Medidas de precisão experimental e número de repetições em ensaios de genótipos de cana-de-açúcar

Abstract

The objective of this work was to evaluate the appropriateness of the statistics selective accuracy and F-test value for genotype as measurements of the experimental precision degree, and to determine the number of measurements (replicates) necessary to predict the performance of sugarcane genotypes (Saccharum officinarum). Data on stalk yield from three harvests in 68 genotype competition trials of sugarcane were used. In all trials, the experimental design was a randomized complete block, with three replicates and number of genotypes between 16 and 28. Correlation and path analyses were performed and the frequency of trials was calculated in each experimental precision class. Repeatability coefficients were estimated and the number of needed replicates was calculated to predict true genotypic values. The statistics selective accuracy and F-test value for genotype are appropriate to evaluate the experimental precision of sugarcane genotype competition trials. Trials with three replicates identify superior sugarcane genotypes, with 80% accuracy of the true prognostic value.