Estimating CSM-CERES-Maize genetic coefficients and soil parameters and evaluating model response to varying nitrogen management strategies under North Carolina conditions

CSM-CERES-Maize has been extensively used to simulate corn growth and grain production in various locations worldwide, but has not been evaluated previously for use in North Carolina. The first objective of this study were to calibrate CSM-CERES-Maize soil parameters and genetic coefficients using Official Variety Trial data from 60 site-years for 53 maize genotypes, and to determine the suitability of the fitting technique and variety trial data for model calibration. A stepwise calibration procedure with grid search algorithm was utilized: (1) two genetic coefficients which determine anthesis and physiological maturity dates were adjusted based on planting date and growing degree day requirements for each hybrid; and (2) plant available soil water and rooting profile were adjusted iteratively with two genetic coefficients affecting yield. Cross validation was used to evaluate the suitability of this approach for estimating soil parameters and genetic coefficients.;Results indicate that the CSM-CERES-Maize model can be used in North Carolina to simulate corn growth under non-limiting nitrogen conditions and Official Variety Trial data can be used to estimate genetic coefficients, although the CSM-CERES-Maize over-estimated yield for low yield environments and under-estimated it for high yield environments for some hybrids.;The second objective of this study was to examine the ability of the CSM-CERES-Maize model to simulate corn response to varying irrigation and nitrogen application strategies. Yield data for a total of 88 irrigation/nitrogen treatments with only one cultivar (Pioneer 31G98) from three fields in Lewiston, North Carolina were available for comparison. Procedures were: (1) develop realistic soil profiles for the three fields; (2) compare simulated CSM-CERES-Maize corn yields to measured yields for all 88 treatments; (3) adjust soil parameters in an iterative process in order to improve simulation of corn yields for these treatments; and (4) determine the importance of each soil parameter to simulated crop yields.;Simulated yields did not match observed yields well using our initial soil profiles, with Relative Root Mean Square Error (RRMSE) values of 17.5, 38.4, and 50.1% for the three fields. The iterative adjustment of soil parameters was successful in determining a set of soil parameters for each field such that the RRMSE values for yield improved to 8.2, 7.8, and 7.4%, respectively. Simulated yield using these optimized parameters generally fell within ±Standard Error (SE) of the measured yield. The soil fertility factor, SLPF, ranged from 1.27 to 1.34 for these fields, much higher than the default value of 1.0. SRGF, the root growth factor, also had a very different pattern than the expected exponential pattern, which begins with a value of 1.0 in the top 15 cm of soil and declines to 0.078 by 135 cm. The optimized pattern of SRGF for all three fields started with a value of 0.1 in the layers above 45 cm, with larger values in the deeper layers.;The importance of each adjusted soil parameter was investigated by setting it back to its starting value while the other adjusted parameters were left at the optimized value. When SRGF was returned to an exponential pattern, simulated yields for irrigated treatments which received a side dressing of N at visual tasseling were lower than those for an irrigated treatment which did not receive this second application. Because new root length is distributed across the soil profile by the model, we recommend necessary changes to CSM-CERES-Maize in order for the model to be used to predict crop response to split applications of N.