Repeatability of elephantgrass forage traits

Abstract

The repeatability coefficient is widely used by geneticists as a measure of the upper limit of the heritability and as an efficient tool for predicting genotype value from successive measurements on individuals. It also allows to estimate whether the number of evaluations is sufficient to determine genotype superiority within a certain degree of reliability. The purpose of this research was to evaluate the coefficient of repeatability of crude protein concentration on stems and leaves, in 77 accesses of the Embrapa's Elephantgrass Germplasm Active Bank. The experimental design consisted of was randomized blocks, with two replicates. Forage was cut on 30, 45, 60 and 75 days of croping, using 15 sprouts of each experimental unit. ANOVA was performed for each trait data set, and genetic and environmental parameters were determined. The repeatability coefficients were estimated by the methods: Analysis of variance, principal components - correlation matrix, principal components - co-variance matrix and structural analysis - correlation matrix. Genetic variability was found among the genotypes for the appraised traits, which allowed the identification of genetically superior material that can be used in breeding programs. The coefficients of estimated repeatability were not different for the traits studied. Four cuts were enough for a coefficient of determination of 0.80, for all the methods. The use of seven cuts may enable predictions with 0.90 reliability for the studied traits, using any method.