Reducing Corn Yield Variability and Enhancing Yield Increases Through the Use of Corn-Specific Growth Models.

Abstract

Crop simulation models (CSMs) are used to evaluate management and environmental scenarios on crop growth and yields. Two corn (Zea Mays L.) crop growth simulation models, Hybrid-Maize, and CERES-Maize were calibrated and validated under Virginia conditions with the goal of better understanding corn response to variable environmental conditions and decreasing temporal yield variation. Calibration data were generated from small plot studies conducted at five site-years. Main plots were plant density (4.9, 6.2, 7.4, and 8.6 plants m-2); subplots were hybrids of differing relative maturity (RM) [early = Pioneer® Brand â 34B97â (108 day RM); medium = Pioneer® Brand â 33M54â (114 day RM); and late = Pioneer® Brand â 31G66â (118 day RM)]. Model validation was generated from large scale, replicated strip plot trials conducted at various locations across Virginia in 2005 and 2006. Prior to model adjustments based on calibration data, both CSMs under predicted corn grain yield in calibration and validation studies. CERES-Maize grain yield prediction error was consistent across the range of tested plant density while accuracy of Hybrid-Maize varied with plant density. Hybrid-Maize-estimated biomass production was highly accurate. Greater leaf area index (LAI) and biomass production were measured than was predicted by the CERES-Maize CSM. Both CSMs were modified based on calibration data sets and validated. Validation results of the calibrated CSMs showed improved accuracy in simulating planting date and environmental effects on a range of corn hybrids grown throughout Virginia over two years. We expect that both modified models can be used for strategic research and management decisions in mid-Atlantic corn production.