A multi-environment trial analysis shows slight grain yield improvement in Texas commercial maize

Grain yield in maize has increased steadily in the Midwestern United States since the 1930s but trends are not clear for the maize production zones in the southern states, of which Texas is the major producer. The objective of this study was to elucidate trends of maize yields in Texas for the last 11 years and to identify any significant patterns in breeding and testing maize that can be exploited for improvement. A linear mixed model applied to 107 corn performance trials grown involving 11 years and 16 locations in Texas identified that the largest G x E variance for the multi-environment trial (MET) dataset was the three-factor interaction model that included hybrid, location, years and their interactions. Repeatability measures showed that two replicates and two years provided accurate estimates of genetic yield potential. Trial data and a separate analysis of county level yield data corroborated that there are two separate production zones in Texas, the High Plains and the rest of the state. Across these production zones there have been no substantial gains in yield and surprisingly some older commercial hybrids appear to have had better yield potential then what is currently commercially available. Plant height, ear height, plant population and test weight were positively correlated with grain yield with stronger effects observed in the rest of Texas than in the High Plains; lodging was negatively correlated with grain yield while days to silk showed little effect. This suggests separate strategies for additional genetic and trial improvement should be used in the High Plains and the rest of Texas. (C) 2013 Elsevier B.V. All rights reserved.