Reconstructing Hydrological Seasonality in the Oubangui River (Congo basin, Africa) Through Stable Isotope Data in Freshwater Bivalve Shells

Abstract:

We examined if freshwater bivalve shells can be used as a proxy of oxygen stable isotope signatures of river water (δ¹⁸O_water) and discharge by comparing daily discharge and δ¹⁸O_water data collected over 5 years (biweekly) from the Oubangui River (at Bangui, Central African Republic) with δ¹⁸O variations along the growth axis of freshwater bivalve shells collected from the same site. Shells of different size were collected between 2011 and 2013. δ¹⁸O_shell values of all samples show a sinusoidal pattern, which correlate with annual hydrographical cycles, as well as with δ¹⁸O_water values, reflecting different sources of water during dry and wet season. Seasonal minima and maxima in δ¹⁸O_shell values agreed well with δ¹⁸O_water values, which ranged between -3.7 ‰ and +2.6 ‰, while discharge ranged between 293 and 8868 m³ s^-1 during the reference time interval (between August 2009 and December 2013). The relationship between discharge and δ¹⁸O_water values fits a logarithmic regression (R²=0.76), which we used to calculate discharge after reconstructing δ¹⁸O_water values from δ¹⁸O_shell data. Because of the logarithmic relationship between discharge and δ¹⁸O_water values, δ¹⁸O_water values are not sensitive to the average to high discharge range, but δ¹⁸O_water values shift significantly towards higher values during low flow conditions, which allows the reconstruction of dry period discharge. Comparing observed discharge data with calculated values indeed show that shell data can mainly be used as records of low discharge situations, and suggest that shell growth may regularly cease during high discharge (above 4000-5000 m³ s^-1). Overall, our data indicate that freshwater bivalve shell isotope data hold high potential to reconstruct long-term patterns in riverine hydrological conditions.