Skip to content

This paper results from a study that was commissioned to contribute to the Program of Accompanying Research for Agricultural Innovation (PARI). The overall purpose was to propose an agenda for supporting sustainable development of seed systems in two Sub-Saharan countries, Kenya and Mali, based on the experiences and insights of seed system actors who contribute to various functions and operate at different scales. The study relied on a mixed methodology, including (1) a desk review of secondary sources; (2) interviews with individual seed system actors and key informants; and (3) workshops in which diverse actors jointly discussed and prioritized options for sustainable seed system development. Staple cereal crops were targeted that are important for the farming and food systems of both countries: maize and sorghum for Kenya and maize, rice, sorghum and pearl millet for Mali.

In Kenya, most breeding for staple cereal crops is done by public breeding programmes, while some private breeding companies are also active. Seed production is mostly based on contracts between seed companies and large-scale farmers, while distribution is organized in the form of multi-level sales networks. In Mali, all breeding for staple cereal crops is done by public breeding programmes, with small farmer-managed seed enterprises being engaged in seed production and dissemination in their geographical areas. Collaboration between breeders and farmer cooperatives is extensive and crucial for the development and spread of new varieties, since the purchase of certified cereal seed by Malian farmers is quite a novelty. Limited choice of new varieties exists in both countries, particularly under conditions where climate variability and low soil fertility prevail. Furthermore, important quality and use-related traits are not systematically considered in breeding programmes. Slow and costly release procedures, limited availability of information about new varieties along with cash-flow constraints at various levels are factors that limit the dynamics of seed system development. Differences in structure, organization and size of the seed markets in Kenya and Mali, and in various actors’ contributions to seed system functions, lead to different outcomes in terms of quality, availability and access to seed. One important hypothesis for further discussion is that business models that include more decentralized models of seed production and distribution have comparative advantages for meeting the highly diverse demands of farmers in countries like Kenya or Mali, with a wide range of agro-ecological conditions and production systems, and could help reduce transaction costs. Furthermore, regulatory systems that provide space for a diversity of approaches for variety development, release, seed production and dissemination, are expected to be more supportive in this particular situation, compared with systems that focus on a narrow range of actors and variety types.

Important conclusions are that sustainable seed system development requires more actor-orientation, with a central focus on farmers’ capacities and needs. Furthermore, strengthening actors’ capacities to collect, share and assess information about varieties and their comparative performances will contribute to dynamic, responsive seed systems. Plant breeding, as the source of value creation, needs to be regarded as an integral component of functioning seed systems and requires joint consideration of what demands for innovations actually exist in order for seed systems to advance. Decentralized seed production and marketing enterprises can serve as nuclei for an emerging locally-based seed industry where market opportunities are limited and preferences diverse. Lastly, seed systems in both Kenya and Mali could benefit from more rigorous assessments of how interventions, new technologies, policies and formal organizations influence seed system innovation and sustainable development.

This study is based on a more detailed report which is available here (3.2 MB).

Published as ZEF Working Paper No. 165.