Plant Traits Related to Productivity of Maize. II. Development of multiple Trait Models1

Abstract

Grain yield and stalk lodging are the result of complex interactions of many traits expressed during maize (Zea mays L.) development. A better understanding of the most important traits influencing grain yield and stalk lodging could lead to improved production practices and more efficient breeding programs. To determine the most important combination of plant traits related to maize productivity, multiple regression models using plant trait means as independent variables, and grain yield or stalk lodging means as dependent variables were developed. Four groups of hybrids with different heterotic patterns and maturity were evaluated separately at 11 to 27 locations for grain yield and stalk lodging in 1984. The same groupings of hybrids were evaluated for 21 plant traits at one location (Urbana or St. Joseph, IL) in 1984 and 1985. Regression analysis showed that, dependent on the maturity group, from two to four of the selection traits would account for 46 to 95% of the variability among the hybrids for grain yield and stalk lodging. Across all maturity groups, a four trait model consisting of stalk dry weight at 15 days after anthesis, leaf P concentration at 15 days after anthesis, kernel number at maturity, and percentage of grain protein accounted for 40 to 93% of the variability among hybrids for grain yield. A five‐trait model consisting of plant carbohydrate to N ratio at the seven‐leaf stage, leaf P concentration at 15 days after anthesis, penetrometer readings of stalk rind strength at 40 days after anthesis, the percentage increase in stalk dry weight following ear removal, and the percentage of grain oil together accounted for 44 to 81% of the variability among hybrid means for stalk lodging. Although the role of a given plant trait varied with the particular hybrid, maturity group, and/or environment, the relatively high coefficient of determination values of the models indicates simultaneous measurement of several traits may prove useful in explaining maize productivity in specific hybrid groups.