On 25 October 2017, researchers paddled beyond the surf near Carpinteria, California, and filled sterile bottles with surface water. No shark entered the bottle. Nothing visible in it proved that a shark had been nearby.
Yet the sample contained a molecular trace specific to the white shark.
Every animal leaks an archive
Organisms leave genetic material through shed cells, scales, mucus, faeces, gametes and decaying tissue. Scientists call the mixture environmental DNA, or eDNA. In water, those fragments can be trapped on a filter, extracted in a laboratory and compared with reference sequences.
A species-specific test asks a narrow question: is a chosen genetic signature present? Metabarcoding asks a wider one. It amplifies a standard genetic region from many organisms at once and compares millions of resulting sequences with a library of known species. One bottle can therefore return a list rather than a single answer.
Dr Lena Ortiz · shoreline assay log · 25 October 2017A clearly labelled fictional research layer of Code Gaia.The water looked empty. That was the problem with looking.
Three tagged juveniles had crossed the receiver line. Our bottles contained no fin, no tooth, no shadow—only suspended matter too small to name by sight.
The first droplets separated in the analyser. Then the target sequence appeared.
The shark had left the frame. It had not left the sample.
The deep ocean can be sampled without catching it
The method becomes more valuable where cameras see only a narrow cone and nets damage fragile organisms. During a 2022 NOAA expedition off Puerto Rico, teams collected one-to-five-litre samples from depths near 1,000 metres. Filters captured DNA released by corals, tunicates, fish and gelatinous animals that might evade the vehicle or disintegrate in a net.
In April 2026, NOAA released its first environmental-DNA dataset from Okeanos Explorer expeditions conducted between 2021 and 2023. eDNA had moved from a specialist experiment into the routine data haul of an ocean-exploration ship.
A detection is not a location pin
DNA can drift from its source, remain after an animal has moved away, or degrade before sampling. Temperature, sunlight, microbes, currents and depth change how long the trace survives. Laboratory contamination may create a false positive; scarce DNA and imperfect primers may create a false negative.
Even a correct sequence match says less than a sighting. It supports recent presence within an uncertain space and time. It usually does not identify how many animals were present, whether they were alive when the water was taken, or exactly where they were.
The database decides what can be named
A sequence without a reliable reference may remain assigned only to a family, a broader group or nothing at all. “Unknown DNA” is not automatic evidence of an unknown species. It may mean the fragment is poor, the marker cannot separate close relatives, or the animal has never been entered correctly into the reference library.
Dr Kaito Mercer · taxonomic review note · sample PR-12A clearly labelled fictional research layer of Code Gaia.Line 884 would not resolve below family. The temptation was immediate: unnamed animal, unrecorded depth, a discovery waiting for a title.
I ran the controls again. Then the reference library.
The sequence was real. The name was missing.
Those are different absences, and the archive must never confuse them.
What is established—and what remains open
| Status | Finding | Boundary |
|---|---|---|
| Confirmed | A species-specific assay detected white-shark DNA in 250 ml seawater samples | The study was a small, confirmatory field test at sites with known shark presence. |
| Confirmed | Water samples can reveal many marine taxa without capture | Results depend on primers, sequencing depth and the reference library. |
| Supported | eDNA expands observation of rare, mobile and fragile organisms | It complements cameras, nets and acoustic tags; it does not replace them. |
| Uncertain | A positive result locates an organism precisely | Transport and persistence blur distance and time. |
| Open | How unidentified sequences map onto undescribed biodiversity | Missing references and technical artefacts must be excluded first. |
The honest conclusion
A glass of seawater does not contain a perfect census. It contains fragments of recent biological history. The power of eDNA lies in revealing organisms that ordinary observation misses; its danger lies in making the molecular trace sound more precise than it is.
The ocean may hold far more readable information than cameras and nets collect. But reading it requires three separate questions: was the sequence real, was the identification valid, and what does presence in the sample actually mean?
What if the ocean itself becomes visible?
CG-036 · OCEAN & DEEP SYSTEMSMilky Seas Can Be Seen from SpaceDNA identifies organisms from fragments left in water. In a milky sea, an entire biological event writes itself in light across an area comparable to a country.
Open the next dossier →What becomes observable when countless microscopic sources begin producing one continuous signal?