Recent range expansion in Australian hummock grasses () inferred using genotyping-by-sequencing.

The Australian arid zone (AAZ) has undergone aridification and the formation of vast sandy deserts since the mid-Miocene. Studies on AAZ organisms, particularly animals, have shown patterns of mesic ancestry, persistence in rocky refugia and range expansions in arid lineages. There has been limited molecular investigation of plants in the AAZ, particularly of taxa that arrived in Australia after the onset of aridification. Here we investigate populations of the widespread AAZ grass to determine whether there is evidence for a recent range expansion, and if so, its source and direction. We also undertake a dating analysis for the species complex to which belongs, in order to place its diversification in relation to changes in AAZ climate and landscapes. We analyse a genomic single nucleotide polymorphism data set from 17 populations of in a recently developed approach for detecting the signal and likely origin of a range expansion. We also use alignments from existing and newly sequenced plastomes from across Poaceae for analysis in BEAST to construct fossil-calibrated phylogenies. Across a range of sampling parameters and outgroups, we detected a consistent signal of westward expansion for , originating in central or eastern Australia. Divergence time estimation indicates that began to diversify in the late Miocene (crown 7.0-8.8 million years (Ma)), and the complex began to radiate during the Pleistocene (crown 1.4-2.0 Ma). This evidence for range expansion in an arid-adapted plant is consistent with similar patterns in AAZ animals and likely reflects a general response to the opening of new habitat during aridification. Radiation of the complex through the Pleistocene has been associated with preferences for different substrates, providing an explanation why only one lineage is widespread across sandy deserts.