EFFECTS OF THE NELLA-DAN OIL-SPILL ON THE FAUNA OF DURVILLAEA-ANTARCTICA HOLDFASTS

The effects of a small spill of light marine diesel on the community structure of invertebrates inhabiting Durvillaea antarctica holdfasts were examined for samples taken from sites which received heavy, moderate and no oiling at Macquarie Island, sub-Antarctic. Non-metric multidimensional scaling (MDS) at both the species and a mixed taxonomic level separated samples collected at oiled sites from the controls. Pre-impact data were only available for 1 of the heavily oiled sites and these samples grouped with those from control sites in MDS analysis. Holdfasts from oiled sites were characterised by heavy sediment loads, a dominance of capitellid, cirratulid and spionid polychaetes and the rarity of the isopod Limnoria stephenseni. In contrast, samples from the control sites were dominated by peracarid crustaceans and in particular by the herbivorous isopod L. stephenseni. Analysis of differences in the size distribution patterns of the 4 most common species (the isopod L. stephenseni, the amphipods Parawaldeckia kidderi and Hyale novaezealandiae and the gastropod Laevilittorina caliginosa) suggests that impacts occurred across all sizes for L. stephenseni and P. kidderi. At the heavily oiled sites, size distribution patterns for the gastropod L. caliginosa were skewed towards larger individuals, suggesting a differential mortality of smaller individuals. We suggest that the rarity of L. stephenseni at the oiled sites may result in the accumulation of sediment within the internal chambers of the holdfast, producing a habitat more suitable for polychaete worms. The residual levels of hydrocarbons within these sediments may further restrict the polychaete species to opportunistic taxa which were not found at the control sites. As this study did not include pre-impact sampling, the interpretation of the results is partially confounded with possible, unmeasured, pre-existing site-specific differences in the community structure. While the availability of pre-impact data from 1 oiled site (from an earlier study) has strengthened interpretation with respect to possible oil effects, further sampling of the sites is needed to resolve some of these uncertainties.