The deficiencies of current approaches to managing marine resources an seen to be related to the almost exclusive use of target reference points under conditions where the level of risk of making dangerous decisions is poorly known. Technical advice is only one component of a resource management system and, even if unbiased, does not necessarily lead to risk-averse decision making. The lack of a clear division of powers and of an internal system of checks and balances is perceived as a deficiency of many current resource management systems. Explicitly encouraging reviews of management decisions appears to offer advantages as an efficient and flexible way of promoting precautionary management approaches. Under the 'standard management format' that currently applies in many areas of the world, the advice offered by a resource assessment group or advisory committee of technical experts is considered by a 'fishery management authority', which takes management decisions on behalf of the State or States concerned and whose role is to decide between one or other of the options presented by the technical advisors, as modified by economic and political considerations. The decisions of the Authority affect, for example, quotas, fleet size/effort limits and their allocation among stakeholders, and are taken at this second level. The question must be asked whether the participants in this decision-making phase always have access to an assessment of the risk that their decisions will adversely affect the future condition of the stock, Several sketches of current management systems are provided which illustrate the 'standard format' and others that diverge from it. Current trends seem to point to another management framework, when limited powers of decision-making at the limits of safe harvesting may be delegated by the political decision-making structure to an overview group referred to in this paper as a 'committee on limits and standards' (CLS), This could have the responsibility of deciding, inter alia, on long-term management objectives, limit reference points for exploitation rates and lower limits to biomass, and the risk of, and tolerance for, overshoots of these. It could also pie-negotiate management actions which are obligatory when these pre-agreed limit reference points are reached or exceeded. Its members should include technicians, harvesters and public representatives, including non-exploitative interest groups, and it should be encouraged to adopt a broad interdisciplinary approach. A second group, referred to here as the 'committee for short-term measures and allocations' (CSMA), roughly corresponds to current stock assessment groups and routine managerial implementation of their advice. It works within the framework provided by the CLS and assesses target reference points and the equivalent values for quotas or access that will result in a low probability of the criteria of the CLS being exceeded. Membership of the CLS and CSMA should ideally be mutually exclusive in order to ensure proper reviews, checks and balances between the two groups and to reduce collusion between the requirement for a long-term precautionary perspective and the need to take management decisions rapidly in the short term. (C) 1997 Elsevier Science B.V.
