Co-infections of parasite species in a host may lead to interspecific associations, which can shape the overall community structure . These associations are particularly likely in systems where numerically abundant parasite species infect the same location in a host. In the present paper, we explored the establishment and interspecific associations in two abundant trematode species, Ichthyocotylurus variegatus and I. pileatus, co-infecting the same organs in their second intermediate fish host, Perca fluviatilis. By sampling young cohorts of fish, we investigated how parasites established and co-occurred in naïve 0+ fish which were exposed to the parasites for the first time, as well as in 1+ and 2+ fish that harboured infections also from the previous years. Parasite transmission to the youngest cohort indicated that the seasonality of transmission was very similar between the species. The first metacercariae appeared in fish in July, which corresponds to the timing of infections in the first intermediate snail hosts (Valvata macrostoma) . Afterwards, abundances of both species increased steadily, but the abundance of I. pileatus was consistently higher compared I. variegatus suggesting interspecific differences in transmission dynamics. This concurs with earlier findings reporting lower abundances in I. variegatus compared with the other Ichthyocotylurus species [6, 17]. One possible factor underlying these differences is the spatial heterogeneity in transmission between the species. In our earlier study , we examined trematode infections in Valvata snails in the same littoral habitats (depth <6 m) where the fish were sampled in the present investigation. However, we did not find infections of I. variegatus from those snails in any of the monthly samples , which suggests that the parasite transmission takes place elsewhere. For example, it is possible that I. variegatus is transmitted to perch in deeper areas of the lake where specific conditions such as lower water temperature could result in lower rate of transmission to fish. Such a small-scale interspecific segregation of the transmission is surprising given that both parasite species mature in birds (see discussion in ), which effectively disseminate parasite eggs to the environment. Nevertheless, the co-occurrence of different developmental stages in the 0+ fish suggests that the fish, caught from one location, move actively between these infection 'hotspots'  and become exposed to both parasite species roughly at the same time.
We observed that the recently established metacercariae of I. variegatus were proportionally more common compared to I. pileatus, while the same was true for the encapsulating stages of I. pileatus. This suggests that metacercariae of I. pileatus become more rapidly encapsulated after establishment, but also that the time between encapsulation and maturation in metacercariae of I. variegatus is relatively short. Exact reasons for these developmental differences between the species are unknown, but they may include processes such as the magnitude of host immune responses. For example, the capsules surrounding metacercariae of I. pileatus were consistently thicker compared to I. variegatus, which suggests stronger host responses against that species. This is somewhat surprising given the significantly larger size of I. variegatus, which intuitively would require more resources taken from the host. However, experimental work is needed to address the different hypothetical scenarios related to interactions between the rate of encapsulation, metacercarial development, and magnitude of parasite-induced damage to the host. Detailed description of the metacercarial development and encapsulation can also be used as a complementary tool in morphological identification of the species, which has previously been based solely on the morphometrics of the fully developed metacercariae. For example, the notable interspecific size difference of the metacercariae already in the early stages of development suggests that examination of the internal morphology is not necessarily needed for separation of these species. Moreover, differences in the diameter and shape of the cyst wall (oval in I. pileatus and round in I. variegatus) provide further means for separation of the species.
We also analysed interspecific associations between the species and investigated how these develop with time and accumulation of the metacercariae in hosts of different age. We found that the species were positively associated and this pattern was emphasised in the older age cohorts. Similarly, both parasite species primarily infected the swim bladder and the associations within this site were also positive. Taken together, these results suggest facilitative rather than competitive associations between the species, possibly emerging from the overlapping temporal transmission dynamics (see above) and common interests in transmission to avian definitive hosts [6, 7, 11, 12, 18]. Similar positive associations have recently been described, for example, in monogeneans infecting marine and freshwater fish [22–24] and trematodes infecting eyes of fish . Thus, our results corroborate with this line of evidence reporting non-competitive but also non-random community structure.
However, the highest parasite abundances nevertheless tended to occur in different host individuals and this pattern was consistent throughout the age cohorts of fish (see also ). In other words, the most heavily infected 10% of the fish harboured parasites mainly from one of the species. Mechanisms underlying such a pattern of infection are unclear, but direct competitive interactions between the species seem unlikely. This is because (i) we observed mainly positive associations in lower abundances, (ii) both species mainly infected the same organ in the host, and (iii) the metacercariae are relatively inactive after encapsulation making direct interspecific interactions unlikely. It is also unlikely that fish individuals most susceptible to infection from one of the parasite species would be among the most resistant to the other species (but see below), especially given the positive association in infection in lower abundances. Moreover, these young infracommunities should be far from saturation as the parasite numbers continue to increase in older age classes and may reach even thousands per individual fish . However, it could be that these differences emerge as a result of spatial heterogeneity in exposure to the parasite species (see above). The risk of infection in trematodes is not spatially uniform [21, 27–29] and similar interspecific heterogeneities in spatial exposure exist also in the present system . Under such circumstances, it is possible that individuals occupying different areas in a lake become exposed to infective stages of different parasite species. Such an unequal or sequential exposure to one species may also lead to responses in the host that will influence the community structure later when the host becomes co-exposed to other parasite species . In the present study, all fish were caught from the same specific location, which indicates that habitats of the fish were at least partly overlapping, although this does not exclude the possibility of past heterogeneities in exposure. Regardless of the underlying mechanism, however, this pattern of infection suggests that analyses conducted on different sub-sets of the host population may lead to different interpretations of the nature of the associations between parasite species.