All Articles

The GEAT department welcomes Dr. David Gillikin, Union College, as our next seminar speaker.

Title: Reconstructing hydrological changes over the past century in Central Africa using freshwater bivalve shells

When: Tuesday, October 3 @ 4:10 pm
Where: 2050 Agronomy Hall

Abstract: Bivalve shell oxygen isotope ratios (δ¹⁸O_s) record both water oxygen isotope values (δ¹⁸O_W) and water temperature. However, in the tropics, temperature variations are minimal and variations in δ¹⁸O_W values dominate the shell signal. Therefore shells can be used as archives of variations in δ¹⁸O_W values as we have shown in two major African river systems. δ¹⁸O_W values in turn predictably vary with changes in the hydrologic cycle – including precipitation and evaporation. Biweekly water samples and in-situ measurements collected over a period of more than 2 years from the Oubangui (Bangui, Central African Republic) and Niger (Niamy, Niger) Rivers allowed a direct comparison with proxies recorded in shells. Seasonal minima and maxima in δ¹⁸O_s agreed well with predicted values calculated based on temperature and δ¹⁸O_W values. My research group then investigated δ¹⁸O_s values of museum-archived shells from the Oubangui River. Archived specimens cover the time periods between 1881 and the 1957. δ¹⁸O_shell analyses on all pre-1950s specimens show that seasonal maxima in δ¹⁸O_shell values are significantly lower (~-2.5 ‰) than in recent shells (+0.2 ‰), while seasonal minima (-4.3 ‰) are in closer agreement with recent shells. This strongly suggests the dry season in the upper Oubangui basin has become drier in recent decades, which confirms decreased river discharge over this time. Interestingly, a shell collected just 160 km to the west in the Ngotto rain forest is similar to the pre-1950s shells. This suggests that the drier conditions in the Bangui region seen post 1950s is a regional phenomenon and is likely driven by precipitation‐evaporation (P-E) budgets and not necessarily by changes in rainfall δ¹⁸O values. Changes in P-E can be caused by land-use change or increased evaporation from warmer air temperatures. This suggests that land‐use change and/or climate change is significantly threatening local water resources in Central Africa. My group is working to expand these results and will apply this method across the African continent.