Effects of land use and land cover on sheet and rill erosion rates in the Tigray highlands, Ethiopia

Sheet and rill erosion rates have been well studied in rain-rich central and south Ethiopia. We conducted runoff plot experiments in the semi-arid north Ethiopian Tigray highlands to investigate whether erosion rates are within the same order of magnitude, to measure the effects of various land uses and covers, and to examine the application of the (Revised) Universal Soil Loss Equation (RUSLE) to the Ethiopian highlands. Average soil loss rate by sheet and rill erosion was 9.7 (+/- 7.8) Mg ha(-1) y(-1) (42 plot-years), with 3.5 Mg ha(-1) y(-1) in exclosures and forest, and 17.4 Mg ha(-1) y(-1) in rangeland. Especially on arable land, the measured rate (9.9 Mg ha(-1) y(-1)) was well below the average erosion rate elsewhere in Ethiopia (42 Mg ha(-1) y(-1)), which is attributed to (1) precipitation depth and total rain intensity, which are lower in north Ethiopia, (2) the widespread use of stone bunds, which decrease runoff length and velocity, and (3) often high rock fragment cover at the soil surface in arable land. The cover-management C-factor of RUSLE for tef (Eragrostis tef) was lower than in central Ethiopia (C = 0.07 against 0.25), probably due to earlier sowing than in the rain-rich regions of Ethiopia. Further values of C in the study area were 0.21 for arable land under wheat and barley; 0.42 for degraded rangeland, and 0.004 for forest and exclosures. Exclosures and forests also trapped upslope eroded sediment.