Résumé
Morphometrical parameters of the attachment apparatus and copulatory organs of 52 Dactylogyrus species parasitizing 17 species of cyprinid fishes were analysed to test for the existence of reproductive barriers among congeneric species. The minimal spanning tree (MST) method was applied in the analyses. The position of "real" parasite communities, based on (1) observed infracommunities, (2) a checklist of parasites for a given host in the morphological space, was compared to the position of randomly generated communities using all Dactylogyrus species. The distribution of species similarity within infracommunities (using both attachment and copulatory measurements) was not significantly different from that obtained by simulation, and this trend was similar for both the checklist and observed infracommunities. When real infracommunities were separated according to host specificity (specialists versus generalists), we found differences reflecting similarities in the shapes of attachment and copulatory organs. Within specialists, more similarities in the shape of the attachment apparatus can be found than within generalists, whereas the similarity in copulatory organ shape seems to be random. When generalists are considered, parasite infracommunities with the greater differences in attachment apparatus are also more different in terms of the shape of their copulatory apparatus. We conclude that specialist parasites possess more similarity in attachment apparatus due to specialisation to their host, whereas the species similarity in copulatory organs within infracommunities exhibits a random pattern, but with the copulatory organs being more variable than the attachment apparatus (which may be due to reproductive isolation). The morphology of the copulatory apparatus seems not to be the single factor explaining reproductive isolation among species.