Fingerprinting sediment sources in mountainous catchments disturbed by plantation and construction activities in southern China

Abstract The growing problems of soil erosion and sedimentation due to plantation and construction activities highlight the importance of reliable information on sediment sources with the objectives of designing and implementing targeted management practices. However, reliable information is unavailable for mountainous catchments in southern China. Accordingly, the fingerprinting approach was adopted to quantitatively identify sediment sources in a typical mountainous catchment disturbed by plantation and construction activities in the Dadingshan catchment, Guangdong Province, China. Sensitivity tests were conducted to evaluate the effect of correction factors on the modified model. The results indicated that geochemical elements could be applied for identifying sediment sources within the study catchment. However, the coarse sediment in the catchment posed challenges to the conservative behaviour of tracers. The findings suggested that the primary sediment source is exposed subsoil (51.5 ± 10.0%), followed by plantation topsoil (35.3 ± 16.2%) and wind farms (13.2 ± 10.2%). The mean goodness‐of‐fit value for all sediment samples was 87.1 ± 5.7%. The sediment contributions of various sources varied significantly due to differences in the extent and intensity of human intervention and the main erosion type of each source. Sensitivity analysis revealed that considering the assumption of source normality improved the source apportionment accuracy. However, particle size correction and tracer within‐source variability weighting in the modified mixing model resulted in unusual estimates of the source contributions. This contribution highlights the significance of considering uncertainties in key fingerprinting‐related aspects and emphasizes the need for implementing sediment control strategies in degraded forest environments.