Application of a Modified Diallel Analysis to Evaluate Recurrent Selection for Grain Yield in Maize1

Abstract

The modified diallel analysis proposed by J. J. Hammond and C. O. Gardner in 1974 and the model proposed by O. S. Smith in 1979 to evaluate progress from recurrent selection were used ta evaluate half‐sib and S1 recurrent selection programs in the maize (Zea mays L.) cultivar, ‘BSK.’ Changes in the mean of selected populations and their crosses were found to be a function of inbreeding depression and changes in the frequencies of alkies with additive effects. Estimates of changes due to alleles with dominance effects were not significant. Terms involving changes due to alleles with dominance effects would be expected to be very small based on quantitative genetic theory and based on the way the terms were defined. Use of the modified diallel analysis for evaluating progress from recurrent selection may be limited to populations that have large effective population sizes and have undergone a large number of cycles of selection.Use of the analysis proposed by Smith showed that half‐sib and S1 selection increased the frequencies of alleles with additive effects and S1 selection was more affective than half‐sib selection. The analysis also provided an estimate of the rate of inbreeding depression and indicated that this rate was the same for both selection methods.