In just a few hours, humans shed millions of bacteria into the world around them — in some cases, enough to identify them from the air surrounding them.

The human body is teeming with bacteria, dubbed the microbiome.

They line every surface, inside and out — the skin, the mouth, the gut.

Most of them are harmless or outright friendly, protecting the body against invading infections, training the immune system or breaking down toxins.

We carry about 10 bacterial cells for every 1 human cell. This makes it hard not to leave traces of bacteria behind us wherever we go.

Everything we touch receives not just our fingerprints and skin oils but also the unique mixture of different bacteria strains that inhabit our skin.

Now, research from the Biology and Built Environment Center (BioBE) at the University of Oregon, has found that it’s not just physical contact that spreads our personal microbiome — we shed it into the very air around us.

And each microbiome carries a unique signature that, if collected properly, can be used to identify the person it came from.

The findings were published today in the journal PeerJ.

Read More: Your Cell Phone Is Covered with a Personal Bacterial Cocktail »

Previous research had found that humans shed about a million sub-micrometer particles every hour, collectively called a bioaerosol. However, the bacterial composition of these particles was, until now, unknown.

The researchers used an experimental chamber at BioBE with sophisticated controls that allowed them to adjust the room’s air flow, temperature, and humidity levels. They sterilized the chamber, lined it with clean-room anti-static plastic sheeting, and set up sterile air filters at the chamber’s intakes and exhaust outputs.

The study included 11 participants, all free of infectious disease, and none of whom had taken antibiotics in the past four months.

Each person wore a tank top and shorts provided by the researchers. They took turns sitting in the room, one at a time, for anywhere from 90 to 240 minutes.

The room was bare save for a chair to sit in, a laptop for entertainment, and an array of petri collection dishes on the floor to gather any bacteria that settled out of the air.

After each test, the researchers gathered the bacterial samples from the air filters and petri dishes, then resterilized everything. From their samples, they extracted a specific gene called 16S ribosomal RNA, which is found in all bacteria and can indicate species and strain.

Between the 11 human subjects, they were able to gather more than 14 million genetic sequences to help identify the thousands of types of bacteria among them.

The species of bacteria that the researchers found weren’t particularly surprising. They included Staphylococcus, Propionibacterium, and Corynebacterium — all commonly found in (or on) humans.

However, these bacteria often appeared in different ratios, or were of a specific strain. Looking at the data, the researchers realized that they could tell apart some of the study’s participants just by looking at their microbiome breakdown.

Of the 11 participants, five could be identified from the collected exhaust air by their unique microbial fingerprint. One person, for example, carried a particular type of Staphylococcus epidermidis at higher levels than the other participants. Another person had a strong Lactobacillus crispatus signature.

For another four participants, air within the testing chamber was enough to tell them apart, but the exhaust air didn’t contain enough bacteria to confirm. And the final two subjects in the study couldn’t be detected by any airborne source.

“As collection and sequencing methods improve, so will these results,” said James Meadow, former postdoctoral research associate for BioME and now lead data scientist at Phylagen, and lead author on the paper, in an interview with Healthline. “DNA sequencing is in the midst of a revolution. Things are changing very quickly. What the human genome project did over a decade can now be done in weeks for a tiny fraction of the cost.”

Meadow acknowledges that his study has limitations, but is hopeful that it will pave the path for future applications.

“A single person sitting in an experimental chamber is not all that realistic,” he said. “So we would love to expand this study to see, for instance, if we can pick a person out of a crowd. I can think of lots of reasons we would want to know if some nefarious character has been in a certain room in the last few hours, and maybe there is a way to use microbes for that.”

Learn More: 80 Million Bacteria Pass Between Partners During a 10-Second Kiss »

On a less Big-Brother level, Meadow also hopes that his research could help explain how dangerous infections like MRSA can spread so quickly.

“We would love to know whether this concept can be used to help study outbreaks in hospitals or other buildings,” he said. “It might be possible to better understand how the people around us are influencing our own microbiome, even without touching each other.”

However, there’s one catch. DNA is a stable molecule and can last long after its host organism has died. This means that Meadow wasn’t measuring the total counts of living bacteria, but instead, the combined counts of bacteria living and dead.

“The presence of a microbe’s DNA does not mean it is alive or active, simply that its DNA was present,” said Lita Proctor, coordinator of the NIH Human Microbiome Project, in an interview with Healthline. “We must be very careful to conduct follow-up studies to verify if these microbial cloud microbes are in fact alive or active.”

But using DNA to identify bacteria, rather than growing cultures of samples, was still the right way to conduct this experiment, Meadow argues.

“The advantage is in the breadth and depth of the data set,” he explained. “Growing bacteria is seriously flawed because most bacteria are not readily grown, so you miss most of the bacteria that are in any sample. DNA sequencing can show you the vast majority of the community.”

Proctor also lays out another avenue of future research.

“The concept of a microbial cloud also means we acquire microbes from the environment,” she pointed out. “However, this study did not measure acquisition, admittedly a more difficult study. Nonetheless, an important companion study would be to evaluate the extent of microbial acquisition in individuals from other microbial clouds.”

If you’re concerned about a new “ick factor” in your life, you can relax, advises Meadow.

“Most of us don’t have to be worried about picking up anything nasty from the microbial cloud,” he assured us. “It’s just a normal interaction, and now, we know more about it.”

Related Reading: Humans Have Lost Many of Our Gut Bacteria Since We Evolved from Apes »