The Canterbury continental shelf, and, therefore, the lithology at Site 1119, are strongly influenced by the terrigenous input from New Zealand. Marked uplift of readily erodible rocks along the South Island Alpine Fault, coupled with a vigorous climate, have ensured delivery of large volumes of sediment to the Canterbury shelf (Gibb and Adams, 1982; Griffiths and Glasby, 1985). Much of the modern fluvial load is entrained within an active, along-shelf transport system, and it is mainly under glacial sea-level lowstands that significant quantities of sediment reached the continental slope (e.g., Carter and Herzer, 1979; Carter and Carter, 1993). Thus, at Site 1119 we can anticipate a eustatically controlled supply of terrigenous sediment. In addition, the Canterbury continental margin is swept by coastal and longslope currents, which results in the transport of sediment and subsequent formation of large sediment drifts along the continental slope (Fulthorpe and Carter, 1991). It is against this background of sea-level-controlled, terrigenous supply to a contour current that Site 1119 was drilled.
The sedimentary sequence is divided into three lithostratigraphic units that are identified on the basis of sediment grain size and composition, together with downhole changes in bedding characteristics. The dominant criteria used for recognition of the units are visual observation of the core and estimates of core composition from smear slides ( "Site 1119 Smear Slides"), supported by shipboard measurements of calcium carbonate, physical properties, and bulk mineralogy using x-ray diffraction (XRD). These data, along with biostratigraphic control (see "Biostratigraphy"), are summarized in Figures F5 and F6. The three lithologic units fit well with the detailed seismic stratigraphy derived from the processed multichannel seismic line run by the Maurice Ewing in 1996 (see "Background and Objectives"). This line also assisted in the subdivision of lithostratigraphic Unit II and in the paleoenvironmental interpretation.
Unit I extends from the present seafloor to ~88 mbsf and has a dominant lithology of greenish gray (5GY 5/1 to 5G 5/1) silty clay interbedded with subordinate silty sand and sandy silt. On seismic profiles it appears as a drape resting on a sequence of sediment drifts. The thickness of individual clayey beds is difficult to quantify because of expansion gaps in the core but, where intact, beds are commonly >2 m thick and reach a maximum thickness of 14 m. Sediments are often faintly mottled and banded with darker (greenish gray, 5G 4/1) and lighter (greenish gray, 5GY 6/1) tones (Fig. F7). Localized dark staining, associated with concentrations of pyrite, is also a common feature. Trace fossils are ubiquitous with Asterosoma, Chondrites, Terebellina, and Thalassinoides as the prominent forms. Smear slides reveal a prevalence of quartzofeldspathic silt grains and clay minerals with nannofossils as a prominent accessory. Siliceous sponge spicules are locally abundant. Mica is also present and gives the silty clay a distinctive speckled appearance in hand specimen. Macrofossils are sparsely scattered throughout the beds and include fragments, single and double valves of the mollusks Tawera spissa and Zygochlamys delicatula (Fig. F8), plus fragments of the brachiopod Neothyris. The presence of Zygochlamys delicatula is particularly important as it signifies cold-water conditions, which is consistent with the subantarctic microfauna of the clays (see "Biostratigraphy"). Another indication that the silty clays were deposited during glacial periods is their close lithologic and stratigraphic similarity to clays from the offshore Canterbury region, whose glacial paleoclimatic signature has been reliably determined by stable isotopes (e.g., Nelson et al., 1986).
Contrasting with the glacial silty clays are scattered beds of olive-gray (5Y 5/1-5Y 5/2) sandy silt and silty sand. Bed thicknesses are typically <1 m, although in Core 181-1119B-6H a bed reaches 1.8 m (42.7-44.5 mbsf). Basal contacts with the glacial clays are usually, but not invariably, sharp and often include sand-filled Chondrites burrows below (Fig. F9). The sand-sized fraction is dominated by quartzofeldspathic grains and varying numbers of foraminifers, with accessory quantities of opaque minerals and rock fragments. Nannofossils are conspicuous in the finer fraction. Macrofossils, especially valves of Tawera spissa, are also conspicuous. Although paleoenvironmental analysis of the microfaunal content was not carried out, the distinctive olive-gray color of the sediments and their stratigraphic position correlate well with documented interglacial deposits in the Canterbury region (e.g., Griggs et al., 1983; Nelson et al., 1986).
Seismic profiles across Site 1119 depict the main bodies of two sediment drifts and the edge of a third. These deposits are represented in the core by lithostratigraphic Unit II, which can be divided into Subunits IIA, IIB, and IIC on the basis of bedding style and lithology.
Subunit IIA, corresponding to the youngest drift, extends from ~88 to ~196 mbsf. Its base corresponds to a prominent reflector marking the top of Subunit IIB drift. Cores show that the Subunit IIA drift is a succession of alternating silty clay and shell-bearing silty sand beds with the same color alternations as Unit I. Silty clays are mainly greenish gray (5GY 5/1) and the silty sands are olive gray (5Y 5/1- 5Y 5/2), suggesting that even though they form part of the drift, they nonetheless represent an interglacial/glacial cyclic sedimentation pattern similar to that described for the overlying sediments.
The greenish gray silty clays are the dominant lithology with beds commonly up to 6 m thick (given thickness inaccuracies caused by post-coring expansion of the sediment). Beds tend to have faint color banding and indistinct mottling that is associated with trace fossils such as Planolites, Thalassinoides, and Chondrites. Like sediments from Unit I, Subunit IIA hand specimens exhibit mica flakes, dark stains associated with pyrite, and scattered shell debris, but the quantities of the last two components are less than in their Unit I counterparts. Nevertheless, there were still sufficiently large fragments to identify the key molluscan species, Zygochlamys delicatula, which makes its last appearance (first time upcore) in Core 181-1119C-18X at 168 mbsf. As noted earlier, the presence of this species confirms the dominance of cold-water conditions at the site. In smear slides, the main constituents are clay, quartzofeldspathic silt, and nannofossils with accessory amounts of opaque minerals (?pyrite), foraminifers, mica, and, near the top of the subunit, radiolarians.
While the greenish gray silty clay beds of Unit I and Subunit IIA are similar, the olive-gray (5Y 5/1) silty sand beds of these units are different. Compared to Unit I lithologies, the underlying sand beds of Subunit IIA are more frequent and, on average, are thicker, typically >1 m. Some beds appear to be <1 m but their tops or bottoms are often missing at core voids, suggesting that the weakly consolidated sediments may have been washed out during drilling. Subunit IIA silty sand beds are further distinguished by their normal grading (Fig. F10), which may sometimes be repeated within a single bed without an intervening erosional contact. Clayey sediments below the bases tend to be strongly bioturbated with olive-gray, sand-filled burrows of Chondrites and Thalassinoides standing out against the greenish gray, silty clay of the host sediment. These graded sands pass upcore to clayey silts, which have a gradational contact with the overlying greenish gray silty clays. Between 120 and 140 mbsf, the olive-gray beds become finer grained (clayey silt to silty clay), with gradational contacts at the top and bottom of each bed. In hand specimen, these fine-grained deposits are most readily distinguished by color from the greenish gray clays. Coarser, graded silty sand reappears below 140 mbsf. Hand specimen and smear-slide data further highlight differences with the Subunit IIA silty sands, being richer in shell debris, benthic foraminifers, and glauconite, compared to Unit I silty sands. Glauconite, however, still remains in accessory (<5%) amounts--this smear-slide observation is confirmed by XRD data.
Subunit IIB corresponds to the second drift that extends from ~196 to ~285 mbsf. Nearly all of this drift (~207-~275 mbsf) is early Pleistocene in age with the base possibly extending into the late Pliocene. It carries many of the lithologic features of Subunit IIA: thick beds of greenish gray (5GY 5/1) silty clay intercalated with thinner but nonetheless distinctive graded layers of olive-gray (5Y 5/2) shell-bearing silty sand with gradational tops and bioturbated bases. There is also a downcore change in grain size to finer grained (clayey silt to silty clay) beds with gradational bottoms and tops (274-280 mbsf). The ichnofossil assemblage of both subunits is similar with Thalassinoides and Planolites as common forms. However, Chondrites is less common than in Subunit IIA. Finally, the gross composition of both subunits is similar.
Despite the similarities, sufficient differences exist to warrant designation of the two subunits. The olive-gray drift sands have higher carbonate contents (60%-70% in Subunit IIB compared to <50% in Subunit IIA) that reflect higher concentrations of nannofossils, molluscan shell debris, and benthic foraminifers. Much of the shell material is too comminuted to allow reliable identification of genera. However, a few unbroken mollusks survive, including the bivalve Limopsis and the gastropod Taniella. In addition, Subunit IIB has higher quantities of mineral carbonate, which coincides with the first development of drilling biscuits, suggesting the sediment is becoming lithified through the formation of carbonate cement.
Of special note is the unique occurrence of a thin fine-sand turbidite in the greenish gray silty clays at 288 mbsf (Core 181-1119C-31X) (Fig. F11). The basal sand is 2 cm thick, but the top of the bed could not be pinpointed, as it merged with the ambient silty clays. Also evident in this core and other silty clay beds throughout Site 1119 is faint color banding. Smear slides of samples from a light-colored band and the overlying greenish gray clay (181-1119C-31X) showed that both sediments had similar compositions, with the notable exception of the light band, which contained only "rare" nannofossils in contrast to the overlying nannofossil-rich clay. The dearth of microfossils in the light band may reflect a turbidite origin either through direct river discharge at lowered sea level or remobilization of rapidly deposited sediment. Whatever the case, the compositional differences and the association with a sand turbidite suggest that the color banding is a reflection of mud turbidite deposition.
Subunit IIC encompasses a sequence of strong reflectors that represent the landward edge of a large drift as it thins toward the moat at the base of the late Pliocene paleoslope. On the basis of lithology, this subunit lies between ~285 and 401 mbsf. Again, alternations of greenish gray silty clay with olive-gray silty sand dominate the sequence. But, in contrast to the overlying subunits, the silty sand beds of Subunit IIC have (1) a higher concentration of carbonate (reaching 75%) (Fig. F5), and (2) a higher frequency of occurrence, with 16 silty sand beds per 100 m of section compared to 12 beds in Subunit IIA and 10 beds in Subunit IIB (note these numbers are based on the lithostratigraphic summary Figure F5 and, in reality, the numbers of individual beds, as recorded on the visual core logging sheets, will be higher). An abundance of silty sand may also be reflected in the poor core recovery of Subunit IIC, which was the worst of the entire Site 1119 core (Fig. F5). Recovered lithologies were mainly well-consolidated greenish gray silty clay, whereas silty sand beds, when present, were often incomplete. Furthermore, the silty sand typically occupied the core catcher, where it presumably was protected from washing out.
The remainder of the cored section, from 401 to 494.8 mbsf, is almost exclusively greenish gray (5GY 5/1) silty clay. This firm sediment has been disrupted by the ubiquitous formation of drilling biscuits. For much of the section, the sediment appears to be moderately bioturbated with Zoophycos and a single occurrence of Chondrites. Despite sediment reworking, some mud turbidites may be present at 475 and 483 mbsf, as evinced by a rhythmic sequence of <3-cm-thick, greenish gray (5GY 6/1) laminae normally grading up to the typical greenish gray silty clay. Unit III has a mixed terrigenous/pelagic biogenic composition of clay, quartzofeldspathic silt grains, calcareous nannofossils, and minor amounts of foraminifers. Although XRD data points are scattered and semiquantitative, silty clays of Unit III also appear to be richer in mica and chlorite, compared to their Units I and II counterparts.
No glacial/interglacial signature could be discerned from the lithology.
Boundaries between the units and subunits at Site 1119 have been identified on the basis of core lithology and with reference to the seismic stratigraphy (Fig. F2; Table T2). Prominent seismic reflectors can be correlated accurately to the cores using seismic velocities measured in the drill hole during the logging operations. The measured velocity increases downhole from 1375 to 1558 m/s.
In Unit I, we see the classic interglacial/glacial cyclicity found elsewhere in continental slope settings around New Zealand (e.g., Griggs et al., 1983; Nelson et al., 1986). The slope sands are interpreted to have accumulated during interglacial highstands, when the terrigenous supply to the outer continental shelf was reduced by both onshore alluviation and diversion along the shelf by the prevailing storm-forced current regime (Carter and Herzer, 1979). As a result, the shelf and slope were starved and the sand was presumably generated either from sediment reworked and redeposited from the shelf edge, from the reworking of older slope deposits, or from both these processes. Off the modern Canterbury margin, the shelf and slope down to ~400 m is subject to the north-flowing Southland Current and Front (Chiswell, 1996). This flow has a mean speed of ~6 cm/s but may exceed 30 cm/s when forced by storms (Carter and Herzer, 1979), when seasonal changes affect the density stratification of the Southland Front (Chiswell, 1996), and when flood tidal currents, which themselves are intensified by the slope topography, reinforce the current. Combine this with the superimposed effects of large surface waves and internal waves breaking against the shelf edge (Carter and Herzer, 1979), then sediment transport will take place. While observations of the flow regime over the upper slope are sparse, the presence of eroded sediments in 3.5-kHz profiles (Carter et al., 1985; Herzer, 1981) and gravel/sand lags on shelf/slope terraces (Carter et al., 1985), attest to the power of modern currents to erode and move bedload.
In contrast, glacial periods were times of high sediment supply to the shelf edge and slope, as witnessed by the thickness of the greenish gray silty clay beds at Site 1119, and by sedimentation rates measured for various parts of the eastern South Island margin, Carter et al. (unpubl. data), for example, indicate a two- to fourfold increase in the noncarbonate sediment flux in the last glaciation, compared to the Holocene. At times of lowered sea level, South Island rivers extended across the emergent continental shelf and discharged their loads close to the shelf edge (Carter et al., 1986). Fine silt and clay, settling through the water column, together with pelagic biogenic detritus and aeolian dust, contributed a hemipelagic component to bottom sediments. Mud turbidites also may have contributed to the glacial sediment budget, but unequivocal turbidites are scarce in the core. Finally, sediment may have been swept into the area from the south by the ancestral Southland Current. It is unlikely that the current generated fine-grained sediment through reworking of the substrate, as there is a lack of moating in Unit I (see "Background and Objectives").
Unit II also accumulated under a climatically controlled sediment supply, as indicated by the alternations of greenish gray silty clays (glacial) and olive-gray silty sands (interglacial). However, compared to the slope drape deposit of Unit I, Unit II accumulated as a series of drift deposits. The seismic architecture of shelfward-migrating, mounded reflectors, with each mound separated from the paleoslope by a moat, identifies the mounds as shelf-parallel drifts (Fulthorpe and Carter, 1991). These authors surmised that the drifts were deposited by a contour current flowing along the Canterbury slope since Miocene times. The moats were identified as zones of either nondeposition or erosion, where the current was topographically intensified against the slope under Coriolis deflection (to the left facing downcurrent in the Southern Hemisphere). Such an inference requires that the current flowed north.
For Site 1119, a key issue is how the lithologic data compare with the seismic interpretation of drift sedimentation by Fulthorpe and Carter (1991). Lithologic characteristics of sandy and muddy drift deposits (contourites) have been summarized by Stow and Holbrook (1984). For "sandy contourites," these authors cite features such as fine irregular layering, extensive bioturbation, poorly sorted silt/sand (although well sorted sands may also occur), and mixed terrigenous and shell-bearing sands. For "muddy contourites," key features are homogeneity with poor or no bedding, rare silt laminae, bioturbation structures, mainly silty mud with <15% sand, and elevated carbonate levels. While these generalized criteria fit the Canterbury slope sediments well, there remains the possibility that some beds, in particular those with well-defined graded intervals, may be (silt or) mud turbidites. Certainly, the presence of a thin (<1 cm), graded, very fine sand bed with a sharp, eroded base indicates an incursion by a turbidity current into the greenish gray silty clay facies of Unit II. Mud turbidites may also have contributed to this facies, as suggested by occasional color banding. However, unless much of the banding has been obliterated by bioturbation, which is unlikely, the influence of turbidites in the glacial silty clay facies appears to be fairly minor. The graded beds in the interglacial silty sand are also discounted as turbidites because (1) they lack any components other than "A" of the classic Bouma sequence, (2) normal graded sequences may be repeated in a single bed without an erosional contact between each sequence, and (3) some normal graded beds are preceded by reverse graded sediments without an obvious break in succession. Such features are more consistent with deposition under a fluctuating current rather than a continuously decelerating turbidity current.
Assuming that a contour current had a significant influence on Unit II sedimentation, the following events are envisaged. During interglacial sea-level highstands, the terrigenous supply of sediment to the upper slope was reduced. Sediment reaching the upper slope was probably a mixture of terrigenous and biogenic shell debris reworked from the shelf edge terraces (e.g., see Carter et al., 1985) and sand winnowed from underlying drift deposits. In each of Subunits IIA and IIB, the resultant olive-gray, shell-bearing, silty sands exhibit a distinctive downcore trend of (1) a succession of single and multiple graded beds with sharp bases that are commonly bioturbated to (2) a succession of finer-grained beds with gradational top and bottom contacts with the glacial silty clays, to (3) a return to the coarser silty sand beds with sharp basal contacts. Such a trend may be related to variations in the local speed of the current and/or to changes in the supply of sediment to the current. Under glacial sea-level lowstands, Site 1119 received large quantities of silt and clay from (1) river discharge near the shelf edge, (2) contour current transport from southern sources, (3) minor turbidity currents, and (4) aeolian detritus swept from the Canterbury Plains (e.g., see Stewart and Neall, 1984; Ives, 1973). Results of XRD suggest a downcore change from calcic to potassic feldspar (Fig. F6), possibly indicating a change in source of the clay component. One possibility is that if the potassic feldspar has in part a plutonic source, then it could mean a southerly provenance (e.g., the plutonic complex of Stewart Island). Detailed mineralogical work is required to confirm this. Whether or not climatic fluctuations affected the speed of the contour current itself has yet to be determined.
Unit III, at least in the core section, carries no obvious lithologic signature of the paleoclimatic fluctuations evident in the overlying units. It is possible that some sand beds may have been lost from Unit III during the coring. Whatever the case, the lithologic and seismic data show the unit is a widespread mud drape overlying a large drift of probable middle to early Pliocene age. Reasons for the change from drift to mud drape require further investigation.