AbstractDifferences in erosion patterns inevitably lead to different soil loss processes. However, the differences in sediment sources between rainfall and snowmelt erosion are not well studied. To address this gap, a study was conducted in a small agricultural catchment (3.5 km2) using the fingerprinting technique to quantify the various sources contributing to sediment, as well as determine the magnitude of sediment total nitrogen (TN), total phosphorus (TP), total potassium (TK) and soil organic matter (SOM) loads from three sources (cultivated land, unpaved roads and gullies) during the rainy season and spring thawing period. Sediment sources from 17 rainfall erosion and seven snowmelt erosion events of varying magnitudes were compared. The Collins and Bayesian mixing models, coupled with geochemical tracers (Cs, V, Cd, Pb and Ag), were applied to assess the contribution proportion of individual sediment sources to suspended sediment. According to the results obtained from the Bayesian mixing model with higher accuracy, cultivated land, unpaved roads and gullies contributed 60.3% (load‐weighted mean contribution [LMC] = 82.8%), 11.0% (LMC = 6.5%) and 28.7% (LMC = 10.7%) of the sediment in the rainy season, respectively. Cultivated land was the dominant sediment source for rainfall erosion, with high rainfall erosivity and sediment concentration. However, for snowmelt erosion, gullies were the primary sediment source (LMC = 66.2%), followed by cultivated land (LMC = 19.8%) and unpaved roads (load‐weighted = 14.0%). Meanwhile, in the spring thawing period, the contributions of unpaved roads to sediment were higher than in the rainy season. During the rainy season, the greatest contributions of TN, TP, TK and SOM were observed from cultivated land, with respective LMC values of 94.2%, 89.4%, 81.1% and 94.8%. With the increase of sediment contribution from gullies in the spring thawing period, its LMC of TN, TP, TK and SOM increased to 42.6%, 58.3%, 65.0% and 38.0%, respectively. The results of the study could be useful for understanding the processes of soil erosion, clarifying the driving mechanism of compound erosion, developing the compound erosion prediction model and optimizing the spatial configuration of erosion control measures.
Read full abstract