Path analysis of the relationships between the performance and its constituents in corn lines

The article presents the results of studying the genotypic variability parameters and relationships between traits affecting the performance in corn lines. In a series of three-year trials in 2006–2020, genotypic levels of the performance and its constituents in mid-early and mid-ripening corn lines, which were bred in the Laboratory of Corn Breeding and Seed Production of the Plant Production Institute named after V. Ya. Yuriev and became a part of its working collection, were determined. It was found that the average group performance of mid-ripening lines was higher than that of mid-early ones with differences depending on kernel type. Path analysis of genotypic correlations between the performance and its constituents allowed for evaluation of direct and side effects. It was found that the coefficients of pair correlation between the performance and its constituents were positive and mostly low or moderate, ranging 0.327 for "kernel number per ear" trait to 0.656 for "ear diameter" trait. However, when the correlations between the studied traits were divided into direct and side effects, the direct path coefficients were only high for the impact of "kernel number per ear" and "1000-kernel weight" traits (0.835 and 0.812, respectively) on the performance. Such traits as the "ear diameter", "number of kernel rows" and "kernel number per row" had very low values of the direct path coefficients (0.041 – 0.098), but affected the performance indirectly through the “kernel number per ear” trait. The "number of kernel rows" (0.622) and "kernel number per row" (0.544) exerted the greatest indirect impact on the performance through the "kernel number per ear" trait. As to the “1000-kernel weight” trait, there were significant negative indirect effects of the "kernel number per ear" and "number of kernel rows" traits, with the path coefficients of -0.350 and -0.362, respectively. It was shown that "ear diameter"-, "number of kernel rows"-, and "kernel number per row"-oriented selections were indirect, as they are components of the "kernel number per ear" trait and are at another level in the performance module. The “kernel number per ear”- and “1000-kernel weight”-oriented direct selections for performance can be most effective, but regarding the latter, one should take into account the negative side effects of the "kernel number per ear" and "number of kernel rows" traits, which can be achieved by building up aggregated indexes. Keywords: corn, inbred line, performance, correlation analysis, path analysis.