Proposed climate change mechanisms are many and various but generally attributable to our part of the solar system. They usually focus on temperature changes driven either by local processes such as variations in oceanic circulation, or, levels of atmospheric greenhouse gases such as carbon dioxide, or by global processes such as variations in received solar energy linked to changes in the parameters of the Earth's rotation and orbit or solar activity. However, two recent papers have suggested that we may need to look outside the Earth System and even outside our local planetary system for the possible origins of climate change, both on a decadal scale and over longer timescales of hundreds of millions of years. In each case, the galactic cosmic ray flux and its potential effects on cloud formation is considered to be the culprit.

Downward fluxes of particulate matter were investigated in the polynya of Terra Nova Bay (western Ross Sea) from February 1995 to December 1997. The main biological components were siliceous phytoplankton (diatoms, silicoflagellates and parmales), abundant faecal pellets of several types and zooplankton (mainly shelled pteropods). Vertical fluxes of particles occurred mainly through diatoms and faecal pellets in the first and second part of the summer, respectively. The highest fluxes were recurrently observed in late summer, when faeces contributed up to 100% of organic carbon. Unusually high fluxes were recorded in winter 1995, when faecal pellets accounted for 84.6% of the organic carbon. Peak fluxes were always driven by the sinking of faecal pellets, that hence appear to be the most efficient vector of export in the polynya of Terra Nova Bay. A major flux component was the pteropod Limacina helicina, which repeatedly sank in high amounts after the growing season. In April–June, L. helicina probably transported biogenic carbon to deep layers as a passive sinker. The inclusion of pteropods in flux estimates resulted in values that were up to 20 (for total mass), 25 (for organic matter) and 48 (for carbonate) times higher than the previously measured fluxes. Fluxes are known to be biased by swimmers, but ultimately attention must be paid to a possible erroneous categorization of some zooplankton as swimmers to avoid severe underestimation of fluxes of total mass (up to 95% in our study), organic matter (up to 96%) and carbonate (up to 100%).

The ecology of two meltwater streams on King George Island, Ornithologists Creek (with penguin rookeries close to its lower reaches) and Petrified Forest Creek (a highly oligotrophic system), was studied during the 1996–97 summer season. To estimate seasonal productivity of the periphyton and to establish which environmental parameters influenced periphyton growth most strongly, two types of artificial substrata (fibreglass nets – ash-free dry weight (AFDW), and microscope slides – Chlorophyll a (Chl a)) were tested in situ. Thus relative periphyton productivity (RPP) reflects algal colonization and growth as well as losses due to cell mortality and abrasive action of moving sediments. The Petrified Forest Creek was more productive (AFDW = 108.63 μg cm−2 d−1, Chl a = 0.35 μg g cm−2 d−1) than the Ornithologists Creek (AFDW = 69.90 μg cm−2 d−-1, Chl a = 0.26 μg cm−2 d−1). RPP differed both along the streams, and during the season. Significant positive or negative relationships (generalized linear models) were found between RPP and streamwater ‘physico-chemical parameters’ and ‘geomorphological-geographical characteristics’ of the streams' catchments. In addition, in the lower reaches of both streams almost no active colonization or growth was recorded. In the Petrified Forest Creek, the periphyton biomass was so high that mainly passive organic matter deposition occurred. By contrast, in the lower reach of Ornithologists Creek, periphyton colonization and growth was around zero, being negatively influenced by penguin excrement. Ornithologists Creek was richer in nutrients (DIN, DRP), which also fluctuated more widely along its length and throughout the season, than in the Petrified Forest Creek. Parameters associated with the inorganic carbon cycle of the streamwater reflect higher RPP in Petrified Forest Creek. Moreover, RPP was higher in stream reaches with higher amounts of gravel boulders on the bottom.

The native terrestrial food web of sub-Antarctic islands is dominated by decomposers with rare herbivores and almost no predators. As a consequence of increasing human activities, the number of alien plants and invertebrates species, including phytophagous species, has been dramatically rising on these islands. These repeated introductions seem likely to have a great impact on the ecosystem functioning. This is the first detailed study on species diversity, host range and spatial distribution of aphids on French sub-Antarctic islands. Six cosmopolitan and polyphagous aphid species have been recorded on these islands. Five species have been found in the wild where they colonized native and introduced plants, and one species was confined to a glasshouse. Aphids colonized a littoral band and were limited to below 200 m a.s.l. Their spatial distribution is constrained by host plant distribution and temperature. The two dominant species, Myzus ascalonicus and Rhopalosiphum padi, are obligately parthenogenetic in these islands and have been observed to be active on plants during winter. The other species are also presumably obligate parthenogens because of the absence of host plants where sexual reproduction can occur. We suggest that polyphagy and parthenogenesis are major biological traits that influence colonization success by aphids in a sub-Antarctic environment.

An ecomorphological study was performed on ten species of the family Artedidraconidae, the most benthic of the suborder Notothenioidei. These species are sympatric on the shelf of the Weddell Sea. The results show that the four genera comprising the family (Artedidraco, Dolloidraco, Histiodraco and Pogonophryne) can be differentiated by just a few morphometric features of their sensory organs and mouth. The genera were also clearly defined by the composition of their diet (benthic and epibenthic), the size of their prey, and their bathymetric distribution. Comparison of the morphological and ecological data shows a very close connection between sensory organs development, the mouth and bathymetric distribution. The morphological and ecological divergence observed can be explained as the result of the rapid adaptive radiation of the artedidraconids. The results confirm that characteristics of the sensory organs and alimentary structures are very suitable for the ecomorphological study of fishes.

Among the benthic lithological samples collected during the Spanish Antarctic survey GEBRAP 1996–97 in the Bransfield Strait, two new oweniid (Annelida, Polychaeta, Oweniidae) species were collected: Myriochele riojai n. sp. and Myriochele robusta n. sp. Myriochele riojai is closely related to M. heeri Malmgren, 1867 but differs in body size, number and distribution of the notosetae and in the position of elongated abdominal body segments. Myriochele robusta differs from all known species of the genus owing to its possession of a biramous setiger between two uniramous setigers in the anterior (thoracic) body region. A key to oweniid species recorded in Antarctic waters is provided.

A floristic and taxonomic survey was made of the diatom communities of sediments and microbial mats in 66 freshwater and saline lakes and pools in the Larsemann Hills, Rauer Islands and Bølingen Islands (continental eastern Antarctica). A total of 31 taxa were distinguished, 10 of which could not be identified to species (nine) or even generic (one) level, either because they have most probably not yet been described or because they belong to species complexes that are in need of revision. Four new combinations are proposed; three species are reported for the first time from continental Antarctica, while another three are confirmed for eastern Antarctica for the first time. Analysis of literature data on Antarctic lacustrine diatoms shows that taxonomic practice has a profound influence on the assessment of distribution patterns. Force-fitting of European and North American names to Antarctic taxa and erroneous identifications have contributed to an underestimation of endemism in the diatom flora of Antarctic inland waters. In addition, changing concepts on species boundaries during the last decade influence the interpretation of biogeographic patterns. The application of a more fine grained taxonomy will almost certainly reveal a higher degree of endemism in Antarctica, and especially continental Antarctica. The present case study shows that in the Larsemann Hills Antarctic endemics account for about 40% of all freshwater and brackish taxa, while the biogeographic distribution of about 26% is unknown, mainly due to their uncertain taxonomic identity. This contradicts the view that cosmopolitanism prevails in Antarctic diatoms.

Although world oceans have been warming over the past 50 years, the impact on biotic components is poorly understood because of the difficulty of obtaining long-term datasets on marine organisms. The Southern Ocean plays a critical role in global climate and there is growing evidence of climate warming. We show that air temperatures measured by meteorological stations have steadily increased over the past 50 years in the southern Indian Ocean, the increase starting in mid 1960s and stabilizing in mid 1980s, being particularly important in the sub-Antarctic sector. At the same time, with a time lag of 2–9 years with temperatures, the population size of most seabirds and seals monitored on several breeding sites have decreased severely, whilst two species have increased at the same time. These changes, together with indications of a simultaneous decrease in secondary production in sub-Antarctic waters and the reduction of sea-ice extent further south, indicate that a major system shift has occurred in the Indian Ocean part of the Southern Ocean. This shift illustrates the high sensitivity of marine ecosystems, and especially upper trophic level predators, to climatic changes.

Accurate assessment of the abundance of pinnipeds from visual surveys requires estimation of both the available (hauled-out) and unavailable (in-water) components of the population (Eberhardt et al. 1979). Continental estimates of the abundance of the four Antarctic seals are based on limited information on haul-out behaviour. In developing continental estimates, Erickson & Hanson (1990) corrected visual surveys of the hauled-out component of the species' populations using data from observational studies of haulout behaviour by Erickson et al. (1989). Erickson & Hanson (1990) point out that, because the observational studies did not account for an unknown fraction of seals that remained in the water during the peak haul-out period, their abundance estimates are minimum values. Further, Erickson & Hanson (1990) corrected the visual surveys for all four species using haul-out data for the crabeater seal only, as observational data for the other species were not available. This assumes that haul-out patterns are constant across species, which is largely untested. Consequently, there is potential for bias, in both a relative and absolute sense, in the estimated abundance of Antarctic seals.

Field observations involving landform evaluation and the physical characterization of sediments, combined with a detailed analysis of the spatial distribution of bedrock and sediment geochemical patterns, suggests a limit to the glacial transport half-distance of c. 3 km in Vestfold Hills. Four morphologically distinct glacial deposits were sampled (small debris ridges, large debris ridges, debris drapes and valley fills) on the basis of field geometry. These landforms were subsequently distinguished by grain size, mineralogy and geochemistry. Since there are no nunataks south of Vestfold Hills, all debris is derived subglacially and sedimentological differences are attributed to the physical weathering of preglacial surfaces in Vestfold Hills and fluvial winnowing during deposition. Given that thrust geometries may occur in large debris ridges, glacial transport distances were short, and fluvial sorting of sediments was an important mechanism, reconstructions of glacial histories based on the stratigraphy of deposits in Vestfold Hills should be made with caution.

Multi-channel seismic data acquired on the South Shetland margin, northern Antarctic Peninsula, show that Bottom Simulating Reflectors (BSRs) are widespread in the area, implying large volumes of gas hydrates. In order to estimate the volume of gas hydrate in the area, interval velocities were determined using a 1-D velocity inversion method and porosities were deduced from their relationship with sub-bottom depth for terrigenous sediments. Because data such as well logs are not available, we made two baseline models for the velocities and porosities of non-gas hydrate-bearing sediments in the area, considering the velocity jump observed at the shallow sub-bottom depth due to joint contributions of gas hydrate and a shallow unconformity. The difference between the results of the two models is not significant. The parameters used to estimate the total volume of gas hydrate in the study area were 145 km of total length of BSRs identified on seismic profiles, 350 m thickness and 15 km width of gas hydrate-bearing sediments, and 6.3% of the average volume gas hydrate concentration (based on the second baseline model). Assuming that gas hydrates exist only where BSRs are observed, the total volume of gas hydrates along the seismic profiles in the area is about 4.8 × 1010 m3 (7.7 × 1012 m3 volume of methane at standard temperature and pressure).

Middle Jurassic continental flood basalts of Vestfjella, western Dronning Maud Land are cut by gabbroic intrusions that represent rare exposures of Karoo-related mafic plutons in Antarctica. The gabbros and numerous associated dolerites indicate high magmatic activity along the continental margin of Dronning Maud Land during the break-up of Gondwana. The scattered nunataks of Utpostane (∼25 km2) are dominated by olivine gabbronorite and olivine mela-gabbronorite, which can be grouped into four zones. The Utpostane intrusion exhibits inclined sheet-like geometry, is at least ∼3 km thick and shows moderate enrichment of incompatible element contents from its base towards the exposed roof contact (e.g. Zr from ∼15 to ∼60 ppm). High MgO contents (∼8–36 wt%) indicate that the parental magma of Utpostane was more primitive than a typical Karoo tholeiite (MgO ∼6 wt%). At Muren, gabbroic outcrops record a cross section of a ∼1.3 km thick inclined sheet-like intrusion. The intrusion can be divided to two main units, the upper and lower zones, which are dominated by olivine gabbro and gabbronorite, respectively. Parental melts of the geochemically differentiated upper zone and the homogeneous lower zone of Muren were typical low-MgO Karoo tholeiites, but they were chemically distinct and were emplaced as separate magma pulses.

This paper describes a major calving of the Ninnis Glacier tongue in January 2000. This event, which took 10 years to complete, produced a major change in the George V Land coastline and a large iceberg (∼800 km2). By grounding or becoming “locked in” by fast ice, bergs produced locally and drifting in from afar reside for long periods (10–20 years) in the region ∼146 to 154°E to have a profound impact on sea ice distribution, both locally and 100s of kilometres up- and downstream. They are responsible for the formation of a lens of thick perennial ice to the east (∼14 000 km2 in area), and polynyas. Iceberg movement is sporadic, with intermittent ungroundings of large bergs occurring every 5–13 years. “Escaping” bergs have a temporary impact on the Mertz Glacier polynya to the west. Although public attention focuses on vast bergs, assemblages of small bergs appear to be equally important in terms of their impact on regional fast and pack ice distribution. Possible global change scenarios are discussed. The need for field observations and improved bathymetric and oceanographic data is emphasized.