Impact of land use and land cover change on stream flow in Nyangores sub-catchment Mara river, Kenya

Abstract

The Nyangores sub-catchment of the transboundary Mara Basin of Kenya and Tanzania is an important upstream area due to its ecological composition. Like many other sub-catchments in Kenya, Nyangores sub-catchment has undergone significant land cover changes arising from various socio-economic reasons. In the Nyangores sub-catchment, these changes are believed to be amplifying stream flows through increased runoff rates during rainfall seasons. Over the years, land use changes have been observed and have contributed to environmental changes which has led to reduction in vegetation cover in the upper catchements hence impacting on the river flow. Hydrological models have proved to be efficient evaluating the effects of land cover changes on stream flow characteristics. These models require high resolution spatio-temporal land cover information which defines the change statistics and directions. The aim of this study was; (i) To determine the physical sub-catchment characteristics. (ii)To assess the spatial and temporal land cover change (1995-2010). (iii) To assess the effects of land use land cover changes on the stream flow in Nyangores sub-catchment. The morphometric characteristics were assessed from a Digital Elevation Model (DEM) based on standard and automated procedures available within ArcGIS to determine physically based catchment characteristics. The land cover changes were obtained through classification of Landsat images for 1995 (Landsat TM) and 2010 (Landsat ETM+) based on a supervised classification methodology employing the Maximum Likelihood Function. The images were classified into four major classes namely, farmlands, tree plantation, natural forest and Shrub lands. The WEAP model was calibrated and validated using observed monthly streamflow data and later used to simulate the land cover change effects on stream flow between 1995 and 2010. From the morphometric analysis, the sub-catchment covers an area of about 933km2. Nyangores River was found to be a 4th order stream with a mean bifurcation ratio of 1.4. Land cover classification reveals that tree plantation declined by about 9.4%, farmland expanded by about 8.7%, Shrub lands increased by about 1.9% while forest declined by 1.2%. After calibration and validation, the simulation results indicated that stream flow increased by 3% in the whole sub catchment. The statistical analysis shows that the increase of streamflow and farmland expansion demonstrates a very strong and positive relationship. Land use change in the study area have contributed to environmental degradation, erosion and deforestation. The analysis of stream flow indicates predominantly low dry season flows and peak wet season flows between the 1995 and 2010. These changes could be attributed to spatial decrease in natural forest and tree plantation land areas that occupied upper sub-catchment over the study period. This situation has strong implications to water resources management in Nyangores sub-catchment where conflicts related to periodic water scarcity are increasing. The results will go a long way providing a useful information to support land-use planning and management. The study therefore calls for a multidisciplinary approach with a comprehensive view towards land use that maintains ecological health and human requirements. This study provides valuable information for sub-catchment management in the efforts to mitigate streamflow impact caused by LULC change in this critical watershed