This study aimed to compare the efficiency in the selection of genotypes and the quality of the estimates of the variance components and heritability using the augmented block designs (ABD), the duplicated augmented block designs (DABD) and the group of randomized blocks design experiments with common treatments (ERBCT). For comparison, it was imposed a condition that the number of genotypes under selection was greater than the number of units available for experimental designs with repetition. Four scenarios were composed of a combination of different values of heritability and parametric residual coefficient of variation. From the 2400 simulations per scenario the ABD had the highest selection efficiency. The designs with repetition (DABD and ERBCT) had similar selection efficiency between them. It is possible to conclude that not using repetitions, and therefore, evaluating a larger number of genotypes, can bring better results to the plant breeding program.
selection; simulation; Saccharum spp
