Phylogenomics, molecular evolution and extinction in the adaptive radiation of murine rodents

Abstract

Adaptive radiation plays a significant role in the generation of biological diversity, and the advent of modern sequencing approaches has unlocked a new genomic perspective on this process. Genomic-scale data from the across the diversity of adaptive radiations can provide unprecedented resolution of the phylogenetic, biogeographic and molecular context of diversification. Murine rodents (Murinae: Rodentia) are a recent and rapid adaptive radiation that make up > 10% of mammal species. Murines have repeatedly colonised new geographic areas and island systems in the Eastern Hemisphere, frequently as a result of overwater transitions. Recurring adaptive radiation, ecological character displacement, and convergent evolution across Murinae make them an ideal model for studying adaptive radiation, especially in the Indo-Australian region.

Within broader Murinae, the Hydromyini are a speciose Australo-Papuan radiation that diversified following an overwater colonisation from Sunda to Sahul ca. 8 Ma. Previous multilocus studies did not provide sufficient phylogenetic resolution of the rapid diversification of Hydromyini, and did not adequately sample taxa to reconstruct their complex biogeographic history. In addition to unresolved biogeography, the endemic Australian clade within Hydromyini has suffered the highest rate of recent mammalian extinction in the world. The rapid decline of Australian rodents is thought to be primarily the result of predation by feral cats, combined with other factors such as anthropogenic land clearing. There is little information about the pace of decline in eight species that went extinct on the Australian mainland in the last 150 years, and it is unclear whether these species had suffered longer term declines that predate the arrival of Europeans into Australia in 1788.

To resolve these outstanding issues, I develop a novel exon capture approach for murine rodents. Firstly, I investigate the degree of congruent and conflicting phylogenomic signal in a rapid radiation, using genus-level relationships in the Hydromyini as a model example. My results show that in a number of cases, strong conflict is not reflected in branch support metrics obtained using either maximum likelihood or summary coalescent approaches. This result is significant, as it suggests that approaches commonly used to estimate support in phylogenomic data can fail to detect uncertainty in the face of underlying genealogical heterogeneity.

Further leveraging this novel exon capture design, I generate a robust phylogenomic tree based on > 350 samples across the Australo-Papuan continent, including extant and recently extinct species in Hydromyini. With these data, I reconstruct the species-level evolutionary and biogeographic history of the Hydromyini across Sahul, recovering numerous examples of overwater colonisation between regions. Consistent with the geomorphological hypothesis that the New Guinea lowlands emerged after the orogeny of the Central Cordillera, I find evidence for increasing ecological opportunity in the Hydromyini from approximately 5 Ma. This first species-level phylogenomic study spanning the entire Sahul region provides a baseline example for future comparative studies that seek to reconstruct the biogeographic drivers of diversification in Sahul at a continental scale.

Using exon capture and whole-exome sequencing data from extinct and extant species, I place recently extinct Australian rodents in a phylogenomic context for the first time. I recover no marked evidence of genetic erosion in five extinct species at the time of specimen collection, in comparison to extant species with present-day low allelic diversity. This indicates that the decline of recently extinct Australian rodents occurred extremely rapidly, and its onset likely did not predate European settlement. Additionally, my results taxonomically resurrect a species from extinction, Gould’s mouse (Pseudomys gouldii), which survived as a single island population in Shark Bay, Western Australia (currently classified as P. fieldi).

Finally, I generate whole exome data from 38 species in the global radiation of Murinae to examine patterns of positive selection and convergent evolution. I uncovered pervasive positive selection across genes associated with diet, digestion and taste across Murinae, and increased rates of adaptive evolution in carnivores compared to omnivores. Limited evidence for molecular convergence in worm-eating specialists Paucidentomys and Rhynchomys suggests a role for developmental phenotypic control in this striking example of ecological convergence. Broadly, my results indicate that the pronounced ecological and phenotypic shifts that are hallmarks of adaptive radiations may also drive corresponding shifts in the pace and pattern of molecular evolution across the genome.

Together, the work in this thesis is fundamental to the understanding of diversification, adaptation and extinction in the Australo-Papuan region, and provides an extensive genomic resource for future studies.