Soil carbon vulnerability to land-cover change and implications for the global carbon cycle

Abstract:

Soil is a major reservoir of carbon (C) that contains more than three times the C in vegetation. While the terrestrial biosphere acts as a sink for approximately 2.5 petagrams C per year (PgC/yr) – equivalent to about 25% of fossil fuel emissions – anthropogenic land-use change reduces the global net land C sink by approximately 1 PgC/yr. Although most assessments of land-use focus on changes in aboveground biomass C, changes to ‘live C’ (above- and below-ground plant biomass) drive lagged, yet substantial, changes to ‘dead C’ (soil, dead wood, and litter) storage, with important implications for the overall land C balance.

Here we provide an observation- and model-based assessment of the impacts of land-cover change on total C stocks (live and dead C) over the last decade and the potential for long-term soil C storage or loss. We find that afforestation in northern latitudes counteracts deforestation in the tropics, due in part to the greater soil C content of boreal and temperate forests. Deriving geospatial estimates of the steady-state ratio of dead to live C and the turnover time of dead C, we demonstrate that, although many recent studies have focused on forests, non-forest ecosystems (particularly shrublands and savannahs) may be responsible for larger changes in total C stocks in response to changes in C input rates due to their high soil C content. Furthermore, we disaggregate our geospatial predictions by biome and explore the depth-resolved vulnerability of soil C globally, with particular focus on the covariation of soil C sensitivity with soil properties and climate. Our findings suggest that the response of soil to changes in plant inputs significantly contributes to regional and global C budgets.