ABSTRACT
We estimated the survival probability of breeding European storm petrels before, during and after a severe oil-spill. We hypothesized that petrels might have deserted the breeding colony to maximize their own survival probability and we expected no major change on adult survival probabilities as a consequence of the spill. We used an information-theoretical approach and multi-model inference to assess the strength of the evidence in favour of different hypotheses. Evidence contained in the data clearly supported the non-effect of the spill on adult survival hypothesis while punctual impact of the spill on survival and expanded (3 years) impact alternatives received less support. The effect size of the spill on averaged survival estimates was negligible in every case. We suggest that petrels minimized the impact of acute pollution by not investing in reproduction. We suggest that short-medium term management actions after oil-spills and similar catastrophes should focus on ecosystem restoration.
