Radiolarian-based paleotemperatures of ODP Site 177-1089 in the Atlantic Sector of the Southern Ocean

Two cores, Site 1089 (ODP Leg 177) and PS2821-1, recovered from the same location (40°56'S; 9°54'E) at the Subtropical Front (STF) in the Atlantic Sector of the Southern Ocean, provide a high-resolution climatic record, with an average temporal resolution of less than 600 yr. A multi-proxy approach was used to produce an age model for Core PS2821-1, and to correlate the two cores. Both cores document the last climatic cycle, from Marine Isotopic Stage 6 (MIS 6, ca. 160 kyr BP, ka) to present. Summer sea-surface temperatures (SSSTs) have been estimated, with a standard error of ca. +/-1.16°C, for the down core record by using Q-mode factor analysis (Imbrie and Kipp method). The paleotemperatures show a 7°C warming at Termination II (last interglacial, transition from MIS 6 to MIS 5). This transition from glacial to interglacial paleotemperatures (with maximum temperatures ca. 3°C warmer than present at the core location) occurs earlier than the corresponding shift in delta18O values for benthic foraminifera from the same core; this suggests a lead of Southern Ocean paleotemperature changes compared to the global ice-volume changes, as indicated by the benthic isotopic record. The climatic evolution of the record continues with a progressive temperature deterioration towards MIS 2. High-frequency, millennial-scale climatic instability has been documented for MIS 3 and part of MIS 4, with sudden temperature variations of almost the same magnitude as those observed at the transitions between glacial and interglacial times. These changes occur during the same time interval as the Dansgaard-Oeschger cycles recognized in the delta18Oice record of the GRIP and GISP ice cores from Greenland, and seem to be connected to rapid changes in the STF position in relation to the core location. Sudden cooling episodes ('Younger Dryas (YD)-type' and 'Antarctic Cold Reversal (ACR)-type' of events) have been recognized for both Termination I (ACR-I and YD-I events) and II (ACR-II and YD-II events), and imply that our core is located in an optimal position in order to record events triggered by phenomena occurring in both hemispheres. Spectral analysis of our SSST record displays strong analogies, particularly for high, sub-orbital frequencies, to equivalent records from Vostok (Antarctica) and from the Subtropical North Atlantic ocean. This implies that the climatic variability of widely separated areas (the Antarctic continent, the Subtropical North Atlantic, and the Subantarctic South Atlantic) can be strongly coupled and co-varying at millennial time scales (a few to 10-ka periods), and eventually induced by the same triggering mechanisms. Climatic variability has also been documented for supposedly warm and stable interglacial intervals (MIS 1 and 5), with several cold events which can be correlated to other Southern Ocean and North Atlantic sediment records.

Neogene radiolarian datum levels in the equatorial Indian and Pacific Oceans

Fifty radiolarian events of early Pleistocene and Neogene age were identified in an E-W transect of equatorial DSDP sites, extending from the Gulf of Panama to the western Pacific and eastern Indian Oceans. Our objective was to document the degree of synchroneity or time-transgressiveness of stratigraphically-useful datum levels from this geologic time interval. We restricted our study to low latitudes within which morphological variations of individual taxa are minimal, the total assemblage diversity remains high, and stratigraphic continuity is well-documented by an independent set of criteria. Each of the five sites chosen (503, 573, 289/586, 214) was calibrated to an "absolute" time scale, using a multiple of planktonic foraminiferal, nannofossil, and diatom datum levels which have been independently correlated to the paleomagnetic polarity time scale in piston core material. With these correlations we have assigned "absolute" ages to each radiolarian event, with a precision of 0.1-0.2 m.y. and an accuracy of 0.2-0.4 m.y. On this basis we have classified each of the events as either: (a) synchronous (range of ages <0.4 m.y.); (b) time-transgressive (i.e., range of ages >1.0 m.y.); and (c) not resolvable (range of ages 0.4-1.0 m.y.). Our results show that, among the synchronous datum levels, a large majority (15 out of 19) are last occurrences. Among those events which are clearly time-transgressive, most are first appearances (10 out of 13). In many instances taxa appear to evolve first in the Indian Ocean, and subsequently in the western and eastern Pacific Ocean. This pattern is particularly unexpected in view of the strong east-to-west zonal flow in equatorial latitudes. Three of the time-transgressive events have been used to define zonal boundaries: the first appearances of Spongaster pentas, Diartus hughesi, and D. petterssoni. Our results suggest that biostratigraphic non-synchroneity may be substantial (i.e., greater than 1 m.y.) within a given latitudinal zone; one would expect this effect to be even more pronounced across oceanographic and climatic gradients. We anticipate that the extent of diachroneity may be comparable for diatom, foraminiferal, and nannofossil datum levels as well. If this proves true, global "time scales" may need to be re-formulated on the basis of a smaller number of demonstrably synchronous events.

(Table T1) Ice rafted debris characteristics of ODP Site 177-1092

We have carried out a multiphase analysis of samples from ODP Site 177-1092, Meteor Rise, subantarctic South Atlantic. Samples were analyzed for ice-rafted debris (IRD [see Table T1]) and stable isotopes from benthic foraminifera [see Murphy et al., 2002, doi:10.1016/S0031-0182(01)00495-3]. Both analyses were performed on the same samples. Additional work was performed to identify the paleomagnetic stratigraphy. The analyzed samples range in age from about 2.6(?) Ma to 4.6 Ma, a time span that saw considerable global warmth, but witnessed overall global refrigeration and the transition to truly bipolar glaciations. IRD arrived frequently during the Early and early Late Pliocene, but only as 'background rafting' (occasional grains per sample). The first identifiable IRD above background rafting is associated with marine isotope stage (MIS) KM4 (~3.18 Ma). Successive IRD peaks become larger, the same pattern as noted at nearby Site 114-704. A very large peak near the top of the record, approximately 2.8 Ma, is considered to represent a hiatus. Peaks below 51.3 meters composite depth (mcd) coincide with positive excursions of the oxygen isotopic record, and with negative excursions of the carbon isotopic curve, a pattern also noted at Site 114-704. However, the reasonably large IRD peak at 51 mcd (tentatively identified with MIS G11) coincides with a positive excursion on the carbon isotopic curve and negative excursion on the oxygen isotopic curve. This relationship suggests a northern hemisphere interglacial, rising sea level, destabilization of the Antarctic margin, and delivery of Antarctic icebergs to the Southern Ocean. Such a mechanism has recently been suggested by Kanfoush et al. (2000, doi:10.1126/science.288.5472.1815) for latest Pleistocene stadial/interstadial oscillations. Here we suggest that such a mechanism may have been in place on glacial/interglacial time scales as early as the Late Pliocene.

(Tables 1,2) Viability of Poa trivialis seeds transported to Antarctica from different locations

Antarctic ecosystems are at risk from the introduction of invasive species. The first step in the process of invasion is the transportation of alien species to Antarctic in a viable state. However, the effect of long-distance human-mediated dispersal, over different time-scales, on propagule viability is not well known. We assessed the viability of Poa trivialis seeds transported to Antarctica from the UK, South Africa and Australia by ship or by ship and aircraft. Following transportation to the Antarctic Treaty area, no reduction in seed viability was found, despite journey times lasting up to 284 days and seeds experiencing temperatures as low as -1.5°C. This work confirms that human-mediated transport may overcome the dispersal barrier for some propagules, and highlights the need for effective pre-departure biosecurity measures.

(Appendix) Relative abundance of planktonic foraminifera in ODP Site 138-846 late Quaternary sediments

High-resolution records of d18O and relative abundances of planktonic foraminifers were generated for ODP Leg 138 Site 846 for the past 800 k.y., with an average sampling interval of 3.6 k.y. The time scale was constructed by correlating the benthic d18O record to the SPECMAP and ODP Site 677 d18O time scales using the mapping function technique of Martinson et al. (1981). Our observations show that variations in the foraminiferal assemblages, although influenced by dissolution, are interpretable in terms of changing characteristics of upper ocean waters. Carbonate dissolution as indicated by fragmentation of planktonic foraminifers shows concentrated variance that is coherent with d18O at the 100 and 41 k.y. orbital periods. At these periods, maximum dissolution occurs during interglacial extremes. This finding differs from previous studies that have indicated that in this region percent carbonate minimum lags global ice volume minimum. N. dutertrei and dextral N. pachyderma dominate the assemblages, but do not show consistent relationships relative to glacial-interglacial cycles. However, less abundant species G. ruber, G. menardii, G. glutinata and G. sacculifer show positive and G. bulloides negative correlation with the d18O record. Q-mode factor analysis of the Site 846 assemblages and comparison with modern assemblages suggest the following. Prior to and during interglacials, the area was considerably warmer and more subtropical than at present; during glacials, the area was colder than at present with greater upwelling and advection off the eastern boundary, and possibly a stronger Peru Current; the equatorial ôcool tongueö was also possibly stronger.

Data compilation of ESOP

The work in this sub-project of ESOP focuses on the advective and convective transforma-tion of water masses in the Greenland Sea and its neighbouring areas. It includes observational work on the sub-mesoscale and analysis of hydrographic data up to the gyre-scale. Observations of active convective plumes were made with a towed chain equipped with up to 80 CTD sensors, giving a horizontal and vertical resolution of the hydrographic fields of a few metres. The observed scales of the penetrative convective plumes compare well with those given by theory. On the mesoscale the structure of homogeneous eddies formed as a result of deep convection was observed and the associated mixing and renewal of the intermediate layers quantified. The relative importance and efficiency of thermal and haline penetrative convection in relation to the surface boundary conditions (heat and salt fluxes and ice cover) and the ambient stratification are studied using the multi year time series of hydro-graphic data in the central Greenland Sea. The modification of the water column of the Greenland Sea gyre through advection from and mixing with water at its rim is assessed on longer time scales. The relative contributions are quantified using modern water mass analysis methods based on inverse techniques. Likewise the convective renewal and the spreading of the Arctic Intermediate Water from its formation area is quantified. The aim is to budget the heat and salt content of the water column, in particular of the low salinity surface layer, and to relate its seasonal and interannual variability to the lateral fluxes and the fluxes at the air-sea-ice interface. This will allow to estimate residence times for the different layers of the Greenland Sea gyre, a quantity important for the description of the Polar Ocean carbon cycle.

Lithogenic and biogenic daily and annual particle fluxes on the Lomonosov Ridge of trap LOMO-2

Investigations of lithogenic and biogenic particle fluxes using long-term sediment traps are still very rare in the northern high latitudes and restricted to the arctic marginal seas and sub-arctic regions. Here, for the first time, data on the variability of fluxes of lithogenic matter, carbonate, opal, and organic carbon as well as biomarker composition from the central Arctic Ocean are presented for a one-year period. The study has been carried out on material obtained from a long-term mooring system equipped with two multi-sampling-traps (150 and 1550 m water depth) and deployed on the southern Lomonosov Ridge close to the Laptev Sea continental margin from September 1995 to August 1996. In addition, data from surface-sediments were included in the study to get more information about the flux and sedimentation of organic carbon in this area. Annual fluxes of lithogenic matter, carbonate, opal, and particulate organic carbon are 3.9 g/m**2/y, 0.8 g/m**2/y, 2.6 g/m**2/y, 1.5 g/m**2/y, respectively, at the shallow trap and 11.3 g/m**2/y, 0.5 g/m**2/y, 2.9 g/m**2/y, 1.05 g/m**2/y, respectively, at the deep trap. Both the shallow as well as the deep trap show significant differences in vertical flux values over the year. Higher values were found from mid-July to end of October (total flux of 75-130 mg/m**2/d in the shallow trap and 40-225 mg/m**2/d in the deep trap, respectively). During all other months, fluxes were fairly low in both traps (most total flux values <10 mg/m**2/d1). The interval of increased fluxes can be separated into (1) a mid-July/August maximum caused by increased primary production as documented in high abundances of marine biomarkers and diatoms, and (2) a September/October (absolute) maximum caused by increased influence of Lena river discharge indicated by maximum lithogenic flux and high portions of terrigenous/fluvial biomarkers in both traps. Here, total fluxes in the deep trap were significantly higher than in the shallow trap, suggesting a lateral sediment flux at greater depth. The lithogenic flux data also support the importance of sediment input from the Laptev Sea for the sediment accumulation on the Lomonosov Ridge on geological time scales, as indicated in sedimentary records from this region.

Mineralogy of ODP Hole 117-722B (Appendix)

Bulk mineralogy of the terrigenous fraction of 99 samples from ODP Site 722 on the Owen Ridge, western Arabian Sea, has been determined by x-ray diffraction, using an internal standard method. The sampling interval, approximately 4.3 k.y., provides a detailed mineralogic record for the past 500 k.y. Previous studies have identified important modern continental sediment sources and the mineral assemblages presently derived from each. These studies have also demonstrated that most of this material is supplied by southwest and northwest winds during the summer monsoon. A variety of marine and terrestrial records and general circulation model (GCM) simulations have indicated the importance of monsoonal circulation during the Pleistocene and Holocene and have demonstrated increased aridity during glacial times and increased humidity during inter glacials. The mineralogic data generated here were used to investigate variations in source area weathering conditions during these environmental changes. Terrigenous minerals present include smectite, illite, palygorskite, kaolinite, chlorite, quartz, plagioclase feldspar, and dolomite. This mineralogy is consistent with the compositions of source areas presently supplying sediment to the Arabian Sea. An R-mode factor analysis has identified four mineral assemblages present throughout the past 500 k.y.: quartz/chlorite/dolomite (Factor 1), kaolinite/plagioclase/illite (Factor 2), smectite (Factor 3), and palygorskite/dolomite (Factor 4). Chlorite, illite, and palygorskite are extremely susceptible to chemical weathering, and a spectral comparison of these factors with the eolian mass accumulation rate (MAR) record from Hole 722B (an index of dust source area aridity) indicates that Factors 1, 2, and 4 are directly related to changes in aridity. Because of these characteristics, Factors 1,2, and 4 are interpreted to originate from arid source regions. Factor 3 is interpreted to record more humid source conditions. Time-series of scores for the four factors are dominated by short-term (10-100 k.y.) variability, and do not correlate well to glacial/interglacial fluctuations in the time domain. These characteristics suggest that local climatic shifts were complex, and that equilibrium weathering assemblages did not develop immediately after climatic change. Spectral analysis of factor scores identifies peaks at or near the primary Milankovitch frequencies for all factors. Factor 1 (quartz/chlorite/dolomite), Factor 2 (kaolinite/plagioclase/illite), and Factor 4 (illite/palygorskite) are coherent and in phase with the MAR record over the 23, 41, and 100 k.y. bands, respectively. The reasons for coherency at single Milankovitch frequencies are not known, but may include differences in the susceptibilities of minerals to varying time scales of weathering and/or preferential development of suitable continental source environments by climatic changes at the various Milankovitch frequencies.