Characterizing land use impacts on carbon stocks and tree cover in the tropics

Abstract:

Tree cover (TC) and biomass carbon stocks (CS) are key parameters for characterizing vegetation dynamics and indispensable for assessing the role of terrestrial ecosystems in the global climate system. Several studies have previously demonstrated human impacts on the global carbon cycle using changes in TC extents to deduce CS changes, especially in data-scarce regions like the tropics. However, such efforts inadequately address the TC-CS relationship and the explanatory role of land use. Here, by combining spatially-explicit estimates of TC and CS in their actual, ie. existing, and hypothetical potential states (in the absence of land use), we analyse their spatial correlations and quantify the impacts of land use across the global tropics (~ 23.4°N to 23.4°S). Land use strongly alters both TC and CS, with ecosystem-specific effects in humid and dry landscapes. While TC and CS are poorly-correlated in the potential states, land use introduces an ordering criterion in the actual states. These observations remain robust even when uncertainties in input datasets are qualified, and we find systematic biases in potential TC states across humid and dry regions in the tropics. We find that the land use-introduced ordering criterion is based on land cover changes, which selectively eliminates both TC and CS, as well as land management changes, ie. changes in land use within the same land cover type. By underlining existing heterogeneities in the TC-CS relationship, we call for particular caution in efforts to analyse changes in carbon fluxes in tropical ecosystems when solely based on TC assessments.