International legislation demands that statutory bodies report on the health of aquatic ecosystems. Traditionally, ecosystem components have been characterised according to species assemblages but with limited success in predicting health. On the other hand, many studies based upon functional groupings that include trophic relationships and bioturbation potential have shown response to pollution. However, these and other functional group responses have not yet been linked to broad scale physical variables. To date this has hindered the development of a predictive model of function based on abiotic factors. In addition, most functional studies ignore any potential role of body size when assessing the importance of each species to overall functional group measures. By weighting all species that belong to the same guild equally, the investigator risks overestimating the true importance of any one guild to the environment. This study compared the ability of different functional group approaches to discriminate between separate estuarine sites, whilst linking biotic data with abiotic factors. Using data for the Tamar Estuary, we show that no two methods of classifying the biotic data, according to function, produce the same groupings of sites; nor did any method produce groupings that matched clusters based on abiotic factors alone. Instead, results show that not only can choice of functional method alter our perception of site associations but also, can influence the strength of similarity relationships between abiotic and biotic datasets. Both the use of bioturbation measures and weighting species abundance data by body size provided better relationships between biotic and abiotic data than the use of trophic groups. Thus both methods merit further research to produce algorithms for modelling studies. |
