Managing the World's Forests Looking For Balance Between Conservation and Development Ed i t e d b y N a r e n d r a P. S h a r m a Managing the World's Forests Looking for Balance Between Conservation and · Development edited by N arendra P. Sharma Principal Economist, The World Bank KENDALL/HUNT PUBLISHING COMPANY 2460 Kerper Boulevard P.O. Box ·539 Dubuque, Iowa 52004-0539 The opinions and recommendations presented in this book are the result of research supported by the World Bank; however, they are solely those of the respective authors, and do not necessarily represent the position or policies of the World Bank, its affiliated organizations, members of its Board of Executive Directors, or the countries they represent. Cover design by Melinda Russell © 1992 by the International Bank for Reconstruction and Development Library of Congress Catalog Card Number 92-54402 ISBN 0-8403-7885-8 All .rights reserved. No part of this publication may be reproduced stored in a retrieval system, or· transmitted, in any form, or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright owner. Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 Dedicated to the World's Forests Printed on Recycled Paper recycled paper Contents Page Foreword ................................................. i Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv 1 Introduction Narendra P. Sharma 1 About This Book Four Questions: Why Are Forests and Trees Important? What Is the Problem? What Are the Causes of the Problem? How Should This Problem Be Addressed? PARTI. BACKGROUND 2 World Forests in Perspective Narendra P. Sharma, Raymond Rowe, Keith Openshaw, and Michael Jacobson ................ 17 World Forest Resources Role of Forests in Natural Systems Role of Forests in Economic Development World Trade in Forest Products Tariffs and Duties on Forest Products Future Demand for Wood Products Summary iii iv MANAGING THE WORLD'S FORESTS 3 Deforestation: Problems, Causes, and Concerns Raymond Rowe, Narendra P. Sharma, and John Browder .................................... 33 Dimensions of the Deforestation Problem Direct Causes Underlying Causes Concerns: Economic, Social, and Environmental Summary PART II. GLOBAL ENVIRONMENTAL CONCERNS 4 Biological Diversity and Forests Daniel B. Botkin and Lee M. Talbot ................... .47 Biological Diversity The Fundamental Reasons to Conserve Biological Diversity The Relationship Between Forests and Biological Diversity, the Trends in Biological Diversity Related to Forests, and the Causes of the Current Situation The Relationship Between Forest Management Techniques, Including Commercial Logging, and Biological Diversity Measures Used or Proposed for Conserving Biological Diversity of Forests 5 The Role of Forests in Climate Change George Woodwell .................................... 75 Factors That Determine the Composition of the Atmosphere Deforestation Changes in Metabolism What Can the International Community Do? Summary PART Ill. POLICY AND ECONOMIC ISSUES 6 Forest-Based Industrialization: A Dynamic Perspective Jeffrey R. Vincent and Clark S. Binkley ................. 93 A Dynamic Theory of Forest-Sector Development Case Studies: Peninsular Malaysia, Ghana, and Chile Summary and Policy Recommendations Annex: Three Scenarios for Forest-Sector Development CONTENTS v 7 Forest Concession Management and Revenue Policies Malcolm Gillis ...................................... 139 Commonalities Current Concession Policies Forest Revenue Systems Conclusions: Options for Reform of Forest Revenue Systems 8 Fuelwood Problems and Solutions D. Evan Mercer and John Soussan .................... 177 The Nature of Fuelwood Problems Fuelwood Policy Analysis Policy Recommendations Summary Annex: A Brief Typology of Fuelwood Situations 9 Policies Contributing to Agricultural Colonization of Latin America's Tropical Forests Douglas Southgate ................................... 215 The Brazilian Amazon: Infrastructure Development, Subsidies, and Tenurial Crisis Ecuador: Tenurial Crisis, Price Distortions, and Weak Agricultural Research and Extension Guatemala: Intense Demographic Pressure Conclusion: The Challenge of Policy Reform 10 Forest Valuation Randall Kramer, Robert Healy, and Robert Mendelsohn .................................. 237 Forest Products and Outputs Measurement of Forest Values Conclusions and Recommendations Summary PART IV. SOCIAL AND INSTITUTIONAL ISSUES 11 Forestry Institutions Marie Lynn Miranda, Olga Marta Corrales, Michael Regan, and William Ascher ................... 269 Definition of the Problem The Colombian Case Study The Honduran Case Study The Malaysian Case Study General Conclusions vi MANAGING THE WORLD'S FORESTS 12 A Sociological Framework: Policy, Environment, and the Social Actors for Tree Planting Michael Cernea ...................................... 301 Environmental Problems and Social Organization The Sociological Perspective on Tree Planting Social Forestry: A People-Oriented Strategy Designing Strategies Around Social Actors Ill-Defined Social Actors and the Failure of Village Woodlots Group-Centered Strategies: Using Alternative Units of Social Organization Summary PART V. FOREST RESOURCE MANAGEMENT 13 The Management of Natural Forests John Palmer and Timothy J. Synnott .................. 337 Extent of and Variation in Natural Tropical Forests Goods and Services from the Natural Forests The Justification for Forest Management Aspects of the Goal of Sustainability Management Is Not Just Harvesting Major Conditions for Sustainable Management Recommendations to the Donor Community 14 Plantation Forestry P.J. Kanowski and P.S. Savill, with P.G. Adlard, J. Burley, J. Evans, J.R. Palmer, and P.J. Wood ............ 375 Introduction Review of Best Management Strategies Financial Characteristics of Trees as Crops The Policy Environment Land Use and Management Planning Elements of Forest Management Review of Environmental Impacts Constraints, Opportunities, and Priorities Summary and Conclusions CONTENTS vii 15 Agroforestry Systems Design: An Ecozone Approach P.K.R. Nair ......................................... 403 The Approach to Agroforestry Agroforestry Practices An Ecozone Approach to Agroforestry Design Summary 16 Production of Forest Products in Agricultural and Common Land Systems: Economic and Policy Issues J.E.M. Arnold ....................................... 433 Forest Products and the Rural Household Economy Common Property Resources Trees in Farming Systems Summary 17 Watershed Management: A Key to Sustainability Kenneth Brooks, Hans Gregersen, Peter Ffolliott, and K.G. Tejwani ..................... .455 Concepts and Definitions Background for Policy Formulation Causes of Watershed Problems A Watershed Management Approach Analysis and Establishment of Priorities Barriers to Adoption of Watershed Management Overcoming the Barriers Conclusions and Recommendations PART VI. TRANSITION TO CONSERVATION AND SUSTAINABILITY 18 Conditions for Sustainable Development Narendra P. Sharma, Raymond Rowe, Mikael Grut, Randall Kramer, and Hans Gregersen ................ .489 Merging Conservation and Development Objectives Balancing Use with Availability of Resources Need for Collaborative Effort Summary viii MANAGING THE WORLD'S FORESTS 19 A Global Perspective on Forest Policy Narendra P. Sharma, Clark Binkley, and Jeffrey Burley ........... , , . , .................. 515 Current Policy: Need for Change Policy Imperatives International Cooperation Conclusions List of Contributors . , , , .......... , , , ......... , , , ......... . 527 Statistical Appendix . , , , ........ , . , ........................ 535 LIST OF EXHIBITS Color Plate 1. World Forest Zones Color Plate 2. Tropical Deforestation 2-1. Forest Area in Developing Countries 2-2. World Forest Distribution, by Type 2-3. Forest Distribution, by Region 2-4. Forest and Woodlands Distribution, by Region 2-5. Production, Export, Import, and Consumption of Wood Products for Developing Countries and the World, 1986-1988 2-6. Average Unit Export Price of Selected Forest Products for Developing and Developed Countries, 1986-1988 2-7. Average Tariffs for Selected Wood Products of Developed and Developing Countries, 1982 2-8. Current and Forecast Demand for Wood Products in Developing and Developed Countries, 1987, 2000, and 2025 5-1. Global Flows of Carbon 5-2. The Net Annual Flux of Carbon from Deforestation in Tropical and Temperate Zones Globally from 1850 to 1980 6-1. Index of Roundwood Production 6-2. Roundwood Consumption Share 6-3. Net Exports Relative to GDP 6-4. Transition from an Old-Growth to a Second-Growth Timber Economy 6-5. Timber Processing in an Open Economy 6-6. Index of Roundwood Production 6-7. Industrial Roundwood Consumption Share of Production, 1960-1987 INTRODUCTION ix 7-1. Illustrative License Fees 7-2. Reforestation Fees 7-3. Export Taxes on Timber, Selected Countries, 1989 7-4. Rent Distribution Under Inefficient Forest Revenue Systems and High Protection to Domestic Processing 7-5. Potential Rent and Government Rent Capture in Tropical Timber Production, Selected Periods 8-1 . Amount of Kerosene Required to Replace All Urban Fuelwood Consumption in Low- and Middle-Income African Countries, and Resulting Increase in Demand for Petroleum and Decrease in Demand for Fuelwood, Selected Years, 1978-1989 14-1. Productivity (Annual Wood Increment of Various Well-Managed Types of Forests) 1 5-1. Main Characteristics of the Most Common Agroforestry Practice in the Tropics 15-2. Agroforestry Interventions Recommended for Major Ecological Zones of the Tropics 15-3. Agroforestry I Agroecological Matrix for the Tropics and Subtropics 16-1 . Some Common Nutrition Problems and the Potential Role of Forest Food 16-2. Measures Adopted by Different Groups in the Face of Decline in Area, Productivity, and Management Systems of Common Property Resources in Dry Regions of India 16-3. Contexts in Which Woodland Management and Tree Planting Occur in the Semiarid Regions of Africa 17-1. Water-Related Problems, Their Causes, and Some Solutions Using a Watershed Approach 17-2. Relationships Between Physical Effects, Environmental Changes, and Downstream Benefits from Watershed Management Practices, as Compared with the ''Without" Practices Condition 17-3. Examples of Watershed Management Practices to Be Considered in Designing a Project, Grouped According to Effects as Illustrated in Exhibit 17-2, and as Described by Gil (1979) 17-4. Examples of Watershed Management Practices to Be Considered in Designing a Project, Grouped According to Location Within a Watershed 18-1. Actions Needed at the Local, National, and Global Levels Foreword We have entered the 1990s under a cloud of concern about depletion of natural resources and degradation of the environment. Saving the remain- ing forests in both tropical and temperate areas has become an important priority of the world community. Sustainable development of forest resources will contribute to economic and environmental security throughout the world. To protect natural systems and to improve human welfare, nations must stabilize existing forests and plant trees. By establishing workable manage- ment systems that are also socially and ecologically acceptable, nations can use forest resources efficiently. Although many nations have already taken steps to achieve sustainable development of forests, more needs to be done. The world community must find a consensus on the appropriate actions to address complex forestry issues. The U.N. Conference on Environment and Development in June 1992 in Brazil represented one opportunity to move toward this objective. The World Bank-as articulated in its new for- est policy-will support global initiatives promoting conservation and sus- tainable development of forest resources. This book provides a timely assessment of the world forestry situation from different perspectives. It offers policy options to decision makers, re- flecting local, national, and global needs. In addition, we expect the book to appeal to a broad audience, enhance people's understanding of forestry issues, and encourage dialogue among people at all levels. VISVANATHAN RAJAGOPALAN Vice President Sector and Operations Policy The World Bank xi Preface The loss of large expanses of forests, especially in the tropics, poses a serious threat to human welfare and the global environment. The increasing rate of deforestation has now become a serious concern of the world com- munity. People everywhere want natural forests in both tropical and tem- perate regions to be protected. And nations must manage forest resources more efficiently to benefit present and future generations. Because of the serious local, national, and global consequences of de- structive deforestation, the World Bank undertook a comprehensive study in mid-1989 to review its existing forestry policies and projects, to assess the forestry situation in a global context, and to define a new forest policy to guide its future operations in the sector. The study, based on an interdisciplinary approach, used expertise asso- ciated with forestry activities and development from within and outside the Bank. Academic institutions, bilateral and multilateral agencies, nongovernment organizations, and researchers-all were invited to discuss policy, institutional, and technical issues and possible solutions for sustain- able use of forest resources. These groups outside the Bank developed nu- merous working papers that provided valuable input for developing the Bank's approach to forestry. The Bank has recently issued its new forest policy, reiterating its com- mitment to helping developing countries to move toward conservation and sustainable development of forest resources. Another important product of the Bank's initiative is this book, which represents diverse views of many authors from social, physical, and biological sciences. The book provides different perspectives on the world forestry situation and options for development. xiii xiv MANAGING THE WORLD'S FORESTS The world community now faces two important challenges: The first challenge is to arrest destructive deforestation and to manage existing for- ests on a sustainable basis. The second challenge is to increase forest re- sources through reforestation and afforestation. These challenges call for appropriate actions and a participatory approach based on local needs, na- tional priorities, and international cooperation. This book addresses these challenges and contributes to the existing knowledge and understanding of forestry issues, including the role of for- ests in natural systems and in economic development. It is hoped that this book will stimulate global dialogue and lead to consensus among people for affirmative actions supporting conservation and development goals. MICHEL J. PETIT Director Agriculture and Rural Development The World Bank Acknowledgments This book is a product of a long process of research, analysis, and dis- cussion. Many people-from within and outside the World Bank-have con- tributed to the book through studies, briefs, and comments. At the World Bank, I extend my appreciation to V. Rajagopalan, Michel Petit, Shawki Barghouti, and Gershon Feder for their encouragement and support to complete this volume. The operational staff of the Bank, espe- cially foresters and environmentalists, provided useful suggestions and com- ments on various studies, issues, and options for development. I am particularly indebted to Raymond Rowe, who provided inspira- tion and objectivity to this study. As a forestry expert, he carefully assessed forestry issues and advocated a pragmatic approach to forestry manage- ment, balancing conservation and development goals. As a friend, he gave support and advice during difficult periods in completing this study. The 37 authors who collaborated on this book brought diverse views and concerns about the current world forestry situation to the exercise. They worked as a team to address complex issues and to propose appropri- ate solutions for better use of forest resources. The contributions of the Food and Agricultural Organization and some bilateral agencies helped in assess- ing development priorities and investment needs. Several nongovernment organizations gave written comments and participated in meetings. Their involvement generated good ideas and influenced some of the policy rec- ommendations outlined in this book. At the early stages of this work, Randall Kramer of Duke University contributed many valuable suggestions that led to various supporting stud- ies. Other people--Lyn Maguire, Margaret McKean, Daniel Richter, John Terborgh, Mimi Becker, and George Dutrow of Duke University; Robert Kellison and Jan Laarman of North Carolina State University; and Nils Chatterjee, Andrew Johns, and Norman Myers, independent consultants- shed light on sociopolitical and ecological considerations, as well as on glo- bal forestry issues. xv xvi ACKNOWLEDGMENTS I extend my appreciation to Priscilla S. Taylor, whose valuable editorial advice and assistance made the book more readable for a wider audience. Special thanks also are due to Michael Jacobson for preparing the statistical appendix and to Rhonda Thomas Benson for typing the manuscript. Finally, I extend my sincere gratitude to my wife, Martha Sharma, a professional geographer, for her valuable insight and perspective on the role of forests in natural systems. The views expressed in this book are those of the authors and do not necessarily represent the views and policies of the World Bank. NARENDRA P. SHARMA Agriculture and Rural Development The World Bank Glossary Agroforestry. Land-use system in which woody perennials are used on the same land as agricultural crops or animals, in some form of spatial arrangement or temporal sequence. Biological diversity or Biodiversity. Diversity of species of plants and ani- mals in an ecosystem. Biomass. Total quantity of living tissue per tree, per unit of area, or in a plant community. Biome. Major ecological community type (e.g., grassland). Biota. Flora and fauna of a region. Broad-leaved species. Also known as hardwood, nonconiferous, or angio- sperm species. This is a botanical classification, based on structure of seed and wood, and not all broad-leaved species have broad leaves. Trees are either broad-leaved or coniferous. Clear-felling or Clear-cutting. Removal of the entire standing crop. Closed forest. Stand density is greater than 20 percent and tree crowns approach general contact with one another. Common property resource. Nonprivate property such as communal land. Can have open or limited access. Conifer or coniferous species. Softwood or gymnosperm species, distinct from broad-leaved (hardwood) species; mostly evergreen. Conservation. Production of goods and services with protection of the en- vironment and the resource base. Conversion forest. Assigned for conversion to agriculture or other nonforest use. Desertification. Degradation of land that ultimately leads to desert-like conditions. Designated forest. Legally set aside for preservation or production. xvii xviii GLOSSARY Dipterocarp forest. Forests in Southeast Asia dominated by trees of the Dipterocarpaceae family. Ecosystem. Complex of living organisms and their environment. Exotic species. Introduced from another ecozone; opposite of indigenous. Farm forestry. People-oriented and carried out on private farmlands. Re- lated but not synonymous terms: agroforestry, community forestry, and social fores try. Forest. (a) Noun (ecological sense of term): A plant community predomi- nantly of trees and other woody vegetation, growing more or less closely together. (b) Verb: to cover with trees; the term includes both afforest and reforest. Forestation. Umbrella term for afforestation and reforestation. Forest concession. Lease or contract for use of forest. Forest fallow (Bush fallow). Woody vegetation resulting from shifting cul- tivation. Forestry. Science, business, and art of creating, conserving, and managing forests and forest lands for the continuing use of their resources, mate- rial or other. Global warming. Rise in temperature of the earth due to increasing emis- sions of greenhouse gases; deforestation is said to be one of the con- tributors. Growing stock. Standing volume of a forest area. Habitat. Place or type of site where organism naturally lives. Hardwood. Conventional term for the wood of broad-leaved trees, and for the trees themselves. Hardwood is thus a botanical category, and the wood may actually be very soft (e.g., balsa wood). Industrial plantation. For the production of forest products other than fuelwood, e.g., sawlogs, veneer logs, pulpwood, poles and pitprops, and wattle bark. Industrial forestry. Industrial plantations and forest industry. Industrial roundwood. Rcmndwood other than fuelwood. Intact forest (Primary or virgin forests). Essentially unmodified by human activity for the past 60 to 80 years. Logging. Felling and extraction of wood, especially as logs. Log pond. Temporary storage area for logs before conversion to sawnwoods. Monocyclic system. Principal fellings on a given area occur only once dur- ing a rotation. GLOSSARY xix National park. Natural area protected for scientific, educational, and recre- ational use. Natural regeneration. Renewal of a tree crop by natural means, as opposed to artificial regeneration by means of planting or sowing. Nature reserve. Area of ecosystem protected for its fauna and flora. Nonwood products. Tangible minor forest products, such as fruit, nuts, and bush meat (edible meat of wild animals). Open forest. Tree canopy layer is discontinuous but covers at least 10 per- cent of the area. Grass layer is continuous, allowing grazing and spread- ing of fires. Other wooded area (Woodland). Has some forest characteristics but is not forest as defined under closed forest or open forest. Includes areas oc- cupied by windbreaks, groups of trees, fallow, and shrublands. Overgrazing. Grazing exceeding the capacity of the site and hence leading to soil erosion. Plantation. Forest crop established by sowing or planting. Polycyclic system. Principal fellings on a given area occur more than once during a rotation. Preservation forest. Designated for total protection of representative forest ecosystems in which all forms of extraction are prohibited. Primary forest. Either essentially unmodified by human activity (intact) or modified only by the hunting and gathering activities of indigenous people. Production forest. Designated for sustainable production of forest prod- ucts. Protection forest. Designated for stabilization of mountain slopes, upland watersheds, fragile lands, reservoirs, and catchment areas. Controlled sustainable extraction of nonwood products could be allowed. Reserve. In this book, and in general U.S. practice, this term is synonymous with preserve, (i.e., land set aside for total protection). However, in the tropics the term forest reserve (Fr.: foret classee) generally means a log- ging reserve that has been designated to be under forest cover forever (U.S. term: national forest). Rotation. Planned number of years between establishment of a forest stand and its felling. Roundwood. Wood in the rough or natural state. Made up of industrial roundwood and fuelwood. Sawlogs. Extracted logs for sawing into lumber. Sawnwood. Lumber. xx GLOSSARY Secondary forest. Subjected to a light cycle of shifting cultivation or to various intensities of logging, but still containing indigenous trees or shrubs. Selection felling or cutting. Harvesting of only a small proportion of the standing crop; opposite of clear-felling. Shelterbelt. Strip of trees providing shelter, generally from wind. Shelterwood system. Old crop is removed in two or more fellings to en- courage regeneration in its shelter. Shifted cultivator. Cultivator who has moved into forest areas. Shifting cultivation (Swidden, Taungya). Farming system in which land is periodically cleared, farmed, and then returned to fallow; synonymous with slash-and-burn or swidden agriculture. Shrubland (Brushland or scrubland). Shrubs or stunted trees cover more than 20 percent of the area; not used primarily for agricultural or other nonforestry purposes, such as grazing of domestic animals. Silviculture. Science and art of cultivating forest crops. Social forestry (Regreening). Farm forestry and community forestry. Softwood. Coniferous wood or conifers. Stumpage or royalty. Fee or price of standing trees before logging. Sustainable development. Meets present needs without compromising the ability of future generations to meet their own needs. Sustainable management. Supports multiple uses (including biodiversity preservation, timber harvesting, extraction of bush meat and other nonwood products, soil and water conservation, tourism, recreation, and enjoyment of natural amenities) based on an ecosystem concept that allows utilization of forests without undermining their use by present and future generations. Different systems of management would be required for each category of forest depending on the intended output. Sustained yield. Yield that a forest can produce continuously at a given intensity of management, hence "sustained yield management." The latter term may be further defined according to the production objective (e.g., water, logs, biodiversity, or bush meat). Tree tenure. Ownership of trees or their usufruct. Tropical dry forest (TDF). Open forest with continuous grass cover; distin- guished from other tropical forests by distinct seasonality and low rain- fall. Includes woody savannas and shrubland. Tropical moist forest (TMF). Situated in areas receiving not less than 100 mm of rain in any month for two out of three years, with a mean annual temperature of 24°C or higher; mostly low-lying, generally closed. Sub- divided into tropical rain forest and tropical moist deciduous forest. GLOSSARY xxi Tropical rain forest. Evergreen, situated in the more humid and low-lying parts of the tropical moist forest zone, and also known as tropical hu- mid forest. The most species-rich part of the tropical moist forest. Con- tains valuable timbers generally growing on poor soils. Underinvoicing of logs. Endemic practice whereby firms obtain tropical hardwood logs from their concessions in the producer country for a fraction of their true value. Understory. Lower levels of shade-tolerant trees in a dense forest. Wasteland. Land that is degraded and unproductive. Watershed. Area draining ultimately to a particular watercourse or body of water. Wood-based panels. Includes veneer, plywood, particle board, and fiber- board. Sometimes called panel products. Woodfuel. Fuelwood and charcoal. 1 Introduction Narendra P. Sharma F orests are a valuable environmental and economic resource for support- ing natural systems and for improving human welfare. Human activities have always modified the forest environment, but in recent years the inten- sity and scale of forest use have increased significantly. Everyone has ben- efited when people have treated forests as renewable resources, protected them to preserve biodiversity, or transformed them to support other eco- nomic activities on a sustainable basis. Conversely, destructive exploitation of forests has caused serious economic, social, and environmental losses. , In recent years, issues relating to forestry have become more complex, and the status of forests is now a subject of worldwide debate. Scores of publications have raised concerns about the demise of world forests, espe- cially in tropical areas, and people throughout the world are increasingly demanding protection of natural forests. At the same time, there are strong differences of opinion among people, as well as among nations, about how forests should be used and managed to support conservation and develop- ment goals. The world community today faces the challenge of achieving a balance between development and maintenance of natural systems and thereby ensuring the integrity and stability of forest ecosystems. People can reverse the tide of destructive deforestation, stabilize the forests of the world, and increase forest resources. This book provides numerous perspectives on world forests and on the political economy of forest management. Authors with broad experience in many fields present diverse views on key forestry issues and solutions needed to accelerate the transition to sustainable use of forest resources. Some of the chapters, in fact, present opposing views, reflecting different perspec- 1 2 MANAGING THE WORLD'S FORESTS lives and opinions. Each study is the outcome of individual research and analysis, discussion among the authors, and dialogue between the authors and representatives from the nongovernment organizations, the donor com- munity, the private sector, and governments. ABOUT THIS BOOK This chapter begins with a brief overview of the other chapters in this book. The subject of managing the world's forests is then examined here in terms of four questions: (1) Why are the forests and trees important? (2) What is the problem? (3) What are the causes of the problem? and (4) How should this problem be addressed? The rest of this volume is divided into six parts. Part I provides background on the world forest situation. Chapter 2 fby Narendra P. Sharma, Raymond Rowe, Keith Openshaw, and Michael Jacobson) describes the world's forest resources: the types and distribu- tion of forests and the different uses made of the forests in developed and developing countries. It also describes the role of the forests in natural sys- tems and in economic development, and the world trade in forest products. Chapter 3 (by Raymond Rowe, Narendra P. Sharma, and John Browder) outlines the dimensions of the deforestation problem and compares the direct causes (agricultural expansion, overgrazing, fuelwood gathering, com- mercial logging, and infrastructure and industrial development) with the underlying causes (market and policy failures, population growth and rural poverty, and the burden of external debt). The chapter ends with an assess- ment of the serious social, economic, and environmental costs associated with destructive deforestation. Global environmental concerns (potential climate change and loss of biodiversity associated with deforestation) are introduced in part IL Chap- ter 4 (by Daniel B. Botkin and Lee M. Talbot) provides an ecological per- spective to natural forest management. It discusses the importance of bio- logical resources and the effects that forest management techniques and changes in land use have had on biological diversity in the past and might have in the future. It assesses the current rate of change in biological diver- sity, outlines the criteria for sustainability of forest ecosystems, and sug- gests ways in which the international community can help to protect more natural forests, particularly tropical moist forests. The authors present the underlying rationale for conserving biological diversity (once lost, a species cannot be regained), as well as the utilitarian, aesthetic, moral, and ecological justifications. They note that of the three general categories of measures for conserving biological diversity of for- ests-protection of natural or near-natural ecosystems, restoration and reha- bilitation of degraded lands, and ex situ protection of individual species- the first is by far the most important. Finally, the chapter presents a new approach to management of natural forests as biological conservation areas. INTRODUCTION 3 Chapter 5 (by George Woodwell) details the role of forests in climatic change. After analyzing the factors that determine the composition of the atmosphere, Woodwell notes that climatologists have grossly underestimated the effects of forests on climate and that the most conspicuous, short-term effect of forests globally is the extent of their control over the carbon dioxide content of the atmosphere. According to Woodwell, deforestation results in the release of carbon into the atmosphere thus contributing to global warm- ing. He addresses a number of questions: whether global warming will produce greater areas of forests; what can be done to slow the accumulation of heat-trapping gases; whether reforestation can be used to remove carbon from the atmosphere at least temporarily; and what steps must be taken to stabilize the composition of the atmosphere. The chapter concludes with some recommendations for action by the international community to stabi- lize forests and to increase reforestation. Parts III and IV focus on policy and institutional issues, as well as on forest valuation and the social dimension of forestry. Chapter 6 (by Jeffrey R. Vincent and Claude S. Binkley) considers whether industrialization based on the forest resource still has a role to play in economic development. The authors believe that many of the past environmental and economic prob- lems can be blamed on government policies that tended to foster an un- stable macroeconomic environment, to keep wood artificially cheap, and to direct investment toward inefficient processing industries. The results have been depletion of forests, inadequate management of natural forests, limited establishment of. plantations, and underused processing capacity. The au- thors argue that with the right policies, forest-based industrialization can become an important source of employment and income, and can promote conservation by enabling forests to outcompete alternative land uses. The chapter explores these issues by examining forest-based industrialization over several decades in Peninsular Malaysia, Ghana, and Chile, and out- lines policy recommendations. Chapter 7 (by Malcolm Gillis) focuses on how government forestry poli- cies provide incentives for rapid exploitation of forest resources and devel- opment of inefficient wood industries. Gillis details current concession poli- cies and forest revenue systems in selected countries in Africa, Asia, and Latin America. He concludes by offering options for reforms of tropical forest revenue systems that vary with administrative strengths and weak- nesses and with the quality of information about forest inventories. All the proposed reform programs aim to capture a greater share of rents for the forest owner, to correct serious underpricing of forest resources, to remove strong incentives for inefficiency in the use of the wood resources of the forest, and to establish forest fees high enough to offset the adverse environ- mental effects of logging. Chapter 8 (by D. Evan Mercer and John Soussan) analyzes the nature and origins of fuelwood problems and past policy approaches to solving them. Fuelwood problems are viewed primarily as one consequence of the . 4 MANAGING THE WORLD'S FORESTS interaction of environmental and economic forces at the local level, which results in numerous resource stresses. The authors state that fuelwood prob- lems and policies should be seen as part of a wider land resource manage- ment policy. They point out that just as there is no one fuelwood problem, there are many potential solutions, and the key is to identify what will work where and why. The authors add that fuelwood problems and intervention opportunities are highly variable and specific to localities. The authors suggest means for tackling fuelwood problems but claim that technical interventions will prove ineffective without the introduction of policy reforms to permit effective markets to operate, to produce prices that reflect real costs, to change land-tenure arrangements, and to promote community participation. The authors conclude that policy reforms may enable markets to solve fuelwood allocation and production problems with- out further public-sector involvement. In other cases, public-sector interven- tion may be required in the short to medium term to assist in this transition. Chapter 9 (by Douglas Southgate) describes the policy environment responsible for excessive deforestation through Latin America, focusing on three countries. In Brazil, substantial agricultural colonization along many of the roads penetrating the Amazon Basin is blamed partly on the subsidi- zation of land clearing, partly on the fact that tax and other policies have made land elsewhere in the country more expensive, and partly on the multifaceted tenurial crisis. The same crisis affects Ecuador, where inad- equate investment in the scientific base underpinning agriculture and other sectors of the rural economy also results in excessive frontier expansion. The role of intense demographic pressure is illustrated by the case of Gua- temala. The chapter closes with some suggestions for appropriate govern- mental response to increasing land scarcity: devolving control over forests, eliminating subsidies for land-use conversion, eliminating policies that arti- ficially enhance hold values in nonfrontier areas, planning and evaluating infrastructure projects more carefully, and accelerating formation of nonland assets in the agricultural sector. Chapter 10 (by Randall Kramer, Robert Healy, and Robert Mendelsohn) describes the importance of proper forest valuation to assess the value of forests, the appraisal of forestry projects, and other projects that affect for- ests, and describes the reasons why forests are often misvalued. The authors point out that forests provide many economic and environmental goods and services. Some of these forest goods are traded in formal and informal markets. But a number of other environmental goods and services, such as biological resources, watershed services, and carbon sequestering, are like public goods for which markets do not exist. Consequently, many noncom- mercial forest products are not valued, or they are misvalued. Moreover, many of the commercial products of forests are misvalued because of mar- ket distortions, such as the existence of externalities, market power, and inappropriate government intervention. The simultaneous production of many products and the lack of information on forest products also result in misvaluation of forests. INTRODUCTION 5 According to the authors, proper valuation of forests is important to the appraisal of investment programs relating to forests and to the development of forests and nonforest policies. Undervaluation of forests can bias policy decisions related to resource allocation, can result in underestimation of the contribution of forests to the national economy, and can result in acceler- ated depletion of forest resources. The authors state that valuation tech- niques are available and applicable to both market and nonmarket forest products. However, they recommend that the donor community take the lead in promoting the refinement and application of valuation techniques for nonmarket goods and services, especially in developing countries, so that planners can formulate effective policies, choose among projects, allo- cate funds for research in forest management, and implement better forest- management systems. Chapter 11 (by Marie Lynn Miranda, Olga Marta Corrales, Michael Regan, and William Ascher) discusses problems plaguing forestry institutions and examines new approaches to forest management in Colombia, Honduras, and Malaysia. These approaches include financing arrangements that pro- vide partial or full administrative funding for the forestry institution di- rectly from forestry operations. The authors note that the new institutional designs have performed reasonably well in generating revenue and captur- ing rent but have taken insufficient account of the diverse environmental and social services provided by the forest. The authors also develop an analytical framework that attempts to link institutional structure to perfor- mance in revenue generation, social forestry, and environmental services. To improve the performance of forestry institutions in forest management, the authors suggest the involvement of local people, provision of appropri- ate financial incentives to both local and government officials, provision of mechanisms to deal with conflicts and complementarities among multiple objectives in forest management, and interagency coordination. Chapter 12 (by Michael Cernea) is based on the premise that it is just as important to address the sociocultural issues involved in tree planting and forest management as it is to resolve the economic and technical issues. The author first discusses the centrality of people, rather than commodities, in formulating environmentally sound development policies. He then de- fines the social actors of deforestation and the failure of markets, the goals of social forestry as a people-oriented strategy, and the need to design it around well-identified social actors. Cernea argues that the ultimate success of any reforestation strategy depends on the social forces that can be sum- moned to accomplish the task. He argues for forestry strategies that go beyond conservation of the remaining forests to trigger massive additional tree planting by the people. He also analyzes the sociological reasons why community woodlot projects have largely failed. He closes by outlining two types of actor-centered strategies in tree planting: family-centered and group- centered. Part V, which deals with forest management, covers several important topics: natural tropical forests, plantations, agroforestry, and watersheds. 6 MANAGING THE WORLD'S FORESTS Chapter 13 (by John Palmer and Timothy J. Synnott) discusses the jus- tification for forest management and concentrates on the factors needed for effective forest management, pointing out that government forestry depart- ments should not aim to be the sole managers of forest in the national interest. The authors argue that industries and wood-using enterprises also have a legitimate interest in ensuring that the source of their raw materials is managed appropriately, and land-hungry farmers on the forest fringes must receive equitable access to forest goods and services and must partici- pate in decisions about management. Consumers of forest products also have a legitimate interest in the fate of forests. Conflict over forests and land is likely to sharpen, the authors note, as public opinion, democratic move- ments, and special interest groups influence policy decisions. The authors also outline major conditions for sustainable management, including the knowledge that people must have for better management. In addition, the authors recommend that the members of the donor commu- nity better coordinate their efforts; encourage better information on tech- niques for management of natural forests; plan and execute development projects in the social, cultural, institutional, and political context; use the established criteria for sustainable management; and use forestry develop- ment projects as demonstrations of good management. Chapter 14 (by P.J. Kanowski and P.S. Savill, with P.G. Adlard, J. Burley, J. Evans, J.R. Palmer, and P.J. Wood) describes the development of planta- tion forestry since 1950 and the important lessons from early trials for today's successful use of tree plantations. The authors review the best management strategies and discuss the financial characteristics of trees as crops, the policy environment, the elements of forest management, and the environmental impacts of plantations. They also assess the constraints, opportunities, and priorities relating to plantations. They conclude with lists of the elements necessary for plantation programs, the technical requirements for successful plantation forestry, the key research approaches necessary for the continu- ing development of plantation forestry, and some ''basic truths" for the formulators of forest policy. Chapter 15 (by P.K.R. Nair) makes the case for the "coming of age" of agroforestry, with the realization of the potentials for sustained yields and conservation benefits and of the multiple-output nature of the age-old land- use systems in which trees and crops, and sometimes animals, are raised on the same piece of land in interacting combinations. The chapter evaluates a few promising agroforestry practices in terms of their potential as well as ecological adaptability, and it develops a matrix of agroforestry practices versus agroecological conditions that can be used as the basis for the design of agroforestry systems. The author also briefly considers some common constraints to adoption of agroforestry practices and argues for more re- search in agroforestry. As the forests in most parts of the developing world become degraded and the demand for fuel, fodder, and similar products continues to increase, INTRODUCTION 7 the importance of nonforest sources of production is growing. And as ex- propriation by the state, privatization, and encroachment reduce common property resources, and overuse degrades the resources that remain outside farming areas, reliance on on-farm resources increases. Thus chapter 16 (by J.E.M. Arnold) examines the economic and policy issues relating to the pro- duction of forest products by rural people along with their agricultural and livestock systems. The two main components of production at this level are (1) the incorporation into its farming system of planted and managed trees of value to the farm household and (2) the management of neighboring common property resources to provide inputs that will complement those available from on-farm resources. Arnold reviews trends in the use of, and rural reliance on, forest products; examines the role of common lands as a source of these products; and reviews trends in the growing and manage- ment of trees in farming systems. Chapter 17 (by Kenneth Brooks, Hans Gregersen, Peter Ffolliott, and K.G. Tejwani) examines the role of watershed management as a component of forestry and related development projects, and provides a practical frame- work that can be used to identify and assess priorities for watershed com- ponents in forestry projects. The chapter also examines the role of trees and forests in meeting watershed management objectives, and provides many examples to illustrate the problems and opportunities associated with inte- grating watershed management into the fabric of development projects. The authors add that appropriate technology based on local practices and re- sources is preferred and argue that single species (monocultures) and quick- fix solutions should be avoided in favor of more diverse systems that can better capture the hydrologic and ecological characteristics of stable sys- tems. The authors emphasize the need for active training and research in watershed management and the importance of conducting pilot studies for projects where considerable uncertainty exists and problems are evident. Generally, the authors favor small-scale projects instead of large, complex, and multifaceted projects. The authors articulate a number of recommenda- tions to overcome barriers to the adoption of improved watershed manage- ment practices. Finally, part VI examines the existing conditions and the policies needed at the local, national, and global levels for managing forest resources in a broader context of sustainable development. Chapter 18 (by Narendra P. Sharma, Raymond Rowe, Mikael Grut, Randall Kramer, and Hans Gregersen) argues for a merger of conservation and development objectives because in the long term they are complementary. Variety in land use-from mainte- nance of the pristine forest preserve to the productive clearing of forest for sustainable agriculture-is consistent with both concepts. Moderation and balance in the use of resources-plus recognition of linkages and variation in the environment-are cornerstones of both conservation and sustainable development. The authors emphasize the importance of improving our knowledge of environment-friendly technologies, land-use planning, and 8 MANAGING THE WORLD'S FORESTS design and implementation of policies that influence forest- and tree-based activities. To reduce the strain on forest resources, the authors point out that countries can undertake a variety of demand-reducing activities, such as improving their efficiency in extracting, processing, and using wood; substi- tuting nonwood fuel sources; promoting paper recycling; expanding mar- kets for diverse species of tropical hardwoods; and pricing their forest prod- ucts more efficiently. On the supply side, countries can improve the management of existing forests for different uses; encourage plantations, woodlots, and agroforestry as alternative sources of wood products; and improve the management of the area devoted to preservation and protec- tion of forests. In addition, the authors state that the diverse interests at the local, national, and global levels must be reconciled to promote wise use of forest resources. The authors emphasize that countries should adopt a multisectoral approach to forest management and create incentives through policy and institutional reforms for sustainable use of forest resources. The authors stress the importance of local participation in forest projects and of recognition of property rights, the interests of women, and the needs and contributions of indigenous people. The authors also describe some actions needed at the global level to find solutions to the problems associ- ated with deforestation, to preserve intact forests, and to expand forestry research. The last chapter (by Narendra P. Sharma, Clark Binkley, and Jeffrey Burley) weaves together the critical concerns of this book into a policy for sustainable forestry, recognizing the critical role of the market, the public sector, and the people. The authors argue that countries need to correct market and government failures and to get people involved in forest man- agement, and to create an appropriate policy environment so that forests retain their essential natural functions while sustaining their capacity to support people. In the short run, especially in the tropics, the extent and quality of existing forests need to be stabilized through the funding of ap- propriate solutions to the causes of deforestation. In the longer term, forest resources need to be augmented-through reforestation and afforestation as well as through sustained, integrated management of existing forests. In addition to policy and institutional reforms, the other main goals proposed for governments are conservation of natural forests through sus- tainable management for multiple uses, implemented through land-use zoning; expansion of protected areas for preservation of diverse forest ecosystems with global efforts; augmentation of forest resources through forestation to meet the demand of forest products and to provide environ- mental services and ecosystem protection; and implementation of programs to intensify agriculture and to promote rural development, especially in densely populated areas adjacent to forests. In short, saving the world's forests for future generations will require greater international cooperation, further developments in global organization, new revenue mobilization and INTRODUCTION 9 allocation mechanisms, availability of concessionary funding, increased re- search, and further improvements in economic tools available to define and analyze environmental externalities relating to forestry.· This book also contains a statistical appendix that presents hard-to-find data, compiled by country and region, on forest resources and forestry activities. FOUR QUESTIONS 1. Why Are Forests and Trees Important? Forests account for almost 30 percent of the earth's total land area. People throughout the world are increasingly recognizing the importance of forests and trees in improving human welfare. Both natural and man-made forests have economic, social, and environmental benefits, and forests play an important role in economic development-providing employment, in- come, and foreign exchange. Forests represent capital when converted to desirable forms of shelter and infrastructure; forests also provide land for food production. They con- tribute to the economy by providing commercial products (sawnwood, ve- neer logs, and logs for pulp), as well as nonwood products (nuts, fruits, gums, fiber, latex, bush meat, and palms). Forests also provide materials for agricultural, industrial, and medicinal uses. The economic benefits arising from the use of nonwood products on a sustainable basis can be substantial. Forests are also an important source of food, fiber, and energy for indig- enous populations and local communities. Nearly half of the world's popu- lation, mainly in developing countries, depends to some extent on forests for consumption goods. Forests are also an integral component of the biosphere, helping to stabilize natural systems. Forests contribute to biological diversity and help maintain air, water, and soil quality. They influence biogeochemical pro- cesses, regulate runoff and groundwater, control soil erosion, influence local climate, and reduce downstream sedimentation and flooding. As carbon sinks, forests sequester carbon dioxide from the atmosphere, thus reducing the greenhouse effect. They have aesthetic value and offer recreational op- portunities. Forests have "nonuse" or "existence" value as well, because people value forests even when they make no direct use of the resource now. The loss of environmental benefits from depletion of forests can be considerable in economic terms (especially when the effect is irreversible), but these costs are difficult to quantify. 2. What Is the Problem? People everywhere are concerned about the rate at which forests are being depleted and the extent of destructive deforestation. In recent decades the pace of deforestation has been increasing because there are strong incen- 10 MANAGING THE WORLD'S FORESTS tives to exploit forests, Deforestation in the tropics is now estimated at nearly 20. million hectares annually, an area almost equivalent to Britain or Uganda, Many developing countries face acute shortages of fuelwood, fod- der, timber, and other forest products, Atmospheric pollution threatens tem- perate forests in many industrialized countries, while many tropical and temperate areas lack forests altogether, By the year 2000 the world population will increase by 1 billion, with developing countries accounting for most of the increase. The rise in popu- lation and income will increase demand for both market and nonmarket forest goods and services-and that demand will place more pressure on existing forests, particularly in developing countries. Deforestation in the tropics is expected to continue to be significant throughout the 1990s. Misuse of forests has significant social, economic, and environmental costs with local, national, and global implications. Depletion of forests has resulted in loss of biodiversity, possible global climate change, degradation of watersheds, and desertification, In many countries, forest-dwellers have been displaced and cultural diversity threatened. Reduced fuelwood sup- plies have significantly influenced how women and children (the primary fuelwood gatherers) spend their time. Deforestation, together with land degradation, exacerbates the problem of poverty in rural areas. Besides hav- ing adverse environmental and social consequences, wasteful deforestation generates economic losses, including the permanent depletion of a renew- able resource, loss of genetic diversity, and reduction of agricultural pro- ductivity. 3. What Are the Causes of the Problem? Economic activities, such as agriculture, cattle ranching, fuelwood gath- ering, commercial logging, and infrastructure development, are perceived as direct causes of deforestation. But these causal factors are driven by economic, social, and political forces in a broader context of political economy. These forces manifest themselves through market and policy failures, popu- lation pressures, and poverty.. The relative importance of these direct and underlying causes of deforestation varies significantly among countries. Social factors (e.g., culture, values, traditional practices, and property rights) influence people's interaction with forests, their access to forests, and their valuation of forests. Economic factors (e.g., the market, incentives, and trade) influence the production of forest goods and services, the role of the forest sector in the national economy, and the distribution of income result- ing from forest activities. Political factors (e.g., the political system and the political process of decision making, government ownership of natural re- sources, and public policies) affect the degree of intervention in the pricing and extraction of forest products, the extension of favorable treatment to interest groups, and the selective provision of forest output as public goods. External factors (e.g., the demand of foreign countries for local resources and products) also influence economic and political considerations in forest INTRODUCTION 11 use. The dynamic interaction of these social, economic, and political factors creates competing demands for forest goods and services and forest lands, which result in either sustainable use of forests or destructive deforestation. Interest groups have an important role in the exploitation of forests, influencing policy decisions and management of forest resources. At the local level, where the concern is for improving human welfare, people use forests for commercial and subsistence purposes, and they clear forest areas for farming and ranching. At the national level, forests often represent an important source of foreign exchange, employment, government revenue, and land for agriculture, mining, or industry. In response to social and political pressures, national interests frequently favor exploitation of forests for short-term economic gains. At the global level, people demand forest products but also seek to preserve forests because of their role in climate and biodiversity. Because of their competing aims and values, local, na- tional, and global interests often conflict. Furthermore, within each level there are competing interest groups. At the local level, for instance, forest dwellers, farmers, landless people, commercial entrepreneurs, and local government compete for the use of forests. The market does not value all the environmental goods and services that have characteristics similar to those of pure public goods. This market failure creates conditions for inefficient use of forest resources. Because en- vironmental costs are not internalized, private and social costs diverge. More- over, the conflict between the time horizon of people now living and the needs of future generations creates a bias in favor of exploiting forests more rapidly. The use of high discount rates in investment decisions discourages conservation and environmental protection projects that have long gestation periods for generating net benefits. Also, the lack of clearly defined prop- erty rights creates market distortions and makes forests vulnerable to presc sures from rapid population growth and poverty. Finally, benefits and costs are often not directly related to the use of forests. Although some benefits from the use of forests (e.g., harvesting of forest for wood products), can accrue directly to some people today, environmental costs (e.g., downstream effects in the form of flooding and soil erosion) may be borne by others in distant places and over time. This situation inhibits individuals and govern- ments from taking costly measures that have intangible benefits. Public policies seldom provide adequate incentives for sustainable man- agement of forests or promotion of reforestation. By distorting the true cost or price of forest resources, perverse public policies have encouraged short- term exploitation of forests. Experience in many countries shows that agri- cultural incentive policies, resettlement, taxation, and trade policies are fre- quently more influential in determining land use than forest-sector policies. Existing agricultural and credit policies and tenurial incentives often en- courage expansion of the agricultural frontier at the expense of forests. In- adequate government response to increasing land scarcity provides incen- tives to people, especially in densely populated regions, to move into forest 12 MANAGING THE WORLD'S FORESTS areas. Inefficient fuelwood policies (pricing, concession policy, and subsi- dies) have made. fuelwood scarce in many areas, leading to depletion of forests. Severe underpricing of tropical timber through deficient royalty and concession policies leads to serious waste of resources. Underpricing also implies that the owners of the resources are not capturing a significant portion of timber rents. Countries reduce benefits from commercial forestry by maintaining an unstable macroeconomic environment, keeping wood artificially cheap, and directing investments toward inefficient processing industries. Other negative consequences include unsustainable management of natural forests, low levels of reforestation, inadequate use of processing capacity, and even the loss of forests. Finally, weak enforcement of existing regulations and concession agreements also has encouraged unsustainable use of forests. In the many countries where the government is the principal holder of forest property rights, traditional systems of providing access to forests and allocating common property resources to local people have broken down. The government's disregard of traditional rights of local communities and tribal groups makes forests more vulnerable to open-access problems. More- over, in many instances, governments lack the capacity to manage forests effectively and to control access to forest land under public ownership. Local people also lack the technology and the legal and institutional frame- work to manage forests sustainably. Forests are undervalued because many of their noncommercial prod- ucts, as well as their environmental goods and services, are not taken into account. Therefore in many countries the contributions of the forest sector to the economy (computed in terms of gross domestic product) is less than the contribution of other productive activities such as agriculture and in- dustry. As a result governments tend to assign a low priority to the forest sector and to make relatively low investments in forest management, research, and plantation programs. And because of a general lack of know- ledge about the ecological effects of human interaction with forests, govern- ments and the private sector often ignore the environmental benefits derived from forests and the environmental costs associated with destructive deforestation. Even though a society may place a high value on environ- mental services, if the goods and services do not generate a monetary re- turn, forests may still be undervalued by the market, the private sector, or the government. Many forest conservation and development programs suffer from weak legal and institutional support. Forestry institutions such as forestry depart- ments usually operate within a larger framework in which overlapping jurisdictions and policy objectives lead to conflict over forest land use. Revenue-earning, development, and conservation priorities conflict. Forestry institutions are frequently pressured to support some objectives to the neglect of others. Governments have also failed to include local communities, tribal groups, and the private sector in the long-term management of forests. INTRODUCTION 13 Finally, intact forests, especially primary tropical moist forests, are in- creasingly viewed as a global environmental good because of their biodiversity and their influence on climate. But the world community has neither the institutional and legal framework nor a special global fund to impose guidelines and "best practice" behavior on countries to ensure sus- tainable management of forests or to finance large-scale management of preservation forests. Each nation retains the sovereign right to manage its forest resources as it wishes, and there is as yet no consensus in the world community on sustainable use of forests. 4. How Should This Problem Be Addressed? The world community and independent nations face two forest-related challenges: to manage existing natural forests (both temperate and tropical) sustainably and to expand forest resources through reforestation and affor- estation. Plantations in tropical and temperate areas, restoration of degraded forests, and trees planted outside forest areas (e.g., farms and urban areas) should provide more forest products and environmental services. Appro- priate local, national, and global actions are needed to meet these chal- lenges. A participatory approach, which takes into account local needs and national priorities and is based on international cooperation, is vital. In the transition to sustainable development of forests, trade-offs be- tween short-term economic gains and long-term development must be made. By balancing conservation and development goals, sustainable development protects the interests of current and future generations in the use of forest resources and links consumption to the needs of the society. Sustainable development also requires reducing population growth and poverty, par- ticularly in areas where natural resources and the environment are already under stress. As countries try to stabilize existing forests and increase forest resources, they face many important questions: How much forest should be main- tained to meet the desired economic, social, conservation, and environmen- tal objectives? How should these resources be classified and managed to reflect both the productive and the protective functions of forests (forest reserves; national parks; protective forests; forests for timber production; wildlife preserves; forests for recreational purposes; and forest areas for mixed cropping, tree crops, agroforestry, and nonwood product extraction)? To what extent should global concerns be reflected in these decisions? Answers to these questions go beyond the scope of economics or the market. Important ecological, ethical, and sociopolitical considerations are involved as well. Economic reasoning and improved scientific information will be helpful, but ultimately each country must decide how much forest to maintain to accommodate current and future needs. Ideally, all remaining natural forests should remain intact, but preserving them all intact would be unrealistic given the needs of many developing countries for social and economic development. Most countries will opt for a second-best solution by considering intact forests for multiple uses, balancing conservation and 14 MANAGING THE WORLD'S FORESTS development objectives. On ethical considerations, a "burden of proof" falls to each country to demonstrate that use of natural forests is necessary and ecologically sustainable. Eventually each country must develop its own forest conservation and development program and create incentives for sustainable use of forest resources. Countries should take the values of conserving forests for the future into account in decisions about land use and management of forest ecosystems for multiple uses, including commercial forestry. Strategies for more efficient use of forest resources should be developed in a larger context of natural resource management policy. Specifically, to establish incentives for sustainable use of forests, countries need to develop a com- prehensive national land-use policy, strengthen forest management systems for single or multiple uses, enhance traditional restrictions on destructive resource use, create property rights and legal restrictions, correct market and policy failures, develop forestry institutions and human resources, involve local people and the private sector in forest management, adopt environment-friendly technology, implement conservation measures in the consumption of forest goods, and expand environmental education. Correcting market and price distortions will significantly improve the use and management of forests, but because of the opposition of interest groups, policy changes will be difficult, requiring strong public support and political will by decision makers. Furthermore, as noted earlier, many pol- icies relating to the use of forests fall outside the forest sector, and the policies in other sectors support competing objectives with broader implica- tions. Loss of forests, for example, is an unintended effect of some agricul- tural policies (such as pricing, taxation, and subsidies) designed to increase food production, income, and employment. Policy reforms could require the sacrifice of some benefits in the short term. For example, preservation of forests could prevent people who depend on this resource for their liveli- hood from having access to forests. In the absence of alternatives for gener- ating employment and income, these people could fall deeper into poverty. Policy changes will also be complex because of other considerations. Externality costs associated with forest use (such as replacement and envi- ronmental costs resulting from harvesting) may have regional and global consequences, but sovereign governments can be expected only to address domestic externalities in their pricing and taxation policies. Global exter- nalities need to be dealt with through international cooperation and, possi- bly, income transfers from rich countries to poor. In addition, forest land is used by people with a wide variety of land tenure arrangements, including indigenous tribal groups with long-held customary rights, illegal squatters, communities managing common land, and freehold farmers with state- granted leases or titles. Customary tenure systems also vary considerably and can be much more complex than open-access systems. The rights that people have over forest lands significantly influence their response to par- ticular incentive policies. INTRODUCTION 15 Moving toward conservation and sustainable use of forest resources also has significant cost implications. Because intact natural forests, espe- cially primary tropical moist forests, are increasingly considered to be a global environmental good, the compelling questions relating to forest con- servation on a large scale are those of cost and compensation. Who should pay for the cost of preserving forests for the benefit of the world commu- nity, now and in the future? Also, what proportion of the cost should be met nationally because benefits accrue locally? How should countries, and affected social groups within countries, be compensated for income forgone as a result of forest preservation that benefits everyone? And how should that compensation be determined? Such questions, however difficult, lie at the heart of the conflict between development and preservation of forest resources. Proper valuation of forests to promote more efficient uses of forest re- sources needs special attention. Accurate valuation is essential for better allocation of resources and for improved design and appraisal of both for- estry and nonforestry projects. Investment decisions among alternative land uses require accurate measures of costs and benefits of different forest goods and services. Undervaluation of forest products as a result of distorted markets and unpriced environmental service.s provided by intact forests may create a bias toward incentive policies favoring nonforest land-use activities. At present, national income accounts reflect only a fraction of the goods and services generated by forests. Current national accounting practices fail to treat the depletion of forest stocks as capital depreciation or to consider the degradation of the environmental services associated with forest destruction. The world community can help countries stabilize natural forests and deal with global environmental concerns. That community also carries the ''burden of responsibility" to support developing countries in their drive to use forest resources more efficiently. During the past decade the world community. has launched a number of important initiatives (such as the Global Environmental Facility, the Tropical Forestry Action Plan, and the establishment of the International Tropical Timber Association) that empha- size preservation and sustainable management of forests, but more needs to be done through international cooperation. The world community urgently needs to develop a global strategy for forest management and to provide funding to help countries. All types of forests, not just primary tropical moist forests, need help; at present, too much attention is directed to the latter. The world community should also support research efforts to improve knowledge of the ecological, biological, and physical processes of tropical forest ecosystems. Additional research should focus on understanding the physical effects of human interventions in tropical forests and on creating sustainable management systems of tropical forests. 16 MANAGING THE WORLD'S FORESTS Funding for the preservation of ecologically diverse forest ecosystems and for reforestation must increase significantly during the 1990s. Because preservation of forests has worldwide benefits, the world community should contribute to the direct and indirect costs of expanding preservation forests. In order to achieve sustainable development objectives, the donor commu- nity should also provide incentives by making forestry lending attractive. More concessionary funding should be made available for reforestation, as well as for investments in large conservation and environmental programs that have significant regional and global benefits. In addition, such funding could be made available for technical assistance, research, training, comple- tion of inventories, development of information systems, and pilot projects. In the immediate future the debate about the status of world forests will continue and perhaps intensify. Throughout the 1990s, deforestation is likely to continue apace, and population growth and poverty will continue to place pressure on existing forest areas. As the rapid loss of natural forests pushes the planet to the threshold of crisis, people will respond more readily to this serious problem. Better management of forest ecosystems will evolve through incremental responses and adjustments, but the problem is serious enough to warrant special attention. As indicated in the following chapters, individual countries are taking steps to improve the use of forest resources for different purposes, but the world community can accelerate the transi- tion to sustainable development through collective action. 2 World Forests in Perspective Narendra P. Sharma, Raymond Rowe, Keith Openshaw, and M ichael Jacobson WORLD FOREST RESOURCES Types and Distribution of Forests F orests account for 3.6 billion hectares-about 28 percent-of the world's land area (see table 1, sta tistical appendix to this volume). Of this total forest area, about 2.9 billion hectares a re closed forests (land covered mostly by trees, with stand d ensity greater than 20 percent), and 700 million hect- a res are open forests (mixed forest-grassland systems, with trees covering at least 10 percent of the ground). In addition, there are 1.7 billion hectares of other wood ed land, including forest fallow and shrubland . Thus the total a rea of "wood y vegetation" is 5.3 billion hectares--40 percent of the world's land area (FAO 1988). 1 The world's original forest area, estima ted a t abou t 6 billion hectares, has been d eclining steadily. About one-third of the forests have been lost during the past few hundred years. Although forests have more or less stabilized in the developed countries, d eforestation in the tropics has in- creased . Exhibit 2-1 shows the rela tive ra te of d ecline of forests in d evelop- ing countries. Types of Forests Forests are influenced by clima te, la ndform, and soil composition . As indicated on color pla te 1, the major natural forest ecozones show regional variation on a global scale. 1 The woody biomass growing stock of the fo rests totals approximately 340 billion 3; m a further 33 billion m 3 of woody biomass grows outside the forests, for a world total growing stock of 373 billion m3 (World Bank estimates). 17 18 MANAGING THE WORLD'S FORESTS EXHIBIT 2-1. Forest Area in Developing Countries Change in Area Since 1850 4 3.5 en 3 ..... 113 t5 2.5 .r::. 0 (/) c 2 .Q .E 1.5 113 ..... ~ 1850 1870 1890 1910 1930 1950 1970 1860 1880 1900 1920 1940 1960 1980 D Tropical Africa - Latin America D SouthAsia - Southeast Asia D China - Middle East & North Africa SOURCE: WRI 1987. Tropical forests. Tropical forests, three-quarters of w hich are broad-leaved species,2 consist of moist and dry forests. Together w ith woodlands, tropical forests account for about 3.1 billion hectares. The moist variety (1.5 billion hectares) can be further divided into rain and deciduous forests. The tropical rain forests, accounting for two-thirds of the tropical moist forests, are rich in biodiversity and contain valuable tropical hardwood. Examples are the lowland evergreen forests of the Amazonian and Zairian river basins. The tropical deciduous forests, which lie along the fringes of the rain forests, are less complex and have more distinct wet and dry periods. Some examples of tropical deciduous forests include the monsoon forests of South Asia, the teak forests of Southeast Asia, and the montane forests of Central America. According to management criteria, tropical moist forests can be classified as Broad-leaved species are also known as hardwood, nonconiferous, or angio- 2 sperm species. IBRD 23446 Color Plate 1. World Forest Zones \ ,, / - D Boreal forest Temperote mixed forest Temperote evergreen forest - D D D Tropical rein fo rest Tropicol deciduous fo rest Tropical dry forest None or limited forest Source: FAO 1976 Forest Di stribution by Region (millions of hectares) fo rest and O ther Wooded Land os 0 Forest a nd Other a Percentage of Region forest Land Other Wooded Landb Wooded Land Total Land Africa 706 633 1,339 45 Asia o nd the 24 568 264 832 Pacifi c Region Latin America 938 353 1,291 63 North America 459 275 734 40 Europe 137 41 178 38 U.S.S.R. (Former) 791 138 929 42 World 3,599 1,704 5,303 40 o Includes closed ond open forests jsee glossary for definitions). b For developing countries, other wooded land includes forest follow and shrubland; for developed countries, it includes open woodland ond shrubland. Source: FAQ 1988, ECE 1985, and WRI 1990. JUNE 1992 WORLD FORESTS IN PERSPECTIVE 19 primary forests (900 million hectares), secondary forests (300 million hectares), and forest fallow (300 million hectares). The tropical deciduous forests (1.6 billion hectares), most of which are in Africa, range from tall deciduous stands to more open grassland and shrubs. Most of the tropical deciduous forests have been altered by human activities such as fire, animal husbandry, and agriculture (Bourliere 1983). Temperate forests. Temperate forests, including woodlands, now cover about 2.2 billion hectares, almost three-quarters of which are classified as closed forests. These forests account for 85 percent of the world's coniferous forests and the bulk of the world's industrial wood production. There are two main groups of temperate forests: boreal and mixed. The boreal forests, which stretch across Canada, northern Europe, and the former U.S.S.R., include large areas of almost pure single-stand species. The smaller group of mixed forests are the temperate broad-leaved or evergreen forests located near or in subtropical climates (Ovington 1983). Plantation forests. Plantation forests are cultivated primarily for industrial use, although in recent years they have been established for nonindustrial purposes, such as conservation, household energy needs, and agroforestry. Plantations, which usually grow faster than natural forests, are often established to produce desired species. A plantation forest is usually composed of a single species, mainly exotic, which has been tested for its wood quality, growth rate, stern form, disease resistance, and manageability. There are now about 100 million hectares of temperate plantations and 35 million hectares of tropical and subtropical plantations (Kanowski et al. 1990). Together they account for less than 3 percent of the total forest area (see table 3, statistical appendix). Plantations provide important industrial wood in Chile, Kenya, South Africa, China, Britain, Ireland, and Spain. But plantations for industrial purposes are less successful in lowland tropical countries; high-altitude grasslands, however, are particularly suited to certain species. Distribution of Forests3 Exhibits 2-2, 2-3, and 2-4 illustrate the distribution of forests and wood- lands according to type of forest and geographical location.4 North America and the former U.S.S.R. account for most of the world's coniferous forests, 3 Forest data are compiled by the use of remote sensing techniques and comple- mented by on-ground surveys. FAO is responsible for global assessment of forest resources. The current data are based on surveys done in 1980 and updated in 1985; new estimates are expected in 1992. 4 Temperate and tropical forests are classified differently. The temperate regions divide forest area into closed forests and other wooded land. The tropical regions classify forests into closed forests, open forests, forest fallow, and shrubland. 20 M A NAGING THE WORLD'S FORESTS EXHIBIT 2-2. World Forest Distribution, by Type Temperate Forest (29.1 %) Tropical Forest (40.0%) Tropical Woodlands (20.0%) Temperate Woodlands (10 .9%) SOURCE: FA O 1988. EXHIBIT 2-3 . Forest Distribution, by Region Africa (19.6%) U.S.S.R. (22.0%) Europe (3.8%) Asia & the Pacific (15.8%) North America (12.8%) Latin America (26. 1%) SOURCE: FAQ 1988. WORLD FORESTS IN PERSPECTIVE 21 EXHIBIT 2-4. Forest and Woodlands Distribution, by Region U.S.S.R. (17.5%) Africa (25.2%) Europe (3.4%) North America (13.8%) Asia & the Pacific (15.7%) Latin America (24.3%) SOURCE: FAO 1988. and Canada, Brazil, and the former U.S.S.R. account for nearly one-half of the world's dosed-forest area. The former U.S.S.R., with 930 million hect- ares (42 percent of land area), is the world's most forested country; Brazil, with 680 million hectares, is second. Africa has about 700 million hectares of forests (24 percent of its land area); most of this area consists of savanna, open forests, or shrubland. Forests in Latin America cover about 940 million hectares. Forests in Central America are largely depleted because of population pressures and agricul- tural expansion, but the Amazonian forests remain the single largest area of primary tropical forest. Asia and the Pacific region have about 550 million hectares of forests. Asia has the largest area of bamboo forests; China's extensive temperate forests include the largest coniferous forest of any re- gion in the developing world. Tropical Asia has 350 million hectares in forests and about one-fourth of the world's tropical moist forests. The tropi- cal forests of Asia are the world's most important source of industrial hard- woods. Europe's forests (about 140 million hectares) cover 35 percent of its total land area. The Nordic countries (Finland, Iceland, Norway, and Sweden) have the highest proportion of land covered with forests (50 percent), fol- lowed by Eastern Europe (28 percent) and Western Europe (23 percent). North America's forests total 460 million hectares, of which Canada has slightly more than 250 million hectares. North America is the largest pro- ducer of wood products, and more plantations are being established as the old-growth forests in this region are drawn down. 22 MANAGING THE WORLD'S FORESTS Utilization of Forests in Developed and Developing Countries Although forests occur in all major geographic regions, their use varies widely. Forests and other wooded areas in developed countries amount to 2 billion hectares-about 40 percent of their total land area. The developed countries have more than 90 percent of the world's temperate forests, and more than 80 percent of their wood extraction is for industrial timber. After experiencing rapid deforestation, developed countries have stabilized and in some cases increased their forest areas. These countries have encouraged multiple-use management of forests, including private investment that has resulted in substantial reforestation since World War II. At present the main concerns in forest management in the industrial countries are acid deposi- tion (also known as acid rain), potential climate change, competing land-use activities, the trade-offs between economic and environmental demands, and the need for more efficient management of forests by public and private owners. Developing countries, by contrast, account for 2.1 billion hectares of the world's forests and 1.2 billion hectares of other woodland. In these coun- tries, the rate of natural regeneration and forestation in the tropics has lagged behind the rate of deforestation. In years past, colonization increased de- mand for primary commodities, including tropical hardwood; more recently, rapid population growth, agricultural expansion, and accelerated economic development have increased the pace of deforestation. The main concern in the developing countries is to slow down uncontrolled deforestation and to expand tree planting, especially in deficit areas. ROLE OF FORESTS IN NATURAL SYSTEMS Forests and trees are important environmental resources. They provide valuable environmental services and help maintain local, regional, and glo- bal natural systems. Environmental Services and Ecosystem Protection Forests and trees protect watersheds. They retard soil loss and erosion, especially in areas of high rainfall, and retain moisture in the soil, ensuring a gradual supply of water to streams and rivers. Forests also improve air quality and help to maintain regional climates, especially patterns of rain- fall. Furthermore, forests maintain the dynamic conditions necessary for their own continued existence and support other ecosystems within the natural system. Trees outside forest areas also provide environmental ser- vices. For instance, a shelterbelt of trees across farmlands protects the soil against the drying effect of winds. WORLD FORESTS IN PERSPECTIVE 23 The diverse species within a forest play an important role within the ecosystem. No single species can create its own food from inorganic mate- rials or completely decompose all its own wastes. Nor can any species main- tain all the necessary characteristics of its habitat for its own persistence (Slobodkin et al. 1980). The maintenance of forest ecosystems requires that chemical elements necessary for life be recycled and made available to indi- viduals and populations, and that usable energy be continually available. Species exist in an intricate web of interdependence. Many trees, for in- stance, depend on a number of species of fungi to derive nutrients from the soil. Free-living soil bacteria fix nitrogen from the air, making it available to trees. Small animals in the soil, such as earthworms and termites, modify the physical characteristics of the soil, improving tree growth. Trees depend on birds, bats, bees, and other insects for pollination of flowers, and on many kinds of vertebrates for the distribution of seeds. The higher the biodiversity, the greater the ecosystem complexities. Forests play a vital role in the global carbon cycle. Forests and their soils contain about three times as much carbon as is currently held in the atmosphere. Trees absorb carbon dioxide from the atmosphere through photosynthesis and emit oxygen. Whereas deforestation and subsequent decomposition of forest organic matter add carbon dioxide to the atmo- sphere, reforestation absorbs carbon dioxide from the atmosphere. Forests and Biodiversity A large proportion of the earth's biodiversity (species, genetic, and eco- system), which may have significant environmental and economic value, is found in forests. Tropical moist forests account for probably more than half of the earth's plant and animal species. Quality of Life Living areas that have valuable natural amenities, such as forests, offer an environmental dividend that can significantly contribute to both eco- nomic development and quality of life. Any increase in market goods pro- duced by diminishing a forest must be weighed against the potential loss of natural amenities. ROLE OF FORESTS IN ECONOMIC DEVELOPMENT Humans have depended on forests and trees for their economic liveli- hood and quality of life for many centuries, and this dependence remains equally applicable today. Forests provide food, fuel, fiber, timber, and other nonwood products. Forests also provide many goods that are used by for- est-dwellers and local communities and never sold in markets. Trees and forest lands have aesthetic value and offer recreational opportunities for 24 MANAGING THE WORLD'S FORESTS both rural and urban dwellers. Forests are also a source of new land when agricultural and urban expansion is necessary. Wood Products Perhaps the best-known single product of forests is industrial round- wood, from which come sawnwood, panels, pulp, and paper. About 80 percent of the industrial roundwood is produced and consumed in the developed countries. The forest sector's gross contribution to economic products in developing countries is $100 billion (1989 US$); fuelwood and roundwood used in the rough account for almost half, with the remainder representing the output of industrial forest products. On average, the forest- products industry contributes about 2.7 percent to the gross domestic product (GDP) of developing countries (FAO 1990c). But the wood indus- try in Malaysia accounts for 5 percent of GDP, and value added in the forestry sector as a percentage of GDP exceeded 5 percent in Liberia and the Ivory Coast and 4 percent in Cameroon and Tanzania (see table 8, statistical appendix). The forest-products industry also contributes significantly to employ- ment. In the United States, 765,000 persons are directly employed in the wood industry, and another 693,000 are employed in paper and allied prod- ucts. In Malaysia, 151,000 people are employed in the wood-products indus- try. In Gabon, Cameroon, and Nigeria, the forestry sector also employs significant numbers of people. These jobs, in turn, stimulate local economies and generate additional secondary employment. The forest-products industry also contributes to government revenue, which traditionally amounts to only a fraction of timber's potential rent but remains sizable. The forest sectors of Indonesia, Malaysia (Sabah), and the Philippines made annual payments to their domestic governments of $400 million, $425 million, and $50 million per year, respectively. In Sabah, forest income accounted for 70 percent of total government revenue. In addition, trade in forest products provides hard currency for exporting nations, in- cluding developing countries. Nonwood Forest Products Although individual nonwood forest products are less economically significant than timber, as a group they can sometimes contribute more to domestic and international economies on a per hectare basis than timber. Most such products are consumed locally, but international markets for some (such as rattan, latex, palm oil, cocoa, vanilla, nuts, spices, gum, and ornamental plants) already exist, and new markets for fruits and medicine are developing. Indonesia, Malaysia, Thailand, Sudan, Brazil, and Guate- mala already have significant exports in nonwood forest products. Forests contribute an enormous amount of recreational services to coun- tries. For example, recreation is probably the single most important com- WORLD FORESTS IN PERSPECTIVE 25 modity supplied on U.S. public land. Recreation, through hiking, hunting, and fishing, is clearly highly valued throughout Europe and other devel- oped nations. Although the subject is less studied, tropical forests also appear to generate significant amounts of recreation. With the growing interna- tional interest in ecotourism, countries with intact natural forests may be able to generate a new source of foreign currency. Forests and scattered trees provide a critical supply of energy for rural people in many parts of the world. An estimated 3 billion people rely on fuelwood for energy. Across all developing nations, fuelwood supplies nearly 20 percent of all energy needs. Most roundwood production in developing countries is for fuelwood. Forests and trees contribute to agriculture, and forest lands provide important inputs for grazing. Trees offer critical soil erosion protection on hillsides and near streams stabilizing large land areas. As mentioned earlier, trees as shelterbelts protect lands from desertification, thus allowing contin- ued farming. The tree canopy provides needed shade for agroforestry crops, and some trees fix nitrogen to fertilize soils and enhance crop productivity. Finally, trees grown in orchards and plantations provide an important food supply for all nations. Lands for Conversion Forest areas are also important to developing countries as a source of new lands for conversion to agricultural or other uses. Countries with rapid population growth and limited food supplies can convert nonacidic forest lands into farms. Through some conversion, properly utilized forest lands can serve to increase food and energy supplies for rapidly growing popula- tions in developing countries. Although tropical forests and woodlands are rapidly being converted to alternative land uses now, less than one-fifth of tropical forests have been converted to date. In contrast, over the centuries one-third of the world's temperate forests have been converted to alternative uses. If current rates of conversion continue, however, large areas of tropical forests also will disappear. By balancing forest needs carefully with conservation and development, countries can determine their ideal rate and direction of conversion. Fur- thermore, by emphasizing sustainable activities that do not undermine the resource base, countries can ensure that conversion leads to sustained eco- nomic development rather than a mining of resources. WORLD TRADE IN FOREST PRODUCTS During the three years from 1986 through 1988, total annual exports of logs and wood products in the world market were estimated at $68 billion, of which the developing countries' share amounted to $10 billion. In this 26 MANAGING THE WORLD'S FORESTS same period, total exports, primarily industrial wood, in roundwood equiva- lent amounted to 630 million m 3 (see exhibit 2-5). During this period, the developing countries also spent more than $13 billion annually on imports. Developed countries dominate world trade in wood products, accounting for more than 80 percent of total industrial wood exports during this period (see table 9, statistical appendix). Tropical hardwood products from natural forests, which represent nearly 70 percent of industrial wood exports from all developing countries, ac- count for just over 10 percent of world timber trade. Indonesia, Malaysia, and Brazil are the principal exporters of tropical hardwood products; their exports were valued at US$2.8 billion, US$2.6 billion, and US$0.2 billion, respectively, in 1988. Malaysia exports about 85 percent of its total indus- trial wood production. Indonesia, the world' s largest exporter of plywood derived from tropical hardwood, exports about 50 percent of its industrial wood production. Brazil exports about 10 percent of its industrial wood production, most of it in pulp and paper products. International markets for nonwood products are still expanding and are a potential source of foreign exchange for developing countries. Sudan al- ready earns about US$60 million annually from exports of gum, and Mada- gascar earns the same amount from exports of vanilla. In 1986 Indonesia earned US$134 million from exports of rattan, resin, essential oils, kapok, and chinchona bark (quinine). EXHIBIT 2-5. Production, Export, Import, and Consumption of Wood Products for Developing Countries and the World, 1986-1988• (Units: million m 3 roundwood equivalent) Fuelwood and Poles Industrial Wood Pro- Net Con- Pro- Net Con- due- Im- Ex- Ex- sump- due- Im- Ex- Ex- sump- Region tion ports ports ports ti on ti on ports ports ports tionb Africa 630 0 0 0 630 53 7.2 9.5 2.3 50.7 Latin America 430 0 0 0 430 127 15.5 19.8 4.3 122.7 Asia 1,290 1 1 0 1,290 256 75.9 75.9 0.0 256.0 Developing 2,350 1 1 0 2,350 436 98.6 105.2 6.6 429.4 Developed 300 3 3 0 300 1,264 531.4 524.8 --6.6 1,270.6 - - - - - World 2,650 4 4 0 2,650 1,700 630.0 630.0 0.0 1,700.0 SOURCE:FAO (1990c), FAO adjusted, which gave figures for imports and exports of saw logs, veneer logs, pulpwood, other industrial wood, sawnwood, panel products, wood pulp, and paper. All these products were converted to roundwood equivalent using standard conversion factors. Some adjustments were made to make imports and exports equal for the three-year period 1986-88. •A three-year average, 1986-88. bConsumption is assumed to equal production minus net exports. WORLD FORESTS IN PERSPECTIVE 27 Export Prices of Forest Products The average unit export price of wood products appears to be about 20 percent less in developing countries than in developed countries, as exhibit 2-6 shows: EXHIBIT 2-6. Average Unit Export Price of Selected Forest Products for Developing and Developed Countries, 1986-1988 (in U.S. dollars) Sawlogs Sawnwood Non- Non- Coniferous coniferous Coniferous coniferous Pulpwood Region ml (roundwood) ml (sawnwood) ml (roundwood) Developing 45.7 79.3 150.6 205.3 28.3 Developed 75.0 117.3 125.0 279.0 37.0 World 74.0 84.7 125.3 231.3 36.3 Difference• 64% 48% - 17% 36% 31 % Printing/ Other Paper Writing and Paper Plywood Wood pulp Newsprint Paper Board Region ml (solid) tons tons tons tons Developing 314.0 472.7 470.7 734.3 598.7 Developed 461.0 494.0 526.3 844.7 705.7 World 351.7 492.3 525.3 838.3 697.6 Difference• 32% 5% 12% 15% 18% SOURCE: FAO (1990c). •Developing average minus developed average divided by developing average. This difference could be attributed to a number of factors, including quantity, quality, shipping costs, marketing expertise, and underinvoicing. Prices of tropical hardwoods both round and sawn have been increas- ing in real terms at a rate of 1 percent to 2 percent per year over the past three decades (FAO 1983, 1990c; World Bank 1991). This increase could reflect the increasing costs of extraction, more intensive use of the product (e.g., as veneer rather than sawnwood), or the use in high-quality products such as furniture as opposed to joinery. However, the prices of all kinds of round and sawnwood appear to have at least kept pace with inflation, so tropical hardwoods are not unique (Bethel 1983; Johnston, Grayson, and Bradley 1967; USDA 1990b). Several tropical countries have export taxes or premiums on round- wood and sawnwood, and some ban log exports-and even rough sawnwood-to encourage the manufacture of finished and semifinished wood products for export. However, these measures will affect the forest indusMany 28 MANAGING THE WORLD'S FORESTS tries of some developing countries relying on log and sawnwood im- ports, including Thailand and Taiwan. TARIFFS AND DUTIES ON FOREST PRODUCTS Developed countries, especially the three largest trading areas of North America, the European Community, and Japan, impose very small duties- or none-on imported forest products (Japan Tariff Association 1990; USDA 1990a). For example, there is no duty on roundwood, pulpwood chips, and woodpulp; tariffs on wood manufactures average 3.5 percent; and tariffs on paper products range from zero to 9 percent (World Bank 1987). In contrast, average tariffs in developing countries range from 2 percent to 27.6 percent (see exhibit 2-7). EXHIBIT 2-7. Average Tariffs for Selected Wood Products of Developed and Developing Countries, 1982 Average Tariff(%) Developed Developing Countries Countries Roundwood (including pulpwood chips and wood pulp) 0.0 8.0 Sawnwood 0.2 13.1 Veneer and plywood 1.7 23.5 Wood manufactures (including paper) 3.5 27.6 SOURCE: World Bank (1987). developing countries have a comparative advantage in growing trees and the manufacture of certain wood products. Chile has used plantation-grown softwoods to build up a large forest-products export industry, and Indone- sia has become the world's largest exporter of plywood using indigenous tropical hardwoods as the raw material. Therefore, tariff reductions in de- veloping countries, especially in wood raw material and in products such as sawnwood and plywood, should promote trade and investment in the in- dustry. As mentioned previously, however, there are many nontariff barri- ers, such as restrictions on log exports and roughly sawn timber. Also, importing countries may impose barriers that discriminate against certain species, quality of wood products, origin, and the like. These barriers should be examined with a view to promoting freer trade. FUTURE DEMAND FOR WOOD PRODUCTS As mentioned earlier, the current demand for wood products, including underrecorded items such as fuelwood, poles, and hand-sawn timber-about 4.4 billion m 3- may reach 6.6 billion m 3 by the year 2025 (see exhibit 2-8). WORLD FORESTS IN PERSPECTIVE 29 This forecast was made using the IIASA model (Dykstra and Kallio 1987) for industrial wood (adjusted for unrecorded industrial wood products) and World Bank estimates for fuelwood and poles. The IIASA model has been criticized as being too high; a model by Resources for the Future analysts forecasts lower figures (Sedjo and Lyon, in press). (This subject is discussed more fully in chapter 16 of this volume.) However, lowering the forecast demand for industrial wood will not greatly affect the overall forecasts, because industrial wood accounts for less than half of the current and esti- mated future consumption. In addition, the forecasting base is somewhat uncertain because of the consumption of unrecorded wood products. Although the world's population will increase by about 70 percent by 2025, demand for wood products will increase only 50 percent as a result of the improvement in end-use efficiency, substitution, and the like. In abso- lute terms, demand for fuelwood and building poles will increase most, but in percentage terms, demand for industrial wood products will increase most. However, the large population gains in developing countries will put a considerable strain on tropical forest land and the demand for forest prod- ucts. If forest land clearing goes on unabated, an additional 170 million to 200 million hectare will have been cleared by the year 2000; by the year 2025 this figure will rise to 600 or 700 million hectares, reducing the forest area in the tropics by about 30 percent. Clearly, unless something is done to curtail land clearing, improve management of existing tree stock, encourage tree planting, and improve end-use efficiency of fuelwood, demand will not be met from the sustainable supply. EXHIBIT 2-8. Current and Forecast Demand for Wood Products in Developing and Developed Countries, 1987•, 2000, and 2025 (Units: million m3 roundwood) 1987" 2000 2025 Fuel- Fuel- Fuel- wood, wood, wood, House- House- House- building building, building, and Indus- and Indus- and Indus- Fencing trial Fencing trial Fencing trial Poles Wood Total Poles wood Total Poles Wood Total Developing 2,350 429 2,779 2,800 530 3,330 3,660 780 4,440 Developed 300 1,271 1,571 320 1,480 1,800 340 1,820 2,160 World 2,650 1,700 4,350 3,120 2,010 5,130 4,000 3,100 6,600 Population (World) million 5,027 6,193 8,491 SOURCE: Dykstra and Kallio (1987) for industrial wood; World Bank estimates for fuel wood and poles. See the statistical appendix for the official 1988 consumption from the FAO Forest Products Yearbook. •Average for 1986-88. 30 MANAGING THE WORLD'S FORESTS SUMMARY Covering nearly 30 percent of the earth's total land area, forests play vital roles in natural systems, as well as in economic development. The growing stock of the world's trees, about 373 billion m3 roundwood equiva- lent, produces some 11 billion m 3 of wood annually. The current demand for wood is about 4.4 billion m3 per annum. Despite the overall surplus, many developing countries experience local shortages of wood. Fuelwood and building poles account for 60 percent of consumption worldwide, but in developing countries these products amount to 85 percent of total wood production. Most of the wood produced in developing countries is for domestic use. The unit price of wood and wood products averages about 20 percent less in developing countries than in developed countries. At present, there is significant potential for increasing trade among developing countries, but the tariffs imposed on wood products are a barrier to trade. To encourage trade, tariff barriers should be eliminated. By the year 2025, demand for wood products may increase by 50 per- cent, to 6.6 billion m3 • This demand could be met from sustainable man- agement of some natural forests for wood production and increased tree planting. If these measures are not taken, there could be a shortfall of wood, particularly in developing countries, and this shortfall could speed the deg- radation of the environment as the remaining forest resources are depleted. REFERENCES Bethel, J.S., ed. 1983. World Trade in Forest Products. Seattle: University of Washington Press. Bourliere, F., ed. 1983. Tropical Savannas, Ecosystems of the World, vol. 13, Amsterdam, The Nether- lands: Elsevier Press. Dykstra, D.P., and M. Kallio. 1987. The Global Forestry Sector. New York: John Wiley and Sons. FAO (Food and Agricultural Organization). 1983. Forest Product Prices, no. 46, 1963-1982. Rome: FAO. _ _ _ . 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Forest Resources Division, Forestry Department, Food and Agricultural Organization of the United Nations, Rome. _ _ _. 1990a. Forest Product Prices, no. 95, 1969-1988. Rome: FAO. _ _ _ . 1990b. "Major Issues Related to Forestry in the 1990s." Unpublished report. _ __. 1990c. 1988 Forest Products Yearbook. _ _ _. 1990d. 1988 Agricultural Statistics Yearbook. _ _ _. 1990e. 1988 Agricultural Products Yearbook. Japan Tariff Association. 1990. Customs Tariff Schedules. Japan Tariff Association. Johnston, D.R., A.J. Grayson, and R.J. Bradley. 1967. Forest Planning. Winchester, Mass.: Faber and Faber. Kanowski, P.J., P.S. Savill, P.G. Adlard, J. Burley, J. Evans, J. Palmer, and Peter Wood. 1990. Plan- tation Forestry. Prepared as a World Bank Forest Policy Issues Paper, Oxford Forestry Institute, Oxford University. Ovington, J.D., ed. 1983. "Temperate Broad-leaved Evergreen Forests," Ecosystems of the World, vol. 10, Amsterdam, The Netherlands: Elsevier Press. Sedjo, R.A., and K.S. Lyon. In press. The Long-term Adequacy of the World Timber Supply. Washington, D.C.: Resources for the Future. WORLD FORESTS IN PERSPECTIVE 31 Slobodkin, L.B., D.B. Botkin, B. Maguire, Jr., B. Moore III, and H.J. Morowitz. 1980. "On the Episte- mology of Ecosystem Analysis." In Estuarine Perspectives, ed. V.S. Kennedy. New York: Aca- demic Press. U.S. Department of Agriculture (USDA). 1990a. Harmonized Tariff Schedule of the United States for Chapter 44 (Wood Products). Washington, D.C.: U.S. Government Printing Office. _ _ _ . 1990b. Wood Products, International Trade, and Foreign Markets. Circular WP 2-90. July. World Bank. 1987. Tile Uruguay Round. Washington, D.C.: World Bank. _ _ _ . 1991. "Commodity Prices and Protections in 1985 Constant Dollars." Unpublished docu- ment. 3 Deforestation: Problems, Causes, and Concerns Raymond Rowe, Narendra Sharma, and John Browder DIMENSIONS OF THE DEFORESTATION PROBLEM B etween 1850 and 1980 about 15 percent of the earth's forests and wood- lands disappeared as a result of human activities. The forests and wood- lands of North Africa and the Middle East, for example, declined by 60 percent; those of South Asia, by 43 percent; of tropical Africa, by 20 percent; and of Latin America, by 19 percent (Houghton et al. in WRI, 1987). Loss of forests continues to be a problem in these regions. Depletion of forests is most significant in the tropics, where about 2.5 billion people depend on natural forest resources for many economic and environmental goods and services. Social, economic, and political fac- tors have created incentives for rapid exploitation of forests and intensified pressure on remaining tropical forests and arid woodlands. Between 1980 and 1985 the estimated annual rate of tropical deforestation was 0.6 percent, or 11.4 million hectares (FAO 1988). Recent studies estimate deforestation in the tropics at a rate of 17 to 20 million hectares annually (see color plate 2). Deforestation is a widely used term, but one with different meanings in different contexts. Disturbance deforestation refers to all man-made distur- bances that seriously alter a forest; conversion deforestation refers to man- made disturbances that subsequently convert forest lands to alternative uses. Many man-made d isturbances, however, permit land to remain in forest use. For example, both sustainable timber production of primary (intact) and secondary forests (still covered by indigenous trees or shrubs) and plan- tations replenish forests after harvest. Also in this category are clearings of secondary forests devoted to sustainable, shifting cultivation (also known as slash and burn) when the secondary forest is part of each rotation. These 33 34 MANAGING THE WORLD'S FORESTS disturbances clearly affect forests, but they do not reduce the aggregate amount of land devoted to forest cover. Wasteful (or destructive) deforestation depletes forests that provide a highly valued flow of goods and services and replaces them with alterna- tive uses that yield lower-value goods and services. It is often assumed that all deforestation must be wasteful, but this generalization is too broad. Some forest lands can be converted to high-value urban development and agricul- ture that far outweigh the lost forest services. Some primary forests can be altered to managed secondary forests or agroforestry uses that yield higher returns. Society clearly benefits from these changes. Nonetheless, wasteful deforestation is widespread, as indicated by the absence of appropriate in- centives to manage lands in the long run; the intrusion of subsidies that favor nonforest activities; and the failure of markets to take into account forest services such as watershed protection, climate control, and biodiversity. Wasteful deforestation reduces a region's welfare and a country's long-term economy, but prohibition of all changes-even beneficial ones-on forest lands may also contribute to loss of social and economic benefits. DIRECT CAUSES The principal direct causes of uncontrolled deforestation in the tropics are agricultural expansion, overgrazing, fuelwood gathering, commercial logging, and infrastructure and industrial development. Agricultural Expansion Subsistence farmers in developing countries account for more than 60 percent of the loss of tropical forests annually, but regional variations are substantial. The expansion of smallholder agriculture in Latin America ac- counts for as much as 35 percent of tropical deforestation in the region; cattle ranching is responsible for most of the remaining 65 percent. Live- stock expansion in Latin America and Africa is directly responsible for much woodland degradation. In Brazil, for example, cattle ranching in the Ama- zon, much of it subsidized, accounted for about 70 percent of the natural forest area converted there (more than 12 million hectares) by 1980 (Browder 1988 and Mahar 1989). In many parts of Central America, where only frag- ments of original natural forests remain, agricultural encroachment is the leading cause of new forest conversion. In some Southeast Asian countries shifting cultivation accounts for up to 50 percent of natural forest conver- sion. More than 60 percent of the annual deforestation in Indonesia is attrib- uted to shifting cultivation on logged-over areas. Shifting cultivation is the leading cause of deforestation in both tropical West Africa and semiarid Africa, accounting for 70 percent of the woodland area converted (PAO 1982). In addition, fires, often set by shifting cultivators and other forest dwellers, are a major cause of forest degradation and impede regeneration of woody plants. IBRD 23447 Color Plate 2. Tropical Deforestation Tropic ol Cone er Equator Tropic ol Capricorn 1 See Inset C Source: Smit hs.onion lnstitiJfion, Tr-oveling Exhib ition Service. 1988. D Original forest cover - Present forest cover " Tropical Rainforests: A Oisoppeoring Treasure." Exhibit, Washington, D.C. C6te A B _!ropic of Conc•r Equator ti Cu!lib- '. :. ,~i - . - ~: Dominican Rep. Jamaica ~ • •. --....~J"OE>:::=...-Honduros N icaragua Pon mo Equoror Tropic of Capricorn ------\ c Tropic of Capricorn \.,Sri Lanka : Equator Australia JUNE 1992 DEFORESTATION: PROBLEMS, CAUSES AND CONCERNS 35 Overgrazing Overgrazing of both natural rangeland and planted pastureland in arid and semiarid regions is another important cause of tropical woodland deg- radation. Along with fuelwood gathering, overgrazing is the main form of deforestation in North Africa, the Middle East, South Asia, and the Sahel area of Africa. World statistics continue to show an increase in the number of grazing animals, and although the animal population has been declining in North Africa and the Middle East, many people in these regions depend on livestock as their main source of food and income. Initiatives to privatize and enclose traditionally regulated communal rangeland also have led to resource degradation . Moreover, livestock man- agement efforts based on rangeland fencing often conflict with wildlife management by preventing seasonal migration of wild-range animals. In upsetting the ecological balance, such enclosures compel both domesticated and wild species to overgraze their respective fodder bases, thereby contrib- uting to further land degrad ation. In the tropics, many cattle owners over- stock woodlands that have been converted to pasture in order to maximize short-term income. Overgrazing accelerates pasture degradation and dimin- ishes the capacity of the forest to restore itself once artificial pastures revert to fallow. Fuelwood Gathering Nearly 3 billion people worldwide depend on wood, primarily from natural forests and trees outside forest areas, as their main or only source of household energy. In developing countries, urban households, especially poor families, often spend 20 to 30 percent of their income to buy fuelwood. Fuelwood gathering contributes to land degradation, especially in agri- cultural regions with limited wooded areas. Fuelwood problems particu- larly affect the Sahel, eastern Africa, and Himalayan range, the Andean plateau, and densely populated Central America and the Caribbean, w here population pressure and inefficient fuelwood use, often combined with other forces, have prompted chronic supply shortages. This trend increases pres- sure on remaining forest resources. Comme rcial Logging Annually, 4 to 5 million hectares of commercially productive closed forests are logged (FAO 1982). There is some regional variation in the rela- tive contribution of commercial logging to loss of tropical closed forests. In Africa about 20 percent of productive tropical forests were logged over by 1985, whereas in Asia and Latin America the figures were about 19 percent and 9 percent, respectively. Latin America's tropical forests are the least affected by commercial logging, but log production is growing rapidly as 36 M ANAGING THE W ORLD'S FORESTS Asian and African forests are depleted of accessible commercial timbers. The forest damage directly caused by logging varies inversely with the degree of selectivity. More important are the indirect effects of logging on opening forest areas to subsistence farmers and the landless. A high propor- tion of the productive closed-forest area subjected to logging later becomes cropland . The direct damage of logging on residual stands is less in Latin America than in Asia, but the higher selectivity of logging in Asia means that larger forest areas are opened to shifted cultivators. Infrastructure and Industrial Development Investments in road building, hydroelectric projects, and mineral extraction, necessary to meet development objectives, often entail envi- ronmental trade-offs. Infrastructure projects that occur without adequate environmental impact assessment cause forest conversion, attract unemployed workers, and provide landless farmers open access to forests. N o precise estimates of the direct impact of infrastructure development on tropical forest conversion in the aggregate are available. Industrialization also can contribute to deforestation. Industrial air pol- lution has damaged or destroyed considerable areas of forests in central and Eastern Europe, as well as in eastern Canada. UNDERLYING CAUSES The direct causes of uncontrolled deforestation just described are driven by market and policy failures, population growth and rural poverty, and the state of the economy. 1 Tenurial traditions and tax policies often provide incentives for inefficient and inappropriate forest land use. Many of these underlying causes are related, and because they are so often embedded in divergent social and economic contexts, their relative importance varies substantially among countries. Solutions to deforestation based on signifi- cant policy reforms are more likely to succeed if measures are also taken to control population growth and to alleviate rural poverty. M a rket a nd Policy Fa ilures Market and public policies are important determinants of how forests are used and managed. The unsustainable rates of deforestation of the world's forests are to a large extent linked to inherent market failures in the free- 1 Accord ing to another strongly held view, public policies promoting agricultural expansion and commercialization play a larger role in deforestation than popula tion growth and poverty. DEFORESTA T/ON: PROBLEMS, CAUSES AND CONCERNS 37 market approach to forest use and management and to policy failures. The inherent market failures include the following: • The existence of externality costs associated with forest use, that is, the divergence between the private and social costs of nonforest land uses and commercial timber harvesting; • The problems of valuing joint products C\nd nonmarket environmen- tal services; • The conflict between the time horizon of people now living and the needs of future generations, which creates a bias in favor of exploit- ing forests now versus conserving for the future; and • Undefined property rights (open access to forests allows squatters and settlement). Inherent market failure has usually been compounded by government policy distortions. Rather than correcting the failure of the market, govern- ment intervention has often aggravated the existing incentives for excessive forest exploitation. More than 80 percent of tropical forest lands are publicly owned, and their uses are subject to diverse, multisectoral, and often con- flicting policy objectives. By distorting the true costs or prices of the forest resources, public policies often provide an incentive for short-term exploita- tion of forests. Public policies have frequently failed to provide adequate incentives for sustainable management of forests or to promote reforesta- tion. Inadequate land policy allows forests to be treated as an "open access" resource, often resulting in degradation of land and loss of forest. Logging concession policies have frequently resulted in inefficient use of forests as renewable resources. Concession periods, usually less than 20 years, are significantly shorter than natural regeneration periods of a timber stand (averaging 20 to 50 years). Also, rents through royalties, license fees, and reforestation taxes are typically much less than the real cost of replac- ing the timber stock and restoring logged-over areas. In addition, the gov- ernments of most timber-exporting nations have been unable to collect more than 50 percent of actual rent from log harvests through taxes and royalties. Aside from contributing to economic and environmental losses, timber con- cessions can promote corruption and result in progressive decapitalization of the country's natural resource base. Policies that increase taxes on log exports to protect fledgling domestic industrial wood processors often lead to production inefficiencies, low log recovery rates, and wasteful harvesting. In Indonesia, for example, plywood mills could have operating costs 2.5 times higher than their counterparts in other Asian countries and still remain profitable under that country's log- export tax policy. Because of these production inefficiencies, Indonesian mills have required 15 percent more log feedstock per cubic meter of ply- wood output than mills elsewhere in the region (Gillis 1990). Nonforest incentives (pricing policies, tax incentives, direct government outlays, and other subsidies) encourage private investments in leading de- velopment sectors such as agriculture, energy, mining, industry, and trans- 38 MANAGING THE WORLD'S FORESTS portation. Incentives that directly and indirectly promote expansion of agri- culture in tropical countries have had the greatest impact on forests. Such policies-by regulating prices and providing tax shelters and subsidized credits to livestock and agricultural producers-have expanded the agricul- tural frontier, often at the expense of natural forests. This problem is height- ened by the relative importance of agriculture in the gross domestic product and total export earnings of most tropical countries. Conserving natural forests at the expense of agricultural expansion entails short-term costs that many developing countries perceive they can ill afford. But many incentive policies that have led to desultory agricultural uses of forest lands would be discredited on economic grounds if the environmental impacts were prop- erly evaluated. Correcting the distortionary effects of inefficient incentive policies through proper regulations and land-use policies will improve forest use and reforestation, but improving market and price policies raises some dif- ficult questions. One is that the externality costs (such as replacement and environmental costs resulting from harvesting) associated with forest use are not borne by the domestic economy alone; some have regional and global consequences. Sovereign governments can be expected to address only domestic externalities in their pricing and taxation policies. Global externalities could be addressed through cooperation among nations and perhaps through income transfers from rich countries to poor ones. More- over, forest land is used by individuals and groups under a variety of land tenure arrangements; the users include indigenous tribal groups with long- standing customary rights, illegal squatters, and freehold farmers with state- granted leases or titles. Customary tenure systems also vary considerably and can be much more complex than open-access systems. The rights that people have over forest lands significantly influence how they will respond to particular incentive policies. Agricultural policies also have created incentives for deforestation. The commercialization of agriculture has typically led to monocultural planta- tions of export crops on fertile lowland areas; traditional farmers who once occupied those areas have been displaced onto more fragile uplands or tropical forests. The concentration of shifting cultivators on a smaller or poorer resource base typically sets in motion a downward spiral of reduced agricultural productivity, increased forest depletion, and worsened rural poverty, as cropland fallow periods are shortened and soil fertility declines. An estimated 60 percent of the world's poorest people have been pushed into ecologically fragile environments (tropical forests, drylands, and hilly areas) as a result of agricultural expansion (Leonard 1989). Public policies have significantly contributed to this process in several developing coun- tries. For example, to mitigate social pressure arising from a growing land- less rural population, Brazil initiated a frontier land settlement program in the Amazon state of Rondonia (Browder 1988). By 1988 deforestation had claimed an estimated 24 percent of the state's natural forest cover, up from only 0.3 percent in 1975 (Mahar 1989). DEFORESTATION: PROBLEMS, CAUSES AND CONCERNS 39 Governments in developing countries often intervene in agricultural markets to keep domestic food prices low, while providing incentives for cash crop production through subsidized interest rate credits, tax holidays, free provision of infrastructure, and land concessions. Such policies tend to reduce agricultural incomes among small farmers and thus their funds to invest in conservation measures. These policies also distort the factor prices associated with commercial production on larger units. Regressive rural property tax structures, which involve higher assessment for smaller prop- erties, favor extensive land uses, such as cattle ranching. Agricultural re- search and extension are largely oriented toward improving export crop production on large farms rather than enhancing income on smaller mixed farming systems. These interventions have tended to penalize smaller pro- ducers, while providing incentives for additional land clearing by small and large rural property owners alike. To relieve population pressures in heavily congested areas or to expand agricultural production, many developing countries (e.g., Indonesia, Nepal, Brazil, the Philippines, and Mexico) have also promoted regional develop- ment programs (such as colonization schemes and large-scale irrigation projects) with primary emphasis on agriculture. Typically these programs, which involve various fiscal and pricing incentives to attract private capital, result in forest conversion and uncontrolled deforestation. In addition, pres- sure to generate foreign exchange earnings has led to an emphasis on quick return and unsustainable land-use practices. Deforestation is also accelerated by inadequate tenurial policy. In many countries, especially in Latin America, governments cannot control access to the forest lands under public ownership. As a result, encroachment on for- est lands is widespread, and traditional systems of allocating common prop- erty resources have broken down. Tenants typically attempt to legitimate their claims of private ownership to public land by demonstrating land "improvements" (usually amounting to forest removal). Tenurial incentives, often codified in legislation, not only encourage forest clearing to establish private property ownership by colonists in forest areas but also oblige in- digenous forest-dwellers, whose common property rights to the same land are violated by these incentives, to establish land claims in the same man- ner. In parts of Africa, in contrast, tree planting may establish a claim to land ownership, and land disputes have traditionally been resolved in favor of the party claiming anct:stral ownership of the most mature trees on dis- puted lands. This tenurial tradition may counter efforts to promote tree planting where land ownership is in question, or encourage deforestation to undermine legitimate land claims. Population Growth and Rural Poverty Rapid population growth often intensifies pressure to convert forest areas to other uses, as well as to exploit forests for short-term benefits. In 40 MANAGING THE WORLD'S FORESTS some countries with high population densities (e.g., Bangladesh, El Salva- dor, Haiti, and Nepal), deforestation can definitively be linked to popula- tion growth. Even in these countries, however, political and social issues (e.g., skewed land distribution, agricultural policies, and poverty) influence the magnitude of the effects of population growth on forest resources. Poor families feel obliged to have many children to provide household help and old-age support. The resultant rapid population growth increases reliance on natural resources for household food and energy, thus perpetuating the trend toward deforestation. Poverty is one of the main underlying factors contributing to deforesta- tion. The majority of the rural poor rely heavily on forests and woodlands for income and subsistence. While many traditional rural communities have developed comparatively sustainable forms of resource use, many others are compelled, by circumstances often beyond their control, to exploit for- ests sustainably for short-term gain. At present, it is estimated that poverty affects about 1.1 billion people, 75 percent of whom live in rural areas (World Bank 1990). State of the Economy Poor economic performance, combined with high external debts, pushes countries to exploit forest resources quickly for short-term gains. External debt among developing countries, which grew from about $570 billion in 1980 to $1.2 trillion in 1988, is another underlying factor driving tropical forest conversion (World Bank 1990). Twelve countries that owe 45 percent of total third-world external debt are responsible for more than 70 percent of global annual deforestation. The debt burden provides an inducement to liquidate forest capital for much-needed foreign exchange. Debt service re- quirements also provide a justification for expanding export crop produc- tion into forest areas. Several countries (e.g., Sudan, Liberia, Burundi, Mauritius, and Myanmar [Burma]) are heavily dependent on a few com- modities for foreign exchange. Therefore, pressure to generate foreign exchange earnings has led to an emphasis on quick return and unsustainable land-use practices. As a result, most developing countries have adopted policies that lead to forest conver- sion to agriculture and short-term exploitation of forest capital. CONCERNS Economic Concerns As experience indicates in many developing countries, extensive con- version of forests and woodlands causes substantial economic losses at local and national levels. Increased sediment deposits resulting from soil runoff from clearings in higher-elevation areas may increase annual floods and DEFORESTATION: PROBLEMS, CAUSES AND CONCERNS 41 reduce growing seasons in cultivated floodplains. Deforestation in upland watersheds often causes flooding of lowland settlements, displaces pop- ulations, and reduces food production (as in Thailand, Bangladesh, and Madagascar). Siltation of major river basins associated with watershed deforestation impedes hydroelectric development and endangers commercial fisheries. The conversion of forests to other land uses often involves wasteful destruc- tion of valuable timber-an opportunity cost that many developing coun- tries cannot afford. The conversion of tropical forests in the Brazilian Amazon, for instance, typically results in the destruction of more than $500 worth of salvageable commercial timber per hectare as well as the loss of biodiversity (Browder 1989). When tropical timber is harvested for export, the revenue that gov- ernments collect from logging companies is much lower than it should be. In the Philippines, official government revenue from log harvests amounted to only 16 percent of actual timber rents, representing a loss exceeding $850 million between 1979 and 1982 (Repetto and Gillis 1988). Substantial but unquantified loss of nonwood forest products also occurs every year as a result of deforestation. In developing countries, accelerating forest degrada- tion threatens industrial and fuelwood production as well. The estimated economic cost of tropical forest depletion ranges from 4 to 6 percent of the gross national product (GNP) in major timber-exporting countries, suffi- cient to offset the economic gains of forest exploitation (Miller, Reid, and Barber 1991). Social Concerns Extensive deforestation in the tropics has serious social consequences, especially for indigenous communities and the rural poor. Over the past several decades, large areas of tropical forests have been brought under government ownership, overriding traditional rights of forest control in many countries. Indigenous communities (including tribal groups) throughout the tropics have been displaced by shifting cultivators, ranchers, and commer- cial loggers. Social concerns about the cultural survival of indigenous groups in the tropics and concerns about the loss of traditional knowledge of forest species and genetic resources that have important economic applications are both altruistic and self-interested. Population pressure, combined with agricultural expansion, has led to reduced supplies of locally available fuelwood in many parts of the dry tropics. Fuelwood depletion counters many positive benefits of rural devel- opment, as women and children spend more time (daily throughout the year) gathering firewood from more distant sources and less time in other vital activities. Poor eating habits often result from fuelwood depletion. Children spend less time in school, and animal dung and crop residues used to fertilize crop fields are diverted to household energy use. Finally, in 42 MANAGING THE WORLD 'S FORESTS many tropical areas undergoing rapid agricultural settlement, extensive deforestation has led to the spread of contagious diseases. Environmental Concerns Extensive deforestation is associated with a loss of biodiversity, climate change, threats to the cultural survival of indigenous population, degrada- tion of watersheds, and desertification in the dry tropics. Because the effects of these environmental externalities are widespread and often far away from the area undergoing deforestation, markets cannot be relied on to send sig- nals to curb deforestation. Instead, some form of collective action, market incentives, or regulations are generally required to reduce these environ- mental problems. Loss of Biodiversity Although tropical moist forests cover only 9 percent of the earth's land surface, they represent the main repository of its biotic resources, contain- ing about one-half of the 1.4 million named species in the entire world biota (estimates of the total number of species range from 5 to 30 million). As mentioned earlier, a single hectare of rain forest typically contains between 100 and 300 different tree species-nearly one-half of the number found in all North America. Between 40 and 60 percent of tropical species in some countries are endemic to specific locations (Gentry, in Myers 1988). The destruction of even small forest areas, therefore, can eliminate entire species. At present, an estimated 10,000 species are extinguished each year because of tropical deforestation (Wilson 1988). Many of the .threatened species have economic value as important sources of food, medicine, genetic material for crop hy- bridization, and other marketable products. Concern about the loss of biodiversity arises from economic, social, and ethical values that society associates with biological resources. Although people disagree about the degree of reduction and the value of biodiversity, a consensus exists that maintaining biodiversity requires protection of habi- tats and that, in light of insufficient knowledge, conserving threatened bio- logical resources and habitat is prudent. Climate Change Because forests help to regulate global atmospheric temperatures and the distribution of moisture, deforestation influences global climate. More- over, about 55 percent of the earth's organic carbon stock is stored in tropi- cal forest biomass (Whitaker and Likens 1973). The reduction of the world's forests, mainly in the tropics, impairs their important atmospheric functions as carbon sinks, and the combustion of forest biomass releases carbon diox- DEFORESTATION: PROBLEMS, CAUSES AND CONCERNS 43 ide into the atmosphere, contributing significantly to the buildup of green- house gases. At present, the atmosphere annually absorbs 1 to 3 billion metric tons of carbon from tropical forest burning and 5.6 billion metric tons from use of fossil fuel, mainly in the industrial countries (Woodwell 1990). In recent years, the share of the buildup of greenhouse gases in the global atmosphere attributable to the burning of tropical forests has been expand- ing more rapidly than the share of fossil fuels use. Although global warm- ing through an atmospheric greenhouse effect has yet to be verified, valid concerns about climate change associated with release of carbon from both fossil fuels and tropical forests point up the need to find more efficient ways to use and conserve these resources. Desertification Desertification causes loss of biological productivity through various forms of land degradation, including erosion and soil and water saliniza- tion. The consequences of desertification include a reduction in the amount of fauna and flora, loss of capacity to retain water, reduced soil fertility, and further land and water erosion. Drylands (including deserts and arid and semiarid areas, especially in North Africa and the Middle East) account for about one-third of the earth's land surface and support more than 700 mil- lion people. Prolonged dry periods, increased human pressure (especially related to fuelwood gathering), and overgrazing further retard natural re- generation and promote desertification (as in Morocco, Algeria, Tunisia, Afghanistan, and Pakistan). Deserts encroach on adjacent arid areas through overgrazing, excessive tree cutting, and clearing of marginal lands. In more humid areas, increased human activity also promotes desertification. Watershed Degradation Many people in developing countries live in villages that depend on the watershed functions provided by forests. Loss of forests can contribute to the degradation of watersheds, with downstream effects within a country and beyond its political boundary. Increasing deforestation in the Hima- layan watershed has caused serious social, economic, and environmental losses in northern India, Nepal, and Bangladesh. In Amazonia, where an estimated one-half of all rainfall results from evapotranspiration, extensive deforestation disrupts local hydrological regimes and promotes soil degra- dation through increased surface temperatures. Also, extensive deforesta- tion quickly disrupts the nutrient cycle in such ecosystems, leading to local soil depletion and reduced agricultural productivity. Weak development policies and the persistence of social inequities that are often aggravated by the development process, foster conditions that encourage destructive de- forestation of shifting cultivators. 44 MANAGING THE WORLD'S FORESTS SUMMARY Deforestation is increasing in the tropics. Although not all forms of deforestation are negative, serious social, economic, and environmental costs are associated with destructive deforestation. Agricultural expansion, over- grazing, fuelwood gathering, commercial logging, and infrastructure and industrial development contribute to uncontrolled deforestation. But these immediate causes are driven by population growth, rural poverty, market and policy failures, and underdevelopment. Humans play an important role in modifying natural systems. Human interaction with the environment is influenced by economic and sociopolitical factors. In the case of forestry, social factors help shape demographic pres- sure on forest resources, ~storical property rights governing access to forest resources, and attitudes toward forest use. Economic forces influence which forest outputs are marketed and which are used for subsistence needs, how important the forest sector is in the national economy, and how income resulting from forest activities is distributed. Political factors affect how much government intervenes in the pricing and extraction of forest prod- ucts, which interest groups receive favorable treatment, and which forest outputs will be provided as public goods. The way in which these social, economic, and political factors interact with one another and with the natu- ral systems determines whether there is sustainable use of forest resources or destructive deforestation. REFERENCES Browder, John 0 . 1988. "Public Policy and Deforestation in the Brazilian Amazon." ln Public Policies and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. Cambridge, Eng.: Cambridge University Press. Browder, John 0., ed. 1989. Fragile Lands of Latin America: Strategies fo r Sustainable Development. Boulder, Colo.: Westview Press. FAO. 1982. Tropical Forest Resources. Rome: Forest Resources Division, Forestry Department. _ __ . 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome: Forest Resources Division, Forestry Department. Gentry, A.H. 1986. "Endemism in Tropical versus Temperate Plant Communities." In Conserva- tion Biology: The Science of Scarcity and Diversity, ed. M.E. Soul. Sunderland, Mass.: Sinauer Associates. Gillis, Malcolm. 1990. "Forest Incentive Policies." Unpublished report. Houghton, Richard A. 1990. "The Global Effects of Tropical Deforestation." Environmental Science and Technology 24(4):414-24. Leonard, H. Jeffrey, ed. 1989. Environment and the Poor: Development Strategies for a Common Agenda. New Brunswick, N.J.: Transaction Books. Mahar, Dennis J. 1989. Government Policies and Deforestation in Brazil's Amazon Region. Washington, D.C.: World Bank. Miller, Kenton R., Walter V. Reid, and Charles V. Barber. 1991. "Deforestation and Species Loss: Responding to the Crisis." In Preserving the Global Environment: The Challenge of Shared Leader- ship, ed. Jessica T. Mathews. New York: Norto n. Myers, Norman. 1988. "Tropical Forests and Their Species: Going, Going . . . ?" In Biodiversity, ed. E.O. Wilson. Washington, D.C.: National Academy Press. DEFORESTATION: PROBLEMS, CAUSES AND CONCERNS 45 Repetto, Robert, and Malcolm Gillis, eds. 1988. Public Policies and the Misuse of Forest Resources. Cambridge, Eng.: Cambridge University Press. Wilson, E.O. 1988. "The Current State of Biodiversity." In Biodiversity, ed. E.O. Wilson. Washington, D.C.: National Academy Press. Whitaker, R.H., and G.E. Likens. 1973. "Carbon in the Biota." In Carbon and the Biosphere, eds. G.M. Woodwell and E.V. Pecan. Washington, D.C.: U.S. AEC. Woodwell, George M. 1990. "The Role of Forests in Determining the Composition of the Atmo- sphere." Unpublished report. World Bank. 1990. World Development Report. Washington, D.C. WRI (World Resources Institute). 1987. World Resources 1987. New York: Basic Books. 4 Biological Diversity and Forests Daniel B. Botkin and Lee M . Talbot T his chapter reviews the issues related to biological diversity and forests, explaining the meaning and importance of biological diversity locally, nationally, and globally. It then discusses the effects that changes in land use have had on biological diversity in the past and might have in the future. It also discusses the current rate of change in biological diversity and the increasing threats to diversity, including those imposed by forest uses. Finally, the chapter shows how the international community can apply mea- sures to conserve biological diversity in forests. The primary focus here is on forests of low latitudes, especially the various types of tropical moist and dry forests, but the chapter also includes some discussion of low-latitude mountain forests and temperate forests of central Europe and the former Soviet Union. BIOLOGICAL DIVERSITY In September 1990 the Los Angeles Times reported that American physi- cians were shocked to discover that a chemical important in human brain activity and valuable in the control of high blood pressure had exactly the same formula as a poison used on the tip of arrows by Amazonian Indians and obtained from one of the Amazonian rain forest trees. Here was another reminder of the potential utility of diverse life forms in tropical rain forests. These threatened forests, among the most biologically diverse of all the earth's natural areas, continue to provide us with a wealth of economically useful products. One of the most striking features of life on the earth is its great diversity of living things. Biological diversity has three main components: species, 47 48 MANAGING THE WORLD'S FORESTS genetic, and ecosystem. Species diversity and genetic diversity have been the primary, and often only, focus for biological diversity conservation. But these two linked considerations are affected by other aspects of diversity that are normally defined under the heading of ecosystem diversity. THE FUNDAMENTAL REASONS TO CONSERVE BIOLOGICAL DIVERSITY1 The Underlying Rationale The underlying rationale for the conservation of biological diversity is that, once lost, a species cannot be regained; its genetic makeup is gone forever. In the past we viewed such a loss from the standpoint of the indi- vidual species involved, and the potential effect of that species on the con- tinuing evolution of other ones. Now, when we can transfer genes from one organism to another, we recognize that not only do we lose forever that species, but also we lose a remarkable collection of unique and potentially immensely useful genes. Many species evolve over long periods of time, and an area of high biological diversity is the result of complex ecological and evolutionary processes, often involving unique histories of local habitats that, once lost, cannot be reproduced. Unlike a machine of our own making, such as an automobile that can easily be replaced and improved upon as technology advances, biological diversity is the result of long processes that we are only beginning to understand. In this sense, biological diversity is a fragile re- source that must be managed conservatively. Species evolve in specific loca- tions. Sometimes, species migrate worldwide, far from their point of origin. In other cases, species remain highly localized and the set of species found in one part of the world may be very different from those found in another. Whereas in the past the phrase conservation of biological diversity was often erroneously used to mean simply protection from any use, we recog- nize today that a purpose of conservation is to ensure future capacity for use. For example, the World Resources Institute defines conservation of biodiversity as "the management of human interactions with the variety of life forms and ecosystems so as to maximize the benefits they provide today and maintain their potential to meet future generation's needs and aspira- tions" (Reid and Miller 1989). Extinction of species is a natural process. In a finite world with some degree of risk, the eventual fate of every species is extinction. What is novel or unnatural about the modern situation is the rate of extinctions, not ex- tinctions per se. Over the history of life on the earth, the natural rate of extinction has averaged slightly less than the rate of evolution of new spe- 1 Important insights and advice were provided by Thomas Lovejoy, Kenton Miller, Peter Raven, Michael Soule, and Edward Wilson. BIOLOGICAL DIVERSITY AND FORESTS 49 cies, so that overall species diversity has, on the average, increased (Myers 1979). There have been periods of comparatively large declines in diversity and periods of comparatively rapid increases. But it is well known that the rate of extinction has rapidly increased since the rise of civilization, and increased even more rapidly since the Industrial Revolution. The current rates of extinction among many groups of organisms including birds and mammals are estimated to be as much as a thousand times what they would be in nature undisturbed by human influences (Wilson 1991), perhaps a thousand times those of the past tens of millions of years (Raven 1987). Comparable rates of extinction have not occurred for 65 million years, when the dinosaurs disappeared and mammals came into the ascendancy. Some authorities estimate that 25 percent of the world's animal and plant species that existed in the mid-1980s may be extinct by the year 2015 or soon there- after (Raven 1988a, 1988b). If we consider only those extinctions caused primarily by tropical deforestation, we may lose between 5 and 15 percent of the world's total species between 1990 and 2020 (Reid and Miller 1989). If there are roughly 10 million species on earth, this rate of loss would account for 15,000 to 50,000 species per year, or 50 to 150 species a day, far .. exceeding any known earlier rates of extinction of animals and plants. Some believe that this rate even exceeds that which occurred during the mass extinction of the dinosaurs (Wolf 1987). These concepts provide insight into a potential flaw in the idea of com- pensatory wildlands (i.e., that when an area of wildland is destroyed by a development project, the loss may be compensated for by protecting an area of equal size and similar ecosystem) and help demonstrate that biological diversity is not a readily exchangeable commodity. Species found in one parcel of land may be different from the species found on another, and a "compromise" that results in the exchange of one parcel for another of equal size may not lead to conservation of the same biological diversity as present in the original parcel. Whereas a farmer may exchange one pasture for another with the government so that a highway can follow a straight line, the same kind of exchange for biological diversity might lead to more rather than fewer undesirable extinctions. Four Basic Justifications Biological diversity has value for four basic reasons: utilitarian, aes- thetic, moral, and ecological. The utilitarian justification for biological diversity means that there are products to be obtained from natural ecological systems that can provide direct economic or social benefits. The pharmaceutical chemical important in controlling hypertension-and used in the tips of arrows in the American tropics- is only one example of the utilitarian reason to value the biological diversity of forests. Fully three-quarters of the people on earth, most of them in the developing nations, depend directly on plants as sources of medicines. At the same time, in the developed nations a substantial part of so MANAGING THE WORLD'S FORESTS the multimillion dollar pharmaceutical industry has been based on materi- als derived from natural sources (plants, vertebrate and invertebrate animals, and microorganisms). For example, some 25 percent of the prescriptions dispensed in the United States between 1959 and 1973 contained active ingredients extracted from vascular plants (Principe in press). Many phar- maceuticals are derived from plants or, if now produced artificially, were discovered as compounds in plants such a quinine and aspirin (the latter from willow bark). Forests are especially valuable for such products. Only a very small percentage of plants have been examined for their uses for drugs and other products, and the high diversity of tropical forests holds great promise for the likelihood of yet-undiscovered pharmaceuticals and other products (e.g., oils and fats) needed by modern industrial societies. Tropical forests (wet and dry, open and closed) already have yielded a long list of original varieties of agricultural plants used for food and plants that provide chemical substances for medicinal and other uses. Well-known medicines derived from tropical forests include anticancer drugs from the rosy periwinkle (Catharanthus roseus), steroids from Mexican yams (Dioscorea composita), and antihypertensive drugs from serpent-wood (Rauwolfia serpentina) (Reid and Miller 1989). Many species that were the origins of present-day horticultural crops, nuts, resins, and other products also have . come out of these forests. The same is true for many species of flowers that are now sold commercially. Of 275 species found in one hectare in a Peru- vian moist tropical forest, 72 yielded products with direct economic value. Of 842 individual trees, 350 yielded products with direct economic value. The market price of the fruit tree and palm products was estimated to average $650 per year, and the net annual revenue at $400 per hectare (Pe- ters, Gentry, and Mendelsohn 1987). Genetic diversity, much of it from forests, has greatly enhanced produc- tion. In the United States, from 1930 to 1980, plant breeders' use of genetic diversity accounted for at least one-half of a doubling in yields of rice, barley, soybeans, wheat, cotton, and sugarcane; a threefold increase in to- mato yields; and a fourfold increase in yields of corn, sorghum, and potato (OTA 1987). Most of the food production in the United States is from plants that originated in foreign lands-a fact that underscores the international dimension of germplasm use. Many indigenous people find the forests essential to sustain life; forests provide their food; wood for shelter, tools, and fuel; wildlife and plant materials for clothing; and native plant materials for medicine. A reduction in biological diversity can reduce the capability of forest areas to support these people. For poor, indigenous people who depend on forests, there may be no reasonable replacement for these benefits other than continual external assistance, which most development projects are supposed to elimi- nate. For urban residents, these forest benefits may not be apparent, or may become apparent only after the forests' biodiversity has been irreparably impaired. BIOLOGICAL DIVERSITY AND FORESTS 51 Biological diversity provides a variety of other utilitarian benefits, of which pollution control is one. For example, carbon dioxide is removed by vegetation, sulfur dioxide is removed by macroscopic plants, carbon mon- oxide is reduced and oxidized by soil fungi and bacteria, and nitrogen oxide is incorporated into the biological nitrogen cycle (Pimentel 1982). Wild ver- tebrates, invertebrates, and microorganisms play major roles in pollination of wild and crop plants, germination, dispersal of seeds and other propagules of plants, soil processes, and nutrient cycling, all of which are vital not only to the maintenance of the ecosystems of which these organisms are a part, but also to human welfare (Talbot 1987). Biological diversity in the form of wildlife has direct utilitarian value for food, other animal products, and income from sport hunting and tour- ism. In many parts of Africa, wild ungulates often offer substantially higher productivity than introduced domestic livestock (Child 1990; Talbot 1972; Talbot et al. 1965). Tourism based largely on wildlife in protected areas is a major and growing source of income for many developing countries. It is, for example, the single largest source of foreign income for Kenya. Although much of the initial wildlife-based tourism was based on the larger forms of wildlife that flourish on the more open savanna lands, increasing attention is now being given to tourism based on the whole array of forest plants and animals. Forests have many indirect uses, too. Forests retard soil erosion, espe- cially in areas of high rainfall, high rates of tectonic uplift, and soft bed- rock-conditions that are found in low-latitude mountainous regions, such as those of India and Nepal (Sidle, Pearce, and O'Laughlin 1985). Forests also stabilize water supply and runoff. This benefit of forests was known to the ancient Greeks, repeated in the 19th century with the awakening of concern with deforestation in Western nations, and quantified in the past 30 years through studies of watersheds in North America. Forests also im- prove air quality and help to maintain regional climates and especially pat- terns of rainfall (Lettau et al. 1979; Salati et al. 1979; Shukla, Nombre and Sellers 1990). It has been calculated, for example, that more than half of the precipitation in the Amazon region is generated by the forests (Villa Nova, Salati, and Matsui 1976; Salati, Marques, and Molion 1978), and some specu- late that the climate in southern Brazil would be so altered by deforesting the Amazon that agriculture might become impossible there (Raven 1991). Because of the role of the Amazonian forests in maintaining climatic condi- tions necessary for their own continuation, complete and rapid destruction of the Amazon tropical forest could be irreversible (Salati, Vose, and Lovejoy 1986; Shukla et al. 1990). The aesthetic justification for biological diversity refers to the value that people place on seeing, hearing, touching-experiencing:-nature and its diversity of life forms. Aesthetic interest, of course, leads to tourism, film- making, and other activities from which an economic return can be ob- tained. Aesthetic appreciation of nature is physiologically deeply rooted in 52 MANAGING THE WORLD'S FORESTS people. Some researchers suggest that appreciation of nature can have im- portant health implications. The moral justification for biological diversity refers to the belief, as stated in the U.N. General Assembly World Charter for Nature 1982, that species have a moral right to exist. Consequently, in their role as global stewards, people have an obligation to assist the continued existence of species, that is, to conserve biological diversity. Although this perspective may not seem to have an economic connection, in fact more and more citizens of the world are asserting the importance of this moral point of view, and more and more people are taking actions to defend this moral position. A recent ex- ample of how the moral argument has had an economic effect is the refusal of many Americans to purchase tuna fish caught by fishing fleets that killed porpoises as part of the fishing. In response, one major tuna company has advertised that it will sell only tuna fish taken without harm to porpoises. Other examples are the boycotts of furs, teak, and ivory. Today, people often contribute to wildlife conservation programs with- out expecting to see the animals, merely from a desire to know that the animals are there. Because moral concerns about biological diversity will probably increase in the future, more economic consequences are likely. Any future forest management policy will encounter groups that wish to conserve biological diversity of forests, and it seems counterproductive to attempt to set policy without understanding the points of view of these groups-in other words, without understanding moral arguments for the conservation of biological diversity. Rather than creating a continual series of confrontations between people who take a moral position in support of biological diversity and those who are intent on harvest, it seems eco- nomically prudent to understand this justification and respond to it constructively. The ecological justification for biological diversity means that diversity is important to the persistence of ecological systems, including forest ecosys- tems. To explain this point, we need to consider the simple question, What is required to sustain life within a forest? We tend to associate life with individuals, and therefore to imagine that the continuation of life depends solely on individuals and their reproductive capacity. But no single species exists that creates its own food from inorganic materials and completely decomposes all its own wastes. Nor does any species single-handedly main- tain all the necessary characteristics of its habitat for its own persistence (Slobodkin et al. 1980). A plantation of rubber trees in a germ-proof green- house would die out and be unable to regenerate without bacteria, fungi, and soil animals to decompose dead material from the trees. Only a set of species of different kinds, interacting together, complete all these processes. Only a set of species and the local environment function to sustain life. The minimum system that can provide cycling of chemical elements and the flow of usable energy necessary to sustain life is called an ecosystem, which is a set of interacting species and their local, nonliving environment. BIOLOGICAL DIVERSITY AND FORESTS 53 To maintain economic production of forests, a forest ecosystem must be sustained. The characteristics of such a minimum ecosystem are currently a source of controversy regarding deforestation of the remaining original forests of the world. The practical consequence of these ideas is that learn- ing more about the minimum ecological requirements to conserve intact forests and their biological diversity, and to restore degraded forests, is worthwhile. Until we know those minimum requirements,•it is prudent for us to treat forests more carefully than we do at present, and to assume that existing characteristics, including all species, are necessary unless proven otherwise. The old idea of forest production was predicated more or less on the notion that the only things a forest needed were inorganic soil, fertilizers, water, and seeds. In fact, most trees survive because they are part of a complex set of interacting species. Many trees depend on a number of spe- cies of fungi in the soil or attached to their roots in order to take up nutri- ents from the soil. Leguminous trees have nitrogen-fixing bacteria living in nodules on their roots; these bacteria convert molecular nitrogen, an essen- tial nutrient, into compounds usable by trees. Free-living soil bacteria fix nitrogen from the air and make it available to trees; other soil bacteria affect the availability of other elements. Small animals in the soil, such as earth- worms and termites, improve the physical characteristics of the soil, stimu- lating tree growth and improving the conditions for the growth of beneficial soil bacteria. As every schoolchild knows, trees depend on birds, bats, bees, and other insects for pollination of flowers and on many kinds of verte- brates for the distribution of seeds. The survival and successful growth of vegetation in forests are therefore the result of a complex set of interactions among species. Hence, a diversity of species is essential to the persistence of forest ecosystems. High biological diversity means greater ecosystem com- plexities, because there are more, and more highly specialized, interdepen- dencies in high-diversity than in low-diversity systems. And the manipula- tion of a high-diversity system requires much more caution to maintain it in the future. Although we know that a set of species is required to sustain life, we know very little about natural ecological systems in general and tropical forests in particular. One of the central controversies about the biological diversity of forests concerns the minimum set of species required for a forest ecosystem to persist. In the past, the typical management approach was to put the burden of proof onto nature and conservationists: forest resources were removed and forest habitats modified; only after the fact, when problems arose, was consideration given to loss of species. When forest resources were abundant and timber supply seemed infinite, that course of action may have seemed economically plausible although it never really was. As forest resources vanish, the burden of proof must be placed on those who plan to exploit forest resources-to demonstrate that a plan of action will not decrease ecosystem stability or add to extinction of species and decline in biological diversity. 54 MANAGING THE WORLD'S FORESTS For example, there is evidence that the supposedly high economic value of a moist tropical forest is completely lost when the forest is converted to other uses. What we do not know is how much we can manipulate that forest before unknown ecological processes drive it past a point of no re- turn, or long-delayed return. It is prudent to be cautious in supporting projects that lead to the manipulation of forests until we know, or at least can approximate, the level of tolerance of the ecosystem. The ecological justification also leads to an extension of the moral justi- fication. If we have a moral obligation to conserve species, we also have an obligation to conserve ecosystems. This assumption underlies the approach, for example, of the U.S. Marine Mammal Protection Act of 1972, the UNESCO Man and the Biosphere Program in ecological reserves, and many other recent conservation programs. Direct Global Connections Many of the values of biological diversity just discussed clearly have implications for the international community. For example, pharmaceuticals are not only of international economic importance, but some can contribute to human health and welfare worldwide. During the 1980s a new global perspective on life and on ecology devel- oped. We now understand that life has changed the global environment and that these life-induced changes, which began more than 3 billion years ago, have affected the atmosphere, the oceans, and solid sediments. Life appears to be a planetary phenomena, and all life on earth seems interconnected. There are two kinds of global connections among species: migratory and chemical. Global Migratory Connections Some species that migrate over large distances are important for the continuation of other species. For example, migrating birds that winter in the tropics and nest in middle and high latitudes can be important in polli- nation and seed dispersal in both nesting and wintering habitats. But these same birds depend on tropical or subtropical ecosystems for their winter survival. Destruction of tropical forests can lead to the extinction of these birds, which, in turn, could threaten some temperate and high-latitude veg- etation. As another example, outbreaks in Canada of spruce budworm, an especially troublesome problem for trees of commercial value, may be de- creased by the abundance of warblers that migrate from the tropics (Holling 1988). BIOLOGICAL DIVERSITY AND FORESTS 55 Global Chemical Connections Life affects global cycles of many chemical elements that are, in turn, important for the persistence of life on earth. Local and regional biological diversity can influence global chemical cycles. The possibility of global warming reveals the global chemical interrelat- edness of life. Global warming refers to a human-induced climate change resulting from the release of carbon dioxide, methane, and other green- house gases from the burning of fossil fuels and from deforestation. De- forestation, and subsequent decomposition of forest organic matter, adds carbon dioxide to the atmosphere. Decomposition of certain moist soils releases methane, as does digestion of wood by termites. Most plants live a short time and do not store carbon for long. Although algae in the ocean remove large quantities of carbon dioxide from the atmosphere, most al- gae die and decompose rapidly, returning the carbon to the atmosphere. Trees, like all green plants and algae, remove carbon dioxide from the atmo- sphere during photosynthesis, so that this removal is part of their process of growth. As the longest lived of all vegetation, trees are capable of storing carbon for decades and centuries, and it is this carbon storage capability that is so important in the global cycle of carbon and in the possibility of global warming. Enough carbon is stored in forests to have a significant impact on the entire atmosphere. At present, anthropogenic activities are adding more than 5 billion metric tons of carbon to the atmosphere each year. Climatolo- gists project that this rate of increase could warm the earth by 2 to 6 degrees centigrade by the end of the next century. There may be as much as 500 billion metric tons of carbon stored in live vegetation (most of which is stored in forests), and a substantially greater amount stored in forest soils. If these estimates are accurate, an increase of 1 percent per year in the net worldwide deforestation rate could double the carbon dioxide released by human activities, assuming that all organic matter in the trees decomposed rapidly. (This assumption, of course, may be wrong because significant amounts of the carbon in trees may end up in buildings or in newspapers in somebody's attic.) The release of carbon dioxide could be much higher if the soils of the cleared forests decompose rapidly. Such soil decomposition is a characteristic result of deforestation. Thus maintenance of forests can be an important part of the solution to the greenhouse effect, and, conversely, deforestation can have a significant negative effect on the future climate of the earth. The conservation of organic matter in forest trees and in soil requires maintenance of forest ecosystems. Because some biological diversity is nec- essary to maintain these ecosystems, the maintenance of biological diversity in forests can have a global effect on chemical cycles. We have already 56 MANAGING THE WORLD'S FORESTS explained that our current state of knowledge is not sufficient to determine exactly what set of species is the minimum required to sustain any forest ecosystem, and we have argued that prudence dictates preserving existing diversity unless evidence were to become available that such diversity is not essential. The global perspective supports the ecological arguments for the conservation of biological diversity and for the reversal of deforestation. THE RELATIONSHIP BETWEEN FORESTS AND BIOLOGICAL DIVERSITY, THE TRENDS IN BIOLOGICAL DIVERSITY RELATED TO FORESTS, AND THE CAUSES OF THE CURRENT SITUATION A large fraction of the earth's biological diversity lives in forests or depends on forests, especially tropical forests. About one-half (some esti- mates place the figure as high as 90 percent [Lovejoy 1988]) of the world's species live in tropical rain forests, even though these forests occupy only about 7 percent of the land area. It is estimated that 2,600 species of birds, about 30 percent of the world's total, depend on tropical forests for some part of their life cycle (Diamond 1985). In one study, 10 hectares in Borneo were found to contain 700 tree species, the same number as found in all of North America. In another study, 570 plant species were found in one hect- are near Kuala Lumpur in Malaysia. More typically, 40 to 100 species of trees can be found in one hectare in South American tropical rain forests (Reid and Miller 1989), and that figure rises to 300 per hectare in Amazo- nian Peru (Raven 1990). In comparison, 10 to 20 might be found per hectare in eastern North America, and as few as 1 to 5 in the boreal forests. Thou- sands of hectares of Siberian boreal forests have a single tree species, Sibe- rian larch. The great diversity of tropical rain forests, along with the current rapid development of regions containing these forests, points up the need for emphasis on the conservation of biological diversity in these areas. Many of the species found in tropical forests occur only in the small area where they evolved. Such species, called endemic, are especially fragile resources because of their localized distribution. The many tree species in tropical rain forests provide habitats for many other organisms. For example, commercial harvests of the Brazil nut (Bertholletia excelsa) are from the wild. This tree is pollinated by several Euglossine bees. The male bees gather organic compounds from certain species of tree-dwelling orchids to attract females before mating, but they depend on other flower species for food. Agoutis, large forest rodents, pro- vide the only known natural mechanism for opening the Brazil nut seed to allow dispersal and germination. Consequently, the ecological processes basic to the Brazil nut industry involve the habitat for the Euglossine bee, includ- ing nesting and feeding areas, certain orchids and the trees on which they grow, insects that pollinate the orchids, and agoutis. Persistence of the Bra- zil nut requires a forest habitat large enough to support agouti and complex enough to support the tree, agouti, and pollinating insects. As long as the BIOLOGICAL DIVERSITY AND FORESTS 57 Brazil nut exists in the forest, it provides a food base and a habitat structure that increases biological diversity. Thus there is a strong relationship be- tween intact forest ecosystems and high biodiversity. Forests also provide habitat for many kinds of species that are not typi- cally associated with a forest habitat. For example, mangrove swamps are important habitats for juvenile stages of commercially important fin fish and shrimp. Indonesia's annual catch dependent on mangrove areas is val- ued at US$194 million. The best way to conserve biological diversity in forests is in natural, wild, intact ecosystems. The next best is in multiple-use areas; third best in silvicultural areas including agroforestry and urban forest gardens. Finally, for extreme cases, such as where habitats have been completely destroyed and species would become extinct if left in the wild, zoos and botanic gar- dens (as part of their broader mission related to research, captive breeding, propagation, education, and public awareness) can maintain genetic rem- nants of wild species. Captive breeding can maintain a species until reintroduction can be achieved. The dramatic reduction of forests is of worldwide concern, particularly because of the consequent loss of biological diversity. The World Bank, United Nations agencies, other intergovernmental bodies, and government and nongovernment organizations throughout the world now routinely list the threat to tropical forests as one of the world's most serious environmen- tal problems. New studies indicate that this threat is far more serious than was previously thought; according to the World Resources Institute (1990- 91), "The world is losing up to 20.4 million hectares of tropical forest annu- ally-79 percent over FAO's 1980 estimate." This is about 56,000 hectares a day, an area almost equivalent to the loss of a nation the size of Britain or Uganda each year. At current rates of tropical deforestation-on the order of 1 percent or greater per year-all intact tropical forests will be cut during the 21st century (Houghton 1990). At current rates of global deforestation, one-quarter of the world's species may be extinct from this cause before the middle of the 21st century (Reid and Miller 1989; World Resources Institute 1990-91; Raven 1988). In Africa, with the exception of the Congo basin and a few parts of West Africa, intact tropical moist forests consist of relatively small and usually rapidly dwindling remnants. These remnants, and the few larger forests (e.g., in Zaire and Gabon), are surrounded by areas of secondary or de- graded forest, often where logging has opened the original forest, and cul- tivation and grazing have followed the loggers. In tropical Central and South America large, moist forests remain, but they are disappearing at about 1 percent per year. Some large, intact tropical forests remain in Southeast Asia, such as those in Indonesia, but most are subject to intense use. Much of the loss of tropical forests is attributable to factors outside the traditional forestry sector. Increasingly, however, scientists are questioning whether sustainability of commercial logging in natural tropical moist for- 58 MANAGING THE WORLD'S FORESTS ests has ever been achieved, and whether it is, indeed, ever possible in other than plantation situations. With increasing stridency environmentalists throughout the world are raising the same questions. They point to the unquestionable loss of tropical forests following lumbering operations, and they say that the claim of "sustainability" is a smokescreen to cover destruc- tion of irreplaceable forests for financial growth. THE RELATIONSHIP BETWEEN FOREST MANAGEMENT TECHNIQUES, INCLUDING COMMERCIAL LOGGING, AND BIOLOGICAL DIVERSITY Forms of Commercial Logging Commercial timber exploitation (as distinguished from subsistence extraction by forest-dwellers) generally takes the form of clear-cutting, selective cutting, enhancement or improvement forestry, or plantations. Plan- tations simply replace the original forest or other previous vegetation, typi- cally with a single species planted at regular intervals to form a grid. Enhancement or improvement forestry seeks to increase the growth rate or numbers of the desired (i.e., economically valuable) tree species by plantings and by cutting or poisoning the "weed" (i.e., uneconomic) species. In either case the effect is to change the original forest into a form of plantation, which is substantially different from the natural forest. The effect of clear-cutting depends on the site, the size and pattern of the clear-cuts, the amount of cut material left on the ground, and the subse- quent treatment of the area. Except for very small cuts surrounded by intact forest, the effect of clear-cutting is devastating to an original tropical forest, at least for many decades. If clear-cutting is conducted simultaneously over a large area, genetic diversity may be diminished, especially when no seed- bearing mature trees are left, or when the habitats of the mature trees are damaged to the extent that mature seed-bearing trees cannot persist. Species diversity declines rapidly in the clear-cut area, although there may be a temporary increase in the local diversity in remaining forest patches as or- ganisms scramble for the remaining habitats. Clear-cutting is carried out in tropical regions to clear the land for agriculture. For example, in the Ama- zon Basin, trees are cut, girdled, or burned; then the cleared land is farmed. Clear-cutting of tropical rain forests has been done in the Philippines and in other areas where the land is simply treated as a resource to be mined. In other areas, w here specific species are of special interest, selecting logging is more common. Selective cutting (including "high-grading" or "creaming") seeks to re- move only a limited number of target trees, such as all those of a given species or those of a particular age of that species. Typically the target species represents a small percentage of the total number of trees present, so that even limited lumbering may effectively remove the species from that BIOLOGICAL DIVERSITY AND FORESTS 59 part of the forest. Where the target species is a key component of the forest ecosystem, even this limited removal may alter the forest in significant ways. In addition, inevitably a substantial percentage (often estimated at more than 50 percent) of the remaining trees and other vegetation will be killed or damaged by the selective cutting and removal of the target timber, and the species composition of the remaining forest will be changed. Consequently, although selective cutting is the least disturbing form of forest exploitation, it still substantially alters the natural forest system. In practical terms, no commercial logging of tropical moist forests has proved to be sustainable from the standpoint of the forest ecosystem, and any such logging must be recognized as using a potentially renewable source as a nonrenewable one- in other words, as mining, not sustaining, the basic forest ecosystem. In fact, the question can be asked whether any extraction has resulted in sustainability of the original level of abundance and productivity, or in the original bio- logical diversity. The methods of cutting and removing logs have an important effect on biodiversity, and in recent decades considerable experience has been ob- tained in techniques that are less damaging to forests than older mechanical methods (Hamilton 1988). The most damaging are bulldozers, tractors, and skidders; the least damaging are aerial cables, balloons, and helicopters in combination with hand felling. International assistance to logging projects should take into account whether the least-damaging methods are to be used. If extraction of timber is permitted, a code of "Best Practices," includ- ing the least-damaging methods of cutting and removal of timber, should be developed. The Question of Sustainability In terms of time-horizons for economic planning and human lifetimes, forests regenerate slowly to their original or predisturbance forms after be- ing severely disturbed. White pine in North American can live more than 400 years, and forests dominated by this species can take that long or longer to reach the biomass and composition of the original forests. The tropical forest of Angkor, Cambodia, which began to regenerate after that city was abandoned in the 15th century, is still structurally different today from nearby original forests. Because the trees that dominate many forests can live so long, it would not be surprising for a forest dominated by such species to take several human lifetimes to reach a stable condition, even under a constant climate. A computer model of forest growth, shown to be realistic and accurate, projects that a forest of the northern temperate zone of North America requires more than four centuries to reach a maturity from a clear-cut (Botkin, Janak, and Wallis 1973). Regeneration of original forests is complicated by climate change. For- ests modify the climate near the ground. Not only is it cooler and more moist under the shade of a forest canopy during the daytime, it is also 60 MANAGING THE WORLD'S FORESTS warmer during the night and during cold seasons. An established forest can persist under climate change, while the same forest may no longer be able to regenerate. Over the centuries the climate can change considerably, so that a forest that developed to maturity in one location in a previous climate may not regenerate, if cut, in the current climate. Some existing forests, perceived to be the kind that will regenerate and assumed to be sustainable, may be remnants of past climatic conditions. For example, some teak forests in Zimbabwe, which regenerate poorly when affected by frost, may persist in the current climate because mature trees protect young ones from frost. Logging mature trees exposes younger ones to frosts, which may kill them. The failure of white pine forests of North America to regenerate may be the result of a similar response by trees to climate change. Such possibilities emphasize the need to establish prior to logging whether the type of forest to be managed can regenerate under current conditions of soils, climate, and vegetation-that is, whether sustainability is possible at all. Forests vary in their susceptibility to logging; tropical rain forests seem to be especially fragile. One reason is that two-thirds of these forests grow on poor and fragile tropical yellow and red soils, which are poor in nutri- ents and, when cleared and exposed directly to sunlight and rain, form a hardened surface from which regeneration of vegetation proceeds poorly. In these forests, most of the chemical elements necessary for life are con- tained in living vegetation and dead organic matter, which decomposes when the forest is cleared by logging; their constituent chemical elements are lost rapidly. Trees in these forests typically have shallow root systems, an adaption to the need to capture nutrients quickly from the shallow soil. Such trees may be more vulnerable when surrounding trees are cleared away. The remaining trees are more vulnerable than deep-rooting trees to windstorm damage and to sun-scalding of their roots. Another reason tropi- cal rain forests regenerate poorly after logging is that many of the seeds of rain forest trees germinate almost immediately and are not stored for long periods in the soil; thus they are not available after logging. Some tropical rain forests do regenerate rapidly from some kinds of natural disturbances, as do, for example, some dipterocarp forests of the Asia-Pacific region. Definitions The question of the extent to which tropical moist forests can be ex- ploited sustainably has been a matter of professional scientific concern for many years, and is an important consideration for the conservation of bio- logical diversity. As already noted, questions have been raised about whether sustainability in commercial logging operations has been achieved, at least in the tropics, and, indeed, whether it is possible under most conditions. Recently, sustainability has come into prominence within the World Bank both in connection with Tropical Forestry Action Plan (TFAP) and with the BIOLOGICAL DIVERSITY AND FORESTS 61 Environmental Assessment (EA) process, particularly in the case of some of the Forest/Environment and Natural Resource Management projects that have a component involving sustainable exploitation of tropical forests. A major reason for the lack of agreement over the concept of sustainability is the definition of the world sustainability itself. There are two main ele- ments in the definition of sustainability in use at present: one refers to sustainability of a forest ecosystem and the other refers to sustainability of the yield of timber. Sustainability of the ecosystem refers to maintaining the integrity of the natural forest in terms of its structure, composition (i.e., species composition and biological diversity), and ecological processes, along with the environmental services it provides. Sustainability of timber yield re- fers to maintaining a yield of timber from the forest area. Sustainability of the Ecosystem The concept and practice of sustainable forestry (in the context of eco- nomically viable, relatively large-scale harvest of timber) were developed in Germany. Subsequently, sustainability of timber harvest has frequently been a stated objective of industrial forest exploitation throughout the world. In some cases in the northern temperate zone such exploitation may have proved sustainable in timber terms, but it is difficult to find cases where sustainability has been maintained in ecological terms and in terms of biological diversity. We know of no cases where enough data have been obtained to be truly informative about long-term ecosystem sustainability. Timber sustainability in the northern temperate zone may be more likely than in the tropics, because temperate tree species have evolved in a frequently disturbed and variable environment, and because there are a small number of tree species and a relatively simple structure and composition of the forest ecosystems involved. Most northern forests have been actively managed (through selec- tive cutting, planting, etc.) for many decades or centuries; hence they are substantially different from their premanagement ecological status and may appear ecologically sustainable when they are not. Whe re old-growth "vir- gin" temperate-zone forests are involved (e.g., in the northwestern United States), opposition to commercial timber exploitation is growing specifically on the ground that it is not compatible with sustaining the ecosystem. Ecological conditions in the moist tropics-especially in terms of diver- sity of trees and associated plants, complex physical structure, soil, and hydrological characteristics- are dramatically different from those in the northern zones. The reasons why tropical forests have so much more diver- sity than northern forests are not entirely clear, but whatever the reasons, the differences are enormous. As a result, industrial forest exploitation is not compatible with maintenance of the ecological integrity of most-more probably, of any-tropical moist forests. 62 MANAGING THE WORLD'S FORESTS Sustainability of Timber Yield Sustaining a yield of timber from a forest area implies maintenance of a forest but not necessarily the original forest. Foresters speak of a "rotation period," which is the time between harvests. Three harvests are generally considered to be the minimum required to determine sustainability in any agricultural system, whether the harvest is of an annual vegetable or a tree. The first harvest establishes a baseline; the subsequent ones indicate whether harvests are steady, rising, or falling. These terms can also be used to distin- guish between a sustainable original harvest and a sustainable disturbance har- vest. In a sustainable original harvest, harvest 3 would equal harvest 1. In a sustainable disturbance harvest, harvest 3 would equal harvest 2. The au- thors of this chapter know of no examples of sustainable original harvest. Because a typical rotation period in a temperate forest (where there is more experience in the attempt to achieve sustainability than in a tropical forest) is 20 years or more, a minimum of 60 or more years would be required to determine whether a harvest appeared sustainable. Data on sustainability in temperate-zone forests are not easily found in open scientific literature. If data exist, they may be in the records of timber companies and forestry agencies. Sustainability of timber yield, even when it is the stated goal of forestry, is rarely achieved. Sustainability of the original forest is especially rare; where forests are grown as sustainable crops, it is a sustainable disturbance harvest, as defined earlier, that is sometimes achieved. As an example, in the state of Michigan more than 7 million hectares of original white pine were clear-cut between 1840 and 1920. The foresters are said to have be- lieved that they would never run out of white pine, because the resource appeared huge, and that by the time the last hectare was cut the first would have regrown. In fact, many of the hectares never regenerated but have become depauperate "stump barrens," open fields of grasses, lichens, and shrubs, where no white pine or any other large trees grow. Where regenera- tion of pines has taken place, the original mature size of the trees is never found. These forests developed over long periods, subject only to natural rates of disturbance by fire and windstorms, and not to clearings of large areas or intense fires produced by the large amounts of fuel, in the form of the parts of the trees considered waste, left on the ground by the loggers. In some of these areas, red pine plantations were established in the 1930s, with great expectations for future sustainable harvest. But as these trees have approached maturity, disease outbreaks in the homogenous, single-species stands have caused serious problems. Thus even in northern temperate- zone forests where sustainability is believe to be achievable there are many counterexamples. In general, sustainability of early successional species is more likely than sustainability of late successional species. In the northern temperate zone, pines and aspen are common early successional species, and these BIOLOGICAL DIVERSITY AND FORESTS 63 species are more likely than others to be maintained in plantations that are harvested regularly. In the moist tropics, some of the most commercially important species are characteristic of light-gaps in the forests, that is, of areas of early or midsuccession. These include the entandrophragmas of Africa, some Dipterocarps of Southeast Asia, and the mahoganies in Central and South America (Sayer 1991). As we have suggested, current knowledge suggests that sustainability of timber is much less likely in the tropics, especially in the moist tropics, than in the northern temperate zones. Plantations offer the best and prob- ably the only proven way to assure such sustainability. Consequently, plan- tations offer the only real option for meeting the future demands for timber (and, in most cases, for fuelwood also) from tropical areas. The degree to which tropical forests, especially moist forests, are sustainable for timber production is a research question that must be studied objectively in the future. If wood production were to be sustainable from natural tropical for- ests-as opposed to plantations or "enhanced" forests-the yield would have to be very low, but the need for the timber enterprise to be financially rewarding may make a low offtake uneconomic. True sustainability from natural tropical forests has yet to be demonstrated. As the 1988 Interna- tional Tropical Timber Organization report stated, "It is not yet possible to demonstrate conclusively that any natural tropical forest anywhere has been successfully managed for the sustainable production of timber. The reason for this is simple. The question cannot be answered with full rigor until a managed forest is in at least its third rotation" (Poore 1988). In practice, the opportunity for a third rotation almost never occurs because the earlier exploitation altered the forest system too much, because the forest manage- ment policy or practice changed, or because the forest was cleared for cul- tivation and settlement. Consequently, most natural tropical forests that have been subject to timber exploitation have ceased to exist, at least as natural forests. The exception may be very low level extraction by indig- enous people, which is discussed later in this book. An additional warning is necessary. Even if a forest appeared sustain- able after three harvests, in the sense that the yield of harvest 3 was equal to or greater than that of harvest 2, the forest might not be sustainable indefinitely. The reason is that some secondary effects of the harvest-such as decreases in the fertility of the soil, decreases in the organic content of the soil, and compaction of the soil-may not be evident after three harvests. The response to logging is better known for temperate forests. Although these forests are usually believed to be more resilient to logging, they un- dergo severe changes after clear-cutting. For example, it has been estimated that the nitrogen available to trees declines by more than half following a clear-cut in a temperate-zone hardwood forest, and that the available nitro- gen may remain below half of the original value for 90 years or more (Aber, 64 MANAGING THE WORLD'S FORESTS Botkin, and Melillo 1978). As a result, timber production is greater for long rotation periods than for short periods (Aber, Botkin, and Melillo 1979). Such damage occurs even if only the main boles of the trees, which are low in nitrogen content, are removed. Any practice that removes leaves, twigs, and roots, where a tree stores most of its nitrogen, leads to greater and longer-lasting damage. The activities of felling, cutting, and removing tim- ber from a forest inevitably result in some compaction of the soil, which is minimized by the most careful techniques, such as the use of balloons, helicopters, and aerial cables to remove the timber. But these methods may be impractical in remote tropical locations, or the forests may not be eco- nomically profitable when the costs of these methods are included in the economic analysis. Regardless of the form of management, any yield from tropical moist forests that is sustainable represents a vanishingly small part of the total existing forestry efforts. The 1988 report by the International Tropical Tim- ber Organization concluded that sustainability was achieved in only 0.125 percent of total area of tropical moist forests managed in theory for sustainability of timber production, and even these management results are hotly disputed (Poore 1988). Sustainability of Other Values Most governments in developing countries view the value of the tropi- cal forests in terms of their potential to earn foreign exchange through log- ging. Most of these governments allow the destruction of this national resource because they. have pressing needs for revenue, and they accept claims that the forestry is sustainable. The many other values of the tropical forests are less visible and often may be obtained only indirectly. But many other values are sustainable under proper forest management. As already mentioned in this chapter and discussed more fully in a variety of other publications (Myers 1979, 1983; Oldfield 1984; OTA 1987; Prescott-Allen and Prescott-Allen 1986; Reid and Miller 1989), biodiversity of tropical forests has great actual as well as potential economic and other values to mankind. Science has barely scratched the surface of knowledge about these forests, and the potential benefits to humanity from research into this kind of complex ecosystem and its component species are incalcu- lable. Nontimber forest products, which already produce substantial eco- nomic returns in many areas, promise substantial increases. Forest ecosystems also provide life support, and, through wildlife and other nontimber forest products, economic livelihood for vulnerable ethnic minorities and other local people both inside and around the edges of the forests. Forests also contribute to regional and even global life support sys- tems by providing a series of ecological services, ranging from watershed protection to amelioration of climate. BIOLOGICAL DIVERSITY AND FORESTS 65 Finally, tourism based on forests offers the possibility of earning valu- able foreign income for government and for private entrepreneurs without destroying the basic forest resource. Properly managed, tourism also can provide substantial economic benefits to people living in or around forests, assisting in their economic development and providing incentives to protect the resource (Kiss 1990). Most nontimber values depend on maintenance of an intact tropical forest ecosystem. It follows that any form of forest man- agement that does not also sustain the ecosystem will not provide most of these benefits. MEASURES USED OR PROPOSED FOR CONSERVING BIOLOGICAL DIVERSITY OF FORESTS There are three general categories of measures for conserving biological diversity of forests: protection of natural or near-natural ecosystems, res- toration and rehabilitation of degraded lands, and ex situ protection of individual species. By far the most important of these is protection of eco- systems, which is probably the only way to assure maximum protection for the full range of biological diversity involved. Under some conditions, mea- sures to restore and rehabilitate degraded lands may conserve biological diversity. These measures range from planting one or a few selected species of indigenous trees to mounting complex efforts to replace a range of the preexisting species of plants and animals. Because abused and degraded lands occupy an ever-increasing area of the earth's surface, these measures are becoming increasingly important, both to restore productivity of lands for direct human use and to conserve some biological diversity. The third category of measures--ex situ protection of species, for example, in zoos, botanic gardens, aquaria, and seed banks-may be the last resort for some species when survival in their natural habitats is no longer possible, but its greatest value is probably in the context of temporary protection with the objective of eventual reintroduction in the wild. Protected Areas Various forms of protection for forest areas are the principal measures that have been used or proposed for conserving biological diversity in for- ests. Types of legally protected status range from national parks and inte- gral nature reserves, where resource exploitation and most human activity other than tourism are excluded, to forest reserves that are maintained pri- marily to maintain forest resources for future harvesting. At present, there are 4,500 legally protected biological conservation areas around the world covering 4.9 million km2-about 3.2 percent of the earth's land area-and including all types of ecosystems. Tropical Africa has 860,000 km2 of totally 66 MANAGING THE WORLD'S FORESTS protected areas, the Central and South American tropics together have a total of 768,000 km2, and Indonesia and Malaysia have 357,000 km2 (Reid and Miller 1989). In all, less than 5 percent of tropical forests lie in protected parks and reserves (Brown 1985). Low as these figures are, however, it is estimated that only 300,000 km2-some 15 percent-of these tropical parks and reserves are actually strictly protected. The rest are used with or without government consent for exploitation of resources including timber, fuelwood, forage for cattle, soils and water for cultivation, and wildlife for hunting. Furthermore, very few of the existing protected areas are self-sufficient ecological units. Most in- digenous birds and mammals require a larger or more varied habitat than the protected area itself provides, and these animals spend part of the year away from the protected area as well. In addition, the smaller the area protected, the greater is the "edge effect" (i.e., the often profound changes in the original ecosystem caused by influences at its edges) and the less likely the area will be to survive intact, regardless of the degree of protec- tion (Lovejoy et al. 1986). The previous discussion suggests that few protected areas are large enough to maintain their integrity in the face of changes in land use around their perimeter. Some important types of forests are not covered by existing protected areas. For example, it has been shown that 88 percent of Thailand's forest bird species occur in the 7.8 percent of its land that is in national parks and wildlife sanctuaries (IUCN/UNEP 1986b)-but that is not to say that the 7.8 percent is sufficient to assure the continued survival of all these species. Therefore, although an important start has been made, much more needs to be done before the conservation of biological diversity, particularly in the tropics, can be considered to be adequate. The International Union for the Conservation of Nature (IUCN) has established a set of eight categories of protected areas. Categories I through II (scientific reserves, national parks, and natural monuments) are "strictly protected areas," where the objective is to maintain biological diversity and natural formations. In categories IV through VIII (managed nature reserves, protected landscapes, resource reserves, anthropological reserves, and multiple-use areas including game ranches, recreation areas, and extractive reserves), the objective is controlled exploitation of resources, plus limited but significant commitments to maintaining biological diversity (IUCN 1984; IUCN/UNEP 1986a; Miller 1975; Reid and Miller 1989). Many nations with a strong commitment to protected areas have appor- tioned 10 percent or more of their land area to the strictly protected areas (IUCN categories I, II, and III). Given the considerations discussed in this chapter, it is simply prudent for a reasonable percentage of the land to be allocated to these uses. The World Bank Wildlands Policy has suggested 10 percent of the land as a reasonable starting point for strictly protected areas, but higher figures also have been proposed (e.g., former World Bank President Robert McNamara proposed maintaining 25 percent of African BIOLOGICAL DIVERSITY AND FORESTS 67 countries' land areas as wilderness areas [McNamara 1990]. Consideration of any figure should be accompanied by the two caveats: 1. Conditions vary greatly from one country to another, and no single figure will be adequate for all countries. In countries with high biological diversity and diverse habitats, a relatively high figure may be needed to secure reasonable conservation coverage, whereas a lower figure may be appropriate in countries with very low, localized biological diversity, or countries where less than 10 percent of the land remains in natural or near-natural ecosystems. 2. Strictly protected areas should not be islands in a sea of totally altered or degraded lands. In such cases much of the biological diversity will eventually be lost. A rule of thumb is that if 90 percent of a habitat is lost, ultimately about half of the original biological diversity will be lost (Raven 1990). Consequently, pro- tected areas should be a central concern in the development of a set of land-use areas as described earlier, which would include a range of uses, from complete protection to some use of biological resources. The first step in managing land for forest biological diversity is to establish a policy for the conservation of intact forests. It is also important to protect certain areas of secondary or somewhat degraded forest where inventories show that these ar- eas also contain essential components of biological diversity. For example, certain forest areas that are considered to be "hot spots" of biological diversity may still contain important elements of such diversity even though they have been somewhat modified. Legal protection alone cannot assure the survival of forest areas. Protec- tion measures must be combined with measures to provide solid benefits to local peoples and governments. Furthermore, rapid conversion of the world's remaining forests is due in large part to the rapid increases in human popu- lation and extension of temporary or low-intensity subsistence agriculture. Consequently, measures to conserve biological diversity ultimately must involve a combination of approaches that stabilize or reduce the human population pressure itself, and reduce pressures on forest areas by provid- ing more intensive production of food and fuel elsewhere. On-Site Factors That Affect Biological Diversity Conservation Because a number of factors tend to decrease biological diversity, a decline in one kind of biological diversity leads to a decline in others. Loss of genetic diversity leads to a decline in species diversity. Reduction in total population size of a species can threaten that species with a loss in genetic diversity or with extinction. Rapid environmental disturbances, such as a series of storms, or a single cold spell or cold season, can decrease total 68 MANAGING THE WORLD 'S FORESTS species diversity. Large-scale disturbances, such as drought over a region for several years, or a volcanic eruption, can cause a temporary decrease in biological diversity. Reduction in size and diversity of habitats are notorious causes of re- cent extinction, especially when habitats are disrupted by human activities. Introduction of technologies that disrupt the soil in ways that are not found under natural conditions disrupts habitats, kills organisms, and decreases biological diversity. Introduction of artificial chemicals, such as biocides, can decrease biological diversity through several mechanisms. First, pesti- cides are toxic to many species. Second, loss of vegetation from herbicides can increase erosion, increase the variability of water runoff, and decrease habitat diversity. Loss of pollinators from use of insecticides reduces biodiversity. Introduction of exotic species (species not native to a location) tends to decrease genetic and species diversity. A new predator will find unwary prey that become easy victims. For example, the extinction of birds on is- lands has been attributed to the introduction of dogs, cats, rats, and goats, as has occurred on the Gallapagos. A tree species that is introduced may win in competition with native trees, which are then lost as habitat to native insects, birds, and mammals. It is often believed that hunting and direct killing of individual animals is the only cause of animal extinctions. In the past these were major causes, but today disruption of habitat or loss and introduction of exotic species, along with direct destruction of an animal population by hunting or by eradication because the animal is believed to be a pest, are the principal ways that human activities lead to a rapid increase in the rate of extinction. A minimum viable population is the smallest number of individual members of a species that can be expected to persist for a specified time. This size is determined by many factors, including genetic diversity, rates of birth and mortality, mobility, environmental variability, likelihood of large and destructive habitat disturbances, as well as the likelihood of human intervention. A minimum viable habitat is a habitat (actually, a set of habi- tats) that is large enough to sustain a minimum viable population and has all the habitat characteristics required for that species. Although these two concepts may seem simple, there are few cases for which the size of either a minimum viable population or a minimum viable habitat is known. Re- search to determine minimum viable populations and habitats is essential if conservation areas are to be of appropriate size. Until research establishes these minimum sizes, management must take a conservative approach and err on the side of establishing conservation management areas that are larger rather than smaller; even then, there is no absolute guarantee of success. BIOLOGICAL DIVERSITY AND FORESTS 69 Off-Site Factors That Affect the Conservation of Biological Diversity In addition to direct, on-site effects, there are other, indirect off-site effects that decrease biological diversity. Pollution and Biological Diversity The potential for pollution to decrease biological diversity was first called to the public's attention by Rachel Carson in her 1962 book, Silent Spring, in which she explained the effects of DDT on birds. In general, pollution by toxic substances simplifies forests, reducing the number of species; severe pollution can destroy forests and all their diversity. A classic example is the area near Sudbury, Ontario, surrounding a large industrial belt. Heavy metals and other pollutants released by industries have killed trees. Some areas are now completely devoid of vegetation. Once the veg- etation cover was lost, the soil eroded away, leaving only exposed bedrock in some places. Forests can never regenerate on exposed bedrock. Although direct destruction of habitat appears to be the main cause of decline in species diversity at present, the effects of pollution on biological diversity can be expected to increase in the future, especially as developing nations increase their industrial capacity and as their standards of living rise. Until recently, little information has been available about the effects of pollution on forests in central Europe and the Soviet Union, but air and water pollutants have severely damaged forests in the European part of the former Soviet Union and in central European countries. For example, Czecho- slovakian forests are among the most heavily affected by acid rain of any forests in the world. Programs to assist biological diversity in the forests of these regions must include reduction of air pollution and restoration of damaged forests. Intense pollution effects are not yet a major problem for tropical moist forests, but as industrial development progresses in tropical nations, such pollution is likely to increase. Global Warming and Biological Diversity If global warming should occur as projected by computer models of climate, it will significantly disrupt the distribution of species and reduce biological diversity (Smith and Tirpak 1989). It will also cause important changes in patterns of climatic phenomena such as hurricanes, which dis- turb community structure in tropical forests. Effects of global warming are projected to be most severe at middle and high latitudes and least severe in the tropics. Nonetheless, projected changes in climate are severe and, in comparison with the time scale of biological evolution and the natural 70 MANAGING THE WORLD'S FORESTS migration of forest trees, rapid. It is estimated that climate will change approximately 40 times faster than the rate at which trees migrated north- ward in the temperate zones at the end of the last ice age. One consequence will be that current parks and reserves may no longer have climate appro- priate for the species they were established to protect. This is a special problem for forest ecosystems, because of the longevity of forest trees and the length of time required for the establishment of forests. Any program to conserve biological diversity in the future must begin now to plan how to revise the boundaries and sizes of parks and reserves, so that effects of rapid climate change can be compensated for in the future. Application of Forest Management to the International Community In the past, attitudes of the industrialized nations toward forest man- agement in developing nations have tended to polarize: at one end is the view that methods applied in industrialized Western-style forest planta- tions were the only practical means, and that techniques involved could be exported intact to developing nations, including those in tropical regions. At the other end was an idealized perception of traditional uses of forests- a belief that whatever was done by an indigenous hunter-gatherer who lived close to the land ipso facto must fit within a balance of nature and must be good. This view sometimes led to a belief that traditional practices, su ch as slash-and-burn agriculture, must always provide for complete pro- tection of the original forest, which is not true. The truth lies somewhere between the two poles; obviously, northern industrial plantations have not always led to sustainable forestry and may not work for tropical forests, especially those with fragile soils. Understanding of traditional practices can provide insight into how Western silviculture might be adjusted for tropical forests. To summarize forestry policy and programs, the international commu- nity should take into account a series of problems with forest resource management in the developing nations: • Many areas set aside as parks and conservation units, with the goal of complete protection and no use, have been poorly managed and subjected to harvest of timber and encroachment of agriculture in- cluding grazing of cattle, usually because neighboring people have borne the costs of protection but have received no benefits from it, or because such encroachment brought significant economic or political benefits to the officials involved. • Although sustainable production based on logging of intact forests has not succeeded in most tropical developing nations, plantation forestry also has had little success, particularly in the moist tropics. BIOLOGICAL DIVERSITY AND FORESTS 71 There are two main reasons: countries have not taken the proper approach to sustainable production and to forest plantations, and they have largely overlooked the potential for promoting the right kinds of plantation development, especially by private individuals. • Government ministries responsible for forests and for protected areas do not coordinate their efforts. • Governments have neglected to consider forest-dwellers and other indigenous people and to gain their cooperation in managing the forests. • Traditional cost-benefit analyses for tropical forest projects have failed to consider economic benefits of forest protection and existing ben- efits realized from harvest of nontimber forest resources. • Tourism offers significant potential for economic benefits, but, even in the best existing situations in the tropics, the relatively small, well- managed, and protected areas visited by tourists tend to be surrounded by larger, poorly managed, or unmanaged areas. Moreover, although tourism has contributed to international awareness of the needs for biological conservation, in only a few cases has tourism provided a large share of the income on which a program in biological conserva- tion could be supported. However, examples such as Kenya and Costa Rica-combined with the international trends in ecotourism-indi- cate that, with proper development, tourism could make biological conservation economically profitable. • Some largely temperate-zone forests that have been subjected to chemi- cal pollution and intense use require restoration for the conservation of biological diversity. • Global warming presents a risk to biological diversity as great as any threat in the past, if not greater. In dealing with these problems, forest management for biological diver- sity must be based on an overall landscape perspective. It is not sufficient to maintain isolated patches of original forests as small ecological islands within an otherwise heavily modified and industrialized landscape. Also, the proportion of land currently allocated to maintenance of bio- logical diversity as the primary goal, including the conservation of intact forests, is entirely too small. At this time, when knowledge is so limited, it seems advisable to maintain at least 10 percent of the land in a country as intact wildlands, so long as that 10 percent contains an adequate represen- tation of the country's ecosystems. Much higher figures have been pro- posed, and for any given country this figure may need to be lower where 10 percent no longer remains intact, or substantially higher-because if only 10 percent of the habitat is protected, roughly half the biological diversity will be lost. 72 MANAGING THE WORLD'S FORESTS A New Approach to Management of Intact Forests as Biological Conservation Areas Under the old approach to preserve management, most intact forests were treated as monuments that would persist indefinitely with only a little maintenance. This approach often failed in the past and will fail in the future. For example, in the United States small stands of original forests set aside in preserves are beginning to decline. Visitors trample the soil and reduce regeneration. Few areas are managed for the next generation of mature forests. Under the new approach to preserve management, forest ecosystems are recognized as dynamic. Not only must the intact mature forest be main- tained, but the preserve must be large enough so that forest succession takes place within it, with major representative stages present to provide habitats associated with these stages. Forest management staffs must include professional ecosystem manag- ers trained in the conservation of biological diversity. Experts on hydrology, soils, erosion, geology, and restoration also must be available. Baseline measurements and ecological monitoring must be integrated into management plans. Usually, a small set of factors need be measured and monitored to determine the status of the forest ecosystem, but a pro- gram of measurement must be sustained over a long time. The adminis- trative nature of these programs may vary from country to country, but national measurement programs could benefit from an international pro- gram in ecological monitoring. Research to improve understanding of ecosystem dynamics and inter- actions among species must be integrated into management plans. A preserve must be planned and managed from a landscape perspec- tive, so that various parts of the preserve are put to their best u ses. A preserve must be established and managed with the involvement of indigenous people, so that their needs are accounted for and so that they benefit from the preserve. Where possible, indigenous people should be involved in the conservation, management, and protection of the preserves. REFERENCES Aber, J.S., D.B. Botkin, and J.M. Melillo. 1978. "Predicting the Effects of Different Harvesting Re- gimes on Forest Floor Dynamics in Northern Hardwoods." Canadian Journal of Forest Research 8:306-15. _ _. 1979. "Predicting the Effects of Different Harvesting Regimes on Productivity and Yield in Northern Hardwoods. Canadian Journal of Forest Research 9:10-14. Botkin, D.B., J.F. Janak, and J.R. Wallis. 1973. "Some Ecological Consequences of a Computer Model of Forest Growth." Journal of Ecology 60:849-72. Brown, R.L., ed. 1985. State of the World, 1985. New York: W.W. Norton. Child, Brian. 1990. "Assessment of Wildlife Utilization as a Land Use Option in the Semi-arid Rangeland of Southern Africa." In Living with Wildlife: Wildlife Resource Management with Local Participation in Africa, ed. A. Kiss. World Bank Technical Paper No. 130, Washington, D.C. BIOLOGICAL DIVERSITY AND FORESTS 73 Diamond, A.W. 1985. "The Selection of Critical Areas and Current Conservation Efforts in Tropical Forest Birds." In Conservation of Tropical Forest Birds, eds. A.W. Diamond, and T.E. Lovejoy. International Council for Bird Preservation, Technical Publication No. 4, Cambridge, England. Hamilton, Lawrence S. 1988. "Minimizing the Adverse Impacts of Harvesting in Humid Tropical Forests." In Ecological Development in the Humid Tropics: Guidelines for Planners, eds. A.E. Lugo, J.R. Clark, and R.D. Childs. Morrilton: Winrock International. Holling, C.S. 1988. "Temperate Forest Insect Outbreaks, Tropical Deforestation and Migratory Birds." Mem. Ent. Soc. Can. 146:21-32. Houghton, R.A. 1990. "The Global Effects of Tropical Deforestation." Environmental Science and Technology 24:414-22. IUCN. 1984. "Categories, Objectives and Criteria for Protected Areas." In National Parks, Conserva- tion and Development, eds. J.A. McNeely, and K.R. Miller. Washington, D.C.: Smithsonian Insti- tution Press. IUCN/UNEP. 1986a. A Managing Protected Areas in the Tropics. Gland, Switz.: IUCN. _ _ _. 1986b. Review of the Protected Areas System in the lndo-Malayan Realm. Gland, Switz.: IUCN. Kiss, Agnes, ed. 1990. Living with Wildlife: Wildlife Resource Management with Local Participation in Africa. World Bank Technical Paper Number 130. Africa Technical Department Series, Washing- ton, D.C. Lettau, H., L. Lettau, and L.C.B. Molion. 1979. "Amazonia's Hydrologic Cycle and the Role of Atmospheric Recycling in Assessing Deforestation Effects." Monthly Weather Review 107(3):227- 38. Lovejoy, Thomas E. 1988. "Will Unexpectedly the Top Blow Off?" Bioscience 38(10):722- 26. May, R.M. 1988. "How Many Species Are There on Earth?" Science 241:1441-49. McNamara, Robert S. 1990. "Africa's Development Crisis: Agricultural Stagnation, Population Ex- plosion, and Environmental Degradation." Address to the Africa Leadership Forum, Ota, Nige- ria, June 21, 1990. Miller, K.R. 1975. "Guidelines for the Management and Development of National Parks and Re- serves in the American Humid Tropics." In Proceedings of a Meeting on the Use of Ecological Guidelines for Development in the American Humid Tropics. Morges, Switz.: IUCN. Myers, N. 1979. The Sinking Ark: A New Look at the Problem of Disappearing Species. New York: Pergamon Press. _ _ _. 1983. A Wealth of Wild Species. Boulder, Colo.: Westview Press. Office of Technology Assessment, U.S. Congress (OTA). 1987. Technologies to Maintain Biological Diversity. OTA-F-330. Washington, D.C.: U.S. Government Printing Office. Oldfield, Margery L. 1984. The Value of Conserving Genetic Resources. Washington, D.C.: U.S. Depart- ment of the Interior, National Park Service. Peters, C.M., Gentry, A.H., and R.O. Mendelsohn. 1989. "Commentary: Valuation of an Amazonian Rainforest." Nature 339:655--56. Peters, R.L., and T.E. Lovejoy. 1991. Biological Diversity and the Greenhouse Effect. New Haven, Conn.: Yale University Press. Pimentel, D. 1982. "Biological Diversity and Environmental Quality." In Proceedings of U.S. Strategy Conference on Biological Diversity. Washington, D.C.: Department of State. Poore, M.E.D. 1988. "Natural Forest Management for Sustainable Timber Production." Yokohama: International Tropical Timber Organization, and London: International Institute for Environ- ment and Development. Prescott-Allen, Christine, and Robert Prescott-Allen. 1986. The First Resource. New Haven: Yale University Press. Principe, Peter. In press. "The Economic Significance of Plants and Their Constituents as Drugs." In Economic and Medicinal Plant Research, vol. 3. New York: Academic Press. Raven, Peter H. 1987. "The Global Ecosystem in Crisis." MacArthur Foundation Occasional Paper. John D. and Catherine T. MacArthur Foundation, Chicago. _ _ _. 1988a. "Biological Resources and Global Stability." In Evolution and Coadaptation in Biotic Communities, eds. S. Kawano, J.H. Connell, and T. Hidaaka. Tokyo: University of Tokyo Press. 74 MANAGING THE WORLD'S FORESTS _ _ _. 1988b. "Our Diminishing Tropical Forests." In Biodiversity, eds. E.O. Wilson and Francis M. Peter. Washington, D.C.: National Academy Press. _ _ _. 1991 . Personal communication. _ _ _. 1990. Personal communication. Reid, W.V., and K. R. Miller. 1989. Keeping Options Alive: The Scientific Basis for Conserving Biodiversity. Washington, D.C.: World Resources Institute. Salati, E., A. Dall'Olio, E. Matsui, and J.R. Gat. 1979. "Recycling of Water in the Amazon Basin: An Isotopic Study." Water Resources Research 15(5):1250-58. Salati, E., J. Marques, and L. Molion. 1978. "Origem e distribucao das chuvas na Amazonian." lnterciencia 3:200-05. Salati, E., P.B. Vose, and T.E. Lovejoy. 1986. "Amazon Rainfall, Potential Effects of Deforestation and Plans for Future Research." In Tropical Rainforests and the World Atmosphere, ed. G.T. Prance. AAAS Selected Symposium 101. Boulder, Colo.: Westview Press. Sayer, Jeffrey A. 1991. Personal communication. Shukla,}., C. Nombre, and P. Sellers. 1990. "Amazon Deforestation and Climate Change." Science 247:1322-25. Sidle, R.C., A.J. Pearce, and C.L. O'Loughlin. 1985. Hills/ope Stability and Land Use. American Geo- physical Union Water Resources Monograph 11, Washington, D.C. Slobodkin, L.B., D.B. Botkin, B. Maguire, Jr., B. Moore III, and H.J. Morowitz. 1980. "On the Episte- mology of Ecosystem Analysis." In Estuarine Perspectives, ed. V.S. Kennedy. New York: Aca- demic Press. Smith, J.B., and D.A. Tirpak, eds. 1989. 'The Potential Effects of Global Climate Change on the United States." EPA-203-05-89-0. Washington, D.C.: U.S. Environmental Protection Agency. Talbot, Lee M. 1987. "The Ecological Va lue of Wildlife." In Valuing Wildlife: Economic and Social Perspectives, eds. J.D. Decker and G.R. Goff. Boulder, Colo.: Westview Press. _ _ _ . 1972. "Wild Animals as a Source of Food." In Perspectives on Human Evolution, eds. S.L. Washburn and P. Dolhinow. Orlando, Fla.: Holt, Rinehart and Winston. Talbot, Lee M., W.J.A. Payne, H.P. Ledger, L. Verdcourt, and Martha H. Talbot. 1965. Review: The Meat Production Potential of Wild Animals in Africa; A Review of Biological Knowledge. Common- wealth Agriculture Bureaux, Edinburgh. Villa Nova, N.A., E. Salati, and E. Matsui. 1976. "Estimative de evaportranspiracao no Bacia Amazonica." Acta Amazonica 6(2):215-28. Wilson, E.O. 1990. Personal communication. _ _ _. 1991. Personal communication. Wolf, E.C. 1987. "On the Brink of Extinction: Conserving the Diversity of Life on Earth." Worldwatch Paper 78, Washington, D.C.: Worldwatch Institute. World Resources. 1990-1991. World Reso11rces Institute, UNEP and UNOP. New York: Oxford Univer- sity Press. 5 The Role of Forests in Climatic Change George M. Woodwell T he accumulation of certain heat-trapping gases, including carbon diox- ide, methane, nitrous oxide, and chlorofluorocarbons in the atmosphere, is contributing to the warming of the earth. Whatever the dominant factors may be at any moment in determining the temperature of the earth, the addition of heat-trapping gases to the atmosphere will make the earth warmer than it would have been otherwise. Since the middle of the 19th century the concentration of carbon dioxide (the most important of these gases) in the atmosphere has increased by about 25 percent. The increase has occurred because human activities, especially the burning of oil and coal, the destruc- tion of forests, and, possibly, the warming itself through effects on metabo- lism of plants and soils, have released larger quantities of carbon dioxide into the atmosphere than have been removed either by diffusion into the oceans or by photosynthesis of plants on land. The increase appears trifling, especially in view of the fact that the volume of carbon dioxide in the atmosphere totals less than 0.04 percent, but carbon dioxide in the atmosphere is important in determining the tem- perature of the earth as a whole. Most of the atmosphere, about 79 percent, is nitrogen; most of the remainder is oxygen. Carbon dioxide and certain other gases that are present in the atmosphere in very low quantities absorb radiant heat (infrared radiation); because the total amount of such gases is small, their concentrations are easily changed. An increase in the amount of carbon dioxide, or any of the other heat-trapping gases, increases the capac- ity of the atmosphere for retaining heat and raises the temperature at which the atmosphere comes to equilibrium with the radiant energy received from the sun. 75 76 MANAGING THE WORLD'S FORESTS In recent years it has become evident that, in addition to carbon diox- ide, the burden of other infrared absorptive gases in the atmosphere is growing, and that these gases add to the warming. By the mid-1980s the carbon dioxide in the atmosphere was expected to cause slightly more than half the warming anticipated at that time. Other gases (methane, nitrous oxide, and the chloroflurocarbons, whose concentrations also are accumu- lating rapidly) were having a combined effect approaching that of carbon dioxide. This discussion emphasizes carbon dioxide and methane because they are now the principal contributors to the warming, because their con- centrations are biotically controlled, and because halting a global warming will require, above all, control of carbon dioxide emissions. The principal biotic factors affecting the composition of the atmosphere are the metabo- lism of plants on land and changes in the area of forests (deforestation and reforestation). FACTORS THAT DETERMINE THE COMPOSITION OF THE ATMOSPHERE The warming of the earth is commonly seen as a physical problem. The heat-trapping gases accumulate in the atmosphere and establish a new, higher temperature for the equilibrium between the incoming solar radiation and the reradiation from the earth into the blackness of space. The higher tem- perature means that more energy is available on the surface of the earth for heating the land and for evaporating water. Continental climates are inten- sified: the higher-latitude continental centers warm and become drier. Greater evaporation of water in the tropics means that more energy is transported to the higher latitudes, which are warmed differentially. The warming in high latitudes may be twice the average warming of the earth. In the tropics the warming is less because the evaporation of water transfers the energy to the atmosphere as water vapor, which is carried in the normal circulation of the atmosphere to the higher latitudes; there the vapor condenses, releasing its energy as heat to accentuate the trend. Climatic zones migrate, and the rate of migration affects the survival of species. No one knows how rapidly the earth can be expected to warm, but the most reasonable assumption is that it will warm somewhat more slowly than the heat-trapping gases accumulate. The evidence from the past cen- tury suggests that in the middle and higher latitudes the warming may proceed at a rate of 0.3 degree to as much as 1 degree centigrade (C) per decade throughout the next decades. If no steps are taken to reduce the accumulation of heat-trapping gases, the warming can be expected to accel- erate. A 1-degree C change in mean temperature in the middle latitudes on land in the northern hemisphere is encountered by moving latitudinally 60 to 100 miles. Such changes are expected to accumulate in these latitudes over the next decades unless steps are taken now to slow or stop the accu- mulation of heat-trapping gases (WMO/UNEP 1985, 1986, 1988; Houghton, Jenkins, and Ephraums 1990; Leggett 1990). THE ROLE OF FORESTS IN CL/MA TIC CHANGE 77 This outline of the physical changes obscures the biotic changes, which are not only cause and effect but also potential cure. For example, biotic changes are, through the destruction of forests, a major cause of the increase in carbon dioxide and methane in the atmosphere. But biotic exchanges are complex. A listing of the stocks and flows of carbon shows the importance of the biota in determining the composition of the atmosphere and therefore the temperature of the earth. The relationships among these stocks are shown in exhibit 5-1. Stocks: • The atmosphere contains about 750 billion tons of carbon. • The vegetation and soils globally contain an estimated 2,000 billion tons of carbon as organic matter in various forms-about three times the amount in the atmosphere. • A large stock of carbon is held in the fossil carbon deposits of oil, coal, and gas. This stock is estimated to exceed 10,000 billion tons. Flows: .__• The annual release into the atmosphere through combustion of fossil fuels is about 5.6 billion tons of carbon as carbon dioxide. • The annual release of carbon into the atmosphere from deforestation was estimated at 1 to 3 billion tons in 1980; the rate is higher now as deforestation proceeds. • Photosynthesis globally on land absorbs into plants from the atmo- sphere about 100 billion tons of carbon annually, nearly one-seventh of the atmospheric burden. • Under normal circumstances, respiration, including the respiration of plants, animals, and the organisms of decay, releases an amount of carbon approximately equal to the amount absorbed in photosynthe- sis, and so the atmospheric composition is maintained at equilibrium. • The annual accumulation of carbon in the atmosphere is 3 to 5 billion tons, which is the amount by which the current releases exceed the transfer into the oceans and the absorption into the terrestrial biota. The stock of carbon that is controlled by biotic processes is about three times the amount in the atmosphere, and a change in the flows between the pool of carbon held in plants and soils and the pool in the atmosphere affects the atmosphere significantly. The potential changes extend well beyond the current release from deforestation. The magnitude of these exchanges is seen in the annual oscillation observed in the carbon dioxide content of the atmosphere. The metabolism of forests caused a seasonal oscillation of 5 parts per million as measured in the Hawaiian Islands. Any shift in the magnitude of the photosynthetic or respiratory flows could be significant. A warming of the earth, especially a rapid warming, might cause such a shift. The pattern is far from certain, but experience suggests that a warming (and almost any other severe disturbance) will stimulate respira- 78 MANAGING THE WORLD 'S FORESTS EXHIBIT 5-1. Global Flows of Carbon (billion metric tons of carbon/year) >100 ATMOSPHERE 750+ 100 >100 100 PLANTS OCEANS SOIL 800 2,000 tion more than it will affect photosynthesis. An increase in temperature of 1 degree C is widely recognized as increasing the rate of respiration by 10 to 30 percent, while having very little effect on photosynthesis. The stimula- tion of respiration increases the rate of release of the products of respiration, including both carbon dioxide and methane. DEFORESTATION Climatologists have grossly underestimated the effects of forests on cli- mate, largely because they have focused too much on the effects of oceans. The distortion persists in the recent report on the scientific sessions of the Second World Climate Conference held in Geneva (WMO 1990). The report acknowledges that forests could contribute to a solution of the problem by storing additional carbon through reforestation, but the report does not emphasize the need to halt deforestation. Deforestation was the dominant source of carbon dioxide in the atmo- sphere until the middle 1960s, when the global surge in the use of fossil fuels brought the emissions from that source above the releases from defor- estation for the first time (Woodwell 1983, 1984). The surge in the use of fossil fuels has continued, and the release of carbon from fossil fuels now exceeds 5 billion tons annually compared with an estimated 1 to 3 billion tons for the release from deforestation. However, the estimates of deforesta- tion are based on appraisals that are often arbitrary and uncertain (Houghton 1991). THE ROLE OF FORESTS IN CL/MA TIC CHANGE 79 What constitutes deforestation is open to interpretation. Forest manag- ers who are interested in the crop of fiber, including lumber and pulp, that can be obtained from a forest do not consider the clear-cutting of extensive tracts as "deforestation" so long as the tracts are expected to return to forest. But there is not much question when forested land is converted to agricul- ture, as is common in the tropics currently, that the region has been defor- ested. As yields drop, agricultural land may be abandoned, may be classi- fied as impoverished, or may revert to forest. Each of these transitions involves a substantial change in the amount of carbon held in plants and soils on land and can be measured as a change in the area of land in various uses. If agricultural land is allowed to revert to forest, that transition, too, can be accommodated in calculations of the carbon flux between the land and the atmosphere. Such calculations, including data on the harvest of forests and their recovery, even when technically called "forest manage- ment," have been made for more than a decade on the basis of records of land use (Woodwell and Houghton 1977; Bolin 1977; Woodwell et al. 1978, 1983; Houghton 1990b). The early estimates were little more than guesses based on limited data (Woodwell and Houghton 1977; Woodwell et al. 1978; Bolin 1977). During the 1980s a body of experience developed in interpreting Food and Agricul- tural Organization data, historical records (Richards, Olson, and Rotty 1983), satellite imagery (Woodwell 1984), and trends that has led to greater confi- dence in the analyses (Houghton et al. 1985). The carbon released in 1980 through changes in the area of forests globally was originally estimated to range between 0.9 and 2.5 billion tons. When Detwiler and Hall (1988) reanalyzed virtually the same data, they extended the lower end of the range for 1980 to 0.4 billion tons. More recent estimates covering the 1980s (Myers 1990; Houghton 1990a,b,c;) suggest that the net flux from these changes is probably closer to the upper end of the range. There is reason to question whether the current flux from higher latitudes remains low, but few direct measurements of changes in land use are available (Melillo et al. 1988). Improved global measurements will come from applications of satel- lite imagery, only now beginning to be applied to this purpose (Grainger 1983; Woodwell, ed. 1984). For more than a century, global deforestation has contributed signifi- cantly to the accumulation of carbon dioxide in the atmosphere. Houghton and Skole (1990) have estimated the total release of carbon globally from deforestation (including reforestation) between 1850 and 1985 at 100 to 130 billion tons- a narrowing of the range of an earlier estimate for the same period (62 to 228 billion tons) made by Houghton et al. (1983) and Richards et al. (1983). The release from combustion of fossil fuels during the same 135 years is thought to have been about 190 billion tons of carbon. The current annual rate of release of carbon from deforestation is calcu- lated on the basis of net deforestation over an area estimated at approxi- mately 11 million hectares, all in the tropics. There is reason to question whether this basis remains accurate, because a substantial harvest of old- 80 MANAGING THE WORLD'S FORESTS growth stands is occurring in the northern coniferous forests of North America, Europe, and Asia. In addition, forests over extensive areas are being impoverished as a result of the cumulative effects of industrial toxins, ozone, and acid rain (Woodwell 1990). All these effects contribute to a loss of carbon stocks from forests and soils, but the magnitude of these changes has not been appraised. Whatever the effects of human activities on temper- ate- and boreal-zone forests, effects on tropical forests globally are more serious and present a threat to the earth as a whole and to human welfare. One estimate of the annual rates of release of carbon from forests be- tween 1850 and 1980 appears in exhibit 5-2. CHANGES IN METABOLISM The most conspicuous, short-term effect of forests globally is the extent of their control over the carbon dioxide content of the atmosphere as re- corded in the data on the seasonal trends in the carbon dioxide content of air at the Mauna Loa Observatory and elsewhere (Keeling et al. 1989). At EXHIBIT 5-2. The Net Annual Flux of Carbon from Deforestation in Tropical and Temperate Zones Globally from 1850 to 1980 2.0 1.8 6 1.6 0) a. I "' I 'I '--' c:: 1.4 I I 0 .0 I (ij 1.2 I () ' 0 1.0 I x I ::J I u:::: «l 0.8 I . ITrop1cal ::J 1 c I c:: 0.6 - ~ z a> 0.4 ------- , .,,,.--- ,_, ............... , .6'; I -,' ; , - .... I ' __ , ', ,... - ' ,... ' J Temperate ' "'\ 0.2 .. ... 0 1860 1880 1900 1920 1940 1960 1980 SOURCE: Houghton and Skole (1990). THE ROLE OF FORESTS IN CL/MA TIC CHANGE 81 Mauna Loa during the summer months the photosynthesis of forests in the northern hemisphere becomes the dominant influence and reduces the car- bon dioxide content of the atmosphere by about 5 parts per million (ppm). During winter, respiration dominates and restores the carbon dioxide re- moved during the summer. The process has been summarized by Woodwell (1983) and explained in greater detail by Houghton (1987a,b), who showed that forests control these shifts. Globally there is an annual flux of carbon through the plants, animals, and decay organisms on land of more than 100 billion tons, approximately one-eighth of the atmospheric burden. Any modification of that flux could appreciably alter the composition of the atmosphere. Most of the flux occurs through forests because forests are so large and carry on such a large fraction of the global metabolism on land. Climatic changes or even changes in the weather can cause rapid changes in the pools of carbon held on land and in the fluxes between major pools. The most important biotic changes may involve immediate shifts in the ratio of gross production (total photosynthesis) to total respiration locally, regionally, and globally. A global warming will affect the balance between the approximately 100 billion tons of carbon that enter green plants annu- ally through photosynthesis on land and the 100 billion tons that are re- leased annually through respiration. Over the past 15 years the annual net accumulation of carbon in the atmosphere has amounted to about 3 billion tons, and a 1 percent change in either metabolic process will have important implications for the composition of the atmosphere in a short time. Because it will appear as a small incremental change in a large stock of carbon distributed very widely, a change in the atmosphere as a result of photosyn- thesis and respiration will be much more difficult to measure than the change resulting from diminished areas of forest. Photosynthesis is affected by many factors, particularly the availability of energy as light and the availability of water and nutrients. Respiration also is affected by many factors, including the availability of water, but respiration is particularly sensitive to temperature. No such sensitivity to temperature exists for photosynthesis. That observation alone suggests that a global warming will speed the decay of organic matter globally without appreciably affecting photosynthesis. A significant fraction of the total res- piration is in the middle and higher latitudes where the warming is ex- pected to be greatest--enough to affect the composition of the atmosphere significantly. If, for example, we assume that a warming of 0.5 degree Chas occurred globally and that in the forested middle latitudes the warming approaches two times the mean for the earth as a whole, these latitudes might experience an increase in respiration of 5 to 20 percent. The increase might apply to 20 to 30 percent of the global total of respiration on land. If there were no compensation through increased storage, the excess release from the warming already experienced could be 1 to 6 billion tons of carbon annually. This release, if it is realized, would be a new release stimulated by the warming itself, and would be of the same order as the other major releases of heat-trapping gases and deforestation. 82 MANAGING THE WORLD'S FORESTS The release of carbon from changes in the rate of respiration could fluctuate appreciably. Between 1963 and 1988 the annual rate of accumula- tion of carbon dioxide in the atmosphere was about 1.5 ppm, equivalent to a global accumulation of about 3.0 billion tons of carbon in the atmosphere. More recently, the rate of accumulation as recorded at Mauna Loa and at the South Pole increased to a peak of about 2.4 ppm per year (about 5 billion tons of carbon), according to the record maintained by Charles D. Keeling of the Scripps Institution of Oceanography. Keeling expected at first that the surge would prove transitory, as a lesser surge did in 1973-74. The rate of accumulation has in fact dropped to about 4 billion tons annually. The surge in this instance is consistent with a stimulation of respiration by the warming that marked the late 1980s. Houghton (1990d) has offered a complicated explanation of hitherto unresolved questions concerning the fluxes of carbon between the biota and the atmosphere, as defined previously by use of oceanic models and isoto- pic ratios as well as by measurements of rates of change in the area of forests. He suggests that the puzzling conflict in interpretations of the glo- bal carbon cycle derived from these different methods, which led to the assumption that there is "missing carbon" in the global cycle, can be re- solved by recognizing (1) that the different measurements address different fluxes of carbon and (2) that carbon fluxes are correlated with temperature (in warmer periods, carbon accumulates in the atmosphere; in cooler peri- ods, carbon accumulates on land). Such a correlation reinforces the conclu- sion that respiration is the most sensitive metabolic process, responds rap- idly to changes in temperature, and determines, throughout a wide range of temperatures, whether the land gains or loses carbon in its continuous ex- changes with the atmosphere. According to these analyses a rapid warming can be expected to release significant additional quantities of carbon as carbon dioxide and methane into the atmosphere, simply by increasing rates of respiration of plants and organic matter in soils, swamps, and bogs. This point, which has previously been made in a different form (Woodwell 1983 and 1988), is important because it means that a warming feeds a further warming: as forests are destroyed, the earth warms more rapidly and the warming destroys additional forests. Will the Warming Produce Greater Areas of Forests? Is there any basis for expecting an increase in the area of forests, and thus an increase in storage of carbon on land, as the earth warms? Will not the warming, for example, open new lands in high latitudes to forests and result in the spread of forests into regions now tundra? Such changes may occur, but forests require decades to centuries to develop, especially where the soil is thin and nutrients are in short supply. Forests also require climatic stability, sources of seeds, and the accumula- tion of sufficient stocks of nutrients in soils and plants to support a forest. THE ROLE OF FORESTS IN CL/MA TIC CHANGE 83 The climatic transitions under way at the moment are rapid by any measure and are expected to continue indefinitely. They do not offer the conditions under which forests develop on new lands and remain for long periods. Is it possible that existing forests or tundra will be stimulated to store additional carbon in plants and soils as the warming progresses? Perhaps. The boreal forest and other coniferous forests may be sufficiently resilient to respond to a warming by increased periods of photosynthesis and increased growth. Whether carbon will be stored or simply released through increased rates of respiration remains an open question. There is also the possibility that the tundra, warmed, will respond in surprising ways, including in- creased primary production and storage of carbon in peat. Much will hinge on the availability of water. A wetter tundra might store additional carbon in soils; a drier tundra might release it through the decay of organic soils long frozen or normally frozen throughout much of the year. Botanist W. Dwight Billings (1987) believes that the warming will speed the decay of peat in soils of the tundra and result in that ultimate breakdown ecologists describe as "thermal karst erosion," which eats away the landscape in acre- size chunks. The result is not only devastation of the tundra but the release of carbon as carbon dioxide and methane from the substantial reserves of organic matter in the deep peat of the tundra. The overall response of the earth to rapid warming could be a net additional release of carbon into the atmosphere, because total respiration is stimulated more than gross photosynthesis. The magnitude of the release will hinge on the rate of warming: the more rapid the warming, the larger the release. This relationship is consistent with, but not proven by, data from the Vostoc Core, a section of ice of the Antarctic glacial cap that dates back 160,000 years (Bannola et al. 1987). Data from this core show that temperature, carbon-dioxide concentrations, and methane concentrations in the atmosphere were closely correlated with one another over that period. As temperatures rose, so did the carbon dioxide and methane concentra- tions; as temperatures dropped, so did the concentrations of these gases (Lorius et al. 1988). The evidence is consistent with the hypothesis that forests determine the composition of the atmosphere in the short run of years to decades (Houghton and Woodwell 1988). A more complicated aspect of global warming is the rapid increase in the amount of methane in the atmosphere-about 1 percent per year. The increase is significant, because methane is, molecule for molecule, 20 times more effective than carbon dioxide in trapping heat. Methane is a product of respiration in places where oxygen is limited. Soils, especially wet soils, including marshes, swamps, and bogs, are probably the main source of methane globally. The warming that has already occurred has without ques- tion stimulated anaerobic decay as well as aerobic decay and is probably the main cause of the increased rate of release of methane. Although other contemporary arguments favor cattle, rice paddies, and dumps as the main sources of methane, ecologists suggest that we are already observing biotic 84 MANAGING THE WORLD'S FORESTS feedbacks from the warming that has already occurred (Woodwell 1988; Khalil and Rasmussen 1989). That soils are the major source of methane is supported by year-to-year differences in the rate of its atmospheric increase. During the strong El Nino/Southern Oscillation (ENSO) of 1982-83, meth- ane concentrations increased less than they had in recent years. Because the areas in rice production and the number of ruminants did not fluctuate during this interval, the variation was probably due to different regimes of temperature and moisture in those ecosystems influencing the air sampled by the monitoring station. A further, much larger, source of methane apparently exists in clath- rates vulnerable to sudden instability as shallow seas warm (MacDonald 1990). What Can Be Done to Slow the Accumulation of Heat-Trapping Gases? The immediate need is to stop the continuous increase in the green- house gas content of the atmosphere. There is no alternative. Delay works rapidly through positive feedbacks to make the challenge more difficult and, ultimately, beyond any control mechanisms now visualized. For more than a decade, no matter what the emissions are from fossil fuels, deforestation, or any of the feedbacks, 3 billion tons of carbon as carbon dioxide have been accumulating in the atmosphere each year. If the current releases were reduced by 3 billion tons, the atmospheric content of carbon dioxide would be stabilized for a few years. The rate at which the oceans absorb atmospheric carbon dioxide is determined by the difference in concentration between the atmosphere and the surface waters of the oceans. If the emissions were reduced, the gradient between atmosphere and oceans would be reduced slowly over several years. As the gradient declined and the oceans absorbed less, there would be a need to reduce emissions further to avoid a further increase in the atmosphere. The largest source of the emissions is carbon dioxide from combustion of fossil fuels-currently about 5.6 billion tons of carbon per year. Because the industrialized nations contribute about 75 percent of these emissions, steps toward stabilizing the composition of the atmosphere must begin in the industrialized world. A recent study by Jose Goldemberg of Brazil and colleagues from various nations, carried out under auspices of the World Resources Institute, suggests that a program of conservation and improved efficiency alone could cut the consumption of energy from fossil fuels in the developed countries in half (Goldemberg et al. 1987a and 1987b). Although the developing nations contribute less carbon dioxide to the atmosphere at present, their contributions are growing, and, if economic development fol- lows conventional patterns, could be much larger. The second step toward stabilizing the atmospheric content of infrared absorptive gases will require innovations in economic development that remove dependence on increased use of fossil fuels. THE ROLE OF FORESTS IN CL/MA TIC CHANGE 85 Can Reforestation Be Used to Remove Carbon from the Atmosphere. at Least Temporarily? Although the suggestion that planting trees will help solve the problem is correct, the objective is the removal of carbon from the atmosphere and its storage for decades to a century or more. Such an objective requires the reestablishment of forests. The mere planting of trees does not build forests with their capacities for storing carbon in soils as well as plants, for absorb- ing and recirculating nutrient elements, and for regeneration after distur- bance. An area of 1 to 2 million square kilometers in forest on fertile soils is required to store 1 billion tons of carbon annually (Woodwell 1988; Marland 1988). The forest will continue to store carbon at that rate for several de- cades, possibly a century or more, until gross photosynthesis is approxi- mately balanced by the total respiration of the forest. The establishment of plantations is a step in the right direction, but plantation forestry is primitive in most applications, the forests are vulner- able to fire and disease, and the expense of planting is so great that the forests must be harvested systematically to warrant the investment. The objectives in establishing plantations for the purposes described here are commonly antithetical to the objective of accumulating carbon stocks in forests and soils and preserving those stocks for a century or more. Planta- tions are commonly established with early harvest in mind, not with the objective of reestablishing a fundamental biotic unit that will stabilize the landscape, preserve biotic resources, and help to control the composition of the atmosphere. The reestablishment of forests designed to persist for a century or more requires more knowledge and effort than are usually avail- able for establishment of plantations of trees. Impoverished land is no better for forests than for agriculture. The richer the land, the more rapid the growth of forest plants and the greater the storage of carbon. Because it is unlikely that large areas of productive land can be made available for reforestation (most productive land in countries such as India is in intensive agriculture), the first objective in management of forests clearly must be the cessation of deforestation, including the de- struction of the world's remaining primary (old growth) forests. However attractive they may appear initially, efforts at reforestation seem destined to play a secondary role. Moreover, efforts assume the stability of climate, an assumption increasingly in doubt. Nonetheless, throughout much of the world, abandoned land is refor- ested in time through normal, natural successional processes, a subject that has been the core of much research and analysis in ecology (Clements 1928; Oosting 1958). Succession, however, requires several conditions: sources of seeds must be available, and soil, mineral nutrients, and climate must be suitable. In large areas of the tropics, deforestation leads quickly to degra- dation of soils, loss of nutrients, and even such profound changes in the site that it will no longer support forest. Forests, once destroyed over large areas, do not necessarily recover naturally. Massive reforestation programs 86 MANAGING THE WORLD'S FORESTS will be expensive, difficult to execute, and slow to become effective. They are in competition with agriculture, which offers earlier financial reward and helps feed a human population that is expanding at 90 million people annually. Reforestation programs cannot compensate for the continued re- leases of carbon from the destruction of the remaining primary forests glo- bally or, under any conditions imaginable at the moment, for the continued use of fossil fuels as the primary source of energy for expanding the indus- trialized world. Reforestation is nonetheless an ameliorating influence and should be encouraged in every way. What Steps Must Be Taken to Stabilize the Composition of the Atmosphere? No one remedy by itself is likely to stabilize the concentration of atmo- spheric carbon dioxide; many efforts will have to be brought to bear simul- taneously. If the surge in accumulation of carbon dioxide in the atmosphere persists, the challenge in stabilizing the atmospheric burden will have risen from 3 billion tons to 5 billion tons. In either instance, a combination of several steps must be taken to reduce or stop the accumulation (WMO/ UNEP 1988): • Global reduction in use of fossil fuels. The burning of oil and coal and gas currently releases about 5.6 billion tons of carbon into the atmo- sphere annually. • Cessation of deforestation. Deforestation, according to current estimates, releases 1 to 3 billion tons of carbon annually. The area deforested annually is thought to be about 11 million hectares, but the estimate is crude. • Reforestation. The absorption of 1 billion tons of carbon annually into a developing forest would require an area of 1 to 2 million square kilometers (100 to 200 million hectares). Each of these steps will have salutary effects quite apart from climatic change. Improved efficiency in the use of energy, for example, which might have been taken long ago with benefits to all, now would bring economic and material advantages to individuals and nations. If the improved effi- ciency enabled a significant reduction in reliance on fossil fuels, as it must, emissions of sulfur and nitrogen oxides would be reduced; so would acid deposition and the release of various other toxins. Halting deforestation would help reduce biotic impoverishment globally (Woodwell 1990), main- tain the genetic diversity of the planet, reduce erosion, stabilize local and regional climates, cleanse water and air, and preserve opportunities for fu- ture generations. Steps taken to reduce one problem help to reduce others. A shift to reliance on enduring sources of energy is inevitable, and steps taken now to speed that transition will be advantageous to everyone. The steps can begin locally, although it is clear that the world must join in the THE ROLE OF FORESTS IN CL/MA TIC CHANGE 87 effort if it is to be effective. International agreements will be appropriate, but any step involving a reduction in use of fossil fuels will put increased pressure on forests. WHAT CAN THE INTERNATIONAL COMMUNITY DO? The time has passed when forests in their entirety can be considered local, regional, or even national resources to be managed for personal or local interests. Forests are essential to the stabilization of climate globally; to the management of water, land, and air regionally and locally; and to the preservation of the terrestrial biota. We have little experience in managing supranational resources and less in recognizing and facilitating the transfer of private or national resources in part or in whole to global status. Yet such a transfer is now required in the common interest. Can it be effected? On the surface the demand appears outrageous, and it will be consid- ered so by landowners and nations, but the transfer of control need not be complete or even specific for a particular forest. What is required is that a forested landscape be preserved with segments in differing degrees of physi- cal and biotic integrity and in varying stages of successional maturity. The challenge is one of management: How can the landscape be managed to keep its essential functions as a part of the biosphere intact while maintain- ing its capacity for supporting people? The challenge is hardly new. The hills of Attica were deforested in Pliny's time and the depredations were deplored then. Is there a way now of managing forests and land to support people without the progressive biotic impoverishment that has marked the advance of civilization throughout time? The issue is highly political as well as technically complex. One of the key issues is the unit of management. An appropriate area for the unit will hinge on the objectives in management as well as on details of ecology. Management to preserve wolves and mountain lions in eastern North America requires much larger areas than management to preserve plant communi- ties or species of trees. The issue becomes one of regional planning on the basis of ecological, as opposed to short-term economic, principles. Such a transition in the basis of management will not come easily. Avarice is in- born; ecology is learned, usually slowly. Although history is not encourag- ing, the need for a transition in management of land and water is so clearly urgent that efforts are long overdue to move toward integrated programs of management of terrestrial resources that will stabilize their yields and as- sure continued availability for human support. Forest management can no longer be separated from the larger question of prevention of further accumulation of heat-trapping gases in the atmo- sphere. The technical issue is straightforward: the earlier the composition of the atmosphere can be stabilized as the first step in stabilizing the surge in the temperature of the earth, the greater the chance of success and the less 88 MANAGING THE WORLD'S FORESTS the disruption of the human enterprise. The longer we wait for effective action, the more difficult the stabilization becomes. Scientists see the transi- tion as urgent, but politicians have been slow to respond to the urgency. Any action that affects the availability of resources is disruptive, but failure to act has the potential for continuous climatic disruptions of even more serious dimensions. The directions in economic development that must be favored are clear enough. The ways in which international development agencies should pursue those objectives are less clear and will probably differ from place to place. Nor is it clear that the objectives can be separated from the necessity for limiting the growth of populations to avoid further mining of essential resources simply to support the new mouths, let alone the people here now. Success will hinge on local support, and that support depends on under- standing both the problem and the solutions. In the pragmatic world of contemporary politics and economics, devel- opment agencies might start with comprehensive grants for education and science, coupled with a search for innovative ways of shifting away from fossil fuels and toward efficiency in the use of energy, and toward increased reliance on solar energy for domestic and industrial use. Such a program will require the development of inexpensive solar energy systems that can be widely distributed at low (or no) cost to the recipients. At the same time there is a need for innovations in land use that will make it clearly desirable and possible to keep forests intact over large areas in the tropics and elsewhere. These innovations include adoption of appro- priate technology that is benign to the environment, improved farming prac- tices, a conservation approach to timber harvesting, and establishment of sustainable production systems. The innovations might also include plan- ning for control of land use to increase the utility of the land in supporting people in the long run through the use of remotely sensed imagery com- bined with geographical information systems. Such innovations are being explored at the moment in Acre, Brazil, in association with the establish- ment of forest tracts as "extractive reserves" to be used by rubber tappers and others. SUMMARY The atmospheric content of the heat-trapping gases, carbon dioxide and methane, has been increasing over the past two centuries as a result of human activities. The major source of carbon dioxide is the combustion of fossil fuels; the major source of methane is less certain, but fossil fuels account for a large proportion. Both gases are products of respiration, the biotic process that results in the breakdown of organic matter and the re- lease of energy, heat, and water. Any activity, including the warming of the earth, that increases rates of respiration above gross photosynthesis will increase rates of emission of those gases. THE ROLE OF FORESTS IN CL/MA TIC CHANGE 89 The composition of the atmosphere is determined by interactions among large pools of carbon in the atmosphere, in forests and soils, in the oceans, and in the fossil fuels now being mined and burned for energy. The ex- changes among these pools are also large: about 5.6 million tons of carbon annually are released into the atmosphere from combustion of fossil fuels; a two-way flow of about 100 billion tons of carbon occurs between the atmo- sphere and green plants on land; and a second two-way flow of about 100 billion tons occurs between the atmosphere and the oceans. The pools of carbon held in forests and their soils are being systemati- cally reduced by human activities. Deforestation is releasing between 1 and 3 billion tons of carbon by accelerating the rate of decay of organic matter in trees and soils globally. The problem is aggravated by the effects of pollution, which also are reducing forests over large areas. Scientists expect the warming itself to increase rates of respiration, including the respiration of plants and the organic matter in soils, thereby speeding the release of carbon dioxide and methane from the large biotically controlled pools on land. This latter effect, which is severe enough to accelerate the warming significantly, adds urgency to efforts to stop the further accumulation of heat-trapping gases in the atmosphere. Forests and their soils contain about three times as much carbon as is currently held in the atmosphere. If additional forests could be established globally, carbon might be removed from the atmosphere and stored. A new forest covering 1 to 2 million square kilometers, possibly more if soils are poor, would be required to store annually 1 billion tons of carbon. The forest would continue to store additional carbon annually throughout its period of rapid growth, as much as 50 years. But the stabilization of the atmosphere at the moment with respect to carbon dioxide would require the reduction of current release rates of carbon dioxide by 3 to 5 billion tons of carbon. In a few years additional reductions in releases would be re- quired. If a devastating, open-ended warming of the earth is to be avoided, the use of fossil fuels should be cut as soon as possible by at least 60 per- cent, deforestation should cease, and reforestation should be vigorously pursued. The sooner these steps are taken, the cheaper they will be and the greater the probability of success in stopping the warming. REFERENCES Barnola, J.M., D. Raynaud, Y.S. Korotkevich, and C. Lorius. 1987. "Vostok Ice Core Provides 160,000- Year Record of Atmospheric C02." Nature 329(6138):408-14. Billings, W.D. 1987. "The Carbon Balance of Alaskan Tundra and Taiga Ecosystems: Past, Present, and Future." Quaternary Science Reviews 6:16~7. Bolin, B. 1977. "Changes of Land Biota and Their Importance for the Carbon Cycle." Science 196:613- 15. Clements, F.E. 1928. Plant Succession and Indicators. Washington, D.C.: Carnegie Institution of Wash- ington. Detwiler, R.P., and C.A.S. Hall. 1988. "Tropical Forests and the Global Carbon Cycle." Science 239:42- 50. 90 MANAGING THE WORLD'S FORESTS Goldemberg, J., T.B. Johansson, A.K.N. Reddy, and R.H. Williams. 1987a. Energy for a Sustainable World. Washington, D.C.: World Resources Institute (WR!). _ _ _. 1987b. Energy for Development. Washington, D.C.: WR!. 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. Special Publication No. 2 of the British Ecological Society. Houghton, J.T., C.J. Jenkins, and J.J. Ephraums. 1990. Climate Change: The IPCC Scientific Assessment. Cambridge, Eng.; Cambridge University Press. Houghton, R.A. 1987a. "Terrestrial Metabolism and Atmospheric C02 Concentrations." BioScience 37:672-78. _ _ _. 1987b. "Biotic Changes Consistent with the Increased Seasonal Amplitude of Atmospheric Co2 Concentrations." Journal of Geophysical Research 92:4223-30. _ _ _. 1990a. "The Future Role of Tropical Forests in Affecting the Carbon Dioxide Content of the Atmosphere." Ambio 19:204-09. _ _ _. 1990b. "The Long-Term Carbon Flux Between Terrestrial Ecosystems and the Atmosphere as a Result of Changes in Land Use." Oak Ridge, Tenn.: Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory. _ _ _. 1991. "Tropical Deforestation and Atmospheric Carbon Dioxide". Climatic Change 19:99- 118. _ _ _. 1990c. "Emissions of Greenhouse Gases." In Deforestation Rates in Tropical Forests and Their Climatic Implications, ed. N. Myers. London: Friends of the Earth. _ _ _. 1990d. "Terrestrial Ecosystems in the Global Carbon Balance: Changes over the Last 135 Years." Unpublished manuscript. Houghton, R.A., J.E. Hobbie, J.M. Melillo, B. Moore, B.J. Peterson, C.R. Shaver, and G.M. Wood well. 1983. "Changes in the Carbon Content of Terrestrial Biota and Soils between 1860 and 1980: A Net Release of C02 to the Atmosphere." Ecological Monographs 53:235-62. Houghton, R.A., W.H. Schlesinger, S. Brown, and J.F. Richards. 1985. "Carbon Dioxide Exchange between the Atmosphere and Terrestrial Ecosystems." In Atmospheric Carbon Dioxide and the Global Carbon Cycle. Washington, D.C.: Department of Energy. Houghton, R.A., and D.L. Skole. 1990. "Carbon." In The Earth Transformed by Human Action, ed. B.L. Turner. New York: Cambridge University Press. Houghton, R.A., and G.M. Woodwell. 1988. "Global Climatic Change." Scientific American (April):36- 44. Keeling, C.D., R.B. Bacastow, A.F. Carter, S.C. Piper, T.P. Whorf, M. Heimann, W.G. Mook, and H. Roeloffzen. 1989. "A Three Dimensional Model of Atmospheric C02 Transport Based on Ob- served Winds: 1. Analysis of Observational Data." Geophysical Monograph No. 55. American Geophysical Union. Khalil, M.A.K., and R.A. Rasmussen. 1989. "Climate-Induced Feedbacks for the Global Cycles of Methane and Nitrous Oxide." Tel/us 418:554-59. Leggett, J. 1990. Global Warming: The Greenpeace Report. London: Oxford University Press. Lorius, C., N.1. Barkov, J. Jouzel, Y.S. Korotkevich, V.M. Kotlyakov, and D. Raynaud. 1988. "Antarc- tic Ice Core: C02 and Climatic Change over the Last Climatic Cycle." EOS, 681-84. MacDonald, G.J. 1990. "Role of Methane Clathrates in Past and Future Climates." Climatic Change 16:247-81. Marland, G. 1988. "The Prospect of Solving the C02 Problem through Global Reforestation." U.S. Department of Energy NBB-0082 ORNL. Melillo, J.M., J.R. Fruci, R.A. Houghton, B. Moore, and D.L. Skole. 1988. "Land Use Change in the Soviet Union between 1850 and 1980: Causes of Net Release of C02 in the Atmosphere." Tel/us 40B:l 16-28. Myers. 1990. Deforestation Rates in Tropical Forests and Their Climatic Implications. London, Eng.: Friends of the Earth. Oosting, H.J. 1958. The Study of Plant Communities. San Francisco: W.H. Freeman and Co. Richards, J.F., J.S. Olson, and R.M. Rotty. 1983. "Development of a Data Base for Carbon Dioxide Releases Resulting from Conversion of Land to Agricultural Uses." ORAU / IEA-82-10(M), ORNL/ TM-8801. Oak Ridge, Tenn.: Oak Ridge National Laboratory. THE ROLE OF FORESTS IN CL/MA TIC CHANGE 91 WMO. 1990. World Meteorological Association. World Climate Conference, Geneva, November. WMO/UNEP. 1985. "Conference Statement: International Assessment of the Role of Carbon Diox- ide and of Other Greenhouse Gases in Climate Variations and Associated Impacts." Villach, Austria, October 9-15, 1985. _ _ _ . 1986. "Report of the International Conference on the Assessment of the Role of Carbon Dioxide and Other Trace Gases in Climate Variations and Associated Impacts." Villach, Austria, October, 1985. WMO 661. _ _ _ . 1988. "Developing Policies for Responding to Climatic Change: A Summary of Discussions and Recommendations of Workshops Held in Villach and Bellagio, 1987." Woodwell, G.M. 1983. "Biotic Effects on the Concentration of Atmospheric Carbon Dioxide: A Review and Projection." In Changing Climate. Washington, D.C.: NAS Press. _ _ _ . 1988. "The Warming of the Industrial Middle Latitudes, 1985--2050: Causes and Conse- quences." Prepared for the Symposium, Developing Policies for Responding to Future Climatic Change, held in Villach, Austria, September 1987. Climatic Change 15:31-50. _ __ .ed . 1984. The Role of Terrestrial Vegetation in the Global Carbon Cycle: Measurement by Remote Sensing. New York: John Wiley and Sons. _ _ _. ed. 1990. The Earth in Transition: Patterns and Processes of Biotic Impoverishment. New York: Cambridge University Press. Woodwell, G.M., J.E. Hobbie, R.A. Houghton, J.M. Melillo, B. Moore, B.J. Peterson, and G.R. Shaver. 1983. "Global Deforestation: Contribution to Atmospheric Carbon Dioxide." Science 222:1081- 86. Woodwell, G.M., and R.A. Houghton. 1977. "Biotic Influences on the World Carbon Budget." In Global Chemical Cycles and Their Alterations by Man, ed. W. Stumm. Dahlem Konferenzen, Berlin. Woodwell, G.M., R.H. Whittaker, W.A. Reiners, G.E. Likens, C.C. Delwiche, and D.B. Botkin. 1978. "The Biota and the World Carbon Budget." Science 199:141-46. 6 Forest-Based Industrialization: A Dynamic Perspective Jeffrey R. Vincent and Clark S. Binkley C an industrialization based on forest resources be used to stimulate eco- nomic growth and development? Once there was great hope that the capital contained in forests, and the forward and backward linkages of the forest-products industry, would make this sector an especially appealing target for industrialization and development aid (Westoby 1962). But many industrialization programs for the sector failed to achieve these objectives (Westoby 1978, 1985; Gillis 1980; Douglas 1983). More recently, even eco- nomically successful forest-based development projects have confronted heightened environmental concerns. These concerns range from the fate of old-growth forests, particularly those in the tropics, to the biodiversity con- sequences of forest plantations, particularly those using exotic species. In view of these economic and environmental problems, does forest- based industrialization still have a role to play in development? We believe it does. Many of the past problems, both environmental and economic, can be traced to inappropriate macroeconomic and sector-specific policies. Government policies have tended to foster an unstable macroeconomic en- vironment, to keep wood artificially cheap, and to direct investment toward inefficient processing industries. The results have been physical (as opposed to economic) depletion of forests, inadequate management of natural for- ests, limited establishment of plantations, and underutilized processing ca- pacity. These policies neither serve economic development nor preserve socially important environmental values. If the right policies are followed, however, forest-based industrialization can provide an important source of employment and income and can promote conservation by enabling forests to outcompete alternative land uses. 93 94 MANAGING THE WORLD'S FORESTS Because we view forest-sector policies as a key determinant of project success, in this chapter we address these issues from the standpoint of the sector rather than that of the individual project. We derive conclusions from both theoretical and empirical analyses.1 A DYNAMIC THEORY OF FOREST-SECTOR DEVELOPMENT The few previous attempts to analyze forest-based industrialization as a development strategy have generally neglected the key dynamic aspects of the sector. The dynamics of forest-based industrial development are inter- twined with the dynamics of the forest itself. We trace resource and indus- try dynamics in a conceptual model of a developing country that is initially forest-rich. We consider closed- and open-economy variants of this model and, in the latter case, explore the implications of the size of the country's forest sector in a global context. The annex to this chapter contains the details of this theoretical analysis. Because the conceptual framework that arises from the theory is essential for making sense of the complex real-world dynamics of the forest sector, we summarize here the main points developed in the annex. Industrialization typically begins with an unexploited old-growth forest. The principles of forest ecology dictate that timber growth in such forests is nil, as mortality just balances any net photosynthetic activity. Under such circumstances, any level of harvest will exceed growth, and the inventory of timber will decline. Ecological capital contained in the form of timber is converted to economic capital which can be used in the development pro- cess. The area of forest also will decline as land is converted to agriculture and other uses yielding higher economic returns. The simple facts that harvests are greater than growth, that the inven- tory is declining, and that deforestation is occurring do not necessarily im- ply that exploitation is unsustainable. What matters are the adjustments in economic and ecological systems that are induced by the decline in timber inventory. Timber (stumpage) prices play a key role in signaling increased scarcity and inducing the necessary adjustments. If the stumpage market operates efficiently, rising stumpage prices create incentives to sustain a permanent forest area, albeit one that is smaller than the country's original endowment and one that may be too small if nonmarket values are not fully considered. Timber growth will increase as inventories decline, possibly increasing to the point at which growth equals harvest and the system tracks 1 This chapter does not review or forecast production, consumption, and trade of forest products. Able reviews of global trends are provided by Sedjo (1987) for commercial forest resources and Francescon, Nagy, and Kornai (1983) for trade flows. Forecasts are provided by Kallio, Dykstra, and Binkley (1987) and, with an emphasis on the Pacific Rim, Cardellichio et al. (1989). Stumpage price forecasts are compared by Binkley and Vincent (1988). FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 95 a stable, sustainable path through time. Simultaneously, the processing sec- tor will adjust to become more efficient in the use of wood. Hence, the economically optimal development of the forest sector re- quires conversion of land and capital embodied in old-growth forests. If sustainable development is defined as nondeclining per capita utility (see Pezzey 1989), then a country's sustainable development is in fact enhanced by conversion of forest land to more highly valued uses and by conversion of ecological capital to economic capital. This dynamic view has several important implications for managing the sector, which the annex discusses in detail. In a closed economy, output will decline, as will the ratio of labor per unit of output. Total employment in the forest sector will decline unless downstream processing industries emerge. The forest sector will shrink relative to the economy as a whole. Management costs will increase, and the total rents available to support other govern- ment activities will decline. Effective policies for the sector must anticipate these dynamics. The dynamics of the sector are just as important in an open economy. "Pulse harvesting" - under which all of the commercially valuable timber is removed over a very short period of time-may be the optimal economic path for a small country facing essentially fixed world prices. Unlike the case in a closed economy, the processing industry can expand using im- ported timber if it enjoys comparative advantage. CASE STUDIES What circumstances call for policy intervention to guide development of the forest sector? According to the theory in the annex, the rationale for intervention is strongest in the case of nonmarket values, particularly those associated with irreversibility. In the real world, however, policy interven- tion is much more pervasive. Ownership of forests in developing countries tends to be centralized under state or national governments. Stumpage fees tend to be administratively determined rather than set by the forces of sup- ply and demand, and those fees tend to understate actual stumpage values. Because governments tend to be too impatient to wait for improvements in comparative advantage, or feel unable to force trading partners to open their markets, they attempt to accelerate expansion of wood-processing ca- pacity through such protectionist measures as export restrictions on logs, import restrictions on processed products, and manipulations of the ex- change rate. Governments may also adopt these measures when they are reluctant to acknowledge, or do not realize, that comparative advantage has been lost-for example, when old-growth timber is running out and the need to disinvest in processing capacity is emerging. Given the prevalence of policy interventions, does our theory of forest- sector development have any basis in reality? In the main body of this paper we explore this issue by examining forest-based industrialization since the 96 MANAGING THE WORLD'S FORESTS mid-1950s in three countries2-Peninsular Malaysia, Ghana, and Chile. These countries collectively represent a range of conditions faced in the develop- ing world today. One case study comes from each of the three major regions of the developing world: Asia, Africa, and Latin America. The countries are at various points along the old-growth/second-growth transition. Peninsu- lar Malaysia is now logging old-growth forests. Ghana has completely logged over its old-growth forests but has not developed its second-growth re- source. Chile produces most of its logs from second-growth forests, espe- cially plantations. Exhibits 6-1 to 6-3 show how the three countries compare in terms of trends in roundwood production, domestic consumption share for roundwood, and net exports of wood products. The case studies illustrate the extent to which the dynamic adjustments predicted by theory can be observed in the real world and, more important, the consequences of policies that interfere with these dynamics. The key questions addressed by the case studies are these: What are the consequences of government policies that are targeted at short-term goals and fail to cre- ate economic or political environments conducive to long-term thinking by economic agents in the forest sector? Of policies that distort the trajectories of stumpage prices? Of policies that promote industries that lack compara- tive advantage? EXHIBIT 6-1. Index of Roundwood Production INDEX (1960 = 1) /-- ..... ,,.,., 4 ········································· ··························· ~-· / - ....... / ' ·················· / .. .............. ' ·········/-··· ··········\ I I ' I ,,.... I 'I 3 ................................................................/ ... . - / / 2 .......... ..................... ··········/ ·/ ,,.,. """------' ___ - / .... ........ ................ ............. .............. ,,--~,' ,, ,_,, ~ ·· · · · ;;.;. ' ........... ........................ .... ................... .... - -- - ----- 0 1960 1965 1970 1975 1980 1985 - - Peninsular Malaysia - - - Ghana -Chile 2 Peninsular Malaysia is a subnational unit (11 of the 13 states of Malaysia), but for convenience we refer to it as a country. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 97 EXHIBIT 6-2. Roundwood Consumption Share SHARE 1.2 / r---- ; _____ , __ ---···---.. --.. -- ........................................... / . ....................... l ., ......... ,.,..". .... .. 0.8 """..- ........ ..._ .,,,,,,/ I .... - \ _"':::.<___ ~-,' .1', .I.I ............ ...,,."''''\ I 0.6 ....., ' - - ........... ....................~................. .f.. ....................... ........ .... .... ................... ............ .. '' I ' I ' I 0.4 ' I I I I 0.2 1960 1965 1970 1975 1980 1985 - - Peninsular Malaysia - - - Ghana -Chile EXHIBIT 6-3 . Net Exports Relative to GDP SHARE 0.05 0.04 .............................................................. } ·········-·· I 11\ ,.. 0.03 --·.;;-~------- . ·- - - - . . ... . .... . . =·7·. ;····i'\ '''"'''j''''""''' ' f··-··\\ I 1I\\ I ' 1' \'--I , \ -/ I 0.02 "--- ----------------------,---- / ----- ·--······················ I I ........ ........... , \.................. . . \ ......... ... ' /,..-, I '- ' , /,*' ', I 0.01 ...............=--=..~... ............ . . ,, ............... ..............,. ___ / / ' ' , __ 0 L--.............----====:-==~~ - .-· - ·-- - ·-· ... . . . ... . .. .... . .":' .,. _ ___. ._ :".'. .. . . . . . _ -0.01 1960 1965 1970 1975 1980 1985 - - Peninsular Malaysia - - - Ghana - Chile 98 MANAGING THE WORLD'S FORESTS Three themes emerge from the case studies: 1. Policies must be made with an awareness of the dynamic con- text of the forest sector. Policies and policy-making processes must be forward-looking and must be flexible enough to accommodate the changing nature of the sector as a country develops. 2. Stumpage prices play a crucial role in facilitating the transition of the forest sector from dependence on old-growth to second- growth, and in coordinating processing capacity with timber stocks. Policies related to timber harvesting and forest management must ensure that the appropriate signals about scarcity are sent and received. 3. Processing inevitably declines in a closed economy, and its appropriate level in an open economy (how far to go into down- stream processing?) depends on comparative advantage. Policies related to processing industries must accommodate, not resist, these constraints. After elaborating the case studies in the next three sections, the final section develops these policy prescriptions in more detail. Peninsular Malaysia 3 The forest sector in Peninsular Malaysia has grown rapidly during the past three decades. Annual rates of increase in physical output between 1960 and 1987 averaged 6 percent for logs, 7 percent for sawnwood, and 20 percent for plywood. During this period, Peninsular Malaysia emerged as the world's leading exporter of hardwood sawnwood and a significant ex- porter of plywood. Old-growth timber was an important source of capital for the growing economy. The stumpage value of forests harvested between 1966 and 1985 (estimated by Vincent 1990b) was equivalent to 7 percent of gross capital formation in Peninsular Malaysia during that period. The percentage ranged from 11 percent to 13 percent between 1969 and 1973, but declined to 3 percent by 1985 as capital was increasingly generated by growth in manu- facturing sectors and by external borrowing. 3 Young, Bussink, and Hasan (1980) and World Bank (1989b) were the primary sources of information on overall economic development. The experience of one of this chapter's authors (Vincent) as a researcher and consultant in Malaysia was the primary source of information on the forest sector. Additional sources include Kumar (1986), Mohd. Shahwahid (1986), and Tang (1987). FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 99 Despite this impressive performance, the outlook for the forest sector is uncertain. Policymakers in Peninsular Malaysia are concerned about three problems: a declining supply of logs from natural forests, overcapacity in primary processing industries (particularly sawnwood), and limited devel- opment of downstream industries (furniture, moldings). Peninsular Malaysia's predicament is a consequence of policies that distort timber prices and sup- press scarcity signals. Insecure concessions, combined with low rent cap- ture, have led to high rates of harvest and low rates of forest management. Restrictions on log exports have kept domestic log prices artificially low and raised the rates of return for investments in primary processing, thus leading to overcapacity in these industries. What appears to be a thriving industry in Peninsular Malaysia is tottering on a resource base that could crumble as suddenly as it did in the neighboring Philippines. Resource dynamics determine industry dynamics, and suppression of the signals of resource scarcity affect both. When processing industries are sheltered by protectionist policies, they generate only a low level of eco- nomic value. This country study illustrates these two themes. Forest Dynamics Forest inventories in Peninsular Malaysia were completed in 1972 and 1982, and a new inventory is scheduled to begin in the early 1990s. In 1972, 62 percent of Peninsular Malaysia was forested, and 58 percent of the for- ests were old growth. By 1982 forests had been reduced to 52 percent of total land area. The proportion of old-growth forests also fell to 52 percent. As of 1985, 36 percent of Peninsular Malaysia was designated as permanent forest estate (PFE) to be maintained under forest cover.4 Sixty percent of the PFE-2.85 million hectares-was classified as productive forest, to be man- aged for commercial timber production. Assuming annual growth rates of 1.0 to 1.5 cubic meters (m3 ) per hect- are, the annual sustained yield from Peninsular Malaysia's productive PFE is in the range of 2.85 to 4.28 million m 3 . In contrast, the harvest in 1989 was 12.l million m 3 . Why is the harvest so much larger than sustained yield? We emphasize two reasons here.5 One is the drawdown of old-growth stocks through the conversion of forests outside the PFE and the harvest of old- growth timber within the PFE. The other is rent seeking. 4 The area of the PFE that is actually forested is unknown. 5 A third reason is an "allowable cut effect" due to overly optimistic assumptions about timber growth rates. The Forestry Department has reduced the recommended rotation for forests managed under the even-aged (monocyclic) Malayan Uniform System (MUS) from 70 to 55 years, and promoted adoption of an uneven-aged (polycyclic) silvicultural system, the Selective Management System (SMS), which has an even shorter cutting cycle of 30 years. These reductions in harvest cycles seem motivated more by a desire to justify high current harvests than by the sound application of the findings of silvicultural research. 100 MANAGING THE WORLD'S FORESTS Conversion is driven by the expansion of rubber, oil palm, and, more recently, cocoa plantations. The area of tree-crop plantations increased by 100 percent between 1966 and 1988, rising to 4.2 million hectares. Even if harvests were restricted to the PFE and only an area consistent with sus- tained yield were harvested each year, harvests would still exceed sustained yield levels as long as the harvested area included some old-growth forests, which carry higher stocks of saleable timber than do second-growth forests. Ignoring environmental externalities,6 a high rate of harvest due to old- growth drawdown is worrisome only if there is evidence that responses to timber scarcity do not occur. Increasing scarcity became evident as early as the mid-1960s. Log prices rose 2.0 percent per year in real terms between 1966 and 1985. Trends in stumpage values would be a better indicator of scarcity, but because government timber charges are determined adminis- tratively rather than by a stumpage market, data for analyzing such trends are not available. Stumpage values can be estimated indirectly, however, from a log supply model presented in Vincent (1990b). Implicit resource rents based on this model indicate that, despite increasing access costs as logging has moved from lowland to hill forests, average stumpage values in Peninsular Malaysia rose 1.7 percent per year in real terms between 1966 and 1985. These rising stumpage values might have been expected to stimulate investment in forest management and to dampen timber demand, but these responses have not occurred because of low rates of government rent cap- ture and insecure concession tenure. Ownership of natural forests in Penin- sular Malaysia is vested in each of the 11 states on the peninsula. Yet the states have not responded like a private owner because the timber fees they receive understate rises in stumpage value. Low rates of rent capture were pointed out in the 1960s and 1970s by Teo (1966) and Sulaiman (1977), and more recently by Gillis (1988b) and Vincent (1990b). Vincent (1990b) esti- mated that forest revenue systems in Peninsular Malaysia, which include a combination of royalties assessed on extracted logs and premiums assessed on concession area, captured only about a fifth of the stumpage value of forests harvested between 1966 an 1985.7 Timber charges in most states of Peninsular Malaysia remained unchanged from the early 1970s (when roy- alties were nominally based on 10 percent of the f.o.b. [free on board] log price) until the mid-1980s. In real terms, the average government timber 6 Of course, environmental costs have been incurred during the old-growth draw- down. For example, little of the country's biologically rich, lowland rain forest re- mains. 7 Concessions are occasionally tendered (auctioned) in several states. In one of these states, Pahang, information on concessions auctioned in 1989 imply that roy- alties and premiums captured only 24 percent of the rent, which corroborates Vincent's (1990b) indirect estimate of rent capture for Peninsular Malaysia as a whole. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 101 charge (sum of royalties and premiums expressed on a per cubic meter basis) between 1966 and 1985 rose only 0.9 percent per year-scarcely half the rate of stumpage value increase. Concessionaires have also failed to act as owners, because concession contracts do not serve as proxy for private ownership.8 Concession contracts tend to be much shorter than harvest cycles; small concessions are granted for periods of 6 months to 10 years, whereas long-term agreements typically last 25 to 30 years. As explained in the annex, however, as long as rights to concessions are exclusive and transferable, the length of the contract is es- sentially irrelevant from the standpoint of forest management. What is more important is that concessions in Peninsular Malaysia are allocated on the basis of political patronage. Low rent capture makes concessions a valuable perquisite for those in positions of political power. It is not uncommon for concessions to be revoked after elections, and criteria for their renewal are unclear. Because concessions are generally not well documented, they can- not easily be enforced legally. Simply increasing the normal length of con- cessions would not increase concessionaires' interest in forest management, as it does not overcome these more fundamental problems. Rent seeking, combined with concession insecurity, provides the sec- ond reason that harvests exceed sustained yield. The failure of timber charges to parallel the rise in stumpage values has increased the returns to rent seeking in the forest. Hence, concessionaires log at higher than optimal rates, as they rush to cash in their increasingly valuable but insecure assets before they are taken away. 9 Because the stumpage price mechanism has been disabled, timber demand intensifies rather than slackens. On the supply side, little investment has been made by either the public or private sectors in timber plantations. The most substantial effort by the government to establish plantations is the Compensatory Plantation Pro- gram (CPP), which was initiated in the early 1980s. The CPP seems to have 8 Concessionaires' reluctance to invest in forest management is also affected by the lack of scientific consensus about silvicultural systems for Peninsular Malaysia's forests. The MUS apparently worked well in lowland forests, but most of these forests have long since been converted to tree-crop plantations. Poor regeneration when the MUS was applied in hill forests and a desire to shorten the harvest cycle promoted development of the SMS. A number of Malaysian silviculturists are now skeptical about growth and yield prospects under the SMS, and they have called for return to an even-aged system like the MUS (Tang 1987). 9 Moreover, uncertainty over concession tenure leads concessionaires to adopt a "repeated entry" approach to harv:i!sting. In an effort to maximize expected profits during the uncertain but anticipated short length of concession tenure, a concession is first creamed for its most valuable trees. During the next political timber cycle, residual but commercially valuable stems will be harvested by the same concession- aire, if he is lucky enough to retain control, or by a new one if he is not. Such repeated entry can devastate regenerating seedlings, saplings, and poles. 102 MANAGING THE WORLD'S FORESTS been motivated more by a misguided "allowable cut effect" rationale and by rivalry between agricultural and forestry agencies than by sound eco- nomics. Although the principal species promoted by the CPP, Acacia mangium, was claimed to be capable of producing utility-grade timber for sawnwood at prodigious rates, its biological suitability for Peninsular Malaysia and the marketability of products made from it were poorly researched. (In other countries, a thorough research program has been found to be crucial to successful industrial plantation projects funded by the World Bank [1986b].) Acacia mangium has proved susceptible to a number of unforeseen pests and pathogens, and it is selling only as low-priced chips. State and federal for- estry departments' eagerness to embrace the unproven plantation program seems to have stemmed from the program's usefulness for justifying contin- ued high rates of harvest of the natural forest. Moreover, high rates of return (on paper, anyway) for the CPP gave these departments a justifica- tion for retaining supervision of logged-over forests rather than seeing them transferred to competing departments responsible for establishing tree-crop plantations. In summary, Peninsular Malaysia demonstrates a policy-induced break- down of the stumpage price mechanism, which is crucial for facilitating the transition from old growth to second growth. Although some old-growth forests still remain, harvest behavior is more consistent with the early draw- down stage in which the only forests are old-growth forests. 10 Industry Dynamics It is convenient to analyze the development of processing industries in Peninsular Malaysia before and after 1972, the year in which log exports began to be restricted. 11 Log exports amounted to only 21 percent of log production between 1970 and 1972. Despite marked increases in log produc- tion, their quantity had actually decreased during this period. Between 1960 and 1972, however, production of sawnwood had increased more than three times and production of plywood (starting from a much lower level) more than 160 times. Consequently, competition for logs became more intense: between 1960 and 1972, log prices increased 5.1 percent per year in real terms. The wood-processing industry sought relief from the government, and in 1972 the government complied by banning the export of logs in 10 of the 10 Effects of the impending timber shortfall might be mitigated by the develop- ment of an unexpected and potentially quite substantial source of timber: rubber plantations. In 1989, 1.5 million m3 of roundwood was harvested from rubber plan- tations. This species has enjoyed a rapid increase in popularity for low- and medium-quality furniture in both domestic and international markets. 11 An export tax of 10 percent was imposed on logs in 1970, but the protective impact of this tax was exactly offset (assuming perfectly elastic demand and a 50 percent recovery rate) by export taxes of 5 percent on sawnwood and plywood. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 103 most popular species groups and by raising the export tax on remaining species to 15 percent. These restrictions immediately cut log exports by more than half. The government strengthened the restrictions in succeeding years. Exports of additional species were banned in 1973, 1978, 1979, and 1983. The export taxes on sawnwood and plywood were repealed in 1974, and the export tax on logs of species still permissible to export was raised to 20 percent in 1980. As of January 1, 1985, exports of all logs other than those of small diameter were banned altogether. Before these policies were enacted, the development of wood-process- ing industries probably had proceeded more or less efficiently. The failure of timber charges to capture resource rents had had little effect on mills, which paid world prices for logs and sold their products at world prices as well. The consequent lack of protection for the solid-wood products indus- try is illustrated by effective rates of protection 12 of -2 percent in 1963, -9 percent in 1965, and -5.5 percent in 1969 (from various sources cited in Vincent (1986]). Although the Industrial Incentives Act (IIA) provided some assistance (in 1972, 19 percent of the industry's output was by HA-assisted firms), the main impetus for growth appears to have been the industry's comparative advantage. The log-export restrictions dramatically slowed the rate of increase in log prices, which rose only 0.8 percent per year in real terms between 1972 and 1985. For comparison, prices for logs exported by East Malaysia, which had been comparable to log prices in Peninsular Malaysia in the years lead- ing up to the restrictions, rose 3.2 percent per year during the same period. This insulation of the Peninsular Malaysia log market partially shifted resource rents from concessions to mills, making investments in primary processing more attractive. Many new mills were established, not because they were efficient entrants into a competitive market, but because of the displaced rents. The number of sawmills increased at a statistically more rapid rate between 1972 and 1985 (2.8 percent per year) than between 1960 and 1972 (1.7 percent), even though sawnwood output increased relatively less rapidly between the two periods (5.0 percent versus 11.1 percent). The rents enabled mills to earn profits even though they operated at well below capacity. 13 12 The effective rate of protection equals the ratio of value added at domestic prices to value added at world prices, minus one. Value added for wood processing is defined as price of processed product minus log price divided by recovery rate. 13 Apparently, the least-efficient mills were state-owned "integrated timber com- plexes" (ITCs). As of December 1985, 12 ITCs had been established in five different states (Forestry Department, Peninsular Malaysia 1986). The rationale for ITCs was to gain economies of scale by establishing large, interrelated mills in a single com- plex, and to provide these mills with an assured source of wood by granting the complex a concession. In several instances the complexes have operated their mills at low levels merely to maintain a front for a more lucrative business: leasing out their timberlands to second-party concessionaires. 104 MANAGING THE WORLD'S FORESTS Published estimates of effective rates of protection have understated the protection that log-export restrictions provided the industry. For example, Mohd. Shahwahid (1986) cited effective rates of protection (estimated by, among others, the World Bank) of 30 to 33 percent for sawnwood and 44 to 58 percent for plywood between 1979 and 1982. These estimates were based solely on the protection provided by export taxes on logs and by import tariffs on processed products (as of 1978, 20 percent on sawnwood and 25 percent on plywood). If the effects of quantitative restrictions on log exports are included, the rates of protection are much higher. For example, in 1988 the price of dark- red meranti logs was M$183.60/ m3 in Peninsular Malaysia, while the world price for such logs was M$334.80/ m3 (M$2.70=US$1.00). Sawnwood of the same species sold for M$697 /m3 on the world market. Because the recovery rate was 56 percent,14 the effective rate of protection was 272 percent. The actual economic value created by processing in Peninsular Malaysia was thus quite low: at the distorted domestic prices, value added appeared to be M$369.14/m3 of sawnwood, but at economic (world) prices, value added was only M$99.14/ m3 . Downstream wood-processing industries do not enjoy comparable pro- tection. Unlike sawmills and plywood mills, they must pay world prices for their raw materials (sawnwood and plywood). The federal government has expressed impatience with the limited development of downstream processing (MIDA/UNIDO 1985), and recently it took action to extend pro- tection downstream. On March 7, 1990, the cabinet announced export restrictions on veneer15 of all species and sawnwood in 21 species groups, which encompass virtually all the species with significant export markets. The restrictions take the form of an export tax of M$60 to M$120, which for sawnwood is equivalent at 1989 prices to an ad valorem tax of 11 to 22 percent. 16 For reasons similar to those in the case of the log-export restric- 14 This rate is 5 to 15 percentage points below recovery rates in log-importing countries. By not having to pay world prices for logs, primary processors in Penin- sular Malaysia have not been driven to economize log intake. Their recovery rates are lower also because they saw a more variable, run-of-the-woods mix than do importers, who saw only export-quality logs. 15 No export takes were announced for plywood, because the Malaysian Plywood Manufacturers' Association persuaded the Ministry of Primary Industries that ply- wood is a downstream product, not a primary product. 16 Initially, taxes on most species other than rubberwood will be assessed only on sawnwood strips, the chief raw material used by existing molding manufacturers. The cabinet announced that quotas might be imposed as well. In fact, it announced that an embargo on air-dried sawnwood-which accounts for more than half the volume and value of Peninsular Malaysia's sawnwood exports-will be imposed on September 1, 1991. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 105 tions, this policy might well succeed in creating large industries, 17 but it will probably also create overcapacity and foster inefficiency. It is not obvious that downstream industries need protection. Although the output of moldings and furniture is still small relative to the output of sawnwood, investment in downstream industries has risen healthily in re- cent years. This situation suggests that the government has a better option for redressing the imbalance in protection between primary and downstream processing: it could reduce the protection enjoyed by primary processors by phasing out log-export restrictions, instead of increasing protection for down- stream industries. Is it credible that the world's largest exporter of hard- wood sawnwood is still an infant industry? The higher log prices that would result from rescinding log-export re- strictions offer several potential benefits. They would increase stumpage values (which would increase the returns to forest management), 18 raise recovery rates (hence, less forest would need to be harvested to produce a given amount of processed product), and induce the closure of inefficient mills. Unfortunately, the federal government appears to be moving in the opposite direction. It has discussed with Sabah and Sarawak the possibility of subsidizing log imports from those two states to enable primary proces- sors in Peninsular Malaysia to continue to buy logs at prices below world levels. At every stage of development-from log exporter to exporter of primary products and embryonic exporter of downstream products-Peninsular Malaysia has promoted industrial development by keeping timber royalties, log prices, and, recently, sawnwood prices artificially low. Processing ca- pacity that exceeds the forest's sustained yield capacity has resulted. For a while, export restrictions can create the illusion that wood is not scarce, but to do so inhibits the economic adjustments needed to prevent the physical depletion of forests and to promote the development of efficient processing industries that truly add value to the economy. 17 This outcome is by no means assured. The government appears to be assuming that if foreign buyers cannot purchase Peninsular Malaysia's sawnwood, they will have no choice but to buy its moldings and furniture. Of course, foreign buyers do have other choices. Abundant substitutes for Peninsular Malaysia's tropical hard- wood sawnwood exist. Other suppliers in the region-Sabah and Sarawak in East Malaysia, and perhaps Papua New Guinea, Myanmar (Burma), and Vietnam-will fill some of the gap in export markets. In addition, European buyers have already begun substituting temperate hardwoods and softwoods for tropical hardwoods in some uses, partially in response to the misguided consumer campaign against tropi- cal hardwoods (Vincent 1990a). 18 Obviously, issues related to concession tenure and rent capture would need to be resolved before higher stumpage values would translate into greater incentives for management. 106 MANAGING THE WORLD'S FORESTS Ghana 19 When Ghana became independent in 1957, it was one of the wealthiest nations in sub-Saharan Africa. Misguided macroeconomic policies and political instability have caused the country to suffer economic decline throughout much of the postcolonial period . These same factors are largely responsible for problems experienced in the forest sector. Like Peninsular Malaysia, Ghana illustrates the importance of stump- age prices in stimulating the transition from old-growth to second-growth forests, and the importance of free trade in stimulating the efficient devel- opment of processing industries. In addition, Ghana illustrates how macroeconomic policies set the stage on which resource and industry dy- namics are played out. Macroeconomic policies followed by Ghana during most of the past 30 years have made it nearly impossible for the forest sector to get onto a sustainable growth trajectory. Forest Dynamics Details about the status of Ghana' s forests are scant: the last forest in- ventory was carried out in the 1950s, although a new one was scheduled to begin in 1986. The forest area is known to have declined rapidly. Al- though approximately 8.2 million hectares of Ghana are ecologically classi- fied as a closed forest zone, only about 1.7 million hectares remain under forest cover (FAO 1981). The area of natural forest was reduced by half during the first half of this century, and virtually no old-growth forest exists today. The explanation for forest depletion is similar to that in Peninsular Malaysia: Forests have been converted to tree-crop plantations, primarily cocoa, and management of logged-over forests has been inhibited by con- cession insecurity and low rent capture. After World War II, improvements in infrastructure, rising international prices for tropical timber, and a concessions act that "regularized proce- dures for alienating common-hold lands to private producers" (Page 1974) stimulated British investment in logging. During colonial times, ownership of timberlands was vested in local tribes, and concessions were allocated by essentially a market mechanism: Guided by the concessions legislation, tribal authorities and timber contractors negotiated the terms of concession con- tracts.20 Agreements were validated by the court system, which gave legal 19 Roemer (1984), supplemented by World Bank (1989a), was drawn on for a historical perspective on economic policies. FAO (1981), Asabere (1987), and Grut (1989) provided information on forest resources and forest management. A trio of papers by Page (Page 1974; Page 1976; Page, Pearson, and Leland 1976) covers industrial use of forests up to the early 1970s extremely well. Gillis (1988c), World Bank (1987b), and World Bank (1988) were used as sources of information on forest- based industrialization in the 1970s and 1980s. 20 There were few restrictions on foreign investment in concessions, except that non-British expatriate firms were excluded. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 107 protection to both concessionaires and tribes. This system probably pro- vided concessionaires with reasonable assurance of exclusive use, and mini- mized concerns about uncompensated expropriation. In 1962, however, in keeping with Premier Kwame Nkrumah's nation- alization policies, the central government assumed authority over allocation of concessions. Thenceforth, as in Peninsular Malaysia, concessions were granted in an arbitrary fashion under a system of political patronage. In addition to taking over concession allocation, the Nkrumah government "standardized" (equalized) royalties across concessions.21 When concession terms were negotiated between concessionaires and tribal leaders, it was at least theoretically possible for royalties to reflect differences in stumpage values. Standardization eliminated this possibility.22 Royalties were set at levels far below stumpage values. As in Peninsular Malaysia, this practice made concessions a tempting political prize. Conces- sionaires who sold logs on the international market were able to reap enor- mous profits, especially before restrictions on log exports were imposed in 1972. Page et al. (1976) estimated that 67 to 88 percent of the value of export- quality logs of leading species was stumpage value, and that the govern- ment captured only 38 percent of this value through the combination of area-based concession fees, royalties, export taxes, and corporate income taxes. Low rates of rent capture were also reported more recently by the World Bank (1987b).23 Would Ghana have been better off if the government had captured more of the rent? The resource rent realized by harvesting Ghana's forests between 1961 and 1970 was an important, but not the primary, source of capital for the country-equivalent to 6 to 8 percent of gross capital forma- tion in the country during this period. 24 In view of the government's squan- dering of public revenue and net foreign exchange reserves during Nkrumah's misguided "Big Push" phase, skeptics might be relieved that the govern- ment did not capture more of the rent. Undoubtedly the 35 percent of the rent that was captured by local concessionaires earned higher returns. Off- setting this possible benefit was Ghana's failure to retain much value from the remaining 27 percent of the rent, which was largely repatriated by for- eign concessionaires. 21 Page (1976) says that royalty rates were equalized in 1962, but Gillis (1988c) states that equalization was not achieved until 1988. 22 Moreover, royalties failed to reflect stumpage value because they were assessed on a per-tree basis. Although they were differentiated somewhat by species, for a given species the same royalty applied regardless of whether a tree had a large or a small commercial volume. 23 On a positive note, another report by the World Bank (1989a) has indicated that Ghana is considering reforms involving higher royalties and market-based alloca- tion of concessions. 24 The estimate of realized rent is based on data from Page et al. (1976, table 11) and Page (1974, table 4.1). 108 MANAGING THE WORLD'S FORESTS What is certain is that no one-government or concessionaire-behaved as owner of the resource and made appropriate investments in forest man- agement as timber became increasingly scarce. The government did not act as owner because royalties did not signal the increasing scarcity value of the forest. It might also be argued that government management efforts were hindered by low rates of rent capture, which reduced the funds available the forestry department, which has suffered chronically from a lack of funds for equipment, transportation, and training. Given the government's zeal to promote manufacturing, however, it is unlikely that the department would have received much of any incremental revenue to the treasury. Concessionaires did not act as owners because, as in Peninsular Malay- sia, they lacked secure concession rights. 25 Reductions in concession size, which apparently ran counter to returns to scale in forest management (World Bank 1987b), might also have reduced concessionaires' economic interest in management. In the early 1960s, nationalistic tendencies coupled with Nkrumah's socialistic bent led the government to adopt a variety of "Ghanaianization" programs. In the forest sector, the government stopped issuing concessions to foreigners and expatriates and reduced the average size of concessions by more than an order of magnitude (Page 1976).26 On the face of it, smaller concessions might appear to be an egalitarian attempt to spread the country's natural wealth, but in fact they were a means of widening the web of political patronage. Did harvests in Ghana exceed the optimal rate? The answer to this question is not clear. On the one hand, the high stakes involved in conces- sions, magnified by insecure tenure and political instability, surely com- 25 Not all the obstacles facing natural forest management have been policy in- duced. Regeneration technologies are still not well established. Between the end of World War II and 1970, the forestry department tried and abandoned three silvicul- tural systems: the tropical shelterwood system, enrichment planting, and a modified selection system (Asabere 1987). In 1971, it reverted to a simple salvage felling of overmature trees in logged-over forests. Silvicultural operations are hindered by low annual growth rates-0.SS m 3 per hectare (World Bank 1987b), as compared with 1.0 to 1.S m3 per hectare in Southeast Asian forests. These low growth rates do not result entirely from natural causes. Diameter cutting limits are extraordinarily high (more than 1 meter), logging is very selective, and much wood is wasted in the forest (33 to SO percent of the harvest according to Grut [1989], SO to 100 percent according to the World Bank [1988]). The first two factors are often blamed on the highly specific demands of international timber buyers, but they might also be caused by insufficient differentiation of royalties, which lead to relatively greater profits on logs that are large and of desirable species. 26 Between 1900 and 1960, 85 percent of concession area was issued to non-Ghana- ians; between 1961 and 196S, 16 percent; and after 196S, zero percent. Concession size declined from an average of 69,000 hectares in the 1900-60 period, to 11,000 hectares in the 1961-65 period, to 4,000 hectares after 1965. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 109 pelled concessionaires to cash in their timber as quickly as possible. On the other hand, the rate at which timber could be cashed in was reduced by the country's macroeconomic policies. These policies-in particular, overvaluation of the cedi and Ghanaianization of logging firms (in addition to conces- sions)-reduced harvests by hindering log exports. After rising rapidly during the late colonial period, log exports fell by about 60 percent during the 1960s and by an order of magnitude between 1970 and 1982. The overvalued exchange rate directly hindered exports by reducing the number of cedis-needed for paying domestic factors of production- that were earned per dollar of exports. Overvaluation indirectly hindered exports by leading to a shortage of foreign exchange that only became worse as exports declined. The shortage of foreign exchange made it difficult to import equipment or spare parts. Logging recovered briefly during the lib- eralization efforts of the post-Nkrumah regime (the 1967 devaluation par- ticularly helped), but logging was hurt again when import controls were reimposed in 1972. Ghanaianization of logging companies was the objective of the Ghana Timber Cooperatives Union (GTCU), which provided loans to Ghanaian logging firms at below-commercial rates. Collection of loan payments was lax, and the GTCU degenerated into a grants scheme for free capital. De- spite free capital, this industrialization strategy raised the industry's aver- age production costs and depressed output, because local logging compa- nies were less well managed than foreign or expatriate firms and tended to be too small to take advantage of increasing returns to scale. Page et al. (1976) found that logging costs for very small Ghanaian firms were more than twice those of large foreign firms. Many Ghanaian logging firms re- mained in business only by virtue of incurring virtually no costs for timber (thanks to low royalties) or capital (thanks to the GTCU). The 1983 devaluation contributed to a recent increase in harvests. Ac- cess to foreign exchange has been improved by provisions that allow timber exporters to retain 20 to 35 percent of export earnings in the form of foreign exchange. Abolition of the Ghana Timber Marketing Board, which had been set up soon after independence to assist the private sector in export promo- tion but instead became a price-setting board, also has helped. Like Peninsular Malaysia, Ghana has shown little success in establish- ing plantations as an alternative source of timber to the natural forests. 27 Concessionaires' interest in plantation establishment is limited as long as they can acquire natural forest timber at low royalty rates. The government initiated an industrial plantation program in the 1960s, but because of a tight budget-which, theoretically, could have been increased by higher rent capture-only 15 percent of the target area was planted. 27 Unlike Peninsular Malaysia, Ghana has not been able to turn to its tree crops as a new source of timber: in contrast to rubber trees, cocoa trees do not yield commer- cial timber. 110 MANAGING THE WORLD'S FORESTS Industry Dynamics The first sawmill in what was to become Ghana was set up in 1903. By 1957, Ghana had 31 export-oriented sawmills financed by European invest- ment and producing for export to European markets. Like log exports, how- ever, sawnwood exports stagnated during the 1960s and fell by three-fourths between 1970 and 1982. Wood processing, primarily sawmilling, declined from 40 percent of Ghana's manufacturing output in 1958 to 13 percent by 1969. The macroeconomic policies that hindered logging were also respon- sible for the poor performance of processing industries. The overvalued cedi decreased exports of processed products for the same reasons it de- creased log exports. Its negative effects overwhelmed policies that might otherwise have stimulated wood exports. These policies included the general income tax incentives and cheap capital (from the National Development Bank) available to manufacturing industries (although import-substituting industries were favored); log-export bans on selected species announced in 1972, 1976, and 1979; and removal of a 5 percent export tax on sawnwood in 1979. The industry's poor performance between 1972 and 1983 can also be ascribed to an ill-fated nationalization effort. Despite an increase in the number of Ghanaian-owned mills, most output in the 1960s continued to come from foreign- or expatriate-owned mills. In 1970, 47 percent of lumber and 64 percent of plywood, as well as 42 percent of logs, were produced by five large, vertically integrated companies. Three were foreign owned, one was owned by expatriates, and one was Ghanaian. In 1972 the government expropriated majority holdings in the foreign and expatriate companies. As an added step, in October 1973, all foreign and expatriate mills were forced to surrender 55 percent of their equity to the government. The nationalized mills have performed abysmally. According to the World Bank (1987b), all but one of the four firms taken over in 1972 "have experienced severe man- agement problems, been closed part time, and are burdened by debts." Although output and exports of sawnwood were stagnant from the early 1960s until the early 1980s, the number of sawmills increased from 43 in 1962 to 61 in 1971 and 95 in 1982. Cheap capital under the government's Ghanaianization programs was largely responsible. Virtually all the new mills were small, Ghanaian-owned mills. Some of the increase in new mills was also related to the government's sporadic efforts to make concession allocation contingent on investment in mills, which induced concessionaires to build mills of a minimum size in order to retain concessions. This expan- sion of wood-processing capacity was costly in terms of diverting capital to unproductive uses: even in 1985, after the industry had begun to revive, rates of capacity utilization were only 57 percent for sawmills, 39 percent for plywood mills, and 57 percent for veneer mills. Studies by Page (1974, 1976) indicate that the country's logging and processing industries were inefficient and provided little net gain to the FOREST-BASED /NDUSTRIALIZA TION: A DYNAMIC PERSPECTIVE 111 economy. Page (1974) analyzed financial value added (revenue minus pay- ments for intermediate goods) and economic net gain (revenue minus social opportunity costs of factors and intermediate goods, minus payments made abroad) for logging, sawnwood, and plywood industries in 1970-71. Al- though logging is seldom considered a value-added activity, Page found that Sl percent of the value of log output was value added. Much of this value was in the form of rents; private-sector profits, depreciation, and re- source rent accounted for more than half, and royalty payments to the gov- ernment for another fifth. Net gain in logging was substantially less than value added; net gain constituted 22 percent of the value of output, and virtually all was resource rent. Net gain could have been increased by about half if the government had captured resource rents that were repatriated abroad. Page's analysis reinforces Gillis's (1980) point about the importance of rent capture in achieving economic gains when foreign-owned firms are involved in commercial timber extraction. Page (1974) found that value added for sawmills and plywood mills was actually less than for logging: 30 percent and 38 percent of the value of output, respectively. Net gains were even less: S percent and 6 percent, respectively. Net gain was actually negative for small Ghanaian wood-pro- cessing firms. Ghana's efforts to promote domestic processing and domestic ownership of mills thus reduced the generation of economic surplus, which is the ultimate source of economic growth. Using the same data, Page (1976) rephrased the analysis in terms of "social profitability," which is the same as net gain except that repatriated payments are not subtracted. Although 86 percent of the logging firms in the sample earned positive social profits, the sample was heavily biased toward the more efficient, vertically integrated, foreign- or expatriate-owned firms. Social profits decreased with level of processing. Although all ply- wood mills and 80 percent of sawmills earned positive social profits, scarcely half (SS percent) of furniture mills did so. This difference between primary and secondary processing is easy to explain. Because (as in Peninsular Ma- laysia before 1972) log exports were permitted, plywood mills and sawmills had to pay world prices for log inputs, and because they were export ori- ented, they sold at world prices as well. This exposure to international competition in both input and output markets forced the plywood mills and sawmills to be more competitive. Furniture mills, by contrast, were highly protected by import barriers against foreign furniture and were one of Ghana's dismal failures with import substitution. Page's estimates of value added, net gain, and social profitability are probably biased upward. Because his studies were carried out during a short-lived period of tentative economic liberalization, they measured per- formance that was better than the performance in the more interventionist periods before 1966 and aft~r 1971. Moreover, Page (1976) mistakenly equated social profits with efficiency. Positive social profits are only a necessary, not a sufficient, condition for efficiency in the normal economic sense, which 112 MANAGING THE WORLD'S FORESTS pertains to maximization of social profits, not their mere existence at a posi- tive level. At given levels of factor consumption, did firms produce the maximum possible output? Did they produce a given level of output at minimum cost? Page ignores these questions, but they are obviously impor- tant, given the distorted prices of capital and timber in Ghana. Excess processing capacity continues to afflict the industry. Processing capacity currently exceeds roundwood intake by a factor of two and the estimated sustained timber yield by 27 percent. As a result the wood-pro- cessing industry has demanded increased log-export restrictions (as of 1989, exports of 18 species were banned). Even the World Bank (1987b) suggested that log-export restrictions might be necessary to limit harvests. The error in this thinking is clear: Log-export restrictions will simply make logs cheaper for already inefficient domestic processors, and restrictions might actually stimulate increased harvests if domestic demand for logs is more elastic than international demand. One year later, the World Bank (1988) was more circumspect about log- export restrictions, noting that the export ban on selected species "does not appear ... [to have] been successful to the extent expected in either discour- aging depletion of the forest or in promoting technical and economic effi- ciency in the processing industry." The World Bank called for a review of the ban (a study was planned for 1989), with a key objective being to deter- mine whether the industry needs infant industry protection. It is hard to believe that an industry that is nearly 90 years old is still in its infancy. Senility-many mills need to be put to rest, now that the old-growth draw- down is complete-is a more likely diagnosis. Chile 28 Chile's economically accessible natural forests were largely depleted by the early 1950s. Two formerly important species, Fitzroya cupressoides and Araucaria araucana, are so depleted that the existence of a single tree of either species per hectare qualifies an area as a legally protected forest. Natural forests remain chiefly in the extreme southern part of the country, but few are economically accessible. In contrast to Malaysia and Ghana, where the key issues pertain to coordinating industrial development with the drawdown of publicly owned natural forests, the issue in Chile is how to build up an efficient industry based on the nation's increasing supply of timber from its privately owned plantations. A principal message of the cas.e study is the importance of 28 Corbo and de Melo (1985), Corbo and Sanchez (1985), and World Bank (1980, 1987a) provided general information on Chile' s economic development. World Bank (1986a), Hunter (1987), Santa-Cruz (1988), and Jelvez et al. (1988, 1989a, 1989b, 1990) provided information on the forest sector. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 113 secure land tenure in encouraging investment in second-growth forests. The case study also reinforces points made in the Ghana case study about the importance of macroeconomic stability. Resource Dynamics Chile had 22 million hectares of forest in 1988, most of it (13 million hectares) in parks and other noncommercial protected areas. The remaining so-called commercial forests were primarily native hardwood forests (7.6 million hectares), but only 3.0 million hectares were considered economi- cally accessible. Plantations covered 1.3 million hectares, mainly Pinus ra- diata (l.l million hectares). Between 1931 and 1974, plantations were promoted by a decree that provided tax exemptions for reforestation and subsidies for seeds and seed- lings. By 1974 more than a quarter of the area of radiata plantations existing in 1986 was already established. Another law, D.L. 701, which took effect in 1974 and continues until 1994, strengthened the promotion of plantations in two main ways: (1) It declared that land of "forest aptitude" cannot be nationalized, thus providing investors with secure, long-term tenure, and (2) it made available several financial incentives, including a subsidy of 75 percent of planting costs (later extended to pruning and management costs). Tax breaks also were provided: income tax rates were reduced by 50 per- cent, and property taxes were eliminated. D.L. 701 provided one of the few subsidies granted to any sector by the military government, and it was the only incentive provided specifically to the forest sector. Although the trend in planting was upward between 1970 and 1974, most of the planting was by the forestry agency, CONAF. In fact, private planting decreased between 1970 and 1973, as a result of concerns about economic stability and potential nationalization of plantations. Since 1974, 74 percent of planting has been by the private sector. Today almost all plantations are privately owned, even those initially established by CON AF. There is little doubt that D.L. 701 has been instrumental in getting the pri- vate sector involved in establishing plantations, although the more stable macroeconomic environment after 1973 certainly contributed. The interest- ing question is, which of D.L. 701's provisions was more important, the subsidy or the assurance of land tenure? Most analysts have assumed the former (e.g., World Bank 1986b), but between 1974 and 1986, more than a third of the area of plantations estab- lished by the private sector did not qualify for D.L. 701 assistance. Accord- ing to Santa-Cruz (1988), D.L. 701 increased the real financial rate of return for plantation investments only slightly, from 9.6 percent to 11.7 percent per year. The actual increase might have been even less: Santa-Cruz (1988) ig- nored the fact that one would expect the subsidies to be capitalized into land values, thus raising land acquisition costs for those who needed to buy land to establish plantations. There is evidence the subsidies did indeed inflate land prices (Hunter 1987; World Bank 1986a). 114 MANAGING THE WORLD'S FORESTS We suspect that D.L. 701 has been successful primarily because it pro- tected forested lands from nationalization, thereby assuring investors that they could reap high enough returns from investments in radiata with or without subsidies. Radiata plantations earned an attractive rate of return because of Chile's comparative advantage in growing wood. Chile's climate and soils provided ideal growing conditions, which yield annual growth rates of more than 30 m 3 per hectare. (Contrast this with Ghana's and Malaysia's rates of growth in the natural forest of 0.55 m 3 and 1.0 to 1.5 m 3 per hectare respectively.) Land was cheap because competition from agri- culture is limited. Easy terrain and a competitive transportation sector put much of the lands suitable for radiata within economic hauling distance of mills and ports. Labor was relatively cheap because the 1974 reforms re- duced underemployment and provided an excess labor pool. This expansion might have occurred decades sooner had land tenure been more secure and economic conditions more stable. Instead, the Chilean forest products industry suffered through an undue hiatus in the transition from natural forests to plantations. Industry Dynamics Radiata provides Chile with remarkably low-cost roundwood for indus- trial processing: according to the World Bank (1986a), prices for roundwood delivered to mills ranged from US$9/m3 for pulpwood to US$12/ m 3 for sawlogs. 29 These are a fraction of corresponding values in North America and Scandinavia. Annual harvest of radiata roundwood is projected to reach 27 million m 3 by the year 2000, with sawnwood and pulpwood comprising roughly equal shares. In contrast, in 1987 roundwood production in Chile was only 12 million m 3 • During the 1990s, Chile faces the challenge of in- creasing its industrial capacity and perhaps its log exports in order to more than double its roundwood consumption. There are obstacles to achieving this goal on both the resource and industry sides. On the resource side, at least two problems will complicate industrial- ization based on this new resource. First, because nearly three-quarters of the plantations have been established in just the past 15 years and because the rotation age is 24 to 30 years, harvests will tend to be uneven over time. This unevenness might be partially overcome by continued establishment of plantations to round out the age-class distribution. Land availability should not be a constraint: Hunter (1987) estimated that radiata plantations would be economically viable on nearly 1 million additional hectares. Second, much of the wood from the plantations is likely to be of low quality because most plantations are unpruned and unthinned. The poor 29The World Bank (1986a) estimated that D.L. 701 financial incentives reduced wood costs by only US$1.20/m3 . FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 115 state of plantations could theoretically be due to the fact that the plantation area has expanded beyond its efficient size, causing owners to be concerned that wood prices will be too low to justify additional management activities. Owners might have a more compelling reason to be concerned about stub- bornly low wood prices: imperfect competition. Chile's two largest forest- products companies, which are primarily pulp and paper producers, own 40 percent of the plantations and account for a comparable portion of the country's total roundwood consumption. These two companies might hold enough market power to collude and keep roundwood prices low. Because they produce only 15 percent of Chile's sawnwood, they might have little concern about inhibiting the pruning and thinning activities needed to pro- duce high-quality sawnwood. On the industry side, growth hinges on conditions in export markets. About two-thirds of the roundwood consumed in the late 1980s was ulti- mately sold in the form of exported products. The country's domestic mar- ket is small, and it uses little wood in housing is low. 30 The nation's Spanish heritage, the cultural association of wooden housing with rural poverty, and concerns about fire safety lead to a preference for masonry houses. Recent history suggests that the Chilean processing industry is capable of taking advantage of increased export opportunities, if they are not pre- cluded by the low quality of radiata roundwood. Between 1977 and 1987, increased exports stimulated a doubling of total roundwood consumption. The forest-products share of Chile's export earnings rose from 4 percent in 1970 to 11 percent in 1987. This growth reflects the fact that wood-process- ing industries were among the sectors that benefited most from the reforms of the 1970s. During the 1960s, sawmills suffered in comparison with other sectors because of low or negative effective rates of protection (21 percent in 1961 and -4 percent in 1967, according to Corbo and Sanchez [1985]). Al- though the pulp and paper industry enjoyed larger effective rates of protec- tion (41 percent in 1961 and 95 percent in 1967, according to the same source), even these were a fraction of the average for all manufacturing industries. Discrimination against wood processing w as eliminated between 1974 and 1980 as rates of protection were made uniform across sectors and reduced. In 1987 sawmills consumed 46 percent of Chile's roundwood output, and mills making panel products consumed another 3 percent. The industry has converted to almost exclusive use of radiata: In 1960 almost 70 percent of its log consumption was native species; today nearly 90 percent is radiata. Sufficient capacity might at first glance appear to be in place to use the additional sawnwood that will become available during this decade; as in Peninsular Malaysia and Ghana, capacity utilization rates were low in 1984- 46 percent for sawmills and 65 percent for plywood mills. These low rates 30 Per capita consumption of sawnwood is only one-seventh that in New Zealand. 116 MANAGING THE WORLD'S FORESTS reflect the existence of a large number of small and portable sawmills, which form a fringe of producers that open and shut according to the business cycle. Out of 1,597 mills operating in 1985, 1,243 produced only 16 percent of output and 700 were portable. It is doubtful whether these small mills, especially the portable ones, can be relied on to pick up the slack in round- wood consumption, especially if it is production for export. The World Bank (1980) commented that sawmilling was inefficient, with low recovery rates (48 percent, according to Hunter [1987)). Several factors suggest that recovery rates at this level might be efficient for Chile: its exceptionally low sawlog prices (which induce mills to substitute toward wood); its low minimum sawlog diameter (20 cm); its limited domestic market for absorbing products that are not export quality; and its policy of permitting log exports (since 1975), which, unlike the situation in Peninsu- lar Malaysia, forces mills to pay world prices for sawlogs. The pulp and paper industry consumes less roundwood (35 percent of the total in 1987) than does the solid-wood products industry, but it ac- counts for more than half of the export value of forest products. Since 1974 the effective rates of protection for the industry have actually run about 20 percent higher than the manufacturing average. Because the industry is export oriented, the moderate level of protection (17 percent in 1979) has probably had limited effect on the industry's efficiency (capacity utilization approaches 100 percent). Moreover, the industry must increasingly compete against international buyers for pulpwood and woodchips. Exports of these raw materials have benefited from an export promotion program, ProChile, which provides grants for increasing exports of "new" products.31 The moderate, essentially uniform protection across sectors in Chile and the largely free trade of both raw materials and processed products suggest that wood processing in Chile has indeed added value to the economy during the past one to two decades. Investment in both Chile's plantations and processing industries has responded primarily to forces of comparative advantage, not to a mirage created by government policies. For growth to be sustained, the government must continue to provide a stable political and economic environment, monitor domestic stumpage prices to investi- gate whether imperfect competition is driving down prices and hindering plantation maintenance, and resist the certain-to-increase calls for log and pulpwood export restrictions as the industry's capacity begins to match the productive potential of the plantations. This last point is crucial if Chile in the early 2000s is to avoid being a reincarnation of Peninsular Malaysia since the early 1970s. 31 "New" products are defined as those with an export value of less than US$2.5 million in 1983-84. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 117 SUMMARY AND POLICY RECOMMENDATIONS Many of the implications of the foregoing analysis have been pointed out for resource-based industrialization in general. Roemer (1979) and Gillis et al. (1987, chaps. 15 and 19) have stressed the importance of enhancing macroeconomic stability, allowing prices to signal scarcity, and using com- parative advantage to guide the development of downstream industries. Lewis (1989) noted several similarities among successful primary exporting countries: • They "have not allowed their domestic currencies to become substan- tially over-valued." • They "have captured a significant amount of the rents accruing from the unique characteristics of their primary products and have rein- vested them in the economy." • They "were generally modest in the levels of protection provided." • "Not only have new exports been promoted but old industries have been allowed to contract." • They have provided "government investment in infrastructure, train- ing, research, and services to support the primary export sector," with "research to support productivity growth seem[ing] especially important." The case studies illustrate the consequences of failing to recognize these points in one or more of the countries analyzed. 32 Auty's (1987) conclusion that the performance of wholly state-owned resource-based industries has generally been disappointing is supported by the poor performance of state- owned wood-processing facilities in Ghana and in Peninsular Malaysia. 33 32 For example, see the discussions of Ghana and Chile regarding the first point, and Peninsular Malaysia and Ghana regarding the second, third, and fourth. We have not addressed the fifth point directly. As the work of Solow (1974) and others has shown, ongoing gains in productivity due to technical change are fundamental to sustained resource-based growth. It is important to realize that simply making improved technologies available to the forest sector-for example, by research at government-supported forestry research institutes-is insufficient. Industries must also have an economic interest in adopting the technologies. For example, sawmills will be less likely to invest in wood-saving equipment as long as export restrictions on logs keep log prices artificially low. This point is not always grasped. For ex- ample, Takeuchi (1983) recommends both log-export restrictions to promote domes- tic processing and investment in research and development to make processing more efficient. 33 The poor financial performance of state-owned enterprises in the wood-based and other sectors in Malaysia was noted by the World Bank (1989b). 118 MANAGING THE WORLD'S FORESTS In addition to these general points, the theoretical analysis and the case studies emphasize four policy-relevant features of the forest sector that are more specific to the sector: land-use competition with agriculture; valuation of nontimber forest values; the old-growth/second-growth transition; and, associated with this last point, the inevitable shrinking of the relative im- portance of the forest sector. The policy implications of our analysis are summarized in the following five recommendations for actions necessary to ensure that forest-based in- dustrialization promotes long-term, sustainable development. 1. Stabilize Macro Policies Industrial development generally demands macroeconomic stability. Without stability, long-term capital investment is impossible. Development of the forest sector-its infrastructure, processing facilities, and the forests themselves-requires unusually large amounts of capital deployed over par- ticularly long periods. As a consequence, macroeconomic stability is unusu- ally important to the efficient development of the sector. 2. Establish Correct Stumpage Prices Stumpage prices must reflect timber's true competitive value. Timber prices signal the relative scarcity of wood and thereby coordinate virtually all of the key facets of forest-sector development. If this important function is disabled, as it is many developing countries where fees paid for timber are administered royalties rather than prices formed in competitive mar- kets, a variety of economic and environmental maladies ensue: • 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 and plantations. • Inefficient processing technologies are installed. • Decisions on log- and lumber-trade policies are incorrect. • Elaborate and counterproductive capital-export controls are needed to ensure that resource rents remain patriated. Privatization of forest resources can produce the appropriate price sig- nals, but so can other institutional arrangements. Because full-scale private ownership is culturally, socially, or politically unacceptable in many places, the international community needs to emphasize the development of insti- tutions that simulate market outcomes. Although tenure arrangements are beyond the scope of this chapter, we emphasize the need not only for secu- rity (exclusivity) of concession rights but for transferability as well. If tenure rights are reasonably secure and concessions can be sold, concession hold- ers will face the "correct" implicit intertemporal timber values, even if offi- cial timber fees bear no relationship to market prices. This situation implies FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 119 that rent capture through efficient government fees is neither a necessary nor a sufficient condition to ensure the economic efficiency of the forest sector. (See the annex to this chapter for a more thorough discussion.) 3. Maintain Free Trade The objective of increasing domestic value added frequently leads coun- tries to abandon free-trade policies on wood products and to restrict exports of logs and imports of processed products. Unfort.unately, such countries often lack comparative advantage in wood processing (although they may gain it over time). These policies might promote impressive development of wood-processing capacity, but at the cost of explicit subsidies or implicit ones such as trade barriers. This cost may exceed any incremental gains to the economy from the additional processing. Economic value may actually be destroyed, not added. Many countries that have adopted this strategy appear to have overes- timated the opportunity to replace log exports with the export of finished products. Importing countries typically have considerable latitude to offset restrictions on the availability of raw materials with similar wood from other countries, with different kinds of wood from other countries, or even with nonwood materials altogether. Because timber is a fairly homogeneous commodity with many close substitutes, log-export restrictions tend to pro- duce changes in raw material sourcing by the importing country rather than sustainable process deepening by the exporting country. Free trade can partially offset distortions induced by the divergence between government timber charges and true stumpage values. If logs can be freely imported and exported, both the logging sector and the processing sector will encounter the correct short-run prices. As long as the country is "small" with respect to world markets- that is, as long as its actions do not affect world prices-the processing sector can develop efficiently even if timber-pricing policies are incorrect. Distortions can still occur, however, if the country is a "large" exporter. For example, excessive development of wood-processing capacity can result if insecure concession rights lead to overly rapid forest depletion and thereby depress log prices. This problem is made worse if log exports are restricted. Conversely, wood-products boycotts by consuming countries will have perverse effects on the development of the forest sector in an open economy. The boycotts will reduce the price of timber and this reduced price will lead the processing and harvesting sectors to waste wood. Investment in forest management will lag, and forests are less likely to be protected from en- croachment and more likely to be converted to other, nonforest uses. Boy- cotts will reduce hard-currency earnings for exporting countries, forcing them to export still more timber to those countries that do not participate in the boycotts. In short, boycotts on the import of tropical timbers are likely to be ineffective and counterproductive. 120 MANAGING THE WORLD'S FORESTS 4. Perform Sector-Level Analysis In considering policies for the forest sector, planners must disentangle the economically optimal drawdown of old-growth timber inventory from overcutting that is due to inappropriate policies or incorrect economic sig- nals. Because there is no simple way to make such a determination, plan- ners need to perform a careful, far-reaching examination of the forest sector in individual countries before proposing individual projects or approving structural loans. At a minimum, such analysis should include tenure insti- tutions, policies for setting stumpage fees, forest management procedures and requirements, factor costs, inducements for processing industries, trade policies, and subsidies for, or limitations on, domestic demand. This critical need for comprehensive forest-sector analysis does not nec- essarily imply the desirability of comprehensive implementation of forest- sector plans. The scarcity of administrative resources in most developing countries demands that change be incremental rather than comprehensive. Planners need to ask, How will a series of incremental changes lead to the necessary reform? This need for careful attention to policy dynamics is reinforced by the economic dynamics of the forest sector. Policies should be adapted to the changing needs of the sector. Policy mechanisms that rely on continuous, incremental adaptation are much more likely to succeed than are ones that depend on sporadic "comprehensive" intervention. 5. Remember the Need for Environmental Protection Many of the environmental problems stemming from forest-sector de- velopment result from one or more of the policy failures noted earlier. For example, administered pricing systems might lead to too rapid depletion of old-growth forests, with consequent deleterious ecological effects. Thus an economically efficient forest sector will be a more environmentally benign one as well. Yet because so many of the products of the forest-clean water, biologi- cal diversity, nontimber products used by indigenous people, and so on- do not enter markets, even an otherwise efficient forest sector will system- atically undervalue and underproduce these products. In theory such concerns can be incorporated into a market economy, but the practical arrangements for doing so have proved to be beyond the policy sophistication of even the most politically advanced countries. ANNEX This annex provides a conceptual perspective on dynamic relationships within the forest sector (resource depletion and processing expansion) and between the forest sector and the developing economy as a whole. We present three scenarios. In all three we begin with a country that is endowed with an old-growth timber resource, and we examine how the forest sector de- velops autonomously, that is, in the absence of policy interventions. We FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 121 assume that secure private property rights prevail. After presenting the basic scenario results, however, we address the issue of designing alterna- tive institutional arrangements to emulate, without full-scale privatization, the efficiency of market solutions. The first scenario concerns a country with a closed economy (autarky). Once this scenario has been developed, we can readily present the other two, which relate to open economies, by highlighting points of difference with the autarky case. In the second scenario, development of the country's timber resources has an insignificant effect on prices in world markets. In the third, the effect is significant. In the two open-economy scenarios we restrict our attention to countries with export surpluses in the forest sector. As is typical in economic analysis, we hold many factors constant to focus on the main issues at hand, with full knowledge that the real world does not offer such constancy. Forest-Sector Development in a Closed Economy The first part of the story is the comparatively old and well-known saga of the dynamic adjustment of the timber stock. 34 The second part links the dynamic adjustment of the timber stock to the dynamics of forest industries and the macroeconomy more broadly. Because economists have attended less to the details of the second part of the story, it has neither the clarity nor the certainty of the first. 35 We begin with the situation where private 34 The problem of dynamic exploitation of a timber stock has been extensively treated in the literature. Clark (1976), Anderson (1976), Heaps (1984), and Heaps and Neher (1979) all treat this problem with an ageless inventory. More realistic models which incorporate the important age-class structure of forests include Wan (1966), Walker (1971), Jungenfelt (1973), Berck (1976), Tedder, Schmidt, and Gourley (1980), Lyon and Sedjo (1983), Johansson and Lofgren (1985), Wan and Anderson (1983), Kemp and Long (1983), Mitra and Wan (1985), and Brazee and Mendelsohn (1990). Some of these models take demand to be perfectly elastic so that prices are fixed, while others relax this assumption to maximize total surplus with a demand curve either fixed in time or shifting in a known manner. None of these models goes beyond partial equilibrium analysis of timber markets, and none discusses issues related to processing or capital and labor markets. 35 Although the economics of forest dynamics has never been formally linked to sector- or economy-level dynamics, several recent papers cover some pertinent ground. Solow (1974) linked resource use to a macroeconomy, to show that contin- ued increases in per capita consumption were possible from a finite resource stock if technical progress occurred rapidly enough. Pezzey (1989) also considered the sustainability of development in economies that include natural resources. Ehui, Hertel, and Preckel (1990) considered the trade-offs between agriculture and for- estry in a developing country, thereby dealing with the land-use aspects of our analysis. Panayotou (1989) examined the trade-offs between investments in resource exploitation (in this case, minerals) and investments in other sectors of a developing economy. 122 MANAGING THE WORLD'S FORESTS property obtains, and then tum to other forms of tenure. Although our exposition focuses on industrial use of forests, toward the end of this annex we comment on connections between industrial development and environ- mental values of forests. Resource Dynamics with Private Property In the early stages of development, dynamic equilibrium in two mar- kets-that for land and that for timber-causes the country's stock of old- growth timber to diminish. Forested land is likely to have higher value for agriculture and other alternative uses, especially when population growth rates are high. Forests are then cleared as a result of demand for land, not timber. In fact, trees might have a negative stumpage value and might sim- ply be burned to lay bare the soil rather than be used as an industrial raw material. On sites that are not converted to alternative uses, harvest exceeds growth because on a unit-area basis the extracted volume of old-growth timber exceeds growth on the logged-over area. Thus when exploitation of old- growth forests begins, harvests necessarily exceed growth. High rates of forest conversion and great discrepancies between harvest and growth do not imply, however, that forest exploitation is on an unsustainable trajectory. The rate of harvest is determined by balancing the opportunity costs of not harvesting the stand against the gains of cutting it now. The opportunity costs of not harvesting a stand in the current period include agricultural demand for land and industrial demand for capital. These costs weigh against the returns, due to growth and stumpage price increases, from holding the stand for harvest in a future period. Initially, the balance favors reducing the inventory. The opportunity costs are high because cleared land is scarce and interest rates are high (as a result of the scarcity of capital in the economy). The value of holding the stand for the future is low because of the lack of net growth in old-growth forests (gross photosynthesis is just balanced by respiratory losses and mortality). Old-growth timber represents natural capital that is not earning a biological return but can be converted into financial capital for fueling economic growth. As the inventory of old growth declines, however, timber becomes scarcer and, assuming that an efficient timber market exists, stumpage prices (or implicit timber values) rise. 36 Rising stumpage prices depress timber de- 36 As discussed later, even if an efficient timber market does not exist, a conces- sion holder with secure, transferable tenure will internalize rising implicit timber valuations and act as though an efficient market existed. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 123 mand and reduce the conversion of forest land to alternative uses as man- agement of second-growth timber becomes economic. Ultimately, stumpage prices and forest area stabilize. This transition from an old-growth to a second-growth timber economy is illustrated in exhibits 6-4 through 6-6. These economic adjustments cause the country's original endowment of forested land to be split into four distinct categories: 1. Some land is converted to nonforest uses, typically agriculture, that earn higher rents. The remaining land is kept under forest cover. 2. The most productive and accessible forest land will be man- aged intensively for second-growth timber. 3. Less-attractive sites will be logged for their original old-growth timber, but they will not be managed actively: Although invest- ments in timber extraction pay off on these sites, investments in timber management do not. These sites will be "pulse-harvested," that is, they will be left to regenerate naturally and will be har- vested again when they have grown a second crop of timber. 4. Finally, the highest-cost old growth will remain unharvested and will become economic wilderness. The country's long-run timber base consists of forests in the second and third categories. Although timber flows from the second category might be expected to be more or less even over time, pulse extraction from the third category could intermittently cause the country's annual harvest to exceed annual growth. At the time, this pattern of harvest might appear to be unsustainable, but from a long-run perspective it is not only sustainable (instantaneous harvest from pulse-harvested forests equals cumulative growth) but economically optimal. (Exhibit 6-4 assumes that no pulse-har- vesting occurs.) Old-growth timber is liquidated through two processes: land conver- sion and logging in permanent forests. When exploitation of old-growth timber begins, harvests rise as the agricultural frontier is opened and infra- structure is built to access the forest (the dotted line for time less than zero). As land conversion slows and logging occurs increasingly in second-growth forests, harvests decline. Stumpage prices (shown here to be negative ini- tially) rise at a decreasing rate and stabilize once the transition is complete. At this point, harvest equals growth. In an open economy, harvests may decline more rapidly if importing wood or wood products is less expensive than exploiting domestic forests. This is shown by the dotted line in the positive quadrant. 124 MANAGING THE WORLD'S FORESTS EXHIBIT 6-4. Transition from an Old-Growth to a Second-Growth Timber Economy , , ,, ,, I I ~ Harvest I <:- E ' 0 ' ' :0 ' (.) ::J ' ' ' ' ' ' ' ' ' ' ' ~ 0 Time Stumpage 0 Time FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 125 EXHIBIT 6-5. Timber Processing in an Open Economy Processing Harvest Time II 111 REGION Net Exports EXHIBIT 6-6. Index of Roundwood Production INDEX (1960 = 1) 5 4 /' ------------- ' I ,..-..., / -- ....... _ ....... ' I --'\---··r I \ I \/ / I " 3 ----------/--- / ,.. ./ 2 ..... ./.. .. / ,,."" .1 - ... :::--.:.··:_ :.:·-.:.--... ...... - - ... ;.;.--~--;,;,··::.··.:.:--... ; ... --... .,,. - ' :.:··------------ -- ----------- 0 1960 1965 1970 1975 1980 1985 - - Peninsular Malaysia - - - Ghana -Chile 126 MANAGING THE WORLD'S FORESTS In a closed economy, output of processed products will more or less parallel the trajectory of harvests shown in exhibit 6-4. The trajectory will be somewhat less steep if recovery rates are not fixed, but processing will inevitably decline. In an open economy, shown in exhibit 6-5, the relationship between harvesting and processing depends on trade as well. Quantity of processed products is depicted in log equivalents. In region r; logs are exported and no domestic processing occurs. In region II, processing begins, but some logs are still exported. In region III, the output of processed products has risen to the point at which it extends the sustainable level of harvest, and logs are imported. The country shifts from being a log exporter to a log importer as it gains comparative advantage in processing. In broad outline, this description of dynamic adjustments in the timber stock is consistent with development of the forest sector in the United States (Clawson 1979; Sedjo 1990; Johnson and Libecap 1980). The area of produc- tive forest land at the time of European settlement has declined by about half; in the late 1800s, the rate of deforestation iin the United States exceeded the current rate in Latin America. Harvests exceeded growth until the middle of this century, when the timber inventory stabilized and began to increase. The annual rate of increase in stumpage price has declined from 4.6 percent before World War II to 3.1 percent after, and it is forecast to decrease to 1.9 to 2.5 percent in coming decades (Binkley and Vincent 1988). EXHIBIT 6-7. Industrial Roundwood Consumption Share of Production, 1960-1987 SHARE 1.2 ---·- ~--~·::.:.;·;;;o.;-·------- / ---- ... 0.8 ' ---- ,, ___ ""::.. ,- ... ... ,,.., ······/··· ... ,,,,,.. / ------,----'·;;.·--- '' ""----· , , ____ _,______ ' ... - .........-----------'------- " \ 0.6 , ' ''' I 0.4 ' 0.2 1960 1965 1970 1975 1980 1985 - - Peninsular Malaysia - - - Ghana -Chile FOREST-BASED INDUSTR/ALIZA TION: A DYNAMIC PERSPECTIVE 127 Resource Dynamics with Other Forms of Tenure We have assumed that private ownership prevails and that property rights are fully specified. In fact, in most developing countries the forests are owned by the state and allocated for use as temporary concessions. Transferring ownership to the private sector (whether local communi- ties, private individuals, or corporations) might not be practical. Are there means of emulating an efficient market outcome while maintaining public ownership? The answer is yes, as long as the tenure arrangement mimics the incen- tives created by fully specified property rights. Efficient tenure arrange- ments will incorporate four fundamental property rights: exclusivity, trans- ferability, divisibility, and enforceability (see, for example, Hartwick and Olewiler [1986)). We discuss only the first three, as enforceability in the present context pertains to the assurance that the first three rights hold. Enforceabil- ity is perhaps the most important right of all, but its importance is obvious. Exclusivity is generally what is meant by "security" of c0 ncession rights. Concessionaires must be assured that they will reap the returns to (1) hold- ing onto timber until it is economically mature (taking into account both biological growth and stumpage price increases) and (2) making invest- ments in forest management, either direct investments such as planting and thinning, or indirect investments such as logging in such a way as not to impair forest regeneration. Exclusivity does not necessarily imply either private ownership or long-term concession contracts. Short-term contracts that include clearly defined and politically well-accepted criteria for renewal can mimic exclusive rights (e.g., making renewal automatic if concession- aires abide by logging regulations and do not relog second-growth sites too soon). Similarly, provisions for compensation for forest improvements achieve the same effect (e.g., reimbursing concessionaires who leave the forest in better condition an amount greater than their security deposit). Tenure se- curity in the sense of exclusivity need be no more (and no less) than the right to collect market value for the concession if it should be expropriated. Transferability, combined with exclusivity, makes the initial allocation of concessions irrelevant from the standpoint of economic efficiency. The argument is analogous to that for the well-known Nobel Prize-winning Coase Theorem (e.g., see Hartwick and Olewiler [1986)). Transferability allows concessions_ to "migrate" to concessionaires with the lowest discount rates, for whom long-term forestry investments have the highest net present value. Hence, these secondary concessionaires can "bid away" concessions from initial recipients who are myopic (perhaps because of political motiva- tions). Allowing concessions to be sold at any point makes forestry invest- ments more liquid and thus more appealing to risk-averse investors, by giving investors the option to collect their return whenever they want (as- suming, of course, that there is a competitive market for concessions). The incentives for efficient forest use created by transferability are obviously reduced by "use it or lose it" provisions and other common features of concession contracts. 128 MANAGING THE WORLD'S FORESTS Although we stress here the efficiency benefits of transferability, we do recognize that the initial allocation of concessions might have important equity ramifications. Divisibility is a subsidiary condition to transferability, but an important one. It allows transfers to take place at their most efficient size. It allows concessions ultimately to have a size consistent with returns to scale in harvesting and management. If returns to scale in harvesting and manage- ment differ, divisibility, combined with transferability, allows concessions to be configured to different sizes for harvesting and managing operations. If these rights are enforced, the holder of a concession will face the correct implicit value of timber even if ownership remains vested in the state and even if formal markets for stumpage do not exist. Resource tenure is thus revealed to be a more fundamental condition for efficient forest use than the state's rent capture: Although an efficient outcome is possible if resource tenure is nonattentuated and rent capture is low, the converse is not true. Rent capture is more an issue of equity than efficiency. 37 Although resource tenure is a more critical and fundamental issue than rent capture, the methods used to capture rent can themselves have per- verse efficiency effects. For example, assessing royalties on extracted logs rather than on standing trees may induce high grading (Gillis 1980, 1988b; Vincent 1990b). Assessing timber fees on timber volume rather than on concessionaire income (profits) or concession area may distort forest man- agement decisions. However, we suspect that these problems are likely to be minor compared with the efficiency issues related to resource tenure. Industry Dynamics How do adjustments in the timber stock affect development of wood- processing capacity? Obviously, if recovery rates in converting timber to processed products are fixed, the country's output of processed products will exactly parallel its timber harvest (remember that we are considering a closed economy). Thus, output of processed products will decline as the transition is made from old growth to second growth. This decline will be mitigated if recovery rates are not fixed, so that mills can alter their processing technology in response to changes in relative factor prices. Assuming that delivered log prices rise along with stumpage prices (that is, technological improvements do not cause logging costs to decline more rapidly than stumpage prices increase38), recovery rates will 37 Discussions of tropical timber concessions have tended to emphasize rent cap- ture rather than resource tenure (Page, Pearson, and Leland 1976, Repetto and Gillis 1988, Vincent 1990b). Resource tenure has generally been discussed superficially, for example, by recommending simply that concessions be made long-term. 38 It is economically optimal to log first the sites that are most accessible and least difficult. Hence, in the absence of technical change, we expect logging costs to rise over time as logging moves to more remote sites and areas with more difficult terrain. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 129 increase to conserve wood. Despite the scope for making more product from less wood, it appears inevitable that at some point production of pri- mary products will decline and the country will face the need to disinvest in primary wood-processing capacity. Because the economy is closed, decreases in domestic production must be matched exactly by decreases in domestic consumption. Why does con- . sumption of wood products decline? Increases in log prices cause increases in product prices, although at a lower rate, because of factor substitution and technical change. Higher prices encourage substitution in end-use mar- kets away from conventional wood-based products. For example, since 1900 the annual consumption of softwood lumber in the United States has re- mained roughly constant at 40 to 50 billion board feet despite a sevenfold increase in economic activity and a more than doubling of population. Per capita consumption has fallen from almost 600 board feet per capita to less than 200. Macroeconomic Dynamics What do these adjustments in the forest sector imply for important macroeconomic variables such as capital, labor, government revenue, and the relative importance of the forest sector? Reduction in the timber stock increases the supply of fungible capital. With the increase in capital comes a reduction in the price of capital-the interest rate-relative to the prices of labor and timber. The reduction in the relative price of capital is accentu- ated by rising wage rates (as the economy develops, per capita income rises) and, as noted earlier, rising timber prices. The lowering of interest rates has two feedback effects on the forest sector: The feasibility of long-term investments in forest management is enhanced, and logging and processing industries substitute capital for labor and for timber. This substitution causes employment per unit of output to decrease (conversely, labor productivity rises, and with it the standard of living for those employed). Employment in the forest sector will be main- tained only if process deepening occurs, that is, if downstream industries develop. Whether these industries develop depends on the returns to labor and capital in other sectors of the economy: downstream wood processing may not be an efficient use of a country's nonwood factors of production. Government revenue from the forest sector might decline. When forests are publicly owned and forest revenue systems fully capture stumpage value, revenue will decrease if the rate of decrease in harvest is greater than the rate of increase in stumpage value. When forests are privately owned or the government does not capture all the stumpage value from public forests, revenue from timber harvests will surely decline if ad rem royalties are used and might decline under ad valorem systems as well. To some extent, decreases in revenue might be offset by taxing the income generated by processing industries. In this scenario-the autonomous development of a forest sector in the absence of policy intervention-the relative, and perhaps even the absolute, 130 MANAGING THE WORLD'S FORESTS economic importance of the forest sector will ultimately decline. As land is shifted from trees to crops, the agricultural share of gross domestic product (GDP) will rise relative to forestry's share. The share of nonwood manufac- turing will increase in response to capital freed up by liquidation of old- growth timber stocks. Forest-Sector Developments in Open Economy Let us now assume that the country has entered the world market be- cause world timber prices have risen sufficiently to create international de- mand for its timber, either directly through log exports or indirectly through exports of processed products. In other words, the country is the newest producer on the world margin. Opening the economy to international trade has at least four potentially profound impacts on development of the forest sector: 1. If other countries have comparative advantage in agriculture, less of the country's forest land will be converted to crops. Con- versely, if opportunities for exporting crops are made available, more forest land will be converted. 2. The addition of international demand will reduce the likeli- hood of negative stumpage values. Timber will be rendered a saleable commodity instead of a costly obstacle to land clearing. In the early phases of exploitation, this situation will lead to more rapid land conversion. 3. If the economy is open not only to :international commodity markets but to international financial markets as well, and if interest rates are lower in the international market than in the domestic market, investments in forest management and wood processing will begin earlier. Of course, these investments will be postponed if higher international interest rates offer greater returns to capital created from cashing in old-growth forests. 4. Trade allows the country to use its factor endowments more efficiently, through (1) taking advantage of economies of scale that might not be achieved if production is limited by a small domestic market and (2) specializing its production according to comparative advantage. If the country does not initially have com- parative advantage in wood processing, it will export most of its log harvest and direct most of its capital and labor to nonwood manufacturing sectors. It might even import processed products to meet domestic demands. Imports will force its processing in- dustry to minimize costs and to innovate to remain competitive. FOREST-BASED INDUSTRIALIZA T/ON: A DYNAMIC PERSPECTIVE 131 Over time, the country might be expected to develop comparative ad- vantage in processing if its capital costs decrease, its labor force becomes more productive, and rising logging costs make location of mills close to the resource more advantageous. Import substitution, followed by export sub- stitution (exporting processed products rather than logs), might then occur as a natural sequence. By following comparative advantage, the country might be able to avoid the inevitable overcapacity in wood processing that occurs in a closed economy: capacity could rise gradually to match sus- tained yields from second-growth forests. These points hold regardless of whether the country's timber resources are small or large relative to global resources. A key difference between the small and large cases is that if the country's forest sector is small-that is, if its level of production has an insignificant effect on international prices- pulse-harvesting is more likely to occur. The best harvest decision will be to harvest instantaneously (or as close to instantaneously as is possible) all the country's forests that are economically exploitable at the current, exogenous world prices. If the country lacks sufficient comparative advantage to rely on imported timber after its domestic timber supplies have been exhausted, the short period of harvest will inhibit investments in wood processing. As the size of the country's timber resources rises relative to the world's, the dynamics of its forest sector resemble more and more the dynamics of the closed economy case. Stumpage prices will be determined endogenously, through the interaction of domestic and foreign supply and demand. They will gradually dampen the (noninstantaneous) drawdown of old-growth stocks and promote transition to a second-growth timber supply. The closed- economy scenario thus provides a useful paradigm for understanding the essential features of the open-economy case, as long as the four points just listed are kept in mind and the country is not too small. An open economy may be influenced by trade sanctions. For the pur- poses of forest policy, boycotts on tropical timber are among the potentially most significant of these (see Hamilton 1990; Vincent 1990a). Pressured by environmental organizations, governments of several European countries have considered or implemented bans on the import of tropical timber. The putative idea of these boycotts is to stop deforestation. Boycotts on timber imports are unlikely to achieve this objective directly. Because boycotts re- duce the value of tropical timber, they may have several perverse side ef- fects that will actually increase the rate of deforestation. In moist tropical forests, land-use conversion, not commercial logging, is the major cause of deforestation. Frequently the timber harvested is sim- ply burned or left to rot on the site rather than being processed for local use or international trade. Unless the timber enters trade, a boycott can obvi- ously have no impact on the economics of land-use conversion. Even if the timber removed is sold into international trade, the increase in land conver- sion costs induced by a boycott is likely to be small compared with the differences in the value of land for timber production and for agriculture 132 MANAGING THE WORLD'S FORESTS (Perez 1990). A boycott may affect those lands where the economics of con- version are marginal, but the total area affected is likely to be small. Conversely, .boycotts will reduce the return from sustainable forestry practices. Governments and landowners will be less likely to protect or manage forests if their economic value has been reduced by timber boycotts. Indirectly, boycotts may actually increase the rate of land-use conversion. Finally, by reducing the price of exported timber, boycotts exacerbate problems related to fiscal deficits, balance of payments, and debt service. Facing a lower unit price, countries may need to harvest and export a greater volume simply to offset declines in revenue and hard-currency receipts. The problem will be particularly significant for developing countries in which timber receipts make up a large fraction of public revenues, and those that have large debt burdens. Environmental Considerations By generating the increases in stumpage prices that enable forests to outcompete alternative land uses, industrial use of forests can play an im- portant role in stabilizing the area of forests and stimulating investments in timber management. Because there is more to forests than just timber, how- ever, relying on timber prices alone to determine the ultimate composition of a country's forest estate will lead to a suboptimal outcome. Most of the externalities associated with natural forests appear to be positive: watershed protection, habitat preservation, micro- and macro-climatic amelioration, aesthetic values, and so forth. To the extent that market-based decisions ignore these values, conversion of forests to alternative uses such as agricul- ture is likely to proceed too far. Not only will the forest estate be too small, but its composition will be suboptimal. The area of remnant old-growth forest will be too small be- cause markets fail to account for its nontimber values. As a result, there may be an irreversible welfare loss: Forest-based industrialization can con- tribute to the overall economic development that ultimately raises the value human populations attach to old-growth forests, but it may destroy the direct source of this value before the future generations that hold these high values have been born. As a country develops, its people's income, educa- tion, and leisure time increase-and so do the aesthetic and recreation val- ues of its forests. Future generations in developing countries will probably place more value on pristine forest environments than do current genera- tions, but they are not present to make their case when the apparently irreversible (relative to the length of human generations) decision to harvest old-growth forests is made .. In the tropics, this market failure is accentuated by the fact that the old-growth forests that are richest in an increasingly recognized nontimber value-biological diversity-tend to be those in the lowlands, which are generally the first to be harvested and converted. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 133 If timber management is complementary to nontimber values, the area of forest that is actively managed will be too small, and the area that is pulse-harvested will be too large. Failure to account for nontimber forest products used by local populations (see Peters, Gentry, and Mendelsohn 1989) will lead to an overemphasis upon timber use in both types of forests. Finally, forests apparently play an important role in the global carbon cycle (Detwiler and Hall 1988; Tans, Fung, and Takahashi 1990). In the short run, harvesting old-growth timber, even if the harvested areas are con- verted to faster-growing second-growth stands, will release carbon dioxide into the atmosphere (Harmon, Ferrell, and Franklin 1990; Dunsworth, Northway, and Hawkins 1990). In this case, "short run" refers to the period of time needed for the conversion to take place, and therefore might last for 30 to 60 years. Conversely, new forest plantations can serve as carbon sinks if they are established on areas that currently support only small carbon pools (Sedjo 1989; U.S. Environmental Protection Agency 1989). Because these concerns are not incorporated into private decisions, the rate of deple- tion of old-growth timber will be too rapid, the regeneration of second- growth forests will be too slow, and the area of new plantations will be too small. REFERENCES Anderson, F.J. 1976. "Control Theory and the Optimal Timber Rotation." Forest Science 22:242-46. Asabere, P.K. 1987. "Attempts at Sustained Yield Management in the Tropical High Forests of Ghana." In Natural Management of Tropical Moist Forests, eds. F. Mergen, and J.R. Vincent. New Haven, Conn.: School of Forestry and Environmental Studies, Yale University. Auty, R.M. 1987. "State Enterprise Performance in Resource-Based Industry." Development Discus- sion Paper No. 244. Cambridge, Mass.: Harvard Institute for International Development. Berck, P. 1976. "Natural Resources in a Competitive Economy." Unpublished Ph.D. dissertation. Deparhnent of Economics, Massachusetts Institute of Technology, Cambridge, Mass. Binkley, C.S., and J.R. Vincent, 1988. "Timber Prices in the U.S. South: Past Trends and Outlook for the Future." Southern Journal of Applied Forestry 12:15-18. Brazee, R., and R. Mendelsohn. 1990. "A Dynamic Model of Timber Markets." Forest Science 36:225- 64. Cardellichio, P.A., Y.C. Youn, D.M. Adams, R.W. Joo, andJ.T. Chmelik. 1989. "A Preliminary Analysis of Timber and Timber Products Production, Consumption, Trade, and Prices in the Pacific Rim until 2000." CINTRAFOR Working Paper No. 22. Seattle: College of Forest Resources, Univer~ sity of Washington. Clark, C.W., 1976. Mathematical Bioeconomics. New York: Wiley InterScience. Clawson, M. 1979. "Forests in the Long Sweep of American History." Science 204:1168-74. Corbo, V., and J. de Melo. 1985. "Introduction." In Scrambling for Survival: How Firms Adjusted to the Recent Reforms in Argentina, Chile, and Uruguay, eds. V. Corbo, and J. de Melo. World Bank Staff Working Paper No. 764. Washington, D.C.: World Bank. Corbo, V., and J.M. Sanchez. 1985. "Adjustments by Industrial Firms in Chile During 1974-82." In Scrambling for Survival: How Firms Adjusted to the Recent Reforms in Argentina, Chile, and Uruguay, eds. V. Corbo, and J. de Melo. World Bank Staff Working Paper No. 764. Washington, D.C.: World Bank. Detwiler, R.P., and C.A.S. Hall. 1988. "Tropical Forests and the Global Carbon Cycle." Science 239:42-47. 134 MANAGING THE WORLD'S FORESTS Douglas, J.J. 1983. A Re-appraisal of Forestry Development in Developing Countries. The Hague: Martinus Nijhoff/Dr W. Junk Publishers. Dunsworth, G., S. Northway, and C. Hawkins. 1990. "A Carbon Dioxide Budget for MacMilan- Bloedel." Paper presented at the Faculty of Forestry, University of British Columbia, Vancouver, November 7, 1990. Ehui, S.K, T.W. Hertel, and P.V. Preckel. 1990. "Forest Resource Depletion, Soil Dynamics, and Agricultural Productivity in the Tropics." journal of Environmental Economics and Management JB,136-54. Ewing, A.J., and R. Chalk. 1988. "The Forest Industries Sector: An Operational Strategy for Devel- oping Countries." World Bank Technical Paper No. 83, Industry and Energy Series. Washington D.C.: World Bank. Fitzgerald, B., 1986. "An Analysis of Indonesian Trade Policies: Countertrade, Downstream Process- ing, Import Restrictions, and the Deletion Program." CPD Discussion Paper No. 1986-22. Wash- ington, D.C.: World Bank. Food and Agricultural Organization. 1981. Forest Resources of Tropical Africa. Part II: Country Briefs. UN32/6.1301-7804, Technical Report No. 2. FAO/UNEP, Rome. Forestry Department, Peninsular Malaysia. 1986. Forest Statistics, Peninsular Malaysia: 1979-1985. Forestry Department Headquarters, Kuala Lumpur. Francescon, A., A. Nagy, and G. Kornai. 1983. "An Historical Analysis of International Trade in Forest Products." Working Paper WP-83-90. Laxenburg, Austria: International Institute for Applied Systems Analysis. Gillis, M. 1980. "Fiscal and Financial Issues in Tropical Hardwood Concessions." Development Discussion Paper No. 110. Cambridge, Mass.: Harvard Institute for International Development. _ _ _. 1988a. "Indonesia: Public Policies, Resource Management, and the Tropical Forest." In Public Policies and the Misuse of Forest Resources, eds. R. Repetto, and M. Gillis. Cambridge, Eng.: Cambridge University Press. _ _ _. 1988b. "Malaysia: Public Policies and the Tropical Forest". In Public Policies and the Misuse of Forest Resources, eds. R. Repetto, and M. Gillis. Cambridge, Eng.: Cambridge University Press. _ _ _. 1988c. "West Africa: Resource Management and the Tropical Forest". In Public Policies and the Misuse of Forest Resources, eds. R. Repetto, and M. Gillis. Cambridge, Eng.: Cambridge Uni- versity Press. Gillis, M., D. H. Perkins, M. Roemer, and D.R. Snodgrass. 1987. Economics of Development, 2d ed. Ne"\v York: W.W. Norton & Co. Grut, M. 1989. "Economics of Managing the African Rainforest." Paper prepared for the Thirteenth Commonwealth Forestry Conferenc_e. Hamilton, L. 1990. "Boycotts of Tropical Timber Products Will Not Stop Deforestation." Washing- ton, D.C.: World Bank. Harmon, M.E., W.K. Ferrell, and J.F. Franklin. 1990. "Effects on Carbon Storage of Conversion of Old-Growth Forests to Young Forests." Science 247:699-702. Hartwick, J.M., and N.D. Olewiler. 1986. The Economics of Natural Resource Use. New York: Harper &Row. Heaps. T. 1984. "The Forestry Maximum Principle." Journal of Economic Dynamics and Control 7:131- 51. Heaps, T., and P. Neher. 1979. "The Economics of Forestry When the Rate of Harvest Is Con- strained." Journal of Environmental Economics and Management 6:297-316. Hunter, L.A.J. 1987. "The Chilean Forestry Sector." Infonnation Paper No. 2. Wellington: New Zealand Forestry Council. Jelvez, A., K.A. Blatner, and C. McKetta. 1988. "A Profile of the Chilean Forestry Sector." CINTRAFOR Working Paper No. 17 (draft). Seattle: College of Forest Resources, University of Washington. Jelvez, A., K.A. Blatner, R.L. Gavett, and P.H. Steinhagen. 1989a. "Chile's Evolving Forest Products Industry. Part 1: Its Role in International Markets." Forest Products Journal 39(10):63-67. Jelvez A., K.A. Blatner, and RL. Gavett, 1989b. "Chile's Evolving Forest Products Industry. Part 2: Investments in an Expanding Industry." Forest Products Journal 39(11/12):76-78. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE 135 Johansson, P-0, and K-G Lofgren. 1985. The Economics- of Forestry and Natural Resources. London: Basil Blackwell. Johnson, R.N., and G.D. Libecap. 1980. "Efficient Markets and Great Lakes Timber: A Conservation Issue Reexamined." Explorations in Economic History 17:372-85. Jungenfelt, K. 1973. "Maloch medel i skogspolitiken [Models for Macroeconomic Planning of Forest Policy]." Swedish Parliament Committee Report SOU 1973:14, Stockholm. Kallio, M., D.P Dykstra, and C.S. Binkley, eds. 1987. The Global Forest Sector: An Analytical Perspec- tive. Chichester, Eng.: John Wiley & Sons. Kemp, M., and N.V. Long. 1983. "On the Economics of Forests." International Economics Review 24:113-27. Kumar, R. 1986. The Forest Resources of Malaysia: Their Economics and Development. Singapore: Oxford University Press. Lewis, S.R., Jr. 1989. "Primary Exporting Countries." Handbook of Development Economics, eds. H. Chenery and T.N. Srinivasan. Amsterdam: North-Holland. Lindsay, H. 1989. "The Indonesian Log Export Ban: An Estimation of Foregone Export Earnings." Bulletin of Indonesian Economic Studies 25(2):111-23. Lyon, K., and R. Sedjo. 198~. "An Optimal Control Theory Model to Estimate the Regional Long- Term Supply of Timber." Forest Science 29:798-812. MIDA (Malaysian Industrial Development Authority)/UNIDO (United Nations Industrial Develop- ment Authority). 1985. Medium and Long Term Industrial Master Plan, Malaysia: 1986-1995. Vol- ume II, Part 4: Wood-Based Industry. Kuala Lumpur: MIDA. Mohd. Shahwahid Haji Othman. 1986. "Sources of Growth for the Timber-Based Manufacturing Industries of Peninsular Malaysia During the Period 1961-1981." Malaysian Forester 49(3):208- 22. Mitra, T., and H. Wan. 1985. "Some Theoretical Results on the Economics of Forestry." Review of Economic Studies 52:263-82. Page, J.M., Jr. 1974. "The Timber Industry and Ghanaian Development. In Commodity Exports and African Economic Development, eds. S.R. Pearson, and J. Cownie. Lexington, Mass.: Lexington Books (D.C. Heath and Co.). _ _ _ . 1976. "The Social Efficiency of the Timber Industries in Ghana." In Using Shadow Prices, eds. l.M.D. Little, and M.Fg. Scott. London: Heinemann Educational Books. Page, J.M., Jr., S.R. Pearson, and H.E. Leland. 1976, "Capturing Economic Rent from Ghanaian Timber." Food Research Institute Studies 15(1):25-51. Panayotou, T. 1989. "Modeling Optimal Resource Depletion in Developing Countries: A· Joint In- vestment-Extraction Decision." Development Discussion Paper No. 286. Cambridge, Mass.: Harvard Institute for International Development. Perez, J.G. 1990. "Deforestation and Agricultural Subsidies: An Analysis of Brazilian Amazonia." Unpublished dissertation. Yale University School of Forestry and Environmental Studies, New Haven, Conn. Peters, C.M., A.H. Gentry, and R.O. Mendelsohn. 1989. "Valuation of an Amazonian Rainforest." Nature 339:655-56. Pezzey, J. 1989. "Economic Analysis of Sustainable Growth and Sustainable Development." Draft paper prepared for the Environment Department, World Bank. Repetto, R. 1987. "Creating Incentives for Sustainable Forest Development." Ambia 16(2-3):94--97. Repetto, R., and M. Gillis, eds. 1988. Public Policies and the Misuse of Forest Resources. Cambridge, Eng.: Cambridge University. Press. Roemer, M. 1979. "Resource-based Industrialization in the Developing Countries." Journal of Devel- opment Economics 6:163-202. _ _ _ . 1984. ''Ghana: Missed Opportunities." In World Economic Growth, ed. A.C. Harberger. San Francisco: Institute for Contemporary Studies Press. Santa-Cruz, Vicente. 1988. "Chilean Forest Sector: An Economic Evaluation of the D.L. 701 on Forestry Incentives." Unpublished manuscript. University of Minnesota. 136 MANAGING THE WORLD'S FORESTS Sedjo, R.A. 1987. "Forest Resources of the World: Forests in Transition." In The Global Forest Sector: An Analytical Perspective, eds. M. -Kallio, D.P. Dykstra, and C.S. Binkley. Chichester, Eng.: John Wiley & Sons. _ _ _. 1989. "Forests: A Tool to Moderate Global Warming." Environment 31:14-20. _ _ _. 1990. "The Nation's Forest Resources." Discussion Paper No. ENR90-07. Washington, D.C.: Resources for the Future. Solow, R. 1974. "The Economics of Resources or the Resources of Economics." American Economic Review 64:1-14. Sulaiman bin Haji Noordin. 1977. "A Method of Forest Revenue Assessment Based on Inventory Data." Malaysian Forester 40(3):144-59. Takeuchi, K. 1983. "Mechanical Processing of Tropical Hard wood in Developing Countries." In Case Studies on Industrial Processing of Primary Products. Vol. 1: Bauxite, Rubber, Tropical Hardwood. Washington, D.C.: World Bank. Tang Hon Tat. 1987. "Problems and Strategies for Regenerating Dipterocarp Forests in Malaysia." In Natural Management of Tropical Moist Forests, eds. F. Mergen, and J.R. Vincent. New Haven, Conn.: School of Forestry and Environmental Studies, Yale University. Tans, P.P., LY. Fung, and T. Takahashi. 1990. "Observational Constraints on the Global _Atmo- spheric Carbon Dioxide Budget." Science 247:1431-38. Tedder, P., J.S. Schmidt, and J. Gourley. 1980. "TREES: Timber Resources Economic Estimation System." Forest Research Laboratory, Research Laboratory, Research Bulletin No. 31. Oregon State University, Corvallis. Teo Peng Chong. 1966. "Revision of Royalty Rates." Malayan Forester 29(4):254-58. U.S. Environmental Protection Agency. 1989. "Policy Options for Stabilizing Global Change." Re- port to the U.S. Congress. Vincent, J.R. 1986. "Growth of the Forest Products Industry in Malaysia: 1961-1985." Malaysian Forester 49(3):223-40. _ _ _. 1988. "Malaysia: Key Player in International Trade." Journal of Forestry 86(12):32-35. _ _ _. 1990a. "Don't Boycott Tropical Timber." Journal of Forestry 88(4):56. _ _ _. 1990b. "Rent Capture and the Feasibility of Tropical Forest Management." Land Economics 66(2),212-22. Walker, J.L. 1971. "An Economic Model for Optimizing the Rate of Timber Harvesting." Unpub- lished Ph.D. dissertation. College of Forest Resources, ·university of Washington, Seattle. Wan, F.Y., and F.J. Anderson. 1983. "Optimal Forest Harvesting with Ordered Site Access." Studies in Applied Mathemetics 68:189-226. Wan, H.Y. 1966. "A Generalized Wicksellian Capital Model:An Application to Forestry." Eco- nomic Papers No. 1. Nankang, Taiwan: Institute of Economics, Academia Sinka. Westoby, J. 1962. "Forest Industries in the Attack on Underdevelopment." The State of Food and Agriculture. Rome: FAO. _ _ _. 1978. "Forest Industries for Socio-Economic Development." Commonwealth Forestry Review 58(2),107-16. _ _ _. 1985. "Forestry and Underdevelopment Revisited." Lecture given at the School of Forestry and Conservation, University of California, Berkeley, November 1985. Reprinted in The Purpose of Forests: Follies of Development, comp. J. Westoby. 1987. Oxford, Eng.: Basil Blackwell. World Bank. 1955. The Economic Development of Malaya. Baltimore: Johns Hopkins University Press. _ _ _. 1978. "Forestry Sector Policy Paper." Washington, D.C.: World Bank. _ _ _. 1980. Chile: An Economy in Transition. World Bank Country Study. Washington D.C.: Latin America and the Caribbean Regional Office, World Bank. _ _ _. 1986a. "Chile: Forest Industries Sub-Sector Study." Report No. 6380-CH. Washington, D.C.: Industry Department, World Bank. _ _ _. 1986b. "World Bank Financed Forestry Activity in the Decade 1977-86." Washington, D.C.: Agriculture and Rural Development Department, World Bank. FOREST-BASED INDUSTRIALIZATION: A DYNAMIC PERSPECTIVE - 137 _ _ _. 1987a. "Chile: Adjustment and Recovery." Report No. 6726-CH. Washington, D.C.: Latin America and the Caribbean Regional Office, World Bank. _ _ _. 1987b. "Ghana: Forestry Sector Review." Report No. 6817-GH. Washington, D.C.: Western Africa Region, World Bank. _ _ _. 1988. "Staff Appraisal Report-Ghana: Forest Resource Management Project." Report No. 7295-GH. Washington, D.C.: Agriculture Operations Division, Western Africa Department, World Bank _ _ _. 1989a. "Ghana: Structural Adjustment for Growth." Report No. 7515-GH. Washington, D.C.: Western Africa Region, World Bank. _ _ _. 1989b. Malaysia: Matching Risks and Rewards in a Mixed Economy. World Bank Country Study. Washington, D.C.: World Bank. Young, K., W.C.F. Bussink, and P. Hasan, eds. 1980. Malaysia: Growth and Equity in a Multiracial Society. Baltimore: Johns Hopkins University Press. 7 Forest Concession Management and Revenue Policies Malcolm Gillis Incentives for forest use are influenced by a variety of government policies and projects. Policies established by government agencies nominally responsible for oversight of national forest endowments or forest-based industrialization are often referred to as forestry policies. This chapter fo- cuses on how these forestry policies affect logging activities and forest- based industrialization in tropical nations. COMMONALITIES The extent and quality of tropical rain forest endowments vary greatly in Africa, Asia, and Latin America. Still, forest use in all three regions to some extent has six elements in common. 1. Accelerating rates of deforestation, with logging as a secondary or tertiary cause of deforestation. In all three regions, as well as in the tropics in general, commercial logging has not been the leading direct cause of deforestation. Worldwide, logging is thought to account for only 21 percent of total deforestation. Nonetheless, damages flowing from inappropriate forest incentive policies governing logging have been sizable: 12,000 square miles- roughly half the size of West Virginia-are deforested annually. In addition, the area degraded (45,000 square miles per year [Myers 1990]) by the type of selective logging typically practiced in the tropics is generally much higher than the area deforested by loggers. Moreover, logging roads in forests degraded by selective logging open up previously inaccessible forest areas 139 140 MANAGING THE WORLD'S FORESTS to shifting cultivators, the principal agents of deforestation. Thus the fires of shifting cultivators often merely administer the coup de grace to already logged-over forests, in a two-stage process of land-use transition that may account for as much as 80 percent of the global conversion rate of tropical rain forests (Myers 1989a). 2. Heavy emphasis on policies geared to extraction of benefits from the productive, rather than the protective, services provided by the forest. Incentives have been directed almost exclusively to exploitation of the productive resources of the forest; the value of protective services has been ignored, although many such services are manifestly economic in the long run. Moreover, incentive policies have been based strongly on the assumption that only two of the productive resources were of any economic value: the timber stands and the generally poor-quality agricultural land lying under the canopy. Consistently overlooked in the formulation of forest incentive policies is a third economic resource: the capacity of intact forests to supply a perpetual stream of -income from commercial and noncommercial nonwood forest products. Virtually all these products can be harvested without degrading the forest, and all but one (rattan) can be harvested without removal of woody stems from the forest. Commercial products that are already known number in the hundreds; they include nuts, oils, fibers, bush meat, cosmetic compounds, dyes, fruits, latex, ornamental plants, spices, and pharmaceutical substances (Jacobs 1988). 3. Declining foreign participation in timber extractive activities and processing activities. The notable reduction of the role of large foreign firms in the harvest and, more recently, in the processing of tropical timber has been particularly apparent in Southeast Asia. Large multinational wood-products enterprises dominated the timber sectors in t:he Philippines and Peninsular Malaysia throughout the 1950s and played a highly visible role in logging in East Malaysia and Indonesia until the early 1970s. Large multinational firms have not been a major factor in timber extrac- tion in Central America in recent years, because much of the timber cut annually is not harvested for commercial purposes. Only in Honduras is commercial timbering a major generator of employment and foreign ex- change (Leonard 1987). With two exceptions, large foreign logging enterprises have had no major presence in South America since at least 1960. Daniel Ludwig's Jari project, which was begun in 1967 on 1.2 million hectares, was sold at sizable losses in 1982 to Brazilian interests (Schmink 1988). And in Colombia, in 1974, Carton de Colombia, a large American-based wood-products enter- prise, was awarded a 30-year concession for 61,000 hectares near the Pacific coast (Cardenas and Pena 1990). In Africa, large foreign-based timber firms, once dominant in Ghanaian timber, virtually abandoned the sector in 1977 and drastically reduced their presence in Liberia after the early 1970s. In Ivory Coast, however, and to an FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 141 even greater extent in Gabon, foreign firms maintained a significant pres- ence through the late 1980s (Gillis 1988c). In the Pacific, the Papua New Guinea government has recently accused large foreign firms of transfer pricing manipulations in timber export (World Bank 1990), a common complaint in Indonesia in the 1970s. 4. Accumulation of property rights to the forest in the hands of central governments. The underlying system of forest property rights has a significant bearing on prospects for reform of forest incentive policies. In the Philippines, Ecuador, and Indonesia, these rights have been constitutionally bestowed on central governments. In Malaysia, property rights to natural forests are vested in each of the 14 states of the Federation. The forests of Africa are also predominantly publicly owned, either wholly by central governments, as in Ghana since 1973 or in other countries by various combinations of central, state, and provincial governments, or villages and communities (Grut, Egli, and Gray 1990). Nevertheless, a small part of the forests of Ivory Coast is owned by the private sector. Transfer of these property rights from the government to the private sector often involves environmentally anomalous activities. In Ecuador, destruction of natural vegetation is usually a prerequisite for formal tenure in colonized parcels. Settlers in the Brazilian Amazon obtain title in a forested parcel only by clearing part of it (Southgate and Whitaker 1989). In eastern Malaysia, any Sabah native may obtain title to forest land by clearing and working it. 5. A shift in concessions policies away from special contracts and toward greater reliance on general law. Until the end of the first decade after World War II, special contracts were widely used in both mineral and timber concession agreements. The largest of these in forest sectors were the Ford concession in 1926 and the Ludwig undertaking in Brazil. These contracts were tailored to a particular mining operation or forest project, containing special tax and royalty provisions governing each operation. In addition, the contracts specified most other regulations to control mining or timber operations, including output minimums or maximums, hiring of host country nationals, safety rules, and length of concession (variable from one agreement to another even within a single country). Some ad hoc contracts even specified the definition of gross income and the specific cost deductions allowed. By 1990 the overwhelming majority of timber operations were subject to provisions of general law. 6. Severe underpricing of· tropical timber by the owner, leading to serious wastage of the resource. Virtually all nations with tropical timber endowments underpriced the wood resources of the forest, at the same time that they have placed heavy emphasis on the productive as opposed to the protective services available from the forest. Underpricing of these resources, through outdated and highly defective systems of forest fees and taxes, has provided false signals of timber abundance, leading to severe waste, not only in 142 MANAGING THE WORLD'S FORESTS harvesting practices but in timber processing (Repetto and Gillis 1988; Gillis forthcoming). Moreover, underpricing of the resource has meant that the resource owner has failed to capture sizable proportions of timber rents. Instead, extractive firms have appropriated a large proportion of uncaptured rent. Worse yet, large amounts of potential rents have been destroyed through waste induced in part by underpricing. CURRENT CONCESSION POLICIES As much as 90 percent of industrial wood is harvested under conces- sion agreements (Schmithusen 1980), by which the government, as forest resource owner, formally grants a company exclusive rights to exploit the wood potential of the forest or to manage a specific forest parcel. The con- ditions of a concession agreement are clearly defined in some countries, less so in others. Absence of concession agreements in any given country does not by any means imply an absence of commercial wood harvests. For ex- ample, when Ecuador reportedly imposed a ban on new logging conces- sions in 1982, agricultural colonists sharply increased harvests,1 switching their land-clearing techniques to an emphasis on harvesting and marketing of commercial species. 2 Attempts to curtail deforestation arising from log- ging might well be focused on efforts to improve the concession system, rather than on more drastic measures such as abolishing concessions or imposing logging bans. Large-scale use of tropical timber resources began nearly four decades ago amid virtually complete ignorance of the ecology of the tropical forest. Now people have come to appreciate the complexity of this ecology, but otherwise little of the ignorance about it has been dispel .d. And even de- cades of field experience and careful research do not always lead to pre- scriptions for reform that protect both economic and ecological values of the forest. The evidence that has become available indicates that certain changes in concession policies (over and above reform of generally applicable forest use fees discussed later) represent moves in the right direction, even if they fail to deal adequately with all issues in all settings. Great care is required in generalizing about reform across diverse tropical timber endowments: Actions that are economically and ecologically appropriate for the forests of Latin America will rarely be as suitable for the forests of Asia and Africa; and some prescriptions for reform in Southeast Asia may be inadvisable or 1 Imports of chain saws were reported to have tripled in the three years after 1982 (Southgate and Whitaker 1989). 2 Ecuador apparently distinguishes between "licenses" and concessions to harvest timber. By 1989, apparently two-thirds of the Amazonian forests were designated for commercial timbering (Winterbottom 1990). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 143 unenforceable in Africa or in other parts of Southeast Asia. Indeed, the reforms examined here are much more likely to be appropriate for the primary forests of Southeast Asia, the region that has dominated world trade in tropical timber for two decades. Even so, before they are imple- mented for all forests in that region, these reforms may require significant modification, particularly in Papua New Guinea, where forests are more heterogeneous and forest property rights less well established than in Indo- nesia or Malaysia. Award, Assignment, and Revocation of Concessions Tropical nations face three basic alternatives in selecting a process of awarding concessions: administrative discretion, competitive auction, or some combination of the two. Administrative Discretion The most common method of award of concession rights has been that of administrative discretion, involving negotiations between government officials and applicants. Prospective concessionaires apply for logging rights on particular parcels in areas where logging is permitted. Eligibility may vary with the country, nationality of applicant, size of firm, commitment to domestic processing of timber, and other factors. Increasingly, eligibility has been restricted either to nationals, as in Ghana since 1973 and in Indonesia since 1974, or (as in most cases) to enterprises with majority participation by nationals. Gabon remains the most notable example of countries that allow concessions to be awarded to firms with 100 percent of equity held by foreigners. Following negotiations, concessions are allocated among eligible con- tenders by administrative decision; guidelines for these decisions are rarely published. The pattern of awards from country to country has sometimes shown evidence of favoritism to particular political groups, and, on occa- sion in Southeast Asia and Ecuador, to military interests (Winterbottom 1990, 54). In any case, reliance on administrative discretion in award of concessions allows ample scope for collusion and financial irregularities (Grut et al. 1990, 41-42). Competitive Auction Auction procedures may apply not only to initial award of concessions but also to reallocation of concessions returned, expired, and revoked. Com- petitive auction of concession rights, instead of administrative award, has been advocated as a method of avoiding the financial irregularities and loss of revenue for governments as resource owners. In addition, proponents of competitive auctions argue that, by enabling governments to capture more 144 MANAGING THE WORLD'S FORESTS rent for the owner, auctions would make wood more expensive, thereby encouraging better use both in logging and processing (Grut et al. 1990, 59). Auctions offer substantial benefits, but many of the benefits may be secured without relying only on auctions. Indeed, auctions are feasible only when there is an inventory of timber resources adequate to permit parcels to be auctioned, a condition not met for most parcels in most tropical nations. 3 Auctions are feasible only when the auction market is sufficiently com- petitive, when there is no collusion, when the costs of gathering and evalu- ating information on stand quality are sufficiently low, and when the auc- tioning agency is sufficiently well staffed. These conditions prevail in few nations with sizable tropical forests. Auctions for granting concessions have been used in a few countries to date, most notably Venezuela and Malaysia, with successful, if not fully documented, results reported in both cases (Gillis and Repetto 1988). Schmithusen (1977) reports that auction premiums in Sarawak were at one point as high as 20 percent of total stumpage revenues. Auctions, through sealed bids (closed tenders), have also been used in many of the 11 states of Peninsular Malaysia. Composite Methods Administrative discretion and auctions are not mutually exclusive ways of granting concessions. Where the information on forest quality is moder- ately good but still far from complete, features of an auction system can be combined with limited administrative discretion in awards. Such composite methods have been most frequently used in the award of exploration rights for oil and gas. For example, Indonesia's production- sharing contracts have long required "signature bonuses" for oil and gas tracts. The company offering to pay the highest signature bonus for a par- ticular tract is awarded exploration and extraction rights. For particularly promising tracts, the bonuses have been sizable, particularly when world oil prices were high. Between 1979 and 1981, several foreign oil and gas firms paid signature bonuses as high as US$75 million to Indonesia. Composite systems of concession awards have much to recommend them for tropical timber, provided enough bidders are present and the scope for collusion by bidders is limited, information on timber stand quality is more than rudimentary, and the quality of on-the-ground supervision by agents of the owner (the government) is sufficiently high. A signature bonus 3 The Peninsular Malaysia Forestry Department already carries out prefelling in- ventories one to two years before harvest. Inventories record stocking (number of trees) per diameter class (12 classes in 5-to-15-cm increments) and by species group. Estimates of standing timber volume are created as part of this process (Vincent 1990b). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 145 becomes the decisive factor in the administrative award of concessions among otherwise similarly situated claimants, thereby reducing the scope for rent- seeking activities. The bonus also serves as one element of a wider strategy of capture of timber rents. But by itself, no bonus can capture all, or even most, of the available rent, or dispense with all arbitrary factors in award concessions. Composite methods of award have worked well in other natural re- source agreements and could be implemented for parcels that have been recently included in careful, systematic inventory of forest resources. In such cases, use rights would be granted to the firm willing to pay the high- est signature bonus, in addition to incurring the other fiscal obligations expected of all concessionaires. Transferability of Concession Rights A limited number of tropical countries have allowed the transfer of concession rights. Allowing concessionaires to sell concession rights, with or without government approval, provides the concessionaires with an in- centive for sustainable use of the resource even if· they do not intend to request renewal. Steps taken to maintain the long-term productivity of the concession, such as avoiding early reentry on logged-over stands or adopt- ing felling methods less injurious to the residual stand, would then redound to the benefit of the concessionaire. The capitalized value of the resources thereby protected will (assuming an active market for transferable rights) tend to be capitalized in the transfer price. And although making rights transferable may, in the absence of auctions, increase the incidence of initial awards to political cronies or favored military groups, transferability would at least make it possible that the concession rights end up in the hands of those for whom exploitation rights are most valuable. At a minimum, transfers should be required to be formally recorded with the government, even when government approval of transfers is not required. This recording requirement is important not only to assure effec- tive monitoring of harvesting activity by concessionaires, but also to safe- guard th.eir forest and nonforest fiscal obligations. Renewal and Revocation of Concessions Many tropical nations have provided for the possibility of renewal of timed concession agreements, but most renewal provisions are perfunctory or vague. As in Indonesia, most stipulate that renewal is conditional on "favorable performance by the concessionaire," or on government evalua- tion of satisfaction of all contractual requirements by the concessionaire. Lack of clarity in the criteria for renewal and uncertainty regarding the process of renewal contribute to insecurity of tenure and therefore to incen- tives for "mining the forest," or to "cut and run." 146 MANAGING THE WORLD'S FORESTS Conditioning renewals on criteria as vague as "performed satisfacto- rily" is not in the interests of the government or private entrepreneur or of the concessionaire. A much better approach is to structure the concession agreement itself to provide at the outset strong incentives for sustainable forest use, so that the owner and the concessionaire will have a shared interest in renewal. The renewal process can be used to induce firms to heed forest-conserving contract terms (Walker 1989, 3). Schmithusen (1977) holds that renewal value could be an important incentive for concession loggers. Revocation of unexpired concessions, once rare, has become more com- mon. In 1979 Indonesia revoked a concession of a Philippine-based firm that originally included 1.2 million hectares, largely on grounds that the firm had taken no steps to establish domestic processing facilities (Gillis 1981). In January 1989 Thailand announced the termination of virtually all 301 log- ging concessions outstanding, in response to massive 1988 floods in the south that were attributed to deforestation (Bangkok, The Nation, January. 11, 1989). And in July 1989 the government of Indonesia ''blacklisted" 114 of the 548 concession holders for nonpayment of reforestation deposits (Jakarta Post, July 17, 1989); 30 of the concessions originally blacklisted were later revoked, and 130 others were fined as much as US$400,000. The threat of revocation, when credible, may be enough to cause errant firms to comply with prevailing rules; actual revocation can be a powerful signal to offer firms tempted to circumvent harvest and financial obligations. Duration of Concessions The question of duration of timber concession rights is linked to the question of concession renewal. In nations with stable political and legal conditions, a concessionaire who has complied with agreement terms might reasonably expect an agreement to gain renewal (Schmithusen 1977; Walker 1990). Prior to 1940, the formal length of tropical timber concessions was in many cases as long as 75 to 100 years. By 1987, concession periods, even for large tracts, had become much shorter. Few governments now allow conces- sions for as long as 20 years; these include the Philippines, Indonesia, Liberia, Gabon, Sabah (for a limited number of licenses) (Gillis 1988c), and the Cen- tral African Republic (Grut et al. 1990). Elsewhere, concession periods are typically 5 to 10 years, or even shorter. Some people favor substantially lengthening the concession period be- cause of considerations pertaining to the long growing cycles (60 to 100 years, depending on the species and the location) for commercial tropical hardwoods. The foresters have long argued that concession lengths at least as long as the harvest cycle (35 years in Southeast Asia) or preferably the growing rotation. (70 years in Southeast Asia) would promote sustainable forest use. Longer concessions would give logging firms more secure, stron- ger financial interests in maintaining forest productivity, either by the use FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 147 of logging methods that involve minimal damage to residual stands or by reforesting (Whitmore 1984; Ross 1984; Setyono et al. 1985). Short conces- sion terms and the uncertainties of renewal give concessionaires little incen- tive for long-term planning and little stake in maintaining long-term forest productivity (Sedjo 1987; Kuswata 1980; Ashton 1988). Other authorities have encouraged longer leases in order to postpone uses of the natural forest that are even more destructive than selective logging. Lengthening the terms of concessions would also tend to magnify any incentives for protection of forest productivity that might arise from having concessions transferable and fully marketable. Governments have paid little heed to proposals for matching concession tenure with rotation cycles, partly be- cause prewar experience with longer concessions yielded little evidence of sustainable development practices in natural moist forests. Skeptics maintain that longer leases could reinforce uncontrolled min- ing of the natural tropical forest (Grut et al. 1990). The skeptics' argument is premised primarily on low growth rates assumed for the physical volume of commercial timber in stands and for the economic value of the stock of residual stands. (Physical growth is asserted to be between 1 to 3 percent per year for a stand remaining after logging of 100 to 200 cubic meters (m3 ) per hectare. Long-term growth rates in value of tropical timber are asserted to have averaged, at most, 1 to 2 percent per year.) Incentives for sustainable use of natural forest endowments will, from this perspective, be attractive only if the growth rate in the value of the biomass in residual stands is greater than the concessionaires' opportunity cost rate of return, which is held to be not lower than 10 to 15 percent and as high as 25 percent per year (Grut et al. 1990). Under these conditions, the incentive to "mine" the forest would be essentially the same under a 5-year, 10-year, or 70-year concession period. This is less an argument against longer concessions than a argument against the wide array of forest policy incen- tives that combine to lead to forest mining. Nevertheless, the argument merits careful attention. If (1) annual growth rates for the volume of bio- mass and the per unit (m3) value of wood are as low as 1 to 3 percent and 1 to 2 percent respectively, and (2) if as many as 100 m 3 to 200 m 3 per hectare are taken upon initial entry, and (3) if the real opportunity cost rate of return for concession holders is as high as 15 to 25 percent, it is indeed likely that concessionaires will always choose to mine the forest, because the annual growth rate in the per-unit value of the biomass in residual stands will be but 2 to 5 percent-well below the rate of return assumed to be required by concessionaires. Although growth rates as low as these may be typical for wood taken from the forests of central and western Africa, they are not representative of all tropical forests, particularly those of Malaysia and Indonesia. In Malay- sia, the net (commercially valuable) annual increment in volume is esti- mated at 2 m 3 per hectare per year. In Indonesia the increment per hectare has been estimated at between 1 m 3 and 1.5 m 3 per year. In both nations, the 148 MANAGING THE WORLD'S FORESTS harvestable stock of commercial species can be as high as 100 m 3 per hect- are, but this is extremely rare. The harvestable stock for all production forests is thought to average 50 m 3 per hectare (yields per hectare have averaged about 40 m 3 to 45 m 3 [Setyono et al. 1987]). In Indonesia, this productivity would translate into an annual growth rate in biomass of be- tween 2 and 3 percent, on the average, while in Malaysia the rate would be about 4 percent. Real rates of growth in per-unit value of tropical hardwoods taken from Malaysia and Indonesia, particularly the high-quality Dipterocarps that have dominated harvest there, have been appreciably higher than 1 to 2 percent asserted earlier. Between 1966 and 1985, real prices of tropical hardwood logs in Malaysia rose at average annual rates ranging from 2.0 percent in Peninsular Malaysia and 2.5 percent in Sarawak to 3.7 percent in Sabah (Vincent 1990a). Real prices of Philippine Lauan logs rose somewhat slower over a longer period: 1.5 percent per year between 1955 and 1986.4 And although it may be true that the real per-unit prices of tropical timber have increased by only 1 to 2 percent on average, over the long term much de- pends on the period used to compute the average. The real price of high- quality tropical wood has been rising faster than the price of low-quality wood. Moreover, there is reason to believe that this pattern will continue indefinitely (Sedjo 1987). Moreover, even in the much richer tropical forests of these two coun- tries, average harvests on initial entry do not approach 100 m 3 to 200 m 3 per hectare. Such a yield would require an intensity of harvest two to four times the harvest in Indonesia since 1970. Finally, the real opportunity cost rate of return for concession holders, while perhaps as high as 15 to 20 percent in some African countries with financial markets insulated from world capital markets, is nowhere near as high as in Malaysia and Indonesia, which together account for more tropical timber harvested under concessions than all of Africa combined. In any case, the dangers of longer concessions are not all that apparent, particularly when the forests of the Malayan archipelago are considered, and particularly when the issue is the effective, not the formal, length of tenure (because renewability can be conditioned on the degree of compli- ance by concessionaires). In the terminology of Grut et al., the growth in the value of the forest biomass would have been on the order of 7.7 percent in Sabah (4.0 percent + 3.7 percent), 6.5 percent in Sarawak, and 6.0 percent in Peninsular Malay- sia. The real opportunity cost rate of return for concession holders in Malay- sia may be no more than 5 to 6 percent, or well above the real interest rate 4 Prices are in 1980 U.S. dollars. Three-year averages for the end points were taken as a rough adjustment for cyclical behavior in the market (Panos Varangis, World Bank !EC/CM, personal communication, 1990). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 149 on short-term deposits in the globally linked financial systems of Malaysia, Singapore, and Indonesia in recent years. 5 Size Few concessions in recent years have been anywhere near as large as the 1 million hectares granted to Henry Ford in Brazil in 1926. In the late 1960s and early 1970s Indonesia granted at least two concessions that origi- nally exceeded 1 million hectares each. Typical concessions granted since 1980 seem to clump into two size categories: those between 80,000 and 100,000 hectares (Indonesia and Liberia) and those between 15,000 and 40,000 hectares (logging-only concessions in Gabon and the Philippines). Whether the smaller category offers greater advantages in policing by government and in manageability by concessionaires is an open question. Concession areas as large as those available in Indonesia (I million hectares) are clearly too large to be easily policed, whether by forestry officials (to ensure that firms comply with regulations) or by the concessionaires (to prevent inva- sion by shifting cultivators and poaching).' In view of the widespread fail- ure of concessionaires to perform according to expectations, whether in terms of protecting forest productivity or payment of forest fees, prudence would seem to argue for smaller concession areas than those typical for Indonesia: perhaps 50,000 hectares at maximum. Smaller concession areas would pro- mote greater competition in bidding in those nations moving toward an auction system. At a minimum, restriction of concession size will discour- age speculation in the form of "stockpiling" of desirable tracts to keep them out of the hands of other firms. Limitation on concession size, of course, can be reinforced by much higher area-based concession fees than now prevail in most countries. Harvest Systems and Annual Allowable Cut Selective cutting methods govern timber harvests in virtually all na- tions. Clear-cutting is ordinarily restricted to lands being converted to cattle ranching, estate crops, resettlement sites, and areas to be submerged as a result of construction of hydroelectric facilities.' The essence of selective logging is the harvest of only a small percent- age of trees above the dbh (diameter breast height) limit, whether40, 50, or 5 World Bank 1989. 6 Concessionaires may not always wish to prevent invasion by shifting cultivators. In Brazil the former may indeed permit such invasions, in order to avoid any legal consequences from failing to adhere to their forest management plans (Anderson forthcoming). 7 The area affected by clear-cutting is much larger than that affected by selective logging. 150 MANAGING THE WORLD'S FORESTS 60 centimeters. Carried to extremes, selec~ive logging involves cutting of very low intensity, which results in the "high-grading" or "mining" of for- est parcels, a practice that is often exacerbated by the structure of forest fees such as timber royalties. The most extreme form of high-grading involves the extraction only of the most marketable primary species (those that have few deformities and are easily transportable),' with no harvest at all of so- called secondary species. High-grading disturbs a larger number of forest parcels to secure a given amount of wood. And for commercial species that are intolerant of shade at critical junctures in their growth, low-intensity cutting may harm immature trees because it does not create sufficiently large openings in the forest canopy. Moreover, high-grading often involves significant depletion of tropical forest resources above the amount of wood removed from (as opposed to cut from) the forest. The reason is that stems not removed on initial entry are often badly damaged by felling, skidding, and associated extractive ac- tivities, particularly when high-lead (winch and cable) logging is the felling method employed.' As a result the second-g:rowth forest will have yields well below expectations, even in the infrequent cases that reentry on re- sidual stands is tightly regulated. Whitmore (1984), a leading authority on tropical forests, maintains that felling trees in the direction that causes least damage, plus well-planned skidding, can reduce damage to residual stands by up to half. And in the Carton de Colombia concession mentioned earlier, a low-intensity cutting system uses aerial :ropeways to reduce damage to residual stands (Poore et al. 1990). The high damage to residual standards attests to the weak signals for protecting forest productivity implicit in the overall structure of incentives facing loggers. Different selective cutting regimes involve different degrees of high- grading. Several systems illustrate the point: the Indonesia Selective Log- ging System (ISLS), variants of the Malayan l)niform System (MUS) first developed in 1949 (Ashton 1988, 195), an experimental system called Inten- sive Dipterocarp Management (IDM), an experimental system called "gap management," and the CELOS (Center for Agricultural Research in Surinam) system. 8 Many deformed and oversize trees die before the next cutting period. Interim reports on FAO plots in Sarawak indicate a mean annual mortality rate of 2.3 per- cent for well-illuminated trees (dominant and co-dominant crown classes) in selec- tively logged but untreated lowland forest. Mortality rates for old-growth forests have been placed at 1 percent (Putz 1990). 9 Logging damage refers to damage to residual stands, in the form of fallen, but unextracted trees, or in the form of damaged or standing dead trees. One study from eastern Mindanao found that 57 to 67 percent of the residual stand was dam- aged or destroyed after high-lead logging, as opposed to 38 percent from tractor logging (World Bank 1988, 19). Another study for Indonesia placed damage to re- sidual stands as high as 72 percent for high-lead logging (using long cables), and between 4 percent and 51 percent for tractor logging (Thiab and Karnasudirja 1981). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 151 The Indonesian Selective Logging System is representative of harvest systems employed in practice, if not by regulation, in many Asian, African, and Latin American nations. ISLS involves a 35-year harvesting cycle and a 70-year growing rotation (see Gillis 1988d; Sedjo 1987). Also known as the "zero silvicultural treatment option," ISLS leaves regeneration to natural processes. The assumption underlying the ISLS is that the harvest will "re- lease" the medium-size trees in the residual stand, encouraging their growth for harvest in 35 years, while the new growth is to wait two harvest cycles before it is felled. This system has little to recommend it in Indonesia or elsewhere. The low economic returns for loggers who adhere to the rules provide incen- tives for circumventing the system, with consequences that are destructive both economically and ecologically (Sedjo 1987). One of the principal meth- ods used to circumvent the system has been unauthorized concessionaire reentry on logged-over stands, sometimes within five years of initial entry. A variant of the Malayan Uniform System is the Complete Harvest and Regeneration (CHR) system. The CHR unifies the harvest and growing rotation at 70 years (Sedjo 1987), so the forest would be allowed to re- main undisturbed for 70 years, in order to rejuvenate. This system permits harvest of all marketable trees. It does not, however, imply clear-cutting, because in a heterogeneous forest large numbers of smaller trees as well as species that are nonmarketable at the time of initial entry would re- main. The system does not require the elimination (through such methods as poison girdling) of "undesired" species. Although the CHR relies on natural regeneration, it causes less damage to the stand and the soil than ISLS, leaving the forest less vulnerable to catastrophic fires (because of less ground litter) and allowing the opening of a smaller area to logging annu- ally, for a given level of harvest. The CHR provides higher economic re- turns than ISLS and is generally less in conflict with the harvest methods that concessionaires are likely to want to pursue in their financial self-inter- est (Sedjo 1987). The Intensive Dipterocarp Management system, still in the experimen- tal stages for the forests of Malaysia and Indonesia, may contain lessons for tropical forest management there and elsewhere. Under this plan, as under the CHR, all saleable trees are harvested on initial entry to a stand. How- ever, under IDM, regeneration is not left to natural processes but to new silvicultural methods. After initial harvest, two separate plantings of seed- lings are undertaken. First, a pioneer species with relatively short maturity (such as Albizza Falacataria) is planted and established in logged-over stands. Three years later, Dipterocarp seedlings inoculated with essential mycorrhizea (see Ashton 1988 for the relationship between this fungus and Dipterocarp regeneration) are planted in the understory, with the pioneer species pro- viding the requisite shade (Sedjo 1987). At year 15, the Albizza are har- vested, "releasing" the Dipterocarps, which in turn are harvested at year 50. Still another experimental silvicultural method, involving the applica- tion of "gap theory," combines elements of selective cutting with clear- 152 MANAGING THE WORLD'S FORESTS cutting (Hartshorn 1989a). This system, which has shown some promise in the tropical forests of La Selva, Peru, features harvests of long, narrow clear cuts. A gap in the canopy is thereby created, as a means of promoting natural regeneration of native tree species. Reportedly, about 63 percent of the tree species that attain the canopy in La Selva, such as those that natu- rally arise from tree falls, depend on gaps for regeneration. Simulation of natural gaps by harvesting of all timber along narrow strips (30 to 40 meters), 100 to 300 meters long and about 150 meters apart, is reported to promote excellent colonization of the harvested strip by native trees (Hartshorn 1989b). This option merits scrutiny for management of forests where gap- dependent tree species are abundant. High-grading of the heterogeneous Amazonian natural forests has generally yielded only 3 m 3 to 5 m 3 per hectare (Hartshorn 1989a), compared with an average of 40 m 3 to 45 m 3 per hectare in Indonesia. Clear-cutting of the narrow strips provides yields of 250 m 3 per hectare; slash is not burned, but scrap wood is converted to charcoal. The Palacazu forest management model iin Peru represents an attempt to integrate economic, ecological, and social principles in a setting where forest property rights are held by local groups with a continuing strong stake in sustainable development of tropical forest resources. As such, it offers the possibility of using tropical forests without destroying them and the biological diversity they harbor. Finally, there is the CELOS (Center for Agricultural Research in Surinam) system, a polycyclic management system that, like the MUS, is based on controlled felling of trees above a certain diameter. The system involves significant, and therefore expensive, silvicultural treatment after felling, 10 including liana cutting and two poison-girdling treatments of noncommer- cial trees (resulting in the death of 40 percent of the biomass that compete with advanced growth of commercial species).11 Sustained yield of high- quality timber (20 m 3 per 20-year felling cycle) is apparently possible, if felling and subsequent silvicultural treatments are gentle (Putz 1988). Thus systems similar to the Indonesian Selective Logging Systems yield the most serious degree of high-grading. For the Dipterocarp forests of the Malayan archipelago, the CHR merits consideration. The promising but still experimental IDM method also deserves consideration for use in Southeast Asia. For the forests of Latin America, the CHR can be recommended with much less confidence because the understory of Latin American forests dif- fers from that of Southeast Asian forests. 10 The CELOS system, while apparently sustainable, reportedly yields wood that is several times more expensive than wood logged by other methods from Surinam's abundant forest (Poore et al. 1989). " The CELOS system treatments are to be applied in three stages: first within 1 to 2 years of felling, then halfway through a 20-year felling cycle, then a few years before the second cut (Putz 1988). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 153 Harvest systems based on application of gap theory to tropical forest management offer some promise in some parts of Amazonian forests. It remains to be seen whether this method can overcome problems associated with the growth of weeds and creepers following gap clearance. Environmental Safeguards Most forestry codes were written before many of the more serious en- vironmental problems of deforestation were apparent to the general public or to government policymakers. By 1990 three decades of intensive logging on an extensive scale, coupled with growing numbers of shifting cultiva- tors, had made readily evident the economic as well as the ecological costs of rapid deforestation. In Thailand forest cover shrank from 66 percent of land area in 1950 to only 29 percent in 1985; as noted earlier, the government has blamed massive forest destruction for disastrous floods in the South. Similarly, in the Philippines, logging of steep slopes around Bacuit Bay, Palawan, has caused siltation that has contributed to smothering the nearby coral reef (Myers 1988). Rapidly growing sedimentation traceable to deforestation also has di- minished the useful lives of hydropower projects in Brazil, India, and the Dominican Republic (Gillis forthcoming). In East Java, sedimentation threat- ens to reduce by two-thirds the economic life of a large multipurpose dam, planned for 100 years (Panayouto 1988). In the Philippines, forest damage from improperly constructed access roads (World Bank 1988, 19), as well as skidding trails and nondirectional felling, has become a major source of long-term sedimentation. In July 1990 the government of the Philippines imposed a substantial "environmental fee" -which takes the form of a uni- form specific royalty charged semiannually-to be used to offset adverse environmental effects of logging. Institutional problems, rather than lack of knowledge, account for much of the failure to curb environmental degradation. Weak enforcement of en- vironmental protection measures in logging is a consequence not only of relying on understaffed, underpaid, or suborned forest agencies but also of placing responsibility for monitoring and control in distant central govern- ments rather than in the hands of local groups with the strongest stake in sustainable development of forest resources. Where property rights for for- est exploitation are in the hands of local groups rather than the central government, effective measures have been more often taken to protect eco- nomic and social values in the forest. Measures requiring directional felling and proper construction of access roads and logging trails contribute not only to the health of residual stands but also to a reduction in the erosion that has led to many downstream sedimentation problems. Where behavior cannot be easily changed to re- duce adverse environmental effects of logging, the forest owner can at least impose fees to offset these costs. But the prospects for protecting stands and 154 MANAGING THE WORLD'S FORESTS the downstream environment from shifting cultivators, firewood gatherers, and charcoal makers who follow legal logging and collectively complete the process of deforestation are dim so long as the property rights to forest land are not vested in local groups. FOREST REVENUE SYSTEMS Forest revenue systems can be made to serve several objectives beyond that of collection of forest fees and taxes for the government as owner of the resource and as sovereign taxing authority. Forest revenue systems can be deployed to reinforce incentives for efficient and sustainable forest and environmental management, which are either implicit or explicit in well- crafted concession agreements and forestry codes. Moreover, by influencing concessionaires' decisions about forest renewal, increased use of trees and species, and forest-based industrialization, forest revenue systems can accomplish objectives for which concession terms may be ill-suited to promote. Few forest revenue systems go far toward satisfying these objectives. Capture of timber rents through forest fees and taxes has been unsatisfac- tory in virtually all nations with tropical timber endowments, sometimes markedly so. Taxes and fees are typically low, partly because of inadequate (or nonexistent) mechanisms for adjusting them to inflation, and partly be- cause of official accommodation to the rent-seeking behavior of politically well connected timber concessionaires. Moreover, generally low fees and taxes nominally imposed on timber operations often are not collected, in part because of inordinately complicated revenue systems and in part be- cause of shortages of forestry officials. Finally, the structure of forest fees and taxes often provides strong signals (or reinforces other incentives) for environmental degradation instead of protection of important economic and ecological forest values. The severity of these problems varies greatly from country to country. But in general, inadequate forest revenue systems in virtually all countries continue to be the prime cause of deep underpricing of the resource, with all that underpricing implies for incentives for sustainable development of forest resources, as well as capture of timber rents. Forest revenue systems consist of two principal elements: requited pay- ments to the owner and unrequited payments from extractive firms to the government. The designs of these two elemenlts reflect two different objec- tives. The owner of a natural resource ordinarily seeks compensation for the taking of the resource; a quid pro quo, or requited payment, is involved. No such quid pro quo is involved in the exercise of sovereign powers of taxa- tion;12 income, sales, and export taxes are unrequited payments levied in 12 Except, of course, in the limited number of instances in which taxes are fully based on the benefit principle, or in the case of local property taxation, when (and if) benefits received from local government services are strongly related to property values. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 155 order to support general government activity, or to achieve social and eco- nomic purposes that may, but usually do not, bear any relation to use of natural resources. Requited Payments to Owners In the overwhelming majority of nations with tropical forests the most important type of requited payment by concessionaires is the timber roy- alty, or stumpage fee. The other significant categories of requited payment are license fees and reforestation fees. A limited number of countries use auction of concession rights as an alternative or, more appropriately, a valu- able supplementary method for collecting royalties or license fees. Royalties Timber royalties have taken several different forms in tropical nations but may be grouped into three general categories: specific royalties based on volume, per-tree royalties, and ad valorem royalties. In turn, the three may be implemented in various ways, with important implications not only for rent capture but for forest-use practices. Specific royalties. Specific royalties based on volume of timber extracted have been imposed in two principal forms: uniform specific royalties and differentiated specific royalties. Uniform specific royalties are those levied on every cubic meter har- vested, regardless of value, or on the costs of extraction, transportation, and marketing. Traditionally, this most common form of royalty has been im- posed on roundwood at the log pond or roadside, although in recent years collection has shifted to the processing site (sawmill or plymill). The royalty is imposed on processed products as they leave the mill, by means of pre- sumed conversion ratios.13 Uniform specific royalties, once used widely in Africa, are now used only in Liberia, Ecuador, Thailand, the Philippines, and the Peninsular Malaysian state of Kelantan. In addition, Indonesia imposes a special "addi- tional royalty" and a reforestation fee, but not a stumpage fee, as a uniform specific royalty. The single advantage of the uniform specific royalty relative to other forms of royalty is administrative: it is the simplest of all timber levies and is therefore conceptually the easiest to collect. But the substantial drawbacks 13 The higher the conversion ratio, the lower is the rate of log use, and presum- ably, the lower is efficiency in procession. In Indonesia, for example, the official conversion ratio for plywood has been 2.3 m3 of logs per 1 m3 of plywood. The ratio for sawn timber was 1.82 (Gillis 1988d). Similar conversion ratios have been used for plywood in Ivory Coast (2.4), but the ratio reported for Gabon in 1980 is much lower: 1.42 (Gillis 1988c). 156 MANAGING THE WORLD'S FORESTS of the uniform specific royalty make it one of the most unsatisfactory forms of forest levy. First, this form of royalty involves high-grading of stands at its most pernicious. The royalty directly increases the "cutoff" grade below which stems will not be harvested because the costs of extraction (inclusive of the royalty) exceed the value of the stem. Lower-quality trees of the most com- mercially desirable species will be bypassed, as well as higher-quality trees of less desirable, or secondary, species (see Gillis 1980, 1988c). Second, the uniform specific royalty performs poorly in capturing rents for the owner: if the royalty is high enough to capture most rent from harvest of the most valuable stems, it will capture little of the potential total rent from a stand, because it will induce loggers to bypass all but the high- est-value stems. (See Vincent [1990a] for an exposition based on Malaysian experience.) Moreover, because in practice this royalty will be set low enough to allow harvest of trees below the highest value, it contributes to the under- valuation and thus the wastage of wood. This form of royalty is poorly suited for rent capture for yet another reason. Uniform specific royalties are nowhere indexed to inflation, whether domestic or world inflation. When expressed in domestic currency, real values of the royalty can depreciate by as much as 18 percent per year, as in the Philippines between 1980 and 1986. And even when expressed in U.S. dollars, the real value of such royalties has declined by as much as 5 percent per year (in the early 1980s), and by as much as 22 percent in the 12 months prior to September 1990. Finally, uniform specific royalties discriminate in favor of accessible stands and stands involving low extraction costs. Specific royalties differentiated by species or location have been used in several countries. In Indonesia before 1979 different royalty rates were ap- plied to nine species (with the highest royalty on, the most valuable species); royalty rates were also uniformly lower for less accessible regions (Gillis 1980). Two Peninsular Malaysia states, Terengganu and Pahang, impose differentiated royalty rates on five species of timber (Vincent 1990b). Sarawak, Thailand, and the Ivory Coast use royalties differentiated by species; the Papua New Guinea royalty is, in effect, differentiated by species and location. The differentiation of specific royalties by species or location of harvest represents a crude attempt to approximate an ad valorem royalty system. Inevitably, the degree of differentiation of volume-based royalties has proved insufficient to approximate the results obtainable under ad valorem sys- tems. This type of royalty suffers to some extent from all the drawbacks of the uniform specific royalty without exhibiting the only merit of the latter, simplicity. While valuation for ad valorem royalties may have presented FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 157 difficult problems in the 1960s or 1970s, this is no longer necessarily the case, and thus these royalties are preferable to specific royalties." Per-tree royalties and variants. Per-tree royalties are employed in Ghana and Nigeria. Collection of this form of royalty is simple; it requires only the verification of the number of trees actually cut, whether by counting stumps (ex post) or actual stems taken. Therefore, the per-tree royalty system requires no grading and scaling of logs, a common source of difficulty in all other types of royalty systems. In practice, per-tree royalty systems tend to be highly differentiated by species or location: Ghana applies 39 different rates, with the highest rate 3.3 times the lowest. In Nigeria the highest rate is 38 times that of the lowest rate (Grut et al. 1990). The per-tree levy has advantages other than simplicity. Because the royalty is based not on volume but on stems cut, loggers might harvest large individual trees that they would have bypassed had the royalty been based on volume harvested, thus providing much larger canopy openings for younger trees left behind. In addition, with significant royalties estab- lished on a per-tree basis, it would not be worthwhile to harvest small trees, which would be left to form the crop for future harvests. Per-tree royalties achieve the purpose of Asian-type selective cutting systems without having their complexity. In addition, this royalty encourages full use of those trees worth harvesting provided the royalty is assessed on all trees killed, whether used or not. Having paid for the tree, a logger will use as much of it as is financially worthwhile (Gray 1983). Related to the per-tree royalty is the system wherein royalty is assessed on the estimated volume of timber in the standing stock, not on the volume actually taken. This is the system used in the Congo. The efficacy of this method of royalty collection depends heavily on the quality of the forest inventory for stands harvested. With perfect and complete inventory infor- mation and sufficient differentiation of royalty rates, the system offers sev- eral potential advantages. If truly imposed on estimated volume in stems above minimum-size girth, rather than volume extracted, it would capture substantial timber rents. Because the royalty would be paid on saleable trees whether taken or not, pressures for high-grading would be much less than under a uniform specific or an ad valorem royalty. Ad valorem royalties. Ad valorem royalties, rare in tropical nations before 1975, are now used in Cameroon, the Central African Republic, Gabon, a few states in Peninsular Malaysia, and Sabah, and have been under consideration in Papua New Guinea (World Bank 1990) and the Philippines. 14 Close and reasonably up-to-date approximations to market values can be easily obtained for virtually all "primary" species and most "secondary" species merely by retaining the services of any of the dozens of reputable international accounting firms. 158 MANAGING THE WORLD'S FORESTS In most of these nations the ad valorem royalty is uniform across all species and locations, with rates between 1 percent and 6 percent, although in Peninsular Malaysia the royalty is differentiated by species. Only in Sabah are ad valorem royalties adjusted to reflect differences in extraction costs and conditions in world markets. Potentially, ad valorem royalty systems can, by their very nature, cap- ture rent better than volume-based royalties can. This is not the case, how- ever, for ad valorem royalties imposed at effective rates as low as in the Central African Republic (0.25 percent). Ad valorem royalties produce a lesser degree of high-grading than would be the case for any administrable volume-based royalty system yielding equivalent revenues (see Gillis 1980, 1988d). Ad valorem royalties can also be differentiated, with lower rates on lower-quality species, if a high premium is placed on minimization of forest mining. Cost-adjusted ad valorem royalty systems can capture rents well for the owner, because royalty rates can be set very high on high-quality timber that has low extraction and transport costs-with much less risk that log- gers will be induced to high-grade than would be the case if costs were not taken into account. This type of royalty has been successfully used in Sabah15 for well over a decade, where it has managed to capture a sizable share of timber rents. The Sabah royalty also contains a progressive feature that contributes to its efficacy in rent capture, a scaler, which increases as log prices reach progressively higher plateaus. The scaler has a value of 0.6 when log prices fall below a certain minimum, 16 and it rises to 0. 9 for the highest-price plateau. Royalties on pfocessed timber. Royalty systems were traditionally applied to roundwood values (log volume or value prior to processing or export). As domestic processing of timber has steadily replaced export of round wood, however, pressures have arisen to apply royalty systems to output of processed products, such as sawn timber and plywood. The nature of these pressures is both administrative (Grut et al. 1990) and financial. Administra- tively, it is argued that the number of processing plants is smaller than the number of sites where logs are collected after harvest (log ponds, roadsides) and that processing plants are more accessible to royalty collectors. Furthermore, basing royalty systems on processed products captures illegal logs and eliminates problems and abuses in grading and scaling of logs. Forest fees based on processed products apply to the volume, not value, of output. Typically, the amount of the royalty is based on royalties previ- ously applied to round wood. For example, a royalty of US$10/m 3 on round- wood is converted to a royalty of US$10/m3 of sawn timber or plywood 15 The Papua New Guinea-differentiated volume-based royalty also allows deduc- tions for presumed costs. 16 The 0.6 scaler applied at US$8 per hoppus foot in the early 1980s. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 159 output; this conversion produces a drastic reduction on royalty collections. In Indonesia, at conversion ratios of 1.8 for logs into sawn timber and 2.3 for logs into plywood, a US$10 royalty on m 3 in log form becomes US$5.65/m3 sawn timber and US$4.35 for plywood. Therefore, in the event that the royalty system is converted to a levy on processed timber, the royalty rate must be changed to reflect conversion ratios. Even when the royalty rate is adjusted accordingly, a system based on output of processed products con- fers an advantage on vertically integrated processing firms, which benefit from postponement of royalty until the product is produced or sold. The chief drawback of this system, however, is that the royalty on pro- cessed products must be based on presumed average conversion rates, which discourages improvements in recovery of wood from roundwood inputs and penalizes relatively efficient processors (Grut et al. 1990). The incentive to waste wood can be very strong. A processing firm that improved the conversion of roundwood into plywood from 2.3 to 2.0 would find that its royalty liability would increase by 15 percent per unit of log input. Impo- sition of the royalty on processed products also facilitates evasion by illegal small mills. In sum, the case for basing royalties on processed output is not strong, notwithstanding the apparent administrative advantages of doing so. License Fees License fees are generally based on total concession area, adjusted for location (accessibility and distance), quality of stand, or both. On occasion, license fees have been based on the value of the annual allowable cost (e.g., the Philippines). The fees may be imposed initially, annually, or both. Li- cense fees may be used to raise revenues from forest activities and to pro- mote other important objectives of public policy toward forest resources and forest land use. A well-designed system of license fees can help to discourage over- exploitation in forest stands that are more accessible to a nation s transport 1 structure. License fees differentiated according to distance and accessibility can help compensate for higher transport costs. Reduced license fees on low-density stands or stands with generally lower-grade composition can be used to promote harvesting of secondary species and lower-grade primary species. High license fees encourage smaller, more manageable concessions. High fees can also discourage "stockpiling" of concessions: holding valu- able stands off the market to keep them out of the hands of competitors (Gillis 1980). High license fees levied annually provide strong incentives for efficiency-higher use and recovery in logging-and further use of second- ary species (Grut et al. 1990). Systems of license fees in force in most tropical nations achieve few of these purposes. Most are clearly too low to play much of a role in rent capture. The initial fee is relatively low in Cameroon, Indonesia, and the 160 MANAGING THE WORLD'S FORESTS Philippines, and the annual fee is extremely low in the first two countries plus Ghana, Ivory Coast, Liberia, and Sarawak (see exhibit 7-1). Conversely, the annual license fees for the Central Africa Republic, Thailand, and Pen- insular Malaysia are 10 to 50 times the fees in the low-fee countries. Furthermore, few nations use license fees to discourage overexploitation of more accessible stands, to compensate for differences in stand intensity, or to compensate for high transport costs. Only Malaysia differentiates by density of stand, and the rates are high enough to be significant: the rate for high-density stands (US$91.91 per year) is 61 percent higher than that for low-density stands. These rates provide strong disincentives for stockpiling of concessions. Reforestation Fees Provision for reforestation fees was uncommon in the early stages of large-scale logging activities in tropical countries. Where reforestation was EXHIBIT 7-1. Illustrative License Fees Country Nature of Fee Cameroona Initial fee: US$0.05/hecfrare Annual fee: US$0.31/hectare Central African Republica Ranges from US$0.043/hectare/year to US$6.80; inversely related to concession length (20 to 5 years) None Permis Temporaire: Rates vary in four zones. Rates range from US$0.01 to US$0.06, according to accessibility of concession. Ghanab Annual fee: US$0.07 /hectare Indonesia Initial fee: US$0.55 /hecil:are Annual fee: US$0.85 /hectare Ivory Coastb Initial fee: US$0.99 /hecil:are Annual fee: US$0.05 /hectare Liberia Annual fee: US$0.63/hectare Peninsular Malaysia Initial fee: US$57.44 for low-density stands (below 25 m 3 /hectare) US$91.91 for high-density stands (25 m 3 /hectare or higher) Philippines' Initial fee: US$1/hectare Annual fee: 5% of value or allowable cut. Sarawakd Auction fee paid per hectare: M$0.20 to M$0.30 (US$0.14) Thailand US$16.70/hectare/year SOURCES: a Grut et al. (1990). b Gillis (1988c). 'World Bank (1988). d Gray (1983). FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 161 considered at all in the design of forest policy in Southeast Asia and Africa, framers of concession agreements relied on reforestation clauses in the agree- ments themselves to assure that forest enterprises would undertake reforestation efforts. These provisions proved unsuccessful, particularly in Indonesia (Gillis 1981) and the Philippines, where the government reported that 90 percent of concessionaires had failed to carry out reforestation pro- grams prescribed in concession agreements (Boado 1988). In response, governments began, usually after a decade of experience with large-scale logging, to resort to reforestation fees that in some cases have been higher than timber royalties. These fees are typically of three types: (1) fees im- posed on roundwood extraction, whether volume based, such as the uniform specific levies in Liberia and Sabah (exhibit 7-2), or flat-rate ad valorem levies, as in Central African Republic, Congo, and Gabon; (2) fees imposed on processed timber output, as in Indonesia since 1989; and (3) reforestation fees based on area, as in the Philippines/' Peninsular Malaysia, and Central African Republic. The various forms of royaltylike reforestation fees have the same effects on timber use as equivalent timber royalties. The fees imposed on processed timber may have the same additional problem that is associated with the conversion formula from round wood to finished lumber. Because fees based on area do not heighten incentives for forest mining, this form of reforesta- tion fee is much superior to fees based on volume, or even on value. The literature on tropical forest policies customarily treats reforestation fees in the same category as timber royalties and license fees. This is clearly appropriate when reforestation fees are viewed as essentially royalties col- lected by another method and when one of three conditions are met: (1) when reforestation fees are imposed in the form of deposits paid by conces- sionaires that are refundable upon completion of reforestation activities sat- isfactory to forest owners; (2) when the entire proceeds of reforestation fees are plowed back into government programs to secure reforestation (includ- ing regeneration) objectives; or (3) when the reforestation fee is high enough to induce concessionaires to undertake serious reforestation efforts. If none of these conditions is met, the reforestation fee differs little from any other form of unrequited payment extracted by government from concessionaires, with little or no relationship to forest policy objectives. Most of the reforestation fees now in force satisfy at least one of the three conditions specified and are therefore appropriately classed as requited forest payments. Fees are refundable in the Philippines and Peninsular Malaysia upon verification of reforestation efforts (exhibit 7-2). The Indone- 17 In the Philippines, however, reforestation fees on concessionaires were not en- acted until 1987, nearly a quarter-century after the beginning of the logging boom in that nation. 162 MANAGING THE WORLD'S FORESTS EXHIBIT 7-2. Reforestation Fees Basis · Country Rate Remarks 1. Roundwood Extraction Liberiaa US$3 Im' on all Conservation tax is imposed a. Volume species, plus a atarateofUS$3/m3 on 11 conservation tax prime species, US$1.50 on on locally processed others. timber Sabahb M$0.15/hoppus ft Earmarked to finance reforestation programs. b. Flat-rate Ad Central lOo/o of valuer Valuer mercuriale = 1I4 Valorem African commerciale price of log Republic mercuriale Congoc 2.5%-3.5% Rate is negotiated for each concession contract. 3.5% Earmarked to finance reforestation program. 2.0utput of Processed lndonesiae US$7/m'of Effective rate in roundwood Products processed output equivalent is US$3.00 for plywood, US$3.85 for sawn timber. Fees are nonrefundable. 3. Concession Area Central Paid on hectare per Rate depends on length of African year in concession attribution period; lowest Republic rate applied to 20-year attribution. Philippinesf US$10,000/hectare Refundable bond for deposit reforestation. Peninsular US$17.24/hectare Refundable deposit (as of Malaysiag 1980). SOURCES: a IITO, preproject report by International Institute for the Environment and Development, drca 1989. b Gillis (1988b). 'Grut et al. (1990). d Gillis (1988c). e Republic of Indonesia (1989). 'World Bank (1988). g Gillis (1988c). sian reforestation fee was initially refundable when created in 1980 but no longer is. 18 There is no record of refunds ever having been paid (Gillis 1988d). As far as the second condition is concerned, reforestation fees are ear- marked to finance reforestation programs in Sabah and Gabon (exhibit 7-2). The Indonesian fee has been loosely related to government programs for investments in tree plantations, on the theory that plantation forests are 18 Payment of these fees in Indonesia was substantially in arrears prior to July 1989. This situation was one of the reasons why Indonesia switched the basis of the fee from roundwood extraction to roundwood utilization in processing mills. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 163 capable of producing substantially more wood per hectare than natural for- ests, and therefore can relieve the pressure on natural forest ecology. Forest analysts have questioned this strategy not only from an ecological but also from an economic and social standpoint (Sedjo 1987). In no case, however, has the reforestation fee been set high enough to induce, by itself, concessionaires to engage in any significant reforestation activities involving regeneration of primary species in selectively logged parcels-the aim of most reforestation programs. 19 In the Dipterocarp for- · ests of the Malayan archipelago, reforestation fees would have to be refund- able and set as high as $20 I m 3 harvested to induce enrichment planting of primary species in cut-over stands, according to cost estimates by Leslie (1983). Unrequited Payments The two most important forms of unrequited payments by concession- aires are export taxes and income taxes. Until the 1970s, export taxes on logs were by far the largest producers of government revenue in this category. Since then, income taxes have become more significant than export taxes. Export Taxes Until recently virtually all tropical timber was exported in roundwood form. Thus export taxes were for all practical purposes merely another form of stumpage fees, easier to collect than the latter because logs were exported through a small number of bottlenecks known as ports. In many cases, export taxes per m 3 were higher than stumpage fees and furnished substan- tially more government revenue,20 in spite of substantial evasion through underinvoicing (Grut et al. 1990). Export taxes on timber were at one time the principal incentive for forest-based industrialization. High export taxes on logs and much lower, usually zero, taxes on processed products (especially plywood) provided powerful inducements for investment in processing facilities. 19 Refundable fees at the level imposed in Indonesia before 1989 were, however, high enough to induce concessionaires to establish tree plantations outside selec- tively logged areas and usually outside the concession area (Sedjo 1987). 20 This was clearly the case in Indonesia from 1968 to 1979. Export taxes on logs were 10 percent of f.o.b. value from 1968 to 1978 and 20 percent from 1978 to the present. The ad valorem equivalent of the highest stumpage fee at no. I-grade meranti harvested in East Kalimantan was 5 percent. Although both stumpage fees and export taxes were widely evaded before 1980 (particularly before 1973), export tax collections on logs were four times larger than royalty collections (Gillis 1980). 164 MANAGING THE WORLD'S FORESTS Export tax structures proved only moderately successful in promoting forest-based industrialization. Some significant investments in domestic processing facilities were achieved, particularly in Indonesia, Malaysia, and the Philippines, but at substantial economic costs. In Indonesia, the export tax structure furnished effective protection rates of 222 percent for plywood manufacture, creating strong signals for waste and inefficiency in opera- tions. Even worse, in sawmilling, the loss in export taxes to the government was not compensated by any gain in value added in sawmilling: measured at world prices, value added in sawmilling was a negative $15.00/m3 (Gillis 1988). Very high rates of effective protection Ito domestic timber processing also prevailed in Gabon, Ivory Coast,21 Papua New Guinea (Gillis 1981), the Philippines, and Sabah. 22 For Papua New Guinea, even as late as 1990 the forcing of forest-based industrialization through export tax differentials might entail high economic costs as well as losses in government revenue. The use of export tax structures to promote forest-based industrializa- tion became largely redundant by 1989, as more and more countries moved to prohibit exports in log form. Log exports were banned in Peninsular Malaysia as early as 1971 (Gillis 1988b), in Ghana in 1979 (Poore et al. 1989), and by 1979 were restricted to 25 percent o:f annual allowable cut in the Philippines. Exports in log form have long been prohibited in Brazil (Schmink 1988), Ecuador, and Indonesia. Several other nations adopted bans on log exports in the 1980s, and a similar restriction was recently proposed in Papua New Guinea (World Bank 1990). Still, export taxes on logs remain in effect in at least eight tax jurisdictions, counting the states of Sabah and Sarawak in Malaysia (see exhibit 7-3), and including some nations where log exports are not allowed. Moreover, Indonesia recently extended the use of the export tax structure to make the export of sawn timber virtually prohibitive. Income Taxes Income taxes are classic examples of unrequited payments to govern- ments. Until recent years, income taxes have been an insignificant source of government revenue, and they will remain so, particularly in Africa, where income tax administration is weak. Governments in Asia, Africa, and Latin America have employed income taxes to promote forest policy goals; these measures have invariably resulted in tax revenue losses. 21 In Ivory Coast, export taxes forgone on exports in log form in order to gain additional domestic value added in sawmilling ranged from 108 percent of addi- tional value added for the low-valued Llomba species to more than 200 percent for the higher-valued Iroku and Acajou species (Gillis 1988c). 22 The royalty rate for domestically processed timber in Sabah is 7 percent-about one-fifth that for roundwood exports. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 165 Before 1975 income tax instruments were widely used to provide incen- tives for investment (particularly foreign investment) in timber sectors, first in logging activities, later in domestic processing of timber. Typically, the incentives took the form of income tax "holidays" (full income tax exemp- tion) for periods ranging from 5 to 10 and even, in Ivory Coast, 15 years EXHIBIT 7-3. Export Taxes on Timber, Selected Countries, 1989 Country Rate Remarks Cameroona 2 % of valuer mercuriale The same source indicates elsewhere that export taxes are 11 % of average f.o.b. prices. Central African Logs: US$11.45/m3 for Tax is imposed on processed wood export. Republica red woods, US$1 l.07 for Roundwood equivalents are much lower. white woods Sawn timber: US$250 I m 3 Veneer: US$2.90/m3 Ivory Coastb Specific rates variable Sapelli: US$57.49 by species Sipo: US$89.20 Assamela: US$138.27 Liberiac Ranges from US$1.44/m3 Tax, which is called the Industrialization for low-valued species Incentive Fee, is imposed only on logs. to US$58.56 for high- valued species (Sipo) Indonesia 20% ad valorem on logsd Ad valorem tax imposed only on some logs from inaccessible regions. Log export ban since 1985 has made the tax irrelevant for other regions. For sawn timber, export Specific export taxes were enacted on sawn taxes are specific and timber in 1989; plywood is exempt from all range from US$250 I m 3 to export taxes. US$2,400/m3.° Papua New 10% of f.o.b. value Tax reported to have been widely evaded Guineaf (logs) through transfer pricing. Sarawakg 15% ad valorem on f.o.b. Tax applies to only one hardwood species. value (logs) See remarks The Sabah timber royalty has a strong export tax feature: the royalty rate for log exports is almost 10 times the rate for logs used domestically. SOURCES: a Grut et al. (1990). b Gillis (1988c). c ITIO, preproject report by IIED (circa 1989). ct Gillis (1988d). e International Monetary Fund (1990). f World Bank (1990). g Gillis (1988b). h Vincent (1990a). 166 MANAGING THE WORLD'S FORESTS (Browder 1988; Gillis 1988b, 1988c, 1988d). After the mid-1970s, eligibility for tax holidays was increasingly restricted to firms agreeing to establish domestic timber-processing facilities. Beginning about that time, governments began to heed a growing body of evidence that tax incentives for natural-resource-based industry in gen- eral and logging operations in particular were not necessary to attract in- vestment in extractive activities, and that, given low national benefits in the form of employment, taxes paid by logging firms were the most significant benefit from logging and the only major benefit from logging activities con- ducted by foreign enterprises (Gillis 1980; Gillis and Repetto 1988). In the early 1980s, governments further restricted tax holidays for forest-based industry and in some cases, abolished them altogether (in Liberia in 1981 and Indonesia in 1984). (Brazil was a notable exception.) Aside from income tax holidays, governments have offered income tax credits to encourage investments in logging and ranching in tropical forests. By 1990 income tax holidays and tax credits for timber concerns had become almost anomalous. They remain in Gabon and Ivory Coast, and, despite ample evidence of the inadvisability of income tax incentives in extractive activity of any kind, Papua New Guinea in 1989 announced a 10-year tax holiday for companies processing at least half of their annual log harvest (World Bank 1990). Income tax incentives have occasionally been used to promote other objectives of forest policy, including reforestation. The most conspicuous example has been in Brazil. In the 1960s Brazilian forest enterprises were allowed to earmark half of all income taxes due on forest operation to spe- cial accounts controlled by the enterprise, rather than remitting the taxes to the government (Extenso Rural 1976). This much-touted incentive report- edly encouraged the replanting of 1.8 billion trees in 820,000 hectares be- tween 1970 and 1975, although much of this "replanting" may have oc- curred in tree plantation projects rather than in selectively logged parcels, and would not be classified as reforestation under some definitions of that term (Gillis and Repetto 1988). Rent Capture Capture of timber rents by governments has become increasingly de- pendent on the level and structure of forest fees, including reforestation fees as well as royalties and license fees. Export taxes on logs, once an important tool for rent capture, have been rendered virtually irrelevant in this respect because of log export prohibitions. As tax holidays and investment tax credits expire in nations that have used these instruments and are abolished for new investments in timber sectors, income taxes may become important tools for rent capture in the small number of tropical countries having reasonably effective income tax administration. Income taxes are particularly attractive supplementary meth- FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 167 ods for collecting tropical timber rents, particularly because income taxes do not provide incentives for mining or high-grading the forest. Moreover, income tax credits can be used to reinforce incentives for reforestation. Efficient and sustainable forest utilization practices will necessarily involve increased emphasis on both license fees (including auctions) and income taxes, rather than stumpage fees (including stumpagelike reforesta- tion fees). These measures are also essential if tropical nations are to improve on their records of capture of timber rent. Discussions of timber rent capture must take into account not only realized rents accruing to the government-as-owner and to timber con- cerns, but also rents destroyed by logging damage, by inefficient methods of processing, and by high-grading (Vincent 1990a). Timber rent, therefore, is best viewed as the potential rent that would accrue if all harvested logs were allocated to uses (direct export, sawmills, plymills) that yield the larg- est net economic rent. In turn, potential rent consists of two elements: (1) rents actually realized by governments and by loggers, including unrecorded rents accruing to loggers as a result of underinvoicing of exports and log poaching; and (2) unrealized rents, composed of residual rent and logging damage (Vincent 1990a; 1990b), as well as rents destroyed by inefficient processing. Rents destroyed by. highly protected timber processing can be sizable. In Indonesia in 1980 and 1981 potential timber rent totaled US$1.72 billion; fully 27 percent of this amount was unrealized, primarily because of ineffi- cacies in processing (Gillis 1988d). Residual rents-living trees left standing in the forest-result from either logging restrictions in prescribed harvest methods (silvicultural rent) or from high-grading where royalty rates are high. Losses from high-grading can be significant. Between 1966 and 1985, an estimated 9 percent of potential rents went unrealized in Sabah because of concessionaires' failure to extract marketable trees. Exhibit 7-4 depicts various realized and unrealized components of rent. Forest revenue systems, concession terms, prescribed harvesting systems, and excessively high incentives for forest-based industrialization all deter- mine not only the division of realized rents between governments and tim- ber concerns but also the proportion of timber rents received by no one (unrealized rents). The interaction of these factors also determines the pro- portion of unrealized rents that are clearly destroyed, for all time, in har- vesting and processing. Some fraction of unrealized silvicultural rent is not destroyed forever, but may be available in future harvests. Exhibit 7-5 presents estimates of potential rents, actual rents, and gov- ernment rent capture in six jurisdictions (including three Malaysian states). In the cases presented, government rent capture approached 50 percent of actual rents only in Sabah; Indonesia managed to capture a third of actual rents, Peninsular Malaysia about a fifth, and the Philippines only a little more than a tenth. Elsewhere, the available evidence indicates large short- falls in rent capture as well. 168 MANAGING THE WORLD'S FORESTS EXHIBIT 7-4. Rent Distribution Under Inefficient Forest Revenue Systems and High Protection to Domestic Pmcessing Potential rent I Realized Unrealized rent rent I I I I Recorded Logging extraction Unrecorded extraction Resi:J rent damage rent (1) I I I Forest revenue Windfall (rent capture Silvicultural High-grading Processing profits rent (1) waste (1) by to firms governments) SOURCE: Vincent (1990a), adapted by the author l:o include the rents destroyed in processing of timber (process waste) by inefficient log conversion in domestic ply- wood manufacture and sawmilling, and by failure to make use of wood chips (for export or domestic use) and sawdust NOTE: As the rent taxonomy is used in Vincent (1990b), high-grading does not include logging damage caused by careless logging methods (e.g., nondirectional felling) or by "cut and run" practices attributable to excessively short duration of concession agreements. Given the numerous avenues through which rent can be dissipated before some of it winds up in the hands of the forest owner, it is not surpris- ing that government capture of tropical timber rent has been so low. Except for royalties in Sabah, license fees in Liberia and Malaysia, and the 1990 environmental fee in the Philippines, rates of levies on forest operations are generally low. In many countries weak administration of income taxation and forest fee systems further erode tax revenues from the forest sector, through unrecorded or misrecorded transactions. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 169 EXHIBIT 7-5. Potential Rent and Government Rent Capture in Tropical Timber Production, Selected Periods Official Potential Actual Govern- Rent Rent ment Column3 + Column3 + from Log from Log Rent Colwnn2 Column 1 Country and Period Harvest Harvest Capture (%) (%) (1) (2) (3) (4) (5) Indonesiaa (1979-82) 4,954 4,409 1,644 37.3% 33.2% (US$ millions) Philippines (1979-82) 1,505 1,033 171 16.5 11.4 (US$ millions) Malaysiab (1966-85): (M$ millions) Sabahc 17,720 16,990 8,190 48.2b 46.2 Sarawak 7,290 7,260 1,340 18.5 18.4 Peninsular Malaysia 11,030 11,000 2,410 21.9 21.8 Ghana (1971-72) 38.0d SOURCES: a Repetto and Gillis (1988). b Vincent (1990a). Malaysian estimates assume a value of 3.0 for the elasticity of the marginal cost curve for logging, which represents the "middle" case for this elasticity. Vincent also presents results of estimates based on elasticities of 1.5 and 4.5. c For the 1979-82 period, Gillis (1988d) estimated that the government of Sabah captured 81 percent of actual rents from log harvests. d The Ghanaian figure is for the proportion of rents caph1red by government, per ID 3• CONCLUSIONS: OPTIONS FOR REFORM OF FOREST REVENUE SYSTEMS Programs for reform of tropical forest revenue systems should have at least four main objectives: (1) to capture a greater share of rents for the forest owner; (2) to correct serious underpricing of forest resources; (3) to remove strong incentives for inefficiency in the use of the wood resources of the forest, not only in wood harvests but in processing industries; and (4) to establish forest fees high enough to offset the adverse environmental effects of logging. As it happens, measures taken to secure these objectives can be mutually supportive. Moreover, gearing reform to these objectives would advance other important goals as well: to reduce the disturbance of natural forests and to provide a sounder, long-term basis for industrialization in tropical countries. Reform should be focused on the establishment of a system of forest fees that will impose higher rates than at present, despite the very weak or inadequate administration of forest fees and taxes. This prescription is not equivalent merely to increasing rates of existing taxes and fees on forest- related operations. Many forms of forest levies in use in many countries are 170 MANAGING THE WORLD'S FORESTS unsuitable to increase rent capture, to corr,ect underpricing of forest re- sources, to enhance efficiency in timber use, and to offset environmental costs of logging. 23 Finally, reform measures must be designed to work well with weak administration, because many proposals for reform involve administrative difficulties and burdens almost as severe as those the reforms are intended to replace. The appropriate reform of forest revenue systems will vary with admin- istrative strengths and weaknesses, and with the quality of information about forest inventories. Reform Under Sufficiently Strong Institutions and Administration and Reasonably Cie>mplete Information Nations with a sufficiently strong administration of forest fees and in- come taxes and reasonably up-to-date and complete forest inventories have a much wider array of workable options than other nations. They are able to shift away from traditional forms of forest fees, such as royalties, toward use of timber auctions, heavier (and differentiated) license fees, and income taxation of timber concerns. Auctions, whether sealed tender or in other forms, may be used not only to allocate parcels to qualified bidders but also to capture for the gov- ernment a greater share of timber rents. But because forest inventories are unlikely to be complete or fully up-to-date, auctions will, for some time, need to be supplemented by other tools. A shift toward heavier and more differentiated forms of area license fees becomes mandatory under these conditions. Initial auction of parcels, coupled with annual license. fees approaching those used in Thailand, is advisable to enhance rent capture, to promote better use of the forest, and to reverse the underpricing of timber. Differentiation of license fees accord- ing to stand quality and accessibility accords with these objectives as well. Where environmental damages from logging can be quantified even approximately, charges to offset these costs can be imposed. In the Philip- pines, for example, the environmental fee enacted in 1990 was set at a level thought to roughly approximate the costs of both on-site and downstream environmental damages for an average operation taking 20 to 30 m 3 per hectare on initial harvests. The Philippine fee, however, is imposed as a uniform specific royalty (adjustable every six months) and at a level high enough (US$20/m3) to induce further forest mining. Environmental charges imposed in the form of license fees geared to area of annual harvest would be far superior to volume-based charges for this purpose. 23 For example, drastic increases in uniform specific royalties in nations where these levies are used would be inappropriate for enhancing rent capture for owners because the resultant increase in high-grading would destroy large amounts of po- tential rent. Or, very sharp increases in export taxes on logs would do nothing to correct serious underpricing of timber used or processed domestically, and would heighten incentives for low recovery in processing. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 171 Finally, where income tax administration is reasonably adequate, as in Colombia, Costa Rica, Malaysia, Surinam, and to a lesser extent Indonesia, greater reliance on this fiscal instrument for capture of timber rents is pos- sible. Because income taxes do not directly impinge on decisions in timber harvesting and processing,24 a shift from output-based forest levies to in- come taxation would reduce incentives for forest mining. If in any given country the combination of auction fees, license fees, and income taxes proves insufficient for reaching rent capture targets, certain forms of timber royalty could continue to be used. The royalty should take the form of an ad valorem lery adjusted for costs. With other levies bearing the brunt of the burden of taxation on timber, the royalty, as in Sabah, could be set at levels well below the royalty imposed on log imports, thus reduc- ing incentives for waste in timber use. If a royalty is used to supplement rent capture, it must take the form of a levy on roundwood inputs to domestic processing facilities. Imposition of the royalty on processed output is inimical to greater recovery in timber processing, and penalizes more efficient firms as well. Finally, use of the export tax structure to capture rent and to promote domestic processing is unnecessary in countries where other tools of rent capture are available, and inadvisable because sharply reduced or, more commonly, zero export taxes on processed products help to heighten the underpricing of timber resources. Moreover, high rates of effective protec- tion resulting from export taxes so structured sharply reduce incentives to minimize costs in domestic processing of timber. Reform Under Weak Institutions and Administration and Poor Information Under some circumstances-where forest agencies are weak,25 under- funded, and immobile, as is claimed for much of central and western Africa; where the quality of forest inventories is poor; and where income tax ad- 24 For a domestic firm, the prospects of income tax obligation may influence the decision to undertake logging or processing investments. For foreign firms based in countries where host-country taxes can be credited against home-country income taxes, host-country income taxes are largely irrelevant if (as is increasingly the case) tax rates are not dissimilar. 25 Weakness in forest agencies is not always attributable to low salaries and low levels of training. It is also often due to institutional factors. For example, through- out much of the 1974-84 period, Indonesian forest officials were relatively well paid and trained. However, the effectiveness of the forest service was severely Weakened not only by lack of cooperation with other departments (especially agriculture and transmigration) but by the pattern of deployment of forest officials. A dispropor- tionate number of staff were deployed in Java rather than in the outer islands where natural forest resources are located. The ratio was one forester per 1,764 hectares of forest in Java versus one to 471,000 hectares for the outer islands (about 267 times higher). The consequences for effectiveness in grading and scaling, as well as polic- ing of concessionaire activity in the outer islands, are obvious (Tarant et al. 1987). 172 MANAGING THE WORLD'S FORESTS ministration is inadequate-the institutional framework for forest policy severely limits options for policy reform and little reliance can be placed on auction and income tax instruments. Nevertheless, improvements are possible. Even this group of countries should emphasize license fees rather than output-based royalties, for the same reasons presented earlier. For many countries in this category, license fees could be increased 5- or 10-fold with- out exhausting the rent capture potential of this instrument (compare the license fees depicted in exhibit 7-1 for Cameroon, Congo, Ivory Coast, and Gabon). At the same time, it is unlikely that the quality of information on stand qualities and accessibility is so poor as to preclude at least a modest degree of differentiation of license fees to reflect these attributes. Most countries in this group must continue to rely on royalties for rent capture, although the efficacy of royalties will continue to be limited by severe problems in grading and scaling. However, the administrative weak- nesses impeding the collection of royalties in many countries are often the main obstacles to implementation of radical reform of forest fees. One reform that has been proposed is to replace all existing forest fees (save perhaps export taxes) with a single, ammal fee for a concession. This annual fee would be subject to adjustment each year, according to a formula made known to concessionaires when they signed a concession contract. The formula would be based on the weighted average f.o.b. value of the relevant forest (Grut et al. 1990). This radical proposal would still require scaling of log volumes as well as on-the-ground inspection to ensure that logging is confined to the approved area and that only trees above legal diameter are taken. Moreover, the proposal requires relatively complete information pertaining to stand density, quality, and accessibility, as well as up-to-date information on f.o.b. prices for all commercial species. While reform of forest fee systems should move in the direction just indicated, timber royalties must remain important elements of forest rev- enue systems in this group of countries. But the royalties should be shifted from volume based to ad valorem. This shift could be facilitated by retain- ing international accounting firms to provide up-to-date information on tim- ber values. A moderate degree of differentiation in ad valorem rates may be administratively feasible, and would assist in rent capture and reduction of forest mining. In any case, any royalties used should be imposed on round- wood inputs, not processed output. Several countries in this group already impose some reforestation fee; in most there is ample scope for increasing the fees, preferably in the form of license fees, rather than royaltylike levies. Countries are likely to have little latitude for use of income taxation on timber operations. Countries with weak income tax administration are par- ticularly vulnerable to transfer price abuses by international firms, or affili- ates of international firms. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 173 REFERENCES Adeyoju, K. 1976. A Study of Forest Administration in Six African Nations. Rdme: FAO. Anderson, Anthony. Forthcoming. "Smokestacks i~ the Rain Forest: Industrial Development and Deforestation in the Amazon Basin." Ashton, Peter. 1988. "A Question of Sustainable Use." In People of the Tropical Rain Forest, eds. J.S. Denslow and C. Paduch. Berkeley: University of California Press. Boado, E.L. 1988. "Incentive Policy and Forest Use in Philippines." In Public Policies and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cambridge University Press. Browder, J. 1988. "Public Policy and Deforestation in the Brazilian Amazon." In Public Policies and the Misuse of Forest Resources, eds. R. Repetto, and M. Gillis. Ne·w York: Cambridge University Press. Bucovetsky, M., and M. Gillis. 1978. "The Design of Mineral Tax Policy." In Taxation and Mining, eds. R. Repetto and M. Gillis. Cambridge, Mass.: Ballantine Press. Cardenas, J.R., and R.P. Pena. 1990. "Alqunas Experiencias Colombianas en El Manejo Sostendido De Los Recursos Naturales Y La Reforescacio." Unpublished manuscript, Bogota, July. Extenso Rural, Staff. 1976. "Tax Incentives and Reforestation." Development Digest 14 (January). Gillis, Malcolm. 1980. "Fiscal and Financial Issues in Tropical Hardwood Concessions." Develop- ment Discussion Paper No. 110. Cambridge, Mass.: Harvard Institute for International Develop- ment. _ _ _. 1981. "Foreign Investment in the Forest-Based Sector of the Asia-Pacific Region." Unpub- lished report prepared for FAO. _ _ _. 1983. "Taxes, Returns and Risks in Petroleum Contracts in LDCs." Development Discus- sion Paper No. 142. Cambridge, Mass.: Harvard Institute for International Development. _ _ _. 1988a. "The Logging Industry in Tropical Asia." In People of the Tropical Rain Forest, eds. R. Repetto and M. Gillis. Berkeley: University of California Press. _ _ _. 1988b. "Malaysia: Public Policies Resource Management and the Tropical Forest." In Public Policy and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cambridge University Press. _ _ _. 1988c. "West Africa: Resource Management Policies and the Tropical Forest." In Public Policy and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cambridge University Press. _ _ _. 1988d. "Indonesia: Public Policies, Resource Management·and the Tropical Forest." In Public Policy and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cam- bridge University Press. _ _ _. Forthcoming. "Economics, Ecology and Ethics: Mending the Broken Circle for Tropical Forests." Gillis, M., and.R. Beals. 1980. "Indonesian Hard Mineral Agreements." Natural Resources Forum 4 (October). Gillis, M., and R. Repetto. 1988. "Conclusion: Findings and Policy Implications." In Public Policy and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cambridge University Press. Goodland, R.J.A., E.0.A. Aisbey, J.C. Post, and M.B. Dyson. 1990. "Tropical Moist Forest Manage- ment: The Urgent Transition to Sustainability.'' Washington, D.C.: World Bank. Gray, John A. 1983. Forest Revenue Systems in Developing Countries. Rome: Food and Agricultural Organization. Grut, Mikael, Nicolas Egli, and John A. Gray. 1990. "Forest Revenue Systems and Concession Management Policies for West and Central Africa Moist Forests." Unpublished draft, World Bank, June 22. 1 Hartshorn, Gary. 1989a. "Application of GAP Theory to T;ropical Forest Management: Natural Re- generation on Strip Clear-Cuts in the Peruvian Amazon." Ecology 70(3). _ _ _. 1989b. "GAP-Phase Dynamics and Tropical Tree Species Richness." In Tropical Forests, eds. L.B. Holm-Nielsen and I.C. Nielsen. New York. Harcourt Brace Jovanovich. 174 MANAGING THE WORLD'S FORESTS Herfindahl, Orris C. 1974. "The Value of Mineral Surveys to Economic Development." In Resource Economics: Selected Works of Orris Herfindahl, ed. D.B. Brooks. Baltimore: Johns Hopkins Univer- sity Press. International Monetary Fund. 1990. Indonesia Recent Economic Development. (Washington, D.C.: International Monetary Fund). ITTO 1989. International Tropical Timber OrganiZation." Preproject Report by the International Institute for Environment and Development. Unpublished circular, London. Jacobs, Marius. 1988, The Tropical Rain Forest: A First Encounter. New York: Springer-Verlag. Kalt, Joseph. 1987. "The Redefinition of Property Rights in American IIldian Reservations." Discus- sion Paper No. E-87-05. Cambridge, Mass.: Kennedy School of Government. Kuswata, K. 1980. ''East Kalimantan: A Comment." Bulletin of Indonesian Economic Studies 16(2). Leonard, Jeffrey. 1987. Natural Resources and Economic Developrnent in Central America. New Brunswick, N.J.: Transaction Books. Leslie, Alf. Food and Agricultural Organization. 1983. Personal communication. Myers, Norman. 1988. "Environmental Degradation and Son1e Economic Consequences in the Phil- ippines," Environmental Conservation 15 (Autumn):205-13. _ _ _. 1989a. "Deforestation Rates-in Tropical Forests and Their Climatic Implications." London: Friends of the Earth. _ _ _. 1989b. "Cut in Production." Review article in Nature 342 (November 16). _ _ _. 1990. ''The World's Forest and Human Populations: The Environmental Interconnections." Population and Development Review 16. Panuyouto, Theodore. 1988. "Economics, Environment and Development" Development Discussion Paper No. 259, Harvard Institute for International Development. Panuyouto, Theodore, et al. 1990. "The Economics of Environmental Degradation." Draft prepared for USAID, October 1989. Pearson, Scott, and John Page. 1972. "Development Effects of Ghana's Forest Products Industry." Accra: USAIP. Poore, Duncan, Peter Burgess, John Palmer, Simon Rietbergen, and Timothy Synnott. 1989. No Timber Without Trees: Sustainability in the Tropical Forest. A study for International Tropical Timber Organization. London, Eng.: Earthscan Publications. - Putz, Francis E. 1988. Reviews of "A Silvicultural System for Natural"Regeneration of Tropical Rain Forest in Suriname, and Three Other Related Volumes.''' Forest Science (September). _ _ _. 1990. Personal communication, August. Repetto, R. 1988. "Overview." In Public Policies and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. New York: Cambridge University Press. Repetto, R., and M. Gillis, eds. 1988. Public Policies and the Misuse of Forest Resources. New York: Cambridge University Press. Republic of Indonesia. 1989. Presidential Decree No. 31, 1989; June 27, The Reforestation Fund. Republic of Philippines. 1990. Department of Environment and Natural Resources. Administrative Order No. 60, June 27. Ross, Michaels. 1984. "Forestry in Land Use Policy in Indonesia." Ph.D. dissertation, Oxford Uni- versity. Schmink, Marianne. 1988. "Big Business in the Amazon." In People of the Tropical Rain Forest, eds. J. Denslow and C. Paduch. Berkeley: University of California Press. Schmithusen, Franz. 1977. Forest Utilization Contracts on Public Lands. 2d ed. Rome: FAO. _ _ _. 1980. "Forest Utilization Contracts." Paper presented at FAO Conference, August, Pattaya, Thailand. Sedjo, Roger. 1987. "Forestry Policies, Incentives and Distortions in Indonesia." Unpublished manu- script, September 15. Setyono, 5., et al. 1985. "A Review of Issues Affecting the Sustainable Development of Indonesia's Forest Land," vol. 2. Unpublished manuscript. Southgate, Douglas, and Morris Whitaker. 1989. "Promoting Resource Degradation in Latin America: Tropical Deforestation, Soil Erosion in Ecuador." Unpublished manuscript, November. FOREST CONCESSION MANAGEMENT AND REVENUE POLICIES 175 Tarant, James, et al. 1987. Natural Resources and Environmental Management in Indonesia: An Overview. Jakarta: USAID. Thiab, Jumali and Saparman Karnasudirja. 1981. "Review of the Application of Selective Cutting Outside Java,'' Indonesian Agricultural Research and Development Journal 3 (no. 3):61--66. Trum-Barina, K. 1981. "Forests of Ohara." In Where Have All the Forests Gone?, eds. V.H. Sutlive et al. Williamsburg, Va.: College of William and Mary. Vihcent, Jeffrey. 1990a. "Rent Capture and the Feasibility of Tropical Forest Management." Land Economics 66 (May). _ _ _. 1990b. Personal communication. Walker, Robert. 1989. "Tropical Deforestation and Contract Dissolution." Unpublished manuscript, Florida State University, Tallahassee. Whitmore, T.C. 1984. Tropical Rain Forests of the Far East. Oxford, Eng.: Clarendon Press. Winterbottom, I. 1990. Taking Stock· The Tropical Forestry Action Plan After Five Years. Washington, D.C.: World Resources Institute. World Bank. 1988. Philippines: Environment and Natural Resources Management Study. Washington, D.C.: World Bank. --~· 1989. World Development Report 1989. Washington, D.C.: World Bank. _ _ _. 1990. Papua New Guinea-the Forestry Sector: A Tropical Forestry Action Plan Review. Wash- ington, D.C.: World Bank. 8 Fuelwood Problems and Solutions D. Evan Mercer and John Soussan C oncern over the "fuelwood crisis" facing the world's poor has been widespread since the late 1970s (Eckholm et al. 1984; Soussan 1988; Agarwal 1986). At first the problem was frequently overstated. In the ex- treme, analysts (foresters, economists, and others) in many countries made erroneous projections of the rapid total destruction of the biomass resource. These projections were usually based on simplistic supply and demand analysis, the so-called gap analysis that was highly influential throughout most of the 1980s (Leach and Mearns 1988). These projections often led to projects that sought to boost fuelwood supplies without regard to local needs, priorities, or resource potentials-or to the economic viability of the plans. Fuelwood problems are now recognized as rarely generalizable. Fuelwood use and scarcity reflect complex and variable interactions be- tween local production systems and the environmental resources on which they are based. The significance and origins of fuelwood problems in, for example, a semiarid area such as parts of the Sahel are very different from those in a mountainous region such as Nepal or a high-density /high-pro- ductivity area such as the Kenyan highlands. A further sharp distinction is evident between rural localities (where fuel wood is. usually a free good gathered locally) and urban areas (where fuelwood is a commodity pro- duced elsewhere). Fuelwood problems have complex causes and take varied forms. In rural areas these problems reflect changes to economic and environmental relationships that affect local supply and demand (Armitage and Schramm 1989). These changes can be gradual, such as erosion of local woodlands as a result of land colonization, increased herd sizes in semiarid regions, in- 177 178 MANAGING THE WORLD'S FORESTS creased exports of fuelwood to meet growing urban demands, or lower quantities of residues available as fuel as a result of changing agricultural practices. In some cases the changes can be sudden and catastrophic: for example, large-scale deforestation associated with giant development schemes; mass influxes of refugees; and environmental collapse associated with droughts, .floods, or other extreme climatic events. Whether gradual or rapid, these changes lie at the heart of fuelwood problems and set clear limits on the opportunities open to confront the problems effectively. This chapter first analyzes the nature and origins of fuelwood problems and past policy approaches to solving them. The analysis focuses on the relationship of fuelwood problems to other development issues and on the various forms the problems take under different environmental and eco- nomic conditions. Fuelwood problems are viewed primarily as one conse- quence of the interaction of environmental and economic forces at the local level, which result in a number of resource stresses. We then suggest an approach to formulation and reform of fuelwood policy based on this analy- sis. The annex to this chapter presents a typology of specific fuelwood situ- ations. Just as there is no one fuelwood problem, there are many potential solutions. The key is to identify what will work where and why. THE NATURE OF FUELWOOIJ PROBLEMS Use of Fuelwood in Rural and Urban Areas Use of fuelwood in the growing cities of the third world contrasts sharply with the more familiar patterns of rural areas. Despite the growth of energy use in other sectors, rural household consumption still dominates the en- ergy budgets of many developing countries, particularly in the poorer na- tions of Africa and South Asia. Fuelwood accounts for more than 75 percent of the energy used in countries such as Nepal, Bangladesh, Ethiopia, Burkina Faso, and even oil-rich Nigeria (Soussan 1988). Cooking consumes most of this energy, and most of the energy is supplied by biomass fuels (agricul- tural residues as well as fuelwood). These fuels are usually gathered freely from the local environment, and their production and use cannot be readily separated from other aspects of land resource management within rural economies. Because rural people rarely fell trees for fuel use and generally depend on trees near their homes, trees outside the forest, within the agri- cultural landscape, are the main source of fuel for rural people. The reliance on local land resources means that tenurial arrangements are important. The poor often have few alternatives to fuelwood to meet their basic subsistence need, and problems associated with access to fuelwood can be considered an integral part of the wider rural development crisis. Whatever happens to energy resources and prices at an international level, rural fuel wood use will be important for the foreseeable future. This essential fact must be used as a starting point for the development of policies to deal with rural fuelwood problems. FUEL WOOD PROBLEMS AND SOLUTIONS 179 As urbanization proceeds, the effects of urban fuelwood use and the problems associated with it are growing rapidly (Soussan et al.1990b; Floor 1987). As in rural areas, most fuelwood in cities is used in the household sector (although the use of fuelwood in small industries such as restaurants, bakeries, and brick kilns can be significant locally). The concept of an en- ergy transition, however, is central to any understanding of household en- ergy in urban areas. As urbanization proceeds and the use of fuel increases, people tend to diversify and switch from wood and charcoal to modern fuels. The stages in this transition are typically not discrete; it is common to find within a household several fuels used for cooking. Recent studies (Leach and Mearns 1988) suggest that even in cities where fuelwood is more expen- sive than the modern alternatives, people prefer fuelwood because the sup- ply is more secure; the fuelwood is available in small, affordable quantities in local markets; and fuelwood requires no expensive initial investment in cooking stoves. Therefore, to understand urban fuelwood problems, it is essential to understand the structure of urban fuel markets. The cost of fuelwood to urban consumers (especially the poor) can be significant, and there is some (though patchy) evidence that in many places the cost is increasing. A more general problem is access to fuel (ETC 1987). Markets for many nonwood fuels are typically poorly developed, particu- larly in peripheral neighborhoods where many poor people live. Govern- ments often restrict imports and the internal distribution of commercial fuels. This situation contrasts markedly with fuelwood markets, which usu- ally reach all corners of the city. Demand for urban fuelwood and charcoal can have a devastating im- pact on the rural areas from which supplies are drawn because urban deal- ers often clear fell woodland areas and make no attempt to conserve the resource base. In effect, they mine the resources and pay only the extraction costs for them. The problems associated with fuelwood use must be dealt with in the context of the control and management of land resource systems. Some of these complexities are discussed in the next two sections. Access and Alternative Uses The existence of fuelwood resources in a locality is not enough to guar- antee that no one experiences fuelwood problems; the resources must be available for use by all who need them. A series of factors that limit access to wood resources have been identified as limitations imposed by the loca- tion of the resources in relation to demand, by land tenure and ownership of biomass resources, and by the way in which biomass resources are man- aged (ETC 1987; Johnson and Tomkins 1989). The locational limitations on access reflect features of the landscape. Most important is the distance between the sources of supply and the point of use. In many localities biomass fuels are gathered freely from the envi- ronment, and the main cost of fuelwood use is the time to collect the fuel 180 MANAGING THE WORLD'S FORESTS (Agarwal 1986). Resources beyond a certain distance will take too long to collect. The time required to collect fuels is also influenced by features of the terrain such as hills, rivers, and gullies. Steep slopes, rugged terrain, and watercourses add significantly to the collection time. Locational constraints on access can be calculated in relation to the benefits accrued for the time and effort taken. Whether people are willing to pay this price depends on the available alternatives to wood, the users' income, and the opportunity cost of the collectors' labor time. A series of access constraints also derive from the legal status of land in an area. The availability of biomass resources inevitably depends on their ownership, which is a function of the ownership of the land on which they grow. Three broad categories of land tenure can be identified for our purposes: I. Land that is owned individually by members of the local com- munity-basically, private farmland; 2. Land that is owned by the state, large commercial producers, or other institutions controlled from outside the area-plantations, commercial farms, state forests, and the like; and 3. "Common" land resources that are legally owned by the state or through customary communal forms but having no proprietorial restrictions on access to them-open rangelands and woodlands, hillsides, grazing lands, and so on. Biomass resources (both residues and trees on farms) from private farm- land are the main source of fuel in many rural areas. Households with little or no land may face severe restrictions on access to fuels even if there appears to be a local surplus. In many cases, land-poor and landless families have traditional rights to collect fuels from the land of larger land owners, but such rights may be eroded when biomass resources are under stress or become commercialized. Rural change, which is marked by greater com- mercialization, new technologies that reduce tree coverage and residue pro- duction, and higher population densities in many areas, has aggravated the inequalities of access to private biomass resources. Access to wood from large-scale commercial farms and plantations is frequently highly restricted or prohibited altogether. Some limited collec- tion rights may be allowed and illegal removal is common, but if the man- agers of these resources use them for fuel (for example, for crop processing) or sell them (for fuel, timber, pulp, or other uses), access is limited by the policing action the managers take. This category of land may contain a significant proportion of the biomass resources of a locality, and even if limited collection is allowed, these resources may be alienated from the effective control of local people. Access to biomass resources on communal land is the most complex of all. A range of traditional customs and practices may regulate access to these resources, but these customs tend to break down as local economies FUEL WOOD PROBLEMS AND SOLUTIONS 181 change and resource pressures increase (e.g., when outside groups, such as urban wood dealers, exploit the resources). In many areas the amount of communal land is rapidly declining, as woodlands and rangelands are cleared and enclosed for agricultural production. In these areas the lack of limita- tions on access can result in the unsustainable exploitation of what are often vital but fragile resources. The final category of access factors consists of those associated with the system of biomass resource management, which reflects the prevailing so- cial structure, resource management and harvesting techniques, nonfuel uses of different categories of biomass resources,· and the rights and obligations of different sections of the community. Depending on their gender, class, and age, people typically have very different approaches to the control and use of biomass resources. Fuel provision is frequently the women's respon- sibility, whereas the men control the resources (especially land and cash) from which the fuel comes. Any attempt to deal with fuelwood problems must take this division of rights and obligations into account. Women and Fuelwoods Collecting fuelwood is physically hard and time-consuming work, an arduous burden on women who typically are also responsible for collecting water, caring for children, doing agricultural work, and handling the myriad other tasks that make up the day of third-world women. As pressures on the local resource base develop, the distances traveled, collection times, and other demands on women also increase. In addition, because fuelwoods are smokier and dirtier than modern fuels, women's health may be impaired by fuelwood use. When fuel shortages lead to changes in the number or type of meals cooked, women (who, even in the absence of fuelwood shortage, eat less well then men) often disproportionately suffer negative nutritional con- sequences. In these and other ways, fuelwood stress hits the health and environment of women harder than those of men in many parts of the third world. Given their central role in fuelwood provision and use, women will often best understand where and in what form fuelwood problems are found, what interventions are more likely to succeed, and which groups in the community should be involved in designing and implementing interven- tions. Therefore the sometimes vague notion of community participation must be structured to identify and create situations and institutions within which women can play a central role in building and implementing fuelwood solutions. Giving women such a role implies challenging or circumventing the social and institutional barriers that limit the scope for integrating women in the planning process. Needless to say, this task will be difficult in most communities where men control the resources (especially land and cash) needed to confront fuelwood stress. Furthermore, men typically dominate local institutions that regulate resource management, provide links to the outside world, and perpetuate gender inequalities. This situation creates 182 MANAGING THE WORLD'S FORESTS particularly difficult problems because the local land, finance, and institu- tions are obviously the resources that need to be harnessed to build local solutions to fuelwood stress. Responses to Fuelwood Stress Pressures on the biomass resource base are not simply a matter of fuel demand and scarcity; they relate to the control over and the range of uses made of biomass resources. Biomass fuel stress is often a product of the disruption of traditional systems of resource management-disruption that is in turn generated by a number of forces that vary in importance from locality to locality. One of the first responses to fuelwood stress is more careful fire man- agement, which can result in dramatic fuel savings. This one-time gain, however, is inevitably limited by the laws of thermodynamics. Conserva- tion consciousness takes extra time, but results in desirable and sustainable efficiency gains. Fuel scarcity may also produce changes in cooking prac- tices that are less clearly beneficial. People may turn to enclosed stoves, losing other functions of the fire such as light, heat, and a social focus. They may reduce the number of meals cooked per day or adopt quicker-cooking foods. The dietary implications of these developments are hard to quantify but may be serious. As biomass resources become scarce, the other, nonfuel uses can be cut back. This may impair other aspects of the production system. For example, increased use of dung as a fuel may reduce soil fertility, and overexploitation of common woodland resources may jeopardize the availability of fodder or construction materials. When resources are abundant, the alternative uses of biomass materials produce no problems, but when resources are scarce, conflicts between alternative needs may emerge. How these conflicts are resolved depends in large part on who controls the resources and who benefits from the alternative uses. The sorts of resource conflicts outlined here are at their most acute in regions where residues have replaced wood as the main household fuel. When wood is unavailable, people may switch "downward" to crop or animal residues (Barnard and Kristofersen 1985). In many areas, such as Bangladesh, northern India, and Lesotho, residues have long been the domi- nant household fuel, because wood is far too valuable to burn. Alterna- tively, people may switch "upward" to nonbiomass fo.els such as kerosene and liquefied petroleum gas (LPG). If this swiltch occurs through choice, it is a good indicator of development, but the switch may occur through ne- cessity, when people have no choice but to spend scarce cash on a basic need. When pressures on biomass fuels become acute, rights and responsibili- ties connected to needs provision and resource management may change to reflect emerging scarcities. Often, men begin to assist in fuel provision, either FUEL WOOD PROBLEMS AND SOLUTIONS 183 by helping to gather fuel or by providing alternative resources. Landless or land-poor families may find that traditional rights to collect fuel on the land of others are eroded (a particular problem when resources acquire commod- ity value). Traditional, sustainable management practices of communal land may break down, depleting the resource base and eventually resulting in the loss of the indigenous knowledge on which these practices are based. These and other changes are symptomatic of the erosion of the social fabric of local communities as systems break down under the traumas of resource scarcity. Finally comes the actual or potential erosion of the land resource base, as conflicts between the different uses and needs of different groups are resolved temporarily by the extraction of biomass materials (not just wood, but fodder, manure, and so on) at a rate greater than the capacity of the local environment to produce these materials. The conflict between the im- . mediate needs for survival and the long-term maintenance of the resource ~ase often produces environmental degradation. Fuelwood, Deforestation, and Land Degradation Much of the international concern over tropical forests has centered on the loss of large tracts of natural forest areas. The causes of this deforesta- tion are associated mainly with the drive to open up and exploit what is seen as one of the last great land frontiers. Commercial logging, clearance for large-scale ranching, in-migration as a result of road construction or through government-sponsored transmigration schemes, flooding from giant hydroelectric power schemes, and other development pressures are all widely cited as contributing to large-scale deforestation (Fearnside 1986; Monbiot 1989; Tyler 1990). The exploitation of forests for fuelwood use contributes little to this process. This is especially true for fuelwood gathered to serve the needs of local rural communities, because where there are large tracts of forest, there are usually few people. Commercial exploitation for urban fuel wood and charcoal markets does have an impact in many regions. Fearnside (1989) cites charcoal production for iron smelting in the Carajas region of Brazil as a threat to the forests of eastern Amazonia; a major study by the Overseas Development Adminis- tration (Bird and Shepherd 1988) has shown that the acacia woodlands of the Bay region of Somalia are being devastated by charcoal production for Mogadishu's markets; Bowonder et al. (1987) detail the impact of urban fuelwood use around a number of Indian cities, and Soussan et al. (1990) provide further evidence from a number of sources. Rural fuelwood use is often cited as a factor in large-scale deforestation, but these assertions are rarely substantiated. Indeed, the evidence points the other way: where the forests are opened up, land clearance leads to massive fuelwood surpluses, and substantial quantities of wood resources are either burnt or left to rot. In such circumstances, talk of stress associated with the overexploitation 184 MANAGING THE WORLD'S FORESTS of forest resources for fuel use obscures the real causes of large-scale de- forestation. Fuelwood use, both for local needs and external markets, does have other environmental implications, There is considerable evidence to suggest that land resources in agricultural areas can experience degradation if overexploited for fuel use. This problem is particularly associated with the small areas of woodland scattered within agricultural landscapes. Most farm- ing areas have woods on steeper hillsides, along river courses, on marshy ground or areas of poorer soils, and in other areas not used for farming. These woodlands, which are often communally owned and managed (al- though the state may have nominal proprietorial rights), provide a range of products and are an integral part of the rural economy. Peasant farming systems are based on the use of both private and communal land resources to produce goods for markets and for subsistence consumption. Growing populations, increasing commercialization of rural economies, and other incentives to clear more land have often led to the incremental. colonization of these resources for cultivation. Cultivation leaves a smaller area to serve growing needs for fuel and other products (especially fodder); the result is overexploitation of biomass resources. The remaining commu- nal areas usually have scant regenerative capacity and declining stocks and become more vulnerable to environmental hazards such as drought or soil erosion. The loss of communal resources may lead to an increasing reliance on externally produced commercial goods, which, in turn, depend on greater cash income. Communal goods are further eroded as they are exploited for private commercial gain. Fuelwood use certainly contributes to the degradation of land resources in agricultural regions where more general resource pressures are felt. This form of degradation, however, is far from universal; indeed in most rural areas, fuelwood gathering for local use has only a marginal, if any, impact on land resource quality. It is a problem precisely where the rural economy and environment is most vulnerable: in localitiles where the resource base is already under threat and where the community has the fewest resources to counter this threat. Many case studies illustrate this form of environmental stress. Saxena (1987), Moench (1989), and Singh et al. (1984) give examples from the foothills of the Himalayas, Christianson (1988) provides evidence from Tanzania, and Johnson and Tompkins (1989) illustrate the pressures on Swazi Na ti on lands in Swaziland. In more comprehensive studies, Smil (1983) cites local fuelwood use as a contributory factor to land degradation in China, and Ryan (1990b) presents a preliminary analysis linking the deg- radation of woodland resources in many parts of India to the pressures of growing fuel demand. When local fuelwood use does contribute to land resource degradation, it is not the sole, or even the main, cause of this stress. Fuelwood use con- tributes to this degradation because it is an integral part of the rural economy I environment relationship, and it is the general pattern of rural development FUEL WOOD PROBLEMS AND .SOLUTIONS 185 in poor, environmentally vulnerable areas that creates stress, not fuelwood use alone. FU ELWOOD POLICY ANALYSIS Since the mid-1970s most governments and donor agencies have ap- proached the fuelwood crisis as an energy demand and supply problem. Both the diagnoses of problems and the designs for solutions have been based in large part on simple models of supply and demand (Foley 1988; Leach 1988; Leach and Mearns 1988; McGranahan 1986; Teplitz-Sembitzky and Schramm 1989). Viewed from this perspective the solutions were self- evident; if projected fuelwood demands exceeded supplies, the solution was to plant more trees and shift the supply curve outward, or to devise policies to reduce demand and shift the demand curve inward. Most of these efforts have failed to have lasting effects on fuelwood scarcity or forest depletion. These failures, however, have led to a rethink- ing of the fuelwood crisis (Teplitz-Sembitzky and Schramm 1989). Although the specific details vary widely between and within regions, fuelwood prob- lems are now more clearly seen as manifestations of more fundamental failures in rural land, labor, and capital markets, urban energy markets, and failures of governments (local and national) to establish the conditions that would allow efficient and sustainable allocation of land and resources be- tween forest and cropland and wood and food production (Dewees 1989; Cline-Cole et al. 1990; Leach 1988; Teplitz-Sembitzky and Schramm 1989). The complex interactions among demand, supply, and market forces and the lack of success of fuelwood interventions indicate that considerable care is required for designing appropriate interventions. 1 For example, some policy interventions may have paradoxical effects. Improving the efficiency of charcoal production may actually increase rather than decrease defores- tation rates by expanding the areas accessible to urban markets (World Bank 1987). Likewise, if consumer disposable income is effectively increased through the subsidization of improved stoves, the quantity of fuel demanded and consumed might increase (Clarke and Shrestha 1989a; Foley 1988). In this section, the common supply- and demand-side policy approaches for solving fuelwood problems are examined. Supply-Side Policies Most supply-side fuelwood policies and programs have concentrated on planting trees or manipulating fuelwood prices. Subsidized tree-planting programs have been frequently proposed and implemented as solutions to fuel wood shortages with few successes. Recent criticisms of these programs 1 More complete models of the interactions of fuelwood demand and supply are presented in Hyde and Mercer (1990) and Mercer (1991). 186 MANAGING THE WORLD'S FORESTS have centered on the analysis (e,g,, fuelwood gap theory) used to justify the expenditures and on the extent to which tree planting for fuelwood produc- tion is economically justifiable or feasible given local fuelwood prices, land tenure and property rights regimes, and the local causes of deforestation, The supply-side pricing policies attempt to influence supply through fuelwood royalties, stumpage fees, and fuelwood taxes or subsidies. Tree-Planting Approaches The most common approach by governments and donor agencies to ameliorate perceived fuelwood problems has been to plant trees. Programs have ranged from establishment of large-scale fuelwood plantations near cities to establishment of village woodlots, subsidies for small-scale private fuelwood enterprises, and private initiatives and incentives for tree planting by small farmers. With planning that takes into account the opportunity costs of the land, land tenure problems, availability and accessibility of markets, and so on, tree planting can be an appropriate response to fuelwood shortages. Unfortunately, too often the decision to spend scarce revenue planting trees has been almost a knee-jerk reaction, taken without consider- ation of other options and the consequences of existing market and policy failures (Ryan 1990a). Three general approaches have been used: large-scale block plantations, rural or social forestry (agroforestry), and natural forest management. Large-scale plantations. The majority of large-scale plantations have used exotic, fast-growing species in state-owned and managed block plantations. Block fuelwood plantations were encouraged for a variety of reasons. Plantation forestry was considered a tested technology, understood and successfully applied by foresters in a variety of locations.2 Plantations were viewed as discrete, highly visible activities for which inputs and yields were easily quantified; as a result plantations lent themselves readily to standard economic analysis. Indirect benefits were thought to include employment generation, environmental protection, reduction of pressures on other forest lands, and demonstrations that governments were actively solving an important problem (Catterson 1984, cited in Freeman and Resch 1986). Unfortunately, the actual cost-benefit ratios usually fell far short of preproject expectations (Floor 1988; Leach and Mearns 1988). Frequently, the predicted benefits were overestimated and the costs underestimated. Because of the shortage of available land and competition with agriculture, fuelwood plantations were usually relegated to marginal lands where bio- mass productivity rarely produced revenues greater than the costs of the planting. French (1986) provides an excellent case study of the financial 2 Experience in Africa and Asia with plantation forestry was particularly signifi- cant (Freeman and Resch 1986). FUEL WOOD PROBLEMS AND SOLUTIONS 187 constraints to using state-run"plantations to solve deforestation and fuelwood problems in Malawi. 3 There have also been some successes. Two examples in Ethiopia are the 20,000 hectares of eucalypti around Addis Ababa and the 50,000 hectares around other cities where fuelwood has become so scarce that demand has driven the price well above replacement costs (World Bank 1986). Also, in the state of Minas Gerais, Brazil, private commercial wood energy planta- tions occupying over 200,000 hectares produce the bulk of the energy for the iron and steel industries, and in Rondoneia, Brazil, a private company sup- plies the city of Ariquemas with electricity produced with plantation-grown fuelwood. Similarly, in the Philippines a successful dendrothermal program was established in 1980 to produce electricity from fuelwood grown in small plantations by farmers (Gregersen et al. 1989). Failures, however, outnum- ber the success stories.' Social forestry. The failure of large-scale plantations led to the popularity of social, or rural, forestry projects emphasizing tree growing for fuelwood by small farmers in woodlots or in agroforestry systems. Because most fuelwood demand is associated with rural households, some observers (e.g., Gregersen et al. 1989) believe the key to solving the fuelwood problem can be found in encouraging farm families to grow enough trees to meet their own requirements and to generate surpluses for sale. The results with social forestry for fuelwood production, however, have generally not been encouraging (Floor 1988). Although there have been some spectacular successes in promoting tree planting by private farmers, particularly in India, the end products are usu- ally higher-valued building poles or pulpwood rather than fuelwood (e.g., see World Bank 1985; Arnold et al. 1987). This situation has led to reevalu- ation of some basic assumptions concerning tree planting and the fuelwood crisis. For example, the highly acclaimed and successful social forestry projects in the Gujarat state of India were based in large part on the assumption that any increase in tree planting .would diminish the pressure on the natural forest and thereby ameliorate the fuelwood crisis (World Bank 1990). The experience in India, however, indicates that this assumption was misleading and often incorrect. Indeed, small farmers planted many trees but few were for producing fuelwood; the stumpage price of fuelwood was too low to compete with products such as poles, pulpwood, or lumber. In some areas, the Indian social forestry projects actually reduced the availa- bility of biomass fuel to landless farm workers who had previously been allowed to collect the agricultural residues following harvests. When farms 3 Also see Freeman and Resch (1986) for an excellent analysis of a case study in the Bandia, Senegal. 4 This statement may apply only to donor- or government-related fuelwood plan- tations, because the successes and failures of private plantations are not well docu- mented. (K. Openshaw, World Bank, personal communication, 1990). 188 MANAGING THE WORLD'S FORESTS were converted to pulp and timber production, not only were these wage- earning opportunities eliminated, but the fuel derived from agricultural residues also was eliminated, creating potentiallly greater pressures on open- access woodlands.' A survey of farmers involved with social forestry projects in eight sites around the world-Bangladesh, Haiti, India (3), Indonesia, Philippines, and Thailand-by Energy/Development International (1986) for the U.S. Agency for International Development (USAID) reported the following findings: • Farmers base tree-planting and management decisions mainly on their expected profits from cash sales of wood. (This is true even of farm- ers who consume more of their tree products than they sell.) • Independent small farmers frequently fail to receive the high prices for trees that they had been led to expect. • In areas where tree growing is a mature, stable enterprise, farmers usually grow several species and produce a variety of products. These findings suggest that although the potential for social and agroforestry is very large, rarely will social forestry schemes based prima- rily on fuelwood production be successful so long as open-access forest resources are available and prices remain low. Where fuelwood-oriented programs have been successful, planners have actively sought advice and help from local communities and have taken into account different uses of trees in farming systems (Gregersen et al. 1989). It should be emphasized, however, that agroforestry and other social forestry initiatives have in general been quite successful (see Spears 1988). Many trees are being grown on farms, and agroforestry, if promoted cor- rectly, provides a means for farmers to stabilize agricultural output without using much artificial fertilizer. In addition, tree planting by small farmers provides considerable economic and environmental returns, such as soil conservation, shade, shelterbelt, and cattle forage. Natural forest management. Natural forest management systems have typically been neglected in favor of ambitious reforestation schemes. The current dissatisfaction with the results of fuelwood plantations has led to increased interest in managing the natural forest for fuelwood and other products. Furthermore, because a large proportion of the fuelwood supply for urban areas comes from woodlands, the proper management and regulation of natural woodlands will likely be essential to establishing a sustainable supply of fuelwood. Advantages to natural forest management include lower investment costs (e.g., no land·dearing, seeding, or planting costs) and the greater adaptation to local conditions. As has been noted, the problem with fuelwood plantations is that they generally supply only relatively low value fuelwood. The goal of natural forest management is to produce a sustained yield of forest products while maintaining ecosystem 5 A. Molnar, World Bank, Washington, D.C., personal communication, 1990. FUEL WOOD PROBLEMS AND SOLUTIONS 189 balance. Thus natural forests provide food, fodder, and other products as well as fuelwood. The natural forest approach recognizes that the fundamental resources (e.g., soil, water, vegetation, wildlife) must be managed as a whole rather than separately. Other benefits are the protection of biodiversity, watershed protection, and reduced soil erosion. USAID has refocused most of its fuelwood activities in Africa on natu- ral forest management, with notable success in Niger and other arid and semiarid Sahelian countries. 6 Studies around Niamey, Niger, for example, have demonstrated that appropriate harvest and management practices can substantially increase fuelwood production from natural forests in the Sahel. Heermans (1986 cited in Heermans and Minnick 1987) demonstrated that it may be possible to achieve growth rates for major fuelwood species almost twice current consumption rates. Although natural forest management for fuelwood production appears promising, considerable research is needed to evaluate its biological, economic, and social potential. Policy constraints to growing trees for fuelwood. Even when countries overcome the technical problems for growing sufficient amounts of fuelwood in large-scale plantations, in agroforestry systems, or under natural forest management systems, as long as enough "free" open-access forest resources remain to dominate market prices, tree-growing approaches are likely to be ineffective. Problems with land ownership and tenure, therefore, are the central issues for developing sustainable fuelwood resources under all three technical schemes. Three direct fuelwood supply effects can be attributed to the availabil- ity of open-access forests (Clarke and Shrestha 1989a): 1. People have no incentive to invest in restocking the biomass in open-access resources, because other people may expropriate the benefits from the investment. Even if property rights are enforced for privately owned tree stocks, the existence of nearby open- access forests tends to make commercial plantations or tree growing on small farms uncompetitive, because commercial tree growing must take into account not only the costs of harvesting but also the land, planting, and maintenance costs. 7 2. Open-access promotes underinvestment in the provision and acquisition of information concerning such things as efficient and sustainable harvesting· practices and techniques or the environ- mental impacts of deforestation. There is no incentive to discover the maximum levels of harvest to ensure sustainable fuelwood supplies or the critical minimum extent of tree cover to prevent environmental degradation. 6 Tim Resch, USAID and U.S. Forest Service, personal communication, 1990. 7 See French (1986) for a case study of these effects in Malawi. 190 MANAGING THE WORLD'S FORESTS 3. The lack of the ability to enforce property rights over forest resources contributes to the general misallocation of lands be- tween forest and farmland. Because individuals can usually reap all the benefits of land converted to annual agricultural crops but must share the benefits of open-access forest with the rest of the community, forests will be converted to private farmland even when its social value is greater as common property forest. Even if the problem of property rights is solved, approaches to increas- ing fuelwood supplies which involve encouraging small farmers to grow trees for fuelwood either in small woodlots or in agroforestry systems face serious constraints because of policies that distort prices in a variety of markets. For example, the existence of distorted prices in labor and nonlabor agricultural input, capital, commodity, and foreign currency markets pro- vides significant disincentives to tree growing for fuelwood by small farm- ers in many developing countries (Mercer 1990; Repetto 1988; Southgate 1988). Fuelwood Pricing Policies One of the major constraints to implementing effective tree-growing programs to increase the supply of fuelwood has been the low value of wood used as fuelwood. In many countries "free" open-access wood re- sources, the bulk of which are government owned, still account for 80 to 90 percent of the supply both for charcoaling and for direct fuelwood use (World Bank 1987). As long as this is the case, the stumpage price at those sites is zero, and market prices for fuelwood and charcoal will not reflect the full scarcity values (i.e., the full social value) of the wood (Hyde and Mercer 1990). Undervalued fuelwood resources may cause waste and inef- ficiency in production and consumption and disincentives for tree growing. In acknowledgment of this situation many observers (e.g., Openshaw and Feinstein 1989) argue that stumpage fees for harvesting fuelwood on government land should be raised at least to its replacement cost. Setting a stumpage fee at the same royalty rate typically applied to crude oil (about US$12 per barrel), for example, could produce revenue of about $2.5 million a year from a city of 250,000 that consumes a million bags of charcoal each year (Leach and Mearns 1988).8 Currently most countries charge some kind of stumpage fee for fuelwood. 9 Most stumpage fees, however, are nowhere near the social value of the in situ resource or the replacement cost, and the fees are rarely imposed on subsistence users. 8 These rates are about the same as for fuelwood produced from government-run plantations in Kenya and Tanzania (World Bank 1987 and World Bank/UNDP 1988). 9 Botswana is an example of a country that imposes no stumpage fee at all on fuelwood except in the Kweneng District (K. Openshaw, World Bank, personal communication, 1990). FUEL WOOD PROBLEMS AND SOLUTIONS 191 One of the main difficulties with stumpage-fee programs is determining the correct amount to charge. There are three principal methods: residual stumpage, alternative fuel substitute, and replacement cost. Implementing these procedures can be quite complicated and costly. (See Openshaw and Feinstein [1989] for a detailed discussion.) Even if the correct stumpage fee could be determined, however, imple- menting effective stumpage fees in places where most fuelwood is obtained from open-access wood resources poses several other administrative, insti- tutional, logistical, and political problems. There are practical problems and large administrative costs in countries where institutional capacity is al- ready weak. For example, in areas where fuelwood comes primarily from operations to clear agricultural land, it may be relatively easy to collect fees from large, commercial farming operations. Collecting stumpage fees from the myriad subsistence farmers who are responsible for the bulk of land clearing, however, would be extremely difficult and costly. Even assuming that the administrative and logistic problems could be overcome and that the government succeeded in raising the price of fuelwood enough to make growing trees for fuelwood a viable option, the distribu- tional effect on low-income urban households could create severe political problems. For example, in Kenya, the World Bank (1987) estimated that a stumpage fee based on replacement values would increase retail prices by as much as 30 percent. Given the fact that any benefits associated with a stumpage fee might not materialize until well into the next century, it is highly unlikely that the political will could be mustered to pass and enact such significant increases in stumpage fees. Furthermore, it is uncertain whether stumpage fees would actually increase private tree growing or enhance the viability of government plan- tations. The situation would depend on such factors as land tenure, the opportunity returns of other land uses, and transportation costs (Hyde and Mercer 1990). Simply setting stumpage fees at the replacement cost in no way guarantees that growing trees for fuelwood would be profitable. Finally., obtaining information to estimate costs and benefits of stump- age fees would be difficult and costly and any potential welfare gains would probably be offset by the costs. In many cases implementing effective stump- age fees might well be more expensive than directly enforcing property rights (Clarke and Shrestha 1989b). However, implementation of stumpage fees has proved successful when enforcement and collection are ceded to local communities that share in the benefits. For example, Leach and Mearns (1988) describe the successful program in Rwanda. Demand-Side Policies Demand-side policies can be separated into two main groups: (1) poli- cies designed to reduce demand by promoting efficient use of wood through improved cook stoves and charcoal kilns; and (2) policies to reduce fuelwood demand by subsidizing the substitution of modern fuels such as kerosene 192 MANAGING THE WORLD'S FORESTS and liquefied petroleum gas. Unfortunately, the experience to date with demand-side policies has been discouraging. Scant evidence exists to sug- gest that the programs have changed demand patterns substantially; some programs (e.g., subsidizing modern fuels) have been expensive failures, the main effect of which has been to provide inexpensive energy to people who least needed it. Increasing Efficiency of Use Fuel-efficient stoves. Programs to design and propagate improved stoves were viewed with great enthusiasm in the 1970s and 1980s. It was believed that if improved stoves became widely used,. firewood consumption and hence deforestation would fall and opportunities for higher incomes would increase because people would spend less time cooking and collecting fuel. In addition, improved stoves were expected to generate dramatic health benefits by reducing indoor air pollution. Rural households were targeted for the new stoves because they typically consumed the bulk of fuelwood and were perceived to need the most help in saving the surrounding forests (Foley 1988). During the mid-1980s, however, improved-stove programs fell into gen- eral disfavor among the donor community as a result of critical reviews by Fdley and Moss (1983) and Gill (1987), which questioned whether improved stoves actually performed any better than the traditional stoves they were replacing. 10 Kirk Smith (1983) added fuel to the fire with his studies indicat- ing that the new stoves also failed to alleviate significantly the ill-health effects associated with indoor air pollution. More recent studies of perfor- mance under normal operating conditions, however, suggest that the new stoves are capable of achieving fuel-use efficiency 30 to 50 percent better than that of traditional stoves (Joseph 1987; Floor 1988; World Bank 1987). The disappointing results of most of the early stove programs can be more accurately attributed to putting "the wrong thiing (heavy-mass mud stoves), in the wrong place (rural areas), in the wrong way (self-built stoves), and with the wrong people" (Floor 1988, p. 128). In contrast, the recent successful programs in Kenya, Niger, Burundi, Rwanda, and Harare have used commercial, charcoal, metal stoves in urban areas. Indeed, in Nairobi it is reported that 80 percent of the traditional stove makers are now producing the new stove with no government or other donor assistance (Foley 1988). In Colombo, Sri Lanka, a ceramic wood- burning stove has been successfully introduced. In all these cases, improved 10 Generalizations about improved-stove programs have ranged from "the major- ity ... were more or less a complete flop" (Foley 1988) to "results ... have been disappointing" (Floor 1988) to "failed to displace traditional modes of cooking to any significant extent" (Gill 1987). FUEL WOOD PROBLEMS AND SOLUTIONS 193 stoves have saved wood resources and cash for the urban poor, and have produced a healthier household environment. They show that improved stoves may be successful when fuelwood is a commercial good produced and sold through the private sector, and when fuel savings mean cash savings. Stove programs are more likely to be successful in urban than rural areas for a number of reasons. First, there are many substitutes for fuelwood in rural areas, and biomass fuels are rarely perceived as scarce (Dewees 1989), at least by men. Women may have very different perceptions. Second, because cash is scarcer in rural areas, the people there have a much lower ability to buy the improved stoves. In many rural communities, labor is in surplus except during planting and harvesting times, so the market value of an individual's time may be close to zero. Therefore it may well make sense for households to use fuelwood inefficiently rather than to spend their scarce cash on improved stoves. Furthermore, the relationship between improved stoves and reduced pressures on forest stocks may be tenuous at best. Clarke and Shrestha (1989a) showed that only when the demand for fuelwood is very price inelastic will improvements in conversion efficiencies reduce the pressures on forest stocks. However, if demand for fuelwood is elastic (as is probably the case in rural areas where there are many substitutes), efficiency im- provements in wood consumption which result in lower prices per unit of energy may provide incentives to harvest existing forest stocks more inten- sively. When energy demand is highly income elastic, if improved fuel-use efficiency frees household members to pursue wage labor or other income- producing activity, use of fuelwood may actually increase. Foley (1988) sug- gests that improved stoves may even delay the shift to modern fuels and hence make fuelwood consumption higher than it would otherwise have been. At this time it is difficult to generalize about the benefits to be obtained from improved stoves. Improved stoves apparently can contribute to solv- ing fuelwood problems, but the stove programs probably should be limited to urban areas. As Foley (1988, p. 72) stated, "As far as demand manage- ment is concerned, improved stove programs are a venture into the un- known." Improved charcoal kilns. Efforts to improve charcoaling methods have been under way for a long time. Improved mound kilns were introduced in India in 1884 and portable steel kilns in 1891 (Cleghorn 1884 and Fernandez 1891, cited in Foley 1988). The Mark V portable steel kiln was introduced in Uganda in the 1960s, the Tropical Products Institute kiln was introduced in a variety of countries in the 1970s, and the Casamance kiln was developed in Senegal in the 1980s (Foley 1988). The usual objective of these efforts is to reduce pressures on forest resources by reducing the amount of wood required to produce a given quantity of charcoal. As long as open-access forest resources exist and charcoalers obtain "free" wood, however, efforts to reduce wood consump- 194 MANAGING THE WORLD'S FORESTS tion by introducing improved kilns are unlikely to succeed. Under open- access situations, there are virtually no incentives for charcoalers to econo- mize on wood use if the cost of the kiln is more than the increased fuelwood harvest costs using inefficient kilns. If collection costs are very small or close to zero, as in open-access situations, there would be little incentive to switch to the improved kilns. Clarke and Shrestha (1989a) found the results for the effect of kilns similar to those for cook stoves. If demand for charcoal is inelastic, im- proved kilns increase revenue from a unit of fuelwood and reduce the ag- gregate costs of wood collection per unit of charcoal, thereby increasing the availability of forest stocks. If demand is elastic, however, improved kilns are likely to reduce forest stocks by increasing the areas that are economi- cally accessible. Where forests are underused and the main problem is ac- cess, this outcome may be desirable because it achieves increased harvest levels with reduced effort. But if the intent is to reduce pressures on over- used open-access forests, charcoal kilns may be counterproductive unless they are accompanied by land tenure reform or other programs to reduce access. In Brazil, the introduction of improved kilns has been very successful in areas where a large portion of the wood for charcoaling is grown and owned by individuals, corporations, or the state, but the efforts have been unsuc- cessful in open-access areas. The successful implementation of improved kilns in Uganda in the 1960s was due to its combination with an effective program to reduce access and to control harvest levels (Foley 1988). Substitution of Modern Fuels Increased reliance on modern fossil fuels and electricity rather than on fuelwood usually accompanies economic development. As national income rises, more and more people are able to pay for the increased cleanliness, convenience, and efficiency associated with modern fuels. For example, fuelwood typically makes up 60 to 95 percent of total energy use in poor developing countries, 25 to 60 percent in middle-income countries, and less than 5 percent in high-income countries (Leach 1988; Leach and Mearns 1988). This energy transition is also central to any understanding of urban fuelwood use. As urbanization proceeds, households tend to use more en- ergy and to switch from wood and charcoal to modern fuels. The Energy Sector Management Assistance Program (ESMAP) (1990a) has discovered this pattern in Indonesia, and Soussan et al. (1990a) discuss a number of other examples. Some countries have initiated policies to encourage a more rapid tran- sition. In most .cases these policies have consisted of price subsidies for modern fuels. To encourage consumers to switch to modern fuels, Indone- sia instituted kerosene subsidies as high as 80. percent of the international FUEL WOOD PROBLEMS AND SOLUTIONS 195 market value. In hopes of reducing charcoal consumption in Dakar, Senegal initiated a massive campaign to subsidize butane. Sri Lanka provided poor families with stamps that could be traded for kerosene (Foley 1988). Thai- land and India have provided subsidies for biogas plants (Clarke and Shresthra 1989b). Unfortunately, most of these policies have been expensive failures. For example, Indonesia's kerosene subsidy reduced the retail price of kerosene to 18 percent of its international price-at a cost of $3.7 billion (roughly 5.4 percent of the GDP) in 1980-81 (Pitt 1985). Household surveys indicate that the main beneficiaries were middle-income households, which would have used kerosene at its nonsubsidized price, and the transportation industry, which substituted kerosene for diesel fuel. There is no evidence to indicate that the subsidies promoted decreased fuelwood consumption or deforesta- tion. In Senegal, subsidization did result in increased consumption of bu- tane, but no evidence exists that charcoal use was significantly affected. Furthermore, a substantial fraction of the subsidized butane was exported to Mauritania. This program also appears to have primarily benefited the middle class, which would have been able to afford the nonsubsidized price (Foley 1988). It is clear that designing effective policies to encourage substitution of modern fuels for fuelwood requires clearly articulating the goals of the policy and understanding the driving forces behind the transition. Goals of interfuel substitution policies typically include reducing high fuel costs for the poor, increasing consumer welfare, saving fuelwood, and protecting the environment. If the goal is to reduce the environmental damage associated with deforestation, success will depend on the source of fuelwood. For ex- ample, if the bulk of fuelwood comes from clearing land for agriculture, subsidizing modern fuels will have little impact. If a large proportion of demand is made up of charcoal produced in open-access forests, substitu- tion policies may well produce environmental benefits, but only if fuelwood use is the chief activity associated with forest exploitation. A variety of household surveys have shown that the two main forces driving the transition from fuelwood to modern fuels are access to depend- able supplies of the modern fuels and income. The positive relationship between fuel choice and income is well established (Alam et al. 1985; Dowd 1989; Fitzgerald et al. 1990; Dunkerley et al. 1990; Leach 1987; Leach and Gowen 1987; Leach and Mearns 1988). The higher the income, the greater the likelihood to choose modern fuels over biofuels. As urban incomes rise, people typically progress from low-grade biofuels (e.g., crop residues, scrap timber, tires) to firewood, charcoal, kerosene, LPG, and electricity or natural gas (Leach and Mearns 1988). In an analysis of fuelwood and modern fuel prices in 19 developing countries, Barnes (1986) reported trends suggesting that rising incomes rather than prices are the predominant factor for fuel substitution and that reduc- ing the prices of substitute fuel (e.g., kerosene) does not reduce fuelwood 196 MANAGING THE WORLD'S FORESTS prices (or demand) unless incomes simultaneously increase or are redistrib- uted toward lower-income groups. This study suggests that subsidies for modern fuels are likely to encourage increased consumption among the high- and middle-income households while having negligible effect on the fuelwood consumption of the poor. Access to dependable supplies of modem fuels is the other driving force behind transitions between fuelwood and modern fuels. Wood and charcoal are often used not because they are cheap, but because they are available in places and quantities that fit in with the life of the urban poor. The access issue has two basic components: differential access between ar- eas (e.g., rural versus urban, small cities versus large cities) and differential access within cities. In rural areas the vast majority of households, even the highest income classes, depend on biofuels for the bulk of their energy needs; supplies of alternative fuels are nonexistent, insufficient, unreliable, or very expensive (Leach 1988). The supply of modern fuels increases as one moves from rural villages to small and medium-size towns and then to large urban areas. For example, in India in 1979, 40 percent of household energy was supplied by modern fuels in towns with populations of 20,000 to 50,000, compared with 58 percent in cities with populations of 200,000 to 500,000 and 75 percent for the largest urban areas (Leach 1987). Access within cities also can affect fuel substitution behavior. Accessi- bility frequently outweighs all other issues, including stove costs and prices, for low-income classes. For example, Leach (1988) discusses a survey in Lucknow, India, which revealed that few of the poor used kerosene, despite the fact that it cost 40 percent less than fuelwood. Major deterrents to kero- sene use were shortages and long lines, rather than the price of stoves or other factors. Factors that limit the effectiveness of policies to promote fuel switching include the paucity of data on demand and supply elasticities, the require- ment for additional foreign exchange, kerosene pricing policies, problems of access in rural areas, inability of poor people to afford the required tech- nologies, and lack of perception of fuelwood crisis by local people. Subsidizing substitute fuels in order to conserve on fuelwood will be cost-effective only when fuelwood is cross-price elastic with respect to the substitute fuel and the demand for the substitute is own-price inelastic (Clarke and Shrestha 1989b). For example, in Indonesia, Pitt (1985) estimated that demand for both fuelwood and kerosene was own-price elastic and cross- price inelastic, resulting in a costly failure (as already noted). However, good elasticity estimates are exceedingly rare in most developing countries. Without these, promoting subsidies for alternative fuels would be shooting in the dark. The requirement for additional foreign exchange is often cited as a se- rious drawback to promoting substitution of modern fuels, particularly for low-income oil-importing countries. Exhibit 8-1 shows the effects of substi- tuting kerosene for all charcoal and fuelwood used in the cities of low- and FUEL WOOD PROBLEMS AND SOLUTIONS 197 EXHIBIT 8-1. Amount of Kerosene Required to Replace All Urban Fuelwood Consumption in Low- and Middle-Income African Countries, and Resulting Increase in Demand for Petroleum and Decrease in Demand for Fuelwood, Selected Years, 1978-1989 Amount of Kerosene Increase in Decrease in to Replace All Total Petroleum Total Fuelwood Urban Fuelwood Consumption Consumption Country" Yeara (103 lons) (%) (%) Low-Income Benin 1983 40 34 26 Burundi 1980 15 46 23 Ethiopia 1982 170 32 26 Ghana 1985 180 26 44 Guinea 1984 110 29 23 Kenya 1985 335 20 29 Malawi 1980 40 29 5 Niger 1980 35 23 20 Nigeria 1980 405 5 12 Rwanda 1978 13 27 8 Sierra Leone 1986 60 34 37 Somalia 1984 75 41 32 Sudan 1980 435 45 32 Tanzania 1981 470 74 29 Togo 1981 40 17 25 Uganda 1982 80 51 15 Zaire 1980 540 91 37 Middle-Income Congo 1985 32 13 50 Ivory Coast 1982 135 14 35 Liberia 1983 50 12 40 Mauritania 1984 13 8 40 Senegal 1981 55 8 27 Zambia 1989b 227 40 21 Zimbabwe 1980 65 11 18 a Byer (1987) in Leach and Mearns (1988), based on ESMAP Country Energy Sector Assignments. bK. Openshaw, World Bank, personal communication, 1990. middle-income African countries. For all the middle-income countries ex- cept Zambia and including Nigeria, replacing all urban fuelwood with kero- sene would raise the oil demand (and the accompanying need for additional foreign exchange) by only.5 to 16 percent, while reducing fuelwood con- sumption by 18 to 50 percent. Petroleum demand in the low-income coun- tries would increase by 17 to 91 percent, and fuelwood demand would decrease by only 8 to 44 percent. These numbers indicate that for those countries where fuelwood use could be greatly reduced without significant increases in oil demand, the substitution option may be appropriate. Whether these percentage changes can be achieved, however, depends on a variety of 198 MANAGING THE WORLD'S FORESTS demand factors. Development planners should analyze the driving factors for fuel demands and reliable elasticity estimates for each local area before governments make large expenditures. Probably the most serious constraints are the distorted fuel markets in most developing countries. Many urban fuelwood markets are highly de- veloped and effective distribution systems. In contrast, the distribution sys- tems for commercial fuels such as kerosene and LPG often are characterized by shortages and black-market pricing and tend not to reach many periph- eral areas where the poor live. Soussan (1990b) provides an example of the effects of poorly developed markets for modern fuel in Mogadishu, Soma- lia. If planners wish to encourage the transition from fuelwood in urban areas, they must take action to ensure that viable alternatives are available to low-income urban consumers. Without such action there is little hope of reducing the use of wood and charcoal in urban areas. POLICY RECOMMENDATIONS Although specific activities will vary from place to place, experience in a wide range of fields suggest four interrelated policy recommendations for governments in every region experiencing fuelwood problems: 1. Improve the information base on which policies are based. 2. Correct market failures and improve :functioning of markets. 3. Develop fuelwood sector strategies. 4. Strengthen fuelwood planning institutions. In some situations, these reforms may enable markets to solve both long-term and short-term fuelwood allocation and production problems without further central government involvement. In other cases, central government planning and allocation may be required in the short term to assist in the transition. Improve Information Probably the single most important step a government can take to im- prove its fuelwood situation is to improve the information base on which policies are formed. Among the most pressing needs are information about . household demand and supply, modern and fuelwood markets, the fuelwood resource base, and forest management and tree-growing systems. • Household Demand and Supply Information. Rigorous estimates of de- mand and supply elasticities are almost nonexistent. Until we can more confidently predict how households will respond to changing parameters, fuelwood policy interventions will be haphazard. Any successes will be fortuitous. FUEL WOOD PROBLEMS AND SOLUTIONS 199 • Market Information. In-depth surveys of modern and fuelwood mar- kets are needed to ascertain how the fuels are supplied, distributed, marketed, and priced in urban areas (and in those rural areas where the fuels have been commoditized). In addition, information is needed on land, labor, agricultural, and other forest product markets in order to identify market and policy failures and to estimate the potential for fuelwood policies to augment supply. Nontraded markets-the amount of time rural households spend collecting fuelwood and crop residues-also should be investigated in the surveys. • Fuelwood Resource Base. Lack of reliable data on fuelwood resource stocking and annual sustainable yields of the local and regional fuelwood catchments for rural villages and urban areas is a major constraint to planning effective fuelwood policies. This information should be coupled with information on household demand and sup- ply in order to identify the nature of the fuelwood problem for spe- cific fuelwood catchments and to tailor policies and programs to local needs. • Forest Management and Tree-Growing Systems. Forest management and tree-growing systems for fuelwood production need further develop- ment and testing. Block fuelwood plantations are well understood, but systems of managing the natural forest for fuelwood and other products need to be developed and tested for biological and eco- nomic productivity potentials in a variety of ecosystems and agroclimates. In addition, agroforestry systems that produce signifi- cant quantities of fuelwood as a by-product need refinement. Finally, methods to use waste by-products from wood production systems unrelated to fuelwood need to be developed and tested. Correct Market and Policy Failures The next step in formulating effective fuelwood policies is to identify the underlying causes of fuelwood problems in specific local areas. Identi- fying the underlying causes of fuelwood problems requires an examination of their economic manifestations. Unfortunately, past fuelwood policies have often been formulated on analyses based on the observable symptoms (physi- cal manifestations) rather than the underlying causes (economic manifesta- tions) of the problem (Panayotou 1989). Many economic manifestations of fuelwood problems can be traced to market and policy failures that produce situations in which benefits are disassociated from costs, prices from scarcity, rights from responsibilities, and actions from consequences. For example, fuelwood harvesters (both commercial and subsistence) are able to extract the benefits from "free," open-access fuelwood resources while shifting the costs of depletion to fu- ture generations. As a result, prices (explicit and implicit) do not reflect the growing scarcities and rapidly increasing social cost of fuelwood depletion. 200 MANAGING THE WORLD'S FORESTS Many fuelwood-related problems can also be traced to policies that intentionally or unwittingly distort prices (incentives). Distorted prices (from both market and policy failures) fail to provide the signals that, in a well- functioning market, would promote the increased conservation, substitu- tion, innovation, and efficiency necessary to bring fuelwood supply and demand into balance. The most important market and policy failures are ill-defined, attenu- ated, unenforceable, or undefined property rights; unpriced or underpriced resources and absent or thin markets; and policy-induced price distortions in capital, labor, and commodity markets. Each of these failures is briefly examined in the paragraphs that follow. Property Rights To achieve efficiency, property rights must be well defined, exclusive, secure, enforceable, and transferable. The lack of one or more of these char- acteristics in local land markets is probably the single most important cause of fuelwood-related problems. Property rights problems are also the root cause of the environmental problems and fuelwood scarcity associated with deforestation, and ill-defined property rights provide perverse disincentives against tree growing. The open-access nature of forestlands in developing countries is probably the most significant impediment to solving fuelwood- related problems.11 Private property rights that are ill-defined or attenuated also contribute to fuelwood scarcity and provide obstacles to implementing successful fuelwood projects, especially those based on encouraging small farmers to use agroforestry systems or otherwise grow trees for fuelwood. Insecure land tenure reduces investment incentives and encourages preference for current consumption over future consumption, (Feder et al. 1988). A number of recent attempts to move land tenure from centralized control to more local or private control have demonstrated the efficiency gains that are possible. For example, Spears (1988) reports a number of 11 For a review of the theory of open-access resource use and abuse, see Magrath (1989). The distinction between open-access and common property resources is not clear in the literature. This chapter uses the definitions of Bromley and Cernea (1989), who defined common property as, in essence, private (or corporate) prop- erty held by a group. All individuals within the group have rights and duties with respect to the property, and nongroup members are excludable. Open access, in contrast, represents situations in which there are :no property rights, no one can be excluded, and no management systems can be enforced. Common property rights may be fully articulated and enforced by cooperative group action to ensure effi- cient management and use. Problems arise when the institutions for enforcing com- mon property rights and management break down. See Magrath (1989) and Runge (1981) for examples. FUEL WOOD PROBLEMS AND SOLUTIONS 201 instances in the Philippines and India when reallocating publicly owned lands to private farmers and local communities increased reforestation ac- tivities. The specific property rights regime that should be implemented, however, will vary for individual locations and situations. In many instances, the best solution may be privatization, especially when the land and pro- duction systems warrant individual or family investment. Privatization may also be best .when shifting cultivation or other agricultural practices have already resulted in de facto privatization. Privatization, however, may not be best when the land area is extensive and cannot be protected from out- siders, when the value of production is too low to warrant individual in- vestments, and when products are diverse and used by a variety of people (but not necessarily cash-valued highly). In small, self-sustaining rural communities where strong traditions of community or tribal management of resources exist and where population and other external pressures are mild, community management of fuelwood resources may be appropriate. Indeed, traditional common property sys- tems have been used successfully throughout history to manage resources on a sustained basis (Ciriacy-Wantrup and Bishop 1975; Runge 1981). Gov- ernment ownership and control of fuelwood resources also can be efficient under certain conditions, as when the country has a strong, well-staffed forestry department with a history of successful management and control of forest resources, and either a relatively small forest resource base to manage or relatively low pressures on the forest resources. The issue is not so much that any particular group should manage the resource but that the group should be able to restrict access and have the necessary human resources available to manage the forest, and that either the land or the proceeds from the land should be distributed equitably among the affected populations. Unpriced or Underpriced Resources and Nonexistent or Thin Markets Because open-access forest resources have no exclusive owners to de- mand a price and deny access when that price is not paid, there are no markets and no prices for these resources. The implicit price of the forest itself is taken to be zero, regardless of the scarcity and social opportunity cost. When open-access forests account for a significant portion of the fuelwood catchment area, markets for fuelwood may exist but the price of fuel wood will only reflect the opportunity cost of labor and capital used in its production and transportation. The price does not include the opportu- nity cost of the scarce natural resource (forests) used in its production, and does not reflect its true social value. Therefore, the fuelwood price that consumers face does not reflect the growing overconsumption and underinvestment in supply. The absence of fuelwood and labor markets in rural communities also contributes to fuelwood problems. Many fuelwood projects have failed be- 202 MANAGING THE WORLD'S FORESTS cause of their reliance on the assumption that fuelwood and the labor used for fuelwood collection have market values, that is, that fuelwood collection time can be converted to income by selling the fuelwood or the labor itself (Clarke and Shrestha 1989a). When fuelwood is primarily a nonmarket sub- sistence good and the value of fuelwood collectors' time is very low, people will continue to use fuelwood if they have to pay cash for alternative fuels. Inadequate markets and supplies of modern fuels in both urban and rural areas often constrain fuel substitution. For low-income households, security of access is more important than price or convenience. Rigid gov- ernment controls on operating oil companies and, where applicable, state- controlled oil monopolies must be removed. In many countries, government policies severely distort kerosene and oil prices, discriminating against low- income consumers and rural areas. These distorting policies should be re- moved and incentives installed to encourage oil companies to distribute and market kerosene in these areas. Price Distortions in Capital, Labor, and Commodity Markets Kerosene pricing policies are one of the most common energy policy failures in developing countries (World Bank 1987). These policies directly compound the difficulties in supplying the urban poor and rural areas with kerosene, the most important potential substitute for fuelwood. Cross- subsidization from gasoline and diesel oil, combined with price controls, results in lower retail prices for kerosene than other petroleum fuels, de- spite the fact that all the import prices are roughly equivalent. 12 The end result is a strong financial incentive at both the supplier and the consumer levels to divert kerosene to the diesel market and away from household energy markets. Moreover, many countries require uniform national pricing of petro- leum products. Although there is usually an attempt to incorporate trans- portation costs in petroleum prices, the pricing structure often gives petro- leum suppliers little or no incentive to market their products in rural areas. Together, petroleum pricing problems and the structure of the kerosene market (in which the poor must buy from secondary retailers or bulk break- ers) have discouraged supply, aggravated scarcity, and raised the secondary retail prices of kerosene. In Kenya, for example, secondary retail prices have been observed to be as much as 75 percent above the primary retail price in urban areas and 100 to 200 percent in rural areas (World Bank 1987). Efforts to increase fuelwood supplies by encouraging small farmers to grow trees either in small woodlots or in agroforestry systems are gaining in popularity. However, the distorted prices in labor and nonlabor agricul- tural input, capital, commodity, and foreign currency markets provide 12 Retail kerosene prices are often 25 to 40 percent lower than diesel oil prices (World Bank 1987). FUEL WOOD PROBLEMS AND SOLUTIONS 203 significant disincentives to tree growing for fuelwood by small farmers (Mercer 1990; Repetto 1988; Southgate 1988). Many of the market failures (especially those arising from insecure prop- erty rights and unpriced resources) directly result from the failure of gov- ernments to establish the legal foundations of markets. The existence of market failures does not necessarily justify eliminating markets or other government intervention. Policy intervention is justified only when the en- suing benefits exceed the costs of intervention. Develop Fuelwood Sector Strategies Once planners have dealt with the prevalent market and policy failures, they can formulate fuelwood sector strategies to manage supply and de- mand. In most countries, planning should be based on the realization that there is rarely a single fuelwood sector. Rather, a variety of fuelwood de- mand centers usually exist with their own specific fuelwood catchment ar- eas, which may or may not overlap. In order to determine the most efficient balance in consumption between modern fuels and wood, planners should identify the actual and potential fuelwood growing stock and the produc- tivity and accessibility (economically and environmentally) of each catchment area. Once this situation has been assessed, the most appropriate combina- tion of demand and supply policies (for each catchment area) can be formu- lated. Supply and demand policies are reviewed in the paragraphs that follow. It should be noted, however, that because the fuelwood sector is only one part of a dynamic household energy system, successful policy intervention requires a strategic planning approach. Supply Management The extent of forest cover that needs to be managed to produce a sus- tainable (economically and environmentally) wood supply must first be determined. Planners can then formulate policies to provide the correct in- centives to ensure sustainable production. The available strategies remain encouraging tree growing or improved forest management, or imposing corrective taxes or stumpage fees. Once property rights are secured and access to "free" resources is prohibited, fuelwood prices will adjust (prob- ably upward). If fuelwood prices rise high enough in comparison with other opportunities, land owners (private or communal) will have sufficient in- centive to plant trees or otherwise manage forests for fuelwood production. If the prices are not high enough to encourage tree growing and if positive environmental effects associated with tree-growing for fuelwood are found to exist, production incentives such as subsidies, extension services, and demonstration projects may be required. Multiproduct agroforestry systems and natural forest management for multiple use under private or commu- nity control are usually the most efficient systems. Because the bulk of com- 204 MANAGING THE WORLD'S FORESTS mercial fuelwood for the foreseeable future will come from natural forests, proper management systems for woodlands will be essentiaL The other major policy to augment supply-stumpage fees-will be successful only when access to the forest resource can be limited. In general, given the inherent political, administrative, and institutional difficulties, attempts to collect stumpage fees are not recommended to central govern- ments. Developing alternative innovative tax systems such as the Malawi kiln tax and solving the open-access problem should come first. In this regard, however, transferring land tenure rights to local communities, ei- ther permanently or in long leases, may prove to be an effective system for collecting the fees. To be successful, local communities or their representa- tives need to have a direct financial interest in collection of the fees. When local communities act as mediators between forest harvesters and the forest service or directly set and collect the fees, the costs may be considerably reduced and the prospects of enforcement enhanced. Demand Management There are two basic approaches for managing demand: speeding up the transition to modern fuels and increasing the efficiency of the use of fuelwood. As already discussed, urbanization and rising urban incomes drive the tran- sition to modern fuels. An unfettered supply system that can guarantee access to the modern fuels in the quantities demanded is essentiaL Experience suggests that subsidization o:f modern fuels is an ineffective way to influence fuelwood consumption, The more appropriate way to en- courage fuel switching is to improve access to the fuels. Sometimes govern- ments fail to provide the institutions, infrastructure, and policy framework to encourage accessibility of modern fuels to low-income groups and rural areas, and sometimes government policies directly distort the incentives for oil companies to provide access. After these policy failures are corrected, several policy approaches remain available. Governments can improve the transportation infrastructure, especially to smaller towns and rural villages. Governments can help increase the supply of LPG cylinders to low-income families and provide incentives to expand relail outlets in low-income areas and rural areas. And when the cost of the modern fuel stoves is a major constraint, governments can provide incentives (e.g., loan programs) or subsidies to help low-income families purchase the stoves. The promotion of efficient firewood and charcoal stoves also has a place in demand management policy, but careful planning and assessments are required before ambitious programs are undertaken. Targeting urban mar- kets is generally most effective. Consumers must be convinced that the stoves will reduce their expenditures on fuel and investments of time, improve the kitchen environment, and reduce disease due to indoor air pollution. Exten- sion efforts will be needed to train consumers and stove producers to pro- duce better stoves; in addition, marketing and distribution networks must FUEL WOOD PROBLEMS AND SOLUTIONS 205 be developed. If stove makers see large profits from producing fuel-efficient stoves and consumers perceive substantial benefits from using improved stoves, urban-market-oriented stove programs will be successful. Strengthen Fuelwood Planning lnstituti.ons A critical issue for fuelwood policy formulation is the performance of the institutions that are involved in forming and implementing biomass energy policy. Unfortunately, because fuelwood does not fall clearly under the jurisdiction of a single ministry, particular attention must be paid to the institutional arrangements that address fuelwood issues. The problem of coordination between different state agencies is exacerbated by the lack of effective channels for the involvement of local people in the planning and implementation process. Many of the policies that appear to offer the best hope for dealing with fuelwood problems are based on the premise that local people will actively participate in the design and execution of projects. However, it is far from clear how this widely advocated participation can be translated into effective and durable institutional structures. The following basic principles are essential for reform of fuelwood plan- ning institutions: • The institutions should be responsive to energy needs. Energy produc- tion capabilities should be based on defined needs of target groups. • The institutions should contain effective channels for the participa- tion of fuelwood users and providers in all stages of planning for local communities. • The institutions should permit multisectoral cooperation. Although the energy and forestry ministries are expected to continue to take a leading role in the planning process, other institutions (particularly those with extension capabilities) may be the most appropriate agency for implementing plans. • The principle of sustainability in environmental, economic, and insti- tutional terms should be fully integrated into the procedures of energy planning institutions. In many cases the public has accepted policies for sustainable energy development, but governments have paid in- sufficient attention to changing the operations of implementing insti- tutions to account for the new policy directions. • In some cases, reforming forestry institutions by creating self-financ- ing forest commissions to supervise commercial operations may in- crease efficiency and effectiveness. In addition, some funds may need to be funneled to an agency that would help rural communities and individuals manage nongazetted forests, provide extension services to farmers on tree-planting and agroforestry techniques, and provide information on the market for forest products. 206 MANAGING THE WORLD'S FORESTS • The government should give control over local resources to local communities and support other local initiatives to increase forest re- sources. • Positive action is needed to create effective management structures and to enhance management skills in parastatal, private, and nongovern- mental organizations as well as government agencies. • Fuelwood planning should become more flexible. Indigenous technical knowledge should be incorporated into the planning to complement the conventional expertise. of national and external personnel. The new policy approach concerns implementation as much as it does principles for sustainable planning. National governments should demon- strate a desire to create these mechanisms for implementation, perhaps with donor support. SUMMARY The fuelwood debate demonstrates that forest-sector policies must ad- dress events far beyond those inside forest areas. Indeed, events outside the forest, in agricultural areas and cities, often dictate the circumstances that create the fuelwood problems and condition the solutions to the problems. Nearly all fuelwood is consumed outside the forest, and a large proportion is produced outside the forest from trees and plants in the agricultural landscape. Similarly, many policy options (including some, such as urban fuel switching, which seek to address problems in woodland areas) are focused in nonforest areas. Fuelwood problems and policies thus should be seen as part of a wider land resource management policy. From this viewpoint, the links between forest policies and policies in other spheres-energy, agriculture, and fiscal, for example-are a central issue. Fuelwood problems cannot be addressed by only one set of profes- sional experts working in one ministry. Effective intersectoral institutional links must be established before viable fuelwood policies can be created and implemented. These links will provide the context in which effective policy reforms can be introduced. Many of the technical inlterventions will prove ineffec- tive without the introduction of policy reforms to permit effective markets to operate, to produce prices that reflect real costs, to change land tenure relationships, and to generate community participation. If the mistakes of the past are to be avoided, the government must review the proposed policy changes before making decisions about project investments. Fuelwood problems and intervention opportunities are highly variable and specific to localities. There is no generic approach to these issues, and individual activities need to be incorporated into a strategic approach that sets the overall policy direction for a country or region, but has the potential to account for locally specific circumstances when detailed, local-level projects are designed. FUEL WOOD PROBLEMS AND SOLUTIONS 207 In addition, fuelwood policies need to be set within the context of a sustainable development approach; the need for small-scale, long-term ac- tivities must be recognized, and principles of local control and participation enshrined in the planning process. Donor organizations must address the procedural reforms needed to permit the funding of small-scale projects over far longer time horizons and with greater devolution of control over project decisions. Small-scale in this sense refers to individual components of projects. The total effort for a country or region need not be small, but ways should be sought to encourage and support numerous small-scale activities by urban and rural communities and entrepreneurs. ANNEX A BRIEF TYPOLOGY OF FUELWOOD SITUATIONS Fuelwood stress needs to be seen as part of a wider resource manage- ment crisis: it cannot be separated from the general development of specific localities. The same is true of solutions to these problems; generic approaches to planning are doomed to failure because they miss the diversity and com- plexity of these problems. This annex advances a typology of fuelwood situations based on the environmental and development characteristics of broadly similar regions within the third world. The criteria for developing this typology are as follows: • The biomass resource potential of different areas, equating to agro- climatic zones; • The characteristic rural economy; • Constraints on access to the resource base for fuel use by different sections of the local community; • Exports of wood resources from the areas; and • Structural changes affecting the fuelwood situation in a locality. The following paragraphs run through a list of characteristic fuelwood situations, based on these factors. 1. High Woody Biomass/Low Population Density: These areas consist mostly of moist tropical forests and woodlands with pockets of clearance but large areas of climax vegetation, such as Amazonia, Zaire Basin, Indonesia, and isolated areas of Southeast Asia, West and southern Africa, and tropical islands. The potential supply of fuelwood in these areas is greater than demand, but fuelwood stress can be associated with large development and sudden popu- lation influxes. Localized problems can be addressed through the management of natural woodland areas or the encouragement of multipurpose tree cultivation on private farmland. 2. High Woody Biomass/Medium to High Population Density: These areas typically have a complex mosaic of land uses; intensive farm- ing may coexist with vestigial woodland areas. The areas include 208 MANAGING THE WORLD'S FORESTS coastal/island localities of South and Southeast Asia, the Carib- bean, West Africa, and Central America, and inland zones such as the East African Highlands and parts of southern China. Most of the farmland is privately owned; large plantations, largely of tree or shrub-based crops, are common. Tree crops are an important part of the production system. Fuelwood problems are common, especially where population densities are high or land is unevenly distributed. Communal areas can be under severe stress as wood- lands are cleared for agriculture. Land tenure and gender are key issues. For those areas with secure land tenure, the best option is farm forestry or agroforestry, based on multipurpose tree produc- tion. For tenants and landless groups, policy reforms to improve access to common property resources, where these exist, or to private farmlands should be examined. 3. Medium Woody Biomass/Medium to High Population Density: These areas consist of river plains and deltas, drier coastal plains, and inland areas, includirtg the Indo-Gangetic Plain, China, South America, and Africa. They are characterized by intensive arable production, mainly through peasant farming, and large-scale cash- crop production. Landlessness is common and poverty is endemic; so are fuelwood problems. Residues are often the main fuel. Policy options aimed at the landed class center .on encouraging private tree planting and better use of residues. For the poor, indirect strategies, conservation, and measures to improve management of land resources are likely to offer the best opportunities. 4. Low Woody Biomass/Low Population Density: These areas include the montane areas of the Himalayas, the Andean and Ethiopian plateaus, Lesotho, and the southern Arabian peninsula. Substan- tial woodland clearance, especially in valleys and lower slopes, is common. So are complex mosaics of land use. Many areas are extremely remote. Fuelwood problems are widespread and acute, especially where agriculture is found on more marginal woodlands and where out-migration has made the maintenance of traditional land management practices, such as terracing, more difficult. The acute poverty and remoteness of these regions exacerbate all prob- lems, including fuel provision. The best policy opportunities are in improved management of existing land uses, both on small farms and in the remaining areas of natural woodland. 5. Low Woody Biomass/Low Population Density: These areas consist of the semiarid and arid areas of Central Asia; northern China; the Indian subcontinent; the Saharan and Sahel/Sudan belts; cen- tral, eastern, and southern Africa; and Laltin America. The economy is dominated by pastoralism, with some isolated peasant farming. There are fuelwood problems in area.ii· where traditional land FUEL WOOD PROBLEMS AND SOLUTIONS 209 management systems are still workable, but resource stresses are common in many places. Fuelwood is a less serious problem than many others. The key to these areas is the management of existing land systems, particularly action to permit the reestablishment of woodlands in areas where they have been degraded and the sup- port of communal management systems. Integration of trees into the spreading areas of cultivation will provide additional fuel resources and have other environmental benefits. 6. Urban Areas: In urban areas, use of fuelwood is common, but it is based on commodity relationships. The importance of wood and charcoal as fuels varies with the size of the urban area, the prosperity of the population, and the availability of commercial alternatives. The use of fuelwood in cities generally has a nega- tive impact on the rural supply areas. Policy options center on fuel-switching strategies, which require action to improve market structures for commercial fuels such as kerosene and LPG. 7. Transition Zones: The boundaries between environmental zones are not sharp lines; there are invariably transition zones through which one dominant set of conditions gives way to another. Within these zones, environmental conditions tend to be particularly vulnerable to disruption, and change in both the economic and environmental landscapes may be rapid. Fuelwood problems are often serious in these transition zones precisely because they are marginal. Moreover, because these areas are so complex and het- erogeneous, attempts to identify prescriptive policy options are futile. In areas affected by urban fuelwood demand, fuel-switch- ing or conservation strategies in the city are often the best way to protect the rural environment. Similarly, in agricultural coloniza- tion zones, agroforestry, extension services, land-tenure policies, and communal management of natural woodlands can play cru- cial roles in developing agricultural systems, both in the home and the colonization areas. REFERENCES Agarwal, B. 1986. Cold Hearths and Barren Slopes. London: Zed Press. Alam, M., J. Dunkerley, K.N. Gopi, W. Ramsay, and E. Davis. 1985. Fuelwood in Urban Markets: A Case Study of Hyderaba_d. New Delhi: Concept Publishing Co. Armitage, J., and G. Schramm. 1989. "Managing the Supply of and Demand for Fuelwood in Af- rica." In Environmental Management and Economic Development, eds. G. Schramm and J. Warford. Baltimore: Johns Hopkins University Press. Arnold, M., A. Bergman, and P. Harris. 1987. "Evaluation of the SIDA Supported Social Forestry Project in Ortissa, India." Swedish International Development Authority. Bajracharya, D. 1980. "Fuelwood and Food Needs Versus Deforestation: An Energy Study of a Hill Village Panchayat in Eastern Nep~" Energy for Rural Development RSI Program Report PR-80~2. Honolulu: East-West Resource Systems Institute, East-West Center. 210 MANAGING THE WORLD'S FORESTS Bajracharya, D. 1983. "Deforestation in the Food/Fuel Context: Historical and Political Perspectives from Nepal." Mountain Research and Development 3(no. 3):227-40. Bandyopadhyay, J., and V. Shiva. 1987. "Chlpko: Rekindling India's Forest Culture." Ecologist 17:26- 34. Barnard, G., and L. Kristofersen. 1985. Agricultural Residues as Fuel in the Third World. London: Earthscan Publications. Barnes, D.F. 1986. "Understanding Fuelwood Prices in Developing Nations." Agriculture and Rural Development Department, World Bank, Washington, D.C. _ _ _. 1990. "Population Growth, Wood Fuels, and Resource Problems in Sub-Sahara Africa." Energy Series Paper No. 26. Washington, D.C.: World Bank. Barrow, E. 1988. Trees and Pastoralists: The Case of Pokot and Turkana. ODI Social Forestry Network, Paper No. 6b. Bird, N., and G. Shepherd. 1988. Charcoal in Somalia. London: ODA. Bowonder, B., S. Prasad, and N. Unni. 1987. "Deforestation Around Urban Centres in India." Envi- ronmental Conservation 14:23-28. Bradley, P., N. Chavangi, and H. Van Gelder. 1985. "Development Research and Energy Planning in Kenya." Ambia 14:228-36. Briscoe, J. 1979. "Energy Use and Social Structure in a Bangladesh Village." Population attd Develop- ment Review 5:615--42. Bromley, D.W., and M.M. Cernea. 1989. "The Management of Common Property Natural Resources: Some Conceptual and Operational Fallacies." World Bank Discussion Paper No. 57. Washing- ton, D.C.: World Bank. Byer, T. 1987. "Review of Household Energy Issues in Africa." World Bank. Unpublished manu- script. Catterson, T.M. 1984. "The AID Forestry Program in Africa: A Status Report." Paper presented at USAID Bureau Forestry Program Evaluation Workshop, Lome, Togo, May 7-11. Christiansson, C. 1988. "Degradation and Rehabilitation of Agropastoral Lands: Perspectives on Environmental Change in Tanzania." Ambio 17:144--52. Ciriacy-Wantrup, S., and R. Bishop. 1975. "Common Property as a Concept in Natural Resource Policy." Natural Resources Journal 15:713-29. Clarke, H.R., and R.M. Shrestha. 1989a. "Traditional Energy Programs and the Theory of Open Access Forest Resources." Energy Journal 10:139-55. _ _ _. 1989b. "Traditional Energy Programs with an Open Access Forest Resource: Policy Impli- cations." Energy Journal 10:45-57. Cline-Cole, R.A., H.A.C. Main, and J.E. Nichol. 1990. -"On Fuelwood Consumption, Population Dynamics and Deforestation in Africa." World Developrnent 18:513-27. Dewees, P.A. 1989. "The Woodfuel Crisis Reconsidered: Observations on the Dynamics of Abun- dance and Scarcity." World Development 18:513-27. Dowd, J. 1989. "The Urban Energy Transition and Interfuel Substitution in Developing Countries: A Review of the Literature." World Bank. Unpublished report. Dunkerley, J., M. Macauley, M. Naimuddin, and P.C. Agarwal 1990. "Consumption of Fuelwood and Other Household Cooking Fuels in Indian Cities." Energy Policy 18:92-99. Eckholm, E., G. Foley, G. Barnard, and L. Timberlake. 1984. Fuelwood: The Energy Crisis That Won't Go Away. London: Earths can Publications. Elkan, W. 1988. "Alternatives to Fuelwood in African Towns." World Development 16:527-33. Energy/Development International. 1986. "Economics and Organizational Aspects of Private Tree Farming in Increase Fuelwood Production in Developing Countries." A report submitted to U5AID per Grant No. LAC-5728-G-55-4075-00. Energy Sector Management Assistane Program (ESMAP). 1.990a. Indonesia: Urban Household Energy Strategy. ESMAP Report No. 107A/90. Washington, D.C.: World Bank. _ _ _. 1990b. Zambia: Urban Household Energy Strategy. ESMAP Report Washington, D.C.: World Bank. ETC. 1987. SADCC Fuelwood Study. Leusden, The Netherlands: ETC Foundation. Fearnside, P. 1986. "Spatial Concentration of Deforestation in the Brazilian Amazon." Ambio 15:74-- 81. FUEL WOOD PROBLEMS AND SOLUTIONS 211 _ _ _. 1989. "The Charcoal of Carajas: A Threat to the Forests of Brazil's Eastern Amazon Re- gion." Ambia 18:141-43. Feder, G., T. Onchan, Y. Chalamwong, and C. Honglabarom. 1988. Land Policies and Farm Productiv- ity in Thailand. Baltimore: Johns Hopkins University Press. Fitzgerald, K., G. McGranahan, and D. Barnes. 1990. "Interfuel Substitution and Changes in the Way Households Use Energy: Estimating Changes in Cooking and Lighting Behavior in Urban Java." Washington, D.C.: World Bank, processed. Floor, W. 1987. "Household Energy in West Africa." Washington, D.C.: ESMAP, World Bank, pro- cessed. _ _ _. 1988. "Why Bother about Household Energy?" In Proceedings of the ESMAP Eastern and Southern Africa Household Energy Planning Seminar, 121-33. Washington, D.C.: World Bank. Foley, G. 1986. "The Economics of Fuelwood Substitutes." Unasylva 38:12-20. _ _ _. 1987. "Exaggerating the Sahelian Fuelwood Problem?" Ambia 16:367-71. _ _ _ . 1988. "Discussion Paper on Demand Management." In Proceedings of the ESMAP Eastern and Southern Africa Household Energy Planning Seminar, 61-84. Washington, D.C.: World Bank. Foley, G., and P. Moss. 1983. Improved Cooking Stoves in Developing Countries. London: Earthscan Publications. Fonzen, P:, and E. Oberholzer. 1989. "Use of Multipurpose Trees in Hill Farming Systems in West- ern Nepal." In Agroforestry Systems in the Tropics, ed. P. Nair. Dordrecht: Kluwer Academic , Publishers. Freeman, P.H., and T. Resch. 1986. "Large Plantations of Rapidly Growing Exotic Species: Lessons from the Bandia, Senegal." Rural Africana 23-24:87-93. French, D. 1986. "Confronting an Unsolvable Problem: Deforestation in Malawi." World Development 14(No. 4):531-40. George, S. 1988. A Fate Worse Than Debt. Harmandsworth: Penguin Books. Gill, J. 1987. "Improved Stoves in Developing Countries: A Critique." Energy Policy, (Apri1):135-44. Goodman, G.T. 1987. "Biomass Energy in Developing Countries: Problems and Challenges." Ambia 16:111-19. Gregersen, H., S. Draper, and D. Elz. 1989. "People and Trees: The Role of Social Forestry in Sustainable Development." Washington, D.C.: EConomic Development Institute, World Bank. Heermans, J. 1986. "Implications for Managing Sahelian Brushland." Paper presented to World Bank Workshop on Desertification Control and Renewable Resources Management, Oslo. Heermans, J., and G. Minnick. 1987. "Guide to Forest Restoration and Management in the Sahel." Niamey, Niger: Ministry of Hydrology and the Environment, Forestry Land Use and Planning Project. Hyde, W.F., and D.E. Mercer. 1990. "Valuing Fuelwood for Urban Areas in Developing Nations: A Methodology for Analyzing Urban Energy Policies." Washington, D.C.: Industry and Energy Department, World Bank. IPC. 1989. "The Malawi Charcoal Project: Experiences and Lessons." Frankfurt: Interdisziplinaere Project Consult. Islam, M.N. 1987. "Rural Energy and Rural Development." University of Reading Geographical Paper No. 99, Reading, England. Ives,}. 1987. "The Theory of Himalayan Environmental Degradation: Its Validity and Application Challenged by Recent Research." Mountain Research and Development 7:189-99. Johnson, V., and R. Tompkins. 1989. "Fuelwood Scarcity in Swaziland." International Journal of Ambient Energy 10:59-102. Joseph, S. 1987. "An Appraisal of the Impact of Improved Wood Stove Programmes: Synthesis of Experience." In Stoves for People. Proceedings of International Workshop, Guatemala, October 1987. Utrecht: Foundation for Woodstove Dissemination (cited in Leach and Mearns 1988). Leach, G. 1986. Household Energy in South Asia. London: International Institute for Environment and Development. _ _ _. 1987. Household Energy in South Asia. London: Elsevier Applied Science. _ _ _. 1988. "Interfuel Substitution." In Proceedings of the ESMAP Eastern and Southern Africa Household Energy Planning Seminar. Washington, D.C.: World Bank, pp. 92-120. 212 MANAGING THE WORLD'S FORESTS Leach, G., and M. Gowen. 1987. Household Energy Handbook: An Interim Guide and Reference Manual. World Bank Technical Paper No. 67. Washington, D.C.: World Bank. Leach, G., and R. Mearns. 1988. Beyond the Woodfuel Crisis: People, Land, and Trees in Africa. London: Earthscan Publications. McGrath, W.B. 1989. ''The Challenge of the Commons: The Allocation of Nonexclusive Resources." Environment Department Working Paper No. 14. Washington, D.C.: World Bank. Mahat, T., D. Griffin, and R. Shepherd. 1986. "Human Impact on Forests of the Middle Hills of Nepal." Mountain Research and Development 6:223-32. McGranahan, G. 1986. "Searching for the Biofuel Energy Crisis in Rural Java." Ph.D. dissertation, University of Wisconsin-Madison. Mercer, D.E. 1990. "Economics of Agroforestry." In Agroforestry and the Social Sciences, eds. W. Bentley and W. Burch. London: Oxford University Press. _ _ _. 1991. "Application of Household Production Theory to Selected Natural Resource Prob- lems in Less Developed Countries." Ph.D. dissertation, Duke University, Durham, N.C. Moench, M. 1989. "Forest Degradation and the Structure of Biomass Utilization in a Himalayan Foothills Village." Environmental Conservation 16:137-46. Monbiot, G. 1989. "The Transmigration Fiasco." Geographical Magazine (May):26--30. Munslow, B., Y. Katerere, A. Firf, and P. O'Keefe. 1988. The Fuelwood Trap. London: Earthscan Publications. O'Keefe, P., and B. Munslow. 1988. "Resolving the Irresolvable: The Fuelwood Problem in Africa." In Proceedings of the ESMAP Eastern and Southern Africa Household Energy Planriing Seminar. Washington, D.C.: World Bank, pp. 25-50. Openshaw, K., and C. Feinstein. 1989. "Fuelwood Stumpage: Considerations for Development Country Energy Planners." Industry and Energy Department Working Paper. Washington, D.C.: World Bank. Panayotou, T. 1989. The Economics of Environmental Degradation: Problems, Causes, and Responses. Washington, D.C.: U.S. Agency for International Development. Pearce, D., and R.K. Turner. 1990. The Economics of Natural Resources and the Environment. London: Harvester /Wheatsheaf. Pearce, D., A. Markandya, and E. Barbier. 1989. Blueprint for a Green Economy. London: Earthscan Publications. Pellek,-R. 1983. "Depletion of the Mauritanian Forest." Journal of Forestry (May):320-21. Perlack, R.D., R.B. Shelton, and A. Sabadell. "Fuelwood Substitutes and Increased Fuelwood Sup- plies: An Examination of Issues and Evidence." Draft. Oak Ridge, Tenn.: Oak Ridge National Laboratory. Pitt, M.M. 1985. "Equity Externalities, and Energy Subsidies:.The Case of Kerosene in Indonesia." Journal of Development Economics. 17:201-17. Redclift, M. 1987. Sustainable Development: Exploring the Contradictions. London: Methuen. Repetto, R. 1988. "Economic Policy Reform for Natural Resource Conservation." Environment De- partment Working Paper No. 4. Washington, D.C.: World Bank. Robinson, P., and I. Thompson. 1989. "Fodder Trees, Nurseries, and Their Central Role in the Hill Farm Systems of Nepal." ODI Social Forestry Network, Paper 9a. Runge, C.F. 1981. "Common Property Externalities/Isolation, Assurance, and Resource Depletion in a Traditional Context.'' American Journal of Agricultural Economics 67:368-72. Ryan, P. 1990a. "ESMAP: Woodfuels Development Strategy;" Washington, D.C.: ESMAP, World Banlc, processed. _ _ _ . 1990b. "India: Woodfuel Supply Situation.'' Washington, D.C.: ESMAP, World Bank, pro- cessed. Saxena, N. 1987. "Commons, Trees and the Poor in the Uttar Pradesh Hills.'' ODI Social Forestry Network, Paper Sf. Shepherd, G. 1988. "The Reality of the Commons: Answering Hardin from Somalia.'' ODI Social Forestry Network, Paper 6d. Shiva, V. 1981. "Social, Economic and Ecological Impact of Social Forestry in Kolar." Bangalore, India: Indian Institute of Management. FUEL WOOD PROBLEMS AND SOLUTIONS 213 Singh, J., U. Pandey, and A. Tiwari. 1984. "Man and Forests: A Central Himalayan Case Study." Ambia 13:80-87. Smil, V. 1983. "Deforestation in China." Ambia 12:226-31. Smith, K. 1983. "Biomass Combustion and Indoor Air Pollution: The Bright and Dark Sides of Small Is Beautiful." In Proceedings of the Conference on Biomass Energy Management in Rural Areas. Paris: UNESCO. Soussan, J. 1984. "Fuelwood Strategies and Action Programmes in Asia." Bangkok: AIT. _ _ _. 1988. Primary Resources and Energy in the Third World. London: Routledge. _ _ _. 1990a. "Common Issues and Processes in African Energy Policy." The Netherlands Minis- try of Foreign Affairs, Directorate General for International Cooperation, The Hague. _ _ _. 1990b. "Alternative Energy Supplies for Urban Areas in Somalia." Report to the Overseas Development Administration (ODA), London. Soussan, J., and P. O'Keefe. 1986. Energy and Urbanization in Kenya. Stockholm: Beijer Institute. Soussan, J., E. Mercer, and P. O'Keefe. 1991. "Funding Local Answers to Fuelwood Problems." Natural Resources Forum. Soussan, J., P. O'Keefe, B. Munslow. 1990. "Urban Fuelwood: Challenges and Dilemmas. Energy Policy, July I August. Southgate, D. 1988. "The Economics' of Land Degradation in the Third World." Environment De- partment Working Paper No. 2. Washirigton, D.C.: World Bank. Spears, J. 1988. "Containing Tropical Deforestation: A Review of Priority Areas for Technological and Policy Research." Environment Department Working Paper No. 9. Washington, D.C.: World Bank. Susman, P. 1985. Energy and Development in Island Economies. Stockholm: Beijer Institute. Teplitz-Sembitzky, W., and G. Schramm. 1989. "Woodfuel Supply and Environmental Manage- ment." Industry and Energy Department Working Paper No. 19. Washington, D.C.: World Bank. Tyler, C. 1990. "The Sense of Sustainability." Geographical Magazine (February):8-13. Wardle, P., and M. Palmieri. 1981. "What Does Fuelwood Really Cost?" Unasylva 33:20-23. World Bank. 1986. "Deforestation, Fuelwood Conservation, and Forest Conservation in Africa: An Action Program for FY86-88." Washington, D.C.: Agriculture and Rural Development Depart- ment, Operations Policy Staff, World Bank. _ _ _. 1987. "Review of Household Energy Issues in Africa." Washington, D.C.: Energy Depart- ment, World Bank. World Bank/ESMAP. 1990. "Niger: Household Energy Strategy." Washington, D.C.: World Bank. World Bank/UNDP. 1988. "Tanzania Woodfuel/Forestry Project." Activity Completion Report No. 086/88. Washington, D.C;: World Bank. World Resources Institute. 1990. World Resources 1990-91. Washington, D.C.: World Resources Insti- tute. 9 Policies Contributing to Agricultural Colonization of Latin America's Tropical Forests Douglas Southgate T hroughout Central and South America, tropical forests are rapidly be- ing converted into cropland and pasture. Some countries have im- plemented projects to relocate farmers to tree-covered hinterlands. More frequently, agricultural colonization is "spontaneous," driven by poverty, population growth, and government policies. The socioeconomic forces causing people to migrate to Latin America's agricultural frontiers are strong. In most countries, population growth ex- ceeds 1.5 percent a year. Since the debt crisis of the early 1980s, per capita economic output has stagnated, and income distribution, which has always been highly skewed in the region, has become even more unequal. Migra- tion to the agricultural frontier helps to vent the pressures that build up under these conditions. Latin American governments provide additional stimulus for agricul- tural land-use conversion. Farmers and ranchers are encouraged to colonize frontier areas in order to strengthen national claims to territory. The pri- mary cause of excessive colonization of Latin America's tropical forests, however, has to do with inadequate government response to increasing land scarcity. Competition for land is becoming more intense in the developing coun- tries of the Western Hemisphere. Urbanization is a major trend in the re- gion. As a result, agricultural land and forests on the margins of expanding cities and towns are being lost. More important, growing demands for food 215 216 MANAGING THE WORLD'S FORESTS and feed are translating into increasing derived demand for agricultural land. Within the region, growth in demand is associated with increases in population. In addition, many countries are removing the distortions that discourage agricultural production for international markets. Provided that natural conditions (e.g., soil characteristics) are suitable and that the opportunity costs of land-use conversion (e.g., forestry rents and the environmental values of tree-covered land) are covered, using more land for crop and livestock production is a rational response to increasing demand for agricultural commodities. As Hayami and Ruttan (1985) point out, efficient agricultural development should begin with outward shifts in agriculture's extensive margin. Once opportunities for geographic expan- sion start to be foreclosed, investments to improve crop and livestock yields are called for. In general, development of North American agriculture has been consistent with this pattern (Cochrane 1979). Until recently, geographic expansion of agriculture was a viable option for many Latin American countries. For example, much of the agricultural land in southern Brazil was still covered with natural vegetation well after the turn of the present century. Likewise, the conversion of tropical forests into cropland and pasture has proved to be beneficial in Ecuador's Rio Guayas Basin and along the Pacific coast of Guatemala. Current prospects for additional geographic expansion of agriculture, however, are not encouraging. In many countries, farmers or ranchers have occupied virtually all soils that are suitable for crop or livestock production. Yields tend to be modest on newly cleared parcels. Furthermore, continued tropical deforestation carries high opportunity costs in terms of lost timber resources and biological diversity. At present, public policy in Latin America does not fully reflect the actual or pending closure of the agricultural frontier. To the contrary, poli- cies that made sense when increasing demands for food and feed could be met by clearing more land for crop and livestock production remain in force. Regardless of low agricultural yields on newly colonized land and the opportunity costs associated with diminished tree cover, frontier expansion is still being encouraged. Frontier property arrangements are a major element of the status quo policy environment. Throughout the region,. government claims on tree- covered land far outstrip the public sector's capacity to control access to these properties. As a result, agricultural encroachment on them is wide- spread. Often, discrepancies between government claims and public sector management capacity are recognized and private parties are allowed to acquire public lands, but the terms of property transfer assure ecosystem destruction. "Improvement" (usually amounting to the removal of natural vegetation) of a colonized parcel is typically a prerequisite for formal ten- ure. In addition to inducing tree clearing, this arrangement discourages the conservation of existing agricultural land (Southgate 1990). FUEL WOOD PROBLEMS AND SOLUTIONS 217 Another element of the tenurial crisis underlying the loss of tree cover in Latin America is suppression of forest-dwellers' property rights. Virtu- ally every public-sector claim on tropical forests conflicts with the rights of indigenous tribes or other long-time inhabitants of tree-covered land. In addition, forest-dwellers have been obliged to claim land· in the same man- ner as newly arrived colonist-through the clearing of natural vegetation. Tenurial incentives for deforestation have been complemented by direct and indirect subsidies for agricultural land clearing. Many governments have extended cheap credit to agriculturalists in frontier areas. The Brazil- ian government has, in the past, also directly subsidized the conversion of forests into pasture. By themselves, tenurial incentives and subsidies are usually insufficient to stimulate agricultural colonization of remote tropical forests. Hence many Latin American governments have constructed roads and otherwise im- proved infrastructure to promote settlement. Where roads traverse areas with relatively good potential for crop or livestock production, colonization is usually rapid. Property arrangements, direct and indirect subsidies, and infrastructure development receive substantial attention in the literature on the causes of tropical deforestation. By contrast, relatively little has been written about the complex linkages between improvements in the scientific base under- pinning crop and livestock production and shifts in agriculture's extensive margin. By improving yields, research and extension enhance agricultural rents, thereby strengthening incentives to convert forests into cropland and pasture. Conversely, if yields stay low, frontier e/(pansion is an inevitable response to increasing demands for food and feed. During the 1960s and 1970s, a major investment was made in the scien- tific base underpinning agriculture in the region, the growth of EMBRAPA in Brazil being a case in point. By contrast, during the 1980s spending on research and extension declined in real terms. As a consequence, increases in agricultural output continue to be achieved primarily by bringing more land, which is often of marginal quality, into production rather than by increasing yields. This chapter describes the policy environment responsible for excessive deforestation throughout Latin America. Specific elements of that environ- ment are clarified through three country studies. In Brazil, substantial agri- cultural colonization has taken place along many of the roads penetrating the Amazon Basin, in part because land clearing has been subsidized and in part because tax and other policies have made land more expensive in the southern part of the country. In addition, deforestation in the Brazilian Amazon is a manifestation of a multifaceted tenurial crisis. The same crisis affects Ecuador, where inadequate investment in the scientific base under- pinning agriculture and other sectors of the rural economy also results in excessive frontier expansion. Finally, the role of intense demographic pres- sure is illustrated by the case of Guatemala. 218 MANAGING THE WORLD'S FORESTS Analysis of the causes of tropical deforestation in Brazil, Ecuador, and Guatemala yields suggestions for appropriate governmental response to increasing land scarcity. These suggestions are summarized at the end of the chapter. THE BRAZILIAN AMAZON: INFRASTRUCTURE DEVELOPMENT, SUBSIDIES, AND TENURIAL CRISIS Appeals to save tropical forests take on. special urgency whenever the Brazilian Amazon seems to be threatened. Satellite images from 1987, when the dry season was unusually pronounced, indicated that deforestation was widespread in South America's largest country. Fearing that the "lungs of the earth" were going up in smoke along with many plant and animal species, environmentalists and government leaders around the world called for quick action, Agricultural land clearing in the Brazilian Amazon, where roughly 75 percent of South America's tree-covered land is located, represents the last stage of deforestation in the country. Of the coastal forest that once stretched from northeastern Brazil almost to Uruguay, only remnants remain. Simi- larly, agricultural land-use conversion is virtually complete in the western parts of Sao Paulo, Parami, and neighboring states in the south. In an area as large as Brazil's Norte Region, which comprises six north- ern and northwestern territories and is bigger than the European Economic Community, all factors contributing to deforestation exist. This section pre- sents a general overview of the forces driving agricultural land clearing in the Brazilian Amazon and then focuses on foru· causes: infrastructure devel- opment, direct and indirect subsidies for deforestation, differences between land values in Amazonia and land values in other parts of the country, and tenurial crisis, Extent and Consequences of Tropical Deforestation The actual extent of deforestation is one olf the major controversies sur- rounding environmental change in the Brazilian Amazon. In a frequently cited study, Mahar (1989) claims that land clearing has accelerated through- out the past 15 years. As he reports, satellite images show that deforested area in the Norte Region and three neighborilng states (Goias, Maranhao, and Mato Grosso) increased from 0.6 percent in 1975 to 2.5 percent in 1980. Extrapolating from Fearnside (1986) and other sources, he estimates that cumulative deforestation in 1988 was 60 million hectares-12 percent of the region. More recent research contradicts the argument that land clearing has continued to accelerate during the last few years. Referring to agricultural FUEL WOOD PROBLEMS AND SOLUTIONS 219 census data, Schneider et al. (1990) report that annual growth in rural popu- lation, cropland, and related indicators has declined during the past few years. In addition, interpretation of recent satellite imagery shows that Mahar' s (1989) estimates of cumulative land-use change in the late 1980s are exaggerated. For example, Brazil's National Institute for Space Research has found that deforested area as of 1988 was between 7 and 8 percent (Schneider et al. 1990). While land clearing in the Brazilian Amazon as a whole is less rapid than some observers have projected it to be, extensive deforestation has occurred along many of the roads penetrating the region from the south and east. For example, farmers and ranchers have occupied more than 10 per- cent of Rondonia state, which is traversed by Federal Highway BR-364. Forests are also being lost because of hydroelectric and industrial develop- ment. Completion of the Tucurui hydroelectric dam, for example, has re- sulted in the flooding of more than 200,000 hectares in Para (Goodland 1985). In the same state, deforestation is also taking place on the margins of the Carajas Iron Ore Project (Mahar 1989). Deforestation in the Brazilian Amazon carries a number of costs, all of which defy precise estimation. Myers (1984) points out that Amazonia is the "single richest region of the tropical biome." On a single hectare near Manaus, for example, Prance (1986) was able to identify more than 400 tree species. Deforestation also affects climate (primarily at the local level) and puts soil resources at risk (Schneider et al. 1990). Causes of Tropical Deforestation Other than swidden cultivation of subsistence crops, which has rela- tively little effect on the natural environment, agriculture was never a major activity in the Amazon before recent times. The first migration into the region in modern times occurred during the Rubber Boom, which collapsed soon after 1910, as Asian plantations came on line. Subsequently, the Ama- zon Basin reverted to its traditional isolation. In 1960, Rondonia, a federal territory at the time, could be reached only by small aircraft or riverboats, and its population numbered only 70,000, mostly prospectors and rubber tappers. Its tropical forests remained largely undisturbed (Mahar 1989). Road construction first broke the Brazilian Amazon's isolation during the 1960s. Initially directed toward Goias, Maranhao, Mato Grosso, and Para, agricultural colonization accelerated during the late 1970s and early 1980s. Subsidies for cattle ranching provided additional stimulus for defor- estation. So did tax and other policies that helped to widen the gap between land prices in southern Brazil and land prices in Amazonia. Finally, as settlement and land clearing have continued both within the Norte Region and along its borders, conflicts over property rights have intensified. 220 MANAGING THE WORLD'S FORESTS Infrastructure Development In 1960 the Norte Region had only 6,000 kilometers (km) of roads, of which less than 300 km was paved (Mahar 1989). Land transportation first became practical in 1964, when Federal Highway BR-010 linked Brasilia and Belem. Katzman (1977) has estimated that as many as 320,000 people mi- grated to the road's zone of influence, in Goias and Para, during the 10 years before the 1970 population census. Rapid deforestation along feeder roads continued in the 1970s and 1980s (Mahar 1989). The military government that came to power in 1964 gave highway construction in the Amazon Basin high priority. Responding to infrastruc- ture development and settlement in the tropical forests of Peru, Venezuela, and other adjacent countries, it launched "Operation Amazonia" soon after taking control (Mahar 1989). Among the first infrastructure projects com- pleted under the auspices of that initiative was the first 1,200 km of the Transamazon Highway, which ran west from Maranhao. A primary objective of the new road, which was finished in 1972, was to facilitate resettlement of 70,000 families from northeastern Brazil, which is impoverished and prone to drought (Mahar 1979). The anticipated ben- efits of the project were summarized by the phrase, "a land without people for a people without land." But colonization objectives were not met, largely because natural condi- tions were highly unfavorable. Smith (1981) has found that only 3 percent of the land along the Transamazon Highway is suitable for agriculture. Ma- laria infection was also a chronic problem for settlers because deforestation created breeding grounds for the Anopheles mosquito (Moran 1983). In addi- tion, because of the remoteness of agricultural settlements, farm-gate prices of crops and livestock tended to be low, and purchased input prices tended to be high. Because government assistance to colonists fell far short of what was needed to overcome these disadvantages, ·a mere 8,000 families had settled by 1980 (Smith 1981). Only about 4 percent of total deforestation in the Brazilian Amazon can be attributed to construction of the Transamazon Highway (Browder 1988). Other federal highways traversing areas ill-suited to crop or livestock production have similarly resulted in relatively little deforestation. This has been the case with westward extensions of the Transamazon Highway, the road linking Porto Velho (Rondonia' s capital) with Manaus, and the high- way between Cuiaba (the capital of Mato Grosso) and Santarem (a city a little more than halfway up the Amazon River from Belem to Manaus). By contrast, where conditions are more hospitable to agriculture, road construction has induced rapid agricultural land clearing, as, for example, in northern Mato Grosso and southern Para. Similarly, the completion of BR-364 has made possible the subsequent colonization of relatively fertile areas in Rondonia state. FUEL WOOD PROBLEMS AND SOLUTIONS 221 Subsidies While infrastructure development was its most visible early manifesta- tion, the military government's Operation Amazonia had other effects on economic activity in the Norte Region and adjacent states. The Superinten- dency for the Development of Amazonia (SUDAM) was established to sub- sidize industrial and agricultural development, and the Bank of Amazonia (BASA) was founded. The offer of financial inducements to firms operating in the Brazilian Amazon predates military rule and the foundation of SUDAM and related institutions. In 1963, for example, legislation was passed stipulating that a Brazilian firm could reduce its income tax payments by one-half if the resulting savings were directed to industrial investments in the Amazon Basin. Operation Amazonia greatly expanded the subsidy program. Most important, agricultural, livestock, and service-sector projects receiving SUDAM approval first became eligible for tax credits in 1966 (Mahar 1989). The livestock sector's response to the investment tax credit scheme was dramatic. By late 1985 SUDAM had approved some 950 projects, 631 of which were cattle ranches (Garda-Gasques and Yokomizo 1986). More than $700 million in tax credit funds administered by SUDAM were channeled to the livestock sector, which has also received a substantial amount of cheap credit during the past two decades (Browder 1988). Subsidies have enhanced the interest of many firms in Amazonian cattle ranches. Citing an enterprise-level study (Hecht, Norgaard, and Possio, n.d.), Mahar (1989) concludes that the typical SUDAM-approved project (which had more than 20,000 hectares of land) would not have been profitable had the investment tax credit not been available. Similarly, Browder (1988) has used enterprise-level analysis to document that credit subsidies amounted to a large part of the accounting profits of many cattle ranches. The effects of direct and indirect subsidies on deforestation throughout the Brazilian Amazon should not be exaggerated. Mahar (1989), for example, posits that tax breaks for cattle ranching, which have been curtailed, explain no more than 10 percent of total land clearing in the entire region, even though they have stimulated considerable deforestation in Para and Mato Grosso. Differences in Land Values Compared with the attention devoted to subsidies for agricultural land- use conversion in the Brazilian Amazon, relatively little has been said about the deforestation caused by steep appreciation of land values in other parts of the country. 222 MANAGING THE WORLD'S FORESTS Land prices in southern Brazil rose rapidly, relative to prices in the Norte Region, during the early 1970s. Major interregional differences were sustained through the early 1980s. In 1970 a farmer from the Sudeste Region, for example, could use the money earned from selling one hectare there to buy two hectares in the Norte Region. Ten years later, the value of land in the southern part of the country was 10 times the value of land in the latter. Schneider et al. (1990) emphasize that increases in agricultural produc- tivity explain a large part of land price trends in southern Brazil during the past two decades. However, they join Binswanger (1989) in pointing out that preferential agricultural taxation and subsidized credit raised land prices more in the Sul and Sudeste regions than in other parts of the country. The same policies encouraged the consolidation of land in large holdings where capital-intensive production practices were applied. Some of the people dis- placed by land concentration and mechanization emigrated to the Amazon (Mahar 1989). Tenurial Crisis Settlement and land clearing in the Brazilian Amazon, stimulated by road construction, subsidies, land price differentials, and other factors, have greatly harmed the region's forest-dwellers. Many indigenous tribes have been hard hit by diseases introduced by recently arrived colonists. In addi- tion, there has been wholesale violation of forest-dwellers' property rights, both those of indigenous tribes and those of the descendants of Rubber Boom immigrants. Conflicts often arise because documentation of forest-dwellers' land rights is spotty to nonexistent in the Brazilian Amazon. What is more, outsiders often use intimidation or actual violence to induce indigenous groups and people who make their living from the extraction of nontimber forest prod- ucts' to surrender their rights (Schwartzman and Allegretti 1987). The importance of violent tactics was vividly demonstrated in December 1988 when "Chico" Mendes, a leader of the rubber tappers' union in Acre, was murdered. Although government authorities do not condone the violent attacks on forest-dwellers, the property regime in Brazil's frontier areas enhances the ability of cattle ranchers and other recently arrived colonists to prevail in land disputes. As elsewhere in Latin America, a person claiming a parcel is obliged to clear up to half of it in order to show that "improvement," which is a prerequisite for formal tenure, has occurred (Mahar 1989). This require- ment, of course, puts those who make a living without greatly disturbing forests at a disadvantage in conflicts over tree-covered land. The Brazilian government is taking measures to contain violation of forest-dwellers' property rights. About 2 million hectares of extractive re- serves, where collection of nontimber forest products is to be the principal economic activity, have been demarcated in the Brazilian Amazon. So have FUEL WOOD PROBLEMS AND SOLUTIONS 223 nearly 200 areas for indigenous groups. The institutional capacity and po- litical will needed to maintain the integrity of many of these lands will be put to the test during the next few years. ECUADOR: TENURIAL CRISIS, PRICE DISTORTIONS, AND WEAK AGRICULTURAL RESEARCH AND EXTENSION With a little more than 26 million hectares on the South American main- land, Ecuador is slightly larger than Brazil's Rondonia state or West Ger- many before unification. Nevertheless, its geography is astonishingly var- ied. Elevations range from sea level, along the Pacific coast, to 6,310 meters at the peak of Mount Chimborazo, a mere 250 km inland. Amazonian low- lands (the Oriente) occupy the eastern third of the country. As elsewhere in the tropics, environmental variety goes hand in hand with dramatic topography. Ecuador can boast of 25 of the 30 "life zones" identified by Holdridge (1967), as well as considerable biological diversity. For example, the number of vertebrate species living in the country ap- proaches the number found in its larger neighbors. The Latin American Division of the Nature Conservancy indicates that 2,950 such species have been identified in Brazil, 2,800 in Colombia, 2,550 in Peru, and 2,450 in Ecuador. More than 10 percent of those living in Ecuador are endemic, and many are threatened by deforestation. Human beings have considerably altered the Ecuadoran landscape. Parts of the Andean highlands were cleared for agriculture long before Colum- bus's journeys, and the Sierra was largely deforested a hundred years ago. Reforestation began in the region during the 1800s, when eucalyptus was introduced from Australia. Today eucalyptus and pine trees cover a small portion of the highlands. Extensive forests remain in lower elevations, but agricultural coloniza- tion is rapid in the western Littoral (or Costa) and in the Oriente. Although that activity is not directly subsidized, as has been the case in Brazil, it is accelerated by high human fertility and extreme land scarcity in other parts of the country. In addition, as this section emphasizes, deforestation and other forms of resource degradation in the Ecuadoran countryside are encouraged by three elements of the policy environment: inappropriate tenurial arrangements, government interference with market forces, and in- adequate investment in the scientific base underpinning agriculture and other sectors of the rural economy (Southgate and Whitaker forthcoming). Extent and Consequences of Tropical Deforestation Agricultural colonization in lowland Ecuador has a long history. Before the country achieved independence from Spain, farmers started to settle the area around Guayaquil. By the 1870s periodic yellow fever outbreaks had been controlled in the Costa, and cacao plantations had been established 224 MANAGING THE WORLD'S FORESTS along waterways, which were the principal mode of transportation, as far as 75 km north and east of the port city. By the turn of the century, at the height of Ecuador's cacao boom, the extensive margin of coastal agriculture had been pushed 150 km north of Guayaquil. The cacao boom collapsed after World War I, and deforestation was relatively slow during the 1920s and 1930s. However, coastal agriculfure resumed its geographical expan- sion after World War II (Bromley 1981). Until recent times, eastern Ecuador's Amazonian lowlands were a back- water. Farmers had penetrated into only a few valleys at the base of the Andes. Development of the region was given a major boost in 1967, when petroleum was discovered near Nueva Loja (:formerly Lago Agrio). Just as an earlier generation of Ecuadoran agricultural colonists had followed navi- gable waterways inland from Guayaquil and :iust as farmers from southern Brazil have followed BR-364 into Mato Grosso, Rondonia, and Acre, immi- grants from the heavily populated Costa and Sierra have settled along roads leading to the oil fields of northeastern Ecuador. Annual surveys conducted. by the Ecuadoran Ministry of Agriculture and Livestock (MAG) from 1965 through 1985 indicate recent trends in crop- land and pasture. No single year's estimates are very reliable, because they are based on information provided by a poorly funded and understaffed network of extension agents. However, the MAG data set clearly shows that rapid agricultural colonization has occurred in lowland Ecuador during the past quarter-century. The area planted to tropical crops, almost entirely in the Costa, increased by 155,000 hectares between the mid-1960s and the mid-1980s. During the same period, pastures in Ecuador's five coastal provinces quadrupled. The Costa's agricultural frontier was also advancing into the western, lowland portions of several Andean provinces. Thus in the past quarter-century de- forestation in western Ecuador has been widespread. Twenty-five years ago land planted to crops was negligible, and pas- tures covered around 225,000 hectares in the Oriente. By the mid-1980s, the area used for livestock production, equal to six times the amount of crop- land, was approaching 500,000 hectares. Some past deforestation in Ecuador has proved to be beneficial. Land- use conversion in the Rio Guayas Basin, extending north from Guayaquil, is a good case in point. Although seasonal flooding is a problem for some farmers and erosion occurs in the upper watershed, soils are generally fer- tile and the basin has become the country's agricultural heartland. However, Ecuador's agricultural frontiers now penetrate far into areas with limited agricultural potential. Cropland and improved pasture cur- rently amount to 4.5 million hectares in the western two-thirds of continen- tal Ecuador (i.e., all land west of the Oriente). This area exceeds the 3.1 million hectares in the Sierra and Costa classified by MAG's National Program for Agrarian Regionalization (PRONAREG) as "prime farmland" because it is fertile, well drained, and not highly erodible. The 4.5 million hectares even exceeds the sum of prime farmlland plus 1.3 million hectares FUEL WOOD PROBLEMS AND SOLUTIONS 225 of land that PRONAREG has determined can be used for sustained crop production once measures to deal with serious erosion or drainage prob- lems have been taken (Southgate and Whitaker forthcoming). Increased settlement of the Oriente does not constitute a satisfactory solution to the problem of scarce prime farmland in the western two-thirds of Ecuador. MAG (1987) estimates, for example, that 84 percent of the soils in the northeastern part of the country should never be cleared for crop or livestock production. Aside from yielding modest agricultural returns, deforestation carries several opportunity costs. Inefficiencies arise, for example, where land with value as a site for timber production is instead used to produce relatively small amounts of agricultural commodities. In addition, disutility is associ- ated with the destruction of plant and animal species. With respect to the latter, Myers (1988) has identified 10 "hot spots" where there is an immi- nent threat that tropical forests with unusually high biological diversity will be cleared. One of those locations lies entirely within Ecuador, and two others extend into other countries. Despite limited (or negative) net social returns to recent deforestation, the pace of agricultural land clearing continues to be rapid in lowland Ecua- dor. Officials of MAG' s National Forestry Directorate (DINAF) indicate that land clearing in the Oriente and northern Costa regions together amounts to 200,000 hectares per annum. Causes of Tropical Deforestation Socioeconomic conditions contribute substantially to tropical deforesta- tion in Ecuador, which is the most densely populated nation in South America and in which income distribution is highly skewed. Per capita gross na- tional product (GNP), which was only US$1,120 in 1988, is stagnating, largely because the country's population is increasing by more than 2.5 percent a year (IBRD 1990). Ecuador's rural poor have responded to overcrowding and poor envi- ronmental conditions by migrating to lowland agricultural frontiers. During the 1970s, for example, an extended drought in Loja province, in the south- ern Sierra, drove small farmers to follow recently constructed oil roads in the northeastern part of the·country. That natives of that province are dis- proportionately represented among recent immigrants is indicated by the new name given to the town where Ecuador's oil industry is centered: Nueva Loja. Explanations of deforestation in Ecuador based exclusively on migra- tion from areas experiencing population growth, resource degradation, and poverty, however, do not suffice. It is interesting to note, for example, that agricultural land clearing has been particularly rapid in Esmeraldas, the northernmost Costa province. Between the 1974 and 1982 censuses, the rural population of Esmeraldas actually declined (!NEC 1982). 226 MANAGING THE WORLD'S FORESTS Instead of being a simple consequence of demographic pressure, exces- sive deforestation in Ecuador is largely explained by the policy environ- ment. The three principal features of that environment are inappropriate tenurial arrangements, government interference with market forces, and inadequate investment in the scientific base underpinning agriculture and forestry. Inappropriate Tenurial Arrangements As in the rest of Latin America, tropical deforestation in Ecuador is largely a manifestation of tenurial crisis. Most of the country's tree-covered land is designated as forest patrimony (patrimonio forestal) or is in a park or reserve. However, the government's ability to control access to these forests is modest. Very few rangers are assigned to the 2.0 million hectares of forest patrimony delimited in the northwestern and northeastern parts of the coun- try (MAG 1987). Not much more is done to protect the 2.1 million hectares of parks and reserves in continental Ecuador. In 1987, for example, a mere two administrators, 25 technicians, and 119 permanent and seasonal rangers were assigned to those areas (DINAF 1988). Implicitly recognizing the discrepancy between its extensive claims and its limited capacity to control access, the Ecuadoran government allows for the transfer of public forests to private parties. However, the terms of that transfer assure ecosystem destruction. Agents of the Ecuadoran Institute for Land Reform and Colonization (IERAC) typically insist that at least half of a SO-hectare claim be cleared before they will "legalize" (or "adjudicate") the claim. The only direct fiscal check on the conversion of forests into agricultural land is an adjudication fee, which was raised in 1989 from a nominal $2 per hectare to a still minor $20 per hectare. Furthermore, tenure insecurity, associated with the many years IERAC requires for adjudication, accelerates deforestation, as Southgate, Sierra, and Brown (forthcoming) have demonstrated in a statistical analysis of the causes of settlement and land clearing in eastern Ecuador. The tenurial regime inducing colonists Ito clear the forest patrimony causes the original inhabitants of Ecuador's tropical forests to follow suit. Macdonald (1981) reports that the periodic fallowing scheme long practiced by the indigenous community of Pasu Urcu, in the eastern part of the coun- try, was abandoned during the 1970s after IIERAC agents informed mem- bers of that community that fallow lands could be claimed by agricultural colonists, who were 50 km away at the time. That land-use change was inefficient, because farming without fallowing is extremely difficult in the Amazon Basin (Denevan et al. 1985). The ban on logging concessions imposed iln 1982 should be added to the list of tenurial factors discouraging efficient development of Ecuador's tropical forests. That ban makes the forest-products industry in northwestern Ecua- dor almost entirely dependent on agricultural colonists for supplies of raw FUEL WOOD PROBLEMS AND SOLUTIONS 227 materials. These colonists' rudimentary harvesting and log transport tech- niques cause significant damage to commercial timber. The degree to which prohibiting logging concessions has enhanced colonists' interest in timber sales is indicated by the increase in chain-saw imports during the early 1980s. In 1981 imports amounted to US$1.8 million (c.i.f. 1988 dollars). In 1985, three years after the ban went into effect, US$5.2 million worth of chain saws were brought into the country, as colonists switched to clearing techniques allowing for the marketing of commercial species (Montenegro and Durini 1989). Price Distortions Along with inappropriate tenurial arrangements, the Ecuadoran gov- ernment's interference with market forces discourages forest conservation. Gasoline prices rarely rose above US$0.10/ gallon during the 1970s or US$0.30/gallon during the 1980s. As a result, agricultural colonization of tree-covered land has been subsidized. The prohibition on log exports also has affected the use and manage- ment of Ecuador's forests. The severe price distortions associated with the ban, which in effect have converted the domestic market for timber into a local monopsony, are illustrated by the difference between the prevailing domestic price for IO-meter eucalyptus logs, $2 to $4 each, and the price offered by an Italian firm for the same product in early 1989, around $20 each. Low timber prices, in turn, reduce private incentives both to protect existing forests and to establish new stands of trees. For example, total reforestation from 1962 through 1985 amounted to 100,000 hectares (AIMA 1985), which is considerably less than the current annual deforestation rate of 200,000 per hectare mentioned earlier. Inadequate Research and Extension Latin American governments have been quick to extend subs_idies, offer protection to industry, and otherwise to interfere with market forces, but they have been slow to develop the scientific base on which growth in agriculture, forestry, and other sectors of the rural economy depends. The consequences of underinvestment in agricultural research and ex- tension are clear in Ecuador. In real terms, spending on research in support of crop and livestock production declined 7.3 percent per annum from 1975 through 1988. At 0.17 percent of agricultural gross domestic product, Ecuadoran research expenditures compare poorly with spending in neigh- boring countries (Whitaker 1990). In addition, limited agricultural research is matched by a weak extension service, and thus Ecuadoran crop yields are below regional standards. As a result, increasing demand for food and feed must be met primarily by bringing additional land, which is usually of marginal quality, into production. The expanding agricultural frontier 228 MANAGING THE WORLD'S FORESTS accounted for two-thirds of the increased crop production occurring in Ec- uador between the mid-1960s and the mid-1980s. Improved productivity explained only the remaining third (Whitaker and Alzamora 1990). And forestry research and extension compare poorly with agriculture's scientific base. Aside from one or two companies undertaking species trials, private-sector forestry research in Ecuador is negligible. Government re- search is correspondingly modest, and, for all intents and purposes, there is no forestry extension. Under these circumstances, no farmer or rancher per- ceives that the conversion of forests into cropland or pasture involves a substantial opportunity cost. GUATEMALA: INTENSE DEMOGRAPHIC PRESSURE With nearly 11 million hectares of terriltory, Guatemala is the third- largest country in Central America. In many ways its varied landscape is comparable to Ecuador's. Corresponding to the Ecuadoran Oriente is Guatemala's Peten, which is part of a limestone tableland extending north- ward to the Gulf of Mexico. The Peten is characterized by high average temperatures, heavy precipitation, and poor soil. Guatemala's highland re- gion, where the country's population has long been concentrated, is similar to the Ecuadoran Sierra. Finally, Guatemala's fertile Pacific and Caribbean lowlands, amounting to a tenth of the national territory, correspond to Ecuador's Rio Guayas Basin. Fourteen Holdridge ''life zones," ranging from humid subtropical forest to dry tropical forest, are found in Guatemala (De la Cruz 1976). By itself this fact would imply that a large number of plants and animals were present in the country. That number is greatly enhanced by Guatemala's location on the land bridge connecting two continents, each with distinct flora and fauna. Ancestors of the country's 600 bird species, 250 kinds of mammals, and 200 species of reptiles and amphibians came from both North and South America. Sixteen coniferous and 450 broad-leaf tree species reflect Guatemala's var- ied plant life (URL 1984). Factors contributing to deforestation in Brazil, Ecuador, and other Latin American countries are also at work in Guatemala. Formal property rights have traditionally been a reward for clearing land in frontier areas that have recently become accessible because of infrastructure development. In addi- tion, because the country invests little in research and extension, expanding demand for crops and livestock tends to be satisfied primarily through ex- pansion of the agricultural frontier. More than in many other countries, however, agricultural land clearing in Guatemala is the direct result of demographic pressure. With a popula- tion of nearly 9 million, it is the second most crowded country in Central America. Furthermore, fertility rates exceed mortality rates by a wide mar- gin in Guatemala: 4.0 percent versus 0.8 percent in 1988 (IBRD 1990). As this section explains, migration to the Peten is a response to these demographic forces. FUEL WOOD PROBLEMS AND SOLUTIONS 229 Extent and Consequences of Tropical Deforestation Compared with agricultural colonization in western Ecuador, tropical deforestation in Guatemala is a recent phenomenon. Occupation of the lower valleys of the Izabal and Motagua rivers (which empty into the Caribbean) and the Pacific lowlands began in earnest well after 1900. Agricultural colo- nization in those parts of the country has run full course; farmers and ranchers currently occupy 90 percent of the land. Deforestation has been even more extreme on the slopes rising from the southwestern plains. More than 95 percent of the Pacific slopes, where most of the country's coffee is produced, is deforested. Even the shade trees on coffee plantations are under threat because of local demand for fuelwood. With agricultural frontiers closed in other parts of Guatemala, defores- tation is occurring rapidly in the Peten. As in other parts of the world, agricultural colonization in the region has been facilitated by infrastructure development. Farmers and ranchers originally settled along roads leading to oil fields and logging sites. More recently, roads have been constructed primarily to facilitate settlement (e.g., in the Franja Transversal del Norte). As of 1979, when the latest agricultural census was conducted, all but 30 percent of the Peten was covered with trees. Satellite images taken in 1987 and 1988, however, suggest that cumulative deforestation has reached 40 percent. Farmers and ranchers have occupied practically all the land in Guate- mala with good or fair potential for agricultural production. In the high- lands, agricultural frontiers extend far into areas that should, because of serious erosion problems and other natural limitations, retain tree cover. Perhaps 50,000 hectares of forest could be converted to cropland or pasture in the Caribbean and Pacific lowlands. Trees also cover some land in the Peten with agricultural potential (Southgate and Basterrechea forthcoming). Agricultural land-use conversion continues to be rapid in Guatemala. Taking into account land clearing in the Peten as well as natural regenera- tion and 6,600 hectares per year of tree plantings in the entire country, Del Valle et al. (1990) estimate net deforestation at 32,500 to 34,300 hectares per year. · Causes of Tropical Deforestation To understand the causes of land-use change in the Peten and other lowland areas in Guatemala, one must recognize social realities in the country's rural highlands. The primary reality is rapid population growth. Demographic Pressure and Land Scarcity Nowhere in Latin America is average fertility higher than in Guate- mala, largely because that country's citizens are less inclined than people in other parts of the Western Hemisphere (with the exception of Haiti) to use 230 MANAGING THE WORLD'S FORESTS contraceptives. Because birth control generally is positively correlated with urbanization and higher incomes, fertility rates among Guatemala's rural poor are particularly high. Accordingly, population growth in the highlands is pronounced despite substantial emigration to Guatemala's cities and ag- ricultural frontiers and to the United States. Rates of increase approach 3 percent in the western highlands. Even in the eastern highlands, where emigration is a major phenomenon, growth rates generally exceed 2 percent (INE 1989). Because of population growth, the amount of land available to the typi- cal highland farmer has declined over time. For the country as a whole, the ratio of arable land to human population fell from 1.11 hectares per person in 1964 to 0.79 hectare per person in 1979. The decline has been especially severe in western highland departments. In 1979 average land endowments were 0.32 hectare in Huehuetenango, 0.22 hectare in Chimaltenango, 0.10 hectare in Totonicapan, and just 0.05 hectare in Solola (Davis et al. 1989). Smallholders bear the burden associated with increased land scarcity. Between the second and third agricultural censuses (conducted in 1964 and 1979, respectively), more than 20,000 new farms with 0.7 hectare to 7 hect- ares were formed, almost exclusively in the highlands. Meanwhile, there was virtually no change in total area divided among all enterprises in that size class. During the same period the number of holdings larger than 45 hectares increased by 55 percent (principally because of agricultural settle- ment in. the Peten, where holdings tend to be larger), and the total area in those holdings rose by nearly one-fourth. Guatemala's Gini coefficient (a measure of concentrated land ownership) has reached 85 percent, which is the highest reading in Central America and one of the highest in the West- ern Hemisphere (Davis et al. 1989). As mentioned earlier, the rural poor of highland Guatemala are re- sponding to population growth and land scarcity by emigrating to the Peten. The latter region's rural population, which is cwrrently estimated to be 158,987, is increasing by 5.5 percent a year, principally because of immigration (INE 1989). A disproportionately large share of the immigrants come from the eastern highlands, where overpopulation, poverty, and resource degrada- tion are especially acute (Southgate and Basterrechea forthcoming). Disincentives for Forest Conservation Once they arrive in the Peten, settlers are presented with strong incen- tives to convert forests into agricultural land. In the past, government has encouraged pasture establishment by extending cheap credit to livestock producers. The property rights regime also induces agricultural land clear- ing. To win a favorable adjudication from the National Institute for Agrar- ian Transformation (INTA), a settler must "improve" his cabelleria of 46 hectares. "Improvement" generally means the removal of natural vegetative cover. Furthermore, settlers are obliged to assert informal agricultural use FUEL WOOD PROBLEMS AND SOLUTIONS 231 rights, because INTA can take years to process an application for formal tenure. Once settlers have deforested a parcel of land, they are in a poor posi- tion to practice sustainable agriculture. Even though prospects for the pro- duction of some perennial crops (e.g., cashew) and for agroforestry appear to be promising in the region, little research on these and other agricultural land-use options is being done. At the same time, an agricultural extension service in the Peten is just being developed. The combination of inappropriate tenurial arrangements and weak ag- ricultural research and extension encourages recent arrivals from Guatemala's highlands as well as long-term residents of the Peten to engage in a cycle of deforestation and nonsustainable farming. CONCLUSION: THE CHALLENGE OF POLICY REFORM Population growth and poverty explain a large part of depletive human interaction with Latin America's natural environment. To appreciate the importance of underlying socioeconomic conditions, one only needs to travel through Haiti and parts of Central America, where the rural poor are mu!- , tiplying quickly and picking the countryside clean. Laying the bfame for resource degradation exclusively on mounting demographic pressure and poverty, however, does not leave much room for optimism. Poor countries' populations are overwhelmingly young. With the number of women capable of bearing children expected to rise for many more years, continued population growth is inevitable, even though fertility rates in much of Latin America are declining. Similarly, educational efforts needed to raise the earning power of the rural poor will take at least a generation to yield results. Because of the immediate threat to Latin America's tropical forests, fertility control and improved education in rural areas cannot be the only elements of a strategy to control deforestation. Policy reform is also .essential. As indicated in this chapter, a number of policies-inappropriate prop- erty arrangements, subsidies, inadequate formation of the rural economy's nonenvironmental assets, and the like-combine to induce excessive agri- cultural encroachment on tropical forests and other natural environments. Accordingly, several reforms are required. • Devolve control over forests. As illustrated by the case of Ecuador (Southgate and Whitaker forthcoming), Latin American governments have not been able to manage, or even to control access to, their extensive holdings of tree-covered land. Another aspect of the tenur- ial crisis underlying deforestation is that destruction of natural veg- etation is a prerequisite for property rights along the region's agricul- tural frontiers. 232 MANAGING THE WORLD'S FORESTS More than anything else, devolution involves strengthening the prop- erty rights of individual settlers, indigenous groups, and others who use tree-covered land. Mechanisms for defining property rights and settling disputes among rival claimants should be streamlined. Obvi- ously, deforestation should no longer be a requirement for formal tenure. • Eliminate subsidies for land-use conversion. Brazil's progress in eliminat- ing subsidies for deforestation needs ito be matched in other coun- tries, where indirect financial inducements for agricultural coloniza- tion continue to be offered. For example, public-sector development banks in many countries should stop disbursing low-interest loans for cattle ranching in frontier areas. • Eliminate policies that artificially enhance land values in nonfrontier areas. Insofar as differences between land prices in nonfrontier areas and land prices in areas undergoing agriculltural colonization are inflated by tax and other policies, as has been the case in Brazil (Schneider et al. 1990), deforestation is excessive. • Plan and evaluate infrastructure projects more carefully. Schneider et al. (1990) recommend a litmus test to be used when evaluating a pro- posed road. Approval should be granted only if those living in the road's zone of influence find it more profitable to engage in environ- mentally sustainable activity and if the income generated by that activity exceeds the cost of building the road. • Accelerate formation of nonland assets in the agricultural sector. Some economists attempting to explain the loss of natural habitats in the developing world fall into a habit of analysis that is nearly as old as the discipline itself. Like those who advocate acreage controls to re- duce agricultural commodity surpluses in the United States, they underestimate the degree to which land and other inputs to the pro- duction of crops and livestock are interchangeable. If the option of substitution is ignored, the predictions of a simple Ricardian model of the agricultural economy hold, that is, frontier expansion is the only possible response to market or demographic "shocks." To be sure, formation of nonland assets should reflect an agricultural economy's factor endowments (Hayami and Ruttan 1985). For ex- ample, investment in yield-enhancing technology is not particularly urgent where land and other natural resources are abundant. Unfor- tunately, investment of that type continues to be marginal in many Latin American countries where the prospects for frontier expansion are limited. Put another way, agricultural underdevelopment and encroachment by farmers and ranchers on fragile environments go hand in hand in the region. In addition to implementing the package of policy reforms just out- lined, Latin American governments could map land-use capabilities in areas that will come under colonization pressure and, through financial and other FUEL WOOD PROBLEMS AND SOLUTIONS 233 inducements, encourage activities consistent with those assessed capabili- ties. Because this action would mark a significant departure from previous practice, "zoning" should probably not be the centerpiece of forest conser- vation strategies in the region. Instead, emphasis should be placed on alter- ing the policy environment that has been described here. The difficulty of overhauling the policies that accelerate deforestation should not be underestimated. Powerful interests support many of the ar- rangements that promote colonization. Hoping to allay national security concerns through the establishment of "live frontiers," for example, many Latin American governments have encouraged farmers to settle in tree- covered hinterlands. Reform is a challenge also because piecemeal changes in policy can be ineffective. For example, if tenurial incentives for deforestation are removed but research and extension are not improved, the effects on agricultural colonization could be minor. Piecemeal reform can even be counterproduc- tive. For instance, if property rights in tree-covered land remain attenuated, lifting log export bans or otherwise deregulating forest-product markets can easily stimulate deforestation. As indicated in this chapter, similar sets of policies that contribute to excessive agricultural encroachment on tropical forests exist in many parts of Latin America. Accordingly, many countries face the challenge of under- taking a thorough reform. REFERENCES Asociaci6n de Industrias Madereras (AIMA). 1985. Diagn6stico Actuali:Zado de[ Sector de la Madera en el Ecuador. Quito: AIMA. Binswanger, H. 1989. "Brazilian Policies that Encourage Deforestation in the Amazon." Environ- ment Department Working Paper No. 16. Washington, D.C.: World Bank. Bromley, R. 1981. "The Colonization of Humid Tropical Areas in Ecuador." Singapore Journal of Tropical Geography 2:15-26. Browder, J. 1988. "Public Policy and Deforestation in the Brazilian Amazon." Public Policies and the Misuse of Forest Resources, eds. R. Repetto and M. Gillis. Cambridge, Eng.: Cambridge University Press. Cochrane, W. 1979. The Development of American Agriculture. Minneapolis: University of Minnesota Press. Davis, S., A. Cornejo, R. Milford, and M. Nelson. 1989. "Guatemala Environmental Issues Paper." Washington, D.C.: Latin American Region, World Bank. De la Cruz, R. 1976. "Clasificaci6n de Zonas de Vida de Guatemala de Acuerdo al Sistema Holdridge." Guatemala City: INAFOR. Del Valle, J., 0. Melendez, E. Medina, A. Azurdia, and M. Basterrechea. 1990. "Incremento de la Productividad Agropecuaria y Conservaci6n de los Recursos Agua, Suelo, y Bosque." (Versi6n Preliminar), Unidad Sectorial de Planificaci6n Agropecuaria y de Alimentaci6n (USPADA), Ministerio de Agricultura, Ganaderla, y Alimentaci6n, Guatemala City. Denevan, W., J. Treacy, J. Alcorn, C. Padoch, J. Denslow, and S. Flores. 1985. "Indigenous Agroforestry in the Peruvian Amazon: Bora Indian Management of Swidden Fallows." In Man's Impact on Forests and Rivers, ed. J. Hemming. Manchester, Eng: MancheSter University Press. Direcci6n Nacional Foresta} (DINAF). 1988. Plan de Acci6n Foresta[ para el Ecuador: Diagn6stico del Sector Foresta!. Quito: Ministerio de Agricultura y Ganaderia. 234 MANAGING THE WORLD'S FORESTS Fearnside, P. 1986. "Spatial Concentration of Deforestation in the Brazilian Amazon." Ambia 15:74- 81. Food and Agriculture Organization of the United Nations [FAO]. 1988a. Production Yearbook, 42. Rome: FAO. _ _ _. 1988b. Trade Yearbook, 42. Rome: FAO. Garcfa-Gasques, J., and C. Yokomizo. 1986. "Resultados de 20 Anos de Incentivos Fiscais na Agropecuaria da Amazonia." XIV Encontro Nacional de Economia, ANPEC 2:47-84. Goodland, R. 1985. "Brazil's Environmental Progress in Amazonian Development." In Man's Impact on Forests and Rivers, ed. J. Hemming. Manchester, Eng.: Manchester University Press. Hayami, Y., and V. Ruttan. 1985. Agricultural Development: An International Perspective. 2d ed. Balti- more: Johns Hopkins University Press. Hecht, S., R. Norgaard, and G. Possio. n.d. "The Economics o:f Cattle Ranching in Eastern Amazonia." Graduate School of Architecture and Urban Planning, ·university of California, Los Angeles. Holdridge, L. 1967. Life Zone Ecology.· San Jose, Costa .Rica: Tropical Science Center. Instituto Nacional de Estadistica (INE). 1989. Guatemala, Poblaci6n Urbana y Rural Estimada por Departainento y Municipio. Guatemala City: INE. Instituto Nacional de Estadistica y Censos (INEC). 1982. Encuesta Nacional de Fecundidad. Quito: !NEC. International Bank for Reconstruction and Development {IBRD). 1990. World Development Report 1990. Oxford: Oxford University Press. Katzman, M. 1977. Cities and Frontiers in Brazil. Cambridge, Mass.: Harvard University Press. Macdonald, T. 1981. "Indigenous Responses to an Expanding Frontier: Jungle Quichua Economic Conversion to Cattle Ranching." In Cultural Transformations and Ethnicity in Modern Ecuador, ed. N. Whitten. Urbana: University of Illinois Press. Mahar, D. 1979. Frontier Development in Brazil: A Study of Amazonia. New York: Praeger Publishers. _ _ _. 1989. "Government Policies and Deforestation in Brazil's Amazon Region." Washington, D.C.: World Bank. Ministerio de Agricultura y Ganaderia (MAG). 1987. Informe Final de la Limitaci6n del Patrimonio Forestal. del Estado. Quito: MAG. Montenegro, F., and M. Durini. 1989. "lEcuador: Una Potencia Foresta!?" Quito: Corporaci6n For- esta! J.M. Durini. Moran, E. 1981. Developing the Amazon. Bloomington: Indiana University Press. _ _ _. 1983. "Government-Directed Settlement in the ]970s: An Assessment of Transamazon Highway Colonization." In The Dilemma of Amazonian D·evelopment, E. Moran, ed. Boulder, Colo.: Westview Press. Myers, N. 1984. The Primary Source: Tropical Forests and Our Future. New York: W.W. Norton. _ _ _. 1988. "Threatened Biotas: Hotspots in Tropical Forests." Environmentalist 8:1-20. Organization of American States (OAS). 1974. Situaci6n y Perspectivas Alimentarias de America Latina. Washington, D.C.: OAS. Prance, G. 1986. "Introduction to Tropical Rain Forests." In Tropical Rain Forests and the World Atmosphere, ed. G. Prance. Boulder, Colo.: Westview Press. Schneider, R., J. McKenna, C. Dejou, J. Butler, and ·R. Barrows. 1990. "Brazil: An Economic Analysis of Environmental Problems in the Amazon." Washington, D.C.: Latin America and Caribbean Region, World Bank. Schwartzman, S., and M. Allegretti. 1987. "Extractive Production in the Amazon and the Rubber Tappers' Movement." Washington, D.C.: Environmental Defense Fund. Smith, N. 1981. "Colonization Lessons from a Tropical Forest." Science 214:755--61. Southgate, D. 1990. "The Causes of Land Degradation Along 'Spontaneously' Expanding Agricul- tural Frontiers in the Third World." Land Economics 66:93-101. _ _ _ . Forthcoming. "Tropical Deforestation and Agricultural Development in Latin America." American Journal of Agricultural Economics. Southgate, D., and M. Basterrechea. In review. "Population Growth, Public Policy, and Resource Degradation: The Case of Guatemala." Ambia. FUEL WOOD PROBLEMS AND SOLUTIONS 235 Southgate, D., R. Sierra, and L. Brown. Forthcoming. "A Statistical Analysis of the Causes of Defor- estation in Eastern Ecuador." World Development. Southgate, D., and M. Whitaker. Forthcoming. "Promoting Resource Degradation in Latin America: Tropical Deforestation, Soil Erosion, and Coastal Ecosystem Disturbance in Ecuador.'' Economic Development and Cultural Change. Universidad Rafael Landivar (URL). 1984. Perfil Ambiental de la RepUblica de Guatemala. Report pre- pared for AID/ROCAP, Guatemala City. Valdes, A. 1986. "Impact of Trade and Macroeconomic Policies on Agricultural Growth: The South American Experience," Economic and Social Progress in Latin America. Washington, D.C.: Inter- American Development Bank. Whitaker, M. 1990. "The Human Capital and Science Base." In Agriculture and Economic Survival: The Role of Agriculture in Ecuador's Economic Development, eds. M. Whitaker and D. Colyer. Boulder, Colo.: Westview Press. Whitaker, M., and J. Alzamora. 1990. "Production Agriculture: Nature and Characteristics." In Agriculture and Economic Survival: The Role of Agriculture in Ecuador's Economic Development, eds. M. Whitaker and D. Colyer. Boulder, Colo.: Westview Press. World Resources Institute. 1990. World Resources 1990~91. Washington, D.C.: WRI. 10 Forest Valuation Randall Kramer, Robert Healy, and Robert Mendelsohn M any environmentalists as well as some economists believe that the world's forests are not valued properly in economic terms. Valuation problems exist for many of the diverse goods and services that forests provide. Some of these forest outputs are traded in commercial markets but are mis- valued. For example, government distortions prevent market transactions from reflecting the true scarcity cost of the traded output (e.g., below-cost timber sales in temperate countries and inefficient concession agreements for logging in tropical countries) (Repetto and Gillis 1988). Furthermore, externalities associated with destructive timber extraction practices can pre- vent timber prices from reflecting the true social returns to timber harvest. The so-called minor forest products such as medicinal plants, nuts, and rubber may be traded in informal markets or in little-studied formal mar- kets and hence not counted in data on the contribution of forests to national economies (Peters, Gentry, and Mendelsohn 1989). Another set of outputs from forests-environmental services such as carbon sequestering and watershed protection-are not traded in markets at all and so their economic value is often ignored. Even when environmental values are recognized, they may not be measured or used to promote effi- cient resource management. This undervaluation of forests has caused deforestation to accelerate and has impaired the ability of residual forests to supply useful environmental services. Undervaluation has also caused governments. to assign a low priority to the forestry sector because of its apparently low contribution to gross national product (GNP). In both developed and developing countries, forest acreage set aside as parks and reserves is growing. In the United States, government decisions to limit timber harvests in order to protect the habitat of threatened species 237 238 MANAGING THE WORLD'S FORESTS (e.g., spotted owl, red-cockaded woodpecker) or to set aside extensive roadless areas as wilderness have caused large amounts of timber revenue to be forgone. Similarly, many developing countries have recently set aside large tracts of forestland as parks and nature reserves. For example, Costa Rica has set aside 9 percent of its total land area (World Resources Institute 1988). Rarely if ever are these decisions subjected to economic analysis, in which the value of the intact forests is compared with value in alternative uses. The values of the world's forests, either as complete ecosystems or as composites of specific output flows, have not been measured precisely, and generally not at all. Without such information, it is difficult to determine whether decisions to augment or reduce forest area, or to emphasize one forest use over another, are economically justified. Yet sound resource valu- ation technologies are available, although some are expensive to implement. Most applications of these technologies to forest valuation have occurred in industrialized countries. Greater application of these valuation methods to forest resources would improve the base of information on the relative val- ues of market and nonmarket outputs of the forest. This improved informa- tion base can contribute to the sustainable management of the world's forests. The Importance of Forest Valuation Proper forest valuation is essential to the appraisal of forestry projects and of other projects that affect forests. Forestry projects, such as establish- ment of plantations or improved management of natural forests, can en- hance the level of goods and services provided by a forest but often change the mix of outputs. For example, replacing natural forest with a plantation may increase timber output, reduce biodiversity, and leave watershed func- tion essentially unchanged. Valuation can en~ble planners to compare out- puts that are enhanced with those that are i'educed by projects. Nonforest projects, such as agricultural land conversions, also can affect forest out- puts. Accurate and comprehensive valuation is essential to complete project appraisal. Valuation is also important for policy analysis. Undervaluation of for- ests and their outputs and services may create a policy bias in favor of competing land-use activities; overvaluation of forests does the opposite. In land allocation policy, for example, lack of appreciation of the multiple services provided by forest ecosystems may lead to the assumption that noncommercial forestland has no economic value and hence should be avail- able for conversion to low-value agriculture. Conversely, without informa- tion on values it is impossible to calculate the timber revenue forgone by setting aside forestland as a park or wilderness. An important related policy issue is whether forestland should be con- trolled by public entities, such as government agencies or state corpora- tions, or by private landowners. Information on the values of forests can FOREST VALUATION 239 help with this decision. If the bulk of the value of forests is in the produc- tion of marketed goods and services, then _forests probably should be pri- vately owned and managed; if most of their value is nonmarketed or public goods, some public role should be considered. A final use of forest valuation is in allocating research funds to forestry research relative to other types of research (e.g., forest versus field crop research) and in guiding allocation among various types of forest research (e.g., wildlife versus timber productivity). Improving the scope and accuracy of forest valuation is particularly timely, because many countries are formulating forest management and land classification and allocation plans. It is also important because in some countries the rate of liquidation of forest capital is rapid, raising fears that unwise and irreversible resource commitments will be made. Much pio- neering economic work on natural resource valuation is being done, and the best of this work must be made available to people now making private and collective decisions about the future of the world's forests. Reasons Why Forests Might Be Misvalued There are at least six reasons to suspect that forests might be misvalued: 1. Forests produce multiple products. Forests produce solid-wood products (lumber, veneer, poles), pulpwood, fuelwood, "minor forest products" (rattan, fruit and nuts, latex, gum arabic, medici- nal plants), and a variety of wildlife products. Indeed, even when a forest is manipulated to produce a single product or service, it is difficult not to produce others simultaneously. Many forest products are sold in regional, national, or international markets; others are sold only locally, and still others are not sold at all. The existence of organized markets not only determines the degree of public knowledge of market values (the price of 1,000 board feet of mahogany or a kilo of latex is more widely known than that of an armload of thatch or a unit of watershed service) but also determines whether changes in demand will induce a subsequent supply response. 2. Forests produce many nonmarket services. Forests also produce service flows, including soil stabilization, watershed services, cli- mate regulation, wastewater treatment, protection of gene pools, forest grazing, recreation, and amenity and existence values. Some of these forest service flows may be marketed or are capable of being marketed (e.g., recreation and forest grazing), but others are public goods. These public goods may have considerable value to society but no private value in terms of an effective return to a producer. 240 MANAGING THE WORLD'S FORESTS 3. Timber and some other forest outputs result from biological processes that require a very long time. The standard economic approach to valuing resources over time is to use a discount rate that reflects society's opportunity cost of capital, but this approach is contro- versial among noneconomists because it weights future genera- tions' needs less than the needs of current generations. In addition, ecologists express concern that forest harvesting or forest land use conversions may lead to irreversible changes, including soil degradation and species extinction. Even if a future owner, or a future government, assigned a high value to these characteristics, they could not be re-created (Krutilla and Fisher 1975). This is particularly true in the case of species extinctions, because the genetic material of an extinct plant or animal cannot be regained, at least not with current technologies. 4. Virgin, no-cost stocks of many forest products, including timber, wildlife, and water, are available, often under open-access conditions. The existence of large virgin stocks owned by various producers means that the price of some forest products may have little rela- tion to the costs of reproduction. Virgin stocks also may initially discourage investments by others in forest plantations and refor- estation. As the stocks begin to decline, however, royalties or scar- city rents should rise (where there are well-functioning markets), thereby encouraging reforestation. A more difficult problem arises in those areas where forest owners, either private or public, are unable to control access to the virgin resource stock. Demand leads to overexploitation. Prices will tend to be low, and other producers will have no incentive to invest in producing the forest products elsewhere. 5. Scientific data on forest production functions are lacking, particu- larly in tropical countries, and knowledge of the market for many forest products, or even of their potential economic uses, is limited. Although there is considerable knowledge of how to manage profitably many types of forests found in developed, temperate countries, there is far less information on how to manage tropical forests. In the tropics, the diversity of tree species (often more than 100 per hect- are) has caused timber harvesters to concentrate on only a hand- ful of species. For the vast majority of species, information on use of the potential management alternative for forest land is lacking. Only recently has much research been done on methods for sustainably managing natural stands in the tropics (Buschbacher 1987; Hartshorn 1989). Moreover, very few tropical species have been tested to determine their potential as plantation crops. FOREST VALUATION 241 There is also much uncertainty about possible economic uses of many nonwood forest products. A number of tropical fruits, nuts, and medicinal plants sold only in local markets, or not at all, may have export potential, provided research is done on their properties and proper market development takes place. For ex- ample, in India nearly 2,500 plants are used for medicinal pur- poses (Myers 1984). 6. Many forest products and services are important to the livelihood of the rural poor, a group whose welfare is a major object of public policy but whose demands have little weight in organized commodity markets. An estimated 300 million people worldwide live in and around forested areas, and many make a substantial part of their liveli- hood from the forest. They include some of the world's poorest people, many of them tribal groups outside the cash economy. Nearly 2 billion people, most of them poor, use fuelwood as a primary energy source. An unknown proportion of their fuelwood production and consumption takes place outside the cash economy and thus is never recorded in national income accounts. Of course, productive systems other than forests also have valuation problems-for example, agricultural use of land also produces watershed, wildlife, and open-space services, and industrial activities produce a variety of positive and negative externalities. But even if considered on a continuum, the long production period, multiple products, nonmarketed products, huge virgin stocks, and other characteristics make forests unusually susceptible to misvaluation. Economic Concepts of Natural Resource Value Economic concepts of the value of natural resources begin with indi- vidual willingness to pay. Economists believe that all consumers have their own values and their own tastes and preferences by which to judge the relative merits of orie good or service over another. Economists believe that aggregate or social values can be derived by adding these individual values. In addition to measuring aggregate values, it is important for analysts to identify gainers and losers from natural resource projects and policies. The World Bank has done considerable work on assigning social weights, but as yet no consensus has emerged about how to introduce social weights formally into economic project analysis (Squire and van der Tak 1975). Even if social weights or shadow prices are not assigned to benefits received by different groups in a society, it seems worthwhile to identify the gainers and losers from development activities and policies, because policymakers are keenly interested in such information. However, this section deals with aggregate values, not the distribution of environmental benefits and costs. 242 MANAGING THE WORLD'S FORESTS Monetary Values Because forests, especially tropical forests, provide such a diverse set of outputs, it is critical to develop an index by which to judge the rel.ative importance of each possible output. Invarilably, different management schemes will provide more of one and less of another. To judge the relative merits of each scheme, one needs a common index by which to judge each combination of outputs. In this chapter, we compare the relative merits of each good or service to a monetary unit (arbitrarily U.S. dollars). In doing so, we do not intend to limit values to commercial ends. For example, people might be willing to pay to preserve an endangered species of monkey, or they might be willing to pay to keep nearby forests in their original condi- tion. We use monetary values simply as an index by which to judge the many market and nonmarket services provided by the forest. If monetary values are used as a common denominator, the cost and benefit streams emerging from various policy or management strategies can be compared. Financial versus Economic Values One important distinction that has been made in the development lit- erature is the difference between financial and economic values (Gregerson 1985). With financial values, only the costs and revenues that accrue to a specific institution are counted. In contrast, with economic values, all costs and revenues to a society that result from a decision are counted. Thus when a bank gives a loan, it may count only revenues and costs that must be paid from that loan. Similarly, when a forestry department manages land, financial analyses may include only the costs and revenues that the department will receive from its land. Costs incurred by others are gener- ally not included in financial analyses. Economic values, in contrast, at- tempt to include all the costs and revenues from a decision. For example, effects that occur off the land being managed, such as the downstream damages from sloppy harvest methods, would be included as an economic cost. Changes in services that are valued but not necessarily paid for also would count in an economic analysis. For example, if local residents his- torically have grazed their livestock in a forest and this practice was affected by a decision, the change in livestock value would be measured under economic values even if the local people historically did not pay for this service. The U.S. Forest Service, as part of a major planning effort for national forests, has developed estimates of forest values using both of the foregoing measures: receipts (user fees and stumpage sales) and economic value (U.S. Forest Service 1990). Because the government does not always charge the full value of the resource, these values can differ greatly. For example, in one region the fee charged under a permit system for grazing on federal lands in 1989 was $1.86 per animal unit month (AUM). The fair market FOREST VALUATION 243 value estimated for this use of forestland was $6.11 per AUM. A decision to remove one A UM would result in a financial loss to the government of $1.86 and an economic loss to society of $6.11. Final versus Intermediate Valuation This chapter focuses on valuing goods and services in the forest, but available data for valuation often apply to final products, which may in- clude processing costs or other value added. Sometimes, the value of inter- mediate products is of interest. For example, a forester who is trying to decide what silvicultural practice to try in a forest might want to know what value to place on the right to cut standing trees. In a competitive market, the forester wants to know a definite relationship between the ob- servable mill price and the stumpage price. The stumpage price will be equal to the mill price minus the expenses it takes to cut the trees and deliver them to the mill. With knowledge about the values of final products, the forester can deduce the value to be assigned to intermediate products by subtracting the production costs. The same logic can be applied to environmental goods and services. For example, suppose a silvicultural practice reduces the number of mice in a forest, which in turn reduces the number of owls. If people valued the reduction in owls, one could measure the dollars they would pay not to lose the owls. Although people may not care at all about the mice for their own sake, the mice indirectly have value in their ability to "produce" owls. Thus, using the ecological interface among species, the analyst can place values on many lower-order species simply through the role they play in providing species and environments we care about. It is important not to use final prices to value intermediate goods. For example, because of their availability, it is tempting to use export prices to value raw materials in the forest. Export prices are an accurate measure of the product delivered to a port. Frequently, however, substantial harvest, transport, and preparation costs are incurred before the raw product can be exported. To measure the value of the material in the forest, these costs must be carefully subtracted from export prices. For example, timber har- vest and transport costs can be 40 percent of mill price. The value of the timber in the forest is thus only 60 percent of mill price. The same principle applies to other forest market products such as medicinals, fruit, rattan, and latex. Flows versus Stock It is also important to distinguish between valuing commodity or ser- vice flows and valuing a stock. A stock accrues from savings and is used traditionally in economics to enhance production and thus yield higher flows over a future time period. The value of a stock can be measured by the 244 MANAGING THE WORLD'S FORESTS value of the long-term flow of goods possible from that stock. A standing forest is a stock. If one harvests the forest, the logs coming from the forest are a flow. For example, suppose a large forest produced 1 million board feet of timber each year at US$0.10 a board foot. The value of the flow would be $100,000 per year. The value of the forest would be the sum of the values of all future flows. However, because future flows are delayed in occurring, they must be discounted to reflect the value of time. At a value of time (interest rate) of 5 percent, the present value of an infinite set of annual flows of US $100,000 is US$2 million. Thus the value of the forest stock is US$2 million. Unfortunately, although flows at least of market goods are measured readily, stocks are rarely valued. This imbalance in information induces a temptation to convert stocks of unknown value to flows of measurable value prematurely. Even if their actions reduce their overall stock or wealth, some countries are tempted to engage in rapid harvest policies simply to increase the magnitude of measured flows. FOREST PRODUCTS AND OUTPUTS Wood Products The output of forests is conventionally measured by their production of wood products. These outputs are almost always measured in physical terms, rather than in terms of financial values. This practice produces some impor- tant measurement problems because physical quantities of, for example, pulpwood, have a financial value very different from the same volume of veneer-quality wood. The product identification problem is reduced some- what by the fact that a fairly large amount of detail is available regarding types of physical output, although within some product categories value varies greatly with respect to species. The most difficult problem for forest valuation is translating product values to stumpage values. Some forest products (e.g., paper) embody a great deal of processing, so that only a portion of their market value can be attributed to the value of wood. Moreover, product prices include extrac- tion costs, transportation costs, and return to risk and entrepreneurship. What is really needed for valuation is the value of stumpage. As with land, this value is a function of site quality, topography, and accessibility, and it varies from site to site. Stumpage is in most cases a relatively small fraction of the value of final wood products, given the allowance that must be made for harvesting, trans- portation, and processing. Ewing and Chalk (1988) calculated that in 10 major regions producing bleached softwood kraft pulp, the cost of logs delivered to the mill averaged 34 percent of the final product. There was a considerable range, from a low of 20 percent in Chile to a high of 54 percent FOREST VALUATION 245 in Finland. Ewing and Chalk calculated (for four countries) that stumpage ranged froni 33 to 50 percent of delivered pulpwood cost. These figures suggest that stumpage is between 6 and 27 percent of the value of pulp production. Stumpage is expected to be a somewhat higher proportion of the value for sawnwood products because there is less processing involved. Ewing and Chalk (1988) estimated that stumpage constituted only a small fraction of the value of hardwood sawn logs in Ivory Coast (22 percent), Indonesia (21 percent), and Congo (10 percent) and only a moderate proportion of the value of softwood sawn logs in the northwestern United States (35 percent), Chile (36 percent), and New Zealand (54 percent). Given the difficulty of using the value of wood products rather than the stumpage price to value annual wood output, what sort of product data exist? The United Nations Food and Agriculture Organization (FAO) pub- lishes annual figures on physical quantities of wood products output, by country, and physical quantities and value data for the international wood products trade (FAO, 1986b). FAO has also published an estimate of the total value of wood products output for the world, estimating its 1985 value (in 1980 prices) at $303 billion-around 2.5 percent of world GDP (Alexandratos 1988). The origin of this figure is unclear, but it was probably produced by applying export prices to the published quantity data. Exports are estimated to have been $50 billion in 1985.1 It is doubtful that stumpage represents as much as 25 percent of the value of wood products, giving a likely maximum world value of US$75 billion for stumpage, of which no more than US$12.5 billion is accounted for by export markets. Another way of estimating stumpage values is by direct observation of royalties and license fees obtained by government forestry departments. The U.S. Forest Service, which controls more than 35 million hectares of commercial forestland, annually (1989) receives $910 million in revenues from timber sales, with stumpage sold by competitive bids .(U.S. Forest Service 1990).2 Repetto and Gillis (1988) offer figures on license fee receipts in several developing countries, but note that collections are generally well below the level necessary to extract the full economic rent attributable to the timber stock. 1 For comparison, 1988 world production and trade figures for some major agri- cultural commodities are maize (production value US$52 billion, trade value US$9 billion); rice (production US$156 billion, trade US$4 billion); cotton lint (production US$28 billion, trade US$8 billion) (FAO 1988a, 1988b). 2 Because of the costs of administering timber sales, actual net revenues to the government are somewhat lower. 246 MANAGING THE WORLD'S FORESTS Other Products and Services Forestland and its associated resources constitute a multipurpose natu- ral resource system. In addition to wood products, a number of other valu- able products and services are produced, onlly some of which are sold in commercial markets. • Minor forest products, including thatch, rattan, fruits and nuts, gum arabic, chicle, latex, and medicinal plants (Debeer and McDermott 1989; Falconer 1989). In some cases the products are sold for cash on organized markets, and they can even represent major export com- modities (e.g., rattan in Indonesia, gum arabic in Sudan). But many minor forest products are sold only in local markets (Peters, Gentry, and Mendelsohn 1989) or are used directly by the harvester. • Watershed services. Forests help to determine water yield (forestland almost invariably averages lower water yield than the same land when cleared) and to reduce land slippage on slopes. Some authori- ties contend that the role often claimed for forests in "smoothing" fluctuations in water yield is nonexistent and that their role in flood control has been exaggerated (Hamilton 1985). • Recreation and tourism. In both developed and developing countries, forests provide a setting for outdoor recreation that ranges from pic- nicking and walking in forests near urban concentrations to long- distance excursions by national or international tourists in search of settings with exceptional qualities. The number of visitors to forests can be quite large. In the United States, for example, the U.S. Forest Service (1990) reported some 260 million "visitor days" (in 1989) of recreational use on National Forest lands. Nature-oriented tourism (sometimes called ecotourism) is among the most rapidly growing segments of the international tourism industry, and forested areas, particularly tropical moist forests, are important destinations (Laarman and Durst 1987). • Wildlife products. Forests support a wide variety of wildlife, including many species used by humans for food, hides, or other products. Wildlife harvesting also has a strong recreational aspect, and many recreational users of forests come specifically for hunting, trapping, or fishing. In some low-income countries, wildlife harvesting pro- vides rural people with an important source of protein and a wel- come source of cash income. • Protection of biological diversity. Forests preserve species or ecosystems that may have current or future commercial value. Several species with known or probable commercial value have been rescued from the brink of extinction; these include teosinte (the ancestral corn plant, recently rediscovered in Mexico) and the dawn redwood (rediscov- ered in China after it was considered extinct). Several species and subspecies of pine growing in very restricted ranges in Mexico and FOREST VALUA T/ON 247 Central America are in danger of extinction because of harvesting and land clearing. The genetic material they contain might be of great value in improvement of important commerdal pine species. Many other uses of genetic material found in forests have yet to be discov- ered. For example, although less than 1 percent of tropical plant spe- cies have been screened in South America, more than 250 have been identified as having potential for human birth control (FAO 1986a). • Climate regulation. Forests have a relatively low surface albedo, a measure of the degree to which the earth reflects sunlight. Convert- ing a forest to an open field tends to raise albedo. Forests also are important in the evaporation and transpiration of rainfall, and large forested areas may affect the amount of rain that falls on them and on adjoining lands (Salati 1987; Lean and Warrilow 1989). Unfortunately, the relationship of albedo and of evapo-transpiration to climate is quite complex, and is poorly understood, so that it is not easy to predict the climate impact of large-scale forest clearing. • Carbon sequestration and release. Trees are approximately 50 percent carbon by dry weight. The annual net clearing of forests releases some of the carbon tied up (sequestered) in trees and forest organic soil material, especially when they are burned. The total amount of carbon sequestered in terrestrial vegetation has been estimated at 557 to 827 billion metric tons (Woodwell 1984). The annual release of carbon from fossil fuels and industrial processes is estimated at 5.5 billion metric tons (World Resources Institute 1988). Releases of even a small fraction of the carbon stored in terrestrial vegetation can ex- acerbate the carbon dioxide problem substantially. Conversely, in- creases in forest stock can offset anthropogenic carbon dioxide emis- sions. For example, establishment of 465 million hectares of tree plantations, at a cost of US$186 to US$372 billion (Sedjo 1989), would offset current emissions of carbon across the world. • Other values. In many cultures, forests play an important role in art, literature, and religion, and they offer material for scientific study and for education. In some cases these values are "nonuse" values, in that they are enjoyed by many people who do not physically visit the forest. These values can also be "nonconsumptive," in that they do not involve the physical removal of any substance from the forest. Our scientific knowledge about the production of these products and services by particular forests is far from complete. Most forests have not been inventoried for the occurrence of minor forest products or of wildlife, and the distribution of biodiversity has not been fully inventoried, particu- larly in the tropics. Of the estimated 5 to 30 million distinct species world- wide, only about 1.4 million have thus far been identified (Wilson 1988). In tropical forests, the diversity of tree species (sometimes 100 or more differ- ent species are found in a single hectare) means that even timber inventories are incomplete. Scientists also have much to learn about specific mecha- 248 MANAGING THE WORLD'S FORESTS nisms through which forests affect the environment. This is particularly true of forest-climate interactions and the effects of various degrees of forest clearing on watershed services. Compatibility of Forest Uses Not all forests can provide all forest outputs. In fact some forest values are incompatible, while others are complementary. If a forest is maintained primarily for watershed values, this use is fully compatible with providing carbon sequestering, wildlife habitat, or scenic beauty (which could gener- ate both recreation value and nonuse value). Other values may have differ- ing degrees of compatibility. Timber production can be compatible with recreation, but these joint products may limit volume or conditions of har- vest. Similarly, forests can be used for both production of minor forest prod- ucts and provision of wildlife habitat, but the intensity of use may have to be regulated. Other uses are clearly incompatible-for example, protection of biodiversity and intensive timber harvesting. Managing forests to provide multiple outputs may involve patterns of specialization rather than attempting to provide all values in all forest areas. The notion of specialization by function is common in economics. Just as nations and regions specialize in the goods and services they produce in accordance with the principle of comparative advantage, some forest areas are best suited for one or more of the array of possible values (Howe 1979), The appropriate specialization will depend on physical characteristics such as topography and rainfall and the structure of the demands for the prod- ucts and services. Some forests do not provide much timber or many minor forest products, but may be well suited to providing recreational services. Other forests are ideal for timber production. Of course, demands for alter- native uses are important for determining optimal use patterns of different forest areas. A forest might have favorable physical characteristics for pro- tection of biodiversity, but might be better used for fuelwood production if located close to densely populated areas. MEASUREMENT OF FOREST VALUES The multiple potential outputs from a forest pose a challenge to valua- tion. Traditionally, only a fraction of the outputs people enjoy from a forest have been quantified. In the past few decades, however, improvements in methodology allow for a substantial expansion of this list. Research is even under way concerning measurement of the nonuse or existence value of forests. Although continued research on methodologies will result in in- creasingly accurate and reliable estimates, methods to measure the value of most forest uses are available today. These methodologies will provide suit- able values for use in forest planning and management. FOREST VALUATION 249 Which valuation methodology is most appropriate to apply depends on the nature of the good being valued. The myriad products from forests can be divided into four groups: market goods, potential market goods, public goods, and nonuse values. Market goods (such as sawnwood, pulpwood, and latex) are actively traded, familiar, and generally well defined. Potential market goods could be traded in markets but may not be because they are distributed by government decree or culturally prescribed rules, or because access to markets is inadequate to warrant their transportation. Many low- quality construction materials and forest foods are potential market goods. Public goods such as water quality, recreation, and carbon sequestering are jointly consumed products or services that involve many independent consumers. The distinguishing features of public goods are that many inde- pendent people can consume them simultaneously and that it is difficulr or impossible to exclude additional users. Unless these simultaneous consum- ers can act as a single buyer, they have difficulty participating in a market for public goods. Thus public goods often are not traded in markets and often are undervalued. The final set of goods from forests consists of nonuse goods. The consumer of nonuse products does not actually come in contact with the good. In fact, there may be no apparent connection between the consumer and the resource. Nonuse values are also called existence values and can be applied to species, habitats, or ecosystems. Market goods can be valued using traditional economic measures such as market prices and consumer surplus. Potential market goods can fre- quently be measured by analyzing the market value of close substitutes that happen to be traded actively. The valuation of public goods and nonuse are more difficult and require more sophisticated valuation tools, which are discussed later. Market Goods The most reliable measures of value are for market goods sold in com- petitive markets. The values of marginal changes in the quantity of market goods are observed every day in the market prices for these goods. People are constantly deciding how much each of these goods is worth. Consumer Surplus Nonmarginal changes in quantity are slightly more difficult to measure, because large changes in quantities relative to market averages will tend to change prices. An assessment of whether a change is large depends on the market context. If the market for timber has an average flow of 2 billion board feet per year, a change in flow of 1 million board feet is small. If a market for fruit averages 5 tons, a change of 1 ton is large. In general, any large increase in the supply of a good, other things remaining equal, will cause its price to decline. Using current price to esti- 250 MANAGING THE WORLD'S FORESTS mate the value of large changes in supply will thus overestimate value. Using current price to estimate large decreases in supply will underestimate values. Accurate valuation of nonmarginal changes requires the use of "con- sumer surplus" (the area underneath a demand function), which takes into account the price change caused by a large change in supply. Each unit of the good that is changing is valued at the price that would have occurred if that quantity had been offered in the market. For many years, natural resource observers have been confused about the correct use of consumer surplus. Analysts have expressed concern that some resources are valued using consumer surplus whereas others are val- ued using just current price. For example, suppose one were considering cutting a vast Brazil nut forest that contained 35 percent of the remaining Brazil nut trees and 2 million board feet of timber. Suppose, further, that the relevant market for timber averaged 2 billion board feet per year. In this case, the timber should be valued using current price, but the Brazil nuts should be valued using consumer surplus. The additional timber resulting from the decision is unlikely to affect timber prices, but the potential loss of a third of the Brazil nuts would certainly affect nut prices. The use of con- sumer surplus is unrelated to whether a good is a market good or whether the consumption will preserve the forest or destroy it. Consumer surplus is appropriate whenever the proposed change involves a nonmarginal change in that market versus a marginal one. Potential Market Goods Potential market goods are more difficult to value than goods actually traded because competitive prices cannot be directly observed for them. A number of possible valuation exercises are possible. When these goods are traded in nearby competitive markets, it is sometimes possible to use those observed prices for valuation. For example, if certain fruits were being con- sumed without being sold in a rural area and these fruits were being sold in a local competitive market, the prices in the nearby market could serve as a reasonable measure of value. In using this approach, however, it is impor- tant to take account of transportation costs required for consumption. Sometimes, low-quality fruits and materials are simply not sold in markets but are nonetheless consumed by local (and often poor) residents. One method of determining the value of these products is to seek counter- part products in the market which these people would have to buy if they did not consume the product in question. That is, one valuation technique is to determine the market price of close substitutes for a product. Care must be taken in this approach to use only close substitutes. Low-quality products must not systematically be assigned the prices of higher-quality counterparts. If quality differences exist, some attempt must be made to adjust prices for these differences. For example, poles are often used in local FOREST VALUATION 251 construction instead of lumber. The equivalent amount of lumber that would have to be purchased to build a home suggests the value of the poles. However, the estimate should take into account the fact that the lumber might be more attractive and effective than the poles for construction purposes. Distorted Markets Sometimes reported prices are not the outcome of competitive markets. For example, some transactions between governments and single buyers are conducted at prices that are not competitive. Similarly, transactions between parts of a vertically organized firm are often arranged for accounting conve- nience at prices that do not resemble competitive rates. Sometimes these prices are altered for a public purpose, such as to stimulate economic devel- opment, and sometimes they are distorted to avoid a tax or charge. For example, if a government charges a firm a royalty based on a percentage of the value of logs taken from public land, the timber harvesting firm has an incentive to understate the value of the logs sold to the mill. If the mill and the harvesting firm are jointly owned or are in collusion, they can under- estimate the value of the cut logs and pay a low royalty. The price of this arranged transaction does not reflect the value of the logs. When sales are between established customers or businesses, it is sometimes possible to find comparable prices in nearby competitive markets. Alternatively, costs can be subtracted from final product prices to estimate raw material values as discussed earlier. Finally, the extent to which a monopolist or monopsonist could have distorted market prices, given the shape of the supply and de- mand functions, can be estimated. With this information, the prices in a noncompetitive market can be adjusted to yield true marginal values. Public Goods Public goods are rarely sold in markets; there is generally no market for scenic beauty, water quality, or wildlife populations. These goods are jointly shared or enjoyed by multiple parties at the same time, prohibiting their effective management in a market setting. To value public goods, it is nec- essary to try to measure the demand function each person has for varying quantities of the good. The value of different quantities of the public good can then be estimated by adding up how much each person values the public good. Determination of the demand for public goods is at the heart of nonmarket valuation. Unfortunately, there are very few examples of public goods being valued anywhere other than in developed countries. Four basic approaches have been successfully applied to value public goods: hedonic, travel cost, avoided cost, and contingent valuation. 252 MANAGING THE WORLD'S FORESTS Hedonic Method The hedonic method was designed to explain the observed variation in price of heterogeneous goods such as housing, cars, and other consumer goods. The model assumes that variations in prices for these goods within the same market must be due to variations in quality or to other observable characteristics of the good. The hedonic method offers a way to value envi- ronmental services provided by forests. For example, some services pro- vided by forests, such as proximity to hiking trails, fishing streams, and · game, can affect the value of proximate properties. Bedonie studies of resi- dences have consistently found that proximity to underdeveloped public spaces increases property values (Brown and Pollakowsky 1977; Freeman 1979). When Brown and Pollakowsky explored the value of homes near lakes with public spaces around them, they found that homes near such spaces have higher values than homes farther away. Moreover, the bigger the strip of public land around the lake, the more valuable was proximity. Another prominent application of the hedonic method could be to value access to higher-quality water. For example, suppose that maintaining a natural forest instead of using the forest for timber increased water quality. Suppose, furthermore, that substantial agricultural activity takes place down- stream of the forest. If high-quality water increases the value of the farms, they should enjoy higher net income and have higher market value than otherwise similar farms that have lower-quality water. An hedonic study of farms in both high- and low-quality drainage areas could control for other differences between farms and determine the marginal contribution (dollar value) of water quality. The value of forest protection of water supplies is approximately equal to the net change in the values of all the affected pri- vate properties downstream. Travel Cost The travel cost method values a recreation site (a forest destination) by estimating the demand for access to the site (Clawson and Knetch 1966). The travel cost model uses the transportation cost to a site coupled with any entrance fees, as a measure of the price of access. Given the price of a visit, the demand for access to a forest can be estimated by looking at how the number of trips taken varies with the transportation costs per trip. As expected, people from greater distances tend to come less frequently, so the demand for visits is downward sloping. With high-valued sites, this demand curve shifts out and is steep, implying that people come often to the site and are willing to come a great distance. With common sites, the demand curve shifts in and is flatter. People who live nearby may use these common sites a lot, but few people would come a great distance to use them. The value of a site is equal to the area underneath the demand function for visits to that site. When elimination of any specific site is being consid- ered, all the visits to that site will be eliminabed. Because elimination is not FOREST VALUA T/ON 253 a marginal change, it must be valued using the consumer surplus associated with the demand for visits to that site. Because the price of a visit is being changed from its current level essentially to infinity, the consumer surplus is the value of what is lost if the site is removed. High-value sites that many people visit, some coming from a long distance, tend to have large con- sumer surplus values. Common sites, which people are not willing to pay much more than their current travel costs to visit, tend to have low con- sumer surplus values. The simple travel cost method is an effective and proven way of valu- ing the lost recreation use that would result from the elimination of any specific measured site. For example, Grandstaff and Dixon (1986) applied the travel cost method to value the benefits associated with a city park in Bangkok. Using consumer surplus, they estimated the value of the park in 1980 at $6 million. Similarly, Sorg and Loomis (1984) estimated the value of cold-water fishing in the Pend Oreille River in Idaho. From a survey of fishermen, they estimated the relationship between trips taken and distance to the site. The researchers found that the log of trips per capita fell by 0.024 per dollar of travel cost. Taking the area underneath this demand function yields an estimated value for this river of US$9,412 per year. Given that this flow is likely to continue indefinitely, the value of the fishing in the river is the discounted sum of all future harvests. Discounting the annual value using a 10 percent discount rate yields a present value of US$94,120. Many times, however, the land planning decision concerns not whether to clear away a forest and lose it permanently, but whether to change it. For example, planners may want to change the forest from an old-growth forest to a younger natural forest, or perhaps from a natural stand to a plantation. To value the change in the quality of a site, one could perform a simple travel cost analysis on the original site, change the site, and perform an analysis on the new site. The drawback of this technique is that the value is evident only after the fact. But many land management agencies find them- selves facing the same decisions over .and over, and this simple before-and- after approach would at least gradually uncover the value of agency actions. Generalized Travel Cost The generalized travel cost model, which compares the characteristics of the different sites, can be used to explain why some sites are visited more frequently than others. If an attribute causes trips to increase given a certain distance, that attribute is increasing the value of the site. The precise contri- bution of the attribute could be measured by the increase in the consumer surplus associated with adding the attribute. For example, it is possible to create a generalized travel cost model for fishermen and estimate how much their demand function shifts if catch rates increase. The increased consumer surplus created by going from an area with 1 catch per hour to one with 1.5 catches per hour could be estimated directly from the demand model. Ex- 254 MANAGING THE WORLD'S FORESTS amples of generalized travel cost include studies of water quality in the United States by Vaughn and Russell (1983) and Smith and Desvousges (1986) and urban recreation centers in Nigeria lby Durojaiye and Ikpi (1988). The generalized travel cost model has the same underlying philosophy as the simple before-and-after experiment discussed earlier, but because different sites are used rather than a single site that undergoes change, this model often suffers from unwanted variation. That is, the sites with and without the change may have other attributes that are different or they may have different substitutes. For example, areas with good hunting often are remote and tend to be near other areas with good hunting. Adding one more good-quality site near the others would be worth less than creating a good site where none now exists. The generalized travel cost model would value a high-quality site where it currently occurs (near others) and, in this instance, would undervalue the benefits of creating a new high-quality site where none now exists. Multiple-Site Travel Cost A third travel cost approach that can sometimes be used to value sub- stantial changes in site quality is the multiple-site travel cost model. The multiple-site model explicitly models the demand and value of a particular type of site. All available sites are grouped into a limited set of types. The technique then estimates the demand and value of each type of site. If one type of site represented the original site and another type represented what the site would look like after the change, this model could be used to value site changes. One would simply determine the value of adding a site of the new type and losing a site of the old type. If there are many differences between the types of sites, this technique cannot specify which quality is important. For example, if natural lakes tend to be surrounded by natural forests and have constant depths, whereas man-made lakes are surrounded by mountains and have fluctuating depths, the technique cannot determine the separate contribution of each character- istic. The multiple-site approach can value only the net effect of all the differences between the types of site. There have been examples published of the multiple-site model: the Burt and Brewer (1971) model of Arkansas lakes and the Cichetti, Fisher, and Smith (1976) analysis of California ski resorts. Another way to value site characteristics is to use a quality-quantity travel cost model across multiple sites. As with the generalized travel cost model, the analyst assumes that choices made across sites are due to ob- served site characteristics. The value of site characteristics can then be teased from the choices that different people make across available packages. There are three major variations of this model: discrete choice, characteristics models, and hedonic travel cost. All three models assume that the allocation of trips to different destinations is influenced by differences in cost and quality. FOREST VALUATION 255 The "discrete choice" and "characteristics" travel cost models make strong assumptions about the nature of substitution between quantity and quality across sites. Both types of models then estimate the value of site character- istics by examining the share of total visits that individual destinations re- ceive. For example, a recreational fisherman may be observed to choose site A slightly more than site B if site A has slightly higher catch rates, even if site B is slightly closer. From such choices, it is possible to infer the value of higher catch rates. The estimations of the characteristics model and the dis- crete choice model can be identical, depending on the functional forms as- sumed: In general, the discrete choice and characteristics models appear to perform best when there are few choices and the substitution among choices is similar. Good examples of the discrete choice model include Smith and Kaoru (1986) and Caulkins, Bishops, and Bowes (1986). The characteristics models have been developed by Morey (1981, 1984). The hedonic travel cost model attempts to estimate the marginal value · of site characteristics by examining the extra distance that people would travel to enjoy a site with a little more of some preferred characteristic. The model estimates the marginal value of site characteristics for each origin (residential area). By examining people across different residential origins who face different choices, it is possible to estimate the demand for site characteristics. By relying on the observed choices that people make be- tween cost and quality, the hedonic travel cost model is like the discrete choice and characteristics models. The hedonic method appears to be most effective when there are many choices and substitution across characteris- tics and sites varies. Successful applications of the hedonic travel cost model include Brown and Mendelsohn (1984), Englin and Mendelsohn (1991), Mendelsohn (1984), and Smith and Kaoru (1986). For example, Englin and Mendelsohn estimate the value of forest char- acteristics in wilderness areas of Washington State. People traveled farther to avoid clear-cuts and to walk through old-growth stands. Given an as- sumed cost per mile of US$0.25, people were observed tci be paying US$2.60 per mile to walk through old-growth stands and US$0.60 to avoid a clear- cut. Views were worth much more, about US$6 per trip. By comparing people who face different prices, the researchers observed that the demand functions for many of these characteristics' are rather steep. What people would pay rather than lose these characteristics outright is thus quite high. For example, people would pay US$71 per trip rather than lose old-growth completely. They would pay US$254 per trip rather than lose all views. Replacement Cost Another distinct approach to valuing public goods is to determine the cost of duplicating a service that the resource might provide (Dixon et al. 1986). For example, if a marsh provides a cleansing service, this service could be valued in. terms of the cost of an equally effective sewage control program. Similarly, if a marsh provides flood control by absorbing flood 256 MANAGING THE WORLD'S FORESTS waters, this service could be valued in terms of the cost of a system of dams and levees that would achieve the same end. This approach is accurate only when the service will clearly be purchased. For example, if the sewage system would not be purchased in the absence of the marsh, the cost of avoiding the sewage system would overestimate the marsh-cleansing service. Contingent Valuation A final valuation technique that has received much attention in the past two decades is the contingent valuation method. Respondents are asked their opinion of the worth of different goods and services, and their answers are averaged. The contingent valuation method is initially appealing be- cause it appears to be simple to design and to analyze and it appears to be highly flexible. Whatever analysts think they want valued can be formu- lated into questions, including the value of goods and services that are not sold in markets. The technique always provides an answer, and the exper- tise required to apply the technique appears to be minimal. Experience with the contingent valuation technique has revealed that attitudinal questions such as these are more complicated than originally conceived (see Cummings, Brookshire, and Schulze 1986 and Mitchell and Carson 1987). Perhaps the most important lesson that has been learned about contingent valuation is that the object being valued has to be well described. Questions that provide poor descriptions of the good being valued yield dubious results. For example, suppose people in the United States were asked, "How much would you be willing to pay to prevent tropical defor- estation?" This question is too vague to be understood: In which country will the deforestation be prevented? What is the cause of deforestation? What would be saved as a result of preventing the deforestation? Questions that value vaguely defined goods such as this elicit what psychologists call "labile" values (Fischoff, Slavic, and Lichtenstein 1980)-values that are easily influenced by the form of the question and the latest news flash. A second lesson learned in the early contingent valuation literature is that people must be familiar with the good and place acquisition of it in a market context. Many times, people may believe that the good does not need to be purchased because they already "own" it. That is, they believe that their property rights to the good are well defined and therefore it is unnecessary for them to purchase it. In this case, many people will not participate in the valuation exercise or will give distorted responses toques- tions. For example, people are asked how much they would be willing to pay for licenses to hunt on public lands on which no license is currently required. If people objected to buying licenses at all, they may answer zero. The zero, however, does not imply that they value hunting at zero but rather that they object to having to pay for it. Similarly, people may be asked whether they are willing to pay for porpoises to exist in the Pacific Ocean. People may care about the marine mammals but have no idea what FOREST VALUATION 257 they would be willing to pay for them to exist because such a payment has never before been suggested to them. Other anomalies are associated with the contingent valuation method. Whenever the survey method has asked what people are willing to pay (WTP) for a good and what they would be willing to accept (WTA) rather than lose it, the WTA responses exceed WTP responses (for a review see Knetsch 1990). The difference is often a factor of five. If these questions were just trying to determine the value of a marginal good, they should have yielded similar results. The exact reason for this consistent discrepancy is not known. People may understate their WTP values for things that they believe they already own. People may be· more afraid to give things up (WTA) than to acquire them (WTP) and so require a bigger compensation for WTA. People may be responding discriminately: WTP would reflect what they would sacrifice to get a good. WTA may reflect what they think someone else would pay for the good. In a hypothetical situation such as this, people reason, why not ask for a lot of money to give up something and then see what happens? One of the remaining debates concerning contingent valuation is whether the survey designer can manipulate the responses. It is clear that poorly designed questions yield vague answers, but that even well-defined ques- tions may yield varying responses, depending on how a question is asked. This issue is still being actively debated in the literature and in practice. The proponents of contingent valuation argue that good protocol eliminates such potential forms of bias, but responses depend on the information provided. For example, if one is trying to value water quality, does it matter whether the polluter is poor or rich, a local company or an international, a local employer or outsider? Does it matter who is being polluted, whether those people benefit from the pollution, and whether they can afford to pay for mitigation? These facts may influence the response given to a contingent valuation question but may appear irrelevant to different parties to the debate. If it can be shown that contingent valuation gives approximately the same result no matter who implements the survey, the technique will gain dramatically in acceptance. A final issue surrounding contingent valuation is the interpretation of the results. Almost all open-ended contingent valuation surveys tend to yield a cluster of zero responses, a frequently chosen response, and a tail of very high responses. The zero responses include the people who truly do not value the good and the people who do not want to participate in the exchange. It is important to find a way to distinguish between the true zero responders and the nonparticipants, or the survey may undervalue the good. Equally problematic is the question of what to do with the large tail of high responses. Are these high respondents giving accurate values, do they mis- understand the questions, or are they strategically affecting outcomes? The answer to this pressing problem is yet to be found. Some analysts believe that they have solved this problem by providing people with only a limited 258 'MANAGING THE WORLD'S FORESTS set of choices, but it is not yet clear whether they have artificially con- strained responses or whether they have obtained accurate answers. The critical discussion raised here should not be interpreted as a con- demnation of contingent valuation. Rather, the purpose of this discussion is to clarify that the contingent valuation method is more complicated than originally conceived. Accurate answers are possible only if careful protocols are followed. The technique is not equally reliable for every good. The more familiar the good in question and the more comfortable people feel trading for the good, the more accurate the response is going to be. Thus the contin- gent valuation method is quite good at valuing goods and services traded in markets. The technique can also be applied to nonmarket goods and ser- vices and is applicable to a wider range of goods and services (e.g., nonuse values) than the travel cost method. Although most applications of contingent valuation have occurred in developed countries, several studies have been carried out in developing countries. The previously mentioned travel cost study of a Bangkok park by Grandstaff and Dixon (1986) also used contingent valuation. Estimates of users' willingness to pay were very similar l:o the estimates derived from the travel cost method. Abala (1987) used contingent valuation to measure willingness to pay to enter Nairobi National Park. He also measured the willingness to pay to prevent the park from being developed for other uses. One of the more interesting contingent valuation studies in developing countries is the work by Whittington et al. (1990) on willingness to pay for improved rural water supplies. This study is of particular interest because it addressed the question, Can contingent valuation surveys be used in devel- oping countries to generate useful empirical estimates of willingness to pay? After testing for several types of biases, these researchers concluded that "it is possible to do a contingent valuation survey among a very poor, illiterate population and obtain reasonable, consistent answers" (p. 37). Nonuse Values The valuation of goods and services that people actually use has re- ceived much attention in the economics literature. Many resource econo- mists also observe that citizens are willing to pay for goods and services that have no use. They argue these nonuse values should be added to use values to obtain an accurate aggregate measure of value (see Fisher and Raucher 1984 or Boyle and Bishop 1987). There are two distinct types of nonuse values: option value and existence value. Option Value Option value is related to a potential future use of a resource. Option value is the difference in expected payments between paying for the good as one consumes it and paying for it up front. For example, one could pay for using a swimming pool each time one goes, or one could make a single membership payment up front and swim free each time. Suppose one ex- FOREST VALUATION 259 pected to go to the pool 15 times a year and the admission charge per trip was US$7. The expected payment, the sum of the maximum pay-as-you-go payments that would be offered by the consumer, is US$7 times 15 (US$105). Option value is positive if the up-front payment is bigger than the sum of the pay-as-you-go payments. Unfortunately, for most goods, people want a discount to pay up front. For example, if the pool offered two kinds of membership-one that required the member to pay US$7 for each use and the other. that required the member to pay US$105 up front-a person who expected to go 15 times would generally prefer pay-as-you-go. Hence, op- tion value is negative for most goods. The empirical literature for option values has not generally reflected the theoretical concept. Instead, it has asked people what they would pay for the option to keep the resource available. People have responded with a figure close to expected-use value, not the precise concept labeled option value. These empirical studies provide a useful check of use value as a measure, but they do not generate a unique value that should be added to use value. For all practical purposes, option value should be dropped from valuation exercises. Existence Value Existence value is a willingness to pay for public services that the re- spondent personally has no plans to use. Empirical contingent valuation studies of existence value have asked people to value environments as di- verse as the Grand Canyon (Tolley and Randall 1985) and whooping cranes (Stoll and Johnson 1984). The analysts define existence value as any positive payment made by someone with no immediate plans to visit the resource. They tend to find that existence value is at least as great as expected-use value and sometimes far exceeds it. These authors conclude that many re- sources are worth at least twice the value of their expected use. The problem with this particular argument is that the nonuse value may simply be double-counting use value. People may be willing to pay for resources that are used by others simply because they are used. In a sense, this is a confirmation of cost-benefit analysis. People support public projects whose benefits exceed their costs even if they do not personally benefit. For example, citizens may be willing to pay for a trout stream because they observe that fishermen enjoy it even if they do not plan to use it themselves. The major question about existence value is whether it represents a value that should be added to use value. If existence value merely confirms a desire for rational use of the nation's resources, it should not be added to use value, but if the existence value represents a hidden pleasure that nonusers get from observing users-a pleasure that is specific to a resource- it should be added to use value. Additional research on this concept is needed. An alternative definition of existence value is a willingness to pay for natural environments or components of those environments that have no 260 MANAGING THE WORLD'S FORESTS human use. There may be no human use because people have found no pleasure in interacting with the resource or because the price of obtaining access to the resource exceeds its use value. Nonetheless, it is theoretically possible that people would sacrifice some part of their budget to avoid destruction of the resource. For example, people may be willing to pay for a salamander or an Arctic wilderness to continue to exist even if there is no particular human use involved in preservation. Philosophers who argue in favor of this kind of existence value object to the anthropocentric nature of use values. They argue that natural things should have a right to exist for their own sake. There are concerns that efforts to measure existence value are unreliable because the good being valued is hard to identify and because people obvi- ously have little experience paying for such goods. Nonetheless, existence value may still be a legitimate concept. Cultures that place a high value on a sacred site or a particular species may place existence value on those components of nature. More work needs to be done on existence value. It may be especially fruitful to do this work cross-culturally, as the cultural context will undoubtedly prove important. Valuation for National Income Accounts Although discussion of forest resource valuation so far has focused largely on the microeconomic level, many of the issues raised also have implications for macroeconomic planning and analysis. In particular, proper forest valuation has several implications for aggregate measures of the per- formance of a count.ry's economy. In this section we argue that forest re- sources are often improperly valued in national income accounts, which are widely used for economic analysis and policy recommendations. National income accounts show changes from year to year in such things as economic activity, savings, and gross and net investment. From these accounts are developed national income measures such as gross national product (GNP) and gross domestic product (GDP). From a resource valua- tion perspective, national income accounts and their summary measures are deficient in accounting for the depletion of natural resources and the degra- dation of the environment (El Serafy and Lutz 1989). Both these deficiencies are relevant to the forestry sector. Consider first the failure to account for forest depletion. In countries with large forestry sectors, the use of forest resources can be an important source of economic growth. The forest resources are natural resource assets that can generate income to finance other development activities. Proper accounting at the national level would reflect depletion of the forest asset to create investment in another sector or to finance current consumption. Rob- ert Repetto and his colleagues at the World Resources Institute (WRI 1989) give the following example: Consider the farmer who cuts timber in his woods to build a new barn. His private accounts would show the addition FOREST VALUATION 261 of one asset, the barn, and the loss of another, the standing timber. Presum- ably, the farmer would choose to trade one asset for the other if he thinks he can earn a higher return from the livestock housed in the barn. However, the national accounts would show increases in income from the barn and the timber cutting, as well as an increase in investment from the barn. A more realistic accounting of the change in resource values would count the removed timber as a reduction in forest assets. Several macroeconomic studies have recently been made or are under way to create new national accounts that reflect resource depletion. These are generally referred to as satellite national accounts. The WRI work has focused on estimating the depreciated value of natural resource stocks in terms of marketed products that could have been generated by that stock. Thus when forests are cut, a capital consumption (depletion) allowance is charged against them in accordance with the loss of the forest's ability to generate market products such as industrial roundwood and fuelwood. For example, it has been estimated that Indonesia's high rate of growth of GDP over the 1970-84 period, 7.1 percent annually, was largely the result of high rates of natural resource depletion (Repetto et al. 1989). After correcting for depletion of the country's stocks of petroleum, forests, and soil, GDP growth was estimated to be only 4 percent annually. The second major shortcoming of the national income accounts from a resource perspective is the failure to reflect changes in environmental qual- ity. The clearing of tropical forests entails economic losses associated with reductions in species diversity, watershed protection, carbon sequestering, and the like. An ideal set of national accounts would also reflect these changes in value of ecological services related to physical changes in forests. Of course, it is unlikely that enough information about these values will be generated soon to make it practical to adjust most countries' national in- come accounts for changes in the nonmarket outputs of forests. Over time, however, as more microeconomic valuation studies of the type described earlier are completed, it may become feasible to extend the forest section accounts into this domain. Our current system of national economic ac- counts is a relatively recent creation, and it involves significant assumptions about value and transactions that are difficult to measure. For example, systems of national accounts have long had to deal with the problem of accounting for subsistence production that does not involve market transac- tions (Blades 1975). As Peskin (1990) argues about resource accounting, "Questions of data and skill adequacy are ultimately empirical matters .... They can only be addressed through experimental case studies [p. 19]." CONCLUSIONS AND RECOMMENDATIONS The demand for better valuation of forest products and services has been a direct consequence of greater recognition of the multiple functions of forests and the rapidity with which the world's current forest endowment is 262 MANAGING THE WORLD'S FORESTS being degraded or converted in various parts of the world. Better forest valuation can help planners formulate better policies, choose among projects, and allocate funds for forest management research. As demonstrated ear- lier, there are theoretically sound and practical valuation techniques that are ap- plicable to a wide range of forest products and services, both market and nonmarket. We need more valuation work on forests, and the tools are available to do the job. However, many of the valuation technologies are expensive to imple- ment. Some techniques require the collection of original data on resource inventories and on market prices for a wide diversity of products. Others require data on travel patterns of forest users or contingent valuations that can be obtained only by sample surveys. Choices must be made regarding which forests should be valued and how much data should be collected on each use and each forest. Choices also must be made as to the emphasis to be placed on valuation of particular forests relative to the estimation of the role of forests as a whole in national or world systems of accounts. Clearly, both micro-level (project) valuation and macro-level (national accounts) valuation are impor- tant. But planners must carefully consider the optimal mix of the two when new resources are made available for valuation studies. One approach that reduces the valuation problem and creates economic incentives for better multiple-resource forest management is to develop inno- vative methods for cost recovery in forest projects and policies. These methods can include reforms in timber concessions, park user fees, carbon emission fees to finance carbon sequestration projects, and contributions by the world community to protect biodiversity in developing countries. Better markets need to be created for noncommercial timber species; minor forest products; and; in some cases, forest wildlife. Some products now sold only on local markets might appeal to national or international commercial markets. Al- ternatively, changes in marketing institutions might bring a greater propor- tion of the final price to forest landowners or harvesters. Improving mar- kets, where appropriate, can establish values that encourage forest owners and users to make more efficient use of a wider variety of forest resources. Although more complete and more accurate valuation can aid decision making, both in the development agencies and in national governments, the information will inevitably be incorporated in a complex, politically influenced, policy process. Improved valuation is likely to encourage greater use of quan- titative decision making procedures, such as cost-benefit analysis. But the same information will be used, sometimes selectively, to advance political agendas of a variety of groups with diverse opinions on the proper ap- proach to resource development (Healy and Ascher 1990). New information can change the terms of political debate, raise new issues, and change the arena in which decisions are made. In some cases, better valuation will lead to a reduction in conflict over policies; in other cases it will exacerbate existing disagreements. FOREST VALUATION 263 SUMMARY . The world's forests produce a number of important goods and services. Some of these forest outputs are commercially marketed, such as industrial roundwood, fuelwood, and fodder. Others, such as medicinal plants, nuts, and latex, are traded in informal markets or in little-known formal markets. In addition, forests provide important environmental services such as biodiversity protection, watershed services, and carbon sequestering. Recre- ation and tourism and wildlife products also are important forest products. Many of these forest products and services are public goods for which markets do not exist. Consequently, noncommercial forest outputs are not valued or are misvalued. Furthermore, many of the commercial products of forests are misvalued because of market distortions, such as the existence of externali- ties, market power, and inappropriate government intervention. The simul- taneous production of multiple products and the lack of scientific data on forest products also result in misvaluation of forests. Accurate valuation of forest resources is essential to the appraisal of projects affecting forests and to the development of forest and nonforest policies. Undervaluation of forests and their outputs and services can bias policy decisions related to resource allocation, can result in underestimation of the contribution of forests to economic growth through inaccurate esti- mates of GNP, and can result in untimely liquidation of forest capital. Many forest products and services are important to the livelihood of the rural poor, a group whose welfare is a major focus of public policy. Here again, accurate valuation is an essential ingredient of effective policy formulation. This chapter has evaluated alternative valuation techniques for the nu- merous forest products and services and suggested improvements in theory and practice. The multiple potential outputs from forests pose a challenge for valuation. Traditionally, only a fraction of the forest outputs people enjoy have been quantified. In the past few decades, however, improve- ments in methodology have allowed for a substantial expansion of this list. Which valuation technique is most appropriate to apply depends on the good to be valued. The forest products can be divided into four groups: market goods, potential market goods, public goods, and goods with nonuse values. Mar- ket goods are actively traded and generally well defined. Potential market goods could be traded in markets but are not, because they are distributed by government decree or culturally prescribed rules, or because there is insufficient access to markets. Public goods are jointly consumed products or services that involve many independent consumers. Nonuse goods are goods and services with which individuals do not actually come in contact but which they still value. Nonuse values, referred to as existence values, can pertain to ecosystems, species, or habitats. Traditional economic measures, such as market prices and consumer surplus, can be used to value market goods. For marginal changes in the quantity of market goods, the market price is the most reliable measure of 264 MANAGING THE WORLD'S FORESTS value, whereas for nonmarginal changes in quantities, the consumer sur- plus is the most appropriate measure. Potential market goods are more difficult to value because competitive prices cannot be observed for them. For such goods a number of possible valuation approaches are possible. When these goods are being traded in nearly competitive markets, one can use those observed prices to value the goods. Another valuation technique is to use the price of a close substitute of the product being valued. Some forest outputs are public goods, Four basic approaches have been successfully applied to value public goods: hedonic, travel cost, avoided cost, and contingent valuation. The hedonic model assumes that variations in prices of certain marketed goods can be attributed to variations in the quality of certain attributes of the good. This method therefore values envi- ronmental (nonmarket) goods by evaluating their contribution to the mar- ket price of goods traded in formal markets. The travel cost method, a technique used to value recreation, values a site (a forest destination) by estimating the demand for access to the site using the transportation cost to a site as the main measure of the price of access to the site and calculating how the number of trips taken to the site varies with the cost. After the demand for visits to the site has been calculated, the value of the site is estimated by measuring the area under the demand curve. With the avoided- cost or replacement cost approach, the value of public goods is established by determining the cost of duplicating a service that the resource provides. A final valuation technique that has been widely applied in the past two decades is the contingent valuation method, which uses carefully designed surveys to elicit responses from consumers about the value of certain public goods to them. Average willingness to pay for public goods, and thus the demand for the goods, is estimated. The total value of the good is estab- lished by considering the area under the demand curve. The values of the world's forests, either as complete ecosystems or as composites of specific output flows, have not been measured precisely, and generally not at all. Yet sound resource valuation techniques are available and applicable to a wide range of forest products, both market and nonmarket. The donor community should take the lead in promoting the refinement and application of resource valuation techniques in developing countries. Better forest valuation can help planners formulate effective policies, choose among projects, and allocate funds for forest management research. REFERENCES Abala, Daniel 0. 1987. A Theoretical and Empirical Investigation of the Willingness to Pay for 11 Recreational Services: A Case Study of Nairobi National Park." Eastern Africa Economic Review 3(2):111-19. Alexandratos, Nikos, ed. 1988. World Agriculture: Toward 2000. For the Food and Agriculture Orga- nization of the United Nations. London: Belhaven Press. Bishop, R., and T. Heberlein. 1979. "Measuring Values of Extra-Market Goods: Are Indirect Mea- sures Biased?" American Journal of Agricultural Economics 61:926-30. FOREST VALUA T/ON 265 Blades, Derek W. 1975. Non-Monetary (Subsistence) Activities in the National Accounts of Developing Countries. Paris: Development Center of the Organization for Economic Cooperation and Devel- opment. Boyle, K.J., and R.C. Bishop. 1987. -,,Toward Total Valuation of Great Lakes Fishery Resources." Water Resources Research 5:943-90. Brown, G., and R. Mendelsohn. 1984. "The Hedonic Travel Cost Method." Review of Economy and Statistics 66:427-33. Brown, Gardner Mallard, Jr., and Henry 0. Pollakowsky. 1977. "Economic Valuation of a Shore- line." Review of Economy and Statistics 59:272-78. Burt, 0., and D. Brewer. 1971. "Estimation· of Net Social Benefits from Outdoor Recreation." Econometrica 39:813-28. Buschbacher, Robert. 1987. "Ecological Analysis of Nahrral Forest Management in the Humid Trop- ics." Paper presented to symposium on "The Application of Ecology to Enhancing Economic Development in the Humid Tropics," American Institute of Biological Sciences, August. Caulkins, Peter P., Richard C. Bishops, and Nicolaas W. Bowes, Sr. 1986. "The Travel Cost Model for Lake Recreation: A Comparison of Two Methods of Incorporating Site Quality and Substitution Effects." American Journal of Agricultural Economics 68:291-97. Cichetti, C., A. Fisher, and V.K. Smith. 1976. "An Econometric Evaluation of a Generalized Con- sumer Surplus Measure: The Mineral King Controversy." Econometrica 44:1259-76. Clawson, M., and J. Knetsch. 1966. Economics of Outdoor Recreation. Baltimore: Johns Hopkins Press. Cummings, R., D. Brookshire, and W. Schulze. 1986. Valuing Environmental Goods. Totowa, N.J.: Rowman and Allenheld. Debeer, Jenne H., and Melanie J. McDermott. 1989. The Economic Value of Nontimber Forest Products in Southeast Asia. Amsterdam: Netherlands Committee for IUCN. Dixon, John A., Richard A. Carpenter, Louise A. Fallon, Paul B. Sherman, and Supachit Manipomoke. 1988. Economic Analysis of the Environmental Impacts of Development Projects. London: Earthscan Publications. Durojaiye, Bamidele 0., and Anthony E. Ikpi. 1988. "The Monetary Value of Recreational Facilities in a Developing Country: A Case Study of Three Centers in Nigeria." Natural Resources Journal 28(2)0315-28. El Serafy, Salah, and Ernest Lutz. 1989. "Environmental and Nahrral Resource Accounting." In Environmental Management and Economic Development. eds. Gunter Schramm and Jeremy J. Warford. Baltimore: Johns Hopkins University Press/World Bank. Englin, J., and R. Mendelsohn. 1991. "An Hedonic Travel Cost Analysis for Valuation of Multiple Components of Site Quality: The Recreation Value of Forest Management." Journal of Environ- mental Economics and Management. 21(3):275-89. Ewing, Andrew J., and Raymond Chalk. 1988. The Forest Industries Sector: An Operational Strategy for Developing Countries. Technical Paper No. 83. Washington, D.C.: World Bank. Falconer, Julia. 1989. "The Major Significance of 'Minor' Forest Products: Local People's Uses and Values of Forests in the West African Humid Forest Zone." Unpublished manuscript. Rome: FAO Forestry Department. Fischoff, B., P. Slavic, and S. Lichtenstein. 1980. "Knowing What You Want: Measuring Labile Values." In Cognitive Processes in Choice and Decision Behavior, ed. T. Wallsten. Hillsdale, N.J.: Erlbaum. Fisher, A., and R. Raucher. 1984. "Intrinsic Benefits of Improved Water Quality: Conceptual and Empirical Perspectives." In Advances in Applied Microeconomics, eds. V.K. Smith, and A.D. Witte. Greenwich, Conn.: JAi Press. Food and Agriculture Organization of the United Nations (PAO). 1986a. "Some Medicinal Forest Plants of Africa." Rome: Forest Resources Development Branch, Forestry Department. _ _ _ . 1986b. 1985 Yearbook of Forest Products. Rome: FAO. _ _ _. 1988a. 1988 Yearbook of Production. Rome: FAO. _ _ _ . 1988b. 1988 Yearbook of Trade. Rome: FAO. _ _ _. 1989. Management of Tropical Moist Forests in Africa. FAO Forestry Paper No. 88. Rome: FAO. 266 MANAGING THE WORLD'S FORESTS Freeman, A. Myrick. 1979. The Benefits of Environmental Improvement: Theory and Practice. Baltimore: Johns Hopkins University Press/Resources for the Future. Grandstaff, Somluckrat, and John A. Dixon. 1986. "Evaluation of Lumpinee Park in Bangkok, Thai- land." In Economic Valuation Techniques for the Environment: A Case Study Workbook, eds. John A. Dixon and M.M. Hufschmidt. Baltimore: Johns Hopkins University Press. Gregerson, Hans M. 1985. "Valuing Goods and Services from Tropical Forests and Woodlands." Report prepared for the U.S. Congress, Office of Technology Assessment, Washington, D.C., June. Hamilton, Lawrence 5. 1985. "Overcoming Myths About Soil and Water Impacts of Tropical Forest Land Uses." In Soil Erosion and Conservation, eds. S.A. ~El-Swaify, W.C. Moldenhauer, and An- drew Lo. Ankeny, Iowa: Soil Conservation Society of America. Hartshorn, Gary. 1989. "Sustained Yield Management of Natural Forests: The Palcazu Production Forest." In Fragile Lands of Latin America, ed. John 0. Browder. Boulder, Colo.: Westview Press. Healy, Robert G., and William Ascher. 1990. "Incorporating Environmental Information in Natural Resources Policymaking: A Policy Process Approach." Paper prepared for annual meeting of Association for Public Policy Analysis and Management, San Francisco, October. Howe, Charles W. 1979. Natural Resource Economics. New York: John Wiley and Sons. Hyde, William F. 1983. Economic Evaluation of Investment in Forestry Research. Durham, N.C.: Acorn Press. Knetsch, J. 1990. "Environmental Policy Implications of Disparities between Willingness to Pay and Compensation Demanded Measures of Value." Journal of Environmental Economics and Manage- ment 18:227-37. Krutilla, John V., and Anthony C. Fisher. 1975. The Economics of Natural Environments: Studies in the Valuation of Commodity and Amenity Resources. Baltimore: Johns Hopkins University Press. Laarman, J.G., and P.B. Durst. 1987. "Nature Travel and Tropical Forest." Journal of Forestry 85 (no. 5):43-46. Lean, J., and D.A. Warrilow. 1989. "Simulation of the Regional Climatic Impact of Amazon Defor- estation." Nature 342 (November 23):411-13. Mendelsohn, R. 1984. "An Application of the Hedonic Travel Cost Framework for Recreation Mod- eling to the Valuation of Deer." In Advances in Applied Econometrics, vol. 3, ed. V.K. Smith, 89- 101. Greenwich, Conn.: JAI Press. Mitchell, R., and R. Carson. 1989. Using Surveys to Value Public Goods: The Contingent Valuation Method. Washington, D.C.: Resources for the Future. Morey, E. 1981. "The Demand for Site-Specific Recreational Activities: A Characteristics Approach." Journal of Environmental Economics and Management 8:345-71. _ _ _ . 1984. "The Choice of Ski Areas: Estimation of a Generalized CES Preference Ordering with Characteristics." Review of Economics and Statistics 66:.584-90. Myers, Norman. 1984. The Primary Source: Tropical Forests and Our Future. New York: W.W. Norton. Nelson, Robert. 1982. "The Public Lands." In Current Issues in Natural Resource Policy, ed. Paul Portney. Baltimore: Resources for the Future/Johns Hopkins University Press. Norgaard, Richard B. 1989. "Three Dilemmas of Environmental Accounting." Ecological Economics 1:303"14. Organization for Tropical Studies. 1989. "Native Tree Species for Reforestation and Land Rehabili- tation in the Lowland Humid Tropics: Phase 2." Durham, N.C. Peskin, Henry M. 1990. "Alternative Environmental and Resource Accounting Approaches." Paper for International Conference for Ecological Economics of Sustainability. Washington, D.C., May. Peters, Charles M., Alwyn H. Gentry, and Robert 0. Mendelsohn. 1989. "Valuation of an Amazo- nian Rainforest." Nature 339 (June 29):655-56. Repetto, Robert, and Malcolm Gillis. 1988. Public Policies and the Misuse of Forest Resources. Cam- bridge, Eng.: Cambridge University Press. Repetto, R., W. Magrath, M. Wells, C. Beer, and F. Rossini. 1989. "Wasting Assets: Natural Re- sources in the National Income Accounts." Washington, D.C.: World Resources Institute. Salati, E. 1987. "The Forest and the Hydrological Cycle." In The Geophysiology of Amazonia, ed. R.E. Dickinson. New York: John Wiley. FOREST VALUATION 267 Sedjo, Roger A. 1983. Comparative Economics of Plantation Forestry. Washington, D.C.: Resources for the Future. _ _ _. 1989. "Forests to Offset the Greenhouse Effect." journal of Forestry 87(7):12-15. Smith, V.K., and William Desvousages. 1986. Measuring Water Quality Benefits. Boston, Conn.: Kluwer- Nijhoff Publishing. Smith, V.K., and Y. Kaoru. 1986. "The Hedonic Travel Cost Model: A View From the Trenches." Land Economics 63:179-192. Sorg, C., and J. Loomis. 1984. Empirical Estimates of Amenity Forest Values: A Comparative Review. Fort Collins: Rocky Mountain Forest and Range Experimental Station, USDA Forest Service. Squire, Lyn, and Herman G. van der Tak. 1975. Economic Analysis of Projects. Baltimore: Johns Hopkins Press. Stoll, J.R., and L.A. Johnson. 1984. "Concepts of Value, Nonmarket Valuation, and the Case of the Whooping Crane." Transactions of the Forty-ninth North American Wildlife and Natural Resource Conference 49:359-75. Tolley G.R., and A. Randall. 1985. "Establishing and Valuing the Effects of Improved Visibility in the Eastern United States." Final Report to the U.S. Environmental Protection Agency for con- tract no. 807768-01-0, Washington, D.C. U.S. Forest Service. 1990. The Forest Service Program for Forest and Rangeland Resources: A Long-term Strategic Plan. Washington, D.C.: U.S. Forest Service. Vaughn, W., and C. Russell. 1983. The National Benefits of Water Pollution Control: Fresh Water Recre- ational Fishing. Baltimore: Johns Hopkins University Press. Whittington, Dale, John Briscoe, Xinming Mu, and William Barron. 1990. "Estimating the Willing- ness to Pay for Water Services in Developing Countries: A Case Study of the Use of Contingent Valuation Surveys in Southern Haiti." Economic Development and Cultural Change 38(2):293-311. Wilson, E.O. 1988. "The Current State of Biological Diversity." In Biodiversity, ed. E.0. Wilson. Washington, D.C.: National Academy Press. Woodwell, G.M. 1984. "The Carbon Dioxide Problem." In The Role of Terrestrial Vegetation in the Global Carbon Cycle: Measurement by Remote Sensing, ed. G.M. Woodwell. New York: John Wiley & Sons. World Resources Institute (WRI). 1988. World Resources, 1988-89. New York: Basic Books. _ _ _. 1989. World Resources, 1989-90. New York: Basic Books. 11 Forestry Institutions Marie Lynn Miranda, Olga Marta Corrales, Michael Regan, and William Ascher T he greatest new policy challenge that has emerged for forestry institu- tions in 'the past two decades is the need to pursue an increasing mul- tiplicity of objectives. This task is complicated by the fact that many of the new objectives are difficult to describe or quantify with the traditional tools of policy analysis. Although the rhetoric at the ministerial level of govern- ment clearly emphasizes all the objectives, in reality conventional forestry institution managers face inordinate pressure to pursue some objectives to the neglect of others. Often, short-term objectives are pursued beyond the point of an efficient trade-off-to the detriment of the long-term health of the forest ecosystem. This chapter addresses the question of what innova- tions in institutional design might provide incentives for government and state-enterprise officials to use and to regulate the forest resource sustainably. The chapter examines three countries where new approaches to forest management have been established (Colombia, Honduras, and Malaysia). These approaches involve financing arrangements that provide partial or full administrative funding for the forestry institution directly from forestry operations. In general, the new institutional designs have performed rea- sonably well in the areas of revenue generation and rent capture, but they have given insufficient attention to the diverse environmental and social services provided by the forest. By its very nature, forestry, especially tropical forestry, is a long-term undertaking, requiring management plans for durations that far exceed the tenure of government policymakers. Because forests are a relatively easy resource to extract, they are an especially tempting target for generating cash flow and foreign exchange. In the dynamic between short-term gains 269 270 MANAGING THE WORLD'S FORESTS and long-term viability, the stability of an ecosystem often loses out to par- tisan concerns and myopic decision making. The long time horizon associated with forests also exacerbates pressure for agricultural conversion. From the government planners' perspective, establishing permanent, settled agriculture helps to meet immediate human needs and to improve food self-sufficiency. Labor-intensive agriculture also better responds to the problems of underemployment and unemployment than does the capital-intensive forestry sector. From the local population's perspective, agricultural conversion, whether through shifting cultivation or permanent agriculture, provides a consistent annual flow of income, which pulp or timber production does not. International agencies and developing country governments now find themselves concerned not only with revenue generation and social forestry goals, but also with issues of biological diversity, environmental benefits, recreation opportunities (both for the local population and for the tourist industry), and mitigation of global and regional climatic changes. As for- estry institutions struggle to deal with this broader agenda of forest man- agement issues, the methodology for evaluating their performance must adjust as well. In this evaluation of forestry institutions, we develop an analytical framework that attempts to link institutional structure to perfor- mance in revenue generation, social forestry, and environmental services. DEFINITION OF THE PROBLEM Forestry institutions are charged with a particularly difficult resource to manage. Sound management practices are especially important in the devel- oping world, where they are essential to the environmental and economic sustainability of many development projects. But high (and unsustainable) extraction rates have been condoned at some time throughout most of the developed world as well. Why then should the developing world overly concern itself with rates of deforestation? Duncan Poore (1989), in a study prepared for the International Tropical Timber Organization (ITTO), suggests several reasons: 1. The process of clearing the temperate forest to make way for agriculture came at the cost of substantial soil erosion. If modern technology and the current deeper understanding of soil and ag- ronomic science had existed, development would probably have taken a very different path. 2. Much of the remaining tropical forest is in areas that are diffi- cult to convert sustainably to agriculture because of soil infertility and incidence of disease. It can be argued that, given those physi- cal constraints on the land, it would be best to leave the forest alone, or at least large portions of it. FORESTRY INSTITUTIONS 271 3. The forests of the wet tropics are especially sensitive to man- agement techniques. Harvesting plans that work well in the tem- perate forests of the world can drive (and have driven) many tropical flora and fauna to extinction. Many of the tropical developing countries have had limited economic activities that could absorb capital from forest exploitation (or take advan- tage of cleared land) so as to yield higher returns. Unlike the situation in the United States, where the heavy logging of the 19th and early 20th centuries provided the materials and capital for the dramatic growth of other sectors, many developing countries have had no such rationale for the transforma- tion of their forest endowment. Forestry departments in developing countries also face a technical chal- lenge. They cannot rely on management techniques used historically in the temperate world. The limited amount of tropical research that has been undertaken means that forest managers often lack the basic information needed to develop good management plans. Even where extensive scientific research has been done, the question of how best to implement policy in accordance with that scientific information remains unanswered. Forestry departments must develop creative and innovative approaches to managing the forest resource sustainably under conditions of extraordinary economic and demographic pressure. At least six problems plague forestry departments in both the devel- oped and the developing worlds, problems that may be exacerbated by the conventional institutional structure of the departments: (I) placement of the forestry institution in the bureaucracy, (2) difficulty in addressing broad definitions of forest management, (3) difficulty in respecting poorly defined property rights, (4) disproportionate emphasis on downstream processing, (5) lack of incentives to extract rents, and (6) view of the forest as a residual land use. Each is discussed in the paragraphs that follow. Placement of the forestry institution in the bureaucracy. The forest resource is affected by policy decisions on a vast array of public-sector issues, including population control efforts, migration policies, agriculture policies, land distribution policies, effective protection in international trade, industrialization efforts, and the enfranchisement of ethnic minorities. An agency with the jurisdiction to make decisions in any of these areas has some effect on the forest resource. The extent depends on the placement, power, and budgetary support of the forestry institution relative to these other agencies or ministries. The lead forestry agency is often placed within a ministry-most com- monly, the agriculture ministry-that has many functions beyond the man- agement of the forest resource. In these situations, forestry is often regarded as less economically and politically important than the other sectors over- seen by the ministry. Ministers often put agriculture far ahead of forest conservation. In fact, the financial connection between timber proceeds and 272 MANAGING THE WORLD'S FORESTS agriculture is sometimes quite direct. Adeyoju (1976, 21) notes that in many African countries, when the forestry departments were under the ministries of agriculture, agricultural extension was financed by timber harvesting and concession proceeds. The budget of the forestry department is formulated as part of the ministry's overall budget, subject to approval by the standard executive bodies and the legislature. This arrangement gives the head of the oversee- ing ministry considerable latitude in deciding how to distribute discretion- ary funds across departments. Top forestry officials everywhere complain about the inadequacy of financial and human resources to manage the day- to-day activities of departments. In Llaurado and Speidel's (1981, 143) sur- vey of 15 Latin American forestry departments, the 28 officials surveyed cited the lack of financial resources as the most serious problem confronting forestry departments, with insufficient staff training a close second. Adeyoju' s African sample of Cameroon, Ghana, Kenya, Liberia, Nigeria, and Tanzania in the mid-1970s revealed that "forest administrations have so far rarely succeeded in getting budgetary funds commensurate with their wide range of responsibilities" (Adeyoju 1976, 32). An organizational structure that puts the principal forestry institution at the mercy of other government agencies leaves forests particularly vulnerable. Attempts to meet the pressure for short-term revenue generation will consistently work to the disadvantage of long-term sustainable management of the resource. The budgetary allocation mechanism also contributes to staffing prob- lems in developing country forestry institutions. First, there is the issue of whether the department can afford to hire enough staff people to handle the far-ranging responsibilities typically covered in its mandate. Second, even if financial resources are plentiful, many developing countries have a limited cadre of forestry professionals with expertise in both forest production and environmental services. Limited staffing combined with a general prefer- ence for living in the capital cities means that many forestry department personnel are physically separated (sometimes by great distances) from the resource they are charged with managing. Difficulty in addressing broad definitions of forest management. Over the past several decades, forestry has come to mean far more than the management of trees. The introduction of social forestry and environmental services to the debate on forest management has greatly complicated the issue, Although most forestry departments have formally articulated their goals and objectives, they have no specific policies for achieving these competing goals. As the responsibilities of forestry departments have expanded, inadequate funding and staffing have made it difficult for the departments to meet the new challenges. This difficulty has presented itself in, among other countries, Malaysia, Nepal, Indonesia, and much of South America. In some cases, the proliferation of forest-related objectives has given rise to the establishment of new institutions. Adeyoju (1976, 19) notes that FORESTRY INSTITUTIONS 273 the functions of protecting or establishing wildlife preserves, game parks, national parks, zoos, and gardens were spun off to separate departments in Ghana, Tanzania, and Kenya. In Latin America, some governments have created semiautonomous agencies or public enterprises to assume some of the responsibilities of forest policy (Llaurado and Speidel 1981, 149). In most countries, however, even where separate conservation institutions have been established, the primary forestry institution has assumed more func- tions and objectives. In addition, several forestry departments have taken on a wider set of development responsibilities, such as the provision of social infrastructure. Not only are forestry institutions facing an increasing number of objec- tives, but many of the new objectives do not lend themselves to the estab- lishment of clear performance criteria. Most nontimber benefits of the forest produce no revenue but are essential to the long-term health of the forest ecosystem. Although methods for valuation of these environmental benefits have been developed, they are expensive to implement. In addition, forestry institutions must consider from whose perspective the environmental ben- efits should be measured. To what extent can local populations, many strug- gling to meet human needs, be expected to concern themselves with the issues of biological diversity, recreational opportunities, and aesthetics? And if the principal concern about these issues comes from the developed world, what sorts of mechanisms can be developed for dealing with these external valuations? This new category of objectives has proved especially problematic for forestry institutions. Most countries give priority to the goals that have clearer policy implications. In addition, forestry professionals who have been trained in traditional forestry schools may lack the professional expertise to deal with the environmental side of the forest resource. Difficulty in respecting poorly defined property rights. Ownership rights over much of the forest resource in developing countries are often poorly understood de facto even if they are clearly defined de jure. The reasons include the inability of the forestry department to enforce use restrictions adequately and legacies of colonial rule under which persons who have "improved" (i.e., cleared) an area of land may formally lay claim to it. When disputes develop over who has a right to the financial and nonfinancial benefits generated by the forest, forestry departments have difficulty formulating policies that local populations will respect. This is especially true if local populations realize that revenue generated by the local forest goes to the general treasury and may not return in the form of local development projects. This difficulty has presented itself in, among other countries, Malaysia, Nepal, Indonesia, and much of South America. Forestry institutions must also deal with the gradations of common property resources that lie between state and private property, balancing resource protection and use, especially for very low income people. The forestry institution must determine which areas of the forest (or potential 274 MANAGING THE WORLD'S FORESTS forest) would best be returned to indigenous groups and which should remain under the control of federal or regional government authorities. The decisions will depend on the level of rural poverty; the degree of pressure from shifting cultivators; the relative productivity of the underlying soils; the value of the forest in terms of logging, recreation, and biodiversity; the historical relationship between the indigenous population and the forest resource; and the availability of adequately trained forestry personnel. Disproportionate emphasis on downstream processing. Many tropical developing countries have restricted the direct exportation of raw materials from the forest in order to foster the development of a local wood-processing sector. This decision, often made outside the forestry department, fundamentally changes how the forest is viewed by various powerful groups within the country. Many downstream processing incentive schemes favor the primary processing sector more than broader downstream. processing possibilities. Essentially the forest becomes a feedstock to the primary processing sector, and the pressure to harvest beyond reasonably sustainable rates is intense. The unbalanced system of incentives has limited the ability of forestry departments to manage the forest resource on a long-term basis. Although the stated national forest policy continues to pay lip service to the environmental services provided by the forest, forestry departments are forced to emphasize "getting the cut out." For example, Thailand, once a major producer of tropical hardwood timber, now must import timber to feed the processing mills. Lack of incentives to extract rents. Forestry departments in many tropical countries generate considerable revenue for the state and local governments they serve, but budgetary allocations are not linked to the amount of revenue generated. The budgetary allocation for the forestry department is determined by the priority that the government leaders, who cater to a host of constituencies, place on the management of the forest resource. The budgetary decision-making process gives the forestry department little incentive to maximize the amount of natural resource rent extracted from private forestry concessions. In addition, forestry department officials may be under pressure to encourage a forest-products industry or to extract personal gains from the logging companies, and thus are even less likely to make rent extraction a priority. View of the forest as a residual land use. Although the important role of the forest in both economic and ecological sustainability has become better understood over the past two decades, agriculture is almost a priori assigned a more important role in rural development than is forestry. Population growth exacerbates the pressure to convert more and more land to agriculture. The forest essentially becomes a residual land use, and conventional forestry departments founder in their attempts to develop broader, more long-term forest management policies. This problem has been especially troublesome FORESTRY INSTITUTIONS 275 in the Latin American nations of Costa Rica, Honduras, Argentina, and Brazil. Conventional forestry departments often lack a mechanism for devel- oping appropriate staff and resource management incentives. Some unconventional forestry departments are introducing design innovations in an attempt to restructure these incentives. Case studies of three countries that have such unconventional approaches-Colombia, Honduras, and Malaysia-follow. THE COLOMBIAN CASE STUDY In 1968 Colombia created the National Institute for Renewable Natural Resources (INDERENA) as a semiautonomous corporation to manage the country's forests and other natural resources. IND ERENA is administra- tively independent from the central government but financially dependent on the annual budgetary process. After INDERENA's forest policies are reviewed by the Ministry of Agriculture, the National Planning Department (DNP), and the National Counsel for Economic and Social Policy, they are reviewed and approved by the president of the republic (Llaurado and . 1981, 15). 1 Speidel ' At the regional level, INDERENA' s field offices share responsibility for forest management with 12 autonomous regional corporations that operate under the supervision of the DNP. Among the functions of these corpora- tions are local resource development, land and water control, integrated agricultural services, power development, and conservation (Thomas 1985, 126). Many of these functions overlap those of INDERENA, and the system has suffered from confused lines of authority. In 1987 the Colombian government joined an ambitious international effort to conserve tropical forests while increasing their productivity. The product of this effort, the Tropical Forestry Action Plan for Colombia (TFAPC), was noteworthy for the participation of INDERENA and both domestic and international groups and interests. The TFAPC provides a thorough diagno- 1 Llaurado and Speidel (1981) also report that autonomous or semiautonomous forestry institutions can be found in Bolivia, Colombia, Guatemala, Brazil, Argen- tina, Honduras, and Chile. The authors note that institutional reform in the region can be described as an attempt to meet certain fundamental needs of public forestry administrations: developing and implementing a unified forest policy; assuming a certain degree of governing authority; taking a minimum amount of initiative and exhibiting the flexibility ordinarily ascribed to private enterprise; being capable of operating at a high level of executive deconcentration; and taking part in integrated or concerted activities, mainly at the national and regional planning levels, in land use, production, and the processing of raw materials derived from forests. 276 MANAGING THE WORLD'S FORESTS sis of the forestry sector and presents a new management strategy that stresses forestry land-use planning, social forestry, forest industry develop- ment, and conservation of forest ecosystems, as well as the strengthening of forest institutions. The TFAPC appears to reflect new prominence for forestry management at the national level, but in fact the institutional reform gives the DNP more responsibility for coordinating forest policy. Hence INDERENA's future role is in question. Although it is far too early to evaluate the performance of this new strategy or to know with certainty its implications for INDERENA' s future, it seems clear that the plan's effective- ness will depend on a high degree of institutional cooperation and policy coordination. INDERENA's Limitations Many of the difficulties confronting INDERENA stem from inadequate budgetary allocations that hamper activities and reflect the low priority given to the forestry sector in the national economy. Throughout most of the 1980s INDERENA and other institutions in the agricultural sector saw their proportion of the government budget decrease (Thomas 1985, 142). In addition, the autonomous regional corporations, which duplicate many of INDERENA's functions, have direct relationships with the central gov- ernment and make their own budget requests without consulting other for- estry institutions. This arrangement undermines the goal of developing a unified forest policy. The TFAPC has noted that forest institutions in Colombia showed "a low capacity to implement projects due to internal problems (e.g., the exces- sive assignment of functions, lack of equipment, and lack of technical exper- tise)" (DNP 1989, 32). Despite the availability of well-trained foresters in Colombia, INDERENA has been unable to attract the personnel required to fulfill various aspects of its broad mandate. INDERENA's Performance The TFAPC identified four critical issues for the forest sector: l. Overexploitation of some resources (deforestation) and underutilization of others have created an imbalance in the use of forest resources. 2. The destruction of important ecosystems has caused the disap- pearance of many species of fauna and flora. 3. The forest industry has not contributed to the national economy or to an increased standard of living in local communities in ac- cordance with its potential. 4. The state has a limited capacity to intervene in the forestry sector. FORESTRY INSTITUTIONS 277 Forest Management During the 1980s, forests with commercial potential covered 34 percent of the national area but contributed only 1.7 percent to the gross national product (DNP 1989, 29). Forest industry, however, has accounted for 11 percent of deforestation in the country. Because the more accessible natural forests have been largely depleted, the nearly insignificant levels of commercial reforestation are startling. For example, reforestation increased from less than 5,000 hectares in 1976 to 32,000 hectares in 1981. Since then, however, it has dropped steadily to 3,700 hectares in 1986 (DNP 1989, 25). Many causes have been mentioned, including insufficient economic incen- tives, poor market penetration, and deficiencies in the development of state policies. The TFAPC notes both the inadequacy of available research and INDERENA's limited planning and management capacity. The concession contract signed with Carton de Colombia illustrates the problems facing INDERENA and others in achieving forestry objectives. Carton de Colom- bia is a private corporation that, since 1944, has produced paper <>nd card- board packaging. On February 27, 1974, the central government and Carton de Colombia signed a 30-year forest concession contract, including a com- mitment to leave the forest with a wood volume equal to that existing at the beginning of the contract. The concession area covers 61,000 hectares in the Bajo Calima district, near Buenaventura on the Colombian Pacific coast. Carton de Colombia has harvested approximately 100,000 cubic meters of timber per year. According to company data, after 15 years the average regeneration rate of the forest has been 7 to 8 cubic meters per hectare, slightly higher than the rate that occurs naturally. While studies that iden- tify the flora and fauna in these regenerated forests have been conducted, extrapolating the studies beyond 15 years of regeneration efforts can only be done subjectively and with some risk. The concession contract is <>lso complicated by the colonization that has come via a new highway from Buenaventura to the new naval base at Malaga Bay on the Pacific coast. The colonists-both landless peasants and specula- tors-have come to clear land adjacent to the highway, to exploit the forest, and to take advantage of the eventual appreciation of these lands. The group of colonists is not large yet, but Carton de Colombia and others have been critical of IND ERENA' s inability to control the colonization or to confiscate illegally felled timber. Forest Conservation Deforestation is considered the principal environmental problem in Colombia, having reached rates of 600,000 hectares per year and accounted for 37.7 million hectares between 1960 and 1984. This loss represents ap- proximately 41.5 percent of all forested areas during the 25-year period (DNP 1989, 15). The principal cause of deforestation is colonization, which 278 MANAGING THE WORLD'S FORESTS includes both shifting cultivation and settled agriculture. The lack of techni- cal resources, the lack of enforcement capabilities, and various government policies contribute to colonization.' The destruction of the natural forests has caused significant erosion in approximaltely 20 percent of the country, disturbing critical watersheds and producing a range of severe environmen- tal consequences. In addition, alternative uses for the land, such as agricul- ture and cattle ranching, are sometimes profitable only because of substan- tial government subsidies. Limited staff and political power have left INDERENA unable to con- trol the colonization process, either through directly enforcing existing laws or through influencing changes in national policies. A new land-titling law was enacted in 1988, restricting land titles to occupants of the land whose activities are consistent with the rational use of the forest, but nobody knows how well the law will be implemented. Revenue Generation Throughout the 1980s Colombia's forests contributed only 1.7 percent to the gross national product (GNP) (DNP 1989, 19). The volume of waste generated in wood processing amounts to 30 percent in extraction and 42 percent in transformation (DNP 1989, 15). With improvements in medium- term extraction practices, such as more efficient processing technologies, it would be possible to increase revenue by 22 to 36 percent. Social Forestry Although INDERENA has not generally been committed to community participation and environmental education, several regional projects dem- onstrate the ability of the regional corporations to work with community organizations to improve forest management. For example, the Asociacion Campesina Integral Del Atrato (ACIA) is an association of indigenous com- munities in the Atrato area (Choco Department) that, for the past five years, has received increased responsibility for managing natural resources in the region. The black communities of this area engage in agriculture, fishing, mining, hunting, and the exploitation of timber. The Atrato area has high rainfall and the ecosystem is fragile, even though the productivity of the soils is quite high. The communities thus must engage in careful silvicul- tural management to avoid washing out the soils. The Atrato area is part of a forestry reserve zone of 800,000 hectares, for which the government has established several important policies. The gov- 2 The DNP (1989) reports the relative importance of various causes of deforesta- tion as colonization (76.3 percent), the consumption of fuelwood (12.7 percent), and the lack of alternative supplies for industry and of efficient technology for wood processing (11 percent). FORESTRY INSTITUTIONS 279 ernment delegated responsibility for the reserve from INDERENA to an autonomous regional corporation, Corporacion Autonoma Regional de! Choco (CODECHOCO). CODECHOCO, in turn, wants to lift the forest reserve limitation on the zone so that peasants can get titles to individual parcels of land and gain access to credit. However, ACIA opposes lifting the reserve status, arguing that it would invite colonization and land speculation. Joint agreements between ACIA and CODECHOCO include territorial classification activities, natural resource management, and a community participation project for sustained· exploitation and preservation of natural resources for the central valley of the Atrato River. CODECHOCO is com- mitted to developing a special regulation for natural resource management for this zone, taking into account the proposals made by ACIA. Innovation and Adaptability Carton de Colombia has developed new technologies for processing short-fiber woods and for reducing the damage to soils from extraction, and INDERENA carried out a national forestry research program (1972-82). INDERENA also has a 50 percent interest in CONIF, a quasi-government organization that promotes forestry research, reforestation, and socio- economic development in concession areas. In addition, the Colombian Corporation for the Amazon (ARARACUARA), an autonomous regional corporation, has launched vari- ous projects to seek alternative means of production that are compatible with the region, its environment, and its people. These projects include agricultural research, pest control research, soil conservation research in the Guaviare region, and an agroforestry project. The TFAPC (1989) noted that the forestry sector's institutional problems stem primarily from overlapping areas of responsibility among INDERENA and the autonomous regional corporations. In sum, INDERENA has insuf- ficient authority to coordinate policy in the forest sector or to influence broader government policy. Conclusions Although reforms are incomplete, several lessons can be drawn from public forestry administration in Colombia: • Uncertain financing-in this case, annual budgetary authorizations- compromises the autonomy of the institution. INDERENA is unable to plan and implement its agenda without interference from those entities controlling the flow of resources. · • The shortage of resources also impairs INDERENA's forestry man- agement; for example, INDERENA lacks the resources to enforce existing laws that prohibit invasions of unoccupied lands. 280 MANAGING THE WORLD'S FORESTS • Decentralization may improve the ability of regional institutions to undertake projects that meet the broader needs of local communities, For example, autonomous regional corporations in Colombia have undertaken several innovative conservation and protection programs to address a wide range of social and economic development objec- tives, Decentralization may also lead, however, to confused lines of authority. The end result can be ineffectual planning and implemen- tation of forest policies, INDERENA's weak performance appears to be a function of resource constraints and political status, as well as internal deficiencies. More effec- tive leadership by its leaders might improve institutional planning, develop domestic political support, and increase support from international donors. THE HONDURAN CASE STUDY The Honduran Forest Development Corporation (COHDEFOR) is a semiautonomous state enterprise created in 1974 to expand the wood-prod- ucts industry and to increase Honduras's retention of the profits from for- estry exploitation. Its original objective was to make Honduras the major exporter of sawnwood in the Central American isthmus. In creating COHDEFOR the Honduran government tried to capture what it hoped would be a solid source of foreign exchange. COHDEFOR was established over the strong and vocal opposition from the private log- ging and processing sector. From its inception COHDEFOR had jurisdiction over all forests and all land classified as having forest cover as its first and best use. Roughly 8 million hectares (representing 75 percent of Honduras's total area) are classified as most appropriate for forestry, but current estimates indicate that only 5 million hectares are actually under forest cover-evenly divided between broad-leaf and pine forest (ITTO 1988c, 97-99). Initially COHDEFOR enjoyed a monopoly on the export of forest prod- ucts (round log exports were banned). COHDEFOR was permitted to retain the profits from exports but was required to finance its entire operation without allocations from the central government. In the mid-1980s the ex- port monopoly was withdrawn, but COHDEFOR was allowed to retain the proceeds from an export levy that it imposes on all timber exports. COHDEFOR' s political position has been bolstered by both its institu- tional structure and its financial autonomy. The Honduran president pre- sides over COHDEFOR, and the ministers of finance, economics, defense, natural resources, and planning make up its board of directors. Unfortu- nately, agricultural policy and refugee policy-both of which have tremen- dous influence on the forest resource-are beyond the scope of COHDEFOR. The concentration of economic resources in an enterprise that is not subject to the standard government oversight and control makes COHDEFOR a tempting target for corruption. FORESTRY INSTITUTIONS 281 Although COHDEFOR has had some success in capturing and retaitting logging rents, its performance with respect to deforestation has been at best mixed. COHDEFOR has narrowed the gap between forest depletion and restoration through afforestation or regeneration, but the rate of deforesta- tion remains high (Projecto de Desarollo 1989, 2). For much of its existence COHDEFOR has been oriented toward commercial and industrial develop- ment. This focus has limited attention to the regulation of extraction, refor- estation, and rural sector participation and to the integrated management of the forest's productive and protective services. COHDEFOR has demon- strated a limited vision of the range and variety of the social and environ- mental aspects of the forest. The financial stability of COHDEFOR has been jeopardized by its direct participation in forest industry. COHDEFOR's US$20 million investment in purchasing existing sawmills, both through joint ventures and sole COHDEFOR ownership,3 is widely acknowledged as disastrous. Adminis- trative errors, operating difficulties, and corruption plagued the undertak- ing from the beginning. From its creation until 1985, COHDEFOR exported primary forest prod- ucts, retaining 16 to 18 percent of the export price. In 1986 this activity was transferred to the private sector. Subsequently, exports of primary forest products declined, because of a reduction in the number of mature trees, a reduction in the cut permitted by COHDEFOR, and an increase in second- ary forest industries by the private sector. In 1986 a study undertaken with the participation of development agen- cies and international organizations recommended a series of technical measures and policies that the Honduran government has begun to imple- ment (Republica de Hondura 1988). The resulting increase in the number of forestry professionals and their growing interest in the implementation of pending recommendations augur well for improvements in Honduran for- est management. COHDEFOR's Performance Forest Management Sustainable forest management was not a priority for COHDEFOR dur- ing its early years. Because its greatest income came from exports, COHDEFOR limited itself to buying sawn timber from commercial saw- mills. In doing so, COHDEFOR fixed a price irrespective of the quality of the timber. This policy discouraged private timber enterprises and encour- aged waste and low productivity. After its attempt at direct intervention in the timber industry failed, COHDEFOR sought to improve the use of the 3 Among these, Foresta! Industrial Agua Fria (FIAFSA), Central de Aserrio Siguatepeque (CASISA), and a pine resin processing plant were acquired in the late 1970s. 282 MANAGING THE WORLD'S FORESTS forest through concession contracts conditioned by detailed management plans. The contracts specify the permissible volume of timber that can be extracted from the indicated land, forest protection measures, roads to be constructed, and so on. Theoretically, COHDEFOR administers both private and state lands in geographically organized management units, but on-site monitoring is in- adequate and illegal exploitation is rampant. Management plans exist for exploitation areas such as La Mosquitia, Comayagua, Francisco Morazan, and Cortes. A detailed study of the Lajas management unit in the Comayagua region in 1985 pointed out COHDEFOR's overcentralization as one of the obstacles to implementing the management plans: The application of sanctions, the resolution of appeals and the signature of timber coritracts are the exclusive prerogative of the General Manager of COHDEFOR. In Las JLajas Management Unit, as well as in other Management Units, it has frequently happened that the central administration has been disregarding the manage- ment plan and approved increases in harvest volumes or increases in industrial capacity. Obviously, the communication at different levels in the forest administration is not satisfactory and decen- tralization of authority would make for great efficiency in man- agement (FAO 1985, 179). Studies undertaken in 1985 demonstrate that although the volume of exported timber has fallen to 1 million cubic meters annually, inefficiency in logging and initial processing of the timber has produced losses of up to 60 percent of the timber (USAID 1982, 76). To combat these inefficiencies, COHDEFOR is now marketing the timber on the basis of standing volume rather than actual removals. This type of pricing scheme reduces high-grad- ing and residual-stand damage. Training of Professionals and Research Honduras has been very successful in developing forest management expertise. The National School of Forestry Sciences (ESNACIFOR) is situ- ated within COHDEFOR, under the direction of a consultative council. Of the 600 Honduran forestry professionals, approximately 420 were trained at ESNACIFOR. Considered one of the best tropical forestry schools, ESNACIFOR also counts on the technical assistance of international organi- zations to provide training to farmers and technical personnel, to undertake studies of genetic improvements of forests, and to augment technical per- sonnel. Honduran and international experts conduct forestry research un- der the National Center for Applied Forestry Research (CENIFA), which publishes the forestry publications El Tatascan and La Floresta regularly. Unfortunately, most Honduran forestry professionals define their roles within the narrow confines of silviculture and commercial logging practices. This professional focus has left many COHDEFOR foresters uninterested in, FORESTRY INSTITUTIONS 283 or even hostile toward, the social forestry, community participation, and environmental aspects of an integrated forestry policy. Rent Extraction COHDEFOR has done a relatively good job of extracting rents from timber concessions. Synott (1990, 87) notes that the fees charged for stand- ing timber are "realistically high," -a situation he attributes directly to COHDEFOR's status as a self-financing parastatal organization. Initially, the greatest part of COHDEFOR's income came from exports. By purchas- ing sawn timber at prices fixed by COHDEFOR and selling it at interna- tional market prices, COHDEFOR earned an average annual income of US$41 million from 1974 to 1979. By 1979, however, total export volume had de- creased to 73 percent of the 1974 level,' and in 1988 the total volume of exports had fallen to 52 percent of the 1979 level. Because COHDEFOR typically offered very low lumber prices, the timber industry opted for production of secondary products. Currently the export of secondary forest products is nearly equal to that of primary products: US$20.8 million from processed sawn timber; US$2 million from processed pine resin and latex; and US$340,000 from seeds, especially Caribbean pine. Although the reduction of exports eased some of the pressure on the forest resource, the reduction of export revenues produced a financial crisis at COHDEFOR in the early 1980s. In 1986 new policies were established, as- suring COHDEFOR a minimum of 14 percent of net income on the f.o.b. value of exports of common sawn timber, and 18 percent of the value of high-grade timber. Social Forestry The law that created COHDEFOR provided a framework for organizing the peasants into cooperatives and other social organizations in order to "guard and protect the forests and promote their regeneration, prevent for- est fires, excessive pasturing, illegal logging and migratory agriculture" (Article 24). The law granted COHDEFOR the power to determine the form- and the area-in which farmers could participate in the benefits of forest exploitation through secondary products. The text of the law, however, clearly excludes farmers from the benefits of primary industry, the highest revenue producer. Currently, 104 agroforestry cooperatives are operating, with 4,080 mem- bers located in eight of the nine forest regions. Activities include pine resin extraction, fuelwood production, manual sawing, extraction of amber sap, and sustainable extraction of fuelwood for small-scale timber production. The average annual income per member is equivalent to US$400. Most 4 From 194.2 million board feet in 1974to142.3 million board feet in 1979 (SECPLAN 1989). 284 MANAGING THE WORLD'S FORESTS cooperatives are located in the broad-leaf forest, where deforestation and regeneration problems are most pronounced. Because social forestry did not contribute to COHDEFOR income, it was relegated to the lowest priority. In 1978, with a total financial support of US$5.65 million, COHDEFOR ear- marked only 0.2 percent to social forestry. The rest was budgeted for the sawmill industry (Cruz Guerrero 1990, 11). Beginning in 1983, COHDEFOR, with the support of the PAO, began a new stage in the social forestry system through the formation of integrated management areas (AMI). AMis consist of forest areas, from 1,000 to 10,000 hectares, managed by the community under the concept of multiple and integrated forest use and directed by a technician living in the community. Since 1987, 50 AMis have been developed throughout the country, and their numbers are expected to increase. Some 2,500 persons benefit directly from this new approach. Through simple techniques, the community is involved in the entire process of sustainable forest use, from the planning stage to the actual exploitation. Another social development model is the industry forestry energy so- cial system (SIPES). SIPES is composed of AMis with special characteristics (location in or near a mature forest, accessibility to markets, etc.). The sys- tem develops industrial activities beyond the AMI conception, including a self-sufficient wood-based energy system, the production of unfinished wood products, and electricity for the industrial area. Charcoal is produced from timber waste and brush. This approach increases employment, broadens forest use, provides electricity to remote areas, and substantially increases the rural income level. In 1986 and 1987 two prototype SIPES units were established.' Innovation and Adaptability The great challenge for COHDEFOR is to change its orientation from an institution that exports timber to one that has an integrated vision of sustainable forest use. Following evaluations of COHDEFOR's institutional structure and policies, COHDEFOR was decentralized; new social forestry policies were instituted; integrated management areas were established; and fiscal incentives for afforestation, reforestation, and forest protection were proposed. 5 Other agroforestry projects include the U.N. High Commission on Refugees project of forest restoration in La Mosquitia. In addition, with the support of a project of the World Food Program of the United Nations, COHDEFOR has tried to improve the living conditions of the agroforestry farmers through silvicultural im- provement, soil conservation, and improvement of communal infrastructure (resto- ration of community centers, construction of potable-water wells, etc.). These projects were reported to have directly benefited some 12,200 persons as of 1986. FORESTRY INSTITUTIONS 285 Col)clusions from the Honduran Case Study COHDEFOR may overemphasize the importance of financial incentives to promote maximum rent extraction at the expense of the other goals of forestry policy. The self-financing arrangement provides no incentive to focus attention on social forestry and forest conservation beyond silvicultural is- sues. To the credit of the integrity and professionalism of the COHDEFOR policymakers, these issues have received at least modest attention. In the hands of more short-sighted policymakers, COHDEFOR could have allowed much greater harvests in order to gain immediate revenues. Concern for the long-term sustainability of the forest and the agency is reflected in ceilings placed on cutting. After a decade of emphasizing expor- tation and industrialization, with disappointing results, COHDEFOR has begun to recognize that the sustainability of the timber industry depends on the participation of the rural population and the improvement of logging techniques. The inclusion of farmers in the COHDEFOR structure is espe- cially noteworthy, considering the highly conservative tradition of the Hon- duran economy. Nevertheless, clashes between forestry specialists and farm leaders persist. COHDEFOR's open attitude toward international and development agencies' suggestions for institutional reform is another distinctive feature of this institution. Recently COHDEFOR has begun to implement recom- mendations of the U.S. Forest Service, the United Nations Program for the Environment, the FAO, and others. These suggestions were developed through an international conference on Honduran forest development, which formulated policies and strategies for the 1986-90 period. The changes introduced into COHDEFOR certainly have promise, but success will depend on whether the institutional and financial structure of the agency can overcome the perverse incentives to focus predominantly on harvesting operations. Although sustainable management of the pine for- ests has improved, the institutional structure provides little incentive for reversing the deterioration of the broad-leaf forest and its greater biodiversity. The performances in social forestry and environmental protection may im- prove through successful decentralization to the AMis, but the institutional incentives for COHDEFOR to devote resources to this initiative are still lacking. THE MALAYSIAN CASE STUDY Malaysia was formed in 1963 when Sabah and Sarawak, located on the island of Borneo, joined the 11 states of Peninsular Malaysia in a federated system of governance. In the original agreement, Sabah and Sarawak were to be on an equal footing with the combined states of Peninsular Malaysia in a tripartite power structure. Sabah and Sarawak still hold to this view, but government officials in the capital, Kuala Lumpur, have tried to shape poli- cies that put each of the 11 states of Peninsular Malaysia on equal footing 286 MANAGING THE WORLD'S FORESTS with Sabah and Sarawak. Because state governments control nearly all mat- ters of forest policy, the political debate over state versus federal power often plays itself out in the formulation of forest policy. Sabah The formal mandate of Sabah' s forestry institutions recognizes the im- portance of a multiple-objective forest policy. Unfortunately, the land-use policy in Sabah emphasizes economic benefits that easily manifest them- selves in the formal marketplace. Roads roughed in for mining and logging operations turn into invitations to the landless poor to enter an area and begin a relentless process of agricultural conversion, through shifting culti- vation or settled agriculture. The problem is particularly acute in Sabah, where vestiges of British colonial law allow citizens to gain title to forested land by clearing and cultivating it. Sabah has 7.4 million hectares of land, of which 4.7 million were under forest in 1981. Annual deforestation rates are estimated at 60,000 hectares for the 1976-80 period and 76,000 hectares for the 1980-85 period (Gillis 1988a, 116). Institutions Sabah has an exceptionally large number of institutional structures re- sponsible for the formation and implementation of forest policy. These in- clude the Forestry Department, the Sabah Foundation, the Sabah Forest Development Authority (SAFODA), the Sabah Rural Credit Development Authority, and the Office of the Chief Minister. Each is discussed in the paragraphs that follow. The Forestry Department. Sabah's Forestry Department has the traditional divisions of a headquarters section, a research section, and a district administration section. District administrative personnel are spread throughout the state. The Forestry Department has the responsibility for interpreting and enforcing forest law; preparing and revising long-term license agreements; supervising all scaling and export operations; providing cartographic services; compiling production, revenue, and expenditure accounts; supervising and enforcing wildlife initiatives; managing mangroves; conducting research on plantations, ecology, silviculture, botany, entomology, and forest use; and administering district offices (Seventh Malaysian Forestry Conference 1979, 291). The director of the Forestry Department now seeks to control harvest regulations while dividing the permanent forests themselves between the Sabah Foundation (in the east and south), SAFODA (in the north), and Sabah Forest Industries (in the west). Whether .this approach will continue beyond the current director's tenure is uncertain. FORESTRY INSTITUTIONS 287 The Sabah Foundation. Perhaps the most unconventional of the many forestry institutions in Malaysia, the Sabah Foundation was created in 1966 to use the forest resource to improve the social infrastructure available to citizens of Sabah. At that time, the foundation was granted a 100-year license agreement for 855,000 hectares of virgin forest. As of 1989, the jurisdiction of the foundation had expanded to include a total of 1,070,866 hectares (Burgess 1989). In 1970 the Sabah Foundation received permission from the state legis- lature to establish companies to undertake commercial activities. The foundation's social development and commercial enterprise activities are officially split into a "people development" initiative and a holding com- pany. Activities of the social development arm include human resource development, educational research and development, resource development and management, product demand and marketing, and social welfare ser- vices. The commercial arm of the foundation, streamlined into the Innoprise Corporation holding company in 1988, is involved in forestry, transporta- tion, wood processing, real estate development, and tourism. The structure of the Sabah Foundation provides a mechanism for ensur- ing that revenue from local natural resources remains within the state and is spent on local social development projects. Thus the state legislature tied the foundation's mission of forest management to the organization's re- sponsiveness to the needs of local populations. The Sabah Foundation also provides periodic cash grants to all registered residents of the state of Sabah. The 100-year concession agreement granted to the Sabah Foundation by the state government should have facilitated long-term planning for the forest resource. The foundation received the authority to contract out the land under its jurisdiction. Unfortunately, the "subconcessions" were for only 1 to 10 years, and thus provided little incentive to the lessee to develop a sustainable management plan. This situation would have been acceptable if the Sabah Foundation had clearly prescribed a management plan for the leased area, but it did not. Although the reason is uncertain, the short-term revenue needs that beleaguer most development agencies may have driven this policy. Total value of timber exports skyrocketed under foundation manage- ment, but the expansion appears to have been bought at the price of rapidly accelerating deforestation rates and to have depended on booming com- modities markets. By explicitly making one of the forestry institutions a development agency, the state government may have crippled the foundation's ability to focus on long-term environmental and economic sustainability. The Sabah Foundation tends to treat the forest as a resource to be mined for maximum foreign exchange, tax, development, and employment ben- efits. Between 1971 and 1982 total value of timber exports increased by 450 percent. Rent capture was the most aggressive and successful in the world, and timber employment stood at 35,000 (or 7 percent of the state 288 MANAGING THE WORLD'S FORESTS population) in the late 1970s (Gillis 1988a). Therefore, although this uncon- ventional structure improved the institution's performance in terms of rev- enue generation and social forestry, it paid little attention to the environ- mental services provided by the forest resource. The Sabah Forest Development Authority (SAFODA). SAFODA is essentially a plantation authority charged with the promotion of rural development on 250,000 hectares of infertile land in western Sabah. The agency oversees shifting cultivation patterns and reforestation plans. To finance SAFODA the state government levied a special tax on sawn timber. Because the tax is based on volume, it tends to exacerbate high-grading problems. In addition, from 1973 to 1980, only 31 percent of the land that could have been replanted was actually reforested. In addition, SAFODA oversees several resettlement programs, the most significant of which is located on Bengkoka Peninsula in northern Sabah. It seems ironic that an agency set up to reforest degraded land would be financed by a tax that, through the promotion of low recovery rates and inefficient processing, actually exacerbates the extraction pressure on the forest resource. While SAFODA is difficult to evaluate because of the scanty information available on the resettlement programs, one suspects that, like the Sabah Foundation, it was established as a forestry institution with broader development agency overtones. Rural Credit Development Authority and the Office of the Chief Minister. Information on the role of the Rural Credit Development Authority and the Office of the Chief Minister is hard to find. The Rural Credit Authority is essentially a lending institution that extends credit to improve the farming techniques of shifting cultivators and to the Sabah Foundation (for unknown reasons). The Office of the Chief Minister has granted concessions from time to time and determines tax and royalty levels. Taxes, Royalties, and Rents Sabah has two main advantages over most of the rest of the timber- producing developing world: its forests are full of highly valued tropical species, and the remaining timber stands are still relatively accessible for logging purposes. This combination of factors has allowed Sabah to assess the world's highest timber royalty (Gillis 1988a, 128). Sabah' s rent capture is the highest in Malaysia and among the highest in the world. Using Indonesian cost data as a proxy for Sabah's logging and milling costs, Gillis (1988a) estimated rent capture to be between 81 percent and 92 percent in the 1979-83 period. A more recent study by Vincent (1990) calculates Sabah' s rent capture to be roughly 46 percent during the same period. Unfortunately, like most ad valorem taxes, the royalty system in Sabah exacerbates high-grading and thus puts additional pressure on the forest FORESTRY INSTITUTIONS 289 resource. In addition, Sabah suffers from extremely high residual stand damage. Fully 72 percent of the remaining trees are damaged by high-grading extraction techniques, compared with 42 percent for neighboring Sarawak (Repetto 1988, 54). Sabah has also used the forestry tax system to encourage investment in sawmilling capacity. Royalties for logs processed domesti- cally are set at only 10 percent of current f.o.b. price (Vincent 1990). The notorious inefficiency of the local processing plants also accelerates defores- tation rates. Perhaps the most striking aspect of the forest tax system is that the Forestry Department does not set the levels of taxes or royalties, nor is its budget directly dependent on revenue levels. Thus whatever good-faith efforts the Forestry Department might make to establish long-term sustainable forestry policies, its efforts can easily be undermined by budgetary uncertainty and by policies established outside the Forestry Department that provide incentives counterproductive to the goals of sound forest management. Environmental Objectives The Sabah Foundation places limited emphasis on the less tangible ser- vices provided by the forest resource. Management of wildlife and national parks is under the jurisdiction of the Forestry Department and thus receives insufficient emphasis. Broader issues of watershed management, erosion, water quality, and the like also tend to get lost in the patchwork system of forest management. To its credit, Sabah did establish Kinabalu Park, a na- tional park and reserve covering more than 700,000 hectares. The park cur- rently averages more than 100,000 visitors per year, thus contributing sub- stantially to Sabah's tourist revenue (Rubeli 1989, 65). Peninsular Malaysia Peninsular Malaysia contains 13.2 million hectares of land, of which 6.3 million were under forest in 1981. The annual deforestation rate held steady at 90,000 hectares over both the 1976-80 and the 1980-85 periods (Gillis 1988a, 116). In general, forest management in Peninsular Malaysia is under- taken by conventional forestry institutions, and the region places a strong emphasis on agricultural conversion for the alleviation of rural poverty. In fact, agricultural conversion caused 90 percent of deforestation during the 1980s (Repetto 1988, 58). The Forestry Mandate Each of the states in Peninsular Malaysia has a state forestry office. In an attempt to standardize management of the forest resource, a nonbinding National Forest Policy was endorsed by the National Forestry Council in 1977 and then by the National Land Council in 1978. The National Forest 290 MANAGING THE WORLD'S FORESTS Policy, which is concerned primarily with technical forestry and the system of forest reserves, includes provisions to designate some 5 million hectares of land in Peninsular Malaysia as the permanent forest estate;' impose maximum annual forest removals of 76,000 hectares of new forest; limit agro-conversion to 68,000 hectares for 1976-.SO; and establish uniformity across states in the granting of licenses, allowable harvesting methods, and forest regeneration policies (Gillis 1988a, 159). The National Forestry Act, which formalized the National Forest Policy, was approved by the federal cabinet in 1984 and has subsequently been adopted by all states in Peninsu- lar Malaysia. The federal government was to bear most of the costs for developing the permanent forest estate. In addition, a "Compensatory Plantation Pro- gram" asks states to set aside acreage for plantations of fast-growing spe- cies, to be financed through interest-free loans extended to the states by the federal government. The government also provides some technical and management support. In addition to the permanent forest estate, as of 1986 there were 2.7 million hectares of state land in forest, of which roughly 30 percent was scheduled for eventual conversion to agriculture. There were also 500,000 hectares of wildlife reserve in Peninsular Malaysia (ITTO Pre- Project Report, vol. II, 47). Land Management Technically, the forest resource in Peninsular Malaysia is supposed to be managed on a sustainable-yield basis to ensure a steady supply of tim- ber. Between 1966 and 1980, however, the government undertook a sizable development plan designed to reduce rural poverty. This plan led to mas- sive conversions of forested land to agriculture. Although still liquidating some forest resources, Peninsular Malaysia appears to be approaching a forestry-agriculture balance. How close the resultant land-use distribution is to the optimal remains unclear, however. Taxes, Royalties, and Rents The forest sector of Peninsular Malaysia has consistently generated more revenue than the government has appropriated for forestry offices, so that forest-sector revenue is being used to finance government activities in other sectors of the economy. While the national headquarters has received a reasonable level of funding, the state forestry offices, which bear the brunt of the responsibility for forest management, seem to be underfunded. Royalty levels for each of five different species groups are fairly uni- 6 Some 3.3 million hectares of this area were considered productive forest; of this productive forest, 2 million hectares were ·Virgin forest. FORESTRY INSTITUTIONS 291 form across all 11 states of Peninsular Malaysia. Although royalties are supposed to be set at 10 percent of the current log price, 7 of the states have not updated their royalty levels since 1972. In addition, the royalties are log- based, and are thus destructive in terms of high-grading and damage to the residual stand. The states also assess a premium on the area of forest under contract which exceeds the area of productive forest. Overall, Vincent (1990) calculates that Peninsular Malaysia manages to capture roughly 22 percent of the rent available from the forest resource. Although the differentiated royalty levels tend to reduce the problems of high-grading and damage to the residual stand, the log-based charges tend to exacerbate the problems. On much of the land being logged in Peninsular Malaysia, high-grading may not be a source of real concern because the land is slated for eventual conversion to settled agriculture. Within the per- manent forest estate, however, better data on residual damage would allow the state governments to formulate a more comprehensive forest policy. In addition to obtaining revenue from taxes and royalties, several states in Peninsular Malaysia have begun tendering concession contracts. Under the reasonably competitive conditions that appear to exist in Peninsular Malaysia, the tendering system can be expected to increase the proportion of rent captured by the governing institution. This type of system greatly simplifies the calculations that the forestry institution must go through in order to set rent capture instruments at the correct levels. Development of Forest-based Industry Peninsular Malaysia has a policy of managing the permanent forest estate for the supply of forest produce for agricultural, domestic, and industrial purposes both within the state and for export. Once domestic requirements have been met, there is no clearly defined level of production targeted for export. As early as 1972 Peninsular Malaysia established poli- cies to encourage the development of forest-based industry. The Malaysian Timber Industry Board (MTIB) was established to promote the exportation of Malaysian timber and timber products. Perhaps the critical policy in terms of promoting the development of wood-based industries was the decision to restrict log exports. Log-export restrictions, coupled with domestic policies that made logs available to domestic sawmills at below interna- tional market prices, led to fairly rapid expansion of domestic milling capability. Once the industry was in place, state governments were hard- pressed to restrict extraction. Despite the infant industry protection that Peninsular Malaysia provided to the primary processing sector, the govern- ment did little to protect secondary processing activities. As might be ex- pected, this situation led to overcapacity in primary processing and undercapacity in downstream processing. 292 MANAGING THE WORLD'S FORESTS This unbalanced pattern of forest industry development placed addi- tional stress on the forest resource. As a result, total timber production in Peninsular Malaysia is expected to be insufficient to meet even domestic requirements in the near future. In response Ito this anticipated shortfall the government has decreased logging rates. In addition, the Forestry Depart- ment has developed plans for silvicultural treatment of logged-over areas, regeneration of deforested areas, and establishment of forest plantations for fast-growing species on a 15-year rotation. Forest industry managers are also trying to develop schemes for importing logs from Sabah and Sarawak. Political tensions between the state and federal governments, however, may make this impossible. Environmental Benefits and Education Peninsular Malaysians have become increasingly aware of the impor- tant role that the forest resource plays in preserving environmental integ- rity. The National Forest Policy calls for community outreach efforts to improve environmental awareness and for schools to foster an interest in forest science; the plan also emphasizes the provision of recreational facili- ties in the forests, which local populations have enthusiastically embraced. The Forest Research Institute, established in Kepong in 1929,7 has a staff exceeding 500, including 52 research scientists, who study a variety of for- estry issues (Salleh 1983). In addition, the forestry faculty at the Agricultural University in Serdang offers a four-year program in forestry and the univer- sity campus in Kuching offers a three-year diploma course. Sarawak Sarawak contains 12.5 million hectares of land, of which 9.4 million were under forest in 1981. Annual deforestation rates are estimated at 80,000 hectares for the 1976-80 period, and 89,000 hectares for the 1980-85 period. All the forested area belongs to the state and is divided into permanent forests and statelands. The most significant pressure on the forest resource is the proliferation of shifting cultivators who, in 1979, were estimated to burn nearly 60,000 hectares of forest annually (Seventh Malaysian Forestry Conference 1979, 311). The Forestry Mandate The basis for forestry management is the 1954 Forest Policy of Sarawak, which was passed by the governor in council No formal land-use or land- classification system has been developed, but the widely accepted norm is 7Sabah and Sarawak also have forest research units, but they are less well devel- oped than the facilities at Kepong. FORESTRY INSTITUTIONS 293 that forested land suitable for agriculture will eventually be converted. As yet, however, land conversion to settled agriculture has not been a signifi- cant source of deforestation. The permanent forest estate consists of forest reserves, protected for- ests, and communal forests. Forest reserves are managed for productive forestry, protected forests are managed for traditional hunter-gatherer ac- tivities with a ban on any logging, and communal forests are designated to supply domestic timber to the surrounding communities. Stateland forests have never been clearly defined. The Forestry Department has the authority to license logging but has no jurisdiction once the timber has been removed; as a result, sustainable management is virtually impossible. This irregular pattern of jurisdiction results from customary rights held by the local Dayak people (Gillis 1988a, 143). Institutions Authority over the forest resource is held by the Ministry of Forestry. Responsibilities are divided between the Forestry Department and the Sarawak Timber Industry Development Corporation (STIDC). The Forestry Department concerns itself with reforestation, protection, operations, and research. It has a conventional structure and is understaffed. The STIDC promotes industrial development and coordinates all industrial activities. Land Management The length of concessions in Sarawak ranges from 1 to 10 years. Al- though these terms seem rather short, they are consistent with the long- range plan of agroconversion. Sarawakians themselves widely acknowledge that corruption abounds in the timber concession-granting process. Logging has long played an important part in the money economy in Sarawak. The sector accounted for 22,000 jobs (9 percent of the total work force) in 1984. In 1989 Sarawak exported more than 15 million cubic meters of timber-a figure that constitutes 35 percent of the world's unprocessed timber exports (Hendrix 1990, Al). The state government has established a series of incentives designed to attract private investment in processing mills. Taxes, Royalties, and Rents The two most important forest charges in Sarawak are the royalty and the export tax. Revenue from forest taxes, royalties, and licenses accounted for 20 to 25 percent of state revenue during the mid-1980s (Gillis 1988a, 147) and for about half the state budget in 1989 (Hendrix 1990, Al). Like Sabah, Sarawak uses tax incentive schemes to encourage local processing. The tax discount on logs processed locally is fully 50 percent of the normal royalty (Vincent 1990). Sarawak also levies a 10 percent export tax. Forest charges in 294 MANAGING THE WORLD'S FORESTS Sarawak are relatively low and tend to be differentiated according to the value of species, reducing high-grading problems. The export tax is the only volume-based tax, so that pressure to high-grade is minimal. As a result there is relatively low residual damage to the remaining stems after an area has been harvested. Rent capture is calculated at about 18 percent (Vincent 1990). Environmental Objectives Despite their importance in the local economy, minor forest products have been essentially ignored during the design and implementation of forest policy, and environmental objectives receive even less emphasis. Conclusions from the Malaysian Case Study The only truly innovative institution in forest management in Malaysia is the Sabah Foundation. The very design of the foundation links the exploi- tation of the forest resource to improvement in the standard of living of native Sabahans. Given the large financial commitment that most develop- ment projects require, this emphasis on social development gives the foun- dation a strong incentive to capture as much rent and generate as much revenue as possible. Unfortunately, environmental objectives-the third important area of forest policy-are lost in the pursuit of the other two objectives. An unbalanced system of incentives to promote downstream forest in- dustry has evolved in all three regions of Malaysia. The system of incentives is most skewed in Peninsular Malaysia and least skewed in Sarawak. This system correlates positively with the extent o:f overcapacity in the primary processing sector. Although Malaysia has a large number of well-trained forestry person- nel and an impressive history of forestry research, forest policy emphasizes the wood production capabilities of the forest resource. Nonmarket benefits associated with the forest resource are not considered beyond the language of the legislation of the forestry mandate. A few areas that have already experienced flood control problems in Peninsular Malaysia, however, are beginning to incorporate the broader role of the forest in their resource management plans. As the local environmental impact of high rates of deforestation becomes more widely understood, local, state, national, and even international envi- ronmental groups are taking a more active role in Malaysian natural re- source politics and economics. These groups include the International Union for Conservation of Nature and Natural Resources, the Worldwide Fund for Nature Malaysia, the Malayan Nature Society, the Environmental Protec- tion Society, and Friends of the Earth-Malaysia. Politically vocal community- based environmental groups may do more for environmental reform than the self-policing functions of the state forestry departments. FORESTRY INSTITUTIONS 29S GENERAL CONCLUSIONS The unconventional structures created to cope with the broad challenges of forestry policy entail a greater separation .between the central govern- ment apparatus and the decisions specifically focused on forestry policy. A common accompaniment of this separation is a financing arrangement that provides partial or full funding directly from forestry operations, or a devo- lution of authority that places financial gains and budgetary decisions in the hands of local or community institutions. Unconventional structures also typically expose the agency and its staff more directly to the consequences of their own policies and actions. Unconventional structures have the poten- tial to tie the fortunes of the agency and its personnel to the agency's own impacts: the rents extracted, the sustainability of forest yields, and the wel- fare of the people served by forest resources. Although some observers tend to regard decentralization and devolu- tion as ends in themselves, we believe that the superiority or inferiority of these structures depends on whether the advantages are worth any problems introduced by deviating from the traditional ministerial model. If the unconventional structures are to be considered successful, they must support policies that take a long view of the forest resource, balance multiple objectives, and respond to the needs of the population. The experiences of unconventional structures provide important lessons in the trade-offs between conventional and unconventional approaches, as well as contributing to a framework for designing forestry institutions that best fit specific contexts. Lesson No. 1: Revenue Generation at the Expense of Social Forestry The forestry agency's responsiveness to the needs of the population is manifested in its ability to extract rents from foreign and local loggers, willingness and ability to police the forest, commitment to addressing the needs of future generations through suitably long time horizons, and ability to promote social forestry development. When the forestry institution's budget depends on rent extraction, as is usually the case with unconventional institutions, greater rent extraction becomes an important incentive. Reasonably good rates of rent extraction prevail in Colombia, Honduras, and Sabah. These rents would be higher if the forestry institutions moved to a system of auctioning off concession agreements. Some progress in this direction has already been made in Pen- insular Malaysia. Because these institutions handle large sums of money from concessions or state timber exports, however, they are susceptible to internal corruption or "raids" by government officials. Widespread accusa- tions of corruption have plagued the forestry institutions in Honduras and Sabah. It is unclear, however, whether conventional structures can boast a better record. 296 MANAGING THE WORLD'S FORESTS In Colombia, decentralization and devolution may have spawned fruit- ful experiments, but INDERENA's ability to provide oversight and policy coherence to the forestry sector has become increasingly tenuous. When forestry exploitation and conservation are basically left in the hands of a corporation like Carton de Colombia, which has better information and expertise than the government has, the state's ability to police the use of forests, should such policing become necessary, is problematic. Whether unconventional structures encourage longer time horizons in the perspectives of forestry officials depends on the combination of the explicit mandate and the specifics of the institution's financing formula. Establishing a "trust fund" operation like the Sabah Foundation, with a mandate to engage in long-term development, can give the forestry institu- tion an incentive to launch long-term development projects. The logic of the COHDEFOR arrangement also comes close to linking a sustainable harvest- ing objective with the sustainability of the institution itself. As the Hondu- ran case suggests, however, leaving all decisions to the discretion of the forestry institution can easily lead to overproduction. One possible solution to this problem is for the agency to charge a user fee that discourages undue exploitation. Unconventional structures can promote responsible social forestry. The Sabah Foundation is in many ways a social forestry institution itself, be- cause its very structure connects resource extraction to poverty alleviation among shifting cultivators. Notwithstanding its problems, it tries to attack the root cause of deforestation directly. In addition, decentralization and devolution may improve the record on social forestry by placing forest management decisions in the hands of the people most directly affected by it. Thus some of the Colombian devolution experiments hold promise for constructive social forestry. This potential can be realized only if the unconventional institution has a social forestry priority deeply rooted in its mandate. Social forestry in Honduras is practically moribund because other priorities, such as interact- ing with concessionaires and conducting other forestry operations for Hon- duras, have squeezed out social forestry. Lesson No. 2: The Autonomy of the Forestry Institution When a forestry institution is established outside the ministerial sys- tem, the institution typically has more autonomy in establishing and pursu- ing forestry policy. It may be the only institution committed to the needs of the forestry sector, so its ability to operate without the constraints of a supervisory ministry is an important advantage. Yet it is typically less con- nected to other policies that affect the forest resource. Except in Peninsular Malaysia, where forestry policy has been part of the strategy of agricultural conversion to high income-yielding cultivation (including agroforestry em- phasizing perennial crops such as palm oil and cacao), forestry policy in FORESTRY INSTITUTIONS 297 Malaysia, Colombia, and Honduras has had little connection with agricul- tural policy. It is difficult for an isolated forestry agency to address the problem of shifting cultivation or grazing, which forestry officials recognize as a major danger to forest sustainability. Although forest development may be subor- dinated when the conventional forestry department is situated within a ministry, at least the conventional department has direct organizational links with other central policy-making entities. Thus a forestry department that is part of an agriculture ministry, for example, may be in a better position to focus the attention of agricultural policymakers on the problems just mentioned. Lesson No. 3: The Forestry Institution's View of the Forest Mandate For both unconventional and conventional forestry institutions, the ob- jectives of revenue generation, social forestry, and resource preservation often conflict. In conventional forestry institutions, officials rank these ob- jectives on the basis of their own predispositions or they seek guidance from higher reaches of government leadership. In unconventional institu- tions, the objectives that are crucial to the agency's institutional interests are pursued over others. Again, the partial or full self-financing arrangements in Honduras and Sabah rank rent extraction far above environmental and wildlife conservation. The institutional interests of a forestry institution could be tied to conservation (perhaps a national parks agency that depends on proceeds from ecotourism would be in this position), but this is not the case for the three countries reviewed. · When separated from the central government, the forestry agency may take too narrow a view of its mandates and responsibilities, particularly if it is understaffed or underfinanced. Restricted finances may lead to an em- phasis on staff-conserving functions like negotiation of large concession contracts rather than development of social forestry programs. A preoccu- pation with downstream industrialization of forest products, even when of dubious economic value, also reduces the resources available for a broader range of activities. When a state forestry institution has complete financial autonomy, the government often cannot impose a better balance among objectives. Lesson No. 4: Ecological and Economic Sustainability A government that is committed to sustainable forestry must ensure that forest policy is consistent with the multiple objectives that sustainability requires. This is particularly true in the case of ensuring adequate attention to preservation. Decisions to set aside areas for biodiversity or wildlife pro- tection, to police against poaching and illegal logging, to provide alternative 298 MANAGING THE WORLD'S FORESTS opportunities to shifting cultivation, or to set the overall harvesting rate all influence the environmental services that the forest resource provides. When unconventional forestry institutions are tied financially to rent extraction or timber operation revenue, they are likely to focus their atten- tion on the overall harvesting rate decision. This decision obviously affects the preservation of species and wildlife as well as the wider environmental services provided by the forest. However, appropriate conservationist prac- tices must be strictly enforced apart from the overall harvest decision. The inability to put a monetary value on the environmental services provided by forests makes it especially difficult to establish incentives to cover less tangible objectives of forest management. In the area of environmental services, unconventional forestry institu- tions offer no real improvement over convenlional institutions. None of the state forestry institutions in Colombia, Honduras, or Malaysia enforced strin- gent conservation measures. In much of the developed world, local activism has evolved to police the preservation functions of the responsible agencies. The solution to the environmental services question may be to create the pressure for reform from without rather than from within. Thus govern- ments should provide sound environmental education programs and en- courage responsible activism in their development plans. Decentralization to subnational authorities and devolution to private citizens and groups can (but do not necessarily) stimulate the development of conservation-minded nongovernment organizations and grass-roots groups. The devolution ini- tiatives in Colombia and the state-based decentralization in Malaysia cer- tainly have this potential. However, the institutional arrangements of semiautonomous agencies (or state enterprises), although decentralized in the sense of not being part of the central ministerial structure, are still cen- tralized if they keep the agency remote from local participation. An alternative approach to structuring the regulation of preservation measures would be to mandate a conventional (ministerial) conservation agency to regulate the actions of the unconventional forestry development institution. Part of the weakness of the COHDEFOR model is that the agency has responsibility for regulating its own actions, even though its institu- tional incentives put no priority on preservation. The Colombian INDERENA model would be better than the Honduran model if the regulating agency had the mandate and staff to enforce such regulations. Despite the limitations of unconventional institutions to pursue the full range of important forestry functions, it is clear that modifying the institu- tional structure can have a significant and predictable effect on how forestry institutions behave. The farsightedness required for sustainable forest ex- ploitation is not guaranteed by deviating from the conventional ministerial structure, but some deviations do offer the prospect of linking institutional interests with the future. The danger is that the most prominently mandated objective will crowd out other objectives, such as long-term conservation and responsiveness to local needs. Whether through a combination of min- isterial and autonomous agencies or through the construction of multiobjective FORESTRY INSTITUTIONS 299 mandates, balancing the multiple objectives of forest management requires balancing the incentives that affect the forest sector. REFERENCES Adeyoju, S. Kolade. 1976. "Forestry Administration Problems in Selected African Countries." Food and Agriculture Organization of the United Nations (FAO), November. Burgess, P. F. 1989 "Asia." In Duncan Poore, No Timber Without Trees. London: Earthscan Publica- tions. Corporacion Hondurena de Desarrollo Forestal. 1989. Political de Desarrollo Forestal de Honduras. COHDEFOR-AID, Proyecto de Desarrollo Foresta!, No. 522-0246. Cruz Guerrer, Arnulfo. 1990. COHDEFOR and Sustainable Development: A Case Study. Durham, N.C.: Duke University Center for International Development Research. Departamento Nacional de Plancacion (DNP). 1989. "Tropical Forestry Action Plan for Colombia (TFAPC)." Bogota, Colombia. Food and Agricultural Organization (PAO). 1985. Intensive Multiple-Use Forest Management in the Tropics. Analysis of Case Studies from India, Africa, Latin America, and the Caribbean, No. 55. Forest Development Authority, Government of Liberia (FDA). 1986. "National Progress Report on Forestry." African Forestry Commission Seventh Session, Bamako, Mali, January. Gillis, Malcolm. 1988a. "Malaysia: Public Policies and the Tropical Forest." In Public Policies and the Misuse of Forest Resources, eds. Robert Repetto and Malcolm Gillis. Cambridge, Eng.: Cambridge University Press. _ _ _. 1988b. ''West Africa." In Public Policies and the Misuse of Forest Resources, eds. Robert Repetto and Malcolm Gillis. Cambridge, Eng.: Cambridge University Press. Hendrix, Kathleen. 1990. "Vanishing Forest Fells a Way of Life." Los Angeles Times, March 18. International Bank for Reconstruction and Development (IBRD). 1977. "Forestry: Sector Policy Pa- per." Report No. 1778. International Tropical Timber Organization (ITTO). 1988a. "Malaysia." In Natural Forest Management for Sustainable Timber Production, Pre-Project Report. Vol. II. London: International Institute for Environment and Development (IIED). October. _ _ _. 1988b. "Liberia." In Natural Forest Management for Sustainable Timber Production: Pre-Project Report, Vol. II. London: IIED. _ _ _. 1988c. "Honduras." In Natural Forest Management for Sustainable Timber Production: Pre- Project Report, Vol. II. London: IIED. Llaurado, J. Prats, and G. Speidel. 1981. "Public Forestry Administrations in Latin America." Food and Agricultural Organization of the United Nations, FAO Forestry Paper No. 25, Rome. Poore, Duncan, Peter Burgess, John Palmer, Simon Rietbergen, and Timothy Synnott. 1989. No Timber Without Trees: Sustainability in the Tropical Forest. A study for International Tropical Timber Organization. London, Eng.: Earthscan Publications. Projecto de Desarrollo Foresta!. "Cambium" in Publication Mensural 1, no. 1, April 1989. Repetto, Robert. 1988. The Forest for the Trees? Government Polices and the Misuse of Forest Resources. Washington, D.C.: World Resources Institute. Republica de Hondura. 1988. Politicas, Medidas, Estrategias y Acciones Prioritarias. Roundtable Discus- sion about International Participation in Honduran Forestry Development. Rubeli, Ken. 1989. "Pride and Protest on Malaysia." New Scientist, October 21. Seventh Malaysian Forestry Conference. 1979. Malaysian Forester 42(4):286--347. Salleh, Mohd Nor. 1983. "Forestry in Malaysia." Journal of Forestry (March):164f. SECPLAN. 1989. Percil Ambiental de Honduras, 1989. Honduras. Synnott, Timothy. 1990. "South America and Caribbean." In No Timber Without Trees, eds. Duncan Poore et al. London: IIED. Thomas, Vinod. 1985. Linking Macroeconomic and Agricultural Policies for Adjustment with Growth: The Colombian Experience. Washington, D.C.: World Bank. U.S. Agency for International Development (USAID). 1982. Honduras: Country Environmental Profile. A Field Study. Vincent, Jeffrey. 1990. Personal communication, October 19. Yusop, Ali. 1982. "The Potential for Forest Industrial Development in Sarawak in the Eighties." Malaysian Forester 45(4):460-67. 12 A Sociological Framework: Policy, Environment, and the Social Actors for Tree Planting Michael M. Cernea ENVIRONMENTAL PROBLEMS AND SOCIAL ORGANIZATION L ike most other global environmental problems, the world's looming de- forestation crisis will not be slowed, let alone arrested, through technical remedies alone. The root causes of deforestation are outside the forest sector itself: these causes are social, demographic, and economic, and countering them requires institutional and policy changes. The ultimate success of any reforestation strategy depends on the social forces that can be summoned to accomplish the task. Such forestry strategies are called not to pursue simply conservation of the remaining forests, but also to trigger massive additional tree planting both inside and outside forests. Thus reforestation provides a convincing illustration of the proposition that the resolution of the environ- mental problems of development is to be searched for in the realm of social organization. For a Jong time the forestry programs of many governments and devel- opment agencies have lacked an explicit action-oriented sociological foun- dation. In fact, these programs have been little concerned with social actors, beyond the forestry departments and their technicians. The programs tended to regard people only as part of the problem rather than as part of the solution. Even the social forestry approaches promoted in the 1980s, par- ticularly those usually called community woodlots, as we shall see in detail later, were not designed around adequate social actors. In addition, they lacked adequate incentives and benefit distribution arrangements. Confused sociological conceptualization and lack of sound social engineering have long impaired the effectiveness of public investments in forests and have 301 302 MANAGING THE WORLD'S FORESTS preempted the positive environmental effects that many social forestry pro- grams could have reached, Improved and enduring forms of social organi- zation-tenure systems, structures for collective action, new institutional arrangements, legal frameworks and enforcement mechanisms-are needed to reverse deforestation processes. The starting premise of this chapter is that correctly addressing the sociocultural issues involved in investing for tree planting and forest man- agement is as critical for success as is resolving the economic and technical issues. In the formulation and implementation of an overall policy for the forest sector, it is necessary to overcome the past underestimation of people's potential role in forest management and to pursue pragmatic step-by-small- step approaches that involve specific social actors and well-identified units of social organization, This chapter examines the sociological components of afforestation pro- grams and, in particular, explores what specific social actors may conceiv- ably be mobilized, organized, and empowered-technically and financially- to carry out reforestation strategies. 1 The first section discusses the centrality of people, as opposed to commodities, in formulating environmentally sound development policies. The second and third sections define the social actors of deforestation and the failure of markets, the goals of social forestry as a people-oriented strategy, and the need to design it around well-identified social actors. The fourth section analyzes the sociological reasons why com- munity woodlot projects have largely failed. The last two sections outline two fundamental types of actor-centered strategies in tree planting: family- centered and group-centered approaches. 2 THE SOCIOLOGICAL PERSPECTIVE ON TREE PLANTING Commodities and People Although the abbreviated formula "forest policy" or "forest-sector policy" is commonly used in development jargon, for the purpose of this chapter I will use a longer name. Indeed, the policy that refers to the forest sector should be defined as the policy on tree planting and forest management. 1 An earlier version of this chapter was presented in the workshop on forestry sector policies organized in December 1990 at the World Bank. The initial paper was written and used as a background sociological framework for the formulation of the World Bank's policy statement on the forest sector (published in September 1991). The current chapter is considerably expanded and draws on the author's earlier work in this domain (see User Groups as Producers in Participatory Afforestation, World Bank Discussion Paper No. 70 1989). 2 Communication with, and comments from, A. Banerjee, 0. Baykal, Ruth Cernea, R. Godoy, S. Guggenheim, Augusta Molnar, W. Partridge, V. Rajagopalan, and J. Spears helped refine some points of this chapter. A SOCIOLOGICAL FRAMEWORK 303 This wording is not chancy: there is more in a name or a concept than first meets the eye. "Forest policy" may simply indicate concern with commodi- ties: wood, timber, fuelwood, and the like, whereas "tree planting" and "forest management" indicate a policy about human activities relating to trees and forests. And indeed, policies are about human activities, not about commodities. The principle expressed in this conceptualization is that the very category of "forest" cannot and should not be the policy object. Policies apply to people and institutions. 3 Thus forestry development policies and programs should explicitly focus on human activities-tree planting and forest management-rather than on timber, fuelwood, poles, or other commodities. Some development practitioners may tend to dismiss this point as mere semantics, but it is not. Vocabulary is germane to substance. 4 The starting point of any discourse on development largely influences the outcome: fo- cusing on commodities tends to lead to a different thinking process than explicitly focusing from the outset on the social actors-the people, the institutions, and their activities. The latter type of thinking process requires constructing the policy approach around the social actors of development- the idea reflected in the title of this chapter. The other components of the policy-the technical, economic, and environmental ones-are integrated into the course of action prescribed by the policy for the multiple social actors it addresses. A Historic Social Transition It is important first to place the current worldwide deforestation crisis in a historical evolutionary context. Although forestry is part of the broader set of agricultural activities-farming, animal husbandry, and so on-that are necessary for human sustenance, there is a fundamental difference in the ways people satisfy their needs for food crops, on the one hand, and for fuelwood, on the other hand. This difference needs to be spelled out explicitly. In its historical evolu- tion, human society long ago emerged from its hunting-and-gathering phase 3 A similar issue came up some three or four years ago, when a "livestock" policy paper was being drafted by a development agency. Yet draft after draft failed to articulate a meaningful course of action for investment lending, of course not only because of this narrow commodity angle but, in my view, largely because of it. A development policy cannot be designed around livestock-cattle, or pigs, or goats, or camels-or any commodity as such. What was and is needed is a policy about investing in the human activities for animal husbandry, and obviously such a policy should focus on the people who do animal husbandry. 4 An interesting discussion of the terminology used in international development practice, particularly in financially induced development activity, is contained in Bare (1991). 304 MANAGING THE WORLD'S FORESTS in virtually all but one domain: the procurement of fuelwood. Initially, humans relied on mere gathering to obtain their vegetable food and on hunting wildlife to get their meat; however, they learned long ago to do- mesticate both plants and animals. Human society shifted from gathering to cultivating for producing its food crops, and shifted from hunting to animal husbandry for producing its meat. Humankind would have disappeared long ago had it not learned to do crop farming and animal husbandry on a regular and gigantic scale, thus meeting its continuously growing food needs. In contrast, gathering, rather than systematic cultivation, has remained an important source of fuelwood. Although many people have substituted other energy sources in heating and cooking, a large proportion-according to some statistics, the majority of the world's population-continues to use wood as fuel. And for the bulk of their fuelwood needs, people continue to rely on Mother Nature-that is, on the spontaneous regeneration of trees, cutting naturally grown trees without systematically planting trees for fuelwood commensurate with their needs.5 A case in point is India, a country with some 800 million people. Only some 10 percent of India's farmers cultivate trees for fuel wood, but all need fuel and most consume fuelwood. The same is true for many other countries. What had been sufficient for millennia ceased to be sufficient in the 20th century. The increasingly intense gathering of trees for fuel and construction, combined with other causes of forest depletion-clearing of forests for agriculture and ranching, industry-driven forest mining, acid rain-have shrunk the globe's tree-covered areas and considerably reduced the wood volume produced by spontaneous natural regeneration (Guggenheim and Spears 1991; Goodland 1991). In Africa, more than 3 mil- lion hectares of tropical forest are lost each year, and in Asia forests are disappearing at a rate of at least 5 million hectares per year. Recent remote- sensing data and ground surveys indicate that a total of about 17 million hectares of forest are lost worldwide every year (World Bank 1991). Awareness is slowly emerging that without massive new tree planting, the current rate of use of forest resources will disastrously accelerate defor- estation and will lead to a worldwide fuelwood scarcity, with far-reaching socioeconomic and environmental consequences. Increasing supply through systematic production of trees in addition to natural regeneration is crucial. Because most wood users in developing countries live in rural areas, the best way to satisfy their demand for wood is to mobilize their own labor and land, as well as public lands, for tree planting and better management of forest and nonforest trees. Thus the global human society must turn from basically gathering fuelwood to basically cultivating and producing trees. The users must become producers. This will be a social transition of historic magnitude. 5 To a certain extent, fishing in the world's oceans, seas, and rivers can be simi- larly seen as gathering, if compared with fish-farming and various forms of aqua- culture. A SOCIOLOGICAL FRAMEWORK 305 Financially induced development programs can help accelerated this historic transition. Of course, such a transition will take generations, but it has started. Users' Participation in Reforestation That such a transition from gathering to cultivating is necessary and, in fact, intrinsic to mankind's evolution is demonstrated also by one subcat- egory of trees that is the major exception to the pattern of gathering: the fruit trees. People learned long ago to domesticate, cultivate, and propagate fruit trees, and fruit trees are now part and parcel of routine agricultural strategies. What was possible for fruit trees-integration into regular pro- duction systems-heralds the trend and future for trees for fuelwood and other uses. In a historic transition of this magnitude, the donor community through its policy and investment lending is only one of ,very many contributing factors. It definitely is not the only or the decisive factor. Yet the donor community's policies and strategic role can make a far-reaching contribu- tion. Defining a policy for tree planting and forest management within this broad historical vision would position its investment programs on the main task of the natural historic process. The need for purposive intervention to accelerate this evolution is made even more urgent by our planet's demographic processes: high birthrates will add 1 billion people to the world's population by the start of the 21st century. This will greatly increase the twin pressures on existing forests, namely, the demand for farming lands and the need for fuelwood and tim- ber. Under such demographic circumstances, improved conservation alone will be insufficient. Only organized increases of supply hold hope for keep- ing pace with the growing social needs. The sociological perspective just outlined leads to at least three conclu- sions: 1. Future investment policies by governments or donors alike must pursue a major change in the agricultural practices of the world's farmers, namely, the incorporation of planting trees as an integral part of regular farming systems. 2. The new planting should be complemented by investment in sound conservation and better management of existing forests, rather than in large-scale logging projects. 3. Accomplishing both these tasks requires identifying and em- powering the social actors who stand to gain from planting trees, on the one hand, and restraining the social agents and vested interests responsible for forest depletion, on the other hand. 306 MANAGING THE WORLD'S FORESTS Out-of-Forest Trees: Regreening "Regreening" is a relatively new term used to distinguish activities aim- ing at planting more out-of-forest trees (on private farms and common and public lands) from standard reforestation activities carried out in or around forests. Regreening is an integral part of the vast effort of reforestation. The current environmental debate about deforestation focuses primarily on tropi- cal humid and dry forests, which are the world's most important forests. But deforestation is not limited to tropical forest depletion; consequently, reforestation should not be confined to forest lands. Much of what follows will discuss regreening activities that aim at planting trees on out-of-forest lands. Conventional government agencies dealiLng with forests tend to give very low priority to out-of-forest trees: they concentrate on classified for- ests. But to meet the world's need for fuelwood, much more attention must be given to out-of-forest trees. For farmers, planting trees out of the forest has enormous potential. Such trees may be planted with relative ease and acceptable survival rates. In many countries, lands under common or public property regimes have a significant tree cover. The scattered trees, bushes, and shrubs, growing on them cannot be classified as "forests," but they belong to the category of out-of-forest trees. Wastelands, shrublands, and abandoned farmlands as a habitat for out-of-forest trees represent key potential resources to be used in social forestry strategies. SOCIAL FORESTRY: A PEOPLE-ORIENTED STRATEGY Market Failure and the Social Act•ors of Deforestation As we will strive further in this chapter l:o identify the potential actors of new tree planting strategies, it is first necessary to define briefly the processes and social actors responsible for deforestation. The types of agents responsible for forest degradation-loggers, ranch- ers, industrial corporations, individual farmers, migrants, farming commu- nities, or government agencies-and the causes of their behavior vary widely across regions and forest types. That behavioir is often in contradiction with the interests of the overall society. But these deforestation agents too must be identified, in each specific country, with accuracy; otherwise their actions cannot be countered adequately. It is a mistake to assume that the same one or another actor is responsible at all times for forest depletion. In different historical stages there are different deforestaltion actors. The sociohistorical analysis of the forest sector in the Philippines (Sajise 1991), for instance, has demonstrated that deforestation has been caused by a succession of social forces acting alternately or concurrently: before 1900, deforestation was caused primarily by agricultural settlements; after 1900, first by logging, then by A SOCIOLOGICAL FRAMEWORK 307 waves of spontaneous migrants from the lowlands, then by government- sponsored settlements, by increased numbers of shifting cultivators, by com- mercial concessions, and the like. It is necessary to pinpoint, with both a sociological and an economic perspective, why these different social actors act-either occasionally or for extended periods-in socially detrimental ways, · and what can be done to slow down or arrest their action. As long as the excessive belief in market omnipotence alone informed government policies or forestry aid programs, it was hardly possible to understand that markets can, in turn, act one-sidedly and that the failures of the market may increase the risks for unanticipated deforestation. Markets place economic value on timber and other wood products. But markets are not able to place economic value on such social intangibles as forests' envi- ronmental protective functions, or on their role as the ha~itat of indigenous forest-dwellers. As a consequence, markets alone cannot economically com- pel the agents of deforestation to pay the costs of such broader societal services. Instead, the market's failure in these respects allows private inter- ests to transfer and impose the economic costs of forest depletion (and of its adverse environmental effects) on the society in its entirety. The agents of deforestation can therefore derive unchecked direct benefits, while others bear the costs. · In turn, by minimizing the costs born by certain agents of deforestation, the markets' failure creates a chain of disincentives for reforestation: it low- ers the benefits that rural people can derive from planting trees and pursu- ing appropriate forest management practices. Market failures can also be compounded by distorted government poli- cies. Binswanger (1989) has demonstrated that in Brazil, for instance, a com- bination of misguided tax policies, inadequate rules of land allocation for settlements, perverse economic subsidies, incentives, and credit systems has clearly accelerated deforestation in the Amazon region. 6 Such distorting policy provisions must be removed before afforestation projects can succeed. Other large-scale social and demographic processes, some of which have already been mentioned, may further aggravate the effects of market failure on deforestation processes worldwide. To correct such market distortions and failures, two main approaches are available: (I) sound public-sector policies and financial interventions, and (2) mobilization of the capacity of rural producers to intensify tree 6 Such distorted policies or incentives do not occur only in developing countries, but are present in industrialized countries as well, leading to similar forest deple- tion effects. In the United States, for instance, the Tongass National Forest in Alaska, which is North America's largest temperate rain forest, is a sad case in point. Be- cause of price subsidies offered to timber companies and pulp mills, the U.S. Forest Service has been supporting unjustified logging rates in a major rain forest. As an editorial article in the New York Times put it, 500-year-old trees are being sold to pulp mills for $2 per 1,000 feet or "about the price of a cheeseburger" (New York Times 1989). 308 MANAGING THE WORLD'S FORESTS planting. To slow deforestation, development-oriented government policies must address a host of factors, from protection of critical natural resources, clarification of legal property rights, elimination of perverse incentives and adjustments in policies outside the forest sector to financial allocations and trade reforms. The core of any such effort, however, will remain a policy focused on intensified, large-scale tree planting, with social forestry as one of its key strategies. Relying on a large spectrum of potential social actors and tailoring socially and technically appropriate tree-planting programs to their needs are imperative in most developing countries. This subject is discussed in the next subsection. The Concept of Social Forestry From their inception, social forestry programs were aimed primarily at helping small farmers and the landless to meet their consumption and in- come needs. 7 The strategic objectives embedded in the very concept of social forestry are as follows: • To encourage large numbers of people to plant trees; • To promote the kind of tree growing that will best supply fuelwood, small timber, grasses, to the small producers themselves; • To increase the income and benefits to poor people from tree grow- ing and forest products; and • To improve environmental protection. In contrast to industrial forestry, the programs that fit the definition of social forestry attempt to influence a key sociocultural variable: people's behavior toward trees. Under conventional industrial forestry programs, business corporations or government agencies hire laborers to establish plan- tations on large tracts of land controlled by private business or agencies; the wood is harvested for use in industry or construction. Social forestry, in contrast, aims to induce a large number of small farmers to plant fuelwood trees systematically for their own needs and on their own (and other avail- able) lands. These programs often attempt to promote collective action for tree planting and protection, institutional development, and establishment of enduring social structures and value systems that activate and energize grass-root actors. By definition, social forestry programs :require massive farmer parti- cipation. Their rise or fall depends on whether the farmers engage in the program. Financial investments alone, however big, cannot make such social 7 When social forestry was first proposed as a concept and policy in India in 1973, its goals were defined as follows: (1) to supply fuelwood to the rural areas and to replace cow dung; (2) to supply small timber; (3) to supply grasses and fodder and provide for grazing; (4) to protect agricultural fields against wind; and (5) to meet recreational needs (see Government of India 1973, 12). A SOCIOLOGICAL FRAMEWORK 309 forestry programs a success. Several social prerequisites must be met for them to succeed. Processes such as reforestation, environmental protection, and watershed rehabilitation depend not only on activities carried out indi- vidually and discretely, but also on collective or coordinated actions. But how can coordinated action be stimulated? Collective actions are most likely to occur and be effective (1) when people belong to organized groups, (2) when they are informed and become conscious that it is in their best interest to act in a coordinated manner, and (3) when the group has developed leadership structures and internal norms and procedures likely to mobilize and manage its members and to overcome conflicts and deviant behavior. The objective situation of many people as users of a certain resource is a propitious circumstance that can help motivate and organize them for producing that needed resource. The purposive construction of user groups is therefore particularly important for husbanding a common pool resource in programs such as afforestation or irrigation, which depend on sustained, long-term consensual action of many individual actors. Designing Strategies Around Social Actors Central among the social prerequisites of any innovative program is a unit of social organization capable of sustaining that program. Therefore, from a sociological viewpoint, afforestation projects must: • start with identifying (or establishing) such a viable unit or group; • seek to engage the rural users of fuelwood in patterns of coordinated action for producing the fuelwood they need; • ensure a match between the silvicultural technologies the projects promote and the social groups they address; and • deal with the issues of social engineering (group formation, leader- ship, participation in decision making, intragroup structures, incen- tives, penalties, communication, benefit distribution) with the same scrupulous attention that the technical or financial elements of the strategy receive. Forming groups is vital for the success of development programs in- volving (even to a small extent) natural resources that are under a common property regime or that lend themselves to group use and management even if they are under a state property regime. To ensure both the short- term use and the long-term renewal and sustainability of a jointly owned natural resource, the owners must act in consensus and subject themselves to the same norms, rights, and restrictions. To act as a group, people need to be a structured social group, not just a set of unlinked individuals. Intragroup connections are forms of influence, help, and mutual control. The absence of structures and strictures leaves the way open to unchecked and counterproductive individual free-riding behavior. 310 MANAGING THE WORLD'S FORESTS Unfortunately, the planners of financially induced social forestry pro- grams often do not yet realize that consideration of these social factors has to be woven into the fabric of such programs from the outset. There is often a contradiction between the theory and the practice of social forestry and, as Fortmann (1988) writes, "many projects that are called social forestry are a far cry from the theoretical vision of social forestry." The penalty for ignor- ing the social factors is project failure. 6ecause of the enormous diversity of country and local conditions, it is important not to search for universal recipes about how to define and select social actors or create units of social organization. Sociologically informed forestation strategies have to be tailored and retailored anew for various countries or socioecological contexts (Guggenheim and Spears 1991). To combine the technical and social approaches into coherent reforestation pro- grams, foresters, planners, sociologists, and anthropologists have to cooper- ate, search, design, test, monitor, learn, and redesign such approaches again and again. Entrusting a social forestry program (and development programs in general) to the wrong social actor will lead to the failure of that program, as in fact has happened repeatedly. The loosely defined concept of community forestry used by some national or international development agencies in the recent past has reflected just such a vague or mistaken definition of the social actors. Some statements or articles are repeating the term community forestry from title to end hundreds of times, as a mantra, without once bothering to discuss what specific social groups, strata, or classes compose this mythical "community" and what accounts for their differential behav- ior vis-a-vis tree planting. In social forestry, it is necessary to disaggregate the broad term people and identify precisely which units of social organiza- tion can do afforestation, and which social units and definable groups can act as sustaining and enduring social struchtres for long-term production activities. Such units of social organization, or social actors, can be (I) natural (existing) social units, such as the individual family household or a tightly knit kinship group or subgroup; (2) groups organized purposively to plant, protect, and cultivate trees; or (3) groups (or organizations) that were estab- lished for purposes other than forestry but are able to undertake forestry- related activities as well. Examples of units of social organizations for each of these three catego- ries are discussed later. Construction of Groups for Resc1urce Mobilization Like any development activity, reforestation requires both public and private investments. Opening up social foresbry strategies to many potential social actors-including discrete families and multiple-group structures-is a way to amplify investments and resources for afforestation. Tree planting A SOCIOLOGICAL FRAMEWORK 311 can be expanded only if farmers' economic incentives and cultivation skills, as well as their organizational capacity for collective action, are enhanced. The establishment of groups as action units opens up opportunities to mobilize resources of land and labor that would not be harnessed other- wise. There are, for instance, significant land areas under state control for which the public sector may not have the investment resources required for tree planting. Leasing such lands to organized groups ready to invest their labor in planting and protecting trees puts those lands to use without the risk of fragmentation or alienation and with relatively low transaction costs. In sum, group-based strategies, combined with some public-sector support, can make better use of available but dispersed resources. In other situations, lands that in principle are common property are often not managed as a group-owned natural resource because the group lacks cohesion, linkages, and authority systems. Such lands slide toward open-access abuse; returns diminish and ecological destruction increases (Bromley and Cernea 1989). The reestablishment or strengthening of the structure and functionality of the group recovers the common property re- source for the group's own benefit. When groups act as economic agents, they can achieve for their mem- bers significant economies of scale in the labor required for tree planting, cultivating, protecting, harvesting, and transporting. Furthermore, when selling the harvest or negotiating with authorities, groups usually can bar- gain more effectively than individuals. Relying on enduring units of social organization as development actors is particularly important in view of the long duration of a tree-production cycle. Even small groups enhance the productive capacity of their indi- vidual members: they maximize the cumulated impact of the contributions of individuals and enable them to perform activities and achieve goals that might not be attained by each one acting separately. Some technological needs may be more easily solved by groups than by individuals (or separate families) alone. For instance, watching and protect- ing tree plantations for a long time and over large areas against theft, fire, or destruction by animals can usually be done more effectively by produc- ers associated in groups than by individual families. Groups can also be powerful psychological motivators for the consensual action of their indi- vidual members. The need to capture the synergistic potential of such social units intro- duces social engineering demands on the activities of forestry departments. Forestry departments are not accustomed to social work and need to be reoriented in light of these social engineering demands. There are also, of course, many hurdles on the road to forming groups other than the lack of social skills among the staff of forestry departments. Political adversities in particular are both numerous and hard to overcome. The establishment of organized groups of small farmers is often perceived as a threat by either the privileged village elites or the state and its agents. Evidence indicates 312 MANAGING THE WORLD'S FORESTS that nongovernment organizations concerned with poverty alleviation and environmental conservation can be instrumental in helping users organize themselves for tree planting (Cernea 1988a). If properly conceived, social forestry projects can become a mechanism for encouraging and forrning groups, thus building up the social capacity for development. Establishing a functional social group means, of course, much more than simply lumping individuals into an artificial entity exist- ing on paper alone. The process involves selection or self-selection of the members, a willingness to associate, the members' perception of self-advan- tage and co-responsibility, and the establishment of an enduring structure with well-defined functions. This process, in turn, helps mold patterned behavior among members and is the essence of grass-root, purposive insti- tution building. Helping users of fuelwood to organize themselves into groups and to undertake production and management functions in forestry would, in fact, restore the "participation equation" to normality: the users of forests and forest products would act as the primary producers and decision makers, and the forestry departments would then participate in the tree growers' activities, rather than the other way around. When the actor is a group of farmers, rather than an individual farm family /household, social forestry programs must address issues of joint dependence on a piece of land and, sometimes, group tenure over trees; issues of group management, labor allocation, and monitoring; and, prob- ably the most sensitive, the issue of benefit distribution. Therefore, organiz- ing and promoting groups as units of social organization for social forestry programs (where groups are pivotal actors or economic agents of such pro- grams) means more than just bringing several individual farmers to one or more meetings: it means designing clear social arrangements for tenure, management, and distribution-arrangements that are known, implemented, and adhered to consensually by the group. The Fit Between Technology and the Social Actor The social arrangements required for group forestry may need to vary with the technologies envisaged for reforestation in different ecological areas. The appropriateness of tree husbandry technologies is not neutral with respect to social structures. The technical and physical characteristics of a forestry program and the social characteristics of its actor should be compatible. When forestry programs are designed, it is essential to realize that the various potential social actors are not equally fit for carrying out all the technical (silvicultural) approaches to forestry, such as site selection, nurs- ery development, species selection, planting technology and configurations, fertilization, plantation management, enclosure or other protection, and marketing. For instance, to determine which of three types of tree arrange- A SOCIOLOGICAL FRAMEWORK 313 ments-block planting, linear planting, or alley cropping-is best in a par- ticular case requires identifying the needs of the farmers themselves and assessing the local land-tenure systems and land availability. The same refers to the choice of tree species. In addition to their bio- physical attributes, trees have socioeconomic attributes "wholly ascribed to trees by people" (Raintree 1991), and the same tree may have different uses and meanings to different people in different cultures. Farmers are inter- ested in planting multipurpose tree species, rather than planting trees just for fuelwood alone. Therefore, recommending tree species adequate to the various needs of specific populations is an important technical and cultural decision, to be made in a participatory manner. Calibrating the overall fit between the technical/biological characteristics of trees and the social actors around whom the afforestation strategy can be built is at the heart of the cooperation among tree growers, foresters, and sociologists. Tenure Rights, Incentives, and Awareness Designing strategies around specific social actors, and constructing or strengthening groups, requires at least two more key elements: (1) tangible economic incentives and benefits to the envisaged social actors and (2) awareness of the need for afforestation. An extraordinarily important incentive is clarifying the land-tenure sys- tems on forested and savanna lands and protecting the land rights of the tree producers. Similarly, tenure on trees must be clarified and secured legally. Customary land-tenure rules often discourage tree planting by ten- ants, because planting and owning trees traditionally entail title to land. Conversely, modern regulations in some developing countries have intro- duced disincentives in other ways-for instance, by limiting farmers' rights to cutting and harvesting tre,es that they planted on their own land. Recog- nizing farmers' tenure on trees and their decision-making rights to harvest, use, or sell their trees whenever they wished would increase the incentives to tree planting. The public sector can introduce additional economic incentives (e.g., contributions of free or subsidized seedlings, technical advice, tax mecha- nisms, policies supporting markets for fuelwood). Fuelwood producers will be even more price-responsive when producing fuelwood becomes a source of income. Because the time lag between planting and harvesting trees is long and tree growers can rarely afford to wait several years for income, particularly where land is scarce, substitute income sources may be tempo- rarily needed to foster behavior change. Altogether, the effectiveness of economic incentives can hardly be overstated. Not only economic but also cultural and symbolic incentives percep- tible to the farm family should be provided, based on an understanding of the local culture and value system. The symbolism of tree planting can be linked to events in the life of the family and the village that are imbued with 314 MANAGING THE WORLD'S FORESTS positive values-from childbirth, or marriage, to receipt of title to land. Such linkage-tree planting to celebrate family events-is being encour- aged, for instance, in Japan, even though it is a country not at all threatened by deforestation (Umebayashi 1991). The articulation of various types of incentives including cultural and symbolic ones (rather than just throwing money at problems), is an important component of responsible social engi- neering, and sociologists must help design multiple-incentive systems. In turn, increased awareness of the need for afforestation may reduce the time lag with which wood shortages are usually perceived and spur individuals to collective action to satisfy their common needs. Better com- munication, extension, and education also can open farmers' eyes toward the more subtle benefits of tree growing, like retaining soil moisture, de- creasing wind effects and soil erosion, or building up savings over time in the form of valuable trees that can be drawn on in times of sudden need (Chambers, Saxena, and Shah 1989). Moreover, promoting broad awareness of the role of trees in averting soil degradation, land slides, or flash floods may help reduce societal demand for products that require cutting trees and spur the search for wood substitutes. There are many potential social actors for forestry projects, but not all are equally adequate: communities, forest villages, local governing bodies, farm families, groups of farmers, cooperatives, private companies, loggers, public agencies, nongovernment organizations, temples, schools. The strengths and weaknesses of some of these actors are analyzed in the sec- tions that follow. Ill-Defined Social Actors and the Failure of Village Woodlots Until recently, the village woodlot was widely promoted as the desir- able model in social forestry, but results have fallen well short of expecta- tions. Many planners and foresters assumed that massive planting of fuelwood could best be induced on communal lands by simply asking people to plant trees. The apparently plausible social assumptions were that com- munities as groups would influence their members to plant on the com- mons and would collectively protect the young plantations on "their" land. The term community forestry became a buzzword. Unfortunately, very few bothered to define the social actor of the woodlot or the composition of the village community. Successful village woodlots in countries such as Korea and China, which were organized authoritatively by those governments, were assumed to be valid models for other social contexts. However, the results of replicating community woodlots in Uttar Pradesh, Karnataka, Gujarat, and other In- dian states, in Niger and other African countries, and elsewhere have been disappointing. Evidence about community woodlots increasingly documents that, when scrutinized, they are not what their name suggests them to be, namely, genuine community undertakings, and therefore do not achieve their stated objectives. A SOCIOLOGICAL FRAMEWORK 315 Over the past 10 to 12 years considerable financial resources have been channeled by both international donor agencies and national governments in many developing countries to social forestry programs that have used the community woodlot model. Between 1977 and 1986 about half of the World Bank's lending for forestry went to 27 projects that included some form of community forestry. Furthermore, during 1987 and 1989 the Bank's lending for social forestry tripled compared with its lending for the full prior de- cade. Major resources came also from bilateral donors like the U.S. Agency for International Development (USAID), Canadian International Develop- ment Agency (CIDA), and Swedish International Development Agency (SIDA). Yet most evaluation reports reveal that the actual plantings accom- plished under the community model fall below targets and do not justify the investments. Structural Limitations of Community Performance The initial assumption~that communities (villages) would be effective agents for implementing community forestry-was not confirmed. This assumption was sociologically naive and exhibited a lack of understanding of the structure and social stratification of village communities. For instance, none of the three large social forestry projects assisted by the World Bank in India-in Uttar Pradesh, Gujarat, and West Bengal (World Bank 1979, 1980, 1981)-came close to its targets for establishing community woodlots.' In Uttar Pradesh the project aimed to establish 3,080 hectares of village woodlots but achieved only a total of 136 hectares (each woodlot averaged two hectares). In Gujarat the self-help village woodlots compo- nent achieved only two-thirds of the 9,200 hectares targeted. Summarizing the causes of such failures, a World Bank report noted: Poor villagers in Uttar Pradesh proved unwilling to contribute their labor as expected by the project in exchange for rather lim- ited potential benefits from a small woodlot, after many years of protection and maintenance .... The social forestry organization lacked relevant know-how and resources to deal with the socio- logical and technical problems associated with densely cultivated areas and very small farms (World Bank 1985). Many of the newly established village woodlots are beset with social, management, and distributional problems that prevent them from accom- plishing their objective to enhance community fuel supply and to alleviate poverty. A sociological field analysis in India concluded that no user-created management system for the protection and maintenance of community woodlots has emerged so far (Salam 1989). Communities as a whole are not 8 However, these projects were effective in other approaches and-to some plan- ners' surprise-even surpassed their targets in farm forestry. 316 MANAGING THE WORLD'S FORESTS getting involved; instead, the village panchayat-or the state forestry de- partment-takes over the administration of the woodlot, often sells the prod- ucts outside the village, and invests the revenue in other assets (World Bank 1988). Among the subsistence farmers, disappointment with the distribution of benefits from these woodlots saps future interest in maintaining or ex- panding them. An evaluation of Orissa Social Forestry project by Arnold and Stewart (1989) found that 82 percent of the villagers did not know how the produce from village woodlots would be distributed; most of the people did not expect any share from the final output and looked on such woodlots as another category of reserved forests. It is therefore not surprising that in such social forestry programs "on village commons and wastelands ... villages have proved most reluctant to manage trees planted as a corporate resource" (Shepherd 1986). Farmers' response to communal woodlots was found to be "ambivalent or negative" in Tanzania as well (Shanks 1990); in West Africa the commu- nity system was evaluated as "ill-suited ... to serve as a vehicle for refor- estation" (Thomson 1980). The system's adequacy was questioned also in most Asian countries. Often forestry departments were asked to set up the village woodlots and then to hand them over to a village committee. This practice deprives the community woodlot of all or most of its social forestry content and belies its social label. Moreover, village committees were often unaware of what they should do with the woodlots. For instance, referring to many government projects in India (not financed by the World Bank), Sen and Das (1987) conclude: One of the most vital problems being faced by the community forestry program is lack of people's participation. The very mecha- nism of raising, maintaining and protecting the community plan- tation ... should be examined carefully.... Villagers are rarely consulted at the preplanting stage ... and selection of site and species is generally done by the local forest officials. The village panchayat or similar agencies offer the land (often with no or half information to their members) for plantation activities by the for- est departments. Similarly, Arnold and Stewart (1989) have synthesized the findings of numerous evaluations of woodlot projects on communal lands in India during the 1980s, and insisted on the missing social arrangements: The communal groups charged with the dialogue with forest de- partments over the planning of woodlots and with their eventual takeover have nearly everywhere been panchayats ... rather than a user group or a body selected by a village specifically for man- aging the woodlot. ... Mechanisms for direct consultation by the forest department with villagers have generally not been put in practice. . . . Benefit sharing agreements are frequently neither finalized nor formalized .... Most of the people did not expect any share from the final output. A SOCIOLOGICAL FRAMEWORK 317 From Thailand, Yaowalak Apichatvullop (1991) reports in the same vein: Many socially-based forestry projects in the Northeast failed to gain adequate participation from local people because local people do not perceive benefits from those projects or they do not believe that the benefits will be distributed locally. Such feeling may be caused by the existing forest laws and regulations .... People are reluctant to participate in a community plantation as the law de- termines that trees in the state forest cannot be cut for personal use. These and many other findings support the conclusion that community woodlot projects initiated during the past dozen years amounted to an ex- traordinary case of national and international programs that pursued envi- ronmentally desirable goals and were intended to be genuinely participa- tory, but nevertheless were launched and financed by agencies that lacked a sound understanding of the social process and system they had to put in motion. Investment in the technical process far outpaced and outweighed the investment in the human and institutional process. The latter was not recognized as the unavoidable learning curve that it must be. Knowledge about how to invest in the institutional components of social forestry was lacking. The technical act of planting trees was generously financed before the cultural construction of new norms and institutions to support such treelots was formulated theoretically or tested experimentally. Although the institutional arrangements should have been the stepping-stones to the suc- cess of the technical process, the financial outlays were rushed into the latter without a prayer for the former. The absence of the basic sociological knowledge to guide social forestry interventions is far more consequential than the routine bureaucratic hin- drances that always appear during the implementation of induced develop- ment programs. Bureaucratic hindrances aside, the weaknesses during project execution were not the primary reason why community woodlots were in- effective. Community woodlot schemes were misconceived and could not be effective because they were inspired by the romantic myth of homoge- neous villages, without new and appropriate social arrangements being put in motion. Eric Wolf (1966) and Louis Dumont (1980), among many others, have demonstrated that corporate peasant villages are disappearing in the contemporary world and that the very term village communities is not ad- equate for India because it conceals the existence of factions and the omni- presence of hierarchies. In light of the empirical results examined earlier, there are at least seven basic sociological reasons why-given their internal social stratification and structure-whole communities are not ready-to-use corporate units of social organization for afforestation programs: 1. Communities and villages are geographic residential popula- tion clusters, but not necessarily corporate organizations. Physical proximity alone is not sufficient to engender the type of long- term collective action required for a woodlot enterprise. 318 MANAGING THE WORLD'S FORESTS 2. Community subgroups often have widely different interests that preclude the kind of collective unified action required by long-term afforestation programs. Most communities are hetero- geneous population clusters, stratified and split into factions and subgroups with fragmented socioeconomic interests. What is ad- vantageous for one subgroup is not necessarily advantageous for another. 3. Community land is so limited that villagers often are reluctant to make it available for tree planting. Tree block sites are small, costs are high. Research has demonstrated that poor households are much more dependent on products from the commons than are better-off households (Jodha 1986), Thus the poorest house- holds have a vested interest in not allowing the commons, which to them are a continuous if meager source of products, to become a closed, inaccessible woodlot. 4. The tenure status of the common lands is often uncertain and engenders insecurity about the tenure on trees. The social body that has jurisdiction over the allocation of common lands is simi- larly unclear.' 5. Authority systems have uneven mobilization power over com- munity subgroups. Local community leaders often appear reluc- tant, or too weak, to mobilize the individuals belonging to differ- ent subgroups to work for establishing woodlots, or to enforce restrictions for tree protection. 6. Distributional arrangements for benefit sharing to ensure that the woodlot products reach the recipients entitled to them are seldom specified at the outset and have not worked in practice. The lack of intragroup rules and guarantees for distribution of benefits commensurate with contributions of labor are lacking, and this problem alone is sufficient to doom the wholesale com- munity approach. Exclusionary rules against noncontributors are missing, too. The length of the three production cycle reinforces doubts in those planting today that they will actually receive wood eight or more years later, and feeds the lingering suspicion that the authorities will appropriate the wood anyway. 7. Last but not least, most communities are not organized as joint producers in other respects and thus do not offer a matrix on which additional activities can be grafted. Externally designed 9 Michael Horowitz (1982), analyzing rural afforestation alternatives in Zimba- bwe, pointed out that "the important issue where communal lands are involved is correctly identifying the locus of authority over land use allocation." A SOCIOLOGICAL FRAMEWORK 319 programs that do not bother to establish grass-roots organizations cannot foster by decree the kind of close interdependence of mem- bers that community-based schemes would require. Because such sociological characteristics tend to be widespread, results are likely to be poor in the future as well, whenever such corporate woodlots would be expected to be sustained by noncorporate communities. Those positive ,results with village woodlots that have been achieved tend to be exceptions linked to particular circumstances in one community or another. 10 When successes occur it is important to identify the specific structural, cul- tural, or political conditions that make them possible or replicable. Alternative Social Actors What alternative vehicles should social forestry adopt in light of the failure of community woodlot approaches? Clearly, future social forestry programs should be built on more care- fully elaborated social and institutional arrangements. The social actors for such programs need to be more precisely defined. Better social engineering should be used to get the poor and the landless, including the women, appropriately involved, while the size of groups that are to be organized should be compatible with the actors' self-management capacity, mastered technology, and available labor and land resources. Two specific questions must be answered: 1. If the community as an entity is usually not a homogeneous collective actor in afforestation, are there other units of social or- ganization that are able to assume and execute such a role? 2. Are there tenurial innovations (in either land tenure or tree tenure, or in the granting of usufruct rights) that can be promoted to mobilize and facilitate performance by such alternative units of social action? To both questions, the answer is positive. There are social groups within the stratified, nonhomogeneous communities, namely, specific subsections of such communities, that can act collectively. Moreover, such groups can be purposively organized. Furthermore, making more refined distinctions between forms of land tenure and various systems of tree tenure, as Fortmann 10 For instance, Mathew S. Ghamser (1987) reported on an interesting community forestry project in Sudan (Um Inderaba) where the village community (some 600 families) effectively in planted, hand-watered, and maintained the trees against a complete lack of rain and large transient animal herds. It appears that the village committee and the local sheikh were able to coordinate the villagers' activities effec- tively while incentives and protection payments were provided, and foresters pro- vided technical advice. 320 MANAGING THE WORLD'S FORESTS proposes (1988b), would permit more imaginative combinations and inno- vations in forming action groups. The challenge is to identify the population subsections able to implement and sustain such innovative approaches. Two fundamental strategies are available and likely to be more success- ful than the previous community woodlot approach: family-centered strat- egies and group-centered strategies, based on groups larger than the farm family. Public investments in social forestry should be made through both, thus enhancing and reinforcing the investments made by the private pro- ducers themselves through labor, land, and inputs. The next two sections of this chapter explore the characteristics of the actors of each of these two basic strategies. THE FAMILY UNIT AS ACTOR IN :SOCIAL FORESTRY Family Agroforestry The accumulating evidence of the ineffectiveness of the community- centered approach mandates a shift in thinking. Foresters and planners must focus on the individual family farm unit as an alternative to the community- based programs in social forestry. The family-centered approach goes by various names, such as farm forestry, family woodlots, and agroforestry. The common denominator in this semantic diversity is that the family household becomes the social unit around which reforestation is planned and financed. The technical approach to family farm programs also differs from the one proposed for community woodlots: it is designed to suit the labor and land opportunities available to the individual family farm. Of course, this is not to say either that all interest in promoting village woodlots has now disappeared, or that promoting tree planting on indi- vidual farms is a totally new orientation. But there is a perceptible shift in emphasis, and a refinement and diversification of social forestry strategies. This shift implies a change in the sociological underpinnings of certain for- estry programs. Various World Bank-assisted forestry projects-in Karnataka, Kerala, Haryana, and other Indian states, as well as in Mali, Tanzania, Nigeria, Nepal, Haiti, and elsewhere-now provide support and incentives for tree planting on small family farms. In the design of India's Jammu and Kashmir and Haryana social forestry project, village woodlots represent only 11.3 percent of the total planting program, while :farm forestry, supported by a distribution of about 47 million seedlings free to individual farmers, repre- sents about 43 percent (World Bank 1982). A similar approach was taken in an ongoing Kerala Project (World Bank 1984a). Some of the most spectacu- lar results in family forestry are being obtained in Gujarat and Himachal Pradesh, where farmers have responded to project-provided incentives (free A SOCIOLOGICAL FRAMEWORK 321 seedlings, etc.) and technical assistance. During the first three seasons of the National Social Forestry Project in India (1985-88) farm families planted on their private lands approximately 500 million seedlings (the equivalent of over 325,000 hectares in block planting), exceeding the already high target by some 18 percent (World Bank 1988). The vast potential of the family farms to incorporate tree planting was dramatically demonstrated in Haiti by the Agroforestry Outreach Project (AOP) funded between 1981 and 1985 by USAID. Guided by prior ethno- graphic knowledge of Haitian tenure systems and cropping patterns, the AOP stands out as one of the social forestry projects that had a clearly conceived sociological strategy, purposively designed around the family farm as its central social actor and accompanied by a technically appropri- ate reforestation package. The project started with the farmers' needs, val- ues, and actual behavior. To reduce the opportunity cost of lands, the project proposed that the family farm (the average small holding in Haiti is 1.5 hectares) plant 500 trees of fast-growing fuelwood and pole-producing species in intercropping and border planting. Light-weight seedlings and technical assistance were provided free to the peasants. Most important, the project guaranteed that "the peasants themselves, and not the government or the project, would be the sole owners of the trees and that the peasants would have unlimited rights to the harvest of the wood whenever they wished" (Murray 1987). Social anthropologists directed the implementation of this social forestry project, in close cooperation with technicians. Nongovernment organizations were involved in carrying the project mes- sages to the farmers. The results were spectacular: whereas the four-year target was to plant 3 million trees on the land of 6,000 peasants, 20 million trees were planted on 75,000 family farms (Murray 1987). Sociological and Technical Variables Sociologically, the advantages of a strategy centered on the family are manifold. Land tenure and tree tenure are much less ambiguous than in community forestry, responsibility and management authority are vested in real persons, and divisive issues of intragroup benefit distribution are elimi- nated. Moreover, the transaction costs entailed by the work for establishing groups are avoided. Technically, on family-owned lands trees can be grown not necessarily in blocks (family woodlots) but also along linear landscape features such as farm boundaries, internal field borders, roads, and watercourses. Under conditions of wood scarcity, the economics of family farming favors ex- panding tree planting. Yet smallholders tend to weigh the opportunity costs of labor and land more than planners and foresters usually realize, as the latter often hold erroneous beliefs about farmers' thinking (Dove 1991). Tree- planting technologies that maximize the use of interstitial locations and 322 MANAGING THE WORLD'S FORESTS other marginal land patches are particularly suitable for individual small farmers because they do not compete with existing land uses and other crops. Even small farms that cannot afford to set aside an arable plot for a tree block can use their hedgerows for planting. Individual trees scattered around the family farm's cultivated areas generally grow faster than plantation trees that compete with one another in dense woodlots (e.g., the volume of free-growing eucalyptus at the age of 10 years is at least five times that of trees grown at a stand density of 1,600 trees per hectare). Moreover, because farmers obtain most of their fuelwood by lopping branches, trees along homestead boundaries can produce more volume per tree and more frequent supplies of small quantities of wood than trees felled on far-away plantations. On-farm tree growing can go a long way toward mitigating fuelwood shortages, because it is easier to per- suade farm families to plant on their own farm boundaries than to persuade communities to provide scarce land for block plantations. Technical options for expanding tree planting are indeed numerous, and they can be put to use if foresters will become better attuned to how small farmers themselves view their farm. Incorporating trees into the farmers' own farming system (rather than leaving them parallel to it, on a remote communal lot) may promote multi- purpose tree species that will respond to several user needs: fuelwood, shade, small timber, and so on. For instance, species suitable for animal fodder, with fuelwood as a secondary rather than primary benefit, often integrate more organically into the overall farming system than species such as the eucalyptus that is widely promoted by many programs. If marketable species are selected, trees can become a cash crop and bring income, and not be merely a source for household subsistence consumption. In favorable ecological circumstances, with reasonable rainfall, an average rural family needs comparatively few mature trees to cover its cooking and heating needs, and some species, if correctly spaced, can help increase agricultural crop yields. It therefore appears that small land holdings need not be regarded as a barrier to family forestry, and forestry can complement rather than com- pete with food crops. Eliciting and motivating such change in users' behavior is a part of the social strategy for reforestation. Every technical approach must incorporate extension for communicating silvicultural information to farmers, influenc- ing their perception of existing opportunities, and activating the little-used but existing thesaurus of farmer knowledge about trees. Successful forestry programs (e.g., in West Bengal, and in Jammu and Kashmir and Haryana) employ special change agents (extension agents called motivators or social forestry workers) to persuade farmers to plant trees and to help them do so. In recent years social scientists in Thailand and India have been refining extension techniques tailored to the specifics of reforestation (PAO 1988; Indian Institute of Management 1988). A SOCIOLOGICAL FRAMEWORK 323 Disincentives An abundance of naturally grown fuelwood, however, may limit the interest of the family farm in tree planting. A recent study of agroforestry in sub-Saharan Africa concluded that as long as rural producers can collect wood from common lands at low cost, there will be little economic incentive to plant trees on family farms to meet fuelwood demands (Cook and Grut 1989). For instance, field investigations in Malawi found that the returns to labor invested in gathering fuelwood are 15 times higher than the returns to labor invested in growing fuelwood and more than 5 times higher than to labor invested in growing trees for poles (World Bank 1984b). In light of these and other findings, the study on sub-Saharan agroforestry pessimisti- cally predicted that "agroforestry for fuelwood will not be widely adopted in Africa until the free wood resources of the commons have virtually dis- appeared" (Cook and Grut 1989). Even though such pessimistic conclusions may be disputable, the facts that led to them are a reminder that agroforestry has its own limits. This is why alternatives to family-centered forestry strategies, as well as conserva- tion and prevention measures, must be promoted. GROUP-CENTERED STRATEGIES: USING ALTERNATIVE UNITS OF SOCIAL ORGANIZATION Group-centered approaches must not be written off because of the inef- fectiveness of the wholesale community approach or because of the diffu- sion of family-centered forestry. It would be akin to throwing out the baby with the bath water if the deemphasis of community woodlots were inter- preted as renouncing all group-centered approaches. Small Groups The social engineering question to be asked is, Which social formations, ranging on the continuum between the entire community and the indi- vidual farmer, would be capable of acting as supporting structures for affor- estation? Is it possible to avoid the weaknesses of the community-based approach, yet elicit and make use of the social synergy of group-powered efforts in forestry? Alternative types of groups definitely can be identified or constructed. Some have already been formed as a result of local social invention under favorable conditions. The key is to identify a group that is free from the inner conflicts of large communities, yet able to generate the synergy that makes a group more effective than the sum of its members. The limitations intrinsic to communities as social actors stem, as discussed earlier, from their large size and internal stratification. Homogeneous cor- 324 MANAGING THE WORLD'S FORESTS porate groups of a manageable size could prove more functional, Their smaller scale would reduce the problems of system maintenance which are sometimes more complex than the tasks the group is called upon to solve. Even the use of lands under common property regimes is not tied exclu- sively to the pattern of whole-village involvement, but can be arranged through contractual arrangements with smaller groups (Bromley and Cernea 1989; Seymour and Rutherford 1990), Small groups can meet a common need more effectively by joint action than individuals acting separately can do. Users of fuelwood can cooperate not only for joint procurement but also for joint production, Furthermore, a simple rule for members' contribution and benefit distribution (e.g., equal shares for all) can eliminate actual disadvantages or misperceptions of ad- vantages. A small group can also enforce rules through peer pressure and mutual control, so as to restrain free-rider behavior. Members of small groups enter into face-to-face contacts simultaneously as users, producers, and en- forcers. Small groups often manage other natural resources (as in the case of a water users' association formed around a small branch of the irrigation system) and could operate a woodlot largely without the conflicts that sur- round community plantations, Also, given their ability to reinforce and speed up dissemination further, groups may become "contact" partners (more ef- fective than individuals) for extension services. Two experiences relevant to the formation and use of small groups are the group farm forestry and the Arabari experiment, both in West Bengal, described in the next subsection, Group Farm Fore!;try The principle underlying this approach is to link specific groups of people who have surplus labor resources with well-defined plots of land that are unused or underused and can be brought under tree cover. As embodied in West Bengal, group farm forestri; is being practiced by a group of landless families to whom the state government leased marginal public land on a long-term basis (99 years) to enable and encourage them to grow trees with security of benefits, The lease was offered to groups of landless people with the guarantee that it can be inherited but with the restriction that the land cannot be alienated (sold or used for nonforestry purposes), The plots of land are contiguous, thus facilitating collective action in plant- ing and protecting (such as taking turns in watching the plantations), be- cause these tasks are performed more effectively than if carried out indi- vidually. However, the ownership of the trees, maintenance obligations (fertilizer application, replacement of dead trees, etc.), and the right to dis- pose of the products are vested in the individual leaseholders. This system also provides for group control over the temptation to change land use or to mortgage the land, The area allotted and the number of trees to be planted guarantee enough wood from lops, tops, branches, and dead trees to meet a substantial part of A SOCIOLOGICAL FRAMEWORK 325 a family's domestic requirements. The stem volume is then available for sale, and the total income meets participants' interests. The group strategy thus not only maximizes land use for forestry but also provides the users with fuel, construction materials, and cash income. These plantations have generated good revenue11 which some families have invested in purchasing land, planting potatoes, and achieving other such gains. As a social innovation, group farm forestry privatizes the use, but not the ownership, of public wastelands, providing an economic incentive to landless people to raise trees as a cash crop. Where surplus labor is avail- able and employment is scarce, this option can significantly benefit the land- less. However, it also requires prudent implementation to avoid depriving other vulnerable households dependent on wasteland products. The Arabari Forest Management Model The innovation accomplished by the Arabari experiment in West Bengal was to stop forest depletion (encroachment, theft, etc.) through making major changes in the prior system of forest management, which had been based on seeing villagers as the forest's enemies and on keeping them at arm's length. A set of specific and interrelated management measures were intro- duced that encouraged villagers' participation in forest management by making them economically interested in planting and protecting the trees, and even by offering revenue-sharing arrangements. The new system pro- vided villagers with an amount of employment in forest protection and replanting that would be at least equivalent to the value of what the villag- ers had earned by sale of stolen forest products. The experiment had several elements: • Villagers were paid to plant trees (acacia, eucalyptus, etc.) and grasses (e.g., mesta, a poor man's jute) on empty patches. • Jobs were phased in and spread throughout the year to match the seasons of most severe underemployment in the area. • Villagers were given responsibility for tree protection, with mini- mum official interference. • The Forestry Department offered a revenue-sharing arrangement under which the villagers received 25 percent of the selling price of the mature trees in cash (this element was introduced later). • The Forestry Department made an intensive effort to explain the in- centives and the experiment rationale to the villagers. 11 The economic analysis of such a land-lease scheme indicates that if some 2,500 seedlings can be given free to each participating family for staggered year-by-year planting over 10 to 12 years, the family would become self-sustaining on tree crop- ping alone when the first year's plantation reaches maturity. The family would satisfy its domestic fuelwood needs from lops, tops, and fallen wood and could sell the main stem volume for cash, replanting anew each year to replace the mature trees harvested. 326 MANAGING THE WORLD'S FORESTS The results confirmed most of the experiment's assumptions (with the exception of relocating grazing outside the forests), in that the villagers enforced total protection of the forest, primarily by refraining from making illegal cuttings, while their employment in replanting generated revenue for them and for the project. The self-imposed and self-enforced reduction of firewood cutting and the patrolling by villagers acted as a "social fencing" around the state forest. The tensions between the villagers and the Forestry Department eased. The upshot of this successful experiment was that the once-degraded forests were rehabilitated specltacularly within five years and have continued to grow since. Recent assessments have confirmed not only the sustainability of the initial Arabari model but also its rapid spread in the mid- and late 1980s to many more areas. Although the experiment started with no formal group formation in each of the small villages involved, the subsequent follow-up took on stronger characteristics of group creation, with the establishment of village protection committees. A.K. Banerjee (1989) reports that some 700 to 800 such groups were formed in the southwest zone of West Bengal, pro- tecting some 70,000 hectares of degraded and replanted forests: In this area, people have formed formal groups composed of one member from each family. These groups meet once in a while, take decisions and keep minutes. Each family provides a forest watcher at regular intervals.... The will to do so developed as these groups believed in the assurance of sustained benefits.... Their collective action is productive as there is an action plan [for the group]. The sociological principle involved in this model is to create a clear economic interest for a well-defined group in the rational management of a well-defined tract of forest land. Group members need to perceive a clear correlation between their contributions and returns. This awareness moti- vates them to cooperate. Authority and benefits must be restricted to the members of the group, not left open to free-riders. Interesting experiences about the creation and productive activities of small user groups have been reported from Nepal (Messerschmidt 1986), Indonesia, Niger, and elsewhere. Analyzing the collective management of hill forests in Nepal, Arnold and Campbell (1986) emphasize user group motivation, organization, and establishment of legal agreements among the government, the panchayat, and the members of the user groups. The core of the legal agreement is the management plan, which regulates the cutting regimes, product collection, and group harvesting. The prevalence and diversity of such arrangements show the potential for varying the degree and forms of group cooperation as called for by the task at hand and by the people's subjective preparedness for cooperation. A SOCIOLOGICAL FRAMEWORK 327 Forest-Dwelling Groups Indigenous and forest-dwelling populations must be recognized as im- portant social actors in forest-related programs. These groups are some- times small, remote, and dispersed, but overall statistical estimates count their numbers in the millions or tens of millions in some countries, and worldwide in the hundreds of millions. The people living in and around forest areas can be grouped into three broad categories: (1) indigenous people who have lived in the area for gen- erations, (2) people who have moved more or less recently into the area (settlers), and (3) nonresident groups who enter periodically to extract se- lected resources (Partridge 1990). The first two groups actually reside in the forest. They often have a low level of social organization, but their involve- ment and participation in forest management is of growing importance. These groups also possess forestry-relevant knowledge and skills (Warner 1991; Warren 1991). Development strategies designed to relieve local-level pressures on forests must be tailored to reflect the needs and to enhance the capabilities of these different types of populations. Communities that have occupied a forest for generations often base their production system on shifting cultivation with long fallow periods. This situation presented little threat to forest areas as long as settlement size and population density remained low, but the environmental effects of slash- and-burn practices become increasingly adverse as forest-dwelling popula- tions grow, fallow periods are reduced, and previously viable production systems become more precarious. Field research among shifting cultivators in Orissa, for instance, has found that they themselves become increasingly aware of the unviability of their farming system, but in order to adopt alternatives they need significant outside assistance (Bogaert et al. 1990). Forest-dwelling populations rely also on what outsiders often call "mi- nor forest products" but what are indeed of major significance for the live- lihood and culture of forest people, as a source of either food or cash. Such products range from usable leaves to medicine, from sponge fibers to build- ing materials, or from small forest animals to edible plants and fodder (Fal- coner 1990). Their need for such products cannot be overlooked in any alternative approach seeking to protect the forests by stabilizing the dwell- ers' production systems. Government agencies have little knowledge about how to enlist forest- dwelling groups in programs for protecting and managing primary forests. Organizing indigenous dwellers to conserve forest areas and parks, as well as to modify and stabilize their own agricultural production systems, is still an unresolved task, for which little successful experience exists (Serageldin 1990; Revilla 1991). Yet it is an imperative. Forestry departments must ex- periment with institutional arrangements that are culturally acceptable to 328 MANAGING THE WORLD'S FORESTS tribal and indigenous groups and would protect them from exploitation. Ways of channeling benefits to indigenous communities include soil im- provement, forest co-management, adequate species selection, extractive reserves, and fair marketing of forest products (Partridge 1990; Guggenheim and Spears 1991). The conventional and painful "solution" to the effects of shifting culti- vation attempted in various places by forestry departments or by the administrations of reserve parks-the involuntary displacement of forest- dwellers-has generated complex new socioeconomic problems and solved none. Ironically, the recently increasing environmental concerns for preserving wildlife and biodiversity through new national parks and enforcement of protective park regulations, have been accompanied by indiscriminate efforts to forcibly evict forest-dwellers from their habitat. Such involuntary resettle- ment should be avoided whenever possible, because it carries with it a high risk of further impoverishment for indigenous populations (Cernea 1988b, 1991; West and Brechin 1991). Involuntary resettlement of forest-dwellers is also rarely practical: alternative lands are ha.rd to find and frequently the outcome is that other settlers quickly move into the cleared areas. Gener- ally, alternative approaches should be tested and developed that would incorporate forest-dwelling people as participants and beneficiaries in the implementation and operation of forest projects. People living outside but near forest areas usually have customary rights to gather fuel, fodder, and nontimber forest products. Farming communities living outside the forest can put more intense pressure on the forest than forest-dwellers do. The failure of traditional restrictive measures through state intervention implies that additional economic and social approaches to stabilizing agricultural production systems near forest areas are needed. To cope with such pressures, agricultural diversification and intensification programs in areas near the forest should mitigate the threats of forest en- croachment by emphasizing food production and income generation. This action should be complemented by regional planning to direct migration away from forest areas. Cooperatives for Tree Planting and Forest Management Even when tree planting is done by farmers on individually owned land, the creation of a farmers' association may be beneficial for specific activities such as the management of adjacent treelots or the marketing of tree products. In some countries, the forestry departments help establish tree growers' associations or similar organizations to help farmers market wood produced under individual family forestry. The forestry cooperative is one such structure. Although forestry coop- eratives are less common than agricultural cooperatives, in some regions (e.g., Scandinavia) forestry cooperatives are numerous and their organizing principles work effectively (Kilander 1987). From Japan, Oya (1991) reports A SOCIOLOGICAL FRAMEWORK 329 rich experiences with two types of forest cooperatives. One is the "coopera- tive of individual forest owners" in which owners of small patches of forest land form associations to obtain economies of scale in purchasing seedlings and in planting, harvesting, and marketing trees. The other is the "coopera- tive of joint owners of forest," which unites into a distinct organization the village households that share entitlement to the forested commons of the village; this cooperative enables the joint owners to carry out production, marketing, and distribution independently, rather than through the local government. Together, these two types of cooperatives account for a consid- erable, and currently expanding, share in Japan's forested land. Oya (1991) also reports that some of these forest cooperatives have recently adopted a profit-sharing arrangement whereby they mobilize fi- nancial resources of urban residents, who are invited to contribute a certain amount of money for tree planting and forest management; in return, urban residents are entitled to a share of the profit accruing from the forest har- vest. These and other experiences demonstrate that with a clearly defined and not too large membership, forest cooperatives can be a more coherent and goal-oriented unit of social organization than the village community as a whole. Cooperative forestry structures are expanding in India as well, as a result of the initiative to form "tree growers' cooperatives" launched by the country's National Dairy Development Board. This attempt to transfer and adjust the pattern used by the well-known Anand dairy cooperatives is obviously risky but promising. Such tree growers' cooperatives are envis- aged to operate on unencroached wastelands in Orissa, Andhra Pradesh, and other states, with each cooperative covering some 100 hectares of pri- vate marginal lands and 50 hectares of common wastelands (National Dairy Development Board 1985). Subgroups Defined by Gender or Age Many traditional societies, particularly in Africa, entrust certain mainte- nance or service functions in the society to subgroups that are defined by age or gender. These groups are accountable to appointed leaders or to the overall village authority structure. Similar groups could also be used for certain forestry development activities. The creation of women's groups to plant trees is expanding in many countries. The gender division of labor in many traditional cultures makes women the primary gatherers of fuelwood. In certain areas of Nepal, for instance, the time a woman spends collecting fuel is estimated to be be- tween 20 and 40 days a year. Therefore, producing rather than collecting fuelwood may save both time and labor. Rural women generally possess a good knowledge of the characteristics and requirements of various tree spe- cies. For both reasons, women are often more interested than men in raising trees for fuelwood. Recent evidence reported and analyzed by Molnar (1991) 330 MANAGING THE WORLD'S FORESTS illustrates the important contribution women are making to many social forestry programs. Although women's associations for various productive or household- related activities have been promoted in many countries, until recently little has been done to encourage women's organized group action for cultivating woodlots. Even in a country such as Kenya, where women's groups are widespread and effective, a sociological field study reported a few years ago that out of 100 women's groups active in the Mbere district, none was directly involved with planting trees (Brokensha, Riley, and Castro 1983). According to statistics for 1989, however, hundreds of women's groups are now involved in forestry in Kenya, and this trend is growing in other coun- tries as well. In India's Himachal Pradesh, multipurpose women's groups called Mahila Mandals, which have been in existence for many years, now frequently include tree planting among their activities (Dioman 1989). Women's groups could become the prototypical grouping of discrete "users turned producers" in forestry. A strategy is needed to facilitate women's tenure (usufruct or custodianship) rights to suitable land tracts and their secure tenure on trees, as well as to help with group creation. One of the notable recent successes has been the involvement of school- age youths in establishing tree nurseries for social forestry, as reported from Kenya, Malawi, India, Haiti, and other countries. The characteristics of such groups are propitious for undertaking certain collective actions: schoolchil- dren form a homogeneous age group, are organized by virtue of their main activity (going to school), and are subject to a built-in leadership system. Although the temporary nature of this age group limits its participation in forestry activities of long duration, the group is well suited to short-term collective efforts such as the production of seedlings. Nongovernment organizations like CARE have been instrumental in enlisting and financing high schools in developing forest tree nurseries. In Ecuador, for instance, work in three agriculture high schools has revealed both the strength and the weaknesses of schools as existing social groups likely to carry out seedling production, as well as their relative advantages over government nurseries (Desmond 1989). In Gujarat, the Forest Depart- ment encouraged schools to raise seedlings Ito respond to the demand cre- ated by a social forestry program in the state, and within three years about 600 schools opened nurseries in which students produced several million seedlings a year. Drawbacks and limitations in this approach often result from the teachers' rather than the students' lack of time, low competence, and weak incentives (factors that can be corrected). It is important that seedling production be combined with strengthened forestry and environ- mental education of both students and teachers. To formalize and expand the support of schools as existing units of social organization to social for- estry, institutional arrangements can be promoted in the form of a "partner- ship between schools, communities, and government agencies" (Chowdhry 1983). A SOCIOLOGICAL FRAMEWORK 331 Temple and Shrine Forests Temples and shrines are not usually thought of as social units likely to play a role in tree planting and social forestry, yet in some cultures they clearly have this capacity. Research in Thailand and Japan has provided evidence that the monks living around many Buddhist temples in rural locations, or around the Shinto shrines in Japan, maintain an adjacent forest area and promote tree planting. The temple or the shrine acts as a unit of social organization that mobilizes resources for tree planting and mainte- nance and enforces protection rules. The areas covered by such temple or shrine forests may range from several hectares to several tens of hectares or even bigger. In Japan, Buddhist temples and Shinto shrines often mobilize volunteer labor among their constituencies to help the monks plant or carry out maintenance and conservation works in the temple forests (Oya 1991). In Thailand, where 9,000 Buddhist temples exist inside National Forest Re- serve lands, some temple-related monks have emerged as supporters of conservation activities and militant opponents of illegal logging or other forms of forest depletion (Traisawasdichai 1991). The potential of such units for intensifying afforestation is large. Watershed Forestry Watersheds are geophysical entities, and the people who inhabit them are not organized as a social unit. But the topographic pattern of the water- shed and its resources shape human activities, division of labor, and settle- ment patterns. Many human societies in different parts of the world have adapted themselves to the watershed landscape in similar ways; as Lovelace and Rambo (1986) note, there are often "parallels between the ways in which human groups are organized and spatially distributed and the physical characteristics of the watersheds." Planning for watershed use, erosion con- trol, and reforestation cannot be effective and sustained unless watershed inhabitants are enlisted in rehabilitation work. Rehabilitation of deforested watersheds demands much more than massive planting of trees. It involves flood control and soil conservation; bench terraces often need to be built, and they require excavation, leveling, and refill work. Changes may be needed in land-use rights, rules of land transmittal, settlement patterns, and number of inhabitants. These changes are beyond the scope of what individual farmers can do as discrete actors, and group action, as well as support from technical agencies, is required. Sometimes groups are formed spontaneously for such activities. In Haiti, for instance, independent groups of small landholders in the Maissade area have voluntarily collaborated to construct conservationist treatments on commonly held ravines in small watersheds and on contiguous private lands (White 1990). As pointed out earlier, however, coordinated action does not automati- cally ensue just because individuals stand to gain from it. People must --- -- ------ - - 332 MANAGING THE WORLD'S FORESTS understand subjectively their common interests, be willing to act consensu- ally, and organize themselves into some kind of group structure, with goals and rules conducive to carrying out the requisite activities. Coordinated social action to manage watershed resources is probably one of the most complex types of collective action, particularly when structured groups have to be organized. Creating organizational structures for social action and engineering the formation of a self-managed group from discrete (and not necessarily inter- active) farmers is a task no less difficult than any of those previously discussed. Watersheds and microwatersheds can be used as the physical subdivisions within which farmers' activities can be aggregated into coherent group efforts. Such groups could get involved in the design of a land-use plan for the watershed and gain the strength to sustain it through jointly enforced rules. SUMMARY The various types of social .units just examined do not exhaust the list of potential social actors for afforestation programs. The same line of think- ing can be continued in order to spotlight other kinds of social units and thus multiply the array of actors able to involve themselves in forestry development. Forestry departments themselves are also a form of social organization created to perform, by using state financial resources, the functions of con- serving, managing, and developing forests. As administrative bodies, for- estry departments have an organization that is different from the types of social units (organizations of people into groups) discussed in this study. By their position, forestry departments and foresters have a critical role infos- tering and encouraging the formation of such groups among users of fuelwood and in providing them with silvicultural, organizational, and eco- nomic assistance to produce trees. Even though foresters are generally un- trained to carry out the social component of !their tasks, they must gradually learn to work with people as well as with trees. In turn, many nongovernment organizations interested in environmen- tal conservation may also organize groups of people and help fuelwood users to structure themselves as producers. Identifying or creating social units is a task that requires a sociologically informed understanding of what is to be done and the methods and skills fo1r social organization. The point is that such social forms do not have to precede the intervention of develop- ment organizations, nor should they all be created from scratch. Enhancing people's capabilities through organizational strengthening, adaptation, and innovation is integral to the development process. Social forestry connotes both a philosophy of development and a prag- matic operational strategy. The philosophy postulates the centrality of people in forestry, the centrality of the masses of users becoming producers. It A SOCIOLOGICAL FRAMEWORK 333 breaks radically with the stereotypical assumption that forest growth is only the business of professional foresters, or of Mother Nature. In turn, the operational strategy to service this philosophy pertains to the how-to questions and is being fashioned with every new project or bold social experiment that attempts to get people involved in tree growing. The practice of social forestry is wide open to multiple approaches for creating diverse patterns of social organization as matrices for action. It is open to imaginative and informed innovations in land tenure and tree tenure, to various forms of ownership or usufruct, to tested or unorthodox tree-grow- ing techniques, and to age-old or novel social structures ranging from the farm family to all types of purposively created groups. No single social approach will prove best for all situations of participa- tory afforestation. The possible strategies span a broad spectrum and should be encouraged as social inventions for accelerating development. Sociologi- cal knowledge is instrumental for conceiving and implementing each ap- proach through testing and continuous learning. REFERENCES Apichatvullop, Yaowalak. 1991. Local Participation in Social Forestry. Paper prepared for the Expert Group Meeting on Social Forestry and Community Development, UNCRD, Nagoya, Japan, January. Arnold, J.E.M., and J. Gabriel Campbell. 1986. "Collective Management of Hill Forests in Nepal: The Community Forestry Development Project." In Proceedings of the Conference on Common Property Resource Management. Washington, D.C.: National Academy Press. Arnold, J.E.M., and W.C. Stewart. 1989. Common Property Resources Management in India. Oxford, Eng.: Oxford Forestry Institute, University of Oxford. Banerjee, A.K. 1989. A Case of Group Formation in Forest Management. June, processed. Bare, Jean Fran~ois. 1991. "Image de la Finance." L'Homme XXX (3)(July-September):119. Binswanger, Hans P. 1989. Brazilian Policies that Encourage Deforestation in the Amazon, World Bank, Environment Department Working Paper No. 16. Bogaert, M.V.D., S.P. Das, L. Ravindran, P. Fernandes and A. Nystrom. 1990. Helping Forest Dwellers of Orissa to Adopt Alternatives to Shifting Cultivation, CENDERET and SIDA, Bhubaneswar. Brokensha, David W., B.W. Riley, and A.P. Castro. 1983. Fuelwood Use in Rural Kenya: Impacts of Deforestation. Binghamton, N.Y.: Institute for Development Anthropology. Bromley, Daniel, and Cernea, Michael M. 1989. The Management of Common Property Natural Re- sources. Some Conceptual and Operational Fallacies, World Bank Discussion Paper No. 57. Cernea, Michael M. 1988a. Nongovernmental Organizations and Local Development, World Bank Dis- cussion Paper No. 40. _ _ _. 1988b Involuntary Resettlement in Development Projects. Policy Guidelines for World Bank Fi- nanced Projects, World Bank Technical Paper No. 80. _ _ _. ed. 1991. Putting People First. Sociological Variables in Rural Development. 2d edition. (See chapter 6 on involuntary population resettlement, and chapter 10 on social issues in reforesta- tion). New York: Oxford University Press. Chambers, Robert, N.C. Saxena, and T. Shah. 1989. To the Hands of the Poor: Water and Trees. New Delhi: Oxford and IBH Publishing Co. Chowdhry, Kamla. 1983. Schools as Partners in Social Forestry. Ford Foundation Discussion Paper Series, Delhi, August. Cook, Cynthia C., and Mikael Grut. 1989. "Lessons from Agroforestry in Sub-Saharan Africa: Some Farmer Perspei:tives." World Bank, Technical Department, Africa Region, processed. 334 MANAGING THE WORLD'S FORESTS Desmond, Dennis F. 1989. Forest Tree Nurseries in Agricultural High Schools: An Analysis of Ecuadorian Experiences, ODI Social Forestry Network Paper 9e. Dioman, D.R. 1989. Involvement of Women in Social Forestry in Himachal Pradesh. Paper presented at the NWDR seminar, New Delhi, February. Dove, Michael R. 1991. Foresters' Beliefs about Farmers: An Agenda for Social Science Research in Social Forestry, EAPI Working Paper No. 28, Honolulu, East-West Environmental and Policy Institute, October. Dumont, Louis. 1980. Homma Hierarchicus. The Caste System and Its Implications, Chicago: University of Chicago Press. Falconer, Julia 1990. "The Major Significance of Minor Foriest Products-Examples from West Af- rica." Appropriate Technology 17, no. 3. Food and Agriculhlral Organization (FAO). 1988. Planning Forestry Extension Programmes. Report of a Regional Expert Consultation, Bangkok, May. Fortmann, Louise. 1988a. "Great Planting Disasters: Pitfalls in Technical Assistance in Forestry." Agriculture and Human Values (Winter-Spring): 51. _ _ _. 1988b. "The Tree Tenure Factor in Agroforestry 1N"ith Particular Reference to Africa." In Whose Trees? Proprietary Dimensions of Forestry, eds. L.. Fortmann and John W. Bruce. Boulder, Colo., and London, Eng.: Westview Press. Ghamser, Mathew S. 1987. Letting the Piper Call the Tune: Experimenting with Different Forestry Exten- sion Methods in the Northern Sudan. ODl Social Forestry Network Paper No. 4a, June 1. Goodland, Robert. 1991. Tropical Deforestation: Solutions, Ethics, and Religion. Environment Working Paper No. 43. Government of India. 1973. Interim Report of the National Commission on Agriculture and Social For- estry. Guggenheim, Scott, and John Spears. 1991. "Sociological Diinensions of Social Forestry Projects." In Putting People First: Sociological Variables in Rural Development, ed. Michael M. Cernea. 2nd edi- tion. New York: Oxford University Press. Horowitz, Michael M. 1982. Zimbabwe Rural Afforestation Project: Social Analysis Working Paper. Binghamton, N.Y.: Institute for Development Anthropology. Indian Institute of Management, 1988. Planning Forestry Extension Programmes in India. Ahmedabad: Indian Institute of Management. Jodha, N.C. 1986. "Common Property Resources and the Rural Poor in Dry Regions of India." Economic and Political Weekly, no. 27, July. Kilander, Kjele. 1987. "Cooperatives for the Promotion of Forestry in Rural Development." Unasylva 39, nos. 3 and 4. Lovelace, G.W. and A. Terry Rambo. 1986. "Behavioral and Social Dimensions." In Watershed Re- source Management, eds. K.W. Easter, J.A. Dixon, and M.M. Hofschmidt. Boulder, Colo.: Westview Press. Messerschmidt, Donald A. 1986. "People and Resources in Nepal: Customary Resource Manage- ment Systems of the Upper Kali Gandaki." In Proceedings of the Conference on Common Property Resource Management. Washington, D.C.: National Academy Press. Molnar, Augusta. 1991. "Women and International Forestry Development," Society and Natural Re- sources, vol. 4. PPR Working Paper Series, May, Washington, D.C. Murray, Gerald F. 1987. "The Domestication of Wood in Haiti: A Case Study in Applied Evolution." In Anthropological Praxis, eds. Robert M. Wulff and Shirley J. Fiske. Boulder, Colo., and London, Eng.: Westview Press. National Dairy Development Board. 1985. "Meeting Rural Fuelwood and Forage Needs Through Tree Growers' Cooperative Societies: A Pilot Project Proposal." Anand, August, processed. New York Times. 1989. "Forest Murder: Ours and Theirs. Tongass Tree At:en't Cheeseburgers." September 20. Oya, Kenji, 1991. A Historical Review of Social Forestry in Japan. Two Cases of Forest Management at the Local Community Level. Paper prepared for the Expert Group Meeting on Social Forestry and Community Development," UNCRD, Nagoya, Japan. January. A SOCIOLOGICAL FRAMEWORK 335 Partridge, William. 1990. "Forest Dwelling People in Bank-Assisted Projects". Environment and Social Affairs Division for Asia, World Bank, September, processed. Raintree, John B. 1991. Socioeconomic Attributes of Trees and Tree Planting Practices. Rome: Food and Agriculture Organization of the United Nations. Revilla Jr., Adolfo V. 1991. "A National Forestry Program for Community Development." In Philip- pine Rural Development: Problems, Issues and Directions. University of the Philippines at Los Banos. Sajise, Percy E. 1991. "The Evolution of Social Forestry in the Philippines: Lessons Learned and Some Questions Asked", lecture at Wageningen Agricultural University, processed. Salam, Abdul. 1989. "Field Report cin Sociological Aspects in the National Social Forestry Project, Gujarat Sub-Project," January, processed. Sen, D., and P.K. Das. 1987. The Management of People's Participation in Commu_nity Forestry: Some Issues. ODI Social Forestry Network Paper No. 4d, June. Serageldin, Ismail. 1990. Saving Africa's Rainforests, A Publication of the Africa Region, World Bank, Washington, D.C. Seymour, Frances J., and Rutherford, Davilyn, 1990. Contractual Agreements in Asian Social Forestry Programs. Paper prepared for the annual mee!ing of the International Association for the Shtdy of Common Property, Durham, N.C. Shanks, E. 1990. Communal Woodlots in Tanzania: Farmers' Response and an Evolving Extension Strategy. ODI Social Forestry Network Paper No. llC, London. Shepherd, Gill. 1986. Forest Policies, Forest Politics. ODI Social Forestry Network Paper No. 3, October. Thomson, J.T. 1980. Bois de Villages (Niger): Report of an Investigation Concerning Socio-Cultural and Political-Economic Aspects of the First Phase of the Project and Design Recommendations for a Possible Second Phase. Montreal: CIDA, February. Traisawasdichai, Malee. 1991. "Buddhist Sanchtary Proposal." Nation, September 19, Bangkok. Umebayashi, Masanao, 1990. Participatory Afforestation: Lessons from the Japanese Experience. Paper prepared for the Expert Group Meeting on Social Forestry and Community Development, UNCRD, Nagoya, Japan, January. Warner, Katherine. 1991. Shifting Cultivators. Local Technical Knowledge and Natural Management in the Humid Tropics. Rome: FAO. Warren, D. Michael. 1991. "Using Indigenous Knowledge in Agricultural Development." World Bank Discussion Paper No. 127. West, Patrick C. and Steven R. Brechin, ed. 1991. Resident People and National Parks, Social Dilemmas and Strategies in International Conservations. Tucson: University of Arizona Press. White, Thomas A. 1990. Personal communication. Also,- Research Prospectus submitted to the Inter- American Foundation on "Peasant Collaboration on Contiguous Land Treatments in Maissade, Haiti: What Are the Motivations and the Implications for Development Planners?" Processed. Wolf, Eric R. 1966. "Kinship, Friendship and Patron-Client Relationship in Complex Societies" In The Social Anthropology of Complex Societies, ed. Michael Batton. London: Tavistock Publications. World Bank. 1979. "Uttar Pradesh Social FOrestry Project." Staff Appraisal Report, World Bank, May. _ _ _. 1980. "Gujarat Community Forestry Project." Staff Appraisal Report, World Bank. _ _ _. 1981. "West Bengal Social Forestry Project." Staff Appraisal Report, World Bank, Septem- ber. _ _ _. 1982. "Jammu and Kashmir and Haryana Social Forestry Project.'' Staff Appraisal Report, World Bank, September. _ _ _ . 1984a. "Kerala Social Forestry Project." Staff Appraisal Report, World Bank, December. _ _ _. 1984b. "Malawi: Forestry Subsector Study." Washington, D.C.: World Bank. _ _ _. 1985. India: '"National Social Forestry Project." Staff Appraisal Report, World Bank, May. _ _ _. 1988. "Mid-term Review Materials for the National Social Forestry Project in India." World Bank, June. _ _ _. 1991. The Forest Sector. A World Bank Policy Paper, Washington, D.C. 13 The Management of Natural Forests John Palmer and Timothy f. Synnott I t is one of the many paradoxes of tropical forestry over the past 30 years that the rise in public interest has been paralleled by a decline in the application of systematic management. It is also paradoxical that the same period has seen a great increase in research on tropical biology but little corresponding incorporation of research results into management practice. Public interest in the industrial countries has been devoted mainly to the wet forests of the tropics; yet deforestation and land-use problems are just as severe in the dry forests, if not more so. The sheer magnitude of problems, as exemplified by the apparently accelerating rates of deforestation in the tropics, has tended to overwhelm the national capacities of the developing countries to combat the current degradation. Some important factors driving the conversion of forest to other forms of land use and causing the degradation of some of the remain- ing forest have origins outside immediate national control. Other significant factors may be national but nonforestry in character, such as the rate of population increase. The reasons for the general failure of tropical countries to implement land-use planning and the causes of forest loss are covered in other chapters of this book. This chapter concentrates on the factors needed for effective forest management. Few reliable data exist on the extent and state of forest resources prior to the creation of most government forestry departments, so it cannot be stated with certainty that these departments have provided protection and management for goods and services that would have become scarcer if the 337 338 MANAGING THE WORLD'S FORESTS forestry departments had not existed. Nevertheless, forestry departments have often made, and continue to make, large claims about their effective- ness as land managers and providers of goods and services to their coun- tries. For much of the time since 1850, the type of argument based on post hoc, ergo propter hoc was either accepted or ignored, because the means for assessing performance had not been de"'.eloped. In the 1980s, analyses by geographers and historians concluded, as might be expected, that although the foresters' claims were often unprovable, the forestry departments did sometimes restrain forest degradation once tradi- tional regulation of local forest use had been replaced with controls by central government over land and other natural resources. Furthermore, the management of forest resources in developing countries that were not colo- nized, or not for long, including Ethiopia, Liberia, and Thailand, has not been conspicuously better than the management in colonized countries. The labors of the historians have shown also that the creation of a national permanent forest estate, with both protective and productive func- tions, was sometimes undertaken in a heavy-handed manner. Local com- munities now frequently dispute the boundaries of the permanent forest estate. However, official records show that the forebears of these communi- ties often assented freely to the creation of forest and other types of government reserve on land that had been traditionally regarded as being communal or open access. Most colonial regimes developed, on the Indian Forest Service model, procedures to ensure that all parties with a reasonable claim to use of an intended reserve were given due hearing. The colonial foresters often com- plained that the reservation procedures were weighted against them, as they were so intended, because traditional practices of land use and land tenure were often governed by factors not easily understood by people from a different culture. For example, forest use might be seasonal rather than continual; important traditional users might be nomadic or transhumant; different social classes might have different traditional rights and claims; some users might have no, or only limited, traditional speaking rights in community meetings; access to or use of the forest might be of recent origin, or be changing as a result of the proscription by a colonial regime of tribal warfare. The government rules were largely respected as long as the colonial forestry departments were essentially field based, with staff living in or at least being known to the community, provided that the stability of rural communities was not affected in other ways. However, the impracticality of some agreements was demonstrated well before the end of the colonial period. Some countries, acknowledging the impossibility of resettling tradi- tional villages and their farms outside forest reserves, demarcated enclaves instead. Any increase in population size and most changes in land use made these enclaves quite unworkable. For example, enclaves based on the needs for farmland for staple crops became too small once farmland was devoted to perennial cash crops such as rubber or cacao. THE MANAGEMENT OF NATURAL FORESTS 339 The replacement of field-based colonial forestry departments by mainly urban-based and urban-oriented forestry departments in the postcolonial period has increased the difficulty of maintaining forest cover inside the permanent forest estate in the face of increased demand for farmland. For- estry departments have often too lightly dismissed the rural perception that urban-based forest users are accorded preferential access to forest goods · and services. Measures by colonial or independent governments to restrict traditional land use by rural communities, however nonsustainable, are bound to be resented. When a government's claims of authority over resources clearly exceed that government's managerial capacity, it is scarcely surpris- ing that the traditional habits of the rural people tend to affect the state of the forest more than the desires of the government do. Latin American countries have not so far attempted to create the kind of rural-based forestry department founded by colonial regimes in the paleotropics. But those countries have been active in the declaration of na- tional parks and biological reserves, with external assistance, and have en- countered much the same problems as the foresters in India and Myanmar (Burma) more than a century ago. It i)s reasonable to suppose that the governments of tropical countries are going to become more and more preoccupied with urban problems. Net population increase plus urban drift are causing some cities to expand at 6 percent annually. Rural populations also are growing. The availability of cheap food in the cities will depend on improved efficiency of production by agriculture in the periurban areas. In addition, governments will seek to balance their economies by expanding agriculture for export markets, de- spite the nonessential nature and plasticity of demand in the markets they seek to supply. There will be great pressure to convert lowland areas into commercial croplands to supply urban and export markets, and uplands will be viewed primarily as a source of raw materials for urban industry and of perennial supplies of clean water. The Forestry and Agriculture Organization of the United Nations (FAO) and the Technical Advisory Committee of the Consultative Group of Inter- national Agricultural Research (CGIAR) both estimate that some 10 million hectares of "new" land are needed annually to support the increase in world population at current levels of nutrition and agricultural yields. Most of this land will come straight from forest, because it is relatively unencumbered legally. Some people claim that as much as 85 percent of the land being converted from forest to agriculture is not being logged systematically be- fore the residual stand is felled and burned; this claim cannot be verified until the International Union for Conservation of Nature and Natural Re- sources (IUCN) completes its surveys of forest condition, which are running somewhat behind the 1990 FAO-executed global forest survey. Logged or not before conversion, the 10 million hectares collectively represent an area substantially greater than that under management for sustainable timber production. The difference between the FAO-estimated current rate of deforestation in the tropics, 17.8 million hectares, and these 340 MANAGING THE WORLD'S FORESTS 10 million represents the inefficiencies of agriculture and land speculation that have little to do with food production. Boycotts of tropical timber im- ports, or calls for the total preservation of remaining tropical forest, are irrelevant responses by the independent sector to this reality. In fact, almost all nongovernment organizations involved with land-management projects, such as World Wildlife Fund for Nature-United States (WWF-US), World Wide Fund for Nature International (WWF--International), and the Inter- national Union for the Conservation of Nature (IUCN), include forest man- agement for production as essential elements of their overall conservation strategies. Foresters have to recognize that government support is likely to become more difficult to obtain in the future, given the likelihood of the predomi- nance of urban problems. Foresters therefore have to ensure: • that their arguments for forest management have a sound factual basis, • that management methods are congruent with needs, and • that institutional backing is appropriately organized. It is important that government forestry departments should not aim to be the sole managers of forests in the national interest. Industries and woodl- using enterprises also have a legitimate interest in ensuring that the source of their raw materials is managed appropriately, but this interest must not be allowed to dominate decision making. There is widespread recognition that land-hungry farmers on the forest fringes must be given equitable ac- cess to forest goods and services and some genuine role in the decisions about management. There is also a growing :recognition that consumers of forest products-including everyone affected by the range of goods, ser- vices, and environmental effects of forests-also have a legitimate interest in the fate of the forests. Some officials in and governments of developing countries do not share these ideas. Conflict over forests and land is likely to sharpen as public opinion, democratic movements, and special interest groups become important factors in influencing policy decisions. In the development of arguments for forest management, the scale and variety of problems have to be appreciated. The following sections outline the extent and types of tropical forest in relation to management, and the goods and services they could provide sustailnably. EXTENT OF AND VARIATION IN NATURAL TROPICAL FORESTS The apparent luxuriance of growth and diversity of species that so attract a foreign visitor to the natural tropical forests are simultaneously obstacles to management of the forests. In comparison with temperate forests and the supposed simplicity of plantations, tropical forests often display complex structure, marked variation from one hectare to the next, relatively unpre- dictable growth rates, and uneven age. THE MANAGEMENT OF NATURAL FORESTS 341 About two-thirds of the world's approximately 250,000 species of flow- ering plants flourish in the humid tropics. Few of these species are common to all three main blocks of tropical moist forest. Alpha diversity (within a vegetation community) seems to be highest in the wet forest on young fertile soils in geologically unstable topography, such as the western slopes of the Amazon Basin and the interior of Borneo. One-acre and one-hectare plots have been used in many parts of the tropics for studying floristic diversity. Yanamomo in Peru holds the record so far, with 283 species of trees with diameters of 10 centimeters or more; in contrast, a typical plot in geologically old Nigeria can muster only 23 species per hectare. A plot of only 100 square meters at Las Horquetas in Costa Rica contained 73 tree species, but the total number of vascular plant species was 233----and bry- ophytes added another 32 species. The plot of 50 hectares at Pasoh in Pen- insular Malaysia holds about 830 tree species, while 711 species stand on 6.6 hectares at Bukit Raya in Sarawak. In contrast, there are 50 indigenous tree species in Europe north of the Alps and west of the former U.S.S.R., and eastern North America has 171. The presence of many plant and animal species indicates that loss of forest is almost certain to result in the loss of biodiversity. Specimens col- lected by early explorers and botanists from some tropical localities have never been recollected because the sites have lost their forest cover. For example, the single collection of Burkilliodendron album from a limestone hill in Parak in Peninsular Malaysia is unlikely to be repeated because the hill has been quarried to the ground for road stone. In contrast, the reduction of natural forest in climatically temperate England to about 4 percent of the country does not seem to have caused the loss of any woody plant species from the flora. The immense richness of some tropical moist forest is not uniform. Infertile soil, bad drainage, or difficult topography support communities that may be much poorer in species. Some forests are adapted to distur- bances such as hurricanes or cyclones or landslides. Salt-tolerant mangroves form communities and graded types on muddy tropical coasts. Increased seasonality of climate causes dramatic reductions in the numbers of species. There are notable examples of forests dominated by one or a few nonpioneer species in areas where otherwise the forest has no dominants. Although it has often been assumed that some mineral deficiency or excess in the soil has caused this situation in such forests, more recent and careful studies have tended to give a verdict of "not proven." Silviculture and management of these species-poor forests are relatively easy if regeneration of the domi- nant species is abundant (as in some Mora excelsa forests in Trinidad), or difficult when regeneration is usually absent (as in forests dominated by Meliaceae in West Africa or the Shorea albida peat-swamp forests in Borneo). Although perhaps half the area of tropical forest has been lost since the beginning of this century, large blocks still remain-about 1,800 million hectares in total. Some one-third are dry forests and two-thirds are wet. 342 MANAGING THE WORLD'S FORESTS About 400 million hectares of closed-canopy moist forest are in the neotropics, 250 million hectares are centered on the botanical region of Melesia, and 180 million hectares are in western and central Africa. About 80 percent of the remaining tropical moist forest is located in only 20 countries. About 33 countries export tropical timber, but possibly as few as 10 will have surplus commercial supplies available for export by the end of the decade. GOODS AND SERVICES FROM THE NATURAL FORESTS Only a few tropical countries have embarked on the huge task of listing their natural products. The Dictionary of Economic Products of India was be- gun in 1884 and completed in 1893, with a selective updating as the Com- mercial Products of India seven years later. The Dictionary of the Economic Products of the Malay Peninsula was published in 1935 after a gestation of almost 14 years. A high proportion of the thousands of items in these dictio- naries are derived from natural forest. Some of the named products have lost their 19th-century commercial markets, displaced by more reliable and unadulterated synthetics produced cheaply by chemical reactions from the waste products of other industries. The former importance of some small (in volume) but commercially valuable products is shown by the quantity of legislation devoted to them. This was especially true of exudates tapped from trees, where careless or greedy tapping could shorten the life of the tree and render it useless for timber. Examples include the damar resin for varnishes from Dipterocarpus in the Philippines, jelutong for chewing gum from Dyera in Malaysia, and chicle for chewing gum from Manilkara (syn. Achras) in Central America. These minor forest products were taxed when they entered commercial trade and formed a useful component of central or local government revenues. The importance of these products now is not so much in their historical roles but as indicators of the potential value of the forest as home to the many thousands of species whose properties are as yet poorly known or completely unknown. The phytochemical survey of India, which has been running for decades, is still far from complete. Few commercial companies are interested in relying on supplies collected directly from the wild state if they can use the natural product to determine the structure and composi- tion of the active component and then synthesize that compound. Thus a major reason for retaining forest as such is for its information content, es- sentially, its genetic material. The example of India shows how long it takes to screen the forest prod- ucts for their human utility. The dictionaries mentioned earlier were com- piled largely from folk experience rather than from laboratory analyses. The moral and emotional arguments in support of a better deal for the often underprivileged tribal peoples who live in the tropical forests could be supple- THE MANAGEMENT OF NATURAL FORESTS 343 mented with the purely material argument: that their historically acquired knowledge and beliefs about forest products provide a preliminary level of screening to guide more formal research. · Although some forest products are currently being touted for their pharmaceutical benefits, other forest products also experience surges in demand from time to time. Bamboo and rattan furniture declined in popu- larity in the industrial countries during the first half of this century. Then American service personnel returning from the war in Vietnam took a fancy to the comfortable and robust furniture made from rattan, and the market has boomed. The value of international trade in rattan canes and rattan products may approach US$100 million annually. The collection and pro- cessing of rattan is labor-intensive. Therefore, producer countries might be expected to be eager to expand research on methods of rattan cultivation and processing, but the research effort in the 1980s was relatively academic and insignificant in relation to the value of the trade. Increasingly important are the services obtained from natural forests, which have been appreciated since the Greeks and Chinese realized over two thousand years ago that soil erosion and declines in crop yields were associated with deforestation. Expanding urban populations consume more, not fewer, forest goods and services per head than rural communities. In the future, water for domestic and industrial purposes (together with the di- verse germplasm) is likely to become the most important tangible forest product. Control of water flows, amelioration of flood risk, and reduction of bed loads of rivers will become important intangibles. The three main kinds of forest goods and services may be summarized as follows: 1. Mass Products • Timber (logs, pit-sawn lumber), • Poles, • Firewood (charcoal), • Shingles, • Bamboo and rattan, • Fodder, and • Thatching grass. 2. Minor Products (mirior by volume, not necessarily by value) • Extractives (bark, dyes, fibers, gums, incense, latexes, oils, resins, shellac, tanning compounds, waxes); • Parts of plants and animals for magic and medicine and decoration; and • Food (bush meat, flowers, fruit, honey, leaves, nuts, seeds, spices). 344 MANAGING THE WORLD'S FORESTS 3. Services • Reduction in airborne pollution; • Perennial flows of clean water; • Reduced erosion by wind and water; • Reduced flooding and sedimentation/bed-loading downstream; • Protection for crops and domestic animals against wind damage, excessively high and low temperatures, blown sand, epidemic preda- tors, and parasites; • Protection of ecosystems and conservation for biodiversity; and • Tourism and recreation, Although most forestry departments are constituted and staffed, in theory, to manage forests, and few, if any, national forest policies actually enjoin forestry departments to engage in revenue collection, foresters tradi- tionally have been viewed as collectors of revenue for government through taxation of products from the natural forests, Conventional forestry educa- tion gives little time to theories of valuation and taxation, and foresters proved to be poor estimators of the holistic values of the forests in their charge, They tend to value the tropical forests largely, if not exclusively, in terms of the taxes on their tangible products. Because the taxes have fre- quently been set in an arbitrary manner, unrelated even to direct market values, Jet alone the real costs of production, the resultant "valuation" has often been disastrous for the forest. Perhaps the best-known example of such a disaster was the economic calculation that followed the land capabil- ity survey in postindependence Peninsular Malaysia. The low value attrib- uted to the forests, based on the royalty rates, contributed to the massive clearance of lowland dipterocarp forest, which had been under formal man- agement for timber production, in favor of oil palm and rubber. These observations may give the impression that foresters in the tropics were unable to make the economic calculations that would have strength- ened their case for better forest management, given that the actual tech- niques for management have been known for some time. In fact, many foresters have attempted to draw the attention of their governments to the anomalies just mentioned but have been discouraged from actions that were viewed as infringements on the profits of timber companies. It is sad that governments of developing countries were unprepared to act until writers from industrial countries repeated and publicized these facts, some decades later. THE JUSTIFICATION FOR FOREST MANAGEMENT The principal justifications for managing the natural forest are that the goods and services continue to be needed by human populations and that the total package of these goods and services can be produced more effi- ciently and effectively by retaining land under forest than by converting the THE MANAGEMENT OF NATURAL FORESTS 345 forest to other forms of land use. Water supplies, for example, are usually cleaner and more reliable from a naturally forested catchment than from one maintained by constant effort under artificial vegetation and engineer- ing structures. This is not to argue that every single hectare of forest is capable of producing efficiently the whole range of potential goods and services. Land capability surveys continue to detect areas now under forest that could provide a higher and apparently sustainable economic yield under some other crop or crops. However, less and less of such fertile and operable soil is available for conversion, and the remaining forest is more and more valu- able for those goods and services that cannot, or cannot easily, be produced by another form of land use. The independent sector has emphasized the increasing importance of the information (genetic) content on the forest-and the moral argument that if we do not know what the forest contains, we should take care, for the sake of the future and more knowledgeable generations, to conserve what evolution has developed over hundreds of millions of years. The task of management is to reconcile the moral imperative to conserve options for the future with the immediate demands and needs of today's farmers, forest industries, and other parties that have an interest in the forest land, goods, and services. In tribal communities, it was, and sometimes still is, a common practice to harvest resources intensively and then move to another area and repeat the process. In evolutionary terms, this is an advantageous strategy if any resulting negative externalities can be ignored or discounted by the harvest- ers. Most of the existing hunter-gatherer forest-dwelling tribes live in this manner, following routes through the forest that have become traditional over long periods of time. It is also the method approved for the organized management of the mass products from forests; but in this case the possible negative externalities should be specifically recognized and countermea- sures should be taken. In both cases, the period of return for a subsequent harvest should be long enough for the harvested product(s) to recover to a sustainable level, on average. Productive management is possible and feasible. It is an observable fact that forests reproduce themselves and that, in the absence of human inter- vention, forests will dominate all but the harshest sites. The main issue is not whether production management is possible but how the human de- mands on the forest can be reconciled with the biological possibilities. Many traditional strategies of resource use have tended to be in the direction of exhaustive harvesting, resulting in boom-and-bust cycles. The "natural" controls of warfare and disease on human populations have now been largely removed, but human society has not yet developed a pattern of behavior to cope with the relatively new conditions. Conventionally, government-the traditional chief of the local tribe, the princely or priestly power, or the de facto local or central national authority 346 MANAGING THE WORLD'S FORESTS operating through civil servants-supplies the controls in order to provide for future years and generations. In some cases, the effective control has been vested in a private or public commercial company, but historically the time scale of forest management has not attracted investment by the pri- vate sector. There are many examples of small-scale forest management by long-settled communities for their own use, but there appear to be no ex- amples of long-term tropical forest management by essentially commercial organizations. At present, local communities in developing countries usually lack both the technology to harvest and manage the forest for multiple goods and services and the social and administrative structures to deal collectively and effectively with the commercial world. Moreover, increasing numbers of the farmers in and near the forests are recent arrivals from other areas that have different soils and climates. These people often have little knowledge of appropriate land use in the more humid climates and strongly leached and erodible soils, and do not understand the characteristics of forested areas in which they now dwell. Not all local communities, farmers, and other set- tlers are interested in long-term forest management. On the contrary, con- version of forest to agricultural land, or rapid sale of all valuable timber in newly colonized areas, is often essential for farming families to survive. For example, many Latin American transmigrants have no intention of settling permanently in the forest or on its fringes. They wish to mine the resource to raise capital so that they can set themselves up as small entrepreneurs in urban areas. To introduce forest management for local benefits and to make a success of it, favorable social circumstances, technical and financial inputs, and appropriate state controls are all required. In most cases, therefore, we conclude that the overall authority for management (or for setting the norms for management) of tropical forests must lie with legally recognized and accountable government. The reason is that a representative government is the agency most likely to have both the long-term interests of the country in view and the most access to the exper- tise necessary to plan and control long-term land use. It is possible for government to devolve powers and to contract out operations to the com- mercial sector and to local communities, but long-term intergenerational management calls for sustained government interest. This conclusion clearly contrasts with our skepticism, noted earlier, about the ability of government to pay sufficient attention to rural matters. This contrast means that national governments, forestry departments, and donor organizations must look closely at the role of the forestry departments. Projects to strengthen for- estry institutions should be better thought out and targeted than has been customary during the past 30 years. As for management of natural tropical forest by the local community or the private sector, the task is not to resurrect ancient systems of manage- ment that were developed when demands on land resources were light. Instead, new social systems must be developed to cope with the increasing THE MANAGEMENT OF NATURAL FORESTS 347 demographic pressure and with the domestic and commercial demand from a shrinking land base. ASPECTS OF THE GOAL OF SUSTAINABILITY Although foresters have been using the concept of sustainability for centuries, the term has only recently passed into general circulation. Its definition remains the subject of much debate. Groups campaigning for the total prohibition of human use of the remaining tropical forest have seized on incautious statements by some foresters to suggest that forestry is too serious a business to be left to foresters. The 1988-89 study by the Interna- tional Tropical Timber Organization (ITTO) of forest management policies and practices in Sarawak failed to point out the valid reasons for the differ- ence between the first harvest from old-growth forest and the considerably lower sustainable yields from subsequent cutting cycles. The same misun- derstanding has been prevalent among groups campaigning in Australia. This section discusses sustainability first in general terms and then in relation to forest management for mass products. The World Conservation Strategy (WCS), developed and promoted by IUCN since 1980 and now being revised, mentions axioms and goals relevant to sustainability. The main axiom is that conservation depends on development, and that lasting development is impossible without conservation. Objectives such as the maintenance of essential ecological processes and the preservation of ge- netic resources would be best served by the exclusion of human influence, but such exclusion is rarely practicable now and demographic pressure will make it increasingly difficult. The compromise that the WCS proposes is the sustainable use of species and ecosystems, which raises the question of the amounts to be used and the amounts to be preserved. Land and water can be used in many different ecologically sustainable ways. It is possible to exploit the natural forest more or less intensively, depending on its type and history of use. The forest can be replaced by an artificial ecosystem, which also can be exploited sustainably at different levels of intensity, according to the energy and skill that are invested in caring for it. The more capital-intensive forms of management usually carry their own costs in higher use of energy, increased risks of pollution, and so on. Demographic pressure requires intensificatio,n of use and management, if the quality of life is to be maintained even at its present low level. The conservation of tropical forest must therefore be concerned, first, with finding an acceptable balance among the different intensities of use. IUCN defined sustainable development as "improving the capacity to con- vert a constant level of physical resource use to the increased satisfaction of human needs," while the World Commission on Environment and Develop- ment preferred "development that meets the needs of the present without compromising the ability of future generations to meet their own needs." The literature on the subject is now extensive, and people generally recog- 348 MANAGING THE WORLD'S FORESTS nize that sustainable forest use should be viewed in the context of sustain- able use of renewable natural resources in the land-use continuum. The subsections that follow touch on the eight most significant aspects of sustainability that apply to a country that still has extensive natural forest but needs to develop economically to meet the requirements of an expand- ing population. These eight aspects do not deal directly with the scale on which sustainability should be sought. We recognize the importance of na- tional and local decisions about scale. Foresters conventionally treat such matters as they develop schemes for the long-term supply of forest goods and services to consumers. 1, Protection The natural forest is a bank of land and of resources, tenurially secure, and maintained at little or no cost. Until it is needed for other purposes, all forest land should be kept under forest. The burden of proof should be on those who wish to alter or remove the forest. If it is necessary to intensify land use, areas that are not under forest should be chosen first if they are available. Selected areas should be permanently dedicated to protective uses, in areas where the loss would be great and irreversible if they were to be converted to other forms of land use. This implies that absolute preserva- tion against human influence is necessary to ensure protective functions. 2. A Place for Indigenous People Indigenous populations, often tribal, live in many tropical forests. People in some countries argue that the self-sufficient way of life of such popula- tions should be preserved partly for sentimental reasons. In contrast, some governments and dominant ethnic groups take the view that the tribal peoples should be enabled or encouraged to adopt the way of life of the mainstream national society, and that there are no good grounds for reserving large areas of forest for extensive use by one segment of society. As noted earlier, the donor agencies do not need to rely on moral or sentimental arguments, because there is a good material argument for the preservation of indig- enous technical knowledge. The recording and dissemination of indigenous knowledge usually depends on the conservation of its social context, and therefore the tribal way of life has a special value. In many regions, a large body of indigenous knowledge remains intact, or is not yet extinct, even where the framework for traditional resource management systems has vanished. The traditional domains of tribal peoples will continue to be threatened by expanding populations, their own and those of other segments of soci- ety. Many tenurial systems of indigenous peoples have evolved under con- ditions quite different from those exiting today. The tenure and the agricul- tural systems are often bound up with the spiritual culture of the society and are not easily adapted to increased populations. THE MANAGEMENT OF NATURAL FORESTS 349 3. Fragile Soils Measures for the conservation of all fragile soils, and the resulting pro- tection of water resources, should take priority over any form of use or development that would cause additional erosion or loss of fertility. Thus all areas with fragile soils must be mapped in order to exclude them from use or development. In cases of extreme fragility, a permanent ban on ex- ploitation may be needed. In other cases, no use or development should be permitted until it has been established that the action could be carried out, and later managed, in such a way that no soil deterioration would take place. It is worth noting that some forms of intensive agriculture on fragile soils may actually improve them; one example is the pig-and-vegetable cycle employed by the overseas Chinese in those parts of Southeast Asia where the immigrants are banned from acquiring tenure except on the worst soils. 4. Genetic Resources Special measures should be taken to preserve the intraspecific variation of species of economic importance. Such measures would include protecting some areas especially for this purpose and applying management constraints in chosen areas elsewhere. IUCN has drafted guidelines for the conserva- tion of biodiversity in tropical forests. 5. Samples of Ecosystems Give our current dearth of knowledge, the conservation of samples of intact ecosystems is prudent. The interactions among species may be as important as the species themselves for the maintenance of essential eco- logical processes. For example, noncommercial trees may be necessary to feed the pollinators or disseminators of commercial species during the sea- sons when the commercial species are not flowering or fruiting. The regen- eration of the commercial species would be blocked if the noncommercial trees were to be removed. The samples of ecosystems should include those that are widespread as well as those exceptional systems that are often on unusual sites and worth preserving for other purposes such as recreation, edu~ation, and tourism. 6. The Full Range of Species Many species will be protected by the measures already cited, but supple- mentary action may be needed to protect areas of special interest because of their exceptional richness or unusual character; the feeding, roosting, and nesting sites of migratory birds and insects; and sufficient territory to en- compass full breeding populations of wide-ranging mammals and birds, especially those at the tops of the food chains. 7. Land for Food When not required for protection and other service functions, the best soils should be earmarked for agriculture through a land capability survey. The conversion of forest to agriculture should be avoided for as long as possible. People in greatest need, especially those who are constrained to 350 MANAGING THE WORLD'S FORESTS overuse and degrade the land on which they live now, should have the first call on the "new" land. Conversion of forests to new farms should be ac- companied by technical extension programs aimed at improving farrning productivity and sustainability. Agroforestry techniques may help to retain the protective characteristics of the forest structure. 8. Forests for Sustainable Production The area of a country that should be devoted to the production of forest goods and services has been the subject of ·much debate. At one time, com- parisons between countries that were in various stages of abundance or deficit in forest products suggested that about 20 percent was desirable for self-sufficiency, but this measure is crude and easily challenged by competi- tors for land. The great potential for single-use mass products from geneti- cally improved and intensively cultivated plantations has forced a more flexible view. An important part of the forestry-sector master plans promoted espe- cially by .the Asian Development Bank has been devoted to estimates of future demand and supply of forest products. Countries that have opted for the Tropical Forestry Action Program approach have devoted somewhat less emphasis to such predictions. These are merely the latest manifestations of numerous attempts to reconcile rising demand for forest goods and ser- vices with the apparently limited potential of the natural tropical forest to increase the turnover rate. of those products that find a human use, espe- cially a commercial human use. Forests in the temperature zone have evolved under the particular stress of winter, for growth in these forests must be concentrated in relatively few months. But these forests have two advantages: During the growing season the temperature does not rise so high that a large proportion of the photo- synthates are immediately lost through respiration, and pests and patho- gens are rendered mostly inactive during the colder months. Conversely, especially in the lowland perhumid tropics, solar radiation is often impeded by cloud cover, persistently high temperatures cause continual losses in respiration, and attack by pests and pathogens is almost .continuous. The typical large trees ("K" type) of the mature phase of the polyspecific tropical forest are adapted to conditions of stability: they are shade bearing and slow growing, and have few and large propagules, long periods to maturity, and heavy investment in phytochemical defense. Given the rela- tively favorable climatic conditions of the lowland wet tropics, this strategy is "rational" for a large proportion of plant species. Most such characteris- tics are contrary to those desired for the satisfaction in bulk of present human needs: rapid growth, many propagules, short periods to maturity, and blandness for human products that need industrial processing. The latter, "R" type, characteristics are found more in species of seral or coloniz- ing forests and in the building phase of the polyspecific forests. THE MANAGEMENT OF NATURAL FORESTS 351 It follows that management for mass products would not seek to restore the forest to an approximation of its original state after logging for timber or cutting for firewood. On the contrary, managers would try to encourage the type of species suitable for industrial processing. That is, the manager would foster a change in floristic composition of future crop trees as well as a change in structure toward a stand of more uniform size for the next har- vests. The logged forest retains and regains many features of natural origi- nal forest, but it is unrealistic to expect to re-create a primary forest after logging or other major disturbance. (Note that this is not necessarily true for dry tropical forests, which are managed primarily for fuelwood and animal fodder by coppicing or lopping trees with fire-tolerant rootstocks.) To encourage these shifts in the forest, the manager would prefer the harvester to remove trees that have reached marketable size as well as smaller trees that appear to be having an adverse effect on the future crop trees. The primary forest (forest that shows no sign of significant human disturbance) often contains huge and possibly ancient trees, some of them highly defec- tive outside and rotten inside. The manager will usually be delighted to see the removal of trees that are accumulating no net growth; indeed they may be rotting internally faster than the wood builds on the outside of the trunk. These huge trees also occupy growing space in which the manager would prefer to see young trees of commercially desirable species. Because of the one-time presence of the huge, old, and often valuable trees, the first systematic harvest or logging of primary forest may well yield, per hectare, gross volumes that are greater than will be possible in subsequent cuts. The. manager will influence replacement crops to provide a more uniform product for domestic use or commercial processing at the second and later harvests. If the trees have produced their own replacement progeny at the time of the harvest and have reached a technical or economic culmination of their mean annual increment, there is usually no reason to leave them to grow especially large. Trees smaller than those in the primary forest can be harvested with lighter machinery, and thus inflict less damage to the soils and residual stand. Again, what applies to tropical wet forests may not apply to the dry woodlands. Fodder produced through frequent light lopping of trees may be more digestible than the fodder obtained by the first systematic harvest from old trees. Likewise, harvests of poles and fuelwood from multiple sprouts on coppiced stumps in dry forests may have more usable biomass than large and heavy logs and branches from a first cut of miombo wood- lands. Optimum piece size from manually harvested fuelwood stands may be small enough to encourage economic cutting well before the culmination of the annual increment of biomass (which is the rotation of maximum volume production). It should be clear that, because of the changes forest managers intend to produce in the second and subsequent harvests, the sustainable level of forest production cannot be calculated from the size of the harvest of the 352 MANAGING THE WORLD'S FORESTS primary forest. In tropical forests, where a variety of goods and services could be provided, sustained production of one good or service may be incompatible with that of another. As already indicated, maximum biologi- cal diversity may be conserved by very light and occasional harvests of genetic material, but attempts to maximize sustained yields of sawlogs may reduce the genetic resources characteristic of mature-phase primary forest. Timber-processing industries in the tropics have been equipped with heavy machinery to cope with the large logs from primary forest. A combi- nation of the natural defects in the logs from primary forest and historically small pressures on the industries to operate at a high level of technical efficiency means that the industries obtain a higher return from large logs than from small. Foresters have arbitrarily set minimum diameters above which trees can be cut, based mainly on industry requirements, because they lacked data on growth rates. Few tropical trees of commercial importance produce reliable annual growth rings, so past growth rates cannot be determined by inspection of cores or disks. Despite many ingenious studies of wood anatomy, there is no currently practicable alternative to repeated measurements of trees in tropical forests for estimating growth on a management scale. The large amount of so far unexplained variation in rates of growth of tropical trees, especially in moist forest, requires long-term measurement of many trees to provide reliable estimates. Although many thousands of plots have been established in the past half-century, only a .small fraction have been main- tained long enough to produce usable results. Foresters have thus often used rules of thumb, such as 1 centimeter in diameter per year or 1 cubic meter of industrial stem wood per hectare per year. These rules are ad- equate for planning at a broad scale but not for discriminating between species and site effects. Individual trees in natural tropical forest may have growth rates as great as those in well-managed plantations. However, the low stocking per hectare of the naturally occurring wide-crowned species means that selec- tive harvesting of each tree as it reaches maturity would cause repeated damage to the residual stand, including the regeneration of the desirable species. Hence, commercial operators prefer periodic harvesting when a sufficient number of technically and economically mature trees can be re- moved in a controlled manner. The yield should be large enough to be financially attractive to loggers but not so great that the remaining stand is seriously damaged. After harvesting, the forest should be closed to loggers until the next period of maturity. The period of recovery and growth-the felling cycle-is long in comparison with agricultural growth cycles and conventional investment cycles. In other words, the biokigical needs of the forest for a long period of recovery always conflict with the desires of hu- man users to return frequently for intermediate harvests. Reentry-the re- turn of loggers to take previously noncommercial species or sizes, which was demonstrated to be highly damaging to regeneration more than 35 years ago-is still a problem in many countries. THE MANAGEMENT OF NATURAL FORESTS 353 The foregoing discussion has concerned mainly the tropical moist forest and the production of large logs. The principles and problems apply equally to the "minor" (small in volume) forest products from moist and dry trop- ical forests. A decline in the quality and quantity of these minor forest products is widely reported, wherever market prices are high enough to stimulate frequent passes through the forest by hunter-gatherers. This de- cline shows that if prices are high enough, individuals and communities can rarely forbear to overharvest now. In order to sustain the yield, controls have to be imposed from outside, especially when the resource is effectively open-access rather than communally regulated common property. To. a large extent, natural forests remain today in the tropics not be- cause harvesters have managed them conservatively, but because the har- vesters have not, until recently, had the marketing opportunities or the technology to cause major damage and decline. The widespread use of re- liable logging machinery, powerful enough to reach, cut, and extract the largest products (logs) from natural tropical forest, dates back less than 50 years, or about the time it takes for a tree to grow from seedling to sawlog size. Because harvesting systems involving caterpillar-tracked bulldozers and rubber-tired skidders arrive during the first felling cycle after entry into the primary forest, and because growth of the residual stand cannot be easily predicted, it is not yet possible to demonstrate unequivocally that production of large-dimension saw- and peeler-logs can be sustained over many felling cycles if heavy machinery is used in harvesting. But the mass of evidence of many types indicates that sustained production is possible when appropriate management procedures are installed and implemented. The goal of sustained production in natural tropical forest is compli- cated by the variety of possible goods and services. That variety is exempli- fied by the various demands supplied by the forests of Peninsular Malaysia between the last decade of the 1800s and the mid-1960s; 70 years was the estimated time to maturity for trees of the light red meranti group of Shorea species, a major component of the canopy of lowland dipterocarp forest in Malaysia. During this time, the forest supplied successively the following major waves of goods: • Gutta-percha (Palaquium spp.) for coating submarine cables and mak- ing golf balls; • Heavy, naturally durable hardwoods for railway sleepers and con- struction, such as Balanocarpus heimii; • Poles and firewood from many species for use in the expanding tin mines; • Medium hardwoods for sawnwood, such as Dipterocarpus, used in the construction of the expanding urban centers; and • Light hardwoods for veneer, plywood, and sawnwood, such as some of the Shorea species. Two world wars, the Depression years of the 1930s, the Communist rebellion of the 1950s, and the upheaval associated with political indepen- 354 MANAGING THE WORLD'S FORESTS dence have contributed to ensure almost continual uncertainty and instabil- ity for forest management in Malaya, But by retaining the forest as forest, Malaya was able to supply the products just listed at negligible cost. This example does not demonstrate sustainability of production, but it does point out the possibilities of obtaining multiple products from maintained forest. None of the species mentioned grows economically in plantations. A strict demonstration of sustainable production would require the monitoring of two rotations after the first systematic harvest in primary forest. The first rotation would set the level believed to be obtainable; the second would confirm that it was actually sustained. If the primary mass products from managed forest are sawlogs and the sawlogs come from species of fast growth, a period of about a century would be needed for this demonstration. This strict definition is clearly unhelpful for taking decisions about land use because, in practice, few countries have forests that were subjected to commercial logging more than ,100 years ago. For practical purposes we need criteria that apply to many typical areas and minimize the chances of error, based on currently available evidence. More practically, "success . , . is achieved when a stand of fine-timber trees has been brought to maturity and is producing natural regeneration, on a site where it has matured before, and where the soil shows no sign of deterioration" (Dawkins 1988). This definition of success reduces the quali- fying period to a single rotation following the systematic harvest from pri- mary forest. Similar criteria can be applied for the sustainable production of any other biological product. It should be obvious from this discussion that claims about sustainable forest management should relate to the specific local objectives, practices, and resources. Arguments about levels of sustainability cannot be resolved now for any of several reasons: because insufficient time has passed since the first systematic harvest in almost all natural tropical forests, because early data were inadequate, or because markets have changed and knowl- edge about management methods has improved. Sustainable management, leading to sustained management, of natural tropical forest is a goal, justi- fied in terms of continued demand, but not a rigid target. At the same time it is a flexible and location-specific form of land use that may be explained and understood in various ways. MANAGEMENT 15 NOT JUST HARVESTING For some loggers and foresters, management means simply opening a forest to logging. As mentioned earlier, foresters in many regions are con- cerned mainly with collecting taxes or royalties on timber sold to or used by loggers and industries. Even though the taxes are low and are levied on only a small proportion of the timber cut and used, foresters are often THE MANAGEMENT OF NATURAL FORESTS 355 involved in few other forest operations, artd often equate the introduction of licensed (even if not well controlled) logging with the introduction of man- agement. Similarly, writers in nontechnical publications sometimes observe that the .presence of government foresters often presages logging. Thus a popular weekly science magazine described the rapid intensification of un- controlled logging in Sarawak, in anticipation of a ban on log exports, as "an outbreak of 24-hours-a-day forestry" (Pearce 1990). Harvesting alone do.es not constitute management, although manage- ment may include harvesting. On the contrary, forests can be and often are managed without logging and even without harvesting of any kind, where management is aimed at biological conservation, watershed or soil protec- tion, research, or tourism. Management may also include silvicultural treatments, or systems di- rected toward ensuring a future harvest. The systems may sometimes be elaborate, entailing controls of sizes, volumes, intensity, and frequency of harvests and specific felling cycles and rotations, as described in the forest management literature. But silviculture bears the same relation to forest management that a farmer's activities (such as milking the cows or plough- ing the fields) bear to farm management. Here, forest management is taken to include both enterprise and re- source management. As an enterprise, forest management involves mobiliz- ing human, financial, and material resources to achieve chosen objectives relating to the forest. The technical elements of silviculture, inventory, and harvesting are ways of achieving some objectives, but other elements of management are important for two simple reasons: . l. In the past, the abandonment or failure of management of tropi- cal forest has seldom been due to technical or ecological factors, but rather to failures of enterprise management (e. g., planning, organizing, staffing, controlling) as well as to broader social, eco- nomic, administrative, and political influences. 2. Much effort has been devoted to improving technical abilities, but it is more important to improve management so that existing technical expertise can be applied more effectively. MAJOR CONDITIONS FOR SUSTAINABLE MANAGEMENT Sustained yield or production depends mainly on technical and ecologi- cal elements, whereas sustained management depends on the powers of the manager and on other elements beyond the manager's control. Four con- ditions have been identified as being necessary for sustained management of long-term goods and services from natural tropical forest: long-term security of operation and tenure, operational control, a suitable financial environment, and adequate information. 356 MANAGING THE WORLD'S FORESTS 1. Long-term Security of Operation and Tenure The main reason forest management has failed in the past century in the tropics has been the lack of any guarantee that forest would remain as forest. The absence of security of tenure discouraged forest managers from investing time and money in management for future production and often led to such investments' being lost. Some examples: • The conversion of managed, regenerating, lowland dipterocarp forest in Peninsular Malaysia from the mid-1960s to perennial cash crops, mainly rubber and oil palm, was as much a political decision to pro- vide more for the land-short Malay farmers as a rational decision based on land capability surveys and market predictions. • The abandonment of a concession system for forest industries in Ec- uador was due to the inability of government to prevent transrnigrants from clearing the selectively logged forest for short-term agriculture. • In Queensland, the system that had belatedly become the best man- aged and documented and most researched in the world for manage- ment of tropical moist forest was closed down by a political decision resulting from a state-federal struggle for supremacy and from a fail- ure of communication between foresters and conservationists. A national, regional, or even local survey of land capability should provide an essential factual basis for decisions about the location and extent of the permanent forest estate. The survey must, of course, be combined with estimates of future demand for the various goods and services that can be obtained from the forest and alternative forms of land use. The arguments for the permanent forest estate must be politically co- gent and understandable as well as based soundly in facts. Several parties need to be convinced: • The economic and planning ministries of governments, that the goods and services to be derived from national forests and forest lands are of high importance for the future well-being of the country as a whole and of parts of the whole; • Local populations (voters), that they can derive greater benefit from well-sited national forests than from alternative forms of land use; and • Environmentalists, that there is benefit in managing some forests for production and that this management can be carried out in a sustain- able and environmentally acceptable manner. A major part of the argument must concern the valuation of the goods and services from the forest in terms that are generally understood and appreciated. Some services, such as biodiversity, may be hard to value in cash terms, but foresters must learn to use the growing literature on the valuation of intangible benefits. Equally, managers must review the val- uation by nonforesters of conversion options for forests; overoptimistic valu- 1---- THE MANAGEMENT OF NATURAL FORESTS 357 ation of perennial cash crops, which has been common, has frequently re- sulted in unwise loss of forest. For crops that are grown largely for export and for nonessential markets, the comparative economic advantage for any one country may last for only a few years. Jn 1988, the International Institute for Environment and Development (IIED) undertook a survey of forest management operations in 17 producer countries to measure sustainability of timber production in tropical forests for the International Tropical Timber Organization (ITTO) (see Poore et al. 1989). Most of the forest management operations reviewed by IIED were carried out by forestry departments. There is a presumption that govern- ments should apply policies more consistently than organizations that are susceptible to short-term market movements. The fact that some companies have a record of medium-term forest management of their concessions bet- ter than that of the forestry department in their country does not invalidate the general argument. Clearly, though, the national forestry department is not the only management model. Forms of communal tenure and manage- ment, or leasing to communities, have been suggested and are now being tried in a variety of countries (for example, in farmer groups in Quintana Roo in Mexico and in areas of integrated management in Honduras), and similar arrangements have been proposed for parts of Indonesia. New mod- els and management systems are urgently required for all renewable natural resources; these systems must be appropriate to deal with the local realities of population pressure, ownership, and increasing demand for goods and services. The development of these models, and of rational policies, is a task in which all concerned governments, nongovernment organizations, local communities, and donor agencies have an important role. 2. Security for Forest Operations For decades, ample documentation has existed on the terms under which a government might grant the usufruct of a national forest for domestic or commercial use by local communities or commercial interests. This wealth of information is still not sufficiently applied, or it has been misunderstood in the context of particular legal systems. The uproar over the "risk" con- tracts in Amazonian Brazil in the late 1970s was an example of poor promo- tion by the forestry department of a technically suitable system for an area in which the government services are likely to remain weak and overstretched. The control of the harvesting operation is the most important condition for sustainable management after the long-term security of the forest itself. As mentioned earlier, governments should be the most reliable agencies for long-term (intergenerational) forest management, but, paradoxically, gov- ernments are Jess and Jess able to provide such services. Many forestry departments are notoriously weak and understaffed, morale is low, and career advancement often depends on activities in an urban and bureau- cratic (not to say political) environment rather than in the forest. Field supervision is inadequate because transport, fuel, and equipment are in 358 MANAGING THE WORLD'S FORESTS short supply. Education and training are insufficient or inappropriate. Research is out-of-date, irrelevant, or insufficient. In extreme cases a con- scientious or zealous forester may put his own life or the lives of his sub- ordinates in jeopardy by upsetting rackets that have been well established at the expense of the forest. Decades of donor-funded projects to strengthen the institutions of for- estry departments, and more recently, to establish communal forest man- agement, have not produced sustained improvement. The accelerating rate of forest loss in the tropics argues rather the reverse. A possible explanation has been the treatment of such projects in isolation, rather than in the con- text of a review and revision of all government services. In this respect, the continued operation of the Tropical Forestry Action Programme (TFAP) as essentially a sectoral activity is disappointing. It is not clear that the country capacity projects proposed by the independent review of TFAP in 1990 will be any improvement over the attempts of previous decades, because no mechanisms for overcoming intersectoral divisions have yet been devised. If it is agreed that intergenerational management is essentially an obli- gation of government, the case for strengthened forestry departments in the context of forestry development programs and other major overhauls of government structures should be strongly emphasized. 3. Suitable Financial Environment The aphorism that tropical forests are overexploited and underused is an expression of the effects of differential accessibility. Forests remote from commercial markets may continue to supply services that are little recog- nized or valued locally, regionally, or nationally. Forests that are accessible by water or road are subject to all kinds of human pressures. The less visible the owner or manager is, the more the forest is treated as an open-access resource to be mined by the user. It can be argued that if forestry departments had been more commer- cially minded and less concerned about forest botany and ecology, the for- ests would now be in better condition. Thousands of tropical timbers have been tested for their anatomical, mechanical, physical, and chemical proper- ties, yet the national and international markets are supplied in bulk by a relative handful of species. A direct consequence of the failure to market together species in use-related groups has been the harvesting of small num- bers of trees per hectare, the most valuable of which can be sold as single species with no marketing effort. Harvesting of only a few trees does not lighten the canopy or reduce the standing basal area to the point at which postlogging silviculture treat- ments are biologically or economically effective. Harvesting of up to 10 trees per hectare from the mature phase of primary tropical moist forest gives the manager and silviculturist much greater flexibility in guiding the postlogging development of the forest. The small-scale shelterwood system in the Arena Forest Reserve of Trinidad and the large-scale uniform system THE MANAGEMENT OF NATURAL FORESTS 359 operated in the South Mengo forests of Uganda are two examples in which the managers considered effective marketing to be just as important as the silviculture of the forest. An obvious way to raise the intensity of harvesting is to make the operation more expensive. An operator who has to pay more for the right of usufruct under controlled conditions will seek to make the logging more efficient. At present, access to forests, even in the permanent forest estate, is often so absurdly cheap that there is no incentive to extract more than is immediately and easily saleable. A simple management charge, such as an annual fee per hectare used, is easy to assess and leaves the field staff entirely free from corruptible influences. Royalties and other fees can be collected by the appropriate tax-specializing government bodies but should not be part of a forestry department's duties. A management charge is more easily retained by a forestry department than less rational fees that govern- ment treasuries can claim. Skepticism about government willingness to im- pose such charges, and about the application of such fees to forest manage- ment, are understandable in view of the record of tropical governments (colonial and independent) in treating their forestry departments almost exclusively as earners of central revenue. But that situation has been as much the result of inept forestry department presentation as of treasury greed. Higher valuation of the forest, with consequent better prospects for management, can be induced by other controls and fees, including: • an obligation to harvest trees of technically suitable but commercially lesser-known timber species; • combined or sequential harvesting of nontimber or nonwood prod- ucts, such as fuelwood from timber tree residues and noncommercial timber species, rattans and other climbing plants, and exudates and forest foods; • differential royalties on timbers of different values, calculated on the basis of prefelling inventories with rebates for allowable defect, and adjusted frequently on the basis of market prices; and • more-efficient capture of forest rent, through systems that do not rely for their application on poorly paid and remotely posted forestry department field staff, who are vulnerable to bribery. 4. Adequate Information Most of the 17 ITTO producer countries covered by the IIED survey in 1988 (Poore et al. 1989) showed critical deficiencies in knowledge about their forest resources, the likely markets for their tangible products, and the information required for effective management. It is perfectly possible to practice a conservative management of the tropical moist forest without having detailed silvicultural knowledge of the behavior of the component species. Such management, based on direct observation by staff with a good 360 MANAGING THE WORLD'S FORESTS biological education and an eye for trees, was practiced in India and Burma from the 19th century and was applied in many other countries in Africa and Asia, and to some extent in the Caribbean. With the exception of Suriname it was scarcely tried in Central and South America. When the desired output is large logs of a small number of species, the yield per hectare from polyspecific forest is, of course, low and the cutting cycle length is long. If the species are sufficiently fast-growing and easy to raise in nurseries, the productivity can be raised by enrichment planting; the increased yield per hectare may justify the increased cost. If the species are amenable to growth in open plantations, the yield per hectare may be further increased. However, the growth rate of individual trees may differ little from one system to another. The difference in yield comes from the number of harvestable trees per hectare. The majority of commercial species from the tropical moist forest today have large-diameter crowns at matu- rity, and there would be little advantage in establishing plantations at the necessarily wide spacings, even if the funds were unlimited. The swing away from polycyclic to monocyclic management regimes, from the late 1940s to the late 1960s, stemmed largely from a realization that repeated cutting without intensive control of exploitation damage would harm the forest. However, logging companies have often prevented na- tional forestry departments from including even minimal damage-limita- tion clauses in logging licenses. The foresters' solution was to encourage a single heavy felling and then close the forest while the next generation grew up from seedlings and small trees. For this system to be effective a sufficient population of desirable species had to survive the logging operation and to produce a commercially adequate stocking for the second and subsequent cycles. To provide more uniform conditions for the second generation, those areas not opened up by the logging were often given a climber-cutting and poison-girdling treatment to remove large trees of undesirable species. The small number of experiments undertaken on established crops indicated that the initial thinning undertaken part of the way through the growth cycle would be commercially rewarding at the end of the cycle. Since the 1970s, some countries have suggested that the market accept- ability of species is changing so quickly that, instead of poisoning the trees left after the first systematic harvest, harvesters should leave them to en- hance the yield or to bring forward the time of the second harvest. This reasoning implies a more heterogeneous yield at the second cutting, in terms of the species composition, and a probable delay in the maturation period of the desirable species, because the available growing space will be partly occupied by currently undesirable trees. The potential of this management scheme could be tested by studying the residual stand structure after log- ging and applying known rates of growth and mortality. Such a study does not appear to have been carried out, except in Queensland (Vanclay 1989), but the argument has been used to stop almost all silvicultural work in Sabah and elsewhere. THE MANAGEMENT OF NATURAL FORESTS 361 Similar simulation studies could be used to test the selection manage- ment system (a selection among silvicultural rather than management sys- tems) now supposed to be in force throughout Peninsular Malaysia except Johore (Chiew and Rashid 1989). Unlike the preceding Malayan Uniform System, the selection management system relies on the survival of a large proportion of the trees left after the first logging and their ability to form a sufficient commercial crop for a second cutting at about half the rotation age. The Forest Department in Peninsular Malaysia is still codifying the field rules for the application of this system. The idea appears to be based on the Philippine system, which was developed in commercially far richer forest than the Malayan hill dipterocarp areas. Where there are large sets and long runs of reasonably reliable data, simulation studies have also been concluded to forecast future stand struc- tures and composition. Some have been done in Ghana (Alder 1989), where the "selection system" developed in the 1950s was based on inadequate calculations, involving many arbitrary groupings and averagings of data. Other large data sets, such as those held by the major logging companies in the Philippines, could be worked or reworked with modern computer-assisted techniques to study the validity of the silvicultural and management sys- tems that are supposedly in force. Desk studies, although complex and te- dious, are much cheaper and faster than new field experiments. Without doubt there is a paucity of well-designed and properly ex- ecuted silvicultural experimentation in the tropical moist forests, and the lack of reliable data has greatly hampered the forestry departments from developing management schemes. At the same time, the substantial experi- mental data that have been collected in recent years have been analyzed incorrectly or not at all because of a lack of staff qualified to do the analysis. The design of silvicultural experiments is not especially difficult. The han- dling of the inevitably voluminous data has become much easier with reli- able commercial computing systems and programs. As long as the implicit objective is the production of large logs, the time required to obtain suf- ficient data for usable results is bound to be lengthy compared with experiments on short-rotation pulpwood and firewood forests. Forestry de- partments need to explore intermediate analyses, probably using multivari- ate methods, but such analyses will require more intensive assessments than has been customary. Necessary work on crop development in large-scale silvicultural experi- ments should be supported by autecological studies on the component spe- cies. Although it is possible to manage the forest conservatively on the basis of general biological knowledge, more intensive yet sustainable systems will be possible only as more information is acquired about the quantitative responses to treatment from formal experiments. The knowledge required is no different from that for plantation species. The current enthusiasm by biologists and ecologists for "gap" studies (studies of natural or artificial holes in the forest canopy) appears to reflect 362 MANAGING THE WORLD'S FORESTS the time scale of research grants rather than a profound analysis of the missing segments in the data. The most serious lack of information refers to the postestablishment phase after harvesting. Data are needed on the growth and survival of the different species in relation to the characteristics of the individual stems and both local and area competition. The characteristics of the stems include crown position in relation to the local canopy, the crown and stem form, and the types and extent of damage sustained during the initial logging. There are several reasons for the paucity of the data from plots of 10 years and older: In Southeast Asia, many were destroyed during World War IL In both Africa and Asia, many were lost in the upheavals following decolonization. And in Peninsular Malaysia, the conversion of much of the regenerated lowland dipterocarp forest to oil palm plantations removed both the yield plots and their direct relevance to the remaining forest estate, which is now concentrated in the silviculturally more difficult hill diptero- carp zone. The very success of some experiments has been another source of loss, in that pirate logging crews have felled plots because of their clearly superior growth. It is impossible to overestimate the importance of protect- ing the plots in postestablishment forest, because they are critical for the development of models from which sound management regimes will be derived. Even now, many old plots and long runs of data remain but have been insufficiently studied. CRITICAL KNOWLEDGE FOR SUSTAINABLE MANAGEMENT Biologists and ecologists tend to agree that the shortage of usable infor- mation is even greater than foresters customarily acknowledge. The posi- tion papers for the UNESCO-IVIC (the Instituto Venezolanode Investigaciones Cientificas) meeting in Venezuela in 1986 suggested that there was almost everything still to learn (Gomez-Pompa, Whitmore, and Hadley 1991). To counterbalance that depressing catalogue, this section highlights the ele- ments of silvicultural knowledge that managers of tropical moist forests need to have. For simplicity this discussion is confined to the primary management objective of the production of valuable timber on a large scale (tens of thou- sands of hectares) to feed a capital-intensive forest industry. The sheer size of areas under the control of tropical silviculturists or managers of tropical moist forest forces them to accept a high level of heterogeneity in their forests, augmented by the effects of logging operations. This difference in scale surely accounts for much of the difference in approach to problems between the forest manager and the ecologist. The discussion is further limited to the permanent forest estate, that is, the forest that is reserved legally for supply forest products to fulfill the national domestic (and per- haps export) requirements, in accordance with the national forest policy. THE MANAGEMENT OF NATURAL FORESTS 363 The elements of managerial and silvicultural knowledge essential for foresters are as follows: 1. National Forest Policy. This policy dictates the broad outlines of work for the national forestry department. 2. Legislative Framework. This framework places the forestry de- partment and forest operations in their legal context. Unfortunately the forest law is too often set aside by short-term considerations of political and personal pecuniary advantage (i.e., bribery and corruption). 3. Land Tenure. The forestry department may find it hard to deter- mine the true traditional right holders or land owners, or there may be no simple answer, or there may be political interference with the operation of the statute law on land tenure. Note that difficulties caused to the forest manager on these three points are often due to the deliberate setting aside of the law by those people who are able to do so. 4. Specific Objectives of Management. These objectives should take into account the historical demands on the forest and the commit- ments to supply local consumers as well as large industries. The objectives are not set in concrete but should be kept under review. Formal forestry working plans usually require revision every 5 or 10 years. There should be feedback from the elements that follow to ensure that objectives are adjusted to cope with long-term market changes and improvements to the knowledge base for silviculture and management. The objectives may imply either large-scale and year-round operations, or small-scale and perhaps seasonal op- erations, or a combination. Multiple phases of operations allow multiple benefits to be obtained. For example, a large forest enter- prise might remove the big logs with heavy machinery, and a second-stage license might permit local people to remove resi- dues for firewood and to collect minor forest products. · 5. Static Inventory. This element may represent the field in which scientific knowledge has had the greatest effect because electronic computers have taken the drudgery out of sample calculations and data sorting and tabulating. Inventories are now multi- purpose, to suit multiple objectives; but just as there has to be a primary objective, so there must be a primary group of variables to be estimated, which determines the sampling scheme. Sub- sampling examines the regeneration banks which nowadays would include the seed bank in the soil. 364 MANAGING THE WORLD'S FORESTS 6. Properties and Uses of Desirable Species and Their Market Prospects Over Time. In all, or almost all, tropical countries, the number of species tested and found to be industrially suitable by forest prod- ucts laboratories exceeds the number of species actually marketed. The difference partly reflects the strongly conservative marketing of the timber trade and partly indicates the pressure that the for- est industries business can bring to bear on the relevant, usually political, authorities. Market studies by properly qualified organi- zations are needed to facilitate more informed negotiations between buyers and sellers. Subsequent studies examine the properties of juvenile wood of the second-crop species, to see whether an early harvest would provide technically adequate timber. More advanced studies look into the possibilities of mixing species in a single processed product, such as chipboard or paper, as well as meth- ods of adding value by secondary processing (such as overlaying printed films and plastics onto plywood, or selling furniture made from mixed species instead of only fine timbers). 7. Dynamic Inventory. Study should concentrate on the population dynamics of the desirable species. These are defined primarily on the basis of timber properties and secondarily on observed ecol- ogy (growth habit, position in relation to the canopy in appropri- ate seral phases after logging, and growth rate). 8. Diagnostic Sampling. Diagnostic sampling is a generalized and improved form of the various linear sampling methods devel- oped in Malaysia, Nigeria, Sabah, Sarawak, and Uganda. As the name implies, it is used to diagnose the current state of the stand, especially the stocking of seedlings, saplings, and young trees. The manager may then prescribe the silvicultural treabnents suited to its condition and to the objectives. There is now an emphasis on early identification of potential final-crop trees ("leading desir- ables") and a concentration on their liberation from competition. Many tropical countries have a more or less explicit land policy of "Use it or lose it," so there is strong pressure for a forestry depart- ment to reestablish its claim to manage the forest after a logging operation. An early silvicultural treabnent-not necessarily poison- girdling of undesirable species-is thus often politically desirable. However, in addition to making a best-guess interpretation of the diagnostic sampling results and prescribing treatments, forestry deparbnents should establish the scientific and productive worth of the "leading desirables." 9. Short-term Studies on Dynamics. As in diagnostic sampling, the emphasis is on those desirable species identified in the dynamic inventory and considered together as a crop, rather than on the individual species. Three classes of studies are especially indi- THE MANAGEMENT OF NATURAL FORESTS 365 cated by forest managers: (1) ecological tolerances, determined by sample surveys in space and time as well as by formal experi- ments; (2) response to silvicultural treatments following diagnos- tic sampling, including response to the major intervention, which is usually the logging operation itself;' and (3) effect of actual and simulated logging damage to standing trees and to the soil. The emphasis should be on experiments rather than on ob- servation. Hypothesis testing is rarely necessary; it is the quanti- fication of the response that is needed. Results of the trials should be incorporated into growth models (based mainly on yield plots discussed next) whose sensitivity is often determined by the qual- ity of data at the extremities of the site/ silviculture/ growth re- sponse surface. 10. Long-term Studies on Dynamics. These studies are intended to provide the bulk of the data needed for modeling growth and yield-control systems. The main data source should be a well- stratified and regularly remeasured system of yield plots; this source should be supplemented by occasional inventories to im- prove spatial coverage and to provide for incorporation of routine diagnostic sampling results. The high rate of turnover that is now known to apply to natural forest previously thought to be very stable applies also to the "leading desirables," and makes model- ing virtually impossible from data collected only from leading desirables. There appears to be no satisfactory alternative to re- cording data from yield plots for all stems of desirable species over a defined minimum size. Much more effort needs to be in- vested in growth studies and modeling, because the failure to make secure predictions of future yield makes forestry depart- ments vulnerable to arbitrary political decisions concerning log- ging operations. Three classes of studies would be particularly helpful to forest managers: (1) critical conditions for pollinators, dispersers, and predators of the crop-tree species, and the interac- tions of each with the trees as well as with each other;2 (2) dynam- ics of the regeneration banks (seed, seedlings, saplings/poles); and (3) the ecology and treatment of climbers (lianes, trepaderos), particularly those of silvicultural importance such as Merremia in southeast Asia and the southwest Pacific and Acacia ataxacantha in 1 Failure to quantify the pre- and postlogging states of the forest before the appli- cation of experimental silvicultural treatments has resulted in a regrettably large number of useless and uninterpretable experiments. 2 Tree species that may have no commercial value may nevertheless play a vital role in sustaining the forest ecosystem, for example, by feeding pollinators or dis- persers during seasons when the commercial species are not flowering or fruiting. 366 MANAGING THE WORLD'S FORESTS western Africa. Several silvicultural schemes have foundered be- cause the regeneration of desirable species needed a canopy open- ing to grow out of the seedling stage, but the greater illumination of the forest floor also stimulated the growth of smothering climbers and tanglers. Clearly, work was needed on the autecological re- quirements of the climbers, their pests and diseases, and their response to selective herbicides. Some managers would add a requirement for studies on nutrient cy- cling. However, because 4 cubic meters per hectare per year is about the best commercial growth rate that could be obtained for naturally regener- ated timber species in lowland tropical moist forest, and because 10 cubic meters per hectare per year seems to be about the rate above which some artificial fertilization would be necessary to sustain yields, nutrient studies are not likely to be urgent. Routine diagnostic samplings may indicate that some areas of forest have fewer than the minimum number of polential crop trees in the regen- eration. Depending on how the critical stocking levels are set in the diag- nostic sampling instructions, the solution may be to enrich the forest with line plantings or to replace it entirely with arllificial plantations. Plantations may also be prescribed if demographic pressure or market demand increase so much that the natural forest must give way to a more directly productive form of land use. Conversion to plantations does not necessarily indicate the failure of management; the forest may have been damaged by natural or human forces before it passed to the control of the forestry department, or the demand may exceed the biological capacity of the natural forest to pro- duce the desired materials, or the trees that regenerate abundantly may happen to be of little commercial importance. Departures from the managerial process just outlined are caused by social, economic, and political pressures. Plantations have suffered as much as tropical moist forests from budget failures and land tenure problems and maybe more, because plantation forestry is the management of intentionally unstable systems and requires timely interventions to prevent the collapse of the system. In summary, although ecological knowledge implicitly underpins forest management, it is only one of a number of factors influencing tropical moist forest management. Ecologists might have more influence if they interpreted their research in terms of potential impact on management, whereas manag- ers should articulate their research needs mo:re clearly and phase them into grant-size products as understood by ecologists. RECOMMENDATIONS TO THE DONOR COMMUNITY This chapter concludes with five recommendations to the donor com- munity for better management of forests. There is nothing new in these recommendatiohs. They have been made repeatedly to and by various inter- THE MANAGEMENT OF NATURAL FORESTS 367 national organizations, especially since the establishment of the Interna- tional Tropical Timber Agreement in 1983. 1. Improve Coordination of Effort. For most of the past 40 years, the FAO' s Forestry Department has been the international body executing the largest group of tropical forestry projects. Despite recent vicissitudes it remains the largest core of professionals for tropical forestry gathered in one organization. The FAO Forestry Depart- ment houses the coordinating unit of TFAP, whereas meetings of the in- formal TFAP Forestry Advisers' Group provide a forum for discussion of problems and donor coordination. The World Bank was a prime mover in the call for action leading to the current version of TFAP and has continued to play a major role. Most major donors have publicly declared their inten- tion to provide the bulk of their aid for tropical forestry through the TFAP process. The charter of the International Tropical Timber Organization, com- posed of countries that participate as producers or consumers in the inter- national trade in tropical timbers, gives this organization a responsibility to promote forest management, and ITTO has developed its own portfolio of projects. ITTO is represented in the informal TFAP Forestry Adviser's Group, but neither the Group nor FAO has status with ITTO. The ITTO secretariat remains minuscule. It is still establishing a mechanism for technical reviews of project proposals submitted for approval, and for monitoring and evalu- ating the field progress of projects. It can be argued that FAO's promotion of TFAP in more than 80 coun- tries has greatly exceeded its managerial capacity and that there is therefore every reason now to look to other bodies to fill gaps in tropical forestry development. It is not clear how ITTO's relative freedom from bureaucracy and its flexibility, compared with FAO, will result in more efficient use of resources. In principle, the long-established FAO is better placed to provide technical support to forestry development projects .. The International Insti- tute of Environment and Development has pointed out that overlaps in mandate between FAO and ITTO are potential sources of confusion and that "it is clearly the responsibility of those governments which participate in the governance of both organizations to decide what the balance should be." The donor community should press for coordination and demarcation of responsibilities between FAO and lTTO. 2. Promote More Reliable and Abundant Flows of Information on Techniques for the Management of Natural Forest. Donor agencies such as the World Bank and FAO face campaigns from pressure groups that concentrate on one or a few issues. Globally valid responses are difficult because of the site-specific nature of most forestry activities. The "wildness" of the genotypes and the long periods to maturity oblige foresters to work with the existing environment; the value of the 368 MANAGING THE WORLD'S FORESTS products and services will rarely be sufficient to justify large expenditures to modify environments. This situation is unlike that of agriculture, in which long-domesticated genotypes are raised mostly in semiartificial and con- trolled environments, relatively speaking, and consequently generalizations can be made more easily. Nevertheless, researchers have collected much reliable information on the conditions and techniques necessary for successful tropical forestry. People unfamiliar with the development of tropical forestry tend to treat particular silvicultural systems as management systems; they also do not understand that the silvicultural systems were developed over decades, usually from field observations rather than experiments, and that modifications were frequent. Because education and training in tropical forestry in developing coun- tries are generally so poor, and because the colonial textbooks are either long out of print or in need of revision, a number of these countries are unfamiliar with the battery of techniques at the disposal of foresters for the development of management systems. The donor community could greatly improve the situation by sponsoring the preparation of: • an authoritative state-of-knowledge report on the management of natural tropical forest (an updating of the 1978 UNESCO /UNEP I FAO report on tropical forest ecosystems with a narrower focus); • a popular version of the state-of-knowledge report for use by nongovernment organizations, the executive directors of the World Bank, journalists, and others;' • thorough, standardized, and impartial case studies on tropical forests that are claimed to be under management for sustained yield, and a selection of management systems that are no longer in use but were (in their time) considered to be sustainable. These case studies would seek to identify the principal features that make the systems appar- ently sustainable and those that appear to be of general application. ITTO is interested in such case studies and may provide funding in the future (ITTO 1990a); • a set of technical handbooks or manuals for practitioners of tropical forest management, on the techniques needed by managers. The guide- lines developed by IIED and adopted by ITTO are a useful beginning. The intention that individual producer countries should adapt the ITTO guidelines to their specific circumstances is, of course, com- mendable, but most countries will require a more extensive and solid basis for such conversion. The technical handbooks are especially necessary because of the frequent calls from urban-based foresters in the developing countries for simple rules for field work. Such sim- plicity is incompatible with the diversity of the natural forest, unless 3 Earthscan published a slim briefing booklet by Catherine Caufield (1982) for a similar audience before Alfred A. Knopf published her book on tropical moist for- ests (1985). THE MANAGEMENT OF NATURAL FORESTS 369 the basis for the simplicity is both sound and well understood. Both FAO and ITTO are interested in sponsoring or cosponsoring such manuals but currently lack funds. 4 3. Plan and Execute Forestry Development Projects in Context. A major cause of the failure of many forestry projects, as well as routine work, in tropical countries has been the failure to plan and execute the work in the social, cultural, institutional, and political context. Forestry has been treated as something apart from. the normal life of a country, perhaps be- cause of the relatively long periods of maturity of most of its products. This sociologically simple approach may be adequate in these countries that have a well-developed national consciousness of the importance of the protective and productive functions of forestry, but such an approach is scarcely likely to be successful in countries where the forest has traditionally been viewed as an obstacle to development or as an open-access and mineable resource. The potential benefits from forests and forestry are not self-evident to most people; foresters have to promote them actively with decision makers at all levels of society, from the president to the individual farmer. To avoid unreasonable expectations, foresters must present the biological and other limitations of the benefits honestly. The futility and impermanence of forestry development projects that are not set in context have been repeatedly demonstrated. Projects that are technically overambitious or socially inadequate can be disastrous for the forest, especially when the project planner has one major aim while the operating logger or national treasury has another. Nongovernment organi- zations have rightly criticized many of the project profiles included in national forestry action plans for the industrial development. of primary forests. The critics note that the plans and profiles do not provide for the establishment or reinforcement of social and legal and institutional measures necessary for the goal of sustainable management of the forest. Project proposals that include forest management by local communities may be better placed contextually than the more traditional projects de- signed by forestry departments and forestry enterprises, but claims by nongovernment organizations that only the forest- and forest-fringe-dwelling communities are competent to manage natural tropical forest should be treated with skepticism. Tropical forestry, as a function of government, was started precisely because local people were not always good guardians of the natural resources. The nongovernment organizations appear to think that the mere existence of forest demonstrates a positive desire on the people who dwell nearby to maintain the forest. The local people may have spared the forest simply because they have so far lacked market incentives or heavy technology to log and clear the area. Certainly, some groups have a special 4 The ITTO Action Plan (1990b) mentions the development of manuals setting out "best practices" in relation to most activities involved in the sustainable manage- ment of natural forests. 370 MANAGING THE WORLD'S FORESTS affinity for the forest and some rely totally on its continued existence for the maintenance of their hunter-gatherer way of life. However, the ancient sys- tems of traditional, indigenous management can now seldom be reintro- duced or maintained unchanged for long in the face of changes in social, political, economic, and even traditional religious systems. Some financial or technical support, as well as environmental guidelines, are usually needed to ensure long-term success and security. 4. Use the Criteria Established by the IIED for Sustainable Management. To demonstrate the sustainability of a production system, even by the simple criteria proposed by Dawkins (1988), referred to earlier, requires perhaps a half-century for confirmation of success. Land-use decisions, es- pecially those pressed by burgeoning human populations, cannot wait so long. The conditions that must be met for sustainable management to suc- ceed are as follows (Poore et al. 1989, 24-26; ITTO 1990c; and Caufield 1982, 207-08): • As part of an overall land-use plan, the government must establish a permanent forest estate with a guaranteed long-term future within the natural forest. • The government must create secure conditions for the managers of the forest. • Governments must set standards for allowable annual cut, cutting cycles, tree marking, harvesting techniques, environmental safeguards, and the like. • Those people and entities that are directly involved must adequately control all aspects of harvesting and the treatment of the forest after harvesting to ensure that future crops are assured and that no unnec- essary environmental damage occurs. • Economic and financial policies must not demand more from the forest than it can yield sustainably .. (This, condition requires a market, a government policy that treats forests as a resource to be managed not mined, and a reasonable distribution of revenues and profits among the various parties involved-government, managers, loggers, local communities, processors, exporters, etc.). • Environmental policies must satisfy a public that is becoming in- creasingly conscious of environmental problems. Other factors that are necessary for the effective operation of all the foregoing conditions include the following: • The government must clearly define its objectives for the future of its forest lands and of timber as a commodity. • Information must be available on the extent and quality of the for- ests, their soils, the quality of their timber, and their environmental values, to enable the firm definition o:f a permanent forest estate for timber production and the associated protection and conservation of forests. THE MANAGEMENT OF NATURAL FORESTS 371 • Growth, yield, and regeneration data from permanent sample plots must be available to facilitate development of models that can be used to determine the pattern of harvesting, the detailed marking of trees for felling and for retention, the silvicultural system to be ap- plied, the length of the cutting cycle, and the nature of the future crop. • Accurate financial data covering all aspects of the operation must be available. • The best available forecasts of future markets and demands must be carried out. • Planners must have knowledge of the environmental impact of tim- ber operations and of the views of any local communities likely to be affected by them. These criteria implicitly recognize that the diversity of the natural for- ests requires flexibility in managerial treatment. The donor community should insist that management procedures in forestry development projects be based on the biological possibilities of the forest, not on some preconceived idea that, for example, the Malayan Uniform System or West African Tropical Shelterwood system should be applied as a set of rigid rules. Indeed, the most durable management and silvicultural systems have been character- ized by their roots in reality and flexibility. In reporting on the IIED survey mentioned earlier, Poore et al. (1989) suggested that five levels of management might be conducive to sustainable production: wait and see, log and leave, minimum intervention, stand treat- ment, and enrichment planting. The need to intensify production above the level sustainable by natural forest might legitimately require conversion of forest to plantations or other use. Some benefits of retaining natural forest would, of course, be lost, and should be justified in terms of the demand for other products and services. 5. Use Forestry Development Projects as Demonstrations of Good Management. The IIED book (Poore et al. 1989) recommends several ways to promote sustainable management. In our second recommendation, we suggested that various written texts on forestry management need to be prepared, but there is no doubt that field demonstrations are much more convincing to many people. Numerous bodies are now promoting or executing projects intended to demonstrate good management, but many of the proposals submitted to ITTO make little provision for the efforts needed to create and promote a demonstration, over and above normal commercial levels of management. For a demonstration to be worth mounting, extra staff and facilities are needed, as well as more intensive monitoring of all aspects of forest man- agement. Line-management foresters could learn much from the managers of national parks and similar conservation areas in the public presentation of their activities. 372 MANAGING THE WORLD'S FORESTS Poore et al. (1989) found that less than 1 million hectares of natural forests were under forms of management that met the IIED criteria listed earlier. This low figure contrasts with the official government views, re- layed by FAO as a result of the 1980 global forest survey, that about 40 million hectares were managed. The FAO figure includes countries that are not signatories of the 1983 International Tropical Timber Agreement, but, even so, the discrepancy is striking. Duncan Poore (1990) has pointed out that the figure of 1 million hect- ares has been widely misunderstood and that several countries have sub- stantial areas of natural forest that do not quite qualify as being under sustainable management. He continues, "A small additional effort would bring the forest under sustainable management. Most logged ares could also be brought into sustainable production if the same conditions were applied." Most logged forest has been converted to other forms of land use because of the widespread belief, from farmers to governments, that man- agement of forest is not sustainable. The reasons for this belief, which have been exhaustively analyzed in the past decade, are not repeated here. Per- haps demonstrations should be concentrated in these areas. It does not follow that external support should be restricted to the rela- tively small areas of logged forest or to regenerating forests on abandoned farmland (both types qualifying as secondary forest). Most of these areas are still in the early stages of their first felling cycle after the harvest in the primary forest and thus are not nearly ready for cutting again. By all means they should be brought under management, but that does not imply that they can be logged again immediately. The maintenance of forest industries is a legitimate national concern. It would be unreasonable to restrict donor agencies from supporting either the industries or the development of forests for the supply of raw materials to those industries. The proviso, of course, is that externally fiminced projects are managed for sustainable production. It is reasonable to conclude with a word on the importance of expanded research and training. Sustainable increases in productivity in tropical moist forest depend, as in every other field, on valid research, maintained over the whole timber rotation. No agronomist is expected to predict maize yields from the first 10 days of a 90-day crop, nor can foresters predict yields of 90-cm diameter trees after only 10 cm of growth. However, much more could be done with existing information if the forestry departments re- cruited their staff from a wider range of disciplines. The general principles of forest management have not changed since organized tropical forestry began more than 130 years ago. Unfortunately, neither principles nor techniques are adequately taught to the foresters in many countries with tropical moist forest. Foresters themselves can redeem this situation and should not look for scapegoats outside the profession. As pressures increase for more intensive use olt the land, the foresters must become more skilled and more adaptable if they are to save the forest on which their livelihoods depend. All countries that aspire to retain their natural tropical forest must improve their direction of forestry operations and their development of career foresters. THE MANAGEMENT OF NATURAL FORESTS 373 REFERENCES Alder, Denis. 1989. "Natural Forest Increment, Growth, and Yield." Seminar notes for the national seminar on the Ghana national forest inventory, Kumasi, March. U.K. Thame: Silviconsult (U.K.)/ Booker-Tate Agriculture International Ltd. Caufield, Catherine. 1982. Tropical Moist Forests: The Resource, the People, and the Threat. London, Eng.: Earthscan Publications. _ _ _. 1985. In the Rainforest: Report from a Strange, Beautiful, Imperiled World. New York: Alfred A. Knopf. Thang Hooi, Chiew and Abdul Rashid bin Mat Amin. 1989. "Forest Management Systems in Malay- sia." In FAO Forestry Paper No. 89, pp. 90-145. Dawkins, H.C. 1988. "The First Century of Tropical Silviculture: Successes Forgotten and Failures Misunderstood." In The Future of the Tropical Rain Forest, ed. Melanie J. McDermott. Proceedings of an international conference at St. Catherine's College, Oxford, June 27-28. Oxford, Eng.: Oxford Forestry Institute. G6mez-Pompa, A., T.C. Whitmore, and M. Hadley, eds. 1991. Rain Forest Regeneration and Manage- ment. Man and the Biosphere series, vol. 6. Paris, France: United Nations Educational, Scientific and Culrural Organization (UNESCO), and Carn.forth, Eng.: Parthenon Publishing Group. International Tropical Timber Organization (ITTO). 1990a. "Elements of the Programme of Work: Reforestation and Forest Management." In ITTO Action Plan: Criteria and Priority Areas for Programme Development and Project Work. Document No. ITTC(IX)/6, Rev. 1, November 22, Appendix 2. Yokohama: International Tropical Timber Organization. _ _ _. 1990b. ITTO Action Plan: Criteria and Priority Areas for Programme Development and Project Work. Document No. IIIC(IX)/6,.Rev. 1, November 22, Yokohama: International Tropical Tim- ber Organization. _ _ _. 1990c. "ITTO Guidelines for Sustainable Management of Natural Tropical Forests." Docu- ment No. PCF(Vl)/16, May 21. Yokohama: International Tropical Timber Organization. Pearce, Fred. 1990. ''Hit and Run in Sarawak." New SciCntist 126 (May 12):46-49. Poore, Duncan, Peter Burgess, John Palmer, Simon Rietbergen, and Timothy Synnott. 1989. No Timber Without Trees: Sustainability in the Tropical Forest. A study for International Tropical Timber Organization. London, Eng.: Earthscan Publications. 1990. Personal communication. UNESCO/UNEP/FAO. 1978. "Tropical Forest Ecosystems: A State of Knowledge Report. Natural Resources Research XIV. Paris: UNESCO. Vanclay, J.K. 1989. "A Growth Model for North Queensland Rainforests." Forest Ecology and Man- agement 27 (March 4):245-71. 14 Plantation Forestry P.J. Kanowski and P.S. Savill with P.G. Adlard, J. Burley, J. Evans, J.R. Palmer, and P.J. Wood P lantation forestry is largely a recent phenomenon; most major new plantings have occurred since 1950. Prior to this period, trial plantings and some industrial plantations had been established in both temperate and tropical regions, but only a few countries had adopted plantation forestry as an important means of implementing forest policy. Lack of significant suit- able natural forest resources was the principal reason in places such as Australia, New Zealand, South Africa, and Britain. Despite the limited need for plantations among previous generations, important lessons from early trials provide the basis for today's successful use of tree plantations. The main conclusions underpinning modern planta- tion forestry are as follows: • Introduced species, especially pines and eucalypts in tropical and subtropical regions, are generally more successful in plantations than native ones if the native flora does not contain these genera. Intro- duced species often have the additional benefit of being largely free of serious pests and disease at least for the initial generations, after which breeding and management techniques can be expected to main- tain acceptable equilibria. This issue is analyzed in some detail by Zobel et al. (1987). • Successful species introductions require careful matching of climates between native habitat and recipient region, especially regarding tem- perature extremes (e.g., occurrence of frost and amount and pattern of rainfall). Selection of suitable provenances usually greatly improves the potential of a species. 375 376 MANAGING THE WORLD'S FORESTS • In general, colonizing species that require light are most successful in plantations, as plantation conditions simulate quite closely their natural growing conditions and tendencies to grow in pure stands. Many slower-growing hardwoods are difficult to propagate and to cultivate in monoculture. • In the tropics, pest and disease problems prevent most valuable hard- woods from growing consistently well in plantations; a notable ex- ception is teak. Most species in the important mahogany family (Meliaceae) have proved impossible to grow in plantations, despite repeated attempts to do so both inside and outside their native ranges. • Most tropical grasslands, including those derived from cleared tropi- cal moist forest and subsequently converted to agriculture and then abandoned, are suitable for plantation establishment provided proper attention is paid to weed control and nutrition. Such sites include those dominated by Imperata cylindrica. • At an early stage of plantation development, two silvicultural fea- tures have been found to be critical for success: (1) the need to inocu- late species with the appropriate soil microorganisms (e.g., pines with mycorrhizae, legumes with Rhizobium, or Casuarina with Frankia) and (2) recognition of the widespread occurrence of certain macro- and micronutrient deficiencies, notably phosphorus and boron. • The principles of management of tree stands-cleaning, thinning, and felling-are much the same in temperate and tropical regions and in managed natural forest and plantations. • Successful plantation projects require considerable time and resources. Stability-in security of land tenure, ownership of the tree crop, and technical input-is essential. Exact figures on areas of plantations are difficult to obtain, but there are roughly 100 million hectares of temperate plantations and 35 million hect- ares of tropical and subtropical plantations (see table 3, statistical appen- dix). Among developing countries, three-Brazil, China, and India-have developed substantial plantation forest resources for industrial, social, and environmental objectives, while other, acutely wood-deficient countries- Lesotho, Swaziland, and Trinidad-now depend almost wholly on planta- tions for their wood supply. It is also difficult to specify the role of plantations in global wood sup- ply. Their contribution has been increasing, in both absolute and relative terms, throughout the 20th century, and is projected to account for up to half of the world's wood supply by the year 2000 (Sedjo 1987). As Kallio, Dykstra, and Binkley (1987) note, however, the contribution of plantations in the developing world to international trade in forest products depends largely on trends in domestic consumption. PLANTATION FORESTRY 377 For a long time, tree planting for industrial uses was the main objective, but since the 1970s the .need to satisfy a wide range of domestic needs (notably for fuelwood, building materials, and livestock fodder) has come to dominate much tree planting in the tropics. The silvicultural principles underlying industrial plantation forestry have proved adaptable to these "new" objectives for tree growing, although, obviously, different tree spe- cies and management regimes may be preferred for particular purposes. Managing systems that comprise both trees and crops requires a more com- prehensive understanding of these systems and their components than we have at present. Because tree planting now includes social and environmental dimen- sions, its importance in rural development has broadened; forestry is no longer a narrow sectoral concern but an important national issue in many countries. Moreover, attention has shifted from the timber or wood in the stem to the many benefits trees can provide, and from a view of plantations as fiber factories to multiple-use forestry, notably agroforestry. These changes have taken forestry from the exclusive preserve of the professional to involvement of local peoples in numerous participatory schemes. Both wood- rich and wood-deficient nations now see plantation forestry as a multi- faceted tool, playing an important part in national development (e.g., Wiersum 1984). These trends are likely to continue into the 1990s with the additional consideration of the role of productive plantations as carbon sinks to miti- gate global warming (e.g., Myers 1989a). The areas required for any appre- ciable impact on atmospheric carbon dioxide levels, however, would be enormous; estimates of 300 million hectares have been suggested (Myers 1989b). The economic, political, and social effects of afforestation on such a scale would be substantial. REVIEW OF BEST MANAGEMENT STRATEGIES This section discusses the principles by which plantation forests might be managed to meet best the objectives for which they were conceived. The definition of forest management has evolved with the development of forest science; although originally implying little more than silvicultural manipu- lation, it is now understood in a much broader sense, encompassing deci- sion making in the use of biological, financial, and human resources. It is therefore apparent that the best management strategy for any unit of plan- tation forest will depend on the objectives of management, the resources available, the inherent or imposed constraints, and the decision-making environment. This discussion is therefore directed to the identification of principles common to the management of plantation forests, which are as- sumed to have wood production as a major objective. Plantation forestry has come to be synonymous with large-scale, mono- cultural, intensive tree growing for the production of industrial wood. This 378 MANAGING THE WORLD'S FORESTS model is typified by the coniferous plantations of both hemispheres, or the temperate and tropical Eucalyptus plantations. The best management strate- gies for such plantations are well developed and sophisticated. Historically, most development agencies have promoted this form of plantation forestry. Muc.h of this enthusiasm can be traced to the original view of the role of forest industries in the attack on economic underdevelopment propagated by Westoby (1962) and the development theories it represented. Since about 1978, development theory and intent have emphasized social and commu- nity forestry (Arnold 1991), although practice seemed often to represent merely a scaled-down version of the industrial forestry approach. This pe- riod has also seen a growing emphasis on the development, institutionaliza- tion, and legitimation of agroforestry. However, plantation forestry and agroforestry have generally been considered to be essentially separate and independent activities. With hindsight, it is apparent that the traditional model of plantation forestry is most appropriate in industrial societies for whom land resources are abundant. It is now acknowledged that, given the circumstances of rural societies and land scarcity prevalent in much of the nonindustrial world, the integration of other crops with plantation forestry is likely to be the only means of successful plantation production (e.g., Sargent 1990). The develop- ment of good management strategies for this form of plantation forestry poses many challenges, some of which cannot be met merely by extension from the more traditional approach. This chapter seeks to summarize the characteristics of trees as crops and the consequent principles of successful plantation management, and to iden- tify the best management strategies for plantation programs. FINANCIAL CHARACTERISTICS 01' TREES AS CROPS In the case of most simple forest plantation crops grown primarily for wood production, the costs-the planning, acquiring, and raising of plant- ing stock; site evaluation and preparation; planting and maintenance-are concentrated in the early phases, and returns from harvesting in the late stages, of the production period. Because financial evaluations based on present-value criteria are strongly influenced by the time between invest- ment and return, this period (approximated for most plantation species by their rotation length) is therefore critical in determining the financial returns from tree growing. The major consequence is that private investment in plantation forestry can probably be expected only in economies that are perceived to be stable over the period of investment (Yoho 1985), and for species and sites from which favorable financial returns are most assured. Governments have, therefore, often been seen to have a leading role in undertaking or facilitating plantation afforestation (Hurditch 1992). How- ever, financial returns from plantation forestry are not necessarily poor: McGaughey and Gregersen (1988) suggested that the return on investment PLANTATION FORESTRY 379 for well-managed, fast-growing, industrial-tree crops should be in the order of 10 to 15 percent, consistent with the 8 to 12 percent internal rates of return reported by Elliott et al. (1989) and Whiteside (1989) for New Zealand Eucalyptus and Pinus radiata plantations. At the higher end of this range, returns from slower-growing plantations are probably more in the order of 2 to 5 percent (e.g., Wilson 1989; Spilsbury 1990), they nevertheless compare favorably with real rates of return from other long-term investments (Leslie 1989). The returns from ariy particular plantation program will depend on myriad biological (e.g., growth rate) and economic (e.g., capital and labor costs) factors, and on the consequent environmental and social costs and benefits. Although the issue of public or private ownership has been hotly de- bated in many countries (e.g., New Zealand, Kirkland [1989]; Britain, Rickman [1991]), it is probably more a matter of politics than economics. The success- ful involvement of both large- and small-scale private investors in planta- tion forestry in many countries (e.g., Chile, Hurditch [1992]; India, Srivastava et al. [1986], Indian Forester [1991]; New Zealand, Elliott et al. [1989]) sug- gests that factors other than the economic characteristics of trees themselves are the primary determinants of the level and source of investment. A se- cure investment environment and sufficient investment capital are probably most critical. Depending on the scale of private firms' or individuals' hold- ings, taxation arrangements that do not discriminate against forestry, or the availability of credit on appropriate terms for tree growing, may be impor- tant in facilitating plantation forestry (Arnold 1983; McGaughey and Gregersen 1988; Brunton 1987). The foregoing discussion assumes that investment decisions are made primarily on the basis of discounted net benefit, which is likely to apply to many large investors. There is considerable debate over the more general relevance and application of this. criterion (e.g., Leslie 1987, 1989; Price in press); in any case, for many small-scale growers, the risks and cash flows associated with plantation forestry are more likely to determine their behav- ior than are evaluations based on net income over a long time period. Schemes by which industrial wood users enter into agreements with private land- holders for the supply of wood are often successful (e.g., Philippines, McGaughey and Gregersen [1988]; Portugal, Cotterill [1992]), and may in- volve the purchaser in establishment and management of the plantations. In such cases, there is usually the added advantage of more efficient and more satisfactory operations, in both environmental and social terms. THE POLICY ENVIRONMENT The decision-making environment for forestry management strategies is largely defined by proclaimed or de facto policies that govern land use, land tenure, forestry activities, and forest industries and markets. The cir- cumstances most favorable to the successful development of forest planta- 380 MANAGING THE WORLD'S FORESTS tions are those in which a coordinated land-use strategy has been formu- lated and agreed, tenure rights are unambiguous and unchallenged, for- estry is treated as more than a residual land use, and forest industries or markets are sufficiently developed and stable to maintain a relatively as- sured demand for forest products. The conflict between legislative and customary or assumed rights has often proved a barrier to the development of forest plantations. A common example is the exercise of common property rights over degraded land typically available for plantation establishment. In these circumstances, lo- cal rights or requirements need to be recognized and accommodated in the design and implementation of plantation forestry projects, and the integra- tion of agroforestry practices into plantation management (e.g., Sargent 1990) is likely to represent the best strategy for plantation forestry development. Although plantations have usually been developed on land owned by the state or forest industries, smaller private landholders have been region- ally important, and are likely to become more so in the future as pressures for land increase. There is a wealth of information describing the character- istics, behavior, and requirements of these forest growers in both industrial and nonindustrial societies (e.g., Arnold 1983; Blatner and Greene 1989; FAQ 1985a; McGaughey and Gregersen 1988). The successful involvement of such landholders in tree growing depends largely on security of tenure over the land, or at least over the trees on it, and on assured access to markets at the time of harvesting. Many schemes by which industrial wood users enter into agreements with private landholders for the supply of wood have been successful. Schemes that involve the purchaser in establishment and management of the plantations usually have more efficient and envi- ronmentally satisfactory operations. Depending on the scale of private firms' or individuals' holdings, taxation arrangements that do not discriminate against forestry, or the availability of credit on appropriate terms for tree growing, also are likely to be important. LAND USE AND MANAGEMENT PLANNING Given a favorable, or at least nonantagonistic, policy environment, re- quirements for the development of successful plantation forestry are well established. The principal requirement is for accurate information describ- ing the land resource. Baseline surveys are necessary to identify the types of soil and vegetation and their distribution, topography, patterns of land use, tenure, and any other information of relevance. Accurate information de- scribing the characteristics of soils relevant to disturbance and tree growth is critical, because it should form the basis for all silvicultural decisions and for plantation and infrastructure design. The development of geographic information systems has greatly facilitated the processing and presentation of this information, which should be mapped at a scale of 1:10,000 or less. These data should form the basis for preparation of a draft management PLANTATION FORESTRY 381 plan for review by appropriate state agencies and by those individuals or groups directly affected by, or with a recognized interest in, the proposed development. ELEMENTS OF FOREST MANAGEMENT Contemporary forest management has a certain dualism: One element is the sophisticated, quantitative management science developed for manag- ing large-scale resources and wood-using industries. Such an approach is characteristic of forest management in industrial societies where method- ologies, usually based on mathematical optimization, have been developed for purposes such as multiple-use planning, harvest schedules, and evalua- tion of silvicultural alternatives. A second element is the recognition of traditional management systems as similarly sophisticated, albeit qualita- tive, and the development of management practices that build on them. Plantations grown solely or primarily to produce wood are usually managed to maximize financial return, subject to constraints such as main- taining a given level of output to maintain industry. Characteristics com- mon to the best management strategies include substantial planning; appro- priate and probably intensive site preparation; establishment with genetically improved planting stock; sufficient fertilization and weed control to pro- mote early growth; and, if products other than pulp are intended, early thinning and pruning to maximize the value of the end product. The devel- opment of the stand is monitored regularly and silviculture adjusted as necessary; fire is managed or excluded; inventory data are used to develop and modify models describing growth and product assortment. In such intensively managed wood-production plantations, there are few opportunities for intermediate returns prior to those from any commer- cial thinnings. One common exception is a return from grazing, particularly from silvicultural systems that use grass cover as a means of limiting ero- sion (e.g., Queensland Department of Forestry 1987). Particularly in the case of short-rotation crops for pulp production, species that can be· coppiced for a number of subsequent rotations may be preferred to those that cannot, although the relative merits of coppice management will depend on its costs and returns compared with those from establishment of new, probably ge- netically superior, plants. In contrast, management strategies for plantations that maximize social benefits rather than wood production per se are still under development. In some cases, a modification of traditional silvicultural principles is appropri- ate (e.g., Gilmour et al. 1990); in others, a synthesis of agroforestry and plantation forestry approaches is desirable (e.g., Sargent 1990); in still oth- ers, replacement of species that produce only wood by those that also pro- duce valuable intermediate returns may offer the best solution. A variety of agroforestry practices (e.g., intercropping, underplanting, or rotational cropping) may be feasible. The relative merits of each will 382 MANAGING THE WORLD'S FORESTS depend on the biological and economic characteristics and interactions of the tree and nontree crops, the sustainable land-use systems with which local people are familiar or which can be introduced, and the risks and cash flows associated with each alternative. The replacement of species whose production is limited to wood often depends on developments in technol- ogy to facilitate multiple uses. A recent example is the progress in the process- ing and marketing of rubberwood (Hevea brasiliensis) in Malaysia, which allow it to be considered as an alternative to more conventional forest plan- tation crops such as Acacia mangium. Whatever management strategy is adopted, it should maintain environ- mental quality and be integrated into local societies and land-use practices. Environmental quality is usually maintained by the definition of operating standards and codes of practice, such as those currently under development for the International Tropical Timber Organization (ITTO 1990a). Integrat- ing the plantation program into the broader social context is somewhat more challenging, although appraisal and diagnosis methodologies (e.g., Formby 1986; Proceedings 1987; Raintree 1987; Molnar 1989) have been developed to offer some guidance. REVIEW OF ENVIRONMENTAL IMPACTS Sustainability The role of plantations in the environment must be judged by the crite- rion of sustainability. Agricultural and forestry systems may be said to be sustainable when productivity is maintained from year to year or rotation to rotation without an increase in the inputs of nutrient and soil working. If the system is causing loss of soil from wind or water, silting and drainage channels, and contamination of rivers, it cannot be said to be sustainable in an ecological sense, even though the market is being supplied with regular quantities of produce. Sustainable production from a site depends on the maintenance of the physical soil, including its organic components and microorganisms. When plantations, or operations in them, lead to a loss of organic matter, a break- down of soil structure, or its removal by wind or water, the system of management is unsustainable and therefore unacceptable. Where the harvest removes more nutrients than are added from the soil and atmosphere, the system is naturally unsustainable, although the nutri- ents may be replaced artificially under specific economic conditions. If the water table is lowered because the trees are transpiring more water than the soil receives in the precipitation, the trees are using up the capital of the site and are therefore not sustainable, unless artificial irrigation is possible. Plantations will affect the hydrological cycle, depending on the natural system they replace, by changing the balance among evapotranspira- tion, interception, runoff, storage, and percolation-and in turn influencing PLANTATION FORESTRY 383 stream flow and climate beyond the plantation area. Undesirable land-use practices contributing to climatic change with uncertain feedback cannot be sustainable. Soil Erosion Undisturbed tree cover reduces erosion, but intensively managed plan- tations may not be good protection forests. They key to soil protection is ground cover, mainly provided by litter; the extent to which litter is pro- vided varies greatly among different plantation species. Eucalyptus, for example, is sometimes thought to provoke erosion as a consequence of the sparse understory found in dense stands. In parts of India the rural poor collect fallen leaves as fuel, leaving a bare and easily eroded forest floor. Not all stands suppress understories, however, and many species produce abundant litter. Lundgren (1978) gave examples of severe surface soil erosion under coniferous plantations in eastern Africa. Undergrowth is suppressed in teak plantations and drip from the large leaves contributes to erosion of surface soil (Champion and Seth 1968). Litter in teak plantations often burns at the end of the dry season, giving a clean floor at the onset of the rains. Much of the literature contains contradictory data, often poorly sub- stantiated. For example, a study in Eucalyptus globulus plantations in Portu- gal (Kardel, Stein, and Fabiao 1984) suggested that E. globulus is well orga- nized to control erosion with, on good soils, a root system consisting of a superficial net of fine roots and deep tap roots to 10 to 15 meters. From the Nilgiri Hills in southern India, however, Mathur, Raj, and Naithani (1984) claimed that the same species "is a shallow-rooted" plant with a root system confined to a depth of about 3 meters. Plantation management is the key to the influence on soil and water conservation of a particular species. Generalizations ignoring the range of variables (climatic, edaphic, and hydrological) are not usually helpful. Har- vesting of plantations on short (8 to 10 year) or very short (3 to 5 year) rotations, as is often done in the tropics, means increased frequency of soil disturbance, destruction of organic matter, soil erosion, and compaction. Where soil erosion is known to be a risk, forestry agencies should give clear guidelines about soil conservation, detailing the specific actions that should be taken. The Queensland Department of Forestry (1987), for ex- ample, states that planting should generally be avoided on unstable slopes greater than 15 degrees and on stable slopes of 20 degrees or greater. Buffer strips of natural vegetation should be retained along streams and selected drainage lines on stable soils downstream from the point where the catchment area exceeds 100 hectares, and on all incised drainage lines on unstable soils. The department emphasizes the important of planting second rota- tions among logging debris, where possible; the need for cultivating along contours; and, where necessary, the importance of controlling runoff, and hence reducing erosion by, for example, sowing crops of millet or oats. 384 MANAGING THE WORLD'S FORESTS Maintenance of Long-term Productivity As in any agricultural system, forestry plantations managed on short rotations require much soil working and the addition of nutrients. When productive monocultures are grown on short rotations, there is always con- cern about the maintenance of yields in successive crops. Experience in some temperate forests, notably in Lower Saxony in Germany and the Landes in France, suggested that yields would decline when soil is poor or species are mismatched. This concern, which was widespread earlier this century, has not been borne out, and the few studies that have been reported (e.g., in Denmark and the Netherlands) have not shown significant changes in yield with successive rotations arising from forestry practice itself (Evans 1976, 1990). In Mediterranean-type climates, a serious decline in productivity was reported in the 1960s for second-rotation Pinus radiata in South Australia and the Nelson area of New Zealand. Most plantations in South Australia were affected, and the typical drop in yield was 30 percent. While pine growing in South Australia was always marginal for climate, principally low rainfall, the cause of the decline between the first and second rotations is now known to result from failure to conserve organic matter at the time of harvesting of the first rotation, and from dense grass competition in the early years of the second crop (Boardman 1988). Indeed, a combination of appropriate silviculture and tree breeding now results in second-rotation stands significantly more productive than the first. Function of Plantations in Reducing Tropical Deforestation Most clearance of rain forests in the tropics is the result of increasing populations and the consequent need to bring more land into agricultural production. The main cause of forest destruction is shifting cultivation. In general, therefore, plantations are unlikely to reduce the levels of forest clearance. They also tend to produce quite different products (industrial sawnwood and pulpwood), so that the opportunities for substituting plan- tation-grown material for the high-quality hardwoods of the natural forests is small. Fuelwood and building-pole plantations can reduce pressure on the natural woodland in relatively arid regions, but it is in these regions that most failures have occurred-mainly for social reasons and lack of technical expertise. Climatic lnfluenice Plantations, like forests of any origin, can have marked effects on micro- climates of neighboring land. Windbreaks and shelterbelts are particularly valuable because they reduce soil erosion by decreasing surface wind-shear stresses, trapping moving soil, and slowing soil drying. Shelter also reduces PLANTATION FORESTRY 385 the rate of evapotranspiration from the leaves of crop plants. Depending on the climate, increases in crop yields of 20 percent to 150 percent are cited in the literature. Among the best-recorded effects of shelter-belts are those on the Great Plains of the United States (where the dust bowl of the 1930s was eliminated), the Danish heathlands, and the Hungarian plains. Shelterbelts are probably just as valuable in the more arid parts of the tropics. It is also well known that the climate inside a forest is markedly differ- ent from that on neighboring open ground. The general effect of tree crowns is to give the space below them a less variable climate than exists in the open: changes in the outside atmosphere are damped down, and corre- sponding changes inside the forest occur only after a considerable time lag. Comparisons of the climate beneath forest canopies with that of adjacent open land usually show the following: • Air and soil temperatures fluctuate much less. • Relative humidity is higher and comparatively stable. • Air moves less. • Light rain does not penetrate to the forest floor. • Little solar radiation reaches the ground. • Carbon dioxide concentrations decrease during the day owing to uptake by photosynthesis (except near the ground, where they are replenished by soil respiration). A topic of much current concern is whether human societies face dis- ruption, and life on earth itself is endangered, by an uncontrolled increase in "greenhouse" gases, particularly carbon dioxide and methane. Because trees are constructed mostly of carbon, plantation forests may also have a role as carbon sinks in mitigating global warming. The complex biological and social issues involved have been extensively reviewed elsewhere (e.g., Grainger 1991), and are further discussed in chapter 5 of this book. CONSTRAINTS. OPPORTUNITIES, AND PRIORITIES Interactions with People An understanding of the interactions between societies and their forest resources is fundamental to any successful forest management policy (van Maaren 1984). Chambers et al. (1989) note that plantation programs have seldom offered participation beyond employment in wage or kind. While pointing out that the success of plantation programs depends strongly on the local sociopolitical environment, they suggest that "where species selec- tion has been right, and benefits assured and direct, people have looked after forests and have cooperated with the Forest Department." Many commentators on the practice of forestry in development (e.g., Douglas 1986; Westoby 1987, 1989) have noted the tendency of foresters to 386 MANAGING THE WORLD'S FORESTS focus on trees and forests per se, rather than on their role in satisfying the people's needs. The ability of governments or their forestry services to ini- tiate or accommodate such changes has been questioned (e.g., Douglas 1983; Leslie 1980), and profound changes in their priorities and practices may be necessary if local needs are to be satisfied sustainably. Land and Tree Tenure Because most plantation schemes are long term and spatially extensive, the tenure of the estate must be clearly defined and legally secure if invest- ment is to be secured. It can often be difficult to determine the true tradi- tional right holders or landowners. Politicians often interfere with the op- eration of the statute law on land tenure or disagree about its interpretation (e.g., ITTO 1990b). Questions about who has de facto or de jure rights to the use of re- sources (e.g., Blaikie 1985; Chapman 1987) are now recognized as funda- mental to the management of natural resources (Byron and Waugh 1988). Resolution of the problems of political economy (as Byron and Waugh [1989] put it, "Whose resource is it?") is therefore a prerequisite for successful establishment, maintenance, and management of plantations. The most ap- propriate tenure arrangements for plantation forestry will vary with the context of the particular program. Corruption In many countries, greed, cynicism, bribery, and corruption are charac- teristic of high officials, who view forests as a short-term source of personal wealth, and aid as a means of personal gain. A lesson that has been learned in many places, including most European countries two centuries ago, is that the value of vanishing forests is never appreciated until their scarcity prompts a crisis that is taken seriously at the highest political levels. Unfor- tunately, as recent history has demonstrated, the lessons from one country are seldom transferred to others; there are numerous examples today of approaching crises, in Africa, Asia, and elsewhere. Politici<1ns pay lip ser- vice to the importance of trees and forestry while becoming wealthy on the illegal proceeds of exploited and often subsequently degraded forests. The scientific and technical approaches to plantation forestry are rela- tively well understood, and the problems are not generally difficult to overcome. Nearly all the constraints relate to people; politicians must be motivated and educated about plantation forestry, while foresters must be paid and equipped. Establishment and Postplanting Management There is a tendency in many countries for plantations to be established and extended too far ahead of scientific observation and research. Any de- PLANTATION FORESTRY 387 velopment is to some extent site specific, and to be successful requires the development, through research, of the site's own unique combination of techniques and their subsequent application. Few plantation schemes re- quire totally new methods or any detailed knowledge of ecology, as .distinct from conventional and prudent silviculture. The early phases of plantation development are critical to eventual suc- cess; robust, vigorous planting stock, be it raised as seedlings or cuttings, is essential. Despite the wealth of knowledge about appropriate methodolo- gies (e.g., Aldhous 1972; Queensland Department of Forestry 1987), poor nursery techniques are responsible for many initial failures. The desirable attributes in the quality of nursery stock are well understood and have been described in detail by Evans (1982) for the tropics, and by Savill and Evans (1986) for temperate regions. Unfortunately, incorrect application of the nursery techniques in the field remains one of the most serious limitations to the success of plantation forestry. Fast-growing trees also require timely silvicultural interventions, sus- tained financing, delegated authority, and an efficient command structure. But failures are common because decision makers set unrealistic targets even when untried species are being used and locally untested techniques applied. One example is the JARI plantations in Brazil {Palmer 1986), but there are many others. Most tropical regions of the world have at least some of the following problems: • Incorrect selection of species and provenances, especially exotics, which have not been adequately tested in the new environment. • Failure to develop suitable silvicultural schedules (ground prepara- tion, weeding~ thinning, pruning, etc.) to achieve the objective of the plantation scheme. • Failure to monitor growth rates, nutritional problems, and health to ensure that the plantation scheme is actually achieving its objectives. This problem is often coupled with a reluctance to seek specialist advice or, if sought, failure to act on it. • Failure to make use of the available literature. In new environments, one can learn expensively through making mistakes, or cheaply by using a good library, where the elements of tropical land develop- ments and plantation forestry are recorded in dozens of texts and thousands of articles (Palmer 1986). Conservation and Amenity Value Plantations.are usually thought of as having relatively low conservation and amenity value (nonmarket utility beyond subsistence needs) (e.g.; Shiva 1987; Tompkins 1989) because of poorly planned afforestation, but some plantation programs have successfully accommodated conservation values. Two essential elements are (1) a prior ecological survey to identify and protect the areas of particular conservation significance and (2) a plantation 388 MANAGING THE WORLD'S FORESTS design that retains a network of natural vegetation and avoids intrusion near to watercourses (e.g., Queensland Department of Forestry 1987; Sargent 1990). Some plantation species totally suppress ground vegetation once the canopy has closed. This situation can impair the conservation of the natural flora and fauna or the integration of other crops with the plantation crop. In Britain, for example, there is much debate about exotic spruce and pine plantations. Conversely, management of some plantations, such as those of Araucaria cunninghamii in Australia, may favor regeneration of the natural flora (Queensland Department of Forestry 1987). The amenity value of plantations is largely determined by cultural per- ceptions and expectations of nature, landscape, and recreation. Some plan- tations are said to have very high amenity values (e.g., Tomkins 1990), but few useful generalities can be made. The more integrated approach to plan- tation design and composition should favor increased conservation and amenity values. Evaluation of the environmental impact of plantation pro- posals, following methodologies, such as those described by Hyman and Stiftel (1988) or Thompson (1990), should be considered a prerequisite for any substantial plantations. Continuity of Staff and Provision of Equipment Many plantations have suitable soils and species and adequate capital but fail because of lack of continuity of staff. No plantation techniques are beyond the ability of reasonably experienced and dedicated foresters, but without such people, the forests cannot be developed and they cannot sur- vive. The forest manager must decide the combination of operations that is required for any particular situation. This skill, which is mostly built up by experience, is worth almost any asking salary. In Germany and France, foresters expect to spend the majority of their working lives in a single forest area, and there are several instances of the third generation of the same family being in charge of the same state forest. But in the tropics, unfortunately, such skill is very rare. Plantation forests also require a relatively long time and technical ex- pertise to achieve high or valuable yields. If forestry officials are badly paid, as they are in many developing countries, and if they lack the vehicles, tools, and chemicals to carry out their work effectively, it is not surprising that success is never achieved or that failure occurs as soon as a donor agency departs. Genetic Resources of Plantation Species The genetic resources of currently or potentially important plantation species are widely distributed. In general, a policy of open access has pre- vailed for most species, although this issue and that of plant variety rights PLANTATION FORESTRY 389 are the subject of increasing debate (e.g., Marshall 1989, 1990). A variety of international or national agencies have established themselves as coordina- tors or agents of the assembly of genetic resources of the major tree genera or species. Some examples are CAMCORE (Central American and Mexico Coniferous Resources Cooperative; Dvorak 1990), CSIRO (Commonwealth Scientific and Industrial Research Organization, Australia; Midgley 1988), CTFT (Centre Technique Forestier Tropical, France; Bellefontaine, Cossalter, and Souvannavong in press), DANIDA (Danish International Development Agency undated), or the Oxford Forestry Institute (Barnes and Burley 1990). In general, these institutions cooperate with local agencies in the conserva- tion of priority species in and ex situ and in the collection, distribution, establishment, and assessment of research quantities of seed. Tree seed are widely traded, and although information and controls concerning their origin and genetic quality have improved (e.g., Midgley 1988), many uncertainties remain for many species about these fundamental requirements for plantations to be established on sound genetic founda- tions. The agencies responsible for seed collection for research purposes can usually provide information about reliable and reputable supplies of com- mercial quantities of seed for plantation programs. The advanced breeding programs developed for many species are usually able to supply large quan- tities of genetically improved seed or propagules, which should be tested on a pilot scale along with accessions from natural populations. Recent ad- vances in bioclimatic analyses (e.g., Booth, Searle, and Boland 1989) have facilitated the selection of species and provenances likely to succeed under particular conditions. It is difficult to overstate the critical importance of appropriate species and provenance selection as the basis for successful plantation forestry; as Zobel and Talbert (1984) stress, "The largest, cheapest, and fastest gains in most forest tree improvement programs can be made by assuring the use of the proper species and seed sources within species." Many early plantation programs, such as those of Eucalyptus in Brazil, Pinus caribaea in the tropics, and Pinus radiata in Australasia, had access in their early stages to only a limited range of genetic material from natural and exotic populations. As more comprehensive collections became available, it has become apparent that many programs developed from suboptimal genetic material, and that plantation health, productivity, quality, and economic viability have suf- fered as a result (e.g., Gibson, Barnes, and Berrington 1983; Ikemori 1990; Moran and Bell 1987; Queensland Department of Forestry 1981). Genetic Improvement Genetic improvement of pure or hybrid species is a feature common to all successful plantation programs. It is helpful to think of forest genetics and tree improvement in terms of the three phases identified by Cheliak and Rogers (1990)-conservation, selection and breeding, and propagation. 390 MANAGING THE WORLD'S FORESTS As a consequence of both their biological characteristics and generally short history of domestication, tree species have high genetic diversity relative to other plants (Brown and Moran 1981; Hamrick 1983), The pressures on re- maining forest resources have resulted in the implementation of numerous coordinated international programs for genetic conservation (e.g., Barnes and Burley 1990; Dvorak 1990; Palmberg and Esquinas-Alcazar 1990). It is also desirable to design breeding strategies that include the maintenance of genetic diversity as an explicit objective (e.g., Burdon 1989a; Kang and Nienstaedt 1987; Kitzmiller 1990; Namkoong, Barnes, and Burley 1980; Namkoong, Kang, and Brouard 1988). Tree-breeding programs based on the recurrent selection cycle described by White (1987) have developed in association with intensive plantation management and therefore have a history of not more than about 50 years. Available information for both angiosperms and gymnosperms under plan- tation conditions suggests that the level of genetic control varies from rela- tively weak, for growth traits, to very strong for disease resistance. Stem quality traits are toward the lower end of the scale, and wood properties toward the higher end. Typical figures for the degree of genetic control would be 10 to 20 percent for growth, 20 to 30 percent for stem quality, 40 to 60 percent for wood properties, and perhaps absolute control in the case of some diseases. In all cases, genetic control and variation are sufficient to permit substantial gains. Typical gains from the first genera tion of breeding are on the order of 30 percent for growth and 50 percent or more in terms of wood recovery. Programs based on short-rotation species have demonstrated that such gains can be maintained over at least several generations (e.g., Reddy, Rockwood, and Meskimen 1986). The increasing use of indices for selecting, and vegeta- tive propagation for multiplying, outstanding genotypes ensures that gains realized by breeding are transferred efficiently to production plantations, although Burdon (1989b) suggested that clonal forestry on an operational scale may be worthwhile only for those taxa that are easily propagated. The additional cost of vegetatively propagated plants varies with the ease of propagation, and is typically a minimum of 30 to 50 percent more than the cost of seedlings (Savill and Evans 1986). The integration of developments in biotechnology into tree-breeding programs offers further opportunities for acoelerated genetic improvement (Cheliak and Rogers 1990; Dommergues 1986). However, the use of these new, sophisticated, and expensive technologies should be seen as comple- menting, rather than substituting for, traditional selective breeding, which is cheap and robust by comparison. The cost of genetically improved mate- rial is a small part-typically about 10 percent-of unit establishment costs; available economic data now suggest that investment in tree improvement as part of a plantation program is well justified (Zobel and Talbert 1984), and empirical evidence certainly supports th:is conclusion. PLANTATION FORESTRY 391 The Threat of Pests and Diseases to Plantations A constant topic of debate is the extent of the risk that pests and dis- eases pose to monocultures. Clonal plantations always present the greatest risk, and with these plantations at least, there is little doubt that selection for pest and disease resistance will become increasingly important in fast- growing plantation species. Poplar-breeding programs in Italy and else- where, for example, have resistance to Marssonia, Melampsora, and Dothiciza as a first step in screening. Only as a second stage are the resistant families screened for growth habit, wood properties, and other characteristics. Among nonclonal plantations, several examples of disasters are quoted in the literature, including devastation by Dothistroma blight of Pinus radiata in Zimbabwe in the 1950s, and more recently the fatal psyllid attack on Leucaena leucocephala in many parts of the tropics. In both these cases and many others, the species had been planted in unsuitable (i.e., stressed) en- vironments, or the advice of geneticists, entomologists, and pathologists had been ignored . .There seems little doubt that with proper care in selecting species, provenances, and families, risks are low. No major plantations have been devastated by pests or diseases. It is perhaps worth noting that most of the spectacular epidemics in forests have occurred in seminatural popu- lations rather than plantations; two examples are chestnut blight in the United States and Dutch elm disease in Europe. Growing High-Quality Timber Many rain-forest species are in great demand by the timber trade but, with a few notable exceptions such as teak, are almost impossible to grow in plantations because of the ecology of the trees. Most species have evolved in highly competitive environments of forests composed of many species and ages of trees. Many species are shade tolerant and will grow only in small gaps in mature forests. Such species are usually ecologically unsuited to growing in pure stands or in open environments when young. Attempts to grow them in such circumstances usually result in chlorosis and decline or in fatal insect attacks, such as Hypsiphila species in the case of Meliaceae. The case of Agathis plantations in Queensland, Australia, devastated by the coccid Conifericoccus agathidis (Queensland Department of Forestry 1987) is typical of many experiences. It is usually unrealistic to expect pure planta- tions of such species to succeed on any scale. If the desired output is a sizable number of large logs from a small number of species that grow in very mixed tropical forests, the only realistic approach to growing many of them may be to carry out some enrichment planting. Unfortunately, experience has also shown that enrichment plant- ing, because of its extensive nature, is seldom successful. The long-term prospects that tropical hardwoods grown in plantations can substitute to a significant extent for the current production of natural or seminatural tropi- 392 MANAGING THE WORLD'S FORESTS cal hardwood forests are, therefore, remote. This reality provides a strong argument for the effective management of seminatural forests, if these spe- cies are to continue to be available in the future. Poore et al. (1989, 237) has suggested that production from natural forests should concentrate on the valuable decorative, joinery, and veneer species, rather than on species for industrial cellulose, which are likely to be better grown in plantations. There is, however, scope for devoting more resources to research on the limited! number of valuable species that appear to have some prospect of success under plantation conditions. Cordia alliodora, for example, has re- cently shown promise as a tree for tropical agroforestry (Greaves and McCarter 1990), and several of the Asian merantis (Shorea species) hold promise for plantations if the problems of seed storage and germination can be overcome, or means of vegetative propagation identified. Several variet- ies of eucalyptus, for example E. pilularis, also provide good veneer timbers. Large areas of well-managed hardwood forests, including plantations, are much more typical of temperate than of tropical countries. They are found, for example, in many parts of Western Europe and in the southeast- ern United States. In anticipation of a continuing decline in and eventual cessation of supplies from the tropics, some European countries have delib- erately decided in recent years to concentrate on the production of high- quality, decorative, native hardwoods, on suitable sites. Plantations or Managed Indigenous Woodland? It is often argued that development of plantations is one means of sav- ing tropical forests, on the ground that growth rates in plantations can be at least 10 times per unit area the rates in natural forests. As Poore (1989, 237) pointed out, though, this saving role is by no means certain. It is often easier to establish plantations on newly cleared forest land where soils are more fertile and there are fewer problems of land tenure, than by planting defor- ested land. In fact, plantations and natural forests both have their place because their purposes and products are different. Even in the narrowest sense of timber production, it is difficult to produce significant quantities of "supe- rior" timbers in plantations. Where discontinuity of both management and finance is a problem, as in most developing countries, many plantations go largely to waste after only a few years of neglect. In such circumstances it would obviously have been better not to have started a plantation scheme at all, and to have concentrated on the management of natural forests, which do not require such high inputs and are more resilient in the face of some forms of neglect. No single approach to forestry, even within quite small regions, is ever likely to work. A variety of management options are usually needed-including a variety of institutiona1 options ranging from a state forest service through community to individual-landowner management. The principal attraction of plantation forests is their productivity rela- tive to that of most natural forest systems (see exhibit 14-1). Most planta- PLANTATION FORESTRY 393 tions have, therefore, been established primarily for industrial wood pro- duction, although Sutton (1991) estimates that only some 14 million hectares can be classified as "fast-growing" (in his terms, of mean annual increment greater than 14 m 3 /ha/year). The relatively small area of plantation forests belies their contribution to world wood supply, which has been increasing, in both absolute and relative terms throughout the 20th century, as planta- tion stands reach maturity. Consequently, plantations currently contribute between 7 and 10 percent of world industrial roundwood production (Gauthier 1991), and that figure is projected to rise to up to half of total world wood supply by the year 2000 (Sedjo 1987). The vast majority of this wood will be best suited to pulp or other relatively low value use (Sutton 1991) and, as Kallio et al. (1987) noted, the contribution of plantations in the developing world to international trade in forest products depends largely on trends in domestic consumption. One of the most important issues for debate in future forest policies of producing countries, especially where a choice exists between plantations EXHIBIT 14-1. Productivity (Annual Wood Increment of Various Well-Managed Types of Forests) Mean Annual Increment Type of Forest (m' /ha/year) Tropics and Subtropics Natural forests Deciduous woodland (range)a 1-2 Rain forest (extensive management) 1.5-15 Plantation forests Lowland hardwoods (range)b 10-20 Lowland eucalypts and confers (range)b 20-35 Montane eucalypts and confers (range)a 30-40 Aracruz, Brazil: Eucalyptus hybridsc 50 Temperate Regions Natural forests Canada: conifers (average)d 1.0 United States: conifers (average)d 2.6 Sweden: conifers (average)d 3.3 Australia: southern eucalypts (average)e 7.3 Plantation forests Britain: broad-leaves (average)d 5.0 Britain: conifers (average)d 11.0 Portugal: Eucalyptus globulus (range)' 3-40 Portugal: Eucalyptus globulus (average)g 13 New Zealand; Pinus radiatae 18-30 soURcES: 'Dawkins (1991); bDawkins (1967); 'Sturm (1991); dSavill and Evans (1986); 'Risby (1987); fPereira and Santos Pereira (1988); gCotterill (1992). 394 MANAGING THE WORLD'S FORESTS and the management of indigenous woodland, concerns the relative roles of plantations and natural forests. This is particularly the case in savanna and more arid regions, where the potential of natural woodland is often ne- glected (Bailly et al. 1982). Extensive plantation areas may have natural or seminatural types of vegetation that are of interest scientifically and need protection-or that need to be retained for landscape, recreational, or soil conservation purposes. In circumstances like those described, for example, by the Queensland Department of Forestry (1987), it is often appropriate to leave strips or connecting corridors of natural vegetation to serve as a buffer zone around a larger gene reservoir needing protection, or to serve a func- tion themselves in, for example, reducing erosion. Plantation forestry requires much more initial investment than natural forest management. For example, Rietbergen (1989) reports results from th.e Ivory Coast that attribute a cost of US$7.40 per cubic meter of plantation- grown timber, compared with a cost of US$5.60 per cubic meter derived from a managed natural forest. Nevertheless, conventional financial analy- ses generally favor the former (Leslie 1987), primarily because of the longer period over which the costs of natural forest management are compounded. However, as Rietbergen (1989) notes, and Barbier (1990), Krutilla (1988), and Leslie (1987, 1989) have discussed, conventional financial analyses typi- cally ignore or underestimate the nonmarket benefits associated with man- agement of natural forests. The valid comparison of plantation and natural forest options demands more sophisticated and comprehensive economic analyses than have been applied in the past, and some slow progress is being made in this direction (e.g., Bowes and Krutilla 1989). The appropriate balance between reliance on and investment in planta- tions and natural forests will depend on the particular demands on the forest resources and the capacity of plantations and natural forests to meet those demands. It is likely that both forms of forest will have a role to play, and that neither one should be emphasized to the neglect of the other. SUMMARY AND CONCLUSIONS Plantation forestry has come to be synonymous with relatively high input, intensive management of monocultures for the production of a rela- tively narrow range of products. In this sense, the development of forest plantation technology has paralleled that of agricultural crops. Although in some circumstances this approach may still be valid, in many others a broader range of plantation objectives and more intimate integration with other land uses are likely to be necessary for successful plantation production. Thus to some extent plantation forestry needs to be redefined to promote an under- standing of its potentially broader role. Techniques for the successful establishment and management of planta- tions are well developed, although informed adaptation is necessary for each specific environment. Plantation programs should be designed to rec- ognize and accommodate environmental values and managed so as to con- PLANTATION FORESTRY 395 tribute to genetic, ecological, and landscape diversity. The political, eco- nomic, and social environments are far more important determinants of the success of forest plantations than issues that are primarily technical. The issues fundamental to the success of plantation forestry are those of political economy; plantations that are not designed and managed to ac- commodate local needs are likely to be of limited value and prospect. The integration of agroforestry practices into plantation programs is one means by which individual and industrial needs may be realized simultaneously. This implies the development of "complex plantation forestry," in contrast to the earlier, simpler concept of plantations managed for a more limited range of products. Complex plantation forestry is likely to integrate other land uses within its boundaries, and to promote the early and continuing production of a wide variety of goods, services, and values. Within these constraints, the principles of sustainable land use and management planning apply as much to plantations as to other land uses. Although plantations of forest trees are relatively long-term investments, the financial returns from them can be as good as those from alternatives. The early and continuing production of as wide a range of products as possible also enhances financial returns and sustainability. Even where a broad range of outputs is achieved, the products of plan- tations are likely to differ substantially from those of natural forests. Plan- tation forestry should be seen as complementary to the management of natural forests rather than as a substitute for it. The following elements are necessary for plantation programs: • For plantation forestry to be successful, it must be recognized that each plantation is to some extent unique and site specific. Although the fundamental principles of successful plantation forestry are well established, their successful application to particular situations re- quires informed interpretation and intelligent adaptation. • The objectives and principles of plantation forestry need to be rede- fined to describe a more integrated form of land use yielding a wider range of products. , • Plantation forestry is most appropriate where land and tree tenure are not contentious, and where plantation systems can be managed sustainably to satisfy both industrial and nonindustrial demands. The incorporation of agroforestry practices will often be appropriate. • Assessment and acceptance of social and environmental impacts are prerequisites for plantation forestry. • The design and management of plantation forests should follow the principles of sustainable land use and should recognize the constraints imposed by nonmarket values. • Investment in plantation forestry should be guided by an economic assessment of plantation performance and alternatives, rather than by the traditional, narrow, financial analyses. The developing eco- nomic understanding of sustainable development has much to con- tribute toward this end. 396 J\/IANAGING THE WORLD'S FORESTS The main technical requirements for successful plantation forestry are as follows: • The availability of genetic resources of potentially useful species. • Selection of the best-adapted species and provenances, their continu- ing genetic improvement, and the raising of planting stock under favorable nursery conditions. • Informed scientific observation by competent staff. The least that should be done is to monitor growth, nutrition, and health. Where necessary, monitoring should lead to-- -timely silvicultural interventions for plantations to achieve their potential. Because these may be expensive, adequate finance must be available on a continuing basis. -adequately paid, well-motivated, and well-qualified staff, available on a long-term basis to manage plantations, with support, when needed, from specialists in relevant disciplines such as anthropol- ogy, sociology, entomology, genetics, land-use planning, or soil science. The key research approaches necessary for the continuing development of plantation forestry are as follows: • The integration of agroforestry and plantation forestry methodolo- gies and technologies, which requires both social and biological re- search. • The identification of the interactions that determine the outputs of complex plantation forestry, and the means by which the multiple benefits might best be obtained. • The development of more satisfactory approaches to economic analy- ses for the evaluation of plantation forestry investments and of alter- native investments. • The means by which relevant information can more effectively be disseminated and adapted to particular circumstances. • The development of methodologies to identify and obtain the most appropriate genetic material for plantation programs, with greater emphasis on high-quality hardwood species. The formulators of forest policy should recognize the following truths: • Investment in simple plantation foreslry is often less likely to satisfy societies' demands from forest resources than is investment in what has been defined here as complex plantation forestry. • Proposals for plantation forestry should be formulated in the broader social context. Methodologies developed for agroforestry interven- tions offer some guidance to relevant approaches. • Planning and implementation of plantation forestry programs must emphasize flexibility and adaptation. Such features should be accom- modated in the terms of finance for plantation forestry. PLANTATION FORESTRY 397 • Investments in plantation forestry complement, rather than substi- tute for, investments in the conservation and management of natural forests. REFERENCES Aldhous, J.R. 1972. Nursery Practice. UK Forestry Commission Bulletin No. 43. London: Her Majesty's Stationery Office (HMSO). Arnold, J.E.M. 1983. "Economic Considerations in Agroforestry Projects." Agroforestry Systems 1:299- 311. _ _ _ . 1991. Community Forestry: Ten Years in Review. FAO Community Forestry Note 7. Rome: Food and Agricultural Organization (FAO). Bailly, C., C. Barbier, J. Clement, J.P. Gaudet, and 0. Hamel. 1982. "Les Problemes de la Satisfaction des Besoins en Bois en Afrique Tropicale Seche. Connaissances et Incertitudes." Bois et For€ts des Tropiques 197:23-43. Barbier, E.B. 1990. Natural Resource Degradation: Policy, Economics and Management. London Environ- mental Economics Center. Gatekeeper Series No: LEEC 90-01. Barnes, R.D., and J. Burley. 1990. "Tropic Forest Genetics at the Oxford Forestry Institute: Explora- tion, Evaluation, Utilization, and Conservation of Genetic Resources." Forest Ecology and Man- agement 35:159~70. Bellefontaine, R., C. Cossalter, and 0. Souvannavong. In press. "L'Experience du CTFT en Matiere des Gaines Forestieres." Bois et For€ts des Tropiques. Blaikie, P.M. 1985. The Political Economy of Soil Erosion in Developing Countries. London: Longman. Blatner, K.A., and J.L. Greene. 1989. "Woodland Owner Attitudes Toward Timber Production and Management." Resource Management and Optimization 6:205-223. Boardman, R. 1988. "Living on the Edge: The Development of Silviculture in South Australian Pine Plantations." Australian Forestry 51:135-56. Booth, T.H., S.D. Searle, and D.J. Boland. 1989. "Bioclimatic Analysis to Assist Provenance Selection for Trials." New Forests 3:225-34. Bowes, M.D., and J.V. Krutilla, 1989. Multiple-Use Management: The Economics of Public Forestlands. Washington, D.C.: Resources for the Future. Brown, A.H.D., and G.J. Moran. 1981. "Isozymes and Genetic Resources of Forest Trees." In Isozymes of North American Forest Trees and Forest Insects, ed. M.T. Conkle. Berkeley, Calif.: Pacific South- west Forest and Range Experiment Station, USDA Forest Service. Brunton, D.P. 1987. "Financing Small-Scale Rural Manufacturing Enterprises." In Small-Scale Forest- Based Processing Enterprises. Rome: FAQ, Forestry Paper No. 79, 117-48. Burdon, R.D. 1989a. "Summing Up." In Breeding Tropical Trees: Population Structure and Genetic Improvement Strategies in Clonal and Seedling Forestry, eds. G.L. Gibson, A.R. Griffin; and A.C. Matheson. Oxford, Eng.: Oxford Forestry Institute, and Arlington, Va.: Winrock International. _ _ _ 1989b. "When Is Cloning on an Operational Scale Appropriate?" In Breeding Tropical Trees: Population Structure and Genetic Improvement Strategies in Clonal and Seedling Forestry, eds. G.L. Gibson, A.R. Griffin, and A.C. Matheson. Oxford, Eng.: Oxford Forestry Institute, and Arling- ton, Va.: Winrock International. Byron, R.N., and G. Waugh. 1988. "Forestry and Fisheries in the Asian-Pacific Region: Issues in Natural Resource Management." Asian-Pacific Economic Literature 2:46-80. Chambers, R., N.C. Saxena, and T. Shah. 1989. To the Hands of the Poor. New Delhi: Oxford and IBH Publishers. Champion, H.G., and S.K. Seth. 1968. General Silviculture for India. Delhi: Government of India Publication Branch. Chapman, M.D. 1987. "Traditional Political Structure and Conservation in Oceania." Ambia 16:201- 05. 398 MANAGING THE WORLD'S FORESTS Cheliak, W.M., and D.L. Rogers. 1990. "Integrating Biotechnology into Tree Improvement Pro- grams." Canadian Journal of Forest Research 20:452-63. Cotterill, P.P. 1992. Personal communication. Obidos, Porhtgal: CELBL Danish International Development Agency (DANIDA). Undated. "DANIDA Forest Seed Centre 1981-86." Dawkins, H.C. 1967. Productivity of Tropical Rain Forests and Their Ultimate Value to Man." In The Ecology of Man in the Tropical Environment. Merges, Switz.: IUCN Publications New Series 4. _ _ _. 1992. Personal communication. Oxford, Eng.: Oxford Forestry Institute. de Beer, J.H., and M.J. McDermott. 1989. The Economic Value of Non-Timber Forest Products in South- east Asia. Amsterdam: Netherlands Committee for IUCN. Dommergues, Y. 1986. "Le Labora~oire de Biotechnologie des Systemes Symbiotiques Forestiers Tropicaux de Nogent-sur-Marne." Bois et Forets des Tropiques 213:41-46. Douglas, J.J. 1983. A Re-appraisal of Forestry Development in Developing Countries. The Hague: Martinus Nijhoff/Dr. W. junk. _ _ _ 1986. "Forestry and Rural People: New Economic Perspectives." In Proceedings Division 4, 18th IUFRO World Congress. Ljubljana, Yugoslavia, September 7-21. Dvorak, W.S. 1990. "CAMCORE: Industry and Government's Effort to Conserve Threatened Forest Species in Guatemala, Honduras, and Mexico." Forest Ecology and Management 35:131-50. Elliott, D.A., R.N. James, D.W. McLean, and W.R.J. Sutton. 1989. "Financial Returns from Plantation Forestry in New Zealand." In Proceedings: 13th Commonwealth Forestry Conference, 6C. Rotorua, N .Z., September. Evans, J. 1976. "Plantations, Productivity and Prospects." Australian Forestry 39:150--63. _ _ _ 1982. Plantation Forestry in the Tropics. Oxford, Eng.: Clarendon Press. _ _ _ 1990. "Long-Term Productivity in Forest Plantations: Status in 1990." Paper IUFRO World Congress. Montreal, Canada. Food and Agricultural Organization. 1985a. Tree Growing by Rural People. FAO Forestry Paper No. 64. _ _ _ 1985b. Tropical Forestry Action Plan. Rome: FAO. Formby, J. 1986. Approaches to Social Impact Assessment. Working Paper 1868/8, Centre for Resource and Environmental Studies, Australian National University, Canberra. Gauthier, J.J. 1991. "Les bois de plantation dans le commerce mondial des produits forestiers." In L'emergence des nouveaux potentiels forestiers dans le monde. Paris: AFOCEL. Gibson, G.L., R.D. Barnes, and J.S. Berrington. 1983. "Provenarice Productivity in Pinus caribaea and Its Interaction with Environment." Commonwealth Foresfry Review 62:93-106. Grainger, A. 1991. "Contraints on Increasing Tropical Forest Area to Combat Global Climatic Change." In Proceedings: Technical Workshop to Explore Options for Global Forestry Management. Bangkok: IIED, ITTO, and ONEB, 196-208. Greaves, A., and P.S. McCarter, 1990. Cordia alliodora: A Promising Tree for Tropical Agroforestry. Tropical Forestry Paper No. 22, Oxford, Eng.: Oxford Forestry Institute. Hamrick, J.L. 1'983. "The Distribution of Genetic Variation Within and Among Natural Plant Popu- lations." In Genetics and Conservation, eds. C.M. Schonewald-Cox, S.M. Chambers, B. MacBryde, and W.L. Thomas. Menlo Park, Calif.: Benjamin/Cummings Publishing Co. Hurditch, W.I. 1992. Problems of Public Forestry and the Socio-economic Impacts of Privatization. Oxford: Oxford Forestry Institute Occasional Paper. Hyman, E.L., and B. Stiftel. 1988. Combining Facts and Values in Environmental Impact Assessment. Social Impact Assessment Series No. 16, Boulder, Colo.: Westview Press. Ikemori, Y.K. 1990. "Gerietic Variation in Characteristics of Eucalyptus grandis (Hill) Maiden Raised From Micro-Propagation, Macro-Propagation and Seed." D.Phil. dissertation, Department of Plant Sciences, Oxford University. Indian Forester. 1991. "Forest Corporations, Special Issue" 117(9). International Tropical Timber Organization (ITIO). 1990a. ITTO Guidelines for the Sustainable Man- agement of Natural Tropical Forests. Yokohama, Japan. ITTO Technical Series 5. _ _ _ 1990b. The Promotion of Sustainable Forest Manage1nent: A Case Study in Sarawak, Malaysia. International Tropical Timber Organization. PLANTATION FORESTRY 399 Kallio, M., D.P. Dykstra, and C.S. Binkley. 1987. "Introduction." In The Global Forest Sector: An Analytical Perspective, eds. M. Kallio, D.P. Dykstra, and C.S. Binkley. London and New York: Wiley-lnterscience. Kang, H., and H. Nienstaedt. 1987. "Managing Long-Term Breeding Stock." Silvae Genetica 36:30-39. Kardell, L., E. Steen, and A. Fabiao. 1984. "Eucalyptus in Portugal: A Threat or a Promise?" Ambia 15:6-13. Kirkland, A. 1989. "The Rise and Fall of Multiple Use Management in New Zealand." New Zealand Forestry 33(1):9-12. Kitzmiller, J.H. 1990. "Managing Diversity in a Tree Improvement Program." Forest Ecology and Management 35:131-50. Krutilla, J.V. 1988. Economics and Public Forestland Management. Twelfth William P. Thompson Me- morial Lecture, School of Forestry, Northern Arizona University. Leslie, A.J. 1980. "Institutional Structures for the Implementation of Forest Policy." Paper presented at 11th Commonwealth Forestry Conference, Trinidad and Tobago, September. _ _ _ 1987. "Economic Feasibility of Natural Management of Tropical Forests." In Natural Man- agement of Tropical Moist Forests, eds. F.R. Mergen and J.R. Vincent. School of Forestry and Environmental Studies, Yale University. _ _ _ 1989. "On the Economic Prospects for Natural Management in Temperate Hardwoods." Forestry 62:147--66. LOfgren:, K-G. 1988. "On the Economic Value of Genetic Progress in Forestry." Forest Science 34:708- 23. Lundgren, B. 1978. Soil Conditions and Nutrient Cycling Under Natural and Plantation Forests in the Tanzanian Highlands. Rapporte Uppsala Institute. Vaxtec och Mark, No. 31. Marshall, D.R. 1989. "Limitations to the Use of Germplasm Collections. In The Use of Plant Genetic Resources, eds. A.H.D. Brown, O.H. Frankel, D.R. Marshall, and J.T. Williams. Cambridge, Eng.: Cambridge University Press. _ _ _ 1990. "Crop Genetic Resources: Current and Emerging Issues." In Plant Population Genetics, Breeding, and Genetic Resources, eds. A.H.D. Brown, M.T. Clegg, A.L. Kahler, and B.S.Weir. Sunderland, Mass. Sinauer Associates. Mather, A.S. 1990. Global Forest Resources. London: Belhaven Press Mathur, H.N., S.F.H. Raj, and S. Naithani. 1984. "Ground Water Quality (pH) under Different Vegetation Covers at Osamund (Nilgiri Hills)." Indian Forester 110:110-16. McGaughey, S.E., and H.M. Gregersen. 1988. Investment Policies and Financing Mechanisms for Sus- tainable Forestry Development, Washington, D.C.: Inter-American Development Bank. Midgley, S.J. 1988. The Australian Tree Seed Centre: A Window to the Resource." Canberra, Australia: Commonwealth Scientific and Industrial Organization. Molnar, A. 1989. Community Forestry: A Review. FAO Committee Forestry Note No. 3. Moran, G.F., and J.C. Bell. 1987. "The Origin and Genetic Diversity of Pinus radiata in Australia." Theoretical and Applied Genetics 83:616-22. Myers, N. 1989a. Deforestation Rates in Tropical Forests and Their Climatic Implications. London: Friends of the Earth. Namkoong, G., R.D. Barnes, and J. Burley. 1980. A Philosophy of Breeding Strategy for Tropical Forest Trees. Commonwealth Forestry Institute, University of Oxford, Tropical Forestry Paper No. 16. Namkoong, G., H.C. Kang, and J.S. Brouard. 1988. Tree Breeding: Principles and Strategies. New York and London: Springer-Verlag. Palmberg, C., and J.T. Esquinas-Alcazar. 1990. ''The Role of United Nations Agencies and Other International Organizations in the Conservation of Plant Genetic Resources." Forest Ecology and Management 35:171-78. Palmer, J.R. 1986. "Case Study: JARI: Lessons for Land Managers in the Tropics." Paper presented at International Workshop on Rainforest Regeneration and Management, Guri, Venezuela, November 24-28. Pereira, H., and J. Santos Pereira. 1988. "Short Rotation Biomass Plantations in Portugal. In Biomass Forestry in Europe: A Strategy for the Future, eds. F.C. Hummel, W. Palz, and G. Grassi. London: Elsevier Applied Science. 509-39. 400 MANAGING THE WORLD'S FORESTS Poore, Duncan, Peter Burgess, John Palmer, Simon Rietbergen, and Timothy Synnott. 1989. No Timber Without Trees: Sustainability in the Tropical Forest. A study for International Tropical Timber Organization. London, Eng.: Earthscan Publications. Price, C. In press. Time, Discounting and Value. Oxford, Eng.: Basil Blackwell. Proceedings: 1985 Conference on Rapid Rural Appraisal. 1985. Khon Kaen University, Thailand. Queensland Department of Forestry. 1981. Research Report 1981. Department of Forestry, Brisbane, Queensland, Australia. _ _ _ 1987. Aspects of Management of Plantations in Tropical and Sub-Tropical Queensland.· Depart- ment of Forestry, Brisbane, Queensland, Australia. Raintree, J.B. 1987. "The State of the Art of Agroforestry Diagnosis and Design." Agroforestry Sys- tems 5:219-50. Reddy, K.V., D.L. Rockwood, and G.F. Meskimen. 1986. "A Strategy For Conversion of a Eucalyptus grandis Base Population into a Seedling Seed Orchard." In Proceedings of IUFRO Joint Meeting of Working Parties on Breeding Theory, Progeny Testing, and Seed Orchards, Williamsburg, Va.: Octo- ber 13-17. Rickman, R. 1991. What's Good for Woods. London Center for Policy Studies, Policy Study No. 129. Rietbergen, S. 1989. "Africa." In No Timber Without Trees, ed. D. Poore, 40-76. London, Eng.: Earthscan Publications. Risby, A.E. 1987. "Realities of Timber Production." Chapter 16 in Prospects for Australian Hardwood Forests, eds. J. Dargavel and G. Sheldon. (Canberra: Center for Resource and Environmental Studies, Australian National University, Monograph 19), 213-20. Sargent, C. 1990. The Khun Song Plantation Project. London: International Institute for Environment and Development. Savill, P.S., and J. Evans. 1986. Plantation Silviculture in Temperate Regions. Oxford, Eng.: Clarendon. Sedjo, R.A. 1987. "Forest Resources of the World: Forests in Transition." In The Global Forest Sector: An Analytical Perspective, eds. M. Kallio, D.P. Dykstra, C.S. Binkley. London and New York: Wiley-Interscience. Shiva, V. 1987. Forestry Crisis and Forestry Myths. Penang, Malaysia: World Rainforest Movement. Spilsbury, M.J. 1990. "The Economic Prospects for the Natural Management of Woodlands in the UK." Forestry 63:379-90. Srivastava, H.C., B. Vatsya, and K.K.G. Menon, eds. 1986. Plantation Crops: Opportunities and Con- straints. New Delhi, India: Oxford and IBH Publishing. Sturm, J. 1991. "Panorama des nouveaux potentals forestiers de l'hbnisphCre sud: origine, problbnatique, impact." L'emergence des nouveaux potentiels forestiers clans le monde. Paris: AFOCEL, 35--50. Sutton, W.R.J. 1991. "Are We Too Concerned about Wood Production?" New Zealand Forestry 36(3):25- 8. Thompson, M.A. 1990. "Determining Impact Significance in EIA: A Review of 24 Methodologies." Journal of Environmental Management 30:235-50. Tomkins, J. 1990. "Recreation and the Forestry Commission: The Case for Multiple-Use Resource Management Within Public Forestry in the UK." Journal of Environmental Management 30:79-88. Tompkins, S. 1989. Forestry in Crisis. London, Eng.: Christopher Helm. van Maaren, A. 1984. "Forests and Forestry in National Life." In Forest Policy. A Contribution to Resource Development, ed. F.C. Hummel. Martinus Nijhoff/Dr. W. Junk. Westoby, J.C. 1962. "The Role of Forest Industries in the A.ttack on Economic Underdevelopment. "In The Purpose of Forests, ed. J.C. Westoby. Oxford, Eng.: Basil Blackwell. White, T.L. 1987. "A Conceptual Framework for Tree Improvement Programs." New Forests 1:325- 42. Whiteside, I.D. 1989. ''Economic Advances Made in New Zealand Radiata Pine Plantation Forestry Since the Early 1980s." In Proceedings: 13th Commonwealth Forestry Conference, 6C. Rotorua, N.Z.: September. Wiersum, K.F., ed. 1984. Strategies and Designs for Afforestation, Reforestation, and Tree Planting. The Hague: Pudoc Wageningen. _ _ _. 1987. The Purpose of Forests. Oxford, Eng.: Basil Blackwell. _ _ _. 1989. Introduction to World Forestry. Oxford, Eng.: Basil Blackwell. PLANTATION FORESTRY 401 Wilson, R.V. 1989. "Financial Returns from Plantation Forestry in Australia." In Proceedings: 13th Commonwealth Forestry Conference, 6C. Rotorua, N.Z.: September. World Bank. 1991. The Forest Sector: A World Bank Policy Paper:Washington, D.C.: World Bank. Yoho, J.G. 1985. "Continuing Investments in Forestry: Private Investment Strategies." In Investment in Forestry, ed. R.A. Sedjo. Boulder, Colo.: Westview. Young, C. 1990. Personal communication. Landel Mills Commodity Studies Ltd. Oxford, Eng. Zobel, B.J., and J.T. Talbert. 1984. Applied Forest Tree Improvement. New York: Wiley. Zobel, B.J., G. van Wyk, and P. Stahl. 1987. Growing Exotic Forests. New York: Wiley Interscience. 15 Agroforestry Systems Design: An Ecozone Approach P.K.R. Nair S ignificant progress has been made since the 1970s in understanding the potential of agroforestry systems to alleviate some of the problems of tropical land management. This "coming of age" of agroforestry is the real- ization of the potentials for sustained yields and conservation benefits, and of the multiple-output nature of the age-old land-use systems in which trees and crops, and sometimes animals, are grown on the same piece of land in interacting combinations. There are many examples of low-input agroforestry systems in various ecological regions of the tropics where woody perennials are deliberately mixed with crops or animals or both in order to derive maximum economic and ecological benefit (Nair 1989; Rocheleau, Weber, and Field-Juma 1988; Cook and Grut 1989; Zhaohua et al. 1991). Although most of these systems have been little studied, there is now enough technical, sociological, and economic information or insight on both the merits and the weaknesses of these systems to provide a basis for recommending the adoption of agro- forestry (Steppler and Nair 1987; MacDicken and Vergara 1990; recent vol- umes of Agroforestry Systems). The next important step in channeling development assistance to agroforestry programs is to provide specific recommendations on the most appropriate agroforestry systems and practices in different parts of the world. The extremely site-specific nature of agroforestry, conditioned by biophysi- cal and sociocultural characteristics, poses serious difficulties in developing precise recommendations of wider applicability. At the same time, develop- ment agencies and donors with a global mandate need information to for- 403 404 MANAGING THE WORLD'S FORESTS mulate policy with wider applicability. This chapter evaluates a few prom- ising agroforestry practices in terms of their potential as well as ecological adaptability, and develops a matrix of agroforestry practices versus agro- ecological conditions that could be used as the basis for the design of agroforestry systems. The common constraints to adoption of agroforestry practices also are considered briefly in order to highlight the issues that need to be considered for developing forestry policy. Readers are advised to refer to a recent World Bank study by this author (Nair 1990) for a detailed discussion on some of the technical and socioeconomic aspects of agroforestry. THE APPROACH TO AGROFORESTRY Following its introduction as a land-use approach in the late 1970s, there was a surge of enthusiasm to define agroforestry and the definition proposed by the International Council for Research in Agroforestry (ICRAF) has gained wide acceptance: Agroforestry is a collective name for land-use systems and technologies in which trees are deliberately used on the same units of land as agricultural crops or animals. In agroforestry systems the different components interact both ecologically and economically. The key concepts of agroforestry are now well established. Agroforestry- • combines production of multiple outputs with protection of the re- source base; • emphasizes the use of indigenous, multipurpose trees and shrubs; is particularly suitable for fragile environments; • is more concerned with sociocultural values than most other land-use systems; and • is structurally and functionally more complex than monoculture. Between 1982 and 1987 ICRAF compiled an inventory of agroforestry systems and practices being used in the developing countries. The inven- tory, financed partly by the U.S. Agency for· International Development (USAID), involved collecting, collating, and evaluating data and publishing the results. It brought together, for the first time, a substantial body of information on many agroforestry systems, their structures and functions, and their merits and weaknesses (Nair 1989). The practice of agroforestry ranges from simple forms of shifting culti- vation to sophisticated hedgerow intercropping systems; from systems with varying densities of tree stands (from widely scattered Faidherbia (Acacia) albida trees in Sahelian millet fields to high-density, complex, multistoried "home gardens" of the humid tropics); from systems in which trees play a predominantly service role (e.g., windbreaks) to those in which they pro- vide the main commercial product (e.g., intercropping with plantation crops). Systems also vary in the species they use, the way they arrange and manage plants, and their outputs. AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 405 AGROFORESTRY PRACTICES The collection of information about existing agroforestry systems from different parts of the tropics demonstrated that there are only a few distinc- tive agroforestry practices-about 20-that constitute the innumerable agroforestry systems. Exhibit 15-1 lists these practices and their essential characteristics. Some of the most promising and common agroforestry prac- tices are examined in the sections that follow. Improved Fallows Traditional shifting cultivation systems are time-dependent sequential arrangements in which the tree (fallow) and the crop phases alternate. The term improved fallows may imply the use of improved tree or shrub species during the fallow phase, but can in practice mean alternatives to the fallow phase of shifting cultivation, leading ultimately to permanent cultivation of some sort. Shifting cultivation cannot be expected to continue in its tradi- tional form. Any realistic approach to improve shifting cultivation would therefore have to be reconciled with a situation of shorter fallow. Unmanaged shorter fallows, however, can cause all the disastrous consequences that are attributed to shifting cultivation (such as soil erosion, loss of soil fertility, weed infestation, and buildup of pests and pathogens). Therefore, managed, permanent cultivation systems that have some advantages of traditional shifting cultivation should be much better than unchecked "fallow-depleted" traditional shifting cultivation. Two types of improved fallows can be envisaged: economically enriched fallows, which increase the economic utility of the fallow vegetation by en- richment with trees valued for cash or subsistence purposes, and biologically enriched fallows, which enhance and accelerate the vegetative regeneration of soil fertility and control of weeds. Long-fallow forest shifting cultivators are unlikely to be interested in techniques to improve soil fertility and weed control because these problems do not yet bother these people, but they may well respond to opportunities for economic benefit from improved fallows. The validity of the economic enrichment approach has been amply proved by such indigenous examples as the fallow enrichment planting of rattan by the Luangan Dayaks of Borneo (Weinstock 1984), of cedar and bamboo by the Lingnan Yao of China (Lebar, Hickey, and Musgrave 1964), of Casuarina by the Siane of Papua New Guinea (Salisbury 1962), of gum arabic in the Sahel (von Maydell 1986) and of multipurpose fallow woodlots whose spe- cies diversity exceeds that of the natural forest by the Ifugao terrace- builders of the Philippines (Conklin 1980). Biologically enriched fallow prac- tices using Acioa baterii, Anthonotha macrophylla, Alchornea cordifolia, Gliricidia sepium, and Leucaena leucocephala have been reported by Benneh (1972); Okigbo and Lal (1979); Getahun, Wilson and Kang (1982); Agboola et al. (1982); Dijkman (1950); Olofson (1983); FAO (1984); and Raintree and Warner (1986). 406 MANAGING THE WORLD'S FORESTS EXHIBIT 15-1. Main Characteristics of the Most Common Agroforestry Practices in the Tropics Agroforestry Arrangement Major Agroecological Practice of Components Components Adaptability Agrisilvicultural Systems (trees and crops, including shrub, vine, tree crops) Improved tree Woody species planted w Fast··growing In shifting cultivation fallows and left to grow during preferably areas the "fallow phase" h Corrunon agricultural crops Modified taungya Combined stand of w Usually All ecological regions, woody and agricultural plantation where taungya is species during early forestry species practiced; several stages of establishment h Corrnnon improvements of plantations agricultural possible crops Alley cropping Woody species in w Fast-·growing Subhumid to humid (and other forms hedges; agricultural leguminous areas with high of hedgerow species in alleys between species that human population intercropping) hedges; microzonal or coppice pressure and fragile strip arrangement vigorously· (productive but h Common easily degradable) agricultural soils crops Home gardens Intimate, multistory w Fruit trees In all ecological combination of various predominate; regions, especially trees and crops also other in areas of high around homesteads woody species, population density vines, etc. Plantation-crop 1. Integrated, dense, w Plantation crops In humid lowlands or combinations multistory mixtures of such as coffee, tropical humid/ sub- plantation crops cacao, coconut, humid highlands 2. Mixtures of plantation and fruit trees, (depending on the crops in alternate or especially in(!); plantation crops con- other regular fuelwood/ cerned); usually in arrangements fodder species, smallholder sub- 3. Shade trees for especially in (3) sistence systems plantation crops; shade h Usually present trees scattered in (4), and to 4. Intercropping with some extent in agricultural crops (!);shade- tolerant species Key w =woody f = fodder h =herbaceous a= anirnals (continued) AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 407 EXHIBIT 15-1. Main Characteristics of the Most Common Agroforestry Practices in the Tropics (continued) Agroforestry Arrangement Major Agroecological Practice of Components Components Adaptability Agrisilvicultural Systems (trees and crops, including shrub, vine, tree crops) Multistory tree Multispecies, multi- w Different woody Areas with fertile soils, gardens layer dense plant components of good availability of associations with no varying forms labor, and high human organized planting and growth population pressure arrangements habits h Usually absent; shade-tolerant ones sometimes present Windbreaks and Trees around w Combination In wind-prone areas shelterbelts, farmlands/plots of tall-growing, livehedges spreading types h Local agricultural species Use of trees in Trees on bunds, w Multipurpose In sloping areas, espe- soil conservation terraces, raisers, etc., and fruit trees cially in highlands, and reclamation with or without grass h Common reclamation of de- strips; trees for soil agricultural graded, acid, alkali reclamation crops soils, and sand-dune stabilization Multipurpose Trees scattered w Multipurpose In all ecological trees on haphazardly or trees and other regions, especially in croplands systemically on bunds fruit trees subsistence farming; terraces, or plot/ h Common also commonly field boundaries agricultural integrated with crops animals Fuelwood lots Interplanting fuel- w Fuelwood In all ecological wood species on or species regions around agricultural h Local lands agricultural crops Key w =woody f = fodder h =herbaceous a= animals (continued) 408 MANAGING THE WORLD'S FORESTS EXHIBIT 15-1. Main Characteristics of the Most Common Agroforestry Practices in the Tropics (continued) Agroforestry Arrangement Major Agroecological Practice of Components Components Adaptability Silvopastoral Syste:ms (trees with pasture/ animal production systems) Trees on range- Trees scattered w Multipurpose; Extensive grazing land or pastures irregularly or of fodder areas systematically value f Present Silvicultural Systems (trees with pasture/ animal production systems) Fodder banks Production of protein- w Leguminous Usually in fairly rich tree fodder on fodder trees densely populated farm/ rangelands for h Present areas cut-and-carry fodder f Present production a Present Plantation crops Example: cattle under w Plantation In areas with less with fodder and coconut crops in crops pressure on plantation livestock Southeast Asia and f Present crops South Pacific a Present Agrosilvopastoral Systems (trees and crops and pasture/ animals) Home gardens Intimate, multistory w Fruit trees In all ecological with aniinals combination of various predominate; regions with high trees and crops, as also other human population well as animals woody species density around homesteads a Present Multipurpose Woody hedges for w Fast-growing Humid to subhumid woody hedge- browse, mulch, green and coppicing areas with hilly and rows around manure, soil fodder shrubs sloping terrain homesteads conservation, etc. and trees h Similar to alley cropping and soil cons·ervation Other (Special) Agroforesh'y Practices Agriculture Trees for honey w Honey Depending on the with trees production producing feasibility of (other apiculture components maybe present) Key w =woody f =fodder h =herbaceous a = anirnals (continued) AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 409 EXHIBIT 15-1. Main Characteristics of the Most Common Agroforestry Practices in the Tropics (continued) Agroforestry Arrangement Major Agroecological Practice of Components Components Adaptability Other (Special) Agroforestry Practices (continued) Aquaforestry Trees lining fish ponds w Trees and Lowlands (trees with shrubs pre- fisheries) ferred by fish (other components may be present) Multipurpose For various purposes w Multipurpose Various woodlots (wood, fodder, soil species; protection, soil location- reclamation, etc.) specific species (other components may be present) Key w =woody f =fodder h =herbaceous a= animals Modified Taungya Taungya, from the Burmese words Taung (hill) and ya (cultivation), refers to the practice of growing agricultural crops along with forestry spe- cies during the early phases of forestry plantation establishment. The prac- tice, initially developed in British colonial India in the 1850s as a means to get farmers involved in the establishment of forest plantations with a view to reducing forest encroachment by these farmers, has since spread to most of the tropics. A distinction is sometimes made between integral and partial taungya systems. Partial taungya reflects "predominantly only the economic inter- ests of its participants (as in some kinds of cash crop, resettlement, and squatter agriculture)," whereas integral systems "stem from a more tradi- tional, year-round, community-wide, largely self-contained, and ritually sanctioned way of life" (Conklin 1957). By analogy with integral shifting cultivation, integral taungya is meant to invoke the idea of a land-use prac- tice that offers a more complete and culturally integrated approach to rural development-not merely the temporary use of a piece of land and a pov- erty-level wage, but a chance to participate equitably in a diversified and sustainable agroforestry economy. The social aims of the proposed approach are high, and they are no- where yet fully realized in practice, although the forest village schemes in Thailand (Boonkird, Fernandes, and Nair 1984) may come closest to the ideal. In some variants of this approach, participation in forestry is made 410 MANAGING THE WORLD'S FORESTS more attractive to traditional shifting cultivators not only by encouraging them to grow long-term perennial cash crops by widening the between-row spacing of the commercial forest species, but also by allocating permanent agricultural plots for them to use as they see fit. In addition, these people are paid decent wages for a variety of work opportunities and are provided with a range of extension and community development inputs such as hous- ing assistance, clinics, schools, and places of worship. Far from being an exploitative practice, this Thai variant of taungya system promises to be- come a model example of integral taungya, although it is inadequately docu- mented in literature. Alley Cropping Alley cropping is a technology about which much has been written, high expectations have been raised, and some research has been done. Two reports that consolidate the information on this promising technology are Kang, Wilson and Lawson (1984) and Kang, Reynolds, and Atta-Krah (1989). In alley cropping, arable crops are grown between hedgerows of planted shrubs and trees, preferably leguminous species, which are periodically pruned to prevent shading to companion crops (Kang, Wilson, and Sipkin 1981, Kang et al. 1984). This production system is classified by Nair (1985) as a zonal agroforestry system. The shrubs and trees grown in the hedgerows retain the same functions of recycling nutrients, suppressing weeds, and controlling erosion on sloping land as those in the bush fallow. Prunings from the trees and shrubs are a source of mulch and green manure. Legu- minous woody species also add fixed nitrogen to the system. The alley-cropping concept is currently being evaluated in many parts of the tropics under different names. ICRAF used the term hedgerow inter- cropping (Torres 1983), while in Sri Lanka the term avenue cropping is used (Wijewardene and Waidyanatha 1984). The criterion most likely used to assess the desirability of alley cropping is the effect of this practice on crops yields. Indeed, most alley-cropping trials produce little data other than crop yield data, and these are usually derived from trials conducted over a rela- tively short period of time. Many trials have produced promising results. An eight-year alley-crop- ping trial conducted by Kang et al. (1989) in southern Nigeria on sandy soil showed that using Leucaena prunings could maintain maize grain yield at a level of 2 tons per hectare, as against 0.66 ton per hectare without Leucaena prunings and fertilizer. Supplementing the prunings with 80 kilograms of nitrogen per hectare increased the maize yield to over 3.0 tons per hectare. Unfortunately, the effect of using fertilizer without the addition of Leucaena pruning was not tested. Yamoah, Agboola, and Wilson (1986) reported that to increase the yield of maize alley-cropped with Cassia, Gliricidia, and Flemingia macrophylla, it was necessary to add nitrogen. However an earlier report by Kang et al. (1981) indicated that an application of 10 tons per hectare of fresh Leucaena prunings had the same effect on maize yield as the AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 411 addition of 100 kilograms of nitrogen per hectare, although to obtain this amount of Leucaena leaf material it was necessary to supplement production from the hedgerows with externally grown materials. Kang and Duguma (1985) showed that the maize yield obtained from using Leucaena leaf mate- rials produced in hedgerows planted four meters apart was the same as the yield obtained when 40 kilograms of nitrogen per hectare was applied to the crop. Results from other alley-cropping trials are less promising. For example, in trials conducted on an acid soil at Yurimaguas, Peru, the yields of all crops studied in the experiment, apart from cowpea, were extremely low, and the overall yield from alley-cropped plots was equal to or less than that from the control plots. Szott (1987) concluded from these data that the main reasons for the comparatively poor crop performance under alley-cropping treatments were root competition and shading. Other possible explanations are that the surface mulch physically impeded the emergence of seedlings, that the decomposing mulch temporarily immobilized the nutrient cycle and thus seriously reduced the amount of nutrients available to new seed- lings at a critical stage of their growth, and that the inherent low levels of nutrients in the soil immobilized the recycling mechanism by tree roots. Other results suggest that alley cropping may not be effective under moisture-stressed conditions. In a four-year study carried out at the Interna- tional Crops Research Institute for Semi-Arid Tropics (ICRISAT) near Hyderabad, India, hedgerow species edged out the crops when there was limited moisture, resulting in reduced crop yields (Corlett, Ong, and Black 1989; ICRISAT 1989; Ong et al. 1989; Rao, Sharma, and Ong 1990). Similar observations have been reported from semiarid areas in northwestern Nige- ria (Odigie, Abu, and Adeola 1989) and Kenya (Nair 1987; ICRAF 1989; Coulson et al. 1989). A six-year study in northwestern India showed that maize, black gram, and cluster bean yields were lower when these crops were alley-cropped with Leucaena hedgerows than when grown in pure stands (Mittal and Singh 1989). The green fodder and fuelwood yields of Leucaena were also lower under alley cropping than under non-alley-cropped hedgerows. However, instead of returning the Leucaena prunings to the soil as green manure, farmers took them away as fodder. Many studies on alley cropping are now being undertaken in various parts of the tropics, and in the next few years there is likely to be a rapid increase in the amount of data available. As more data become available, the interpretation of the data will doubtless become more refined and con- sistent. In interpreting the results that have been produced so far, some experts have gone to great lengths to use the data to defend alley cropping, others to denigrate it. But the review of research results just presented indi- cates clearly that the merits or demerits of alley cropping cannot be judged on the basis of any single criterion or of short-term results. Benefits other than crop yield, such as the improvement of soil fertility and the yield of fuelwood and fodder, must be carefully weighed against drawbacks, such as labor requirements or pest management problems. 412 MANAGING THE WORLD'S FORESTS A key issue is ecological adaptability. Many research results suggest that although alley cropping offers considerable potential in the humid tropics, it is not a suitable crop production technology for the semiarid tropics. The provision of nutrients through decomposing mulch, a basic feature of alley cropping, depends on the quantity of the mulch as well as on its quality and time of application. If the ecological conditions do not favor the production of sufficient quantities of mulch, there is no perceptible advantage in using alley cropping. Because of such limitations, alley cropping as it is known today is un- likely to be widely adopted in the semiarid tropics. This does not mean that agroforestry in general is unsuitable for these regions. Indeed, some of the best-known agroforestry systems are found iin the semiarid tropics; two examples are the system based on Acacia (Faidherbia) albida, found in the dry areas of Africa (Felker 1978; Miehe 1986) and the system based on Prosopis cineraria found in the dry areas of India (Mann and Saxena 1980; Shankarnarayan, Harsh, and Kathju 1987). Alley cropping can be appropriate for both low and high levels of pro- ductivity; if higher levels of crop productivity are the goal, fertilizer appli- cation is necessary under most conditions, and the efficiency of fertilizers can be substantially increased under alley cropping (Kang et al. 1989). In extremely acidic sandy soils, such as those in the Peruvian Amazon basin (TropSoils 1988; Szott, Fernandes, and Sanchez 1991), the success of alley cropping depends on the extent to which external inputs such as fertilizers are used. The choice of hedgerow species that can adapt to harsh conditions also is an important management consideration under such circumstances. By integrating small-ruminant production with alley cropping, the In- ternational Livestock Centre for Africa (ILCA) project in Ibadan, Nigeria, has developed the alley-farming concept (Sumberg et al. 1987) in which prunings from the hedgerows provide high-quality supplementary fodder. So alley farming can be defined as the planting of arable crops between hedgerows of woody species that can be used for producing mulch and green manure to improve soil fertility and produce high-quality fodder. Home Gardens The term home garden has been used rather loosely to denote practices ranging from growing vegetables behind houses, to complex multistoried systems. Various authors have used numerous terms to denote these prac- tices, including mixed-garden horticulture (Terra 1954), mixed garden or house garden (Stoler 1975), home-garden (Ramsay and Wiersum 1974), Javanese homegarden (Soemarwoto et al. 1976), compound farm (Lagemann 1977; Okafor and Fernandes 1987), kitchen garden (Brierley 1985), house- hold garden (Vasey 1985), and homestead agro:forestry (Nair and Sreedharan 1986; Leuschner and Khalique 1987). The Javanese words pekarangan and Talun-kebun are often used interchangeably with home garden. In an analysis AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 413 of the structure and function of some tropical home gardens, Fernandes and Nair (1986) defined home gardens as "land-use practices involving deliber- ate management of multipurpose trees and shrubs in intimate association with annual and perennial agricultural crops and invariably livestock within the compounds of individual houses, the whole crop-tree-animal unit being intensively managed by family labor." Although Javanese home gardens provide an excellent example of the diversity of the structure and function of tropical home gardens, there are several other types of home gardens in other geographical locations, each with its characteristic features. In fact, home gardens can be found in almost all tropical and subtropical ecozones where subsistence land-use systems predominate. Although the choice of plant species in home gardens is determined largely by environmental and socioeconomic factors, dietary habits of people, and market demands of the locality, the species composition is remarkably similar among home gardens in various places. Fruit trees and other food- producing trees predominate. Apart from providing a steady supply of various types of edible products, these fruit-and-food trees also combine well, biologically and environmentally, with other components of the sys- tem (Nair 1984). Fruit trees such as guava, rumbutan, mango, and mango- steen, and other food-producing trees such as Moringa and Sesbania grandi- flora dominate the Asian home garden, whereas indigenous trees that produce leafy vegetables (Pterocarpus), fruit for cooking (Dacroydes edulis), and condi- ments (Pentaclethra macrophylla) dominate the West African compound farm. Food production is the primary function of most home gardens, and a large proportion of the production from the home gardens is consumed by the gardeners themselves. In fact the various food crops from the home gardens may provide a substantial proportion of the calorific and nutritive requirement of the local diet. Moreover, the combination of crops with dif- ferent production cycles and rhythms provides an uninterrupted supply of food products throughout the year. Depending on the climate and other environmental characteristics, there may be some peak and slack seasons for harvesting the various products, but generally there is something to harvest daily from most home gardens. Any marketable surplus can pro- vide a safeguard against failure of, as well as a security for the interval between, the harvests of other agricultural crops of the home gardens (e.g., rice in Java and Sri Lanka, coffee and maize in Tanzania, coconut and rice in southwestern India). All these harvesting and other operations require a relatively small amount of time from members of the family. Plantation-Crop Combinations Tropical perennial plantation crops occupy about 8 percent of the total arable area in developing countries. Some play a minor role in national economies; others produce high-value products for the international market 414 MANAGING THE WORLD'S FORESTS and are important, economically and socially, to the countries that produce them. The latter group includes oil palm, rubber, coconut, cacao, tea, cashew, and black pepper. Sisal and pineapple, although major crops, are not con- sidered here, because they differ from the other crops in terms of growth habits and duration. Traditionally, most of the major plantation crops were developed as monocultural production enterprises that required high labor input during harvesting and, in some cases, during processing. As a result, modern com- mercial plantations of crops such as rubber, oil palm, coffee, and tea are well-managed, profitable, land-use enterprises in the tropics, supported by· excellent research. Contrary to popular belief, substantial proportions of these crops are grown by smallholders (Ruthenburg 1980; Nair 1983; Watson 1983; Nair 1989). This is true of- • most of the cacao production in Ghana and Nigeria; • smallholder rubber plantations in Southeast Asia and Nigeria; • smallholder systems in which coffee is integrated with other crops and livestock in Ethiopia and East Africa; • smallholder coconut production in Ind:ia, the Philippines, Sri Lanka, and the Pacific islands in which the coconut crop is integrated with a large number of annual and perennial crops; and • cashew grown in a wide range of ecological situations, including wastelands where few other species thrive, in India, Tanzania, and Mozambique. Although research on these practices has been carried out since the 1970s, before agroforestry came of age, few results have been published. Most of the data that are available concern coconut-based systems in India (Nair 1979; Nelliat and Bhat 1979), Sri Lanka (Liyanage et al. 1989) and the Far East and the South Pacific (Plucknett 1979; Smith 1983; Smith and Whiteman 1983; Steel and Whiteman 1980). The rationale for integrating coconut palm with other crops is that a number of shade-tolerant and economically useful species can be grown between or under coconut during different stages of growth of the coconut crop. Apart from the period between the 8th and 25th year of the palm's growth, the light reaching the understory in coconut stands is enough to permit the growth of other compatible species. The palm's rooting pattern is such that most of the roots are near the bole, and thus there is minimal overlap between the rooting system of the palm and that of the other crops. On the basis of these factors, Nair (1979) proposed several possible crop combinations with coconut palms of different age groups and evaluated their performance. Considerable research has also been done on coffee/shade and cacao/ shade tree combinations, largely by CATIE in Costa Rica. Much of this research has concentrated on nutrient-related issues. A long-term replicated experiment, established in 1977 and known as "La Montana," has produced significant data on topics such as organic matter, nutrient-related cycles, AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 415 litter fall, and water infiltration. The tree species used in this experiment are Erythrina poeppigiana, which is periodically cut back, and a valuable timber species, Cordia alliodora, which is periodically thinned (Alpizar 1985; Alpizar et al. 1986; Fassbender et al. 1988; Heuveldop et al. 1988; Imbach et al. 1989). In a study comparing the two species, Beer (1987, 1989) showed that E. poeppigiana, when pruned two or three times a year, can return the same amount of nutrients to the litter layer as are applied to coffee plantations via inorganic fertilizers, even at the highest recommended rates for Costa Rica (270 kilograms of nitrogen per hectare per year, 60 kilograms of potassium per hectare per year). The annual nutrient return in this litter fall represents 90 to 100 percent of the nutrient store in the aboveground biomass of E. poeppigiana. In the case of C. alliodora, which is not pruned, nutrient storage in the tree stems, particularly of potassium, may limit both crop and tree productivity. Thus in fertilized plantations of cacao and coffee, litter pro- ductivity of shade trees is even more important than nitrogen fixation. Among other plantation-crop combinations that have been described are crops grown with cashew and coconut on the Kenyan coast (Warui 1980); crops grown with plantation crops in northeastern Brazil (Johnson and Nair 1984) and in Bahia, Brazil (Alvim and Nair 1986); crops grown with babassu palm in Brazil (May et al. 1985); crop associations with arecanut palm in India (Bavappa, Nair, and Kumar 1982); and crop associations with oil palm and rubber in West Africa (Watson 1983). Most of these are quali- tative and analytical descriptions of existing systems, and thus do not con- t<>in quantitative data based on rese<>rch investigations. Multistory Tree Gardens Multistory tree gardens are mixed tree plantations consisting of forest species and other commercial tree crops, forming a forestlike system. As opposed to home gardens, which surround individual houses, these tree gardens are usually established away from houses, and most are on com- munally owned lands surrounding villages with dense cluster of houses as in Indonesia (Java and Sumatra). Various forms of tree gardens can be found, including the following: • tree gardens (kebun or talun) of Java (Wiersum 1982) and agroforestry garden systems of Sumatra (Michon, Mary, and Bompard 1986); • compound farms (gardens) of southeastern Nigeria (Okafor and Fernandes 1987); • Kandyan Forest Gardens of Sri Lanka (Jacob and Alles 1987); and • crop combinations with cacao and other plantation crops in southeast Bahia, Brazil (Alvim and Nair 1986). The characteristics and functions of all these tree-gardening systems are often similar, although their relative importance may change from one sys- 416 MANAGING THE WORLD'S FORESTS tern to another. Wiersum (1982) lists the following common characteristics of tree gardens: • The tree gardens are characterized by a large variety of mostly multi- purpose plants in various vegetation layers, which provide for good use of water, nutrients, and sunlight. This variety ensures production of different materials throughout the year. • Most systems are dominated by perennial rather than annual crops; as a result there is a relatively high ratio of nutrients stored in the vegetation to those stored in the soil, which ensures an effective nu- trient cycle and relatively small hazard for leaching and erosion. An effective nutrient status is further maintained by the uptake of min- erals through deeply rooted perennials from deeper soil layers and effective catchment of mineral inputs by rain and by nitrogen fixation of leguminous species. • Most tree gardens form a part of a whole farm system, which also comprises annually cultivated fields. Normally, the latter are used to produce staple, high-calorie food stuffs (rice, maize, cassava), while the tree gardens are used to produce highly nutritious supplemen- tary products (proteins, vitamins, minerals), medicinal plants and spices, firewood, forage crops, and construction wood. Fruit trees also are an important component of tree gardens. • Most tree gardens are used to produce a small, continuous flow of these supplementary products for subsistence and a possible small surplus for sale to local markets. Higher production and marketing levels may be attained in a time of sudden need. • Although the general cultivation practices are rather standard, tree gardens vary with climate and soil, as well as with socioeconomic conditions. • Tree gardens contribute to the general cash economy of the farmers through the sale of various (mostly nonedible) commercial products. • Tree gardens often form an efficient buffer zone around protected forests. Windbreaks and Shel1terbelts Farmers throughout Africa use windbreaks to protect crops, water sources, soils, and settlements. Hedgerows.of Euphorbia tirucalli protect maize fields and settlements in the dry savannas of Tanzania and Kenya. Tall rows of Casuarina line thousands of canals and irrigated fields in Egypt. In Chad and Niger, multispecies shelterbelts protect wide expanses of cropland from desertification. These practices are not new, but the design of multipurpose windbreaks for smallholdings will require new agroforestry skills. The benefits of windbreaks at a given site depend on the distance be- tween windbreaks, the species used, and other site-specific management AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 417 practices; the value of the subsidiary products also should be taken into account. Before windbreaks are planted, however, the benefits (increased crop yields, soil improvements, and economic by-products) must be weighed against direct costs (such as labor and planting materials) and other disad- vantages (such as the amount of land, land that windbreaks will take up, land that would otherwise be used for crop production, and the competition between crops and windbreak species for water, light, and nutrients). The distance between windbreaks is determined mainly by the height of the tallest trees in the row. A properly designed windbreak can protect a field at least 10 times as long as the height of the tallest trees. The protec- tive influence will diminish with distance from the windbreak. A more per- meable windbreak will shelter a longer stretch of cropland than a dense windbreak. The most effective windbreaks provide a semipermeable barrier to wind over their full height. Because tree shapes change as they grow, it is usually necessary to mix several species with different growth rates, shapes, and sizes in three or more rows. Some fast-growing species, such as Eucalyptus, Cassia, Prosopis, Leucaena, and Casuarina, should be used to establish the desired effect as rapidly as possible. However, no tree will grow if it is ill- adapted to the environmental conditions of the site. In addition, some of these species are shorter-lived than slower-growing trees. Fast- and slow- growing species should be mixed to extend the useful life of the windbreak. Mixing species also provides protection against attack from diseases or pestS that can easily destroy the single-species stands. A fully developed windbreak can yield wood, fruit, fodder, fiber, and honey for sale and home use. Where animals are allowed to graze nearby, some of the lower, outer trees or shrubs should be unpalatable; for example, the leaves of neem, Azadirachta indica, are unpalatable, and this species has been used in Niger to protect windbreaks from livestock damage. Fodder species should be grown near the center or along an inside edge, where they are not exposed to animals but can be cut by hand. Species should be used selectively, even those that have been used widely in windbreaks. Eucalyptus should not be planted alone, because it has a sparse understory and may limit water availability and crop productivity in the vicinity. Azadirachta indica is known to shade crops and thus to reduce the land available for crop production. Successful windbreaks have incorpo- rated such unlikely trees as cashew and indigenous Acacia. The species se- lected must fit together as a group into the overall design; this design, in tum, must suit the local landscape and land-use system. Although diversity is important, the choice of species must also take into account the form, size, and growth rate required to establish an effec- tive windbreak, as well as the production priorities of the local people. Environmental hazards such as insect pests (especially termites), wild and domestic animals, poor soil, and drought will narrow this choice. Water management, especially during establishment, will be important, particu- 418 MANAGING THE WORLD'S FORESTS larly in dry environments; microcatchments, hand watering, or irrigation should be planned. Although little information is available on the quantities of wood pro- duced from trees growing in windbreaks for use as fuelwood, building poles, and other purposes, some preliminary results are encouraging. In windbreak tests in l\t!ajjia Valley in Niger, which has favorable soils and a mean annual precipitation of 125 millimeters, the yield of usable firewood from Azadirachta indica averaged 5 kilograms per year. Thus two rows of trees, each row 100 meters long and trees spaced 4 meters apart, would provide 250 kilograms of fuelwood (5 kilograms times 25 trees times 2 rows), or enough to meet the needs of a family of five for almost two months. This windbreak would protect about 1 hectare of cropland. If extended to protect 6 hectares, the windbreak would produce enough fuelwood to meet the family's annual requirements. However, the wood cannot be harvested until several years after planting. Cashew trees used in a windbreak in Senegal are yielding a fair amount of fruit and nuts. Although not in sufficient quantity and quality to be commercially viable on a large scale, these by-products are an important addition to local diets. Acacia scorpioides trees planted in windbreaks in Niger are now producing seed pods used for traditional leather tanning. Because there is a steady market for this product, the windbreaks make a modest but important contribution to the local economy. In other cases where Prosopis species are used in windbreaks, seed pods are collected daily to supplement livestock feed and some are sold in the local markets. In northwestern China, shelterbelts of Paulownia have been planted to stop desert encroachment. A 21 to 55 percent decrease in wind speed was measured in the. protected area, together with an increase of 12.5 percent in air humidity and 19.4 percent in soil humidity (in the top 50 centimeters). Maximum summer and winter temperatures were reduced, crop yields increased, and wood was produced. The reported effects of windbreaks on crop yields vary considerably. In some cases, grain yields have increased significantly; in others, the compe- tition for water and light, the land "lost" to the tree planting, or changes in the microclimate have slightly reduced crop yield. The effect on yield de- pends to a large extent on the design of the windbreak, the particular crop involved, and the environment. As a resull, multiple tree products and long-term soil conservation should be considered as the primary benefits. In the Sahel it appears that millet and sorghum yields in fields protected by windbreaks of Azadirachta indica can be as much as 23 percent higher than in the unprotected fields nearby (Bognettean-Verlinden 1980). In a year with poor rainfall, even relatively small differences in crop yields can be signifi- cant for the local population. It was estimated that pollarding these wind- breaks every four years would bring Majjia Valley residents US$800 worth of construction poles and wood per kilometer of windbreak (USAID 1987). AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 419 Use of Trees in Soil Conservation and Reclamation Agroforestry can help alleviate soil erosion through the supplementary or direct use of trees and shrubs. In supplementary use, the trees and shrubs are not the primary means of checking erosion, but support other soil con- servation structures through the stabilizing effect of the tree root system, while making some productive use of land. Direct use encompasses increased soil cover by trees and shrubs, permeable hedgerow barriers, natural terrace formation by soil accumulation upslope of hedgerows, and increased soil resistance to erosion by maintenance of organic matter. Young (1989) provides an excellent review of the role of agroforestry in soil conservation and identifies the agroforestry practices that can be used in erosion control under various conditions. The most widely applicable technologies are alley cropping (and other forms of hedgerow intercrop- ping), plantation-crop combinations, multistory tree gardens, home gardens, and shelterbelts and windbreaks; the ecological adaptability of these differ- ent practices was shown on exhibit 15-1. Other Agroforestry Practices The other agroforestry practices listed in exhibit 15-1 are not discussed in detail here mainly because there is, as yet, little scientific information to draw upon. Substantial qualitative descriptions, however, are available on the silvopastoral systems and practices (Le Houerou 1980, 1987; von Maydell 1987; Rocheleau et al. 1988; Baumer 1987), the role of indigenous trees as food producers in agroforestry systems (Nair 1989, 1990), and the potential of agroforestry in combatting desertification (Baumer 1987). The exploita- tion of these practices and of the indigenous knowledge concerning their production and processing has wide implications, not only in terms of food security and environmental protection but also with regard to the conserva- tion and use of genetic resources to meet current and future needs. The main difficulty in basing large-scale development plans on these practices is that the available qualitative descriptions by themselves often do not pro- vide the necessary technical basis for preparing sound projects that envis- age large-scale investments. The analysis of the state of the art of agroforestry practices can be sum- marized as follows: • Because limited scientific studies on agroforestry have been done, the potential of agroforestry remains vastly underexploited. • The little scientific material available supports the merits of these indigenous practices and points to several possibilities for improve- ment. • Prominent agroforestry practices demonstrate distinct patterns of ecological adaptability. 420 MANAGING THE WORLD'S FORESTS AN ECOZONE APPROACH TO AGROFORESTRY DESIGN One of the essential characteristics of agroforestry systems is their site- specificity, Although only a few distinct agroforestry practices make up the innumerable agroforestry systems in different places, a single practice takes various forms in different places depending on site-specific biophysical and socioeconomic conditions, It is possible, however, to identify certain types of land-use constraints that are characteristic of a broad ecological region and to devise agroforestry interventions to address those constraints, This is the essence of the ecozone approach to agroforestry design. Agroecological Analysis of Tropical Agroforestry Systems · On the basis of an evaluation of the ecological spread of agroforestry systems in different parts of the tropics, Nair (1989) concluded that the type of agroforestry system found in a particular area is determined to some extent by agroecological factors. However, several socioeconomic factors, such as human population pressure, availability of labor, and proximity to markets, also come into play, resulting in considerable variation among systems operating in similar or identical agroclimatic conditions. Sometimes, socioeconomic factors take precedence over ecological considerations. Even in the case of systems that are found in most ecological and geographical regions, such as shifting cultivation and taungya, numerous variants are specific to certain socioeconomic contexts. As a general rule, it can be said that although ecological factors determine the major type of agroforestry system in a given area, the complexity of the system and the intensity with which it is managed increase in direct proportion to the population inten- sity and land productivity of the area. The multispecies, multistoried home garden systems serve to illustrate some of these points. Although these systems are found mainly in humid lowlands, they are also common in pockets of high population density in other ecological regions. In their analysis of the structural and functional aspects of 10 home garden systems in different ecological regions, Fernandes and Nair (1986) found that although the average size of a home garden unit is less than 0.5 hectare, the units generally consist of a large number of woody and herbaceous species. Moreover, the unit is carefully structured so that the species form three to five canopies at varying heights, with each component having a specific place and function within the overall design. Agroecological factors have a considerable bearing on the functional emphasis on agroforestry practices. For example, the primary functions of agroforestry practices in sloping lands are erosion control and soil conserva- tion. In wind-prone areas, the emphasis is on windbreaks and shelterbelts; in areas with a fuelwood shortage the emphasis is on fuelwood production. There are also specific agroforestry approaches for the reclamation of de- graded lands or wasteland (e.g., land that has been badly eroded or over- AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 421 grazed, or contains high levels of salinity or alkalinity). The preponderance of home gardens and other multispecies systems in fertile lowlands and areas with high agricultural potential at one end of the ecological scale and extensive silvopastoral practices at the other end, with various systems in between, indicates that the ecological potential of an area is the prime deter- minant of the distribution and extent of adoption of specific agroforestry practices. The ecological and geographical distribution of the major agroforestry systems in the world has been schematically presented by Nair (1989). How- ever, caution must be exercised in interpreting such "agroforestry maps," because they aim to show general distribution patterns and thus include only those areas in which specified agroforestry systems are abundant. There are innumerable location-specific agroforestry practices in the tropics that, although important in certain respects, are not significant enough in terms of overall economy and land-use pattern of the area in which they operate to warrant inclusion on a global map. Conversely, some practices, such as "multipurpose trees on farmlands," are found in almost all ecological and geographical regions, but only a few of them-for example, the arid zone systems involving Acacia (Faidherbia) albida (Miehe 1986; Poschen 1986) and Prosopis (Shankarnarayan et al. 1987)-are classified as agroforestry systems and included on an agroforestry map. A significant feature that emerges from this type of ecological and geo- graphical analysis of tropical agroforestry systems is that, irrespective of the sociocultural differences in different geographical regions, the major types of agroforestry systems are structurally similar in areas with similar or iden- tical ecological conditions. Main Ecological Zones of Agroforestry Significance The main physical parameters that determine the ecology of a location in the tropics are rainfall (quantity and distribution) and temperature. Alti- tude is also important because it influences temperature and land relief characteristics. The Agroecological Zones Inventory of the Food and Agri- cultural Organization (FAO) divides the tropics into "warm," "moderately cool," and "cold" zones depending on the climatic (mainly temperature) regime during the "growing period." These growing periods refer to a few selected agricultural crops that are seasonal and therefore do not necessarily reflect the growing periods of perennial species (although it can be argued that in most cases the agricultural growing period also represents the peak growing seasons of the perennial species). There are a number of ways to delineate agroecological zones, but for the purpose of this chapter three broad zones can be considered: (1) humid and subhumid lowland tropics (humid tropics); (2) semiarid and arid lands (drylands); and (3) tropical highlands (highlands). The main characteristics of these zones are described in the sections that follow. 422 MANAGING THE WORLD'S FORESTS A Matrix of Agroecological Zones Versus Agroforestry Practices By far the most important ecological region-in terms of the total hu- man population it supports and its area and diversity of agroforestry and other landcuse systems-is the humid lowlands, characterized by hot and humid climates for all or most of the year and an evergreen or semievergreen vegetation. These climatic conditions favor rapid growth of a large number of plant species. Various forms of home gardens, plantation-crop combina- tions, and multilayer tree gardens are common in densely populated areas. In tropical lowlands with low population density (e.g., the low selvas of Latin Americas), trees on rangelands and pasltures, other silvopastoral sys- tems, improved fallows, and multipurpose woodlots are common. The com- mon land-use problems in this zone are rapid deforestation, resource-de- pleting shifting cultivation, acidic and strongly leached soils, and declining soil and crop productivity. Given the rapid growth of vegetation in this zone, improved tree fallows and alley cropping (and other forms of hedgerow intercropping), especially when contour-aligned on slopes, are promising technologies. The subhumid tropics are also grouped under the broad category of humid lowlands in this discussion, but these areas have a distinct dry sea- son, extending sometimes to several months. Thus the main land-use prob- lems include the clearance and degradation, through overcutting, of natural woodland, declining soil fertility, pasture degradation on grazing lands, erosion, shortages of fuelwood and fodder, and periodic drought, especially toward the drier margins of this ecozone. Trees are particularly important for maintaining soil fertility. Systems of trees on cropland, boundary plant- ing, trees on erosion-control structures, live fences, and fodder banks are the technologies most applicable in this zone; the apparent potential of hedgerow intercropping is the subject of current research. There appears to be some potential for developing home gardens, which are found in this zone but are not widespread. The semiarid tropics and subtropics, including the Sahel in Africa, cerrado of South America, and large areas of the Indiian subcontinent, have severe problems of drought, pasture degradation, and fuelwood shortage. Here, systems of trees on cropland, boundary planting (especially for fuelwood), windbreaks and shelterbelts, live fences, fodder banks, and trees on range- land or pastures are important. Hedgerow int,ercropping also may be appli- cable in the less drought-prone parts of this zone, although the problem of competition for soil moisture will need special attention. In tropical and subtropical highlands,, the areas with significant agroforestry potential are humid or subhumid; areas with dry climates have low potential and are not considered here. In the humid and subhumid highlands, where human and animal populations tend to be. dense, the of- ten inherently favorable natural resource base is rapidly deteriorating. Acute problems associated with erosion and declining soil fertility are compounded by drought and fuelwood shortages in areas where degradation is already AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 423 advanced. However, the highland environment varies considerably, and farm forestry is traditional in many of the remaining high-potential areas. Tech- nologies such as fruit trees, live fences, and boundary planting are appro- priate, while adapted forms of hedgerow intercropping on the contours are needed for the sloping land that is characteristic of this zone. Various forms of terracing, and the use of trees on grass or stone bunds, may be appropri- ate on such land. Exhibit 15-2 summarizes the types of agroforestry systems appropriate for the three major ecological regions of the tropics. Designing appropriate agroforestry systems for each ecozone involves dealing with many issues, ranging from the species to be used and their management to socioeconomic considerations. Once the ecological constraints have been analyzed and the type of agroforestry practices that can address the constraint identified, it is important to choose the components (trees and, mostly, crops) that offer the best potential and adaptability for the project area. Substantial information is available on the attributes of various woody components (e.g., Nair et al. 1984; von Carlowitz 1987) and herba- ceous crops (e.g., Nair 1980). Considering the complexity and diversity of agroforestry systems on the one hand, and the need for recommendations on site-specific agroforestry practices to be available easily for a vast num- ber of decision makers at all levels on the other, it is doubtful whether any expedient approach other than the computer-aided, knowledge engineering application of expert systems would be feasible. Warkentin et al. (1990) have developed a Knowledge-Based Expert System (KBES) for design of EXHIBIT 15-2. Agroforestry Interventions Recommended for Major Ecological Zones of the Tropics Humid Lowlands Improved fallow Modified taungya Alley cropping Home gardens Plantation-crop combinations Multilayer tree gardens Multipurpose trees on farmlands Semiarid Lowlands Silvopastoral systems Windbreaks and shelterbelts Multipurpose trees on farmlands Fuelwood lots Fodder banks Highlands Agroforestry for soil conservation Silvopastoral systems Plantation-crop combinations Multipurpose tree woodlots 424 MANAGING THE WORLD'S FORESTS alley cropping, which illustrates the opportunities and possibilities in ap- plying this technique in agroforestry systems design. Exhibit 15-3 illustrates an agroforestry I agroecological matrix for the tropics and subtropics. EXHIBIT 15-3. Agroforestry/Agroecological Matrix for the Tropics and Subtropics Altitude (m) LEGEND 500 1,000 1,500 IA] lmfcroved tree e .§. I 1,500 !ID @] IQ] IHI III > t: f/) z fal ows [!] Improved taungya [Q] Alley cropping c 0 !RI "' c z Multipurpose [QJ trees and shrubs E ·a !ID Q] 0 on crop lands ·r; ~ ...J [[] Planti;ltio~-crop !! ::> combinations Q. 1,000 !ID @ 11. Oi 0 Home gardens, 11. [fJ multilayer tree " c c !RI gardens ". Q] !QI Trees in s~il ."' I! ~ 500 conservation [H] Shelterbelts/ 1 ITJ windbreaks l Fodder banks QJ FuelwoOd lots IRl Trees on pasture/ C~ rangelands Humid Semiarid & Highlands [h] Multipurpose tree woodlots lowlands arid lands Constraints to Agroforestry llmplementation The constraints to agroforestry implementation are many, but they can be grouped into three broad categories: (1) technological, (2) economic and financial, and (3) institutional, each comprising a number of issues. These constraints, discussed in detail in a recent World Bank study (Nair 1990), are summarized in this section. Technological Constraints The technological constraints originate partly in the historically sepa- rate development of agriculture and forestry, and in the consequent re- search emphasis on monocultural production systems of agriculture and AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 425 forestry. Farmers have always used trees and shrubs for a variety of pur- poses on farms and grazing lands in developing countries. These types of farming systems were, however, largely ignored by conventionally trained experts, who are usually interested only in the disciplines in which they are trained. Only recently have some experts begun to take a multidisciplinary approach toward resolving these problems, but many professionals still scorn agroforestry for its lack of scientifically proven research results. This relatively recent-and still partial-realization of the importance of agroforestry by the scientific community is reflected in the state of the art of agroforestry technologies. Despite the commendable work that has been done by institutions such as ICRAF, CATIE, and several national institu- tions, and despite the very valuable descriptions of successful agroforestry systems everywhere, hard facts about the structure, function, dynamics, and response to manipulation of most agroforestry systems and technologies are very scarce. Hence it is difficult to extrapolate agroforestry systems from one place to another. The multipurpose trees and shrubs that are fundamen- tal to all agroforestry systems are rather primitive in a scientific sense, in that they represent completely unknown germplasm. Rarely have scientists attempted to explore systematically the genetic variation of these species and to collect and improve them for use in agroforestry. This lack of a reliable set of agroforestry technologies seriously con- strains agroforestry development. More research and trials are needed on optimal (in biophysical, social, and economic terms) combinations of woody and nonwoody components of land-use systems. Researchers need to apply scientific methods that are already being used in agriculture and forestry to improve agroforestry components and systems. Economic and Financial Considerations Economic considerations are among the most important factors that will determine the value of agroforestry to the land user. In fact, the great majority of whatever agroforestry research has been done to date has con- centrated on the biological and physical factors that affect productivity. Inadequate attention has been paid to the economic value of directly quantifiable agroforestry outputs including fodder, green manure, fuelwood, timber, and significant indirect effects such as enhanced soil fertility and watershed protection. There is, therefore, a serious lack of reliable informa- tion based on actual farm and field conditions. Furthermore, whereas traditional agroforestry systems may have proved economically viable under the conditions in which they evolved, increasing land pressures, changing social perceptions, and new land-use options all call for new economic evaluations of many existing systems. Yet another problem that adds to the complexity of the issue is the tendency to apply the term agroforestry to many types of existing or poten- tial farming systems, ranging from "pure" agriculture to "pure" forestry. For example, an appraisal of World Bank funding for agroforestry projects 426 MANAGING THE WORLD'S FORESTS from 1978 to 1987 (Spears 1987) refers to many projects that can be argued to be only peripherally relevant to agroforestry (Jickling 1989). My own review (Nair 1990) shows that, among all agroforestry tech- nologies, alley cropping has received the most attention in economic stud- ies, too, as with biological investigations. These studies have confirmed the economic feasibility of alley cropping in humid and subhumid tropics, and its relative infeasibility in areas with high labor cost, low annual rainfall, or extended dry seasons. The other major agro:forestry technologies on which economic studies have been made include silvopastoral systems and live fences, multistory cropping and plantation-crop combinations, and home gardens. Despite their shortcomings, these studies have shown that although there are some situations in which agroforestry might not have economic advantages, there are many others in which economic benefits are clear. In general, however, economic and financial studies in agroforestry have been grossly inadequate in relation to their indicated potentials, and this is a serious constraint to agroforestry implementation. Institutional Constraints Another important constraint to the implementation of agroforestry projects is the lack of a defined institutional niche for agroforestry. Institu- tions dealing with research, education, training, and extension in land-use disciplines such as agriculture, forestry, and range science have all been set up along disciplinary lines, linked to the respective government ministries and departments; as a result, each institution is charged with maximization of production, management of one or a set of commodities, or alleviation of a specific problem. Even the International Agricultural Research Centers of the Consultative Group of International Agricultural Research (CGIAR) system are no exception in this regard. Thus, the land-use institutions of today are usually not mandated to deal with interdisciplinary activities such as agroforestry. This problem also afflicts the operation of large development projects, which often are attached to a particular ministry or department and thus required to adhere to the mandate of that ministry or department. Legal and social issues form another set of institutional constraints. Land (and tree) tenure issues, forest (and tree) protection laws, and legal and societal norms for management and exploitation of community lands and woodlots all seriously affect the implementation of agroforestry projects. The lack of clear and definite government policies or the lack of coordi- nation of existing policies also can seriously impede agroforestry programs. A clear example is the experience of the USAID-funded Agroforestry Out- reach Project in Haiti, which is implemented by private voluntary organiza- tions with little or no involvement of the Haitian government. Although there are clear advantages in having few official ties with the government, AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 427 there are also serious disadvantages. The lack of strong government policies reflects a lack of understanding of the issues by the policymakers and their technical advisers. Educating these policymakers about the importance of agroforestry and enlisting their strong support for agroforestry programs are therefore of paramount importance. SUMMARY Agroforestry holds considerable promise as a practical land-manage- ment approach in developing countries to deal with issues of production (of multiple outputs and benefits) and sustainability (of production base). Agroforestry systems are abundant throughout developing countries in most ecological and geographical regions. These systems are complex, diverse, and extremely site specific. Because only a few of the traditional agroforestry systems have been scientifically studied, such systems are not well understood. The little re- search that has been done indicates their scientific merits and points to several possibilities for improving them. Despite the differences among the innumerable agroforestry systems throughout the world, distinct similarities are evident among systems oper- ating in areas with similar ecological characteristics. On the basis of these similarities, it is possible to id7ntify about 20 distinct practices that consti- tute most of these systems. The scale and extent of adoption of the different agroforestry practices in any particular system will vary with biophysical and socioeconomic characteristics, but each practice has distinct patterns of ecological spread and adaptability. Considering the ecological spread of the different practices and the variety of major land-use problems in different ecological regions, a matrix of eco- logical conditions versus agroforestry practices can be a useful approach to designing agroforestry systems. The matrix provides a framework for de- veloping agroforestry recommendations in a wide variety of conditions. It can also be used for developing such models for any of the agroforestry practices, using different components that have varying ecological adapt- ability and uses. It is now well recognized that agroforestry is an extremely useful ap- proach to land management in all situations that warrant increased partici- pation of local people and combine traditional and modern technologies. However, organized research in agroforestry is still in its infancy. It is con- strained by the lack of unclear methodologies and institutional constraints, and by the sheer multiplicity of factors that need to be taken into consider- ation. Some commendable research efforts are under way in various places, yet these initiatives are so new and so few in number that direct, field-tested research results are limited. The clear indication of the vast potential of agroforestry, coupled with limited scientific evidence of a direct nature, points to the need for increased 428 MANAGING THE WORLD'S FORESTS funding for agroforestry research. The matrix of agroecological conditions versus agroforestry practices could be a starting point for developing knowl- edge-based expert systems for the design of agroforestry systems. The use- fulness of this approach has been demonstrated in a recent work on the application of the expert system technique to alley-cropping design. The approach can be extended to various other agroforestry practices as well. REFERENCES Agboola, A.A., G.F. Wilson, A. Getahun, and C.F. Yamoah. 1982. "Gliricidia sepium: A Possible Means to Sustained Cropping." In Agroforestry in the African Humid Tropics, ed. L.H. MacDonald. Tokyo: United Nations University. Alpizar, O.L.A. 1985. "Untersuchungen uber den Stoffaushalt einigeragroforstlicher Systeme in Costa Rica." Doctoral dissertation, Georg-August University, Goettingen, Germany. Alpizar, L., H.W. Fassbender, J. Heuveldop, H. Foelster, and G. Enriquez. 1986. "Modelling Agroforestry Systems of Cacao (Theobroma cacao) with Laurel (Cordia alliodora) and Poro (Erythrina poeppigiana) in Costa Rica." Part I. "Inventory of Organic Matter and Nutrients." Agroforestry Systems 4:175-90. Alvira, R., and P.K.R. Nair. 1986. "Combination of Cacao with Other Plantation Crops: An Agroforestry System in South-East Bahia, Brazil." Agroforestry Systems 4:3-15. Baumer, M. 1987. "Agroforesterie et desertification." CTA, Wetteren, Belgium/ICRAF, Nairobi, Kenya. Bavappa, K.V.A., M.K. Nair, and T.P. Kumar, eds. 1982. "The Arecanut Palm (Areca catechu Linn.)." Central Plantation Crops Research Institute, Kasaragod, Kerala, India. Beer, J. 1987. "Advantages, Disadvantages, and Desirable Characteristics of Shade Trees for Coffee, Cacao, and Tea." Agroforestry Systems 5:3-13. _ _ _. 1989. "Litter Production and Nutrient Cycling in Coffee (Coffen arabica) or Cacao (Theobroma cacao) Plantation with Shade Trees." Agroforestry Systems 7:103-14. Benneh, G. 1972. "Systems of Agriculture in Tropical Africa." Economic Geography 48(3):244-57. (Accra University, Legon, Accra, Ghana). Bognettean-Verlinden, E. 1980. "Study of Impact of Windbreak in Majjia Valley, Niger." Agricul- tural University, Wageningen, The Netherlands. Boonkird, S.-A., E.C.M. Fernandes, and P.K.R. Nair. 1984. "Forest Villages: An Agroforestry Ap- proach to Rehabilitating Forest Land Degraded by Shifting Cultivation in Thailand." Agroforestry Systems 2:87-102. Brierley, J.S. 1985. "The West Indian Kitchen Gardens: A Historical Perspective with Current In- sights from Grenada." Food and Nutrition Bulletin (UNU) 7:52-60. Conklin, H.C. 1957. "Hununoo Agriculture." FAO Forestry Development Paper No.12. Rome: FAO. _ _ _. 1980. Ethnographic Atlas of Ifugao: A Study of Environment, Culture, and Society in Northern Luzon. New Haven, Conn.: Yale University Press. Cook, C.C., and M. Grut. 1989. Agroforestry in Sub-Saharan Africa: A Farmer's Perspective. Washing- ton, D.C.: World Bank. Corlett, J.E., C.K. Ong, and C.R. Black. 1989. "Modification of Micro climate in Intercropping and Alley-Cropping Systems." In Meteorology and Agroforestry, eds. W.S. Reifsnyder and T.Q. Darnhofer. Nairobi, Kenya: ICRAF/WMO/UNEP/GTZ. Coulson, C.L., D.N. Mungai, C.J. Stigler, P.W. Mwangi, and D.M. Njiru. 1989. "Studies of Sustain- able Crop Yield Improvement through An Agroforestry Intervention." Kenya Forest Research Institute, Nairobi. Unpublished. Dijkman, M.J. 1950. "Leucaena: A Promising Soil Er6sion Control Plant." Economic Botany 4:337-49. _ _ _. 1985. "Tree Growing by Rural People." FAQ Forestry Paper No. 64. Rome: FAQ. AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 429 Fassbender, H.W., L. Alpizar, J. Heuveldop, H. Foelster, and G. Enriquez. 1988. "Modelling Agroforestry Systems of Cacao (Theobroma cacao) with Laurel (Cordia alliodora) and Poro (Erythrina poeppigiana) in Costa Rica." Part III. "cycles of Organic Matter and Nutrients." Agroforestry Systems 6:4~2. Felker, P. 1978. "State-of-the-.Art: Acacia albida as a Complementary Intercrop with Annual Crops." Report to USAID. University of California, Riverside. Fernandes, E.C.M., and P.K.R. Nair. 1986. "An Evaluation of the Structure and Function of Tropical Homegardens." Agricultural Systems 21:279-310. Food and Agricultural Organization (FAO). 1984. Food and Fruit-Bearing Forest Species. Part 2. Ex- amples from South-eastern Asia. Rome: Food and Agricultural Organization. Getahun, A., G.F. Wilson, and B.T. Kang. 1982. "The Role of Trees in Farming Systems in the Humid Tropics." In Agroforestry in the African Humid Tropics, ed. L.H. MacDonald. Tokyo: United Na- tions University Press. Gregersen, H.M., and A. Contreas. 1979. ''Economic Analysis of Forestry Projects." FAO Forestry Paper No. 17. Rome: PAO. Heuveldop, J., H. Fassbender, L. Alpizar, G. Enriquez, and H. Foister. 1988. "Modelling Agroforestry Systems of Cacao (Theobroma cacao) with Laurel (Cordia alliodora) and Poro (Erythrina poeppigiana) in Costa Rica." Part II. "Cacao and Wood Production, Litter Production and Decomposition." Agroforestry Systems 6:37-48. Imbach, A.C., H.W. Fassbender, R. Borel, J. Beer, and A. Bonnemann. 1989. "Modelling Agroforestry Systems of Cacao (Theobroma cacao) with Laurel (Cordia alliodora) and Poro (Erythrina poeppigiana) in Costa Rica." Part IV. "Water Balance, Nutrient Inputs, and Leaching." Agroforestry Systems 8:267-87. International Council for Research in Agroforestry (ICRAF). 1989. Annual report for 1988. Nairobi, Kenya. International Crops Research Institute for Semi-Arid Tropics (ICRISAT). 1989. Annual report for 1988. Hyderabad, India. Jacob, V.J., and W.S. Alles. 1987. "Kandyan Gardens of Sri Lanka." Agroforestry Systems 2:281-92. Jickling, A.J. 1989. "Short-term Economic Benefits of Agroforestry: A Survey." Report to USDA Forestry Support Program (RSSA-BST-5519-R-2188). Washington, D.C. Johnson, D.V., and P.K.R. Nair. 1984. "Perennial-Crop-Based Agroforestry Systems in Northeastern Brazil." Agroforestry Systems 2:281-92. Kang, B.T., and B. Duguma. 1985. "Nitrogen Movement in Alley Cropping Systems." In Nitrogen in Fanning Systems in the Humid and Subhumid Tropics, eds. B.T. Kang, and J. van den Heide. Haren, The Netherlands: Institute of Soil Fertility. Kang, B.T., L. Reynolds, and A.N. Atta-Krah. 1989. "Alley Farming." Advances in Agronomy 43:315-- 59. Kang, B.T., G.F. Wilson, and T.L. Lawson. 1984. "Alley Cropping: A Stable Alternative to Shifting Cultivation." International Institute of Tropical Agriculture, Ibadan, Nigeria. Kang, B.T., G.F. Wilson, and L. Sipkens. 1981. "Alley cropping maize (Zea mays L.) and Leucaena (Leucaena leucocephala LAM) in Southern Nigeria." Plant and Soil 63:165--79. Lagemann, J. 1977. "Traditional African Farming in Eastern Nigeria." Africa Studien, Welforum Verlag, Munich, Germany. Lebar, F.M., G.C. Hickey, and J.K. Musgrave, eds. 1964. "Ethnic Groups of Mainland South-East Asia." Human Resources Area Files. New Haven, Conn.: Yale University Press. Le Houerou, H.N., ed. 1980. Browse in Africa: The Current State of Knowledge. Addis Ababa, Ethiopia: International Livestock Center for Africa. Le Houerou, H.N. 1987. "Indigenous Shrubs and Trees in the Silvopastoral Systems in Africa." In Agroforestry: A Decade of Development, eds. H.A. SteppIer and P .K.R. Nair. Nairobi, Kenya: ICRAF. Leuschner, W.A., and K. Khalique. 1987. "Homestead Agroforestry in Bangladesh." Agroforestry Systems 5:139-51. Liyanage, L.V.K., H.P.S. Jayasundera, D.T. Mathew, and D.N.S. Fernando. 1989. "Integration of Pasture, Fodder, and Cattle in Coconut Smallholdings." CORD (Sri Lanka) 5(2):53-59. 430 MANAGING THE WORLD'S FORESTS MacDicken, K.G., and N.T. Vergara, eds. 1990. Agroforestry: Classification and Management. New York: John Wiley. Mann, H.S., and S.K. Saxena, eds. 1980. "Khejri (Prosopis cineraria) in the Indian Desert: Its Role in Agroforestry." Monograph No. 11, Central Arid Zone Research Institute, Jodhpur, India. May, P.H., A.B. Anderson, J.M.F. Frazao, and J.M. Balik. 1985. "Babassu Palrnin the Agroforestry Systems in Brazil's Mid-North Region." Agroforestry Systems 3:275-95. Michon, G., F. Mary, and J. Bompard. 1986. "Multistoried Agroforestry Garden System in West Sumatra, Indonesia." Agroforestry Systems 4:315-38. Miehe, S. 1986. "Acacia albida and Other Multipurpose Trees on Fur Farmlands in the Jebel Marra Highlands, Western Dafur, Sudan." Agroforestry Systmts 4:89-119. Mittal, S.P., and P. Singh. 1989. "lntercropping Field Crops Between Rows of Leucaena leucocephala Under Rainfed Conditions in Northern India." Agroforestry Systems 8:165--72. Nair, P.K.R. 1979. Intensive Multiple Cropping with Coconuts in India: Principles, Programmes, and Prospects. Berlin/Hamburg, Germany: Verlag Paul Parey. _ _ _. 1980. Agroforestry Species: A Crop-Sheets Manual. Nairobi, Kenya: ICRAF. _ _ _. 1983. "Agroforestry with Coconuts and Other Tropical Plantation Crops." In Plant Research and Agroforestry, ed. P.A. Huxley. Nairobi, Kenya: ICRAF. _ _ _. 1984. Soil Productivity Aspects of Agroforestry. Nairobi, Kenya: ICRAF. _ _ _. 1985. "Classification of Agroforestry Systems." Agroforestry Systems 3:97-128. _ _ _. 1987. "Soil Productivity Under Agroforestry." In Agroforestry: Realities, Possibillties, and Potentials, ed. H.L. Gholz. Dordrecht, The Netherlands: Martinus Nijhoff. ______Jed. 1989. Agroforestry Systems in the Tropics. Dordrecht, The Netherlands: Kluwer Academic Publishers. _ _ _. 1990. "The Promise of Agroforestry in the Tropics." Technical Report. Washington, D.C.: World Bank. Nair, M.A., and C. Sreedharan. 1986. "Agroforestry Farming Systems in the Homesteads of Kerala, Southern India." Agroforestry Systems 4:339-63. Nelliat, E.V., and K.S. Bhat, eds. 1979. ''Multiple Cropping in Coconut and Arecanut Gardens." Technical Bulletin No. 3. Kasaragod, India: Central Plantation Crops Research Institute. Odigie, G.A., J.E. Abu, and A.O. Adeola. 1989. "Effect of Alley Width and Pruning Height on Crop Yield in Gambari, Southwestern Nigeria." Paper presented to the International Conference on Agroforestry, University of Edinburgh, July 1989. Okafor, J.C., and E.C.M. Fernandes. 1987. "Compound Farms (Homegardens): A Predominant Agroforestry System Involving Food and Fruit Trees with Crops and Small Livestock in the Humid Lowlands of Southeastern Nigeria." Agroforestry Systems 5:153-68. Okigbo, B.N., and R. Lal. 1979. "Soil Fertility Maintenance and Conservation for Improved Agroforestry Systems of the Lowland Humid Tropics." In Soils Research in Agroforestry, eds. H.A. Mongi, and P.A. Huxley. Nairobi, Kenya: ICRAF. Olofson, H. 1983. "Indigenous Agroforestry Systems." Philippine Quarterly of Culture and Society 1!:149-74. Ong, C.K., J.E. Corlett, R.P. Singh, and C.R. Black. 1989. "Above- and Below-Ground Interactions in Agroforestry Systems." Paper presented to the International Conference on Agroforestry, Uni- versity of Edinburgh, July 1989. Pickering, D.C., ed. 1988. "African Agricultural Research and Technological Development." Pro- ceedings of a high-level meeting in Feldafing, Germany, September 24--27, 1987. Washington, D.C.: World Bank. Plucknett, D.L. 1979. Managing Pasture Under Coconuts. Boulder, Colo.: Westview Press. Poschen, P. 1986. "An Evaluation of the Acacia albida-Based Agroforestry Practices in the Hararghe Highlands of Eastern Ethiopia." Agroforestry Systems 4:129-43. Raintree, J.B., and K. Warner. 1-986. "Agroforestry Pathways for the Intensification of Shifting Cul- tivation." Agroforestry Systems 4:39-54. Ramsay, D.M., and K.F. Wiersum. 1974. "Problems of Watershed Management and Development in the Upper Solo River Basin." Conference on Ecologic Guidelines for Forest, Land, and Water Resources, Institute of Ecology, Bandung, Indonesia. AGROFORESTRY SYSTEMS DESIGN: AN ECOZONE APPROACH 431 Rao, M.R., M.M. Sharma, and C.K. Ong. 1990. "A Study of the Potential of Hedgerow Intercropping in Semi-arid India Using a Two-Way Systematic Design." Agroforestry Systems 11:243-58. Roberts, N., ed. 1989. "Agricultural Extension in Africa: A World Bank Symposium." Washington, D.C.: World Bank. Rocheleau, D., F. Weber, and A. Field-Juma. 1988. Agroforestry in Dryland Africa. Nairobi, Kenya: ICRAF. Ruthenberg, H. 1980. Farming Systems in the Tropics. 2d ed. London: Oxford University Press. Salisbury, R.F. 1962. From Stone to Steel. London: Cambridge University Press. Shankarnarayan, K., i...N. Harsh, and S. Kathju. 1987. "Agroforestry Systems in the Arid Zone of India." Agroforestry Systems 5:69--88. Smith, M.A. 1983. "Evaluation of Tropical Grasses in Increasing Shade Under Coconut Canopies." Experimental Agriculture 19:153--61. Smith, M.A., and P.C. Whiteman. 1983. "Rotational Grazing Experiments Under Coconuts at Ligatu Estate, Russell Islands." Technical Report, Deparbnent of Agriculture, University of Queensland, Australia. Soemarwoto, 0., Z. Soemarwoto, Karyono, E.M. Soekartadredja, and A. Ramlan. 1976. "The Javanese Homegarden as an Integrated Agroecosystem." In Science for a Better Environment. Tokyo: Sci- ence Council of Japan. Spears, J. (1987). "Agroforestry: A Development Bank Perspective." Agroforestry: A Decade of Devel- opment, eds. H.A. Steppler and P.K.R. Nair. Nairobi: Kenya: ICRAF, 53--66. Steel, R.J.H., and P.C. Whiteman. 1980. "Pasture Species Evaluation: Pasture, Fertilizer Require- ments, and Weed Control in Solomon Islands." Technical Report, Department of Agriculture, University of Queensland, Australia. Steppler, H.A., and P.K.R. Nair, eds. 1987. Agroforestry: A Decade of Development. Nairobi, Kenya: ICRAF. Stoler, A. 1975. "Garden Use and Household Consumption Patterns in a Javanese Village." Ph.D. dissertation, Columbia University, Department of Anthropology, New York. Sumberg, J.E., J. Macintire, C. Okali, and A.N. Atta-Krah. 1987. Economic Analysis of Alley Farming with Small Ruminants. Addis Ababa, Ethiopia: International Livestock Centre for Africa. Szott, L.T. 1987. "Improving the Productivity of Shifting Cultivation in the Amazon Basin of Peru Through the Use of Leguminous Vegetation." Ph.D. dissertation, North Carolina State Univer- sity, Raleigh. Szott, L.T., E.C.M. Fernandes, and P.A. Sanchez. 1991. "Soil-Plant Interactions in Agroforestry Sys- tems." Forest Ecology and Management (in press). Terra, G.T.A. 1954. "Mixed-Garden Horticulture in Java." Malaysian Journal of Tropical Geography 4:33-43. Torres, F. 1983. "The Role of Woody Perennials in Animal Agroforestry." Agroforestry Systems 2:131- 63. TropSoils. 1988. Annual Report of the TropSoils Project for 1987. Soil Science Department, North Caro- lina State University, Raleigh. U.S. Agency for International Development (USAID). 1987. "Windbreak and Shelterbelt Technology for Increasing Agricultural Production." Washington, D.C.: USAID. Vasey, D.E. 1985. "Household Gardens and Their Niche in Port Moresby, Papua New Guinea." Food and Nutrition Bulletin (UNU) 7(3):37-43. von Carlowitz, P.G. 1987. "ICRAF's Multipurpose Tree and Shrub Information System." Agroforestry Systems 5:319-38. von Maydell, H.-J. 1986. "Trees and Shrubs of the Sahel: Their Characteristics and Uses." Eschborn, Germany: GTZ. _ _ _. 1987. ''Agroforestry in the Dry Zones of Africa: Past, Present, and Future." In Agroforestry: A Decade of Development, eds. H.A. Steppler, and P.K.R. Nair. Nairobi, Kenya: ICRAF. Warkentin, M.E., P.K.R. Nair, S.R. Ruth, and K. Sprague. 1990. "A Knowledge-Based Expert System for Planning and Design of Agroforestry Systems." Agroforestry Systems 11:71-83. Warui, C.M. 1980. "Research on Agroforestry at the Coast, Kenya." In Proceedings of the Kenya National Seminar on Agroforestry, ed. L. Buck. Nairobi, Kenya: ICRAF. 432 MANAGING THE WORLD'S FORESTS Watson,-G.A. 1983. "Development of Mixed Tree and Food Crop Systems in the Humid Tropics: A Response to Population Pressure and Deforestation." Experimental Agriculture 19:311-32. Weinstock, J.A. 1984. "Rattan: A Complement to Swidden Agriculture." Unasylva 36(143):16-22. Wiersum, K.F. 1982. "Tree Gardening and Taungya in Java: Examples of Agroforestry Techniques in the Humid Tropics." Agroforestry Systems 1:53-70. Wijewardene, S.R., and P. Waidyanatha. 1984. "Conservation Farming for Small Farmers in the Humid Tropics." Colombo, Sri Lanka: Department of Agriculture. World Bank. 1978. Forestry Sector Review. Washington, D.C.: World Bank. Yamoah, C.F., A.A. Agboola, and G.F. Wilson. 1986. "Nutrient Composition and Maize Performance in Alley Cropping Systems." Agroforestry Systems 4:247-54. Young, A. 1989. "Agroforestry for Soil Conservation." C.A.B. International, Willingford, Oxford, England, and ICRAF, Nairobi, Kenya. Zhaohua, Z., C. Mantang, W. Shiji, and J. Youxu, eds. 1991. "Agroforestry Systems in China." Chinese Academy of Forestry, Beijing and International Development Research Center, Singapore. 16 Production of Forest Products in Agricultural and Common Land Systems: Economic and Policy Issues J.E.M. Arnold T his chapter examines economic and policy issues relating to the produc- tion of forest products by rural people in conjunction with their agri- cultural and livestock systems. Production at this level has two main components: (1) the incorporation into its farming system of planted and managed trees of value to the farm household and (2) management of neigh- boring common pool resources to provide inputs needed to complement those available from on-farm resources. As the forests in most parts of the developing world become degraded and the demand for fuel, fodder, and similar products continues to increase, the importance of nonforest sources of production is growing. As expro- priation by the state, privatization, and encroachment reduce common prop- erty resources, and overuse degrades those resources that remain outside farming areas, reliance on on-farm resources increases. Production and use outside the forests are determined primarily by the factors that shape local agricultural patterns, which include agroecosystem characteristics, cropping and livestock patterns, population pressures, land tenure, income-earning opportunities, and agrarian transition. The discus- sion here, therefore, is organized within a framework that focuses on agri- cultural rather than forest systems. 433 434 MANAGING THE WORLD'S FORESTS The chapter does not deal with local production systems that take place wholly within the forest, namely, systems based on shifting cultivation on forest land, or on forest land that is locally owned but operated for forest outputs rather than agricultural inputs. These local systems are influenced more by policies and practices related to the use of forest land and forest resources than by agricultural policies and practices. The first part of this chapter reviews trends in the use of, and rural reliance on, forest products. The next part examines the role of common lands as a source of these products. The last part reviews trends in the growing and management of trees in farming systems. FOREST PRODUCTS AND THE RURAL HOUSEHOLD ECONOMY Although the categories overlap, forest products fall into roughly three groups: (1) household inputs, such as fuel, fodder, and food, that are used directly by households; (2) inputs into the agricultural system such as fod- der and mulch; and (3) products that are sources of rural household income and employment. Household Inputs For most rural people, foods derived from forests or from trees they maintain in their farming system add variety to diets, improve palatability, and provide essential vitamins, protein, and calories (Falconer 1989). The quantities of forest foods consumed may be Jess than the quantities of the main food staples, but forest foods often form an essential part of otherwise bland and nutritionally poor diets (exhibit 16-1). Forest and farm tree products are also widely used as snack foods be- tween meals, eaten while working in fields or herding, for example. They are particularly valued during the peak agricultural labor period, when less time is available for cooking and people consume more snack foods. In addition, forest and farm tree foods may be used to help meet dietary short- falls during particular seasons of the year; they help bridge hunger periods when stored food supplies are dwindling and the next harvest is not yet available. And in emergencies such as floods, droughts, famines, and wars, energy-rich foods such as roots, tubers, rhizomes, and nuts can provide an important buffer. Where people have had relatively unrestricted access to forests, forest food is often particularly important for poorer groups within the commu- nity. Although forest-gathering activities are not restricted to the poor, the poor are most likely to be affected by a reduction in the availability of such foods as the forest resource is reduced, degraded, or becomes inaccessible to them. PRODUCTION OF FOREST PRODUCTS. 435 EXHIBIT 16-1. Some Common Nutrition Problems and the Potential Role of Forest Food Forest Food with Potential Nutrient-Related Problem to Combat Deficiencies Protein-energy malnutrition Energy-rich food that is available during due to inadequate food consumption seasonal or emergency food shortages, (Causes reduced growth, especially nuts, seeds, oil-rich fruit and susceptibility to infection, tubers; e.g., the seeds of Geoffroea and changes in skin, hair, and decorticans, Ricinodendron rautanenil, and mental facility.) Parkia sp.; oil of Elaeus guineensis, babassu, palmyra, and coconut palms; protein-rich leaves such as baobab (Adansonia digitata); as well as wild animals (e.g. snails), insects, and larvae. Vitamin A deficiency Forest leaves and fruit that are good (In extreme cases causes sources of Vitamin A; e.g., leaves of blindness and death; Pterocarpus sp., Moringa oleifera, Adansonia responsible for blindness digitata, the gum of Sterculia sp., palm oil of 250,000 children per year) of Elaeus guineensis, bee larvae, and other animal food. (In addition, fats and oils are needed for the synthesis of Vitamin A.) Iron deficiency Wild animals including insects such as (In severe cases causes anemia, tree ants, mushrooms (often consumed weakness, and susceptibility to as meat substitutes), as well as forest disease,especially in women and leaves such as Leptadenia hastata, children.) Adansonia digitata. Niacin deficiency Forest fruit and leaves rich in niacin such (Common in areas with a maize as Adansonia digitata, fruit of Boscia staple diet; can cause dementia, senegalensis and Momordica balsamina, seeds diarrhea, and dermatitis.) of Parkia sp., Iruingia gabonesis, and Acacia albida. Riboflavin deficiency Forest leaves that are especially high in (Common throughout Southeast riboflavin, notably Anacardium sp., Asia; causes skin problems Sesbania grandiflora, and Cassia obtusifolia, among people with rice diets.) as well as wild animals, especially insects. Vitamin C deficiency Forest fruit and leaves that are especially (Common to people who good sources, including fruit of Ziziphus consume monotonous diets; mauritiana, Adansonia digitata, and increases susceptibility to Sclerocarya caffra. (Leaves such as Cassia disease, weakness.) obtusifolia and the gum of Sterculia sp. are also good sources of this vitamin.) SOURCE: Adapted from Falconer and Arnold (1989). 436 MANAGING THE WORLD'S FORESTS As new food products have penetrated rural markets and tastes have changed, however, the role that forest food plays in household nutrition has diminished. Some studies indicate that emergency uses of forest resources are dwindling, as people rely to a greater extent on purchased food. Thus, in many regions, forest food is no longer consumed, and knowledge about its use is vanishing. Elsewhere, the market for forest foods-for example, that for bush meat in West Africa-has grown rapidly. But even where consumption is not declining, the nutritional diversity of the gathered food may have decreased. The effects of declining consumption of forest food vary. In some cases these changes have led to a poorer-quality diet, most notably, as greater reliance on purchased food reduces dietary diversity. Perhaps the worst effect is the progressive reduction of poorer people's food options, espe- cially during seasonal and emergency hardship periods (Falconer 1989). Fuel shortages may influence the amourtt of food supplied or cooked. When there is less fuel or time for cooking, consumption of uncooked and reheated food may increase, perhaps causing a serious rise in disease inci- dence. Cooking is necessary to remove parasites, and few uncooked foods can be digested properly. A decrease in the number of meals provided may have a particularly damaging effect on child nutrition, because children may be unable to consume enough of the often overstarchy staple food in one meal (Cecelski 1987). However, in some places food is so scarce that fuel shortages play only a minor role in determining diets. A decreasing availability of wood does not necessarily lead to shortages of fuel. People respond spontaneously to decreases in fuelwood supplies through a number of adjustments. Those who have land may use more of the woody material grown on their own land, and change their cropping patterns to include species such as pigeon pea (Cajanus cajan), which pro- vides woody residues that can be used for fuel. Others may collect fuelwood from farther afield. Still others use available supplies more economically and shift to other readily available biomass fuels such as crop residues and dried dung (Dewees 1989; Leach and Mearns 1988). Agricultural Inputs Many systems rely on tree cover to restore nutrients to the upper layers of the soil, either by intercropping suitable tree species with the farm crops or by gathering green mulch from trees off-farm. Crop cultivation in the Himalayas, for example, is dependent on access to a substantial area of forest from which to cut and carry leaf mulch to maintain soil fertility. Eventually the forest can no longer sustain the repeated offtake. In dryland systems, where ploughing and sowing have to be compressed into a short rainy season, the number of animals needed is considerably higher than the number that can be sustained from feed produced within the farm system; hence farmers must have access to grazing or fodder off- PRODUCTION OF FOREST PRODUCTS 437 farm. Forests, woodland, and areas of scrub are usually the principal comple- mentary sources, and arboreal fodder may be the main source of livestock feed in the dry season and in periods of drought. Numerous pressures have combined to reduce the availability of live- stock feed. The irrigation of land previously under dryland crops or pas- ture, the shift to short-stemmed high-yield grain crops, and shifts away from cereal crops are but some of the changes occurring within agriculture. At the same time, privatization and overuse have widely reduced availabil- ity on public lands. The available responses-irrigated fodder crops, stall feeding, the substitution of tractors for animals-tend to require more inten- sive use of capital or labor and therefore are unavailable to the poor. It has been argued that fodder shortages can become more serious than fuelwood shortages because alternative sources of livestock feed are less available to the poor than alternative ways for maintaining supplies of fuel. Shortages of fodder are likely to mean that poor farmers are unable to upgrade to the higher-quality animals needed for dairying, and the poor are often forced to dispose of livestock. Sources of Income and Employment Rural people gather, produce, and trade a wide range of forest products to derive income. Gathered products include fuelwood, rattan, bamboo, fi- bers, medicines, gums, and wild foods (Falconer and Arnold 1989). The main groups of traded products, which first undergo simple processing at the household or small-enterprise level, are furniture and other products of wood; baskets, mats, and other products of canes, reeds, and grasses; and handicrafts. The first two product groups serve predominantly rural house- hold and agricultural markets, while much of the handicrafts output goes to urban markets (Fisseha 1987). The predominance of small-enterprise operations in the forest sector reflects the size of rural markets for forest products and the dispersion of these markets across large areas with a relatively poor transport infrastruc- ture. At a time when rural households are having to look to nonfarm em- ployment and income for a growing share of their total livelihood (Kilby and Liedholm 1986), small forest-based gathering and processing enterprises provide one of the largest sources of nonagricultural employment and in- come to rural people. Many people depend on year-round sale of products such as fuelwood and rattan to supplement their farm income. Others engage in such activi- ties seasonally, to exploit raw materials or markets available only at particu- lar periods, to exploit the labor available in slack agricultural months, or to meet seasonally induced cash needs such as agricultural loan payments or school fees. Others resort to such activities to tide them over emergencies; for example, more people become involved in gathering and sale of fuelwood in years when agricultural conditions are bad. 438 MANAGING THE WORLD'S FORESTS The seasonality of some activities is dictated by the availability of the product or raw material, and of some others by the demands of other activi- ties such as agriculture. For example, in northern Brazil, babassu palm ker- nels are gathered and processed during the agricultural slack period, when the income earned from these activities represents more than a third of the family's overall budget (May et al. 1985). Because the markets for many locally processed forest products depend on rural people's purchasing power, these markets, too, are tied to the cyclic nature of agricultural incomes. As with forest foods, the poor find forest-based income and employ- ment opportunities particularly important, because of the ease of access and low thresholds of capital and skill needed to enter and engage in most such activities. Poor women often dominate activities such as mat and basket making, which may be performed in or near the home and thus may be combined with other household tasks. But returns to labor from many forest-based activities are marginal, and markets for the products may be vulnerable to introduced substitutes. More- over, diminishing forest stocks are a particular threat to small enterprises, because small producers are seldom able to create or conserve their own tree resources. Their access to forest raw material is often impeded by unfa- vorable harvesting controls, exclusive allocation to large users, complicated licensing or auctioning procedures (plus demands for heavy deposits or other insurmountable preconditions), high prices due to state monopolies, and monopoly distribution systems. Thus, although forest-based activities provide a way for many rural poor people to earn income, many such activities may not be sustainable in the future (Falconer and Arnold 1989). COMMON PROPERTY RESOURCES Patterns of Common Property Use As already noted, rural people draw many of their forest products from areas of forest, woodland, and "waste" land to which they have access as common property resources. These outputs may constitute important parts of the overall agricultural system, filling gaps in the resource and income flows from other resources and providing complementary inputs often criti- cal to the continued functioning of agricultural and household systems. The nature and magnitude of the relationship vary with the character- istics of the surrounding ecological and agricultural systems. In India, for example, three broad categories have been identified (Arnold and Stewart 1991): 1. In the arid and semiarid regions, land allocation and encroach- ment have reduced common lands to a small area, which is typically heavily degraded and under open-access usage. A large percentage of the draft animals that are needed for dryland agri- culture are maintained on these common lands, and fodder and PRODUCTION OF FOREST PRODUCTS 439 fuel are the two most important outputs. The relative importance of the common lands varies significantly among households, with the poor being much more dependent on them and the rich more interested in privatization. Most of the common lands that were not privatized earlier .exhibit soil, erosion, or fertility problems and cannot sustain intensive agriculture. 2. In the hills, common property can comprise 60 to 80 percent of the area, predominantly in the form of forests under the control of forestry departments. In contrast to the dry-plain regions, com- mon land areas per household can be substantial, and all house- holds have similar patterns of common land use. Green mulch and fodder are the main outputs. Increased commercialization of common land products has led to heavy use in many areas. None- theless, the forest, though degraded in terms of timber content, may still be capable of producing sustained supplies of common land products. 3. In the forest belt across central India, common property re- source management traditionally covered most of the land. Indig- enous people mainly collect minor forest products as a source of income. Expropriation of local rights by the state-plus rapid and continued privatization, much of it by outsiders-has seriously weakened traditional institutions and practices. In Africa, with its traditional patterns of clan and lineage control of the use of land, management of tree-covered land, and of the tree stocks that the land supports, is intimately linked with fallow management, range management, and crop management (Shepherd 1990). Forest products are drawn principally from tree cover, which serves the primary function of fallow within rotational cropping systems, or from seasonal grazing in live- stock systems. Use of the tree resource is therefore managed as part of the control and management of the use of the land. Trends in the Availability and Use of Common Property Resources In modern times, common property resources have been massively re- duced nearly everywhere. Privatization, encroachment, and government appropriation have been the main processes taking resources out of com- mon use. Increasing pressures on what is left have frequently led to its progressive degradation. This process is now so heavily entrenched in policy and practice in many countries as to make further privatization or appro- priation seem inevitable or desirable, or both. The strong pressure to bring the use of common resources under pri- vate or government control has been widely influenced by the thesis of the "tragedy of the commons," which argues that the increasing pressures on individual users prevent effective cooperation and group control (Hardin 440 MANAGING THE WORLD'S FORESTS 1968). However, much usage of common land is open access-characterized by unrestricted entry and unregulated use rather than by collective manage- ment. The breakdown of common property resource management has all too often been erroneously declared in situations in which the unregulated use under an open-access regime actually caused the deterioration. This misunderstanding has been compounded by a tendency to over- look the factors that encourage collective action, the self-regulating capabili- ties of groups of users, and the reasons why the alternatives of private or state control may themselves not be sustainable or efficient. Because exclu- sion from common property is difficult, it may not be feasible to privatize it. Private use can also lead to overuse and degradation. Equally, the state may not be able to control, manage, or prevent degradation to a resource it has expropriated (Berkes et al. 1989). Common property may not have the same degree of support in law, or elicit the same response when threatened, as private property (Bromley and Cernea 1989). Privatization, by transferring control of the resource to a limited num- ber of individuals who thereby acquire the social and legal sanction to ex- clude others, is likely to exacerbate the problems of people who lack access to private property (Bromley and Cernea 1989). Lack of complementary resources to develop and cultivate the land, or immediate needs for cash for other purposes, can compel a large proportion of those of the poor who are allotted land under privatization programs to sell, mortgage, or lease it (Jodha 1990). Another reason to continue trying to manage the land remaining in common use is that it is too poor to support crop cultivation. The uses for which such land is best suited, such as grazing and fuel production, can be best served by keeping it in the common property domain. Land-use policies that have diminished the availability of, and access to, common property resources have frequently contributed to the weak- ening or collapse of the institutional arrangements within which common property was controlled and managed in the past. A study of common property resource management in the dry regions in India found that of the communities that had exercised controls such as rotational grazing, sea- sonal restrictions, and watchmen in 1950, only 10 percent had such controls in 1980, and use of fines, taxes, and fees had ceased altogether. Most com- mon property had become an open-access type of resource (Jodha 1990). Increasing population pressure, greater commercialization, and techno- logical change all contribute to this breakdown. In addition, increasing internal differentiation of rural communities as a result of privatization, which favors wealthier and more powerful people, has undermined interest in the main- tenance of common property. Possibly the most important factor in under- mining communal control has been the replacement of local leadership and authority with centralized political control-'' the ever increasing tendency of the state to expropriate the initiatives and activities which belong to people" (Jodha 1990). The consequence of the breakdown of local control PRODUCTION OF FOREST PRODUCTS 441 over the use of, common property resources in the dry land areas of India are summarized in exhibit 16-2. Comparable patterns of change have been occurring in other regions. In forest communities in Southeast· Asia, traditional methods of controlling access, usufruct allocation, and conflict resolution have become ineffective or disappeared, undermined by political, economic, and social changes within the village and nation. State assertion of control, first over the resource, then over the land, reduced access, and rights of usage. Differentiation within EXHIBIT 16-2. Measures Adopted by Different Groups in the Face of Decline in Area, Productivity, and Management Systems of Common Property Resources in Dry Regions of India Rural Rich 1. Withdrawal from common property resources as user of products (Opportunity cost of labor higher than the value of products from common property) 2. Increased reliance on alternative options: • Use of biomass supplies (stall feeding, etc.) • Use of nonrenewable/ external resources (e.g., substituting stone fencing for thorn fencing, rubber tires for wooden ones on carts, and iron tools for local, wooden ones) 3. Private squeeze on common property resources as assets: • Grabbing common lands • Preventing others from using seasonal common property resources (private croplands during off-season) 4. Approach to common property resources management: • Indifference to decline of common property resources • Participation in legal and administrative changes that undermine local control of community resources Rural Poor 1. Use of common property resources as an important source of sustenance (complementarity of common property resources and private property resource-based activities) 2. Acceptance of inferior options (opportunity cost of labor lower than value of products of degraded common property resources) 3. Measures reflecting desperation: • Premature harvesting of common land products • Removal of roots/base of product • Overcrowding and overexploitation of common property resources • Use of hitherto unusable inferior products Rural Community (general) 1. Acceptance of common property resources as open-access resources (overexploitation without users' obligations or regulations) 2. Selective approach to specific common property resources units (despite general neglect of common property resources, concern demonstrated for some units) 3. Focus on "other" uses of common property resources (in seeking government subsidy I relief, in running factional quarrels, in supporting populist programs, etc.) 4. Part of nonoperating legal and administrative measures 5. Structural changes and focus on alternative sources: • Changes in livestock composition (replacement of cattle by sheep or goats, etc.) • Agroforestry initiative (revival of indigenous agroforestry, etc.) SOURCE: Adapted from Jodha (1990). 442 MANAGING THE WORLD'S FORESTS the community and in-migration of outsiders asserting claims to use the resource have widely thwarted efforts to maintain or reestablish local con- trol systems (Poffenberger 1990). In Africa, large areas of land have been transferred from communal to state control. Management has changed from use-rights based on clan mem- bership to the exercise of state-granted privileges and management by re- striction and exclusion. The authority of the traditional kin-group has been undermined, allowing an increasingly unregulated exploitation of land. Now that legislation is necessary for any change to established practice, groups are discouraged from organizing to manage their local resources (Shepherd 1990). Despite these trends, many people, partkularly poor people, still de- pend on common property resources. Even in the heavily reduced and de- graded dryland communal areas of India, Jodha (1990) found that the poor obtained the bulk of their fodder and fuelwood and 14 percent to 23 percent of their income from common property resources. The poor were also found to benefit considerably from the employment created by common property resource management activities. Interventions in Management uf Forest-Related Common Property Resources Social Forestry Woodlots One of the largest interventions designed t:o increase the productivity of forest-product common property resources, and to strengthen local man- agement institutions, has been the communal woodlot projects established under the social forestry program in India. Most of the woodlots have been established in the dryland areas of the country, where common property resources have been shrinking and local control has been breaking down. Productivity was to be increased by raising woodlots on uncultivable public land, and control was to be exercised through the panchayat system (the lowest level of the state administrative structure). Trees have been planted on village lands or uncultivated public lands temporarily transferred to the forestry department for this purpose. The panchayat, or some other designated community-level body, was to take over responsibility for management from the forestry department after the woodlot was established. Initiated in most Indian states in the early 1980s, the programs have expanded very rapidly. For example, during the first four years of the project in the state of Orissa, woodlots were established in about 3,200 villages (SIDA 1988). Under forestry department management the projects have created pri- marily tree stocks and wood products, rather than the intermediate prod- ucts such as fuelwood and grass that were previously harvested from the areas. In this way, use of the common resource is inadvertently being shifted PRODUCTION OF FOREST PRODUCTS 443 from products for local use to higher-value wood products for sale outside the community. Consequently, benefits are, in practice, being transferred from those who earlier used the common land to those who will gain from the income accruing to and spent by the community as a whole. In addition, the transfer of responsibility for management of woodlots on common land to the community that was planned is rare. The rules established by the government for forest management are more closely tai- lored to foresters' than to villagers' skills and experience and are not always compatible with local needs and cannot be changed locally; in addition, because planning and control are centered in local government bodies rather than user groups, there is a widespread lack of local confidence in the out- come. As a result of these institutional weaknesses, the projects may be unwittingly converting common property resources into state-controlled resources (Blaikie, Harriss, and Pain 1986; Chambers, Saxena, and Shah 1989; Arnold and Stewart 1991). Although the projects have increased the produc- tivity of the sites used, the interventions have not been well adapted to the institutional framework of the region Uodha 1990). Joint Management on Forest Land In common with most other "woodlot" programs, the social forestry interventions in India are predominantly on land outside the forest. Success seems more likely where the interventions seek to bring about joint man- agement of forest land, building on the mutual benefits to be obtained from greater access to forest products by local people and reduced protection costs for the forestry department. In northern India and Nepal, substantial government land available for common property resource use and similar patterns of use of forest products throughout the community are two fea- tures that appear to explain their relative success. Other positive features include the following (Arnold and Stewart 1989): • Management by the user group, or groups, rather than by the village or panchayat as a whole; • Security of tenure to the user group, with the state playing an active role in defining and protecting boundaries against outside use and encroachment; • Use regulations that have evolved and are enforced locally, and are marked by rules that are understandable and easily adjusted to meet new challenges; • Community management of benefit allocation to reflect the interests of the elite and the powerful as well as the interests of people depen- dent on the common property resources; and • A management focus on low-value products of local importance. The greatest success seems to have occurred in situations where the technical knowledge already existed at the village level, and the missing 444 MANAGING THE WORLD'S FORESTS ingredient was an effective agreement between village-level institutions and local representatives of the government. Experience suggests that such insti- tutions and working arrangements can mature relatively quickly, at least in appropriate situations in South Asia. Elsewhere, success with joint management remains the exception rather than the rule. A recent survey of experience in Southeast Asia (Seymour and Rutherford 1990) reports two main impediments to progress. One is the reluctance, or inability, of forestry departments to proceed with or imple- ment devolution of responsibility to local level, particularly where the de- partments perceive that such action will threaten their control over a timber resource. Improved access to use of forest products therefore tends to be concentrated on degraded forest. The other constraint arises from pressures from within the community, such as from in-migration by outsiders, which undermine or overwhelm agreed systems of local control. Problems tend to be more pronounced where access to forest products is to be combined with rights to cultivate land--as on tree pattas (leases) in India, stewardship contracts in the Philippines . land entitlement certificates in Thailand, and the forest-management agreements between the State For- est Corporation and groups of farmers practicing taungya (tumpang sari) on forest land in Java, Indonesia. These arrangements can stimulate concern on the part of governments or government departments that a concession al- lowing temporary use of forest land for cultivation will lead to permanent alienation of the land from forest to agriculture. Such schemes may also founder because participants lack the resources to bring the degraded land assigned to them into productive use. Allocation to individuals of land that previously was available for common use also tends to raise problems of choice and exclusion among the previous users. Weaknesses in the legal framework have proved common to nearly all experience, both promising and unpromising. For local or joint control to be effective, the government must be willing and able to legitimize and em- power the local controlling institutions and to help them enforce their rights. Government failure in this respect tends to undermine even the most prom- ising approaches. Most governments are slow to amend or to implement laws. Interventions may be at variance with existing legislation, or imple- mentation of enabling legislation, where it exists, may be neglected. Often, forestry departments do not honor their own obligations or enforce those of right holders. Because the poor are usually at a disadvantage in legal cases, recourse to the law can further undermine their situation, putting even their existing de facto rights at risk (Seymour and Rutherford 1990). It is therefore necessary to recognize the factors that may prevent col- lective management from succeeding. When local institutions have broken down under the pressures of change, it is unrealistic to expect that new village institutions capable of controlling resource allocation and use can be easily created. Interventions that increase the productivity and value of a common property resource may attract interest in its privatization, which PRODUCTION OF FOREST PRODUCTS 445 could undermine the current level of control. The low returns and high social cost associated with trying to control common property resources may prove so unacceptable to users that they will prefer to leave those resources to the state to manage. TREES IN FARMING SYSTEMS Patterns of Farmers' Tree Management As common property resources disappear or are degraded, farmers everywhere have sought to shift the production of outputs of value onto their own land by protecting, planting, and managing trees of selected spe- cies. Many farmers now depend on their own tree stocks for some products and on common property resource sources for others. Recently, the process of adding trees to farming systems has been accelerated or transformed by the growing commoditization of fuelwood and other tree products, and the consequent emergence of the growing of trees as a cash crop. Some of the changes in agricultural land use may result in the elimina- tion of trees from farming systems rather than their retention or establish- ment. Prominent among such pressures are competition with crops for light, water, and nutrients on intensively used cropland; new agricultural tech- niques (e.g., use of tractors); broader land-use practices (burning, free graz- ing); changes in control of the land (privatization, nationalization); and re- duction in the rotational cycle to the point at which desirable trees are no longer able to reestablish. Other changes tend to reduce or eliminate the need for trees. Irrigation of dry land, for example, is likely to reduce the need for draft animals, and hence for fodder; irrigation is also likely to create new and more productive sources of fodder than fodder trees could provide. Moreover, some alterna- tives may present a lower opportunity cost to the farmer than creating supplies of tree products-hence the widespread use of dung and crop residues in place of fuelwood. Other economic options available to the farm household, off the farm as well as on it, may offer a better use of its re- sources than adding or intensifying tree management. The balance between supplies from tree stocks off-farm and from man- aged and planted on-farm stocks varies widely with the agroecosystem and with the patterns of land use, population pressure, and level of activity associated with different agroecological situations. Exhibit 16-3 outlines contexts in which woodland management and tree planting occur in the semiarid regions of Africa. Within a continuum from low rainfall, low popu- lation, and considerable common resources to higher rainfall, high popula- tion, and very little remaining common land, supply of forest products shifts steadily toward dependence on farmer-managed tree resources (Shepherd 1990). 446 MANAGING THE WORLD'S FORESTS EXHIBIT 16-3. Contexts in Which Woodland Management and Tree Planting Occur in the Semiarid Regions of Africa General Situation Type 1 Situation Type 2 Situation Type 3 Situation Type 4 Low rainfall ............................................................. High rainfall Far from town ........................................................... Near to town Low population density ........................................· High population density Type of Land Use Extensive Intensive • Pastoralism or • Settled agriculture, • More intensive agri- • Highly intensive agri- • Settled home base some on registered culture. Most land is culture with all farms + migrant animals or land. Some open demarcated as per- contiguous. • Shifting cultiva- land between farms. manent registered • Increasing land prices tion, long fallows. Fallows shortening. plots. Dung or other + plot fragmentation. • Labor is the key • Animals important fertilizer is bought. • Landlessness. constraint, so poly- but grazing pressure • Animals fewer, • Off-farm employ- gyny is often found increasing; kept on kept on farms. ment increases. in this type of nearby common. • Animals stall-fed situation. or sold off. Extent of Common Property Resources Lots of common Common land All common land Scraps of waste land land; traditional getting scarcer. gone except for hill- may still exist. Man- management rules Management rules tops, etc. Common ageinent forgotten. still extant. causing conflict. property resource Open access only. rules no longer thought workable. Likely Villager Interests: Promising Tree Project Interventions • Only homestead • Mostly homestead • Interest in field- • All tree needs are planting; shade, planting for shade, boundary planting farm grown except fruit, hedges. fruit, hedges, and of poles, timber, and high-quality timber. • Only small num- perhaps poles. maybe fuel. •Good markets exist bers of trees wanted. • Animal damage to • Interest in all the for high-value farm • Tree-related cash planted trees a homestead options. tree products. from bush products common problem. • Cash sales of fruit, • Trees provide fodder such as browse, • Cash from farm- poles, and fann- for stall-fed animals. honey, charcoal. grown fruit, grown fuelwood if • Whole farm can • Only here is fuelwood gath- no competition from be put under trees. woodland manage- ered from remoter common ment with villagers common property. property resources. worth investigation. • Alley cropping and mulching are practiced. NOTE: The chart represents four points along a continuum, from low population density and low interest in tree planting (on the left) to a high population density and high interest in tree planting (on the right). SOURCE: Adapted from Shepherd (1990). PRODUCTION OF FOREST PRODUCTS 447 Similar patterns are found in other regions. In parts of the middle hills of Nepal where population growth has put existing resources under in- creasing pressure, there has been a major increase in tree cover on private land over a 24-year period. Trees were added first to stream beds and banks, then to uncultivated land and the walls of rain-fed terraces (Carter and Gilmour 1989). In very arid areas of Rajastan, India, where fodder and fuel resources on common land have been severely depleted over 30 years, there has been an increase in the density of some of the woody shrubs inter- cropped on farmland (Jodha 1988). In much of the humid tropics, there has been an increase in the proportion of farmland under home gardens, which include trees. Factors Influencing Change in Farmers' Tree Management Within a particular agroecosystem, farmers' involvement in tree grow- ing appears to be largely related to changes in the availability and employ- ment of land, labor, and capital, and to the progressive commoditization of tree products such as fuelwood and poles. Variations in tree-growing patterns within this framework seem to reflect inter alia variations in the efficiency of operation of factor markets, different stages in the process of agrarian transition, and different patterns of tenure. Factor Availability and Use It is generally argued that, as land holding declines, the more intensive use of land for the cultivation of food crops will preclude the growing of trees. However, where agroecological conditions favor vertically structured joint tree, crop, and livestock systems as potentially the most productive use of the site, farmers may respond to declining land availability through more intensive intercropping of trees and other perennial and annual crops. Such an evolution has been quite widely observed in the humid tropical belt in Asia and Africa within which home gardens feature as an important part of farming systems (Arnold 1990). Where labor resources are limited, as farm households are forced to turn increasingly to off-farm employment, low-input tree crops may be employed as a way of keeping land in productive use. Tree crops are likely to be grown where poorly functioning labor markets prevent farmers from adopting more productive labor-intensive uses for their land and where farmers do not want to lease out the land they have to withdraw from crop production. Farmers may also grow trees where lack of access to capital prevents farmers from adopting more capital-intensive crops. In highland areas in Kenya, for example, lack of capital appears to be one of the principal deter- minants of farmers' choice between tea and woodlots (Dewees 1990). In areas where livestock are not used as a way of maintaining a reserve of capital, farmers may grow trees for this purpose. 448 MANAGING THE WORLD'S FORESTS Farmers also use trees to help manage risk where repeated drought threatens other crops. Trees are also grown to help diversify farm produc- tion, to provide products and income in the period between the main har- vests, and to help bridge the peaks and troughs in seasonal demands for labor (Chambers and Leach 1987). Commoditization Growth in the markets for short-rotation wood products has often en- couraged many farmers to grow trees. This is particularly true in wood- short areas, where site and tenure conditions are favorable for tree growing, where land uses that require little labor are favored, and where the transi- tion from predominantly subsistence-oriented agriculture toward involve- ment in commodity markets has occurred. The expansion of the cultivation of trees as field cash crops has caused concern, notably in India, that land is being diverted from production of essential foods and is reducing rural employment (e.g., CSE 1985). This concern tends to overlook factors that are causing farmers to withdraw land from coarse-grain production and to find less labor-intensive forms of land use, and the features of tree growing that make it a logical response to these pressures (Saxena 1990). Crops of trees such as eucalyptus may not be appropriate where they could put household food security at risk. Producing only a single product, they are potentially vulnerable to market fluctuations, and thus to income fluctuations, and they provide income in "lumps" followed by periods with little or no income. Multipurpose trees and multispecies systems such as home gardens are more likely to contribute to a sound mixed-subsistence/ cash-crop household economy. Tree monocropping is likely to be appropri- ate only if the household has access to other sources of income or food, and if there are reasonably stable markets for the tree products (Falconer and Arnold 1989). Tenure Security of tenure has obvious implications for decisions about invest- ments in a long-term crop such as trees. It has been widely assumed that leasing, sharecropping and other forms of tenancy, systems of customary tenure under which land is a common pool resource, and customary and legal rights associated with the presence of trees all tend to inhibit farmers from engaging in tree growing. In some of these situations, however, tenure may be less important than other factors in influencing tree growing. In customary land-use systems in Africa, existing tenure often appears to pro- vide decisions that are determined more by considerations of profitability (Cook and Grut 1989; Shepherd 1990). Although individualization of hold- ings tends to result in an increase in tree planting, this could be as much a PRODUCTION OF FOREST PRODUCTS 449 reflection of the loss of access to common pool resources, and to other changes in agricultural use of the land, as of a perception of increased security. A recent review of evidence from upland Java, Indonesia, where tree cover plays an important conservation role, concluded that the principal factors influencing farmer decisions were the productive potential of the land, the size of total land-holding, and the presence of good local markets for perennial crops and of off-farm employment opportunities. While share- cropping and leasing arrangements may slow the rate of adoption, "tenure status per se is probably less important than related factors such as access to credit and the fragmentation, isolation and minute size of landholdings of many rural households" (Mackie 1989). The need to increase security of tenure in order to encourage tree grow- ing may therefore occur less often than tends to be assumed in project design. Indeed, attempts to change tenure can prove to be counterproduc- tive. Past changes have often engendered strong local distrust of govern- ment in this matter, so that the prospect of change to the tenurial situation can itself inhibit investment in long-term activities such as tree growing. Where tenure is a constraint, change is often difficult to accomplish, and interventions that require such alterations in order to succeed may prove less realistic than solutions that can be effected within the existing legal and tenurial framework. Interventions in Private Tree Management In the first generation of projects designed to encourage private tree growing by farmers, there was a widespread tendency to develop projects as though they were effectively isolated from many of the key influences on them, particularly economic forces. The assumption that farmers plant trees to meet subsistence or environmental needs, which are not bought or sold in the marketplace, was reflected in projects designed as though they ·were divorced from and immune to market forces. Some projects even attempted to prevent participants from selling their produce, on the ground that this was contrary to the service function assumed to be the goal of community or social forestry. This approach reflected the priority that projects in the late 1970s and early 1980s gave to increasing supplies of fuelwood, in response to what were perceived to be widespread shortages in the subsistence sector. More exhaustive studies of a range of situations in which the fuelwood supply situation had been identified as worsening, have usually disclosed that domestic fuel shortages were much less serious than had been initially un- derstood. As was noted earlier, users have adjusted to declining fuelwood supplies primarily by using more of woody stocks other than forest trees (bushes, shrubs, prunings, etc.) and other biomass stocks (crop residues and dung), which were often renewing at rates able to sustain current levels of use. 450 MANAGING THE WORLD'S FORESTS Where rural fuel shortages do exist, the issue may be less one of physi- cal scarcity than of labor shortages, constraints on access, or culturally de- termined patterns of behavior (Dewees 1989). Thus the task of obtaining fuel may be becoming a greater burden because women have more to do, as is widely reflected in fluctuations in gathering seasonally which coincide with seasonal cycles in agricultural or other pressures on their time (Cecelski 1987). This fact makes the task of obtaining fuel no less a burden, but alters the likelihood that tree planting will be a sufficient, or even appropriate, solution. The emphasis on meeting subsistence needs tended to be accompanied by underestimating of the influence of market demands for wood products, including urban demands for wood fuels. As forest products such as fuelwood, fodder, and fruits become progressively commoditized, and farm households increasingly depend on income to meet at least part of their needs, the distinction between production for subsistence or sale loses mean- ing. Not only will producers sell what is surplus to their subsistence needs, but they will sell a commodity needed in the household if the opportunity cost of doing so is advantageous-hence the widespread phenomenon of households that are short of fuelwood selling wood. One result of promoting tree growing as though it were outside the forces of the market system has been a failure to match project production to market possibilities, or to link producers to markets. The collapse in building-pole prices in northwestern India as large quantities of farmer- grown material entered the market in the late 1980s (as a consequence of social forestry support programs) reflected lack of market information and lack of attention to the functioning of this emerging market. Most Indian states still have in place restrictions on the harvesting and sale of wood products by private producers, which severely hinder the efficient function- ing of these markets (Chambers et al. 1989; Saxena 1990). In addition, farm- ers had to compete with fuelwood supplied to urban markets from state forests at subsidized prices. Many farmers are now withdrawing from tree growing in the areas affected (Saxena 1990). Similarly, projects have generally neglected to put producers in touch with sources of higher-level inputs, such as credit, available to those seek- ing to produce for the market. Indeed, provision of credit, which has fea- tured prominently in government programs to encourage tree-crop cultiva- tion, has been notable by its absence in farm forestry projects (an exception being the PICOP (Paper Industries Corporation of the Philippines smallholder tree-growing project in the Philippines [Hyman 1983]). Project interventions have centered on provision of subsidized planting stock and cash payments to offset establishment and maintenance costs. Originally intended to encourage pursuit of 'essentially social and environ- mental goals, in practice these interventions generally support production for the market. Recent evaluations of projects in India suggest that there is a danger that this type of intervention is encouraging tree cash crops in PRODUCTION OF FOREST PRODUCTS 451 situations where they are unlikely to be profitable. In Bihar, farmers ap- peared to be planting in response to the short-term returns from the cash payments provided rather than the longer-term returns from investment in trees, leading to distortions in land use, such as displacement of sharecrop- pers and reduction in small-farmer subsistence production to the point at which household food security is jeopardized (SIDA 1990). Technical prescriptions in early farm forestry projects also have been weak. Raintree and Haskins (1988) reported that "the project record abounds with examples of projects that have foundered because of inappropriate species choice" and that "few social forestry project documents ever pro- vide any systematic rationale whatsoever for the matching of tree species to the needs of the target community!" This situation is partly due to the lag in applied research, and the relative neglect within the research of on-farm work. Thus, even with alley cropping, which has benefited from one of the most intensive and thorough research efforts of any innovation in the field of agroforestry, it is still unclear to what extent farmers will find this prac- tice appropriate. In addition, the pressures often placed on forestry services to achieve planting or seedling distribution targets have too often resulted in priority being accorded to quantity rather than to quality (or appropriateness). And poor communication with farmers and their families-due to a shortage of people trained in communication and extension skills-has meant that even projects that are now vigorously trying to remedy this weakness are bur- dened with project objectives and designs that were developed without the benefit of involvement of the target population. SUMMARY Production managed as part of agricultural systems is meeting an in- creasing proportion of the demand for forest products at the rural house- hold level in developing countries. In many of these agricultural systems, common property resources, managed to complement farm resources, play a critical role. Government policies that progressively replace local with state control over natural resources are widely undermining collective management of common lands. Privatization, encroachment, and government expropriation- and legislation that gives greater support to private property-reinforce this process. Nevertheless, there are many situations in which common land management would be a preferred solution, given an appropriate policy and implementation environment and more accurate targeting and design of interventions. With the deterioration in forest and common property resources, on- farm tree stocks will become progressively more important. Interventions are likely to be most needed to remove impediments stemming from gov- ernment policies and actions, to strengthen farmers' access to markets, and 452 MANAGING THE WORLD'S FORESTS to provide improved technology and inputs. Lack of information about the role of trees in particular farming systems, and about current constraints to tree growing, can still hamper the definition and formulation of appropriate interventions. REFERENCES Arnold, J.E.M. 1990. "Tree Components in Farming Systems." Unasylva 41:160. Arnold, J.E.M, and W.C. Stewart. 1991. Common Property Resource Management in India. Tropical Forestry Paper No. 24, Oxford Forestry Institute, University of Oxford. Berkes, F., D. Feeny, B.J. McCay, and J.M. Acheson. 1989. "The Benefits of the Commons." Nature 340 (July 13). Blai~e, P., J. Harriss, and A. Pain. 1986. "The Management and Use of Common Property Resources in Tamil Nadu." In Proceedings of the Conference on Common Property Resource Management. Wash- ington, D.C.: National Academy of Sciences. Bromley, D.W., and M.M. Cernea. 1989. The Management of Common Property Natural Resources: Some Common Conceptual and Operational Fallacies. World Bank Discussion Paper No. 57, World Bank. Carter, A.S., and D.A. Gilmour. 1989. "Increase in Tree Cover on Private Farm Land in Central Nepal." Mountain Research and Development 9(4). Cecelski, E. 1987. "Energy and Rural Women's Work: Crisis Response and Policy Alternatives." International Labour Review 1216(1):41--64. Chambers, R., and M. Leach. 1987. "Trees to Meet Contingencies: Savings and Security for the Rural Poor. IDS Discussion Paper No. 228. Institute of Development Studies at the University of Sussex, Brighton. Chambers, Robert, N.C. Saxena, and Tushaar Shah. 1989. To the Hands of the Poor: Water and Trees. New Delhi: Oxford and IBH Publishing Co. Cook, C.C., and M. Grut. 1989. Agroforestry in Sub-Saharan Africa: A Farmer's Perspective. World Bank Technical Paper No. 42. Center for Science and Environment (CSE). 1985. "Social Forestry." In The State of India's Environ- ment, 1984-85: The Second Citizens' Report. New Delhi: CSE. Dewees, P. 1989. "The Woodfuel Crisis Reconsidered: Observations on the Dynamics of Abundance and Scarcity." World Development 17(8). _ _ _. 1990. "Tree Growing, Labour Availability and Land-Use in Smallholder Agriculture in Kenya: Preliminary Results from a Field Survey in Murang's District." Oxford Forestry Institute (processed). Falconer, J. 1989. Forestry and Nutrition: A Reference Manual. Rome: FAO. Falconer, J., and J.E.M. Arnold. 1989. Household Food Security and Forestry: An Analysis of Socioeco- nomic Issues. Rome: FAO. Fisseha, Y. 1987. ''Basic Features of Rural Small-scale Forest-based Processing Enterprises in Devel- oping Countries." In Small-scale Forest-based Processing Enterprises. Forestry Paper No. 79. Rome: FAO. Hardin, G. 1968. "The Tragedy of the Commons." Science 162. Hyman, E.L. 1983. "Pulpwood Tree Fanning in the Philippines from the Viewpoint of the Smallholder: An ex post Evaluation of the PICOP Project." Agricultural Administration 14:23-49. Jodha, N.S. 1988. "Fuel and Fodder Management Systems in the Arid Region of Western Rajasthan." Report for the Study Group on Fuel and Fodder, Planning Commission, Government of India. _ _ _. 1990. "Rural Common Property Resources: Contributions and Crisis." Economic and Politi- cal Weekly Quarterly Review of Agriculture 25(26). Kilby, P., and C. Liedholm. 1986. "The Role of Nonfann Activities in the Rural Economy." Employ- ment and Enterprise Policy Analysis Discussion Papers, Cambridge, Mass.: Harvard Institute for International Development. PRODUCTION OF FOREST PRODUCTS 453 Leach, G., and R. Mearns. 1988. Beyond the Woodfuel Crisis: People, Land and Trees in Africa. London, Eng.: Earthscan Publications. Mackie, C. 1989. "Land Tenure and Conservation Practices in the Upper Watersheds of Java." A report submitted to the Asia Technical Office, Environment Division, World Bank (processed draft). May, P.H. 1985. "Babassu Palm in the Agroforestry Systems in Brazil's Mid-North Region." Agroforestry Systems 3(39). Poffenberger, M. 1990. Keepers of the Forest: Land Management Alternatives in Southeast Asia. Washing- ton, D.C.: Kumarian Press. Raintree, J.B., and M.W. Hoskins. 1988. "Appropriate R&D Support for Forestry Extension." In Planning Forestry Extension Programmes. Bangkok: PAO. Saxena, N.C. 1990. Economic Aspects of Fann Forestry in North-West India. New Delhi: Ford Founda- tion. Seymour, F.J., and D. Rutherford. 1990. "Contractual Agreements in Asian Social Forestry Pro- grams." Paper presented at the First Annual Meeting of the International Association for the Study of Common Property, September 27-30, Durham, North Carolina. Shepherd, G. 1990. "Communal Management of Forests in the Semi-Arid Regions of Africa." Report prepared for FAQ Forestry Department. Social Forestry Network, Overseas Development Insti- tute, London (processed draft). Swedish International Development Authority. 1988. "Evaluation of the SIDA-Supported Social Forestry Project in Orissa, India." Stockholm: SIDA. _ _ _. 1990. "Evaluation of the SIDA-Supported Bihar Social Forestry Project for Chotanagpur and Santhal Parganas, India." Stockholm: SIDA. 17 Watershed Management: A Key to Sustainability Kenneth N. Brooks, Hans M. Gregersen, Peter F. Ffolliott, and K.G. Tejwani F orest lands serve as an economic and environmental resource, with both perspectives needed to achieve sustainable development (World Bank 1978). The economic resource includes wood products (fuelwood, lumber, building poles, etc.), plus food, animal, and other products from forest lands that are consumed by humans. The environmental resource includes biodiversity, wildlife habitat, climatic benefits, water quantity and quality, and aesthetics. Watershed management provides a framework for manag- ing such multiple resources in an integrated manner. The World Bank's 1978 Forestry Sector Policy Paper (World Bank 1978) stated, "The ecological usefulness of forests is most readily observed in their beneficial effect on water catchment areas, where they have a regulatory effect on streamflows and where they protect soils from erosion and pre- vent silting of dams and canals." Although it may be argued that biodiversity and other ecological features are equally important, natural forests are hy- drologically the most stable systems on earth. Forestry and related develop- ment projects should attempt to capture the hydrologic and erosion control benefits of natural forest systems. Environmental components in projects, including watershed manage- ment, often have a low priority in developing countries because many envi- ronmental benefits are hard to measure, equity problems are associated with some of the environmental benefits, and the benefits take a long time 455 456 MANAGING THE WORLD'S FORESTS to accrue. However, it is in developing countries where environmental is- sues are foremost and need to be addressed in development projects. The challenge is to protect upland watersheds, which provide benefits to local, regional, and countrywide economies. · With appropriate planning, design, and management, forestry and re- lated projects can have tangible watershed benefits. But developing coun- tries need operational guidelines and developmental strategies that help rural low-income groups without leading to the environmentally and eco- logically destructive patterns of forest removal. These countries also need training, education, and forestry research that emphasize the environmental linkages and benefits associated with development projects. In high-population areas, land degradation is directly linked with food security, both in terms of upland watersheds and downstream effects (FAO 1985). The loss of soil and productive capacity in uplands is directly evi- dent, but the downstream effects may not be immediately apparent. How- ever, sedimentation and excessive streamflow from uplands can disrupt downstream transportation systems and water resource facilities that are needed for irrigation, hydropower generation, and the like; these disrup- tions have social and economic implications. Developing countries need to consider the multiple use of natural re- sources in uplands and downstream areas. The sustainable production of food, fuelwood, fiber, forage, water, and other products requires the recog- nition of watershed boundaries and linkages. Because the multiple resources of land cannot be managed by individuals and agencies at cross-purposes, institutional mechanisms are needed to achieve successful projects. This chapter examines the role of watershed management as a compo- nent of forestry and related development projects, and provides a practical framework that can be used to identify and assess priorities for watershed components in forestry projects. The chapter also examines the role of trees and forests in meeting watershed management objectives. Selected examples are presented to illustrate problems and opportuni- ties associated with integrating watershed management into the fabric of development projects. CONCEPTS AND DEFll\llTIONS Development agencies have had difficulty in transforming their recog- nition of the importance of environmental effects into meaningful actions that (1) lead to tangible environmental benefits and (2) improve productiv- ity for rural inhabitants-both of which are needed for sustainable develop- ment. Taking a watershed management approach allows for the explicit accounting of certain environmental benefits associated with forestry devel- opment projects and helps identify the linkages between environmental improvement and productivity increases over the long term. The following definitions, taken from Working Group on Watershed Management and PRODUCTION OF FOREST PRODUCTS 457 Development (1988) and Brooks et al. (1991), point to the usefulness of such an approach: • A watershed is a topographically delineated area that is drained by a' stream system (i.e., the total land area above some point on a stream or river that drains past that point). The watershed is a hydrologic unit that often is used as a physical-biological unit and a socio- economic-political unit for the planning and management of natural resources. A river basin is similarly defined but is larger. For example, the Amazon Basin, the Mississippi Basin, and the Mekong Basin include all lands that drain through these rivers and their tributaries into the ocean. • Watershed management is the process of guiding and organizing the use of the land and other resources on a watershed to provide de- sired goods and services without harming soil and water resources. The interrelationships among land use, soil, and water and the link- ages between uplands and downstream areas are recognized in this concept. • Watershed management practices are. changes in land use, vegetative cover, and other nonstructural and structural actions that are taken to achieve specific watershed management objectives. These objectives can be rehabilitation of degraded lands; protection of soil and water systems for land being managed to produce food, fiber, forage, and other products of the land; and enhancement of water quantity or quality (Black 1990; Brooks et al. 1991). • A watershed management approach incorporates soil and water conser- vation and land-use planning into a broader, logical framework by focusing on the following concepts: -People are positively and negatively affected by the interaction of water with other resources, and, in turn, people influence the na- ture and severity of such interactions by the ways in which they use resources and the quantities they use. -The effects of these interactions follow watershed boundaries, not political boundaries; water flows downhill regardless of how people define their political boundaries. Thus what is done in the high- lands of one political unit (country, community, or landowner) can significantly affect another political unit occupying a downslope or downstream position in the watershed. -Because such interactions cut across political boundaries, what may be sound use of resources from the point of view of one political unit may not be sound use of resources from a broader, societal point of view, because of undesirable downstream effects (i.e., what economists call externalities). This approach brings such externali- ties into the analysis by considering watershed boundaries. 458 MANAGING THE WORLD'S FORESTS -Given the existence of externalities, ecologically sound manage- ment becomes good economics for all concerned only if the costs and benefits are appropriately distributed among the political units, communities, and individuals that carry out watershed manage- ment practices and those that benefit from them. The integrated concepts of watershed management provide a frame- work for sustainable development, whereas watershed management prac- tices provide the tools for making the framework operational. Various insti- tutional mechanisms-regulations, market and nonmarket incentives, public investment-provide the means for implementing the practices. A common misconception is that watershed management is based only on physical interrelationships. However, the concepts just described indi- cate that sound watershed management involves economic and institutional interrelationships as well. These concepts also illustrate the focus of a wa- tershed management approach, and they guide the design of the practices and institutional mechanisms needed to implement the approach on the ground. BACKGROUND FOR POLICY FORMULATION Academics and development agency personnel have often discussed the concept of watershed management as an integral part of forest manage- ment. When development projects are implemented, development agencies rarely include watershed management practices, and when they do, the benefits and other effects are rarely quantified. So what difference does it make if one does or does not consider watershed management in forestry and agricultural development projects? It may be argued that watershed management is an issue that has little relevance in actual projects, because, in practice, it cannot be implemented for several reasons: the urgent needs of the rural poor, the lack of appropriate institutions, and the fact that watershed boundaries do not coincide with political boundaries. Such argu- ments can lead to short-term solutions and a continuation of nonsustainable land-use practices; opportunities to enhance the welfare of the rural poor over the long term can then be lost. Policies are needed that provide conser- vation of soil and water resources in a manner that fits within the water- shed management approach. Watershed management problems and some opportunities for solutions are discussed in the sections that follow. CAUSES OF WATERSHED PROBLEMS What are normally observed as evidences of watershed degradation are the physical manifestations of problems such as soil erosion, polluted water, evidence of frequent flooding, sediment-filled channels and reservoirs, and shortages of potable water. Natural phenomena, human activities, or some PRODUCTION OF FOREST PRODUCTS 459 combination of the two often cause watershed degradation and resulting upland and downstream impacts. The causes must be understood so that appropriate measures are taken, and, conversely, so that inappropriate or ineffective measures are avoided. The main causes of watershed degrada- tion are as follows (adapted from FAO 1986; Brooks et al. 1991): 1. Natural Causes • Geologic instability • High-intensity, long-duration rainfall • Steep river gradients • Shallow soils on steep slopes • Fire 2. Human Causes • Deforestation -Unwise and poorly designed logging, usually carried out with no guidelines or requirements to protect the soil and riparian (streambank) systems. The cause of such practices sometimes can be traced to the need for log exports to deal with foreign debt. Govern- ments. are unwilling or unable to impose controls or to enforce guide- lines on loggers. -Repeated shifting cultivation without adequate fallow periods, caused by high population densities in uplands. Alternative means of pro- ducing food and fuelwood are not available. -Fuelwood cutting due to fuel shortages and under conditions of large rural populations: -Conversion of forests to grazing lands or cultivated croplands to meet the demand for food by an expanding population. Such changes can result from illegal encroachment or legal land settlement. -Forest fires set by local inhabitants. • Inappropriate farming practices -Uncontrolled land-use changes -Inappropriate cultivation practices • Road construction on fragile lands • Overgrazing by livestock • Improper collection, transportation, treatment, and utilization of water • Socioeconomic and institutional problems -Land-tenure problems -Inadequate policy and legislative support -Scarcity of skilled workers -Lack of unified planning and extension for integrated watershed management -Inadequate community participation 460 MANAGING THE WORLD'S FORESTS Meeting the resource needs of the rural poor is also critical to the devel- opment of viable and sustainable solutions. After all, these people are the land users who can have the most widespread influence as managers of watersheds. Even projects that focus on downstream benefits-such as re- ducing sediment delivery to a reservoir or improving water quality-must consider the means of meeting the resource needs of upland inhabitants, Otherwise there will be little incentive to carry out practices aimed at down- stream protection. Thus socioeconomic and institutional issues as well as the restoration of the physical and biological system must be part of the solution. A WATERSHED MANAGEMENT APPROACH It is important to adapt a watershed management framework to the political and economic realities of the world. Watershed boundaries are useful in planning and management ·because soil and water are basic ingre- dients in sustainable development, and because the watershed is a logical unit for considering the effects of land use on soil and water. Doolette and Magrath (1990) emphasized that "watersheds as hydro- logic units provide appropriate units for conceptualizing and implementing development investments." This is a key point, because forestry, water re- source, and agricultural development projects can affect one another and therefore should not be developed in isolation. Development projects often are designed to solve problems associated with water, food, energy, and other natural resource problems, As exhibit 17-1 illustrates, watershed management practices are an integral part of the solution to many of these problems. Exhibit 17-2 is a guide to the specific effects of watershed management practices on the production, protection, rehabilitation, and enhancement of natural resources for a given physical-biological and socioeconomic setting. Watershed management practices involve the application of vegetative mea- sures, land-use guidelines (e.g., guidelines for logging, maintenance of buffer strips of vegetation along stream channels), and in some cases structural (engineering) measures. Watershed management practices and their pur- poses are summarized in exhibits 17-3 and 17-4. Later in this chapter we consider where such practices apply and what conditions are necessary to achieve overall production or protection goals. We suggest that planners consider the strategies and types of forestry and watershed management interventions for countries or regions from the standpoint of climate, topography, and population pressures. Forestry and watershed management interventions can differ in the humid tropics, dry- land forests (tropical and subtropical), and temperate forests, recognizing PRODUCTION OF FOREST PRODUCTS 461 that many watershed management practices apply in more than one of these climate-topography-population situations. Specifically: • Each climatic and topographic unit would have its own set of bio- physical conditions and, therefore, capability. • Opportunities and options will be influenced by both topography and population density. Role of Forests and Trees in Achieving Watershed Management Objectives Forests and trees affect the hydrologic behavior of a watershed, includ- ing the quantity and quality of streamflow, erosion, and sedimentation. In general, natural forests yield the highest quality of water of any ecosystem. The lowest erosion and sedimentation rates are usually associated with for- ested watersheds in natural conditions (Brooks et al. 1991).'Strearnflow from forested watersheds tends to be more uniform, with peak flows lower than those from watersheds with other vegetative cover. Given this background, the role of trees and forests can be viewed in terms of watershed protection, enhancement of water resources, and watershed rehabilitation. Watershed protection is an objective under special conditions, such as fragile, steep slopes, and for watersheds above municipal reservoirs. In some cases, may be prohibited on these watersheds, but they may provide mul- tiple benefits in terms of wildlife habitat, aesthetics, recreation, production of high-quality water for municipal use, and protection of biodiversity. Enhancement opportunities include manipulating vegetative cover to increase or decrease water yield. For example, replacing species of high consumptive use with species of low consumptive use may be desirable in drylands, other factors being considered. In contrast, to lower water tables in lowlands that have excessive water, planners may consider drainage and forestation with trees that have high annual consumptive use. In drylands, or even in the humid tropics where there are extended dry periods, planners should consider the effects of vegetation changes on wa- ter yield. For example, planners of watershed rehabilitation efforts using reforestation to reduce erosion should recognize effects on water yield. In contrast with other types of vegetative cover, forests tend to yield less water because forests evapotranspire more water than other types of vegetation, the exception being cloud forests (Brooks et al. 1991). Along coastal areas or in high-elevation zones in the tropics where dense fog or low clouds are prevalent, the interception of fog by forest canopies results in a net increase in moisture to the watershed, in contrast to lower-growing vegetation with less surface area. Thus reforestation can mean a reduction in water produc- tion from watersheds. Where water is in short supply, species that use water conservatively would be preferred. 462 MANAGING THE WORLD'S FORESTS EXHIBIT 17-1. Water-Related Problems, Their Causes, and Some Solutions Using a Watershed Approach Problems and Causes Structural and Associated Watershed Nonstructural Solutions Management Objectives 1. Inadequate water a. Reservoir storage and a. Minimize sediment supplies transport delivery to reservoir site (due to population b. Water harvesting maintain watershed density, competing uses, c. Reduce vegetation vegetative cover pollution of existing manipulation- b. Develop localized collection water supplies, lack of evapotranspiration and storage facilities access to supplies of d. Development of c. Convert from deep-rooted water, inability to store groundwater to shallow~rooted species or water during periods of e. Cloud seeding from conifers to deciduous excess, and extended or trees frequent dry periods) d. Maintain high infiltration rates in groundwater recharge zones • e. Maintain vegetative cover to minimize erosion 2. Flood damage/flooding a. Reservoir storage a. Minimize sediment (due to occupancy of b. Construct levees, delivery to reservoir site; flood-plain areas, improve channels maintain watershed increased surface runoff c. Manage flood plain vegetative cover from watersheds, and d. Revegetate disturbed b. Minimize sedimentation of inability to store and areas downstream channels regulate stormflows) c. Zone lands to restrict human activities in flood- prone areas. Minimize sedimentation of channels d. Implement afforestation or reforestation of denuded watersheds; encourage natural revegetation 3. Degraded watersheds; a. Erect erosion control a. Maintain life of structures high rates of erosion and structures by revegetation and sedimentation b. Build contour terracing management c. Revegetate b. Revegetate, stabilize slopes and terraces, and institute land-use guidelines c. Protect vegetative cover until site recovers; use reseeding, fertilization, etc. 4. Polluted drinking water a. Develop alternative a. Protect groundwater from (due to improper supplies from wells, contamination development of wells, springs b. Protect catchments from improper sewage b. Treat water supplies contamination treatment facilities, and contamination of surface water supplies) PRODUCTION OF FOREST PRODUCTS 463 EXHIBIT 17-1. Water-Related Problems, Their Causes, and Some Solutions Using a Watershed Approach (continued) 5. Polluted streams/reduced a. Establish/maintain a. Develop buffer strips along fishery production vegetation cover on stream channels and (due to inappropriate land watershed manage to sustain use, improper treatment b. Treat wastewater vegetative cover of uplands of wastewater, and c. Protect streambank b. Use natural systems streambank degradation) vegetation (forests and wetlands) as secondary treatment systems of wastewater c. Control grazing and develop guidelines for riparian vegetation zones 6. Food shortages a.Develop agroforestry a. Maintain site production; (due to erosion and practices minimize erosion (nutrient reduced production b. Increase cultivation losses); develop crops capacity of the soil, c. Increase livestock compatible with soils and population increases, production climate of area waterlogging, and salinity d. Import food from b. Restructure steep hill resulting in losses of outside slopes and other areas agricultural productivity) e. Drain waterlogged soil susceptible to erosion; use contour plowing, terraces, etc. c. Develop herding-grazing systems for sustained yield and productivity d. Develop forest resources for pulp, wood products, etc., to provide economic base e. Maintain drainage ditches 7. Energy shortages a.Develop fuelwood a. Develop sustainable (due to loss of fuelwood resources fuelwood plantations and supplies-reduced forest b. Encourage agroforestry guidelines that maintain cover, population systems that include productivity and protect increases, loss of fossil multipurpose fast- soils fuels, inadequate growing trees b. Develop agroforestry infrastructure to distribute c. Develop hydroelectric systems that reduce erosion fuel, accumulation of power projects (e.g., and produce fuelwood sediment in reservoirs minihydro projects, c. Minimize sediment that provide hydropower, dams/reservoirs) delivery to channels and and reduced efficiency of reservoir pools; use turbines) vegetation management and structural solutions SOURCE: Adapted from Brooks et al. (1990). 464 MANAGING THE WORLD'S FORESTS ' EXHIBIT 17-2. Relationships Between Physkal Effects, Environmental Changes, and Downstream Benefits from Wa1tershed Management Practices, as Compared with the "Without" Practices Condition DIRECT OUTPUTS LAND FROM lAND USE: USE food, fiber, fuel, forage, water WATERSHED MANAGEMENT PRACTICE'S AND PRINCIPLES are applied, then there are opportunities to: Increase direct'f-~----------.,'----------~ outputs INCREASE IMPROVE _ _ _ _ _ STREAMFLOW _ _ _ _ _ IMPROVE SOIL PATIERN/ WATER STABILITY QUALITY VOLUME Reduce downstream ENVIRONMENTAL sedimentation CHANGES Reduce Increase reservoir/ Reduce: sedimentation channel capacity Reduce • nutrient loading effects on and capability to maximum • toxic substances infrastructure regulate flow flow • thermal pollution • undesirable organisms Increase ,,,__ ___.., streamflow during i---...__ critical low -flow periods Increase Increase Increase Increase Reduce waterborne Increase industrial disease and hydropower navigability irrigation water fish increase potable potential capacity production watersu I SU lies Increased Greater use Increased Increased crop and Increased Increased Increased of floodplain hydropower tonnage livestock industrial fish productiviity production shipped roduction production haivest and health DOWNSTREAM BENEFITS SOURCE: Adapted from Gregersen et al. (1987). PRODUCTION OF FOREST PRODUCTS 465 EXHIBIT 17-3. Examples of Watershed Management Practices to Be Considered in Designing a Project, Grouped According to Effects as Illustrated in Exhibit 17-2. SOIL STABILIZATION PRACTICES Structural Alternatives Gully plugs Bench terraces for crops Broad-base terraces • with soil risers • with stone walls Contour furrows/trenches-use of native vegetation Orchard terraces Hillside ditches with benches for upland crops Grassed waterways and ditch systems to convey water and protect against gully formation Nonstructural Alternatives Reforestation: protection forest, industrial forest, fuelwood plantation Agroforestry: array of practices to provide for soil stabilization and for food and wood production (windbreaks, alley cropping, etc.) Establishment of grass cover (quick soil protective cover and forage production) Management Practices Zoning of land use according to soil-slope criteria; land capability classification on hilly lands Promotion of agroforestry practices over intensive cultivation of steeper lands Best management practices-livestock Best management practices-forests Criteria for controlled road construction Protection of riparian vegetation by development and maintenance of buffer strips of vegetation along stream channels WATER YIELD/STREAMFLOWPATTERN Structural Alternatives Water storage facilities, such as dams and reservoirs Water harvesting systems and storage tanks Channel improvement/ channelization Water spreading and irrigation diversions Development of wells Nonstructural Practices Changes in type of vegetation • reforestation/afforestation to reduce water yield and surface runoff • conversion from deep-rooted to shallow-rooted species to increase water yield Delineation and zoning of flood plain IMPROVEMENT OF WATER QUALITY See practices above that reduce sediment delivery into streams Management of fertilizers and pesticides to prevent entry into water bodies Establishment of best management practices for logging and cultivation Establishment and enforcement of regulations to reduce human/livestock waste from entering surface and groundwater supplies SOURCE: Details of these practices are described by Gil (1979). 466 MANAGING THE WORLD'S FORESTS EXHIBIT 17-4. Examples of Watershed Management Practices to Be Considered in Designing a Project, Grouped According to Location Within a Watershed Location on Watershed Watershed Practices Management Purpose Lowlands Croplands Cross-ditch system Improve surface drainage Bedding system Improve surface drainage Contour cultivation Conserve soil and water Stubble mulching Conserve soil and enhance infiltration Infiltration galleries Irrigate Other lands Well and spring development Increased water supplies Dams, ponds, reservoiis Water development and flood control and fisheries Channel improvement Flood control (levees, diversions, channelization) Uplands Croplands Strip cropping/ alley Conserve soil and water cropping Broad-base terraces Conserve soil Contour banks Conserve soil Interception ditches Reduce water flow Diversion ditches Reduce water flow Improved natural drainage/ Prevent erosion grassed waterways Bench terraces Conserve soil and water Pasture/ Fencing Control grazing/protect water Rangelands systems Reseeding Conserve soil Planned grazing Improve range/pasture condition/ soil conservation Water harvesting Increase water supplies Contour furrows/seeding Conserve soil and water Forest lands Reforestation and Stabilize soil/reduce surface management runoff/streamflow Shelterbelts Reduce wind erosion Agroforestry Stabilize soil Steep and Gully control structures Reclaim gullies degraded lands Revegeta tion/ reforestation Stabilize soil and control runoff Zoning of hazard areas Protect against landslides · Elimination of use or Allow soil and vegetation to protection recover ReforestatiOn in headwater Control torrent channels and protection Roads Construction guidelines/ Reduce landslides and road regulations established washouts Minimized cut and fill Reduce erosion Drop structures with Control runoff and erosion vegetated waterways SOURCE: Details of these practices are described by Gil (1979). PRODUCTION OF FOREST PRODUCTS 467 Watershed rehabilitation projects have received the greatest emphasis in the past and sometimes are viewed as the only reason for watershed management. Such projects have been carried out in countries and regions that fall in all types of climatic regions, topography, and population catego- ries, but much effort has been directed toward mountainous regions in the tropics, subtropics, and drylands. For example, the Himalayan region, which ranges from humid temperate to humid tropical in the high-population cat- egory, has experienced considerable development efforts. Countries such as Nepal, India, Pakistan, and China would fall in this category. Mountainous tropical areas in South America (e.g., Colombia, Ecuador) and Africa (Rwanda, etc.) similarly have received attention. These areas are characterized by steep slopes and erosive soils, high rainfall amounts and intensities, and high populations and intensive land use. Such areas require special measures to stabilize soils, to reforest steep slopes (deep-rooted species are preferred for slope stabilization), to control road/trail construction on fragile slopes, and to develop means of provid- ing food and resources for the numerous rural inhabitants. Measures that have been employed with varying success include stall-feeding of animals and hand-lopping of forage from steep slopes to reduce grazing pressures, control of road construction, reforestation and protection of the most fragile slopes, and the promotion of appropriate agroforestry practices. Human pressures to cultivate forest lands, to graze with livestock, or to meet fuelwood needs can compromise reforestation and rehabilitation ef- forts, leaving compacted soils that do not have the hydrologic stability of a natural forest, and mixtures of vegetation (e.g., sparse tree cover with thin canopies, shrubs, grasses, croplands). Even when trees and forest planta- tions can be established on steep uplands, the overall soil stability may not be the same as that of natural forests unless the understory and soil surface conditions are managed appropriately (Wiersum 1984; Brooks et al. 1991). In sum, the mere planting of trees does not necessarily improve watershed conditions unless it is accompanied by watershed management practices that reflect sound land use. The challenge in populated, mountainous regions in the tropics is to build in management strategies so that conservation measures will continue to be practiced following rehabilitation measures. In many fragile upland watersheds, particularly steep mountainous lands, the carrying capacity, in terms of livestock and human populations, must be kept in balance with land capability to support those populations. Even though improving tech- nologies permit the limits of such systems to be stretched to some extent, developers cannot continue to overtax these ecosystems indefinitely. In regions of high populations and watershed degradation, whether mountainous or not, planners should consider the opportunities to integrate trees into agricultural development projects with multiple benefits. Agroforestry practices that improve crop productivity, reduce soil erosion, and provide multiple products from the trees themselves (e.g., fuelwood, fodder, poles, lumber, fruit, nuts) stand a greater chance for sustainabil- ity because they can provide products or income relatively quickly to local 468 MANAGING THE WORLD'S FORESTS farmers while improving long-term productivity. However, indigenous spe- cies and local know-how should be preferred over the use of exotic, monoculture systems, which often are touted as a quick fix to problems. The oversell, and subsequent failure, of Leucaena leucocephala for dendrothermal plantations in the Philippines is a good lesson for using diversified, native species whenever possible. Agroforestry practices in dryland-lowland areas have reduced wind erosion, enhanced crop production, and furnished fuelwood supplies; an excellent example is the Majjia Valley Reforestation Project, Niger, West Africa (Harrison 1987). Strategies for such areas should focus on improving fuelwood, fodder, and water supplies, which can be as critical as food to the rural poor. Details of forestry practices for dryland environments are pre- sented in a recently developed guide by FAO (1989). In all the cases just mentioned, no matter how serious the problems of watershed degradation, the level of population density dictates the water- shed management strategy and practices. In most cases, trees are desirable components of soil stabilization practices, along with gully plugs, gabions, and other structural measures. Stabilization is required before development can proceed. Sometimes grasses and shrubs may be used to get a quick vegetative cover established to control surface soil erosion, so that forest vegetation can become reestablished later. If the area has low population density, stabilization characteristics became more important and species with the best rooting characteristics are preferred. If population pressures are great, trees that produce fruits, nuts, or other products may -be given pref- erence over trees with better stabilization characteristics but with less pro- duction value. A good understanding of the various interrelationships is critical to the development and coordination of projects so that activities are not carried out that are at cross-purposes. For example, to encourage reforestation and protection of steep-sloped watersheds above reservoirs and to compensate for wood products that previously came from these fragile lands, commu- nity forestry programs may need to be implemented elsewhere on the wa- tershed. If fuelwood needs are not addressed, the protection forests above the reservoir will likely fall prey to these needs. Conversely, community forestry programs also should recognize the need to provide protection forests for fragile lands; planting trees near villages does not necessarily help stabilize steep uplands. Doolette and Magrath (1990) imply that watershed rehabilitation of uplands solely for improving downstream environmental quality (i.e., sedi- mentation of reservoirs, flooding) cannot be economically justified. How- ever, they emphasize correctly that by using soil and water conservation principles to intensify efforts that improve the productivity of upland farm- ers in an environmentally sound manner, benefits usually will accrue in downstream areas. This is true in terms of erosion, but sometimes there can be unwanted changes in water flow if the hydrologic effects of land-use practices are not properly understood (see example 1). PRODUCTION OF FOREST PRODUCTS 469 EXAMPLE 1. CHANGES IN WATER YIELD ASSOCIATED WITH AFFORESTATION (FROM DRYSDALE [1981], AS REPORTED BY GREGERSEN ET AL. [1987]) Unanticipated changes in water yield were encountered in Fiji when 60,000 hectares of Pin us caribaea were planted on the leeward zones of Viti Levu and Vanua Levu, areas that previously had mission grass cover. On the island of Viti Levu, a water supply dam with hy- dropower capabilities was developed coincident with the afforestation. Pine plantations were expected to yield an increased and more uniform water supply than the vegetative cover of mission grass did. Although annual rainfall in the windward zones can exceed 4,800 millimeters, rainfall during the dry season (May-October) on the leeward slopes is 300 to 500 millimeters, with long dry periods. After six years, the pine plantations reduced dry season flows by 50 to 60 percent. Further reductions would be expected when the pine reaches maturity. In this instance, there was apparently coordination between the forestry and water agencies, but the hydrologic linkages between upland forestry and downstream interests were not properly evaluated. Watershed Management as an Integral Part of Forestry and Related Projects Any project that involves the manipulation of soil, vegetation, or water must contain built-in guidelines that minimize unwanted impacts on the environment and productivity of uplands and downstream areas. Two situ- ations can be considered: 1. Projects that concentrate on improving forest productivity, and 2. Projects that include trees but focus on improving agricultural and multiple-resource productivity, such as agroforestry practices (whereby trees can stabilize soil, enhance crop production by nitrogen fixation, etc.). In either case, the issue is how to meet production objectives in an environmentally sound manner. Management practices that sustain produc- tion by preventing excessive erosion and other environmental degradation include some practices that might not initially be thought of as watershed management-for example, maintaining livestock in numbers that do not exceed carrying capacities of rangelands, or specifying restrictions on types of logging-skidding operations for certain types of soil and slopes. Other practices include requirements that vegetation buffer strips be left along stream channels or wetlands to protect aquatic ecosystems from logging without constraining production. 470 MANA.GING THE WORLD'S FORESTS Guidelines for management that promote soil conservation and reason- able water resource management goals, with a theme of preventing losses over the long term, are referred to as best management practices (BMPs). In the United States, BMPs are aimed primarily at protecting water quality and related environmental characteristics of watersheds that are undergoing management. Such practices are legally the responsibility of each state, which is responsible for specifying environmentally sound management practices for forestry, agricultural, and mining activities. BMPs include logging guide- lines, restrictions on use of riparian vegetation areas, grazing management guidelines, road construction and road development guidelines, and waste- water treatment requirements. The enforcement of such practices can im- pose an increased cost in operations over the short term to avoid losses over the long term. It is important to emphasize that losses avoided have as much social and economic value as do increases in productivity. Establishing BMPs or a similar approach requires more than determin- ing the appropriate techniques for land management. The importance of improving land management requires that the physical-biological and insti- tutional linkages of a watershed be understood at all levels of society. Pub- lic education and professional educational and training programs can en- hance this awareness. Neither local communities nor governments can ignore the physical linkages. Nor can planners of sustainable projects in rural areas ignore land tenure, institutional capacity, and the culture of watershed in- habitants. More emphasis is needed on the development of human resources and institutional capacity and less on infrastructure. The International Tropical Timber Organization, in its 1990 program, prepared guidelines for best-known practices in the sustainable develop- ment of tropical forests. These guidelines include conservation and environ- mental benefits as well as economic benefits. The guidelines also take into account the issues of equitable distribution of costs and benefits and the need to adapt land-use practices to local ecological and socioeconomic con- ditions (ISTF 1990). ANALYSIS AND ESTABLISHMENT OF PRIORITIES As already mentioned, watershed interventions can be considered in terms of projects aimed primarily at rehabililtation and those aimed prima- rily at improving production, both in terms of forest and agricultural products. This section discusses an approach for analyzing the options and establishing the priorities for the two different types of projects. Forestry Projects Aimed at Watershed Rehabilitation In formulating watershed rehabilitation projects with forestry compo- nents, analysts need to determine the biophysical potential of the uplands PRODUCTION OF FOREST PRODUCTS 471 receiving watershed management interventions, the potential to enhance downstream benefits, and the capability to implement .the project. This does not mean that economically significant impacts downstream are essential in all cases, but there always must be tangible benefits in the project area. Furthermore, benefits must accrue to the rural inhabitants of watersheds, usually subsistence farmers. Population pressures are considered to be a major determinant in developing an appropriate strategy. Using field and remote sensing techniques, analysts need to examine the project area for the degree and extent of watershed degradation and the potential for watershed practices to improve on-site productivity. The latter requires knowledge of soil erosion-productivity relationships in the physi- cal watershed setting; such information often is not known. For example, if most of the soil has been eroded and the site is unproductive parent mate- rial, the potential to improve productivity (or avoid losses over time) would be poor. The effects of such interventions on downstream areas then can be considered. The analyst can rank watersheds or areas within the watershed (subwatersheds) that should receive interventions. Only an experienced professional can carry out such an assessment. The next step in the analysis is to consider the capability to implement the interventions that have been identified in the preceding step. The capa- bility to implement a project involves aspects of three factors: (1) policies, which reflect social, economic, and political needs; (2) institutions, which reflect social, economic, and political conditions and are determined by operational needs; and (3) operational capabilities, which relate to the re- sources available to implement the project and are influenced by policies laid down and institutions created to support a program. The overall capability to implement a successful project is only as good as the weakest of the three factors. If there is a strong policy with the necessary strong institutions but a weak operational capability, the overall capability is weak. Similarly, if operational capabilities are good but institu- tions are weak, the overall capability is weak. Watershed projects need to consider strengthening the appropriate links as a matter of priority. The major issues in terms of policies, institutions, and operational capa- bilities to carry out watershed management components of projects are as follows: Policies • Natural resource management (soil, land, water, forest, minerals, ten- ure, user rights, etc.) • Population • Development (roads, markets, food and energy security, industries, etc.) • Fiscal • Education 472 MANAGING THE WORLD'S FORESTS Institutions Within a country • Interagency agreements to coordinate upland and downstream inter- ventions • Extension and other capabilities to motivate farmers • Laws and regulations that promote best management practices and restrict or control overuse of natural resources • Training and educational opportunities and capabilities • Planning, monitoring, evaluation Donor related • Coordination between and among bilateral and multilateral agencies, nongovernment organizations, and others • Institution building versus local resource consumption • Technology promotion; high-tech approaches versus local and tradi- tional approaches Operational Capabilities • Technology • Facilities • Availability of maps, soil surveys, aerial photography, and the like • Personnel • Fiscal situation • Sociological factors, such as the role of women and the participation of local people • Others Problems of user rights to forests and land rights often lead to conflicts between local farmers and government agencies. Planners of reforestation projects for purposes of rehabilitation should fully consider tenure arrange- ments. Population pressure comes into the analysis in terms of appropriate strategies and methods. Of course, population pressure depends on the ca- pability or carrying capacity of the land and should include considerations of people and livestock. For example, in some tropical areas a figure of 250 persons per square kilometer can be used to demarcate high population pressure. If there are large populations living in the watersheds (e.g., Indian Himalayas, South Asia, Java in Indonesia, Philippines), the strategy should emphasize increasing production (e.g., food, fuelwood) in its rehabilitation options. The participation of local people must be stressed in the planning and implementation of the project. The goals of the project must be the same as the goals of the local farmers. When population pressures are low (e.g., Kalimantan, Indonesia), more emphasis can be placed on conservation measures that use local species and protect or maintain some of the natural flora and fauna while meeting the -- - -- ------------------ - - - - - - - - PRODUCTION OF FOREST PRODUCTS 473 production needs of local farmers. Production goals and local participation remain important here as well, but planners may have greater flexibility to combine objectives such as maintaining biodiversity, stabilizing or protect- ing steep slopes, and improving productivity. Watershed management activities can produce competitive economic returns on investment of a country's scarce capital, even without adding in many of the nonmonetary environmental downstream and upstream benefits that can flow from such projects (see example 2). Oftentimes the upstream benefits alone can justify watershed management activities and investments. When downstream benefits are added in the results can be even more favorable. Strategies for forestry projects should be directed to those with poten- tial to succeed and those with serious problems. To be successful, projects must meet the needs of people whether in areas of high or low population. Ultimately, in all cases, efforts should be directed toward establishing and adhering to best management practices for rural inhabitants. EXAMPLE 2. ECONOMIC APPRAISAL OF A WATERSHED MANAGEMENT PROJECT IN MOROCCO (BROOKS ET AL. 1982) A watershed management project was proposed for the 182,000- hectare Loukkos watershed in northern Morocco with the main pur- pose of reducing soil erosion and sediment delivery into the down- stream Oued El Makhazine reservoir that was completed in 1979. More than 50 percent of the watershed was considered to exhibit severe soil erosion. Land use consisted of 42 percent in cultivation, 28 percent in rangeland, 21 percent in forest cover, and 9 percent in urban and industrial use. The multipurpose reservoir provides irrigation water to 25,200 hectares downstream as well as hydroelectric power generation, municipal and industrial water supplies, and flood control benefits. Protecting the functional life of this reservoir could therefore have considerable economic benefit. The project included the following prac- tices: road construction, channel stabilization, gully control, reforesta- tion and pasture management, and pl'J_nting of olive trees, affecting an area of 40,000 hectares. An economic analysis was performed in which the upstream and downstream effects of watershed management practices designed for the project were examined and contrasted to the effects that would be expected without the implementation of the project. The analysis ap- proach involved assessing the situation "with" and "without" the project The analysis suggested that even estimating project effectiveness (continued) 474 MANAGING THE WORLD'S FORESTS Example 2 (continued) on erosion and sediment<1tion conservatively, and considering the sub- sequent changes associated with upland productivity and downstream effects (as indicated in exhibit 17-2), the project could increase the life of the reservoir over the conditions without the project. Project benefits included the avoidance of loss of irrigated crops downstream and increased productivity of olives in the uplands. Other benefits that were recognized, but not quantified in the analysis, included improved wood production, sustained hydroelectric power generation and extended life of turbines, sustained municipal and industrial water supplies, sus- tained flood control benefits, and reduced costs of fertilizers that result from soil erosion protection. Nevertheless, the internal rate of return (!RR) from the project was 15.9 percent, and the net present worth of the project, using a 10 percent discount rate, was $18.8 million. Recognizing the uncertainty that is associated with such analyses, the costs of the project were assumed to be 25 percent higher than estimated and the benefits to be 25 percent lower than estimated. The resulting /RR was still 12. 1 percent, indicating that even with rather significant changes in costs and benefits that are unfavorable to the project, the /RR was above the 10 percent social discount rate consid- ered to be relevant for Morocco. Forestry Development Projects with Primarily a Production Objective For forestry projects that are aimed principally at improving productiv- ity, the key watershed management considerations are to avoid unwanted environmental consequences and, in some cases, to consider how soil and water production can be enhanced. The following guidelines determine the watershed management components of a project: 1. For areas with proposed forestry I agroforestry interventions, identify the watershed boundaries and evaluate the possible con- sequences of the project in terms of soil and water relationships (Gregersen et al. 1987). 2. Consider the implications of trees in te:rms of watershed values and production of forest and agriculture products: • When, where, and under what condiitions does the introduc- tion of trees and forestry advance the opportunities for im- proving human welfare on a sustainable basis? (Consider erosion control and the provision of fuelwood, green ma- nure, fodder, structural wood, poles, etc.) PRODUCTION OF FOREST PRODUCTS 475 • Where should priority be given to protection? (Identify criti- cal areas that cannot sustain intensive cropping, logging, grazing, and the like.) 3. Consider whether the policies, institutions, and operational capabilities exist to introduce, promote, and sustain project ele- ments: • What are the incentives that will motivate local farmers? • Do the project staff have the education, training, extension, and in some cases mechanisms to offset inequities that arise? 4. Concentrate on strengthening those components that offer the greatest obstruction to the achievement of a successful and sus- tainable project. BARRIERS TO ADOPTION OF WATERSHED MANAGEMENT Ironically, the basic concepts underlying a watershed management ap- proach also explain in part why this approach has not been more widely adopted (Working Group on Watershed Management and Development 1988). Because the main effects of the water and land-use practices of one political unit often are felt by people outside that unit or by future genera- tions who cannot vote now, there has been little incentive to consider the concepts of watershed management that account for these interactions. A common question from upstream land users is, Why should we carry out watershed management practices to benefit those downstream? Indeed, why should developers expect them to, if they are not compensated for the costs of such activities? An attempt to deal with the inequities of who pays and who benefits in a watershed management project is described in example 3. Although the concept of taxing downstream beneficiaries to help pay for upland conser- vation works is a good one, the institutional capacity must exist to imple- ment the program. 476 MANAGING THE WORLD'S FORESTS EXAMPLE 3. UPPER MAGDALENA WATERSHED PROJECT, COLOMBIA, SOUTH AMERICA (WORLD BANK 1990) Denudation of forested hillsides and the resulting erosion, water pollution, and the sedimentation effects on downstream hydroelectric power generation led to the World Bank project. A pilot operation was developed with the overall objective of developing viable farming and tree crop systems on the Upper Magdalena Watershed, with the follow- ing specific objectives: • Developing the institutional mechanisms for carrying out appro- priate soil conservation and erosion control measures; • Determining feasible and low cost ways of introducing soil con- servation to farming systems for overall watershed protection; and • Determining the incentives required to motivate land users to adopt conservation measures and to determine the appropriate role for the public sector in protecting Colombia's water re- sources. Two key features in this project were to provide credit to farmers for improving soil conservation practices ancf to establish a Soil Conser- vation Fund with contributions from downstream beneficiaries (rice growers and electric utilities) to be used by upstream farmers to control erosion. The results of this project were not. well documented. In terms of production and protection goals there was inadequate monitoring of the project to determine whether components of the project were successful. The project was not fully implemented; for example: • Forests were to be planted on 660 hectares of publicly owned lands for watershed protection and commercial exploitation, but only 148 hectares were planted. • Although 17 civil works and river control schemes were desig- nated, only 2 were constructed. • Only a fraction of the planned agriculture and tree-crop expan- sion was carried out. (For example, 200 hectares of fruit trees were planned, but only 37.5 hectares were planted.) • Credit to farmers could not be mobilized because of in-country constraints on loans and institutional problems associated with collateral requirements. Only a small portion of the $8.9 mil- lion for on-farm investments was extended by financial inter- mediaries. (continued) PRODUCTION OF FOREST PRODUCTS 477 Example 3 (continued) The key lessons learned from this project were as follows: • The viability of watershed management systems was to be evalu- ated on the basis of the final return to farmers. This was too idealistic; intermediate steps needed to be monitored as well. There are many reasons why the final returns can be high or low that are independent of the project. • Downstream entities were reluctant to contribute for up-stream work without seeing evidence of tangible benefits. Monitoring was not adequate and research or demonstration projects were not established to quantify such benefits. The scale of this project is probably too large to detect direct downstream benefits in any event, unless subwatersheds were monitored as well. • Although environmental protection control units were to be established for purposes of monitoring and protecting forest reserves, none were established. • The technical proficiency to coordinate was inadequate and the monitoring and evaluation systems were not in place. Other barriers to watershed management include the following: • The lack of awareness or understanding of watershed management concepts and practices by development professionals and by the pub- lic has limited the application of watershed management practices. • Technical experts in this field have not cooperated adequately with development practitioners and administrators. Only recently have technical experts made a concerted effort to explain, in language understood by the pragmatic development professional, how water- shed management can aid in development programs aimed at in- creasing food security, employment opportunities, economic growth, and poverty alleviation, all within a sustainable development frame- work. • There has been skepticism about downstream benefits of watershed management, and projects have provided little quantitative informa- tion concerning downstream effects (and, in some cases, no informa- tion about impacts on productivity). Projects have not been moni- tored sufficiently so that the effectiveness of land-use changes can be evaluated. For example, Doolette and Magrath (1990) found that 27 of the 34 completion reports on World Bank projects contained "no quantitative information" that could be used to evaluate the projects. • There has been a general misunderstanding of forest hydrology, water quality, and sedimentation relationships; many professionals in de- velopment work do not understand and cannot separate human-caused i 478 MANAGING THE WORLD'S FORESTS effects (from land-use practices) froin effects that are natural phe- nomena (see example 4). Erosion, flooding, landslides, and so forth are naturally occurring phenomena that will occur no matter what we do. Jn areas of active mountain building, natural geologic erosion results in unstable slopes, landslides, and high levels of sedimenta- tion. To some extent, however, human actions on the watershed can affect the frequency of occurrence and severity of natural phenom- ena. Deforestation followed by improper cultivation practices or over- grazing can influence flooding and sedimentation. The key is to be able to decide whether there is potential to improve conditions through changes in land use and the introduction of watershed management practices. EXAMPLE 4. MUDSLIDES AND FLOODING: UNDERSTAND/NG THE LIMITS OF WATERSHED DEGRADATION AND MANAGEMENT EFFECTS The challenge in assessing the potential impacts of forestry-water- shed interventions in many instances is to be able to understand the magnitude of the naturally occurring processes and impacts. The rash of landslides, mudslides, and flooding and the resulting loss of 200 lives in southern Thailand in 1988 were attributed in some reports to defores- tation and conversion from natural forest ito rubber plantations (Rao 1988). The publicity of this event led to the banning of all logging in Thailand. A hydrologist, however, would quickly point out that when a rainfall of 1,022 millimeters occurs in a three-day period, mudslides and floods will occur no matter what type of vegetative cover exists. • There has been a Jack of technical expertise in the planning, design, and implementation of watershed management projects. Trained spe- cialists who are experienced in watershed management, hydrology, geology, and so on are needed early in the project identification and appraisal phase and in the design and monitoring of projects. Such expertise has rarely existed on teams that have developed projects in the past (see example 5). PRODUCTION OF FOREST PRODUCTS 479 EXAMPLE 5. LACK OF TECHNICAL EXPERTISE IN PROJECT FORMULATION AND IMPLEMENTATION Technical expertise is often Jacking in at least one phase of a project. No matter how well planned a project is, if the expertise is not available to implement the project then success is unlikely. likewise, if the operational capability exists, but the project is inappropriately planned, the project will likely fail. The following examples illustrate these points. 1. As part of a larger watershed management project aimed at reforestation. range improvement, and erosion control, numerous rock- wire (gabion) gully plugs were installed in the Tleta watersheds in northern Morocco. The opportunities to measure the rates of sedimen- tation behind these structures provided an excellent chance to quantify erosion-sedimentation rates at the initial part of the project and to follow through with annual measurements during the project to assess the effectiveness of management (structural and nonstructural) mea- sures. However, within weeks after the gully control structures were installed, the people who constructed the gully plugs were found to have backfilled the structures, rendering them useless. Proper project management and field checking should have ensured the proper instal- lation of these structures. Although data concerning the costs of these structures were not available (such structures are costly to construct), the benefits would be negligible without additional efforts and costs to remove the inappropriately backfilled soil material. Furthermore, there were no plans to control grazing in the area after the gully plugs were constructed. Therefore, even had they not been backfilled, the gully plugs would not have functioned for long. 2. While working in the Rif mountains of northern Morocco in 1979, one of the authors encountered a team from the World Bank investigating the needs and opportunities for a watershed project. The World Bank team members were surprised to learn that watershed management specialists were working in the area. In fact, they were unaware that such specialists existed. No specific expertise in watershed management soil conservation, or hydrology was represented in the team membership. • Unrealistic expectations from watershed and forest interventions have led to suggestions that watershed management does not work. For example, any developers planning forest plantings and associated watershed management practices aimed at reducing sedimentation of 480 MANAGING THE WORLD'S FORESTS downstream reservoirs (gully plugs, changes in farming practices, range management, etc.) need to consider- -the amount of watershed area affected (i.e., the scale of the prob- lem) and the proximity of forestry and other activities in the water- shed to the reservoir; -the natural erosion-sedimentation processes; -existing levels of sediment in stream channels that feed the reser- voir; and -other practices occurring on the watershed. The watershed management project above the Mangla Dam in Pakistan illustrates the problems associated with failure to consider scale and other factors carefully (see example 6). EXAMPLE 6. MANGLA DAM, PAKISTAN (FROM MITCHELL 1986, FAO-UNDP 1982)1 A watershed management project was initiated in 1960 to reduce the high levels of sedimentation into the Mang/a Reservoir. Structural and nonstructural rehabilitation measures were undertaken to control erosion and reduce sedimentation from 40,000 hectares of critically denuded areas. The watershed project was determined to be ineffec- tive in reducing sediment delivery to the reservoir. However, inspection of the watershed and the Jhelum River that drains into the Mang/a Reservoir indicated that (1) the watershed management rehabilitation efforts were carried out on only a small percentage of the large water- shed above the dam, (2) there is a high level of natural (geologic) erosion and sedimentation, and (3) road construction above the reser- voir has caused numerous landslides that deposit sediment directly into the Jhelum River that feeds the reservoir. The third point is significant; had there been efforts to reduce road construction across steep slopes and had there been adequate monitor- ing, some reductions in sedimentation might have been detected. The high level of natural erosion and sedimentation in this region plus the large scale of the watershed, however, would not provide the best opportunities for reducing sediment delivery to the reservoir. Some people might use this example to suggest that watershed management practices are ineffective in protecting the life of a reservoir. Unfortu- nately, such a conclusion could lead to inappropriate decisions in smaller upland watersheds and reservoirs where the possibilities of reducing sedimentation are better. 1Some of the information is based on personal observations by Kenneth Brooks. PRODUCTION OF FOREST PRODUCTS 481 OVERCOMING THE BARRIERS The barriers to wider adoption of a watershed management approach are being broken down slowly but surely. Many decision makers now rec- ognize the imperative of environmentally sound and sustainable develop- ment. Ignoring the boundaries and interrelationships set by the forces of nature will inevitably lead to serious, if not disastrous, problems. The challenge in introducing a watershed management approach is not to change the world by replacing current land-use practices with some cure- all watershed managementformula. Development within a watershed man- agement framework does not mean populating the world with professional watershed managers who direct projects and activities of people living in a given watershed; nor does it mean establishing a great number of isolated watershed management projects. Rather, watershed management concepts and practices should be introduced mainly as integral components in for- estry, agricultural, hydropower, irrigation, and related rural development projects. These components have to be implemented and sustained by people other than professional watershed managers-by general project adminis- trators, foresters, agriculturalists, sociologists, hydrologists, and, most im- portant, by farmers and other land users. The implication is that these types of people need to understand why and how watershed management should be incorporated into everyday activities. Nongovernment organizations (NGOs) have taken a small-scale approach toward watershed management solutions with a strong emphasis on local people. For example, in Asia, such organizations have been promoting en- vironmental awareness and encouraging local-level afforestation. Their projects have played a crucial role in representing local interests, and act as intermediaries between governments and local communities (FAO 1987). Formulators of watershed projects should consider questions such as these about the socioeconomic and cultural setting: What are the indigenous abilities and potential of the rural farmers? How can local people become partners in the planning and implementation of projects? What are the insti- tutional or regulatory factors that can change land-use practices? What are the financial resources of local people? What is the history of the use of mechanisms such as educational programs, subsidies, and other incentives to get changes in technology for the local area, the region, and the country? What markets exist? What infrastructure is present? How can projects get the support of the local rural people (see example 7)? - - - -- - - ----- 482 MANAGING THE WORLD'S FORESTS EXAMPLE 7. BACKGROUND INFORMATION AND DEFINITION OF THE PROBLEM FOR THE MAE CHAEM WATERSHED DEVELOPMENTPROJECT, CHIANG MAI, THAILAND (FROM PUNYATRONG [1985]) The 4,200-square-kilometer Mae Chaem watershed has about 40,000 inhabitants belonging primarily to two ethnic groups, the Skaw Karen (47%) and the northern Thai (45%). More t:han three-fourths of these people live below the poverty level, they experience poor health and educational services, and they Jack most basic government services. Agricultural extension and credit facilities 5 0.0 0 0 0.0 56 Korea, DPR 280 280 100 100 0.0 0 0 0.0 80 80 0.0 57 Korea, Rep. 4,677 5,348 65 75 3.9 35 25 -3.9 1,680 3,659 10.4 58 LaoPDR 43 21 96 76 -7.8 4 24 9.9 0 0 0.0 59 Malaysia 7,450 7,706 86 82 -0.6 141 18 2.6 71 70 1.0 60 Mongolia 474 474 99 99 0.0 1 1 o.p 0 0 0.0 61 Myanmar 444 498 97 97 -0.8 3 3 1.5 10 8 0.9 62 Nepal 220 220 100 100 0.0 0 0 0.0 2 2 0.0 63 New Zeal. 2,381 2,445 87 80 -0.8 13 20 5.4 674 700 -0.4 64 Pakistan 103 140 54 39 0.0 46 61 8.6 63 81 3.3 65 P. N. Guinea 207 136 90 86 -3.6 10 14 -0.4 0 0 0.0 66 Philippines 2,288 1,573 67 66 -3.4 33 ' 34 -3.0 323 335 0.6 67 Saudi Arabia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 68 Sri Lanka 41 30 64 68 -4.4 36 32 -5.6 18 28 5.1 69 Syria 36 36 25 25 0.0 75 75 0.0 3 19 22.5 70 Thailand 1,693 1,295 91 81 -1.3 9 19 4.5 338 514 6.1 71 Turkey 5,096 5,704 91 86 1.3 9 14 9.7 478 400 -2.4 72 Viet Nam 491 394 96 90 -3.2 4 10 10.5 37 56 6.5 73 Yemen 0 0 0 0 0.0 0 0 0.0 0 0 0.0 CENTRAL& SOUTH AMERICA 30,189 35,679 86 85 2.2 14 15 2.2 7,730 11,287 2.5 74 Argentina 1,297 1,805 68 80 5.4 32 20 -1.1 713 974 5.4 75 Belize 21 14 100 100 1.7 0 0 0.0 0 0 0.0 76 Bolivia 236 103 93 92 -9.0 7 8 -7.8 1 2 17.6 77 Brazil 17,363 21,165 86 86 2.4 14 14 2.1 3,361 4,639 5.7 78 Chile 2,412 2,981 95 91 5.8 ' 0 9 12.2 356 449 4.3 79 Colombia 1,081 834 90 86 -3.6 10 14 0.1 351 501 4.4 80 Costa Rica 592 573 89 90 0.4 11 10 -2.0 12 17 0.5 81 Cuba 118 255 95 46 0.7 5 54 47.5 73 141 9.7 82 Dom. Rep. 0 0 0 0 0.0 0 0 0.0 9 10 1.1 83 Ecuador 992 1,451 91 88 4.6 9 12 8.5 26 54 8.0 (continued) STATISTICAL APPENDIX 579 TABLE 14. Processed Wood Production (continued) Wood-based Mechanical Sawnwood Panels Paper & Paperboard Wood Asa%of As a% of Growth Growth (thousands of mechan- Growth mechan- rate o/o Volume rate o/o cubic meters) ical wood rate% ical wood 0- (000 MT) 198 1980- 1980 1988 1980 1988 1980-88 1980 1988 88 1980 1988 88 84 El Salvador 37 44 100 100 0.9 0 0 0.0 16 16 0.0 85 Guatemala 103 91 91 91 -4.2 9 9 2.0 32 17 -5.6 86 Guyana 70 57 100 100 1.8 0 0 0.0 0 0 0.0 87 Haiti 14 14 100 100 0.0 0 0 0.0 0 0 0.0 88 Honduras 571 446 98 98 -3.3 2 2 -3.0 0 0 0.0 89 Jamaica 29 48 87 92 5.5 13 8 0.0 10 22 10.4 90 Mexico 2.595 3,212 77 75 3.6 23 25 2.9 1,979 3,375 6.4 91 Nicaragua 416 236 97 94 -8.5 3 6 -1.8 0 0 0.0 92 Panama 67 57 79 79 -2.6 21 21 -1.8 20 20 -5.0 93 Paraguay 724 1,013 90 89 4.4 10 11 5.8 13 11 -4.8 94 Peru 695 601 88 90 --0.9 12 10 -3.7 205 260 -1.6 95 Suriname 102 74 77 86 1.8 23 14 1.4 0 0 0.0 96 Uruguay 115 70 86 82 -3.7 14 18 -2.2 52 70 5.1 97 Venezuela 485 494 72 66 1.4 28 34 2.8 501 708 4.9 NORTH AMERICA 159,462 210,963 81 80 2.3 19 20 2.4 70,229 86,225 2.2 98 Canada 49,126 67,095 90 90 6.2 10 10 5.4 13,390 16,638 3.1 99 United States 110,336 143,868 76 75 4.8 24 25 5.0 56,839 69,587 2.8 EUROPE 124,015 120,597 73 71 1.9 27 29 2.0 50,105 64,331 2.3 100 Albania 212 212 94 94 0.0 6 6 0.0 8 8 0.0 101 Austria 8,199 8,041 82 81 --0.8 18 19 1.6 1,616 2,650 6.4 102 Belgium 2,475 3,180 28 32 2.5 72 68 2.2 864 848 2.0 103 Bulgaria 2,010 1,872 71 72 --0.6 29 28 -1.1 395 476 1.8 104 Czechoslov. 6,081 6,573 81 78 0.6 19 22 2.8 1.185 1,266 0.8 105 Denmark 1,155 1,212 69 71 1.0 31 29 1.1 225 326 4.2 106 Finland 12,022 9,279 85 84 -2.5 15 16 -2.3 5,919 8,652 5.1 107 France 12,859 13,175 76 78 0.8 24 22 -1.8 5,152 6,313 2.1 108 Germafly 21,279 21,809 61 59 0.4 39 41 1.1 8,822 11,916 2.0 109 Greece 767 832 50 49 2.2 50 51 0.8 307 282 --0.5 110 Hungary 1,739 1,564 76 78 --0.4 24 22 -1.1 440 554 2.7 111 Ireland 214 536 67 56 7.9 33 44 29.3 55 29 -4.9 112 Italy 5,271 4,993 52 42 -3.5 48 58 1.3 4,934. 5,513 1.0 113 Netherlands 490 484 69 81 5.3 31 19 -6.0 1,701 2,462 4.9 114 Norway 3,083 3,038 80 79 --0.5 20 21 1.6 1,373 1,670 3.0 115 Poland 9,403 8,143 79 74 -1.5 21 26 1.5 1,277 1,448 2.4 116 Portugal 2,742 2,931 83 70 --0.8 17 30 9.3 463 627 4.1 117 Romania 6,287 4,093 74 67 -6.6 26 33 -3.0 822 819 --0.1 118 Spain 4,030 4,608 53 57 1.8 47 43 0.6 2,566 3,418 3.7 119 Sweden 13,202 12,586 86 90 0.5 14 10 -4.0 6,182 8,161 4.0 120 Switzerland 2,501 2,377 70 70 -1.0 30 30 --0.3 914 1,216 3.9 121 U. Kingdom 2,423 3,189 71 60 1.9 29 40 10.3 3,788 4,296 2.9 122 Yugoslavia 5,570 5,871 76 78 1.3 24 22 --0.3 1,097 1,381 3.0 123 U.S.S.R. 108,818 117,239 92 0 0.7 8 100 4.1 8,733 10,216 2.2 0 Zero or less than half the unit of measure. 580 MANAGING THE WORLD'S FORESTS TABLE 15. Processed Wood Consumption Wood-based Sawnwood Panels Paper & Paperboard Mechanical Asa%of As a % of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousand rate% (thousands m3) ical wood rate o/o ical wood 1980- metric tons) 1980- 1980 1988 1980 1988 198D-88 1980 1988 88 1980 1988 88 WORLD 488,713 625,103 83 80 1.6 17 20 3,0 1,350,508 230,803 4.2 AFRICA 8,372 13,218 87 85 5.0 13 15 6.0 1,793 4,074 7.0 1 Algeria 606 693 76 89 -2.1 241 11 ...f>.7 174 299 7.3 2 Angola 16 12 63 42 -9.5 37 58 3.0 15 16 15.8 3 Benin 11 13 86 89 3.6 141 11 0.0 1 1 -2.6 4 Botswana 6 8 100 100 5.1 0 0 0.0 1 7 55.2 5 Burkina Faso 21 27 93 93 0.1 7 7 6.1 1 5 16.5 6 Burundi 2 4 67 83 20.1 33 17 0.4 0 1 3.0 7 Cameroon 294 605 84 92 10.6 16 8 2.0 19 49 15.7 8 C. Afr. Rep. 37 31 91 91 -3.0 9 9 -5.2 1 0 -24.2 9 Chad 8 2 89 53 -25.5 11 47 0.0 0 0 -3.9 10 Congo 44 66 73 65 --0.7 27 35 7.2 1 2 7.2 11 Cote d'Ivoire 521 483 74 65 -3.5 26 35 5.5 21 38 9.3 12 Egypt 1,123 1,703 85 85 5.6 15 15 7.2 320 1,253 11.2 13 Eq. Guinea 16 48 94 88 8.3 6 13 13.0 0 0 0.0 14 Ethiopia 79 60 82 75 -4.2 18 25 0.7 13 23 10.2 15 Gabon 221 302 41 41 4.0 59 59 5.4 2 2 1.5 16 Ghana 225 346 69 83 8.3 31 17 -1.8 7 7 0.9 17 Guinea 92 93 98 98 0.0 2 2 0.0 0 1 -14.4 18 Guinea-Bissau 10 14 100 100 3.5 () 0 0.0 0 0 0.0 19 Kenya 234 247 73 72 0.3 27 28 3.1 71 138 11.0 20 Liberia 151 397 96 100 10.6 ~I 0 18.1 1 2 16.8 21 Libya 232 188 90 60 -12.7 10 40 6.2 14 22 4.1 22 Madagascar 236 236 99 99 -0.1 1 1 -4.6 6 7 8.0 23 Malawi 67 38 81 82 -'7.4 19 18 -8.4 26 10 -7.4 24 Mali 10 18 90 97 7.3 10 3 -10.2 0 1 4.7 25 Mauritania 0 10 0 95 0.0 0 5 0.0 0 0 0.0 26 Morocco 456 698 78 84 3.8 22 16 0.6 170 183 1.3 27 Mozambique 52 39 88 92 -2.4 12 8 -4.1 20 3 -25.6 28 Namibia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 29 Niger 2 5 79 93 12.9 21 7 0.0 0 2 10.5 30 Nigeria 3,077 2,968 91 91 --0.4 9 9 -4.0 141 154 -4.4 31 Rwanda 6 16 54 80 18.1 46 20 2.3 0 0 --0.8 32 Senegal 23 35 72 86 8.7 28 14 -2.7 10 17 --0.8 33 Sierra Leone 27 10 95 90 -13.8 5 10 ...f>.4 0 1 14.8 34 Somalia 17 18 91 84 -2.3 9 16 5.5 2 0 -20.0 35 South Africa 2,164 2,455 84 84 0.7 16 16 1.0 1,232 1,534 2.2 36 Sudan 27 76 79 90 8.7 21 10 -2.2 20 23 --0.1 37 Tanzania 125 161 95 94 6.2 5 6 9.5 24 50 8.8 38 Togo 5 5 100 100 --0.3 () 0 0 2 11.3 39 Tunisia 305 375 81 69 1.3 19 31 9.1 58 129 12.0 40 Uganda 25 32 94 89 1.0 6 11 12.8 3 3 7.3 41 Zaire 99 147 91 69 0.5 9 31 12.6 8 5 -3.6 42 Zambia 126 108 74 72 -3.7 26 28 -3.0 21 11 ...f>.6 95 85 3.0 ,. 15 17.2 38 80 6.9 43 Zimbabwe 106 214 ·' (continued) STA T/STICAL APPENDIX 581 TABLE 15. Processed Wood Consumption (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a% of As a o/o of Growth Volume Growth Wood mechan- Growth mechan- rate o/o (thousand rate o/o --- (thousands m3) ical wood ,rate% ical wood 1980- metric tons) 1980- 1980 1988 1980 1988 1980-88 1980 1988 88 1980 1988 88 ASIA& PACIFIC 98,250 139,426 88 84 0.0 12 16 6.0 24,724 59,693 9.2 44 Afghanistan 486 486 82 82 o.o 18 18 0.0 1 2 -3.8 45 Australia 5,207 5,992 83 81 0.7 17 19 2.7 2,029 2,290 1.5 46 Bangladesh 198 96 94 87 -11.9 6 13 0.1 55 85 5.0 47 Bhutan 0 5 0 JOO 0.0 JOO 0 0.0 0 0 0.0 48 China 2,144 31,503 77 87 4.6 23 13 18.6 4,092 14,818 J0.8 49 India 11,230 17,892 98 98 6.4 2 2 5.9 1,253 2,177 6.0 50 Indonesia 4,370 7,298 82 97 7.4 18 3 -J0.2 436 883 8.2 51 Iran 681 407 74 65 -8.6 26 35 -1.5 291 218 -4.1 52 Iraq 460 256 90 85 -6.6 JO 15 -1.5 71 112 7.0 53 Israel 349 377 59 60 -1.0 41 40 -1.5 213 372 6.5 54 Japan 52,919 50,570 80 76 -lJ.8 20 24 2.5 17,981 25,164 4.6 55 Kampuchea 45 45 96 96 o.o 4 4 0.0 6 0 -21.5 56 Korea, DPR 280 280 JOO JOO 0.0 0 0 0.0 88 104 0.9 57 Korea, Rep. 3,335 5,696 80 78 6.4 20 22 7.9 1,557 3,458 J0.5 58 LaoPDR 18 19 91 74 -1.6 9 26 9.9 1 0 -5.8 59 Malaysia 3,768 2,552 87 86 -5.0 13 14 -8.4 317 471 5.0 60 Mongolia 475 475 99 99 0.0 1 1 0.0 8 13 5.9 61 Myarunar 326 467 96 97 2.0 4 3 1.5 32 28 1.9 62 Nepal 220 220 JOO JOO o.o 0 0 0.0 2 11 31.7 63 New Zeal 1,672 1,945 88 83 0.8 12 17 6.8 395 622 5.3 64 Pakistan 191 229 69 58 -2.8 31 42 8.2 158 254 8.1 65 P.New Guinea 154 129 92 85 -1.1 8 15 7.5 13 9 -3.9 66 Philippines 1,112 623 71 65 -5.5 29 35 -1.9 442 527 2.7 67 Saudi Arabia 1,501 984 75 65 -9.J 25 35 -4.3 J05 112 -lJ.l 68 Sri Lanka 71 46 50 61 -4.7 50 39 -4.2 57 59 -1.1 69 Syria 309 358 84 83 5.3 16 17 -2.2 82 91 5.1 70 Thailand 2,026 1,763 93 89 -1.1 7 11 0.1 457 738 5.8 71 Turkey 5,091 5,706 91 86 1.5 9 14 9.6 577 430 -3.3 72 Viet Nam 550 394 99 9a -3.8 1 JO 16.7 53 64 2.2 73 Yemen 17 31 94 94 5.8 6 6 15.1 0 a a.a CENTRAL& SOUTH AMERICA 17,556 35,271 85 86 3.5 15 14 5.a 6,276 15,669 5.1 74 Argentina 1,963 2,a83 78 84 2.2 22 16 -2.2 945 1,050 2.7 75 Belize 8 14 63 57 8.6 38 43 2.3 1 2 5.3 76 Bolivia 142 45 91 85 -14.9 9 15 a.a 18 13 -5.4 77 Brazil 16,780 19,99a 87 88 2.4 13 12 0.7 3,425 3,751 3.5 78 Chile 1,081 2,a14 92 90 11.2 8 JO 12.9 323 356 1.3 79 Colombia 1,a92 830 88 86 -3.5 12 14 -2.9 495 632 2.3 80 Costa Rica 57a 553 92 93 0.4 8 7 -().1 86 57 -8.1 81 Cuba 730 840 90 71 1.7 JO 29 16.1 231 313 5.2 82 Domin.Rep. 115 39 94 69 -16.8 6 31 11.0 82 7a -2.3 (continued) 582 MANAGING THE WORLD'S FORESTS TABLE 15. Processed Wood Consumption (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a o/o of As a % of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousand rate% (thousands m3) ical wood rate% ical wood 1980- metric tons) 1980-- 1980 1988 1980 1988 1980--88 1980 1988 88 1980 1988 88 83 Ecuador 914 1,411 98 90 4.6 2 10 26.8 154 229 4.6 84 El Salvador 45 56 92 96 1.6 8 4 -14.5 59 53 -1.6 85 Guatemala 80 63 83 86 -<>.7 17 14 -3.8 106 85 -1.5 86 Guyana 28 48 100 100 5.5 0 0 0.0 2 1 -2.9 87 Haiti 24 15 100 100 -8.0 0 0 0.0 6 7 2.6 88 Honduras 301 276 97 96 1.6 3 4 7.1 43 51 1.1 89 Jamaica 68 163 86 85 3.7 141 15 -1.3 41 74 4.5 90 Mexico 3,180 3,920 79 79 3.4 21 21 2.3 2,468 3,509 3.5 91 Nicaragua 407 224 98 95 -8.8 2 5 -1.5 13 13 92 Panama 82 59 81 79 -4.1 19 21 -2.6 93 75 -1.0 93 Paraguay 352 864 105 92 9.5 -o ,. 8 0.0 29 28 -1.6 94 Peru 690 599 89 90 -1.0 11 10 -1.9 223 350 1.4 95 Suriname 11 67 63 91 -1.2 37 9 5.4 9 9 0.2 96 Uruguay 143 90 87 86 -4.3 F·' 14 -2.4 44 69 2.3 97 Venezuela 445 567 81 67 0.6 19 33 6.6 536 955 4.0 NORTH AMERICA 148,186 191,549 79 79 3.5 21 21 2.6 51,900 82,906 1.7 98 Canada 20,303 25,076 82 81 6.6 18 19 5.3 4,094 6,058 4.3 99 United States 128,597 168,672 79 77 5.2 21 23 5.3 48,460 76,848 3.7 EUROPE 114,010 137,527 74 70 1.1 26 30 1.9 38,248 64,656 3.0 100 Albania 210 210 95 95 0.0 5 5 0.0 13 12 -0.4 101 Austria 3,702 3,996 79 80 0.1 21 20 0.5 944 1,086 3.3 102 Belgium 2,339 3,391 64 68 2.8 36 32 1.9 987 1,906 5.0 103 Bulgaria 2,044 1,987 76 75 -0.1 24 25 -0.3 518 563 1.0 104 Czechoslov. 5,121 5,507 76 73 0.1 24 27 2.8 1,089 1,225 2.0 105 Denmark 2,519 2,903 76 74 4.1 24 26 5.1 728 990 4.3 106 Finland 4,094 3,521 82 80 -2.0 18 20 -0.2 1,105 1,467 2.9 107 France 15,785 15,295 78 78 -0.2 22 22 -0.3 6,197 8,333 3.3 108 Germany 23,364 26,681 66 63 0.8 341 37 1.4 8,355 13,734 3.9 109 Greece 1,290 1,467 71 63 2.1 29 37 5.5 423 500 2.3 110 Hungary 2,577 2,509 81 81 0.5 19 19 0.1 676 739 1.0 111 Ireland 762 886 77 59 -2.1 23 41 13.0 254 272 1.2 112 Italy 11,590 10,979 73 69 -1.1 27 31 0.9 5,294 6,434 2.5 113 Netherlands 4,537 4,630 73 69 1.8 27 31 2.6 2,183 2,773 3.2 114 Norway 3,122 3,365 79 79 1.5 21 21 1.9 489 616 2.8 115 Poland 9,011 7,486 76 73 -1.6 241 27 0.1 1,433 1,533 1.6 116 Portugal 1,690 1,421 82 78 -4.1 18 22 -1.4 396 621 6.1 117 Romania 4,845 3,298 76 65 -5.7 24 35 -0.7 694 715 0.3 118 Spain 4,570 6,103 69 66 3.0 31 34 4.7 2,700 3,921 4.8 119 Sweden 7,198 6,024 79 78 -1.8 21 22 0.1 1,738 2,016 3.4 120 Switzerland 2,819 2,875 76 78 0.4 24 22 -1.0 1,036 1,314 3.6 121 U. Kingdom 11,308 17,156 73 69 4.5 27 31 6.7 6,831 9,272 3.9 122 Yugoslavia 3,872 4,838 76 76 2.6 24 24 2.1 770 1,181 3.6 123 U.S.S.R. 116,896 108,112 93 88 -1.4 7 12 3.4 8,249 9,519 1.3 0 Zero or less than half the unit of measure. STATISTICAL APPENDIX 583 TABLE 16. Processed Wood Exports .Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a% of As a o/o of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousand rate% --- (thousands m.3) ical wood rate% ical wood 1980- metric tons) 1980- --- 1980 1988 1980 1988 1980-88 1980 1988 88 1980 1988 88 WORLD 95,948 125,029 . 83 80 23 17 20 2.6 35,041 51,119 2.5 AFRICA 1,234 1,236 73 76 2.0 27 24 1.9 158 258 2.6 1 Algeria 0 0 0 0 0.0 0 0 0.0 0 0 0.0 2 Angola 0 0 0 0 0.0 0 0 0.0 0 0 0.0 3 Benin 0 0 0 0 0.0 0 0 0.0 0 0 0.0 4 Botswana 0 0 0 0 0.0 0 0 0.0 0 0 0.0 5 Burkina Faso 0 0 0 0 0.0 0 0 0.0 0 0 0.0 6 Burundi 0 0 0 0 0.0 0 0 0.0 0 0 0.0 7 Cameroon 191 115 85 81 -6.5 15 19 -8.0 0 0 0.0 8 C. Afr. Rep. 40 25 91 96 -5.0 9 4 -14.0 0 0 0.0 9 Chad 0 0 0 0 0.0 0 0 0.0 0 0 0.0 10 Congo 104 62 36 39 -3.1 64 61 -6.4 0 0 0.0 11 Cote d'Ivoire 339 558 82 82 8.3 18 18 8.2 0 0 0.0 12 Egypt 0 0 0 0 0.0 0 0 0.0 0 0 0.0 13 Eq. Guinea 0 13 0 69 0.0 0 31 0.0 0 0 0.0 14 Ethiopia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 15 Gabon 79 53 23 2 -23.9 77 98 -1.7 0 0 0.0 16 Ghana 78 192 89 88 17.5 11 12 17.5 0 0 0.0 17 Guinea 0 0 0 0 0.0 0 0 0.0 0 0 0.0 18 Guinea-Bissau 6 2 100 100 0.0 0 0 0.0 0 0 0.0 19 Kenya 11 3 96 100 -9.7 4 0 0.0 19 1 -37.5 20 Liberia 49 20 92 73 -17.8 8 27 -20.6 21 Libya 0 0 0 0 0.0 0 0 0.0 0 0 0.0 22 Madagascar 0 0 0 0 0.0 0 0 0.0 0 0 0.0 23 Malawi 0 0 0 0 0.0 0 0 0.0 0 0 0.0 24 Mali 0 0 0 0 0.0 0 0 0.0 0 0 0.0 25 Mauritania 0 0 0 0 0.0 0 0 0.0 0 0 0.0 26 Morocco 5 .3 0 0 0.0 100 100 -3.8 0 0 0.0 27 Mozambique 20 2 100 6 -53.l 0 94 -2.0 0 0 0.0 28 Namibia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 29 Niger 0 0 0 0 0.0 0 0 0.0 0 0 0.0 30 Nigeria 2 1 100 100 -11.0 0 0 0.0 0 0 0.0 31 Rwanda 0 0 0 0 0.0 0 0 0.0 0 0 0.0 32 Senegal 0 0 0 0 0.0 0 0 0.0 0 0 0.0 33 Sierra Leone 0 3 0 100 57.4 0 0 0.0 0 0 0.0 34 Somalia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 35 South Africa 134 58 55 43 -10.7 45 57 -4.5 138 253 10.2 36 Sudan 0 0 0 0 0.0 0 0 0.0 0 0 0.0 37 Tanzania 3 5 90 91 19.4 10 9 --0.6 0 0 0.0 38 Togo 0 0 0 0 0.0 0 0 0.0 0 0 0.0 39 Tunisia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 40 Uganda 0 0 0 0 0.0 0 0 0.0 0 0 0.0 41 Zaire 50 28 48 72 1.3 . 52 28 -9.4 0 0 0.0 42 Zambia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 43 Zimbabwe 23 10 91 81 -5.9 9 19 4.4 2 4 27.9 (continued) 584 Mil.NAG/NG THE WORLD'S FORESTS TABLE 16, Processed Wood Exports (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a% of As a % of Growth Volume Growth Wood mechan- Growth rnechan- rate o/o (thousand rate% (thousands m3) ical wood rate % ical wood 1980- metric tons) 1980- 1980 1988 1980 1988 1980-88 1980 1988 88 1980 1988 88 ASIA& PACIFIC 12,160 20,035 66 50 2.3 34 50 3.4 1,655 2,624 2.6 44 Afghanistan 0 0 0 0 0.0 0 0 0.0 0 0 0.0 45 Australia 105 45 78 60 -15.6 22 40 -1.3 95 121 2.5 46 Bangladesh 0 0 0 0 0.0 () 0 0.0 19 18 4.5 47 Bhutan 6 0 100 100 -35.8 0 0 0.0 0 0 0.0 48 China 929 12 5 70 16.4 9'' ,) 88 -27.2 149 564 126 49 India 15 19 12 12 11.8 88 88 3.5 4 1 -7.5 50 Indonesia 1,459 9,509 83 32 13.4 17 68 45.3 7 199 61.3 51 Iran 0 0 0 0 0.0 0 0 o.o 0 0 0.0 52 Iraq 0 0 0 0 0.0 0 0 o.o 0 0 0.0 53 Israel 41 21 0 0 0.0 100 100 -4.7 7 3 -8.8 54 japan 205 129 28 9 -18.3 72 91 -2.5 763 594 -0.7 55 Kampuchea 1 1 0 0 0.0 100 100 o.o 0 0 0.0 56 Korea,DPR 0 0 0 0 0.0 0 0 0.0 0 0 0.0 57 Korea, Rep. 1,342 359 28 72 -1.4 72 28 -26.2 154 376 14.0 58 LaoPDR 25 2 100 100 -22.8 0 0 0.0 0 0 0.0 59 Malaysia 3,924 5,188 85 79 2.5 15 21 8.2 5 0 0.0 60 Mongolia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 61 Myanmar 118 31 100 100 -15.6 0 0 0.0 0 0 0.0 62 Nepal 0 0 0 0 0.0 0 0 o.o 0 0 0.0 63 New Zealand 740 561 84 73 -5.3 16 27 1.9 323 209 -5.3 64 Pakistan 0 0 0 0 0.0 0 0 0.0 0 0 0.0 65 PapuaN. G. 53 7 85 97 -20.2 15 3 -43.5 0 0 0.0 66 Philippines 1,177 950 63 66 -1.5 37 34 -3.8 1 1 5.3 67 Saudi Arabia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 68 Sri Lanka 1 0 100 0 0.0 0 0 0.0 0 0 0.0 69 Syria 19 0 100 0 0.0 0 0 o.o 1 0 0.0 70 Thailand 17 149 9 62 89.1 91 38 24.3 15 58 28.3 71 Turkey 7 73 100 94 20.5 0 6 4.8 1 42 49.8 72 Viet Nam 0 0 0 0 0.0 0 0 0.0 0 0 0.0 73 Yemen 0 0 0 0 0.0 0 0 o.o 0 0 0.0 CENTRAL& SOUTH AMERICA 3,483 2,890 82 70 1.7 18 30 2.4 397 1,326 4.3 74 Argentina 23 31 5 0 28.9 95 100 1.2 15 10 -1.4 75 Belize 5 6 100 100 -2.2 0 0 o.o 0 0 0.0 76 Bolivia 95 58 96 98 -1.6 4 2 -16.9 0 0 77 Brazil 1,133 1,314 71 50 -3.5 29 50 7.7 198 1,104 19.7 78 Chile 1,331 967 98 94 -0.6 2 6 10.5 87 135 10.1 79 Colombia 13 15 85 39 -10.6 15 61 18.0 26 24 1.9 80 Costa Rica 26 22 0 12 14.3 100 88 -4.7 16 17 -4.1 81 Cuba 0 0 0 0 0.0 0 0 0.0 0 0 0.0 82 Dominican Rep. 0 0 0 0 0.0 0 0 o.o 0 0 0.0 83 Ecuador 78 39 10 39 -0.4 90 61 -11.7 0 0 0.0 (continued) STATISTICAL APPENDIX 585 TABLE 16. Processed Wood Exports (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical Asa%of As a% of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousand rate% (thousands m3) ical wood rate % ical wood 1980- metric tons) 198Cl- 1980 1988 1980 1988 198Cl-88 1980 1988 88 1980 1988 88 84 El Salvador 0 0 0 0 0.0 0 0 0.0 2 2 -4.2 85 Guatemala 28 29 97 100 4.1 3 0 0.0 18 5 -16.5 86 Guyana 14 9 100 100 -1.6 0 0 0.0 0 0 0.0 87 Haiti 0 0 0 0 0.0 0 0 0.0 0 0 0.0 88 Honduras 273 174 98 100 -7.0 2 0 0.0 13 0 0.0 89 Jamaica 0 0 0 0 0.0 0 0 0.0 1 0 -27.5 90 Mexico 24 40 2 62 ·. 78.2 98 38 13.7 9 19 9.2 91 Nicaragua 10 12 50 73 4.4 50 27 -2.7 0 0 0.0 92 Panama 1 1 0 0 0.0 100 100 -4.2 2 0 -22.2 93 Paraguay 372 148 77 75 -6.9 23 25 -8.0 0 0 0.0 94 Peru 23 3 71 93 -22.9 29 7 -40.5 2 4 16.0 95 Suriname 42 7 57 43 -16.2 43 57 0.0 0 0 0.0 96 Uruguay 0 0 0 0 0.0 0 0 0.0 9 8 1.5 97 Venezuela 0 0 0 0 0.0 0 0 0.0 0 0 0.0 NORTH AMERICA 37,472 57,272 94 92 2.5 6 8 2.9 13,675 16,109 2.2 98 Canada 30,670 44,439 96 95 5.9 4 5 6.5 9,555 11,514 3.0 99 United States 6,803 12,833 86 83 6.0 14 17 7.9 4,120 4,595 1.8 EUROPE 33,400 34,302 76 74 2.0 24 26 2.1 18,120 29,550 2.6 100 Albania 2 2 0 0 0.0 100 100 0.0 1 0 0.0 101 Austria 5,146 4,922 85 81 -1.0 15 19 4.3 886 1,984 9.4 102 Belgium 1,589 2,003 14 17 -2.5 86 83 2.5 0 0 0.0 103 Bulgaria 142 60 39 43 -8.3 61 57 -8.0 6 13 16.2 104 1,185 Czechoslov. 1,217 95 97 1.7 5 3 -5.9 234 143 -7.9 105 Denmark 307 392 68 68 -1.3 32 32 -1.5 121 152 2.4 106 Finland 7,980 5,863 87 86 -26 13 14 -3.0 4,868 7,309 5.6 107 France 1,496 1,717 53 60 3.9 47 40 --0.8 1,064 1,850 7.2 108 Germany 1,712 3,218 1 1 12.1 0 0 8.3 1,862 4,310 7.8 109 Greece 32 55 11 6 -4.1 89 94 8.2 44 24 -8.1 110 Hungary 105 142 71 71 5.9 29 29 6.0 78 90 1.8 111 Ireland 27 113 78 82 15.8 22 18 19.5 22 9 -11.3 112 Italy 246 364 33 31 7.1 67 69 6.7 501 952 8.9 113 Netherlands 347 593 61 75 9.2 39 25 4.0 953 1,822 8.9 114 Norway 554 522 78 80 --0.9 22 20 0.0 1,048 1,315 3.6 115 Poland 959 1,087 74 67 -1.0 26 33 8.3 30 129 20.9 116 Portugal 1,071 1,563 84 64 3.4 16 36 22.4 144 164 2.2 117 Romania 1,445 962 69 82 -6.2 31 18 -13.1 178 144 -2.6 118 Spain 688 428 32 49 -1.3 68 51 -9.2 160 426 10.1 119 Sweden 6,504 7,219 91 94 2.4 9 6 -4.7 4,626 6,405 4.3 120 Switzerland 427 546 35 19 -1.7 65 81 7.7 239 492 10.0 121 U. Kingdom 193 152 26 26 -3.2 74 74 -5.7 466 798 9.7 122 Yugoslavia 1,242 1,162 89 86 0.5 11 14 4.8 109 300 15.5 123 U.S.S.R. 8,198 9,294 88 88 1.8 12 12 2.2 1,035 1,252 3.1 . . Data not available. 0 Zero or less than half the unit of measure. 586 MANAGING THE WORLD'S FORESTS TABLE 17. Processed Wood Imports Wood-based Sawnwood Panels Paper & Paperboard Mechanical Asa% of As a o/o of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousands rate% --- (thousands m3) ical wood rate% ical wood 1981}- of MT) 1981}- --- 1980 1988 1980 1988 1981}-88 1980 1988 88 1980 1988 88 WORLD 92,798 119,087 83 79 2.2 17 21 2.6 33,601 50,475 2.5 AFRICA 3,104 3,658 83 88 2.3 17 12 1.8 927 1,899 3.0 1 Algeria 544 630 83 96 -2.1 17 4 -13.1 87 179 10.6 2 Angola 0 0 0 0 0.0 0 0 0.0 2 1 -13.3 3 Benin 2 2 12 35 23.9 88 65 0.0 1 1 -2.6 4 Botswana 6 8 100 100 5.1 0 0 0.0 1 7 55.2 5 Burkina Faso 19 26 93 93 1.2 7 7 6.1 1 5 16.5 6 Burundi 1 1 0 0 0.0 100 100 0.4 0 1 3.0 7 Cameroon 1 0 0 0 0.0 100 0 0.0 14 44 20.2 8 C. Afr. Rep. 0 0 0 0 0.0 0 0 0.0 1 0 -24.2 9 Chad 7 1 87 0 13 100 0.0 0 0 0.0 10 Congo 6 10 100 100 -4.0 0 0 0.0 1 2 7.2 11 Cote d'Ivoire 0 .0 0 0 0.0 0 0 0.0 21 38 9.3 12 Egypt 1,082 1,660 88 87 5.6 12 13 8.5 198 1,093 13.6 13 Eg.Guinea 0 0 0 0 0.0 0 0 0.0 0 0 0.0 14 Ethiopia 0 0 0 0 0.0 0 0 0.0 5 13 16.6 15 Gabon 0 0 0 0 0.0 0 0 0.0 2 2 1.5 16 Ghana 0 0 0 0 0.0 0 0 0.0 7 7 0.9 17 Guinea 0 1 0 100 -5.3 0 0 0.0 0 1 -14.4 18 Guinea-Bissau 0 0 0 0 0.0 0 0 0.0 0 0 0.0 19 Kenya 3 0 0 0 0.0 100 100 -9.6 29 33 4.0 20 Liberia 47 0 98 0 -38.6 -· '· 100 -13.1 1 2 16.8 21 Libya 201 157 88 52 -15.4 ]"' 48 6.2 9 16 4.4 22 Madagascar 1 0 0 100 -37.7 '· 100 0 0.0 2 1 -9.3 23 Malawi 11 1 62 0 -31.4 38 100 -19.8 26 10 -7.4 24 Mali 4 5 74 90 7.5 26 10 -10.2 0 1 4.7 25 Mauritania 0 10 0 95 0.0 0 5 0.0 0 0 0.0 26 Morocco 248 447 91 98 4.5 9 2 -10.7 74 74 0.2 27 Mozambique 4 0 0 0 0.0 100 0 0.0 17 1 -41.5 28 Namibia 0 0 0 0 0.0 0 0 0.0 0 0 0.0 29 Niger 2 5 79 93 12.9 21 7 0.0 0 2 10.5 30 Nigeria 169 23 0 0 -37.7 100 100 -26.9 123 59 -15.3 31 Rwanda 1 2 0 0 0.0 100 100 2.2 0 0 -0.8 32 Senegal 12 24 46 80 18.4 54 20 -2.7 10 17 -0.8 33 Sierra Leone 3 1 52 0 0.0 48 100 -6.4 0 1 14.8 34 Somalia 2 3 93 46 -10.9 7 54 23.4 2 0 -20.0 35 South Africa 326 242 93 91 -4.4 7 9 -0.7 172 151 -0.4 36 Sudan 19 62 78 90 9.3 z-· 10 '· -2.3 11 13 -0.8 37 Tanzania 0 0 0 0 0.0 0 0 0.0 24 22 -1.3 38 Togo 0 0 0 0 0.0 0 0 0.0 0 2 11.3 39 Tunisia 262 268 93 96 1.4 7 4 -6.5 33 59 8.7 40 Uganda 1 0 66 0 0.0 34. 100 0.0 3 1 -9.4 41 Zaire 2 1 0 0 0.0 100 100 -3.4 6 4 -0.9 42 Zambia 27 5 51 31 -24.5 49 69 -14.2 19 9 -4.9 43 Zimbabwe 13 1 73 40 -28.9 27 60 -22.8 19 2 -17.2 (continued) STATISTICAL APPENDIX 587 TABLE 17. Processed Wood Imports (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a% of As a % of Growth Volume Growth Wood mechan- Growth mechan- rate o/o (thousands rate% (thousands m3) ical wood rate% ical wood 1980- of MT) 1980- 1980 1988 1980 1988 1980-88 1980 1988 88 1980 1988 88 ASIA& PACIFIC 13,960 22,089 85 73 2.4 15 27 3.8 5 8 2.6 44 Afghanistan 85 85 0 0 0.0 100 100 0.0 1 2 -3.8 45 Australia 1,076 1,621 92 94 4.2 8 6 2.0 695 710 0.6 46 Bangladesh 10 5 100 100 10.2 0 0 0.0 0 7 0.0 47 Bhutan 0 0 0 0 0.0 0 0 0.0 0 0 0.0 48 China 189 2,180 74 55 11.2 26 45 10.5 399 1,614 12.5 49 India 17 10 99 80 -9.8 1 20 39.9 295 238 -5.2 50 Indonesia 2 0 4 0 0.0 96 0 -34.8 212 108 -8.9 51 Iran 430 190 80 54 -15.9 20 46 0.0 213 140 -6.0 52 Iraq 450 247 90 85 -6.7 10 15 -1.6 43 84 10.4 53 Israel 252 250 81 90 -1.0 19 10 -11.3 111 204 6.6 54 japan 5,889 11,192 95 76 6.7 5 24 38.7 656 1,133 6.0 55 Kampuchea 0 0 0 0 0.0 0 0 0.0 6 0 -21.5 56 Korea, DPR 0 0 0 0 0.0 0 0 0.0 8 24 6.8 57 Korea, Rep. 0 707 0 100 41.5 0 0 0.0 32 175 26.4 58 LaoPDR 0 0 0 0 0.0 0 0 0.0 1 0 -5.8 59 Malaysia 241 34 88 73 -21.9 12 27 -17.6 251 401 5.6 60 Mongolia 1 1 0 0 0.0 100 100 0.0 8 13 5.9 61 Myanmar 0 0 0 0 0.0 0 0 0.0 22 20 1.8 62 Nepal 0 0 0 0 0.0 0 0 0.0 0 9 67.0 63 New Zealand 30 61 83 86 11.4 17 14 4.9 44 131 20.0 64 Pakistan 88 89 87 87 -4.3 13 13 6.3 95 173 11.1 65 P. New Guinea 0 0 0 0 0.0 0 0 0.0 13 9 -3.9 66 Philippines 0 0 100 0 0.0 0 0 0.0 120 193 7.4 67 Saudi Arabia 1,501 984 75 65 -9.1 25 35 -4.3 105 112 --0.1 68 Sri Lanka 32 16 33 48 -6.0 67 52 2.4 39 31 -4.6 69 Syria 292 322 92 89 4.4 8 11 -2.9 80 72 3.1 70 Thailand 350 617 98 99 1.4 2 1 -2.6 134 282 7.9 71 Turkey 1 75 8 100 160.6 92 0 2.5 100 72 0.0 72 Viet Nam 31 1 99 50 -31.9 1 50 0.0 5 9 4.8 73 Yemen 0 0 0 0 0.0 0 0 0.0 0 0 0.0 CENTRAL& SOUTH AMERICA 3,744 2,469 87 84 1.6 13 16 1.8 2,395 1,817 1.8 74 Argentina 689 310 93 98 -6.2 7 2 -21.4 247 85 -12.6 75 Belize 3 6 33 0 1.0 67 100 4.0 2 3 2.5 76 Bolivia 0 0 0 0 0.0 0 0 0.0 17 11 -7.1 77 Brazil 549 139 84 81 -8.3 16 19 -10.6 261 216 --0.1 78 Chile 0 0 0 0 0.0 0 0 0.0 54 43 -3.4 79 Colombia 25 11 26 16 -11.7 74 84 -23.3 170 154 -2.8 80 Costa Rica 3 2 10 0 0.0 90 100 --0.6 89 57 -8.9 81 Cuba 612 585 89 82 1.9 11 18 5.9 158 172 2.4 82 Dominican Rep. 115 39 94 69 -16.8 6 31 11.0 73 61 -2.8 (continued) 588 MANAGING THE WORLD'S FORESTS TABLE 17. Processed Wood Imports (continued) Wood-based Sawnwood Panels Paper & Paperboard Mechanical As a% of As a % of Growth Volume Growth Wood mechan- Growth mechan- rate% (thousands rate% (thousands m3) ical wood rate % ical wood 1980- of MT) 1980- 1980 1988 1980 1988 198Q-88 1980 1988 88 1980 1988 88 83 Ecuador 0 0 0 0 0.0 0 0 0.0 128 175 3.7 84 El Salvador 8 12 55 80 7.4 45 20 -14.5 45 38 -2.5 85 Guatemala 6 I 12 0 0.0 88 100 -27.2 92 73 -2.1 86 Guyana 2 2 0 0 0.0 100 100 2.0 0 0 0.0 87 Haiti 11 1 100 100 -32.6 0 0 0.0 6 7 2.6 88 Honduras 3 3 0 0 0.0 100 100 0.0 56 51 -1.8 89 Jamaica 40 115 85 82 3.4 15 18 -1.5 32 52 1.6 90 Mexico 609 748 88 94 3.0 12 6 -3.9 498 153 -17.2 91 Nicaragua 1 0 0 0 0.0 100 100 -17.1 13 13 --0.3 92 Panama 16 3 83 60 -18.2 17 40 -8.8 75 56 1.0 93 Paraguay 0 0 0 0 0.0 0 0 0.0 16 16 0.9 94 Peru 18 2 100 100 -28.2 0 0 0.0 19 94 16.0 95 Suriname 0 0 0 0 0.0 0 0 0.0 13 9 1.7 96 Uruguay 0 20 95 98 -12.2 5 2 -17.5 23 7 -16.7 97 Venezuela 337 73 78 73 -21.0 22 27 -19.0 220 196 -2.3 NORTH AMERICA 26,909 40,057 91 88 2.4 9 12 3.0 8,118 12,455 2.5 98 Canada 1,846 2,420 87 76 1.0 13 24 10.3 260 933 IS.I 99 United States 25,063 37,637 91 89 6.8 9 11 9.4 7,858 11,522 6.8 EUROPE 44,538 50,389 78 73 2.1 22 27 2.4 16,246 25,808 2.6 100 Albania 0 0 0 0 0.0 0 0 0.0 6 4 0.0 101 Austria 649 877 90 79 2.2 10 21 13.6 214 420 8.5 102 Belgium 2,031 2,239 79 73 2.0 21 27 2.4 975 1,503 2.5 103 Bulgaria 175 175 96 97 2.2 4 3 -3.3 128 100 -I.I 104 Czechoslov. 225 150 48 43 -5.3 52 57 -2.3 138 102 -2.8 105 Denmark 1,672 2,083 79 74 5.6 2:1 26 6.5 625 816 4.0 106 Finland 52 105 61 43 5.4 39 57 19.4 53 124 12.1 107 France 4,422 3,837 78 69 -2.8 22 31 3.5 2,109 3,870 7.9 108 Germany 9,142 8,090 75 70 1.3 25 30 3.5 4,057 5,718 3.5 109 Greece 555 690 98 76 2.0 2 24 55.0 160 242 4.6 110 Hungary 942 1,088 90 84 2.5 10 16 5.5 314 275 -1.5 111 Ireland 575 463 81 68 -5.4 19 32 2.7 222 252 1.5 112 Italy 6,565 6,350 88 88 0.0 12 12 1.4 861 1,873 12.4 113 Netherlands 4,394 4,739 72 69 2.1 28 31 3.6 1,435 2,133 5.4 114 Norway 593 849 72 80 8.2 28 20 1.6 165 261 6.0 115 Poland 566 430 33 34 -3.6 67 66 -3.0 186 214 1.7 116 Portugal 20 53 93 89 5.0 7 11 27.0 77 159 10.7 117 Romania 4 167 74 100 46.9 26 0 -7.8 50 40 -3.5 118 Spain 1,228 1,923 99 84 4.9 I 16 58.3 294 929 13.6 119 Sweden 501 657 67 42 0.3 33 58 16.9 181 260 5.3 120 Switzerland 745 1,044 73 66 3.9 27 34 7.4 361 590 7.5 121 U. Kingdom 9,078 14,118 73 70 5.1 27 30 5.2 3,510 5,774 5.6 122 Yugoslavia 267 129 89 77 -7.9 11 23 8.7 87 100 1.8 123 U.S.S.R. 543 324 66 67 -6.7 34 33 -4.0 905 555 -7.2 .. Data not available. 0 Zero or less than hall the unit of measure. STATISTICAL APPENDIX 589 TABLE 18. Basic Economic Indicators GNP GDP Distribution of Gross Domestic Product (percent) per Capita Growth 1989 Rate Agriculture Industry Service etc. (US$) 1980-89 1980 1989 1980 1989 1980 1989 WORLD AFRICA 1 Algeria 2,221 3.5 10 16 54 45 36 40 2 Angola 611 3 Benin 382 1.8 48 46 12 12 40 42 4 Botswana 1,603 11.4 12 3 41 57 47 40 5 Burkina Faso 320 5.0 29 36 25 26 46 39 6 Burundi 217 4.3 63 56 13 25 7 Cameroon 996 5.4 28 27 26 27 46 46 8 Cent. Afr. Rep. 388 1.4 40 42 20 15 40 43 9 Chad 188 5.8 53 36 12 20 35 44 10 Congo 942 4.1 12 14 47 36 42 51 11 Cote d'Ivoire 794 1.2 33 46 20 24 47 30 12 Egypt 634 5.4 18 19 37 30 45 52 13 Ethiopia 123 1.9 51 42 16 16 34 42 14 Eq.Guinea 430 59 15 Gabon 2;963 --0.2 7 10 62 47 31 43 16 Ghana 386 2.1 58 49 12 17 30 34 17 Guinea 428 30 33 38 18 Guinea-Bissau 198 47 19 Kenya 368 4.1 33 31 21 20 47 49 20 Liberia 451 -1.3 36 28 36 21 Libya 5,453 2 76 22 22 Madagascar 228 0.8 30 31 16 14 54 54 23 Malawi 177 2.7 37 35 19 19 44 45 24 Mali 267 5.3 58 50 9 12 32 38 25 Mauritania 495 1.4 30 37 26 24 44 39 26 Morocco 876 4.2 18 16 31 34 51 50 27 Mozambique 81 -1.4 54 64 27 22 19 14 28 Namibia 29 Niger 291 -2.5 43 36 23 13 35 51 30 Nigeria 249 --0.4 26 31 42 44 32 25 31 Rwanda 317 1.1 46 37 22 23 33 41 32 Senegal 654 3.3 19 22 25 31 57 47 33 Sierra Leone 220 0.6 33 46 22 12 45 42 34 Somalia 290 3.0 68 65 8 10 24 26 35 South Africa 1,870 1.5 7 6 51 44 42 50 36 Sudan 471 2.2 34 36 14 15 53 49 37 Tanzania 120 2.6 44 66 17 39 38 Togo 394 1.7 28 33 25 23 48 44 39 Tunisia 1,260 3.4 16 14 36 33 48 53 40 Uganda 260 2.5 72 67 5 7 24 26 41 Zaire 257 1.9 28 30 33 32 39 38 42 Zambia 391 0.7 14 13 41 47 45 40 43 Zimbabwe 652 2.7 14 13 39 39 47 49 (continued) 590 MANAGING THE WORLD'S FORESTS TABLE 18. Basic Economic Indicators (continued) GNP GDP Distribution of Gross Domestic Product (percent) per Capita Growth 1989 Rate Agriculture Industry Service etc. (US$) 1980-89 1980 1989 1980 1989 1980 1989 ASIA & PACIFIC 44 Afghanistan 220 45 Australia 14,443 3.3 5 4 36 58 46 Bangladesh 178 3.5 50 44 15 14 35 41 47 Bhutan 150 8.1 57 0 12 0 31 0 48 China 357 10.3 36 32 49 15 49 India 344 5.3 38 30 26 29 36 41 50 Indonesia 503 5.1 24 23 42 37 34 39 51 Iran 7,974 3.4 17 23 33 15 50 62 52 Iraq 2,400 53 Israel 9,753 3.2 54 japan 23,734 3.6 4 3 42 54 55 Kampuchea 56 Korea,DPR 910 57 Korea, Rep. 4,400 9.8 15 10 41 44 44 46 58 LaoPDR 177 59 Malaysia 2,160 4.6 22 38 40 60 Mongolia 61 Myanmar 212 47 59 13 11 41 30 62 Nepal 179 4.7 62 58 12 14 26 28 63 New Zealand 11,797 2.2 11 10 31 58 64 Pakistan 365 6.4 30 27 25 24 46 49 65 PapuaN.G. 894 1.8 33 28 27 30 40 42 66 Philippines 700 0.8 23 24 37 33 40 43 67 Saudi Arabia 6,394 -3.3 1 8 78 45 21 48 68 Sri Lanka 433 4.0 28 26 30 27 43 47 69 Syria 1,645 0.5 20 38 23 57 70 Thailand 1,229 7.0 23 15 31 38 46 47 71 Turkey 1,369 5.1 23 17 30 35 47 48 72 Viet Nam 200 3.8 73 Yemen 625 24 20 17 26 59 54 CENTRAL& SOUTH AMERICA 74 Argentina 2,157 -0.3 9 14 37 33 55 54 75 Belize 1,600 19 76 Bolivia 616 -0.8 18 32 35 30 47 38 77 Brazil 2,538 3.0 11 9 44 45 78 Chile 1,765 1.9 7 37 56 79 Colombia 1,194 2.9 19 17 32 36 49 47 80 Costa Rica 1,778 1.5 18 17 27 27 55 56 81 Cuba 82 Dominican Rep. 787 2.2 20 15 28 26 52 59 83 Ecuador 1,021 1.4 12 15 38 39 50 47 84 El Salvador 1,070 0.0 28 12 21 21 52 67 85 Guatemala 913 -0.2 18 26 56 86 Guyana 310 25 (continued) STATISTICAL APPENDIX 591 TABLE 18. Basic Economic Indicators (continued) GNP GDP Distribution of Gross Domestic Product (percent) per Capita Growth Agriculture Industry Service etc. 1989 Rate (US$) 1980-89 1980 1989 1980 1989 1980 1989 87 Haiti 359 --0.2 31 88 Honduras 903 2.3 25 21 25 25 50 54 89 Jamaica 1,257 --0.8 8 6 38 54 90 Mexico 1,994 0.6 8 9 33 32 59 59 91 Nicaragua -2.8 23 31 45 92 Panama 1,764 --0.5 9 11 21 15 70 75 93 Paraguay 1,033 1.9 29 30 27 22 44 48 94 Peru 1,007 1.2 10 8 42 31 48 62 95 Suriname 3020 11 96 Uruguay 2,623 0.1 11 11 33 28 56 61 97 Venezuela 2,451 0.9 5 6 46 46 49 48 NOR1H AMERICA 98 Canada 19,023 3.4 99 United States 21,099 3.3 3 34 64 EUROPE 100 Albania 930 101 Austria 17,360 1.9 5 40 56 102 Belgium 16,389 1.7 2 36 62 103 Bulgaria 2,317 14 11 54 59 32 29 104 Czechoslovakia 3,455 7 60 34 105 Denmark 20,511 2.2 106 Finland 22,056 2.9 107 France 17,835 2.1 ·4 34 62 108 Germany 20,754 1.9 2 43 55 109 Greece 5,342 1.6 18 31 51 110 Hungary 2,582 1.6 17 14 41 36 42 50 111 Ireland 8,497 1.7 112 Italy 15,155 2.1 6 39 55 113 Netherlands 16,011 1.8 4 33 64 114 Norway 21,850 3.5 4 40 57 115 Poland 1,778 116 Portugal 4,264 0.8 10 39 51 117 Romania 13 60 27 118 Spain 9,151 2.5 7 39 54 119 Sweden 21,712 1.8 120 Switzerland 30,268 2.0 121 U. Kingdom 14,566 2.9 122 Yugoslavia 2,919 1.3 12 10 44 42 44 48 123 U.S.S.R. . . Data not available. 592 MANAGING THE WORLD'S FORESTS TABLE 19, Social Indicators Energy Con- Popu- sump- Urban la tion Forest tion per Population Population Density Area Capita Growth o/o of Growth (per per (kg of Total rate total rate square Capita oil (mil- (per- popula- (per- kilo- (hec- equiva- lions) cent) tion cent) meter) tares) lent) --- ----- --- 1989 1980-89 1989 1980-89 1987 1989 1987 WORLD 5292.2 1.7 43 2.6 40 0.8 1,300 AFRICA 647.5 3.0 35 5.0 20 1.1 300 1 Algeria 24.5 2.9 44 3.9 10 0.1 1,035 2 Angola 9.7 2.9 28 5.2 7 5.5 203 3 Benin 4.6 3.2 41 6.0 38 0.8 46 4 Botswana 1.2 3.1 23 6.3 2 26.8 431 5 Burkina Faso 8.8 2.7 9 4.6 30 0.5 18 6 Burundi 5.3 3.0 7 6.9 179 0.0 21 7 Cameroon 11.6 3.3 48 ~).7 23 2.0 144 8 Cent. Afr. Rep. 3.0 2.8 46 4.4 4 12.2 30 9 Chad 5.5 2.5 32 ~).7 4 2.4 18 10 Congo 2.2 3.2 42 4.2 6 9.7 223 11 Cote d'Ivoire 11.7 4.3 46 5.9 33 0.8 172 12 Egypt 51.4 2.4 48 3.4 49 0.0 588 13 Eq.Guinea 0.3 2.4 64 3.6 12 3.8 73 14 Ethiopia 48.9 3.1 13 4.7 38 0.6 21 15 Gabon 1.1 2.6 45 4.4 4 18.6 1,142 16 Ghana 14.4 3.2 33 3.9 57 0.6 129 17 Guinea 5.5 2.7 25 5.0 21 1.9 59 18 Guinea-Bissau 1.0 2.1 30 4.1 25 2.2 37 19 Kenya 23.3 3.6 23 6.4 37 0.1 99 20 Liberia 2.5 2.8 43 ll.5 21 0.8 169 21 Libya 4.4 3.5 69 5.1 2 0.0 2,671 22 Madagascar 11.2 2.9 24 5.1 18 1.2 39 23 Malawi 8.2 3.3 14 7.3 65 0.5 41 24 Mali 8.2 2.8 19 3.9 6 0.9 24 25 Mauritania 2.0 2.5 41 S.6 2 0.3 113 26 Morocco 24.6 2.5 48 4.1 52 0.1 242 27 Mozambique 15.4 2.7 26 ?.7 18 1.0 86 28 Namibia 1.3 3.2 56 4.9 1 14.2 29 Niger 7.5 3.0 19 5.9 6 0.3 42 30 Nigeria 113.7 3.2 35 5.3 115 0.1 133 31 Rwanda 6.9 3.5 7 6.7 244 0.0 42 32 Senegal 7.2 3.0 38 3.9 35 1.5 155 33 Sierra Leone 4.0 2.6 27 4.4 54 0.5 77 34 Somalia 6.1 3.1 37 4.9 9 1.5 81 35 South Africa 34.9 2.7 58 3.4 27 0.0 2,464 36 Sudan 24.4 2.7 22 4.0 9 2.0 58 37 Tanzania 25.6 3.6 31 8.1 25 1.6 35 38 Togo 3.5 3.7 25 6.1 57 0.5 52 39 Tunisia 8.0 2.3 54 2.8 47 0.0 496 (continued) STATISTICAL APPENDIX 593 TABLE 19. Social Indicators (continued) Energy Con- Popu- sump- Urban la tion Forest tion per Population Population Density Area Capita Growth %of Growth (per per (kg of Total rate total rate square Capita oil (mil- (per- popula- (per- kilo- (hec- equiva- lions) cent) tion cent) meter) tares) lent) --- --- --- 1989 1980-S9 1989 1980-S9 1987 1989 1987 40 Uganda 16.8 3.4 10 5.0 66 0.4 26 41 Zaire 34.4 3.2 39 4.4 14 5.2 73 42 Zambia 7.8 3.7 55 5.5 10 3.8 379 43 Zimbabwe 9.6 2.8 27 4.7 23 2.1 512 ASIA & PACIFIC 3135.0 2.0 30 3.2 117 0.2 800 44 Afghanistan 19.9 2.3 21 5.4 29 0.0 71 45 Australia 16.8 1.4 86 1.4 2 2.5 1,614 46 Bangladesh 111.6 2.5 13 5.1 737 0.0 47 47 Bhutan 1.4 2.2 5 6.2 29 82.0 48 China 1105.1 1.5 53 7.8 112 0.1 570 49 India 832.5 2.1 28 3.9 243 0.1 208 50 Indonesia 178.2 1.9 28 4.4 90 0.0 216 51 Iran 50.2 3.3 54 4.4 29 0.0 955 52 Iraq 18.3 3.5 73 4.4 39 0.0 732 53 Israel 4.5 l.8 91 2.1 211 5.3 1,967 54 Japan 123.0 0.4 77 0.4 323 0.1 3,232 55 Kampuchea 8.0 2.4 11 3.8 42 0.6 59 56 Korea,DPR 21.1 1.5 67 2.5 170 0.2 2,165 57 Korea~ Rep. 42.4 1.0 71 2.9 420 0.3 1,475 58 LaoPDR 4.1 3.1 18 6.1 16 5.2 37 59 Malaysia 17.3 2.4 41 4.4 50 0.5 771 60 Mongolia 2.1 2.7 51 2.9 1 15.0 1,181 61 Myanmar 40.8 2.0 24 2.6 58 0.1 73 62 Nepal 18.4 2.5 9 6.7 124 0.1 23 63 New Zealand 3.3 1.5 84 1.7 12 2.2 4,216 64 Pakistan 110.0 3.4 32 4.8 129 0.3 207 65 P. New Guinea 3.8 2.4 16 4.3 8 2.5 229 66 Philippines 61.2 2.3 42 3.6 195 0.0 241 67 Saudi Arabia 13.6 4.0 76 5.1 6 0.1 3,292 68 Sri Lanka 16.8 1.2 21 1.4 249 0.0 160 69 Syria 12.1 3.7 51 4.6 61 1.3 901 70 Thailand 55.2 1.3 22 3.9 104 0.4 330 71 Turkey 54.9 2.1 48 2.9 68 0.2 763 72 VietNam 65.8 2.4 190 0.0 88 73 Yemen 11.2 3.4 28 6.7 20 0.0 CENTRAL& SOUTH AMERICA 448.3 2.0 65 3.7 22 2.1 500 74 Argentina 31.9 1.2 86 1.6 11 1.4 1,472 75 Belize 0.2 2.6 51 3.2 8 7.8 447 76 Bolivia 7.1 2.8 51 4.2 6 9.4 258 (continued) 594 MANAGING THE WORLD'S FORESTS TABLE 19. Social Indicators (continued) Energy Con- Popu- sump- Urban la tion Forest tion per Population Population Density Area Capita Growth %of G:rowth (per per (kg of Total rate total rate square Capita oil (mil- (per- popula- (per- kilo- (hec- equiva- lions) cent) tion cent) meter) tares) lent) --- ------ --- 1989 1980-89 1989 1980-89 1987 1989 1987 77 Brazil 147.3 2.0 76 3.1 17 3.5 825 78 Chile 13.0 1.7 85 2.2 17 0.6 821 79 Colombia 32.3 2.0 70 2.8 27 1.6 757 80 Costa Rica 2.7 2.4 45 2.4 51 0.7 581 81 Cuba 10.5 1.0 74 1.8 93 0.1 1,112 82 Dominican Rep. 7.0 2.1 59 3.7 138 0.1 335 83 Ecuador 10.3 2.5 56 4.2 35 1.4 625 84 El Salvador 5.1 2.2 44 3.0 235 0.0 217 85 Guatemala 8.9 2.9 33 2.9 77 0.5 169 86 Guyana 0.8 0.1 34 1.5 4 23.4 585 87 Haiti 6.4 1.8 30 3.9 221 0.0 50 88 Honduras 5.0 2.9 43 4.8 42 0.8 192 89 Jamaica 2.4 0.7 52 1.8 215 0.0 852 90 Mexieo 85.4 2.1 72 2.9 42 0.6 1,299 91 Nicaragua 3.7 3.2 59 4.3 27 1.2 256 92 Panama 2.4 2.0 54 2.9 30 1.8 1,626 93 Paraguay 4.2 2.9 47 4•. 2 10 4.7 224 94 Peru 21.1 2.2 70 3.0 16 3.3 485 95 Suriname 0.4 2.1 47 2.9 3 34.4 1,635 96 Uruguay 3.1 0.6 85 0.8 17 0.2 765 97 Venezuela 19.2 2.6 84 3.1 20 1.8 2,394 NORTH AMERlCA 275.7 1.0 75 I.I 15 2.7 6600 98 Canada 26.3 1.3 76 1.5 3 10.0 9,155 99 United States 248.2 0.8 74 0.8 26 0.8 7,265 EUROPE 497.7 0.2 73 I.I 108 0.3 3000 100 Albania 3.2 ,1.9 35 2.7 107 0.3 1,178 101 Austria 7.6 0.0 57 0.6 90 0.5 102 Belgium 9.9 0.1 97 0.2 0.1 103 Bulgaria 9.0 0.2 70 1.2 81 0.4 4,705 104 Czechoslovakia 15.6 0.2 68 l.1 122 0.3 4,954 105 Denmark 5.1 0.0 86 0.3 119 0.1 3,889 106 Finland 5.0 0.5 60 0.5 15 4.0 5,579 107 France 56.1 0.4 74 0.6 101 0.2 3,729 108 Germany 78.0 0.0 84 0.0 220 0.1 4,852 109 Greece 10.0 0.3 62 l.1 76 0.3 1,970 110 Hungary 10.6 -0.2 60 0.5 114 0.2 3,062 111 Ireland 3.5 0.0 59 0.7 50 0.1 2,505 112 Italy 57.5 0.2 68 0.5 190 0.1 2,676 113 Netherlands 14.8 0.5 88 0.5 393 0.0 5,198 (continued) STA T/STICAL APPENDIX 595 TABLE 19. Social Indicators (continued) Energy Con- Popu- sump- Urban la tion Forest tion per Population Population Density Area Capita Growth o/o of Growth (per per (kgof Total rate total rate square Capita oil (mil- (per- popula- (per- kilo- (hec- equiva- lions) cent) ti.on cent) meter) tares) lent) --- --- ------ 1989 1980-89 1989 1980-89 1987 1989 1987 114 Norway 4.2 0.5 74 0.9 13 1.8 8,938 115 Poland 38.1 0.5 62 1.1 120 0.2 3,385 116 Portugal 10.3 0.4 33 1.7 111 0.3 1,321 117 Romania 23.1 0.5 49 0.5 97 0.3 3,465 118 Spain 39.2 0.4 78 1.1 77 0.2 1,938 119 Sweden 8.5 0.6 84 0.7 19 2.9 6,488 120 Switzerland 6.5 -0.5 60 1.2 158 0.1 4,107 121 United Kfagdom 57.3 0.3 92 0.5 232 0.0 3,805 122 Yugoslavia 23.7 0.6 49 2.2 92 0.4 2,115 123 U.S.S.R. 287.7 0.7 67 1.3 13 3.2 4,500 . . Data not available. 596 MANAGING THE WORLD'S FORESTS TABLE 20. Countries Not Included in Tables l-19 Total Other Land Forest Wooded Area Population Area Area (ODO ha) (thousands) (ODO ha) (000 ha) 1 American Samoa 20 32 0 14 2 Antigua and Barbuda 44 75 9 16 3 Bahamas 1,007 210 324 0 4 Bahrain 62 347 0 0 5 Barbados 43 249 0 5 6 Bermuda 5 71 0 1 7 Brunei 527 196 323 237 8 British Virgin Is. 15 12 3 2 9 Cape Verde 403 296 4 1 10 Cayman Islands 26 17 0 6 11 Comoros 217 381 17 28 12 Cook Islands 23 18 0 0 13 Cyprus 924 629 153 40 14 Djibouti 2,198 310 71 44 15 Dominica 75 73 41 14 16 Fiji 1,827 629 851 6 17 French Guiana 8,915 69 7,833 85 18 French Polynesia 366 148 0 115 19 Gambia 1,000 584 216 560 20 Greenland 34,170 52 0 10 21 Grenada 34 107 5 1 22 Guadeloupe 176 327 94 0 23 Hong Kong 100 5,039 0 13 24 Iceland 10,025 228 0 100 25 Jordan 9,718 2,923 71 75 26 Kiribati 71 59 0 2 27 Kuwait 1,782 1,375 0 0 28 Lebanon 1,023 2,669 39 45 29 Lesotho 3,035 1,339 3 16 30 Macao 2 323 0 0 31 Maldives 30 155 0 1 32 Martinique 106 326 0 28 33 Mauritius 185 955 14 32 34 Monserrat 10 12 3 1 35 Netherlands Antilles 96 252 0 7 36 New Caledonia 1,876 139 705 630 37 Niue 26 3 0 6 38 Oman 21,246 984 0 0 39 Puerto Rico 886 3,199 284 12 40 Qatar 1,100 225 0 0 41 Reunion 250 510 90 42 42 Saint Helena 31 5 2 8 43 Saint Lucia 61 120 8 29 44 Sao Tome 96 85 56 o 45 Seychelles 27 65 4 o 46 Singapore 57 2,415 o 3 47 Solomon Islands 2,754 255 2,457 40 (continued) STATISTICAL APPENDIX 597 TABLE 20. Countries Not Included in Tables 1-19 (continued) Total Other Land Forest Wooded Area Population Area Area (000 ha) (thousands) (000 ha) (000 ha) 48 St. Vincent 34 99 12 1 49 Swaziland 1,720 559 176 0 50 Tonga 67 97 0 8 51 Trinidad and Tobago 513 1,095 224 63 52 Turks Is. 43 7 0 0 53 Tuvalu 16 8 0 0 54 United Arab Emirates 8,360 980 0 0 55 Vanuatu 1,476 117 236 0 56 Western Samoa 283 155 142 31 0 Zero or less than half the unit of measure. Technical Notes to the Statistical Tables Table 1. Forest Area Resources souRcEs: FAO 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome. ECE/FAO. 1985. The Forest Resources of the ECE Region. Geneva. FAO is responsible for compilation of forest area data. Countries usu- ally report directly to FAO, or FAO uses country sources to determine forest area. These data are the most current, but new FAO data are expected in early 1992. This table's data are from the FAO/UNEP GEMS Forest Re- sources Assessment for the tropical regions and the 1985 FAO /ECE report on forests in the ECE region. All the data were published in 1982 and up- dated in 1985 in the Interim Report, which lists all countries' total forest and other wooded areas. However, the Interim Report only lists forest types for the developing countries. The ECE/FAO report is used for the ECE coun- tries' forest types. For Australia, New Zealand, South Africa, Japan, and both Koreas, only total forest and. other wooded areas are reported. Some of the smaller countries are not listed (less than 1 million population or less than 100,000 hectares [ha] of forest), but the totals reflect all the regions' or world's forest. Regional and world totals include "other countries" not listed in the table. Total forest and other wooded land is a summation of closed and open forests. All forests in North America, Europe, and the former U.S.S.R. are temperate forests. Temperate forests in Africa (the North African coun- tries, South Africa, Lesotho, and Swaziland) total 6.9 million ha. Temperate forests in Asia and the Pacific (China, Mongolia, Japan, and both Koreas, total 134 million ha. In Latin America (Uruguay, Chile, and Argentina) the temperate forests total 54.1 million ha. 599 600 MANAGING THE WORLD'S FORESTS Generally, closed forests have trees that cover most of the ground (greater than 20 percent density). The ECE region (temperate forests) includes all plantations in this category. For the tropical regions and temperate regions in developing countries the closed forests refer only to broad-leaved, conif- erous, or bamboo forests that cover most of the ground. Open forest is a term used only in the Interim Report to describe discon- tinuous forest stands with grass cover (usually crown cover of 5 to 20 per- cent) of area. The ECE countries call these forests open woodlands, and they are listed in "other wooded areas." Other wooded areas are divided into tropical and temperate categories. The tropical region's classification of other wooded areas includes planta- tions, forest fallow, and shrubland. Plantations come under this category to keep them separate from natural open and closed forests. The temperate region's other wooded areas are open woodlands, isolated groups of trees, and shrubland. Forest fallow area in Africa (177.7 million ha), Asia and the Pacific (73.3 million ha), and Latin America (172 million ha) totals 423 million ha. Table 2. Deforestation Rates SOURCES: FAO. 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome. WR!. 1990. World Resources Report, 1990-91. Washington, D.C. Deforestation data are based on 1980 FAO data, and updated between 1981 and 1985. New data are continuously being reported. See WRI 1991, "World Resources Report," and IUCN 1991, "Conservation Atlas of Tropical Forests." The FAO recently reported that the 1990 tropical deforestation rate was 16.8 million ha for 62 countries with mostly tropical moist forests. The FAO report gave regional annual deforestation totals for Africa (4.8 million ha), Asia (4.7 million ha), and Latin America (7.3 million ha), but did not report by country. Deforestation refers to the permanent clearing of forest lands for use in shifting cultivation, permanent agriculture, or settlements. It is not clear whether these data are net or gross deforestation. However, in most devel- oping countries, deforested land is not replanted. The estimated ratio for deforestation to reforestation is 14:1. Regional and world deforestation totals reflect only the countries that reported. Note that some totals do not have a breakdown by open or closed forest. Countries in the developed world are assumed not to have any net deforestation, but because no data are available, these countries are not listed as having zero deforestation. STATISTICAL APPENDIX 601 Table 3. Reforestation and Plantations SOURCES: FAO. 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome. WRI 1990. World Resources Report, 1990-91. Washington, D.C. FAO 1985. Forest Resources. Rome. The FAO Interim Report includes a table on reforestation for the devel- oping countries. It shows the areas of plantations by 1980, and the estimated area between 1981 and 1985. Breakdown by type of plantation is given only for the developing countries. Some countries report only totals. An annual reforestation rate is determined from the estimated extent from 1981 to 1985. For the developed countries, the FAO produced a table of forest re- newal in 1985. An annual renewal rate is equivalent to the reforestation rate. Average annual reforestation refers to the annual rate of plantations established. Reforestation does not include regeneration of old tree crops (through natural regeneration or forest management), although some coun- tries may report regeneration as reforestation. Also, many of the trees planted in farm or village woodlot or outside forest areas are not included in this category. In the developed countries the terms renewal and regeneration are synonymous with reforestation. Forest plantations are established artificially on lands that previously did not carry forest, or on lands that have carried forest in living memory or during the past 50 years. Regional and world totals reflect only the countries reporting data. Table 4. Land Use and Change SOURCES: FAO. 1988. Production Yearbook. Rome. WRI 1990. World Resources Report, 1990-91. Washington, D.C. Every year FAO updates the distribution of land by use. These data are from FAO as cited by WRL Land area excludes all inland water bodies. Greenland is included in the world total only. Cropland includes land under temporary and permanent crops, tempo- rary meadows, market and kitchen gardens, and temporary fallow. Permanent pasture is pastureland used for five years or more for for- age; it includes wildlands used for pasture. Forest and woodlands are land under natural or planted trees; they also include logged-over areas that will be reforested in the near future. Forest and woodlands data will not equal the data in table 1, because these data for forest and woodlands include land from which the forests have been cleared but which will be reforested in the foreseeable future. Other land includes uncultivated land, grassland not used for pasture, built-on areas, wetlands, wasteland, and roads. 602 MANAGING THE WORLD'S FORESTS Tables 5 and 6, Productive Forests SOURCES: ECE/FAO 1985. The Forest Resources of the ECE Region. Geneva. FAO/UNEP(GEMS). 1982. Tropical Forest Resources Assessment, 1980. Rome. FAO 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome. Most forest activities will occur in the so··called exploitable or produc- tive forests. The GEMS report uses the term productive, whereas the ECE uses the term exploitable. These forests do not include protected areas or forest land that is physically inaccessible. For the developing countries and non-ECE countries, total productive forest area is from the Interim Report. The area of productive forests for the ECE countries is from the ECE report. Volume of productive forests was only available for tropical developing countries (from the GEMS report) and the ECE countries. Breakdown for type of productive forest area was not available for Egypt, Libya, Mauritania, China, either Korea, Mongolia, Saudi Arabia, Al- bania, or Canada. The data for Turkey are from the ECE report. Undisturbed, productive, closed broad-leaved forests (used by FAO) are the same as primary tropical moist forests. The logged-over and man- aged, closed broad-leaved forests constitute the secondary forests of the tropical moist forest zone. The ECE countries do not distinguish broad-leaved or coniferous by management class, because all the exploitable forest in these regions is managed. The coniferous forests are divided into unmanaged and managed be- cause they are not as abundant in the tropical zone. Regional and world totals reflect only those countries reporting. Table 7. Protected Forests SOURCES: FAO 1988. Production Yearbook. Rome. WRI 1990. World Resources Report, 1990-91. Washington, D.C. IUCN 1990. United Nations List of National Parks and Protected Areas. Geneva. IUCN, ECE/FAO 1985. The Forest Re- sources of the ECE Region. Geneva. FAO 1988. An Interim Report on the State of Forest Resources in the Developing Countries. Rome. Total protected forest includes all legally protected forest. Closed pro- tected forest for developing countries is from FAO's Interim Report. The report's term for protected forest is "unproductive and legally protected" closed forests. Closed protected forest for the ECE region is from the ECE/ FAO report. The ECE' s term is closed forest that is unexploitable and is a national park or reserve. National Protection Systems data are from World Conservation Monitoring Center (WCMC), which uses five categories of protected forest. STATISTICAL APPENDIX 603 Table 8. Forest Sector Production SOURCE: FAO 1991. Unpublished data. This table shows the relationship between the value of forest-sector production and gross domestic product of countries. The comparison is done for 1961 and 1989, and all values are stated in 1980 prices. The overall GDP values, which are not shown, are derived from an assembly of UN, World Bank, and OECD sources. Table 18 shows GDP for agriculture, industry, and services sectors. The gross value of forest-sector product is compiled from statistics on production and net exports of forest products valued at average world unit values of the component products. The percentages of forest sector, forest industry, and nonindustrial for- estry and logging of total GDP are shown. The forest-sector percentage of GDP is the sum of the forest industry and the nonindustrial forestry and logging percents. Forest industry is the gross value of industrial production: sawnwood, wood-based panels, and paper, plus net exports of industrial roundwood, and pulp. Nonindustrial forestry and logging is the gross value of fuelwood and charcoal plus "other industrial roundwood," i.e., industrial roundwood not consumed in sawnwood, panels, or pulp and paper manu- facture or exported, but used in unprocessed form domestically. In many cases the resulting percentages are overestimates (1) because the forest-sector statistics are values of gross product, not value added, or (2) because forest products have been valued at world average border prices, which may be out of line with the pricing used in generating estimates of GDP. Table 9. Forest Products Trade Value SOURCE: FAO 1990. Forest Products Yearbook, 1977-1988. Rome. Compilation of forest products is based on country reports. Nonreporting countries' data are from unofficial publications. Estimates of trade for Alba- nia, China, Korea DPR, Mongolia, and Viet Nam are from information pro- vided by trading partners. Data under South Africa include Namibia. Data for Belgium include Luxembourg. Estimates for China include the province of Taiwan. A comparison is made between 1980 and 1988 data. At the time of compilation of the data, 1989 data were not available. Growth rates using the least squares method are used to show change in exports and imports over the nine years. Forest products (both coniferous and nonconiferous) include fuelwood, charcoal, sawlogs, veneer logs, pulpwood, chips and particles, wood resi- dues, other industrial roundwood, sawnwood, veneer sheets, plywood, par- ticle board, fiberboard, mechanical wood pulp, chemical wood pulp, dis- 604 MANAGING THE WORLD'S FORESTS solved wood pulp, newsprint, printing and writing paper, and other paper and paperboard. Values are in current U.S. dollars. Total merchandise ex- ports are from the World Bank data base. Exports and imports are not equal. The difference could be explained by a number of factors including quantity, quality, shipping costs, marketing expertise, and underinvoicing. Totals are based only on those countries reporting. Countries that did not report are assumed to be zero. Percentages of total trade for some coun- tries are not given because total trade figures for those countries were un- available. Quite a few of the smaller developing countries that show zeros have a trade value of under half a million dollars. Tables 10-13. Roundwood SOURCE: FAO 1990. Forest Products Yearbook, 1977-1988. Rome. All commodities include coniferous and nonconiferous species. Round- wood is "wood in the rough or natural state." It comprises all wood ob- tained from removals or recovered from forest and nonforest areas. Indus- trial roundwood includes sawlogs, veneer logs, pulpwood, and other wood products. Woodfuels are wood in the rough to be used as fuel for purposes such as cooking. Woodfuel includes fuelwood and wood for charcoal, pit kilns, and portable ovens. Totals are based only on those countries reporting. Countries that did not report are assumed to be zero. A comparison is made between 1980 and 1988 data. At the time of compilation of the data, 1989 data were not avail- able. Growth rates using the least squares method are used to show change in production over the nine years. Tables 14-17. Processed Wood SOURCE: FAO 1990. Forest Products Yearbook, 1977-1988. Rome. Mechanical wood is the sum of sawnwood and wood-based panels. Sawnwood (and sleepers) is timber derived from industrial roundwood (sawlogs). With few exceptions, sawnwood is greater than 5 mm in thick- ness. Wood-based panels include veneer sheets, plywood, particle board and fiberboard, and noncompressed board. Paper and paperboard, derived from wood pulp, include newsprint, printing and writing paper, and other paper products. Totals are based only on those countries reporting. Countries that did not report are assumed to be zero. A comparison is made between 1980 and 1988 data. At the time of compilation of the data, 1989 data were not avail- able. Growth rates using the least squares method are used to show change in production over the nine years. STA T/STICAL APPENDIX 605 Table 18. Economic Indicators SOURCE: World Bank 1991. World Bank Atlas. Washington, D.C. Unpublished data from World Bank, various years. Gross national product (GNP) is the sum of two components: the gross domestic product (GDP) and the net factor income from abroad. GNP per capita is calculated according to the World Bank Atlas method. Data for the following countries are from 1987 reports: Liberia, Libya, Somalia, South Africa, Uganda, and Bhutan. Data for Afghanistan, Iraq, Syria, the former U.S.S.R., and Albania are from a CIA report, The World Factbook, 1988. Gross domestic product is the final output of goods and services pro- duced by the domestic economy, including net exports of goods and nonfactor services. The GDP growth rate is for 1980-88 for the following countries: Botswana, Cameroon, Gabon, Ghana, Liberia, Morocco, Senegal, Zambia, China, Indonesia, Israel, Malaysia, Nepal, Saudi Arabia, Syria, New Zealand, Chile, Dominican Republic, El Salvador, Guatemala, Hungary, Portugal, and Spain. Distribution of GDP is based on current price series in U.S. dollars. Agriculture covers forestry, hunting, and fishing. In developing countries, with high levels of subsistence farming, much of agricultural production is either not exchanged or not exchanged for money. This situation increases the difficulty of measuring the contribution of agriculture to GDP and re- duces the reliability and comparability of such numbers. Industry comprises value added in mining, manufacturing, construction, electricity, water, and gas. Services constitute value added in other branches of economic activity, including imputed bank charges, import duties, and any statistical discrep- ancies noted by national compilers. Table 19. Social Indicators SOURCE: World Bank 1990. Social Indicators of Development 1990. Washington, o.c. Population data are from World Bank sources; some information from the UN and Population Council is used. Urban population countries may use different definitions that may affect comparability. Population growth rates are derived using the annual least squares growth rates. Population density is derived from World Bank population information and FAO land area data. Forest area per capita is taken from FAO data. Energy consump- tion per capita is from World Bank sources. This indicates the annual con- sumption of commercial primary energy (coal; lignite; petroleum; natural gas; and hydro, nuclear, and geothermal electricity) in kilograms of oil equiva- lent per capita. Managing the World's Forests is based on a study of fo rest utilization around the world. It discusses the concerns that must be considered for conservation and sustainable development. • Nearly half of the world's population depends to some extent on forests goods. • Exploitation of forests can lead to climate change, degraded lands, destruction of ecosystems, and loss of species. The conclusions explored in the book-variety of land use: from maintenance of the old growth , pristine forest to the clearing of forests for sustainable agriculture-are consistent with conservation and development goals. Actions at the local, national , and global levels can lead to destructive exploitation of forests or to good and productive management strategies. This excellent compilation of essays from the world's leading ecologists , economists , and scientific researchers will answer some of the questions on everyone's mind. ·Why are forests and trees so important? •What is the problem? • What are the causes of the problem? • How should this problem be addressed? Narendra P. Sharma is presently a Principal Economist at the World Bank in Washington, D.C. and an adjunct professor in the School of Environment, Duke University, North Carolina , U.S.A. ISBN 0-8403-7885-8 90000 KENDALL / HUNT PUBLISHING COMPANY 9 8084 2460 Kerper Boulevard P .O . Box 539 Dubuque. Iowa 52004 -0539