Mapping of QTL for seed size and shape traits in common bean

Our objective was to identify quantitative trait loci (QTL) for seed weight, length, and height segregating in a recombinant inbred line population derived from the common bean (Phaseolus vulgaris L,) cross 'PC-50' x XAN-159. The parents and progeny were grown in two separate greenhouse experiments in Nebraska, and in field plots in the Dominican Republic and Wisconsin. Data analysis was done for individual environments separately and on the mean over all environments. A simple linear regression analysis of all data indicated that most QTL appeared to be detected in the mean environment. Based on these results, composite interval mapping (CIM) analysis was applied to the means over environments, For seed weight, strong evidence was indicated for five QTL on common bean linkage groups (LGs) 3, 4, 6, 7, and 8, Multiple regression analysis (MRA) indicated that these QTL explained 44% of the phenotypic variation for the trait, Weaker evidence was found for three additional candidate QTL on bean LGs 4,,5, and 8, All eight markers associated with these QTL were significant in a MRA, where the full model explained 63% of the variation among seed weight means. For seed length, CIM results indicated strong evidence for three QTL on LG 8 and one on LG 2, Three additional putative QTL were detected on LGs 3, 4, and ii, The markers associated with the three seed length QTL on LG 8, and the QTL on LGs 2 and 11 were significant in a MRA with the full model explaining 48% of the variation among seed length means. For seed height, three QTL on LGs 4, 6, and 11 explained 36% of the phenotypic variation for trait means. Four of the seven QTL for seed length and two of three QTL for seed height also appeared to correspond to QTL for seed weight, Four QTL for common bacterial blight resistance [Xanthomonas campestris pv, phaseoli (Smith Dye)] and for smaller seed size were associated on LGs 6, 7, and 8, The implications of these findings for breeders is discussed.