Evaluating short-season soybean management adaptations for cover crop rotations with a crop simulation model

Cover crop fall biomass production and thus successful provisioning of ecosystem services depend on the previous cash crop harvest date. We used a process-based eco-physiological model to investigate the potential of short-season soybean maturity groups (MG) to lengthen the cover crop growing window while achieving yields similar to full-season MG cultivars. Cultivar coefficients for MG 0–4 cultivars for the DSSAT – CROPGRO model were calibrated with data from 13 site-years (in 2017 and 2018) across Kentucky, Nebraska, and Ohio. The model was efficient in predicting differences in soybean harvest maturity date (R8; Model efficiency [ME] = 0.61; Root Mean Square Error [RMSE] = 7.4 days) and yield (ME = 0.38; RMSE = 0.452 Mg ha⁻¹) for an independent set of soybean cultivars in the same site-years. Thereafter, a multi-factor sensitivity analysis across 30-yr of historical weather data was conducted. Simulated results showed that MG 3 cultivars would not reduce yield and would advance cover crop establishment compared to MG 4 cultivars. For planting dates in May and conditions of no water stress, adaptating cultivar choices to MG lower than 3 would reduce yields by 55 to 567 kg ha⁻¹ per unit decrease in MG. Under water stress or when planting date was delayed, adapting cultivar choices to MG lower than 3 had a less detrimental effect on yield. Overall, switching to earlier cutlivar maturities would advance soybean harvest by 7–11 days MG⁻¹ (May 15 planting date) or 1–7 days MG^-1 (Jul 1 planting date), and lengthen the cover crop growing season in the fall by 95–198 °C day MG⁻¹ (May 15 planting date) or 19−104 °C day MG⁻¹ (Jul 1 planting date). The greater potential to increase the cover crop growing season with short-season MG cultivars was also associated with a greater soybean yield penalty in the warmest locations in our study. Using crop coefficients calibrated by MG rather than by specific cultivar provided a way to increase model application within a study region to study cultivar maturity adaptations for crop rotations while reducing the need for calibration. Further studies that analyze the tradeoffs from soybean cultivar adaptation on C, N, and water balance, and other indirect ecosystem services from cover crops are necessary.