Toward Sustainability: Soil and Water Research Priorities for Developing Countries (1991)

A collaborative, integrated research strategy will not develop without a conscious effort on a number of fronts. Changes are needed in the mechanisms and structures used by research organizations and funding agencies. Mechanisms need to be in place for various reasons:

• To link clients with researchers;

• To link components of research;

• To evaluate the effects of research;

• To set priorities as an ongoing task; and

• To deal with the problem of obtaining site-specific data to help solve site-specific problems.

A basic issue in targeting research to the needs of users is the pattern of communication and feedback. These chains of communication can be complex. Because a great deal of research relevant to the tropical context is conducted outside the areas of application, researchers in universities and in international organizations need to be linked with the users of the research in effective ways. Traditionally, crop research went through a hierarchial sequence from basic research to field testing to promotion by extension agents. But this structure does not work well in many developing countries and is not adaptable to soil and water management issues. Intermediate

agents who can transfer research findings to the field and who can also communicate farmers' needs to the researchers, increasingly important in the process of research and technology transfer. Private voluntary organizations and cooperative host country universities are institutions that are particularly important in this regard. Special efforts are required to encourage networks that link institutions in a two-way exchange involving researchers, research institutions, universities in host countries, private voluntary organizations, cooperatives, and farmers.

Many of the existing research structures will need to be modified to support the needs of sustainable agriculture and natural resource management. Currently, much agricultural research still is done within the confines of disciplinary boundaries or as dictated and supported by various commodity groups and state funding sources. The reward system for the researcher is based on some of the same constraints; for instance, professional recognition is tightly bound to within disciplines. Little acknowledgement is available for a researcher involved in interdisciplinary efforts. Trends away from rural sociology and cultural geography within the social sciences have provided fewer interfaces with agricultural disciplines on rural development problems. Likewise, efforts to bridge the natural resource disciplines and agricultural disciplines at universities have been weak at best, and almost adversarial in some cases. New efforts should

be made to change the administrative configurations of diverse groups of scientists working on common problems. The institutional reward system should reflect these changes. Rewritten mission statements, goals, and objectives would contribute to harmony among agriculturists, natural scientists, and social scientists working on common strategies for sustainable development. This new vision could help lead to more responsive technologies.

In summary, linking clients with researchers will entail a sensitive restructuring of the mechanisms and procedures now used by the research establishment. It is, however, the sense of the committee that awareness of this need has grown rapidly in the last few years and ways of enhancing better communication between researchers and research users will find a ready audience.

A second need is for the establishment of networks to facilitate the integration of different research components and to then focus these efforts on problems identified in the field. To be successful, such structures would require a strong client-driven component and assigned resources to facilitate the process. Many networks now exist that support the component approach to agricultural research in the tropics. Some of the better known ones focus on fast-growing nitrogen-fixing trees, biological nitrogen fixation, the winged bean, amaranth, and different soil types. The Collaborative Research Support Programs (CRSP) also support networks for their specific topics, for example, small ruminants, bean-cowpea, sorghum, millet, tropical soils, and other single issues. The International Agricultural Research Centers have also developed networks with National Agricultural Research Centers for different commodities and problems. These networks work because they link people with common interests and there is a framework within which they can operate. Networks that serve to draw components together to focus on more complex field problems are harder to develop but all the more needed. In fact, strong links between research organizations and users, as suggested in the preceding section, are a necessary element for successful networks (Table 4.1).

Client involvement is essential to the research priority-setting process. One example of a client-need-driven network is the Women in International Development (WID) network. The network was developed to focus more attention on gender issues in development, for instance, equal access to credit, land, and services. The network in the United States has membership dues structure, annual meetings, and publications. It also interacts with similar networks around the world and with a focal point in the United Nations. It grew from a number of initiatives in Third World and devel

oped countries. The network was not imposed on the members—in fact, its success is based in part on a comradery formed of common interests and problems. It is potentially self-supporting but attracts funds from foundations and governments.

A network similarly structured and motivated might help address the needs for soil and water research. Although the WID network is still in its infancy, it has already identified important research needs, has had significant impacts on projects, and has developed an impressive body of literature. A similar organization could help define research priorities, provide feedback on the applicability of research findings, and integrate soil and water research into efforts to solve other resource management problems.

An often neglected element in the setting of research priorities is the evaluation of existing research efforts to get feedback about the relative success of current directions. Evaluators should try to assess whether there is evidence of progress toward sustainability as a result of the research. While sustainability—a slippery concept, at best—is difficult to measure, a range of questions could be posed that could serve as indicators. For instance, evaluators should be able to ask whether the research is actually making some difference in the field. Are people doing or trying different things as a result of the research? If they are not changing, why not? Are soil and water properties improving? Are rates of degradation slowing?

Some mechanism is needed to allow evaluators to go into the field and assess impacts and effectiveness. It should be possible to evaluate the degree of dissemination of research results and at least speculate on the potential impacts in the field.

Many research programs include external evaluations of how the activities are administered.¹ However, in addition, field-level evaluations are also needed to ascertain the actual level of dissemination of the relevant research. These field evaluations should be carried out by interdisciplinary teams that include at least one social and one physical scientist. The teams should focus on analyzing farmers' awareness of the new information being generated and disseminated as well as on the apparent effectiveness of any new practices that farmers are using as a result of research. This evaluation of field studies would also provide useful feedback to help guide the evolving research process.

Although chapter 3 presented an agenda of research priorities as requested by AID, the committee unanimously agrees that research priorities will change, sometimes dramatically, over time, and the list only represents a first view of a longer research trajectory. Long-term benefits can accrue only if some mechanism is established to guide an ongoing evaluation and review of priorities.

Such priority-setting processes should involve a standing oversight committee that would evaluate existing field research and make judgments about the relative merits of the research efforts, in light of current needs and philosophies. Periodic workshops could be held, perhaps every three to four years, to assess progress and redefine priorities. A number of possibilities for such an oversight committee exist, such as a committee of the National Research Council, or an organization like the Board for International Food and Agriculture Development and Economic Cooperation. An alternative might be to involve the Collaborative Research Support Program, which was recently recommended (NRC, 1991b) that focuses on

sustainable agriculture and natural resource management.² A committee associated with that CRSP could be assigned the task of reviewing and revising research priorities.

When matching crop requirements to the characteristics of the land and its owner, agriculture is part art and part science. The success of agriculture depends on the degree to which the match between crops and land is successful, and one of the roles of the farmer is to rectify any mismatches in a way that is economically feasible and environmentally sustainable. In modern agriculture, the performance of a crop planted at a particular site and time of year generally can be judged if the requirements of the crop and the characteristics of the land are known. The two greatest influences on production are soil and weather. To predict production at a particular site, therefore, a minimum set of soil, crop, and weather data is required. (It bears repeating, however, that in Third World areas where labor availability, price fluctuations, and transport problems exist, such efforts have only modest predictive powers.)

For instance, in 1982 an international group of agricultural systems scientists met at the International Crop Research Center for the Semi-Arid Tropics in Hyderabad, India, to specify the minimum data set needed to predict crop performance in any agroclimatic zone. The participants used systems analysis and crop simulation models as the primary means to match crop requirements to land characteristics. The aim was to develop a foundation for dealing with the soil-plant-atmosphere continuum so that strong links could later be forged between the biophysical and socioeconomic processes.

The scope of work was limited to 10 food crops including 4 cereals (maize, rice, sorghum, and wheat); 3 grain legumes (dry beans, groundnut, and soybean); and 3 root crops (aroid, cassava, and potato). The report

defines the minimum data set needed to match crop requirements to land characteristics and serves as a guide for designing field experiments for model validation and model application (International Benchmark Sites Network, 1988). If a particular variety is found unsuitable to a farm, the crop genetic coefficients allow other crops and crop varieties to be examined. An economic assessment can be made for each management option as long as input costs and market prices are available. Although this minimum data set needs to be expanded to include a wider range of soil and water characteristics such as data needed to assess salinity, sodicity, metation toxicities, and nutrients such as phosphorus, calcium, magnesium, sulfur, potassium, and trace elements, it is an example of one careful attempt to improve our capability to identify appropriate crops for specific sites.

Dynamic, process-based crop-environment simulation models can be developed by interdisciplinary teams. These models are generic and are designed to operate anywhere in the world. The minimum set of soil, crop, and weather data enables the models to generate site-specific information for decision support. Models focused on human systems also offer potential, and the process of designing them could help delimit the critical human behaviors and beliefs that influence agricultural decision making and practice. The collection or compilation of the minimum data set for systems simulation should be the responsibility of the user nation. Countries unable to establish a data base for their soil and climate will require assistance from development agencies. The compilation of genetic coefficients should be the responsibility of plant breeders (Hunt et al., 1989). Although these types of efforts are certainly not panaceas for bringing about sustainability, they are useful tools in any broad effort to develop and implement better agricultural management.

Promoting sustainable agriculture and the wise use of soil and water resources will require more than scientific innovation. Any integrated research strategy will need to find ways to enhance communication among the different actors, from researchers to farmers. As we have learned from experience in the U.S. land-grant system, research alone cannot be expected to carry the load of agricultural development and natural resource conservation. The adoption of new policies, technologies, and farming systems will require social change—the kind of change that can be encouraged by effective communication. Certain key communication issues must be considered (Caudle, 1990):

1. Those who most urgently need information do not all speak the same language—literally or figuratively. They do not see resources the same

To be sustainable, agricultural research strategies must include dynamic interaction and communication between scientists and farmers. These Kampuchean rice farmers are taking a training course at the International Rice Research Institute in the Philippines. Credit: International Rice Research Institute.

way, they do not enjoy equal access to productive resources, and tenure and access rights are mediated through complex social relationships of kin, cohort, and obligation. They number in the hundreds of millions, and their cultures, levels of education, needs, and expectations are exceedingly diverse. These audiences, often illiterate, cannot always be reached through the traditional channels used to communicate with farmers in the United States because extension systems in developing countries often are inadequate. These farmers, many of whom are women, do not always toil individually. They often are members of different kinds of institutions ranging from extended families to peasant leagues that could be used as vehicles for transmitting information.

3. Information about agricultural technologies and soil and water resources primarily consists of formal research results from individual projects or experiments and vast informal practical knowledge maintained and transmitted by farmers. While the "language of science" may be universally understood in the community of researchers, the ways local knowledge is encoded and transferred is tied to local idioms and institutions. To provide a place for these formal and traditional knowledges to meet will require scientists to make the effort to understand their clients. Many well-intentioned scientists have gone about their work assuming that their clients

are basically similar to their U.S. counterparts. Cultural, economic, class, and gender differences that affect the utility of research were often overlooked, and the logic of existing practices was often misunderstood. Any strategy for sustainability must rely in the long run on a much more dynamic interaction and communication between scientists and farmers. Special mention should be made of the importance of women farmers and farming systems, so that research strategies include programs that substantially address their concerns.

There are several means for achieving these newly focused efforts, including careful analysis of local societies and economies so that how decisions are made can be understood in light of gender concerns, economic strategy, and household strategy. Another means for promoting this holistic approach is the use of computer-based decision-support systems, including simulation models and expert systems. These systems not only have the potential to store, organize, and access huge amounts of data, they also provide the opportunity and impetus for applying scientific information.

Whatever the medium of communication, we should recognize that disciplinary specialization is not always the ally of development. In some cases, this specialization has led to a confusion of tongues. When we approach the people of a foreign country and offer them our help, we must all speak the same language: theirs. Although the importance of using local languages seems obvious, it is too often overlooked when time and money are short.

3. In an era of intense competition for diminishing public support, there is a premium on "success stories"—on results that may be used to tout a program to constituents and thereby increase its funding or prolong its life. Unfortunately, the results of soil and water research typically do not lend themselves to the development of simple, easily understood "products." Such research tends to produce incremental and sometimes subtle improvements, not the dramatic breakthroughs that attract publicity. Agencies should recognize that in the world of science communication, not all subjects are created equal. While investments in medical science regularly pay off with dramatic stories and media attention, investments in soil and water science may never yield a single headline. Professionals could advise program managers and funding agencies on alternative strategies for public relations.

   Most of the resources now expended on communications in agriculture and natural-resource programs are devoted to "in-house" publications such as newsletters and project reports, and the bulk of these are directed at the funding agency to meet documentation requirements. Typically, communications professionals are involved only in the final production of a publication, not in its planning or development. For almost every program there is a bale of brochures, reports, papers, proposals—many of them unread, ill-conceived, and ineffective in the cause of helping developing countries.

Communications dollars should be spent on products that will put information in the hands of people who need it, in a form people can use. This means placing professional communicators on staff, and making them integral to the operation, using their input in all aspects of program development and management. It does not mean hiring an occasional consultant. Agencies should have one staff communicator review and comment on every program or project proposal.