Strategies for recurrent selection programs aiming upright common bean plants

Abstract In common beans (Phaseolus vulgaris L.), upright plant architecture is highly valued for agriculture. As plant architecture is controlled by many genes, recurrent selection is the optimal strategy for breeding programs. Additionally, conducting mass selection before flowering to intercross selected plants in the field increases efficiency by reducing the number of harvest cycles needed to complete a selective breeding cycle. For this to be effective, the plant's architecture at flowering must accurately predicts its trait at harvest. Based on that, the objective of this study was to verify whether there is an association between plant architecture at flowering and at harvest and to study the relationship between traits of plant architecture and grain yield. The traits assessed included plant height and plant architecture scores at flowering and harvest, stem diameter, grain yield, and the mass of one hundred grains. Sixty-two progenies from each S0:2, S0:3 and S0:4 generation, with two commercial cultivars as controls, were evaluated across three environments in a triple lattice experimental design. Genetic and phenotypic parameters and correlation coefficient were estimated. Progenies selection was carried out using index Z, and subsequently, the correlated response in the traits was determined. It was found that the correlation between heights and architecture scores at flowering and harvest were of high magnitude. There was also a low correlation between plant architecture traits and grain yield. The results of this study suggest that selecting upright plants during the flowering stage is feasible, and that plant architecture and grain yield are expressed independently.