Abstract
Soil organic carbon (SOC) models currently in widespread use omit known microbial processes, and assume the existence of a SOC pool whose intrinsic properties confer persistence for centuries to millennia, despite evidence from priming and aggregate turnover that cast doubt on the existence of SOC with profound intrinsic stability. Here we show that by including microbial interactions in a SOC model, persistence can be explained as a feedback between substrate availability, mineral protection and microbial population size, without invoking an unproven pool that is intrinsically stable for centuries. The microbial SOC model based on this concept reproduces long-term data (r(2) = 0.92; n = 90), global SOC distribution (rmse = 4.7 +/- 0.6 kg C m(-2)), and total global SOC in the top 0.3 m (822 Pg C) accurately. SOC dynamics based on a microbial feedback without stable pools are thus consistent with global SOC distribution. This has important implications for carbon management, suggesting that relatively fast cycling, rather than recalcitrant, SOC must form the primary target of efforts to build SOC stocks.