RICHARD R. HARWOOD
Michigan State University

The challenge for the coming decades of sustainable agricultural development is to transform existing farms (agricultural businesses) into new industries that meet evolving human needs, use resources efficiently, and are environmentally stable. Over the next 25 years, efforts to provide sustainable agricultural development will be confronted with a massive increase in the world's population (perhaps a doubling), which, when coupled with economic growth, will greatly increase the demand for goods and services generated from the land and water base. Although the world's food supply will be plentiful and will see only modest price increases over the next decade, after that as yet undiscovered technologies and new resources must be brought on line to sustain growth.

Also over the next 25 years, global trade patterns, together with ongoing development, will open markets for industries located wherever production resources are available. For agriculture, the emergence of biological technologies for the transfer of the ability to produce new industrial precursors, essential oils, and flavor compounds into agricultural crops will open many new markets and create new agricultural industries. These new industries will permit production of more products directly from solar energy, helping the transition to a solar economy, but at the same time they will put more pressure on land resources because of the increased demand for additional products from agriculture.

The “frontier” period in freshwater and marine fisheries has ended; natural supplies of fish and their products are declining while global demand is increasing. Thus as limits are placed on harvest from native stocks in an effort to optimize and sustain natural productivity, new industries devoted to “fish farming” will be required to meet growing demand. Also during the next decade the earth will have reached the end of its “frontier” period in the harvest of virgin forest. The forest industry, then, will have to undergo transformation on a global scale to meet the growing demand for its products. As for water, demands will increase, with ever-greater emphasis on quality. Its value (and cost) will therefore continue to increase.

FIGURE 1 Agricultural productivity of land worldwide (million hectares). SOURCE: Rockefeller Foundation, 1993 Annual Report (New York: Rockefeller Foundation, 1994).

Overall, the environmental interactions among geographical areas and between industrial and human activities within areas will continue to intensify. Industry will be under increasing pressure to control outflow to the environment during production and to recycle the post-consumption product or materials. The problem of control of materials during production is particularly relevant to agriculture, which, to be economical, must manage pests and mobilize nutrient movement at high rates.

As for the land itself, the next 25 years will see a leveling off in the availability to agriculture of the land most productive for cultivated crops, followed by a decrease. The most productive agricultural areas, on a global basis, are only a small portion of the total land base, accounting for 447 million hectares out of a total 13,900 million hectares, or 3.2 percent (Figure 1). The “slightly” and “some-what” productive lands add another 1,087 million hectares.¹ With advances in technology, some of the soils with “lower potential” can be made productive and converted to arable farming, especially in meeting the growing demand for products from perennial crops which do not require frequent soil tillage.

From 1950 to 1981, the global grainland area increased about 0.7 percent a year, while from 1981 to 1992 the grainland area decreased at an annual rate of 0.5 percent. This decrease is caused by many factors, including lower world prices for grains, the conversion of land to nonagricultural uses, and the transfer of the less productive lands from cereals to more suitable crops. Irrigated land in developing countries increased at a rate of 2.17 percent a year from 1961 to 1971, 2.09 percent a year from 1971 to 1981, and 1.24 percent a year from 1981 to 1990.² The most optimistic outlook for an eventual increase in irrigated area in developing countries is 59 percent above present acreage,³ but also noteworthy is the fact that the per hectare costs of such irrigation development increase as the quality of land and the availability of water decrease.

FIGURE 2 Responses of resistant, resilient, fragile, and marginal soils to stress (see note 4). SOURCE: Adapted, with permission, from R. Lal, G. F. Hall, and F. P. Miller, “Soil and Degradation: I. Basic Processes,” Journal of Land Degradation and Rehabilitation 1 (1989): 5-69. © 1989 by R. Lal, G. F. Hall, and F. P. Miller. Reprinted by permission of John Wiley & Sons, Ltd.

The pockets of hunger that continue in developing countries caused by the plethora of social, political, climatic, and other disturbances will require globally available stocks of food. The poverty evident in both rural and urban areas is, of course, an ongoing problem. Rural poverty interacts with agriculture on several fronts. Its greatest long-term impact is seen in the overexploitation and degradation of fragile soils by production enterprises geared toward short-term human survival rather than optimal long-term productivity. The land areas in developing countries that are receiving the greatest portion of increases in rural population are those areas having lower soil and water availability. With their increased fragility, these poorer soils have far less tolerance for stress before reaching an irreversibly degraded phase (Figure 2), and many if not most of the less-productive soils fall into the “fragile” and “marginal” categories as defined by Lal. ⁴ Most of these lower-potential soils will not permanently support cultivated crops on large, contiguous areas. Annual crops such as cereal grains are not highly responsive to inputs when soil quality and water are the primary limiting factors.

TABLE 1 Prevalence of Poverty in the Developing World, 1990 and 2000

NOTE: The poverty line is defined as an annual income per capita of $370 in 1985 purchasing power parity dollars.

SOURCE: Reprinted, with permission, from World Bank, World Development Report 1992 (New York: Oxford University Press, 1992). © International Bank for Reconstruction and Development.

Millions of poor people depend on these low-productivity areas for their subsistence. While the overall percentage of poor people in developing countries is projected to decrease during the decade of the 1990s (Table 1), the combination of population growth and migration to low-resource areas may actually cause major population growth in those areas. In the mid-1980s nearly half of the developing world's poorest people-some 370 million-lived in low-potential rural areas (Table 2). But development in high-resource areas and subsequent outmigration from poor areas will not, by themselves, solve the poverty problem in the near term for most of these people. Sustainable development must include optimizing resource use in these low-productivity areas through use of appropriate production strategies.

THE BROADENING (AND GREENING) OF THE AGRICULTURAL DEVELOPMENT PARADIGM

Against this rapidly changing global development picture the public demands on agriculture (the development paradigm) are evolving. Ironically, the quality of life of an increasing proportion of the earth's inhabitants depends on the structure and activities of nearby agriculture at the same time that the number of people making their living from farming is decreasing. Public agendas for agricultural change will continue to broaden as social and environmental dimensions are added to economic demands. The breadth of the public agenda depends on three factors: (1) food sufficiency-if food is scarce or the supply is insecure, the agenda will focus narrowly on production, as in the 1960s during the early days of the green revolution when new rice and wheat varieties were combined with the rapid development of rural roads, electrification, irrigation systems, and production credit using models defined by such development giants as Art Mosher; ⁵ (2) the degree of a country's political and social pluralism-centrallycontrolled and planned economies and societies tend to have narrow agendas; and (3) the overall affluence of the majority of people. Superimposed on all of this is the “global” agenda of the various development agencies and institutions, both large and small, emanating largely from the developed countries. As a result, an environmental or social ethic may be forced on a developing nation well ahead of its evolution at the public level. Finally, in an increasingly populous and interdependent world, the agricultural agenda will inevitably broaden to include ever greater levels of: human utility-measured in terms of production, employment, safe food, dependability, recreation, green space, and a host of other factors; efficient use of resources-measured as land use, return on investment, and use of inputs; preservation of nonrenewable resources, including soil, water, and genetic diversity; environmental impacts favorable to humans and most other species; and macro structure in harmony with local and national economic, social, and political goals. This broadened agenda is spelled out in great detail in the Agenda 21 report of the 1992 UN Conference on Environment and Development (UNCED). ⁶

TABLE 2 Distribution of the Poorest of the Poor within Lowand High-Potential Areas, Mid-1980s (millions)

NOTE: The poorest of the poor are defined as the poorest 20 percent of the total population of all developing countries.

SOURCE: Reprinted by permission of Transaction Publishers from H. J. Leonard, “Overview-Environment and the Poor: Development Strategies for a Common Agenda,” in Environment and the Poor: Development Strategies for a Common Agenda, ed. H. J. Leonard and contributors (New Brunswick and Oxford: Transaction Books, 1989). © 1989 by H. J. Leonard; all rights reserved.

One might debate the rate of change called for in this agenda, but its overall direction cannot be altered significantly. Most development agencies, including the international agricultural research centers, have reviewed their programs and developed detailed responses. ⁷ Individual centers such as the International Board on Soil Research and Management (IBSRAM) have responded with specific reference to their subject areas, ⁸ and these responses are quite typical of the dynamic changes under way.

TECHNOLOGICAL BREAKTHROUGHS

Because the world is facing a shrinking land base and growing demand for agricultural products, the output per unit area of food and feedgrains, as well as starchy vegetables, must more than double over the next 25 years. While there is considerable scope for increasing yields within the existing genetic potential, scientific breakthroughs will be needed to fully achieve the required yields. The recently announced achievement of an up to 30 percent increase in the genetic yield potential of rice is exactly the kind of progress needed. But many more years of highly complex (and expensive) research and development will be necessary to bring this new rice technology to farmers' fields. And, just as important, yet to be developed are the production systems for managing pests, water, and the high nutrient flows needed to achieve such yields without additional pesticide, fertilizer, and plant or animal waste loss to the environment.

Yields of rice in Asia, when adjusted for climate, average 57 percent of the present genetic yield potential of 8.0 tons per hectare. But the combination of pest management, disease control, and soil quality and nutrient management technologies that are required to close the yield gap are not yet available. Most important, the sustainable yield potential must be increased. Yet it is not feasible to expect average production to rise to more than perhaps 70-80 percent of maximum potential yields because farmers are unable to control the many yield-reducing variables to the extent that researchers can.

The rapidly evolving science of production ecology is giving new insight into mechanisms for better management of biological process, which should dramatically help in this regard. Researchers' abilities to manage soil microbial populations to better synchronize seasonal nutrient flux with crop demand, to manage genetic shifts in pest resistance to control measures, and to genetically engineer new characteristics into key crop-influencing organisms are but a few examples.

The genetic transfer of a plant's capacity to produce essential oils, industrial products' medicines, or other useful products represents a new wave of scientific breakthrough. The ability of plants to produce a chemical precursor for biodegradable plastic is an excellent example. Because this chemical is linked to the production of carbohydrate, this trait will be transferred first to such crops as sugar beet. If an investment were made in transforming oil palm into a plastics producer, a new industry could form in the humid tropics, where much of the nonarable land is ideally suited to sustainable tree crop production. If land tenure, credit schemes, and a development strategy for an appropriate scale of farm enterprise were in place, such an industry could have a major impact on both the productivity of the land resource and rural poverty.

In some cases, the new technologies, such as the genetic transformation of fungi or bacteria to produce high-value products, may move traditional agricultural products from a land-based to a “factory”-based system. Among other products, chocolate and vanilla production may be affected, with potentially disastrous social and economic consequences for the people and countries now producing and exporting these items. Ghana, for example, now earns a major portion of its foreign exchange from exports of cocoa and will be forced to shift a significant part of its agriculture toward alternative products.

RAPIDLY EVOLVING AGRICULTURAL INDUSTRIES

Several industries such as fisheries, farm forestry, and selected animal enterprises are evolving rapidly on a global scale to meet changing market demands. In Indonesia, because the harvest of virgin forest has passed its peak, the price of logs of certain timber species has risen by 20 percent a year. ⁹ This is prompting a significant shift in Indonesia's timber industry toward land management and timely production rather than continued reliance on native stands of trees. Private forest land held on long-term lease is beginning to encroach on what was once the exclusive plantation domain of the government.

The “industrialization” of swine and poultry production in Asia over the past decade illustrates the direction that fish and other animal production is taking. Nonruminant animals and fish, in particular, respond dramatically to a combination of quality genetic stock, control of disease and parasites, and proper nutrition. In fact, productivity per animal can be increased up to fivefold over that of animals foraging untended in a village-level environment. But while vertical integration of feed, veterinary care, production, processing, and marketing has brought extremely high economic efficiency, it also has brought with it significant problems, even in developed countries. Indeed, in many developing countries where public institutions have little influence on environmental protection, the situation can be intolerable, when, for example, the uncontrolled use of antibiotics to maintain animal health threatens food safety. Nutrient containment and recycling from animal wastes is a serious concern. Nitrogen and phosphorus loss from manure can be a major contaminant. As agriculture intensifies, nitrate (the most troublesome soluble form of nitrogen) contamination of drinking water will become one of the most serious environmental problems. The solution to this problem will depend on the proper location of animal confinement facilities with respect to water sources. Higher locations provide greater nutrient management options. The huge amounts of feed and water used in these operations create a very large volume and tonnage of nutrient and bacteria-rich water. A facility located in the upper part of the landscape can distribute much of the waste by gravity to the agricultural land where it has great benefit. A poorly situated facility must undertake expensive containment, pumping, and long-distance hauling operations. In short, few developing country animal confinement facilities are now sustainable. Moreover, it is highly probable that regulatory control will be essential for long-term public acceptance. Fly and odor problems, which are related and dependent on location, require specific attention. Livestock operations in Japan are facing the enormous social and environmental pressures from flies, odor, and waste disposal soon to be felt in most countries, where agriculture coexists with an increasingly intensive and diverse use of lands.

TABLE 3 Comparison of the Biophysical, Social, and Economic Attributes of
Land-Use Systems in the Humid Tropics
Biophysical Attributes

NOTE: L (low), M (moderate), and H (high) refer to the level at which a given land would reflect a given attribute, using widely available technologies for each land use system. Symbols in parentheses indicate potential of best technologies assuming continued short-term (5- to 10-year period) research and extension,

SOURCE: National Research Council, Sustainable Agriculture and the Environment in the Humid Tropics (Washington, D.C.: National Academy Press, 1993), 140-141.
Social Attributes Economic Attributes

FIGURE 1 Evolution of modern electronics manufacturing.

Modern industries are just beginning to evolve around many nonfood agricultural products, many of them included in a broad range of crop, livestock, and tree-producing systems that are better suited to lands unfit for cereal crops. Specialty wood such as rattan for furniture, bamboo for construction, palm fronds for basketry and handicrafts, medicinal herbs, wood for charcoal, fodder trees for animal feed, landscape nursery plants, and even fresh market cut flowers are derived from farming systems that have intricate combinations of annual and perennial crops and a variety of animals. Farm output is used for family consumption and for sale to local markets. Many high-value products from such systems-such as handicrafts, spices, and a large number of animal products - even reach national and global markets as national economies develop.

These production systems have a wide range of biological, social, economic, and environmental impacts (Table 3). Many have a high level of environmental sustainability on fragile soils-that is, they have high nutrient recycling and retention and high biological stability, thereby requiring few if any pesticides - and many are labor-intensive, with a quite high economic return on labor. They are usually small scale, are well adapted to small land holdings, and are reasonably specific in their environmental, economic, and sociopolitical applications. Finally, they usually produce small amounts of each product and so are adaptable to specific market opportunities. Because many plant species grow only in certain environments, there is no fixed formula or pattern for how the thousands of plant and animal species should be assembled in a “pre-designed” system.

Any national agricultural system is composed of a mixture of such production systems (Figure 3), which today are proliferating in response to the dramatically increasing markets in the industrially developing countries such as Indonesia, Thailand, and Korea. Increasing population pressure on the dwindling supply of agricultural land in these countries is forcing an evolution toward intensive, highly diverse systems at the expense of less-intensive grazing or shifting cultivation. These land-use changes are not without conflict. Population pressures in the industrialized countries cause breakdown of traditional (often tribal) land tenure arrangements. In countries such as Brazil with a large land base, the evolution from forest to slash and burn and the move toward either large-scale cattle ranching or intensive smallholder use is often fiercely contested, with political legal, and economic powers pitted against smallholder interest.

FIGURE 3 Area balance of agricultural and forestry land-use types that have evolved over past decades from native forest in a typical developing country. Land clearing is usually by slash and burn, and moves toward alternatives uses are based on soil type, land tenure, farm size, market availability, and other factors such as domestic security and availability of capital.

To meet a range of market and social needs and to operate within local environmental and resource constraints, farm structures in developing countries must take several forms. l[ndustrial enterprises, operated under appropriate social and environmental controls, will play an important role, as will the limited-diversity, mostly small farms in the high soil and water resource areas, which will continue to serve as their countries' breadbaskets. Appropriate systems for the increasingly populous, environmentally sensitive areas that are now under great stress will have high biological diversity and a high proportion of perennial plant species and animals, and will be highly integrated both biologically and socially. Because of the requirements for perennial species, such systems must maintain greater amounts of carbon in their plant and animal biomass than do systems in high soil and water areas. Carbon accumulates in the system as a result of photosynthesis and plant growth, being “fixed” from carbon dioxide in the air as part of trees, roots' organic matter in the soil, and in plant and animal residues. In fact, these systems, at least during the preindustrial stages of economic development, would operate very much as a carbon-based economy; carbon accumulated in its many diverse forms in the system would act as “biological capital” - much like the financial capital that must be accrued by other businesses to prosper. Economic and social conditions must favor a long-term outlook for farmers to permit carbon stocks to accumulate to levels that form a base for nutrient recycling and high productivity.

SOLUTIONS TO KEY PRODUCTION PROBLEMS

The technological breakthroughs and the industrial systems emerging in agriculture share many real but solvable problems, illustrated by a few of the more pressing examples, in addition to those mentioned earlier.

Single-commodity, “industrial” crops are associated with heavy pesticide use, soil erosion, and waste disposal problems (processing waste-plastic materials in the case of bananas). The solutions to these problems-integrated pest management, some degree of landscape diversity, and careful soil management - are complex and often costly. Thus regulation and subsidies or incentives probably will have to be a driving force behind the adoption of such solutions. As large-scale, vertically structured industries come under increasing international and local regulatory pressure, they will develop and employ the corrective technologies, but production costs will increase. Publicly funded research or correction should not be necessary in most instances.

Intensive, small-scale systems in high soil and water resource areas fall prey to problems stemming from pest management (pesticide loading), crop nutrient availability and loss, soil erosion, and salinity (rising water tables). This area will require considerable public technical assistance through traditional public research and extension channels. The breeding of pest- and disease-resistant crops, predator management, and farmer training in integrated pest management (IPM) methods are crucial to reducing the need for pesticides.

Nutrient containment and recycling will become increasingly important as systems continue to intensify and as nutrient inputs are increased to raise productivity. Work in this area has focused on the economics of higher efficiency, but as water supplies become increasingly threatened by high nitrate levels, public pressure to halt such contamination will increase. Current work in the management of soil biota for nutrient mobilization and recycling (containment) within soil shows considerable promise. Such processes are common to the intensive, mixed-culture systems of traditional agriculture. The exciting news is that these very efficiencies seem to work under the high nutrient flow and turnover rates of the high production systems in developed countries. This is one of the most promising lines of research toward sustainability and cost effectiveness in highly productive systems.

Water management, in terms of both supply and drainage, often has a large public component. During the construction of water systems, budget constraints often lead to inadequate drainage systems. This results in rising water tables, thereby turning large areas into wetlands, and to the accumulation of soluble salts, which reduce the production potential of the land. Unfortunately, the measures needed to correct these problems after the damage occurs are extremely expensive and may take years.

Mixed systems for lower soil and water resource environments suffer from lack of land tenure permitting long-term investment, soil erosion and lack of nutrients, as well as lack of infrastructure, social stability, and security in the community. In fact, problems in areas having these systems are as much social and political as they are technical. These areas often are newly settled with no land tenure history as in the better soil areas. Social and political institutions are weak, frequently with little security. The productivity of these areas depends on a mix of perennial crops, livestock, and land improvement, all of which require adequate conditions for long-term investment. A farmer's investment in long-term crops (the accrual of biological capital) will require more labor than cash but will require the same investment environment as will the control of soil erosion.

Ultimately, the cycling of nutrients becomes limiting in such systems. Thus eventually the ecological processes for containing and recycling applied nutrients must be mastered so that such mixed systems can be sustained. These systems are, after all, critical to the support of rural populations, to national productivity, and to overall political and social stability.

In addressing the problems of how to use resources efficiently and how to maintain the environmental stability of the many highly promising technologies and emerging industrial systems, scientists have realized that they know little about the ecological processes involved. Enormous potential exists for harnessing soil biota to more effectively release soil nutrients at low levels of availability ¹⁰ and to contain nutrients at high flow rates by achieving greater synchrony between soil nutrient release and crop uptake. ¹¹ Likewise, the potential for enhanced pest and disease management by manipulating ecological processes has barely been tapped. But even though more effective harnessing of these processes will improve the efficiency of nutrient recycling, it will not eliminate the need for fertilizers. Similarly, more effective pest management will reduce but not eliminate the need for pesticides. Ecological methods thus hold great promise for increasing production efficiency and reducing the environmental damage caused by high-productivity agriculture, but the development of the scientific basis for production systems ecology will require a major input from the public sector.

PATHWAYS OF CHANGE FOR DEVELOPING COUNTRY SYSTEMS

Three basic, often overlapping pathways for the generation and movement of capital, technology, knowledge, and production resources are commonly found throughout agriculture. Each pathway has a range of options, depending on the stage of agricultural development and the specific circumstances encountered, but the three main pathways are each suited to particular agricultural situations. Most rural development uses combinations of the three.

The private sector industrial-dominant pathway is the most narrowly focused and specific. In this model, technology is highly privatized and, in many cases, may be proprietary. The industries tend to be vertically integrated economically and capital-intensive. This model is most applicable where standardization and control of the production process are possible, where local environmental and social interaction are modest or low (or are disregarded), or where a homogeneous environment exists. Large-scale confinement animal operations and plantation crops for export such as sugar and oil palm are examples.

The public sector land-grant, research and extension-dominant model with its many variations is the second pathway. In this model, a high proportion of technology is generated within public sector institutions, with delivery to farmers heavily dependent on public channels. This pathway has a high degree of centralization of both technology and of knowledge. It is the most applicable to the small number of crop and animal commodities that constitute most agricultural value and are widely produced. Traditionally, this pathway has served well in production areas of high homogeneity and high production potential where a modest number of research and development sites are representative of broad areas. Its reach extends generally toward a greater diversity of production environments than that of the industrial pathway, but, because of its centralization, it has significant weaknesses in dealing with a high level of biological integration in very diverse production systems or with high levels of social integration. The well-known green revolution approaches are excellent examples of the effective use of this pathway.

The third pathway is a decentralized, networking-type model in which the energies for change are concentrated at the community-based social and political levels, with a much greater proportion of technology and information flow being horizontal. (Pathways for change that operate with a high degree of farmer-to-farmer linkages in no way imply communal ownership of land.) The decision-making process for the selection and application of technologies is more decentralized than in the first two pathways, with technologies and resources entering the system at widely scattered points. Thus the system requires a higher level of social infrastructure and networking than do the more centralized pathways. In such systems, the portion of indigenous, or community-based, knowledge is high in relation to new knowledge or technology entering from outside. Outside sources are usually needed for solving specific problems or understanding a process.

These three pathways for agricultural development are systems-specific. Nevertheless, the proportionate mix of each pathway must be carefully determined for each type of agricultural system (each agricultural sector).

Development of the industrial sector is highly dependent on the investment climate, which, among other things, depends on social and political stability and on government policy. Although there is considerable international experience in this area, the environmental and social acceptability of such operations, given the relatively undeveloped regulatory frameworks existing in most developing countries, is sometimes abysmal. Some multinational companies, such as the Cargill poultry processing plant in Thailand, have done a good job in treating and recycling wastewater, but, overall, industry's record is not good. As a result, the developing countries, like many of the more developed countries, will undertake greater regulation to correct the most flagrant abuses. Generally, the creation of new technologies through research and development for these industries will continue to be in the private sector, and their spread will be through market channels.

The high soil and water resource areas will continue to depend heavily on public sector international and national research programs for their genetic resources and nutrient and integrated pest management technologies. Recently, however, this international research sector has been under enormous financial constraints, brought about in part by the global recession, by alternative demands for resources, and by a global complacency about food sufficiency for the next decade or more. ¹² Within the high soil and water resource areas, considerably more location- and time-specific practices must be used to achieve more efficient pest and nutrient management and to reduce environmental loading. This can be accomplished by adding a much broader base of local area networking and farmer-to-farmer approaches to the more traditional public sector pathway. In addition, the technological inputs from private sector agribusiness must continue to grow.

The highly integrated systems of the lower productivity areas present a much greater challenge, and public sector institutions, especially the International Centre for Research in Agroforestry (ICRAF), have been reaching out to them. ¹³ One farmer-to-farmer, local community initiative in agroforestry has been described as “a new path.” ICRAF's Southeast Asia program is documenting the application of three types of agroforestry systems to serve as a basis for local adaptation by farmers. ¹⁴ A technology flow pathway based on farmer and community networking and empowerment is crucial to these highly integrated systems. Market forces as well as pricing policy must be favorable.

In contrast, scientific understanding of the highly diverse, very location-specific farming systems is extremely difficult. Their improvement without such knowledge is nearly impossible. In a farmer-participant research model now “fashionably” popular with scientists, farmers are considered members of the research team. They operate the system and work with scientists to collect and interpret descriptive data. This approach has had mixed success in both understanding the systems and changing them. Thus the value of this approach, as it has been practiced for the last 20 years, must be questioned. There have been few common themes and virtually no hypotheses suggested for systems structure or function, and the vast array of systems descriptions has not generated a body of knowledge useful to either researchers or farmers. The ICRAF Southeast Asia approach, based on hypotheses of systems structure, is exemplary in its innovation. ¹⁵

Any effective strategy for the agricultural development of the low soil and water resource areas must seek the participation of the public sector in (1) creating an enabling environment-public security, guaranteed access to land, and appropriate pricing and economic incentives; (2) assisting in the development of a progressive rural infrastructure-private rural institutions, a framework for farmer interaction, and a modest physical infrastructure; (3) evolving indigenous lead technologies in land and soil management, efficient use of water, and carbon husbandry; and (4) providing new technologies in such areas as tree production, animal husbandry, and high-value crops for market.

CONCLUSION

The greening of agriculture will require that its structure continue to evolve, with strong guidance and help from public policy and investment. Appropriate resources, incentives, and controls must be applied not to a single pathway but to the three overlapping development pathways. Far more emphasis must be given to the empowerment of local community groups, particularly in areas of complex and highly integrated farming systems. In so doing, their access to the scientific community must be strengthened. If indigenous knowledge alone were able to evolve quickly enough to meet changing social, economic, and environmental demands, the developing countries would not be facing such overwhelming problems. But at the same time, top-down solutions to these problems are illusory. Research needs will continue to increase for the near term, although, unfortunately, the public's interest in research is waning.

The technological breakthroughs and the adjustment of rapidly evolving production systems for greater sustainability will continue to require major public sector support, much of which will have to come from multilateral agencies.

The agricultural development future, with its plethora of options and the increasingly urgent demands of a rapidly growing population, is far more complex than it Was during earlier decades of the green revolution. But in spite of the challenges, there are many reasons for optimism.

NOTES

1. Rockefeller Foundation, 1993 Annual Report (New York: Rockefeller Foundation, 1993), 21.
2. P. Pinstrup-Anderson and R. Pandya-Lorch, “Alleviating Poverty, Intensifying Agriculture, and Effectively Managing National Resources,” International Food Policy Research Institute, Washington, D.C., 1994, 7.
3. P. Crosson and J. R. Anderson, “Resources and Global Food Prospects: Supply and Demand for Cereals to 2030,” World Bank Technical Paper No. 184, World Bank, Washington, D.C., 1992.
4. R. Lal, G. F. Hall, and F. P. Miller, “Soil and Degradation: I. Basic Processes,” Journal of Land Degradation and Rehabilitation I ( 1989): 5-69. To further explain what is shown in Figure 2, the productivity of agricultural soils is dependent on a layer of geologically-formed topsoil, which may range in depth from a few inches to several feet. Degradation of the topsoil can include loss resulting from wind, water movement, or landslide; compaction from heavy machinery use, particularly when the soil is wet; loss of plant nutrients and organic matter; or buildup of salt or other unwanted toxins. A resistant soil is often thick, flat in topography-and thus less subject to erosion - and rich in nutrients, allowing it to tolerate heavier and lengthier stress. Once such a soil is severely degraded, its agricultural productivity is nearly zero, like that reached sooner by marginal soils that have been abused.
5. These leaders successfully set forth a development paradigm that was adopted rapidly in a climate of extreme urgency stemming from a perceived imminent global food scarcity. See A. Mosher, Getting Agriculture Moving: Essentials for Development and Modernization (New York: Praeger, for the Agriculture Development Council, 1966).
6. United Nations, Agenda 21, Report of the United Nations Conference on Environment and Development, Rio de Janeiro (New York: United Nations, 1992).
7. Consultative Group on International Agricultural Research, A CGIAR Response to UNCED Agenda 21 Recommendations (Washington, D.C.: World Bank, 1992).
8. D. J. Greenland et al., “Soil, Water, and Nutrient Management Research-A New Agenda,” International Board for Soil Research and Management, Bangkok, Thailand, 1994.
9. D. P. Garrity, “Agroforestry: Getting Smallholders Involved in Reforestation,” International Centre for Research in Agroforestry, Bogor, Indonesia, 1994.
10. P. L. Woomer and M. J. Swift, eds., The Biological Management of Tropical Soil Fertility (Exeter, England: Tropical Soil Biology and Fertility Programme and Sayce Publishing, 1994).
11. R. R. Harwood, “Managing the Living Soil for Human Well-Being,” in Reinventing Agriculture and Rural Development, ed. S. A. Breth (Morrilton, Ark.: Winrock International, 1994).
12. R. O. Blake et al., “Feeding 10 Billion People in 2050: The Key Role of the CGIAR’s International Agricultural Research Centers,” Action Group on Food Security, Washington, D.C., 1994; Consultative Group on International Agricultural Research, A CGIAR Response; M. C. Ageaoili and M. W. Rosegrant, “World Supply and Demand Projections for Cereals, 2020,” International Food Policy Research Institute, Washington, D.C., 1994; M. C. Ageaoili and M. W. Rosegrant, “World Production of Cereals, 1966-199O,” International Food Policy Research Institute, Washington, D.C., 1994; and Pinstrup-Anderson and Pandya-Lorch, “Alleviating Poverty.”
13. International Centre for Research in Agroforestry, “Agroforestry for Improved Land Use: ICRAF's Medium-Term Plan, 1994-1998,” ICRAF, Nairobi, Kenya, 1993.
14. D. P. Garrity, “ICRAF Southeast Asia: Implementing the Vision,” International Centre for Research in Agroforestry,” Bogor, Indonesia, 1994.
15. Ibid.