CCAMLR

Multifrequency echosounder data were collected during the 2006 BROKE-West summer survey of Division 58.4.2 for the purposes of estimating the unexploited biomass (B0) of Antarctic krill (Euphausia superba) and its associated coefficient of variance (CV). This paper updates the version submitted to WG-EMM in 2006 (WG-EMM-06/16) because a reanalysis of the data has resulted in amendments to the acoustic estimates of krill mean biomass density, biomass and variance. The mean acoustic biomass density of krill, integrated to 250 m depth across the entire survey stratum (1.31 million km2), was 9.48 g m-2. B0 was estimated to be 12.46 million tonnes with a CV of 15.15%. Krill were widely distributed at relatively low densities throughout the survey area; only 13% of the 2-km-alongtrack echo-integration intervals were devoid of krill, 50% of intervals registered densities of 1 g m-2 of krill or less, and 80% of intervals registered densities of 10 g m-2 or less. Mean densities were highest in the waters to the south of the Southern Boundary of the Antarctic Circumpolar Current, particularly in waters to the west which were within the influence of the Weddell Gyre. Half of the cumulative density across the survey was found within 120 km of the 1000 m isobath (the shelf-break/ slope), and 40% within 50 km. This was mostly due to very high densities (up to 1400 g m-2) around the shelf break on 3 of the 11 transects surveyed. The majority of acoustic krill detections were in the top 100 m of the water column, centred around 50 m depth. The krill distributions inferred from both the acoustic data and from net catches were considered in the context of the physical oceanography, from which a case is presented for the subdivision of Division 58.4.2 into smaller, more biologically homogeneous areas. A qualitative critical appraisal of the methods is included by way of contribution to ongoing discussions about acoustic survey and analysis methods for krill.

A field key to early life stages of Antarctic fish caught along with the Antarctic krill is produced. The key includes 8 families and 28 species mainly from the Atlantic sector of the Southern Ocean and uses distinguished characters which permit rapid field identification. In some cases, however, it is impossible to discriminate among species of the same family by remarkable characters. A species key is not shown for such resemble species and a brief summary of the main morphological features of species and genera is provided.

Antarctic krill biomass trends in the South Shetland Island region of Area 48 are presented. Updated time series using the Stochastic Distorted Wave Born Approximation, and a dynamic ?Sv krill delineation model are used to derive krill biomass. This paper provides updated (through summer 2007), acoustic biomass estimates previously presented at the 2006 WG-EMM meeting in Namibia (WG-EMM-06/32). In 2007, biomass in the South Shetland Islands region exceeded 19 million tons. This increase from

The total abundance of krill in the Scotia Sea was estimated from an international echosounder and net survey (CCAMLR 2000) to be 44.3 million metric tons (Mt; CV=11.4%), prompting the Antarctic Treaty's Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) to revise the precautionary catch level for krill in the area from 1.5 to 4 Mt (SC-CAMLR, 2000). By incorporating recent improvements in the remote identification and target strength (TS) of krill, a range of krill biomass was estimated, 108.0 Mt (CV=10.4%) to 192.4 Mt (CV=11.7%), depending solely on the expected distribution of krill orientations. The new methods were then reviewed by CCAMLR, and revised protocols based on the Stochastic Distorted Wave Born Approximation model (SDWBA) were adopted. Here, the protocols are applied to again re-analyze the CCAMLR 2000 data. Using the 120 kHz echosounder data, the resulting estimates of krill biomass in the Scotia Sea are 197.78 Mt (CV=11.06%) and 37.29 Mt (CV=21.20%), depending on whether two- or three-frequencies are used for krill identification, respectively. At 38 kHz, estimates of krill biomass range from 65.64 Mt (CV= 11.50%) to 10.39 Mt (CV= 15.25%); and at 200 kHz from 343.09 Mt (CV= 12.91%) to 38.73 Mt (CV= 14.86%). CCAMLR uses the estimates derived from the 120 kHz data. Results of the three-frequency method are likely less biased owing to better rejection of non-krill species; also the patchiness of krill is better elucidated, resulting in higher CVs. Thus, the revised estimates of krill biomass in the Scotia Sea during the CCAMLR 2000 survey are 37.29 Mt (CV=21.20%), or 15.8% lower than the original estimate, but with a larger CV.

In Antarctic krill, Euphausia superba, sampled by a Japanese scientific observer onboard a krill fishing vessel in the winter of 2003 and 2006 in the South Georgia region, the Antarctic Ocean, approximately 2-5% of sub-samples of 100 krill bore small black spots. The black spots were most often found on the cephalothorax of the body. Three bacteria were isolated from these black spots, and classified into either Psychrobacter or Pseudoalteromonas by the sequences of 16S rRNA genes. Histological observations revealed that the black spots were melanized nodules. A single melanized nodule often contained more than one type of morphologically distinct bacterial cell. More than three bacterial species or strains were also confirmed by in situ hybridization for 16S rRNA. The melanized nodules were almost always accompanied by a tumor-like mass of unknown large heteromorphic cells, which seemed to be derived from a gonadal tissue. These results suggest that the krill were affected by bacterial infections, whereas the presence of multiple bacterial species suggests that the infections were likely to be secondary. The development of the tumor-like cell mass in the gonad may be the primary condition, although this requires further detailed study.

Net selection of the scientific RMT8 plankton net and a commercial sized pelagic trawl has been studied. Selection curves indicate that krill length classes smaller 20 mm are underrepresented in the RMT8. For the commercial trawl with a liner mesh size of 12mm the length classes smaller 35 and 45 mm are undersampled, probably depending on the size composition of the catch. A preliminary study on krill damaged during trawling operations indicates an effect of trawling duration and total catch per haul.

This paper outlines the difference in fishing patterns between regions (Subareas) and seasons by analyzing CPUE (catch per hour) patterns towards the end of their sets of operations before leaving a SSMU, as well as its timing of the day, diurnal pattern of CPUE and fishing depth. Vessels tend to move between SSMUs frequently in summer, especially in SSMUs in Subarea 48.1 mostly spending less than a day to explore the fishing grounds from multiple options for the start of the fishing season. Vessels move around less frequently in Subarea 48.2 during the summer season because of only one viable SSMU, but in autumn, there appear to be some opportunistic fishing on the way to Subarea 48.1 to 48.3. Continuous operations in a single SSMU are longer later in the fishing season, especially for winter operation in Subarea 48.3, however forced to be relatively short in Subarea 48.1 due to changing ice conditions etc. Skippers tend to allow a day to decide before leaving a SSMU after over days of continuous operations. This behaviour suggests there are two types of tactics employed within an SSMU. The first set of tactics is to optimize factory supply. When factory supply diminishes, the skipper may initiate a second set of pre-departure tactics over one day to determine if factory supply can be regained. CPUE and towing depth generally showed clear diurnal patterns. It was not possible to outline vessel patterns of pumping methods due to small amount of data accumulated so far. More detailed data is need to be obtained to improve these analyses and models in the future.

The workload of the tasks required in the CCAMLR Observer manual is assessed. The total time needed for the minimum amount of daily routine tasks was well above the capacity that an observer to undertake. The manual must be structured as its entirety so that the observer, only by following instructions in the manual, can produce the report and logbook data which is collected systematically and allows CCAMLR to achieve its objectives.

This interim protocol has been developed in cooperation with our UK and Japanese colleagues/operators/observer coordinators in response to these recent requests by the CCAMLR Scientific Committee (SC-CAMLR 2006). The purpose of developing this protocol is to standardize the larval fish by-catch observation among observers and vessels so that we can later use it for quantitative analysis. The fish identification can also be verified by the CCAMLR fish experts by using systematically archived images. It is also designed so that the data can be validated later by using randomly kept samples by the flag states. This protocol will be used only for Fish and Larval fish by-catch observation, and therefore the rest of the observation must be undertaken using the usual CCAMLR observer manual. This manual was distributed to all the fishing krill fishing nations prior for use in 2006/07 fishing season.

Ecological risk management is increasingly being applied to marine fisheries worldwide as an aid to developing management strategies to avoid, mitigate, or manage adverse outcomes. Risk management encompasses four major steps: recognition of risk; assessment of risk; development of strategies to avoid, mitigate, manage or tolerate risk; and monitoring of risk. Here we begin the development of an ecological risk assessment for Antarctic Toothfish (Dissostichus mawsoni) longline fishery in the Ross Sea, Antarctica. We propose that, by defining risks and quantifying potential impacts, the limited research and management resources can be prioritised so as to meet the objectives of Article II of CCAMLR. Risks are considered in 4 categories: 1. Target species harvest: Risks of depletion of Antarctic Toothfish to below a level that ensures stable recruitment. 2. Bycatch species harvest: Risks of depletion of other harvested species to below a level that ensures stable recruitment. 3. Ecosystem impacts: Risks of changes to the marine ecosystem relationships due to the removal of harvested and bycatch species. 4. Exogenous effects: Risks of change in the marine ecosystem due to, or exacerbated by, exogenous effects (e.g., the introduction of alien species, effects of associated activates on the ecosystem, and effects of environmental change). The assessment of risk is based on combining the likelihood of an adverse outcome occurring and the consequence should it occur. Numerical models, such as stock or ecosystem mass-balance models can provide insights into these factors for some risks. In addition, semi-quantitative and qualitative estimates are needed because of a lack of knowledge and inability to predict the future dynamics of some parts of the system. It is also recognised that some risks (e.g., impacts of climate change) may not be able to be well predicted. The uncertainty arising from the complexity of the system and external factors acting on it means that risk management and ongoing monitoring will be required to ensure that the fishery is managed according to the conservation principles of Article II of CCAMLR.