REPETIBILITY IN GUAVA: HOW MANY EVALUATIONS IS NECESSARY FOR SELECTION THE BEST GUAVA TREE?
In fruit breeding, especially in the case of perennial species, is questioned what the ideal number of evaluations. The objective main of the present study was to estimate what the best number of harvest to be evaluation and select productive genotypes, with good adaptability and stability, by method of harmonic mean of the relative performance of predicted genetic values (HMRPGV), for production related traits by mixed models. We evaluated ten full-sib progenies, over four harvests, for fruit weight, number of fruits per plant, and total yield. Five predicted measurements from the same individual were sufficient for the selection of promising guava genotypes in the early stages of the breeding program. The genetic gain predicted for the 30 best individuals was higher than that achieved with the progeny. These genotypes can be used both as parents in new crosses and in the recommendation of clonal propagation for conventional planting, since there was a significance increase in yield.