Dry permafrost on Earth is unique to the Antarctic and is found in the upper elevations of the McMurdo Dry Valleys. Despite its widespread presence in the Dry Valleys, the factors that control the distribution of dry permafrost and the ice-cemented ground below it are poorly understood. Here I show, by a combination of theoretical analysis and field observations, that the recurrence of snow can explain the depth of dry permafrost and the location of ice-cemented ground in Antarctica. For data from Linnaeus Terrace at 1600–1650 m elevation in Upper Wright Valley a recurrence intervals of about two years explains the presence of ground ice at 25 cm depth, under 12.5 cm of dry permafrost. Snow recurrence periods longer than 10 years would create only dry permafrost at this site. The snow gradient in University Valley resulting from the windblown snow from the polar plateau creates a corresponding gradient in the depth to ice-cemented ground. On the floor of Beacon Valley, the presence of dry permafrost without underlying ice-cemented ground indicates snow recurrence intervals of more than 10 years and implies that the ancient massive ice in this valley is not stable. Snow recurrence may also set the depth to ground ice on Mars.

In situ cosmogenic 10Be exposure dating, radiocarbon determinations, salt and sediment geochemistry, and rock weathering observations indicate that parts of Larsemann Hills, East Antarctica have been subaerially exposed throughout much of the last glacial cycle, with the last glaciation occurring prior to 100 ka bp. Salt-enhanced subaerial weathering, coupled with a paucity of glacial erratics, made exposure age dating challenging. Rapid subaerial surface lowering in some places means that some exposure ages may underestimate the true age of deglaciation. Despite this uncertainty, the data are consistent with the absence of overriding by a thick ice sheet during the Last Glacial Maximum ∼20–18 ka bp.

During the late 19th and early 20th centuries, when blubber oil fuelled house lamps, the king penguin population at Macquarie Island was reduced from two very large (perhaps hundreds of thousands of birds) colonies to about 3000 birds. One colony, located on the isthmus when the island was discovered in 1810, was extinct by 1894 and it took about 100 years for king penguins to re-establish a viable breeding population there. Here we document this recovery. The first eggs laid at Gadget Gully on the isthmus were recorded in late February 1995 but in subsequent years egg laying took place earlier between November and February (this temporal discontinuity is a consequence of king penguin breeding behaviour). The first chick was hatched in April 1995 but the first fledging was not raised until the following breeding season in October 1996. The colony increased on average 66% per annum in the five years between 1995 and 2000. King penguins appear resilient to catastrophic population reductions, and as the island's population increases, it is likely that other previously abandoned breeding sites will be reoccupied.

Passive dispersal has traditionally formed a fundamental component of biogeographical theories of the origin of the fauna that occupy the ice-free habitats of mainland Antarctica. But in the context of an emerging picture of endemism for many Antarctic terrestrial invertebrates, is there still a place for such stochastic processes in Antarctic biogeography? The case of the Antarctic fairy shrimp, Branchinecta gaini Daday 1910, may provide an answer - or, at least, an important exception to the rule. Although passive dispersal is certainly a stochastic and contingent phenomenon in Antarctica, the occurrence of B. gaini on the Antarctic Peninsula can only be explained satisfactorily by resort to this explanation. It is, at present, probably the best example of an Antarctic invertebrate with a biogeographic signature of passive - in particular, zoophoretic - dispersal.

Zoophycos is a complex three dimensional trace fossil that is abundant in deep ocean sediments worldwide, but has not been described previously from Cenozoic continental margin deposits of Antarctica. In the ANDRILL 1B core drilled through the north-west McMurdo ice shelf, Zoophycos occurs in a 17 m thick unit of interglacial sediments bounded above and below by glacial surfaces of erosion. This unit was deposited during the transition from the relatively warm Early Pliocene characterized by productive open waters to the cooler Late Pliocene with fluctuating subpolar ice sheets. Globally, Late Cenozoic Zoophycos are most abundant at great depths (> 1000 m), and where sedimentation rates and TOC levels are low; the Zoophycos producer, probably a worm-like animal, was (is) a slow colonizer. Application of these preferences to the ANDRILL 1B core indicates that the Zoophycos-bearing unit was deposited episodically, with sufficient time between events to allow for the slow processes of colonization and construction. The foray of Zoophycos producer into the relatively shallow ANDRILL 1B depths (200–1000 m) during the Pliocene documents “emergence” of benthic animals, supporting suggestions that the unique modern Antarctic and Southern Ocean faunas result from both “emergence” and “submergence” during the Cenozoic.

The Weddell Sea is one of the key regions of the Southern Ocean with respect to climate as most of the Antarctic Bottom Water (AABW) that occupies the world ocean deepest layers is likely to originate from this region. This study applies the Optimum Multiparameter water mass analysis to the Weddell deep waters in order to investigate their distribution and variability. The dataset used is based on the WOCE repeat sections in the area (SR04 and A12) from 1984 to 1998. The mean water mass distribution is consistent with previous knowledge of the region, along with high interannual variability. Regarding the temporal variability, it seems that the years of maximum Weddell Sea Deep Water (WSDW) contribution correspond to the lowest levels of Weddell Sea Bottom Water (WSBW), and vice versa. In order to identify possible forcing mechanisms for such variability, the water mass temporal anomalies were compared with oceanic and atmospheric modes of variability in that region such as the Southern Annular Mode (SAM). An apparent correlation between the SAM index temporal gradients and WSBW anomalies indicate that the Weddell Sea export of dense waters to the world ocean may be linked to that index on several time scales.

Ixodes uriae White is the only tick species found in Antarctica, and in our research area at Palmer Station it feeds predominantly on Adélie penguins. At the beginning of the summer the ticks leave their off-host aggregation sites, feed on penguins, then subsequently return to their off-host habitats, where they remain until the next summer (Benoit et al. 2007). These ticks have been implicated as a competent vector for Borrelia spirochetes (Olsen et al. 1993) and tick-borne viruses (Nuttall 1984), and are thought to impact development, cause anaemia, alter the thermoregulation, and even lead to the death of penguins (Gauthier-Clerc et al. 1998, Mangin et al. 2003). In this report, we note a striking increase in the number of fed ticks found near Adélie penguin rookeries in 2007 compared to 2006, suggesting that the longer and warmer summer of 2007 may have permitted more ticks to find their hosts. High levels of parasitism could be detrimental to the already declining Adélie penguin populations near Palmer Station (Fraser & Patterson 1997).

This study estimates energy and mass balance at the edge of the Antarctic ice sheet close to a non-glaciated area. An automatic weather station was installed on the ice sheet, near an ice free area of Schirmacher Oasis in Dronning Maud Land, East Antarctica. Hourly snow-meteorological parameters were recorded and observed during the summer of the year 2007–08. Hourly radiative and turbulent energy fluxes were estimated at the ice surface. An ultrasonic sensor was used to measure accumulation or ablation at the glacier surface. Ground Penetrating Radar was also used to measure the changes in ice thickness at the observation point. The net radiative flux was the main heat source and the latent heat flux was the main heat sink for the ice sheet with seasonal average values of 98 W m-2 and -86.7 W m-2 respectively. There was a high ablation rate for the ice sheet near the non-glaciated area with a seasonal mean of 0.0172 m w.e. per day. Over the period 10 November 2007–7 February 2008 the mass balance was -1.53 m w.e. Good correlation (r2 = 0.97) was observed between estimated and observed hourly ablation of the glacier. Sublimation and melt processes contributed 16.5% and 83.5% respectively to the net summer ablation.

Deception Island is the largest volcano in the actively extending Bransfield Basin, a marginal basin situated behind the extinct South Shetland Islands arc. Deception Island has been well studied but its submerged flanks have not. A multibeam bathymetry survey was conducted around the island in 2005. Data from the flooded caldera show no evidence for recent localized resurgence. The gently-sloped bottom of the caldera basin is consistent with either a broad zone of resurgence on its east side associated with trap door deformation or with higher rates of sediment supply from the east side of the island. Around the island, numerous tectonic and volcanic features on the volcano's east and west flanks are nearly all aligned with the regional strike (~060°) of the Bransfield rift and there is very little evidence for the other fault populations that have been identified on the island. We infer that models that link the ongoing tectonic development of Deception Island to complex regional tectonics are less likely than models in which the dominant regional extension in Bransfield Strait is modulated by the local effects of caldera collapse and possibly a small right-lateral transfer zone offsetting the primary extension axes in the Central and Western Bransfield Basins.

Sediment trap arrays were deployed in Brialmont Cove and Andvord Bay, eastern Gerlache Strait, from December 2001–March 2003. The recovered sediments (representing instantaneous deposition from the viewpoint of luminescence dating) encompass all the annual and local glaciomarine depositional processes. Magnetic susceptibility profiles were used to infer seasonality in the trap cores, and thus to select subsamples for luminescence measurements. Multi-aliquot infrared stimulated luminescence (IRSL) apparent ages were used to assess the effectiveness of ‘clock zeroing’ (by daylight) of light sensitive luminescence within fine silt polymineral samples from each trap depth. IRSL apparent ages for 24 samples indicate that the largest age-depth differences occur with the autumn season samples at both trap sites, suggesting a previously unrecognized and regional (within the Gerlache Strait) change in depositional controls in the autumn compared to other seasons. The apparent ages also indicate some differences between the fjords, and a more complex oceanographic regime at Andvord Bay than at Brialmont Cove. Dry-mass sediment fluxes varied from 0.4 to 0.7 g cm-2 yr-1, with the largest flux at Brialmont Cove (∼0.7 g cm-2 yr-1) occurring in the bottom trap, whereas at Andvord Bay, the largest flux (∼0.6 g cm-2 yr-1) occurred in the middle trap (∼45 m above seafloor).

Understanding of glacially driven sedimentary transport systems across the shelf to the slope and subsequently to deep sea sediment bodies along the Pacific continental margin of Antarctic Peninsula is crucial for interpreting ice sheet dynamics. Here we quantify slope-failure frequencies recorded in Pliocene core intervals of ODP Site 1095. We used the relationship between long-term sedimentation rate and marine carbon burial efficiency to calculate glacial or interglacial specific sedimentation rates. Using the decompacted average length of glacial-interglacial cycles it was possible to solve a set of linear equations to derive average half-periods of 61.59 and 59.77 kyr respectively for the time interval 5.8–3.2 Ma. The resulting frequency distribution of slope failures reflects short and rapid but cyclic ice advances every ∼375 years. Short retention times between slope loading and slope failure are supported by biogenic silica dissolution analyses. This study demonstrates the potential of the rise record to improve models of orbitally controlled size variations of the West Antarctic ice sheet and confirms the hypothesis of a highly dynamic ice sheet during the early Pliocene warm period.

A high resolution record of water column temperatures was measured in a coastal meltwater pond on the McMurdo Ice Shelf, Antarctica. The maximum temperature gradient measured through the water column was 35°C, with an annual temperature range of 52.1°C within the pond. For most of the year the pond shows reverse temperature stratification with the lowest temperatures measured at the surface of the pond, with the exception of brief periods of normal stratification over winter caused by regional warming events. During freezing, the freezing front propagated downwards from the pond surface, excluding major ions and releasing large amounts of latent heat, both of which had a dramatic effect on the thermal and compositional evolution of the pond. Thawing is dominated by changes in surface air temperatures and the differential absorption of solar radiation. A new conceptual model of the physical freeze-thaw process has been developed that explains the presence of an ‘ice plug’ during melting, which reduces wind-induced mixing, forms a physical barrier to chemical processes, and encourages thermal and chemical stratification. It may also explain the persistence of anoxic and hydrogen sulphide bearing basal brines in summer stratified ponds that are otherwise fully oxidized.

Except for the rich record from the Neuquen Basin (e.g. Gasparini & Fernández 2006), Jurassic southern Gondwanan marine reptiles are relatively rare. A tooth discovered in the Bean Peaks, Ellsworth Land, Antarctic Peninsula (Fig. 1) represents the southernmost, and only the second record of Jurassic marine reptiles from the Antarctic Peninsula. Comprising a single, incomplete tooth, the specimen is unable to be assigned to a species, but the paucity of Gondwanan Jurassic marine reptile material means this find adds significant palaeobiogeographical information.

New whole rock Rb-Sr and zircon U-Pb geochronological data and Sm-Nd isotopic data are presented from the central magmatic arc domain of the Antarctic Peninsula in the area of north-west Palmer Land and south-west Graham Land, Rb-Sr isochrons indicate an age of 169 ± 6 Ma for basement orthogneisses and 132 ± 9 to 71 ± 9 Ma for plutons. A U-Pb age of 183 ± 2.1 Ma, with no detectable inheritance, on zircons from an orthogneiss from Cape Berteaux provides the first reliable age for the orthogneisses, which are interpreted as metamorphosed silicic volcanic rocks, and Sm-Nd data indicate derivation in a mature volcanic arc. The age indicates they may be correlatives of the Jurassic ‘Chon Aike’ volcanism of the eastern Antarctic Peninsula. A U-Pb zircon age of 107 ± 1.7 Ma on a terrestrial volcanic sequence overlying an uncomformity strongly suggests a mid-Cretaceous age for the extensive volcanic cover of north-west Palmer Land that was previously thought to be Jurassic. The unconformity is interpreted to have been a result of compressional uplift related to the Palmer Land event. This is the first date for the event in the western part of the central magmatic arc terrane of the Antarctic Peninsula.

The oldest record of the hexanchiform sharks from the Southern Hemisphere and the second chondrichthyan report known from Carboniferous to Early Cretaceous beds in Antarctica is given. The material was collected in late Aptian rocks of the Kotick Point Formation outcropping in the western part of James Ross Island, near Antarctic Peninsula. It consists of an isolated tooth assignable to a hexanchiform different from the other described genera. The tooth shows putative plesiomorphic cusp (few cusps, no serrations) and apomorphic root characters (relatively deep, quadrangular). It could be related to a species close to the origin of Hexanchus (unknown in beds older than Cenomanian).

New stable isotope and noble gas data obtained from fumarolic and bubbling gases and hot spring waters sampled from Deception Island, Antarctica, were analysed to constrain the geochemical features of the island's active hydrothermal system and magmatism in the Bransfield back-arc basin. The 3He/4He ratios of the gases (< 9.8 × 10-6), which are slightly lower than typical MORB values, suggest that the Deception Island magma was generated in the mantle wedge of a MORB-type source but the signature was influenced by the addition of radiogenic 4He derived from subducted components in the former Phoenix Plate. The N2/He ratios of fumarolic gas are higher than those of typical mantle-derived gases suggesting that N2 was added during decomposition of sediments in the subducting slab. The δ13C values of -5 to -6‰ for CO2 also indicate degassing from a MORB-type mantle source. The H2/Ar- and SiO2 geothermometers indicate that the temperatures in the hydrothermal system below Deception Island range from ~150°C to ~300°C. The δD and δ18O values measured from fumarolic gas and hot spring waters do not indicate any contribution of magmatic water to the samples. The major ionic components and δD-δ18O-δ34S values indicate that hot spring waters are a mixture of local meteoric water and seawater. Mn and SiO2 in spring waters were enriched relative to seawater reflecting water-rock interaction at depth.