Site locations are shown in Figures F5 and F11. Three sites (1095, 1096, and 1101) were occupied on sediment drifts on the continental rise, and four sites were on the outer continental shelf (1100, 1102, and 1103 on the "shelf transect" of a depositional lobe and 1097 in an interlobe area). Two short sites (1098 and 1099) were drilled inshore, at Palmer Deep. For a brief summary of drilling at these sites, see Table T1, in the "Leg 178 Summary" chapter.

Sites 1095 and 1096 (prospectus sites APRIS-02A and 01A) were located on Drift 7 and were intended to be complementary. The main site (1096), close to the crest of Drift 7, was intended to sample the expanded upper part of the sedimentary section within the drift. The lower part of the drift section was more accessible at Site 1095, in a more distal position on the northwest flank, where the upper section was known from seismic reflection profiles to be significantly thinner. The intention was to double sample the upper section at each site, using the advanced hydraulic piston corer (APC), then drill with the extended core barrel (XCB) to refusal or to target depths and log the deep holes. Together, these two sites were intended to cover the last 6 to 10 m.y., the likely extent of Antarctic Peninsula glacial history.

The drifts are assumed to be formed largely of the hemipelagic, bottom-current-transported fine fraction of the unstable component of glacially transported sediment deposited on the upper continental slope (Fig. F3). The nature and similarity of seismic reflection profiles across the drifts support this view. The nature and similarity of sediments from short piston cores also support the conclusion that the unstable component of upper-slope deposition has a short residence time compared to a glacial cycle. The sites should therefore contain a high-resolution record of ice-sheet grounding to the continental shelf edge. There should also be sufficient biogenic interbeds (mainly during interglacials when sea ice cover was less) for biostratigraphic control and a sufficient fine-grained terrigenous component to assure a stable magnetic remanence. However, the record will be indirect: it will be necessary to correlate events on the rise and on the shelf in order to interpret the rise-drift record in terms of the state of continental glaciation. An additional objective at the rise sites was to try to identify any effects of the Pliocene Eltanin impact. A tsunami-like disturbance of the seabed, especially if accompanied by air-fall redistribution of both asteroid fragments and Eocene-to-Pliocene biogenic sediments, could be detectable at an elevation 1 km above the surrounding ocean floor, as the crest of the drift would have been.

Site 1101 was an additional site, on Drift 4, that became possible because environmental limitations on drilling on the continental shelf made time available. It was drilled by APC and XCB to a little more than 200 m to resample the upper part of the section sampled at Sites 1095 and 1096. Certain features of those sites could have been heavily influenced by local events (e.g., the development of an additional channel system). We therefore thought it necessary to distinguish between local and regional developments before trying to correlate events on the rise and on the shelf in search of paleoclimate inference. Site 1101 lay on a different drift, some distance along the margin from Drift 7 and within GLORIA coverage. It was a new site, roughly halfway between existing alternate sites, chosen on the basis of results from Sites 1095 and 1096.

Leg planning included four sites on a dip transect along the axis of Lobe 1, off Anvers Island (prospectus sites APSHE-01 through -04). There was also an additional site between Lobes 3 and 4 (APSHE-05) where the lower part of the section (Sequence S3) was less deeply buried and thus accessible to drilling. Because of time lost waiting for ship heave to drop below the specified 2-m limit, it was only possible to access the interlobe site (1097) and three shelf transect sites (1100, 1102, and 1103). The aim of shelf drilling was to date the major changes in the geometry of the glacial prograded wedge; these changes were assumed to mark major shifts in the development of the ice sheet. The main shelf transect lay along the line of Figure F12. It provided the opportunity to date the main changes in the geometry of the two known glacial sequence groups (S1 and S2) and to examine the S3/S2 boundary (which may or may not have coincided with the onset of Antarctic Peninsula glaciation). Sites were planned (1) at the shelf edge (APSHE-01) to sample slope foresets; (2) with two close together (APSHE-02 and 03) to sample the beginning and end of the process that eroded the topsets and truncated the foresets of S2; and (3) with one (APSHE-04) to sample the S3/S2 boundary, farther inshore. In fact, the ship visited the shelf edge (Site 1100) and the inner of the two S2/S1 sites (Site 1100), but to little effect because of ship heave. The inner end of the shelf transect at an alternate site (APSHE-10) was then drilled. At this location S2 is pinched out completely, and it is possible to drill through a thinner sequence of S1 topsets into the lower part of S3 (Site 1103). Earlier, it had been possible to drill the interlobe site (1097).

Sites 1098 and 1099 (prospectus sites APSHE-13 and 15) are located in Basins I and III of Palmer Deep, on the inner continental shelf directly south of Anvers Island (Leventer et al., 1996). Three linked basins contain an ultrahigh-resolution Holocene record of Antarctic Peninsula climate. Short piston cores from Basin I show a pronounced 200- to 300-yr periodicity in paleoproductivity that is also seen in some Antarctic Peninsula fjords. The expanded, presumed pelagic section may be compared with recently acquired records from low and intermediate latitudes (Santa Barbara Basin, Saanich Inlet, and Cariaco Basin) and ice-core records from Greenland and Antarctica to examine decadal and millennial variability on a global scale. This record may also provide opportunities to examine magnetic secular variation and, for the inshore environment, the time variability of the ¹⁴C "reservoir effect," which is large but uncertain for waters south of the Polar Front. Site 1098 in Basin I, presumed to contain a much lower component of locally derived turbidites, is triple APC-cored to basement. Site 1099 in Basin III is a single APC core to more than twice the depth of 1098, in an effort to sample older sediments, to below a prominent high-amplitude reflector that was assumed to mark the beginning of the high-resolution biogenic section.