MODELING LAND DEGRADATION IN UPPER TURKWEL WATERSHED AND IMPLICATIONS ON THE SUSTAINABILITY OF THE TURKWEL DAM RESERVOIR

Abstract

Sediment accumulation in water reservoirs with consequent shortening of the reservoir life is a growing global problem,the worst scenarios are common in tropical countries like the study area Turkwel watershed in Kenya. The study aimed to determine the land cover changes in the watershed. These would further answer how it affected land degradation and sediment accumulation in the reservoir and proposed mitigation measures. The study also considered the changing land use and land degradation patterns that have taken place in the catchment by studying periodic Landsat/satellite imagery. The studies established that bare land in the watershed between the last four decades of 1987 to 2017 increased insignificantly by (1.05%)/62 km2 (F 0.05 (1, 2) =12.49, p=0.0716), reinforcing the reality of global warming and climate change as a precursor to more soil erosion. The forest and agricultural land declined by 73 km2 (-1.24%) and 116 km2 (-1.97%), respectively, attributed to climate change and land pressure from the increasing population. The grassland increased insignificantly between 1987 to 2017 by 120.26 km2 (F 0.05 (1, 2) =1.95, p=0.2975). More farmers embraced livestock production through enclosures as a source of livelihood while abandoning crop farming. SWAT model studies in the watershed revealed that the area under medium to very high erosion rates rose to 1940 km2 (33%) in the 5900 km2 compared to 1290 km2 (22%) in 1982. Soil erosion was increasing and called for more conservation efforts on erosion hot spots and a function of soil vulnerability. The soil loss and Best Management Practices (BMPs) determined using standard runoff plot experiments on a 10.5% land slope farm showed clear benefits of integrated approaches. The Land Use/Land Cover changes (LUCC) studied using Landsat imagery, remote sensing, and Arc View GIS established that Runoff was reduced by retention ditches more than pasture reseeding. The highest significant difference in mean soil losses (using STATGRAPHICS) was between runoff plots with terraces and reseeded with Pokot Rhodes grass in comparison with no conservation (F0.05 (3, 220) =13.23, p=0.0001) with a maximum mean significance difference between the plot and the control plot/or no treatment plot with the variation of (78.81±29.22 kg/ha/storm). The field with Pokot Rhodes only had the second-lowest soil loss and indicated, that growing pasture was a cheap single conservation option. The best management Practice involving enclosures with an integrated ecosystem approach using agronomic, physical, and cultural measures was most effective with the extra benefits of ensuring sustainable production and environmental conservation. The study on the perception and adoption of introduced Soil and Water Conservation (SWC) strategies suggested that 98% of the farmers knew the accruing benefits of soil conservation but low adoption and sustainability. There was a need to intensify farmer's agricultural extension services. Some 32% of the farmers acknowledged stopping conservation activities at the end of the project support and this called for further research to establish the reasons. Using the four indices in modelling river flow at Turkwel, the flow data was consistent with the criteria for a monthly time step (R 2 =0.58, NSE=0.56, RSR=0.66 & PBIAS=10.9%). The SWAT model was therefore applicable to the catchment and the impact analyses.