RENDICONTI LINCEI-SCIENZE FISICHE E NATURALI, vol.28, no.1, pp.11-23, 2017 (SCI-Expanded)
Soil erosion is one of the major threats to the conservation of soil and water resources. For that reason, predictive erosion models are useful tools for evaluating soil erosion and developing soil erosion management plans. For this aim, the revised universal soil loss equation (RUSLE) function is a widely used erosion model. This research integrated the RUSLE with a geographic information system (GIS) to investigate the spatial distribution of annual soil loss potential in the Alaca catchment in north central Black Sea region, Turkey. The rainfall erosivity factor was developed from local annual precipitation data using a modified Fournier index; the topographic factor was developed from a digital elevation model; the land cover factor was generated from satellite imagery and forest inventory maps; and the soil organic carbon level and the erodibility factor were developed from systematically collected soil samples and the application of the geostatistical method, respectively. From the model, more than the half of the total study area as in the very low and low erosion risk classes (0-12 t ha(-1) year(-1)), whereas 4.4 % (723.6 h) of the total area was at high and very high erosion risk (35-150 and >150 t ha(-1) year(-1)), respectively. In addition, soil organic carbon density values were between 0.18 and 4.92 kg m(-2) across the catchment. Moreover, the distribution of soil organic carbon losses was closely correlated with the distribution of soil erosion risk classes in the study area. Soils and topographical properties of the watershed had a greater influence than land use/land-cover type on the magnitude of potential soil and soil organic carbon losses, because the erosivity factor did not change substantially in the study area.