Microbial activity and root carbon inputs are more important than soil carbon diffusion in simulating soil carbon profiles

It is well-known that soil carbon composition and turnover rate vary with soil depth, and the responses of soil carbon to global change in deeper soil layers may differ from those near the surface. Therefore, vertically resolved soil carbon models are needed for accurately predicting future soil carbon under global warming. In this study, we developed a vertically resolved soil carbon model by including vertical transport of soil carbon and using Michaelis-Menten kinetics for soil carbon decomposition by microbes. The model was calibrated against six sites with the observed profiles of both soil carbon concentration and 14C measurements, and against 91 forest sites with soil carbon concentrations across a wide range of climate and soil conditions for four forest types in Europe and China. Results of independent model validation at another 93 sites showed that the calibrated model explained 40%–94% of the observed variance of soil carbon concentrations at different depths. Model sensitivity analysis showed that microbial activity and root carbon inputs are more important than soil carbon diffusion in simulating soil carbon profile. Results from our study highlight the need for detailed measurements of soil microbial activities and root carbon input at different soil depths in the field.