Poor core recovery at Site 1097 allows a partial picture of the older sedimentary history of the continental shelf to be established (Fig. F5). The total depth of the hole was 436.6 mbsf, and ~57 m of sediment was recovered from the lower part of the hole below ~80 m. Only cobbles were recovered in the upper part of the hole. In the absence of a continuous sedimentary record, lithostratigraphic subdivision is avoided. Nonetheless, a number of glacial and glacially influenced marine sedimentary facies can be identified from recovered core intervals.

Subglacial environments are recorded by massive diamicts (defined as clasts supported by a muddy matrix; see "Lithostratigraphy"  in the "Explanatory Notes" chapter) that lack any internal structure or bioturbation and contain reworked marine microfauna. These sediments are interpreted as deformation tills produced by the subglacial reworking and transport of marine muds and other sediments below grounded ice. Massive diamicts are the most common facies recovered at Site 1097 (65% of recovered sequence).

Glacially influenced continental shelf environments can also be recognized. Stratified diamicts, which occur in close association with normal and inversely graded diamicts, are interbedded with laminated marine muds containing a well-preserved in situ marine fauna. These diamict facies are interpreted as sediment gravity flows (debrites and turbidites) deposited in a proglacial marine setting characterized by low salinity and high deposition rates, most likely within a few kilometers of an ice margin. Such glaciomarine diamicts are characterized by a soupground ichnofacies assemblage, indicating that diamicts were poorly consolidated during deposition.

Laminated and massive marine muds, which contain varying numbers of dropstones and are bioturbated, have an in situ marine microfauna, indicating that deposition occurred in both ice-proximal and ice-distal settings. Ice-rafted debris is present throughout; some marine muds contain sufficient ice-rafted debris to generate thin (<30 cm) intervals of massive rainout diamict.

Massive diamict (Facies Dmm; Figs. F6, F7) is the most common facies recovered at Site 1097 comprising ~65% of total recovery. The maximum recovered interval through such facies is 2.95 m (Core 178-1097A-37R). The term diamict refers to poorly sorted admixtures of clasts (defined as larger than sand sized) and matrix (see "Lithostratigraphy" in the "Explanatory Notes" chapter). No internal structure or organization can be recognized. Dip angles of elongate clasts >0.5 mm in length were measured from split core surfaces; clasts do not appear to be systematically oriented (Fig. F8). Matrix grain size varies from silty mud (10%-20% sand, 50%-75% silt, and 10%-40% clay) to sandy silty clay (as much as 30% sand), and clast content varies from clast-rich (>20% gravel) to -poor (10%-20%). The maximum clast size recovered intact in matrix was 6 cm, but it is limited by the throat diameter of the drill bit. The largest boulder drilled was at least 50 cm in diameter; the presence in some poorly recovered intervals of several freshly fractured pieces of the same lithology also indicates large boulders. Clast lithologies include volcanics (basalt, volcaniclastics, andesite, and rhyolite) and plutonic igneous rocks (mafic, granite, and diorite) that are all derived from local Antarctic Peninsula sources (Fig. F9; e.g., Pope and Anderson, 1992). The mineralogy of the sand and silt fraction is dominantly quartz (30%-80%) and feldspar (5%-20%), with minor lithic fragments, mica, and hornblende. Trace amounts of tephra are present in Core 178-1097A-13R, and opaque micronodules (manganese?) of as much as 2% are present in massive diamict facies in Cores 178-1097A-10R and 12R. Matrix color varies from dark gray (5Y 3/1) to green (5GY 3/1). Diamicts are not cemented except for Core 178-1097A-40R, which is indurated and therefore described as diamictite. Massive diamict in Cores 178-1097A-16R and 17R show a horizontal fissility marked by very thin silt laminae (<2 mm thick) (Fig. F7E).

With a single exception (Core 178-1097A-44R; see below), massive diamict and diamictite facies recovered at Site 1097 contain large amounts of reworked and abraded diatoms, foraminifers, and sponge spicules (see "Biostratigraphy"). In Core 178-1097A-44R, massive diamict is transitional to bioturbated sandy mud with dropstones and shows a poorly defined mottling consisting of patches of lighter colored silt-rich mud (Fig. F6). Sand and granules are concentrated in these mottles, which suggests concentration by burrowing organisms. In contrast, diamict of Core 178-1097A-44R contains well-preserved marine microfossils (see "Biostratigraphy").

These facies show internal structure created by size sorting of clasts and account for 10% (~5 m) of recovered sediments at Site 1097. Beds are between 20 and 50 cm thick and show steep dips and deformation. Graded diamict facies (Facies Dmg; Fig. F6) identifies those beds that show either an upward decrease in clast size (normal grading) or upward increase (inverse grading). Core 178-1097A-27R shows a normally graded diamict bed as much as 40 cm thick (Fig. F6) capped by burrowed muds containing dropstones, well-preserved sponge spicules, and well-preserved marine microfossils. An overlying inverse to normal graded diamict bed rests with a sharp erosive contact on underlying muds and contains a rip-up clast of altered tephra (Fig. F10B, F10C).

Stratified diamict (Facies Dms; Fig. F6) shows slight and nonsystematic variation in matrix texture and clast content, giving rise to weak lamination and crude bedding (e.g., Core 178-1097A-25R; Fig. F11A, F11B). These facies are interbedded with thin beds of mud (as much as 10 cm thick) that contain in situ marine microfauna and flora.

Mud facies account for 25% of recovered sediment at Site 1097. Good sections through muds were recovered in Cores 178-1097A-34R and 36R (Figs. F6, F12). Muds are either weakly laminated or massive and range from silty clay (e.g., Core 178-1097A-36R; 10% sand, 40%-50% silt, and 30%-50% clay) to clayey silt (e.g., Core 178-1097A-34R; 5% sand, 25%-69% silt, and 30%-75% clay). Mud facies contain dropstones and shell fragments and are bioturbated. In Cores 178-1097A-34R to 36R, the content of gravel-sized, ice-rafted clasts in several thin (<10 cm) intervals is sufficiently high (<10%) to warrant description as diamict (e.g., intervals 178-1097A-34R-3, 122-100 cm; 35R-1, 140-150 cm; 36R-2, 80-90 cm; and 36R-3, 15-25 cm; Fig. F6). In Core 178-1097A-34R, laminated and bioturbated muds show broad folds as much as 80 cm in amplitude that are too large to be related to drilling disturbance (Fig. F6).

Interpretation of massive diamict is not straightforward and relies on contextual information from associated facies and the nature of bed contacts. These data are limited at Site 1097. Massive, matrix-supported diamicts at Site 1097 lack any coherent in situ assemblage of marine microfauna and flora but contain large amounts of abraded and fragmented material. The lack of any in situ microfossils, internal stratification, and bioturbation appears to preclude an origin by the rainout of suspended sediment and ice-rafted debris. A debris-flow (debrite) origin is possible, although diamicts are unstructured and do not show the chaotic bedding or flow fabric seen in debrites at Site 1103 (see "Lithostratigraphy"  in the "Shelf Transect [Sites 1100, 1102, and 1103]" chapter). Massive diamicts at Site 1097 are probably of subglacial origin, given the presence of reworked and abraded marine microfossils. Diamicts are likely to have been deposited as deformation till where marine sediment has been reworked subglacially as a "soft" or deforming bed (e.g., Boulton, 1990; Fig. F13). Subglacial incorporation of pre-existing sediment has been widely invoked in explaining massive diamict with reworked marine microfossils on glaciated continental surfaces and shelves (e.g., Boulton, 1996; Clark et al., 1996). Fissile and weakly stratified diamict found as thin (30 cm) intervals within massive diamict in Cores 178-1097A-16R and 17R (e.g., Fig. F7E) may record subglacial shear or limited sorting by subglacial water; on the other hand, such fissility could be the result of unloading and swelling after coring.

Contextual information is available for massive diamicts in Cores 178-1097A-36R, 35R, and 44R. These diamicts lack sharp bed bases and tops and are transitional to bioturbated muddy sands with ice-rafted clasts (Facies Fmd; Fig. F6). They also contain intact microfaunal and floral assemblages indicating deposition in shallow water (100-300 m) (see "Biostratigraphy"). Burrows with poorly defined walls are also present, indicating soft sediment at the time of burrowing. Because of these characteristics, massive diamicts in Core 178-1097A-44R are interpreted as rainout facies produced by suspension settling of mud with coarser material rafted in by floating ice. Variation in the flux of ice-rafted debris through time gives rise to transitions from mud with sufficient numbers of clasts to be identified as diamict (Facies Dmm), to mud with isolated ice-rafted clasts (Facies Fmd).

Stratified diamict and graded diamict (Facies Dms and Dmg) form distinct beds associated with marine muds containing well-preserved marine microfauna and flora (e.g., Core 178-1097A-25R; Figs. F6, F11, F12). These diamict facies are interpreted as sediment gravity flows (debrites). Interbeds of mud in Core 178-1097A-25R, for example, probably record pauses between debris flows, which allow mud accumulation. Graded facies (Facies Dmg; Fig. F6) result from turbulent downslope flow resulting in the segregation of different size fractions (see Walker, 1992). An ice-proximal marine setting for stratified and graded diamicts is suggested by marine microfossils present in diamict and associated muds. The microfossils denote deposition within a few kilometers of a glacier terminus in an area of high sedimentation and reduced salinity caused by meltwater input (see "Biostratigraphy"). Given this setting, downslope resedimentation is likely to have been near to the ice front where subglacial debris (deformation till; see above) was released and moved downslope by mass flow (Fig. F13). Consequently, stratified and graded diamict facies are recognized as glaciomarine in origin.

Massive muds accumulate in a wide range of settings on glaciated continental shelves. The most common depositional process is suspension settling of fines carried by meltwater plumes. Tidal and bottom-current interaction with plumes generates laminated muds (e.g., Griffith and Anderson, 1989; Domack, 1990). Mud recovered in Cores 178-1097A-36R and 44R contains marine biota indicating water depths between 100 and 300 m and salinity characteristic of an open shelf distal to ice margins (see "Biostratigraphy"). Muds are bioturbated and contain variable quantities of ice-rafted debris including thin intervals of diamict (Fig. F6). Core 178-1097A-44R contains a thick (3 m) section of rainout diamict (see above) produced by ice rafting and suspension settling of mud (see above and Fig. F13). The deformed mud in Core 178-1097A-34R (Fig. F6) may result from iceberg scour of the substrate or downslope slumping.

Because of an inability to collect continuous stratigraphic information at Site 1097, recovered facies provide a partial picture of glacial and glacially influenced deposition on the continental shelf. Inferred environments range from subglacial through glaciomarine to open-marine shallow shelf (Fig. F13).

The lithostratigraphy at Site 1097 can be divided into two thick intervals with differing characteristics (Fig. F5). The lower, below ~180 mbsf, shows an alternation between subglacial conditions, when grounded ice extended across the shelf, and glaciomarine conditions, when the shelf was partially free of grounded ice and Site 1097 lay seaward of the ice margin. Biofacies data from the upper part of Site 1097, above ~155 m, indicate that subglacial conditions were dominant throughout this upper interval. This possibility suggests a major change in style of glacial deposition along the Antarctic continental margin (see "Seismic Stratigraphy"). This may have implications for changing ice volumes through time.

Within the glaciomarine intervals at Site 1097, weakly bioturbated, massive, and laminated muds with high ice-rafted debris content suggest influxes of meltwater, suspended sediment, and icebergs to Site 1097 during early Pliocene(?) time (Fig. F5). These deposits suggest a subpolar or temperate glacier regime and the release of meltwater plumes derived from subglacial meltwaters that intercept easily erodable marine sediment or deformation till. The biogenic component of the mud within these glaciomarine intervals is generally <10%, which is much less than the biogenic component of sediments accumulating on the modern Antarctic Peninsula shelf. For example, many of the cores described by Pope and Anderson (1992) are capped by diatomaceous muds, and the upper part of cores analyzed by Pudsey et al. (1994) contain from 30% to 60% diatoms. The low values recorded in muds at Site 1097 could reflect an influx of fine-grained, meltwater-derived, terrigenous sediment as well as environmental factors such as reduced salinity.

Intervals of expanded glaciation, when grounded ice reached the shelf edge, are recorded at Site 1097 by massive diamicts that are interpreted as deformation tills. During glacier expansion, fine-grained marine sediments within fjords and inner shelf basins were probably reworked subglacially and transported below ice streams to the outer shelf. The recovery of deformation till at Site 1097, if correctly interpreted, is important because these are the first such sediments to be recovered from the Antarctic continental margin. Work on Pleistocene mid-latitude ice sheets has shown that the ice-sheet margins rested on "soft" deformable beds giving rise to a low surface profile and fast ice flow ("streaming"). It has been suggested that such conditions existed on the Antarctic Peninsula shelf (e.g., Pudsey et al., 1994; Bart and Anderson, 1995), and there has been much discussion of depositional mechanisms (Alley et al., 1989; Vanneste and Larter, 1995). To date, such facies have not yet been described anywhere along the continental margin of Antarctica. The occurrence of drumlin and flute fields on the shelf adjacent to Site 1097 (Pudsey et al., 1994) may be further evidence of "soft bed" conditions below ice sheets extending onto the shelf (e.g., Boulton and Hindmarsh, 1987; Boyce and Eyles, 1991).