Arresting gully formation in the Ethiopian highlands

Over the past five decades, gullying has been widespread and has become more severe in the Ethiopian highlands. Only in very few cases, rehabilitation of gullies has been successful in Ethiopia due to the high costs. The objective of this paper is to introduce cost effective measures to arrest gully formation. The research was conducted in the Debre-Mewi watershed located at 30 km south of Bahir Dar, Ethiopia. Gullying started in the 1980s following the clearance of indigenous vegetation and intensive agricultural cultivation, leading to an increase of surface and subsurface runoff from the hillside to the valley bottoms. Gully erosion rates were 10–20 times the measured upland soil losses. Water levels, measured with piezometers, showed that in the actively eroding sections, the water table was in general above the gully bottom and below it in the stabilized sections. In order to develop effective gully stabilizing measures, we tested and then applied the BSTEM and CONCEPT models for their applicability for Ethiopian conditions where active gully formation has been occurring. We found that the model predicted the location of slips and slumps well with the observed groundwater depth and vegetation characteristics. The validated models indicated that any gully rehabilitation project should first stabilize the head cuts. This can be achieved by regrading these head cuts to slope of 40 degrees and armoring it with rock. Head cuts will otherwise move uphill in time and destroy any improvements. To stabilize side walls in areas with seeps, grass will be effective in shallow gullies, while deeper gullies require reshaping of the gullies walls, then planting the gully with grasses, eucalyptus or fruit trees that can be used for income generation. Only then there is an incentive for local farmers to maintain the structures