RESULTS

Site 1094 (53°10.81´S, 5°7.82´E; 2807-m water depth) is located in a small sedimentary basin north of Bouvet Øya south of the present day position of the Antarctic Polar Front (Fig. F1). The interval from 95.0 to 121.19 meters composite depth (mcd) was sampled continuously (nominally at 5-cm intervals) at Site 1094 for stable isotopic analyses and subsequent lithic counts. Beyond 121.19 mcd there is not a composite spliced section available. The abundance of N. pachyderma (sinistral) allowed the planktonic record to be acquired at a 5-cm (~357 yr) resolution, whereas the relative scarcity of benthic foraminifers rendered it impossible to construct a benthic stable isotope record from this site with only 20-cm³-sized samples. The ¹⁸O record from Site 1094 is plotted vs. depth in Figure F2. The abundance of planktonic foraminifers remained fairly constant throughout the studied interval, with only minor gaps that do not influence the continuity of the isotope record. Of the 522 samples analyzed, only 28 were barren of planktonic foraminifers (Table T1).

The Matuyama/Brunhes (780 ka) and top Jaramillo (990 ka) paleomagnetic boundaries initially defined the age control for the mid-Pleistocene interval at Site 1094 (Channell and Stoner, in press.). In this study, we have refined this chronology by tuning the ¹⁸O signal at Site 1094 to the ice volume simulation model of Imbrie and Imbrie (1980). We chose the latter as our tuning target because it has proved to be a powerful target for the late Pleistocene (Bassinot et al., 1994; Martinson et al., 1987) as well as for the early and mid-Pleistocene (Channell and Kleiven, 2000; Shackleton et al., 1990). Our procedure was to match the ¹⁸O record from Site 1094 to the target using linear interpolation between tie points (Table T2). This procedure was fairly straightforward because precession-related oscillations are well expressed in the ¹⁸O record of Site 1094. When assigning the tie points, we gave the glacial-interglacial transitions higher priority than the centers of glacial or interglacial intervals. We assumed constant sedimentation rates between tie points, resulting in a change in sedimentation rate more or less abruptly at these points. This assumption may be realistic when the tie points correspond to glacial-interglacial transitions, when environmental changes might be expected to effect the sedimentation rates. Moreover, assigning tie points at high-amplitude transitions is probably more accurate than the designation of any particular peak or trough in intervals displaying considerable high-frequency, low-amplitude variability. In Figure F3, we have plotted our new Site 1094 ¹⁸O record vs. the Ice Volume Model on their common age scale. Based on our correlation, we conclude that our ¹⁸O record spans the time interval from 740 to 993 ka.

In Figure F4, we have plotted our new Site 1094 planktonic ¹⁸O record vs. the mid Pleistocene benthic ¹⁸O record from North Atlantic Site 983 (Channell and Kleiven, 2000; Kleiven et al., in press) as comparison. Oxygen isotopic Stages 18 through 26 can easily be identified in the Site 1094 ¹⁸O record by matching to the benthic ¹⁸O record from Site 983. Orbital variability in ¹⁸O is clearly present in the Site 1094 record, indicating that Milankovitch-forced ice volume changes are controlling the large-scale fluctuations. Superimposed on this glacial-interglacial ¹⁸O signal is large suborbital-scale variability. It is interesting to note that many of the same high-resolution features are common and well defined in the records from both the South and North Atlantic. Significant millennial-scale variability, with pacing indistinguishable from that of the last glacial cycle, characterize all observed climate states during the mid-Pleistocene interval from these sites, suggesting that suborbital variability has been a fundamental part of the climate system in both regions. Further paleoceanographic interpretation of the oxygen isotope results and lithic proxy records from Site 1094 are presented in Kleiven and Jansen, (submitted [N1]).

Site 1091 (47°5.68´S, 5°55.12´E; 4363-m water depth) is located in the Polar Front Zone (PFZ) on the western flank of the Meteor Rise, ~2° north of the present-day Polar Front (Fig. F1). Site 1091 lies well north of the present day Antarctic PFZ, but during the course of glacial cycles, the lithology of the site alternates between calcareous (interglacial) and diatomaceous (glacial) ooze. The sedimentation rates vary between ~10 to ~30 cm/k.y. At Site 1091 there is a continuous spliced section down to 234 mcd (~1.7 Ma). For our study, we sampled the interval from 93.03 to 158.01 mcd continuously at 5-cm intervals at Site 1091 for stable isotopic analyses and subsequent lithic counts.

As for Site 1094, benthic foraminifers at Site 1091 were to absent to scarce for analyses and we have focused on obtaining a continuous isotope signal based on the abundance of the polar planktonic foraminifers N. pachyderma (sinistral) selected from the >150- to 250-µm size fraction. The ¹⁸O record from Site 1091 is plotted vs. depth in Figure F5. The abundance of planktonic foraminifers has remained fairly constant back to 119 mcd, which is as far as we have analyzed at this site to date. There are, however, some intervals barren of foraminifers from 98.6 to 99.4, 114.1 to 115.8, and 117.2 to 117.7 mcd, which have led to minor gaps in our record (Table T3).

We have not yet constructed a final age model for our Site 1091 interval, as work on these sections is still in progress. However, to make a correlation between our results at Sites 1094 and 1091, we have tied the ¹⁸O record from Site 1091 from 93.03 to 119 mcd to that of Site 1094 using the AnalySeries program (Paillard et al., 1996). In Figure F6, we have plotted the ¹⁸O records from Site 1094 and 1091 on their common age scale together with the benthic ¹⁸O record from Site 983. Based on this correlation, we conclude that our record from site 1091 spans the interval from 750 to 950 ka (marine isotope Stages 18-25).

Site 1093 (49°58.60´S, 5°51.92´E) lies just beneath the present Antarctic PFZ, at the northern edge of the modern diatomaceous ooze belt (Fig. F1). At Site 1093 there is a continuous spliced section down to 255 mcd. For our study, we sampled the interval from 190 to 255 mcd continuously at 10-cm intervals for stable isotopic analyses and subsequent lithic counts. We have so far analyzed the first 12-mcd section of our interval and have found every sample to be barren of planktonic or benthic foraminifers. Based on shipboard data (Shipboard Scientific Party, 1999) we know that our interval begins in a cool interval at the MIS 17/18 transition and remains cold until the MIS 18/19 transition. In recent work by Hodell et al. (Chap. 9, this volume), they find that Site 1093 is barren of foraminifers in every cold (glacial) interval for the last 600 k.y. We believe that this pattern continues into the mid-Pleistocene and hope to be able to obtain stable isotope measurements as we move into the warmer interglacial intervals of MIS 19 and MIS 21.