Shallow carbonate geochemistry in the Bahamas since the last interglacial period

Abstract

The carbon isotope composition (δ13C) of ancient shallow-water carbonates frequently is used to reconstruct changes to Earth's global carbon cycle and to perform chemostratigraphic correlation. However, previous work demonstrates that local banktop processes also exert an important control on shallow carbonate δ13C as well as other isotope systems like δ18O. To effectively interpret ancient δ13C records, we must understand how both global carbon cycle perturbations and changes to local conditions are translated to the stratigraphic record. Modern environments, while imperfect analogues, can serve as a guide for interpreting physical and geochemical records of more ancient environments. Shallow carbonate strata from the most recent Pleistocene glacial cycles, which drove high-amplitude perturbations to sea level, temperature, and pCO2 without significantly altering the δ13C of global-mean seawater DIC, present an opportunity to begin untangling signals of global and local processes. However, the geochemistry of Pleistocene platform carbonates largely was overprinted by dissolution and meteoric diagenesis during glacial sea level lowstands. To understand how shallow carbonate geochemistry has changed during the Pleistocene, we instead look to the periplatformal slope and proximal basins. These deep environments serve as a refuge for carbonate produced on the shelf and exported to the slope, and contain a record of shallow carbonate geochemistry that persists across glacial cycles. We study 21 short piston cores from around Bahamian platforms to quantify differences in banktop production and geochemistry between the Holocene interglacial, the last glacial period, and the last interglacial (LIG) period. We show that mud production persists on the periplatformal slopes during the last glacial period, but differences in geochemistry between glacial and interglacial carbonates are a complex function of surface conditions and diagenesis. In contrast, Holocene and LIG carbonates show no evidence of post-depositional alteration, and offer the chance to study differences in δ13C and δ18O between interglacials. We find that while the δ13C of aragonite mud is the same during the Holocene and LIG, the LIG carbonate factory may have delivered more aragonite mud to the periplatform. In addition, the mean δ18O of this mud is elevated compared to the Holocene. We posit that these differences are caused by changes to regional climate and LIG surface conditions.