Two cycles of S1 per se recurrent selection were conducted in four populations of corn (Zea mays L.). The primary selection criterion was a performance index: grain yield divided by percent moisture at harvest. The original (C0) source material and first cycle (C1) of selection for each population were evaluated in a favorable (1977) and an unfavorable (1978) growing season. Second cycle (C2) materials were also included in the 1978 comparisons. In 1977, C1 yielded significantly more grain than C0 in three of the four populations. In contrast, performance of C1 and C2 materials in 1978 was inferior to C0 in all populations but one. The advanced cycles of only one population showed improvement over C0 in both years. Comparisons of the 1977 data for days to silking, grain yield, and percent ear moisture at harvest suggest that rate of grain filling in C1 was higher than in C0 for all populations. The higher rates of grain filling in the advanced cycles may have caused a carbohydrate deficiency under stress (1978) so that the leaves no longer functioned normally and kernel filling was terminated prematurely. In general, the four populations showed inherent differences in their respective responses to selection, to environmental stress, and to inbreeding.
Half Sib Family Selection for Population Improvement in Maize (Zea Mays L.)