Decision support systems for aquaculture licensing

Abstract

Models used to assess the environmental impacts of aquaculture are becoming increasingly numerous and complicated. It is therefore becoming more and more difficult to present these models to non-scientists; even though the ultimate clients of research on a aquaculture impacts are administrators and producers who have to deal with practical considerations and have little time or inclination to deal with the complexities of scientific models. The Aquaculture Research Group within the Habitat Ecology Division has therefore been exploring the development of a decision support system (DSS) as a tool for communicating scientific advice to managers, specifically addressing the use of models to evaluate environmental impacts in order to assess whether the licensing of finfish aquaculture sites is likely to lead to degradation of natural marine habitat. The proposed DSS will incorporate simplified versions of several models along with a geographical data base of relevant hydrographic and other environmental information. The user will be able to enter various scenarios in a simple Fashion (for example, a mouse can be used to specify site locations) and see an evaluation of the proposal based on several scientific models. Although a computer program cannot be expected to represent more than a fraction of the expertise of real scientists, the DSS approach appears to have several advantages; these include the ability to deliver a degree of expertise in remote and isolated regions, and, perhaps most important, a chance for managers to access scientific resources in a private environment which lets them ex lore various options without having to justify their eventual actions to scientists who may not fully appreciate all the pressures which bear on their decisions. The material developed so far includes a prototype of the proposed DSS in form of a computer program which demonstrates what the user interface for a working DSS would look like. The program has only a crude graphical user interface and does not actually interact with a real database, but the models are realistic and the output offers a simplified representation of what a real Decision Support System might provide