This paper aimed to evaluate the efficiency of circulant diallel in relation to the complete ones, with data simulation based on estimates of general combining ability (GCA) and specific combining ability (SCA). Fifty lines with 100 independent genes of same effects and without epistasy were simulated. Heritabilities of 10%, 20%, 50% and 75% were considered, in models with addictive and complete dominant allelic interaction. Real data experiments presented in literature were used. They were compared by means of correlation between estimates of GCA and SCA, and also by means of proportions of estimates in the circulant that occurred in the confidence intervals of the complete ones. It was concluded that the efficiency is the same in both diallels, considering the parents classification in relation to GCA and SCA, and the magnitude of those parameters; that the number of crossings (s) affects the estimates of GCA and SCA, although with a small value of s it is possible to obtain good agreement with the estimates of the complete diallel; and, that with low heritability it is advantageous to increase the number of each parent crossings, although with the half of the number of parents involved, high efficiency of the circulant diallels is reached.
breeding methods; progeny testing; diallel crossing; combining ability; simulation; quantitative genetics
