Substantial funding has been directed towards building up formal seed supply organizations in developing countries - both parastatal bodies and private companies. But it is becoming very clear that often a much more effective network of informal seed diffusion based on farmer-to-farmer seed exchange exists. Elizabeth Cromwell of London's Overseas Development Institute reports on the findings of some recent research.
Mary Mphande willingly laid down her hoe to talk to us when we asked about her seed problems, but she was clearly becoming more and more irritated as we worked through our questions about the national seed company's service. "You 're surprised I don 't get all my seed from the company," she interrupted eventually, "when their new varieties can yield so much better than my local seed, and the price is quite cheap at the moment. But let me explain". Mary told us how she liked having the high yields from the new varieties, so she could sell some of her harvest to pay her children's school fees, but in years of poor rains she couldn 't rely on getting any harvest at all -- and using the new varieties meant she had to plant them all early, just at the time when she wanted to concentrate on her maize garden. Neither did they cook as nicely as her local varieties, she added, and, most importantly, she could never rely on finding seed at the company's depot when she took the time to make the long journey there. Why should she take all these risks, she asked, when she could save her preferred local seed quite easily at home, and she could always find neighbours with some seed to spare if she ran short.
Mary Mphande is just one of the farmers ODI interviewed in Malawi last year, but her story is repeated all the time throughout the developing world. Instead of trying to improve the service national seed companies provide to small farmers, why isn 't farmer-to-farmer seed saving and diffusion being supported? After all, in most countries farmer diffusion systems are well established, easily locally accessible and they deal with the seed varieties farmers actually want: less than one third of developing country farmers have ever used seed provided by national seed companies.
Small farmers and the formal seed industry
Since the seed sector became the target of widespread official support by Western donors and Third World governments in the late 1960s attention has been focused on formally organized national seed programmes. Activity has been widespread; the FAO Seed Improvement and Development Programme has spent $80 million on 120 projects in sixty countries and the World Bank funds 13 national seed projects and 100 other seed-related projects. But despite this substantial funding, the seed needs of small-scale farmers have rarely been met. Where the organization of the seed sector has been reexamined, particularly in the last decade of emphasis on IMF-style structural adjustment, the most common option considered has been the privatization of the seed companies. But when private companies have responsibility for seed supply, the service provided to small farmers is frequently even worse because of the conflict between generating profits and serving small, poor farmers with the special varieties and small quantities of seed they require. Some companies in Zimbabwe and India, for example, refuse to supply the small farmer market for some crops at all.
Private and national seed companies share several shortcomings. Firstly, they usually produce a limited range of modern high potential yield varieties, often hybrids, because this is the cheapest production strategy for them and because it guarantees regular replacement of seed by farmers. Small farmers, though, do not want to replace seed regularly, and they cannot afford the fertiliser which modern varieties need to yield well. In fact, farmers may not need modern varieties at all but simply local seed that is of better physiological quality. But the seed distribution chains of seed companies are often very long and this means the companies ' seed is often badly weevilled or rotten by the time it reaches farmers. Timely delivery of seed is also very important for small farmers because they do not have sufficient cash flow to buy seed in advance of planting and because without irrigated water or mechanical cultivators, they must plant at the optimum time. But in practice, seed companies rarely have enough resources to make sure that all the large number of often remote small farmers are reached with seed on time.
Increasing small farmers ' access to seed provided by companies -- national or private -- seems to be neither feasible in the medium term nor helpful in contributing to more sustainable farming systems in the long run. More and more evidence shows that small farmers ' own community-based seed diffusion mechanisms are very finely tuned, well-used, independent and sustainable, and they provide a far more viable means of getting better seed of both traditional and modern varieties to farmers in the quantities they require, on time and in ways which they can afford. And at the same time, the quality of seed reaching farmers in this way is comparable to anything the national seed companies can produce.
Farmer-to-farmer seed diffusion
In the Eastern Hills of Nepal "pokhreli masino" is a traditional rice cultivar of unknown origin which had been actively distributed by the Pakhribas Agricultural Centre through the formal extension service. It produces good grain and straw yields, is early maturing and has proved to be much more popular with local farmers than the available modern rice varieties: the area sown to this variety between 1976 and 1986 increased 50-fold. But, although the Agricultural Centre played a crucial role in initially making "pokhreli masino" available to the farmers, 94% of farmers obtained the variety not from the "mini-kits" promoted by the Centre but from other farmers. For over two thirds of the farmers, this takes the form of seed exchange, without any cash changing hands and so the seed was not restricted to the richer farmers but became available to a broad range of socioeconomic groups. A few key individuals who have a personal commitment to their community are of particular importance in promoting the spread of new seeds. This example clearly shows that farmer-to-farmer seed exchange can be a highly effective means of seed diffusion even in remote hill regions with poor communications infrastructure.
In Kenya's Eastern Province, farmers have considerable knowledge about on-farm selection, storage and treatment of seeds and in the past there was little effort at organized seed production and distribution. However, new seeds brought in after periods of drought have been widely and rapidly spread by the informal farmer-to-farmer system. The "Serena" sorghum variety brought in after the 1984 drought for example, is now regarded by farmers as "local", and has been used in on-farm selection to improve the local seed stock. Since the mid 1980s the formal sector has adopted a more coordinated approach to produce and distribute new varieties of maize, small grains and food legumes through the "Dryland Farming Project". This has not been very successful due to the high cost, poor quality control and low demand for modern varieties which are unproven in the local farming systems. However, close study has shown that farmers do use new varieties, but they subject them to rigorous on-farm experimentation and observation before deciding whether to incorporate them into the cropping system. Informal, farmer-to-farmer seed diffusion mechanisms seem to work effectively in meeting their needs. Most farmers want new varieties primarily to experiment with and to multiply up themselves; they do not feel the need for the formal sector to provide new seeds in large quantities on a regular basis for replacement.
In Peru, the focus of the Ministry of Agriculture on introducing improved potato seed through the commercial seed system has provided little benefit to the small farmers in the Central Highlands, as 90% of the highland potato area is linked up to informal seed diffusion mechanisms instead. These use special seed selection and storage techniques in designated areas, and potato seed is traded in the market network that is already an integral part of the local economy. The farmer-to-farmer seed diffusion networks maintain a very high quality of seed through a system of informal "neighbour certification". The farming community makes use of the large variations in altitude within the farming system to isolate different seed stocks and so maintain a renewable seed flow. They have also developed considerable expertise in seed multiplication, selection and conservation such that the quality of seed coming out of the system is not in the least inferior to that a formal certification system could ensure. In fact, the National Special Seed Project found that seed quality of new potato varieties they had distributed had been well maintained even after three years of farmer multiplication and distribution.
These examples demonstrate that farmer-to-farmer seed exchange can be very effective mechanisms for seed diffusion of both local and "new" varieties. Informal seed diffusion has several advantages over the kind of service provided by seed companies. Firstly, farmer-to-farmer systems are traditional; that is not to say that they have remained unchanged over time but that they are well-established, often highly elaborate, and rooted in the needs of the community. Secondly, they are informal or semi-structured in their organization, changing between locations and over time, and not subject to the same rigidities of ownership and control as formal sector organizations. Thirdly, they operate mainly, although not exclusively, at the individual community level, between households within one community, although lines of supply may extend over a relatively wide geographical area. Fourthly, a wide variety of exchange mechanisms are used to transfer seed between individuals and households, including cash sales, barter and transfers based on social obligations. And lastly, the individual quantities of seed thus exchanged are often very small compared to the amounts formal sector organizations deal in.
Sustaining the sustainable
The existence of farmer-to-farmer seed diffusion mechanisms and their importance to small farmers opens up a whole new set of options for supporting small farmers in their search for better seed, through helping to sustain community-based seed diffusion instead of struggling to adapt cumbersome and cost-conscious seed companies. But the kind of support that is most helpful will be very location-specific because of the way farmer-to-farmer seed diffusion works in practice.
In the Kenyan case, for example, what is required from the formal sector is not active participation in seed distribution on a wide scale but help with specific constraints within the seed system: improved on-farm seed storage technologies and the maintenance of security seed stocks for times of drought. Like in Nepal, the formal sector can play a useful role in the initial provision of new genetic material from outside provided that the new varieties are well adapted to their local needs. The inclusion of traditional varieties in formal breeding programmes and the active participation of the small-scale farmers themselves can help to ensure this. However, farmer-to-farmer exchange works well for subsequent diffusion and for replacement of the seed stock each season, and there should be a move away from an extension service which controls seed distribution towards one to facilitate informal seed exchange. The right of farmers ' to use saved seed is crucial to maintaining informal seed diffusion networks, and, where necessary, their capacity to do so must be enhanced.
A new emphasis on sustaining farmer-to-farmer seed diffusion will require new approaches to development support. Funders must be willing to support different kinds of mechanisms, according to what best fits local needs, and they must be willing to deal directly with the local level community organisations which control seed diffusion at the grassroots. Above all, such a strategy requires funders to broaden their focus dramatically, from how best to re-structure formal sector national seed companies to how best to support what goes on in the informal sector at community level. There are four main areas where new initiatives are needed:
• training in low-cost on-farm seed selection and storage techniques: - too often, extension services ignore the wealth of knowledge and experience farmers themselves have and teach instead that stocks must be replaced regularly with certified seed from the formal sector. As this is often impossible in practice, training in on-field selection, and in how to apply low-cost seed treatments is a more realistic way of improving the quality of seed farmers use;
• decentralising seed multiplication and distribution, and breeding work: - to involve community organisations and to build on the strengths of existing seed systems, instead of trying to replace them with parallel formal sector systems;
• broadening the genetic base of formal sector breeding programmes: - to include traditional cultivars, through making collections and selections of local material rather than relying only on parent material provided by the International Agricultural Research Centres. And orienting breeding objectives more towards farmers ' felt needs, away from rigid Distinctness, Uniformity and Stability criteria and towards proven Value for Cultivation and Use;
• increasing dialogue: - between formal sector breeders and seed companies and farmers on the one hand, so that communities participate in selection and multiplication decisions, and between different NGOs and community organisations involved in seeds on the other hand. In particular, this requires greater institutional flexibility on the part of national agricultural research systems, the IARCs and the international donors to make it possible for them to work with and to support community initiatives.
Many farmer groups and other non-governmental organizations are already acting practically. In Zimbabwe, for example, Environment and Development Action (ENDA) are working to multiply and distribute traditional varieties of sorghum and millet collected from small farmers in the communal areas. More initiatives like these need to be supported to ensure that seed systems for farmers become genuinely sustainable in the long run, both by strengthening traditional systems -- rather than replacing them -- and by giving local communities control over their own genetic resources. Farmers ' informal seed exchange networks are a vital component of this.
This article is based largely on two recent reports of the Overseas Development Institute: Cromwell, E.A. (ed) (1990) "Seed Diffusion Mechanisms in Small Farmer Communities: Lessons from Asia, Africa and Latin America", ODI AA(R&E) Network Paper No.21, ODI, London; and Cromwell, E A (1991) "The Performance of the Seed Sector in Malawi: An Analysis of the Influence of Organisational Structure", ODI Working Paper. For further information, contact: Elizabeth Cromwell, ODI, Regent's College, Inner Circle, Regent's Park, London, NW1 4NS, UK. Tel: (44-71) 487 7413. Fax: (44-71) 487 7590.