Development of a massively parallel, genotyping-by-sequencing assay in American badger (Taxidea taxus) highlights the need for careful validation when working with low template DNA

Non-invasive DNA sampling to identify and enumerate species is critical to population monitoring and for developing effective management strategies. However, individual DNA identification is often limited by degraded and low template DNA (LT-DNA) that routinely yields partial profiles prone to technical artifacts, thus limiting their utility/reliability. Massively parallel, genotyping-by-sequencing (GBS) assays present an opportunity to amplify not only a large suite of molecular markers simultaneously, providing higher resolution to identify individuals, but also higher levels of sequence redundancy to enable quality metric evaluations of profiles from LT-DNA. Taxidea taxus jacksoni is an endangered badger subspecies in Canada, with low levels of genetic diversity, complicating individual identifications from closely related DNA sequences. Challenges arise from the small number of hairs collected from snag traps set in badger burrows that rarely provide full profiles. We designed a GBS assay to obtain microsatellite profiles compatible with pre-existing databases generated with conventional capillary electrophoresis (CE) genotyping. We assessed the assay’s reproducibility via a dilution series to mimic LT-DNA and tested if the assay produced similar CE-generated results. While GBS offers the potential to genotype large numbers of individuals and markers at the same time, we found low concordance between GBS- and CE-based genotypes from DNA templates < 250 pg. We recommend existing wildlife genetic databases focus on tetra-nucleotide microsatellite or SNP markers to reduce or eliminate sequencing artifacts (i.e., stutter) that present challenges for GBS genotypes from degraded and LT-DNA, and the use of sample replicates to form consensus genotypes.