PS 80-175 - Evaluating confidence in eDNA results: A case study detecting an invasive amphibian

Friday, August 11, 2017
Exhibit Hall, Oregon Convention Center
Emily A. Wilson, Ecology, Evolution and Marine Biology, UCSB, Santa Barbara, Tom L. Dudley, Marine Science Institute, University of California, Santa Barbara, Santa Barbara, CA, Cheryl J. Briggs, Ecology, Evolution and Marine Biology, UC Santa Barbara and Edward L. Ervin, Merkel & Associates Inc., San Diego, CA

Environmental DNA (eDNA) detection methods are becoming a powerful tool to determine the presence of species in complex aquatic environments, replacing more traditional labor-intensive surveys. As these methods become more common it is imperative to include multiple controls to ensure correct identification of positive and negative samples.

We tested a previously published species-specific primer and its ability to detect an invasive amphibian, the African clawed frog (Xenopus laevis), at 40 sites throughout southern California. DNA was collected by filtering water through 1.2μm glass fiber filters. The DNA was extracted from the filter and tested in triplicate with the species-specific primer and probe using qPCR.

A series of positive and negative controls were included to increase the confidence of the sample results. The specificity of the primer was tested in qPCR using tissue extracts from the target and non-target species. Tissue extracts were also run on gels and bands sequences for comparison to target X. laevissequences. Known positive and negative sites were included in the field surveys. Sites that showed positive in qPCR were confirmed further by sequencing the bands from a gel. Samples were also checked for inhibition with an internal positive control and purified if needed.


The eDNA field survey results appeared to confirm the presence or absence of X. laevis at the majority of sites. The eDNA results matched the known positive and negative sites and no negative controls were positive. Four of the eDNA sites were still partially inhibited after treatment, which makes it difficult to draw conclusions at those sites.

Some late cycle amplification was observed with high concentrations of some non-target species’ tissue extracts. This suggests that there is some non-target binding with the primers, which was further confirmed with gel images that displayed multiple bands from the tissue extracts of non-target amphibian species. However, sites with high densities of non-target amphibians did not amplify, suggesting that high eDNA levels of non-target DNA will not create a false positive.

These results show the importance of multiple levels of confirmation for positive and negatives when performed an eDNA survey. The primers should be tested with tissue extracts to confirm specificity. Positive and negative field sample sites should be included along with negative controls at each step of analysis.