The determination of the plot size and the repetition number is important to the correct evaluation and differentiation of genotypes. The experimental planning can be made using previous results in already conducted assays. The objective of this work is to verify the repeatability of the heterogeneity index and the experimental plan, in the same experimental area with beans culture, during seven years (2000 to 2006). Fourteen competitive assays of elite genotype beans had been tested in Santa Maria, State of Rio Grande do Sul, Brazil - in two growing seasons (normal and off season crops), with different treat numbers (advanced lines and variety). The experiment was carried out in a randomized complete block design with three replications. For each assay, the mean, variation coefficient, intraclass correlation coefficient and heterogeneity index of the grain yield of beans were determined. Also the true differences between two means of treatments by Hatheway methods, using for different values of the heterogeneity index, repetition number had been estimated together with the plot size. The frequency of the heterogeneity index between 0.2 and 0.7 was 71%, evidencing that the experimental plan, on the basis of the results of previous experiments, has validity in 71% of the cases. It does not seem to have b predominance in function of the growing seasons - normal and late season crops, or of the type of experiment - VCU or variety, showing a good repeatability or persistence in elapsing of the years.
plot size; replications number; repetibilidade; crop breeding
