Simulating Dryland Cotton and Sorghum Production over the U.S. Southern High Plains

Over semi-arid agricultural regions such as the U.S. Southern High Plains (SHP) producers of dryland crops need to know which management practices increase yields and decrease production risk. Here, an in-silico modelling approach is used to explore management options (MO) that increase dryland cotton and sorghum yields and estimate those practice's yield risk effects under current SHP climate conditions. To simulate current dryland yield variability, dense distributions of yield outcomes for both crops were generated using a crop model driven by weather inputs from 21 SHP weather stations during 2005-2016. Management effects were explored by repeating simulations over 32 MOs defined by 4 planting dates, 4 planting densities, and applying or not applying nitrogen. For both crops decreased plant density increased median simulated yields, although these effects are inconsistent with those found in regional field studies. By contrast, cotton and sorghum planting dates effects are consistent with SHP climate conditions. Mid-May cotton planting dates increased lint yields relative to early-June planting, which is consistent with field studies and is basically a consequence of the SHP region's cool growing conditions and short cotton growing season. The sorghum simulations show that early July planting dates increase median sorghum yields by causing a crop's peak water-demand period to coincide with the region's early-Fall wet period. By generating dense distributions of yield outcomes consistent with current SHP summer climate conditions, this simulation approach provides a basis for estimating climate-related risk in dryland agriculture.