Résumé
Investigating the relationships that link marine top predators and their prey is crucial for an understanding of the mechanisms that operate within marine food chains. Many seabird species capture their prey underwater, where direct and continuous observation is difficult. However, in a captive setting, predator–prey interactions can be studied under controlled conditions and in great detail. Using an underwater video-array, we investigated the prey–capture behaviour of a foot-propelled pursuit diver, the double-crested cormorantPhalacrocorax auritus, targeting juvenile rainbow troutOncorhynchus mykiss. We tested the effects of prey density, prey size, light conditions and prey behaviour (schooling vs. solitary trout) on the foraging performance of 9 cormorants. As predicted, prey density exerted the strongest influence on cormorant foraging success. While we found an apparently linear relationship between prey density and prey capture rate, a density below the threshold of about 2 g m–3resulted in disproportionately lower catch per unit effort (CPUE) values. If such a threshold density exists in a natural setting, it could have important implications for birds confronted with a decline in food abundance, when density levels will be reduced. We also demonstrate the marked impact of fish behaviour on the predatory performance of cormorants. Capture success of cormorants was significantly lower and pursuit duration significantly higher when birds attacked schooling rather than solitary trout. By contrast, prey size and light conditions did not have a measurable effect on cormorant prey–capture performance. Our study is an experimental investigation into the prey–capture performance of an avian pursuit diver within a captive setting. We provide input values that should be incorporated into ecological models, which might help to understand predator requirements in a changing environment.