generation mean analysis for grain yield and its related traits in lentil

in order to estimate heritability and gene action for grain yield and its related traits in lentil, six basic generations were evaluated in a randomized complete block design with three replications in a field experiment. besides seed yield per plant, plant height, pod length, and 100-seed weight, the number of pods per plant, primary branches, clusters per plant, nodes per main stem, secondary branches, and the number of seeds per pod were recorded. generation mean analysis using a, b, c and joint scaling tests indicated that additive [a], dominance [d] and at least one of the epistatic effect (additive×additive [aa], additive×dominance [ad] and dominance×dominance [dd]) were involved in the inheritance of the studied traits. however, simple additive-dominance model was sufficient only for pod length. significant dominance [d] and dominance×dominance [dd] interactions with opposite sign indicated duplicate epistasis for all traits except pod length. narrow-sense heritability was low for seed yield per plant, pod length, number of seeds per pod and 100-seed weight and moderate for other traits. average dominance ratio was more than unity for seed yield per plant, number of primary and secondary branches, pod length, and 100-seed weight, which showed the high importance of dominance gene effect in control of these traits. due to the presence of greater non-additive gene effects combined with low narrow-sense heritability, selection for almost all of the studied traits in this cross, especially in early generations, would be complex in conventional methods.