Multigeneration Index in the Selection of Common Bean Inbred Families

Abstract

For the common bean ( Phaseolus vulgaris L.) crop, most recurrent selection breeding programs use, at the selection stage, inbred families evaluated across two or three generations. This study applied a multigeneration index to produce a best linear unbiased predictor (BLUP) of family effects, aimed at assessing its efficiency in the selection of common bean families. Data from 380 multigenerational families were evaluated across different sowing times and years, and across generations F 2:3 and F 2:4 . The evaluated traits were: plant architecture, yield, and grain appearance. Bivariate BLUP with genotype × environment interaction was fit using a compound symmetry model. We also selected 90 families and evaluated the abovementioned traits in the F 2:6 generation. Comparing the accuracy of the multigeneration index using all three generations simultaneously with the one using only the F 2:6 generation, the increases in genetic gain were 21, 28, and 3% in plant architecture, yield, and grain appearance, respectively. Therefore, we concluded that the multigeneration index including data from all available generations is the most efficient method for selecting common bean families. Additionally, genotypic values were predicted to be free from interactions with the sowing times and years of planting.