The Antarctic Dry Valleys represent a unique environment where it is possible to study dry permafrost overlaying an ice-rich permafrost. In this paper, two opposing mechanisms for ice table stability in University Valley are addressed: i) diffusive recharge via thin seasonal snow deposits andii) desiccation via salt deposits in the upper soil column. A high-resolution time-marching soil and snow model was constructed and applied to University Valley, driven by meteorological station atmospheric measurements. It was found that periodic thin surficial snow deposits (observed in University Valley) are capable of drastically slowing (if not completely eliminating) the underlying ice table ablation. The effects of NaCl, CaCl and perchlorate deposits were then modelled. Unlike the snow cover, however, the presence of salt in the soil surface (but no periodic snow) results in a slight increase in the ice table recession rate, due to the hygroscopic effects of salt sequestering vapour from the ice table below. Near-surface pore ice frequently forms when large amounts of salt are present in the soil due to the suppression of the saturation vapour pressure. Implications for Mars high latitudes are discussed.

Ecosystem services are the benefits that mankind obtains from natural ecosystems. Here we identify the key services provided by the Southern Ocean. These include provisioning of fishery products, nutrient cycling, climate regulation and the maintenance of biodiversity, with associated cultural and aesthetic benefits. Potential catch limits for Antarctic krill ( Dana) alone are equivalent to 11% of current global marine fisheries landings. We also examine the extent to which decision-making within the Antarctic Treaty System (ATS) considers trade-offs between ecosystem services, using the management of the Antarctic krill fishery as a case study. Management of this fishery considers a three-way trade-off between fisheries performance, the status of the krill stock and that of predator populations. However, there is a paucity of information on how well these components represent other ecosystem services that might be degraded as a result of fishing. There is also a lack of information on how beneficiaries value these ecosystem services. A formal ecosystem assessment would help to address these knowledge gaps. It could also help to harmonize decision-making across the ATS and promote global recognition of Southern Ocean ecosystem services by providing a standard inventory of the relevant ecosystem services and their value to beneficiaries.

Measurements of dissolved N2, O2, Ar, CO2, and CH4 were made in perennially ice-covered Lake Hoare. Results confirm previous reports that O2 concentrations in the upper water column exceed atmospheric equilibrium and that N2 and Ar are supersaturated throughout the water column. The mean supersaturation of N2 was found to be 2.0 (+/- 0.37) and Ar was 3.8 (+/- 1.1). The ratios of N2/Ar (20.3 +/- 13.8), and O2/Ar (22.5 +/- 4.0) at the ice-water interface are consistent with those previously measured, suggesting that bubble formation is the main process for removing gas from the lake. However, the saturations of N2 and Ar greatly exceed those previously predicted for degassing by bubble formation only at the ice-water interface. The data support the hypothesis that removal of gas by bubbles occurs in the water column to a depth of 11 m in Lake Hoare. CO2 concentration increases from near zero at the ice-water interface to 80-100 times saturation at and below the chemocline at c. 28 m. There is considerable variability in the gas concentrations throughout the water column; samples separated in depth by one metre may vary by more than 50% in gas content. It is likely that this phenomenon results from the lack of turbulent mixing in the water column. Methane (c. 2 micrograms l-1) was detected below the chemocline and immediately above the sediment/water interface at a depth of 30 m. Samples from lakes Vanda, Joyce, and Miers, also show supersaturations of O2, N2, and Ar at levels similar to levels found in Lake Hoare.

Year-round temperature measurements at 1600 m elevation during 1994 in the Asgard Range Antarctica, indicate that the mean annual frost point of the ice-cemented ground, 25 cm below the surface, is -21.7 +/- 0.2 degrees C and the mean annual frost point of the atmosphere is -27.5 +/- 1.0 degrees C. The corresponding mean annual temperatures are -24.9 degrees C and -23.3 degrees C. These results imply that there is a net flux of water vapour from the ice to the atmosphere resulting in a recession of the ice-cemented ground by about 0.4-0.6 mm yr-1. The level of the ice-cemented permafrost is about 12 cm below the level of dry permafrost. The summer air temperatures would have to increase about 7 degrees C for thawing temperatures to just reach the top of the subsurface ice. Either subsurface ice at this location is evaporating over time or there are sporadic processes that recharge the ice and maintain equilibrium over long timescales.

A process-oriented study was carried out in White Smoke lake, Bunger Hills, East Antarctica, a perennially ice-covered (1.8 to 2.8 m thick) epishelf (tidally-forced) lake. The lake water has a low conductivity and is relatively well mixed. Sediments are transferred from the adjacent glacier to the lake when glacier ice surrounding the sediment is sublimated at the surface and replaced by accumulating ice from below. The lake bottom at the west end of the lake is mostly rocky with a scant sediment cover. The east end contains a thick sediment profile. Grain size and delta 13C increase with sediment depth, indicating a more proximal glacier in the past. Sedimentary 210Pb and 137Cs signals are exceptionally strong, probably a result of the focusing effect of the large glacial catchment area. The post-bomb and pre-bomb radiocarbon reservoirs are c. 725 14C yr and c. 1950 14C yr, respectively. Radiocarbon dating indicates that the east end of the lake is >3 ka BP, while photographic evidence and the absence of sediment cover indicate that the west end has formed only over the last century. Our results indicate that the southern ice edge of Bunger Hills has been relatively stable with only minor fluctuations (on the scale of hundreds of metres) over the last 3000 years.