Survey of plant density tolerance in U.S. maize germplasm

Global demand for cereal crops like maize is rising at a rapid pace as the world population expands beyond 7 billion people. To meet these needs, productivity (i.e. grain yield) per unit area must be increased. A survey of U.S. maize germplasm was conducted to identify sources of favorable alleles for plant density tolerance and better understanding the genetics involved. Hybrids created using a genetically diverse set of inbreds representing parentage of key heterotic sub-groups were evaluated at plant densities ranging from 19,000 plants per acre (ppA) to 54,000 ppA. Five categories of traits were hypothesized to be associated with plant density tolerance: photosynthetic capability, growth responses, source-sink relationship, general stress tolerance, and plant architecture. Fifty phenotypic traits from these five categories were evaluated in three environments that differed for levels of moisture availability. The relationship between plant density and grain yield was assessed for each hybrid, with a wide range of responses observed. Five hybrids showed substantial tolerance to plant densities ≥47,000 ppA based on grain yield. Phenotypic trait correlations revealed a subset of traits associated with grain yield. Further analysis provided insight into relationships among traits that ultimately influence grain yield. All 5 categories of traits were found to have an association with grain yield directly and indirectly. Analysis of environments with differing moisture levels suggested that the 5 top-performing hybrids at high plant density have exceptional capacity for light utilization and translation of that energy into kernel mass. Estimates of heritability for grain yield at high plant densities were found to be similar to those at other plant densities, therefore requiring no alteration with breeding strategies used for new and improved maize lines. Results of this work will be used to create plant materials for further characterization of the trait through QTL mapping and candidate gene approaches