Responses of soil microbial community diversity to thinning and replanting, and their effects on microbial necromass carbon accumulation efficiency.

Thinning and replanting of broadleaf trees (TRBT) significantly affect stand structure and soil functions, but their impacts on soil microbial community diversity and enzyme activity remain poorly understood. This study evaluated five different TRBT treatments: unthinning plot (NT), thinning without replanting (NP), thinning with replanting of Schima superba (CS), thinning with replanting of Liquidambar formosana (CL), and thinning with replanting of both Schima superba and Liquidambar formosana (CSL). We analyzed soil microbial community structure, diversity, and enzyme activity, and explored their relationships with microbial necromass carbon (MNC) accumulation efficiency. TRBT mainly affected bacterial diversity, while having minimal impacts on fungal diversity and richness. TRBT significantly altered the β-diversity of both bacterial and fungal communities. The Acidobacteriia, Alphaproteobacteria, Deltaproteobacteria, Ktedonobacteria, and Gammaproteobacteria were the dominant bacterial classes, while the Agaricomycetes were the dominant fungal classes. TRBT reduced the coefficients of β-1,4-glucosidase (BG), cellobiohydrolase (CBH) and acid phosphatase (AP), but increased the coefficients of β-1,4-N-acetylglucosaminidase (NAG) and leucine aminopeptidase (LAP). Meanwhile, the soil bacterial diversity and coefficients of NAG and LAP in CS and CL treatments were higher than those in CSL, whereas the coefficients of BG and CBH in CS treatment were lower than those in Cl and CSL. The MNC accumulation coefficient was significantly positively correlated with bacterial diversity and NAG coefficient, but negatively correlated with coefficients of BG, CBH, and AP. Mantel test revealed that microbial biomass carbon (MBC), phosphorus (MBP), enzyme activity coefficients (CBH, AP, NAG), bacterial diversity, and Alphaproteobacteria (bacterial) were key factors influencing the MNC accumulation coefficient. Random forest analysis further confirmed that MBP, coefficients of AP and NAG, and MBN were important variables for predicting the MNC accumulation coefficient. Our findings provide a valuable reference for optimizing tree species selection to improve soil carbon sequestration in plantation forests.