Effects of China's Rice-animal Co-farming system on Rice Yield and Greenhouse gas emissions: a meta-analysis

Rice paddies are significant sources of greenhouse gas emissions. Rice-animal co-culture systems play an important role in improving rice yield and mitigating greenhouse gas emissions, with effects influenced by factors such as rice cultivation patterns, aquatic animal species, density, and fertilization methods. Based on 76 papers and 248 field experiment datasets published between 2000 and 2025, this study used Meta-analysis to investigate the effects of rice-animal co-culture systems on rice yield and greenhouse gas emissions in China. A random forest model was employed to quantify the contributions of various influencing factors. Compared with monoculture rice, the rice-animal co-culture system significantly increased rice yield (by 8.17%, P<0.05) and reduced CH₄ emissions (by 14.64%, P<0.05), but showed no significant effect on N₂O emissions (P>0.05). Among the co-culture systems, the duck model performed better in yield increase and emission reduction than the shrimp and fish models. Nitrogen input, stocking density, and soil and climate conditions varied in their effects on yield and emission reduction. Moderate nitrogen input (120~240kg∙hm^-2) achieved better yield-increasing and emission-reducing effects compared to low nitrogen input (0~120kg∙hm^-2) and high nitrogen input (≥240kg∙hm^-2). High stocking density of ducks (>450) might lead to decreased rice yield, while high stocking density of shrimp (≥500 kg) contributed to increased yield. An initial soil organic carbon content of 0~20g∙kg^-1 favored rice yield increase but was unfavorable for CH₄ emission reduction. Rice production and emission reduction were enhanced when precipitation was between 1000~2000mm, but temperature rise (mean annual temperature>20℃) was detrimental to both. According to the random forest model, the top three factors influencing rice yield were nitrogen input (31.02%), initial soil organic carbon content (19.13%), and stocking density (18.77%). The top three factors influencing CH₄ emissions were mean annual temperature (55.02%), annual precipitation (22.65%), and initial soil organic carbon content (12.14%). The results suggest that rice-animal co-culture systems in China are influenced by climate, soil properties, and field management practices. Under suitable soil and climate conditions, optimizing rice cultivation patterns and stocking density and applying nitrogen fertilizer reasonably can help achieve a dual win of rice yield increase and greenhouse gas emission reduction, thereby ensuring food security.