eh ARTICLE IN PRESSU N

OChemical alarm cues released from injured tissue are not released under any other context and therefore reliably inform nearby prey of the presence of a predator. Laboratory and field studies have demonstrated that most aquatic taxa show antipredator responses to chemical alarm cues. Ostariophysan fish (e.g. minnows) possess specialized skin cells that contain an alarm chemical. Magurran et al. (1996, Proceedings of the Royal Society of London, Series B, 263, 1551e1556) were the first to use underwater video to carefully document the behavioural response of free-ranging wild populations of minnows to minnow alarm cues. They found no evidence of an antipredator response, and challenged the assumption that the contents of these cells indicate risk in the field. They proposed that alarm responses are context dependent in that they are an artefact of enclosed environments such as laboratory aquaria and field traps. Here, we repeat their experiment on free-swimming field populations of littoral fish and report a significant decrease in the number of fish in areas where chemical alarm cues of blacknose shiners, Notropis heterolepis (Ostariophysi: Cyprinidae) were released. The effect of these chemical cues was equal in magnitude to the effect of the presentation of a model predator. The response to the approach of a model predator (visual cue) was intensified by pre-exposure to chemical alarm cues. We corroborated this interaction between chemical and visual indicators of predation risk in a laboratory study using glowlight tetras, Hemigrammu