A half diallel set of crosses involving 8 inbred lines of maize along with two commercial hybrids (SC 10 and SC Pioneer 30k8) were evaluated at two different sowing dates (2/5/2014 and 1/6/2014) using RCBD design with three replications at the Agricultural Research and Experimental Station of the Faculty of Agriculture, Moshtohor, Benha University to determine combining ability and gene action for grain yield and its components. Significant planting date mean squares were detected, with a higher magnitude at early sowing date compared to that at late one. Genotypes as well as their components (parents, crosses and parents vs crosses) mean squares reached the significance level of probability for most studied traits. Significant mean squares due to general and specific combining ability (GCA (a) and SCA (b)) were obtained for all the studied traits using Griffing's methods (M 2 and M 4) and method of Jones at early, late and across sowing dates for most studied traits. Also, low GCA/ SCA ratio using Griffing's methods and a/b ratio using Jones method showed that non-additive type of gene action was the more important part of the total genetic variability for most studied traits in both and across planting dates. The parental inbred lines P3 and P7 seemed to be the best combiner for grain yield plant-1 as well as one or more of its components. The most desirable inter- and intera-allelic interactions were presented in both methods of Griffing by the F1 hybrids; P1xP7, P2xP8, P3xP4, P3xP7, P1xP8, P2xP3, P2xP8, P3xP4, P3xP7, P4xP7, P5xP6, P5xP8 and P6xP7 for grain yield plant-1. The crosses P3xP4, P3xP7, P6xP7, P5xP6, P2xP3, P1xP7, P4xP7, P2xP8, P1xP6 and P1xP7 gave significantly positive superiority in grain yield over SC10 and SC 30K8 in the combined analysis. Also, the three crosses P3xP7, P3xP4 and P5xP6 in the combined analysis were out yielded by (26.37 and 36.56%), (22.03 and 31.87%), (16.16 and 25.56%) over SC 10 and SC Pioneer 30K8, respectively. Highly significant and larger in magnitude values of dominance component (H1) were obtained for all traits in both sowing dates. Significant (h2) values were detected for all the studied traits. Low heritability values were detected for all studied traits. Values of H2/4H1was largely deviated from 0.25 were obtained for 100-kernel weight in both planting dates and No. of kernels row-1 at early sowing date. The same trend was detected by F values. Over-dominance was obtained for all traits. The parental inbred line P5 at early sowing date and P8 and P1 at late planting date for grain yield plant-1 seemed to carry the most dominant genes responsible for the expression of this trait. However, P4 for grain yield plant-1 had recessive genes. The correlation analysis revealed a strong relation between the four methods used in this study for estimation of error variance for all studied traits. |
