Evaluation of APSIM Next Generation for Simulating Winter Wheat Growth, Yield Response to Nitrogen, and Nitrogen Dynamics

Jose Cesario Pinto^1*Guillermo R Balboa^1*Nathan D Mueller²Glen Slater¹Katherine Frels¹Fernando E Miguez³Caio Dos Santos³Romulo Pisa Lollato⁴Laila Alejandra Puntel^1*

• ¹University of Nebraska-Lincoln, Lincoln, United States
• ²Mueller Ag Consulting, Lincoln, Nebraska, United States
• ³Iowa State University of Science and Technology, Ames, United States
• ⁴Kansas State University, Manhattan, United States

Process-based crop models, such as Agricultural Production Systems sIMulator Next Generation (APSIM-NG), can simulate crop growth, phenology, and yield under diverse conditions, making APSIM-NG a valuable tool for optimizing management practices and supporting climate-smart agriculture (CSA) strategies that enhance productivity and resilience. We evaluated APSIM-NG accuracy in simulating winter wheat cultivars' response to nitrogen (N) rates using field trials in Nebraska during the 2020/21 and 2021/22 growing seasons. The randomized complete block design included two cultivars (LCS and WB), four N rates (0, 56, 112, 168 kg N ha-1), and three replications. Wheat phenology, grain yield, protein content, shoot biomass, carbon-to-nitrogen ratio, soil nitrate (NO3-), ammonium (NH4+), soil moisture, and weather data were measured at each site. Data was used to calibrate cultivar phenology, biomass, yield, and protein content. The model was validated using yield data from 29 site-years combinations, representing five counties across Nebraska evaluated over six growing seasons (2017-2022). Accuracy was assessed through root mean square error (RMSE), relative root mean square error (RRMSE), and mean bias error (MBE). Calibration improved accuracy, showing well to moderate performance across cultivars for phenology (RRMSE = 2.1-2.2% and RMSE = 3-5 days), grain yield (15-24%), protein (8-11%), and grain N uptake (11-13%). The model moderately captured differences in leaf N uptake between cultivars, with RRMSE of 27% and 33% for LCS and WB, respectively. Validation performed well for both LCS (RRMSE = 14%) and WB (19%). APSIM-NG simulated yield response to N well for LCS (RRMSE = 18% for economical optimum nitrogen rate; EONR) but moderately for WB (32%). APSIM-NG showed well to moderate performance for predicting phenology, yield, and grain N dynamics across cultivars, indicating its value for guiding N management decisions. These results underscore APSIM-NG's potential to inform climate-smart management strategies that enhance N use efficiency and adaptation to climate variability in wheat systems.