Soil detachment and transport on field- and laboratory-scale interrill areas: erosion processes and the size-selectivity of eroded sediment

Field- and laboratory-scale rainfall simulation experiments were carried out in an investigation of the temporal variability of erosion processes on interrill areas, and the effects of such variation upon sediment size characteristics. Poorly aggregated sandy soils from the semiarid environment of Senegal, West Africa, were used on both a 40 m(2) field plot and a 0-25 m(2) laboratory plot; rainfall intensity for all experiments was 70 mm h(-1) with a duration of 1 to 2 hours. Time-series measurements were made of the quantity and the size distribution of eroded material: these permitted an estimate of the changing temporal balance between the main erosion processes (splash and wash). Results from both spatial scales showed a similar temporal pattern of runoff generation and sediment concentration. For both spatial scales, the dominant erosional process was detachment by raindrops; this resulted in a dynamic evolution of the soil surface under raindrop impact, with the rapid formation of a sieving crust followed by an erosion crust. However, a clear difference was observed between the two scales regarding the size of particles detached by both splash and wash. While all measured values were lower than the mean weight diameter (MWD) value of the original soil (mean 0(.)32 mm), demonstrating the size-selective nature of wash and splash processes, the MWD values of washed and splashed particles at the field scale ranged from 0(.)08 to 0(.)16 mm and from 0(.)12 to 0(.)30 mm respectively, whereas the MWD values of washed and splashed particles at the laboratory scale ranged from 0(.)13 to 0(.)29 mm and from 0(.)21 to 0(.)32 mm respectively. Thus only at the field scale were the soil particles detached by splash notably coarser than those transported by wash. This suggests a transport-limited erosion process at the field scale. Differences were also observed between the dynamics of the soil loss by wash at the two scales, since results showed wider scatter in the field compared to the laboratory experiments. This scatter is probably related to the change in soil surface characteristics due to the size-selectivity of the erosion processes at this spatial scale. Copyright (c) 2006 John Wiley & Sons, Ltd.