Spatial modelling of soil and water conservation activities for a catchment in the Ethiopian Highlands
MetadataShow full item record
Ellison, J. 2016. Spatial modelling of soil and water conservation activities for a catchment in the Ethiopian Highlands. MSc thesis in Sustainable International Agriculture (Specialization: Tropical Agriculture). Göttingen, Germany: University of Göttingen.
Permanent link to cite or share this item: http://hdl.handle.net/10568/78138
This thesis focused on the use of soil and water conservation (SWC) planning tools to address two ecosystem services (ES’s), erosion control (EC) and dry season baseflow enhancement (BF). The study site was a headwater catchment encompassing Gudo Beret town located in the central sub-humid highlands of Ethiopia. The aims of the thesis were to 1) model the current soil loss risk for the catchment, 2) to simulate a spatial allocation of recommended SWC activities throughout the catchment while giving different weights to the ES objectives, and 3) to estimate potential changes in soil loss and in the water balance as a result of those simulated activity scenarios. In the study, a number of tools and procedures were used: field observations to map erosion hotspots, participatory dialogues with focus groups, remote sensing to generate a land use/land cover (LULC) map, a water balance calculation, and GIS-based spatial modelling tools. The current soil loss risk was predicted for the Gudo Beret catchment using the Revised Universal Soil Loss Equation (RUSLE). Erosion at Gudo Beret was found to far exceed safe limits, with soil loss from rainfed cropland estimated at 47 t ha-1 yr-1 (=82). Estimates of soil loss compared favorably with measurements and estimates from other sub-humid highland catchments in Ethiopia, but are believed to have significantly under-predicted total losses due to the prevalence of gullies in the study area. The Resource Investment Optimization System (RIOS) tool was used to locate the most “responsive” sites to SWC. Three scenarios were tested in which EC and BF were weighted according to the ratios 1:1, 2:1, and 1:0. RIOS performed a spatial allocation of the activities and produced a hypothetical post-SWC LULC map to represent changes in biophysical parameters where activities were allocated. Soil loss was then estimated for the entire catchment for the hypothetical scenarios and a simplified water balance was performed for the rainfed cropland LULC class to assess the potential impact on baseflow, using the soil water storage/drainage term of the water balance as a proxy. The analysis that followed found that soil loss and soil water storage/drainage were not significantly different between the scenarios. The lack of significance between outcomes of the scenarios was attributed to low data quality for some inputs and the relatively small catchment size – both of which suppressed spatial variability which is needed to produce contrasting relative rankings for the ES objectives in RIOS. Recommendations for improving RIOS were given. Despite the results, the model’s unique approach draws attention to both the need to target multiple ES objectives in future conservation goals as well as to account for the offsite benefits of SWC.