Rill erosion is one of the main forms of soil erosion of the hillslopes on the Loess Plateau of China. Few studies have evaluated the effects of land use on the distribution and morphology of rills under extreme rainstorms. A total of 50 rill survey sites were selected in a typical watershed of the Loess Plateau of China to investigate the rill distribution and morphology of three land uses (fallow land, FL; slope farmland, SF; and annual grassland, AG) under extreme rainstorm conditions (called “7·26” rainstorm). Seven parameters (rill width, rill depth, width-depth ratio, cross-section area, rill density, degree of rill dissection and rill erosion modulus) were used to characterize rill morphology, and the basic soil and vegetation properties (bulk density, particle size distribution, and vegetation coverage) and topographic factors (slope length, gradient, direction) were measured. It was found that rills developed on three land uses, and those with widths ranging from 5 to 20 cm accounted for >80%. Rills developed on FL were deeper than those on SF and AG. Almost all rills on FL had a depth of 5–20 cm (90.3%), while the depth of rills on SF and AG were mainly smaller than 15 cm (91.5% and 91.9%, respectively). In addition to width, rill depth and cross-section area of SF and AG were also significantly less than those of FL. SF and FL had a relatively high width-depth ratio and dissection degree. However, land use had no significant effect on rill density. Crop plantations (SF) and grass restoration (AG) significantly reduced rill erosion modulus (28.70% and 45.07% reduction, respectively). Land use, topography, soil, and vegetation factors significantly affected rill morphology (p = 0.001) based on redundancy analysis (RDA) and can explain 48.0% of the total variability of rill morphology, with slope length, vegetation coverage and slope gradient having the greatest influences. Most of the rill morphological parameters significantly decreased as the vegetation coverage increased but they significantly increased with increasing slope length and slope gradient. This study clearly illustrates the severe rill erosion caused by this rainstorm, and indirectly indicates that crop harvesting (FL) aggravates rill erosion and that revegetation (AG) can effectively contain rill erosion. Therefore, rehabilitation efforts should be focused on revegetation (grassland and/or woodland) on steep slopes to control soil erosion.
Read full abstract