Paddy soil fertility shifts microbial metabolic quotient by regulating the selective enrichment of specific bacterial groups

• Soil microbial metabolic quotient (MMQ) is lower under high-fertility conditions (HF). • HF soils exhibit greater enrichment of carbon-conserving groups compared to low-fertility soils. • Soil pH, AP, stoichiometric ratios, BD, and SWC as environmental predictors of carbon-conserving groups in HF. The heterogeneous distribution of specific microbial taxa with diverse physiological traits, caused by changes of soil fertility level, controls community-level functions. However, how specific microbial taxa mediate community-level carbon (C) metabolic efficiency across various fertility conditions in paddy soils is limited. Herein we surveyed 114 paddy soils with distinct fertility conditions to assess the potential roles of specific bacterial taxa in regulating the soil microbial metabolic quotient (MMQ). Lower MMQ values were observed in high-fertility soils compared to low-fertility soils, indicating higher metabolic efficiency in high-fertility conditions. Our study further advanced that MMQ can be significantly predicted by the abundance of 30 specific bacterial operational taxonomic units (OTUs) using random forest mechine learning model. Compared to low-fertility conditions, high-fertility conditions exhibited an enrichment in the abundance of predictive OTUs classified into C-conserving groups, while depleting the abundance of OTUs classified into C-consuming groups. Furthermore, higher soil stoichiometric ratios (C:N, C:P, and N:P ratio), soil organic C (SOC) content, soil water content and lower pH, bulk density favored the enrichment of C-conserving bacterial groups under high-fertility conditions. Additionally, SOC content and cation exchange capacity significantly correlate with the abundance of C-consuming groups in low fertility. Taken together, this study reveals significant correlations between specific bacterial taxa and soil MMQ, highlighting how these taxa may influence C cycling processes under different fertility conditions, also suggesting their potential utility as bioindicators for precision paddy soil management strategies.