Optimizing recommended nitrogen fertilizer rates for flue-cured tobacco using the DNDC model

Achieving precise and efficient nitrogen fertilization for flue-cured tobacco remains a critical challenge, as recommendations derived from traditional field trials often have limited applicability across different regions. To develop a more efficient approach for determining nitrogen fertilizer recommendations, this study calibrated and validated the DNDC model using six site-years of field data (2019–2023) from Weifang, Shandong, to optimize N fertilization strategies. The calibrated DNDC model accurately simulated tobacco yield across varying N rates (R 2 = 0.903). A linear-plus-plateau model fitted to the simulated yield–N rate relationship effectively captured key agronomic parameters, including maximum yield, yield increase efficiency, baseline yield, and the site-specific recommended N rate. The strong agreement between DNDC-simulated and field-observed parameters (R 2 = 0.892, 0.848, 0.902, 0.718) confirmed the model's robustness in representing N–soil–yield interactions. Subsequently, 2100 scenario simulations were conducted, encompassing five soil organic carbon (SOC) levels, seven initial inorganic N levels, and twelve N application rates under conditions of five different experimental years. Regression analysis of the simulation dataset revealed a linear relationship between the recommended N rate and both SOC and initial inorganic N content：The optimum nitrogen fertilizer rate declines as soil organic carbon content and soil organic matter increase. The derived regression equation was validated against independent field measurements, demonstrating high predictive accuracy (R 2 = 0.91, RMSE = 9317 kg ha⁻¹, NRMSE = 11.60%). We conclude that the localized DNDC model is a reliable tool for simulating flue-cured tobacco yield, and the empirical equation provides a robust, site-specific method for determining precise N fertilization rates in the study region and under similar agroecological conditions.