Quantifying water erosion hazard in central Tunisia A case study: the Leben watershed

Abstract

This study was performed in order to quantify soil erosion in the arid lands of central Tunisia. To do this we used geographic information systems, remote sensing and the RUSLE prediction model. While the study area is characterized by reduced rainfall quantities and therefore a low R factor, other factors such as high soil erodibility, low vegetation cover and steep slopes resulted in locally high erosion rates. Indeed, the calculation of RUSLE factors revealed values between 0 and 731 tones×ha-1×year-1 with an average of 3.84 tones×ha-1×year-1. Whereas most catchment lands have low levels, the relief surrounding the catchment has relatively high rates due to the high value of slopes, the skeletal soils that compose them, but also due to the absence of protection, reforestation or crop management. This study also showed that despite their small extension compared to the watershed, main streams sectors, and especially their banks, record the highest erosion rates mainly due to the flow accumulation in the downstream sectors. Relative protection of the piedmont, which has saved them from higher erosion, can only be considered efficient and sustainable if the upstream hillsides are also protected. The comparison with previous studies carried out in Tunisia shows similarities in the results of some of them, considering the natural geographical variation of the R factor.