Late Pleistocene Emergence of Crystalline Canadian Shield Sources in Sediments of the Northern Gulf of Mexico

The regolith hypothesis attributes the inferred growth of a thicker and more voluminous Laurentide Ice Sheet (LIS) across the mid-Pleistocene transition (MPT) to the removal of softer subglacial substrates of either regolith or sedimentary cover and subsequent exposure of crystalline bedrock with higher basal friction at the base of the LIS. This hypothesis predicts that sediment derived from the LIS was dominated by weathered sources prior to the MPT and then transitioned to fresh Canadian Shield sources during and after the MPT. Here, we present clay mineralogy, modified indices of chemical alteration, K/Ar ages and radiogenic isotopes (Sr, Nd, Pb) of Pleistocene clay detrital fractions from Ocean Drilling Program Site 625 in northeastern Gulf of Mexico to evaluate changes on glacial-interglacial and longer-term timescales in sediment composition during the Pleistocene and test this prediction. Mineralogical and geochemical proxies show a predominance of chemically weathered clays (kaolinite and smectite with clay ratios >= 3.0) sourced from the Appalachians as determined by radiogenic isotopes and K/Ar ages throughout the site's three million-year sediment record. This is consistent with extensive pre-Pleistocene weathering of Caledonian-Appalachian (similar to 450 Ma) material that covered North American cratonic sources within the Mississippi drainage basin. However, following the MPT, we observe a unique addition of physically weathered clays (chlorite and illite with clay ratios <3.0) with radiogenic isotopes and K/Ar ages indicating contributions from more ancient geological sources, most simply explained as sediment derived from physical erosion from the crystalline Canadian Shield. These observations are generally consistent with predictions made by the regolith hypothesis.