Effects of long-term combined application of organic and chemical fertilizers on bacterial community characteristics and soybean yields

Returning organic materials is the main agricultural practice used to improve soil microbial diversity and quality. However, the mechanisms underlying the effects of organic material addition on key soil bacterial taxa and the relationship between their effects and soybean yield have not yet been fully elucidated. The correlations between different fertilization practices and soil chemical properties, bacterial communities, and soybean yields were comprehensively evaluated based on a 44-years long-term experiment in Mollisols of Northeast China. The results showed that organic fertilization (M and CFM treatments) significantly increased soil nutrient content, microbial biomass carbon content and microbial biomass nitrogen content, in contrast to the no fertilization treatment (NoF treatment). Compared to chemical fertilization (CF treatment), organic fertilization plus chemical fertilization (CFM treatment) significantly increased soybean yield by 61.91%, and the S treatment significantly decreased by 9.54%. Soil bacterial communities showed significant correlations with total carbon (TC), total nitrogen (TN), total potassium (TK), ammonium nitrogen (NH₄⁺-N), nitrate nitrogen (NO₃^–-N), available potassium (AK), microbial biomass carbon (MBC), microbial biomass nitrogen (MBN), and the C/N ratio. The S and CFM treatments significantly increased the richness and diversity of soil bacterial communities compared to the CF treatment. The richness and diversity of soil bacterial communities in chemical fertilization plus straw returing (CFS treatment) were lower than those in the S treatment, whereas they were higher in the CFM treatment than in the M treatment. The application of fertilizers had a significant impact on the structure of bacterial communities. Significant differences were observed between the M and CFM, S and NoF, and CFS and CF treatments. The bacterial community composition showed higher similarity between the S and NoF treatments, as well as between the CFS and CF treatments. The results of LEfSe analysis showed that there were five, one, seven, four, and four distinct groups in the NoF, S, M, CF, and CFS treatments, respectively. These groups were identified as norank_f_norank_o_Saccharimonadales, Alphaproteobacteria, Vicinamibacterales, Proteobacteria, and Acidobacteria. Fertilization altered the life strategies of soil microorganisms and significantly increased the bacterial abundance of the r-strategy taxa, resulting in a higher ratio of eutrophic bacterial taxa (r-strategy) to oligotrophic bacterial taxa (K-strategy) compared with the NoF treatment. Long-term straw returning (S and CFS treatments) and chemical fertilization (CF treatment) increased bacterial network complexity, whereas it decreased with organic fertilization (M and CFM treatments). Structural equation modeling (SEM) revealed a significant positive correlation between soil ammonium nitrogen, C/N ratio, bacterial network structure complexity, abundance of key taxa (Nitrospira and Gemmatimonas), and soybean yields. Soil available potassium can indirectly affect soybean yield by altering the complexity of bacterial networks and the abundance of key taxa. The taxa Nitrospira and Gemmatimonas play crucial roles in nitrification and phosphate metabolism, respecabolism, respectively, and those of two taxa can be used as indicator species to assess soil fertility. In conclusion, organic and chemical fertilization increased the abundance of key beneficial microbial taxa, bacterial community diversity, and soybean yield.