Hornbills (Aves: family Bucerotidae) are a charismatic group of Palaeotropical birds recognised for their distinctive morphology (casque) and behaviour (nest-sealing). Hornbill diet, habitat use, distribution and social system display pronounced interspecific variation, and their mutualistic interactions with tropical fruits provide vital ecosystem services. A wide range of species of hornbills across the Afrotropics, Indo-Malaya and Melanesia are of conservation concern. However, the evolutionary history of hornbills remains unclear and there are conflicting hypotheses about their origin, tempo of diversification and biogeography. Despite a comprehensive cladistic analysis of phenotypic data, there are unresolved taxonomic uncertainties within the family, and although a gradual accumulation of molecular data has revealed interesting phylogenetic relationships, methodological limitations, and incomplete sampling, has left gaps and produced incongruent results. Hence the evolutionary framework against which to interpret the diversity of this group is incomplete. The aim of this thesis was to construct a well-resolved molecular phylogeny of hornbills, and to use it to address longstanding questions about the evolution and diversification of these remarkable birds. This thesis presents a broad phylogenetic and biogeographic analysis of the family Bucerotidae, based on a coalescence of molecular methods and comparative analyses, largely using historical samples and recently-developed bioinformatic approaches. I provide the first complete species-level molecular phylogeny of hornbills, derived from nuclear adenylate kinase 1 intron 5 and mitochondrial (mtDNA) cytochrome b genes, and also a comprehensive mtDNA phylogeny covering 98% of the taxa, with extensive sampling of Asian geo-isolates. Using these two phylogenetic trees, I sought to determine the tempo of divergence, trace the evolution of traits, identify ancestral areas and colonisation routes, and also calculate genetic divergence. In this part of the work, I stress the importance of (1) using historical samples, (2) calibrating time-trees with fossil and molecular anchor-points, and (3) the use of a complete tree to test models of diversification and reconstruct ancestral states. My findings confirm the monophyly of Bucerotidae, showing nearly distinct African and Asian lineages, with relatively congruent topologies across different phylogenetic methods and genes. In turn, these gene trees were comparable with previous cladistic analysis based on phenotypic data. As a result, I was able to resolve some taxonomic issues and propose generic changes. Comparative analyses of social behaviour revealed that cooperative breeding is an ancestral trait, and its evolution in hornbills (in contrast with some other taxa) is associated with stable environments in combination with frugivory, territoriality and reduced body size. Analysis of the evolution of diet with diversification showed that the shift to frugivory from faunivorous African ancestors influenced the rapid diversification of Afro-Asian forest hornbills, facilitated by their preference for humid forests and mutualistic interactions with , hornbill-dispersed fruits (HDF). This dispersal of frugivorous lineages via the Palaeotropical Biotic Interchange promoted colonisation of Asia, with Sundaland being the centre of radiation for continental and insular Asian species. The gradual eastward colonisation from India to Melanesia matches palaeo-tectonic events that allowed dispersal across Huxley's, Wallace's and Lydekker's lines, and was congruent with historical biogeography of some HDFs. Finally, I used a combined analysis of genetic divergence and a standard scoring system for phenotypic data of Asian geo-isolates to provide evidence for quantitative delineation of species and propose changes in conservation status. My findings reveal the evolutionary history of horn bills from their emergence in the Late Oligocene from African origins, with a switch to frugivory influencing successful colonisation of hornbills and HDFs in Asian forests, which combined to promote sociality. Overall, this thesis demonstrates that access to novel environments and innovations to ecological niche facilitate rapid diversification in an avian lineage, and that this process is further promoted by the interplay of these birds in complex mutualistic interactions with their food, as well as palaeo-climatic and palaeo-tectonic changes.

Cooperatively breeding societies, in which one or more non-parental individuals (“alloparents”) care for young alongside the parents, show considerable variation in social structure. Traditionally, such societies have been thought to comprise small, kin-based family groups where offspring from previous broods delay dispersal and help raise closely-related offspring to gain indirect fitness benefits when independent breeding opportunities are unavailable or yield lower fitness outcomes. However, genetic evidence is increasingly revealing cooperatively breeding species whose social groups comprise unrelated individuals as co-breeders or alloparents or both (for e.g., 45% of all avian cooperative breeders). Such social groups exhibit complexity in social structure such as large group size, multiple breeders, and low and varied group kin structure. To understand why such complex societies form and how are they maintained when the opportunity to gain indirect benefits via kin selection is low and variable, I investigated the direct and indirect benefits driving a key demographic process, dispersal, and the resulting variation in group social structure on the individual, group, and population levels in an obligate, avian cooperative breeder, the superb starling (Lamprotornis superbus). I used a combination of long-term, individual-level data spanning 15 years from nine groups monitored at the Mpala Research Centre in Kenya and fine-scale genetic and environmental data sampled across 22 social groups that included the long-term study population. In Chapter 1, I show that (i) dispersal decisions in superb starling males are driven by temporal environmental variation experienced by their parents pre-laying, (ii) both dispersal and philopatry result in equivalent lifetime inclusive fitness outcomes, and (iii) oscillating selection due to high temporal variability in the environment likely maintains the two alternative dispersal tactics, resulting in the formation of mixed-kin groups. In Chapter 2, I show that (i) immigrants are vital to the stability of superb starling social groups in light of low and variable offspring recruitment in a harsh, unpredictable environment, (ii) plural breeding likely arises as a result of reproductive concessions provided by group members as joining incentives to recruit immigrants, and (iii) despite smaller groups providing more reproductive concessions, immigrants gain higher fitness in larger social groups and thus prefer to immigrate into larger groups which are found in higher-quality territories. Finally, in Chapter 3, I find genetic signatures of directional dispersal from social groups in low- to high-quality territories across an environmental gradient which likely generates considerable within-population variation in group social structure. Overall, my dissertation underscores the importance of direct benefits derived from group augmentation in the formation and maintenance of cooperative social groups with a complex social structure in a harsh and unpredictable environment.

The early environment impacts many aspects of an individual’s developing phenotype. In particular, early environmental conditions are important for shaping the hypothalamic-pituitary-adrenal (HPA) axis, which coordinates an individual’s stress response. These developmental changes are likely mediated by epigenetic modifications, functional changes to the genome that can alter gene expression in response to environmental variation, resulting in significant phenotypic differences (Kundakovic and Champagne 2015; Richards 2006). Determining how early life variation alters epigenetic modifications (such as DNA methylation) of genes throughout the HPA axis, and how these marks change over time, in wild organisms is important for understanding their potential long-term fitness consequences. Here, I examine DNA methylation modifications in the HPA axis in relation to early environmental variation in free-living house sparrows (Passer domesticus). In Chapter 1, I show a relationship between natural variation in the early environment and DNA methylation marks of numerous genes related to HPA axis function, which in turn predict growth trajectories. In Chapter 2, I show that early life stress in particular impacts DNA methylation in genes critical to HPA axis function, but does so differently depending on the life history stage in which stress is encountered. Finally, in Chapter 3, I find that these early life marks have long-term effects past the developmental period, predicting longevity as well as lifetime reproductive output in a sex-specific manner. Overall, my dissertation adds to a growing understanding of the dynamic role of epigenetic modifications in mediating phenotypic responses to the early life environment in wild birds, and demonstrates the potential long-term fitness outcomes of these changes.