Abstract

Jonas Chianu, Bernard Vanlauwe , John Mukalama , Akin Adesina , Nteranya Sanginga

In order to determine the improved soybean varieties that if recommended to the farmers would have a high probability of adoption, a farmer participatory approach was used to evaluate 12 soybean varieties at full podding in five locations (Oyani, Riana, Kasewe, Akiites, and Mabole) in western Kenya. These comprised of 11 improved varieties (TGx1871-12E, TGx1895-4F, TGx1895-33F, TGx1895-49F, TGx1878- 7E, TGx1893-7F, TGx1893-10F, TGx1740-2F, TGx1448- 2E, NAMSOY 4m, and MAKSOY 1n) and one local variety (Nyala). Farmers generated all the 17 criteria for use in the evaluation, with researchers only facilitating. One hundred and two farmers (52% females) participated in the evaluation. A scoring matrix was employed to articulate the results. Data analysis was done using Microsoft Excel. This paper shows that of the seven dual-purpose varieties tested in all the five locations, only TGx1740-2F was acceptable in all. Some varieties were acceptable in specific locations: TGx1895-49F in Oyani, Nyala in Kasewe, TGx1448-2E in Akiites, and TGx1893- 7F in Mabole. This result shows that to avoid low adoption, a blanket recommendation of varieties that were accepted only in selected locations must be avoided. TGx1740-2F was the only variety that could be recommended across locations and that was clearly better than the existing farmers’ own variety, Nyala.