During some, but not all, Cenozoic glacial episodes, the Lambert Glacier advanced to various points on the shelf, prograding the shelf and building a large trough mouth fan that records these major advances since the late Miocene-middle Pliocene (Figs. 1B, 4). Interglacial sediments are probably preserved on the slope foresets; thus, the Prydz Channel Fan contains a measure of the major sediment pulses caused by peaks in Antarctic ice volume over the last 4-5 m.y.

The continental rise adjacent to Prydz Bay exhibits large sediment drifts deposited under the influence of turbidity currents from the continental shelf and deep currents in the Southern Ocean (Fig. 4). These drifts are a fine-grained distal equivalent to shelf and upper slope sediments and record fluctuations in the ratio of continent-derived terrigenous sediments to oceanic material and record the fluctuations of oceanic current activity (Fig. 5). The amount of terrigenous material rises strongly with major ice expansions so that interbedding of terrigenous-rich and biogenic-rich horizons tend to reflect glacial-interglacial cycles. The longevity of these drifts and the presence of seismic horizons that can be projected back to the continental slope and shelf mean that these drifts can provide a link between continental glaciation and changes in the ocean back through time to the Paleogene.

Regional Setting
Prydz Bay is a re-entrant in the East Antarctic coastline between 68°E and 78°E (Figs. 2, 4). The bay shape and glacier flow patterns reflect the underlying geological structures, the Lambert Graben and Prydz Bay Basin (Fedorov et al., 1982; Stagg, 1985). The Lambert Graben extends about 600 km inland, is as much as 200 km wide, and contains more than 5 km of sediment, based on magnetic and seismic refraction data (Fedorov et al., 1982). The Lambert Graben-Prydz Bay Basin fill consists of two lower sequences of parallel bedded units that onlap or are faulted against basement beneath the northwestern and southeastern sides of the bay (Fig. 6). These sequences were penetrated at Ocean Drilling Program (ODP) Sites 740 and 741 drilled during Leg 119 and consist of Cretaceous coal-bearing nonmarine sediments overlying nonmarine redbeds (Turner and Padley, 1991; Turner, 1991). Cenozoic sequences overlying the Cretaceous have foreset and topset beds that prograde the continental shelf (Cooper et al., 1991a, 1991b).

West of Prydz Bay, the Mac Robertson Shelf is a passive margin that probably formed during the Mesozoic. It is narrow compared to the Prydz Bay shelf and is rugged, having experienced erosion by glaciers during glacial episodes and by iceberg scouring and geostrophic currents during interglacials (Harris and O'Brien, 1996).

Sedimentary Environments
The Prydz Bay continental slope and rise are underlain by thick (more than 6000 m) post lower Cretaceous sediments. Some of the sediment drifts in Prydz Bay are elongated ridges aligned along the margins of deep channels, others have no clear correlation with channels, but all of them are elongate approximately orthogonal to the continental margin (Fig. 4). The features and seismic patterns of sediment drifts suggest that they have been basically deposited as a result of the interaction of downslope mass flow and strong bottom (contour) currents (Fig. 5). The drifts are composed of a mixture of sediment derived from the continent and biogenic material. Drilling of drifts beneath the continental rise during ODP Leg 178 showed that they preserve alternating clastic-rich and biogenic-rich intervals that reflect alternations of glacial and interglacial conditions. Such records can be compared to the proximal records of the continental shelf and upper slope to understand the relationship between oceanographic conditions and the advance and retreat of the ice sheet.

The most conspicuous sediment drifts are developed in the western part of Cooperation Sea between Wilkins and Wild Canyons and are referred to as the Wilkins and Wild Drifts (Figs. 4, 7). Kuvaas and Leitchenkov (1992) recognized two major seismic unconformities (P1 and P2). Additional data and reinterpretation have allowed the mapping of a third surface younger than P1 and P2 and a better understanding of the likely age and paleoceanographic significance of the surfaces (Fig. 8). Surface P1 within these sediments marks the transition from a lower homogenous part of the section with mostly irregular reflectors to an upper heterogenous one in which a variety of well-stratified seismic facies are present. New more distal data suggests that P1 may be as old as Cretaceous. Surface P2 marks a change to submarine canyons and related channel and levee deposits and chaotic seismic facies. This transition appears to reflect a dramatic change in the continental margin depositional environment that resulted from the onset of continental glaciation in the Eocene or the arrival of grounded ice sheets at the shelf edge in the early Oligocene, as indicated by ODP Sites 739 and 742 (Barron et al., 1991). This sedimentation change produced thick, prograding foresets above the P2 unconformity beneath the Prydz Bay outer shelf (Kuvaas and Leitchenkov, 1992).

Surface P3, above P2, represents the base of deposits containing abundant, well-stratified sediment drift facies, including sediment waves. Such features imply that strong, presumably westerly flowing, bottom currents played a significant role in drift formation. The changes at this level could have been related to initiation of the Antarctic Circumpolar Current after the opening of Drake Passage around the Oligocene/Miocene boundary or may relate to the a major ice expansion during the Oligocene or Miocene.

A major change in Prydz Bay shelf progradation took place in the late Miocene to middle Pliocene when a fast-flowing ice stream developed and excavated a channel across the shelf on the western side of Prydz Bay, where a surface can be mapped from the shelf to the continental rise (Surface PP15, Fig. 9; Surface A of Mizukoshi et al., 1988). Basal debris carried to the shelf edge deposited in a trough mouth fan on the upper slope (Fig. 10; Boulton, 1990; Larter and Cunningham, 1993). This change may reflect the earliest growth of thick ice on the Antarctic coast deflecting the Lambert Glacier when it advanced (O'Brien and Harris, 1996). This trough mouth fan must contain a reasonably complete record of glacial history because it received siliciclastic sediment when the shelf eroded during major ice advances and hemipelagic material during interglacials and smaller glaciations. Preliminary optical stimulate luminsecence (OSL) dating of sediments from the surface of the fan and ¹⁴C Atomic Mass Spectrometer (AMS) dating of last glacial maximum grounding zone deposits on the shelf (Domack et al., 1998) suggest that the Lambert Glacier last grounded at the shelf edge at 80-120 ka, indicating that not every glacial episode produces a major expansion of the East Antarctic Ice Sheet. This raises questions as to which glacial episodes produced a major advance and what conditions were necessary to produce a major advance.

The Antarctic continental shelf displays many glacially excavated valleys that reach depths of several hundred to more than 2000 m. Some of these valleys have acted as sediment traps in areas of high biological productivity and have accumulated Holocene sedimentary sections with resolution approaching that of ice cores. These sedimentary sections typically consist of biosiliceous oozes with a minor terrigenous component. These oozes record changes in phytoplankton that reflect sea-surface conditions such as temperature and sea-ice cover (Domack and McClennen, 1996). The first of these sections to be cored by ODP in the Palmer Deep off the Antarctic Peninsula provide a high-latitude record comparable to other important ODP Holocene cores such as the Santa Barbara Basin (Kennett, Baldauf, et al., 1994).

The Iceberg Alley shelf valley is a U-shaped glacial valley crossing the Mac Robertson Shelf at 63°E offshore from Mawson Station (Fig. 11). It is as deep as 500 m, and part of the valley on the outer shelf is floored by about 50 m of unconsolidated sediments (Fig. 12). Gravity cores from this area recovered siliceous mud and ooze, some of which show decadal-scale fluctuations in diatom floras (Taylor, 1999).

To 188 Scientific Objectives

To 188 Table of Contents