The genetic control of flag leaf length in normal and late sown durum wheat

Parental, F 1 , F 2 , BC 1 , BC 2 , BC 11 , BC 12 , BC 21 , BC 22 , BC 1 self-pollinated and BC 2 self-pollinated generations of three crosses involving six cultivars of durum wheat ( Triticum turgidum var. durum ) were studied for flag leaf length under normal and late sown environments to analyse the nature of gene effects. For most crosses the 10-parameter model was the best descriptor of the data to account for the variability in generation means of this trait but in cross HI 8062×JNK-4W-128 the 6-parameter model was the best (normal sown) while in cross Raj 911×DWL 5002 the 3-parameter model was the best (late sown). Of the epistatic interactions, dominance×dominance ( l ) and dominance×dominance×dominance ( z ) played significantly greater roles in controlling the inheritance of this trait. Absolute totals of non-fixable gene effects were much higher than the fixable gene effects in all the crosses in both the environments, indicating the greater roles of non-additive effects in controlling the inheritance of flag leaf length in durum wheat cultivars. Significant partial dominance was observed frequently in most of the crosses. Significant heterosis was attributed to combined effects of dominance×dominance ( l ), additive×additive×dominance ( x ) and dominance×dominance×dominance ( z ) epistatic interactions in the cross Cocorit 71×A-9-30-1 under late sown environment. Biparental mating and/or diallel selective mating, which exploit both fixable and non-fixable components, have been suggested for the improvement of this trait in durum wheat cultivars.