Document Type




Degree Granting Department

Marine Science

Major Professor

Amelia E. Shevenell


Ba/Ca, d18O, meltwater, Orca Basin, redox, Younger Dryas


During the last deglaciation (ca. 24-10 ka thousand years ago (ka)), the North American Laurentide Ice Sheet (LIS) was a major source of meltwater to the Arctic Ocean, North Atlantic Ocean, and the Gulf of Mexico (GOM), and it is hypothesized that meltwater routing played an important role in regulating Late Quaternary millennial-scale climate variability, via its influence on Atlantic Meridional Overturning Circulation (AMOC). For example, the meltwater routing hypothesis predicts that a rerouting of meltwater from the GOM to the North Atlantic and/or Arctic Oceans resulted in a decrease of North Atlantic Deep Water (NADW) formation and subsequent cooling in the northern North Atlantic region, at the onset of the Younger Dryas (ca. 13 ka). The GOM was an important outlet for meltwater that likely originated from the southern margin of the LIS. Northern GOM sediments document episodic LIS meltwater input via the Mississippi River throughout the last deglaciation, and further study may provide insight to the evolution of LIS deglaciation and the hydrological response of meltwater flux to the marine depositional environment of GOM. Here, a multi-proxy geochemical study, based on marine sediments from Orca Basin, in northern GOM, aims to 1) reconstruct high-resolution records of deglacial (ca. 24-10 ka) LIS melting history to assess linkage between meltwater input to the GOM and deglacial climate change; 2) investigate the relationship between marine-based records of meltwater input and terrestrial evidence for continental deglaciation to reconstruct LIS drainage patterns within the Mississippi River watershed; and 3) reconstruct the redox state of Orca Basin sediments to evaluate the potential role of turbidity flows as a means of meltwater transport into the northern GOM. All data for this study is from core MD02-2550, a 9.09 m long giant box core, recovered from 2248 m water depth from the Orca Basin, approximately 300 km southwest of the modern Mississippi River delta. High sedimentation rates (45 cm/thousand years (kyr)) and 0.5 to 2 cm sampling resolution allow for sub-centennial sampling resolution. An anoxic hypersaline brine lake currently occupies the bottom 200 m of Orca Basin; yet, visible laminations and color changes that suggest episodic suboxic to anoxic sedimentary conditions during deglaciation, possibly related to LIS meltwater input and/or local biologic productivity.

In chapter one, paired d18O and Mg/Ca-sea-surface temperature (SST) analyses on two varieties of the surface-dwelling planktic foraminifera Globigerinoides ruber (G. ruber; (white and pink, separately)) are used to reconstruct deglacial changes in GOM seawater d18O (d18Osw). Once corrected for global ice volume, the ice volume-corrected d18Osw (d18Oivc-sw) record is primarily influenced by LIS meltwater. d18Oivc-sw records document negative excursions at ca. 19-18.2, 17.5-16.2, 15.3-14.8, and 13.7-13 ka, interpreted as four LIS melting events, followed by the cessation of meltwater at the onset of the Younger Dryas (12.9 ka). Additionally, LIS melting at ca. 17.5 ka suggests that enhanced seasonality in the North Atlantic produced mild summers sufficient for ice sheet retreat during the Mystery Interval (17.5-14.5 ka) despite extremely cold winters.

Because of the inherent difficulties in quantifying meltwater flux using d18Oivc-sw data, foraminiferal (G. ruber) Ba/Ca data are generated in chapter two to assess the influence of LIS meltwater on GOM salinity (a function of meltwater flux) during deglaciation. Ba concentrations in the Mississippi River are elevated relative to GOM seawater and are negatively correlated to sea-surface salinity. Because foraminiferal Ba/Ca (Ba/Caforam) exhibits a predictable relationship to the Ba/Ca of seawater (Ba/Casw), it may be used to calculate changes in salinity arising from deglacial variations in Mississippi River discharge. A complicating factor for Ba/Ca-based salinity interpretations is that Ba concentrations vary spatially throughout the Mississippi River watershed. For example, modern Missouri and Upper Mississippi River Ba concentrations (633 and 436 nM, respectively) are higher than that of the Ohio River (253 nM). Thus, GOM Ba/Ca variability could reflect changes in total Mississippi River input and/or shifts in the dominant region of LIS melting. Applying the modern spatial variability of Ba, we can gain insights into the pattern of ice retreat along the southern margin of the LIS during the last deglaciation.

d18Oivc-sw and Ba/Ca results suggest that meltwater, originating from the Great Lakes region, entered the GOM at ca. 19.0 ka and may have contributed to global sea level rise. A melting event at ca. 17.5 ka coincided with Lake Erie Lobe retreat and may have preconditioned the North Atlantic for AMOC instability during the Mystery Interval (ca. 17.5-14.5 ka). Elevated GOM Ba/Ca (ca. 15.6 to 14.0 ka) suggests greater meltwater input from the Ba-rich Missouri and Upper Mississippi River watershed during the second half of the Mystery Interval (ca. 16.1-14.5 ka), when wet climate conditions prevailed in the southwestern United States and Central America. Overall, Ba/Ca and d18Oivc-sw data suggest large variations in the delivery of meltwater to the Mississippi River and GOM during the last deglaciation.

In chapter three, a suite of redox sensitive trace metals (Mo, Re, U, Mn) from bulk sediment samples are analyzed to reconstruct the redox state of Orca Basin sediments, from the Last Glacial Maximum through the early Holocene (24-7 ka). Variations in the redox state of Orca Basin sediments during deglaciation may be due to changes in local biologic productivity, sediment transport, and/or regional/global physical oceanography. Laminated sediments enriched with authigenic Mo, Re, and U, suggest suboxic to anoxic conditions coincident with high total organic carbon fluxes and LIS meltwater input at ~17.0 ka. Low authigenic trace element concentrations, high quantities of terrigenous material, and abundant Cretaceous-age nannofossils in a 19-cm homogenous interval indicate a turbidite in Orca Basin at ca. 14.4 ka. This stratigraphic unit correlates with evidence from Pigmy Basin, and the Louisiana Shelf, suggesting increased meltwater flux may reflect LIS contribution to Meltwater pulse 1a (MWP-1a) sea level rise. Trace element records coupled with analyses of Orca Basin sedimentary structures will likely improve understanding of deglacial water column stratification, how meltwater entered the GOM (i.e. as a buoyant cap or at depth via sediment-laden hyperpycnal plumes), and the affects of glacial meltwater on marine biologic productivity.