Adjustments of error mean square in corn cultivar competition trials by Papadakis method

Abstract

The objective of this work was to evaluate the different ways to calculate the yield index, used as covariable for the Papadakis method, and to verify if this method is efficient to increase the experimental precision in corn (Zea mays L.) cultivars competition trials with high number of treatments. The following analysis was performed using grain yield data: mathematical model assumption, analysis of variance in the randomized complete block design (conventional analysis), and analysis of covariance in the completely random design using yield index as a covariable. The yield index covariable was estimated using five different variants of the Papadakis method. The comparisons between the conventional analysis and the Papadakis method was done with the mean square error, coefficient of variation, least significant differences of Tukey test, group numbers of Scott-Knott cluster analysis, and differentiation index of Fasoulas. Papadakis method improved the experimental precision from 10.5% to 48.1% in all different ways used to estimate index yield covariable. The four hills neighboring the reference hill was the best index to increase experimental precision.