Site 1172 contains only 11 m of upper Eocene sediment deposited on the ETP shelf and small hiatuses (Stickley et al., submitted [N1]) caused by current action and nondeposition. As the water deepened from shelf to upper slope depths and current action increased, the sediment changed from diatomaceous mudstone to glauconitic siltstone. Smectite is the dominant clay. Site 1171 contains ~80 m of sediment laid down in a local basin on the southern STR, almost all being diatomaceous mudstone overlain by a few meters of glauconitic quartz-bearing sandstone. Smectite is the dominant clay at this site also. The water was deepening and current activity increasing on this part of the former land bridge and at Site 1170 in the Ninene Basin. Site 1170 contains ~50 m of upper Eocene sediments: marine mudstone overlain by glauconitic diatomaceous siltstone. Illite and smectite alternate as dominant clays; the illite content suggests proximity to an area of active tectonism (Robert, in press).
Site 1168 contains ~130 m of upper Eocene siliciclastic mudstone deposited in a restricted embayment on the broad west Tasmanian shelf in the AAG, which becomes increasingly more open marine upcore. Unlike the situation at other sites, spores are much more abundant than dinocysts, suggesting closer proximity to land during deposition. Also, the dinocyst taxa are largely cosmopolitan with some low-latitude forms, and (unlike those at Site 1172 to the southeast) the assemblages generally lack endemic high-latitude taxa, suggesting relatively warmer conditions (Brinkhuis, Munsterman, et al., this volume). The lowermost sequence consists of sandy mudstone, possibly largely nonmarine. Kaolinite dominates the clay assemblage, suggesting a source region marked by tectonic activity and intense chemical weathering (Robert, in press) or perhaps a change in source rocks. The bulk of the sediments are laminated, organic-rich, pyritic shallow-marine to paralic mudstone, indicating sluggish circulation and poor ventilation. Kaolinite decreases upcore and smectite increases, suggesting reducing tectonic activity but continuing chemical weathering (Robert, in press). Anoxic to dysoxic depositional environments extended up onto the continental shelf. Above this mudstone sequence are shallow-marine sandstones and mudstones containing quartz and sponge spicules derived from nearby beaches and banks (Exon et al., in press b).
In the transitional upper Eocene interval, sedimentation changes uphole from more muddy to more sandy, with increased glauconite and quartz, reflecting an increase in bottom currents and winnowing. At Sites 1170, 1171, and 1172 the upward gradation is from mudstone into glauconitic siltstone. Despite evidence for general winnowing and the presence of hiatuses in the glauconitic siltstone, some levels containing angular quartz indicate episodes marked by little reworking. The glauconitic sediments are strongly bioturbated and were deposited in well-oxygenated bottom waters. Stepwise changes in dinocyst, pollen, and spore assemblages indicate environmental changes and deepening water (Sluijs et al., this volume). Diatoms and benthic foraminifers indicate deepening from shallow to deeper neritic or possibly uppermost bathyal environments.
At Site 1172, dinocyst assemblages continue to be dominated by endemic forms (e.g., Brinkhuis, Sengers, et al., this volume). Of course, Southern Ocean stable isotopic records generally indicate progressive cooling through the middle and late Eocene (Shackleton and Kennett, 1975; Stott et al., 1990; Kennett and Stott, 1990). During the late Eocene at Site 1171 and, especially, Site 1170, cooler continental conditions are indicated by increased illite relative to smectite, suggesting a reduction in continental chemical weathering. Pollen and spore records suggest diverse and cool temperate late Eocene plant communities in the hinterland. Floras were dominated by Nothofagus and podocarps with an understory of ferns, similar to a floral assemblage of similar age in the Weddell Sea sector of Antarctica (Mohr, 1990). Although the late Eocene pollen assemblages indicate cooling, they also show that the Tasmanian part of the Antarctic margin was still relatively warm compared to the distinctly cooler Oligocene. This contrasts with the Prydz Bay margin far to the west, where clear evidence exists for late Eocene glaciation close to sea level (Barron et al., 1991a, 1991b; Cooper and O'Brien, 2004). There is also convincing evidence for early Oligocene growth of a significant ice sheet on at least parts of East Antarctica (Zachos et al., 1996).
Various lines of evidence suggest that Antarctica and the South Tasman Rise separated fully during the Eocene–Oligocene transition (Fig. F7B). All four deep sites contain a fairly continuous record over this interval until the earliest Oligocene, after which there are hiatuses at all sites except Site 1168 until ~30 Ma (Pfuhl and McCave, this volume; Fuller and Touchard, in press). In the early Oligocene, deposition at the Pacific sites changed to open-water pelagic carbonates. This lithologic change reflects the shift from siliciclastic to biogenic sedimentation, from a poorly to a well-oxygenated benthic environment, from tranquil to moderately dynamic environments, and from relatively warm to cool climatic conditions. This paleoenvironmental change in the oceans was the most profound of the entire Cenozoic (Kennett, 1977; Zachos et al., 1993). An early Oligocene cooling and dissolution episode is recorded widely in deep-sea carbonates and is associated with the well-known positive oxygen isotopic shift (Oi-1) at ~33 Ma (e.g., Shackleton and Kennett, 1975; Miller et al., 1991; Zachos et al., 1994).
In the earliest Oligocene, similar open-ocean conditions began to develop on both sides of the former Tasmanian land bridge. We argue that a shallow-water proto-ACC was established at the time of final separation and that the cool countercurrent that had reached Sites 1170–1172 from the southeast no longer did so (Fig. F7B). Currents continued to circulate clockwise in the AAG and westward along the Antarctic coast in the Pacific Ocean. By the late Oligocene, nearly all of the former land bridge south of Tasmania had submerged. The Tasmanian Gateway south of the STR was hundreds of kilometers wide and continuing to widen, and water depths were abyssal. The ACC, flowing from the west and accommodating an ever-increasing circumpolar flow, was effective in all water depths and eroded and dissolved older sediments. The Drake Passage, south of South America, may have opened to deep water in the early Oligocene (Lawver and Gahagan, 1998, 2003) or at the Oligocene/Miocene boundary (Barker and Burrell, 1977). The expansion of the Antarctic cryosphere during the middle and late Cenozoic, and its effect of strengthening thermohaline circulation at deep and intermediate water depths, contributed to very widespread deep ocean erosion and the formation of hiatuses.
Why was there such a change from siliciclastic to carbonate sedimentation at the Eocene/Oligocene boundary rather than early in the late Eocene? A very broad, shallow Australian-Antarctic continental shelf had been supplied with siliciclastic sediments since early in the Cretaceous. Although rifting, subsidence, and compaction had commenced then, sedimentation had kept up, and shallow-marine sediments were deposited rapidly until the end of the middle Eocene. Australia and Antarctica were almost completely separated when fast spreading began in the middle Eocene (~43 Ma), and this could be expected to increase the rate of subsidence. In the Tasmanian region, slower siliciclastic sedimentation continued in deepening but largely shelfal water at Sites 1170, 1171, and 1172 in the late Eocene until the Eocene/Oligocene boundary (~33.5 Ma), some 10 m.y. after fast spreading started. We suggest that the slower sedimentation resulted from current winnowing, bypassing, and probably also falling sediment supply. Subsidence curves (Hill and Exon, in press) suggest faster subsidence at the Eocene/Oligocene boundary in the Tasmanian region, like that which formed the Victoria Land Basin in nearby Antarctica (Cape Roberts Science Team, 2000). Such subsidence would have rapidly reduced the area of potential erosion in the Tasmanian region and drastically reduced sediment supply. However, only in the earliest Oligocene did pelagic carbonate sedimentation rapidly replace the siliciclastic and diatomaceous sedimentation at the eastern sites. The change in the biogenic component of sedimentation, from diatomaceous in the Eocene to calcareous in the Oligocene, must have been related to the changes in oceanography and latitude. At the western Site 1168, the transition began at the same time but continued through the entire Oligocene.
The Eocene/Oligocene change to carbonate sedimentation probably was also related to contemporaneous climatic cooling, which would have greatly reduced rainfall, and thus weathering and erosion, reducing siliciclastic supply. Thereafter, slow deposition of pelagic carbonate completely dominated off southern and eastern Tasmania and was increasingly important west of Tasmania. The sequence of changes in the sediments over the transition is remarkably consistent over the STR and ETP, as determined from our deep cored sequences. Differences in detail are clearly related to individual setting at the time of deposition (such as latitude) and proximity to the ocean and landmasses. Sequences from elsewhere on the Antarctic margin show a similar drastic reduction in siliciclastic sedimentation and increase in biogenic sedimentation during the Eocene–Oligocene transition. The earliest Oligocene is often marked by an increase in biogenic sediments or components in otherwise relatively slowly deposited siliciclastic sediments, including diamictites (Diester-Hass and Zahn, 1996; Kennett and Barker, 1990; Salamy and Zachos, 1999). However, outside the Tasmanian region, the biogenic component is usually biogenic silica (diatoms) rather than biogenic carbonate (calcareous nannofossils and foraminifers). On the shallow (probably neritic) northwest margin of the Weddell Sea, carbonate-free diatom ooze was deposited during the earliest Oligocene, suggesting significant cool-water upwelling (Barker, Kennett, et al., 1988). On the margins in Prydz Bay and the southern Ross Sea, diatoms became an important component in diamictites (Barron et al., 1991a, 1991b; O'Brien, Cooper, Richter, et al., 2001).
Evaluation of the sedimentary sequences cored in the Tasmanian Gateway region (Stickley et al., this volume) suggests opening of the Tasmanian Gateway to cool shallow-water flow occurred during the latest Eocene, with intensifying current flow toward the Eocene/Oligocene boundary. This was followed in the earliest Oligocene by expansion of the Antarctic cryosphere and deepwater interchange between the southern Indian and Pacific Oceans. This interchange heralds the ACC in this part of the Southern Ocean. Although planktonic microfossils in the Leg 189 cores indicate climatic cooling, there is no evidence of glaciation in these sequences. Indeed, the calcareous nannofossil assemblages suggest somewhat warmer conditions at equivalent latitudes elsewhere in the Southern Ocean (Wei and Wise, 1990; Wei and Thierstein, 1991; Wei et al., 1992).
The late Eocene glauconitic siltstones in the sites closest to Antarctica are overlain, with little gradation, by ooze and chalk of early Oligocene age, whereas near western Tasmania there is more gradation upward into the Oligocene. From the earliest Oligocene onward, sedimentation at the eastern Leg 189 sites was completely dominated by deposition of nannofossil ooze. Sedimentation rates of these oozes were faster than those of the glauconitic silts. At the eastern sites, Oligocene rates were slower than those of the lower and middle Eocene siliciclastic sediments, but later rates were comparable. In contrast, at Site 1168 rates were slower across the Eocene/Oligocene boundary but comparable in the upper Eocene siliciclastic and upper Oligocene to lower Miocene marly sequences. Although the age of the base of the carbonates requires better constraint, existing stratigraphic data suggest deposition commenced at ~30 Ma (Stickley et al., submitted [N1]), following the oxygen isotope shift, which is dated at ~33.5 Ma. The isotopic shift represents major cooling and the initial major cryospheric development of East Antarctica (Shackleton and Kennett, 1975; Miller et al., 1991; Wei, 1991; Zachos et al., 1994) and major expansion of the psychrosphere with its deep ocean circulation (Kennett and Shackleton, 1976). On northwest Tasmania there is an alpine glacial tillite, dated palynologically as latest Eocene or earliest Oligocene (Macphail et al., 1993). In summary, the synchronous commencement of biogenic carbonate deposition appears to reflect major tectonic, climatic, and oceanographic changes that affected broad regions in the Southern Ocean near Tasmania. At the eastern Leg 189 sites, these changes created a more dynamic, well-ventilated ocean with increased upwelling and higher surface water biogenic productivity, which increased rates of sedimentation of calcareous nannofossils and diatoms and decreased preservation of organic carbon. Open-ocean planktonic diatoms replaced neritic diatoms, reflecting this deepening and also suggesting initiation of limited coastal upwelling on the STR and Tasmania during the earliest Oligocene (Stickley et al., this volume). Furthermore, associated cooling of the Antarctic and Australian continents apparently decreased weathering rates and transport of siliciclastic sediments to the margins. In addition, subsidence rapidly reduced land areas, which also became more remote from the depocenters on the continental margins, dramatically decreasing the transport of siliciclastic sediment to those depocenters. The environment of deposition was thus transformed from the late Eocene to the earliest Oligocene from siliciclastic to deep-sea carbonate sediments. At the relatively nearshore western Site 1168 there was a long transition, extending from the earliest Oligocene until the early Miocene.