The Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) was a pioneer in the development of what has become known as the ‘ecosystem approach’ to the regulation of fisheries. An ecosystem approach does not concentrate solely on the species fished, but also seeks to avoid situations in which fisheries have a significant adverse effect on ‘dependent and related species’. CCAMLR has to develop management approaches which assess the status of the ecosystem and its health. In the application of this ecosystem approach, CCAMLR has tackled with the difficulty of describing the full complexity of marine ecosystems by assuming that the system is dominated by the complex of species most important in the food chain. This paper describes where the Scientific Committee of CCAMLR stands 17 years after the convention came into force in 1982.

The objectives of the Convention, as laid down in its Article II, were translated into scientific meaningful working hypotheses as a first step. The precautionary approach to management was found to be most appropriate to guide all CCAMLR activities towards the regulation of exploitable resources given the considerable uncertainty associated with data collected in vast and largely unknown regions, and the complexity of the underlying marine systems. In addition to the core problem of developing multispecies exploitation strategies for the marine resources, CCAMLR is currently faced with three other problems: the incidental mortality of seabirds in fisheries, particularly longline fisheries; the entanglement of animals in marine debris; and the impact of fishing on the seabed.

CCAMLR has several approaches to regulating exploitation in the Southern Ocean. It collects data to follow as closely as possible the development of exploited stocks and newly developing fisheries. It also develops models to deal specifically with uncertainty in data collection. CCAMLR draws on five main sources to improve data collection: fisheries catch and effort statistics; biological information and data on by-catches of fish in commercial fisheries; seabirds and marine mammals caught during commercial operations and collected by national and international scientific observers; biological information collected during scientific and fishery-independent surveys; and biological information on krill and dependent species collected as part of the CCAMLR Ecosystem Monitoring Program.

A number of models have been or are currently being developed. The ‘Krill Yield Model’ was developed to provide precautionary limits for annual yields. By multiplying the estimate of krill biomass by a factor to take account of many uncertainties in data, and to thereby derive a precautionary catch limit. The factor is currently fixed at 0.116. A very similar approach, termed the ‘Generalised Yield Model’, has been applied to fisheries for finfish to account for the fact that CCAMLR lacks the time series of catch, effort, length and age available to many other fisheries organisations. Other areas of model development concern functional relationships between krill and krill predators. The development of models has just started. As a first step, the ‘Critical Period Distance’ model was developed but more complex models are likely to follow in the near future.

Decision rules have been developed for objective scientific analysis. Decision rules, specifying the set of decisions that are made in setting, removing, or varying management measures, using the results of assessments of the status of a harvested resource, are under development. They have so far been applied to catches in the krill fishery and the fisheries on Patagonian toothfish.

All the above steps are part of what is called the ‘multispecies approach’. CCAMLR’s use of the multispecies approach to this problem is innovative, so there is little experience in this type of assessment in fisheries conventions elsewhere. The first step in developing a sustainable harvesting strategy for krill was the single-species model of potential krill yield described above. The next step was to develop a model which took the needs of the krill-dependent predators into account. This is then followed by the choice of appropriate parameter estimates for the model. Finally, the functional relationships need to be defined. Unfortunately there are yet no integrated datasets available to test the model as a whole. However, subsets of data are already available which will allow testing of various aspects of the model.

Given the complexity and dynamics of the Southern Ocean, CCAMLR is still far from reaching its ultimate goal, the ‘ecosystem approach’ to the regulation of fisheries. However, in its young history, CCAMLR has made important steps towards the development of an integrated approach to fisheries management and has, in many respects, been the fisheries organisation that has taken the lead in developing such approaches.