Molecular and vocal evolution in loons (Aves: Gaviiformes).

Abstract

This dissertation research examined evolutionary questions focused on a small but enigmatic order of birds, the loons (Aves: Gaviiformes), ranging from inquiries on the proximate function of conspicuous vocal behaviors, to building theoretical connections between molecular and behavioral evolution. Taking a synthetic approach, I used both molecular genetic and vocalization data from several species of birds to test various predictions generated from hypotheses based in the fields of molecular evolution, phylogenetics and animal behavior. Mitochondrial and nuclear DNA sequence data were used to test hypothesized relationships among avian orders. These data supported a sister relationship between loons and penguins (Aves: Sphenisciformes). Various lines of evidence indicated that the diversification of these avian orders was ancient and rapid. Among loon species, there was strong support for a non-conventional hypothesis of relationships where <italic>Gavia pacifica</italic> is sister to <italic> G. immer</italic> and <italic>G. adamsii</italic>, rather than sister to the morphologically similar, <italic>G. arctica</italic>. I examined the significance of the male-specific territorial yodel of <italic>G. immer</italic> through statistical analyses of audiospectrographic parameters and field experiments. These analyses showed that yodels were identifiable to individual male loons, and that introductory and repeat phrases encode similar amounts of information. I further showed with a series of playback experiments that territorial birds discriminate between yodels of familiar and unfamiliar males and I discuss the role of natural selection in establishing these behavioral phenomena. I discuss the evolution of territorial yodeling behavior in all loon species, and propose explanations for its presence in <italic>G. stellata </italic> females and its absence in females of the four other species. Several analyses further evaluated the evolution of loon vocal behavior, including a cladistic analysis of vocalizations based on discrete vocal characters, an inter-observer agreement technique based on overall similarity of audiospectrographs, and an application of mapping behavioral characters onto the mitochondrially-derived phylogeny. I showed that one seemingly unique vocalization, the tremolo of <italic>G. immer</italic> and <italic>G. adamsii</italic> is a duplicated and modified version of another call, the wail. These data demonstrated that general biological processes (i.e. duplication, mutation, etc.) that generate novel phenotypic characters from ancestral characters (e.g. paralogous genes, serially homologous organs) also generate novel behavioral traits.