Simulation of soil erosion and deposition in a changing land use:: A modelling approach to implement the support practice factor

Using the USPED (Unit Stream Power Erosion Deposition) model, three land use scenarios were analysed for an Italian small catchment (15 km(2)) of high landscape value. The upper Orme stream catchment, located in the Chianti area, 30 km south of Florence, has a long historical agriculture record. Information on land use and soil conservation practices date back to 1821, hence offering an opportunity to model impacts of land use change on erosion and deposition. For this study, a procedure that takes into account soil conservation practices and potential sediment storage is proposed. The approach was to calculate and model the flow accumulation considering rural and logging roads, location of urban areas, drainage ditches, streams, gullies and permanent sediment sinks. This calculation attempts to assess the spatial variability, especially the impact of support practices (P factor). Weather data from 1980-2003 were taken into account to calculate the R factor. However, to consider the intense pluviometric conditions in terms of the erosivity factor, the 0.75th quantile was used, while the lowest erosivity was modelled using the 0.25th quantile. Results of the USPED model simulation show that in 1821 the mean annual net erosion for the watershed was 2.8 Mg ha(-1) y(-1); in 1954 it was 4.2 Mg ha(-1) y(-1); and in 2004 it was 5.3 Mg ha(-1) y(-1). Conservation practices can reduce erosion processes by >= 20 Mg ha(-1) y(-1) when the 1821 practices are introduced in the present management. On the other hand, if the support practices are not considered in the model, soil erosion risk is overestimated. Field observation for the present-day simulation confirmed that erosion and associated sediment deposition predicted by the model depend, as expected, on geomorphology and land use. The model shows limitations that are mainly due to the input data. A high resolution DEM is essential for the delineation of reliable topographic potential to predict erosion and deposition especially in vineyards. (C) 2007 Elsevier B.V. All fights reserved.