Turbinal bones are key components of the mammalian rostrum that contribute to three critical functions: (1) homeothermy, (2) water conservation and (3) olfaction. With over 700 extant species, murine rodents (Murinae) are the most species-rich mammalian subfamily, with most of that diversity residing in the Indo-Australian Archipelago. Their evolutionary history includes several cases of putative, but untested ecomorphological convergence, especially with traits related to diet. Among the most spectacular rodent ecomorphs are the vermivores which independently evolved in several island systems. We used 3D CT-scans (N = 87) of murine turbinal bones to quantify olfactory capacities as well as heat or water conservation adaptations. We obtained similar results from an existing 2D complexity method and two new 3D methodologies that quantify bone complexity. Using comparative phylogenetic methods, we identified a significant convergent signal in the rostral morphology within the highly specialised vermivores. Vermivorous species have significantly larger and more complex olfactory turbinals than do carnivores and omnivores. Increased olfactory capacities may be a major adaptive feature facilitating rats’ capacity to prey on elusive earthworms. The narrow snout that characterises vermivores exhibits significantly reduced respiratory turbinals, which may reduce their heat and water conservation capacities.

The Pig-footed Bandicoot, Chaeropus ecaudatus, an extinct arid-adapted bandicoot, was named in 1838 based on a specimen without a tail from the Murray River in New South Wales. Two additional species were later named, C. castanotis and C. occidentalis, which have since been synonymised with C. ecaudatus. Taxonomic research on the genus is rather difficult because of the limited material available for study. Aside from the types of C. castanotis and C. occidentalis housed at the Natural History Museum in London, and the type of C. ecaudatus at the Australian Museum in Sydney, there are fewer than 30 other modern specimens in other collections scattered around the world. Examining skeletal and dental characters for several specimens, and using a combination of traditional morphology, morphometrics, palaeontology and molecular phylogenetics, we have identified two distinct species, C. ecaudatus and C. yirratji sp. nov., with C. ecaudatus having two distinct subspecies, C. e. ecaudatus and C. e. occidentalis. We use palaeontological data to reconstruct the pre-European distribution of the two species, and review the ecological information known about these extinct taxa.

Natural history specimens are widely used across ecology, evolutionary biology and conservation. Although biological sex may influence all of these areas, it is often overlooked in large-scale studies using museum specimens. If collections are biased towards one sex, studies may not be representative of the species. Here, we investigate sex ratios in over two million bird and mammal specimen records from five large international museums. We found a slight bias towards males in birds (40% females) and mammals (48% females), but this varied among orders. The proportion of female specimens has not significantly changed in 130 years, but has decreased in species with showy male traits like colourful plumage and horns. Body size had little effect. Male bias was strongest in name-bearing types; only 27% of bird and 39% of mammal types were female. These results imply that previous studies may be impacted by undetected male bias, and vigilance is required when using specimen data, collecting new specimens and designating types.

Australia has the highest historically recorded rate of mammalian extinction in the world, with 34 terrestrial species declared extinct since European colonization in 1788. Among Australian mammals, rodents have been the most severely affected by these recent extinctions; however, given a sparse historical record, the scale and timing of their decline remain unresolved. Using museum specimens up to 184 y old, we generate genomic-scale data from across the entire assemblage of Australian hydromyine rodents (i.e., eight extinct species and their 42 living relatives). We reconstruct a phylogenomic tree for these species spanning ∼5.2 million years, revealing a cumulative total of 10 million years (>10%) of unique evolutionary history lost to extinction within the past ∼150 y. We find no evidence for reduced genetic diversity in extinct species just prior to or during decline, indicating that their extinction was extremely rapid. This suggests that populations of extinct Australian rodents were large prior to European colonization, and that genetic diversity does not necessarily protect species from catastrophic extinction. In addition, comparative analyses suggest that body size and biome interact to predict extinction and decline, with larger species more likely to go extinct. Finally, we taxonomically resurrect a species from extinction, Gould’s mouse (Pseudomys gouldii Waterhouse, 1839), which survives as an island population in Shark Bay, Western Australia (currently classified as Pseudomys fieldi Waite, 1896). With unprecedented sampling across a radiation of extinct and living species, we unlock a previously inaccessible historical perspective on extinction in Australia. Our results highlight the capacity of collections-based research to inform conservation and management of persisting species.