Depth ranges of calcareous nannofossil datums from ODP Holes 113-689B and 113-690B

Magnetostratigraphic studies of Paleogene sediments piston-cored on Maud Rise, Weddell Sea (ODP Sites 689 and 690), are a cornerstone of Southern Ocean Paleogene and Neogene chronostratigraphy. However, parts of previous magnetostratigraphic interpretations have been called into question, and recent reinvestigation of the upper Paleocene-middle Eocene portion of Site 690 suggested that the records might be contaminated by spurious magnetizations, which raises doubts about the reliability of these important records. We undertook a high-resolution magnetostratigraphic study of Eocene-Oligocene u-channel samples from ODP Holes 689B, 689D, 690B, and 690C in order to address these concerns. A pervasive overprint appears to be present below the middle Eocene, which compromises magnetobiostratigraphic interpretations for the upper Cretaceous and lower Paleogene. Nevertheless, our new results provide a robust record of geomagnetic field behavior from 38.5 to 25 Ma and confirm the reliability of these sediments for calibration of biostratigraphic datum events during a crucial phase of earth history when major Antarctic ice sheets developed. Also, comparison of magnetozone thicknesses in multiple holes at the same site indicates that ~1.2-1.8 m of the stratigraphic record is missing at each core break, which corresponds to time breaks of 120-360 k.y. Lack of a continuous record within a single hole renders useless spectral analyses for investigating long geomagnetic and paleoclimatic time series. This observation reinforces the need for coring of multiple offset holes to obtain continuous paleoceanographic records. Sedimentary hiatuses have been identified only at the deeper of the two investigated sites (Site 690), which could mark a local response to the onset of the Antarctic Circumpolar Current.

Benthic foraminifera and stable isotope composition in Paleocene-Eocene sediments

In the late Paleocene to early Eocene, deep sea benthic foraminifera suffered their only global extinction of the last 75 million years and diversity decreased worldwide by 30-50% in a few thousand years. At Maud Rise (Weddell Sea, Antarctica; Sites 689 and 690, palaeodepths 1100 m and 1900 m) and Walvis Ridge (Southeastern Atlantic, Sites 525 and 527, palaeodepths 1600 m and 3400 m) post-extinction faunas were low-diversity and high-dominance, but the dominant species differed by geographical location. At Maud Rise, post-extinction faunas were dominated by small, biserial and triserial species, while the large, thick-walled, long-lived deep sea species Nuttallides truempyi was absent. At Walvis Ridge, by contrast, they were dominated by long-lived species such as N. truempyi, with common to abundant small abyssaminid species. The faunal dominance patterns at the two locations thus suggest different post-extinction seafloor environments: increased flux of organic matter and possibly decreased oxygen levels at Maud Rise, decreased flux at Walvis Ridge. The species-richness remained very low for about 50 000 years, then gradually increased. The extinction was synchronous with a large, negative, short-term excursion of carbon and oxygen isotopes in planktonic and benthic foraminifera and bulk carbonate. The isotope excursions reached peak negative values in a few thousand years and values returned to pre-excursion levels in about 50 000 years. The carbon isotope excursion was about -2 per mil for benthic foraminifera at Walvis Ridge and Maud Rise, and about -4 per mil for planktonic foraminifera at Maud Rise. At the latter sites vertical gradients thus decreased, possibly at least partially as a result of upwelling. The oxygen isotope excursion was about -1.5 per mil for benthic foraminifera at Walvis Ridge and Maud Rise, -1 per mil for planktonic foraminifera at Maud Rise. The rapid oxygen isotope excursion at a time when polar ice-sheets were absent or insignificant can be explained by an increase in temperature by 4-6°C of high latitude surface waters and deep waters world wide. The deep ocean temperature increase could have been caused by warming of surface waters at high latitudes and continued formation of the deep waters at these locations, or by a switch from dominant formation of deep waters at high latitudes to formation at lower latitudes. Benthic foraminiferal post-extinction biogeographical patterns favour the latter explanation. The short-term carbon isotope excursion occurred in deep and surface waters, and in soil concretions and mammal teeth in the continental record. It is associated with increased CaC03-dissolution over a wide depth range in the oceans, suggesting that a rapid transfer of isotopically light carbon from lithosphere or biosphere into the ocean-atmosphere system may have been involved. The rapidity of the initiation of the excursion (a few thousand years) and its short duration (50 000 years) suggest that such a transfer was probably not caused by changes in the ratio of organic carbon to carbonate deposition or erosion. Transfer of carbon from the terrestrial biosphere was probably not the cause, because it would require a much larger biosphere destruction than at the end of the Cretaceous, in conflict with the fossil record. It is difficult to explain the large shift by rapid emission into the atmosphere of volcanogenic CO2, although huge subaerial plateau basalt eruptions occurred at the time in the northern Atlantic. Probably a complex combination of processes and feedback was involved, including volcanogenic emission of CO2, changing circulation patterns, changing productivity in the oceans and possibly on land, and changes in the relative size of the oceanic and atmospheric carbon reservoirs.

Foraminiferal Mg/Ca and Mn/Ca ratios across the Eocene-Oligocene transition

Constraining the magnitude of high-latitude temperature change across the Eocene-Oligocene transition (EOT) is essential for quantifying the magnitude of Antarctic ice-sheet expansion and understanding regional climate response to this event. To this end, we constructed high-resolution stable oxygen isotope (d18O) and magnesium/calcium (Mg/Ca) records from planktic and benthic foraminifera at four Ocean Drilling Program (ODP) sites in the Southern Ocean. Planktic foraminiferal Mg/Ca records from the Kerguelen Plateau (ODP Sites 738, 744, and 748) show a consistent pattern of temperature change, indicating 2-3 °C cooling in direct conjunction with the first step of a two-step increase in benthic and planktic foraminiferal d18O values across the EOT. In contrast, benthic Mg/Ca records from Maud Rise (ODP Site 689) and the Kerguelen Plateau (ODP Site 748) do not exhibit significant temperature change. The contrasting temperature histories derived from the planktic and benthic Mg/Ca records are not reconcilable, since vertical d18O gradients remained nearly constant at all sites between 35.0 and 32.5 Ma. Based on the coherency of the planktic Mg/Ca records from the Kerguelen Plateau sites and complications with benthic Mg/Ca paleothermometry at low temperatures, the planktic Mg/Ca records are deemed the most reliable measure of Southern Ocean temperature change. We therefore interpret a uniform cooling of 2-3 °C in both deep surface (thermocline) waters and intermediate deep waters of the Southern Ocean across the EOT. Cooling of Southern Ocean surface waters across the EOT was likely propagated to the deep ocean, since deep waters were primarily sourced on the Antarctic margin throughout this time interval. Removal of the temperature component from the observed foraminiferal d18O shift indicates that seawater d18O values increased by 0.6 ± 0.15 per mil across the EOT interval, corresponding to an increase in global ice volume to a level equivalent with 60-130% modern East Antarctic ice sheet volume.

Paleogene planktonic foraminifera from the southern Kerguelen Plateau

An essentially complete Paleogene record was recovered on the Central and Southern Kerguelen plateaus (55°-59°S) in a calcareous biofacies. Recovery deteriorated in the middle Eocene and down to the upper Paleocene because of the presence of interbedded cherts and chalks. The stratigraphic distribution of about 70 taxa of planktonic foraminifers recovered at Sites 747-749 is reported in this paper. Faunas exhibited fairly high diversity (approximately 20-25 species) in the early Eocene, followed by a gradual reduction in diversity in the middle Eocene. A brief incursion of tropical keeled morozovellids occurred near the Paleocene/Eocene boundary, similar to that recorded on the Maud Rise (ODP Sites 689 and 690). The high-latitude Paleogene zonal scheme developed for ODP Leg 113 sites has been adopted (with minor modifications) for the lower Eocene-Oligocene part of the Kerguelen Plateau record. A representative Oligocene (polarity chronozones 7-13) and late Eocene-late middle Eocene (questionably polarity chronozones 16-18) magnetostratigraphic record has allowed the calibration of several biostratigraphic datum levels to the standard Global Polarity Time Scale (GPTS) and established their essential synchrony between low and high latitudes.