Diversidade fenotípica e molecular, correlações entre caracteres, adaptabilidade e estabilidade de genótipos de soja

Abstract

This study aimed to: a) assess the genetic diversity of soybean genotypes using agronomic traits and microsatellite markers b) assess the phenotypic and genotypic correlations between agronomic traits of soybean lines derived from different crosses, c) study interaction between genotype and environment and evaluate its effect on grain yield and d) evaluate the adaptability and stability of early and late soybean genotypes by AMMI, GGE Biplot and factor analysis. To study genetic divergence, we conducted a field experiment at Fazenda Capim Branco (phenotypic analysis) and performed molecular analysis at the Laboratory of Genetics, Institute of Genetics and Biochemistry, UFU. We evaluated 35 soybean genotypes using seven agronomic traits and nine microsatellite markers. These genotypes were then grouped by UPGMA and Tocher cluster analyses, which showed that agronomic traits and microsatellite markers can be used concomitantly to detect potential parents for soybean breeding programs and that the G11, G12, G16, G22, G26 and G33 genotypes could be used to obtain segregating populations with superior genetic variability. Correlations among 71 soybean lines and 7 agronomic traits were evaluated. It was found that selecting late-flowering plants produced taller plants at maturity and that selecting plants with the highest pod numbers favored breeding for grain yield. To study environmental stratification, adaptability and stability, four experiments were conducted in five municipalities of the state of Mato Grosso (Alto Taquari, Lucas do Rio Verde, Sinop, Querencia and Rondonopolis) using the 2011/2012 and 2012/2013 crops. Twenty-seven early-cycle and thirty late-cycle genotypes were evaluated to determine grain yield. These evaluations showed significant and complex interactions between genotype and environment regarding grain yield. The AMMI2 model was more efficient at retaining most of the variation in the first two principal components for early and late-cycle genotypes, respectively. The Lucas do Rio Verde site (2012/2013 crop) was the most stable and highest yielding environment. The G5 and G10 genotypes were the most stable of the early and late-cycle genotypes, respectively, based on AMMI2, factor analysis and GGE Biplot.