Selection Criteria for Grain Yield and Moisture in Maize Yield Trials

Abstract

Grain yield and moisture in maize (Zea mays L.) are positively correlated, and must be considered simultaneously. Typically, breeders desire to identify better‐yielding hybrids within each of several maturity groups. Data from 28 multiple‐location yields trials were used to (i) compare five indices for ability to identify hybrids that are high yielding for their maturity; (ii) investigate the validity of assumptions used in developing estimated indexes; and (iii) investigate possible relationships between effectiveness of indexes and violations of assumptions. Selection criteria evaluated included: BSE, a base index; S‐H, a Smith‐Hazel index; Y/M, the ratio of yield and moisture; REG, yield adjusted for the regression of yield onto moisture; and DSG, a desired gains index designed to hold moisture constant. Departures from normality for distribution of grain yield means and for moisture were detected in 18 and 32% of the experiments, respectively. The regression of yield onto moisture varied considerably across experiments, with a mean coefficient of 12.5 kg ha−1 yield per g kg−1 moisture. Selection differentials of REG, BSE, and DSG for yield were 92, 91, and 89% respectively of that resulting from direct selection for yield, but changes for moisture were considerably less than those for selection based on yield. The Y/M selection differential for yield was only 64% of that of yield, and had a large decrease in grain moisture. No association between the success of an index and nonnormality was observed. Both BSE or REG are recommended for identification of hybrids that are high yielding for their maturity. However, these or other indices employed should be carefully evaluated in the context of the breeder's circumstances and objectives.