Reciprocal Recurrent Selection Compared to within-Strain Selection for Increasing Rate of Egg Lay of Tribolium under Optimal and Stress Conditions.

A replicated comparison of reciprocal recurrent selection (rrs) based on crossbred performance and within strain-selection (wss) based on purebred performance was made in three diverse environments over ten generations for the improvement of a heterotic trait, 4-day virgin egg lay of Tribolium castaneum. A selection intensity of 20% based on performance in either an optimum (33 degrees ), a mild stress (38 degrees ), or a severe stress (28 degrees ) environment was applied uniformly. Periodically, the performance of each population was measured in all three environments to provide both direct and correlated responses.-Heritability of egg lay in the base population ranged from 0.36 +/- 0.03 in optimum to 0.26 +/- 0.03 in severe stress. Estimates of dominance effects assumed significant proportions in severe stress only. Genetic correlations for egg lay in diverse environments were large and positive (.6 to.8).-Only in severe stress did the rrs response significantly exceed that for wss. Quadratic adjustments fitted to response curves revealed that small initial genetic gains under rrs were followed by significantly increasing rates of gain in late generations of selection. The reverse was true for wss. This and evidence from realized heritabilities and genetic correlations suggested that rrs had utilized both additive and dominance effects, but wss response was limited to additive effects.-These results agree with selection theory in demonstrating that purebred selection is more efficient than crossbred selection in utilizing additive gene effects. The latter method has merit when non-additive effects assume significant proportions, and this is the more probable case for severe stress conditions.