Compound-specific δ13C isotopes and Bayesian inference for erosion estimates under different land use in Vietnam

• CSSI analysis was applicable to discriminate erosion sources based on land use types.
• Stable isotope analysis in R (SIAR) allowed improved data processing of erosion source discrimination.
• Bayesian point-in-polygon test increased the reliability of the SIAR output.
• The proposed combined approach improved the application domain of CSSI.

Recent studies have pointed out the potential of the compound specific stable isotope (CSSI) technique based on long-chain fatty acids methylester (FAME) to identify hot spots of soil erosion by means of land use types. We tested the applicability of the CSSI technique on the basis of soil and sediment samples derived from a small agriculturally used catchment in Vietnam which is exemplary for many mountainous areas in Southeast Asia. Following CSSI analysis we set up a statistical decision sequence to identify hot spots of soil erosion by i) testing for significant differences between δ13C values of fatty acids (FA) of different contributing land use types and thereafter ii) examining the data using a Monte Carlo simulation of mixing polygons to provide a quantitative basis for model rejection and exclusion for sediment samples which violate the point-in-polygon assumption and iii) applying a Bayesian model with a Markov chain Monte Carlo (MCMC) model fitting using “SIAR‟ (Stable Isotope Analysis in R), which produces simulations of plausible values and therefore representing a true probability density for the proportional contribution of source soils. Our results confirmed that there were significantly different δ13C values for identical FAMEs extracted from soils under different land uses. Most fatty acids with significantly different δ13C values were found between soils under C3 (protected and secondary forest, teak and fruit plantations) and C4 (maize) plants but also within different soils of land use types which consisted only of C3 plants (e.g. protected forest, fruit plantation and teak). The resulting soil proportions were plausible for the six investigated sedimentation areas and suggested that fields under crop production such as maize and cassava, but also teak plantations were the main sources of eroding soil in the upland area surrounding the Chieng Khoi reservoir. Based on our data, we can conclude that the developed integrated Bayesian SIAR-CSSI approach represents a unique tool to identify and apportion soil sources to major land use types in small heterogeneous catchments by linking biomarkers of land use types to the sediment in deposition zones.