Assessment of soybean yield variability in the southeastern U.S. with the calibration of genetic coefficients from variety trials using CROPGRO‐Soybean

Soybean (Glycine max [L.] Merr.) is a major crop cultivated in the United States, and it is well adapted to different latitudes and weather conditions. However, grain yield variability of rainfed soybean across early and late sowing dates and among maturity groups (MGs) has not been well evaluated in the southeastern United States. This study uses crop variety trials and soil reports to calibrate the CROPGRO‐Soybean model and then assesses yield variability for different sowing dates and MGs across sites within the southeastern United States. After model calibration and evaluation, soybean yield was simulated across 36 yr at eight sites for 16 sowing dates and three cultivars from MGs V, VI, and VII. The results show that the model accurately simulated soybean yield for MGs across different soil types, where the root mean square error (RMSE) across cultivars ranged from 423 to 589 kg ha−1 and across sites ranged from 219 to 684 kg ha−1. This facilitated the assessment of sowing date effects on yield variability, where simulations showed high yields for sowing dates from April to May, but with higher variability for April sowing dates and low grain yield for sowing dates in July. Our approach successfully illustrates the use of soil reports and crop variety trials to assess the best sowing dates and MGs for soybean production in this region. Core Ideas Cultivar and soil parameters were calibrated in the DSSAT with variety trials and soil reports. Long‐term yield simulations were conducted with CROPGRO‐Soybean for multiple maturity groups and sowing dates. Yield variability was mitigated with cultivar and sowing date selection in the southeastern U.S.