EnviroCoder logoEnviroCoder

eDNA Science

How accurate is eDNA testing, really?

February 10, 2025 · 6 min read

Every user of eDNA eventually asks: how much can I trust this list? The honest answer is that eDNA is more reliable than it seems at first blush and less than the excited early press releases claimed. Here is how professionals interpret an eDNA report.

What high confidence looks like

A hit is high-confidence when the barcode match is near-perfect against a well-populated reference, the species is plausible for the region, and the read count is comfortably above the noise floor. Common regional species almost always meet this bar.

Detection sensitivity for well-studied groups such as North American fish, amphibians, and mammals is generally in the 85–98% range under good sampling conditions — better than most traditional single-visit surveys.

Where false positives creep in

The main sources are lab contamination, reference database errors, and primer artifacts. A hit for a species that does not occur in the region is a red flag — reputable labs flag those for review rather than passing them through.

Cross-sample contamination during collection is another culprit. Reusing gloves between sites, dipping the same vial into two ponds, or storing samples together without seals all invite crossover.

Where false negatives creep in

The three big causes: sample too small to catch rare DNA, degradation between shedding and collection, and PCR inhibition. A species can be genuinely present and still not show up in one sample.

This is why serious surveys take multiple replicates and, when detection matters, use species-specific qPCR as a follow-up on suspected negatives.

How to read a species list honestly

Treat common regional species with high match scores as confident. Treat rare, out-of-range, or low-match hits as leads that need corroboration. Treat 'absent' as 'not detected in this sample,' not 'not present on the land.'

eDNA replaces guesswork with data, but the data still needs an ecologist's eye. A good report frames each result in that context.

Frequently asked questions

Will eDNA find every species on my property?

No. It finds the species whose DNA reached the sample, in enough copies to survive PCR, in a form that matched the primers. That is a large fraction of what is present, not all of it.

How do I know a surprising hit is real?

Look at match quality, whether the species occurs in the region, and whether independent samples repeat the hit. Reputable labs will annotate these.