Genetic potential of soybean populations and progenies in tropical conditions

Abstract

Commercial soybean breeding programs have thousands of segregating populations, and the careful choice of which populations to concentrate the selection efforts on is of fundamental importance to reduce the costs and time of launching a new cultivar. Studies seeking the improvement of selection methods in autogamous plants from the application of the mixed models approach, have been carried out using strategies that associate the merit and performance of the populations, making the selection process more efficient. Thus, the objective was to compare different selection strategies of soybean progenies, and to verify the implications of these selection methods in the ranking of superior progenies. The progenies were obtained during the harvest of segregating populations in the 2018/2019 season. After selecting and trailing individual plants, they were separated into two groups: one group containing progenies from early populations and another from late populations, totaling 8 different populations. Thus, each group comprises an experiment (early and late) keeping the same number of progenies in each. The F 4:5 and F 4:6 progenies were evaluated for grain yield and absolute maturation in the municipalities of Lavras and Ijaci. The F 4:5 progenies were evaluated during the 2019/2020 agricultural year. Plots consisted of a two-meter line with two replications in a simple 13 x 13 lattice design (160 progenies + 9 controls). The F 4:6 progenies selected in the previous generation were evaluated in the 2020/2021 agricultural year. The plots consisted of two rows of four meters with three replications in a 9 x 9 tripod lattice design. The data were analyzed via the mixed model approach through the following strategies: considering and disregarding the effect of population on generations, and considering and disregarding the genotypes not selected in both strategies. Genetic and phenotypic parameters were estimated in each of the analyses. The expected and realized gain with the selection, the Spearman correlation, the coincidence index and the genetic progress were used to verify the efficiency of the used strategies. Differences were observed in the ranking and coincidence of the selected progenies when considering and disregarding the population effect, as well as when including the non-selected genotypes in the analyses. However, it was possible to identify coincident progenies in the different selection strategies that associate high grain yield and lower absolute maturation.