Application of a battery of biomarkers in mussel digestive gland to assess long-term effects of the Prestige oil spill in Galicia and Bay of Biscay: correlation and multivariate analysis

According to published reports, selected cell and tissue-level biomarkers revealed alterations in the health status of mussels collected after the Prestige oil spill (POS) in Galicia and the Bay of Biscay (2003-2006). Presently, univariate and multivariate analysis provides a comprehensive view of the whole data set. Correlation and principal component analysis (PCA) were performed using biomarkers corresponding to or derived from those data, say: induction, AOX(exp), and inhibition, AOX(eff), of Acyl-CoA oxidase enzyme activity; membrane labilisation period, LP, volume density, Vv(LYS), and surface-to-volume ratio, S/V(LYS), of digestive cell lysosomes; volume density of neutral lipids, Vv(NL); volume density of basophilic cells, Vv(BAS), mean luminal radius to mean epithelial thickness in digestive gland epithelium, MLR/MET; gonad index, GI; cumulative intensity of inflammatory responses, CI(IR) and parasitic infestation, CI(PI). The PCA results were used to classify sampling times and localities by means of hierarchical cluster analysis. Biomarkers were more frequently correlated with naphthalene than with total PAH concentration in mussel tissues. Many biological parameters were correlated with each other in agreement with their involvement in toxic processes or their role in the mussels' response against environmental insult. Naphthalene appears to be the most effective toxic herein, which supports that the reported effects may be mainly due to POS. According to the PCA, 4 groups of biomarkers explained 67.37% variability and the factors that most strongly contributed were S/V(LYS), CI(IR), Vv(BAS) and MLR/MET. These factors were used in the hierarchical cluster analyses carried out. The cluster analysis performed to classify sampling times discriminated 3 significantly different periods after POS: (a) a high affection period (until April-04); (b) the beginning of the recovery; and (c) an advanced recovery status in which geographical variability increased (from April-05 onwards). As regards the clustering of localities, major geographical differences were not detected but three outliers were identified: (a) early and highly impacted localities in Galicia and Bay of Biscay; (b) Agui?o that was characterised by extremely high Vv(BAS) during 2003; and (c) localities that were subjected to persistent chronic pollution.