Biological Reviews

Review Article

Modelling Southern Ocean ecosystems: krill, the food-web, and the impacts of harvesting

S. L. Hill a1, E. J. Murphy a1, K. Reid a1, P. N. Trathan a1 and A. J. Constable a2
a1 British Antarctic Survey, Natural Environment Research Council, Madingley Road, High Cross, Cambridge, CB3 0ET, UK
a2 Australian Antarctic Division, Australian Department of Environment and Heritage, Channel Highway, Kingston, Tasmania, 7050, Australia

Article author query
hill sl   [PubMed][Google Scholar] 
murphy ej   [PubMed][Google Scholar] 
reid k   [PubMed][Google Scholar] 
trathan pn   [PubMed][Google Scholar] 
constable aj   [PubMed][Google Scholar] 


The ecosystem approach to fisheries recognises the interdependence between harvested species and other ecosystem components. It aims to account for the propagation of the effects of harvesting through the food-web. The formulation and evaluation of ecosystem-based management strategies requires reliable models of ecosystem dynamics to predict these effects. The krill-based system in the Southern Ocean was the focus of some of the earliest models exploring such effects. It is also a suitable example for the development of models to support the ecosystem approach to fisheries because it has a relatively simple food-web structure and progress has been made in developing models of the key species and interactions, some of which has been motivated by the need to develop ecosystem-based management. Antarctic krill, Euphausia superba, is the main target species for the fishery and the main prey of many top predators. It is therefore critical to capture the processes affecting the dynamics and distribution of krill in ecosystem dynamics models. These processes include environmental influences on recruitment and the spatially variable influence of advection. Models must also capture the interactions between krill and its consumers, which are mediated by the spatial structure of the environment. Various models have explored predator-prey population dynamics with simplistic representations of these interactions, while others have focused on specific details of the interactions. There is now a pressing need to develop plausible and practical models of ecosystem dynamics that link processes occurring at these different scales. Many studies have highlighted uncertainties in our understanding of the system, which indicates future priorities in terms of both data collection and developing methods to evaluate the effects of these uncertainties on model predictions. We propose a modelling approach that focuses on harvested species and their monitored consumers and that evaluates model uncertainty by using alternative structures and functional forms in a Monte Carlo framework.

(Received October 7 2005)
(Revised July 10 2006)
(Accepted July 11 2006)
(Published Online September 21 2006)

Key Words: ecosystem approach to fisheries; ecosystem model; Southern Ocean; Euphausia superba; food-web effects; model uncertainty; CCAMLR.