Abstract
The introduction of nonnative species and reductions in native biodiversity have resulted in substantial changes in vector and host communities globally, but the consequences for pathogen transmission are poorly understood. In lowland Hawaii, bird communities are composed of primarily introduced species, with scattered populations of abundant native species. We examined the influence of avian host community composition, specifically the role of native and introduced species, as well as host diversity, on the prevalence of avian malaria (Plasmodium relictum) in the southern house mosquito (Culex quinquefasciatus). We also explored the reciprocal effect of malaria transmission on native host populations and demography. Avian malaria infection prevalence in mosquitoes increased with the density and relative abundance of native birds, as well as host community competence, but was uncorrelated with host diversity. Avian malaria transmission was estimated to reduce population growth rates of Hawai'i 'amakihi (Chlorodrepanis virens) by 7-14%, but mortality from malaria could not explain gaps in this species' distribution at our sites. Our results suggest that, in Hawaii, native host species increase pathogen transmission to mosquitoes, but introduced species can also support malaria transmission alone. The increase in pathogen transmission with native bird abundance leads to additional disease mortality in native birds, further increasing disease impacts in an ecological feedback cycle. In addition, vector abundance was higher at sites without native birds and this overwhelmed the effects of host community composition on transmission such that infected mosquito abundance was highest at sites without native birds. Higher disease risk at these sites due to higher vector abundance could inhibit recolonization and recovery of native species to these areas. More broadly, this work shows how differences in host competence for a pathogen among native and introduced taxa can influence transmission and highlights the need to examine this question in other systems to determine the generality of this result.