Mammal declines across northern Australia are one of the major biodiversity loss events occurring globally. There has been no regional assessment of the implications of these species declines for genomic diversity. To address this, we conducted a species-wide assessment of genomic diversity in the northern quoll (Dasyurus hallucatus), an Endangered marsupial carnivore. We used next-generation sequencing methods to genotype 10,191 SNPs in 352 individuals from across a 3220 km length of the continent, investigating patterns of population genomic structure and diversity, and identifying loci showing signals of putative selection. We found strong heterogeneity in the distribution of genomic diversity across the continent, characterised by (1) biogeographic barriers driving hierarchical population structure through long-term isolation, and (2) severe reductions in diversity resulting from population declines, exacerbated by the spread of introduced toxic cane toads (Rhinella marina). These results warn of a large ongoing loss of genomic diversity and associated adaptive capacity as mammals decline across northern Australia. Encouragingly, populations of the northern quoll established on toad-free islands by translocations appear to have maintained most of the initial genomic diversity after 16 years. By mapping patterns of genomic diversity within and among populations, and investigating these patterns in the context of population declines, we can provide conservation managers with data critical to informed decision-making. This includes the identification of populations that are candidates for genetic management, the importance of remnant island and insurance/translocated populations for the conservation of genetic diversity, and the characterisation of putative evolutionarily significant units.

We use next-generation sequencing methods to genotype 10,191 SNPs in 352 northern quoll individuals from across a 3220 km length of the continent. We investigate patterns of population genomic structure and diversity, and identify loci showing signals of putative selection. Detailed methodological information is provided in the raw data, metadata files, scripts, and published manuscript that relate to this dataset.

Bioinformatic scripts are mostly run in command line format on Linux systems, using BWA v0.7.17, SAMTOOLS v1.7-1, ANGSD v0.93.

R scripts including most genomic analyses and figure production were run using R version 4.1 in RStudio 2022.07.0+548 "Spotted Wakerobin" Release for Windows.

Demographic and temporal trends were analysed using BEAST, SNeP v1.1, StairwayPlot v2.