Established in 1971, the Consultative Group on International Agricultural Research- CGIAR-is an association of countries, international and regional organizations, and private foundations dedicated to supporting a system of agricultural research centers and programs around the world. The purpose of the research effort is to improve the quantity and quality of food production in developing countries. The World Bank, the Food and Agriculture Organization of the United Nations (FAO) and the United Nations Development Programme (UNDP) are cosponsors of this effort. The World Bank provides the CGIAR's chairman and secretariat. CGIAR is advised by a Technical Advisory Committee (TAG) whose secretariat is provided by the three cosponsors and located at FAO headquarters. CGIAR has 51 members, of which 39 were donors in 1986. Total contributions were about US$ZKS million. GGIAR-supported international agricultural research centers CIAT Centro International de Agricultura Tropical Cali, Colombia CIMMYT Centro International de MeIoramiento de Maiz y Trig0 El Batan, Mexmo CIP Centro International de la Papa Lima, Peru. IBPGR Internatronal Board for Plant Genetic Resources Rome, Italy ICARDA Internatronal Center for Agricultural Research m the Dry Areas Aleppo, Syria ICRISAT International Crops Research Institute for the Semr-Arid Tropics Hyderabad, India IFPRI International Food Policy Research Instrtute Washmgton, D C , United States. IITA International Instrtute of Tropical Agriculture Ibadan, Nigena ILCA International Livestock Center for Africa Addis Ababa, Ethiopia ILRAD International Laboratory for Research on Animal Diseases Nairobi, Kenya IRRI International Rice Research Institute Los Banes, Philippines ISNAR International Service for National Agricultural Research The Hague, Netherlands. WARDA West Africa Rice Development Association Monrovia, Liberia. ISSN 0257-3156 Pubhshed by the Secretanat of the Consultatwe Group on Internatronal Agricultural Research (CGIAR) 1818 H Street, N W , Washmgton, D C ,20433 United States, September 1987 Cltatlon. Annual Report-Consultative Group on International Agricultural Research. 1986-87 19864987 Annual Report Consultative Group on International Agricultural Research CGIAR Secretariat 1818 H Street, N.W. Washington, D.C. 20433 United States Foreword I write this foreword with reserve and not a little awe. As a result of the reorganization of the World Bank, I have been asked to assume respon- sibility for the CGIAR as part of the Bank's new Senior Vice Presidency for Policy, Planning and Research. As I believe that the responsibility for electing a new CGIAR chairman lies in a careful canvass of the views of the CGIAR members (a canvass that must take place at International Centers' Week in October), I write with the reserve of an interim chair- man. Having been associated with the founding of the CGIAR system, I find myself rather awed by the turn of events that has given me the privilege of writing a foreword to this significant publication. The honor of writing the foreword has fallen to me from S. Shahid Husain, who on June 1,1987 assumed the responsibility of the Bank's Vice President for Latin America and the Caribbean Region. This report covers the last of his stewardship of the Group, that is, the period from mid-1986 to mid-1987. The range and excellence of the CGIAR activi- ties that this report presents is an accurate reflection of the quality of leadership Mr. Husain brought to his exercise of the chair's functions. At this time of transition, it is useful to review the role and work of the past chairmen of the CGIAR system. Richard H. Demuth, as the first CGIAR chairman, and Sir John Crawford, chairman of TAC, helped create an atmosphere which persists to this day: A high standard of professional judgment, an informality of discussion with an emphasis on individual participation, an enthusiasm for the common enterprise. Warren Baum, who held the position of chairman for 10 years, saw the Group establish procedures consistent with major growth both in the number of centers and volume of funds. During his tenure, he led two reviews of the system. He is responsible for the present institutional shape of the CGIAR and for having established many of its traditions. Much of this experience is recorded in his book, Partners Against Hunger, which is a standard reference for those interested in knowing how the CGIAR came to be what it is. S. Shahid Husain had a large impact on the outlook of the CGIAR. He led the Group in substantive meetings and insisted that strategic concepts be defined clearly and that hard issues be faced. He recog- nized the need for special action related to Africa and for a clear focus on sustainability and the issues of resource management. He will be remembered in the CGIAR for progress in these and other areas. The next chairman will face a world circumstance that is very different from the early days of the CGIAR. In 1971, there was a specter of famine over many parts of the globe, and the need to raise the production of food in virtually all the world's developing countries was a major force in the formation of the Group. Today, in no small measure because of the extraordinary success of the research findings of the CGIAR centers, there is a global surplus of food (at prevailing market prices) coexisting with continued malnutrition of massive proportions, especially, but not exclusively, in Asia and sub-Saharan Africa. While fears of the massive famines of the 1960s have abated, there remains a nagging concern for the future. Population growth has not declined significantly and, in too many nations, growth in the need for food and better diets threatens to outstrip an increasing capacity to produce. The view of the present is one of abundance amidst want. The vision of the future is one of an uncertain capability to remain ahead of population growth while ensuring greater equity of access to all who are hungry. This view and this vision set two broad challenges for the CGIAR: 0 The first challenge is to discover how the CGIAR can best con- tribute to the enhancement of income and, through more income, the enhancement of food availability for poor people throughout the devel- oping world. The effort must be on understanding better the policies and programs that will increase the incomes and earning opportunities of the poor, especially the rural poor, thereby opening the markets for agricultural produce to their participation. If consumer subsidies or welfare payments are not to drain the fragile development budgets of developing nations, income enhancement must rest on finding and disseminating ways of improving the economic productivity of indi- viduals and families-a task that is both challenging and worthy of the CGIAR's affirmation. 0 The second challenge is how best to maintain the research drive to find and exploit new technologies of producing basic food materials for growing world demand. The goal is not new; it is the reason the CGIAR was formed. The challenge now is how best to effect its accom- plishment. The means we use to undertake the fundamental task of the CGIAR system must be examined in the light of the accelerating revolu- tion in biological research findings and methodologies that has taken place over the past two decades. The frontiers of applied science have advanced more broadly and rapidly than the CGIAR centers have been able to follow. This has implications for the efficiency of CGIARresearch, for the quality of the results, and, most importantly, for the centers' longer-term capacity to attract first-rate scientists to CGIAR endeavors. The third annual report shows that much progress is being made. New successes are reported; major research results have been obtained in such specialized fields as agro-ecological analysis and food policy, and the system remains the world's unique instrument for providing hope for a future free of nutritional want. One cannot doubt, in the light of the vitality, substance and shared purpose revealed in the text that follows, that the CGIAR and the centers it supports will continue to contribute to the conquest of world hunger and poverty. W. David Hopper Washington, D.C. Interim Chairman, September 1987 CGIAR Contents Foreword.................................. i 1. Agricultural research: Still a good investment? A Commentary by G. Edward Schuh. . . . . . . . . . . 1 2. Research: Fitting technology to the physical environment. . . . . . . . . . . . . . . . . . . 16 3. Policy: Influences in developing technology . . . . . . . . . . . . . . . . . . . . . . 36 4. Impact: From farmers' fields to national policies. . . . . . . . . . . . . . . . . . . . . . . . . 43 5. Key CGIAR events.. . . . . . . . . , . . . . . . . . . . . . 58 6. The financial situation . . . . . . . . . . . . . . . . . . 65 Annexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Annexes Page 1. About the CGIAR. 71 2. CGIAR major crops and activities. 73 3. CGIAR organization, May 1987. 74 4. Donor contributions to center programs, 1972-86 (in US$ million). 78 5. CGIAR-supported center expenditures, 1971-86 (current US$ million). 79 6. Regional origin of internationally recruited staff and board trustees, 1986. 80 Text boxes 2.1. An environmental catalogue: climate, soil, water, and vegetation. 19 2.2. Retrieval of agroecological data by CGIAR centers. 21 2.3. The CIAT land-systems study. 32 3.1. Labor and organic fertilizers in China. 40 3.2. On-farm client-oriented research. 42 4.1. Millets, a crop for marginal areas. 44 4.2. Breeding triticale for marginal environments. 46 4.3. ISNAR: Recommendation and response in Costa Rica to a diagnostic review. 53 4.4. Strengthening regional cooperation: the Central African potato network. 55 4.5. Incomes to Masai pastoralists from livestock production. 57 5.1. Meetings in 1986 through May 1987. 58 5.2. New publications about the CGIAR and its work. 61 5.3. Prizes and honors. 64 Text figures 2.1. Predicted yield for irrigated rice based on minimum temperatures and solar radiation during ripening. 29 2.2. Simulation of grain yield by two hypothetical wheat varieties at three locations in Syria. 34 4.1. Relative yield performance of triticales under differing agroclimatic conditions. 48 4.2. ISNAR's involvement with research systems in 32 countries, 1981-86. 52 6.1. CGIAR core funding, 1972-86. 65 6.2. Core expenditures, 1986. 67 6.3. Core research expenditures, 1986. 68 6.4. Core expenditures by region, 1986. 69 iv Text tables Page 2.1. Estimated worldwide harvested rice area by five major rice environments. 23 2.2. Estimated potential gains in rice production by ecology in West Africa over the next 20 years. 33 2.3. IRRI research efforts on major rice-growing environments, compared with anticipated economic returns from productivity increases in each area. 35 4.1. Potential triticale area in developing countries. 47 4.2. Documented impact of improved bean varieties from CIAT germplasm network, 1986. 49 4.3. Comparison of bean yield data and farmers' varietal preferences in Rwanda. 51 6.1. CGIAR funding, 1982-86. 66 6.2. Center operating expenditures in constant terms, 1982-86. 67 6.3. Core expenditures by program (percent), 1984-86. 68 6.4. Core expenditures by research commodity/activity [percent], 1984-86. 68 6.5. Core expenditures by region (percent), 1984-86. 68 6.6. Center core program expenditures, 1982-86. 69 6.7. Non-core program expenditures, 1984-86. 69 Editor's note: This report covers financial information in detail for calendar year 1986, the latest available. On other matters, the report deals with events through mid-1987. 1. Agricultural research: Still a good investment? A Commentary by G. Edward Schuhl From Malthus to surplus A little over a decade ago, the world was in the midst of one of its periodic Malthusian scares. Commodity prices were rising rapidly in U.S. dollar terms. National governments were scrambling to gain access to available supplies. Some public speakers in countries such as the United States were speaking of triage-the expected need to let some people die in order that others could live. Some speakers were urging Americans to eat one less hamburger a day to release the grain needed to feed starving people in faraway lands, not seeming to realize that hamburgers tend to be made from grass-fattened cattle. And predic- tions of doom were all around. Today, in the second half of the 1980s conditions could hardly be more different. If adjustments are made for inflation, commodity prices are as low as they have been since the Great Depression of the 1930s. Governments of the developed countries are spending huge sums to dispose of their surplus production. Developing countries find their export earnings shrinking as commodity prices decline. Farmers in the developed countries protest the World Bank's lending for agricultural projects and the purported harm done to their markets by bilateral foreign assistance to farmers in developing countries. And many observ- ers question investments in agricultural research, which in their view will only aggravate the current problems of low prices. I would like to address these issues in this essay. To anticipate my findings, the essay makes essentially four points: 0 Contrary to popular belief, agricultural production is not grow- ing at an accelerating pace in Third-World countries. 0 Economic growth is the key to expanding global markets, and a more productive agriculture worldwide is the key to obtaining that expanded growth. 0 Agricultural research has a long gestation period; turning the flow of resources for agricultural research off and on can be counterproductive. 0 The agenda for research has broadened significantly to include such things as greater efforts for maintenance of previous pro- ductivity, the need to address emerging environmental prob- lems, and the need to facilitate diversification and adjustment. `Schuh is director of the Agriculture and Rural Development Department of the World Bank. The views expressed herein are the author's alone and in no way should be construed as official views of the World Bank or the CGIAR. Perceptions about low commodity prices It is widely believed today that agricultural commodity prices are declining in large part because of accelerating production in develop- ing countries. The basis for this idea seems to rest on several percep- tions: (1) that agricultural output grew rapidly in China in response to policy reforms, and in the process China virtually withdrew from international markets; (2) that India, by means of Green Revolution technology, has some 30 million tons of grain in stock and is exporting modest amounts; (3) that some former grain-importing countries such as Indonesia (for rice) have become self-sufficient; and (4) that experi- mental results show that cereal yield potential has increased significantly, thus the world food problem has been virtually solved. These "facts" need to be put in perspective. The notion that agri- cultural production in the developing countries is growing at an accel- erating pace is incorrect. In fact, in the 1980s the trend of agricultural production in the developing countries as a whole is little different from the trend that prevailed in the 1970s. During the 197Os, the volume of production in developing market economies grew at an average annual rate of 3 percent; during 1980-85, at an annual average of 2.9 percent. Population growth rates in the developing countries have declined modestly since about 1975, and that has caused their agricultural production per capita to increase modestly. But that development has hardly been of sufficient significance to create a surplus in commodities. The important point is that growth in agricultural production in the developing countries is on the same trend line in the 1980s that it was in the 197Os, despite the Green Revolution and the higher yields that improved varieties brought with them. This steady growth proba- bly reflects the fact that production in many parts of the Third World is being pushed onto marginal lands. The benefits of past research may thus be doing little more than offsetting the lower returns from new areas being brought into production. The CGIAR's impact study indi- cated that the additional output that could be attributed to the CGIAR system was sufficient to feed 500 million people.' The true value of past investments in the CGIAR can perhaps best be realized by thinking about the consequences if such investments had not been made: World food prices would now be much higher, per capita incomes around the world would be much lower, and many more people would have died from starvation and the debilitating effects of malnutrition. `CGIAR Secretariat, 1985. Summary of International Agricultural Research Centers: A Study of Achievements and Potential. Washington, D.C.: Consulta- tive Group on International Agricultural Research. Turning to the perceptions about rising agricultural production in developing countries as the cause of depressed international commod- ity prices, it is true that China's agricultural output grew rapidly as a consequence of policy reforms and that its food imports declined, in part, as a consequence. Prior to the reforms, however, China's agricul- tural policy had induced wasteful use of the country's agricultural resources. Especially harmful were policies that forced local self- sufficiency at the commune level. The reforms not only led to a more efficient use of China's agricultural resources, they also freed China's farmers to respond to market forces. The result was an unprecedented increase in agricultural output. But will China's agricultural output continue to grow at its recent fast pace? Probably not. Shifting from an inefficient use of agricultural resources to a more efficient use is a cheap and fairly easy way to obtain growth in output. Once the use of resources becomes more efficient, however, production tends to increase at a rate consistent with the pre-reform trend. Moreover, more efficient resource use is accompanied by a significant increase in per capita incomes. Higher per capita income raises the demand for agricultural output. And at least part of that increase in demand involves greater consump- tion of more resource-using commodities such as livestock products, fruits, and vegetables. Rising demand might well have caused China to increase its imports of agricultural goods, especially feed grains, even though its domestic output was growing rapidly. However, China has faced a serious import constraint in recent years, due to trade barriers placed on its exports by other countries and sluggish economic growth worldwide, both of which have reduced its foreign exchange earnings. Thus it has not been able to import what it otherwise might have. As growth in agricultural production moves back to its longer-term trend, as per capita incomes continue to climb as a consequence of reforms in the rest of the econ- omy, and as import constraints are eased, it is likely that China will be back in the market for agricultural imports, possibly on a significant scale. India's recent agricultural history also has to be put in perspective. One reason India has accumulated large stocks of grains is that it has subsidized production, especially of wheat, by setting producer prices significantly above what would be market-driven levels. This is a costly way to expand output, and the stocks can be disposed of only at considerable additional cost to the government. It is doubtful whether India's fiscal resources can sustain such policies. In any case, this year's drought will probably cause these stocks to be drawn down significantly. Despite the accumulation of large cereal stocks, India has hun- dreds of millions of malnourished and underfed people. If poverty were reduced and the price of wheat in India permitted to decline to near- international levels, the expansion of domestic demand would absorb the so-called surplus production in short order. The fact that the poor in India live mostly in rural areas indicates the importance of agricultural modernization and development in reducing poverty. There is no argument with the statement that some countries are reaching food self-sufficiency. But one important reason is that they are undertaking reforms of policies that in the past discriminated against agriculture. Discrimination through economic policy was especially serious in the 197Os, when an abundance of liquidity in the interna- tional economy generated by the flood of petrodollars made it easy for developing countries to over-value their currencies. Overvaluation con- stituted a tax on domestic agriculture and a subsidy for imports. It is little wonder that imports by developing countries grew as rapidly as they did. The serious international debt problem of the 1980s has caused many countries to reverse these policies, and thus provide more incentives to agriculture. New production technology makes it possible for the incentives to have a stronger effect. Economic discrimination through overvalued currencies is not in the interest of the world as a whole. In fact, the sluggish growth of the 1980s is largely a consequence of trying to undo the effects of those unwise policies. However, with policy reforms, a basis is laid for sound economic expansion in the future. Increases in demand for agricultural commodities will be a consequence of increases in per capita incomes and population growth, not of distorted economic policies. Finally, evidence of increased cereal yields under experimental conditions is a weak basis for assuming that the world food problem is solved. In the first place, there is always a significant disparity between experimental yields and the average yield a nation's farmers obtain in their fields. This is because many farmers lag in their adoption of the new technology, and because the new technology may not be economi- cally profitable for all farmers, despite the fact that it increases yields under experimental conditions. A further complicating factor is that much new technology requires modern inputs such as fertilizers and pesticides to make it effective. These inputs may be unavailable to many farmers. The infrastructure -roads, communications, banks, lending agencies-and marketing arrangements may be too rudimentary to bring inputs to farmers and to accommodate increased output. What this means is that the translation of experimental yields into increased yields on farmers' fields is a difficult process, often requiring costly investments in rural infrastruc- ture and marketing arrangements, Genuine causes: Weak demand and policy distortions If the above "facts" do not explain today's low commodity prices, what does? Two major factors appear to be at work: weak demand and policy distortions in industrialized countries that have led to an export-subsidy war between the United States and the European Community. Weak demand is partly a consequence of sluggish global economic growth that has prevailed throughout the 198Os, with the exception of the rapid expansion of the U.S. economy in 1984. This problem has been complicated since 1982 by the international debt crisis. Develop- ing countries significantly increased agricultural imports in the 1970s when the international economy was awash in petrodollars. In the 198Os, liquidity dried up, and countries with serious debt problems have had to reduce their imports in order to generate a surplus in their balance of trade in order to service debt. Associated with this develop- ment, many developing countries have had to devalue their currencies to bring their external accounts into order. These devaluations have significantly reduced their import demand in the short term. As the international economy recovers, demand should once again increase. The chances for recovery in the relatively near future are good, since many large and severe adjustments have already been made. Policy reforms in the developing countries may reduce imports in the short run, but to the extent such reforms ultimately engender a more rapid rate of growth in countries undertaking them, demand will recover and at the same time be on a more solid economic base. Next, consider the issue of agricultural policies in the developed countries. The great distortions created by these policies are having serious consequences for international commodity markets. The Euro- pean Community, Japan and the United States all support substantial portions of their agricultural sectors by holding internal prices well above international levels. These high prices stimulate agricultural production. In each of these economic entities, the disparity between prices paid to farmers and those prevailing in international markets is quite great. For example, the price of rice in Japan is 10 times that for Thai rice. In the United States, the target price for wheat (the basis for defining payments by the government) is US$4.28 per bushel; the international market price is around US$2.50. The costs of these government subsidies are very large. In the United States, aid to farmers may reach US$27 billion in 1987. The 12 countries of the European Community spent US$23 billion in 1986 and Japan spent US$15 billion. Because of these policies, the European Community, for example, has shifted from being a net importer of agricultural commodities to being a net exporter-and by a wide margin. These distortionary policies have led to a costly export-subsidy war between the United States and the European Community. Govern- ments try to dispose of the large accumulated stocks by dumping them abroad. This is good for the consumers in importing countries, but it is devastating to producers in those countries. The important point, how- ever, is that the low commodity prices of recent years are not a conse- quence of accelerating production in the developing countries. They result largely from the dumping of agricultural produce by industrial- ized countries. Will these countries continue such policies into the future? There are significant pressures to reduce the high costs of these programs and to move to more liberal domestic and trade policies. Chances for liber- alization look better now than they have for some time. The current Uruguay Round of Multilateral Trade Negotiations provides a critical opportunity to move in this direction. Why agricultural research is important For developing countries, agriculture is the foundation of economic growth since the bulk of their resources is in agriculture. Moreover, these resources are characterized by very low levels of productivity. Simply put, agricultural research is vital because it is the source of new production technology, and new production technology is the source of economic growth. An important justification for continued investment in agricul- tural research is that such investments, if they are in well-managed and relevant research programs, generate large returns to society. Numer- ous studies consistently show rates of return that range between 25 and 100 percent a year in perpetuity-even studies that cover all of the agricultural research in a country, successful as well as unsuccessful ventures. No country that seeks to grow can ignore such high rates of return. Similarly no donor country that is seriously concerned about the welfare of those in the developing countries can fail to contribute to such investments. To appreciate the value of new production technology for agricul- ture, it is useful to consider it in a somewhat different way than is currently fashionable. Most people view new production technology as a source of expanded production. Hence, they tend to think of it in the context of production programs and to see its effects only in terms of supply or output. This focus on output also frequently causes research programs to be directed toward self-sufficiency goals. (Usually this goal is not a rational policy objective, nor is it consistent with promoting a liberal international trade regime.) New production technology does tend to increase agricultural output by raising the technical efficiency with which conventional resources of land and labor are used. But it is more fruitful and insightful to think of new production technology as a source of new income streams. The fact that the rate of return to investments in agricultural research tends to be so high is another way of saying that the increased streams of income that new technology creates come at a relatively low cost. To put it more formally, new production technology is a cheap source of economic growth, and policy-makers would be wise to invest their scarce development resources in such a cheap source of growth. These new income streams take on a variety of forms and can be found in several parts of the economy. For example, the early adopters of new production technology, such as improved varieties, receive increased incomes from their invested resources in land and labor. This is because output rises due to the gains in productivity, and in the short term the price of the commodity is unaffected. If the new technology is not widely adopted, or if the country is a net exporter but unimportant in international markets, this flow of increased income will continue into the future and be capitalized into higher land values and into the incomes of the entrepreneurs making the production decisions. Farm owner-operators tend to capture most of the benefits. Under these circumstances, demand for labor may also expand, so landless workers may experience increased employment or higher wages. If the gains in production lead to greater exports, then the increase in foreign exchange can finance a higher rate of growth for the economy as a whole or lead to a rise in the value of the nation's currency. In either case, the new technology creates expanded income streams throughout the economy. More commonly, however, the benefits of new production technol- ogy are passed to consumers in the form of lower prices. An increase in output tends to depress prices, and they fall to a level equivalent to the lower costs of production made possible by the new technology. This is probably the most pervasive way that the benefits of new technology are realized in an economy, but the expanded foreign exchange earn- ings it makes possible may still be significant. Consider now the sense in which lower prices for agricultural commodities constitute an important source of expanded income streams. Consumers can now purchase their food or other agricultural goods at less cost. In effect, their real income increases-an increase that will continue as long as the low prices prevail. This is the true "miracle" of investing in agricultural research as the basis of economic growth. The benefits of the new technology it produces are widely distributed in the economy. Moreover, they are distributed in favor of the poor in a relative sense, since the poor tend to spend a larger share of their budget on food. Few means of economic development spread their benefits in such a broad way, and so much in favor of the poor. The final advantage of thinking about new production technology as a cheap source of income streams, rather than as a source of expanded output, is that it underscores that the introduction of new production technology has both demand effects and supply effects, These demand effects can in fact outweigh the supply effects. Perhaps even more important, thinking about income streams rather than out- put conveys a different and more efficient set of research priorities, leading to faster economic growth. Is agricultural research threatening? Should producers in developed countries fear agricultural research in developing countries? The short answer to this question is, "In some cases yes, but in general no." A faster rate of agricultural modernization can make a country more competitive in international markets. But that problem can be reduced or eliminated by increasing the investment in research in countries that are threatened by the competition. In the final analysis, the spread of knowledge and new technology cannot be halted by political boundaries or policy actions. The only answer to such competition is to keep one's own productivity growing and to make efficient use of one's resources by reducing or eliminating distortionary policies. More compelling, however, is the fact that agricultural moderniza- tion, as a major source of income growth in developing countries, can fuel increased demand for agricultural output. The analysis can be summarized in four propositions: 0 Future foreign markets for the agricultural products of developed countries will be the developing countries, not the industrial- ized or centrally planned countries (China is included among the developing countries). Developing countries were an impor- tant locus of expanding markets in the 1970s. Agricultural imports by these countries grew as fast as imports by the centrally planned economies, and by the end of the decade the volumes were about equal. As discussed below, with economic advances in developing countries, their imports can be expected to resume the rapid pace of growth of the 1970s. 0 Developing countries will constitute a growing import market for agricultural commodities only if they experience significant eco- nomic development. The experience of the 197Os, when devel- oping countries' imports of wheat and coarse grains increased from 20.4 million to 58.6 million tons, provides an object lesson in this regard. Over 70 percent of the imports were by upper middle-income countries wherein rapid increases in per capita incomes were occurring. Poor countries, which exist in near- Malthusian conditions, simply do not have the means to pay for imports. o The development of agriculture in developing countries is the key to their economic growth. The bulk of developing countries' resources are in the agriculture sector, typically characterized by low productivity. Increasing productivity (and incomes) in the agriculture sector is the fulcrum for raising per capita incomes in the economy as a whole, and in the short run often the only means by which countries can earn the foreign exchange to further economic development. 0 Rising productivity in agriculture in the developing countries need not, as a general proposition, pose a competitive threat to producers in developed countries. In most developing countries, population is growing by 2 to 3 percent per year. Given their low level of per capita incomes, the income elasticity of demand for agricultural commodities in the aggregate tends to be much higher than in industrialized countries. Assuming this income elasticity is, plausibly, 0.6, the result of a modest 3 percent growth rate in per capita income, combined with a 2 percent population growth rate, would be a 3.8 percent growth rate in demand for agricultural output. Assuming a more optimistic growth rate of 5 percent in per capita income and a 3 percent population growth rate, the growth rate in demand for agricul- tural output would be 6 percent. Two points must be made about these growth calculations. First, obtaining sustained growth in agricultural output of even 3.5-4 percent per year is not easy. Few countries have done it in the past, except when there have been extensive new lands to bring into production, as in Brazil. Not many developing countries still have such stocks of land available. For those that do, the costs of bringing them into production, including the needed physical infrastructure, tend to be quite high. Evidence of this lack of available land is that in country after country population pressure is pushing cultivation onto lands that are at best marginally suited for agricultural production. Second, increases in per capita income on the order of 3-5 percent per year are not unusual in countries that have lagged in their develop- ment and thus can play catch-up by adopting technology from abroad. Japan, South Korea, and other newly industrialized countries, such as Brazil and Mexico, have all experienced higher rates of increase in per capita income for long periods as a result of this phenomenon, There is another feature of increases in per capita incomes in these countries that is important for developed countries. Rising per capita incomes lead not only to an upgrading in the quality of the diet, but to increased demand for commodities that require less time for household preparation. Both of these changes favor developed-country agricul- ture. The upgrading of diets means more rapid growth in the demand for poultry, livestock, and livestock products. This, in turn, implies an increase in the demand for feed grains, commodities for which the United States, a country whose producers have complained most about development assistance to developing countries, has an obvious com- parative advantage. Increased demand for commodities that need less household preparation involves a shift away from commodities such as rice towards commodities such as wheat. In the aggregate, this change Upland rice in Brazil is undergoing evaluation for favors the United States, the European Community, Canada, Australia tolerance to aluminum toxicity, a common soil prob- and Argentina. lem in Latin America, to find higher-yielding varieties. South Korea and Taiwan provide other striking examples. Returning to the relationship between general economic development and the emergence of import markets, in 1981 alone, South Korea bought US$Z.l billion in farm products from the United States, exceed- ing the total value of U.S. food aid to Korea between 1955 and 1979. Similarly, in the early 195Os, Taiwan exported more grain than it imported. Although Taiwan has increased food production very rap- idly since then, it now imports 60 percent of its cereals. Virtually all of these imports are feed grains, because of great demand for grain-fed livestock products. This analysis suggests that producers in developed countries should be lobbying to increase investments in agricultural research in the developing countries, not to reduce them. But these producers have some important counter-examples. It is useful to consider some of them. U.S. wheat farmers often view India as a lost market because of new technology used in its wheat production. The Green Revolution has been a key factor in reducing India's wheat imports, but most of the 10 imports replaced had been subsidized by U.S. taxpayers through such programs as P.L. 480. In addition, India is now importing more oilseeds than earlier. As noted above, moreover, India's problem is weak demand associated with low incomes, and its still high cost of produc- tion make it unlikely that it will become a strong competitor in interna- tional wheat markets. Brazil is another example of a developing country in which increased agricultural production, especially the rapid expansion of soybeans, is believed to have harmed U.S. agricultural exports. From 1970 to 1981, Brazil's agricultural production grew almost 70 percent, or 5 percent a year, one of the highest growth rates in the world. Although Brazil did emerge as a strong export competitor vis a vis U.S. soybean meal and oil, its imports of U.S. farm products as a whole increased substantially-by 15 percent per year in quantity and 25 percent per year in value. Brazil's imports of U.S. farm products also became almost 100 percent commercial (that is, private cash pur- chases) during the 1970-81 period, compared with earlier periods when 64 percent of imported farm products were subsidized by U.S. taxpayers through food aid. The case of Brazil also illustrates another important point, which is the significance of international specialization. If Brazil should, in fact, have a comparative advantage in soybean production relative to the United States (and that is not yet evident), it could well make economic sense for Brazil to specialize in soybeans and the United States in maize and other feed grains, with Brazil importing feed grains from the United States. Both countries could gain from such specializa- tion, even though specific groups of farmers in each country could be harmed. The record of the 1970s provides another example of such interna- tional specialization in production. Those countries which at the beginning of the decade were the largest exporters of agricultural com- modities were also the largest importers of such commodities. More- over, those whose agricultural exports grew the most during the decade were also the countries whose agricultural imports grew the most. That is what international specialization is all about, and the United States is a prime example of such specialization. Although it is the largest exporter of agricultural commodities, it also tends to be the second largest importer (and on occasion the largest). Making new production technology available to competitors is also viewed by some in developed countries as an unsound policy. The competition provided by wheat and maize producers in Argentina is often cited. But, the movement of new technology is a two-way street. Just as developed countries may lose a competitive edge from transfers of technology to developing countries, they also stand to benefit from improved technology produced by developing countries. This has 11 already occurred in the case of some semi-dwarf wheat and rice varie- ties. With a growing research capability in many developing countries, the potential for such gains is increasing. Thus, continued investment in agricultural research is vital, despite current low commodity prices, because the effect of better technology on demand can be equal to, or greater than, the effect on supply. Perhaps a more important point is that the best way to expand markets is to make the total economic pie greater, rather than to squab- ble over a fixed pie. Investing in agricultural research in the developing countries is the key to enlarging the economic pie. Other important issues Accumulation of knowledge. Man's struggle with the natural envi- ronment is interminable. No one knows when a new disease or pest will emerge that can wipe out the production of a nation's output, or have even wider effects. Some years ago the cotton industry in Brazil was virtually wiped out by an infestation of cotton wilt. An intensive research program, based on resistant lines from other countries and the advice of knowledgeable people outside Brazil, yielded locally adapted resistant varieties in a few years and revitalized the Brazilian cotton industry. The accumulation of knowledge is the only real defense against such episodic events. Sustaining gains. A significant proportion of agricultural research is maintenance research, that is, aimed at maintaining already achieved increases in yield. Much of the research on wheat rust, for example, is of this nature. As new strains of the fungus emerge, new resistant lines of wheat have to be identified and developed. More generally, as crop yields and productivity of livestock are pushed upward, the plants and animals become more susceptible to natural pests because of more intensive cultivation conditions. Significant resources are needed to deal with these problems. In fact, there is some evidence that as yield potentials are raised, the need for maintenance research grows and tends to take up a larger share of the total research budget. Finally, insects and diseases generally can become more resistant to pesticides over time. New pesticides and new means of control continually have to be developed. Environment. Environmental issues loom increasingly large on the agricultural scene worldwide. There are at least two important dimensions to this problem. First, the intensive use of fertilizers and pesticides can have important environmental effects. Research is needed to identify ways to reduce such environmental effects without sacrificing yield and productivity gains. Integrated pest management is one means, but research on this approach is still in its infancy. Another form of environmental damage or degradation occurs in countries where rapid population growth and limited labor absorption by the 12 non-farm sector forces farmers onto marginal lands. Cultivation of such An agroforestry nursery at ICRISAT's Sahelian Cen- lands can lead to permanent damage, or to damage that can be repaired ter in Niger is a testing ground for potential increases only over a lengthy period. And the effects may harm adjacent produc- in farming system productivity. tive areas as well. Research is needed to develop production systems that limit such environmental damage. And economic research is needed to identify policies that enhance the labor-absorptive capacity of the economy. Diversification. Research is seriously needed to facilitate the diver- sification process. As economic development proceeds and per capita incomes rise, consumers shift their preferences towards commodities associated with higher per capita income levels. Eventually, the pro- duction mix needs to shift in the same direction as the consumption mix. Research is needed to facilitate the adjustment of resources in these new directions without sacrificing productivity and income. 13 Another diversification problem arises when there is a technologi- cal breakthrough in a commodity that has low price and income elasticities of demand. As productivity rises in those cases? consumers tend to receive the major share of benefits. If resources do not shift from this activity to other agricultural activities or to non-farm activities at a sufficiently rapid rate, producers may actually lose as a consequence of the breakthrough. The important point is that unless such adjustments are facilitated, society's benefits from the original breakthrough will not be fully realized. Economic equity. Successful research efforts have implications for economic equity. One issue is that new production technology can significantly alter the distribution of income, both within the agricul- ture sector and the economy as a whole. Economic policy will have to deal with the problems associated with such redistribution, but agri- cultural research can also contribute by providing alternative produc- tion activities for those displaced by innovation and by changing the proportion of resources invested to offset such effects. New production technology also tends to favor regions that have high-quality land and good climate. The disparity in incomes between these regions and marginal areas will thus tend to grow. New produc- tion technologies are needed to assist the regions that are disadvan- taged, as well as economic policies to facilitate resource adjustment. Internationally, some countries will benefit, relative to others, from particular advances in research. The disadvantaged countries will need to make concerted efforts to catch up or to develop new produc- tion alternatives. Such changes on the international scene are always occurring. Strong agricultural research programs can help keep disparities from growing. Competence and capacity. Research is not an activity that can be turned off and on without serious consequences. The development of trained researchers is a slow process. And the development of new technologies takes time. The estimated time from the beginning of an agricultural research initiative until its benefit shows up as increased yields in farmers' fields is 7-10 years. If the research program is shut down, or even cut back, a response may not be forthcoming when a new challenge arises. A minor problem may thus become a disaster before a solution is at hand. Finally, all nations of the world can benefit from additions to the stock of knowledge. The key to economic development worldwide is to maintain the flow of new knowledge. The benefits are so great, as exemplified by the high rates of return to investments in agricultural research, that it is unlikely that the costs of adjusting to new technology will outweigh the benefits. 14 Expansion of CGIAR research is in order In the next three decades, the world will experience the most dramatic increases in demand for food and agricultural products in history. The only things that could keep that from occurring would be a nuclear war that destroyed a significant share of the world's population, an epi- demic that did the same thing, a sustained collapse of the international economy on the scale of the Great Depression of the 193Os, or some combination of them. Assuming that any of these catastrophies might occur is not a good basis for planning for the demands to be placed on the world's agricul- tural economy. That is why support for the global agricultural research system, including both the CGIAR system and national agricultural research systems, must continue, even though commodity prices are currently so low. For the CGIAR system in particular, both the scope and the scale of its programs need to be expanded. Introducing new production technology, the output of organized research programs, is an imperative if the developing countries are to experience increased and sustained rates of economic growth. The scope of the CGIAR system could productively be expanded to include a greater range of food crops, attention to the growing environmental problems around the world, making more effective use of natural resources, and the emerging problems of diversification and adjustment associated with the successes of its programs. There is also much to be gained from expanded work on cash crops, which are so important in generating the income and employment for the rapidly growing agricultural labor force around the world. 15 2. Research: Fitting technology to the physical environment Characterization of agricultural environments is a research undertak- ing of increasing importance to CGIAR centers. The knowledge gained through this process-often expressed as maps or computer models- promises to sharpen understanding of the different ecologies in which crops and livestock are and can be produced. It is shaping research programs, changing priorities, and in general, is a vital tool in devising more productive and stable farming systems. Moreover, the technology used is as sophisticated and demanding as any found at work in the CGIAR system. Achieving stable production Higher agricultural productivity on a stable and sustainable basis depends on identifying crops, livestock, and systems of production that are adapted to their environments-whether those environments are well or poorly endowed with resources. Finding the right fit of technology The ability of selected millet seeds to emerge through crusted red soil is being investigated by ICRISAT and environment requires detailed information about both. In the ini- scientists in India. tial years of the Green Revolution, scientists labored to create broadly adapted technology, mostly for high-potential areas. The early successes of CIMMYT with wheat and IRRI with rice were based not so much on varieties that performed well in different environments, but on ones that gave good returns to water and nitrogen fertilizer under rather similar conditions in different places. With the broadening of the commodity and geographical interests of the CGIAR system, it has become clear that environmental circum- stances are so diverse that greater sensitivity to variation is required, especially as agricultural production in marginal areas, frought with adverse climate and soils, is targeted for improvement. Plant breeders and other scientists must specify the range of agroecological conditions for which a particular variety is best suited. For harsher climates and fragile soils, specially adapted varieties must be developed. Greater attention to yield stability has been the major emphasis in activities of CGIAR centers during the 1980s. Broad adaptability and high-yield potential remain important goals, as does maintenance of yields through stronger and broader resistance to diseases and pests. But research increasingly turns to less-favored areas where the produc- tivity revolution has reached only a fraction of farmers. Significant gaps exist in describing environments with specific needs, especially hard- pressed marginal environments. Stable production in these environ- ments calls for stress-tolerant varieties, whose growth requirements must closely match temperature and available moisture. However, data on temperature, length of growing season, rainfall, and soil characteris- tics in marginal environments are lacking, thereby making it difficult to classify environments and to identify ecological stumbling blocks to increased production. 16 It may not be possible to moderate uncertainty and risk in all In Senegal, West Africa, pearl millet and sorghum marginal environments. In areas subject to large swings in weather, are intercropped among Acacia albido to enhance productivity and maintain soil fertility in a harsh extreme difficulties are customarily resolved by movement of people environment. and animals and by storage of water, food, and marketable valuables. In such environments, the concept of sustainable or stable yields may not be supportable, and there may be no alternative except to cease relying on them for production. Potential uses of characterizing agricultural environments Agroecological characterization aims at systematically describing a region's resource endowment and its potential use. The goal is to determine where crops and livestock can be grown, in what production context, with what inputs and constraints, and at what levels of yield or 17 productivity. The first task is to develop an inventory of environmental resources for agriculture and to understand how the resources are being used. The second task is to broaden knowledge of the potential uses of the resources and to suggest how the potentials might be achieved. To be cost-effective, the commodity programs of the CGIAR centers require systematized information about their target environments in three main phases of researc'h activity: formulation of research strategy, evaluation of technology and its potential for transfer, and improving the efficiency of international networks. Among its numerous potential uses! therefore: the CGIAR centers are chiefly interested in applying agroecological characterization to: @ Define broa.d research priorities and breeding strategies: @ Identify comparable environments for the transfer of technology; +B Improve the design of international trials and collaboratis:e research networks and the interpretation of results from them. The centers' special competence and comparative advantage lie in s-ustained research on commodities and production systems in devel- oping countries, and in analyzing and interpreting-across many loca- tions -the adaptation of crops and livestock to their environments. Standard terms and methods of observation and measurement are required. For crops, the tools encompas c eco-physical and analytical agronomy, including studies of phenology [crop growth patterns) and the development of plant structure over time, and the estimation of the so-tailed genetic coefficients for use in computer models of growth and yield. Compihg aspagssecsiogical inventory A vast quantity of data is needed to provide an integrated picture of diverse production environments: The data fall into four main groups: @ Edaphic factors-topography, terrain, soils, and hydrology. 8 Aerial factors--weather and climate. 0 Vegetation factors. Q I-Iumanfactors-land use, farming systems, societal organization, The data become more complex-and less stable--in that order. Except for catastrophic erosion soil tends to stay in place year after year. Although weather fluctuates, and over long periods may change, its broad characteristics are repeated year after year in most regions. Vegetation and other biological factors may differ radically within short distances and can undergo substantial change quickly. Land use and farming sh;stems are highly location-specific. They ma;; be altered suddenly and dramatically in the face of population increase, improved communications, or enhanced economic opportunity. The inherent instability and complexity of such data suggest that the compilation of an agroecological inventory is a formidable task, but major advances by international institutions have made global 18 19 Brazil's tropical rain forest, with high inherent fer- agroecological datasets conceivable. Among the notable developments tility, is being increasingly cleared for cultivation. are the unified description and mapping of soils such as has been done CIAT scientists are using land surveys to specify productive land-use options. in the FA~KINESCO Soil Map of the World, and the U.S. Department of Agriculture's Soil Taxonomy; standardized recording and communica- tion of weather information as promoted by the World Meteorological Organization; quantitative assessment of the potential output and population-supporting capacity of natural resources in FAO's Agro- Ecological Zones project and more recent developments stemming from it in several countries; remote sensing; and recent advances in assessing natural resource endowment with computers. At a CGIAR inter-center workshop on agricultural characteriza- tion, classification, and mapping, held at FAO in Rome in April 1986, participants called for the cooperative study of agricultural environ- ments by the centers, international institutions and developing country researchers. The participants recognized the need to: 0 Describe agricultural production environments in standard terms, based on common standards of observation and measurement; @ Compile accessible datasets on aerial, edaphic, biological, and human factors (although the centers are not yet prepared to attempt to incorporate human factors into the data sets) for each environment; and 0 Involve the CGIAR centers primarily, because of their special advantage, in the collection of data on food crops and livestock and their adaptation across environments. 20 21 They underlined, based on recognition that there is no single universal set of agroecological zones, that the overriding aim is to char- acterize agricultural environments, rather than to impose a classifica- tion system on them. Varied approaches to classifying environments The CGIAR centers' approaches to agroecological zoning are as differ- ent as their mandates. Centers whose research pivots on crops or livestock seek more detailed understanding of the regions that produce those commodities, as well as adjacent areas that might be affected by new technology for the commodity. Some centers, however, in addition to working on certain commodities, are responsible for areas that have special resource problems or potentials, e.g. the semi-arid tropics (ICRISAT), dry areas of West Asia and North Africa (ICARDA), the humid and subhumid tropics (IITA), certain zones of livestock produc- tion in Africa (ILCA), and the acid soils of Latin America (CIAT). These centers must deal with resource-base zoning problems, as well as zoning needs for their mandated commodities. Examples of the appli- cation of zoning by CIAT (see Box 2.3), IRRI, IITA and CIMMYT demon- strate how centers fit the approach to their goals. At CIAT, the highly variable tropical ecosystems in Latin America led researchers to gather extensive climatic, edaphic, and cropping systems data (on beans, cassava, pasture species, and rice) for Latin America in order to define land systems and microregions. Land sys- tems are areas with recurring patterns of climate, landscape, and soils. CIAT's analysis relied on satellite and radar imagery and occasionally aerial photography to provide a geographical base. Microregions are defined from data describing the growth and development of a particu- lar crop-sowing dates, harvest dates, varieties used, incidence of disease, pests, and weeds, areas sown, associated crops and cropping sequences, and yields. Microregions for different crops, or even the same crop in a different cropping system, can be superimposed on the land system map. As an example, CIAT scientists have defined environments for upland rice which is seeded under dry conditions and dependent on rainfall for moisture. One finding that arises from this example of agroecological zoning is that soil fertility is not necessarily a strong indicator of a favored upland ecosystem. Rather, climatic stresses, such as the probability of mid-season drought, are often more important in sorting out preferred environments. Classifications can be a useful tool to clarify thought in complex situations, but there can be limitations, as indicated by an example from CIAT's cassava program. In cassava, growth slows markedly below ZOO-21°C. While it would be informative to trace that isotherm using the standard climate mapping classification developed by Koppen, for 22 example, the demarcation of temperature regimes in that system is at 18°C and 24"C, making it unsatisfactory for cassava studies. IRRI-with a global mandate for rice-has organized its breeding program in relation to five rice-growing environments throughout the world. For favorable rice environments, the center's major aims are to raise and stabilize yields, raise the yields of irrigated rice, and develop earlier-maturing varieties so that cropping can be intensified. In less- favored environments, IRRI hopes to increase productivity and stabi- lize yields (Table 2.1). T&k 2.1. Xsfimated worldwide harvested rice area by five major aice environments. -~ __-.---~~--. __ Proportion [%) Earvested Global area Harvested rice Rice environment (million ha) area output Irrigated 78 53 73 Wet season 69 Dry season 9 Rainfed lowland 33 23 17 Drought-prone 13 Submergence-prone 4 Drought- and submergence-prone 3 No single severe constraint 6 Medium deep water (0.25-0.50 m] 7 Tidal wetlands 5 3 1* Deep water 11 8 5 Deep (0.5-1.0 m) 6 Very deep ( > 1.0 m) 5 Upland 19 13 5 Long growing season, fertile soils 2 Long growing season, infertile soils 7 Short growing season, fertile soils 5 Short growing season, infertile soils 5 Total 146 100 lOOk Source: IRRI *Ike to rounding. 23 IRRI is refining priorities of its rice breeding pro- For several Asian countries, IRRI has developed a specific classifi- gram, bCased on assessment of potential productivity cation based on the growing period for paddy rice and upland crops. In by rice environment. In Indonesia, a farmer in a govern= lent transmigration scheme is growing swamp Indonesia, paddy rice is the main food crop and agroclimatic zones rice in i Lchallenging environment. were determined by the amount of monthly rainfall and the number of consecutive wet months (at least 200 millimeters of rainfall] and length of dry period (number of consecutive months with less than 100 milli- meters a month). So far, 12 agroclimatic zones have been delineated on Java, Sulawesi, and Sumatra. 24 In Africa, IITA is concerned with diverse crop production environ- ments of the western and central regions. For IITA scientists, it is important to understand the range of environmental variation in a region and the effect of such variation on the needs for a specific technology in different areas. There have been numerous approaches to classifying the region agroecologically. The most suitable system depends on the questions being posed and the nature of the specific area being considered. In most classification schemes of the region, the general pattern is similar even when different variables are used, but there are numerous variations in the details of boundary locations and terminol- ogy. Generally, agroecological regions are classified either on the basis of vegetational or climatic variables, including various possible expres- sions of moisture balance. For West and Central Africa, a vegetational classification is often used because of the marked distinctions between the forest and savanna zones, though there are also numerous subzones within them and transitional areas between them. At CIMMYT, scientists have characterized the major production environments for wheat, triticale, and maize, starting with a general definition of macroenvironments for each crop. A more refined demar- cation of environments is evolving from country-by-country estimates of such variables as moisture stress and soil toxicities. Six sub- environments for wheat and 20 sub-environments for maize have been fCIATt,elmproveddietandproductivityofLatinAmeri- is introducing improved pasture species (left) ort . described so far. CIMMYT's most detailed compilation of environmental can livestock (right). 25 South Asia: agroclimate Number of Drymonths Wetmonths 11-12 0 9-10 0 7-8 0 !zzIl 6-10 1 -'=1 8-9 2 rII.zz 5-7 2 7 3 7-8 3 6-8 4 5-6 5 4-6 6 or more Dry month = < 100 mm. precipitation K'et month= > 200 mm. precipitation Peak precipitation at least one month > 500 mm. r-7 i-i-ii South Asia: the dry season Length of dry season (Month counted as dr when Thornthwaite's PE > 5 ) 12 dry months 10 or 11 dry months 9 dry months I I 8 dry months 7 dry months 6 dry months <6 dry months Total water deficit (Z of 12 monthly figures of Thornthwaite's PE - AE] > 1000 mm zzm 750-1000 mm tzz2 501- 750 mm ZZB k/&d,3 vt?p 251- 500 mm lOl- 250 mm LLL4 vm r----j Cl01 mm L -- -1 26 AFGHANISTAN -1. 2 South Asia: rice area planted by culture type 8A Y 01 A RAE/AN BENGAL SEA 1 dot = 3,000 hectares ,w $j Deepwater ( >lm.) : @j Irrigated (dry season] *,if w Irrigated (wet season] d m Intermediaterainfed(SO-100cm.) a Shallow rainfed (O-30 cm.) @j Dryland 0 `9% sa*:4,EoQwoar*cnL-4551 "0 IRK1 maps for South Asia capture complexities of agroecological data for evaluation: Rainfall pattern (at left, top), compared to definition of dry season [left, bottom) using evapotranspiration index and current area under rice production (above). The current areas under rice production indicate how the zoning picture is complicated by the presence of ground water (rivers) and the use of irrigation in otherwise low-potential environments. Editor's note: Maps do not reflect positions by IRRI, the CGIAR or its cosponsors regarding political boundaries. 27 At WARDA's experimental station in Mali, deep- data is for sub-Saharan Africa where three major agroecological water rices are being evaluated for traits leading to zones and nine sub-environments have been described. The relative higher productivity. uniformity and broad distribution of sub-environments globally sug- gests that germplasm is likely to perform well across comparable loca- tions. For example, high-performing maize germplasm at test sites in Central America is likely to do well in corresponding sub-environments in West Africa, provided that resistance to maize streak virus is added, or in Southeast Asia, provided that downy mildew resistance is incorporated. Matching cropping patterns with climate The design of biologically feasible cropping patterns is a process that matches the crop's physical characteristics (over their growth duration) to the area's physical conditions (over the year). Climatic information can be efficiently used for predicting land suitability for specific crops, different cropping patterns, and perform- ance of alternative patterns over time. Improved climatic analysis increases the likelihood of success for cropping patterns chosen for a given agroecologica1 base. It also helps simplify the description of experi- 28 mental sites, so researchers can concentrate on the physical parameters that really matter. The International Rice Testing Program, which includes IRRI, CIAT, IITA, and WARDA, has conducted a rice-weather nursery where a broad range of weather data are carefully recorded. This collaborative activity includes the World Meteorological Organization, the United Nations Development Programme, and the Netherlands. By allowing scientists to compare the performance of a standard set of varieties under care- fully measured weather conditions in many locations, these nurseries have illuminated the effect of weather variables on yield. This informa- tion has also been used to construct a model from which the perform- ance of an adequately fertilized irrigated rice crop in different conditions can be developed. Included in the rice-weather studies were efforts to predict rice yield on the basis of solar radiation and minimum tempera- tures (Fig.2.1). In association with several countries of the Sahel, ICRISAT uses rainfall information for marking off semi-arid areas, for delineating zones for the transfer of improved groundnut technology, and for defin- ing problems requiring interdisciplinary collaborative research. Based Fig. 2.1. Predicted yield for irrigated rice based QII minimum temperatures and solar radiation during ripening." Predicted yield (Ton/ha) 8 6 .I8 19 20 21 22 23 24 25 26 27 28 Minimum temperature ("C) Source: IRRUIRTP nurseries. "The shaded area for each radiation value is thego-percent confidence interval. 29 on an analysis of weather and production in the region, ICRISAT scientists have confirmed the link between falling groundnut produc- tion and declining rainfall over the past two decades. How to stabilize and restore yields in the face of a shorter growing season has been a specific problem for Sahelian farmers. ICRISAT's response is the devel- opment of earlier-maturing and drought-resistant groundnut varieties. ILCA scientists have been employing remote sensing over the past three years to predict drought and to determine the length of plant growing seasons and the production of plant biomass over large, often inaccessible, areas of Africa. Two ILCA teams have been involved-one in West Africa focusing on annual grasslands and the other in East Africa studying the more diverse perennial grasses. In Niger, ILCA is collaborating with the government, the U.S. Agency for International Development, and Tufts University (United States) to develop an early warning system for drought. The field work involves measurements of canopy reflectance and biomass, and is combined on the ground with low-level aerial photography (and reflectance characteristics). Devel- opment of precise estimates of grass production during seasons and over several years is hampered by the complexities of calibrating satel- lite information. But estimation of the length of growing periods from satellite data seems feasible, using techniques developed by ILCA in Ethiopia. The results could be useful for environmental monitoring of crop and biomass production. In Kenya, initial results of collaborative work involving ILCA and the United Nations Environment Programme suggest that satellite information can be directly used for predicting climatic disasters, but further research on calibration and correction for atmospheric interference is required. Such techniques have considera- ble promise as cost-effective and reliable tools for predicting and monitoring rangeland production. ICRISAT's computation of the length of the growing season for 160 locations in West Africa, using the world soils map and a review of existing criteria for climatic zonation in West Africa led, in 1986, to the significant research finding that two rainfed crops could be grown in sandy soils in parts of the Sahelian region. The date of the onset of rains is strongly correlated with the length of the growing season, according to recent findings, and tactical use of crop choices could enable farmers to be more productive under certain rainfall regimes. Efficiency of internatisnaall trials Once agroecological zones are delineated, the placement of research trials pertaining to specific environments can be planned more ration- ally. Results obtained in one environment can be used to predict the results in another that is similar or differs from it in known aspects. The vast network of multilocational testing trials used by the CGIAR centers and national programs would benefit from this efficiency. 30 The well-established international nurseries and yield trials ILCA's use of satellite imagery to estimate biomass conducted by the CGIAR centers in conjunction with collaborators in and vegetation in Africa is supplemented by on-ground national programs test the performance of varieties in different envi- surveys to confirm observations. ronments. The centers' current approach to international testing is to distribute trials widely at the request of cooperators. Repeated field trials over several years are costly. Also the data returned often vary greatly from location to location, and because the locations often change from year to year, it is not easy to draw firm conclusions about perform- ance and stability. If, however, locations for testing were more carefully grouped in well-defined environments, data on genotype-environment interactions would be more valuable to researchers. For national pro- grams, reliable maps of sub-environments would also help in identify- ing relevant field trials in other countries and in applying the findings in analogous environments within their own borders. IFPRI and the Australian Centre for International Agricultural Research have found in a recent study that there are substantial spill- over effects from regions where research is conducted to other regions with similar agroecologies and infrastructure. Research networks can enhance the mutual benefits to be gained from such spillover effects. 31 Box 2.3. The CIAT land-systems study. CIAT began land-resource survey were described as "land facets" and l soil-physical conditions affect- work in 1877 to gain better under- helped bridge the gap between land ingmoisture-holdingcapacity; I standing of the varied landscapes systems and soil units. 0 relative levels of soil fertility. for farming in tropical America. It Whenever information was avail- The picture that emerges from targeted the lowlands of South able, physical and chemical proper- the land-system evaluation, and by America east of the Andes-covering ties of topsoils (top 20 centimeters) inference, priorities in research on i 820 million hectares and extending and subSoils (21-50 centimeters genetic traits in pasture species, is 1 frompanama to southernBrazil-for depth) of the individual land facets considerably different from the one I agroecological zoning. A major objec- were recorded, tabulated, and previously inferred from generalized / tive was to narrow the gap between coded. small-scale maps. Phosphorus fixa- the actual and expected perform- Based on computer printouts of tion is not a potential problem over antes of improved varieties of tropi- land-system groupings that integrate much of the region, nor is alumi- cal crops (beans, cassava, rice, pasture the broad climatic, topographic, and num toxicity as widespread as species]. natural vegetation classes, five thought (though both are significant CIAT collected information on agroecological zones were selected in warm savannas). Thus, phosphate 1 climate (radiant energy received, tem- to define the humid lowlands of cen- rock and minimal lime applications perature, potential evapotranspira- tral tropical South America. Three may be low-cost solutions for a defi- 1 tion, water balance, other climatic of the `zones are savanna (poorly ciency in phosphorus or for toxicity factors], landscape features (land- drained, hot, and warm) and two- problems, respectively. Pasture plants form, hydrology, vegetation), and soil comprise forest (semi-evergreen and tolerant to high aluminum satura- (soil chemical and physical charac- tropical rain). tion are still desirable, but not all teristics). Data were then used to The five zones represent a first- pasture germplasm need be screened create a mosaic of zones that incor- approximation to -put gross climate for tolerance to high-level satura- 1 porated nuances in moisture and soil and landscape differences into per- tion, as has been the practice in the i regimes as well as vegetation. After spective. In order to t+r this broad past. However, potassium, calcium climatic analysis, the landscape was view into a framework for establish: and magnesium levels are low on a subdivided into land systems, which ing research Priorities and to define Iarge proportion of the soils, were delineated on 1:1,000,000 sat- conditions for selecting, testing and suggesting that pasture plant ellite imagery and, for some areas, transferring new pasture plantvari- germplasm that copes with such low on side-looking radar imagery. The eties, CIAT had to look-much more 1eveIs would be desirable. land surface was mapped to illus- closely at a number of factors, in&d- The study found substantial fer- trate areas sharing similar ecologi- ing for example: tility reserves in forest vegetation. cal characteristics. * rainfall patterns, which are In the tropical rainforest ecosystems, 1 Field work studied variations erratic in some areas: particularly, it suggests that the ' within land systems and helped 0 soil chemical conditions, search for pasture germplasm 1 : standardize descriptive criteria. specifically deficienciesin adapted to low soil fertility need ; These variations, although not phosphorus, potassium and not be a top priority. mapped because of scale limitations, calcium; 32 Careful site analysis also plays an important role in farming sys- tems research. Choice of research site, targeting of areas to be studied, and understanding of the farmers' environment all can be enhanced by improved environmental characterization. A real need in farming sys- tems research is better methods for using secondary data on climate, soils, and land capability. Priorities and strategies of individual centers Agroecological zoning efforts have helped the CGIAR centers refine their priorities. In West Africa, IITA, in ordering priorities for rice research over the next 20 years, defined the major ecologies and deter- mined their respective relative importance to future rice production. An important conclusion from this analysis is that the inland valley swamps, hydromorphic [poorly drained) environments and favorable upland areas account for almost two-thirds of the projected increase in production (Table 2.2). Table 2.2. Estimated potential gains in rice production by ecology in West Africa over the next 20 years. Yield (t/ha) Contribution Proportion of ecology to of total Projected increased Ecology area (%) Current Target gain production (%) Upland Unfavorable 10 0.5 0.5 0 2 Favorable 24 1.0 1.5 0.5 18 Hydromorphic 24 1.5 2.0 0.5 23 Lowland Inland valley 19 1.5 2.3 0.8 22 Mangrove swamp 8 2.0 2.8 0.8 11 Kiverine 9 1.5 2.3 0.8 10 Irrigated 6 3.0 4.0 1.0 14 Average 1.4 2.0 0.6 Source: IITA Simulating crop performance The application of modern methods of systems analysis, in combina- tion with recently developed techniques for simulating crop growth and yield, offers alternative and complementary approaches for evalu- ating and predicting the effects of environmental variation on opportu- nities for increased crop and livestock production. Descriptions of an environment's aerial and edaphic factors, while necessary, are not 33 Fig. 2.2, Simulation of grain yield by two hypothetical wheat varieties at three locations in Syria." Cumulative probability of yield sufficient for researchers seeking to simulate crop growth. Center scientists need to be able to associate-conceptually, and preferably 450 m abovesea level) quantitatively-variations in these factors with crop variations in struc- ture, growth cycle, and yield. Until modelers can develop coefficients to predict variations in different environments, they have to rely mainly on such measurements as weather and soils data, phenological observa- tions, and plant dry weight and harvest index (in cereals, ratio of grain weight to total plant weight). Genetic coefficients are factors that affect the accumulation of biomass over time and the time patterns of radiation received, tempera- ture, and water balance. Coefficients are used in models that can pre- dict performance in environments in which they have been validated and hence lessen the need for formal experiments. Modeling is particu- larly useful for marginal and unpredictable environments where many field experiments are wasted because the variance in experimental conditions is small relative to the variance across locations. IRRI, in collaboration with the Centre for Agro-Biological Research (the Netherlands), has organized training in crop modeling and sys- tems analysis for eight teams from seven Southeast Asian countries. It is also working with other countries in the region to set up small groups [soil scientist, agro-meteorologist, agronomist, often from different national agencies) to work on specific questions in agroecological characterization. In Syria, ICARDA scientists have developed a simulation model, SIMTAG (simulation of Triticum aestivum genotypes), to predict how climatic variability affects wheat yields. With climatic data supplied by the national meteorological service for three sites, the model compared the ability of two hypothetical varieties to provide high and stable yields. The hypothetical varieties were "early-big" [maturing early, with few but big kernels and a high kernel-filling rate) and "late-small" [late-maturing with many small kernels and a lower filling rate). Both were "grown" in a "good" (Luvisol) soil. The simulation addressed the problem of high seasonal variability in rainfall and yields. The ratio of rainfall between the driest and wettest season is 1:3, while the ratio in wheat yields for the respective rainfall conditions is 1:15. SIMTAG also .- Early-big provides day-to-day accounts of water balance and crop development > ______ La&malt as part of the analysis. Over 25 seasons, late-small was superior to 0.0 L early-big at two sites-at one because of higher rainfall and at the other, I I I I 1 1 2 3 4. 5 6. 7 which has a higher elevation, because of lower evaporative demand (Fig. 2.2). At the third site, which had low rainfall and high evaporative Saurce: ICARDA demand, the early-big variety was superior. The probability of grain "The lines reflect the relative ability of the variety to cone with climatic risk. The curve predominantly yield under environmental risk can help determine the most desirable on-right represents the "best-bet"-ifminimum risk germplasm traits for field testing. ICARDA scientists found, based on in variable climate is first nrioritv. The "late-small" the simulations, that six seasons of field results were not sufficient to variety (late maturing, small kernels) is less risky at Kamishly and Sweida, and "early-big" is preferable ensure correct interpretation of varietal characteristics most desirable at Muslimieh. for a specific location over the long run. 34 Looking ahead At the CGIAR centers, environmental priorities are shifting. This stock- taking and redefining of priorities has become possible because the gains achieved for major crops in favorable environments have enabled scientists to address a wider circle of agricultural environ- ments. But as a wider research net has been cast, new challenges - physical, biological and socioeconomic - have been encountered. Characterizing the environment in detail becomes vital as the centers confront two continuing challenges: safeguarding and stabilizing food production in line with the needs of an ever-increasing global popula- tion and contributing to improved productivity of cropping systems in less-favored environments. The job requires the collaboration of national and international agencies, some of which are not a part of the agricul- tural research system and yet whose work is essential to this endeavor. Based on agroecological characterization of production environ- ments, the sources and magnitude of increased food production can begin to be identified (Table 2.3). The result is an important tool for research scientists and policy-makers, as they grapple with priorities for deploying resources and strategies for food security. Table 2.3. IRRH research efbrts on major rice-growing environments, compared with anticipated economic returns from productivity increases in each area. Staff efforts Projected economic Current Projected benefits Rice environment WI (%I (%I Irrigated 51 42 58 Rainfed shallow 22 26 27 Upland 14 16 4 Deepwater and 12 15 10 tidal/adverse Total 100* 100* 100" Source: IRRI *Due to rounding. 35 3. Policy: Influences in developing technology Physical resources are the foundation of agriculture. Climate and soils set the range of crop and animal production opportunities available to farmers in specific localities. However, tying technology into an envi- ronment demands a broader perspective, because farmers' choices of what to grow and what techniques to use are guided by overlying economic, institutional and policy factors. More detailed agroecological characterization improves the ability of CGIAR centers to set priorities and target their research efforts towards high-potential payoffs. At the national level, the clear identifi- cation of research goals and the mobilization of emerging technologies require positive policies (avoiding distortions of comparative advan- tage) and effective institutions including research organization, input and credit channels, transport and marketing services. The importance of these factors to the global agricultural research system's ability to move advances in agricultural technology to poor farmers is illustrated here by examples of the CGIAR centers' research on policy and institutions. Comparative advantage One of the main methods used in policy research by IFPRI and other centers is domestic resource cost analysis, a technique which values production inputs and outputs, even those imported or exported, in terms of domestic resources. It looks across the physical, economic, policy, and institutional environments in which farmers operate to identify the comparative advantage of a country, or a region of a coun- try, in producing a particular product rather than another. The method highlights policies necessary to reconcile farmers' interests with the national interest. It often identifies existing policies on taxes, subsi- dies, or exchange rates that keep farmers from exploiting the compara- tive advantage of their physical resources and basic economic circumstances. IFPRI, in collaboration with IRRI, recently completed a study of the price and investment policies needed to accelerate the expansion of food-crop production in the Philippines. The study found that the Philippines' basic agricultural comparative advantage is to meet domes- tic demand for rice and maize (as food and animal feed) and to use these two crops as major determinants of economic growth through agricul- ture. According to the study, while maintaining stable rice and maize prices at somewhat above world-market price trends, the government should greatly expand investments in productivity-enhancing pro- grams, such as irrigation and technological development, to maintain long-term growth rates in rice yields and to accelerate the growth of maize yields. As productivity rises, price levels can be eased at a rate that balances increased consumer benefits with improving agricultural incomes. 36 Investments in irrigation should emphasize construction of new IFPRI policy study suggests investment in irrigation systems to bring irrigation to farmers dependent on rainfall, rather than schemes in the Philippines in which the community takes responsibility for organizing use of water. reinforcing the advantages of farmers who already have irrigation. Also, since the study found no significant regional differences in the compar- ative advantage of growing rice, investments in irrigation should stress the lower-productivity regions. Both actions would help make new irrigation systems instruments for reducing income disparities between rich and poor farmers. As part of the study, IFPRI examined competition between two key inputs, fertilizer and irrigation, for government budgetary resources. The study notes that "the massive subsidies that government would have to provide to lower the fertilizer price by 50 percent would be better spent on irrigation and other productivity-enhancing invest- ments." The study also evaluated the irrigation options for government investments. It recommended organizing new irrigation systems com- munally. Despite the fact that crop yields in communal schemes are lower than in other types, initial investment and operating costs are very much lower, giving them better and more consistent economic performance. Although the study identified new irrigation systems as the most efficient and equitable investment for government funds, it also exam- ined options for rehabilitating existing systems, to which significant 37 budget resources are already committed. The major problem in existing systems is distribution of water. Diversion irrigation schemes with no storage serve two-thirds of the irrigated area in the Philippines and usually operate by continuous and simultaneous water supply to the whole area under command. There are few difficulties when water is ample; but in years when rainfall is low, inequities occur in the sharing of water. On the whole, these absolute shortages cannot be remedied in a cost-effective way through rehabilitation or by the rotational manage- ment of water. In new irrigation systems, however, the analysis suggests that rotational management should be instituted to help avoid the income disparities that have arisen from uneven access to water where distribution is continuous. Domestic resource cost analysis has not, so far, been widely applied to the allocation of resources for agricultural research. Based on a series of case studies, done in collaboration with national agricultural research systems, CIMMYT's economics program is developing a manual on domestic resource cost analysis for use in assessing the national com- parative advantage of alternative research programs. The manual will help research managers identify discrepancies between farmers' inter- ests under current policy and under local comparative advantage which may be distorted by that policy. For example, a recent study, conducted in collaboration with INIA, the national agricultural research agency of Mexico, investigates the comparative advantage of irrigated and rainfed wheat in two areas of Mexico. Mexico faces increasing consumer demand for wheat products and the trade and financial situation favors expanding domestic wheat production rather than imports. Greater output in irrigated areas will have to come from yield increases unless wheat replaces other crops. This raises the question of whether policy-makers should give more attention to rainfed wheat production. The study analyzes the comparative advantage of wheat produced in the irrigated Yaqui Valley of the state of Sonora, the most important wheat-growing area in Mexico, and the rainfed highland area in the states of Tlaxcala and Hidalgo. It shows the substantial influence of government in setting both input and output prices for Mexican wheat. In general, producers have been receiving below world-market prices for wheat. This is particularly the case in Tlaxcala which is adjacent to major consuming centers and, hence, has markedly lower transporta- tion costs than alternative sources of supply. Policy has also intervened to change price relationships with other crops. In most cases, farmers receive higher-than-world prices for competing crops, especially maize and oilseeds. To some extent, government policy has compensated producers through subsidies on inputs. Subsidies on fertilizer, diesel fuel, credit, seed (in rainfed areas) and water (in irrigated areas) combined, exceeded 38 50 percent of production costs in 1979-82, for example. Since then, direct government subsidies have been reduced in relative terms, and credit and water remain the major sources of assistance. The high-level subsidies have encouraged the intensive use of inputs and incurred high costs in terms of national resources. The production of oilseeds and maize seems to have no compara- tive advantage over wheat in Sonora. For irrigated wheat, the major research opportunity is to find ways to reduce production costs. With government's commitment to reduce subsidies, there is a need to look for ways to use credit, water, fuel, and fertilizers more efficiently. In Tlaxcala, the study indicated that wheat production also has a compar- ative advantage relative to other crops, especially maize. This potential for wheat has not yet been realized, partly because pricing policies have generally not encouraged its production and partly because more pro- ductive wheat varieties have yet to be extended to local farmers. The analysis of the comparative advantage of wheat production in these two regions in Mexico suggests that rainfed wheat in the central highlands may be as competitive as irrigated wheat in the northwest. So far, very little research and extension have been devoted to wheat grown under rainfed conditions in Mexico. The case for allocating resources to the rainfed crop is strengthened by the CIMMYTiINIA study. Fertilizer policies: Bangladesh and China Bangladesh. In association with the Bangladesh Institute of Develop- ment Studies, IFPRI recently completed a study on fertilizer pricing policy and food-grain production strategy. In Bangladesh in the mid-197Os, the fertilizer subsidy absorbed up to 27 percent of the government's development budget for agriculture. Since then, the government has gradually raised the price of fertilizers to farmers. Even so, fertilizer use expanded so rapidly that the absolute amount of the subsidy was not much reduced, though the proportion of the agricultural budget absorbed fell to 13.6 percent by 1984. At that time, the government asked IFPRI to identify the price increases required to completely remove the fertilizer subsidy by 1988. IFPRI found the subsidies to vary considerably, depending on type of fertilizer and source of supply. The average economic subsidy was 17.7 percent for urea, 29.3 percent for triple superphosphate, and 24.4 percent for muriate of potash. The study projected that across-the- board price increases of 16 percent for two years would compensate for anticipated cost inflation and also eliminate subsidies. Further analysis by IFPRI indicated that these price increases would reduce fertilizer use by about 22 percent, causing a 2.2 percent drop in foodgrain output, which would be equivalent to 350,000 tons of rice. Recovery of the subsidy would also represent a 2 percent tax on 39 farm household incomes. In 1986, the Bangladesh government completely eliminated the subsidy on urea, though phosphate and potash are still partially subsidized. China. Completed in 1986, IFPRI's work on fertilizer in China analyzed the historical development of the fertilizer industry in order to identify factors that would help shape policies for meeting national needs in the 1990s. The study identified several distortions in past fertilizer policy and distribution that will have significant implications for the future. First, the study found that fertilizer use is concentrated in the rich agricultural areas, often close to urban markets. To a large extent, these better agricultural areas have captured the distribution system. To a significant degree, fertilizer has been obtained only in exchange of marketable crop surpluses, which provide food for the cities and raw materials for industry. Because marginal areas tend to produce close to the subsistence level, they cannot generate the surpluses necessary to barfer for fertilizer. Estimates for one province suggest that only 3 kilograms of nutrients per hectare were being applied in the less fertile areas, while application rates in the richer lands were up to 46 times as great. Other nationwide estimates indicate that by the late 1970s. the value of the paddy rice produced from a marginal application of fertil- izer had fallen to a ratio of 2.5-5 to 1 on fertile lands, compared with a ratio of 15-25 to 1 in areas of more depleted soils. A second observation is the imbalance in the proportions of nitro- gen, phosphorus and potassium (NPK) applied. The 1984 figures for domestic manufacture show a ratio lOON:ZlP:0.3K. The actual con- sumption ratios for 1983 have been estimated at 100:30:5, the balance improved by imports. However, in 1983 Chinese planners were indicat- ing a suitable ratio to be 10:50:20 or even 10:60:30. [The world con- sumption ratio is 10:52:40). The shortfalls in phosphates and potash indicate that on good agricultural land receiving high levels of nitro- gen, response rates may have been held down by the poor balance in the types applied, the nitrogen being unable to express itself in the absence of adequate phosphate and potash. The IFPRI research further indicates that the future of China's traditional use of organic fertilizer is in doubt. The use of organic fertilizer is highly labor-intensive with low rewards. Aspirations for higher incomes in the future are likely to encourage a shift from organic supplies to chemical fertilizers. The implications are far-reaching in terms of possible pollution and, more important, the development of soil structural problems known to follow the reduction of the organic matter content of the soil. 40 In China, efficiency in the use of fertilizer will be at a premium in the 1990s. Near Beijing, a farmer hand-meters fertilizer into irrigation water. The research concludes that without large public investments in new fertilizer plants or a firm commitment to import fertilizer, much of the growth in agricultural production for the 1990s will have to come from more efficient fertilizer use. IFPRI's analysis of the fertilizer indus- try suggests several routes to higher efficiency: l Improve the balance of nutrients used on high-yielding agricul- tural land. 0 Reallocate fertilizers from good to less-good agricultural areas to improve the average response of crops to fertilizer use at existing levels of supply. l Conduct localized research to determine more precisely the quan- tities and times of application appropriate for particular soil and water conditions. 0 Develop better placement methods and fertilizer formulations to ensure that crops recover more of the nitrogen applied. Plant uptake of nitrogen by plants varies with soil type; estimates of uptake in some areas are as low as 22 percent, which could probably be doubled. 0 Improve packaging, storage, and distribution. Losses due to vola- tilization before application may be 20 percent nationally and up to 40 percent in some areas. Responding to concern over slow rates of fertilizer uptake in coun- tries south of the Sahara, which are experiencing major food supply problems, IFPRI has shifted its focus increasingly to Africa. Its research emphasis is to seek policies that will increase fertilizer use by smallholders. Work began in Rwanda in 1986. 41 I / The institutional environment / Box 3.2. On-farm dient-orient&l j ISNAR addresses both the policy and the institutional environments / research. through its advisory service to national agricultural programs, which , An ISNAR study -identifies sever& inter- aims at improving their capacity to respond to farmers' needs and to related fun&ions which on-farm client- develop useful technologies; its new research program which is devel- oriented research should perform when fully oping concepts, tools and analytical methods in research policy, organ- ! integrated into a national agricultural ization, and management appropriate to developing countries; and its research system: training program. * Incorporate a problem-solving In 1986, it launched a study of the integration of on-farm research approach. 0 Contribnte to an-inter-disciplinary into national agricultural research systems-the first in a series of perspective. in-depth studies on critical management issues. Its purpose is to eluci- l Characterize major-farming systems date the critical organizational and managerial factors that allow for and client groups. effective integration of on-farm research into the broader research endeavor l Adapt existing technologies or con- so that it enhances research capacity and the generation and dissemina- tribute to the developmerit of alter- tion of technology. --nativetechnologieSfoftarge~ed~oups .I Studies were begun in Africa, Asia, and Latin America of eight of farmers, -, national agricultural research systems that have had sufficient time to 0 Promote farmer participation in experiment with, and develop, diverse organizational arrangements research as collabdrators, experi- and management systems for implementing and integrating on-farm menters, testers, and evaluators of client-oriented research. The case studies are expected to be completed alternative technologies. by mid-1987. l Provide information for- setting- research pr;oritiies: planning, and- These examples of CGIAR research emphasize the importance of programming. policy and institutional environments in making good use of the under- * Promote collaboration With &tension lying physical environment of climate and soils. They demonstrate that and development agencies to make appropriate policies and effective institutions are a prerequisite to the the generation and diffusion of tech- mobilization of CGIAR-related technologies by partners in national nology m&e efficient. programs. 42 4. Impact: From farmers' fields to national policies Individual CGIAR centers continue to gather and analyze information about the impacts of technical change as one important aid in charting the course of future research. The examples of impact presented here are chosen to illustrate two important facets of the evolving role of the centers in the global agricultural research system. 0 The impact of CGIAR activities and products will be apparent more at intermediate levels of the global agricultural research system rather than directly in farmers' fields where national institutions play a central role. 0 The assessment of potential impact will be increasingly vital to the CGIAR centers in choosing research priorities. The closeness of national agricultural research systems to resource- poor farmers makes national systems the foundations of the global system. Improving national capacity includes increasing their ability to signal the real needs of their client farmers to the CGIAR. Both these components of the global system need strong competence in assessing the potential impact of technologies as an aid to more efficient priority- setting. It is widely acknowledged that "miracle" technologies are few and far between. The mainstay of agricultural research is the purpose- ful anticipation, targeting and solution of high-priority problems that affect large numbers of poor farmers. In this regard, CGIAR efforts on agroecological characterization are an important step towards better estimation of the potential impact of research activities. For example at ILRAD, which focuses on developing methods of immunization for trypanosomiasis and East Coast Fever in African cattle, the centers' new social science unit will try to anticipate the ramifications of successful development of such methods on the economy and ecosystem, already struggling with dramatic population increase. Prior assessment of the impact of immunization can help to ensure that innovations in animal health are of real benefit and do not add to the socioeconomic or ecological strains on Africa. Several of the following examples demonstrate the essential part- nership between national research systems and the CGIAR centers in improving agricultural productivity. The national systems increas- ingly take responsibility for the use, adaptation and refinement of products coming from the CGIAR centers to meet the specific needs of their farmers and institutions. Other examples demonstrate the research investment in time and resources required before fruition. 43 Managing Vertisol soils Burgeoning populations have heightened the interest of the develop- ment community in the management and sustainability of resources. Not only will land currently under cultivation need to be used more intensively-the traditional focus for agricultural research, but mar- ginal lands will increasingly be brought into cultivation as rural people are compelled to move to find a means of support. ICRISAT is one of two CGIAR centers specifically mandated to focus on the problems of agricultural production in low rainfall and poorer soil areas; ICARDA is the other. A third, IITA, is specifically mandated to identify an alternative to shifting cultivation. This tradi- tional land-use management system for the fragile soils of the humid tropics is being rapidly overwhelmed by high rates of population increase. Vertisols are marginal soils, common in valley bottoms. These fertile soils are easily waterlogged and their high clay content makes them difficult to plough. ICRISAT's long-term experiments with Vertisol soils in India have established that improved drainage to reduce waterlogging is a crucial principle for managing these soils. The broadbed- and-furrow technology developed by ICRISAT reduces waterlogging and facilitates faster, earlier planting of crops. The broadbeds support the plant above the water table. The furrows provide drainage to pre- vent the water table from intruding into the root zone and causing waterlogging. ICRISAT's techniques are now being applied in Ethiopia in a collaborative project with ILCA and four national institutions. The project includes the development of a modified Ethiopian plow to prepare broadbeds and furrows. In 1986, field trials resulted in grain yield increases of 25-63 percent for breadwheat, 144 percent for durum wheat, 297 percent for horse beans, and 84 percent for finger millet. In trials with 34 farmers around Debre Zeit, the technology also reduced labor inputs by 40 percent for teff (a local cereal) and maize, thus substantially increasing labor productivity on these test farms. There is huge potential impact on the Vertisols-deep black cotton soils -widespread in sub-Saharan Africa. They account for about 70 percent of all highland African soils with slopes between 0 and 8 percent and cover 100 million hectares of the continent, including 7.6 million hectares in the Ethiopian highlands alone. The indications in Ethiopia are strongly favorable. Nevertheless, even if farmer acceptance is immediate, the CGIAR system will have well over ten years of work invested in the technology. 44 Broadbed-and-furrow technology is improving drainage on deep black cotton soils (Vertisols) in Ethiopia. 45 Triticale Other centers also devote significant effort to the problems of toxicities, drought, and low fertility that characterize marginal lands. For exam- ple, CIMMYT's 20 years of work on triticale has heightened awareness of the crop's potential in marginal environments. As more favorable pricing policies are introduced, the area devoted to triticale in developing countries is likely to expand further. The potential is very significant. Over 3 million hectares currently in wheat, barley, or rye could be sown to triticale because of its superior produc- tivity, including significant semi-arid areas of Asia, the Middle East, and North Africa. Other important potential production areas are the Brazilian cerrados, northern Zambia, and tropical highlands in many Southeast Asian countries. Triticale clearly outperforms wheat in these environments and some 15 million additional hectares could come under the crop (Table 4.1). This man-made cereal crop recently celebrated its 100th birthday, having first appeared in the 187Os, although it was treated more as a biological oddity than a crop for most of its history. A product of a cross between wheat, a member of the genus Triticum, and rye, a member of the genus Secale, it takes half its name from each parental genus. By 1986, however, triticale plantings worldwide passed the million-hectare mark, a tenfold increase in a decade. Much of this expansion is the 46 Table 4.1. Potential triticale area in developing countries (thousand ha). : ,,.. Highland/acid soil conditions Semi-arid .i_i_ ;. .Crop New _ I. Crop Location 5-_?i$ :- ,su@itutio~~l cultivation ;->Tota& substitution Source: CIMMYT result of commercial varieties that have been developed by national programs from CIMMYT materials and released to farmers since 1968. Of the 32 nations that have released triticale varieties, 11 are develop- ing countries, and it is in their more marginal areas that the new crop is beginning to find a valuable niche. In its early years, the CIMMYT work on triticale, done in collabora- tion with the University of Manitoba in Canada, used sites in both Mexico and Canada to advance two breeding generations each year and benefitted from the spontaneous outcross of a triticale with an unknown Mexican semi-dwarf breadwheat, which came to be called Armadillo. Its good agronomic traits proved to be highly heritable, and by 1970, most CIMMYT triticales had Armadillo in their pedigrees and Arma- dillo materials were distributed to breeders around the world. (Arma- dillo was later found to be a substitution of the 2D chromosome from breadwheat for the 2R chromosome of rye. Triticales having D chromo- somes are called substituted types and triticales having all seven R chromosomes are termed complete types.) Theoretically, the genetic constitution of triticale provides a built-in potential for adaptation to a wider range of soil and water conditions than other small grain crops. The International Triticale Yield Nurseries have been distributed continuously since 1969 and currently are grown in 71 countries at 115 locations. An analysis of the international nursery results in 1982 and 1983 across agroclimatic and edaphic conditions shows that triticales are comparable to breadwheats in favorable environments-the irri- gated subtropics and the Mediterranean region-and superior in dryland conditions, in the tropical highlands, and on acid soils (Fig. 4.1). Since 47 Fig. 4.1. Relative yield performance of triticales under differing agroclimatic conditions." Yield (kg/ha) Irrigated subtropics Mediterranean climate Dryland conditions Tropical highlands Acid soils Average of all 60 reporting locations Source: CIMMYTI14th ITYN data 1982-83. "Compared to the average of all 60 reporting locations and the long-term check varie- ties (1982-83). rye has high tolerance of acid soils, it is probably the source of triticale's generally superior performance in such soils. In one experiment, 10 of CIMMYT's best triticales were compared to 10 CIMMYT breadwheats most tolerant of acid soils, and the least productive triticale yielded more than the best-performing breadwheat. Although CIMMYT's triticale program commits some resources to breeding for optimal crop production environments where an alterna- tive to breadwheat or durum wheat might be needed, the outstanding performance of triticale in difficult environments has prompted the program to place emphasis on geographical areas where wheat is not a real competitor. 48 Bean varieties take hold in Latin America and Africa Growing beans is the major means of support for an estimated 5 million people in Latin America and Africa. For many more families, beans are an important part of their cropping system and diet. In parts of central Africa, for example, bean consumption reaches 50 kilograms per per- son yearly, and contributes as much as a third of total proteins in the diet. At CIAT, the adoption of improved bean varieties is monitored through farm surveys that also allow scientists to identify persisting production constraints that may warrant further research or a modifica- tion of selection procedures. By 1986, national programs in Latin America had released over 100 new bean varieties derived from CIAT germplasm. Studies have been undertaken in five Latin American coun- tries reporting widespread adoption. Table 4.2 includes an estimate of the additional production due to the new varieties. From 1975 to 1983, bean production in Costa Rica was stagnant, ranging between 11,000 and 16,000 tons a year. Then, starting in 1984, Costa Rica enjoyed three successive record bean harvests, and from 1985 it has had no bean imports, although it had been importing about half its bean requirements for more than a decade. In southern Costa Rica, the availability of new varieties and viable, more intensive, management practices has provided incentives for farmers to move away from the traditional shifting cultivation in which beans are broadcast and no inputs are purchased. The new technology is significantly more labor-intensive than the old one and therefore Table 4.2. Documented impact of improved bean varieties from CIAT germplasm network, 1986. Production Area in Percent increase improved area in due to new varieties improved varieties Country (ha) varieties [tons] Argentina 90,000 40 26,300 Costa Rica 21,700 62 5,300 Cuba 16,900 80 7,100 Guatemala 12,300 13 4,100 Nicaragua 14,000 17 3,500 Total 154,900 43 46,300 49 In Costa Rica, improved varieties and management suited to farmers with plentiful labor, but limited land. Small farmers, techniqu es for the common bean (Phaseolus vulgaris) have hoi istered production, particularly for small those with fewer than 10 hectares, were found most likely to couple the farmers. new varieties with more intensive management, weed control, and application of farm chemicals. The new planting system spread rapidly throughout Costa Rica. The potential benefits from better varieties and more intensive management have encouraged land-use methods to evolve in response to the increasing population pressures placed upo~ the traditional system. 50 CIAT's bean program in central Africa is much newer than the Latin American one. Its studies on Africa are therefore still concentrat- ing on identifying farmer and consumer preferences in beans. Table 4.3 shows varieties grown in on-farm trials in Rwanda as ranked by yield and by farmers' evaluation. The differences are striking and emphasize the importance of the identification of varietal types attractive to farm- ers and consumers alike. Farmers downgraded the highest-yielding variety, Ikinimba, because of its sprawling plant type, which makes weeding difficult. Farmers also felt that its tendency to lie on the ground would affect grain quality by increasing the danger of rotting, and they noted that the variety was difficult to thresh. In Rwanda, preferences by farmers and consumers for bean varieties differ from yields realized. Taible 4.3. Comparison of bean yield data and farmers' varietal preferences in Rwanda. Farmers' Yield preference Variety Yield (t/ha] rank ranking Ikinimba 1.65 A 197 1.22 ISAR mixture 1.22 Kirundo 1.07 Kiiyumukwe 1.05 Local mixture 1.05 Rubona 5 1.04 Umutikili 1.02 Source: CIAT 51 Fig. 4.2. ISNAR's involvement with research systems in 32 countries, 1981-86. Type of Overall system review Institutional review Human resources study Research strategy and plan study by : (performance potential, (internal management and , (personnel management development: giving a new country structure and organization linkages with operating including recruiting, sense of direction. and sometimes policy environment). compensation, performance, Year environment). and sometimes training). 1981 l Costa Rica l Indonesia `_ l Kenya 1982 l Burkina Faso l Rwanda l Pakistan l Fiji l Solomon Islands l Guyana l Ivory Coast l Malawi l Papua New Guinea 1983 l Dominican Republic e Madagascar l Cameroon l Western Somoa l Somalia l Sudan l Sri Lanka l Thailand l Zimbabwe 1984 l Zaire l Morocco l Bangladesh e Kenya l Jordan 1985 e The Gambia l Fiji e Panama l Sri Lanka l Tunisia 1986 l Niger l Costa Rica 4 Ethiopia l Zimbabwe Changing institutions ISNAR's role in the development of the national agricultural research systems shares an important characteristic with the development of improved germplasm-the pay-off is not immediate. Changing poli- cies, processes, and attitudes requires a no less patient and painstaking approach than plant breeding. Fig. 4.2 outlines the nature of ISNAR's involvement in diagnostic reviews with 32 countries since its founding in 1980 and indicates countries in which reviews have been followed up by direct ISNAR assistance in the development of research strategy and planning. The impact of the earlier collaborations between ISNAR and national agricultural research systems is only now becoming evident. Coincidentally with CIAT's work on the introduction of new bean varieties, in Costa Rica, ISNAR was involved in helping to develop a more efficient national research organization in the country. ISNAR had undertaken a review in Costa Rica in 1981, but there was little apparent reaction to its review report for several years. In 1986, Costa Rica invited ISNAR to conduct a second review, and in planning for the review, the Ministry of Agriculture and Livestock documented actions, begun in 1985, that stemmed from the 1981 report (Box 4.3). 52 53 Developing regional cooperation Among CGIAR centers, ISNAR is solely concerned with assisting national agricultural research systems to build managerial capacity, although most other centers contribute to the improvement of research organiza- tion at the national and regional levels. Through its potato program in Burundi, for example, CIP is gradually developing regional coopera- tion in potato research in eastern Africa. In 1979, the government of Burundi organized a national program for potato improvement with financial aid from Belgium, under the auspices of theInstitut des Sciences Agronomiques du Burundi (ISABU). After the necessary infrastructure was established, the program was reinforced in 1983 by the addition of a CIP scientist, also funded by Belgium. The program's principal objectives are to identify better-adapted cultivars and to ensure their multiplication and diffusion in adequate quantities. At the same time, the necessary agronomic and storage techniques are to be developed to exploit the varietal potential. The varietal improvement program is based on the evaluation in Burundi of improved tubers produced at CIP headquarters in Peru. Clones showing moderate late-blight resistance and other suitable agronomic character- istics, such as earliness, are put through a series of on-station trials, followed by multilocational trials throughout the country, ranging in altitude from 1,260 to 2,250 meters. Trials are conducted, as much as possible, under conditions found in farmers' fields. The program also evaluates some varieties or clones at an advanced stage of selection from CIP, as well as from sources in other developing countries, including Kenya, the Philippines, and Rwanda. Such mate- rial, particularly from neighboring countries like Rwanda that have similar agroecological conditions, has helped the program to move immediately into the later stages of evaluation and to begin multiplica- tion. Reciprocally, adapted material identified by the program is now being freely exchanged with neighboring countries. The final stage of the varietal evaluation program consists of a series of on-farm trials, which follow procedures developed by CIP. In these trials, the program tests the suitability of advanced clones for farmers' conditions and, perhaps equally important, gathers informa- tion from farmers on production constraints and characteristics desired. In 1983, a survey of the potato multiplication centers showed that almost all sites were unsuited to producing clean seed pototoes, prima- rily because bacterial wilt, root-knot nematodes, and powdery scab were present. Since no ideal site was found, the program decided to enlarge a small center at Mwokora, an isolated area in the north (2,300 meters elevation). This center has become the focal point for multiply- ing and diffusing quantities of high-quality seed tubers to other agricul- 54 tural development projects, which in turn further multiply the seed before passing it to other smaller projects or "farmer-multipliers." The farm at Mwokora has also proved successful for testing and demonstrating cultural techniques with application to many crops and regions throughout the country. Most of the farm is situated on steeply sloping land in an area where rainfall is high (1,800 millimeters a year) and often intense. The soils are acid (pH 3,9-&l), have over 45 percent silt and clay, and are suitable for cultivation only because the organic matter content is above 20 percent. The need for careful soil manage- ment of these delicate soils gave the program an opportunity to demon- strate conservation methods applicable to a high-rainfall area. A 2-1/Z year rotation has been established in which potatoes are grown only once during the five cropping seasons. In the other seasons, maintaining the level of organic matter in the soil by using green manure crops is emphasized. As an additional way to maintain soil fertility on the farm, beef cattle have been introduced to produce manure. Although animals are traditional in Burundi agriculture, the long dry season results in considerable weight losses if stored feed is not availa- ble. The program has demonstrated that it is possible to produce cheap fodder throughout the dry season by introducing various fodder crops, particularly swedes and kales, new to the Central African region. Trials to study the effect of farmyard manure on potato yields indicate that responses of up to 10 tons per hectare may be expected on the poorer soils-a significant boost to the seed potato multiplication scheme. 55 The generation of new ideas and the production of seed alone do not ensure the future of a crop; thus the program regards training of its own staff and the extension and research staff from other national agricultural development projects as an important task. Regular train- ing courses in both French and Kirundi are given by the program staff with support from ISABU staff, agricultural workers from neighboring countries, and CIP staff from Peru or its regional headquarters in Nairobi. Price controls A recent example of the impact of policy-related research is ILCA's work on meat pricing, in collaboration with the Ministry of Livestock Development in Kenya. The program also helped institutionalize the collection of market information to allow for more informed pricing decisions. Prices of basic commodities in Kenya have long been con- trolled by the government. Meat prices have been among the lowest in the world (retail prices were less than US$l.SO per kilogram in 1986). In 1982, ILCA economists reported that the low meat prices were discour- aging livestock production, and they suggested that the government Masai pastoralists in Kenya are benefitting from could not effectively set meat prices in the absence of systematic infor- deregulation of beef prices. mation on market prices, supply and demand, and production and marketing costs. They recommended a low-cost method of collecting such market information. In 1986, for the first time, the government set prices for beef above those for mutton and goat meat. ILCA's analysis of production and marketing information revealed a decade-long deterioration in the prices of animals sold by pastoral producers relative to the costs of resources needed to produce them. The important policy and institutional constraints were found to be low producer prices for cattle, an underdeveloped market for sheep and goats, and the ineffectiveness of group ranch management. The ILCA research emphasized that the pastoralists suffered from a declin- ing resource base in the face of rapid population growth, which was exacerbated by the government policy of keeping beef prices low for urban consumers. The study estimated that beef prices would have to rise 25 percent to maintain parity for the pastoralists. Consequently, the ministry reassessed the price-fixing process and decided to eliminate it, leaving prices to be determined by supply and demand. ILCA is evaluating the impact of the deregulation by monitoring livestock prices received by Masai pastoralists at Emali, a major regional livestock market. So far, producers' prices have risen 30 percent due to deregulation. The commodity centers working with national programs have a comparative advantage in this type of policy research, which depends on an intimate knowledge of their product in a particular local environ- ment. The contribution to the policy-making process, in this case, the 56 low-cost approach to the collection of livestock market information, is a permanent contribution to national decision-making. The immediate and observed impact is an increase in meat prices to producers. Over the medium term, producers would be stimulated by the higher prices to sell more animals. In the pastoralist system of the Masai, if this happens, it will have a favorable impact on resource conservation by reducing the stocking rates and the pressures on grass and soils. These later steps in the impact chain are much more difficult to monitor, particularly when the effects of the price increase are confounded with many other factors in a constantly changing production environment. It is impractical for the national systems and the CGIAR centers to continue to monitor in detail the widening impact of all the technologies they produce. It is, however, important that where macroeconomic effects are significant-self-sufficiency in rice in Southeast Asia and the need for diversification into other crop and livestock enterprises may be such a case-that the global agricultural research system is tuned to pick up such a shift and change research priorities. 57 5. Key CGIAR events The dominant theme of the CGIAR since mid-1986 has been elabora- tion of strategy in many fields. At its Ottawa meeting in May 1986, the CGIAR endorsed most elements of the research strategy proposed by TAC. It also approved the concept that setting priorities and deciding on strategies is not a periodic matter that can be done and set aside for a number of years. It is, rather, a continuing process in which new questions are raised as a result of fresh appraisals of experience, changed circumstances in the developing world, or opportunities opened by scientific discovery. As a consequence, the Group gave TAC a number of immediate questions for further strategic investigation. Shortly after the Ottawa meeting, TAC organized several subcom- mittees to help deal with its large agenda. * One subcommittee is exploring the global role of the CGIAR, that is, the relations of the centers among themselves, with other public and private research institutions wherever located, and with the national agricultural research systems of developing countries. 0 A second is considering the overarching issues of sustainability of agricultural production systems, an area of urgent concern to the centers, but which still requires a carefully designed research agenda as well as identification of practical institutional steps. 0 A third deals with the processes by which TAC monitors the implementation of CGIAR priorities and draws together the diverse product of continuing strategic analyses into a coherent whole. 58 Other aspects of TAC's work continue outside the subcommittee structure, including a proposal for bringing aquaculture research into the CGIAR and a similar proposal related to vegetables. TAC and the center directors are collaborating in the preparation of an overall policy for CGIAR involvement in the collection, preservation, evaluation, and utilization of plant genetic resources. The CGIAR secretariat is TAC's partner in the development of a new approach to the review process within the CGIAR. This work benefitted from Vernon Ruttan's study of the existing system, which was discussed at length at the Group's mid-term meeting in 1987. The Group strongly supported the study's contention that future reviews of center performance should give major consideration to the validity and quality of center research strategies. The issue of strategy is also at the heart of a new resource allocation process approved by the Group at the May meeting. In light of each center's strategy and results of external program and management reviews, TAC will undertake an overall look at the program of each center and then recommend its essential programs/activities for Group approval for a five-year period. Such center programs/activities will not be subject to further review until the end of the five years unless circumstances change significantly. TAC will also recommend other desirable activities for funding, which may be subject to more frequent review. One purpose of the new resource allocation system, which will be applied for the first time to three centers in 1987, is to ensure that center programs are congruent with system priorities. Others are to reduce both the detailed attention TAC must give to the annual budget exercise and the burden placed on centers in preparing for that exercise and to enable TAC to identify funding requirements for center programs over a five-year period, thus giving donors a sense of overall direction and requirements as a basis for planning. The new procedure does not imply that donors will be asked to make firm funding commitments for five years. A request for such commitments would cause both constitu- tional and practical difficulties for many donor members of the Group. External reviews An external review of ILCA was completed and considered by the Group. Continued attention was given to WARDA and IBPGR, follow- ing up on earlier reviews. ILCA. The external program review of ILCA, which was considered along with the management review at International Centers' Week 1986, found much progress had been made in resolving problems identified in the first such review, five years earlier. The review team also found that the basis had been laid for an effective program dealing 59 with livestock systems in Africa, but that the program should be better focused. Acting on the recommendation of TAC, the Group asked ILCA to prepare such a program strategy for consideration, with TAC recom- mendations at International Centers' Week 1987. The management review of ILCA likewise found significant improvement in the support base and management infrastructure, but recommended further improvement in the mechanisms for program delivery, particularly the process for establishing and implementing center strategy. The review closely examined the role of the center board. This emphasis was a change from previous reviews and was commended during the Group's discussion of the report. Although the team praised much of the board's work, it was critical of performance in providing leadership to management, specifically in the area of devel- oping research strategy. The center made a positive response to the suggestions of the review, and the CGIAR secretariat was asked to monitor the implementation of the various suggestions and keep the Group informed. WARDA. A remarkable transformation in the situation of WARDA occurred during the year covered by this report. In May 1986, the Group concluded that CGIAR support could not continue without major changes in the form of governance of WARDA's research program. At the request of CGIAR chairman S. Shahid Husain, a negotiating mission was undertaken by Moise Mensah, assistant president of the Interna- tional Fund for Agricultural Development (IFAD), under the auspices of the International Development Research Centre. It laid the ground- work for decisions by WARDA's member governments to transform their organization into a research center controlled by a board with powers comparable to those of boards of other CGIAR centers. With the diplomatic leadership of Minister IF. Sagna of Senegal, chair of the WARDA council, the new board was appointed, half from West Africa and half chosen by the CGIAR from the rest of the world. Sufficient funds were received from member states to liquidate WARDA's overhanging debt. Several donors gave strong support by providing funds to meet costs during the transition period; at the same time, staff and other expenses were reduced sharply. Meeting in June 1987, the board chose Eugene Terry, of Sierra Leone, as WARDA's first director general, and he was confirmed by the council of ministers. A strategy for rice research in West Africa is under preparation, and the board is seeking funds for the construction of a main research station at a location to be determined. IBPGR. For IBPGR, major progress was made in settling longstand- ing management issues through an agreement signed in December 1986 by the deputy director general of FAO, Declan Walton, and the board 60 chairman of IBPGR, W.J. Peacock. The agreement resolves issues identified-in the external review of the center, including a system for hiring of scientific personnel at appropriate grades, provision of adequate office space, and the responsibility of IBPGRstaff to the board. The agreement provides a framework for IBPGR to operate within FAO, as desired by all parties, through 1988. Before the agreement expires, its implementation will be appraised to determine whether it works and whether it should be extended with or without changes. Changing relationships with national systems The centers collaborate with national agricultural research systems that are at various stages of evolution. The pace at which national systems could assume responsibility for types of research and training done by the centers has been debated for a number of years. The discussion gains urgency as the rapid pace of change in biological science puts pressure on the centers to shift part of their limited resources to the application of new scientific tools to developing-coun- try problems. In January 1987, CIMMYT and IRRI staff participated in a forum arranged by IFAD, in collaboration with several other donor organiza- tions, on the prospect and willingness of several developing-country national agricultural research systems involved in rice and wheat research to assume more prominent roles in global research on those 61 crops. As reported at the Group's meeting in May, the results were positive: while the first concern of national systems obviously is to respond to national priorities, there are opportunities for expanded collaborative research, particularly in ecologies different from those at the centers, and in training. Recognizing that much more exploratory work needed to be done, the Group urged TAC and th e centers to follow up on this initiative and include specific proposals in their programs in the near future. One early step will be getting the views of the national systems whose needs would be potentially served in part through the collaboration of other national systems. Collaboration among centers A characteristic of center activities in recent years has been increased collaborative work both among centers and with other scientific insti- tutions. The need for centers to work together in Africa was recognized early in 1986 as being particularly critical and a task force was created, including TAC and donor representatives, as well as leaders of national systems in Africa. The task force has since met several times under its chair, TAC chairman Guy Camus. It works closely with a subcommittee of center directors that has a parallel purpose. The task force and the centers have identified three initiatives which they are pursuing: @ Mounting a program dealing with maize-based cropping systems in the mid-altitude regions of southern Africa (in collaboration with the Southern African Centre for Cooperation in Agricul- tural Research (SACCAR); o Determining how-in a to-be-designated African country-the need for external assistance in agricultural research can best be met from the CGIAR and other sources; and @ Studying the existing research programs on cassava and maize in West African countries and identifying further needs. Centers are compiling an inventory of their activities throughout Africa (through ISNAR) and are preparing a research strategy for Africa [through IFPRI). Collaboration among centers was also manifested in numerous other ways, among them: 0 A meeting in Kenya at the International Centre for Research on Agroforestry in September 1986 concerned with agroforestry research and involving ICRISAT, IITA and ILCA. Participants identified a number of potential areas for collaboration. e A series of meetings between center experts in plant genetic resources and TAC to work out an overall appraisal and policy statement for the Group. 62 0 A meeting of representatives of national agricultural research systems from Latin America at CIAT in August 1986 to discuss research priorities and practical lines of collaboration in the region (with representatives of CIMMYT, CIP, ICRISAT, and ISNAR) . l Creation of inter-center committees on public relations and on information and communication, both of which are working on strategies and programs for joint action. Scientists and institutions in industrialized countries Collaboration with European scientific institutions was a particularly appropriate topic for special attention at the Group's 1987 mid-term meeting at Montpellier, France, the center of French research on prob- lems of the tropical world. The preliminaries of this meeting offered many opportunities to appreciate the work being done by French insti- tutions involved in tropical agriculture. At the meeting, the results of a two-year study on existing and planned scientific collaboration between CGIAR centers and European institutions were presented. Conducted by Rudolf Binsack, on secondment to the CGIAR secretariat from the Gesellschaft fur Technische Zusammenarbeit (GTZ) of the Federal Republic of Germany, the study indicated a high level of pres- ent collaboration and future needs that are substantial in both range and volume. Also at the meeting, representatives of countries in other regions called attention to similar types of cooperation with institu- tions elsewhere and to possibilities of enhancing such relationships to the benefit of CGIAR programs. At the same time, questions were raised about modes of financing, particularly contributions that are restricted to expenditure through host-country institutions, The Group agreed that the principal test of such collaboration shpuld be the value to center research, and that the centers should be the initiators, lest they be overwhelmed with offers of collaboration. Personnel changes At the May 1987 meeting, CGIAR chairman Husain announced that because of a reorganization in the World Bank, he would be shifting to regional duties and would no longer be able to serve as chair of the CGIAR. The Group responded to the news with surprise and regret. A resolution of gratitude to Mr. Husain for his contribution over the past three years was signed by all present. The chair of the Group is being assumed by W. David Hopper, newly appointed senior vice president of the World Bank for policy, planning and research. This step will assure continuity of leadership pending consultations about the Group's needs from its chairman and the process of filling the position. A change in leadership occurred at ILCA in November 1986. Peter Brumby resigned as director general and was replaced by John Walsh. 63 I j Box 5.3. Prizes and honors. IITA was awarded the 1986 CGIAR W. Herdt with Beth Rose, received development of techniques to grow King Baudouin Award for its work the American Agricultural Eco- trypanosomes in the laboratory. with maize streak, one of the princi- nomics Association's 1986 Award King Carl Gustaf of Sweden pal cereal diseases in tropical for "superior achievement as exem- presented the International Inven- Africa (see CGIAR Annual Report, plified by quality of communica- tors Award in June 1986 to IRRI's 1985.) The award is presented every tion" for -the book, The Rice Amir U. Khan for the development / other year to one of the CGIAR cen- Economy ofAsia.(Resources for the of the axial flow thresher, a machine ters for a particular technology that Future/ IRRI, 1986). Barker and that processes high-moisture paddy improves the welfare of farmers in Herdt are former IRRI staff mem- rice and reduces threshing losses. developing countries. Streak-resis- bers. Herdt subsequently served as Versions of the thresher are pres- 1 tant varieties and hybrids are being a CGIAR secretariat scientific advi- ently manufactured in eight ' grown or multiplied in Benin, sor and is currently with the countries. Ghana, Nigeria, Tanzania, Togo, Rockefeller Foundation, John W. Mellor, along with and Zambia. By 1990, the Nigerian Henry M. Beach& andGurdev Bruce F. Joh=ton, feed the , government expects that approxi- S. Khush were awarded the 1987 American Agricultural Economics mately 2 million hectares of maize. Japan Prize, the nation's top scien- Association's 1986 Award for a will be seeded with streak-resistant tific award. Beachell, the former "publication of enduring quality" varieties-nearly all of the maize head of plant breeding at IRRI, and for "The Role of Agriculture in ECO- grown by Nigerian farmers. Khush, the current leader, were *omit Development" which 1 IITA's achievement was based cited for their roles in the develop- appeared in The American Eco- on the development of simple ways ment of the semi-dwarf rice varie- nomic Review in 1961. Mellor, to identify resistant plants; the evo- ties that launched the Green Revo- director of IFPRI, also was the first : lution of research methodologies lution in rice farming. They were social scientist to be awarded that helped scientists to understand the first a&culturists to receive the Finland's Wilhuri International the virus and its vector, the Ieafhup- award. The prize was awarded to Prize. He was cited for having per Cicudulina spp; and the cooper- Beachell and Khush in April 1987 ". . . furthered and developed the ation of national programs and in the presence of Crown Prince cultural and economic progress of other international centers, princi- Akibito and Princess Michiko. mankind." pally CIMMYT Studies show that Former IRRI director Robert F. IRRI director general M.S. the resistance in- the 1ITA lines is Chandler .received the U.S. Presi- Swaminathan received the World multigenic and is thus less likely to dential End Hunger Award in Cultural Council's Albert Einstein break down than lines containing Washingtan in October 1986. The Award in November 1986 in Guada- only a single source of resistance. award citation notes Chandler's lajara, Mexico, for his "outstanding The significant reduction in. the "continued, demonstrated vision, contribution to scientific research incidence of the disease is expected initibtive, and leadership in the and the appIication of science to to lead to greater stability of maize effort to achieve a world wsthout human welfare." Swaminathan also yields, encouraging farmers to use hunger." was one of five people to receive the more inputs, thus increasing total ILRAD senior scientist H. 1986 Krishni Ratna Awards from i maize production. Hirumi received the 30th annual the India's Farmers Welfare Trust Also in 1986, several CGIAR Noguchi Prize for Medical Research Society. The awards, which were staff and managers received inter- in Tokyo in November 1986. Himi presented by India's president, national awards for their work. is a member-of the team of scientists Giani Zail Singh, are conferred to I Among them: responsible for one of BRAD's people "devoted to the welfare of Randolph Barker and Robert major research achievements-the the Indian farming community." 64 6. The financial situation Contributions to the CGIAR in 1986 increased by US$ZS million, includ- ing US$22 million in core contributions and nearly US$4 million for special projects. Total funding amounted to US$235.5 million, or 12 percent more than in 1985. However, only about one-third of the increase was the result of higher donor pledges. About two-thirds of the growth in core contributions resulted from the weakening of the U.S. dollar relative to other currencies, and the system thus realized the effect of increases in non-dollar pledges which had occurred in previous years. One new donor, Austria, joined the group in 1986 and another donor, the African Development Bank, renewed its contributions, bring- ing the total number of donor members in 1986 to 39. Actual core funding in 1986 was US$192.2 million, US$7 million more than estimated at International Centers' Week in late 1985, an increase that was also mainly due to weakening of the U.S. dollar. After a net set-aside of US$3.8 million for the stabilization fund, the amount available from donor contributions to the centers' core programs was US$189.4 million (Table 6.1). Fig. 6.1. CGIAR core funding,l972-86 (US$million). 65 At the start of 1987, based on firm commitments from donors and informal estimates, the CGIAR secretariat projected that core contribu- tions would be US$191 million. This amount would finance about 97 percent of most centers' programs (provided that the World Bank does not set aside its normal contribution of up to US$5 million in the stabilization fund), which were appproved for funding at an aggregate level of US$196 million. The 1987 core funding estimate reflects a significant decline in U.S. dollar contributions, which is fortunately offset by an increase in the dollar value of non-dollar contributions. Table 6.1. GGHAR funding, 1982-86 (current U$$ million/. 1982 1983 1984 1985 1986 Total donor core funding" 143.8 164.7 173.2 170.2 192.2e (stabilization mechanism included) - l2.31 lo.91 L4.41 L3.81 Total core expendedb 147.0 163.8 177.9 176.4 189.4 Operations 136.0 150.9 157.9 163.3 175.2 Capital 11.0 12.9 20.0 13.1 14.2 Non-core (special project) donor fundingb 28.0 23.6 29.8 39.6 43.3 Total non-core expended Operations 27.0 23.7 28.5 35.7 41.3 Capital - 0.2 1.0 3.9 1.2 Total donor funding 171.8 188.3 203.0 209.8 235.5 Percent change from previous year in core funding 10 14c 5 -2 13 in non-core funding 37 -15d 26 32 9 in total funding 13 10 8 3 12 a Funding represents donor contributions only; centers also finance programs from income, carry-overs, and changes in working capital. A stabilization mechanism was initiated in 1984 to buffer center budgets against exchange rate and inflation rate fluctuations. b Core programs are those recommended by TAC and approved annually by the Group. Special projects are activities within the overall scope of each center, but not part of the currently approved program. c Including a transfer of about US$S million from special projects to core. Excluding this transfer, the growth rate in 1983 would have been about 8 percent. d Excluding the transfer of special projects to core. Including this transfer, the growth rate in 1983 would have been about 17 percent. e Including $3.0 million funding of capital projects for which commitments were made in 1986. Actual expenditures will be made in 1987. With continuing decline of the U.S. dollar, exchange rate gains are likely to raise 1987 core funding to US$199 million. Special project contributions are projected in the US$40 million-$45 million range, bringing total funding for the year to US$239 million-$244 million. 66 During the 1986 mid-term meeting in Ottawa, Canada, there was Figure 6.2. Core expenditures, 1986. general agreement among CGIAR members that active publicity and promotion are required on behalf of the centers' activities and accom- plishments. In this context, national support organizations are being established in Australia, Japan, the United Kingdom and the United States. Expenditure trends In 1986, the CGIAR's recent decline in operational expenditures in real terms began to reverse (Table 6.2). After taking into account an inflation rate of 6 percent, expenditures grew 1.3 percent over the 1985 level. Capital expenditures were US$14.2 million, slightly more than LJS$l million above 1985. In addition, centers' operatingfunds increased by about US$6 million, and centers are carrying forward about US$3 million for commitments on capital that were funded but not expended in 1986. Table 6.2. Center operating expenditmes in constant terms, 1982-86. [In constant 1986 US$ million] Center 1982 1983 1984 1985 1986 CIAT 19.7 23.2 23.1 21.4 21.3 CIMMYT 20.6 21.8 22.7 22.2 21.4 CIP 10.7 11.8 12.0 12.0 12.4 IBPGR 3.8 5.5 4.7 4.6 4.8 ICARDA 16.2 17.9 17.6 17.5 18.0 ICRISAT 16.1 19.9 18.1 19.6 20.6 IFPRI 4.1 4.8 5.0 4.4 4.5 IITA 22.6 20.9 20.3 18.9 17.5 ILCA 11.2 13.3 14.3 13.8 13.7 ILRAD 8.5 10.0 9.6 9.3 9.3 IRRI 24.0 23.7 23.1 22.9 23.6 ISNAR 3.0 4.4 4.2 4.4 4.4 WARDA 3.2 2.8 2.3 2.0 3.7b Total 163.6a 180.0 1 76.ga 173.0 175.2 Percent change 6.5 10.0 -1.7 -2.2 1.3 Additional expenditures: [current US$ million] Capital 11.1 12.9 20.0 13.1 14.2 Non-core expenditures 27.0 23.9 29.5 39.6 42.5 - "Due to rounding. bWARDA's total research program. Prior amounts relate to its core research program only. Core operating expenditures in constant dollars remained roughly the same or increased at 11 centers in 1986, but declined at CIMMYT Figure 6.3. Core research expenditures, and IITA, due to devaluations of the Mexican peso and Nigerian naira, 1986. respectively. Analysis of the 1986 core operating expenditures by program, commodity/activity, and region indicate consistency in general pattern over the past three years (Tables 6.3-6.5). Table 6.3. @ore expenditures by program (percent), 1984-86. Program 1984 1985 1986 Research 47 45 46 Research managementa 26 26 25 Strengthening national research capacityb 15 17 18 Research support 12 12 11 a Comprises general operations and administration. b Through information, training and institution building. Commodity/activity 1984 1985 1986 Cereals 41 40 39 Legumes 12 10 12 Roots and tubers 10.3 12 11 Livestock 19 20 20 Farming systems and genetic resources 14.3 14 14 Food policy 3.4 4 4 100 100 100 Table k5. Core expenditures by regions (percent), 1984-86. Region 1984 1985 1986a Sub-Saharan Africa 39 39 38 Asia 25 25 25.5 Latin America 23 22 22 North AfricalMiddle East 13 14 14.5 100 100 100 a Without an exchange rate gain of US$3.5 million at IITA, due to devaluation of the Nigerian naira, the distribution would have been 40 percent for Africa, 25 percent for Asia, 21 percent for Latin America, and 14 percent for North AfricaiMiddle East. 68 Direct research expenditures account for about half of operating Figure 6.4. Core expenditures by region, expenditures; the balance is divided among research management, strengthening of national research capacity, and research support. The proportion of expenditure on research management and research sup- port declined marginally in 1986, with a matching increase in funds for research and strengthening national research capacities (Fig. 6.2). Table 6.6. Center core program expenditures, 1982436. (In constant 1986 US$ million) Average annual change (percent) Program/activity 1982 1983 1984 1985 1986 1982-86 Research Cereals 29.8 33.5 32.2 31.1 31.6 1.5 Legumes 7.9 8.9 9.6 8.7 9.5 4.8 Food policy 2.2 2.9 3.0 3.2 3.0 7.3 Livestock 14.6 15.5 16.1 16.0 15.6 1.7 Farming systems and genetic resources 10.1 11.7 11.6 10.6 11.4 3.1 Roots and tubers 8.6 8.5 8.7 9.1 8.9 1.0 Subtotal 73.1 81.0 81.1 78.7 80.0 2.3 Strengthening national agricultural research 23.3 27.7 28.2 29.4 32.1 8.4 Research support 19.1 21.3 21.6 20.6 19.9 1.1 Research management 48.2 50.0 46.1 44.3 43.2 -2.6 Total operations 163.6a 180.0 176.ga 173.0 175.3 1.7 "Due to rounding. Table 6.7. Non-core program expenditures, 198486. (In current US$ million) % of Program 1984 1985 1986 Total Research 15.1 23.0 27.3 66 Strengthening national research capacity 12.9 11.0 12.0 29 Research support 0.4 0.7 0.9 2 Research management 0.1 1.0 1.1 3 Subtotal 28.5 35.7 41.3 100 Capital 1.0 3.9 1.2 Total 29.5 39.6 42.5 - 69 Stabilization mechanism The CGIAR stabilization mechanism was created to guarantee the U.S. dollar exchange rates of donor contributions prevailing on the pledging date at International Centers' Week and to finance extra costs when budgeted assumptions about inflation are exceeded. In 1986, coverage of the mechanism was expanded to include capital costs. The fund was used to compensate for adverse price movements in operating expendi- tures at CIP, ICARDA and ISNAR and for increased construction costs at ICRISAT's Sahelian center. These increases were due to the weakening of the U.S. dollar relative to the Dutch guilder (ISNAR) and the CFA franc (ICRISAT), and higher-than-budgeted local inflation for CIP and ICARDA. By contrast, IITA benefitted by US$3.5 million as the result of the Nigerian currency devaluation in October 1986. Normally, a center would pay such a gain into the stabilization fund. In this case, however, the CGIAR secretariat and IITA agreed that the gain should be used to replenish the center's working capital, which had decreased due to extraordinary expenditures in 1985. The stabilization fund started with a balance of US$6.9 million at end-1985 and rose to US$15.2 million in 1986, following a World Bank contribution of US$7.4 million and interest earned of US$O.8 million. In total, the mechanism paid out US$3.6 million for inflation-related adjustments, resulting in a balance of $11.5 million at end-1986. The CGIAR secretariat commissioned a study on the optimal size of the stabilization fund by the Financial Policy and Analysis Depart- ment of the World Bank. Its report, circulated to members at Interna- tional Centers' Week 1986, recommended that the fund be maintained at a level of US$lO million-$16 million to cover exchange rate and inflation risks within each calendar year. Other financial matters Following discussion of the recommendations of the finance and budget study, an accounting practices manual was prepared and circulated to Group members. An auditing practices manual and financial manage- ment handbook have been drafted and are being discussed with cen- ters' managers and boards. Medium-term resource allocation process According to the new five-year resource allocation process (see Chapter 5, Key CGIAR events), IFPRI and ISNAR discussed their medium-term proposals at TAC's March 1987 meeting, followed in June by three other centers (CIP, IBPGR, and ILRAD). Medium-term budgets for IFPRI, ILRAD, and ISNAR were recommended for approval to the Group. TAC expects to act on the proposals for the remaining centers in 1987 and 1988. 70 Annex 1. About the CGIAK. The Consultatrve Group on Internatronal Agrrcultural Research(CGLAR) IS an informal assocration of countries, mternatronal organizations, and prrvate mstrtutrons, cosponsored by the World Bank, the Food and Agrmulture Organizatron of the Unrted Nations (FAO), and the United Natrons Development Programme (UNDP) The threecosponsors brought donors together m 1971 to get the mternatronal commumty to earmark a small proportion of its concesslonal aid for agrmulture m the develop- mg countries to support, on a sustamed basis, a well-defined and cIosefy monitored program of research on food commodrtres CGIAR operates without a formal charter, relying on the consensus derrvmg from a sense of common purpose CGIAR started w&h a nucleus of four exrstmg mternatronal agrr- cultural research centers-CLAT, CIMMYT, IOTA,and IRRL-established by the Rockefeller and Ford Foundatrons m Colombra, Mexico, Nrgeria, and the Philippines, respectively At the start, there were 15 donors provrding about US$ZOm&on. The number of centers has now mcreased to 13, supported by 39 donor members and other contrrbutors provid- mg about US$235 5 million in funding m 1986 Each center supported by the CGIAR 1sindependent and autono- mous, with a particular structure, mandate and oblectrves, and with oversrght by rts own board of trustees Some centers focus on one commodity for which they have a global mandate, whrle others have a regional or ecologrcal mandate with, m some cases, a global mandate for one or more commodrties St111others perform specrahzed func- tions m the fields of food pohcy research, genetic resource conserva- non, and strengthenmg natmnal agricultural research m developmg countries The programs of the commodrty-orrented centers vary, but com- mon components mclude genetrc resource conservation and classifrca- tron. brologmal research to improve yields or mcorporate resistance to pests and diseases, farming systems studies to gain an understanding of farm-level constramts, and traming and other actrvrtres to strengthen natronal research systems The CGIAR's objectrves may be summarrzed as follows. "Through internatronal research and related actrvrtres, to contribute to increasing sustainable food productron m developmg countrres m such a way that the nutritional level and general economic well-being of the low-income people are improved " The CGIAR rs serviced by an executive secretariat, located m Washington, D C and provrded by the World Bank A Technical Advr- sory Committee (TAG), comprrsing a chauman and 14 screntrsts drawn equally from developed and developmg countrres, makes recommen- dations on research programs and priorities and monitors performance through annual program and budget reviews and perrodrc external 71 reviews by independent scientists mvrted to serve on specially constl- tuted panels TAC is supported by a secretariat, provided by the three cosponsors of CGIAR and located at FAO headquarters m Rome The CGIAR secretariat coordinates fund raismg m support of TAC- recommended work at the centers. In keeping with the concept of accountability to danors, the secretariat assists the centers m maintammg and disseminating sclentifm and financial mformatron on then work on a comparable basrs through reporting systems drawn up in consulta- tion with them The secretariat also provides some common services and overall admmistratlve support CGIAR meets twice a year, once in Washington, II C m October/ November and once elsewhere in May. The meetings recerve and dis- cuss recommendations on overall strategy, budgetary needs, and management rssues pertammg to the centers as a group Reports from mdividual centers, as well as independent externa1 evaluations, are presented periodically at these meetmgs. Present at these twice-yearly meetings are representatives of 10 developing countries, selected by regronal conferences of FAO, to represent the developing regions. The funds provided by donors are not pooled, but go dnectly to centers according to allocations made for each by indrvidual donors The World Bank uses its contribution to balance the amount each center receives in relatron to approved budget levels. Global locatzon of the 13 CGfAR-supported centers, Annex 2. CGIAR majm props and activities. ObJectives Center Regional focus Barley CIMMYT Latin America ICARDA Developing countries Cassava CIAT Developing countries IITA Sub-Saharan Africa Chickpea ICRISAT Developing countries ICARDA North AfricaiIvhddle East cocuyam IITA Developing countries Cowpea IITA Developrng countries Faba bean ICARDA Developmg countries Groundnut ICRISAT Developing countrres Lentil ICARDA Developing countries Maize CIMMYT Developmg countries IITA Sub-Saharan Africa Millet ICRISAT Developing countries Pigeonpea ICRISAT Developrng countries Potato CIP Developing countries Pastures CIAT Lattn America ILCA Sub-Saharan Africa Phase&us (field bean) CIAT Developing countries Rice Developing countries CIAT Latin America IITA Sub-S&ran Africa WARDA West Afnca Soybean IITA Sub-Saharan Africa Sorghum ICRISAT Developing countries Sweet potato CIP Latin America IITA Developrng countnes Tnticale CIMMYT Developing countrres Wheat CIMMYT Developing countries ICARDA North Africa/Middle East Yam IITA Developing countries Livestock ILCA Sub-Saharan Africa Theileriosis ILRAD Sub-Saharan Africa Trypanosomiasis ILRAD Sub-Saharan Afrrca Food policy IFPRI Developing countries Plant genetic resources IBPGR Global National research ISNAR Developing countries systems 73 Annex 3. CGfAR organization, May 1987. Continuing members: Australia Germany, Fed. Rep Phrhppmes Austrra Indra Saud1 Arabia Belgium Ireland Sweden Brazil Italy Swrtzerland Canada Japan Umted Kmgdom China Mexico United States Denmark Netherlands Finland Nrgerra France Norway Afrman Development Bank Arab Fund for Economm and Socral Development Asran Development Bank Commlssron of the European Communrtres Food and Agriculture Orgamzatron of the Umted Natrons Ford Foundation Inter-Amencan Development Bank International Bank for Reconstructron and Development {World Bank Internatronal Development Research Centre International Fund for Agrmultural Development Kellogg Foundation Leverhulme Trust OPEC Fund for Internatronal Development Rockefeller Foundation Umted Nations Development Programme United Nations Envrronment Programme Fixed-term members of developing countries: Africa-Guinea and Zambia Asia and Pacifm-Bangladesh and Thailand Latin America-Argentina and Venezuela Near East-Egypt and Turkey Southern and Eastern Europe-Poland and Portugal CGIAR Interim Chairman: W David Hopper World Bank 1818HSt,NW Washmgton, D C 20433, United States 74 CGIAR Executive Secretary: CUFtlS FarFar World Bank 1818 H St, N W Washmgton, D C. 20433, United States Technical Advisory Committee: TAC Chairman: Guy Camus do World Bank 66 Avenue d'Iena 75116 Pans, France TAC Members: Michael H Arnold Charan Chantalakhana CT de Wit Ola Heide Alexander McCalla Amlr Muhammed Ibrahlm Nahal Thomas R Odhlambo Ernest0 Paterniani Abdoulaye Sawadogo Winfried von Urff E. T York Tomlo Yoshtda TAC Executive Secretary: John H. Monyo TAC Secretariat Food and Agriculture Organization of the Unrted Nations Via delle Terme di Caracalla Rome 00100, Italy CGIAR-supported Centers: CfAT Centro Internaclonal de Agricultura Tropical Apartado Aereo 6713 Cali, Colombia DIrector General. John L Nickel Chair: W&nn E. Tossell' 75 CIMMYT Centro International de Meloramiento de Maiz y Trig0 PO Box6-641 Mexico 06600, D F Mexico Director General: Donald L Wmkelmann Chair Guy Vallaeys CIP Centro Internacionat de la Papa Apartado 5969 Lima, Peru Director General Richard L Sawyer Chair: John W Nleagher IBPGR International Board for Plant Genetic Resources Food and AgrrcuIture Organization of the United Natrons Via delle Terme di Caracalla Rome 00100, Italy Director. J. Trevor Wilhams Chair Wilham J Peacock ICARDA International Center for Agricultural Research in the Dry Areas P 0 Box 5466 AIeppo, Syria Director General* Mohamed A Nour Actmg Director General. G. Jan Koopman3 Chair. Jose I Cuber0 ICRISAT International Crops Research Institute for the Semi-And Tropics ICRISAT Patancheru P.0 Andhra Pradesh 502 324, India ICRISAT Sahehan Center BP. 12404 Niamey, Niger (via Pans) Director General. Leslie Swlndale Chair: Fenton V MacHardy IFFW International Food Pohcy Research Institute 1776 Massachusetts Avenue, N W Washmgton, D.C , 20036, Umted States Dnector John W Mellor2 Chair. Dick de Zeeuw 76 IITA Internatronal Institute of Tropmal Agriculture PMB 5320 Ibadan, Nrgerra Marhng address. IITA, Ibadan, Nigeria c/o MS Maureen Larkm L W Lmnbourn b Co. Carolyn House, 26 Dmgwall Road Croydon CR9 3EE, Unrted Kingdom Director General Laurence D. Strfel Charr: Lawrence A. Wrlson ILCA International Livestock Center for Africa P.0 Box 5689 Addis Ababa, Ethiopia Drrector General John Walsh Chair Ralph Cummnrgs, Sr. ILRAI3 International Laboratory for Research on Animal Diseases P. 0. Box 30709 Narrobr, Kenya Director General. A. R. Gray Chair* Hans E Jahnke IR.IU International Rice Research Instrtute P 0. Box 933 Manila, Phrhppmes Director General. M. S. Swammathan Chair: Kenzo Hemmr ISNAR Internatronal Service for National Agncultural Research P 0. Box 93375 2509 AJ The Hague Netherlands DIrector General* Alexander von der Osten Charr M Henrr Carsalade WARDA West Africa Rice Development Association F'.O.Box 1019 Monrovia, Liberia Acting Executive Secretary: Aheu Jagne Dtrector General. Eugene Terry3 Chair. Moctar Toure `Chaxr, Board &a= Gxmp Thair, Dwectors General Group 3Assumes poshon September 1987 77 Annex 4. Donor contributions to center programs, 1972-86 (in US$ million). Core Programs Total (Core + Non-core] Donor 1972-76 1977-81 1982 1983 1984 1985 1986 ' 1983 1984 1985 1986 Australia 400 13 28 377 406 400 4.18 452 411 403 4 27 485 Austria - - - - - - 100 - - - 101 Belgium 348 13.70 185 188 172 2 01 177 246 2 31 266 248 Brazil - - - - 1 00 - - - 1 00 - 01 Canada 1737 3614 8 29 9 94 1003 9 70 1066 10.74 1158 12 74 1426 Chma - - - - 050 0 50 48 - 0.50 050 48 Denmark 171 469 096 0 95 124 112 165 095 124 126 167 Fmland - - - - 050 060 99 - 050 060 99 France 105 3 14 0 90 101 0 88 123 207 1 10 0 94 139 215 Germany, Fed Rep 1327 3906 784 789 667 615 8 03 868 7 39 8 14 890 India - 0 50 050 0 50 050 0 49 50 0 50 0 50 049 50 Iran 1 98 300 - - - - - - - - - Ireland - 038 021 034 041 040 58 034 041 040 58 Italy 0 10 190 159 610 662 649 a33 6 10 662 6 78 9 73 Japan 249 26.25 a 85 913 9 72 1109 15 89 948 1046 12 05 1892 Mexico - 145 0 10 015 122 037 20 015 144 047 25 Netherlands 411 1154 3 24 3 58 328 389 665 412 3 79 453 788 New Zealand 0 11 014 002 002 0 02 0 01 -01 002 002 0 01 01 Nlgerra 1.30 536 113 100 1 00 0 85 19 140 160 1 29 38 Norway 3.33 927 187 219 1 92 2 27 3 12 219 192 2 27 340 Phrhppmes - 065 045 0 35 032 023 27 0.35 032 023 27 Saudi Arabia 100 100 - 150 150 - - 150 150 - - Spam - 050 046 0 52 052 050 50 052 052 0 50 50 Sweden 7 19 'I480 3 18 305 307 302 4.20 305 307 302 4 21 Switzerland 1.87 947 2 76 489 6 70 517 731 5 91 a21 780 908 Umted Kmgdom 902 2751 6 34 592 5 66 632 840 5 98 5 74 633 855 United States 4160 12809 4079 4455 45 25 45 16 4625 55 02 5685 6019 60 22 Country Subtotal 11498 35182 95.11 10952 11423 111.74 13336 12467 13246 13790 16130 Ford 16 79 620 0 ai 131 0.99 090 90 175 137 168 173 Kellogg 132 063 - 063 034 - - 0 69 041 - - Kresge 075 - - - - - - - - - - Leverhulme - 108 065 075 081 0.60 62 0 75 081 0 60 62 Rockefeller 1710 667 080 0.50 0 50 0 80 93 054 055 0 99 122 Foundation Subtotal 3596 14 58 226 319 2 64 230 245 3 72 3 14 3 27 3 57 ADB 0 30 120 - - - - - 017 045 064 71 AFDB - 0.15 002 - - - 59 - - - 59 AFESD - 112 0 24 0 23 0 23 0 34 34 0 23 023 0 34 34 EC - 1738 4 72 5.16 4 72 658 714 625 601 7 95 847 IDB 1115 3219 810 816 8 73 817 9 39 al6 8 73 817 944 IDRC 3 95 568 120 i a0 1.01 130 118 245 2 78 312 3 51 IFAD - 1105 594 837 7 02 315 45 1031 a67 5 26 122 OPEC - 190 3 58 2.25 219 1 00 47 2 25 219 105 87 UNDP 742 2159 619 686 806 7,49 842 7 16 912 885 887 UNEP 094 049 018 013 003 - - 017 003 002 03 WORLD BANK (IBRD) 1615 5333 1630 19 00 2430 2810 2840 1950 2468 2887 29 61 InternatronaI Donor Subtotal 39 91 14608 4647 5196 56 29 5613 5639 , 5665 62 88 6427 63 66 Other Donors - - - - - - - 3 29 460 437 7 01 TOTAL 19085 51248 14384 16467 17316 170.17 19220 18833 20308 20981 23554 78 Annex 5. ~~~~~-supp~r~~~ center expenditures, W71-86 (current US$ dltion). Core Operating Expenditures Center 1971-76 1977431 1982 1983 1984 1985 1986 CIAT 24 5 616 17 9 208 21.4 206 21.3 CIMMYT 34 0 715 178 179 203 210 21.4 CIP 87 313 89 93 99 96 12.5 IBPGR 1.4 120 31 45 41 43 48 ICARDA 1.4 32.8 115 138 14.8 16.0 18.0 ICRISAT 11.7 43.0 140 17 7 16 8 1%8 20.6 IFPRI - 103 31 3% 43 41 4.5 ETA 316 654 188 19 0 200 202 17.4 ILCA 44 337 82 101 11 8 126 137 ILRAD 25 28 3 75 84 85 8.8 93 IRRI 242 669 203 19 9 205 216 23 6 ISNAR - 24 23 33 3 3 3 8 4.4 WARDA 18 86 27 24 21 19 37 Total 1462 467% 1360 1509 157.9 163 3 1752 197146 Cumulatwe Center Operations Capital Special Projects Total CIAT 1881 186 186 225 3 CIMMYT 2039 96 342 247 7 CIP 90.2 93 55 1050 IBPGR 34.2 - 0% 350 ICARDA 108 3 37 6 11.1 1570 KRISAT 1426 412 246 2084 IFPW 299 07 8.2 38.8 IITA 2924 27.8 695 2897 ILCA 93.7 179 81 119 7 ILRAD 73 0 204 14 948 IRRI 195 2 166 570 2688 ISNAR 190 07 24 22 6 WARDA 234 16 170 42 0 Total 1,393 9 202 0 2589 1,854 8 79 Region Staff % Trustees % Asia 154 18 3 37 19 7 Sub-Saharan Africa 111 13 3 35 18.6 N AfrlcalM East 34 42 10 5.3 Latin AmencaKanbbean 95 114 23 12 2 Europe 212 25 4 44 23 4 North America 196 23 5 30 16 0 Au&al&New Zealand 33 40 9 48 Total 835 100 188 100 Center ASla Sub-Saharan N Africa/ L AmerIcai Europe N America Au&alla/ Total Africa M East Caribbean New Zealand CIAT 5 - 1 20 19 24 5 74 CIMMYT 13 6 3 23 17 35 8 105 CIP 7 4 3 34 27 16 - 91 IBPGR 2 1 1 3 7 2 - 16 ICARDA 9 4 19 3 14 12 3 64 ICRISAT 27 9 3 1 24 19 5 88 IFPRI 13 1 1 2 2 10 - 29 IITA 33 32 1 4 21 28 1 120 ILCA 2 11 - - 33 7 4 57 ILRAD 2 6 - - 33 10 1 52 34 - - 3 5 23 5 70 ISNAR 3 2 1 2 10 10 1 29 WARDA 4 35 1 - - - - 40 Total 154 111 34 95 212 196 33 835 Amex 6~. ~e~i~na~ origin of board trustees by center, 1986. Center Asia Sub-Saharan N Africa/ L America1 Europe N America Australia/ Total Africa M East Caribbean New Zealand CIAT 2 1 - 7 2 4 16 CIMMYT 4 2 1 3 2 3 1 16 CIP 3 - - 3 1 1 1 9 IBPGR 5 2 - 1 6 2 1 17 ICARDA 1 - 7 1 6 1 1 17 ICRISAT 4 2 - - 4 4 1 15 IFPRI 4 2 1 2 2 4 1 16 IITA 1 6 - 3 4 2 - 15 ILCA - 5 - - 5 2 - 12 ILRAD - 5 - - 3 3 1 12 IRRI 9 1 - 1 2 2 1 16 ISNAR 3 2 - 1 4 1 1 12 WARDAl 1 7 1 1 3 1 - 14 Total 37 35 10 23 44 30 9 188 `WARDA Saentlftc and TechnlcaI Committee 80 Consultative Group on International Agricultural Research CGIAR Secretariat 1818 H Street, N.W. Washington, D.C. 20433 ISSN0257-3156 United States