| Harvest index: The ratio of grain dry weight to the total aboveground weight at
maturity. Grain yield: Weight of grains in grams harvested from each plant. Spike length: Length in centimeters of the spike on the tallest culm (awns excluded). 100-kernel weight: Weight in grams of 100 random kernels of each plant. Weight of kernels, number of spikelets and kernels per spike were based on the spike of tallest culm. Analysis Procedures Analysis of variances including all generations as well as separate analysis for each population were conducted for each trait. Scaling tests have been calculated to test the adequacy of additive-dominance model using Mather's formulae (1949). A joint scaling test attributed to Cavelli by Mather and Jinkes (1971) was also conducted to evaluate the goodness of fit of the three-parameter model. The means of all generations were analysed by the method outlined by Hayman (1958) to provide a six-parameter model. Those parameters using Gamble's notation (1962) were the F2 population mean or mean effects (m), the pooled additive effects (a), the pooled dominance effects (d), the pooled additive x additive epistatic effects (aa), the pooled additive x dominance epistatic effects (ad) and the pooled dominance x dominance epistatic effects (dd). Heritability in the narrow sense (h2 ns) and their standard errors were calculated for each trait using the method of Warner (1952). The expected genetic advance from selection (G.S) was estimated following Allard (1960) at 5% level of selection. Results and Discussion Means and variances Analysis of population means and variances are given in Table (1). The results proved that the six generations were differed significantly for all traits under study. Mean values for different agronomic traits as shown in Table (2) indicated that Super-X cultivar was earlier than Samma cultivar and larger in grain yield per plant, weight and number of kernels per spike, 100-kernel weight and harvest index. The F1 showed to be lower than the parental mean value for all traits except plant height (Table 2). These significant deviations of F1 indicating a sizable amount of nonadditive gene action for all traits except date headed and harvest index. It can be concluded that there were a preponderance of recessive genes plus modifiers at the loci which the parents differed. Similar results were obtained for kernel weight by SUN et al. (1972), for spikelets and kernels number per spike and 1000-kernel weight by HASSANIEN et al. (1974) and for spikelet number per head by HALLORAN (1975). Date headed, plant height and 100-kernel weight in this investigation were in agreement to the findings of KETATA et al. (1976). WIDNER and LEBSOCK (1973) found that the F2 was higher in mean grain yield than the F1. |
| <-- Back | | | --> Next |