The role of intermediate-depth currents in continental shelf–slope accretion: Canterbury Drifts, SW Pacific Ocean

The Late Oligocene to Recent Canterbury Drifts were deposited in water depths between c. 400 and c. 1500 m by northward-flowing, cold, intermediate-depth water masses: Subantarctic Mode Water (SAMW), Antarctic Intermediate Water (AAIW) and their predecessor current flows. Drift accumulation started at c. 24 Ma, fed by terrigenous sediment derived from the newly rising Alpine Fault plate boundary in the west, which has built a progradational shelf–slope sediment prism up to 130 km wide at rates of eastward advance of up to 5.4 km Ma. Gentle uplift associated with the nearby plate boundary has exposed older Late Oligocene and Miocene drifts onland (Bluecliffs Formation). Ocean Drilling Program Site 1119, located 100 km offshore at a water depth of 394 m, penetrated a 428 m thickness of midPliocene to Pleistocene (0–3.9 Ma) drift located just seaward of the eastern South Island shelf edge. Uniquely, these large (.60 000 km), regionally extensive, intermediate-depth sediment drifts can be examined in outcrop, in marine drill-core and at the modern sea bed. The drifts comprise planar-bedded units up to several metres thick. Some sand intervals have sharp, erosive bases and normally graded tops into overlying siltstone; others are symmetrically graded with reversegraded bases and normally graded tops. Bioturbation is moderate and rarely destroys the pervasive background, centimetre-scale, planar or wispy alternation of muddy and sandy silts displayed by Formation Micro-Scanner imagery. These features are consistent with deposition from rhythmically fluctuating bottom currents. Texturally, the drifts are polymodal quartzofeldspathic silty sands, sandy silts, silts and silty clays, with varying admixtures of benthic and biopelagic carbonate and silica. Miocene samples are mostly dominated by coarse silt (45–60 mm) and very fine sand (70–105 mm) grain-size modes, whereas strong fine silt (11–13 mm) and very fine silt– clay (,5 mm) modes become dominant after c. 3.1 Ma in the Late Pliocene, consistent with an increasing input of glacially ground material. Over the Plio-Pleistocene part of the succession, the sand–silt lithological rhythmicity occurs in synchroneity with Milankovitch-scale climate cycling, with periods of inferred faster current flow (sand intervals) mostly corresponding to warm, interglacial periods. Northward drift dispersal has helped cause the seaward growth of the eastern South Island shelf–slope system since the Late Oligocene probably by clinoform progradation and by episodic welding of mounded slope drifts onto the pre-existing sediment prism. Such along-slope, contourite drift accumulation occurs even in the absence of mounded drifts on seismic profiles, and represents a previously underemphasized mechanism for the progradation of shelf–slope clinoforms, worldwide. The Canterbury Drifts vary in thickness from c. 300 m near the early Miocene shoreline, where they were accumulating in limited shallow-water accommodation, to c. 2000 m under the modern shelf edge. Mounded drifts first occur in the Middle Miocene, at c. 15 Ma, their appearance perhaps reflecting more vigorous intermediate water flow consequent upon the worldwide climatic deterioration between 15 and 13 Ma. At Site 1119, a further change from large (.10 km wide) to smaller (1–3 km wide) mounded slope drifts occurs at c. 3.1 Ma, marking further cooling and perhaps the inception of discrete SAMW flows and initiation of the Subantarctic Front. The concept that deep ocean currents play a major role in shaping the continental slope originated from seismic observations of migratory abyssal sediment waves (Ballard 1966; Lonsdale & Hollister 1979), and sea-bed photographs of nearby current-influenced bedforms. Similar features were later shown to occur worldwide beneath the path of contourhugging thermohaline currents (e.g. Ewing et al. 1971; Hollister et al. 1974; Gardner & Kidd 1987; Howe et al. 1997), and also laterally to turbiditycurrent pathways (e.g. Damuth 1979; Normark et al. 1980; Carter et al. 1990). The accompanying ‘contourite’ sediments, with characteristic sand– mud sedimentary structures and textures (Stow & Lovell 1979), have been described both from sea-bed cores and from ancient sedimentary basins (summarized by Stow et al. 1998). Long Deep Sea Drilling Project (DSDP) core samples through wellknown sediment drifts, the Hatton, Gardar and Feni sediment drifts in the North Atlantic, were described by Laughton et al. (1972), Montadert et al. (1979), McCave et al. (1980) and Kidd & Hill (1987). From: VIANA, A. R. & REBESCO, M. (eds) Economic and Palaeoceanographic Significance of Contourite Deposits. Geological Society, London, Special Publications, 276, 129–154. 0305-8719/07/$15.00 # The Geological Society of London 2007. The 1980s saw a widening interest in marine sediment drifts, extending to include those developed in shallower intermediate water depths. The Faro Drift, a 300 m thick, 50 km long body located at depths of 500–700 m along the path of the deep Mediterranean outflow in the eastern Atlantic ocean (Faugères et al. 1984; Stow et al. 1986; Nelson et al. 1993), was a subject of particular study about the time that the first detailed facies models for drift sediments were emerging (Stow et al. 1998; Viana et al. 1998). Although it was implicit in many papers that the deposition of sediment drifts helped build up the continental margin (for instance, McCave & Tucholke (1986) referred to the ‘plastering and decorating’ of the sides of the North Atlantic Ocean basin), most previous research has been focused on the description of drift geometry and sedimentary facies, and the inference of current pathways. Latterly, it has become apparent that some drift deposits play a determining role in the progradational building of the continental shelf and slope (Fulthorpe & Carter 1991; Seranne & Abeigne 1999). This paper discusses the sedimentary texture, composition, structure and origin of one such field of drifts, cored to a depth of 513 m (c. 3.9 Ma) offshore at Ocean Drilling Program (ODP) Site 1119 (Carter et al. 1999). Site 1119 is located at 394 m water depth on the upper continental slope, about 100 km east of Timaru, New Zealand (Fig. 1a and b). The sediments there represent the most recent part of a long-lived, c. 24–0 Ma, succession of terrigenous drifts that underlie the eastern South Island coastal plain– shelf–slope sediment prism (Fig. 2) (Carter, R. M., et al. 1996) and form an important part of the Eastern New Zealand Oceanic Sedimentary System (ENZOSS; Carter, L., et al. 1996). This paper summarizes the available published information on the Canterbury Drifts, both offshore (ODP Site 1119) and onland (Bluecliffs Formation). New sediment textural analyses provide insights into the evolution of the drift succession since the Early Miocene, and comparison between onshore and offshore sites contributes to our understanding of the climatic and oceanographic history of the region. Sediment analyses were conducted according to the laboratory protocols described in the Appendix. For discussion, textural data have been aggregated into three grain-size classes, cohesive mud (cM; ,8.70 or ,9.48 mm), sortable silt (sZ; ,60.65 or ,56.09 mm) and sand (sS; .60.65 or .56.09 mm). These classes represent the bin-boundary grain diameters on the laser particle sizer that, for the 2000 or 600 mm lenses, respectively, approximate to the conventional cohesive–noncohesive and silt–sand boundaries of c. 10 and 62.5 mm. It should be noted also that the non-specific allocation of clay–very fine silt modes to ,5 or ,10 mm is because these grain sizes fall at the lower end of the Mastersizer grain-size spectrum (using, respectively, the 600 or 2000 mm lens), the accuracy of mode identification in these ranges being degraded by instrumental edge effects. Previous research on the Canterbury