2015: Evaluating the accuracy of DNA tests for specimen identification: a bandicoot case study

Poster from Genetics Society of AustralAsia conference 2015

DNA detection of species from environmental samples is an effective means of studying rare and cryptic wildlife, and diagnostic DNA tests are increasingly applied to management questions. Taxon-specific DNA tests rely upon PCR primers that selectively amplify DNA from target species. Where DNA from several species might be amplified, or where false positives have important management implications, DNA sequencing is also required. To ensure the accuracy of DNA tests for specimen identification it is important to develop appropriate reference sequence databases for each study system, and to assess potential sources of error.

We have developed a test to detect bandicoot DNA from trace samples. Bandicoots have declined dramatically since European settlement and are at risk from introduced predators. We designed primers that amplify 134bp of the ND2 gene, targeting sequences specific to the six extant species of Isoodon and Perameles. To evaluate the limitations of these primers and the likelihood of erroneous species assignments, we conducted a series of tests in silico, in vitro and in vivo. Specifically, we sequenced the ND2 gene for all Australian bandicoots and, using a distance-based analysis, determined that this amplicon sequence provided sufficient resolution to distinguish among all species, except for some ambiguity between I. auratus and some mainland Australian I. obesulus. We also determined that these bandicoot-specific primers did not amplify DNA from 42 other mammal species, demonstrating substantial specificity. Finally, we used the bandicoot-specific primers to test DNA from 22 predator scats collected in Tasmania. DNA from the eastern barred bandicoot (P. gunnii) was detected from two scats, which were previously shown to originate from Tasmanian devil and cat.