The sentiment of online content containing Antarctic krill fishing and related search terms 12 was analysed to evaluate public perception of the fishery, as a consideration in management 13 decisions. The Antarctic krill fishery is managed by the Commission for the Conservation of 14 Antarctic Marine Living Resources (CCAMLR) which aims to manage the fishery with 15 minimal ecosystem impacts. A combination of recent increases in catches and perceived 16 negative environmental impacts of fishing for Antarctic krill has caused concern for the 17 future of krill populations among stakeholders and has resulted in an ongoing public debate. 18 The tenor of this debate is inevitably guided by access to information about Antarctic krill, 19 typically through online searches and social media. Public perception of the fishery was 20 analysed through sentiment and relevant keyword searches from three online platforms 21 (Factiva®, Google and Twitter). The analysis revealed an overall neutral to positive 22 sentiment of Antarctic krill fishing related content across all search platforms. This study 23 forms a baseline result for future monitoring of sentiment regarding Antarctic krill fishing as 24 it continues to operate in a changing environment as well as a providing a method for using 25 online content sentiment analysis to gauge the public’s perception of other fisheries.

As part of the German contribution to Convention for the Conservation of Antarctic Marine Living Resources (CCAMLR), we will conduct an acoustic krill (Euphausia superba) biomass survey in CCAMLR division 48.1 during April 2018. The survey will be part of a larger research cruise (PS112) investigating the role of krill and salps in Southern Ocean carbon cycling and the temperature adaptation capacities of both species in the context of climate change. Acoustic survey and physiological experiments will be conducted in conjunction with a detailed description of the biological and physical environment of the krill habitat. The overall objective of the German contribution is to provide an overarching assessment of the effect of climate change on krill and associated ecosystem-processes. The aim of this particular proposal however is to discuss the proposed design and methods of the acoustic biomass survey to ensure they align with CCAMLR standard procedures and are comparable with previous krill biomass surveys.

Marine ecosystems are becoming increasingly threatened by anthropogenic disturbances, and understanding where marine top predators forage is vital to ecosystem based marine spatial planning. Using habitat models to understand which variables influence the distribution of predators at-sea can enable us to predict the distributions of un-tracked populations, and thus help to identify any potential threats they may face. In the Southern Ocean the expanding krill fisheries have the potential to impact penguin populations. We show that the distribution of chinstrap penguins can be predicted using two simple variables; the distance to the colony and a measure of the direction of travel towards the shelf edge, whilst avoiding high densities of Pygoscelis penguins from other colonies. Additionally, we find that the chinstrap penguins breeding on the South Orkney Islands use areas which overlap with frequently used krill fishing areas. If the krill fisheries continue to expand, then impacts on the penguin populations may become evident. With this in mind, we recommend the implementation of monitoring schemes to investigate the effects of prey depletion on predator populations, and to ensure that management is addressed at spatial and temporal scales relevant to ecosystem operation.

Summary of the SOOS West Antarctic Peninsula Working Group meeting held in Cambridge, UK, May 15-16th 2017

Inter-annual variability in diet during crèche (December to February) over 22 years (1989 to 2010) was examined for gentoo penguins Pygoscelis papua breeding at Bird Island, South Georgia (54 º0’S, 38º2’W). Overall, diets comprised 51 % crustaceans and 49 % fish by mass. Crustaceans were present in 89 % of samples and were the main prey (> 50 % by mass) in 10 years of the study. Antarctic krill Euphausia superba were present in 85 % of all diet samples. Fish were present in 79 % of samples and were the main prey in 12 years, with Champsocephalus gunnari and Lepidonotothen larseni the most frequently recorded species, in 51 % and 33 % of samples respectively. The energy or mass of krill in the diet was the most reliable predictor of breeding success (the number of chicks fledged per breeding pair); the correlation between model-predicted and observed values was 0.58. We compared annual patterns of gentoo penguin diet variability with those of macaroni penguins Eudyptes chrysolophus breeding at the same location. Our results suggest that the availability of krill is a key source of diet variability for both species, but their diets indicate that gentoo penguins are generalist predators (feeding on pelagic and bentho-pelagic prey), while macaroni penguins are krill specialists (feeding on pelagic prey). Differences in the response to variability in key prey species is an important factor separating the ecological niches of these two sympatric krill predators.

We show that the preferred habitat models developed by Warwick-Evans et al. (2017) for chinstrap penguins breeding at the South Orkney Islands, can be modified to predict the at-sea distribution of chinstraps breeding at the South Shetland Islands, based on the distance from the colony, and the bearing to the shelf edge, adjusting for the at-sea density of Pygoscelis penguins from other colonies. The models performed well, even when using lower resolution tracking data. Cross validations between colonies and between regions indicated that our models perform better if local tracking data were available. Nevertheless, we conclude that these models can, with some confidence, be extrapolated to predict the foraging locations of animals during the breeding season from untracked colonies, even in different regions and across different archipelagos.

Based on mean summer (January to March) near-surface (averaged over 50 m to 150 m) oceanographic flows derived from a high resolution oceanographic model developed using NEMO, we report that chinstrap penguins prefer habitats with slow-flowing, nearshore waters over shallow bathymetry, but that they also sometimes move towards the edge of the continental shelf where the faster-flowing waters of the Coastal Current or fronts of the Antarctic Circumpolar Current exist. The preferred habitats are also locations preferentially used by the commercial krill fishery. Cross-shelf submarine canyons occur near to a number of preferred krill fishing locations, suggesting that these canyons potentially play an important role in krill transport onto the shelf. In the relatively slow-moving shelf waters krill depletion is probably more likely and krill replenishment less likely. Our results suggest that better understanding of krill retention and depletion in areas used by natural predators and the krill fishery are needed, and that  management strategies for the fishery should include assessment of how krill movement can satisfy the demands of both predators and the fishery across a range of spatial and temporal scales.

Effective management of the Antarctic krill (Euphausia superba, hereafter krill) fishery requires an understanding of the controls on the distribution and abundance of krill.  Particularly important are the pathways and timescales of movement and retention affecting krill distribution at scales relevant to the fishery and predators.  This paper reports the initiation of a project that will use numerical ocean-sea ice models at differing resolutions to improve current understanding of the regional and local-scale processes that influence the distribution of krill in CCAMLR Area 48.  Modelling will focus on the South Orkney Islands region, an area that is crucial for determining the overall availability of krill across the Scotia Sea region.  The results will help inform WG-EMM activities on the development of feedback management procedures and provide the present-day context for considering the potential impacts of climate change on this regional system.

The Integrating Climate and Ecosystem Dynamics in the Southern Ocean (ICED) programme is undertaking integrated circumpolar analyses to improve our understanding of Southern Ocean change, the implications for ecosystems and for management. ICED’s current major focus is to build on this work to more comprehensively assess (and where possible quantify) key impacts of change on Southern Ocean ecosystems. This will be achieved through the analysis and integration of available data together with development of models, scenarios and projections. These activities and their outputs will provide valuable information for ecosystem-based management in the region. Following discussions between ICED and CCAMLR to develop joint activities in this regard, planning is underway for an ICED Workshop focussed on projections of change in Area 48 - Developing projections of the future state of Southern Ocean ecosystems: incorporating uncertainties associated with climate variability and change in CCAMLR’s decision making. This Workshop will be held in Hobart in April 2018 in association with the MEASO2018 conference. This paper provides a brief update of recent ICED activities of relevance to WG-EMM and presents current plans for the Workshop, to which we welcome further input from the WG-EMM.

The shelf regions around South Georgia and the South Orkney Islands are highly productive, supporting large colonies of higher predators, and with a history of commercial exploitation. Many of the key oceanographic and ecological processes determining the operation of these marine ecosystems operate over small spatial scales of <10 km. Understanding the drivers of variability in such processes requires the development of high-resolution ocean models. Here we describe the development of regional models for the South Georgia and South Orkney Islands shelves and surrounding regions, and present the results of preliminary analyses. The models are regional applications of NEMO with a ~3 km horizontal resolution, and include key physical processes of relevance to the local ecosystems including tides, atmospheric forcing from reanalysis, glacial melt, and with sea ice processes incorporated using LIM3. The models have been used to generate 20-year hindcast time series of oceanographic flows and water mass properties. This valuable resource is providing the underlying physical environment for detailed examinations of the controls on the distribution of krill and fish around the islands, their interactions with predators and availability to fisheries. Insight from such studies will help inform WG-EMM activities aimed at developing spatial and feedback management procedures.