The effectiveness of soil conservation measures at a landscape scale in the West Usambara highlands, Tanzania

The adoption of soil and water conservation (SWC) technologies among small holder farmers in the East African highlands is an area which poses many challenges. When adoption occurs across a vast landscape, the locations and effectiveness of the adopted measures are often not adequately known. For this reason, the majority of SWC studies in the highlands of East Africa employed field surveys and experiments to locate and estimate effectiveness of the installed technologies. This approach however has certain limitations when applied at a landscape scale. Potentially, remote sensing techniques could be used for the purpose of locating soil conservation structures, while modeling can help in estimating the effectiveness of the implemented measures. This study therefore employed remote sensing and GIS techniques to 1) to locate SWC structures in two 100km2 areas in the West Usambara highlands of Tanzania, and 2) to determine the effectiveness of the implemented measures in reducing soil erosion at a landscape scale. The study was conducted in the west Usambara highlands of north-eastern Tanzania as a paired plot design in which two blocks of 100km2 each were studied using pixel (Maximum Likelihood Classification) and object-based image analysis (OBIA) remote sensing techniques to detect land use patterns and adoption of soil conservation technologies. Soil losses were modeled using the Universal Soil Loss Equation (USLE)-model while effectiveness of the measures was estimated from calculations. Results indicate that there are large differences in the adoption of soil conservation technologies between the two blocks. The study also finds the Maximum Likelihood Classifier to be reliable in generating land use thematic layer maps from which soil conservation measures can be studied with ease in mountainous areas. The OBIA-technique was found to be effective in identifying, classifying and mapping of the adopted SWC technologies. Effectiveness of the installed technologies remained comparable across the blocks but with higher indices for Sunga area. The study concludes that adoption of the SWC technologies in the two blocks is largely influenced by biophysical conditions within and between the two blocks and is not related to the quality of the technologies being implemented in either block.