ABSTRACT
Biomarkers of fish health are recognised as valuable biomonitoring tools that inform on the impact of pollution on biota. The integration of a suite of biomarkers in a statistical analysis that better illustrates the effects of exposure to xenobiotics on living organisms is most informative; however, most published ecotoxicological studies base the interpretation of results on individual biomarkers rather than on the information they carry as a set. To compare the interpretation of results from individual biomarkers with an interpretation based on multivariate analysis, a case study was selected where fish health was examined in two species of fish captured in two ports located in Western Australia. The suite of variables selected included chemical analysis of white muscle, body condition index, liver somatic index (LSI), hepatic ethoxyresorufin-O-deethylase activity, serum sorbitol dehydrogenase activity, biliary polycyclic aromatic hydrocarbon metabolites, oxidative DNA damage as measured by serum 8-oxo-dG, and stress protein HSP70 measured on gill tissue. Statistical analysis of individual biomarkers suggested little consistent evidence of the effects of contaminants on fish health. However, when biomarkers were integrated as a set by principal component analysis, there was evidence that the health status of fish in Fremantle port was compromised mainly due to increased LSI and greater oxidative DNA damage in fish captured within the port area relative to fish captured at a remote site. The conclusions achieved using the integrated set of biomarkers show the importance of viewing biomarkers of fish health as a set of variables rather than as isolated biomarkers of fish health.
