Lamy, F., Winckler, G., Alvarez Zarikian, C.A., and the Expedition 383 Scientists
Proceedings of the International Ocean Discovery Program Volume 383
publications.iodp.org

https://doi.org/10.14379/iodp.proc.383.101.2021

Expedition 383 summary¹

G. Winckler, F. Lamy, C.A. Alvarez Zarikian, H.W. Arz, C. Basak, A. Brombacher, O.M. Esper, J.R. Farmer, J. Gottschalk, L.C. Herbert, S. Iwasaki, V.J. Lawson, L. Lembke-Jene, L. Lo, E. Malinverno, E. Michel, J.L. Middleton, S. Moretti, C.M. Moy, A.C. Ravelo, C.R. Riesselman, M. Saavedra-Pellitero, I. Seo, R.K. Singh, R.A. Smith, A.L. Souza, J.S. Stoner, I.M. Venancio, S. Wan, X. Zhao, and N. Foucher McColl²

Keywords: International Ocean Discovery Program, IODP, JOIDES Resolution, Expedition 383, Dynamics of the Pacific Antarctic Circumpolar Current, Site U1539, Site U1540, Site U1541, Site U1542, Site U1543, Site U1544, Southern Ocean, South Pacific, Chilean margin, paleoceanography, Antarctic Circumpolar Current, oceanic fronts, Circumpolar Deep Water, Antarctic Intermediate Water, marine carbon cycle, dust, biological productivity, iron fertilization, southern westerly winds, Patagonian ice sheet, West Antarctic ice sheet

MS 383-101: Published 18 July 2021

Abstract

The Antarctic Circumpolar Current (ACC), the world’s strongest zonal current system, connects all three major ocean basins of the global ocean and therefore integrates and responds to global climate variability. Its flow is largely driven by strong westerly winds and is constricted to its narrowest extent in the Drake Passage. Fresh and cold Pacific surface and intermediate water flowing through the Drake Passage (cold-water route) and warm Indian Ocean water masses flowing through the Agulhas region (warm-water route) are critical for the South Atlantic contribution to Meridional Overturning Circulation changes. Furthermore, physical and biological processes associated with the ACC affect the strength of the ocean carbon pump and therefore are critical to feedbacks linking atmospheric CO₂ concentrations, ocean circulation, and climate/cryosphere on a global scale. In contrast to the Atlantic and Indian sectors of the ACC, and with the exception of drill cores from the Antarctic continental margin and off New Zealand, there are no deep-sea drilling paleoceanographic records from the Pacific sector of the ACC. To advance our understanding of Miocene to Holocene atmosphere-ocean-cryosphere dynamics in the Pacific and their implications for regional and global climate and atmospheric CO₂, International Ocean Discovery Program Expedition 383 recovered sedimentary sequences at (1) three sites in the central South Pacific (CSP) (U1539, U1540, and U1541), (2) two sites at the Chilean margin (U1542 and U1544), and (3) one site from the pelagic eastern South Pacific (U1543) close to the entrance to the Drake Passage. Because of persistently stormy conditions and the resulting bad weather avoidance, we were not successful in recovering the originally planned Proposed Site CSP-3A in the Polar Frontal Zone of the CSP. The drilled sediments at Sites U1541 and U1543 reach back to the late Miocene, and those at Site U1540 reach back to the early Pliocene. High sedimentation rate sequences reaching back to the early Pleistocene (Site U1539) and the late Pleistocene (Sites U1542 and U1544) were recovered in both the CSP and at the Chilean margin. Taken together, the sites represent a depth transect from ~1100 m at Chilean margin Site U1542 to ~4070 m at CSP Site U1539 and allow investigation of changes in the vertical structure of the ACC, a key issue for understanding the role of the Southern Ocean in the global carbon cycle. The sites are located at latitudes and water depths where sediments will allow the application of a wide range of siliciclastic-, carbonate-, and opal-based proxies to address our objectives of reconstructing, with unprecedented stratigraphic detail, surface to deep-ocean variations and their relation to atmosphere and cryosphere changes.

Introduction

Our prior knowledge of southern high latitude paleoceanography comes from conventional sediment coring and Deep Sea Drilling Project (DSDP)/Ocean Drilling Program (ODP) drilling in the Atlantic and Indian sectors of the Antarctic Circumpolar Current (ACC). A prominent example is ODP Leg 177, which drilled sites along a north–south transect across the major ACC fronts and documented the Cenozoic history of the Atlantic sector of the Southern Ocean. These sites revealed, for example, linked changes in dust supply, marine productivity, and biological nutrient consumption (Martínez-Garcia et al., 2009, 2011, 2014), opposing changes in productivity in the Antarctic compared to the Subantarctic Zone on glacial–interglacial cycles (Jaccard et al., 2013) and in the ACC compared to the Antarctic margin on longer timescales (Cortese et al., 2004), marked steps in sea-surface temperature (SST) evolution since the Pliocene (Martínez-García et al., 2010), and Antarctic Ice Sheet (AIS) dynamics during the late Quaternary (Kanfoush et al., 2002; Teitler et al., 2010, 2015; Venz and Hodell, 2002; Hayes et al., 2014).

In contrast, studies of the Pacific sector of the ACC have been limited mainly to regions close to the Antarctic continental margin south of the ACC (e.g., Ross Sea during DSDP Leg 28 and Antarctic Peninsula during ODP Leg 178). Integrated Ocean Discovery Program Leg 318 drilled Cenozoic sediments off Wilkes Land (eastern Indian Southern Ocean) and revealed significant vulnerability of the East Antarctic Ice Sheet (EAIS) to Pliocene warming (Cook et al., 2013; Expedition 318 Scientists, 2010). A substantial advance in constraining the history and stability of the West Antarctic Ice Sheet (WAIS) and Ross Ice Shelf came from the Antarctic Geological Drilling (ANDRILL) project (e.g., Naish et al., 2009; Florindo et al. 2008), which documented ice sheets that were expanded and stable between 13 and 11 Ma and more dynamic in the late Miocene–late Pliocene (10–2.5 Ma) with cyclic expansion of grounded ice sheets over the coring site after 2.5 Ma.

These on- and near-shore drilling programs have significantly advanced our understanding of AIS dynamics during the Cenozoic. However, to understand how these ice sheet changes are ultimately linked to climate, atmospheric CO₂ levels, and ocean circulation requires constraints on the strength, latitudinal extent, and overall nature of the ACC (i.e., the current system that connects Antarctica to the rest of the globe). In this context, the results of Expedition 383 in the Pacific ACC will be closely linked to the recently completed International Ocean Discovery Program (IODP) expeditions to the Ross Sea (Expedition 374), Amundsen Sea (Expedition 379), and Scotia Sea (Expedition 382). These expeditions targeted Antarctic near-shore records, which will likely be incomplete due to glacial erosion, and the potential for analyses of paleoceanographic proxies may be limited. Therefore, Expedition 383, Dynamics of the Pacific Antarctic Circumpolar Current (DYNAPACC), pelagic sites were selected to provide critical paleoceanographic baseline information, including rates of change, for improving the understanding of reconstructed AIS changes and testing ice sheet models.

Background

The ACC, the world’s largest current system, connects all three major basins of the global ocean (i.e., the Pacific, Atlantic, and Indian Oceans; Figure F1) and therefore integrates and responds to climate signals throughout the globe (e.g., Talley, 2013). Through deep mixing, upwelling, and water mass formation, the ACC is fundamentally tied to the global Meridional Overturning Circulation (MOC; Marshall and Speer, 2012) and the stability of Antarctica’s ice sheets.

Figure F1. Antarctic Circumpolar Current.

The ACC has long been recognized as a key player in regulating atmospheric CO₂ variations and therefore global climate based on the tight coupling between Southern Hemisphere temperatures and atmospheric CO₂ concentrations observed in Antarctic ice cores (e.g., Parrenin et al., 2013). Strong, zonally symmetric westerly winds drive a northward Ekman component of flow that promotes ventilation of deep water, which provides a direct hydrographic link between the large deep-ocean reservoir of dissolved inorganic carbon and the surface ocean (e.g., Kuhlbrodt et al., 2007; Marshall and Speer, 2012). The ACC therefore acts as a window through which the interior ocean and atmosphere interact and as a key player in regulating atmospheric CO₂ variations and therefore global climate.

Biological utilization of nutrients in the Southern Ocean is particularly important in relation to changes in atmospheric CO₂ concentration because it regulates the preformed nutrient inventory for most of the deep ocean and therefore the global average efficiency of the biological pump (e.g., Sigman and Boyle, 2000; Sigman et al., 2010). Nutrient utilization is inefficient in the Southern Ocean today in part because phytoplankton growth is limited by the scarcity of bioavailable iron (e.g., de Baar et al., 1995). Martin (1990) proposed that dust-borne iron fertilization of Southern Ocean phytoplankton caused the ice age reduction in atmospheric CO₂. However, the role of iron in explaining variations in opal export to the Southern Ocean sediments is complex. Chase et al. (2015) found a significant correlation between annual average net primary production and modeled dust deposition but not between dust deposition and opal burial. On glacial–interglacial timescales, proxy records from the Subantarctic region support a positive relationship between dust flux and opal production (e.g., Chase et al., 2003; Bradtmiller et al., 2009; Lamy et al., 2014; Kumar et al., 1995; Anderson et al., 2014). The opposite is true south of the Antarctic Polar Front (APF) (e.g., Chase et al., 2003; Jaccard et al., 2013), indicating that factors other than dust regulate the production and export of opal in the Southern Ocean and, by association, the strength and efficiency of the biological carbon pump. Leading candidates for these other factors include changes related to shifts in the hydrographic fronts, the flow of the ACC, the spatial and temporal variability of wind-driven upwelling that supplies nutrients and CO₂ to surface waters, and sea ice extent.

Oceanographic setting

Expedition 383 drilled sites located in the ACC system in the central South Pacific (CSP), the eastern South Pacific (ESP) and at the southern Chilean margin close to the Drake Passage. The flow of the ACC is largely driven by Southern Westerly Winds (SWW) and is constricted to its narrowest extent in the Drake Passage. This so-called “cold-water route” through the Drake Passage is one important pathway for the return of fresh and cold waters from the Pacific to the Atlantic, which strongly affects the strength of the Atlantic MOC (AMOC), in concert with the “warm-water route” inflow of warm and salty Indian Ocean water masses through the Agulhas Current system (Beal et al., 2011; Gordon, 1986). The Drake Passage is ~800 km wide, and it is located between Cape Horn and the western Antarctic Peninsula (Figure F1). Numerous hydrographic surveys across the Drake Passage since the 1970s have contributed to the understanding of seasonal and interannual variability in ACC transport through the Drake Passage and the mechanisms forcing physical and biological changes in the Southern Ocean (e.g., Meredith et al., 2011). Despite the insight gained from model simulations (e.g., Meijers, 2014) and from hydrographic time series reaching back 20 y (Koenig et al., 2014) in this relatively well constrained region of the ACC, important issues such as the role of zonal winds in forcing ACC transport remain controversial (Watson et al., 2015).

The total ACC flow transported through the Drake Passage is estimated to be between ~130 and 150 Sv (Koenig et al., 2014; Meredith et al., 2011; Renault et al., 2011). Three major ACC oceanographic fronts (Orsi et al., 1995), the Subantarctic Front (SAF), the APF, and the Southern ACC Front (SACCF), occur from north to south in the Drake Passage. The modern winter sea ice (WSI) margin does not extend significantly north into the Drake Passage. Oceanographic studies show that more than half of the total Drake Passage transport occurs at and north of the SAF (i.e., in the Subantarctic Zone), with the majority of the remaining transport in the Polar Frontal Zone (between the SAF and the APF) (Donohue et al., 2016). The relative contribution of Drake Passage transport south of the APF is generally <20%.

At slightly lower latitudes along the southernmost Chilean margin, the southward flowing Cape Horn Current (CHC), a northern branch of the ACC that continues toward the Drake Passage, provides a major fraction of the present day northern Drake Passage transport (Well et al., 2003). Satellite-tracked surface drifters reveal that after crossing the East Pacific Rise (EPR), Subantarctic surface water of the ACC is transported northeastward across the ESP toward the Chilean coast at ~45°S, 75°W (Chaigneau and Pizarro, 2005). Subsequently, only a minor fraction of the ACC is deflected northward into the Humboldt Current System (HCS), whereas the major fraction is deviated southward toward the Drake Passage within the CHC. The CHC thus transports a significant amount of northern ACC water toward the Drake Passage in a narrow belt ~100–150 km wide off the coast (Chaigneau and Pizarro, 2005) (Figure F2).

Figure F2. Surface buoy trajectories.

The CSP drill sites are located at Subantarctic latitudes between 56° and 50°S. Sites U1539 and U1540 are located east of the EPR, whereas Site U1541 lies west of the EPR. Unfortunately, the originally proposed drill site south of the APF could not be reached during Expedition 383 because of persistent bad weather conditions. The drilled sites (U1539–U1541) lie in the pathway of the Subantarctic Pacific ACC, 100–170 nmi north of the modern mean position of the SAF in a zonal transition zone of the ACC. These ACC fronts form the core of the ACC and are partly constrained by the bathymetry of the EPR in this area. As a result of seafloor topography, frontal positions in the ACC close to large bathymetric features remain relatively stable at seasonal and interannual timescale, whereas over abyssal plains their location appears to be fairly unstable (Sokolov and Rintoul, 2009). The SAF at the western and eastern flank of the EPR, where our sites are located, is therefore partly constrained by topography and the location of major fracture zones (e.g., Eltanin-Tharp Fracture Zone). Interannual and decadal oceanographic conditions in the South Pacific are influenced by the Southern Annular Mode, by the Amundsen Sea Low, and by tropical variability and the El Niño-Southern Oscillation (Thompson et al., 2011; Turner et al., 2009; Yuan, 2004). At these timescales, SST and sea ice changes are amplified in the Pacific sector of the Southern Ocean; for example, observations of trends from 1979–2011 indicate a distinct increase in sea ice north of the Ross Sea embayment and a strong sea ice decrease in the Bellingshausen-Amundsen Sea (Maksym et al., 2012; Stammerjohn et al., 2012).

A portion of the deep water that upwells south of the APF becomes Antarctic Bottom Water (AABW) close to the Antarctic margin, while Antarctic Intermediate Water (AAIW) and Mode Water is derived from the same deep water mass north of the APF (e.g., Sloyan and Rintoul, 2001; Talley, 2013). The Southeast Pacific is known as the major AAIW formation area where fresh, cold, high-oxygen and low-nutrient waters are formed. Originally, it was thought that AAIW originates from the subduction of Antarctic surface waters below the SAF (e.g., Deacon, 1937). Later, it was suggested that the formation of AAIW is directly linked to the formation of Subantarctic Mode Water (SAMW) (McCartney, 1977), which has similar properties in the Southeast Pacific. More recent studies confirm AAIW formation occurs primarily in the ESP and is related to deep mixed-layer development leading to SAMW formation and finally to AAIW formation (Bostock et al., 2013). The Southern Ocean’s major water mass, the Circumpolar Deep Water (CDW), is divided into upper and lower branches. Lower CDW (LCDW) is identified by a salinity maximum inherited from North Atlantic Deep Water (Reid and Lynn, 1971; Orsi et al., 1995). Upper CDW (UCDW) is typically identified as an oxygen-minimum layer derived from oxygen-depleted, nutrient-rich Indian Deep Water and North Pacific Deep Water (Callahan, 1972). LCDW and UCDW are transported into the North Pacific, where they are transformed into NPDW, and return to the Southern Ocean in the East Pacific along the South American continent at 1500–3500 m water depth (Kawabe and Fujio, 2010). North of the SAF, UCDW is centered at ~1500 m water depth and LCDW covers water depths below 2000 m. Below LCDW, northward-spreading dense AABW formed in the Weddell and Ross Seas and along the Adélie Coast covers the abyss around Antarctica. In the South Pacific, the expansion of Ross Sea Bottom Water, the coldest and saltiest variety of AABW, is restricted by bottom topography to areas south of the Pacific-Antarctic Ridge (Orsi et al., 1995). Miocene to recent Sites U1539–U1541 and U1543 are currently bathed by LCDW, and Chilean continental margin Sites U1542 and U1544 are currently bathed by AAIW and a mixture of CDW and NPDW, respectively.

Geological setting

The drill sites in the CSP are located on oceanic crust near the EPR. Site U1541 is located ~160 nmi west of the crest of the EPR, north of the Eltanin-Tharp Fracture Zone. The basement age (Figure F3) (Eagles, 2006) at this location is about 10 Ma (late Miocene). Sites U1539 and U1540 are located at a similar distance to the fracture zone but ~210 nmi east of the EPR crest on slightly older crust with a middle Miocene basement age of ~15 Ma.

Figure F3. Basement ages of South Pacific oceanic crust.

One of the ESP drill sites (U1543) is located west of the Chile Trench on a topographically elevated ridge ~300 m above the trench axis (~4200 m water depth). The basement at this location is oceanic crust older than ~20 Ma (Eagles, 2006). The other two ESP drill sites (U1542 and U1544) are located at the Chilean margin on continental crust. The southern Chilean margin is characterized by a complex geodynamic setting with oblique convergence between plates, transcurrent motion, and tectonic rotation on land (Polonia et al., 2007). This portion of the South American margin is influenced by the relative movements between three main plates (Antarctica, Scotia, and South America), resulting in a strong segmentation of the southern Chilean margin, which is clearly visible in bathymetric and multichannel seismic (MCS) data (Polonia et al., 2007). Two major fore-arc basins at ~53°–54.5°S and 55°–56.5°S are characterized by thick sediment infill of as much as ~3 km (Figure F4). We drilled one site in the southern fore-arc basin (U1544), where Seismic Line IT97-171 (Figure F4) documents ~3 km of remarkably undeformed sediments (Polonia et al., 2007). This southern basin most likely formed after ridge subduction (14–10 Ma) when convergence resumed, contributing to the construction of the outer bathymetric high that provided the barrier for the accumulation of continent-derived sediments (Figure F4).

Figure F4. Tectonic setting.

North of ~53°S, subduction in the Chile Trench is more orthogonal. Here, the transition between the undeformed Antarctic abyssal plain and the lower slope is very abrupt and results in a steep slope incised by canyons that run perpendicular to the trench axis. Recent sediment cover is minor at the lower slope, but a small-scale sediment depocenter occurs at the upper continental slope near the western entrance to the Strait of Magellan. Site U1542 is located at this sediment drift, which contains a sediment sequence as thick as ~700 m that most likely does not extend beyond the early Pleistocene because of high sedimentation rates.

Scientific objectives

The overall goal of Expedition 383 is to improve our knowledge of late Miocene to recent atmosphere-ocean dynamics of the ACC in the Pacific and their implications for regional and global climate and atmospheric CO₂ using sediment records with the highest possible resolution. The sites (Table T1; Figure F1) are located at latitudes and water depths where sediments will allow the application of a wide range of siliciclastic-, carbonate-, and opal-based proxies for reconstructing surface to deep-ocean variations and their relation to atmosphere and cryosphere changes with unprecedented stratigraphic detail. The cores recovered during Expedition 383 will allow us to test two major scientific hypotheses.

Hypothesis H1: ACC dynamics and Drake Passage throughflow conditioned the global MOC and high–low climate linkages on orbital and submillennial timescales since the Pliocene.

The role of the Drake Passage throughflow in driving changes in global MOC on glacial–interglacial and millennial timescales is not well constrained (Knorr and Lohmann, 2003; McCave et al., 2014). For the last glacial cycle, Lamy et al. (2015) reconstructed Drake Passage throughflow based on a sediment record (MD07-3128) from the Chilean margin close to Site U1542 and another record from the southernmost Argentinean continental slope and found a significant glacial decrease in Drake Passage throughflow (Figure F5). This finding is consistent with generally reduced current velocities of the ACC during glacials and concomitant reduction in interbasin exchange in the Southern Ocean that is most likely regulated by variations in the SWW field over the Subantarctic Zone and changes in Antarctic sea ice extent. However, the link between ACC flow through the Drake Passage and changes in the AMOC are still not well established.

Figure F5. Central and eastern South Pacific paleoceanography and Drake Passage throughflow.

Past studies suggest that there are likely regional differences in how the ACC varies on glacial–interglacial timescales. For example, downstream of the Drake Passage, a meridional section of sediment records from the Scotia Sea revealed little overall bottom current speed variations between the Last Glacial Maximum (LGM) and the Holocene (McCave et al., 2014). In the Indian sector of the Southern Ocean, a 500 ky record suggests that the ACC was weak during warm stages and strong during glacial intervals (Mazaud et al., 2010). Likewise in the southwest Pacific, the deep western boundary current east of New Zealand intensified during glacial periods over the past 1.2 My (Hall et al., 2001), a change that has been related to a northward extension of the ACC along the New Zealand continental margin. Finally, a recent density reconstruction (based on benthic foraminifers) across the ACC south of Australia suggested a moderate increase of the ACC flow during the LGM (Lynch-Stieglitz et al., 2016). Taken together, there is substantial disagreement about the ACC flow speed. Given the evidence for substantial regional differences in ACC behavior, it has been particularly challenging to make the link between Drake Passage throughflow and changes in the AMOC dynamics on glacial–interglacial timescales.

Superimposed on the long-term glacial reduction of Drake Passage throughflow, prominent high-amplitude, millennial-scale variability was observed, implying that the northern ACC in the Southeast Pacific has a strong sensitivity to, or role in, suborbital climate variations (Lamy et al., 2015). Enhanced Drake Passage throughflow during Antarctic warm periods is consistent with the previously suggested impact of the bipolar seesaw mechanism on the Southern Ocean (e.g., Anderson et al., 2009; Lamy et al., 2007), which includes surface water warming, enhanced upwelling, and a stronger ACC caused by southward-shifted SWW. The glacial reduction of the cold-water route occurred in concert with the well-documented decrease of the warm-water route (Agulhas leakage; Beal et al., 2011). The data currently available thus indicate that both of these oceanographic corridors are critical for the South Atlantic contribution to glacial AMOC strength changes.

Our sites along the south Chilean margin (U1542 and U1544) and in the ESP (U1543) (Figure F2) provide a unique opportunity to extend the existing ESP 60 ky record (Lamy et al., 2015) back into the early Pleistocene and Pliocene to investigate the role of long-term, orbital-scale, and millennial-scale changes in ACC strength and Drake Passage throughflow on Antarctic and global climate change, including warmer-than-present interglacials (e.g., Marine Isotope Stages [MISs] 5e, 11, and 31) and Late Neogene warm periods.

In the Pacific sector of the Southern Ocean, pronounced past changes in surface-ocean properties at both millennial and orbital timescales, particularly SSTs, indicate an exceptional climate sensitivity to external and internal forcing. Glacial–interglacial amplitudes in Subantarctic SSTs are generally >4°C in the Atlantic (Gersonde et al., 2005; Martínez-García et al., 2010), Indian (e.g., De Deckker et al., 2012; Lorrey et al., 2012), and Pacific Southern Ocean (Caniupán et al., 2011; Ho et al., 2012; Pahnke et al., 2003). Millennial-scale fluctuations with amplitudes of ~2°–3°C are observed at continental margin sites in the Subantarctic Pacific during the last glacial (Barrows et al., 2007; Caniupán et al., 2011; Lamy et al., 2004; Pahnke et al., 2003). Furthermore, there is evidence from studies of Subantarctic surface ocean variability that high, middle, and low latitude ocean conditions are strongly linked. High-amplitude glacial–interglacial SST changes in the Subantarctic Pacific sector of the Southern Ocean (Ho et al., 2012) are closely coupled to those observed in the subtropical ESP, implying large shifts of the northern ACC system and export of Subantarctic surface water masses into the HCS. This variability is probably directly related to variations in the strength and position of the ACC fronts, the SWW, and Drake Passage throughflow (Lamy et al., 2015) (Figure F5). Along the Chilean margin, glacial cooling implies a stronger northward extension of the Antarctic cold water in the ESP (Ho et al., 2012) compared to the central Pacific (Figure F5). The significant glacial decrease in Drake Passage transport accompanied by a reduction of the ACC transport strongly reduced the interbasin water mass exchange and enhanced the high–low latitude exchange in the Pacific Eastern Boundary Current system. The reacceleration of Drake Passage transport during the deglaciation matches major warming in the Subantarctic Southeast Pacific and South Atlantic (Barker and Diz, 2014), suggesting a close coupling of Drake Passage throughflow to the initiation of Southern Hemisphere warming.

Evidence for the link between the ESP and far-field oceanographic changes can also be inferred from millennial-scale changes in Drake Passage throughflow (Lamy et al., 2015) that are paralleled by SST changes in the Southeast Pacific (ODP Site 1233) (Kaiser and Lamy, 2010) that extend north to the cold tongue in the eastern tropical Pacific (Lea et al., 2006). These data imply that substantial changes in the oceanographic dynamics of the South Pacific Gyre are related to the northward deflection of ACC waters at millennial timescales. The suggested enhanced glacial northward export of cold-water masses to the tropical Pacific reinforces modeling studies showing the importance of southerly derived surface and intermediate water masses in the tropical Pacific for global cooling during the LGM (Liu et al., 2002). Paleoceanographic reconstructions based on our sites at the Chilean margin (U1542 and U1544) and in the ESP (U1543) in combination with ODP Leg 202 sites (Figure F6) (e.g., Dekens et al., 2007; Lamy et al., 2004; Pena et al., 2008; Rincon-Martinez et al., 2009) and the JR100 expedition (http://iodp.tamu.edu/scienceops/expeditions/JR100_379T.html) sites will allow us to examine, in unprecedented detail, changes in the coupling between the ACC and the tropical Pacific over various timescales and at warmer-than-present climate states.

Figure F6. Pacific ACC sites and cores.

Our drill sites will allow us to better constrain the latitudinal shifts of the Southern Ocean fronts over time to understand the resulting changes in ACC transport, intrabasin mixing, and AAIW production over a wider range of climate states. For example, reconstructions of past changes in the latitudinal position of the SAF suggest as much as 9° latitude northward shifts during glacials, whereas other areas suggest no shift at all (Kohfeld et al., 2013). Smaller shifts probably occurred where the front is bathymetrically confined, like in the vicinity of the EPR (CSP sites). Large shifts likely occurred in the ESP with the SAF potentially located north of the Drake Passage during peak glacials (Gersonde et al., 2003; Ho et al., 2012; Lamy et al., 2015). As stated above, these changes may substantially affect the Drake Passage throughflow and low–high latitude exchange of heat and nutrients with major implications for the global climate development. Farther south, Benz et al. (2016) reconstructed a glacial WSI expansion by ~2°–3° in latitude in the western part of the Pacific sector and an even stronger shift (~5° in latitude) of both the SACCF and the WSI edge in the central part (Figures F6, F7). This west–east differentiation in the winter SST pattern and sea ice extent can be related to differences in topographic and atmospheric forcing (Benz et al., 2016).

Figure F7. LGM winter sea-surface temperatures and sea ice extent.

Despite the potential importance of the Subantarctic Southern Ocean in global climate change, there have been very few studies that focus on the role of the ACC in the evolution of the cryosphere and of the low–high latitude climate system connectivity from the warm Pliocene and into the Pleistocene. Only one SST record is available that reaches back to the Pliocene (ODP Site 1090 in the Subantarctic Atlantic) (Figure F8). Here, the major cooling phase after ~1.6 Ma coincides with the emergence of the eastern tropical Pacific cold tongue and an intensification of the meridional temperature gradients (Martínez-García et al., 2010) (Figure F4). The ANDRILL records suggest that high latitude climate cooling between 3.0 and 2.5 Ma drove both the WAIS and EAIS toward their present expanded cold polar state (Naish et al., 2009). The intensification of Antarctic cooling resulted in strengthened SWW and northward migration of Southern Ocean fronts, which would restrict water exchange between ocean basins beginning after 3.3 Ma (McKay et al., 2012). Warm early Pleistocene and Pliocene SSTs are consistent with the poleward expanded warm pool (Brierley et al., 2009).

Figure F8. Pliocene to recent SST and %C_37:4changes.

Our sites from the CSP (U1539–U1541) will allow a more comprehensive assessment of its temperature evolution and relation to atmospheric variations of the SWW, sea ice, AIS variability, and low latitude climate change during cold and warmer-than-present (e.g., MISs 5, 11, and 31 and the Pliocene warming) time periods. Information on warm periods is critical to the assessment of the Southern Ocean’s role under future warming; this is especially true for the AIS melting and associated sea level rise because ice sheet modeling and modern observations show that oceanic forcing is the most important determinant of AIS melting (Pollard and DeConto 2009; Paolo et al., 2015).

Moreover, temperature records from our high-resolution sites at the Chilean margin (U1542 and U1544) and CSP (U1539 and U1540) will allow us to compare millennial-scale variations between the Pacific Southern Ocean and Antarctica for a time interval that is well beyond the reach of ice cores. Even on a global scale, such millennial-scale temperature records back into the Pliocene are rare (e.g., Integrated Ocean Drilling Program Site U1313 [Naafs et al., 2013] and ODP Site 980 [McManus et al., 1999]).

Hypothesis H2: variations in the Pacific ACC determine the physical and biological characteristics of the oceanic carbon pump and atmospheric CO₂.

Atmosphere-ocean-cryosphere interactions and teleconnections between high and low latitudes play an important role in past and future climate change, and the Subantarctic Southern Ocean provides the major link between Antarctica and the low latitudes. In the Southern Ocean, atmosphere-ocean interactions are believed to control sea ice cover, AIS dynamics, upper ocean stratification, biological nutrient utilization, and exposure rates of deep water. Thus, they have been considered to play a key role in explaining the variability in atmospheric CO₂ concentrations, which are controlled by biogeochemical and physical processes.

The SWW strongly impact the upwelling of carbon-rich deep-water masses in the Southern Ocean that affect atmospheric CO₂ (e.g., Toggweiler 2006, Anderson et al., 2009) and also control the return flow of intermediate waters to the tropics (Figure F9). Moreover, the SWW exert an important control on upwelling of relatively warm deep water at the West Antarctic continental shelf, where it causes sub–ice shelf melting and mass loss of the WAIS (e.g., Steig et al., 2012). Thus, the SWW are thought to play a key role in past and future climate change, yet their strength and position in colder and warmer climates relative to today remain a wide-open question (Kohfeld et al., 2013).

Figure F9. ACC overturning circulation.

In the modern Southern Ocean, deep waters with high dissolved inorganic carbon and major nutrient concentrations are carried to the surface by wind-driven upwelling and density-driven overturning. However, because phytoplankton growth is limited by the scarcity of iron, a substantial fraction of these nutrients return to the subsurface before they are completely consumed. This incomplete utilization of nutrients represents a substantial leak in the modern global biological pump because it allows the escape of deeply sequestered carbon back to the atmosphere, keeping atmospheric CO₂ levels high during interglacial stages. An increase in the efficiency of the global biological pump can be accomplished either by decreasing the physical cycling of deep water through the surface of the Southern Ocean or by increasing the degree to which surface nutrients are consumed by marine organisms (e.g., as a result of increasing iron availability relative to other nutrients) (Sigman and Boyle, 2000; Sigman et al., 2010).

Paleoceanographic records (mainly from the Atlantic sector of the Southern Ocean) indicate that ice age productivity was lower (Anderson et al., 2009; Kohfeld et al., 2005; Jaccard et al., 2013) in the Antarctic zone (i.e., south of the APF) and nutrient consumption was higher (Robinson and Sigman, 2008; Studer et al., 2015). These data indicate that the Antarctic zone was more strongly stratified during ice ages, inhibiting the physical mixing of deep carbon-rich deep water to the surface and therefore contributing to the lowering of atmospheric CO₂ during glacial times.

In the Subantarctic Atlantic, higher dust flux, increased productivity and nutrient consumption, and lower atmospheric CO₂ co-occur during peak glacial times and Antarctic millennial cold events, which has been associated with iron fertilization (e.g., Frank et al., 2000; Kumar et al., 1995; Martínez-García et al., 2009, 2011, 2014). Martínez-García et al. (2009, 2014) suggested that around 40–50 ppmv of the glacial atmospheric CO₂ decline can be attributed to iron fertilization (Figure F10), which is in good agreement with some model estimates (Hain et al., 2010; Chalk et al., 2017; Brovkin et al., 2007). At ODP Site 1090, the first significant rise in ice age dust and iron deposition over the past 4 My occurred at ~2.7 Ma and the most pronounced rise in dust deposition occurred across the mid-Pleistocene transition (Martínez-García et al., 2011).

Figure F10. Paleoproductivity patterns.

For the past 1 My, higher dust deposition during glacial periods compared to interglacials has also been documented for the Pacific Southern Ocean, suggesting an eastward extension of the influence of Australian and/or New Zealand dust sources (Lamy et al., 2014). In this area, dust fluxes follow the same glacial–interglacial fluctuations, but absolute glacial fluxes in the Subantarctic Pacific are about half of those in the Atlantic Southern Ocean (Figure F11). As in the Atlantic sector, export production in the Subantarctic Zone of the Pacific sector over the past ~450 ky is higher during glacial times, but in contrast to the Subantarctic Atlantic (Martínez-García et al., 2014), the relationship between productivity and nutrient consumption has not yet been evaluated in detail. Preliminary data suggest that nutrient consumption increased in response to iron fertilization (A. Martínez-García, unpubl. data).

Figure F11. Lithogenic mass accumulation rates.

Expedition 383 drill cores from the Pacific sector of the Southern Ocean will allow exploration of the glacial–interglacial changes in biological productivity both north and south of the APF in more detail. Published studies, including those from ODP Leg 177, show that during glacial periods of the late Quaternary in the Atlantic sector of the Southern Ocean, biological productivity increased in the Subantarctic Zone and decreased in the Antarctic Zone south of the APF (Diekmann, 2007; Jaccard et al., 2013). The glacial productivity decrease south of the APF is under debate and may have been caused either by stratification of the water column (Sigman et al., 2004; Jaccard et al., 2005) or expansion of Antarctic sea ice coverage (Hillenbrand and Cortese, 2006).

Our drill sites in the CSP (U1539–U1541) will provide robust data to test to what extent processes that are active in the Atlantic sector can be translated to the Pacific sector, thus allowing the construction of a more global picture of the Southern Ocean’s role in nutrient distribution and biogenic export production and their impact on CO₂ variations. It is necessary to solve these issues in the Pacific before the impact on the ocean carbon cycle under different climate boundary conditions can be modeled more realistically.

Moreover, the ESP site (U1543) and Chilean margin sites (U1542 and U1544) will provide a long-term record of variability of the Patagonian Ice Sheet (PIS). Both terrestrial records from Patagonia and records of ice-rafted debris (IRD) in marine sediment cores from the Chilean margin (Core MD07-3128 at Proposed Site CHI-4A) indicate substantial millennial-scale variability of the PIS paralleling dust input changes recorded in Antarctic ice cores. This covariance implies a causal link between ice sheet advances and dust availability (Caniupán et al., 2011; Sugden et al., 2009). Thus, the Chilean margin sites will provide important baseline information for dust variability in the Atlantic sector of the ACC.

Our two hypotheses contain a large number of challenging research targets critical for paleoceanography and paleoclimatology on a global scale. CSP Sites U1539–U1541 represent the first drill cores in the Subantarctic Pacific sector of the Southern Ocean. The deep Pacific ESP site (U1543) is located in a particularly sensitive ocean area where glacial–interglacial changes are strongly amplified. Finally, the Chilean margin sites provide information on changes at shallower water depths together with changes in PIS extent and potential links to dust input to the Atlantic sector of the Southern Ocean.

Site summaries

Site U1539

Background and objectives

Site U1539 (Proposed Site CSP-2B) is located in the CSP at 56°09.0655′S, 115°08.038′W, ~1600 nmi west of the Strait of Magellan at 4070 m water depth (Figure F12). The site sits at the eastern flank of the southernmost EPR, ~220 nmi from the modern seafloor spreading axis, and is underlain by oceanic crust formed between ~10 and 12 Ma (Eagles, 2006). Assuming a constant seafloor half-spreading rate of ~4.5 cm (Pitman and Heirtzler, 1966), the plate tectonic backtrack path of Site U1539 moves the site westward. The extrapolated backward trajectory translates to an early Pliocene position ~100 nmi closer to the crest of the EPR at a water depth shallower by several hundred meters. At a smaller scale, the site is located in a northeast–southwest oriented, ~5 nmi wide trough that parallels the orientation of the EPR. The adjacent ridges rise to ~3000 m water depth northwest of the site and to ~3500 m water depth to the southeast (Figure F12).

Figure F12. Oceanographic and bathymetric setting, Site U1539.

Site U1539 is located on MCS Line AWI-201000013 (Figure F13A) close to the intersection with Line AWI-2010000011 (Gersonde, 2011). The seismic cross-lines indicate ~650 m thick sediments above oceanic basement. Sediments are mostly well stratified with flat-lying reflectors. Low to moderately reflecting layers become stronger deeper than ~100–120 m sediment depth. Sediment echo sound (Parasound) profiles (Gersonde, 2011) reveal excellent penetration (>150 m) with distinct layering (Figure F13B), suggesting a succession of fine-grained soft sediments with varying lithologic composition.

Figure F13. MCS and Parasound profiles, Site U1539.

Site U1539 lies in the pathway of the Subantarctic ACC, ~100 nmi north of the modern average SAF, in a zonal transition zone of the ACC. West of the site, the ACC and associated fronts are strongly steered by the topography of seafloor spreading systems (Udintsev and Eltanin-Tharp Fracture Zone systems), whereas to the east the vast Amundsen Sea basin does not influence the ACC strongly.

SSTs seasonally vary between ~2°C (July–September) and ~6.5°C (January–March). The area is located west of the main AAIW and Mode Water formation regions in the ESP. The 4070 m water depth places Site U1539 within LCDW, close to the boundary with AABW (Figure F14). This hydrographic setting makes it ideal to evaluate past changes in frontal position, associated export production, ACC current speed and position, and aeolian dust and IRD input during the Pleistocene.

Figure F14. Modern salinity and oxygen distribution.

The main objectives at Site U1539 were to

Recover a moderate- to high-resolution Subantarctic Pliocene–Quaternary sediment record close to the SAF;
Characterize the stratigraphy of siliceous and calcareous oozes, allowing for a wide range of paleoceanographic reconstructions;
Reconstruct high-amplitude Subantarctic SSTs and sea ice variations;
Provide a record of lowermost CDW and glacial AABW;
Reconstruct productivity (opal versus carbonate), nutrient distribution, and dust-productivity coupling; and
Recover a potential far-field record of WAIS variability.

Operations

The first core of Expedition 383 was taken from Hole U1539A at 0445 h on 1 June 2019 from 4071.1 meters below sea level (mbsl). We used the advanced piston corer (APC) system to penetrate from the seafloor to 107.6 m core depth below seafloor, Method A (CSF-A) (Cores 1H–12H) and recover 104.75 m of core (97%). The hole was terminated after Core 12H because of rough weather and sea conditions. Shattered liners were recorded on Cores 5H and 11H. Shear pins were sheared prematurely on Cores 1H, 11H, and 12H.

After 27 h of waiting on weather, Hole U1539B was spudded at 0540 h on 3 June. Cores 1H–3H recovered 21.9 m (78% recovery), but Cores 2H and 3H were misfires, so we decided to terminate the hole and wait for seas to further improve before continuing operations.

By 1245 h, heave had fallen to 3.0 m and we resumed coring. Hole U1539C was started with Core 1H returning a full 9.7 m core barrel. This allowed us to determine a seafloor depth of 4070.2 mbsl. Full-length APC coring continued to refusal at 240.1 m CSF-A (Core 26H). Half-length APC (HLAPC) coring was then used to deepen the hole, and coring terminated after Core 32F at 268.1 m CSF-A. A total of 32 cores were taken in this hole using the APC and HLAPC systems. Formation temperature measurements were taken with Cores 4H, 13H, and 16H. Partial strokes were recorded on Cores 24H–27F with high overpull on Cores 17H–26H. A misfire was recorded on Core 31F.

Hole U1539D was spot cored to fill in coring gaps from the previous holes. The hole was spudded at 1405 h on 5 June, drilled ahead without recovery to 47.5 m CSF-A, where coring began, and continued to 198 m CSF-A with seven drilled intervals totaling 68.7 m. A total of 14 APC cores were taken over the 129.3 m cored interval with a recovery of 128.15 m (99%). The hole reached maximum depth at 1500 h on 6 June, and the drill string was recovered, ending the hole and Site U1539. Misfires were recorded on Cores 2H, 3H, 6H, 8H, and 9H and a partial stroke was recorded on Core 18H. No temperature or orientation measurements were taken in this hole. After recovering the drill string, the rig floor was secured for transit at 0130 h. The thrusters were raised, and the sea voyage to Site U1540 began at 0224 h on 7 June.

Principal results

At Site U1539, a 268.1 m thick continuous sequence of Holocene to early Pleistocene sediments was recovered with a bottom-hole age of ~1.3 Ma and a mean sedimentation rate of ~20.6 cm/ky. Hole U1539C recovered the deepest cored sediments at 268.1 m CSF-A. The recovered 247.50 m of core spans one lithologic unit with two subunits consisting of interbedded calcareous and siliceous biogenic ooze (Figure F15). Site U1539 sediments can broadly be divided into two categories: (1) diatom ooze with varying amounts of calcareous and biosiliceous components and (2) biogenic carbonate ooze with varying amounts of biosiliceous and other calcareous components. Siliceous microfossils show relatively good preservation and are abundant throughout the record. Preservation of calcareous microfossils varies from moderate to good.

Figure F15. Hole summaries, Site U1539.

Five primary lithofacies were identified, and they mostly differ in their proportions of calcareous and biosiliceous components (Table T2). Secondary defining attributes include degree of diagenetic alteration, bioturbation, and in the case of Lithofacies 5, microfossil size and sedimentary structure. Lithofacies 1a is composed of distinct centimeter-scale diatom mats in a biogenic ooze matrix. This facies can be distinguished from Lithofacies 1b, which is characterized by sequences of more pronounced and stacked diatom mats with a near absence of diatom ooze matrix. Lithofacies 2 is the dominant lithofacies at Site U1539. It consists of light greenish gray to gray carbonate-bearing or carbonate-rich diatomaceous ooze that exhibits moderate to heavy bioturbation. Lithofacies 3 consists of white to light gray diatom-bearing to diatom-rich nannofossil or calcareous oozes that occur as thin beds generally <3 m thick with moderate to heavy bioturbation. Lithofacies 4 is nearly pure nannofossil ooze. Lithofacies 5 consists of foraminiferal oozes and sands that occur sporadically as centimeter- to decimeter-scale layers characterized by a sharp basal contact and distinct cross-bedding observable in X-ray images.

We divided the Site U1539 stratigraphic sequence into Subunits IA and IB based on the relative abundance of the five defined lithofacies. With the exception of Lithofacies 5 in Subunit IB, all lithofacies are present in both subunits, but the thickness, frequency, and overall occurrence differs between the two. Subunit IA is characterized by the frequent occurrence of ~1.5 m thick beds of calcareous and nannofossil ooze (Lithofacies 3) and limited occurrence of diatom mats (Lithofacies 1b). In contrast, Subunit IB has only a few occurrences of Lithofacies 4 and is largely dominated by beds of thick and continuous diatom mats (Figure F16). One massive homogenous ~10.7 m thick bed of diatom-rich nannofossil ooze that exhibits an erosional basal contact overlain by foraminifer sands (Lithofacies 5) interrupts the normal depositional sequence of Subunit IB and most likely represents a single rapidly deposited bed. Carbonate-bearing and carbonate-rich diatomaceous ooze (Lithofacies 2) is the dominant lithofacies in both subunits and does not provide a useful means to divide the stratigraphy because of its widespread occurrence. The prominent white to light gray calcareous beds associated with Lithofacies 4 drive the high-amplitude variations in color reflectance and spectrophotometry values (red-green-blue color space [RGB] blue and L*). In general, we observed high-amplitude, low-frequency variations above the subunit boundary at 125 m core composite depth below seafloor, Method A (CCSF-A), and more frequent low-amplitude variations below it (Figure F16).

Figure F16. Primary lithostratigraphic variations, Site U1539.

Biostratigraphic age assignments are in agreement with each other and are mainly derived from diatom, radiolarian, and calcareous nannofossil datums. The preliminary shipboard age model derived largely from diatom, radiolarian, and nannofossil datums indicate no major hiatuses, indicating that a continuous sediment sequence from the early Pleistocene through the Holocene was recovered at this site.

We could not observe unambiguous evidence for geomagnetic reversals in any of the holes at Site U1539. This is probably due to low natural remanent magnetization (NRM) intensities that could be related to the high diatom content of the sediments. Therefore, paleomagnetic reversals were not included in the preliminary shipboard age model for Site U1539.

We analyzed samples for headspace gas, interstitial water chemistry, and bulk sediment chemistry at a resolution of three per core for Hole U1539A and three per core from 78.9 to 120 m CSF-A, two per core from 120 to 150 m CSF-A, and one per core from 150 m CSF-A to the bottom of Hole U1539C at 266.41 m CSF-A. Headspace samples were not collected from HLAPC cores. Methane concentration is low at this site overall. It gradually increases downhole, averaging 12.3 ppmv and never exceeding 18.4 ppmv. Ethane and propane remain below detection limit throughout the entire hole. In Hole U1539A, alkalinity increases in the uppermost 10.6 m to ~5 mM, likely because of anaerobic organic matter remineralization. pH values also increase from 7.7 at the surface to ~8 at 2.94 m CSF-A and remain constant with depth. Interstitial water magnesium and lithium concentrations appear conservative throughout the sediment column. Dissolved manganese concentrations increase sharply from below detection limit in the mudline sample to a maximum of 60 µM at 7.57 m CSF-A, indicating anaerobic, reductive dissolution of manganese oxides within the uppermost 3 m of sediment.

The composite carbonate record at Site U1539 (Holes U1539A and U1539C) shows downhole variability ranging from 2.9 to 92.8 wt% and a strong correlation with RGB blue and reflectance L* measurements (Figure F17) and gamma ray attenuation (GRA) bulk density. Total organic carbon (TOC) concentration has a mean value of 0.40 wt%, and values range between 0.18 and 0.75 wt% (n = 86). Total nitrogen (TN) concentrations vary between 0.004 and 0.079 wt% with a U1539 average of 0.039 wt% (n = 90). The mean value of TOC/TN is often >10 from 0.8 m CSF-A in Hole U1539A to 115.2 m CSF-A in Hole U1539C.

Figure F17. Bulk sedimentary carbonate content, RGB blue, and L*, Holes U1539A and U1539C.

Physical properties measurements at Site U1539 included nondestructive whole-round measurements of GRA bulk density, magnetic susceptibility (MS), Whole-Round Multisensor Logger (WRMSL) P-wave velocity, and natural gamma radiation (NGR) on core sections from Holes U1539A–U1539D. Additional physical properties collected include thermal conductivity on whole-round cores from all holes. Downhole changes in physical property characteristics overall are in good agreement with the lithofacies defined based on sedimentologic characteristics (Figure F18). In general, diatom ooze lithofacies (1a and 1b) correspond to relatively higher MS, low bulk density, and high porosity. The lowest densities and highest porosities occur in diatom mats within diatom ooze, which is the recurring dominant Lithofacies 1b. In contrast, intervals that are predominantly calcareous nannofossil ooze (Lithofacies 4 and 5) correspond with decreased NGR, high bulk density, low P-wave velocity, low to moderate MS, and low to moderate porosity. Over most of the record, GRA bulk density shows an anticorrelation with the NGR-derived potassium concentrations and MS. The NGR record also exhibits 10–15 m scale cyclicity, particularly from ~10 to 135 m CSF-A. NGR data also imply that total counts and the derived potassium concentrations can be used as a semiquantitative proxy for the abundance of terrigenous material delivered by dust or sea ice and iceberg transport.

Figure F18. Physical properties, Hole U1539A.

Correlations between holes at Site U1539 were accomplished using the Correlator software (version 3.0). We constructed a splice for the upper portion of the site from 0 to 219.80 m CCSF-A using Holes U1539A–U1539D. However, the splice contains several gaps, as detailed below. Below 219.80 m CCSF-A, only a single hole (U1539C) was drilled; thus, Cores 383-U1539C-22H through 32F were appended to the splice with gaps set between each core to the bottom of the hole. The composite depth scale is anchored to the mudline of Core 383-U1539A-1H, which is assigned as 0 m CCSF-A.

We constructed a preliminary age model based on biostratigraphic results, mainly including diatom and nannofossil age markers. These data suggest that the sedimentary sequence recovered at Site U1539 covers the past ~1.3 My (Figure F19). For a pelagic setting, sedimentation rates are extraordinarily high, reaching on average 17 cm/ky back to 0.9 Ma and even higher values of ~30 cm/ky from ~0.9 to 1.3 Ma. This biostratigraphic age model is generally consistent with preliminary stratigraphic tuning performed onboard based on physical property data such as color measurements (RGB blue) and GRA bulk density.

Figure F19. Preliminary age model, Site U1539.

The combination of nearly continuous recovery, very high sedimentation rates driven by high diatom productivity and potential sediment focusing, clear patterns in physical properties and sediment color, and a rich array of well-preserved diatoms combined with calcareous microfossils will provide unprecedented opportunities for improving our understanding of the dynamics of the ACC and its link to global carbon cycle changes at orbital and suborbital timescales.