Evaluating Nitrogen Reduction Under Combined Rice Straw Biochar and Milk Vetch Application: A Multi-Objective...

Co-incorporating rice straw biochar (RSB) and milk vetch (MV) is a promising strategy for sustainable rice production. However, the appropriate nitrogen (N) reduction rate under RSB + MV, particularly when balancing multiple production objectives, remains unclear. This study evaluated five N-reduction rates (0-70% conventional N fertilizer) under RSB + MV. Grain yield, quality, nutrient uptake, and partial factor productivity of N fertilizer (PFP_N) were assessed in early and late rice. Entropy-weighted Technique for Order Preference by Similarity to Ideal Solution (EW-TOPSIS) model was used to integrate all indicators, when Gaussian model was applied to estimate the appropriate N-reduction rate. Pearson correlation and random forest analysis identified primary predictors of grain quality. Results showed that appropriate N-reduction (≤30%) under RSB + MV maintained or increased grain yield, quality, and PFP_N compared to conventional N fertilization alone. Among all treatments, N₈₀BM and N₇₀BM achieved the highest comprehensive scores in EW-TOPSIS analysis, and Gaussian model-estimated N-reduction of 18.2% and 21.0% for early and late rice, respectively. Nutrient uptake, effective panicle number and filled grains per panicle were identified as primary predictors of grain quality. Therefore, under the tested condition, RSB + MV with 20% N-reduction achieved the highest overall performance. However, these findings require further validation across multiple-year and multi-location.