Spatio-temporal dynamics of forest cover and its impact on streamflow in River Muzizi Catchment, Western Uganda

Abstract

Despite the linkage between forests and river water flow, the relationship between the two does not make unanimity. This is partly because their interaction is dependent on multiple factors, including but not limited to scale (spatial and temporal), slope, soil, climate, and forest management practices. The objectives of this study were to (i) To determine the spatial patterns of forest cover in River Muzizi catchment from 2000-2019; (ii) To determine the spatial variability of streamflow in River Muzizi catchment from 2000-2019; and (iii) To evaluate the causal relationship between forest cover change and water yield in space and time. Land use land cover change was assessed for the years: 2000, 2005, 2010, 2015, and 2019 using Landsat imagery with a resolution of 30m*30m. The images were clustered into eight land use land cover types namely; forest, woodland, wetland, agriculture, bushland, grassland, bare land, and built-up areas using the Maximum Likelihood Classification of supervised classification approach. Streamflow variations were evaluated using SWAT-CUP 2012 and its Sequential Uncertainty Fitting Algorithm (SUFI-2) was used for model calibration, validation, sensitivity, and uncertainty analysis. Calibration was done for three years from 2003 to 2005 while validation was also done for three years from 2008 to 2010. The values of statistical indices NSE and R2 were 0.88 and 0.81 for calibration and 0.87 and 0.82, for validation respectively. The causal relationship between forest cover change and water yield variations was evaluated using Pearson’s correlation and regression analysis to determine the significance of the relationship. Findings showed a significantly decreasing trend in forest cover from 2000 to 2019 (p<0.05) with a deforestation rate estimated at an average of 6.8% annually. The decline in forest cover is explained by the high population growth rate caused by the influx of refugees which has increased demand for land for agriculture and settlement. Contrary to the forest cover, streamflow within the catchment showed an insignificant increasing trend for the same period (p>0.05). The relationship between forest cover change and streamflow was found to be insignificantly negative (70%, p>0.05) with only 47.7% of the variations in flow attributed to changes in forest cover.