The importance of Water Quality and Quantity in the Tropical Ecosystems, Tanzania
Abstract (Summary)This thesis looks at the importance of water quality and quantity in the tropical ecosystems in relation to movement of animals who depend on that water for their daily lives and existence. Seasonal fluctuations of rainfall were pronounced, with marked wet and dry seasons. Surveys of water quality in the surface waters in the Serengeti and Tarangire National Parks in Tanzania, East Africa were undertaken throughout the year, in both the wet and dry seasons, from 1996 to 2003. Surveys of water quality in the surface waters of Tarangire National Park were carried out from 1999 to 2001. In the Serengeti most of the rivers were ponded, with ponds having a flushing time of 1 month in the wet season and zero flushing in the dry season. The parameters used in the water quality included temperature, pH values, dissolved oxygen (DO), salinity (S) and visibility. pH values varied spatially from extremely alkaline conditions (pH>10) in the southern plains of the Serengeti to acidic conditions in the northern region (pH=5.9). In the southern plains at the end of the dry season the salinity of the surface waters was high (5-100/00) while there was still abundant fodder, the zebras and wildebeest had started to migrate away, suggesting that excessive salinity may be the trigger initiating the annual migration. Most surface waters were heavily eutrophicated as a result of animal dung. Because of this animal dung, the dissolved oxygen concentration near the surface fluctuated widely between 1 and 200% of saturation, the smaller values occurring deeper in the water column. Stirring and mechanical aeration by various animals in the water prevented the formation of anoxic conditions. The oxygen stress was less in wetlandfringed water bodies, because of the filtering effect of wetlands. Light penetration was high (>10 cm) in saline waters because of settling of suspended matter was accelerated by flocculation caused by bacteria and vegetation detritus. Elsewhere, the euphotic zone was less than 1 cm thick and the waters generally inhospitable to aquatic life.Tarangire National Park, on the other hand, showed similar situation as in the Serengeti. In the dry season, the only drinking water available for wildlife was the Tarangire River and a number of small, scattered wetland-fringed water holes. The salinity was often high (>8 ppt) and was higher in the dry years than in wet years, as well as at the start of the wet season. Water quantity and quality also appeared to control the annual migration of wildebeest, zebras, elephants and buffaloes. These animals aggregated in the dry season in areas with the least salty water. The timing of seasonal variations in rainfall was largely predictable and controlled annual migration. All wildebeest and most zebras migrated out of Tarangire National Park and into the wider Tarangire ecosystem at the start of the wet season, and they returned into the park in the dry season. Some elephants and buffaloes also migrated in and out of the park and a larger resident population remained, whose size may vary interannually depending on surface water quantity and quality. The extent of the migration zone may also vary interannually.The study also focused at the vertical distribution of temperature and dissolved oxygen over 24 hours in a pond of the Seronera River inhabited by hippos in the Serengeti. Findings showed that the waters were very turbid (visibility <2 cm). The high turbidity was from animals trampling sediment and a permanent surface algal bloom sustained by faecal matter. Direct solar heating was restricted to the top few centimetres. This resulted in a strong thermal and density stratification inhibiting aeration of the water column. Waters at the mid-depth were aerated only when hippos stirred the water. Anoxic conditions were common in bottom water; these were occasionally ventilated in day time by mixing due to bottom heating from decaying organic matter and at night by convective cooling. Poor water quality in hippo pools may affect wildlife.In the Serengeti ecosystem, the Mara River is the only source of permanent water supply throughout the year. Other rivers such as the Grumeti and Mbalageti dry out in the dry season with only pools of water scattered in a series. The Mara River catchment is the dry weather refuge for more than 1 million wildebeest and zebras of the Serengeti ecosystem. The Mara River flow is affected by developments in Kenya, including deforestation, water diversion for irrigation, and the proposed Ewaso Ng’iro (South) Hydropower project. An ecohydrology model was developed to predict the inter-annual fluctuations of the wildebeest and lion populations as a function of the hydrology. The model was calibrated against observations of rainfall and wildebeest and lions numbers in the period 1960-2000. This model used to predict the likely impacts of these developments on the Serengeti ecosystem. The model was forced by observed monthly rainfall in the period 1900-2000 and calibrated against observations of the number wildebeest and lions also in the period 1960-1999. The projects are predicted to have little effect on the number of migrating wildebeest in the Serengeti until a drought occurs; historically a drought occurs about every 7 years, and a severe drought occurs every 15 years. At that time 20-80% of the migrating wildebeest may die, according to the severity and duration of the drought. With a 50% die-off, it may take twenty years for the population to recover; with an 80% die-off there may be no population recovery. In practice the economic benefits would go to Kenya while Tanzania would suffer most of the economic costs, i.e. the negative impact on the tourist industry and socio-economic benefits to communities living along the Mara River. To ensure sustainable developments for both Kenya and Tanzania, a transboundary Mara River Management Plan needs to be implemented and be compatible with ecohydrology principles for the sustainable use of aquatic resources.
Source Type:Master's Thesis
Date of Publication:01/01/2004