(Brassica carinata A. Braun) varieties at Guder and Meta-Robi, West Shewa Zone

This paper summarizes the results of on-farm evaluation of Ethiopian mustard varieties grown under farmers' conditions at two locations (Guder and Meta-Robi) in west Shewa zone of Ethiopia. The study was made to test and identify varieties that can fit to the needs and circumstances of different target groups in a participatory manner for two years using nine advanced lines from the breeding program and farmer's own cultivar as a local check. Seven of the tested varieties gave seed yield advantages ranging from 5% - 18% over the local check. The entry PGRC/E 21237 gave the highest yield (2324 kg/ha) followed by PGRC/E 21223 (2271 kg/ha) while the local check gave the third lowest yield (1977 kg/ha). Farmers evaluated the varieties using their own pre-and post-harvest evaluation criteria. Leaf cook-ability (for food purpose) was the most important pre-harvest criterion they considered whereby the standard check 'Yello Dodolla' incidentally stood as the top ranking variety in this regard. They also gave weight for each of the post-harvest criteria: seed color, seed yield, marketability, seed weight, seed size and seed filling in the same order of importance. With respect to all the criteria set as post-harves, PGRC/E 21237 was ranked to be the first most preferred variety followed by 'Yellow Dodolla' whereas PGRC/E 21263 was the least of all. Local oil millers on the other hand, insited that 'Yellow Dodolla' is their first choice for having higher oil content. Considering the overall criteria used by the stakeholders (both pre- and post-harvest criteria), PGRC/E 21237 and PGRC/E 21256 appeared to be the first and third ranking varieties respectively. The standard check Yellow Dodolla ranked second in this regard and all other varieties except PGRC/E 21170 and PGRC/E 21263 performed better than the local check in fulfilling stakeholders' preferences. The study has generally established that mustard varieties with best agronomic performance are not necessarily of high farmer's choice and also that ranking of varieties differs with target groups; but the key stakeholders participation from problem identification to technology generation is indispensable and future breeding and selection of genotypes will include the selection criteria proposed by the stakeholders.