The impact of land use change on soil organic carbon pools: A multi-method assessment

Soil organic carbon (SOC) is significantly affected by land use change (LUC), which can lead to either SOC losses or gains. In consequence, LUC is a major controlling factor of total SOC contents and its dynamics. In general, LUC from forest or grassland to permanent cropland results in losses of SOC, while the reverse can result in long-term gains. Several methods have been developed to assess distinct SOC pools. This includes particle size separation, thermal analysis and soil reflectance spectroscopy. The responses of such defined SOC pools to LUC have rarely been studied comprehensively, while doing so is a straightforward way to reveal their biogeochemical relevance. Here we used 23 sites covering six different LUC (i.e. forest-cropland, grassland-cropland, grassland-forest, cropland-grassland, cropland-forest and cropland-perennial Miscanthus) to assess SOC pool changes. We used particle size fractionation to obtain coarse (>50µm) and fine (<50µm) fractions and Rock-Eval 6 analysis to estimate active and stable SOC pools. Additionally, we used mid-infrared spectroscopy (Diffusive Reflectance Infrared Fourier Transformed Spectroscopy (DRIFT)) on bulk soils and particle size fractions to determine the relative SOC composition.

All methods identified kinetically different SOC pools across all LUC. The fine particle size fraction, representing a stabilized and slow cycling SOC pool, was more responsive to LUC compared to the stable pool estimated using Rock-Eval. Assessing the relative changes of total SOC and organic carbon contents of the fractions across all LUC, showed that the fine particle fraction follows closely the total SOC changes (R²=0.91 with slope of 0.77). In comparison, the stable pool extracted by Rock-Eval was less dynamic (R²=0.72 with a slope of 0.43). Absolute changes in bulk SOC were well described by the absolute organic carbon change of extracted Rock-Eval pools (active: R²=0.99 and stable: R²=0.91), while organic carbon changes of the particle size fractions were less sufficient to describe bulk SOC changes (coarse: R²=0.77 and fine: R²=0.33). The SOC composition, determined by DRIFT, revealed that changes in the relative composition of fast cycling aliphatic to slow cycling aromatic compounds can well explain relative total SOC changes (R²=0.81). This shows that three conceptually different methods (physical, thermal and spectroscopic) are suitable to determine SOC pool changes for a large diversity of LUC, but the sensitivity of the individual pools can differ strongly, depending on the method.