Lineage diversity within a widespread endemic Australian skink to better inform conservation in response to regional-scale disturbance

This dataset was used to examine the phylogeographic genetic structure of Eastern three lined skink Bassiana duperreyi. It comprises  SNP data used for population genetics and phylogenetic reconstruction. The data were used to provide foundational work for the detailed taxonomic re-evaluation of this species complex and to reinforce the need for biodiversity assessment to include an examination of cryptic species and/or cryptic diversity below the level of species. Such information on lineage diversity within species and its distribution in the context of disturbance at a regional scale can be factored into conservation planning regardless of whether a decision is made to formally diagnose new species taxonomically and nomenclaturally.

Methods
Briefly, samples of tissue were collected from across the range of the species, Bassiana duperreyi, including from Australian Museums, DNA was extracted, double digested and genotyped for SNP markers using the technology of Diversity Arrays Technology (DArT, Canberra). The data were analysed in the software package dartR available on the CRAN repository, as per the script provided. Structure across the landscape was used to inform assessment of the impact of regional scale disturbance.

Skin tissues and extracted DNA were provided to DArT for processing, sequencing and informative SNP marker identification using DArTseqTM (Kilian et al., 2012). DArT performed a genome complexity reduction technique using double digestion of genomic DNA with two restriction endonucleases PstI (5′- CTGCA|G- 3′) and SphI (5′- GCATG|C- 3′), fragment-size selection and next-generation sequencing on an Illumina HiSeq2500 (CA, USA). Sequences were processed using proprietary DArT analytical pipelines (for full details refer to Georges et al. (2018). Initial filtering was based primarily on average and variance of sequencing depth, average allele counts and call rate across samples. Approximately one-third of samples were sequenced twice as technical replicates, with scoring consistency identifying high quality SNP markers with low error rates. We applied further quality control filtering using the R package dartR 2.7.2 (Gruber et al., 2018; Mijangos et al., 2022).  These filters were for reproducibility across technical replicates (< 99%), call rate removing both loci and individuals with > 5% missing data, read depth (< 8x and above > 50x) to remove low coverage SNPs and potential paralogs and by removing all but one of multiple SNPs per locus.