Historical biogeography and community ecology of avian malaria in the Lesser Antilles

Avian malaria is a caused by a diverse group of apicomplexan blood parasites of the genera Haemoproteus and Plasmodium. Because of their strong impact on some bird populations, these parasites have been an important model system for studies of host-parasite relationships. The taxonomy of malaria parasites, which forms our understanding of the distribution of parasites across hosts, has been based largely on morphological traits, which are difficult to distinguish and do not correlate well with recent molecular analyses. As a result, little is known about the distribution of evolutionarily independent parasite lineages with respect to their hosts and geography. Here, I present a series of studies that employ molecular techniques to identify and describe the distributions of genetically distinct parasite lineages. This research characterizes host-parasite relationships in avian malaria across multiple spatial and temporal scales to assess the primary ecological and evolutionary factors that influence the distribution of malarial parasites both geographically and among their host species.; Avian parasite lineages exhibited strong associations with their hosts. While the majority of parasite lineages found in these studies were recovered from more than one host species, they typically were restricted in their distribution to a small number of closely related hosts. This constraint is reflected in a broad phylogenetic survey of the parasites revealing a long coevolutionary history between the lineages and their hosts. Switching to new hosts does occur, though it appears to be limited in part by evolutionarily independent dynamics of host resistance and parasite virulence and additionally by competition with other parasite lineages. Contrary to classic host-parasite literature, neither host specialization nor generalization were associated with parasite abundance or geographic distribution. Furthermore, host breadth was phylogenetically independent within each genus of parasite indicating that specialization is evolutionarily labile.; Despite the strong effect of host species on the distribution of parasite lineages, geographic patterns revealed additional limitations to parasite dispersal. In a global survey, closely related parasite lineages had a high probability of being from the same geographic region, although this signal was weaker at greater depths in the parasites phylogeny, indicating historical long distance dispersal events. A regional analysis of parasite lineages in the Lesser Antillean avifauna revealed some widely distributed lineages, spanning from North America, through the Caribbean region and reaching South America. Nonetheless, restricted ranges of some parasite lineages and gaps in the distributions of some of the widespread lineages indicate that dispersal is limited and that extinction of island populations can outpace recolonization.; A temporal comparison of two insular parasite communities suggests that the frequency of these extinctions may occur as rapidly as one lineage per decade for assemblages with 14 parasite lineages. However, endemic parasite lineages within the same island system exhibit long and stable histories in the region, indicating that lineage turnover is not as frequent for all parasites. Though the composition of parasite communities changes over space and time, the overall prevalence of infection remains relatively constant providing support for the idea that parasite lineages compete within one another and engage their hosts in dynamic coevolutionary interactions. In sum, patterns of infection by avian malaria parasites reflect historical evolution of these parasites with their hosts in a local community context, as well as occasional shifts between host species and dispersal to distant locations, resulting in a complex and ecologically dynamic host-parasite system.