Soil carbon stocks across tropical forests of Panama regulated by base cation effects on fine roots

Tropical forests are the most carbon (C)-rich ecosystems on Earth, containing 25–40% of global terrestrial C stocks. While large-scale quantification of aboveground biomass in tropical forests has improved recently, soil C dynamics remain one of the largest sources of uncertainty in Earth system models, which inhibits our ability to predict future climate. Globally, soil texture and climate predict ≤ 30% of the variation in soil C stocks, so ecosystem models often predict soil C using measures of aboveground plant growth. However, this approach can underestimate tropical soil C stocks, and has proven inaccurate when compared with data for soil C in data-rich northern ecosystems. By quantifying soil organic C stocks to 1 m depth for 48 humid tropical forest plots across gradients of rainfall and soil fertility in Panama, we show that soil C does not correlate with common predictors used in models, such as plant biomass or litter production. Instead, a structural equation model including base cations, soil clay content, and rainfall as exogenous factors and root biomass as an endogenous factor predicted nearly 50% of the variation in tropical soil C stocks, indicating a strong indirect effect of base cation availability on tropical soil C storage. Including soil base cations in C cycle models, and thus emphasizing mechanistic links among nutrients, root biomass, and soil C stocks, will improve prediction of climate-soil feedbacks in tropical forests.