|
Résumé :
|
One of the classic ecological questions is how predators affect population size. This is often assessed by measuring how many individuals are killed by a predator, yet such direct effects may only be a relatively minor part of population dynamics. Predators frequently affect prey populations indirectly, with the fear of predation resulting in costly behavioural compensation that has the potential to lead to large population and community effects. Large observable lethal effects may then just represent the most easily observed "special" cases of the effects of predation on populations, with the costs of non-lethal effects being ubiquitous and usually dominant. This review explores these two ideas: that both cases where there are no population effects due to predation and those where lethal effects dominate are unusual and involve special circumstances. First, systems in which predation effects appear not to arise include (1) complete avoidance of predators by prey; (2) when other environmental factors limit populations so that predation is not additive to mortality; (3) when there are other more vulnerable prey for a predator; (4) when predators interact; (5) because the relationship of perceived predation risk with predator abundance is usually a non-linear function; (6) for the simple reason that non-lethal effects have not been considered. Second, lethal effects tend to dominate over non-lethal effects when (1) there is a high cost of compensating for predation risk associated with either a resource constraint or a particularly vulnerable niche or life-history stage (e.g. the nest stage generally for birds); (2) prey are the most popular prey of a predator or linear trophic chains operate; (3) there is evolutionary lag, such as introduced predators and na
|