Monsieurs, E., Dessie, M., Verhoest, N. E., Poesen, J., Adgo, E., Deckers, J. & Nyssen, J. 2016. ‘Impact of draining hilly lands on runoff and on-site erosion: a case study from humid Ethiopia’. Earth Surface Processes and Landforms 41: 513-525. John Wiley & Sons, Ltd. URL: http://onlinelibrary.wiley.com/doi/10.1002/esp.3845/abstract I.F. 3.505.
Article dans une revue scientifique / Article dans un périodique
The use of drainage ditches on farmland has an impact on erosion processes both on-site and off-site, though their environmental impacts are not unequivocal. Here we study the runoff response and related rill erosion after installing drainage ditches and assess the effects of stone bunds in north Ethiopia. Three different land management systems were studied in 10 cropland catchments around Wanzaye during the rainy season of 2013: (1) the exclusive use of drainage ditches (locally called feses), (2) the exclusive use of stone bunds, and (3) a mixture of both systems. Stone bunds are an effective soil and water conservation technique, making the land more resistant against on-site erosion, and allowing feses to be installed at a larger angle with the contour. The mean rill volumes for the 10 studied cropland catchments during the rainy season of 2013 was 3.73 ± 4.20 m3 ha−1 corresponding to a soil loss of 5.72 ± 6.30 ton ha−1. The establishment of feses causes larger rill volumes (R = 0.59, N = 10), although feses are perceived as the best way to avoid soil erosion when no stone bunds are present. The use of feses increases event-based runoff coefficients (RCs) on cropland from c. 5% to values up to 39%. Also, a combination of low stone bund density and high feses density results in a higher RC, whereas catchments with a high stone bund density and low feses density have a lower RC. Peak runoff discharges decrease when stone bund density increases, whereas feses density is positively related to the peak runoff discharge. A multiple linear relation in which both feses and stone bund densities are used as explanatory variable, performs best in explaining runoff hydrograph peakedness (R2 = 83%)