Effects of parasites on fish behaviour: a review and evolutionary perspective

Fish serve as hosts to a range of parasites that are taxonomically diverse and that exhibit a wide variety of life cycle strategies. Whereas many of these parasites are passed directly between ultimate hosts, others need to navigate through a series of intermediate hosts before reaching a host in (or on) which they can attain sexual maturity. The realisation that parasites need not have evolved to minimise their impact on hosts to be successful, and in many cases may even have a requirement for their hosts to be eaten by specific predators to ensure transmission, has renewed interest in the evolutionary basis of infection-associated host behaviour. Fishes have proved popular models for the experimental examination of such hypotheses, and parasitic infections have been demonstrated to have consequences for almost every aspect of fish behaviour. Despite a scarcity of knowledge regarding the mechanistic basis of such behaviour changes in most cases, and an even lower understanding of their ecological consequences, there can be little doubt that infection-associated behaviour changes have the potential to impact severely on the ecology of infected fishes. Changes in foraging efficiency, time budget, habitat selection, competitive ability, predator-prey relationships, swimming performance and sexual behaviour and mate choice have all been associated with – and in some cases been shown to be a result of – parasite infections, and are reviewed here in some detail. Since the behavioural consequences of infections are exposed to evolutionary selection pressures in the same way as are other phenotypic traits, few behavioural changes will be evolutionarily neutral and host behaviour changes that facilitate transmission should be expected. Despite this expectation, we have found little conclusive evidence for the Parasite Increased Trophic Transmission (PITT) hypothesis in fishes, though recent studies suggest it is likely to be an important mechanism. Additionally, since the fitness consequences of the many behavioural changes described have rarely been quantified, their evolutionary and ecological significance is effectively unknown.