Combining Ability in Durum Wheat: I. Agronomic Characteristics1

Abstract

The parents and 45 F1 and F2 populations from a diallel cross among 10 genetically diverse varieties or lines of durum wheat (Triticum durum Desf.) were evaluated at two locations in North Dakota in 1965. Parental, Fu and F2 populations of 17 of the 45 crosses were evaluated again in 1966 at one location. The traits studied were grain yield, number of tillers per unit area, kernels per spike, kernel weight, test weight, seedling vigor, maturity, height, and lodging.Heterosis percentages for grain yield in 1965, based on means of the higher parent of each cross, ranged from —19 to 84% for the F1's and from —23 to 26% for the F2's. Seventeen hybrids displayed significant heterosis for yield, but none was significantly higher than the check variety, ‘Wells.’ The best hybrid yielded 16 and 9% more than Wells when Wells was sown at 33 and 67 kg/ha, respectively. Heterosis also occurred for the other character studied. Twenty‐seven F2 populations yielded significantly less than their respective F1's.An estimate of genotype ✕ environment interaction was calculated for grain yield on the means of the 17 F1's, F2's, and parents for the crosses grown in 1965 and 1966. The analyses were based on mean values from three environments, Fargo 1965, Langdon 1965, and Fargo 1966. Genotype ✕ environment mean squares were significant for the F1 hybrids and parents, but nonsignificant for the F2 populations.General and specific combining ability mean squares were obtained from F1 and F2 populations grown in 1965. General combining ability mean squares were highly significant among F1's and among F2's for all nine characteristics.Specific combining ability mean squares among F1's were significant for kernel weight, test weight, and seedling vigor, and were significant among F2's for kernel weight, test weight, and lodging score. Maximum yields of durum wheat may be attainable only with a system that can exploit both additive and nonadditive genetic effects.