| heal.abstract |
The Pan-European Soil Erosion Risk Assessment (pesera) model was evaluated using existing soil erosion data collected under various types of climate, vegetation, landscape and soil conditions. The data used represent a variety of typical Mediterranean land uses such as winter wheat, vines, olives and bare, stony land prevailing in hilly areas. Using this data, the model was calibrated for sediment transport by overland flow and results compared to measured soil erosion values from runoff plots and a watershed on a monthly basis. The performance of the model was assessed statistically, showing that it can be satisfactorily used for predicting soil erosion rates under the conditions included in the study. The overall model estimate including all the available experimental data was 0.69 t ha(-1) yr(-1) with a maximum error of 1.49 t ha(-1) yr(-1). After validation, the model was applied to a small watershed (60 ha) of great ecological importance for the sea turtle Caretta caretta. For this purpose, soil and vegetation maps were compiled from all the necessary data for applying the model. The model was run for three years using daily data from an existing nearby meteorological station. The predicted and measured soil erosion rates for a 7-month period were 0.31 t and 0.18 t, respectively. Application of the model to each mapping unit showed the over-riding importance of land use for sediment generation under the given climatic conditions. Bare land, occupying 5.5% of the watershed area, generated up to 69% of the total sediments estimated for the watershed. It is concluded that the pesera model can be used as a regional diagnostic tool under a range of soil, topographic and climatic conditions for identifying the best land use type and vegetation cover to protect hilly areas from soil erosion. The calculated overall root mean square error for the model is 0.06 t ha(-1) yr(-1), compared to a soil erosion rate of 0.04 t ha(-1) yr(-1), which can be tolerated for protecting the area for the sea turtle. |
en |