Involving farmers in maize variety selection and seed production: experiences from the Lake Victoria region of western Kenya.

G.D. Odhiambo, M. Siambi, R. Apamo, E. Ochieng, P. Mbai, S. Mugo & H. DeGroote

Maize is the staple food for the Lake region of western Kenya. Hybrid varieties could increase productivity but environmental stress, especially low soil fertility and the parasitic weed Striga make hybrids unattractive to farmers. Farmers prefer open pollinated maize varieties (OPV), the seeds of which can be recycled for up to three seasons before farmers commonly seek new seeds.  OPVs can be improved for grain yield potential, foliar diseases, and tolerance to drought, low soil nitrogen and Striga. Seed of improved maize OPVs are often unavailable as they are unattractive to seed companies due to their low profit margins.  One way to help ensure adequate supply of high quality seeds is to have farmers produce and market maize OPV seeds. A collaborative project between CIMMYT and KARI was developed to expose farmers to improved maize varieties during their last phases of development, while allowing farmers to select material based on their own criteria. Farmers are then trained in seed production, hopefully leading to an increased capacity of the farming community to produce their own seed. The varieties are evaluated on-farm using the mother and baby trial design which enables a large number of farmers to access variety evaluation sites and participate in varietal selection. Through this approach, 28 maize OPVs from different maize breeding institutions were evaluated during the 2001 long rains season at 13 sites located in six districts of Nyanza province. The mother trials were grown under optimum and low soil fertility management. Farmers participated in the evaluations and selection of the varieties at two stages including flowering and harvest. Farmers developed and used their own criteria for selection during these periods. Desirable traits identified by farmers at flowering were earliness and large potential ear size. At harvest, desirable traits included large cobs, kernels, and grain yield. Farmers scored the varieties for each criterion and ranked the best three varieties at each site. Farmer preferences often differed among the sites, which often reflected differences in the agro-ecological zones. Since grain yield components were the main criteria used by farmers, different varieties were selected under high fertility and low fertility management as the varieties performed differently under these conditions. Two farmer groups are already participating in on-farm seed production and more communities are planning to participate in production of seed of varieties of their choice.  Several are in the process of release by Kenyan authorities. Our study shows that farmer participation in the evaluation process is a useful mechanism to expose farmers to new and improved maize varieties hopefully leading to increased maize variety adoption and eventual productivity.