Time‐calibrated phylogeny and ecological niche models indicate Pliocene aridification drove intraspecific diversification of brushtail possums in Australia

Abstract

Major aridification events in Australia during the Pliocene may have had significant impact on the distribution and structure of widespread species. To explore the potential impact of Pliocene and Pleistocene climate oscillations, we estimated the timing of population fragmentation and past connectivity of the currently isolated but morphologically similar subspecies of the widespread brushtail possum ( Trichosurus vulpecula ). We use ecological niche modeling (ENM) with the current fragmented distribution of brushtail possums to estimate the environmental envelope of this marsupial. We projected the ENM on models of past climatic conditions in Australia to infer the potential distribution of brushtail possums over 6 million years. D‐loop haplotypes were used to describe population structure. From shotgun sequencing, we assembled whole mitochondrial DNA genomes and estimated the timing of intraspecific divergence. Our projections of ENMs suggest current possum populations were unlikely to have been in contact during the Pleistocene. Although lowered sea level during glacial periods enabled connection with habitat in Tasmania, climate fluctuation during this time would not have facilitated gene flow over much of Australia. The most recent common ancestor of sampled intraspecific diversity dates to the early Pliocene when continental aridification caused significant changes to Australian ecology and Trichosurus vulpecula distribution was likely fragmented. Phylogenetic analysis revealed that the subspecies T. v. hypoleucus (koomal; southwest), T. v. arnhemensis (langkurr; north), and T. v. vulpecula (bilda; southeast) correspond to distinct mitochondrial lineages. Despite little phenotypic differentiation, Trichosurus vulpecula populations probably experienced little gene flow with one another since the Pliocene, supporting the recognition of several subspecies and explaining their adaptations to the regional plant assemblages on which they feed.

The admixed brushtail possum genome reveals invasion history in New Zealand and novel imprinted genes

“…Of these, Trichosurus vulpecula fuliginosus from lutruwita Tasmania, and Trichosurus vulpecula vulpecula from south-east Australia, have well-documented importation into New Zealand 22 . The extent to which these subspecies are hybridising has not yet been fully described—elsewhere in New Zealand, certain populations maintain significant substructure, whereas others are freely interbreeding 44 – 48 . In all cases, genetic tracing to reveal the source Australian populations remains incomplete.…”

“…3a ). The other two haplotypes aligned to D-loop sequence from possums sourced from separate mainland populations (99.1% and 97.0% identity, respectively) 48 . We assessed nuclear genetic diversity from our RNA-sequencing data and reference samples 49 , 50 , finding standardised multilocus heterozygosity (sMLH) is higher on average in Dunedin possums than in the native Australian range, even when restricting comparisons to possums sharing the same mitochondrial haplotype.…”

“…Clades were obtained from the maximum likelihood phylogenetic reconstruction of mitogenomes, utilising reference samples from New South Wales (NSW, light green), Kangaroo Island (KAN, beige), and Tasmania (TAS, dark green). Haplotypes from NSW and TAS, but not KAN, were present in Dunedin samples, as well as two likely mainland Australian haplotypes, identified by homology to D-loop sequence 48 (red and blue). b Nuclear admixture analysis for Dunedin samples relative to reference samples from TAS (dark green) and NSW/KAN (light green), as well as another likely mainland source (blue).…”

“…The reference samples were used to define the origin of mitochondrial haplotypes within our Dunedin samples as either NSW, TAS, KAN, vs. ‘other well-supported clades’ not represented in our Australian full-length reference samples (note, although these latter clades did not match full-length mitochondrial reference sequence we had, they could be identified as being most similar to D-loop sequence originating from south-east mainland Australia 48 ). Only one discrepancy between runs/methods was detected: Peninsula2018_022 fell within the NSW clade in both RAxML runs, but differed between the BEAST runs (NSW and TAS), with the combined BEAST run finding this sample falling within the TAS clade.…”