Prey detection by bottlenose dolphins, Tursiops truncatus: an experimental test of the passive listening hypothesis DAMON P. GANNON* , NE ´ LIO B. BARROS †, DOUGLAS P. NOWACEK †, ANDREW J. READ * , DANIELLE M. WAPLES * & RANDALL S. WELLS †‡ *Duke University Marine Laboratory, Beaufort, North Carolina, U.S.A. yMote Marine Laboratory zChicago Zoological Society, Chicago, Illinois, U.S.A. (Received 23 October 2003; initial acceptance 26 February 2004; final acceptance 2 June 2004; published online 25 January 2005; MS. number: A9736) Bottlenose dolphins possess a sophisticated echolocation system, but evidence suggests that they use this sensory modality sparingly in the wild. Several authors have noted that soniferous fish are prevalent in the diet of bottlenose dolphins, leading to the hypothesis that these predators detect their prey by passive listening. We tested this hypothesis by performing controlled acoustic playback experiments with free- ranging dolphins in Sarasota Bay, Florida. We used recorded calls of prey fish and sounds of snapping shrimp as experimental and control treatments, respectively, and measured the dolphins’ direction of travel and rate of echolocation as response variables. Dolphins changed their direction of travel significantly, turning towards the sound source when fish sounds were played. In addition, dolphins significantly increased their rate of echolocation immediately following playbacks of fish sounds. The sounds of snapping shrimp elicited neither directional nor echolocation responses. The occurrence of echolocation sounds was low, except following playback of fish sounds. We conclude that bottlenose dolphins use passive listening extensively during the search phase of the foraging process. By listening passively, dolphins may obtain useful information on the identity, number, size and location of soniferous prey. Once dolphins discover prey by passive means, they then appear to use echolocation to track the prey during the pursuit and capture phases. Such judicious use of echolocation suggests that this sensory modality incurs significant energetic or ecological costs. These findings have implications for coevolution of dolphins and their prey with regard to sound production and detection. Ó 2004 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. Bats and toothed whales possess highly developed echo- location systems. More is known about the physiology, neurology and behavioural ecology of echolocation in bats than in dolphins (see Thomas et al. 2003) in part because bats are easier to study. Research on dolphin echolocation has focused on the acoustic characteristics of their sonar signals and on their target detection, ranging and discrimination abilities. These studies show, for example, that trained bottlenose dolphins can detect a 7.62-cm sphere at a maximum range of 113 m and can distinguish subtle differences between objects with regard to size, shape, material composition and wall thickness (reviewed by Au 1993). Despite the remarkable sensitivity of their echolocation system, however, the degree to which dolphins rely on this sensory modality in nature is not clear. Wood & Evans (1980) discovered that recently captured dolphins are often unable to use echolocation for simple tasks, such as obstacle avoidance. Odontocetes also become entangled in fishing nets that they should be able to detect (Au 1994; Hatakeyama et al. 1994), and in most cases, efforts to enhance the acoustic reflectivity of these nets have been unsuccessful at reducing bycatch (Evans & Awbrey 1988; Dawson 1991, 1994; but see Trippel et al. 2003). In the few cases where bottlenose dolphin echolocation has been studied in the wild, individual animals appear to use this sensory system sparingly. Jones & Sayigh (2002) found that bottlenose dolphin echolocation rates vary by geographical region, activity state and group size. Echoloca- tion rates (echolocation bouts per individual) of bottlenose dolphins are lower in Sarasota Bay, Florida than in three sites off North Carolina, are higher during socializing and Correspondence and present address: D. P. Gannon, Mote Marine Laboratory, 1600 Ken Thompson Parkway, Sarasota, FL 34236, U.S.A. (email: damon@mote.org). D. P. Nowacek is now at the Department of Oceanography, Florida State University, Tallahassee, FL 32306, U.S.A. 709 0003–3472/04/$30.00/0 Ó 2004 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. ANIMAL BEHAVIOUR, 2005, 69, 709–720 doi:10.1016/j.anbehav.2004.06.020