Many approaches have been developed to estimate the optimal N application rates and increase nitrogen use efficiency (NUE). In particular, in-season and variable-rate fertilizer applications have the potential to apply N during the time of rapid plant N uptake and at the rate needed, thereby reducing the potential for nitrogen fertilizer losses. However, there remains great challenges in determining the optimal N rate to apply in site-specific locations within a field in a given year. Additionally, the benefit to in-season N application has not been well quantified in some regions. Crop growth models may be used to explain dynamics that influence optimal N rate and make predictions of future N need. 

Two non-irrigated, on-farm experimental sites were established in 2020 to evaluate N rate and timing. One site is a bottom field with shallow water table and tile drainage while the second site is an upland field with contour farming and terraces. Within each site, soil type, elevation, landscape position, yield history, and historical imagery were used to delineate zones. Zones were characterized with detailed soil sampling for texture and organic matter by depth. Intensive sampling consisting of soil moisture, temperature, ammonium, and nitrate, and leaf area index (LAI) was conducted to calibrate the APSIM model.  

In this work, we quantified the APSIM model prediction accuracy to predict the economic optimum N (EONR) in contrasting zones within a field in Nebraska and we described the steps to differentiate zones within the field in APSIM setup. The calibrated, zone-differentiated model was used to explain zone specific crop and soil dynamics that influenced variability in yield, yield response to N, N losses and mineralization, and EONR. Additionally, we evaluated the ability of the model to predict in-season N need and the calibrated model was used to simulate yield response to N over 30 years and determine the benefit of in-season applications.