Soil loss from small rangeland plots under simulated rainfall and run-on conditions

In order to predict and mitigate soil erosion it is necessary to understand the mechanism of detachment by raindrops and by shallow flow. The relative contribution of these drivers to the overall erosion process on arid rangelands is not well understood and field experimental data is limited. The experiment was conducted using simulated rainfall and overland flow on 56 small (2 m x 6 m) natural plots located on 7 rangeland sites in Arizona, USA. The goals of the study were (i) to compare raindrop driven and shallow flow driven erosion rates on arid rangeland, and (ii) assess the role of flow hydraulics, vegetation, and cover on attenuation of erosive impact of raindrops. A total of 520 measurements of steady state flow under two treatments (rainfall and run-on) were obtained. Flow discharge on the plots varied between 3 and 355 mm h(-1) and unit sediment yield varied between 0.02 and 30 g m(-2) min(-1). Sediment yield was best predicted by flow discharge for both rainfall and run-on treatments on all sites explaining 18% to 75% of its variability. There was statistically significant difference between two treatments. Rainfall treatment generated 2 to 44 times more sediment than run-on at the same discharge rate. We found no strong evidence of raindrop impact affecting overland flow velocity. Among 19 variables related to surface conditions a weak correlation was found between sediment yield and plant foliar cover, structure, and surface litter. However, there was no single best cover predictor common for all ecological sites tested.