During the ice stations, sea ice, brine/slush, snow and under-ice water sampling were collected for oxygen isotopic ratio. Ice cores were collected using a Kovacs 9 cm diameter ice corer. The ice core for oxygen isotopic ratio was cut directly after retrieval with a stainless steel folded saw. The core was cut generally into 10 cm sections (20 cm when ice cores were higher than 200 cm) and put into zip-lock polyethylene bags. Care was taken to use laboratory gloves when collecting the cores. For brine sampling, partial core holes were drilled into the ice (so called sackholes), usually to a depth of 25 cm and 50 cm. At site with flooding, brine collection was not possible, and samples of the surface slush were collected instead. Slush was collected by plastic shovel. Snow samples were also collected. Under-ice water was collected with a Teflon water sampler (GL Science Inc., Japan) 1, 3, 5 m below the bottom of the sea ice. In addition, CTD water sampling was examined at each station. The cores were taken back to the ship, and transferred to the gas tight bag (GL Science Inc., Japan), and then ice was melted at about +4 degrees C in a refrigerator. Melted samples were sub-sampled for each component. The snow samples were treated in the same manner as the sea ice samples for further analysis. Oxygen isotopic ratio was determined with a mass spectrometer (DELTA plus; Finnigan MAT, USA) in Hokkaido University. Oxygen isotopic ratio in per mil (parts per thousand) was defined as the deviation of H2 18O/H2 16O ratio of the measured sample to that of the standard mean ocean water (SMOW). The precision of oxygen isotopic ratio analysis from duplicate determinations is plus or minus 0.02 parts per thousand (Toyota et al., 2007). Data available: excel files containing sampling station name, dates, and oxygen isotopic ratio.

Sediment cores were collected from the East Antarctic margin, aboard the Australian Marine National Facility R/V Investigator from January 14th to March 5th 2017 (IN2017_V01; Armand et al., 2018). This marine geoscience expedition, named the “Sabrina Sea Floor Survey”, focused notably on studying the interactions of the Totten Glacier with the Southern Ocean through multiple glacial cycles. The cores were collected using a multi-corer, allowing to sample the surface of the sediment (top ~ 30cm). The cores were then sliced every centimetre, wrapped up in plastic bags, and stored in the fridge. Radiocarbon (14-C) ages were measured to build an age model for future paleo-reconstructions. Sediment samples were pre-treated in the IMAS Sediment Lab (UTAS, Hobart, Australia). Samples (~ 2 g) from the multi-cores MC01, MC03 and MC06 were dried, ground and acidified with HCl for carbonate removal using sterilised beakers. Dried and ground samples were then packed into sterilised aluminium foil and sent to DirectAMS (Radiocarbon Dating Service, USA) for 14C analysis by Accelerator Mass Spectrometer (AMS). Results were corrected for isotopic fractionation with an unreported δ13C value measured on the prepared carbon by the accelerator. References L.K. Armand, P.E. O’Brien and On-board Scientific Party. 2018. Interactions of the Totten Glacier with the Southern Ocean through multiple glacial cycles (IN2017-V01): Post-survey report, Research School of Earth Sciences, Australian National University: Canberra.

Continuous underway measurements of sea surface (7 metres depth)dissolved gasses (co2, o2, argon, nitrogen)by quadrupole mass spectrometry (Electron Impact Mass Spectrometry - EIMS). ASCII encoded. 1 file per 24 hours. Naming convention: YYMMDD. Excel readable format. Column data (0/0 refers to ion mass, 7 ION masses detected in total): Cycle Date Time RelTime[s] '0/0' '0/1' '0/2' '0/3' '0/4' '0/5' '0/6' '0/7' '1/0' '2/0' '2/1' '2/2' '2/3' '2/4' '2/5' '2/6' '2/7' Measurements were made on the CEAMARC voyage of the Aurora Australis - voyage 3 of the 2008-2008 summer season.

40Ar/39Ar geochronology data of basalt samples from the Kerguelen Plateau and Broken Ridge The samples include basalts from ODP drilling cores and dredge sites. The drilling core samples were stored in the Kochi Core Centre, Japan and the dredged samples were stored in the National Museum of Natural History, France. Analytical methods of the 40Ar/39Ar geochronology data: Samples were crushed and minerals/groundmass were separated using a Frantz magnetic separator. Plagioclase, pyroxene, amphibole, sericite, and basaltic glass crystals and groundmass were separated from either the 125–212 μm or the 212–355 μm size fractions using a Frantz isodynamic magnetic separator. Minerals and groundmass were subsequently hand-picked grain-by-grain under a binocular stereomicroscope. Plagioclase and groundmass were further leached using diluted HF (2N) for 5 minutes and thoroughly rinsed in distilled water. Samples were loaded into several large wells of 1.9cm diameter and 0.3 cm depth aluminium discs. The discs were Cd-shielded to minimise undesirable nuclear interference re-actions and irradiated for 40 hours in the Oregon State University nuclear reactor (USA) in the central position. The samples were irradiated alongside FCs and GA1550 standards, for which ages of 28.294 ± 0.037 Ma and 99.738 ± 0.100 Ma were used, respectively. The 40Ar/39Ar analyses were performed at the Western Australian Argon Isotope Facility at Curtin University. The samples were step-heated using a continuous 100 W PhotonMachine© CO2 (IR, 10.4 µm) laser fired on the crystals during 60 seconds. Each of the standard crystals was fused in a single step. The gas was purified in an extra low-volume stainless steel extraction line of 240cc and using one SAES AP10 and one GP50 getter. Ar isotopes were measured in static mode using a low volume (600 cc) ARGUS VI mass spectrometer from Thermofisher© set with a permanent resolution of ~200. Measurements were carried out in multi-collection mode using four faradays to measure mass 40 to 37 and a 0-background compact discrete dynode ion counter to measure mass 36. We measured the relative abundance of each mass simultaneously using 10 cycles of peak-hopping and 33 seconds of integration time for each mass. Detectors were calibrated to each other electronically and using air shot beam signals. The raw data were processed using the ArArCALC software. The criteria for the determination of plateau are as follows: plateaus must include at least 70% of 39Ar released. The plateau should be distributed over a minimum of 3 consecutive steps agreeing at 95% confidence level and satisfying a probability of fit (P) of at least 0.05. Plateau ages are given at the 2σ level and are calculated using the mean of all the plateau steps, each weighted by the inverse variance of their individual analytical error. Uncertainties include analytical and J-value errors.

Owing to the fact that the principal investigator died before data were able to be archived, the only available data are in the form of the referenced paper, which is available as a PDF download to AAD staff only. From the referenced papers: Macquarie Island is an exposure above sea level of the Macquarie Ridge Complex, on the boundary between the Australian and Pacific plates south of New Zealand. Geodynamic reconstructions show that at ca. 12-9.5 Ma, oceanic crust of the Macquarie Island region was created at this plate boundary within a system of short spreading-ridge segments linked by large-offset transform faults. At this time, the spreading rate was slowing (less than 10 mm/yr half-spreading rate) and magmatism was waning. Probably before 5 Ma, and possibly before the extinct spreading ridge had subsided, the plate boundary became obliquely convergent, and crustal blocks were rotated, tilted, and uplifted along the ridge to form the island. Planation by marine erosion has exposed sections through the oceanic crust. The magmatism that built the oceanic crust produced melts similar in composition to the widespread normal to enriched mid-oceanic ridge basalt (N- to E-MORB) suite found in many spreading ridges, but the melts ranged beyond E-MORB to primitive, highly enriched, and silica-undersaturated compositions. These compositions form one end member of a continuum from MORB but seem not to have been derived from a MORB-source mantle, despite sharing a Pacific MORB isotopic signature. The survival of these primitive melts may be due to their origin in a slow-spreading system that must have been closing down as extension along the plate boundary gave way to transpression, putting a stop to the upwelling of asthenosphere and decompression melting. In a more energetic, faster-spreading system, mixing would have been more efficient, the presence of this end member could not easily have been inferred from its isotopic composition, and the igneous rocks would have resembled a typical N- to E-MORB suite. Macquarie Island may therefore provide a type example of magmatism at a very slow spreading ridge and a clue to the origins of E-MORB. Macquarie Island is an exposure above sea-level of part of the crest of the Macquarie Ridge. The ridge marks the Australia-Pacific plate boundary south of New Zealand, where the plate boundary has evolved progressively since Eocene times from an oceanic spreading system into a system of long transform faults linked by short spreading segments, and currently into a right-lateral strike-slip plate boundary. The rocks of Macquarie Island were formed during spreading at this plate boundary in Miocene times, and include intrusive rocks (mantle and cumulate periodites, gabbros, sheeted dolerite dyke complexes), volcanic rocks (N- to E-MORB pillow lavas, picrites, breccias, hyaloclastites), and associated sediments. A set of Macquarie Island basaltic glasses has been analysed by electron microphobe for major elements, S, Cl, and F; by Fourier transform infrared spectroscopy for H2O; by laser ablation-inductively coupled plasma mass spectrometry for trace elements; and by secondary ion mass spectrometry for Sr, Nd and Pb isotopes. Macquarie Island basaltic glasses are divided into two compositional groups according to their mg-number-K2O relationships. Near-primitive basaltic glasses (Group I) have the highest mg-number (63-69), and high Al2O3 and CaO contents at a given K2O content, and carry microphenocrysts of primitive olivine (Fo86-89.5). Their bulk compositions are used to calculate primary melt compositions in equilibrium with the most magnesian Macquarie Island olivines (Fo90.5). Fractionated, Group II, basaltic glasses are saturated with olivine + plagioclase + or - clinopyroxene, and have lower mg-number (57-67), and relatively low Al2O3 and CaO contents. Group I glasses define a seriate variation within the compositional spectrum of MORB, and extend the compositional range from N-MORB compositions to enriched compositions that represent a new primitive enriched MORB end-member. Compared with N-MORB, this new end-member is characterised by relatively low contents of MgO, FeO, SiO2 and CaO, coupled with high contents of Al2O3, TiO2, NaO2, P2O5, K2O and incompatible trace elements, and has the most radiogenic Sr and Pb regional isotope composition. These unusual melt compositions could have been generated by low-degree partial melting of an enriched mantle peridotite source, and were erupted without significant mixing with common -MORB magmas. The mantle in the Macquarie Island region must have been enriched and heterogenous on a very fine scale. We uggest that the mantle enrichment implicated in this study is more likely to be a regional signature that is shared by the Balleny Islands magmatism than directly related to the hypothetical Balleny plume itself.

Mesopelagic fish bulk stable isotope data from the Kerguelen Axis ecosystem study (AAS_4344): These data are based on samples collected as part of the Kerguelen Axis marine ecosystem study (AAS_4344), chief scientist Andrew Constable. This research was supported by the Australian government under the (i) Cooperative Research Centre Program through the Antarctic Climate and Ecosystems Cooperative Research Centre (ACE CRC), (ii) Australian Antarctic Science Program (Projects 4343, 4344, 4347 and 4366), and (iii) Australian Research Council’s Special Research Initiative for Antarctic Gateway Partnership (Project ID SR140300001). The preferred citation is: Woods, B., Walters, A., Hindell, M.A., Trebilco, R. (2019) Isotopic insights into mesopelagic niche space and energy pathways on the southern Kerguelen Plateau. Deep Sea Research Part II: Topical Studies in Oceanography Samples for stable isotope analysis were collected on board the R.S.V Aurora Australis during the austral summer of 2016 (22 January – 17 February) as part of the Kerguelen Axis marine ecosystem study (AAS_4344). Samples were collected from 15 sampling stations along two transects from the Antarctic continental shelf to the BANZARE Bank over the Kerguelen Plateau and in an east to west direction across the Kerguelen Plateau. Mesopelagic fish were sampled from the surface to 1000 m depth using an IYGPT (International Young Gadoid Pelagic Trawl) net equipped with a MIDOC (Mid-water Open Close) multiple cod-end device. Analyses focused on an assemblage including genera from the family Myctophidae (Electrona, Gymnoscopelus, Krefftichthys and Protomyctophum), and the genus Bathylagus from the family Bathylagidae, as these are dominant genera in the Southern Ocean (Pusch et al., 2004; Hulley and Duhamel, 2011; Collins et al., 2012). Muscle tissue from each fish was analysed for δ15N and δ13C. Collins, M. A., Stowasser, G., Fielding, S., Shreeve, R., Xavier, J. C., Venables, H. J., . . . Van de Putte, A. (2012). Latitudinal and bathymetric patterns in the distribution and abundance of mesopelagic fish in the Scotia Sea. Deep-Sea Research Part Ii-Topical Studies in Oceanography, 59, 189-198. doi:10.1016/j.dsr2.2011.07.003 Hulley, P. A., and Duhamel, G. (2011). Aspects of lanternfish distribution in the Kerguelen Plateau region. The Kerguelen Plateau: marine ecosystems and fisheies. G. Duhamel and DC Welsford, Editors, 183-195. Pusch, C., Hulley, P. A., and Kock, K. H. (2004). Community structure and feeding ecology of mesopelagic fishes in the slope waters of King George Island (South Shetland Islands, Antarctica). Deep-Sea Research Part I-Oceanographic Research Papers, 51(11), 1685-1708. doi:10.1016/j.dsr.2004.06.008

Marine sediment samples were obtained from box corer, Smith-MacIntyre and Van Veen grabs. Samples were named by: 1. CEAMARC site (e.g. 16) 2. Instrument (e.g. box corer = BC; Smith-MacIntyre = GRSM; Van Veen = GRVV) 3. Sequence of sample at each site (e.g. first sample = 01; second sample = 02) So 16BC02 is the second sample at CEAMARC site 16, using the box corer. From each successful sample, a sub-sample was obtained: 1. 200 g surface scrape (labelled A) 2. short (20 cm) push core (labelled B) 3. bulk (labelled Bulk) 4. rocks-only (labelled Rocks) e.g. 16BC02A is a 200 g surface scrape subsample from 16BC02. 16BC02B is a push core subsample from 16BC02 16BC02Bulk is a bulk sediment subsample from 16BC02. 16BC02Rocks is a rocks-only subsample from 16BC02. Post-cruise analyses: 1. Grain size 2. Total organic carbon 3. Total organic nitrogen 4. Carbon and nitrogen isotopes 5. Biogenic silica and carbonate 6. Physical properties of cores 7. Zircon dating 8. X-rays for infauna and sedimentary structures Added by Alix Post - March 2010: Seabed samples were collected from 52 sites across the George V Shelf. Most samples were collected with a box corer (BC), though more gravelly sediments required a Smith-McIntyre (GRSM) or Van-Veen grab (GRVV) as indicated by the station name in the spreadsheet. A small volume of sediment was frozen following collection and later analysed for organic carbon and nitrogen content, in addition to carbon and nitrogen isotopes. Organic carbon and nitrogen values are express as percent of the total sediment, and have been corrected back to the total sediment volume. Isotopic values are expressed as values per mil. Where sufficient volume of sediment was collected, a mini-core was pushed into the sediment to provide a depth profile of the sample, and a bulk surface sample was also taken. Surface sediment samples analysed for sieve grainsize, calcium carbonate and biogenic silica content. All values are expressed as percentage values. The naming convention of the samples describes the type of gear used and the nature of the sediment analysed: e.g. 01BC01Bulk is a bulk sediment sample collected with a box core; 38GRVV02B/0-1 is a slice taken from 0 to 1 cm at the top of a van veen grab.