Local predation risk assessment based on low concentration chemical alarm cues in prey fishes: Evidence for threat-sensitivity

Abstract

Individuals that can reliably detect and avoid potential predators should gain significant fitness benefits (Lima and Dill, 1990). Early detection and avoidance of potential predation threats should, therefore, be strongly selected among prey species. Predator avoidance, however, comes with a significant cost associated with the loss of time available for other fitness related activities such as foraging, mating and/or territorial defence (Lima and Dill, 1990; Lima and Bednekoff, 1999). Thus, antipredator behaviour can be conceptualized as a series of 'threat-sensitive' trade-offs between the benefits associated with predator avoidance and those of other fitness related activities. Individuals that are able to reliably assess local predation risk should be at a selective advantage, as they would be capable of maximizing these threat-sensitive trade-offs through context appropriate behavioural responses (Brown, 2003). One of the underlying assumptions of the threat-sensitivity model is that individuals should have sufficient flexibility to exhibit graded responses to a perceived predation threat (Helfman, 1989). Individuals should exhibit stronger responses to a higher perceived predation risk, as this would maximize potential survival benefits. Conversely, if the perceived predation risk is low, individuals should exhibit a less intense antipredator response, as this would allow them to continue gaining benefits from other activities. By adjusting the intensity of antipredator responses, individuals would be able to continue, for example, foraging while still gaining some antipredator benefits. Assessment of local predation risk based on damage-released chemical alarm cues is widespread among freshwater fishes (Chivers and Smith, 1998; Smith, 1999; Wisenden, 2000). Such alarm cues are typically released following mechanical damage to prey, as