Evaluation of new soybean varieties for market traits and adaptation in Uganda
This study was carried out to identify superior soybeans genotypes that are adapted to different farming environments in Uganda and most desired by farmers and processors through Participatory Crop Improvement approach. The field experiments were set up on-farm and on-station in the 2008 – 2009 cropping seasons. Researcher designed but farmer managed trials were set up in districts of Mayuge, Pallisa, Lira, Apac, Masindi, Hoima and Kamwenge, and run for three consecutive cropping seasons. Several univariate and multivariate methods were used to analyze qualitative (farmer and processor preference) and quantitative (yield) data. Preference scores of 82 farmers were recorded at pre and post harvest stages. Soybean genotype stability studies were conducted at five locations and analyzed with AMMI. Preference for new soybean genotypes varied from farmer to farmer however, genotypes DXTPYT06A8.11, DXTPROGENIES4.17-4, MAKSOY1N, DXTPYT06A7.10 and DXTBLP(SRB)12.4 were generally acceptable because of being early maturing. Data from on-farm trials also showed that 60% of the most farmer preferred new soybean genotypes were not the highest yielding genotypes. A case in point is genotypes DXTPROGENIES4.7, BSPS17B, BSPS48A and BSPS85 which were found to have yielded highest but were not selected because they were late maturing. Processors generally preferred genotypes DXTSPS4.19, NGDT8.10-10 and DXTPYT06A4.22 for their big seed size and clear/white hilum colour. Genotype DXTPROGENIES4.17-4 scored well for both farmers and processors and may thus be successfully grown and marketed. With AMMI analysis, genotypes BSPS48A, DXTPROGENIES4.7, DXTPYT06A8.3, NGDT8.10-10, DUIKER and DXTPROGENIES4.17-4 were identified as most stable across the five locations. Similarly, genotypes NAM11XGCBLP11.3, BSPS43, NAM11XGCBLP20.2 and BSPS48A were tolerant to soybean rust. Therefore involvement of end-users in the development of new soybean varieties helped in the identification of attributes valued by end-users but unknown to breeders.