Comparison of methodologies (REML/BLUP vs. GGE Biplot) for soybean selection in the Brazilian Cerrado
Rafael de Moura, Rafael Novais de Miranda, Cleiton Renato Casagrande, Franco Romero Silva Muniz, Rafael Storto Nalin and Aurinelza Batista Teixeira Condé
Abstract: The selection of adapted and stable soybean genotypes for the Brazilian Cerrado is hindered by the complex genotype x environment (G x E) interaction. This study aimed to compare the effectiveness of the REML/BLUP and GGE Biplot. Yield data from 33 genotypes in 44 environments (2019-2020 and 2020-2021 growing seasons) were analyzed. The results indicated a significant G x E interaction and high selective accuracy (0.86), validating the selection precision. Both methodologies proved to be effective and concordant in identifying a group of elite genotypes (commercial checks G3 (BASF), G6 (BASF), G4 (BASF), and the experimental lines G17 (BASF) and G18 (BASF)). REML/BLUP provided the quantitative basis for selection, whereas the GGE Biplot elucidated the formation of mega-environments and the positioning of genotypes relative to the ideotype. It is concluded that the methodologies are complementary, enhancing decision-making and maximizing genetic gains in soybean breeding programs for the Cerrado.