Skip to main content

    Predicting the presence and abundance of Antarctic krill (Euphausia superba) in the waters of the South Orkney Island Archipelago

    Request Meeting Document
    Document Number:
    WG-EMM-2022/16
    Author(s):
    J.J. Freer, V. Warwick-Evans, G. Skaret, B.A. Krafft, A. Lowther, S. Fielding and P.N. Trathan
    Submitted By:
    Dr Vicky Warwick-Evans (United Kingdom)
    Approved By:
    Dr Chris Darby
    Abstract

    Antarctic krill (Euphausia superba) is a fundamental species within the Antarctic ecosystem yet also target of the largest commercial fishery in the Southern Ocean. An understanding of the distribution and abundance of krill across appropriate spatial and temporal scales is necessary to implement a new management framework for the krill fishery, endorsed by the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR). As a contribution towards this, we create a dynamic krill distribution model for the waters surrounding the South Orkney Islands Archipelago (SOI) and the wider CCAMLR Statistical Subarea 48.2, a region of long standing importance for krill fishery operations. This study uses data from a spatially and temporally consistent krill-targeted acoustic survey (2011-2020) and year-specific environmental predictors within a two-part hurdle model. This approach combines a binomial regression to model the processes driving the presence of krill with a general additive model to model the processes that influence krill density. Predictors found to be important in both hurdle components were distance from shelf break, distance from summer sea ice extent, and salinity. Year-specific projections of krill distribution revealed that the shelf break surrounding the SOI, particularly the northern shelf break, was a consistently important area for krill. This location is an area of intensive krill fishery activity and chinstrap penguin foraging. Inter-annual variability in predictions is consistent with reported fisheries catch data. Finally, models were used to project krill presence and density under conditions for the year 2021. Projections reveal low probability of krill presence and the combined hurdle model estimated krill densities to be an order of magnitude lower than previous years. This aligns with reports of poor breeding success in krill predators at SOI but remains untested against similar acoustic data. Overall, the model we present provides improved understanding about krill habitat and distribution within this important region.