Multivariate approach for identifying environmental indicator species in estuarine systems

摘要

Over time, several indices have been developed for assessing estuarine environmental conditions. In many cases, these indicators were chosen at least partly based on human value judgements that may or may not be justified or defensible. We present a new, 2-step multivariate approach that impartially identifies a suite of ecological indicators and then determines their relationship to environmental conditions in San Antonio Bay, Texas. A total of 12 fisheries indicator species were identified by the PRIMER BEST analysis procedure, including Leiostomus xanthurus, Bairdiella chrysoura, Pogonias cromis, Portunus gibbesii, Callinectes similis, Ictalurus furcatus, Harengula jaguana, Polydactylus octonemus, Loliiguncula brevis, Macrobrachium ohione, Opisthonema oglinum, and Libinia dubia. This suite of indicators showed the highest rank correlation with variation in the environmental variables measured. Subsequent redundancy analysis of the relationship between these indicators and environmental variables revealed salinity as the most influential variable shaping the indicator assemblage, with turbidity the second most influential. When freshwater inflow increased or salinity was low, I. furcatus, M. ohione, and O. oglinum were at their highest relative numbers in the assemblage; conversely, when salinity was high, L. brevis, L. dubia, and C. similis were at their highest relative abundances. Additionally, the euryhaline species L. xanthurus and B. chrysoura were negatively related to water turbidity. The 2-step analysis presented provides a statistically robust way to impartially identify biological indicators most responsive to the environmental variables of interest, and thus provides managers with a robust method for monitoring the effects of their management actions.