Bi-segmented regression, factor analysis and AMMI applied to the analysis of adaptability and stability of soybean

Abstract

The aim of this study was to determine the phenotypic adaptability and stability of the soybeans grown in different environments of Rio Grande do Sul through bissegmented regression analysis, factor analysis and AMMI. The experiment was held in 2013/2014 season in a randomized block design factorial (six environments cultivation 20 genotypes of soybean), arranged in three replications. The results showed that methods of adaptability and stability have similar responses to the rank of environments. The Tenente Portela -RS, Arroio Grande -RS lowland and coxilha were favorable environments, while Santa Rosa -RS and Campos Borges -RS have been recognized unfavorable for soybean genotypes. Differential responses were obtained for the Sarandi -RS environment, where the factor analysis and regression bissegmented revealed that this is a favorable environment, while the method AMMI indicated negative scores and unfavorable. Genotypes TMG 7161 RR and NA5909 RR were indicated as high productivity and recommended for favorable environments for both methods. To reconcile bissegmented regression, factor analysis and AMMI accurately predicted the best genotypes and environments for soybean.