Interação entre genótipos e ambientes, adaptabilidade e estabilidade e correlações canônicas em caracteres de milho

Abstract

The objectives of this work were to verify if there is interaction between genotypes and environments, to study the behavior of the genotypes through the analysis of adaptability and stability and to verify if there is a linear dependence between phenological, morphological, productive and tassel traits in maize genotypes, in seven experiments. Seven experiments with 16 maize genotypes were controlled in a randomized block design with three replications. Phenological (two), morphological (three), productive (four) and tassel (11) traits were evaluated. Individual and joint analyzes of variance and the F test at 5% significance were performed. Genotype effects were considered as fixed and environment as random. Afterwards, analyzes of adaptability and stability were carried out, for the traits that showed interaction between genotypes and environments, using the methods of Yates and Cochran (traditional method), Plaisted and Peterson, Wricke, Annicchiarico, Schmildt et al., Eberhart and Russell, Silva and Barreto, Cruz, Torres and Vencovsky, Huehn, Lin and Binns modified by Carneiro and the AMMI analysis. Agreements and/or disagreements between estimates of adaptability and stability parameters were performed for nine traits, which were verified using Spearman's rank correlation coefficient, and their significance was verified using Student's t test. The matrix of phenotypic correlation coefficients between the traits that showed significant differences was estimated by the joint analysis of variance F test. The diagnosis of multicollinearity was made in each traits group through the condition number and the variance inflation factor. In the analysis of the canonical correlations, the associations between the groups of phenological versus tassel, morphological versus tassel, and productive versus tassel traits were verified. There is interaction between genotypes and environments, with regard to phenological, morphological, productive and tassel traits in maize, except for the morphological character ear height. In general, there is no indication of an ideal genotype when considering the four groups of traits, as these did not simultaneously present the appropriate parameters for the adaptability and stability methods used. There is agreement between the adaptability and stability methods of Plaisted and Peterson with Wricke and of Annicchiarico with Schmildt et al. and/or Lin Binns modified by Carneiro being unnecessary the concomitant use of these methodologies. There is a need to use more than one methodology for consolidating genotype indications, and it is interesting to use methods that simultaneously assess mean, stability and adaptability in general, favorable and unfavorable environments. Based on the phenotypic correlation matrix and on the canonical correlation analysis, it can be inferred that there is a linear dependence between the groups of phenological, morphological, productive traits and the tassel traits. Tassel traits, number of branches and tassel dry matter mass can be used for indirect selection of maize plants.