Earth system models vary in their predictions of carbon (C) uptake and release by the terrestrial biosphere, partly due to great uncertainties in the response of soils, one of the largest C reservoirs. The world’s soils play a major role in the exchange of greenhouse gases with the atmosphere, in sustaining primary production, and in providing food security. Despite this, the sensitivity of soils to disturbance is highly uncertain. One reason for this is geographic variability in the importance of different mechanisms regulating soil C turnover. Most of our understanding of factors influencing soil organic C dynamics comes from research in temperate soils, despite the major role of tropical soils in the global C cycle. Even in the tropics, the diversity of soil environments is grossly underrepresented in the literature. This has important implications for predictions of soil C change across latitudes. We discuss results from the response of soil C pools and microbial communities to land use legacies on two contrasting tropical soil environments.
Uncertainties in the response of soil C to disturbance also stem from a historic focus on shallow depths and the assumption that deep soil C is unreactive to landscape change. Growing evidence indicates that soil C pools in deep mineral horizons can be sensitive to changes in land cover and climate. This realization highlights the need to reassess the source of soil C at depth and the processes contributing to its stabilization. We discuss results from the interaction between multiple disturbances: drought, fire and erosion, on the accumulation of soil C at depths beyond those typically included in regional or global inventories. Our data show that deep soil C can be reactive and be a potential source of C if reconnected to the atmosphere. A deeper, mechanistic appreciation for a landscape’s history of disturbance is critical for predicting feedbacks between the terrestrial biosphere and the climate system.