12 6 SO fVodVe*[ 1qqg Agri culture and Enrironmental Challenges Proceedings of the Thirteenth Agricultural Sector Symposium Jitendra P. Srivastava and Harold Alderman, editors i~~~~~ \ \6 ~ F - Z-Dw Agriculture and Environmental Challenges Proceedings of the Thirteenth Agricultural Sector Symposium Jitendra P. Srivastava and Harold Alderman, editors The World Bank Washington, D.C. Copyright C) 1993 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1993 These symposium proceedings are not a fornal publication of the World Bank and are circulated to encourage discussion and comment and to communicate information quickly to the development community; citation and the use of these proceedings should take account of their provisional character. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to Director, Office of the Publisher, at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncornmercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. Jitendra P. Srivastava is senior agriculturalist in the Agriculture and Natural Resources Department at the World Bank, and Harold Alderman is senior economist at the Policy Research Department. Library of Congress Cataloging-in-Publication Data Agricultural Sector Symposium (13th: 1993: World Bank) Agriculture and environmental challenges: proceedings of the Thirteenth Agricultural Sector Symposium / Jitendra P. Srivastava and Harold Alderman, editors. p. cm. Includes bibliographical references. ISBN 0-8213-2585-X 1. Agriculture-Environmental aspects-Congresses. I. Srivastava, Jitendra, 1940- . II. Alderman, Harold, 1948- III. International Bank for Reconstruction and Development. IV. Title. S589.75.A365 1993 338.1'09172'4-dc20 93-30489 CIP FOREWORD The tradition of the Annual Agricultural Symposium is well established and offers an occasion when World Bank agricultural staff have the opportunity of meeting to consider matters of current importance to Bank agricultural policy and implementation. The focus of this year's Symposium - Agriculture and Environmental Challenges -- is no exception. In fact, it is an arena where concern for environmental issues along side those for productivity is most critical. The wealth of information and practical applications of these premises presented in the papers provided a substantial foundation for future project assessments and implementation of new and exciting ideas. We were again honored by the presence of Bank President, Mr. Lewis Preston, who opened the Symposium. Newly appointed Vice President for Environmentally Sustainable Development, Ismail Serageldin delivered an excellent opening address, "Agriculture and Environmentally Sustainable Development," which was followed by an enthusiastic debate among Bank staff. By making these papers available in the Proceedings, it is our intention to share this wealth of knowledge and examples with those within the Bank who were unable to attend the Symposium and those outside the Bank who share our concerns for the challenges ahead in advancing the agricultural development process in an environmentally sustainable way. I wish to acknowledge the outstanding efforts of the coconveners, Jitendra Srivastava and Harold Alderman, their staff members, and the staff of the Training Division under the direction of Surinder Deol, who orchestrated this year's Symposium. I also wish to acknowledge the special efforts of Mary Horne in editing this Symposium, and for her work in previous years. Michel Petit Director Agriculture and Natural Resources Department TABLE OF CONTENTS Opening Session The Challenge to Sustainability Lewis T. Preston 1 Agriculture and Environmentally Sustainable Development Ismail Serageldin 5 Technical Considerations for Sustainable Agriculture Richard Grimshaw 17 Socioeconomic and Policy Considerations for Sustainable Agricultural Development Per Pinstrup-Andersen 35 Agriculture and Environment: Responsible Management Does Population Growth Inevitably Lead to Land Degradation? John English 45 Conservation Tillage as a Tool to Conserve Soil, Moisture, Energy, and Equipment in Large and Small Crop Production Systems John Hebblethwaite 59 Moisture Management in Semiarid Temperate Regions B.A. Stewart 67 Soil Fertility Management for Intensive Agriculture in the Humid Tropics Christian Pieri 81 Preserving the Options International Research for Sustainable Agriculture Hubert Zandstra 101 Biological Nitrogen Fertilization: Present and Future Applications Ralph W.F. Hardy 109 Making IPM Work: Developing Country Experience and Prospects Jeff Waage 119 Changing Perceptions and Practices of Central American Smallholders Keith Andrews 135 Women in Agricultural Resource Management Raising the Productivity of Women Farmers in Sub-Saharan Africa Katrine Saito 147 Agricultural Extension for Women - Experience from Nigeria Ndanusa Mijindadi 157 Women in Agricultural Resource Management Aruna Bagchee 167 Targeting Women in Extension Willem Zijp 175 Women and Fuel Issues Augusta Molnar 183 Poverty and Agricultural Resource Management The Population, Environment, and Agriculture Nexus in Sub-Saharan Africa Kevin Cleaver 199 Economic Stagnation and Deforestation in Costa Rica and the Philippines Maria Concepcion Cruz 215 Credit and Enviromnental Rehabilitation: A Case for Supply-Led Credit? Monica Fong 235 Does Conservation Condemn the Poor to Perpetual Poverty? A Nongovernmental Organization Perspective Jim Nations 245 Closing Session World Bank Experience in Renewable Resource Management: Implications for Lending and Evaluation Work Alfredo Sfeir-Younis 251 Implications of the Earth Summit for Sustainable Agriculture and World Bank Activities Mohamed Al-Ashry 271 The World Bank's Water Policy: Executive Summary K. William Easter, Gershon Feder Guy Le Moigne and A.M. Duda 277 Challenge to Bank Agricultural Staff Michel Petit 285 OPENING SESSION The Challenge of Sustainability Lewis T. Preston* Introduction I'd like to welcome everyone to the Thirteenth Agricultural Symposium. In opening this Symposium last year, I said that I didn't know enough about agriculture to address the major issues in the sector. It's twelve months later -- and I'm still not an agricultural expert. However, I have become increasingly aware of the central role of agriculture in our work - and I'd like to talk about that this morning. It's also important for me to have your views. We have agreed to present an agriculture sector review and strategy paper to our Executive Directors. This Symposium can make a valuable contribution to that effort. Agricultural Challenges The world's population is projected to expand to about nine million over the next forty years. Demand for food and fuel will rise tremendously. Meeting that demand will require more intensive use of many natural resources -- especially agricultural land, forest, water, and fisheries. In that respect, agriculture is at the center of development -- cutting across the major themes of poverty reduction, economic growth, and environmental protection. And yet, Michel Petit tells me that he feels agriculture has been treated with "benign neglect" by the Bank in recent times. He says that we often don't give it the priority it deserves in the policy dialogue, and we don't give enough attention to helping shape long-term sectoral strategies in our borrowing countries. He has also pointed out that agriculture's share of the lending program has dropped from about one- third in the 1970s to around 17 to 18 percent today. While recognizing that Mr. Petit is not exactly an "unbiased source" on this matter, I do think he raises some questions that we ought to consider. We know that agriculture has been abused and even plundered in many developing countries. But are we doing enough through the policy dialogue, public expenditure reviews, and other instruments to help remedy the situation? We know that there has been a severe lack of coordination among donors. My own "favorite" example is Tanzania where 120 different kinds of agricultural research projects were discovered -- each supported by a donor. Are we using our analytical skills and knowledge to achieve better focus? Increasing our lending is not necessarily the answer. As you may know, the Wapenhans Report found that 42 percent of Bank-supported agricultural projects have "major problems." Weak policy frameworks, institutional shortcomings, and external factors have all contributed to poor performance. But we also know -- and Wapenhans confirmed it -- that our projects have often been too ambitious and complicated; too "top-down"; and that we have too often failed to ensure local ownership and follow-up. I'm aware of some of the efforts being made to improve performance in the sector - the emphasis on helping to support smallholder farmers and, particularly, the role of women in * President, World Bank. 2 agriculture. The innovative work being done in social forestry, integrated pest management, and in involving local communities in environmental protection is also important and exciting. The name of the game, however, must be implementation. There are just too many examples of perfectly sensible policies and programs that have not been followed through in order to have the needed impact. Why is that? Is it a lack of commitment on our part? Or on the part of our borrowers? Are our agriculturalists and country economists communicating with each other? I hope you can devote some time to discussing these questions -- because they're vital to our future work in the sector. Environmental Challenges The link between agriculture and the environment is also very important -- and it will be the major theme of your discussions over the next few days. In the long run, development and environmental protection are mutually reinforcing objectives. In the short run, however, there are many difficult tradeoffs and issues to be resolved, including the effects of pricing and incentives, population growth rates, the role of women, and the balance between agricultural productivity and environmental sustainability. We are far from having all the answers. We need more research at the national and international level -- through the CGIAR, for example. We need new technologies. We need to place the concept of sustainability at the center of our work. Meeting the Challenge The Reorganization should help us meet the challenge. In particular, the new Vice Presidency for Environmentally Sustainable Development specifically links agriculture with the environment. It will give added focus to our research and operations. The recent completion of IDA-10 at a level of US$18 billion, a remarkable achievement under the circumstances, should also help. The IDA Deputies emphasized, for example, that resources should be linked to the preparation of National Environmental Action Plans in our borrowing countries. The Global Environment Facility (GEF), which has recently been placed on a more permanent footing and opened up to universal membership, can also help by working to integrate environmental and development objectives at the global level. Conclusion: Challenge for the 1990s A major challenge facing us in the 1990s is going to be keeping up with the increasing demand from our developing member countries for assistance in helping them address sustainability issues. Frankly, I don't think we yet have the right "skills mix" to meet that demand. We haven't yet developed all the necessary tools. We need to do more to incorporate the lessons of our successes and failures. And, again, we need to improve on-the-ground implementation, which is where development really counts. We're still on the steep end of the learning curve. 3 However, I believe that there is already considerable knowledge and experience within this institution on which successful programs can be based. I believe that we have begun to move in the right direction. This Symposium is an important step along the way. Agriculture and Environmentally Sustainable Development Ismail Serageldind Introduction Mr. Preston has painted a broad-brush canvas of the challenges we face in the rest of this decade if we are to fulfill our developmental mission, to help reduce poverty, and to improve well-being for billions of unfortunate souls on this planet. Mr. Preston said: "We need to place the concept of sustainability at the center of our work." Because that is very much the raison de'tre of the new Environmentally Sustainable Development (ESD) Vice Presidency, let me start my remarks by discussing the concept of sustainability and how we can operationalize the concept. Defining Environmentally Sustainable Development For a first cut at an operational definition of sustainable development, we may begin with the Brundtland Commission's definition: 'Sustainable development is development that meets the needs of the present without compromising the ability offuture generations to meet their own needs. ' It contains within it two key concepts: * The concept of 'needs,' in particular the essential needs of the world's poor, to which overriding priority should be given * The idea of limitations imposed by the state of technology and social organization on the environment's ability to meet present needs.2 Or put another way: 'To say that a development path is 'sustainable' means, at least, that its patterns ofJproduction and consumption can be reproduced indefinitely without doing increasing or irreparable damage to essential natural ecosystems. 3 These two definitions are cautious in that they are static: they suggest holding the line on degradation of future opportunities but do not necessarily promise any hope for improvement. Challenges for Conventional Neoclassical Economic Analysis To be meaningful guides for the design of policies and investments, these definitions must be translated into the language of economic analysis. This is not an obvious task because mainstream neoclassical economics has had little to say about sustainability.4 In fact, one can argue that, at least until about a decade ago, it has been singularly shortsighted. Indeed, * While most of us readily recognize the obvious links between economic behavior and environmental quality, mainstream neoclassical economics treats these as externalities. The author is Vice President, Environmentally Sustainable Development, The World Bank. 6 * While most of us readily accept the idea that economic behavior both affects and is affected by changing technologies, time horizons, and values, tastes, and preferences (including those regarding family size), mainstream neoclassical economics continues to treat these as exogenous variables.5 So, where do we start? Perhaps the most obvious point is the concept of capital, and the need to sustain the capital base of humanity, including * Manmade capital, both physical and immaterial (such as information) * Natural capital, with proper valuation of the environmental services that natural resources provide (not just as inputs into productive processes) * Social capital, not only in the conventional terms of investing in people but also in terms of the institutional and cultural framework that makes civilized transactions possible. The disintegration of societal institutions and structures in Yugoslavia and Somalia are stark reminders of what extreme neglect of that aspect can lead to. (This links up with concerns of poverty reduction and equity.) Today, certainly in terms of natural capital, possibly in terms of social capital, we are depleting it rapidly. For example, tropical forests are being lost at the rate of 17 million hectares annually, 60 percent of that due to agriculture (shifting cultivation).6 This means that we are living off our capital, not just our income. Therefore, our children will have less capital from which to derive their income. Therefore, once we have begun to feel the impact of this capital depletion, and until the healing effects of a new path can take hold, there will be a downward trend. Even after the "right" policies and technologies are in place, it will take time to build up a capital stock compatible with a new way of doing things. Incidentally, this is not a revolutionary departure from conventional economic analysis, which has always maintained the distinction between capital and income. But there, capital (only manmade capital) was to be kept at a level to produce income in the next reporting period. This definition merely extends that one from manmade to natural and social capital.' I will not belabor the point here, but more generally, our economic accounting must adequately reflect these environmental and social concerns if the policies and the investments we recommend are to effectively promote sustainable development.8 Let me now move back to another part of the Brundtland definition -- the concept of needs. This undergirds the link between environmentally sustainable development and poverty reduction in a rapidly changing world. Environmentally Sustainable Development and Poverty Keeping the poor in their misery, while protecting the environment and promoting economic growth, is not sustainable development. We must strike at the roots of poverty. What is poverty? It is not just having low incomes. It is not just the absence of means of social participation. It is not just the persistence of massive inequities. It is not just deprivation of decent shelter, clothing, and diet; too frequently, it is much more than that. Too frequently it is oppression, hunger, sickness, and death-extreme conditions that defy our capacity to imagine them much less measure them. I could speak at length on the issues of poverty and the design of strategies to deal with it.9 But my purpose here today is simply to remind us all that doing our job right in this area is not 7 merely a matter of a few additional percentage points of income growth per year. It is literally a matter of life and death for millions of people. * Literal reminders of the terrible cost of failed policies shocked a complacent world a decade ago with images of starvation in Ethiopia. * That it happened at all was a tragedy. * That it happened again in Somalia was beyond belief. * And that it could still happen again elsewhere would be an unspeakable crime. But even if outright famine is averted, the problems of poverty are no less pernicious for the unpublicized way in which they claim, at the margin, hundreds of thousands of lives. How much of infant mortality and morbidity bred of ignorance and malnutrition could be averted by better attention to basic health and nutrition? By better attention to the meaning of poverty and the means of its alleviation?'" Our subject is indeed a matter of life or death for countless thousands whose silent, hopeful presence should permeate our discussions here. The Role of Agriculture Agriculture is the key to address these issues of * Reduction of poverty * Food security * Natural resource management * Environmental sustainability. In most parts of the world, poverty and malnutrition are still far more pronounced in rural areas than in urban areas. The bulk of the world's poor are in rural areas. Thus the percentage of the poor found in rural areas is: 91 percent in Indonesia, 79 percent in India, 67 percent in the Philippines, 86 percent in Cote d'Ivoire, and 52 percent in Peru." In the rural areas, the poor account for significant numbers of the total population:'2 60 percent in Latin America and the Caribbean, 60 percent in Sub-Saharan Africa, 31 percent in Asia, and 26 percent in the Middle East and North Africa. Between the late sixties and the late eighties, food staples self-sufficiency declined from 99 percent to 94 percent in Asia (excluding China and India), from 98 percent to 93 percent in Sub- Saharan Africa, from 92 percent to 76 percent in the Middle East and North Africa, and from 112 percent to 93 percent in Latin America and the Caribbean.'3 The fact that the title of the department concerned with agriculture no longer refers to rural development cannot imply that these poverty concerns have been reduced. An entire new vice presidency has been created to emphasize this point. Within the ESD vice presidency, many opportunities will arise for much better coordination of strategies for agricultural production, for conservation and development of the land and water resources available to support rural livelihoods, and for improved access of the rural poor to clean water, roads, and other infrastructure necessary for a productive and healthy life. We will also strive to improve coordination with our colleagues working on human resources to improve badly neglected rural education, population, health, and nutrition programs. You must take the lead in this cross-unit endeavor to reach out to the rural poor of the world. 8 Agriculture and Environment What about agriculture and the environment? Agriculture, in many ways, embodies the idea of interaction among people, land, water, and air (climate). That is the same immediately accessible conceptual framework for environment. Environmental activism was initially labeled "green." The kind of environmentalism we are concerned with is one that guarantees people the right to pure water, to clean air, and to fertile soil."4 In many parts of the world, these rights are being denied. Today: * 1 billion poor people are denied easy access to clean water. * 1.7 billion poor people live without adequate sanitation. (fhe diseases stemming from these conditions kill 2 to 3 million children every year.) * 1.3 billion people in the cities of the developing world -- many of whom are rural migrants fleeing pervasive rural poverty -- suffer from inhaling soot and smoke in the air at levels considered dangerous by the World Health Organization. * 700 million women and children are suffering from indoor air pollution from biomass burning stoves -- with health impacts equivalent to smoking three packs of cigarettes a day. - Hundreds of millions of poor farmers are suffering the cause of declining soil fertility -- not because they are greedily exploiting the land but because they simply do not have the resources to make the simple investments to protect and replenish the soils, their source of livelihood."5 These rights to pure water, to clear air, and to fertile soil have to be given to those groups who currently lack the power of voice: the poor and the generations to come. T'he key is empowering the poor, especially women. When poor farmers are given legal tenure or legally binding user rights to their land, rather than keep it as state-owned collective farms or open access,"6 they are more able (because they get access to credit) and willing to invest in protecting the soil. Examples abound: from hill farmers in Kenya, to slash-and-burn smallholders in Thailand, to pastoralists in Senegal and Burkina Faso. But more often, the needed empowerment relates to access to services and resources. * Access to education has a remarkably positive impact on incomes and the environment. The agricultural productivity effect of education is well documented. But there is much more. Providing education to slash-and-burn farmers in Northern Thailand was found to be a most powerful policy to reduce deforestation. Education of girls is especially important. Better educated mothers provide better nutrition in Brazil even if their income is the same. Education is the strongest form of empowerment. This is why the World Bank is providing over US$2 billion per year for education programs. Ninety percent of these education programs have special provisions for female education. * Other crucial forms of empowerment include access to extension services and credit. Again, providing these to the poor and to women has extraordinary economic human and environmental returns. Example: women manage half of the natural resources in Africa (70 percent of the farms in Congo), yet are often denied extension and credit services, education, and legal protection. No wonder investment in soil protection is inadequate! * The right to clean air and the absence of pollution requires a different form of empowerment. There must be accountable government responsive to the needs of all 9 citizens, and competent to put in place cost-effective policies and investments. This is difficult -- but very encouraging developments are taking place. The World Bank is working with about forty countries to strengthen environmental institutions, including major efforts in Chile, China, Indonesia, Mexico, Nigeria, and Poland. Beijing's industrial output has doubled in the past eight years, while its output of hazardous waste has halved. Malaysia, Mexico, and Thailand have introduced lead-free gasoline. Air pollution from particulates (soot and smoke) and from sulphur dioxide is improving on average in middle income countries. (But not enough and not in low income countries.)"' How can the rights offuture generations be protected? By protecting the natural capital stock so it can continue to provide incomes and services into the indefinite future. For example: * Make sure thatfish stocks are not depleted irreplaceably. (The Bank is supporting a new fishing law in Chile to ensure a sustainable harvest.) * Protect aquifers from irreversible depletion or pollution. (The Bank is assisting dozens of countries in improving water management.) C Protectforest ecosystems so they can provide economic, social, climatic, and ecological services for generations to come."8 (Forest protection and enhancement is at the core of the Bank's forest sector work around the world.) * Protect the fertility of the soils from depletion, erosion, and water-logging.'9 _ Protect the atmosphere from excessive accumulation of greenhouse gases and from ozone depletion. (The Bank chairs the Global Environment Facility` and is the major executing agency of the Montreal Protocol Fund.)2' While in many instances pro-poor development policies and projects also increase sustainability, as in the case of soil conservation via vegetative barriers and agroforestry programs, such is not always the case. Conflicts between development and conservation are likely to occur, at least in the short-term. Conflicts are embedded in most programs aimed at indigenous groups. Conflict arises in buffer zones of national parks. And there are potentially large conflicts if development of wetlands is to be restricted in areas with rapid population growth, because this is the main opportunity to add high quality land to agriculture. I believe that in such instances we must first check whether the poor would truly benefit from the proposed development, or whether the rich will reap the benefits. If the poor will reap significant benefits, then some compromise that does not sacrifice high-priority conservation objectives can be found. This will be an important area of intellectual research for the Development Economics Vice Presidential Unit (DEC) in the Bank and of policy analysis for ESD. The earlier comments I made about improving our economic analysis can and should be brought to bear on such issues. However, it is important to note that with 100 million people added to the planet's population each year, simply protecting the environment is not enough. We need to improve the productivity of all kinds of "capital" -- natural, manmade, and human. This is why it is essential to change the focus from the environment towards sustainable development.' The Promise of Technology Are we shortchanging the promise of technology, of a new "green revolution," while listening to the Cassandras who preach the "limits to growth" gospel? I do not think so. What I have sketched out are only prudent and responsible approaches to management of natural resources in a period of rapid 10 change and future uncertainty. You will be devoting some time in the next few days to this issue so I shall not expand on it here. But what about technology and what it can do for the future? In an excellent essay Neva Goodwin raises the question of the role of technology in helping humanity out of its apparent bind.3 "Malthusian pessimists" see a global system of diminishing marginal returns, which "technological optimists" are convinced will be overcome by new (and unspecified) technologies that will maintain the momentum of ever-greater productivity of all (classical) factors of production (land, capital, and labor). Both may be right! Goodwin suggests that the concept of technology be unbundled to differentiate between two types of technologies: * The mostly material inputs group of technological resources whose marginal returns are declining, most markedly in modern agricultural systems such as those spawned by the green revolution. These include chemicals, machinery, imported (that is, off- farm) energy, and other purchased material inputs. * The information-intensive, immaterial inputs group of technological resources that appear to retain significant potential to enhance the productivity, and reduce the intensity of use, of most or all other factors. Examples of such information-intensive agricultural technologies include: * Pest control strategies that employ natural interactions of plants and pests' * Crop rotation and diversity' i The selection and creation of improved animal and plant varieties (including modern techniques of bioengineering as well as older breeding methods)' * :Complex farming systems, including staple grains, legumes, fodder, and, importantly, livestock * On-farm reuse of organic materials, including composted plant byproducts and animal wastes * Agroforestry and other types of three-dimensional design to maximize use of sunlight in plant "layers" * "Fine-tuning" of inputs, for example, in timing as well as in quantity of applications of water and fertilizer. Many of these areas are not likely to be promoted by the private sector alone because they cannot be appropriated and marketed. A conscious government effort to promote and disseminate such approaches is needed.27 A Vision for the Future Without pinning too much hope on untested new technologies, but without dismissing their potential benefit, we can outline a vision for the future that builds on the concepts of ESD, namely, recognizing that: * Long-term sustainability is at the heart of any development strategy. * The linkages among poverty, environment, and agriculture are important. * The links between economic activity and environmental quality are powerful. * The key to effective change on the ground -- implementation -- resides in empowering the poor, especially women, to take effective action to improve their well-being. Embracing these concepts would lead us toward a new and more "holistic" vision of development.? A vision that would revolve around a broad development strategy whose elements include the following prescriptive items: 11 * People are both the means and the ends of the development process - human resource development and capacity building are priorities. 41 There is no alternative to continuous economic reform and adjustment. * Economic growth is necessary but not sufficient to bring about improvements in human well-being. ° Aggressive antipoverty policies must go hand in hand with pro-growth policies. * Increased agricultural productivity and food security are essential, as are diversification and competitiveness of exports. * An enabling environment -- political and legal -- for individual initiative and private enterprise is fundamental. * The most efficient use of scarce resources is imperative, including a review and reduction of military expenditures. H Empowerment of people through good governance and accountability is vital. * Adequate population policies must be pursued. * A larger role for women is essential. * Immediate action on environmental issues is critical. a Increasing economic integration is necessary. Short-term measures must be embedded in a long-term perspective. * Adequate external financing and imaginative treatments of outstanding debt are needed. Next Steps This group here today has a massive responsibility to make this lofty vision a reality. I have already highlighted the centrality of agriculture to the problems of poverty and environmental quality. It is also clear that in many countries -- especially Sub-Saharan Africa - the key to growth is effective modernization of agriculture. Yes, what you will be doing in the next few years will be decisive. How decisive? The World Bank currently administers a portfolio of 413 agricultural projects, representing US$26 billion in loans and US$60 billion in total investments. It is further planned that over the next five years, US$23 billion in new commitments will be made for this sector. But this is not all. Not only is the question of implementation essential - as Mr. Preston underlined in his opening remarks -- but the framework within which these investments take place is, and must be, a prime concern of yours. This will involve giving much greater emphasis in the Bank's policy dialogue and lending programs to a country's overall strategy for the sector. It will also require much greater attention to the quality of a country's entire expenditure for agricultural and natural resource development and for food assistance to the poor. It is not sufficient to focus only on the expenditures directly associated with Bank-assisted projects. One implication of such an approach is a far greater need for donor coordination. Another implication is that the policies advocated for agriculture must be compatible with -- and indeed part of -- national policies aimed at creating an enabling environment to empower the poor, spur investments, and promote private sector development. All this is very much in line with the proposed Board paper on agriculture, which you all have discussed at length.' * We must now move from words to action. * We must make use of the knowledge and experience we have gained. We must: 12 * Learn from our mistakes * Build upon our successes * Spread "best practice" approaches to different parts of the whole. The new ESD Vice Presidency enables us to do so better: * We can strengthen the links between the staffs of Agriculture, Environment, and infrastructure. * We can promote bridge-building with other thematic vice-presidencies. * We can help cross-fertilization across regions and across sectors. * We can provide the strongest and most effective support for the work of the regional staff to make this strategy a reality. Whatever we do at the management level, the challenge of the future is in our hands: * How can we best integrate the concerns of Agriculture and ESD into macro and sectoral strategies for our borrowers? * How can we develop responsible approaches to Natural Resource Management that will improve agricultural production and sustainability? * How can we improve our economic measurement and analysis to take into account ecological and social concerns? * How can we ensure that what is already known in terms of technology is effectively disseminated and used among the poor farmers of the world? * Do we have the right skills and knowledge and experience among the staff of this institution, or do we need to revisit the issues of skill mix and staffing in Agriculture? Your work will define the future. The Bank is committed to put up about US$4.5-$5 billion per year to back up your recommendations based on your analysis. Envoi The future is on our doorstep. What we do today and tomorrow will help define the world of the third millennium for hundreds of millions of people currently living in rural poverty, without voice or power, in a deteriorating and deleterious environment. We must bring to bear the rigorous discipline of analysis and the wisdom of vast and pragmatic experience in dealing with these issues. But I beseech you at all times not to forget the precarious reality of the human condition: * The vulnerability of unskilled labor * The soul-destroying impact of poverty and homelessness * The ease with which the rich and powerful subvert law enforcement to their own ends. Hard and complex as the issues are, there is a compelling and crushing reality out there: * Every passing day of misguided policies deepens the misery of wretched millions of human beings. * Every incomplete package adopted represents a lost opportunity to reach out to those millions of kindred souls. We cannot claim to have all the answers and must be humble about the scope of our possible interventions -- but we must dare to be bold, dare to be imaginative. With vision and commitment we can help empower the long-suffering rural people to take charge of their own destinies for, ultimately, progress lies in enabling the weak and the marginalized 13 to become not the beneficiaries of aid or the recipients of charity, but the producers of their own bounty and welfare. Endnotes 1 . Brundtland Commission (World Commission on Environment and Development, Jim MacNeil, principal architect and chief author), Our Common Future (New York: Oxford University Press, 1987). 2. Brundtland Commission (World Commission on Environment and Development, Jim MacNeil, principal architect and chief author), Our Common Future (New York: Oxford University Press, 1987), 43. 3. Frank Ackerman, "The Natural Interest Rate of the Forest: Macroeconomic Requirements for Sustainable Development" (Unpublished manuscript, April 22, 1992), cited in Neva Goodwin, "On a Period of Global Transitions: Issues of Sustainable Development" (Paper presented at the Roundtable on Global Change, United Nations Development Programme, Development Study Programme, 1992). 4. Some notable exceptions include the pioneering work of A.C. Pigou on social welfare and the handling of externalities, The Economics of Welfare, Ist ed. (London: Macmillan and Co., 1920); H. Hotelling's 1931 article, "The Economics of Exhaustible Resources," Journal of Political Economy 39 (1931): 137-75; S.V. Ciriacy-Wantrup's influential 1952 book, Resource Conservation: Economics and Policy, which introduced the concept of the safe minimum standard (Berkeley: University of California Press, 1952); and John Krutilla's 1967 classic article on renewable resource issues, "Conservation Reconsidered," American Economic Review 57 (1967): 787-96. 5. See among others Herman Daly and John Cobb, For the Common Good: Redirecting the Economy toward Community, the Environment, and a Sustainable Future (Boston: Beacon Press, 1989); and Herman Daly, "From Empty-World to Full-World Economics: Recognizing an Historical Turning Point in Economic Development," in Robert Goodland, Herman Daly, and Salah El-Serafy, eds., "Environmentally Sustainable Development: Building on Brundtland," World Bank Environment Department Working Paper No. 46 (Washington, D.C.: World Bank, 1991). 6. Narendra Sharma, Managing the World's Forests: Looking for Balance between Conservation and Development (Dubuque, Iowa: Kendall/Hunt Publishing Co., 1992). 7. See World Bank Sustainability Group with Johan Holmberg, Consultant, "Operationalizing Sustainable Development" (Unpublished memorandum, April 24, 1992), 5-6. 8. Jan Tinbergen and Roefie Hueting, "GNP and Market Prices: Wrong Signals for Sustainable Economic Success That Mask Environmental Destruction," in Population, Technology, and Lifestyle: The Transition to Sustainability, ed. Robert Goodland, Herman E. Daly, and Salah El 14 Serafy (Washington, D.C.: Island Press, 1992), 52; and Salah El Serafy, "Sustainability, Income Measurement, and Growth," in Population, Technology, and Lifestyle: 7he Transition to Sustainability, ed. Robert Goodland, Herman E. Daly, and Salah El Serafy (Washington, D.C.: Island Press, 1992), 63. 9. World Bank, World Development Report 1990: Poverty (New York: Oxford University Press for the World Bank, 1990); Ismail Serageldin, Poverty, Adjustment and Growth in Africa (Washington, D.C.: World Bank, April 1989); and Ismail Serageldin and Michel Noel, "Tackling the Social Dimensions of Adjustment in Africa," Finance & Development (September 1990): 18-20. 10. Ismail Serageldin, "For the Future of Africa," Development: Journal of the Society for International Development (Issue on "Human-Centered Economics: Environment and Global Sustainability") (1990: 3/4): 147-51. 11. World Bank, World Development Report 1990: Poverty (New York: Oxford University Press for the World Bank, 1990); and I. Jazairy, M. Alamgir, and T. Panuccio, 7he State of World Rural Poverty: An Inquiry into Its Causes and Consequences (Rome: International Fund for Agricultural Development, 1991). 12. World Bank, World Development Report 1990: Poverty (New York: Oxford University Press for the World Bank, 1990); and I. Jazairy, M. Alamgir, and T. Panuccio, The State of World Rural Poverty.- An Inquiry into Its Causes and Consequences (Rome: International Fund for Agricultural Development, 1991). 13. The number of food deficit countries in Sub-Saharan Africa almost doubled from 28 to 41 between 1965-67 and 1986-88. The IFAD study of 108 developing countries found that the proportion of countries whose per capita energy supply fell below requirements has declined significantly in all regions. Between 1965 and 1985, of the 23 developing countries in Asia, the number whose per capita energy supply fell below requirements declined from 21 to 9 countries. In Sub-Saharan Africa, the number declined from 33 to 30 countries. In the Near East and Africa, it declined from 9 to 2 countries while in the Latin American and Caribbean countries, it declined from 19 to 10 countries. 14. World Bank, World Development Report 1992: Development and the Environment (New York: Oxford University Press for the World Bank, 1992). 15. See Paul Harrison, The Greeninig of Africa: Breaking through in the Battle for Land and Food, an International Institute for Environment and Development/Earthscan Study (New York: Penguin Books, 1987); and J. Patrick Madden and Thomas L. Dobbs, "The Role of Economics in Achieving Low-Input Farming Systems," in Sustainable Agricultural Systems, ed. Clive A. Edwards, Rattan Lal, Patrick Madden, Robert H. Miller, and Gar House (Ankeny, Iowa: Soil and Water Conservation Society, 1990). 16. This issue of land tenure and land user rights should be looked at carefully, especially where communal (as opposed to state) lands exist. See Kevin Cleaver and Gotz Schreiber, The Population, Agriculture, and Environment Nexus in Sub-Saharan Africa (Washington, D.C.: 15 Agriculture and Rural Development Series No. 1, Africa Technical Department, World Bank, 1992), 30-35, 111-20. 17. World Bank, World Development Report 1992: Development and the Environment (New York: Oxford University Press for the World Bank, 1992). 18. Narendra P. Sharma, Managing the World's Forests: Looking for Balance between Conservation and Development (Dubuque, Iowa: Kendall/Hunt Publishing Co., 1992); and Ismail Serageldin, Saving Africa's Rainforests (Washington, D.C.: World Bank, 1991). 19. World Bank, World Development Report 1992: Development and the Environment (New York: Oxford University Press for the World Bank, 1992). See also Clive A. Edwards, Rattan Lal, Patrick Madden, Robert H. Miller, and Gar House, ed., Sustainable Agricultural Systems (Ankeny, Iowa: Soil and Water Conservation Society, 1990). 20. The Bank is the Administrator of the GEF, the repository of its two trust funds, and undertakes investment projects. For the official delineation of the Bank's role in the GEF see "Procedural Arrangements among the International Bank for Reconstruction and Development, the United Nations Environment Programme, and the United Nations Development Programme for Operational Cooperation under the Global Environment Facility," signed on October 28, 1991. 21. The Ozone Projects Trust Fund, or Montreal Protocol Fund, provides grants and concessional loans to developing countries for environmental clean-up. For the official designation of the Bank's responsibilities see "Procedural Arrangements among the International Bank for Reconstruction and Development, the United Nations Environment Programme, and the United Nations Development Programme for Co-Operation and Assistance in Protecting the Ozone Layer in the Context of the Vienna Convention for the Protection of the Ozone Layer and Its Montreal Protocol on Substances That Deplete the Ozone Layer," signed on June 19, 1991. 22. Randolph Barker and Duane Chapman, "The Economics of Sustainable Agricultural Systems in Developing Countries," Sustainable Agricultural Systems, ed. Clive A. Edwards and others (Ankeny, Iowa: Soil and Water Conservation Society, 1990). 23. Neva Goodwin, "On a Period of Global Transitions: Issues of Sustainable Development" (Paper presented at the Roundtable on Global Change, United Nations Development Programme, Development Study Programme, 1992). 24. See Sam Engelstad, William M. Coli, and Jeffery L. Carlson, ed., Innovations in Pest Management: Proceedings from an International Forum on Integrated Pest Management, Biological Controls, and Other New Approaches to Controlling Pests in Our Environment, forum sponsored by the Massachusetts Department of Food and Agriculture; the University of Massachusetts, Cooperative Extension; and the World Bank, March 7-8, 1988 (Sturbridge, Mass.: University of Massachusetts, Cooperative Extension, 1988); and Agnes Kiss and Frans Meerman, Integrated Pest Management and African Agriculture, World Bank Technical Paper No. 142 (Washington, D.C.: World Bank, 1991). 25. See C. A. Francis and M. D. Clegg, "Crop Rotations in Sustainable Agricultural Systems"; and Benjamin R. Stinner and John M. Blair, "Ecological and Agronomic Characteristics of Innovative 16 Cropping Systems," both in Sustainable Agricultural Systems, ed. Clive A. Edwards and others (Ankeny, Iowa.: Soil and Water Conservation Society, 1990). 26. Holly Hauptli and others, "Biotechnology and Crop Breeding for Sustainable Development," in Sustainable Agricultural Systems, ed. Clive A. Edwards and others (Ankeny, Iowa.: Soil and Water Conservation Society, 1990). 27. For all these issues see National Research Council, Board on Agriculture, Committee on the Role of Alternative Farming Methods in Modern Production Agriculture, Alternative Agriculture (Washington, D.C.: National Academy Press, 1989). 28. Ismail Serageldin, "The Challenge of a Holistic Vision: Culture, Empowerment, and the Development Paradigm," Culture and Development in Africa (Proceedings of an International Conference), interim document, ed. Ismail Serageldin and June Taboroff (Washington, D.C.: Africa Technical Department, World Bank, 1992), 11-28. 29. Agriculture and Natural Resources Department, "Agricultural Sector Review Paper" (Unpublished Board paper, February 1993). Technical Considerations for Sustainable Agriculture Richard G. Grimshaw, Christopher J. Perry, and James Smyle* For the purpose of this paper a sustainable system of agriculture is one that can maintain production at a sustained and profitable level, if necessary with moderate levels of support from external inputs, without leading to significant environmental damage. To achieve sustainability most agricultural production systems depend on a policy and institutional environment (including pricing, regulation, and tenure) that is conducive to the producer; infrastructure (both physical and support services) that allows effective access to services and the market; and four critical technical criteria -- nutrition, health, genetic material, and water. This paper addresses these technical criteria. In recent years the World Bank has rightly come under criticism because some projects have not performed well. In looking for leads to improve that performance, it is worth reviewing some successful projects and schemes developed by agencies other than the Bank, examples include: * Global 2000's project in Ghana that emphasizes soil fertility and the introduction of improved varieties, primarily maize, and other inputs, clearly demonstrates that African farmers are responsive to inputs. = SIDA's Machakos Integrated Development Project in Kenya that gave priority to increasing arable area and unit area productivity through improvements in soil fertility, land management, and water conservation. * FAO's integrated pest mangement project for the control of Brown Plant Hopper in rice in Indonesia. WIMCO Ltd's introduction of high-yielding poplar clones to India's northwest farmers that has led to a highly successful venture in private sector wood production. _ The World Food Program's successful soil and water conservation program in the Loess Plateau Region of China that reduces sediment flows in the Yellow River, expands cultivable land, and improves soil fertility, soil moisture, and crop production. 3 CARE International's Dry Land Farming projects on the outer islands of Indonesia that focus entirely on water and soil moisture conservation technologies resulting in important land-use change from degrading, near monocropped cassava land, to lush perennial farm gardens supporting cocoa, coffee, bananas, coconut, and so forth, on a sustainable basis. These projects have a common feature -- they all focus on a narrow band of technologies that are fundamental to agricultural growth and the sustainability of the agricultural resource base. In the discussion of technical considerations for sustainable agriculture it is worth dwelling on development strategies of the past. Two thousand years ago in China it was imperative for the Han government to increase agricultural production (and thus its tax receipts), for not only was the population expanding, but expensive wars were being fought. An intensive campaign was therefore mounted both to improve agricultural methods and to expand the agricultural area. The campaign was specifically designed to benefit smallholders, and to improve peasant agriculture by providing the necessary technical inputs and physical infrastructure. Wu-Ti was the first emperor of unified China Richard Grimshaw is Agricultural Adviser in Asia's Technical Department, Chris Perry is an irrigation economist at the New Delhi Office, and Jim Smyle is a watershed hydrologist in Asia's Technical Department. 18 to realize the importance of water control, and he carried out an enormous program of canal building... that irrigated over a million acres of arable land. The government also provided peasant farmers with seed-grain, tools and draught animals, sometimes on credit, sometimes on loan, and sometimes as outright gifts (Needham 1984). It is interesting to note that China's approach to agricultural development hasn't changed much since Emperor Wu-Ti's time, and today China has the world's fastest sustained growth of agriculture averaging 4.5 percent since 1950, and about 6.5 percent per annum since 1980 following market reforms and return to privatization of production. It is estimated that research generated technology has, since 1965, contributed to 20 percent of this production growth (Fan and Pardey 1992). In the history of agriculture, sustained agricultural growth has always been associated with new technology introductions that increase factor productivity. At this time when rural populations are increasing in relation to virtually static land resources, and when more is expected of agriculture in the alleviation of rural poverty, there is a need to take a new look at those four critical areas of technology described in the opening statement of this paper. Nutrition -- Soil Fertility The failure to apply adequate afnounts and/or balanced levels of nutrients leads to loss of actual and potential production and income. In addition an increasingly occurring phenomena, especially in irrigated areas, is the misuse of fertilizers leading to ground and surface water contamination from phosphates and nitrates. There are no magic solutions to maintaining soil fertility -- soil nutrients removed in the form of crops, residues, trees, and livestock and livestock products have to be replaced. The farmer is looking for low-cost and efficient replacement. That is why farmyard manure, and in some countries human manure, when suitably composted, continues to be an important replacer of exported soil nutrients. Various green manures, including the leaves of leguminous hedge plants, that have other uses as well, are also becoming more widely applied. In rainfed areas where production risks are higher and the opportunity cost of labor is low, farmers are reluctant to use purchased inorganic fertilizer. The need to produce and use organic manures under such conditions is important. High-yielding forage (such as napier grass) production for "resident" livestock, particularly draft animals, is probably the key to maintaining a sustainable fertility cycle. Where such practices are used, as in Machakos District of Kenya, livestock can be supported on small areas of land, typified by the average smallholder farm that rarely exceeds one hectare. The contribution of livestock, as part of a small farm system is essential for draft power and for manure (Tiffen 1992). In China, pig manure has a similar key role in maintaining soil fertility. It recycles soil nutrients eroded from the uplands that support lowland crop (rice) production, the byproducts of which are recycled to pigs, as well as for mulch on upland farms. The ammoniation of rice straw, a technology being widely used on Bank projects in China, enhances its feed value and can undoubtedly contribute indirectly to increased soil fertility. Soil fertility issues, particularly in the tropics, are of importance to rural people, especially for those associated with approximately 1.7 billion hectares of acid tropical soils in seventy-two developing countries. These soils, commonly yellow or red in color, are often characterized by high acidity, aluminum toxicity, and, when mismanaged, high levels of degradation. Reducing acidity through deep liming; using aluminum tolerant plants; the introduction of mycorrhiza that increase the uptake of soil nutrients, particularly phosphates (the inoculation of tree seedlings with mycorrhiza greatly enhances tree growth on acid degrading soils) (Malajczuk, Jones, and Neely 1992); using a better balance of NPK; all add to increasing crop yields. 19 Currently there is a mass of investigations relating to tropical soils, including outstanding work by the University of North Carolina's TropSoils in association and collaboration with other international research groups such as IBSRAM and ICRAF, as well as national research agencies. A good summary of this work is found in TropSoils Technical Report 1986/87 (1989). A detailed investigation at Sitiung, Indonesia, is described in a recently published TropSoils paper (Arya, and others 1992), and establishes rather nicely the soil structure/moisture/fertility relationships under the conditions at that site. It confirms the difficulty of establishing shallow rooted food crops on highly acid, aluminum toxic soils, and demonstrates why deep rooted trees, shrubs, and grasses are successful. It recommends breeding and selection of crops that can grow in aluminum toxic soils, and the deep application of lime if food crops are to be successful. For example, Amaranth as a grain and fodder crop, and Vetiver grass for soil conservation, fodder, and mulch, both deep rooting and adapted to aluminum toxic soils, may be useful plants for this type of condition. Most country policies and programs relating to soil nutrients are deficient. In areas of high soil acidity liming programs are rudimentary, and often nonexistent; fertilizers are rarely matched to soil type or crop demands; farmers take what fertilizer is offered -- the offering often being determined by inappropriate national fiscal and foreign exchange policies, as well as inadequate distribution systems. For example, because phosphates and potash are generally imported, in contrast to locally manufactured nitrogen, farmers often receive unbalanced, nitrogen dominant, fertilizer. Additionally micronutrients such as sulphur, copper, boron, and so on, are generally not taken account of, with resultant loss of potential yields. Table 1 summarizes the yield response and potassium (K) balance in corn-wheat-corn cropping system at an experimental site in north China to different levels and types of nutrients (Lin, Ji-Yun, and Dowdle 1989) and demonstrates the incremental physical gains from a balanced fertilizer approach on these soils. Data from China's National Network of Chemical Fertilizer Experiments revealed that on the basis of field trials conducted during 1981-83, 74 percent of China's cultivated land was deficient in phosphate (available P less than 10 ppm), about 40 percent was severely deficient in P (available P less than 5 ppm), and about 23 percent was deficient in potash (available K less than 70 ppm). This scenario of imbalance is, outside of the large alluvial plains and deltas, typical of most developing countries, and is one of the reasons for stagnating yields, poor quality crops, increasing incidences of certain diseases, and soil degradation. It should be corrected through wider soil nutrient analysis, as well as appropriate actions including policy changes to modify fertilizer subsidies, and investment actions that promote better fertilizer production mixes, distribution, and technical support to farmers in relation to fertilizer application (introduction of low cost seed/fertilizer drills where appropriate and general soil management practices including a greater use of organic manures). With few exceptions, organic manures alone will not support the yield increases in crop production that are required to meet grain demands and enhanced farm incomes. To achieve sustained agricultural growth where land is limited there is no option but to use inorganic fertilizers. Chinese data show close correlation to the use of inorganic fertilizer and increased grain production as in the following graph (Lin, Ji-Yun, and Dowdle 1989). In China over the past twenty-five years the use of inorganic fertilizers has increased twelve times, as compared to only a 30 percent increase in organic manure (180 kilograms per hectare application rate) (Fan and Pardey 1992). The Bank should pay much more attention to soil fertility; detailed country by country subsector analysis is required to assess production foregone due to inappropriate fertilizer strategies and policies, and analyze the economics of doing things right. Annual budget savings from reduction in fertilizer subsidies could be reinvested in, for example, improved soil testing facilities, decentralized fertilizer blending and packaging plants, and improved technical advice. 20 Table 1: Yield Response and K Balance in Corn-Wlzeat-Corn Cropping System K K Year Crop Treatment' Yield Increase (Applied) (Removal) Balance Kg/ha % kg K/ha kg Klha kg Klha 1st Crop 1987 Spring NP 4,546 - 0 32.4 -32.4 Corn NPK 7,588 66.9 93.4 69.9 +23.5 NPM 6,235 37.1 21.0 42.7 -21.7 NPKM 7,921 74.2 114.4 72.1 +42.3 2nd Crop 1987/88 Winter NP 2,665 - 0 49.5 -49.5 Wheat NPK 3,466 30.1 93.4 79.5 +13.9 NPM 2,799 5.0 14.4 54.4 -40.0 NPKM 3,750 40.7 107.8 86.7 +21.1 3rd Crop 1988 Summer NP 1,948 - 0 24.0 -24.0 Corn NPK 4,471 129.5 93.4 95.1 - 1.7 NPM 3,154 61.9 0 34.9 -34.9 NPKM2 5,179 165.8 93.4 89.3 + 4.1 Data adapted from Jin Ji-yun, 1989. For NPM, NPKM treatment, cattle compost manure (M) was applied at rate of 13.5 t/ha dry weight. 2 No manure applied for this crop. Figure 1. Grain Production and Fertilizer Consumption in China 1952 to 1987 420- 400 -19.5 Grain Production 380 -18.0 >---Fertilizer / 360 -16.5 Consumption // 340- -15. E, _ -300- 12.0 ' / .=280- -10.5 -, 260-- 9.0 / °240-- 7.5 _/ ._ 220-- 6.0 c 200-- 4.5 , 180-- 3 160- -.5 - F i I I I I I i 1 50 56 62 68 74 80 8687 Year 21 Health -- Crop Protection As crop yields increase, and cash crops (particularly fruit and vegetables) are introduced, so does the use of protective chemicals. This is leading in some cases to: overuse of chemicals; and increasing insect and plant resistance to insecticides and herbicides, as for example the cotton bollworm in China. The cotton bollworm has four generations every year. The second and third generations are most destructive in the Yellow River cotton belt. Cotton bollworm developed resistance to DDT due to excessive application over a long period of time (1950s-1970s). Pyrethroids were introduced for control in 1982; by 1987 the cotton bollworm had developed high levels of resistance to pyrethroid insecticides. Dependency on chemicals can be reduced through such management practices as Integrated Pest Management (IPM). The successful Indonesian/FAO IPM program for the control of Brown Plant Hopper in rice is just one such example where spraying regimes have been reduced by two-thirds, which together with yield increases has resulted in incremental income of US$100 per hectare per season (World Bank 1992a). As a direct result of this program, pesticide subsidies were removed, saving US$150 million a year, and extension methods were reversed from a top down to bottom up approach, with farmers fully participating in technology transfer and research. In Brazil, IPM trained farmers have reduced their insecticide applications on soybean, from nearly six to less than one on average, with no loss of yield. The introduction of disease resistant soybean varieties has virtually eliminated the need for fungicide. IPM is an approach that requires a range of appropriate technologies (tools) so that the mix can be adjusted according to the perception of the problem. It is not an easy option to implement in practice. It often needs time, money, and research. Most importantly, it requires knowledge and understanding of pests. Most Bank projects are large and complex and are unable to give proper treatment to plant protection and related issues. There is a need to review the overall plant protection subsector, assess critical problem areas, review alternative pest control methods, develop effective regulatory systems, and use savings in pest control subsidies for directed research activities that can support targeted IPM programs and for better communications with farmers. Currently the Bank is attempting this approach in Indonesia, and with modification, it might be a model for other countries as well. Germplasm -- Crop Varieties, Seeds, and Planting Materials The Green Revolution was a result of the breeding of high yielding varieties (H4YV) of rice and wheat that were introduced to the irrigated areas, and, because they were self pollinated and were responsive to management, they spread rapidly from farmer to farmer with minimum support from extension (Antholt 1990). The challenge of the future will be the maintenance as well as the improvement of HYV germplasm for irrigated agriculture (as is being undertaken in China's hybrid rice program that has increased yields from 4.8 tons per hectare to 6.5 tons per hectare); and the introduction of high quality plant material, and its dissemination to enable improved levels of production in rainfed areas and to support crop and livestock diversification. The majority of perennial species planting stock (including in some instances industrial crops such as rubber and oil palm) supplied to many farmers today are generally not genetically optimal. Neither is sufficient attention or rigor given to production of top quality seeds nor to culling of substandard perennial plants in the nursery. As a result, farmers forego annual income and make long-term investments in perennials that may only achieve a fraction of their full potential. It is important that special emphasis is given to addressing the technical issues involved. 22 Some are complex using modem biotech approaches; on the other hand many are simple and involve primarily low-cost management techniques such as the selection of good seed or clonal material, the application of known nursery technologies, and improved field establishment techniques. Development planners and researchers need to revisit past experiences. Farmer demand for crop varieties is often very different to that perceived by the scientist. In Machakos, Kenya, supposedly a semiarid region, with an average rainfall probability of 600 millimeters, farmers showed little interest in improved varieties of traditional drought tolerant crops such as sorghum and millet, but preferred (in response to market driven forces), and were able, through adapting their farm system with the aid of effective conservation systems, to successfully grow higher value cash crops that included maize, fruit trees, coffee, vegetables, cotton, sisal, and high-yielding fodder for livestock (Tiffen, Mbogoh, and Ackello-Ogutu 1992). The introduction of higher value crops led to better livestock control and management, and perhaps may be a key factor in the successful land stabilization that occurred in the district. Water - Soil Moisture Last of the critical criteria, and the most important is soil moisture. If soil moisture is inadequate the other three criteria can never contribute optimally to a sustainable agriculture system. It is an indisputable fact that without adequate soil moisture agriculture is not sustainable. As one of the oldest and most successful agricultural economies, China has depended on technology innovation (Needham 1984). In particular Chinese agriculture has given great attention to soil fertility and water for crop growth. Irrigation systems, still in use today, were developed over 2,000 years ago, as was soil moisture conservation, including unique soil mulching and terracing systems (in China terraced fields are known as 'three fold conservers," that is, they prevent erosion and conserve nutrients and moisture). On the semiarid Loess Plateau, where terraces have been used for 3,000 years, grain yields on nonirrigated terraces are 200 to 400 percent higher than those on unterraced hillsides, while in times of drought they may be even ten times as high (Needham 1984). Seed placement (the famous Chinese official, Chao Kuo, introduced the seed drill to northern China in 89 BC with one objective to maximize production from available soil moisture) has been an important part of Chinese agriculture. China had the equivalent of this century's "Green Revolution" during the Sung Dynasty (AD 1012) with the introduction of new varieties of quick ripening rices in irrigated areas of the Yangste Delta. Seeds, together with written instructions, were distributed to master farmers (today's equivalent of village farmer technicians). At the same time land improvement capital was made available to farmers, and major improvements were made in flood control, drainage, and irrigation distribution systems (Needham 1984). Likewise in Bali and the Philippines unique terracing for rice has been developed over the millennia. Today's Temple-managed proportional distribution systems of Bali are renowned for their simplicity and effectiveness. Crucial to these developments were farmer involvement in innovative measures both in farmer managed research and irrigation, terracing, and other developments. Such capability exists today when farmers are given the opportunity and encouragement. Irrigation The Bank is currently wrestling with its "Water Policy Paper." There has been much debate, and no doubt much more in the future, but clearly institutional reform that addresses a more comprehensive 23 approach to water, its allocation, price, and management is necessary if the irrigation sector is to perform its critical function in the long term. The large water carriers, as found in a number of Asian countries, not only provide vital irrigation water that results in 70 percent of Asia's food production, but also supply water to meet the ever increasing demands of towns and cities. Similarly large carriers in the Middle East, Central Asia, and the western United States serve both irrigation and municipal and industrial (M&I) use and face similar crisis. Asia Region's recent Asia Water Resources Study (World Bank 1992b) shows the growing demands for water for nonirrigation purposes. In some areas, such as Haryana in India, as much as 30 percent of current supplies to irrigation will be diverted to M&I in the foreseeable future, while in Madhya Pradesh, a substantial proportion of carry over storage in many dams is now preempted on a regular basis for M&I needs. It is not only important to know when these reallocations will be necessary, but also what investments and improvements have to be made in current use of agricultural water so that production levels can be sustained or even improved. The following paragraphs describe just two of many areas where technology and technology choices are important for Asian irrigation. While the objective of most Bank-financed irrigation projects in India is to assure reliable and predictable deliveries to all farmers, the reality is that usually few farmers take water at will, while others get little or none. Irrigation infrastructure (particularly the smaller canals, gates, and other structures) are seen at best as inadequate, and at worst an obstruction to farmers realizing individual needs, and are frequently 'modified,' or simply destroyed, by the project beneficiaries. An early attempt to directly address such performance failures through technology innovation was the Uttar Pradesh (UP) Public Tubewell Project in India. Well operators of existing traditional well designs were able to manipulate supplies to preferred customers -- water was not reaching tail enders due to preference of upper reach farmers and seepage from the farm ditches. The proposed new design automated the well operation, stabilized discharges in the distribution system, and made operators redundant. In a subsequent development the Government of West Bengal has, with Bank support, introduced a highly successful system of making public investment in ground water a cooperatively managed and maintained operation. Simple technology combined with careful attention to defining the responsibilities of the farmers and the irrigation department has yielded immediate benefits (200 percent plus irrigation intensity in the first year). The importance of matching technology to environment (social, technical, and economic) is clearly demonstrated by the different experiences of Uttar Pradesh and West Bengal. A great deal of attention has been devoted to the improved operation of large surface systems. Technology can make significant contributions. A common flaw in the design of surface systems is to plan complex management with very limited infrastructure. The infrastructure (even when not subjected to farmer modification) is inadequate to give steady flows to relatively large contiguous areas, let alone varying flows to small individual plots. The technical challenge is to achieve a match between the plan of operation, and the physical performance of the system. This can be achieved by reducing the complexity of the plan, improving the performance of infrastructure, or both. In practice a combination of both approaches has been successful, based on careful diagnosis and definition of those parts of the system which the authorities must operate (for example, the head regulator at the dam), those parts of the system that the farmers must be responsible for (distribution among individual fields), and those parts of the system which can be designed to be self managing (for example, proportional dividers, which provide planned allocations between channels without operator intervention, and long crested weirs, which maintain flow levels independently of flow rates). In each case the choice of technology is critical to success. 24 The simple intervention of ensuring automatic distribution (as has been done in Bali for thousands of years) of water in proportion to area - reducing the load on the authorities, and consequently increasing the reliability of supplies to farmers - has shown remarkable success. One project, funded under the Bank's National Water Management Project (NWMP) in southern India has shown productivity increases per unit of water of 30 percent over a three-year period. Water now reaches the tail ends, because headenders are more confident that future delivery schedules will be met (and peer pressure operates far more effectively when the supply is clearly defined). Increases in area irrigated for project and nonproject distributaries are shown in figure 2. Figure 2. Comparison of Percentage Increases in Irrigated Areas Between Improved and Unimproved Mangement of Distribution Systems % INCREASE IN AREA IRRIGATED 40 3 5 3 0 2 5 2 0 UKharif 1 5 0Rabi 10 5n 0 5~~~~~~~~~~~~~~~~~~~~~~~~ --NWMP DISTRIBUTARIES--- ---- NON-NWMP DISTRIBUTARIES--- 25 A point that should be stressed is that without careful analysis and discussions with farmers, and a full understanding of the broad range of technical considerations from hydrology to distribution system design, the results would not have been possible. These examples show that a careful match between operational objectives and technology is a key element of successful irrigation. Generally, technology exists within or is external to developing countries; a critical issue to be addressed is to improve application and transfer of technologies to countries such as India and China. For example while a large number of computer-based systems for hydrological analysis are well proven and accepted throughout much of the world, none is widely available to water planners and operators in India. Thus flood analysis, as well as real time operations, continues to be done by hand, or using locally developed rules of thumb. In Situ Moisture Conservation Farmers who do not have access to surface and groundwater are dependent on rainfall for successful farm production. Seventy percent of India's cultivated land is rainfed, as is 50 percent of China's. All but 3 percent of African agriculture relies on rainfall only. The dependency on rainfed agriculture is growing as potentially new irrigation land becomes scarce, and water supplies for irrigation are restricted by availability. Rainfed farmers are cultivating marginal lands that are often classified as nonarable. They will continue to cultivate them until they are either destroyed or a system of management is established that will assure sustainability. In most countries with high rural populations, unless sustainable systems are established there will be little hope in preserving forested lands. In western Java and some of the smaller islands of Indonesia, land is being cultivated from the bottom to the top of hills and mountains that have slopes often exceeding 25 degrees. Forest has been cleared and has been replaced by a system dominated by cassava and as a consequence erosion is extreme -- up to 400 tons per hectare. Drought is often a reflection not of lack of total annual rainfall, but of within-season shortages and the failure to conserve rainfall in situ. Asia Region's study "Watershed Development in Asia" (Doolette and Magrath 1990) reviews research studies relating to soil and water conservation, and links soil conservation closely with water conservation. For at least 3,000 years, farmers from ancient Mesopotamia to China have resorted to bench terracing with a high degree of success, particularly in respect to moisture conservation, whether in the wet tropics or in semiarid areas. Bench terracing has induced dramatic changes in societies and in the economics of their agriculture. Subsequently the technology has fallen victim to economic and social changes. A recent case of the rapid degeneration is the Yemen's "Hanging Gardens of Arabia" because of the out migration of labor (responsible for terrace maintenance) to the oil rich Gulf states. In other countries such as Kenya, India, Indonesia, south China, Thailand, and elsewhere terracing is becoming less effective and more difficult to introduce because of the high cost of labor, the increased subdivision of land that makes the planning of engineered structures extremely difficult, poor maintenance, and neglect. In some countries such as north China it is still possible to mobilize the "masses" to create new terraces as on the Loess Plateau; elsewhere, such as in Indonesia, large financial incentives (US$300 per hectare) (FAO/World Bank Cooperative Program 1992) have to be given to farmers to assure terrace construction. As a large-scale intervention, terracing is therefore generally not appropriate; yet too often, it is still viewed as such and is promoted as a key conservation activity with poor results. Clearly there are technology options that can be applied to enhance soil moisture conservation. First, such techniques as traditional mulching (rice straw in particular is a very 26 effective mulch and has a positive impact on soil structure), stubble mulching, and contour cultivation are often easily applicable at minimum cost, and all are very effective for in situ soil conservation. Compared to bench terracing, cultural and vegetative systems are cheap, involve minimum movement of top soil, and can be carried out by the farmer as a part of normal farm management practice. However, vegetative systems are not always able to survive the stress of uncontrolled high density livestock grazing as in the drier areas of India and Africa. All systems of conservation, some more than others, need a level of management to assure long-term sustainability. As a substitute for constructed terraces, vegetative barriers have been developed as important conservation technologies. As a key approach a wide range of trees, shrubs, and grasses can be utilized as hedgerows, as long as the farmers are prepared to manage them. On very steep slopes a small farm may require one kilometer or more of hedgerow per hectare to provide adequate protection. Potentials for high labor inputs; sensitivity to grazing; occasional drought, fire, pests, and diseases; reduced yields from crop area reduction and hedgerow/crop competion; and disruption from normal farming practices all led to hegerows being a technological approach whose successful introduction requires a level of flexibility and support far beyond project or agency capability. This is particularly true where hedgerows are not an indigenous technology on which initial efforts can be based. In the mid-1980s John Greenfield, of the Bank's New Delhi staff, identified a grass whose history of traditional use in south India seemed to indicate a good potential for its application as a hedgerow species. Subsequent investigations found the species to be pan tropical both in its distribution and its traditional use as a conservation plant. This grass, Vetiveria zizanioides (Vetiver grass), displayed a number of characteristics which avoided many of the problems (mentioned above) associated with hedgerow introduction and management. The grass and its application are described in a number of papers (Grimshaw 1991; Smyle and Magrath 1990). Further experience showed the species' potential for extending hedgerows as a key technology. Because Bank staff's early work with trial, demonstration, and dissemination of this technology, follow on research by IARCs, universities, and government agencies has confirmed Vetiver's utility for soil and moisture conservation. Interest by individuals, nongovernmental organizations (NGOs), governments, and most recently in a publication by the U.S. National Academy of Science (1992) confirms the widespread applicability and potential of the species. In reference to the latter, a scientific audit was carried out by Dr. Norman Borlaug primarily to confirm World Bank staffs' contentions regarding Vetiver as a key technology. By utilizing its technical expertise, Bank efforts resulted in the identification, verification, and dissemination of a new technology within a short span of eight years (it is generally recognized that the introduction of a new cereal variety can take twelve years or more). A comparison, by Bharad and Bathal (1990) of GKVK University in Akola, India, between contour cultivation, graded bunds, tree (Leuceana sp) hedges, and grass (Vetiver) hedges, showed, respectively rainfall runoff percentages as follows 26, 24, 22, and 17. More recently work at ICRISAT (Rao, Cogle, and Srivastava 1992) that measured rainfall runoff and soil losses demonstrated, under a rainfall of 680 millimeters, the advantages of Vetiver grass hedges over other grasses and stone barriers. Vetiver grass hedges reduced runoff by 57 percent, followed by lemon grass with 29 percent. Under the site conditions stone bunds were not effective in reducing runoff (in Africa the use of "digettes" as a means of moisture conservation may need to be revisited), although they reduced the transportation of erosion sediments. These results are summarized in table 3. 27 Table 3. The Effect of Slope and Barriers on Runoff and Soil Loss During Natural and Simulated Rainfall on an Alfisol in 1991 Period/ Runoff Soil loss rainfall Treatment (mm) (t ha) Simulated 2.8% slope 20.9 0.96 rainfall' 0.6% slope 18.6 0.18 SE +1.33 +0.18 Control 24.2 1.21 Stone bund 21.7 0.33 Lemon grass 19.3 0.48 Vetiver grass 13.8 0.26 SE +1.19 +0.31 Slope x barrier ** NS Natural 2.8% slope 293.1 9.8 rainfall2 0.6% slope 182.8 3.3 SE +13.9 +1.2 Control 307.1 10.7 Stone bund 294.6 6.6 Lemon grass 218.3 5.4 Vetiver grass 131.7 3.6 SE +19.7 +1.7 Slope x barrier NS NS ** Significant at P < 0.01 level; NS = Not significant ' Between Feb-Mar 1991 with 80 mm hi intensity 2 Between Apr-Oct 1991 with 689 mm rainfall Note the relation between runoff and soil conservation. The lower the soil loss the greater the in situ moisture conservation. In stark terms the area, without conservation, had an actual effective rainfall of only 382 millimeters, but with a stiff grass hedge, Vetiver in this case, had an effective rainfall of about 557 millimeters. This level of incremental soil moisture can make the difference between a good crop and a crop failure. Nonagricultural surveys (Ochs 1992, personal communication) indicate that improved moisture conservation in south India has resulted in substantial improvements in the level of the groundwater table and the consequent benefit to rural water supply (farmers are digging wells where they have been unable to do so before). This can be reinforced from the same set of experiments that 28 generated the data in figure 2. Comparative hydrographs (figure 3), were developed which showed how the different barriers affected rate of runoff, and the total amount during individual storm events. It should be noted that even on near flat ground, 0.6 percent slopes (0.25 of a degree), release of water was slowed down significantly, and infiltration increased substantially. Currently, conventional wisdom relating to soil conservation in the United States and elsewhere does not require or even recommend conservation structures on slopes of less than 3 percent. It is recommended that agencies involved in dealing with groundwater recharge problems take a hard look at the possibility of using stiff grass hedges as a means of improving ground water recharge. Figure 3. ICRISAT Experiment - Impact of Porous and Vegetative Barriers on Storm Hydrograph a.6% slope E 0.8 50.4 02 0 #2.8% Slope 0.8 E S 0.8 0.4 a Veqve 0 10 20 30 40 50 Time (mvi) In Indonesia indications are that moisture conservation is of greater importance than soil conservation in rehabilitating degraded slopes, and in establishing farm systems that are sustainable and based on perennial crops. Since 1987 CARE International has been managing a project on the island of Lombok (eastern Indonesia). This area has 1,800 millimeters of rainfall and an eight-month dry season. Population pressure is sending farmers up the steep mountain slopes, destroying forests and creating erosion hazards and increasing runoff. The CARE project has recognized the need for moisture conservation through antisoil erosion measures. CARE has concentrated on the establishment of tree hedges using mainly Gliricidia sp. (other species include Calliandra, Albizia, Leuceana sp). Government has provided farmers with perennial fruit tree seedlings such as mango, avocado, cocoa, jack fruit, and coffee. As a result, some 5,000 farmers, have established over 3,000 hectares sustainable farms that now are mainly perennial cash crop based, and are not unlike the tropical garden farms of eastern Java and Kerala. Farmers report threefold increases in yield of maize and soybean; and fodder is 29 available for improved livestock feeding. In addition the improved situation on farm has contributed locally to increased dry season stream and spring flows. The project clearly demonstrates the primary role of moisture, and the secondary impact of soil conservation. Farmers are too poor to use pesticides and fertilizer, and from all appearances hardly need them, under the particular site conditions, including the dominance of perennial cash crops. The stabilization that has occurred on this project clearly demonstrates the need to stabilize farm land prior to, or in parallel with, action to maintain the status of adjacent forest lands, in fact without stabilization of farm land there is little chance of protecting forests. The latter can only be best dealt with on a community basis to assure that the benefits of the resource remains available to the local community. Another case demonstrating very successful moisture conservation techniques was the Bank's "Watershed Development Project in Rainfed Areas" located in a number of south Indian States. The introduction of contoured 'V' ditches that spread the water evenly over large areas of land gave remarkable increases in tree survival rates and growth, even in the worst years, when compared to other methods. As a result, the technique is spreading widely in India. Figure 4 demonstrates the results from Maheshswaram Watershed in Andhra Pradesh, India (World Bank 1991). Figure 4. Comparison of Survival and Growth of Trees in Relation to Different Water Conservation Treatments Second Year Srvival As A FuncUon Of Planing Technuque Mman Height Al Second Year As A Function Of Planting Technique 150 8 60 A AA Pi Cs r ca itPM C C j l L L nh CS s C 20 - T~~~~~~~~~~~Ai 50 M 0 0 Pfanting Technique Planting Technique DS = Dalbergia sissoo; Al =Azadirachta indica; AA = Acacia auricuiformis; SJ =Syzgiun jambolana; LL = Leuceana leucocephala; CS = Cassia siamea; AM = Aegle mannelos; TI Tamarindus indica; PE = Phylanthus embilica. Recent studies of Machakos District in Kenya (Tiffen 1992) suggest that moisture conservation, through soil conservation measures, may be the key reason why farmers have been able to intensify production and maintain production levels under mounting population pressures. 30 Development strategies that fail to address satisfactorily soil moisture conservation are likely to be more risky and are less likely to achieve sustainability objectives. Nowhere is this more important than in the rainfed areas of Asia and Africa. Technology Transfer and Generation In the review of these critical technology areas one is struck by the fact that once a farming community has accessed technology, either though internal generation or external acquisition, the technology seems thereafter, primarily farmer driven, self generating, and needs little outside support. This is supported by experience in Lombok andonesia), Machakos (Kenya), and the Loess Plateau (China) where agencies focused primarily on conservation technologies, and once the technology, which in these cases was an agricultural sustaining technology, was established, the agencies were basically able to withdraw. A role for the external agency, together with the client farmer, should be to identify critical technologies suitable for sustainable agricultural development, demonstrate them to a level of reasonable acceptability, and then see them internalized by the community. There are many ways of getting information to farmers. Public extension services are a common method, though not always successful. Other methods include: farmer training, farmer experimentation, market penetration, films, school programs, demonstrations (universally "seeing is believing"), written communications, master farmers, village technicians, NGOs, and farmer-to- farmer word of mouth. All have a place, and all under the right circumstances have documented successes; unfortunately the Bank has not paid enough attention to this menu of delivery, and has instead probably put undue attention on government extension services. One of the transfer areas that has not been sufficiently used is direct knowledge transfer to farmers through written material. The Chinese (because of the widespread nature of their written language) have depended on farmer publications. One thousand years ago prominent Chinese agriculturists (landowners) manualized farming instructions so that the small tenant farmers might be more productive. China does the same thing successfully today. A survey in the United Kingdom showed that 70 percent of farmers, mostly the smaller farmers, were dependent on publications for their technical information. It should be noted that in some countries the intransigent nature of the bureaucracies in transferring new technology to the user can be a real constraint. Technology change often impacts on individual agendas and vested interests. Nowhere is this more apparent than in some soil conservation departments in their reluctance to move away from past engineering practices toward more environmentally benign and low-cost vegetative systems. Research The four technical areas that are described in this paper could be better served by targeted and competitive research that would confirm farmer practices as being sound, and research other relevant technologies that could be offered to farmers and to the specific areas concerned. The World Bank should be instrumental in identifying some of these topics, and directing funds toward them. In some instances client countries would probably be better served if the Bank desisted in underwriting national research programs, but instead supported targeted research though a selective and competitive approach. Incentives and rewards are part of this process. There still remains a lot of repetitive and rather irrelevant research being undertaken because of lack of innovation in identifying useful topics, 31 lack of accountability for performance and failure by scientists to work closely with farmers and extension agents. Linking technically sound and practical information networks to research systems can trigger off new lines of research. If one accepts the farmer as the ultimate researcher it would seem that scientific information networks should consider including selected "master farmers" in their networks. An area of research that has been badly neglected in relation to the four technological areas discussed in this paper is that effecting community lands -- grazing and forest lands in particular. Most of these lands are badly degraded and have no real identifiable owners. To put them back into production requires appropriate tenurial policies to be enacted and development of technologies that will be effective under what are normally ecological adverse conditions. Conclusions This paper has highlighted some important features in technology and sustainable agriculture. The combination of the technologies has in nearly every case led to a sustainable form of agriculture. Mary Tiffin's review of Machakos District concludes that the land use and agriculture is better now than sixty years ago. The area supports five times the number of people, incomes have improved, and the degradation process has been reversed (Gichuki 1992). In Lombok, CARE International is able to withdraw from a specific conservation area after only three years of support knowing that the conservation works have changed a degrading land-use situation into a sustainable farming framework based on long-term perennial tree crops (Gibney 1992, personal communication). In China, conservation works on the Loess Plateau have for hundreds of years created stable agricultural conditions, and with the help of new technologies (varieties, tools, and fertilizer) have enabled significant production increases to be achieved on a sustainable basis (World Food Program 1991). These are just three of many programs, mostly small, that are making progress in reaching sustainable agricultural objectives for the small and poor in many parts of the world. The Bank has an important role in learning from these successes and in assisting governments to develop appropriate policies and investment programs that can extend such technologies to much wider areas and to many more beneficiaries. The capture of rainfall to improve soil moisture in upland areas, and the improved performance of irrigation systems for the lowlands are critical to sustainable agriculture. Improved crop water management enables the farmer to move to a higher level and value of production, a step that often leads to better overall resource management and conservation. Soil moisture is linked closely with soil fertility. Both relate to the effectiveness of adequately protected and quality germplasm. The trick to sustainability is to keep them in balance. Too much of one input can have a detrimental impact elsewhere. Basically the need is to practice good husbandry; good husbandry equates to sound resource management by the user. It has been achieved in the past and can no doubt be achieved in the future. References Antholt, Charles. 1990. "Strategic Issues for Agricultural Extension in Pakistan: Looking Back to Look Ahead." Proceedings, Productivity Extension in Pakistan, Islamabad, March 1990. 32 Arya, L.M., T.S. Dierolf, B. Rusman, A. Sofyan, and I.P.G. Widjaja-Adhi. 1992. "Soil Structure Effects on Hydrologic Processes and Crop Water Availability in Ultisols and Oxisols of Sitiung, Indonesia." Soil Management CRSP Bulletin No. 92-03. North Carolina State University. Bharad, G.M., and B.C. Bathal. 1990. "Role of Vetiver Grass in Soil and Moisture Conservation." In the Proceedings of the Colloquium on the use of Vetiver in Sediment Control. 25 April, 1990 Dehra Dun, India: Watershed Management Directorate. Doolette, J.B, and William B. Magrath. 1990. "Watershed Development in Asia - Strategies and Technologies." World Bank Technical Paper 127. Washington, D.C. Fan, Shenggen, and Philip G. Pardey. 1992. "Agricultural Research in China. Its Institutional Development and Impact." The Hague: International Service for National Agricultural Research (ISNAR). FAO/World Bank Cooperative Program. 1992. "Forestry Institutions and Conservation Project." Wonogiri Watershed Component Mid-Term Evaluation. Draft. Rome. Gichuki, F.N. 1992. "Environmental Change and Dryland Management in Machakos District, Kenya. 1930-90." Conservation Profile. ODI. Washington, D.C.: World Bank. Grimshaw, R.G. 1991. "The Establishment of Vetiveria zizanioldes in Low Rainfall Areas." In Desertified Grasslands: Their Biology and Management. The Linnean Society of London. Lin, Bao, Jin Ji-Yun, and S.F. Dowdle. 1989. "Soil Fertility and Transition from Low-Input to High-Input Agriculture." Better Crops International. Potash and Phosphate Institute. Malajczuk, Nicholas, Norman Jones, and Constance Neely. 1992. "The Importance of Mycorrhiza to Forest Trees." World Bank. Asia Technical Department, Agricultural Division. Land Resources Series No 2. Washington, D.C. Needham, Joseph. 1984. Science and Civilization in China. vol.6, part II: Agriculture, Francesca Bray. Cambridge: Cambridge University Press. Rao, K.P.C., A.L. Cogle, and K.L. Srivastava. 1992. "Conservation Effects of Porous and Vegetative Barriers." Hyderabad: International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). Smyle, J.W., and William B Magrath. 1990. "Vetiver Grass - A Hedge Against Erosion." American Society of Agronomy annual meetings in San Antonio Texas. Tiffen, Mary. 1992. "Environmental Change and Dryland Management in Machakos District, Kenya 1930-1990." Farming and Income Systems. ODI Working Paper 59. Washington, D.C.: World Bank. 33 Tiffen, Mary, S.G. Mbogoh, and Ackello-Ogutu. 1992. "Environmental Change and Dryland Management in Machakos District, Kenya 1930-1990." Production Profile. ODI Paper 55. Washington, D.C.: World Bank. TropSoils. 1989 . Technical Report 1986-1987. North Carolina State University: TropSoils. U.S. National Academy of Science. 1992. "The Thin Green Line." Washington, D.C. World Bank. 1992a. "Indonesia: Integrated Pest Management Project." Appraisal Report. 11377-IND. Washington, D.C. World Bank. 1992b. "Asia Water Resources Study." Draft Report. Asia Technical Department. Agricultural Division. Washington, D.C. World Bank. 1991. Vetiver News Letter No.5. Asia Region Technical Department. Washington, D.C. World Food Program. 1991. Mizhri Project. Project Completion Report. I I I Socioeconomic and Policy Considerations for Sustainable Agricultural Development Per Pinstrup-Andersen* During the next twenty to thirty years, the agricultural sector of developing countries will be faced with the biggest challenge ever. In fact, several big challenges. First, it must provide food at affordable prices for almost 100 million more people every year, the largest annual population increase in history. Second, a large and increasing share of the production increases must originate from higher yields per unit of land. Third, the increasing production capacity must be sustainable, that is, the additional food must be produced at the same time as the future productive capacity is enhanced. Fourth, in most low-income developing countries, the agricultural sector must provide employment for a rapidly increasing labor force, either directly or indirectly through linkages with nonagricultural sectors, and thus serve as the lead sector for general economic growth and poverty alleviation. Add to these future challenges, the current challenge of reducing food insecurity among the 700 million people who do not now have access to enough food for a healthy and productive life. This paper will focus on one of the many facets of the above: the socioeconomic and policy aspects of the sustainability of future agricultural development. Without entering into a discussion of the merits of the various definitions of sustainable development currently found in the literature, I propose that agricultural development should be considered sustainable if it assures that the productive capacity of the agricultural sector will be sufficient to meet current and future needs. This corresponds to the more general definition suggested by the Brundtland Commission: "Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs" (World Commission on Environment and Development 1987). This definition implies that sustainability is compatible with the reduction of the stock of natural resources if and only if the productivity of the stock of human-made resources increases sufficiently to compensate for the loss of natural resources and, together with natural resources, to meet future needs. Because future needs cannot be predicted with great certainty, reductions in the stock of nonrenewable natural resources is a very serious matter, even when future generations are fully compensated for the loss in productive capacity through enhanced human-made capital. However, both current and future generations are best served by an agricultural development strategy that enhances sustainable productive capacity of the combined set of resources, that is, natural and human-made resources, even when that implies a reduction in the stock of natural resources. An overriding goal of conserving the stock of natural resources may result in stagnation while a focus on opportunities for substitution between natural and human-made resources and expanded productivity through research and technology is likely to result in a higher rate of sustainable growth in agricultural output. In addition to the productive capacity, two other factors are important in considerations regarding the future stock of natural resources. First, scarce natural resources may have an intrinsic value over and above their productive value due to their mere existence. This "existence value" increases with increasing scarcity. Closely related is the desire to avoid reduced biodiversity not only for the existence value but also to assure future access to biological materials for further enhancement. Second, changes in the stock of natural resources may be associated with externalities not immediately reflected in agricultural growth, for example, the effects of deforestation on * Director General of the International Food Policy Research Institute, Washington, D.C. 36 climatical variables. I believe these two factors should be dealt with explicitly as premiums placed on the implicit or explicit discount rates. Production Considerations Sustainability aspects of agricultural development must be considered within the context of past and current production performance and the future demand for agricultural products. Although food production increased at an impressive rate in developing countries during the 1980s, it failed to keep pace with population growth in two-thirds of the developing countries and in more than 80 percent of African countries. Most of these countries do not have the foreign exchange to import the production shortfalls. On average, per capita food production in developing countries increased by 9 percent during the first eight years of the 1980s, while it decreased by 6 percent in Africa. Although net food imports increased, the 1980s saw an expansion in the number of people, currently in excess of 700 million, who do not have access to enough food to meet energy and protein requirements for a healthy and productive life. Keeping up with increasing needs due to population growth and increasing resource requirements in food production due to income increases and dietary changes is a formidable challenge to the agricultural production and marketing sectors. Opportunities for sustainable increases in food production through area expansions are gradually being replaced by the need to depend exclusively on yield increases as the source of production increases. Future perspectives for per capita food production are grim for Africa because of the expected continuation of high rates of population growth and slow transition from area to yield dependence for production expansions. Between now and the year 2000, the African population will grow at more than 3 percent per year while food production will grow annually by only 2 percent. The International Food Policy Research Institute (IFPRI) projects that the difference between food production and demand in Sub-Saharan Africa will be 50 million tons by the year 2000. Today this gap is 12 to 14 million tons. It is extremely unlikely that the region will have the necessary foreign exchange to import such large amounts of food. It is equally unlikely that the governments will be able to count on food aid in these amounts. If the current food production and population growth trends continue, the World Bank estimates that the food shortage for Africa will be 250 million tons by the year 2020. That is more than twenty times the current food gap. The World Bank also projects that if the annual increase in food production could be doubled to 4 percent and the annual population growth reduced by half -- or 1.5 percent -- then the region could be self-sufficient by the year 2020. While opportunities for bringing new land under cultivation have compensated for slow crop yield increases in the past, continued attempts to expand agricultural land will entail increasing investments, accelerated deforestation, and land degradation. Thus, as has already happened in most of Asia, Africa must rely on increased yields to meet most of its future production expansions. But even in Asia there is trouble ahead. Rice yields increased at an annual rate of about 3 percent between the mid-1970s and the early 1980s. Those increases have dropped to less than 2 percent in the 1980s. In Southeast Asia rice yields increased at an annual rate of 3.2 percent during the late 1970s. These increases are now down to half of that -- or 1.6 percent per year. In China annual yield increases dropped from above 4 percent in the late 1970s to about 1.6 percent during the 1980s. Growth in incomes and urbanization is expected to accelerate current diet diversification in many Asian countries toward more livestock products and related rapid increases in the demand for feedgrain. IFPRI projects that developing countries will need 90 million tons more grain in the year 2000 than they will produce. Some will be able to fill the gap through imports. Most will not. 37 The international community has become complacent about future food supplies. We have convinced ourselves that there is enough food to go around and that the problem is one of distribution. Real food prices in the world market have decreased, the European Common Market and North America have excess capacity, and some speculate that Eastern Europe and the former Soviet Union will, over the next ten years or so, greatly expand their food production. Unfortunately, current and expected future excess production capacity in Europe, North America, and the Commonwealth of Independent States (CIS) will be of only limited support to low- income countries with little foreign exchange available for food import. While food self-sufficiency is not an appropriate goal for all countries, most low-income developing countries must rely on their own agricultural sectors for meeting most of their food demands for a long time to come. The agricultural sector may also serve as a major contributor of export commodities needed to generate foreign exchange to support overall economic growth. Furthermore, agricultural development is likely to be the most effective way to generate general economic growth in these countries. Except for those few low-income developing countries possessing large amounts of exportable minerals, a stagnant agricultural sector is a prescription for poor economic growth and accelerated rates of poverty, food insecurity, and malnutrition. Although these relationships are widely accepted, external assistance to agricultural development and enhanced food production decreased significantly during the 1980s, both in absolute terms and relative to other assistance. Thus, support to agriculture decreased from 22 to 14 percent of all external assistance from Organisation for Economic Co-operation and Development (OECD) countries to developing countries during the 1980s. Similar trends are found in the allocation of public funds in many low-income developing countries. Sharp reductions in both external and national support of agricultural research during the 1980s has resulted in serious deteriorations in the capacity to generate improved agricultural technology, without which the needed productivity increases will not materialize. The trends of the 1980s must be reversed -- and soon. Pursuing a Triple Goal It should be clear from the above that sustainability in food production is not merely a matter of conserving our natural resources. It is rather a matter of finding a way to meet rapidly increasing food demands without compromising the ability of the total stock of resources (natural and human- made) to meet even larger demands by future generations. But, although usually referring to intergenerational distribution, sustainability is also related to spatial distribution, that is, distribution among population groups at a given point in time. The prevalence and severity of poverty is of particular importance in this regard, not only from the point of view of social justice but also because contemporary poverty will both weaken the justification for conservation of resources for future generations and make such conservation more difficult. The poverty link will be further discussed below. Suffice it to say at this point that sustainable agricultural development should pursue the triple goal of assuring sufficient increases in food production to meet future demand, strengthening the productive capacity of the total stock of resources for agricultural production, and alleviating poverty. While any one of the three goals can be pursued at the expense of the others, the three can be complementary. For example, rural poverty and low productivity in food production contribute to land degradation and deforestation. Similarly, land degradation contributes to low productivity and poverty. On the other hand, high productivity brought about by inappropriate water management and excessive use of chemicals may result in degradation of natural resources. The role of policy is to pursue the three goals simultaneously, enhance the complementarities among them, and avoid unacceptable tradeoffs. 38 Socioeconomic and Policy Considerations Five critical socioeconomic and policy considerations, that is, poverty, externalities, input use, market and policy failures, and population growth are considered in this next section. Poverty Poverty is probably the most important source of environmental degradation in low-income developing countries. Poor people have a high internal discount rate. They consume capital, that is, future productive capacity, when it is necessary to survive. It makes much sense for low-income rural people to cultivate fragile soils and to clear forest land for agricultural production, even though they are fully aware that such practices are not sustainable. When survival is at stake, conservation of natural resources for future generations takes on a lesser importance. This is particularly the case when the poor cannot assure that their children will in fact benefit from such conservation. Regulations and land planning that contradict the poor's survival strategies are unlikely to be successful simply because they are difficult or impossible to enforce. This is one of many examples where incentives are preferable over regulations. Externalities One major reason why poverty results in environmental damage is the existence of externalities, that is, a situation where costs and benefits of a particular behavior or action are not borne by the same person or persons. Externalities resulting from inadequate property rights are of particular importance. We may distinguish among four types of property rights to land, water, and forests: open access, communal property, private property, and state property. Resources with open access are particularly prone to exploitation because exploiters may benefit without paying the costs associated with reduced future productive capacity. Regulations may reduce exploitations but enforcement is likely to be difficult and expensive. The institutions and traditions surrounding common property rights are complex and poorly understood by those not directly involved. Efforts by governments and international institutions to privatize common property are often based more on ideology than on a thorough understanding of how best to achieve sustainability goals. Common property ownership is frequently confused with open access. While private land ownership may be expected to be most effective in achieving the food security, poverty, and sustainability goals in many cases, it should not be assumed that it is always superior to common ownership. Efforts to assure appropriate ownership patterns should begin by understanding what exists and improve on it. Such improvement may but need not necessarily imply a transfer to private ownership. Incentive policies, including taxes and subsidies, combined with common ownership should be considered as an alternative to private ownership. Incentive policies are also likely to be needed to guide private resource owners such as small farmers toward sustainability goals. These policies should focus on compensating for poor farmers' high discount rate, reducing or compensating for the risk and uncertainty with which they are faced, compensate for or correct poorly functioning land markets, and provide information to assist farmers in avoiding large errors in their expectations regarding future land and output prices. 39 State ownership may also result in environmental degradation and failure to achieve the three goals mentioned above. In comparison to the privatization of common property, privatization of state-owned agricultural production and removal of government monopolies in input and output markets is much more likely to enhance food security and sustainability while reducing poverty. Population Growth The importance of population growth in efforts to meet future food needs was mentioned above. Because the pivotal role of population pressures in current and future attempts to avoid environmental degradation and to alleviate poverty is well known, suffice to say that failure to reduce population growth rates drastically in Africa and South Asia and significantly in many developing countries in other regions over the next twenty years will render all other efforts to achieve universal household food security, poverty alleviation, and sustainability in agricultural production futile. In addition to growth rates, a number of other demographic factors are closely related to sustainability. During the last fifteen years, Africa and parts of Asia and Latin America have experienced movements of people of magnitudes never previously seen. Such movements are frequently part of poor people's survival strategy. Ecological disasters, including drought, along with wars and social strife, have been significant reasons for these movements. Large-scale migration both within and between countries may result from environmental degradation or it may cause such degradation. As populations become more mobile (in part a result of desperation and attempts to survive), the risk of environmental degradation increases because of rapidly increasing population concentrations in areas not yet degraded and associated externalities. Areas benefiting from improvements in rural infrastructure, natural resources, and agricultural productivity are particularly prone to large inflows of migrants from other less productive regions. Productivity and Input Use Farmers who do not have access to yield enhancing inputs are more apt to degrade land, water, and forests than farmers who do. Low-input agriculture clearly has a role to play, low-productivity agriculture much less so. Improved cultivation practices, use of organic wastes, enhanced biological nitrogen fixation, integrated pest management, and other means to reduce the need for chemical inputs should be pursued. However, the limitations of these practices should be recognized. Where current uses are low, chemical fertilizers must necessarily play an increasing role in efforts to expand productivity. This is the case in most of Africa, where average fertilizer use is around ten kilograms per hectare. Given the low levels of current use, the environmental risks associated with expanded fertilizer use in most of Africa and much of the rest of the developing countries are low. Efforts to expand fertilizer use are essential to achieve productivity and poverty goals. Furthermore, the environmental effects are likely to be positive, partly because soil fertility will be maintained and partly because farmers will be less likely to cultivate new fragile lands and clear forests for agricultural production. In some regions of the developing countries, notably areas in Asia with highly intensified rice and wheat production, excessive fertilizer use poses serious environmental risks. While these risks should be effectively dealt with, it is important that they not be confused with the situation on the large majority of developing country farms, where the problem is insufficient rather that excessive fertilizer use. 40 Inappropriate water management in irrigated areas has resulted in waterlogging and salination. Such land degradation, which fortunately is reversible, contributed to the above mentioned reduction in the rate of increase in rice yields in Asia. Inappropriate water management has also resulted in lowering of the water table and groundwater pollution. Excessive use of pesticides is another highly location-specific environmental risk. Biological control measures, integrated pest management, and research to incorporate tolerance or resistance into plants should be promoted. However, where such measures are unable to provide the necessary protection from pests, efforts to reduce the use of pesticide may result in unacceptable productivity and poverty effects. Policy and Market Failures Policy and market failures are an integral part of the problems already mentioned. Poorly functioning land markets and policies that effectively prohibit land markets from functioning are common. Excessive state intervention in input and output markets currently is being replaced with a stronger private sector. One of the major challenges in these privatization efforts is that of identifying the appropriate role of the state both during and after the transition. A variety of policies that promote environmental degradation are found in developing and developed countries. Some of these are justified on productivity and poverty grounds, for example, subsidies on chemical pesticides for small farmers, subsidies on chemical fertilizers in areas with excessive use, free access to water, payment to farmers for deforestation, subsidies for logging where such practice is not sustainable, taxes on alternatives to firewood, and so forth. The challenge to policymakers is to design and implement policies with positive effects on food security, poverty, and sustainability rather than policies benefiting one to the detriment of the others. Designing and Implementing Appropriate Policies The design and implementation of appropriate policies to simultaneously achieve the three goals must be based on a sound understanding of household and community behavior as it relates to food security, poverty, and the management of natural resources. Such understanding is essential because the effects of policies will be determined or heavily influenced by household and community behavior. Without it, policies may be ineffective or their effects may be contrary to expectations. Goals and preferences of households, time allocation, gender-specificity in decisionmaking and power structures, risk behavior, expectations, and implicit discount rates are all behavioral issues of importance for policy choice and implementation strategy. Policies that guide and modify the behavior of households and communities through incentives are much more likely to succeed than policies that conflict with behavior and attempt to achieve their objectives through regulations. That implies that policies should focus on incentives rather than regulations although there are some cases where regulation is likely to be most appropriate. The point is that because there are so many decisionmakers with influence over the environment and because even simple policies are difficult to enforce among people fighting for survival, governments are unlikely to be able to assure sound management of natural resources by fiat. Another important point related to the design and implementation of policies to assure sustainability, poverty alleviation, and food security is that both agriculture and nonagricultural sectors should be considered. In some cases, the most appropriate solutions to protecting fragile 41 environments may lie outside agriculture, for example, out-migration or the strengthening of nonagricultural income sources in the regions. Policies should consider both intergenerational and spatial distribution. Efforts to assure conservation of natural resources for future generations at the expense of the survival of part of the current generation are hard to justify on moral grounds and will be difficult to implement. The potential tradeoffs between alleviation of current poverty and the needs of future generations must be considered and attempts must be made to identify policies and strategies that achieve both goals. As markets, particularly capital markets, become more effective, one would expect that the needs for corrective policies to assure sustainability in agricultural production will decrease. However, it is naive to believe that market signals will be sufficient to create the right institutions and to assure the right kinds of research and technology in a timely fashion. The public sector will continue to play an important role in guiding research and technological and institutional change. To facilitate the needed increases in crop yields in developing countries, there is an urgent need for accelerated investment in agricultural research and technology aimed at yield enhancement, stabilization, and reduced dependence on chemical inputs. The dramatic impact of research and technology on the yields of most crops grown in temperate zones and on wheat and rice in Asia and Latin America is well known. Less dramatic, but significant, impact has been made in other crops for developing countries, notably maize, and results from current experiments on various crops are promising. Accelerated investment in agricultural research and technology is also urgently needed to protect natural resources from degradation. As population, poverty, and food demands continue to grow, failure to develop and implement appropriate technology in production will lead to either more food insecurity and hunger, for which the current generation of poor people will pay, or to further degradation of natural resources, for which future generations will pay. While development and use of appropriate technology is necessary to meet future food demands, it is not sufficient. Investment in rural infrastructure, institutional change, and appropriate government policy are needed to strengthen savings and credit institutions, to facilitate access by farmers to modern inputs, improve farm management, develop a marketing system capable of assuring access to sufficient food by the rapidly growing urban population, provide the necessary production incentives, promote economic growth in rural areas, and enhance the exchange of goods and services between urban and rural areas. Public sector investment should also be accelerated to pursue short-term poverty alleviation, including labor-intensive public works programs; credit programs targeted on the rural poor; poverty relief programs; and long-term human resource development, including education, primary health care, nutrition, water and sanitation, and family planning. Renewed emphasis must be placed on efforts to reduce population growth in the developing countries in general and in Africa in particular. Although the rate of population growth is falling in developing countries as a whole, the decrease is insufficient to counter absolute increases. Thus, over the next ten to twenty years, the world population will increase by almost 100 million people annually, the largest annual increase ever. The pressures on food production and distribution will be immense, particularly because the majority of the population growth will occur in urban areas of developing countries. Failure to significantly reduce the current high population growth rates in Africa within the next twenty years will render all other development efforts insufficient to avoid future famines, degradation of land and forest resources, poverty, and human misery of much greater magnitudes than experienced to date. Incentives and -- where they are unlikely to achieve the objectives -- regulatory policies must be strengthened to endogenize or compensate for externalities related to natural resources. The nature of such policies will vary across countries and over time and may include appropriate water pricing 42 and watershed management, elimination of exploitation of land and forests resulting from free access, and a variety of other policy measures. While usually preferable to regulatory measures, subsidies, taxes, and other incentives should be used selectively and advisedly because of possible unintended market distortions, opportunities for rent-seeking, and high fiscal costs. Market reforms, institutional change, and investment in infrastructure to assure effective and efficient markets and reforms of property rights to eliminate externalities in the management and use of natural resources will play a major role in successful efforts to assure long-term sustainability in food and agricultural production. While this paper has attempted to identify a number of socioeconomic and policy issues to be considered in efforts to assure sustainability in future agricultural development, the design and implementation of a sustainable strategy in a particular country will require national capacity in short supply in most low-income developing countries. It is essential that both institutional and human capacity be strengthened through policy research, training, short- and long-term technical assistance and other relevant measures. Sustainable agricultural development will be successful only if based on sound analysis and location-specific understanding of the relevant factors. References Anderson, Dennis. 1990. Environmental Policy and the Public Revenue in Developing Countries. Environment Working Paper No. 36. Washington, D.C.: World Bank. . 1987. The Economics of Afforestation: A Case Study in Africa. Baltimore and London: Johns Hopkins University Press for the World Bank. Anderson, Robert C., Lisa A. Hofmann, and Michael Rusin. 1990. The Use of Economic Incentive Mechanisms in Environmental Management. Research Paper No. 051. Washington, D.C.: American Petroleum Institute. Bromley, Daniel W. 1987. "The Common Property Challenge." In Proceedings of the Conference on Common Property Resource Management, April 21-26, 1985. Prepared by the Panel on Common Property Resource Management, Board on Science and Technology for International Development, Office of International Affairs, National Research Council. Washington, D.C.: National Academy Press. Brown, Lester R., Alan Durning, Christopher Flavin, Hilary French, Jodi Jacobson, Nicholas Lenssen, Marcia Lowe, Sandra Postel, Michael Renner, Linda Starke, Peter Weber, and John Young. 1993. State of the World 1993: A Worldwatch Institute Report on Progress Toward a Sustainable Society. New York and London: W. W. Norton and Company. Brundtland, Gro Harlem (Chairman). 1987. Our Common Future: World Commission on Environment and Development. Oxford: Oxford University Press. Chambers, Robert, N.C. Saxena, and Tushaar Shah. 1989. To the Hands of the Poor: Water and Trees. London: Intermediate Technology Publications. 43 Cleaver, Kevin. 1992. "A Strategy to Develop Agriculture in Sub-Saharan Africa and a Focus for the World Bank." Agriculture and Rural Development Series No. 2. Washington, D.C.: World Bank. (Draft for Discussion). Cleaver, Kevin M. and Gotz A. Schreiber. 1992. The Population, Agriculture, and Environment Nexus in Sub-Saharan Africa. Agriculture and Rural Development Series No. 1. Washington, D.C.: World Bank. Crosson, Pierre, and Jock R. Anderson. 1992. "Global Food - Resources and Prospects for the Major Cereals: An Assessment of the State of Agriculture with Commentary on Population, Demand, Resources, Technology, and Supply to 2030." World Development Report Background Paper No. 19. Washington, D.C.: World Bank. Eskeland, Gunnar S. and Emmanuel Jimenez. 1991. "Choosing Policy Instruments for Pollution Control: A Review." Policy, Research, and External Affairs (PRE) Working Papers No. WPS-624. Washington, D.C.: World Bank. Gregersen, Hans, Peter Oram, and John Spears, eds. 1992. Priorities for Forestry and Agroforestry Policy Research: Report of an International Workshop. Washington, D.C.: International Food Policy Research Institute. Joshi, P. K., and Dayanatha Jha. 1990. "Environmental Externalities in Surface Irrigation Systems in India." In Environmental Aspects ofAgricultural Development, Policy Briefs No. 6. Washington, D.C.: International Food Policy Research Institute. Krutilla, John V., and Anthony C. Fisher. 1975. The Economics of Natural Environments. Washington, D.C.: Resources for the Future. Leach, Gerald and Robin Mearns. 1988. Beyond the Woodfuel Crisis: People, Land and Trees in Africa. London: Earthscan Publications Ltd. Leonard, H. Jeffrey, Montague Yudelman, J. Dirck Stryker, John 0. Browder, A. John De Boer, Tim Campbell, and Alison Jolly. 1989. Environment and the Poor: Development Strategies for a Common Agenda. New Brunswick, New Jersey: Transaction Books. Markandya, Anil, and David W. Pearce. 1991. "Development, the Environmnent, and the Social Rate of Discount." Research Observer, Vol. 6 (No. 2, July):137-152. Washington, D.C.: World Bank. Pearce, David, Edward Barbier, and Anil Markandya. 1990. Sustainable Development: Economics and Environment in the Third World. London: Earthscan Publications Ltd. Platteau, J-Ph. 1992. Formalization and Privatization of Land Rights in Sub-Saharan Africa: A Critique of Current Orthodoxies and Structural Adjustment Programmes. Development Economics Research Programme DEP No. 34. London: London School of Economics, January. 44 Reardon, Thomas, and Stephen A. Vosti. Forthcoming. "Issues in the Analysis of the Effects of Policy on Conservation and Productivity at the Household Level in Developing Countries." In Quarterly Journal of International Agriculture. Schramm, Gunter, and Jeremy J. Warford, eds. 1989. Environmental Management and Economic Development. Baltimore and London: Johns Hopkins University Press for the World Bank. Tietenberg, Tom. 1988. Environmental and Natural Resource Economics. Glenview, Illinois: Scott, Foresman and Company. Vosti, Stephen A., Thomas Reardon, and Winfried von Urff, eds. 1991. Agricultural Sustainability, Growth, and Poverty Alleviation: Issues and Policies. Proceedings of the IFPRI/DSE Conference held from 23 to 27 September, 1991 in Feldafing, Germany. Germany: Deutsche Stiftung fUr internationale Entwicklung. Wheeler, Joseph C. 1990. Development Co-operation: Efforts and Policies of the Members of the Development Assistance Committee. Paris, France: Organisation for Economic Co-operation and Development (OECD). World Bank. 1992. World Development Report 1992: Development and the Envronment. Oxford: Oxford University Press. World Bank. 1991. The Forest Sector: A World Bank Policy Paper. Washington, D.C. World Commission on Environment and Development. 1987. Food 2000:. Global Policies for Sustainable Agriculture. A Report of the Advisory Panel on Food Security, Agriculture, Forestry, and Environment to the World Commission on Environment and Development. London and New Jersey: Zed Books Ltd. World Resources Institute. 1992. World Resources 1992-93. A Report in collaboration with the United Nations Environment Programme and the United Nations Development Programme. Oxford and New York: Oxford University Press. AGRIcuLTuRE AND ENvIRoNMiENT: RESPONSIBLE MANAGEMENT I Does Population Growth Inevitably Lead to Land Degradation? John English* Introduction In this paper I intend to discuss this issue by drawing on the results of two studies that we have undertaken in Sub-Saharan Africa, one drawing on data from Ethiopia, and the second in Kenya. These were very different types of studies, undertaken under quite different circumstances, but which have bearing on the issue. I will argue that the answer to the question is "no," and that the real concern should be on those steps that can be taken to ensure that land degradation does not follow population growth. The paper briefly reviews the results of these two studies and discusses the general view on the links between population growth and environmental degradation, and the importance of human capital in avoiding degradation under conditions of population growth. It then discusses possible preconditions for successful adaptation to increased population, relates the two case studies to the model outlined, and finally discusses the implications of the findings for development policy. The Ethiopia Study This study was undertaken because of concerns over possible population-degradation links in Ethiopia and to take advantage of the considerable data on land degradation and related indicators developed by the Food and Agriculture Organization of the United Nations (FAO) in the mid-1980s in their Ethiopian Highlands Reclamation Study. As a part of the FAO work, a Soil Erosion Severity Index (SESI) was developed. For our study it was possible to prepare estimates of the SESI for each of the forty-seven "highland awrajas" (districts) where most of the population live. A data set covering a range of physical and socioeconomic variables was prepared for these districts (see appendix 1). The objective of the study was to see if the variations of the SESI could be explained by variations in the other variables. Starting from the hypothesis that erosion would be positively related to population pressure, a simple correlation of the variables indicated a relationship in this direction. More detailed analysis was handicapped by the fact that SESI was a polychtomous ordinal variable. This is a constraint on the direct application of least squares methods, because the basic methodological assumptions, on which such methods are based, would be violated. An ordinal logit analysis was therefore chosen. In the model with the most satisfactory results, the variables that were significant (see appendix 1) were (in declining order of significance): animal pressure; population pressure; farming system; and the proportion of forest in total land use. In this model, population and animal pressure variables were expressed as negative reciprocals. This means that there is a hyperbolic relationship and that the probability of a given area being in a higher SESI class increases at an accelerating rate with increased population pressure. The farming system variable is a dummy distinguishing between those districts with populations John English is Principal Economist in the Policy and Research Division of the Environment Department. 46 predominantly in the Ethio-semitic or Cushitic language groups. The former are concentrated in the northern highlands and practice an agriculture based on animal traction and a predominance of seed- based annual crops. The latter, concentrated in the south, is based on a hoe culture dependent more on perennial and root crops. In the model used, the annual cropping system is linked to greater erosion severity, but this may also capture some differences in settlement history and population density. In this study, the estimate of the population pressure variable was based on a notional carrying capacity reflecting the existing low productivity agricultural systems being used, and essentially assuming static technology. This may be a reasonable assumption in this case. At the time the data was generated (the mid-1980s) the country had been under the Mengistu regime for about a decade. In this period there was little effort to foster technological change and growth of commercial production. There was also great uncertainty regarding land rights. In fact, general conditions in the countryside fostered subsistence production and discouraged innovation and investment. Under such conditions, increased population pressure does appear to be accompanied by land degradation. The Machakos Study In contrast the Machakos study was 'longitudinal' rather than cross-sectional and studied land management over an extended period. In the late 1930s Machakos District, a semiarid area of East Central Kenya, inhabited by the Akamba people, was considered by the colonial administration to be degrading alarmingly and to be rapidly approaching, if not exceeding, its capacity to support its inhabitants and their livestock. Today the area has a population five times as great and the value of agricultural output per head (at constant prices) is estimated to be three times larger than it was then. At the same time food production in the area is less susceptible to drought than before, although it is still subject to it. The objective of the study was to review the experience of this district in relation to management of its land resources over the past fifty years and to attempt to identify the factors responsible for these changes. While population has grown, there also has been some redistribution in the district. Movement has occurred into the lower, more arid area, previously tsetse infested and virtually unpopulated, which now houses about one-third of the population. The area under cultivation has expanded by four or five times and there has been a corresponding reduction in the area of bush, scrub, and general grazing area. Much of the area used is now under continuous cultivation, and almost 100 percent of the area cultivated is now subject to some form of terracing. In the 1930s the main focus of the population was on livestock herding with subsistence cropping to meet basic needs. This has now changed to a primarily crop-oriented production of which a significant proportion is sold. One of the objectives of the study was to assess the current status of the land resource base. This indicated that the rate of erosion has been sharply reduced, although it does still occur. While soil analyses do show that the chemical content of the soils is lower than in the soils under natural vegetation, their productive capacity has clearly been raised substantially, and there is no evidence to suggest that their quality is declining under current use practices. In fact, the farmers of Machakos have made a very large investment in their land resources. More than 200,000 hectares has been terraced in some way, most without external support. The rangeland areas appear to have a higher proportion of woody species than earlier. This could indicate poor grazing practices and reduced capacity. However, it is not clear that this is the cause, as prohibitions on bush fires, reductions in game populations (especially of elephants), and actions by users to encourage tree growth have all 47 played a part. There are certainly more trees than before and they are being actively managed by farmers. Projections made in the 1950s, the 1960s, and again in the 1970s all foresaw severe fuel wood shortages, but there is no evidence that such have occurred. Therefore, the conclusions of the study are that the agricultural growth, which has occurred, has not been accompanied by resource degradation. The increased per capita value of agricultural production has almost entirely been in the form of cash crops. Initial emphasis was on coffee in particular, and cotton, with a subsequent shift over the past decade into fruit and horticultural crops as the relative price of coffee collapsed. Staple food production appears to have stabilized at about the level required for basic subsistence, about 200 kilograms of maize equivalent per head per year. Trade in these items does take place between those with surpluses and deficits. Substantial changes have occurred in agricultural practices. The study enumerated about forty-five new technologies that have been adopted, half of which are new products. In addition to terracing and the main cash crops, some of the most important of these have been use of ox-drawn plows requiring only two animals for traction (and their modification for weeding), early maturing maize varieties, use of crop residues for forage and use of animal manure (that is, development of mixed farming systems), and monocropping the main annual crops in rows to facilitate weeding. These types of developments are closely in line with the hypotheses of Boserup, Ruthenberg, and others on the changes and intensification induced by population growth. Since the 1950s, there has been an almost continuous process of agricultural innovation and change in the district. Some of this has resulted directly from governmental efforts, for example, breeding early maturing varieties. Others, such as use of ox plows and many of the introductions of new crops, have had virtually no official support. These agricultural changes also have not taken place in a social vacuum. Major changes have occurred in the social structure of the society. The importance of the nuclear family has increased, relative to the extended one. The absence of men in the 1940s and 1950s required women to take a more active leadership role and this has continued, and also influenced family roles, including traditional agricultural tasks. Traditional self-help groups have been modified (partly under the influence of the early terracing programs) to have more development- oriented goals and women became more active in leading them. The Akamba have always placed emphasis on education and development of local, vocationally oriented schools has been marked. This has helped widen the range of available skills and assisted in the broadening of the range of small-scale commercial and artisanal activity and in technological innovation, and undoubtedly had a significant impact on the capacity of the society to change. The Common View of the Link Between Population Growth and Degradation We have here then two very contrasting examples. In one, under conditions that did not foster social and economic change, the analysis suggests that greater population pressure is accompanied by a greater likelihood of land degradation. In the other, great technical and social changes have occurred and the area studied has accommodated a markedly increased population while reducing the perceived level of land degradation. The view that population growth is an important factor in land degradation is held widely. Its impact is frequently seen as being linked to that of poverty. This type of view has been well summarized in a recent report (Shammugaratnam and others 1992), which includes figure 1 as an illustration. "At the macro level, the cumulative effects of historical factors and policy and market failures lead to a situation characterized by four sets of factors: unequal access to resources, conditions favoring demands for large families, undeveloped human capital, and technological 48 stagnation. These factors interact among themselves. At the micro level, individuals act within the institutional and resource constraints set by these factors. That is, their actions are conditioned by structures of incentives and disincentives generated by the larger environment. An important determinant of the status of individuals or households and their behavior is their entitlements and capabilities" (Shammugaratnam, p. 2-3). Population growth does not have to lead to natural resource degradation. What it does automatically do is increase the ratio of population to land in the aggregate. Because in the initial stages of population growth, not all of the available land is occupied, the group can continue to exist as they have by expanding their area of occupation. Sooner or later, however, they will come up to some limit of land, and the ratio of population to effectively available land will start to rise. In all likelihood this change will begin, in some way, to degrade the resource base being used by the group. In the simplest case a hunting society begins to hunt out the species on which they have relied for food. In order to survive they have to make changes in their diet to utilize other, previously ignored or less preferred, species. In other words they have to innovate, and possibly reorganize themselves in some way in order to successfully hunt or gather other foods or useful products. No doubt if one looked, one could find a nexus of problems these primitive societies had to overcome. Some did overcome them. Some, no doubt, did not. Thus, it seems to me that, in a general sense, one can look at human history as being a process of population growth leading to innovation designed to avoid degradation of the resource base. In most cases societies have managed to adjust to their circumstances and one should perhaps consider that successful adaptation is to be expected. However, the implication of the views typified by figure 1 is that success is not expected, or at least is virtually impossible to achieve. The problem with this perspective is that it tends to lead to the conclusion that only a massive, externally initiated, effort will be able to overcome these problems and enable a more benign process to get underway. The General Issue What enables groups or societies to reach conditions consistent with support of higher rates of population growth? Perhaps we should consider the key questions to be, "what results of population growth increase the likelihood of innovation and successful adaptation to the changing circumstances /environment (that is, to development), and therefore, avoid a descent into environmental degradation and destitution?" And, conversely "what circumstances (including government actions) are likely to reduce the likelihood of successful adaptation?" An attempt has been made in figure 2 to indicate a set of interacting changes, which flow from population growth within a limited area, that is, the growth leads to an increase in the density of population. Some of these effects have been discussed by Esther Boserup (1965) and others, with specific reference to agricultural change, others in the more general context of overall economic and social change, by authors such as Julian Simon (1977, 1992). Most obviously and directly the increase in population results in an increase in the number of bodies to be fed, (that is, an increase in the demand for the products of agriculture) and an increase in the number of workers (that is, in the supply of labor to agriculture). The initial increase in demand will be for the traditional basket of goods used for subsistence in the region. However, for the reasons noted above, the region will come up against constraints in production of some elements of this package and the consumption pattern will have to change unless substitute sources of supply can be found. 49 Indirectly an increase in the density of population means that the number of people who can be served from an individual location (other things being equal) increases and conversely, the cost of providing a service to a given number of people falls. At the same time the number of other individuals, with whom one person is likely to interact, will rise, which will tend to increase the exchange of ideas and flow information and increase the likelihood of new ideas being generated. This in turn will assist in solving the supply problems generated by population growth. This increased population also is likely to increase the appeal of the area for external parties and thereby improve the access of the area to the wider world and, thus to external markets. The latter effect will increase the demand for some of the goods the region is able to produce. The result will be that the region has a wider range of potential combinations of products, each of which will make different demands on the labor supply and the resource base. This will allow farmers to change their farming system with little or no economic cost. In these circumstances there is no reason to suppose that a group will not take long-run resource conservation into account when choosing what product combinations to adopt. Under normal circumstances, depletion of the resource, if it is occurring, will be gradual and allow for some experimentation in methods to avoid it. Many of the relationships in figure 2 are mutually reinforcing. Once external sales begin, this will allow for some purchase of previously home-produced items and other new items. This begins to develop new aspirations, which begin to increase the emphasis on further improvement in the future, and in consequence, the development of longer-term perspectives by the population. This may be reflected in less emphasis on maintaining the traditions of the past and more on the means of supporting the group in the future; less emphasis on fate and more on conscious planning for the future. Thus, investment will be undertaken both to offset problems resulting from past changes and to improve future capacities. However, it quickly becomes apparent that the continuation of this process is not a result of population per se but of the reaction of the society to the changes required to accommodate the growth in population. This is a function of the human capital (in the broadest sense of the term) of the population, that is, their technical and organizational skills. These issues are clearly complex and a simple model may help to give us a useful starting point. Supporting an increased population means (if living standards are not to decline) that production must be increased. This requires investment in order to increase productive capacity. Stein Hansen (1992) has noted in a recent paper that Haavelmo, in a 1961 paper, outlined a simple aggregate model with only labor and capital as the two inputs to explain the ability to support population growth. The model assumes that each unit of labor needs a fixed amount of capital to produce, so that output is proportional to capital. The more capital is needed, the more savings are needed. Per capita output increases only if savings, and thus capital accumulation, are great enough to boost output faster than the growth in population. The table above shows the maximum population growth rate that can be sustained at constant per capita incomes, under varying assumptions about capital requirements and current savings rates. The left-hand column shows different incremental capital- output ratios (ICORs), which measure the amounts of additional capital needed to produce an extra unit of output. Then the body of the table shows the maximum population growth that can occur, if output per capita is not to decline, for a range of ICORs and savings rates. Thus, the higher the savings rate and the lower the ICOR, the faster the rate of population growth that can be accommodated. The incremental capital output ratio is a function of the innate productivity of the natural resources being used (that is, the natural capital) and the extent of the human capital, which can be drawn on. The circumstance illustrated in figure 1 is likely to result in a low savings rate and a high ICOR. That is, a poor population with a high dependency rate, probably poor nutrition and health, is attempting to scratch a living from poor soil using manual 50 Figure 1. Ile Interacting Nexus Factors strical Factors+ l ~~~Policy &-Market | ] ~~~Faflures. Unual access Condlt;ons Tehooia FE reo ces &- FavourinUndevelope S ,Itaghnatono ser-vices . lrze Famzilies Hurman pi>I1 a aato \ / ~~~~~~~~~~~~~~~~~Institut2or-a_ apid opula on novet/ Failures d - |Unhfygeruc | * v § I s < ~~~~~~~~waste di;sposall t| La3ofnputs |f1,t<- Ersive _Lw Produc vty _t\ S SshGopina \ | . tAnd~~~~~~Fragarnen taS |t f\ = |Short| > LR) | g ~~Fall 1 T DemDand astu>Mrenllns, ; I I~~~~~Watershe-ds. Reserves1 / I I / Ckss~~~~~~~~~~~of i } f g t W~~~~~~~~~~~bodiver-sit 1 l | ~~~~~Deforestation 51 Figure 2. Interacting Changes Flowing from Population Growth POPULATION GROWTH Reduced Cap'lal Costs / Id) of Physlcal, Ccmmorclal and Social In rdstrucluro (°) / More Accessible Easior Interchange / MarKels for Goods \ More Idea / And Labor \ {^) Generations increased Incroas9d Moro Accosiblo Local rnal ~~~~Changed Knowledge (d Demand Demand Aspliations Kowo /(d| Increased Labor i \/ / 1 X X / I ~~~~~~~~~~~~~~~~~~supply1 Longer Torm Gonorallon of Perspectives Now Technology / A I \ \ / / ~~~~~~~ ~ ~ ~~~~~~~~~security/ / / + ; \ I / (g} ~~~~~~~~~~~~~~~Land/ SIncrasod Changed Pattern of Involtmont Domand More Purchased / Producilve Land Increased I inputs Income Dlvrsilicatio n a, Changod Farming ot Loca Syolems 4 \Economy Increasod Output Note: (a), (b), etc. refer to points made in the paper. 52 Table 1. Maximum Population Growth That Can Be Achieved Compatible with Constant per Capita Income Under Constant Returns to Scale Net domestic savings rate Incremental 0.05 0.10 0.15 0.20 0.25 Capital- Output ratio 2.0 0.0250 0.0500 0.0750 0.1000 0.1250 3.0 0.0167 0.0333 0.0500 0.0667 0.0833 4.0 0.0125 0.0250 0.0375 0.0500 0.0624 5.0 0.0100 0.0200 0.0300 0.0400 0.0500 10.0 0.0050 0.0100 0.0150 0.0200 0.0250 Source: Hansen (1992). methods and a traditional cropping system in conditions where contact with the wider world and new ideas are limited. Such an area would have a minimal ability to support population growth. A lower ICOR is fostered by the types of relationships outlined in figure 2 and the key factor is the buildup of human capital. There are then two sets of issues that need to be addressed: what types of conditions are likely to foster the types of positive and mutually reinforcing change outlined in figure 2; and what are the public policies and programs that can impact positively on the relationships outlined? Preconditions A first attempt to define some areas in which such change is likely is outlined in table 2. These can be subdivided into three broad categories, physical, economic, and social. Physical Factors Perhaps most basically, physical conditions vary and some create barriers to successful adaptation. Under very arid tropical conditions the ability to adapt is restricted in two ways. First, conditions are highly variable from year to year, and droughts (that is, seasons when rainfall is insufficient to sustain a plant fully through its normal life cycle) are common. Thus, what works one year may not the next. Second, the range of plants or animals adapted to the conditions is relatively small. Thus, because risks are high a spread of alternatives has to be produced in order to offset it, and specialization, an implication of the processes outlined in figure 2, is usually too risky to be effective. On the other hand, under very humid conditions in the tropical rainforests, ecological systems are very complex. While the range of species that can be used is high, their degree of 53 Table 2. Conditions Conducive to a Positive Impact of Population Growth on Agriculture and Land Resource Management Physical High basic carrying capacity (a) Reasonably reliable rainfall in the intermediate range (b) Moderate or better soils Economic National (a) Significant internal market (urban) or access to external markets (b) Opportunities for nonfarm income sources Local (a) Active private trading network in the region (b) Reasonable physical infrastructure in the region (significant share of farmers can get products to market at appropriate seasons) Social Reasonable social and political stability Forward looking society Entrepreneurial-development oriented leadership National Factors interdependence is also high and attempts to increase use of preferred species may soon come up against these limits. Thus, while complexity and variety of the system may be high, flexibility is low. Some indigenous crops of rather low productivity have been cultivated in these areas for several thousand years, but, in general, the climate had been unhealthy for both introduced cereal crops and for livestock and, therefore, for dense sedentary human populations. This leaves the intermediate areas as the most conducive to adaptation, that is, the subhumid to humid regions and the montane regions at higher elevations. These are characterized by savannah/forest fringe, or woodland/pasture, vegetation types. Here rainfall is more reliable and a moderate range of species are adapted to the conditions. 54 Economic Factors These should, perhaps, be further subdivided into national and local factors, which are quite different from a policy point of view. National Factors The basic economic factors linked to the ease of innovation are related to the ability to trade and thus to widen the potentials for both production and consumption. At the national level this means that there should be a significant internal market (that is, an urban area) or access to external markets. Thus, the process is difficult in a situation like the highlands of central Africa (Burundi, Rwanda, and Kivu), where soils and climate are good but lack of market access severely restricts the economically viable options open to farmers. At the same time there should also be a demand for labor for nonfarm purposes. This allows for increased flexibility in production and is likely to encourage new skills that may be applicable to agricultural use. Local Factors Even if markets are accessible from a region in the physical sense, there is still need for an active trading network within a region for that potential demand to be made real at the farm gate, and the traders of the region need to have effective access to the market. To back up this commercial infrastructure there also needs to be adequate physical infrastructure, in particular the ability to actually ship goods to or from the region at reasonable cost. This does not have to be in the form of all season motorable access to every ten-acre plot, but lack of access for long periods clearly reduces the effectiveness of potential demand. Thus, a significant share of producers should be able to ship products out or obtain inputs at appropriate seasons. Social Factors Physical and economic conditions as outlined above are not likely to lead to successful innovation and development of sustainable farming systems in the absence of a society that has the ability to identify and solve the problems that confront it. First, there needs to be reasonable social and political stability. Individuals should not feel that anything they do may be countermanded or destroyed at any moment or their lives put at risk. Second, there needs to be an entrepreneurial or development oriented leadership for the society, that is, not one that is primarily concerned with maintaining its own power or with expanding its area of control by conquest. The latter is likely to result in the commandeering of labor or resources for the military effort and in placing emphasis on maintenance of discipline and the established order, and not on fostering change. Similarly, theocracies or other societal structures mainly concerned with maintaining an existing order or mode of life will not provide much scope for innovation. Furthermore, a societies' leadership should be forward looking so that problems are identified quickly, before they have become crises, and responses developed before disruption is caused. 55 The Two Cases in the Context of the General Model How do the findings of the two case studies relate to the model suggested in figure 2 and table 2? The physical preconditions in Machakos were not encouraging. Even in the more settled areas rainfall per crop season is relatively low (about 500 millimeters) and highly variable. Drought seasons, or sequences of them, are common. In the nineteenth and early twentieth century, severe droughts, which resulted in heavy loss of livestock and people, were not infrequent. If the rainfall had been in the 1,100 millimeters to 1,600 millimeters range, as in the higher potential districts of Kenya, the task of overcoming the incipient disaster faced in the 1930s would certainly have been mitigated. In this case the economic conditions undoubtedly helped to offset the difficulties of the physical environment. While transport within the district was not all that it might have been, external links and access have been good. Proximity to Nairobi gave access to a significant market as did the accessibility to external markets assisted by the internal road and rail and external sea and air transport links developed over the course of this century, and by the open trading regime fostered by successive administrations in Kenya. Looking at the social situation, it is intriguing that the 1930s, 1940s, and 1950s in Machakos were far from stable, although few may have feared for their lives or property. During World War II significant numbers of younger men were absent in military service and this imbalance continued after the war as external employment was sought both for income and to avoid either forced labor for land conservation or getting involved in the insurrection of the time. This caused great disruption, but did have some positive effects. Those who went abroad were exposed to new ideas and a number of new technologies resulted from that experience. In addition the situation meant that women had to take a more active role in the society and as a result the nature of some of the traditional self-help groups changed reflecting this. Women's groups were particularly active in soil conservation work. Overall, however, there has been a fair degree of continuity in overall government economic and social policy with a general orientation toward private business, commercialization of agriculture, and individual rights over land. In this context it has proven possible for overall leadership in society to evolve and broaden in a direction conducive to development. Turning to the Ethiopian case, it should be noted that the analysis was based on the assumption of a low, and generally constant level of technology across the country, which was used as the basis of the estimation of carrying capacity. In other words, it showed that, in conditions where carrying capacity is fixed, when population exceeds this carrying capacity, erosion is increasingly likely. As noted earlier, at the time the data were generated the general conditions in the countryside fostered subsistence production and discouraged investment and innovation. In that sense the conditions may have approximated a constant technology across regions much more closely than would be the case in most countries. The Machakos study indicated the ability of a society, under appropriate conditions, to respond to population pressure and adopt new technologies, increase the carrying capacity, and modify the farming system in ways which do reduce degradation. Some authors (for example, Boserup 1965; Simon 1977, 1992) have gone so far as to suggest that population growth itself, in creating the pressure for intensification, is a factor impelling agricultural change and increased productivity. If such a process had worked perfectly, one would not expect to have found a significant relationship between population pressure and land degradation, even in a cross-sectional analysis such as that carried out in Ethiopia. As noted above, the overall policy environment in Ethiopia was hardly conducive to agricultural innovation in the period prior to the analysis. In addition, a sociological survey carried out among Ethiopian farmers indicated that farmers were aware of the degradation problem, but not the underlying causes (Constable 1984). Awareness of the causes 56 of degradation is indispensable for the development of effective technologies and, under such conditions, there is a clear gap for extension efforts to fill. Conclusions and Implications for Public Policy The results of the Ethiopia study suggest that under the set of conditions, which have recently been experienced there, there is a probability of erosion with increasing population pressure. However, the Machakos study showed that under appropriate conditions populations can adapt to increasing population pressure and avoid land degradation. The two studies suggest strongly that, in a low income region where population pressure is raising concerns about land degradation, this problem is unlikely to be resolved by a narrow focus on improved land management technologies. More general economic and social changes will be necessary to create conditions under which there are clear benefits to land users in changing technologies. This paper has in part discussed the types of physical, economic and social/political conditions that appear to be conducive to a positive outcome. In our report on land resource management in Machakos (English, Mortimore, and Tiffen 1993), based on the results of the study, we recommended a number of policies to foster improved land resource management. These may be summarized: e Primary emphasis should be given to measures which will assist in raising the value of farm products at the farm gate (for example, improved road access, and elimination of marketing bottlenecks or unnecessary controls and costs), and in widening the range of economically and technically viable land use options for farmers (for example, through experimentation on potential new crops and livestock, including focus on their impact on land resource management). * These agricultural research and extension efforts might be better placed on a range of possible technologies rather than a very limited number of 'best' technologies in order to increase the range of choice available to farmers. * Research by other than official government stations should be encouraged. * Emphasis should be placed on other measures that will facilitate the development of other economic activities, for example, the extension of transport, electricity, and telecommunications networks. X Where, because of remoteness or physical limitations to production, few if any viable land uses can be developed, emphasis should be placed on facilitating outmigration on a seasonal or permanent basis, or on other measures to improve economic and social links with the rest of the economy. To what extent can policy further encourage such a situation? If we return to figure 2, many of the changes posited in that model are clearly impacted upon by policies or actions of the type outlined in the previous paragraph. For example (the letters of the following paragraphs refer to the notations in figure 2): (a)--The development of greater accessibility to goods and labor markets will be dependent on government action on: transportation infrastructure; and fostering the trading system (avoidance of impediments such as restrictive licensing, security road blocks, and so forth, and development of standards and improvement and dissemination of market information). (b)--The translation of the reduced per capita costs of infrastructure provision into more accessible knowledge and wider skills is a function of educational and training policy. (c)--The translation of access to increased demand outside the region and internationally will be encouraged by the avoidance of impediments to interregional or international trade. 57 (d)--Support to agricultural research and extension will assist in the generation of new technology, especially once increased accessibility to ideas and skills has helped in the acceptance of the notion of the potential benefits of change. (e)--The real increase in labor supply resulting from increase in population will be dependent on the health and nutritional status of the population. Public efforts could be significant here at a minimum in reducing outbreaks of epidemics or mitigating impact of droughts or other disasters. (f)--Increased sales of local products will translate into increased diversification of the local economy if this is not impeded by excessive registration, licensing, or other barriers to start- up. (g)--Investment in land is impeded by lack of security of occupancy. This will require appropriate tenure policy. These examples undoubtedly do not exhaust the list of possible government actions which might encourage or impede the process of change outlined in figure 2, but illustrate the pervasiveness of the potential impact of government. A number of interesting points may be noted: (a) comparing these impact areas with the preconditions outlined in table 2, it is noticeable that the impacts relate primarily to what have been defined here as the economic preconditions, to a lesser extent to the social ones, and very little to the physical conditions (or their amelioration); and (b) that the points of impact are clustered in the early stages of the chain, that is, education, basic health, transport, and the stimulation of trade, are essential to get the process moving and are highly susceptible to governmental action or inaction. An approach of this type has a number of positive aspects from the point of view expressed in this paper. First, it is aimed at producing a framework that will facilitate change and not protect a status quo. Second, it is geared toward creating developmentally oriented interactions between individuals and groups, in which the relationships between parties are balanced and freely arrived at, but facilitate a move in a commercial direction. Third, they encourage the development of an open structure which is the effective basis of development and of the ability to adapt to change. References Boserup, Esther. 1965. The Conditions of Economic Growth. London; Allen and Unwin. Constable, 1984. "Resources for Rural Development in Ethiopia." EHRS Working Paper 17. Rome: FAO. English, John C., Michael Mortimore, and Mary Tiffen. 1993. "Land Resource Management in Machakos District, Kenya, 1930-1990." World Bank Environment Paper No. 5. Washington, D.C. Grepperud, Sverre. 1993. "Population-Environment Linkages: The Case of Ethiopia." ENVPE Working Paper forthcoming. Hansen, Stein. 1992. "Population and the Environment." Unpublished Paper presented to the African Development Bank. 58 Shanmugaratnam, N. and others. 1992. "From Natural Resource Degradation and Poverty to Sustainable Development in Malawi." Unpublished Draft Report for World Bank. Simon, Julian. 1992. Population and Development in Poor Countries. Princeton: Princeton University Press. . 1977. The Economics of Population Growth. Princeton: Princeton University Press. Conservation Tillage as a Tool to Conserve Soil, Moisture, Energy, and Equipment in Large and Small Crop Production Systems John F. Hebblethwaite* Introduction Driven by the 1985 Farm Bill, a part of the 1985 Food Security Act, farmers in the United States are progressing on schedule toward implementation of a soil conservation plan when planting crops on highly erodible acres. Complete implementation of this plan is required by 1995 if farmers are to participate in U.S. Department of Agriculture benefit programs. The plan requires that farmers use reduced tillage systems or no-till to maintain adequate plant and crop residue on the soil surface at all times so that annual soil loss will be reduced to a level where soil productivity is maintained. During the course of implementation farmers are discovering that conservation tillage and no-till planting, in addition to soil and moisture conservation, also have significant economic benefits. These benefits include significant reductions in labor, equipment, and fuel costs, together with more timely planting of crops. On average, yields in both conservation and no-till systems can be similar to those under traditional cultivation systems using the plow. Conservation tillage is a growing trend worldwide. No-till is already being applied on large acreages of soybeans in Brazil and Argentina and reduced tillage in small grains in Canada and Australia. Improvements in seed drills and planters and the availability of more effective herbicides for weed management have made this possible. Conservation tillage and no-till can be effectively applied in both large and small production systems. The smaller investments needed in tractors and equipment for conservation tillage could help greatly in revitalizing crop production in the Commonwealth of Independent States (CIS). In both the CIS and emerging countries of Africa, conservation tillage could play a major role in reducing rampant erosion of soil. Sustainability of Agriculture The major threat to sustainability of worldwide agricultural production is soil erosion and resultant infertility. It is estimated that soil erosion and infertility are degrading 30 percent of rainfed cropland in Central America, 17 percent in Africa, 20 percent in Southwest Asia, and 36 percent in Southeast Asia. In the former Soviet Union it is estimated that two-thirds (about 152 million hectares) of total arable land is affected by erosion and that over 50 percent of the soil profile has been eroded on 64 million hectares of this land. Crop yields in affected areas are estimated to remain, on the average, 20 percent below their yield potential due to erosion. It is estimated that the United States still loses roughly 3 billion tons of topsoil from cropland every year. John Hebblethwaite, Director, Conservation Tillage Systems, The Agricultural Group, Monsanto Company. 60 Conservation Compliance and Soil Productivity In 1985 the U.S. government introduced legislation that mandates the biggest attack on soil erosion caused by water and wind in the history of the United States. The 1985 Farm Bill, as part of the Food Security Act, mandates that farmers develop and implement a soil conservation plan when planting crops on highly erodible acres if they are to continue receiving U.S. Department of Agriculture program benefits such as price and income supports, crop insurance, Farmers Home Administration loans, and so on. About 148 million acres out of 280 million acres of cropland are considered highly erodible (table 1). Farmers by 1990 had to develop a soil conservation plan for these acres for complete implementation by 1995. To comply, the maximum average annual soil loss in tons per acre must be reduced to a level where soil productivity is maintained. Driven initially by conservation compliance regulation, we have seen a fast growing trend to conservation tillage in the United States. Conservation tillage is any tillage or planting system that leaves at least 30 percent of ground covered with crop residue after planting. In the United States this is achieved through mulch-till (leaving at least 30 percent residue), ridge-till, or no-till (direct drilling). In ridge-till one-third of the soil surface is tilled to prepare planting ridges four to six inches higher than the row middles. In no- till the soil is left undisturbed prior to planting and the seed is planted directly into the crop residue. In 1992 it is estimated that about 60 percent of the highly erodible crop acres (88 million) are under compliance in the United States. The remaining 59 million acres is expected to be in compliance by 1995. Among the compliance methods mentioned, no-till has shown the most dramatic growth (table 2). Among the major crops, corn and soybeans have shown the most dramatic growth in no-till with steady growth on small grain acres (table 3). It is estimated (by Monsanto - with inputs from the Conservation Tillage Information Center (CTIC), Soil Conservation Service (SCS), dealers and farmers) that no-till soybean and corn acres could grow to 46 million by 1996 with soybeans contributing 25.5 million acres, and corn 20.8 million acres. What is driving this trend? There is no doubt that conservation compliance is the catalyst. However, in implementation farmers are discovering substantial economic benefits such as savings in equipment cost, time, and fuel usage. These savings, together with yield equivalence (to tillage) or even yield increase, have resulted in increased profit. Purdue University compared the cost of various tillage systems with the following results shown in table 4. In a separate study, the CTIC demonstrated the usage of fuel and labor with various tillage practices (table 5). It is clear from this study that no-till results in substantial savings in both fuel and labor input. Replacement of the plow with shallower tillage implements such as the chisel, which leaves as much as 50 to 75 percent of the crop residue on the soil surface (conservation tillage), will also result in considerable savings in fuel and labor. Practical farmer experience has shown substantial reductions in equipment and horsepower requirements in no-till planting and drilling systems. For example, requirements for a 1,000 acre no- till farm in the Midwest are as follows: an 80 Horsepower tractor, a 120 horsepower tractor, a spot sprayer, a forty-foot sprayer, a nitrogen tool bar with coulters, a light disc, a no-till drill, an eight- row planter with coulter and starter, and a combine. Gone is the major tillage equipment such as the plow and also the 280-350 horsepower tractors needed to pull this equipment. They have been replaced by smaller tractors, no-till planters, and light equipment that can cover the same number of acres faster. In addition to savings in input cost, it has been shown that equivalent yields, or even increased yields, can be obtained from conservation tillage systems. In a study conducted over several years by Purdue University, continuous corn and corn in a soybean/corn rotation in Indiana 61 Table 1. Crop Acreage in the United States (millions of acres) Highly Erodible Conservation Tillage Year Planted Acres Acres Acres 1989 280 141 72 1990 264 148 74 1991 281 148 79 1992 282 148 88 Source: CTIC - Cedar Package Table 2. Conservation Tillage in All Crops in the United States (millions of acres) Year No-till Ridge-till Mulch-till 1989 14.2 2.7 54.9 1990 18.1 3.0 52.4 1991 20.6 3.2 55.3 1992 28.1 3.4 57.0 Source: CTIC - Cedar Package Table 3. No-Till of Major Crops in the United States (millions of acres) Year Corn Soybeans Small Grains 1989 5.36 4.82 2.43 1990 6.41 5.99 2.82 1991 7.56 7.92 3.25 1992 10.85 11.31 3.64 Source: CTIC - Cedar Package Table 4. Cost per Acre for Various Tillage Systems System $IAcre No-till 75.50 Mulch-till 87.00 Conventional-till 86.00 Source: Purdue PEPS Program - Successful Farming (1991). 62 Table 5. Comparison of Inputs for Various Tillage Practices Tillage Fuel Labor Practice gallons per acre hours per acre No-till 1.50 0.45 Till-plant 2.75 0.90 Rotary-till 3.00 0.88 Disc 3.00 0.85 Chisel 3.30 0.89 Plow 5.25 1.22 Source: CTIC gave yields in no-till that were equivalent to, or even slightly higher, than those obtained in conventional tillage systems when applied to a poorly structured silt loam soil for four years (table 6). In contrast, on a poorly drained silty clay loam, ridge-till planting or reduced tillage (chisel) was necessary as the conservation tillage practice to obtain equivalent yields. Farmers with a number of years of experience have obtained excellent yields of corn and soybeans by no-till on well-drained soils. They have also successfully no-tilled poorly drained soils using good management techniques such as winter cover crops to drain moisture from the soil in spring. The cover crop is then killed in spring with a herbicide. Soybeans are very responsive to no- till, and yields, as demonstrated by the CTIC, University of Missouri, and University of Kentucky are similar to those obtained in conventional tillage practice. Table 6. Corn Response to Tillage and Rotation on Two Soil Types in Indiana (average yields 1983-86 in bushels per acre) Tillage System Silt Loam Silty Clay Loam Continuous corn: Plow 122.50 161.03 Chisel 128.00 156.10 Disk 123.30 - Ridge-till 122.40 156.10 No-till 131.00 132.40 Corn following soybean: Plow 121.90 178.03 Chisel 120.80 176.70 Disc 127.70 - Ridge-till 130.10 180.30 No-till 138.60 169.62 Source: Griffiths and others (1988). In a five-year study in Manitoba, Canada a small grain and canola rotation showed dramatic profit improvements for minimum-tillage and no-till over conventional tillage (table 7). Slightly higher expenses in no-till and minimum-tillage plots are attributable, largely, to the additional 63 herbicide cost. However, recent reductions in herbicide price have narrowed these cost differences and make average net returns from no-till and minimum-tillage even more attractive. Increased yields and improved gross returns in no-till and minimum-tillage were attributable to improved soil moisture conservation. In the Western Plains of the United States and the Canadian prairies, soil moisture conservation is an important component for maximizing yield. Leaving upright stubble after harvest is important for trapping snow in the winter and surface crop residue reduces evaporation in the summer. Table 7. Five-Year Manitoba Study of a Small Grain and Canola Rotation Min-till Conventional Year Crop No-till BulAc tillage 1986 Wheat 56.00 56.00 54.00 1987 Barley 77.30 73.20 66.20 1988 Barley 77.10 69.60 63.20 1989 Canola 11.10 8.20 6.50 1990 Wheat 64.20 61.30 54.90 $/Acre Average gross return 146.91 136.60 124.18 Average total expense 116.46 113.63 110.78 Average net returns 30.45 22.97 13.40 Source: Zero Till Production Manual - Manitoba/North Dakota. Zero Tillage Farmers Association (1991). Conservation Tillage in Other Countries Soybeans produced by no-till is a rapidly growing trend in both Brazil and Argentina. It is estimated that as many as 3 to 3.5 million acres are already under no-till. No-till is an important tool to control soil erosion on the sandy slopes of Rio Grand Du Sol and Parana of Brazil. In Australia it is estimated that as much as 12 million acres of small grains and fallow are under conservation tillage practice, while in Canada conservation tillage is practiced on 17 million acres or 24 percent of the cropland. Nearly 5 million acres is already under no-till in Canada. Factors Contributing to the Success of No-Till and Conservation Tillage There is no doubt that we have seen tremendous benefits from conservation tillage and no-till farming; but what has made this possible? Two critical factors have contributed to this: (a) availability of well-constructed no-till planters and drills capable of handling and planting into surface crop residue; and (b) availability of new herbicide technology able to manage weeds effectively in high residue situations. Most of these herbicides have very attractive environmental properties and the newer ones can be applied at lower dosages; not several pounds per acre, but in some cases grams per acre. These 64 properties, combined with less water runoff in conservation tillage or no-till, will greatly contribute to much reduced surface or groundwater residues. Roundup0 Herbicide by Monsanto has been the cornerstone of no-till and conservation tillage practice in the United States, Australia, Canada, Brazil, Argentina, and Europe. Because of its broad-spectrum activity (annuals and perennials) and versatile application timing to weeds, it has become the preferred herbicide for cleanup of weeds prior to no-till or reduced tillage. In no-till it allows the crop to get off to a clean start or its burndown of weeds helps facilitate shallow tillage. Glyphosate, the active ingredient of Roundup0, is rapidly biodegraded in the soil and has low toxicity. These desirable properties, together with the fact that it is tightly bound by soil particles, makes it an environmentally attractive choice in conservation tillage. Wide use is not expected to result in surface or groundwater residue problems. RoundupO can be mixed with residual herbicides or be used in a herbicide system with other herbicides for complete preplanting and post-crop emergence weed control in conservation tillage. Conservation Tillage in Developing and Less Developed Agricultural Systems There is no doubt that the conservation tillage and no-till technology, which has been successfully applied in the United States, Canada, Brazil, Argentina, and Australia can also be successfully applied in the inefficiently resourced and declining agricultural systems of the former Soviet Union and Eastern Europe. As in the United States, these systems could greatly reduce these countries' dependence on heavy tillage equipment and their large and inefficient high horsepower tractors, while at the same time reducing serious soil erosion to sustainable levels. While people and expertise are available to transfer this technology to the former Soviet Union and Eastern Europe, a major challenge stands in the way of its adoption; namely, availability of hard currency to purchase smaller tractors and no-till drills and planters. It would be best if these countries could produce the planters and drills locally, as well as smaller tractors. Outside investment in these production facilities, accompanied by expertise, could yield very productive results. However, such investment would need to be tied to the availability of other inputs such as herbicide and fertilizer. Availability of no-till planters and smaller tractors and the reduction of heavy tillage equipment could also facilitate a move to smaller farms and, ultimately, privatization of agricultural production. No-till, or conservation tillage, can also be applied to smallholder production systems in Africa. We have seen successful applications in Kenya, particularly in pasture and corn rotations. In this system, kikuyu pasture (Pennisetum clandestinun) is killed with RoundupO Herbicide just prior to seeding corn with a hand hoe. Corn can be planted into a narrow strip prepared with a hoe in the dead kikuyu pasture. The dead grass in the interrow protects the soil from wind and water erosion and acts as a barrier to weed germination. This system of conservation tillage could, with further research, be adapted to many annual cropping situations in Africa. Herbicides have also contributed to tillage elimination in the interrows of plantation crops in smallholder production. For example, RoundupO has been successfully used to control annual and perennial weeds in the interrows of coffee on the slopes of Mount Kenya. The soil is left undisturbed and the decaying vegetation protects the soil from erosion. 65 Conclusions Conservation tillage is a growing trend worldwide. In the United States legislation has been the catalyst. However, in the process farmers have discovered that input cost can be reduced dramatically through reductions in labor, equipment, and fuel costs. At the same time farmers have discovered that corn and soybean yields can be maintained at levels which will result in improved average net returns per acre from conservation tillage and no-till. In the drier Western Plains of the United States, the Canadian prairies, and Australia conservation tillage and no-till have resulted in improved soil moisture conservation that has translated into increased yield. Modem no-till planters and new herbicides, which effectively manage weeds in no-till and conservation tillage situations, have undoubtedly contributed to this growing trend. These economic benefits together with dramatically reduced soil erosion could also apply to the vast acreages of the former Soviet Union and Eastern Europe. The challenge in these countries is the availability of hard currency to purchase no- till drills and planters. The best solution might be local production through foreign investment and here the World Bank could play a key role. Conservation tillage can also be applied to smallholder production systems. Planting by hand into vegetation managed with herbicides with a minimal amount of surface tillage has been successfully practiced in Kenya and Indonesia. This practice needs development in other countries through such organizations as Winrock International, Global 2000, and so on. Because of the successful application of conservation tillage and its potential contribution to sustainable agriculture, the World Bank needs to give serious consideration to investments in this area. References The Citizens Network for Foreign Affairs. 1992. International Markets and the Environment: The Stake of U.S. Agriculture. CTIC (Conservation Tillage Information Center). 1992. National Crop Residue Management Survey. Griffith, D.R, E. J. Kladivko, J. V. Mannering, T. D. West, and S. D. Parsons. 1988. "Long-Term Tillage and Rotation Effects on Corn Growth and Yield on High and Low Organic Matter, Poorly Drained Soils." Agronomy Journal 80:599-605. Hayes, William A., and H. M. Young. 1982. No-tillage and Minimum 7illage Farming. No-Till Farmer, Inc., Brookfield, Wisconsin. Manitoba/North Dakota Zero Tillage Farmers Association. 1991. Zero Tillage Production Manual. Soybean Digest. 1988. "Countdown to Compliance." Special Supplement. World Bank. 1992. "Review of Food Policy Options and Agricultural Sector Reforms." Joint Report to the Russian Federation and Members of the Commonwealth of Independent States. I I Moisture Management in Semiarid Temperate Regions B.A. Stewart, O.R. Jones, and P.W. Unger* Introduction The temperate regions are located between the Tropic of Cancer and the Arctic Circle or between the Tropic of Capricorn and the Antarctic Circle. These regions contain vast areas of land in semiarid climatic zones, often classified as areas where the precipitation/potential evapotranspiration ratio is between 0.2 and 0.5 (UNESCO 1977). Dryland farming is often practiced in these regions, but water conservation practices are essential for successful cropping systems. The Canadian Prairie Provinces, U.S. Great Plains, Southern Australia, and parts of Argentina, China, and the Commonwealth of Independent States are examples of dryland farming areas in semiarid temperate regions. Dryland farming systems emphasize water conservation, sustainable crop yields, limited inputs for soil fertility, and wind and water erosion constraints. The three components of a successful dryland management system are (a) retaining the precipitation on the land, (b) reducing evaporation, and (c) utilizing crops that have drought tolerance and fit the rainfall pattern. Although these components have been known for centuries, new technologies are developing that improve moisture management in these water deficient areas. Some of these technologies will be presented and the principles on which they are based will be discussed. Semiarid Climate Although semiarid zones are often classified by the ratio of precipitation/potential evapotranspiration, semiarid is a comparative term implying a moisture state intermediate between truly arid conditions and others that are more humid. Semiarid regions typically receive substantial precipitation for at least a few months of the year, enough to bring soil moisture up to levels sufficient to produce amounts of biomass that far exceeds that produced in arid regions. The water balance values for annual cropping of wheat at three semiarid locations are presented in table 1. An understanding of these data will provide the base for discussing the development and implementation of new technologies for improved cropping and soil management systems. The percentage of total rainfall that was used for evapotranspiration was similar for all three locations, approximately 65 percent. Evapotranspiration is the combined loss of water from transpiration and evaporation from the soil surface during the period when the crop is growing. The fallow period is the time between harvesting the crop and seeding the subsequent crop. For the data presented in table 1, evapotranspiration values were calculated by adding the growing season precipitation amounts to the change in the amount of available water held in the soil at seeding time and at harvest time. In all locations, soil water was decreased significantly during the growing season, and increased during the fallow period. However, the change was considerably less for the Laboratory director and soil scientists, respectively, at the U.S. Department of Agriculture Conservation and Production Research Laboratory, Agricultural Research Service, Bushland, Texas. 68 Table 1. Water Balance Values for Annual Cropping of Wheat at Three Semiarid Locations Wheat Fallow Total Wheat Fallow Total Wheat Fallow Total ----Texas, U.S.A.---- ---Shaanxi, China--- --New South Wales, Australia-- Precipitation (mm) 256 202 458 181 213 394 280 280 560 Evapotranspiration (mm) (ET) 293 293 264 264 360 360 Soil water change (mm) -37 37 -83 83 -80 80 Evaporation and runoff (mm) 165 165 130 130 200 200 Potential evapotranspiration 1,140 740 1,880 475 408 883 (mm) (PET) ET/PET (%) 26 56 Precipitation/PET (%) 24 45 ET/Precipitation (%) 64 67 64 Yield 0.90 1.25 2.40 Water use efficiency .33 .47 .67 (WUE) kg_m3 _M3 Source: Adapted from unpublished data, O.R. Jones, Bushland, TX; Shan Lun and others (1992); and Comish and Pratley (1991). 69 Texas location. There was less precipitation at this location during the fallow period, and thepotential evapotranspiration was very high, resulting in only 37 millimeters of storage, compared to storage at the other two locations of 80 millimeters or more. The Texas location is the most arid of those presented in table 1. Although total precipitation was more for the Texas site than for the China site, the amount of actual evapotranspiration was only 26 percent of the potential evapotranspiration for the Texas location, compared to 56 percent for the Shaanxi site. The China location had a much higher yield, and a water use efficiency of 0.47 kg m-3, compared to 0.31 kg mn' for the Texas location. The yield and water use efficiency values were low for both sites, but the Texas site values were extremely low. Water use efficiency values for wheat grown in humid regions or under irrigation often exceed 1.25 kg mt' and values as high 1.9 m-3 are reported in the literature (Musick and Porter 1990). The data presented in table 1 are average values for several years. One of the difficulties with crop production in semiarid regions is the extreme variation of precipitation and, therefore, crop yields among years. Annual precipitation in these regions commonly ranges from about 50 percent of average for a dry year to about 200 percent of average for a wet year; yields vary from 0 to about three times average. Much of the precipitation in semiarid regions also occurs during high intensity storms and runoff can be significant. Runoff, combined with evaporation from the soil surface during the fallow period, resulted in a loss of one-third of the precipitation for all three locations discussed in table 1. The development and implementation of technologies that reduce these losses can greatly increase yields. This is illustrated in figure 1 showing the relationship between yield of wheat grain and seasonal water use (evapotranspiration) for a location in Texas and one in northwest China. Data for both locations indicate that about 200 millimeters of water use is required before any grain is produced, but for each additional millimeter of water use, about 12 kg ha7' of grain is produced at the Texas site and about 25 kg ha-' for the China location. As already discussed the Texas site is much more arid than the China site, resulting in a lower water use efficiency. These relationships clearly show the great impact that technologies, which increase the amount of water available for crop use, can have on grain yields. Technologies for Increasing Plant Available Water Several technologies for increasing plant available water are discussed in this section. They include lengthening the fallow period, mulches, tillage, and runoff control and conservation. Lengthening the Fallow Period One of the oldest, and most controversial, technologies for increasing plant available water is lengthening the fallow period. This is generally called summer fallow, defined as a practice wherein no crop is grown and all plant growth is controlled by cultivation or chemicals during a season when a crop might normally be grown. Proponents have emphasized the water conserving, weed controlling, and crop yield stabilizing virtues, whereas critics have emphasized the inefficiency in soil water storage and the wind and water erosion and declining organic matter problems associated with fallow. Only about 15 to 20 percent of the precipitation that occurs during the fallow period is stored in the soil profile. The remainder is lost as runoff and evaporation and, on some soils, as drainage below the root zone. The data presented in table 2, from Bushland, Texas, illustrate the effect that lengthening the fallow period has on increasing soil water storage. When wheat is grown annually, 70 the fallow period is about three months to four months and, on average, 37 millimeters, or 18 percent, of the 202 millimeters rainfall that occurs during the fallow period is retained as stored soil water as previously discussed in table 1. By changing the cropping system from annual wheat to wheat/sorghum fallow, the length of the fallow period between crops is increased to about eleven months. However, only two crops, one wheat and one grain sorghum, are produced every three years as compared to two wheat crops for the annual wheat system. In this system wheat is seeded about October 1 and harvested about July 1. Grain sorghum is seeded about June 1 the following year and harvested in November. After an additional eleven-month fallow period, wheat is again seeded. The average amount of water stored at time of seeding wheat was increased to 86 millimeters as compared to 37 millimeters for the annual wheat system. The fallow period is even longer when a wheat fallow system is used, producing only one crop every two years and lengthening the fallow period to fifteen months to sixteen months. In this system the average amount of plant available soil water at seeding time is increased to 98 millimeters. While these fallow systems increase soil water contents at seeding and, therefore, increase the amounts of water consumed by the wheat crop from 293 millimeters for annual cropping to 329 millimeters and 354 millimeters for the wheat/sorghum fallow and the wheat fallow system, respectively. Although these amounts are relatively small, the associated yield increases are very significant because of the yield-water use relationship presented in figure 1 and discussed earlier. Each millimeter of additional water use has the potential of increasing yield about 12 kg ha-', so lengthening the fallow period often increases yields by 50 to 75 percent, and in some cases can double yields. The water use efficiency, expressed as grain/evapotranspiration is significantly increased, but a much smaller amount of the total precipitation is actually used by the crop. For example in the wheat fallow system, only 354 millimeters of the total 916 millimeters of precipitation that occurred during the two-year system was actually used by the growing crop. The other 562 millimeters was lost as evaporation, runoff, and perhaps some drainage. Therefore, summer fallow is very inefficient for conserving precipitation as stored soil water, but efficient for increasing and stabilizing grain yields. Because of these divergent effects, summer fallow remains a subject of controversy. Figure 1. Relationship Between Wheat Yield and Seasonal Evapotranspiration for Two Semiarid Regions. 8- N. W. China 6- 6 - // Texas High Plains 0- - / Cd 2 - 0 100 200 300 400 s00 600 700 800 Seasonal Evaporation (mm ) 71 Perhaps the biggest concern about summer fallowing is its effect on soil degradation. Until herbicides became available in recent decades, tillage was the only means of controlling vegetative growth during the fallow period. Consequently, it was not uncommon for a field to be tilled eight to ten times during the fallow period. Intensive and frequent tillage buries most of the crop residues and hastens the decomposition of crop residues and soil organic matter. Cultivation increases biological activities in the soil, often as a result of better soil aeration. But cultivation also exposes fresh topsoil to rapid drying and, after each drying, a burst of biological activity occurs for a few days following rewetting (Allison 1973). This is because the drying process releases organic compounds, probably from the breakdown of soil aggregates that are bound together by humic materials. Considerable organic nitrogen is mineralized as ammonia and later oxidized in large part to nitrates. Other nutrients are also made available from the decomposition of organic matter. This is particularly true for phosphorus because much of the phosphorus in soils is present in organic forms. The nutrients released as a result of tillage are readily available to growing plants and increased yields are generally obtained. Therefore, in addition to increasing water storage, summer fallowing also increases available soil nutrients. However, unless the organic matter supply is replenished by plant residues or manures, the system is not sustainable. This is the situation for many soils of the world located in arid and semiarid regions and increased attention to the problem is critical. It is also the underlying principle that resulted in the infamous "Dust Bowl" that occurred in the U.S. Great Plains during the drought years of the 1930s and considered by many as the worst ecological disaster ever exacerbated by man. T'he Great Plains region was largely settled in the early 1900s by farmers who migrated from the humid areas of the eastern United States and brought with them their clean-tillage tools and experiences. These worked well the first few years after cropping began because the native soil organic matter content was high and the precipitation during the period of the "big plowout" was above average. However, when annual precipitation decreased to average and below, the annual net loss of soil organic matter accelerated and led to increased vulnerability to wind erosion. The moldboard plow, and many other intensive tillage implements, were developed in Europe where soil organic matter content of soils is high, and the organic matter level can be maintained at a high level because of relatively high precipitation amounts that produce large amounts of biomass and cool temperatures that slow the rate of decomposition. In arid and semiarid regions, high temperatures accelerate the rate of decomposition and the lack of precipitation severely limits biomass production so organic matter loss can be rapid and severe. Summer fallow, particularly during years of above average precipitation, can infiltrate much more water than can be stored in the soil profile. This can result in substantial amounts of water moving through the profile removing nutrients and, if salts are present, they will be leached and cause saline seeps in certain situations. This has been a significant problem in parts of the northern Great Plains of the United States where spring wheat is the dominant cropping system. The length of the fallow period in spring wheat fallow is about twenty months during each two-year cycle. Saline seep problems are also widespread in parts of Australia. Summer fallowing has also been used extensively in Australia and China. Cornish and Pratley (1991) stated that fallows have had a long and often sorry history in Australia. The practices described above that were so successful when first implemented in the U.S. Great Plains were imported by Australian farmers in the early 1900s. The primary practices involved deep plowing and frequent harrowing to produce a dust mulch. The plowing was thought to increase the waterholding capacity of the soil, while the dust mulch supposedly prevented water rising to the soil surface by capillary action and evaporating. Subsequent research showed that the major loss of water from soils was through transpiration by weeds and that the benefits of dust mulching were largely due to weed control. These technologies were used to extend the limits of wheat growing into the marginal 250- 72 Table 2. Water Balance for Various Cropping Systems at Bushland, Texas Continuous wheat (one crop annually)a Wheat Fallow Total --------------------- mm--------- - Precipitation 256 202 458 Evapotranspiration 293 293 Soil water change -37 37 Evaporation (and runoff) 165 165 Two crops in three yearsb Wheat Fallow Sorghum Fallow Total ----------- --_-----mm----------------- Precipitation 256 462 241 416 1,375 Evapotranspiration 329 286 615 Runoff 13 25 27 43 108 Soil water change -86 86 -72 72 Evaporation 351 301 652 One crop in two yearsc Wheat Fallow Total -------------- mm---------------- --- Precipitation 256 660 916 Evapotranspiration 354 354 Soil water change -98 98 Evaporation (and runoff) 562 562 Fallow period between crops is three to four months. Runoff was not measured but would be minimal under annual cropping. , b Fallow periods between crops are about eleven months. c Fallow period between crops is fifteen months to sixteen months. Runoff was not measured but was a minor portion of the total. Source: 0. R. Jones, personal communication; Johnson and Davis (1972). 73 millimeter to 400-millimeters rainfall zone of the South Australian, Victorian, New Soulth Wales, and Western Australian Mallee. Long fallow periods (fifteen months) were used and the frequent cultivation of these light-textured soils resulted in soil structural breakdown, fertilitv decirie, anid ultimately, catastrophic erosion. Li Shengxiu and Xiao Ling (1992) sunmmarized some cf the results from the Loess Plateau Region of China and concluded that fallowing was a good nrace for the drylands. Summer fallowing was usually combined with summer deep plowing for c.ontrolling weeds, keeping the soil loose, and increasing soil infiltration. As a result soil water storage was increased. In addition available plant nutrients, especially nitrates, accumulated in the profile and stimulated growth of the subsequent crop. They did not stress the negative effects sucn as organic m?ater decline and deteriorating soil structure that were discussed above. This may be because Chinese farmers have historically used organic wastes on their fields and this may offset the otherwise negative effects. The steppe area of northern Kazakhstan is another region where fallow has been widely practiced. However, Souleimenov (1992) concluded that fallow in this region of about 350 millimeters of precipitation was not justified. Research showed that the available water storage prior to seeding wheat was only slightly higher for the fallow fields than those for whe-a after fallow or for continuous wheat (table 3). Fallow was adopted in this region in 1966 based largely on s ome selected data of the state farms for extremely dry years (1962, 1963, and 1965). The decision aso was influenced by data and experiences from the Canadian prairies. Souleimenov (1992) recommended that most fallow be discontinued with the more marginal lands being returned to grass and the better lands cropped annually. He also pointed out the benefits that such a system would have oa t-ie environment. Weed infestation, wind erosion, and other soil degradation processes have been widely experienced in the region where fallow systems were the dominant practice. Table 3. Available Moisture Storage in the Layer of 0-100 Centimeters Prior to :pring Vi 'nea Sowing Depending on Preceding Crop and Cultural Practice, Millimeters (Average for 1986-1990) Cultural practice Preceding crop Simplified Conumon irproved Bare fallow 110 137 5 Wheat after fallow 98 133 2 153 Continuous 102 135 155 Source: Souleimenov (1992) Mulches The Dust Bowl of the 1930s, described earlier, led to the development of stu,bble rmulc&hirg. S&aubbie mulching uses V-shaped sweeps or blades that are pulled flat about 10 centimeters beneath the soii surface. This operation cuts plant roots and kills the weeds but does not invert the s0il. The.re fore, most of the crop residue is left on the surface where it can serve as a mulch to prevent wind and water erosion, and slow evaporation losses. Only about 15 percent of the residue is burled hy a sweep tillage operation, so there is substantial residue remaining on the surface even after toree to four operations, which is often done between the time a crop is harvested and a subsequent crop is seeded. A rodweeder, a square rod about one inch thick that turns about five centimeters to ten 74 centimeters beneath the surface as it is pulled, is another tool that is sometimes used to kill weeds without intensively tilling the soil. A rodweeder operation can sometimes bury less that 10 percent of the residue present on the soil surface. In recent years herbicides have been used to partially replace tillage with reduced tillage, and completely replace tillage with no-tillage. Although stubble mulching was developed to address the wind erosion problem, it soon became evident that mulches had beneficial effects on soil water storage. The increase in soil water storage generally is attributed to increased infiltration and reduced evaporation. However, the degree that each of these factors contribute varies with specific conditions. Cornish and Pratley (1991), working on clay soils in Australia, found that plant residues on the soil surface caused a major reduction in runoff, principally by protecting soil surfaces that were prone to crusting from raindrop action. Fallow efficiencies in Queensland were increased from about 21 to 29 percent, almost entirely because of reduced runoff. They reported that only about 4 t ha-' of crop residues were needed to gain the maximum improvement in infiltration (figure 2). A crop of wheat yielding about 2 t ha7' of grain will produce about 4 t ha71 of residue. Because the national average wheat yield of grain is about 1.5 t ha-', there is sufficient residue on most fields to gain most of the potential benefit from increased infiltration if the residue is left on the soil surface. Unger (1978), working on a clay loam soil in the southern U.S. Great Plains, also found very significant increases in soil water storage when crop residues were maintained on the soil surface. The residues enhanced water infiltration and suppressed evaporation, thus providing more water for the subsequent crop (table 4). In contrast to the Australian study, however, soil water storage values continued to increase with each additional amount of residue. Following the eleven-month fallow period, dryland grain sorghum was grown and the yields reflected the increases in soil water storage. Table 4. Straw Mulch Effects on Soil Water Storage During an Eleven-Month Fallow, Water Storage Efficiency, and Dryland Grain Sorghum Yield at Bushland, Texas Water Storage Grain Total Mulch rate storage2 efficiency2 yield water use WUEb (mg ha1') (mm) (percent) (Mg ha7') (mm) (kg m3) 0 72 c 22.6 c 1.78 c 320 0.56 1 99b 31.1 b 2.41 b 330 0.73 2 100 b 31.4 b 2.60 b 353 0.74 4 116 b 36.5 b 2.98 b 357 0.84 8 139 a 43.7 a 3.68 a 365 1.01 12 147 a 46.2 a 3.99 a 347 1.15 a Water storage determined to 1.8-meter depth. Precipitation averaged 318 millimeters. b Water use efficiency (WUE) based on grain produced, growing season precipitation, and soil water changes. c Column values followed by the same letter are not significantly different at the 5 percent level (Duncan's multiple range test). Source: Unger (1978). 75 Figure 2. Effect of Crop Residues on Infiltration and Soil Water Storage in an Australian Soil (from Cornish and Pratley 1991) so E so 20 / 2 :1 0 2 4 6 8 10 12 14 16 ResIdues t/ha3 Li Shengxiu and Xiao Ling (1992) summarized studies from China and results are similar to those reported above. Straw mulch significantly increased soil infiltration and reduced water loss by evaporation, thereby increasing water storage both in summer and winter (table 5). In addition to increasing soil water storage, the mulch decreased bulk density and increased the number of earthworms and soil organic matter content. Bulk density in the top zero to ten centimeters was 1.36 Mg m-3 without a straw mulch compared with 1.29 and 1.23 with 4.5 and 6.0 Mg ha-' mulch, respectively. The earthworm number per square meter in the top fifteen centimeters was two without straw mulch, but twelve, thirty-two, and thirty-four with straw mulch of 3.0, 4.5, and 6.0 Mg ha7', respectively. Organic matter content in the top layer was 1.61 percent with no straw mulch, and 1.67 and 1.76 percent with 4.5 and 6.0 Mg ha"', respectively. Soil temperature was also affected, being cooler in the summer and warmer in the winter when there was a straw mulch on the soil surface. Table 5. Amount of Water Stored (millimeters) in Different Soil Layers with Different Amounts of Straw Mulch (kilogram/hectare) Depth Without mulch Mulch 3,000 Mulch 4,500 Mulch 6,000 (centimeters) Amount of water stored 0-30 66.1 71.2 74.4 76.2 30-100 191.5 194.5 198.3 201.1 100-200 248.9 252.5 259.9 271.1 0-200 506.5 518.2 532.6 548.4 Note: Two years' average values in two locations. Source: Han Siming and others (1988). 76 Smika (1976) studied the specific effect of surface residues on soil water evaporation during a thirty-four-day period following 13.5 millimeters of rainfall in the central U.S. Great Plains. One day after the rain (figure 3), water contents to a fifteen-centimeter depth were similar with no tillage (herbicides only), minimum-tillage (combination of stubble mulch tillage and herbicides), and conventional-tillage (stubble mulch) treatments. After thirty-four days with no additional rain, water contents were greatest with no-tillage and least with conventional-tillage. Surface residue amounts during the study were 1.2, 2.2, and 2.7 Mg ha-' with conventional-tillage, minimum-tillage, and no- tillage treatments, respectively. Shikula and others (1992) reported that mulches were very useful for increasing water storage in the semiarid steppe regions of the Ukraine. On no-plow plots, plant available soil water storage values were 170 millimeters to 186 millimeters in the 0 centimeter to 150-centimeter soil profile, compared to only 132 millimeters to 154 millimeters on the conventionally plowed plots. Wheat yields were also higher on the no-plow plots. Soil loss as a result of water erosion was eighteen to twenty-three times less than for conventional tillage. In addition expenditures for labor was reduced by 40 percent, fuel by 45 percent, and total expenditures by 48 percent for the no-tillage plots. They concluded that no-plow technologies should be adopted in the semiarid steppe zones of the Ukraine as soon as possible. Although surface residues clearly reduce runoff and evaporation, it should be appreciated that these benefits diminish as the soil becomes wetter. Toward the end of a fallow, when the soil approaches maximum waterholding capacity, residues have little effect. Figure 3. Soil Water Content to a 15-Centimeter Depth One Day (A) and Thirty-Four Days (B) after 13.5 Millimeters Rainfall as Influenced by Tillage Treatments (CONV-TILL, Conventional-Tillage; MIN-TILL, Minimum-Tillage; NO-TILL, No-Tillage) (from Smika 1976) A SOIL WATER -m3/m3 B SOIL WATER - m3/m3 . a .2 .3 . t .2 .3 1 S gCONV -TILL 2 2 l 4 3 MIN - TILL E 6 _ OCNO - TILL 17 10 11 1J 15 15 77 7i11age The practice of tillage dates back to the beginning of history, and was well-established in Mesopotamia at least as early as about 2000 B.C. The first tillage tools were crude implements of stone, wood, and possibly bones and shells used by man to eliminate weeds and to chop or dig a few centimeters into the soil so seed could be planted. Later, animals were used to pull stick plows. Modern tillage systems had their origin in the 18th century when the moldboard plow was invented. A moldboard factory was opened in Scotland in 1760. Until recently, frequent and thorough cultivation was considered the mark of good farming. The seal of the U.S. Department of Agriculture has a moldboard plow as the focal point. Because the more prosperous farmers cultivated their soils very often, it was generally assumed that this was of major importance. Tillage was important because it controlled weeds and, more importantly, it released nutrients from the soil, mostly from soil organic matter. Conservation tillage systems, consisting of reduced tillage and no tillage, have received increasing attention in recent years because maintaining residues on the surface greatly reduces water and wind erosion. Conservation tillage is defined as any tillage or planting system that maintains at least 30 percent of the soil surface covered by residue after the crop has been seeded to reduce soil erosion by water. When wind erosion is the primary concern, residues or plants of other crops equivalent to at least 1.1 t ha7T of flat, small grain residue must be maintained on the surface during the critical erosion period. These amounts of residue are generally sufficient to keep soil erosion at an acceptable level. Conservation tillage systems also require less fuel than more intensive tillage systems, and increasing fuel costs also have increased interest in these systems. In semiarid regions, the emphasis on reduced and no-tillage systems has focused on water conservation. As already discussed, mulches increase infiltration and reduce evaporation, and the most practical way to create a mulch is to reduce tillage. Greb, Smika, and Welsh (1979) summarized more than sixty years of progress in wheat production in fallow systems in the central U.S. Great Plains (table 6). As the number of tillage operations decreased, marked increases in the amount of water stored during the fallow periods occurred with dramatic increases in yield. These positive effects tend to accumulate because higher yields result in more residue and increased residue results in more water storage, which translates into higher yields, creating an upward spiral. Soil physical properties also are improved. Reduced tillage also is important in semiarid regions as a means of maintaining organic matter. As discussed earlier, cultivation increases biological activities in the soil. Cultivation also exposes fresh topsoil to rapid drying and, after each drying, a burst of biological activity occurs for a few days following rewetting. This is because the drying process releases organic compounds, probably from the breakdown of soil aggregates that are bound together by humic materials. Considerable organic nitrogen is mineralized as ammonia and later oxidized in large part to nitrates. Other nutrients also are made available from the decomposition of organic matter. This is particularly true for phosphorus because much of the phosphorus in soils is present in organic forms. The nutrients released as a result of tillage are readily available to growing plants and increased yields are generally obtained. Tillage also increases the infiltration rate of most soils and this reduces runoff and often increases storage of water in the soil profile so it can be used later for plant growth. However, unless the organic matter supply is replenished, the system is not sustainable. This is the situation for many soils of the world located in arid and semiarid regions and increased attention to the problem is critical. 78 Table 6. Progress in Fallow Systems with Respect to Water Storage and Wheat Yield at Akron, Colorado Fallow water storage Wheat yield Years Tillage during fallow' (millimeter) (percent of (Mg ha-') precipitation) 1916-30 Maximum tillage; plow, harrow (dust 102 19 1.07 mulch) 193145 Conventional tillage; shallow disk, 118 24 1.16 rod weeder 1946-60 Improved conventional tillage; began 137 27 1.73 stubble mulch in 1957 1961-75 Stubble mulch; began minimum 157 33 2.16 tillage with herbicides in 1969 1975-90 Projected estimate; minimum tillage; 183 40 2.69 began no-tillage in 1983 Based on fourteen-month fallow, from mid-July to second mid-September Source: Adapted from Greb, Smika, and Welsh (1979) Runoff Control and Conservation Although precipitation is lacking in semiarid regions, high intensity rainfall events are common and runoff can be significant. Runoff can be particularly high where fine-textured soils are dominant because of their low infiltration rates. As infiltration rates decrease, runoff increases, thus accelerating erosion. Terraces are often used to control runoff and reduce erosion, but they are expensive to construct and sometimes interfere with cultural practices, particularly in regions where large machinery is used. Furrow diking, sometimes called tied-ridging, is also an effective cultural practice for retaining surface water until it can infiltrate. Furrow diking is even more effective if it is done on the contour, and it also lessens the erosion potential associated with a very large precipitation that could result in more surface water than can be retained in the basins. Seeding crop rows on the contour is a practice that is adaptive to all types of tillage, reduced tillage, and no-tillage systems and is highly recommended. A major obstacle to farmer adoption of furrow-diking technology is that there are many years in which positive results are not obtained. Data from Bushland, Texas show that average runoff during the sorghum growing season was 25 millimeters, adequate to increase grain yields about 375 kg ha (Stewart and Steiner 1990). However, over half the years had little or no runoff, and often two or three of these years occurred in sequence. Farmers often discontinue the practice before enough favorable response is obtained to convince them that they must utilize the practice each year, knowing 79 very well that they will not get any benefit in half or more of the years. As already discussed the most effective tillage methods for increasing infiltration are those that maintain crop residues on the soil surface. In general furrow diking is not needed in those situations. Effective erosion control and runoff conservation may involve a combination of practices. As mentioned earlier maintaining all or most residues on the surface can increase infiltration rates and reduce erosion if sufficient residues are present. However, for erosion control, terraces may also be recommended, depending on soil type, slope, and slope length. Fewer terraces may be required if conservation tillage systems are used. Summary Crop production in semiarid temperate zones requires specialized moisture management practices that impact one or more of the three components of successful dryland management systems identified in the introduction. * Retaining precipitation on the land by using surface management techniques such as terracing, contouring, and furrow diking; by using tillage to break crusts or relieve compacted layers in the soil, thus improving infiltration; or by adopting reduced or no- tillage management systems to retain some or all crop residues on the surface. * Reducing evaporation by using tillage systems that retain crop residues on the surface; by reducing tillage frequency and intensity; and by using more intensive cropping systems with less fallow. By retaining rainfall and reducing evaporation, yields of currently grown adapted crops can be increased and it may be possible to grow alternative crops not previously grown in the region. References Allison, F.E. 1973. "Soil Organic Matter and Its Role in Crop Production." In Developments in Soil Science. Amsterdam: Elsevier. Cornish, P.S., and J.E. Pratley. 1991. Tillage Practices in Sustainable Farming Systems. In Victor Squires and Philip Tow, eds., Dryland Farming: A Systems Approach. South Melbourne, Australia: Sidney University Press, in association with Oxford University Press. Greb, B.W., D.E. Smika, and J.R. Welsh. 1979. "Technology and Wheat Yields in the Central Great Plains: Experiment Station Advances." Journal of Soil and Water Conservation 34:264-268. Han Siming, Yang Chungfeng, Shi Juntong, and Pang Huancheng. 1988. 'Stubble Mulching in Dryland on the Loess Plateau in China." Agriculture Research Arid Areas 3:1-12. Johnson, W.C., and R.G. Davis. 1972. "Stubble-Mulch Farming of Winter Wheat; A History of 28 Years' Experience at USDA Southwestern Great Plains Research Center, Bushland, Texas." U.S. Department of Agriculture. Agricultural Research Service Conservation Research Report 16. Washington, D.C. Li, Shengxiu, and Xiao Ling. 1992. "Distribution and Management of Drylands in the People's Republic of China." Advances in Soil Science 18:148-302. 80 Li, Yushan. 1982. "The Relationship Between Soil Water Dynamics and Wheat Production." Shaanxi Agricultural Science 2:5-8. Musick, J.T., O.R. Jones, B.A. Stewart, and D.A. Dusek. 1993. "Relationships of Evapotranspiration, Grain Yield, Water-Use Efficiency and Stored Water at Planting for Irrigated and Dryland Winter Wheat Production." Agronomy Journal. Musick, J.T., and K.B. Porter. 1990. "Wheat." In B.A. Stewart and D.R. Nielsen, eds., Irrigation of Agricultural Crops. Agronomy 30. Madison, Wisconsin: American Society of Agronomy, Crop Science Society of American, and Soil Science Society of America. Shan Lun, Liu Zhongmin, Deng Xiping, and Xin Yequan. 1992. "Field Water Balance Under the Different Crop Rotation Patterns in the Loess Plateau, China." In Conservation Tillage Practices for Grain Farning in Semi-Arid Regions. Proceedings International Symposium, July 7-9, 1992. Shortandy, Kazakhstan, CIS. Shikula, N., A. Gnatenko, and L. Petrenko. 1992. "Soil Protective Technologies of Winter Wheat Cultivation in Arid Regions of Ukraine." In Conservation 7illage Practices for Grain Farming in Semi-Arid Regions. Proceedings International Symposium, July 7-9, 1992. Shortandy, Kazakhstan, CIS. Smika, D.E. 1976. "Seed Zone Soil Water Conditions with Mechanical Tillage in the Semiarid Central Great Plains." In Proceedings of Seventh International Soil Tillage Research Organization Conference, Uppsala, Sweden. Uppsala: College of Agriculture in Sweden. Souleimenov, M.K. 1992. "Development of Soil Conservation Farming Practices for Steppe Areas of Northern Kazakhstan." In Conservation Tillage Practices for Grain Farming in Semi-Arid Regions. Proceedings International Symposium, July 7-9, 1992. Shortandy, Kazakhstan, CIS. Stewart, B.A., and J.L. Steiner. 1990. "Water-Use Efficiency." In R.P. Singh, J.F. Parr, and B.A. Stewart, eds., Dryland Agriculture: Strategies for Sustainability. Advances in Soil Science vol. 13. New York: Springer-Verlag. UNESCO (United Nations Educational, Scientific and Cultural Organization). 1977. World Map of Desertification. United Nations Conference on Desertification. A/Conf. 74/2. Rome: FAO. Unger, P.W. 1978. "Straw-Mulch Rate Effect on Soil Water Storage and Sorghum Yield." Soil Science Society of America Journal 42:486-491. Soil Fertility Management fbr Intensive Agriculture in the Humid Tropics Christian Pieri Introduction Soil fertility management is or may become a major cozisideration as it is expected that 75 percent of future increase in production (Crosson and Anderson 1992) will have to come from increased productivity of the 730 million hectares of these soils already cultivated. It is, of course, recognized that soil productivity is affected in many ways by the widely varying socioeconomic circumstances under which farming is practiced in the tropics. Rates of population growth and the associated pressure on the laild, land titles, and difficult access to inputs may, in many cases, mean that sound practices of soil fertility management are simply not economical, at least given the resources to which the concerned farmers have access. This paper focuses on technical and related matters concerning the contribution of soil fertility per se and its management in intensive agricultural production in the humid tropics. The main aim of soil fertility management for intensive agriculture in the humid tropics from a technological perspective is the maintenance of the newly upgraded soil qualities to raise the inherent low soil productivity and to prevent more severe degradation symptoms such as erosion. Conservation tillage might well offer the best opportunity to reach that aim. In the first section, through selected examples, the nature of the soil fertility problem in the humid tropics is identified, together with the major issues to be addressed in managing the soils under intensive, permanent methods of rainfed agriculture. The second section analyzes the potential of some conservation tillage practices in soil fertility management and the maintenance of a high level of crop productivity. The case of Brazilian cerrados will be used as an example to analyze the constraints and potential of mulch farming in maintaining the soil fertility resource in a sustainable form for long-term productivity. The last section summarizes the significance and application of the scientific findings for the World Bank in terms of actions and research agenda. The Problem of Soil Fertility Management in the Humid Tropics The humid tropics refer to areas where the average annual temperature is higher than 18°C and the length of the growing season exceeds 270 days in a year (Buringh and Dudal 1987) or according to Greenland and Lal (1977) where the rainfall exceeds evaporation for at least 7½h months in the year. Oxisols and ultisols comprise two-thirds of the humid tropical land mass (Sanchez 1989). The primary constraints to plant production on these soils are low nutrient reserves and aluminum toxicity (table 1). The following section discusses some of the challenges of managing soil fertility. The concept of soil fertility is discussed as well as building the fertility of acid soils and maintaining soil fertility in the humid tropics. 82 Table 1. Extent of Major Soil Constraints in the Humid Tropics Soil constraint 10f hectares Percent of humid tropics Low nutrient reserves 980 66 Al toxicity 850 57 High P fixation 565 38 Acid, but not Al toxic 270 18 High erodibility 255 17 Poor drainage 195 13 Low CEC 165 11 Source: Sanchez 1989. Soil Fertility Concept In agriculture, soil is best thought of as a renewable resource. As the soil resource is used over time, it may increase or diminish in its inherent productivity. Soil fertility is a concept for the evaluation of the inherent plant production capacity of the soil resource and its evolution under given climatic conditions and current farming practices. The literature in soil science and related disciplines is rich in providing attributes of soil systems that are influential in controlling the soil fertility experienced by plants growing on such soils. In essence, these attributes are: (a) nutrient reserves and availability; (b) physical factors such as structure, porosity, and compaction that control the access of the plant roots to water, air and nutrients; and (c) the systems of physico-chemical regulation (that is acidity-alkalinity, redox potential) and the systems of biological regulation at rhizosphere and soil organic matter level (Swift 1987), which facilitate the synchronization of moisture and nutrients released from the soil to the plant (Ingram and Swift 1989). Soil fertility is not a static concept; it is the dynamic interaction of these attributes, as influenced by climate and cultural practices. Thus, from an initial static concept of soil richness, referring principally to available nutrient reserves in soils, soil fertility is best described now as an evaluation of the status of the "replenishment system" (Tilton and Skinner 1987; Sebillotte 1989), which maintains the soil renewable resource in a sustainable form for long-term productivity. In practice, the problem of soil fertility management for intensive agriculture in the humid tropics should be analyzed from two different and complementary aspects; (a) the initial detection and correction of the inherent soil constraints as listed in table 1, and (b) the maintenance of improved soil qualities for sustainable crop production. Building the Fertility of Acid Soils A vast amount of scientific and practical knowledge has been accumulated for 30 years on identification, distribution, and extent of the major constraints of acid and infertile oxisols and ultisols in the humid tropics. Appropriate technology to overcome soil chemical constraints is available and has been implemented (Sanchez 1976). Practical recommendations for the determination of lime requirement and the control of aluminum toxicity based on soil analysis and crop tolerance (Kamprath 1972) have been defined and 83 adapted to local conditions. Scientists have demonstrated that aluminum activity in the soil solution is the best estimate of potential aluminum toxicity. It can be easily determined by measuring the ratio of exchangeable aluminum over the sum of exchangeable cations (s). One practical consequence is that aluminum toxicity can be decreased not only by the pH effect of liming products, but also by application of organic and inorganic fertilizers increasing the concentration of calcium (Ca), magnesium (Mg), and potassium (K) in soils, resulting in a decrease of AVEl ratio (Mengel and Kirby 1978). Intensive work also has been carried out on phosphate availability and phosphorus (P) requirement in soils, particularly for oxisols and ultisols. These soils are rich in sesquioxides of iron or aluminum and have a high capacity for phosphate fixation, so that the availability to plants of phosphorus (native or applied) is usually very low (Fox and others 1974; Sanchez 1976). Thus, the P requirements have been assessed consequently in reference to P content in soil solution as influenced by crop, soil, and fertilizer management (Fox 1988). In addition, the potential use of different forms of P fertilizers has been studied, including the assessment of the agronomic value of unacidulated and partially acidulated indigenous rocks phosphate (Hammond, Chien, and Mokwunye 1986). Similarly, specific recommendations are available for the correction of other common nutrient deficiencies in cultivated oxisols, such as potassium (Cooke 1985), sulphur (Blair 1979), and zinc (Giordano and Morvedt 1972), the three most common other than phosphate. From the perspective taken in this paper, I conclude that for a long time the attention of scientists and agronomists has been focused on the building of soil fertility and cannot be considered now as a major technical issue, even if local adaptation of the recommendations is sometimes required. The main problem for most farmers in the humid tropics, who want to improve the low inherent productivity of their fields, is clearly limited access to the required inputs attributable to unfavorable economic conditions and lack of infrastructure for efficient procurement (FAO 1987; Schultz and Parish 1989). Strengthening the fertilizer sector in tropical countries, and particularly in Africa (Yates and Kiss 1993), is certainly an answer to this problem. It is beyond the scope of this short paper to address this issue, which has received full attention in many Asian countries with significant success (Soedjias 1991). Maintaining Soil Fertility in the Humid Tropics Once restored or rebuilt, favorable soil qualities should be maintained for sustainable intensive plant production. Three major causes influence the evolution of soil properties in humid tropical conditions. Water Balance The water balance is positive during most of the growing season, resulting in periodic excess of water in the soil surface, inducing runoff and drainage. The impact on the nutrient losses have been measured at field level under different cropping systems and reported by several authors (Sanchez 1976; Roose 1977). The loss of soluble nutrients by leaching has a significant impact on the nutrient balance of amended and fertilized fields, as illustrated by an example from Madagascar in table 2. Losses of nutrients by runoff and sheet erosion assessed during a fifteen-year period in Madagascar resulted mainly from the nearly 1 percent annual loss of soil organic matter from the soil surface (412 kilogramslhectare/year during 1958-83). Such soil organic matter loss may exceed 5 percent on recently cleared forested areas (Sanchez 1976). 84 Table 2. Nutrients Balance in Heavily Fertilized Maize Fields in Ampangabe (Madagascar) for a Three-Year Period (1975-78) N K20 CaO Fertilizer application + 480.0 + 180.0 + 1,400.0 Crop residues rest. + 149.5 + 109.1 + 56.9 Crop uptake - 251.7 - 142.0 - 59.2 Leaching - 196.0 - 40.4 - 81.6 Balance (3 years) + 182.0 + 107.0 + 1,316.0 Source: Arrivets 1978 (modified by Pieri 1983). In oxisols and ultisols with limited nutrient reserves, the resulting negative nutrient balance is quickly reflected by soil chemical characteristics and by soil pH. Szott, Palm, and Sanchez (1991) observed that 40 percent of the available reserve of calcium and magnesium in the top layer (0-15 centimeters) was lost within 12 months under alley cropping systems introduced in the Yurimaguas region (Peru). They measured similar changes in the total nutrient stocks (living biomass, litter, and soil) under different fallow management systems, and then noted that the "lost calcium and magnesium may have accumulated below the 45 cm sample depth and may eventually be recycled by deep-rooted species.. .however, this would require a relatively long period time." Finally, the authors concluded that "such apparent losses seriously call into question the long term sustainability" of alley cropping and managed fallow systems, unless periodic applications of lime fertilizers are made as required by annual crops. Thus, the maintenance of soil pH and soil nutrients reserve in the humid tropics has to be implemented through the periodic application of amendments, which is based not only on the calculation of nutrient uptake by crops but also on the assessment of mineral losses by leaching, principally for nitrates, calcium, and magnesium. Such an assessment requires more than the limited data now available, such as the estimates made recently by FAO on the nutrient balance in Sub- Saharan African countries. More basic data are required to assess, particularly in the humid tropics, the contribution of nutrients lost by leaching and erosion to the change of the nutrient balance, as measured by the model that has been set up (Stoorvogel and Smaling 1990). Soil Organic Matter Oxidation The increase of soil organic matter oxidation is well documented and annual decomposition rates in the humid tropics range from less than 1 percent to more than 10 percent according to some soil and crop management systems (Sanchez 1976). Very high "decomposition rates' (12.8 percent reported in Zaire by Sanchez) result from the combination of loss by erosion, decrease in annual addition of fresh organic matter, and oxidation of the soil organic matter (2 to 4 percent on average). In contrast with traditional shifting cultivation with long periods of fallow (Greenland 1970), settled agriculture increases the rate of soil organic matter loss, not only on the surface but also in the deeper layer of the soil profile. Annual cropping compared to perennial cropping systems, increases the rate of organic matter decomposition as illustrated by figure 1. This phenomenon is quite general: most of the soil organic matter is lost at the inception of cultivation cycles. This pattern is well illustrated by recent data from 85 the Yurimaguas experiments (figure 2). Moreover these data prove that soil fauna management, as experienced by earthworm inoculation, does not fundamentally change this pattern (Lavelle, Gilot, and Fragoso 1992). In conclusion, if it is generally admitted that the maintenance of soil organic matter is of fundamental importance for no-fertilizer agriculture, it is also true that organic matter is a key soil attribute for the maintenance of soil fertility in low-cation exchange capacity (CEC) soils, such as oxisols and ultisols. Most of the negative charges (holding cations) are in the organic radicals. The soil structure and porosity of these soils are also largely dependent on the presence of organic binders, which bridge the aggregates and particles in soils dominated by low activity clays (Greenland and Lal 1977; Sanchez 1976). Thus, for intensive agriculture under rainfed conditions and annual cropping, soil fertility management should be first targeted toward the maintenance of soil organic matter in such a way that the "replenishment system" of the soil resource is maintained for the long term. Soil Tillage Soil tillage is a major cause of changes in soil characteristic, because it influences directly the physical properties (structure, porosity, mechanical resistance) of the soil surface, and indirectly the whole hydric and temperature soil regimes controlling soil organic matter balance, root development, and nutrient availability. A typical example is given in figure 3, which illustrates the impact of soil preparation methods on soil mechanical resistance to penetration in the context of the mechanized agriculture in the Brazilian cerrados region. The soil compaction induced in this case by continuous disking has favored water runoff and erosion, which resulted in the loss of most of the initial costly corrective application of lime and phosphate, causing finally poor root development (table 3), and poor yields (Seguy, Bouzinac, and Matsubara 1992). Table 3. Changes in Soil Bulk Density and Soybean Root Density Under Different Tillage Methods in the Brazilian Cerrados Alvorado, Goias, Brazil 1987 Depth Soil bulk density Soybean root density3 (cm) Disking Plowing Disking Plowing (g/cc) (g/dm3) 0-15 1.19 1.01 0.466 1.168 15-30 1.30 1.23 0.031 0.266 30-40 1.42 1.22 0.010 0.153 40-50 1.20 1.20 0.010 0.155 a- Variety DOKO, 50 days after planting. Source: Seguy, Bouzinac, and Matsubara 1992. 86 Figure 1. Change in Soil Carbon Content After a Fifteen-Year Period, Under Different Farming Systems, Akpanadoure, Benin C %O 0 5 10 15 20 0 20 - ~S -0 -. |' I ----- Native Forest -60 Figure _. Efe aI _ _ Forest Plantation -80 . *........ Oil Palm Tree Plantation with Mulch 100cI n - -i- Oil Palm Tree Plantation without Mulch l I ~~~~~~~~~~~Annual Food Cropping System 120 (Source = IRiHO, 1992) Figure 2. Effects of Earthworm Inoculation on Grain Production in a Continuous Maize Crop at Yurimaguas 502 0 Worm - 40 .U om B ' ~~~~~~~~~~~~Application of fertilizers _ 10 10 12 3 4 5 Crop Number 87 Figure 3. Effects of Soil Preparation Methods on Mechanized Resistance Penetration, Goiania (Oxisol, Latosol, "Vermelho Oscuro"), Brazil, 1989 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 W kgm cm1 I I I11 10 - 20- 30- Q o-o Continuous disking .-.---. Continuous ploughing 40 . (rice 86/legumes 85) 40- \sn r/ ° ° Continuous ploughing Chisel (rice 86/maize 85) Chisel ploughing 50- (rice 86/maize 85) In summary soil fertility management cannot be thought of only in terms of nutrient application in inorganic or organic forms to maintain the mineral and organic balance of cultivated soils. Tillage methods are of fundamental importance for the maintenance of good physical qualities of cultivated soils and the subsequent efficiency of the nutrients applied for improved plants. Tillage methods must be considered as indispensable to any soil fertility management, particularly in soils prone to compaction and erosion such as oxisols and ultisols. What Is the Main Problem? The major chemical soil constraints limiting plant growth are well known and the basic scientific and technical knowledge is available to overcome the soil acidity problem and the nutrient deficiencies, principally phosphorus. The implementation of this investment in soil fertility building is seriously limited by a broad web of socioeconomic factors, which are beyond the scope of this paper. The main problem of soil fertility management for intensive agriculture in the humid tropics, from a technological point of view, is principally the maintenance of the newly induced favorable soil qualities that have to be created in oxisols and ultisols to raise their inherent low productivity. This improved fertility must be maintained in spite of the difficulties of leaching induced by surplus rainfall, of the rapid turnover of soil organic matter, and of the poor buffering capacity of these soils -- significant decreases in pH and soil nutrient reserves are observed within months following lime and fertilizer applications. 88 From this technical perspective, it can then be noted that there is not a clear boundary between soil fertility decline and severe soil degradation symptoms such as erosion, which are both influenced by the same climatic and edaphologic conditions. The possible linkage between soil fertility decline and further soil erosion is supported by scientific analysis (Hudson and Jackson 1959; Lal 1990; Pieri 1992) in different agroecological circumstances. It may also be consistent with the Bank's past experience. Among the 180 agriculture and forest projects analyzed by the Operations Evaluation Department (OED), entitled Renewable Resource Management in Agriculture (1989), it is very obvious that there are virtually no references to soil fertility. Dramatic symptoms of land degradation, such as erosion, deforestation, and weed infestation have been recognized as a major issue for only 8 percent of the projects. However, if soil fertility is never cited as an issue in this survey related to past Bank-financed projects, there is an implicit acknowledgment that it could be or become a potential problem. Such a concern is now clearly reflected in several of the Bank's more recent publications (Carr 1989; Meyers 1989; Anderson and Thampapillai 1990; World Development Report 1992; Yates and Kiss 1993). Improved Soil Fertility Management and Conservation Tillage In humid tropics, shifting agriculture based upon partial deforestation and complex mixed cropping has provided, and will continue to provide, a sustained soil productivity and a satisfactory maintenance of soil fertility, as long as low population density and available good farm land allows a long fallow period (Greenland 1970; Sanchez 1976). Is there currently any sound alternative for intensive and settled agriculture? In the context of this paper, it has not been possible to address all the potential forms of intervening in soil fertility management in the humid tropics. Among a broad range of techniques generally known as "conservation tillage" (Lal 1989), it has been suggested that direct planting in crop residues and/or permanent mulch cover with minimum or no-tillage and proper crop-rotation have the highest potential for sustainable soil fertility management under intensive agricultural conditions. Attention was first directed to no-tillage in the late 1940s with the appearance of growth regulators, and more intensively in the 1960s with the commercial availability of contact herbicide (Paraquat), with no tillage planters, and fertilizer applicators (Phillips, Thomas, and Blevins 1980; Phillips and Phillips 1984). Several attempts to develop corresponding conservation tillage methods for the tropics have been reported in recent publications (Crovetto 1992), and especially for the humid tropics (Lal 1989; Sheng and Meiman 1991). The case of Brazilian cerrados where such methods have been recently implemented by commercial soybean and grain producers on a significant scale (more than 20,000 hectares) is relevant. The following briefly reviews the major technical findings from this experience to demonstrate the potential of mulch farming as a soil fertility management system eventually suitable for intensive agriculture in the humid tropics. The information presented here has been principally drawn from Seguy, Bouzinac, and Pieri (1991) and Seguy, Bouzinac, and Matsubara (1992). Technical Constraints to Direct Planting Methods on Mulch Cover in the Brazilian Cerrados Three problems have to be solved for implementing no tillage methods in the humid tropics. First, one has to find a suitable mulch cover where warm and humid conditions favor the rate of decomposition of crop residues in the field, resulting in an insufficient protection of the soil top layer. 89 Tlirty days after the onset of rain, the soil coverage measured on farmers' fields (Fazenda Progresso, Mato Grosso State) was no more than, 54, 46, and 16 percent respectively for maize, rice, and soybean residues. Two months later the soil coverage was equal to 30, 38, and 7 percent, respectively. Among a large set of technical experiments at field level, a mixed sowing of a commercial crop (40 to 50 kilograms/hectare of upland rice seeds) and a cover crop (4 to 6 kilograms/hectare of Calopogonium mucunoides) was found to be the best alternative. When planted, Calopogonium (a) has an initial slow growth rate without competitive effect for upland rice; (b) ensures a good soil coverage (more than 80 percent) over time; (c) has a high potential for recycling nutrients and for biological nitrogen fixation (Giller and Wilson 1991); and (d) has a strong allelopathic impact on weeds such as Digitaria horizontalis, Digitaria insularis, Eleusina indica, and Enchinochloa colonum. At harvest time, the calopogonium seeds can be easily separated from rice grain (mechanical and/or manual harvest) producing 500 to 900 kilograms/hectare of seeds. Other mulch covers can be used, such as Macroptylium atropurpureum, Stilozobium aterrinum, and Dolichos lab. A new and promising development is now considered by using some native weeds such as Paspalum notatum for mulch cover (and/or fodder to feed animals in mixed farning). Incidentally it can be noted that "weed management," appears to be a new field of research for scientists who have observed and assessed that total weeding, which is always recommended in modern agriculture, may not be the best alternative for soil fertility maintenance in the humid tropics (Ramakrishnan 1992). Results suggest that weeds below a particular density level (up to 20 percent of the total weed biomass) have positive impact on soil fertility by lowering nutrient leaching and runoff erosion. Herbicide management is another key problem in no tillage systems. In general terms, two successive herbicide applications are required, one for weed cleaning before sowing, and one to control eventual weed development before the full establishment of the planted crops. Herbicide technology is evolving rapidly and new chemical formulas are tested regularly. The last potential problem is the availability of appropriate equipment for sowing and fertilizer application through a thick layer of mulch cover. Such equipment is now available in Brazil, adapted to all conditions. Hand tools, oxen drawn and motorized implements for direct seeding, and fertilizer and lime localization are locally produced (Santa Catarina and Parana States) and are commercially available. Impact on Soil Characteristics For the Brazilian cerrados, data on soil characteristics as influenced by direct planting methods on mulch cover compared to other tillage methods are limited and are still incomplete. Emphasis has been given to some soil physical parameters controlling root development, water movement, and biological activity. The soils analyzed are fertilized adequately and amended, which is usually done by commercial farmers in the cerrados zone. Table 4 shows some results from the Fazenda Progresso (Mato Grosso State) where different tillage methods have been compared since 1985-86, on a significant scale (130 hectares). 90 Table 4. Comparison of Average Infiltration Rate for Different Tillages Practices (Fazenda Progresso, Mato Grosso, Brazil, 1989) Tillage practices Infiltration rate (Double rings infiltrometer) centimeters/hour Continuous disking 28.4 No tillage without permanent mulch cover 29.6 Chiselling 34.0 Plowing (onset of the rainy season) 45.0 No tillage on Calopogonium + crop residues 48.2 Plowing (following harvest) 54.3 Source: S6guy, Bouzinac, and Pieri 1991. Direct planting on permanent plant cover provides satisfactory physical soil characteristics. Soil biological activity also is increased (table 5). The few chemical data available indicate that the pattern of nutrient distribution in the soil profile is modified more by conventional plowing than by direct planting methods or continuous disking (figure 4). Economic and Technical Results The Fazenda Progresso experience indicates that sustainable high yields are achievable as long as continuous disking and monocropping are replaced by deep plowing or no-tillage systems on permanent vegetative cover, such as Calopogonium mucunoides, associated with crop rotations (table 6). Table 5. Measurements of Microbiological Activity for Two Types of Soil Preparation (Fazenda Progresso, Mato Grosso, Brazil, 1989) lillages practices Centimeters Microorganisms per gram of soil Dry season Rainy seasona Plowing (after harvest) 0-10 1.2 x 105 10.7 x 105 10-20 1.2 x 105 5.1 x 10' No tillage on Calopogonium 0-10 4.8 x 105 50.0 X 106 and crop residues 10-20 1.6 x 105 6.4 x 105 a- One month after the beginning of the rainy season (240 millimeters; 12 percent annual rainfall). Source: Seguy, Bouzinac, and Pieri 1991. 91 Figure 4. Content of Soil Exchangeable Calcium and Magnesium for Different Tilage Practices, Compared to Initial Content in Native Cerrado Soil (Oxisol in Fazenda Progresso (Mato Grosso) Brazil, 1988) 1 2 3 4 5 6 7 Ca + Mg I0 20- E.-.--o Uncultivated cerrado soil 40 / -- Continuous disking ! ! / Direct planting (2nd year) I}I i ^ A Ploughing (Source: Seguy et al, 1990) On average, during the period 1986-92, these two systems provided similar yields of soybean (table 7) and maize. In contrast, upland rice is a crop not adapted to direct planting because it is very dependant on soil macroporosity and deep plowing (figure 5). Costs of production have been calculated during the same period. No decisive advantage in favor of one or the other method of tillage has been observed. The average cost of land preparation (including the pre-emergence herbicide application) is approximately the same: US$45 per hectare, representing only 10 percent of the total cost of production. An additional advantage cited by Brazilian farmers using these new techniques is the flexibility provided by direct planting methods. Field trafficability is increased on mulch cover and time of land preparation and seeding is substantially decreased per hectare -- direct planting - 1.4 hour; disking - 3.6 hour; and plowing - 4.3 hour -- as measured in Fazenda Progresso. From this commercial experience in the Brazilian cerrados, it appears that no-tillage and permanent vegetative cover is an attractive alternative to the traditional disking and monocropping systems. The maintenance of favorable physical and biological soil properties is easier to maintain. More work certainly has to be done to monitor the chemical status resulting from direct planting methods under humid tropical conditions. High yields of soybean, upland rice, and maize can be achieved with both deep plowing system, and no-tillage systems. At this stage of experience it should be noted that both systems induce specific risks when the managerial and technical capacity of farmers are not sufficient. There is the risk of erosion induced by an inadequate plowing, the risk of pests (mulch cover is a niche for insects), and of weed infestation linked to improper use of pesticides. Two surveys to assess the acceptability of the different components of conservation tillage initiated in Fazenda Progresso have been carried out in 1990 and in 1991. The first survey was run in the immediate surroundings of the Fazenda (total area surveyed 17,123 hectares; 57 farms); the second survey covered 2.6 million hectares (Mato Grosso State, center and southeast). Remote sensing analysis (Spot image) complemented a ground survey. Among the chief conclusions, it appears that from 1985 on (a) deep plowing progressively replaced disking (42 percent of soybean Table 6. Effect of Conventional and Direct Planting (No Tillage) Practices in Presence of a Cover Crop (Calopogonium mucunoides) on the Yields of Crops in Rotation, Fazenda Progresso, Mato Grosso, 1987-90 1986-87 1987 1987-88 1988-89 1989-90 Crop and soil Crops Yield Crops Yield Crops Yield management (kg/ha-') (kg/ha-') (kg/ha7') Sowing of a mixture of Deep plowing Fertilized with Soybean 1,215 Maize 4,700 Soybean 1,775c rice and C. NPKa at mucunoides. seeding Maize 4,030 Maize 2,678 Rice yield = 3,225 kg + calopogonium ha7' Deep plowing Thermo- Soybean 1,440 Maize 6,500 Soybean 900c phosphate Yoorin Maize 4,226 Maize 3,068 Bz 1500 +calopogonium kg ha' b Soil cover at the end of Direct planting Fertilized with Soybean 2,040 Maize 5,200 Soybean 2,460 the dry season (straws NPK& at of rice + C. seeding Maize 4,360 Maize 5,200 mucunoides) = 12.5 t ha-' There is natural Direct planting Thermo- Soybean 2,486 Maize 6,400 Soybean 2,947 dissemination of C. phosphate mucunoides the Yoorin Maize 4,940 Maize 5,830 following years Bz 1500 kg ha-' b ' NPK fertilizer placed at seeding * soybean: 350 kg ha-' 0-25-25 * maize: 350 kg ha-' 5-30-15- + 100 kg ha-' urea b. Fertilizer thermophosphate * 1,500 kg ha-' of Yoorin Bz applied in 1987 for three years complimented with N and K to achieve same level as C. Plots dominated partly or totally by Calopogonium sp 93 Table 7. Impact of Different 7illage Practices and Crop Rotation on Soybean Production (Fazenda Progresso, Mato Grosso, Brazil 1992) Soybean seeds Kilograms/hectare Crop rotation Disking Plowing No tillage Monocropping 1,675 2,120 1,990 (100) (127) (119) Soybean + rice 2,564 3,092 3,041 (153) (185) (182) Soybean + maize 2,850 3,010 3,057 (170) (180) (183) Source: Seguy, Bouzinac, and Matsubara 1992 seedbeds are plowed); (b) crop rotations replaced monocropping particularly in maize growing farms; and (c) direct planting methods were used only on 4 percent of the total area, but many farmers (65 percent) are trying them. These first surveys suggest that direct planting methods are raising expectations among farmers. More and more scientists (Kang, Wilson, and Lawson 1984; El Swaify, Singh, and Pathak 1987; Lal 1987; Sanchez, Palm, and Smyth 1990; Crovetto 1992) also have come to the conclusion that direct planting methods on mulch cover are cost effective in containing soil degradation, maintaining soil fertility level, and sustaining soil productivity in the humid tropics. However, much of this work has not gone beyond scientific investigation and predevelopment. It must be concluded that more experience is required and should be encouraged to implement and assess such methods in both commercial and, even more so, in present smallholder agriculture. Conclusions What then can we learn from a more focused attention on the question of soil fertility management in World Bank agricultural project work? The socioeconomic circumstances of particular farming systems are so diverse that there can be few general conclusions and what makes most sense in any particular circumstance must be adapted to the local situation (Tourte 1984; Binswanger 1980). It is thus a complex policy issue to look at systems of effective soil fertility intervention to maintain the soil resource in a sustainable form for long-term productivity, particularly, in the context of intensive settled agriculture under rainfed conditions (IBSRAM 1991). However, from the scientific findings and practical experiences presented in this paper, it is necessary to stress three implications for the Bank's actions and research agenda. First, a few principles should guide our thinking about the soil fertility dimension in project work. The documented experience is substantial in indicating that, in many tropical situations, it is most cost- effective to intervene early in terms of maintaining soil "chemical fertility" and related structural 94 elements, such as organic matter content, because the longer such intervention is delayed, the more expensive, and thus sometimes impossible, any such replenishment of the system becomes. Figure 5a. Effects of Tillage Practices on Maize Yields (Average Six Years, 1986-92), Fazenda Progresso, Brazil (Seguy, Bouzinac, and Matsubara 1992) 4500 _ 4037 4062 4000 3874 YIELD 3500 Kg/ha 3000 2500 2000 - _ DISKING PLOUGHING NO TILLAGE Figure 5b. Effects of Tillage Practices on Upland Rice Yields (Average Five Years, in 1986-91), Fazenda Progresso, Brazil (Seguy, Bouzinac, and Matsubara 1992) 3500 3093 3000 YIELD 2500 Kg-/ha 2000 1835 1500 DISKING PLOUGHING NO TILLAGE 95 With this guiding principle, the possibility should be considered during project design that the soil fertility resource is declining early in the project cycle and even before the symptoms of extreme degradation, for example, erosion becomes highly evident, as suggested by the Brazilian cerrados experience. This is presented here as hypothesis and it would be instructive if more Bank projects could be examined carefully to explore its validity, under differing farming conditions of the humid tropical agroecological world to begin with. Second, a related question is, do we know how to measure change in soil fertility? As already noted, the literature of soil science and related disciplines is rich in defining the attributes of soil systems that influence soil fertility as measured by plants growing on such soils on experimental plots or sites. The problem is that there is as yet no successful integration of these diverse measurable aspects that leads to effective low cost implementation over large project areas where soil fertility interventions may well be contemplated. Recently, Australian scientists (Hamblin 1992) have addressed this issue on an agroecological basis "to improve the ability of providing land holders and managers with early warnings of ecological deterioration and productivity loss." For each region, they strive (a) to select only three indicators on which the others depend; (b) to develop the reasons (hierarchy) for the selection; and (c) to identify the process or methodology for obtaining and using the information. Table 8 illustrates the selection made for a region dominated by mixed cropping under rainfed conditions. The first two indicators are related to the productivity of soils as influenced by climatic and edaphic factors. Certainly nutrient balances calculated at regional levels can be a relevant indicator of possible change in soil fertility (Stoorvogel and Smaling 1990; Andre 1990). Experience in Burkina Faso is certainly noncontroversial in this regard (table 9) but just how relevant that set of findings is to other agroecological zones must, for the moment, remain an open question. Table 8. Three Most Important Primary Indicators of Sustainable Agriculture in Rainfed Crop and Animal Production Region Indicator Why Process/Measurement 1. Water use efficiency Easy to relate production to Yield (kilograms/ rainfall; useable for both millimeters/hectares); at crops and pastures scales of farm, shire, and region to detect trends in time and between areas 2. Soil health A direct expression of Soil analysis of pH (trends) sustainability of system nutrient balance; direct measurement of worm counts, surveys of microflora 3. Farm management Understanding needed Farm records and farm skills before change occurs; surveys; units of cash flow, good level required for debt equity, whole-farm financial survival planning Source: Hamblin 1992 96 Thus, there seems to be a major gap in the knowledge of soil fertility management systems pertaining to workable indicators for guiding interventions at the project level. Hence, there is at first sight a case for a research program dedicated to identifying such pragmatic indicators and testing them in project implementation. I suggest that we organize a working group to clearly identify the specific needs related to soil fertility management and soil fertility monitoring in the Bank's current lending program, to summarize the scientific state of knowledge in these matters, and to identify ways of bridging the gap between what is known and what is needed at operation level. Third, in the last World Development Report 1992, it is clearly indicated that preserving soil fertility cannot be envisaged in isolation from the management of other natural resources: water, natural vegetative covers, and fauna, which are jointly exploited on the same surface area. In order to progress beyond the relatively simple step of building up soil fertility, and to foster intensive agriculture in the humid tropics it will be necessary to develop transitional cropping and farming systems for each specific condition. Conservation farming might well offer the best opportunity to reach that aim. The technical dimensions of conservation farming run the full gamut of mechanical innovations and hence, mechanization policy in general, to biological control systems, including a more effective use of herbicides and farmers' access to them, and finally to advances in biotechnology (such as the development of herbicide-resistant crop cultivars). The most promising of these new technologies is direct planting in permanent vegetative cover. This has been substantiated by research and is being adopted by increasing numbers of farmers around the world. More experience has to be gained and analyzed at an operational scale to identify the technical and socioeconomical constraints to the practical implementation of conservation tillage methods. This experience surely will be influential in determining more effective intervention for maintaining soil fertility and a high level of soil productivity in the humid tropics. Table 9. Course of Yields with Fertilizer Treatment and Fertilizer Efficiency at Six Young Farner Centers in Burkina Faso Rotation no. Yield Fertilizer rate Efficiency (Course no.) (kilograms (kilograms/hectares) (yield/N + P205 + K20) /hectares) N P205 K20 1 (1st) 2,060 20 31 30 25.4 2 (4th) 1,546 20 31 30 19.1 3 (7th) 1,437 69 82 66 6.6 4 (10th) 1,514 107 116 96 4.7 5 (13th) 1,445 107 116 96 4.5 Source: According to Hien in Pieri 1992. References Anderson, J. R., and J. Thampapillai. 1990. Soil Conservation in Developing Countries: Project and Policy Intervention, World Bank Policy and Research Services Paper No.8, Washington, D.C. 97 Andre, M. 1990. Approvisionnement, Commercialisation et Demande en Engrais en Republique du Togo, Lome, Togo: International Fertilizer Development Center. Binswanger, H. P. 1980. "Attitudes Toward Risk: Experimental Measurement in Rural India." American Journal of Agricultural Economics 62(1):395-407. Blair, G. J. 1979. Sulphur in the Tropics, Technical Bull, T12, Muscle Shoals, Alabama: International Fertilizer Development Center. Buringh, P., and R. Dudal. 1987. "Agricultural Land Use in Space and Time." In M.G. Wolman, and F.G.A. Fournier, eds. Land Transformation in Agriculture, SCOPE 32. Wiley and Sons. Carr, S. 1989. Technology for Small-Scale Farmers in Sub-Saharan Africa: Experience with Food Crop Production in Five Major Ecological Zones, World Bank Technical Paper No. 109, Washington, D.C. Cooke, G. W. 1985. "Potassium in the Systems of the Humid Tropics." In Proceedings of the 19th Colloquim International Potash Institute. Crosson, P., and J. R. Anderson. 1992. "Resources and Global Food Prospects: Supply and Demand for Cereals to 2030." World Bank Technical Paper No.184, Washington, D.C. Crovetto, C. L. 1992. Rastrojos Sobre el Suelo: una Introduccion a la Zero Labranza. Santiago de Chile: Universidad Autonoma. El-Swaify, S. A., S. Singh, and P. Pathak. 1987. "Physical and Conservation Constraints and Management Components for SAT Alfisols." In Alfisols in the Semi-arid Tropics. Patencheru, India: International Crops Research Institute for the Semi-Arid Tropics. FAO (Food and Agriculture Organization). 1987. Fertilizer Strategies. Land and Water Development Series, No.10. Rome. Fox, R. L., R. K. Hashimoto, J. R. Thompson, and R. S. de la Pena. 1974. "Comparative External Phosphorous Requirements of Plant Growing in Tropical Soils." In Proceedings of the Tenth International Congress of Soil Sciences, Moscow. Fox, R. L. 1988. "Phosphorus: A Basic Nutrient for Soil Improvement." In Proceedings of the International Conference on the Management and Fertilization of Upland Soils in the Tropics and Sub-Tropics. Nanjing, P. R. of China. Giller, K. E., and K. J. Wilson. 1991. Nitrogen Fixation in Tropical Cropping Systems, London: C.A.B. International. Giordano, P. M., and J. J. Morvedt. 1972. "Agronomic Effectiveness of Micronutrients in Macronutrient Fertilizer." In Micronutrients in Agriculture. 98 Greenland, D. J., and R. Lal. 1977. Soil Conservation and Management in the Humid Tropics. Chichester: John Wiley and Sons. Greenland, D. J. 1970. "The Maintenance of Shifting Cultivation Versus the Development of Continuous Management Systems." In Semi-Traditional Systems of African Agriculture. Ford Foundation, IRAT, Ibadan: International Institute of Tropical Agriculture. Hamblin, A. 1992. "Environmental Indicators for Sustainable Agriculture." Report on a national workshop, November 28-29, 1991. Bureau of Rural Resources, Canberra: Land and Water Resource Research and Development Corporation. Hammond, L. L., S. H. Chien, and A. U. Mokwunye. 1986. "Agronomic Value of Unacidulated and Partially Acidulated Phosphate Rock Indigenous to the Tropics." In Advances in Agronomy Vol. 40. London: Academic Press. IBSRAM (International Board for Soil Research and Management). 1991. Evaluation for Sustainable Management in the Developing World. Proceedings No. 12, Vols. I-III. Ingram, J. S. I., and M. J. Swift. 1989. "Tropical Soil Biology and Fertility (TSBF) Programme." Report of the Fourth TSBF Interregionnal Workshop, Special Issue-20, International Union of Biological Sciences (IUBS) News Magazine. IRHO (Institut de Recherche sur les Huiles et les Oleagineux). 1992. Aptitude des sols a la culture du palmier a huile et du cocotier. Oleagineux 47:6. Kamprath, E. J. 1972. "Soil Acidity and Liming." In Soil of the Hwnid Tropics, Washington, D.C.: National Academy of Sciences. Kang, B. T., G. F. Wilson, and T.L. Lawson. 1984. "Alley Cropping a Stable Alternative to Shifting Cultivation." Ibadan: International Institute for Tropical Agriculture. Lal, R. 1987. Tropical Ecology and Physical Edaphology. Chichester: John Wiley and Sons. Lal, R. 1989. "Conservation Tillage for Sustainable Agriculture: Tropics Versus Temperate Environments." In Advances in Agronomy Vol.42.London/New York: Academic Press. Lal, R. 1990. Soil Erosion in the Tropics: Principles and Management. New York: McGraw-Hill. Lavelle, P., C. Gilot, and C. Fragoso. (forthcoming). "Soil Fauna and Sustainable Land Use in the Humid Tropics." Symposium on Soil Resiliency and Sustainable Land Use, Budapest, 28 September-2 October 1992, London: CAB International. Mengel, K., and E. A. Kirkby. 1978. Principles of Plant Nutrition. Bern: International Potash Institute. Meyers, L. R., ed. 1989. "Innovation in Resource Management: Proceedings of the Ninth Agricultural Sector Symposium." World Bank Policy Paper, Washington, D.C. 99 Phillips, R. E., G. W. Thomas, and R. L. Blevins, eds. 1980. "No-Tillage Research: Research Reports and Review." Lexington: University of Kentucky, College of Agriculture and Agricultural Experiment Station. Phillips, R. E., and S. H. Phillips. 1984. No-Tillage Agriculture: Principles and Practices. New York: Van Nostrand Reinhold Company. Pieri, C. 1992. Fertility of Soils: A Future for Farming in the West African Savannah. Berlin: Springer Verlag. Ramakrishnan, P. S. 1992. Shifting Agriculture and Sustainable Development, UNESCO. Paris: The Parthenon Publishing Group. Roose, E., 1977. Dynamique actuelle des sols ferrallitiques et ferrugineux tropicaux d'Afrique occidentale. Etude experimentale des transferts hydrologiques et biologiques de matiere sous veg6tations naturelles ou cultivees. Travel Document No.78. Paris: ORSTROM. Sanchez, P. A. 1976. Properties and Management of Soils in the Tropics, New York: John Wiley and Sons Publishers. Sanchez, P. A. 1989. "Biologeographical and Ecological studies." In H. Leith, and M. J. A. Werger, eds. Tropical Rainforest Ecosystems, Amsterdam: Elsevier. Sanchez, P. A., C. A. Palm, and T. Smyth. 1990. "Approaches to Mitigate Tropical Deforestation by Sustainable Soil Management Practices." In H. W. Scharpenseel, M. Schomarkar, and A. Ayoub, eds., Soils on a Warmer Earth: Developments in Soil Science. Schultz, J. J., and D. H. Parish. 1989. Fertilizer Production and Supply Constraints and Options in Sub-Saharan Africa. Muscle Shoals: International Fertilizer Development Center. Sebillotte, M. 1989. Fertilite et Systemes de Production. Ecologie et Amenagement Rural. Paris: INRA. Seguy, L., S. Bouzinac, and C. Pieri. 1991. "An Approach to the Development of Sustainable Farming Systems." In Evaluation for Sustainable Land Management in the Developing World, 15-21 September 1991, Chiang Rai, Thailand. Proceedings No.12. Bangkok: IBSRAM. Seguy, L., S. Bouzinac, and M. Matsubara. 1992. "Gestion des sols et des cultures dans les zones de frontieres agricoles des cerrados humides du Centre-Ouest bresilien: Highlights 1992." Montpellier: CIRAD-Ca. Sheng, T. C., and R. Meiman. 1991. "The Taiwan Experience: A Workshop Summary." In Development of Conservation Farming on Hillslopes. Ankeny, Iowa: Soil and Water Conservation Society. 100 Soedjias, Z. 1991. "Fertilizer Sector Development in the Asian Region." In Proceedings IFA- FADINAP Regional Fertilizer Conference for Asia and the Pacific, New Delhi, 20-22 Nov. 1991. Spain, J. M. 1971. "El manejo de Oxisoles en el Oriente de Colombia." In IV Congreso Latinoamericano de la Ciencia del Suelo, Maracay. Stoorvogel, J. J., and E. M. A. Smaling. 1990. Assessment of Soil Nutrient Depletion in Sub-Saharan Africa: 1983-2000, Report 28, Vol. I, Wageningen: The Winand Staring Center. Swift, M. J. 1987. Tropical Soil Biology and Fertility (ISBF). International Regional Planning Workshop. Biology International. Special Issue No.13. Paris: IUBS. Szott, L. T., C. A. Palm, and P. A. Sanchez. 1991. "Agroforestry in Acid Soils of the Humid Tropics." In Advances in Agronomy, Vol.45. San Diego: Academic Press California. Tourte, R. 1984. "Introduction." In P.J. Matlon, ed. Coming Full Circle: Farmers' Participation in the Development of Technology. Publication No.189. Ottawa: International Development Research Centre. Tilton, J. E., and B. J. Skinner. 1987. "The Meaning of Resources." In Resources and World Development. Dalheni Workshop. Berlin: John Wiley and Sons. World Bank. 1989. Renewable Resources Management in Agriculture. Operation Evaluation Study. Washington, D.C. World Bank. 1992. World Development Report 1992. New York: Oxford University Press. WRI (World Resources Institute). 1992-93. A Guide to the Global Environment, New York: Oxford University Press. Yates, A, and A. Kiss. 1993. Using and Sustaining Africa's Soils. Summary of the proceeding of a seminar held in Washington, D.C. in January 1992. Washington, D.C.: World Bank. Preserving the Options International Research for Sustainable Agriculture Hubert G. Zandstra* Somehow the global food production system must keep pace with the demand that 90 million new mouths place on it every year. In addition the natural resource base -- land, water, and plant and animal species -- must be preserved, and eventually upgraded to satisfy food needs 100 years from now. Increased productivity in agriculture is central to reducing damage to the resource base. The technologies required to raise productivity while conserving natural resources will come from agricultural research, the domain of national research systems and the CGIAR centers. The CGIAR centers have formulated their response to Agenda 21 of the United Nations Conference on Environment and Development (UNCED). They will increase emphasis on ecologies that are particularly threatened. Improved germplasm and management techniques will, however, remain the source for technologies to improve land-use systems. The author expects that commodity, factor, land use, and policy research for increased productivity will become more participatory and will include measures for conservation and rehabilitation of biotic, land, and water resources for specific watershed or farm conditions. The CGIAR is developing an ecoregional approach for this purpose. It will facilitate the combination of capabilities of several CGIAR centers to address problems in threatened ecologies. It will also evolve the CGIAR's modus operandi toward one of increased participation of national research and development organizations with a wide range of capabilities including ecology, farmer organizations, natural resource policy, and nature conservation. The FAO estimates that the land's carrying capacity is severely challenged in many of the more marginal ecologies of the world, such as parts of the humid tropics and the tropical mountain regions of the Andes, the Himalayas, and the Eastern African highlands. In much of the industrialized world, artificially high prices have led to excessive use of agricultural chemicals causing environmental pollution. In developing countries the lack of fertilizer of any kind is reducing crop yields, ground cover, and the rate of regrowth of native vegetation during fallow periods. This increases soil loss and people migrate to new, marginal, highly erodible land in much of the developing countries. And as the natural resource base deteriorates, it becomes ever more difficult to provide for an increased population. The impact of greater demand for food and fuel for the rapidly growing number of poor is not confined to land and water resources. It also threatens the diversity of plant and animal life. Genetic diversity of crops has been the main source for progress in agriculture over the last seventy- five years, and holds the promise for reducing future dependency of agriculture on toxic chemicals for pest control. * Hubert Zandstra is Director General of the International Potato Center in Lima, Peru and presently chairman of the Consultative Group on International Agricultural Research's (CGIAR) committee on sustainability and environment. 102 Interconnected Challenges Three of the major problems confronting humankind -- poverty, the environment, and population growth -- are closely interrelated (Murqueito 1992). Poverty is demonstrably toxic to the environment, for example, as the poor press on forest margins and fragile lands in pursuit of food. And while there are programs focused on health and education for women, which have reduced population growth rates, clearly population growth is slowed most effectively by higher incomes. Also higher incomes provide the resources for health and education programs. Finally, in the longer run, environmental degradation will limit income growth through its impact on the resource base. The three problems, then, form a complex nexus with poverty as the pivotal dimension. Agricultural research can play a central role in resolving these problems through its impact on agricultural productivity and on the environment. Solutions to the poverty--environment--population nexus of problems in developing countries require economic growth. Growth alone may not be sufficient but it will be necessary. Growth, the direct answer to the poverty problem, is favored tremendously by declining real prices for foodstuffs as these lubricate the complex interactions that lead to economic growth. Indeed there are very few examples where growth has occurred without such price decline. Increased productivity in agriculture is central to lowering food prices. As well, increased productivity brings higher incomes to the sector, propelling ever widening demands and income streams. The resulting higher incomes will both slow degradation of the environment and population growth. Moreover, reduced population growth will itself favor the environment. Finally, improved management of agricultural activities, through the use of resource-conserving technologies, will directly reduce the impact of agriculture on the local, nearby, and global environments. Sustainable Agriculture and Productivity Of the many definitions of sustainable agriculture I would like to share the one the CGIAR has defined in 1987 as: "The successful management of resources for agriculture to satisfy changing human needs while maintaining or enhancing the quality of the environment and conserving natural resources." More recently the CGIAR committee on Sustainability and Environrment has adopted the following definition of sustainable agriculture: "A sustainable agriculture is one that over the long term enhances environmental quality and the resource base on which agriculture depends, provides for basic human food and fibre needs, is economically viable and enhances the quality of life for farmers and society as a whole."' This definition stresses the changing human needs. It implies that the choices we make today about using resources will not be permanently maintained. Sustainable agriculture, however, will allow us to meet our present needs while maintaining or increasing our future options. Between 1950 and the end of the century the average area of cropland per person is expected to drop from 0.24 to 0.13 hectares. Maintaining per capita food production at the present level will therefore require increases in yields. In addition the productivity of arable land is deteriorating through desertification, waterlogging and salinity, loss of topsoil, compaction, and accumulation of toxins. Scientists also are predicting significant climate changes. It is well established that 25 percent of the world's population in the industrialized world generates nearly 75 percent of total carbon dioxide emissions thus contributing to eventual global warming. Rising temperatures due to atmospheric buildup of carbon dioxide and other greenhouse gasses have serious implications for 103 agriculture. Furthermore, a thinning of the earth's protective ozone layer is allowing harmful ultraviolet radiation to strike the earth's surface. The result is that farmers must do more with less, not simply increase food production but also assure that the resource base is upgraded. Somehow the global food production system must keep pace with the demand that 90 million new mouths place on it every year. Second, the natural resource base -- land, water, and plant species -- must be preserved, and eventually upgraded, to fuel the agricultural giant that will have to satisfy the food demands of the future. Agriculture therefore faces a double challenge -- not only to increase productivity to meet food needs, but to assure that the resource base is preserved for the future. World population has more than doubled since 1950 to reach about 5.4 billion. Growth is expected to continue until the end of the next century, when world population could reach from 10.2 to 14 billion. Nearly all this growth will take place in the developing countries. The technologies required to raise productivity while conserving natural resources will come from agricultural research. It follows then that those who would reduce poverty, conserve natural resources, and slow population growth should see a well-functioning global agricultural research system as an effective vehicle for achieving their aims. But increases in food production cannot be achieved by research alone. Economic and social conditions, and especially government policies, are limiting factors. Many governments now recognize that the present cost of protecting resources to conserve soil, water, and forests is lower than the cost of trying to restore them later after much environmental damage is done. The CGIAR centers have put a heavy emphasis on research designed to help poor farmers, despite the evidence that returns in production are greater from research done on prime farmland. 'Tis is because the introduction of agriculture in less favorable areas has frequently resulted in environmental damage. Subsistence farmers cannot afford conservation programs. For them the greatest concern is survival. Yet increased demand for food will have to be met where the demand exists: not by developing the most favorable environment at the expense of the less favorable, but by helping the poorest become more productive. Methodological Consequences of Natural Resource Management Research for Sustainable Increases in Agricultural Production The increased attention for sustainability issues largely stems from the need to avoid degradation of the resource base on which agriculture depends and the deterioration of the quality of the environment, while assuring a continued increase in agricultural production to meet the demands of the growing world population on. From its definition follows that sustainability is a comprehensive quality of a complex system. It, therefore, is not a simple parameter that can be directly measured.2 However, one could quantify the physical and biological processes occurring in a particular agroecosystem and quantify interactions between processes and components of the system. Researchers could then relate these processes and interactions to environmental and social conditions, and try to predict the behavior of such a system over time. Such approaches will require the use of comprehensive simulation models, at specified levels of input and based on certain assumptions with respect to social and environmental conditions, crop varieties, the incidence of pests and diseases, and so forth.3 All physical and biological processes occurring in agroecosystems are functions of time and space. Therefore, time and space scales have to be defined before systems can be described and the sustainability of such systems can be assessed. It is important to note that the mathematical equations 104 or analytical models and measurement approaches used to describe physical and biological processes in agroecosystems may be fundamentally different between different time or space scales. This implies that information obtained from a specific set of time and space scales (for example, minutes and millimeters, or days and meters) is not necessarily relevant for another set of scales (for example, years and kilometers). For this reason, the assessment of sustainability for different scales would have to be distinguished. For example, if nutrients are transferred from surrounding grazing lands to agricultural lands in the form of animal manures, this may not be a sustainable system, as far as the grazing lands are concerned. However, this imbalance would not be noticed if one would only consider the farmer's field as the relevant unit. Similarly, on the scale of the region or country one could easily miss this nonsustainable practice. In addition to the time and space scales, one has to define the agricultural systems in terms of inputs, outputs, soil, crop, environmental, and social conditions. If the initial and boundary conditions of a system are not specified, the behavior of that system over time cannot be described. For example, one has to indicate whether carbon dioxide contents, mean atmospheric temperatures, and crop genetic potentials can be taken as constant, and so on. Of course predictions can be made for different scenarios, but these examples illustrate that long-term predictions on the behavior of agricultural systems may be of limited value. The long-term impact on the natural resources of a region from changes in agricultural practices or land-use systems is therefore difficult to ascertain. Nonetheless, there are obvious improvements that can be made to avoid deterioration of land qualities and biological diversity, without major challenges of measurement. These include among others, reduction of soil loss, prevention of toxins in the environment, improved nutrient use and cycling, and the use of multiple genetic backgrounds in cultivars. For many of these practical interventions directed at the correction of obvious resource use problems, measurement can be relatively simple in terms of specific resource productivity such as yield or net returns per unit of soil phosphorus or millimeters of water (Zandstra and others 1986). These concepts have been extended to those of total factor productivity (TFP) which when appropriate care is taken in capturing and costing all resources should reflect the comparative sustainability of land use alternatives. Pricing of resources and products would have to compensate for distortions such as subsidies, and demand induced changes in product prices. In most situations the best objective measure is therefore not a monetized value of production, but rather a physical one, such as yield, calories, protein, or grain. The new approach to resource management research is based on four major considerations: (a) the need to integrate resource management research with research on crop and tree improvement and livestock husbandry, (b) the need to address human and technical dimensions in an integrated way, (c) the need to adopt a systems level approach and to plan and evaluate the component research from this viewpoint, and (d) the strategic need to link policy formulation to technology development and diffusion. This systems approach requires multidisciplinary teams of scientists in several disciplines of social, biological, physical, and/or engineering sciences, as well as capabilities: (a) soil sciences, including soil physics, chemistry and biology, and soil conservation and land management, (b) economics and social sciences, including human nutrition, (c) crop sciences, including agronomy, physiology, and plant nutrition, (d) plant pathology and entomology, (e) hydrology and water management, (f) forestry, including agroforestry and social forestry, (g) aquatic biology, (h) crop, tree, and livestock improvement, and in addition, (i) in mathematics, statistics and biometrics, simulation modeling, meteorology, and climatology. The research methodology can be described under three headings: (a) diagnostic research, (b) building inventories of component technologies, and (c) policy management for technology diffusion. 105 The starting point for resource management research is the diagnosis of the position of existing farming systems on the transition path from extensive to intensive agriculture. This diagnostic process includes the following steps: * The characterization of the ecoregion by both physical and socioeconomic parameters to create a sampling framework to know where and to whom results are relevant and to where they can be extrapolated. a The utilization of the diagnostic research in priority setting, including the selection of research thrusts based on urgency and generality of problems. * The design of strategic experimental and laboratory research. Building an inventory of resource management components requires a farming systems perspective in both formal agricultural experimentation and for the evaluation of indigenous technologies identified by farmers themselves. Formal experimental work as well as sources of indigenous technical knowledge will feed into inventories of resource management components, each with descriptors on the circumstances of farming systems conducive to its adoption and use by farmers in their systems. While some of the component technology research can be conducted on station, much of it will require community level participation in a well-defined location. Such "Heritage Research Sites" for natural resources management research have been prepared by CIP. For the policy management for technology diffusion it is important to spell out the factors conditioning the behavior of small, resource poor farmers. All farmers, including resource poor ones, operate in a production environment in which policy is a key component in shaping their decisions. When policy changes create appropriate conditions to intensify, farmers shift their production strategy. Policy instruments of all types, but particularly market prices, input prices and subsidies, and credit access and subsidy, are useful to draw new technology into local farming systems to preempt falls in labor productivity. In most cases advances will be made in a stepwise fashion by the introduction of a sequence of new components that accumulate into sustainable and productive resource management systems. Institutional Consequences of Natural Resources Management Research for Sustainable Agriculture The research will have to be highly multidisciplinary and should therefore combine a range of institutions at the national, regional, and international level. Spatial Coverage The activity is generally focused on a major ecological condition, but it may cover ecological systems (recurrent combinations of ecologies that are strongly interdependent in their use of, and impact on natural resource systems). It may also focus on a predominant land-use system (for example, livestock production in the savannahs on acid weathered soils in South America). The activity is developed within a region, or a contiguous area covering parts of more than one region. While this is so, consideration should be given to spill over effects to similar ecologies in different regions (for example, from the Andes to other areas of the cool mountainous tropics). 106 Commodity, Enterprise, and Factor coverage The approach is designed to address major constraints to increasing sustainable production. It includes work on priority commodities, livestock, agroforestry, nutrient and soil management, and water management as part of research on improving land use systems, actual or potential. The work also includes a strong emphasis on policy and institutional aspects that affect the sustainability of land use. Institutional Construct The activity will be as structured as an interinstitutional initiative4 involving several international centers and a wide range of national research organizations selected for their capacity, commodity coverage, and location. The activity can be hosted by any international center that has advantage because of location, commodity, or other expertise. The consortium will seek substantial involvement of national systems, their government, nongovernment, academic, and, where appropriate, commercial institutions. This involvement will result in a considerable devolution of research activities to institutions with advantage over others in terms of capability, location, or access. Participation will be open and based on merit. The coverage, priorities, outputs, and institutional participation will be defined in a research planning process that involves all participating IARCs and NARS. The funding of agreed upon activities will predominantly be from existing sources, but will be supplemented from international and national sources as contract research or grants-in-aid to participants for collaborative (shared) work. Research will be planned in a way so that to the extent possible existing physical and human capabilities will be used. The activity will have a governance that complements that of the participating institutions. It will normally have a steering committee, which is small, but represents members of all important participating groups. The steering committee assists the host institution in decisions involving the allocation of research tasks and funds. It approves a program of work and budget on a yearly basis and helps develop procedures for monitoring, internal and external reviews, and reporting. Ecological Suicide Traditional farming systems designed to feed small populations are failing to meet the needs of increasing populations. Unable to increase food output from limited resources, the poor are driven to farming practices that amount to ecological suicide: shortening bush fallows, extending cultivation to forest areas, and grazing more animals than the land can support. The intensification of agricultural production to meet growing needs can have undesirable environmental or ecological consequences and is contributing to the deterioration of natural resources and living conditions in the developing world. There are two sides to the issue. First, we need to develop technologies that give farmers higher income and minimize enviromnental damage. Second, we need to produce as much food as possible on the most stable land so that people will not be literally driven into the hills because of lack of food. Experience has shown that it is possible to develop responsible agriculture. Though worries about the environment have brought about changes in the world's research agenda, reports of declining rice and cereal yields may eventually create a backlash. It is still uncertain whether a leveling off of yields is a short-term aberration due to unfavorable weather 107 conditions, or whether it is the beginning of a long-term trend that could have disastrous consequences. The loss of production momentum in Asia and the failure under stress of traditional systems of land use, combined with the depletion of nonrenewable natural resources, pose tremendous problems. If the production lid is not lifted, the implications for the future are horrendous given the problem of feeding twice as many people just a few years from now. At the same time, if we don't invest in our natural resources, we may one day lift the lid to find that the world's food basket is nearly empty. References Figueroa, A. 1991. "Commentary on Agricultural Sustainability, Growth, and Poverty Alleviation: Conditions for Their Compatibility in the Tropical Highlands." In Agricultural Sustainability, Growth, and Poverty Alleviation: Issues and Policies. DSE, International Food Policy Research Institute. Feldafing, Germany. Murqueito, E. 1992. Sistemas Sostenibles de Produccion Agropecuaria para Campesinos. En Agroecologia y Desarrollo. CLADES. Numero Especial 213. Santiago, Chile. Zandstra, Hubert. 1981. "Preserving the Options: Food Productivity and Sustainability," available from the Secretariat of the Consultative Group on International Agricultural Research, 1818 H St., NW, Washington, DC 20433, USA. Zandstra, H. G., E. C. Price, J. A. Litsinger, and R. A. Morris. 1986. Metologia de Investigacion de Sistemas de Cultivo en Finca. International Development Research Centre, Ottawa, Canada. Endnotes 1. For a more extensive list of definitions, for example, Vernon W. Ruttan 1991. Sustainable growth in agricultural production: Poetry, Policy and Science. In: Agricultural Sustainability, Growth and Poverty Alleviations: Issues and Policies, DSE, IFPRI. 2. Some believe it cannot even be achieved (Figueroa 1991). 3. This analysis and much of this section is based on unpublished materials of Karl Harmsen, Head Natural Resources Management, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), India and Louise Fresco, Professor of Tropical Agriculture, University of Wageningen, The Netherlands. 4. The most common mechanism suggested is that of a research consortium. This provides for a flexible, open, and participatory process that allows a wide range of IARCs and NARS to participate and benefit. It allows for a sharing of tasks of problem identification, priority setting, task allocation, implementation of research monitoring, and governance. It I Biological Nitrogen Fertilization: Present and Future Applications Ralph W.F. Hardy* Introduction I'm pleased to share with you a very timely opportunity in component technology -- biological nitrogen-fertilization systems. This component technology is relevant to crop production and productivity, to food, to economics, to environment, and to sustainable agriculture, which was discussed by the previous speaker. I plan to provide an overview of the status and identify possibilities for investment at this time. It's timely for an expanded, focused effort on applications of biological nitrogen fertilization. Some of you may refer to this area as biological nitrogen fixation, but I will use fertilization rather than fixation in parallel with synthetic nitrogen fertilization. During the last thirty years an outstanding scientific and technological base has been developed from biological nitrogen research. Probably there are in excess of a thousand scientists worldwide that work in this area. I estimate that we're spending worldwide between US$50 and 100 million a year to support this research activity. We're spending most of this money on the front end of the process -- science and early technology development. Little emphasis is placed on the conversion of this substantial science base to useful products and processes. The major opportunity at this time is a focused, high-quality application effort to produce products and processes. My recommendation to the World Bank and other governmental and nongovernmental organizations (NGOs) is to invest in an integrated focused program to develop the most promising near-, mid-, and long-term applications of biological nitrogen-fertilizing systems. To be successful in this area it's going to take a substantial investment -- US$10 million in year one, $20 million in year two, $30 million in year three, $40 million in year four, and $50 million in year five (in 1992 dollars) continuing for about fifteen years. I believe it will need a consortium of organizations to support the proposed activity. Investors should expect an appropriate return on their investment. It's too long term for the private industrial sector (based on my twenty-two years in the DuPont Company). This area was too long term in the 1980s for the private sector and is even less acceptable in the 1990s where the private sector has an even shorter time horizon in research. I project that products and processes can be expected to be commercialized from as early as three years * up to fifteen years. The major returns are expected from the high-reward products that tend to be the longer-term opportunities. This proposed development should not be centered in universities; nor should it be done in government laboratories. The private sector has the experience to focus, integrate, and manage this application type of program. The private sector is not a leader in front-end research, but the private sector is effective in converting science to useful products and processes. Some of the work may be contracted to academic and government laboratories. President of the Boyce Thompson Institute for Plant Research, Inc., Ithaca, New York. 110 Opportunities Biological nitrogen-fertilizing systems (BNFS) are a huge economic opportunity. The annual cost of fertilizer nitrogen is US$20 to $60 billion worldwide. It is equal to or greater than the US$25 billion agrochemical industry. In addition, there are productivity improvements in legumes, tree crops, and possibly cereal grains. BNFS will enable decentralized manufacturing as opposed to highly centralized huge nitrogen-fertilizer plants. Decentralization would be advantageous to rural communities. BNFS are an integral component of sustainable agriculture and forestry. It enables improved environmental compatibility with decreased groundwater pollution and greenhouse gas production. Synthetic nitrogen fertilization is one of the key components of crop yield improvement during the last thirty to forty years (figure 1). Several years ago I plotted crop yield versus nitrogen- fertilizer application per area. From the 1950s through the 1970s there was s strong correlation. World fertilizer consumption has gone up almost linearly from 1950 to 1990 (figure 2). We were using about 3 million tons of synthetic nitrogen fertilizer in 1950 and increased this twenty-seven fold over the forty-year period so that 80 million tons were used by 1990. Annual global fertilizer nitrogen addition now represents about 50 percent of the nitrogen fixed by natural biological and physical processes. Pest perturbation of a major global cycle by 50 percent is substantial and may be of concern although no specific problem has been identified. Figure 1. Total Nitrogen Fertilizer (Kilograms)/Area under Cereal Cultivation 25C00 971 x x 1960- x 2000 _ X X lS0 XX D - -1966 7 ;j 1 196 1 1 956 10o I I X 25 50 75 100 Total N fertilizer (kg)/area Under cereal cultivation (ha) 111' Figure 2. World Fertilizer Consumption Nitrogen Totals, 1950-1989 70- 4 60- 50 .o 40- = 30-/ 20- 10- 1950 1960 1970 1980 Crop Year A strong case can be made for the need for alternatives to synthetic nitrogen fertilizers (table 1). Fossil energy and economic costs were a major concern in the 1970s. There is a large capital cost required to build the big thousand-plus ton-per-day nitrogen-fertilizer-manufacturing facilities. About a third of the fossil energy cost for corn production up to the farm gate in the United States is used for the production of synthetic fertilizer nitrogen. In addition, there are transportation, storage, and application costs as well as purchase costs. The about 50 percent inefficiency of fertilizer- nitrogen use is a major limitation. There are environmental concerns. Probably the most significant environmental problem in high productivity agriculture is not agrochemical or pesticide contamination of groundwater and soils, but nitrate contamination of groundwater from inefficiently used fertilizer nitrogen. Conversion of nitrate to nitrous oxide results in one of the most potent greenhouse gases; nitrous oxide is 180 times as potent a greenhouse gas per molecule as is carbon dioxide. The increase in tropospheric gaseous nitrogen is correlated more with the increase in fertilizer nitrogen use in the 1980s than it is with the amount of fossil fuel combustion. Then there is the sustainability issue. Synthetic nitrogen fertilizer is not a sustainable practice while BNFS are more sustainable. There are biological nitrogen-fertilizing or nitrogen-fixing alternatives; nature has provided them (table 2). In general, there are two types. In one type a microorganism and a plant form a symbiotic or associative relationship with various degrees of intimacy to produce a system that is able to take the abundant but unusable molecular nitrogen from the air and provide useable nitrogen directly to the plant with 100 percent efficiency of transfer, not the 50 percent efficiency of fertilizer nitrogen. In the best known example there is a class of microbes called rhizobia that form nodules on roots or occasionally on stems of well-known legume plants such as peanuts, soybeans, peas, beans, and alfalfa and some less well-known legumes -- Sesbania and Aeschynomene. There also are legume trees. About 30 to 70 percent of tropical forests are legume trees. 112 Table 1. Need for Biological Alternatives to Synthetic Nitrogen Fertilizer * Fossil Energy and Economic Cost -- Capital Cost for Thousand + Tons/Day Nitrogen-Fertilizer Plants -- Fossil Fuel (35-40,000 feet3 of Natural Gas/Ton Nitrogen Fertilizer) Major Fossil Energy -- Derived Input for Grain Corn Production up to Farm Gate -- Transportation/Storage/Application -- 80 MM Tons - $20-$60MMM Cost * Inefficiency of Fertilizer Nitrogen Use by Crops -- About 50 Percent * Environmental Concerns -- Nitrate Contamination of Ground Water -- N20 from NO3- Denitrification Is Potent (180 x CO) Greenhouse Gas (Troposphere Gaseous Nitrogen Increase Correlated with Fertilizer Nitrogen Increase in the 1980s) 3 Agricultural Sustainability -- Biological Nitrogen Fertilization More Sustainable -- Synthetic Nitrogen Fertilization Less Sustainable Table 2. Biological Nitrogen-Fertilizing Systems Symbiotic Relationships (Many with Agricultural/Forestry Significance) Microbe Relationship Plant Rihizobium Legume Root/Stem Nodules Legumes, for example, Glycine, Arachis, Pisum, Phaseolus, Medicago, Sesbania, Aeschynomene, and so forth. 30 to 70 percent of Tropical Forests are Legume Trees Frankia Nonlegume Angiosperm Nonlegume Angiosperms, for example, Alnus Root Nodules (Alder), Casuarina, Ceanothus, Elaeagnus, Hippophae, Myrica, Shepherdia Azospirillum Other Associations Grasses, Sugarcane Anabaena Intracellular Associations Azolla in Paddy Rice Free-Living (Limited Agricultural/Forestry Significance) Anabaena, Nostoc, Azotobacter, Bacillus, Clostridium, Rhodospirillum, and so forth 113 In another type, a microorganism called Frankha forms a symbiotic relationship with nonlegume plants such as alder and other nonlegume angiosperms. There are other, less intimate associations where organisms such as Azospirillum associate with the roots of certain grasses such as sugarcane. There are relationships where certain algae associate with a fern called Azolla in paddy rice. The top of table 2 is more important for agriculture and forestry and the bottom is less important. Microbes living by themselves fix or fertilize such small amounts of nitrogen that they're really not worth considering for agricultural or forestry purposes. The top part of the chart is the important area, and we need to expand that capability beyond those limited crop and tree plants that now form symbiotic or associative relationships to fix nitrogen. The biological system also has limitations (table 3), but we're living in the era of biology, not the era of the physical sciences which ended ten to twenty years ago. There are few major fundamental discoveries occurring in the physical sciences. Biology is where the action is, and there are major new tools for biology. We have the ability to do things that were previously impossible in biology, and we're very, very early in the learning curve of new biological capabilities. Some early benefits of the new-biology era will be used in the agricultural area in the next few years, and the agricultural base is building. There are real technological opportunities today to address these limitations in BNFS that previously were difficult to impossible. Table 3. Limitations/Technology Opportunities for Biological Nitrogen-Fertilization Systems * Limited Crops and Trees/No Cereal Crops * High Specificity of Plant Host Range of Rhizobia with Legumes, for example, Bradyrhizobiumjaponicum/fredii - Soybeans Only * Poor Competitiveness of Added Rhizobia with Endogenous Soil Rhizobia * Inadequate Formulation/Stabilization * Limited Rate of Nitrogen Fertilization, for example, 100 kilograms of Nitrogen/hectare * Inhibited by Fertilizer Nitrogen * Chemical/Physical Stress Sensitivities * High Energy Requirement -- 12 lb CH20/lb Nitrogen * Oxygen Sensitivity of Biological Nitrogen-Fertilizing Catalyst * Complexity of Biological Catalyst The limitations are the following: * Limited Crops -- For example, BNFS can be used with legumes and some tree crops but not with cereal grains. * Specificity -- There is high specificity, a microbe is specific for a given plant, so you have to have different microbes for different plants. 114 * Poor Competitiveness -- The microbes that are already in the soil tend to compete very strongly with improved microbes that are added to the soil. We need to find ways to improve the competitiveness of added microbes. * Stabilization -- We aren't able to formulate these microbes for stability throughout production, storage, and field use. * Limited Rate -- The bean is an example where there is a very limited rate of nitrogen- fertilizing capability for most of the microbes that are used. Beans probably fix only a few kilograms of nitrogen per hectare in most cases. * Inhibition -- These biological nitrogen-fertilizing systems are inhibited by fertilizer nitrogen. We'd like to be able to combine the use of fertilizer nitrogen with BNFS, for example, integrated nitrogen management analogous to integrated pest management. * Stress -- These organisms are very sensitive to chemical and biological stresses. * Energy Requirement -- BNFS require a high amount of energy to operate; 12 pounds of carbohydrate per pound of nitrogen provided. * Oxygen Sensitivity -- They are extremely sensitive to oxygen. * Catalyst Complexity -- The enzyme nitrogenase is a very complex catalyst. The above limitations, in fact, are best viewed today as opportunities; our knowledge base and biological tools make it possible now to do several key things that were impossible heretofore. Assessment of Possibilities Nine possible product/process applications for BNFS are shown in Figure 3. They are positioned on the basis of low-to-high reward and low-to-high risk with respect to projected success. For example, improved microbes for beans are fairly low risk with modest return opportunity. Other examples include expanding production of microorganisms in areas where they don't have quality production at this time; alternatives to green manuring in rice; improving the characteristics of macrobes; improving the formulation of microbes; giving the capabilities that legumes have for BNF to grain cereals -- one of the major opportunities at this particular time; inventing catalysts that work at room temperature instead of high pressures and high temperatures, and developing associative systems that provide useful amounts of nitrogen fertilization. The ultimate solution is transgenic self-nitrogen- fertilizing plants so that only the plant, not the plant and microbe, are needed. The farmer would only have to use the seed. However, at this stage an effort to produce transgenic self-nitrogen- fertilizing crops is very high risk. There are too many barriers to be overcome. Fifteen to twenty- five years of additional knowledge may enable us to overcome these barriers. I will make a few comments on each objective. The first concerns improved rhizobia for beans. Low rates of biological nitrogen fertilization occur at this time. Some recent data from Mexico and Canada suggest that rhizobias can be found with improved rates of nitrogen fertilization for beans; the amount of nitrogen fixation is actually substantial. This is a fairly low-risk situation with a low plus reward. The major benefit would be for South American food protein production. Another near-term application is the establishment of production facilities for rhizobia and Frankia in developing countries. Inoculation with effective microbes usually produces 20 to 30 percent yield increases. NIFTAL has developed such technology. Stem-nodulated crops may be developed as green-manure crops. The International Rice Research Institute (IRRI) has shown that stem-nodulated crops accumulate 100 kilograms of nitrogen per hectare in a forty to sixty day period. 115 Reincorporation of the vegetation may need to be solved. Overall, the above are low risk and low reward. Figure 3. Biological Nitrogen-Fertilizing Applications High * Tronogenic Self N- Fertilizing PLants Legume sym Genes (Legumes, Cereals, to Cereal Graizs Trees, eo.) for SeLf N- Fertilization Useful Associative Fertilizing Symbiosis for Cereal Grains * Improved Microbial * mbient Abiological N- 31 Formulation Fertilizer Syntliesis Catalyst Rhizobia with Various Imaproved Characteristics Stem Nodule Legumes as Green Manure Expand Rhizobia.'Fro,okia Production, etc. Improved Rhizobia for Beans Low I Low High RISK In the intermediate risk area there are intermediate reward opportunities. These are mid term, which is five to ten years to obtain products. Through selection and genetic engineering approaches rhizobia could be improved with respect to competitiveness, amount of nitrogen fixation, expanded host range, and decreased sensitivity to fertilizer nitrogen. Some of the scientific discoveries are in hand. Improved formulations are needed to stabilize rhizobia. The objective is desiccation tolerance, storage stability, and stability as preinoculated seeds. The opportunities would involve material for formulation, coupled with microbial selection. Ability to formulate is critical for any biological organism and is as important to biological pest control agents as it is for biological nitrogen-fertilizing agents. There are long-term -- fifteen plus year -- applications. Probably the most attractive, based on its projected medium risk and high reward, is the extension of the rhizobial nitrogen-fertilizing system to major cereal grains such as corn, wheat, rice, barley, sorghum, oats, and so on. The approach is to transfer the necessary genes that enable a legume plant to form a nitrogen-fertilizing symbiosis with rhizobia to cereal grain plants. One might describe the approach as teaching the cereal grain what the legume has learned about forming a nitrogen-fertilizing symbiosis. To do this we need to discover what the legume plant has learned or what genes the legume plant has that enables it to form the symbiosis. We call these legume genes necessary for the symbiosis the sym genes. Thomas A. LaRue at Boyce Thompson Institute has, over more than ten years, produced pea mutants and from them identified thirty sym genes. The next steps are to microlocate these sym genes on the pea chromosomes, clone the sym genes, and then make transgenic cereal grains with the 116 sym genes and agronomically develop these sym transgenic cereal grains for self nitrogen fertilization through a rhizobial symbiosis and also, of course, for high yield. The successful outcome would be self nitrogen-fertilizing grains where the need for synthetic nitrogen fertilizer would be eliminated. The key fundamental scientific answer -- the pea plant mutants for the sym genes -- is in hand. A major development effort is needed for the cloning and subsequent steps and is projected to require up to fifteen years with a focused, intensive effort to produce seed that could be introduced into farmers' fields. The opportunity is large and will require substantial investment and continuity. There is a reasonably high probability of success. The reward for success is very large for plant productivity, the environment, and sustainability. Investors should obtain a favorable return. Another longer-term application is to seek useful associative symbioses for nonlegume crops. The recent observation of nitrogen-fixing endophytic microbes in sugarcane by Johanna Dobereiner in Brazil may be such an opportunity. Additional quantitative data are needed to assess the potential of these organisms. Earlier examples of associative symbioses, such as that of Azospirillum and grass sugarcane discovered in the 1970s, have limited utility because of their low rates of biological nitrogen fertilization. The approach here would be to survey and collect organisms with greatest potential and modify or select these organisms for improved characteristics. At this time, the risk is high because of the experience to date. The return could be significant if successful. Undiscovered organisms in nature could be very useful. For example, Allan R. J. Eaglesham at Boyce Thompson Institute discovered in the 1980s a novel rhizobia from potting material obtained from Potomac sand. This novel rhizobia is able to both photosynthesize and fix nitrogen. It forms nodules on stems of a legume called Aeschynomene and has the potential to be energy self-sufficient in biological nitrogen fixation. Furthermore, it is genetically similar to the rhizobia that nodulates soybeans. This nitrogen-fixing photosynthetic rhizobia may be the answer to concerns about the energy cost to the plant of symbiotic nitrogen- fertilizing systems. Transgenic self nitrogen-fertilizing plants, containing the nif genes for biological nitrogen fixation, are a very high-reward opportunity, but the knowledge base makes the risk so great at this time that developmental focus on this objective is not justified. Barriers, such as the need for compartmentalization or protection of the highly oxygen-sensitive nitrogenase enzyme, will be needed. Another opportunity is the design of new catalysts to enable nitrogen- fertilizer-production plants to operate at low temperatures and pressures and to be small in size. Major scientific advances during the last year have provided an atom-by-atom structure of the biological catalyst that fixes nitrogen. It is a complex molecule. I do not see this information enabling the design of a low-cost process. The biological catalyst requires large amounts of a sophisticated molecular form of energy called ATP. It is too high risk to pursue for nitrogen-fertilizer production. It's a very exciting area if you are a chemist trying to figure out how to do new things with nitrogen, a very abundant raw material. Recommendations A summary of my recommendations are presented in table 4. There is a major opportunity for a focused, high-quality effort in application to develop products and processes for biological nitrogen- fertilizing systems for agriculture and forestry. The outstanding scientific and technological base established during the last thirty years in this field and the new tools of biology make this opportunity timely. 117 Table 4. Recommendations * Integrated Focused Program on Developing the Most Promising Near-, Mid-, and Long-Term Applications of Biological Nitrogen-Fertilizing System * Investment of US$10, $20, $30, $40, and $50MM/Year for years one, two, three, four, and five and $50MM thereafter in 1992 US$ for fifteen years. Support by Consortium of Government/Private Organization Supporters. Investors Should Expect an Appropriate Return on Investment * Products/Processes Expected to be Commercialized from as soon as Three Years up to Fifteen Years * Major Return from High Plus Reward Products * Application Program Directed by a Board Comprised of Private Sector Entrepreneurs. Should not be Managed by Government or Academe but by the Private Sector * Economic Opportunity -- Reduced Capital and Operating Costs of Alternative Products/Processes for US$20-$60MMM Fertilizer Nitrogen Market. Productivity Improvements for Legume and Tree Crops and Cereal Grains * Decentralized Manufacturing * Integral Component of Sustainable Agriculture and Forestry * Improved Environmental Compatibility -- Ground Water, Greenhouse Gases I Making IPM Work: Developing Country Experience and Prospects Jeff Waage* Introduction Agricultural production in developing countries is significantly affected by the action of pest insects, nematodes, plant diseases, and weeds. Together these pests cause losses to crop production of the order of US$300 billion annually, about 30 percent of potential global food, fiber, and feed production. Preharvest and postharvest losses to pests have been higher than average in many developing countries. Broadly speaking this can be associated with a tropical or subtropical climate favorable to pest increase and an under-resourced national crop protection capability. But it is possible to be more specific. Pest problems in developing countries today are substantially associated with deliberate efforts to intensify agricultural productivity in order to meet a growing demand for food and for revenue from the export of agricultural products. Agricultural intensification aggravates pest problems through the creation of large monocultures, the introduction of genetically uniform plant varieties, the reduction of intervals between cropping, and the use of agrochemicals. Any crop grown in greater abundance and over larger areas appears to accumulate pest species (Strong, Lawton, and Southwood 1984), and often greater populations of these pests. Programs to intensify crop production in developing countries have often extended areas under cultivation and encouraged large monocrops, because this is more efficient. In many cases these large areas are planted to genetically uniform crop varieties which are easily exploited by adapted pests. Intensification also has involved the reduction of intervals between the planting of the same crop. Around the developing world, periods of fallow, or periodic removal of crop residues or even crops (for example, rampasan) have long been effective means of interrupting the buildup of pest populations. Intensification often leads to overlapping of crops and more continuous resources for pests. Even where a range of crops are grown in sequence this can be a problem, because some of the worst pest species are polyphagous (Moran 1983). Thus, the well-known pest control crisis in the Gezira cotton scheme in Sudan was fueled by the introduction of vegetable crops in the intercrop period between cotton, which allowed the polyphagous cotton bollworm, Helicoverpa armigera, to maintain numbers between cotton crops (Griffiths 1984; Kiss and Meerman 1991). Subsequent intensification of pesticide use greatly aggravated this problem. Finally, an unfortunate side effect of agricultural intensification in developing countries has been the introduction of new exotic pests. Increasing world trade, the movement of plant genetic material by development programs, and even the movement of food aid has brought to the tropics many species from other regions. In favorable new habitats, free of their specific natural enemies, these species often become serious pests. Recent and notable introductions of exotic pests in developing countries include the golden snail on rice (from Latin America to Southeast Asia), the larger grain borer on maize (from Central Director, International Institute of Biological Control, an Institute of CAB International. 120 America to Africa), the mango mealybug (from South Asia to West Africa), the cassava mealybug and mite (from South America to Africa), itch grass, a weed of tropical cereals (from South Asia to Central America), the cypress aphid (from Europe to Africa), and the coffee berry borer (from Africa to South America). In passing it is worth noting that many of these exotic pest problems are amenable to biological control, and most are the subject of successful or ongoing programs of the International Institute of Biological Control (IIBC) and its collaborators. For all of the above reasons, we can expect agricultural intensification to create new and greater pest problems. Therefore, development programs should provide in their design for a specific effort in pest management. To the limited extent that countries and development assistance agencies have made this provision in the past, they have done so largely by making chemical pesticides cheaply or freely available to farmers. Recent experience has shown that this approach can sometimes lead to more problems than it solves. The Pesticide Experience When introduced into pest control programs in the 1950s, synthetic organic pesticides had some remarkably desirable properties. They were extremely toxic to pest insects, weeds, and diseases, not very toxic to man, effective against a range of pests, and relatively cheap. Since then chemical pesticides have encountered a number of problems. Intensive pesticide use has led to the selection of pesticide-resistant strains of many pest species. This has necessitated the introduction of new pesticides, some of which have been more toxic to humans. Cross resistance between old and new products have shortened the life of the latter. Today over 500 species of insects and mites are reported to be resistant to one or more pesticides, and a few are resistant to all major pesticide groups (Georghiou 1990). Herbicide and fungicide resistance is less prevalent but growing (Davies 1992). Pesticides have had undesirable side effects on the environment and on human health in developing countries (Rengam 1992). Early forms, particularly organochlorine insecticides, accumulated in food chains. Newer insecticides have been developed to be less persistent in the environment, but sometimes this has only resulted in their more frequent application. But perhaps the most striking consequence of continued pesticide use has been its capacity to create rather than control problems with serious insect pests. For a number of reasons the natural enemies of insect pests, that is the predators, parasites, and pathogens that limit their numbers naturally, are often more susceptible to pesticides than the pests themselves. Thus, it is possible that an application of pesticide may confer a degree of control by killing pests, but remove a degree of future control, by killing natural enemies (Waage 1989). In some cases this can lead to wasted effort and expense, replacing a free, natural control with an expensive chemical one. In others, where natural enemies are particularly important, it can cause pest problems to get worse. Not only can the target pest resurge, but other crop-feeding species, which have posed no problem before, suddenly become pests because the insecticides have eliminated their natural enemies too. In varying degrees, resistance of insect pests to insecticides and the quite separate effect of insecticides on natural enemies combine to create a phenomenon called the "pesticide treadmill," which goes something like this. Pest resurgence following pesticide applications poses farmers with a dilemma. Having little or no experience with nonchemical approaches to pest control, they have little option but to respond to new or greater pest problems with the application of more pesticides. This further aggravates the problem by making pests more resistant to chemicals and killing even more of 121 their natural enemies that would limit their numbers. Insecticide use increases further. Characteristically farmer's costs increase, as insecticide use goes up, and income declines, as uncontrollable pests reduce yield. Ultimately the cropping system may be abandoned. The pesticide treadmill represents an extreme situation, but one which has been experienced again and again around the world with a range of insect pests in a number of key crop systems including fruit trees, vegetables, cotton, and rice. IPM -- An Alternative Approach Economic problems and crop failures associated with the pesticide treadmill attracted widespread attention and led to research into a new approach to pest control, which earned the name integrated pest management or IPM. Integrated pest management is an alternative to reliance on pesticides, which draws from a range of pest control methods to achieve the most effective, economical, and sustainable combination for a particular local situation. While scientists speak broadly of IPM, it has been largely a response to insect pest problems. In principle, IPM is relevant to weeds and plant diseases, and indeed should integrate management of all crop pests. In reality, pesticide use on weeds and plant pathogens has caused fewer problems, alternatives are less thoroughly explored, and insect IPM proceeds as a model with the potential of future, broader relevance. Among the various methods employed in IPM for insect pests, the most important is biological control (Lim 1990), and particularly the natural control, which already exists in the crop system, conferred by the many predators, parasites, and diseases of insect pests, weeds, and other species. It is this self-renewing, baseline of control, which the farmer can conserve and enhance, and to which he/she can add other IPM components, always mindful of the need not to interfere with the natural enemies that provide it, lest he/she creates problems with pest resurgence, as illustrated above. Cropping practices can be modified to discourage pests or encourage natural enemies, and thereby counteract some of the problems created by large-scale monocultures. Resistant plant varieties can be used. Natural enemies can be augmented at critical periods, so that they build up to controlling levels when pests are abundant. Attractants and repellents, often natural products from insect pests or plants, can be used to influence pest behavior, to trap pests in order to kill them, or to monitor their numbers as a basis for decisions about other control methods. With varied and vested interest in pest management, the precise role of pesticides in IPM has had a number of interpretations. Thus, the argument that IPM should minimize (that is, rather than rationalize) pesticide usage is more easily accepted by environmentalists than by the agrochemical industry. However, for developing countries, a history of pesticide misuse and resulting pest, human, and environmental problems, together with the relatively high cost of chemicals, make minimizing pesticide use a sensible objective of IPM in many cases. Pesticide use in IPM can be reduced by observing changes in pest numbers and using chemicals only when pests reach economically damaging levels, when the consequence will not be more pest problems, and when no other economical control alternatives exist. At the same time, alternatives to so-called broad-spectrum chemicals, which can kill pests and natural enemies alike, are encouraged, including biological pesticides, which are formulations based on specific pathogens of the pest, and more selective chemicals. The assembly of pest control methods into an IPM strategy is best illustrated by example. I will take an example from work done by IIBC at its Station in Pakistan, where we operate as part of 122 the Pakistan Agricultural Research Council (PARC) to help develop IPM on a range of crops including sugarcane, vegetables, cotton, apples, and mangoes. The following study is not yet published. Case Study: Development of IPM Methods for Insect Pests of Mangoes in Pakistan Mangoes are grown in a number of regions of Pakistan, largely for local consumption. Insect pests are a major problem, and of these there are four kinds that have become the targets of pesticide application: mealybugs, fruit flies, scale insects, and leafhoppers. Farmers apply insecticides about five times per year, but still suffer problems with these species. During the 1980s, staff of PARC-IIBC worked with cooperating mango growers in the Punjab to develop IPM methods which would give good, cost-effective pest control. This required developing IPM methods for each pest, while ensuring that these were compatible with methods developed for other pests. Mealybugs (Drosicha stebbingi) feed on the growing shoots and flowers of mangoes and thereby limit fruit production. Studies on their ecology revealed that females lay their eggs in the soil around trees and young larvae move up to the leaves in the spring. As the season progresses, a number of predators, particularly a ladybird (Sumnius renardi), reduce numbers of mealybugs dramatically. On the basis of this understanding, farmers were encouraged to hoe around the base of trees in the winter, to expose and kill eggs. Further studies on ladybirds revealed that they spend the winter in shelters such as rough tree bark. The smooth trunks of mango did not provide this kind of refuge, and hence they have to emigrate from the crop at the end of the season. To see if this was responsible for their late appearance, artificial shelters, simple bands of rough sacking fastened around mango trunks, were put in the orchard. As anticipated, ladybirds used these shelters for overwintering and thus became active in the mango crop earlier in the season, giving better control of mealybugs. The biology of the predators was explained to farmers, and the shelter bands adopted. Fruit flies (Bactrocera dorsalis species complex) received the majority of insecticidal sprays to mango because their maggots, laid in the mango fruit, can greatly reduce the market value of the crop. As an alternative to chemicals, attractant traps were made from cheap local materials and baited with an imported fruit fly attractant, methyl eugenol. These traps proved highly effective, reducing infestations from 35 percent of fruit to 3 percent. Scale insects (primarily Aspidiotus destructor) caused problems in sprayed orchards, where they covered leaves and produced a honeydew that attracted mold. However, experimental studies revealed that they were a secondary pest, brought about by the use of insecticides against fruit flies and mango hoppers, which eliminated their effective natural enemies. Hence, with the use of traps for fruit fly control, the scale insect problem decreased. Mango hoppers (Amritodus and Idioscopus spp) remained the only pest requiring insecticide applications. Ecological studies revealed that the several species involved had a range of different natural enemies, but despite these damaging levels were still reached. This made it difficult to abandon insecticide applications, but careful study of hopper distribution on plants revealed that insecticides only had to be applied to the lower part of the trees, up to 5 meters, to get effective control. This reduced the amount of chemical applied, and also the risk of upsetting biological control of mealybugs and scales. Farmers were encouraged to modify their spraying accordingly. As a result of this program of research in farmers orchards, experimentation, and integration of methods for the four insect pests, annual sprays were reduced from five to one, with a 123 fourteen-fold reduction in cost to the farmer, which more than compensated for the costs of IPM methods. Roughly 25 percent of the 13,000 hectares of mango in the Punjab presently use this IPM method. This case study identifies some typical and important properties associated with development of effective IPM systems: * Research was done on-farm, where farmers' practices and the nature of pest problems could be understood. * A good understanding of the local ecology of the pests and natural enemies, their populations, distribution, and movements was essential to developing IPM methods. * IPM methods involved a mix of technologies, some imported, like pesticides and attractants, but some local as well, like hoeing and shelter bands for ladybirds. * Farmers were closely involved in the process of IPM development, and trained to understand the pests, their natural enemies, and the IPM methods for their control. o The IPM methods returned a net benefit to the farmer. Constraints to IPM Development and Implementation Despite 20 years of research into IPM methods and examples such as the one just described, which show the economic benefits of IPM, IPM is not widely practiced today in developing countries. Pesticide use in developing countries is rising rapidly, and now accounts for about 20 percent of global pesticide sales. Insecticides remain the major kind of pesticide used in developing countries. Some countries like India and China, despite a long tradition in biological control and IPM research, have recently developed national insecticide production industries, concentrating on production of broad-spectrum pesticides particularly harmful to IPM. In recent years considerable effort has been made to analyze the record of IPM in order to identify constraints on its wider application. To chronicle all of these efforts would take pages, and therefore I will cite only two recent ones. In 1989 a group of donor agencies and international organizations concerned about IPM established an IPM Task Force to evaluate IPM development and constraints on its implementation on a global basis. This group prepared a report on IPM in developing countries, with the aid of expert consultants (NRI 1992) and has subsequently organized a series of regional workshops to identify constraints and opportunities for IPM, which are still ongoing. Building on the results of the first such regional workshop in Malaysia, the Asian Development Bank (ADB) and CAB International organized in 1991 a Conference on IPM in the Asia-Pacific Region, to which ministerial-level delegations from twenty-one developing countries in the region were invited. The objectives of the conference were to share experience in pest management and IPM, to develop policy guidelines for IPM implementation by governments, and to develop regional action plans for IPM in key crop systems (Ooi and others 1992). These initiatives, involving scientists, donor representatives, and national policymakers, have considered constraints and opportunities for IPM in considerably more detail than I can here. Instead, I will highlight a few conclusions common to these efforts. The Need for Greater Awareness and Training While the problems associated with pesticide misuse are attracting increasing attention worldwide, the nature and benefits of an IPM approach remain poorly understood outside the scientific community. IPM successes developed in one part of the world are little known elsewhere. Policymakers, crop 124 protection specialists, farmers, and the general public all need to be the subjects of an awareness-raising exercise, presented in an appropriate form for each constituency. The message needs to identify clearly for particular crops the problems with present pest control methods, some simple IPM options, and the economic and social benefits of this approach. Those directly involved in the development of IPM methods; the researcher, extension specialist, and farmer, have a forty year tradition of nearly exclusive reliance on chemical pesticides. Raising their awareness of IPM alone is insufficient to empower them to develop IPM. They need training in understanding pest ecology, monitoring pests numbers, recognition and evaluation of natural enemies, and the use of various IPM techniques (for example, resistant plant varieties, attractants, and need-based pesticide application). Training methods and materials must be designed to be appropriate to these different groups. Better Integration of Researcher, Extension Specialist, and Farmer By its very nature, IPM development and implementation defies the conventional process by which technology-driven research is carried out on the experimental plots of a research institution, packaged, and passed through extension specialists to farmers. This model was developed in pest management for the extension of pesticide technologies. At the outset, at least, it worked fairly well because these products appeared relatively simple to use and highly effective in a range of situations requiring little local adaptation. Furthermore, in the process of research and extension of pesticide technologies, under-resourced national services were augmented by the more substantial advisory services of agrochemical companies. IPM, by contrast, is highly specific to crop, pest, and location. Thus, in the case study on mango pests in Pakistan, a number of IPM methods were closely linked to the local ecology of the pest. The need to provide shelters for ladybirds reflected the composition and design of the particular orchards, as did the designation of 5 meters as the height limit for insecticide application against hoppers. A direct consequence of the site-specificity of IPM is the need to develop IPM methods on-farm. IPM research in centralized research institutions too frequently fragments into single-technology efforts, selects pests that are convenient to study rather than important to the farmer, and develops inappropriate, relatively high technology, and expensive methods. On-farm research not only compels the researcher to consider all pests and their local ecology together, but makes possible as well the involvement of farmers, and the inclusion of their knowledge, perception, and practical economic constraints in the design of IPM methods. Once trained to recognize pests and natural enemies, farmers have a unique capability to be research partners and innovators because of their knowledge of the crop environment. A paper given by Andrews in this symposium presents one of the most exciting examples of farmer involvement and innovation in pest management. Needless to say, the involvement of extension specialists and farmers in IPM development also facilitates subsequent adoption of IPM practices, a point at which many top-down, science-driven IPM efforts fail. The Need for Government Support at the Policy Level Much of the difficulty in developing and implementing IPM stems from a lack of support at the policy level. Certain government policies can be directly antagonistic to IPM, such as government pesticide subsidies that reduce the cost of pesticides to the farmer and distort true economic comparisons with 125 alternative approaches like IPM. In developing countries, government pesticide subsidies, often supported by donor agencies, have been extensive and substantial (Repetto 1985). Like scientists and farmers, many governments have a long experience of strict reliance on pesticides for supporting programs aimed at food security and agricultural export. Hence, many government crop protection recommendations are concerned exclusively with chemical pesticides. Regulation of pesticide use, which could benefit IPM, is frequently ineffective or absent. It would be fair to say that IPM on a national scale is only possible in countries where the government adopts IPM as a central component of agricultural projects. In so doing governments must be prepared to structure their research and extension system to facilitate IPM development and implementation (see above), and to provide the training and resources to support this effort. It will often be necessary to establish interministerial committees to guide IPM implementation and, particularly to coordinate research and extension activities, which are presently poorly linked in many developing countries. Having laid out a daunting set of challenges and prerequisites for development and implementation of IPM, it is time for another case study to demonstrate that this can, indeed, be done! This study concerns the implementation of IPM against insect pests of rice in Indonesia. Much of the value of this study is to be found in its history, which is drawn mostly from Kenmore (1991). Other relevant references, besides those given in the text below, include Whitten and others (1990), Barfield and others (1991), and Fox (1991). Case Study: IPM of Insect Pests on Rice in Indonesia In the 1970s, the Indonesian government, supported by development assistance agencies, embarked on a rice intensification program which involved the widespread use of insecticides and the introduction of high-yielding rice varieties. Production increased, but with it increased a new insect pest, the brown planthopper, Nilaparvata lugens. This tiny insect lives at the base of rice plants and sucks their sap. The planthopper is capable of rapid reproduction and, unchecked by its natural enemies, it can reach numbers that cause rice plants to brown and wither. By 1977, 700,000 hectares of rice production were infested by brown planthopper. In response pesticide application intensified -- by now pesticide use was directly subsidized by the government. In retrospect we know that pesticides removed the important natural enemies of the brown planthopper, such as spiders, and that this loss of natural control outweighed any controlling effect of the pesticides. Not surprisingly, therefore, intensified pesticide application increased the pest problem further. In 1980 rice varieties bearing new planthopper-resistant genes developed at IRRI were widely planted. Following this and a large increase in fertilizer use in irrigated areas, the planthopper problem diminished for four years, although pesticide use was still high. Indonesia reached self-sufficiency in rice. But this achievement was short-lived. Soon, planthopper outbreaks began again in areas planted to new varieties and damage rose to levels reported in the 1970s. It was at this crisis point, with the apparent failure of a program based on existing pest control technologies, that the government of Indonesia bravely chose the alternative path of IPM. The convincing argument for IPM was based on the results of international and national research, which demonstrated that commonly used insecticides induced planthopper outbreaks by eliminating their natural enemies, and that reducing pesticide application, in field trials, led to no loss in yield. This timely understanding and its promotion at a policy level was facilitated by the FAO Intercountry Rice Integrated Pest Control Project and started in 1980. This FAO project supported communication and research by Asian scientists on the role of pesticides and natural enemies in 126 ,brown planthopper control, and thereby facilitated and accelerated IPM promotion and uptake with national programs. It also played a major supporting role in the IPM success, which was to follow in Indonesia. The Indonesian National IPM Policy was announced in November 1986 as a Presidential Instruction, which banned fifty-seven trade formulations of insecticides from use on rice, ordered that resistant varieties be grown in affected areas, more than doubled the number of government field pest observers assigned to rural extension centers and ordered that observers, extension staff, and farmers be trained in IPM. Banning pesticides, and the consequent removal of pesticide subsidies on rice, were a prerequisite for IPM implementation, but alone insufficient. It was the program of training that succeeded in empowering farmers to be practitioners of IPM. Because IPM of brown planthopper relies largely on a knowledge of the local ecology of the pest and its natural enemies, training was focused less on technology transfer than on gaining a personal understanding of the rice ecosystem and how to grow a healthy crop. The traditional extension model of training farmers to use new equipment or pesticides was traded for a more fundamental educational approach, where farmers were encouraged to discover the role of natural enemies, to learn by doing, and to ultimately become IPM experts (Gallagher 1992). The program works through applying these training methods first to pest observers, who are employees of the Ministry of Agriculture. Pest observers receive more than 500 hours of training, over two rice seasons, which includes all aspects of rice production and alternating crops. These trainees then become trainers themselves, and use their skills to train farmers and extension specialists, two for twenty-five trainees. Farmers receive more than 50 hours of training over one rice season, based entirely in the field, and involving the growing of a rice crop using the IPM method. Trained farmers then take this knowledge to their villages and train other farmers. Training methods require little equipment: containers to collect natural enemies and paper and crayons to draw and redraw the rice ecosystem, as knowledge grows. By mid-1992 the numbers of people trained in this method were estimated at 1,000 observers, 3,000 extension specialists, and 150,000 farmers. An additional 300,000 farmers are thought to have been trained subsequently by IPM farmers returning to their villages. By this method, it is planned to extend this training to 2.5 million farmers by 1995 (Wardhani 1992). Today in Indonesia, IPM on rice is a cross-sectoral policy implemented under a steering committee consisting of representatives of the Ministries of agriculture, population and environment, health, education, local government, finance, economic coordination and the Central Bureau of Statistics, all coordinated by the development planning agency, BAPPENAS. The benefits of IPM implementation can be expressed in a range of ways, but the focus of the program has been on increasing profit to the farmer. During the period 1987-90, total application of formulated pesticide on rice has fallen over 50 percent while rice yield and production has risen. Yields of IPM-trained farmers are usually the same or higher than those of untrained farmers. The primary benefits come from reduced pesticide use and its associated costs. At the farm level, the program has so far reduced average insecticide applications per season from over 4 (1986) to 2.1 (1989) to 0.8 for IPM farmers in 1991. As a result, farmers' net profits increase by about US$18 per IPM-trained farmer per season. Dollar for dollar, the increase in farmers' profits is estimated to give a return on training investment of 4.6 to 8.6. (Kenmore 1991). Other benefits of IPM have been the effect of training on improved crop production, as farmers become more observant of changes in their crops, and improvements to occupational health and the local environment from reduced use of pesticides. Also, IPM protects and makes more sustainable those inputs that do contribute substantially to yield, in this case fertilizers and improved crop varieties. Without IPM, and with consequently higher pest populations, the benefits of these inputs would be short-lived, as experience has shown. 127 Perhaps the most convincing measure of success has been the continuing and increasing commitment to the effort at all levels in Indonesia. The federal government, which benefitted substantially from a reduction of 85 percent in pesticide subsidies, has begun to realign the research agenda to support IPM training and field implementation (Wardhani 1992). District administrators are diverting funds to accelerate farmer training, and at the farm level, villages are raising funds to pay for their own training programs (Kenmore 1991). The commitment of the Indonesian government to IPM has been paramount to its success. Development assistance has played a role, initially through support to the FAO Intercountry Rice IPM Project from AIDAB (Australia) and DGIS (Netherlands). In 1989, USAID made a policy development grant of US$10.5 million to Indonesia for the IPM program, and further funding for implementation is anticipated from the World Bank in 1993. By far the largest component of Indonesian and donor investment has gone to farmer training. The scale of this exercise is unprecedented in pest management extension -- in the last two years alone, 2 million person-days of high-quality IPM training have been given to Indonesian rice farmers. The management challenge to national programs of a project of this scale is considerable (Useem, Setti, and Pincus, forthcoming), and will usually require donor support. The Indonesian rice IPM program has proven responsive to change, which is perhaps the greatest evidence of its sustainability. In 1990 a long-standing, but usually minor pest of rice, the white stem borer, Scirpophaga innotata, reached outbreak levels in Java. Calls to relax restriction on pesticide use were resisted by the federal government, and national funds were allocated to training farmers in pest recognition, which permitted manual control by removing the pests' egg masses. Pesticides suspected of actually contributing to the problem were avoided and losses in the following year were much reduced. Research was initiated, parallel to training, through the existing IPM program. Rather than weakening the commitment to IPM, this new pest problem has strengthened it, and demonstrated the power associated with broadening the IPM constituency from crop protection specialists to policymakers, extension specialists, and farmers. Getting IPM Going The history of the Indonesian rice IPM program illustrates the need for policy support and change, training and awareness raising, and the reorganization of the research-to-extension-to-farmer model. It also provides some insight into how IPM programs can be initiated. Making the Most of IPM Entry Points Crises in pest management, such as the pesticide treadmill experienced in Indonesia, provide entry points for IPM. In such situations it is often possible to quickly improve the situation of farmers simply by pesticide reduction, thereby gaining support at the farm and government level for further development of IPM. The identification and exploitation of IPM entry points is the most effective way to initiate the adoption of IPM. Once developed for a particular pest, IPM tends to spread to cover other pests in the crop system, as we have seen for white stem borer in rice. This is due in part to the fact that farmers now have an incentive to avoid pesticide use that might disrupt existing IPM. More importantly, however, an IPM approach and associated training makes farmers, researchers, and extension specialists more aware of opportunities to develop IPM for other pests. 128 The self-spreading effect of IPM can also be seen to operate between crops. In Indonesia, where farmers alternate rice with other crops (palawija), including vegetables and soybean, a training/research program has begun on IPM against pests of these crops, which after all support many of the natural enemies that will be important in next seasons' rice production. As already observed, however, little spread of IPM is likely unless it is adopted as a national crop protection policy. The recent Conference on IPM in the Asia-Pacific Region identified particular entry points for IPM in rice, cotton, vegetables (particularly brassicas), sugar cane, fruits, and plantation crops. All of these crops experience excessive pesticide use in some areas, which could be rapidly reduced at a saving to the farmer. For all of them, experience of successful IPM exists for key insect pests in certain areas that could form the basis for developing IPM in new areas, through adaptive, local research. The Value of Intercountry IPM Networking The FAO Intercountry Rice IPM Project was an essential catalyst to the Indonesian program, as it continues to be to equally promising national rice IPM programs in other Asian countries. As a regional, donor-funded effort, it fostered research and information sharing in the region, generating a considerable knowledge base from relatively modest efforts of national and international research institutions. It was this activity and the participation of Indonesian scientists in it, that provided at the critical time, convincing evidence from the field that IPM would work, and that farmers could be trained to use it. A similar, regional activity is necessary to backstop efforts in other commodities that can serve as IPM entry points. Such an effort should involve a large component of the technical cooperation between developing countries (TCDC), because there already exist successful or at least promising IPM programs somewhere in the world for most developing country crop systems. Over the past year, efforts of this kind to stimulate uptake of IPM in developing countries have begun. As a direct result of the Conference on IPM in the Asia-Pacific Region, the Asian Development Bank will support CAB International to manage an Asian regional cotton IPM program, to involve India, Pakistan, and China. This will be coordinated with a larger regional initiative by FAO on cotton in Asia, which will include a greater range of countries. In addition to this, and to the continuing Asian rice IPM program, FAO is also planning regional vegetable IPM initiatives in Asia and Africa. Distinctly missing from these initiatives is regional support for IPM in plantation crops, like sugar cane, coffee, coconut, citrus, cocoa, oil palm, and so forth. This undoubtedly reflects a view that private industry can support IPM development, and there is some precedent for this -- indeed some of the first IPM programs were developed for plantation tree crops in Asia (Liau 1992). However, many plantation crops are grown by smallholders who have poor access to the IPM experience through government or company-based extension. Furthermore, the commercial nature of these commodity industries has itself limited the exchange of information between countries. If anything, the relatively organized nature of the plantation sector, and the ease of implementing pest management decisions in it, would ensure a high return on the relatively small investment necessary to raise awareness and exchange information on the IPM experience. Regional programs of technical networking and support logically focus on particular commodities. However, governments can develop pest management policy, such as that regulating pesticide use, to cover all crop systems. It is important, therefore, that support to developing and 129 implementing IPM in particular commodities take into account the need to provide governments with the information and stimulation to extend IPM over all crops. Extending IPM Further -- Being Proactive In the arguments and case studies presented so far, the emphasis has been on crop systems where chemical control of insect pests is presently causing problems, thereby creating opportunities for IPM. But many crops in developing countries are grown at a subsistence level, and receive little pesticide application. What is the role for IPM here? A number of experts have considered this question and arrived at similar conclusions (Teng 1989, Lim, 1990; Kiss and Meerman 1991). For subsistence crops IPM should aim to optimize the resources inherent within the system to improve yield, identifying and making use of natural biological control, cultural controls, improved crop varieties and other methods that involve low, realistic inputs. IPM should be developed as part of a general approach to healthy crop production as in the Indonesian case. Many farming systems sit today poised between subsistence and intensification. Returning to the argument with which I began this paper; if we have as a priority the intensification of agricultural production in developing countries, then we can expect, for good ecological reasons, an intensification of pest problems. If we are to propose a more sustainable alternative to complete reliance on chemical pesticides, we should be developing an IPM approach in these crops now in advance of problems which demand this. As Kiss and Meerman (1991) point out for the situation in Africa, where intensification is anticipated in a range of crops, farmers not yet used to chemical control may be more responsive to alternatives. Perhaps the challenge will be more one of convincing governments and donors to invest in the development of an IPM approach in an intensification program in advance of perceived pest problems. A particularly good example of a crop system suitable for the proactive development of IPM is the rapidly expanding horticultural and fruit export industry of developing countries. In South America, Africa, and Asia, this industry is rapidly outstripping conventional agricultural exports as a major generator of foreign exchange income. The crops are frequently new -- exotic fruits, vegetables, spices and flowers -- as are many of the pests. Indeed many if not most of the insect pests are exotic, brought in through incautious movement of planting material. Without their indigenous natural enemies, and with a poor national capability in diagnosing or managing these new problems, pesticide use can be expected to be high. The need to provide unblemished produce for many of these commodities will further aggravate this situation. Opposing this tendency, however, will be the increasingly stringent minimum residue levels for pesticides imposed by North America, Europe, Japan, and other importers. Developing country producers are, in effect, stuck. Without the rapid development of an IPM approach, it is difficult to see how a crisis in pest management in this important new agricultural sector can be avoided. Virtually all of the insect pests associated with tropical high value horticulture are well known to field or protected cultivation in industrialized countries. Virtually all are prone to resurgence after pesticide application, and virtually all can be effectively controlled using nonchemical methods. With this experience, and the organized nature of the horticultural industry, prospects for preventing another, costly pesticide treadmill are good. An appropriate first step would be a modest regional IPM program, of a form similar to that described above, aimed at supporting and communicating on farm research into IPM methods. 130 Implications for Development Assistance Agencies Of the many implications that IPM holds for development assistance, and particularly for its support to agricultural intensification to support food security and wealth creation in the developing world, I would like to close by drawing out three. Environment, Agriculture, and IPM Donor agencies are increasingly aware of the problems of pesticide use and the opportunities presented by IPM to reduce these. To a large extent, this awareness comes from their national constituencies, in the form of concern for the environmental implications of pesticide use. In the North, the movement toward IPM is driven largely by conservation and health issues. In the South, as we have seen, it is presently driven largely by economics, although environmental and health concerns are rapidly gaining ground. Thus, while developing country representatives prioritize for pesticide reduction and IPM for those crops where it will improve farm economies (Ooi and others 1992), developed countries like Denmark and the Netherlands can comfortably make decisions to reduce pesticide use by 50 percent over all crops! The Northern, environmental slant on IPM has had implications for the way IPM is viewed in development assistance programs. Presently, for instance, guidelines for pesticide use in donor agencies stress human and environmental safety. The emphasis is on identification of acceptable and unacceptable chemicals and safe use and disposal of pesticides. While guidelines of this kind pose extremely important constraints on the provision of pesticides to agricultural intensification programs, they often fall short of specifying an IPM approach, that is, the development of the most economical and sustainable pest management strategy, without the presumption of a reliance on chemical pesticides. There is much scope for reconsidering donor guidelines for pesticide use in the context of recent experience of IPM in agriculture in developing countries. Not only can a stronger case be made now for IPM, but experiences like that in Indonesia have revealed the extent to which IPM is not simply a technical intervention in a crop production program, but an exercise in human resource development intimately linked with other such activities in the agricultural sector of developing countries. IPM stands out as both economically sensible and environmentally sound, and therefore can draw strength from a wide constituency in donor agencies and developing countries. As an environmental initiative, it is one of the more easily realized challenges. In Agenda 21, for instance, it is proposed that substantial uptake of IPM at the farming community level can be achieved before the end of this century, and indeed this is possible. Research to Develop IPM Traditionally, donor support to IPM development has passed through national and international agricultural research institutions. The growing consensus that technology-driven, top-down research does not lead quickly to successful IPM should therefore give donors pause to think. Success appears to be more closely associated with a more decentralized, on-farm research and training effort involving scientists and farming communities. 131 While the farm-level focus of IPM research, and the necessary restructuring of the research-extension-farmer model might be seen to threaten established research institutions, nothing could be further from the truth. Rather, the development of IPM will help to give these institutions a more meaningful agenda in the context of national agricultural development. By investing more in IPM research and development at the farm level, and encouraging the participation of local governments, community organizations, and other nongovernmental organizations (NGOs), donors can ensure that the problems addressed by national and international research institutions will, in future, be more demand-driven and less technology-driven, which can only make these institutions more sustainable. IPM Implementation and Its Need for Donor Support As the Indonesian rice IPM case has indicated, the scale of IPM implementation at the farm level involves substantial commitment from governments and donors, particularly to training, and it takes time, even at the remarkable rate that has been achieved in Indonesia. Furthermore, at the level of implementation, IPM becomes a program of human resource development, intimately linked to other aspects of crop production. In developing programs of agricultural intensification, donors and governments should therefore view IPM not as a technical component for expert consultation, but as a broad educational program to involve farmers and other groups, backed up with resources for local research. As such, IPM will not only be most successful, but it may come to serve as a model for bringing together farmers, govermnent, scientists, and others to effect positive change in agricultural development. References Barfield, C., J. Mangan, J. Markie, R. Robers, S. Padmanagara, and S. Soehardjan. 1991. "Mid-Term Review Mission of the Government of Indonesia and FAO in Association with USAID - Training and Development of Integrated Pest Management in Rice-Based Cropping Systems." UTF/INS/0678/INS. Mission Report, July 1991. Davies, W.P. 1992. "Prospects for Pest Resistance to Pesticides." In Abdul Aziz, S.A. Kadir, and H.S. Barlow, eds., Pest Management and the Environment in 2000. Wallingford: CAB International. Fox, J. 1991. "Managing the Ecology of Rice Production in Indonesia." In J. Hardjono, ed., Indonesia.- Resources, Ecology and Environment. Oxford: Oxford University Press. Gallagher, K.D. 1992. "IPM Development in the Indonesian National Program." In Abdul Aziz, S.A. Kadir, and H.S. Barlow, eds., Pest Management and the Environment in 2000. Wallingford: CAB International. Georghiou, G.P. 1990. "Overview of Insecticide Resistance." In M.B. Green, H.M. LeBaron, and W.K. Moberg, eds., Managing Resistance to Agrochemicals. Washington, D.C.: American Chemical Society. 132 Griffiths, W.T. 1984. "A Review of the Development of Cotton Pest Problems in the Sudan Gezira." MSc Thesis, Imperial College, University of London. Kenmore, P.E. 1991. "Indonesia's Integrated Pest Management: Policy, Production and Environment." Paper presented at USAID-ARPE Environment and Agriculture Officer's Conference, 11 September, Colombo, Sri Lanka. Kiss, A., and F. Meerman. 1991. "Integrated Pest Management and African Agriculture." World Bank Technical Paper 142. Washington, D.C. Liau, S.S. 1992. "The IPM Experience in Plantation Crops." In P.A.C. Ooi, G.S. Lim, T.H. Ho, P.L. Manalo, and J.K. Waage, eds., Integrated Pest Management in the Asia-Pacific Region. Wallingford: CAB International. Lim, G.S. 1990. "Development and Implementation of Integrated Pest Management in the Developing Tropics." In Proceedings of the Third International Conference on Plant Protection in the Tropics, Vol. 5. Kuala Lumpur: Malaysian Plant Protection Society. Moran, V.C. 1983. "The Phytophagous Mites and Insects of Cultivated Plants in South Africa - Patterns and Pest Status." Journal of Applied Ecology 20: 439-450. NRI (Natural Resources Institute). 1992. Integrated Pest Management in Developing Countries. Experience and Prospects. Chatham, UK. Ooi, P.A.C., G.S. Lim, T.H. Ho, P.L. Manalo, and J.K. Waage. 1992. Integrated Pest Management in the Asia-Pacific Region. Proceedings of a Conference on Integrated Pest Management in the Asia-Pacific Region, 23-27 September, 1991, Kuala Lumpur. Wallingford: CAB International. Rengam, S. 1992. "IPM: The Role of Governments and Citizens' Groups." In P.A.C. Ooi, G.S. Lim, T.H. Ho, P.L. Manalo, and J.K. Waage, eds., Integrated Pest Management in the Asia-Pacific Region. Wallingford: CAB International. Repetto, R. 1985. "Paying the Price: Pesticide Subsidies in Developing Countries." Research Report 2. Washington, D.C.: World Resources Institute. Strong, D.R., J.H. Lawton, and T.R.E. Southwood. 1984. Insects on Plants: Community Patterns and Mechanisms. Oxford: Blackwell Scientific Publications. Teng, P.S. 1989. "Integrated Pest Management in Rice: An Analysis of the Status Quo with Recommendations for Action." Report prepared for a Task Force of the Technical Advisory Committee of the Consultative Group on International Agricultural Research reviewing international activities in IPM. Useem, M., L. Setti, and J. Pincus. (forthcoming). "The Science of Javanese Management: Organizational Alignment in an Indonesian Development Program." Public Administration and Development. 133 Waage, J.K. 1989. "The Population Dynamics of Pest-Pesticide-Natural Enemy Interactions." In P.C. Jepson, ed., Pesticides and Non-Target Invertebrates. Wimborne, Dorset: Intercept. Wardhani, M.A. 1992. "Developments in IPM: The Indonesian Case." In P.A.C. Ooi, G.S. Lim, T.H. Ho, P.L. Manalo, and J.K. Waage, eds., Integrated Pest Management in the Asia-Pacific Region. Wallingford: CAB International. Whitten, K., K. Eveleens, L. Brownhall, K. Sogawa, G.S. Lim, and A. Khan. 1990. "Mid-Term Review of the FAO Intercountry Program for the Development and Application of IPC on Rice in South and Southeast Asia." GCP/RAS/092/AUL, GCP/RAS/101/NET, GCP/RAS/108/AGF. Mission Report, November 1990. I Changing Perceptions and Practices of Central American Smallholders Keith L. Andrews, Jeffery W. Bentley, Rafael Diaz D., Elias Sanchez and Fransisco Salinas* Jeff Waage (1993) has given us a comprehensive and insightful look at what integrated pest management (IPM) is, and he provided us with interesting examples from Asia. Several other important successes in Latin America should be mentioned (Andrews and Quezada 1989). Among the most notable cases are bananas in Central America, sugar cane and coffee in many of the countries, soybeans in Brazil, and melons in Honduras. As with the Asian rice example which Jeff Waage presented, the Latin American successes are with intensively produced cash crops. Before discussing our smallholder or hillside farmer IPM program in Central America (especially Honduras) we need to look at the recent history of the soil conservation and regenerative agriculture movement in the region. For nearly twenty years, Asociaci6n Coordinadora de Recursos para el Desarrollo (ACORDE) headed by Elfas Sanchez, Finca Loma Linda, has taught thousands of Honduran smallholders about organic gardening and soil management and conservation. In 1982 ACORDE helped a small group of Guatemalan extension agents find asylum in Honduras. This was the core of people that started the World Neighbors program in Honduras. World Neighbors, the Peace Corps, Partners of the Americas, Catholic Relief Services, and dozens of other private and public sector extension agencies eventually began extending appropriate technologies of organic fertilizers and soil management as elements of community and human development projects. Despite their low budgets, these agencies have enjoyed high adoption rates of their technologies, because of their participatory, grassroots approach, which is ably summarized by Bunch (1982). These nongovernmental organizations (NGOs) have been effective in the challenging context of marginal, hillside farm communities. A tribute to their success is that the Honduran public sector extension programs began emphasizing similar soil conservation technologies. In 1983 Zamorano (Escuela Agrfcola Panamericana) started the Integrated Pest Management Project in Honduras (MIPH) with the U.S. Agency for International Development (USAID) funding. The project evolved into the Crop Protection Department (DPV). From the beginning, Zamorano's IPM program collaborated with farmers. Together they developed a menu of different pest control practices for most pest complexes. The technologies were transferred to farmers during extension trials on their farms (Goodell, Andrews, and L6pez 1990; Bentley and Andrews 1991). Later Zamorano developed a biological control center to enhance natural control of the major pests in Central America (Cave 1992). Zamorano now has a massive educational program for rational pest and pesticide management to reduce health and environmental hazards from pesticide abuse; this program has been active in all Central American countries as well as in the Dominican Republic and Bolivia. Zamorano's maize, bean, and sorghum IPM program has demonstrated the ability to develop nonchemical solutions to most of the common problems that rural families try to solve with pesticides. Keith L. Andrews, and Jeffrey W. Bentley are with Zamorano, Apartado 93, Tegucigalpa, Honduras; Rafael D(az D. is with World Neighbors, Apartado 3385, Tegucigalpa, Honduras; Elfas Sanchez is with Granja Loma Linda, Apartado 163-C, Tegucigalpa, Honduras; and Fransisco Salinas is with Catholic Relief Services, Apartado 257, Tegucigalpa, Honduras. 136 Zamorano's IPM program worked with consulting anthropologists from the beginning and hired one full-time (Bentley) in 1987 to study linguistic and cultural aspects of smaliholder pest control. The anthropologists have developed culturally sensitive research and training techniques with a group of agronomists at the DPV (Bentley 1989, 1990, 1991a, Bentley and Melara 1991). These three threads of soil management/conservation, IPM, and anthropology came together in 1990 when World Neighbors personnel in Honduras asked Zamorano to help train their staff. The combination of Zamorano and World Neighbors was natural and complementary. Zamorano had technical skills in pest management and the ability to discuss and modify them with small farmers, but did not have a large cadre of effective, low-cost extension agents. World Neighbors had many extension agents from rural villages, good communication skills with farmers, and the winning technologies of soil conservation and organic fertilizers, but not the ability to develop alternatives to inappropriate pesticide use and increasing pest problems associated with cropping system intensification. Because the extension agents are from the people themselves, unless they are given new skills and training, the extension agents have little or nothing new to teach the farmers. Without the extension capabilities of the NGOs, Zamorano's research and development capabilities can't reach the real world. In late 1991, Zamorano, World Neighbors, ACORDE, and Catholic Relief Services formed the national IPM consortium and in mid-1992 started an extension program with their own resources. Despite having "bootlegged" all support up to now, the results are so encouraging that we have begun to enlarge the Honduran program and expand it to other Central American countries. We expect to obtain support from the United Nations Development Programme (UNDP) in the very near future. The project, which is explained in this paper allows the cooperating organizations to develop a synergistic approach to a complex set of problems. Our technical collaboration with smallholder farmers is based on learning what the people know and what they don't know, figuring out what they need to know, teaching it to them in a way that is consistent with what they know, and then learning from them as they synthesize new information with old knowledge. Our program includes a series of short courses that emphasize hands-on learning and introduction of new concepts and information that increase farmers' abilities to innovate. The first course has been given to 500 change agents in three countries, and focuses on insect reproduction and biological control (Bentley 1992a, b). Future courses already being developed will cover applied plant pathology, soil pest and disease control, and pesticide safety. Our pest management goal is not to simply eliminate pesticides from the farms. We must help farmers to develop alternative pest control, and to increase and stabilize food production. This is especially important in those cases where farmers face new pest problems as a result of having adopted soil management practices like cover crops, reduced tillage, and not burning crop residues. To use Jeff Waage's terminology, the "entry point" for this IPM project is the NGOs' desire to protect and enhance the gains they have made in soil management. IPM specialists have a responsibility to keep soil management programs from losing momentum because of new pest problems with which they are associated. In other words, from the start this effort places IPM within a more holistic, regenerative agriculture initiative with poor, hillside farmers as the primary clients. From the beginning of this program we assumed that a participatory research and extension methodology was essential. Marginalized Central American smallholders are keen observers and experimenters who know a great deal and constantly innovate. They should be good research collaborators. However, in practice their contributions have been limited because they have little notion of some key elements in the bioecology of pests, biological causes of disease, biological pest control, and other important aspects of sustainable pest control (Bentley 1989). The gaps in their knowledge contribute to a feeling of helplessness in the face of pest problems and in the past helped create dependency on synthetic pesticides. Earlier extension efforts often met with polite puzzlement and empty promises to implement technologies that the farmers simply did not understand. When 137 asked to suggest priorities for cooperative research, most farmers proposed studies of pesticide efficacy (Bentley and Andrews 1991). We now believe that farmers will be best able to innovate and adapt the practices they need for each of their many natural and economic environments if they are previously trained in the key concepts that they have lacked (for example, biological control, insect life cycles, germ theory of disease). In order to understand our training program, we need to look at the results of the anthropological studies carried out by Bentley and his cooperators to document the nature of campesino or peasant knowledge. On some topics this knowledge is encyclopedic, but in other areas it can be distressingly incomplete, and from the point of view of a positivist scientist, even wrong. For example, farmers are often experts at handling crops and livestock, but do not understand what causes disease, and think that insects are spontaneously generated. While farmer participation is now widely promoted, most descriptions of "participation" are more rhetoric and wishful thinking than substance. They appear to be based too much on a naive desire to work together than on a businesslike understanding of the complimentary strengths of the partners. In order to make farmer-scientist interactions more fruitful, we need to honestly confront the limitations as well as appreciate the strengths of indigenous technical knowledge -- just as we need to be realistic about what researchers understand and can do. Bentley (1991b) presented a simple scheme (figure 1) to help us understand why peasant ethnoscience is so variable. This model can be applied to explain the knowledge of Bombay bus drivers, Bank bureaucrats or any other group of people, but in this paper we will concentrate our discussion on its application to Central American smallholders' pest management understanding. The level of understanding of any subject is a function of two factors: the perceived importance of a thing or phenomenon and the ease with which it can be observed. Importance is defined as meaning of perceived value or harm to the local people; the concept includes economic utility and potential to cause physical pain. For example, humans perceive bees as important because they produce food and their stings hurt and occasionally kill people. Whether or not a creature is conspicuous or easily observed depends on its size, color, movements, time of activity, and perceived risk to the observer, and is also influenced by cultural attitudes (such as, "spontaneous generation is common," or "all insects are bad"). We assume that there is a strong relation between knowledge and behavior and that by acquiring knowledge the learner may adopt more effective behaviors. Figure 1. Four Classes of Farmer Knowledge IMPORTANCE E 0 Social wasps A B Mud dauber wasps Bees S S + Earwigs Weeds E E Spiders Farm tools R Insect predation Plant growth stages O V F A A- _ T Parasitic wasps Bean diseases I _ Nematodes Lepidoptera reproduction 0 N 138 Importance and ease of observation can be diagrammed as two axes that divide folk knowledge into four cells (figure 1), with different taxonomic structures and unique classes of knowledge. In the upper right-hand cell of the figure are the important, easily observed topics like social insects, weeds, farm tools, and plant growth stages. These domains are rigorously classified and well understood. The upper left-hand cell includes easily observed but unimportant entities like large, nonstinging arthropods. Earwigs, spiders and mud dauber wasps are good exarnples; these animals are named but are neither highly differentiated taxonomically nor connected with much cultural lore. The lower right-hand cell includes important but difficult-to-observe topics such as many plant diseases and most aspects of insect reproduction. These are named and, although not split into many folk categories, are the focus of cultural beliefs that may be at odds with western science. The lower left-hand cell holds those topics that peasants consider to be both unimportant and difficult- to-observe. Topics like parasitic wasps, which campesinos are generally unaware of, are not even named. Although many of the prime examples presented here come from insect taxonomy, this philosophical scheme can account for much indigenous technical knowledge about the natural environment. This scheme is about ideas rather than biological organisms per se. Some organisms are easy to classify according to one of the four classes of knowledge, while others must be teased apart. Honduran folk knowledge of ants, for example, falls in at least two classes: stinging behavior and seed eating are in the "important, easy to observe" class, while ant reproduction and predation fall in the "unimportant but easy to observe" cell. Classes of Knowledge The various kinds of knowledge have very different formal characteristics and, as we will see, present vastly different opportunities for participatory research and extension work. Conspicuous and Important:` "Thick Taxonomies" Phenomena that are conspicuous and considered to be important tend to be organized into many folk categories in a taxonomy five or six layers deep. Conspicuous and important organisms are often labeled at the biological species level. Explanations of these phenomena - the quality of honey, the painfulness of wasp stings -- are often couched in "positivist" terms, that is, the explanations are consistent with scientific knowledge and acceptable to scientists. With the help of our colleague, entomologist Ronald Cave, we have found that Honduran campesinos generally categorize social bees and certain wasps to the species level. Campesinos must gauge bee defence strategies and honey quality to decide whether to chop a tree down and split it open for honey. Campesinos describe various kinds of honey as medicinal, good to eat, nasty, and potentially poisonous. Conspicuous but Unimportant: Strings of Folk Genera Conspicuous but unimportant phenomena are often classified in a taxonomic structure with many categories, but few levels -- shallow strings of dozens of names with no subordinate and few superordinate categories. Conspicuous but unimportant organisms are often labeled at the biological 139 family or order level. There is little attempt at explanation, positivist or otherwise, for phenomena in this group. As much as anthropologists like to portray traditional rural people as able taxonomists, exhaustive studies of folk taxonomies often reveal many animal names with little paradigmatic structure. Hunn (1977) found that for Tzeltal speakers of Mexico, 106 of 335 individual names for animals were classed as birds, another 45 as mammals, while 184 names, many of which were insects, were not included in higher taxonomic levels (except for that of animal) and most of the 335 names include no subdivisions. Honduran campesinos do not think of any "bugs" (terrestrial arthropods) other than honey producers as beneficial, so most insects are classified in a shallow taxonomy and are given folk genus names with no species subdivisions. Campesinos lump the entire order of Dermaptera (earwigs) together as tijerillas (little scissors), just as most spiders are undifferentiated araflas and all dragonflies (order Odonata) are merely caballitos del diablo, "the devil's little horses." Being conspicuous is no guarantee of even a unique name for animals with no perceived economic importance. The mud dauber wasps (family Sphecidae) are highly conspicuous, building nests shaped like organ pipes, footballs, and mud clods on houses and other buildings. Campesinos see the wasps hauling spiders or grasshoppers into the nests and know that they rear their young there, but because sphecids are considered to be useless and harmless they are merely lumped into the residual category "just wasps," sharing the name avispa with the vespids and other wasps. Many campesinos claim that sphecids have no name, or that they do not know it. Important but Difficult-to-Observe: the Enigmas Nothing is more maddening than a real problem with no obvious solution, like many insect pests and crop diseases. The voracious worms that seem to appear fully grown from nowhere, others that descend on a field by the thousands overnight and diseases that suddenly wipe out whole fields rank high on the importance scale, but are hard to observe. Magic-religious explanations or other 'odd' unscientific sounding beliefs about insects and other organisms are likely to occur in the important but difficult-to-observe cell. Important but difficult-to-observe phenomena may or may not have complex taxonomies, depending on biological factors. For example, bean diseases in Honduras are poorly classified, with viral, bacterial, and fungal disease, nutritional deficiencies, certain insect damage, and other ailments all grouped together. Some insect pests are classified at the biological species level, although knowledge of their behavior, especially of their reproduction, may be poorly understood. 'Folkloric' explanations (for example, spontaneous generation), often at odds with positivist science, are much more common than they are for other kinds of knowledge. Multiple diseases are more difficult to observe and differentiate than one disease. Campesinos confuse many bean diseases (Bentley 1991 a) but because there is only one major maize disease in Honduras, maize ear rot, farmers are able to focus on the disease and acquire a body of knowledge comparable in some aspects to that of plant pathologists. Honduran campesinos have formed many of the same hypotheses as specialists for solving this disease problem, including increased soil fertility, quicker drying of the grain, burning crop residues, and bending the maize plant over (Bentley 1990). Although insect pests are some of the few insects other than bees and wasps, which campesinos classify at the biological species level, farmers have a poor understanding of caterpillar reproduction. The cogollero or whorlworm, Spodopterafrugiperda, is an endemic maize pest which campesinos perceive as chronically lowering crop yields. Because it is very tiny when it first hatches 140 and glides through the air on a silk thread, landing inconspicuously on the earth and making its way to maize plants, campesinos do not notice the cogollero in its early instars. They notice the little windows the tiny larvae carve in maize leaves, eating off the green tissue and leaving a transparent film in the center, but many fail to distinguish those windows from the damage of leaf miners, a host of small insects of different orders that work in the completely opposite way, by eating out the interior of the leaf. Campesinos notice whorlworms when they are large caterpillars eating the tender new tissue of the maize whorl, and burying themselves in their own feces. Farmers believe that the worms are generated spontaneously by the corn plant itself, citing as evidence the fact that smashed whorlworms are green, like maize plants (Bentley and Andrews 1991). Not only material factors (size, mobility, and so on) influence how easily organisms can be observed. Cultural attitudes also affect how people see the world around them, even though those attitudes may have been shaped in part by the biological structure of that world. For example, Hondurans, both campesinos and most of the middle class, believe that all insects are bad except bees. Virtually all insects are thought to be herbivorous. While this belief may have a basis in the observation of abundant plant eating insects in the tropics, it also affects campesinos' vision of their fields as being virtually under siege to insect pests; plant protectionists know that many of the pesticide applications made are not justified economically (Shaxson and Bentley 1991). After learning to recognize the beneficial aphid lion, Chrysoperla spp., one farmer said "and I used to apply pesticides when I saw those in my field!" We have known university trained producers who have applied pesticide in eighty hectares of maize to control earwigs that are effective whorlworm predators. Difficult-to-Observe and Unimportant: the Empty Quarter Because difficult-to-observe and unimportant phenomena are not usually categorized, they fit into no folk taxonomies and are not labeled at any levels of biological classification. They are accompanied by no folk explanations. Many organisms are neither named nor paid any attention to, because they are both difficult to observe and not perceived as important. Because they are so small, none of the wasps in the four major families of parasitic Hymenoptera in Honduras is even recognized by farmers, let alone seen as pest controllers. Each herbivorous insect has at least one parasitoid wasp, and sometimes dozens, as well as nematodes, flies, and other tiny organisms whose lives are intertwined with the host they feed on and kill. If not for these little creatures, Central American farmers would starve; yet the wasps are neither named nor known. Some additional examples of farmers' knowlege of pesticides provides a review and helps illustrate the value of this scheme. * Important and easy to observe. Acute pesticide poisoning and the quick die off of pests following an application are important and easy to observe, and farmers understand them. * Easy to observe but unimportant. Farmers could easily see that earwigs, spiders, wasps and other beneficial insects are killed by pesticides, but as long as the biocontrol role of these animals is not appreciated, their deaths are considered unimportant. * Important but hard to observe. Many farmers and NGOs are starting to think of botanical insecticides as important and desirable. Many odd, almost mythical, beliefs about the characteristics and efficacy of botanical insecticides are spreading rapidly. But the many plants with their complicated and ephemeral chemical structures are hard to study. We are far from filling the research demand for this topic. 141 * Hard to observe and unimportant. Although chronic pesticide poisoning is important, it is so difficult to observe that for many years farmers didn't notice it, and didn't realize its importance. Implications for Technology Generation and Transfer The anthropological model presented above is useful for describing how farmers perceive some of IPM specialists' most important ideas and concepts. This scheme of Central American peasants' knowledge is also valuable in orienting outreach and participatory research and development efforts (figure 2). Figure 2. Style of Participation According to Class of Knowledge IMPORTANCE E 0 A B Scientists teach new ideas to Scientists learn from farmers S S + farmers, then learn from farmers E E who synthesize new information R with old o V F A T Farmers learn from scientists Scientists expand existing folk I - taxonomies, enhance farmer 0 observations, challenge existing N beliefs Important and Easy-to-Observe As this class includes farmers' most familiar topics, this is where scientists can learn the most from them. Rhoades' (1989) familiar example of diffused light potato storage falls into this class. Scientists learned about the technology from farmers in Kenya and successfully spread it to farmers around the world. Just a few of the other topics where plant protection specialists can learn much from traditional peoples include: * Intercropping, crop rotation schemes and other cropping practices * Behavior of other large social insects * Natural history, pharmaceutical, and nutritional value of "weeds" and other wild plants * Control of vertebrate pests. Farmers' knowledge should especially be relied on to set research agendas instead of allowing scientists' often esoteric disciplinary interests and unrealistic expectations to drive research. Unused maize-drying buildings in Honduras, abandoned water harvesters in Arizona's Papaguerfa (Bentley 1987) and failed, large-scale, capital-intensive irrigated rice schemes in West Africa are just some of the monuments to planners' and scientists' arrogance. 142 Not Important but Easy-to-Observe There are great opportunities for collaborative research on topics in this class that the scientist considers to be of importance. By teaching farmers things they do not know about certain easily observed organisms, farmers may gain an enhanced perception of some of the species and processes around them, and then learn more about them by continued observation. By changing peasants' attitudes regarding the importance of some easily observed phenomenon an important chain of creative innovations may be unleashed. Scientists can help shift farmers' notions of insect predation from the unimportant to the important side of the chart, by teaching them about it. Because earwigs, social wasps, ants, certain true bugs (Hemiptera), and praying mantises are easy to observe, if we let farmers know that these creatures help control crop pests, they can teach themselves how to conserve and manipulate these natural enemies. Farmers often gratuitously destroy wasps and ants to avoid being stung. The unimportant but easily observed class of knowledge is especially suited for extending precepts. Teaching farmers that ants eat insects gives people a reason to see ants in a new light, re-evaluate them as natural enemies, and then learn how to manipulate them. This appears to be the approach taken by the Asian rice IPM outreach program that highlights the role of spiders and other easily observed natural enemies as biocontrol agents. In Honduras we use bee, wasp, and ant reproduction as a starting place for discussing insect metamorphosis with farmers: explaining fly reproduction (which they partially understand) and moth and beetle reproduction (which they do not understand) in terms of hymenopteran reproduction (which they do understand). As farmers blend new information with old knowledge and new observations, they may create new, synthetic ideas and technologies, which scientists would not have invented. We experienced one such case in Honduras. One of us (Andrews) experimented with the predatory wasp, Polybia spp., moving hives onto maize fields, but was frequently stung and most of the wasp colonies soon absconded. We abandoned the idea in the early 1980s. Not long afterward Andrews explained wasp predation to a group of farmers, and one of the farmers, Wilfredo Flores, began moving nests. In 1989 we discovered that Flores was successfully moving nests on his own. Campesinos traditionally move nests from brush to avoid being stung while clearing land. They start learning about wasp relocation as children, bringing hives into rural schoolrooms and releasing them, hoping to terrorize the teacher and other students. This experience helped Flores overcome some of the problems that had discouraged our early attempts. Farmers who have become aware of insect predation through our biological control course have invented many new methods for stimulating the activity of predators in their fields. One farmer placed boxes of worm-infested potatoes on top of a fire ant (Solenopsis geminata) colony, removing the box just as the ants finished removing the insects but before they began to feed on the tubers. Another farmer set out raw sugar to attract ants and other predators into his field. One herded chickens through his cornfield to eat caterpillar pests. Others have hand collected ladybird beetles to release them in their fields. It appears in retrospect that the soil conservation and management programs carried out by our NGO colleagues fit initially in this cell. Soil erosion and depletion are easily observed. By helping farmers to see the importance of these phenomena, and by giving them a few prototype technologies, farmers move the subject to the conspicuous and important cell; they then concentrate their efforts to learn more on this subject. 143 Important but Dillcult-to-Observe This class represents the greatest challenge to scientists because it sometimes implies changing beliefs rather than simply adding new information. It is a heterogeneous class that permits at least three styles of intervention (expanding existing taxonomies, enhancing farmer observation, and challenging existing beliefs). Expanding Existing Taxonomies Some agronomists have ridiculed the campesinos' use of the word "ice" (hielo) for plant disease (Bentley 1991a). The agronomists mistakenly thought that farmers believed that their crop froze. However ice labels most plant diseases much like the English term "cold " labels a set of human ailments. Farmers do not believe that the plants actually freeze. Now many of our extensionists use the term. They explain to farmers "as you know, there are many kinds of ice" and then explain the different symptoms and causes of various diseases, essentially filling in a broad traditional category, ice, with a series of new subordinate categories: fungus, virus, bacteria, and so on. Associated with the new terms are many useful etiological and management concepts. Enhancing Farmer Observation If farmers are interested in a topic and lack the tools of observation to fully appreciate it, one tactic is to share novel methods of observation with them. In a study of maize ear rots, the major disease of maize in Honduras, we found that campesinos know virtually all that phytopathologists know, except for the causal agent (Bentley 1990; del Rfo 1990), so in over a dozen villages in the remote interior we set up a microscope and showed the fungus to campesinos, and we explained how this kind of fungus was like mushrooms that they were familiar with, but smaller. We then invited the campesinos to suggest possible control tactics. They proposed dozens of ideas, of which we eventually tested three for control of the disease: burning crop residues, bending the maize plant or removing leaves or tassel at physiological maturity, and trials of (native) maize varieties. As a result of the extension effort they carried out a series of novel epidemiological and control experiments. Challenging Existing Beliefs This may be more difficult and is risky. It requires great sensitivity. For example, many Honduran campesinos believe that agrochemicals spontaneously generate insect pests. They say that the first pests were seeded in chemical fertilizer so the people would be forced to buy insecticide, but each one they bought contained the seeds of yet another pest, trapping the farmers on a conspiratorial chemical treadmill. It is our belief that when farmers realize the true relations between pesticide and pest populations, they will be better able to wean themselves off agrochemical dependency. Farmers understand very well that physiological traits are inherited by the offspring of people, livestock, and crops -- and they readily grasp the idea of insects being selected for genetic resistance to pesticides. Farmers also accept the idea that natural enemies are killed by insecticides. They are fascinated by insect reproduction and can be helped to jettison the concept of spontaneous generation. We spend days carefully building a logical framework for changing the idea of spontaneous generation that is 144 nevertheless consistent in most respects with the local culture. At the beginning of our short courses, less than one-third (32 percent) of the participants understand that insects come from eggs while 81 percent know this at the end. We rely heavily on experiential learning. We show farmers caterpillars hatching from eggs. We watch mature caterpillars pupate and emerge as moths. Farmers learn that they can capture caterpillars and care for them in bottles until the adult emerges. By giving people tools for learning they continue to teach themselves after the course. Unimportant and DiJficult-to-Observe On the other hand, adding completely new concepts is easier. Although initially campesinos do not know about parasitic wasps, they enjoy the topic. (Only 7 percent of the farmers know of parasitoids at the outset of our biocontrol course, while 93 percent of them understand the concept and can identify these natural enemies at the end.) We use photographs and live parasites in bottles to expose campesinos to the subject. We also find it easy to introduce farmers to the notion of entomopathogens by analogy with humans: just as people get sick and sometimes die because of disease, so do insects. We show farmers cadavers of insects killed by disease. Farmers' knowledge of entomopathogens goes up to 79 percent from a mere 9 percent at the beginning of the course. This subject offers promise because of the growing importance of biological insecticides as alternatives to chemicals. Basic knowledge about disease may help farmers accept the biological control agents, even though they take days instead of minutes to kill pests. Concluding Remarks Human resource development is the key to the success of IPM; education, empowerment, and effective implementation go together. Our technical collaboration with smallholder farmers is based on learning what the people know and what they don't know, figuring out what they need to know, teaching it to them in a way that is consistent with what they know, and then learning from them as they synthesize new information with old knowledge. Our program includes a series of short courses that emphasize hands-on learning and introduction of new concepts and information that increase farmers' ability to innovate. The notion that natural enemies can be manipulated and conserved improves from 29 to 77 percent during our short course. Approximately 80 percent of the farmers who have participated in the courses have adopted or adapted at least one of the technologies included in the course. Fifty percent report that they no longer use insecticides. This acceptance of the biological control message is encouraging for its own sake and suggests that "farmer participation" can be more than rhetoric. We expect that further experience will help farmers to go beyond any all- or-nothing attitude toward pesticides to one that is more in line with the IPM idea that all tactics, if used correctly, can be important elements in a balanced program. Scientists can help farmers increase the rate of endogenous innovation. Training, if it is to be effective, should be based on focused, sophisticated social science. Anthropologists' involvement cannot be limited to post mortems or token visits. Pest management has to be recognized as a human enterprise. Men and women have to be seen as the complex centerpiece of all pest management activities. The entomologists, plant pathologists, ecologists, and their biological and agronomic scientist colleagues should play more supportive and less directive roles. 145 Extension must put as much emphasis on concepts as on facts and concrete technologies. The model for understanding the strengths and weaknesses of local knowledge presented here helps demonstrate the value of this approach. Almost half the participants interviewed four to six months after completing the biocontrol course have invented a new technique based on the new concepts presented in the course. IPM research should be site specific, responsive, and strongly influenced by the farmers. Given the severe human and financial limitations facing IPM today and the site specificity of pest problems, outreach takes at least temporary precedence over research. While research and extension are best viewed as concurrent and complementary activities, we feel strongly that there is a serious misallocation of resources in Central America today that results in unused research findings and farmers unable to enjoy the benefits afforded by the application of scientific ideas. Pesticide problems are only one of many motivations for carrying out an IPM program. Addressing pesticide abuse should be a component of almost any IPM program. However to treat IPM as a tool to correct pesticide abuse while ignoring rural families' production needs is inappropriate. A consortium approach that builds on the strengths of several very different private organizations creates valuable synergisms. Effective, focused public sector organizations can play important roles too. IPM is such a multifaceted undertaking and developing world institutions are so small and specialized that the union of their capacities is essential to the progress needed in IPM. Not all of our ideas could be developed in the space available, but are discussed in detail in the references given elsewhere in the text. We would enjoy corresponding with interested parties at any time. References Andrews, Keith L., and Jose R. Quezada (eds). 1989. "El Manejo Integrado de Plagas Agrfcolas: Estado Actual y Futuro." Tegucigalpa, Honduras: Zamorano Press. Bentley, Jeffery W. 1987. "Water Harvesting on the Papago Reservation: Experimental Agricultural Technology in the Guise of Development." Human Organization 46(2):141-146. . 1989. "What Farmers Don't Know Can't Help Them: The Strengths and Weaknesses of Indigenous Technical Knowledge in Honduras." Agriculture and Human Values 6(3,Summer):25-31. . 1990. "Conocimiento y Experimentos Espontaneos de Campesinos Hondurenios sobre el Mafz Muerto." Manejo Integrado de Plagas 17(September):16-26. . 1991a. ",Qu6 Es Hielo? Percepciones de los Campesinos Hondurenios sobre Enfermedades del Frijol y otros Cultivos." Interciencia 16(3):131-137. . 1991b. "The Epistemology of Plant Protection: Honduran Campesino Knowledge of Pests and Natural Enemies." In R.W. Gibson, and A. Sweetmore, eds., Proceedings of a Seminar on Crop Protection for Resource-Poor Farmers. Isle of Thorns, East Sussex, UK 4- 8 November, 1991. Chatham: Technical Centre for Agricultural and Rural Co-operation (CTA) and Natural Resources Institute (NRI). 146 . 1992a. "Learning about Biological Pest Control." ILEIA Newsletter 8(4):16-17. . 1992b. "Alternatives to Pesticides in Central America: Applied Studies of Local Knowledge." Culture and Agriculture 44:10-13. Bentley, Jeffery W., and Keith L. Andrews. 1991. "Pests, Peasants and Publications: Anthropological and Entomological Views of an Integrated Pest Management Program for Small-Scale Honduran Farmers." Human Organization 50(2): 113-124. Bentley, Jeffery W., and Werner Melara. 1991. "Experimenting with Honduran Farmer- Experimenters." ODA Agricultural Administration (Research and Extension) Network Newsletter 24(June):31-48. Bunch, Roland. 1982. Two Ears of Corn. Oklahoma City: World Neighbors. Cave, Ronald. 1992. "Center for Biological Control in Central America." Zamorano: Escuela Agrfcola Panamericana. (In English and Spanish). del Rfo, Luis E. 1990. "'Mafz Muerto' en Honduras Provocado por el Complejo Diplodia y Fusarium." Manejo Integrado de Plagas 18:42-53. Goodell, Grace, Keith L. Andrews, and Julio I. Lopez. 1990. "The Contributions of Agronomo- Anthropologists in Integrated Pest Management." Agricultural Systems 32:321-340. Hunn, Eugene S. 1977. Tzeltal Folk Zoology: The Classification of Discontinuities in Nature. New York: Academic Press. Rhoades, Robert. 1989. "The Role of Farmers in the Creation of Agricultural Technology." In Robert Chambers, Arnold Pacey, and Lori Ann Thrupp, eds., Farmer First: Farmer Innovation and Agricultural Research. London: Intermediate Technology Publications. Shaxson, Louise, and Jeffery W. Bentley. 1991. "Economic Factors Influencing the Choice of Pest Control Technology by Small-Scale Honduran Farmers." Chatham: Natural Resources Institute. Waage, Jeff. 1993. "Making IPM Work: Developing Country Experience and Prospects." In J.A. Srivastava, and Harold Alderman, eds., Agriculture and Environmental Challenges: Proceedings of the Thirteenth Agricultural Sector Symposium. Washington, D.C.: World Bank. WOMEN IN AGRICULTURAL RESOURCE MANAGEMENT I Raising the Productivity of Women Farmers in Sub-Saharan Africa Katrine A. Saito* Introduction Women farmers in Sub-Saharan Africa (SSA) now dominate the smallholder sector and account for more than three-quarters of the food produced in the region. Yet the economic, social, and cultural environment in which they work, rear their children, and manage their households is not supportive and, in some respects, actually hostile. Given the widespread food insecurity in SSA, this inhospitable environment results in large private and social costs. Governments and donors alike are increasingly realizing that one of the critical factors in revitalizing agriculture in Africa is to raise the productivity of women farmers. Increasing the productivity of women farmers will contribute directly to higher output and improved household food security. It is likely that the greatest gain from raising women farmers' productivity would come in the form of improved child nutrition, increased capacity for education and more generally, an enhancement of the welfare of rural households, an increasing number of which are female-headed. This paper summarizes the findings of a study, funded by the United Nations Development Programme (UNDP), of women farmers in SSA and the problems they face in raising their productivity.' It is based primarily on four country studies -- Burkina Faso, Kenya, Nigeria, and Zambia -- which document women's roles in agriculture, identify and evaluate the key constraints they face in attempting to raise their productivity, and recommend measures to relieve these constraints. All four country studies draw on fieldwork, specialized studies undertaken by local researchers, direct experience in World Bank agricultural operations, and extensive household surveys in Kenya and Nigeria, with 720 and 750 randomly selected rural households surveyed in each country, respectively. An important feature of these surveys was the collection of data on a plot- specific basis. Data was collected in this way because of the common practice in Africa of men and women managing their own plots. Based on the plot-specific information, farmer level observations were constructed by aggregating outputs and corresponding farm inputs on all his/her plots. Enumerators living in the villages undertook the fieldwork over the cropping season and sections of the questionnaire were administered four times corresponding to the main phases of agricultural activity. The country studies have also benefitted from discussion within the countries concerned. In Burkina Faso, for example, the draft report was reviewed and revised during a month-long process of discussion at the village level among farmers and government officials, culminating in a workshop with senior government officials. The Key Findings of the Study The African rural household is changing and traditional farming systems are breaking down. In response to evolving social and economic circumstances, particularly growing population pressure on an increasingly degraded land, men are migrating off the farm in search of more remunerative activities elsewhere. As a result, the traditional pattern of intrahousehold rights and obligations is * Senior Economist, AF5AG 148 changing. The gender-specific nature of African farming is disappearing as women are growing crops (such as coffee and other cash crops) and taking on tasks (such as land clearing) traditionally the responsibility of men, and making decisions on the daily management of the farm and household. Although men are generally still more involved than women in cash crop production, the Kenya and Nigeria surveys show that all crops are grown by men and women. In Kenya, for example, 33 percent of the women surveyed grew coffee compared to 26 percent of the men, and in Muranga District more female than male respondents had a plot of coffee and more women than men decided on the inputs and controlled the proceeds from coffee plots. These evolving circumstances have changed the role of women in African agriculture. In all four countries studied, smaliholders are the core of the agricultural sector, and women now comprise the majority of smallholder farmers. In Kenya, for example, 61 percent of women surveyed cited farming as their main occupation compared to only 24 percent of men. In Nigeria, farming was the main occupation of 88 percent of female and 58 percent of male household heads. The survey data also show the preponderance of female labor input in farming; women are engaged on a more regular basis than men in all farming activities and phases of the production cycle. Both men and women in the rural household make decisions on what to farm, how to farm it, and how to dispose of the proceeds, but these decisions are usually specific to the plot they manage and the revenue it yields. While some men and women do make certain decisions on each others' plots, essentially they manage their own separate plots (table 1). Table 1. Decisionmaking Responsibilities by Gender in Kenya and Nigeria (as percentage of decisions made) Head of Household' Male Female Farmer using plot Head Wife Head Kenya Nigeria Kenya Nigeria Kenya Nigeria Who decides what to plant? (703) (2213) (236) (789) (140) (436) - Husband 88 73 40 23 1 - - Wife/Woman 10 21 59 76 94 98 - Other 2 6 1 1 4 2 Who decides to use fertilizer? (252) (1080) (115) (258) (47) (164) - Husband 86 94 25 24 0 1 - Wife/Woman 12 5 66 71 94 97 - Other 3 1 1 5 6 2 Who decides to use improved seeds? (270) (456) (113) (110) (54) (78) - Husband 87 79 43 37 4 - - Wife/Woman 10 18 56 62 87 95 - Other 3 2 1 1 10 5 Who decides to sell the crop? (202) (1536) (55) (508) (41) (304) - Husband 93 68 25 21 0 - - Wife/Woman 6 25 75 66 93 98 - Other 1 7 0 13 7 3 Note: Actual sample sizes given in parentheses. They show responses for those plots where the decisions applied and were made. Defined as households headed by single, divorced, or widowed women, and households where the spouse has been absent for six months or more. Source: WAPIA survey. 149 Female-headed households (FIH) are becoming increasingly common in SSA. In Zambia, for example, they comprise about one-third of all rural households and up to 51 percent in the Northern Province. FHH are far from homogeneous, and include both autonomous households recognized and accepted as headed by women (mostly widows or single women), and households headed de facto by wives during the male head's absence. Although the practice of male outmigration is common, in some societies the wife who remains behind may not head the household. Despite heavy male outmigration in Burkina Faso, for example, FHH are uncommon because Burkina Faso's polygamous society absorbs these "lone" wives into the extended family. Within the family compound, however, wives manage their farming activities on their own plots. Female-headed households tend to be particularly disadvantaged as farmers (table 2). Land -- Landholdings of households headed by women are much smaller than those headed by men. In Nigeria, for example, the mean size of land farmed by FHH was one-third that of male-headed households (MHH). Labor -- FHH also tend to be smaller in size and have fewer farming adults than MHH. In the Northern Province of Zambia, for example, FHH had an average of 3.4 persons compared to 6.5 persons in MHH. The smaller number of farming adults in FHH point to a lower family labor supply than for MHH. In Oyo State in Nigeria, for example, the supply of family labor was much more restricted than in MHH, and women's use of hired labor, often paid for with remittance income, was insufficient to compensate. Table 2. Characteristics of Male- and Female-Headed Households in Kenya and Nigeria Kenya Nigeria Male Female Male Female (Sample size) (508) (199) (633) (117) Gender of head as % of all HH 72 28 84 16 Religion of HH head % Christian 72 84 56 85 % Muslim 10 12 40 11 Marital Status of HH head % married 95 67 99 19 % widowed, divorced 4 31 1 75 % single 1 2 - 6 Mean age (years) HH heads 53 47 50 47 All family members 24 22 23 22 % HH heads > 60 years 16 9 22 9 Education (no. of years) HH head 3.8 3.4 3.0 1.6 Children in HH 2.4 2.6 3.0 4.0 No. of individuals/HH 8.6 8.0 7.6 4.9 No. of individual farming/HH 2.3 1.7 2.1 1.5 Dependency ratio* 0.7 0.7 1.0 0.8 Note: Dependency ratio is the number of individuals aged under fourteen and over sixty divided by the total number of adults in the household. Source: WAPIA Survey. 150 Capital -- FHH are relatively undercapitalized. The survey data show that, in Kenya, for example, the total value of farming tools and equipment in FHH is less than half that in MHH, and 92 percent of female farmers used only hand tools compared to 72 percent of male farmers. Extension contact -- Women heading households also had much less contact with extension agents (only 4 percent of FHH were in contact with extension compared to 14 percent of MHH). Education -- Women heading households also had lower levels of education than men heading households, and in both Kenya and Nigeria, children of FHH had more years of schooling than those of MHH. Women work considerably longer hours than men on both agricultural and other tasks (table 3). The range of tasks on and off the farm that SSA women farmers, especially those heading households, are required to perform is very broad, and calls for an application of time and energy that tests human endurance. The four country studies show rural women working, on average, 50 percent more hours per day than men. There is a marked seasonal pattern in the type of work performed, with women's labor input exceeding men's in both the rainy and dry seasons. Table 3. Average Daily Hours in Agricultural and Nonagricultural Economic Activities by Gender Burkina Faso Kenya Nigeria Zambia Men Women Men Women Men Women Men Women Agriculture 7.0 8.3 4.3 6.2 7.0 9.0 6.4 7.6 Nonagriculture 1.7 6.0 3.8 6.1 1.5 5.0 0.8 4.6 Total 8.7 14.3 8.1 12.3 8.5 14.0 7.2 12.2 Source: Volume II Country Studies. Clearly, there is a finite limit to the time and energy that women farmers can apply. Given the already large and growing contribution women farmers are making to agricultural production in general, and to feeding their households in particular, policymakers must recognize that any strategy to improve agricultural productivity that increases the demand for labor, especially female labor, must take the consequent opportunity costs fully into account. Gender Differences in Agricultural Productivity Within a given agroecological environment, agricultural productivity is determined by the amounts of land, labor, capital, and other inputs that are used, and by the quality of these factors. Providing technologies and managerial skills are the same, farmers who have identical access to identical factors will produce identical outputs of a given crop. That is, their productivity will be identical. If they use different technologies or different quantities of these factors, their productivity will differ. Differences in the productivity of men and women farmers are likely. As noted above, men and 151 women in the African rural household pursue their own-account activities both on and off the farm and they also have different endowments (such as land rights and education), and different access to technologies, to factors of production (such as labor and capital), and to support services (such as extension and credit). Using data from the rural household surveys of Kenya and Nigeria, men's and women's productivity in agriculture and the relative contributions of different inputs and factors of production were examined. A Cobb-Douglas production function was used. The dependent variable was the gross value of output because intercropping is a common farming practice in both Kenya and Nigeria; both conventional and nonconventional inputs were used as explanatory variables. The main conventional inputs were land cultivated, family labor and hired labor disaggregated by gender, and capital stock; the main nonconventional inputs (represented by dummy variables) were an index of tenurial status (constructed by considering the extent of control a farmer has over a plot in terms of the ability to improve, sell, rent, mortgage, and lend the plot), use of insecticide, tractor use, gender of farmer, extension contact, and soil fertility. District dummies were also included to reflect different agroecological conditions. Education and age variables were represented in years. The results highlight the importance of female family labor for both male and female farmers: on male-managed plots in Kenya, for example, female family labor is the most important factor of production, with an elasticity of more than twice that for male family labor (table 4). Similarly on female-managed plots, female family labor is positive and significant with a relatively high elasticity. Another noticeable result is that hired female labor significantly and positively affects output on male- managed plots while male-hired labor significantly and positively affects output on female-managed plots. This cross-gender effect of hired labor may be explained by the difference in task -- hired male labor tends to be used mostly for land clearing while hired female labor is used mostly for weeding and harvesting. In other words unlike family labor, which shows little gender-specialization by task, hired labor tends to be used for the traditional tasks of men and women. Table 4. Land Holdings of Farmers Surveyed in Kenya and Nigeria by Gender (hectares) Kenya Nigeria M F M | F Households Bty gender of household head Mean total size of household holding 2.6 1.7 2.6 0.8 Total number of people per household 8.6 8.0 7.6 4.9 Hectares per person 0.3 0.21 0.34 0.16 Holdings By gender of land user Mean holding size of household head 3.1 1.3 2.2 0.7 Mean holding of household members 1.9 0.6 1.8 0.7 Source: WAPIA survey. Another important finding is that contact with extension did not increase the output of female farmers as it did for male farmers. Possible reasons for this difference include a lack of 152 complementary inputs, incomplete adoption, poor explanation, and hence understanding of technologies (women complained in both Nigeria and Kenya that the technology was "too technical"). The male-dominated research and extension staff frequently lack understanding of women's roles and constraints, and cultural norms hinder effective communication between male extension agents and female farmers. Furthermore, agricultural technical messages concentrate on the resources, commodities and tasks of more interest to men than women, while extension to women often revolves around home economics subjects. Finally, women's attendance at extension activities is constrained by their lack of time and mobility. In other words although the provision of extension services to women farmers has improved in recent years, women's contact with extension does not have the same impact on output as men's contact with extension. Women's disadvantaged access to factors of production and support services results in considerable loss in productivity. Simple comparative evidence from the Kenya survey found that men's gross value of output per hectare is 8 percent higher than woman's. However, from the econometric analysis it was found that if women had the same human capital endowments and used the same amounts of factors and inputs as men, the value of their output would increase by some 22 percent. Thus women are quite possibly better, more efficient, farm managers than men and their productivity is well below its potential. Capturing this potential productivity gain by improving the circumstances of women farmers would substantially increase food production in SSA thereby significantly reducing the level of food insecurity on the region. If these results from Kenya were to hold in SSA as a whole, and recalling that women produce an estimated 75 percent of the region's food, simply raising the productivity of women to the same level as men could increase total production by 10 to 15 percent. In addition to the production function analysis, influences on the adoption by men and women of selected agricultural technologies, such as the use of fertilizer and agrochemicals, were analyzed. These results highlighted the importance of extension contact in raising the probability of both men and women adopting such technologies, and also the positive effect of education, especially for women. For example, it was found that one year of education raised the probability of women using agrochemicals by 4 percent and of men by only 1 percent. The factors affecting the use of credit were also analyzed, and one especially interesting result was that women's use of credit was negatively affected by the distance to the financial institution, a highly plausible finding given the severe constraints on women's time and mobility. What This Means for Policymakers Governments are beginning to realize that raising agricultural output and productivity means a greater focus on women farmers. However, the pace of implementing the requisite supportive measures has been all too slow, resulting in considerable loss in potential output. As discussed above, potential agricultural output is reduced by as much as 22 percent because of women's disadvantaged access to inputs and support services. This potential productivity gain can only be realized by substantially improving women's access to inputs and support services such as land, labor, technology, extension, and credit. Some of the key measures can be summarized as follows: * Land Rights: Since the 1960s, some attempts have been made to improve women's rights to land, but in practical terms, the situation has frequently worsened: growing population pressure on increasingly depleted land has further weakened women's land rights, and as good agricultural land has become more scarce, women are managing even smaller plots. As pressure on the land increases and efforts to improve agricultural productivity intensify, it will be even more important to ensure 153 that women have access to and control over adequate land. Women's legal rights to land throughout Sub-Saharan Africa must be expanded and secured so that they can be implemented in practice. * Farm Size: Within the context of a growing shortage of good quality farming land in SSA, women are particularly disadvantaged compared to men in the size of plots they farm (table 5). However, given existing farming technologies, smallholders are faced with a situation where available family labor and insufficient income to hire labor constrains the productive use of additional land. Because smallholder technology is labor intensive, and because of acute seasonal labor shortages, more land, even if available, would not be a solution. Hence, smallholders, especially female, must gain access to more inputs and better technology so that the returns to the land they have is increased, in short, their productivity is raised. Table 5. Summary of Labor Use Per Hectare on Male- and Female-Managed Plots: Kenya and Nigeria Source of Labor Kenya Nigeria Male plots Female plots Male plots Female plots No. of % of No. of % of No. of % of No. of % of hours total hours total hours total hours total Family labor Male 524 32.7 486 23.2 709 34.9 767 27.6 Female 722 43.1 1,135 54.2 857 42.1 1,368 49.2 Child 82 5.1 114 5.4 204 10.0 364 13.1 Hired labor Male 128 8.0 160 7.6 105 5.2 144 5.2 Female 136 8.5 176 8.4 107 6.8 107 3.8 Child 8 0.6 24 1.1 28 1.0 28 1.0 Total 1,600 100.0 2,095 100.0 2,034 100.0 2,778 100.0 Hired as % of total 17.0 17.0 13.0 10.0 Source: WAPIA Survey. * Labor: The survey findings clearly show that women are the "work horses" of SSA agriculture. In both Kenya and Nigeria, women provide most of the family labor on plots they manage as well as on plots managed by men. Averaged over all plots, Kenyan women provide 84 percent more family labor than Kenyan men, while Nigerian women provide 33 percent more than Nigerian men. On a per hectare basis, the use of labor on women's plots is higher than on men's plots (31 percent more in Kenya and 37 percent more in Nigeria). All four countries face the paradoxical situation of a rural labor shortage within a labor-surplus economy with high population growth rates and high rates of unemployment. This has to do with the generally low level of labor productivity, reflecting smallholders, especially female smallholders, limited access to information and resources that would enable them to adopt different technologies, and increase labor productivity. With low average and marginal returns to labor, male family members in particular seek employment possibilities off the farm. That a high proportion of male heads and members of rural households are not engaged in farming is reflected in the findings of the 154 surveys. This reduces family labor supply and highlights the lack of cash or credit with which to hire labor. As a result, households adjust cropping patterns and farming systems to fit labor availability. They do this by limiting the area cultivated and planted, the amount of weeding or fertilizer applied, or by growing less labor-intensive crops such as cassava, and thus reducing labor value added. The solution lies in raising output by generating and employing superior technology. * Technology: Labor- and energy-saving technologies are women farmers' greatest needs. In addition they require production technologies for their commodities, constraints, and objectives, which are not always exactly the same as those of male farmers. No matter how technically feasible recommendations may be, they cannot increase productivity unless they are implemented. Certain technologies may be less easily adoptable by female than male farmers because, as the four country studies demonstrate, male and female farmers do not operate under the same conditions. If gender- related problems are allowed to constrain adoption, women farmers will be further disadvantaged and efforts to increase national agricultural output and productivity will be compromised. * Agricultural Research: To address these technology needs, agricultural research must focus more on the needs of the majority of farmers -- women -- by concentrating on the farming and household system, by increasing participatory research with male and particularly female farmers, and by improving feedback from gender sensitive extension agents and systems. Gender sensitive technology generation and promotion is possible. An understanding of women's farming roles and constraints, including cultural constraints, is a prerequisite to devising suitable strategies. Evidence from the country studies suggests that appropriate technology equipment for women farmers should be economically accessible and viable. In addition the necessary infrastructure and facilities should be available. Women should be included in the planning and trained in the operation of the technology, and the technology must be targeted at the person who will use it. * Extension: The production function analysis showed that contact with extension significantly and positively affects the output of male farmers, but not of female farmers. To address this problem, three strategies need to be pursued simultaneously. * More female farmers need to be contacted by agricultural extension agents. Female heads of households are most in need of extension contact because relatively few are in contact with extension agents and are less likely to receive information from close relatives. They also have considerable decisionmaking authority on whether or not to implement the advice. * The quality of communication must be improved: male agents must be sensitive to the needs of women farmers, and efforts should be made to increase the number of female agents (for example, by integrating retrained home economics agents into the extension service, as has been done in Nigeria). * The messages must be suitable for the objectives and constraints of women farmers. To do this requires better diagnosis of gender differences in agricultural activities and constraints. * Monitoring and evaluation should routinely be on a gender-disaggregated basis. The country studies show that there are many useful and practical examples of how to improve extension for women farmers in Africa2, but a more intensified effort is needed. * Credit: Women's access to formal credit must be increased. Cost-effective and sustainable financial services are critically needed by African smallholder farmers, both men and women. As the country studies show, they are presently quite inadequate, especially for women. In Nigeria, for example, only 5 and 11 percent of women surveyed had obtained credit from a bank and a cooperative, respectively, compared to 14 and 24 percent of men. Availability of inputs and technologies is to no avail unless farmers have the means to obtain or use them, and the seasonal surpluses of agricultural income may not be invested to full advantage. Financial innovations aimed 155 at providing such services in a sustainable way should be identified, particularly a greater effort is needed to explore and identify the informal savings and credit systems that are working for smallholders in Africa, together with ways of linking them to formal financial systems. Specific examples of such innovations are drawn from the Kenya and Burkina Faso studies. Conclusion Women are so important to African agriculture that initiatives to raise the sector's productivity cannot afford to ignore them. As this study shows, women do most of the work on the farm and increasingly have become the key decisionmakers. Despite this additional responsibility, however, women's access to agricultural inputs and support services has not improved commensurately. This results in a considerable loss in agricultural productivity and output, more than 20 percent according to the Kenyan analysis. The recommendations set out in this study are consistent with well- established tenets of agricultural development. Tenurial rights to land, land and labor productivity, cost-effective extension advice, appropriate technologies, and viable financial services are all important for effective agricultural development strategies. However, what this report emphasizes is that agricultural development strategies have not adequately focused on the clients, and in Sub- Saharan Africa at least, the clients increasingly are women. If SSA is to revitalize the agricultural sector and improve household food security -- goals assigned high priority by all countries in the region -- raising the productivity of women farmers must be made the centerpiece of agricultural strategy. Endnotes 1. For a more extensive presentation of this study, see "Raising the Productivity of Women Farmers in Sub-Saharan Africa," by Katrine A. Saito (with contributions by Hailu Mekonnen and Daphne Spurling), World Bank Discussion Paper (forthcoming). 2. A fuller presentation of the subject is given in Saito and Spurling "Developing Agricultural Extension for Women Farmers" World Bank Discussion Paper, No. 156, March 1992. I Agricultural Extension for Women - Experience from Nigeria Ndanusa B. Mijindadi* Introduction Agricultural extension aims at providing farmers the necessary education, skills, and technologies to enable them to make effective farm management decisions. However, in Nigeria and indeed in other African countries, the primary focus of agricultural extension efforts with respect to farm production traditionally has been on men. This is not surprising as African conventional wisdom is that men make the key farm production decisions - not women. As a result of the above reasoning, there had been a tendency to neglect women in the delivery of extension messages in the area of farm production. Instead, extension advise directed at women traditionally had concentrated entirely on home economics related topics -- home management, child care, family nutrition - all those aspects that have to do with the domestic roles of women. In recent years, however, the dominant role of women as farmers who make key farm management decisions has become increasingly clearer. That women play a dominant role in agricultural production is valid not only for those African countries where there have been an out- migration of men to other countries to seek employment (Lesotho, Malawi, Botswana); it is also valid for those countries where political unrests and civil wars have resulted in a large number of deaths of males (Ethiopia, Uganda). More importantly, it is equally true for several African countries in normal times. In Nigeria, for example, women involvement in farming ranges from situations where they assist in certain operations such as planting, weeding, harvesting (as in parts of the northern states) to situations in which certain crops are designated as "women-crops" -- especially food crops such as cocoyams and cassava -- for which women have exclusive production responsibilities (as in parts of the southeastern states). There are also areas in Nigeria where women farmers own and manage farms without any restriction on the crops they cultivate (as in the middle belt states). It needs to be mentioned that the recognition of the dominant role of Nigerian women in farm production and the need to modify the extension system to address the issue was brought effectively to the attention of Nigeria's agricultural authorities after a series of World Bank Study Missions to a number of loan-assisted projects in various parts of the country. These studies indicated that women were indeed responsible for as high as 70 percent of actual farm work, and in some cases constituted up to 60 percent of the farmers; while they receive little or no information from extension agents. This has led to the redesign of all World Bank loan-assisted Agricultural Development Projects to accommodate a Women In Agriculture (WIA) Program, which should ensure extension service support to women farmers. Thus, the aim of this presentation is first to highlight the objectives of the WIA program; second, describe its organizational structure and the implementation strategies employed; and third, summarize its major achievements as well as constraints. It is hoped that the lessons from Nigeria's experience in her efforts to provide extension services for women farmers could be of some use to other countries who may be in similar situations. * Head, Federal Agricultural Coordinating Unit, Ibadan, Nigeria. 158 The Agricultural Development Projects and Objectives of Women in Agriculture Program Before discussing the objectives of Nigeria's Women In Agriculture Program, it would be in order to briefly discuss the extension service system in Nigeria, otherwise called the Agricultural Development Projects (ADPs). The ADPs were started in the mid-1970s as enclave integrated agricultural development projects with funding assistance from the World Bank. The broad objective is to increase the production of both food and industrial crops. The core elements include a systematic extension program, adaptive research, input delivery system, a rural infrastructural program (rural feeder road, rural water supply), and an autonomous project management unit. The relative success of the first enclave projects encouraged the Nigerian Government to accept the ADP system as the main strategy for encouraging agricultural production at the small farmer level. Thus, the projects have been established on a statewide basis in all the thirty states of the country including the Federal Capital Territory of Abuja. At present the Training and Visit (T&V) system of extension is the strategy for providing extension service support to farmers within the ADPs. The main features of the T&V system are the fact that the number of operating farm families that can be effectively covered by a village extension agent is so assigned to him/her; the entire state is divided into zones, blocks, and cells for field level operation; and effective supervision is ensured. There are field visits and trainings conducted by Subject Matter Specialists (SMS) and other officials for extension staff and farmers. Extension officials are constantly kept up to date with research innovations through adaptive research activities and monthly technology review meetings (involving researchers from research institutes and universities). The training sessions provide opportunities for field extension staff to acquire knowledge on technologies required to meet identified field problems while at the same time providing researchers with a better grasp of field problems. Other features of the T&V extension approaches include farm trials on farmers' fields (which act as demonstrations to farmers); regular and fixed schedule of farm visits; and the use of the print and electronic media to enhance dissemination of farm innovations. The operations of the ADPs are undertaken through activities of the core subprograms of (a) technical services (covering adaptive research on crops, livestock, fisheries, and agroforestry); (b) extension services; (c) engineering services - feeder roads, rural water supplies; and (d) commercial services -- input supply, marketing, and credit. There are also support units -- administration, finance/accounts; planning, monitoring, and evaluation; personnel development and training. The WIA program, which may be part of the extension services or technical services subprogram, has the primary objective of increasing the productivity and incomes of women farmers. Specifically the program objectives include the following: (a) identify the constraints faced by women farmers; (b) source and where necessary collaborate with research institutions to develop suitable technologies to meet identified constraints and needs; (c) ensure timely extension support to women farmers in the area of agricultural production, processing and utilization (with greater emphasis on production); (d) provide advice to women on the formation of groups so that they can gain access to farm inputs and credit; (e) encourage diversification of women farming activities to small-scale production enterprises such as small ruminants, poultry, fisheries, and piggeries; and' (f) introduce labor saving technologies in the activities of women farmers. 159 Organization and Operational Procedures The extension services in every state in Nigeria now has women extension workers at all levels of its operation, from the headquarters at the state capital city to the villages. Thus at the headquarters there is the Head of the Women In Agriculture Program who has the rank of Deputy Director and works under the Director Extension Services or Director Technical Services. Head WIA has a university degree in agriculture, extension, or other related discipline with at least five years relevant working experience. She has overall responsibility for the planning and implementation of the program. She is assisted at the headquarters by a Subject Matter Specialist who may also act as Deputy Head of the program. At the zonal level, there are SMS (WIA) who supervise and monitor the implementation of the WIA programs in the zone. More specifically they liaise with research institutions to source relevant technologies, develop production messages, participate in field problem identification surveys and trainings, and provide overall support to Block Extension Agents (BEAs). At the Block level, there are WIA Block Extension Agents who are essentially Village Extension Agents and work mostly with women farmers. They are advised to spend about 70 percent of their time on agricultural production related matters and 30 percent on post harvest technology related problems. BEAs (WIA) have specific responsibilities to identify and organize women into groups in the eight cells in the block, and to liaise with cooperative society inspectors to register women groups into cooperative societies. The BEAs, however, report to the Block Extension Supervisors. At the circle or village level, there are no separate Extension Agents (EAs) for WIA. However, an understanding has been reached that at least 30 percent of all EAs in an ADP are expected to be female. In addition all female EAs are to ensure that at least 60 percent of their contact farmers are women farmers. Furthermore, all EAs (men and women) are to disseminate information to all farmers where no religious or customary barriers prevent such contacts. Implementation Procedures The operational procedures of the Training and Visit system of extension, which is currently in use in the ADPs, also provide the basic strategies for extension support to women farmers. The strategies include the use of the following: * Surveys or studies -- the objective here is to assess women farmers' priority needs, the constraints they face, and potential development opportunities in the area. Such studies are undertaken by resource persons from research institutes and universities working in collaboration with subject matter specialists in WIA. From the results of the surveys, needed technologies can be identified and problems requiring solutions passed to the appropriate agency. * Adaptive research -- while a number of agricultural technologies are gender neutral, it is important that technologies identified to meet women's needs are relevant to their circumstances. Thus, this strategy ensures that technologies directed at women are, where necessary, adapted specifically for use by women and made to suit the environment in which they operate. * Training -- this is undertaken at various levels. For example, at the monthly technology review meetings, resource persons from research institutions and universities in collaboration with Head WIA select and treat topics on technologies for 160 which the skills of subject matter specialists and BEA supervisors need updating. The topics and technologies treated are based on problems identified during studies and on results arising from adaptive research; other relevant technologies may also be treated. In addition, fortnightly trainings based on recommendations and technical messages ready to be passed on to farmers are organized for Extension Agents. Such recommendations are presented in simple to understand language using learning aids as necessary. The ADPs are also encouraged to establish skill development centers in various parts of the zones for the direct training of both field extension staff and women farmers on improved production technologies. At such trainings, practical demonstrations of the technologies are given. Finally, in passing messages to women farmers, Extension Agents may pay visits to individuals farmers, that is, direct face-to-face contact, or work through women's groups. Field visits are undertaken regularly by SMS (WIA) sometimes in the company of officials from the research institutes, Federal Agricultural Coordinating Unit, and the World Bank to supervise implementation of the program. Performance of the Program In the last three years emphasis has been placed on four activity areas (a) ensuring that the organizational structure for providing needed extension services to women farmers is in place; (b) getting block and village level extension workers to establish necessary contacts and rapport with women farmers; (c) sourcing and demonstrating improved technologies that attempt to solve the identified constraints faced by women farmers; and (d) establishing appropriate linkages between WIA program of the agricultural development projects and other women's programs being undertaken by other agencies. Viewed against the above areas of emphasis, some modest achievements have been recorded. Staffing With regards to staffing, table I indicates the following. All states' ADPs except one now have in place qualified Heads of WIA program. Subject Matter Specialists are also operating at the headquarter level for all but six of the thirty states. The requirement for Zonal Subject Matter Specialists has been met to some degree by all but four states where none was in place as of October 1992. At the critical block level only three states, Rivers, Taraba and Yobe, have less than 25 percent of the required BEAs. However, in the Rivers State the shortage of BEAs (WIA) should not be a matter of concern as there are no customary nor religious barriers preventing male extension workers from passing technical messages to women farmers. In the other two states, the staff shortage problem has arisen as a result of the recent creation of states. Taraba and Yobe States were carved out of the older states of Adamawa and Borno, both of which have acceptable degrees of staffing at the block level. 161 Table 1. Women in Agriculture Program in Nigeria: Staffing Situation in Agricultural Development Projects (1992) ADP Head WIA HQ SMS Zonal SMS BEAS In In In In _ Rq Post S Rq Post S Rq Post S Rq Post S Lagos 1 1 - 1 - 1 4 - 4 5 5 - Ogun 1 1 - 1 1 - 4 - 4 30 18 12 Oyo 1 1 - 1 1 - 4 - 4 34 27 7 Osun 1 1 - 1 1 - 3 3 - 29 19 - Ondo 1 1 1 1 - 5 5 - 37 24 13 Edo 1 1 - 1 1 - 2 2 - 23 8 15 Delta 1 1 - 1. 1 - 3 3 - 28 16 12 Kwara 1 1 - 1 1 - 4 4 - 23 14 9 Kogi 1 1 - 1 1 - 3 3 - 32 14 18 Abia 1 - 1 1 1 - 3 3 - 20 9 11 Akwalbom 1 1 - 1 1 - 6 6 - 54 7 47 Anambra 1 1 - 1 1 - 5 2 3 22 15 7 C/River 1 1 - 1 1 - 3 2 1 30 15 15 Enugu 1 - 1 1 - 3 1 2 24 20 4 Imo 1 1 - 1 1 - 3 3 - 45 17 28 Rivers 1 1 - 1 1 - 4 2 2 47 0 47 Benue 1 1 - 1 1 - 4 3 1 40 31 9 Plateau 1 1 - 1 1 - 4 4 - 47 47 - Taraba 1 1 - 1 - 1 4 4 - 40 7 33 Adamawa 1 1 - 1 1 - 4 3 1 47 46 1 Bauchi 1 1 - 1 1 - 4 4 - 55 16 39 Yobe 1 1 - 1 - 1 4 - 4 40 2 38 Bomo 1 1 - 1 1 - 5 5 - 61 55 6 Abuja 1 1 - 1 1 - 2 2 - 9 9 - Niger 1 1 - 1 1 - 3 3 - 37 37 - Kaduna 1 1 - 1 1 - 3 3 - 72 81 9 Kebbi 1 1 - 1 1 - 3 3 - 32 10 22 Kano 1 1 - 1 - 1 2 1 1 110 152 42 Jigawa 1 1 - 1 - 1 2 2 - 80 49 31 Katsina 1 1 - N/A 3 2 1 63 55 8 Sokoto 1 1 - 2 - 1 2 2 - 27 10 17 Note: Rq = Required; S = shortfall; N/A = not available. 162 Table 2. Summary of Selected WIA Activities in Nigeria (1991-92) ADP No of Women No. of Women No. of Demonstrations Groups Contact Farmers Lagos 66 NA 330 Ogun 123 1,543 269 Oyo/Osun 1,385 2,297 3,694 Ondo 155 936 783 Edo/Delta 197 911 1,149 Kwara 448 1,200 1,070 Kogi N/A l Benue 257 1,105 1,222 Plateau 81 3,098 1,465 Taraba N/A N/A N/A Adamawa 20 3,726 1,250 Bauchi 157 493 3,300 Yobe N/A NA l Bomo 43 4,321 3,690 Abuja 25 350 102 Niger 161 559 462 Kaduna 370 1,620 6,053 Kebbi N/A N/A N/A Kano 745 1,665 3,474 Jigawa N/A N/A N/A Katsina 20 N/A 15 Sokoto 128 1,520 29 Establishing Contacts The first column of table 2 provides some data on the number of women WIA groups which are in operation in the states, and for which extension service support is provided. As indicated earlier, women farmers are encouraged to form groups that may get registered as cooperative societies. Such WIA farmers' groups provide for a more rapid dissemination of agricultural innovations, and easier access to farm inputs and credit, which may be more difficult for individual farmers to obtain. Column 2 of the same table shows the number of women contact farmers that the Block and Village Extension Agents are working with in the various states. Technology Sourcing and Demonstration The constraint diagnostic studies and the monthly technology review meetings earlier discussed reveal the problems faced by women farmers. WIA Subject Matter Specialists respond to such needs by identifying improved technologies to resolve such constraints and enhance farm productivity. Such technologies are then demonstrated to women farmers at various times, for example, small plot adoption trials on farmers plots, and practical demonstrations at group meetings held in skill development centers or during individual contact sessions. Table 3 provides a list of the types of 163 technologies that were extended to women farmers in 1992 in the southeastern states of Nigeria. It is clear from the table that crops, livestock, fisheries and postharvest technologies were covered. The last column of table 2 shows the total number of small plot adoption trials (SPATS) and technology demonstrations that were undertaken for the benefit of women farmers in the states by BEA (WIA). Linkage with Other Relevant Organizations The WIA program has been able to establish close links with other women organizations in the state. This has helped to avoid duplication of efforts and encouraged support for one another's activities. For example, WIA Subject Matter Specialists and Block Extension Agents have been requested on a number of instances to provide technical trainings and demonstrate the use of improved technologies to other groups in the area of agricultural production, processing, and utilization. Similarly the WIA women's groups have received valuable technical assistance from organizations such as Nigeria's Better Life for Rural Women and donor agencies such as International Labor Organization, FAO/UNDP, and UNICEF. Constraints The above modest achievements not withstanding, both women extension officials and women farmers have faced a number of constraints in the implementation of the WIA program. With respect to Women Extension officials, the following problems have been most critical: * The number of women extension agents working at the village and block levels had their basic training and earlier job experiences in home economics. This has necessitated their retraining in the area of agricultural production, processing and utilization. This retraining exercise has not been as fast as desirable. * Subject Matter Specialists WIA and Block Extension Agents WIA are required to use group approaches in implementing their programs. Yet they have had little or no training in group formation and management to be able to give effective advise to the women's groups and cooperative societies being encouraged. * In a sizeable number of cases there has been instability in the staffing situation brought about by conflicts between commitment to work and family responsibilities on the part of women extension staff. For example, the marriage of female extension staff and transfers of husbands to duty posts outside cities where their wives work have tended to lead to resignations of WIA extension staff. Thus, in general, it has been very difficult to retain the services of female extension workers over long periods. On the part of women farmers the most recurrent problems are the following: * Shortage of farm implements that are gender specific or engineered with considerations of women's particular environment in mind. However, attempts are now made to involve women farmers in needs identification, selection of innovations, and in decisions on necessary adaptations to ensure relevance of the final product. * Extension officials have complained of delayed decisionmaking by women on whether or not to accept some technologies. This has been attributed to the need to clear such decisions with their husbands. This has mostly been in the southeastern states. In 164 Table 3. Technologies Extended to Farmers in Southeast Nigeria in 1992 TECHNOLOGIES NO. A Crops Yam/maize/cassava/egusi intercrops Alternative row Cassava/maize/egusi intercrop Yam miniset/maize intercrop Yam minisett sole Late maize/cassavalcowpea intercrop Cocoyam minisett Plantain sucker production Swamp and upland rice B Livestock Rabbit production - Brood selection - Hutch construction - Feeding and feeds - Kindling boxes - Handling of young rabbits C Fisheries Construction of fish ponds - Fish selection - Stocking methods - Feeding and Ponds - Cultural practices - Checking overflow - Checking weeds D Postharvest Processing of soybean - Soymilk, moi-moi, akara, pancakes - Fortification of soups, jollof rice etc. Cassava processing - Instant fufu, chips 165 order to reduce the time lag in such decisions, joint training sessions are now held for both male and female farmers on the same technologies. Similarly, collective decisionmaking at women's group meetings has tended to reduce the lag in adoption of innovations. * There have been reports that women farmers face difficulties when they decide to expand their cultivated landholdings. Women farmers are also hesitant to accept innovations that involve long gestation periods, for example, tree crops production, and soil conservation practices. The problem is that women farmers may only have use rights and not ownership rights for the land they operate. It is fair to add that local government authorities and chiefs have not hesitated to allocate unoccupied community lands to women farmers' groups for long-term agricultural production purposes, when they are approached. Lessons and Conclusions A few lessons can be drawn from Nigeria's experience with extension services for women farmers. First, there is the need to put in place an organizational system that extends to the village level and ensures a constant flow of innovations between the extension service and women farmers. Such a system must include a feedback mechanism such that the needs of women farmers are taken into consideration by Subject Matter Specialists and researchers. Second it is helpful to undertake studies aimed at identifying women farmers' specific needs. This should aid in setting priorities that satisfy the multiple roles that women farmers play, for example, their roles within the family, in the village economy, and in agriculture. Third, strong professional linkages with research institutions, universities and polytechnics, where technologies are developed, is helpful for the constant generation of new ideas. Thus, the involvement of resource persons from such institutions is an indispensable input for an effective WIA program. Fourth, before the transfer or dissemination of identified innovations, there is a need to test and where necessary adapt such innovations to the environment and conditions under which women farmers operate. This will ensure that such technologies are technically usable by women, low cost, profitable, and backed by locally available spare parts and repair facilities. A participatory approach, which involves the women in the selection and adaptation of technologies, has been found useful. Fifth, Nigeria's experience shows that encouraging the formation of groups by women and working with such groups helps enhance innovation dissemination and decisionmaking by women farmers. It also provides for easier access to farm inputs and credit. Sixth, following on the last point, Subject Matter Specialists and Extension Agents in WIA programs, require training in techniques of group formation and management, cooperatives, and credit institutions and their operation. This is in addition to regular training in agricultural production, processing, and utilization. Finally, Nigeria's experience shows that the linkage of the WIA program with other women's programs in the area, as well as with donor agencies who could provide support for the implementation of WIA activities is mutually beneficial to all parties. In conclusion, given the dominant role that women play in agriculture in developing countries, it is becoming clear that if such countries are to make a success of their agricultural development strategies, the specific needs of women farmers must be addressed. In particular extension service support must design programs that enhance farm productivity and incomes of women. Nigeria's experience in these respects indicates that if appropriate education and guidance are provided, women farmers are responsive and willing to adopt relevant innovations. 166 Acknowledgement This paper is based on field reports submitted by Federal Agricultural Coordinating Unit's Women In Agriculture program and Extension Specialists in the four regional offices at Enugu, Benin, Kaduna, and Jos. I wish to acknowledge the specific inputs of the following: J. Abdullahi; V.C. Agu; R.O.C. Chukwunta; C.Y. Akinboade, P.O.S. Abeywardena, and A.D. Onyia. Women in Agricultural Resource Management Aruna Bagchee' Women in Indian Agriculture While Indian agriculture, as a whole, has shown quite substantial gains especially since the mid- 1960s, the recognition of women's roles in this sector has come more slowly and more recently, only in the 1980s. There is, today, more awareness of the fact that farm women carry a significant part of the responsibility not only in crop production and homestead gardens, but also in tending livestock, poultry, and dairy. Not only is the contribution of female labor obvious and important in certain farm operations like the transplanting of paddy, or harvesting and postharvest management of most crops, their role as owners and decisionmakers of the farms is also significant. The latter is particularly true in hill farming situations, among tribal populations, and elsewhere too where the men try to get at least parttime, off-farm regular employment, or in cases of more or less permanent out- migration of the males. Even in the case of more apparently male-dominated agriculture (as in some states of the northern plains region), women, nevertheless, take responsibility for fuel collection and management of livestock, as well as a number of other allied activities, including market gardening. The fact that rural women are thus intimately involved in the management of agricultural resources has now come to be well accepted. However, it has to be admitted that this recognition of the role of farm women in Indian agriculture, has yet to make a significant impact in reorienting either agricultural research priorities or the developmental schemes. A determined effort is thus needed to translate this slow recognition of farm women's contributions to agriculture, into positive support and encouragement to their role in the further development of the country's agriculture. How exactly this positive support should be given is, however, still a matter of debate. One strategy attempted (through bilateral assistance from DANIDA) has been to implement separate women-farmer oriented projects. In these, female supervisory and field extension staff identify the production constraints of women farmers, arrange for short-term in-house training, and for follow-up extension services thereafter. But these are isolated microlevel projects without any large-scale visible impact or wider replicability. Another strategy, which has been considered, is to appoint more women contact farmers and village extension workers (VEWs) under the training and visit extension system. This would imply creating a parallel delivery system, which implies high costs, and in Kerala, where it has been tried, the impact is not very encouraging. A third approach sees the necessity of involving women farmers in mainstream development programs, such as the watershed development program, for example. This paper argues that involving women farmers in mainstream development programs is a sound strategy. Because the watershed development program is a major thrust area in the eighth five-year plan, a beginning can be made by strengthening women's roles in agricultural resource management through this program. Secretary of the Department of Tribal Development, Government of Maharashtra, Bombay. 168 The Watershed Development Programs Watershed development aims at restoring a denuded catchment area to a properly managed system of replanted forest cover in the upper reaches, of adequate pastures and range lands, and of measures for maximum harvesting of rainfall. It is thus an important means of increasing the productivity of rainfed farming systems. Given the extensive area under rainfed cultivation in the country, the Government of India has, in all its development plans, implemented various schemes for dryland agriculture. In the eighth five-year plan, 1990-95, particularly, it has embarked on a revised, more comprehensive program for improving rainfed farming systems on a watershed basis. The scheme's title is the National Watershed Development Programme for Rainfed Areas (NWDPRA). Besides the NWDPRA, there are operations research watersheds (ORP), sponsored by the India Council for Agricultural Research (ICAR) and agricultural universities, as well as several state level schemes. The watershed approach also is followed in some other important central schemes like the National River Valley Project, the Drought Prone Areas Programme, the Western Ghat Development Programme, and so forth. However in all of these schemes, the emphasis so far has been on developing the physical infrastructure of the watershed. Consequently, the programs largely consisted of traditional soil and water conservation measures, such as construction of check dams and diversion drains, contour and graded bunding, terracing, and so on. Much less effort has gone into training the farmers (both men and women) in on-farm scientific crop production systems (that is, in extension efforts to popularize in situ moisture conservation methods), and to encourage them to cooperatively manage the resources in a watershed on a sustainable basis. However, in the eighth plan period, some of these shortcomings are sought to be corrected by: * Placing greater emphasis on an integrated approach, which covers both arable and nonarable land treatment, as well as caters to the farmers' multiple needs for food, fodder, fuel, and income-generating activities. * Relying more on low-cost and vegetative conservation measures. Thus, in situ moisture conservation would replace the earlier dependence on more expensive, earth and stone masonry engineering structures for water harvesting. * Stimulating and promoting people's participation in project planning, implementation. and the management of community assets. For this, three contact farmers (Mitra Kisans) from the watershed villages are to be trained, and "care will be taken to select farm women as well" (Government of India 1990). These revisions are welcome, and give great scope to strengthen the role of women in managing the resource base of the watershed. However, this paper argues that merely giving women a quota representation on the beneficiaries' committee will not be enough, and would amount to only tokenism. Much more serious thought and efforts are required to meaningfully involve women farmers in watershed management. Strengthening the Role of Women in Watershed Management Discussed below are some of the key interventions that seem required from this point of view. These include (a) underscoring the need to develop the watershed as a whole; (b) paying as much attention to institutional development as to the physical restoration of the watershed; (c) measures to train particularly the women in the watershed through functional literacy subprojects; (d) introducing 169 farming systems research (FSR) subprojects that have a predetermined focus on the role of women; and (e) involving more voluntary agencies and documenting cases of success and failure at involving women in watershed management activities. An explanation of how these interventions would help women farmers follows. Underscoring the Need for Integrated Development of the Watershed as a Whole There is a need to reemphasize the concept of watershed as natural geohydrological units of planning in which land management practices at different gradients -- upland, midland, and lowland -- are interconnected. Because the watershed is, by definition, the unit of planning under this program, the need to develop the watershed area as a whole, through integrated planning, should be obvious. Unfortunately, in actual execution, this does not happen, partly because several agencies are responsible for implementing different components of the watershed development plan. Water harvesting and drainage management are the responsibility of the soil conservation department, the on-farm moisture conservation practices are to be explained to the farmers by the extension staff, and afforestation in the upper reaches is the task of the social forestry department. In the absence of coordination between these agencies, the linkages between land use in the upland, midland, and lowland regional, and that between individual farming units and the wider resource base of a watershed are not very well recognized. In the absence of such integrated development, many current watershed projects are reduced merely to the construction of a series of drainage management structures, called nala bunds, leaving on-farm cultivation practices and social forestry efforts untouched. However, this approach is bound to fail. Unless there is adequate tree and grass cover in the upper reaches of the watershed to "harvest" some of the rainwater and decrease the velocity of the runoff, there is little point in nala bunding farther downstream. Because individual landholdings are small and uneconomical in the rainfed areas, these farmers are dependent on the larger resource base of the watershed as a whole, which provides many of the free goods such as sisal for rope making, wood poles for housing, tools, and fuel, besides free grazing for the animals, and so on. There is, thus, an intimate linkage between the individual farm enterprise and the wider resource base of the watershed. Women are particularly concerned with such linkages, as many of these allied activities concern them specifically. For example, most women from smallholder families not only work at crop husbandry in private farmlands, they are also charged with responsibilities for the grazing of livestock in the common pastures, fodder and fuel wood collection from the community lands, and gathering nutrition supplements (tubers, roots, honey, fruit, and pods) and other free goods (sisal, bamboo) from the nearby forests and making them into marketable items like rope and woven baskets. They are, therefore, critically interested that the entire watershed be treated as a resource * base and as an integrated unit of planning. Such an approach thus would be directly in the interest of women in the watershed. Institutional Development Second, the current concept of watershed development tends to be limited to the physical restoration of a degraded environment. Thus, the components of watershed development are generally understood as planting more trees, constructing nala bunds, and developing pasture lands. However, a most critical ingredient of the farming system, namely human resource development (community 170 development), is completely ignored in such a concept. Yet, human resource development, including the development of appropriate institutions in the community, is the very cornerstone of scientific management of the resource base of a watershed. Unless people living in the watershed understand the linkages mentioned above, and unless they are motivated for collective action to manage the watershed for sustainable agriculture, government expenditure in afforestation or nala bunding per se is bound to be wasted, in the absence of maintenance and care, and proper institution s to manage the community asset. A few examples from the successful experiment of watershed development in village Ralegaonsiddhi will make the point clear (Bagchee and Bagchee 1990). All watershed projects include a component of afforestation and pasture development. But in most cases the experience is that grasslands are fences off and trees planted, at government cost, only to be encroached in by the local villagers. Only in Ralegaonsiddhi, because of local leadership, the newly forested commons are examples of "social fencing" -- the villagers have strictly followed the discipline of not letting in cattle for free grazing. Not only this, but when the Ralegaon villagers found that villagers from a neighboring village were stealing wood and grass form this area, they went as a group and apprehended the trespassers. Each was forced to pay a fine of Rs. 111/-; the sum collected was, however, returned to the neighboring village as a lump sum for its school. This gesture so impressed the villagers in that village that they too have started thinking of community action along similar lines. The reason that pasture development and afforestation works have succeeded in Ralegaon is that the economics of their management have been clearly worked out and a consensus built about the use and benefits to be shared from the assets so created. First of all, any alternate use that marginal croplands are converted to, such as pasture or grassland development, can be viable only in the context of the demand for fodder in the watershed, and the availability of foodgrains for the owners of these marginal croplands. Such lands are generally owned by the poorest cultivators, who grow minor millets (hulga, varai) in the kharif (monsoon season) for domestic consumption. They cannot forego this option merely because the soil conservation department declares these lands as unfit for crop cultivation, and recommends that they be developed as grasslands. In Ralegaonsiddhi the problem of the food security of the poorest villagers was first solved through a "grain bank" -- a buffer stock of cereals contributed by the better-off farmers. Thereafter, land-use planning according to the land's capability became feasible, as the smallholders who were cultivating food crops on marginal lands were willing to use this land for grass and fodder production. The activity of fodder development also became worthwhile as all the farmers came together to work out the demand for fodder within the watershed. There has to be a balance between the livestock maintained by the farmers in the watershed village and the fodder budget worked out for them. In Ralegaon when it was decided to take up such works, the villagers realized that if some of the common lands had to be temporarily closed for regeneration, the number of animals (then about 1,900) needed to be brought down to the carrying capacity of the available grazing grounds. Many of the small stock (goats) were sold off and some of the cattle were gifted away, until the number was reduced by half to 1,000. As the private marginal lands and the common lands started yielding grass, a fodder budget was worked out for each farmer. It was decided that the cultivators should, by and large, maintain as many heads of livestock as they could feed from their own grasslands and crop residues. The right to exploit the common grasslands was given only to the landless families who had no land of their own to maintain their livestock. Even in their case, there are no free rights: each family is charged Rs. 15 per month for the right to cut and carry grass from the common pasture lands and each one is allowed only one headload a day. This has brought in a modest income (Rs. 3,000 annually) to the Gram Panchayat (that is, the Village Council). A set of rules have been worked out that serve the interests of all parties involved. It is this consensus regarding the ground rules for sharing the benefits of the newly created community assets that explains why "social fencing" of these pastures and grasslands has been possible and successful here.It is these social or 171 institutional arrangements that are generally lacking in merely executing the physical targets of a development program. Innovated Functional Education (IFE) Subprojects As mentioned above institutional development -- encouraging men and women farmers to collectively manage the watershed resources -- has so far been a neglected aspect in the program. It needs to be emphasized and moreso in the case of women. In many villages, however, in order that women can come forward and participate in such community action, they need some special training. Therefore, an important component for strengthening the role of women in ongoing watershed projects would be to start innovated functional education subprojects for the women in these watersheds. Nongovernment organizations (NGOs) and other agencies (agricultural schools, Krishi Vigyan Kendras, and so forth) can be involved to organize innovated functional education subprojects for the women in the watersheds. The main thrust of the subprojects should be to evolve a need- based curriculum and a teaching model to train the women in the selected watersheds. They should have the following salient features: * The curriculum should be need-based and locality specific (in some watersheds, there may be scope for introducing scientific beekeeping, in others, tussor silkworm rearing, in still others, for improving vegetable gardening or floriculture). * The medium of instruction should be the local language, or even the specific dialect or subdialect that the women speak. * Skill training should be an integral part of the literacy component, and the literacy material should be developed around the skills to be imparted. * The skills and competencies taught should, besides being site-specific, be income generating so as to be meaningful to the participants. * Emphasis should be placed on teaching nontraditional skills, such as the use of the simple A-frame, biogas plant installation, and even operation of electrical and diesel pumpsets. The IFE subprojects can be run in clusters of twenty to thirty selected watersheds. Each subproject should start with a need assessment survey, development of site-specific curriculum, and literacy material as well as a plan for the skills training program. Farm System Research/Extension (FSR/E) Subprojects in Selected Watersheds Agricultural research that is oriented predominantly to varietal improvement and organized for specific commodities in watertight disciplines within university departments is not likely to serve the interests of women farmers, especially of women in smallholder families in rainfed areas. These women generally are engaged in mixed farming, on smallholdings and marginal holdings, and with many constraints of time, labor input, and cash. What is, therefore, essential is more field problem oriented research, that is moreover interdisciplinary, in that it looks at the whole farming system, rather than only at the main activity or dominant crop. There is a need for farming systems oriented agricultural research that has a predetermined focus on women farmers. There have been a few attempts in this direction, such as the Ramakrishna Mission work in the Sunderbans areas, and the Ford Foundation-sponsored Eastern India farming systems network, but these are not specifically oriented to developing the watershed as a resource base. It is expected that women's interests in agricultural development would be better served by having a farming systems research and extension subproject linked to the watershed development program. In order to give the FSR/E subproject a consciously predetermined focus to improve the productivity of women farmers, 172 gender analysis must precede the drawing up of the research agenda, as well as the monitoring and field testing of new technologies. This would involve: * Bringing together women farmers, extensionists, and scientists from the nearest agricultural universities, Krishi Vigyan Kendra (KKVKs), or National Agricultural Research Management (NARM) centers, to do a joint diagnostic survey. A useful tool in this exercise is a gender differentiated activities calendar that lists all the activities in which the local farmers are engaged over the year. To provide a more accurate picture of the entire range of enterprises involved and of the seasonal constraints of labor and other inputs, the activities calendar should include all production activities, not just major crops. It should also include: (a) domestic production or home maintenance activities such as collecting fuelwood and water, childcare, cooking, house construction and repair, and so on; (b) any gathering activity, such as collection of minor forest produce; (c) nonfarm production of goods and services through wage labor and other work; and (d) home processing of farm produce or collected goods made into marketable items like woven baskets (Feldstein and Poats 1990). * Such baseline data can then be used to design a research agenda, which gives high priority to addressing the identified constraints to increasing the productivity of women in the watershed. Unless this is done, quite frequently, gender analysis stops with diagnosis, that is, with charts showing men's and women's tasks. However, if there is an ongoing FSR project, with a predetermined focus on women, then gender analysis can form the basis for actual research work. It will be useful also in the field testing and evaluating of new technologies. For instance in activities like fertilization, foraging, pesticides, crop protection, postharvest home processing, and so forth, gender analysis would help in answering the key questions of who is affected and who must be taught techniques of application when new technologies are recommended. Are the new tasks labor saving or intensive? Are the opportunity costs correctly estimated according to the gender of who is doing the task? In particular, we need to commission special studies with respect to: (a) conducting time- allocation studies, in different agroclimatic regions, to determine how women's time is utilized before and after the completion of a watershed development project; (b) evaluating the extent of increase in the employment and income, which accrues to women as a consequence of watershed development activities; and (c) documenting the case studies of success and failure of women getting involved in watershed development activities. Involving NGO's and Documenting Case Studies Besides training and research support another crucial element is mobilization for collective action. As mentioned above, the term watershed development has to be correctly conceptualized, so that human resource development (HRD) forms an integral part of it. HRD includes grassroots level efforts to motivate and organize the local farmers for collective action to manage their resource base. In this, voluntary agencies can play a very important role. Indeed more than government, it is voluntary agencies who can effectively play the role of catalyst agent to motivate farmers for collective action. Unfortunately, very few of the voluntary agencies working in rural India have yet turned their attention to watershed management activities. Even fewer have worked specifically with the women in the watersheds. Yet, the few exceptions that exist have shown very dramatic successes in making the watershed development activities an effective measure for changing the pattern of opportunities available to the local farmers. It is, therefore, necessary to support the involvement of voluntary 173 agencies in watershed development activities, particularly those addressing the problems of rural women. Such support should be through (a) networking of the agencies already working in the field, (b) making available to these agencies published material and slides and films bringing out the interlinkages between all of the activities going on in the watershed, (c) encouraging them to more directly address the problems of women in the watershed, and (d) encouraging more voluntary agencies in the rural areas to focus on watersheds and particularly on the women in watersheds. The experience of NGOs so far has shown that considerable motivational and organizational work has to be done to involve women in watershed development activities. Despite their considerable contribution in almost all aspects of farming, rural women have been handicapped in participating in development schemes because they lack direct access to information or to government or bank officials. Yet, there have been some quite remarkable experiences where women have learned to play leadership roles, have even formed all-women cooperatives, and are in a position to make demands on the system. More needs to be known regarding these initial mobilization efforts, the dynamics of group formation, emergence of leadership, and the interaction of women's groups within the given system. Unfortunately, very little is documented about these efforts, even by those NGOs working in the field with women's issues. There is an urgent need for documentation of case studies that throw some light on these aspects. Indirectly and in the long run, these case studies of success or failure will help us in refining further efforts to strengthen the role of women in the management of agricultural resources. Conclusions In brief I have discussed that women are obviously and significantly engaged in agricultural resource management in India. At the same time it has to be recognized that rural India is a highly segmented society with clear-cut barriers of class status, hereditary caste occupations, and community and sex determined rules of interaction and tradition. All of these have important consequences for the organization of agricultural production, and these elements have to be considered while planning further agricultural growth. In such a situation to effectively reach the women farmers with a viable strategy to support their agricultural activities is not an easy task. It requires determined efforts matched by a lucid perspective and clear-cut goals. I have argued, further, that given this need, the strategies of reaching women in agriculture through separate microprojects or through a parallel extension and delivery system seem inadvisable on two counts. First, they do not appear to be cost-effective solutions. Second, they run the risk of marginalizing women's contributions, for the women's programs will, with certainty, be marginalized as far as both funding and staffing issues are concerned. On the other hand, I argue that it is advisable instead, to conceive of ways in which the mainstream development programs reach and address the production constraints of women, as well as men, engaged in the farming system. In this paper, I have outlined how the government's major development program for the rainfed areas -- the watershed development program -- needs to be reconceptualized in order to serve the interests of the women engaged in this sector. This includes: * Reemphasizing the need for integrated development of the catchment as a whole, so that individual farming units (particularly the women) can derive benefits from the wider resource base of a restored watershed. 174 * Stimulating and encouraging collective action at the village level to develop appropriate institutions (cooperatives, committees, whatever) for the management of common property resources developed in the restored watersheds. * Introducing functional literacy subprojects specifically developed