Skip to main content

    Using the EPOC modelling framework to assess management procedures for Antarctic krill in Statistical Area 48: evaluating spatial differences in productivity of Antarctic krill

    Request Meeting Document
    Document Number:
    WG-EMM-06/38 Rev. 1
    Author(s):
    A.J. Constable (Australia)
    Agenda Item(s)
    Abstract

    Spatially explicit simulation models of the Antarctic marine ecosystem are needed to evaluate management procedures for the Antarctic krill fishery. The key issues to be resolved in the development of a harvest strategy are whether (i) spatial differences in the productivity of krill give rise to differential affects on predators in different locations, (ii) movement of krill between locations ameliorate any local fluctuations in krill abundance and (iii) fishery behaviour could be constrained by regional differences in krill dynamics and cause differential impacts on predators as a result, particularly as the fishery expands to take the large-scale catch limit. A fourth issue is to determine whether climate change will impact on the krill-based food web and whether the potential for achieving conservation objectives for predators could be affected by those changes. Productivity of krill can be impacted by sea temperature and available production. Similarly, survivorship and successful recruitment of juvenile krill is likely to be dependent on the dynamics of sea ice. This paper uses the Ecosystem Productivity, Ocean and Climate (EPOC) modelling framework to develop a spatially explicit model of Antarctic krill, Euphausia superba, within a wider ecosystem context (ocean, productivity, krill and predators) in the southwest Atlantic in order to explore the potential for spatial differences in krill productivity and their affects on predator productivity and fisheries. It uses satellite data as proxies for the key physical environmental drivers that may affect productivity. Illustrative results show that spatial and temporal variability of krill productivity is likely and that attention needs to be given to appropriately parameterising models to explore the sensitivity of management outcomes to these differences.