Nitrogen management on coarse Sandy soils: A DSSAT-based mechanistic evaluation of controlled-release and conventional fertilizers for maize

CONTEXT Controlled-release fertilizers (CRFs) and conventional fertilizers both influence nitrogen (N) availability in coarse-textured soils, but their effectiveness depends on how fertilizer release or transformation aligns with crop N uptake and weather-driven N-loss processes. However, their agronomic and N-loss trade-offs across contrasting historical weather conditions remain insufficiently quantified in irrigated, coarse-textured soils. OBJECTIVE This study evaluated the ability of CSM-CERES-Maize to reproduce observed crop growth, grain yield, aboveground crop N, and soil NO₃ − –N responses under contrasting controlled-release fertilizer (CRF) and split-applied UAN management strategies and examined how these integrated N-management strategies affected simulated N partitioning and agronomic–environmental trade-offs across 36 independent historical weather-year simulations. METHODS The model was calibrated using 2023 field observations and independently evaluated using 2022 observations from irrigated maize, including leaf area index, aboveground biomass, yield, aboveground crop N, and 0–90 cm soil NO₃ − –N. CRFs were implemented with DSSAT's product-informed logistic controlled-release function. The evaluated model was applied to 36 independent historical weather-year simulations to derive crop N uptake, NO₃ − –N leaching, model-derived gaseous N losses, soil mineral N change, apparent N recovery efficiency and crop N capture relative to total modeled N input. RESULTS AND CONCLUSIONS Model agreement was strongest for grain yield and aboveground crop N, whereas biomass and soil NO₃ − –N performance was more variable during independent evaluation. Lower CRF rates generally maintained higher simulated N-use metrics and lower modeled NO₃ − –N leaching, whereas higher CRF rates produced greater grain yield and crop N uptake. However, scenario-based simulated NO₃ − –N leaching increased with higher CRF rates, from 17.0 ± 19.3 to 49.4 ± 47.7 kg N ha −1 . For CONV 269, simulated gaseous N loss was strongly dependent on UAN placement representation, decreasing from 64.05 ± 8.34 kg N ha −1 under the surface-band representation (S0; 0 cm) to 1.10 ± 0.11 kg N ha −1 under the shallow subsurface representation (S2; 2 cm). Thus, the magnitude and ranking of the simulated CONV 269 gaseous-loss response were conditional on the assumed UAN placement representation. SIGNIFICANCE Process-based modeling reveals trade-offs among integrated N-management strategies, showing that CRF performance under the evaluated site-specific conditions reflected the modeled interaction among the assumed fertilizer-release pattern, N rate, application schedule, and hydrologic variability. These findings may inform N-management decisions for irrigated maize grown under site conditions comparable to those evaluated in this study but should not be generalized to all coarse-textured production systems.