found that even in the Amazon basin there are marked and regular annual
cycles of leaf flushing and flowering.
Many tropical trees are mast-fruiting, that is they produce seed in large
quantities at irregular intervals. One probable explanation for this phe-
nomenon is that it holds the population levels of predators down and per-
mits pulses of regeneration. Some trees are monocarpic, that is they flower
only once in their lifetime, the so-called 'big bang'. They are mostly mono-
cotyledons such as bamboos and palms. Flowers are usually ephemeral,
but some trees are characterized by having very persistent flowers.
Monsoon forest occurs in the Indo-Malayan region, and reflects the
irregular rainfall. It is more or less leafless during the dry season, especially
towards its termination, and is rich in woody lianas and herbaceous epi-
phytes, but poor in woody epiphytes.
In the Amazon there are regular annual floods to a depth of over 3 m
(10 feet), and several species such as egg-laying turtles and ground-nesting
birds including plovers and skimmers are dependent on its regularity.
1.5.4 Aunimals
The animals of rain forest are as diverse as the plants. They show strong
specialization (such as preferences for different foods), spatial and time-of-
day partitioning. In Maraca, Brazil, 450 species of bird and 45 species of
bat have been recorded (Robinson in Hemming et aI., 1988). The rivers are
often the richest areas, and in the case of Maraca have been described as
being like aviaries. In Malaysia 203 species of land mammal have been
recorded, 83 species of bat, 54 rodents and 29 carnivores. Other taxa such
as herpetiles and fish are less well documented but are known to be compa-
rably rich. Malaysian rain forest has 1200 species of termite, and over 1000
species of butterfly, of which over 200 species are birdwings (a kind of swal-
lowtail). In Muly National Park in Sarawak, there are 4000 species of bee-
tle. It is difficult to extrapolate to produce national and world totals. We
can say only that the biodiversity is impressive. Fungi beat even the animals,
with 1.5 million species estimated to occur in the forests of South-East Asia.
In most areas of tropical rain forest, birds (such as horn bills, plantain-
eaters and fig-birds) and bats are important for the dispersal of large-
seeded, often fleshy fruits (Janzen, 1975; Richards, 1996). In Guyana, for
example, up to 90% of tree seeds are dispersed by mammals and birds
(Richards, 1996). The seeds are usually embedded in pulp which is an
attractive food. Odours are known to attract fruit bats, whilst bright
colours attract birds. Many bird species are highly specialized for eating
tropical seeds, e.g. parrots, toucans and hornbills.
However, in Madagascar the role of birds has been largely taken over by
pro simian lemurs (Birkinshaw, 1995). He found that the black lemur,
Eulemur macaco, eats the fruits of 70 species of tree. Undamaged seeds of
57 species were found in droppings; 29 species of tree were tested and
some seeds of all species germinated. Between 67 and 77% of trees in
Lokobe Forest were found to be dispersed by the black lemur. Most seeds
were deposited away from parent trees, and seed shadows were mostly het-
erogeneous. (Seed shadows are the area within which most of the seed of
an individual falls, effectively the population recruitment surface). Thus
Birkinshaw concluded that black lemurs are important seed dispersers. The
fruits dispersed by them tend to have a dull colour, thick husk and a rather
large size.
Table 1.1 shows the species richness of vertebrates (excluding fish) in ten
well-investigated tropical study sites (Bourliere, 1993).
Animals are also important in pollination and there is great variety in
taxonomic affinity, size and behaviour. Bees, wasps, Lepidoptera, Diptera,
Table 1.1 Species richness of vertebrates (fish excluded) in ten well-investigated tropical study sites (reproduced from Bourliere, 1983)
Mean Number Number Number Number

annual Surface of of of of
rainfall Altitude area mammal bird reptile amphibian
Study sites (mm) (m) (km2) species species species species
Lowland rain forest

Barro Colorado island, Panama 2600 164 14.8 97 366 68 32
Kartabo, Guyana 2500 10 0.6 73 464 93 37
Pasoh forest, west Malaysia 1900 72-150 7.8 212 >20 25
Bukit Langan forest, west Malaysia -2300 -40 6.4 >90 >119 50 23
Makokou area Gabon 1730 500 2000 119 342 63 38
Gogol Forest study sites, Papua New Guinea 3800 40-60 10.0 27 162 34 23
Analamazoatra (Perinet) forest, east Madagascar 1708 900 8.0 29 73 >47 54
Forest-savanna mosaic
Lamto, Ivory Coast 1300 110 2000 128 263 63 38
Savannas
Serengeti National Park, Tanzania 500-1000 1200-1700 25 000 130 381 61 23
Fete Ole Sahelian savanna, north Senegal 213 40 100 36 112 13 8
How serious is fragmentation? 11
birds (humming birds in the Neotropics), bats, marsupials, lemurs and
small monkeys are all involved.
It would be wrong to suggest that either the functionally important
organisms or the biological interest in tropical rain forest are all in the ver-
tebrate fauna. As Janzen (1983a) and others have indicated, about 12.5%
of the leaf machinery is consumed by invertebrates and only 2.4% by ver-
tebrates. Some of the most spectacular organisms within the forest are but-
terflies, and the caterpillars of moths are extremely important consumers
(Holloway, 1989). Lieth and Werger (1989) also have chapters on termites
and leaf-cutting ants.
1.5.5 Soils
Soils under tropical rain forests are many and varied, and have so far just
about defied classification. They are usually highly weathered with sub-
stantial quantities of sequioxides, mostly kaolisols and some podzols where
leaching is high. Here the horizons may be deeper than in temperate
regions (Baillie, 1996). Locally, after landslides or drastic human activity,
there may be immature soils, poorly drained ones in valley basins, and
montane ones at high altitudes (see Ross, this volume; Swaine, 1996).
1.6 WHY SO MUCH DIVERSITY?

The classic explanation for so many species is that the tropics have optimal
temperatures and thus high productivity; time-partitioning (day/night,
especially well-described for vertebrates but presumably equally important
to invertebrates) providing more niches; spatial partitioning; resource
(food) partitioning; mutualism and co-evolution; and stability over a long
time scale (although the latter point is debatable). The high level of diver-
sity of trees leads to greater diversity of epiphytes, lianas, pollinators, seed
predators and parasites. Individual pollinators co-evolve with individual
trees, for example. Such are the interdependencies between species that an
evolutionary move in one group is followed by an evolutionary move in
another, resulting in still more species. So we end up with a wide range of
functionally different groups including autotrophs, heterotrophs, parasites,
symbionts and saprophytes, and of species within such groups.
But here we meet another paradox: not all tropical rain forests are
species-rich. Some possible mechanisms for producing low-diversity rain
forests are discussed by Connell and Lowman (1989).
1.7 HOW SERIOUS IS FRAGMENTATION?

Wholesale destruction of rain forest is of course disastrous, but the loss of
extensive areas leaving small fragments also leads to the loss of local
species. Turner (1996), in a comprehensive essay review based on 21 stud-
ies, reports that the fragmentation of tropical rain forest presents a major
threat to local biodiversity, though the increased pressure on wildlife from
greater human activity in fragmented landscapes may make a major contri-
bution to the negative impact. Various mechanisms are responsible for the
local extinction of species in fragmented forest but it is not possible as yet
to quantify the relative importance of different factors such as restriction of
population size, forest edge effects or invasion of exotic species. Rare and
patchily distributed species, and those with requirements for a large range
or specialist habitats, seem particularly susceptible to fragmentation.
Animals that are large, or very specialized and intolerant of the vegetation
surrounding fragments, are also particularly prone to local extinction.
Turner (1996) goes on to say that it is imperative to prevent fragmenta-
tion if at all possible. Evergreen tropical forests are particularly sensitive to
fragmentation because a large portion of their species are both very
sparsely distributed through the community and also intolerant of open
sites. Thus, both the absolute number of species and the proportion of the
complete biota under threat from fragmentation are greater than for any
other biome. Nowhere else is there such a sharp distinction between highly
diverse primary forest and the species-poor early successional matrix.
However some areas of old secondary forest, especially where they are in
close proximity to primary forest, are known to be very species-rich. Africa
is different from South America and South-East Asia in that there is little
primary rain forest; it has nearly all been 'socialized', i.e. cleared or frag-
mented (see Sharpe, this volume).
Bierregaard et al. (1992) report on an experimental fragmentation of
forest 30-60 km north of Manaus, established by the WWF-Worldwide
Fund for Nature and Brazil's National Institute for Research in Amazonas
(INPA) in 1979. The Smithsonian Institution has also recently been heavily
involved. The objective was to determine the critical size of rain forest
ecosystem that should be left after clearance and to study edge effects. The
experimental cutting was not completed until 1990 and it may be too early
to judge the results conclusively. The cut area was converted to pasture but
some regrowth occurred. There were 12 fragments left of size 1-200 ha,
along with a 1000-ha control. The results are complex and should be
viewed as indicative rather than conclusive. A break of as little as 80 m can
have serious effects. With decreasing size of fragment, numbers of individu-
als decrease, species change, and ecosystem function changes. Tree mortal-
ity was greatest at the edges of the larger plots. The results suggest that if
200-ha remnants are left in future, as proposed, they would suffer signifi-
cant alteration due to edge effects. Another study suggests a further para-
dox: small, isolated fragments of lowland tropical rain forest can have a
high conservation value (Turner and Corlett, 1996). See too Laurance and
Bierregaard (1997).
Extinction rates have been modelled by Reid (1992) and co-workers for
Case study: Tiwai island 13
the World Resources Institute. Three scenarios were used: 5, 10 and
15 mlha/year deforestation between 1990 and two future dates, 2015 and
2040. The model was based on a species-area relationship with a z con-
stant of between 0.15 and 0.35. The percentage rates of extinction pre-
dicted ranged from 1-53%. There is still debate over whether this model is
appropriate: some people argue that the reality is likely to be worse than
this prediction, whereas others suggest the opposite.
1.8 QUANTITATIVE STUDIES IN RAIN FOREST

Greig-Smith (1971, 1996) provides a summary of vegetation studies in
tropical rain forests. Sample plots of relatively small size have been found
to be adequate (20X20 or 25x25 mi. In multivariate approaches such as
classification and ordination, less-common species can be eliminated from
the data without affecting the information return. Thus those species
which occur only once, or even a few times, can be deleted prior to analy-
sis. In one study a reduction from 818 to 65 large trees produced a fully
satisfactory classification. Thus, in terms of analyses, studies of rain forests
are no more difficult than those of temperate regions, although taxonomic
and logistical problems remain challenging.
1.9 CASE STUDY: TIWAI ISLAND

Tiwai is in Sierra Leone, West Africa, in a zone known as Upper Guinea
Forest. It is a tiny, riverine island only 11 km 2 on the Moa River, about
65 km from the sea. It is a sacred place of the Mende people, most of
whom are Moslem, and hunting pressures have been low. The primate
fauna is very rich, consisting of about 2000 individuals of 11 species,
including Camp bells, Diana, Spot-nosed and Green Monkeys;
Chimpanzees; Red, Black and White Colo bus and Prosimian Potto (Davies,
1989; Anglia TV, 1996).
Other animals are also spectacular, including birds such as the large
plantain-eater, mammals such as pangolin and pygmy hippo, palm civet,
African civet, brush-tailed porcupine, spot-necked otter and several species
of antelope (see Allport et aI., 1989 for an account of the birds of the
neighbouring Gola Forest).
The local Paramount chiefs on either side of the river are keen to see the
area conserved and developed for research and low-intensity tourism; how-
ever, the war has now spilled over from neighbouring Liberia, where mon-
keys are an item of diet. The future is dismal, if not desperate. With 97%
of the primary forest gone from Sierra Leone (this figure can be challenged
as most of the country has been 'socialized' for a very long period, see
Sharpe, this volume; Leach and Fairhead, 1996), the importance and value
of such remnants becomes overwhelming, but what can be their fate?
Possible solutions are the subject of much of the content of this book. The
answers lie in wider recognition of the importance and value of tropical
forest. We need to know more about the cultural and human values and
the edaphic benefits of protecting rain forest; also about policies and insti-
tutions and how they can be encouraged to provide an enhanced status,
sustainable use and effective long-term protection. These are the subjects of
subsequent chapters.
1.10 SKETCH FROM THE AMAZON

A Survival Special television programme (Anglia TV, 1996) showed the
wildlife around the confluence of the River Amazon and the Rio Negro in
Brazil. The links between different component species were also neatly
illustrated. For example, the agouti has teeth which can break the hard
shells of castanea nuts (Bertholletia excelsa), which we call Brazil nuts, but
the spectacularly spotted jaguar preys upon the agouti. The beautifully
spotted jaguar has been vulnerable to the whims of western fashion, but
fortunately this demand has recently decreased, the jaguar is making a
comeback and numbers are recovering. The Brazil nut tree is common and
widespread in the forest canopy and regenerates after disturbance of the
canopy. It has an unusual habit of concentrating strontium in the
endosperm; during the 1960s, after nuclear tests, seeds became perceptibly
radioactive.
Caciques (Icteridae) are sexually dimorphic colonial birds which build
their nests around wasp nests for protection. How do the wasps know that
they should not sting the caciques and do they derive any benefit in return
for the protection offered? The three-toed sloth is another bizarre occupant
of the forest. It hangs upside down most of the time in the forest, has
reverse hair direction to help shed water, and algae grow on its coat and
provide camouflage. Sloths are slow, obligate leaf-eaters but can swim well
in order to move from one area of forage to another.
The waters of the area are no less important. One-fifth of all fresh water
on the earth occurs in the Amazon Basin at the height of the rainy season.
The forest buffers its flow by absorbing it at times of high precipitation.
The waters of some rivers are nutrient-deficient and described as 'white',
whilst others are peaty and 'black'. Within the rivers, plants and detritus
from the forest feed fish which in turn feed the species of higher trophic
levels. Spectacular giant turtles weighing up to 23 kg (501b) are mostly
herbivorous and grow up to 1.5 m (5 feet) long. The seasonality of rainfall,
and consequent spate river flows, ensure that even here sandbanks occur
regularly and provide sites for turtles to lay their eggs. Young turtles are
the food of vultures and piranha. Anaconda are huge water snakes, and the
manatees are bizarre aquatic herbivores weighing up to 450 kg (1000 lb),
hunted for their tough skin which was used to make fan-belts.
Acknowledgements 15
Sadly the rivers are also the main highways for transportation, and the
only sewage system. Mercury used for extracting gold from ore, and diesel
leaks from the engines, fuel tanks and bilges of boats, pollute the rivers.
1.11 OTHER REGIONAL STUDIES

There have been several classic regional studies of tropical rain forest. A
few of these, together with selected key references, are listed below:
• Amazonas, Brazil (Pires and Prance, 1985)
• Mato Grosso, Brazil (Smith, 1971; Ratter et al., 1973, 1978)
• Maraca, Roraima, Brazil (Hemming et aI., 1988)
• Barro Colorado Island, Panama (Leigh and Windsor, 1983; Hubbell and
Foster, 1986)
• LaSelva, Costa Rica (Hartshorn, 1983; Janzen, 1983b; McDade et al.,
1994)
• Jamaica (Tanner, 1985)
• Korup, Cameroon (ODNRI, 1989)
• Okomo, Omo, Nigeria (Jones, 1955; Richards, 1996)
• Ghana (Swaine and Hall, 1983)
• Belalong, Brunei (Cranbrook and Edwards, 1994; Richards, 1996)
• Malaysia and Singapore (Ashton, 1969; Whitmore, 1984, 1991; Aiken
and Leigh, 1992)
1.12 CONCLUSIONS
In tropical rain forest it does not always rain and it is not always exciting.
The case studies selected here are from exceptional areas of rain forest
rather than typical ones. As indicated by Sharpe (this volume), much rain
forest in West Africa is probably secondary and people have had a consid-
erable impact even in the distant past. Old secondary rain forest, such as at
least part of Tiwai in Sierra Leone, can be very diverse. But this applies
only if there are long time intervals between fellings and recolonization can
occur from adjacent areas.
ACKNOWLEDGEMENTS
I would like to thank the late Paul Richards for arranging for me to see my
first area of rain forest as part of the Royal Society/Royal Geographic
Society Mato Grosso expedition in 1969. Since then I have enjoyed shorter
visits to Nigeria, Ghana, Sierra Leone, India and Australia. I would also
like to thank PhD students Jo Pires, Chris Birkinshaw and Joe Comiskey;
students on the Forest Monitoring Course at the Wildlife Institute of India
and the MSc Course in Conservation, Tropical Rain Forest Module
(Department of Biology, University College, London), for stimulating me to
think more about the real issues relating to tropical rain forests, and to
challenge some of the conventional wisdom.
I would also like to thank Drs Guy Cowlishaw, Barrie Sharpe and Gopal
Rawat for reading the manuscript and making useful comments.
REFERENCES
Aiken, S. R. and Leigh, C. H. (1992) Vanishing Rain Forests: the Ecological
Transition in Malaysia, Clarendon, Oxford.
Allport, G., Ausden, M., Hayman, P. V., Robertson, P. and Wood, P. (1989) The
Conservation of the Birds of the Gola Forest, Sierra Leone, ICBP Study Report
No. 39, BirdLife International (formerly International Council for Bird
Preservation), Cambridge.
Anderson, A. B., Magee, P., Gely, A. and Jardin, M. A. G. (1995) Forest manage-
ment patterns in the floodplain of the Amazon estuary. Conservation Biology, 9,
27-61.
Anglia TV, 1996. Island of the Apes. Survival Special Programme, 10 July 1996,
Anglia Television, Norwich.
Archibold, O. W. (1995) Ecology of World Vegetation, Chapman & Hall, London.
Ashton, P. S. (1969) Speciation among tropical forest trees: some deductions in the
light of recent evidence. Biological Journal of the Linnean Society, 1, 155-196.
Baillie, I. C. (1996) Soils of the humid tropics, in The Tropical Rain Forest, (ed. P.
W. R. Richards), Cambridge.
Bierregaard, R. O. Jr, Lovejoy, T. E., Kapos, V., dos Santos, A. A and Hutchings, R.
W. (1992) The biological dynamics of tropical rainforest fragments. Bioscience,
42, 859-866.
Birkinshaw, C. R. (1995) The importance of the black lemur, Eulemur macaco, for
seed dispersal in Lokobe Forest, Madagascar. PhD thesis, University College
London.
Bourliere, F. (1983) Animal species richness in tropical forest, in Tropical Rain
Forest Ecosystems, (ed. F. Golley), Elsevier, Amsterdam.
Brown, K. S. Jr. and Brown, G. G. (1992) Habitat alteration and species loss in
Brazilian forests, in Tropical Deforestation and Species Extinctions, (eds T. C.
Whitmore and G. Sayer), World Conservation Union (IUCN)/Chapman & Hall,
London,pp.119-142.
Bullock, S. H., Mooney H. A. and Medina, E. (eds) (1995) Seasonally Dry Tropical
Forests, Cambridge University Press, Cambridge.
Champion, H. G. and Seth, S. K. (1968) General Silviculture for India,
Government of India, Delhi.
Connell, J. H. and Lowman, M. D. (1989) Low diversity tropical rainforests: some
possible mechanisms for their existence. American Naturalist, 134, 88-119.
Cranbrook, Earl and Edwards, D. S. (1994) A Tropical Rain Forest, Royal
Geographical Society, London.
Davies, A. G. (1989) The Gola Forest Reserves, Sierra Leone; Wildlife
Conservation and Management, World Conservation Union (IUCN), Gland and
Cambridge.
References 17
Davies, A. G. and Oates, J. F. (1994) Colobine Monkeys - Their Ecology,
Behaviour and Evolution, Cambridge University Press, Cambridge.
Evans, J. D. (1984) Plantation Forestry in the Tropics, Oxford University Press,
Oxford.
Golley, F. B. (ed.) (1983) Tropical Rain Forest Ecosystems: Structure and Function,
Elsevier Ecosystems of the World No. 14A, Elsevier, Amsterdam.
Gomez-Pompa, A., Whitmore, T. C. and Hadley, M. (eds) (1991) Rain Forest
Regeneration and Management, Man and the Biosphere Series, Vol. 6, UNESCO,
ParislParthenon, Carnforth, pp. 91-118.
Greig-Smith, P. (1971) Application of numerical methods to tropical forests, in
Statistical Ecology 3, (eds G. P. Patil, E. C. Pielow and W. E. Waters),
Pennsylvania State University Press, Philadelphia, pp. 195-206.
Greig-Smith, P. (1996) Application of numerical methods in rain forest, in The
Tropical Rain Forest, (ed. P. W Richards), Cambridge University Press,
Cambridge.
Hall, J. B. and Swaine, M. D. (1980) Seed stocks in Ghanaian forest soils.
Biotropica, 12, 256-263.
Halle, F., Oldeman, R. A. A. and Tomlinson, P. B. (1978) Tropical Trees and
Forests: An Architectural Analysis, Springer-Verlag, Berlin.
Hartshorn, G. S. (1983) Plants - introduction, in Costa Rican Natural History, (ed.
D. A. Janzen), University of Chicago Press, Chicago, pp. 118-157.
Hemming, J., Ratter, J. and dos Santos, O. A. (1988) Maraca, Edicoes Artisticas,
Brazil.
Holloway, J. D. (1989) Moths, in Tropical Rain Forest Ecosystems:
Biogeographical and Ecological Studies, (eds H. Lieth and M. J. A. Werger),
Elsevier Ecosystems of the World No 14B, Elsevier, Amsterdam.
Hopkins, B. (1974) Forest and Savanna, Heinemann, London.
Hubbell, S. P. and Foster, R. B. (1986) Canopy gaps and the dynamics of a neotrop-
ical forest, in Plant Ecology, (ed. M. J. Crawley), Blackwell, Oxford.
Janzen, D. H. (1975) Ecological Studies in the Tropics, Institute of Biology Studies
in Biology No. 58, Edward Arnold, London.
Janzen, B. J. (1976) Seedling patterns in tropical trees, in Tropical Trees as Living
Ecosystems, (eds P. B. Tomlinson and M. H. Zimmermann), Cambridge
University Press, Cambridge, pp. 83-128.
Janzen, D. H. (1983a) Food webs: who eats what, why, how, and with what effects
in a tropical forest?, in Tropical Rain Forest Ecosystems: Structure and Function,
Elsevier Ecosystems of the World No. 14A, (ed. Golley, F. B.), Elsevier,
Amsterdam, pp. 167-182.
Janzen, D. H. (1983b) Costa Rican Natural History, University of Chicago Press,
Chicago.
Johns, A. D. (1992) Species conservation in managed tropical forests, in Tropical
Deforestation and Species Extinctions, (eds T. C. Whitmore and G. Sayer),
World Conservation Union (IUCN)/Chapman & Hall, London, pp. 15-53.
Jones, E. W. (1955) Ecological studies on the rain forest of southern Nigeria.
Journal of Ecology, 43, 564-594.
Kricher, J. C. (1989) A Neotropical Companion, Princeton University Press,
Princeton, 436pp.
Laurance, W. F. and Bierregaard, R. O. (1997) Tropical Forest Remnants,
University of Chicago Press, Chicago, 616pp.
Leach, M. and Fairhead, J. (1996) Misreading the African Landscape: Society and
Ecology in a Forest-Savanna Mosaic, Cambridge University Press, Cambridge.
Leigh, E. G. Jr and Windsor, D. M. (1983) Forest production and regulation of pri-
mary consumers on Barro Colorado Island, in The Ecology of a Tropical Forest:
Seasonal Rhythms and Long-Term Changes, (eds E. G. Leigh Jr, A. S. Rand and
D. M. Windsor), Oxford University Press, Oxford, pp. 111-122.
Lieth, H. and Werger, M. J. A. (eds) (1989) Tropical Rain Forest Ecosystems:
Biogeographical and Ecological Studies, Elsevier Ecosystems of the World No.
14 B, Elsevier, Amsterdam.
Mabberley, D. J. (1992) Tropical Rain Forest (3rd edn), Blackie, Glasgow.
McDade, L. A., Bawa, K. S., Hespenheide, H. A. and Hartshorn, G. S. (1994) La
Selva: Ecology and Natural History of Neotropical Rain Forest, University of
Chicago Press, Chicago.
Meggers, B. J., Ayensu, E. S. and Duckworth, W. D. (1973) Tropical Forest
Ecosystems in Africa and South America: A Comparative Review, Smithsonian,
Washington, DC.
Myers, N. (1979) The Sinking Ark: A New Look at the Problem of Disappearing
Species, Pergamon, Oxford.
Myers, N. (1989) Deforestation Rates in Tropical Forests and their Climatic
Implications, Friends of the Earth, London.
Odum, H. T. (ed.) (1970) Tropical Rain Forest, US Atomic Energy Commission,
Oak Ridge, Tennessee.
ODNRI (1989) Korup Management Plan, Overseas Development Natural
Resources Institute, Chatham, Kent.
Peres, C. A. and Terborgh, J. W. (1995) Amazonian nature reserves: an analysis of
the defensibility status of existing conservation units and design criteria for the
future. Conservation Biology, 9, 34-46.
Perry, D. A. (1995) Forest Ecosystems, Johns Hopkins University Press, Baltimore.
Pires, J. M. and Prance, G. T. (1985) The vegetation types of the Brazilian Amazon,
in Key Environments: Amazonia, (eds G. T. Prance and T. E. Lovejoy),
Pergamon, Oxford, pp. 109-145.
Pires-O'Brien, M. J. and O'Brien, C. M. (1995) Ecologia e Modelamento de
Florestas Tropicas, FCAP Servico de Documentacao e Informacao, Belem.
Prance, G. T. (1982) Forest refuges: evidence from woody angiosperms, in
Biological Diversification in the Tropics, (ed. G. T. Prance), Columbia University
Press, New York, pp. 137-156.
Prance, G. T. (1989) American tropical forest, in Tropical Rain Forest Ecosystems,
(eds H. Lieth and M. J. A. Werger), Elsevier, Amsterdam, pp. 99-132.
Prance, G. T. and Campbell, D. G. (1988) The present state of tropical floristics.
Taxon, 37, 519-548.
Ratter, J. A., Richards, P. W., Argent, G. and Gifford, D. R. (1973) Observations on
the vegetation of northeastern Mato Grosso. 1. The woody vegetation types of
the Xavantina-Cachimbo expedition area. Philosophical Transactions of the
Royal Society of London, Series B, 266, 449-492.
Ratter, J. A., Askew, G. P., Montgomery, R. F. and Gifford, D. R. (1978)
Observations on the vegetation of northeastern Mato Grosso. II. Forests and
References 19
soils of the Rio Suia-Missu area. Proceedings of the Royal Society of London,
Series B, 203,191-209.
Reid, W. V. (1992) How many species will there be?, in Tropical Deforestation and
Species Extinctions, (eds T. C. Whitmore and G. Sayer), World Conservation
Union (IUCN)/Chapman & Hall, London, pp. 55-73.
Richards, P. W. (1996) Tropical Rain Forest, 2nd. edn, Cambridge University Press,
Cambridge.
Sayer, J. A. and Whitmore, T. C. (1990) Tropical moist forests: destruction and
species extinction. Biological Conservation, 55, 199-214.
Simberloff, D. (1992) Species-area relationships, fragmentation and extinction in
tropical forests. Malaysian Nature Journal, 45, 398-413.
Smith, A. (1971) Mato Grosso: Last Virgin Land, Michael Joseph, London.
Smuts, B. B. (1987) Primate Societies, Chicago University Press, Chicago.
Sutton, S. L., Whitmore T. C. and Chadwick A. C. (eds) (1983) Tropical Rain
Forest Ecology and Management, Blackwell, Oxford.
Swaine, M. D. (1996) Rainfall and soil fertility as factors limiting forest species dis-
tributions in Ghana. Journal of Ecology, 84,419-428.
Swaine, M. D. and Hall, J. B. (1983) Early succession on cleared forest land in
Ghana. Journal of Ecology, 71, 601-627.
Swaine, M. D., Hall, J. B. and Lock, J. M. (1976) The forest-savanna boundary in
west-central Ghana. Ghana Journal of Science, 16, 35-52.
Symington, C. F. (1974) Foresters' Manual of Dipterocarps, Malayan Forest
Records No. 16, Penerbit Universiti Malaya, Kuala Lumpur.
Tanner, E. V. J. (1985) Jamaican montane forests: nutrient capital and cost of
growth. Journal of Ecology, 73, 553-568.
Turner, I. M. (1996) Species loss in fragments of tropical rain forest: a review of the
evidence. Journal of Applied Ecology, 33, 200-209.
Turner, I. M. and Corlett, R. T. (1996) The conservation value of small, isolated
fragments of lowland tropical rain forest. Trends in Ecology and Evolution, 11,
330-333.
UNESCO (1978) Tropical Forest Ecosystems: A State of Knowledge,
UNESCO\UNEP\FAO Natural Resources Research No. 14. UNESCO, New York.
Walsh, R. P. D. (1966) Climate, in The Tropical Rain Forest, (ed. P. W R.
Richards), Cambridge University Press, Cambridge.
Walter, H. W. (1971) Ecology of Tropical and Subtropical Vegetation, Oliver and
Boyd, Edinburgh.
Waring, R. H. and Schlesinger, W H. (1985) Forest Ecosystems: Concepts and
Management, Academic Press, Orlando.
Whitmore, T. C. (1983) Secondary succession from seed in tropical rain forests.
Forestry Abstracts, 44, 767-779.
Whitmore, T. (1984) Tropical Rainforests of the Far East, 2nd edn, Clarendon
Press, Oxford.
Whitmore, T. C. (1990) An Introduction to Tropical Rain Forest, Oxford.
Whitmore, T. C. (1991) Tropical rain forest dynamics and its implication for man-
agement, in Rain Forest Regeneration and Management, (eds A. Gomez-Pompa,
T. C. Whitmore and M. Hadley), Man and the Biosphere Series, Vol. 6,
UNESCO, ParislParthenon, Carnforth, pp. 91-118.
Whitmore, T. C. and Sayer, J. (eds) (1992) Tropical Deforestation and Species
Extinctions, World Conservation Union (IUCN)/Chapman & Hall, London.
Wyatt-Smith, J. (1963) Manual of Malayan Silviculture for Inland Forests (2 vols),
Malayan Forest Records No. 23, Forest Research Institute, Malaysia.
-2------
Indigenous non-timber benefits from
tropical rain forest
G. T. Prance
2.1 INTRODUCTION
In the rush for commercial exploitation of rain forest areas, one aspect that
is often neglected as timber is extracted or forest is felled for agricultural
land is the multitude of non-timber forest products that are available. Non-
timber forest products (NTFPs) have been a vital resource for the survival
of forest dwellers, whether indigenous or settlers, for many generations. It
is the NTFPs that provide food, medicine, fibres, arrow poisons, shelter
and many other essentials to life. As indigenous peoples have come into
contact with the more developed world and as settlers have occupied the
rain forest, NTFPs have entered the economy in a large way. They already
provide considerable income to forest peoples, and even more to those
people who prey upon them. For example, de Beer and McDermott (1989)
state that 29 million forest dwellers in South-East Asia depend on NTFPs
for their livelihood. It is therefore important to consider the potential of
NTFPs in this volume, and to discuss their role in the sustainable manage-
ment and conservation of rain forest.
2.2 INDIGENOUS USE OF THE FOREST

Most of the NTFPs that are traded today have entered into the market
because they were originally put to use by indigenous peoples, but not nec-
essarily for the same thing. For example, Amazonian arrow poisons have
become the medicine curare. Recent studies in quantitative ethnobotany,
for example by Boom (1985); Prance et al. (1987) and Toledo et al. (1995),
have shown the extent to which indigenous peoples make use of the rain
forest.
22 Indigenous non-timber benefits from tropical rain forest
Table 2.1 Number and percentage of useful species and individuals (10 cm or
more dbh) of the Shuar Indians, arranged in types of use, for a I-ha plot at the
Jatan Sacha Biological Reserve in Amazonian Ecuador (from Bennett, 1992a)
Species Individuals
Use category Number Percentage Number Percentage
Construction 191 77.0 637 87.4

Crafts and 7 2.8 29 4.0
dye/paint
Fibre 1 0.4 1 0.1
Fishing and 3 1.2 11 1.5
hunting
Food 56 22.6 278 38.1
Food processing 2 0.8 10 1.4
Fuel 59 23.7 270 37.0
Medicine 17 6.9 215 29.5
Ornamental 0 0 0 0
Personal 1 0.4 14 1.9
Ritual/Mythical 5 2.0 13 1.8
Veterinary 0 0 0 0
Miscellaneous 12 4.8 51 7.0
The Chacobo Indians of Bolivia have uses for 82 % of the species and
92 % of the individual trees of 10 cm or more in diameter on a sample
hectare studied by Boom (1989). Similarly the Ka'apor people of eastern
Amazonian Brazil use all 136 trees and lianas of over 10 cm diameter in
the 1-ha plot studied by Balee (1986), and the Pan are Indians of Venezuela
use 48.6% of the species on a hectare studied by Boom (1990). Bennett
(1992a) found that the Quijos Quichua of Ecuador use over 90% of the
species and 96% of the individual trees on a number of different 1-ha plots
which he sampled.
Table 2.1 shows the major use categories of the forest products used by
the Shuar Indians of Amazonian Ecuador as recorded by Bennett (1992a).
It shows a wide range of uses, and that these forest products are vital to
their lives. The studies cited above have recorded only the use of trees of
over 10 cm diameter at breast height. All indigenous peoples also use many
of the other plants, the herbs, shrubs, small trees, epiphytes and lianas.
Bennett (1992b) also recorded the uses of epiphytes, lianas and parasites
by the Shuar.
Toledo et al. (1995) compiled a data bank of plant uses by the indigen-
ous peoples of Mexico. The 1330 species catalogued are used for a consid-
erable variety of applications (Table 2.2).
These data are sufficient to show that indigenous peoples use a large
number of the plant species around them for a wide range of uses. It is not
Indigenous management of plant resources 23
Table 2.2 Use of products by indigenous groups of
Mexico from a data bank of 1330 species (From
Toledo et at., 1995)
Use category Number of species
Medicinal 780
Food 360
Construction 175
Timber 102
Fuel 93
Nectar for honey 84
Forage 73
Domestic use 69
Crafts 59
Poisons 52
Work items 51
Ritual use 50
Fibres 38
Colour and dyes 34
Shade 31
Gums 20
Flavours 24
Stimulants 16
Fertilizers 15
Living fences 15
Insecticides 11
Tannins 12
Aromatics 11
Chicle gums 10
Saponins 10
'Tutor' 10
surprising therefore that so many of their uses have been more widely
applied to a market economy. Often the intermediate stage has been the use
of these products by settlers into a region. These people have learned from
the indigenous peoples and have used many plants in the same way or have
adapted them for uses more appropriate to their life style and to a certain
amount of trading activity, such as, in Amazonia, the extraction and mar-
keting of rubber latex and Brazil nuts.
2.3 INDIGENOUS MANAGEMENT OF PLANT RESOURCES

The lessons on the use of NTFPs from both indigenous and local peoples
are not just connected with the listing of uses and their adaptation. One of
the most important aspects is the management of these resources to provide
continuity or sustainability. The study of the ethnoecology of indigenous
and local peoples is providing many insights into both management and
mismanagement of rain forest resources. Examples of studies of manage-
ment systems include the Huastec Mayans (Alcorn, 1984a, b), the Bora
Indians of Peru (Denevan et al. 1984, 1985; Denevan and Padoch, 1988),
and the Kayapo Indians of Brazil (Hecht and Posey, 1979; Posey, 1982,
1983, 1984). In all these cases it has been shown that there is considerable
management of NTFPs and that the cultivation of many of these products
has been incorporated into their agroforestry systems. Figure 2.1 is an
example from the Bora Indians of Peru (Denevan and Padoch, 1988).
Similar systems have been incorporated by local peoples. For example, the
agroforestry systems described for local riberenos in Peru (Padoch et aI.,
1985; Padoch, 1990) are very similar to those of the Bora Indians. These
were also described in some detail by Hecht (1982). Figures 2.2-2.13 show
examples of some NTFPs and their extraction by local people.
2.4 EXTRACTIVE RESERVES

The attention given to NTFPs has led to the concept of extractive reserves,
particularly in Amazonian Brazil (Schwarztman and Allegretti, 1987;
Fearnside, 1989; Prance, 1989; Schwartzman, 1989; Allegretti, 1990).
These are reserves where the local peoples are allowed to extract products
from the forest (usufruct) but not to clear-cut except for small areas of
agricultural crops to sustain them (up to 5% area only). Table 2.3 lists the
principal extractive reserves of Amazonian Brazil.
There is no doubt that the creation of these reserves, due mainly to
grass-roots pressure from local communities, has been useful to slow down
deforestation, especially in the Brazilian states of Acre and Rondonia.
However, various authors (e.g. Browder, 1990, 1992; Homma, 1994;
Prance, 1994) have shown that there are severe limits to the concept of
extractivism as a permanent solution for the sustain able use of the
Amazon rain forest. So far most of the extractive reserves have depended
upon the gathering of rubber latex and Brazil nuts, the price of both of
which has fluctuated greatly in the world market. There are many other
extraction products that are used to a lesser extent, as shown in Tables 2.4
and 2.5, and sustainable extractivism will work only if a greater variety of
products are used.
One of the major criticisms of extraction is that it generates an extremely
low income and is therefore perpetuating a low standard of living for the
extractivists. Is the promotion of extractivism holding some people in peonage?
The poverty of most extractivists is increased by the fact that they do not
have direct access to the market. In most parts of the world where extrac-
tivism is important there is a long chain of intermediaries between the pro-
ducer and the market. It is these people who became rich, not the forest
dwellers. This was clearly demonstrated in the study by Padoch (1992) and
Extractive reserves 25
¥
•
Banana
Peach palm
~
j
~ Uvilla
'( , ;(II,
Cf Caimito band _- .~
.r§' -'( - - - ~ - ~
~ Macambo
+
~'I>~ _ .... "
.,...
tfP.'. Fruit-pineapple $
+ Guaba ,'0$ - - ,~; '~~----- v
y +
Gl Barbasco '~ .;- Cf:e::. ,~~ ,-,\ I
d; ,r T T ' , ~,/ T I .,.. )
~ Peanuts Jf , Fruit-,,~,
'<J
/ Fruit fi3/ /
• ~ -' _
Y ,'pineapple,. :::',~T 1jI),/_ 'Y
~ Pineapple
,~
e> \""','
_
'T" ~ " - eO'
• - ,.< - ~ _ 'r"-es~
,{l\
'¥'
'( Manioc If) -~ 9J "';/Fruit T ~;~e~~ ~
'T"
r Coca
_.
~ ~.- ~ - - - ~_1' ~/ ~ - 'f
"i'Plneapple:/: --~--: ",JQJanioc rl1I
$ Forest -. __ . Coca, y <l<>
(a) _._._._._._._._._. Trail
~ Weedy plant -2700 m2 -'-'--'-
(j Avocado
,.. Old pineapple
Trail
~
6-year field
) Trail
Original field
perimeter-
Old fallow
30 years
Old fallow
30 years
(c)
(b) -3500 m2
Cleaned fruit zone: -700 m2

-2100 m2
Figure 2.1 Stages of regeneration of Bora Indian fields after felling. There is a
gradual transition from their agroforestry system to a managed secondary forest.
(Diagrams from Denevan et al., 1984.)
Table 2.3 Extractive reserves in Amazonian Brazil (source IBAMNCNPT, 1994)
Name State Area (ha) Population Principal products
Alto Jurua Acre 506 186 6000 Rubber

Chico Mendes Xapuri Acre 970570 7500 Rubber
Brazil nut
Copaiba oil
Rio Cajari Amapa 481650 5000 Brazil nut
Rubber
Copaiba oil
A<;:af palm
Rio Ouro Preto Rondonia 204583 3410 Brazil nut
Copaiba oil
Rubber
Ciriaco Maranhao 7050 1150 Baba<;:u
Mata Grande Maranhao 10480 1500 Baba<;:u
Fish
Quilomobe de Freixal Maranhao 9542 900 Baba<;:u
Fish
N. Tocantins Tocantins 9280 2000 Baba<;:u
Fish
Table 2.4 Extraction and agroforestry products from native plants which enter
the Brazilian economy (Source IBGE, 1992)
Production (tonnes)
Value 1988
Product 1988 1989 (Cz$)
Vegetable oils
Pequi (Caryocar brasiliensis) 1394 1593 785221
Tucum (Astrocaryum tucuma) 5109 5092 273490
Baba<;:u (Attalea phalerata) 200031 195 378 14004499
Copaiba (Copaifera spp.) 54 49 14780
Cumaru (Dipteryx odorata) 15 9 3603
Licuri (Syagrus coronata) 3632 12 421 1196 527
Oiticia (Licania rigida) 10277 15968 332641
Andiroba (Carapa guianensis) 363 38730
Fibres
Pia<;:ava (Leopoldinia) 959 1444 47326
Other Fibres 2023 364 139363
Continued
Table 2.4 Continued
Production (tonnes)
Value 1988
Product 1988 1989 LCz$2
Tanins
Angio (Piptadenia peregrina) 1557 1185 59014
Barbatimao (Stryphnodendron barbatiman 1527 1387 32508
Others 19 15 1502
Latex
Caucho (Castilla ulei) 39 42
Rubber (Hevea spp. coagulant) 23035 22990 171 488
Rubber (Hevea spp. liquid) 2409 1784 743454
Mangabeira (Hancornia speciosa) 2 1 55
Balata (Manilkara bidentata) 21 21 4938
Ma..aranduba (Manilkara huberi) 192 127 40531
Sorva (Couma spp.) 1059 1106 63980
Food and beverage
A.. ai (Euterpe oleracea fruit) 117119 114304 16623748
Cashew (Anacardium occidentale) 12 716 8870 1 775592
Brazil nuts (Bertholletia excelsa) 23391 25672 1351852
Erva mate (!lex paraguariensis) 145064 145649 20023399
Mangaba fruit (Hancornia speciosa) 939 988 100343
Palmito (Euterpe spp. other palms) 190314 202439 12 230169
Pinhao fruit 3118 2919 313 648
Umbu fruit (Phytolacca dioica) 19555 18999 1638624
Wax
Carnauba wax (Copernicia cerifera) 7373 7372 2730850
Carnauba powder 10734 11011 3349219
Fibres
Buriti (Mauritia flexuosa) 972 991 34116
Carnauba (Copernicia cerifera) 2544 2876 96 597
Medicines
Ipecacuana (Cephaelis) 17 3 3196
Jaborandi (Pilocarpus spp.) 1765 1676 244793
Colourant
Urucu (Bixa orellana) 793 845 71497
Miscellaneous
Other minor products 2388 934 70787
Table 2.5 Some important Amazonian products that are marketed but do not
enter the official economy, with further development potential
Common name Species name
Fruit
Abiu Pourouma caimito
Ara..a-boi Eugenia stipitata
Ata Annona squamosa
Bacuri Platonia insignis
Bacupari Rheedia spp.
Biriba Rollinia mucosa
Camu-camu Myrciaria dubia
Cupua.. u Theobroma grandi{lorum
Cubiu Solanum sessili{lorum
Graviola Annona muricata
Macambo Theobroma bicolor
Pupunha Bactris gasipaes
Sacha Manga Grias peruviana
Sapota Matisia cordata
Sapoti Manilkara sapota
Tapereba Spondias mombin
Ucuqui Pouteria ucuqui
Umari Poraqueiba sericea
Uvilla Pourouma cecropiifolia
Oils
Castanha de Cutia Acioa edulis
Castanha de galinha Couepia longipendula
Curupira Curupira tefeesis
Other Products
Tagua vegetable ivory Phytelephas macrocarpum
Titara cana fibre Desmoncus spp.
Aruma cane fibre Ischnosiphon aruma
Padoch and de long (1989) of the markets of Amazonian Peru, and by

Peluso (1992) in a study of the exploitation of rattan cane in East
Kalimantan, Indonesia. There have been some commendable efforts to
shorten this chain of intermediaries, such as that of Cultural Survival
Enterprises setting up a Brazil nut-processing plant operated by local
extractivists (Clay, 1992) and the Body Shop Company buying Brazil nut
oil directly from the Kayap6 Indians for use in their cosmetics.
Another factor is that extractivism is often the pathway to cultivation.
This in itself is not necessarily a bad thing, but it does cause one to ques-
tion the longer-term sustainability of extractive reserves. For example,
cocoa (Theobroma cacao) and guarana (Paullinia cupana var. sorbilis)
Figure 2.2 Indigenous peoples such as this Paumari Indian use many of the plants
in the forest; the stem fibre of this Ischnosiphon aruma is used to make baskets.
both began as Amazonian extraction products and are now major crop
plants in the region. Extraction and the experimentation which it fosters is
one of the ways in which new crops are introduced to the market.
Table 2.3 clearly indicates a further reason for caution about extraction,
which is that it requires a large area to sustain each person. The much-discussed
Figure 2.3 Rubber from Hevea spp. has for many years been the mainstay of
extraction. Much of it is still coagulated using acid smoke.
Figure 2.4 Brazil nut gatherer chopping open the fruit to extract nuts.
Figure 2.5 View of a Brazil nut factory.
Chico Mendes extractive reserve at Xapuri has 7500 people on

970 570 ha, i.e. 129 ha per person. Agricultural and agroforestry systems
require much less space per person. As pressure on land increases it is
unlikely that any individual will be allowed to use 129 ha of land.
Another important factor is that the extraction of forest products seldom
yields enough cash to entirely support the extractivist. Even the occupants
Figure 2.6 A little-exploited fruit, Annona squamosa, the ata.
of extractive reserves are allowed to use 5% of the area for agriculture.

Most of the indigenous people who extract so much from the forest are
also agriculturalists. Their crops of cassava or bananas are as important as
forest products for their survival.
2.5 SUSTAINABILITY
Some forms of extraction, such as the gathering of Brazil nuts and the tap-
ping of rubber trees, are sustainable and do little damage to the forest.
However, the extraction of other NTFPs can be predatory. Rosewood
(Aniba rosaeodora) is almost extinct because every tree is cut to distil the
essential oil linalol. Many species of palms have been over-exploited to
gather palm hearts as a vegetable. Sometimes unsustainable harvesting
practices are adopted both by indigenous and local peoples. There are
Sustainability 33
Figure 2.7 An oligarchic forest of babar,;u, Attalea phalerata.
Figure 2.8 Local people extracting the kernel of babar,;u to obtain the oil.
Figure 2.9 Local boy on Maraj6 Island extracting the heart of palm from a~ai
(Euterpe oleracea).
Figure 2.10 A local heart-of-palm processmg plant on Maraj6 Island, Para,

Brazil.
Sustainability 35
Figure 2.11 The fruit of Carapa guianensis, a timber tree which also yields a
useful oil.
many stands of the dioecious aguaje palm (Mauritia f/exuosa) in western

Amazonia which contain almost exclusively male trees. This is because in
order to gather the fruit, the female, fruit-bearing trees have often been
felled rather than climbed. The latex of some trees is also extracted
through felling, for example, sorva (Couma spp.) and ma~aranduba
(Manilkara spp.). Until methods are developed to extract these latexes
without felling, their extraction is unsustainable. In East Kalimantan,
Indonesia, trees of Litsea are felled to extract a resin used in mosquito-
repellent coils (Salafsky et aI., 1993).
The harvesting of the fruits of a species can lead to reduction in the num-
ber of individuals because it inhibits the establishment of new plants.
Sustainable extraction of fruits requires studies of the recruitment of new
individuals into the population and knowledge of their reproductive biol-
ogy. Much existing extraction is leading to depletion of the very resource
upon which it depends. This fact has also led more rapidly to cultivation of
some products. One example where the sustainable harvest of fruits has
been taken into consideration is a study by Peters and Hammond (1990)
on three fruit species of the Peruvian Amazon: the sacha mangua (Grias
peruviana), camu camu (Myciaria dubia) and breadnut (Brosimum alicas-
trum). Peters and Hammond make recommendations about the amount of
Figure 2.12 Guarana, Paullinia cupana var. sorbilis, is an example of an extrac-

tion product that has now become an important cultivated crop as the source of a
caffeinated beverage.
fruit that could be harvested without altering the population structure of

these species.
In many extraction forests, hunted animals form an important part of
the diet of extractivists, and in some cases animals are also part of the com-
mercial side of extraction. Many game animals, for example deer and paca,
eat the same fruits that are extracted, so excess harvesting may have an
adverse effect on game.
There are relatively few examples of sustainable extraction of forest
products. Extraction has been popular with environmentalists in Amazonia
because the gathering of the two most-used products, rubber and Brazil
nuts, is sustainable if correctly managed. Most Amazonian rubber tappers
know how to tap a tree without causing it to die. The extraction of the
leaves of the date palm (Chamaedora spp.) from the forests of Guatemala
for ornamental purposes appears to be sustainable (Heinzman and
Reining, 1988; Salafsky et al., 1993).
2.6 THE CONVENTION ON BIOLOGICAL DIVERSITY

Many large companies are using indigenous knowledge derived from their
use of NTFPs to search for commercial products, especially pharmaceuticals
and agrochemicals. Historically indigenous and local peoples have gained
The convention on biological diversity 37
Figure 2.13 Cupua<;u, Theobroma grandiflorum, is both an extraction product

and a commonly cultivated tree for the flavourful pulp which is used for ice creams,
drinks and desserts.
nothing from this use of their knowledge. What tribal person has gained
from the profits made out of curare or pilocarpine? The establishment of
the Convention on Biological Diversity at the Rio de Janeiro Earth Summit
of 1992 should address this important issue. It is dubious whether extrac-
tion of NTFPs will sustain or provide a greater standard of living for
extractivists. However, if royalties were to be channelled to their benefit
from a biochemical product they could be greatly helped. Some botanical
gardens, such as the New York Botanical Garden and the Royal Botanic
Gardens at Kew, were conducting their chemical work in a way that would
provide royalties to the originators of information long before the
Convention on Biological Diversity was drafted, and it is to be hoped that
some commercial products will emerge from their research. It is most
important that both the use of forest products by industry and the
exploitation of biological information gained from local peoples should be
to the benefit of local peoples. The equitable sharing of the profits gained
from NTFPs is an area that requires much more study and action.
2.7 OLIGARCHIC FORESTS

Most of the data on extraction are based on studies of species-diverse rain
forest and most extractive reserves have been set up in that habitat. There
are however, even within the Amazonian rain forest region, certain areas
where for environmental reasons of ecological stress the forest is much less
diverse. In a few cases almost pure stands of a species occur. These were
termed oligarchic forests by Peters et al. (1989). In many cases oligarchic
forests are composed of particularly useful species. Because of the density
of individuals and the consequent ease of harvest, these are areas where
extraction can work on a long-term sustainable basis and economic returns
can be high. Some examples of oligarchic forests are given below.
The best known oligarchic forests are the large stands of the baba\u
palm (Attalea phalerata) in the transition zone between the Amazon rain
forest and the cerrado of central Brazil. Some of these stands are natural,
but the area of baba\u forest has greatly increased as forest is felled. Since
baba\u is fire-resistant it is able to survive where other species die out. This
palm has been put to many uses (Balick, 1987; May, 1990) especially for
its comestible oil that is similar in fatty acid constituents to coconut oil,
and as a source of charcoal. Four of the extractive reserves listed in Table
2.3 exist mainly for the exploitation of baba\u. These reserves have a high
possibility of sustainability because they are based on a species which
occurs in an oligarchic forest.
Camu-camu (Myciaria dubia) is a common flood-plain shrub that can
withstand up to 10 months of flooding. Few other species can grow in this
habitat and consequently camu-camu occurs in almost pure stands along
many lakes and rivers, especially in Amazonian Peru. It has an edible fruit
with a growing market because of its exceptionally high quantity of vita-
min C. The ecology of this species was studied in detail by Peters (1990)
and Peters and Hammond (1990) who showed that it can easily be har-
vested in a sustainable manner.
There are several other examples of oligarchic forests in Amazonia, par-
ticularly of palm species such as the aguaje palm swamps (Mauritia f/exu-
osa) and the large patches of a\ai (Anderson, 1988, 1990). A good,
detailed account of Amazonian oligarchic forests is given by Peters and
Hammond (1990). The forests should be much further researched and
given high priority for use because of the likelihood of their sustainability
under exploitation of their products.
Conclusions 39
2.8 TIMBER AND NON-TIMBER FOREST PRODUCTS
Certain timber trees can also be the source of NTFPs, and so it is some-
times quite possible to combine the production of timber and NTFPs. For
example the Meliaceae species Carapa guianensis or andiroba produces a
fine quality timber. The fruit of this species is also harvested by local people
for its medicinal oil. It would therefore be possible to exploit the fruit from
plantations until the timber is ready for harvest. Various indigenous and
local systems of management do this for various species of tree. At
Tamshiyacu in Peru the agroforestry system includes Brazil nut trees, some
of which are harvested for their timber at the end of each 35-year cycle
(Padoch et aI., 1985). One of the most popular fruit trees, umari
(Poraqueiba sericea), is finally harvested as a source of charcoal. It is there-
fore important not to place timber and non-timber forest products in two
completely separate categories.
Extractive reserves are likely to be of far greater long-term viability if the
extraction of NTFPs is combined with the sustainable management and
extraction of timber. As with NTFPs the key to success is the level at which
the resource is exploited. Knowledge of the population ecology is equally
important for the management of timber and NTFP species.
A good example of the combination of the use of timber with NTFPs is
given by Salick et al. (1995) for the forests of Rio San Juan, Nicaragua.
This appears to work well there, but the forest has the advantage of being
far less species-diverse than most of those in Amazonia. A compelling case
for the combined management of timber and NTFPs in the forests of tropi-
cal Asia is also made by Panayotou and Ashton (1993).
2.9 CONCLUSIONS
Although the rain forest offers a cornucopia of non-timber forest products,
their sustainable use is not easy. The use of many products leads to their
depletion because people are not able to maintain harvesting at sustainable
levels. Local peoples are often more exploitative because they are not gov-
erned by the taboos and conventions that often produce rigid rules about
the exploitation of species of both plants and animals.
The demand for popular products usually leads either to their extermin-
ation or to production in cultivation. Most extraction has been carried out
inequitably with the benefits going to a long chain of intermediaries and to
industry. One of the greatest needs in order to achieve sustainable extrac-
tion of NTFPs is to create more equitable marketing and to ensure that the
terms of the Convention on Biological Diversity are implemented to ensure
equitable sharing of any profit.
Extractivism is not the miracle that will save the rain forest. It is, how-
ever, a shorter-term measure that is buying time for a longer-term system of
forest management based on diversity to be developed.
REFERENCES
Alcorn, J. B. (1984a) Development policy, forests, and peasant farms; reflections on
Huastec-managed forests' contributions to commercial production and resource
conservation. Economic Botany, 38, 389-406.
Alcorn, J. B. (1984b) Huastec Mayan Ethnobotany, University of Texas Press,
Austin.
Allegretti, M. H. (1990) Extractive reserves: an alternative for reconciling develop-
ment and environmental conservation in Amazonia, in Alternatives to
Deforestation, (ed. A. B. Anderson), Columbia University Press, New York,
pp. 252-264.
Anderson, A. B. (1988) Use and management of native forests dominated by a~al
palm (Euterpe oleracea Mart.) in the Amazon estuary. Advances in Economic
Botany, 6,144-154.
Anderson, A. B. (1990) Extractivism and forest management by inhabitants in the
Amazon estuary, in Alternatives to Deforestation: Steps towards Sustainable Use
of the Amazon Rainforest, (ed. A. B. Anderson), Columbia University Press,
New York, pp. 65-85.
Balick, M. J. (1987) The economic utilization of the Baba~u palm: a conservation
strategy for sustaining tropical forest resources. Journal of the Washington
Academy of Science, 77, 215-223.
Balee, W. (1986) Analise preliminar de inventario florestal e a etnobotanica
Ka'apor (Maranhao). Boletim do Museu Paraense Emilio Goeldi. Serie Botanica
2(2): 141-167.
Beer, J. H. de and McDermott, M. J. (1989) The Economic Value of Non-Timber
Forest Products in Southeast Asia, World Conservation Union (IUCN),
Amsterdam, The Netherlands.
Bennett, B. C. (1992a) Plants and people of the Amazonian rainforests. Bioscience,
42, 599-607.
Bennett, B. C. (1992b) Uses of epiphytes, lianas and parasites by the Shuar of
Amazonian Ecuador. Selbyana, 13, 99-114.
Boom, B. M. (1985) Amazon Indians and the forest environment. Nature, 314,
324.
Boom, B. M. (1989) Uses of plant resources by the Chacobo. Advances in
Economic Botany, 7, 78-96.
Boom, B. M. (1990) Useful plants of the Panare Indians of the Venezuelan
Guayana. Advances in Economic Botany, 8, 57-76.
Browder, J. O. (1990) Extractive reserves will not save the tropics. BioScience, 40,
626.
Browder, J. O. (1992) The limits of extractivism. BioScience, 42,174-182.
Clay, J. (1992) Some general principles and strategies for developing markets in
North America and Europe for non-timber forest products: lessons from
Cultural Survival Enterprises, 1989-90. Advances in Economic Botany, 9,
101-106.
Denevan, W. M. and Padoch, C. (eds) (1988) Swidden-fallow agroforestry in the
Peruvian Amazon. Advances in Economic Botany, 5, 1-107.
Denevan, W. M., Treacy, J. M., Alcorn, J. B., Padoch, c., Denslow, J. and Paitan, S.
F. (1984) Indigenous agroforestry in the Bora Indian management of swidden fal-
lows. Interciencia, 9, 346-357.
References 41
Denevan, W. M., Treacy, J. M., Alcorn, J. B., Padoch, c., Denslow, J. and Paitan, S.
F. (1985) Indigenous agroforestry in the Peruvian Amazon: Bora Indian manage-
ment of swidden fallows, in Change in the Amazon Basin, Vol. I: Man's Impact
on Forests and Rivers, (ed. J. Hemming), Manchester University Press,
Manchester, pp. 137-155.
Fearnside, P. M. (1989) Extractive reserves in Brazilian Amazonia. BioScience, 39,
387-393.
Hecht, S. B. (ed.) (1982) Agroforestry systems in the Amazon Basin, in Amazonia:
Land Use and Agricultural Research, Centro Internacional de Agricultura
Tropical (CIAT), Cali, Colombia, pp. 271-372.
Hecht, S. B. and Posey, D. A. (1979) Preliminary results on soil management tech-
niques of the Kayapo Indians. Advances in Economic Botany, 7, 174-188.
Heinzman, R. M. and Reining, C. C. S. (1988) Sustained Rural Development:
Extractive Forest Reserves in the Northern Peten of Guatemala, USAID,
Guatemala.
Homma, A. K. O. (1994) Plant extractivism in the Amazon: limitations and possi-
bilities, in Extractivism in the Brazilian Amazon: Perspectives on Regional
Development, Man and the Biosphere, Digest No. 18, UNESCO, Paris,
pp.35-57.
IBGE-Fundaciio Brasileiro de Geografia E Estatfstic (1992) Anmirio Estatistico do
Brasil. Rio de Janeiro.
May, P. H. (1990) Local product markets for baba~u and agroindustrial change in
Maranhao, Brazil. Advances in Economic Botany, 8, 92-102.
Padoch, C. (1990) Santa Rosa: the impact of the forest products trade on an
Amazonian place and population. Advances in Economic Botany, 8, 151-158.
Padoch, C. (1992) Marketing of non-timber forest products in Western Amazonia:
general observations and research priorities. Advances in Economic Botany, 9,
43-50.
Padoch, C. and De Jong, W. (1989) Production and profit in agroforestry: an exam-
ple from the Peruvian Amazon, in Fragile Lands of Latin America, (ed. J. O.
Browder), Westview Press, Boulder, Colorado, pp. 102-112.
Padoch, c., Chota Inuma, J., de Jong, W. and Unruh, J. (1985) Amazonian agro-
forestry: a market-oriented system from Peru. Agroforestry Systems, 3, 47-58.
Panayotou, T. and Ashton, P. S. (1993) Not by Timber Alone: Economics and
Ecology for Sustaining Tropical Forests, Island Press, Washington, DC.
Peluso, N. L. (1992) The rattan trade in East Kalimantan, Indonesia. Advances in
Economic Botany, 9,115-127.
Peters, C. M. (1990) Population ecology and management of forest fruit trees in
Peruvian Amazonia, in Alternatives to Deforestation: Steps Toward Sustainable
Use of the Amazonian Rain Forest, (ed. A. B. Anderson), Columbia University
Press, New York, pp. 86-98.
Peters, C. M. and Hammond, E. J. (1990) Fruits from the flooded forests of
Peruvian Amazonia: yield estimates for natural populations of three promising
species. Advances in Economic Botany, 8, 159-176.
Peters, C. M., Balick, M. J., Kahn, F. and Anderson, A. B. (1989) Oligarchic forests
of economic plants in Amazonia: utilization and conservation of an important
tropical resource. Conservation Biology, 3, 341-361.
Posey, D. A. (1982) The keepers of the forest. Garden, 6, 18-24.
Posey, D. A. (1983) Keeping of stingless bees by the Kayapo Indians of Brazil.
Journal of Ethnobiology, 31, 63-73.
Posey, D. A. (1984) A preliminary report on diversified management of tropical for-
est by the Kayapo Indians of the Brazilian Amazon. Advances in Economic
Botany, 1, 112-126.
Prance, G. T. (1989) Economic prospects from tropical rainforest ethnobotany, in
Fragile Lands of Latin America: Strategies for Sustainable Development, (ed. j.
0. Browder), Westview Press, Boulder, Colorado, pp. 61-74.
Prance, G. T. (1994) Amazonian tree diversity and the potential for supply of non-
timber forest products, in Tropical Trees: the Potential for Domestication and
the Rebuilding of Forest Resources, (eds R. R. B. Leakey and A. C. Newton),
HMSO, London, pp. 7-15.
Prance, G. T., Balee, W., Boom, B. M. and Carneiro, R. L. (1987) Quantitative eth-
no botany and the case for conservation in Amazonia. Conservation Biology, 1,
296-310.
Salafsky, N., Dugelby, B. L. and Terborgh, j. (1993) Can extractive reserves save
the rain forest? An ecological and socioeconomic comparison of non-timber for-
est products extraction systems in Peten, Guatemala and West Kalimantan,
Indonesia. Conservation Biology, 7, 39-52.
Sa lick, j., Megia, A. and Anderson, T. (1995) Non-timber forest products inte-
grated with natural forest management, Rio San juan, Nicaragua. Ecological
Applications, 5, 878-895.
Schwartzman, S. (1989) Extractive reserves: the rubber tappers' strategy for sus-
tainable use of the Amazonian rainforest, in Fragile Lands of Latin America:
Strategies for Sustainable Development, (ed. J. 0. Browder), Westview Press,
Boulder, Colorado, pp. 150-165.
Schwartzman, S. and Allegretti, M. H. (1987) Extractive Production and the
Rubber Tappers' Movement, Environmental Defense Fund, Washington, DC.
Toledo, V. M., Batis, A. I., Becerra, R., Martinez, E. and Ramos, C. H. (1995) La
selva util: etnobotanica cuantitativa de los grupos indigenas del tropico humedo
de Mexico. Interciencia, 20,177-187.
-3------
Degeneration versus regeneration -
logging in tropical rain forests
N. Brown
Department of Plant Sciences, University of Oxford, South Parks Road,
Oxford OX1 3RB, UK
'Naturce, nisi parendo non imperatur' (Obedience is the only way to com-
mand nature)
Francis Bacon
3.1 REGENERATION OF TROPICAL RAIN FORESTS

Tropical rain forests are widely regarded as being very ancient climax
ecosystems composed of highly specialized organisms. This has encouraged
a view that they are potentially extremely fragile and that human interven-
tion will inevitably lead to their destruction. As a consequence, many of
those who are concerned about the loss of this important ecosystem have
campaigned against the use of tropical rain forests for timber production.
Despite this pressure, it is almost certain that within the next few decades
virtually all tropical rain forest not set aside exclusively for conservation
will be logged. This chapter examines the assertion that tropical rain
forests are fragile and that their use for timber production is ecologically
incompatible with their conservation. The process of natural regeneration
in tropical rain forests is reviewed in order to understand the likely conse-
quences of a logging operation.
3.1.1 Environmental instability

In the first half of the 20th century ecologists believed that within a partic-
ular climatic region a stable plant community would develop which was
composed of those species of plant that are dominant under those climatic
conditions (Clements, 1916). Clements believed that this climax community
would be stable because it was in equilibrium with a constant environment
44 Degeneration versus regeneration - logging in tropical rain forests
and that only humans could destroy this stability. Clements's model seems
so intuitively logical that it still exerts considerable influence over the ways
in which people view plant community dynamics, despite the fact that a
modern vision of vegetation dynamics has substantially revised it.
It is now understood that almost no aspect of the natural environment is
ever stable for long periods of time. It is therefore improbable that the sort
of stable climax plant community that Clements envisaged would ever
occur. The global environment is, and always has been, extremely dynamic.
Not only do extreme natural events frequently disrupt forest ecosystems,
but climate is changing continuously and at a rate which is likely to have
ecological significance within the life-span of most trees. Condit et al.
(1996) report that a drying trend in the climate of Barro Colorado Island
over the last 25 years and an associated increase in severity of droughts is
having an obvious impact on forest composition. Sixteen species which are
restricted to mesic sites are going extinct. At least nine glacial periods have
significantly reduced rainfall in humid tropical areas during the Quaternary
period. It is likely that the area of closed forest in West Africa was
restricted to small refuges in Liberia, western Ghana and Cameroon.
Glaciers only began to retreat from Africa's mountains 15 000 years ago
and palynological evidence suggests that rain forest did not return to many
lowland areas before 12 500 BP (Livingstone, 1982). There is no indication
that the recent past is in any way exceptional or that the global climate has
ever been stable for longer periods.
Superimposed on this pattern of constantly changing climate are natural
extreme events of a wide range of magnitudes that also have a significant
impact on forest structure and composition. These include:
• Lightning: has been shown to have caused gaps in an Amazonian forest
of up to 680 m2 in extent and was thought to have been responsible for
55% of the total gap area in the forest (Magnusson et aI., 1996).
Bruenig (1964) estimated from aerial photographs that up to 2.6% of a
forest in Sarawak was affected by lightning gaps.
• Storms: Nelson et al. (1994) reported the frequent occurrence of large-
scale forest blow-downs in the Amazon basin, caused by wind squalls.
These are commonly up to 500 ha in extent but some are in excess of
3000 ha. Hurricanes frequently affect tropical rain forests in both the
Caribbean and the Pacific and can cause almost complete defoliation.
Emergent trees are more likely to be killed or damaged (Brokaw and
Grear, 1991).
• Volcanic eruptions: Whittaker et al. (1992) concluded that a series of
ash-falls from the Krakatau volcano had been the principal determinant
of the successional process in forests in the region.
• Drought: whilst droughts are typically of short duration in most tropical
rain forest regions, their impact on a vegetation which has little adapta-
Regeneration of tropical rain forests 45
tion to water stress can be severe. Walsh (1996) has demonstrated that
droughts of longer than 3 months occur with a frequency which is sig-
nificantly shorter than most tree life-spans in northern Borneo. He cites
unpublished data collected by Mark Leighton (Department of
Anthropology, Harvard University) which show that 37-71 % of canopy
trees on slopes and ridges were killed by the 1983 drought in
Kalimantan, Indonesia. Small trees and vines have relatively higher rates
of survival. Baillie (1976) pointed out that even if droughts are not
strong or long enough to kill adult trees, the phenology of many humid
tropical forest tree species in South-East Asia appears to be strongly
influenced by them. The magnitude and frequency of droughts may
therefore be of great importance in controlling when and where many
tree species reproduce. Droughts will also be important in determining
how vulnerable forests are to fire. Conventionally it was believed that
primary rain forest was not susceptible to fire. However there is a grow-
ing body of evidence that fires have been widespread in tropical rain
forests around the globe long before significant human impact. In 1982
nearly 1 million ha of primary lowland rain forest were burnt on the
island of Borneo in one of the largest forest fires ever recorded
(Malingreau et al., 1985).
• River dynamics: erosion and deposition by lateral migration of streams
is constantly reworking the topography of the Amazon rain forest.
Satellite photography of the Amazon basin has shown that 27% of the
lowland forest has characteristics of recent erosion or deposition by
rivers, and that 12% of the Peruvian lowland forest is in successional
stages along rivers (Salo and Kalliola, 1991).
• Forest-dwelling human disturbance: the effect of humans on forest com-
position is not new. It is now known that populations of forest-dwelling
humans have had a profound effect on the forests in wbich they live. The
apparently natural landscapes of many tropical rain forests may in fact be
the product of thousands of years of manipulation by aboriginal popula-
tions (Posey, 1994). Because aboriginal populations were often much
larger than those remaining today, it is easy to underestimate their influence.
• Individual tree life-cycles: at the very smallest scale is a type of distur-
bance that occurs in almost all plant communities. Individual plants are
born, grow and then die, leaving a hole in the plant canopy that can be
filled by new plants growing up from below. In a tropical rain forest
many of the trees are in excess of 50 m in height and may have a trunk
and limbs weighing many tonnes. The death of a tree from old age can
result in considerable damage to the surrounding forest as the tree
crashes to the ground. Anyone who has camped for a night in a tropical
rain forest will be all too aware of how frequently the peace is shattered
by the thunder of an unexpected tree fall. Much more common, and
considerably less dramatic, are the small holes left by falling branches or
the death of an under storey tree.
Disturbance is particularly important to the process of regeneration in
tropical rain forest. The forest canopy, unlike that of many temperate
forests, is multi-layered and absorbs incoming solar radiation extremely
effectively. As a consequence very little radiation that is of any value for
photosynthesis filters through to the forest floor. For most plants the quan-
tity of photosynthetically active radiation on the forest floor is too low for
growth, and for many it may be below the compensation point at which
the demands of respiration exceed gains from photosynthesis. It is there-
fore of little surprise to learn that many tree species depend on canopy dis-
turbance to create conditions under which there is sufficient light for their
seedlings to grow.
A growing appreciation of how common and widespread natural distur-
bance is in plant communities has led to a significant change of view.
Ecologists now view vegetation composition as a balance between the
processes of competition between plants and environmental disturbance,
rather than seeing a plant community as stable unless disrupted.
Disturbance releases resources by killing off the dominant plants and pro-
vides opportunities for new plants to grow. The magnitude and frequency
of disturbance exert a crucial influence over the types of plant that are
likely to survive and grow.
3.1.2 Plant competition

An important paradigm in ecology maintains that two or more resource-lim-
ited species having identical patterns of resource use cannot co-exist. One
species will be better adapted and will out-compete the others (Gause,
1934). Such a view poses great problems for the study of successional
processes in tropical rain forests, where many hundreds of tree species
appear to coexist without any obvious differences in their ecology. Co-
existence among animal species has been partly explained by differences
among species in the use of resources, such as differences in feeding niches
(Hutchinson, 1957). Similarly, it has been proposed that plants growing at
the same place at the same time may occupy different niches by the devel-
opment of physiological and morphological traits which allow them to sur-
vive or grow best at different levels of resource availability (Grime, 1977).
A patchy distribution of resources (such as the changes in soil depth and
moisture availability down a soil catena) might provide opportunities for a
large number of species of different ecology to co-exist.
Once established, mature trees may differ little in their resource require-
ments. Indeed most large trees will simultaneously experience a huge range
of environmental conditions across the full width and depth of their
canopy and across a wide-spreading root network. However small-scale
patchiness in resource availability may be much more important to a
seedling which is confined to one microenvironment. Grubb (1977) has
Regeneration of tropical rain forests 47
suggested that species may differ in the environmental conditions they
require for establishment, but once established may become increasingly
tolerant of a wide range of conditions.
Trees clearly differ at the seedling stage in their ability to capture and
utilize different levels of resource. It is possible to characterize tropical rain
forest trees as belonging to one of two crude categories (Swaine and
Whitmore, 1988). Pioneer species are those that germinate and establish
seedlings only in highly disturbed environments. Their ecology is designed
to exploit rapidly the abundance of resources made available by the death
of dominant plants. Pioneer species typically exhibit very fast growth rates
and become sexually mature relatively quickly. Few have sufficient com-
mercial value to warrant their extraction from natural tropical rain forest,
though their ecology is well suited to plantation forestry. At the other
extreme, many climax species have an ecology which is designed to tolerate
resource scarcity. Most have seeds that are able to germinate in the dark
forest understorey and can tolerate canopy shade for long periods, until
disturbance allows light to penetrate to the forest floor and creates an
opportunity for growth. Shade-tolerant species reach their peak rates of
photosynthesis at much lower light levels than pioneer plants (Riddoch et
al., 1991).
There is some evidence to suggest that there is a wide spectrum of physi-
ological and life-history characteristics amongst tropical rain forest trees.
Present understanding of seedling growth responses to different forest light
environments is based largely on extrapolation from short-term growth
trials or instantaneous measurements of photosynthesis (Press et al., 1996).
These do not adequately reflect the range of plant responses to environ-
mental variability (Zipperlen and Press, 1996). However, even on this very
limited basis, studies have shown that different seedling species reach max-
imum rates of photosynthesis and growth at different light levels (e.g.
Oberbauer and Strain, 1984; Freedon and Field, 1991; Ashton and Berlyn,
1992).
In view of the measurable differences between different tree species in
their response to light, it has been suggested that the great spatial and tem-
poral variability in light within a tropical forest provide opportunities for a
variety of different species to grow. Denslow (1980) has proposed that dif-
ferent species of climax rain forest tree seedlings may be adapted for opti-
mum growth at different levels of irradiance. As the forces that disturb the
forest canopy are essentially haphazard, a wide spectrum of light levels is
maintained. Species specialization would give a competitive advantage in a
disturbance that created just the right light conditions, but would be likely
to involve adaptive compromises that would restrict success where the light
was at a different level. Different levels of resource availability therefore
give different plants the upper hand. In this way a variety of different types of
disturbance may be central to maintaining a diversity of different tree species.
The relative competitive status of plants is not, however, determined
solely by species-specific adaptations. The consequences of the competitive
interaction between plants are not felt equally by the competitors. The
example of competition for light provides a useful illustration. A tall plant
will intercept incoming sunlight and cast shade on those plants beneath it.
The shaded plants may have inherently faster growth rates under full illu-
mination but are at a competitive disadvantage as a consequence of their
inferior position. Competition between plants is therefore highly asym-
metrical. Plants of different size have highly unequal access to resources.
Differences in competitive ability between species will become progres-
sively less important as some individuals dominate a microsite and exclude
others from access to resources. The practical consequence of this for
forestry is that often the first trees to colonize a site may pre-empt
resources (especially light) and prevent the establishment of better-adapted
species for many decades.
3.2 SILVICULTURAL SYSTEMS

One of the most important factors that controls the nature of forest regen-
eration is the type and extent of disturbance. The damage inflicted on a
tropical rain forest by a logging operation is both quantitatively and quali-
tatively different from that which the forest experiences naturally. The pre-
cise nature and extent of forest disturbance depends on the silvicultural
system that is used, on the number of commercially valuable trees that are
available to be harvested, and on the skill and care exercised by those car-
rying out the logging.
Whilst many different natural forest silvicultural systems have been
described (e.g. Troup, 1952; Baur, 1968; Matthews, 1989; Gomez-Pompa
and Burley, 1991), they differ significantly in only three ways (Pryor and
Savill, 1986):
• whether they use natural or artificial regeneration to produce the future
crop;
• the proportion of the commercial stock of trees that is extracted at each
rotation;
• the age distribution of the commercially valuable trees.
One option for the management of natural tropical rain forest is to fell all
saleable trees above a given size at once. This results in a uniform opening
of the canopy and it is intended to create open conditions in which an
even-aged crop of young seedlings of light-demanding tree species grow
rapidly to maturity. Such an approach is variously described as a uniform,
monocyclic or shelterwood system. This intensive operation often causes
extensive damage to residual trees, which are frequently killed off by
poison-girdling to avoid their suppressing seedling regeneration. The key to
Silvicultural systems 49
success is to make sure that there is widespread survival after logging of the
seedlings of commercially valuable species. The detailed silvicultural pre-
scriptions for the Malayan uniform system (Wyatt-Smith, 1963) stress the
importance of ensuring that there is adequate seedling regeneration before
felling occurs. If few trees have produced seed in the recent past and there
is an inadequate number of seedlings, felling should not be permitted.
In a polycyclic or selection system, a small number of desirable trees are
removed over a series of felling operations. At its most basic level such a
selection system can amount to little more than the cutting of a certain
number of trees above a chosen diameter. Matthews (1989) has pointed
out that this type of rather crude exploitation requires little or no silvicul-
tural skill and does not ensure regeneration. It does not guarantee a sus-
tained yield and often results in serious deterioration in the timber value of
the forest. This results in an unevenly aged pattern of regeneration.
Scattered, small gaps are made in the forest canopy where trees are
extracted. Saplings and young adult trees already present in the undis-
turbed forest will be released by the influx of light and, if undamaged by
. felling operations, may grow rapidly to commercial size. However, after a
small number of felling cycles there will no longer be any residual saplings
left from the undisturbed forest. Regeneration will then depend on
seedlings that have established and grown into saplings and young adults
in logged forest. A silvicultural system, whether monocyclic or polycyclic,
can therefore be judged to be truly sustainable in the long term only if it
creates conditions under which seedlings of commercial species continue to
establish and grow. It is essential that logging creates environmental condi-
tions favouring the germination, establishment, growth and survival of
seedlings of commercially important tree species.
Most natural tropical rain forest management relies on natural regenera-
tion for two reasons. Firstly, it is often the simplest and cheapest type of
forest management. It requires little investment and frequently involves no
further manipulation of the forest beyond that which occurs as a conse-
quence of timber extraction. All too often natural regeneration is seen as
being synonymous with laissez-faire management. Planting is much more
expensive as it is a labour-intensive operation and requires the production
of planting stock. The seeds of many important tropical timber trees are
difficult both to collect from the wild and to store. Seeds of many species
may last only a few months when stored under ambient conditions. Those
that are fleshy or have a high oil content may have an even shorter period
of viability. Such seeds, which are described as being recalcitrant, fre-
quently begin to germinate soon after they have fallen from the tree and do
not tolerate drying. Most species of dipterocarp, for example, fruit only
once every few years and the seeds remain viable for a matter of a few days
after they have fallen from the tree. Planting stock cannot therefore be pro-
duced on demand. As a consequence, foresters are forced to use whatever
50 Degeneration versus regeneration -logging in tropical rain forests
source of planting stock they can obtain regardless of its quality. They are
often dependent on wild seedlings transplanted from undisturbed forest or
old nursery-grown stock. Foresters from temperate countries with small,
well known tree floras often argue against natural regeneration on the
grounds that the dissemination, genetic quality and even the species which
regenerate are largely out of the control of the manager (Pryor and Savill,
1986). However, as we have seen, these are the very problems that may
afflict planting in many tropical forest areas.
The second important reason for the use of natural regeneration in nat-
ural tropical rain-forest management is that it may result in the production
of better quality trees than is likely from planting. The stocking levels of
naturally regenerated seedlings are usually much higher than it is feasible
to achieve by planting. This provides a broader basis for the selection of
high quality trees. As the parent trees are likely to be well adapted to local
conditions it is probable that many of the seedlings will also be adapted to
the site. However, this does not necessarily mean that they will be the most
productive trees on that site. Introduced species and provenances may
grow faster or have better form. However, virtually no information exists
on the relative performances of different provenances of most climax rain
forest tree species, and little or no selection is typically made for quality trees.
3.2.1 Harvesting intensity

The precise number of commercially valuable trees varies from forest to
forest. FAO (1993) estimated that, on average, only 33 m3 of logs were
extracted from each hectare of moist natural forest in the Asia-Pacific
region in the period 1986-90. This is roughly equivalent to seven large
trees. The figures for Africa and the Latin American-Caribbean region were
13 m3 (three large trees) and 8 m3 (two large trees), respectively. There are
two important reasons why so few trees are cut.
(aJ Most tropical trees have no commercial value

Although typically over 100 different species of tree may be found growing
on a hectare of tropical moist forest, only a minority of these trees have
any commercial value. This is because the timber industry is a highly con-
servative one. The prestige and durability of tropical hardwoods are highly
valued characteristics. Timber users are usually reluctant to experiment
with a costly material with no established prestige value and of unproven
performance. The timber industry is unwilling to market woods for which
there is no obvious demand. As a consequence, whilst the prices paid for a
small number of very well known tropical timbers are high, the majority of
tropical tree species cannot be sold. In Belize the profit on selling mahogany
and cedar for export is 10-20 times that of most other hardwoods (Plumptre,
1993). Timber companies prefer to extract relatively small quantities of
high-value timber from extensive areas of forest than large quantities of
low value timber from smaller areas.
The lowland forests of South-East Asia contain the highest number of
commercially exploited species. Many forests in this region have very high
stocking densities of large, valuable trees. Simultaneous extraction of all of
these can result in devastation of the residual forest. Many so-called selec-
tive logging systems merely prescribe a minimum diameter above which
commercially desirable trees may be felled. Up to 30 large trees per hectare
have been extracted from some Malesian dipterocarp forests. This situa-
tion is unusual elsewhere in the humid tropics. In Ghana only 10 species
are regularly used; in Latin America and the Caribbean the variety of
species used is even lower. This is because many of the common species of
tree have timber which is dense, dark and siliceous (Whitmore and Silva,
1990) making it difficult to machine and utilize profitably. The tropical
timbers that are most extensively utilized today are of low-to-medium den-
sity.
(b) Most tropical tree species are rare

The corollary of high species diversity in tropical forests is that the forests
are also rich in species occurring at exceptionally low densities. One of the
most detailed enumerations of a large area of tropical rain forest was car-
ried out in Panama (Hubbell and Foster, 1986). When compared with
plant communities from other types of habitat, the tropical forest was
found to have very many more rare species. More than one third of the
303 tree species found on a 50 ha plot had an average density of less than
one individual per hectare. A similar pattern is found in most other humid
tropical forests. The implication for timber harvesting is that in order to
accumulate a significant volume of timber of anyone species, an extensive
area of forest will have to be exploited. The distribution of individuals of
anyone species is typically neither random nor uniform. Most species have
been found to be patchily distributed. This means that the intensity of tim-
ber extraction is also very patchy.
There are a small number of humid tropical forests where commercially
valuable timber trees are not rare. In Trinidad, for example, Mora excelsa
may comprise up to 95% of the canopy (Beard, 1946). In the north-eastern
tip of Borneo over 60% of the timber trees may be of Parashorea
malaanonan (Whitmore, 1984). Extraction of timber volumes in excess of
100 m3 per hectare can lead to excessively high levels of damage to the
residual forest.
3.2.2 Damage levels

Until the development of powerful vehicles in the Second World War, log-
ging of tropical rain forest was restricted to the cutting of isolated individuals
of highly valuable trees. The damage caused was similar in both frequency
and extent to natural forest disturbance, and therefore had a negligible
impact on forest structure and composition. Modern forestry methods
have a much higher impact. Selective logging of tropical rain forest rarely
results in the complete destruction of the forest canopy. Surprisingly, the
damage caused by a logging operation is not always in direct proportion to
the volume of timber cut. Damage is caused in several ways.
(a) Extraction routes

Traditionally logs were dragged from the forest by hand or using animals.
These methods have been almost entirely replaced by extraction with trac-
tors, bulldozers and trucks. Tracks and roads are cut through the forest in
order to extract logs. These include:
• roads along which heavy trucks carry logs from the forest to the timber
mills or port;
• log-loading bays where logs recently extracted from the forest are col-
lected, graded and temporarily stored;
• skid tracks along which the logs are dragged from the felling site to the
loading bay.
The methods for planning and construction of an extraction network have
traditionally constituted an important part of a forester's education.
However, it is unusual to find a logging operation in humid tropical forests
where these skills are put into practice. Roads are expensive to construct
and maintain, but a well-planned road system will minimize transport dis-
tances and reduce the need for skidding logs through the forest. However,
in order to plan an efficient road system the forester needs full information
on the topography and position of those trees that are to be extracted.
Rarely are detailed harvest plans required and few logging companies will
afford the significant cost of accurate mapping. As a consequence most
road and skid track construction is ad hoc. This is inefficient from an oper-
ational perspective and can lead to unnecessary damage where many roads
are constructed to access a limited area. Verissimo et al. (1995) report that
over 10% of an Amazonian forest logged for mahogany was destroyed by
roads and skid tracks. MaImer and Grip (1990) found skid tracks to cover
24 % of a forest plot harvested by bulldozers. Logs are, in some cases,
extracted several days after felling. Bulldozer drivers will find felled trees
by driving towards apparent openings in the forest canopy (Gerwing et al.,
1996). The result is that skid tracks may wander through the forest.
Frequently trees that have been felled are not relocated. Gerwing et al.
(1996) found that as many as 20% of felled trees were never extracted.
Careful planning and construction of roads and skid tracks are often iden-
tified as the most important measures that can be taken to reduce damage
(d'Oliveira and Braz, 1995; Pinard et al., 1995). In mixed dipterocarp hill
forest in Sarawak, careful planning of skid tracks resulted in a noticeable
increase in extraction efficiency, fewer commercial trees were left unfelled,
fewer logs were lost after felling and there were fewer accidents. In addi-
tion costs dropped by 23% (Marn and Jonkers, 1982).
Logs are usually dragged from the felling site to the roadside by a bull-
dozer. The bulldozer was designed for moving large volumes of earth and
not for forestry work. It is an efficient workhorse because it is powerful,
manoeuvrable, and its caterpillar tracks allow it to maintain traction over
steep and soft ground (Figure 3.1). However, bulldozers can be extremely
damaging to both trees and soils. Passage of a bulldozer creates a skid
track that is at least 3.5 m in width. Along this track all plants are crushed
and soils are heavily compacted. Frequently all vegetation and topsoil are
bulldozed from tracks in order to improve traction. This exposes subsoil
that is naturally more compact. Compaction and topsoil removal by bull-
dozers resulted in a 70% increase in soil bulk density in Amazonian rain
forest (Woodward, 1996). Jusoff (1988) measured an average increase in
soil bulk density of over 50% after a single pass of a D8 bulldozer over
soils in a Malaysian forest. Compaction significantly reduced soil porosity
and resulted in a ten-fold increase in resistance to penetration.
(b) Felling damage

Natural tree-fall gaps are formed by the death of trees from all size classes.
Indeed, Swaine et al. (1987), in a review of studies of tropical rain forest
dynamics, found that the risk of mortality among trees was independent of
tree age. Large emergents were no more likely to die than main canopy or
understorey trees. However, when a forest is logged for timber the trees
that are felled are typically some of the larger individuals in the forest.
Hence large gaps will be much more common in logged forest than in
unmanaged forest. In Guyana felling gaps were found to be at least three
times larger than average natural gaps (ter Steege et aI., 1996). When it is
felled, the crown of a large tree, which may be 20-30 m in diameter, will
sweep down through the forest. It is inevitable that many smaller trees in
its path will be crushed or snapped (Figure 3.2). Skilled directional felling
can help to minimize this damage and avoid injuring trees that may be of
high ecological or commercial value. Larger trees frequently carry heavy
vine infestations that tie their canopies to those of the trees around them.
Unless these vines are cut before the tree is felled, neighbouring trees may
also be pulled down. Vine-cutting before felling has been shown to reduce
the number of trees dragged down or snapped by up to 50% (Fox, 1968;
Appanah and Putz, 1983; Pinard et aI., 1995) but is rarely practised
because of the cost. Further extensive damage is frequently caused at the
felling site as a log is manoeuvred into position for extraction. Skilled log-
ging crews are able to fell trees in the direction from which they will then
be extracted to minimize the need for turning logs.
Figure 3.1 The caterpillar-tracked bulldozer is an extremely efficient but highly

damaging work-horse for dragging timber from tropical rain forests.
(c) Post-logging tree mortality

The mortality of residual trees after logging is often higher than in undis-
turbed forest. This is primarily because many individuals may have been
injured but not destroyed during logging. These are subsequently more
prone to disease as trunk damage may weaken a tree and allows access to
fungi. Logging will also create significant changes to the forest structure,
leaving a larger number of isolated trees and forest edges where tree falls
are more likely. Kasenene and Murphy (1991) report that tree falls were
over twice as common 20 years after heavy logging in a rain forest in
Uganda than in an un logged control area. In a logging experiment in
French Guiana, post-logging tree mortality increased markedly in compari-
son with unlogged control plots (Madron, 1994). However, this increase in
mortality was relatively short-lived and within 5 years little difference
could be detected between the two treatments.
Despite the fact that, in general, relatively small volumes of timber are
removed from tropical rain forest, damage can be considerable. Figure 3.3
compares the proportion of trees felled with the percentage of all residual
trees that were killed or damaged in a number of studies where comparable
methods were used for assessing damage. Gaps in the forest canopy created
by logging are more frequent and tend to be larger than those in most nat-
ural forests. In Uganda mean gap size was found to have increased from
Figure 3.2 It is common for between five and ten trees to be damaged or killed
when one tree is felled in natural tropical rain forest.
approximately 250-1300 m 2 in heavily logged forest (Skorupa and
Kasenene, 1984). This situation is exacerbated where commercially valu-
able trees grow as conspecific groves. Mahogany trees occur in widely scat-
tered patches. Whilst only a small proportion of the whole forest is
disturbed by their extraction, locally the damage is intense (Verissimo et
aI., 1995). Greenheart (Ocotea rodeii) trees grow gregariously in forests in
Guyana (ter Steege et aI., 1996). Logging leaves those parts of the forest
without groves of greenheart more or less intact. Where a group of large
greenheart trees is felled the resultant gap is much larger than those that
form naturally. Large-scale natural disturbance is rare in Guyana.
Logging may indirectly make forests more vulnerable to natural distur-
bance. Malingreau et al. (1985) concluded that though up to 800000 ha
of primary lowland rain forest was burned during the extreme droughts of
1982-83, 1.2 million ha of selectively logged forest suffered comparatively
greater damage. This was attributed to the accumulation of organic debris.
3.3 THE EFFECTS OF LOGGING ON FOREST STRUCTURE AND

COMPOSITION
Logged forest is a mosaic consisting of areas that have been totally
denuded of all vegetation, areas occupied by residual trees that have sur-
vived disturbance, and areas where the original forest is virtually intact.
Logging leads to changes in the structure and composition of tropical
rain forest because:
• the largest individuals of commercially valuable species of tree are felled;
• other species and size classes are destroyed in proportion to their abun-
dance;
• the high levels of disturbance caused by logging create widespread
opportunities for regeneration.
3.3.1 Consequences of a loss of large trees of commercial species

Forests that have not experienced large-scale disturbance in the recent past
typically show an inverse exponential size-class distribution with large
numbers of the smallest sizes of tree and progressively fewer individuals of
larger trees. Whilst large trees are few in number they constitute a signifi-
cant component of the total forest biomass. Logging selectively removes
the largest trees and they are replaced by large numbers of seedlings and
saplings in the very smallest size classes. As a consequence, secondary for-
est is characterized by a higher total stem density, lower biomass, shorter
stature and lower density of large trees (Brown and Lugo, 1990). It is
unlikely that the forest will recover its pre-logging structure. Many of the
largest trees removed the first time that a forest is logged may be many
The effects of logging on forest structure and composition 57
70
• Indonesia (1)
60
• Malaysia (2)
;e
~ 50
'C • Indonesia (3)
.!!!
:si • Amazonia (4)
0 40
'C
CD
Ol
• Amazonia (4)
('II
E 30
('II
• • {r.\\
'C .Amazonia (4).Amazonia (5)

rn
E
20 uueenslana (IS) •
c75
• Amazonia (6)
• Amazonia (7)
10 "'Amazonia (6)
•• Amazonia (7)
o
0.0 1.0 2.0 3.0 4.0 5.0 6.'
Stems extracted (%)
Figure 3.3 Relationship between the proportion of all trees with trunks ~ 10 cm
diameter which were felled and the proportion accidentally killed or damaged dur-
ing this operation. Data from (1) AbduIhadi et al. (1981); (2) Johns (1988);
(3) Bertault and Sist (1995); (4) Verissimo et al. (1992); (5) UhI and Vieira (1989);
(6) UhI et al. (1991); (7) Verissimo et al. (1995); (8) Crome et al. (1992).
hundreds of years old. Even where forests are managed sustainably, very
few trees will be allowed to grow to such great age before they are cut.
Timber tree species populations suffer disproportionate damage. A num-
ber of important timber trees have been exploited so intensively that they
have become commercially extinct in large parts of their natural range.
Large-sized trees of Ceiba pentandra have been completely logged out of
the Peruvian Amazon (Gentry and Vasquez, 1988). Seedlings of this species
are, however, plentiful and it is likely that a population could re-establish.
Where large-sized trees are the only representatives of a population of a
commercially valuable species, even selective felling may lead to local
extinction. Unfortunately this situation appears to occur quite commonly.
Perhaps the best-known example is that described by Jones (1956) in the
rain forest of southern Nigeria. Here he noted that an African mahogany
(Entandrophragma spp.) had virtually no regeneration. It is therefore very
vulnerable to local extinction as a consequence of exploitation. Appanah et
al. (1990) reported that middle-sized dipterocarps are often rare in hill
forests of Peninsular Malaysia, resulting in a shortage of trees to grow on
and form the next crop after logging. Verissimo et al. (1995) discovered a
complete absence of mahogany trees of between 1 and 30 cm breast-height
diameter (dbh) in an Amazonian rain forest from which large mahogany
trees were being extracted. Mahogany seedlings require large, well-lit open-
ings in order to grow. The present cohort of very large trees is likely to be a
relic of disturbance that occurred several hundred years ago. Mahogany
populations may no longer be replacing themselves in undisturbed forest.
Although the logging of large mahogany trees causes extensive disturbance
and therefore might create brightly lit conditions in which seedlings would
thrive, there is then a scarcity of mature trees to produce seed.
The sudden decline in numbers of large trees of a small number of tim-
ber species can have a direct impact on other components of the ecosystem.
Epiphytic and strangling figs (Ficus spp.), for example, are prone to severe
damage in Malaysian rain forests because they tend to grow on the largest
trees of commercially valuable species (Lambert, 1991). The fruits are of
enormous importance to a wide range of frugivorous birds that rely on this
source of food when others are scarce. Their loss may therefore have a dis-
proportionately large impact on bird conservation.
Concern is frequently raised over the potential for selective logging to
lead to dysgenic selection in the regenerating stand. A selection system will
preferentially harvest the largest trees with the best form. Smaller trees and
those with poor form will therefore constitute a larger proportion of the
residual population of commercial species. Should selective harvesting con-
tinue over a number of rotations, it is likely that there will be progressive
deterioration in the quality of the crop. This effect is well proven in tem-
perate forests. Jones (1952) demonstrated that the selection system used in
the Chiltern beech woods caused a significant decline in tree quality.
However, the evidence for tropical rain forests is less convincing. There are
well-known examples of extreme levels of exploitation having led to con-
siderable reduction in genetic diversity. Styles (1972) claimed that true
mahogany (Swietenia mahogani) now exists only as a small number of
shrubs because of selective harvesting of the best trees. A combination of
almost complete exploitation of adult trees and very heavy seedling her-
bivory reduced the population of an island endemic, Trochetiopsis ery-
throxylon (St Helena redwood) to a single individual. The species is now
suffering severe in-breeding depression (Rowe, 1995).
However, for most tropical trees exploitation has not yet been so severe.
Regrettably, there are no data on pre-exploitation levels of diversity for any
tropical timber tree and it is not possible to determine whether unique
alleles have been lost (Ledig, 1992). It is likely that, given the very high lev-
els of genetic diversity in most tropical tree populations, there will be negli-
gible genetic degradation so long as only the finest trees are removed from
a forest. Under a logging regime where all saleable trees apart from the
very poorest are removed there could be a rapid degradation of form and
growth rate (Ledig, 1992). Dysgenic selection may be avoided by ensuring
that high-quality trees have established a vigorous seedling bank before
they are felled and that slow-growing and poorly formed trees are killed.
Unfortunately, in practice many selection systems are no more than forest
exploitation and little heed is paid to ensuring that there is an adequate
seedling stock before logging commences. As a consequence, future tree
crops have to seed from a small number of poor quality trees. The reduc-
tion in numbers of sexually mature adults can have further significant con-
sequences for the genetic structure of subsequent seedling populations.
Murawski et al. (1994) compared two adjacent populations of an impor-
tant timber tree, Shorea megistophylla, in a Sri Lankan rain forest. They
demonstrated that the reduction in adult population size in the logged for-
est caused a significant increase in the proportion of seeds produced
through in-breeding. This may have some effect on adaptive genetic diver-
sity, but more seriously it is likely to lead to in-breeding depression result-
ing in reduced reproductive output.
3.3.2 Consequences of non-timber tree destruction

Quite contrary to expectation, it is frequently found that the number of
tree species in a logged forest differs little from the pre-logging number. For
example, Crome et al. (1992) report that no tree species were eliminated by
the logging of an upland rain forest in Queensland, although almost all
species became less abundant. A detailed study of the whole plant commu-
nity in a Nicaraguan rain forest showed that species richness increased
after logging (Salick et aI., 1995). Similarly, a study of forest regeneration
after logging in Ghana (Hawthorne, 1993), which included herbaceous as
well as woody plants, showed a larger number of species in logged forest.
The reasons for these counter-intuitive results are simple. It would appear
that despite extensive damage, residual populations of most species survive
forest logging. Many species may be reduced to population levels at which
their future viability may be at risk, but in the short term they continue to
be represented in the forest. However, logging provides opportunities for a
variety of species that were not present in the undisturbed forest to colon-
ize disturbed areas. Frequently these include common herbaceous pioneer
species such as Eupatorium, a pan-tropical genus of weeds that proliferate
along roadsides and around log-loading areas. The new arrivals more than
compensate numerically for the loss of some rare forest species but clearly
are of much lower conservation value.
3.3.3 Consequences of forest regeneration

The structure of a forest regenerating after logging reflects the heterogene-
ity of damage inflicted and the condition of the forest prior to logging.
There are three common types of environment in logged forest and each is
colonized by a characteristic flora.
(a) Log-loading bays

Most pioneer species thrive in the bright, open conditions found where the
forest canopy has been heavily disturbed. However, loss of forest topsoil
on abandoned logging roads and loading bays leads to a drastic reduction
in the numbers of viable seeds available. Putz and Appanah (1987) found
131 viable seeds per m2 in the top 10 cm of soil under the mature forest
canopy of lowland dipterocarp rain forest in Peninsular Malaysia, but
almost none in soil from between 10 and 20 cm depth. As all plants and
virtually all topsoil is lost from roads and log-loading bays, forest regener-
ation on such sites may only proceed by a process of primary succession
(Figure 3.4). The pattern of succession is primarily determined by which
seeds are dispersed to the site and are then able to germinate and establish.
Very few forest species are able to colonize large, cleared areas that have
been totally denuded of topsoil. There are two important reasons for this.
Firstly, seeds of many tree species may only rarely be dispersed to cleared
areas. Many climax species that produce large seeds are notoriously poorly
dispersed. Gravity or wind disperses seeds of dipterocarp species in
Malesian rain forests. Burgess (1970) reported that 80% of Shorea curtisii
fruits fell less than 40 m from the parent tree. Animals are the most effi-
cient dispersal vectors in a humid tropical forest. Hammond et al. (1996)
found that 51 % of timber trees in a Guianan rain forest had mammal-dis-
persed seeds and 21 % were dispersed by birds. Although many animal
species will forage along forest edges, they rarely venture out into the cen-
tre of large cleared areas where there is little cover and few perch sites are
available. Nepstad et al. (1991) found that seeds dispersed from nearby
forest by birds and bats were the most important source of new trees estab-
lishing on cleared and degraded land in the Amazon. Seeds of 18 tree
species were collected from beneath scattered small trees and bushes in the
cleared areas where birds and bats could perch and roost. Only two species
were found in the open.
The second reason why many species find great difficulty in colonizing
cleared areas is that loss of seed and seedlings may be very high. Heavily
compacted subsoil has a low infiltration capacity and intense tropical rain-
fall will almost always result in run-off. Silt, sand, seeds and leaf litter are
all washed away by this run-off. Seeds that are dispersed onto the site are
carried away before they are able to take root. Removal of plant litter
means there is no accumulation of organic matter in surface soil horizons.
This, in combination with loss of fine sediments, depletes the soil of the
majority of its cation-exchange capacity, permitting the leaching of plant
nutrients from the soil.
Climax forest species do not germinate or establish well in large clearings.
Figure 3.4 There are significant barriers to the re-colonization of roads and log-
loading bays by vegetation; rapid erosion and low nutrient status are among the
most important.
Raich and Gong (1990) found that seeds from 43 species of canopy tree
native to lowland forests of Malaysia germinated well in a small forest gap
but failed to germinate in a large clearing. This implies that regeneration of
many trees may be poor in large clearings, even when their seeds are pre-
sent. Poor germination of climax species in open conditions is commonly
reported (Turner, 1990; ter Steege et al., 1994; Swaine et aI., 1997). Climax
forest tree species have conventionally been thought to be adapted to life in
the cool, shaded forest interior. This led to the supposition that they are
unable to cope with the harsh microclimate of large cleared areas.
However, a growing number of studies of seedling physiology have shown
that many forest species are highly adaptable and can acclimate to condi-
tions in the open (Press et al., 1996). Scholes et al. (1997) have found that
dipterocarp seedlings develop systems for protecting leaves from excess
energy through increased thermal dissipation when grown in the open.
This increased protection reduces the efficiency with which leaves are able
to photosynthesize, but in some species is reversible once conditions ame-
liorate. Many indigenous tree species, whilst physiologically capable of
withstanding high irradiance and low relative humidity, are not able to tol-
erate very oligotrophic soils. Nussbaum et al. (1995) have shown that the
very poor growth of both pioneer and climax tree seedlings on log-loading
bays is a consequence of nutrient deficiency rather than soil compaction or
an extreme microclimate. Seedlings that were given a large nutrient dose
showed a dramatic increase in growth. Similar results were obtained when
forest topsoil was replaced around seedlings. In addition to increased
seedling growth, this treatment also resulted in dense germination of other
pioneer plants from the topsoil seed bank.
A small number of herbaceous and grass species are able to cope with
the low nutrient status, exposure to full sun and severe soil compaction of
road and loading-bay centres. Vines may also colonize these areas by
spreading out across the soil surface from the forest edge. A climber blan-
ket may well overwhelm any seedlings that do find a foothold. The ecology
of climbers is poorly understood, yet it is clear that logged forest provides
an ideal habitat. Scrambling bamboos and vines such as Merremia spp. cre-
ate a serious silvicultural problem, suppressing tree regeneration in heavily
logged forests throughout South-East Asia (Figure 3.5). Yap et al. (1995)
found that 20 years after logging, 76% of trees (>10 cm dbh) were infested
with climbing bamboo and other vines in a Sabah forest.
Lateral penetration of sunlight from large open areas into the surround-
ing forest often encourages dense pioneer growth along road verges and
around the margins of loading bays. These areas may also benefit from an
accumulation of topsoil and logging debris pushed from the cleared area.
Large quantities of nutrient-rich tree bark are often stripped from logs
before transport and will provide local soil enrichment.
(b) Skid tracks

Bulldozers create a narrow corridor of heavily compacted soil, bordered by
a ridge of unconsolidated topsoil and woody debris that has been scraped
from the track. The regeneration of plants on the track itself is constrained
by nutrient-deficient subsoil. However, the debris mounds that border these
tracks are rapidly colonized by pioneer plants. Skid tracks rarely create
large gaps in the forest canopy and many pioneer trees may struggle to sur-
vive once the forest canopy begins to close over again after logging.
Seedlings of climax species may survive at the sides of a track and will
grow rapidly in response to increased light levels. Abandoned skid tracks
are often adopted as animal pathways. This may mean that plants growing
along these routes suffer disproportionate herbivore damage.
(c) Felling gaps

The greatest opportunity for the regeneration of climax forest tree species
will be found in felling gaps where the soil surface has been little disturbed
and seedling banks remain largely intact.
Felling of large trees inevitably creates large gaps which give a competi-
tive advantage to species that are responsive to high light levels and are
able to grow rapidly. The most important factor, therefore, determining the
Figure 3.5 Vines often proliferate in logged forest and smother regenerating trees:
here, in Sabah, Merremia is overwhelming even fast-growing pioneer species.
composition of regeneration is which species are already present when log-

ging occurs. This is because any felling gap is likely to contain only a small
subset of the species present locally in the forest, either as dormant seed in
the soil seed bank or as a bank of seedlings. A forester can maximize the
chances of regeneration of a desirable species by ensuring that the seedling
bank is well stocked with these species before trees are cut. Competition
between plants is highly asymmetrical and those that are already well
established when a gap is opened have an enormous head start. Where the
desirable species have a light-demanding ecology, regeneration is likely to
be profuse if a relatively simple rule is observed: the forest should be
logged only when seedlings are abundant. Conversely, if there are few
desirable seedlings present at the time of felling it is almost inevitable that
trees of little commercial value will dominate the regenerating forest. It will
be difficult for seeds of climax species to germinate in open-gap conditions
and those which do establish will be at a significant disadvantage when
competing with large pre-existing plants.
One of the reasons why many modern logging operations fail to promote
adequate regeneration of commercially valuable species is that it is rarely
practical to postpone felling during periods when seedling stocks are low.
The timber processing industry is dependent on continuous supplies of logs
from the forest and would not be able to go without supplies until the next
fruiting occurred.
Where the forest canopy has been extensively disturbed, pioneer tree
species may grow more rapidly than most climax trees and suppress their
growth for many years. However, high irradiance alone may be insufficient
to trigger pioneer germination. Kennedy and Swaine (1992) found that on
average only 2.3% of the estimated seed bank in the top 5 cm of soil ger-
minated when gaps were opened in a Sa bah forest. Seed germination was
increased markedly when leaf litter was removed or the soil scarified. A
similar effect was noted by Raich and Christensen (1989) who found that
seedlings of many pioneer tree species were restricted to gaps in which soil
erosion or disturbance had occurred. Climax species may therefore avoid
widespread competition from pioneer trees in those parts of the forest
where the topsoil has not been disturbed.
3.4 THE HYDROLOGICAL CYCLE IN TROPICAL RAIN FORESTS

It is dangerous to generalize about much in the humid tropics. It is a cli-
matic region characterized by diversity of soils and topography. However,
there is one unifying feature: frequent and heavy rainfall. Intense rain-
storms contribute a very large proportion of total annual rainfall. Typically
over 40% of rainstorms have an intensity that exceeds 25 mmlh, compared
with less than 5% in temperate areas. Intensities exceeding 150 mmlh are
common. Comparisons by Douglas (1969) of tropical and extra-tropical
catchments lead to the conclusion that contrasts in the quantity and distri-
bution of rainfall, rather than temperature or any inherent site factor, are
the most important influences on denudation rates.
There are two important sources of erosion by rainfall: rain-splash and
surface-wash erosion. The former is caused when raindrops impact with
the soil surface and soil particles are dislodged; the latter when soil is
entrained by water flowing over the soil surface.
The leafy canopy of an undisturbed forest intercepts almost all of the
rainfall. The majority of this then drips or flows its way down the trees to
the forest floor. However, a significant proportion (typically 10-25%) is
evaporated back into the atmosphere from the canopy. The proportion lost
in this way depends very much on the intensity and duration of rainfall. A
smaller proportion of water from large or intense storms is lost in this way.
Although the thick forest canopy intercepts the high-intensity rainfall,
contrary to popular belief it does not necessarily reduce the erosive power
of incoming rain (Bruijnzeel and Critchley, 1994). Water accumulates on
leaf surfaces where it coalesces to form drops that are up to twice as large
as the original raindrops. Drips falling from the canopy will have higher
kinetic energy than rainfall if they fall more than about 6 m. In undis-
turbed rain forest there is often very sparse ground vegetation due to the
low light levels. It is therefore principally the litter layer that absorbs the
The effects of logging on forest hydrology and erosion 65
impact of rainfall. This emphasizes the importance of keeping soil distur-
bance to an absolute minimum during logging.
On reaching the forest floor a proportion of rainwater infiltrates the soil
via the leaf litter layer. The topsoil of undisturbed rain forest is typically
rich in organic matter. Water that infiltrates the soil surface slowly per-
colates down slope. Infiltration rates vary considerably with soil type but
almost all rates are low when compared with rainfall intensity. However,
the sponge effect of the litter layer and the ponding of water in and around
surface roots prevents water from coalescing and starting to flow rapidly
(Walsh and Voigt, 1977). Surface flow of water occurs only over limited
areas in undisturbed forest after intense rainfall of long duration.
A significant proportion of water that enters the soil profile is lost due to
root uptake. The rate of water loss by this mechanism is highly variable
and dependent on local weather conditions. It would not be unusual for in
excess of 1000 mm per year to be taken up in forests where there is no sig-
nificant seasonal water shortage (Bruijnzeel and Critchley, 1994).
Measurements by Leopoldo et al. (1995) near Manaus in the Brazilian
Amazon indicate that over 67% of a mean annual rainfall of 2200 mm was
lost through evapotranspiration.
3.5 THE EFFECTS OF LOGGING ON FOREST HYDROLOGY AND

EROSION
It is likely (though the experimental evidence is very thin on the ground)
that due to the decrease in evapotranspirative losses of water there will be
a greater volume of water draining from logged catchments (Bruijnzeel and
Critchley, 1994). Where vegetation cover and organic rich topsoil have
been lost and soils have been compacted, heavy rainfall is likely to generate
storm-water surface run-off. As much as 70% of the rain that falls on such
compacted soils will run off rather than soaking in. Roads and skid tracks,
especially those on steep slopes, will act as gutters and collect run-off in
fast-flowing rivulets, channelling it quickly into rivers and streams. As a
result, logging is almost always shown to increase peak storm discharges.
However, rapid re-growth of secondary vegetation typically stabilizes the
hydrology of logged forests after 5-10 years (Bruijnzeel and Critchley,
1994).
Leaf litter and a dense surface root mat give undisturbed tropical rain
forest one of the lowest rates of surface erosion of any land use. Disruption
of litter and root mats by logging inevitably increases sediment losses.
Splash erosion will increase dramatically where vegetation has been
removed. An increase in water flowing over the soil surface will also
increase wash erosion. Mass wasting may increase where removal or dam-
age of trees on steep slopes damages the root network, but removal of large
trees may actually reduce the loading on deeply weathered slopes and
reduce the risk of deep-seated slope failure. It would appear that com-
pacted roads and loading bays often form a permanent source of increased
run-off and erosion. Rain-splash erosion is dramatically reduced once bare
soil is protected by vegetation. Skid tracks and felling gaps may temporar-
ily increase erosion rates by up to 20 times, but this might well be a rela-
tively short-lived phenomenon due to rapid regeneration of a plant canopy.
Not all eroded material is immediately lost from the forest. Erosion will
frequently redistribute sediment in a catchment to a temporary store fur-
ther down slope. Douglas et at. (1992) have shown that in a Bornean
catchment much sediment is redeposited in small-order stream channels
after logging. It may take many years of extreme events for this sediment to
be evacuated. It may well be the reduced channel capacity due to sedimen-
tation that has triggered speculation about increased risk of flooding fol-
lowing logging, rather than an increase in water yield. Temporary storage
of sediment makes it extremely difficult to estimate how quickly erosion is
reduced by forest regeneration. Sediment may be flushed from temporary
stores by extreme events for decades after the forest has ceased to be a sub-
stantial sediment source.
A number of simple precautions can significantly reduce erosion losses.
Roads and skid tracks should not cross streams as this creates a major sed-
iment source. Wherever possible, tracks and roads should not climb steep
gradients. Forests along the banks of streams and on steep slopes should
not be harvested because of their importance in protecting water and soil
resources.
Concern is often expressed that logging will lead to the loss of plant
nutrients from a catchment and threaten the long-term sustainability of a
forestry operation. One of the few generalizations that can be made from
the literature is that rates of input of many important plant nutrients are
extremely low in undisturbed forest. Atmospheric sources may be more
important than inputs from mineral weathering. It is clearly not sustain-
able for nutrient loss due to timber extraction, erosion and accelerated
decomposition of damaged vegetation to exceed the nutrient gains that
may be made from weathering and atmospheric inputs over the rotation
period. In areas that are geologically young, geomorphologically active and
close to the sea (for example much of South-East Asia), rates of nutrient
input may be sufficient to counterbalance extraction losses. Nitrogen and
organic carbon may recover rapidly after logging, but other nutrients such
as calcium and potassium make take much longer.
Most of the nutrient capital in the above-ground biomass is found in
leaves, young stems and bark. Leaves and twigs are always left in the forest
and often bark is also stripped from logs in the forest, as it can present a
disposal problem at the sawmill. The amounts of nutrients removed per log
are therefore small. However, there are very few humid tropical environ-
ments where the nutrient capital will not be significantly depleted by the
Conclusions 67
removal of large volumes of timber on a short rotation. Nutrient supply
may therefore apply a very significant long-term constraint to highly pro-
ductive forest management systems.
3.6 IMPACT ON ANIMAL POPULATIONS

Many animal species have been found to be surprisingly resilient to log-
ging. This is because many are relatively mobile and are able to avoid areas
which are being actively logged, whilst others are able to seek refuge in
residual patches of undisturbed forest. Almost all vertebrate species present
prior to logging had survived in or re-colonized a Malaysian dipterocarp
rain forest 12 years after it was logged (Johns, 1988). However, significant
changes in population structure have been observed which reflect changes
in resource abundance in logged forest. A small number of species tend to
become abundant. Many generalist herbivore species find young leafy
growth both abundant and accessible in a regenerating forest. Struhsaker
et al. (1996) found that areas of a Ugandan forest that had been logged
very heavily attracted elephants. The elephants browsed on the herbaceous
tangle and suppressed regeneration of forest trees. Although conditions in
logged forest may favour generalist species, species with highly specialized
habitat requirements may decline. Lambert (1992) reported that nectar-
ivorous and opportunistic frugivorous birds became significantly more
abundant in logged forest in Sabah. In contrast, understorey flycatchers,
woodpeckers and trogons were prone to local extinction.
Almost all species are likely to survive a single logging event if a signifi-
cant area of a coupe is set aside as a reserve. These residual undisturbed
stands create refuges from which animals and plants can re-colonize the
harvested area. This might include steep slopes and buffer zones around
streams, set aside for soil and water conservation but which can also act as
important wildlife corridors for the re-colonization of recently logged
areas.
3.7 CONCLUSIONS
In most parts of the humid tropics, if close attention is paid to the biologi-
cal constraints on forest regeneration, it is entirely possible to log tropical
rain forest without causing permanent ecosystem degeneration. Sadly, it is
very rarely ecology and almost always the demands of the market that
determine forest practice. There can be no question that current forestry
practices have substantial impacts on the structure and composition of
tropical rain forests. These may be reduced or partially ameliorated by
good practices, but they can never be avoided altogether. Logged forest will
inevitably have a structure and composition that reflects the processes of
recovery from this high-magnitude disturbance event. However, the value
of even badly logged forest for conservation is high.
Most remaining undisturbed tropical rain forest is almost certainly des-
tined to be logged. The world-wide network of protected tropical forest
areas provides inadequate protection for the conservation of a very large
proportion of tropical rain forest species. Foresters and production forests
should and must form a vital part of any strategy for the conservation of
tropical biodiversity. In many tropical countries, significant changes in atti-
tude to the exploitation of natural resources will be needed before the
widespread adoption of good forestry practices is likely. Most logging con-
tinues to degrade the conservation quality of production forests. Bruenig
(1996) has lamented the fact that devastatingly harmful logging practices
have succeeded in perverting the image of production forestry to the point
where it is now widely perceived as a threat both to the environment and
humanity. A significant change of attitude is also needed from the conser-
vation community, to avoid continuing conflict in which foresters are pil-
loried for having destroyed forests which they have logged. It is vital that if
tropical rain forests are to continue to exist outside formal reserves they
must be shown to be more valuable than deforested land. If logged forest
continues to be perceived as destroyed and therefore of no conservation
value, a vital ecosystem may well be allowed to spiral into irreparable
degeneration.
REFERENCES
Abdulhadi, R., Kartawinata, K. and Sukardjo, S. (1981) Effects of mechanized log-
ging in the lowland dipterocarp forest at Lempake, East Kalimantan. Malaysian
Forester, 44, 407-418.
Appanah, S. and Putz, F. E. (1983) Climber abundance in virgin dipterocarp forest
and the effect of pre-felling climber cutting on logging damage. Malaysian
Forester, 47, 335-342.
Appanah, S., Weinland, G. and Cheah, L. C. (1990) Will the management systems
for hill dipterocarp forest stand up? Journal of Tropical Forest Science, 3,
140-158.
Ashton, P. M. S. and Berlyn, G. P. (1992) Leaf adaptations of some Shorea species
to sun and shade. New Phytologist, 121, 587-596.
Baillie, I. C. (1976) Further studies on drought in Sarawak, East Malaysia. Journal
of Tropical Geography, 43, 20-29.
Baur, G. N. (1968) The Ecological Basis of Rain Forest Management, Forestry
Commission of New South Wales, Sydney.
Beard, J. S. (1946) The Mora forests of Trinidad, British West Indies. Journal of
Ecology, 33,173-192.
Bertault, ]. G. and Sist, P. (1995) The effects of logging in natural forests. Bois et
Forets des Tropiques, 245,5-20.
Brokaw, N. V. L. and Grear,]. S. (1991) Forest structure before and after Hurricane
References 69
Hugo at three elevations in the Luquillo Mountains, Puerto Rico. Biotropica, 23,
386-392.
Brown, S. and Lugo, A. E. (1990) Tropical secondary forests. Journal of Tropical
Ecology, 6, 1-32.
Bruenig, E. F. (1964) A study of damage attributed to lightning in two areas of
Shorea albida forest in Sarawak. Commonwealth Forestry Review, 43, 134-144.
Bruenig, E. F. (1996) Conservation and Management of Tropical Rainforests: An
Integrated Approach to Sustainability, CAB International, Wallingford, UK.
Bruijnzeel, L. A. and Critchley, W. R. S. (1994) Environmental impacts of logging
moist tropical forests, IHP Humid Tropics Programme Series No.7, UNESCO,
Paris.
Burgess, P. F. (1970) An approach towards a silvicultural system for the hill forest
of the Malay Peninsula. Malaysian Forester, 33, 126-134.
Clements, F. E. (1916) Plant Succession: An Analysis of the Development of
Vegetation, Carnegie Institution, Washington.
Condit, R., Hubbell, S. P. and Foster, R. B. (1996) Changes in tree species abun-
dance in a Neotropical forest, impact of climate change. Journal of Tropical
Ecology, 12,231-256.
Crome, F. H. J., Moore, L. A. and Richards, G. C. (1992) A study of logging dam-
age in upland rainforest in north Queensland. Forest Ecology and Management,
49,1-29.
Denslow, J. S. (1980) Gap partitioning among tropical rain forest trees. Biotropica,
12(suppl.),47-55.
Douglas, I. (1969) The efficiency of tropical denudation systems. Transactions of
the Institute of British Geographers, 46, 1-16.
Douglas, I., Spencer, T., Greer, T., Bidin, K., Sinun, W. and Wong, W. M. (1992)
The impact of selective commercial logging on stream hydrology, chemistry and
sediment loads in the Ulu Segama rain forest, Sabah, Malaysia. Philosophical
Transactions of the Royal Society of London, Series B, 335, 397-406.
FAO (1993) Forest Resources Assessment 1990, Tropical Countries, FAO Forestry
Paper 112, Food and Agriculture Organization, Rome.
Fox, J. E. D. (1968) Logging damage and the influence of climber cutting prior to
logging in the lowland dipterocarp forest of Sabah. Malay Forester, 31,
326-347.
Freedon, A. L. and Field, C. B. (1991) Leaf respiration in Piper species native to a
Mexican rainforest. Physiologica Planatarum, 82, 85-89.
Gause, G. F. (1934) Experimental analysis of Vito Volterra's mathematical theory
of the struggle for existence. Science, 79, 16-17.
Gentry, A. H. and Vasquez, R. (1988) Where have all the Ceibas gone? A case his-
tory of mismanagement of a tropical forest resource. Forest Ecology and
Management, 23, 73-76.
Gerwing, J. J., Johns, J. S. and Vidal, E. (1996) Reducing waste during logging and
log processing, forest conservation in eastern Amazonia. Unasylva, 187, 17-25.
Gomez-Pompa, A. and Burley, F. W. (1991) The management of natural tropical
rain forests, in Rain Forest Regeneration and Management, (eds Gomez-Pompa,
A., Whitmore, T. C. and Hadley, M.), Man and the Biosphere Series No.6,
UNESCO, ParislParthenon, Carnforth, UK, pp. 3-18.
Grime, J. P. (1977) Evidence for the existence of three primary strategies in plants
and its relevance to ecological and evolutionary theory. American Naturalist,
111,1169-1194.
Grubb, P. J. (1977) The maintenance of species richness in plant communities: the
importance of the regeneration niche. Biological Review, 52, 107-145.
Hammond, D. S., Gourlet-Fleury, S., van der Hout, P., ter Steege, H. and Brown, V.
K. (1996) A compilation of known Guianan timber trees and the significance of
their dispersal mode, seed size and taxonomic affinity to tropical rain forest man-
agement. Forest Ecology and Management, 83, 99-116.
Hawthorne, W. D. (1993) Forest Regeneration after Logging: Findings of a Study
in the Bia South Game Production Reserve, Ghana, ODA Forestry Series No.3,
Overseas Development Administration, London.
Hubbell, S. P. and Foster, R. B. (1986) Commonness and rarity in a neotropical for-
est, implications for tropical tree conservation, in Conservation Biology, the
Science of Scarcity and Diversity, (ed. M. E. Soule), Sinauer Associates,
Sunderland, Massachusetts, pp. 205-231.
Hutchinson, G. E. (1957) The multivariate niche. Cold Spring Harbor Symposium
on Quantitative Biology, 22, 415-421.
Johns, A. D. (1988) Effects of "selective" timber extraction on rain forest structure
and composition and some consequences for frugivores and folivores.
Jones, E. W. (1952) Natural regeneration of beech abroad and in England.
Quarterly Journal of Forestry, 46, 75-82.
Jones, E. W. (1956) Ecological studies on the rain forest of southern Nigeria. IV.
The plateau forest of the Okomo Reserve. Journal of Ecology, 43, 564-594.
Jusoff, K. (1988) Soil compaction from off-road transportation machine on tropical
hill forest land. Pertanika, 11, 31-37.
Kasenene, J. M. and Murphy, P. G. (1991) Post-logging tree mortality and major
branch losses in Kibale Forest, Uganda. Forest Ecology and Management, 46,
295-307.
Kennedy, D. N. and Swaine, M. D. (1992) Germination and growth of colonizing
species in artificial gaps of different sizes in dipterocarp rain forest. Philosophical
Transactions of the Royal Society London, Series B, 335, 357-366.
Lambert, F. (1991) The conservation of fig-eating birds in Malaysia. Biological
Conservation, 58, 31-40.
Lambert, F. (1992) The consequences of selective logging for Bornean lowland for-
est birds. Philosophical Transactions of the Royal Society London, Series B, 335,
443-457.
Ledig, F. T. (1992) Human impacts on genetic diversity in forest ecosystems. Gikos,
63,87-108.
Leopolda, P. R., Franken, W. K. and Nova, N. A. V. (1995) Real evapotranspira-
tion and transpiration through a tropical rain forest in central Amazonia as esti-
mated by the water balance method. Forest Ecology and Management, 73,
185-195.
Livingstone, D. A. (1982) Quaternary geography of Africa and the refuge theory, in
Biological Diversification in the Tropics, (ed. G. T. Prance), Columbia University
Press, New York.
de Madron, L. D. (1994) Tree mortality in a primary forest in French Guiana. Bois
et Forets des Tropiques, 239, 43-57.
References 71
Magnusson, W. E., Lima, A. P. and Lima, O. de (1996) Group lightning mortality
of trees in a Neotropical forest. Journal of Tropical Ecology, 12, 899-903.
Malingreau, ]. P., Stephens, G. and Fellows, L. (1985) Remote sensing of forest
fires, Kalimantan and north Borneo, in 1982-83. Ambia, 14,314-321.
Maimer, A. and Grip, H. (1990) Soil disturbance and loss of infiltrability caused by
mechanized and manual extraction of tropical rainforest in Sabah, Malaysia.
Forest Ecology and Management, 38, 1-12.
Marn, H. M. and ]onkers, W. B. ]. (1982) Logging damage in tropical high forest,
in Tropical Forests, Source of Energy through Optimization and Diversification,
(ed. P. B. L. Srivastava), Universiti Pertanian, Malaysia.
Matthews,]. D. (1989) Silvicultural Systems, Oxford University Press, Oxford.
Murawski, D., Gunatilleke, I. A. U. N. and Bawa, K. S. (1994) The effects of selec-
tive logging on inbreeding in Shorea megistophylla (Dipterocarpaceae) from Sri
Lanka. Conservation Biology, 8, 997-1002.
Nelson, B. W., Kapos, V., Adams, ]. B., Oliveira, W. ]., Braun, O. P. G. and do
Amaral, I. L. (1994) Forest disturbance by large blowdowns in the Brazilian
Amazon. Ecology, 75,853-858.
Nepstad, D. c., Uhl, C. and Serrao, E. A. S. (1991) Recuperation of a degraded
Amazonian landscape, forest recovery and agricultural restoration. Ambia, 20,
248-255.
Nussbaum, R., Anderson, ]. and Spencer, T. (1995) Factors limiting the growth of
indigenous tree seedlings planted on degraded rainforest soils in Sabah,
Malaysia. Forest Ecology and Management, 74, 149-159.
Oberbauer, S. F. and Strain, B. R. (1984) Photosynthesis and successional status of
Costa Rican rain forest trees. Photosynthesis Research, 5, 227-232.
d'Oliveira, M. V. N. and Braz, E. M. (1995) Reduction of damage to tropical moist
forest through planned harvesting. Commonwealth Forestry Review, 74,
208-210.
Pinard, M. A., Putz, F. E., Tay,]. and Sullivan, T. E. (1995) Creating timber harvest
guidelines for a reduced-impact logging project in Malaysia. Journal of Forestry,
93,41-45.
Plumptre, R. A. (1993) A Survey of Sawmilling and Related Industries in Belize,
Consultancy Report No.7, Forest Planning and Management Project, Ministry
of Natural Resources, Belize.
Posey, D. A. (1994) Environmental and social implications of pre- and post-contact
situations on Brazilian Indians, the Kayapo and a new Amazonian synthesis, in
Amazonian Indians from Prehistory to the Present, Anthropological
Perspectives, (ed. A. Roosevelt), University of Arizona Press, Tucson.
Press, M. c., Brown, N. D., Barker, M. G. and Zipperien, S. W. (1996)
Photosynthetic responses to light in tropical rain forest seedlings, in The Ecology
of Tropical Tree Seedlings, (ed. M. D. Swaine), Man and the Biosphere Series
No. 17, Parthenon, Carnforth, UK, pp. 41-54.
Pryor, S. N. and Savill, P. S. (1986) Silvicultural Systems for Broadleaved Woodland
in Britain, OFI Occasional Papers No. 32, Oxford Forestry Institute, Oxford.
Putz, F. E. and Appanah, S. (1987) Buried seeds, newly dispersed seeds and the
dynamics of a lowland forest in Malaysia. Biotropica, 19, 326-333.
Raich, ]. W. and Christensen, N. C. (1989) Malaysian dipterocarp forest, tree
seedling and sapling species composition and small-scale disturbance patterns.
National Geographic Research, 5, 348-363.
Raich, J. W. and Gong, W. K. (1990) Effects of canopy opening on tree seed germi-
nation in a Malaysian dipterocarp forest. Journal of Tropical Ecology, 6,
203-217.
Riddoch, I., Grace, J., Fasehun, F. E., Riddoch, B. and Ladipo, D. O. (1991)
Photosynthesis and successional status of seedlings in a tropical semi-deciduous
rain forest in Nigeria. Journal of Ecology, 79, 491-503.
Rowe, R. E. (1995) The population biology of Trocheitiopsis, a genus endemic to
St Helena, PhD thesis, University of Oxford.
Salick, J., Mejia, A. and Anderson, T. (1995) Non-timber forest products integrated
with natural forest management, Rio San Juan, Nicaragua. Ecological
Applications, 5, 878-895.
Salo, J. S. and Kalliola, R. J. (1991) River dynamics and natural forest regeneration
in the Peruvian Amazon, in Rain Forest Regeneration and Management, (eds A.
Gomez-Pompa, T. C. Whitmore and M. Hadley), Man and the Biosphere Series
No.6, UNESCO, ParislParthenon, Carnforth, UK, pp. 245-256.
Scholes, J. D., Press, M. C. and Zipperlen, S. W. (1997) Differences in light energy
utilisation and dissipation between dipterocarp rain forest tree seedlings.
Oecologia, 109,41-48.
Skorupa, J. P. and Kasenene, J. M. (1984) Tropical forest management, can rates of
natural tree falls help guide us? Oryx, 18, 96-10l.
ter Steege, H., Bokdam, c., Boland, M., Dobbelsteen, J. and Verburg, I. (1994) The
effects of man made gaps on germination, early survival and morphology of
Chlorocardium rodiei in Guyana. Journal of Tropical Ecology, 10,245-260.
ter Steege, H., Boot, R. G. A., Brouwer, L. c., Caesar, J. c., Renske, C. E.,
Hammond, D. S., Haripersaud, P. P., van der Hout, P., Jetten, V. c., van Kekem,
A. J., Kellman, M. A., Kahn, Z., Polak, A. M., Pons, T. L., Pulles, J.,
Raaimakers, D., Rose, S. A., van der Sanden, J. J. and Zagt, R. J. (1996) Ecology
and Logging in a Tropical Rain Forest in Guyana, with Recommendations for
Management, Tropenbos Series 14, Tropenbos Foundation, Wageningen.
Struhsaker, T. T., Lwanga, J. S. and Kasenene, J. M. (1996) Elephants, selective log-
ging and forest regeneration in the Kibale Forest, Uganda. Journal of Tropical
Ecology, 12,45-64.
Styles, B. T. (1972) The flowering biology of the Meliaceae and its bearing on tree
breeding. Silvae Genetica, 21,175-182.
Swaine, M. D. and Whitmore, T. C. (1988) On the definition of ecological species
groups in tropical rain forests. Vegetatio, 75, 81-86.
Swaine, M. D., Lieberman, D. and Putz, F. E. (1987) The dynamics of tree popula-
tions in tropical forest, a review. Journal of Tropical Ecology, 3, 359-366.
Swaine, M. D., Agyeman, V. K., Kyereh, B., Orgle, T. K., Thompson, J. and
Veenendaal, E. M. (1997) Ecology of Forest Trees in Ghana, ODA Forestry
Series No.7, Overseas Development Administration, London.
Troup, R. S. (1952) Silvicultural Systems, 2nd edn (ed. E. W. Jones), Oxford
University Press, Oxford.
Turner, I. M. (1990) The seedling survivorship and growth of three Shorea species
in a Malaysian tropical rain forest. Journal of Tropical Ecology, 6, 469-478.
Uhl, C. and Vieira, I. C. G. (1989) Ecological impacts of selective logging in the
Brazilian Amazon, a case study from the Paragominas region of the state of Para.
Biotropica, 21,98-106.
References 73
Uhl, c., Verissimo, A., Mattos, M., Brandino, Z. and Vieira, 1. (1991) Social, eco-
nomic and ecological consequences of selective logging in an Amazon frontier,
the case of Tailandia. Forest Ecology and Management, 46, 243-273.
Verissimo, A., Barreto, P., Mattos, M., Tarifa, R. and Uhl, C. (1992) Logging
impacts and prospects for sustainable forest management in an old Amazonian
frontier, the case of Paragominas. Forest Ecology and Management, 55,
169-199.
Verissimo, A., Barreto, P., Tarifa, R., and Uhl, C. (1995) Extraction of a high-value
natural resource in Amazonia, the case of mahogany. Forest Ecology and
Walsh, R. P. D. (1996) Drought frequency changes in Sabah and adjacent parts of
northern Borneo since the late nineteenth century and possible implications for
tropical rain forest dynamics. Journal of Tropical Ecology, 12, 385-407.
Walsh, R. P. D. and Voigt, P. (1977) Vegetation litter, an underestimated variable in
hydrology and geomorphology. Journal of Biogeography, 4, 253-254.
Whitmore, T. C. (1984) Tropical Rainforests of the Far East, 2nd edn, Clarendon
Press, Oxford.
Whitmore, T. C. and Silva, J. N. M. (1990) Brazilian rain forest timbers are mostly
very dense. Commonwealth Forestry Review, 69, 87-90.
Whittaker, R. ]., Walden, J. and Hill, J. (1992) Post-1883 ash fall on Panjang and
Sertung and its ecological impact. GeoJournal, 28, 153-171.
Woodward, C. L. (1996) Soil compaction and topsoil removal effects on soil prop-
erties and seedling growth in Amazonian Ecuador. Forest Ecology and
Management, 82,197-209.
Wyatt-Smith,]. (1963) Manual of Malayan Silviculture for Inland Forests (2 vols),
Malayan Forest Records No. 23, Forest Research Institute, Malaysia.
Yap, S. W., Chak, C. V., Majuakim, L., Anuar, M. and Putz, F. E. (1995) Climbing
bamboo (Dinochloa spp.) in Deramakot forest reserve, Sabah, biomechanical
characteristics, modes of ascent and abundance in logged-over forest. Journal of
Tropical Forest Science, 8, 196-202.
Zipperien, S. W. and Press, M. C. (1996) Photosynthesis in relation to growth and
seedling ecology of two dipterocarp rain forest tree species. Journal of Ecology,
84, 863-876.
-4------
Forest people and conservation
initiatives: the cultural context of rain
forest conservation in West Africa
B. Sharpe
GEC/ESRC Rainforest Research Group, Department of Anthropology,
University College London, UK
In an influential account of rain forest, the botanist P. W. Richards has sug-

gested that Africa is the 'odd man out' in terms of human impact on rain
forest ecology (Richards, P. W., 1996). In this view, the rain forests of trop-
ical America and Asia are pristine or primary forest, whereas those of
Africa (and especially West Africa) show clear signs of human disturbance
and management. Ecological differences are paralleled by differences in the
relationships between forest people and the environment in Asia, America
and Africa. In Asia and tropical America, forest societies are commonly
perceived as 'indigenous people', racially or culturally distinct from the
encroaching wider modern society. Following on from this, the anthropolo-
gist's role in conservation has been primarily one of advocate for indigen-
ous rights, and of documenting cultures which are, like the forest, under
threat. Thus in Amazonia, anthropologists have long been engaged in
ethnobotanical inventory work, in establishing first contacts with tribal
peoples, and in documenting their plight to the outside (western) world of
non-governmental organizations (NGOs) and international agencies. (For
a useful review of anthropology and the conservation of biodiversity, see
Orlove and Brush, 1996.)
More recently, however, the pristine condition of the forest, the demo-
graphic history of its population, and the role of anthropologists in mediat-
ing relations between forest people and the outside world have all come
under some challenge. Thus the population density of much of Amazonia
appears to have been far higher before the conquest, whilst anthropologists
have begun to point out the existence of distinctive 'creolized' cultures in
Amazonia: cultures such as Caboclo society (Nugent, 1993) which are not
indigenous but have considerable historical depth. Recently, anthropologists
76 Forest people and conservation initiatives
and others (e.g. Colchester, 1993) have turned their view from the analysis
of the communities or tribes of indigenous people to the interplay of insti-
tutions (government and private companies) and social movements (such
as environmental NGOs, as well as land-rights movements, etc.) and how
these impact upon the forest and its people (Burnham et al., 1995; Sharpe,
1996). Using the insights from this new perspective, Africa is perhaps not
so much the odd man out, but rather one regionally specific 'socio-ecology'
amongst many. Perhaps the perspective adopted in this chapter will serve
as a bridge or middle ground between often excessive over-generalizations
and extrapolations from global processes to particular countries or regions
(wildly variable estimates of deforestation that are characteristic of much
environmental planning may be compared with the equally obsessive con-
centration on the local by many anthropologists and green campaigners).
One aim of the chapter is to suggest that anthropology has sufficient com-
parative data to generate the same kinds of regional generalizations as, for
example, forestry (see Guyer and Lambin, 1994). Relations between people
and forest in West Africa and Asia are as different from each other as are
the multi-species forest of Africa from the dipterocarp forest of South-East
Asia. A further aim is to argue that, in the rush to establish conservation in
West and Central Africa, the cultural uniqueness of forest-society relations
in the region was lost. What is needed for effective conservation is a recog-
nition of West African forest peoples' world views, and a realization of
how these cosmologies have incorporated an understanding of the modern
world.
Conservationists are relatively new entrants in the 'fantastic invasion' of
the West African rain forests by Europeans (Marnham, 1980; Colchester,
1993). Large areas of West and Central Africa were nominated as national
parks and protected areas in the 1980s (see Figure 4.1). Unlike the savanna
game parks of East Africa, conservation projects in the rain forest have rel-
atively little historical experience either of managing the environment or of
dealing with local people. Nowhere in forested West Africa is there a
National Park that is more than 15 years old. Thus it was perhaps
inevitable that when rain forest conservation became an internationally
prominent issue, spurred on by the perceived urgency of the task of pre-
serving biodiversity, conservation organizations acted without paying suffi-
cient attention to local West African society.
In large part, conservation organizations brought to West Africa a series
of myths about forest peoples and the forest environment which are drawn
from other places in the world, or from other times, rather than the reality
experienced by West African forest peoples themselves. Such views are the
stock-in-trade of both conservation campaigns and the western media.
Perhaps the key theme in these outsiders' stereotypes is the idea of 'indige-
nous people' as custodians living in harmony with the rain forest: a theme
deriving from Amazonian and South-East Asian experience. A second
major theme is that rain forest is pristine and natural, a survival of a world
Forest people and conservation initiatives 77
Figure 4.1 Location of major rain forest conservation projects.
before human interference in the environment. A third is that the forest is

threatened by the primitive agricultural methods of an ever-expanding pop-
ulation. All of these themes effectively ignore biological, archaeological
and anthropological work On West African forest societies and cultures
which document a rich and long-standing relationship between forest and
human activity (Keay, 1971; Shaw, 1978).
The proliferation of rain forest conservation projects, and their apparent
indifference to the actual cultures of forest people, can be illustrated from
research carried out in Cameroon and Sierra Leone. These two countries
represent, in microcosm, almost all of the ecological, socio-cultural, colo-
nial and post-colonial processes which have affected the forest societies of
West and Central Africa.
Ecologically, Sierra Leone and Cameroon represent the Guinean and
Congo forest regions (see Figure 4.1). Each country straddles two or more
wider cultural divisions: Manding and non-Manding cultures in Sierra
Leone; Cross River and Congo Basin societies in Cameroon (Vansina,
1990). Each contains contrasting forms of local political and social organi-
zation (such as chiefdoms and uncentralized societies, divergent forms of
ethnic diversity or different processes for incorporating strangers). Sierra
Leone and western Cameroon were subject to British colonial rule, whilst
eastern Cameroon was ruled by the French. (Cameroon was originally a
German colony, but little of the heritage of German colonialism survives
apart from plantation concessions, mission stations and government build-
ings.) In each case colonial rule firmly established francophone and anglo-
phone legal frameworks for ownership and control of the forest. Sierra
Leone is largely logged secondary forest with important enclaves of older
secondary forest in the Gola region (Fyfe, 1962; Davies and Richards,
1991) whilst Cameroon is amongst the top ten sites in Africa for large-
scale logging for export, as well as having a large number of high-visibility
conservation projects. Finally, in spite of war, Sierra Leone may (as of
November 1996) be beginning to climb out of the political and economic
mess of the 1980s, whereas Cameroon is undergoing rigorous economic
restructuring under World Bank tutelage.
In each country, our research (see Acknowledgement) sought to capture
the whole range of interests and cultures concerned with forest: from local
villagers in forest enclaves through to urban popular culture, from the
views of the rural poor through to political elites and other groups active in
civil society, from studies of local peoples' use of forest resources through
to case studies of the activities of conservation NGOs, timber companies,
ministries and international agencies concerned with forest exploitation
and conservation. This chapter, however, concentrates on local cultures,
traditional and modern, and how these articulate with the general premises
of rain forest conservation as enunciated by conservation organizations
and by project workers on the ground.
As we will see, the West and Central African situation does not marry
easily with the stereotypes employed by conservation projects. The purpose
of this chapter, therefore, is to outline some key themes regarding West
African cultural realities and their relevance for the future of the forest.
4.1 WHO ARE THE INDIGENES?

Rain forests of west and central Africa contain large populations ranging
from the modern remnants of large pre-colonial states (Ashanti in Ghana,
Dahomey in Benin), city states (the Yoruba cities and Benin in southern
Nigeria), through small chiefdoms in Sierra Leone and Liberia (Mende,
Gola), clan-based villages (Eastern Nigeria, West Cameroon), lineage-based
village societies without political leaders (eastern Cameroon) to band societies
of the archetypal indigenous people (so-called 'pygmies' such as the Baka).
All of these are societies and cultures whose history has developed in intim-
ate relation with the forest (see Figure 4.2). Yet few of this enormous popu-
lation are treated by conservationists or outsiders as 'indigenous people'.
Thus, recent World Bank documents on southern Cameroon class only the
pygmies as indigenes - 40-60 thousand people out of a population of 4-6
million inhabiting the forest zone (Winterbottom, 1992). Similarly, one of
the best general introductions to the rain forest in West Africa (specifically
dealing with Ghana and Cote d'Ivoire) nonetheless offers a model of
indigenous people living in harmony with the forest environment, but
swamped by immigrant slash-and-burn agriculturalists (Martin, 1991).
Such stereotypes penetrate the media and even the Internet: recently,
for example, there has been much e-mail posted on the threats to the
Who are the indigenes? 79
Figure 4.2 Cultures and countries cited in the text. 1, Mende (Sierra Leone);
2, Gola (Sierra LeonelLiberia borderlands); 3, Ashanti (Ghana); 4, Yoruba states
(Nigeria); 5, Cross River groups (Nigeria/Cameroon borderlands); 6, Mvae
(Cameroon); 7, Beti, Bulu, Ewondo, Fang (Cameroon, Equatorial Guinea, Gabon);
8, Makaa, Mkako, Ngoumba, Djem (Cameroon); 9, Baka (Cameroon);
10, Bakongo (Congo, Zaire); 11, Azande (Central African Republic, Zaire, Sudan).
environment posed by Liberian and Sierra Leonean refugees in Guinea (see

Black and Mohammed Sesay, 1996). In contrast to these views, ethno-
graphic and historical studies throughout the forest of West and Central
Africa document waves of migration or 'brownian motion' of villages for
at least a couple of millennia (Alexandre and Binet, 1958; Vansina, 1990).
Evidence of settlement in the forest often predates the historical memories
of the societies that live there now. In the Cross River valley, for example,
sculpted standing stones are included in local ritual even though their mak-
ers are unknown. Elsewhere in West Africa, forest cultures have myths of
pre-existing inhabitants (often dwarfs or fairies) who have been driven
deeper into the forest, or into hiding, by the current inhabitants. In Sierra
Leone, chimpanzees are considered almost the 'indigenous people' of the
forest, with a society and culture that is the mirror image of later human
settlers (Richards, 1995). Wherever they exist, the pygmies are viewed as
the 'first people' of the forest by Bantu-speaking villagers, though this does
not prevent villagers from both exploiting the pygmies for their forest
knowledge and despising them at the same time. But the pygmies are not
the only indigenes, from the viewpoint of groups which have arrived later.
Thus for example, in east Cameroon, linguistic groups such as Makaa,
Ngoumba or Djem are indigenous from the viewpoint of the Beti, Bulu,
Ewondo and Fang who migrated through the area in the 18th and 19th
centuries in their quest for trade on the coast (Burnham, 1994). In southern
Cameroon, as elsewhere in Central Africa (Evans Pritchard, 1971), the
later settlers attained political ascendancy. Perhaps more typically, the pro-
gressive settlement over time of groups in a region established an elaborate
hierarchy, from founders who first settled the area, to the most recent
immigrants. Such processes can still be seen at work in expanding towns of
central Ghana (Sharpe, 1991) or Sierra Leone (Davies and Richards,
1991). In parts of Zaire that are now almost ungoverned by the state,
immigrant diamond diggers, smugglers and refugees have actually sup-
ported the reinvention of the chiefly powers of the 'original' inhabitants, in
order to achieve a measure of political and communal order (MacGaffey,
1991). (Zaire became the Democratic Republic of Congo in April 1997;
there is evidence that the rebel movement led by Laurent Kabila was aimed
in the first instance at gaining control of the diamond areas.)
The opposite process can also occur, as many of the most densely
forested and least populated areas of west and central Africa are former
boundary wilderness - buffer zones between competing communities or
states in which depopulation resulted from war and colonial conquest
(Richards, 1996) or from epidemics such as sleeping sickness.
Contemporary immigrants to such regions bring their own cultures, assim-
ilating remnants of the former population. West African cultures' recogni-
tion of degrees and types of indigenousness has long been used as an
organizing principle for creating civilizations in a hostile environment.
4.2 FIRST THE FOREST: VILLAGES AND THE SOCIALIZATION

OF SPACE
Rain forest is a highly diverse mosaic of micro-environments. Due to tree
falls and ecological successions, edaphic and geological variation, and
human management, rain forest is characterized by patch ecology: com-
petition amongst species can lead to highly varied succession trajectories.
Parts of the forest may be almost mono specific in terms of the tree cover.
Elsewhere, highly varied botanical assemblages can result from growth of
species which kill competing members of their own species. (For more
details of these processes and the patch ecology that results see Goldsmith,
this volume.)
Humans, like other large mammals (gorilla, elephant), utilize these
environments selectively and maintain or manage them in the process.
Hunter-gatherers such as the Baka pygmies of south-east Cameroon follow
complex migratory patterns depending on the availability of fruits, insects,
honey and game, but also selectively manage wild yams and other staples,
thus increasing yields (Dounias, 1993). Settled agriculturalists utilize a sim-
ilar range of resources, either through symbiotic relations with pygmy
groups, or through expeditions from villages to the borders of village
First the forest: villages and the socialization of space 81
territories. There is evidence that apparently wild resources, such as oil-
palm groves and wild yams, may in fact have been maintained through
human activity. In many parts of West and Central Africa abandoned vil-
lage sites and farms continue to be managed for decades for their useful
crops. In the process, humans manage the regeneration of forest in
ecologically significant ways. Thus, for example, large areas in the north of
Korup National Park are abandoned settlements and farm sites or man-
aged wild resources (Unwin, 1915).
Village territories extend far beyond the boundaries of recognizable
farmland and bush fallow, and often define rights to utilize specific trees
(especially oil seeds such as moabi, Baillonella toxisperma) to gather fruits
and honey, or to hunt. Traditional West African legal systems typically
define ownership in terms of first use, once again emphasizing the pre-
eminence of the earliest settlers in a locality.
Entering the forest is never easy, however. Whilst old forest may be
almost devoid of understorey vegetation, the forest edge is often a tangled
thicket of saplings, lianas, bamboo and razor grass. For villagers with hand
tools the newest forest is often the most difficult to penetrate. For this rea-
son many societies value elephants, both for their trails into the forest and
as sources of culturally valued resources. Mende village histories, for
example, often begin with a hunter who kills an elephant and plants a new
rice variety found in the animal's stomach (Richards, 1996). In south-west
Cameroon, important men were believed to be able to turn into elephants,
perhaps reflecting their ability to keep open the complex trade paths which
linked settlements in the pre-colonial period, as well as the ambivalence of
power. Elsewhere, the human act of finding, or making, an entry into the
forest is often celebrated as the beginning of history. In southern
Cameroon, Gabon and Rio Muni (Equatorial Guinea), a widespread his-
tory traces the origins of the Beti, Bulu, Ewondo and Fang peoples to entry
into the forest by cutting a hole in the adzap tree (Guyer, 1984; d.
Thikalul, undated, which identifies adjap as B. toxisperma). In central
Ghana (Wenchi and Techiman) settlements dating to the 1870s were estab-
lished in former boundary wilderness on naturally occurring grass clearings
in the forest (Sharpe, 1991). In Cote d'Ivoire, government policy of the
1950s-1980s emphasized forest conversion to plantations as a means to
create a new society (parren and de Graaf, 1995). In 1993 a settler in old
forest in western Cameroon described to me how cutting down huge and
unstable Bombax and Ceiba pentandra (silk cotton) trees was an effective
means for protecting houses and clearing bush around the site of a new vil-
lage. In the mythological past and in the present day, clearing a new village
from the forest has heroic qualities.
Deep-forest villages are founded by pioneers on the frontiers of society
(Kopytoff, 1987). Such frontiers may not be unoccupied: almost all forest
societies claim to have met with previous (human or non-human) inhabitants
or evidence of their settlements. Hence new settlements occupy social
rather than geographical boundary zones. In the colonization of such
areas, hunters are the archetypal pioneers in myth and history, but detailed
ethnographic work documents various processes by which societies gener-
ate pioneer groups - most commonly groups which are disadvantaged in
village political and economic structures. In lineage-based societies, for
example, junior lineages are excluded from power and are thus predis-
posed to strike off into the forest and establish new settlements (Copet,
1977). Similarly, losers in contests for chiefly office have hived off and
established new chiefdoms, whilst the pre-colonial states of forest West and
Central Africa (Ashanti, Kongo) had quite systematic policies of establish-
ing aristocratic lineages amongst the surrounding peoples. The origins of
many forest societies can be traced to runaways and refugees during peri-
ods of war and economic collapse, a process that may be reappearing
today in Zaire, Sierra Leone and Liberia.
These patterns in the foundation of society have key cultural and envi-
ronmental correlates, which are expressed in spatial organization and in
village politics and concepts of power. Establishment of a village and the
clearing of farms is achieved through the socialization of space: the incor-
poration of forest powers into human society. Archetypally, the village
becomes surrounded by farms, then a penumbra of supernaturally mar-
ginal bush, beyond which lie the unknown spirits and powers of the forest
(Figure 4.3).
These concentric zones reflect degrees of human activity: from the fully
socialized space of the village, through farm and fallow, to the partly
socialized forest in which non-timber forest products, game and many
medicines are collected, and finally on into the unsocialized and little-
utilized forest itself.
In the world views of forest peoples, the forest is not wild or natural, in
opposition to society. Rather society and forest are mutually interdepen-
dent on one another. Whilst the forest certainly contains dangerous, some-
times whimsical and tricky powers, human beings can connect to these
powers through luck or through ritual and can harness them for social pur-
poses: human health and fertility, social order and security, the proper rela-
tions between men and women, youth and elders, indigenes and
immigrants. In order to achieve these aims, cultures of West and Central
Africa have developed a variety of means for bringing the forest into the
village. Amongst the Beti in Cameroon, women are considered to be 'of the
forest' whilst men, paradoxically, are not - even though the men are the
most likely to enter it. Accordingly the dangerous, potentially disruptive
powers of women must be controlled. The forest thus penetrates even the
most basic human gender relations. Elsewhere, the entry of the forest into
the human world is managed through masked figures ('bush-devils' in
Liberia and Sierra Leone; the elephant mask or Male in south-west
First the forest: villages and the socialization of space 83
Unsocialised space: the wild
Village =:>
Figure 4.3 Space and cosmology in forest society.
Cameroon), through rituals and cults (ekpe, the leopard society, in the
Niger Delta), or through adherence to regular calendars of agriculture and
hunting. Thus the forest is not simply implicated or represented in social
life. It is necessary to human society, so that even in regions where the for-
est has disappeared (to European eyes), locations for forest powers are pre-
served as secret society bush, through planted trees, or even as an imagined
forest at the margins of society: cross-roads or rubbish tips (Barber, 1990).
Given the centrality of the forest to human society, the cultures of the
region attach high value to the activities through which it is socialized, and
to the people who carry out this process, whether as heroes or through the
more mundane and pragmatic process of converting forest to farm bush.
Karin Barber's work on Yoruba oral praise poems (oriki) documents these
values. A successful chief can be commended with this verse:
'Turns forest to settlement, turns bush to town,
Oba (chief) who turns a rubbish heap into a market.'
(Barber, 1990, my parenthesis)
Similarly, conversations with Cameroon villagers show that they experi-
ence aesthetic pleasure as well as technical mastery of the environment
when they clear new farms.
The obverse of the socialization of the forest is the constant threat that
its powers will overwhelm society: that it will grow back. Abandoned vil-
lages and overgrown farms evoke despair. Settlements with declining popu-
lations are conceived to have lost their ability to manage relations with the
power of the forest. Around Mount Cameroon the ideal village is said to
be 'smoking' from the sight of the steam which rises from the sodden
thatch of crowded, lively households. Fernandez (1982), writing on
Gabon, documents how the Fang themselves contrast the 'hot' villages of
the pre-colonial past with the colonially resettled villages and 'cold' houses
that they were forced to live in after the French conquest. Similarly another
of Barber's oriki reflects the fear of the regenerative power of the forest:
'The monkey robbed the farm, it swaggered,
Jayeola, Oyedokun (names), my father the ferocious leopard'
(Barber, 1990, my parenthesis)
The forest is not however the only locus of unsocialized powers. Such powers
come also from the outside (non-Mrican) world. Historically, forest societies
have been far more successful at socializing the powers of the forest itself
than in socializing the power of the outside world, whether in the form of
colonial rule and missions, or of the representatives of the modern state -
gendarmes and armies, timber companies, forestry departments or conser-
vation organizations. Indeed the recent Royal Academy exhibition Africa -
Art of a Continent (Philips, 1995) was, for those who know these objects,
a powerhouse of forest socialization. But the exhibition was also testimony
to the failure of those socializing objects in the face of the outside power of
19th-century colonialism and Christianity. In the face of this failure,
'socializing the forest' becomes all the more urgent to West Africans of the
modern day.
The socialization of space demands the conversion of some of the forest
to farm. At the same time, however, the persistence of a cultural logic
about power and the necessary interaction of forest and society demands
that forest remains available in some shape or form. Our research docu-
ments how the ideas of socializing the forest continue to inform popular
urban culture (another kind of forest, indeed), but again with paradoxical
results so far as urban youth considers the responsibility of their society
and state toward forest conservation.
4.3 FOREST DESTRUCTION: WHO ARE THE VILLAINS?

The third myth which conservationists carried to West Africa has been the
danger to forest posed by the primitive agriculture of an explosively
expanding population. According to this stereotype, slash-and-burn agri-
culture leads to biodiversity decline, and the replacement of endemic
species and valuable, potentially sustainable timber species by fast-growing
secondary bush. Such a view fits in well with the agricultural policies of
many West African countries, which emphasize development of permanent
agriculture as a prerequisite for establishing title to land.
Contrary to this stereotype, there is now considerable evidence that
African forest farming systems are highly adapted to ecological conditions,
and that farmers manage fallows both to maximize useful and valuable
species and to accelerate regeneration (Amanor, 1994). Our own research
has also called into question many of the scientific issues on which the
stereotype rests. Thus, for example, there is considerable evidence that the
expanse of old forest actually increased in the 19th and early 20th cen-
turies as a result of precolonial insecurity and colonial epidemics, forced
Forest destruction: who are the villains? 85
labour and population collapse (Davies and Richards, 1991; Colchester,
1993; Fairhead and Leach, 1996; d. Unwin, 1915). In parts of West
Africa, and especially in Central Africa, populations may only recently
have reached the levels of the early 19th century. Elsewhere, forest villages
suffer from net population loss, as the young seek their fortune in the city.
Moreover it is not necessarily the rain forest and its biodiversity which is
under greatest threat. A number of researchers have questioned the value,
in terms of biodiversity, of the narrowly endemic rain forest species. From
this evolutionary viewpoint, valuable intra-specific genetic diversity is more
likely to occur in widespread common species inhabiting a range of
environments - species of the forest-savanna boundary. It is precisely this
area of West and Central Africa which seems under the greatest environ-
mental pressure, partly as a result of forest conservation and logging activities
further south. Thus the transition zone of Ghana is now the expanding
agricultural frontier in which World Bank-led economic reforms are having
a major impact on farming methods and fallow management. On a larger
scale, a study using satellite imagery of fire incidence over the 5 years
1984-89 concludes that the greatest impact of fire on vegetation is occur-
ring in areas of lowest population (often as low as 3-5/km2 ) in eastern
Cameroon, Central African Republic and southern Sudan (Lambin and
Ehrlich, 1995). Combined with detailed local-level ethnographic data, the
authors conclude that fires are well managed and controlled in areas of
high population density, and are far more destructive in areas with few
farmers (Nyerges, 1994; Tiffen et ai., 1994). Thus whilst large populations
of farmers certainly clear rain forest and replace it with more or less con-
tinuous farm bush, this is only one dynamic in transformations of the
African environment. Historically, in regions such as the Ghanaian cocoa
belt, Yorubaland and Iboland in southern Nigeria, and Sierra Leone, such
transformations have taken a century or more to completely replace forest
with farm bush.
Contrary to the received wisdom which underpins claims for urgent
action to conserve rain forest against slash-and-burn farmers, every case
which we investigated in Sierra Leone and Cameroon showed that timber
exploitation and export to Europe or the Far East was the prime mover in
transformation of rain forest to a less useful, often more hostile ecology.
Sierra Leone was logged over, largely in response to the needs of the British
navy, by the 19th century (Fyfe, 1962). In Ghana and Nigeria, as well as
French West Africa, colonial forestry departments actively encouraged
exploitation by expatriate companies. (Though it must be added that land
clearance for cocoa was a major impetus for local farmers to clear bush.)
In most colonies, the Forestry Department was dependent for its expansion
on increasing timber exports (parren and de Graaf, 1995; Sharpe, 1996).
However, the real impact of timber companies on the rain forest began with
mechanization in the late 1950s (see Brandler, 1993 for an autobiographical
account from a logging company perspective). It has accelerated ever since,
often in response to short-term political and economic trends. Peaks of log-
ging activity in West Cameroon, for example, closely follow political
changes: independence, the unification of the two parts of Cameroon,
change of regime in the early 1980s and, most recently, multi-party elec-
tions imposed by western donors. In Cameroon and Gabon, a free-for-all
period of logging up to 1994 took place to beat the deadline imposed on
non-sustainable timber by the European Union. Most recently, partly as a
result of international money markets awash with funds, large-scale, highly
capitalized Far-Eastern timber companies have entered the Cameroonian
industry. In Ghana, the World Bank and other agencies are attempting a
twin-track approach of greater control over logging and sustainable forest
management combined with expansion of the local timber industry. It
remains to be seen whether this approach will sustain existing rain forest.
Timber companies in West and Central Africa do not, generally, carry
out clear-felling. Given the species diversity and unknown marketing
potential of many species, logging typically removes only 2-5 treeslha.
Ostensibly, this pattern leaves smaller trees, and trees of secondary quality,
to regenerate. In practice though, construction of logging roads, and the
destruction of understorey vegetation by felling and skidding logs,
accounts for a minimum 10% of the area logged, and often far more.
Arguments for sustainable logging are problematic to say the least. But
such arguments typically ignore the on-the-ground reality of forest man-
agement.
In theory, at least, concessions are subject to management plans based
upon known regeneration rates, and rates of logging are controlled by the
Forestry Department. Concessions theoretically include both long-term
restocking and the responsibility of the companies to manage the forest for
decades. In practice, none of these conditions is fulfilled, especially in the
remote forests of highest ecological and economic value. Regeneration
rates for prime species are often unknown. Forestry officials have no trans-
port to supervise logging, and rely on logging companies for transport to
the forest. Even for the same forest, management plans vary from wildly
optimistic 20-year logging cycles, through to 80 years or more. Companies
log an area for the 5 years initially allowed by law and then go bankrupt.
Concessions are granted repeatedly over the same area of forest, which is
then logged until no saleable timber remains (Burnham, 1994).
New international and national laws are attempting to remedy some of
these problems. Throughout forested West and Central Africa, laws have
been enacted under pressure from donors and NGOs involved in conserva-
tion, demanding more stringent management plans and policing, longer
concessions, institutional strengthening of the forestry department and
greater professionalization of timber exploiters. The irony is, though, that
such programmes have commonly been formed under a bizarre alliance
Myths revisited: environmental management plans and local realities 87
between the largest multinational companies and environmental NGOs,
both seeking to protect the rain forest from local farmers. Such an alliance
is based upon the argument that large expatriate companies are the only
reliable managers of the forest, that such companies therefore need large
long-term concessions, and that the forest can be zoned into areas for log-
ging and areas set aside for conservation.
Such plans for zoning of forest commonly assume that it is possible both
to log the forest sustainably and to establish permanent agricultural sys-
tems for the inhabitants. However, a survey of the literature on agricultural
and forestry research throughout West and Central Africa discovered no
research which offered either feasible permanent low-input agricultural
systems or practices for sustainable management of timber trees in species-
diverse rain forest. Whilst this is not the place to review all of this data (see
also Parren and de Graaf, 1995), the main justifications for this statement
are:
• exclusionary long-term concessions for expatriate logging companies are
neither politically nor socially realistic;
• plantations are too expensive (despite extensive French research and
work by various agencies at Mbalmayo, Cameroon);
• no feasible agroforestry packages have been developed which could
make up for the quantities of forest timber being extracted or which
could enable farmers to cultivate their land permanently;
• community forest management is in its infancy (but see below).
Perhaps in recognition of the absence of feasible alternatives, logging con-
cerns and conservation NGOs seek to control local farmers, for whom log-
ging companies are the new 'elephants'. Logging roads allow farmers to
penetrate the forest without the arduous work of clearing the matted
vegetation of the forest edge. Recognizing their role, some timber com-
panies routinely demolish bridges and block tracks after their operations
have finished. Similarly, the World Bank, at the instigation of the World
Wide Fund for Nature (WWF), has successfully blocked a number of roads
proposed for the Cameroon, Gabon and Congo borderlands. Local people,
it seems, remain the villains so far as rain forest conservation is concerned.
4.4 MYTHS REVISITED: ENVIRONMENTAL MANAGEMENT

PLANS AND LOCAL REALITIES
One of the aims of this chapter was to attempt to define how West Africa is
different. This is not to argue that conservation in forest West Africa is
impossible, but rather that each of the major rain forest regions of the
world has its own history and hence its own future conservation trajectory.
Only three of the relevant particularities of the West African case have
been reviewed here. The first section outlined how different degrees of
indigenousness and strangerhood have been long-standing features of West
African cultures and politics. Contemporary migration into (and out of)
the forest is only an aspect of long-running historical dynamics. The second
section outlined cosmological relations between humans and the forest,
and how society has been recurrently constructed through the socialization
of the forest and its powers. Finally, the third section summarized research
findings regarding the source of the threats to the forest. The facts about
indigenous people, the socialization of space and the causes of deforesta-
tion are somewhat at variance with attempts to demarcate the lands and
resources which communities may use in conservation areas.
As we have seen, conservation organizations and projects are new
entrants to the contests around forests. Early attempts to set up national
parks or faunal reserves envisaged moving people out. Such forced resettle-
ment is still enshrined in the plans of the Korup National Park in
Cameroon, for example. In the face of local hostility, however, and perhaps
in recognition of the sheer impossibility of policing rain forest, most con-
servation projects now claim to have adopted a more participatory 'people
and parks' approach to conservation, an approach which respects peoples'
rights to forest resources. On the face of it, community participation in
conservation is an advance upon earlier authoritarian policies.
Unfortunately, however, these attempts to co-opt local people into conser-
vation are themselves situated in a wider political economy in which non-
local stakeholders (timber firms, conservation organizations and state
institutions) have both greater power and access to scientific knowledge.
Often the power of different stakeholders is enshrined in Environmental
Management Plans.
Environmental Management Plans (also sometimes termed Environ-
mental Action Plans) are land-use plans which have been developed in
most West African countries under the aegis of, and with funding from, the
World Bank and UN agencies. Such plans attempt to achieve a scientifically
founded balance between conservation and utilization of forest resources.
Typically such plans assess forest regions or entire countries according to
environmental resources and problems, producing intricate maps of areas
zoned for conservation, timber and farming. Whilst such plans are often
based upon extensive survey work by foresters and ecologists as well as
techniques using geographical information systems and high-tech analytical
methods, they are also liable to be negotiated amongst only the more pow-
erful stakeholders in the region concerned: the national government, tim-
ber companies, conservation organizations and external donors. The plans
which emerge from the negotiations can have serious unintended con-
sequences, both at the local and regional levels.
Locally, for example, management plans take little cognisance of actual
patterns of resource use by forest villagers. In Cameroon, the Plan de
Zonage for the entire forest region allocates corridors for farming and
Forest destruction: who are the villains? 89
settlement based on aerial photo-interpretation of present cultivation and
'rule-of-thumb' estimates for future land needs of villagers. On paper these
appear adequate to the extremely low populations of the forest. However,
detailed ethnographic classification of resource use and mapping of human
activities on the ground shows villagers losing control of up to 60% of
their territory (see Figure 4.4).
Penelon and Mendouga Mebenga estimate that in this single village terri-
tory, 56.4% of the forest is to be classified as State forest, in which local
use is prohibited. It is certain that a similar situation is occurring through-
out the forest zone of Cameroon.
Unintended consequences of Environmental Management Plans are also
likely to be occurring at the regional level. Whilst there are, as yet, no data
from West Africa comparable to data from the East African savanna, it is
almost certain that attempts to protect rain forest are displacing and inten-
sifying economic activities (fuel wood, non-timber forest products, farm-
ing) in unprotected areas, especially the forest-savanna transition (Sharpe,
1991; Amanor, 1994). Whilst part of this dynamic is certainly driven by
global economic processes and demographic change, much of it is the
direct result of the failure to include forest societies in the development of
Environmental Management Plans. Outright 'protection' of rain forest
from people, whether for conservation or sustainable management,
impacts on the environment elsewhere.
For conservation (and for the world) the concern is that large areas of
species-diverse old forest are being lost. Earlier I have suggested that this
may be only one of the dynamics at work in the transformation of the West
African landscape. The fact remains, however, that West African people are
converting forest to farm fallow at an increasing rate. As we have seen,
conservationists have portrayed this as an invasion of environmental
refugees using primitive slash-and-burn techniques which are unsustain-
able. Environmental management plans attempt to control this, making
various heroic assumptions about sustain ability in the process. Per contra,
this chapter suggests that West and Central African cultures have been well
aware of how to manage the forest-farm-forest dynamic, and that the cul-
turallogic of societies that have developed in the forest does value the for-
est as an essential part of the relationship between humans and the world.
Given this fact, attempts at control miss the point.
From an African perspective, the problem is not saving the rain forest,
but rather under what circumstances people may continue to interact with
it. As we have seen, civilization has been built on the resources of the for-
est. Many townspeople as well as forest dwellers do not doubt the cultural
value of forest, but they are also well aware of the origins of the pressures
to ignore cultural logic and expand farm-bush into the last remaining for-
est. In the contemporary context the costs of constructing and maintaining
society and community are placing overwhelming demands on the forest.
..............................
................... o
(/
:::>
o
o
o
. . . . . ..
1
i....:····
o (S'"
<::,:: .. /./
.::>
.........".l"/""""· o ..../ ................
.••.....
o Village
~ Limits of territory used by the village of Bimba
Approximate limit of forest zoned as Permanent Domain of the State
D Area to be excluded from village use
Figure 4.4 Conflict between local land use and Plan de Zonage based on the
Bimba community mapping and forest management study, Eastern Province,
Cameroon (from Penelon and Mendouga Mebenga, 1995). Over 50% of village
territory will be declared State land zoned as Production Forest.
People cite economic collapse, increasing needs, demands for export-led

growth, the crisis of the educated youth with few prospects of professional
employment, as well as environmental and political decline. Without in any
way subscribing to some notion of traditional 'Merrie Africa' (Davidson,
Not all doom and gloom 91
1992), the misunderstandings between conservationists and local views on
the forest are the result of quite basic differences in what the forest means.
For West African cultures, the idea is one of socializing the forest, manag-
ing it for its useful properties and then leaving it to regenerate. In contrast,
rain forest is seen, in the West, as simply a timber or biodiversity 'bank'.
Perhaps a useful parallel here is in European and African responses to local
economic decline. In the face of currency instability, collapsing producer
prices, venal and corrupt services, or the virtual disappearance of the state,
West African communities have redeveloped and elaborated circulating
credit associations (esusu, njangi) and non-monetary exchange systems
(MacGaffey, 1991). In contrast, their rulers, often working closely with
multinational firms, have transformed rain forest resources into literal
bank deposits. The state is bankrupt (though its rulers aren't).
In this context of the 'bankrupt state', current conservation solutions
such as environmental management plans and legislative change, timber
certification and sustainable management tinker with the real environmen-
tal problem. Unfashionably, West Africans want a state that works, rather
than a 'free' market in which the majority realize little of the value of the
natural resources which are of such importance to the world. Arguably, the
key issue for research and for conservation efforts is to comprehend how
the historical dynamics of migration and the creation of society will inter-
act with environmental dynamics stemming from global environmental and
socio-economic change. The burden of this section, therefore, is that rain
forest conservation in West Africa should concentrate on understanding
and developing more feasible and successful ways of working with forest
dwellers based on the cultural logic of interactions between society and the
forest.
4.5 NOT ALL DOOM AND GLOOM - ENGAGEMENT OF

CONSERVATION WITH LOCAL ASPIRATIONS
This chapter has been written from a perspective arising out of a research
project funded by the Economic and Social Research Council's Global
Environmental Change programme. Our research has gone beyond simply
criticizing current conservation policy and practice. As part of the research
project, we collected data on attitudes to the forest and its future, on which
it might be possible to base future developments. For Sierra Leone, which
descended into war during our research, this work on knowledge and atti-
tudes has become crucial to post-war reconstruction (Richards, 1996). In
Cameroon, on the other hand, our research uncovered a rich political and
cosmological debate around forest. These views were gathered in surveys,
participant observation, interviews and participation in the work of village
councils and ministries. They are not official rhetoric, but rather express
how West Africans, from all age groups and levels in society, actually discuss
and argue passionately about the forest and the future. One clear theme in
these discussions is distrust of the new concern of expatriate organizations
with 'saving' the forest. More enlightening, though, are the positive views
of how, and by whom, preserving the forest might be possible. One clear
theme is that this should be a major function of the state, closely linked to
popular aspirations for more honest and effective government. Another is a
quite strongly authoritarian concern to control timber extraction and to
link the proceeds from logging to development of the forest areas them-
selves. A third theme, and perhaps the most crucial, is despair that the
long-standing cultural dependence of people on the forest is about to
become untenable. Whilst much of this research has yet to be written up,
the basic conclusion is that the best prospect for 'saving' the rain forest is
for conservation and development organizations to engage with this popu-
lar discourse and the legitimate political and economic aspirations in which
it is embedded. Hopefully dissolving the myths outlined in this chapter will
go some way to facilitating this process.
4.6 THE DCL/ODA COMMUNITY FOREST MANAGEMENT

PROJECT
Since 1995, our research team at University College London has been
working with the UK Overseas Development Administration on a project
to develop ethnographically informed procedures for community forest
management in Cameroon. The aim of this project is to develop procedures
through which communities apply for the right to manage their forest,
including timber resources and non-timber forest products, for their own
benefit.
Cameroon, like most countries in West and Central Africa, has recently
passed new forestry legislation. In Cameroon the 1994 Law envisages a
new category of forest - foret communautaire - which was not previously
included in either the French colonial forestry law or the Cameroonian
laws which succeeded it. (It is important to note, however, that the British
Colonial Forestry Ordinance of 1917 did vest forest reserves in designated
local councils, so-called Native Authorities, though these laws were super-
seded by the 1974 Forestry Law of the United Republic of Cameroon.) The
new legal framework has only recently been supplemented by the publica-
tion of a code of procedures for carrying out an inventory and preparing
Management Plans, through which communities may come to exercise
control over their forest. Whilst many outside agencies seem to have
assumed that simply changing forestry policy would lead to the successful
creation of community management, there are in fact considerable prob-
lems in putting the legislation into practice. Many of these problems stem
from the particularities raised in this paper: especially that community for-
est might be 'captured' by the powerful (in the villages or from elsewhere);
Conclusion: forest futures. forest cultures 93
that allocation of community forest might conflict with other management
plans and uses (as shown in Figure 4.4); or that there was inadequate scien-
tific knowledge and institutional capacity to put community forest manage-
ment in place. All three of these problems have (as of October 1996) arisen
as predicted, as conservation organizations and politicians have rushed to
implement community forest law. At the present time, all applications for
community forest have been delayed by the Cameroonian Government
pending clarification of the legal status of the groups involved and redefini-
tion of simplified procedures for inventory and management plans.
The anthropological input from University College London into the
ODA Community Forestry Development Project is attempting a rather dif-
ferent development process, fully informed by the cultural and institutional
context which has been outlined here. This involves a number of prepara-
tory steps, of which the most important are listed below.
• Thorough investigation of the legal status of forest areas at the local
level. In many areas there is little room for manoeuvre as forest has been
granted as contiguous (or even overlapping) logging concessions. Given
this, there is no point in raising local aspirations unless concession
licences can be revoked.
• Recognizing and analysing the problem of defining 'community', given
the sociological processes through which commumtles form.
Community forestry development is thus in many ways congruent with
community development, and revenue from managed forest will be cru-
cial to the sustainability of the programme.
• Developing the institutional capacity of the Forestry Department to
incorporate local social analysis in the assessment and development of
community management plans.
Although it is early days yet, and some observers have expressed consider-
able pessimism about the feasibility of community forestry in Cameroon,
the project is consolidating its work in two of the forested provinces (South
and South-West). Ultimately though, as shown by Liberia, Sierra Leone
and Zaire, there is no sustainability without security, and community forest
management needs to be complemented with development measures to
improve economic and welfare conditions for forest dwellers. Not the least
of these is reducing the demands on the forest by 'the North', whether in
the form of tropical logs, debt repayments or export cash crops.
4.7 CONCLUSION: FOREST FUTURES, FOREST CULTURES

Many of the processes affecting rain forest in Africa are shared by other
tropical regions of the world: overexploitation, population growth, politi-
cal and institutional incompetence (at best) and corruption (at worst), com-
munal insecurity, cultural change. Analyses similar to that in this chapter can
be applied elsewhere, in South America and South-East Asia. As noted
above, Africa is different from these other regions insofar as the forest indi-
genes are not a culturally distinct minority in a larger state. African forest
cultures are unlikely to disappear, as is happening elsewhere. For that rea-
son, however, the African case raises important ethical, sociological and
biological issues for conservation. Anthropology is fundamentally relativist
in seeking to understand societies and institutions. In Africa, understanding
the motivations and actions of businessmen, conservation organizations
and international agencies is almost as significant as understanding local
cultures, so far as preserving the forest is concerned.
Ethically, we might question whether co-opting local communities to
conservation efforts which are essentially western in conception is the best
solution, or whether preservation of the rain forest would be better served
by controlling the outside forces which are destroying it: issues such as
western over-consumption, excessive demands for debt repayment, and tol-
erance of third-world governments which are undemocratic and often
actively intolerant of the legitimate demands of their own population
(including forest dwellers).
Sociologically, the African situation raises the issue of just how much
social knowledge, and of what kind, is required for an effective dialogue
between conservationists and forest people. In all parts of the world, west-
ern organizations assume that indigenous people are different, but that set-
tlers, farmers (and loggers?) are distressingly just like us. Beyond its
biological rationale, conservation arises from our disenchantment with the
modern world. An Amazonian gold digger seeking a stake to set up a small
store, or a Vietnamese commercial hunter saving for a satellite TV, are
assumed to have taken up the worst aspects of our culture, whilst aban-
doning or destroying their own. The aim of conservation then becomes to
preserve the forest from them (and by extension from ourselves). But as the
African situation shows, this is a caricature of the real cultural changes
occurring in the forest, which demand analysis in their own right. At pre-
sent, given the squeeze on research and demands for urgent action, deep
social analysis is notably lacking even in recent books on conservation and
culture (see, for example, Furze et al., 1996). Detailed socio-economic and
cultural analysis of forest societies reveals some of the pressures leading to
biodiversity destruction. Conservation could probably be better served by
addressing these issues, especially if the aim is to build a conservation con-
stituency in poor countries. But that is unlikely to happen unless deep
social knowledge is acquired and employed.
Historical experience in the west shows how biodiversity decline follows
deforestation, but it also shows how rich nations have been able, at least
partly, to redress or reverse this decline through the establishment of con-
servation areas, reforestation and environmental management. It is con-
ceivable that the emerging economies of South America and South-East
References 95
Asia may be able to follow this route, should that prove culturally desir-
able and economically possible. But such possibilities are unlikely to be
available in the areas where logging is currently expanding - Central Africa
and the South Pacific/Papua New Guinea. The burden of this chapter is
that in Africa, at least, continuation of current exploitation patterns will in
fact lead to irrevocable extinction of many forest species.
ACKNOWLEDGEMENT
Research contributing to this chapter was carried out by a team of anthro-
pologists from University College London, under a grant from the
Economic and Social Research Council (UK) Global Environmental
Change Programme, whose support is gratefully acknowledged. (Grant:
The Cultural Context of Rain Forest Conservation in West Africa, P.
Burnham, P. Richards, M. Rowlands, B. Sharpe, with research assistant R.
Malleson.)
REFERENCES
Alexandre, P. and Binet, J. (1958) Le Groupe dit Pahouin: Fang, Bulu, Beti, Presses
Universitaire de France, Paris.
Amanor, K. (1994) The New Frontier: Farmer Responses to Land Degradation: a
West African Study, UN Research Institute for Social Development (UNRISD),
Geneva.
Barber, K. (1990) I Could Speak Until Tomorrow: Orib, Women and the Past in a
Yoruba Town, International African Library, University Press for the
International African Institute, Edinburgh.
Black, R. and Mohammed Sesay (1996) Contesting claims on forest land in Guinea:
refugees, development and environmental change, paper presented to a work-
shop on Contested Terrain: West African Forestry Relations, Landscapes and
Processes, Centre for West African Studies, Birmingham, April 1996.
Brandler, J. L. (1993) Out of Nigeria: Witness to a Nation's Toils, Radcliffe Press,
London.
Burnham, P. (1994) The cultural context of rain forest conservation in Cameroon,
paper presented to the Annual Conference of the African Studies Association,
USA.
Burnham, P., Richards, P., Rowlands, M. and Sharpe, B. (1995) The Cultural
Context of Rain Forest Conservation in West Africa, final report to the Global
Environmental Change Programme, Economic and Social Research Council, UK.
Cleaver, K., Munasinghe, M., Dyson, M., Egli, N., Peuker, A. and Wencelius, F.
(eds) (1992) Conservation of West and Central African Rainforests. World
BanklInternational Union for the Conservation of Nature, Washington, DC.
Colchester, M. (1993) Slave and enclave: towards a political ecology of Equatorial
Africa. Ecologist, 23, 166-73.
Copet, E. (1977) Nguelebok: Essaie d'Analyse de I'Organisation Sociale de Mkako
Mbogendi, Universite de Paris, Ecole Pratique des Hautes Etudes, Paris.
Davidson, B. (1992) The Black Man's Burden: Africa and the Curse of the Nation-
State, James Currey, London.
Davies, G. and Richards, P. (1991) Rainforest in Mende Life: Resources and
Subsistence Strategies in Rural Communities around the Gola North Forest
Reserve (Sierra Leone), Report to the UK Overseas Development
Administration, Department of Anthropology, University College London.
Dounias, E. (1993) Dynamique et gestion du systeme de production a dominante
agricole des Mvae du Sud Cameroun forestier, PhD thesis, University of
Montpellier.
Evans Pritchard, E. E. (1971) The Azande: History and Political Institutions,
Oxford University Press, London.
Fairhead, J. and Leach, M. (1996) Misreading the African Landscape: Society and
Ecology in a Forest-Savanna Mosaic, Cambridge University Press, Cambridge.
Fernandez, J. (1982) Bwiti: an Ethnography of the Religious Imagination in Africa,
Princeton University Press, Princeton.
Furze, D. (ed.) (1996) Culture, Conservation and Biodiversity, Wiley, Chichester.
Fyfe, C. (1962) A History of Sierra Leone, Oxford University Press, London.
Guyer, J. (1984) Family and Farm in Southern Cameroon, African Studies Centre,
Boston University.
Guyer, J. and Lambin, E. (1994) The Complementarity of Remote Sensing and
Anthropology in the Study of Complex Human Ecology, African Studies Centre,
University of Boston.
Keay, R. (1971) Human Ecology in the Tropics, Pergamon, Oxford.
Kopytoff, I. (ed.) (1987) The Africa Frontier: the Reproduction of Traditional
African Societies, Indiana University Press, Bloomington.
Lambin, E. and Ehrlich, D. (1995) Broad scale maps on Africa's land cover and its
dynamics, 1982-1991, presented to a workshop African Farmers and their
Environment in Long Term Perspective, Leiden, Netherlands, March 1995.
Mac Gaffey, J. (1991) The Real Economy of Zaire: the Contribution of Smuggling
and other Unofficial Activities to National Wealth, James Currey, London.
Marnham, P. (1980) Fantastic Invasion: Despatches from Contemporary Africa,
Cape, London.
Martin, G. (1991) The Rainforests of West Africa: Ecology, Threats, Conservation,
Birkhauser Verlag, Basel.
Orlove, B. and Brush, S. (1996) Anthropology and the Conservation of
Biodiversity, Annual Review of Anthropology, Annual Reviews Inc., Palo Alto.
Nugent, S. (1993) Amazonian Caboclo Society: an Essay in Invisibility and Peasant
Society, Berg, Oxford.
Nyerges, A. Endre (1994) Deforestation History and the Ecology of Swidden
Farming, African Studies Centre, Boston University.
Parren, M. and de Graaf, N. R. (1995) The Quest for Natural Forest Management
in Ghana, Cote d'Ivoire and Liberia, Tropenbos Series 13, Tropenbos,
Wageningen.
Penelon, A. and Mendouga Mebenga, L. (1995) Carte 3: Carte du Plan Provisoire
de Zonage sur Ie Terroir de Bimba, in Connaissance et Fonctionnement du
References 97
Terroir Villageois de Bimba, project document for Projet d'amenagement pilote
integre de Dimako (mimeo.).
Philips, T. (ed.) (1995) Africa - Art of a Continent, Royal Academy exhibition cata-
logue, Prestel, Munich.
Richards, P. (1995) Local understandings of primates and evolution: some Mende
beliefs concerning chimpanzees, in Ape, Man, Apeman: Changing Views since
1600, (eds R. Corbey and B. Theunissen), Department of Prehistory, Leiden
University.
Richards, P. (1996) Fighting for the Rainforest: War, Youth and Resources in Sierra
Leone, International African Institute/James Currey, London.
Richards, P. W. (1996) The Tropical Rain Forest: an Ecological Study, 2nd edn,
Cambridge University Press, Cambridge.
Sharpe, B. (1991) Wenchi Farmer Training Project, report to the Overseas
Development Administration, London (mimeo.).
Sharpe, B. (1996) Forestry institutions and identities in south-west Cameroon, pre-
sented at a seminar on Institutions and the Environment, Institute of
Development Studies, University of Sussex.
Shaw, T. (1978) Nigeria: its Archaeology and Early History, Thames and Hudson,
London.
Thikalul, S. (undated) Manual of Dendrology, Cameroon, Groupe Poulin,
Theriault, Quebec, CanadalForestry Institutional Support Project, Centre for
Forestry Development, Cameroon.
Tiffen, M., Mortimore, M. and Gichuki, F. (1994) More People, Less Erosion:
Environmental Recovery in Kenya, Wiley, Chichester.
Unwin, A. H. (1915) Report on the Forests of West Cameroon, Cameroon
National Archives, Buea.
Vansina, J. (1990) Paths in the Rainforest: Towards a History of Political Tradition
in Equatorial Africa, James Currey, London.
Winterbottom, R. (1992) Tropical Forestry Action Plans and indigenous people: the
case of Cameroon, in Conservation of West And Central African Rainforests,
(eds Cleaver et al.), World Bank Environmental Paper No.1, World
BanklInternational Union for the Conservation of Nature, Washington,
pp. 222-228.
-5------
Forest and environmental
degradation
M. J. Eden
Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK
Recurrent concern has been expressed over recent decades regarding tropi-
cal deforestation. Concern has focused on the unsustainable nature of
many of the agronomic systems that have been established on cleared for-
est land and on the broader biophysical feedbacks that are associated with
the loss of forest cover; the latter include climatic and hydrological feed-
backs and loss of biodiversity. The concern with tropical deforestation,
dating particularly from the early 1970s (Gomez-Pompa et aI., 1972;
Denevan, 1973; Richards, 1973) has been an important element of the
global environmental movement and it has provoked considerable scientific
investigation of tropical forests.
Much of this investigation has focused on the causes, extent and impact
of deforestation, and has usefully alerted many individuals, agencies and
governments to the need for forest management strategies that accommo-
date both development and conservation. However, this attention to defor-
estation has lately been perceived as too crude and partial for current
purposes, and some beneficial broadening of the perspective on tropical
forests has emerged which, on paper at least, enhances the possibility of
more effective forest management.
A first advance has been the recognition that the present status of previ-
ously deforested land is a relevant variable. An earlier tendency to assume
that cleared forest land stayed cleared, and was permanently deflected to
some herb-dominated cover, has been succeeded by increasing recognition
of the renewability of the forest and of the significance of the post-
clearance land cover, not least in respect of the climatic and hydrological
feedbacks of deforestation (Eden, 1996a). As a consequence, secondary
forests and their ecological dynamics have lately received increasing atten-
tion (Chazdon, 1994; Corlett, 1995). Secondly, it has been recognized that
100 Forest and environmental degradation
attention to forest damage, as distinct from deforestation (clear cutting), is
also desirable. Forest damage, typically associated with selective logging,
usually leaves a substantial canopy in place but causes a significant reduc-
tion in biomass and changes in species composition. Forest damage is less
easily monitored than deforestation, but increasingly deserves the attention
of forest researchers and managers.
Emerging from the above is the useful concept of forest degradation,
which encompasses both deforestation and forest damage (Eden, 1996b;
Grainger, 1996). Forest degradation is in turn a component of the broader
concept of land degradation which, as perceived by Blaikie and Brookfield
(1987), usefully emphasizes the social as well as the environmental aspects
of degradation, and also stresses the idea of 'net degradation' as a product
of counteractive 'degrading' and 'restorative' processes. In the present con-
text, forest degradation is seen as a valuable generalizing concept that
incorporates deforestation, extends therefrom to consider the nature of the
post-clearance land cover, and also includes forest damage. In all cases, the
environmental processes involved and their management implications are
of concern.
The present chapter firstly examines the broad status of forest degrada-
tion in the tropics, with particular attention to remote sensing which is
increasingly used to measure and monitor it; secondly, it reviews the physi-
cal and biological impacts of forest degradation; and thirdly, it discusses
associated aspects of forest management, especially forest conservation
that aims to minimize the biophysical impacts of forest degradation.
5.1 FOREST DEGRADATION

Forest degradation is an element of the broader phenomenon of land
degradation. The latter has lately emerged as an integrative concept applic-
able to a range of physical and biological processes that cause land to suf-
fer 'a loss of intrinsic qualities or a decline in capability' (Blaikie and
Brookfield, 1987). Forest degradation, particularly when involving defor-
estation, has latterly received much attention and been seen as one of the
more serious global environmental issues. As a result, considerable efforts
have been made in recent decades to monitor deforestation, particularly
using remote sensing techniques. Although useful data have accrued at
national level, reliable pan-tropical information has not readily been
obtained since the early 1970s, despite the availability of Landsat imagery.
Only during the present decade has the long-standing objective of assem-
bling a reliable, pan-tropical data set on deforestation approached reality.
Monitoring forest damage, as opposed to deforestation, is a more challeng-
ing task and one which is only now beginning to be investigated.
Ultimately, however, integrated data for forest degradation are required as
an aid to formulating sound technical and management strategies of issues
Forest degradation 101
ranging from global warming to loss of biodiversity.
One attempt to integrate the impact of forest degradation at the
regional level is that of Grainger (1996). His pioneer study in South-East
Asia, based on degrees of biomass reduction below theoretical forest cli-
max values, is a starting point for assessing potential forest rehabilitation.
The specific context is that of mitigating global climate change by seques-
tering atmospheric carbon, but the approach is also relevant in assessing
and managing other feedbacks of forest degradation (Eden, 1996b).
5.1.1 Deforestation monitoring

Monitoring the extent of tropical deforestation essentially involves tem-
poral comparisons of forest cover. This is relatively easy to undertake for
small areas over recent, shorter periods, but difficulties exist when large-
area, longer-term monitoring is attempted. One issue that arises is what
constitutes tropical forest. For most researchers, the utility of the term
'tropical moist forest' has been recognized. Tropical moist forest is
broadly defined as comprising all closed-canopy forests occurring in the
humid tropics (Sommer, 1976; Grainger, 1993). It thus includes such cate-
gories as tropical rain forest and tropical moist deciduous forest, but
excludes the closed-canopy dry forests that exist in parts of the tropics, as
well as open-canopy forests such as savanna woodlands. The concept of
tropical moist forest is widely, if not universally accepted, and offers the
most convenient basis for assessing human impacts on the forest.
A second issue that arises in the context of forest monitoring is the orig-
inal extent of the natural forest as a baseline for comparison with the cur-
rent extent. The original extent is not easily estimated for a variety of
reasons. According to Sommer (1976), the climatic zone suitable for tropi-
cal moist forest as defined is 1600 million ha, although part of the area is,
for ecological and other reasons, thought historically to have been covered
by other vegetation types (Reid, 1992). Simberloff (1986) suggests that,
prior to significant human impacts, 10% of that area was not forested,
giving an extent for the original tropical moist forest of 1440 million ha.
Against this background, various estimates of the contemporary tropi-
cal moist forest cover have been made. For the most part, they have been
compilations of national data from various sources and of varying quality.
An early study was that of Sommer (1976), who estimated the total area
of tropical moist forest at 935 million ha. This compared with Myers
(1980) who gave an area of 972 million ha, and with the 1980
FAOIUNEP Tropical Forest Resource Assessment (Lanly, 1981) which
estimated an area of 1081 million ha. Later studies were undertaken by
Myers (1989) and FAO (1993). Myers (1989), whose compiled data were
"heavily dependent" on remote sensing sources, both satellite and air-
borne, gave an area of tropical moist forest of 800 million ha. FAO
(1993), who also made additional use of remote sensing, likewise esti-
mated the forest cover, but their chosen forest classes, i.e. tropical rain
forest, moist deciduous forest, montane forest, etc. precluded comparison
with earlier figures for tropical moist forest.
Some estimate can thus be made of total deforestation. Using
Simberloff's (1986) figure for the extent of natural tropical moist forest, it
appears that some 25-45% of the original forest cover has already been
cleared (Lanly, 1981; Myers, 1989). As far as rates of deforestation are
concerned, recent estimates are also more notable for their variability than
their consistency. They include Sommer's (1976) oft-quoted figure for
deforestation of 11 million halyear, and Myers' (1989) figure of 14.2 mil-
lion ha/year; other researchers are more conservative, with Grainger
(1983) suggesting a figure of 6.1 million ha/year.
These studies are important because they recognize the need for precise
data on the extent of the forest cover and on rates of clearance. However,
the quality of the existing compilations leaves much to be desired, and it
has been widely recognized that more consistent methods of forest moni-
toring are needed. In this respect, a critical role exists for satellite remote
sensing and for a centralized or top-down approach to its use. The poten-
tial for remote sensing of the forest cover has already been demonstrated
at national level. Useful examples include studies of the Philippines
(Kummer, 1992) and of India (Ravinandrath and Hall, 1994) where
sequential imagery, mainly Landsat from the 1970s and 1980s, has been
employed. Likewise for Brazilian Amazonia, Landsat images have been
used since the 1970s to monitor forest cover; indeed, since 1988 annual
estimates of deforestation have been produced for the region, which con-
tains about one-third of the remaining tropical forest. In the case of
Brazilian Amazonia, whose original forest cover was approximately 400
million ha, the extent of deforestation has been estimated, for 1990, at
10.2% (Eden, 1994).
Although some large tropical territories are being monitored by remote
sensing, application of such techniques at the pan-tropical level has been
slow. Landsat imagery, which despite cloud problems has commonly been
used for forest monitoring at national levels, has been available since the
early 1970s, but only belatedly have this and other, newer satellites, both
optical (e.g. the National Oceanographic and Atmospheric
Administration's Advanced Very High Resolution Radiometer; NOAA
AVHRR) and microwave (e.g. European Radar Satellite; ERS), been
adopted as the basis for pan-tropical forest monitoring. Among major
projects of this kind is NASA's Landsat Pathfinder project which aims to
assemble multiple Landsat coverages for most of the tropical forest zone.
Retrospective coverage is being sought of the early 1970s, 1980s and
1990s, with a view to documenting the extent and rate of deforestation.
This approach builds on that originally developed by the Brazilian
National Institute for Space Research (INPE) for Brazilian Amazonia, but
Forest degradation 103
the project is intended to be pan-tropical (Malingreau, 1993;
Chomentowski et ai., 1994; Skole et ai., 1994).
A similar project is the TREES project established in 1990 as a joint
initiative of the European Commission and the European Space Agency
(ESA). The basis of the project is imagery from the optical NOAA
AVHRR, with a spatial resolution of 1.1 km, for which pan-tropical 'wall-
to-wall' coverage for the period 1989-92 is being sought. This imagery is
being supplemented by higher-resolution Landsat TM and SPOT imagery
for calibration and validation purposes. In addition, high-resolution
microwave ERS imagery has lately been incorporated in the project,
where it will perform a useful gap-filling function in space and time, par-
ticularly in areas of more-or-less permanent cloud cover (Malingreau,
1993; Malingreau and Duchossois, 1995).
As well as attempts to determine the extent of deforestation, attention
has focused on the pattern of deforestation and in particular on the level
of fragmentation of the residual forest, i.e. how much it is broken up into
smaller units. Fragmented units are likely to suffer higher levels of species
loss and are more accessible to further human damage (FAO, 1993;
McCloskey, 1993; Skole and Tucker, 1993). According to McCloskey
(1993), only one-third of the remaining tropical rain forest is found in
large 'wilderness' blocks, with two-thirds of it being fragmented and
especially vulnerable. Fragmentation is most advanced in South-East Asia
and Oceania, where only 12 % of the remaining tropical rain forest
reportedly comprises large wilderness blocks.
In addition to concern with the scale and pattern of deforestation, it has
lately been recognized that the post-clearance land cover is a critical vari-
able. In the past, it was commonly assumed that cleared forest was non-
renewable and was generally replaced by a low-biomass, herb-dominated
community of some kind. Gomez-Pompa et al. (1972) explicitly described
the tropical forest as a 'non-renewable resource', arguing that across most
of its area it was incapable of regenerating under current land-use prac-
tices. Particular concern was expressed over large-area clearance where, as
a function of the reproductive characteristics of many primary forest
species, natural forest regeneration was considered unlikely to occur. In a
more specific Amazonian context, similar views were expressed by Uhl
(1988).
While cleared forest land in the tropics sometimes remains cleared, sig-
nificant forest regeneration is reported on cleared land in Brazilian
Amazonia. This includes the majority of pastures along the Belt~m-Brasilia
highway, which are typically abandoned after 6-12 years and thereafter
develop secondary woody regrowth (Fearnside, 1979; Uhl et aI., 1988).
Similar conditions are described by Lucas et al. (1993) in northern
Amazonia, where extensive forest clearance for crops and pasture
occurred as a result of the opening of the Manaus-Boa Vista highway in
the late 1970s. From examination of Landsat imagery, it is apparent that
by 1991 'much of the previously cleared land north of Manaus supported
secondary forest' (Lucas et aI., 1993). Likewise, in the vicinity of Boa
Vista itself, Dargie and Furley (1994) report that large areas of cleared
forest land, visible on 1978 Landsat imagery, had reverted to forest by
1985.
Attempts are being made to quantify the status of the post-clearance
land cover. Mausel et al. (1993), for example, have examined the land
cover along the Transamazon highway near Altamira, where extensive
clearance for crops and pasture was undertaken during the 1970s. Within
a study area of 267 000 ha astride the highway, Landsat images reveal
that, by 1991, approximately 100000 ha of forest had been cleared. Of
this area, 13.5% remained under cultivation, but the rest consisted of sec-
ondary regrowth, classified as initial succession (29.2%), intermediate
succession (41.4%) and advanced succession (15.9%). Both the intermedi-
ate and advanced categories showed 'multicanopy' development, with
trees reaching heights of 8-12 m and over 20 m, respectively (Mausel et
al., 1993). This constitutes a significant level of forest regeneration.
At the regional level, researchers are also seeking to characterize the
stage of regeneration of tropical forests, using coarse-resolution imagery
such as NOAA AVHRR as well as other, higher-resolution systems.
Studies have been undertaken in both Amazonia and West Africa (Foody
and Curran, 1994; Curran et aI., 1995), and are of particular importance
for clarifying the global carbon budget.
5.1.2 Forest damage

Forest damage, as distinct from deforestation, has been variously
described but is here taken to refer to modification or disturbance of the
natural forest cover without involving clear cutting. Forest damage may
be apparent in various ways, including reduced biomass density, reduced
canopy cover and/or reduced species diversity. In all cases, the implicit
comparison is with the natural forest climax for the area in question
(Grainger, 1996). Various human impacts cause forest damage, but the
most significant and widespread is that of harvesting forest products, par-
ticularly timber which is widely extracted by selective logging. Other for-
est damage can be caused by harvesting non-timber forest products or by
local fire or pollution impacts.
The main cause of forest damage, namely extractive logging, is now-
adays widespread in the tropics, but is of variable impact. Traditional log-
ging commonly involved the harvesting of some 5-10 m 3/ha of timber,
with further damage arising from the cutting of logging trails and from
the timber extraction itself. Of late, higher levels of extractive logging
have occurred. According to FAO (1993), average logging intensities in
Biophysical implications of forest degradation 105
insular South-East Asia are 63 m 3/ha and in continental South-East Asia
20 m 3/ha. Average values for tropical Africa and tropical America are
lower, at less than 20 m3/ha (FAO, 1993) although in places they clearly
exceed that level (Uhl and Vieira, 1989; Verfssimo et aI., 1996).
While experimental mixed forest silviculture, based on selective logging,
has been widely undertaken and reported in many parts of the tropics
(Eden, 1996c), fewer studies exist of the precise impact of extractive log-
ging. It is increasingly evident, however, that such operations seriously
damage the forest, particularly when careless timber extraction occurs. In
eastern Amazonia, for example, Uhl and Vieira (1989) describe the
removal of eight large trees per ha, totalling some 52 m3 of timber, which
killed or damaged 26% of the remaining trees and reduced the canopy
cover from an original 80% to 43%. In general, logging damage reflects
the cutting and extraction of the timber itself, but secondary damage
arises when the logging trails in the forest provide access for other land
users, often farmers, who enter the area and continue the process of clear-
ance. Once significant canopy opening has occurred, fires can also acceler-
ate forest degradation, given the combustible slash that can accumulate at
ground level and the attendant drier microclimatic conditions (Uhl and
Kauffman, 1990; Nepstad et al., 1991). In recent years, widespread burn-
ing of damaged forests has been reported on several occasions in
Kalimantan, Indonesia (e.g. Malingreau et aI., 1985).
As yet, remote sensing has been little applied to monitoring forest dam-
age as a result of selective logging or other impacts. However, with the
development of high-resolution, cloud-penetrating microwave sensors,
such as the Canadian Radarsat launched in 1995, the possibility exists of
large-area monitoring of selective logging, by the detection of small gaps
where individual trees have been removed, or of associated logging trails
(Ahern et aI., 1994; Sanden, 1994).
5.2 BIOPHYSICAL IMPLICATIONS OF FOREST DEGRADATION

Forest degradation has impacts at both local and regional levels. Local
impacts mainly involve the on-site effects of clearance for agricultural or
pastoral purposes, impinging directly on individual or corporate land
users. The effects involve loss of available nutrients on cleared forest land,
notably through increased leaching or phosphorus fixation, and soil physi-
cal changes such as topsoil compaction and increased bulk density that
adversely affect soil aeration and root development and increase rates of
erosion. In addition, the productivity of cleared land may be impaired by
crop pests and diseases that arise in plantations and other simplified crop-
ping systems which displace the natural forest. Likewise, herbaceous and
woody weeds constantly compete with cultivated plants in crop and pasture
systems. Other local impacts also contribute to forest degradation. They
include extraction of timber and other forest products, mineral extraction,
and various urban-industrial developments.
Broader climatic, hydrological and biotic effects also derive from the
cumulative impact of multiple local clearance and damage. As well as
impinging on local land users, such effects have longer-term socio-eco-
nomic implications for human populations within and beyond the tropics
and, in the present context, warrant more detailed consideration.
5.2.1 Physical feedbacks of forest degradation

The broader physical effects of forest degradation include the climatic and
hydrological feedbacks of large-area forest clearance and damage.
Hitherto, such effects have been considered mainly in terms of deforesta-
tion, but forest damage is a contributing factor. In general, the initial
effects of forest clearance are reduced surface roughness, increased albedo
and modified run-off, which induce more general changes in regional
energy and water balances (Lean and Warrilow, 1989). Given a natural
forest cover, it is apparent from isotope studies in Amazonia that although
water vapour is essentially derived by air flow from oceanic sources, some
50-60% of the regional rainfall, averaging 2200 mm/year across the
Amazon basin, is the product of local evapotranspiration and re-precipita-
tion (Molion, 1976; Salati et aI., 1979). The presence of a natural forest
cover, enhancing evapotranspiration at the expense of run-off, promotes
local recycling and thus increases rainfall levels. Conversely, where forest
is degraded, terrestrial evapotranspiration and rainfall are reduced.
At what scale of clearance such changes become apparent is less clear.
In the Amazonian context, Franzle (1979) considers that localized defor-
estation 'limited to the areas adjoining transport routes and settlements'
would have little influence on precipitation levels, but given substantial
clearance, significant rainfall reductions may well occur. As yet, there is no
general evidence for such reductions, which is not surprising in view of
the modest scale of current deforestation (ca 10% in Brazilian Amazonia)
and the paucity of rainfall records in the area. Computer simulations that
assume the general replacement of forest by pasture or degraded grassland
provide some test of these assumptions, although initial results have been
somewhat ambivalent. However, Lean and Warrilow (1989) show a sig-
nificant lowering of both regional evapotranspiration and rainfall after
deforestation.
In addition, it has long been assumed that modified evapotranspiration
and run-off rates due to deforestation will translate into modified dis-
charge and sediment loads in tropical rivers (Richards, 1977; Sioli, 1980).
As indicated, the local response to deforestation is likely to be increased
run-off and erosion, but the situation becomes more complex at the
regional level, when postulated rainfall reductions attributable to large-
scale forest clearance are taken into account. It is possible that a new
equilibrium will arise involving a higher run-off ratio from a reduced total
rainfall, but what this means in terms of fluvial discharge is unclear;
increased peak flooding may still occur, but will depend as much on rain-
fall seasonality as annual total in the newly degraded landscape. Another
critical variable is the nature of the post-clearance land cover, which has
typically been assumed to be a herb-dominated community (see e.g. Lean
and Warrilow, 1989).
Evidence of increased peak flooding in parts of South and South-East
Asia as a result of deforestation has been cited by Myers (1988a, 1992),
with emphasis placed on the human cost in downstream agricultural areas
subjected to an increased range and intensity of flooding. Similar concerns
over increased annual flood levels have been expressed in Amazonia
(Gentry and Lopez-Parodi, 1980). However, confirmation of the links
involved is by no means straightforward and, in both the South Asian and
Amazonian contexts, doubts have been raised regarding the downstream
flood impacts of deforestation (Richey et aI., 1989; Ives, 1991).
Temperature increase has also been predicted as a function of large-
scale deforestation, with several factors again involved. The increased
albedo of cleared forest land will presumably reduce surface temperature,
but parallel reductions in evapotranspiration and hence evaporative cool-
ing will have a contrary effect. According to the simulation of Lean and
Warrilow (1989) for Amazonia, a net surface warming of 2.4° C would
result, with a complete replacement cover of pasture. In the past, it has
also been assumed that such warming would be reinforced by more gen-
eral global warming, at least partly due to a net flux to the atmosphere of
carbon dioxide (C0 2 ), nitrous oxide (N2 0) and possibly methane (CH4 )
as a result of forest degradation. While global warming may be occurring,
tropical forests are no longer generally perceived as net sources of carbon
dioxide and may indeed be net carbon sinks, at least partly as a result of
current forest regeneration after earlier clearance and damage (Lugo and
Brown, 1992; Battle et aI., 1996; Grace et aI., 1996).
Equally critical for some forest areas will be the effects of any related
rise in sea level. Few precise data exist, but extensive coastal and estuarine
areas in parts of the Old and New World tropics are likely to suffer
increased flood risk and associated socio-economic damage (e.g. Sattaur,
1990).
5.2.2 Biotic feedbacks of forest degradation

The biotic impact of large-scale forest degradation is relatively straight-
forward in that increasing numbers of plant and animal species will
surely be extinguished as degradation proceeds. What is unknown is the
likely scale of extinctions, not least because of inadequate data concern-
ing current species diversity; there may be millions or tens of millions of
species in the tropical forest, depending mainly on the diversity of insects
(Erwin, 1982, 1988; Myers, 1992). Any biome is of course prone to
extinctions when disturbed by humans, but the tropical forest is espe-
cially vulnerable as a function of its composition and dynamics. It is
highly diverse and thus prone to multiple extinctions, while many of its
species are specialized and inter-related and thus intolerant of general
forest degradation. Evolutionary factors have also resulted in localized
distributions for many species, which are correspondingly vulnerable to
local forest degradation.
The pattern of forest degradation will also affect the level of extinc-
tions, with particular reference to forest fragmentation. As indicated,
fragmented units of forest are likely to suffer higher levels of species loss.
This is partly because fragmented forest, as opposed to large residual
units, is more accessible and thus more prone to further human penetra-
tion and degradation. Fragmented forest will also suffer natural species
loss as a result of 'island' effects, i.e. the reduced species density associ-
ated with smaller units of a particular habitat (Diamond and May, 1976;
Terborgh, 1992). Both effects mean that the spatial pattern of forest
degradation, especially deforestation, will influence extinction rates and,
by implication, such effects are highly relevant to conservation planning
where consideration is given to the size of individual national parks, for
example, as well as to their total extent.
Attention to species extinction and indeed to any loss of biodiversity
has underpinned much of the recent scientific and popular concern over
tropical forests. Attention has also extended to concern about tribal pop-
ulations who inhabit forest areas and whose knowledge of the ecosystem
and its species is seen as a valuable resource in its own right (Sharpe, this
volume). While concern commonly focuses on the human rights of these
groups and their entitlement to land, the groups are also seen as sources
of information on forest products of potential commercial value (Eden,
1990; Alcorn, 1995). Conversely, any loss of forest species or indeed of
tribal populations puts any such broader benefits at risk.
It is thus important to emphasize the wide range of useful natural
products that exist in the tropical forest. Numerous plant and animal
products are already commercially exploited, but many other species of
potential value exist there. They include potential crops of value for the
production of fibres and fodder. Equally, the number of forest trees that
produce edible fruit or nuts is immense and although a few, such as
cocoa (Theobroma cacao) and Brazil nut (Bertholletia excelsa) are
already of major commercial importance, many others have unrealized
potential as food products.
Also of potential value are phytochemical products, which exist in
individual plants or taxa and have widespread industrial and other uses.
The products include latex, oils, waxes, resins, dyes, tannins and other
natural compounds (Prance, this volume). Latex, an emulsion of hydro-
carbons and water, is familiar as the raw material for the production of
rubber from Hevea brasiliensis, but many other species of similar poten-
tial exist in the forest and warrant investigation (Eden, 1990). Forest
plants also contain natural compounds of medicinal value. The latter
have been widely used by tribal groups, and some have been exploited by
modern society. Thus, the traditional hunting poison curare, derived
from several alkaloid-bearing plants, contains an active ingredient that is
employed as a muscle relaxant in surgery (Mors and Rizzini, 1966; Reis
Altschul, 1977). Meanwhile, screening of other forest plants needs to
continue in a search for natural compounds that will serve as direct con-
stituents of medicines or drugs, or can provide starting points for their
laboratory synthesis.
Loss of forest species will also put established products at risk. Species
that have already been brought into commercial cultivation still depend
on the genetic resources of their wild relatives and primitive cultivars for
breeding purposes. Even when major plant improvements have been
made in respect of energy and nutrient conversion or of pest resistance,
continuing plant improvement is required and depends on the availability
of a large and diverse gene pool. This applies as much to established
commercial species, such as rubber and cocoa, as it does to any new
domesticates, plant or animal, that may emerge from the forest. As mod-
ern land exploitation extends, such material is as vulnerable to genetic
erosion as wild forest species and needs to be conserved (Eden, 1990).
Despite the general concern over loss of forest biodiversity, it is diffi-
cult to estimate the risks involved. Some researchers express extreme con-
cern regarding forest degradation and its impact on biodiversity. Myers
(1988b), for example, describes a 'mass-extinction episode' that is
already under way and largely centred on tropical forests. Others, such
as Brown and Brown (1992), are more sanguine; with regard to the
Atlantic forests of Brazil, which have been massively reduced in area and
fragmented over recent centuries, they argue that little or no species
extinction has yet occurred. Such optimism is not generally shared across
the tropics, but limited data exist upon which to predict levels of species
extinction. As a result, there has lately been increasing resort to the 'pre-
cautionary principle' which expresses the desirability of risking too much
action rather than too little in the face of perceived threats (Heywood
and Stuart, 1992). The principle is a valid response to the uncertainty
involved in predicting extinction rates in tropical forests, but in practice
does need to be used as more than merely an excuse for further environ-
mental propaganda.
As in the case of the physical feedbacks of forest degradation, discus-
sion of biological feedbacks has focused mainly on deforestation.
However, forest damage also affects species composition and diversity,
and particularly in the case of selective logging that targets only a few
commercial species, may lead to their local decline or disappearance.
Indirect impacts on forest composition can also occur as a result of hunt-
ing activities which modify patterns of seed dispersal by particular ani-
mal species (Maury-Lechon, 1991).
5.3 STRATEGIES FOR FOREST CONSERVATION

In the present context, the main aim of forest conservation is to minimize
the biophysical feedbacks of human-induced forest degradation. As such,
conservation is part of the broader strategy of forest management, which
acknowledges the inevitability of continuing forest colonization and
development, but also stresses the need for explicit conservation measures.
Such conservation should be seen as a positive use of land and resources.
By forest conservation in this context is meant attention to protecting
both the physical and biological integrity of the forest environment. Much
conservation activity currently has biological objectives, notably conserv-
ing biodiversity, but it is equally important that attention is paid to physi-
cal objectives, namely minimizing the edaphic, climatic and hydrological
feedbacks of forest degradation. Both biological and physical objectives
may be met simultaneously, as when national parks are created. Although
primarily intended to protect flora and fauna, they also maintain a tree
cover that protects the integrity of the physical environment. Indigenous
reserves supporting low-density tribal populations may perform similar
functions.
The creation of indigenous reserves draws attention to the long-term
involvement of humans in the forest zone. Despite common assumptions
regarding the 'naturalness' of the tropical forest, it is increasingly evident
that most, if not all of the forest area has been exploited and its composi-
tion modified by humans over time. In many areas, shifting cultivators
have surely 'turned over' the entirety of the forest cover on a millennial
time scale, in parallel with the natural process of renewal driven by nat-
ural gap formation. As Balee (1989) asserts in Brazilian Amazonia, the
long-term indigenous impact has converted at least 12% of the forest
cover into an 'anthropogenic' forest. Elsewhere, the indigenous impact
may have been less, but is unlikely to have been absent.
What is needed today is for humans to recognize that they should co-
exist with the forest, some parts of which will admittedly be displaced,
but other parts of which need to remain more or less natural in character.
What this means in terms of actual area of national parks and the like on
the one hand, and of allowable development for agricultural and other
purposes on the other, is not easily determined. Estimates exist of the current
Strategies for forest conservation 111
extent of designated protected areas across the tropics, amounting on
paper to approximately 5% of the original forest cover (Table 5.1). Such
an area is surely insufficient to conserve the species diversity of the forest,
but the question arises as to what is the appropriate level of biological
conservation. On the basis of island biogeography theory, Myers (1979)
suggests for Amazonia that conserving some 10% of the forest area
should safeguard most of its species. Such data are no more than 'an
informed first guess' (Myers, 1979), and will in any case be affected by
how the protected areas in question are distributed (Eden, 1990). But the
data are a starting point for assessing the conservation requirement, and
do expose the inadequacy of existing provisions (Table 5.1).
Further issues arise. Firstly, few data exist on the overall status of
indigenous reserves in the forest zone and on their potential contribution
to conserving flora and fauna. To some extent, indigenous reserves com-
plement the protective function of national parks and the like, and addi-
tional investigation of this secondary role deserves more explicit
investigation. Secondly, it must be emphasized that the existence of 'paper
parks' often belies what is happening on the ground. Recurrent examples
exist in the tropics where national parks and indigenous reserves are inad-
equately marked or patrolled on the ground, and where invasion and
exploitation by shifting cultivators, ranchers or loggers recurs with
impunity (Fearnside and Ferreira, 1984; Tchamie, 1994). The reasons for
this may be acute shortages of land for impoverished peasant cultivators
or the greed and corruption of large corporations, but in either case the
environmental impact is much the same. Thirdly, it is evident that, even if
10-20% of the original forest cover is conserved and a high level of
species protection achieved (Myers, 1979), a much larger area of forest, or
at least tree cover, is needed to provide adequate physical protection of the
landscape. While physical protection is qualitatively less demanding, in
that a managed tree cover, as well as a natural forest, serves to protect the
soil, minimize run-off, and maximize evapotranspiration, such protection
is more demanding in quantitative terms, i.e. in the area of land needing
tree cover.
In these circumstances, 'compromise' land-use systems (Odum, 1969;
Eden, 1978) such as agroforestry or mixed-forest silviculture, which pro-
tect the land as well as produce from it, can contribute to broader conser-
vation objectives. This is especially so if they are located as buffer zones
around national parks. Even commercial tree plantations, if broadly sus-
tainable, will contribute to physical conservation. In parts of the tropics
where substantial deforestation has already occurred, managed tree covers
of this kind are often seen as means of rehabilitating land that has previ-
ously been cleared and degraded.
What sort of assemblage of national parks, indigenous reserves, tree-
covered buffer zones and other land uses is environmentally appropriate
Table 5.1 Protected areas within moist tropical forests (after Sayer, 1991)
Total area
Number of areas (million hay
Afrotropical 62 9.4
Indomalayan 386 18.3
Australian 72 8.2
Neotropical 149 30.1
Total 669 66.0
for forest areas has not been clearly identified. Some attention has been
given in recent decades to 'ecological zoning' of the landscape as a basis
for pursuing parallel development and conservation objectives (Eden,
1990), and this is an approach that has some merits in the present con-
text. Ultimately, however, the use of forest land in the tropics is a political
issue, and in practice environmental considerations have commonly
received short shrift. Nevertheless, the environmental case nowadays
receives greater attention at the political level, as was symbolized by the
UN Conference on Environment and Development at Rio de Janeiro in
1992. Some hope thus exists of improving the conservation prospects of
tropical forests in the longer term, and although many obstacles exist to
such progress, it is sensible to have in place conservation strategies that
are technically sound in character and available as and when they are
needed. Among the technical aids thereto is remote sensing, with its capa-
bility of monitoring forest change, which provides a means of quantifying
forest degradation. Its data are of direct technical value, but they also pro-
vide a critical basis for a continuing and necessary dialogue on forest
degradation and conservation between international agencies, national
governments, non-governmental organizations, the media and local popu-
lations.
REFERENCES
Ahern, F. J., Kux, N. H. H., Salcedo, R. and Pietsch, R. W. (1994) An evaluation
of C-band radar to detect openings in moist tropical forests, in South
American Radar Experiment, SAREX-92, Workshop Proceedings, 6-8
December 1993, (eds M. Wooding and E. Attema), European Space Agency,
Paris, pp. 57-69.
Alcorn, J. B. (1995) The scope and aims of ethnobotany in a developing world, in
Ethnobotany: Evolution of a Discipline, (eds R. E. Schultes and S. von Reis),
References 113
Chapman & Hall, London, pp. 23-39.
Balee, W. (1989) The culture of Amazonian forests, in Resource Management in
Amazonia: Indigenous and Folk Strategies, (eds D. A. Posey and W. Balee),
New York Botanical Garden, New York, pp. 1-21.
Battle, M., Bender, M., Sowers, T. et at. (1996) Atmospheric gas concentrations
over the past century measured in air from firn at the South Pole. Nature, 383,
231-235.
Blaikie, P. and Brookfield, H. (1987) Land Degradation and Society, Methuen,
London.
Brown, K. S. ]r and Brown, G. G. (1992) Habitat alteration and species loss in
Brazilian forests, in Tropical Deforestation and Species Extinction, (eds T. C.
Whitmore and]. A. Sayer), Chapman & Hall, London, pp. 119-142.
Chazdon, R. L. (1994) The primary importance of secondary forests in the trop-
ics. Tropinet, 5, 2.
Chomentowski, W., Salas, B. and Skole, D. (1994) Landsat Pathfinder project
advances deforestation mapping. GIS World, 7, 34-38.
Corlett, R. T. (1995) Tropical secondary forests. Progress in Physical Geography,
19, 159-172.
Curran, P. ]., Foody, G. M., Lucas, R. M. and Honzak, M. (1995) A methodol-
ogy for remotely sensing the stage of regeneration in tropical forests, in
TERRA 2 - Understanding the Terrestrial Environment. Remote Sensing
Data Systems and Networks, (ed. P. M. Mather), Wiley, Chichester,
pp.182-202.
Dargie, T. and Furley, P. A. (1994) Monitoring change in land use and the envi-
ronment, in The Forest Frontier: Settlement and Change in Brazilian Roraime,
(ed. P. A. Furley), Routledge, London, 68-85.
Denevan, W. M. (1973) Development and the imminent demise of the Amazon
rain forest. Professional Geographer, 25, 130-135.
Diamond, ]. M. and May, R. M. (1976) Island biogeography and the design of
natural reserves, in Theoretical Ecology: Principles and Applications, (ed. R.
M. May), Blackwell Scientific, Oxford, pp. 163-186.
Eden, M. ]. (1978) Ecology and land development: the case of Amazonian rain-
forest. Transactions of the Institute of British Geographers, 3, 444-463.
Eden, M.]. (1990) Ecology and Land Management in Amazonia, Belhaven Press,
London.
Eden, M. ]. (1994) Environment, politics and Amazonian deforestation. Land
Use Policy, 11, 55-66.
Eden, M. ]. (1996a) Environmental degradation and forest renewability in
Amazonia, in Land Degradation in the Tropics: Environmental and Policy
Issues, (eds M.]. Eden and]. T. Parry), Pinter, London, pp. 48-60.
Eden, M. ]. (1996b) Forest degradation in the tropics: environmental and man-
agement issues, in Land Degradation in the Tropics: Environmental and Policy
Issues, (eds M.]. Eden and]. T. Parry), Pinter, London, pp. 41-47.
Eden, M. ]. (1996c) Sustainable silvicultural systems in the British
Commonwealth, with particular reference to Guyana, in Global Change and
Commonwealth, (ed. Y. Yeung), Hong Kong Institute of Asia-Pacific Studies,
Hong Kong, pp. 195-213.
Erwin, T. L. (1982) Tropical forests: their richness in Coleoptera and other
arthropod species. Coleopterists' Bulletin, 36, 74-75.
Erwin, T. L. (1988) The tropical forest canopy: the heart of biotic diversity, in
Biodiversity, (ed. E. o. Wilson), National Academy Press, Washington, DC,
pp. 123-129.
FAO (1993) Forest Resources Assessment 1990. Tropical Countries, FAO
Forestry Paper 112, Food and Agriculture Organization, Rome.
Fearnside, P. M. (1979) Cattle yield prediction for the Transamazon highway of
Brazil. Interciencia, 4, 220-225.
Fearnside, P. M. and Ferreira, G. de L. (1984) Roads in Rondonia: highway con-
struction and the farce of unprotected forest reserves in Brazil's Amazonian
forest. Environmental Conservation, 11, 358-360.
Foody, G. M. and Curran, P. J. (1994) Estimation of tropical forest extent and
regenerative stage using remotely sensed data. Journal of Biogeography, 21,
223-244.
Friinzle, O. (1979) The water balance of the tropical rain forest of Amazonia
and the effects of human impact. Applied Sciences and Development, 13,
88-117.
Gentry, A. H. and Lopez-Parodi, J. (1980) Deforestation and increased flooding
of the upper Amazon. Science, 210, 1354-1356.
G6mez-Pompa, A., Vazquez-Yanes, C. and Guevara, S. (1972) The tropical rain
forest: a non-renewable resource. Science, 177, 762-765.
Grace, ]., Lloyd, ]., Mcintyre, ]., Miranda, A., Meir, P. and Miranda, H. S.
(1996) Carbon dioxide flux over Amazon Forest in Rondonia, in Amazonian
Deforestation and Climate, (eds]. H. C. Gash, C. A. Nobre, J. M. Roberts and
R. L. Victoria), Wiley, Chichester, pp. 307-318.
Grainger, A. (1983) Improving the monitoring of deforestation in the humid trop-
ics, in Tropical Rain Forest Ecology and Management, (eds T. C. Chadwick
and A. C. Chadwick), Blackwell Scientific, Oxford, pp. 387-395.
Grainger, A. (1993) Rates of deforestation in the humid tropics: estimates and
measurements. Geographical Journal, 159, 33-44.
Grainger, A. (1996) Degradation of tropical rain forest in Southeast Asia: taxon-
omy and appraisal, in Land Degradation in the Tropics: Environmental and
Policy Issues, (eds M.]. Eden and J. T. Parry), Pinter, London, pp. 61-75.
Heywood, V. H. and Stuart, S. N. (1992) Species extinctions in tropical forests, in
Tropical Deforestation and Species Extinction, (eds T. C. Whitmore and J. A.
Sayer), Chapman & Hall, London, pp. 91-117.
Ives, ]. (1991) Floods in Bangladesh: who is to blame? New Scientist, 130, 1764,
34-37.
Kummer, D. M. (1992) Remote sensing and tropical deforestation: a cautionary
note from the Philippines. Photogrammetric Engineering and Remote Sensing,
58,1469-1471.
Lanly, J. P. (1981) Tropical Forest Resources Assessment Project (GEMS):
Tropical Africa, Tropical Asia, Tropical America. FAOIUNEP, Rome.
Lean,]. and Warrilow, D. A. (1989) Simulation of the regional climatic impact of
Amazon deforestation. Nature, 342, 411-413.
Lucas, R. M., Honzak, M., Foody, G. M., Curran, P. ]. and Corves, C. (1993)
References 115
Characterizing tropical secondary forests using multi-temporal Landsat sensor
imagery. International Journal of Remote Sensing, 14, 3061-3067.
Lugo, A. E. and Brown, S. (1992) Tropical forests as sinks of atmospheric car-
bon. Forest Ecology and Management, 54, 239-255.
McCloskey, M. (1993) Note on the fragmentation of primary rainforest. Ambio,
22,250-251.
Malingreau, J.-P. (1993) Satellite monitoring of the world's forests: a review.
Unasylva, 44, 31-38.
Malingreau, J.-P. and Duchossois, G. (1995) The TREES/ERS-1 SAR '94 Project.
Earth Observation Quarterly, 48, 1-5.
Malingreau, J.-P., Stevens, G. and Fellows, C. (1985) 1982-83 forest fire of
Kalimantan and North Borneo: satellite observations for detection and moni-
toring. Ambio, 14, 314-346.
Maury-Lechon, G. (1991) Comparative dynamics of tropical rain forest regenera-
tion in French Guyana, in Rain Forest Regeneration and Management, (eds A.
Gomez-Pompa, T. C. Whitmore and M. Hadley), UNESCO, ParislParthenon,
Carnforth, pp. 285-293.
Mausel, P., Wu, Y., Li, Y., Moran, E. F. and Brondizio, E. S. (1993) Spectral iden-
tification of successional stages following deforestation in the Amazon.
Geocarto International, 4, 61-71.
Molion, L. C. B. (1976) A Climatonomic Study of the Energy and Moisture
Fluxes of the Amazonas Basin with Considerations of Deforestation Effects,
Instituto de Pesquisas Espaciais, Sao Jose dos Campos.
Mors, W. B. and Rizzini, C. T. (1966) Useful Plants of Brazil, Holden-Day, San
Francisco.
Myers, N. (1979) The Sinking Ark: A New Look at the Problem of Disappearing
Species, Pergamon, Oxford.
Myers, N. (1980) Conversion of Tropical Moist Forests, National Academy of
Sciences, Washington, DC.
Myers, N. (1988a) Environmental degradation and some economic consequences
in the Philippines. Environmental Conservation, 15,205-214.
Myers, N. (1988b) Threatened biotas: 'hot spots' in tropical forests.
Environmentalist, 8, 187-208.
Myers, N. (1992) The Primary Source: Tropical Forests and our Future, 2nd edn,
W. W. Norton, New York.
Nepstad, D. c., Uhl, C. and Serrao, E. A. S. (1991) Recuperation of a degraded
Amazonian landscape: forest recovery and agricultural restoration. Ambio, 20,
248-255.
Odum, E. P. (1969) The strategy for ecosystem development. Science, 164,
262-270.
Ravinandrath, N. H. and Hall, D. O. (1994) Indian forest conservation and trop-
ical deforestation. Ambio, 23,521-523.
Reid, W. V. (1992) How many species will there be? in Tropical Deforestation
and Species Extinctions, (eds T. C. Whitmore and ]. A. Sayer), Chapman &
Hall, London, pp. 15-53.
Reis Altschul, S. von (1977) Exploring the herbarium. Scientific American, 236
(5),96-104.
Richards, P. W. (1973) The tropical rain forest. Scientific American, 229 (6),
58-67.
Richards, P. W. (1977) Tropical forests and woodlands: an overview. Agro-
Ecosystems, 3, 225-238.
Richey,]. E., Nobre, C. and Deser, C. (1989) Amazon river discharge and climate
variability: 1903 to 1985. Science, 246, 101-103.
Salati, E., Dall'Olio, A., Matsui, E. and Gat, ]. R. (1979) Recycling of water in
the Amazon basin: an isotopic study. Water Resources Research, 15,
1250-1258.
Sanden, ]. van der (1994) Textural discrimination of forest types in Guyana.
Earth Observation Quarterly, 44, 2l.
Sattaur, O. (1990) Guyana's test at high tide. New Scientist, 125,46-49.
Sayer, J. (1991) Conservation and protection of tropical rain forests: the perspec-
tive of the World Conservation Union. Unasylva, 42, 40-45.
Simberloff, D. (1986) Are we on the verge of a mass extinction in tropical rain
forests? in Dynamics of Extinction, (ed. D. K. Elliott), Wiley, New York,
pp. 165-180.
Sioli, H. (1980) Foreseeable consequences of actual development schemes and
alternative ideas, in Land, People and Planning in Contemporary Amazonia,
(ed. F. Barbira-Scazzocchio), Centre of Latin American Studies, Cambridge,
pp. 257-268.
Skole, D. L. and Tucker, C.]. (1993) Tropical deforestation and habitat fragmen-
tation in the Amazon: satellite data from 1978 to 1988. Science, 260,
1905-1910.
Skole, D. L., Chomentowski, W. H., Salas, W. A. and Nobre, A. D. (1994)
Physical and human dimensions of deforestation in Amazonia. BioScience, 44,
314-322.
Sommer, A. (1976) Attempt at an assessment of the world's tropical moist forests.
Unasylva, 28, 5-24.
Tchamie, T. T. K. (1994) Learning from local hostility to protected areas in Togo.
Unasylva, 45, 22-27.
Terborgh, J. (1992) Diversity and the Tropical Rain Forest, Scientific American
Library, New York.
Uhl, C. (1988) Restoration of degraded lands in the Amazon basin, in
Biodiversity, (ed. E. O. Wilson), National Academy Press, Washington, DC,
pp. 326-332.
Uhl, C. and Kauffman, ]. B. (1990) Deforestation, fire susceptibility, and poten-
tial tree responses to fire in the eastern Amazon. Ecology, 71, 437-449.
Uhl, C. and Vieira, I. C. G. (1989) Ecological impacts of selected logging in the
Brazilian Amazon: a case study from the Paragominas region of the State of
Para. Biotropica, 21,98-106.
References 117
Uhl, c., Bushbacher, R. and Serrao, E. A. S. (1988) Abandoned pastures in east-
ern Amazonia. 1. Patterns of plant succession. Journal of Ecology, 76,
663-681.
Verissimo, A., Barreto, P., Mattos, M., Tarifa, R. and Uhl, C. (1996) Impactos de
actividade madeireira e perspectivas para 0 manejo sustentavei da floresta
numa veiha fronteira da Amazonia: 0 caso Paragominas, in A Evolur;ao da
Fronteira Amazonica, Oportunidades para urn Desenvolvimento Sustentavel.
Edi~oes Caraveia, Porto Alegre, pp. 7-37.
-6------
Soil and vegetation effects of tropical
deforestation
s. M. Ross
SGS Environment, Yorkshire House, Chapel Street, Liverpool, Merseyside
L3 9AG, UK
'Nature nowhere abhors a vacuum more strongly than on a patch of bare

soil in the humid tropics.'
Whitmore (1991)
'Widespread deforestation is creating a landscape in which mature tropical
forests are islands in a sea of successional vegetation.'
Uhl (1987)
6.1 INTRODUCTION
The regrowth of vegetation after deforestation provides a paradox: of
rapid growth but with species composition and diversity at odds with the
original primary forest. Secondary species which regenerate from seed in
the soil after clearance are totally different from the primary species which
dominate 'climax' tropical rain forests. Regrowth of primary forest after
disturbance depends on distance and access to seed sources of primary for-
est species. On-site biological effects of deforestation are also controlled by
alteration of the physicochemical environment in which plants and animals
live and grow, combined with the resilience to perturbation shown by plant
and animal associations, both above and below the soil surface. Resilience
is partly determined by reproductive abilities of plants, both vegetative and
sexual, the volume, frequency, dispersal, dormancy and germination of
their seeds. Resilience is also determined by the speed of vegetation
regrowth, as this affects the rate of alteration of soil and microclimate con-
ditions of the site. A range of research studies world-wide, in the neo-
tropics, Africa and the Far East, have charted specific ecological changes
taking place when tropical forests are removed. Some are site-specific,
120 Soil and vegetation effects of tropical deforestation
others provide more process-related information which helps in predicting
how newly deforested sites might respond over time.
A key problem in tropical rain forest regeneration is the degree to which
below-ground biological processes control regeneration and the degree to
which they are disrupted by deforestation practices. Many authors have
discussed the nutrient-conserving abilities of tropical forests (e.g. Snedaker,
1980; Jordan and Herrera, 1981; Medina and Cuevas, 1989), including
rapid leaf litter decomposition mechanisms linked to immediate nutrient
uptake in the root mat (e.g. Stark and Jordan, 1978); fungal decay of
woody tissues under high humidity conditions; mycorrhizal associations
with feeding roots; and nitrogen-fixing symbiotic and non-symbiotic bac-
teria. Jordan et al. (1979) hypothesized that the suppression of nitrifying
bacteria by both low pH and high concentrations of tannins in the root
mat results in nitrogen conservation in the rain forest. The degree and
longevity of interruption of soil biological processes after clearing may be
key issues in the re-establishment of primary forest species. The interrup-
tion of these processes may be one of the main edaphic differences between
the germination conditions of a post-clearance seedbed and those of an
undisturbed forest soil. There is still a dearth of good research data on
these issues.
This chapter aims to examine the effect that tropical deforestation has
on soil properties, soil processes and vegetation, and to explore the poten-
tial for regeneration and ecosystem sustainability after deforestation. To do
this adequately it is first necessary to distinguish between the effects of dif-
ferent types of deforestation practices, including clear felling, selective log-
ging and timber haulage, bulldozing and slash-and-burn. It is also
necessary to consider the size of cleared areas, as this will determine the
degree of 'connectivity' between regenerating areas and their sources of
seed, vegetative growth and soil biology, such as soil organisms and mycor-
rhizae. Secondly, it is valuable to differentiate between those effects caused
by alteration of in situ environmental conditions, particularly changes in
microclimate and soil, and more direct effects caused by death or damage
to the original vegetation. A third part of the discussion must consider to
what degree, if any, increasing knowledge of 'natural' forest gap processes
and regeneration patterns can aid an understanding of vegetation recovery
after deforestation of larger areas. It is also relevant to question the idea of
the so-called irrecoverability of deforested areas in light of the view that
natural forest disturbance provides the means for maintaining maximum
site and species diversity.
There is a need for a clear understanding of the factors governing the
recovery of forest biomass, structure and species diversity, as this provides
the basis both for intervening to limit site alteration and for making
rehabilitation more effective.
Deforestation effects on microclimate 121
6.2 EFFECTS OF DIFFERENT DEFORESTATION PRACTICES ON
SOIL AND VEGETATION
The choice of land-clearing method depends primarily on whether timber,
land or both is required for use after deforestation. Wholesale clear-felling
may be employed only where timber is required for energy generation or in
the production of charcoal, pulp or chipwood. In many tropical regions
where the requirement for land outweighs the requirement for timber,
slash-and-burn deforestation techniques prevail. These practices have gen-
erally been used in shifting cultivation and pasture production. Sometimes
slash is burnt in situ, sometimes it is bulldozed into rows or heaps and then
burnt. The effects of fire include nutrient losses in smoke, ash deposition,
altered soil surface albedo and sometimes hydrophobicity of the soil sur-
face, causing reduced infiltration which can lead to droughty soil condi-
tions. Mechanized forest clearance methods often involve the use of
bulldozers for the removal of timber and brash and for the levelling of the
site. These techniques not only cause severe soil rutting and compaction
but sometimes also scrape off the topsoil which contains the bulk of the
feeding roots, soil organic matter and soil nutrients. Selective logging, for
the extraction of desired timber from a dense stand of 'undesirable' trees,
does not disrupt the entire site, but does create large gaps of disturbance,
caused by toppling of groups of adjacent trees whose crowns have been
linked through liana and vine entanglement. Resultant timber haulage usu-
ally removes litter and topsoil.
Deforestation practices thus cause direct site alteration and damage
through:
• canopy removal, with resultant changes in both above-ground and
below-ground microclimate;
• soil compaction, loss of soil structure and even topsoil removal, with
resultant changes in soil physical and chemical properties;
• volatilization of plant nutrients by fire, followed by return of nutrients
in deposited ash.
In addition to these physicochemical changes, deforestation has a direct
effect on soil biology and vegetation through:
• canopy removal and loss of both seed and litter influx;
• damage to in situ regenerating seedlings;
• damage to soil surface root mat, soil microbial population and seed
bank.
6.3 DEFORESTATION EFFECTS ON MICROCLIMATE

Vegetation regeneration after forest clearance depends partly on sprouting
of old plants not killed by felling, and partly on the germination of new
seedlings. Changes in microclimate, in light, temperature and moisture
regime, are vital triggers for germination of seeds. Regenerating vegetation
itself progressively alters the microclimate of clearings, rendering them
more amenable to shade-loving and humid plants and animals, including
soil organisms.
Surface albedo (reflectivity) of forests is low in the visible wavelengths,
as much of the incident solar radiation of these wavelengths is absorbed by
the canopy. Surface albedo is defined as the ratio of solar radiation
reflected by the surface to the total incident solar radiation. Potter et al.
(1981) suggest that deforestation and the regrowth of herbaceous vegeta-
tion after clearance could increase the albedo loS-fold, while change from
savanna to desert could increase the albedo by a factor of 3.4. The upper
forest canopy in the wet tropics receives less radiant energy than the
canopy of other tropical forests as water molecules in the high humidity
conditions absorb radiation (Bazzaz, 1991). The removal of forest cover in
the humid tropics is thus likely to cause a greater increase in albedo than
elsewhere. These changes in radiation balance at the soil-plant interface
cause changes in the three main elements of microclimate which affect
plant growth: light, temperature and relative humidity. The leaf area index
and leaf absorbency of the upper rain forest canopy determine the quantity
and quality of light reaching the understorey and the forest floor. Chazdon
and Fetcher (1984a) compared the light characteristics of a rain forest
understorey with that of an adjacent O.S-ha clearing in Costa Rica.
Photosynthetic photon flux density (PPFD) showed a maximum of greater
than 1000 J.lmollm 2/s in the clearing, but only 90 J.lmollm 2/s in the forest
under storey (Figure 6.1a). The proportion of PPFD reaching the forest
under storey, calculated as a percentage of that received in the clearing, was
low at 4-6%, with higher values recorded under cloudy, overcast condi-
tions, compared to sunny conditions. As plant foliage absorbs red light,
radiation received in the understorey is depleted in red wavelengths, as
shown by Chazdon and Fetcher (1984b) for forest and clearing conditions
in Costa Rica (Figure 6.1b). These differences in quantity and spectral
quality of light received by regenerating plants in clearings affect seed ger-
mination and the type and rate of vegetative and seedling growth. These
issues are discussed in section 6.9 below.
Consistently higher soil temperatures have been recorded after forest
clearance compared to undisturbed forest, and the differences appear to
depend both on the method of forest clearance and on the season. Ghuman
and Lal (1987), for example, report consistently lower air and soil temper-
atures under Nigerian rain forest compared to adjacent clearings. They
found that mean air temperatures, measured at 1 m above ground, were
3-6° C lower under forest, while mean forest soil temperatures at a depth
of Scm were around 3° C lower. Air diurnal temperature ranges at 1 m
were higher in clearings (26-34° C) compared to forest (26-30° C), as
were soil temperature ranges at Scm depth (26-28° C in forest; 28-34° C
Deforestation effects on microclimate 123
~
·iii \~/\
c:: 1000
Q) \
'"\400 m2
-C
Q)
x
~
;;::::
cuu 200 m2
(/) \ gap
c::
(5
0 CJ)
g \
..c:: 100 \...\
Q.Ci)
u;:;-.
- --Q) E
~
..c:: "0
c::
>- E
- 2;
(/)
0 10
0
..c::
a..
1
0600 0800 1000 1200 1400 1600 1800
(a) Time of day (h)
0.35 10
Q)
-C
I"!S
..c::
~ 0.28 8
>-
.~ .-...
(/) (/)
5i
-C "'~
__ 0.21 69
(1)
I\)
x "0
-
~
::l
;;:::: ~ 0.14 4~·
-
E
::l
C 0.07 2
I"!S
a
::l
Shade
0 0
3 4 5 6 7 8 9 10 11
(b) Wavelength (nm x 10.2 )
Figure 6.1 (a) Daily pattern of photosynthetic photon flux density in tropical rain
forest in Costa Rica for a sunny day in clearing (5000 m2 ); 400 m2 gap; 200 m2
gap; and forest understorey (from Chazdon and Fetcher, 1984a). (b) Spectral distri-
bution of radiant energy in a clearing and in the shade of a tropical rain forest in
Costa Rica (from Chazdon and Fetcher, 1984b).
in clearing). Lal and Cummings (1979) noted that maximum surface soil
temperatures in areas which had been cleared mechanically were, on aver-
age, 5° C higher than areas which had been cleared manually. Burning and
the deposition of ash and charcoal on the soil surface also alter soil temper-
atures through changes in the albedo. de Souza et al. (1996) compared soil
temperatures for different seasons and at different depths under forest and
under adjacent pasture developed after forest clearance in eastern
Amazonia. The surface 20 cm of soil was, on average, 4° C higher in pas-
ture during the dry season, declining to 2° C difference in the wet season.
Although elevated temperatures in the surface soil of clearings can have
a pronounced effect on germination, early growth and survival of
seedlings, vegetation regrowth rapidly causes a reversion of microclimate
conditions. Fetcher et at. (1984) found that 2 years after deforestation, the
microclimate of a 0.5-ha clearing at seedling height was very similar to that
of the adjacent forest understorey.
Temperature differences between forests and clearings have important
knock-on effects on relative humidity after clearance. In Nigeria, maxi-
mum air temperatures of 26 and 37° C on forested and cleared sites,
respectively, corresponded with minimum relative humidities of 87 and
49% (Ghuman and Lal, 1987). These authors also report reductions of
20-40% in daytime relative humidity in cleared areas in the dry season.
Even under these dry conditions, relative humidity on cleared sites persis-
tently rose overnight to 95%. In the same study, open-pan evaporation in
cleared areas, although variable, averaged around 3-4 mm/day, while the
much more consistent values for adjacent forest averaged 0.5-1.5 mm/day.
Micrometeorological changes due to deforestation affect seedling germi-
nation, as cleared sites are characterized by greater ranges of air and soil
temperature and humidity, as well as an altered spectrum of solar radiation
receipt (see section 6.9).
6.4 DEFORESTATION EFFECTS ON SOIL PHYSICAL PROPERTIES

Perhaps one of the most documented results of tropical forest clearance is
soil compaction resulting from the use of heavy, mechanized felling and
haulage techniques. Soil compaction is due to loss of soil structure and
causes reduced infiltration, ponding of surface water on flat land and
increased surface run-off on slopes, poor aeration and mechanical imped-
ance of growing roots. All of these soil changes have a detrimental effect
on the ability of seeds to germinate and of seedlings to establish in cleared
areas.
Dias and Nortcliff (1985), studying the effects of deforestation in central
Amazonia, showed that there was a good relationship between the number
of tractor passes and the degree of soil compaction. Soil damage increased
with higher soil moisture contents and vehicle impacts were greater where
manoeuvres involved wheel slippage. Mello Ivo et at. (1996), working on
selective logging in central Amazonia, found that the majority of vehicle
tracks showed surface water ponding and localized soil anaerobic conditions,
Deforestation effects on soil physical properties 125
due to rutting and soil compaction. Similar compaction conditions are
reported by Nussbaum et al. (1995) for log-landing areas and skid trails
associated with timber harvesting in Sa bah, Malaysia. Soil compaction in
haulage tracks does not appear to improve significantly, even after years of
non-use (e.g. MaImer and Grip, 1990).
Alegre et at. (1986) and Alegre and Cassel (1986) compared the effects
on soil physical properties of several different kinds of mechanized forest
clearance and of slash and burn at Yurimaguas in the Amazon region of
Peru. For all properties studied, including soil bulk density, infiltration rate,
aggregate stability and soil organic carbon, more severe damage was
caused by mechanical procedures. Compared to undisturbed forest, the
mean weight diameter of soil aggregates after clearance was reduced by
12.4,26.0 and 39.7%, respectively, for slash and burn, bulldozing with a
shear blade and bulldozing with a straight blade (Alegre and Cassel, 1986).
These changes were significantly correlated with reductions in soil
organic carbon of 0, 16.3 and 21.2% for slash and burn, shear-blade bull-
dozing and straight-blade bulldozing, respectively. The loss of soil structure
which contributes to soil compaction after clearance is partly caused by
direct crushing by heavy vehicles, and partly by decline in soil organic mat-
ter due to rapid decomposition after clearance, combined with a lack of
replenishment after tree removal due to the loss of litter influx. Also work-
ing at Yurimaguas in Peru, Seubert et al. (1977) reported that infiltration
rates were 12 times higher after slash and burn than after bulldozing, and
that this paralleled an increase in topsoil bulk density of 0.9% and 14% at
29 and 98 weeks respectively after slash and burn, and 8.6% and 22% at
29 and 98 weeks respectively after bulldozing with a straight blade (Alegre
et aI., 1986).
The main effect of these changes in soil physical properties on plant
growth is through loss of topsoil macroporosity. Van der Weert (1974)
showed significant decreases in the volume of soil meso- and macropores
(>30 J..Lm pore diameter) and increases in the volume of micropores
(<30 J..Lm pore diameter) after bulldozer clearance in Suriname. The result-
ing impaired soil aeration, together with mechanical impedance, restricted
the downward growth of roots. Van der Weert (1974) reported severe
reductions in root density at a depth of 10-40 cm, associated with soil
compaction. He noted that the effect of compaction on rooting depth is
more temporary for perennials than it is for annuals, which have to form
their root systems in a much shorter time.
Jordan et al. (1956) and, more recently, Chauvel et al. (1991), illustrated
the value of post-clearance ground-cover vegetation in reversing the
adverse effects of compaction on topsoil pore-size distribution and in rein-
stating soil structure. Both studies report the effects of the prolific ground-
cover legume Pueraria. Jordan et al. stress the value of intercepting
above-ground foliage, while Chauvel et al. illustrate the importance of
Pueraria roots in altering the pore-size distribution, particularly reduction
in volume of micropores, input of organic matter and improvement of soil
structure. Chauvel et at. (1991) suggest that soil structure cannot be
restored after forest clearance until vegetation colonization becomes vigor-
ous. Surface mulching with crop residues after forest clearance has also
been shown to be beneficial for a whole range of soil physical properties,
including reduced soil bulk density and improved topsoil porosity, which
result in increasing infiltration rates at the soil surface (Lal et al., 1980).
Mulching also helps to increase the number of soil organisms, particularly
earthworms, with knock-on positive effects on soil structure. Improved soil
aggregation and stability of aggregates after mulching (Lal et at., 1980) not
only make soils more water permeable, but also make them less susceptible
to drought.
Loss of soil structure and resulting soil compaction after forest-clearance
operations have a serious impact on soil hydrology, particularly infiltration
capacity, water-holding capacity and hydraulic conductivity. MaImer and
Grip (1990) found that mechanical clearance of forest in Sabah signifi-
cantly reduced infiltration capacity, leading to increased rates of surface
water run-off and accelerated soil erosion. The altered soil pore distribu-
tion described by Van der Weert (1974), particularly the loss of macro-
pores and increase in the number of micropores, results in reduced rates of
unsaturated hydraulic conductivity, especially under wetter soil conditions,
and greater storage of soil moisture at higher soil moisture tensions (Jetten,
1994). As a result, soils after deforestation are generally wetter in the rainy
season and take longer to dry out, but conversely become dryer in the dry
season and take longer to wet up. Both conditions are adverse for plant
growth. The effects of forest clearance and subsequent regeneration on soil
moisture conditions are reviewed in detail by Bruijnzeel (1996).
Plants rely on two main mechanisms for obtaining access to nutrients in
soil: root extension and mass flow/diffusion processes. Both of these mech-
anisms may be adversely affected after mechanical forest clearance, as
(i) compaction physically slows down or even restricts root growth, and
(ii) the more tortuous water flow path of soils consisting of micropores
means a longer route for mass flow and diffusion of ions, combined with
reduced rates of saturated and unsaturated hydraulic conductivity.
Deterioration of the physical condition of soil after forest clearance, along
with direct effects of deforestation on soil chemistry, can thus play an
important part in subsequent poor plant nutrition.
6.5 DEFORESTATION EFFECTS ON SOIL EROSION

Forest tree cover, understorey vegetation canopies and fallen leaf litter pro-
tect the soil surface from soil erosion. Once they are removed during forest
clearance, the soil surface is vulnerable to energy dissipated by raindrop
impact and soil particles can be easily detached and transported.
Deforestation effects on soil erosion 12 7
Rainsplash and the breakdown of soil organic matter after deforestation
cause disaggregation of surface soil material, blockage of micropores,
reduced infiltration and enhanced surface run-off. Any soil surface charac-
teristics, such as impermeable clay texture, surface crusting, or even a shal-
low slope, will enhance overland flow rather than infiltration, and will
encourage soil erosion.
Deforestation encourages on-site erosion (splash, sheet and rill erosion),
larger-scale gully erosion and other forms of mass movement. A wide range
of soil erosion rates have been reported for very different deforestation, soil
and slope conditions. Wiersum (1984) reviewed around 80 studies of sur-
face erosion in tropical forest and agroforestry systems. Natural forests
showed the lowest rates of erosion, with a range of 0.03-6.2 t/ha/year. The
fallow period of shifting cultivation, with a range of 0.05-7.4 t/ha/year, is
not significantly different, as the ground vegetation protects the soil sur-
face. The highest rates of soil erosion were found in clean-weeded tree
crops and forest plantations where the litter had been removed, which had
erosion rates of 1.2-183 and 5.9-105 t/ha/year, respectively. These high
erosion rates are due to lack of protection of the soil surface, in combina-
tion with throughfall from the high tree canopy. Soil erosion rates from
clear-felled areas which are subsequently cropped can rise as high as
70 t/ha/year due to the exposure of bare soil surfaces associated with
tillage.
Soil loss per se excavates established root systems, and unstable soil sur-
face conditions cause problems for vegetation establishment. The losses of
soil organic matter and nutrients in soil erosion after deforestation, how-
ever, can have a lasting effect on re-establishment of vegetation as soils
become infertile and droughty. The two main factors influencing soil ero-
sion and nutrient loss are slope angle and vegetation cover. Lal (1976)
found very high rates of organic matter and nutrient losses under bare fal-
low conditions in Nigeria. Erosive losses of organic matter rose from
416-3780 kg/ha/year with an increase in slope angle from 1-15%. These
losses amounted to 36.7 and 313.5 kg/ha/year of soil nitrogen, rates which
are higher than fertilizer nitrogen applications for intensive, arable agricul-
ture. Quantities of organic matter and nutrients lost in soil erosion depend
on the original fertility of the soil. In a sandy and infertile forest soil in
northern Brazil, Ross et al. (1990) reported losses of 203-386 kg/ha/year
of soil organic matter, containing only 3-12.5 kg/ha/year nitrogen. It is in
the first 6 months after forest clearance, when vegetation cover is at a mini-
mum, that the bulk of soil erosion and nutrient loss occurs. For this reason,
many authors recommend different methods of soil-surface protection
immediately after forest clearance, including the retention of leaf litter on
the forest floor, mulching, brashing, the rapid sowing and establishment of
ground-cover vegetation, or leaving the understorey vegetation as intact as
possible while removing timber.
6.6 DEFORESTATION EFFECTS ON SOIL CHEMICAL
PROPERTIES
Whittaker's (1975) general observation that 'tropical rain forest has a rela-
tively rich nutrient economy (phytomass) perched on a nutrient-poor sub-
strate (soil)' would lead even the inexperienced to suspect that
deforestation combined with timber and vegetation removal, whether by
bulldozer or fire, will result in severe nutrient losses from the whole system.
Failure to produce properly sustainable agricultural systems after forest
removal has been attributed to the infertility of tropical forest soils prior to
clearing, combined with removal of the vegetation in which the bulk of the
system's nutrients were stored. This is certainly not the universal story:
almost the reverse, according to Whitmore (1984) who illustrated the soil
and vegetation compartment distributions of nutrients for several tropical
forests and noted (as did Proctor, 1987) that in many cases a large propor-
tion of the ecosystem nutrients lie in the soil (see also Figure 6.5 below).
This is certainly true for nitrogen and phosphorus, where as much as
81.5% of the ecosystem nitrogen (Bernhard-Reversat et aI., 1978) and
89.1 % of the phosphorus (Poels, 1987) is present in the soil (see Table
6.1). High levels of nitrogen in some tropical soils have been attributed to
the large numbers of leguminous trees and plants in some tropical forest
systems. Since nitrogen and phosphorus are the most critical nutrients for
regeneration and subsequent plant growth, biomass removal alone cannot
be the only reason for infertility after deforestation. Physical processes of
erosion and leaching are major factors in post-clearance soil deterioration.
Apart from the nutrients lost in removed forest biomass, two direct influ-
ences of deforestation on soil chemistry are the removal of regular organic
matter inputs and the effects of fire on plant and soil nutrients. Loss of for-
est canopy results in curtailed or severe reduction of leaf litter influx above
ground, accompanied with the death of tree roots and loss of organic mat-
ter input below ground. Burning results in volatilization of both organic
matter and large proportions of associated plant nutrients.
6.6.1 Effects of forest removal on soil organic matter and plant nutrients
Total annual litter influx in tropical forests ranges from around
3.6-12.6 tlha/year, depending on forest type and geographical location,
with generally higher values on more fertile soils and lower values in mon-
tane forests (Proctor, 1984). These amounts represent very significant regu-
lar inputs of plant nutrients to forest soils, ranging from (kg/ha/year)
nitrogen, 28-224; phosphorus, 1.1-15; potassium, 6.68; calcium, 6.5-290;
magnesium, 1.64. Calculated as the percentage of all aerial nutrient inputs
by Ross et al. (1990) (Table 6.2), it can be seen that litter fall provides as
much as 78.9% of the nitrogen input, 78.4% phosphorus, 33.6% potassium,
Table 6.1 Proportions of nutrients present in the soil compartment of tropical rain forest systems, expressed as a percentage of total
ecosystem nutrients (excludes litter and roots)
Location N(%) P(%) K(%) Ca(%) Mg(%) Reference
Venezuela (San Carlos) 37.6 33.8 10 26.2 7.8 Herrera (1979)

Brazil (Manaus) 56.5 50.6 10.3 0 5.8 Klinge (1976)
Surinam 77.2 89.1 45.3 n.a. n.a. Poels (1987)
Panama n.a. 11.7 11.3 84.3 83.7 Golley et at. (1975)
Columbia 74.5 79.2 n.a. n.a. n.a. FoIster et at (1976)
Puerto Rico n.a. n.a. 15.6 28.3 55.0 Ovington and Olson (1970)
Ivory Coast 81.5 65.5 n.a. n.a. n.a. Bernhard-Reversat et al. (1978)
Ghana 68.5 7.7 n.a. n.a. n.a. Greenland and Kowal (1960)
New Guinea 95.4 24.8 31.7 68.2 71.9 Edwards (1982)
Table 6.2 Proportions of nutrients returned to the soil of tropical rain forests through litter fall, expressed as a percentage of total
aerial inputs (after Ross et at., 1990)
Location N(%) P(%) K(%) Ca(%) Mg(%) Reference
Brazil (Manaus) 72.4 58.3 n.a. 52.8 62.2 Brinkmann (1985)

Venezuela 78.9 61.5 31.1 77.4 62.9 Grimm and Fassbender (1981)
Ivory Coast (Banco) 68.0 78.4 30.1 61.0 55.4 Bernhard-Reversat (1975)
New Guinea 75.9 67.1 28.3 83.3 63.3 Edwards (1982)
N .E. Australia n.a. n.a. 33.6 78.9 46.1 Brasell and Sinclair (1983)
83.3% calcium and 63.3% of the magnesium inputs to tropical forest sys-
tems. It is much less easy to find information on nutrient uptake or deple-
tion during regeneration after forest clearance. Nevertheless, it is clear that
the curtailment of leaf litter influx in clearings represents not only a deficit
to the soil nutrient pool, but also an important loss of organic matter
which, like the residue mulches described by Lal et al. (1980), positively
influences soil physical conditions and promotes good soil structure.
At the time of forest clearance, a total above-ground biomass of around
100 tlha/year, comprising around 10 t/ha/year leaf litter, can be input to the
forest floor (Anderson and Spencer, 1989). As the deforestation leaf litter
component will not have undergone the senescence of normal litter fall, its
labile nutrient content will be higher than regular litter. Turnover rates for
natural litter in tropical forests are generally less than 1 year (see kL values
in Proctor, 1987). Although we might expect the higher nutrient content of
deforestation leaf litter to speed up its decomposition rate, Maheswaran
and Gunatilleke (1988) have shown that the altered microclimate of
cleared areas acts to reduce decay rates. They found that leaf litter decom-
position in a cleared area dominated by ferns in Sri Lanka was only 66.1 %
of that of the adjacent forest.
Observations show that the decay of abandoned tree boles and tree slash
material is also reduced in clearings compared to the more humid environ-
ment of the forest, but quantification is still scarce.
Root biomass estimates for tropical forests range from 11.2-132.2 tlha
(e.g. Vitousek and Sanford, 1986). It is likely that as much as 50-100 tlha
of dead root organic matter will be added to the soil after forest clearance
and will decay in situ. For tropical rain forest in Surinam, Poels (1987) esti-
mated a mean root biomass of 108.5 tlha, comprising (kg/ha) nitrogen,
857; phosphorus, 58; potassium, 380; calcium, 445; and magnesium, 79.
These amounts represented 36, 38, 23, 15 and 25% of the total phytomass
nitrogen, phosphorus, potassium, calcium and magnesium, respectively,
and 6, 14, 15, 10 and 11 % of the total ecosystem nitrogen, phosphorus,
potassium, calcium and magnesium. These significant quantities of organic
matter and nutrients are much less likely to be removed during clearance
operations than is the above-ground phytomass component.
6.6.2 Effects of burning on soil organic matter and plant nutrients

Depending on forest type and the completeness of the burn, as much as
50% of the biomass may remain on site, at least 5% of it remaining as
charcoal which is resistant to further oxidation. Lal and Ghuman (1989)
described windrowing and burning which followed forest clearance in
Nigeria. The windrow contained around 49.4-60 kg/m2 of fuel and
deposited a layer of ash 30-50 cm thick on the soil surface. Alegre et al.
(1988) estimated that 6600-7500 kg/ha of dry matter was returned to the
Deforestation effects on soil chemical properties 131
soil as ash after burning of a secondary evergreen tropical forest at
Yurimaguas in Peru. This material contained (kg/ha) nitrogen, 64; phos-
phorus, 14-19; potassium, 31-89; calcium, 88-131; and magnesium,
33-68. The high pH and high levels of basic cations in ash returned to the
soil surface can elevate soil alkalinity by up to 1.3 pH units (Lal and
Ghuman, 1989). Basic cations in ash deposits at the soil surface can be
rapidly lost by surface run-off and through leaching. Silva (1979) showed
that nitrogen and potassium were the elements most quickly lost after the
burn, presumably because of the mobility of N0 3 and potassium in soil.
Despite this, several authors report elevated soil fertility for as much as 1-2
years after burning. One rainy season after windrow burning, Lal and
Ghuman (1989) found greatly elevated levels of phosphorus, potassium,
calcium and magnesium in the top 30 cm of soil. While the results of
Alegre et al. (1988) support the observation that phosphorus, potassium,
calcium and magnesium remain elevated in topsoils 98 weeks after burn-
ing, these authors also note significant down-profile transport of potas-
sium, calcium and magnesium over the same period.
High temperatures may be reached in burning slash, up to 620 0 C in the
fuel and >100 0 C in the topsoil (Ewel et aI., 1981). Under these conditions,
large but very variable amounts of plant nutrients become volatilized and
are lost from the ecosystem. These nutrients may subsequently contribute
to the nutrient economy of adjacent ecosystems as atmospheric dry fall-out
or rain-out. Salas and FoIster (1976) found that 25 kg C/ha and 673 kg
Nlha were lost in smoke during a tropical rain forest fire in Columbia. This
accounted for 11-16% of the system's carbon and 20% of the total ecosys-
tem nitrogen, but they stress that these are minimum values, probably
accounting mainly for the burning of the soil surface organic matter. Ewel
et al. (1981), working in a wet tropical forest in Costa Rica, recorded
volatile losses (kg/ha) of calcium, 16000; nitrogen, 490; and sulphur, 130.
After the fire, they found that 38% of the system's carbon remained in or
above the top 3 cm of soil, 57% of the nitrogen, 20% of the phosphorus,
only 10% of the potassium, 40% of calcium and magnesium, and 44% of
the sulphur (see Figure 6.2). It is interesting to compare the losses of cal-
cium and nitrogen from the above-ground biomass during the fire with
those losses from the soil surface plus the above-ground biomass. Much
larger amounts of both calcium and nitrogen are lost from the vegetation
component than from the soil organic matter and litter component. This is
probably due to the sharp temperature gradient generated in soil during
the fire. In Thailand, for example, Zinke et al. (1978) found temperatures
of 650 0 C in slash fuel 2 cm above the soil surface, decreasing to 70 0 C at
only 2-3 cm below the soil surface. These results represent significant
retention of nutrients in tropical forest soils, apart from phosphorus and
potassium, which are lost in large amounts during slash-and-burn prac-
tices. Ewel et al. (1981) measured substantial losses of the non-volatile
elements after the burn, through erosion and leaching: 51 % phosphorus,
45% calcium and 40% magnesium (Figure 6.2.).
At the Jari forestry plantation in the Eastern Amazon Basin, Russell
(1989) carried out studies to determine the nutrient loss associated with
selective logging of the native forest, followed by forest burning and plan-
tation establishment. He estimated that around 40% of the ecosystem's
nitrogen was lost, 30% of phosphorus, 75-80% of potassium, and
60-70% of calcium. Only a small proportion of the ecosystem phosphorus
was lost, as the bulk of the system's phosphorus is held in a fixed form in
the soil. These nutrient losses subsequently affected soil fertility of the new
plantations and artificial fertilizing is now practised at Jari, even for less
nutrient-demanding Pinus caribaea.
6.6.3 Changes in soil organic matter and nutrients with time after forest
clearance
Allen (1985) carried out a statistical examination of the effects of forest
clearance on soil properties in both temperate and tropical conditions, and
found that few soil properties returned to pre-clearance levels. She showed
that the main nutrients to decrease after clearance were organic carbon,
total nitrogen and exchangeable magnesium. The only nutrient to return to
original soil concentrations with time after forest clearance was phospho-
rus. Allen's (1985) work highlighted post-clearance differences between
tropical soils developed on 'old', highly weathered parent materials and
those developed on 'new' parent materials. Significantly greater decreases
occurred in organic carbon, total nitrogen and cation-exchange capacity in
response to forest clearance on old parent materials in the tropics. One of
the reasons for this is that the mineral fraction of highly weathered soils is
dominated by clays with low cation-exchange capacity, and the soil organic
fraction provides an important source of ion-exchange sites. These soils
thus respond significantly to organic matter loss. These results suggest that,
even under a similar climatic regime, soil organic matter in andosols, mol-
lisols and some inceptisols may be more resilient than in oxisols or ultisols
under cropping of forestry plantation in the tropics (Anderson and
Flanagan, 1990).
Changes in soil organic matter and nutrients with time after forest clear-
ance have been studied in a range of subsequent land uses, including
forestry plantations (e.g. Russell, 1989), subsistence agriculture (e.g.
Sanchez et aI., 1983), pasture (e.g. Serrao et al., 1979; Hecht, 1981), fallow
(e.g. Jaiyeoba, 1988) and natural forest regeneration (e.g. Uhl et al., 1982).
In virtually all cases, a similar pattern in soil fertility is seen, with peaks in
mineralization immediately after deforestation, followed by a rapid decline
in all soil nutrient levels which may later be followed by a gradual
Deforestation effects on soil chemical properties 133
Carbon Nitrogen
Sulphur
(5300) (219)
(15)
To 3cm
in soil
(a) (14)
(3000)
Above
ground
Phosphorus Calcium Magnesium

(2.2) (87) (17)
To3cm
in soil
(b)
c=J Wood harvest n Decomposition . . Burned
D Remaining
Figure 6.2 Fates of plant nutrients after tropical forest burning. Preburn initial
nutrients shown in (glm 2 ). (a) Carbon, nitrogen and sulphur, when the budget
includes only above-ground biomass or extends to a soil depth of 3 cm.
(b) Phosphorus, calcium and magnesium, when the budget includes all above-
ground biomass plus soil to a depth of 3 cm. (From Ewe! et aI., 1981.)
accretion of all soil nutrients over a period of up to 10 or more years. In

traditional, post-clearance cropping systems, a general decline in soil fertil-
ity with successive seasons of cropping is characteristic (Nye and
Greenland, 1960). The rapidity of the decline varies with soil type and
geographic region, but in a range of examples given by Nye and
Greenland, noticeable deterioration in crop yields were recorded in as little
as 1 year after clearance. Sustainable cropping systems and agroforestry
systems are currently being researched in the Tropical Soil Biology and
Fertility Programme (TSBF) of UNESCO, with the prime objective of
maintaining and improving soil organic matter and nutrients (e.g. Swift
and Sanchez, 1984).
A majority of studies have indicated that it is rare in any type of post-
clearance land use for nutrient levels to rebound to levels present in the
original forest soil. It is likely that this is because most studies are short
term. In Caribbean forest zones classified as wet, moist and dry, Brown and
Lugo (1990) found that the time for recovery of soil calcium levels after
deforestation was around 40-50 years and for soil nitrogen, around 15-20
years. The rate of soil calcium recovery was faster in wet forest zones
where conditions are optimal for fast rates of production of soil organic
matter, while the rate of soil nitrogen recovery was faster in dry forest
zones.
Most deforestation studies have also shown very variable soil nutrient
levels after clearance, particularly after burning, partly due to the hetero-
geneity of ash distribution and composition. The studies of Serrao et al.
(1979) and of Hecht (1981) on changes in soil fertility with time after pas-
ture creation in eastern Amazonia clearly show that different soils respond
to forest clearance in different ways. Generally, the speed of nutrient
decline is faster in clayey tropical oxisols than in loamy oxisols (Serrao et
aI., 1979). This is particularly true for soil organic matter, total nitrogen,
extractable calcium and magnesium and for available phosphorus. The
decline in available phosphorus is related to the high phosphate-fixation
capacity of clayey oxisols, which have high levels of iron oxides.
There is some debate concerning soil-plant-nutrient relationships after
forest clearance. From examination of a number of studies, it seems
unlikely that soil nutrient status could playa major role in tropical succes-
sion recovery unless soil erosion occurs after clearing (Harcombe, 1980).
There are also several studies which indicate that secondary vegetation is
rather good at regenerating soil fertility. Harcombe (1977), for example,
suggests that while nutrients may be lost from topsoils after clearance, they
can accumulate in the subsoil where deep-rooted plants may still be able to
tap them. This 'nutrient pumping' action of regenerating vegetation and
crops has been demonstrated by Nye and Foster (1961) who found that as
much as 30% of a plant's phosphorus requirement could be taken up from
deeper than 25 cm in the soil.
With nearly all land uses after clearance, a large amount of slash mater-
ial, burnt and unburnt, remains on the ground. These woody tissues,
mainly the trunks and branches of trees together with the bark, store as
much as 40-55% of the above-ground biomass nitrogen, potassium, cal-
cium and magnesium (Grubb and Edwards, 1982). The decay of tree boles
Changes in soil conditions with time after deforestation 135
and slash in situ during subsequent land use will release a significant nutri-
ent input to the soil. Buschbacher et al. (1988) estimated that decomposi-
tion of un burnt woody residues may provide as much as 50% of the
nutrients taken up by vegetation during 8 years of recovery following light
pasture after deforestation in Para, eastern Amazonia.
6.7 CHANGES IN SOIL CONDITIONS WITH TIME AFTER

DEFORESTATION
6.7.1 Soil changes with pasture creation
Toledo and Serrao (1982) illustrate a model for soil fertility degradation
following forest clearance and pasture establishment (Figure 6.3). Serrao et
al. (1979) recorded patterns similar to this for phosphorus, potassium, cal-
cium and magnesium in deforested oxisols of southern Para in the eastern
Amazon. The initial sharp increases in soil phosphorus, potassium, calcium
and magnesium on burning, and steep declines in all four elements after 6
months to 2 years, are followed by a levelling-off of exchangeable potas-
sium, calcium and magnesium at levels higher than those prior to clearance
(Figure 6.4) (Serrao et aI., 1979). Hecht (1981) found good comparisons
between her data for southern Para and data of Serrao et al. (1979). Hecht
identified a decrease in available phosphorus with time after pasture estab-
lishment as the main reason for drastic reductions in pasture productivity,
leading to pasture abandonment.
Changes in soil organic matter and total soil nitrogen in the Serrao et al.
study are rather less dramatic than for other nutrients. Initially high above-
ground and below-ground productivity of pasture grasses such as Panicum
maximum soon after clearance and burning has a high organic matter
input which can increase soil organic carbon to levels higher than before
clearance (Hecht, 1981). Soil organic matter decomposition can also
become limited by nitrogen availability, leading to a build-up of organic
carbon. Soil nitrogen levels during this perturbation period are the result of
a dynamic balance between inputs from nitrogen fixation and organic mat-
ter decomposition, with subsequent losses through denitrification, leach-
ing, erosion and plant uptake.
Buschbacher et al. (1988), studying abandoned pastures in the eastern
Amazon, found that age since pasture abandonment had no significant
effect on soil nutrient concentrations, irrespective of the degree of grazing
activity. This was corroborated by Eden et al. (1990) who examined soil
nutrients in young (2-4 years) and old (6-25 years) pastures in the north-
ern Amazon Basin. The majority of ecosystem nutrients in abandoned pas-
tures (Bushchbacher et al., 1988) were found in the soil compartment.
Even the oldest pasture with the lightest grazing activity and highest above-
ground biomass, conditions where we would expect maximum uptake of
'- Seeding of pastures
Clearing
and
burning
!
Fertility level of crops
i --------~=~------j----------------
Pasture with rational management
t
Traditional pasture
Natural fertility level t
+
---Forest---o 5 10
Years after burning
Figure 6.3 Model showing probable trends in soil fertility on changing from for-
est vegetation to pasture (from Toledo and Serrao, 1982).
nutrients by the actively growing vegetation, had 97% of the ecosystem

phosphorus, 89% of the nitrogen, 60% of the magnesium and 48% of the
calcium present in the soil compartment. Potassium was predominantly
found in the biomass (75%) with the slash. Compared to primary forest,
pastures show little nutrient degradation apart from calcium (Figure 6.5)
(Buschbacher et al., 1989). The maintenance of soil fertility over time after
the first few years in pastures established after deforestation depends on
soil type (Hecht, 1981) and very probably on the ability of pasture grasses
to recycle nutrients, particularly phosphorus. Phosphorus deficiency affects
the competitive ability of grasses such as P. maximum, so that forest regen-
eration begins to invade and take over the pasture. This eventually leads to
pasture abandonment.
8 5
l 4
~
~ 3
:r: E
c.
.~ 2 ,,
4 o
(1)
C)
1
-_
'''''....... ......... -,,-----, ......... ' ........ _-
3~+"1--'---~-'--~~~
3 5 7 9 11 13 1 3 5 7 9 11 13
Burning Burning
0.25 E 10
;g ci.
~ 0.20 .9, 8
c: til
Ol :::l
g
C)
0.15 ~ 6
c.
'c til
o
S 0.10 ,, .s::
c.
~
0.05 ~- - ...... - ... -_........... -- -- .... -", :0
Ol
2
,-- ~
~
3 5 7 9 11 13 1 +
O"'----..,.----,r---,---.--r---,---,---,
3 5 7 9 11 13
0.5
Oi E 10
:5.- 0.4 :::l
.~ 8
g c:~
S 0.3
C)C)
~:5 6
~ 0.2 ".-c:c- 4
(1)(1)
~O~ 0.1
ES
'5iii 2
a.
o O"'----.-r--.-,--,---,---.-.
1 3+ 5 7 9 11 13
Age of pasture sampled (years)
Figure 6.4 Changes in topsoil properties of Panicum maximum pastures of

known age in eastern Amazonia (Cochrane and Sanchez (1982) from the data of
Serrao et aI., 1979). - Clayey oxisol, Paragominas, Para; ---loamy oxisol, north-
ern Mato Grosso.
6.7.2 Soil changes with slash and bum, followed by cropping

Traditional slash-and-burn agriculture differs little throughout the tropics.
Four general stages are: hand cutting of underbrush; cutting of trees at
waist height followed by branch removal to promote drying; after a period
of drying, the fuel is burnt; followed fairly soon by direct planting. Seubert
et al. (1977) found that bulldozer clearance was more detrimental to soil
2000
1500
Ca K Mg N/5 Px5
r/)
r/)
co
E
o
:.0 1000
Ol
c:
.s;
:.:::i 500
(E x tractable)
500
'0
(J)
1000
• Mature forest
2000 D Low-intensity-use pasture

~ Moderate-Intenslty-use pasture
till! High -intensity- use pasture
Figure 6.5 Stocks of major nutrient elements in the living biomass and soil of
mature forest and abandoned pasture, following three intensities of use (from
Buschbacher et al., 1989).
properties than traditional slash-and-burn techniques (Figure 6.6).

Exchangeable bases and soil pH were all higher in slash-and-burn plots, as
were soil organic carbon, total nitrogen and available phosphorus. Perhaps
the main advantage of slash and burn is the initial sharp increase in avail-
able phosphorus, although this drops rapidly after about 6 months.
For the maintenance of soil fertility for subsistence cropping after defor-
estation, many authors have stressed the need to maintain stocks of soil
organic matter. Only a fraction of the fresh organic matter produced by
new vegetation, whether pasture grasses, crops or weeds, will become
incorporated into the soil humus. A significant proportion will be lost by
exposure to wind and water through leaching and erosion. Young (1976)
claims that as much as 30-60% of tropical soil organic matter typical of
0.15, ~ 17 0.35
E
a.
~ 1.4 ~
:.e 'j: oS
~ II)
c: ~ 0.13 2 13
8
CD 0
I i 0.25
~ Cl &.
e a.
II) §.
~ 1.2 0 P ....
0 ~
'2
,, -a. : ..... , ............ ' , E ,",,
0
111
~
~ 0.11
F
"I ,,
,,
CD
~
9 .~ 0.15
!Q
,' , ,
' " ,,
1.0 0
,, " a.. '----
0.09 , " " ! 5 0.05 ---
6 10 6 10 6 10 6 10
4.6. 7. 3.5
I
::=- 1.0 ~ 0.5
E
8 0.4 4.4
~'- E .. '
i ------ ., 8 3.1
~
CD E ..- ... ' " "
§. 0.6 - :I: ..... "
E a.
:::l
.. lE ,"
E , ...... , ' 'iii I
:::l CD 0.2 I ,
c: 4.2 U 2.7
~ ........ ------------ w
()
,,
~
() 0.2 ::E
I I
4.0 2.3
6 10 6 10 1 6 10 6 10
Months after clearing
Figure 6.6 Effect of land-clearing methods on topsoil (O-tO em) chemical properties in the Peruvian Amazon at Yurimaguas, measured at
1,6 and 10 months after clearing (from Seubert et at., 1977). - Burned; --- bulldozed. Error bar, LSD at P=O.OS.
natural vegetation can be lost under continuous cultivation. These effects
can be reduced with judicious management practices. In an early study of
the effect of shade on the maintenance of soil organic matter, Cunningham
(1966) recorded an important improved retention of organic carbon in
soils which were artificially shaded. He explained the results in terms of
reduced organic matter decomposition rates under shade.
Sanchez et al. (1983) found slight decreases in soil organic carbon and
total nitrogen in the first 8 years of rotational cropping with rice, corn and
soya bean/peanut, both with and without liming and fertilizing. Fertilized
plots, however, consistently showed higher pH (0.25-1.0 units) and higher
levels of extractable calcium, magnesium, potassium and phosphorus.
Sanchez et al. (1983) assessed these cropped soils at Yurimaguas in Peru to
be deficient in all the major nutrients and marginally deficient in several of
the micronutrients. A balance sheet of crop uptake in response to fertilizer
additions for two rotations (8 years) showed that fertilizer additions
resulted in accumulation of phosphorus, calcium, zinc and copper in the
topsoil. Nitrogen was taken up in amounts exceeding fertilizer inputs, the
deficit probably being made up from symbiotic nitrogen fixation by soy-
abeans. Continuous cropping without complete fertilization in the same
study resulted in an incomplete vegetation canopy, soil compaction and
erOSIOn.
Experiments with organic matter inputs and mulches have shown that it
is possible to arrest soil degradation. Ayanaba and Jenkinson (1990) found
that organic carbon and total soil nitrogen declined rapidly 2 years after
forest clearance on andosols in Nigeria. When low-quality maize residue
was retained on site, the decline was slower and organic matter increased
in some cases.
Several types of slash-and-burn agriculture involve forest clearance, fol-
lowed by cropping for a few years, followed by a fallow spell to allow
regeneration of fertility, followed by further burning, cropping and fallow
cycles. Such is the jhum system in India. Different lengths of fallow regen-
eration were studied by Toky and Ramakrishnan (1983b). They found that
the annual rate of return of nutrients in litter fall increased with age of the
fallow (up to 20 years), with highest returns for nitrogen, followed by
potassium and calcium (Table 6.3). Stocks of soil nutrients decline sharply
when the jhum cycle is shortened to 5 years (Figure 6.7) (Ramakrishnan
and Toky, 1981), illustrating depletion of the system during cropping
phases, followed by insufficient time for nutrients to accumulate during the
fallow.
Saldarriaga (1989) examined the long-term recovery in ecosystem struc-
ture, biomass and nutrients after shifting cultivation in the Rio Negro
region of western Amazonia. Over a period of 80 years, nitrogen, phos-
phorus, potassium and calcium accumulated in soil, litter, stem and leaf
compartments of the system. Over this relatively long period, all soil
Table 6.3 Percentage of total ecosystem nutrients present in the soil compartment
of successional fallow, following slash-and-burn agriculture in north-east India
(calculated from the data of Toky and Ramakrishnan, 1983b).
Age of fallow (years)

Nutrient 1 5 10 15 20
N 99.7 98.7 98.3 96.7 95.4

P 50.0 33.3 40.0 33.3 33.3
K 96.1 75.0 32.1 54.6 46.3
Ca 98.5 90.9 86.1 79.0 70.6
Mg 98.5 93.9 87.8 88.2 84.1
nutrient stocks were similar to those in the soil of adjacent mature forest
(Figure 6.8),
6.7.3 Soil changes with plantation forestry

Although soil and nutrient cycling data for tropical forest plantations are
few, planting with Pinus caribaea, Eucalyptus deglupta and Gmelina
arborea has met with varying degrees of success in many different parts of
the tropics. The sequence of soil and vegetation nutrient losses after forest
clearance and plantation establishment at Jari in eastern Amazonia was
studied by Russell (1989).
Patterns of nitrogen, phosphorus, potassium and calcium in different
ecosystem compartments at Jari before and after clearance and planting
with P. caribaea are illustrated in Figure 6.9. Leaching caused an important
loss of nutrients only in the first few years of plantation growth. Nutrient
stocks in the soil litter layer were depleted at burning, but built up again
under the pine canopy. In the overall ecosystem, only phosphorus accumu-
lates to pre-clearance levels. All other elements measured, particularly
nitrogen, fail to reach levels in primary forest. Only a small portion of soil
nitrogen is assimilated into pine biomass. For subsequent rotations of tree
plantation, potassium may be the limiting element as removal of the above-
ground biomass will remove around 87% of the ecosystem's potassium
stocks.
6.7.4 Soil changes with fallow and natural forest regeneration

Most regeneration to secondary forest occurs after abandonment of other
land uses, such as subsistence cropping or pasture. Thus a fallow period
ensues, during which time woody species invade. In some quantitative
studies of soil nutrients in fallows, a decline in fertility is seen in the early
stages. Other studies indicate gradual accretion of soil nutrients over time.
1.400 4.5
1.200 4,0
2 3.5
-ao 3.0
~ N 2.5
;:9
.c E
2.0
]I 1.5
'iii
~ 1.0
0.5
Sampling lime (days)
30 10 5
Cycle (years)
i
Dept,.. (em) within 40 em .oR column
~~4 480
14-28 420
28-40
380
480
420
360
E 300
SoE
::l~
'0 240
~~ 180
120
60
o
Samping lime (days)
30 10 5
Cycle (years)
Figure 6.7 Changes in total quantity of soil nutrients after burn and cropping
during jhum (fallow) cycles of 5, 10 and 30 years (from Ramakrishnan and Toky,
1981).
In studies of differently aged fallows in Nigeria, Aweto (1981) and

Jaiyeoba (1988) found a gradual build-up of all soil nutrients in topsoils
over a 14-year sequence (Figure 6.10). In 10 years, the tree density in the
fallows increased from 56/ha to 2670/ha. Tree height and breast-height
diameter (dbh) over the same period increased from 1.3-5.8 m and from
0.7-7.0 cm, respectively, as compared to means for mature forest of
lOA m tree height and 20.0 cm dbh (Aweto, 1981). Soil fertility after 10
years of fallow more closely approximates forest soil conditions than
fallow tree structure approximates forest tree structure. Aweto (1981)
3,000
60
50
~«;-2'000i'~. reb.. :::J o
o.r; ~ «;- 40
~~ . ~.r;
...- 0.-
z- :g ~ 30
.r;-
1,000 a..
10 20 30 40 50 60 70 80
Age (years) Age (years)
400
300
«;-
.~0.r;
~ E-
:::I! ....
0_ :go,
~~2oo tU,:,e.
0- Q-
a..
1 ...
~'9IUJ"U IIJIU 111'1 1UU.IJUUI IJbLHd • 5oI;;
01
.•. ·(. . :. .}),. ;. ....... .!J.!,·. .!:,:,~-·-~·<··,>,<·,:·-..:;~~·,>;:=A;·:·<: ',"'",0:':..
~J'~!lI!111111111111·-iii!lJ~,1IWJ~)!,~'~.~!I{~L,,·
IFf r Matlft
10 20 3b 40 50 60 70 80 Maue
Age (years) Age (years)
Figure 6.8 Total stocks of nutrients in different ecosystem compartments of succession forest as a function of age
since clear felled, compared to mature forest in western Amazonia (from Saldarriaga, 1989).
s .....O & plant
i'( ..~ \ Decomposng slash

~- 1
_
<U
10000
f
Losses 10 harvest.
PAIMARY : vcMatilization , ~ 1- losses '0 ha.yes'
E
~
FOREST ~aching & . Harvest & burn
CIenItnhcatl()f) I
C . .
1 ••••• } PO,. ~ash
g>
Q) 5.000
,'.Q . LOsse:s to harvesl
2
- PINE '''~.l ~~I&
denitnflCatlon
~""""'==Q~~~"" ; li1Ter
SOIL SOl'
o '0 '0
Age of plantat.on (years)
&In & plant

'.500 i'(
cr/ ~ ~..I
Blsn & plant
Harvest & burn
1,000 ,000 ....

>Losses 10 harvest
:w 01 FOREST
I1
J~ .I~-
c;, I
=. ~~ts~y [' Losses to harvest .. ..0 Decomposng slash
Harvest & tun ,losses to harvest
E
•
:?I
:J 500 o )
'iii ) .... leachng Iosse,
<II leac~ losses
} Net .. losses to harvest
.!!! O~onr'sng PINE } Net leactWlg lOSses
0........ }Pi1e slash
0
c..
SOft. o.~i. !~erslaSh ~~~~~~~~~- /l.''''
.0 '0
Figure 6.9 Total stocks of nutrients in different ecosystem compartments after

clear felling and planting with Pinus caribaea at Jari, eastern Amazon (from
Russell, 1989) . • Measured; 0 estimated; --- projected.
calculated a soil fertility index for fallows, using five variables: soil organic
matter, total nitrogen, cation-exchange capacity, available phosphorus, and
water-holding capacity. The mean value of each for the fallow was
expressed as a percentage of that in mature forest. The mean value of this
index for 1-, 3-, 7- and 10-year fallows was 48, 42, 65 and 71 %, respec-
tively, illustrating a sequence of fertility restoration. A similar pattern was
reported for a sequence of forest fallows in north-east India, where both
soil organic carbon and total nitrogen gradually increased over a 50-year
period (Ramakrishnan and Toky, 1981). A different pattern of soil nutrient
status is apparent for forest fallows in South-East Asia. Nakano and
Syahbuddin (1989) report generally decreasing soil fertility with time dur-
ing the fallow, with soil organic carbon, total nitrogen and cation-exchange
capacity reaching lowest values after 3-4 years. These results may be more
typical of fallows initiated on already degraded soils.
Compared to primary forest, regenerating clearings in southern
Venezuela showed higher leaching rates of all nutrients except nitrogen
(Uhl et aI., 1982), irrespective of whether the clearance technique had been
cutting, or cutting followed by burning. Three years after clearance, both
cut and cut-and-burn plots showed enhanced levels of calcium, potassium,
6
0~--T2---'4----6r---T8---'10----lr2---1T4--/Mature
forest
Fallow age (years)
0.6
~
- 0.5
i
-'2
~
'c
0.4
0.3
iii
i2 0.2
0.1
OL---T2---'4---'6r---TS---'10---'12r---1T4--(Mature
forest
Fallow age (years) /
~0.6
/
( 60;
o
.,.... o.,....
gO.S ?--......
..... ~
/ sg
-------,.......
/
g 1"/ / g
E 0.4 / 4 01
..... ::!:
"...
::J
'0 "0
// ,
:3 0.3
/
3 ai
& as
o
0.2 ,~~"" 2 ~
~ ., ... ------..... -~~-~~"""~!!!
Q)
"'II' ,,' \II
~ 0.1 .. ____ ........ 1 ~

~ ~
~ ~--~--~--_r--~--_,----r_--~I ~
w 0 2 4 6 S 10 12 14 Mature 0 w
forest
Fallow age (years)
Figure 6.10 Changes in soil nutrients in forest fallows compared to mature forest
with time after clear felling in Nigeria (from the data of .. Aweto, 1981 and.
Jaiyeoba, 1988). -- calcium; ----- magnesium; - potassium.
magnesium, nitrogen and phosphorus, while bulldozed plots still showed
severely reduced levels (Table 6.4). The removal of topsoil by bulldozing
created site heterogeneity, resulting in rapid woody regeneration of the
debris mounds to the detriment of the plots whose topsoil seed banks had
been removed.
6.8 DEFORESTATION EFFECTS ON SOIL BIOLOGY

Over the past decade great strides have been made in understanding how
soil biology influences soil fertility and nutrient cycling. Accordingly, there
is a rapidly accumulating literature on the effects of tropical deforestation
on soil organisms and how this affects soil-nutrient dynamics. Altered soil
microclimate, particularly increased ranges of soil temperatures and
decreased relative humidity, is detrimental to soil organisms which require
moisture for processes such as movement, respiration, digestion and repro-
duction. Compaction of soil by mechanized clearance decreases root pene-
tration and microbial activity, either through physical impedance or
through impaired aeration. Fire and the use of pesticides in forest clearance
sterilize the topsoil for significant periods after their use.
The high biological diversity of tropical ecosystems results in more eco-
logical functions and greater ecological diversity. Heterogeneity of physical
and chemical qualities of leaf litter results in variations in decomposition
rates and nutrient release. Diversity of soil decomposer organisms increases
the variation in timing of nutrient release. Lavelle (1987) uses these logical
arguments to point out that the greater the diversity of plants and decom-
posers, the more diverse is the timing of nutrient release. In diverse systems
there is less risk of an imbalance between nutrient production and nutrient
uptake by plants, hence minimizing losses by leaching. Where system diver-
sity is reduced, as occurs immediately after deforestation and in succeeding
regeneration phases, nutrient losses are more likely to occur through asyn-
chronous nutrient release and uptake.
There are a few direct studies of plant roots and of soil microorganisms
after deforestation and there is also increasing work on soil microbial bio-
mass. A more sizeable literature exists on bacterially mediated soil
processes, but concentrating mainly on the nitrogen cycle and including
mineralization, nitrification, denitrification and nitrogen fixation.
Speculative suggestions on the role of mycorrhizae in nutrient cycling and
in regeneration processes after forest clearance simply whet the appetite for
field verification.
6.8.1 Damage to root mat and new root growth

There are many anecdotal reports of damage to the surface root mat of
tropical forests during clearance (e.g. Seubert et al., 1977) and implications
Deforestation effects on soil biology 147
Table 6.4 Soil characteristics of cut, cut-and-burn and bulldozed sites after 3
years of regeneration in southern Venezuela (from Uhl et al., 1982)
Forest clearance technique

Soil Primary Cut Cut and
characteristic forest burned Bulldozed
Percentage soil
organic matter 4.5 4.4 3.2 0.1
pH (H 2O) 3.8 3.4 4.0 4.8
Total N (mg/100 g) 145.0 546.9 384.8 21.4
Total P (mg/100 g) 4.3 22.9 16.9 0.2
Exchangeable bases
(mg/100 g)
Ca 24.0 33.7 42.0 1.7
Mg 2.8 24.3 17.2 0.3
K 12.9 18.4 14.7 0.4
that this effect will severely inhibit subsequent nutrient cycling. There are
rather differing quantitative results of fine root biomass and regrowth after
deforestation. Raich (1980), at La Selva in Costa Rica, found that the total
fine root biomass in l-year-old forest regrowth to a depth of 50 cm was
92% of that in mature forest. However, in the Darien of Panama, Ewel
(1971) found that it took more than 4 years of forest regeneration after
clearance before the total root biomass increased significantly. As the fine
roots are responsible for active nutrient absorption, Raich (1980) sug-
gested that forest regeneration after only 1 year may have a nutrient-
absorption capacity similar to that of mature forest, and thus an ability to
limit nutrient losses through leaching. A different result was found in forest
gaps caused by hurricane damage in Puerto Rico (Silver and Vogt, 1993)
where the fine root biomass was reduced by up to 77%. As 48-65% of the
dead fine roots were still present after 1 year, these authors attributed lim-
ited nutrient losses in forest gaps to slow decomposition and nutrient
release from the dead roots. This slow nutrient turnover from the below-
ground phytomass after disturbance may indicate a nutrient conservation
mechanism until the new fine root biomass re-establishes.
6.8.2 Soil microorganisms and microbially mediated processes

There are few quantitative studies of the responses of different soil organ-
ism groups to forest clearance. The influence of fire was examined by
Meiklejohn (1955), who reported substantial reduction or complete elimin-
ation of a range of soil microflora after burning of soil under Acacia in
Kenya. During the first 3 months after the burn, the main soil organism
groups affected were the fungi (Actinomycetes), aerobic nitrogen-fixing
bacteria and nitrifying bacteria. Exposure and desiccation are the major
threats to soil organism survival after canopy removal. Collins (1980) dis-
covered that the main types of termite to survive after logging and clear
felling in Sarawak are those which maintain fungus combs inside their
nests which help them to feed on fresh litter. Arid post-clearance conditions
adversely affect decomposer fungi and bacteria which require damp, humid
conditions. This has a knock-on harmful effect down the saprophytic food
chain, as other soil organisms depend on fungal and bacterial decay prod-
ucts as a food source.
Studies of the whole soil organism population after forest clearance have
focused on measurements of respiration or microbial biomass. The activity
of the total soil microbial population is usually assessed using soil respira-
tion indices (evolved carbon dioxide or oxygen uptake), but these measure-
ments can produce variable results and may include root respiration or
physicochemical gas exchange. Determination of the amount of carbon in
the microbial biomass using a fumigation technique provides an alternative
method of assessment. As microbial biomass represents an important labile
pool of soil nutrients and may play an important role in preventing nutri-
ent losses, decline in microbial biomass numbers may provide an early indi-
cation of slower, less easily detected changes in more resistant soil organic
matter.
In the same way that forest clearance harmfully affects root growth,
through compaction and impaired aeration, the soil microbial population
is also reduced. Ayanaba et al. (1976) found decreased soil microbial bio-
mass carbon after forest clearance and cultivation, but the decrease was
less when crop residues had been added as a mulch. The mulch organic
matter not only helps to stabilize topsoil structure and hence maintain
good aeration, but also provides another source of available energy in the
form of decomposable organic matter. After slash-and-burn clearance in
central Amazonia, Luizao et al. (1992) found lower soil microbial biomass
carbon in the top 5 cm of slash-and-burn areas (829 f.1g carbon/g), com-
pared to topsoils under adjacent pasture (1290 f.1g carbon/g) and forest
(1287 f.1g carbon/g). In the burnt areas, microbial biomass carbon was a
little higher at depth (933 f.1g carbon/g), while values under pasture and
forest declined with depth. This pattern of microbial biomass may repre-
sent an early stage in topsoil recovery after surface sterilization during the
fire.
Following deforestation and conversion to savanna and pasture in India,
microbial biomass carbon was reduced by up to 57-59% (Srivastava and
Singh, 1991; Basu and Behera, 1993). In Costa Rica, Henrot and
Robertson (1994) found similar patterns of microbial biomass carbon
decline after forest removal. Rapid decline occurred in the first 6 months
after forest clearance, eventually stabilizing at around 35% of its initial
value. Examining the effects of pasture age after forest clearance, Bonde et
al. (1992), working in the same area as Luizao et al. (1992), found a
decrease in microbial biomass carbon with time, so that 8-year-old pasture
had only about half of that of 2-year-old pasture. An adjacent 8-year-old
legume pasture maintained a higher microbial biomass carbon than did the
2-year-old pasture, presumably due to higher nitrogen availability assisting
decomposition processes and allowing the maintenance of a larger decom-
poser population.
Generally high nitrogen mineralization and low phosphorus mineraliza-
tion rates are characteristic of lowland wet tropical forests (Vitousek and
Denslow, 1986). Vitousek (1984) suggests that phosphorus most com-
monly limits productivity in late-succession tropical forests. Robertson's
(1984) suggestion that losses of mineral nitrogen through leaching and
denitrification after clearance may make lack of available nitrogen an
important limiting factor in early stages of succession, is considered below.
Rates of nitrogen mineralization, nitrification, leaching and denitrification
after deforestation are the key processes controlling nitrogen availability at
this critical time, but there are very limited quantitative data on these
processes after tropical deforestation.
Forest clearance in temperate regions has been shown to increase nitro-
gen mineralization rates (e.g. Vitousek and Matson, 1985) and also to
increase the losses of nitrate through leaching to groundwater and through
run-off to streams (Bormann and Likens, 1979). After tropical deforesta-
tion, mineralization rates for soil nitrogen and phosphorus are also
increased and have been found to depend on soil organic matter content.
Mueller-Harvey et al. (1989), working on cleared plots in Nigeria, found
significant increases in nitrogen mineralization to be related to increasing
amounts of organic carbon present in the soil, with maximum nitrification
rates occurring during April at the onset of the rainy season, in plots with
highest organic carbon contents (1.60-2.67%). After tropical deforestation
a flush of nitrogen mineralization might be expected due to an input pulse
of biomass nitrogen. This pattern was observed by Matson et al. (1987),
working in Costa Rica, who found that nitrogen mineralization, nitrifica-
tion and amount of ammonium and nitrate in the topsoil increased sub-
stantially during the first 6 months following burning, then returned to
background levels. Rapid vegetation regrowth after clearance, with high
nitrogen-uptake capacity, may be one reason why elevated levels of mineral
nitrogen are not seen longer in these disturbed soils.
Regenerating vegetation thus acts to conserve nitrogen and to retain it
on site. Robertson's (1984) report of lowest rates of nitrogen mineraliza-
tion and nitrification in most recently cleared sites, with higher ammonium
and nitrate production over time in an 8-year succession (also in Costa
Rica) is rather anomalous and may explain his suggestion of nitrogen
deficiency in early successional stages. Robertson and Sollins (1987) later
report more expected results from cleared secondary forest in Costa Rica,
with much higher soil nitrate concentrations and nitrification potentials
during the first 60 weeks after clearance. Over the longer term, Robertson's
(1984) work indicated a general pattern of change in rate of soil nitrifica-
tion after forest clearance, with a steady increase in the process during the
first 15-20 years, followed by a levelling off in later successional stages at a
nitrification rate of around 3-4 J.Lg NO/g/day. The flush of decomposable,
high carbon: nitrogen ratio organic matter at clearing and the concurrent
pulse of soil microbial activity immobilize nutrients, including mineralized
ammonium. Nitrifiers are able to obtain ammonium for nitrification only
when decomposable organic substrates dwindle, carbon: nitrogen ratios
decline and microbial biomass nitrogen is returned to the soil. Nitrifiers
then have to compete with plant root uptake for the available ammonium
nitrogen.
Both time elapsed since clearance, and the season in which mineraliza-
tion and nitrification measurements are taken, may be important in deter-
mining rates of these processes. Marrs et al. (1991), for example, working
on newly cleared forest gaps in the northern Amazon basin, found much
higher mineralization and nitrification rates in undisturbed forest com-
pared to felled gaps at the start of the study, during the dry-to-wet season
transition period. During the wet season, mineralization and nitrification
rates were slightly higher in felled gaps than in the forest, and during the
dry season mineralization and nitrification were low in both habitats. A
peak in mineralization and nitrification processes in the early wet season
may be due to seasonal microclimate conditions in open, cleared ground
when high temperature variation and wetting/drying cycles are optimal for
decomposition and ammonium production.
We would expect environmental conditions in primary rain forest, with
regular organic matter influxes and high humidity conditions, to be ideal
for denitrification (the reduction of soil nitrate by facultative anaerobic
bacteria to evolve nitrous oxide and nitrogen gases). McElroy and Wofsy
(1986) concluded that tropical forests were the major natural source of
nitrous oxide globally, contributing 7.4 Tg (7.4X10 9 kg) of N 2 0 annually
(excluding evolution during forest burning). Around three quarters of
background stratospheric N 2 0 flux could be attributed to tropical eco-
systems.
The processes governing nitrate production in soil also affect denitrifica-
tion rates: a ready supply of nitrate is a prerequisite for denitrification to
occur. Robertson and Tiedje (1984) found that N0 3 production in soil was
one of the best predictors of denitrification nitrous oxide evolution in old-
succession deciduous forests in Michigan. Despite high soil nitrate levels in
newly cleared plots in Costa Rican rain forest, Matson et al. (1987) found
denitrification rates to be low and unresponsive to further nitrogen inputs.
This indicates that the cleared Costa Rican plots were not nitrate-limited
and denitrification must have been controlled by carbon availability. In
experimental sites in central Amazonia, Luizao et al. (1989) found that
mature forest and recently cleared and burned sites on a yellow clay latosol
did not differ greatly in nitrous oxide evolution. A nearby pasture site on
the same soil showed nitrous oxide fluxes as much as five times higher than
in the forest, particularly during February to June, which is the wet season
in the Manaus region of Amazonia. The authors identify three possible
mechanisms for elevated denitrification in pastures established after forest
clearance: (i) soil compaction due to forest clearance and cattle trampling,
which elevates soil wetness at a time when ambient temperatures are high-
est; (ii) regular oxidizable carbon inputs to the soil from grass roots; and
(iii) fertilization.
The significance of these bacterially mediated denitrification processes in
cleared areas established as pasture should not be underestimated.
Quantitative studies in shifting agriculture and subsistence farming are
urgently required. The conversion of forests to pastures or agricultural
land transforms as much as 11 million ha every year (WRI, 1985). This
amounts to a very substantial potential increase in the flux of N 2 0 (a so-
called 'greenhouse' gas) to the atmosphere, with associated catalytic effects
on ozone breakdown.
6.8.3 Disruption of interacting biological relationships

Two important symbiotic relationships contribute to 'tight' nutrient cycling
in tropical forest soils: nitrogen-fixing bacteria and mycorrhizal fungi.
Nitrogen fixation occurs in soil through both symbiotic and non-symbiotic
bacteria. The predominance of leguminous plants in many tropical forests
and in clearance and fallow successions has prompted many authors to
speculate that a substantial proportion of the nitrogen accumulation in
these systems after clearance is due to biological fixation processes. A small
proportion (around 12.5%, according to Jordan et aI., 1983) of total rain
forest nitrogen fixation occurs in epiphylls such as algae, lichens, liverworts
and bacteria on leaf and bark surfaces in the rain forest canopy, so these
inputs of nitrogen to the system are lost with deforestation.
In West Africa, Robertson and Rosswall (1986) found much higher
nitrogen fixation rates in early rain forest succession than in late succession
or in mature forest. They reported rates of 100, 60 and 15 kg
nitrogen/ha/year for early succession, late succession and mature forest,
respectively. On the other hand, in the nutrient-poor forests of San Carlos
in southern Venezuela, Jordan et al. (1983) and Jordan (1989) found
extremely low nitrogen fixation rates in terra firme forest (16 kg
nitrogen/ha/year) and caatinga >35 kg nitrogen/ha/year). Phosphorus
deficiency may be a major reason for inhibited nitrogen fixation in some
soils (Cole and Heil, 1981), such as the later stages of tropical forestry suc-
cessions after felling.
Mycorrhizae are known to be ubiquitous on tropical trees, with the
majority of species forming vesicular-arbuscular mycorrhizae (VAM) and
an important minority, such as those in nutrient-stressed caatinga systems,
forming ectomycorrhizas (EM) (Singer and Silva Araujo, 1979; Alexander,
1989).
Root biomass differences between topsoils and subsoils under many low-
land tropical wet forests tend to result in higher mycorrhizal infection in
the soil surface mat, although deeper roots can also be infected. Thus any
clearance mechanism which effectively removes or severely damages the
root mat, such as bulldozing or burning, will damage or even eliminate the
mycorrhizae. Most VAM infection in tropical forests is transmitted via
root-root contact (Janos, 1983). With clear felling of vast areas, reinfection
of cleared soil could take much longer than for higher plants whose seeds
may be transmitted by wind or animals, as it may have to rely on spread
from remaining VAM fragments. Tommerup and Abbot (1981) have
shown that VAM hyphae can survive in dead root fragments for at least
6-9 months. VAM survival is thus likely to be higher after clear felling
than after slash and burn. Tropical VAM have unrestricted host ranges
(Janos, 1980), so contact with isolated individuals of particular species is
not required and spread of infection may be quite rapid once a good root
network has been initiated on cleared sites. Tropical species which form
EM associations may be good pioneers and colonizers of cleared sites
because their fungal associates have wind-dispersed spores which allow
earlier colonization than VAM hyphae (Janos, 1980). Many pioneer species
on nutrient-poor sites are obligately mycorrhizal, while facultatively myc-
orrhizal species are most common in fertile soils (Janos, 1980). Herrera et
al. (1991) offer evidence from tropical forest and forest clearings in Cuba
which indicates that the incidence of VAM infection in pioneer species of
clearings, such as heliophilous plants, is higher when these plants are
grown in high isolation conditions than when they are grown in shade,
where photosynthesis rates are low. This type of response illustrates how
rapid symbiotic relationships may be achieved in early regeneration stages
of forest recovery. Janos (1980) illustrated a hypothetical colonization pat-
tern by tropical, YAM-dependent species after abandonment of shifting
cultivation (Fig 6.11). These post-disturbance theories raise many exciting
research questions and call for urgent verification.
The role of mycorrhizae in tropical plant nutrition and nutrient cycling is
still in its infancy. A sizeable body of temperate research suggests that myc-
orrhizae exploit labile soil phosphorus more efficiently than do uninfected
roots, as they possess a mycelium of hyphae with a large surface area
which proliferates a large volume of soil. A second theory is that there are
iron phosphates in tropical soils which are chemically more available to
Deforestation effects on vegetation 15 3
100
(a)
80
(b)
..... ,~--
--. -------- ----------------- ss
............
V
,,
,/ - - - - - - - - _____
(c) _
20
,/ "
"',
',~~--------------------------~~----
O~--r_----------------._----------------,_~O
10 100
Age since disturbance (years)
Figure 6.11 Predicted changes in relative abundances of plants that depend on

vesicular-arbuscular mycorrhizae (VAM) to different extents, after forest distur-
bance. (a) Obligately mycotrophic plants; (b) VAM fungus; (c) facultatively
mycotrophic plants; (d) non-mycorrhizal plants. Proportions on a scale of 0 =
absent to SS = steady state. (From Janos, 1980.)
mycorrhizal roots than to uninfected roots (e.g. Bolan et aI., 1984). A final
suggestion is that the proliferation of mycorrhizae in organic substances
may indicate their ability to use organic phosphorus or organic forms of
other nutrients such as nitrogen (e.g. St John et al., 1983). The work of
Herrera et al. (1978) on an oxisol at San Carlos in southern Venezuela
appears to provide some evidence to support this. They suggested that
mycorrhizae may provide a direct pathway for the movement of nutrients
from decomposing litter to roots by showing that 32p could be transferred
from leaf litter to ectomycorrhizal roots isolated in petri dishes on the for-
est floor. The primary criticisms of this proposal are that one cannot be
sure that the fungus was responsible for release of isotope from the litter,
or that hyphae transported it to the root. Despite these criticisms, the work
has sparked much interest in the nutrient-conserving potential of tropical
mycorrhizal systems in mature forests after clear felling and burning, and
in regeneration successions.
6.9 DEFORESTATION EFFECTS ON VEGETATION

As far as the ability of vegetation to recolonize cleared sites is concerned,
there are two critical effects of deforestation on in situ ecological conditions
caused by canopy removal: the loss of seed influx, and the loss of organic
matter and nutrient influx in leaf litter. Loss of organic matter and nutrient
inputs have been considered earlier. Immediately after deforestation, the
loss of seed influx places reliance on the native soil seed bank for initial
germination. Three other important revegetation strategies are: advance
regeneration (undamaged seedlings which had existed under the original
forest); recruitment of seeds from outside the clearing; and sprouting (or
coppicing) forms of vegetative growth from cut stumps or damaged roots.
Uhl et al. (1989) charted the differing importance of these processes in
undisturbed forests and in natural and anthropogenic gaps and clearings
(Figure 6.12). The regeneration mechanism most vulnerable to clear felling
and burning is advance regeneration.
Much tropical forest regeneration research, so-called gap-phase dynam-
ics, has been carried out on naturally occurring gaps caused by individual
treefalls, lightning, landslides, hurricanes and other natural hazards (e.g.
Denslow, 1980; Brokaw, 1985a). Man-made clearances are usually differ-
ent because burning and cultivation result in soil degradation and loss of
forest-tree seed bank, causing differences in forest succession after aban-
donment. Nevertheless, gap-phase research has produced many results
which are highly relevant for ecologists who study deforestation effects,
including data on environmental controls on seed dormancy and germina-
tion, seed recruitment, nutrient mineralization patterns and forest succes-
sion. A whole research field has opened up, examining light requirements
of different tropical woody species. Ecologists classify tropical tree species
into shade tolerants or light demanders, with wide varieties of preferences
in between, related to growth strategy (Oldeman and Van Dijk, 1991).
Tropical forest clearance successions are dominated by fast-growing, light-
demanding pioneers such as Cecropia. Primary forest trees tend to be slow-
growing shade tolerants which are uncommon in the open, light
environment of early successional stages. Oberauer and Strain (1984) show
that species adapted to differing light conditions actually have different
growth rates. Light-tolerant plants in clearings had a photosynthesis rate
which was four times higher than that of shade-tolerant plants. This extra
productivity capacity of pioneers possibly accounts for their early regenera-
tion success.
Comparisons of photosynthesis and growth of light- and shade-tolerant
tree species in the neotropics (King, 1994) and in dipterocarps in Sabah
(Zipperlen and Press, 1996) have shown how shade-tolerant plants maxi-
mize survivorship and net carbon assimilation under low light regimes.
Light-demanding plants were out-competed in the shade, but had a greater
ability to acclimatize to high light intensity and had greater assimilation
rates at high instantaneous photon flux densities.
The different light intensities of differently aged and sized forest gaps has
a profound effect on the forest-floor understorey herb community. Dirzo et
Deforestation effects on vegetation 155
Advance {
regeneration
Sp,o",'''9 {
Seed book {
[1 1
Seed
dispersal { ;:;::::::; .::::::
'----v---' '-----.,v,----'
Natural "
TreefalVfirel Farm pasturel
forest natural anthropogenic
Figure 6.12 Four main forest regeneration processes, as influenced by size and
severity of forest disturbance. (From Uhl et aI., 1989.)
al. (1992), studying the herbaceous vegetation of a wide range of forest

gaps in south-west Mexico, found that the flora of a number of 3-5-year-
old gaps was more similar than either an equivalent number of young «2
years) or old (>5 years) gaps. They attributed this to the fact that light con-
ditions in a gap become more similar as gap closure occurs and suggest
that the ecology of 'building phases' of the gap is critical for determining
the structure, composition and dynamics of the forest herbaceous commu-
nity.
As with light and shade tolerance, regenerating tropical saplings have
differing abilities to cope with extremes of soil moisture availability.
Wright (1992) and Grubb (1996) hypothesized that, in comparison to
plants in gaps, understorey plants will be more affected by seasonal
droughts. In forest gap conditions in Ghana, Veenendaal et al. (1995) con-
firmed that shaded, understorey saplings experience more drought stress
during the dry season than do seedlings growing in the forest gap. The dif-
ference was attributed to significantly sharper reductions in stomatal con-
ductance of shaded plants and associated reduced leaf-water potentials
compared to those in the forest gap. Several authors have reported that
understorey plants show both morphological and physiological adapta-
tions to cope with drought stress. In West Africa, some understorey
saplings have sclerophyllous evergreen leaves with thick cuticles
(Veenendaal et aI., 1995), while in dipterocarps in Brunei, Booth (1996)
noticed that leaves of understorey saplings had lower stomatal densities
than their counterparts regenerating in forest gaps. Observations such as
these are beginning to explain the higher mortality of understorey seedlings
compared to gap seedlings under drought conditions, despite the seemingly
harsher microclimate of forest clearings.
6.9.1 Seed ecology after clearance

Critical issues of seed ecology after clearance are seed production and dis-
persal, their dormancy abilities once in the soil, how susceptible they are to
predation and decay and which key factors trigger germination. These
issues and many others are reviewed by Whitmore (1983). Initial stages of
regrowth are dominated by shade-intolerant pioneer species whose seeds
were already present in the soil under the existing forest. Changes in light
intensity and spectral distribution when the canopy is removed (Fig 6.1a,
b) acts as a trigger for germination. There has been some speculation that
soil temperature changes or the wetting-drying cycles which accompany
clearance could be responsible for germination of some seeds. Most
authors, however, agree with Vazquez-Yanes (1976) that the main trigger is
the removal of the photosynthesizing forest and vegetation canopy, which
results in increases in the red wavelengths of light received at the soil sur-
face in clearings. Shade acts to maintain seed dormancy. Hall and Swaine
(1980) found that nearly all seeds from Ghanaian forest soil seed banks
failed to germinate in the shade. In Panama, Putz (1983) found enhanced
germination after soil disturbance caused by upturned root boles of fallen
trees. This may have been partly due to exposure of buried seed and partly
to enhanced nutrient mineralization in the zone of disturbance, indicating
that disruption of the soil surface by clearing or logging may not interrupt
this type of regeneration.
Viability studies have indicated that seeds of primary forest trees have
little capacity for dormancy and are usually obligate immediate germina-
tors. Pioneer species, on the other hand, may survive dormant in the soil
for up to 6 years (Lebron, 1979). Most seed-bank germination studies indi-
cate the dominance of one or two pioneer species. Putz (1983) found that
the fast-growing colonizers Miconia argentea and Cecropia insignis
accounted for 54 and 15%, respectively, of the germinated seeds in seed
banks from Barro Colorado Island in Panama. Uhl and Clark (1983) found
that a different species of Cecropia, C. ficifolia, dominated the seed banks
of both terra firme oxisol seed banks (70%) and caatinga spodosol seed
banks (55%) at San Carlos in Venezuela. It is important to point out, how-
ever, that the viable seed bank in forest soil can be very diverse. Ewel et al.
(1981) managed to germinate 67 species from only 0.11 m 2 of soil,
amounting to nearly 8000 individuals/m2 , in an 8-year-old forest in Costa
Rica.
There is very clear evidence that fire and high soil-surface temperatures
kill viable seeds. Uhl and Clark (1983) found that a heat treatment of only
80° C for 5 min reduced or eliminated the germination of all seeds. Ewel et
at. (1981) and Brinkmann and Vieira (1971) have shown that fire very sig-
nificantly reduces the seed bank. Forest clearance changes the size and
composition of the seed bank. Germinated soil seed banks from forest and
cleared sites at San Carlos were different in size and showed quite different
composition (Table 6.5) (Uhl and Clark, 1983), with dominance of grasses
and forbs in disturbed sites and successional woody species in forest sites.
Reviewing a range of studies world-wide, Uhl and Clark (1983) noted dif-
ferences in viable seed numbers in disturbed and undisturbed forest seed
banks: numbers in forest soil ranged from 180-860/m2, while numbers in
disturbed sites are much higher, at 500-8000/m2.
Within-site seed dispersal increases very rapidly as the site is colonized.
Uhl et at. (1981a), using seed traps in slash-and-burn areas at San Carlos,
found no seed influx over a 2-week period on the freshly burnt site, but
found 197 seeds/m2 influx over the same period on a 16-month-old site.
With time, after a clearing has been opened up, physical and biological vec-
tors operate to bring primary forest seed to the new site. In clearings, grav-
ity, wind and water become less important dispersers of seeds from
primary forest trees; birds, bats and to a lesser extent larger animals
become more important dispersers, often dropping seeds in faeces, which
provides an ideal germination medium. Uhl et at. (1981a) found that seeds
of primary forest species began to arrive and grow as early as 4 months
after clearing and burning. Much seed predation in the forest is carried out
by animals which are unwilling to venture into open fields or pastures.
Early stages of succession can thus often show reduced rates of seed preda-
tion (Janzen and Vazquez-Yanes, 1991). Remnant forest trees which sur-
vive deforestation provide an important seed source in isolated clearings.
Sometimes mature trees are left to provide shade for grazing animals; occa-
sionally trees are not killed by burning. Isolated trees attract birds which
drop or regurgitate seeds which then accumulate in the vicinity of the par-
ent tree. Guevara et at. (1986) pointed out that these isolated trees became
regeneration nuclei. After 1 year of regrowth following farm abandonment
in Amazonia, Uhl (1987) reported that areas in the shade of isolated fruit
trees or slash had significantly more woody colonizers than did more open
areas.
Clear felling and burning are destructive of seedlings and saplings on the
forest floor, but after logging some may survive to quickly take over the
clearing. Uhl et al. (1988b) report as many as 10-20 seedlings or
saplings/m2 of forest floor. These survive in a state of arrested development
in the understorey and grow rapidly when sunlight becomes available (Uhl
et aI., 1989). Almost all of these trees were primary forest species. It is
important to point out that these are results from tree-fall gaps and that
advance regeneration is much less important in deforested areas (see Figure
6.12).
Most tropical tree species are not solely dependent on seed for their
Table 6.5 Seed bank size and composition in forested and cleared sites at San Carlos, Amazonia (after Uhl and Clark, 1983)
Undisturbed forest Disturbed forest

3-year-old farm, 6-year-old
Seed bank characteristics Caatinga Terra Firme just abandoned pasture, degraded
Mean number of germinations per m2 200 180 580 1250

Total germinations growing to
identifiable stage 78 69 74 100
Proportions of germinated seed:
Forested tree species (%) 2 1 0 0
Successional tree species (%) 98 95 8 0
Forbs (%) 0 1 20 54
Grasses (%) 0 0 72 45
regeneration after felling and burning, as most are capable of vegetative
sprouting or coppicing. Sprouts rarely survive burning. They are common-
est in early stages of succession following clearance. In a sequence of regen-
erating forest stands in Amazonia on farms which had been abandoned for
9 to >80 years, Saldarriaga et al. (1988) found that the number of sprouts
fell rapidly in plots older than 20 years. Up to 20 years, there were as many
as 2000-4500 individualslha. In the Gogol Valley in Papua New Guinea,
Saulei (1985) reported seedling growth rates to be faster than coppice
growth rates. Two years after logging, the average height of seedlings was
10.7 m, while that of coppice was 3.1 m. This difference he attributed to
the fact that most seedlings were secondary succession species with fast
growth rates, while coppices were predominantly primary tree species.
Despite these differences, Bellingham et al. (1994) have shown that sprout-
ing is an extremely important mechanism for survival and recovery of
Jamaican forests after severe hurricane damage. They also found high
sprouting frequencies in eight common tree species rarely observed to ger-
minate or establish. These results parallel those of Yih et al. (1991) study-
ing hurricane damage in Nicaragua, and Unwin et al. (1988) reporting
forest recovery after cyclone damage in Queensland, and show that sprout-
ing plays a major role in maintaining forest species diversity and recovery
after damage.
6.9.2 Accumulation of vegetation biomass and nutrients during

regeneration
Biomass accumulation is seen in both above- and below-ground compart-
ments of regenerating forest with time after clear felling. In the first 5 years
of forest succession after slash-and-burn agriculture in Amazonia, Uhl
(1987) measured a steep increase in stem biomass and a gradual increase in
root biomass (Figure 6.13a). He estimates that a mean of 677 glm2 of
above-ground biomass accumulated each year during the first 5 years of
succession. Over a longer successional period of forest regeneration on
abandoned farms in the same area, Saldarriaga (1989) reported a gradual
biomass increase both above and below ground (Figure 6.13b), until over
80 years, when the magnitude of stem, leaf and root biomass compart-
ments approached those of mature forest. In the first 5 years of forest suc-
cession reported by Uhl (1987), branch plus stem biomass increased from
29-35% of the living biomass of the system in year 1, to 79-85% in year
5. A similar successional trend in biomass distribution is reported by Toky
and Ramakrishnan (1983a) in northern India, with branch and stem bio-
mass accounting for as much as 93-95% of the total in 10-20-year-old
forest fallows. The stem component of the mature forest in the study
reported by Saldarriaga (1989) made up 77% of the total system biomass.
In section 6.7, the recovery of soil fertility after forest clearance was
40
30
In 20
In
<U
E
o
co 10
(a) 2 3 4 5
Time
(Years following farm abandonment)
400
300
~
co
--
~
~
In 200 leaves
In
(Il +
E
0
CO
100
Stems
(b)
Figure 6.13 Biomass accumulation in different compartments of forest succession

after slash-and-burn agriculture in the San Carlos region of Venezuelan Amazon.
(a) Short-term changes (0-5 years, from Uhl, 1987); (b) long-term changes (0-100
years, from Saldarriaga, 1989).
discussed for several subsequent land uses. A large number of studies show
that the total stock of nutrients biologically immobilized does increase with
successional development, mainly as a result of increasing vegetation bio-
mass (Snedaker, 1980). The quantity of nutrients immobilized in the leaves,
however, does not increase after early levelling off of foliar concentrations.
The vegetation nutrient stock, particularly foliar concentrations, is an
important control on soil fertility as litter fall and below-ground organic
inputs are the prime route whereby nutrients are pumped back into top-
soils. In the abandoned pastures described by Buschbacher et al. (1989) in
eastern Amazonia (see Figure 6.5), and in the forest fallows described by
Aweto (1981) in Nigeria, soil fertility recovery appears to occur more
rapidly than forest structure, biomass and vegetation nutrient stocks. In
Amazonia (Figure 6.5), stocks of calcium, magnesium, nitrogen and phos-
phorus in vegetation fall far short of those in mature forest, and potassium
levels are only half those of forest. Clearly after 8 years of regeneration
there is still a long way to go to approach forest maturity. However, Jaffre
(1985), reporting data from a slash-and-burn system in the Ivory Coast
studied for 15 years of fallow following 1 year of post-deforestation rice
cultivation, found that the above-ground biomass accumulated greatest
amounts of nutrients in the early stages of succession, up to 1 year. Eighty
years of successional regeneration after shifting cultivation in the upper
Rio Negro in Amazonia is clearly long enough for nutrient stocks in all
ecosystem compartments to very closely approximate those in mature for-
est close by (see Figure 6.8) (Saldarriaga, 1989).
6.9.3 Regeneration patterns after deforestation

With four main processes contributing to the forest regeneration pattern
after deforestation (see Figure 6.12), one might expect to see great varia-
tion in succession patterns, depending on local environmental and ecologi-
cal conditions. Despite this, there are clearly recognizable trends in the
growth of secondary and primary tree species with time after clearance.
The density of primary and secondary tree species during the early stages
of successions is illustrated by Swaine and Hall (1983) for cleared forest
land in Ghana (Figure 6.14). The exponential decline in secondary species
after 2 years is due to mortality caused by competition between adjacent
trees. This general pattern of a steep recruitment curve for secondary
species, reaching maximum tree density after 1 year, followed by a mortal-
ity decline due either to competition or environmental stress, such as
drought, is quite typical of post-clearance successions.
It is not unusual for many of the primary tree species to be present in the
succession as sprouts, while secondary species tend to be newly germinated
seedlings. In cleared forest land in Kalimantan, Kartawinata et al. (1980)
found different proportions of resprouts and new seedlings in the new for-
est successions, depending on whether the slash had been burnt. In un burnt
plots, successions showed 42.2 % sprouts, while in burnt plots, sprouts
accounted for only 28.7% of the regrowth. Although sprouting from cut
stumps was found to be the most common form of regeneration after forest
---- --- ---

Primary species __
----- ----
.--
"."",.....-- ------ ."".,..,....
2 3 4 5
Time after clearing (years)
Figure 6.14 Density of secondary and primary trees in the first 5 years following
forest clearance in Ghana. (From Swaine and Hall, 1983.)
cutting at San Carlos in Amazonia, making up 870/0 of the plant density

(Uhl et al., 1981a), sprouting was completely curtailed after burning.
The earlier suggestion that soil disturbance influenced regeneration
prompted Uhl et ale (1981a) to examine the microhabitat preferences of
germinating tree seedlings after slash and burn in Amazonia. They identi-
ned three main types of germinating habitats after slash-and-burn clear-
ance (root mat, charcoal and bare soil), and studied the environmental
conditions of each with and without slash. Soil-surface temperatures were
highest on sites without slash, particularly on charcoal, which reached a
temperature of 54° C. High daytime temperatures elevated rates of soil
moisture evaporation, and the sites with slash covering all showed higher
moisture retention and resulted in better seed germination. Some species
were found to 'prefer' bare soil surfaces (e.g. the grass Panicum laxum),
while the woody species Cecropia ficifolia and Vismia lauriformis germin-
ated preferentially on charcoal and base soil sites covered with slash. This
variety in germination habitat availability and preferences is certainly one
reason for spatial diversity during the early stages of forest succession.
Conclusions 163
Nearly all studies of forest regeneration after clearance or after pasture
and agriculture abandonment report the early growth of grasses and
sedges, of understoreys of Solanum, Heliconia and Calathea, and of the
fast-growing woody genera such as Cecropia, Trema, Inga and Miconia.
Some very fast growth rates have been reported for these pioneer woody
species, with Brokaw (1985b), working on Barro Colorado Island in
Panama, reporting the amazing height extensions of 6.75, 5.0 and
4.9 mlyear for Trema micrantha, Zanthoxylum sp. and Cecropia insignis,
respectively. Most primary species are slow-growing, but some, such as the
Albizzia spp. reported by Swaine and Hall (1983) in Ghana, can put on
around 1 m height extension/year.
The species diversity of regenerating tropical forest increases over the
initial stages of succession. Swaine and Hall (1983) noted an increase in
total number of different species per 800 m2 up to 50-60 after 2 years,
after which time numbers levelled off gradually. They showed that the rate
of immigration of new species was sharply negatively exponential, with the
bulk of new species, both primary and secondary, arriving in the first year
after forest clearance. This severe slowing down in the rate of recruitment
of new woody species is one reason why tropical forest recovery takes so
long.
6.10 CONCLUSIONS
Evidence from post-clearance forest regeneration on a range of different
soil types in different geographical regions indicates that, given sufficient
time, the biomass, structure, nutrients and even the species diversity of
mature tropical forest may be regained. A major caveat would be that
regenerating areas must be within contact of primary forest seed sources.
Clearly the elimination of primary forest from an area would prevent sub-
sequent regeneration of primary species unless seeds or seedlings were
imported from elsewhere. Concern that subsistence agriculture after defor-
estation quickly becomes unsustainable may be dispelled if judicious man-
agement is employed. An example of this would be the jhum fallow cycles
of India (Ramakrishnan and Toky, 1981) which generate some degree of
soil fertility restoration, depending on fallow length. Many woody species
in regenerating forest employ special techniques to cope with particularly
nutrient-poor conditions. Uhl (1987) quotes examples of crops and sec-
ondary forest species which have low root-to-shoot ratios in enriched soil
conditions after the clearance burn, but which produce higher root bio-
mass under infertile conditions. Mycorrhizal associations are another
mechanism for maximizing nutrient uptake in degraded soils.
There are three major requirements for successful forest regeneration
after disturbance: availability of microhabitats for seed germination and
for seedling establishment, and availability of nutrients (Uhl, 1987). After
shifting cultivation, a major problem for forest succession after farm aban-
donment is that continuous weeding during the lifetime of agriculture
depletes the plot of seeds and seedlings which would initiate forest
regrowth. Reliance is then placed on seed dispersal and influx for regenera-
tion. To assess the influence of farming on subsequent forest regeneration,
Uhl (1987) compared the structure, biomass and species composition of
successions growing on immediately cleared land at San Carlos in
Venezuela with regrowth on sites at an abandoned farm. After 5 years of
succession, the immediately abandoned plot had twice the density of stems
as the abandoned farm, with 37% of the stems representing forest species,
compared to 7% on the abandoned farm. The immediately abandoned plot
also had higher species diversity, with 25 speciesl100 m2 , while the aban-
doned farm had only nine species per 100 m2 • These differences were
attributed to weeding during the cultivation period rather than to loss of
soil fertility. Uhl et al. (1988a) noted that the growth of vines in abandoned
pastures provides major competition for regenerating forest species. They
found 68 vine species, of which one third has small, windblown seed,
potentially giving them very early access to new sites. These examples illus-
trate how post-clearance land use, as well as the deforestation procedures
themselves, influence subsequent forest succession.
Much recent and current research on gap-phase dynamics yields useful
information for ecologists interested in post-clearance processes. It seems
surprising, however, that ecologists have concentrated on gap size as a
determinant of forest regeneration pattern when it is only a general sur-
rogate for the interaction of specific environmental conditions. It seems
more logical to look at the influence of individual environmental and eco-
logical factors in more detail, such as light, temperature and humidity char-
acteristics; soil physical and chemical characteristics, particularly soil
hydrology; animal behaviour, including predation and dispersal mechan-
isms; and soil biological relationships and processes. Many such studies are
now in their infancy and should help greatly in providing management
advice for remedial treatment of existing deforested areas, and for mini-
mizing habitat degradation during clearance. It is already apparent that
mechanized clearance using bulldozers causes excessive damage to the
soil's physical condition and removes surface soil litter and roots. Fire also
removes soil organic matter and kills seeds and sprouts. Clearance tech-
niques which avoid these practices and which cause minimum damage to
the soil litter layer and root mat will retain maximum soil nutrients and
provide the best conditions for regrowth.
Tropical forest regeneration is a slow process. Less than 10% of defor-
ested land is replanted to forestry each year (Gomez-Pompa et aI., 1991).
In many cases replanting with non-native species proves unsuccessful, often
due to infertility and sometimes due to disease. A change in tropical forest
policy is needed to place heavier emphasis on the use of native species in
References 165
forest plantings, or to plan forest clearance patterns and operations in such
a way that manageable natural regeneration is encouraged.
REFERENCES
Alegre, J. C. and Cassel, D. K. (1986) Effect of land clearing methods and post-
clearing management on aggregate stability and organic carbon content of a soil
in the humid tropics. Soil Science, 142,289-295.
Alegre, J. c., Cassel, D. K. and Bandy, D. E. (1986) Effects of land clearing and
subsequent management on soil physical properties. Soil Science Society of
America Journal, 50,1379-1384.
Alegre, J. c., Cassel, D. K. and Bandy, D. E. (1988) Effect of land clearing method
on chemical properties of an Ultisol in the Amazon. Soil Science Society of
America Journal, 52,1283-1288.
Alexander, 1. (1989) Mycorrhizas in tropical forests, in Mineral Nutrients in
Tropical Forest and Savanna Ecosystems, (ed. J. Proctor), Blackwell Scientific,
Oxford, pp. 315-324.
Allen, J. C. (1985) Soil response to forest clearing in the United States and the
tropics: geological and biological factors. Biotropica, 17, 15-27.
Anderson, J. M. and Flanagan, P. W (1990) Biological processes regulating organic
matter dynamics in tropical soil, in Dynamics of Soil Organic Matter in Tropical
Ecosystems, (eds D. C. Coleman, J. M. Oades and G. Uehara), NiffAL Project,
Department of Agronomy and Soil Sciences, College of Tropical Agriculture and
Human Resources, University of Hawaii, pp. 97-123.
Anderson, J. M. and Spencer, T. (1989) Carbon, nutrient and water balances of
tropical rainforest ecosystems subject to disturbance and implications for man-
agement, paper commissioned under UNESCO Contract No RP 268127.9 and
presented at a Man and the Biosphere workshop on Ecological and Economic
Sustainability of Tropical Rainforest Management, Paris, September 1989.
Aweto, A. O. (1981) Secondary succession and soil fertility restoration in south
western Nigeria. 1. Soil fertility restoration. Journal of Ecology, 69, 609-614.
Ayanaba, A. and Jenkinson, D. S. (1990) Decomposition of carbon-14 labelled rye-
grass and maize under tropical conditions. Soil Science Society of America
Journal, 54, 112-115.
Ayanaba, A., Tuckwell, S. B. and Jenkinson, D. S. (1976) The effects of clearing
and cropping on the organic reserves and biomass of tropical forest soils. Soil
Biology and Biochemistry, 8, 519-525.
Basu, S. and Behera, N. (1993) The effect of tropical conversion on soil microbial
biomass. Soil Biology and Fertility, 16, 302-304.
Bazzaz, F. A. (1991) Regeneration of tropical forests: physiological responses of
pioneer and secondary species, in Rain Forest Regeneration and Management,
(eds A. Gomez-Pompa, T. C. Whitmore and M. Hadley), Man and the Biosphere
Series, Vol. 6, UNESCO, ParislParthenon, Carnforth, pp. 91-118.
Bellingham, P. J., Tanner, E. V. J. and Healey, J. R. (1994) Sprouting of trees in
Jamaican montane forests, after a hurricane. Journal of Ecology, 82, 747-758.
Bernhard-Reversat, F. (1975) Recherches sur l'ecosysteme de la foret subequatoriale
de basse Cote d'Ivoire. Les cycles des macroelements. La Terre et la Vie, 29,
229-254.
Bernhard-Reversat, E, Huttel, C. and Lemee, G. (1978) Structure and functioning
of evergreen rain forest ecosystems of the Ivory Coast, in Tropical Forest
Ecosystems, UNESCO, Paris, pp. 557-574.
Bolan, N. S., Robson, A. D., Barrow, N. B. J. and Aylmore, L. A. G. (1984) Specific
activity of phosphorus in mycorrhizal and non-mycorrhizal plants in relation to
the availability of phosphorus to plants. Soil Biology and Biochemistry, 16,
299-304.
Bonde, T. A., Rosswall, T. and Victoria, L. (1992) The dynamics of soil organic
matter and soil microbial biomass following c1earfelling and cropping of a tropi-
cal rainforest soil in the Central Amazon. Soil Science Society of America Journal
Booth, W. E. (1996) Comparative morphology and physiology of gap and under-
storey seedling leaves from a mixed dipterocarp forest, in Tropical Rainforest
Research: Current Issues, (eds D. S. Edwards, W. E. Booth and S. C. Choy),
Kluwer, The Hague, pp. 367-375.
Bormann, E H. and Likens, G. E. (1979) Biogeochemistry of a Forested Ecosystem,
Springer Verlag, New York.
Brasell, H. M. and Sinclair, D. E (1983) Elements returned to forest floor in two
rainforest and three plantation plots in tropical Australia. Journal of Ecology,
71, 367-378.
Brinkmann, W. L. E (1985) Studies on the hydro biogeochemistry of a tropical low-
land forest system. Geojournal, 11, 89-101.
Brinkmann, W. L. E and Vieira, A. N. (1971) The effect of burning on germination
of seeds at different soil depths of various tropical tree species. Turrialba, 21,
77-82.
Brokaw, N. V. L. (1985a) Treefalls, regrowth and community structure in tropical
forests, in The Ecology of Natural Disturbance and Patch Dynamics, (eds S. T.
A. Pickett and P. S. White), Academic Press, New York, pp. 53-69.
Brokaw, N. V. L. (1985b) Gap-phase regeneration in a tropical forest. Ecology, 66,
682-687.
Brown, S. and Lugo, A. E. (1990) Effects of forest clearing and succession on the
carbon and nitrogen content of soils in Puerto Rico and US Virgin Islands. Plant
and Soil, 124, 53-64.
Bruijnzeel, L. A. (1996) Predicting the hydrological impacts of land cover transfor-
mation in the humid tropics: the need for integrated research, in Amazonian
Deforestation and Climate, (eds J. H. C. Gash, C. A. Nobre, J. M. Roberts and
R. L. Victoria), John Wiley, Chichester, pp. 14-55.
Buschbacher, R., Uhl, C. and Serrao, E. A. S. (1988) Abandoned pastures in eastern
Amazonia. II. Nutrient stocks in the soil and vegetation. Journal of Ecology, 76,
682-699.
Buschbacher, R., Uhl, C. and Serrao, E. A. S. (1989) Large-scale development in
eastern Amazonia. Case study No. 10: Pasture management and environmental
effects near Paragominas, Para, in Amazonian Rain Forests: Ecosystem
Disturbance and Recovery, (ed. C. E Jordan), Ecological Studies Vol. 60,
Springer Verlag, New York, pp. 90-99.
Chauvel, A., Grimaldi, M. and Tessier, D. (1991) Changes in soil pore space distri-
bution following deforestation and revegetation: an example from Central
Amazonia. Forest Ecology and Management, 38,259-271.
References 167
Chazdon, R. L. and Fetcher, N. (1984a) Photosynthetic light environments in a
lowland tropical rainforest in Costa Rica. Journal of Ecology, 72, 553-564.
Chazdon, R. L. and Fetcher, N. (1984b) Light environments of tropical forest, in
Physiological Ecology of Plants of the Wet Tropics, (eds E. Medina, H. A.
Mooney and C. Vazquez-Yanes), W. Junk, The Hague, pp. 27-36.
Cochrane, T. T. and Sanchez, P. A. (1982) Land resources, soils and their manage-
ment in the Amazon Region: A state of knowledge report, pp. 137-209, in
Amazonia: Agriculture and Land Use Research, (ed. Hecht, S. B.) Proc. Int.
Conf. sponsored by the Rockerfeller Foundation.
Cole, D. W. and Heil, R. D. (1981) Phosphorus effects on terrestrial nitrogen
cycling. Ecological Bulletin, 33, 363-374.
Collins, N. M. (1980), The effect of logging on termite (Isoptera) diversity and
decomposition processes in lowland dipterocarp forests, in Tropical Ecology and
Development, (ed. J.1. Furtado), pp. 113-121.
Cunningham, R. K. (1966) The effect of clearing a tropical forest soil. Journal of
Soil Science, 14, 334-345.
Dias, A. C. C. P. and Nortcliff, S. (1985) Effects of tractor passes on the physical
properties of an oxisol in the Brazilian Amazon. Tropical Agriculture (Trinidad),
62,137-141.
Denslow, J. S. (1980) Gap partitioning among tropical rainforest trees. Biotropica,
12 (suppl.), 47-55.
Dirzo, R., Horvitz, C. c., Quevedo, H. and Lopez, M. A. (1992) The effect of gap
size and age on the understorey herb community of a tropical Mexican rain for-
est. Journal of Ecology, 80, 809-822.
Eden, M. J.,McGregor, D. F. M. and Vieira, N. A. (1990) Pasture development on
cleared forest land in northern Amazonia. Geographical Journal, 156, 283-296.
Edwards, P. ]. (1982) Studies of mineral cycling in a montane rainforest in New
Guinea. V. Rates of cycling in throughfall and litterfall. Journal of Ecology, 70,
807-828.
Ewel, J. J. (1971) Biomass changes in early tropical succession. Turrialba, 21,
110-112.
Ewel, J. J., Berish, c., Brown, B., Price, N. and Raich, J. (1981) Slash and burn
impacts on a Costa Rican wet forest site. Ecology, 62,816-829.
Fetcher, N., Oberaurer, S. F. and Strain, B. R. (1984) Vegetation effects on microcli-
mate in lowland tropical forest in Costa Rica. International Journal of
Biometeorology, 29,145-155.
Foister, H., De Salas, G. and Khanna, P. (1976) A tropical evergreen forest site with
perched water table, Magdalena Valley, Columbia. Biomass and bio-element
inventory of primary and secondary vegetation. Oecologia Plantarum, 11,
297-320.
Ghuman, B. S. and Lal, R. (1987) Effects of deforestation on soil properties and
microclimate of a high rain forest in Southern Nigeria, in The Geophysiology of
Amazonia, (ed. R. E. Dickinson), John Wiley, Chichester, pp. 225-244.
Golley, F. B., McGinnis, ]. T., Clemmentes, R. G., Child, G. I. and Duver, M. J.
(1975) Mineral Cycling in a Tropical Moist Forest Ecosystem, University of
Georgia Press, Athens, Georgia, USA.
Gomez-Pompa, A., Whitmore, T. C. and Hadley, M. (eds) (1991) Rain Forest
Regeneration and Management, Man and the Biosphere Series, Vol. 6, UNESCO,
Greenland, D. J. and Kowal, J. M. L. (1960) Nutrient content of the moist tropical
forest of Ghana. Plant and Soil, 12, 154-173.
Grimm, V. and Fassbender, H. W. (1981) Ciclos bioquimicos en un ecosistema fore-
stal de los Andes Occidentales de Venezuela. I. Inventario de los reservas organi-
cas y minerales (N, P, K, Ca, Mg, Mn, AI, Na). Turrialba, 31, 27-37.
Grubb, P. J. (1996) Rainforest dynamics: the need for new paradigms, in Tropical
Rainforest Research: Current Issues, (eds D. S. Edwards, W. E. Booth and S. C.
Choy), Kluwer, The Hague, pp. 215-233.
Grubb, P. J. and Edwards, P. J. (1982) Studies of mineral cycling in a montane rain
forest in New Guinea. III. The distribution of mineral elements in the above-
ground material. Journal of Ecology, 70, 623-648.
Guevara, S., Purata, S. E. and Van der Maarel, E. (1986) The role of remnant forest
trees in tropical secondary succession. Vegetation, 66, 77-84.
Hall, J. B. and Swaine, M. D. (1980) Seed stocks in Ghanaian forest soils.
Harcombe, P. A. (1977) Nutrient accumulation by vegetation in the first year of
recovery of a tropical forest ecosystem, in Recovery and Restoration of
Damaged Ecosystems, (eds J. Cairns, K. L. Dickson and E. E. Herricks),
University of Virginia Press, Charlottesville, USA, pp. 347-378.
Harcombe, P. A. (1980) Nutrient loss as a factor in early tropical secondary succes-
sion. Biotropica, 12 (suppl.), 8-15.
Hecht, S. B. (1981) Deforestation in the Amazon Basin: magnitude, dynamics and
soil resource effects. Studies in Third World Societies, 13, 663-684.
Henrot, J. and Robertson, G. P. (1994) Vegetation removal in two soils of the
humid tropics: effects on microbial biomass. Soil Biology and Biochemistry, 26,
111-116.
Herrera, R. (1979) Nutrient distribution and cycling in an Amazon caatinga forest
on spodosols in Southern Venezuela. PhD thesis, University of Reading, UK.
Herrera, R., Capote, R. P., Menedez, L. and Rodriguez, M. E. (1991) Silvigenesis
and the role of mycorrhiza in the natural regeneration in Sierra del Rosario,
Cuba, in Rain Forest Regeneration and Management, (eds A. Gomez-Pompa, T.
C. Whitmore and M. Hadley), Man and the Biosphere Series, Vol. 6, UNESCO,
Herrera, R., Merida, T., Stark, N. and Jordan, c. F. (1978) Direct phosphorus
transfer from leaf litter to roots. Naturwissenschaften, 65, 208-209.
Jaffre, T (1985) Composition mineral et stocks de bioelements dans la biomasse
epigee de recrus fore stiers en Cote d'Ivoire. Acta Oecologia Planta, 6, 233-246.
Jaiyeoba, I. A. (1988) Build-up of organic matter and nutrients under fallow in a
tropical rainforest environment, Nigeria. Malaysian Journal of Tropical
Geography, 18, 10-16.
Janos, D. P. (1980), Mycorrhizae influence tropical succession. Biotropica, 12
(suppl.),56-64.
Janos, D. P. (1983) Tropical mycorrhizas, nutrient cycles and plant growth, in
Tropical Rain Forest: Ecology and Management, (eds S. L. Sutton, T. C.
Whitmore and A. C. Chadwick), Blackwell Scientific, Oxford, pp. 327-345.
Janzen, D. H. and Vazquez-Yanes, C. (1991) Aspects of tropical seed ecology of
References 169
relevance to management of tropical forested wildlands, in Rain Forest
Regeneration and Management, (eds A. Gomez-Pompa, T. C. Whitmore and M.
Hadley), Man and the Biosphere Series, Vol. 6., UNESCO, ParislParthenon,
Carnforth, pp. 137-157.
Jetten, V. G. (1994) Modelling the Effect of Logging on the Water Balance of a
Tropical Rain Forest: A Study in Guyana, Tropenbos Series 6, Tropenbos
Foundation, Wageningen.
Jordan, C. E (1985) Nutrient Cycling in Tropical Forest Ecosystems: Principles and
their Application in Management and Conservation, John Wiley, Chichester.
Jordan, C. E (ed.) (1989) An Amazonian Rain Forest: The Structure and Function
of a Nutrient Stressed Ecosystem and the Impact of Slash and Burn Agriculture,
Man and the Biosphere Series, Vol. 2, UNESCO, ParislParthenon, Carnforth.
Jordan, C. E and Herrera, R. (1981) Tropical rain forests: are nutrients really criti-
cal? American Naturalist, 117, 167-180.
Jordan, C. E, Todd, R. L. and Escalante, G. (1979) Nitrogen conservation in a
tropical rain forest. Oecologia, 39, 123-128.
Jordan, C. E, Caskey, W., Escalante, G., Herrera, R., Montagagnini, E, Todd, R.
and Uhl, C. (1983) Nitrogen dynamics during conversion of primary Amazonian
rain forest to slash and burn agriculture. Oikos, 40,131-139.
Jordan, H. V., Crockett, S. P. and Bardsley, C. E. (1956) Some effects of Kudzu ver-
sus continuous corn on soil properties and crop yields. Soil Science Society of
America Journal, 20, 225-227.
Kartawinata, K., Riswan, S. and Soedjito, H. (1980) The floristic change after dis-
turbances in lowland dipterocarp forest in east Kalimantan, Indonesia, in
Tropical Ecology and Development, (ed. J. I. Furtado), pp. 47-54.
King, D. A. (1994) Influence of light level on the growth and morphology of
saplings in a Panamanian forest. American Journal of Botany, 81, 948-957.
Klinge, H. (1976) Bilanzierung von Hauptnahrstoffen in okosystem tropischer
Regenwald (Manaus) - vorlaufige Daten. Biogeographia, 7, 59-76.
Lal, R. (1976) Soil erosion on alfisols in western Nigeria. VI. Nutrient element loss
in runoff and eroded sediments. Geoderma, 16, 403-417.
Lal, R. and Cummings, D. J. (1979) Clearing a tropical forest. I. Effects on soil and
microclimate. Field Crops Research, 2, 91-107.
Lal, R. and Ghuman, B. S. (1989) Biomass burning in windrows after clearing a
tropical rainforest: effects on soil properties, evaporation and crop yields. Field
Crops Research, 22, 247-255.
Lal, R., De Vleeschauwer, D. and Malafa Nganje, R. (1980) Changes in properties
of a newly cleared tropical Alfisol as affected by mulching. Soil Society of
Lavelle, P. (1987) Biological processes in the productivity of soils in the humid trop-
ics, in Geophysiology of Amazonia, (ed. R. E. Dickinson), John Wiley,
Chichester, pp. 175-214.
Lebron, M. L. (1979) An autecological study of Palicourea riparia Bentham as
related to rain forest disturbance in Puerto Rico. Oecologia, 42, 31-46.
Luizao, E J., Matson, P. A., Livingston, G., Luizao, R. and Vitousek, P. M. (1989)
Nitrous oxide flux following tropical land clearing. Global Biogeochemical
Cycles, 3, 281-285.
Luizao, R. c. c., Bonde, T. A. and Rosswall, T. (1992) Seasonal variation of soil
microbial biomass - the effects of clearfelling a tropical rainforest and establish-
ment of pasture in the Central Amazon. Soil Biology and Biochemistry, 24,
805-813.
Maheswaran, J. and Gunatilleke, A. U. N. (1988) Litter decomposition in a low-
land rain forest and a deforested area in Sri Lanka. Biotropica, 20,90-99.
MaImer, A. and Grip, H. (1990) Soil disturbance and loss of infiltrability caused by
mechanized and manual extraction of tropical rainforest in Sa bah, Malaysia.
Forest Ecology and Management, 38,1-12.
Marrs, R. H., Thompson, J., Scott, D. and Proctor, J. (1991) Nitrogen mineralisa-
tion and nitrification in terra fume forest and savanna soils in Ilha de Maraca,
Roraima, Brazil. Journal of Tropical Ecology, 7, 123-137.
Matson, P. A., Vitousek, P. M., Ewel, H. J. and Mazzarino, M. J. (1987) Nitrogen
transformation following tropical forest felling and burning on a volcanic soil.
Ecology, 68, 491-502.
McElroy, M. B. and Wofsy, S. C. (1986) Tropical forests: interaction with the
atmosphere, in Tropical Rain Forests and the World Atmosphere, (ed. G. T.
Prance), Westview, Boulder, Colorado, pp. 33-60.
Medina, E. and Cuevas, E. (1989) Patterns of accumulation and release in
Amazonian forests of the upper Rio Negro basin, in Mineral Nutrients in
Tropical Forest and Savanna Ecosystems, (ed. J. Proctor), Blackwell, Oxford,
pp. 315-324.
Meiklejohn, J. (1955) The effect of bush burning on the microflora of a Kenya
upland soil. Journal of Soil Science, 6, 111-118.
Mello Ivo, w., Ferreira, S., Biot, Y. and Ross, S. (1996) Nutrients in soil solution
following selective logging of a humid tropical 'terra fume' forest north of
Manaus, Brazil. Environmental Geochemistry and Health, 18,69-75.
Meuller-Harvey, I., Juo, A. S. R. and Wild, A. (1989) Mineralisation of nutrients
after forest clearance and their uptake during cropping, in Mineral Nutrients in
Tropical Forest and Savanna Ecosystems, (ed. J. Proctor), Blackwell, Oxford.
Nakano, K. and Syahbuddin (1989) Nutrient dynamics in forest fallows in south-
east Asia, in Mineral Nutrients in Tropical Forest and Savanna Ecosystems, (ed.
J. Proctor), Blackwell, Oxford, pp. 325-336.
Nussbaum, R, Anderson, J. and Spencer, T. (1995) Factors limiting the growth of
indigenous tree seedlings planted on degraded rainforest soils in Sa bah,
Malaysia. Forest Ecology and Management, 74, 149-159.
Nye, P. H. and Foster, W. N. M. (1961) The relative uptake of phosphorus by crops
and natural fallows from different parts of their root zone. Journal of
Agricultural Science, 56, 299-306.
Nye, P. H. and Greenland, D. J.(1960) The Soil under Shifting Cultivation,
Technical Communication No. 51, Commonwealth Bureau of Soils, Harpenden,
UK.
Oberbauer, S. F. and Strain, B. R. (1984) Photosynthesis and successional status of
Costa Rican rain forest trees. Photosynthesis Research, 5, 227-232.
Oldeman, R. A. A. and van Dijk, J. (1991) Diagnosis of the temperament of tropi-
cal rainforest trees, in Rain Forest Regeneration and Management, (eds A.
Gomez-Pompa, T. C. Whitmore and M. Hadley), Man and the Biosphere Series,
Vol. 6., UNESCO, ParislParthenon, Carnforth, pp. 21-65.
References 171
Ovington, J. D. and Olson, J. S. (1970) Biomass and chemical content of El Verde
lower mountain rainforest plants, in A Tropical Rain Forest, (ed. H. T. Odum),
Division of Technical Information, US Atomic Energy Commission, Washington,
DC, pp. H53-H77.
Poels, R. L. H. (1987) Soils, water and nutrients in a forest ecosystem in Surinam,
PhD thesis, Agricultural University, Wageningen.
Potter, G. L., Elasser, H. W., MacCracken, M. C. and Ellis, J. S. (1981) Albedo
changes by man: test of climatic effects. Nature, 291, 47-50.
Proctor, J. (1984) Tropical forest litterfall. II. The data set, in Tropical Rainforest:
The Leeds Symposium, (eds S. L. Sutton and A. C. Chadwick), Leeds
Philosophical and Natural History Society, Leeds, UK, pp. 83-113.
Proctor, J. (1987) Nutrient cycling in primary and old secondary rainforests.
Applied Geography, 7, 135-152.
Putz, F. E. (1983) Treefall pits and mounds, buried seeds, and the importance of soil
disturbance to pioneer trees on Barro Colorado Island, Panama. Ecology, 64,
1069-1074.
Raich, J. W. (1980) Fine roots regrow rapidly after forest felling. Biotropica, 12,
231-232.
Ramakrishnan, P. S. and Toky, O. P. (1981) Soil nutrient status of hill agro-ecosys-
terns and recovery pattern after slash and burn agriculture (jhum) in north east-
ern India. Plant and Soil, 60, 41-64.
Robertson, G. P. (1984) Nitrification and nitrogen mineralisation in a lowland rain-
forest succession in Costa Rica, Central America. Oecologia (Berlin), 61,
99-104.
Robertson, G. P. and Rosswall, T. (1986) Nitrogen in West Africa: the regional
cycle. Ecological Monographs, 56,43-72.
Robertson, G. P. and Sollins, P. (1987) Biological control of soil charge chemistry in
a humid tropical ecosystem in Costa Rica, Central America. Ecological Society
of America, Bulletin, 86, 46.
Robertson, G. P. and Tiedje, J. M. (1984) Denitrification and nitrous oxide produc-
tion in successional and old-growth Michigan forests. Soil Science Society of
Ross, S. M., Thomes, J. B. and Nortcliff, S. (1990) Soil hydrology, nutrient and
erosional response to the clearance of terra firme forest, Maraca Island,
Roraima, northern Brazil. Geographical Journal, 156,267-282.
Russell, C. E. (1989) Plantation forestry: Case Study No 9: The Jari project, Para,
Brazil, in Amazonian Rain Forests: Ecosystem Disturbance and Recovery, (ed.
C. F. Jordan), Ecological Studies, Vol. 60, Springer Verlag, New York,
pp.76-121.
Salas, G. de las and FoIster, N. (1976) Bioelement loss on clearing a tropical rain-
forest. Turrialba, 26, 179-186.
Saldarriaga, J. G. (1989) Recovery following shifting cultivation: Case Study No 2:
A century of succession in the upper Rio Negro, in Amazonian Rain Forests:
Ecosystem Disturbance and Recovery, (ed. C. F. Jordan), Ecological Studies, Vol.
60, Springer Verlag, New York, pp. 24-33.
Saldarriaga, J. G., West, D. c., Tharp, M. L. and Uhl, C. (1988) Long-term
chronosequence of forest succession in the upper Rio Negro of Columbia and
Venezuela. Journal of Ecology, 76, 938-958.
Sanchez, P. A., Villachica, J. H. and Bandy, D. E. (1983) Soil fertility dynamics after
clearing a tropical rainforest in Peru. Soil Science Society of America Journal, 47,
1171-1178.
Saulei, S. (1985) The recovery of tropical lowland rain forest after clear-fell logging
in the Gogol Valley, Papua New Guinea, PhD thesis, University of Aberdeen.
Serrao, E. A. S., Falesi, r. c., Veiga, J. B. and Texeira, J. F. (1979) Productivity of
cultivated pastures in low fertility soils of the Amazon of Brazil, in Pasture
Production in Acid Soils of the Tropics, (eds P.A. Sanchez and L. E. Tergas),
Centro Internacional de Agricultura Tropical (CIAT), Cali, Colombia,
pp. 195-226.
Seubert, C. E., Sanchez, P. A. and Valverde, C. (1977) Effects of land clearing meth-
ods on soil properties of an Ultisol and crop performance in the Amazon jungle
of Peru. Tropical Agriculture (Trinidad), 54, 307-321.
Silva, L. F. (1979) Influencia do manejo de urn ecossistema nas propriedades edafi-
cas dor Oxisols de 'Tabulairo', Resumenes de Ponencias, X Reunion ALCA,
Acapulco, Mexico, Centro de Pesqujisas do Cacau, CEPLAC, Itabuna, Bahia,
Brazil.
Silver, W. L. and Vogt, K. A. (1993) Fine root dynamics following single and multi-
ple disturbances in a subtropical wet forest ecosystem. Journal of Ecology, 81,
729-738.
Singer, R. and Silva Araujo, r. J. (1979) Litter decomposition and ectomycorrhiza in
Amazon forests. Acta Amazonica, 9, 25-41.
Snedaker, S. C. (1980), Successional immobilisation of nutrients and biologically
mediated recycling in tropical forests. Biotropica, 12 (suppl.), 16-22.
de Souza, J. R. S., Pinheiro, F. M. A., de Araujo, R. L. c., Pinheiro, H. S. Jr and
Hodnett, M. G. (1996) Temperature and moisture profiles in soil beneath forest
and pasture areas in eastern Amazonia, in Amazonian Deforestation and
Climate, (eds J. H. C. Gash et al.), Wiley, Chichester, pp. 125-137.
Srivastava, S. C. and Singh, J. S. (1991) Microbial C, Nand P in dry tropical forest
soils: effects of alternative land-use and nutrient flux. Soil Biology and
Biochemistry, 23, 117-124.
St John, T. V., Coleman, D. and Reid, C. P. P. (1983) Association of vesicular-arbus-
cular mycorrhizal hyphae with soil organic particles. Ecology, 64, 957-959.
Stark, N. and Jordan, c. F. (1978) Nutrient retention by the root mat of an
Amazonian rain forest. Ecology, 59,434-437.
Swaine, M. D. and Hall, J. B. (1983) Early succession on cleared forest and in
Ghana. Journal of Ecology, 71, 601-627.
Swift, M. J. and Sanchez, P. A. (1984) Biological management of tropical soil fertil-
ity for sustained productivity. Nature and Resources, 10, 1-10.
Toky, O. P. and Ramakrishnan, P. S. (1983a) Secondary succession following slash
and burn agriculture in north eastern India. r. Biomass, litterfall and productiv-
ity. Journal of Ecology, 71, 735-745.
Toky, O. P. and Ramakrishnan, P. S. (1983b) Secondary succession following slash
and burn agriculture in north eastern India. II. Nutrient cycling. Journal of
Ecology, 71, 747-757.
Toledo, J. M. and Serrao, E. A. S. (1982) Pasture and animal production in
Amazonia, in Amazonia: Agriculture and Land Use Research, (ed. S. B. Hecht),
Centro Internacional de Agricultura Tropical (CIAT), Cali, Colombia,
pp.281-309.
References 173
Tommerup, I. C. and Abbot, L. K. (1981) Prolonged survival and viability of VA
mycorrhizal hyphae after root death. Soil Biology and Biochemistry, 13,
431-433.
Uhl, c. (1987) Factors controlling succession following slash and burn agriculture
in Amazonia. Journal of Ecology, 75, 377-407.
Uhl, C. and Clark, K. (1983) Seed ecology of selected Amazon basin successional
species. Botanical Gazette, 144,419-425.
Uhl, c., Clark, K., Clark, H. and Murphy, P. (1981) Early plant succession after
cutting and burning in the upper Rio Negro region of the Amazon basin. Journal
of Ecology, 69, 631-649.
Uhl, c., Jordan, c. E, Clark, K., Clark, H. and Herrera, R. (1982) Ecosystem
recovery in Amazon caatinga forest after cutting, cutting-and-burning, and bull-
dozer clearing treatments. Gikos, 38, 313-320.
Uhl, c., Buschbacher, R. and Serrao, E. A. S. (1988a) Abandoned pastures in
eastern Amazonia. I. Patterns of plant succession. Journal of Ecology, 76,
663-681.
Uhl, c., Kauffman, J. B. and Cummings, D. L. (1988b) Fire in the Venezuelan
Amazon. 2: Environmental conditions necessary for forest fires in the evergreen
rainforest of Venezuela. Gikos.
Uhl, c., Nepstad, D., Buschbacher, R., Clark, K., Kauffman, B. and Subler, S.
(1989) Disturbance and regeneration in Amazonia: lessons for sustainable land-
use. Ecologist, 19,235-241.
Unwin, G. L., Applegate, G. B., Stocker, G. C. and Nicholson, D. L. (1988) Initial
effects of tropical cyclone "Winifred" on forests in north Queensland.
Proceedings of the Ecological Society of Australia, 15,283-296.
Van der Weert, R. (1974) Influence of mechanical forest clearing on soil conditions
and resulting effects on root growth. Tropical Agriculture (Trinidad), 51,
325-333.
Vazquez-Yanes, C. (1976) Seed dormancy and germination in secondary vegetation
tropical plants: the role of light. Comparative Physiology and Ecology, 1, 30-34.
Veenendaal, E. M., Swaine, M. D., Agyeman, V. K., Blay, D., Aberese, I. K. and
Mullins, C. E. (1995) Differences in plant and soil water relations in and around
a forest gap in West Africa during the dry season may influence seedling estab-
lishment and survival. Journal of Ecology, 83, 83-90.
Vitousek, P. M. (1984) Litterfall, nutrient cycling and nutrient limitation in tropical
forests. Ecology, 65, 285-298.
Vitousek, P. M. and Denslow, J. S. (1986) Nitrogen and phosphorus availability in
treefall gaps of a lowland tropical rainforest. Journal of Ecology, 74,
1167-1178.
Vitousek, P. M. and Matson, P. A. (1985) Disturbance, nitrogen availability and
nitrogen losses in an intensively managed loblolly pine plantation. Ecology, 66,
1360-1376.
Vitousek, P. M. and Sanford, L. (1986) Nutrient cycling in moist tropical forest.
Annual Review of Ecology and Systematics, 17, 137-167.
Wiersum, K. E (1984) Surface erosion under various tropical agroforestry systems,
in Proceedings of a Symposium on Effects of Forest Land Use on Erosion and
Slope Stability, (eds C. L. O'Loughlin and A. J. Pearce), International Union of
Forestry Research Organizations (IUFRO), Vienna and East-West Centre,
Honolulu, Hawaii, pp. 231-239.
Whitmore, T. C. (1983) Secondary succession from seed in tropical rain forests.
Forestry Abstracts, 44, 767-779.
Whitmore, T. (1984) Tropical Rainforests of the Far East, 2nd edn, Clarendon
Press, Oxford.
Whitmore, T. C. (1991), Tropical rain forest dynamics and its implication for man-
agement, in Rain Forest Regeneration and Management, (eds A. Gomez-Pompa,
T. C. Whitmore and M. Hadley), Man and the Biosphere Series, Vol. 6,
UNESCO, ParislParthenon, Carnforth, pp. 67-89.
Whittaker, R. H. (1975) Communities and Ecosystems, Macmillan, New York.
WRI (1985) Tropical Forests: A Call for Action, Report of a Task Force of World
Resources Institute, World Bank and UN Development Programme, World
Resources Institute, Washington, DC.
Wright, S. J. (1992) Seasonal droughts, soil fertility and the species density of tropi-
cal forest plant communities. Trends in Ecology and Evolution, 7, 260-263.
Yih, K, Boucher, D. H., Vandermeer, J. H. and Zamora, N. (1991) Recovery of the
rain forest of southeastern Nicaragua after destruction by Hurricane Joan.
Young, A. (1976) Tropical Soils and Soil Survey, Cambridge University Press,
Cambridge.
Zinke, P. J., Sabhasri, S. and Kunstader, P. (1978) Soil fertility aspects of the Lau
forest fallow system of shifting cultivation, in Farmers in the Forest, (eds P.
Kunstader, E. C. Chapman and S. Sabhasri), University of Hawaii Press,
Honolulu, pp. 134-159.
Zipperlen, S. W. and Press, M. C. (1996) Photosynthesis in relation to growth and
seedling ecology of two dipterocarp rainforest tree species. Journal of Ecology,
84, 863-876.
-7------
Causes of tropical deforestation
and institutional constraints to
conservation
K. S. Bawa and S. Dayanandan
Deforestation rates in the tropics remain high despite global concern for
loss of forests and wildlands. The amount of deforestation increased from
75000 km 2 annually in 1979 to 132000 km 2 in 1991 (Myers, 1994).
Deforestation in the tropics has severe biological and economic conse-
quences. Tropical forests harbour a disproportionate amount of biodiver-
sity relative to their area. For example, more than one half of all species are
estimated to occur in the tropics, which occupy only 7% of the earth's land
area (Myers, 1984). Approximately 34000 endemic species of plants,
which represent 13% of the world's plant species, occur in 3.5% of the
area of the tropics (Myers, 1988). Because of the magnitude and the irre-
versibility of the losses, the actual and potential loss of these species and
the ecosystems in which they occur constitutes one of the greatest environ-
mental disasters of our time.
The economic consequences of the loss of tropical rain forests are
equally deleterious. Forests provide important services including a wide
variety of goods. The economic value of ecosystem services, such as regula-
tion of climate, retention of water and flood control, conservation of soils
and nutrients, carbon sequestration, pest control, pollination, conservation
of genetic resources and detoxification, is considerable (see Myers, 1997a,
b for a recent detailed description of these services). For example, Pearce
and Brown (1994) calculate the cost of converting primary forest to agri-
culture to be US$4000-4400Iha, assuming that every ton of carbon
released causes global warming damage worth US$20. The value of the
natural pest control systems is estimated to range from US$54 billion to
one trillion annually (Naylor and Ehrlich, 1997). The annual value of wild
corn as resource for new genes is estimated to be US$6.8 billion (Fischer
and Hanemann, 1984). Forests also provide a wide variety of goods worth
billions of dollars and directly or indirectly support the livelihoods of mil-
lions of people around the world.
The biological and economic value of forests notwithstanding, the
world's forests (particularly in the tropics) continue to be converted to
other land uses at a rapid rate. This chapter reviews rates and causes of
tropical deforestation and examines institutional constraints to the conser-
vation of forests. Our analysis shows that there is considerable uncertainty
about the rates of deforestation. We find that both the causes of deforesta-
tion and the institutional constraints to conservation are varied and com-
plex. We conclude by offering suggestions that might improve the precision
of our understanding of the rates, patterns and causes of deforestation in
the tropics, so that our efforts at conservation might be more effective.
7.1 DEFORESTATION RATES

Deforestation refers to the clear felling and conversion of forests into some
other land use. Forest degradation can also result in transformation of the
forested landscape into a habitat largely devoid of trees. According to the
classification of the UN Food and Agriculture Organization, land with less
than 10% tree cover is considered as non-forest land (FAO, 1993). The
10% figure is arbitrary, but it is important to note that severely degraded
forests, even with canopy cover less than 10%, are often classified as
forests in official statistics.
The best available and most widely used sources of information about
the state of the world's forests are the reports by FAO (1993, 1996), which
provide data on forest cover in various countries in 1980 and 1990. FAO's
estimates are based on country-wide forest inventories and Landsat satel-
lite images, covering 3.4 million ha of land area, taken approximately 10
years apart. The images comprise 47, 40 and 30 sampling units in Africa,
Latin America and Asia, respectively. Resulting data are used to estimate
the global tropical forest cover. According to this report, the pan-tropical
mean annual deforestation rate for 1980-90 was 0.6-0.7% of the 1980
forest cover.
Deforestation rates range from 0.09% for Suriname to 5.3% for
Jamaica. Table 7.1 lists 10 countries each with the highest rates of defor-
estation in each of the three regions covered by FAO. Recently, deforesta-
tion rates as high as 6.5% have been reported for the Ivory Coast
(Chatelaine et al., 1996). High rates have also been detected for Costa
Rica. For example, for the wet evergreen forests of Costa Rica, Sanchez-
Azofeifa (1996) has estimated an annual deforestation rate of 4.2%
between 1986 and 1991; the FAO estimate for the whole country for
1980-90 is 2.9%. Sanchez-Azofeifa used Landsat imagery for the area cov-
ered by wet forests in Costa Rica, whereas FAO estimates are based on for-
est inventory of the country and a sample of regional Landsat data. It is
Deforestation rates 177
Table 7.1 Countries with the highest annual percentage deforestation in each
region (WRI, 1995)
Africa Asia Latin America

Country Rate Country Rate Country Rate
Togo 1.39 Bangladesh 3.28 Jamaica 5.29

Malawi 1.31 Pakistan 2.93 Haiti 3.95
Ghana 1.26 Thailand 2.88 Costa Rica 2.57
Benin 1.24 Philippines 2.88 Dominican Republic 2.46
Tanzania 1.16 Malaysia 1.84 Paraguay 2.38
Guinea 1.15 Vietnam 1.42 El Salvador 2.07
Zambia 1.01 Sri Lanka 1.34 Honduras 1.95
Sudan 1.01 Myanmar 1.22 Trinidad 1.93
Ivory Coast 0.99 Indonesia 1.00 Panama 1.71
Uganda 0.92 Nepal 0.98 Nicaragua 1.71
not known if the higher rates reported by Sanchez-Azofeifa reflect the dif-
ferences in the methodology of the two studies, temporal differences in
rates, or differences between deforestation rates of wet evergreen forests
and dry deciduous forests. As dry deciduous forests represent less than 1 %
of the country's forests, it is unlikely that disparities are due to the exclu-
sion of dry deciduous forests in Sanchez-Azofeifa's study. It is interesting
that for both Ivory Coast and Costa Rica, detailed extensive studies utiliz-
ing improved techniques and approaches show rates higher than those
reported before.
There is considerable variation in deforestation rates among regions. The
1996 FAO assessment shows that in Africa and Latin America the annual
rate of deforestation between 1980 and 1990 was 0.66 and 0.59%,
whereas in Asia the rate was 1.13%. Detailed studies at the local level also
have revealed variations in causes and rate of deforestation. For instance,
Barraclough and Ghimire (1996) have reported an increase in forest cover
in the highly populated western Usabaras area of Tanzania, and rapid
deforestation in the less populated Rufiji area. In the Usabaras area, land
alienation due to plantations, optimized agriculture and government sub-
sistence have led to a reduction in deforestation; in the Rufiji region, defor-
estation has escalated due to government-assisted settlements and in order
to meet the wood demands of the urban population.
Deforestation rates also vary over time. In the Philippines there has been
a steady increase in the rate from 0.5% in the early 1900s to 3.5% by
1980 (Table 7.2; Figure 7.1). In the Agastyamalai region of the Western
Ghats in south-west India, the rate increased from 0.07% during the
period 1920-60 to 0.33% for 1960-90 despite the fact that the area has
4
3.5
;g 3
~
Q) 2.5
~
c 2
0
1.5
~~
S!
•
•
Q)
0 0.5
0
-0.5
1780 1820 1860 1900 1940 1980 2020
Year
Figure 7.1 Rate of deforestation in the Philippines from the 16th century to the
present (Liu et aI., 1993; Kummer, 1992).
Table 7.2 Annual percentage deforestation based on local studies in selected

regions
Region Duration Deforest- Method Reference

ation (%)
Northern
Guatemala 1986-90 0.4% Landsat Sader et al., 1994
Costa Rica 1986-91 4.2% Landsat Sanchez-Azofeifa,
1996
Rondonia, 1973-78 <0.001 Landsat Stone et al., 1991
Brazil 1978-80 0.26% andAVHRR
1980-86 1.77%
Philippines 1700-1900 0.001% Land use maps Liu et aI., 1993
1900-34 0.5%
1934-88 1.07
Philippines <1969 <2% Landsat SPOT Kummer, 1992
1970-80 2.5% and historical
1976-80 3.5% historical records
Malaysia 1972-82 1.7% Land use maps Brown et aI., 1992
Agastyamalail 1920-60 0.07% Land use maps, Ramesh et al., 1997
India 1960-90 0.33 Landsat and IRS
Madagascar 1950-85 1.4% Landsat and Sussman et al., 1994
vegetation maps
been increasingly brought under protection from 1960 onwards (Ramesh
et al., 1997).
According to FAO (1996), over 50% of deforestation during the period
between 1980 and 1990 occurred in closed forest (Figure 7.2), followed by
open forest (12.1 %), shrub land (11.5%) and fragmented forests (8.8%).
Almost 40% of the deforestation in closed forest resulted in a complete
loss of tree cover (Figure 7.3). The rest of the change in closed forests led to
shrub land (26.2%), open forests (19.8%), fragmented forests (9.9%) and
plantations (4.3%).
Concomitant with deforestation is the increasing fragmentation of the
habitat. However, fragmentation statistics are available for only a few
regions. In the wet evergreen forests of Costa Rica, the number of frag-
ments in the size range 3-100 ha increased from 3516 in 1986 to 4085 in
1991 (Sanchez-Azofeifa, 1996). In the Sarapiqui region of Costa Rica, dur-
ing the same time period, the number of fragments in the size range
3-50 ha increased from 379 to 470, and fragments of 50-100 ha in size
increased from 11 to 22 (Sanchez-Azofeifa, 1996). In the Western Ghats,
the number of forest fragments ranging from 10-100 ha increased from 47
in the early 1900s to 396 in 1990 (S. Menon and K. S. Bawa, unpublished
data).
Apart from deforestation and forest fragmentation, tropical forests also
suffer from degradation. The figures for degradation are more scarce than
those for fragmentation. In India, 40% of the country's estimated
600-639 km 2 of forests have canopy cover between 10 and 40% (FSI,
1995). In many other countries, selective logging has led to severe degrada-
tion of forests.
Fragmented and degraded forests are subject to severe losses of biodiver-
sity and ecosystem function. The effects of fragmentation on the structure
and function of tropical forests have been documented by several recent
studies. In general, fragmentation is expected to result in reduction of
species as well as genetic diversity within species, disruption of biotic inter-
actions, changes in nutrient fluxes, and alteration of ecosystem processes
(Schelhas and Greenberg, 1996; Laurance and Bierregaard, 1997). The
incidence of fire can also increase in fragmented forests. Fragmented and
degraded forests are also more easily invaded by exotics and weeds than
are large contiguous forests. For example, in the Biligiri Rangaswamy
Temple wildlife sanctuary in India, the weedy shrub Lantana camara
occurs in almost the entire 540 km2 area (Figure 7.4).
In many parts of the world, particularly in Latin America, old cut-over
forests have been replaced by second-growth forest. World-wide or even
within countries, the extent of such forest is not known. Moreover, there is
little information about differences between primary and secondary forests
with respect to species composition, structure and ecosystem functions.
In summary, the rate of deforestation in the tropics remains high and
Closed forest (51.6%)
Figure 7.2 Percentage deforestation in different land-cover classes over the period
1980-90 (FAO, 1996).
Permanent agriculture
Cattle ranching
Water reservoirs
(39.8%)
Figure 7.3 Conversion of closed forests into other land-cover classes over the
period 1980-90 (FAO, 1996).
there are indications that at least in some parts, the rate is accelerating. In
addition, tropical forests are undergoing qualitative changes. There is a
substantial amount of fragmentation and degradation of large contiguous
forests, and fragmented and degraded tropical forests are subject to inva-
sion by weeds and exotics. Furthermore, large tracts which are currently
being classified as forest are actually secondary forests. If left undisturbed,
these areas of secondary growth may revert to a state similar to primary
forest. Overall, the public debate about tropical forests has been largely
Percentage cover
D 1-15
016-30
31-45
.46-60
. 6 1 -70
Figure 7.4 Distribution of Lantana camara, a weedy shrub species, in the Biligiri
Rangaswamy Temple Wildlife Sanctuary in the Western Ghats of India. Hatched
areas indicate agriculture and plantations. Podus (settlements) shown as diamonds.
Scale, 1 : 200 000.
confined to deforestation; there is little information and discussion about
the health of the remaining forests, or the size of the 'healthy' forests.
7.2 CAUSES OF DEFORESTATION

The causes of deforestation are many, varied and complex. In order to
understand the drivers of deforestation, it is useful to distinguish between
proximate and ultimate causes (Perrings et a!., 1992; Barbier et a!., 1994).
Proximate factors such as logging, collection of fuel wood and conversion
to agriculture explain how deforestation occurs, whereas ultimate or
underlying causes (see also Perrings et a!., 1992; Pearce and Brown, 1994;
Panayotou and Sungsuwan, 1994) explain why deforestation occurs.
However, such a distinction is often not made, leading to confusion about
the factors underlying deforestation. Moreover, there is considerable varia-
tion in the relative importance of various drivers of deforestation among
regions, across and within countries. Finally, the causes are not static; the
relative importance of factors varies over time. The discussion below takes
into account all these distinctions and caveats.
7.2.1 Proximate and ultimate causes

Logging, shifting agriculture, migration of human populations, clear felling
for alternative land uses, and plantations are often cited as causes of defor-
estation (see various papers in Brown and Pearce, 1994). Each of these fac-
tors alone, however, may not be ultimately responsible for deforestation.
For example, pressure for logging may be due to undervaluation of the for-
est resource or the necessity to quickly raise revenues faced by a govern-
ment required to meet domestic expenditure or to meet payments related to
external debt. Shifting agriculture is assumed to be a major cause of defor-
estation (Myers, 1994), but over millennia, due to low population densities
and long fallow cycles, shifting agriculture by itself has not caused deple-
tion of forests. The migration of human populations may be an immediate
cause of deforestation, as for instance in Amazonia (Fearnside, 1993), but
the reasons for migration may be high unemployment in urban areas or
inequities in the control of resources. Clear felling for alternative land use
may also have its origin in rural or urban unemployment or in inequitable
distribution of land, as in much of the tropical world, particularly in Latin
America. Similarly, plantations per se are not responsible for deforestation.
Economic policies that lead to the establishment of plantations are the pri-
mary cause of changes in land use.
Recently, attention has shifted from immediate or proximate to ultimate
causes of deforestation. Even in the ultimate sense, there are several causes
of deforestation, with their relative importance varying from time to time
and from place to place. Competition for land (Panayotou and Sungsuwan,
Causes of deforestation 183
1994); external debt (Kahn and McDonald, 1995); undervaluation of
forests (Panayotou and Ashton, 1993); market and policy failures (Barbier
et aI., 1994); population pressure (Allen and Barnes, 1985; Cropper and
Griffiths, 1994); poverty (Gillis, 1996); and inequity in benefit/cost ratio of
conservation (Wells, 1992) have all been suggested as ultimate or underly-
ing causes of deforestation. These factors are interrelated with linkages in
the form of feedback loops, making it difficult to estimate the relative con-
tribution of each factor. It is also important to note that fundamentally,
each of these factors is the outcome of human behaviour, culture, values,
social institutions and government policies, making it even more difficult to
identify the 'ultimate' or the basic drivers of deforestation.
Competition for land is a major driver of deforestation (Panayotou and
Sungsuwan, 1994). Competition arises not only from intrinsic require-
ments for various products, but also from demands extrinsic to a country.
Extrinsic demands, such as demands for banana, timber, beef, pineapple,
coffee, tea, or other cash and export crops in tropical countries, also point
to the role of international trade in deforestation. The use of land for
export crops creates a wide chasm between the consumer and the place
where the land is used as a production site. Consequently, forest land is
likely to be undervalued, and therefore degraded by logging or converted
to agriculture, because ecosystem services are not taken into account in
assigning value to land (Folke et ai., 1993). A distant consumer, or a multi-
national corporation representing such a consumer, is not likely to assign
value to or pay for ecosystem services provided by the land. For the same
reason, the land is likely to be more severely degraded and subject to ill
effects of pesticides and herbicides. Unfortunately, the distant consumer is
often unaware of environmental costs including the costs of land conver-
sion, and is very rarely asked to bear such costs. The consequences of land
degradation and pollution are left to the victims who are least able to cope
with the ill effects or to bear the costs (Gadgil, 1993). As a result, forest
conversion continues unabated.
Undervaluation of forest ecosystems as a cause of deforestation is thus
linked with decisions concerning land use on the one hand, and interna-
tional trade on the other. The true value of forests, however, is also dis-
counted by consumers and governments. Decisions involving land use have
until recently taken into account the value of standing timber alone, and
have failed to fully realize even that value (Repetto and Gillis, 1988; Gadgil
and Guha, 1993). The economic value of non-timber forest products and
the benefits accruing from ecosystem services have not been considered in
economic valuation of tropical forests. Only recently have economic tools
been applied to a comprehensive valuation of ecosystem services (Daily,
1997).
Market and policy failures occur only when markets do not take account
of the value of ecosystem services and policies are not modified to correct
market failures (Barbier et aI., 1994). Discounting of ecosystem services is
also due to distortions called 'missing markets' (see Pearce and Brown,
1994). Basically, it is the undervaluation of forests that is responsible for
market failures and missing markets. Moreover, market failures also occur
when the distant consumer does not pay for all environmental costs associ-
ated with the conversion of distant forest lands and protection of goods on
these lands, as noted above.
In many cross-country analyses, external debt has been shown to be cor-
related with deforestation (Kahn and McDonald, 1995; Bawa and
Dayanandan, 1997). The connection between external debt and deforesta-
tion is indirect. Countries suffering from heavy external debt must raise
cash quickly to service debt or to pay loan instalments (Kahn and
McDonald, 1995). Cash can be rapidly generated by liquidating forest
stocks or converting forest lands to agriculture to raise cash crops for
exports.
Population pressure is widely assumed to be a major factor underlying
deforestation. Undoubtedly, an increasing number of humans exert a heavy
toll on forests. However, it is simplistic to assume that population pressure
per se is responsible for deforestation. As noted by Barbier et al. (1994),
the impact of population depends on the size of the population in relation
to the size of the affected ecosystems, factors influencing population
growth, and consumption patterns. If population growth is affected by
poverty, which in turn contributes to resource degradation (Dasgupta,
1992; but see Ferraro and Kramer, 1994), then poverty, not population,
would be the basic factor underlying deforestation. Just as poverty leads to
natural resource depletion, so does affluence: consumption patterns are
dependent on affluence. Small populations with a high consumption of for-
est resources can have a more severe effect on forest resources than rela-
tively large populations with low consumption levels.
Moreover, in discussions about the role of population it is important to
distinguish between population density and population growth rate on the
one hand, and forest cover and deforestation rate on the other. The rela-
tionship between population density and population growth rate and
deforestation rate is difficult to test due to a lack of historical data on rates
of forest loss (Pearce and Brown, 1994). Thus most analyses of the effect of
population on deforestation are based on the relationship between popula-
tion density or population growth rate and the amount of forest cover,
rather than the deforestation rate (Kummer and Sham, 1994).
The evidence for a positive association between population density or
growth rate and deforestation rate is ambiguous. Pearce and Brown (1994)
list a number of studies that show the relationship to be either positive or
negative. Using cross-national data, Rudel (1994) found the relationship to
be strongly positive, but using similar data, Bilsborrow and Geores (1994)
note that the relationship is weak. More recently, Rock (1996) observed
Causes of deforestation 185
that the impact of demographic factors is mediated through technological
changes in agriculture, and that an increase in the number of agricultural
workers at the household level without intensification of agriculture does
contribute to deforestation. Using cross-national data, Bawa and
Dayanandan (1997) found that relative to other variables, population den-
sity is the most important correlate of deforestation, but when a different
model was used, population density was replaced by crop-land area as the
most significant factor. In the FAO assessments (1993, 1996), population
density was considered to be an important predictor of forest cover change.
However, in a broader context, Forester and Machlis (1996) using a model
to examine the effect of a multitude of factors on biodiversity loss found
no significant association between population and depletion of biodiver-
sity.
The role of poverty in environmental degradation is well recognized.
Myers (1994) and Gillis (1996) have argued that much of the deforestation
in the tropics is due to slash-and-burn agriculture practised by poor farmers.
Landless farmers playa major role in encroachment on forests. However,
Fearnside (1993) has shown that in Amazonia, 30% of the clearing of
forests in 1991 was due to small-scale farming and 70% due to medium or
large ranching operations. Moreover, Rock (1996) has argued that the rela-
tionship between agriculture practised by poor farmers and deforestation is
complex. Poor agricultural households can have two opposite effects on
deforestation. Surplus labour, when used for intensification of agricultural
production from the same area, can lead to a decrease in the deforestation
rate, but in the absence of intensification it can result in conversion of for-
est to agricultural land. The outcome may depend upon the size of the land
holding and the rootlessness of the rural poor. Household behaviour may
be a key to understanding causes of deforestation, but such studies are
scarce.
Inequity in the control of resources, which also contributes to poverty, is
ultimately one of the major, if not the most important, factors underlying
deforestation. Inequity in the cost-benefit ratio of conservation, which also
stems from unequal distribution and control of resources, is a major factor
responsible for deforestation. As Wells (1992) has argued, the benefits of
conservation are widely dispersed whereas costs are borne by local com-
munities. For the rural poor, costs are further accentuated by insecurity of
tenure over land, small landholdings, and a lack of capital for intensive
agriculture.
7.2.2 Regional differences

Policies to combat deforestation are often discussed in general, global
terms. However, there is considerable spatial and temporal variation in the
factors underlying deforestation. Although a number of studies have
examined causes of deforestation for specific countries or regions, there are
few analyses that take into account variation among regions. Two recent
studies reviewed below focus specifically on regional differences (Table
7.3). These studies are based on correlations between a large number of
socio-economic variables and deforestation rates. Moreover, the studies use
explicit models to examine the relative importance of various factors. It is
important to emphasize that correlation does not mean that there is a
causal linkage, thus inferences about causes are indirect.
Rudel and Roper (1996) performed a qualitative comparative analysis
using data from several sources representing 68 countries for two time
periods, 1976-80 and 1981-90. They selected 11 variables (population
density, rural population growth rate, proximity to coast, hilly or inaccessi-
ble area, roads, per capita debt, gross GNP, amount of protection, size of
forest, wood production and political unrest). Rudel and Roper coded the
variables as either high/low or present/absent, arranged the values in truth
tables, and analysed the underlying patterns using Boolean algebraic tech-
niques. They showed that causes of deforestation vary over time and
among regions. In Asia, rural population growth was important through-
out the study period, but the effect of population density and wood pro-
duction became more pronounced in the 1980s. For the African region,
population density, external debt and GNP played a significant role in the
1970s, but in the 1980s rural population growth and population density
were more important. In the Latin American region, rural population
growth and population density were important in the 1970s, but the effect
of rural population density in the 1980s became less pronounced.
However, overall population density, along with gross national product
(GNP) and wood production, remained important factors during this
period.
Bawa and Dayanandan (1997) performed a quantitative comparative
analysis using data from 70 countries from the World Resources Institute
database. Fourteen variables were grouped under five categories:
• population density: rural, urban and total;
• economic: per capita GNP and per capita external debt;
• forest products: fuel wood and charcoal, industrial roundwood, panel
wood;
• land use: cattle density, cropland, pasture, land in other use;
• energy consumption: traditional fuel and commercial energy.
The correlations between each variable and deforestation rate were
examined. There were strong regional differences among correlations
between selected variables and deforestation rate. For the African region,
deforestation rate was more strongly correlated with rural population than
with urban population density, but an opposite trend was observed for
the Asian region. Pasture land and land in other use were significantly
Table 7.3 Causes of deforestation identified through qualitative (Rudel and Roper, 1966) and quantitative (Bawa and Dayanandan,
1977) comparative studies
Qualitative Quantitative FAG, 1996

Region 1976-1980 1981-1990 1981-1990 1981-1990
Asia Population growth Population growth Cropland * Plantations
Population density Population Density Rural population
Wood product Wood product Resettlement programmes
Other land
Africa Population density Population density Population density* Rural population
GNP Population growth Cattle density Other and
Debt Debt Cropland Fragmentation
Forest size Coast
Hills Civil unrest
Latin America GNP GNP Cattle density Cattle ranching
Population density Population density Population density Other land
Population growth Wood product Wood product Resettlement programmes
Roads Roads Cropland Hydroelectric schemes
Debt Debt Pasture land
Forest size Forest size Other land
In the qualitative analysis, variables which were associated with deforestation are listed. In the quantitative analysis, variables that showed significant
correlations with deforestation rate are listed, and variables with the highest direct effects (~) on deforestation in a given region are indicated by '. Causes
identified by a recent FAO report (FAO, 1996) are also given. GNP = gross national product; Coast = proximity to coast.
correlated with deforestation only in the Latin American region, and cattle
density was significant in Africa and Latin America, but not in Asia. Panel
wood production was significantly correlated with deforestation only in
the Asian region, and none of the forest product variables was significantly
correlated with deforestation in Africa. The relative effect of selected vari-
ables was evaluated through standard and forward stepwise regression
models. The relative importance of a given variable was dependent on the
model used. Standard regression analysis, using a model comprising one
variable from each group, showed that crop land is important in deforesta-
tion in Asia and Latin America, and population density is important in
deforestation in Africa. However, forward stepwise regression analysis
using independent variables which were significantly correlated with defor-
estation in tropical countries (p{{0.05) showed that crop land in Asia, cattle
density in Latin America and population density in Africa are the most
important variables associated with deforestation in each region.
A recent report (FAO, 1996) also highlights regional differences in
causes of deforestation (see Table 7.3). Conversion to plantations in Asia,
forest fragmentation in Africa, and cattle ranching and hydroelectric pro-
jects in Latin America are considered as important factors associated with
deforestation. Rural population pressure seems to play an important role
in deforestation in Asia and Africa, and government-sponsored resettle-
ment programmes in Asia and Latin America.
A comparison of the results of the above studies suggests that cross-sec-
tional analysis at the macro level may be useful in understanding the causes
of deforestation and their regional variations to a certain extent. More pre-
cise information on deforestation rates assessed through remote sensing
and data on socio-economic variables at the local scale are needed for a
better understanding of the causes of deforestation in the tropics.
7.2.3 Temporal variation

The causes of deforestation vary over time. Economic and demographic
parameters, as mentioned above, have a profound effect on deforestation.
As these parameters change, pressure on some forest products is relieved as
demands are created for other products or services. In India, for example,
much of the deforestation in the last century occurred to meet the needs of
the British Royal Navy and the extensive railroad network of the country
(Gadgil and Guha, 1993). At the beginning of the 20th century, large tracts
of forest land were converted into tea and coffee plantations. After inde-
pendence in 1947, the replacement of natural forests by plantations of tim-
ber trees accelerated. During the last 50 years, forests have also been lost to
developmental activities such as logging and hydroelectric projects. In
1980, the rapidly shrinking forest cover forced the government to curtail
clear cutting and logging in most states. Although deforestation due to
Institutional constraints 189
development projects still continues, much of the forest degradation today
is due to grazing and collecting of fodder and fuel wood. Deforestation due
to conversion to agriculture has probably occurred throughout the 20th
century.
In the Sarapiqui region of Costa Rica, much of the deforestation during
the last 5 or 6 decades occurred as a result of cattle ranching operations,
agriculture and banana plantations, but during the last few years banana
plantations have been the major source of deforestation.
In summary, we have a poor understanding of the causes of deforesta-
tion. Forests are being lost due to a number of factors, and the causes are
apparent only in some cases. However, because of high correlation among
various factors, it is difficult to determine the relative importance of vari-
ous drivers of changes in forest cover. There is also considerable spatial and
temporal heterogeneity with respect to underlying causes. Most of our con-
clusions about causes are drawn from gross cross-national surveys that use
questionable data and analytical models. Thus, unless primary data from
relatively small-scale and site-specific studies are carefully analysed, the
debate concerning the most important drivers of change will continue.
7.3 INSTITUTIONAL CONSTRAINTS

As pointed out in the previous section, the causes of deforestation are var-
ied and not well understood. Without precise knowledge of causal mechan-
isms, it is difficult to identify institutional constraints to the conservation of
forests. However, correlates of deforestation do provide clues to the institu-
tional mechanisms that have broken down and need to be restored to cur-
tail deforestation. These mechanisms are discussed below at various levels
of operation.
7.3.1 Local constraints

Until the advent of commercial extractive aCtiVItieS, tropical forests
throughout the world were managed by local communities. These commu-
nities had customs and institutions developed over time to discourage over-
exploitation and habitat degradation (Gadgil et al., 1993). The existence of
sacred groves, forests and ponds indicates that certain areas were set aside
for conservation purposes. However, in many countries, as for example in
South Asia, towards the end of the 19th century many of the community-
owned forests were incorporated into the government-held reserve system
(Gadgil and Guha, 1993). But with the shift in control from private to pub-
lic ownership, community institutions that regulate extraction were not
replaced by effective mechanisms for the utilization and management of
resources. In other places, increasing control by commercial organizations
disrupted local institutions that might have existed. Now in many countries
there is considerable conflict between local communities, commercial inter-
ests and government organizations for the control of forest resources, but
no institutions to effectively regulate the access or to serve the interests of
the various constituencies.
Institutional mechanisms which reduce the conflict between people's
reliance on local natural resources and the conservation of biodiversity also
need to be strengthened. Local communities inside and outside protected
areas depend on forest biomass and forest products for their sustenance
and often have open access to protected areas. Unless the dependence on
local ecosystems for the fulfilment of basic human needs is reduced by
developing alternatives to forest resource use and by curtailment of
poverty, the integrity of the protected areas will continue to suffer. In many
parts of the world attempts are being made to provide economic incentives
to local communities to increase the prospect of sustainable use and to seek
the involvement in conservation of people around the protected areas
(BeN, 1995). Such efforts, however, are confined to very few areas and are
hampered by a number of constraints (Alpert, 1996; Bawa et ai., 1997).
There are no institutional mechanisms to reward local communities for
conservation of biodiversity. The benefits of forests extend well beyond the
area in which these ecosystems occur, yet the cost of conservation is largely
borne by local communities. Recently, Panayotou (1994) and Gadgil and
Rao (1994) have advocated the use of incentives to encourage the judicious
use of biological resources.
Another major institutional failure at the local and other levels is the
standardization and homogenization of the curricula of schools, colleges,
and universities which does not take account of education based on local
ecology, natural resources and biodiversity. Traditional knowledge is being
lost and is not replaced by modern text books. The people exposed to the
prevailing curricula know less about local environmental assets and their
impending fate than their predecessors who tilled soil, herded cattle, logged
forests and caught fish in ponds or rivers. This problem is not peculiar to
tropical countries; the biodiversity crisis seems to have passed by the cur-
ricula and text books used by middle and high schools also in western
countries.
7.3.2 National constraints

At the national level, there are several types of institutional constraint: the
absence of appropriate institutions; the lack of coordination among institu-
tions that influence the use and conservation of forests; the centralization
of policy making and authority; and the absence of appropriately trained
personnel to deal with the deforestation crisis.
The lack of appropriate institutions is well illustrated by the general
absence of country-wide information about the rates of deforestation, forest
fragmentation and forest degradation. Few tropical countries have the
institutions or the institutional capability to monitor forest resources and
their state of health. Only a handful of countries are using technologies
such as remote sensing imagery and geographical information systems to
examine changes in land use on a continuous basis. In addition, long-term
monitoring of ecosystems on the ground is almost non-existent. A lack of
resources and a lack of appreciation of the true value of the forests are the
main factors responsible for the absence of institutions devoted to the sur-
vey and monitoring of forests. Unless there is information and awareness
of the rate and magnitude of changes in forest area, the spatial and tem-
poral patterns of change, and the ecological and economic consequences of
change, it is difficult to expect sound policy and institutions for the conser-
vation of forest resources.
Several other issues related to forests also lack frameworks. For exam-
ple, it is well known that national accounts disregard forests and many
other natural resources as capital assets (Barbier et al., 1994). As men-
tioned earlier, there are almost no institutional mechanisms at the national
level to reward local communities for conservation of forests, though such
ecosystems contribute greatly to national economies. The absence of insti-
tutional mechanisms to resolve critical issues in conservation and manage-
ment of forests reflects a lack of awareness of the true value of forest
ecosystems.
Coordination among different institutions is critical to the curtailment of
losses in forest cover. As mentioned above, the causes of deforestation are
many and varied. Land use is governed by policies influencing population
control and resettlement, agriculture, forestry, animal husbandry, energy,
urban development and transportation. Yet policies for each of the sectors
are formulated independently. Ten years after the World Commission on
Environment and Development urged governments to force coordinating
bodies to deal with environmental problems, there is little coordination,
and actions continue to be taken in isolation from each other (Sayer, 1992).
In many cases, there is little interaction even among agencies that deal with
forest resources and biodiversity. For example, in India a number of agen-
cies deal with issues related to forestry and biodiversity: Forest Survey of
India, Botanical Survey of India, National Bureau of Plant Genetic
Resources, Zoological Survey of India, Wildlife Board, Indian Council of
Forestry Research and Education, and several others. Although all of these
organizations operate under the jurisdiction of the Ministry of
Environment and Forests, there is not enough coordination towards for-
mulation of an effective national plan to assess losses of biodiversity and
forests, to analyse causes of such losses, and to develop a comprehensive
plan for conservation of biodiversity in general and of forests in particular.
Even in a relatively small country such as Costa Rica, the Departmento
General de Forestal is in charge of forest resource assessment and forest
policy, whereas much of the work on inventory and assessment of biodiver-
sity is being done by the Instituto Nacional de Biodiversidad, a non-gov-
ernmental organization (NGO), with little interaction between the two
organizations. There a pressing need to coordinate the actions not only of
the institutions and agencies dealing with related and overlapping areas,
but also of the organizations that directly or indirectly have an impact on
forests.
Centralization of authority and policy formulation are major institu-
tional constraints operating at all levels, from local and county or district
to state and national government and NGOs. The colonial legacy still per-
meates institutions in most tropical countries. These highly structured insti-
tutions confine decision-making power to a few hands even in democratic
countries. Much of the deforestation in many countries is due to the failure
to meet the energy and biomass needs of the rural poor (Gadgil, 1993). Yet
local populations have little influence on the way that local resources are
utilized and conserved. Participatory processes for planning and policy for-
mulation which involve rural populations or small local organizations in
the assessment of forest resources and development of conservation strategies
are almost non-existent.
The development of appropriate human resources to improve research
and development efforts in forestry, to formulate effective policies and
strategies for conservation, and to seek the participation of a broad section
of society in the resolution of environmental problems is critical to the
deforestation problem. Most countries lack the appropriate institutions.
7.3.3 International constraints

Recent years have witnessed an unprecedented proliferation of interna-
tional institutions devoted to the use and conservation of forests and bio-
diversity. These institutions have played an exceedingly important role in
highlighting deforestation and loss of biodiversity. International institu-
tions have been major catalysts in identifying key policy issues in the devel-
opment of agendas for research and action programmes. Many of these
institutions, such as the World Resources Institute, the International Union
for the Conservation of Nature, the World Conservation Monitoring
Center, the World Bank and FAO, have assembled and developed large
databases that provide critical information about the indicators and direc-
tion of change. Finally, the technical expertise available with these institu-
tions has been key to the resolution of environmental problems in many
tropical regions of the world.
The international institutions, however, are also burdened with con-
straints. Firstly, many of these institutions have a narrow, sectoral focus,
although the problems that they seek to address are complex and of multi-
ple origin. For example, land-use change, deforestation, loss of biological
diversity and climate change are interrelated. Yet there are few institutions
that have programmes designed to explore linkages among the critical
areas, the multiplicative effects of these changes, and the policy modifica-
tions necessary to deal with such effects (Bawa and Menon, 1997).
Although international trade (particularly in agricultural commodities and
forest products) greatly influences the way land is used in the tropics, most
international organizations dealing with forestry and biodiversity have
poor representation in negotiations concerning international trade and tar-
iff treaties. Secondly, many institutions have similar programmes. There is
considerable overlap in the goals and approaches as well as in the geo-
graphical areas of operation. In many cases, the objectives and implemen-
tation strategies are identical. Yet there is little coordination, just as in the
case of national agencies. Thirdly, the international institutions usurp con-
siderable resources from the third world institutions. A great deal of inter-
national assistance flows through these institutions, which have very high
overhead costs. Ironically, while these institutions deal with the problems
of the developing world, their operational bases are in cities such as
Geneva, London, New York, Paris, Rome and Washington, where opera-
tional costs are so high that only a small proportion of international assis-
tance ends up at the target site. Moreover, with the long distance between
the operational base and the target site, there is often an unrealistic and
superficial assessment of the issues and problems. It is true that many
developing countries do not have the institutional and human resources to
undertake the type of work being done by the North-based international
organizations, but costs can be considerably reduced by establishing strong
regional centres and by subcontracting the work to local institutions, at
least in some countries. Finally, although the international institutions have
considerable technical expertise, their record in developing the institutions
and the human resources necessary to deal sufficiently with environmental
problems has been abysmal.
Other institutional failures include the discounting of international bene-
fits of forests and inadequate protection of intellectual property rights.
Although forests have global benefits, the international community has
provided few incentives for conservation. Funding from the Global
Environmental Facility is one of the few concrete financial incentives, but
the extent to which the availability of such funds has actually led to conser-
vation is not yet known. The issue of intellectual property rights is com-
plex, but if such rights were to include traditional knowledge and the
genetic value of resources (Swanson, 1995), the benefits may extend well
beyond those who obtain patents or discoveries from tropical forest prod-
ucts without compensating those who provide the initial knowledge and
who conserve resources in the wild.
In summary, institutional constraints are pervasive at all levels, but are
most apparent at the local and national levels. Existing institutions need to
review their goals and develop plans to coordinate their activities with
other institutions. However, in many instances new institutional mechan-
isms are required to address the complex issues of deforestation and nat-
ural resource management.
7.4 CONCLUSIONS
Tropical forests are undergoing major transformations, with serious conse-
quences for biodiversity and the well-being of human societies. However,
the nature of changes in tropical forests is not fully understood. At the
local level, or even at the country level, precise measures of deforestation
are lacking. The extent of fragmentation is not known, nor is there infor-
mation about the impact of disturbance and the degree of forest degrada-
tion. Geographical information systems and remote sensing imagery offer
inexpensive means for the study of land-use change. These technologies,
combined with monitoring on the ground, can provide data for rapid
assessment of the magnitude, rates and patterns of changes. Long-term
ecological monitoring can enhance the capability to assess the health of
tropical forests and the impact of human-induced disturbance. Monitoring
of land-use change must be coupled with collection of information on the
socio-economic variables influencing alterations in land-use patterns (Bawa
and Menon, 1997). The causes of tropical deforestation remain obscure.
Information about the socio-economic determinants of forest resource use
is needed at the local level, just as information about the rate and magni-
tude of deforestation is required at relatively small spatial scales. However,
the actions of the rural poor in degradation of forest resources are rooted
in social and economic inequalities. Tropical deforestation is also linked
with regional and national development plans, international trade and con-
sumption patterns in the developed world. Policies to stem deforestation
must be based on the relative impact on forests of different segments of
society, and the cost of conservation for each segment must be commensu-
rate with benefits accrued from forests and biodiversity.
The institutional constraints on forest conservation are overwhelming
and in part reflect policy failures. Institutions, however, can contribute to
the formation of new policies and the modification and implementation of
existing policies. Although considerable progress can be achieved by reori-
entation of goals and coordination on the part of existing institutions, new
institutions are required if we are to be successful in saving the remaining
tropical forests.
ACKNOWLEDGEMENTS
This paper represents contribution number 57 of the program in conserva-
tion of biodiversity and the environment coordinated by Tata Energy
References 195
Research Institute, New Delhi and the University of Massachusetts at
Boston, supported by the MacArthur Foundation. The work is also sup-
ported in part by grants from the US National Science Foundation. Shaily
Menon, Leah Gorman and Ricardo Godoy offered comments on the man-
uscript.
REFERENCES
Allen, J. C. and Barnes, D. F. (1985) The causes of deforestation in developing
countries. Annals of the Association of American Geographers, 75,163-184.
Alpert, P. 1996. Integrated conservation and development projects. Bioscience, 11,
845.
Barbier, E. B., Burgess, J. C. and Folke, C. (1994) Paradise Lost? The Ecological
Economics of Biodiversity, Earthscan, London.
Barraclough, S. L. and Ghimire, K. B. (1996) Deforestation in Tanzania: beyond
simplistic generalizations. Ecologist, 26, 104-109.
Bawa, K. S. and Dayanandan, S. (1997) Socioeconomic factors and tropical defor-
estation. Nature 386, 562-563.
Bawa, K. S. and Menon, S. (1997) Biodiversity monitoring: the missing ingredients.
Trends in Ecology and Evolution, 12, 42.
Bawa, K. S., Lele, S., Murali, K. S. and Ganesan, B. (1997) Extraction of Non-
Timber Forest Products in Biligiri Rangan Hills, India: Monitoring of a
Community Based Project. Biodiversity Support Program, Biodiversity
Conservation Network, WWF-Worldwide Fund for Nature, Washington, DC (in
review).
Bilsborrow, R. and Geores, M. (1994). Population, land-use, and the environment
in developing countries: what can we learn from cross-national data? in The
Causes of Tropical Deforestation: The Economic and Statistical Analysis of
Factors Giving Rise to the Loss of the Tropical Forests, (eds K. Brown and D. W.
Pearce), University College Press, London, pp. 106-133.
BCN (1995) Biodiversity Support Program, Biodiversity Conservation Network.
1995 Annual Report, Biodiversity Conservation Network, WWF-Worldwide
Fund for Nature, Washington, DC.
Brown, K. and Pearce, D. W. (eds) (1994). The Causes of Tropical Deforestation:
The Economic and Statistical Analysis of Factors Giving Rise to the Loss of the
Tropical Forests, University College Press, London.
Brown, S. L., Iverson, R. and Lugo, A. (1992) Land use and biomass changes of
forests in peninsular Malaysia during 1972-1982; use of GIS analysis, in Effects
of Land-Use Change on Atmospheric CO 2 concentrations: Southeast Asia as a
Case Study, (ed. V. H. Dale), Springer, New York.
Chatelaine, c., Gautier, L. and Spichiger, R. (1996) A recent history of forest frag-
mentation in southwestern Ivory Coast. Biodiversity and Conservation 5, 37-53.
Cropper, M. and Griffiths, C. (1994) The interaction of population growth and
environmental quality. American Economic Review, 84, 250-254.
Daily, G. (ed). (1997) Nature's Services: Societal Dependence on Natural
Ecosystems, Island Press, Washington, DC.
Dasgupta, P. (1992) Population, resources, and poverty. Ambio, 21, 95.
Fearnside, P. M. (1993) Deforestation in Brazilian Amazonia: the effects of popula-
tion and land tenure. Ambio, 22, 537-545.
Ferraro, P. and Kramer, R. A. (1994) Affecting Household Level Behavior to
Promote Rain Forest Conservation in Madagascar, paper prepared for USAID
EPAT Project, Winrock International, Arlington, Virginia (mimeo.).
Fischer, A. C. and Hanemann, W. M. (1984) Option Values and Extinction of
Species, Department of Agricultural Economics, University of California,
Berkeley.
Folke, c., Perrings, c., McNeeley, J. A. and Myers, N. (1993) Biodiversity conser-
vation with a human face: ecology, economics, and policy. Ambio, 22, 62-63.
FAO (1993) Forest Resources Assessment: Tropical Countries, FAO Forestry Paper
112, Food and Agriculture Organization, Rome.
FAO (1996) Forest Resources Assessment: Tropical Countries. Food and
Agriculture Organization, Rome.
Forester, D. J. and Machlis, G. E. (1996) Modeling human factors that affect the
loss of biodiversity. Conservation Biology, 10, 1253-1263.
FSI (1995) The State of the Forest Report, Forest Survey of India, Government of
India, Dehra Dun.
Gadgil, M. (1993) Biodiversity and India's degraded lands. Ambio, 22,167-169.
Gadgil, M. and Guha, R. (1993) This Fissured Land: An Ecological History of
India, University of California Press, Berkeley.
Gadgil, M. and Rao, P. R. S. (1994) A system of positive incentives to conserve bio-
diversity. Economic and Political Weekly, August 6, 2103-2107.
Gadgil, M., Berkes, F. and Folke, C. (1993) Indigenous knowledge of biodiversity
conservation. Ambio, 22,151-156.
Gillis, M. (1996) Tropical deforestation: poverty, population, and public policy.
Vital Speeches, 62, 373-375.
Kahn, J. R. and McDonald, J. A. (1995) Third-world debt and tropical deforesta-
tion. Ecological Economics, 12, 107-123.
Kummer, D. M. (1992) Deforestation in the Postwar Philippines, University of
Chicago Press, Chicago.
Kummer, D. and Sham, C. H. (1994) The causes of tropical deforestation: a quanti-
tative analysis and case study from the Philippines, in The Causes of Tropical
Deforestation: the Economic and Statistical Analysis of Factors Giving Rise to
the Loss of the Tropical Forests, (eds K. Brown and D. W. Pearce), University
College Press, London, pp. 146-158.
Laurance, W. F. and Bierregaard, R. o. (eds) (1997) Tropical Forest Remnants:
Ecology, Management, and Conservation of Fragmented Communities,
University of Chicago Press, Chicago.
Liu, D. S., Iverson, L. R. and Brown, S. (1993) Rates and patterns of deforestation
in the Philippines: application of geographic information system analysis. Forest
Ecology and Management, 57,1-16.
Myers, N. (1984). The Primary Source: Tropical Forests and our Future, W. W.
Norton, New York.
Myers, N. (1988) Threatened biotas: "hotspots" in tropical forests.
Environmentalist, 8, 1-20.
Myers, N. (1994) Tropical deforestation: rates and patterns, in The Causes of
Tropical Deforestation: The Economic and Statistical Analysis of Factors Giving
References 197
Rise to the Loss of the Tropical Forests, (eds K. Brown and D. W. Pearce),
University College Press, London, pp. 27-40.
Myers, N. (1997a) The world's forests and their ecosystem services, in Nature's
Services: Societal Dependence on Natural Ecosystems, (ed. G. Daily), Island
Press, Washington, DC, pp. 215-235.
Myers, N. (1997b) Biodiversity's genetic library, in Nature's Services: Societal
Dependence on Natural Ecosystems, (ed. G. Daily), Island Press, Washington,
DC, pp. 255-273.
Naylor, R. L. and Ehrlich, P. R. (1997) Natural pest control services and agricul-
ture, in Nature's Services: Societal Dependence on Natural Ecosystems, (ed. G.
Daily), Island Press, Washington, DC, pp. 151-174.
Panayotou, T. (1994) Economic Investments for Environmental Management and
Sustainable Development, International Environment Program, Harvard
Institute for International Development, Cambridge, Massachusetts.
Panayotou, T. and Ashton, P. S. (1993) Not By Timber Alone: Economics and
Ecology For Sustaining Tropical Forests, Island Press, Washington, DC.
Panayotou, T. and Sungsuwan, S. (1994) An econometric analysis of the causes of
tropical deforestation: the case of Northeast Thailand, in The Causes of Tropical
Deforestation: The Economic and Statistical Analysis of Factors Giving Rise to
the Loss of the Tropical Forests, (eds K. Brown and D. W. Pearce), University
College Press, London, pp. 192-210.
Pearce, D. and Brown, K. (1994) Saving the world's tropical forests, in The Causes
of Tropical Deforestation: the economic and statistical analysis of factors giving
rise to the loss of the tropical forests, (eds K. Brown and D. W. Pearce),
University College Press, London, pp. 2-26.
Perrings, c., Folke, C. and Miiler, K.-G. (1992) The ecology and economics of bio-
diversity loss: the research agenda. Ambio, 21, 201-211.
Repetto, R. and Gillis, M. (1988) Public Policies and the Misuse of Forest
Resources, Cambridge University Press, Cambridge.
Ramesh, B. R., Menon, S. and Bawa, K. S. (1997) A landscape ecology approach to
biodiversity conservation in the Agastyamalai region, western Ghats, India.
Ambio (in press).
Rock, M. T. (1996) The stork, the plow, rural social structure, and tropical defor-
estation in poor countries. Ecological Economics, 18, 113-131.
Rudel, T. (1994) Population, development, and tropical deforestation, in The
Causes of Tropical Deforestation: The Economic and Statistical Analysis of
Factors Giving Rise to the Loss of the Tropical Forests, (eds K. Brown and D. W.
Pearce), University College Press, London, pp. 96-105.
Rudel, T. and Roper, J. (1996) Regional patterns and historical trends in tropical
deforestation, 1976-1990: a qualitative comparative approach. Ambio, 25,
160-166.
Sader, S. A., Sever, T., Smoot, ]. C. and Richards, M. (1994) Forest change esti-
mates for the northern Peten region of Guatemala - 1986-1990. Human
Ecology, 22,317-331.
Sanchez-Azofeifa, A. (1996) Assessing land use/cover change in Costa Rica. PhD
thesis, University of New Hampshire.
Sayer, J. A. (1992) Institutional arrangements for forest conversion in Africa, in
Conservation of West and Central African Rainforests. World Bank
Environmental Paper Number 1, World BankJIUCN, Washington, DC,
pp.310-317.
Schelhas,]. and Greenberg, R. (eds) (1996). Forest Patches in Tropical Landscapes,
Island Press, Washington, DC.
Stone, T. A., Brown, I. F. and Goddwell, G. M. (1991) Estimation, by remote sens-
ing, of deforestation in central Rondonia, Brasil. Forest Ecology and
Sussman, R. W., Green, G. M. and Sussman, L. K. (1994) Satellite imagery, human
ecology, anthropology, and deforestation in Madagascar. Human Ecology, 22,
333-354.
Swanson, T. M. (ed.) (1995) Intellectual Property Rights and Biodiversity
Conservation, Cambridge University Press, Cambridge.
Wells, M. (1992) Biodiversity conservation, affluence and poverty: mismatched
costs and benefits and efforts to remedy them. Ambio, 21, 237-243.
WRI (1995) World Resources 1994-1995. World Resources Institute, Washington,
DC/UN Environmental Programme/UN Development Programme, Oxford
University Press, Oxford.
-8------
The economics of the tropical
timber trade and sustainable forest
management
E. B. Barbier
Environmental Economics and Environmental Management Department,
University of York, Heslington, York YOl SDD, UK
8.1 INTRODUCTION
In recent years, concern about the destruction of tropical forests has led to
increased interest in the role of the international timber trade in promoting
deforestation in the tropics. At the same time, concerted international
efforts led by global institutions such as the International Tropical Timber
Organization (lITO), the Food and Agriculture Organization (FAO) of the
United Nations and the Intergovernmental Panel on Forests (IPF), are
searching for ways to improve the role of trade in providing incentives for
more sustainable, long-term management of tropical forests. This linkage
between sustainable forest management and the international trade in for-
est products was also enshrined in the global Forest Principles Accord of
the 1992 United Nations Conference on Environment and Development
(UNCED) at Rio de Janeiro.
This chapter examines current trends in the tropical timber trade and
deforestation and their implications for the sustainable management of
tropical forests. The chapter explores important issues of market access
and trade barriers, including the recent effects of the Uruguay Round on
the global forest products trade. Examples of key export and import
restrictions as 'new barriers' to the tropical timber trade are also discussed.
The potential competition from non-wood substitutes, substitution
between tropical and temperate products and the possible role of lesser-
used species of tropical timber are also examined. The chapter goes on to
review current efforts to employ certification and labelling schemes as a
means for discriminating between 'sustainably' and 'unsustainably' pro-
duced tropical timber products, and discusses the implications of such
schemes for trade and sustainable forest management. Finally, the chapter
considers the costs of tropical countries making the transition to sustainable
200 The economics of the tropical timber trade
forest management of their forest industries, and explores various interna-
tional financial mechanisms for raising the additional revenues required.
Because the term 'sustainable' forest management has been interpreted in
many different ways, it is important from the outset to be clear how this
term is being used in this chapter. Here, the chief concern will be with sus-
tainable production and environmental management of those forests cur-
rently or potentially exploitable for timber, although there is also concern
about the impact of timber production and trade on the 'sustainability' of
forest resources generally and the potential values of their goods and ser-
vices. Thus the working definition of sustainable forest management
employed in this chapter follows the definition and criteria of sustainability
adopted by the ITTO:
'Sustainable forest management is the process of managing permanent
forest land to achieve one or more clearly specified objectives of manage-
ment with regard to production of a continuous flow of desired forest
products and services without undue reduction in its inherent values and
future productivity and without undue undesirable effects on the physi-
cal and social environment.'
(From the IITC Decision 6(XI), Quito, 8th Session, May 1991; see also
lITO, 1990).
Equally, it is important to be clear as to definitions and data coverage with
respect to the tropical timber trade. Most available information on the
international trade in wood products has been assembled by the FAO, and
it is this information that will be used primarily in this chapter. Following
FAO convention, tropical wood or timber products will also be referred to
as forest products. The FAO segregates timber products into five general
categories: roundwood; sawnwood and sleepers; wood-based panels; wood
pulp; paper and paperboard. The sub-category roundwood is composed of
both industrial roundwood (i.e. saw logs and veneer logs, pulp wood and
particles) and non-industrial roundwood (i.e. fuel wood and charcoal). All
further processed products (such as furniture, window frames and doors)
are referred to as higher processed products. Trade statistics on the latter
products are not routinely provided by FAO. Tropical timber is defined
here as non-coniferous timber coming from developing countries lying pre-
dominantly between the Tropic of Cancer and the Tropic of Capricorn,
unless specified otherwise. However, the term tropical timber trade will be
used as a shorthand expression to describe the international trade in all
forest products of tropical origin.
8.2 TRENDS IN FOREST PRODUCTS TRADE AND DEFORESTATION

As a background to examining the linkages between forest products trade
and the environment, this section provides an overview of tropical and
Trends in forest products trades and deforestation 201
world production and trade in forest products, as well as future trends in
the supply and demand for forest products. Finally, the available evidence
on the state of closed tropical forests and deforestation is briefly reviewed.
While charcoal and fuel wood production is significant as a percentage
of the volume of total roundwood production, especially in developing
countries (i.e. 55%), only a very small percentage (less than 0.3%) of non-
industrial roundwood production enters the international trade (see Table
8.1a). The value of imports of fuel wood and charcoal compared to the
value of imports of all timber products is also very small - less than 0.2 %
(Table 8.1b). For these reasons, and because fuel wood rarely qualifies as a
traded good due to its low value per unit of volume, it is excluded from
further consideration in this chapter. In the following section, which pro-
vides an overview of the global trade in forest products, lack of data has
also precluded discussion of statistics and trends in higher processed prod-
ucts.
8.2.1 Global forest products trade

Global forest products trade and production have been expanding
throughout the last few decades (Table 8.1a). The volume of industrial
roundwood produced has grown steadily to reach 1600-1700 million m 3
in the early 1990s, of which around 6-7% enters international trade. In
contrast, just under 20% of wood pulp and around 20-25% of sawn-
wood, wood-based panels, paper and paper boards are traded internation-
ally. Although since the 1960s the proportion of wood pulp production
that is traded has remained fairly constant, the share of sawnwood and
wood-based panels traded has doubled (from 11.8 to 22.8% and 12.0 to
25.5%, respectively). The proportion of paper products traded has also
gone up substantially, from 16.5 to 24.9%.
While the value of industrial round wood imports has been rising steadily
over time, the share of logs in the total forest products trade has nearly
halved from a peak of around 20% in the 1970s to 10-12% in the early
1990s (Table 8.1b). The shares of wood pulp and sawnwood in the global
forest products trade have also declined steadily since the early 1960s,
from 19.4 to 11.3% and from 27.2 to 21.4%, respectively. In contrast,
over this period the share of wood-based panels doubled from 6.4 to
12.3%, and the share of paper and paper board increased from 33.1 to
42.6%. These trends reflect the increasing importance of higher valued for-
est products to the global trade.
8.2.2 Tropical timber production and trade

Despite popular perceptions, the global trade in forest products is not
dominated by tropical timber. A large proportion of the timber harvest in
Table 8.1 World production in forest products
Product 1961 1970 1980 1990 1993

(a) Volume of world production and trade (million, cumulative)a
Fuelwood and charcoal
Production 1041 1186 1480 1790 1876
Imports 4 3 3 4 5
Imports as percentage of production 0.3 0.2 0.2 0.3 0.3
Industrial roundwood
Production 1018 1278 1452 1720 1529
Imports 38 93 118 105 106
Sawnwood
Production 346 415 451 505 432
Imports 41 56 78 91 99
Wood based panels
Production 26 70 101 124 133
Imports 3 10 16 30 34
Wood pulpa
Production 62 102 126 154 152
Imports 10 17 21 25 29
Paper and paperboardsa
Production 77 126 170 240 254
Imports 13 23 34 56 63
(b) Value of world trade (US$ million, current)

All forest products (AFP)
Imports 6778 14170 62377 110038 106742
Fuelwood and charcoal
Imports 31 29 122 192 232
as percentage of AFP 0.5 0.2 0.2 0.2 0.2
Industrial roundwood
Imports 908 2693 12 316 11269 12 867
Sawnwood
Imports 1841 3020 13 952 20429 22864
Continued
Trends in forest products trades and deforestation 203
Table 8.1 Continued
Product 1961 1970 1980 1990 1993

Wood based panels
Imports 437 1184 5236 11012 13 077
Wood pulpa
Imports 1316 2650 9777 17321 12047
Paper and paperboardsa
Imports 2244 4567 20845 49606 45431
a Wood pulp, paper and paper boards in million mt.
Source: FAO Forest Products Yearbook, various years.
tropical countries is consumed domestically and does not enter the interna-
tional trade. Only 17% of total tropical timber production is used for
industrial purposes; the remainder is consumed for fuel wood and other
non-industrial uses. Out of the total volume of industrial timber produced
by tropical countries, approximately 31 % is exported in round or product
form. In other words, only about 6% of total tropical roundwood produc-
tion enters the international trade (Barbier et aI., 1994). In addition, tropi-
cal timber plays a minor role in the global timber market; tropical
countries account for approximately 15% of the total volume of global
timber production, and 11 % of the value of global exports.
While the production of tropical round wood has been steadily increasing
over the past few decades, timber consumption in tropical countries has
been growing at an even faster rate due to population and income growth.
The rapid expansion of timber demand in many tropical countries has led
to increased domestic consumption, reduced timber exports and increased
timber imports. As a result, many tropical timber-producing countries are
becoming net timber importers.
Although exports of industrial roundwood, sawnwood and sleepers con-
tinue to dominate tropical timber exports, producer countries are increas-
ingly exporting more highly processed products (e.g. wood-based panels,
wood pulp and paper and paperboard) in an effort to capture a higher pro-
portion of the value added in timber processing. This pattern is especially
apparent in the Asia region where the percentage of wood-based panels
exported relative to total production has risen from 40% in the 1960s-70s
to 90% in 1990 (Barbier et aI., 1994).
8.2.3 Forest product price trends
Although forest product prices have risen steadily, the real forest product
price index has been fairly stable since the 1960s (see FAO, 1992 for more
details). The index fell gradually in the early 1980s but has been increasing
since 1985. Within the broad forest products category, the real price of
tropical logs has followed a rising trend since the early 1970s; prices briefly
declined between 1979 and 1985, only to pick up again thereafter (FAO,
1992). The real price of tropical sawnwood has followed a similar trend,
although with larger fluctuations and a steeper decline in the mid-1980s
and a more rapid rise recently. The real price of other industrial timber
products, including wood-based panels, pulp, and paper and paper board,
have been increasing since the early 1970s and have sustained this rise
throughout the 1980s. The real price increases in tropical products may
reflect increasing product scarcity due to declining forest inventories and
increasing demand, whereas the recent downturn in some real tropical
product prices may reflect the depressed state of the global economy during
the late 1970s and early 1980s.
8.2.4 Regional forest products trade

Some evidence concerning the regional pattern of global forest products
trade is depicted in Table 8.2, which shows the direction of trade between
the major exporters and importers of forest products. The pattern and
direction of trade suggests that the global forest product market is still
largely dominated by developed countries, in terms of both exports and
imports. (The FAO (1995) classification of developing and developed coun-
tries is used here: the latter are Australia, Canada, Israel, Japan, New
Zealand, South Africa, USA, all East and West European countries, and the
former USSR countries; all other countries are considered to be 'develop-
ing'.) A recent review of the world forest products trade suggests that two
distinct trends have emerged (Barbier, 1996).
Firstly, the trade in forest products is highly regionalized within three
important trading blocs: the Pacific Rim, North America and Europe
(mainly Western Europe). Within each trading bloc the major importers are
mainly developed countries, such as Japan, USA, Canada and the
European Union. However, in recent years developing countries, particu-
larly in Asia, have been increasing their share of global imports. Much of
this demand reflects the increased growth in consumption of industrial
wood products in developing countries. Newly industrialized countries
with limited forest resources have also been increasing their imports of logs
and semi-finished wood products as raw materials for the export-oriented
processing industries.
Secondly, the major global exporters of forest products still tend to be
Table 8.2 Forest products direction of trade, 1993 (US$ million a )
Maior exporters
~
~ :t
~ "<:!
~
·s ·s .5"
"<:! ~
iIi:
{l '" ~ ;1, " "'-
..st;
~ ~ \.) .S :t ..Sl
-{j '" ..s "
~
·S 1i ~
i! ..s E .g ..c- ;§> ~ ~
~ ~
~ :'! ~ ";:.
J ~ ~ " -. ~ .t -.:: ;g '<l ~ ""Z <>: '<l
" -{j -{j r:;
Maior importers '"
USA 13 401 471 113
"211 474 147 123 39 71 465 122 38 20 62 451 532 16740
Japan 2436 4162 177 60 31 1777 1411 13 9 41 124 16 3 637 4 882 1497 13 280
Germany 462 572 1294 1683 45 30 1029 818 529 55 238 591 115 450 1407 255 9573
UK 515 485 1269 1432 594 86 140 427 126 144 187 265 150 216 839 297 7172
Italy 307 413 287 411 473 72 27 480 750 65 49 34 187 122 874 368 4919
France 185 226 652 511 836 29 41 168 322 73 193 245 59 377 604 331 4852
Netherlands 126 299 523 470 609 54 169 263 83 65 31 138 52 317 386 188 3773
China 157 245 54 22 34 508 365 6 17 4 12 13 3 61 6 187 1431 3125
Belgium-Luxembourg 192 243 222 256 393 107 67 435 56 74 59 134 342 0 97 91 2768
South Korea 209 662 45 20 17 565 365 13 7 6 46 11 1 2 0 324 188 2481
Canada 2039 32 6 31 24 5 28 6 17 18 30 9 2 3 20 15 2285
Hong Kong 56 256 68 24 38 252 165 21 5 14 9 40 16 0 5 416 634 2019
Spain 46 203 317 196 188 2 4 329 46 117 24 38 41 22 43 200 192 2008
Taiwan 161 318 33 55 27 375 295 16 6 9 9 10 7 0 0 242 109 1672
Austria 14 15 107 135 534 1 4 62 64 12 18 32 29 20 351 12 1410
Australia 125 254 140 24 86 58 30 26 9 62 19 37 12 0 10 420 40 1352
Thailand 58 73 21 29 17 30 279 13 2 5 20 14 18 35 2 137 463 1216
Mexico 40 982 18 12 10 33 0 7 13 3 37 3 5 0 0 19 19 1201
Switzerland 31 12 140 122 312 1 1 162 192 54 7 28 25 1 22 60 16 1186
Denmark 2 29 278 466 143 10 0 27 10 7 9 27 22 0 0 67 4 1101
Sweden 14 26 179 107 1 2 28 9 6 8 36 20 44 15 262 11 768
Turkey 15 54 58 55 88 5 0 41 25 36 1 9 7 135 10 156 62 757
Singapore 32 72 37 44 15 127 208 11 6 5 6 20 13 0 0 115 22 733
Saudi Arabia 46 89 52 92 27 139 0 8 49 13 51 10 5 0 0 33 13 627
Norway 4 8 95 316 35 7 3 7 1 3 4 15 23 5 12 40 13 591
Other developed 123 282 472 484 593 9 4 176 240 198 60 333 83 182 20 633 203 4095
Other developing 538 1382 504 446 322 365 402 246 122 136 585 349 64 38 27 672 969 7167
Total 19295 13 401 7545 7484 5771 5156 4164 3997 2814 2005 1996 1931 1924 1776 1743 9894 7975 98871
• Contains estimated and unofficial figures. Source: FAP, 1995. Based on trade data collected by the UN Statistical Office .
Table 8.4 Global tropical deforestation trends, 1980-90
Annual deforestation
Forest cover 1981-90
Number of Land area 1980 1990 million % per
Region countries (million haY (million hay (million hay ha annum
Africa 40 2236.1 568.6 527.6 4.1 0.7

West Sahelian Africa 6 528.0 43.7 40.8 0.3 0.7
East Sahelian Africa 9 489.7 71.4 65.5 0.6 0.9
West Africa 8 203.8 61.5 55.6 0.6 1.0
Central Africa 6 398.3 215.5 204.1 1.1 0.5
Tropical southern Africa 10 558.1 159.3 145.9 1.3 0.9
Insular Africa 1 58.2 17.1 15.8 0.1 0.8
Asia and Pacific 17 892.1 349.6 310.6 3.9 1.2
South Asia 6 412.2 69.4 63.9 0.6 0.8
Continental South East Asia 5 190.2 88.4 75.2 1.3 1.6
Insular South Eastern Asia 5 244.4 154.7 135.4 1.9 1.3
Pacific 1 45.3 37.1 36.0 0.1 0.3
Latin America and Carribean 33 1650.1 992.2 918.1 7.4 0.8
Central America and Mexico 7 239.6 79.2 68.1 1.1 1.5
Caribbean 19 69.0 48.3 47.1 0.1 0.3
Tropical South America 7 1341.6 864.6 802.9 6.2 0.7
Total 90 4778.3 1910.4 1756.3 15.4 0.8
Table 8.5 Trends in the incidence of trade barriers affecting forest products
Direction of movement
Import restrictions 1960s-1979 1979-1985 Since 1985
Tariff Declining Declining Declining

Tariff quota Increasing Static Static
Total prohibition Increasing Static Static/increasing
Conditional prohibition Increasing Static Static/increasing
Quota Increasing Static Static/increasing
Import licence Increasing Static/increasing Static/increasing
Import procedures NA NA NA
Variable levy NA NA NA
Anti-dumping/countervailing
duties Increasing Increasing Static
Voluntary export restraints NA NA NA
Price control NA NA NA
Standards Increasing Static/declining Increasing
Government procurement Increasing Static/declining Static/increasing
Marking and packing Increasing Static/declining Static/increasing
Export restrictions 1960s-1979 1979-1985 Since 1985
Price controls, levies, etc Increasing Increasing Increasing
Quotas, prohibitions Increasing Increasing Increasing
NA =little or no importance.
Source: Barbier et al. (1994) The assessment does not involve any weighting by the
volume of trade.
imports (85%) without duty in developed-country import markets -

almost double the proportion of imports of all other industrial goods that
have zero tariffs (GAIT, 1994). A second major contribution of the
Uruguay Round has been to reduce further the degree of tariff escalation
faced by forest products in developed-country markets.
The implications of the Uruguay Round for the non-tariff barriers
increasingly faced by the tropical timber trade are less clear. However, two
special agreements, the Agreement on the Application of Sanitary and
Phytosanitary (SPS) Measures, and the Agreement on Technical Barriers to
Trade (TBT), do provide the basis for tackling certain non-tariff measures
that have been used as trade barriers against forest products. The SPS
agreement could reduce the use of inspection, quarantine and treatment of
imported forest products as prohibitive measures beyond what is necessary
to protect domestic human, animal and plant populations from pests or
diseases. The TBT agreement could limit the use of technical regulations on
forest products as non-tariff restrictions rather than for legitimate purposes of
protecting human health and safety, preventing environmental degradation
and ensuring adequate product quality and design standards.
Table 8.6 Uruguay round tariff reductions in developed countries of industrial productsa
Imports all sources Imports from developing markets

Import Import
value Pre-UR Post-UR Reduction value Pre-UR Post-UR Reduction
Product category (US$bn)b Tariff (%) Tariff (%) (%) (US$bn)b Tariff (%) Tariff (%) (%)
Fish and fish products 18.5 6.1 4.5 26 10.6 6.6 4.8 27
Forest productsC 40.6 3.5 1.1 69 11.5 4.6 1.7 63
Textiles and clothing 66.4 15.5 12.1 22 33.2 14.6 11.3 23
Leather, rubber, footwear and
travel goods 31.7 8.9 7.3 18 12.2 8.1 6.6 19
Metals 69.4 3.7 1.4 59 24.4 2.7 0.9 67
Chemicals and photographic
supplies 61.0 6.7 3.7 42 8.2 7.2 3.8 47
Transport equipment 96.3 7.5 5.8 23 7.6 3.8 3.1 18
Non-electric machinery 118.1 4.8 1.9 58 9.8 4.7 1.6 66
Electric machinery 86.0 6.6 3.5 47 19.2 6.3 3.3 48
Mineral products and precious
stones/metals 12.9 2.3 1.1 52 22.2 2.6 0.8 69
Manufactured articles n.e.s. 76.1 5.5 2.4 56 10.9 6.5 3.1 52
Industrial tropical products 32.8 4.2 2.0 52 14.4 4.2 1.9 55
Natural resource products 80.2 3.2 2.1 34 33.4 4.0 2.7 33
All industrial products
(excluding petroleum) 736.9 6.3 3.8 40 169.7 6.8 4.3 37
UR, Uruguay Round; n.e.s., not elsewhere specified.
Tariffs based on weighted averages on imports from all sources.
a Excluding petroleum products.
b Import data from 1990 or latest available year, and for all developed country participants in the Uruguay Round, excluding imports from free-trade
area partners (e.g. intra-European Union trade) and from contractural preferential arrangements.
C Wood, pulp, paper and furniture.
Source: GAIT, 1994.

Other provisions of the Uruguay Round that may also help to reduce
barriers to the tropical timber trade include limitations and clarifications
on the use of anti-dumping and countervailing duties, customs valuation
and licensing procedures, and market access restrictions. A possible indi-
rect impact of the Uruguay Round is that it may have prompted other
long-standing non-tariff barriers in individual markets to be lessened. For
example, the European Union has recently proposed eliminating for
1995-96 its plywood tariff/quota scheme for GSP beneficiaries.
A recent estimate has been made of the likely impacts of the Uruguay
Round tariff reductions on the forest products trade for selected products
and markets (Barbier, 1995, 1996). The results of the analysis indicate that
the total trade effects of the Uruguay Round tariff changes on selected for-
est products may range from US$340- 472 million in key developed- and
developing-country markets (Table 8.7). In comparing the two scenarios
depicted in Table 8.7, it is clear that Scenario B offers the more globally
Table 8.7 Estimated trade effects on selected forest products of Uruguay Round
tariff changes in key developed and developing country markets (1991 trade levels,
US$1000)
Developing country imports
Trade Creation Trade Diversion Total Effect
Importer A B A B A B
Australia 2042 5073 -256 o 1786 5073

Canada 19 689 -77 o -58 689
European Union o 65746 -201756 o -201 756 65746
Japan 36701 74759 -16383 o 20318 74759
New Zealand o 128 -3 o -3 128
South Africa" 36 36 o o 36 36
United States o 24472 -604 o -604 24472
All developed countries 38798 170902 -219079 o -180281 170902
China 4402 4402 o o 4402 4402
India 303 303 o o 303 303
Republic of Korea 1402 1402 o o 1402 1402
Malaysia 569 569 o o 569 569
Thailand 5337 5337 o o 5337 5337
All developing countries 12013 12013 o o 12013 12013
All counries 50811 182915 -219079 o -168268 182915
Scenario A assumes a medium elasticity of demand for forest product imports; pre-Uruguay
Round tariffs on imports in developing country markets are at MFN rates; and pre-Uruguay
Round tariffs on imports in developed country markets are at GSP rats for developing
country imports and MFN rates for developed country imports. Scenario B assumes a
Market access and trade barriers 215
beneficial results for the Uruguay Round tariff reductions for the forest
products and importing markets analysed. In Scenario A, although there is
a net global gain of $340 million in the forest products trade, developed-
country exporters are the overwhelming beneficiaries from both the trade
creation and diversion effects, whereas developing countries are net losers.
In Scenario B, not only is the net gain in global forest products trade larger,
at around $472 million, but also both developed and developing countries
benefit from the trade expansion - although the former gain by almost
60% more in relative terms. The actual post-Uruguay Round outcome is
likely to lie somewhere between Scenarios A and B. Although prior to the
agreement not all developing country forest product imports into devel-
oped-country markets may have benefited from GSP rates, a number of
these imports have. In addition, as will be discussed in the following sec-
tion, recent evidence suggests that tropical timber products in importing
markets are subject to strong substitution effects from imports of compet-
ing products (e.g. temperate wood products) and from different sources of
Table 8.7 Continued
Developed country imports Percentage

of forest
Trade Creation Trade Diversion Total Effect Total change products imports
A B A B A B A B A B
10871 10871 256 0 11127 10871 12913 15944 1.1% 1.3%

3596 3596 77 0 3673 3596 3615 4285 0.2% 0.2%
198591 198591 201756 0 400346 198591 198591 264336 0.4% 0.6%
37783 37783 16383 0 54166 37783 74484 112 543 0.6% 0.9%
560 560 3 0 563 560 560 689 0.3% 0.4%
30 30 0 0 30 30 66 66 0.0% 0.0%
24317 24317 604 0 24921 24317 24317 48788 0.2% 0.4%
275748 275748 219079 0 494827 275748 314546 446650 0.4% 0.5%
3570 3570 0 0 3570 3570 7972 7972 0.2% 0.2%
651 651 0 0 651 651 954 954 0.2% 0.2%
8486 8486 0 0 8486 8486 9889 9889 0.4% 0.4%
146 146 0 0 146 146 715 715 0.1% 0.1%
827 827 0 0 827 827 6164 6164 0.5% 0.5%
13 682 13 682 0 0 13 682 13 682 25694 25694 0.3% 0.3%
289430 289430 219079 0 508509 289430 340240 474345 0.4% 0.4%
medium elasticity of demand for forest product imports, and pre-Uruguay Round tariffs on
imports in all markets are at MFN rates.
a No GSP rates for developing country imports in pre-Uruguay Round tariffs.
Source: Barbier, 1996.
origin (e.g. developed countries). If both conditions hold, then some trade
diversion away from developing country imports is likely to occur -
although the magnitude of the diversion effect as represented in Scenario A
is probably an over-estimate. Thus the estimates provided by Scenarios A
and B can be considered a lower and upper bound, respectively, on the
impacts of the tariff reductions in the Uruguay Round on the forest prod-
ucts trade for the commodities and markets analysed. See Barbier et al.
(1994) for a review and further discussion of these studies.
However, these impacts amount to only 0.4% of total 1991 forest prod-
ucts imports in the markets analysed, which had an aggregate value of
US$8S.6 billion. This suggests that, although the real trade gains from the
tariff changes are positive and significant, they may not have a substantial
impact on the global forest products trade. One factor limiting the gains
from the Uruguay Round for forest products is that the pre-Uruguay tariff
rates for most of these products in major importing markets were already
very low.
Perhaps more important than the estimated gains in trade for forest
products, now that the Uruguay Round has led to international commit-
ments to reduce forest product tariff rates in major markets, it is unlikely
that tariff rates will be increased unilaterally. The new and often lower
rates in developed-country markets will be bound, as will an increasing
proportion of tariffs in developing-country markets. Binding a tariff means
that a country has agreed officially to a ceiling on the tariff rate.
8.3.3 'New' barriers to forest products trade

In recent years there has been a proliferation of additional policies and reg-
ulations that have the potential of becoming 'new' barriers to the forest
products trade. These barriers include:
• export restrictions by developing countries to encourage domestic pro-
cessing of tropical timber for export;
• environmental and trade restrictions on production and exports in
developed countries that affect international trade patterns;
• quantitative restrictions on imports of 'unsustainably produced' timber
products;
• the use of eco-Iabelling and 'green' certification as import barriers.
Although only the last two measures could be strictly defined as new, all of
these trade measures have been increasingly employed in recent years and
have the potential to affect tropical timber trade flows significantly. The
issue of certification will be discussed in more detail in the next section.
The first three barriers are briefly discussed below.
Developing countries are continuing to use export restrictions on wood
in rough and semi-processed products to support domestic processing
industries and improve export prospects for higher-valued forest products.
Several authors have recently reviewed the role of export taxes and bans in
encouraging forest-based industrialization and sustainable timber manage-
ment in tropical forest countries - see for example Barbier (1994); Barbier
et al. (1994, 1995); Constantino (1990); Gillis (1990); Vincent and Binkley
(1991); Vincent (1992).The general conclusion is that tropical timber
export taxes and bans have proved only moderately successful in achieving
the desired results in South-East Asia. For example, although expanded
processing capacity was established in Malaysia, the Philippines and
Indonesia, it was achieved at high economic costs, both in terms of the
direct costs of subsidies as well as the additional costs of wasteful and inef-
ficient processing operations.
Despite the losses in terms of economic inefficiencies and the implica-
tions for the management of their forest resource base, developing coun-
tries are unlikely to end such policies but may instead employ them more
extensively. Many log-producing countries see the use of export taxes and
bans as the means to compensate domestic processing industries for import
barriers faced in developed-economy markets. However, with the post-
Uruguay Round decline in tariff escalation and barriers generally in import
markets for forest products, this argument is less valid.
Developed countries are increasingly employing a variety of environmen-
tal regulations in their forest industries - both alone and in conjunction
with export restrictions - that may have significant trade implications.
Whether or not such regulations are being used intentionally for this pur-
pose, they may lead to trade distortions and discrimination. For example,
the combination of trade and environmental restrictions on logging in the
Pacific Northwest of the USA (such as the spotted owl reservations coupled
with the state-level log bans) produced significant domestic and global
trade impacts, including increases in global sawlog prices and regional
shifts in production with related effects in major sawnwood and plywood
markets (Flora and McGinnis, 1991; Perez-Garcia, 1991).
In most cases, the lack of international consensus and agreement on
environmental criteria and standards lies at the heart of disputes as to
whether unilaterally imposed measures are restricting market access. For
example, Brazilian manufacturers and exporters of pulp have expressed
their concern that the criteria developed by Denmark would largely benefit
European paper producers at the expense of foreign exporters of pulp and
paper products (Jha and Zarrilli, 1994). Potential problems exist with
other environmentally oriented regulations, such as the increasing restric-
tions on trade in wood panels using formaldehyde glue; regulations ban-
ning or controlling certain timber preservation processes and materials;
and controls on processing materials such as the use of chlorine in bleach-
ing pulp (Bourke, 1995).
Many developed countries are also under pressure to adopt quantitative
restrictions to limit the import of 'unsustainably' produced forest products
or to impose countervailing duties on imported products that benefit from
an environmental export subsidy - i.e. 'unsustainable' forest management
that leads to lower harvesting costs and thus lower export product prices.
The arguments against imposing unilateral sanctions, such as singling out
tropical timber products for trade bans or other restrictions on environ-
mental grounds, are fairly formidable and now generally accepted (Barbier
et al., 1994). However, the establishment of the Trade and Environment
Committee of the World Trade Organization (WTO) has shifted the debate
over environmental restrictions on trade to the multilateral arena
(Bhagwati, 1995; Eglin, 1995).
Although the rationale behind WTO's work programme on trade and
environment is to link the benefits of trade liberalization more directly to
policies to achieve better environmental protection and sustainable devel-
opment, the popular perception is that the real objective should be
'Greening the GATT', i.e. subjugating trade rules to environmental criteria.
This in turn has led to repeated calls to amend GATTIWTO rules to allow
for quantitative restrictions and countervailing duties where there are envi-
ronmental justifications for employing these measures (Arden-Clarke,
1991). If such a broadening of the GATTIWTO rules is achieved, then the
use of quantitative restrictions and even countervailing duties by importing
countries, e.g. to restrict imports of timber products that are not 'sustain-
ably' produced or to counter perceived environmental export subsidies,
will increasingly become a feature of the forest products trade.
Although there are legitimate uses of all the trade policy measures dis-
cussed above, the rate at which they are being implemented and the fre-
quency with which they have led to trade distortion and discrimination
suggests that their use must be examined carefully. International agree-
ments and rules governing their use should also be negotiated, and the
interface and possible conflicts between multilateral environmental agree-
ments and trade rules need to be explored through the auspices of the
WTO. What clearly needs to be avoided is indiscriminate and widespread
application of 'new' barriers to the forest products trade that could easily
override the gains in market access resulting from the recently concluded
Uruguay Round.
8.3.4 The relative competitiveness of tropical timber products

In addition to trade barriers, the degree to which forest products from dif-
ferent regions compete among themselves and with non-wood substitutes
for import markets is an important determinant of the long-run returns to
forest products. Changes in long-run returns may in turn influence the
incentives for sustainable forest management.
The degree of substitution between tropical and temperate products in
consumer markets illustrates the extent to which the markets for these two
types of product are inter-related, or whether there are essentially two dif-
ferent markets for two distinct commodities. The available evidence indi-
cates that the elasticities of substitution between temperate and tropical
wood products are very low (Barbier et ai., 1994; Constantino, 1988; NEI,
1989). This suggests that there are two distinct markets, and tropical pro-
ducers of these products would have difficulty in penetrating the larger
temperate market. That is, temperate softwoods from different regions are
still closer substitutes than softwoods and tropical hardwoods; equally, for
tropical timber products there are strong substitution effects between prod-
ucts from different tropical regions or countries. In general, substitution by
origin for tropical sawnwood and plywood in certain importing countries
appears to be very high, especially for plywood. There is also evidence that
in some major processing markets, imports of tropical logs are subject to
substitution by domestic softwood logs and by technical change (Vincent et
ai., 1991; Youn and Yum, 1992).
Tropical timber products may also be substituted by non-wood products
in end uses and final markets. Although there is increasing anecdotal evi-
dence of this occurring in many consumer markets, particularly in con-
struction and furniture industries, estimating the magnitude or scale of this
effect has proven more difficult. However, for specific products this substi-
tution effect may be significant (NEI, 1989). For example, plywood is
believed to face severe competition from solid synthetic panels, with price
strongly influencing the choice of product in the construction industry. In
addition, substitution may be more of a problem for wood-based compos-
ites such as particle board, fibreboard and reconstituted panels, and wood
pulp.
To summarize, empirical studies suggest that substitution between tropi-
cal and temperate timber products in importing markets has not been very
significant. However, in response to export log bans by tropical producers,
some importers are increasingly diversifying the source of their supply.
Substitution of non-wood products for timber may be occurring, but the
evidence is largely anecdotal. Substitution between tropical timber prod-
ucts originating from different countries or regions does appear to be very
high, particularly for plywood. This would suggest that importers can sub-
stitute between sources of origin with relative ease but also that exporters
can easily capture market share through price competition.
8.3.5 Lesser-used species

Provided that markets exist, there is substantial potential for expanding
utilization of existing tropical production forest resources by exploiting
commercially lesser-used species. For example, recent FAO statistics sug-
gest that only about 26% of the potential standing volume in tropical
harvest areas is being felled. In general, the forest products industry is
based on the utilization of large sawlogs and logs for plywood and veneer.
Over the 1988-92 period, saw and veneer logs comprised 92 % of the total
industrial roundwood harvest in Indonesia, 97% for Malaysia and 93%
for Papua New Guinea (Drake et aI., 1995).
If lesser-used species are to play an expanded role in the tropical forest
products industry, it is unlikely that this additional supply would be suit-
able for traditional solid-wood products such as sawnwood and plywood.
Instead, the likely use of lesser-used species would be for wood pulp and
reconstituted wood products such as fibreboard, particle board and recon-
stituted panels. At least one set of projections for the Asia-Pacific region
suggests that the potential gains to producers of developing capacity to
manufacture composite panel products and other engineered wood prod-
ucts might be substantial, and could offset the negative impacts on the for-
est products industry of declining supplies of forest resources for
sawnwood, plywood and veneer (Drake et aI., 1995).
However, there are important considerations for any global strategy to
promote lesser-used species. Firstly, as discussed above, reconstituted wood
products and wood pulp are currently some of the most competitive and
volatile markets for timber products. Reconstituted wood panels are highly
susceptible to substitution in import markets by semi-wood composites,
such as cement fibreboard, composites made from agricultural and other
recycled wastes, and a variety of non-wood substitutes. Any new form of
virgin wood pulp would be competing in terms of quality and price with
more traditional sources, but also with recycled pulp. As noted above, the
problem of market access for any new products generated from lesser-used
species is further exacerbated by the increasing proliferation of environ-
mental, health and other regulations specifying the composition and qual-
ity of both reconstituted wood and paper products in consumer markets.
Secondly, in many tropical countries the potential for identifying, let
alone exploiting, lesser-used species is limited by lack of basic information
on the availability and commercial viability of these species. This in turn
reflects the limited human and technical resources devoted to inventorying
and assessing the forest resource base of the timber industry. Forest
resource accounting has been promoted as a means for tropical producer
countries to assess their forests regularly, including the identification of
lesser-used species for potential commercial exploitation through sustain-
able forest management (Kemp and Phantumvanit, 1995). Basic assessment
of this kind is essential if lesser-used species are to be utilized more in the
tropical forest products industry.
Thirdly, although it is generally assumed that exploitation of lesser-used
species is advantageous as it would increase stand harvesting yields and
thus reduce per unit costs, and possibly lower regeneration costs by removing
more of the residual stems, greater utilization of the standing volume of
timber is not always compatible with sustainable forest management. A
critical issue is whether there is a considerable loss of forest services, such
as biodiversity and watershed protection, if more species and hence stems
are removed. Clearly, more needs to be learned about the potential ecologi-
cal effects on tropical forests of increasing the rate of utilization of lesser-
used species and the resulting impacts of greater stem removal from stands.
Again, forest resource accounting might be a first step in determining the
capacity for exploiting lesser-used species through sustainable management
practices. However, further analysis of market demand and the costs of
exploitation and utilization is required to determine the commercial viabil-
ity of these species.
8.4 CERTIFICATION AND LABELLING

The number of eco-labelling and certification initiatives applied to the for-
est products trade has increased rapidly in recent years. Generally, the aim
of these initiatives is to distinguish 'sustainably' produced forest products
or to ensure that forest product imports conform to domestic environmen-
tal standards and regulations. As noted in the previous section there is con-
siderable concern, particularly among tropical producer countries and
forest-based industries, that certification and labelling will be used as non-
tariff barriers limiting access to key import markets. To avoid being
employed as trade barriers, such regulations and schemes must be non-
discriminatory, transparent and justified. They also should be agreed mutu-
ally between trading partners or through multilateral negotiations, comply
with GAIT rules and conform with internationally recognized guidelines.
The following section examines the possible role of voluntary and non-
discriminatory certification in promoting sustainable forest management in
tropical regions.
8.4.1 Timber certification

The term certification has been used indiscriminately to cover a wide range
of processes. In this document the term timber certification will be used to
mean a process which results in a written statement, i.e. a certificate, attest-
ing to the origin of wood raw material and its status and/or qualifications,
often following validation by an independent third party (Baharuddin,
1995). To be effective in reassuring consumers that wood products origi-
nate from sustainably managed sources, timber certification requires both
certification of the product process and certification of the sustainability of
forest management practices. The latter requires verification of the forest
management system in the country of origin, including the environmental
and social impacts of forestry practices, against specified sustainable man-
agement criteria and standards. The former involves inspection of the
entire product processing chain of supply from the forest to final product,
through domestic and export markets if necessary.
Proponents of certification argue that it can potentially assist in promot-
ing sustainable forest management while simultaneously reassuring con-
sumers. A properly designed, voluntary and independently accredited
certification scheme at the global level can be a means by which the various
stakeholders can hold producers accountable; it can provide a market-
based incentive to the individual producer to improve management; it can
meet consumer demands for wood from well-managed forests without
creating trade discriminations; and it can be a mechanism for monitoring
multiple factors involved in forest use (Dubois et aI., 1995).
However, others suggest that the evidence for considerable additional
demand for certified wood products is unproven, and then only in certain
small niche markets may customers be willing to pay more for certified
timber (Varangis et aI., 1995). In fact, there is concern that the impacts of
certification on production and distribution costs might reduce the compet-
itiveness of wood products in consumer markets. It is also argued that,
although certification requires sustainable forest management as a neces-
sary prerequisite, implementation of sustainable forest management does
not require certification to take place (Kiekens, 1995). The promotion of
certification globally should not either displace or divert resources from
ongoing efforts in the major timber supplying countries to implement
national forest policies, regulations and standards in accordance with inter-
national and national commitments to sustainable forest management. It is
argued that the necessary but stringent conditions required for an accred-
ited global certification scheme are bound to have only a limited impact on
a small proportion of global timber production, and equally, on the sus-
tainable management of a limited area of forests (Baharuddin and Simula,
1994; Kiekens, 1995).
Despite the proliferation of certification schemes, the evidence to date
suggests that certified timber is currently having only a limited impact on
the global market. In 1993, around 1.5 million m3 of timber products and
35 suppliers were certified. This amounted to less than 0.5% of the global
trade (Baharuddin and Simula, 1994). Recent estimates of the total global
forest area certified are depicted in Table 8.8. Total production amounts to
around 3.5 million m 3 from about 5.1 million ha of certified forests. In
fact, certified production accounts for only 0.23 % of the world's industrial
roundwood production. It is unlikely that the supply of certified timber
and timber products will expand very fast. Even under optimistic projec-
tions, only 15% of traded wood products are expected to be affected by
certification by the year 1999 (Baharuddin and Simula, 1996).
Certification and labelling 223
Table 8.8 Certified global forests a
Percentage of
Production Area of total Percentage of
Region (m 3) (ha)b productionC total area d
Africae NA 6000 NA 0.45%

Plantation NA 6000 NA
North America 502,408 617,000 0.09% 0.14%
Natural 502,408 617,000 0.14%
Latin America f 101,350 672,940 0.12% 0.10%
Natural 101,350 597,000 0.09%
Plantation NA 75,940 1.05%
Asia g 738,810 2,847,394 0.85% 1.66%
Natural 1310 12,610 0.01%
Plantation 737,500 2,834,784 31.8%
Europeh 2,135,000 916,720 9.7% 8.43%
Semi-natural 2,135,000 916,563 8.43%
Plantation NA 157 NA
World Total 3,477,568 5,060,054 0.23%i 0.14%i
N aturaUsemi-natural 2,740,068 2,143,173 0.06%
Plantation 737,500 2,916,881 10.12%
NA, not applicable or not available.
a Based on Baharuddin and Simula (1996), unless specified otherwise.
b In some cases includes non-forest land as well.
e Percentage of total industrial roundwood production in 1993 for the countries specified,
from FAO (1995).
d Percentage of total forest area, natural forest area and plantation area for the countries
specified, from WRI (1994).
e Certified forest is located in South Africa.
f Certified forests are located in Costa Rica, Brazil, Guyana, Honduras and Mexico.
g Certified forests are located in Indonesia, Malaysia, Papua New Guinea and Solomon
Islands.
h Certified forests are located in Poland and the United Kingdom.
i Total production from certified forests expressed as a percentage of total world industrial
roundwood production in 1993, from FAO (1995).
j Total certified forest, naturaUsemi-natural forest and plantation areas expressed as a
percentage of total world forest, natural forest and plantation areas in 1980, from WRI
(1992). Total world plantation area excludes plantation area from Europe, North America
and the former Soviet Union.
Nevertheless, from recent experience with timber certification it is possi-

ble to obtain a better idea of the benefits and costs involved. Firstly, certified
timber sales in export markets are expected to attract a 'green premium' in
consumer markets: that is, certified timber could be sold at a higher price
than uncertified timber. Secondly, in the absence of certification some tim-
ber products may lose substantial market share in consumer markets of
importing countries that are currently developing legislated or voluntary
schemes, and even quantitative restrictions and environmental regulations
on imports, which will ultimately affect non-certified timber products (see
section 8.3 above). Thus one of the additional benefits of certification is
that it will allow timber exporters to avoid losses of market share and rev-
enues.
An attempt to estimate these benefits for tropical timber certification has
been conducted by Varangis et al. (1995). The results are depicted in Table
8.9. A key assumption of the analysis is that certification of tropical timber
products would attract a green premium of 10% in certain niche markets
in North America and Europe. This leads to a gain of US$62 million. On
the other hand, in the absence of certification, these markets would be lost
to uncertified tropical timber products, although some of this loss could be
recouped from diversion of these exports to non-EuropeanINorth
American markets. Thus the avoidance of net revenue losses without cer-
tification amounts to US$366 million. The total gain from certification of
tropical timber products is US$428 million, or 4% of current developing-
country timber product exports. An interesting aspect of this estimate is
that the vast majority of the gains from certification arise from avoiding
losses in markets and revenues in the absence of certification, and not from
the additional gains of any green premium - despite the generous assump-
tion that the latter might be as high as 10% and with no substitution
effects.
Such an estimate is of course not without controversy. Although there is
now some evidence from surveys that there is a 'willingness to pay' a
higher price for certified products in certain importing markets, the actual
size of such a premium is debatable. More importantly, the higher the pre-
mium and thus final product price, the greater the likelihood of substitu-
tion by competing non-wood products and domestic wood-based
substitutes (see section 8.3 above). The estimated gains in Table 8.9 do not
take into account such substitution effects in importing markets.
In addition, there is considerable debate about the size of the niche mar-
kets ultimately affected by certification. Table 8.9 indicates that, in the
absence of tropical timber certification, the markets affected would result
in a revenue loss of around 6% of developing-country export markets.
However, this would amount to only 0.64% of the global forest products
trade being affected. Others have suggested that if both tropical and timber
products were certified, more markets would certainly be affected, and as a
result around 15-25% of the total share of global forest trade could be
influenced by certification (Baharuddin and Simula, 1996). However, it is
not clear on what basis the latter estimates are made. The ultimate impact
on revenues and markets is also not clear. As certification is extended to
Table 8.9 Estimated revenues from tropical timber certification in European and
North American markets
Export Share of
revenues developing
gained country wood
Source of revenue (US$ million) exportsa
(1) Incremental revenue from European and

American 'niche' markets b 62 0.58%
(2) Avoidance of net revenue losses without
certification (a-b) 366 3.43%
(a) Loss of European and North
American markets C 622 5.83%
(b) Gains from diversion to other
markets d 256 2.40%
(3) Total net revenue gain (1+2) 428 4.02%
Based on 1991 trade flows and values.

a Developing countries excluding China, Argentina, Chile and Near-eastern countries. Total
value of forest product export revenues in 1991 estimated at around US$10.66 billion, of
which US$9,02 billion from non-coniferous logs, non-coniferous sawnwood and wood-
based panels and US$1.64 billion from furniture and processed wood products.
b Assumes a 10% increment in revenues due to a 'green premium' from certification affecting
10% of the North American and 20% of the European markets.
C Assumes that the markets in footnote a will be lost in the absence of certification.
d Assumes that developing country exports would increase by 8.8% to non-EuropeanINorth

American markets in the absence of certification.
Source: Varangis et al., 1995.
more temperate and tropical products, then the green premium differential
between certified and non-certified wood products will be substantially
smaller and may eventually disappear. However, if global certification
results in higher prices for wood products generally in consumer markets,
then there may be a problem of substitution by non-wood alternatives on a
large scale.
The issue as to whether timber certification inevitably leads to higher
prices for timber products in final consumer markets is also controversial,
and relates to the evidence concerning the overall costs of certification. It is
useful to distinguish two costs: the direct costs of certification in terms of
implementing such schemes, and the indirect costs of certification through
any trade losses and diversion in final consumer markets as a result of sub-
stitution between certified and non-certified products.
The direct costs of certification include both the costs of assessing or
auditing forest management practices in the country of origin, and the
costs of identifying, monitoring and assessing the entire processing chain of
supply from forest to final product. These costs will clearly vary depending
on the type and size of forest being certified, the type of final product being
produced, and the location of processing activities and their degree of verti-
cal integration both domestically and internationally. For example, for
tropical products the costs of assessing or auditing have been estimated at
about US$O.3 and US$llha/year in developing countries, and the costs of
certifying the chain of supply for processing could be up to 1 % of border
prices (Baharuddin and Simula, 1994). Dubois et al. (1995) suggest that, as
a rough approximation, the minimum costs of certifying forest manage-
ment would be a fixed assessment cost of US$500 plus US$OAOlha for the
initial assessment, and US$O.15Iha for each subsequent visit.
Many analysts suggest that an additional direct cost of certification is the
incremental cost of improving current forest management practices to sat-
isfy the criteria and standards specified to obtain certification. These costs
are likely to be considerably higher than the costs of the certification
process itself (Dubois et aI., 1995). Much of this cost is assumed to result
from reduced timber yields and higher operating costs, although the latter
could actually be reduced by better planning, improved residual stands,
and reduced-impact harvesting techniques (Baharuddin, 1995).
However, to the extent that the costs of improved forest management are
really to do with the transition of forest practices to sustainable forest
management and not certification per se, then it is not completely accurate
to attribute all these additional costs to certification. Instead, it is probably
best to turn this viewpoint around: the costs of assessing and auditing
forests and the processing chain of supply from forest to final product are
the additional direct costs of certifying sustainably produced timber, on top
of the costs of transforming existing forest practices to sustainable forestry
(see section 8.5 below).
There may also be indirect costs of certification as a result of higher
prices for certified tropical timber products in import markets. As dis-
cussed above, these costs are related to the size of price differentials for cer-
tified products, substitution effects, and the extent of products and markets
affected. So far, the available evidence suggests that the number of wood
products and markets affected by certification will remain fairly small for
the foreseeable future. Moreover, as will be discussed in the next section,
higher sustainable tropical forestry costs plus any additional certification
costs in the country of origin may not necessarily translate into substan-
tially higher final product prices in importing markets. On the other hand,
the analysis of the impacts of Uruguay Round tariff reductions in section
8.3 suggests that only minor changes in prices for wood products in import
markets are required for trade creation (or loss) and diversion to occur.
Thus there are likely to be some trade losses and diversion for certified tim-
ber products in importing markets - although the precise magnitude of
these indirect costs of certification is difficult to determine at present.
Importantly, there is now an emerging international consensus that an
adequate international framework is needed both to ensure harmonization
and mutual recognition of certification systems, and to ensure an effective
international accreditation process of certification bodies. The important
criteria for any internationally accredited certification body is that it is
independent, impartial, and able to demonstrate that its organization and
personnel are free from any commercial, financial or other pressure
(Dubois et aI., 1995). Equally, to achieve harmonization and mutual recog-
nition, a voluntary international certification system must (Baharuddin and
Simula, 1996):
• be comprehensive and cover all types of forests and wood products -
not just tropical products;
• be based on objective and measurable criteria;
• produce reliable assessment results and thus be fully independent from
any vested interests;
• be transparent and involve a balanced participation of the interested
parties and stakeholders, thereby ensuring their commitment;
• represent all involved parties;
• be goal-oriented and cost-effective.
Establishment of such an international framework covering all existing and
proposed timber certification schemes is clearly a long-term process. The
purpose of any agreed timber certification rules should be to reinforce the
positive incentives for sustainable forest management and not to penalize
or restrict production and trade in timber not meeting standards.
8.4.2 Country certification

Given the proliferation of certification schemes and the increasing threat of
wood products being targeted for quantitative restrictions and other mar-
ket barriers in major consumer markets, the need to develop internation-
ally agreed, accredited and transparent timber certification rules is great.
However, it is important to emphasize that timber certification is currently
influencing only a very small proportion of the global trade in forest prod-
ucts and an equally limited area of the world's production forests. Even
under very optimistic scenarios for the expansion of certification schemes,
this is unlikely to change in the foreseeable future. Thus timber certification
cannot be considered the main instrument for promoting sustainable forest
management globally; in fact, given the pressing need to promote sustain-
able forest management, it is imperative to develop urgently other instru-
ments complementary to timber certification that are more directly aimed
at wholesale improvements in forestry management policies and regula-
tions in producer countries.
One such approach is the concept of country certification. Originally
proposed in a report to lITO (Barbier et aI., 1994), country certification
involves certifying through explicit bilateral or multilateral recognition all
timber products from a country that can prove it is complying with an
internationally agreed objective, such as a sustainable forest management
target. Such a scheme could be enacted for all timber producer and con-
sumer countries through an international agreement on global forests. As
discussed further in section 8.6, it may require additional assistance for
poorer countries with inadequate financial resources to achieve, implement
and monitor the key policy objectives.
The main purpose of country certification is to ensure an international
commitment by both producer and consumer countries to adopt policies
and practices towards encouraging sustainable management of production
forests and timber products while simultaneously improving international
market access of these products. The overall objective is to tackle the two
major obstacles to sustainable forest management identified in this chapter:
restrictions on market access (see section 8.3) and forest-sector policy fail-
ures (section 8.5). It would also hopefully assist with the problem of mar-
ket transparency (see section 8.6). To be effective, country certification
would require two broad sets of policy commitments, from timber-produc-
ing and timber-consuming countries.
The first set of policies would require producer countries to undertake
substantial reviews of their forest-sector policies to determine the implica-
tions of their existing domestic forestry policies and regulations on timber-
related deforestation, and the extent to which their timber export policies
may also be affecting deforestation, either directly or through exacerbating
problems caused by poor domestic forestry policies and regulations.
Producer countries ought to correct those policy distortions that work
against sustainable timber production objectives, as such distortions are
believed to be at the heart of inefficient and unsustainable forest sector
development and timber-related deforestation (see section 8.5).
The second set of policies would require a commitment by consumer
countries to remove any remaining tariff and non-tariff barriers to timber
imports into domestic markets, particularly for those producer countries
that demonstrate a commitment to forest sector policy reform. For exam-
ple, the removal of specific tariff and non-tariff barriers on forest product
imports could take place on a case-by-case basis, depending on demonstra-
ble progress by each exporting country in promoting sustainable forest
management policies and forest sector policy reform. This could occur
through normal bilateral trade negotiations or through multilateral agree-
ments and organizations. In addition, consumer countries should actively
promote, through information and market intelligence campaigns, the use
of tropical timber imports from exporting countries that are implementing
sustainable management policies. Consumer countries should also under-
take not to resort to the use of any of the 'new barriers' (identified in sec-
tion 8.3) to the imports from participating producer countries.
With its emphasis on policy reform, country certification is not necessarily
an alternative to timber certification but is complementary to it. The wide-
spread adoption of sustainable forest management and practices in pro-
ducer countries would make it much easier for more forests and timber
products to qualify for timber certification. The global impact of existing
timber certification schemes and their coverage of forests and products
would probably increase. Moreover, implementation of a global country
certification process could initiate increased cooperation between producer
countries, consumer countries, timber traders and independent monitors to
develop more internationally recognized, transparent and detailed criteria
for establishing and evaluating forest management practices globally. As it
may be easier to negotiate and implement, not only may country certifica-
tion lay the groundwork for more comprehensive international timber cer-
tification agreements, but also it may 'buy time' by initiating immediate
incentives for action on sustainable forest management while the details of
more complicated agreements on timber certification for all global forest
products from natural and plantation forests are concluded.
As with timber certification, any country certification scheme needs to be
voluntary and internationally agreed. If poorly implemented without suffi-
cient international transparency, recognition or commitment, a country
certification scheme would have little impact on improving sustainable for-
est management globally. It would neither take advantage of the trade-
related incentives needed for encouraging sustainable management of
forests, nor provide the stimulus for fostering further cooperation in
related areas, such as timber certification. A country certification process
that encourages further cooperation by major tropical timber-exporting
countries, while at the same time being sufficiently comprehensive to cover
major temperate producers, would be an important objective. However,
where technical and financial constraints limit the ability of some poorer
tropical producers to implement forest-resource assessments, management
regulations and policy reforms, then additional financial assistance may be
required for these countries as part of a global scheme (see section 8.6).
8.5 THE COSTS OF IMPLEMENTING SUSTAINABLE FOREST

MANAGEMENT
It is generally assumed that most of the world's tropical production forests
are not being harvested on a sustainable basis, and that the cost of timber
produced under sustainable forest management will be generally higher
than for timber produced under current, less sustainable practices.
Although this appears to be a reasonable assumption, given the harvesting
regulations and management practices that need to be implemented to
improve the sustain ability of tropical forest operations, there is generally
little discussion as to why 'unsustainable' practices are occurring in the first
place, and what are the likely impacts of such practices on tropical forest
systems and services. To understand more fully the rationale for internaliz-
ing the costs of sustainable tropical forest management, it is necessary to
determine the exact nature of these costs and why they are not normally
'internalized' under current forest management practices in the tropics.
Only then is it possible to determine how best to 'correct' the economic
incentives that are continuing to encourage unsustainable forestry practices
and what the likely magnitude of these costs might be.
8.S.1 The costs of unsustainable timber extraction

Unsustainable extraction of timber from tropical forests may not only lead
to a decline in standing timber stocks but may also have wider environ-
mental effects. These external effects include the loss of other consumptive
uses (e.g. harvesting and hunting other forest resources, and recreational
uses), the loss of ecological functions (e.g. watershed protection, carbon
storage and microclimatic role) and the loss of other non-consumptive val-
ues (e.g. ecotourism, genetic or biochemical information, and 'existence'
values) of the tropical forest. The production of timber can therefore affect
the environment, both:
• directly through the removal of trees and other damage incurred to sur-
rounding forest during timber extraction; and
• indirectly through opening up and improving access to the forests which
has an impact on other socio-economic factors which may degrade the
environment.
An obvious direct economic cost of unsustainable tropical forest manage-
ment and timber-related deforestation is that the future standing stock of
timber declines. That is, the cost of increasing harvest and profits today by
depleting timber stocks will be less timber and thus profits from harvesting
in the future. Provided that good government policies and regulations are
in place to ensure that private individuals making the harvesting decisions
that affect the availability of timber resources over time are doing so in
accordance with social interests, then the rate of timber stock depletion
should be socially 'optimal'. However, as will be discussed further below, it
is doubtful whether domestic policies in many tropical producer countries
are operating in this manner.
In addition to the direct depletion of stocks, timber extraction can incur
external environmental costs by degrading other tropical forest resources
and functions which are of value to people. Much of the criticism of tropi-
cal deforestation stems from scientists' claims that closed tropical forests
are estimated to hold between 50 and 90% of the world's biodiversity
The costs of implementing sustainable forest management 231
(Reid and Miller, 1989). Evidence of the importance of the sustainable uses
of this biodiversity for subsistence forest products, pharmaceutical and
crop-breeding research and ecotourism, and as a key component in global
warming, local watershed protection, microclimatic functioning and other
environmental functions, dictate that timber extraction imposes significant
external costs in the tropics.
The extent of these external environment impacts from timber extraction
depends largely on the type, and success, of forest management practices.
Poore et al. (1989) notes that successful tropical forest management
depends on certain conditions being met, including the long-term security
of the investment, operational control, a suitable financial environment
and adequate information. Although there are a few cases of successful sus-
tained-yield management of tropical forests for timber, these tend to be the
exception.
Poorly designed and implemented management regimes for selective log-
ging of natural forests are likely to have serious implications for the envi-
ronment. For example, on average only about 8-33 m 3 of merchantable
wood are extracted per hectare of tropical closed broad-leaved forests
(FAO, 1993). However, these small commercial volumes relative to total
standing timber can lead to disproportionate damage to the forest due to
careless use of equipment and inefficient logging practices. Typically, at
least half of the remaining stock, including immature trees of commercial
value and harvestable stocks of less desirable varieties, are damaged
beyond recovery (Repetto, 1990). Measuring the wider environmental
impacts and costs of forest management practices is complicated, and few
empirical studies have been undertaken.
The impact of timber production on the environment may be offset to
some extent by investments in reafforestation. However, if investments to
offset the degradation of natural forest are channelled into plantation
forests then only part of the full environmental costs of timber extraction
may be compensated. That is, investments in plantations may counteract
the decline in stocks of tropical timber, but may not always compensate for
the wider environmental costs of natural forest degradation. In many tropi-
cal countries, natural forests have been cleared and replaced with planta-
tion forests, with the loss of wider environmental values such as
biodiversity. Even if the compensatory investment is in management of a
natural forest area, some of the environmental benefits of the degraded nat-
ural forest may still have been irreversibly lost, e.g. biodiversity or non-
timber forest products. Nevertheless, Lugo et al. (1993) have argued that
even plantation forestry and secondary forest management in the tropics
can be used effectively to maintain much of the indigenous understorey
biodiversity.
In addition to the direct impacts of timber extraction on the environment,
timber production can influence environmental degradation indirectly. This
indirect impact may occur through the opening up and improvement of
access to the forests, which may then interact with other socioeconomic fac-
tors, encouraging activities that degrade the environment. However, due to
the intricately interconnected relationship of the various causes of tropical
deforestation, it is extremely difficult to identify how much of the defor-
estation process is due to timber production.
In many developing countries where there are still areas of previously
unexploited forest, and there exist no formal property rights for this land,
timber production may encourage open-access exploitation at the forest
frontier and rapid forest conversion (see studies in Brown and Pearce,
1994). Timber extraction is often the first step towards opening up the
tropical forest and clearing the land for agricultural production. Timber
extraction usually involves extensive road building which benefits other
activities, such as agriculture and hunting, by improving access to the for-
est and reducing costs of transporting produce to market. What is more, in
many developing countries property law establishes deforestation as a pre-
requisite of formal claim over the land for those settling in forested areas.
The environmental impact of forest conversion to agriculture has grown as
logging has progressively opened up more remote, hilly and ecologically
vulnerable areas.
8.5.2 Public policies and the incentives for sustainable forest

management
Many of the impacts of timber extraction involve losses of environmental
goods and services that are essentially non-marketed and would not be
normally taken into account by the private individuals making the harvest-
ing decisions. Although some of the resulting environmental values lost
(such as declining biodiversity, loss of carbon storage and even damage to
watershed protection function) may benefit individuals in other countries,
many of the foregone values affect the welfare of domestic populations
within tropical timber-producing countries. It is the role of public policy in
producer countries to ensure that the latter welfare impacts are fully incor-
porated, or 'internalized', in the timber-harvesting decisions undertaken by
private individuals. Unfortunately, in many tropical producer countries,
wider economic and forestry policies not only fail adequately to 'correct'
economic incentives to account for the costs of unsustainable forest man-
agement, but actually encourage such practices through distorting influ-
ences on markets.
As outlined by Hyde and Sedjo (1992), public policies influence the envi-
ronmental effects of tropical timber forest management through their
impacts on:
• the level of privately efficient harvests;
• the level of socially efficient harvests when accounting for environmen-
tal externalities;
• alternative royalty, contract and concessional arrangements and their
implications for trespass, high-grading and other environmental losses;
• the level of rent distribution.
(Trespass is a forestry term referring to losses due to logging theft, which
could also be extended to include losses due to graft; high-grading refers to
the removal of high-valued timber leaving a degraded timber stand.)
Many domestic policies do not even begin to approximate the appropri-
ate incentives required to achieve a socially efficient level of tropical timber
harvesting. More often than not, pricing, investment and institutional poli-
cies for forestry actually work to create the conditions for short-term har-
vesting by private concessionaires, and in some instances, even subsidize
private harvesting at inefficient levels.
For example, section 8.3 above discusses how government policies in
Malaysia and Indonesia to encourage switching from the export of raw
logs to processed timber products have led to substantial economic losses,
the establishment of inefficient processing operations and accelerated
deforestation. Throughout South-East Asia the allocation of timber conces-
sion rights and leasing agreements on a short time scale, coupled with the
lack of incentives for reforestation, have contributed to excessive and rapid
depletion of timber forests (Poore et aI., 1989). In the Philippines, the
social gain from logging old-growth forest was found to be negative, once
the social costs of timber stand replanting, the costs of depletion and the
costs of off-site damages were included (Paris and Ruzicka, 1991; see Table
8.10).
Short-term concessions and poor regulatory frameworks, coupled with
inappropriate pricing policies, often contribute to excessive rent-seeking
behaviour in tropical timber production (Repetto and Gillis, 1988; Gillis,
1990; Barbier et aI., 1994). That is, concessionaires have an incentive to
open up additional stands for harvesting in order to 'mine' timber for high
short-term profits. By not charging sufficient stumpage fees and taxes or by
selling harvesting rights too cheaply, most governments have allowed the
resource rents to flow as excess profits to timber concessionaires and specu-
lators, often through short-term harvesting operations.
For example, in the Philippines, if the government had been able to col-
lect the full value of actual rents, its timber revenues would have exceeded
US$250 million - nearly six times the US$39 million actually collected.
Instead, excess profits of at least US$45001ha went to timber concession-
aires, mill owners and timber traders (Repetto, 1990).
As noted by Hyde and Sedjo (1992), rent capture per se may not be as
fundamental to an efficient outcome as ensuring proper internalizing of the
Table 8.10 Private and social efficiency -logging in the Philippinesa
Source Net returns (P/ha/year)
Value of log harvestb 5720

Road building, harvesting and transportation costs -2369
Net financial return to short-run timber harvesting 3351
Internalizing cost of sustainable timber management:
Cost of protection, timber stand improvement and
enrichment planting C-1000
Net returns to 'privately efficient' timber harvesting 2351
Shadow pricing adjustments:
Adjustment to market price of logsd 572
Adjustment to local harvesting costse 474
Social costs of degradation of non-timber values:
cost of marginal offsite damage to downstream activities f -6245
Net returns to 'socially efficient' timber harvestingg -2848
US$l = etc.
• Old growth forest selectively logged and subsequently protected.
b Legal operations using existing selective logging systems assumed. Private profits of illegal
operators will be higher. Different combinations of yield and price are possible to capture
the variations in the quality of standing forest. Assumption is that one hectare of old
growth forest of 30-50% slope sustainably yields 100 cu m every 35 years or 2.86 cu m per
year. Market price is 2000 Philippino Pesos per cu m.
c Pl,OOO/ha for one year to ensure sustainability of production on the one hectare in
question.
d Equal to market price adjusted upwards by 10% to account for low cost illegal supplies.
e Standard conversion factor 0.8 applied.
f P2600 for 3 years discounted at 12%. Off-site damage assumed to be limited in duration
instead of being sustained in perpetuity. Damage resulting from the selective logging of a
single plot. The high magnitude of the estimate is a result of logging on very steep slopes.
g The overall resturns are only a rough indication and are based on estimates prepared in
1990 as part of the formulation of the Philippine Master Plan for Forestry Development.
Adapted from Paris and Ruzicka, 1991.
user costs of timber exploitation through appropriate contractual and con-

cession terms. That is, even if government rent capture is low, it can still
ensure through proper concession arrangements that the stand is harvested
at the private long-run efficient level. However, in many countries poor
rent capture and poor concession policies go hand in hand, combining to
produce short-term and rent-seeking behaviour in concessionaires.
Although the total area of production forest in the Philippines is 4.4 mil-
lion ha, the total area under timber concessions exceeds this at nearly 5.7
million ha - almost 90% of the entire forest area. Concessions are awarded
for 25 years but some for as little as 5 years, even though the minimum
realistic felling cycle is 30 years and the rotation 60 years. Almost all the
large logging companies have senior politicians on their boards, and it is
generally the politicians, not forestry officials, who ultimately determine
concession policy and allocation (chapter 5 in Poore et al., 1989).
Hyde and Sedjo (1992) point out the difficulty experienced by forestry
departments in developing countries in administrating and collecting vari-
ous timber fees and taxes. Much of the problem may have to do with the
complexity of fees and concession arrangements, which makes enforcement
and supervision of revenue collection difficult. In a review of forest pricing
and concession policies in West and Central Africa, Gmt et al. (1991) sug-
gest replacing the multiplicity of forest fees with an annual concession rent,
set by competitive bidding, and replacing logging concessions with forest
management concessions that should be regularly inspected.
Public policies also have an important influence on the pattern of forest-
based industrialization and its implications for long-term economic devel-
opment and deforestation. Vincent and Binkley (1991) note that stumpage
prices (i.e. the prices of harvested logs at the stand) have a crucial role to
play in the interrelated dynamics of timber reserve depletion and process-
ing expansion, particularly in facilitating the transition of the forest sector
from dependence on old-growth to second-growth forests and in coordi-
nating processing capacity with timber stocks. Unfortunately, in most
developing countries stumpage prices tend to be administratively deter-
mined rather than set by the forces of supply and demand, thus understat-
ing stumpage values and failing to reflect increasing scarcity as old-growth
forests are depleted. A number of economic and environmental distortions
result:
• old-growth forests are depleted too rapidly;
• forest land is inappropriately cleared for agriculture or other uses;
• inadequate and inappropriate investment is made in second-growth
forests;
• inefficient processing facilities are installed;
• decisions on log and lumber trade policies are inefficient and encourage
unsustainable management practices;
• elaborate and counter-productive capital-export controls are needed to
ensure that resource rents are not repatriated.
Vincent and Binkley (1991) explore three case studies of long-term forestry
policy and development: logging in Peninsular Malaysia which is mainly of
old-growth tropical forests; in Ghana which logged over most of its old-
growth tropical forest but has poorly developed its secondary forest; and in
Chile which produces timber mainly from second-growth (e.g. plantation)
temperate forests. In Malaysia, wood prices (timber charges, log prices and
sawnwood prices) have been kept artificially low at every stage of forestry
development, from log exporter to exporter of primary products to embry-
onic exporter of downstream products. The result has been the development
of processing capacity that exceeds the forests' sustained yield capacity.
Like Malaysia, Ghana has shown little success in establishing plantations
as an alternative source of timber to the natural forest. Artificially low roy-
alty rates for natural forest timber mean that concessionaires' incentives to
invest in plantations are limited, and at the same time, through their
impact on delivered log prices, have helped encourage the overexpansion
of domestic processing capacity. In contrast, in Chile the crucial policy
issue has been how to build up an efficient industry based on the nation's
increasing supply of timber from privately owned plantations (mainly of
Pinus radiata). Initial success seems to have come from providing private
investors with secure, long-term tenure for plantations. Another key policy
has been to permit the export of logs, which forces domestic mills to pay
world prices for sawlogs, pulpwood and woodchips and to be more effi-
cient.
8.5.3 Internalizing the costs

Several important points emerge from the above discussion. Firstly, inade-
quate and often distortionary public policies are a major barrier to sustain-
able forest management in tropical producer countries. The result is
inappropriate economic incentives at the stand level that lead to inefficien-
cies in timber harvesting and create the conditions for short-term extrac-
tion for immediate gain, while at the same time failing to 'internalize' the
direct and indirect environmental impacts of forestry operations. Improper
policies also have a more long-term and wide-scale effect on the pattern of
forest-based industrialization and its implications for the management of
the tropical forest resource base as a whole, including the conversion of
forest land to agriculture and other uses. Thus policy reform to improve
sustainable tropical forest management may not only reduce the direct and
indirect environmental impacts of forestry operations, but may also be jus-
tified on economic efficiency grounds for long-term development of the
forestry industry and the use of forest resources. The result is that tropical
producer countries may incur significant short-term costs from encourag-
ing policy reforms and regulations to support sustainable forest manage-
ment, but they are also likely to gain substantially in the long run from a
more efficient forestry sector. Even in the short run, the reduction in subsi-
dies, preferential tax breaks and other inducements may be an additional
financial benefit of policy reform.
Equally, the transition to sustainable tropical forest management may
impose additional costs at the stand level for residual stand management
and increased environmental protection. For example, Baharuddin and
Simula (1996) suggest that the increased costs may derive from five differ-
ent sources: (i) setting aside of areas; (ii) lower harvesting yields; (iii) addi-
tional silvicultural and harvesting costs; (iv) additional costs of planning
and monitoring; and (v) different distribution of costs and benefits over
time. Low-intensity harvesting will generally mean less timber extracted
per hectare in the short term. However, these costs can be at least partially
offset by improved harvesting techniques and better planning which will
reduce operating costs. In addition, the current income initially foregone
with reduced yields may be more than compensated over the long run from
improved stand productivity and yields as a result of reduced residual dam-
age and better stand regeneration and recovery. Too often, assessment of
the costs of sustainable forest management focuses on the short-term costs
of implementing improved management and fails to take into account the
potential long-term gains in stand productivity and income.
Estimating the additional costs both to timber operations at the stand
level and to forestry industries at the national level of implementing sus-
tainable forest management practices is therefore extremely difficult.
However, the available evidence does suggest that on the whole:
• the transition to sustainable management is likely to impose some
increase in production costs in the short term, both at the industry-wide
level and the stand level;
• the additional costs are likely to be significant for many tropical pro-
ducer countries;
• it may no longer be economically worthwhile to harvest some tropical
forests, and large areas of some countries' forest resource base may have
to be 'set aside' from production which could result in some income
losses;
• increases in costs and stumpage prices at the stand level do not necessar-
ily mean significantly higher prices for final tropical timber products.
The costs of implementing sustainable forest management are likely to vary
significantly across tropical forests, countries and regions. Additional pro-
duction costs ranging from 5-50% are possible (Baharuddin and Simula,
1996). Most estimates suggest that the costs of sustainable forest manage-
ment per m3 of log produced lies between 10 and 20% of the current aver-
age international tropical log price of about US$350 (Varangis et al.,
1995).
The higher costs of sustainable tropical forest management on overall
timber production are likely to make it unfeasible to harvest some forest
areas that would have otherwise been logged. This makes perfect sense in
cases where the failure to internalize the environmental and the long-run
costs of timber operations have meant that these operations remain finan-
cially profitable even though they are socially inefficient.
An estimate of the private and social returns to selective logging on
steeply-sloped (30-50%) old-growth forest in the Philippines illustrates the
result (see Table 8.10). Selective logging yields positive private returns
when the revenues from selling logs (i.e. the value of the log harvest) are set
against the costs of road building, harvesting and transportation. After
accounting for the costs of protection, timber stand improvement and
enrichment planting to ensure the sustainability of production, the private
returns are reduced but still positive. However, when the loss of non-timber
values is taken into account the net economic returns are substantially neg-
ative. In this case the magnitude of the estimated damage to downstream
activities indicates that the Philippines would be better off not harvesting
old-growth forests on such steep slopes. Whereas it is in the direct eco-
nomic interest of the private concessionaire to continue unsustainable har-
vesting on the steep terrain, it is not in the interests of the Philippines.
However, the widespread implementation of sustainable forest manage-
ment across many tropical regions and countries could result in the
removal of many forest areas from potential production. Although there
would no doubt be substantial environmental gains, the economic costs to
producer countries could be significant, particularly for major tropical tim-
ber-exporting countries.
A recent policy simulation was conducted using a global forest-sector
trade model to indicate the additional economic impacts to tropical forest
countries of 'setting aside' some of their forest resource base (Perez-Garcia
and Lippke, 1993). Essentially, this was simulated by a reduced timber sup-
ply scenario where the inventory of commercial tropical hardwood
resources is reduced by 10%, which is equivalent to land being taken out
of production forest and permanently protected. The result is that the
underlying trend of increasing timber scarcity and higher sawlog prices is
accelerated in tropical forest regions, notably in Malaysia and Indonesia.
The model indicates that such reductions in supply would result in a loss of
wealth for tropical timber-producing countries. Over the long run, per-
manent set asides would mean that the remaining production forest inven-
tory could not support as high a level of sustainable harvest as under base
case projections.
It is sometimes argued that the higher additional costs of sustainable for-
est management will make many timber products uncompetitive in final
markets. However, although harvesting costs are often a large proportion
of the stumpage value of logs, for most processed forest products the cost
of the wood raw material is a small proportion of the total costs of harvest-
ing. This is particularly the case for products traded globally; for example,
typical stumpage values in tropical countries of US$6-30/m 3 log equivalent
end-product often represents less than 1 % of the final value of the product
being sold in foreign consumer markets (Barbier et aI., 1994). As a conse-
Financing sustainable forest management 239
quence, even reasonably large increases in harvesting costs and the
stumpage value of timber can have little or only a modest impact on the
final product price in consumer markets. Thus the available evidence for
tropical wood products suggests that a doubling of harvesting costs may
lead to an increase of 10-15% of the costs at the importer or wholesaler
level and less than a 10% increase in the retailer's cost (OFIfTRADA,
1991; Barbier et al., 1994; Dubois et aI., 1995).
A model of Indonesia's forestry sector developed by Barbier et al. (1995)
was used to simulate a policy initiative by Indonesia to implement more
'sustainable' management of its remaining production forests. Correcting
forest management policy failures would improve the use of Indonesia's
remaining tropical forests and thus reduce any timber-related deforesta-
tion, but would also mean higher harvesting costs per m3 of wood
extracted. As shown in Table 8.11, scenarios depicting 25 and 50%
increases in harvesting costs across Indonesia's forestry sector were exam-
ined.
Although domestic log prices are affected significantly by the increased
harvest costs, any resulting impacts on the rest of Indonesia's forestry sec-
tor seem to be somewhat dissipated. There appear to be several factors at
work. Firstly, the price elasticity of log supply for Indonesia is very inelastic
in the model (see Barbier et al., 1995 for further details). This is not sur-
prising given that Indonesian policy has been to devote log production
solely to supplying domestic processing capacity, which of course has
expanded considerably over the 1968-88 period covered by the model.
Secondly, increased log costs are only one component of the total factor
costs of Indonesia's processing industries, and have increasingly become
the least important component in recent years. Thirdly, Indonesia's sawn-
wood and plywood exports seem to be the least affected by the increased
harvest costs, which would suggest that external demand factors exert an
important counteracting influence. The simulation results shown in Table
8.11 do not show the direct reductions in timber-related deforestation due
to improved sustainable forestry practices; any reductions in deforestation
are purely the result of higher harvesting costs reducing logging (see
Barbier et aI., 1995 for further details). Interestingly, these results suggest
that a transition to sustainable forest management would have the addi-
tional benefit of encouraging value-added domestic processing industries
rather than exporting raw logs, which would suggest that such a policy is
consistent with Indonesia's overall forest-based industrialization strategy.
8.6 FINANCING SUSTAINABLE FOREST MANAGEMENT

Although in the long run producer countries ought to obtain sufficient
returns from timber production, from value-added processing and from
exports to cover the additional harvesting costs and other economic
Table 8.11 Indonesia - timber trade and tropical deforestation simulation model
25% rise in 50% rise in

Key variables harvest costs harvest costs
Prices (Rp/m3)
Log border-equivalent price (unit value) 41.59% 83.06%
Sawnwood export price (unit value) 4.04% 8.09%
Plywood export price (unit value) 2.86% 5.72%
Quantities (thousand m 3)
Log production -0.94% -1.87%
Log domestic consumption -1.37% -2.73%
Sawnwood production -1.89% -3.77%
Sawnwood exports -1.03% -2.05%
Sawnwood domestic consumption -2.28% -4.55%
Playwood production -0.87% -1.73%
Plywood exports -0.38% -0.75%
Plywood domestic consumption -3.12% -6.24%
Deforestation (km2 )
Total forest area 0.02% 0.04%
Annual rate of deforestation -2.28% -4.23%
Policy scenario - sustainable timber management (percentage change over base case).
Source: Barbier et ai., 1995.
impacts of sustainable timber management, in the short and medium term

the transition to sustainable forest management may be costly for many
producer countries, particularly developing economies. In addition, some
of the additional costs that producer countries may be asked to bear would
be associated with maintaining certain forest services, such as carbon
storage, biodiversity preservation and watershed protection, which yield
mainly global and regional benefits. To undertake some of the required
basic improvements in forest-resource assessment, monitoring and planning
of forest management, chain-of-supply inspections of processing activities,
independent certification and even basic forest-sector policy analysis may
require a transfer of resources, skills and even technology to build up local
capacity in many developing countries.
Thus additional financial and technological assistance (including human
skills) may be required from the international community to assist some
developing countries in the transition to sustainable forest management.
The main rationale for assisting developing countries in moving towards
sustainable forest management is that two important principles of interna-
tional compensation may be at stake.
• Implementation of the forestry policies and regulations required to
ensure the proper implementation and monitoring of sustainable man-
agement of production forests will impose substantial additional costs
on some developing countries which they will find difficult to afford.
There are also likely to be skills and even technology capacity con-
straints on many of these countries' abilities to implement and monitor
sustainable forest management in the short term.
• To the extent that all nations benefit from the global external benefits
resulting from sustainable management of large tracts of tropical forest
lands, then the international community should compensate developing
countries for the loss of potential income that they would incur by
reducing deforestation, timber sales and conversion of forest land to
other uses.
It is now generally accepted, as well as enshrined in the Forest Principles
Accord of UNCED (1992) as well as in the deliberations of the Inter-
governmental Panel on Forests, that compensating developing countries for
their role in maintaining forest resources that have value on a global level
is a fundamental basis of multilateral policy action. It should also be recog-
nized that compensation is needed by developing countries for the income
they may forego in protecting their forests and for the additional costs
incurred in implementing sustainable management practices for their pro-
duction forests.
However, actual assessment of the additional financial and technological
assistance required to compensate developing countries is extremely diffi-
cult to undertake empirically. Estimating the global values generated by
sustainable forest management will in many cases require information on
biophysical relationships and human welfare impacts that is not readily
available. Lack of information, market transparency and adequate local
skills and assessment capacity also make it difficult to determine accurately
the income losses and additional costs of implementing sustainable forest
management in many developing countries. Some financial and technical
assistance is already being provided to developing countries through multi-
lateral and bilateral sources to promote and implement sustainable forest
management. As discussed in section 8.5, any additional economic benefits
due to improved stand-level management and policy reform also need to be
accounted for.
Nevertheless, there are some preliminary indications of the additional
costs required to implement sustainable forest management policies and
regulations in developing countries. Based on broad estimates made for
lITO and UNCED, additional funds required by all producer countries to
implement sustainable management of their tropical forest resource could
be anywhere in the range US$O.3-1.S billion annually (Barbier et aI., 1994).
Although these figures suggest a need for additional financial assistance
for producer countries, the real issue is whether the financing ought to be
raised from the tropical timber trade or from other sources. There are
essentially three policy options available:
• re-direction of existing revenue from the trade;
• appropriation of additional revenue from the trade;
• additional funding from sources external to the trade.
A recent study has argued the case for a tax transfer of revenue from the
trade between consumer and producer countries (OFIffRADA, 1991). The
main implications of a tax transfer would be to change significantly the dis-
tribution of economic rents from the timber trade as an incentive for sus-
tainable management in developing countries. The main advantage would
be that additional funds for this purpose could be raised with little or no
effect on final product prices. Preliminary estimates suggest that just under
US$1.S billion in additional funds could be raised by producer countries
through this means - closer to the 'upper bound' of the estimated financing
for sustainable management required by these countries (Barbier et aI.,
1994).
However, consumer countries are likely to be concerned about the fiscal
and political implications of a tax transfer scheme. Based on the above cal-
culations, their governments would have to forego nearly US$3.7 billion in
tax revenues from the trade - more than 2.5 times what producer country
governments gain in increased revenues (Barbier et aI., 1994). This implies
a substantial net loss in revenue captured from the trade. Secondly,
exempting tropical timber products from VAT or other taxes could prove
politically problematic in that it could set the precedent for other goods
being exempted from taxes on environmental grounds. Forest industries in
temperate forest countries and their governments could also press for simi-
lar treatment on the same grounds - the need for additional investment for
sustainable management or for compensation for past investment. Thirdly,
establishing an appropriate and internationally agreed monitoring and
enforcement system would be a formidable challenge.
One possible variant on the tax transfer scheme would be a revenue
transfer scheme. Rather than lower their VAT or other taxes on the trade in
timber products from developing countries, consumer-country govern-
ments could instead transfer directly some proportion of the revenue raised
through these taxes to developing countries. Although the consumer-coun-
try governments would still forego substantial revenues, they could ensure
more direct control, and thus leverage, over the allocation of funds to sus-
tainable management, for example indicating which producer countries are
to receive them. In addition, consumer-country governments would most
likely have to forego less revenue if it were directly transferred to develop-
ing countries to help them meet their US$0.3-1.5 billion target than under
the tax transfer scheme (Barbier et aI., 1994).
However, a revenue transfer scheme would still require an internation-
ally agreed monitoring and enforcement system. Developing-country gov-
ernments would probably be more concerned about the direct control and
conditions that consumer-country governments could assert over a revenue
transfer scheme. International agreement on the establishment, implemen-
tation and monitoring of any such scheme would have to be carefully nego-
tiated.
In recent years there has also been renewed interest in the use of trade
instruments, such as a small 1-5% surcharge on tropical timber imports
into developed country markets, to appropriate additional revenue from
the trade for sustainable forest management. If endorsed by a global
forestry agreement, an import surcharge would be within GAITIWTO
rules through sub-article XX (h). A differentiated surcharge could also be
imposed so that imports of processed tropical hardwood products face less
discrimination than logs, thus reducing any remaining distortions from
escalating tariffs. The funds raised could most probably be transferred to a
mutually recognized international body, such as the lITO, for distribution
to tropical producer countries, possibly through specific projects and pro-
grammes.
One of the major concerns of developing countries is that any revenue-
raising import surcharge, even at very low levels, would be distortionary.
In particular, if the tax was levied by all importers on a wide range of trop-
ical timber products, then there could be a more significant impact on total
world trade in these products through substitution effects (see section 8.3).
Moreover, such a tax would discriminate in favour of temperate forest-
based industries of the developed market economies. Thus as an alterna-
tive, developing countries may prefer a tax on exports rather than an
import surcharge. This would give producers more direct control over the
proceeds of the tax. In addition, an export tax would affect all import mar-
kets rather than just one, thus spreading the costs of sustainable manage-
ment to all producers and consumers (Buongiorno and Manurung, 1992;
Barbier et aI., 1994).
Finally, there is the issue of whether the amount of funds raised through
any trade surcharge would be adequate for the task. The studies under-
taken so far suggest that the amount of net funds raised from a trade sur-
charge of 1-5% may fall short of the approximate target of US$0.3-1.5
billion required annually by developing countries as additional resources
for sustainable forest management (NEI, 1989; Buongiorno and
Manurung, 1992; Barbier et aI., 1994).
All the above proposals for raising the additional funds for sustainable
forest management from the global forest products trade are clearly
controversial and pose great difficulties for comprehensive international
agreement on mutually recognized and transparent schemes. It is also
unclear why the international trade in forest products should be used to
raise revenues to cover the costs of sustainable forest management, when
the vast majority of timber production world-wide does not even enter into
trade (see section 8.2). Schemes to raise revenues from the world forest
products trade to provide financial assistance for sustainable forest man-
agement could be deemed to involve unnecessary, and possibly inappropri-
ate, discrimination against the timber trade.
Thus there is a strong rationale for additional funds to be made available
to developing countries for sustainable forest management from sources
outside the forest products trade, rather than raising revenues through
redirecting existing or appropriating additional revenues from the trade.
Comprehensive international agreements, targeted financial aid flows and
compensation mechanisms to deal with the overall problem of sustaining
global forest resources may ultimately eliminate the need to consider inter-
ventions in the forest products trade to secure funds for sustainable forest
management.
It is unlikely in the current global economic climate that there will be a
concerted international effort to substantially increase bilateral or multilat-
eral aid flows for sustainable forest management globally. Nevertheless,
there still remains the possibility of designing new sources of financial
assistance that are separate from existing developing-country aid budgets.
The Forestry Principles signed at the 1992 UN Conference on Environment
and Development are effectively a step in that direction, and international
commitments through the National Forestry Action Plan (NFAP) and
Global Environmental Facility (GEF) continue to reinforce the global inter-
est in forestry and biodiversity protection. The case could be made that a
comprehensive international agreement on global forest management
should include provisions for additional funds and technical assistance
over the short and medium term to support the transition to sustainable
forest management in developing countries. Such assistance could take the
form of compensation payments made from an international 'rain forest
fund' (Amelung, 1993). Alternatively, such payments could arise through
the establishment of a global system of marketable forest protection and
management obligations that could be initially limited to mainly bilateral
agreements with little trading before being implemented globally (Sedjo et
aI., 1991). Other schemes that essentially provide the mechanism for trade
in forest services include debt-for-nature swaps, carbon offsets, internation-
ally tradeable carbon dioxide permits and tradable development rights.
Improvements in technological capacity and international competitive-
ness of the forestry sector in developing countries may be important for
improving the ability of these countries to benefit from trade-related
Summary and conclusions 245
incentives for sustainable forest management. This suggests that there may
be great scope for encouraging appropriate and complementary interna-
tional investments from the private sector to promote efficient and sustain-
able forestry operations and industrialization in developing countries. In
particular, these countries should explore ways of attracting greater private-
sector investment through joint ventures to facilitate the transfer of tech-
nologies, capacity building and funds to finance sustainable forest
development.
8.7 SUMMARY AND CONCLUSIONS

Trade is still a relatively small proportion of global timber production.
Approximately one quarter of wood-based panels and paper products and
one fifth of sawnwood and wood pulp are traded internationally. Only
6-7% of global and tropical industrial roundwood output is currently
traded. In addition, tropical timber plays a minor role in the global timber
market: tropical countries account for approximately 15% of the total vol-
ume of global timber production, and 11 % of the value of global exports.
The trade in forest products is highly regionalized within three impor-
tant trading blocs, the Pacific Rim, North America and Europe (mainly
Western Europe), dominated by developed country importers such as
Japan, the USA, Canada and the European Union. However, in recent
years developing countries and newly industrializing countries, particularly
in Asia, have been increasing their share of global imports. The major
global exporters of forest products still tend to be developed countries with
temperate forest resources and processing industries. However, developing
countries such as Indonesia and Malaysia have emerged as dominant world
exporters of certain forest product exports, such as non-coniferous wood-
based panels, logs and sawnwood. Other developing countries, notably
Brazil, Chile and the Asian newly industrializing countries, are beginning
to have an impact on the international trade in wood pulp and paper prod-
ucts. In general, the trade in forest products has shifted towards value-
added processed products.
Recent short- and long-term projections of the forest products trade sug-
gest that there may be an increasing scarcity of tropical hardwood
resources, particularly given current rates of exploitation and consumption
in South-East Asia. However, temperate resources, secondary forests and
plantations, and new sources of tropical hardwood resources in Latin
America and Africa, are expected to meet any global shortfalls in supply.
The extent to which the scarcity of tropical hardwood resources may
become a constraint on processing activities, as reflected in increased real
prices for tropical logs and sawnwood, will depend on the commitment of
major developing-country producers to pursue sustainable management of
remaining old-growth production forests and coordinate processing capac-
ity with supply.
International market access will be a key factor in determining the long-
term success of a global policy to encourage sustainable management of
tropical forests. Although not all tropical forest products are by any means
traded internationally, there is considerable evidence that, provided appro-
priate domestic policies and forestry regulations are in place and enforced,
the proper economic incentives for sustainable tropical forest management
are enhanced by potential access to international markets that can add
value to forestry options. Certainly, n:strictions on the tropical timber
trade have a knock-on effect of reducing forest product prices and thus the
stand-level returns to forestry management. Such interventions will reduce
rather than increase the incentives for sustainable forest management.
By continuing progress towards improved market access for forest
products, the Uruguay Round Agreement has been an important and posi-
tive step towards encouraging sustainable forest management globally. The
Agreement has achieved even further reductions in most tariff barriers for
forest products, and perhaps more importantly, has reduced uncertainty by
the binding of tariffs in major importing markets and reducing the degree
of tariff escalation. In addition, some progress has been made in curtailing
some important non-tariff and other barriers that have affected the trade in
forest products in the past.
However, the potential benefits of the Uruguay Round in improving
market access for tropical timber products could be overshadowed by
some 'new' trade barriers discussed in this chapter: export restrictions by
developing countries to encourage domestic processing of tropical timber
for export; environmental and trade restrictions on production and exports
in developed countries that affect international trade patterns; and quanti-
tative restrictions on imports of 'unsustainably produced' timber products.
A particularly thorny but important issue, well-recognized by the WTO, is
determining whether a domestic regulation to restrict market access on
environmental or health grounds is a legitimate policy to ensure environ-
mental protection or whether it is a form of disguised protectionism (Eglin,
1995).
Competition between different wood products and between products
from different regions of origin and between wood and non-wood substi-
tutes is inevitable. The available evidence suggests that such competition
should not unduly hamper efforts to improve sustainable forest manage-
ment in tropical countries. To the contrary, if producer countries as a
group instigate sustainable forest management and this leads to higher
prices for tropical timber products across the board, then there may not
necessarily be any significant loss of market share. However, if only some
producers, such as a few tropical timber countries, move to more sustainable
practices and there is a significant price change as a result, some degree of
substitution away from their products may occur.
The increased utilization of lesser-used species from tropical production
forests is probably inevitable as some currently commercially exploited
species become increasingly scarce relative to their demand. However, it is
essential that any policy for exploiting lesser-used species and increasing
the volume of timber removed from stands is compatible and consistent
with improvements in the overall sustainable management of production
forests. For many tropical forest countries, international assistance in
establishing forest resource assessments might be an important initial step
in determining not only the potential commercial viability of lesser-used
species but also the capacity for exploiting these species through sustain-
able management practices.
The development of an internationally agreed, transparent and voluntary
timber certification process whereby final temperate and tropical forest
products can be traced back through the chain of supply to sustainably
managed forests is a long-term process, which is currently being hampered
by the proliferation of schemes and standards. Moreover, there is a genuine
fear in tropical forest countries that such schemes could be used selectively
as barriers against their tropical timber exports. There is a need to achieve
international harmonization and mutual recognition of standards, and to
encourage cooperation and common agreement between competing inter-
national accrediting schemes. The purpose of timber certification should be
to reinforce the positive incentives for sustainable forest management and
not to penalize or restrict production and trade in timber not meeting standards.
In fact, the current international focus on timber certification needs to be
put into perspective. To date, only a tiny proportion of the global trade in
forest products and a small area of the world's forests is influenced by tim-
ber certification. Even under extremely optimistic projections, this impact
is not likely to increase significantly in the foreseeable future. Thus not
only is timber certification only one of many instruments for promoting
sustainable forest management globally, but also its role in unilaterally pro-
moting sustainable forest management in tropical countries in the near
future is clearly limited.
Given the pressing need to promote sustainable forest management glob-
ally, it is imperative to develop urgently other instruments complementary
to timber certification that are more directly aimed at wholesale improve-
ments in forestry management policies and regulations in producer coun-
tries. One possible alternative is the country certification approach, which
would involve an international commitment by both producer and con-
sumer countries to adopt policies and practices towards encouraging sus-
tainable management of production forests and timber products while
simultaneously improving international market access of these products.
Such a scheme could be enacted for all timber producer and consumer
countries through an international agreement on global forests. The overall
objective would be to tackle directly the two major obstacles to sustainable
forest management identified in this chapter: restrictions on market access
and forest sector policy failures. It would also hopefully foster more
detailed agreements on international harmonization and mutual recogni-
tion of standards on timber certification.
The major costs associated with unsustainable tropical forest manage-
ment are the losses in potential timber stocks and future income from
short-term extraction decisions that do not take into account these deple-
tion costs, and the losses in valuable non-wood products and environmen-
tal services that are affected or lost through timber production. In tropical
forest countries, timber operations can have an important indirect impact
on the environment by opening up previously inaccessible forest areas to
other economic activities, such as frontier agricultural expansion, that
cause widespread deforestation.
It is the role of public policy in all tropical producer countries to ensure
that the domestic welfare impacts of these direct and indirect costs of
forestry operations are fully incorporated, or 'internalized', in the timber-
harvesting decisions undertaken by private individuals. Unfortunately, in
many producer countries, wider economic and forestry policies not only
fail adequately to 'correct' economic incentives to account for the costs of
unsustainable forest management, but actually encourage such practices
through distortionary influences on markets.
Thus public policies are a major barrier to sustainable forest manage-
ment in tropical producer countries. The result is inappropriate economic
incentives at the stand level that lead to inefficiencies in timber harvesting
and create the conditions for short-term extraction for immediate gain,
while at the same time failing to 'internalize' the direct and indirect
environmental impacts of forestry operations. Improper policies also have
a more long-term and wide-scale effect on the pattern of forest-based
industrialization and its implications for the management of the forest
resource base as a whole, including the conversion of tropical forest land to
agriculture and other uses. Policy reform to improve sustainable forest
management may not only reduce the direct and indirect environmental
impacts of forestry operations but may also be justified on economic effi-
ciency grounds.
Although the transition to sustainable forest management is likely to
impose significant costs on timber operations and forestry industries in
tropical regions, the overall economic impact may be less than initially
feared. The cost burden is likely to be substantial for many tropical coun-
tries as they are likely to face higher production and harvesting costs. It
may no longer be economically worthwhile to harvest some tropical
forests, and if this occurs on a large scale, significant proportions of some
countries' forest resource base may have to be 'set aside' from production.
If this occurs then it is likely to result in some losses in forestry income and
export earnings.
However, some of the increases in stand-level and industry-wide costs
resulting from the transition to sustainable forestry management and pol-
icy reforms could be offset by long-term improvements in forestry practices
and harvesting as well as greater efficiencies throughout the forestry indus-
try. Moreover, any increases in costs and stumpage prices at the stand level
do not necessarily mean significantly higher prices for final tropical timber
products. Although harvesting costs are often a large proportion of the
stumpage value of logs, for most processed forest products the cost of the
wood raw material is a small proportion of the total costs of harvesting.
Even substantial increases in harvesting costs and the stumpage value of
timber may result in only a modest impact on the final product price in
consumer markets. As discussed above, substitution by non-wood products
in final markets may therefore not pose a significant threat to global initiatives
to improve sustainable forest management.
The main rationale for assisting developing countries in moving towards
sustainable forest management is that these countries may require interna-
tional compensation to implement and monitor sustainable forest manage-
ment in the near term, and to the extent that all nations benefit from the
global external benefits resulting from sustainably managing large tracts of
tropical forest lands, then the international community should compensate
developing countries for the loss of potential income that they would incur
by reducing deforestation, timber sales and conversion of forest land to
other uses.
Although it is possible that the funds needed for such compensation
should come from either the re-direction of existing revenue from the trade
in forest products or the appropriation of additional revenue from the
trade, international financial and technical assistance to developing coun-
tries should come largely from sources outside the trade. Any agreement on
global forest management should include specific provisions for additional
funds and technical assistance over the short and medium term to support
the transition to sustainable forest management in developing countries.
Additional schemes that essentially provide the mechanism for trade in for-
est services include debt-for-nature swaps, carbon offsets, internationally
tradable carbon dioxide permits and tradable development rights.
Developing countries should also explore ways of attracting greater pri-
vate-sector investment through joint ventures to facilitate the transfer of
technologies, capacity building and funds to finance sustainable forest
development.
REFERENCES
Amelung, T. (1993) Tropical Deforestation as an International Economic Problem,
in Economic Progress and Environmental Concerns, (ed. H. Giersch), Springer-
Verlag, Berlin.
Arden-Clarke, C. (1991) The General Agreement on Tariffs and Trade,
Environmental Protection and Sustainable Development, WWF Discussion
Paper, WWF International, Geneva.
Arnold, M. (1991) Forestry Expansion: A Study of Technical, Economic and
Ecological Factors, OFI Paper No.3, Oxford Forestry Institute, Oxford.
Baharuddin, Hi. G. (1995) Timber certification: an overview. Unasylva, 46,18-24.
Baharuddin, Hi. G. and Simula, M. (1994) Certification Schemes for All Timber
and Timber Products, Report for the International Tropical Timber
Organization, Yokohoma.
Baharuddin, Hi. G. and Simula, M. (1996) Study of the Development in the
Formulation and Implementation of Certification Schemes for all Internationally
Traded and Timber Products, Draft Report to the International Tropical Timber
Organization, Yokohoma.
Barbier, E. B. (1994) The environmental effects of the trade in the forestry sector, in
The Environmental Effects of Trade, OECD, Paris.
Barbier, E. B. (1995) Trade in timber-based forest products and the implications of
the Uruguay Round. Unasylva, 46,3-10.
Barbier, E. B. (1996) Impact of the Uruguay Round on International Trade in
Forest Products, Food and Agriculture Organization, Rome.
Barbier, E. B., Burgess, J. c., Bishop, J. T. and Aylward, B. A. (1994) The
Economics of the Tropical Timber Trade, Earthscan Publications, London.
Barbier, E. B., Bockstael, N., Burgess, J. C. and Strand, I. (1995) The linkages
between the timber trade and tropcial deforestation - Indonesia. World
Economy, 18,411-442.
Bhagwati, J. (1995) Trade liberalisation and 'fair trade' demands: addressing the
environmental and labour standards issues. World Economy, 18,745-759.
Bourke, I. J. (1988) Trade in Forest Products: A Study of the Barriers Faced by
Developing Countries, FAO Forestry Paper 83, Food and Agriculture
Organization, Rome.
Bourke, I. J. (1995) International trade in forest products and the environment.
Unasylva, 46,11-17.
Brown, K. and Pearce, D. W. (eds) (1994) The Causes of Tropical Deforestation:
The Economic and Statistical Analysis of Factors Giving Rise to the Loss of the
Tropical Forests, University College Press, London.
Buongiorno, J. and Manurung, T. (1992). Predicted Effects of an Import Tax in the
European Community on the International Trade in Tropical Timbers,
Department of Forestry, University of Wisconsin, Madison (mimeo.).
Constantino, L. F. (1988) Analysis of the International and Domestic Demand for
Indonesian Wood Products, Report for FAO, Department of Rural Economy,
University of Alberta, Edmonton, Alberta.
Constantino, L. F. (1990) On the Efficiency of Indonesia's Sawmilling and
References 251
Plymilling Industries, D.G. of Forest Utilization, Ministry of Forestry,
Government of IndonesialFAO, Jakarta.
Drake, P., Haid, D., Hammons, E. J. and Williams, D. (1995) Analysis of
Macroeconomic Trends in the Supply and Demand of Sustainably Produced
Tropical Timber from the Asia-Pacific Region, Phase II, Report to the
International Tropical Timber Organization, Yokohoma.
Dubois, 0., Robins, N. and Bass, S. (1995) Forest Certification, Report to the
European Commission, International Institute for Environment and
Development, London.
Eglin, R. (1995) Trade and environment in the World Trade Organization. World
Economy, 18, 769-779.
Flora, D. F. and McGinnis, W. J. (1991) Effects of Spotted-Owl Reservations, The
State Log Embargo, Forest Replanning and Recession on Timber Flows and
Prices in the Pacific Northwest and Abroad, Draft Review, Trade Research,
Pacific Northwest Research Station, USDA Forest Service, Seattle.
FAO (1988) Forest Products: World Outlook Projections: Projections of
Consumption and Production of Wood-Based Products to 2010, FAO Forestry
Paper 84, Vols 1 and 2, Food and Agriculture Organization, Rome.
FAO (1992) Forest Product Prices 1971-1990, Food and Agriculture Organization,
Rome.
FAO (1993) Forest Resource Assessment: Tropical Forests, Food and Agriculture
Organization, Rome.
FAO (1995) FAO Forest Products Yearbook 1993, Food and Agriculture
Organization, Rome.
GAIT (1994) Results of the Uruguay Round of Multilateral Trade Negotiations.
Market Access for Goods and Services: Overview of the Results, Secretariat,
General Agreement on Tariffs and Trade, Geneva.
Gillis, M. (1990) Forest Incentive Policies, Paper prepared for the World Bank
Forest Policy Paper, World Bank, Washington, DC.
Grut, M., Gray, J. A. and Egli, N. (1991) Forest Pricing and Concession Policies:
Managing the High Forests of West and Central Africa, World Bank Technical
Paper No. 143, Africa Technical Department Series, The World Bank,
Washington, DC.
Hyde, W. F. and Sedjo, R. A. (1992) Managing tropical forests: reflections on the
rent distribution discussion. Land Economics, 68, 343-350.
lITO, (1990) Guidelines for the Sustainable Management of Natural Tropical
Forests, International Tropical Timber Organization, Yokohama.
Jha, V. and Zarilli, S. (1994) Eco-labelling initiatives as potential barriers to trade,
in Life-Cycle Management and Trade, Organization for Economic Cooperation
and Development (OECD), Paris.
Kemp, R. H. and Phantumvanit, D. (1995) 1995 Mid-Term Review of Progress
Towards the Achievement of the Year 2000 Objective, Report to the
International Tropical Timber Council, XIX Session, 8-16 November,
Yokohoma.
Kiekens, J.-P. (1995) Timber certification: a critique. Unasylva, 46, 27-28.
Lugo, A. E., Parotta, J. A. and Brown, S. (1993) Loss in species caused by tropical
deforestation and their recovery through management. Ambio, 22,97-105.
NEI (1989) An Import Surcharge on the Import of Tropical Timber in the
European Community: An Evaluation, Netherlands Economic Institute,
Rotterdam.
OFIffRADA (1991) Incentives in Producer and Consumer Countries to Promote
Sustainable Development of Tropical Forests, lITO Pre-Project Report, PCM,
PCF, PCI(IV)I1/Rev. 3, Oxford Forestry Institute (in association with the Timber
Research and Development Association), Oxford.
Pajuoja, H. (1995) The Outlook for the European Forest Resources and
Roundwood Supply, UN-ECFlFAO Timber and Forest Discussion Papers,
ECFlTIMIDP/4, United Nations Economic Commission for Europe, Geneva.
Paris, R. and Ruzicka, I. (1991) Barking Up the Wrong Tree: The Role of Rent
Appropriation in Sustainable Forest Management, Occasional Paper No.1,
Asian Development Bank Environment Office, Manila, Philippines.
Perez-Garcia, J. M. (1991) An Assessment of the Impacts of Recent Environmental
and Trade Restrictions on Timber Harvests and Exports, Working Paper No. 33,
Center for International Trade in Forest Products, University of Washington,
Seattle.
Perez-Garcia, J. M. and Lippke, B. R. (1993) Measuring the Impacts of Tropical
Timber Supply Constraints, Tropical Timber Trade Constraints and Global
Trade Liberalization, LEEC Discussion Paper 93-03, London Environmental
Economics Centre, UK.
Poore, D., Burgess, P., Palmer, J., Rietbergen, S. and Synnott, T. (1989) No Timber
Without Trees: Sustainability in the Tropical Forest, Earthscan, London
Reid, W. V. and Miller, K. R. (1989) Keeping Options Alive: The Scientific Basis for
Conserving Biodiversity, World Resources Institute, Washington, DC.
Repetto, R. (1990) Deforestation in the tropics. Scientific American, 262,36-45.
Repetto, R. and Gillis, M. (1988) Public Policies and the Misuse of Forest
Resources, Cambridge University Press, Cambridge.
Sedjo, R. A. and Lyon, K. S. (1990) The Long-Term Adequacy of World Timber
Supply, Resources for the Future, Washington, DC.
Sedjo, A., Bowes, M. and Wiseman, C. (1991) Toward a Worldwide System of
Tradeable Forest Protection and Management Obligations, Washington, DC.
Varangis, P. N., Crossley, R. and Primo Braga, C. A. (1995) Is There a Commercial
Case for Tropical Timber Certification? Policy Research Working Paper 1479,
World Bank, Washington, DC.
Vincent, J. R. (1992) The tropical timber trade and sustainable development.
Science, 256,1651-1655.
Vincent, J. R. and Binkley, C. S. (1991) Forest-Based Industrialization: A Dynamic
Perspective, Development Discussion Paper No. 389, Harvard Institute for
International Development (HIID), Cambridge, Massachusetts.
Vincent, J. R., Brooks, D. J. and Gandapur, A. K. (1991) Substitution between
tropical and temperate sawlogs. Forest Science, 37(5).
WRI (1992) World Resources 1992-93, World Resources Institute/Oxford
University Press, New York.
References 253
Youn, Y. C. and Yum, S. C. (1992) A study on the demand and supply of timber in
South Korea, paper presented at a Symposium on Forest Sector, Trade and
Environmental Impact Models: Theory and Applications, Centre for
International Trade in Forest Products, University of Washington, Seattle, 30
April-1 May 1992.
-9------
Can non-market values save
the tropical forests?
D. Pearce
Centre for Social and Economic Research on the Global Environment
(CSERGE), University College London, Gower Street, London
WC1E 6BT, UK
9.1 INTRODUCTION: THE FOREST CRISIS

It seems fair to say that, from a world standpoint, forestry is in crisis.
While it would be foolish to be complacent - the historical evidence of
social collapse due to deforestation is all too stark (Perlin, 1989) - there is
little evidence that the current supply of timber is scarce, and little evidence
that it will become scarce in the near-term (Sedjo and Lyon, 1990; Hyde et
aI., 1991). But there is a concern that the non-timber functions of forests
are becoming increasingly scarce (Krutilla, 1967; Panayotou and Ashton,
1993). Those non-timber functions include the maintenance of biological
diversity, carbon sequestration, local non-timber products, environmental
protection, broader life support functions, recreational use, and 'passive
value' - the value of forests independent of any use now or in the foresee-
able future. The clue to the simultaneous condition of non-timber scarcity
and, if not timber abundance then timber adequacy, lies in the fact that
non-timber functions of forests are as much a product of the type of forest
as they are of the geographical extent of forests. Biodiversity, for example,
would appear to be a function of heterogeneous forests rather than uni-
form ones, although that is not a hard-and-fast rule. The forest crisis is,
then, a crisis of non-timber function loss, and it is easy to see that such a
crisis could emerge almost unnoticed in a world where, until recently, for-
est policy was determined by commercial timber interests alone.
If non-timber values are disappearing, why would we not expect the
same economic feedback forces to apply to them as to timber? In other
words, non-timber scarcity should raise the price of non-timber functions,
encouraging new supply. That it does not happen this way arises from two
fundamental features of the forest context.
The first is that, in the developing world, some non-timber functions
256 Can non-market values save the tropical forests?
have prices but the relevant markets are often the domain of the poor, the
vulnerable and the powerless. Whether wild meat, nuts, honey or fuel
wood, the market agents involved in these products cannot compete with
the superior power of the forest concessionaires: governments and the
multinationals. Even if market prices for non-timber products yield higher
per hectare revenues than timber (a debatable issue in itself, as we shall
see), insecurity of property rights makes for uneven competition.
The second feature of the forest context is common to both developed
and developing worlds: many functions have no market. If markets are
missing, then the implicit ruling price is zero. Non-market functions count
for nothing. It is hardly surprising that in a competition between land uses
(and it is this competition that defines the problem for the forest sector just
as it does for land-based biological resources generally; Brown et al., 1993;
Swanson, 1994), forests with high non-market value will lose out.
The recipe for reversing this situation or, more realistically, for slowing it
down, appears obvious. It consists of the two stages of policy advice that any
practising environmental economist would give: 'demonstrate and capture'.
Demonstration means demonstrating the underlying economic value of
non-timber values. In the case of marketed non-timber values, the issue is
more one of bringing to the attention of policy-makers and opinion-makers
the ways in which forests serve local communities who already market and
utilize forest functions. In the case of non-market values the issue is more
complex, as it is necessary first to impute the economic values. The extent
to which these non-market valuation exercises have been successful is of
course open to debate, but even in the course of a decade economic
research has made major inroads into this problem of demonstrating value.
The questions that arise then are (i) how far are these estimated non-mar-
ket values representative of the broad spectrum of forest values? and (ii) do
the economic value estimates we already have justify the avoidance of
deforestation and/or afforestation?
The issue of value capture is the second stage of the exercise. There is
little point in demonstrating non-market value if those values cannot be
converted into flows of real resources. Otherwise they are just 'paper
money', entries on an albeit interesting balance sheet, but of no practical
significance. The essential reason for this is that the battle over land use is a
battle of real economic values. It is about who makes most profit from the
different ways in which land might be used. As much as we might like the
battleground to be a different one, for example an issue of morality, of
ethical norms, it is the conflict of economic values that defines the real
world and, given world population growth and a finite supply of land, it is
a conflict that will remain for a very long time. Thus non-market value
must be captured and appropriated. Not only must there be a conversion
to real resource flows, but at least part of that flow must accrue to those
who surrender a preferred land use for the sake of non-market values. If,
What do we know about non-market values? 257
for example, the world wants forests as carbon stores, the world must pay
those capable of providing that function but who currently receive no
income from conserving carbon-store functions. Such a blanket statement
hides many complex issues. For example, by highlighting carbon values we
are saying that forest owners have an asset they did not hitherto realize
they had. This opens the way for them to threaten its destruction unless
they are compensated for foregoing timber benefits in favour of carbon-
store benefits. This 'threat' context is germane to the evaluation of com-
pensatory mechanisms involving payments for the incremental cost of
conserving non-market values (see Cervigni and Pearce, 1995). Capture
mechanisms vary widely but they certainly involve definition and enforce-
ment of property rights where their absence or insecurity makes it impossi-
ble for landowners or users to capture non-market values. Beyond that,
markets need to be created through, for example, forest tourism, optimized
entry charges, debt-for-nature swaps, transferable development rights, and
even global benefits markets through e.g. the Global Environment Facility
and joint implementation for carbon reduction.
The model of economic valuation is summarized in terms of a flow dia-
gram (Figure 9.1).
9.2 WHAT DO WE KNOW ABOUT NON-MARKET VALUES?

The task of summarizing the state of the art in estimating non-timber val-
ues in the tropical forest context has been made comparatively easy by a
number of recent surveys: Godoy et al. (1993); Pearce and Moran (1994);
Southgate (1996); Lampietti and Dixon (1995). There are of course sub-
stantial difficulties in reaching general conclusions, primarily because
appropriate guidelines for carrying out such studies, such as those set out
by Godoy et al. (1993) and Godoy and Lubowski (1992) have not been
followed. The result has been a mixture of legitimate and illegitimate valu-
ation procedures. The types of mistake made have included generalization
from studies of a small area of forest to wider areas, with little regard for
the fact that the area in question will not be typical of the whole forest area
simply because of variations in distance to market, and ignoring the fact
that, in a hypothetical world where the whole forest was exploited for non-
timber products, the prices, and hence the profitability, of non-timber pro-
duction would fall. (A rare attempt to secure spatially varied valuations
using geographical information systems for Rio Bravo in Belize is made by
Eade, 1995.)
Another methodological issue is the extent to which values are based on
maximum sustainable yield or on actual harvests, which are often very
much less, i.e. the values that emerge are sensitive to what is assumed
about the management regime in place. Godoy et al. (1993) also point out
that some studies value the stock and some the flow, the former being an
Demonstrate value
of non-timber functions
Compare to alternative
land use value
Bigger?
I
+
Yes
+
•
No
Capture economic value
*
Consolidate Create
property rights markets
Figure 9.1 The model of economic valuation.
interesting measure for wealth accounting but of little value when compar-
ing competing land-use values. Studies also vary as to whether they report
revenues, or revenues net of labour and other costs. Finally, little account
has been taken in many studies of the extent to which the relevant non-
timber activity is itself sustainable, so that what is being compared may
well be two non-sustainable land-use options.
Lampietti and Dixon (1995) divide non-timber values into extractive,
non-extractive and preservation values. Extractive values involve an actual
harvest, e.g. of nuts or rattan. Non-extractive values should be more cor-
rectly titled non-extractive-use values as they involve use but not harvest of
the forest. They include recreation and tourism, but also the indirect eco-
logical functions of forests such as watershed protection and carbon
storage. Preservation values are what most now call non-use or passive-use
values.
9.2.1 Extractive values

Taking extractive values first, Lampietti and Dixon (1995) note that most
studies relate to Central and South America (14 studies out of 20
analysed). Average per hectare values are US$86-101 for Central and
South America and US$60-65 for Asian countries. The Central and South
American results are exaggerated by the Peters et al. (1989) study which
has been severely criticized (Godoy et aI., 1993; Southgate, 1996). Godoy
et al. (1993) report 23 different estimates from studies which only partially
overlap with those reported by Lampietti and Dixon (1995). The authors
resist the temptation to average the results as they are more concerned to
identify differences in methodology and errors as factors accounting for the
variation in values. Ignoring the caveats, an average of US$50lha/year is
obtained (Pearce and Moran, 1994). Some more recent studies suggest
higher extractive values. Thus, Adger'et al. (1995) report values from just
2 US cents/ha up to US$1537 for te'lom grove (groves in rain forests) man-
agement and coffee growing, and around US$6/ha for pharmaceuticals in
Mexico. Boja (1993) reports extractive values in private woodlands in
Zimbabwe of US$39/ha. An extensive study by Kramer et al. (1995) of the
Mantadia National Park in Madagascar found that villagers would lose
around US$91 per household per year from foregone forest products (rice,
fuel wood, crayfish, crab, tenreck and frogs). This converts to just
US$3.2lha ($91 X 351 households = $31 941 across an area of 9875
hectares = $3.2lha). In contrast, Houghton and Mendelsohn (1996) find
present values of fodder, fuel wood and timber (mainly the first two) of
US$2200-3600lha for the Nepalese Middle Hills, or around
US$176-288lha in annuity form (at a 5% discount rate).
With regard to pharmaceutical products, the subject of extensive debate,
Pearce and Puroshothaman (1995) suggest values of US$0.01-21lha/year,
based on established probabilities of finding a successful drug from plant
species currently at risk. This assumes a tropical forest area of 1 billion ha.
Ruitenbeek (1992) has rough estimates of medicinal plant value in the
Korup forest, Cameroon, which translate to around US$0.2-0.7lha. Using
a very different approach, Simpson et al. (1994) suggest that, taking an
optimistic point of view, a pharmaceutical company's willingness to pay
would be a maximum of US$20lha in Western Ecuador and very much
less, perhaps US$lIha, elsewhere. Thus, adopting different approaches,
these studies produce very low values for pharmaceutical values.
Mendelsohn and Balick (1995) suggest a value of undiscovered tropical
forest drugs to the pharmaceutical companies of US$2.8-4.1 billion. They
divide this by 3.1 billion ha of tropical forest to obtain average values of
US$0.9-1.3/ha. The 3.1 billion ha figure appears to be an exaggeration,
whereas the estimate of Pearce and Puroshothaman (1995) appears too
low. Using a figure of 1.7 billion ha of total tropical forest, the Pearce-
Puroshothaman figures would reduce further to a range of nearly zero to
US$12.3lha, and the Mendelsohn-Balick figures would rise to
US$l.6-2.4lha. In contrast, Balick and Mendelsohn (1992) suggest annual
net revenues of US$19-61lha for Belize. The high values of Pearce and
Puroshothaman (1995) reflect values to society rather than values to drug
companies, i.e. values based on lives saved and the value of a statistical life.
The comparable Mendelsohn-Balick figure for social values would be
around 50 times the private willingness-to-pay figure. It seems clear that
pharmaceutical values will not 'save' tropical forests unless the social value
of genetic material is translated into private willingness to pay.
There is an alternative argument which would express non-timber values
as a percentage of household income: Kant et al. (1996) show that house-
hold incomes in West Bengal are increased by 20-30% because of income
from non-timber products, and that the effect is biggest for the poorest
households. Overall, the conclusions on extractivism appear to be that, in
some circumstances, there are high values to be obtained and these may
help the case for conservation. Average values have little general meaning
but, such as they are, the estimates suggest that US$sO/ha may be a very
rough rule of thumb, but there are clearly situations in which higher values
can be achieved and others where US$SO will seriously exaggerate the net
revenues. As a general rule, however, limited faith can be put in non-timber
extractive values to save tropical forests. This contrasts with some of the
grander claims made in the past (Myers, 1984; Peters et aI., 1989).
Southgate (1996) warns against the exaggerated view that South American
extractivists can live by non-timber products alone - they will invariably
deforest as well- and against the assumption that extractivism is inevitably
sustainable. Extractivists tend to be poor. Net returns to vegetable ivory
collection in Ecuador and rubber tapping in the Amazon basin, for exam-
ple, tend to be only just above the opportunity cost of labour. Southgate
(1996; p. 45) concludes that:
, ... although it might provide limited amounts of supplementary income
for forest dwelling populations, commercial non-timber extraction com-
prises a very shaky foundation for an integrated strategy of habitat con-
servation and local economic development".
9.2.2 Non-extractive values

Non-extractive values tend to consist of recreation and indirect ecological
functions such as watershed protection and carbon storage. Of these, recre-
ation and carbon storage have attracted the most study.
Recreation
Adger et al. (1995) suggest ecotourism values for Mexican forests of some
US$8/ha/year, whilst Tobias and Mendelsohn (1991) use the travel-cost
method to obtain values of US$s2/ha for Monteverde in Costa Rica. One
would expect high values for rare ecosystems such as Monteverde. Kumari
(1995) estimated a potential recreational value of M$s7/ha for her study
site in Malaysia, but in present value terms and at 8% discount rate. The
cash flows suggest an annual income of about US$5Iha. For 'conventional'
tropical forest, then, values of US$5-1 Olha might seem appropriate.
Ecological functions
Lampietti and Dixon (1995) find a limited number of studies dealing with
erosion prevention which are capable of estimation of benefits on a per
hectare basis. Magrath and Arens' (1989) study of soil erosion in Java sug-
gests minimum estimates of damage of US$2-7Iha. Cruz et al.'s (1988)
study in the Philippines suggests US$17-28Iha; Ruitenbeek's (1992) Korup
study implies US$141ha for fisheries protection and US$2Iha for flood con-
trol. To these estimates we can add Kumari's (1995) detailed analysis for
Malaysia. This suggests hydrological benefits in terms of conserved agricul-
tural output equal to US$25Ihalyear. Domestic water benefits and fisheries
protection would add a further US$2-3 in each case. Overall, then, water-
shed protection functions do seem to have values which cluster around
US$30lhalyear once a reasonably wide range of functions is considered. (In
contrast, however, Adger et aI's (1995) study for Mexico suggests just 4 US
centslha for watershed protection.)
Carbon storage
Unquestionably the largest value dominating the use values of tropical
forests is that relating to carbon sequestration. The values of Lampietti and
Dixon (1995) for this function are too low due to the adoption of some-
what outdated estimates of the marginal damage from carbon dioxide
releases. All forests store carbon, and if they are cleared for agriculture
there will be a release of carbon dioxide which will contribute to the accel-
erated greenhouse effect and hence to global warming. In order to derive a
value for the 'carbon credit' that should be ascribed to a tropical forest, we
need to know both the net carbon released when forests are converted to
other uses, and the economic value of one tonne of carbon released to the
atmosphere.
Carbon will be released at different rates according to the method of
clearance and subsequent land use. With burning there will be an immedi-
ate release of CO2 into the atmosphere, and some of the remaining carbon
will be locked in ash and charcoal which is resistant to decay. The slash not
converted by fire into CO 2 or charcoal and ash will decay over time, releas-
ing most of its carbon to the atmosphere within 10-20 years. Studies of
tropical forests indicate that significant amounts of cleared vegetation
become lumber, slash, charcoal and ash; the proportion differs for closed
and open forests. The smaller stature and drier climate of open forests
result in the combustion of a higher proportion of the vegetation.
If tropical forested land is converted to pasture or permanent agriculture,
then the amount of carbon stored in secondary vegetation is equivalent to the
carbon content of the biomass of crops planted, or the grass grown on the
pasture. If a secondary forest is allowed to grow, then carbon will accumu-
late, and maximum biomass density is attained after a relatively short time.
Table 9.1 illustrates the net carbon storage effects of land-use conversion
from tropical forests (closed primary, closed secondary or open forests) to
shifting cultivation, permanent agriculture or pasture. The negative figures
represent emissions of carbon; for example, conversion from closed pri-
mary forest to shifting agriculture results in a net loss of 194 t carbon/ha.
The greatest loss of carbon involves change of land use from primary
closed forest to permanent agriculture or pasture. These figures represent
the once-and-for-all change that will occur in carbon storage as a result of
the various land-use conversions.
The data suggest that, allowing for the carbon fixed by subsequent land
uses, carbon released from deforestation of secondary and primary tropical
forest is of the order of 100-200 t carbonlha.
The carbon released from burning tropical forests contributes to global
warming, and we now have several estimates of the minimum economic
damage done by global warming, leaving aside catastrophic events. Recent
work suggests a central value of US$20 of damage for every tonne of car-
bon released (Fankhauser and Pearce, 1994). Applying this figure to the
data in Table 9.1, we can conclude that converting an open forest to agri-
culture or pasture would result in global warming damage of, say,
US$600-1000/ha; conversion of closed secondary forest would cause dam-
age of US$2000-3000/ha; and conversion of primary forest to agriculture
would give rise to damage of about US$4000-4400/ha. Note that these
estimates allow for carbon fixation in the subsequent land use.
There are problems with these values for the indirect carbon storage
functions of tropical forests. Firstly, the science of global warming is uncer-
tain and this suggests that the values need to be multiplied by some unspec-
ified probability that the effects are certain. The Intergovernmental Panel
on Climate Change'S Second Assessment Report (IPCC, 1996) states that
"the balance of evidence suggests that there is now a discernible human
influence on climate" but its quantification is still limited. Put another way,
the context is pure uncertainty rather than risk. Secondly, the value of
US$20/t carbon is itself uncertain. It is the product of a Monte Carlo simu-
lation so that it encompasses a good deal of the uncertainty about impacts
and values, but it does not deal with the potential for surprises or extreme
events (Fankhauser, 1995). Thirdly, even if the values are broadly correct
and global warming is a real phenomenon, the avoidance of deforestation
or investment in afforestation may not be cheapest ways of reducing car-
bon emissions. The opportunity cost of conservation is clearly the develop-
ment benefit foregone, i.e. the returns to forest clearance for agriculture,
timber or livestock. It seems very likely that these foregone values are
indeed very low in many cases. For example, Schneider (1992) reports
Table 9.1 Changes in carbon with land use conversion
Original Shifting Permanent Pasture

carbon agriculture" agriculture
Original carbon
79 63 63
Closed primary 283 -204 -220 -220

Closed secondary 194 -106 -152 -122
Open forest 115 -36 -52 -52
a Shifting agriculture represents carbon in biomass and soils in second year of shifting
cultivation cycle.
Source: Brown and Pearce, 1994.
upper bound values of US$3001ha for land in Rondonia, Brazil. The figures
suggest carbon credit values 2-15 times the price of land in Rondonia.
These carbon credits also compare favourably with the value of forest land
for timber in, say, Indonesia, where estimates are of the order of
US$1000-2000. If land is worth US$3001ha in a development use, then the
cost of conservation on global warming grounds becomes, say, US$3/t car-
bon ($300 divided by 100 tiha, say). If the land is worth US$2000, then
carbon conservation costs US$201t carbon. The latter cost is certainly not
the cheapest way of conserving carbon, and, if some commentators, includ-
ing the IPCC, are correct, even US$3/t carbon could be quite expensive. (It
is not exactly clear what IPCC believes with respect to low-cost, zero-cost
and even negative-cost options for reducing emissions as the statements are
not very clear: they appear to suggest, however, that 10-30% gains in
energy efficiency over and above existing or near-term projected levels are
feasible.) Compared to investments undertaken by the Global Environment
Facility in its Pilot Phase, however, carbon reduction at US$3/t carbon
might be relatively cheap.
9.2.3 Preservation values

The final category of value suggested by Lampietti and Dixon's (1995) sur-
vey is preservation value, by which is meant passive or non-use value. The
only estimate of such value for tropical forests is that of Kramer et al.
(1994), who report average willingness to pay of US citizens for protection
of an additional 5% of the world's tropical forests. One-time payments
amounted to US$29-51 per household, or US$2.6-4.6 billion. If this will-
ingness to pay was extended to all OECD households, and ignoring income
differences, a broad order of magnitude would be a one-off payment of
US$11-23 billion. Annuitized, this would be, say, US$1.1-2.3 billion/year.
Taking 1.7 billion ha as the area for total tropical forest, 5% of it would
come to 85 million ha, so that annual willingness to pay would be
US$13-27Iha. Obviously, the assumptions being made here are fairly
heroic, but they bear comparison with some of the use values identified
above, and also pale into insignificance when compared to the carbon-storage
values.
Pearce (1996) looks at other potential estimates of global value. One
approach is to see what the values for similar assets would imply.
Willingness to pay studies for the conservation of biological resources sug-
gest average payments for the individual of perhaps US$10/year. This
would produce a fund of US$4 billion/year when applied to OECD house-
holds. This would translate to around US$2.3lha if applied to all tropical
forest. An alternative is to look at implicit prices in debt-for-nature swaps,
however the procedure of estimating implicit prices of this kind is open to
some doubt, although it has been used by some writers (see Ruitenbeek,
1992; Pearce and Moran, 1994). The range of implicit values is from
around 1 centlha to just over US$4Iha (Pearce, 1996).
The estimates of non-use value are clearly very speculative and it is not
even clear that the methodologies in question are eliciting non-use rather
than some mixture of use and non-use values. As we have seen, the only
direct approach based on contingent valuation suggests fairly significant
values of US$13-27Iha for a small part of the total forest stock. The more
indirect approaches suggest very much lower values, of perhaps one tenth
of the direct values.
9.2.4 Conclusions on tropical forest values

Extracting some kind of consensus from the above estimates is clearly haz-
ardous. More as a guide for future research than anything else, we can
speculate on the following annual values (US$lha):
Extractive values 50
Non-extractive values:
recreation 5-10
ecological 30
carbon 600-4400
Non-use values 2-27
Whichever way the analysis is done, the major role of carbon values is
revealed. Should, for some reason, global warming not remain a serious
issue of concern, then tropical forests might be found to have measured
environmental value of around US$100Iha, far from enough to justify con-
servation on economic grounds.
References 265
9.3 WILL NON-TIMBER VALUES SAVE THE FORESTS?
If the future of the world's forests rested solely on the demonstration and
capture of non-timber value, we might expect some forests to be saved and
many to disappear. Despite early claims that non-timber values will prove
to be highly significant, and often in excess of the alternative land-
clearance values, non-timber values may well be insufficient to justify the
prevention of deforestation and the planting of new forests. The most
favourable circumstances appear to be where recreational values are high
and where they are consistent with some degree of continued extractivism,
and where carbon-storage values can be appropriated. Pearce (1996)
details the kinds of emerging markets for carbon sequestration and, pro-
vided the scientists do not change their minds about the threat of global
warming, one can expect 'trades' in carbon to grow. Perversely, if the
major environmental threat of global warming is removed, the fate of the
forests would appear to be compromised. But this need not be a matter for
despondency. What it tells us is that the economic arguments based on val-
uation are perhaps less powerful than we originally thought. But, even if
we discount the moral arguments for forest conservation, and they appear
not to have prevented an accelerating trend in deforestation, it remains the
case that a substantial amount of deforestation arises from perverse incen-
tives such as subsidies to land clearance, and from insecure or barely
defined property rights. Pressures on state revenues may ultimately see the
former decline, whilst the need to protect land against a rapidly rising pop-
ulation should force the pace on conferment of property rights. If so, the
future of the forests is reasonably secure. Let us hope so.
ACKNOWLEDGEMENT
An earlier version of this paper was presented at the International
Symposium on the Non-Market Benefits of Forestry organized by the
Forestry Commission in Edinburgh, UK, June 1996.
REFERENCES
Adger, N., Brown, K., Cervigni, R. and Moran, D. (1995) Total Economic Value of
Forests in Mexico. Ambio, 24, 286-296.
Balick, M. and Mendelsohn, R. (1992) The Economic Value of Traditional
Medicine from Tropical Rain Forests. Conservation Biology, 6,128-139.
Bojo, J. 1993. Economic Valuation of Indigenous Woodlands, in Living With Trees:
Policies for Forest Management in Zimbabwe, (eds P. Bradley and K.
McNamara), Technical Paper 210, World Bank, Washington, DC.
Brown, K and Pearce, D. W. (1994) The economic value of non-market benefits of
tropical forests: carbon storage, in The Economics of Project Appraisal and the
Environment, (ed. J. Weiss), Edward Elgar, London, pp. 102-123.
Brown, K., Pearce, D. W., Perrings, C. and Swanson, T. (1993) Economics and the
Conservation of Global Biological Diversity, Working Paper No.2, Global
Environment Facility, Washington, DC.
Cervigni, R. and Pearce, D. W. (1995) North-South Resource Transfers,
Incremental Cost and the Rio Environment Conventions, Centre for Social and
Economic Research on the Global Environment, University College
LondonlUniversity of East Anglia (mimeo.).
Cruz, W., Francisco, H. and Conway, Z. (1988) The on-site and downstream costs
of soil erosion in the Magat and Pantabangan watersheds. Journal of Philippine
Development, XV, 85-112.
Eade, J. (1995) Spatial Economic Valuation: Benefits Transfer Using Geographical
Information Systems, World Conservation Monitoring Centre, Cambridge
(mimeo.).
Fankhauser, S. (1995) Valuing Climate Change, Earthscan, London.
Fankhauser, S. and Pearce, D. W. (1994) The social costs of greenhouse gas emis-
sions, in The Economics of Climate Change, OECD, Paris, pp. 71-86.
Godoy, R. and Lubowski, R. (1992) Guidelines for the economic valuation of non-
timber tropical forest products. Current Anthropology, 33, 423-433.
Godoy, R., Lubowski, R. and Markandya, A. (1993) A method for the economic
valuation of non-timber tropical forest products. Economic Botany, 47,
220-233.
Houghton, K. and Mendelsohn, R. (1996) An economic analysis of multiple use
forestry in Nepal. Ambio, 25, 156-159.
Hyde, W., Newman, D. and Sedjo, R. (1991) Forest Economics and Policy
Analysis, World Bank Discussion Paper 134, World Bank, Washington, DC.
IPCC (1996) Intergovernmental Panel on Climate Change, Second Assessment
Report, Volume I, Cambridge University Press, Cambridge.
Kant, S., Nautiyal, S. and Berry, R. (1996) Forests and Economic Welfare. Journal
of Economic Studies, 23, 31-43.
Kramer, R., Mercer, E. and Sharma, N. (1994) Valuing Tropical Rain Forest
Protection Using the Contingent Valuation Method, School of the Environment,
Duke University, Durham, North Carolina (mimeo.).
Kramer, R., Sharma. N. and Munasinghe, M. (1995) Valuing Tropical Forests:
Methodology and Case Study of Madagascar, Environment Paper No. 13, World
Bank, Washington, DC.
Kumari, K. (1995) An Environmental and Economic Assessment of Forest
Management Options: A Case Study in Malaysia, Environment Department
Papers No. 26, World Bank, Washington, DC.
Krutilla, ]. (1967) Conservation reconsidered. American Economic Review, 57,
777-786.
Lampietti,]. and Dixon, J. (1995) To See the Forest for the Trees: A Guide to Non-
timber Forest Benefits, Environment Department, World Bank, Washington, DC
(mimeo.).
Magrath, W. and Arens, P. (1989) The Costs of Soil Erosion on Java: A Natural
Resource Accounting Approach, Working Paper No. 18, Environment
Department, World Bank, Washington, DC.
Mendelsohn, R. and Balick, M. (1995) The value of undiscovered pharmaceuticals
in tropical forests. Economic Botany, 49, 223-228.
References 267
Myers, N. (1984) The Primary Source: Tropical Forests and our Future, W. W.
Norton, New York.
Panayotou, T. and Ashton, P. (1993) Not by Timber Alone: Economics and
Ecology for Sustaining Tropical Forests, Island Press, Washington, DC
Pearce, D. W. (1996) Global environmental value and the tropical forests: demon-
stration and capture, in Forestry, Economics and the Environment, (eds W.
Adamowicz, P. Boxall, M. Luckett, W. Phillips and W. White), CAB
International, Wallingford, UK, pp. 11-48.
Pearce, D. Wand Moran, D. (1994) The Economic Value of Biodiversity,
Earthscan, London.
Pearce, D. Wand Puroshothaman, S. (1995) The economic value of plant-based
pharmaceuticals, in Intellectual Property Rights and Biodiversity Conservation,
(ed. T. Swanson), Cambridge University Press, Cambridge, pp. 127-138.
Perlin, J. (1989) A Forest Journey: The Role of Wood in the Development of
Civilization, W. W. Norton, New York.
Peters, C, Gentry, A. and Mendelsohn, R. (1989) Valuation of an Amazon rainfor-
est. Nature, 339, 655-656.
Ruitenbeek, J. (1992) The rainforest supply price. Ecological Economics, 1,57-78.
Schneider, R. (1992) Brazil: An Analysis of Environmental Problems in the
Amazon, Report 9104-BR, World Bank, Washington, DC
Sedjo, R. and Lyon, K. (1990) The Long Term Adequacy of World Timber Supply,
Resources for the Future, Washington, DC
Simpson, D., Sedjo, R. and Reid, J. (1994) Valuing Biodiversity: An Application to
Genetic Prospecting, Discussion Paper 94-20, Resources for the Future,
Washington, DC
Southgate, D. (1996) What Roles Can Ecotourism, Non-timber Extraction,
Genetic Prospecting, and Sustainable Timber Play in an Integrated Strategy for
Habitat Conservation and Local Development? Report to Inter-American
Development Bank, Department of Agricultural Economics, Ohio State
University (mimeo.).
Swanson, T. (1994) The International Regulation of Extinction, Macmillan,
London.
Tobias, D. and Mendelsohn, R. (1991). Valuing ecotourism in a tropical rain-forest
reserve. Ambio, 20,91-92.
-10------
The role of policy and institutions
James Mayers and Stephen Bass
Forestry and Land Use Programme, International Institute for
Environment and Development (lIED), 3 Endsleigh Street, London
WC1HODD, UK
10.1 WHY POLICY MATTERS: DEALING WITH MULTIPLE

INTERESTS AND CHANGE
People have very different interests in tropical forests. Whilst some assert
that forests are needed to store carbon, others are looking to carve off a
piece of forest to convert into a farm. Interests in forests range from those
which are central to an individual's livelihood or a corporation's viability,
to the opinions and aesthetic preferences of those who live far away.
Interests may be backed by strong political influence, legal rights and
resources, or by none of these. The actual impact of these different interests
on forests therefore varies from place to place. The job of policy includes
balancing such interests, and charting paths to solve problems created by
the excessive influence of some, and by the lack of influence of others. This
chapter examines how policy, and the institutions charged with developing
and implementing it, currently fare in the rapidly changing context of trop-
ical forests.
Replacement or alteration of forests has produced economic and social
benefits in many countries. In particular, the clearance of forest for agricul-
ture has been looked upon as advancing possibilities for development,
either by increasing food security or by providing cash crops to fuel income
growth. Yet there are environmental and social costs of forest use or
replacement.
The wasting of tropical forest assets is occurring on an international
scale, and the causes of this are to be found at all levels: local, national and
global. The problem is that those who reap the benefits generally do not
pay the costs. It is ordinary people in developing countries who bear the
brunt of most of these costs.
270 The role of policy and institutions
Getting policy right is a particularly important challenge for the next few
years. Current policies in many countries send signals that favour only a
few interests in forest use or replacement. They do not require these inter-
ests to cover the associated environmental and social costs. Furthermore,
stakeholder interests and perceptions are not static, and there will be envi-
ronmental and economic change beyond the control of stakeholders; yet
policy processes are not well set up to make decisions about how to deal
with rapid change.
BOX 10.1: FOREST PROBLEMS

Deforestation in tropical forests is calculated by the FAO to have con-
tinued throughout the 1980s at an average rate of 15 million ha (0.8%
of forest cover) per year. Millions of additional hectares were degraded
by inappropriate logging practices, fuel wood collection and pollution.
Whilst these immediate effects are on trees, problems are caused for dif-
ferent groups of people.
The main physical forest problems are:
• Declining quantity and quality of forests and forest products. This is
because wood, fuel and food are being harvested at rates faster than
forest regeneration; because remaining growing stock is often poorly
managed; and because many forests are being cleared to make way
for other land uses.
• Environmental degradation of forest areas. Forest exploitation and
clearance can create interlinked problems: notably soil erosion,
watershed destabilization and micro-climatic change. These threaten
the soil and water base for agriculture. Industrial pollution and poor
waste management also reduce forest health.
• Loss of biodiversity. These problems are contributing to a rapid
reduction in ecosystem, species and genetic diversity. This lowers the
world's biological potential for improving material, food and medi-
cine production. With tropical forests being perhaps the major repos-
itory of biodiversity, forest abuse in tropical regions is causing
particular concern.
• Climate change. Forests play a major role in carbon storage. It is
probable that the cumulative effect of global forest loss and environ-
mental degradation will be a net contribution to regional and global
climate change. This could bring many problematic side-effects: sea-
level rise, risks to food supply, declining productivity of soils and the
quality and quantity of fresh water, inviability of current land uses
and protected areas, and human health problems.
Why policy matters 271
The main costs for people are:

• Loss of cultural assets and knowledge. The culture and knowledge of
many peoples, which have evolved through long periods of nurturing
the forest, and which are not always documented, are diminishing as
forest area, access and traditional rights are reduced.
• Loss of security of forest goods and services. Different stakeholders
will find it relatively difficult to meet their needs for forest products,
especially those who do not have access to marketed produce.
• Loss of livelihood. Forest loss and degradation are affecting the
livelihoods of forest-dependent groups, particularly poorer groups
who may not have significant agricultural land, and who depend on
forests for 'social security'.
• Inequality. Increasing concentration in fewer hands of control of and
access to forest wealth is removing development options for the
majority of people in many countries. Those who lose their forest
livelihoods and become marginalized may be forced to create social
and economic problems elsewhere, for example in cities.
It is important to distinguish the direct agents of forest problems from

their underlying root causes. Policy and institutional failures are at the root
of many forest problems.
The direct agents which bring about problems for tropical forests
include: slash-and-burn cultivation by small-scale farmers; clearance and
burning for grazing; settled agriculture and cash crop production; timber
harvesting; wood fuel harvesting; clearance for forest plantations; infra-
structure construction (mining, dams, roads, etc.); and air or water pollu-
tion. Often, several factors work together. For example, road development
encourages timber exploitation, which opens the forest for agricultural set-
tlement and fuel wood salvaging. About half of all logged tropical forests
are eventually cleared by farmers. The importance of these agents may dif-
fer, according to region. In Africa, the major agent of deforestation is shift-
ing agriculture after timber harvesting; in Latin America, agriculture often
precedes logging operations; in South-East Asia, logging may be the central
agent of deforestation due to its intensity.
The root causes of forest problems are complex. They often arise far
away from the forest itself. These are the policy, market and institutional
'signals', to which the direct agents of deforestation and forest degradation
respond. Many of these signals arise at national level, either within the
forestry sector or outside - notably in agriculture and industrial develop-
ment. Or they are a result of national macro-economic policies. Other sig-
nals are set at local levels, often due to fundamental inequities in rights and
access to resources. Further signals are set internationally, such as trade
policies, debt and structural adjustment requirements.
These signals either 'push' groups into the forest, by creating difficult
economic or social conditions outside the forest; or they 'pull' groups into
the forest, by attracting them with excess profits. Many policies and mar-
ket conditions effectively undervalue forests. That is, they do not provide
long-term incentives for the groups 'pushed' or 'pulled' into forests to look
after forest assets.
BOXI0.2: POLICY AND INSTITUTIONAL FAILURE IS AT THE

ROOT OF MANY FOREST PROBLEMS
Policy failures
• Policies do not signal scarcity of all forest goods and services, and do
not require markets to do so
• Undervaluing forests e.g. low stumpage (Ghana approximately 1 %),
low land price (Costa Rican forest, worth US$20001ha locally, sells
for US$500Iha), tax breaks on logging (Russia)
• Overvaluing forest removal e.g. subsidized agricultural pnces
(Brazilian agricultural loans, input subsidies and tax breaks)
• Social and environmental costs are not included in prices, nor in
measurement of economic growth (in Costa Rica in 1989, US$400
million counted as agricultural income was, in fact, natural resource
depreciation)
• Investments in sustainable forest management are rendered unprof-
itable or risky e.g. high interest rates and lack of long-term financial
stability
• Trade terms are against primary production (typically the primary
producer receives only 10% of the finished goods price)
• Conflicting foreign policies (international demands for debt payment
and structural adjustment clash with international policies promoting
conservation)
Institutional failures
• Poor information and monitoring of forest stocks and flows, and the
demands they face
• Weak means to reconcile differing priorities between stakeholders -
different demands on forests; and different perceptions about
good/sustainable forestry
• Poor participation and consensus in seeking solutions acceptable to
most stakeholders
• Uncoordinated decision-making
• Unclear or outdated institutional roles
Policy influences on forests 273
• Lack of learning mechanisms and skills in forest authorities for han-

dling multiple interests
• Government control mechanisms either too weak or too strong, such
as poorly-handled decentralization
• Misdirected international assistance which merely reinforces the
above
• Inequity between and within nations, in access to forest benefits, and
to resources for their management and use
• Misapplied wealth and poor control of major corporations
• The 'invisible' institution - covert interactions amongst individuals
determine many inequitable outcomes
• Lack of political will - influential stakeholders unwilling to compro-
mise between their and others' demands on forests
• Low political influence of forest authorities in relation to other land
use authorities and some forest users; and consequently few resources
at their command
Weak/inappropriate tenure and rights and their conflicts with policies
• Tenure systems which encourage deforestation in order to obtain title
(e.g. Latin American land 'improvement')
• Tenure systems which discourage afforestation in order to keep title
(e.g. West Africa)
• Governments enforcing tenure extremes, such as nationalization or
privatization; and not recognizing property rights continuum
• Poor recognition of access and user rights for the landless and people
with traditional claims to forests
10.2 POLICY INFLUENCES ON FORESTS

There is a range of policies which may influence the ways in which land
stays as forest, is deforested, or is afforested, and how forests are used
(Figure 10.1). Forest policy itself is just one of these; in many places it is
becoming relatively insignificant. Other policy influences on forests may be
far more powerful, such as those concerning government revenues, agricul-
tural pricing or land ownership. Indeed the nature of forest policy may be a
consequence of the dominance of these other policies.
Tropical forests in some countries are experiencing profound changes
that are not affected one way or another by formal forest policy interven-
tions, whilst macro-economic policy influences are fundamental. For exam-
ple, in Costa Rica, historically a 'frontier' society like many others in Latin
America, removal of trees was necessary to secure land tenure. The frontier
has now gone but a web of conflicting land claims continues to encourage
Macroeconomic
policies Employment
Policies affecting
land use
Policies influencing
demand for forest
products
Concessions
Forest management policy
policies
Figure 10.1 Types of policy influences on forests and people.

BOX10.3: MIXED SIGNALS IN COSTA RICA

Policies weighted in favour of pasture and agriculture relative to forest
management
Forest law is based on regimen forestal - a contractual relationship
between government and landowner. In theory, government protects
and assists landowners who maintain their land under forestry, but in
practice the regimen dictates cumbersome obligations, including logging
permits, over-complicated management plans and mandatory reforesta-
tion. The combination of insecure land title and the possibility of gov-
ernment regulation or intervention (expropriation for protected areas or
imposition of forestry regimen) has many impacts including:
• burning forest to avoid intervention
• lower value of forest land compared to cleared land
• illegal logging
• evasion of forest taxes
• costly and ineffective attempts at enforcement.
Other policies weighted against forest management include:
• Land titling. Despite the passing of 'frontier' approaches to forest
use, tree felling is still recognized as a means to secure land holdings.
• Livestock credit. Still represents about 25 % of all agricultural credit
and has benefited from debt rescheduling.
• Log export bans. The 1987 ban on exporting logs and unprocessed
lumber depressed domestic prices, leading to conversion of forest
lands at the margin, lower reforestation, wasteful milling etc.
• Agricultural export preferences. Policies in this area have established
Costa Rica's strong comparative advantage in banana growing
Problems with economic incentives for forestry

Economic incentives have been the Government's major policy tool in
forestry since the 1970s. They include subsidies for plantation develop-
ment and natural forest management in the form of income tax deduc-
tions, subsidized credit schemes and transferrable reforestation bonds.
The combined effect of these include about 97 000 ha reforested since
1973 and the involvement of about 57 farmers' organizations in bond
schemes. However there are problems:
• large landowners have benefited most but do not need the incentives
as they will invest in plantations without subsidy
• bonds may contribute to conversion of natural forests at the margins,
and plantations established are low in diversity
• incentives are not tied to other benefits of forests - plantations maxi-
mize income not social benefits of forests
• distorting macroeconomic policies continue to make incentives

poorly effective
• high fiscal cost
A new Forest Law in 1996 deregulates many forest activities and intro-
duces a new incentive for natural forest management based on compen-
sating forest owners for maintaining environmental values. Hence
incentives are more strongly linked to environmental performance.
However, some of the conflicts with pasture and agriculture policy
remain (Watson et ai, forthcoming).
tree felling as a means to establish secure land holding. Other policies, such
as credit availability, still promote clearance on lands which may lack the
capacity for sustained agriculture or livestock.
In many countries, however, the workings of the forest sector have had
profound impacts on the condition of forests, and indeed on the nature of
society. Most tropical countries have 'formal' forest policies, although until
relatively recently, many of these were not formally expressed and could
only be implied from the operations and decisions of forestry institutions.
The following paragraphs outline a number of strikingly common themes
in the history of these forest policies.
Many current forest policies were defined to serve narrow and static ends,
and by-and-Iarge concentrated on government control of forest land. They
tend to be centralized expressions of government intent prepared by profes-
sionals, with little participation beyond a few powerful interests, and no
broad consensus. Forest policies have treated forestry as a 'sector' only, not
dealing with complex systems and, as noted above, not integrated with
other policies. These have largely concentrated on forests as a land reserve
for protection and ultimate allocation to agriculture, and on timber produc-
tion. Many such policies were instituted by colonial powers; their current
relevance is now often questionable. A large government-controlled land
base was the basis of forest authority power in British, French and Dutch
colonies. To this day, forest departments are often in nominal control of
huge areas of land which have been designated as forest 'reserve', for gener-
ating national revenue or for later land allocation. For example, in India,
the country on which many British colonial forest policies were developed,
the forest authorities have formal responsibility for 22% of national land.
In Indonesia, the comparable figure is about 74% (Hobley, 1996a).
To create such areas, local people were generally divested of control over
land. Later, forest areas were leased out to private industry as logging con-
cessions. Today, the relatively small bureaucracies charged with administer-
ing and controlling the forests are often poorly able to control logging and
prevent public access. The resulting combination of open access and
destructive logging can lead to uncontrolled exploitation. Governments
often then respond to these pressures as they arise by imposing more for-
mal control, which they have little capacity to exert.
In this process, the potential partners in multiple-use management
become marginalized, leaving little to build on but government control
mechanisms. Meanwhile the policy bias against local people and in favour
of timber-based economies has handed de facto control over many forest
areas to the large-scale private sector. 'Patronage politics', which tends to
pervade the tropical forest industry, has not only served to ensure that the
benefits are captured by very few in society, but it has also severely weak-
ened the evolution of democratic institutions (Colchester, 1994). For exam-
ple, a comprehensive judicial inquiry into the forest industry in Papua New
Guinea, in 1988-89, revealed widespread malpractice and a decay in stan-
dards of public service due to the influences of the logging industry.
Subservience of national and local government 'policy implementers' to
external corporate interests and capital owners is a particularly common
theme. Throughout the tropics (and in temperate areas) big companies are
able to make alliances with government officials to shape policy in their
favour, or to over-ride or undermine it. In Guyana and Suriname, South-
East Asian-owned logging companies have recently negotiated forest con-
cessions over several hundred thousand hectares with massive tax breaks
on their investments (Sizer and Rice, 1995; Sizer, 1996). Thus, 'actual' pol-
icy - that is, the decisions which define the trade-offs in practice - often
reflects power structures and domination by certain actors and interests.
In Indonesia, a hugely powerful timber industry was created by policy
interventions, ostensibly to stimulate the economy in the 1980s. About 25
wealthy 'timber kings' control the timber industry and are amongst the
strongest players in setting Indonesian forest policies. Their wealth depends
on cheap timber supplies (Barber et ai, 1994). This leads them to exert con-
siderable force against any policy initiative aimed at redistributing benefits
more equitably and towards collaborations between groups who might be
in a position to manage forests more sustainably. In the 1990s, policy inter-
ventions have been opening the way for a major pulp/paper industry
dependent in large part upon the clearance of rain forests (lIED, 1996).
Figure 10.2 compares the potential log export value for two case studies,
Indonesia and Ghana.
The part played by industrial forestry in triggering forest degradation
and creating social problems varies from country to country and is often
linked to other contributory factors. A few forestry companies are demon-
strating that timber harvesting from tropical forests can be well-managed
and meet recognized standards of quality forest management. Currently,
however, companies practising good forest management are still in the
minority, whilst some of the worst are some of the biggest, with the great-
est influence on policy.
Potential log export value Distribution of stumpage value
US$145/m 3 (weighted mean log export value) US$95/m3 (Total value - US$2.4 billion)
Normal return to Ucence fee, property

capital (30% 0/ tax and royalty
logging cost)
Reforestation fee
"Unrealized rent"
(Total - US$2 billion)
(a)
Potential log export value Distribution of stumpage value

US$161/m 3 (weighted mean log exportvalue) US$96/rn3 (Total value - US$105 million)
Normal return to capital
(20% of logging cost) Profits, informal
Service charges 40;' payments,
(port, documents etc.) 0
unexplained
......... ' Unrealized

rent" (Total -
........ US$I00m)
Wastage in forest
and in processing
Stumpage value
(potential economic rent)
(b)
Figure 10.2 Where does the log value go? (a) Indonesia, 1990 (data from Barber et
al., 1994; (b) Ghana, 1991 (data from Richards, 1995).
Formal planning, legislative and regulatory mechanisms are overly com-

plex, incoherent and therefore do not function very well. There are, for
example, 22 volumes of forest laws in Indonesia, many of them overlap-
ping and contradictory. This mass of planning and regulation is often so
impenetrable that few can make sense of it, and powerful interests can
manipulate it.
The closeness of powerful interest groups to national 'policy-makers'
partly explains the continuing gulf between local forest actors and national
'policy-making'. National policy rarely seems able to respond to local
dynamics and innovations. On the other hand, the decisions of local orga-
nizations and projects are frequently not made with a view to being consis-
tent with national policy priorities. 'Meso' level institutions can sometimes
make the linkage, particularly those which are engaged in supporting local
Policy challenges for sustainability - national and international 279
BOX 10.4: PRIVATE BENEFITS, SOCIAL COSTS: CORPORATE

INFLUENCES ON POLICY IN PAPUA NEW GUINEA
A judicial inquiry into the forest sector in the late 1980s produced 21
volumes of detail on malpractice in the largely foreign-owned forest
industry and in government. This led to wide-ranging national debate
on the 'forest crisis' and to the National Forestry and Conservation
Action Programme involving much proposed policy and institutional
reform. However, the timber industry resisted at every turn, notably one
company, Rimbunan Hijau, which:
• controls between 60 and 80% of the timber market through a net-
work of affiliated companies;
• formed an association with timber permit landowner 'representa-
tives' to oppose proposed changes in national forest policy;
• started a second national daily newspaper, and supported a concerted
media campaign against various progressive policy initiatives;
• persistently flouted the law, e.g. in disregarding Department of
Labour requirements for employment and training, not adhering to
environmental conditions, and in ignoring orders from law courts to
stop logging activities in particular areas.
"They have their fingers in every pie, they can block or change every
move we make"
(General Manager of the National Forest Authority, personal communi-
cation, 1995.)
livelihoods and in recognized policy processes, such as local authorities and

recognized developmental non-governmental organizations (NGOs).
Successful examples in forestry remain rare: some NGOs have played this
role effectively in the development of joint forest management in India
(Saxena, 1996), but in many countries these are poorly developed.
10.3 POLICY CHALLENGES FOR SUSTAINABILITY - NATIONAL

AND INTERNATIONAL
Many of the old forest policies are disintegrating in today's context. In the
light of increasing demands for multiple forest goods and services, govern-
ment foresters' attempts to control the division of rights and responsibili-
ties are increasingly challenged. Forestry's traditional domain, management
for wood and trees, is just one use of forest land. There are many other
'layers' of use, each with its associated rights and responsibilities, linked to
forest goods and services such as wildlife, medicines, water, soil, local
climate stability and cultural environment (Hurditch, 1996).
Policy which fosters sustainable forest management will have to accom-
modate such multiple layers of use, and must attempt to balance the ele-
ments of several key dimensions: economic, social and environmental
objectives; global, national and local demands; and present and future con-
ditions. Changes will be of greater magnitude, speed and uncertainty. Yet
typical forest policy attempts to deal with certainties, not change.
Integrating these different interests, making appropriate trade-offs
between interests where necessary, and dealing with uncertainty, is a com-
plicated balancing act. Nowhere can sustainable forest management simply
be 'switched on' through re-writing policy statements. Rather, phased
changes in policy formulation and implementation processes may be
needed which enable progress to be made:
• from little forest management to informed management;
• from single use/user to multiple use/user;
• from government domination to civil society management;
• from coercive means to incentives and informational means;
• from an injudicious use of forest to obtain products to a mix of forests,
imports and substitutes.
Flexible processes will be required which allow adaptive change in policy,
and continuous improvement in forest management, by:
• enabling genuine participation in defining forestry goals, and re-negotia-
tion of rights and responsibilities;
• building on local knowledge and values;
• optimizing external inputs;
• allowing for uncertainties;
• monitoring and learning.
Figure 10.3 illustrates a conceptual framework for achieving sustainable
forest managements and thereby securing a mix of forest goods and ser-
vices. This will be described further in section 10.5. Some policy progress
has been made towards such a framework in many countries. Since con-
cern for the perceived problems of tropical forests emerged as a major
international issue in the 1980s, some national and international initiatives
have tried to identify the elements of sustainable forest management and to
capture this in policy (Humphreys, 1996). In the 1980s and early 1990s
these efforts generally focused on a continuous yield of timber because the
social, economic and political dimensions of sustainability were so difficult
to pin down. They also tended to imply top-down (including supra-
national) control of forests to achieve sustainability.
In the later 1990s, at the international level, bodies such as the
International Tropical Timber Organization (lITO), an inter-governmental
association, and the Forest Stewardship Council (FSC), an industry-NGO
Policy challenges for sustainability - national and international 281
Setting goals
for forests
Information Planning/
systems institutions
Effective national forest
strategies are:
- Participatory
- Cyclical and
continuously improving
Sustainable
Monitoring forest
management
in the field
Figure 10.3 Sustainable forest management: processes for its achievement.
alliance, have made considerable progress in developing consensus on some

common elements of sustainable forest management. These debates are
highly charged as very varied perspectives are involved (Thomson, 1996).
It has become clear that there are no universal blueprints for sustainable
forest management - rather, there are certain principles, the articulation of
which needs to be locally agreed. Any workable policy and institutional
structure aiming to bring it about must be negotiated within the particular
circumstances of a nation, and be compatible with its range of stakeholder
perspectives, norms of participation and decision-making. Notions of
supranational control have been replaced by a general acceptance that
means are required to ensure that local stakeholders can be effective stew-
ards of forests, with society's best interests at heart. It will be some years
before nations can agree the elements that should be subject to suprana-
tional regulation through e.g. a forests convention (Bass and Thomson, 1997).
Nonetheless, some international agreement on the principles and criteria
common to sustainable forest management everywhere is needed.
Markets and international obligations increasingly demand wise forest use.
Monitoring of such achievement at forest level (for the sale of certified
products) and at national level (for possible ultimate compensation for
producing 'global' benefits such as carbon storage, and for information
sharing on national progress towards sustainable forest management) all
require means of mutually recognising locally different approaches to sus-
tainable forest management. Hence the current lITO and FSC emphases
on principles and criteria. The history of major international forest policy
initiatives to date is presented in Table 10.1.
10.4 INSTITUTIONS AND STAKEHOLDERS

Because better decision-making is at the core of better policy processes for
forest management, we need to understand the institutions in which deci-
sion-makers operate. The strengths and weaknesses of institutions can be
looked at in two main ways: their functional formal mandates; and their
internal dynamics and characteristics.
10.4.1 Institutional roles - who should do what?

As already noted, formal forestry institutions in tropical regions were
mostly founded on policies to control a large government land-base in the
form of forest reserves for revenue generation through industrial activities.
These policies were generally based on those developed for the very differ-
ent context of European forests (which were generally privately owned and
largely uninhabited). They were applied to India in the 19th century, and
then throughout the British Empire and colonies of other European coun-
tries (Shepherd, 1992). As former colonies became nation states, industrial
forestry continued to be prioritized, particularly as foreign exchange earn-
ing/saving objectives became key.
Typically, government forest agencies combine elements of:
• authority - setting regulations;
• user - with its own commercial enterprise;
• monitor - checking on other users and controlling information.
In many countries, these functions are either in mutual conflict or are inad-
equately performed through lack of capacity (Morrison and Bass, 1996).
Key questions which need to be addressed in such contexts include the fol-
lowing.
• For which functions do overriding national or global interests demand
that government is in charge? Securing inter-generational benefits of
forests and conserving significant biodiversity are commonly stated
examples.
Institutions and stakeholders 283
• Which functions are incompatible, requiring their separation or special
procedures to minimize conflict and ensure transparency? A typical
example of this is the housing under one roof of regulatory and enter-
prise functions of state forestry agencies such as commissions.
• For which functions does government have a comparative advantage;
and conversely, where are there other more suited bodies, e.g. in the pri-
vate sector, communities or NGOs?
Three partners, the private sector, communities and government are
increasingly recognized as necessary for achieving sustainable development
- the development triad, as lIED terms it. But these partners are not well
integrated into policy processes, except by virtue of strong lobbies. State-
run commercial enterprises have been vital revenue sources in some tropi-
cal countries, but have become anachronistic in the face of more effectively
run private-sector enterprise. NGOs can be much more effective than gov-
ernment in monitoring and 'watchdog' functions. Furthermore, NGOs can
lever change towards better forest management by: catalysing local-level
organization, participation and partnerships; mediating conflicts; obtaining
multiple perspectives from the field; raising public awareness; advocating
alternatives to the status quo; and brokering information and services
between 'vertical' and 'horizontal' levels. A model of levels of participation
of the various players is presented in Figure 10.4.
To varying degrees the above questions are behind a number of coun-
tries' initiatives for institutional change. A major focus of donor-supported
national forestry action programmes, forest sector master plans and vari-
ous bilateral projects has been institutional restructuring. However, the rel-
atively fragile reforms which often result from such programmes have led
to efforts to achieve greater understanding of internal dynamics of institu-
tions, their histories and the 'actors' within them.
10.4.2 Institutional histories, internal dynamics and actors

As in other types of institution which are responsible for implementing
government-dominated policy, forest authorities have tended to 'shape'
forest policy into that over which they have practical decision-making
power. Over time, a web of incremental changes defining operational pro-
cedures gradually forms and obscures the (often) more lofty or ambitious
formal goals of policy. As they respond to pressures on the forest resource
and constraints on their budgets, forest authorities tend to become more
rigid and 'top-down'. Procedures become objectives, and the emphasis on
technical planning and professionalism leads to a legitimation of isolation
within narrower confines of forestry (Hobley, 1996b). Internal incentive
structures tend to be based on observing the hierarchical norms of behav-
iour, not on rewarding individual innovation.
Table 10.1 Major international forest policy initiatives
Date Event Detail
The 1960s were dominated by large-scale industrial-commercial forest harvesting and

plantations, based on Northern models, to earn revenues and foreign exchange. In the
latter part of the decade, scepticism about the ability of prevailing kinds of economic
growth to deliver development was increasing.
The 1973-74 oil price shocks led to the 'wood fuel crisis'. Much of the world was
gripped by the energy crisis of modern fuels, and against a background of deforestation
and declining tree-stocks the assumption that both types of crises were essentially
similar was an obvious one to make. The answer to the woodfuel crisis was seen as
large-scale afforestation.
1980 World Conservation Important contributions of the document were
Strategy published conceptual 'sustainable development' in a public
document
Two useful generalizations emerged in the mid 1980s: (a) the driving forces of
deforestation and forest degradation lay mainly outside the forest sector; (b) the general
cause of forest destruction lay in the undervaluation of goods and services provided
from the forest, and overestimates of the benefits derived from forest clearance.
1985 Tropical Forestry Action TFAP set up to curb deforestation, promote the
Plan (TFAP) launched sustainable use and conservation of forest
at World Forestry resources and to increase aid to forestry sector
Congress, Mexico City
1985 International Tropical An agreement in a class of its own among
Timber Agreement commodity agreements, concerned not only with
(ITTA) signed and timber trade but resource upon which it depends,
International Tropical signed by main producers and consumers of
Timber Organization tropical timber
(ITTO) established
1987 Brundtland Report Environment should be seen not as an obstacle to
published growth but as an aspect to be reflected in policy if
growth is to be sustained
1988 No Timber Without International Institute for Environment and
Trees report Development team commissioned by ITTO
concluded that less than 1 % of global tropical
timber came from sustainable sources
By 1992 three major issues were at the forefront of the North/South division in
International forestry issues: sovereignty; double standards; and paying for revenue
foregone to promote global benefits
1992 Centre for International A non-profit, autonomous scientific research
Forestry Research organization with key areas of focus for
(CIFOR) established research
1992 ITTO Criteria and The C&I debate continues in the search for
Indicators (C&I) for reliable repeatable standards to be set and
sustainable forest accepted
management
Table 10.1 Continued
Date Event Detail
1992 UN Conference on Discussion on forestry moved to a new level

Environment and though no convention on forests was agreed at
Development (UNCED), the conference
Rio de Janeiro
UNCED provided the stage for issues in forestry, biodiversity, desertification and
climate change to converge. The Conventions on Biological Diversity and Climate
Change both affect forests. Agenda 21's Chapter 11 on Combating Deforestation
recognises the broad role played by forests. The Forest Principles indicate consensus on
the global importance of forests but they represent a political statement, not an
operational document.
1993 Forest Stewardship An NGO-industry alliance set up to combat the
Council problems of lITO and focus attention on the forest
management unit level as opposed to the
intergovernmental level; promotes development of
standards and accredits forest certifiers.
1994 World Commission on As a result of the lack of global framework for
Forests and Sustainable action on forests, WCFSD set up to fill void
Development (WCFSD)
1994 Pan-European Helsinki Three separate processes emerged after the
Process renegotiation of IITA: the Helsinki and
Montreal Processes focus on C&I for boreal and
1994 Non-European temperate forests, and the Tarapoto on
Montreal Process Amazonian forest; each covers principles and criteria
for the sustainable managements of forests.
1994 Tarapoto declaration
of the Amazon
1995 UN Commission on The CSD session considers forestry: IPF is the
Sustainable highest international body ever to consider
Development (CSD) sets forest issues, created for generating consensus
up Intergovernmental and to propose actions for implementation of
Panel on Forests (IPF) UNCED's forest-related agreements at the national
and international level
1997 IPF reports back to IPF demonstrates the possibilities and problems
the CSD for institutionalized intergovernmental cooperation
on forestry: many 'lowest common denominator'
proposals made; main decision is to reconvene as
the Intergovernmental Forum on Forests; forest
convention on its agenda for debate.
1997 WCFSD completes "Our Forests, Our Future" focuses on civil society:
regional hearings calls for direct involvement of poor people and
business; and tackling inequity and corruption.
International initiatives for forests which burgeoned after UNCED have made
significant progress in defining principles for sustainable forest management; securing
broader support from beyond the forest sector; and obtaining a better balance of
representation of rich and poor country interests. Contentious issues which continue as
challenges for international agreement include dealing with international companies
impacting on forests; payments for 'global services'; and supporting national efforts at
sustainability and obligations to stakeholders.
Source: adapted from Poore (1995) and OD! (1996).
Breadth of Participation - - - - - -
Government. ....... Academia ........ Sectors ........ Communities
National
Many National Policies
11 11 11
Province
c:
0
~
a.
'0
'E 'Meso' level - links are needed to bring right information 'up'
al
a.. and to send appropriate policy signals 'down'
"0
V V V
.s:::
is.
(J)
Cl
Local
Individuals Local Participatory

Projects and Strategies
Marginal
groups
Figure 10.4 Participation 'map'. (Adapted from Bass et al., 1995.)
Individuals within institutions interact in ways which relate to the insti-

tution's formal mandate, e.g. legislation, human resource development and
budgets. These interactions are 'visible' and can be planned and managed.
But there are also hugely complex, less visible interactions which collec-
tively define the institutional culture. These are more difficult to incorpo-
rate into plans for change, but without analysis and understanding they
will render plans mere paper exercises.
Institutional actors may have diverse motivations which include:
• to enhance the standing of agencies in which they work;
• to pursue a particular policy objective, such as forest protection, above
all others;
• to adhere to professional standards, either for the sake of professional-
ism, or to gain respect from professional peers;
• to pursue ideological or party political objectives;
• to promote their own careers;
• to honour familial and kinship obligations;
• to seek financial gain, both within the agency and with outsiders.
The combination of such motivations, when linked to both the visible and
invisible institution, sets the boundaries for the scope and rate of institu-
tional change and may explain many policy outcomes. Many efforts to
make policy responsive to the new demands of sustainable forest manage-
ment have concentrated their recommendations on broad goals, without
quite knowing how such changes could be brought about. Without basing
such prescriptions on motivations for change within existing institutions,
frustration at the lack of subsequent action is a common result.
10.4.3 Who matters most?

The outcomes of forest policies in the past have often borne little relation
to their stated objectives because success for some has only been possible at
the expense of others, i.e. policy creates both winners and losers. Different
groups have different rights, responsibilities and interests in forests. As we
have noted, some people have greater power and influence to directly affect
the forests, whilst others, particularly those living in forests (who may have
the knowledge and capacity to bring about sustainable forest management)
have been adversely affected by the more powerful.
A key question is: whose needs for forest goods and services matter
most? For most goods, the question might be left to the market to decide,
as an efficient allocator of resources. But the market deals with demands;
the job of policy is to deal also with needs, especially of those who cannot
operate in markets due to poverty, for example. In addition, there are
social and environmental consequences of the market's allocation of
resources; incentives or regulations may be needed to ensure a more opti-
mal outcome than can be achieved by the market alone. Some groups with
strong environmental and social needs are also actively involved in forest
management. A further question is therefore: who contributes most to sus-
tainable forest management?
In negotiating the answer to both questions at the level of a particular
forest, dimensions to consider for each stakeholder group are likely to
include: existing stakeholder rights; dependency on forest goods and ser-
vices for livelihoods; proximity to the forest; indigenous knowledge; and
n
HIGH POTENTiAl CONTRIBUTION TO SUSTAINABLE FOREST MANAGEMENT
Indigonous groops (Oroanized user grotJ~ -I"

( ) r-- j
'7 I!
( localfarmors) ( SOliJorsltXllCN'alsLS ) I j ~
~ ~
r-- ~ {
.lI
ii
! E &
( FOrQsl dE!'ParlffiQnllielclSl6fl ) S!
~
OJ (commUnEtv-based organisationS) r--

J, i
:; OJ
z
" ~( Small-scale " OJ
OJ
r--
":>z
~;;; r--
J It
~
A
l- tI- I--
C~ ~E ~ "'~ I-- b
0 ~
1 (!)
~ '---' z
N.OIo081 NGOs
! ~ ( !
J:
(!)
~ :;;
l!l
1.E
~
I
ConsutnOr'S ')'-- i J
~
AlIltlclans ~
J )
NationaJ ahzoos ') '-
LOW POTENTIAL CONTRIBUTION TO SUSTAINABLE FOREST MANAGEMENT
LJ
Figure 10.5 Some matter more than others: a stakeholder 'map' of power and potential contribution to sustainable forest management (hypothetical).
The four quarters of the diagram (A-D) show stakeholder groups which will have to be treated differently if sustainable forest management is to be
achieved. (A) Stakeholders have great potential positive contribution to sustainable forest management but have limited power and influence; specialist
initiatives to develop and secure their interests will be required. (B) Stakeholders have both high levels of influence and high potential contribution to
sustainable forest management; effective collaboration between these groups and with group A stakeholders will be needed. (C) Stakeholders have high
levels of power and influence but have interests which may hinder the pursuit of sustainable forest management; these stakeholders may be a source of
significant risk and will need careful monitoring and management. Actions may be required to mitigate their impacts. (D) Stakeholders have limited
power or potential to contribute to sustainable forest management; limited monitoring may be needed but in general this group is of low priority.
Emerging lessons on 'policy that works' 289
cultural integration with the forest environment (Colfer, 1995). Questions
of the relative power of different groups must also be tackled in this
process (Figure 10.5).
10.5 EMERGING LESSONS ON 'POLICY THAT WORKS'
10.5.1 Forestry is changing (whether foresters like it or not) and good

policies deal with change
Figure 10.6 is an attempt to put together the main elements in understand-
ing how policy changes: the key elements of policy context; the important
dynamics within groups of policy actors; the resulting policy characteris-
tics; and the impacts which these policies have (Mayers, 1997).
In Figure 10.6, four types of forest policy change are characterized.
Whilst we are currently testing this schema against six country case studies,
current indications suggest that 'politics as usual' perpetuates inequitable
and unsustainable roles. Generating a sense of crisis, as did the early inter-
national Tropical Forest Action Plan (TFAP) campaign, results in a limited
incrementalist agenda. Rather, the moves towards more cooperative
approaches, and building on local skills and control mechanisms, are more
promising in terms of genuine transformation towards sustainability.
However, the precise 'value added' of participation and subsidiarity is not
yet fully proven in policy contexts (as opposed to local projects), and the
degree and manner in which they should be implemented needs further
research. In this section, we therefore focus on some lessons learned to date
about the kinds of processes which lead to policy change for better forest
management.
10.5.2 Participation - a forum and a process

The key element in making policy more workable is the existence of mech-
anisms which link policy development efforts to on-the-ground realities,
and which result in outcomes acceptable to all parties. Mounting evidence
suggests that forms of wider participation in decision-making for policy
formulation and implementation can bring major benefits:
• improving stakeholder relations through greater transparency, account-
ability and political credibility;
• uncovering information that matters and learning from it;
• broadening the base of ideas, skills and inputs;
• expressing disagreements and reaching better consensus on trade-offs;
• mutual understanding of realities and practicality of resulting objectives;
• developing capabilities for negotiation collaboration;
• strengthening partnerships and commitment to implementation;
• cost savings in the longer run.
Policy Actors
Ideological predispositions
Position and control of resources
Professional expertise & experience
PoliticaVinstitutional loyalties
Personal attributes and goals
Stakeholder/societal
pressure
History of forest use

and policy
Institutional capacity
Forest ownership
conditions
Economic conditions!
changes
Forest resource
conditions
Determinants of I Single or Strength of state Contingen~ of Strength of

. group of implementing budget national support
Implementation policy changes structures! incentives on
finances cooperation
Degree of elite Level of Level of De~ree of

consensus! bureaucratic cross-sectoral devo ution of
control control consensus power
Policy Impacts
Social Economic
Figure 10.6 Process and power in changing forest policy. (Adapted from Mayers, 1997.)
The appropriate scope of participation will vary between contexts. For
example, in quite homogeneous societies presided over by a few widely
respected leaders, a premium may be put on avoidance of conflict, while
consensus supports policy decisions made by the elite few. In more hetero-
geneous societies, only an inclusive process involving all key forest actors
in decision-making is likely to produce policy that stands the test of time. It
will be noted that examples of the former context are increasingly scarce in
the contemporary forestry world.
BOXI0.5: PARTICIPATION IN POLICY

How people participate
The viability and impact of policy has much to do with the level of
agreement amongst actors on the form of participatory practice.
Varying depths of participation can be identified, each of which may
involve a few actors only, or all major groups and government:
1 Being told what has been decided
2 Listening and giving information
3 Being consulted
4 Contributing analysis and agenda-setting
5 Reaching consensus and negotiating trade-offs
6 Decision-making on policy or its components
7 Implementing, learning from experience and adapting policy.
Systems for participation in policy

There is scope for building participation on forest issues into existing
systems:
• Planning systems
• Political systems
• Education/academic systems
• Extension systems
• Media and the arts
• Traditional systems - village-based, religious
• Personal networks cutting across institutional boundaries
For many of the more recent attempts to revise forest policy, new sys-
tems have been developed. These were often one-off, time-bound, or
linked to a project. However, some may prove to be transitional to more
permanent new systems. For example:
• Special committees (e.g. National Forestry Programme Steering
Groups in Papua New Guinea and Tanzania)
• Round tables (e.g. developing Canada's Forest Strategy)
• Working groups (e.g. developing National Forest Standards for certi-
fication in Cameroon, Ghana, Gabon and Brazil)
There are many challenges and risks in responding to increasing pres-
sures by developing forms of participation and negotiation in forestry.
Where participation is 'new', how do you start, how do you phase it in
effectively? In some instances, participatory structures are institutionalized,
replacing old hierarchical structures but without providing the flexibility
which is crucial for workable adaptation to local circumstances.
Issues of political representation and power in wider society present
major questions for any actions targeted at forests alone. Can forestry
institutions function democratically in the absence of democratic systems
of governance? Can long-term planning for forests compete with short
electoral cycles? Fundamental problems of inequity in access to forest
goods and services may rule out any strategy of nurturing participation for
better forest management. Indeed, some forms of participation may be cyn-
ically manipulated to legitimize or enhance existing power structures.
As the credibility of national-level approaches to forestry is typically
being challenged from all sides, re-negotiation of a national vision for
forests is needed. Sustainable forest management will require key stake-
holder perspectives and needs to be addressed, and conflicting interests to
be balanced and managed. Some kind of national forum will be necessary.
Government has the comparative advantage in convening such a forum,
which may utilize a wide range of organizational mechanisms. Collectively,
these mechanisms should constitute a cyclical process, aiming at continu-
ous improvement, dealing with uncertainties and learning from them
(Figure 10.3). Unattainable wish-lists are replaced by objectives which
become strategic and tactical with iteration.
BOX10.6: CORPORATE TACTICS AIMED AT UNDERMINING

CHALLENGES TO POWER
Powerful interests, often state agencies or big companies, typically wish
to resist policy changes towards more equitable forest use, in order to
maintain their power base and protect their claims. Tactics used by
powerful interests to resist policy changes include:
• Directing debate into safe areas, diverting pressure for changing fun-
damental policies, for example by promoting debate at local level to
keep things bogged down in internecine struggles between local com-
munities;
• Institutionalizing active groups, by co-opting them into phoney con-
sultative status, thus reducing their ability to openly challenge policy;
• Influencing events from a reputation for power, whereby issues are
kept off the policy agenda by decision-makers fearful of the conse-
quences from powerful interests - called 'non-decision making'.
Whilst many national fora are being planned or formed at present, there
is not much experience. Canadian round tables on forestry and the paper
industry offer some lessons. They have developed many ideas, but them-
selves were constructed as neutral bodies and therefore have no powers,
relying on the influence of their members through other institutions (Bass
et al., 1995). The same is true of the various committees set up by TFAPs.
The national working groups set up to define national forestry standards,
often at the instigation of the FSC, will be worthy of investigation soon -
they are resulting in some important 'soft' policy such as certification stan-
dards and procedures, and have been useful in-country focal points for dis-
cussing what is good forest management, who should be accountable, and
how to monitor them.
Capable and creative management of the participation process appears
to be crucial, to know how to organize it, to generate the trust of key
stakeholders, and to ensure that there is learning from successes and fail-
ures. A process like this never really ends, but there are milestones along
the way. The following sections note some of these.
10.5.3 Setting and agreeing national forest targets - introducing the goal
of security of forest goods and services
A key stage in shaping a national vision for forests is to agree which forest
goods and services are scarce, and to define how much is needed, both now
and in the future. Security of forest goods and services can be provided by
a wide spectrum of land uses, not just high forest. Also, some goods and
services can be provided by imports, whilst others can be substituted from
non-forest sources such as concrete, metals and plastics.
In practice most countries will aim for a mix of these options for achiev-
ing forest security. Most will decide to invest in a 'permanent' forest estate,
as some forest goods and services have limited substitutability (e.g. water
supplies, biodiversity and support for food security), and as some needs
cannot be met by market means such as purchasing imports or substitutes.
However, an estate consisting of the vast majority of a country's forests
'reserved' under government control is unlikely to prove a viable option in
the long run. Many ownerships need to be included under the permanent
forest estate, with procedures to alienate forests if key social benefits result.
Current policy debates on 'sustainable forest management' frequently
become lost amongst differences in perceptions, definitions and values. So
also do those with a focus on forest area, such as those 'against deforesta-
tion', or 'planting target' policy initiatives. We propose that the concept of
forest security provides a better focus on the actual goods and services,
who needs them, and their most efficient means of production (Bass and
Thomson, 1997). Policies which concentrate mainly on achieving security
of forest goods and services ensure that the focus is on people and their
needs rather than forests per se, and are more likely to lead to a shared
vision of the future and a path to get there.
10.5.4 Linking forest policy with other decision-making cycles and

extra-sectoral influences
Understanding the root causes of forest problems is a prerequisite to their
solution. As we have seen, many root causes are far from the forest. Some
influences on forests will remain intractably out of reach of even much
more inclusive policy approaches to forests in land use. Yet there is much
that can be done to integrate the forest policy cycle with national develop-
ment plans and treasury budgets, rather than remaining a stand-alone sec-
toral activity. Doing so will help forestry to respond more rapidly to
extra-sectoral change. Approaches to national strategic planning, environ-
mentaVsocial impact assessment and forest valuation can assist this engage-
ment.
Articulation with strategic processes like national strategies for sustain-
able development, where they have genuine popular and high-level sup-
port, can enable building on current mechanisms and networks such as
those initiated by the best national conservation strategies, national envi-
ronmental action plans and national forestry programmes (Carew-Reid et
al., 1994). Valuation of forests, and assessment of the costs and benefits of
alternative forest uses (including social and environmental aspects), can
demonstrate the economic contribution of intact forests to other sectors,
and assist the beneficial transfer of forest land between sectors (Eaton,
1995).
10.5.5 Tackling power imbalances and institutional inertia

An active approach to agreeing who matters most has already been
described. By recognising such factors as forest dependency, proximity, pre-
existing rights and indigenous knowledge as important, the imbalances in
power to put into practice the potential contributions to better forest man-
agement become apparent. The challenge is then to mitigate the impacts of
those with negative influence and empower those who can best balance
social, economic and environmental objectives.
Along with an increasing body of international law, some principles
which have gained currency through international experience can provide
'levers' for tackling power imbalances in forestry. For example, the princi-
ple of subsidiarity implies that management decisions should be made by
those closest to their consequences. Policy requires negotiation at levels
with the greatest potential contribution to better forest management. This
often requires devolution of powers to those who live in forests, or are
most dependent on them for their livelihoods. Balancing private-sector
dependence on forests against household-level dependence is a key policy
dilemma.
Disabling institutional cultures are characterized by hostile working rela-
tionships, closed communications, few team interactions, weak learning
mechanisms, and low rewards for performance. These features are com-
mon in government forestry agencies, but are by no means limited to them.
Realistic assessments of the rate and type of possible change are needed.
Reorienting an institutional culture depends on supporting those who truly
want to make improvements. There are always some highly motivated peo-
ple in an institution with areas of responsibility within which small positive
changes can occur. Increasing capacities to generate more room for
manoeuvre for these 'change agents' can eventually bring significant
results.
As the corporate forestry sector has found, quality/environmental man-
agement system approaches allow a step-wise improvement in performance
on multiple fronts - by improving communications, setting realistic targets,
and helping to reward efficient means of achieving targets. Such
approaches are beginning to be used by government agencies.
10.5.6 Democracy of knowledge about forest assets, demand and use

Understanding the condition of forests, and the nature of changes taking
place, involves more variables than in the past. Information that is truly
useful for policy and management needs to cover not only the state of
forests, their extent, quality and ownership, but also the pressures on the
resource in terms of demands and obligations, and the use of forest assets -
the flows of goods and services and the management practices applied.
Traditional scientific approaches to generating forest information, such
as permanent sample plots and remote sensing, need to be complemented
by a mix of more people-oriented approaches such as participatory inquiry.
Many forest departments are currently hamstrung by the detail of silvicul-
tural and biophysical data they are supposed to collect and manage. This
attempted comprehensiveness can lead to unviable management prescrip-
tions. With a focus on making existing information sources more useful for
policy and management processes, information systems can be made more
strategic than comprehensive, and can accommodate the range of social
and economic as well as ecological dimensions.
Approaches which are contributing to meeting the need for a broader
base of understanding, accountability and policy-relevance of information
include: valuation of forest assets, benefit flows and costs; multi-factor,
dynamic information systems and watchdogs that use them (e.g. in Ghana
and Brazil); cross-sectoral analysis and research on forests (e.g. the Land
Use Working Group in Vietnam); environmental and social impact assess-
ments of major forest plans (far from routine in forestry yet, but there are
scattered examples e.g. in Guyana and Indonesia); and participatory forest
assessment and monitoring methodologies (progress being made e.g. in
India, Ghana and Papua New Guinea) (Carter, 1996).
Forest resource accounting is a strategic forest information system which
focuses on stakeholders tracking changes in production and protection in
forests, and sharing information to build up a national forest balance sheet
of asset reductions, additions and product flows. It was recently designed
for lITO as a methodology to organize information for improving forest
policy and management (IIEDIWCMC, 1996). Forest resource accounting
systems have been designed for Pakistan, Ecuador and Guyana, and there
is interest elsewhere. On paper, the approach is efficient. The principal con-
straint is the lack of incentives for heads of forest authorities to share infor-
mation, and for managers to assess actual forest condition. These
constraints are, however, being gradually whittled away by certification at
the forest level, and by possibilities of international compensation for for-
est conservation at the national level. The time for routine forest resource
accounting may yet arrive. Until something like it is in place, however, few
nations will have an adequate picture of their forest asset base and how it
is being used.
Empowerment for effective forest stewardship also requires a democracy
of knowledge about forests, and about what management works and what
does not. 'Good forest management practice' can no longer be the sole pre-
serve of a professional elite. Sharing goals and knowledge between levels is
needed, with a commitment to building elements of sustainable forest man-
agement into all key actors' tasks. One of the major challenges at present is
in convincing the highest authorities that more participation is needed, not
only in contributing the many types of information required, but in the
right to access such information in useful forms (as with forest resource
accounting). A high priority has traditionally been placed on such a
democracy of knowledge in Sweden, with the result that there is a much
stronger unity of vision, knowledge and purpose, from minister and cor-
porate boss to villager and forest worker (lIED, 1996).
10.5.7 Agreement on developing capacity: roles, rights and

responsibilities
Many of the pressures on the policy agenda now emanate from the market
and from initiatives in civil society. Recognition is increasing that capacity
for better forest management resides with a range of actors -local commu-
nities, NGOs and the private sector - as well as governments. The assertion
that government by necessity has to carry out all tasks of forest manage-
ment is fading with governments' declining capability to do so, and with
demands for broader forest objectives. Yet government cannot effectively
divest itself of responsibilities by decree. A renegotiation of roles is
required, one which is based on a shared analysis of the key issues, which
sorts out the authority, watchdog and user functions, makes the linkages
between top-down and bottom-up efforts, and generates consensus on the
rights, responsibilities and relationships of forests institutions and actors.
10.5.8 Framework of coherent but adaptable policy instruments

The forms of policy content that promote more sustainable forest manage-
ment are becoming better understood. We mention these policy contents,
or policy instruments, at the end of our outline of emerging lessons to
emphasize the fact that policy instruments are only as good as the process
which generates and implements them. Conversely, however, a framework
of coherent policy instruments is required to manage change - a frame-
work which is iterative and continuously improving.
BOX10.7: USING THE 'TOOL-KIT' OF POLICY INSTRUMENTS

Five main types of instrument for implementing policy can be distin-
guished.
• Regulatory e.g. laws on forest practices, rights, tenure, trade
• Economic/market e.g. taxes, subsidies, fees/rebates, tradable permits,
compensation, auctions
• Informational e.g. research (technical, marketing, socioeconomic),
training, consumer information, land-use planning
• Institutional e.g. mechanisms for dialogue, partnership, intersectoral
integration, common property regimes and local management, exten-
SIOn
• Contracts/agreements e.g. access/management agreements, direct
works
Criteria for choosing the appropriate mix of instruments might include
• timeliness
• relevance
• credibility to stakeholders
• equity amongst stakeholders
• ease of understanding
• level of uncertainty and controversy
• information needed to operate instrument
• cost effectiveness - economic efficiency
• time requirements
• reliability
• reproducibility
The past reliance on regulatory policy instruments alone has become
nonsensical in many contexts. Newer initiatives seek to emphasize infor-
mation and incentives, rather than coercion, to meet incremental costs of
improved management. Some new instruments are being developed by
multi-stakeholder groups, notably the private sector and NGOs working
together, with rather less involvement of government. These include
national forest management standards and codes of practice, with or with-
out means of assuring accountability. Such initiatives are helping to gener-
ate vision and commitment to better management where they recognize
dependencies between forest actors.
Such instruments include forest management certification and labelling
of the products of certified forests. These are potentially powerful tools to
improve the standard of forest management by generating market incen-
tives in its favour (Upton and Bass, 1995). They are thought to offer some
key advantages over regulatory approaches, including cost savings, by
allowing forest managers to determine the most appropriate ways of meet-
ing given external standards; and an incentive to continuously improve,
and in so doing to encourage the development of new techniques and tech-
nologies. They encourage improvement without requiring new legislation
and regulations, and relative efficiency in terms of resource pricing. Several
international certification programmes are gaining credibility, notably
those connected to FSC and the International Organization for
Standardization (ISO), whilst significant national initiatives are under way
in a number of tropical countries, including Brazil, Cameroon, Costa Rica,
Ghana, Indonesia and Malaysia.
However, certification and labelling, along with other market instruments,
can only work where forests are producing for a market which discriminates
in favour of production methods which are environmentally and/or socially
sound (mainly north-western Europe and some markets in North America),
and where these markets more than compensate for the extra costs involved.
The size, longevity and behaviour of such markets, and their reaction to cer-
tified products, are not yet fully proven in the forestry sector.
By mid 1997, only about 5 million ha, producing less than 1 % of glob-
ally traded forest produce, was certified, but many more forests were in the
process of being certified (initial audits and corrective action plans, etc).
Some major challenges and dilemmas stand out. Can certification do more
than create islands of good management in a sea of bad practice? Can the
benefits of certification be secured by small-scale forest producers?
Indications are that answers to these will be in the affirmative, but experi-
ence is still in its infancy. One thing is certain. Certification cannot work on
its own; it is only part of the policy 'tool kit'. It can do little, for example,
to tackle the problem where weak policies, laws and forest pricing encour-
age forest asset stripping. Fundamental policy changes are needed in such
instances, for which certification is no substitute.
Signs of change 299
10.6 SIGNS OF CHANGE
In some parts of the tropics, policies and institutions are slowly becoming
more enabling of sustainable forest management. Progress is being made:
through decentralization approaches in which local-level collaborations are
given more support; through personal networks cutting across institutional
boundaries; through semi-formal or temporary institutions such as round
tables, which offer a level playing field for many forest actors; and through
the design of new instruments, such as certification, which can link forest
producers with demands for good social and environmental performance.
Initiatives like these have brought the forest debate into a new light, and
have often done much to rehearse the ingredients (sections 10.5.2 to
10.5.8) of improved policies. (However, the temptation in many cases has
been to view the initiative in question as necessarily being a resolution to
the debate at hand - the 'tyranny of panaceas'.)
We, and our colleagues in several countries, have observed policy
changes which are generally agreed to have been good for both forests and
people, some examples of which are given below.
• In Brazil, alliances of rubber tappers, academics and international pres-
sure have gradually changed the 'perverse' policies which had caused so
much deforestation and favoured so few.
• In Ghana, consultative processes over several years have generated con-
siderable momentum for positive change and produced: a new national
forest policy; with innovative informational and regulatory measures
which are bringing improved forest management inside and outside for-
est reserves; working groups developing a national approach to certifica-
tion; and gradual adoption by the forestry department of collaborative
approaches with other sectoral agencies and with local communities,
NGOs and the private sector (Smith, 1996; Koley et ai, forthcoming).
• In Papua New Guinea, NGOs are now represented on the board of the
national forest authority and have made their presence felt by opening
up the debate on appropriate levels of control over PNG's forests - a
result of the TFAP initiative. The role of state intervention is gradually
being re-positioned to seek negotiated contracts with other stakeholders
to balance biodiversity conservation, sustained yield and revenue distri-
bution objectives. Participatory working approaches in a number of
agencies are increasing capacity to facilitate, listen to and respond to
bottom-up decision making (Mayers and Peutalo, 1995; Filer, 1997).
• In Zimbabwe, forestry policy and institutions are in a state of flux. The
government has recognized that it is no longer tenable to combine the
functions of forest authority, enterprise and extension agency under one
state roof. Commercial functions are being hived off, and the forest
agency is experimenting with resource-sharing approaches in some
reserves. A government commission of inquiry into land tenure has
raised the level of debate on complementary issues of resource manage-
ment. Although fundamental inequities in access to forest goods and ser-
vices remain, the involvement of more stakeholders in development of
natural resource management policy is likely. An NGO alliance is prov-
ing influential in pressing for development of a clearer national vision
for sustainable woodlands based on experience with approaches to
locally controlled resource management (Nhira et ai, forthcoming).
• In India, an unparalleled level of experience with participatory forest
management has developed over the last 20 years. There have been sig-
nificant enabling changes in land tenure and rights; development of
improved resource management practices; new collaborative relation-
ships between government, NGOs and academic institutions; and the
incorporation of participatory methods into the practices of the forestry
sector (Hobley, 1996b). A new national forest policy in 1988, and sub-
sequent national- and state-level resolutions, have formalized many joint
forest management agreements between forest departments and local
people. In some locations this has translated into nothing more than a
new strategy for government to reassert control over forest land. In
many others, however, the level of learning and institutional reorienta-
tion is such that a genuine interface between local people and govern-
ment staff has developed. Such relationships can lead to a flexible match
of policy and institutions to the ecological and social environments in
which they operate (Khare et ai, forthcoming).
• In Pakistan, legal changes have finally been made to allow communities
to play their part in joint forest management, reinforcing a trend away
from governmental control alone and towards reinstating community
management mechanisms and rules. Fifteen years of NGO- and donor-
led participatory forestry projects helped to change the policy climate
(Ahmed and Mahmood, forthcoming).
Progress will be fragile for as long as tropical forestry remains character-
ized by: the predatory tactics of logging companies seeking and supporting
weak policies and national authorities; by over-concentrated control and
inequitable access to forests; by 'fortress forestry' institutions; and by an
ill-informed public and consumer. Yet resolve to achieve sustainable forest
management continues to strengthen. The processes which make, apply
and learn from better policy for forests are becoming clearer. And there are
increasing opportunities for key forest actors to recognize and develop
these policy processes.
REFERENCES
Ahmed, J. and Mahmood, F. (forthcoming) Pakistan Country Study, Policy That
Works for Forests and People, Series No.1, International Institute for
Environment and Development, London.
References 301
Barber, C. V., Johnson, N. C. and Hafild, E. (1994) Breaking the Logjam:
Obstacles to Forest Policy Reform in Indonesia and the United States, World
resources Institute, Washington, DC.
Bass, S., Dalal-Clayton, B. and Pretty, J. (1995) Participation in Strategies for
Sustainable Development, lIED Environmental Planning Issues, No.7,
International Institute for Environment and Development, London.
Bass, S. and Thomson, K. (1997) Forest Security: challenges to be met by a global
forest convention, lIED Forestry and Land Use Series No. 10, International
Institute for Environment and Development, London.
Carew-Reid, J., Prescott-Allen, R., Bass, S. and Dalal-Clayton, B. (1994) Strategies
for National Sustainable Development: A Handbook for their Planning and
Implementation, Earthscan, London.
Carter, J. (1996) Recent Approaches to Participatory Forest Resource Assessment,
ODI Rural Development Forestry Study Guide 2, Overseas Development
Institute, London.
Colchester, M. (1994) Sustaining the forests: the community-based approach in
South and South-East Asia. Development and Change, 25, 69-100.
Colfer, C. J. P. (1995) Who Counts Most in Sustainable Forest Management?
CIFOR Working Paper No.7, Center for International Forestry Research, Bogor.
Eaton, D. (1995) Improving Economic Information for Forest Policy: Exploring
the Linkages between Forest Resource Accounting and Natural Resource
Accounting, International Institute for Environment and Development, London
(mimeo.).
Filer, C. (ed.) (1997) The Political Economy of Forest Management in Papua New
Guinea, NRI Monograph 32, National Research Institute Papua New Guinea
and International Institute for Environment and Development, London.
Hobley, M. (1996a) Institutional Change within the Forest Sector: Centralised
Decentralisation, Working Paper 92, Overseas Development Institute, London.
Hobley, M. (1996b) Participatory Forestry: The Process of Change in India and
Nepal, ODI Rural Development Forestry Study Guide 3, Overseas Development
Institute, London.
Humphreys, D. (1996) Forest Politics: The Evolution of International Cooperation,
Earthscan, London.
Hurditch, B. (1996) Changing roles in forest management and decision-making: an
overview, in Making Forest Policy Work: Conference Proceedings of the Oxford
Summer Course Programme 1996, (ed. K. L. Harris), Oxford Forestry Institute,
Oxford.
lIED (1996) Towards a Sustainable Paper Cycle, International Institute for
lIEDIWCMC (1996) Forest Resource Accounting: Strategic Information for
Sustainable Forest Management, International Institute for Environment and
Development, LondonIWorid Conservation Monitoring Centre, Cambridge.
Khare, A., Bothla, S., Polit, S., Sarin, M. and Saxena, N.C. (forthcoming) India
Country Study, Policy That Works for Forests and People Series No.3,
International Institute for Environment and Development, London.
Kotey, N. A., Amanor, K. S., Antwi, L., Francois, J., Owusu, J. G. K. and Yeboah,
R. (forthcoming) Ghana Country Study, Policy That Works for Forests and
People Series No.4, International Institute for Environment and Development,
London.
Mayers, J. (1997) Policy for forests - what makes it change and what makes it
work? Examples from countries of the South, in Proceedings of the International
Symposium on the Non-Market Benefits of Forests, (eds A. Park and C. S.
Raper), Forestry Commission, Edinburgh, June 1996, HMSO, London (in press).
Mayers, J. and Peutalo, B. (1995) NGOs in the Forest: Participation of NGOs in
National Forestry Action Programmes: New Experience in Papua New Guinea,
lIED Forestry and Land Use Series No.8, International Institute for
Morrison, E. and Bass, S. (1996) Forestry Commissions: A Review of Experience,
International Institute for Environment and Development, London (mimeo.).
Nhira, c., Baker, S., Gondo, P., Mangono, J.]" Marunda, C. and Matose, F. (forth-
coming) Zimbabwe Country Study, Policy That Works for Forests and People
Series No.5, International Institute for Environment and Development, London.
OD! (1996) Events Affecting the Course of Recent International Forestry History,
Paper prepared for the Forestry Programme, Overseas Development Institute,
London (mimeo.).
Poore, D. (1995) Forestry and nature conservation - changing perspectives.
Commonwealth Forestry Review, 74, 5-19.
Richards, M. (1995) Role of demand side incentives in fine grained protection: a
case study of Ghana's tropical high forest. Forest Ecology and Management, 78,
225-241.
Saxena, N. C. (1996) Joint Forest Management in India: Empowering the People or
Another Development Bandwagon? Centre for International Forestry Research
(CIFOR), Bogor.
Shepherd, G. (ed.) (1992) Forest Policies, Forest Politics, OD! Agricultural
Occasional Paper 13, Overseas Development Institute, London.
Sizer, N. (1996) Profit Without Plunder: Reaping Revenue from Guyana's Tropical
Forests Without Destroying Them, World Resources Institute, Washington.
Sizer, N. and Rice, R. (1995) Backs to the Wall in Suriname: Forest Policy in a
Country in Crisis, World Resources Institute, Washington, DC.
Smith, E. K. (1996) The Evolution of Policy Into Practice: A Case Study of Ghana's
Forest Policy, paper presented to the African Forestry Policy Forum, Nairobi,
August 1996, World Bank, Washington, DC.
Thomson, K. (1996) Global initiatives in forest policy: context setting, in Making
Forest Policy Work: Conference Proceedings of the Oxford Summer Course
Programme 1996, (ed. K. L. Harris), Oxford Forestry Institute, Oxford.
Upton, C. and Bass, S. (1995) The Forest Certification Handbook, Earthscan,
London.
Watson, V., Cervantes, S., Castro, c., Mora, L., Solis, M., Porras, I. T. and
Cornejo, B. (forthcoming) Costa Rica Country Study, Policy That Works for
Forests and People Series No.6, International Institute for Environment and
Development, London.
-11------
Modelling tropical land use change
and deforestation
A. Grainger
School of Geography, University of Leeds, Leeds LS2 9JT, UK
11.1 INTRODUCTION
Little attention was paid to modelling deforestation until the late 1980s,
since it was viewed more as a problem to be curtailed than as a universal
phenomenon to be understood. But this meant that policies to control
deforestation were based on partial or anecdotal evidence, and so it was no
surprise that they had little effect (Grainger, 1993a). Controlling deforesta-
tion will be difficult until we know more about the processes involved.
Modelling is the key to this, for once a suitable theoretical framework has
been devised it can form the basis for mathematical models that use empir-
ical data to test hypotheses about deforestation. Models can also be used to
predict alternative future scenarios for long-term trends in deforestation
and its impacts, and test how these could be influenced by policy-makers
(Lambin, 1994). This chapter looks in turn at theoretical and mathematical
models of the causes of deforestation, long-term trends in forest cover, spa-
tial trends in deforestation, and spatial trends in logging. It focuses on
deforestation in tropical moist forests - the closed forests of the humid
tropics - rather than forests in the montane or dry tropics where deforesta-
tion processes are rather different.
11.2 BACKGROUND
11.2.1 Deforestation
Deforestation is defined here as the temporary or permanent clearance of
forest for agriculture or other purposes (Grainger, 1993a). This is compati-
ble with the definition used by the UN Food and Agriculture Organization
(FAO) in estimating rates of deforestation (Lanly, 1981).
304 Modelling tropical land use change and deforestation
11.2.2 Forest degradation
Deforestation is just one of the human impacts on tropical forests, and
excludes other impacts which do not involve clearance. These are better
described as causing forest degradation (Lanly, 1981), defined here as a
'temporary or permanent reduction in the density, structure, species com-
position or productivity of vegetation cover' (Grainger, 1996a).
Deforestation is an extreme case of degradation, temporarily removing
vegetation cover. Vegetation cover may increase afterwards, due to plant-
ing or natural regeneration, but if the result is inferior to what was there
before, in terms of such attributes as biomass density, canopy cover, struc-
ture and species composition, then the long-term impact on the forest may
be described as degradation.
A prime cause of degradation is selective logging. This, and not clear
felling, is the almost universal forestry practice in tropical rain forest,
which is the most extensive type of tropical moist forest. Only a few of the
thousands of tree species in this forest type are commercially marketable,
and so from one hectare just 2-10 trees are generally removed out of a
total of up to 300 trees above 10 cm diameter. This does not clear the for-
est and so does not constitute deforestation as defined here (although log-
ging may be an indirect cause of deforestation because logging roads may
later be used by farmers to gain access to the forest which they then clear
for cropping). But forest is degraded when the canopy is disturbed by tree
felling; trees not removed may be damaged and sometimes killed, and some
deforestation occurs during construction of logging roads and log-loading
areas. There may also be a long-term change in species composition caused
by continued removal of a few popular species.
11.2.3 Rates of deforestation and degradation

The mean rate of deforestation in tropical moist forest between 1976 and
1980 was estimated at 6.1 million ha/year (Grainger, 1983). This compares
with a mean rate of forest degradation by logging in the same period of 4.0
million ha/year (Grainger, 1986). As some degraded forest was later defor-
ested it would be incorrect to combine the rates of deforestation and degra-
dation to estimate the overall human impact.
11.2.4 Land use and land cover

Deforestation will lead to a change in land use and some degradation of
land cover - the physical and biotic characteristics of the land surface
(Meyer and Turner, 1992) - compared with the climax ecosystem. But land
cover is an attribute derived from the type and intensity of land use prac-
tised in an area, and its degradation does not necessarily result in soil
A theoretical model of the causes of deforestation 305
degradation if the new land use is sustainable. So forest degradation is a
descriptive, not a pejorative term. Contrary to Bilsborrow and Geores
(1994), neither forest degradation nor deforestation can be equated auto-
matically with environmental degradation at the local level. But forest
degradation does have cumulative environmental impacts at the global
level, e.g. by contributing to a loss of biodiversity and the transfer of car-
bon from the biota to the atmosphere.
11.2.5 National land use morphology

Each country has a distinct national land use morphology: its spatial distri-
bution of different land uses and covers, including various types of forest,
other types of natural vegetation, agriculture, and urban and industrial set-
tlements (Grainger, 1995a). The national land use morphologies of most
countries in the humid tropics are dominated by the forest and agriculture
sectors, and as forest originally occupied a large proportion of total land
area, the spread of farming has required the transfer of land from the forest
sector, i.e. deforestation.
11.3 A THEORETICAL MODEL OF THE CAUSES OF

DEFORESTATION
11.3.1 Immediate versus underlying causes

A good first step when building a theoretical model of deforestation is to
distinguish between immediate and underlying causes. A range of land uses
can replace forest after deforestation but they do not spread by themselves.
These land uses may well be the immediate causes of deforestation (or
proximate causes; Turner et a!., 1990) but they only expand in response to
social, economic or political driving forces which (with the negative feed-
back factors that can later control deforestation) are the underlying causes.
The two sets of causes together form a complex national land-use system,
portrayed in a causal loop diagram (Richardson and Pugh, 1981) in Figure
11.1. State intervention is regarded as exogenous to this as it can influence
many of the variables. In this section various assertions are made about the
causes of deforestation, some of which are tested in the mathematical mod-
els reviewed below.
11.3.2 Immediate causes

The immediate causes of deforestation may be divided into three main
types of land use: shifting agriculture, permanent agriculture and other
land uses (Table 11.1). They differ in their sustainability, the area of forest
cleared, and the timing of clearance (Grainger, 1993a).
Investment in plantations
y~+
+-- Timber t
~ ~+
Ti~ber Efficiency of
7.
pnce availability utilization
External - / ~
demand ~ ~ +/ / _ Timber -----=. Forest
. ~ reserves
-1 ~
Demand + protection
-/L0991n9~
t
for timber
+ + Logged Forest area
Forest sector
P!,!r capita + forest ! / /-
Income ~~ _ y/
+
Permanent •
+/
cropland • Deforestation
Agricultural
sector V/+ +
Shifting cropland
+ ~/ ,-
Food ,r+ Degraded
production Forest land
fallow t+
~ Conversion of
-/ ' \+
'"
Degradation
t-
forest fallow + Fallow period
Yield
~ .~
+ /
Degradation
per ha
~
+
_______ Technical
input -
~
.
Food pnce +
¥j. Demand for
/food
+
............ '/'" + +
............ Soil /
- fertility Food consumption
+ per capita
'--H,--=--=--=--=--=--=--=--=--=--=--=--=--=--=--=-~'" ~er capita/'+
Income
+ +
Population
Figure 11.1 A systems model of national land use.
There are three main types of shifting cultivation. The traditional prac-
tice involves clearing small patches of forest (1-2 hal on long rotations
(15-20 years or more) that allow considerable forest regeneration. It is one
of the most sustainable ways to use the poor soils widely found in the
humid tropics (Sanchez, 1976). Short-rotation shifting cultivation is now
the most common practice, with rotations typically of 6-15 years, but as
low as 3 years in some areas. It supports higher population densities and
allows more settled lifestyles, but is unsustainable if rotations become too
short (Young, 1989). The brief fallow period allows only a low, bushy,
Table 11.1 The immediate causes of deforestation: land uses which replace forest
in the humid tropics
A. Shifting agriculture
Traditional shifting cultivation
Short-rotation shifting cultivation
Encroaching cultivation
B. Permanent agriculture
Permanent field crop cultivation
Government-sponsored resettlement schemes
Commercial ranches
Cash-crop plantations
C. Other land uses
Mining
Hydroelectric power schemes
Narcotic plant cultivation
degraded forest cover, called forest fallow, to develop. The third type,
encroaching cultivation, is typically carried out by landless people who
have little intention of long-term sustainability. They spread out in waves
from roads, clearing forest and cropping land for 3 years or more until fer-
tility is exhausted or weeds are rampant. Only then do they move on to
repeat the process elsewhere, leaving behind a wasteland. So the overall
impact on forest cover is similar to that of permanent agriculture. In the
dry and montane tropics, nomadic pastoralism is also an important imme-
diate cause.
Permanent (or sedentary) agriculture requires the clearance of larger
areas of forest at one time. Cultivation of staple crops, such as rice, maize
and cassava, is widespread in the humid tropics. Wet (paddy) rice cultiva-
tion is very sustainable but not universally applicable, requiring fertile, eas-
ily flooded and well drained land, e.g. in river valleys and coastal areas.
Government resettlement programmes that transfer people from over-
crowded areas to forested areas have caused significant deforestation, par-
ticularly in Indonesia and Brazil, clearing up to 10000 ha per scheme in
Indonesia. These were intended as sustainable alternatives to the
unplanned deforestation of encroaching cultivators, but poor site selection
and other factors have often prevented sustainability, forcing settlers to
move on to clear more forest elsewhere.
Large estate agriculture in Latin America is dominated by cattle ranching,
but its productivity and sustainability is usually low as poor soils cannot
sustain the growth of pasture grasses for long without artificial fertilizers.
Brazilian ranches average 23 000 ha in size, with some as large as 1 million
ha. In Africa and South-East Asia, plantations of tree crops such as oil palm,
rubber and cocoa are the dominant form of large agricultural estate, usually
requiring major forest clearances (1000-10000 hal. Plantations can be sus-
tainable alternatives to forest cover if managed well, and do restore tree
cover, though it is not ecologically equivalent to forest.
The extraction of coal, gold and the ores of iron, aluminium, copper, tin,
etc. can clear thousands of hectares of forest per mine, and only the best
mining companies restore the forest when mining ends. The building of
reservoirs to tap hydroelectric power can clear and/or submerge even larger
forest areas, e.g. 200000 ha for the Tucurui dam in Brazilian Amazonia.
Illegal narcotics (such as cocaine and opium) are, as in shifting cultivation,
grown in small clearings, but in this case farmers move elsewhere when dis-
covered by government agencies. In the dry and montane tropics, gathering
of fuelwood for domestic or commercial use and/or conversion into char-
coal is an important immediate cause of deforestation.
11.3.3 Underlying causes

Population growth
The most prominent of the underlying causes of deforestation is population
growth, which increases demand for food and other commodities. In princi-
ple, this could be supplied entirely by intensification, i.e. raising yields by cut-
ting shifting cultivation rotations, increasing the ratio of permanent
cultivation to shifting cultivation, or using high-yielding varieties.
Intensification was fundamental to Boserup's (1965) theory of how mainly
subsistence societies adjust to population pressures. But the ability to do this
Table 11.2 The underlying causes of deforestation
A. Socia-economic factors
Population growth
Economic growth
Poverty and inequality
B. Physical factors
Proximity of rivers and roads
Fragmentation of forests
Topography
Soil fertility
C. State intervention
Agriculture policies
Forestry policies
Other policies
D. External factors
Demand for exports
Financing conditions
sustainably is limited by poor soil fertility and lack of capital to invest in fer-
tilizers and other inputs, so much of the extra demand is often supplied by
extensification - expanding the farm land area. Cash crop farmers will,
according to Induced Innovation Theory, choose between extending or inten-
sifying cropping according to market opportunities and the relative costs of
factors of production (Binswanger and Ruttan, 1978). If a population cannot
adjust in time to increased pressures on the land, this can lead to involution:
land degradation and a decline in per capita food output (Geertz, 1963).
Some of the extra demand for food could of course be supplied by imports.
Opinions differ on how to model the influence on deforestation of
changes in population. One view is that deforestation is linked with a phys-
ical movement of people into forested areas, so that rural, rather than over-
all, population should be the independent variable in regression tests. For
example, according to Bilsborrow and Geores (1994): 'What is important .
... is not overall population density.... but where the population is concen-
trated'. An alternative view is that it is the overall population growth in a
country which raises demand for food. In many developing countries the
dominant flow of people is from rural to urban areas, and not to the agri-
cultural frontier. So population density is quite low in areas being defor-
ested, although it increases afterwards.
Economic growth and development

Economic growth, which is a rise in the annual production of goods and ser-
vices, also leads to deforestation as it is usually associated with an increase in
agricultural and non-agricultural production and urban-industrial area.
Growth is not always accompanied by economic development, which strictly
requires a sufficiently wide distribution of the benefits of growth to improve
the standard of life of all citizens (Simpson, 1987): mean per capita income
may decline if population growth outpaces economic growth, or income may
be concentrated in the hands of a few. When a real rise in per capita income
does occur it is generally accompanied by an increase in per capita food con-
sumption, which can lead to further deforestation.
On the other hand, economic growth can also reduce the need for defor-
estation by providing capital to invest in more productive forms of agricul-
ture, potentially leading to a general trend from shifting to permanent
agriculture (though there is debate on this), and an ability to improve the
productivity of permanent agriculture. But income generated by growth
may be used instead for other purposes: Thailand is a major exporter of
farm products, but in this century there has been consistent under-invest-
ment in its agricultural sector (Hirsch, 1993).
Poverty and inequality

Westoby (1989) claimed that poverty and inequality, not population
growth, are the principal causes of deforestation. They certainly motivate
people to clear forest in societies in which the distribution of income,
wealth and property rights are skewed in favour of a privileged few, and it
is wrong to assert that just reducing population growth will halt deforesta-
tion. But it is difficult to disentangle poverty and inequality from popula-
tion growth because, according to demographic transition theory, fertility
rates (and net population growth rates) stay high in a developing country
until poverty is alleviated (Notestein, 1945). So the results of regression
tests need careful evaluation, especially as poor data inhibit modelling the
effect of poverty.
Physical factors
Population growth and economic growth alone will not ensure coloniza-
tion of forested areas. Good access by road is essential if settlement is not
to be constrained by proximity to rivers which, until recently, were the
main transport arteries in the humid tropics. Free diffusion of settlers is
possible only if forests are de facto open-access resources, i.e. the state does
not enforce its claim of ownership. Those who clear forest establish a
strong claim to the land. Access is also easier if the remaining forest is frag-
mented. Sloping land is less accessible than flat land, and together with
poor soil fertility makes an area less attractive for farming.
State interoention
In neoclassical economics theory, land and other scarce resources should be
allocated according to their market value. In practice, allocation is often
conditioned by the intervention of the state in order to implement its poli-
cies on agriculture and forestry. For example, from 1960 onwards the
Brazilian Government invested huge sums of money in building a highway
network to facilitate colonization of Amazonia and subsidize the expan-
sion of farming. Sometimes the state merely assists what would otherwise
be an economically efficient allocation of land, but it can also promote the
spread of a land use (such as cattle ranching) which is physically and eco-
nomically unsustainable. As market allocation of land does not take
account of the environmental services provided by forests, since these lack
market values (Pearce and Brown, 1994), the state can play an important
role in compensating for this market failure.
Land-use change may also be influenced by policies which promote
regional development and influence population growth and economic
development. Policies that fail to tackle land tenure inequality, poverty and
unemployment can cause deforestation if they force people to become
encroaching cultivators. Indeed, spontaneous or planned colonization may
be viewed by governments as an easy alternative to land reform (Mahar
and Schneider, 1994). If governments keep food prices down to appease
urban dwellers then farmers have no economic incentive to invest in raising
productivity.
Regression models of the causes of deforestation 311
External influences
Land use also changes as a result of influences from abroad. Thus, rising
demand for exports of plantation crops and minerals can lead to deforesta-
tion, and demand for tropical hardwood exports has driven the expansion
of logging in tropical rain forests. Another influence comes from conditions
attached to the provision of overseas finance. It has been alleged that defor-
estation increased in the 1980s because of the need to raise export income
to repay mounting foreign debts, but there is conflicting evidence for this.
Restructuring schemes imposed by international agencies to control gov-
ernment spending may have had even more effect.
11.4 REGRESSION MODELS OF THE CAUSES OF

DEFORESTATION
11.4.1 Variables and methods

Of the various modelling techniques available (Jeffers, 1975), multiple lin-
ear regression models have been widely used to test the influence of the
underlying causes of deforestation proposed above, by estimating func-
tional relationships between a dependent variable and one or more inde-
pendent variables. Variables must be carefully chosen so that relationships
between them correspond to those seen in practice. Thus, models using
annual deforestation rate as a dependent variable are typically linked with
a number of other flow variables, e.g. annual population growth rate.
Forest cover (the percentage of a country, region or province covered by
forest) is commonly used instead of forest area in cross-sectional models
containing data from a number of spatial units having different land areas.
Some of the models reviewed here test a specific combination of indepen-
dent variables proposed on the basis of theory. Others use a stepwise
approach to test in turn different combinations of independent variables
(also identified by theoretical models) until the one that best explains varia-
tion in the dependent variable is found. The stepwise approach is particu-
larly appropriate in situations such as this, where there is uncertainty
about the processes operating and a large number of potentially influential
variables. In searching for the 'best' equation, modellers take into account,
among other things:
• the coefficient of variation (R2), which estimates the explanatory power
of the model;
• the statistical significance of independent variables, estimated using the t
statistic at significance levels of 5% (*) or 10% (+);
• the F statistic, which estimates the significance of R2.
Models must also conform with the assumptions of the classical linear
regression model (Hebden, 1981). For example, dependent variables should
have normal distributions, and so variables with skewed distributions are
often transformed, e.g. into their logarithms (In = log to the base e).
All models have their limitations. They can only really test the reliability
of prior hypotheses advanced on the basis of a theoretical model, and evi-
dence of a statistically significant relationship does not imply that a causal
relationship also exists (Lambin, 1994). If unexpected relationships appear,
they must be thoroughly scrutinized by reference to the theoretical model
before any inferences are made.
11.4.3 Data and specification problems

The phrase 'garbage in, garbage out' is well known to modellers but is too
often ignored. The quality of data merits careful scrutiny before modelling
starts, otherwise the results will not be reliable. This applies particularly to
models of tropical deforestation, since the quality of international forest
data is so poor, and this is likely to limit all of the models reviewed here.
However, some have been affected more than most.
A common error is to use annual statistics on forest and woodland area
from the FAO Production Yearbook (FAO, 1991) as if they were a reliable
time series of forest area for all countries. For example, they were used by
Shafik (1994) as the basis for an international panel regression study, and by
Osgood (1994) to give a time series for just one country: Indonesia. Even
Allen and Barnes (1985), whose regression study was the first to be published
and is still widely quoted, used this source, claiming that it was available for
more countries and over a longer time period, and referred to all woody
cover. This was surprising given their statement that it conveyed a picture of
deforestation drastically different from that of other sources, and that 'using
it in an analysis could seriously misrepresent the deforestation picture'.
In fact, the statistics in the FAO Production Yearbook are merely those
routinely supplied to FAO's Agriculture Department by governments, and
as in most countries the frequency of national forest inventories is at best
every 10-15 years, they cannot be used as an annual time series. The only
reliable international compilations of estimates of forest area are those
made by FAO's Forestry Department, which does check the quality of esti-
mates before publishing them. It uses national statistics it regards as accu-
rate, but has increasingly commissioned its own surveys for countries
where data quality is poor. It also reports the source and quality of its data.
FAO published four World Forest Inventories using data sent to it by gov-
ernments, but the fifth, due to be published in 1975, was abandoned as
official data were judged to be too unreliable (Persson, 1974). There was
then a gap until FAO, in conjunction with the UN Environment
Programme (UNEP), produced a comprehensive assessment of tropical for-
est resources for 1980 (Lanly, 1981), followed by another for 1990 (FAO,
1993). There is no other reliable, comprehensive, primary global data set,
and even its accuracy is limited (Grainger, 1984, 1993b, 1996b). Other
estimates are either derived from it (Grainger, 1983; WRI, 1992) or are less
comprehensive (Myers, 1989).
Another error is the use of projections instead of estimates of actual
trends. The accuracy of FAO's estimates of deforestation rates for 1976-80
(Lanly, 1981) has been widely questioned, but they were an attempt to esti-
mate change that had already happened. Also included in the report were
projections of possible deforestation rates for 1981-85. Some studies,
including that of Kahn and McDonald (1994), have used these as if they
were estimates of actual forest change.
The results of regression models are also questionable if inappropriate vari-
ables are used. Capistrano (1994), for example, tried to estimate the influence
of the global economy on tropical forest depletion from 1967-85, but repre-
sented the deforested area by the area logged, not the area actually cleared.
11.4.4 Cross-sectional global studies

Cross-sectional regression models use data from a number of countries,
regions or provinces to simulate a trend in the dependent variable that
would be evident for a single geographical unit if sufficient time-series data
were available. In the regression models summarized in this chapter, a uni-
form set of acronyms is used to save space. For ease of reading, where vari-
ables are regressed in logarithmic form this is mentioned in the tables but
not in the text. A reference to a model as Table 11.3.1, for example, refers
to Equation 1 in Table 11.3.
Two cross-sectional regression models were tested by Allen and Barnes
(1985), with the mean national annual Deforestation Rate of closed forest
and open woodland for 1968-78 in 39 developing countries as dependent
variable. Independent variables in the first model were annual Population
Growth Rate 1970-78; annual Per Capita Gross National Product (GNP)
Growth Rate 1960-78; annual Change in Cultivated Land 1970-78; and
annual Change in Fuelwood Consumption and Log Exports 1968-78. The
result was disappointing: R2 was just 11 % and Population Growth Rate the
only significant variable (and at only 10%) (Table 11.3.1a). There could be
three reasons for this. Firstly, poor quality of forest data: estimates of open
woodland area are known to be inaccurate, as it does not receive the same
attention from government forest inventories as the more commercial closed
forest (Lanly, 1982). Secondly, the attempt (acknowledged by the authors)
to cover two different sets of deforestation processes in one model: those in
closed forests (where commercial timber logging is important) and open
woodlands (where fuelwood extraction is important). Thirdly, the inclusion
of change in cultivated land as an independent variable, even though the
authors knew this was correlated with population growth and deforesta-
tion. This violates the assumption of no collinearity in regression models.
Table 11.3 Principal regression models of the causes of deforestation
1. Cross-sectional regressions for 39 developing countries, 1968-78 by Allen and Barnes (1985)1
(a) DEF = 0.0042 - 005 PGR 70 _77 - 0.140 CULTGR 68 _78 + 0.0009 GNPpcGR - 0.007 WOODPR0 68
t= (1.477)+ (0.734) (0.699) 60-78 (0.141)
R2 = 0.113, F = 1.091
(b) DEF = -0.144 + 0.045 LnFC68 + 0.0013 PDENS 68 + 0.0001 GNPpc - 0.0001 WOODPROpc - 0.016 PTNC68
t= (1.337)+ (0.529) (0.932) 68 (2.204)* 68 (2.791)*
R2 =0.353, F = 3.614
2. Cross-sectional regression for 30 countries in the humid tropics, 1976-80 by Grainger (1986)2
LnDEF = -1.611 + 0.379 LnPGR 70-79 + 0.489 LnAL 70_79
t= (3.614)* (2.702)*
R2' =0.52, S.E. = 1.286
3. Cross-sectional regressions for 36 tropical countries, 1976-80 by Rudel (1989)3
(a) Unweighted
LnDEF = 0.133Dd + 0.667 LnFA75 + 0.180 LnPGR 6 0-75 - 0.014 LnGNPpc
S.E. = (0.223) (0.093)* (0.086)* (0.235) 75
R2' =0.77
(b) Unweighted
LnDEF = -3.85 + 0.311 Dd + 0.046 LnFC75 + 0.537 LnPGR 60-75 + 0.216 LnGNPpc
S.E. = (0.313) (0.014)* (0.103)* (0.352)*75
R2' =0.89
(c) Weighted
LnDEF = -0.354Dd + 0.418 LnFA75 + 0.355 LnPGR 6 0-75 + 0.520 LnGNPpc
S.E. = (0.224) (0.105)* (0.069)* (0.172)* 75
RF =0.83
4. Cross-sectional regressions for 55 developing countries, 1981-85 by Kahn and MacDonald (1994)
(a) DEFpc = 26.26 - 13.82 DAF -13.71 DCA - 19.6 D As + 0.00054 FApc - 0.27 LABP -0.012 SPpc + 67.8 DEBT
t = (2.66)* (2.49)* (3.26) (1.92) (1.07) (0.88) (2.31)*
R2 =0.292
(b) DEFpg = -0.022 - 0.0024 DAf -0.011 DCA -0.013 D As + 0.0012 FApc - 0.000012 LABp -0.0034SPpg + 0.26 DEBT
t = (2.66)* (2.49)* (3.26) (1.92) (1.07) g (0.88) (2.31)*
R2 =0.304
5. Cross-sectional regression for 17 provinces in north-east Thailand, 1973-88 by Tongpan et al. (1990)
DEF =-222.526 - 0.099 INC + 17.829 Pc + 0.439 VFP + 0.133 FA + 178.547 NRR
t= (3.28)* (2.99)* (1.84)* (5.56) (2.12)*
RF =0.75
6. Pooled cross-sectional regression for 17 provinces in north-east Thailand by Panayotou and Sungsuwan (1989)
LnFC = -0.411 LnPwd -0.32 LnPup/rice + 0.42 LnINC - 1.51 LnPOP* + 0.7D Bk -0.11 LnRDENS -0.22 LnAin + 0.38Y\ice
_ (4.1)* (1.7)* (4.0)* (9.7)* (4.8)* (1.4) (1.02) (1.9)*
R2- = 0.80 F = 32
A irr Area of irrigated land LABp Labour force as percentage of population

AL70--79 Ease of access (forest area logged 1970-79) LABg Labour force per unit GNP
CULTGR Annual change in cultivated land NRR Access - new rural roads in last 5 years
DAF Dummy variable for Africa Pc Cassava producer price
D As Dummy variable for Asia Pup/rice Price of upland crops/price of rice
DCA Dummy variable for Central America and Caribbean Pwd Mean wood price
DBk Distance from Bangkok PDENS Population density
Dd Dummy variable for data quality PGR Average annual population increment
DEBT Debt service/exports POP Population
DEF Average annual rate of deforestation RDENS Road density
DEFpc Average annual rate of deforestation per capita PTNC Plantation cover (percentage of national land area under
DEFpg Average annual rate of deforestation per unit GNP plantations)
FA Forest area Spc Government spending per capita
FApc Forest area per capita Spg Government spending per unit GNP
FC Forest cover (Percentage of national land area under forest) VFP Value of forest products harvested
GNPpc GNP per capita WOODPRO Annual change in fuel wood consumption and log exports
GNPpcGR Annual growth rate in per capita GNP Yrice rice yield
INC Per capita income
* significant at 5%; + significant at 10%

IAllen and Barnes (1985) did not specify the countries in the dataset for these models.
2See Table 11.4 for dataset.
3The dataset for Rudel (1989) includes: Benin, Burundi, Cameroon, Gabon, Gambia, Ghana, Guinea-Bissau, Ivory Coast, Kenya, Liberia, Nigeria,
Rwanda, Sierra Leone, Tanzania, Togo, Zaire, India, Malaysia, Nepal, Papua New Guinea, Philippines, Sri Lanka, Thailand, Bolivia, Brazil, Colombia,
Costa Rica, Ecuador, Haiti, Mexico, Nicaragua, Panama, Paraguay, Peru, Trinidad and Tobago, and Venezuela.
In the second model, the mean annual Deforestation Rate for 1968-78
was expressed as a function of state variables for 1968, in order to test the
delayed impacts of logging and other factors on deforestation (Table
11.3.1b). Per Capita Fuelwood Consumption and Wood Exports and
Percentage of country under plantation crops were significant at 5%, and
Forest Cover at 10%. Population Density and Per Capita GNP were not
significant, and R2 was still only 35%. The authors could not explain why
plantation area should be a significant variable in this model.
Grainger (1986) focused on closed forests in the humid tropics only, test-
ing various combinations of dependent variables for 30 countries based on
prior hypotheses about the underlying causes of deforestation, and using
data from the FAOIUNEP 1980 tropical forest resource assessment (Lanly,
1981). The dependent variable was Annual Deforestation Rate 1976-80
(in logarithmic form to give a more normal distribution), and the indepen-
dent variables were in logarithmic form too. He tested a fixed combination
of variables to match a theoretical model in which annual Deforestation
Rate 1976-80 was a function of Population Growth Rate, GNP Per
Capita, Access to Forest (represented by Mean Forest Area Logged
1970-79, a surrogate for the length of new logging roads), and
Environmental Constraints on Access. This had an R2 of 49% or 51 %,
depending on whether River Density or Topography was used as the
Environmental Constraints on Access variable, though neither was signifi-
cant, and nor was GNP Per Capita (Table 11.4.6, 11.4.7). He also used a
stepwise approach to find the best combination of variables (Table
11.4.1-5). R2 was greatest (52%) for a model including only Population
Growth Rate and Access to Forests. The addition of GNP Per Capita
decreased the Standard Error but R2 fell too and the variable was not sig-
nificant. The lack of significance of GNP Per Capita could have been due to
its conflicting effects on deforestation, i.e. a rise in income can both cause
deforestation (by increasing per capita food demand) and control it (by
providing the basis for investment in improving agricultural productivity).
This problem could have influenced Allen and Barnes' results too.
Rudel (1989) also rejected data from the FAO Production Yearbook in
favour of more reliable data on deforestation rates for closed forests in
1976-80 from Lanly (1981). He tested two types of regression models on a
set of 36 tropical countries for which data quality was good. This excluded
Burma (Myanmar), Indonesia, Laos and Madagascar, which have quite
large deforestation rates, and his models only had 20 countries in common
with Grainger's. In the first model he found that annual Deforestation Rate
was a function of Population Growth 1960-75 and Forest Area 1975. Per
Capita GNP was included but not significant (Table 11.3.3). In the second
model, he weighted countries by their forest area so that equal units of for-
est area were the basic unit of analysis, and this made Per Capita GNP
1975 significant. When Rural Population Growth was used instead of
Population Growth, it was also significant but the model had a lower R2.
All of Rudel's models had an R2 of about 80%. Population Growth was
lagged to reflect delays in deforestation until people become adults, but the
timing of lags deserves more study. Another question is whether to include
Forest Area in such regressions; it does explain a lot of variation in the
dependent variable, and represents the potential supply of new forest land
to complement the potential rise in demand caused by population growth.
An alternative view is that the deforestation rate should reflect only the
demand for extra farmland, not how much is available, at least until forest
becomes scarce. In a model containing Forest Area and Population Growth
as independent variables (Grainger, 1986), Forest Area was only just signif-
icant at 5% and its R2 was inferior to other models (Table 11.4).
Moreover, as countries with large forest areas tend to have large absolute
deforestation rates - though they are usually not a large proportion of for-
est area (Allen and Barnes, 1985) - this could distort the regression analy-
sis. Rudel responded to this latter concern, in particular, by repeating his
regressions using Forest Cover instead of Forest Area. It was significant in
the unweighted but not the weighted model, and in the former model Per
Capita GNP also became significant (Table 11.3.3c).
A test of the hypothesis that excessive foreign debt promotes forest clear-
ance found that the annual Deforestation Rate Per Person for 1981-85 was
indeed a function of Forest Area and Mean Debt for 1981-85, and annual
Deforestation Rate per Unit of GNP for 1981-85 was a function of the
same variables, plus Labour Force (Kahn and McDonald, 1994). However,
R2 was poor (30%), so including debt did not improve the explanatory
power of previous regressions (Table 11.3.4). If such a relationship does
exist it should probably be lagged, not contemporaneous. The reliability of
the deforestation data used here was queried above.
11.4.5 Sub-national regional cross-sectional studies

The influence of poverty on deforestation was tested by a pooled cross-sec-
tional regression model of the 17 provinces in Thailand's north-east region
from 1973-88, when regional forest cover fell sharply (Tongpan et aI., 1990).
Deforestation data derived from Landsat satellite images were included for
1973-76, 1976-78, 1978-82, 1982-85 and 1985-88. The model showed
that provincial Deforestation Rate was related to poverty (Per Capita income
had a negative coefficient), and to access (New Rural Roads Constructed in
the Last 5 Years). Population Growth Rate was excluded from this regres-
sion, on the grounds that it was strongly correlated with road density,
although in theory deforestation should be related both to the rise in popula-
tion and to ease of access to forest. Cassava Producer Price was significant,
reflecting the huge clearance of forest for cassava cultivation (for export) in
that period. Other variables were Value of Forest Products Harvested, and
Table 11.4 Regression tests for 30 countries in the humid tropics
A. Stepwise Regression
1. LnDEF = -1.473 + 0.581 LnPGR
S.D. 0.128
t= (4.556)"
S.E. = 1.332 R2 = 0.426 D.W. = 1.897
2. LnDEF = - 3.596 + 0.613 LnPGR + 0.330 LnGNPpc

S.D. 0.128 0.253
t = (4.777)* (1.306)
S.E. = 1.316 R2 = 0.460 R2 = 0.420 D.W. = 1.840
3. LnDEF = -2.448 + 0.488 LnPGR + 0.309 LnFA

S.D. 0.135 0.181
t = (3.614)* (1.711)
S.E. = 1.289 R2 = 0.482 R2 = 0.444 D.W. = 1.661
4. LnDEF = -1.611 + 0.379 LnPGR + 0.489 LnAL70-79

S.D. 0.137 0.181
t = (2.759)* (2.702)*
S.E. = 1.286 R2 = 0.548 R2 = 0.515 D.W. = 1.853
5. LnDEF = -2.554 + 0.408 LnPGR + 0.148 LnGNPpc + 0.453 LnAL70-79

S.D. 0.147 0.247 0.453
t= (2.771)* (0.601) (2.345)*
S.E. = 1.218 R2 = 0.554 R2 = 0.503 D.W. = 1.896
B. Complete model
6. LnDEF = -2.554 + 0.406 LnPGR + 0.124 LnGNPpc + 0.480 LnAL70-79 - 0.110 LnRIV
S.D. 0.149 0.254 0.453 0.203
t= (2.715)* (0.489) (2.374)* (-0.540)
S.E. = 1.235 R2 =0.559 R2 = 0.488 D.W. = 1.964
7. LnDEF = -1.992 + 0.537 LnPGR + 0.085 LnGNPpc + 0.496 LnAL70-79 - 0.001 LnTOP
S.D. 0.178 0.267 0.204 0.231
t= (3.015)* (0.320) (2.429)* (-0.006)
S.E. = 1.186 R2 = 0.581 R2 = 0.505 D.W. = 1.853
*significant at 5%. Number of data points in Equation 7 is 27.

AL70-79 Ease of access (forest area logged 1970-79)
DEF Average annual rate of deforestation 1976-80
FA Forest area 1980
GNPpc GNP per capita 1980
PG R Average annual population increment 1970-80
RIV Average density of rivers (per 10 000 km2)
TOP Area above 3000 ftlnational area as percentage
Source: Grainger (1986)
NB. Dataset includes: Cameroon, Central African Republic, Gabon, Ghana t , Ivory Coast,
Liberia, Madagascar, Nigeria, Uganda, Zaire, Burma, Indonesia, Malaysia, Papua New
Guinea, Philippines, Thailand, Bolivia, Brazil, Colombia, Coast Rica, Dominican Rep.,
Eduador, French Guiana t , Guatemala, Guyana, Honduras, Nicaragua, Panama, Peru,
Surinamt, Venezuela. tNot included in Equation 7.
Theoretical models of trends in forest cover 319
forest Area (Table 11.3.5). An earlier cross-sectional study by Panayotou and
Sungsuwan (1989), using separate demand functions for industrial wood,
fuelwood and agricultural land, also showed that forest cover was lowest in
provinces with low income and high road density (Table 11.3.6).
11.5 THEORETICAL MODELS OF TRENDS IN FOREST COVER

Regression models of contemporaneous links between deforestation and its
underlying causes say little about how deforestation rates are likely to
change over time. This section outlines two theoretical models that build
on agrarian change theory to explain long-term trends.
11.5.1 Modelling agrarian change

The evolution of national land use morphology will depend both on the
particular combination of extensification, intensification and involution
responses to the underlying causes of land-use change, and on wider socio-
economic trends. For example, a country's demographic transition, in
which economic development leads to a lower mortality rate and hence a
rise in net population growth rate, is not curtailed until later when people
feel economically more secure and fertility rates fall (Notestein, 1945; Ness
and Ando, 1984). Another influence will be the structural transformation
of agriculture: as an economy develops and urban demand for marketed
food increases, farming becomes more productive, rural people migrate to
urban areas, and the proportion of the labour force employed in farming
declines. There is usually a delay before the absolute size of the farming
population declines (Grigg, 1982), and this will affect the speed at which
the deforestation rate falls.
These factors will lead to a trend in agricultural area that is unique for
each country but is generally similar to the logistic curve found in the clas-
sic model of population growth under constraints (Figure 11.2). Just as
population should move to a limiting value as its growth rate declines, so
too will agricultural area. The latter is ultimately constrained by territorial
boundaries, but the rate of spread will slow down as remaining land is
increasingly inaccessible and less fertile. Agricultural area should eventu-
ally decline as marginal land is abandoned (Grigg, 1980). Other influential
factors include the rate of diffusion of farming innovations to peripheral
areas (Lambin, 1994); the relative shares of investment in agricultural and
non-agricultural sectors (many developing countries are not investing
enough in farming as they try to bypass their agricultural revolution in
favour of an industrial revolution); and bifurcation of land-use evolution in
colonial times, which stunted development of indigenous agriculture and
promoted the growth of a parallel plantation economy (Blaikie and
Brookfield, 1987).
l!?
.3
"S
o
~ 80····················································...................................... .
Qj
"C
c:
::J
as 60
l!?
«
"C
c:
j 40
iii
c:
o
~
z 20
E
~
rf.
Time
Figure 11.2 A model of agrarian change.
11.5.2 The national land use transition

The spread of farming in countries that were mainly forested before human
settlement will lead to a period of substantial reduction in forest cover,
called the national land use transition (Grainger, 1995a). This is charted by
a transition curve, which is the inverse of the agrarian expansion curve just
described (Figure 11.3). The transition ended long ago in most temperate
countries, but is still in progress in many tropical countries (Williams,
1989).
In the early-to-middle stages of the transition, forest protection is likely
to be minimal and deforestation trends largely influenced by processes in
the agriculture sector. In later stages, other forces operate to control defor-
estation so it becomes increasingly independent of population growth.
Each country's transition curve is unique but can be compared with two
alternative models. In the normative model, the transition ends with a sub-
stantial amount of forest cover remaining (Figure 11.3). Deforestation is
controlled largely by market forces in the agricultural sector, which by rais-
ing farm yields reduce demand for new farm land. Pro-active government
policies to protect forest for sustainable timber production and conserva-
tion may also support this. The stability of the eventual normative equilib-
rium between the agriculture and forest sectors will, however, be
threatened if agriculture is unsustainable. In the critical model, deforesta-
tion is controlled only when forest cover falls to a much lower level, and
rising timber prices, the growing cost of timber imports and popular con-
cern about the environment lead to government intervention (to conserve
Theoretical models of trends in forest cover 321
Forest
cover
Normative
- - - - - - - - - - Critical
Time
Figure 11.3 The national land use transition: normative and critical models.
forest for environmental reasons) and market-based protection (to main-

tain domestic wood supplies). Environmental limitations will also come
into playas remaining land becomes increasingly inaccessible for logging
or unsuited to farming. An arbitrary lower limit of 0.1 ha of forest per
capita in critical transitions was proposed by Grainger (1986), based on
the forest area needed to supply a mean per capita annual wood consump-
tion of 1 m 3 • In the Philippines, for example, which had a forest per capita
of 0.13 ha in 1987, deforestation does seem to be coming under control.
11.5.3 The forest transition

The national land use transition may be followed by an increase in forest
area, owing to the need both to supply demand for wood and to give
Forest
Cover Forest
Transition
Forest Replenishment
National Land Use Transition Period
Time
Figure 11.4 Unified model combining the national land use transition, forest transi-
tion and forest replenishment period.
greater priority to conservation, and to make use of marginal land that has
become uneconomic for farming. The shift from a shrinking to an expand-
ing forest area was termed the forest transition by Mather (1992). He
acknowledged that there can be a delay between the end of deforestation
and the beginning of forest replenishment, but his theory could not explain
this, and assumed that the mechanisms which control deforestation were
the same as those leading to the forest transition. Grainger (199 5b) argued
that: the forest transition cannot happen until the national land use transi-
tion has ended; they have different mechanisms; and a considerable time
may elapse between the two events. But it is possible to integrate the two
transitions in a single model that includes a delay between the end of the
national land use transition and the start of forest replenishment (Figure
11.4). A similar V-shaped curve was proposed to link local forest cover
and population density in Asian social forestry projects (Romm, 1991).
11.5.4 Deforestation and sustainable development

National land use transition theory could also be used in sustainable devel-
opment assessment. In a sustainably developing country (Pearce, 1993),
each resource stock should ideally follow an optimum depletion path as
natural capital is transformed into human capital. The normative transi-
tion curve could correspond to this path for the forest resource component,
though it is debatable whether the normative curve should be one that
maximizes the total net return from forestry and agricultural uses of
national land area, based on market values, or should also take account of
non-market values (Barbier et a!., 1991), which would lead to a less rapid
forest depletion and more forest remaining at the end of the transition.
11.6 MODELLING TRENDS IN FOREST COVER

Hypotheses about long-term trends in forest cover can be tested by either
national time-series regressions or cross-sectional international and
national regressions. The latter were the first to be attempted, owing to the
general lack of historical data on trends in tropical forest cover.
11.6.1 Cross-sectional international models

Population density is often used as a surrogate for the level of national
development in cross-sectional tests. A test for 42 countries in the humid
tropics found that Forest Cover did indeed decline as Population Density
rose, but levelled off at high population density (Figure 11.5). This sup-
ported the idea of a bounded national land use transition, so that forest
cover will not decline much at high population density, even if the popula-
tion continues to rise. Curves for the three main tropical regions were not
Modelling trends in forest cover 323
100
• •
~
~ 80
Q;
~
() 60
iii
~
.r 40
20
0
-5 -4 -3 -2 -1 o
Log POP Den
Figure 11.5 National forest cover versus logarithm of population density for 43
countries in the humid tropics, 1980.
identical: the limiting value of forest cover was 15-20% for Asia-Pacific,
but below 10% for Africa and Latin America. This difference could be
partly explained by Asia's higher climax closed forest cover (Grainger,
1986). A linear regression model of Forest Cover as a function of Log
(Population Density) was estimated and had an R2 of 0.38 (the distribution
of Forest Cover was relatively normal) (Table 11.5.1).
Palo et al. (1987) used correlations between different variables to test
hypotheses about the causes of deforestation and cross-sectional regres-
sion, in order to estimate a relationship between national Forest Cover and
Population Density (Table 11.5.2), regarding the latter as a surrogate for
domestic demand for forest and forest products, not the level of develop-
ment. This functional relationship varied from one region to another, so
when they built models to project future trends for 60 tropical countries
they divided them into four regional groups (Tables 11.6, 11.9). Forest
cover included both closed and open forest, so the reliability of their pro-
jections was affected by the inaccurate estimates of open forest areas which
had to be used.
11.6.2 Cross-sectional national models

Valuable insights also emerge from cross-sectional analysis within individ-
ual countries. This requires many data points: the Philippines, with over 70
provinces, is ideal for such a test and was the subject of three cross-sec-
tional analyses for 1957, 1970 and 1980 by Kummer (1992). He found
that Forest Area was a function of (a) Road Density in 1957; (b) Road
Density and Population Density in 1970; (c) Road Density, Length of New
Table 11.5 Principal regression models of long-term trends in deforestation
1. Cross-sectional regression for 42 countries in the humid tropics, 1976-80 by Grainger (1986)
FC = 21.77 - 12.54 LnPDENS R2 = 0.38
t = (5.04)*
2. Cross-sectional regression for 60 tropical countries by Palo et al. (1987)
FC = 71.21 - 0.22 PDENS R2 = 0.60
t= (9.5)*
3. National time-series regression of demand for agricultural land in Thailand, 1961-89 by Panayotou and Parasuk (1990)
LnLAND, = 0.0811n P Agt _ 1 + 1.337 LnPOP Agt - 0.279 Y t- I - 0.155 LnDIVt - 0.308 LnIND t + 0.351 D
t= (1.99) (12.82)* (2.43)* (3.95)* (3.37)* (4.12)'
Rr = 0.8975, D.W. = 1.998
4. National time series regression of the trend in forest cover in Thailand, 1976-89 by Katila (1992)
LnFC = -0.237 - 1.1203 LnPDENS + 0.2319 LnGDPpc - 0.1788 LnPwd - 0.28Y
t = (2.14)+ (0.959) (3.94)* (2.28)*
R2 = 0.9667, S.E. = 0.018, F = 988.5, D.W. = 1.95
• significant at 5%; + significant at 10%.

D Dummy variable
DIV Crop diversification index
FC Forest cover (percentage of national land area under forest)
IND Degree of industrialization
GDPpc GDP per capita
LAND Demand for agricultural land
PDENS Population densiry
POPAg Agricultural population
PAg Mean agricultural price
PWd Mean wood price
Y Mean agricultural productiviry
Table 11.6 Three alternative regression models of forest cover (FC) versus population density (PD) for four tropical country groups,
1980
Country group B C D
Africa (moist) FC = 80.404 - 0.449 PD FC = 98.46 - 11.14 LnPD LnFC = 4.763 - 0.210 LnPD
Africa (semi-moist) FC = 61.442 - 0.210 PD FC = 89.35 - 12.22 LnPD LnFC =5.178 - 0.466 LnPD
Asia FC = 82.695 - 0.225 PD FC = 169.80 - 25.39 LnPD LnFC = 6.257 - 0.516 LnPD
Latin America FC = 73.821 - 0.232 PD FC = 111.31 - 16.57 LnPD LnFC = 5.320 - 0.471 LnPD
Country groups:
Africa (moist): Benin, Cameroon, Congo (Brazzaville), Gabon, Ghana, Guinea, Guinea-Bissau, Equatorial Guinea, Ivory Coast, Liberia, Nigeria, Sierra
Leone, Togo, Congo (Zaire).
Africa (semi-moist): Angola, Burundi, Botswana, Burkina Faso, Central African Republic, Ethiopia, Gambia, Malawi, Rwanda, Senegal, Tanzania,
Uganda, Zambia, Zimbabwe.
Asia: Brunei, Myanmar, India, Indonesia, Kampuchea, Laos, Malaysia, Philippines, Sri Lanka, Thailand, Vietnam.
Latin America: Belize, Bolivia, Brazil, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, French Guiana, Guatemala, Guyana,
Honduras, Haiti, Jamaica, Mexico, Nicaragua, Panama, Peru, Paraguay, Suriname, Venezuela.
Source: Palo et al. (1987).
Roads 1970-80, and Population Density in 1980 (Table 11.7). Provinces
lacking forest cover in 1970 were excluded. Kummer suggested that the
presence of Population Density only in the 1970 and 1980 models might
indicate increasing migration to frontier areas from the 1960s onwards,
but he acknowledged that it was not clear whether his road data included
the logging roads that provided migrants with the closest access to forest.
The use of Forest Area as dependent variable in a cross-sectional analysis
of provinces of different sizes was questionable.
Table 11.7 Regression tests for the Philippines 1957, 1970, 1980
1. LnFA s7 = -0.87 LnRDENSS7

t= (12.6)*
R2 = 0.76, F = 159
2. LnFA70 = -0.54 LnPDENS70 - 0.28 LnRDENS 70
t = (4.7)* (2.4)*
R2 =0.58, F =48
3. FA so =0.32 NEWROADS70-S0 - 0.29 PDENS so - "" 0.38 LnRDENS so
t= (3.5)* (3.5)* (2.6)*
R2 =0.41, F = 17
FA Forest area
NEWROADS 70-80 Length of new roads 1970-80
PDENS Population density
RDENS Road density
Source: Kummer (1992).
11.6.3 National time-series models

The Philippines and Thailand seem ideal for testing the above theories of
long-term trends in forest cover, because they have a long history of defor-
estation and good data are available on forest trends over the last 30 years.
Forest cover in both countries fell by a half between 1960 and 1990, reach-
ing 24% in 1987 and 28% in 1988, respectively, and appearing to
approach limits of 20 and 22 % as deforestation came under increasing
control (Figure 11.6).
Panayotou and Parasuk (1990) modelled deforestation in Thailand by
estimating Demand for Cultivated Land for 1961-89 (Table 11.5.3). Their
model implies that Agricultural Price and Agricultural Population were dri-
ving forces of deforestation, and Agricultural Productivity, Crop
Diversification and Industrialization were controlling forces. The latter
would have been expected if industrialization was a surrogate for the level
of development, but industrialization also raises demand for land for non-
agricultural uses, which can lead to more deforestation. The explanatory
power of the model seems high, but R2 for time-series regressions generally
exceeds that for cross-sectional models.
80
60
'.
~
..
~ '. " .
~
....
~
()
40
U; .... ....
~
~ Philippines
20
..................... Thailand
0
1900 1920 1940 1960 1980 2000
Figure 11.6 The national land use transitions of the Philippines and Thailand.
Katila (1992) identified the Price of Wood and Agricultural Productivity

as significant variables influencing Forest Cover in a national and time-
series regression for Thailand for 1976-89 (Table 11.5.4), although Forest
Cover fell as Agricultural Productivity rose, which was not expected.
Population Density was significant only at the 10% level.
To test for the influence of controlling forces, Francisco et al. (1996) esti-
mated national time-series models for the Philippines and Thailand for the
entire period 1960-90, and for 1975-90 only (Table 11.7). If the transi-
tions had followed the normative model then farm productivity should
have been a significant independent variable for 1975-90, but it was signif-
icant only in the Philippines: mean Rice Yield rose faster than population
there but not in Thailand, where it is only 75% of that in the Philippines.
Corn Yield was a controlling factor in the Philippines, which was expected
as most deforestation now takes place in the uplands where corn is the sta-
ple crop. But Rice Yield appeared to be a driving, not a controlling vari-
able. This was either an anomaly, or showed that the Green Revolution
and land reform measures exacerbated deforestation by making many
small-scale Filipino farmers landless.
If the transitions had followed the critical model they would be controlled
by state intervention and environmental constraints. State intervention has
been crucial in both countries but it is difficult to test this quantitatively.
Environmental constraints were not evident in the national regressions, but
did occur in provincial cross-sectional regressions for both countries, which
showed that deforestation rates were highest on poor quality land or
uplands. Characteristics of the critical transition model were therefore found
in both countries, but a key feature of the normative model was present in
the Philippines.
Table 11.8 Time-series regression models of trends in forest cover in the
Philippines and Thailand, 1960-90 and 1975-90
Philippines
1. 1960-90
FC = 50.2 - 0.145PDENS + 4.73Yeorn - 0.0002WOOD - 14.61RDENS + O.13Psuj:ar
t= (12.0) (3.42) (3.52) (13.66) (2.43]
R2 =0.999, D.W. =2.15
2. 1975-90
FC = - 17.37 + 0.69PRUR - 2.69Yriee +3.21 Yeorn + 0.0073Priee
t= (4.8) (2.89) (2.28) (2.58)
R2 =0.997, D.W. =2.6
Thailand
3. 1960-90
FC = 106 - 0.266PDENS - 2.918PUR + 0.306PSCPC
t = (7.34) (5.41) (3.98)
R2 =0.992, D.W. 1.229
4. 1975-90
FC = 45 518 - 0.299PDENS + 1.67PSCPC - 0.988PNFOOD + 0.000755WOOD
t = (18.04) (12.51) (3.58) (4.56)
R2 =0.994, D.W. =2.69
PC Percentage forest cover
PDENS Population density
PNFOOD Percentage area of non-food crops
PRUR Percentage rural population
PSCPC Ratio of shifting cultivate/permanent cultivation areas
PUR Percentage urban population
Price Price of rice
Ps~r Price of sugar
RuENS Road density
WOOD Annual roundwood production
Yeorn Mean corn yield
Yrice Mean rice yield
Source: Francisco et al. (1996)
11.6.4 Projecting long-term trends in deforestation
Regression models
One way to project future trends in forest area is to use regression models
of past trends. By using a regression model estimated for 1961-89 (see sec-
tion 11.6.3) and making assumptions about future trends in independent
variables, Panayotou and Parasuk (1990) projected future demand for agri-
cultural land in Thailand between 1990 and 2010. This predicted a decline
of 11-25% in agricultural area over that period, depending on such factors
as the rate of rise in farm productivity and the level of government subsi-
dies to farmers.
In the absence of sufficient time-series data at an international level, Palo
et al. (1987) used three regional cross-sectional regression models of the
relationship between forest cover and population density in 1980 for 60
tropical countries (Table 11.9) to predict alternative scenarios for regional
forest areas in 2000 and 2025, and the corresponding deforestation rates.
Compared with 1980, total forest cover (including both closed and open
forest) fell by 18% by 2025 using the linear model (scenario B), and by 30
and 45% using the two non-linear models (scenarios C and D). Merely
continuing the projections of Lanly (1981) of deforestation rates for
1981-85 led to an 18% fall (Table 11.9). FAO (1993) later used this
approach in their 1990 tropical forest resource assessment to adjust the
most recent empirical estimates for all countries to the same year (1990).
Projecting deforestation trends in this way has major limitations. Firstly,
there is likely to be tremendous variation in forest cover remaining in differ-
ent countries, even at equivalent stages of their socio-economic development,
owing to differences in climax forest cover and historical experiences.
Secondly, this function will be mainly determined by farming, not forestry,
activity. Population-supporting capacity varies greatly between and within
countries, according to land capability (based on soils, topography etc.) and
how this is enhanced by artificial inputs (Higgins et ai., 1983), so a determin-
istic relationship between forest cover and population density would not be
expected at the national level. Both Palo et al. (1987) and FAO (1993) were
careful to minimize the effects of these limitations. Palo et al. (1987) used
regional, not global curves for their predictions, and made projections only
for those regions, not for individual countries. FAO did make national pro-
jections, but only for a short period, typically less than 10 years.
Simulation models
An alternative approach is to simulate alternative future scenarios using a
model whose variables include the key underlying causes of deforestation.
Grainger (1986, 1990a) devised a simple model, based on the systems
model in Figure 11.1, to simulate trends in national forest areas and defor-
estation rates. This model links the expansion of total agricultural area (A)
to population in year t (Pt ), and to annual growth in population (p), per
capita food consumption (a), and yield/ha (f3). Population was assumed to
grow logistically, initially at average 1970-80 rates (p) but only up to a
limit of Pmax' a hypothetical stable population estimated by the World
Bank (1982). With k a constant for each country, the area of farmland A in
year t (At) was given by:
A .P . e(a-{J)t
A= 0 max
[1]
t Po. (1 + k.e- pt )
The annual deforestation rate was calculated as the additional area of farm-
land required in that year, except in countries where forest area per capita
reached an arbitrary limit of 0.1 ha (see above), when it was set to zero.
The model was simulated from 1980 to 2020 for 43 individual countries
containing 95% of all tropical moist forest. Two alternative scenarios, 'high'
Table 11.9 Four scenarios for trends in total forest area in 60 tropical countries for 2000 and 2025 (thousand sq. km)
Country group Scenarios

A B C D
1980 2000 2025 2000 2025 2000 2025 2000 2025
Africa (moist) 3020 2818 2565 2616 1929 2421 2133 2276 2005
Africa (semi-moist) 2688 2664 2634 2721 2431 2335 1953 1743 1305
Asia 2855 2516 2092 2800 2044 2493 2062 2338 2039
Latin America 9014 8191 7161 8892 8324 7241 6220 5375 4324
Scenarios B, C and D are projected using the corresponding models in Table 11.6. Scenario A is a linear projection of deforestation rates for the same
countries using projections for 1981-85 by Lanly (1981). Country groups as in Table 11.6.
Source: Palo et al. (1987)
and 'low', were constructed with reference to mean regional values of a and
f3 for 1970-80 (Table 11.10). The results are shown in Table 11.11, with
national trends aggregated by region. The simulated annual deforestation
rate in 1980 for all tropical moist forest was 4.1 and 6.6 million ha in the
Low and High scenarios, respectively, which compares with the estimated
actual deforestation rate in those countries of 5.6 million ha (Lanly, 1981).
Between 1980 and 2020 the annual deforestation rate fell from 6.6 to 3.7
million ha in the High scenario and from 4.1 to just 0.9 million ha in the
Low scenario (Figure 11.7). Total forest area fell by about 20% in the High
scenario and by less than 10% in the Low scenario. (If deforestation rates
had continued unchanged at estimated 1976-80 levels, the reduction would
have been 23%.) Despite the decline in regional deforestation rates, national
deforestation rates rose initially in about half of the countries.
The model was later adapted to simulate trends in the emissions of car-
bon dioxide from deforestation that contribute to global climate change.
Mean forest biomass density in tonnes of carbon per ha was estimated for
each country, deforestation was divided evenly between permanent and
shifting agriculture (split 30:20% between short-rotation shifting cultiva-
tion and encroaching cultivation), and the loss and regeneration of soil car-
bon was also included. Simulated emissions in 1980 were 0.4 and 0.7
gigatonnes of carbon (1 Gt C = 10 15 g C) per annum in the Low and High
deforestation scenarios, respectively, compared with an estimate of
0.4-1.6 Gt C per annum for that year by Detwiler and Hall (1988).
Emissions fell to 0.1-0.4Gt C per annum by 2020 (Table 11.12). Total
emissions for the tropics as a whole were simulated by adding 0.1 Gt C per
annum from the deforestation of dry forests and assuming that this
remained constant throughout the 40 years (Grainger, 1990b).
Future rates of deforestation and carbon emissions could be higher if:
(a) agriculture is not sustainable, resulting in declining yields and more defor-
estation; (b) construction of logging roads catalyses deforestation; (c)
governments promote deforestation; (d) external demand for cash crops
Table 11.10 Assumed increases in per capita food consumption (a) and yield per
ha (~) percentage per annum
Region High scenario Low scenario
a /3 a-/3 a /3 a-/3
Africa 0.5 1.0 -0.5 0.0 1.0 -1.0
Asia-Pacific 1.5 1.5 0.0 1.5 2.0 -0.5
Latin America 1.5 2.0 -0.5 0.5 1.5 -1.0
Modified values of <X-~ for the high and low scenarios, respectively: Gabon, -0.2, -0.5;
Madagascar, -1.0, -1.5; Brazil, -0.5, -0.8; Colombia, -0.5, -0.8; French Guiana, +0.5,
-1.0.
Table 11.11 Regional trends in regional deforestation rates and forest areas in the humid tropics 1980-2020 based on national
simulations
High scenario Low scenario

1980 1990 2000 2010 2020 1980 1990 2000 2010 2020
Deforestation rate (million ha per annum)
Africa 1.552 1.486 1.207 0.885 0.866 1.036 0.935 0.670 0.559 0.427
Asia-Pacific 1.731 1.505 1.219 1.192 1.149 1.128 0.940 0.707 0.526 0.417
Latin America 3.275 3.065 2.705 2.221 1.670 1.971 1.599 1.111 0.575 0.036
Humid tropics 6.558 6.056 5.131 4.298 3.685 4.135 3.474 2.488 1.660 0.880
Forest area (million ha per annum)
Africa 198.9 183.5 170.3 160.4 151.6 198.9 188.9 181.1 175.0 170.1
Asia-Pacific 239.4 222.8 209.5 197.5 185.8 239.4 228.9 220.8 214.8 210.2
Latin America 598.0 566.2 537.5 513.0 493.8 598.0 580.1 566.7 558.6 555.8
Humid tropics 1036.3 972.6 917.2 870.8 831.1 1036.3 997.9 968.7 948.5 936.1
These are actual simulated values for the years specified, not averages. Estimates of deforestation rates for 1980 by Lanly (1981) for the countries
simulated were: Africa 1.1; Asia-Pacific 1.3; Latin America 3.1 (million ha per annum).
Source: Grainger (1986, 1990a).
7
I )I: High Scenario - Low Scenario I
6
~ 5
II:
c:: 4 --- ---. --
--- --- --
0
E
-- --- ---
I/)
~
3
.$2
-- ---
Q)
--- -- -- ---
Cl 2
O+-----------r---------~----------_+----------~
1980 1990 2000 2010 2020
Figure 11.7 Two scenarios for deforestation rates in the humid tropics, 1980-2020
(million ha per annum).
leads to a further expansion of plantations - but simulated deforestation rates

were close to actual rates even though the model did not include export demand.
A more comprehensive model of tropical deforestation's contribution to
global climate change was built by Rotmans and Swart (1991) as part of
IMAGE (Integrated Model to Assess the Greenhouse Effect). It included
both closed and open forests, accounted for transfers between forests and
arable land, grassland and other land, and estimated the impact of trans-
fers on the global carbon budget. Demand for new land for permanent cul-
tivation, shifting cultivation and pasture was projected separately. For
example, cumulative demand for permanent cultivation land in year t
(D Il')'t) was modelled using:
[2]
where P is population, Ce cereal consumption, SSR food self-sufficiency

ratio, Y yieldlha, ex the fraction of cereals used for food, ~ the fraction of
area for food from non-cereals, and y the fraction of area for other farm
products. The model improved on that of Grainger (1986) by including the
impact of soil degradation on deforestation and allowing agriculture to
expand preferentially onto the most productive lands. The spread of log-
ging was projected by logistic curves. Alternative scenarios for areas of
closed and open forests were simulated from 1980-2100 for the three trop-
ical regions, and trends in net carbon emissions were calculated. Because
fossil fuel emissions are so dominant, changing from the most optimistic
scenario to the most pessimistic - in which all tropical forest disappeared
by 2050 - only raised total carbon emissions by 10%. To make it feasible
Table 11.12 Trends in carbon emissions from tropical deforestation (Gte per annum)
High scenario Low scenario

1980 1990 2000 2010 2020 1980 1990 2000 2010 2020
Africa 0.192 0.187 0.158 0.131 0.125 0.132 0.122 0.902 0.756 0.609
Asia 0.211 0.193 0.164 0.157 0.151 0.138 0.120 0.949 0.712 0.557
Latin America 0.298 0.281 0.250 0.208 0.158 0.179 0.150 0.106 0.578 0.089
Humid tropics 0.702 0.661 0.572 0.495 0.434 0.449 0.391 0.291 0.205 0.126
All tropics 0.797 0.757 0.668 0.591 0.530 0.544 0.486 0.387 0.300 0.221
The total for All tropics includes the following constant values of emissions from the dry tropics: Africa 0.043, Asia 0.009, Latin America 0.044, Total
dry tropics 0.096 (Gte per annum).
Source: Grainger (1990b).
Modelling long term spatial deforestation trends 335
to model such a large number of processes comprehensively, however,
required simulation of land-use change on a regional rather than a national
scale, and obscuring the large differences between trends in different coun-
tries probably had an impact on overall trends.
11. 7 MODELLING LONG-TERM SPATIAL DEFORESTATION

TRENDS
The models reviewed so far have all been very aggregated, covering an
entire country or sub-national region, but there is growing interest in mod-
elling spatial patterns of deforestation in ways that reflect spatial variation
in socio-economic, political and environmental factors.
11.7.1 Theories of land allocation and development
Land allocation and infrastructure

National land use morphology evolves through an iterative allocation of
land to new uses as appropriate to its economic value at a given time.
Ideally, the flatter the topography and the more fertile the soil, the more
attractive an area should be for farming and the greater the chance of
deforestation, although desperate landless people do not always follow this
rule. Allocation is also influenced by ease of access and the distance to the
nearest market. The relationship between land allocation and transport
costs to an urban centre was the basis of Von Thiinen's theory of circular
land-use zones, in which the economic rent captured by land users is a
function of distance to the centre (Hall, 1966). Zones nearest to the centre
will be allocated to more intensive uses than those further away.
Uneven development
Land allocation will also be politically conditioned. Core-periphery theories
of political economy attempt to explain the uneven patterns of development
found on a global scale by dividing the world into a dominant core, com-
prising the major industrialized countries, and a less powerful periphery,
comprising the developing world. In Dual Economy Theory, for example,
the growth of the core leads to the growth of the periphery (Hicks, 1969),
but in Modern World System Theory it leads to the underdevelopment of
the periphery, whose surplus production it expropriates (Taylor, 1989).
Political ecology theories go further in regarding environmental degradation
at the local level in developing countries as linked through a series of nested
relationships to the global core (Blaikie and Brookfield, 1987). Areas of a
country that are distant from the main urban centres tend to be the most
socially, economically and politically marginal. People migrating there may
be escaping from the constraints of unequal land tenure in more fertile core
areas, or have lost their jobs as farm labourers when farms there were mod-
ernized. Environmental management in peripheral areas is often poor,
owing to lack of local capital and inadequate state control.
Dynamic allocation patterns
In the early-to-middle phases of national land-use evolution, the spread of
settlement into forested areas may be characterized by a visible frontier, but
this varies between countries. For while in Amazonia an agricultural fron-
tier preceded the spread of logging; in South-East Asia the spread of settle-
ment and infrastructure was usually pioneered by a logging frontier, with
farmers following on behind.
After passing through a number of stages of land allocation it might be
expected that national land-use morphology would ultimately move closer
to an asymptote represented by the national land suitability classification,
as modified by economic considerations relating to proximity to urban cen-
tres. The country would then be as evenly developed as natural environ-
mental conditions permit, and each area of land would be used in the most
sustainable way.
11.7.2 Modelling spatial trends in deforestation

Various spatial statistical models have been developed to test hypotheses,
such as those proposed in sections 11.3 and 11.7.1, about the influence of
access, infrastructure and land capability on deforestation. Thus, changes
in land cover were associated with slope, elevation, and proximity to
houses, roads and rivers in Thailand (Nualchawee et aI., 1981) and
Honduras (Ludeke et aI., 1990). The spread of deforestation was linked to
proximity to roads in Costa Rica (Sader and Joyce, 1988) and the
Philippines (Liu et aI., 1993), and to the degree of fragmentation of exist-
ing forest in Peninsular Malaysia (Brown et aI., 1993).
Such models can also be used to make short-term predictions about pos-
sible future spatial trends. For example, Hall et al. (1994) linked deforesta-
tion patterns in Asia and Central America to elevation and slope, and
Pontius and Hall (1994) extended this to predict the spread of new land
uses into forest in Africa according to cloud cover, rainfall, temperature,
soil, aspect, and proximity to cities and rivers. These models have a supe-
rior predictive capability to Markov models, which can only predict future
spatial trends on the probability of past change. One of the earliest spatial
modelling studies of tropical deforestation used a Markov model to predict
future patterns of change in north-east Thailand, but the inflexibility of
constant transition probabilities was very apparent (Miller et aI., 1978).
The expansion of the agricultural frontier depends in practice on decisions
made by land managers on the ground, and they should receive economic
and environmental signals as remaining land becomes increasingly marginal
and leads to diminishing economic returns. Panayotou and Panasuk (1990)
estimated that for every 10% rise in cultivated land in Thailand between
1961 and 1987, agricultural output rose by only 4.8%, owing to the lower
productivity of newly cleared land. Southgate (1990) built a partial equilib-
rium model of trade-offs made by spontaneous settlers between the need to
Modelling long term spatial deforestation trends 337
Forest within designated areas
Forest outside designated areas
Designated areas outside forest
200km
Figure 11.8 Comparison of remaining forested area in Thailand with land desig-
nated for forest cover by a land suitability classification.
clear land and the need to capture forestry and non-market rents, e.g. by soil
conservation, and found that settlers gave priority to clearance so they could
capture agricultural rents and establish property rights.
Thailand provides an interesting example of long-term spatial trends in
land use. Most of its remaining forest cover is now concentrated in periph-
eral upland regions. Only 58% of the country is suitable for sustainable agri-
culture, according to a national land suitability classification (GOTIFAO,
1972), and the government has formally adopted 40% as the optimum forest
cover. This is much higher than the present value (see above). The distribu-
tion of forest land allocated sub-optimally was mapped by Grainger et al.
(1996), using geographic information system (GIS) techniques to compare
digital maps of current forest distribution and land suitability (Fig. 11.8).
11. 7.3 Modelling spatial trends in logging
Factors detennining spatial patterns of tropical hardwood allocation

The second major human impact on tropical moist forest, selective logging,
is a response to an entirely different demand function from that driving
deforestation, i.e. demand for wood rather than demand for land. Much of
tropical hardwood production is consumed in tropical countries, but at
least half is exported overseas. So it is not possible to project demand for
tropical hardwood in a country purely on the basis of national factors.
Moreover, the global distribution of commercial tropical hardwood species
and their share of total wood volume in forests is not uniform. This,
together with differences in transport costs, labour costs and government
taxes, makes forests in some countries much more economically attractive
for logging than those in others, and so explains why South-East Asia has
supplied 80% of all tropical hardwood exports since the 1960s.
A model of the tropical hardwood trade 1980-2020

These factors were taken into account when constructing the TROPFORM
model to simulate long-term trends in logging in the humid tropics. It pro-
jected demand for tropical hardwood in five main importing regions:
Japan, Europe, USA, Other Asian Processing Countries (South Korea,
Taiwan and Singapore), and the Rest of the World; and domestic demand
in the three main producing regions of Africa, Asia-Pacific and Latin
America. This demand was satisfied by using a linear programming model
to minimize the cost of producing wood in 21 producing regions (19
individual countries and two groups representing 11 countries in Africa
and Asia) and shipping it to the consuming regions. Each producing region
had its own forest model accounting for wood reserves lost by logging and
deforestation and gained by natural regeneration and growth of planta-
tions. The unit logging cost was estimated by a universal extraction cost
function and the shipping cost was the sum of internal and international
transport costs (Grainger, 1986).
Conclusions 339
In the Base scenario, which assumed the Low deforestation scenario
described above (Table 11.11), tropical hardwood exports doubled by the
end of the century and then fell below current levels by the year 2020,
owing to depleted reserves and rising domestic consumption (Table 11.13).
Plantations were unable to compensate for the shortfall in supplies from
natural forests in this period. There was also a shift from Asia-Pacific to
Latin America as leading exporter at the start of the next century, as low-
cost Asian timber reserves were depleted (Figure 11.9). To assess the impact
of deforestation, the Base scenario was modified by including the High
deforestation scenario instead. This caused exports to peak earlier and then
decline parallel to the Base scenario trend, so they were 12% lower by 2010
and 35% lower by 2020 (Figure 11.10). If deforestation rates are indeed
higher than assumed in the Base scenario, the prospects for tropical hard-
wood exports could therefore be much worse than anticipated, and this makes
controlling deforestation all the more important (Grainger, 1986, 1987).
11.8 CONCLUSIONS
This chapter has shown how theoretical models can be built to give insight
into contemporary and historical deforestation processes, and lead to
hypotheses that can be tested by applying mathematical modelling tech-
niques to empirical data. Models can also be used to simulate future trends
in deforestation and its impacts more rigorously than before, and to sug-
gest how to make deforestation control programmes more effective.
Current models are still quite primitive, however, and need to be improved
by undertaking more empirical studies of deforestation processes. A new
Table 11.13 Three TROPFORM scenarios (Base, High, HighDef) for general
trends in the tropical hardwood trade 1980-2020 (million m3 )
1980 1985 1990 1995 2000 2005 2010 2015 2020

Removals
Base 86.7 109.2 137.6 173.1 195.6 216.8 155.4 126.6 106.1
High 86.7 115.6 153.7 181.0 215.0 151.8 127.0 86.9 68.6
HiDef 86.7 109.2 137.6 173.2 180.7 163.8 131.4 99.4 76.6
Exports
Base 50.6 63.4 79.2 98.4 121.9 151.0 88.8 62.6 35.9
High 50.6 66.5 86.9 112.7 145.4 90.7 56.2 21.0 15.4
HiDef 50.6 63.4 79.3 98.4 122.0 111.3 78.2 43.0 23.3
Domestic consumption
Base 36.1 45.8 58.4 74.7 73.7 65.8 66.6 64.0 70.2
High 36.1 49.1 66.8 68.3 69.6 61.1 70.8 65.9 53.2
HiDef 36.1 45.8 58.3 74.8 58.7 52.5 53.2 56.4 53.3
Total consumption = total removals; total imports = total exports.
Source: Grainger (1986).
120
100
Latin America
80
!
~
60
Asia-Pacific
40
20
O+---~----~--~----~----r---~----+---~
1980 1985 1990 1995 2000 2005 2010 2015 2020
Figure 11.9 TROPFORM simulation of regional trends in tropical hardwood

exports, 1980-2020, Base scenario (million m3 ).
160
140
120
Base
.l!l 100
0
c.
)(
w
80
60
40
20+-----~--~~--_+----~----~----+_----+_--~
1980 1985 1990 1995 2000 2005 2010 2015 2020
Figure 11.10 Effects on tropical hardwood exports of a higher deforestation rate:

Base scenario (includes low deforestation scenario) versus HiDef scenario (includes
High deforestation scenario) (million m3 ).
References 341
generation of spatial deforestation models will be required to meet increas-
ingly sophisticated demands for information about the environmental
impacts of deforestation and its role in global environmental change, but as
these will be even more data-intensive than the present generation, mod-
ellers are still likely to be constrained by lack of data. So the reliability of
modelling will continue to depend on available data being used wisely.
REFERENCES
Allen, J. C. and Barnes, D. F. (1986) The causes of deforestation in developing coun-
tries. Annals of the Association of American Geographers, 75,163-84.
Barbier, E., Burgess, ]. C. and Markandya, A. (1991) The economics of tropical
deforestation. Ambio, 20, 55-58.
Barbier, E., Bockstael, N., Burgess, J. and Strand, I. (1994) The timber trade and
tropical deforestation in Indonesia, in The Causes of Tropical Deforestation, (eds
K. Brown and D. W. Pearce), UCL Press, London, pp. 242-270.
Bilsborrow, R. and Geores, M. (1994) Population, land use and the environment in
developing countries: what can we learn from cross-national data?, in The Causes
of Tropical Deforestation, (eds K. Brown and D. W. Pearce), UCL Press, London,
pp.106-133.
Binswanger, H. and Ruttan, V. (eds) (1978) Induced Innovation: Technology,
Institutions and Development, Johns Hopkins University Press, Baltimore.
Blaikie, P. and Brookfield, H. (eds) (1987) Approaches to the study of land degrada-
tion, in Land Degradation and Society, Methuen, London, pp. 27-48.
Boserup, E. (1965) The Conditions of Agricultural Growth: The Economics of
Agrarian Change Under Population Pressure, Aldine, Chicago.
Brown, K. and Pearce, D. W. (eds) (1994) Saving the world's tropical forests, in The
Causes of Tropical Deforestation, UCL Press, London, pp. 2-26.
Brown, S., Iverson, L. R. and Lugo, A. E. (1993) Land use and biomass changes in
Peninsular Malaysia during 1972-82: a GIS approach, in Effects of Land Use
Change on Atmospheric Carbon Dioxide Concentrations, (ed. V. H. Dale),
Springer-Verlag, New York.
Capistrano, A. D. (1994) Tropical forest depletion and the changing macroeconomy,
in The Causes of Tropical Deforestation, (eds K. Brown and D. W. Pearce), UCL
Press, London, pp. 68-85.
Detwiler, R. P. and Hall, C. A. S. (1988) Tropical forests and the global carbon cycle.
Science, 239, 42-47.
FAO (1991) Production Yearbook 1990, Food and Agriculture Organization of the
United Nations, Rome.
FAO (1993) Forest Resources Assessment 1990. Tropical Countries. FAO Forestry
Paper No. 112. Food and Agriculture Organization of the United Nations, Rome.
Francisco, F., Tirasawat, P., Grainger, A. and Villanueva, E. (1996) Quantitative
analysis of factors controlling deforestation in the Philippines and Thailand, in
Controlling Deforestation in the Philippines and Thailand, Report to UK
Economic and Social Research Council, pp. 75-88.
Geertz, C. (1963) Agricultural Involution: The Process of Ecological Change in
Indonesia, University of California Press, Berkeley.
GOTIFAO (1972) General Land Capability Map of Thailand, Government of
ThailandIFood and Agriculture Organization, Bangkok.
Grainger, A. (1983) Improving the monitoring of deforestation in the humid tropics,
in Tropical Rain Forest - Ecology and Management, (eds S. L. Sutton, T. C.
Whitmore and A. C. Chadwick), Blackwell Scientific, Oxford, pp. 387-395.
Grainger, A. (1984) Quantifying changes in forest cover in the humid tropics: over-
coming current limitations. Journal of World Forest Resource Management, 1,
3-63.
Grainger, A. (1986) The future role of the tropical rain forests in the world forest
economy, DPhil thesis, Department of Plant Sciences, University of Oxford.
Grainger, A. (1987) TROPFORM: a model of future tropical hardwood supplies, in
Proceedings of CINTRAFOR Symposium on Forest Sector and Trade Models,
University of Washington, Seattle, November 3-4 1987, pp. 205-212.
Grainger, A. (1990a) Modelling deforestation in the humid tropics, in Deforestation
or Development in the Third World?, vol. III, (eds M. Palo and G. Mery), Bulletin
No. 349, Finnish Forest Research Institute, Helsinki, pp. 51-67.
Grainger, A. (1990b) Modelling future carbon emissions from deforestation in the
humid tropics, in Proceedings of the Intergovernmental Panel on Climate Change
Conference on Tropical Forestry Response, Options to Global Climate Change,
Sao Paulo, Brazil, 9-12 January 1990, Report No. 20P-2003, Office of Policy
Analysis, US Environmental Protection Agency, Washington, DC, pp. 105-19.
Grainger, A. (1993a) Controlling Tropical Deforestation, Earthscan Publications,
London.
Grainger, A. (1993b) Rates of deforestation in the humid tropics: estimates and mea-
surements. Geographical Journal, 159, 33-44.
Grainger, A. (1995a) National land use morphology: patterns and possibilities.
Geography, 20, 235-45.
Grainger, A. (1995b) The forest transition: an alternative approach. Area, 27,
242-5l.
Grainger, A. (1996a) The degradation of tropical rain forest in Southeast Asia: tax-
onomy and appraisal, in Land Degradation in the Tropics, (eds M. E. Eden and J.
T. Parry), Mansell, London, pp. 61-75.
Grainger, A. (1996b) An evaluation of FAO's Tropical Forest Resource Assessment
1990. Geographical Journal, 162, 73-79.
Grainger, A., Malayang, B. S. III, Francisco, H. A, Domingo, L. J., Mehl, C. B. and
Tirasawat, P. (1996) Controlling Deforestation in the Philippines and Thailand,
Report to UK Economic and Social Research Council.
Grigg, D. (1980) Population, Growth and Agrarian Change. Cambridge University
Press, Cambridge.
Grigg, D. (1982) The Dynamics of Agricultural Change, Hutchinson, London.
Hall, P. G. (ed.) (1966) Von Thunen's Isolated State, Pergamon, London.
Hall, C. A. S., Uhlig, J., Qui, Y., Tian, H., Falero, E. and Brown, S. (1994) A GIS-
based model for predicting the spatial pattern of tropical land use change and its
application to Peninsular Malaysia and northern Thailand.
Hebden, J. (1981) Statistics for Economists, Philip Allan, Oxford.
Hicks, J. (1969) A Theory of Economic History, Oxford University Press, Oxford.
Higgins, G. M. et al. (1983) Potential Population Supporting Capacities of Lands in
the Developing World. Technical Report of Project FPAlINT/513.
FAOIUNFPAlIIASA, Rome.
Hirsch, P. (1993) Political Economy of Environment in Thailand. Journal of
Contemporary Asia Publishers, Manila.
Jeffers, J. N. R. (1975) An Introduction to Systems Analysis, with Ecological
Applications, Edward Arnold, London.
References 343
Kahn, J. and McDonald, J. (1994) International debt and deforestation, in The
Causes of Tropical Deforestation, (eds K. Brown and D. W. Pearce), UCL Press,
London, pp. 57-67.
Katila, M. (1992) Modelling Deforestation in Thailand. Report prepared for the
Thai Forestry Sector Master Plan, Forest Products Marketing Subteam, July 1992.
Royal Forest Department, Bangkok.
Kummer, D. M. (1992) Deforestation in the Postwar Philippines, Ateneo de Manila
University Press, Manila.
Lambin, E. F (1994) Modelling Deforestation Processes. A Review, TREES Series B,
Research Report No.1, Institute for Remote Sensing Applications, European
Commission Joint Research Centre, Ispra, Italy.
Lanly, J. P (ed.) (1981) Tropical Forest Resources Assessment Project (GEMS):
Tropical Africa, Tropical Asia, Tropical America, UN Food and Agriculture
Organization/UN Environment Programme, Rome.
Lanly, J. P. (1982) Tropical Forest Resources, FAO Forestry Paper No. 30, UN Food
and Agriculture Organization, Rome.
Liu, D. S., Iverson, L. R. and Brown, S. (1993) Rates and patterns of deforestation in
the Philippines: application of geographic information systems analysis. Forest
Ecology and Management, 57,1-16.
Ludeke, A. K., Maggio, R. C. and Reid, L. M. (1990) An analysis of anthropogenic
deforestation using logistic regression and GIS. Journal of Environmental
Mahar, D. and Schneider, R. (1994) Incentives for tropical deforestation: some
examples from Latin America, in The Causes of Tropical Deforestation, (eds K.
Brown and D. W. Pearce), UCL Press, London, pp. 159-17l.
Mather, A. S. (1992) The forest transition. Area, 24,367-79.
Meyer, W. B. and Turner II, B. L. (1992) Human population growth and global land-
use/land cover change. Annual Reviews in Ecology and Systematics, 23, 39-6l.
Miller, L. D., Nualchawee, K. and Tom, C. (1978) Analysis of the dynamics of shift-
ing cultivation in the tropical forest of northern Thailand using landscape model-
ling and classification of LANDSAT imagery, in Proceedings of the 20th
International Symposium on Remote Sensing of Environment, ERIM, Ann Arbor,
Michigan, pp. 1167-1185.
Ness, G. D. and Ando, H. (1984) The Land is Shrinking, Johns Hopkins University
Press, Baltimore.
Notestein, F. (1945) Population - the long view, in Food for the World, (ed. T. W.
Schultz), University of Chicago Press, Chicago.
Nualchawee, K., Miller L., Tom, c., Christenson, J. and Williams, D. (1981) Spatial
Inventory and Modelling of Shifting Cultivation and Forest Land Cover of
Northern Thailand with Inputs from Maps, Airphotos and Landsat, Remote
Sensing Centre, Technical Report No. 4177, Texas A&M University, College
Station.
Palo, M., Mery, G. and Salmi, J. (1987) Deforestation in the tropics: pilot scenarios
based on quantitative analyses, in Deforestation or Development in the Third
World?, vol. III, (eds M. Palo and G. Mery), Bulletin No. 349, Finnish Forest
Research Institute, Helsinki, pp. 53-106.
Osgood, D. (1994) Government failure and deforestation in Indonesia, in The
Causes of Tropical Deforestation, (eds K. Brown and D. W. Pearce), UCL Press,
London,pp.217-225.
Panayotou, T. and Panasuk, C. (1990) Land and Forest: Projecting Demand and
Managing Encroachment, Research Report No.1, TDRI 1990 Year-End
Conference, Thailand Development Research Institute Foundation, Bangkok.
Panayotou, T. and Sungsuwan, S. (1989) An Econometric Study of the Causes of
Tropical Deforestation: the Case of Northeast Thailand, Discussion Paper No.
284, Harvard Institute of International Development, Cambridge, Massachusetts.
Pearce, D. W. (ed.) (1993) Blueprint 3. Measuring Sustainable Development,
Earthscan Publications, London.
Persson, R. (1974) World Forest Resources, Research Notes No. 17, Department of
Forest Survey, Royal College of Forestry, Stockholm.
Pontius, G. R. and Hall, C. A. S. (1994) Modelling tropical land use change: detailed
methods and application to Sub-Saharan Africa.
Richardson, G. P. and Pugh, A. L. (1981) Introduction to System Dynamics
Modeling with DYNAMO, MIT Press, Cambridge, Massachusetts.
Romm, J. (1991) Exploring institutional options for global forest management, in
Proceedings of the Technical Workshop to Explore Options for Global Forestry
Management, Bangkok, 1991, (eds D. Howlett and C. Sargent), International
Institute for Environment and Development, London, pp. 186-92.
Rotmans, J. and Swart, R. J. (1991) Modelling tropical deforestation and its conse-
quences for global climate. Ecological Modelling, 58, 217-247.
Rudel, T. K. (1989) Population, development, and tropical deforestation: a cross-
national study. Rural Sociology, 54, 327-338.
Sader, S. A. and Joyce, A. T. (1988) Deforestation rates and trends in Costa Rica.
Sanchez, P. A. (1976) Properties and Management of Soils in the Tropics, Wiley,
New York.
Shafik, N. (1994) Macroeconomic causes of deforestation: barking up the wrong
tree?, in The Causes of Tropical Deforestation, (eds K. Brown and D. W. Pearce),
UCL Press, London, pp. 86-95.
Simpson, E. S. (1987) The Developing World. An Introduction, Longman, London.
Southgate, D. (1990) The causes of land degradation along 'spontaneously' expand-
ing agricultural frontiers in the Third World. Land Economics, 66, 93-10l.
Taylor, P. J. (1989) Political Geography, 2nd edn, Longman, London.
Tongpan, S., Panayotou, P., Jetanavich, S., Faichampa, K. and Mehl, C. (1990)
Deforestation and Poverty: Can Commercial and Social Forestry Break the
Vicious Circle? Research Report No.2. The 1990 TDRI Year-End Conference on
Industrializing Thailand and its Impact on the Environment. Thailand
Development Research Institute, Bangkok.
Turner, B. L. II, Kasperson, R. E., Meyer, W. B. et at. (1990) Two types of global
environmental change. Global Environmental Change, 1, 14-22.
Westoby, J. (1989) Introduction to World Forestry, Blackwell, Oxford.
Williams, M. (1989) Deforestation: past and present. Progress In Human
Geography, 13, 176-208.
World Bank (1982) World Development Report 1982, Oxford University Press,
Washington, DC.
Young, A. (1989) Agroforestry for Soil Conservation, CAB InternationalJICRAF,
Wallingford, UK.
-12------
Communicating the message: a case
study from the Royal Botanic Gardens,
Kew
Colin Clubbe
12.1 INTRODUCTION
The Earth is thought to be 4600 million years old. If we condense this
to an understandable time-scale, the Earth is a 46-year-old person.
This person is a late developer and nothing is known about its first 7
years and only sketchy information exists for the next 35 years. Earth
began to flower at the age of 42. Dinosaurs and the great reptiles did
not appear until a year ago, when this planet reached the age of 45.
Mammals arrived only 8 months ago. In the middle of last week,
human-like apes evolved into ape-like humans, and at the weekend the
last ice age enveloped the Earth. Modern humans have been around
for 4 hours and discovered agriculture during the last hour. The indus-
trial revolution began just a minute ago. During those 60 seconds of
biological time, humans have made a rubbish tip of Paradise. We have
caused the extinction of many hundreds of species of animals and
plants, and have ransacked the planet for fuel. Now we stand like
brutish infants gloating over this meteoric rise to ascendancy, poised
on the brink of the final mass extinction and of effectively destroying
this oasis of life in the solar system (Greenpeace membership leaflet).
The analogy of Earth as a middle-aged person of 46 has been adopted as a
mantra for the destruction of our environment and now appears in many
publications. Our common future, if it is to be healthy and peaceful,
depends on our acting together to protect, preserve and improve our envi-
ronment. This action is required now, and must be taken at the highest lev-
els, by governments and decision makers, but with the full support of and
pressure from all people. We all need to press for faster and more effective
measures to protect the environment, and help bring about a new course of
346 Communicating the message
sustainable development through actions in our daily life. It is important to
keep global environmental issues in mind when making decisions and taking
action at the local level. Friends of the Earth have captured this idea in the
phrase, 'Think Globally, Act Locally', and this should be our guiding light.
This chapter reviews the ways in which conservation messages can be
communicated, illustrated by examples from the range of activities cur-
rently in operation as part of the education and conservation programmes
offered by the Royal Botanic Gardens, Kew. These are related to broader
issues and complementary activities in related institutions.
12.2 WHAT IS THE MESSAGE AND WHY SHOULD WE BOTHER?

The species-richness of tropical forests is unrivalled. Covering only 6% of
the Earth's land surface they harbour at least 50%, and possibly as much
as 90% of all of the Earth's species (Myers, 1989; Elliot, 1991; Goldsmith,
this volume). In addition to providing a habitat for so many of the world's
species, tropical forests are of tremendous importance to humankind for
the wide range of goods and services that they provide. These positive val-
ues and benefits of tropical forests are described in detail in other chapters
of this book, and the wider issues of biodiversity loss have recently been
reviewed by Club be (1996). Despite these numerous measurable benefits,
approximately 10% of the world's tropical forests were destroyed during
the 1980s, and only 1 % replanted. This represents a loss of 154 million ha
over that decade, an area the size of France, Germany, Italy and the UK put
together (Heywood and Watson, 1995). It is for good reason that many
describe this as 'The Global Chainsaw Massacre'.
So we have a dilemma on our hands. On the one hand tropical forests
are of inestimable value, whilst on the other they are being destroyed at an
unprecedented rate. What is the future for these forests and what can we
do to try and influence this? How do we communicate this message effec-
tively? Our goal should not only be to raise awareness, but to convert this
awareness to action. For some this may be campaigning (see Juniper, this
volume). For others it may be changing their purchasing policy on goods
and materials derived from tropical forest, so exercising their consumer
power. The destruction of tropical forests is symptomatic of the wider envi-
ronmental destruction that is under way. Similarly, communicating the
message about tropical forest destruction is an element of the wider need
for environmental education across a much broader front.
For the biodiversity world, a seminal event took place in June 1992 in
Rio de Janeiro. The United Nations Conference on Environment and
Development (UNCED) brought together more than 140 national delega-
tions and 100 heads of state of UN member countries, together with
around 3000 support staff and 1000 people from the media, for an Earth
Summit. UNCED was by far the biggest conference ever convened in the
Types of audience 347
history of international relations and the first time that biodiversity, as an
issue, had been comprehensively addressed. It was described at the time as
'The Biggest Show on Earth' concerning the future of our planet. Much
was achieved at the Earth Summit, including a solemn Rio Declaration,
Agenda 21 (a blueprint for sustainable development), two conventions, on
climate change and biodiversity, and a set of principles for managing
forests (Grubb et aI., 1993). The real challenges are still being tackled as
we try to implement the Rio decisions at a national level, on the basis of
close international collaboration.
UNCED produced a new vocabulary which was transmitted across the
world by the huge media presence in Rio. Biodiversity, the condensation of
biological diversity first coined by Wilson and Peter (1988), came into pop-
ular usage. Sustainable development, first coined at the Stockholm Summit
20 years earlier and in the Brundtland Report (WCED, 1987) was a strong
focus of debate and is the central theme of Agenda 21. The Brundtland
Report defined sustainable development as 'development which meets the
needs of the present without compromising the ability of future genera-
tions to meet their own needs'. In a world reduced to the sound bite and in
capitalist societies with a 'me-generation' attitude, the challenge is how we
persuade people to look beyond their own noses and to think about future
generations.
Ideas of sustainable development are emerging in people's consciousness
with the appearance of Local Agenda 21s (LA21). The Earth Summit set a
1996 target date for local authorities to .consult their population and
achieve a consensus on an LA21 for the community which is about
improving the quality of life for individuals without reducing or damaging
the quality of life and environment of others. Action across the London
boroughs is well advanced. Every borough has an LA21 contact and most
have begun their consultation process (Anon., 1997).
LA21s can provide a mechanism for converting awareness to action, so
long as the consultation process is wide enough and people take responsi-
bility for their own actions. We still need to articulate more clearly the idea
of the environment and development being two sides of the same coin.
12.3 TYPES OF AUDIENCE

'If you are thinking one year ahead, plant rice
If you are thinking ten years ahead, plant trees
If you are thinking one hundred years ahead, educate the people.'
(Chinese proverb)
Education is an important component of all of the major outputs of the
Earth Summit, even if it didn't get much attention at the time. Educational
issues are clearly addressed in Articles 12 (Research and Training) and 13
(Public Education and Awareness) of the Convention on Biological
Diversity (Glowka et al., 1994). The word 'education' appears in Agenda
21 more than 600 times, second only to the word 'government', and
Agenda 21 clearly identifies education, public awareness and training as a
vital cross-sectoral issue (Boulton and Knight, 1996).
Chapter 36 of Agenda 21 states that:
'Both formal and non-formal education are indispensable to changing
people's attitudes so that they have the capacity to assess and address
their sustainable development concerns. It is also critical for achieving
environmental and ethical awareness, values and attitudes, skills and
behaviour consistent with sustainable development and for effective pub-
lic participation in decision making'
(UNCED, 1992).
Within the formal education sector, three groups are focused on the Royal
Botanic Gardens, Kew: school children, particularly primary school
groups; undergraduate and postgraduate groups studying botanical and
conservation-related disciplines; and professionals from overseas who are
seeking specific training as part of a capacity-building programme. Non-
formal education is aimed at the general public.
12.3.1 Formal education
Schools education
Around 800 million young people attend school world-wide, providing a
vital opportunity for developing environmental awareness (Boulton and
Knight, 1996). In England environmental education is one of five official
cross-curricular themes. The knowledge and attitudes about the environ-
ment that pupils should gain from environmental education as a cross-cur-
ricular subject within the National Curriculum are fully documented
(NCC, 1990). Whether environmental education should be a cross-curricu-
lar theme, or whether it should be taught as a separate subject has been
debated for some time. However, with a very prescribed and full National
Curriculum there is always going to be strong competition for discrete sub-
ject areas.
A cross-curricular approach has been adopted by the Royal Botanic
Gardens, Kew in an effort to raise the profile of another under-represented
subject area in the school curriculum: plants, and botany in general. The
development of a new teacher's pack Plants across the Curriculum has pro-
vided a much-needed resource for teachers and students who want to use
plants and botanic gardens as a focus for curriculum development (Figure
12.1). The pack, developed for work at Key Stages 2 and 3 of the National
Curriculum for England and Wales, provides information, pictorial resources,
Figure 12.1 The Kew teacher's pack Plants across the Curriculum. © The Trustees
of the Royal Botanic Gardens, Kew.
ideas, activities and worksheets. There are eight subject wallets: art, eng-
lish, geography, history, mathematics, music, science and technology, which
can be used independently or within a topic or cross-curricular theme. By
working with teachers directly and providing resources such as Plants
across the Curriculum a wide audience can be reached. Approximately
70 000 children visit Kew annually. Large numbers of children visit other
botanic gardens in the UK with good tropical collections such as
Edinburgh, Oxford, Chelsea and Birmingham. All have thriving education
programmes which highlight rain forests as a component of their pro-
grammes.
Another focus for schools activity is the Government-sponsored Science,
Engineering and Technology (SET) week. Staged in mid-March each year
since 1994 it gives research institutions and others the opportunity to com-
municate their own message across all levels of the formal education sector.
For SET97, the Royal Botanic Gardens, Kew teamed up with the
Rainforest Foundation's Roadshow to provide a series of activities for
school children in the Princess of Wales Conservatory at Kew (Figure
12.2). Creating the feel of a traditional tropical forest community within
the Conservatory allows children to get a feel for forest living. A range of
activities which highlight the many ways that cultures indigenous to tropi-
cal forests use plants formed the focus of the programme. This is supported
by a grant from the Committee on the Public Understanding of Science
(COPUS).
Figure 12.2 Science, Engineering and Technology Week (SET97), activities in the
Princess of Wales Conservatory, Royal Botanic Gardens, Kew. © The Trustees of
the Royal Botanic Gardens, Kew.
Outreach at Kew is developed using a vehicle to bring information about

plants to schools who are unable to get to the Gardens. The vehicle is a
fully mobile classroom, towed by a Land Rover, which is packed with plant
specimens and artefacts allowing the outreach officer, plus school teacher
where appropriate, to conduct a full series of workshops on the economic
importance of plants and conservation issues, based around the tropical
rain forest and other themes. The vehicle is usually established in each site
for a week and the outreach officer conducts a series of classes for local
schools who have pre-booked these sessions. The outreach vehicle can also
be taken to county or borough shows.
Tertiary education
Biology has always been one of the most popular science subjects studied
at undergraduate level. Science applications to university, particularly for
biological sciences and physics, have recently seen a significant upturn
(Utley, 1997). The University and Colleges Admissions Handbook lists
more than 100 biology/biological sciences and over 100 environmental sci-
ences courses available at undergraduate level (UCAS, 1997). Many of
these have conservation-based and tropical ecology type modules.
Undergraduate groups are frequent visitors to the Royal Botanic Gardens,
Kew, making use of the tropical collections as substitute tropical biomes.
Undergraduate groups form a significant proportion of the approximately
80 overseas university expeditions that are organized annually, the major-
ity being to tropical ecosystems. The Royal Geographical Society has rec-
ognized the importance of these activities by organizing a series of seminars
on expedition planning, including obtaining the right permits to work in-
country, and they publish an Expedition Planners' Handbook (Winser and
McWilliam, 1992). More recently the Royal Geographical Society's
Expeditions Advisory Group have started to run Biodiversity Assessment
Workshops to ensure that data collected on these expeditions can be used
by the wider science community. Experts in the major taxa (mammals,
birds, plants, etc.) are on hand to advise. Staff from the Natural History
Museum and Kew provide the main botanical input.
At a postgraduate level, there has been an increase in the number of mas-
ter's programmes with a significant conservation theme since the MSc in
Conservation at University College London (UCL) was established in
1960. The prospects web site maintained by the University of Manchester
<http://www.prospects.csu.man.ac.uk> lists all of the postgraduate courses
offered in the UK along with information on funding. It currently lists
seven master's courses with a strong conservation component. Botanic gar-
dens and zoos can provide valuable resources for these courses, both in
terms of their collections and for access to current conservation projects
and initiatives on the world stage (see Prance, this volume). The Royal
Botanic Gardens, Kew provides a number of taught modules for a variety
of these courses, including UCL's tropical forest module. This chapter is
based on one of these modules. In addition they provide excellent project
opportunities for the dissertation component of MSc courses, and Kew
regularly has 10-15 MSc students carrying out significant elements of
their projects at Kew during the summer research period. Despite the
sustained interest in conservation, funding for these MSc courses is still a
major problem. Lack of funds could spell disaster for this valuable
source of trained personnel and for the next generation of committed
conservationists.
Professional training
Article 12 is one of the cornerstones of the Convention on Biological
Diversity as it addresses human capacity-building (scientific and technical
training), research and international co-operation to apply biodiversity-
related research. It charges the Contracting Parties (149 countries as of
August, 1996; <http://www.unep.chlbio/ratifica.htmb), taking into
account the special needs of developing countries, to:
'Establish and maintain programmes for scientific and technical educa-
tion and training in measures for the identification, conservation and
sustainable use of biological diversity and its components and provide
support for such education and training for the specific needs of develop-
ing countries'
(Glowka et ai., 1994).
The Royal Botanic Gardens, Kew and other international research organi-
zations committed to the goals of the Convention on Biological Diversity
have met the requirements of Article 12 for many years through a variety
of informal attachments or internships, as well as more formal postgradu-
ate and professional courses. The Royal Botanic Gardens, Kew now offers
a suite of professional training courses in its core activity areas, as part of
its capacity building programme. These are mostly 8-week summer schools
aimed principally at people from the tropical developing world, the areas
of Kew's primary research focus. Individual courses are usually offered in
alternate years so that two, or sometimes three courses are run in anyone
year. Courses are offered in Herbarium Techniques, started in 1987;
Botanic Garden Management and Plant Conservation Techniques, both
started in 1993; Botanic Gardens Education, started in 1995; and
Cultivation and Conservation of Threatened Plant Species, started in 1996.
New courses are being explored as demands arise, and a new training
course is currently being developed on the Convention on Trade in
Endangered Species of Wild Fauna and Flora (CITES), which came into
force in 1975, and to which 140 states were party as of October, 1996;
<http://www.unep.ch.cites/parties.htmb. Since 1990, 139 trainees from 66
countries have participated in this programme.
Details of these courses are available on the Kew Web site:
<http://www.rbgkew.org.uk>. Several of these courses are supported by
funds from the UK Government's Darwin Initiative for the Survival of
Species. A number of other organizations run similar capacity-building
programmes in their own fields of expertise, such as the World
Conservation Monitoring Centre with its major focus on data manage-
ment: <http://www.wcmc.org.uk>.
12.3.2 Non-formal education

There are many informal educational programmes currently in operation,
aimed at raising awareness of conservation issues at all levels of society
from children to the public at large. Botanic gardens and zoos, as major ex
situ conservation facilities, are ideally placed to take an active role in this
process by explaining the reasons that species are threatened in the wild
and the value of ex situ collections. The potential audience is very large.
Approximately 1 million people visit Kew Gardens each year. Botanic
Gardens Conservation International estimate there to be approximately
1600 botanic gardens world-wide attracting 150 million visitors annually
(Heywood and Watson, 1995). There are approximately 1000 zoos world-
wide which attract over 600 million visitors annually, more than 10% of
the world's population (Worley, 1996). This places particular challenges on
interpretation, and lessons are being learned from the museums' world as
to how best to put these messages across. The job for the zoos is somewhat
easier in that animals are more contained and there are obvious places for
labels, which are relatively easy to update. The Jersey Wildlife Preservation
Trust has developed a particularly good approach where the visiting public
are kept fully informed of the threats to the animals' habitats, how the cap-
tive breeding programme is being managed, and when animals are to be re-
introduced to the wild ('going home'). This is all done via unobtrusive
labels at the corner of a cage or enclosure. The problem faced by botanic
gardens is to ensure that the labelling doesn't overshadow the plants them-
selves. This is more easily done in themed, habitat plantings such as the
Princess of Wales Conservatory at Kew, which provides 10 different habi-
tat zones, each of which has an explanatory board, e.g. wet tropics, dry
tropics, cloud forest. Again a standard format has now been adopted, a
large colour photograph of the habitat or plant, together with some
explanatory text which includes details on the habitats, the major plants
and the threats faced by that habitat (Figure 12.3). Topic, rather than habi-
tat, themes can provide very powerful messages. For example the Garden
of World Medicine at Chelsea Physic Garden is an ethnobotanical garden
exhibiting plants used by a wide variety of indigenous peoples including
North American Indians, Maoris and Aboriginals (Minter, 1993).
Special events can provide a forum for focused, co-ordinated activities.
World Environment Day, established under the auspices of the United
Nations Environment Programme (UNEP) in 1987, is an annual celebra-
tion held every June 5th. It enables the global community to reflect on the
progress made in the defence of the environment and to celebrate and rec-
ognize acts of courage and determination in conserving the environment,
and provides an opportunity to renew the call for change in our attitude
and actions towards the environment. The theme of World Environment
Day, 1995 was 'We the People, United for the Global Environment' (Figure
12.4), reflecting the need for solidarity by the global community in its fight
against poverty and other social ills, and for the protection of the environ-
ment (Dowdeswell, 1995).
At a local level Kew's Orchid Festival has become an annual event to
highlight this fascinating group of plants. It is one of the largest families of
flowering plants with an estimated 20 000 species which is being added to
at the rate of about 100 species per year as more orchids are discovered
and described. Orchids have been a focus of research activities at Kew for
many years. Many species threatened in the wild are grown at Kew, such as
Figure 12.3 A sample interpretation panel (wet tropics) in the Princess of Wales
Conservatory, Royal Botanic Gardens, Kew, which contains 10 different habitat
zones. © The Trustees of the Royal Botanic Gardens, Kew.
Phragmipedium besseae, a beautiful red slipper orchid, known only from

one population of a few individuals in Ecuador. The family is a focus of
systematic research as well as conservation (Anon., 1996). A whole series
of events including special exhibitions, study days on growing orchids,
schools activities and public lectures is organized during the 6-week festi-
val. It also provides an opportunity to highlight the threats to orchids in
the wild, including habitat loss from tropical forest clearance and over-col-
lecting of wild individuals. From the marketing point of view, visitor num-
bers are up in what is otherwise a quiet time of the year, February/March,
and the increased revenue can be ploughed back into the scientific, educa-
tional or horticultural work of the organization. Significantly, the Orchid
Festival draws visitors from groups of people who might not normally visit
Kew, so provides an opportunity to spread our message to a wider audi-
ence. Like most big organizations in the late 20th century, the Royal
Botanic Gardens, Kew has developed a mission statement to guide its activ-
ities and to communicate it a sense of purpose. The mission of the Royal
Botanic Gardens, Kew is:
Resources available 355
\\ 'II.! 111\If~lllI1ll'UlI),1\
I lIlt'"
Figure 12.4 The logo of World Environment Day, 1995.
'To enable better management of the Earth's environment by increasing

knowledge and understanding of the plant kingdom - the basis of life on
Earth.'
12.4 RESOURCES AVAILABLE

When Marshall McLuhan published The Medium is the Massage he put
forward a compelling argument that society has always been 'shaped more
by the nature of the media by which men communicate than by the content
of the communication' (McLuhan and Fiore, 1967). It was an exposition
of McLuhan's prophecies of the electronic age and was not a treatise but a
happening, not a message but a massage. McLuhan's predicted electronic
revolution is certainly with us and the pace of change in the communica-
tions field is still very rapid. However, the accessibility of the Internet and
World Wide Web, satellite television and CD-ROM has not yet replaced
more traditional means of communication. Rather these act to widen the
choice of methods of communication available to the educator. If
McLuhan's ideas are still relevant today, then the challenge for today's
communicators is to choose the right form of massage to ensure that the
message gets through.
Printed media
The printed word can still be regarded as the most permanent, easily acces-
sible record. Book sales are buoyant, newspaper sales holding. Printed
leaflets are a favoured means of getting a message across cheaply and effi-
ciently, however it is very important that these are distributed to a targeted
audience for maximum effectiveness.
Since the publication in 1962 of Rachel Carson's Silent Spring, which
alerted the public to the insidious and widespread effects of pollution, we
have seen the power of popular books to generate concern. Norman Myers
was probably the first to expose the wholesale destruction of tropical
forests and bring it to the wider public's attention in The Primary Source,
first published in 1983 and recently updated (1990). Edward Wilson's
(1992) classic The Diversity of Life and David Attenborough's marvellous
trilogy The Living Planet (1984), Life on Earth (1989) and The Private
Life of Plants (1995), all best sellers around the world, have served to
make people marvel at the variety of life on our planet, whilst also raising
awareness of the threats it is facing through unsustainable practices and
exploitation.
Articles in newspapers and magazines have immediate impact. Headline
grabbers such as those below convey a startling message, but how long
does this last?
• What on Earth Are We Doing? (Time, January 2, 1989, when Time took
the unusual step of naming Endangered Earth as Planet of the Year in
lieu of the usual Man or Woman of the Year)
• Burning of the Amazon Forest is Worst in Years (International Herald
Tribune, October 13, 1995)
• The Rape of an Island Paradise (The Observer, September 29, 1996 on
logging in the Solomon Islands)
• Billion Dollar Drugs are Disappearing in the Forest (New Scientist, July
8, 1995, on potential drugs undiscovered before rain forests are
destroyed)
• Mahogany Is Murder, Don't Buy It (Friends of the Earth, Mahogany
Campaign Leaflet, December, 1994)
How often do these need reinforcing and at what level do they make an
impact? We need to try and answer these questions to ensure maximum effect.
The Royal Botanic Gardens, Kew produces a series of Information Sheets
in order to communicate information about its own activities, plant biol-
ogy in general and important conservation issues relating to plants and the
role of the Gardens. These are provided for educational groups studying at
the Gardens and are available to the general public on request. Themed
collections are sold through Kew's book shop. The Rainforest Pack
includes such titles as Description of Rainforests; Destruction of
Rainforests; Economic Potential of Rainforests; CITES and the Royal
Botanic Gardens, Kew; and Role of the Living Orchid Collection.
Posters can provide a very potent message. Kew Gardens has been por-
trayed on London Transport Posters displayed on platforms along the rail-
way lines leading to Kew since 1908. Collections of these posters have been
exhibited at Kew and the collection has recently been published (Riddell
and Stearn, 1994). However, it is difficult to quantify the specific role that
these have contributed to the increased visitor numbers to the Gardens.
The University of Oxford Botanic Garden has recently produced an innov-
ative series of educational posters which highlight the importance of plants
and the role of botanic gardens in plant conservation. Cleverly titled Sex,
Drugs and Botanical Fulfilment, the first one illustrates how plants have
provided the active chemicals for condoms, the pill, aphrodisiacs, drugs
used to fight cancer and more. However, we do need to consider how
posters can be used most effectively.
Leaflets are the currency of many organizations and groups trying to sell
something, whether it be an idea or a commodity. Untargeted, they run the
risk of being filed with the rest of the junk mail that flows through the sys-
tem, in the dustbin, or preferably the paper recycler. However, membership
applications for organizations such as WWF-Worldwide Fund for Nature,
Friends of the Earth and Greenpeace, which are regular inclusions in the
Sunday magazines, undoubtedly do yield some positive results.
Newsletters are a favourite way of keeping members of a particular
organization or group informed of activities and are a regular feature of
any membership package. The more successful ones are also available for
sale to non-members at selected outlets. Members of Friends of the Earth
receive Earth Matters, which celebrated 25 years with a special anniversary
issue in summer 1996. Members of Fauna and Flora International receive
Fauna and Flora News, and Friends of the Royal Botanic Gardens, Kew
receive Kew magazine. Many organizations use a newsletter as a semi-for-
mal record of their activities and send it out to interested parties such as
the Smithsonian's Man and the Biosphere Program's Biodiversity News,
and the UK Tropical Forest Forum News.
Companies with a strong environmental ethic, such as The Body Shop,
produce a variety of leaflets highlighting their activities and their commit-
ment to the environment in an effort to influence purchasing habits. Trade
Not Aid is a business ethic which has been developed by a number of orga-
nizations including The Body Shop and the charity Oxfam. The basis of
this is the operation of fair trading agreements with communities in devel-
oping countries where more of the income generated goes directly to the
communities involved in production. The coffee 'Cafe Direct' has recently
made the transition from the specialist outlet (Oxfam) to the general super-
market shelves and sales are rising.
Visual media
Visual images have tremendous power to move, to enthral or to revolt. It is
no surprise that improvements in technology over the last 40 years have
been paralleled by the degree with which the natural world has been
brought into our living room. In the UK this has almost been synonymous
with one name, that of David Attenborough. From his Zoo Quest films in
the early 1960s to his ground-breaking trilogy The Living Planet, Life on
Earth and The Secret Life of Plants, Attenborough and his film crews have
redefined television natural history. The programmes have attracted huge
audiences world-wide and the videos of these programmes are used in for-
mal education and viewed for enjoyment throughout the world. The BBC
Natural History Film Unit and Oxford Scientific Films have made an art of
filming the new, unique and 'unfilmable' secrets of the natural world. These
act as the shop window for conservation. However, they are so profes-
sional that we now take them for granted and there is a danger that we
may become immune to the conservation message, the accompanying com-
ment about destruction of forests and extinction of species. Under-cover
programmes to highlight threats and illegal practices can provide powerful
messages and can result in legal action being taken. Members of organiza-
tions such as the Environmental Investigation Agency and TRAFFIC risk
their lives to obtain film footage exposing illegal trade in animals, clearing
of tropical forests and exploitation of indigenous communities.
Photographic exhibitions such as the Wildlife Photographer of the Year,
an annual competition organized by the BBC Wildlife Magazine and spon-
sored by British Gas, add to the variety of ways that conservation messages
can be conveyed. Winning and commended photographs are exhibited
each year at the Natural History Museum. The introduction of categories
such as species in trade and species threatened in the wild have provided
some very chilling images of the abuse of nature and natural resources.
The 1990s have brought about a revolution in electronics. The explosion
in the uptake and use of the World Wide Web provides us with a brand
new arena for communication. In the first few years of availability the
number of Web sites exploded (Table 12.1), and all the evidence points to a
continued increase.
Conservation organizations have been quick to put their message on the
Web. Some of those with a strong focus on tropical forest conservation
include:
• Friends of the Earth at <http://www.foe.co.uk>
• Rainforest Alliance at <http://www.rainforest-alliance.org>, particularly
the chilling 'digital clock' on its home page showing the amount of rain
forest lost since January 1996, which increases by the second as you
watch it. Population increase is also shown in real time, as is the amount
of rain forest conserved, a much slower-moving figure (Figure 12.5)
• Rainforest Action Network at <http://www.ran.org>
• Missouri Botanic Garden at <http://www.mobot.org>, particularly its
schools pages
• World Conservation Monitoring Centre at <http://www.wcmc.org.uk>,
with a particularly comprehensive listing of related sites
• Smithsonian Natural History Museum at <http://www.nmnh.si.edu>
• WWF-Worldwide Fund for Nature at <http://www.panda.org>
Table 12.1 Web site figures
Date Number of active web sites Percentage

available on internet commercial sites
June, 1993 130 1.5

December, 1993 623 4.6
June, 1994 2738 13.5
December, 1994 10022 18.3
June, 1995 23500 31.3
January 1996 100000 50.0
June, 1996 230000 (estimate) 68.0
January, 1997 650000 (estimate) 62.6
(Source: Matthew Gray of the Massachusetts Institute of Technology)
<http://www.mit.edu:8001lpeoplelmkgray/netlweb-growth-summary.html>
• Natural Environment Research Council at <http://www.nerc.ac.uk>

• Natural History Museum at <http://www.nhm.ac.uk>
• Greenpeace International at <http://www.greenpeace.org>
• United Nations Environment Programme at <http://www.unep.ch>
• Royal Botanic Gardens, Kew at <http://www.rbgkew.org.ub, with links
to a wide range of other botanical and horticultural servers
Many of these have excellent links to other relevant sites making them par-
ticularly useful starting points for more extensive searches.
As more schools are hooked up to the Internet, the World Wide Web will
become an increasing important source of information, particularly as
more databases and library catalogues become live and are searchable via
the Web. More private homes are subscribing to one of the many on-line
service providers which are competing for customers. The World Wide
Web is set to become an ever more powerful means of communication for
the future with increasing computing power to transmit videos and sound.
Aural media
Despite the rise to supremacy of the television and the computer, the spo-
ken word can still deliver a powerful message whether via radio, the public
lecture, or the personal story quietly exchanged between individuals or
small groups.
BBC Radio 4's Natural History Programme and Costing the Earth both
cover topical issues with a conservation theme and have carried features on
tropical forests in recent times. Costing the Earth also deals with develop-
ment issues and can be very hard-hitting when dealing with these difficult
questions. Occasional documentaries are produced which highlight signifi-
cant events such as the murder of the Brazilian rubber tapper Chico
~~ Acr~ LO)I SlncQ 1/96: ',·1. ·1', /. (",h ~l. ~r~ W~ Con~Qor;~d: H .'-IHII. 'm •
'Ii!- World Populaloon: ',.11 j~.", I .fI. I
Figure 12.5 Home page of the Rainforest Alliance (April 1997), <http://www.rain-
forest -alliance.org>.
Mendes. Scientists have recently become regular guests on Melvyn Bragg's

weekly Radio 4 programme, Start the Week.
Public lectures have always been a favoured way of communicating new
ideas or revisiting old ones. Whether in a learned institution or an open
public forum, words spoken with feeling and commitment can enthuse. An
inspirational lecture can change the listener for life. Mark Plotkin, the
American ethnobotanist and writer, vividly explains in the opening of his
marvellous book, Tales of a Shaman's Apprentice, the moment he decided
how to spend the rest of his life. It was whilst studying at Harvard, during
an evening lecture by Professor Richard Schultes, the world's pre-eminent
authority on ethnobotany (Plotkin, 1993). Plotkin has devoted his career
to the study of tropical rain forests of South America. He founded the
Shaman's Apprentice programme, a collaboration between shamans and
ethnobotanists to pass on traditional plant knowledge, itself very much an
oral tradition.
Stories about plants are fascinating and stories of their demise are tragic.
Educators at all levels at Kew have built up a wealth of plant stories that
Is the message getting across? 361
communicate a message to school children, postgraduate students and the
general public; it's all in the telling. In the Temperate House at Kew there is
a single specimen of the cafe marron tree (Ramosmania rodriguesii).
Related to the true coffees, this specimen is a cutting from the only individ-
ual left in the wild, on the Indian Ocean island of Rodrigues (Verdcourt,
1996). This individual tree flowers, but due to an incompatibility problem
cannot set seed. Although it can be propagated from cuttings, these are all
genetically identical and so there is no hope for this species. It is on the
verge of extinction, one of the 'living dead' according to many. Also in the
Temperate House, close to Ramosmania, is another tree, Sophora toromiro
from Easter Island, known to be extinct in the wild. The last-known indi-
vidual on Easter Island survived on the inner slopes of Rano Kau crater
until 1960. However, a world-wide search has located several other indi-
viduals of this species in botanic gardens around the world, in addition to
those held at Kew. Representatives from those botanic gardens together
with representatives from Easter Island have formed a Toromiro
Management Group which is developing a strategy to re-introduce S.
toromiro back to Easter Island (Maunder, 1994). Prance (1997) lists 22
plant species thought to be extinct in the wild which are held in British
botanic gardens. There are many other stories to tell (see Prance, this vol-
ume).
The Environmental Lecture Series at the Royal Botanic Gardens, Kew is
an annual lecture which provides a forum for raising topical environmental
issues. The first in the series, The Rainforest Lecture was given by Prince
Charles in 1990 (HRH The Prince of Wales, 1990). Sir Maurice Strong,
Secretary General of UNCED, gave his views one year on from the Earth
Summit (Strong, 1993), and David McDowell gave the 1996 lecture
shortly after becoming Director General of the World Conservation Union-
IUCN (McDowell, 1996).
12.5 IS THE MESSAGE GETTING ACROSS?

There is no shortage of information about the environmental crisis and in
particular the threats faced by the tropical forest biome. What is in doubt
is whether we are getting this message across well enough. How can we
evaluate whether the message is getting through? Are people more aware,
and if they are, has this raised awareness resulted in changed attitudes and
actions? There is a danger of information overload and replacement func-
tion, the feeling that because we read about something we equate that with
having done something about it. The message becomes the action rather
causing a reaction. It is perhaps analogous to the tendency for today's stu-
dents to photocopy references, but not get around to reading them: the
feeling created is that possession of the photocopy equals the reading of the
reference.
It's not just the charismatic megafauna that can draw a crowd: the
flower that enthralled a nation
The titan arum (Amorphophallus titanum) produces the largest inflores-
cence in the world, rising up to 3 m from the floor of the moist tropical
forests of Sumatra. Its habitat is under threat from tree clearance. We have
no real idea of how rare it is in the wild, but it is under threat. The inflores-
cence is cut down by rubber tappers who think, incorrectly, that it attracts
malarial mosquitoes. The swollen underground stems, which can weigh up
to 50 kg, are dug up for the Japanese who use them as a starchy food.
When the titan arum flowered at Kew in the summer of 1996 it caused
a sensation (Figure 12.6). The event in itself was amazing enough - the
species last flowered at Kew in 1963 - but what was even more remark-
Figure 12.6 The titan arum, Amorphophallus titanum. © The Trustees of the
Royal Botanic Gardens, Kew.
Is the message getting across? 363
able was the way that the event captured the public imagination. 49 000
visitors entered the Gardens between Tuesday 30 July and Sunday 4
August, 21 000 more than anticipated. Twenty eight film crews, 14 from
the UK and 14 from overseas, beamed live and recorded footage around
the world, to Europe, Australia, Japan, North America, South Africa and
Brazil. The story was covered in 15 national and 112 local newspaper in
the UK as well as many papers throughout the world. Thirty-one radio
interviews were conducted for world-wide broadcast. Amorphophallus
certainly was a media megastar, a role seldom played by a plant, and
more commonly associated with a large cuddly, furry, mammal.
The Indonesians call the plant 'the corpse flower' because of the nau-
seating smell that the plant produces to attract its pollinators, thought
to be sweat bees. Whether it was the potential smell, described as a mix-
ture of rotten fish and burnt sugar, that attracted the crowds or whether
it was a genuine fascination with a very unusual flower, is difficult to
ascertain. What is certain is that the press leader writers lost no time in
coining the puns:
• Lily the stink loses its 33-year reputation by a nose (The Times, front
page, August 1, 1996)
• It's Phew Gardens (Daily Star, August 1, 1996)
• Scents Assail Sensibilities (Daily Telegraph, August 1, 1996)
K£w GARD[N~ WARNlN(;' I - - - - . / _ v

1~1~ PLANT CAUS~~ .
ANAgsoWTf GTINK !
Figure 12.7 Reproduced with permission from The Times, 31 July 1996.
• Sitting Pretty But Not Heaven Scent (Glasgow Herald, August 1,

1996)
• Giant Cannibal Plant Gets Up British Noses (Wall Street Journal, 25
July, 1996)
Arguably the clearest sign that our publicity was working was when the
titan arum appeared in a political cartoon. The Times of 31 July carried
a memorable cartoon showing Chancellor Kenneth Clarke emerging
from a titan arum bloom, with a gaggle of Eurosceptics looking on,
holding their noses against the stink Clark was making on European
monetary union (Figure 12.7). The challenge for the Kew press office is
to top that one.
Recent visitor surveys at Kew have resulted in an increasing number of

people answering that they thought Kew was a scientific institution
involved in conservation activities. Previously a majority had thought of
Kew as a nice park with interesting plants. However, in the wider world
the clearance of tropical rain forests continues unabated. The economic
pressures are still there. The Observer's 'Forests for Life' supplement (The
Observer, 29 September 1996) made for very depressing reading. As David
Bellamy said in the conclusion of his introduction: 'Again I ask if you are
sitting comfortably? If the answer is yes then you are condoning a pretty
bleak future for your family. If the answer is no, read on, then get off your
complacent backside and join WWF and all the other NGOs that are
working to protect our forests and so help save the world'. Evidence sug-
gests that people are taking this advice. One in 10 people belong to an
environmental organization (quote from David Bellamy at the Tourism for
Tomorrow Awards, 27 February, 1997). WWF-Worldwide Fund for
Nature recognize 4.7 million supporters and Greenpeace about 5 million.
People are beginning to ask searching questions of retailers. We want to
know the origin of the wood, especially if it is a tropical hardwood. Does it
carry the Forest Stewardship Council's (FSC) seal of approval signifying the
wood comes from a well-managed, sustainable forest? As more and more
influential western retailers switch to FSC-certified wood imports, timber-
exporting countries that fail to join this ecological club will bec.ome mar-
ginalized (Upton and Bass, 1995; Barbier, this volume). However, we need
to keep up the pressure from the consumer end, with more accessible infor-
mation and by exercising our buying power. In a market economy busi-
nesses respond quickly to falling profits. The UK is still the world's second
largest importer of Brazilian mahogany. However, since 1985 when Friends
of the Earth launched its 'Mahogany is Murder' campaign, UK imports of
rain forest timber have fallen by over 50% (Friends of the Earth web site
information: <http:www.foe.co.uklpubsinfo/infosystlmahog-murder.htmb).
References 365
12.6 CONCLUSIONS
Although conservation and biodiversity are in the political arena, the global
environment does not seem to be very high on the domestic political agenda
in the UK. The 'green policies' of the main political parties seem more like
vote-winning tools than real efforts to make world changes. However, with
the implementation of Local Agenda 21s and increased consumer choice for
goods and services derived from sustainable sources, we may be able to stem
the tide of non-sustainable practices and give future generations some hope.
The conservation of the Earth's biological resources can be achieved only by
a well informed and educated global community who are able to place the
conservation of biological resources into a political, social and economic
context at local, national and global levels. As individuals we must all be able
to relate our own needs and desires for a fulfilling life to the environmental
constraints placed on the expression of this lifestyle. This will involve some
hard decisions, but it is only when these decisions are taken that true envi-
ronmentally sustainable development can be achieved. Education and com-
munication are activities central to achieving this. Using both formal and
informal education, and by involving everyone, only then can we achieve our
goals and there will be hope for the future of Planet Earth. Our goal must be
'Education for Sustainability', the theme of the recent Botanic Gardens
Conservation International Education Congress (BGCI, 1996).
REFERENCES
Anon. (1996) Celebration of orchids. Kew Magazine, Spring, 1996.
Anon. (1997) The Green Guide for London, The Green Guide Company Ltd,
London. {{http://www.cerbernet.co. uk/greenguide/:
Attenborough, D. (1984) The Living Planet, BBC Books, London.
Attenborough, D. (1989) Life on Earth, BBC Books, London.
Attenborough, D. (1995) The Private Life of Plants, BBC Books, London.
BGCI (1996) Education Congress Success, BGC! News, 2, 16.
Boulton, M. N. and Knight, D. (1996) Conservation education, in Conservation
Biology (ed.I. F. Spellerberg), Longman, Harlow, UK.
Carson, R. (1962) Silent Spring, Houghton Mifflin, Boston.
Clubbe, C. (1996) Threats to biodiversity, in Global Environmental Problems, (eds R.
Blackmore and A. Reddish), Hodder and Stoughton/Open University,
London/Milton Keynes.
Dowdeswell, E. (1995) Editorial. Our Planet, 7, 2.
Elliot, C. (1991) Tropical Forest Conservation, WWF Position Paper 7, Gland, WWF
International.
Glowka, L., et at. (1994) A Guide to the Convention on Biological Diversity, mCN,
Gland and Cambridge.
Grubb, M., Koch, M., Thomson, K., Munson, A. and Sullivan, F. (1993) The 'Earth
Summit' Agreements: A Guide and Assessment, Earthscan, London.
Handley, M. and Crowcroft,]. (1995) The World Wide Web: Beneath the Surf, UCL
Press, London
HRH The Prince of Wales (1990) The Rainforest Lecture, Royal Botanic Gardens,
Kew.
Heywood, V. H. and Watson, R. T. (eds) (1995) Global Biodiversity Assessment,
Cambridge University Press, Cambridge.
Maunder, M. (1994) Report of Meeting of the Sophora toromiro Management
Group, Royal Botanic Gardens, Kew (mimeo.).
McDowell, D. (1996) Conservation at the Cutting Edge, Royal Botanic Gardens,
Kew.
McLuhan, M. and Fiore, Q. (1967) The Medium is the Massage, Penguin,
Harmondsworth, UK.
Minter, S. (1993) The Herb and Medicinal Gardens at the Chelsea Physic Garden,
Chelsea Physic Garden, London.
Myers, N. (1989) The future of forests, in The Fragile Environment (eds L. Friday
and R. Laskey), Cambridge University Press, Cambridge.
Myers, N. (1990) The Primary Source: Tropical Forests and our Future, 2nd edn, W.
W. Norton, London.
NCC (1990) Curriculum Guidance 7: Environmental Education, National
Curriculum Council, York, UK.
Prance, G. T. (1997) The conservation of botanical diversity, in Plant Genetic
Conservation: the in situ Approach, (eds N. Maxted, B. V. Ford-Lloyd and J. G.
Hawkes), Chapman & Hall, London.
Plotkin, M. J. (1993) Tales of a Shaman's Apprentice, Viking, New York.
Riddell, J. and Stearn, W. T. (1994) By Underground to Kew, Studio Vista, London.
Strong, M. F. (1993) Beyond Rio: New World Order or Lost Opportunity? Royal
Botanic Gardens, Kew.
UCAS (1997) UCAS Handbook 1997, Universities and Colleges Admissions Service,
Cheltenham, UK.
UNCED (1992) The Global Partnership for Environment and Development: A guide
to Agenda 21, United Nations Conference on Environment and Development,
Geneva.
Upton, C. and Bass, S. (1995) The Forest Certification Handbook, Earthscan,
London.
Utley, A. (1997) Students return to science. Times Higher Education Supplement,
1268, February 21, p. 1.
Verdcourt, B. (1996) Ramosmania rodriguesii. Curtis's Botanical Magazine, 13,
204-209.
WCED (1987) Our Common Future, World Commission on Environment and
Development/Oxford University Press, Oxford.
Wilson, E. O. (1992) The Diversity of Life, Allen Lane/Penguin, London.
Wilson, E. 0 and Peter, F. M. (eds) (1988) BioDiversity, National Academic Press,
Washington, DC.
Winser, S. and McWilliam, N. (eds) (1992) Expedition Planners' Handbook and
Directory, 1993-94, Expedition Advisory Centre, Royal Geographical Society,
London.
Worley, D. (1996) Ex situ conservation, in Conservation Biology (ed. I. F.
Spellerberg), Longman, Harlow, UK.
-13------
Effective campaigning
A. Juniper
Friends of the Earth, 26-28 Underwood Street, London Nt 7JQ, UK
This chapter charts the development of campaigning to save the rain

forests. It identifies which issues have been prioritized by the campaigners
and what tactics they have used, and identifies where they have been suc-
cessful. It concludes with proposals for campaign priorities for the future.
Due to space constraints it is inevitably limited in scope and detail.
13.1 THE CAMPAIGNERS

Environmental campaigning organizations operate at the interface between
mainstream politics and ecology. Large international organizations such as
Friends of the Earth, the World Wide Fund for Nature (WWF) and more
specific networks such as the World Rainforest Movement (WRM), the
Rainforest Action Network (RAN), Rainforest Information Centre (RIC)
and the Rainforest Foundation, have sought to attack the causes of tropical
deforestation through campaigning activity. Ultimately, the campaigners'
aim is to influence mainstream policy-making, consumer behaviour and the
attitudes and activities of companies so that the pressures on the rain
forests are reduced.
The campaigning organizations use a wide variety of tactical approaches
to achieve change. Consumers are encouraged not to use destructively pro-
duced tropical hardwoods, so that producers move towards sustainable
forestry. Rain forest communities have gained support in their resistance
against logging companies; exposure in the media of destructive activities
has been designed to encourage companies to adopt best practices; whilst
research projects have revealed, for example, how some overseas aid
intended to slow deforestation has actually accelerated it. These various
tactical activities have been combined in strategic campaigns aimed at radi-
cally altering behaviour and attitudes at all levels towards rain forests.
These and other tactics have placed rain forests on the political agenda,
raised awareness and mobilized citizens' action.
Tropical Rain Forest: A Wider Perspective. Edited by F. B. Goldsmith

Published in 1998 by Chapman & Hall, London. ISBN HB 0 412 81510 9 and PB 0 412 815206
368 Effective campaigning
The campaigners have been able to draw on a range of political and
technical strengths. Very importantly, the campaigning organizations are
from all parts of the world. The World Rainforest Movement and the
Friends of the Earth International networks, for instance, combine the
information-gathering potential and expertise of campaigners in dozens of
countries from both the developing south and industrialized north.
Political intelligence and lobbying skills have enabled the campaigners to
exploit their networks and various manifestations of political legitimacy to
influence decision-making fora that would be inaccessible to anyone orga-
nization or group of organizations. Experience and creative use of the
media through thorough and carefully targeted research have expanded
public understanding and placed pressure where it has been most needed.
The campaigners' understanding of the issues behind deforestation and
familiarity with the ecological debate has led to widespread support in the
scientific community as well as among the general public.
13.2 THE CAMPAIGN - ORIGINS AND IMPACTS

Initially, the rain forest campaign agenda arose against a backdrop of
growing disquiet expressed during the 1970s by leading figures in the sci-
entific and tropical forestry communities over the worsening overall condi-
tion of the tropical forests. Official figures on tropical deforestation rates
became available in the early 1980s (see for example FAO, 1981) generat-
ing a growing realization that serious ecological and economic conse-
quences would accompany continuing and accelerating deforestation. In
1985, Tropical Forests: a Call for Action was published by the World
Resources Institute (WRI), the World Bank and the United Nations
Development Programme (UNDP) (WRI, 1985), heralding a period of
intense international interest in tropical forests.
Before this time, the principal environmental and conservation organiza-
tions had identified habitat loss as the main force behind rapidly accelerat-
ing rates of species extinction, but had left forestry issues (at least those in
the tropics) to the foresters and development specialists. However, as envi-
ronmental organizations began to look more carefully at the issues behind
deforestation, it became clear that a campaigning approach was necessary
to stimulate the required action.
A growing body of ecological literature on the dynamics of rain forest
ecosystems had revealed their enormous biological diversity and therefore
the extent to which they represented a priority focus for conservation
action. Estimates of species richness were combined with accumulating
information on levels of endemism and details on the rate and distribution
of deforestation, ultimately to expose a biological crisis of global propor-
tions (see for example Myers, 1985).
The campaigning organizations went public with their first rain forest
initiatives during the mid 1980s. After initial internal uncertainty, Friends
The campaign - origins and impacts 369
Figure 13.1 A growing body of evidence on the ecological implications of tropical

rain forest destruction has spurred campaigners to press for political change.
Source: Dylan GarcialFriends of the Earth.
of the Earth launched its first campaign in Britain in 1985. WWF-UK fol-
lowed soon after. At this time and in the years to follow, other global envi-
ronmental questions were racing up the political, media and scientific
agendas. Ozone depletion, global climate change and the explosion of one
of the Chernobyl nuclear reactors in the Ukraine were commanding front-
page headline news on an almost daily basis (Lamb, 1996). The environ-
ment had become a matter for mainstream public and political life, and the
media soon seized on the images of destruction brought home from the
tropics as one of the most serious and pressing green issues. The statistics
on deforestation rates fuelled media interest and enormous amounts of
print and broadcast space were devoted to the issue. The effect of this was
exactly as the campaigners had hoped, at least in Europe and North
America: the generation of intense public concern leading to pressure on
politicians to take action. The political reaction to media and growing
interest in the new scientific evidence and ecological predictions was two-
fold.
Firstly, as public concern in developed nations became manifest, the
diversion and increase of overseas aid for tropical forestry characterized
the principal official response in many industrialized nations and interna-
tional institutions such as the World Bank. This approach is typified in the
former UK Prime Minister Margaret Thatcher's famous address to the UN
General Assembly in November 1989 in which she committed an extra
£100 million of UK taxpayers' money to tropical forest activities. Overall,
during the latter part of the 1980s and early 1990s, several billion dollars
worth of aid was earmarked for tropical forests. Whether this injection of
finance helped to slow deforestation is a matter of continuing debate, as
discussed below.
The other main political reaction arising from the initial burst of public-
ity and public interest was from the tropical developing countries them-
selves. These nations had recently become the focus of media attention and
pressure from campaigners demanding action to stop the destruction, and
were forced to develop their own strategies to improve their international
standing. In some cases the reaction was positive with, for example, a
series of tax and land reforms in Brazil during the late 1980s and early
1990s that were intended to reduce the incentives to clear forest. In other
cases, the measures taken were cosmetic, with rhetorical commitments to
sustainable forest use not matched by controls on the often illegal activities
of timber companies. For some countries, especially Malaysia, the interna-
tional attention became a source of frustration, especially as many of the
industrialized countries had witnessed the loss of virtually all their own
natural forests several hundred years earlier. These perceived double stan-
dards would be a matter for debate during the negotiation of a global for-
est accord at the Earth Summit several years later (see below).
There was also public mobilization in defence of the rain forests.
Consumers placed pressure on retailers of wood and wood products to
stock only material from well-managed sources, whilst firms engaged in
destructive development projects in forested areas became the subject of
consumer boycotts and negative media exposure.
The successes of the campaigners during this initial period of mobiliza-
tion were impressive. Politicians of all colours from all parts of the world
committed themselves to saving the rain forests; companies withdrew from
damaging project proposals and established codes of conduct and sustain-
able development targets; international development institutions
announced new environmental safeguards; new international rain forest
conservation initiatives were launched; and indigenous peoples in many
tropical countries gained a new political confidence and dialogue with gov-
ernments.
Many of these gains were achieved through campaigns targeted at differ-
ent aspects of the pressures behind tropical deforestation. Broadly speak-
ing, the campaigning organizations between them have taken on four main
clusters of issues and their work has sought to:
• render the tropical timber industry sustainable;
• influence other private-sector interests (in the mmmg sector, for
example) such that their activities cause minimum damage;
The timber trade - importing deforestation 371
Figure 13.2 Media coverage is vital for effective campaigns. Demonstrations that
present images of interest to newspaper and television editors are more likely to get
their message across to the public. Source: David TowendlFriends of the Earth.
• influence the policies and activities of international development and

financial institutions so that their programmes promote sustainable for-
est use;
• campaign for effective international agreements and accords for wise
tropical forest use and conservation.
The following sections briefly explain how and why the campaigners focus
on these issues. This analysis provides the background to the third and
final part of this chapter which includes thoughts on the future develop-
ment of rain forest campaigning.
13.3 THE TIMBER TRADE - IMPORTING DEFORESTATION

The extraction of tropical timbers from natural and semi- natural forests
has been identified as a principal force behind deforestation. Some 45 000
(Given, 1990) to 50000 km2 (WRI, 1985) of tropical moist forest is
affected annually by logging, mostly with serious negative consequences for
the forests. For example, in a major study for the International Tropical
Timber Organization (lITO), it was stated that 'it is not yet possible to
demonstrate conclusively that any natural tropical forest anywhere has been
successfully managed for the sustainable production of timber' (Poore,
1989). The study adds that 'when a primary forest is first logged ... quality
and volume will probably not be matched in future cuts ... unless the forest
is closed to further exploitation for a century or more'. Many professional
foresters and World Bank advisers have agreed with this view or at least
expressed a precautionary approach (see for example Lamprecht, 1989;
Talbot, 1990). There was in addition a substantial and growing body of
information pointing to massive illegal operations, flouting of forestry codes
and several political forces driving forestry policy that were major blockages
to sustainable management (e.g. Rice, 1991; Callister, 1992; Mombiot,
1993).
Beyond the broad professional consensus that developed around the
unsustainability of tropical timber exploitation, it became clear that the
industry was a powerful precursor of other pressures, especially in provid-
ing access for landless farmers to areas of otherwise inaccessible forest
where they then engaged in agricultural activities that often led to total for-
est loss. The pivotal role of the timber industry and the global trade con-
nections embedded in it made the industry a priority focus for the
campaigners.
Friends of the Earth's initial attack against the industry sought to popu-
larize the issue and bring it to the public and media agenda. In 1985, a list
of retailers and products was published identifying which tropical woods
were being sold, where and by whom. Local campaign groups picketed
shops and used stickers bearing slogans such as 'a monkey lost its nuts for
this product' to warn would-be purchasers of the damage caused by the
demand for tropical timber. Simultaneously, timber trade bodies were lob-
bied to accept a code of conduct and in the process were placed very much
on the defensive.
The campaign continued with various actions, photo stunts and distribu-
tion of leaflets and posters. The media were very interested and covered the
campaign well. The reaction of industry was mixed. Some companies
claimed the industry has a beneficial impact on the tropical forests because
the developing countries exporting the timber could (because of the rev-
enue earned) afford to conserve them rather than having to clear them for
agriculture. Other industry and retail players saw which way consumer
preference was heading and ceased to trade in tropical woods, or commit-
ted themselves to ensure that their products were derived from sustainable
sources. The majority reacted somewhere in between, accepting that some-
thing needed to be done but that it should be based on voluntary rather
than official regulation.
The campaigners followed the technical and economic debates very
closely and publicized information that aided their cause. For instance, a
technical report carried out with support from the UK Overseas
Development Administration (ODA) found that most of the profits from
tropical timber exports go to the rich countries and that only a small per-
centage remains with the tropical forest country government itself (Barbier
Figure 13.3 The public is increasingly bombarded with complex messages. If cam-
paigners are to penetrate the popular consciousness and alter public opinion, sim-
ple and interesting messages and images are needed. Source: Friends of the Earth.
et al., 1992). Research and publicity of this kind enables campaigners to

cut through the inaccuracies that lay behind the exhortations and rhetoric
of the industry, to good effect.
Similar campaigns grew up, with varying degrees of success, in several
European countries including Germany, the Netherlands and Switzerland.
A new campaign group also sprung up in Japan, the country using more
tropical timber than any other. The Japan Tropical Forest Action Network
was (and is) small with sparse funding and lacks the mass support enjoyed
by some groups in the west, but lent tremendous legitimacy to organiza-
tions working to influence the Japanese Government and corporations
from outside. Similarly, the work of Friends of the Earth campaigners in
Sarawak, Malaysia allowed the campaign to be internationalized, further
increasing the political stakes, and the likelihood of meaningful policy
change.
Simultaneous with calls for industry to take action to reduce the impacts
of the tropical hardwood trade were campaigns urging government action,
principally through trade restrictions. However, against the backdrop of
the negotiation of the Uruguay Round of the General Agreement on Tariffs
and Trade (GAIT) and other moves toward trade liberalization (such as
the Treaty of Maastricht in the European Union), trade restrictions were
simply not politically possible. Countries that did try to impose restrictions
(for example Austria) were quickly subject to international pressure
(Malaysia announced a boycott of Austrian chocolate, for instance) and
were forced to reverse their plans. Industrialized-country governments
reacted to campaign pressure with additional funding for tropical forest
aid projects and by seeking agreement with exporting countries as to how
sustainable use could be achieved. In this latter respect, the institution with
a leading responsibility was (and is) the International Tropical Timber
Organization (lITO).
lITO was set up in 1986 with a Governing Council of representatives
from the principal countries exporting and importing tropical timbers.
Essentially lITO oversees a commodity agreement (the International
Tropical Timber Agreement, IITA) which seeks to 'expand and diversify'
the trade in tropical timbers, but also to 'encourage the development of
national policies aimed at sustainable utilisation and conservation of tropi-
cal forests and their genetic resources'. The role of the ITTO (especially in
respect of its mandate to promote sustainable utilization) has undoubtedly
been enhanced through the activities of campaigners.
The principal lITO initiative that is relevant in this regard is the so
called Target 2000 which informally commits its member governments to
ensure, by the year 2000, that all tropical timber in international trade will
be from sustainable sources. However, no government has said how it
intends to achieve this target and, as with various guidelines for sustainable
timber production produced by lITO, it seems unlikely that any govern-
ment will meet it. Counsell (1993) provides a detailed critique of ITTO and
its failure to implement its own agreements.
The campaigners, aware that the lITO would prove to be an effectual
cover for inaction by industry and lack of regulation by government,
sought to expose its failings. In the UK alone, Lynda Chalker MP, Minister
for Overseas Development, received over 22 000 postcards from members
of the public calling for reforms of the lITO, whilst detailed reports
(Counsell, 1993) and media exposure on lITO's activities limited the
extent to which the organization could be used as a cover for the timber
industry.
This last point in many respects sums up the development of the cam-
paigns against the mining of hardwood trees from tropical rain forests.
Industry and governments have acknowledged the existence of a serious
problem and (whilst there has been dispute over the actual impacts of log-
ging and its relationships with other pressures) timber companies and offi-
cial bodies have pledged, both informally and through agreements like the
lITO, to achieve sustainable utilization. These agreements and commit-
ments have been insufficient to reduce the impacts on the forests, and cam-
paigners have been at pains to expose this.
However, once industry has acknowledged the problem and it and gov-
ernments have pledged action, there is little for the media to report. As the
issues have become further complicated, the more limited media profile on
the issue leads to reduced public interest and diminishing political pressure.
This is exactly what has happened in recent years as the news outlets have
found it more and more difficult to find novel angles on the tropical timber
issue.
Despite diminishing media coverage, campaigners have not been idle. A
strong body of consumer preference evidently remains for sustainably pro-
duced timber and a great deal of effort has been devoted to finding ways of
harnessing this for conservation benefit. In 1996, a new scheme was
launched with this aim. The Forest Stewardship Council (FSC) scheme is
the most reliable independent certification measure yet conceived and
could, in years to come, prove an important means for promoting sustain-
able forest use.
The FSC essentially seeks to provide reliable consumer information and
support for sustainable wood-production operations. The FSC evaluates,
accredits and monitors well-managed forestry operations against a set of
criteria and principles in order to identify for consumers which wood (and
eventually paper) has been derived from sources that are deemed to be well
managed. The FSC's work basically involves FSC-approved certifying bod-
ies visiting forestry operations that might meet the standards set out by the
FSC. If such an operation is approved, wood from it can be marked with
the FSC logo so that consumers can select it over wood of unknown (and
probably unsustainable) origin. The FSC has a large membership of non-
governmental and forest industry representatives and enjoys increasing
official support. The FSC could provide a vehicle for pursuing many of the
aims of the forest campaigners and is set to be very influential in harnessing
market forces to promote positive forestry operations in the coming years.
In parallel with moves towards reliable wood certification, other initia-
tives have taken forward a sustainable forestry agenda with industry. For
example, WWF-UK has for several years worked with a large sector of the
wood-using industry in the UK to develop voluntary commitments aimed
at encouraging more sustainable forestry practices. The WWF's 1995 Plus
Group comprises over 70 companies and includes several of the UK's lead-
ing retail chains and large manufacturers. Membership of the group
requires that they take steps to purchase forest products solely from
Figure 13.4 'Mahogany is murder!' Directly confronting key decision makers with
a strong campaign message can help to advance debate and bring about change.
Influencing leading companies can assist in changing the policies of less forward-
thinking companies. Source: Jonathan Rose/Friends of the Earth.
well-managed forests. The companies establish internal databases to moni-

tor sources of forest products and are committed to using independent cer-
tification and the framework provided by the FSC.
Perhaps indicative of the campaigners' success in putting sustainable
management of forests into the mainstream of public life is a brochure pro-
duced by Sainsbury's, one of the largest retailing chains in the UK. Their
publication Sustaining the Forests (Sainsbury's, 1996) explains the firm's
support for the FSC and makes several statements on the company's com-
mitment to sustainable forest use, including: 'We want our customers to
know that when they buy products containing wood or paper from
Sainsbury's, they are not contributing to the loss of the world's forests'.
Whilst there will inevitably be questions over the specific environmental
impacts arising from individual products, the fact that a company as large
as this is willing to make such a bold commitment and openly to support
an independent wood certification scheme is a strong indication of progress
in the direction the campaigners intended. That Sainsbury's might be seri-
ous about sustainable management is further suggested in the fact that six
of the firm's major wood suppliers were axed in 1996 for not complying
with FSC criteria - progress indeed.
The private sector in the rain forests 377
13.4 THE PRIVATE SECTOR IN THE RAIN FORESTS
Campaigners advancing a sustainable development agenda in the formula-
tion of policies that affect rain forests have sought to influence the attitudes
of other private-sector players (outside the timber industry). Pressure
directed at a number of large companies over the last decade has led to
dramatic shifts in company rhetoric and to some extent in their actual pro-
grammes and operations.
A great deal of emphasis has been placed on the mining and oil and gas
sectors because of the potential for large-scale impacts arising from explo-
ration and production and related infrastructure development. One tropi-
cal forest country where especially strong campaigns have developed is
Ecuador. Here, local indigenous peoples' groups and other campaigners
have joined with organizations operating overseas (especially in the UK,
USA and Scandinavian countries) to challenge the activities of foreign firms
searching for and producing oil in the country's tribally-inhabited eastern
lowlands (Kimerling, 1991).
Companies including British Gas, Conoco and Texaco have been among
the larger multinational companies targeted in media and by public letter-
writing campaigns in their own countries and subject to pressure in tropi-
cal countries through, for instance, office occupations by tribal groups and
claims to recover compensation for damage to tribal lands (Rice, 1994).
The combination of local campaigners working with groups in the coun-
tries where companies have their global head offices can be powerful. Local
campaigners can provide up-to-the-minute and accurate information for
news releases, they can take legal action in the countries where the com-
pany is operating (if it fails to take proper account of environmental con-
trols, for instance) and can formulate demands to be used by campaigners
outside. Campaigners pressuring companies internationally can give politi-
cal cover to those 'on the ground' who may be subject to intimidation and
pressure to keep quiet, they can mobilize the company's customers and
other interested parties, and generate media coverage. This kind of rela-
tionship has developed in many campaigns against destructive development
in rain forest areas, including against dams, roads and mines as well as oil
and gas developments, and has proved very powerful and often successful.
For instance, both British Gas and Conoco finally withdrew from opera-
tions in the Amazonian lowlands of Ecuador. Although neither company
cited the environmentalists' and tribal peoples' campaign as the reason, the
mobilization of the public in the UK and USA along with opposition from
tribal people undoubtedly played a critical role in their analysis of business
prospects.
A wider reaction to the campaigners' activities has been the publication
of guidelines for oil companies exploring and producing in tropical rain
forest areas. Voluntary industry guidelines and corporate environmental
Figure 13.5 Indigenous communities have joined with campaign groups from the
developed countries to campaign against environmental damage caused to their
homelands by private-sector companies. Source: Simon CounselUFriends of the
Earth.
awareness have been correctly, but cautiously, welcomed by campaigners

as part of an overall process of rendering development in tropical forest
areas sustainable. However, no matter how green they appear on paper, all
large firms involved in large-scale developments in sensitive ecosystems will
inevitably be faced with potential financial benefits which, from a narrow
short-term financial perspective, outweigh the global good. A case in point
is the proposal by the giant mining firm RTZ-CRA to extract mineral
sands from beneath some of the few remaining fragments of eastern coastal
forests in the south-eastern corner of Madagascar.
RTZ is perhaps the most criticized company in the world in terms of its
impacts on the environment and human communities in the areas it oper-
ates (see e.g. Moody, 1991). However, it has responded with high-level
commitments to sustainable development and environmental protection
through environmental and social impact assessment. In the case of the
Madagascar mine, preliminary assessments by the company have revealed
that some 30 species of plant endemic to the coastal forests exist in the
mining area. Various mining scenarios have been sketched out that lead to
the removal of between one half and two thirds of the forest. It is predicted
that some of the endemic species will be rendered extinct as a result. Others
will undoubtedly be hastened to the same fate through the reduction
The private sector in the rain forests 379
and fragmentation of their habitat. The company points to the fact that the
local people are clearing the forest anyway and that money generated by
the mine can be used for the more effective conservation of what remains.
Friends of the Earth remains vigorously opposed to the mining and has tar-
geted RTZ shareholder meetings and the media in attempts to dissuade the
firm from pressing ahead with the project. Friends of the Earth maintains
that RTZ has not shown any overall benefit to Madagascar's economy, that
the forest cannot be properly restored, and that the international commu-
nity should mobilize to save the forests rather than accepting that they
have to be mined to have an 'economic' value.
The fact that RTZ continues to stand its ground behind environmental
impact assessments and rehabilitation packages in the face of impending
species extinctions directly caused by its proposed activities shows how
sophisticated 'corporate environmentalism' has become, and how resistant
some firms are prepared to be in the face of negative publicity, especially if
they have no high-street products for the public to boycott.
Campaigns against specific company operations have also had a wider,
less tangible but perhaps more important impact, as company directors
have become aware of the need to reflect sustainable development princi-
ples in corporate policy and practice. There is no doubt that some board-
room environmentalism can be rightly dismissed as 'greenwash', but the
fact that the environment (including rain forests) is on the agenda can
probably be traced to two things. The first is the negative exposure of com-
pany after company implicated in the perpetration of environmental dam-
age, and the second is the threat of regulation. Campaigners have sought to
limit the damage caused by private-sector operations in rain forests
through exposing individual companies and, using the individual case stud-
ies they have worked on, calling for some form of regulation. Fearful of
regulation and cognisant of the expense involved in the public relations
work needed to recover the company's position, the environmental impacts
of operations are taken on board.
Sustainable development still remains secondary to profitability and
competitiveness. This is understandable - if companies are not competitive
they will not survive. Whilst campaigners can have some bearing on com-
petitiveness through boycotts and other tactics, it is also necessary to
develop and express a broader analysis of how societies achieve sustainable
development. If markets are indeed set increasingly to determine company
behaviour, then the so called 'externalities' of environmental and social
impacts must be brought into the equation and reflected in the price of
products which are presently subsidized by environmental damage (for
instance deforestation). Whilst this debate is beyond the scope of this chap-
ter, it will be vital to show how the private sector's impacts on the rain
forests (and the environment in general) are to be resolved.
13.5 AID, DEVELOPMENT, THE INTERNATIONAL FINANCIAL
INSTITUTIONS AND THE RAIN FORESTS
By the time the world became aware of the tropical forest crisis, development
institutions were already major players in determining the direction of eco-
nomic policy in many tropical developing countries, and their influence was
growing. One of the principal reasons for the expanding role of these institu-
tions during the 1980s, especially the World Bank and the International
Monetary Fund (IMF), was because of the importance attached by the inter-
national community to resolving the so-called debt crisis.
The debt crisis started in the early 1980s, when successive developing
countries (chiefly because of interest rate increases) were unable to meet
repayments on their international loans. Hundreds of billions of dollars
worth of loans had been extended to the developing countries of the south
during the 1970s, with the stock of debt roughly equally shared between
official creditors such as the World Bank (especially in Africa) and com-
mercial banks such as Lloyds and Barclays (principally in Latin America).
Many developing countries simultaneously found themselves unable to
meet repayment schedules and experienced a sharp increase in their
reliance on overseas assistance, both for economic management pro-
grammes and for more practical purposes of infrastructure development.
Rather than arrange the cancellation of debts which were causing tremen-
dous environmental and social damage, the strategy of the international
development community was to increase the ability of the debtor nations
to pay. This has been mainly promoted through economic reforms that
encourage export-led growth and reduce domestic expenditure. Almost
invariably, increased exports mean greater exploitation of natural
resources such as minerals, soils, fisheries and forests. The campaigners
soon realized that the unfolding economic strategy for the developing
countries, linked to the debt crisis and the use of overseas aid, was (and is)
a major factor in the nexus of issues that lie behind forest loss (see e.g.
George, 1989; Counsell et al., 1992 for a fuller explanation of these
issues).
One of the first salvos from the non-governmental organizations (NGOs)
came from the Sierra Club in the USA. In 1986 it launched one of the first
high-profile attacks on the World Bank with its report Bankrolling Disaster
(Sierra Club, 1986). This detailed the World Bank's involvement with envi-
ronmentally destructive projects and set a campaign rolling that soon
involved a wide range of development and environmental organizations,
and even a visit to the World Bank by rain forest activist Chico Mendes
(Mendes, 1989), who was murdered soon after.
The campaign led swiftly to announcements by the World Bank of its
intention to restructure and introduce safeguards that would prevent a rep-
etition of the kinds of impacts outlined by the Sierra Club. For instance,
Aid, development, the financial institutions and the rain forests 381
Frind. 01 IIw F.anh
\X!hat arc the

High ~trt:ct banks
doing In
the rainforc"t\?
Figure 13.6 The engine of destruction. The massive burden of third world debt has
helped drive large-scale tropical deforestation. The public is drawn closer to the
issue of forest loss through the high street banks who provide financial services for
millions of people in the UK. Source: Friends of the Earth.
projects would be categorized and screened on the basis of their potential

environmental consequences, whilst the organization's 1991 forestry policy
committed the World Bank not to fund logging projects in primary rain
forests (World Bank, 1991).
One of the most important characteristics of the campaign to reduce the
World Bank's impact on the rain forests has been the simultaneous raising
of the Bank's profile in several of its main funding countries (such as the
UK and USA). Without strong political support from those nations (who
fund it through decisions taken by Parliament or Congress), its future
would not be secure. Caught in the headlights of media exposure and pub-
lic condemnation, it had no choice but to react decisively to factual infor-
mation with which it could not argue.
Similar campaign actions exposing projects with severe environmental
impacts have also influenced other aid and development bodies such the
UK's ODA and the African Development Bank. Virtually all aid agencies
are now explicitly committed to 'sustainable development' and most
employ environmental impact screening of some sort on their projects. As
ever, though, the extent to which commitments on paper shape actual pro-
jects is questionable. The recent exposure of World Bank support for
potentially damaging roads in forested parts of eastern Cameroon without
an environmental impact assessment provides a case in point (Daily
Telegraph, September 9th, 1996).
In terms of aid specifically aimed at tropical forests, the most ambitious
aid and development programme attempted to date is the Tropical Forestry
Action Plan (TFAP). This multi-billion-dollar sustainable development plan
for the tropical forests was conceived in the mid 1980s as a means of coor-
dinating the many and growing sources of funding for tropical forest pro-
jects. These funding sources include multilateral sources (such as the World
Figure 13.7 Making the global connections. Decisions made by aid agencies have
had a tremendous influence on the status of tropical forest resources. Source:
Friends of the Earth.
International agreements 383
Bank), bilateral (ODA or USAID) and agencies of the UN such as FAO and
United Nations Development Programme (UNDP).
Coordination of development assistance for forestry projects under the
TFAP process was to be achieved through the agreement of national fund-
ing priorities set out in National Forestry Action Plans (NFAPs). NFAPs
would be agreed through a process of meetings involving interested donor
agencies, government departments and, to a limited extent, NGOs. The
TFAP soon became a focus for campaigners who saw it as addressing the
wrong priorities through the wrong sorts of projects on the basis of a
flawed decision-making process. The World Rainforest Movement and oth-
ers published a report in 1990 called The Tropical Forestry Action Plan:
What Progress? which concluded that in several countries the implementa-
tion of NFAPs had actually increased the rate of deforestation (Colchester
and Lohmann, 1990).
This carefully researched critique, along with other material placed in
the public domain by the campaigners, led to an official review of the
TFAP that identified shortcomings in its management. A revamping process
was initiated and the campaigners sought to reform the TFAP such that it
became a vehicle to conserve forests. However, because of in-fighting
between the international agencies charged with coordinating the TFAP
and because of decreasing political support from its principal donors, most
projects were never funded and most national plans completed under the
TFAP process now gather dust on foresters' and development planners'
shelves. The TFAP now wallows in the doldrums with little enthusiasm
from the international development community to breathe new life into it.
13.6 INTERNATIONAL AGREEMENTS

The notion of an international convention on the conservation and sustain-
able use of forests was first mooted in the late 1980s. At this time, the
potential global impacts of other environmental threats such as global
warming, ozone depletion and species extinction had led to the opening of
the United Nations Conference on Environment and Development
(UNCED) which was to meet at an international conference in 1992 (the
Earth Summit) to conclude agreements on biodiversity, climate, sustainable
development and forests.
The preparatory meetings leading up to the Earth Summit soon exposed
the deep differences existing between nations on how best to reach interna-
tional consensus for sustainable forest use. For instance, some rich countries
(such as Sweden) wanted to negotiate a legally binding forest convention
emphasizing sustainable use, whilst some of the developing nations (such as
Malaysia) saw such a plan as a potential block to their economic develop-
ment. Some wanted to emphasize sustainable trade in forest products whilst
others demanded only free trade. With complications added by the parallel
UNCED debates on indigenous peoples, biodiversity, climate change and
the funding of sustainable development, it became clear early on that no
forest convention would be negotiated during UNCED.
The principal split between governments was on North-South lines, with
the developing nations of the South arguing that the industrialized nations
of the North were in no position to argue for rain forest conservation when
they had cut down their own forests centuries before to facilitate economic
development. The southern nations (led by Malaysia) insisted that the right
to development and sovereignty over countries' natural resources were fun-
damental principles that should underpin any global accord. They
demanded commitments for new and additional financial resources to fund
forest programmes and held out strongly against any restrictions on the
timber trade. A split between developed and developing countries,
although never really explicit, also emerged between those countries that
believed they could gain economically from a convention, and those that
felt they would lose out because of their lower forest management stan-
dards. In the end, the deep divisions between North and South prevented
any clear agreement. Instead, a rather weak set of 'forest principles' was
negotiated.
The forest principles, although the first-ever world-wide agreement on
what must be done to conserve and achieve the sustainable use of forests,
are very weak and vague, and commit participating nations to actually do
very little. The ineffectiveness and ambiguity of the forest principles docu-
ment, in addition to the fundamental North-South stand-off, can be attrib-
uted to the very widely differing perspectives on what needs to be done and
by whom to conserve forests, what are the legitimate issues for interna-
tional (rather than only national) debate, and the fact that forests link with
a great many other sustainable development issues, including biodiversity
conservation and combating global climate change.
The complexity of the sum of the political perspectives that lay behind
each country's position on forests at UNCED also left the NGOs with rela-
tively little influence. The broad-brush campaigning that had mobilized
public concern several years earlier is not appropriate for influencing the
detail of closely negotiated international agreements. This is partly because
the politicians involved in the negotiations (especially if they are held
behind closed doors) are not accountable. They can (and frequently do)
blame other countries, or the complexity of the issues, when decisions fail
to meet public expectations.
Whilst campaigners followed every stage of the forest debate in UNCED
with great interest, the strong agreement many of them wanted to see on
protected area networks, the roles and rights of forest peoples, and decisive
action to achieve equitable and sustainable use, were not there at the end.
Most governments openly admit that the forest principles agreed at the
Rio de Janeiro Earth Summit are unsatisfactory and in need of considerable
Rain forests - campaign priorities for the future 385
strengthening. This view has been expressed in the post-Earth Summit
process (convened under the auspices of the UN's Commission on
Sustainable Development) whereby countries will seek to develop a consen-
sus on what should be done next to halt global deforestation through the
report of the Intergovernmental Panel on Forests. This has looked at fund-
ing questions, whether a forest convention is appropriate, and the matter
of certification of forest products from sustainable sources. Given the evi-
dent differences that remain between countries over fundamental questions
such as the necessity for a convention, the Intergovernmental Panel on
Forests (IPF) is unlikely to deliver any new initiative that will provide the
action needed to save what is left of the rain forests. The process that fol-
lows the IPF (which submitted its final report to the UN in April 1997) will
undoubtedly be followed by many forest campaigners with great interest
(see below).
13.7 RAIN FORESTS - CAMPAIGN PRIORITIES FOR THE FUTURE

The rain forests continue to be cleared, and if the most recent FAO figures
are to be believed (FAO, 1993a), the rate of loss has accelerated consider-
ably during the 1980s, the period when the international community ral-
lied to the forests' defence. Whilst there have been impressive
achievements, they have clearly not yet been sufficient to turn the tide.
Whilst opinion surveys show continuing public concern over tropical
deforestation (see for example DoE, 1994), political commitment and
media interest in the issue have clearly faded since the time of the Earth
Summit. It is even possible that some of the policy gains and campaign vic-
tories of the past years are in jeopardy. This stagnation has occurred not
because the campaigners have been less active, but largely because the
media have failed to maintain their initial interest, apparently because
news editors see few new stories to tell. As a result, politicians feel no pub-
lic pressure and therefore no urgent need for additional action. When there
is a brief flurry of public or media interest, support for the lITO, the aims
of forest principles and assurances that overseas aid is helping the problem
are usually sufficient to confuse most journalists and members of the public
into silence.
This decline in the profile of tropical rain forests is a serious problem
and a major block to effective campaigning. Every official or corporate
body has a mass of competing demands for its limited time and the issues
that are given attention first are usually those seen as most pressing or
urgent. Placing something high up on the media's agenda, especially if it
concerns the work of the government agency or company in question, is
one very effective way of gaining that urgent attention.
However, putting tropical forests back in the media is easier said than
done. It may require a new period of interest in the environment and green
issues of the kind seen in the late 1980s and early 1990s. If this is the case,
then campaigners can expect relatively slow progress, certainly slower than
that seen in the years before the Earth Summit. Having said this, and irre-
spective of whether tropical forests are successfully put back into the public
and political mind via extensive media coverage, there are a number of
areas where campaigners should focus more specific efforts over the com-
ing few years.
Before considering these, it is worth noting that whilst rain forest cam-
paigns remain an integral part of most environmental organizations' work,
the issue is generally conceived within a wider forest campaign agenda
linked to timber certification, wood and paper consumption, indigenous
rights and biodiversity conservation in all forests - boreal, temperate and
tropical. This broadening of emphasis can be generally traced to the Earth
Summit and the double standards raised there by the developing nations of
the South, and the fact that any future international discussions will need
to cover all forests, not just those in the moist tropics. The points set out
below should be read with this context in mind.
The new rain forest campaign agenda

• Press for agreement to convert the UNCED and post-UNCED forest
consensus into more tangible national-level action plans. The divisions
and differences that lie behind the failure of countries to gain clear inter-
national agreement for the conservation and sustainable development of
forests will not go away. Because of this, there are probably no global
solutions to the problem of tropical deforestation. In order to establish a
framework for debate, priority-setting and consensus-building that can
result in action, the IPF (or whatever body succeeds it) should agree a
process whereby each country commits to tackle its own situation in
ways that are most appropriate to it. If such national-level processes are
clear enough, they could provide a means by which international fund-
ing agencies identify priorities and perhaps even provide a basis for
regional agreements or cooperation based on cultural, biological or
other affinities. The TFAP went some way down the road of national-
level planning, but addressed the wrong issues with the wrong
approach. Even where positive action was proposed, this remains
largely unimplemented. If sufficiently robust agreement, based on a
fuller and more frank analysis of what are the real underlying causes for
deforestation, progress can undoubtedly be made. For this reason, cam-
paigners should focus on gaining a new international agreement on the
root causes for deforestation and promote national action to tackle
these. The UN Convention on Biological Diversity (the Biodiversity
Convention) requires that each signatory country complete a national
biodiversity plan. The translation of this complex international agree-
ment into tangible national plans can be seen as comparable to the pro-
Rain forests - campaign priorities for the future 387
posal set out here on forests. UK NGOs working under the Biodiversity
Challenge banner have certainly found the national-level framework to
implement the Biodiversity Convention in the UK a useful means to
improve policy (see Wynne et at., 1995) and campaigners in other coun-
tries could usefully exploit the opportunities created by the Biodiversity
Convention at the national level.
• Focus on impacts of economic policy reform.
Whilst the campaigners have been very successful in highlighting the
impacts of institutions such as the World Bank, they have been less suc-
cessful in getting the environmental impacts of the structural economic
reforms promoted by the World Bank and IMF subject to scrutiny.
These reforms, generally advanced under the heading of structural
adjustment, have huge implications for the direction of economies and
the extent to which the environment is subject to excessive exploitation.
Whilst the period of structural adjustment is evidently already far
advanced, campaigners should press for a consideration of environmen-
tal issues in the conception of economic reforms promoted by the inter-
national financial institutions.
• Promote and improve the Forest Stewardship Council.
The FSC has some way to go before it will satisfy many environmental
campaigners, but it remains at present the only potential means to
enhance consumer purchasing power to influence forest management on
a large scale. The FSC could demonstrate the feasibility of wood
labelling and certification (official bodies still express scepticism), and
whilst it identifies sustainably produced wood, those still engaged in the
sale of destructively- (or even illegally-) produced wood products will be
at a distinct market and moral disadvantage. For the FSC to gain the
credibility and political standing it needs to succeed, campaigners
should constructively scrutinize its work and contribute to the continu-
ing debate to render it fully effective.
• Focus on wood consumption levels
Having won an international consensus that forest management for
wood production should be sustainable, it is now necessary to achieve
policy change to support reductions in overall levels of wood consump-
tion. Present consumption levels of wood products (including timber and
paper) are set to increase sharply (FAO, 1993b) and are already in excess
of what forests can supply on a sustainable basis. If sustainable manage-
ment is to become a reality, consumption levels must fall. Rice (1995)
calculates global per capita wood consumption limits based on estimates
of potential maximum sustainable yield from the world's production
forests and concludes that UK wood consumption levels are about three
times higher than what is regarded as sustainable. The situation is
comparable or worse in other industrialized nations (see for example
RAN, 1995). The rather unglamorous business of promoting wood recy-
cling, re-use and consumption reduction should feature in future cam-
palgmng.
In terms of what increasing levels of timber and paper consumption
mean specifically for the tropical forests, there is an emerging consensus
that one of the main impacts in the near future will be a massive expan-
sion in the area of plantations. Lohman and Carrere (1996) provide a
summary of the forces driving this process with an analysis of expected
impacts.
• Pursue corporate interests through financial and investment routes
The recent relative decline in the role of official development aid com-
pared to private-sector finance and investment suggests the need for a
renewed focus on companies. One consequence of the liberalization of
financial markets and trade during the 1990s has been a proliferation in
the connections between consumers and investors in the North and
companies undertaking business activities in the rain forests (see Korten,
1995 for a full explanation of the growing role and impacts of the pri-
vate sector in the South). Research to establish who is connected to who
by what financial or investment relationships will reveal rich new seams
of potential influence. The enormous and growing number of people
with private pension schemes, mortgages, shares and bank accounts
could prove to hold a constituency of green consumers with massive
leverage on policy. Campaigners should research potential investment
relationships with companies wrecking the rain forests and mobilize
investors to place pressure. This campaigning point is of course not
specifically of interest to rain forest campaigners and would equally
assist those fighting global climate change.
Critical to effective campaigning against companies damaging rain
forest areas is a parallel critique of the World Bank's role as a facilitator
for private-sector activities in the South. The Bank increasingly empha-
sises the role of the private sector in the delivery of its development
objectives, but sometimes in the absence of the environmental and social
safeguards necessary to prevent serious negative impacts. For this rea-
son, the World Bank should remain under close scrutiny by rain forest
campaigners and subject to robust challenge when its lending operations
depart from the principles that underline sustainable forest use.
• Networking and global connections
The stronger consensus that exists between the campaigners from North
and South compared to that between governments is a tremendous
strength. Continued dialogue and sharing of ideas and campaign strate-
gies must continue. The globalization of the world economy presents
many challenges and dangers that can only be effectively addressed by a
References 389
coherent campaigning agenda shared by peoples' movements from both
North and South. Whilst discussion, dialogue and information-sharing
leading to joint approaches may not seem top priority in the midst of a
campaign, they form a vital part of the effectiveness of campaigning that
represents a very worthwhile long-term investment.
13.8 CONCLUSIONS
Decisions affecting the rain forests, in common with most decisions affect-
ing the management of environmental assets, are mainly political. Whilst
the scientific and technical debates help to define the limits of what is politi-
cally possible, these factors generally do not determine what is decided by
governments, international institutions or companies. These decisions are
most often based on wider economic concerns, what issues and policies will
win votes, and the balancing of the various interest groups who pressure
governments and other decision-makers. Campaigners can influence the
political debate so that scientific, technical and moral concerns are reflected
in the approach taken by key players, to the extent that the sustainable use
of the tropical rain forests moves from rhetoric to achievement.
ACKNOWLEDGEMENTS
Many thanks to Jean McNeil and Daniel Carpenter for their assistance in
getting this chapter finished, finally.
REFERENCES
Barbier, E. B., Burgess, J., Bishop, J., Aylward, B. and Bann, C. (1992) The
Economic Linkages between the International Trade in Tropical Timber and the
Sustainable Management of Tropical Forests, International Institute for
Callister, D. J. (1992) Illegal Tropical Timber Trade: Asia-Pacific, TRAFFIC,
Sydney.
Colchester, M. and Lohmann, L. (1990) The Tropical Forestry Action Plan: What
Progress? World Rainforest Movement, Penang, Malaysia.
Counsell, S. (1993) The International Tropical Timber Agreement; Conserving the
Forests or Chainsaw Charter? Friends of the Earth, London.
Counsell, S., Juniper, T. and LeMarchant, M. (1992) Whose Hand on the
Chainsaw? Friends of the Earth, London.
DoE (1994) Digest of Environmental Protection and Water Statistics, No. 16,
Department of the Environment, HMSO, London.
FAa (1981). Tropical Forest Resources, Food and Agriculture Organization,
Rome.
FAa (1993a) Forest Resources Assessment, Food and Agriculture Organization,
Rome.
FAO (1993b) Year Book of Forest Products, Food and Agriculture Organization,
Rome.
George, S. (1989) A Fate Worse than Debt, Eanhscan, London.
Given, D. (1990) Conserving botanical diversity on a global scale. Annals of the
Missouri Botanical Gardens, 77, 48-62.
Kimerling, j. (1991) Amazon Crude, Natural Resources Defence Council, USA.
Konen, D. (1995) When Corporations Rule the World, Eanhscan, London.
Lamb, R. (1996) Promising the Earth, Routledge, London.
Lamprecht, H. (1989) Silviculture in the Tropics, Deutsche Gesellschaft fur
Technische Zusammenarbeit, Germany.
Lohman, L. and Carrere, R. (1996) Pulping the South, Zed Books, London.
Mendes, C. (1989) Fight for the Forest, Latin American Bureau, London.
Mombiot, G. (1993) Mahogany is Murder! Friends of the Earth, London.
Moody, R. (1991) Plunder! PanizanslCAFCA, London.
Myers, N. (1985) The Primary Source: Tropical Forests and our Future, W. W.
Nonon, London.
Poore, D. (1989) No Timber without Tree - Sustainability in the Tropical Forest,
Eanhscan, London.
RAN (1995) Cut Waste not Trees, Rainforest Action Network, San Francisco.
Rice, T. (1991) Plunder in Ghana's Rainforests for Illegal Profit, Friends of the
Eanh, London.
Rice, T. (1994) Crude Operator: the Environmental, Social and Cultural Effects of
Texaco Oil Operations in the Tropical Forests of Ecuador, Friends of the Earth,
London.
Rice, T. (ed.) (1995) Out of the Woods: Reducing Wood Consumption to Save the
World's Forests, Friends of the Earth, London.
Sainsbury's (1996) Wood Policy, Sainsbury's pIc, London.
Sierra Club (1986) Bankrolling Disasters: International Development Banks and
the Global Environment, Sierra Club, San Francisco.
Talbot, L. M. (1990) A Proposal for the World Bank's Policy and Strategy for
Tropical Moist Forests in Africa, World Bank, Washington, DC.
World Bank (1991) Forestry Sector Lending Policy, World Bank, Washington, DC.
WRI (1985) Tropical Forests: a Call for Action, Repon of an International Task
Force convened by World Resources Institute, World Bank and UN Development
Programme, World Resources Institute, Washington, DC.
Wynne, G., Avery, M., Campbell, L., Gubbay, S., Hawkswell, S., juniper, T., King,
M., Newberry, P., Smart, j., Steel, c., Stones, T., Stubbs, A., Taylor, j., Tydeman,
C. and Wynde, R. (1995) Biodiversity Challenge, 2nd edn, Royal Society for the
Protection of Birds (RSPB), Sandy, UK.
-14------
Synthesis
F. B. Goldsmith
Department of Biology, University College London, Gower Street, London
WC1E 6BT, UK
In the preceding chapters we have covered a considerable number of ideas

and pages. It is now time to present some sort of structure or synthesis of
the preceding material. I will do this with the minimum of comment, as
that has been offered earlier (Chapter 1).
In future, in order to understand, value and protect tropical rain forest
we must work on six fronts, most of which should be considered in combi-
nation. Forests are for the production of materials, the protection of func-
tions such as those associated with watersheds and for the conservation of
biodiversity and other values. Forests also provide services for people. We
need to increase people's perception of the value of forests, to develop poli-
cies and institutions for their protection and sustainable development, and
in parallel we need to increase or enhance our knowledge of the contents
and functioning of these most complex ecosystems. Here are the main headings:
14.1 PRODUCTION
The main categories are:
• timber
• fuel wood and charcoal (approx. 80% of harvest)
• paper
• non-timber forest benefits (food, fodder, traditional medicines).
Goods can come from a range of different types of forests in which yields
can vary enormously. Similarly, conservation values may vary substantially,
and not necessarily with any relationship to productivity:
• natural forests
• extractive reserves
• managed forests
• plantations.
392 Synthesis
14.2 PROTECTION
Tropical rain forests have a variety of protective functions, including:
• Watersheds, e.g. from flooding
• Soil erosion (loss of particulates, nutrients)
• Climate, e.g. carbon balance, regulation of global warming
• Forests suffer from losses due to deforestation, degradation and frag-
mentation (especially of biodiversity).
14.3 CONSERVATION
There are many conservation benefits associated with rain forests, especially:
• biodiversity safeguard/enhancement
• non-timber forest benefits
• landscape and other values.
14.4 SERVICES FOR PEOPLE

Forests provide the items above and also employment, as well as places and
materials of religious significance:
• food
• materials, e.g. for housing
• employment: wardening, surveymg, interpretation, logging, other
forestry operations
• traditional medicines
• religious artefacts and places, e.g. sacred groves
• recreation - tourism.
14.5 POLICIES AND INSTITUTIONS

Much greater attention needs to be paid to the role of policies and institu-
tions in the protection of rain forests. These include:
• information - knowledge base, interpretation
• participation - community forestry, participatory rural appraisal/rapid
rural appraisal
• cooperation - fewer cooperatives than in agriculture
• financial incentives (subsidies, compulsory investment)
• market development
• laws and regulations; contracts and agreements
• trade restrictions (boycotts)
• protected area design and management
• debt-for-nature swaps
• campalgnmg
• promote agencies such as the FSC (labelling, certification).
Knowledge base 393
14.6 KNOWLEDGE BASE
This is still sadly lacking, and the old maxim 'the more we know, the more
we find that we need to know' has never been more appropriate:
• survey: rates of losses (e.g. remote sensing), forest inventory (i.e. tim-
ber), biodiversity
• research: forestry: extraction techniques, increasing yields, efficiency of
utilization, appropriate technology, role of plantations, recycling, substi-
tutes for timber
• research: social: cultural practices, participatory rural appraisaVrapid
rural appraisal
• research: environmental: hydrology, soils, climate, biology/ecology.
In the future in developing countries, including those with tropical rain for-
est, there will be an increase in the number of people, at least to the year
2050. This will necessitate more land for food production, so considerable
areas of forest are likely to be converted to arable and pasture. At the same
time, areas of forest on steep slopes, on fragile soils or with high biodiver-
sity will need to be protected in some way. Corridors will also be needed
for migration, especially of large vertebrates such as forest antelope and
elephant. Most of these will need buffer zones to protect them from fire,
poachers, firewood collectors and graziers.
As plantations can yield considerably more wood than can harvesting
natural forests, their selective use to relieve pressures on natural forests
should be promoted in a restrained manner, at least until such time as tim-
ber recycling and increased efficiency of use make further exploitation
unnecessary.
The involvement of local people in decision making and forest manage-
ment needs to be fostered and, where appropriate, extractive reserves, mul-
tiple use of forests and agroforestry need investigating. The valuation of
forests by local people and officials should be promoted by the provision of
more information and increased participation. Internationally and nation-
ally this can be achieved by more and better campaigning. Local people
will protect areas of significance or value to them, e.g. sacred groves, and
in turn they need protection from multinational or large national forestry
operations. Environmental Action Plans, as proposed by agencies such as
the World Bank and UN agencies, may sound attractive but only involve
the more powerful stakeholders. Maybe the use of land designated as 'foret
communial', as used in some former French territories, should be explored.
In most developing countries, forestry regulation needs developing by
the use of incentives, such as those for better forest management and more
efficient extraction. Compulsory investment can be effective, for example
in Norway companies have to pay a proportion of profits into a tax-free
fund which can be reclaimed if spent on approved types of forest improve-
ment. The regulation of harvesting and of forest operations could minimize
394 Synthesis
damage to non-target tree species and non-timber products. Possibilities
include allowing cutting only of trees of minimal diameter; rationing areas
or volumes for cutting each year, i.e. an annual allowable cut; and checks
on the efficiency of harvesting. Finally, legislation can be used to enforce
these approaches, but often financial incentives such as grants and tax
allowances can be more effective.
Buffer zones need to be developed and more effective eco-development
practised such that more efficient means of producing fodder and food
relieve pressure on core areas. A combination of effective protection of core
areas combined with local peoples' cooperation with the eco-development
of buffer zones may offer the most realistic and practicable way forward.
In developed countries we need to campaign for effective international
agreements; influence the policies and activities of international companies
and institutions; extend certification schemes (considering country certifica-
tion as well as product certification); and generally ensure that tropical
forestry is more sustainable. No single action is going to save tropical rain
forests but the further empowerment of local people, a raising of their stan-
dards of living, their knowledge base and so their valuation of these
resources, the control of national and multinational companies, reduced
demand for products because of better efficiency of utilization and recy-
cling, more effective policies and institutions, incentives and controls will
gradually lead to a wider recognition of the importance of these unique
resources.
As well as the preceding chapters, the following references have been
influential in my choice of topics for this synthesis: Osmaston, 1968;
Smith, 1968; Helliwell, 1982; Harris, 1984; Anon., 1985; Grainger, 1993;
Harris, 1996; Seip, 1996.
REFERENCES
Anon. (1985) The State of India's Environment, 1984-5, Second Citizens' Report,
Cen~re for Science and Environment, New Delhi.
Grainger, A. (1993) Controlling Tropical Deforestation, Earthscan, London.
Harris, K. L. (1996) Making Forest Policy Work, Conference Proceedings, Oxford
Forestry Institute, Oxford.
Harris, L. D. (1984) The Fragmented Forest, University of Chicago Press, Chicago.
Helliwell, D. R. (1982) Options in Forestry, Packhard, Chichester.
Osmaston, F. C. (1968) The Management of Forests, George Allen & Unwin,
London.
Seip, H. K. (1996) Forestry for Human Development, Scandinavian University
Press, Oslo.
Smith, D. M. (1968) The Practice of Silviculture, Wiley, New York.
List of acronyms
CINTRAFOR Center for International Trade in Forest Products

CITES Convention on Trade in Endangered Species of Wild
Fauna and Flora
EM ectoOlycorrhizas
ERS European Radar Satellite
ESA European Space Agency
FAO Food and Agriculture Organization of the United
Nations
FPA Forest Principles Accord of UNCED
FSC Forest Stewardship Council
GAIT General AgreeOlent on Tariffs and Trade
GEF Global EnvironOlental Facility
GIS geographic inforOlation systeOl
GNP gross national product
GSP generalized systeOl of preferences (tariff rate)
lIED International Institute for EnvironOlent and
DevelopOlent
IMAGE integrated Olodel to assess the green effect
IMF International Monetary Fund
INPA National Institute for Research in ADlazonas, Brazil
INPE Brazilian National Institute for Space Research
IPCC IntergovernOlental Panel on CliOlate Change
IPF IntergovernOlental Panel on Forests
ISO International Organization for Standardization
mA International Tropical TiOlber AgreeOlent
mo International Tropical TiOlber Organization
mCN World Conservation Union
LAl1 Local Agenda 21
MFN Olost favoured nation (tariff rate)
NASA National Aeronautics and Space AdOlinistration
NFAP National Forestry Action Plan
NGO non-governOlental organization
NOAAAVHRR National Oceanographic and AtDlospheric
AdOlinistration's Advanced Very High Resolution
RadioOleter
NTFPs non-tiOlber forest products
396 List of Acronyms
ODA Overseas Development Administration, UK
OECD Organization of Economic Co-operation and
Development
PPFD photosynthetic photon flux density
RAN Rainforest Action Network
RIC Rainforest Information Centre
SET Science, Engineering and Technology Week (UK)
SPS Sanitary and Phytosanitary Measures (GAIT, Uruguay
Round)
TBT Technical Barriers to Trade (GAIT, Uruguay Round)
TFAP Tropical Forestry Action Plan
TSBF Tropical Soil Biology and Fertility Programme of
UNESCO
UKTFF UK Tropical Forest Forum
UNCED UN Conference on Environment and Development
UNDP United Nations Development Programme
UNEP United Nations Environment Programme
UNESCO United Nations Educational, Scientific and Cultural
Organization
VAM vesicular-arbuscular mycorrhizae
WCFSD World Commission on Forests and Sustainable
Development
WRI World Resources Institute
WRM World Rainforest Movement
WTO World Trade Organization
WWF Worldwide Fund for Nature
Index
Note: page numbers in italics refer to tables, those in bold refer to figures.
aboriginal populations 45 animals 9,10, 11

see also indigenous people logging impact on populations 76
a\=ai 34, 38 Annona squamosa (ata) 32
action plans, national-level 386 arum, titan 362-4
affluence 184 ash 130-1
African Development Bank 382 carbon release 261
agouti 14 deposition 121, 124
agrarian change modelling 318,319, ata 32
320 Attalea phalerata (baba\=u palm)
Agreement on the Application of 33,38
Sanitary and Phytosanitary authority, centralization 192
Measures 212
Agreement on Technical Barriers to baba\=u palm 33, 38
Trade 212 bacteria
agricultural area, forest cover nitrifying 148, 150
simulation models 330 nitrogen-fixing 148, 151
agricultural frontier expansion 336 Baillonella toxisperma (moabi) 81
agricultural productivity 327 Baka pygmies 78, 80
agriculture bankrupt state 91
optimized 177 Barro Colorado Island 44
settled 271 benefit flows 295
see also farming; shifting cultivation; Bertholleta excelsa (Brazil nut) 14
slash-and-burn biodiversity
agrochemicals 36 commercialization xiii
agroforestry systems 24 loss 270
managed secondary forest 25 deforestation 179
products from native plants 26-7 exploitation of lesser-used species
aguaje palm swamps 38 222
aid 369-70, 380-3 forest degradation 108
albedo 106 non-timber function 255
definition 122 political arena 365
fire effects 121 understorey 232
Amazon Basin 14-15 Biodiversity Convention 36-8, 39,
Amorphophallus titanum (titan arum) 351-2,386,387
362-4 biomass
amphibians 10 above-ground 130
anaconda 14 accumulation during regeneration
andiroba 39 159-61
Aniba rosaeodora (rosewood) 32 microbial 148-9
398 Index
needs and poverty 192 Cameroon
biotic feedback, forest degradation community forest management
107-10 project 92-3
birds 9, 10 World Bank support for roads
logging effects 67 scheme 382
black waters 14 campaign 368-71
Body Shop Company 28 agenda 368
books 355-6 government action pledges 375
Bora Indians (Peru) 24, 25 information publication 372-3
botanic gardens 253 issues 370-1
Brazil nuts 14,23,24 launching 368-9
extraction 30 media coverage 372, 375
factory 31 political lobbying 374
gathering 32 problem acknowledgement by
oil 28 industry 375
processing plants 28, 31 retailer involvement 376
sustainable extraction 36 specific company operations 379
British Gas 377 spread to other countries 373-4
Bruntland Report 347 success 370
buffer zones campaigners, local 377, 378
eco-development 394 campaigning, effective 367-89
tree-covered 111 campaign origins/impacts 368-71
bulldozers 53, 54, 62 international agreements 383-5
bulldozing priorities for future 385-9
site heterogeneity 146 profile decline 385
soil effects 125 wider agenda 386-9
soil structure 164 campaigning organizations 367-8
topsoil chemical properties 139 tactics 367
burning camu-camu 38
carbon dioxide release 261 canopy 5
deforestation 130-2 disturbance 64
for grazing 271 opening 105
nutrients 130-2, 133 rainfall interception 64
slash 130, 131 solar radiation absorption 46
soil microorganism effects uniform opening 48
147-8 capital
soil organic matter 130-2 investment 309
soil temperature 122-3 owners 277
topsoil chemical properties 139 capture, non-timber values 256-7
see also fire Carapa guianiensis (andiroba) 35, 39
carbon
caatinga, nitrogen fixation 151-2 burning 133
caciques 14 credit values 263
cafe marron tree 361 emissions and deforestation 263
calcium 128, 129, 130 reduction 263
burning 131, 132, 133 release 175
pasture 136, 138 sequestration 255
recovery after deforestation 134 emerging markets 265
Index 399
shade effects 140 Chacabo Indians (Bolivia) 21-2
carbon dioxide 107 charcoal 202
emission trend simulation 332, 333 carbon release 261
release 261 deforestation 308
carbon storage 261-3 extraction 4
land-use conversion 262 production 201
payment for function 257 chlorine, pulp bleaching 218
sustainable forest management 282 clear felling 176
values 264 alternative land use 182
cash crop 309 clearance
production 271 for agriculture 269
castanea nuts see Brazil nuts fallow 140, 142
cauliflory 5, 7 farming systems 185
Ceiba pentandra 57 for grazing 271
cement fibreboard 221 habitat degradation 164
Center for International Trade in Forest mechanized 121, 124-5
Products (CINTRAFOR) 206 mulching 126
certification 222-30 nitrogen mineralization 149
campaign agenda 387 nutrients 132-5, 144, 146
constraint removal 296 plantations 307
country 228-30, 248 rate of loss 385
demand for wood products 223 rates 102
direct costs 226-7 regeneration 144, 146
estimate of benefits 225, 226 relative humidity 124
Forest Stewardship Council 364, 375 root mat effects 146-7
global market impact 223 seed ecology 156-7, 158, 159
global promotion 223 soil fertility 159-61
green 217 soil hydrology 126
premium 224, 225 soil microorganisms 248
harmonization 227-8 soil-plant-nutrient relationships 134
indirect costs 227 climate change 270
international programmes 298 global 332, 334
internationally agreed 230 cocoa plantations 307
mutual recognition 227-8 commercial exploitation, potential
niche markets 223, 225 species 108
non-tariff barriers 229 commercial functions 299
policy instruments 298 commercial interests, environmental
recent management practice improve- ethic 357
ment 227 commercial species
substitution of non-wood alternatives improvement 109
226 large tree loss 56-9
sustainable forest management 248 local extinction 57
practices 229, 230 logging out 57
promotion 223 wild relatives 109
tariff barriers 229 commercial value
trade losses 227 forests xii
verification 222 tropical trees 50-1
voluntary 230 communication
400 Index
education 348-55 culture disappearance 94
getting the message across 361-4 cupua\a 37
resources 355-61 curare 37
communities, local 189-90
community forest management project debt
92-3 crisis 380
community-owned forests 189 external 184
competition for land, deforestation 183 foreign 311, 317
competitiveness, international 245 decay 130, 134-5
composition decision making, local people involve-
forest structure 56-64 ment 393
logging 56-64 deforestation xiii, 99
concessions 86, 87 carbon emissions 263
arrangements 234-5 causes 175-6, 177, 182-6, 187,
tax breaks 277 188-9
conservation 76-7, 392 proximate 182-5
benefits xii ultimate 182-5
costs with global warming 263 changing parameters 188-9
cultural pressures 94 charcoal 308
facilities 253 competition for land 183
institutional constraints 175-6 definition 303
international compensation 296 degradation 179, 304
national strategies 294 development projects 188, 189
political arena 365 economic growth/development 309
responsibility xiv estimates 102
species protection 111 extent 270
strategies 110-12 external debt 184
consumers external influences 311
distant 183 extractive reserves 24
hardwood imports 364 forest type 179
pressure on retailers 370 fragmentation 179
consumption levels 387-8 fuel wood 308
contract arrangements 234 global trends 211
Convention on Biodiversity 36-8,39, Green Revolution 327
351-2,386,387 immediate causes 305-11
corporate environmentalism 379 inequality 309-10
corporate interests 277, 388 land
corporate tactics 292 allocation 334-5
credit associations 91 development 335
crops reform measures 327
cash 271, 309 use after 305
plantation 311 land-cover classes 180
staple 307 land-use change 304-5, 310
cultivars 109 logging 271
cultivation, extractivism 28-9 spatial trend modelling 338-9,
cultural asset loss 271 340
cultural dependence on forest 92 long-term spatial trend modelling
Cultural Survival Enterprises 28 334-6,337,338-9
Index 401
long-term trend projection 328, 329, social collapse 255
330,331,332,333,334 soil 119-20
market failures 183-4 biology effects 146-53
microbial biomass 148-9 chemical properties 128, 129,
microclimate effects 121-4 130-5
modelling 303 effects 121
over time 177, 179 microorganisms 147-51
pattern 103 organic matter 128, 129, 130
permanent agriculture 306, 307 physical properties 124-6
physical factors 310 spatial trends 335-6, 337
plant nutrients 128, 129, 130 state intervention 310
policy 310 temporal variation 188-9
failures 183-4 theoretical model 305-11
population timber trade model 239-40, 241
growth 308-9 tropical timber extraction 371
pressure 184-5 underlying causes 308-11
poverty 309-10 vegetation 119-20
rate 161, 176-80, 182,304,330, effects 121, 153-7, 158, 159-63
332,385 degradation 100-5, 179, 204
FAO estimates 313 biodiversity loss 108
mean annual 315, 316 biophysical feedback, minimization
regeneration 110
effects 120 biophysical implications 105-10
patterns 155, 161-3 biotic feedback 107-10
regional differences 185-6, 187, 188 damage 104-5
regional effects 177 downstream flood impact 107
regional trend projections 331 environmental impact 305
regression models 311-13, 314-15, industrial forestry 277
316-17,318,319 monitoring 100, 101-4
cross-sectional global studies 313, physical feedbacks 106-7
314-15,316-17 rate 304
data 312-13 remote sensing 101, 102
foreign debt 317 temperature increase 107
forest cover 319 demographic transition theory 310
forest transition 321-2 demonstration, non-timber values 256
GNP per capita 316, 317 denitrification 149, 150-1
methods 311-12 developing countries
national land-use transition 320-1, debt crisis 380
322 reaction to campaigns 370
poverty 317 development 380-3
specification 312-13 countries' rights 384
sub-national cross-sectional studies projects 188, 189
317,319 disturbance 46
sustainable development 322 canopy 64
variables 311-12 natural 56
resource control inequity 185 regeneration 48
selective logging 304 soil 65
shifting agriculture 271, 305-6, 307 diversity 11
402 Index
drip tips 5, 6 environmental management
drought 44-5 peripheral areas 335
fires 56 plans 87-91
understorey plants 155-6 environmental protection 255,345-6
drugs, value of undiscovered 259 environmental questions, public 369
environmental restrictions on trade
Earth, human analogy 345 217,218
Earth Summit (Rio) 37, 347, 383 environmentalism, corporate 379
forest principles 384-5 epiphytes 5, 7, 8
post-summit process 385, 386 large tree loss 58
preparatory meetings 383-4 erosion 45, 64
eco-labelling 217 deforestation 126-7
ecological diversity 146 forest degradation 106-7
ecological function value 261 logging 65
ecological literature 368 loss prevention 66
ecological monitoring 194 on-site 127
ecological zoning 111 prevention 261
economic growth/development 309 protection 126
economic incentives 275-6 rain-splash 64, 66
economic policy reform 387 rate 127
ecosystem splash 65
benefits 183 estates
function loss with deforestation 179 government control 293
undervaluation 183 large agricultural 307
ecotourism 260-1 ethnobotany 21, 353, 360
ectomycorrhiza 152 Eucalyptus deglupta 141
edge effects 12 Eupatorium 59
education 347-55 Europe, forest products trade 204, 246
formal 348-52 Euterpe oleracea (a\rai) 34
funding 349, 351 evapotranspiration 106
non-formal 352-5 evergreen forest fragmentation 12
postgraduate 351, 352 exchange systems, non-monetary 91
professional training 351-2 exotics, invasions 179
standardization 190 expeditions, overseas 351
teachers' packs 348-9 exploitation xii
tertiary 350-1 open access 233
elephants 67, 81 export
energy needs, poverty 192 bans 218
enforcement attempts 275 restrictions 217
Entandrophragma spp. (mahogany) 57 tax 218, 244
Environment Action Plans 393 tropical hardwoods 338, 340
environmental degradation 270, 305 expropriation of local peoples' rights of
poverty 185 access xiii
environmental impact assessment extensification 309
295-6 extinctions
environmental importance of forests forest degradation 108
Xl-Xli fragmentation 108
environmental instability 43-6 localS7,67
Index 403
mass 109 financial institutions 380-3
mining activities 378-9 financial market liberalization 388
rates 12-13 fire 56
extraction 52 canopy opening 105
extractive reserves 24, 26-7, 28-9, 30, controlled 85
31-2 fragmentation of forest 179
agriculture needs 32 nutrient loss 121
Amazonian Brazil 26 seed mortality 156-7
area required to sustain people 29, soil temperature 122-3
31 vegetation impact 85
sustainability 28-9 vulnerability 45
extractive values 258-60, 264 see also burning
extractivism, cultivation 28-9 fisheries protection 261
extractivists flood
diet 36 control 261
poverty 24 downstream impact 107
extreme events 44 flooding 9
flowering, synchronized 8
fair trade 357 Food and Agriculture Organization
fallow 140, 141 (FAO) 192, 199,200
nutrients 145 Production Yearbook 312, 316
soil changes 141-2, 143-4, 145, 146 forest
soil fertility index 144 alternative use costs/benefits 294
tree density/dimensions 142 area 323, 326
farm bush 85 total 2
farming trend projections 329
clearance 185 assets
rate of diffusion of innovations 319 knowledge 295-6
see also agriculture valuation 295
faunal reserves 88 wasting 269
felling boundaries 5
damage 53 certified global 224
directional 53 characteristic features 7
gaps 53, 62-4 climate 1
monocyclic 48 community-owned 189
polycyclic 49 conservation 392
regeneration 25 conversion
relocation of trees 52 to farm fallow 89
uniform 48, 49 to other land use 176
vine cutting 53 to other land-cover classes 180
fertility rates 310 crisis 255-7
fibre baskets 29 damage 99-100, 104-5
Ficus spp. (fig) 58 clearance by farmers 105
fig, strangling 58 monitoring 100
films 357-8 secondary 105
finance definitions 4-5
overseas 311 degradation 270
private-sector 388 destruction 84-7, 346
404 Index
economic consequences of loss 175 species
farming systems 84-5 diversity 2
floor inhabiting range of environments
herbaceous vegetation 154-5 85
light reaching 5 narrowly endemic 85
global 370 structure 5-7
goods composition 56-64
needs 287 threats 3-4
security 293 transition 321-2
institutions 392 tropical moist 100-2
knowledge base 393-4 undervaluation 182, 183,272
law 275, 276, 278 unexploited 233
West Africa 81 valuation 294
location 1, 2 zoning 87
monsoon 8 see also canopy; degradation; under-
non-market value 256 storey
non-timber functions 255 forest cover 319
oligarchic 38 agrarian change modelling 318,319,
participatory assessment 296 320
penetration 81 extent 102
physical problems 270-2 land use spatial trends 336, 337
plans modelling 322-34
environmental impact assessment cross-sectional international
295-6 models 322-3, 324-5
social impact assessment 295-6 cross-sectional national models
policy 392 323,326
influences 273, 274, 275-9 long-term deforestation trend
power locations 83 projection 328-34
principles 384-5 national time-series models 324,
production 391 326-8
profile 6 simulation models 330,331,332,
protective functions 392 333,334
qualitative changes 180, 182 national 323
quantitative studies 13 theoretical models of trends 319-22
rate of disappearance 3-4 forest management
regeneration 7-8 community 87
replacement for other land use improvement for certification 227
269 local people involvement 393
saving ideas of expatriate multiple-use 277
organization 92 participatory 300
secondary 99,180,207 practice improvement 227
security concept 293-4 resource use 88-9
services see also sustainable forest manage-
needs 287 ment
for people 392 Forest Principles Accord (UNCED) 242
security 293 forest products
size class distribution 56 harvesting 104
society's mutual dependence 82 indigenous use 21-3
Index 405
forest products trade 200-1, 202, fruits
203-4,205,206-7,246 medicinal oil 39
barriers 207-10, 211, 212, 213, 214, sustainable extraction 35-6
215,216-22 fuel wood 4, 202
certification 222-30 deforestation 308
eco-labelling 217 production 201
environmental restrictions 217, 218 fungi 9,148
export 218 mycorrhizal 151, 152
forest resource base assessment 221
future trends 206-7 game animals 36
generalized system of preferences gap
(GSP) 210 ecology 8
global 201 logging 54
exporters 205-6 root mass 147
labelling 222-30 size 53, 55
lesser-used species 220-2 gap-phase dynamics 154, 164
market access 207-10, 211, 212, Garden of World Medicine (Chelsea
213,214,215,216-22 Physic Garden) 353
most favoured nation (MFN) 210 gas companies 377
new barriers 216-19 General Agreement on Tariffs and
preference system 210 Trade (GATT) 208-9,219
price trends 204 special agreements 212
quantitative restrictions on imports tariff reductions 213, 214, 215, 216
217 Uraguay Round 208, 210, 213, 214,
regional 204, 205, 206 215,216,247
tariff sustainable forest management 247
barriers 208-9 trade restrictions 374
elimination 210 generalized system of preferences (GSP)
escalation 209-10 210
value-added processing 209 genetic degradation 58
forest society genetic material, social value 260
cosmology 83 geographical information systems 194
fear of forest regeneration 83-4 germination, habitat availability/prefer-
mutual dependence with forest ence 162
82 glacial periods 44
powers from outside world 84 global connections 388-9
space 83 Global Environmental Facility xiii, 183,
see also indigenous people 245,257
Forest Stewardship Council (FSC) 280, global forest 370
282,298 global warming 107, 175,262-3
certification 364, 375 Gme/ina arborea 141
promotion 387 GNP per capita 316, 317
forest-society relations 76, 77 government
forestry codes, flouting 372 control 276
(oret communale 393 education sponsorship 349
fragmentation 11-13, 103, 179 forest management responsibility 296
extinctions 108 functions 283
Friends of the Earth 367 national forum 292, 293
406 Index
resettlement programmes 307 West Africa 78-80
subsistence 177 world view 82-3
government agencies 86, 282 see also forest society
coordination 191 indigenous reserves 111
disabling institutional cultures 295 induced innovation theory 309
green premium, certification 217, 224, industrialization, forest-based 236
225 inequality 271,309-10
Green Revolution 327 information sheets 356
greenwash 379 information systems, dynamic multi-
guarana 28-9, 36 factor 295
institutional constraints 189-94
habitat international 192-4
fragmentation 179 local 189-90
plantings for educational purposes national 190-2
353 institutional cultures, disabling 295
hardwood institutional failure 272-3
allocation 338 institutional inertia 294-5
resource scarcity 246 institutional mechanisms, resources
harvesting 190, 191
intensity 50-1 institutions 282-3, 286-7, 288, 289
unsustainable practices 32 consensus 297
harvests, value 257 coordination 191
Hevea brasiliensis (rubber) 30, 109 individuals within 286
high-grading 3, 4, 234 international 192-4
household income 260 motivations 286-77
human activity 89 participation of players 283, 286
hunter-gatherers 80 rain forest protection 392
hydroelectric power 308 roles 282-3
hydrological benefits 261 sustainable forest management 299
hydrological cycle 64-5 intellectual property rights 183
hydrology, logging effects 65-7 intensification 308
inter-governmental associations 280
imports Intergovernmental Panel on Forests
barriers 217 199,385
surcharges 244 international agreements 383-5, 394
in-breeding 58 international institutions 192-4
incentive structures, internal 283 assessment of issues 193
Indian Forest Service xii constraints 192-3
indigenous people 45, 75-8 environmental record 193
campaign involvement 377, 378 fail ures 193
devolution of power to 294 international market access 246
ecosystem knowledge 108 International Monetary Fund 380
involvement policy reform 387
decision making 393 International Organization for
forest management 393 Standardization (ISO) 298
policy attitudes 276, 277 International Tropical Timber
management of resources 23-4 Organization (lITO) xiv, 199,
use of forest 21-3 200,280,282
Index 407
commodity agreement 374 Lantan camara 179, 181
logging study 371-2 latex extraction 35
sustainable use agreements 374 leaching, nitrogen 149
International Union for the leaf litter
Conservation of Nature 192 nutrients 130
Internet 358-9 removal 60
investment 309 leaflets 357
compulsory 393-4 leaves 6
relationships 388 allelopathic exudates 6
tax breaks 277 structure 5
Ischrosiphon aruma 29 synchronized production 8
lectures 359, 360, 361
jaguar, spotted 14 legal systems in West Africa 81
jhum system 140, 142 legislative mechanisms 278
see also forest, law
Ka'apor people (Amazonian Brazil) 22 lemurs 9
kaolisols 11 lesser-used species 247
knowledge lianas 5, 8
base 393-4 life support functions of forest 255
democracy 296 light 122
forest assets 295-6 intensity 154-5
loss 271 response of trees 47
light-demanding plants 154
labelling 222, 298 lightning 44
campaign agenda 387 linalo132
land livelihood loss 271
alienation 177 livestock credit 275
allocation 335 lobbying 372
dynamic patterns 335 Local Agenda 21s (LA21) 347, 365
claims 273 local people see indigenous people
development 335 log-loading bays 60-2
post-clearance cover 103-4, 107 logging 4
reform measures 327 activity 85-6
tenure 273, 299-300 damage 48
titling 275 deforestation 182, 271
land-use erosion effects 65
change forest
deforestation 304-5, 305, 310 composition 56-64
modelling 303 hydrology 65-7
circular zones 335 structure 56-64
compromise systems 111 zoning 87
conversion 262 gaps 54
forest replacement 269 hardwood
national morphology 305,335 allocation 338
national transition 320-1, 322 trade projection 338-9, 340
spatial trends 336, 337 illegal operations 275, 372
Landsat imagery 102-3, 104 intensity 104-5
landscape, ecological zoning 111 management plans 86
408 Index
natural disturbance 56 Mexico, product use by indigenous
rates 86 people 23
regeneration 59-60 microbial biomass 148-9
seedling microbially mediated processes 147-51
growth 157, 159 microclimate, deforestation effects
survival 49 121-4
selective 49, 100 migration 88
degradation 304 mining 308, 378-9
forest damage 104-5 export demand 311
system 51 private sector 377
spatial trend modelling 338-9, 340 mission statements 354-5
sustainable 86 moabi 81
timber exports 85 modelling, deforestation 303
trails 104 monocarpy 8
see also concessions; roads monsoon forest 8
logs Morea excelsa 51
extraction routes 52-3 most favoured nation (MFN) 210
prices and harvest costs 240 mulching 126
value 278 Myciaria dubia (camu-camu) 38
macro-economic policies 271 National Curriculum 348

magazines 356 national fora 292, 293
magnesium 128, 129, 130 National Forest Programmes xi
burning 133 National Forestry Action Plan 245,
pasture 136, 138 383
mahogany 57-8 National Forestry and Conservation
mammals 9, 10 Action Plan (Papua New Guinea)
mapping 52 279
market access 207-10, 211, 212, 213, national parks 88, 111
214,215,216-22 national targets 293-4
international 246 negotiation 292
tariff escalation 209-10, 212 networking 388-9
markets newsletters 357
demands 287 newspapers 356
failures 183-4 niche markets, certified wood products
initiatives 296 223,225
resource allocation 287 nitrates 150-1
mass-extinction episode 109 nitrification 149
mast-fruiting 8 nitrogen 128, 129, 130
Mauritia flexuosa (aguaje palm) burning 131,132
38 conservation with regeneration
maximum sustainable yield 257 149-50
media fixation 151
aural 359-61 leaching 149
campaign use 368 mineralization 149
printed 355-7 pasture 136, 138
visual 357-9 nitrogen dioxide 151
Merremia spp. 62, 63 non-extractive values 258-60, 264
Index 409
non-governmental organizations 192, loss 66
279,380 oxisols 134, 135
industry alliance 280-1 pasture 136, 138
policy involvement 299 creation 135-6, 137, 138
non-market values 257-64 plantation forestry 141, 144
extractive 258-60, 264 succession forest 143
non-extractive 258, 260-3, 264
preservation 258 odours 9
values 263-4 oil companies 377
sustainability 258 guidelines for exploration 377-8
non-timber forest products (NTFPs) 21 oil palm plantations 307
agrochemicals 36 oligarchic forests 38
Brazilian economy 26-7 Orchid Festival (Kew) 353-4
combined management with timber outreach vehicle 350
39 Overseas Development Administration
economic value 183 (UK) 372-3, 382
indigenous management 23-4 oxisols
local 255 deforested 135
outside Amazonian economy 28 nutrients 134, 135
pharmaceuticals 36
predatory extraction 32 Pacific Rim, forest products trade 204,
timber trees 39 246
non-timber functions 255 palm hearts 32, 34
capture 256-7 palms
demonstration 256 harvesting 35
markets 255-6 oligarchic forests 38
price 255 ornamental leaves 36
non-timber value over-exploitation 32
carbon storage 261-3 Panicum maximum 135, 137
ecological functions 261 paper products 201, 202
expression as household income 260 Parashorea malaanonan 51
importance 265 participation 289, 291-3
recreational 260-1 development 292
North America, forest products trade forest management 300
204,246 management 293
nutrients passive value of forests 155
above-ground biomass 66 pastoralism, nomadic 307
accumulation during regeneration pasture
159-61 abandonment 135
burning 130-2, 133 creation 135-6, 137, 138
clearance 132-5, 144, 146 denitrification 151
conservation 120 nutrients 136, 138
deficiency effects on seedling growth soil fertility 136
62 patch ecology 80
deforestation effects 128, 129, patronage politics 277
130 Paullinia cupana var. sorbilis (guarana)
fallow 145 28-9,36
leaf litter 130 Paumuri Indians 29
410 Index
permanent agriculture, deforestation policy 269-73
306,307 certification 298
pest control natural systems 175 challenges for sustain ability 279-82
pharmaceutical products 36, 259-60 changes289,290,299
phosphorus 128, 129, 130 civil society initiatives 296
burning 131, 132, 133 colonial 276
mycorrhizae 152-3 control of forests 276-7
pasture 136, 138 corporate resistance to change 292
slash-and-burn 138 decision-making cycles 294
photographic exhibitions 358 deforestation 310
photosynthesis devolution of power 294
maximum rate 47 economy stimulation 277
pioneer species 47 extra-sectorial influences 294
shade-tolerant species 47 failure 272
photosynthetic photon flux density 122, forest use 270
123 government control of forest lands
Phragmipedium besseae 354 276
phytochemical products 108-9 implementation 297
phytomass 128, 130 influences 274
pilocarpine 37 institutional constraints on
Pinus caribaea 141, 144 formulation 192
pioneer species 47 instruments 297-8
canopy disturbance 60 international initiatives 282, 284-5
irradiance level adaptation 47 leasing of forest areas 276
logging impact 59 market initiatives 296
seed dormancy 156 needs aspects 287
piranha 14 outcomes 287
Plan de Zonage (Cameroon) 88-9,90 participation 289, 291-3
planning, formal 278 political forces 372
plant powerful interest group representa-
indigenous management of resources tion 278
23-4 rain forest protection 392
litter removal 60 reform 387
size 6 replacement of forest 270
plantations 87, 307-8, 393 sustainable forest management 280,
export demand for crops 311 299
land use alteration 177 see also public policies
non-native species 164 political representation 292
privately owned 237 politics
resources 207 commitment to campaign 370
soil changes 141, 144 environmental questions 369
West Africa 81 pollination 9, 11
planting stock 50 pollution of rivers 15
Plants across the Curriculum 348-9 population density 184-5
plywood population growth 308-9
logs 221 deforestation regression models 317
synthetic panel substitution 220 forest cover simulation models 330
podzols 11 rate 184-5
Index 411
population pressure 308 reafforestation impact 232
deforestation 184-5 recreation 260-1
Poraqueiba sericea 39 use of forest 255
post-Earth Summit process 385, 386 values of forest 265
posters 356-7 regeneration 7-8, 43-8, 59-60
postgraduate education 351, 352 advance 154, 155, 157
potassium 128, 129, 130 below-ground processes 120
burning 131, 132 biomass accumulation 159-61
poverty 184, 309-10 clearance 103-4, 144, 146, 147
biomass needs 192 clearings 61
deforestation regression models 317 deforestation effects 120
energy needs 192 disturbance 46, 48
environmental degradation 185 environmental instability 43-6
power imbalances 294-5 growth rate 163
preservation values 258, 263-4 guilds 8
Princess of Wales Conservatory (Kew) human management 81
349,353,354 logging 59-60
private sector 276, 277, 377-9 natural 49-50
professional training 351-2 nitrogen conservation 149-50
profit nutrient accumulation 159-61
compulsory investment 393-4 patterns after deforestation 155,
equitable shares 39 161-3
short-term 234 plant competition 46-8
property rights 233 processes 154, 155
public concern 369-70, 385 requirements 163-4
public ownership 189 saplings 155
public policies 249 soil disturbance 162
domestic 234 species diversity 163
forest-based industrialization 236 suppression 62, 63
Pueraria spp. 125 regulatory mechanisms 278
pygmies 78, 79 reintroductions 361
relative humidity 122, 124
radiation balance 122 remote sensing imagery 194
radio 359-60 renewability of forest 99
rain forest see forest rent
rain-splash 64, 66, 127 capture 234-5
rainfall 1 distribution 234
erosion 64 economic 335
forest degradation 106-7 rent-seeking behaviour 234
root uptake 65 reptiles 10
run-off 60 reservoirs 308
Rainforest Action Network (RAN) 367 resettlement, forced 88
Rainforest Foundation 367 resilience 119
Rainforest Information Centre (RIC) resource-limited species 46
367 resource-sharing 299
Ramosmania rodriguesii (cafe marron resources
tree) 361 accounting 221,296
ratan cane 28 allocation 287
412 Index
availability 47 production 201
capturing at tree stages 47 Royal Botanic Gardens, Kew
control inequity 185 education programmes 348-55
global status 207, 208 information sheets 356
institutional mechanisms 190 mission statement 354-5
institutional resources for monitoring orchid festival 533-4
191 posters 356-7
maintenace of wild 81 Princess of Wales Conservatory 349,
management 300 353,354
plantations 207 public perception of role 364
requirements 46 titan arum 362-4
scarcity 207 royalties 37
small-scale patchiness in availability arrangements 234, 237
46 RTZ 378-9
sustainability 200 rubber 30
sustainable development xiii latex 23, 24, 109
temperate 207 plantations 307
tropical 207 sustainable extraction 36
unequal access 47 tree tapping 32
use by villagers 88-9 run-off rate 106
retailers, campaign involvment 376
revegetation strategies 154 Sainsbury's 376
revenue transfer schemes 244 sapling regeneration 155
ribereitos (Peru) 24 sawlogs 221
rights of access, expropriation xiii sawnwood 201, 202
rill erosion 127 price 204
Rio Declaration 347 Science, Engineering and Technology
see also Earth Summit (Rio) week 349
Rio Negro (Brazil) 14 seasonal rhythms 8-9
flvers sediment erosion 66
animal species 9 seed bank 158
dynamics 45 seedlings
pollution 15 germination after slash-and-burn
spate flow 14 agriculture 162
roads 4,52, 104 growth after logging 157, 159
access for settlement 310 loss 60
blocking 87 natural regeneration 49, 50
forest area relationship 323 survival after logging 49
state investment 310 seeds 9
World Bank support 382 dispersal 60, 157
roots dormancy 156
biomass 130 eating 9
growth 146-7 ecology after clearance 156-7, 158,
soil compaction 125 159
mat damage 146-7 fire effects 156-7
rosewood 32 germination 47,64,122,156
round tables, Canadian 293 heat treatment 156
roundwood 202 loss 60
Index 413
recalcitrant 49 compaction 53, 124-5
viability 49 cropped 140
selection deforestation 119-20, 121
dysgenic 59 biology effects 146-53
see also logging, selective physical property effects 124-6
semi-wood composites 221 disturbance 65
set aside 249 regeneration 162
sustainable forest management 238, fallow 141-2,143-4,145,146
239 fertility
settlement 310 after clearance 159-61
shade, seed dormancy 156 with cropping 133-4
shade-tolerant plants 47,154 index 144
sheet erosion 127 organic matter 138
shelterwood system 48 pasture 136
shifting cultivation 3, 4, 271 slash-and-burn 138, 140
deforestation 182, 305-6, 307 hydrology 126
successional regeneration 161 microorganisms
types 307 clearance 248
Shuar Indians (Ecuador) 22 deforestation 147-51
Sierra Club (USA) 380 nutrient
silvicultural systems 48-54, 55, 56 leaching 60, 61
damage levels 51-4, 55, 56 loss 66
extraction routes 52-3 organic matter
felling damage 53 breakdown 127
harvesting intensity 50-1 burning 130-2
post-logging tree mortality 54, 55, 56 clearance 132-5
skid tracks 52, 53, 62 deforestation effects 128,129, 130
slash material 134-5 pasture creation 135-6, 137, 138
slash-and-burn 4, 85,271 plantation forestry 141, 144
biomass accumulation 159, 160 slash-and-burn 137-8, 139, 140-1
environmental refugees 89 slope angle loss 127
fallow 140, 141, 142 temperature 122-4
phosphorus level 138 vegetation cover loss 127
soil effects 125, 13 7 see also topsoil
stages 137 soil-plant-nutrient relationships with
tree seedling germination 162 clearance 134
slope angle loss 127 solar radiation 46
sloths 14 Sophora toromiro 361
social collapse 255 species diversity, regeneration 163
social impact assessment 295-6 splash erosion 127
social value, genetic material 260 stakeholders 282-3, 286-7, 288, 289
socialization 84 map 288
society lineages 82 relations 289, 292
soil 2, 11 state
aggregates 125 forestry agencies 283
bare surface prefence of seedlings 162 intervention 310, 327
biological process interruption 120 see also government
chemical properties 128, 129, 130-5 stewardship, empowerment 296
414 Index
Stockholm Summit 347 implementation 241
storms 44 public 233-7, 249
story telling 360-1 political forces blocking 372
strategic planning, national 294 private returns 235, 238
stumpage promotion 223
fees 234, 238 regulations implementation 241
prices 249 revenue transfer 243
subsidiarity 289, 294-5 set aside 238, 239, 249
succession social returns 235, 238
nutrients 143 stakeholder
primary 60 map 288
tree species density 161, 162 perspectives 292
sulphur, burning 133 standards 248
surface roughness 106 tax transfers 243
surface-wash erosion 64 technological assistance 242
sustainability 32, 35-6 technological capacity improvement
extractive reserves 28-9 245
forest management 200 trade surcharge amounts 244
resources 200 transition costs 237, 240, 249
indigenous management 23-4 value-added processing industries
sustainable development 322 240
competitiveness relationship 379 voluntary commitments 375-6
forest resources xiii Swietenia mahogani 58
profitability relationship 379
Tropical Forestry Action Plan 382-3 tariff
sustainable forest management xiv, binding 216
199-200 escalation 209-10, 212
additional costs 239 tax 234, 236
certification 248 breaks 277
cost increases 249 evasion 275
developing countries 241 export 218, 244
enablement 299 transfer scheme 243
environmental value loss 233 VAT 243
export tax 244 taxonomic affinity 9, 11
financing 240-5 technological capacity improvement
policy options 242 245
global external benefits 241-2 temperate softwood market 219-20
implementation costs 230-40 temperature 1, 122
import surcharges 244 air 122-3
institutions 299 clearings 124
internalizing costs 237-40 increase with forest degradation 107
international agreement 281-2 soil 122-4
international competitiveness 245 terra firme forest 151-2
international issue 280 Theobroma cacao (cocoa) 28-9
international market access 246 Theobroma grandiflorum 37
lesser-used species 221-2 timber
non-wood product substitution 250 combined management with NTFPs
policy 280, 299 39
Index 415
consumption 203 regeneration cycle 8
cutting 3 size 6
demand 203 structure variety 5
exports 85, 203 water status 8
extraction 104 trespass 234
control 92 . Trochetiopsis erythroxylon 58
environmental effects 248 tropical forest see forest
unsustainable 231-3 Tropical Forest Action Plan xiv, 289,
fees 236 293,382-3
industry resistance to regulation 279 national level planning 386
preservation processes 218 tropical moist forest 100
production 201, 203 climatic zone 101
uses 203 estimates of cover 101-2
see also wood turtles 14
timber products
export restrictions 217 UCUODA community forest manage-
relative competitiveness 219-20 ment project 92-3
substitution 220 umari 39
temperate 219-20 UN Convention on Biological Diversity
tropical 219-20 386
Tiwai Island (Sierra Leone) 13-14 undergraduate studies 350-1
topic plantings, educational 353 understorey
topsoil biodiversity 232
chemical properties with drought stress adaptations of plants
burninglbulldozing 139 155-6
macroporosity loss 125 United Nations Conference on
removal and site heterogeneity 146 Environment and Development
trade 201, 203, 371-6 (UNCED) 346-7, 383-5, 386
barriers 207-22 unsustainable forest management, costs
bodies 372 248
economics 199-200 unsustainable timber extraction 231-3
international 193 access to forests 231
liberalization 388 damage during extraction 231, 232
restrictions 374 depletion of stocks 231
Trade Not Aid 357 environmental effects 231, 232
trees
commercial value 50-1 valuation
cover in landscape protection 111 forest assets 295
density 51 legitimate/illegitimate procedures 257
with succession 161, 162 value-added processing industries 240
dimensions/density in fallow 142 value-added products 206, 209
diversity 11 VAT exemption 243
lesser-used species 247 vegetation
life-cycles 45 cover loss 127
loss of large 56-9 deforestation 119-20, 121, 153-7,
non-timber 59 158,159-63
post-logging mortality 54, 55, 56 post-clearance ground-cover 125-6
rarity 51 regrowth 119
416 Index
veneer 221 panelling
vesicular-arbuscular mycorrhizae formaldehyde glue 218
(VAM) 152, 153 lesser-used species 221
videos 358 price 327
village product competition 247
clearing forest 81 recyclinglre-use 388
deep-forest 81-2 see also timber
territories 81 wood pulp 201
vmes bleaching 218
colonization of gaps 62, 63 lesser-used species 221
cutting prior to felling 53 wood-based composites
visual media 357-9 panels 201, 202
volcanic eruptions 44 substitution 220
vulture 14 World Bank 192, 369-70, 380
Environment Action Plans 393
wasps 14 policy reform 387
water stress 44 private sector activity facilitation 388
waters, Amazon Basin 14 restructuring 380-1
watershed protection loss 222 roads in Cameroon 382
weed invasions 179 safeguard introduction 380-1
weeding 164 scrutiny requirement 388
West Africa World Commission on Forests and
conservation 76-8 Sustainable Development
culture 78, 79 (WCFSD) xiii
ecology 77-8 World Environment Day 353, 355
forest settlements 79 World Forest Inventories 312
indigenous people 78-80 World Rainforest Movement (WRM)
legal systems 81 367,368,383
societies 78 World Resources Institute 192
white waters 14 World Trade Organization (WTO) 219
willingness-to-pay World Wide Fund for Nature (WWF)
annual 264 367
private 260 1995 Plus Group 375-6
windrowing 130, 131 voluntary commitment scheme 375-6
wood World Wide Web 358-9, 360
certified products 223
consumption levels 387-8 zoos 352, 353

Tropical Rainforest

Uploaded by

Copyright:

Available Formats

Tropical Rainforest

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tropical Rainforest

Uploaded by

Copyright:

Available Formats

C O N S E R V A T I O N B I O L O G Y SERIES

The aim of this Series is to provide major summaries of important topics in

1. Monitoring Butterflies for Ecology and Conservation

2. Insect Conservation Biology

3. Monitoring for Conservation and Ecology

4. Evaluation and Assessment for Conservation: Ecological Guidelines

5. Marine Protected Areas: Principles and Techniques for Management

6. Conservation of Faunal Diversity i n Forested Landscapes

7. Ecology and Conservation of Amphibians

8. Conservation and the Use of Wildlife Resources

9. Conservation Management of Freshwater Habitats

10. Tropical Rain Forests: A Wider Perspective

11. Valuation of Costs and Benefits of Wildlife i n Africa

12. Primate Conservation Biology

Wetland Ecology and Management

Grazing as a Tool i n Conservation Management

© 1998 Springer Science+Business M e d i a Dordrecht

Thomson Science is a division of International Thomson Publishing I(T)P

Typeset in 10/12pt Sabon by Columns Design Ltd, Reading

ISBN 978-0-412-81520-1 I S B N 978-94-011-4912-9 (eBook)

Library of Congress Catalog Card Number: 97-69618

1 Tropical rain forests - what are they really like? 1

2 Indigenous non-timber benefits from tropical rain forest 21

3 Degeneration versus regeneration - logging in tropical rain forests 43

4 Forest people and conservation initiatives: the cultural

5 Forest and environmental degradation 99

6 Soil and vegetation effects of tropical deforestation 119

7 Causes of tropical deforestation and institutional constraints

8 The economics of the tropical timber trade and sustainable

9 Can non-market values save the tropical forests? 255

10 The role of policy and institutions 269

12 Communicating the message: a case study from the Royal

13 Effective campaigning 367

14.6 Knowledge base 393

List of acronyms 395

The international perspective for this book is the unprecedented level of

influenced by more powerful short-term sectional interests. There is

Hardly a week passes without a new report of threats to another area of

1.1 WHERE ARE RAIN FORESTS?

1.2 HOW FAST IS RAIN FOREST DISAPPEARING?

1.5 SO WHAT ARE THEY REALLY LIKE?

1996). Secondly, attention is drawn to the size of the trees: 46-55 m

Figure 1.3 A range of features characteristic of rain forest.

invest in so much wood (capital - or is it essential infrastructure?). The

1.5.3 Seasonal rhythms

Mean Number Number Number Number

Lowland rain forest

1.6 WHY SO MUCH DIVERSITY?

1.7 HOW SERIOUS IS FRAGMENTATION?

1.8 QUANTITATIVE STUDIES IN RAIN FOREST

1.9 CASE STUDY: TIWAI ISLAND

1.10 SKETCH FROM THE AMAZON

1.11 OTHER REGIONAL STUDIES

2.2 INDIGENOUS USE OF THE FOREST

Construction 191 77.0 637 87.4

Use category Number of species

2.3 INDIGENOUS MANAGEMENT OF PLANT RESOURCES

2.4 EXTRACTIVE RESERVES

Cleaned fruit zone: -700 m2

Name State Area (ha) Population Principal products