Soil and water conservation measures reduce soil and water losses in China but not down to background levels: Evidence from erosion plot data

The efficacy of soil and water conservation measures (SWCMs) on reducing the loss of soil and water has been widely tested at the plot scale in China. However, an integrated comparison of the efficacy and an overall investigation of the controlling factors of the efficacy is absent. Therefore, an extensive dataset compiling erosion plot measurement of applying SWCMs on controlling soil loss and runoff in China was established in this study. In total, 848 erosion plots representing 2494 plot years were retained in this dataset which compiled from 108 publications and 2 data reports. The 15 most frequently tested SWCMs were identified and described. Soil loss ratio (SLR) and runoff ratio (RR) of each SWCMs was calculated to assess the efficacy of an SWCMs on reducing soil and water loss. The potential controlling factors, such as slope, slope length, annual precipitation, were evaluated for the controlling factor for both SLR and RR. Our results suggested that in general, applying the SWCMs can reduce soil loss by ca. 70% (overall SLR is 0.39 ± 0.02) and runoff by ca. 50% (overall RR is 0.54 ± 0.05), respectively. However, erosion rates on plots with SWCMs in China were, in general, still significantly higher than erosion rates measured under permanent vegetation cover land and, on average, ca. one order of magnitude higher than soil formation rates which required for a soil-neutral agriculture. On overall, the SLR presented a significant relation with topographic factors: positive with slope gradient and negative with slope length. This suggests that the support practice factor (P factor) of RUSLE model might be overestimated when values measured from plot scale are directly applied to larger areas (catchment or region scale). The SLR and RR values presented in this study can be applied for the planning of SWCMs in China and can be used to estimate the P factor when applying RUSLE model whereby the effect of topographic factors on SWCM efficacy can be accounted for.