Effects of the Exxon Valdez oil spill on marine bird communities in Prince William Sound, Alaska

The supertanker Exxon Valdez ran aground on 24 March 1989, spilling 41 X 10(6) L of oil into Prince William Sound, Alaska. To examine effects of this oil spill on the marine bird community, we analyzed data from 11 survey cruises between June 1989 and August 1991. Cruises were conducted in 10 study bays differing in the magnitude of initial oiling. We gauged bird responses to the spill in terms of habitat use, measured by frequency of bay occupancy and species abundances as functions of initial bay oiling. We focused on community-level measures to obtain a broader perspective than can be obtained from studies directed toward individual species of concern. Effects of the oil spill on community measures were most apparent shortly after the spill but diminished rapidly. Species richness was significantly lower in 1989 than at the same season 1-2 yr later, especially in heavily oiled bays. Species diversity (log-series alpha) was also significantly reduced in more heavily oiled bays in early summer 1989 and 1990, but impacts evident in midsummer and fall 1989 were absent 1 yr later, and there were no significant relationships between diversity and bay oiling after midsummer 1990. Species occurrence in bays was more restricted immediately following the spill than 1-2 yr later, and widespread species were less abundant in early summer and fall 1989 than at the same seasons 1 yr later. This latter pattern was reversed in the midsummer surveys, perhaps because spill clean-up activities attracted large numbers of nonbreeding gulls. We used cluster analysis to define six avian guilds based on ecological characteristics of the species. Species richness of several guilds of birds feeding on or close to the shoreline was negatively related to initial oiling level until early or midsummer 1990, but not thereafter. Of these guilds, the richness of a guild of winter visitant and resident species showed the greatest negative association with initial oiling. However, the richness of guilds of solitary or colonial species that dive and/or feed on fish showed no significant relationships with oiling at any time. Correspondence analysis based on bird community composition indicated clear differences between heavily oiled and unoiled bays in 1989, but overall community composition converged between these sets of bays in subsequent years. Our analyses indicated that the Exxon Valdez oil spill had significant initial impacts on marine bird community structure, although they were not evenly distributed among ecological guilds. Even during the first survey, many species were present in the most heavily oiled bays. Although a few species continued to show spill impacts in late 1991, none of the community measures indicated continuing negative oiling effects. This suggests that, at the community level, recovery was well underway, consistent with observations that seabird habitat had apparently returned to normal in all but a few localized areas by mid-1991. Seabird communities appear to have considerable resiliency to such severe but relatively short-term perturbations, possibly because birds move over a regional scale. It may, therefore, be important to consider regional processes in evaluating recovery following environmental accidents.