Twelve Cycles of Mass Selection for Prolificacy in Maize I. Direct and Correlated Responses

Abstract

Selection for prolificacy in maize, Zea mays L., has been recommended as an effective method to improve grain yield. The objective of this study was to evaluate selection response from 12 cycles of biparental mass selection for prolificacy in the Golden Glow maize population as measured under different plant densities and fertility levels. Genotypes evaluated were Cycles 0, 3, 6, 9, and 12, along with testcrosses of Cycles 0, 6, and 12 with three inbred lines, A632, W64A, and W9. The evaluation was conducted in 1985 in low‐fertility plots at two locations and three plant densities (38 271, 53 597, and 68 887 plants ha−1). The evaluation was repeated in 1986 at two fertility levels (no added fertilizer and 168 kg N ha−1 with 37 kg K ha−1). Number of ears plant−1 in the cycles per se increased by 2.4 and 3.3% cycle−1 in 1985 and 1986, respectively. Corresponding increases in grain yield plant−1 were 2.0 and 3.0% cycle−1, and increases in grain yield ha−1 were 2.0 and 2.8% cycle−1. Grain moisture, flowering dates, and the period between silk emergence and anthesis decreased with selection for cycles per se. Testcross responses were approximately one‐half those of cycles per se for ears plant−1 and either measure of grain yield. Grain moisture did not change with selection, and days to flowering increased by 1 to 2 d in the testcrosses. The correlations between ears plant−1 and grain yield plant−1 were 0.90 and 0.79 for cycles per se and testcrosses, respectively (both significant at P ≤ 0.05). Response to selection for ears plant−1, as reflected in the regression of ears plant−1 on cycle of selection, was greater under optimum fertility (b = 0.032 ± 0.002) than low fertility (b = 0.022 ± 0.003). Selection response for ears plant−1 was progressively greater as plant density decreased (b = 0.016 ± 0.003, 0.027 ± 0.003, and 0.036 ± 0.003 for 68 887, 53 597, and 38 271 plants ha−1, respectively).