Modeling Long-Term Corn Yield Response to Nitrogen Rate and Crop Rotation

Area of Science:

• Agronomy and Crop Science
• Agricultural Modeling
• Soil Science

Background:

• Optimizing nitrogen (N) fertilizer rates for corn (Zea mays L.) is critical for reducing environmental losses and enhancing farm profitability.
• Accurate simulation of crop yield and economic optimum N rate (EONR) requires robust crop modeling approaches.
• Understanding factors influencing year-to-year variability in N requirements is essential for adaptive management.

Purpose of the Study:

• To evaluate the Agricultural Production Systems sIMulator (APSIM) model's accuracy in simulating corn and soybean yields and EONR.
• To quantify model prediction accuracy before and after calibration using a 16-year field experiment dataset.
• To compare crop model-based techniques for estimating corn's EONR and identify factors driving yield and EONR variability.

Main Methods:

• Utilized a 16-year field experiment dataset from central Iowa, USA, with continuous corn and soybean-corn sequences.
• Applied five N fertilizer rates (0, 67, 134, 201, 268 kg N ha⁻¹) to corn and simulated yields and EONR using APSIM.
• Quantified model prediction accuracy using relative root mean square error (RRMSE) before and after model calibration.

Main Results:

• APSIM demonstrated good simulation of long-term crop yield response to N (RRMSE 19.6% before, 12.3% after calibration).
• EONR prediction showed higher uncertainty (RRMSE 44.5% before, 36.6% after calibration) than yield prediction.
• Calibrated APSIM is necessary for accurate year-by-year EONR simulation, revealing precipitation's impact on N loss and EONR.

Conclusions:

• Long-term experimental data are invaluable for refining and validating crop models like APSIM.
• Calibrated APSIM can serve as a decision-support tool for N management in the US Midwest.
• The model aids in understanding N loss mechanisms and supports agronomic, economic, and environmental sustainability goals.