Shedding light on rakali: genetic and morphological differentiation in the Australo-Papuan golden-bellied water rat (Hydromys chrysogaster), with notes on the Barrow Island population

The Australo-Papuan golden-bellied water rat or rakali (Hydromys chrysogaster Geoffroy 1804), is a large, carnivorous, semi-aquatic rodent endemic to Australia, New Guinea and islands in eastern Indonesia. It has a unique history in Australia, being the first native rodent described and up until mid last century was hunted intensively as a commercial fur-bearer. Sixteen subspecies described in past nomenclature were consolidated during the 1980s into the singular taxon it is known by today. Despite this history, known polymorphisms, and populations persisting in urban areas, H. chrysogaster remains a cryptic and relatively poorly researched component of the regions’ fauna assemblages.

To date, there has been a lack of investigation into spatial patterns of morphological variation, phylogenetic relationships, genetic structure and genetic differentiation across the species’ range. While considered secure nationally with an IUCN ranking of “Least Concern”, the need for genetic research and intraspecies taxonomic clarification has been highlighted as central to the species’ conservation. Additionally, there has been limited study of disjunct but potentially conservation significant Western Australian (WA) populations, despite the species’ priority status, population declines and numerous ongoing threats such as climate change.

The core aim of this study was to examine differentiation across the range of H. chrysogaster, in order to clarify its taxonomy and conservation status. Specific aims were to examine phylogenetic relationships in a temporal context, test whether all specimens resolved in a single Australian and New Guinean clade, examine genetic population structure and differentiation across the range and within Western Australia's southwest bioregion and document spatial patterns of morphological variation. In the process, information on population ecology and the natural history of WA’s remote Barrow Island population was collated.

Phylogenetic relationships were determined by partial sequencing of 1,115 base-pairs of the cytochrome b gene of mitochondrial DNA in 82 specimens, supplemented by eight Hydromys and Parahydromys sequences. Sequences were analysed using Bayesian and Maximum Likelihood methods, with a basic molecular clock reconstruction undertaken by Bayesian inference. To investigate gene flow and structure, primers were used to amplify six microsatellite markers in 93 individuals across the species’ range, with 37 samples assessed for sub-regional structure in southwest WA. External morphometric, pelage colouration and tail tip albinism data was collected from between 172 and 340 specimens. Ecological information relating to the Barrow Island population was obtained during targeted surveys conducted in July 2011 and August 2012 utilising a combination of live trapping, remote cameras and passive survey methods.

The majority of hypotheses were confirmed. Phylogenetic reconstructions revealed a complex biogeographic history, well-supported genetic structuring and divergence, an early divergent Tasmanian lineage and reciprocal monophyly between distribution extremes. Australian and New Guinean clades were evident, as well as an additional sister New Guinean Hydromys clade, containing two potential new taxa that appeared paraphyletic with the monotypic waterside rat (Parahydromys asper). Analysis of nuclear DNA revealed six major genetic clusters aligning with geographic regions. Southwest WA and Barrow Island populations showed a high level of differentiation, while remaining groups demonstrated contemporaneity of gene flow. Four distinct genetic clusters were evident in the former with a significant Isolation by Distance (IBD) effect. Morphology results showed male-bias body size sexual dimorphism, highly significant differences in tail tip albinism and correlations between body weight and latitude. A range of colour morphs were evident, including a melanistic phenotype in southwest WA and a pale silver morph restricted to Barrow Island, where albino individuals were also recorded. Observed morphology patterns were largely concordant with outcomes of genetic analysis. Targeted surveys of Barrow Island indicated that H. chrysogaster occurs in low abundance, prefererntially uses rocky headlands and is an opportunistic predator feeding upon a range of aquatic and terrestrial prey.

This thesis represents the first research to document divergence across the range of H. chrysogaster and addresses voids in understanding Western Australian populations. Results are discussed in context of ecological traits and geographic and temporal isolation, driven by aridification of continental Australia and changing sea levels during Pliocene and Pleistocene epochs. Findings support taxonomic revision within Hydromys to at least subspecies level, delineation of a new Hydromys/Parahydromys species, and due to high genetic structuring, treatment of certain regional populations as evolutionarily significant units (ESUs). These outcomes will help guide effective conservation and management of the species and all its forms into the future.