Potencial genético de populações de soja para produtividade de grãos e maturação absoluta

Abstract

Hybridization is the main breeding method used to increase genetic variability in soybean cultivation in order to obtain superior cultivars to existing ones. In order to reduce the costs and release time of a new cultivar, it is necessary to use methods in order to identify the best segregating populations even in early generations. For this, the use of quantitative genetics has helped breeders in choosing the best populations. Estimates of m + a '(contribution of homozygous loci) were successfully used to identify and select segregating bean populations, but there are only few reports of their use for soybean crop. Thus, studies are necessary to obtain useful information for the selection of populations in quite early generations. Therefore, the objective of this study was to predict the genetic potential of segregating soybean populations for the traits grain yield and absolute maturity; to estimate phenotypic parameters for agronomic traits in soybean segregating populations and to select segregating populations with better genetic value. The experiments were conducted in Lavras and Ijaci municipalities in the 2018/2019 crop. Twenty F3 and F4 populations obtained through hybridization of 10 commercial soybean cultivars were used. The experimental design in band with three replications was used. The plots consisted of four lines five meters long and spaced 0.50 meters between lines. The agronomic traits evaluated were grain yield and absolute maturity, and afterwards, for the prediction of the genetic potential of the segregating populations, the estimates of m + a', from each plot were obtained through the variables grain yield and absolute maturity. The data obtained were submitted to individual and joint analysis of variance with the aid of SISVAR software and the averages obtained compared by the Scott-Knott test (1974). The sources of individual variation were significant for all evaluated characters, showing differences between the segregating populations and the generations of these populations. Pearson Linear Correlation estimates for grain yield and absolute maturity with their respective m + a' were significant, assuming values of 0.66 and 0.90 respectively. Estimates of m + a' for the evaluated traits show differences in the contributions of the homozygous loci of the populations, allowing the selection of superior strains. Populations from FMT1 x FMT2, FMT1 x TMG123RR and CD250RR x BRSMG780RR crosses have good potential for strains extraction. The FMT1 x TMG123RR population stands out for including all favorable characters.