The lipid composition of microalgal communities dominated by diatoms collected from the sea ice at three locations within McMurdo Sound during the austral spring bloom of 1989/90, was determined using gas chromatography (GC) and GC mass spectrometry. A range of C27-C29 sterols were detected. The major sterols found at the three sites were 24-methylcholesta-5, 22E-diene-3β-ol (Cape Armitage); trans-22-dehydrocholesterol, 24-ethylcholesterol and 24-methylenecholesta (Erebus Ice Tongue); and 24-methylenecholesterol (Cape Royds). The difference in sterol profiles is believed to reflect the differing species composition at each site. The high relative levels (as % of total) of 24-ethylcholesterol at the Erebus Ice Tongue site (possibly related to Amphiprora kufferathii) supports the proposal that diatoms are a more probable source of C29 sterols in Antarctic lakes than are other algal groups or cyanobacteria. Changes in sterol composition over the course of the bloom were evident at the Cape Armitage site, particularly within the cellular free-lipid fraction. The major fatty acids identified were 14:0, 16:0, 16:1ω7c, 16:4ω1 and 20:5ω3 (Cape Armitage and Erebus Ice tongue sites); 16:0, 16:1ω7c and 20:5ω3 (Cape Royds site). All sites demonstrated high levels of PUFA (40–50% of total fatty acids), with an average 20:5ω3 level of 21% Erebus Ice Tongue, 20% Cape Royds, and 17% Cape Armitage. Variation was also observed in the percentage of 20:5ω3 for the Cape Armitage community over the sampling period. Levels of 22:6ω3 were between 0.4 and 1% of total fatty acids for the three sites. A C25:2 isoprenoid hydrocarbon was present in samples from all sites, adding further evidence to the proposal that diatoms are probably a source of this and related isoprenoid alkenes in marine and coastal sediments.

Twenty five species of Polychaeta were recorded in soft bottom samples collected from 4–30 m. Total abundance of polychaetes ranged from 60 to 3300 m-2. High abundance values were locally recorded for Microspio moorei, Tharyx epitoca and Ophelina syringopyge. These species, together with more regularly distributed Capitella capitata and Scoloplos marginatus, constituted over 70% of all specimens. Total biomass value of the polychaetes varied between 3.8 and 46.4 gm-2. Travisia kerguelenensis and Aglaophamus ornatus constituted over 75% of total biomass. Species composition, richness and diversity of the polychaete assemblage varied with depth. Two parts of the investigated bottom section, differing in the polychaete assemblages structure, were distinguished; the first one in the depth range from 4–20 m and the second one at the depths of 25–30 m. On the basis of both new and previously published data two types of polychaete assemblages of the shallow soft bottom of the Antarctic sublittoral were distinguished. The type of sediment seems to be the main factor influencing the composition of polychaete assemblages.

New cytogenetical data are provided for the three Ranunculus species occurring on Iles Kerguelen. Chromosome morphometrical analysis (idiograms) strongly argues for a relationship between them, with some indications of a more recent origin for R. moseleyi. Aneuploidy was found within the three species for the first time. Whilst only a slight aneuploidy exists in R. pseudotrullifolius and R. moseleyi, R. biternatus shows a tendency to hyperaneuploidy. Vegetative reproduction is likely to increase the number of plants showing such chromosome number irregularities. Aneuploidy could explain some of the differences between the species in ecological tolerance and distributional area.

Data collected on four large-scale surveys around the subantarctic island of South Georgia provide information on the variability in the distribution of chlorophyll and inorganic nutrients during the austral summer and winter. During three summer surveys, surface water cholorophyll and nutrient concentrations were highly patchy over scales ranging from a few to hundreds of kilometres. The highest measurement of chlorophyll a was 8 mg m−3 and a wide range of nutrient concentrations were found; 5–32 mmol m−3 NO3−N, 1.1–2.2 mmol m−3 PO4−P and 8–60 mmol m−3 Si(OH)4−Si. In winter, chlorophyll and nutrient levels were far more uniform, with chlorophyll concentrations lower and nutrient concentrations generally higher than in summer. The spatial variability in nutrient concentrations was due to a variety of factors acting over a range of scales, however biological processes appeared most important in creating the mesoscale patchiness around the island. Although phytoplankton abundance and nutrient concentrations were not directly correlated, the scales of variability were clearly similar.

Adamussium colbecki is one of the dominant members of the nearshore benthic community around Antarctica (Nicol 1966, Dell 1974). It has been studied from a systematic and biogeographic point of view since the beginning of this century (Smith 1902, Pelseneer 1903). However, detailed knowledge of its biology and ecology has been only acquired during the last decade (Stockton 1984, Berkman 1990, Berkman et al. 1991, Mauri et al. 1990, Nigro et al. 1992, Nakajima et al. 1982).

The fatty acid profiles of Euphausia superba, the Antarctic prymnesiophyte, Phaeocystis pouchetii, and a temperate diatom, Phaeodactylum tricornutum were analysed and compared. The lipid content, lipid class, fatty acid and sterol composition of E. superba fed on each cultured phytoplankton and a mixed diet of both species, were determined. No significant difference was found between total lipid levels of E. superba reared on each of these different diets. Phaeocystis pouchetii, although deficient in a number of the essential fatty acids, is apparently an adequate food source for E. superba. The proportion of polyenoic fatty acids varied within lipid classes although there was no significant difference between levels of the long chain polyunsaturate 20:5(n-3) in the total lipid of E. superba fed on these diets. This acid was found to be less than 1% of the total lipid in Phaeocystis pouchetii compared to 37% in Phaeodactylum tricornutum. This suggests that krill may possess the ability to convert exogenous shorter chain fatty acids to 20:5(n-3) and 22:6(n-3). Significant differences were detected in the isomeric ratio of 16:1(n-7c)/16:0 between krill fed the diatom compared to the prymnesiophyte diet. Significant differences were also detected in several shorter chain fatty acids and between fatty acids within their lipid classes. Such differences may have the potential to be used as biochemical signatures to provide information on food sources and possible feeding grounds of E. superba. Phaeocystis pouchetii in a very late stationary phase, although not used in this feeding study, was found to contain 11% of 22:6(n-3) for which there are few substantive sources in natural algal populations.

Sea ice provides a habitat for a conspicuous and productive assemblage of autotrophic microalgae and for heterotrophs ranging from bacteria to vertebrates (Horner 1990, Garrison 1991). With the exception of a reference to chytridiaceous fungi that were found infecting Arctic ice diatoms (Horner 1977) and a note in a cruise report (Schnack-Schiel 1987, p. 153), it appears that fungi and similar organisms have until now not been mentioned as members of the heterotrophic sea ice community. In the present short note we report on the abundant occurrence of apparently thraustochytrid fungus-like protists associated with mucilage tubes of pennate diatoms, encountered in the lower section of a fast ice core drilled close to the southern shelf ice margin of the Weddell Sea.

Twelve saline lakes 5-35m above sea level in Bunger Hills (66°10'S, 101°00'E) were investigated from January to April in 1987-89. Some lakes may be relict and all were subject to wind-borne marine salts with present salinities between 3.4-79.0°/‰ and δ18O values mostly between -10 to -14%. Temperatures up to 17.7°C were measured at the bottom of Lake Polest where a sharp thermohalocline was observed during the period of open water. Mg2+ dominated over Ca2+ in all the lakes. Phosphates concentration was 3-10 μgP I-1 and total phosphorus 8-16 μgP I-1. The concentrations of nitrates and nitrites were often equal, ranging between 1-4 μgN I-1. Dissolved O2 was usually near saturation levels but peaked at more than 230% of saturation in the hypolimnion of Lake Polest as a result of temperature-enhanced photosynthesis with an assimilation rate of 23.9 mgC (mg Chl a)-1h-1. The values for chlorophyll a were 0.26-1.93 mg m-3 and for primary production 0.013-0.171 gCm-3d-1, the latter being an order of magnitude higher than in the fresh water lakes of the Bunger Hills.

The various life history stages of the Weddell Sea population of Pleuragramma antarcticum were sampled to investigate life-history patterns of this ecologically important polar fish. Otoliths were examined for size, morphology and microstructure. Age determination of Antarctic fishes has proved to be difficult because of small ambient temperature fluctuations. External and internal examination of otoliths by scanning electron microscopy revealed internal increments (assumed to be daily) and hatching marks. Back calculation of hatching dates from otolith increments, suggested a hatching season from September-November, with recruitment to the adult population at three to five years of age. Growth data conformed well to the von Bertalanffy equation. Fish grew slowly, with the largest fish attaining ages of more than 30 years. A multivariate mathematical model relating age to otolith morphometrics and fish size proved reliable, making it possible to age large sample sizes of fish. Limited elemental microprobe data obtained from two otoliths demonstrated patterns which may be useful in indicating the environmental life history of individual fish. These collective data suggest that the ecological importance of populations of Pleuragramma antarcticum is most likely due to a long life span and high lifetime reproduction rate.

Thurston Island, and the adjacent Eights Coast and Jones Mountains, record Pacific margin magmatism from Carboniferous to Late Cretaceous times. The igneous rocks form a uniformly calc-alkaline, high-alumina, dominantly metaluminous suite; some relatively fractionated granitoids are mildly peraluminous. The magmas were hydrous, a result of subduction. Gabbros have compositions outside the range of mafic volcanic and hypabyssal rocks, as a result of cumulate processes. Trace element compositions of the mafic magmas range from a low La/Yb, Th/Ta end-member close to E-MORB in composition, perhaps contaminated by crust, to a high La/Yb, Th/Ta end-member, close to shoshonite, with strong magmatic arc trace element character. This variation may be a result of mixing of tholeiitic and shoshonitic end-members. Most silicic rocks could have been generated batch-wise from mafic magmas by fractional crystallization of a phenocryst assemblage dominated by plagioclase, pyroxene ± amphibole, as seen in the cumulates. Cessation of magmatism at about 90 Ma approximately coincided with collison of a spreading centre between the Phoenix and Pacific oceanic plates with the continent margin subduction zone. The rifting of New Zealand from West Antarctica and associated extension probably was responsible for emplacement of a coast-parallel Cretaceous dyke swarm.

The Larsemann Hills represent a low-pressure granulite terrain with a complex structural-metamorphic history that comprises two parts: 1) granulite facies D1 structures transposed within an early form surface that probably formed at 1000 Ma, and 2) a sequence of progressive, upper amphibolite to lower granulite facies D2–D6 structures that formed during the Pan-African at 500 Ma and were associated with the emplacement of granites and pegmatites with high-grade alteration zones. D2–D6 events comprise an early form surface that has been tightly folded and sheared twice after which it was warped and transected by discrete mylonites. D2–D6 assemblages are associated with decompression textures on D1 peak-assemblages, such as cordierite coronas on garnet + sillimanite in metapelite and plagioclase coronas on garnet in metabasite. This suggests that D2–D6 formed at slightly lower pressures than D1 structures. However, the spatial correlation between the coronas and alteration zones around pegmatitic intrusives indicates that the apparent decompression textures may have partly resulted from transient fluxes in water pressure following melt crystallization. Throughout East Antarctica tectonic provinces have been recognized in which the 1000 Ma tectonothermal events are identified as the main stage in the evolution, and Pan-African events are dismissed as a minor thermal overprint. Although the Larsemann Hills are small in area, they are representative of a great many granulite terrains in East Antarctica, and suggest that great care is needed in the structural-metamorphic analysis of such terrains to ensure the separation of tectonic stages before an interpretation of the tectonic path is attempted.

The Lagrelius Point Formation (?Barremian–Aptian) is the basal unit of the Gustav Group and crops out on the north-west coast of James Ross Island. It consists of about 250 m of coarse-grained siliciclastic rocks. The type section of the Lagrelius Point Formation is defined here from just south of Lagrelius Point. The measured section comprises the uppermost 80 m of the unit and mainly consists of clast-supported, boulder, cobble to pebble conglomerates; very coarse to medium-grained sandstones occur rarely. Four sedimentary facies are recognized. A disorganized conglomerate facies (1) is interpreted as having been deposited from non-cohesive debris flows and high density gravelly turbidity currents. Inversely graded conglomerate facies (2) and normally graded to graded stratified conglomerate and pebbly sandstone facies (3) reflect sedimentation from high density gravelly turbidity currents. Massive and parallel stratified sandstone facies (4) is thought to record deposition from high density sandy turbidity currents. Two types of facies assemblages have been recognized. A major channel assemblage, represented by the lower part of the measured section and the minor channel assemblage forming the upper part of the section. The total succession is thought to represent the aggradation of a major submarine braided channel followed by the establishment and subsequent infill of a series of minor channels in a marginal terrace.

The internal structure of ice cores from western Ross Sea pack ice floes showed considerable diversity. Snow-ice formation made a small, but significant contribution to ice growth. Frazil ice was common and its growth clearly occurred during both the pancake cycle and deformation events. Congelation ice was also common, in both its crystallographically aligned and non-aligned varieties. Platelet ice was found in only one core next to the Drygalski Ice Tongue, an observation adding to the increasing evidence that this unusual ice type occurs primarily in coastal pack ice near ice tongues and ice shelves. The diverse internal structure of the floes indicates that sea ice development in the Ross Sea is as complex as that in the Weddell Sea and more complex than in the Arctic. The mean ice thickness at the ice core sites varied between 0.71 m and 1.52 m. The thinnest ice generally occurred in the outer pack ice zone. Regardless of latitude, the ice thickness data are further evidence that Antarctic sea ice is thinner than Arctic sea ice.

A compositionally varied swarm of mafic dykes in the Jetty Peninsula area was emplaced about 320 Ma ago (K-Ar age). There are three major groups: Group 1 dykes range from transitional-alkaline dolerites to camptonites, Group 2 are trachydolerites, and Group 3 are diorite to quartz diorite porphyries. Group 1 dykes have very similar ratios of most incompatible elements and were derived from the same (or a very similar) enriched lithospheric mantle source region (∈Nd −0.18 to −3.05) with high Nb and Ta (i.e., OIB, ocean island basalt, characteristics). However, the presence of several distinct subgroups with different incompatible element abundances implies significantly different degrees of melting. Group 2 trachy dolerites are much more fractionated (mg 22–36), but were apparently derived from a similar, although somewhat more enriched (∈Nd −2.26 to −4.63) source. Group 3 diorites are compositionally quite distinct and may have been derived by intracrustal melting. Enrichment of the mantle source(s) of Groups 1 and 2 dykes apparently occurred about the same time as high-grade metamorphism in the area, and may have been coeval with crust formation in nearby parts of Gondwana.

The application of zircon U-Pb geochronology using the SHRIMP ion microprobe to the Precambrian high-grade metamorphic rocks of the Rauer Islands on the Prydz Bay coast of East Antarctica, has resulted in major revisions to the interpreted geological history. Large tracts of granitic orthogneisses, previously considered to be mostly Proterozoic in age, are shown here to be Archaean, with crystallization ages of 3270 Ma and 2800 Ma. These rocks and associated granulite-facies mafic rocks and paragneisses account for up to 50% of exposures in the Rauer Islands. Unlike the 2500 Ma rocks in the nearby Vestfold Hills which were cratonized soon after formation, the Rauer Islands rocks were reworked at about 1000 Ma under granulite to amphibolite facies conditions, and mixed with newly generated felsic crust. Dating of components of this felsic intrusive suite indicates that this Proterozoic reworking was accomplished in about 30–40 million years. Low-grade retrogression at 500 Ma was accompanied by brittle shearing, pegmatite injection, partial resetting of U-Pb geochronometers and growth of new zircons. Minor underformed lamprophyre dykes intruded Hop and nearby islands later in the Phanerozoic. Thus, the geology of the Rauer Islands reflects reworking and juxtaposition of unrelated rocks in a Proterozoic orogenic belt, and illustrates the important influence of relatively low-grade fluid-rock interaction on zircon U-Pb systematics in high-grade terranes.

A varied suite of sapphirine-bearing and quartz-undersaturated granulites, the Taynaya Paragneiss, occur as boudins and enclaves within the c. 2500 Ma old felsic orthogneisses of northern Vestfold Hills. Highly magnesian varieties with XMg (=100x(Mg/(Mg+Fe)) near 95 preserve the assemblage sapphirine + enstatite + spinel, whereas sapphirine + cordierite + sillimanite + corundum occurs in aluminous and feldspathic types with XMg near 90. Phase equilibria and relative thermometry based on Al2O3 solubility in enstatite indicate equilibration of these assemblages at c. 830–880°C and 0.35–0.85 GPa. There is no evidence for the extreme temperatures (1000–1100°C) previously proposed for early metamorphism in the Vestfold Hills, and no indication in the metamorphic assemblages of isobaric cooling prior to 2500 Ma. Two types of metasomatism have altered the bulk rock compositions near boudin and enclave margins. Cordierite rinds locally formed on corundum-sillimanite granulites reflect interaction with magmatic precursors to the enclosing felsic gneisses, as supported by the isotopic and chemical compositions of cordierite channel volatiles. More extensive metasomatism producing schistose phlogopite + sapphirine rinds on all boudins involved infiltration of a LILE-enriched fluid which introduced K2O, H2O, Fe, Rb, Ba and minor Sr along the boudin margins. Whole rock geochemistry of Taynaya Paragneiss unaffected by this metasomatism is consistent with their derivation from evaporitic mudstones, and implies the existence of a basement older than the dominant 2500 Ma orthogneisses.

Terrestrial sedimentary rocks at Hope Bay, northern Graham Land are well known for their diverse but poorly-preserved fossil flora, previously assigned ages ranging from Early Jurassic-Early Cretaceous. The beds form part of the Botany Bay Group, which comprises several outcrops of terrestrial sediments in northern Graham Land and the South Orkney Islands. A latest Jurassic or earliest Cretaceous age for the Hope Bay plant bearing sequence (and by extension for the rest of the Botany Bay Group) has been adopted in most recent publications dealing with Mesozoic volcanic arc evolution and palaeogeography of the northern Antarctic Peninsula region. New evidence, based upon the study of extensive collections of previously undescribed fossil plants from Hope Bay and nearby Botany Bay, indicates that they should be assigned an Early Jurassic age. The new palaeobotanical findings, combined with recently-published radiometric data from overlying volcanic sequences, show that a Cretaceous age is no longer tenable for these floras nor, therefore, for the Botany Bay Group in Graham Land. Interpretations of Mesozoic volcanic arc evolution and palaeogeography in this region are revised accordingly.

Multichannel seismic data acquired over the northern Pacific margin of the Antarctic Peninsula, have shown the presence of high-amplitude sub-bottom reflections across part of the South Shetland accretionary prism. Detailed seismic data analysis, such as true amplitude signal recovery, reflection coefficient determinations and closely-spaced semblance velocity analysis, have been carried out in order definitely to assign this bottom simulating reflector to the base of the stability field for methane hydrate. This represents the first evidence of gas hydrate layers in the Antarctic Peninsula region.

Temperature records in the Antarctic Peninsula have shown a climatic warming of 1.5°C over the past 30 years and a number of ice shelves have retreated. The most dramatic retreat has been that of Wordie Ice Shelf which has undergone a catastrophic disintegration since the 1960s. Understanding the cause and mechanism of the break-up may provide important clues to the fate of ice shelves farther south which, it has been suggested, help to stabilize the West Antarctic Ice Sheet. The break-up of Wordie Ice Shelf has been analysed using Landsat and SPOT imagery. These observations show that the relative contribution of the various input glaciers to the grounding line flux has not altered during the break-up. This means that the effect of the rapid and almost complete removal of the ice shelf has not been transmitted upstream and is not causing a rapid increase in velocities on the input glaciers. The volume of grounded ice in the catchment of Wordie Ice Shelf will thus, be largely unaffected by the break-up and there will be no significant contribution to sea level change. Since other ice shelves around the Antarctic Peninsula are also fed by relatively steep mountain glaciers the effect of the loss of the ice shelves on sea level would be likely to be similarly small.

Chemical, isotopic and crystallographic characteristics of marine ice formed at the base of the Hells Gate Ice Shelf, Terra Nova Bay, allow a better understanding of the dynamics of marine ice accretion under small ice shelves. The observed properties of the different types of frazil ice found in the area immediately behind the ice shelf front, result from a progressive evolution of the individual frazil ice crystals initially accreted at the base of the ice-shelf. Basal melting caused by the descending plumes of water masses at a temperature above their local freezing point, initiates partial melting of the frazil ice crystals. This dilutes the interstitial water and initiates chemical sorting effects as diffusion proceeds from the normal sea water in the free water column to the diluted interstitial water in the loose frazil layer. Different environmental conditions will result in contrasting properties. Where the subglacial interface is sculptured with domes or inverted channels, it will favour the accumulation of thick units of frazil ice, in a calm environment, that will be further protected from convection mixing over long time periods. This will result in the formation of orbicular frazil showing c-axes at random, strong dilution and important sorting effects. On the contrary, where no channel or dome exist, or where those are already filled with frazil, rectangular or wave-like banded frazil will form with properties showing interfacial streaming effects induced by water currents. Strong c-axes concentration at a single maximum, less dilution and weaker chemical sorting effects are then observed. These findings provide a tentative explanation for the apparent contradiction between the very low salinity levels detected in marine ice at the base of ice shelves and the comparatively minor salinity fluctuations in sea water profiles near ice shelves.