Abstract
Using genetic information to develop and implement conservation programs is vital for maintaining biodiversity and ecosystem resilience. Evaluation of the genetic variability within and among remnant populations can inform management of both natural and translocated populations to maximise species' adaptive potential, mitigate negative impacts of inbreeding, and subsequently minimise risk of extinction. Here we use reduced representation sequencing to undertake a genetic assessment of the golden bandicoot (Isoodon auratus), a threatened marsupial endemic to Australia. The currently recognised taxon consists of three subspecies distributed among multiple natural and translocated populations. After confirming the genetic distinctiveness of I. auratus from two closely related taxa, I. fusciventer and I. macrourus, we identified four genetic clusters within I. auratus. These clusters exhibited substantial genetic differentiation (pairwise F(ST) values ranging from 0.18 to 0.65, pairwise D(XY) ranging from 0.1 to 0.168), reflecting long-term isolation of some populations on offshore islands and the influence of genetic drift. Mainland natural populations in the Kimberley region had the highest genetic diversity and the largest contribution to overall allelic and gene diversity compared to both natural and translocated island populations. A population translocated to Guluwuru Island in the Northern Territory had the lowest genetic diversity. Our data suggest that island populations can appear genetically unique due to genetic drift and this needs to be taken into account when considering genetic diversity in conservation efforts to maintain overall genetic diversity of the species. We effectively demonstrate how genomic information can guide practical conservation planning, especially when declining species are represented by multiple isolated populations.