Genotypic correlation and path analysis in early and super-early maize genotypes.

Abstract

The aims of this study were to assess the linear relationships between agronomic and nutritional traits and to identify promising traits for indirect selection in early and super-early maturing maize genotypes. Two trials were run in the 2009/2010 agricultural year, each consisting of a randomized block design with three replications. One trial was run on 36 early maturing maize genotypes and the other on 22 super-early maturing genotypes. Six agronomic traits, 11 protein-nutritional traits, and four energetic-nutritional traits were measured. Error normality assumptions were verified using the Kolmogorov-Smirnov test and residual variance homogeneity assumptions using the Levene test. Analysis of variance and the F-test were run for each of the 21 traits. Next, the genotype correlation coefficient matrix was estimated for the 21 traits and each trial. Ridge path analysis was based on values of k = 0.00 and k = 0.10 on the diagonal of X'X correlation matrix, taking the nutritional traits as principal variables and agronomic traits as explanatory variables. The number of days from sowing to male flowering, the number of days from sowing to female flowering, plant height, ear insertion height, ear weight and grain yield can be used in indirect selection as indicators of grain nutritional quality.