Soil aggregate fragmentation reduces microbial diversity but enhances oxidative metabolism that drives carbon loss

Widespread soil organic carbon (SOC) loss under accelerating anthropogenic activities threatens both soil biodiversity and carbon persistence. Carbon decline is often accompanied by soil structure disruption, reorganizing the microhabitats for microbe-carbon interactions toward smaller aggregate domains. A key question is how such structural reorganization affects microbial regulation of soil carbon turnover. Here, we used twelve Mollisol croplands along an agricultural chronosequence of SOC decline to trace microbial diversity and carbon decomposition in relation to changes in aggregate structure. As SOC declined, aggregate-size distribution shifted progressively toward a greater contribution of microaggregates, which consistently harbored lower microbial diversity than macroaggregates. Aggregate-specific diversity responses differed markedly between aggregate fractions: macroaggregates showed persistent diversity loss, whereas microaggregates exhibited delayed but accelerated loss under stronger SOC depletion. These contrasting trajectories underlay the nonlinear decline in bulk-soil diversity, with limited change at higher SOC levels but sharp losses below a threshold of 27‒29 g C kg −1 . Despite declines in microbial diversity and SOC stocks, mineralization of both added 13 C-labelled glucose and vanillin increased in bulk soils. Notably, this acceleration occurred primarily within microaggregates, where genes associated with oxidative carbon degradation and respiratory metabolism were enriched under low-SOC conditions, while substrate mineralization within macroaggregates remained relatively insensitive to SOC decline. Overall, our findings reveal that aggregate fragmentation drives SOC-depleted soils toward a low microbial diversity yet high carbon turnover state, primarily by increasing the importance of microaggregate domains with enhanced oxidative metabolism. This shift may increase the vulnerability of remaining soil carbon to further loss. Incorporating microscale soil structure and associated biological processes is therefore critical for predicting soil carbon persistence and designing management strategies to conserve biodiversity and rebuild SOC stocks.