Document(s) 22 of 29

Introduction

Farmers are usually perceived as uneducated, conservative, superstitious, and illogical people who need instruction in effective farm management for increased agricultural production. Extension agents adopt the role of expert advisers, and breeders develop and release hybrid seed varieties without considering the farmer's seed-selection practices or other local knowledge of production systems. Dependency theories have been propounded to explain the reasons for this arrogant superimposition of the international community and, by implication, their agents (scientists, planners, and implementers) on peripheral peoples. These theories, however, portray local people as passive recipients unable to resist attempts by the core to obliterate local cultural practices and impose their supposedly advanced forms of social organization and technological development (Dos Santos 1969; Bernstein 1973; Todaro 1977).

What modernists fail to realize is that no society wholly denies its cultural values or environmental orientation to fully embrace the values, norms, technological innovations, and social institutions of another culture. This fact was aptly summarized by Apter (1965, p. 8), who noted that "cultures never give way completely to the new, no matter how ruthless the impact of innovation."

Many traditional societies draw insights from the knowledge, cultural values, practices, and perceptions learned and passed on through the generations (Thrupp 1989). They selectively adopt innovations, according to their needs, while adapting or rejecting those that do not fit into their cultural orientation (Rogers 1983; Thrupp 1989). Indeed, contemporary anthropological studies argue that the farmer's so-called conservative or backward practices are often rational responses to local conditions and are logical adaptations to perceived risks, based on practical experience.

This study examines the adaptation of new agricultural technology by a subsistence-oriented rural community in the semiarid regions of Kenya. The implications of the national Sorghum and Millet Improvement Programme (SMIP) for the Akamba community of Kitui District, in Eastern Province, and its seed-selection and seed-adoption practices are analyzed. Emphasis is placed on the farmers' adaptive responses to the program, which result from a combination of indigenous knowledge and selective adoption of innovation to suit their environment and particular sociocultural and economic circumstances.

Theoretical assumptions

the particular forms of technology that develop and bed into society are shaped by cultural meanings and social negotiations. Equally, the experience of technology and its impact on everyday life are shaped according to cultural meanings that reside within the wider society.

The second assumption is that farmers' reactions to agricultural innovations depend to a large extent on environmental factors. For subsistence farmers, who live in marginal environments characterized by harsh climatic conditions, limited and erratic rainfall, poor soil nutrients, perpetual crop failures, and chronic famine, a bad crop would not imply simply that some land or livestock has to be sold to provide the much needed subsistence. In extreme cases, it would also mean starvation and death (O'Leary 1980). This is consistent with the theory of the "subsistent ethic" of Myint (1969), Scott (1976), and Ellis (1988), who argued that the cost of failure for farmers near the subsistence margin is such that safety and reliability take precedence over long-term profit. The fear of food scarcity provides justification for the farmers' holding on to their traditional practices, rather than gambling with innovations and plunging their families into starvation and death. In the event that they adopt new technologies, they do so cautiously, marrying tradition and modernity and selecting only those items that best suit their circumstances.

Rationale for the project

As a result, the Government of Kenya, in collaboration with the Kenya Research Institute and its various research stations, established the national SMIP in 1978. This program developed and released improved sorghum and millet seed varieties for adoption by farmers in the Eastern Province in the early 1980s. The improved seeds were early maturing and drought and pest resistant, had acceptable food-processing qualities, and could withstand fluctuating environmental conditions. The technological package also included complementary methodological procedures, such as improved farming techniques (e.g., row planting, timely weeding, and terracing), use of fertilizers and pesticides, and improved harvesting and post-harvest techniques.

Kitui District, which is one of the areas where the program was implemented, is the largest district in the semiarid zones of Eastern Province. The district has a total land mass of 6030 km² and a population of more than 700 000 (Republic of Kenya 1979–83). Because most of the district lies within ecological zones IV, V, and VI, which are the areas with the lowest potential in Kenya, it is drier than other semiarid regions in the country. Thus, the other semiarid regions have some choice in the crops they grow, but Kitui District has none — it must grow sorghum, millet, and other dryland crops or face starvation. Agricultural-development implementers, therefore, felt that SMIP would result in an overall increase in food production in the area.

At the time SMIP was implemented, the farming systems of the Akamba community were predominantly subsistence oriented. Emphasis was placed on family labour at the expense of hired labour. Division of labour was based on sex, age, and status. Rudimentary technology, rather than mechanized farming, was used. Traditional farming systems predominated: seed-selection practices, land preparation, planting, weeding, and harvesting were based on indigenous technical knowledge accumulated over generations. Although the community was fully monetized, the farmers usually ignored the neutral and impersonal mechanisms of the money economy, preferring to use kinship networks, reciprocal exchange systems, communal structures, supernatural sanctions, and rituals to organize their production, consumption, and distribution of agricultural products.

SMIP, therefore, aimed at transforming the production system by promoting hybridization, modern farm methods, and the use of farm inputs, which are all components of the green-revolution model. Within this modernistic approach, indigenous knowledge and farmers' experiences and perspectives on the environment were considered obstacles to development and were omitted during the design, planning, and implementation of the program.

Methodology

Farmers' adaptive responses to the program

Table 1. Sorghum seed varieties grown by respondents.
	Respondents (n)	Rate (%)
Hybrid varieties Mutune Serena 2k × 7	50 82 97	7.2 11.7 13.9
Traditional varieties Kaveta Katengu Other local varieties	148 148 173	21.2 21.2 24.8
Totala	698	100.0
a There were only 415 respondents, but some grew more than one variety.

From Table 1, it is evident that the traditional sorghum varieties were preferred (67.2%). However, it was difficult to clearly distinguish the adopters of hybrid seed from the nonadopters because all the adopters also grew traditional varieties. Their reason for growing both kinds was that the local varieties and the hybrid varieties had different characteristics, so they would be avoiding the risk of total crop failure (usually caused by lack of rain fall and by other environmental hazards). The choice and use of particular seed varieties largely depended on their compatibility with the existing norms, their environmental adaptability, and the farmers' cultural perception. The farmers considered taste, colour, storability, grain size, processing qualities, weight (whether the flour is filling when eaten), and resistance to pests and drought. These selection criteria led to the continuous use of katengu, a traditional sorghum seed that is highly palatable, is easy to prepare, has good processing qualities, and is white, a characteristic that makes it preferable to red sorghum varieties. Kaveta, another traditional variety, was mixed with salt and eaten raw. It was also harvested and eaten before it was completely ready. This seemed particularly important in this area, where food scarcity is a chronic problem and alternative foods are rare. It was also believed that the nutritious qualities of kaveta increased breast milk in lactating mothers.

The unpopularity of hybrid seeds was due precisely to their lack of the qualities that farmers preferred in traditional varieties. Only 82 of the respondents adopted serena seed. Despite its high-yielding qualities and resistance to birds, serena seed was largely rejected because of its red colour, which was offensive to the eye, according to the local community, and had a bitter taste as a result of its high tannin content. Those who adopted it did so for marketing purposes rather than for consumption. Adopting it for marketing would appear rational to modernists and agricultural scientists, but for many local farmers this was not an option because food scarcity, starvation, and death were prevalent in the area. The first consideration, therefore, was to attain food security; the surplus could be marketed later. In fact, findings reveal that those who grew serena seed in abundance had a higher socioeconomic status and could afford to buy alternative food for consumption.

The only hybrid seed that compared favourably with local varieties was 2k × 17. It was white, palatable, used in a variety of local recipes, early maturing, and drought resistant. It had the additional advantage of producing higher yields than some of the local varieties and, therefore, could be successfully marketed. However, farmers complained about stomach upsets when it was poorly processed. It was also highly susceptible to birds, a common problem with all white varieties of sorghum.

Table 2 shows the millet seed varieties grown by the respondents. The limited adoption of bulrush millet (14.4%) resulted from the farmers' inability to eat it raw, which was due to its spiky covering. It was also more susceptible to weevils when in storage than the local varieties. The local millet seed variety, referred to as mwee wa kikamba or ute wa makova, was grown for the same cultural reasons as kaveta. It was also preferred (69.7%) over the hybrids because it was more pest resistant and could be stored for several years without any pest infestation.

Table 2. Millet seed varieties grown by respondents.
	Rate (%)
Mwanza Nguunu (undefined variety) Bulrush Local variety	7.3 8.6 14.4 69.7
Total	100.0

Questions relating to sources of information about hybrid seed revealed that only 19.9% of the respondents who usually had direct contact with the scientific agents had learned about hybrids from them. The rest obtained information informally from churches, nongovernmental organizations (NGOs) working in the area, friends, or relatives. None of the respondents indicated having had any contact with the Katumani Sorghum and Millet Breeding Research Station, whose area of operation and testing sites covered the district. The farmers indicated no knowledge of collaborative efforts between other farmers and research station or extension staff working locally to produce hybrids for multiplication and distribution. This finding seemed to be consistent with information received through interviews with the staff of the Katumani Sorghum and Millet Breeding Research Station, who indicated that before any seeds are released to farmers for adoption, breeders develop them at the research station. Cluster farmers are then identified in the station's area of operation, and farm trials are undertaken to obtain feedback from these farmers. Once the required characteristics are developed in a seed variety, the Farming Systems Section collaborates with extension staff, who in turn transmit the information to local farmers. The fact that none of the respondents had had any contact with the breeding station suggests that either they were not cluster farmers or they had not heard of the stations. This meant that the cluster farmers were few and far between or that the rest of the farmers had little contact with extension staff and, therefore, were unaware of research activities relating to seed development, multiplication, and distribution.

For more than 60% of the respondents, the most reliable source of seed was themselves: they selected healthy, untainted seeds from the harvest and preserved them, usually by the smoking method or by applying ash and pepper for pest resistance (Table 3). Seed selection was largely a specialty of the older women, who would gradually teach this to the eldest daughters-in-law, to be used when the elders were either away, too old, or dead.

Table 3. Seed preservation measures.
	Respondents (n)	Rate (%)
Ash and mitaa (traditional herb) Ash Nothing Pesticides Ash and pepper	47 60 69 89 150	11.3 14.5 16.6 21.5 36.1
Total	415	100.0

Cost and availability of the hybrid seed also played a major role in determining farmers' choice and adoption of a particular seed. More than 50% of the respondents could not afford the price of seeds available in the shops; the rest noted that seed was unavailable at the shops and other distributing centres during planting season. Instead, they depended on sporadic sources, such as NGOs working in the area, government personnel, chiefs, churches, neighbours, relatives, and others in the kinship network. Family affiliations and friendships played an important role in ensuring that seeds were available to those who could not get any through the reciprocal-exchange systems. Seed was freely given and received, and many farmers were without any specific reason for planting a specific seed variety, except that it was available or was given to them by a friend, relative, or neighbour. The unavailability of hybrid seeds reduced the choice and encouraged the farmers to continue planting traditional varieties or to replant hybrid varieties.

Qualitative information obtained from focus-group discussions revealed a curious event, which resulted from repeated replanting of hybrid seeds. Questions about the significant differences between repeatedly replanted hybrids and traditional varieties revealed that traditional varieties retained their original characteristics, but the seed of the hybrids shrank, its stalk grew taller, and its yields decreased. For serena seed, the red colour faded into white, making the seed attractive to birds. The spikes of bulrush millet fell off, also exposing the seed to birds. In time, the farmers saw no difference between their traditional seed and the hybrid seed.

It is clear that seed adoption cannot be encouraged without making the seeds available on a continuous basis and providing adequate information about them. Furthermore, one cannot assume that after the farmers obtained hybrid seed, they would use modern farm methods to plant, weed, and harvest them, which would be needed to ensure increased yields. On the contrary, the type of farm operations chosen, whether traditional or modern, depended on the availability of resources, such as farm inputs and modern farm equipment, and on the farmers' awareness of modern storage techniques. If the farmers did not have the resources or lacked the scientific knowledge they needed to use modern techniques (as was usually the case), they would broadcast the hybrid seed or plant it in the same hole with traditional seed to minimize the risk of losing both seed crops if the rains failed. Often, they also used traditional weeding methods, pest-control strategies, and storage techniques for hybrid seeds. Obviously, this idiosyncratic use of planned agricultural practices made it difficult to determine the impact of hybrid seeds on overall crop yields. As a matter of fact, the complexities resulting from the interplay between technology and human action made it quite difficult to directly correlate the adoption of hybrid seed with the expected overall yields in the long run.

The mixture of traditional and modern farm operations by the farmer could also be indicative of the weak extension–farmer linkage evident in the area of study. Although 67.2% of the farmers knew that there was an extension agent working in their locality, only 46.3% had met him. However, the majority (60%) had only met him once in 6 months and often in public forums rather than on their farms. The implication here is that many of the farmers who adopted hybrid seed had to rely on information from neighbours, NGOs, friends, and other sources or fall back on the traditional practices used for generations.

Strategies for sustainable agricultural development in arid regions

In Kenya, research directors and their staff typically set up research priorities with the expectation of feedback from farmers through farm trials and extension input. However, as Thairu (1985) and Audi (1985) rightly noted, the research agenda is set by the plant breeders, but rarely in collaboration with the farmers. This weak linkage extends to the interaction between research stations and extension personnel: financial constraints and other logistical problems prevent them from holding regular workshops and other forums of communication for disseminating research findings (Audi 1985). Extension–farmer linkage faces a similar setback because of chronic problems that plague the Kenyan Ministry of Agriculture, such as few extension personnel (estimated at one for every 500 farmers), limited transportation available, lack of incentives for staff, and a bureaucratic structure that stifles staff initiative at the local level (Uma Lele 1975; Chitere 1980; Ogum 1982; Benor and Batter 1984; Oduol 1991).

The obvious way to overcome these weak linkages is to change the agricultural-development approach from one that is top down to one that is bottom up, from centralized standardization to local diversity, and from blueprint to learning process (Chambers 1980). The adoption of the bottom-up approach by planners would require that the research agenda be determined by farmers' preferences, rather than those of the experts. This means that a needs assessment of the farmers' circumstances should be undertaken before hybrid seeds are developed, rather than developing the seed first and then testing it with farm trials. The major advantage of the bottom-up approach is that farmers' local knowledge of crop-development practices and their adaptation of farming practices to ecological and genetic diversity would be harnessed to supplement scientific knowledge for sustainable crop development. Linked with this is the need to consider farmers' seed-selection criteria and their effectiveness, as well as the sociocultural and economic setting in which farming practices are undertaken. This knowledge becomes particularly important where hybrids are broadly adaptive and uniform but rarely sustainable in the long term, particularly in arid environments (Berg 1993).

Although a lot of indigenous knowledge is presented too generally by social scientists to be of much use to plant breeders, this knowledge can make a big difference in agricultural development if harnessed by a multidisciplinary team of scientists, social scientists, and farmers. Through their combined efforts, they could obtain information on consumer preferences, as well as the sociocultural, economic, and ecological characteristics of the plant environment. Prain (1993) provided the example of an interdisciplinary team that successfully explored ways of collecting and documenting sweet potato germ plasm and associated indigenous knowledge in western Java. Dorp and Rulkens (1993) also described how the combination of a multidisciplinary team, participatory rural appraisals, and villagers with local knowledge of seed-selection criteria provided information that was then integrated into breeding programs in Indonesia. Worede and Mekbib (1993) noted similar experiences in northeastern Shawa and southeastern Walo, where farmers maintain crop diversity and improve genetic performance with the assistance of breeders and other scientists. Such examples clearly illustrate the importance of taking a participatory approach in agricultural development programs in the Third World to ensure sustainability.

The weak extension–farmer linkages could be strengthened if the Kenyan central Ministry of Agriculture extension agents networked with the NGOs working in the rural areas. One advantage of networking with local NGOs and churches is that they already have extension programs and resources that facilitate their interaction with local populations. Because the Ministry of Agriculture lacks precisely those facilities needed for effective work, it could form collaborative networks with NGOs and provide the additional skills, experiences, and expertise that local NGOs may require to strengthen their extension programs. The extension training component of the Kenya Energy and Environmental Organization and the efforts of World Neighbours help to strengthen farmers' capacity to analyze agricultural problems and make appropriate changes (Wellard 1993). As well local farmers chosen by the local community could be trained to provide extension services. The use of local farmers chosen by the community could help alleviate problems with contact farmers, who are isolated by local communities because they are identified with extension personnel and, by implication, with the administration, rather than with the villagers.

The study also revealed that agricultural development is implemented selectively, rather than in an integrated and holistic manner. Although hybrid seeds are developed for adoption, adequate plans are not made to ensure they remain available to farmers, and the purchasing power of the farmers and other economic constraints on farmers' access to seed are not taken into account. This conclusion was reinforced by Oduor (1991), who revealed that most of the improved varieties are unavailable at the Kenya Grain Growers Cooperative Union, one of the main seed-multiplication centres, because most of the projects that used to produce seeds in bulk are no longer operational. Furthermore, the Kenya Seed Company, the major commercial distributor of seeds, may not be stocking the shops and other retail outlets effectively in arid regions because of their remoteness and inaccessibility from cosmopolitan centres (Oduor 1991).

The difficulties agricultural implementers face in seed multiplication and distribution leads us to explore the possibilities of using an important but previously ignored resource — the farmers themselves. Case studies abound with examples of local communities using available resources to adapt to a quickly changing environment (Brokensha et al. 1980; Chambers 1980; Thrupp 1989). The challenges experienced in difficult environments such as arid regions are usually dealt with through group formations based on kinship affiliations, family networks, friendships, and neighbourhood linkages. Communal linkages are used as moral- and social-support systems in labour camps and in development-project activities, such as road construction, bridge building, and soil conservation. The Mwethya groups among the Akamba, the Saga among the Luo, and the Bulala among the Luyia communities serve to illustrate this point.

The local capacities that have been eroded by interaction between local communities and the international market economy should be rebuilt and strengthened for sustainable agricultural development. Although traditional breeding systems could certainly benefit from scientific input, they need not be wholly replaced by modern technology. Instead, a participatory breeding program, with farmers, plant breeders, and scientists identifying appropriate seed-selection criteria, along with seed development, distribution, and multiplication, would go a long way in empowering local communities to be self-sustaining in seed production. In their study of a participatory approach to breeding for improved varieties, Muarya et al. (1988) recommended a similar adaptation strategy for Zimbabwe. They argued that for small-scale farming systems to become the central source of crop genetic diversity, the systems must incorporate four steps:

1. Selection of germ plasm by breeders in consultation with farmers
2. Preliminary on-station selection, based on criteria identified by farmers, and breeding for improved varieties, with frequent feedback from farmers via agricultural extension agents
3. Release of improved varieties for evaluation and selection by farmers
4. Coordination of farmers' evaluation by agricultural scientists and breeders and use of the farmer-selected material for subsequent trials

In conclusion, this paper analyzes the adoption strategies developed by Akamba farmers in response to new agricultural technology. The major finding of the study is that farmers largely persist in using traditional seed varieties because agricultural scientists and extension agents hardly consult with them before developing hybrid seeds. The farmers are also not adequately involved in hybrid-seed selection, production, multiplication, and distribution, and this results in low adoption rates.

The conclusions derived from this study are that the deterministic and top-down approach used by agricultural experts to transmit information ignores the fact that farmers also influence technological processes and their development cycle. This omission results in the development of inappropriate technologies and policies during the design, planning, and implementation of agricultural programs. Sustainable agricultural development can only occur if a participatory approach, involving plant breeders, extension agents, social scientists, and farmers, is used in future program activities.

Acknowledgments

References

• Apter. 1965. The politics of modernization. University of Chicago Press, Chicago, IL, USA.
• Audi, P.O. 1985. Training and visits (T&V) and development of small scale and subsistent agriculture in the semi-arid areas of Kenya: A Katumani experience. Presented at the Workshop on Agricultural Research and Extension for the Drylands of Kenya, 2–6 Dec. 1985, Embu, Kenya.
• Benor, O.; Batter, M. 1984. Training and visits extension. World Bank, Washington, DC, USA.
• Berg, T. 1993. The science of plant breeding: Support or alternative to traditional practices? In de Boef, W.; Amanor, K.; Wellard, K.; Bebbington, A., ed., Cultivating knowledge: Genetic diversity, farmer experimentation and crop research. Intermediate Technology Publications, London, UK. 72–78.
• Bernstein, H. 1973. Underdevelopment and development: The Third World today. Penguin Books, Harmondsworth, UK.
• Brokensha, D.; Warren, D.; Werner, O. 1980. Indigenous knowledge systems and development. University Press of America, Washington, DC, USA.
• Central Bureau of Statistics. 1980. Statistical abstracts: Nairobi. Government Printer, Nairobi, Kenya.
• Chambers, R. 1980. The small farmer is a professional. CERES, 13(2).
• Chitere. 1980. Decentralization of decision-making process: Its implications for the implementation of the Crops Improvement Programme in Kenya. University of Nairobi, Nairobi, Kenya. PhD dissertation.
• Cooley, M. 1987. Architect or bee? The human price of technology. Hogarth Press, London, UK.
• Dorp, M.; Rulkens, T. 1993. Farmer crop-selection criteria and genebank collections in Indonesia. In de Boef, W.; Amanor, K.; Wellard, K.; Bebbington, A., ed., Cultivating knowledge: Genetic diversity, farmer experimentation and crop research. Intermediate Technology Publications, London, UK. 119–128.
• Dos Santos. 1969. The crisis of development, theory and the problem of dependence in Latin America. Sigho, 21.
• Ellis, F. 1988. Peasant economics: Farm households and agrarian development. Cambridge University Press, Cambridge, UK.
• Ellul, J. 1962. The technological order. Technol. Cult, 394–421.
• Hill, S. 1988. The tragedy of technology. Pluto Press, London, UK.
• Hyden, G. 1983. No short cuts to progress: African development management in perspective. Heinemann, London, UK.
• Inkeles, A. 1973. Making men modern: On the causes and consequences of individual change in six developing countries. In Etzioni, A.; Etzioni, H., ed., Social change: Sources, patterns and consequences. Basic Books, New York, NY, USA.
• Langdom. 1975. Technology transfer by multinational corporations in Africa: Effects on the economy. Presented at the 18th annual meeting of the African Studies Association, 19 Oct. – 1 Nov. 1975, San Francisco, CA, USA.
• Maurya, D.H.; Bott Val, A.; Farrington, J. 1988. Improved livelihoods, genetic diversity and farmer participation: A strategy for rice breeding in rainfed areas in India. Experimental Agriculture, 24(3), 311–320.
• Moore, W.E. 1963. Social change. Prentice-Hall, Englewood Cliffs, NJ, USA.
• Myint, H. 1969. The peasant economics of today's underdeveloped areas. In Wharton, C.R., ed., Subsistence agriculture and economic development. Alden, Chicago, IL, USA.
• Oduol, W. 1991. Planned social change: A case study of agricultural extension officers in South Nyakach, Kisumu District, Kenya. University of Nairobi, Nairobi, Kenya. MSc dissertation.
• Oduor, J. 1991. Constraints to production of sorghum and pigeon pea and strategies to increase production in Eastern Province. Presented at the International Crops Research Institute for the Semi-Arid Tropics meeting, 18–19 Sept. 1991, Nairobi, Kenya.
• Ogum, J.L. 1982. Modernity and adoption of farm innovations: The case of agricultural modernization in Torit District, southern Sudan. University of Nairobi, Nairobi, Kenya. MSc dissertation.
• O'Leary, M. 1980. Responses to drought in Kitui District, Kenya. Disaster 4(3), 315–327.
• Pfaffenberger, B. 1988. Fetished objects and humanised nature: Towards an anthropology of technology. Man — The Journal of the Royal Anthropological Institute, 23(2), 236–252.
• Prain, G. 1993. Mobilizing local expertise in plant genetic resources research. In de Boef, W. Amanor, K.; Wellard, K.; Bebbington, A., ed., Cultivating knowledge: Genetic diversity, farmer experimentation and crop research. Intermediate Technology Publications, London, UK, 102–111.
• Republic of Kenya. 1979–1983. Kitui District development plan. Government Printers, Nairobi, Kenya.
• Rogers, E. 1983. Diffusion of innovations. 3rd ed. The Free Press, London, UK.
• Scott, C.J. 1976. The moral economy of the peasant. Yale University Press, New Haven, CT, USA.
• Thairu, D. 1985. Research and extension liaison with special reference to transfer of technology. Presented at the Workshop on Agricultural Research and Extension for the Drylands of Kenya, 2–6 Dec. 1985, Embu, Kenya.
• Thrupp, A. 1989. Legitimizing local knowledge: Scientized packages empowerment for Third World people. In Warren, M.; Slikkerveer, J.; Titlolo, O., ed., Indigenous knowledge systems: Implications for agriculture and international development. Iowa State University, Ames, IA, USA. Studies in Technology and Social Change, No. 11.
• Todaro, M.P. 1977. Economics for a developing world: An introduction to principles, problems and policies for development. 2nd ed. Longman, London, UK.
• Uma Lele. 1975. The design of rural development: Lessons from Africa. The Johns Hopkins University Press, Baltimore, MD, USA.
• Wellard, K. 1993. Linking local knowledge systems and agricultural research: The role of NGOs. In de Boef, W.; Amanor, K.; Wellard, K.; Bebbington, A., ed., Cultivating knowledge: Genetic diversity, farmer experimentation and crop research. Intermediate Technology Publications, London, UK. 136–143.
• Worede, M.; Mekbib, H. 1993. Linking genetic resource conservation to farmers in Ethiopia. In de Boef, W.; Amanor, K.; Wellard, K.; Bebbington, A., ed., Cultivating knowledge: Genetic diversity, farmer experimentation and crop research. Intermediate Technology Publications, London, UK. 78–85.

Document(s) 22 of 29