IMPACTS OF CLIMATE AND LAND USE/ COVER CHANGES ON SELECTED ECOSYSTEM SERVICES IN THE ABBAY RIVER BASIN, ETHIOPIA

Abstract

Land, together with its water bodies, delivers manifold ecosystem services which are the foundation for human lives and well-being. However, unmanageable land use/ cover and climate change are putting a strain on the flow of essential ecosystem services. Therefore, this dissertation was organized to investigate the spatiotemporal impact of climate and land use/ cover change on selected ecosystem services such as soil erosion regulation and water yield of the Abbay River Basin Ethiopia. This study used a stepwise approach to achieve the intended objectives. First, the integration of the Multi-Layer Perceptron Neural Network and Markov Chain Model in the Land Change Modeler was used to model and analyze land use/ cover change. Secondly, the spatiotemporal trends and rates of change of NDVI and climate variables were analyzed using Mann-Kendall Monotonic Trend and Theil-Sen slope statistical methods respectively. The correlations, and relationships between NDVI and climate variables were executed using partial correlation coefficient, and multiple linear regression methods, respectively. Thirdly, the Revised Universal Soil Loss Equation Model was used to execute the annual rate of actual soil loss and potential soil loss. Then soil erosion prevention service was computed as the difference between structural impact and mitigated impact. Lastly, the spatiotemporal impacts of climate and land use/ cover change on water yield were analyzed using the InVEST model. The relative contributions of climate and land use/ cover change on water yield variation were computed by designing three model scenarios. The initial scenario represented actual conditions including climate and land use/ cover change. The second and third scenarios were the actual scenario with constant climate change, and the actual scenario with constant land use/ cover change respectively. The findings of the study revealed that ARB has experienced significant land use/ cover change. Between 1994 and 2021, agriculture, bareland, shrubland, and wetland showed a decreasing trend of 34.5 km2, 1 km2, 200 km2, and 0.1 km2 annual rate of change while forest, grassland, settlement, and waterbody showed an increasing trend with 166 km2, 0.5 km2, 20 km2 and 50 km2 annual rate of change respectively. Between 2021 and 2056, agriculture and shrubland will continue to decrease by 161 km2 and 113.6 km2 annual rate of change while forest and settlement will increase by 116 km2 and 31 km2 annual rate of change respectively. The spatiotemporal trend and rate of change analysis showed that interannual NDVI increased significantly in 78% of the basin's total area. Of the 78% of the basin, 31%, and 47%, of the total area showed an extremely significant increase (Zmk = 4.706, p ≤ 0.01), and significant increase trends (Zmk = 2.378, p ≤ 0.05) respectively. The interannual variation of NDVI was well explained (R2 = 0.88, Adjusted R2 = 0.84) by the climate variables in the eastern, southeastern, and central subbasins where agriculture, grass, sparse vegetation, and barelands are the predominant. The increasing trends of NDVI resulted in increasing of the soil erosion regulation services provisioning capacity of Beressa watershed from 0.55 in 1994 to 0.72 in 2018 by an annual rate of 0.085 (R2 = 0.75). This in turn resulted in decreasing mitigated impact from 24.2 t ha-1yr-1 in xxi 1994 to 14.5 t ha-1yr-1 in 2018 by an annual rate of 4.9 t ha-1yr-1 (R2 = 0.87). Finally, this resulted in an increase in the mean actual ecosystem service provisioning from 33.2 t ha-1yr-1 in 1994 to 34.6 t ha-1yr-1 in 2018 by an annual rate of 0.75 t ha-1yr-1 (R2 = 0.999). However, the increasing forest cover and potential evapotranspiration were found inversely correlated with water yield services in the Beressa watershed . Temporally, the water yield decreased significantly from 231.7 MCM in 1994 to 77.5 MCM in 2018 with a 48.0 MCM annual rate. Spatially, the water yield showed decreasing trend with the annual rate of change ranging from 337.7mm to 256.9 mm yr-1 in the Beressa watershed where the forest (epecially eucalyptus tree plantation) is predominant and mean annual potential evapotranspiration is high. In the parts of the watershed where sparse vegetation and bare land are predominant and where mean annual rainfall is high, water yield showed an increasing trend with the annual rate of change ranging from 28.8 mm to 164.8 mm. The analysis of the three scenarios showed that climate change contributed a 57% impact on water yield variation whereas land use/ cover change contributed a 43% impact. The increasing forest (epecially eucalyptus tree plantation) cover and mean annua potential evapotranspiration showed detrimental impact on the water yield services of the watershed. Therefore, integrated watershed management that takes thoroughly this scenario into consideration should be implemented to match the shifting hydrological regimes at various hydrometric stations in light of ongoing land use/ cover and climate change. Moreover, environmental policy should be developed and put into place to address the potential fluctuations in water yield services of Beressa watershed.