Knowledge of the glacial history of the Antarctic Peninsula region, including the South Shetland Islands, has become much enhanced of late. Recent years have seen an increasing number of expeditions of various nationalities studying glacial and glaciomarine deposits on the land and the sea floor (see overviews and discussion in Björck et al. 1996, Lopez Martinez et al. 1997, Bentley and Anderson 1998, lngólfsson et al. 1998). One conclusion reached by these recent studies, based on geological data from the currently ice-free land areas in the region (Hjort et al. 1998), is that the on-the-spot Antarctic glacial and climatic evolution since the end of the Last Glacial Maximum (culminating same 18 000 years ago) has not always been synchronous with developments in the Northern Hemisphere, and only to a limited extent over the last ca. 100 000 years. A more or less total north-south synchronicity has usually been advocated by the ice-core community (Ciais et al. 1994), basing their conclusions mainly on oxygen isotope data obtained from high elevation ice cores from East Antarctica and Greenland. The simplest explanation for this geologically documented discrepancy between the hemispheres seems to be that the solar insolation curve of, for example, the Antarctic Peninsula, lags that of important parts of the northern hemisphere by about 5 000 years. This lag is due to the so-called Milankovich effect. The similar climatic curves derived from both the Antarctic and Greenland ice-cores may thus mirror same overruling global signal, rather than describing the actual ground-truth situation along the coasts of Antarctica itself.

One line of evidence for such hemispheric asynchronies is that the climatic optimum, the warmest period within the present (Holocene) interglacial, seems to have culminated as recently as 4500-3000 calendar years ago (equivalent to 4000-3000 radiocarbon years before the present; only calendar years are used below in both the Antarctic Peninsula region (Björck et al. 1996) and the Ross Sea area (Baroni and Orombelli 1994). This fits well with the peak in solar insolation for these areas and agrees with most other local geological information (see overview in lngólfsson et al. 1998). However, it does not agree with certain results from far south on the Antarctic Peninsula. Evidence from Alexander Island, between 69°-72°S off the west coast of the peninsula, seemingly told a different story from that told further north (Clapperton and Sugden 1982). Radiocarbon dated barnacle shells from ice-pushed marine sediments in moraines along the George VI Ice Shelf, which fills the sound between Alexander Island and the mainland, indicated only one period with seasonally open water in the sound, around 7 000 years ago. This is some 2 500 years earlier than the warmest and most ice-free period further north. The question was whether this indicated a true difference in climatic development and in the timing of the climatic optimum between the northern and southern parts of the Antarctic Peninsula, or whether the difference was simply an artefact of the sampling, as the Alexander Island Holocene chronology was in fact based on only one sample.

To solve this problem a small expedition went in to the field in January-February 2000, with representatives from both the Lund University and University of Gothenburg, which had previously worked with these problems further north along the Antarctic Peninsula, and from the United Kingdom-group which did the original study on Alexander Island.

Field-work

The field -work on Alexander Island took place between 20 January and 3 February. It was based on British logistics (e.g. transport in the field with Twin Otters), and worked out of the British Antarctic Survey’s main base at Rothera On Adelaide Island, which is reached by British Antarctic Survey’s flights from Port Stanley on the Falkland Islands. The geological work was done at two main localities: at Fossil Bluff on 20-26 January, with a one day excursion on 24 January to the Batterbee Mountains on the mainland, and at Two Step Cliffs between 27 January and 3 February.

The work at Fossil Bluff and in the Batterbee Mountains mostly concerned former glacial overriding of the landscape, the elevation reached by the glaciers and the glacial chronology. The latter problem was approached by collecting samples for exposure dating (measuring gradually accumulated cosmogenically derived isotopes like  and  on deglaciated rock surfaces and glacially transported boulders). The ice-shelf moraines, created by the movement of the shelf-ice from the east on to the coast of Alexander Island (Sugden and Clapperton 1981), were not well exposed here and did not contain any marine shells.

Much of the work at Two Step Cliffs also concerned the former glacial overriding, with the collecting of samples for exposure dating, etc. But here the ice shelf’s push moraines were well exposed, largely consisting of marine sediments bulldozed from the bottoms off the coast. The moraines were studied at two places–the Mars and Ares Oases. The sites studied by Clapperton and Sugden (1982) in the late 1970s were revisited and re-sampled by us, and several new sites with ice-pushed shell-bearing marine sediments were found. The contents represent former benthic (bottom dwelling) communities in George VI Sound (bivalves, gastropods, serpulid worms, barnacles, echinoderms), which were sampled for absolute (radiocarbon) and relative (amino acid racemization) dating.

Excellent weather throughout supported the work, which therefore could be conducted in record time.

The results

Glacial history

Glacial geomorphology in combination with the occurrence at various altitudes of erratic (transported from a great distance) boulders of various origins both at Fossil Bluff and Two Step Cliffs show that the current nunataks along the east coast of Alexander Island were earlier overridden by glaciers which came both from the mainland and from central Alexander Island. These glaciers probably, to varying degrees, contributed to an ice stream in George VI Sound (Payne et al. 1989). The glacial events are not yet dated, but an exposure age program is underway in the United Kingdom.

Ice-shelf history

The ice-shelf moraines in the Two Step Cliffs area were thoroughly searched for shell-bearing displaced marine sediments, and several new (relative to the original study by Clapperton and Sugden 1982) exposures with such material were found.

The shells from the Ares Oasis-including bivalves like Laternula elliptica, Adamussium colbecki and Hiatella sp., a Nacinna gastropod, serpulid worms and Echinoderms­ – are being dated, and results obtained so far indicate infinite radiocarbon (AMS) ages, i.e. the shells are >30-40,000 years old. Clapperton and Sugden (1982) had already described and dated Hiatella shells from a glacially displaced marine sediment from a hillside above the Ares Oasis at ca. 150 m; we re-sampled the site. These shells also have infinite radiocarbon ages, but in addition the amino acid ratios (the measured racemisation, i.e. the degree Of peptide bond disintegration since death) indicates an age of perhaps several hundred thousand years. Most probably an ice stream in George VI Sound brought them from the bottom of the sea to where they now rest.

The shells found in the ice-shelf moraines at Mars Oasis all belong to the barnacle Bathylasma corolliforme and date from the present (Holocene) interglacial. The two radiocarbon dates so far obtained for our material indicate ages within a very brief (300-600 years) interval ca. 7 000 years ago. This is the same age as Clapperton and Sugden (1982) obtained for shells from the same location, and indicates that George VI Sound has been seasonally ice-free-that is without a glacier or ice-shelf in it – for only a short period during the Holocene.

Preliminary conclusions

Our results, combined with those of our predecessors, have so far revealed several important matters:

  1. George VI Sound, currently filled by a floating ice-shelf, was earlier – on one or several occasions – instead filled by a true glacier, most likely an ice stream moving along the sound (Payne et al. 1989), filling the basin to the bottom and overriding most of the present nunataks.
  2. The present situation with a floating ice-shelf, fed both by glaciers from the mainland (the dominating ones) and from Alexander Island, and by precipitation directly on to its surface, represents an intermediate situation. It is halfway between full glaciation and more strictly marine conditions with seasonally open water allowing a diverse benthic ecosystem in the sound.
  3. The shells in the moraines indicate that there were times when the sound was without glaciers or large ice shelves, indicating relatively seen warm interglacial or interstadial periods with seasonally open water. Radiocarbon dating and amino acid analyses help us to date these episodes. The oldest one documented probably dates from several hundred thousand years ago, whereas the most recent occurred only ca. 7 000 years ago.

Our present working hypothesis is that during the Last Glacial Maximum, George VI Sound was much more heavily glaciated than today, although the exact thickness and extension of the ice is still not well known, exposure datings pending. The sound was not deglaciated until well in to the Holocene interglacial. This de glaciation was probably an effect both of gradually rising global sea levels (mainly resulting from melting ice sheets in the Northern Hemisphere) which floated coastal glaciers and ice streams, temporarily converting them in to ice shelves (Kennedy and Anderson 1989), and of increasing interglacial warming which melted the ice. Around 7000 years ago, George VI Sound then experienced a brief marine period with seasonally open water, which perhaps lasted only for 300-600 years. Soon thereafter the glaciers expanded again-at least enough to re-create an ice shelf in the sound. This may have already happened around 6000 years ago, contemporaneously with a period of cooling which led to the brief Brandy Bay glacial re-advance documented further north on the Antarctic Peninsula (Hjort et al. 1997). Perhaps the George VI Ice Shelf has remained intact since then. It is, after all, situated far to the south, in a markedly colder environment than one encounters further north along the Antarctic Peninsula, where the climatic optimum with the warmest weather and the smallest glaciers occurred as recently as 4 500-3 000 years ago.

South of Two Step Cliffs lie the truly barren lands, like the Coal Nunatak and a few similar desolate places, surrounded by glaciers in all directions. From there not even rumours of workable Quaternary deposits have ever been heard! Thus we may finally have reached the end of the line?