Improved δ13C method to assess soil organic carbon dynamics on sites affected by soil erosion

The assessment of soil organic carbon (SOC) loss and input after land-use change is based largely on comparisons between disturbed sites and an undisturbed reference site. At sites affected by soil erosion this comparison neglects the interactions between soil erosion, decomposition and SOC input. Our aim was to develop a stable carbon isotope-based method (CIDE; carbon input, decomposition and erosion) to characterize SOC dynamics on sites affected by soil erosion. Using the CIDE approach, rates of (i) SOC loss by erosion, (ii) decomposition of forest-derived SOC and (iii) maize-derived net SOC input were determined. The approach was tested on a steep slope maize cultivation chronosequence of up to 20years following deforestation of primary forest. An uncertainty analysis showed that even small erosion rates (0.3kgsoil m(-2) year(-1)) underestimated SOC loss (sum of erosion and decomposition) and overestimated real SOC input using unadjusted measurement data. The misrepresentation was particularly large under conditions of severe erosion, small changes in SOC concentration with depth, large changes in C-13 with depth, and small changes in SOC stocks or C-13 with time. Twentyyears after land-use change CIDE-derived SOC loss was 89% and 52% larger than the unadjusted SOC loss estimates based on the RUSLE (revised universal soil loss equation) erosion model and the RothC carbon turnover model, respectively. The unadjusted maize-derived SOC input was 36% and 14% larger when derived with the RUSLE and RothC model, respectively, 20years after land-use change. This study demonstrates the care that must be taken when assessing SOC dynamics on erosion prone sites and provides a framework to address these challenges.