Article · Field note 45
A water sample can carry a biological trace
Animals release cells and biological material into the water. Environmental DNA, or eDNA, is genetic material collected directly from environmental samples rather than from a captured organism. A small bottle of deep-ocean water can contain traces from animals that were too small, hidden, or quick for a camera. Researchers filter the water, extract DNA, amplify selected genetic markers, and compare the sequences with reference databases.
NOAA Ocean Exploration describes eDNA as a way to complement ROV video, which can miss cryptic, tiny, or visually similar species (NOAA field overview). A camera shows an organism in place; a water sample detects genetic material in the sample. The two methods see different parts of biodiversity and should be interpreted together.
From water to sequence
An expedition can collect water with bottles on a CTD rosette or with samplers mounted on an ROV. The sample is filtered so that cells and fragments are retained. In the lab, researchers extract DNA, select a marker region, and use sequencing or targeted assays to identify what is present. Metabarcoding can detect many taxa in one sample, while quantitative PCR targets a specific species or group.
Each step affects the result. The filter pore size influences which material is retained. DNA extraction can be more efficient for some organisms than others. Primer sequences may amplify one taxon more readily than another. A database match may resolve a sequence only to family or genus if close reference records are missing. The technology is powerful because it can survey many taxa, but the output still depends on sampling and analysis choices.
A detection is not a head count
Finding a DNA sequence does not automatically reveal how many animals were present. DNA production varies by species, size, activity, and life stage. Water volume, temperature, currents, and degradation affect how much DNA remains available. A larger signal may reflect more animals, more shedding, a recent event, or transport from another location.
Marine currents can move genetic material away from its source. NOAA Fisheries describes how tides, depth, temperature, and distance influence eDNA distribution in a chum salmon study (NOAA Fisheries report). In the open ocean, physical transport makes it difficult to assign every detection to an animal directly beneath the sampling point. An eDNA result is evidence of genetic material in the sample, while the spatial relationship to a living animal must be inferred carefully.
DNA can also degrade, producing false negatives if the target was present but not detected. Laboratory contamination can create false positives. Field blanks, filtration controls, replicate samples, and negative PCR controls help identify contamination and procedural problems. Positive controls show that the assay can detect its target under the test conditions. Good reporting includes these controls rather than only publishing the species list.
Reference databases matter
Once researchers sequence a marker, they compare it with reference records. If the target species has a reliable sequence in the database, identification can be strong. If related species share similar markers or the reference library is incomplete, the best supported name may be broader. A sequence can be exact and still lack a trustworthy taxonomic label.
The database problem is especially relevant in the deep sea, where many animals have not been sampled for genetic reference libraries. A detection that cannot be matched may indicate an undescribed lineage, a known species with no reference, or a technical artifact. More sequencing is not automatically more certainty. Researchers may need to collect a specimen and link its DNA to a verified physical voucher.
Peer-reviewed reviews of marine eDNA emphasize the influence of DNA production, persistence, transport, marker choice, and reference data on interpretation (marine eDNA methods review). A NOAA-supported paper on communicating eDNA science cautions against overstating exploratory results and recommends making uncertainty explicit (communications paper).
The best survey has more than one window
Combining methods can reveal why they disagree. A species may appear in eDNA but not video because it is small, hidden, or outside the route. A camera may record an animal whose DNA is not detected because the sample missed its plume or the marker did not amplify. The disagreement is informative when investigators consider the method and sampling scale.
For biodiversity surveys, researchers may collect water at several depths and locations, repeat samples across time, and compare results with camera transects. Replication helps estimate how reliably a species is detected. If a DNA signature appears in only one sample, scientists can examine controls and resample before making a strong distribution claim.
Deep-sea sampling changes the workflow
Collecting a sample several kilometers below the surface presents practical challenges. The water has to be brought up or processed in place, protected from contamination, and stored at a temperature and chemical condition suitable for DNA. On a ship, scientists use clean workspaces and separate equipment to reduce contamination between samples. A pressure-retaining sampler may preserve conditions better, but it adds design complexity.
Natural samplers add another collection method
A 2024 study of North Atlantic sponges analyzed environmental DNA retained in 97 tissue samples. Unlike a bottle of seawater, a sponge integrates material through its own filtration. That biological collection process introduces species-dependent selectivity. The result extends the sampling toolkit, but a detected sequence still cannot establish that the organism lived on the sampled sponge. Direct observation remains necessary for that spatial relationship.
