MEPS 299:205-215 (2005)  -  doi:10.3354/meps299205

ABSTRACT: Examples of behavioural manipulation by parasites are numerous but the consequences of these phenomena on the ecology of host populations remains unclear. Through its effect on Gammarus insensibilis behaviour, the salt marsh trematode Microphallus papillorobustus splits its host population into 2 discrete subunits, one living at the surface (infected individuals) and the other living near the bottom (uninfected individuals). Here, we compare several biological characteristics of gammarids from these 2 spatially segregated subunits. Infected females were smaller and thus less fecund than uninfected females. They also had a longer intermoult duration. The mean body size of infected and uninfected males was not significantly different. However, energy reserves were significantly different: infected males possessed higher glycogen and lipid contents and lower glucose content compared to uninfected males. Interestingly, uninfected males living at the surface (those paired with infected females) also displayed higher energy reserves compared to uninfected conspecifics living in the bottom subunit (those paired with uninfected females). In both of the 2 subunits, there was no significant relationship between male sexual performance (measured through the size of their partner) and their levels of energy reserves. The pattern of size-assortative pairing was not significantly different between pairs from the surface and from the bottom, but male sexual performances were on average smaller for males from the surface subunit. Our results also indicated that the benefits of preferring large more fecund females are however likely to be counterbalanced by the higher mortality rate of these females compared to smaller ones during the subsequent intermoult. Because such a phenomenon directly influences the number of juveniles produced, it reduces the inequalities between the true sexual performances (i.e. number of descendants) achieved by males from the surface and from the bottom. This study supports the idea that infected gammarids are not simply normal hosts with an aberrant behaviour, they are more deeply modified in ways that may substantially alter host population ecology.

KEY WORDS: Parasitism · Host population · Trematode · Amphipod · Life history traits