In the drought-prone Upper Blue Nile River (UBNR) basin of Ethiopia, soil erosion by water results in significant consequences that also affect downstream countries. However, there have been limited comprehensive studies of this and other basins with diverse agroecologies. We analyzed the variability of gross soil loss and sediment yield rates under present and expected future conditions using a newly devised methodological framework. The results showed that the basin generates an average soil loss rate of 27.5thayr and a gross soil loss of ca. 473Mtyr, of which, at least 10% comes from gully erosion and 26.7% leaves Ethiopia. In a factor analysis, variation in agroecology (average factor score=1.32) and slope (1.28) were the two factors most responsible for this high spatial variability. About 39% of the basin area is experiencing severe to very severe (>30thayr) soil erosion risk, which is strongly linked to population density. Severe or very severe soil erosion affects the largest proportion of land in three subbasins of the UBNR basin: Blue Nile 4 (53.9%), Blue Nile 3 (45.1%), and Jema Shet (42.5%). If appropriate soil and water conservation practices targeted ca. 77.3% of the area with moderate to severe erosion (>15thayr), the total soil loss from the basin could be reduced by ca. 52%. Our methodological framework identified the potential risk for soil erosion in large-scale zones, and with a more sophisticated model and input data of higher spatial and temporal resolution, results could be specified locally within these risk zones. Accurate assessment of soil erosion in the UBNR basin would support sustainable use of the basin's land resources and possibly open up prospects for cooperation in the Eastern Nile region.

No abstract

This paper reviews Ethiopia’s experience and research progress in past soil and water conservation (SWC) efforts and suggests possible solutions for improvement. Although indigenous SWC techniques date back to 400 BC, institutionalized SWC activity in Ethiopia became significant only after the 1970s. At least six national SWC-related programs have been initiated since the 1970s and their focus over time has shifted from food relief to land conservation and then to livelihoods. The overall current soil erosion rates are highly variable and large by international standards, and sheet, rill, and gully erosion are the dominant processes. The influence of human activities on the landscape has traditionally been deleterious, but this trend seems to have recently reversed in some parts of the country following the engagement of the communities in land management. The efficiency of SWC measures show mixed results that are influenced by the type of measures and the agro-ecology under which they were implemented; in general, the relative performance of the interventions is better in the drylands as compared with humid areas. Methodological limitations also occur when addressing the economic aspects related to benefits of ecosystem services and other externalities. Although farmers have shown an increased understanding of the soil erosion problem, SWC efforts face a host of barriers related to limited access to capital, limited benefits, land tenure insecurity, limited technology choices and technical support, and poor community participation. In general SWC research in Ethiopia is fragmented and not comprehensive, mainly because of a lack of participatory research, field observations, and adoptable methods to evaluate impacts. A potentially feasible approach to expand and sustain SWC programs is to attract benefits from global carbon markets. Moreover, a dedicated institution responsible for overseeing the research–extension linkage of SWC interventions of the country should be established.

Catchment management in the developing world rarely include detailed hydrological components. Here, changes in the hydrological response of a 200-ha catchment management in north Ethiopia are investigated. The management included various soil and water conservation measures such as the construction of dry masonry stone bunds and check dams, the abandonment of post-harvest grazing and the establishment of woody vegetation. Measurements at the catchment outlet indicated a runoff depth of 5 mm or a runoff coefficient (RC) of 1.6 % in the rainy season of 2006. Combined with runoff measurements at plot scale, this allowed calculating the runoff Curve Number (CN) for various land uses and land management techniques. The pre-implementation runoff depth was then predicted using the CN values and a ponding adjustment factor, representing the abstraction of runoff induced by the 242 check dams in gullies. Using the 2006 rainfall depths, the runoff depth for the 2000 land management situation was predicted to be 26.5 mm (RC = 8 %), in line with current RCs of nearby catchments. Monitoring of the ground water level indicated a rise after catchment management. The yearly rise in water table after the onset of the rains (∆T) relative to the water surplus (WS) over the same period increased between 2002-2003 (∆T/WS = 3.4) and 2006 (∆T/WS >11.1). Emerging wells and irrigation are other indicators for improved water supply in the managed catchment. Cropped fields in the gullies indicate that farmers are less frightened for the destructive effects of flash floods. Due to an increased soil water content, the crop growing period is prolonged. It can be concluded that this catchment management has resulted in a higher infiltration rate and a reduction of direct runoff volume by 81 % which has had a positive influence on the catchment water balance.

Since two decades, stone bunds have been installed in large areas of the Tigray Highlands, Northern Ethiopia, to control soil erosion by water. Field studies were conducted to quantify the effectiveness, efficiency, side effects and acceptance of stone bunds. Based on measurements on 202 field parcels, average sediment accumulation rate behind 3-21 year old stone bunds is 58 t ha À1 year À1 .The Universal Soil Loss Equation's P-factor for stone bunds was estimated at 0.32. Sediment accumulation rates increase with slope gradient and bund spacing, but decrease with bund age. Truncation of the soil profile at the lower side of the bund does not lead to an important soil fertility decrease, mainly because the dominant soil types in the study area (Regosols, Vertisols and Vertic Cambisols) do not have pronounced vertical fertility gradients. Excessive removal of small rock fragments from the soil surface during stone bund building may lead to a three-fold increase in sheet and rill erosion rates. Negative effects of runoff concentration or crop burial by sediment deposition due to bunds were only found over 60 m along 4 km of studied bunds. As the rodent problem is widespread and generally not specific to stone bunds, it calls for distinct interventions. On plots with stone bunds of different ages (between 3 and 21 years old), there is an average increase in grain yield of 53% in the lower part of the plot, as compared to the central and upper parts. Taking into account the space occupied by the bunds, stone bunds led in 2002 to a mean crop yield increase from 0.58 to 0.65 t ha À1 . The cost of stone bund building averages s13.6 ha À1 year À1