Document Type



Doctor of Philosophy



Date of Defense


Graduate Advisor

Prof. Robert E. Ricklefs


Patricia Parker

John Blake

Bette Loiselle

Ravinder Sehgal


I characterized a lowland Amazonian assemblage of haemosporidian (“malaria”) parasites (Haemoproteus and Plasmodium) of understory birds by analyzing variation in prevalence (proportion of infected host individuals) among years and host species, documenting diversity of haemosporidian evolutionary lineages, and quantifying host specialization. Using standard molecular methods to screen for haemosporidia in 2488 individual birds from 104 species in the Tiputini Biodiversity Station, Ecuador, I found 21.7% to be infected. Prevalence ranged significantly among years and host species. Forty-five putative evolutionary lineages of haemosporidia were identified, by sequencing part of the cytochrome b (cyt b) gene. Based on a comparative analysis, among host species variation in haemosporidian prevalence related positively to level of sexual dimorphism and negatively to foraging height. I assigned 385 parasite individuals to cyt b lineages. These exhibited a wide range of abundance (one to 91 individuals) and host specialization (one to 23 host species). I quantified host specificity by incorporating both phylogenetic relationships (based on genetic data) and frequency distribution among hosts. Based on null model comparisons, six haemosporidian lineages were more specialized than expected by chance. The hosts of these six haemosporidian lineages were on average more abundant than hosts of generalist lineages, but average body size and survival rate did not differ between hosts of specialists and hosts of generalists. Host specificity was also phylogenetically conserved among haemosporidia. Consequently, I performed a comparative regression analysis, controlling for the effect of parasite phylogeny, and found no relationships between host specificity and host abundance, body size, or survival rate. Finally, I applied network analysis in combination with null models to test whether the level of reciprocal specialization (where one parasite lineage associates with only one bird species, which harbors no other parasite lineages) is greater in this tropical assemblage than it is in an equivalent temperate assemblage. Assuming coevolution proceeds towards reciprocal specialization, it should be greater in the tropics, where coevolution has historically been hypothesized to be more important in species diversification. I found no evidence for this hypothesis; instead, reciprocal specialization was greater in the temperate site.

OCLC Number


Included in

Biology Commons