Paleoceanographic Evolution of the Antarctic Southern Ocean since the Mid-Pleistocene Transition

Abstract

The Pleistocene epoch was characterized by orbitally-forced climate oscillations between warm stages and ice ages. The concentration of atmospheric CO2 (pCO2) has varied in step with these so-called glacial-interglacial cycles over at least the last 800 thousand years (kyr), with consistently 80–100 parts per million per volume (ppmv) lower pCO2 during ice ages. The Southern Ocean, a large water body that entirely encircles the Antarctic continent, exerts a dominant control on the partitioning of CO2 between the ocean interior and the atmosphere through its leverage on the efficiency of the biological pump. In the modern Southern Ocean, nutrient- and CO2-rich deep waters ascend to the surface ocean where iron limitation restricts the fixation of the major nutrients by phytoplankton, allowing for the evasion of deeply sequestered carbon to the atmosphere. In the Antarctic Zone of the Southern Ocean, south of the Antarctic Polar Front, the evasion of CO2 was reduced during ice ages by increased sea- ice cover and/or by a cooling-induced increase in stratification. In the northward Subantarctic Zone, a glacial increase in dust-derived iron was suggested to have stimulated marine export production, thereby contributing to enhanced deep ocean sequestration of carbon. While these two regions of the Southern Ocean provide a coherent two-part mechanism to explain the bulk of the glacial-interglacial pCO2 variations, the specific combination of processes modulating atmospheric pCO2 on longer time scales are not fully understood. The main focus of this thesis is on the mid-Pleistocene transition (MPT; ~1.2 to 0.7 million years ago (Ma)), when the climate cycles shifted from 41- to ~100-kyr periodicities in the absence of any substantial changes in the orbital parameters that control the amount of incoming solar radiation. Many of the proposed hypotheses are related to internal feedbacks within the climate system, and involve global cooling and an associated decline in glacial atmospheric CO2. Whereas evidence suggests that the ice ages prior to the transition were characterized by atmospheric CO2 concentrations 30–40 ppmv higher than today, the mechanisms accounting for the CO2 drawdown remain elusive. Using a variety of proxies from marine ODP Site 1094, situated in the Atlantic Sector of the Southern Ocean (53.2 °S, 5.1 °E, 2807 m), this thesis assessed if changes in the Antarctic Zone might have contributed to the presumed pCO2 decline across the MPT, and investigated the coupling between deglacial mechanisms and oceanic CO2 evasion during the past 1.5 Million years (Myr). The chronology of ODP 1094 is based on a newly acquired stratigraphy derived from oxygen isotope ( δ18O) variations recorded by bottom-dwelling foraminifera. One of the main goals of the thesis was to find evidence for potential changes in the stratification across the ice ages of the MPT by reconstructing the vertical density gradient of the Antarctic Ocean. The combination of paired Mg/Ca and δ18O measurements on surface- and bottom-dwelling foraminifera allows estimation of the temperature and δ18O of the surface and deep waters throughout the past 1.5 Myr, both of which exert primary control on the density gradient (seawater δ18O is closely related to salinity). The temperature gradient was determined by Mg/Ca paleothermometry applied to planktonic Neogloboquadrina pachyderma (sinistral) and benthic Melonis pompilioides. The occurrence of manganese-rich coatings had a significant effect on their bulk Mg/Ca ratios, thereby biasing seawater temperature reconstructions. We have corrected for the Mg in the Mn coating by determining the coating Mg/Mn ratio that was assumed to be relatively stable throughout the past 1.5 Myr. Importantly, the samples covering the glacial maxima are largely unaffected by contamination from Mn-rich coatings, and have a negligible influence on the reconstructed vertical gradients. To convert the M. pompilioides Mg/Ca values to deep water temperature, we have compiled a novel core top data set based on new and published data that allowed the establishment of a Mg/Ca-temperature calibration of improved quality. Focusing on the glacial maxima, we have reconstructed a contraction of the thermal gradient and a glacial expansion of the salinity gradient at around MIS 16 (~650 ka). This mode switch makes the glacial density gradient to be primarily dominated by salinity that tends to vertically stabilize the water column in the polar oceans. The intensification of Antarctic Ocean stratification occurred late in the MPT when the glacial pCO2 level was already at its minimum. Instead, the presumed reduction in glacial pCO2 across the MPT might have been primarily caused by a gradual increase in dust-borne iron deposition to the Subantarctic Ocean starting at ~1.2 Ma, which was suggested to have increased the sequestration of carbon into the deep ocean. Concurrently, glacial sedimentary carbonate from ODP 1094 started to decrease, possibly indicating a long-term shoaling of the lysocline as a response to increased glacial carbon sequestration. The reconstructed increase in glacial stratification coincided with an increase in ice volume and the emergence of the dominant high-amplitude ~100 kyr glacial cycles. We speculate that the decrease in overturning at the end of the transition might have further enhanced the efficiency of carbon sequestration by focusing the accumulation of regenerated carbon in the abyss. The reduced evasion of CO2 through the Antarctic surface ocean might have been crucial to drive increased build-up of ice on the northern hemisphere, which could have allowed the ice sheets to survive periods of obliquity-paced summer insolation maxima on a more regular basis than at the beginning and during the MPT. Hence, while the gradual increase in Subantarctic iron fertilization starting early in the MPT, and the weakening of the thermohaline circulation at ~900 ka might have helped to cross a threshold that allowed the development of the 100-kyr climate signal, we propose that the intensified Antarctic Ocean stratification at the end of the transition was crucial in “locking in” the switch to a colder climate with more prolonged and severe ice ages. While the Subantarctic mechanisms likely had a higher leverage on the carbon cycle during glacial maxima, we found confirming evidence that the deglacial mechanisms in the Antarctic Zone are crucial for controlling the interglacial pCO2 level. This is best illustrated by the high correlation between deglacial export production (primarily related to open ocean overturning), deglacial deep ocean oxygenation (primarily related to coastal ventilation) and interglacial pCO2 level. The robust coupling between the deglacial Antarctic mechanisms and the evasion of CO2 from the deep ocean is supported by the changes observed during the lukewarm interglacials. During MIS 13 and 17, when peak pCO2 concentrations were lowest within the past 800 kyr, deglacial Antarctic export production and deep ocean oxygenation were very low, indicating that Antarctic overturning and ventilation both ceased dramatically. The lower oxygen levels in the deep Southern Ocean coincide with reduced deglacial carbonate peaks, suggesting that proportionally less CO2 has escaped from the deep ocean. Assuming a primary dependence of deglacial Antarctic overturning on the interglacial pCO2 level, the deglacial export production peaks were used to predict the interglacial pCO2 concentrations of the past 1.5 Myr. The results of this thesis contribute to our understanding of the Southern Ocean’s role in regulating atmospheric pCO2 since the mid-Pleistocene transition. Using a variety of proxies, the 1.5 Myr-spanning records complement previously published data from the same core site, providing a holistic picture of the Pleistocene climate variability in the Antarctic Southern Ocean.