3.2 Phylogeny and demographic history
To construct a time-calibrated species tree (Figure 2a ), we identified 7,116 high-confidence single-copy orthologs from nine marsupial species and a monotreme, the platypus (Ornithorhynchus anatinus ). Extant Australasian and American marsupials diverged from a South American ancestor approximately 80 million years ago (Kumar et al., 2017; Nilsson et al., 2010), emphasising the deep evolutionary history of marsupials. We estimated that the lineages leading to extant Dasyuridae (e.g., Antechinus ) and Didelphidae diverged 81.7 million years ago [77.7-85.8 Mya 95% credibility interval (CI)]. The dasyurid subfamily Phascogalinae comprises three genera: the semelparous Australian Antechinus and Phascogale , which have male die-off, and the iteroparous New Guinean sister genusMurexia . Small dasyurids from mainland Australia are likely to have dispersed into New Guinea 9-11 million years ago (K. J. Mitchell et al., 2014). We found the divergence time of Antechinus andMurexia to be about 7.1 Mya (95% CI 5.9-7.3). Our analysis show that the three antechinus species split from a common ancestor ~4.6 Mya (95% CI 3.8-4.8), while A. flavipes andA. argentus share an ancestor ~2.4 Mya (95% CI 2.0-2.6). Our phylogeny and divergence times concord with trees generated from the PHYLACINE database (Faurby et al., 2018) (Figure S6 ) and with recent estimates from a small set of mitochondrial and nuclear genes (Mutton et al., 2019).
The pairwise sequentially Markovian coalescent (PSMC) method (Li & Durbin, 2011), which can infer changes in the effective population size (Ne ) over the Pleistocene, was used to evaluate the four sequenced dasyurid species (Figure 2b,c ). Until the end of the Pleistocene, approximately 11,700 years ago, mainland Australia, New Guinea, and Tasmania formed a single continent, Sahul (Barrows, Stone, Fifield, & Cresswell, 2002). Glaciation of Sahul was restricted to the Snowy Mountains of southeastern mainland Australia and the Tasmanian highlands, and the continent experienced cycles of aridity and sea-level fluctuations (Barrows et al., 2002). Until about 250,000 years ago, the species in higher-elevation, wetter habitats; A. argentus , A. arktos , and M. melanurus , had larger effective population sizes than A. flavipes , which were found in lower-elevation, drier habitats than today (A. M. Baker & Van Dyck, 2013). After 250,000 years ago, the species in higher-elevations declined in population size, while A. flavipes populations expanded (Figure 2a,b ). This pattern persisted into the Late Pleistocene ~11,700 to 126,000 years ago, a period when temperature fluctuations were large, and many mesic-adapted species declined, became restricted to wet refugia such as mountain-tops or became extinct, and more arid-adapted species radiated (Saltre et al., 2016). While the PSMC method cannot infer population dynamics after ~20,000 years ago, A. flavipes had the largestNe in the most recent PSMC estimate, consistent with the status of the yellow-footed antechinus being one of the most common and broadly distributed species in genus Antechinus today (A. M. Baker & Van Dyck, 2013; Lada, Thomson, Mac Nally, & Taylor, 2008; Mutton et al., 2019). This conclusion is in agreement with a recent study showing that aridity and geographical isolation led to dasyurid speciation (García‐Navas, Kear, & Westerman, 2020). In conclusion, our analysis suggests that climate change during the Early (~2.5-0.78 Mya) (Lisiecki & Raymo, 2005) and Mid Pleistocene (0.78-0.126 Mya) (Elderfield et al., 2012) shaped the evolution of the dasyurid subfamily Phascogalinae by isolating populations in different habitats.
We found that all small dasyurid taxa have much higher (100 orders of magnitude) genome-wide heterozygosity (π from 0.048 to 0.096) than the endangered S. harrisii (π=0.00032 (Miller et al., 2011)) (Figure S7 ). Moreover, the heterozygosity of A. flavipesis twofold higher than A. arktos , A. argentus , andM. melanurus . This may reflect its larger historical and current population size (Figure 2b,c ). Although we are comparing individuals from single populations, combined with factors such as environmental change, loss of habitat, and population fragmentation, the lower genetic diversity of A. arktos and A. argentus may reflect their present-day low population and threatened species status (Mutton et al., 2019; Riordan, Pearce, McDonald, Gynther, & Baker, 2020). Future studies, including assessment of geographic population structures and the inclusion of historical and current specimens (Diez-Del-Molino, Sanchez-Barreiro, Barnes, Gilbert, & Dalen, 2018) are required to assess the threat level of extant small dasyurids. Our high-quality Antechinus flavipes assembly serves as an ideal reference genome for population genomic studies of A. flavipesand the 14 other antechinus species.