9706522 The large lakes of East Africa are sensitive monitors of continental-scale tropical hydrology that provide long continuous decadal-scale archives of water balance and temperature to test evolving models of tropical ocean moisture transport. Our problems is to read and quantify such signatures. Prior studies in the region have lacked the tools necessary to unequivocally separate temperature and precipitation effects. In contrast, this study proposes to apply new techniques of oxygen and hydrogen isotopic analyses of aquatic cellulose and lipids that will allow us to differentiate the precipitation signal. The objectives are to quantify the magnitude and timing of shifts in the water balances of Lakes Albert, Edward, and Victoria over the last 15,000 years and determine their implications for the local and regional biosphere. The time-series of water dynamics in these lakes will also be compared with the history of vegetation and civilization along the Nile. The workplan includes compilation of a modern terrestrial and aquatic cellulose/ 18O database to confirm the temperature-independent and species-independent oxygen isotopic biochemical fractionation factor of aquatic cellulose. This database will be used to constrain paleohydrologic interpretations. Analyses of available core samples from all three lakes at less than 300 year increments will provide aquatic cellulose and hence lake water oxygen isotopic time-series profiles over the last 15,000 years. These primary oxygen isotopic signals of the lake waters will derive from shifts in P:E ratios and residence times with negligible groundwater and high mountain water effects. Additional analyses in aquatic lipids are expected to reflect D values of the lake water. Careful consideration of interactions between the watershed and lake water hydrologies as well as of past ocean isotopic values may allow testing for the history of moisture sources (Atlantic vs. Indian Ocean) to these basin. This study should provide sensi tive new approaches to the study of paleohydrology in tropical regions and help constrain paleoclimate model predictions.
