We Filtered Sixty Litres Of Creek. The Unknown Reads Were A Vole.
Environmental DNA picks up every animal that has been near the water. We sampled eleven sites across a watershed with a long sighting record. The lab found forty-one vertebrate species. The unmatched sequences turned out to be a vole.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
SKAMANIA COUNTY, Wash. — Every animal that goes near water leaves cells in it. Skin, hair, saliva. Filter enough water and you can list what has been there.
We filtered sixty litres at each of eleven sites in a watershed with a long record of sightings. The people who report those sightings helped choose where.
Forty-one vertebrate species came back.
What "unassigned" means
Dr. Marisol Kwan-Petrakis made us agree to print this before any result, and she was right to.
Sequencing gives you fragments. The software compares each fragment to a reference library and names it. Anything it cannot match is filed as unassigned.
Unassigned does not mean unknown to science. It overwhelmingly means not in the library — a species nobody has bothered to sequence, or a sequence too degraded to match, or a stretch of DNA that is identical across a dozen relatives.
"Every time a newspaper runs one of these, somebody reports the unassigned fraction as a mystery animal," she said. "It is a mystery filing cabinet."
Our unassigned fraction
1.4 per cent.
Most of it resolved on a second pass with a wider library. The largest single component was a vole — a perfectly well-known vole, living exactly where a vole should live, whose reference sequence was simply not in the database we started with.
A residue would not resolve. The honest description of that residue is degraded sequence, and Dr. Kwan-Petrakis would not let us call it anything else.
The bears
Black bear at nine of the eleven sites.
That is not a throwaway. Work comparing sighting reports against black bear populations finds the two track each other closely, which is exactly what you would expect if a large share of encounters are bears seen briefly, at distance, in poor light, standing up.
Our correspondent, who has spent eighteen years on this beat, would like to add that a standing black bear at forty metres in failing light does not look like a bear. He has done it. He wrote what is that in his notebook and then spent ninety seconds being wrong.
The four who helped
Two of the four people who chose sites with us walked them on the day.
Neither has changed their mind, and both said so carefully. One of them put it in a way our correspondent has been thinking about since.
"You have tested the water," he said. "I did not see a thing in the water. I saw a thing on the ridge, once, in 1994, for about four seconds, and I have never said it was anything except that."
Then: "I helped you pick the sites because I would rather know. If you had found something I would have wanted the same standard of proof. I do not get to only want it in one direction."
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
Pre-registered before a single sample: the eleven sites, chosen with the help of people who report sightings there; the filtration volume; the lab; and an undertaking to publish the entire species list whatever was in it. Sixty litres per site through a 0.45-micron filter. The full sequence data is in a public repository.
Who we spoke to
- Dr. Marisol Kwan-Petrakis, Molecular ecologist; ran the sequencing. Commissioned; fee $22,800, disclosed. Protocol published before sampling. April–July 2026 Insisted we print the explanation of unassigned reads before any result. We have.
- Washington state wildlife biologists, Two, named in our notes. Advised on site selection; took no fee March 2026
- Four people with sighting reports in the watershed, Named with consent. Interviewed; two walked the sites with our correspondent April 2026 All four helped choose sampling locations. Two have since said the result does not change what they saw.
- The full sequence data, Deposited in a public repository. Uploaded on publication August 2026
Documents
- PX-2029 — eDNA survey, eleven sites, one Washington watershed — protocol, species list and raw sequence accepted
What we could not confirm
- Whether a larger survey would find something else. Eleven sites, one watershed, one season. This is a small study and says so.
- What made the 1.4% unassigned. Most resolved on a second pass; a residue did not, and the honest description of that residue is 'degraded sequence', not 'unknown animal'.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Hollow PostNEWSPAPER FINDS UNKNOWN DNA IN BIGFOOT COUNTRY
Reported the 1.4% unassigned fraction as an unidentified species.
Had that we did not
Nothing.
Left out
What unassigned means, which the lab explained to us in one sentence and which is in our fourth paragraph.
-
They Buriedthis newspaperWe Filtered Sixty Litres Of Creek. The Unknown Reads Were A Vole.
Pre-registered, published the whole species list, and had the lab explain what an unassigned read is.
Had that we did not
The full list and the unassigned fraction, resolved.
Left out
Eleven sites in one watershed in one season is a very small survey and our headline sounds more conclusive than eleven sites can be. — V. Ashcombe-Doyle, standards editor
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.