METHODOLOGIES

The bulk of the remanence measurements made during Leg 188 were carried out using a 2G Enterprises (model 760-R) pass-through cryogenic magnetometer equipped with pickup coils that enable measurement of the magnetic signal over an interval of ~8 cm (Shipboard Scientific Party, 2001). Natural remanent magnetization (NRM) was routinely measured before and after alternating-field (AF) demagnetization on all archive-half core sections at 4-cm intervals. Time constraints permitted analysis with only two or three AF demagnetization steps at 10-, 20-, and 30-mT peak values for most of the core sections. The low maximum peak AFs ensured that the archive halves remained useful for shore-based paleomagnetic studies. In a few intervals, the presence of strong magnetic overprints necessitated progressive demagnetization of the archive halves up to 60-80 mT. Measurements at the end of each core section and those within intervals of drilling-related core deformation and in the vicinity of obvious metamorphic and/or igneous pebbles were removed during data processing.

Discrete samples (standard 8-cm³ plastic cubes) were collected from the working halves of the cores at ~1-m intervals and were analyzed to verify the reliability of the whole-core measurements on the archive core halves. If possible, these samples were taken from fine-grained horizons and sampling was adjusted to avoid intervals with drilling-related core deformation and pebbles. Most discrete samples were AF demagnetized at 10, 20, 30, 40, 50, 60, 70, and 80 mT using the in-line demagnetizer installed on the 2G Enterprises pass-through cryogenic magnetometer on the ship. A subset of samples was thermally demagnetized on the ship using a Schonstedt TSD-1 oven. All of the samples subjected to thermal demagnetization were measured at steps of 20°, 100°, 200°, 300°, 330°, 360°, 400°, 500°, 550°, 600°, 650°, and 700°C. The magnetic susceptibility was measured after each heating step to monitor for thermal alteration of magnetic minerals.

The lack of azimuthal orientation for these cores does not pose a problem for determination of paleomagnetic polarity in our magnetostratigraphic studies because the geomagnetic field at the latitude of Site 1165 (64.4°S) has a steep inclination (±76.5°, assuming a geocentric axial dipole field). The paleomagnetic inclinations were determined using the 20- to 30-mT steps from the long-core measurements and using principal component analysis (Kirschvink, 1980) for data from multiple demagnetization steps for discrete samples. The maximum angular deviation (MAD) was calculated to provide an estimate of the precision for each best-fit line. Samples were only included in this study if MAD values were <10°.

Mineral magnetic analyses were conducted on a set of representative discrete samples after they had been subjected to AF demagnetization in order to estimate downcore variations in the composition, concentration, and grain size of magnetic minerals. Low-field magnetic susceptibility (k) was routinely measured for all the discrete samples using a Bartington Instruments MS2 magnetic susceptibility meter. Further analyses were made on a selected subset of discrete samples. These analyses included (1) stepwise acquisition of an isothermal remanent magnetization (IRM) in fields up to 1.3 T; (2) determination of the coercivity of remanence (Bcr) and S-ratio (-IRM [-0.3 T]/IRM [1.3 T]) by progressively applying increasing backfields up to 300 mT after application of a forward-field IRM at 1.3 T; and (3) anhysteretic remanent magnetizations (ARMs) imparted with a 100-mT AF and a 0.05-mT direct current (DC) bias field. For a few samples, we also carried out a stepwise thermal demagnetization of a composite IRM (Lowrie, 1990) at steps of 20°, 100°, 200°, 300°, 330°, 360°, 400°, 500°, 550°, 600°, 650°, and 700°C. Fields of 1.3, 0.5, and 0.12 T were applied along the x-, y-, and z-axes of samples to distinguish between high-, intermediate-, and low-coercivity magnetic phases, respectively. Temperature dependence of magnetic susceptibility was also measured for selected samples from room temperatures up to 700°C, using a furnace-equipped Kappabridge KLY-3 magnetic susceptibility meter (Hrouda, 1994).

Grain-size separations were carried out using standard methods, as described in Allen and Warnke (1991). The >63-µm fraction was dry-sieved into the 63- to 150-µm, 150-µm to 2-mm, and >2-mm fractions. For the purposes of this report, the 63- to 150-µm and 150-µm to 2-mm fractions were combined in the accompanying graphs. Analyses of samples in the depth range of 0-15 mbsf were performed at Stanford University. Analyses of samples in the interval 15-50 mbsf were carried out at California State University, Hayward.

For the HiRISC section in Hole 1165B, we determined the calcium carbonate content for closely spaced (~10 cm) samples provided by HiRISC. We used the vacuum gasometric technique of Jones and Kaiteris (1983) to determine weight percent calcium carbonate content for all HiRISC samples in Hole 1165B from 0 to 54 mbsf (Fig. F2) (see Damuth and Balsam, their table T1, this volume, for sample intervals and carbonate values). With this technique, a small (~0.25 g) powdered sample is digested in concentrated phosphoric acid under vacuum and the pressure generated by the release of CO₂ is recorded on a vacuum gauge. Weight percent CaCO₃ is calculated by relating the pressure increase in the sample to the pressure increase in reagent carbonate after correcting for temperature and pressure. The Jones and Kaiteris (1983) technique has an accuracy of about ±1%. As with most techniques used to determine carbonate, the Jones and Kaiteris (1983) technique combines both polymorphs of CaCO₃, calcite, and aragonite, as well as carbonate from biogenic and nonbiogenic sources. In most marine settings dolomite is not a significant contributor to the carbonate component. The Pleistocene and uppermost Pliocene sediments in this section (0-10 mbsf) show wide fluctuations in carbonate content, ranging from 0 to 37 wt%. Below 10 mbsf the carbonate content is generally zero and rarely rises to a few weight percent. Most of the values in this interval are <0.4 wt% and are below the accuracy of the instrument.

NUV/VIS/NIR spectral data were measured from all samples taken from Sites 1165 and 1167 with our laboratory-grade PerkinElmer Lambda 6 spectrophotometer at the University of Texas at Arlington (UTA) (Damuth and Balsam, this volume). Sample preparation followed the procedures described by Balsam and Deaton (1991). Reflectance spectrophotometers such as ours are designed to scan different wavelengths of light reflected from a sample's surface and record the intensity of that reflected light relative to a white standard (e.g., barium sulfate) used to set the 100% reflectance level. The PerkinElmer Lambda 6 spectrophotometer uses a reflectance sphere, which is a diffuse reflectance attachment that allows total reflectance measurements to be made from the near infrared (250 nm) through the visible into the near ultraviolet (850 nm). The Lambda 6 contains two light sources, a tungsten lamp for 350-850 nm and a deuterium lamp for 250-350 nm, a moving grating (to separate light into different wavelengths), and a photomultiplier tube (to measure the intensity of light reflected from the sample surface). Data from the spectrophotometer are recorded directly on a floppy disk at 1-nm intervals from 250 to 850 nm, the analytical range of the Lambda 6 in the reflectance mode. Samples were analyzed using a slit width of 2 nm at a scan rate of 600 nm/min. Details and data tables are presented in the companion paper by Damuth and Balsam (this volume; their table T2). In the accompanying figures, we display curves showing brightness downhole as percent brightness, which is simply brightness values rescaled to 100%.

Left-coiling foraminifers of the planktonic species Neogloboquadrina pachyderma (s.) dominate surface sediment foraminifer assemblages in polar regions (Bé, 1977) and are used widely in paleoclimatic and paleoceanographic studies (e.g., Mackensen et al., 1989; Charles and Fairbanks, 1990; Hodell, 1993). At Site 1165, N. pachyderma (s.) composes the large majority of preserved foraminifers from the Pliocene-Pleistocene section and is present in sufficient numbers for stable isotopic analysis in the upper 16.36 mbsf. Foraminifers are generally well preserved, although there is evidence of dissolution (broken tests, holes in tests, etc.) in some samples that may affect their stable isotopic values.

Methods

Sediment samples, taken at 20-cm intervals, were soaked in a dilute Calgon solution, wet sieved, and separated into several fractions (>2 mm, 150 µm-2 mm, 63-150 µm, and <63 µm). All fractions were dried in an oven at 50°C. N. pachyderma (s.) specimens were picked from the >150-µm fraction. Samples in the upper 16.36 mbsf contained sufficient foraminifers for stable isotopic measurements. All remaining samples (to 50 mbsf) were either barren of foraminifers or, in rare cases, had poorly preserved foraminifers that were deemed unsuitable for analysis. All measurements were made at the Stanford University Stable Isotope Laboratory. Analytical methods and precision are described in Theissen et al. (this volume).

X-ray diffraction (XRD) was performed on oriented clay samples. The sediment samples were first decomposed by ultrasonic vibration then centrifuged for 1 min to suspend clay minerals (<2 µm). Samples were then concentrated by centrifugation for 15 min. The separated clay minerals were treated by standard methods: air-dried, glycolated, and heated following the technique described by Hardy and Tucker (1988). Clays were fed to the XRD (Siemens D 5000) at angles from 2° to 32°2 (0.02° 2/s) immediately after the treatments. The four principal clay mineral groups have basal spacings at 7 Å (kaolinite and chlorite), 10 Å (illite), 12-15 Å (smectite), and 14 Å (chlorite), and mixed-layer minerals give intermediate or higher values. Ethylene glycol treatment was used to separate smectite from chlorite. Kaolinite collapses when heated. In this study chlorite (004) was identified at 3.54 Å and kaolinite (002) at 3.58 Å, and this proportion was used to calculate quantities of kaolinite and chlorite from the joint peak at 7 Å. MacDiff software version 4.25 (Petschick, 2001) was used to quantify clay minerals, which were then used to calculate percentages using weighting factors (Biscaye, 1964). Since no internal standards were used, the exact accuracy is not known; however, the quantitative analyses justify interpretations of fluctuations which are around ± 2%.

Materials and Methods

The section studied is that contained in intervals 188-1165B-1H-1, 0.17-0.21 cm, through 2H-6, 0.66-0.69 cm (0.17-14.96 mbsf). The lower depth limit is that below which foraminifers are virtually absent. The following should be considered an amplification of analyses contained in Quilty (this volume).

This study is designed to detect changes in distribution of foraminifers and to relate these to other parameters measured as part of the HiRISC project. Although some quantification is attempted, this should not be interpreted as a productivity signal, as it probably is more influenced by carbonate dissolution patterns than by productivity.

The site appears to lie within the Antarctic Divergence Zone where currents are westerly flowing, but the oceanography probably has varied considerably within the timescale under study and it is possible that, at different times, the area has been both north and south of the divergence.

Samples studied are those in the 150-µm to 2-mm mesh range, prepared by K. Theissen at Stanford University, California. Sample numbers included on Table T1 are those supplied by Ocean Drilling Program (ODP).

As many samples are very highly dominated by N. pachyderma (Ehrenberg), all samples were split to smaller, more manageable, volumes. All foraminifers were manually separated from the smaller subsamples, glued down on separate slides, counted, and identified. Data were compiled in the accompanying table (Table T1), and basic analyses were carried out.

Factors affecting the accuracy of counts include human error and, especially, the effects of dissolution and breakage, which influence the content of fragmented specimens, leading to occasional double counting or even noncounting when fragments pass through the finer sieve size.

Whereas the object of the study concerned the foraminifers, records were kept of other aspects of the residues, such as dominant components and presence or absence of trace amounts of sponge or echinoid remains or of glauconite or volcanic glass. These data also are included on the accompanying table.

Taxonomy

Taxonomy of foraminifers follows, where possible, that of Jones (1994). Osangulariella umbonifera is thus used in place of Nuttallides umbonifera. The generic name Cibicides is preferred to Cibicidoides where relevant. A few species are left in open nomenclature. Unless readily identifiable, species of Lagena and Fissurina are not differentiated.