Hybrid Performance of Sorghum Parental Lines Developed by Mass Selection and S1 Yield Testing

Experiments were conducted co obtain information on the effects of two methods of developing parental lines from a random-mating population of sorghum (Sorghum bicolor L. Moench) on hybrid performance. Gridded mass selection for threshed panicle weight was used in developing IAP1R(M)C3: and a sister population, IAP4R(S1)C3, was advanced each cycle on the basis of grain yield of S1 families in replicated yield trials. Hybrids with male parents developed by mass selection did not differ significantly for grain yield and panicles/plant from those with parents chosen on the basis of S1 yield tests. The S1-selection hybrids were significantly (P ≤0.01 or P ≤0.05) shorter and later to bloom, had smaller seed and more seeds/panicle, but the differences usually were small and of little practical consequence. Estimates of generic variance, heritability, and expected gain from selection among hybrids developed with mass or S1 selected male parents were similar for grain yield and most other characters. Collectively, the results of our experiments indicated that developing sorghum parental lines by using mass selection for panicle weight or replicated S1 yield tests should be equally effective