BackgroundPrevious studies have shown that haemosporidian parasites (Haemoproteus (Parahaemoproteus) and Plasmodium) infecting passerine birds have an evolutionary history of host switching with little cospeciation, in particular at low taxonomic levels (e.g., below the family level), which is suggested as the main speciation mechanism of this group of parasites. Recent studies have characterized diverse clades of haemosporidian parasites (H. (Haemoproteus) and H. (Parahaemoproteus)) infecting non-passerine birds (e.g., Columbiformes, Pelecaniiformes). Here, we explore the cospeciation history of H. (Haemoproteus) and H. (Parahaemoproteus) parasites with their non-passerine hosts.MethodsWe sequenced the mtDNA cyt b gene of both haemosporidian parasites and their avian non-passerine hosts. We built Bayesian phylogenetic hypotheses and created concensus phylograms that were subsequently used to conduct cospeciation analyses. We used both a global cospeciation test, PACo, and an event-cost algorithm implemented in CoRe-PA.ResultsThe global test suggests that H. (Haemoproteus) and H. (Parahaemoproteus) parasites have a diversification history dominated by cospeciation events particularly at the family level. Host-parasite links from the PACo analysis show that host switching events are common within families (i.e., among genera and among species within genera), and occasionally across different orders (e.g., Columbiformes to Pelecaniiformes). Event-cost analyses show that haemosporidian coevolutionary history is dominated by host switching and some codivergence, but with duplication events also present. Genetic lineages unique to raptor species (e.g., FALC11) commonly switch between Falconiformes and Strigiformes.ConclusionsOur results corroborate previous findings that have detected a global cospeciation signal at the family taxonomic level, and they also support a history of frequent switching closer to the tips of the host phylogeny, which seems to be the main diversification mechanism of haemosporidians. Such dynamic host-parasite associations are relevant to the epidemiology of emerging diseases because low parasite host specificity is a prerequisite for the emergence of novel diseases. The evidence on host distributions suggests that haemosporidian parasites have the potential to rapidly develop novel host-associations. This pattern has also been recorded in fish-monogenean interactions, suggesting a general diversification mechanism for parasites when host choice is not restricted by ecological barriers.