Classification of juvenile flatfishes to estuarine and coastal habitats based on the elemental composition of otoliths

The elemental composition of otoliths from juvenile English sole Pleuronectes vetulus and speckled sanddab Citharichthys stigmaeus living in estuaries and sandy coastal habitats were compared to determine if a chemical habitat tag existed that could be used to differentiate fish living in the alternative juvenile habitats. Juveniles of both species were collected from seven estuaries and 11 coastal sites along 500 km of the central California coast. Collections occurred in three years: 1998, 1999 and 2000. The chemical composition of the otoliths was analyzed using solution-based inductively coupled plasma mass spectrometry. The multi-elemental composition (Sr, Li, Ba, Mn) of otoliths from estuarine and coastal fish differed significantly for all models (global, region and years models) examined in this study. For each species, discriminant function analysis (DFA) for the global model, which pooled juveniles collected from three regions over multiple years, classified fish into estuarine and coastal groups with close to 80% accuracy. The two main elements in the discriminant models, Sr and Li, differed consistently between habitats; Sr was higher and Li was lower in estuarine fish. Classification success was modestly improved by generating separate discriminant functions for each region (regions model) because of some regional differences in Sr and Li concentrations. Classification success also was improved in some cases by generating separate discriminant functions for each year (years model) because Ba and Mn differed between habitats in only some years. Despite this variability in the concentration of elements in some regions and years, a chemical habitat tag was present in each species that could be used to discriminate coastal and estuarine juveniles over a large geographic area and over three years with very different oceanographic conditions (e.g., El Nin˜o, La Nin˜a). In addition, I found that English sole and speckled sanddab had striking similarities in their chemical habitat tags and that, in some cases, one species could be used as a proxy to classify juveniles of the other species without compromising the accuracy of the habitat tag. The ability to use a proxy classification model would significantly reduce the number of juvenile fish that would need to be collected and analyzed in order to classify members of an ecologically similar species. The chemical habitat tags found in this study appears to be promising tools for identifying fish that have lived in alternative juvenile habitats and, ultimately, the proportional contribution of estuarine and coastal habitats to the central California populations of English sole and speckled sanddab.