Adaptabilidade e estabilidade de genótipos de soja por RELM/BLUP e GGE Biplot

Abstract

With the rapid expansion of Glycine max (L) soybeans. in Brazil the study of genotype x environment interaction has become increasingly essential given the high influence that culture suffers on this interaction, where in different regions with the same genotype the most different phenotypes are observed, hindering the work of breeders, noting that there are changes in the ranking of genotypes from one environment to another, thus not making a perfect correlation between genotype and phenotype. Therefore, it is necessary to conduct multi-environment tests in order to test and select the genotypes in the region of interest. The data obtained by this range of sites are generally unbalanced and voluminous, making it necessary to use precise statistical methods. The use of methodologies such as REML/BLUP and GGE Biplot are highly used for the purposes of genotype classification, selection, test sites and formation of mega-environments. Therefore, the objective of this work was to select genotypes superior in terms of adaptability, stability and productivity in different environments throughout the Brazilian cerrado during the agricultural years 2018-2019 and 2019-2020. In all assays, a randomized block design was used, with three replications, evaluating mainly the grain yield. Joint analyses of variance were performed, and the genotype x environment interaction was highly significant. The most promising genotypes in relation to grain yield were G3, G6 and G4, the most discriminative and representative environment wasQuerência – MT 2020, and six mega-environments were identified serving as allocation of sites for future tests. Thus, it is observed that the REML/BLUP and GGE Biplot methods are highly correlated for selection and recommendation purposes.