Describing benthic impacts of fish farming with fuzzy sets: theoretical background and analytic methods

Abstract

Summary Diver observations have been recorded over a period of several years at a site severely impacted by finfish niariculture. The logged data were analysed using fuzzy logic in order to address the problem of interpreting sometimes qualitative and sub- jective information. The underlying methodology based on fuzzy set theory is developed and described in this paper, along with the techniques used to quantify the relationship between the information recorded by the divers, which was usually of a verbal descriptive nature, and a numerical index of benthic impact. Fuzzy logic proved to be a valuable tool for translating the dive log data into quantitative form. Introduction A 3-year time series of more than 100 dive logs describing the status of the benthic environment beneath a net cage farm in the northern Gulf of Aqaba (Red Sea) has been collected and analysed. These logs cover the productive life of a fish farm site, including observations made after the cages were removed. The original purpose of these logs was an informal collection of qualitative observations made during the frequent dives under the fish farm, in the course ofa geochemical and microbiological study on benthic implications of net cage aquaculture in the oligotrophic Gulf of Aqaba (Angel el al. 1995a). During this study it became evident that valuable information about the sediment dynamics and benthic processes involved in organic enrichment and recovery of the site could be gleaned from the observations. These observations include information that cannot be easily collected quantitatively due to the dynamic nature of the environment. Moreover, the use of diver obser- vations enables sampling of larger areas than would be tech- nically feasible by conventional quantitative measurements of such variables as sediment organic carbon or porewater nutri- ents. Owing to the complexity of events occurring in the benthos, it is not possible to give a clear-cut diagnosis of the health of a site by just looking at the dive logs. This is due to several factors, including the extreme patchiness of the benthic environment, occasionally incomplete dive logs and the apparent con- tradictory nature of certain observations regarding the health of the benthos on any one day. Faced with informative obser- vations that are also semiquantitative, sometimes incomplete, or subjective and conflicting in nature, we decided