A Science Story
The technician called it "dirty air." Not in a rude way — in a technical way. The filters from the radionuclide station north of the Arctic Circle were so clogged with biological material that the sequencing machine kept choking.
Per Stenberg stared at the error reports and laughed. Dirty air. Thirty years of nuclear surveillance had accidentally created the most comprehensive biodiversity archive on the planet, and nobody had noticed.
He was about to change that.
The year was 2015, and Stenberg was a molecular biologist at Umeå University in Sweden with a problem and a seminar invitation. The problem was conventional. His research into environmental DNA — the genetic traces animals leave in water and soil — had hit a wall. Water samples gave you a snapshot, but a blurry one. You caught what swam past your net, nothing more.
The seminar was different. It was about Sweden's radionuclide-detection network, built in the late 1950s to catch the sniff of nuclear weapons testing. Twenty-five stations scattered across the country, each one sucking hundreds of cubic metres of air through glass-fibre filters every hour, around the clock, for seventy years.
"And the filters?" Stenberg asked during the Q&A.
"They're stored in Stockholm," the speaker said. "Tens of thousands of them."
Stenberg's mind did something it hadn't done since grad school. It went completely quiet. In that silence, a thought crystallized with the clarity of ice forming on a still pond: Every organism that ever drifted past those stations is still there, frozen in glass fibre, waiting.
He called his colleague Mats Forsman at the Swedish Defence Research Agency the next morning.
The first results took four years. Four years of asking for access, of convincing committees, of arguing that yes, this was legitimate science and not some backdoor into classified materials. Four years of designing new statistical methods because the old ones assumed you were looking at a water sample, not a compressed history of the atmosphere.
When the data finally came back, Stenberg sat in his office and wept.
"Viruses, bacteria, fungi, plants, animals, birds, fish … the intestinal parasites of moose," he told an interviewer later, his voice still thick with wonder. "Whatever was out there and had a reference to match it — we could see. Every single organism that is not extremely rare in the ecosystem."
Not a snapshot. Not a blurry approximation. A film.
He was looking at seventy years of ecological history, encoded in the dust that Swedes had breathed in and out every day without knowing it.
The discovery landed quietly in the scientific world — a 2019 paper that most journalists overlooked in favour of whatever feud was trending on Twitter. But within the ecology community, it detonated.
Other researchers had been approaching the same idea from different angles. In 2013, biologists Matt Clark at the Natural History Museum in London and Richard Richard Leggett at the Earlham Institute in Norwich had taken air samples in a greenhouse, mostly on a lark. "We were just wondering whether we would get anything," Clark said later. "Actually, we got dozens — hundreds — of things turning up."
Meanwhile, at Texas Tech University, ecologist Matthew Barnes was adapting waterborne eDNA techniques for air, and finding them teeming with DNA from leaves, flowers, and pollen that had never been designed to travel by wind. At York University in Toronto, Elizabeth Clare and Joanne Littlefair at University College London were running what they called a "zoo proof of concept" at a small wildlife park in Cambridgeshire — a contained environment where they knew exactly which animals should be present. They collected air samples, extracted the DNA, and amplified it.
They found tigers two hundred metres from their enclosure.
Not just tigers. Their food — chicken, horse, pig. Wildlife they hadn't expected: hedgehogs, bats, squirrels. DNA from twenty-five species of mammal and bird, including seventeen that were actually kept at the zoo.
"Airborne animal DNA has always been there," said Simon Creer, a molecular ecologist at Bangor University. "It's just that we've never looked for it."
A parallel study near Copenhagen Zoo found the same thing. The scientific community had been breathing in biodiversity data for decades without realising it.
By 2025, the implications had grown strange and large.
James Allerton, a physicist at the National Physical Laboratory in London, suggested a collaboration with the UK Heavy Metals monitoring network — twenty-five pumps in cities, countryside, and industrial sites. Clare and Littlefair analysed samples from fifteen of these stations and published what they called the world's first national survey of terrestrial biodiversity using airborne eDNA.
They found common UK animals. They found exotic pets — parrots that had apparently escaped or been released. They found an invasive fish species, silver carp, that had never been reported in the region.
They found 1,100 taxa. Vertebrates to single-celled protists.
To check their work, they compared it against iNaturalist, the vast citizen-science database where millions of people record what they see in nature. iNaturalist had missed half of what the eDNA had found. In turn, eDNA missed forty-three percent of what iNaturalist observers had recorded.
The reason was illuminating. Humans tended to see beautiful, visible things near where they lived — birds in gardens, foxes in suburbia, deer in country lanes. Airborne DNA caught what humans didn't see: the small, the cryptic, the nocturnal. Fungi. Lichens. Invertebrates. Plants that weren't trees.
"These are really the powerhouses of ecosystem function," said Littlefair. She meant it as a scientific statement, but it sounded like a eulogy for the overlooked.
Stenberg's group was now modelling cause and effect in ecosystems — something ecologists had dreamed about for a century but never had the data to attempt.
"We know that foxes eat rabbits, and rabbits eat some plants and so on," Stenberg said. "But the full ecosystem — when we talk about the bacteria, the nematodes, the insects, the plants, the animals — we basically have no idea."
Airborne eDNA was beginning to provide that idea.
His team had documented weekly, seasonal, and cyclic fluctuations in species abundance, matching them to climate variations. They had tracked long-term community changes: the rise and fall of pine trees driven by shifting forestry practices, and the concomitant decline in other trees, mosses, lichens, and fungi that depended on them. They had verified ancient ecological relationships — flies and bacteria, predators and prey — and stumbled onto new ones that nobody had predicted.
Europe was dotted with radionuclide-detection stations. If other countries followed Sweden's lead in opening their archives, scientists would have access to an unprecedented network for reconstructing ecological history and watching change happen in something approaching real time.
But there was a shadow in the data.
Some of the genetic material the researchers were pulling from the air came from humans. Exhalations. Skin cells. Hair. And with it came information that went beyond species identification.
When Clare and Littlefair shared their zoo findings with the research ethics board, the board flagged a concern that nobody had raised before: airborne eDNA could inadvertently reveal a person's ethnicity. Whether someone carried genetic markers for disease. Even, theoretically, their identity.
"We should be concerned about this," said one researcher who asked not to be named. "The same technique that can tell you a hedgehog lives in your garden can tell you a lot about the humans who live there too."
There were no easy answers. The filters were already collected. The data already existed. The technology to read it was getting cheaper and more powerful every year.
Stenberg's response was measured. "Science has always faced ethical questions when it develops new capabilities," he said. "The important thing is to face them openly, not to pretend they don't exist."
Meanwhile, in Australia, Erin Hahn was thinking smaller.
Hahn, a conservation geneticist at the Australian National Wildlife Collection in Canberra, had designed and 3D-printed passive air samplers — devices that didn't need electricity, that anyone could deploy. She had given them to landholders across New South Wales, in a pilot programme that felt more like a farmers' market than a research lab.
"There's heaps of variables around airflow, light exposure, proximity to game trails," Hahn said. "We're just starting to chip away at them."
What she wanted, ultimately, was nimble. A distributed network of cheap sensors that could flag an invasive species before it established, that could notice a crashing population before the die-off made the news. No lab required. No expertise required. Just people with a patch of land and a device the size of a coffee can.
"If we can get these into the hands of landholders, we can do biodiversity monitoring at a scale that's never been possible before," she said. "Not just the sites where scientists have time to go. Everywhere."
On a quiet evening in Stockholm, Per Stenberg stood outside his institute and looked up at the sky.
It was a clear night, stars visible between the clouds. He thought about the fact that every breath he took contained fragments of DNA — from the trees along the street, from the birds roosting on the roof, from the bacteria growing in the rain puddles, from the seven billion humans on the planet exhaling their genetic selves into the wind.
We are absolutely surrounded by information in the form of DNA and RNA, at all times, Ryan Kelly had said. Stenberg had read that quote and felt it settle into his chest like a truth he had always suspected but never had the words for.
The air was not empty. It had never been empty. It was full of ghosts — the inherited echoes of every organism that had ever passed through, each one leaving a whisper of code behind.
And now, for the first time in history, we had learned to listen.
Story based on discoveries by Per Stenberg (Umeå University), Elizabeth Clare (York University), Joanne Littlefair (UCL), Matt Clark (Natural History Museum London), and researchers cited in Nature (2026).