Playing the waiting game: Predator and prey in a test environment

Optimal behaviour of a predator depends on what its prey are doing and vice versa. Yet empirical tests of optimal foraging generally take place with only one of the two species free to adjust its behaviour. Question: Do predators and prey follow predictions of optimality when they are interacting and each is free to adjust its behaviour? Organisms: Commercially purchased common goldfish (Carassius auratus) and wild-caught little egrets (Egretta garzetta). Experimental arenas: Covered circular enclosures (radius 3.5 m), each with three artificial pools (radius 0.76 m). Each pool had a habitat where the goldfish could hide, and an open habitat where they could collect their food but were exposed to predation. Methods: An egret was allowed to forage in an arena each of whose pools contained 15 goldfish (replaced daily according to the number consumed). We measured the time goldfish spent in the open and under cover; the time it took for them to emerge from cover after an egret visit (i.e. the refractory time, rf ); the number of fish eaten; and the time an egret took to return to a pool after leaving it (i.e. the return time, rt). Predictions: Qualitative: during an egret's visit to a pool, fish should spend a greater proportion of time hiding and thus become less and less vulnerable to predation. Quantitative: An egret will maximize its chance to encounter exposed fish by adopting a return time equal to fish refractory time (rt = rf ). An egret should maximize its catch of fish if rt = rf. Results: In the presence of an egret, goldfish significantly reduced the time they spent in the open (2.09% vs. 65.4%). The average goldfish refractory time was 20.31 min. The average little egret return time was 21.09 min. These values are not significantly different. Egrets catch the most fish if rt = 18.6 min (not significantly different from that observed).