Multisensor track (MST) and color reflectance data (650-750 nm) collected from Holes 1089A-1089C were used to determine depth offsets in the composite section. Magnetic susceptibility, color reflectance, and gamma-ray attenuation (GRA) bulk density measurements were the primary parameters used for the core-to-core correlation at Site 1089. The data used to construct the composite section and determine core overlaps are presented on a composite depth scale in Figures F14, F15, and F16, respectively. The depth offsets that comprise the composite section for Holes 1089A-1089D are given in Table T4 (also in ASCII format in the TABLES directory).
GRA bulk density and magnetic susceptibility data were collected at 2-cm intervals on all cores recovered at Site 1089. Color reflectance data were collected at 4- to 6-cm intervals on all cores from Holes 1089A and 1089B, Cores 177-1089C-1H through 9H, and Cores 177-1089D-2H, 4H, 6H, and 9H (see "Physical Properties" for details about these MST and color reflectance data).
The composite data show that the cores from Site 1089 provide a continuous overlap to a depth of 94 mcd (base of Core 177-1089C-10H). Between 94 and 156 mcd (base of Core 177-1089C-10H to Section 177-1089B-16H-5), the sedimentary section is marked by numerous highly disturbed intervals consisting of flow-in, slumping, and microfaulting. The severity and amount of disturbance in this interval makes it impossible to correlate cores between holes with any degree of certainty without further postcruise analysis. Below 154 mcd, the cores were placed into a composite depth frame, but most of the gaps between cores could not be covered by cores from other holes.
Generally, the reference core for the composite section (in this case Core 177-1089A-1H) will have a 0-mcd offset. At this site, the reference core and Cores 177-1089C-1H and 177-1089D-1H have negative offsets. The negative offset results from using the first stratigraphically intact portion of the reference core as the mudline reference. The uppermost 24 cm of Core 177-1089A-1H consists of washed-in sediment that is probably not in place. As a result, this core was shifted 24 cm up to remove this disturbed interval.
Stretching and compression of sedimentary features in aligned cores indicate distortion of the cored sequence. Because much of the distortion occurred within individual cores on depth scales of <9 m, it was not possible to accurately align every feature in the MST and color reflectance records by simply adding a constant to the mbsf core depth. Within-core scale changes will require postcruise processing to align smaller sedimentary features. Only after allowing variable adjustments of peaks within each core can an accurate estimate of core gaps be made.
Following construction of the composite depth section for Site 1089, a single spliced record was assembled for the aligned cores over the upper 94 mcd primarily by using cores from Holes 1089B and 1089C. The cores were aligned relative to the mcd scale so that tie points between adjacent holes are present at exactly the same depths (mcd). Intervals having significant disturbance or distortion were avoided when possible. The Site 1089 splice (Table T5, also in ASCII format in the TABLES directory) can be used as a sampling guide to recover a single sedimentary sequence between 0 and 94 mcd. A spliced record was not constructed below 94 mcd because of the highly disturbed nature of the sediments between 94 and 156 mcd and the lack of core overlap among holes below 156 mcd. Spliced records of magnetic susceptibility, color reflectance, and GRA bulk density for the upper 94 mcd are presented in Figure F17.
Sediments recovered from Site 1089 provide a nearly continuous late Pliocene-Pleistocene record. Calcareous nannofossil abundances vary from abundant to barren, and moderate to poor preservation is observed in the assemblages. Dissolution is a common feature in placoliths and, in several cases, makes identification at the species level difficult. At least two samples per section were examined to establish an accurate age assignment of Holes 1089A and 1089B. Significant numbers of reworked nannofossils are recorded at certain intervals. This reworked material is mainly Pliocene in age, but lower Miocene and upper Eocene-Oligocene species have also been identified. Moreover, it is difficult to accurately establish the reworking of Pleistocene species, especially through intervals in which the biozonal events are last occurrences. To improve the Martini (1971) and Okada and Bukry (1980) standard zonations, we have considered some additional events according to Pujos (1988), Raffi et al. (1993), and Wei (1993) (see "Explanatory Notes" chapter).
Several events are recognized in the Pleistocene interval (Table T6, also in ASCII format in the TABLES directory; Fig. F18), providing a good biostratigraphic resolution and biochronology, when preservation permits. The first occurrence (FO) of Emiliania huxleyi at the base of Subzone NN21a is identified in Holes 1089A and 1089B between 39.56 and 40.56 mcd. However, the above-mentioned interval is affected by dissolution, and in several samples it is difficult to identify the presence of E. huxleyi (Table T6). The last occurrence (LO) of Pseudoemiliania lacunosa is present between 67.54 and 69.37 mcd, defining the base of Zone NN20 (Figs. F18, F19). Comparison of the FO of E. huxleyi and the LO of P. lacunosa with the interpreted reflectance curve (Fig. F18) allows us to infer that these events occurred during marine isotopic stages (MISs) 8 and 12, respectively3. The top and base of the small Gephyrocapsa acme are not clearly recognized, although a dominance of "very small Gephyrocapsa" complex is observed at the top of Zone NN19 (Table T6). The LO of Reticulofenestra asanoi is present at different depths in Holes 1089A and 1089B, suggesting that this event cannot be used reliably at Site 1089 (Table T6). The varying range of R. asanoi may have been caused by reworking, which is very intense in several intervals as noted previously. Alternatively, this species may have an extended range in the region. A quantitative study, as well as the correlation of this event at other sites, will provide a more reliable evaluation of these results.
The reentrance of medium Gephyrocapsa (4-5.5 µm) is present between 160.36 and 161.26 mcd. The FO of R. asanoi is present between 171.56 and 172.05 mcd. The LO of large Gephyrocapsa (>5.5 µm) is observed from 176.65 to 178.01 mcd, whereas the FO of this species is well defined between 195.56 and 197.83 mcd.
The FO of medium Gephyrocapsa (4-5.5 µm) is present between 223.90 and 224.80 mcd (Table T6) and marks the Pliocene/Pleistocene boundary. The LO of Calcidiscus macintyrei (>11 µm) is defined between 220.30 and 220.09 mcd.
The LO of Discoaster brouweri is identified between 236.70 and 238.53 mcd (Zone NN19). The age assigned to the lower part of Hole 1089B, below 222.5 mbsf, is younger than 2.54 Ma on the basis of the presence of D. brouweri (Zone NN18) (Table T6; Fig. F19). In Hole 1089B, late Pliocene species (Zones NN17 and NN18) are identified in well-preserved assemblages in some samples from resedimented clasts in Core 177-1089B-27H, below 257.11 mcd. Unfortunately, Samples 1089B-27H-CC, 28H-CC, and 29H-CC are barren, and we cannot establish whether the identified upper Pliocene assemblages are reworked (Table T6).
The planktic foraminifer abundance in the four holes drilled at Site 1089 is relatively high for the Pleistocene, but it decreases significantly downhole (Table T7, also in ASCII format in the TABLES directory). The preservation of the planktic foraminifer fauna is generally moderate to poor. The degree of fragmentation is high in all studied samples. There is a distinct trend of decreasing preservation associated with the downhole decrease in abundance of planktic foraminifers. The sediments recovered from Holes 1089A-1089D span the Pliocene/Pleistocene boundary; however, as a result of the low abundance of Globorotalia truncatulinoides at this site, it is not possible to distinguish between the G. truncatulinoides and Globorotalia inflata Zones (Jenkins and Srinivasan, 1986). The planktic foraminifer assemblages are dominated by Globigerina bulloides, G. quinqueloba, Globorotalia puncticuloides, Globigerinita glutinata, G. inflata, and Neogloboquadrina pachyderma (sinistral). In addition, a few species (e.g., G. truncatulinoides, Globigerina woodi, Globigerinella calida, Globigerinita uvula, Globorotalia crassaformis, G. scitula, Neogloboquadrina humerosa, N. pachyderma [dextral], and Orbulina universa) make smaller contributions to the planktic foraminifer fauna at Site 1089. In the deepest hole (1089B), the studied core-catcher (CC) samples from Samples 177-1089B-19H-CC, 16-21 cm (175.33 mbsf), through 29H-CC, 62-67 cm (264.91 mbsf), are barren or contain only traces (a few broken specimens) of planktic foraminifers (Table T7). Thus, it is possible that the faunal changes observed at Site 1089 are linked to changes in carbonate preservation.
Benthic foraminifers were present in all the CC samples above 247.88 mcd at Site 1089, generally comprising between 5% and 10% of the total foraminifer fauna from the >63-µm fraction studied. Quantitative estimates of relative species abundance were made from Hole 1089A, with counts of as many as 300 specimens per sample (Table T8, also in ASCII format in the TABLES directory). Absolute foraminifer abundances, while variable, exhibit a clear trend toward higher values between ~100 and 150 mbsf, reaching a maximum of 95 specimens/cm3 in Sample 177-1089A-16H-CC, 9-14 cm. Changes in benthic foraminifer abundance appear to reflect the general pattern of Pleistocene glacial-interglacial cyclicity at Site 1089, from generally high abundances during the glacial intervals to lower abundances during the interglacial intervals. This could be explained by the observation that sedimentation rates at Site 1089 are higher during interglacials than glacials, possibly resulting in some dilution of benthic foraminifers during interglacials. However, the low stratigraphic resolution of the CC samples studied and possible problems of flow-in associated with Hole 1089A make it difficult to determine the exact relationships that exist between benthic foraminifer abundances and the variability of the Pleistocene glacial-interglacial cycles. Preservation is generally good above 190 mcd, although there may be some variability associated with carbonate abundances through the well-developed glacial-interglacial cycles. Samples 177-1089B-26H-CC through 29H-CC are barren. Species richness is variable, with a maximum of 43 taxa recorded in Sample 177-1089A-17H-CC, 0-8 cm, and a minimum of 21 taxa recorded in Samples 177-1089A-2H-CC, 16-21 cm, and 8H-CC, 9-14 cm. Some of this variability can be accounted for by sample size as counts of 250-300 specimens are required before no correlation between the number of species and number of specimens is achieved (see "Explanatory Notes" chapter).
The benthic foraminifer assemblages of the Pleistocene sequence at Site 1089 (Table T8) are dominated by relatively few species, notably Alabaminella weddellensis, Epistominella exigua, Oridorsalis umbonatus, and Pullenia quinqueloba. These taxa, notably E. exigua and A. weddellensis, are known to respond to seasonal changes in phytodetritus supply within the flocculent layer of the ocean floor (Gooday, 1988; Mackensen et al., 1990) and may, therefore, provide a valuable tool in assessing surface-water productivity and hydrographic conditions at a very high temporal resolution through the late Pleistocene glacial-interglacial cycles. Preliminary results indicate that the glacial intervals are characterized by higher abundances of these phytodetritus species. These data, combined with benthic stable isotopic records, should provide an excellent opportunity to explore inferred changes in the nutrient status of CDW throughout the late Pleistocene at Site 1089.
Biostratigraphic differentiation of Site 1089 is limited to the LO of Stilostomella lepidula (e.g., Thomas, 1987) in Sample 177-1089A-10H-CC, 0-5 cm, which supports a late Pleistocene age of the sediment sequence above 103.16 mcd. The latter is in good agreement with the LO of Actinocyclus ingens (0.64 Ma) at a mean depth of 90.05 mcd in Holes 1089B and 1089C (see "Diatoms").
Holes 1089A-1089D recovered a continuous late Pliocene to Pleistocene record at high sedimentation rates ranging between 84 and 128 m/m.y. (Fig. F20; Table T9, also in ASCII format in the TABLES directory). This allows the application and testing of the revised biostratigraphic zonation recently proposed by Gersonde and Bárcena (1998) for the northern area of the Southern Ocean. The latter provides biostratigraphic age control at a resolution of ~0.2 m.y. Improved temporal resolution can be achieved particularly for the last 700 k.y. by the combination of color reflectance logs, which reflect glacial-interglacial variability in carbonate content, and the abundance pattern of selected diatom taxa, which allow the identification of specific MISs. In addition to the CC samples obtained from all holes, we have examined smear slides from Hole 1089B (Tables T10, T11; both also in ASCII format in the TABLES directory). All diatom stratigraphic information from the four holes is combined and converted to the mcd scale. This is crucial because significant disturbance of the sedimentary record by slumping is present in all holes between 100 and 150 mbsf (see "Lithostratigraphy"), resulting in strong offsets of isochronous intervals in the different holes (see "Composite Depths"). However, the slumping did not result in disturbance of the stratigraphic sequence.
Diatoms are common to abundant in most samples and are generally moderately preserved. Diatoms are most abundant in the upper 120 mcd, representing the last 800 k.y. (Fig. F21). In addition to planktic marine taxa, a few benthic marine taxa and Paralia sulcata, a taxon that preferentially dwells in near-shore marine environments, are also recorded. These diatoms most likely originated from the shallow waters off the African continent. The consistent presence of near-shore and shallow-water diatoms throughout the entire sedimentary record indicates a near-continuous input of sediment particles derived from African near-shore environments by deep- and/or bottom-water advection. More detailed, shore-based studies may determine if this input is controlled by glacial-interglacial sea-level variations. Lateral advection of reworked sediments by bottom currents is also documented by sparse diatoms stratigraphically older than the sedimentary record.
The biostratigraphic analyses of the diatom assemblages indicate that the base of the Thalassiosira lentiginosa Subzone c, defined by the LO of Hemidiscus karstenii, can be placed around 30 mcd (Fig. F21). The presence of this datum within MIS 6.5 is corroborated by the interpretation of the color reflectance record (Fig. F18). The presence of H. karstenii in the T. lentiginosa Subzone b, displaying abundance peaks in MISs 11, 9, and 7 (Burckle, 1982) (Fig. F18), permits recognition of these stages as reflected in the color reflectance record. MIS 11.3 is marked by the first abundant reentry of H. karstenii that has its FO in the late Miocene (see Gersonde and Bárcena, 1998). Assemblages from MIS 11 are typically characterized by abundant occurrences of Azpeitia tabularis (Fig. F21), a pattern that has also been observed in other cores from the Atlantic sector of the Southern Ocean (R. Gersonde, unpubl. data). The LO of Actinocyclus ingens, which marks the top of the A. ingens Subzone c within the lower and warmer portion of MIS 16, is recognized at 90 mcd. This event, which occurred at 0.64 Ma, supports the identification of MISs 17, 16, and 15 in the color reflectance record (Fig. F18). Within the A. ingens Subzone c, the LOs of Thalassiosira elliptipora and T. fasciculata are encountered at 120 and 110 mcd, respectively, at or close to the Brunhes/Matuyama boundary (Fig. F21), as reported by Gersonde and Bárcena (1998).
The base of the A. ingens Subzone c, marked by the first abundant appearance datum of T. elliptipora and correlated to the lower boundary of the Jaramillo Subchron (C1r.1n) in the Matuyama Chron at ~1.1 Ma, is present at ~162 mcd (Table T9). The base of the A. ingens Subzone b is marked by the LO of Fragilariopsis barronii. This datum was found around 200 mcd. However, reworking and rare occurrences of this taxon might indicate that the depth assignments of this datum are not reliable at Site 1089. The nominate taxon of the Proboscia barboi Zone, which nearly coincides with the Olduvai Subchron (C2n), was found below ~233 mcd (Table T9). The low number of samples studied in this sedimentary interval hinders the exact identification of the lower and upper boundaries of this zone. This datum is corroborated, however, by the recognition of the Olduvai Subchron between 230 and 242 mcd (Table T9). Below 247 mcd, we found Fragilariopsis matuyamae, a taxon that is characteristic of the Thalassiosira kolbei/F. matuyamae Zone, in the lower reversed portion of Subchron C2r of the Matuyama Chron. T. kolbei, however, was only encountered in trace to rare amounts.
The correlation of diatom ranges with the magnetostratigraphic record at Site 1089 indicates that the biostratigraphic zonation proposed by Gersonde and Bárcena (1998) is well adapted to sediments of the northern area of the Southern Ocean. In addition, an acme of A. ingens reported by Gersonde and Bárcena (1998) from the middle to lower portion of the Matuyama Chron is present at Site 1089. The biostratigraphic study of Site 1089 also allows preliminary definition of the ranges of warm-water taxa such as Fragilariopsis reinholdii and F. doliolus for further biostratigraphic use in the northern realm of the Southern Ocean. The LO of F. reinholdii at Site 1089 is apparently synchronous with its LO reported from low latitudes (Barron, 1992). It occurred around 0.65 Ma, close to or synchronous with the LO of A. ingens. The FO of F. doliolus, at ~1.8 Ma in the equatorial Pacific (Barron, 1992), was encountered in trace numbers near the Olduvai Subchron at Site 1089. However, the first consistent occurrences were found only in the late A. ingens Subzone b around 1.1 Ma (Fig. F21). Fragilariopsis kerguelensis, a taxon which dominates surface sediment assemblages in the Southern Ocean (Zielinski and Gersonde, 1997), displays high numbers during the last ~800 k.y. It is, thus, prominent in sediments deposited during the late Pleistocene when climatic variability was marked by 100-k.y. cycles. The first significant occurrence of this taxon is found at Site 1089 in the P. barboi Zone (Fig. F21). Alveolus marinus, a warm-water dwelling taxon, is encountered throughout most of the sedimentary record at rare or trace amounts. This taxon does, however, display prominent abundance peaks in mid- and late Pleistocene warm stages such as MISs 15, 11, 9, and 5 (Fig. F21).
Silicoflagellates of the genera Distephanus and Dictyocha are only found in small numbers.
Radiolarian biostratigraphy at Site 1089 is based on examination of 31 CC samples (Table T12, also in ASCII format in the TABLES directory). All samples yielded well-preserved, highly diverse radiolarian assemblages. Co-occurrences of warm-water species (e.g., Dictyocoryne truncatum and Spongaster tetras) and Antarctic species (e.g., Antarctissa spp. and Saccospyris antarctica) are recognized in many samples, which is probably a result of the proximity of Site 1089 to the STF.
All samples contain small numbers of Cycladophora pliocenica and Saturnalis circularis, which are characteristic below the Psi Zone. Further indication of reworking comes from the presence of the Cretaceous Dictyomitrella sp. and Miocene Actinomma golownini in Samples 177-1089C-19H-CC and 177-1089D-1H-CC, respectively. This impedes biostratigraphic zonation based on radiolarian occurrences at Site 1089. An exception is the LO of Stylatractus universus, which marks the base of the Omega Zone at 0.46 Ma that was found in Samples 177-1089A-7H-CC, 177-1089B-7H-CC, 177-1089C-6H-CC, and 177-1089D-7H-CC, consistently around 60 mcd (Tables T11, T12).
Only one specimen of Cycladophora davisiana, which is abundant to common in most samples, was present on a strewn slide in Sample 177-1089B-29H-CC. This may suggest that this horizon (279.65 mcd) is very close to the first appearance datum of C. davisiana at 2.61 Ma.
Archive halves of APC cores recovered at Site 1089 were measured using the shipboard pass-through magnetometer. Measurements were made at 5-cm intervals. Sections obviously affected by drilling disturbance were not measured, although intervals affected by slumping from the base of the Brunhes Chron to the top of the Jaramillo Subchron were measured (see "Lithostratigraphy"). All cores from Holes 1089A and 1089B, Cores 177-1089C-1H through 8H and 11H through 21H, and Cores 177-1089D-11H through 12H were measured after alternating-field (AF) demagnetization at peak fields of 0 (natural remanent magnetization [NRM]), 5, 10, 15, and 20 mT. Cores 177-1089D-1H through 10H and Cores 177-1089C-9H through 10H were measured after AF demagnetization at peak fields of 0, 10, and 20 mT.
NRM intensities are ~1 × 10-2 A/m at the top of each hole, decrease through the upper 50 m to ~1 × 10-3 A/m, and then remain fairly uniform throughout the cored interval. After AF demagnetization at peak fields of 20 mT, intensities generally decreased to ~3.5 × 10-4 A/m. NRM inclinations are typically steep down as a result of a magnetic overprint that has been attributed to the drill string. The drill-string magnetization is generally removed by peak demagnetization fields of 10 mT. At demagnetization fields above this value, the resulting characteristic inclination values are consistent with those expected for the site location (60°) (Fig. F22). Declinations are consistent within core sections and within individual cores. The Tensor orientation tool was used below the third core at all holes except Hole 1089B, where weather conditions were deemed unsuitable for its deployment.
Unfortunately, only the record from Hole 1089B approaches completeness (Fig. F22). At Hole 1089A, recovery was particularly poor and the quality of the recovered core was compromised by the failure of core liners. The Brunhes/Matuyama boundary is present in the interval from 105 to 114 mbsf in Hole 1089B (Fig. F22; Table T13); it is not well preserved, however, because of slumping that affects an interval from the earliest Brunhes Chron to the top of the Jaramillo Subchron. Sharper polarity transitions are found for the lower boundary of what is interpreted as the Jaramillo Subchron and at the upper and lower boundaries of the Olduvai Subchron (Table T13).
A 264.9-m-thick sedimentary section spanning the interval from the Holocene through the late Pliocene was recovered at Site 1089. The basal age was estimated to be ~2 Ma. Holes 1089A-1089D were cored with the APC to 216.30, 264.9, 194.4, and 118.0 mbsf, respectively. A continuous sedimentary section could be documented to 94 mcd. Between 94 and 156 mcd (base of Core 177-1089C-10H to Section 177-1089B-16H-5), the sedimentary section is marked by numerous highly disturbed intervals that made it impossible to correlate between holes. Below 156 mcd, the cores were placed into a composite, albeit incomplete, depth framework.
Age assignment and calculation of sedimentation rates for Site 1089 are based on calcareous nannofossil and diatom biostratigraphies, as well as measurements of geomagnetic polarity changes (Table T9). Because of reworking and difficulties in the identification of stratigraphically useful species, only one stratigraphic data point was established using radiolarian biostratigraphy. Despite apparent disturbance of some cored intervals, reworking of stratigraphically older species, and bottom-water advection of microfossils, the different stratigraphic marker groups and polarity reversal stratigraphy that were obtained allow the establishment of a consistent age assignment for Site 1089 (Fig. F19). In the upper 100 m of the recovered section, MISs were identified based on fluctuations of color reflectance combined with stratigraphically useful diatom and calcareous nannofossil occurrences. Biostratigraphic events resulted in the establishment of age control points separated by ~100 k.y. or less between 0 and 0.5 Ma (Fig. F20). In the older section of Site 1089, the age-depth points are separated by intervals spanning ~200 k.y. Of the 25 datum levels, 23 were chosen as control points (Table T9) to establish an age-depth model and estimate sedimentation rates. Depth uncertainty of the calcareous nannofossil datums corresponds to 0.7 m, or one-half core section. Depth uncertainties in the assignment of geomagnetic polarity changes and diatom datums are indicated by vertical error bars in Figure F20.
The resulting age-depth relationship shows a rather continuous sedimentation (Fig. F20). Calculated sedimentation rates average ~128 m/m.y. in the upper 94 mcd (~0.7 Ma). Below the disturbed section, the sedimentation rates are slightly lower (between 84 and 11 m/m.y.). The interval of relatively high sedimentation corresponds to the disturbed section between 95 and 154 mcd. Within the upper 60 m of the cored interval, corresponding to approximately the last 400 k.y., there is a tendency for increased sedimentation during interglacial intervals and relatively lower sedimentation during glacial intervals (Table T9). On the basis of the age model derived from the control points, the Pliocene/Pleistocene boundary can be placed at ~229 mcd.
3 Note in proof: Preliminary oxygen isotopic data suggest that variations in color reflectance do not strictly reflect glacial-interglacial cycles; therefore, the MIS assignments should be viewed with caution at Site 1098.