The presence of a DWBC off the eastern end of the Chatham Rise was first captured in the classic SCORPIO hydrographic transect of Warren (1973). He identified a major northward flow with a volume transport of up to 20 Sv in water depths greater than ~2000 m. Using more recent hydrographic data, together with seismic profiles, bottom sediments, and photographs, Carter and McCave (1994) and McCave and Carter (1997) showed that the DWBC swings around the eastern end of Chatham Rise and travels northwest to the Kermadec Ridge where it alters course to the north. En route, the current has helped shape an extensive, fine-grained drift on the northern flank of the Rise. Termed the North Chatham drift by Carter and McCave (1994), this deposit appears on seismic profiles as a succession of acoustically laminated Pliocene-Pleistocene sediments above a more massive seismic unit that was interpreted to be Miocene in age. The base of this unit and the base of the proposed hole were considered to be an unconformity at the top of the Oligocene marked by seismic reflector X of Carter and McCave (1994). Profiles from a 3.5-kHz system reveal that much of the drift surface is covered by irregular, nonmigratory sediment waves, which were probably formed under the influence of the DWBC (Fig. F3). Thus, the sedimentary sequence at Site 1123 potentially offers a record of DWBC inflow to the Southwest Pacific under varying paleoclimatic conditions that influenced sediment input and possibly flow intensity.
Cores from Site 1123 recovered a succession of clay-rich nannofossil ooze, chalk, and limestone. The sedimentary succession is divided into four basic lithologic units that are identified on the basis of changes in the calcareous biogenic and noncarbonate components, along with variations in bedding and color. The division of the lithologic units is supported by estimates of core composition from smear slides, together with shipboard measurements of calcium carbonate, physical properties, light reflectance, and bulk mineralogy using X-ray diffraction (XRD). The resulting lithologic logs, together with physical properties and biostratigraphic controls, are summarized in Figures F4 and F5.
Except in a few cases, calcium carbonate concentration is always >50% (the criterion for ooze designation), and on occasion reaches ~80%. The strength of the correlation between light reflectance at 550 nm and percentage calcium carbonate concentration (Fig. F6) suggests that reflectance is an accurate proxy for carbonate, and, thus, can provide detailed information on the carbonate contents (Mix et al., 1995a).
Unit I extends from the present seafloor to 256.59 mbsf and represents an interglacial/glacial cyclic sedimentation pattern that can be divided into Subunits IA and IB on the basis of bedding style and lithology.
Subunit IA corresponds to the Pliocene to Pleistocene drift sequence and extends from 0 to ~182 mbsf. Cores show Subunit IA to be a succession of alternating light greenish gray (5GY 7/1) to greenish gray (5GY 6/1) and white (5Y 8/1) clayey nannofossil oozes (i.e., the dominant biogenic component is nannofossils, with clay mostly as a subordinate constituent, although still occasionally rivaling nannofossil abundance). Beds are distinguished by color variations with layers typically between 1 and 1.5 m thick. Contacts are generally bioturbated and gradational. Accessory components include foraminifers, diatoms, sponge spicules, pyrite (found as smears or frequently as aureoles around and in-filling burrows), and trace amounts of mica. Compositional differences between beds are minimal, but smear-slide data suggest the greenish gray layers have a higher abundance of biogenic silica (within the "Present" category for both diatoms and sponge spicules) and possibly more mica and quartz as well. Beds tend to have faint color banding (particularly pale green laminae) and indistinct mottling associated with pervasive bioturbation. Identified ichnofauna include Zoophycus, Chondrites, Planolites, Thalassinoides, and Terebellina. Of note is the occasional presence of large (>5 cm long) pyritized Teichichnus burrows (e.g., Fig. F7) and centimeter-scale unidentified burrows filled mainly with foraminifers.
Numerous tephra layers are present in Subunit IA (Table T3; Fig. F8) and range in thickness from <1 cm to >20 cm. They typically have sharp bases, normal grading, and bioturbated upper contacts with the overlying ooze. Reworked tephra often occurs as burrow fill (mainly Thalassinoides) below a bed's basal contact (Fig. F9). Above ~90 m the tephra is typically pinkish gray (5YR 6/2), whereas, at deeper levels, the tephra layers are darkened by an increased presence of diagenetic pyrite. In the upper few cores of each hole a number of tephra-"blobs" and lenses are present that are unlikely to be in situ and may reflect drilling disturbance. Tephra layers commonly display dark green laminae at or close to their bases. These bands, of probable diagenetic origin, were used to determine orientation of the tephra layers, some of which were overturned by drilling.
Core disturbance is minimal through Subunit IA, with only a small amount of flow-in present in cores from Hole 1123A. The switch from APC to XCB coring and the coincident inception of drilling biscuits occur at around 155 mbsf (Core 181-1123B-18X), suggesting that the sediment is becoming lithified at that level through the formation of carbonate cement.
Subunit IB encompasses a similar succession of alternating interglacial/glacial couplets as seen in Subunit IA. The sequence is of late Miocene age and extends from ~182 to 256 mbsf. It carries many of the lithologic features of Subunit IA except that IB is sufficiently indurated to be classified as chalk. The sequence, therefore, comprises mainly white (5Y 8/1) clayey nannofossil chalk interbedded with greenish gray (5GY 6/1) clayey nannofossil chalk.
As in Subunit IA, layers are typically between 1 and 1.5 m thick, and contacts between beds are bioturbated. Smear slides of samples from the white and greenish gray-colored chalk showed that both sediments have a similar composition compared to the oozes in Subunit IA. Bioturbation is again abundant, with an ichnofaunal assemblage (Chondrites, Planolites, Thalassinoides, Zoophycus, and Terebellina) similar to that in Subunit IA. The harder sediments of Subunit IB permit better preservation of individual burrow structures than in Subunit 1A (e.g., Fig. F10), and, in some cases, these are highlighted by pyrite stains.
A noticeable difference between Subunits IA and IB is the infrequent occurrence of tephra layers in the latter. The chalk sequence of Subunit IB contains only two tephra layers, whereas at least 33 are present in the ooze of Subunit IA (Fig. F8). The younger late Miocene tephra have abundant pyrite toward their sharp bases, normal grading and bioturbated upper contacts with the overlying chalk. However, the firm sediment has been disrupted by pervasive drilling biscuits. Therefore, potential loss of the more poorly consolidated tephra layers (grain size of the tephra is in the silt to sand range) by washing out during drilling cannot be discounted.
The sedimentary sequence exhibits a fairly uniform lithology between ~257 and ~451 mbsf. This uniformity is exemplified by the color recorded in hand specimen, which does not alter from light greenish gray (5BG 7/1). However, light reflectance data over this subunit show cyclicity too subtle to be recognized with the naked eye, and calibration of the spectrophotometer to proxy carbonate suggests that the percentage carbonate varies rhythmically between ~50% and 60% (Fig. F11). Nevertheless, moderate to abundant bioturbation has ensured extensive but subtle mottling of this subunit, associated with trace fossils such as Chondrites, Planolites, and Skolithos, together with abundant and conspicuous Zoophycus burrows. Beds have occasional very faint centimeter-scale color banding in pale green and pink. In smear-slide observations, these bands appear identical to the major lithology, except that they may contain slightly higher clay contents and possibly diatoms and silicoflagellates (concentrations in the "Present" category). Like the overlying unit, Unit II lithology is a clayey nannofossil chalk with a similar accessory assemblage. The noncarbonate mineralogy is dominated by clay and pyrite, with trace amounts of mica also present. The pyrite occurs as smears and occasionally within burrow fills.
At least nine tephra beds occur in Unit II. The layers are typically 1-5 cm thick with a maximum of 7 cm, have sharp basal contacts, bioturbated tops, and exhibit pyritization toward their bases. One tephra layer (Section 181-1123B-29X-7) contains fine planar laminae toward its base. Some tephra beds are disrupted by the ubiquitous formation of drilling biscuits.
To fill the single coring break in the otherwise continuous record from Hole 1123B, Core 181-1123C-17X was recovered from 230-239.6 mbsf (within Subunit IB) during the wash-down phase of Hole 1123C. Rather than the containing the glacial/interglacial couplets of Subunit IB, as might be expected, the core contained a light greenish gray (5BG 7/1) clayey nannofossil chalk, with heavy bioturbation and common pyrite smears reminiscent of Unit II. The core was damaged during retrieval and required some reconstruction on the drill floor. Thus, the stratigraphy (particularly of Section 181-1123C-17X-4) is suspect. Nevertheless, with circumspection, we place Core 181-1123C-17X within Unit II.
Unit III consists mainly of sediments reflecting a similar interglacial/glacial cyclic sedimentation pattern to that described for Subunit IB. The unit extends from 450.8 to 587.2 mbsf, where it is bounded below by the mid-Oligocene Marshall Paraconformity. A ~7-m-thick debris flow occurs within this sequence, allowing the division of the unit into three subunits (Fig. F5), with Subunit IIIB itself corresponding to the debris flow.
Subunit IIIA, corresponding to the early to middle Miocene, extends from 450.8 to 542.9 mbsf. Its base corresponds to the upper contact of the deformed sediments of the underlying debris flow. Alternating light greenish gray (5BG 7/1) and greenish gray (5G 6/1) interbeds again dominate the sequence, with layers typically <1 m thick. Contacts are generally bioturbated, but the abundant biscuiting and brecciation makes the positions of many contacts subjective. Color differentiation is not as clear as in Unit I. The major difference in composition between this sequence and Unit I is the abundance of terrigenous clay present in the sediment. In addition to the light greenish gray clayey nannofossil chalk, clay concentrations are sufficient to classify the greenish gray layers as nannofossil mudstone, with a dominant nonbiogenic component of terrigenous clay, with nannofossils and micrite as a subordinate constituent, although not less than 40% of the total abundance.
Unlike the oozes and chalks of Unit I, the sediment appears comparatively barren of siliceous fauna, with only "Present" concentrations of sponge spicules in the smear slides. The noncarbonate mineralogy is dominated by clays with "Common" concentrations of diagenetic pyrite, and shows few compositional differences between layers. Only mica appears richer (concentrations from "Trace" to "Present") in the greenish gray layers. Pyrite occurs throughout the subunit as rare smears. Nannofossils are starting to recrystallize, and there is incipient micritization.
Most of Subunit IIIA is moderately bioturbated, with occasional intervals of more intense bioturbation. Mottling is therefore typically faint. Identified trace fossils include well-preserved examples of Zoophycos, Planolites, Chondrites, Thalassinoides, Cylindrichnus, Terebellina, Teichichnus, Palaeophycos, and Skolithos.
A chaotic assemblage of plastically deformed clasts of clayey nannofossil chalk extends from 542.9 to 550.5 mbsf. The clasts are greenish gray (5GY 6/1) and light greenish gray (5BG 7/1) and of 0.5- to >7-cm diameter. The sediment matrix shows strongly contorted laminations, which are possibly flow lineations (e.g., Fig. F12). An intact block of white (5Y 8/1) foraminifer-bearing micritic limestone fills the core between 547.8 and 548.6 mbsf. Although the upper contact of this subunit is conspicuous, the position of the basal contact is somewhat subjective, as the chaotic assemblage feathers out and grades to an intact sediment of a similar color but with disturbed bedding, probably caused by emplacement of the overlying bed. The sediment structure of Subunit IIIB is consistent with the instantaneous deposition of a debris flow originating from further upslope on Chatham Rise.
Subunit IIIC extends from ~550 to 587 mbsf and is identical in lithology to Subunit IIIA. Of note is the paraconformity between early Miocene and early Oligocene ages (~12 m.y. gap; see "Biostratigraphy"), which marks the base of Subunit IIIC (Fig. F13). This paraconformity, at 587.2 mbsf in Section 181-1123C-29X-2, manifests itself as a simple burrowed contact between greenish gray (5GY 6/1) nannofossil mudstone of Subunit IIIC and white (5Y 8/1) micritic limestone below (Unit IV). Beneath the contact (strictly Unit IV) are particularly well-preserved Zoophycos and Chondrites, containing fill from the overlying greenish gray mudstone. The contact closely resembles the shallow-water, onland occurrences of the Marshall Paraconformity (e.g., Concord Greensand on the Burnside Marl, at the abandoned Burnside cement quarry, Dunedin, New Zealand [R.M. Carter, pers. comm., 1998]).
Below the paraconformity at 587.2 mbsf, the remainder of the cored section, to its base at 632.8 mbsf, is alternating white (5Y 8/1) and light gray (5Y 7/1) to light greenish gray (5BG 7/1) micritic limestone. Bioturbation is variable through Unit IV, ranging from rare to abundant. Observed ichnofauna include Zoophycos, Skolithos, Chondrites, Teichichnus, and Planolites. Core disturbance is variable, with moderate to extreme formation of drilling biscuits together with limited brecciation of some biscuits. Of special note is the presence of fine clay residues that mark out stylolitic surfaces through Core 181-1123C-30X (594.3-603.9 mbsf; Fig. F14). Such a feature indicates advanced diagenesis and pressure dissolution within the limestone.
The alternations of light and dark colors in the sedimentary sequence of Site 1123 probably represent cool/warm climate cyclicity, as noted elsewhere off eastern New Zealand (e.g., Griggs et al., 1983; Nelson et al., 1986b; Weaver et al., 1998). The cyclicity is not always self-evident in hand specimen, as is the case for Unit II, but it is almost invariably recorded by either or both the multisensor track and light reflectance measurements. Carbonate proxy values alternate between ~50% and 80% in the oozes of Unit I. These values are lower than those found at the isolated Campbell Plateau Site 1120 (see "Carbonate and Organic Carbon" in the "Organic Geochemistry" section in the "Site 1120" chapter), suggesting a greater terrigenous accumulation at Site 1123. The mica tracer work of Carter and Mitchell (1987) suggests that at least part of this terrigenous component was carried to Chatham Drift by the DWBC. This flow transported suspended load from the region of the Bounty Fan, which, in turn, derived sediment ultimately from the Southern Alps of the South Island.
A short kasten core from the vicinity of Site 1123 (Core CHAT 1K, from NIWA station S924), contains a succession of alternating light-colored, clayey nannofossil ooze and dark colored silty clay from the late Quaternary. On the basis of oxygen isotope curves for benthic foraminifers, foraminiferal assemblage data, and calcium carbonate profiles, Weaver et al. (1998) and Lean and McCave (1998) concluded that the light and dark layers represented interglacial and glacial phases of sedimentation, respectively. Interglacial sedimentation rates ranged over 1.8-2.3 cm/k.y., whereas glacial rates were 2.3-2.9 cm/k.y. Unpublished mass accumulation rates for Core CHAT 1K show that the slightly higher glacial sedimentation rates were largely caused by an increase in the noncarbonate flux of terrigenous and biogenic siliceous components. The nature of the elevated noncarbonate accumulation in glacial periods is not clear, but may result from an influx of waterborne and aeolian detritus together with increased biogenic silica production, supported by upwelling under increased glacial wind stress (R.M. Carter et al., unpubl. data). Organic carbon fluxes are distinctly higher during glacial Stages 2, 4, and 6, caused by higher productivity (Lean and McCave, 1998). Initial indications are that the recovered sediment from Site 1123 contains a record of Milankovitch-frequency colder/warmer cyclic sedimentation since at least the early Miocene.
The average late Quaternary sedimentation rate for Site 1123, estimated from the position of the Brunhes/Matuyama boundary (31.72 m; 0.78 Ma), is around 4 cm/k.y. The discrepancy between this rate and that reported above for Core CHAT 1K may reflect the different coring techniques; the kasten, being a gravity corer, probably compresses the section, whereas the piston of the APC may increase the natural thickness of the section by ~10% (see "Composite Depths").
Well-preserved trace fossils occur throughout this site. Their identification and assemblage provide some limited information on the paleoenvironment, which is summarized in Figure F15. The alternating sequences of clayey nannofossil ooze and clayey nannofossil chalk of Subunits IA and IB have similar ichnofauna, containing Zoophycos, Chondrites, Planolites, Thalassinoides, Terebellina, and Teichichnus. The ichnofauna assemblage alternates between Cruziana and Zoophycos ichnofacies in the upper ~80 mbsf of Subunit IA and throughout Subunit IB. Between these cyclic ichnofacies at ~80 and 180 mbsf, low diversity and a high abundance of dominant Zoophycos are consistent with a typical Zoophycos ichnofacies. Such an ichnofacies indicates a fairly quiescent deep-water environment with low mass accumulation rates, and moderate to low oxygen saturation of the overlying water (Pemberton and MacEachern, 1995). In contrast, the Cruziana ichnofacies is indicative of a deep-water environment that is more energetic and better ventilated (for a detailed discussion of the recurring archetypal trace fossil associations [ichnofacies] and their common environmental implications, see Pemberton and MacEachern [1995]). The alternating ichnofacies in Units IA and IB are in a sequence consistent with the interglacial (Zoophycus)/glacial (Cruziana) cyclic sedimentation pattern of Unit I and may similarly reflect variability in the DWBC flow.
The monotonous clayey nannofossil chalk of Unit II, which has no visible color cyclicity, has an ichnofauna with abundant Zoophycos and lesser Chondrites, Planolites, and Palaeophycos continuing to ~470 mbsf. The resulting Zoophycos ichnofacies suggests that the middle to late Miocene was a period of relative quiescence, with slow DWBC flow and low oxygen concentrations. An isolated zone of Skolithos occurs in 181-1123B-43X and probably represents opportunistic colonization.
Subunit IIIA contains well-preserved Zoophycos, Planolites, Chondrites, Skolithos, Thalassinoides, Terebellina, Palaeophycos, Teichichnus, and Cylindrichnus. The presence of abundant Skolithos, Cylindrichnus, and Thalassinoides just below the Unit II/Unit III boundary at around 470 mbsf suggests an abrupt transition from the overlying Zoophycos to a robust Cruziana ichnofacies. The structure of the latter suggests a higher energy environment than is seen in the equivalent ichnofacies of Unit I. There is a gradual change through to Subunit IIIC with the Cruziana ichnofacies alternating with a lower energy Zoophycos ichnofacies and eventually becoming only a Zoophycos ichnofacies at 538 mbsf toward the Marshall Paraconformity.
These data suggest that DWBC activity was minimal in the early Miocene, but showed a gradual increase into the middle Miocene and then an abrupt decrease and relatively quiescent conditions through the late Miocene. It is tempting to suggest that this pattern may, to some extent, reflect a middle Miocene cooling event in Antarctica (e.g., Kennett, 1977).
In the upper Eocene and lower Oligocene micritic limestone of Unit IV, the trace-fossil assemblage is nearly identical to the Zoophycos ichnofacies in Subunit IIIC with Zoophycos, Chondrites, Planolites, and Teichichnus. Other observed trace fossils are generally robust burrows of opportunistic colonizers (e.g., Teichichnus and Skolithos).
The zone of strongly deformed clayey nannofossil and nannofossil ooze with matrix-supported intraclasts (Subunit IIIB) is interpreted as part of a debris flow (e.g., Prior et al., 1984). The regional bathymetry suggests a flow source from the northern flank of Chatham Rise. The multichannel seismic line through Site 1123 sheds little light on the debris flow. It appears to coincide with a zone of short, discontinuous reflectors that are indistinguishable from the reflectors of in situ chalk higher up in the sequence. As the debris flow is only 7.2 m thick, it may simply be too small to detect on the large scale of the seismic line.
In contrast, the physical properties are a little more conclusive with respect to identifying the flow deposit. Magnetic susceptibility above and below the debris flow tends to decrease downcore, whereas susceptibility within the flow is fairly uniform. A prominent spike between 547.7 and 548.5 mbsf coincides with the block of white limestone incorporated within the debris flow. The light color of this block is also detected in the reflectance profile, although the remainder of the deposit did not produce a distinctive response. Similarly, the gamma-ray attenuation porosity evaluator (GRAPE) density profile failed to define the deposit clearly. By comparison, paleomagnetic measurements detected a marked change in the debris flow. This change was a sharp departure from the normal polarity of sediments above (Chron C6n) and below (Chron C6n or C6 An) the flow. The change was a response to a fabric that was neither clearly normal or reversed. Changes were also revealed in the field in the underlying 4 m of sediment extending to 554 mbsf. Either the flow deposit is thicker than can be detected visually, or the magnetic fabric of the underlying beds has been disturbed slightly by the emplacement of the flow. The first option is less likely because the apparent bedding of the underlying layers is consistent with that of in situ clayey chalks. Furthermore, the paleomagnetic change in the underlying beds is less marked than that of the debris flow proper.
Plastic deformation affects the debris-flow matrix and some of the intraclasts. Deformed laminae are a feature of the matrix and probably reflect flow lines rather than original bedding, as the chalks above and below the debris flow are bioturbated and, therefore, poorly bedded. Intraclasts include soft clayey nannofossil chalk and an 80-cm-long cored section of a foraminifer-bearing micritic limestone. The microfauna within this block suggests an age that is contemporaneous with, or up to 1 m.y. older than, the flow. The lithified character of the block indicates it is older than the flow, which is tentatively assigned an age of 19.5 Ma on the basis of the magneto- and biostratigraphy.
We can only speculate upon the mechanism initiating the mass movement. The debris flow coincides with a period of active deformation that accompanied the onset of the present New Zealand collisional plate boundary (e.g., Rait et al., 1991). An increase in seismicity may have been sufficient to trigger the mass movement on what is normally a passive margin.
A notable feature of Site 1123 is the presence of numerous macroscopic tephra layers from the middle Miocene to present (Fig. F8). Age estimates for individual tephra layers are given in "Paleomagnetism." Many of the tephra layers found at Site 1123 contain fresh glass and phenocrysts (plagioclase) suitable for radiometric dating (Pillans et al., 1996) and thus may become numerical stratigraphic calibration points (Shane et al., 1996). Since New Zealand is the only area in the Southwest Pacific with abundant rhyolitic volcanism during the late Cenozoic (Nelson et al., 1986a), it is the probable source for the tephra. For the late Quaternary, Carter et al. (1995) have shown that widespread downwind dispersal of tephra occurred from the Taupo Volcanic Zone (North Island, New Zealand). Voluminous eruptions have resulted in deposition of macroscopic tephra at least 1400 km from the source (e.g., the Kawakawa Tephra). For pre-Quaternary rhyolitic tephras, the probable eruptive source is the Coromandel region of northern North Island, which was active during the Neogene (Ballance, 1976; Suggate, 1978).
The frequency and level (in meters composite depth) of the recovered tephra layers show some differences between the three holes (Fig. F8). This includes a 22-cm-thick tephra layer recovered in Hole 1123C (Section 181-1123C-11H-6; 103.17-103.39 mbsf) of which no evidence is found in either Hole 1123A or Hole 1123B. This remains enigmatic. Original spatial variability in burial (note the holes were each 30 m apart) or artifacts induced by the drilling are possible explanations.
The best-fit tephra stratigraphy is depicted against the composite section for Site 1123 in Figure F8. Forty-five tephra beds are recognized, the oldest of middle Miocene (~12 Ma) age. Naming individual tephra on the basis of stratigraphic position awaits onshore chemical analysis and dating. However, the two youngest tephra layers are almost certainly the well-documented Kawakawa Tephra (note this tephra is heavily bioturbated and its position is picked from magnetic susceptibility), which is dated at 22,590 radiocarbon years (Carter et. al., 1995), and the distinctive white-pink Omataroa Tephra (28,000 radiocarbon years; Froggatt and Lowe, 1990).
The composite tephra section suggests an event frequency of 1/27 k.y. for major eruptions. However, the tephra beds are not evenly distributed through time, and a major period of explosive activity occurred at ~1.60-1.35 Ma (~51-64 meters composite depth [mcd]). This could be related to the period of extensive volcanism inferred between 1.79-1.60 Ma from tephras of the southern North Island (Shane et al., 1996).
Additional lesser eruptions may be represented by the numerous green laminae in cores from Site 1123. These probably represent smectitic (i.e., bentonitic) alteration products of former tephra (Gardner et al., 1986). As Nelson et al. (1986a) suggest for Leg 90 sediments, the preservation potential for such inferred more basic material in highly calcareous sediment is lower than it is for the New Zealand-derived rhyolitic shards. Overall, however, the macroscopic tephra recorded at Site 1123, together with the continuous, astronomically tuned stratigraphy that may result for the site, will provide a unique record of Cenozoic volcanism in New Zealand and the Southwest Pacific region.