DETECTION OF EPISTATIC, ADDITIVE, AND DOMINANCE VARIATION IN WINTER-WHEAT (TRITICUM-AESTIVUM-L EM THELL)

The role of epistasis in the genetic systems for selected traits among several T. aestivum cultivars was determined, and the importance of additive vs. dominance gene action for traits not influenced by epistasis was evaluated. In each of 2 replicated experiments, 2 cultivars (referred to as testers L1 and L2) and their F1 (L3) were crossed to each of 10 other cultivars and evaluated in a single field experiment. The deviations (cultivar .times. L1 + cultivar .times. L2 - 2 cultivar .times. L3) were analyzed to provide a test for epistasis. An additive-dominance model was fitted to the data for those characters not affected by gene interactions. Results of the 2 experiments were generally consistent with respect to the detection of epistasis. Epistasis affected the expression of heading date, kernels/spikelet and grain yield, but not of plant height, protein content, tiller number, spikelets/spike and kernels/spike in both experiments. Epistasis was indicated for kernel weight in 1 experiment. Expression of epistasis was influenced by particular cultivars, indicating that a limited number of cultivars may not be sufficient to detect nonallelic interactions for a character within this species. Additive gene action was predominant in both experiments for protein content, spikelets/spike and kernels/spike, suggesting that improvement of those characters can be achieved through standard selection procedures. Evidence for dominance gene action was obtained for plant height and kernels/spike in 1 experiment. Epistasis may be a nontrivial factor in the inheritance of agronomic traits. Genetic models neglecting epistasis may result in biased information.