Peihua Shi ¹ Yuan Wang ² Congfei Yin ^1* Kaiqing Fan ^1* Yinfei Qian ^3* Gui Chen ^4*

• ¹ Jiangsu Vocational College of Agriculture and Forestry, Jurong, China
• ² Institute of Soil Science, Chinese Academy of Sciences (CAS), Nanjing, Jiangsu Province, China
• ³ Soil and Fertilizer & Resources and Environmental Institute, Jiangxi Academy of Agricultural Sciences, Nanchang, China
• ⁴ Institute of Biotechnology, Jiaxing Academy of Agricultural Science, Jiaxing, China

Nitrogen is vital for rice growth and yield, but traditional methods of assessing nitrogen status are labor-intensive or inaccurate at high nitrogen levels due to saturation effects. This study evaluates the effectiveness of flavonoid content (Flav) and the Nitrogen Balance Index (NBI), measured with a Dualex sensor and combined with machine learning models, for estimating rice nitrogen status. Field experiments involving 15 rice varieties provided Dualex measurements from the top five leaves. An incremental analysis was used to quantify saturation effects in nitrogen estimation. Results showed that chlorophyll measurements saturated at high nitrogen levels, limiting their reliability. In contrast, Flav and NBI remained sensitive across all nitrogen levels, accurately reflecting rice nitrogen status. Machine learning models, particularly random forest and extreme gradient boosting, achieved high accuracy in predicting leaf and plant nitrogen concentrations (R² > 0.82). SHAP analysis identified NBI and Flav measurements from the top two leaves as the most significant predictors, enhancing model interpretability. Combining Flav and NBI measurements with machine learning effectively overcomes the saturation limitations of chlorophyll, offering precise estimation of rice nitrogen status under varying conditions. These findings suggest promising approaches for precise nitrogen management in rice cultivation, enhancing agricultural efficiency and sustainability.