Abstract
INTRODUCTION: Soybean-based intercropping has emerged as a key sustainable agricultural practice, which enhances system productivity and improves soil health. Although numerous studies have investigated soybean yield under intercropping systems, the responses of soil microbial community traits and their associations with yield remain unclear. METHODS: Based on 323 observations extracted from 89 peer-reviewed articles, we conducted a meta-analysis to investigate the responses of soil microbial community traits and crop yield to soybean-based intercropping. RESULTS: Compared to monoculture, our analysis revealed that soybean-based intercropping did not significantly affect soybean yield (effect size = -0.020, 95% CI: -0.105 to 0.065), but significantly changed soil microbial communities. The practice significantly enhanced microbial community α-diversity indices, including Shannon (effect size = 0.036, 95% CI: 0.020 to 0.053), Chao1 (effect size = 0.034, 95% CI: 0.002 to 0.067), and richness (effect size = 0.102, 95% CI: 0.026 to 0.178), while also significantly altered microbial community structure (effect size = 0.741, 95% CI: 0.629 to 0.852). Random forest analysis identified nitrogen fertilization rate as the primary predictor of α-diversity indices. And nitrogen fertilization rate showed a significant negative correlation with Chao1 (R (2) = 0.051, p = 0.079) and a marginally significant negative correlation with richness (R (2) = 0.141, p = 0.018). Nitrogen fertilizer type significantly affected soil microbial α-diversity, with mixed nitrogen fertilizers demonstrating greater effects on both Chao1 (effect size = 0.111, 95% CI: 0.034 to 0.188) and richness (effect size = 0.192, 95% CI: 0.038 to 0.345) than mineral fertilizers. Intriguingly, while we only found a marginally significant positive correlation between Chao1 and soybean yield (R (2) = 0.131, p = 0.082), yield advantages over monoculture emerged within a specific Shannon index range from 0.008 to 0.401. Given the lack of a direct association between nitrogen fertilization rate and soybean yield (r = 0.056, p > 0.05), its influence on soybean yield in soybean-based intercropping may mediated by the changes in microbial community diversity. CONCLUSION: Collectively, these findings reveal soil microbial responses to soybean-based intercropping and highlight the potential for optimizing microbial communities to enhance soil health and agricultural sustainability in intercropping systems.