Description: Antarctic krill (Euphausia superba) from South Georgia comprise one of the most northern and abundant krill stocks. South Georgia waters are undergoing rapid warming, as a result of climate change, which in turn could alter the oxygen concentration of the water. We investigated gene expression in Antarctic krill related to aerobic metabolism, antioxidant defence, and heat-shock response under severe (2.5% O2 saturation or 0.6 kPa) and threshold (20% O2 saturation or 4 kPa) hypoxia exposure compared to in situ levels (normoxic; 100% O2 saturation or 21 kPa). Biochemical metabolic and oxidative stress indicators complemented the genic expression analysis to detect in vivo signs of stress during the hypoxia treatments. Expression levels of the genes citrate synthase (CS), mitochondrial manganese superoxide dismutase (SODMn-m) and one heat-shock protein isoform (E) were higher in euphausiids incubated 6 h at 20% O2 saturation than in animals exposed to control (normoxic) conditions. All biochemical antioxidant defence parameters remained unchanged among treatments. Levels of lipid peroxidation were raised after 6 h of severe hypoxia. Overall, short-term exposure to hypoxia altered mitochondrial metabolic and antioxidant capacity, but did not induce anaerobic metabolism. Antarctic krill are swarming organisms and may experience short periods of hypoxia when present in dense swarms. A future, warmer Southern ocean, where oxygen saturation levels are decreased, may result in smaller, less dense swarms as they act to avoid greater levels of hypoxia.

Description: The South Georgia region supports a large biomass of krill that is subject to high interannual variability. The apparent lack of a locally self-maintaining krill population at South Georgia means that understanding the mechanism underlying these observed population characteristics is essential to successful ecosystem-based management of krill fishery in the region. Krill acoustic-density data from surveys conducted in the early, middle and late period of the summers of 2001 to 2005, together with krill population size structure over the same period from predator diet data, were used with a krill population dynamics model to evaluate potential mechanisms behind the observed changes in krill biomass. Krill abundance was highest during the middle of the summer in 3 years and in the late period in 2 years; in the latter there was evidence that krill recruitment was delayed by several months. A model scenario that included empirically derived estimates of both the magnitude and timing of recruitment in each year showed the greatest correlation with the acoustic series. The results are consistent with a krill population with allochthonous recruitment entering a retained adult population; i.e. oceanic transport of adult krill does not appear to be the major factor determining the dynamics of the adult population. The results highlight the importance of the timing of recruitment, especially where this could introduce a mismatch between the peak of krill abundance and the peak demand from predators, which may exacerbate the effects of changes in krill populations arising from commercial harvesting and/or climate change.

Description: Acoustic backscatter data were collected at five frequencies (18, 38, 70, 120 and 200 kHz) across two visits to site P3 (P3A, P3B), South Georgia, aboard the RRS Discovery during DY086. Acoustic backscatter was measured with a Simrad EK60. The data consistently shows no evidence of synchronised diel vertical migration (Cook et al. 2023).

Description: Data derived from net catches for zooplankton and micronekton during the COMICS cruise DY086 in November to December, 2017. Raw catch counts and biomass measurements have been used alongside published values to provide biomass, respiration and ingestion data between 0 and 500 metres depth (Belcher et al. 2022, Cook et al. 2023, Stowasser et al. 2020). Data values are from multiple net deployments and the number of deployments for each value are provided in the dataset. Bongo, Multiple Opening/Closing Net and Environmental Sensing System (MOCNESS) and Rectangular Midwater Trawl (RMT) nets collected small (100 μm mesh; day only), medium (330 μm mesh; day and night) and large (4000 μm mesh; day and night) samples, respectively.

Description: Physical, chemical and biogeochemical measurements derived from CTD-rosette deployments during three visits to site P3 (November to December, 2017) in the South Atlantic. Measurements were made during COMICS cruise DY086 on the RRS Discovery using a trace metal free Titanium Rosette (events 4, 7, 15, 19, 24, 26, 29) and a Stainless Steel Rosette (all other events). Physical parameters include temperature, salinity, density, photosynthetically active radiation and turbulence; chemical parameters include dissolved oxygen, dissolved oxygen saturation, nitrate, phosphate and silicate; biogeochemical parameters include turbidity, beam transmittance, beam attenuation, fluorescence, particulate organic carbon (POC), dissolved organic carbon (DOC), chlorophyll-a, net primary productivity (NPP), ambient leucine assimilation and bacterial cell count. To determine turbulence, a downward facing lowered acoustic doppler current profiler (LADCP, Teledyne Workhorse Monitor 300 kHz ADCP) was attached to the CTD frame. Shear and strain, which are obtained from velocity and density measurements, were used to estimate the dissipation rate of turbulent kinetic energy and the diapycnal eddy diffusivity from a fine-scale parameterisation. Estimates are calculated by parameterising internal wave-wave interactions and assuming that wave breaking modulates turbulent mixing. A detailed description of the method for calculating diffusivity from LADCP and CTD can be found in Kunze et al. (2006). Two datasets with different vertical resolutions were produced: one in which the shear is integrated from 150 to 300 m and the strain over 20-150 m, and one in which the shear is integrated from 70 to 200 m and the strain over 30-200 m. Nutrients (nitrate, phosphate, silicate) were determined via colourimetric analysis (see cruise report, Giering and Sanders, 2019), POC was determined as described in Giering et al. (2023), DOC and DOC flux were determined as described in Lovecchio et al. (2023), NPP was determined as described in Poulton et al. (2019), and ambient leucine assimilation and bacterial cell count were determined as described in Rayne et al. (2024). Bacterial abundance and leucine assimilation were made from bottle samples of six CTD casts of the stainless-steel rosette. Water was collected at six depths (6 m, deep-chlorophyll maximum, mixed layer depth + 10, 100, 250 and 500 m). Acid-cleaned HDPE carboys and tubing were used for sampling. Samples were then stored in the dark and at in-situ temperature prior to on-board laboratory sample preparation or analysis. Flow cytometry was used to measure bacterial abundance. Room temperature paraformaldehyde was used to fix 1.6 ml samples for 30 minutes. Then, using liquid nitrogen, the samples were flash frozen and stored at -80°C. Samples were then defrosted before being stained using SYBR Green I and run through the flow cytometer (BD FACSort™). The method of Hill et al. (2013) was applied to determine prokaryotic leucine assimilation using L-[4,5-³H] leucine which has a specific activity of 89.3 Ci/mmol­. In the mixed and upper layers of the water column, the protocol in Zubkov et al. (2007) was followed. Below the mixed layer, adaptions to the method included reducing the concentration of ³H-Leucine to 0.005, 0.01, 0.025, 0.04 and 0.05 nM; increasing experimental volumes to 30 ml; enhancing incubation times to 30, 60, 90 and 120 min. These adaptions were made to improve accuracy where lower rates of leucine assimilation were expected. Data were provided by the British Oceanographic Data Centre and funded by the National Environment Research Council.