The question as to whether predators preferentially kill sick or disabled individuals has been addressed by many ecologists working with different predator–prey systems. Rau & Caron (1979) showed that heavily infected moose were more susceptible to hunting. Kruuk (1972) observed that hyenas appeared to select sick animals in the Serengeti. Vorisek et al. (1998) demonstrated that voles infected with a species of Frenkelia were taken more frequently by buzzards. As a broad generalization, wherever prey capture is difficult and involves large energy expenditure a greater proportion of sick animals seems to be captured (Fitzgibbon & Fanshawe 1989, Holmes & Bethel 1972, Temple 1987). In a host–parasite association where the prey species is an intermediate host and the predator is the definitive host, the capture of the prey is often an essential part of the life cycle. Therefore any mechanism which makes the prey susceptible to predation would enhance the parasite's fitness. In such relationships the susceptibility induced by the parasite can be very specific to the predator host (Levri 1998). Freedman (1990) suggested that a mutualistic association between the predator and parasite might exist. A mutualistic relationship can be said to exist between a predator and a parasite if the cost of harbouring the parasite is less than the benefit of greater success in catching the prey. There is perhaps no demonstrated example of such a mutualism in natural populations since it is difficult to weigh the parasite cost against the predation benefit.