A Quantified Reconstruction of Amazon River Discharge for the Last 40 kyr

The glacial moisture history of the Amazon Basin is comparatively poorly known. Previous estimates have been inferred from highly localised and qualitative indicators of effective moisture, and debate exists as to whether the glacial Amazon Basin was humid or dry. Reconstructing glacial Amazon aridity is essential for three main reasons: 1) Aridity is a key physiological control on the distribution of vegetation and therefore provides a means of testing the Pleistocene tropical rainforest refuge hypothesis; 2) Amazonian wetlands represent a major source of atmospheric methane and may exert a primary control on the ice core methane records; 3) the Amazon Basin is a major source of tropical atmospheric water vapour, another important greenhouse gas. Here we present unequivocal evidence to suggest widespread aridity within the Amazon Basin during the Last Glacial Maximum (LGM) and also the Younger Dryas (YD). We have used δ¹⁸O of planktonic foraminifera from ODP Site 942 on the Amazon Fan to quantify Amazon River discharge for the last 40 kyr, employing an age model constructed around 35 ¹⁴C AMS dates. Our calculations suggest outflow to have been c. 70% relative to modern during the LGM (i.e. c. 30% reduced), and c. 60% relative to modern during the YD. Sedimentation rates are relatively more elevated during the YD with two distinct peaks around 11200 and 13500 Cal yr BP. Centennial and millennial-scale variability in the data are most likely climate-driven.