Genetics of Yield and Yield Components in Maize Composites

Half-sib family groups were developed in three composite populations namely, JI, Cuba IIJ and F₃ generation of JI × Cuba IIJ, using design I mating system of Comstock and Robinson (1948, 1952). Data were recorded in two environments on grain yield, days to silk, plant height, ear height, ear number per plant, ear length and ear diameter were subjected to genetic analysis to find out the relative magnitude of additive genetic variance, dominance variance and the dominance ratio between them.

The dominance ratio for grain yield and for most of the other characters under study was in the no dominance to partial dominance range. In all the three populations, estimates of additive genetic and dominance variance were biased by genotype × environment interaction in respect of the characters studied. Bias was larger for σ²_A in F₄ population in respect of the character grain yield.

In the F₄ population, estimates of additive genetic variance were used for calculating predicted gains for the seven characters from one cycle of selection. The results from full-sib family selection represented considerable improvement, The predicted gains at Pantnagar and Delhi were 27% and 25.1% of fhe mean respectively, while based on the combined analysis the predicted advance was 16.5%. Similar gains were also indicated for the other characters. As compared to mass selection, full-sib family selection was about three times more effective for the characters studied. Full-sib family selection, however, needs three times as many years as would one cyle of mass selection and would need more facilities.