A splitting headache: relationships and generic boundaries among Australian murids

Genera should ideally make sense to someone not familiar with the constituent species. Recent authors have supported recognition of the large Australian rodent genus Pseudomys simply on the basis of its monophyly, but have not posed the reasonable question: are there smaller groups within the genus that are typified by unique biological or morphological traits warranting generic recognition? Indeed, Pseudomys is only rendered monophyletic in molecular phylogenies with the inclusion of Mastacomys , an arrangement not supported by local authorities due to the clear differences between Mastacomys and other Pseudomys . Here I review the phylogeny of Australian rodents, and suggest a ‘common‐sense’ generic re‐description of Pseudomys based on previous work, new mitochondrial and nuclear DNA phylogenies, and a critical appraisal of the biological attributes of each species group within Pseudomys . I propose that if species groups such as the pebble‐mound mice, typified by behaviour not observed in any other mammals, are not recognized as genera, the taxon becomes meaningless. I suggest criteria by which future morphological and genetic studies can test the utility of characters in identifying systematic relationships among the conilurine rodents, and propose groups to be fully described as genera following morphological review in a forthcoming paper. © 2006 The Linnean Society of London, Biological Journal of the Linnean Society , 2006, 89 , 117–138.