Hydrogeology and water quality of wetlands in East Africa : case studies of a floodplain and a valley bottom wetland

This thesis is embedded in the interdisciplinary BMBF-funded project “GlobE – Wetlands in East Africa”, which focuses on reconciling future food production in East African wetlands with the concurrent environmental protection of these valuable ecosystems. In this context, the aim of the work presented here was to characterize and evaluate the hydrogeological situation of those wetlands. To this end, two study sites were selected, which represent the major wetland types prevalent in East Africa, floodplains and valley bottom wetlands. The Ifakara study site is located in the northeastern Kilombero Catchment in southcentral Tanzania, which comprises a huge floodplain wetland. The Namulonge study site is represented by asmall valley bottom wetland that is part of a huge system of wetlands in southcentral Uganda. Both wetlands are agriculturally used, and an intensification of crop production is expected in future.As agricultural production and other ecosystem services in wetlands highly depend on water quantity and quality, the assessment of hydrological processes within wetlands is mandatory. Wetlands are open systems, which interact with other water components in their catchments. Groundwater has been recognized as one important component, interacting with wetlands on different temporal and spatial scales. This thesis provides the first hydrogeological characterization of thetwo study sites, focusing on the analysis of hydrogeological and hydrochemical processes. Considering the lack of reliable dataabout both sites, in a comprehensive approach aquifer structure, water dynamics, and water composition were assessed thoroughly. A combination of geological, hydraulic, and tracer-based methods was applied, to provide an overview of all processes and to outline the systems’ interconnections. The results were summarized and visualized in hydrogeological conceptual models. Hydrochemical modeling was performed to quantify the processes and to enhance the conceptual models. These models help scientists and stakeholders as information and decision tools. They arefurthermore necessary for other research activities within the GlobEproject in terms of boundary conditions and general hydrogeological conclusions. The status quo of the water quality was evaluated for its suitability as drinking water, for domestic use, and irrigation.Additionally, current as well as future mutual effects betweenagricultural production and water quality were identified and discussed. Both studied wetlands are located on alluvial sediments that are variable in terms of geological and hydraulic properties. The sediments are derived from a saprolite weathering profile ofcrystalline rocks. However, deposition environments, sedimentological features, and geomorphology are highly different between the two study sites. The Ifakara wetland is located on thick sediments south of an escarpment made up of crystalline rocks. The weathering profile above these rocks has been extensively eroded. At Namulonge, the sediments are thinner and cover the weathering profile of crystalline rocks only in valley bottoms. Here, the weathering profile is much thicker due to prolonged weathering with minor tectonic movement. Hydrochemistry at both study sitesis mainly influenced by silicate weathering, which occurs within each lithological unit. Groundwater level measurements and tracer-based assessments revealed that both wetlands interact with groundwater, though these interactions are highly variable in space and time. At Ifakara, groundwater flows from a fractured rock aquifer in the mountains north of the wetland through an unconfined sedimentaryaquifer to the wetland in the south most of the year. Finally, itdischarges into the wetland, made up of stream and flooding water. Wetland water does not contribute significantly to groundwater recharge. Main recharge occurs due to direct infiltration of precipitation in the mountains and in an alluvial fan north of the floodplain. Stream water is derived from precipitation of the entirecatchment and groundwater discharge. Flooding water near the main stream is in direct contact to the stream and originates from over-bank flow and little groundwater discharge. In contrast, flooding water at a greater distance to the stream comprises precipitation and slight groundwater discharge. Evaporation rates of flooding water increase with greater distance to the stream. In the Namulonge area, regional groundwater flows through aweathering profile aquifer from the upper to the lower catchment as well as from the hilltops to the valley bottoms. This regional aquifer mainly receives direct recharge from heavy rainfall events on top of the hills and at the slopes. Local groundwater flow within a smaller confined porous aquifer of valley sediments follows the same flowpattern from the upper to the lower catchment and from the valley fringes to the valley bottoms. This local groundwater is recharged by infiltrating precipitation and surface runoff in the valley of the upper catchment. Moreover, it receives inflow from groundwater of the weathering profile aquifer by upwards discharge below the valley and subsurface runoff from the slopes. Groundwater discharges to the main stream of the wetland in the upper catchment. In the lower catchment, groundwater is separated from wetland water by an impermeable clay layer. Percolation of soil water to groundwater is hampered by the clay layer here, inducingconfined conditions, which can be of an artesian character during rainy seasons. Flooding water mainly originates from over-bank flow and precipitation. Inorganic chemical groundwater quality in terms of drinking water is comparatively good at both study sites. However, at Ifakara, a geogenic contamination with manganese was detected. Based on the spatial analysis of this contamination, it is highly recommended to inform local population about associated hazards and possible solution strategies. Currently, groundwater in the study sites is notsignificantly influenced by the input of agrochemicals. For example, only few groundwater samples showed elevated concentrations of nitrate and nitrite. In the Ifakara study site, such elevatedconcentrations either originate from waste water discharge of Ifakara City or from leaching of nitrogen from agriculturally used soils in the uplands surrounding the wetland. The latter one is also true for the Namulonge study site, indicating that upland agriculture is more hazardous to groundwater quality than wetland agriculture in both study sites. Regarding irrigation, many water components showed low electrical conductivities combined with high sodium concentrations, inducing a decrease in the infiltration capacity of the soil. In terms of big scale irrigation schemes, which are planned forIfakara, the irrigation water needs to be chosen with precaution, considering water quality aspects, to ensure a sustainable crop production. Based on the developed conceptual models and the evaluation of water quality, recommendations are given for stakeholders, in order to develop a first framework for a sustainable management of wetlands. Until now, only a small influence of agricultural production on water quality was found and water purification of both wetlands is considered as good. Nevertheless, the plannedintensification of agricultural production in the wetlands will most probably include an increased amount of applied agrochemicals and maybe lower the groundwater table due to water consumption forirrigation. This might in the future lead to degradation of water quality within and in the vicinity of the wetlands. This thesis emphasizes the need for more case studies aboutgroundwater in East African wetlands, due to the high variability ofhydrogeological and hydrochemical settings and processes. The applied methods can be transferred to other East African wetlands, as it was proven that a combination of quantity and quality related assessments yields reliable results in regions with very limited data