Gaining insights into interrill erosion processes using rare earth element tracers

- Sediment transport process controlled interrill soil erosion rate.
- Erosion-transport-deposition occurred simultaneously in the form of bedload.
- Soil erodibility estimated with erosion models was sediment transportability.
- Erosion per unit area varied with slope length due to changes in transport modes.
- REE tracers provide a unique means to validate process-based erosion models.

Increasing interest in developing process-based erosion models requires better understanding of the relationships among soil detachment, transportation, and deposition. The objectives are to 1) identify the limiting process between soil detachment and sediment transport for interrill erosion, 2) understand the dynamic relationships between transport and deposition by tracking sediment fate with multiple tracers, and 3) verify the effects of slope length on interrill soil erosion. Five rare earth element (REE) tracers were applied in five bands or segments to track sediment dynamics on a 10%, 4 × 4-m long uniform slope under simulated rainfall. A silt loam soil with 8% clay and 87% silt was used. Six rain events were applied, the first four with 60 mm h− 1 and the last two with 90 mm h− 1. Sediment in runoff and deposition along the slope were measured. Results confirmed that interrill soil erosion was controlled by the transport process. Thus soil erodibility estimated with many interrill erosion models was actually sediment transportability. As the slope length increased, soil erosion rates initially increased and then decreased, indicating that the upper section was dominated by erosion and the lower section by transport. Raindrop-driven creeping prevailed in the upper section while flow-driven rolling dominated in the lower section. Sediment influx from upslope clearly suppressed soil detachment downslope. Moreover, the steady state sediment discharge from a segment was positively correlated to the amount of sediment deposited downslope from that segment, implying that 1) re-detachment, transport, and deposition occurred simultaneously in the system and 2) sediment was transported in a form of bedload, mainly rolling or creeping on interrill areas. Sediment delivery ratios of each segment tended to slightly increase downslope, suggesting the flow-driven transport was more efficient than the raindrop-driven transport. The sequence of sediment discharged from different slope positions was inconsistent with the transport-distance theory. However, interrill erosion processes vary with soil, slope, and rainfall properties, and the findings here need to be verified in different conditions.