SELECTION STRATEGIES OF PROGENIES IN A COMMON BEAN RECURRENT SELECTION PROGRAM

INTRODUCTION Breeders biggest challenge is to identify the best progenies/lines or hybrids that will remain the best ones in future years and under different conditions from the experimental stations ones. The interaction progenies x environment is expressive for common beans in Brazil (MENDES; RAMALHO; ABREU, 2009; LIMA; RAMALHO; ABREU, 2012; LIMA et al., 2015), consequently the selection success can be hampered. Therefore, the aim with this work was to evaluate a progeny selection strategy that considers not only the progenies means, but also the mean from the population from which the progeny came. It uses several inbreeding generations information in order to mitigate the interaction effect. MATERIAL AND METHODS Progenies used in this work came from Universidade Federal de Lavras (UFLA) recurrent selection program for grain yield, XV cycle. 439 S0:1 progenies, 322 S0:2 and 79 S0:4 progenies were evaluated in 2015/2016. All progenies were obtained by a recombination system that is similar to a “top cross”. 20 S0:3 progenies are selected in each cycle and used as female in crossing with the 19 other ones. Grain yield was the trait under evaluation. The data were analyzed by mixed model approach. BLUP’s effects were estimate considering the original population effect and also progenies of previous inbreeding generations effect. The genetic variance among progenies (VP), the progenies x environments interaction variance (VPA), and heritability (h2) among progenies were estimated. Coincidence index was also estimated among the 10% best progenies considering one or more generation and the reference generation, S0:4