écologie des virus influenza aviaires en Camargue - IRD

Annexe 2between species, types of phenotypic traits that are altered, proportion of manipulativeparasites in the parasitic community, etc). Exploring the idea that ecosystem dynamics andfood web complexity contribute in return to the evolution of manipulation by parasites isanother exciting direction.A full understanding of the roles played by manipulative parasites in ecosystemswould additionally require detailed investigations of the complexity of manipulated hosts,focusing in particular on phenotypic traits other than those that are the most obviouslyaltered. Parasitically modified hosts are not simply normal hosts with one (more or lessstrong) trait alteration (e.g. behaviour); instead they are deeply modified organismsdisplaying a range of subtle modifications [14,51,52]. Certain of these trait alterations,although not spectacular, might be quite relevant in terms of ecosystem consequences. Forinstance, certain acanthocephalans have recently been shown to increase salinity tolerance incrustacean amphipods [53]. Although this physiological change may be less spectacular thanthe behavioural changes generally induced by these parasites (e.g. Figure 1.b), it can behighly relevant from an ecological viewpoint since salinity tolerance is a key determinant ofthe invasion potential of amphipod species [54]. Similarly, for reasons that still need to bedetermined, several studies suggest that aquatic crustaceans whose behaviour is manipulatedby helminths of aquatic birds concomitantly display significantly higher levels of lipids andglycogen than uninfected conspecifics (trematodes: Ref [14]; acanthocephalans: Ref [55];cestodes: Ref [56]). Manipulated hosts are therefore not only easier to capture, they may alsorepresent prey of higher nutritive value for foraging predators [57]. Whatever the exactproximate and ultimate reasons for higher lipid reserves in manipulated hosts, theimplications at the ecosystem level and for conservation are undoubtedly important,particularly when variation in parasite prevalence exists between habitats. We thus encourageother researchers to examine a broader range of phenotypic traits in manipulated hosts thanthe ones of immediate relevance to transmission.Furthermore, research into this area would benefit greatly from an emphasis onquantitative measurements of the impact of manipulators on ecological processes (e.g.predation rates, competition, and energy flow along food chains). The study of hostmanipulation has its roots in a natural history approach that did not encourage rigorousquantitative analyses beyond those directly linked to the manipulation itself. Indeed, for some202