Abstract

The advantage of farmers' participation in breeding programmes depends on the effectiveness of the evaluation methods used. There is scarce information on the farmers' evaluation methods and their effectiveness as compared to the breeder’s methods. Three districts representing the different agro-ecologies, inhabiting ethnic communities, cassava production and utilization niches were selected. Focused group discussion was used to determine the variety evaluation methods. The effectiveness of farmers and breeders evaluation methods were tested by evaluating cassava trials consisting of 15 varieties planted in the three districts. High storage roots yields, resistance to diseases and pests were the most preferred traits across the districts. Preference for traits related to plant type and storage roots quality, and variety ranking differed between districts indicating the differences in agro-ecologies, production and utilization niches. Farmers' and breeder's evaluation methods significantly correlated for related traits and elicited differences between varieties for most traits evaluated. Though low cyanide content was among the farmers' preferred traits, their evaluation method did not either elicit genotypic differences between varieties or correlate with the breeder's evaluation. The breeder lacked evaluation method for taste of boiled storage roots despite it being preferred by farmers. Despite the differences in traits preferences and variety ranking between farmers and the breeder and between districts, farmers and the breeder have effective evaluation methods. However, both evaluation methods have deficiency in evaluation of some of the preferred traits indicate the need for participatory variety selection.

Key words: Farmer preferred traits, farmers' qualitative evaluation method, breeder's quantitative evaluation method, participatory variety selection (PVS), variety ranking.