Abstract:
Understanding the dynamics of multihost parasites and the roles of different host species
in parasite epidemiology requires consideration of the whole animal community. Host communities
may be composed of hundreds of interacting species, making it necessary to simplify the problem.
One approach to summarizing the host community in a way that is relevant to the epidemiology of the
parasite is to group host species into epidemiological functional groups (EpiFGs). We used EpiFGs to
test our understanding of avian malaria (Plasmodium
and Haemoproteus) dynamics in four communities
of wetland-associated
birds in southern Africa. Bird counts and captures were undertaken every
2–4 months over 2 yr and malaria was diagnosed by nested PCR. One hundred and seventy-six
bird
species were allocated to a set of EpiFGs according to their assumed roles in introducing and maintaining
the parasite in the system. Roles were quantified as relative risks from avian foraging, roosting,
and movement ecology and assumed interaction with vector species. We compared our estimated
a priori risks to empirical data from 3414 captured birds from four sites and 3485 half-hour
point counts.
After accounting for relative avian abundance, our risk estimates significantly correlated with the observed
prevalence
of Haemoproteus but not Plasmodium. Although avian roosting height (for both malarial
genera) and movement ecology (for Plasmodium) separately influenced prevalence, host behavior alone
was not sufficient to predict Plasmodium patterns in our communities. Host taxonomy and relative abundance
were also important for this parasite. Although using EpiFGs enabled us to predict the infection
patterns of only one genus of heamosporidia, our approach holds promise for examining the influence of
host community composition on the transmission of vector-borne
parasites and identifying gaps in our
understanding of host–parasite interactions.
