Environmental DNA lets scientists census a kelp forest from a few litres of seawater. How eDNA surveys work, what they miss, and why they matter.
Every fish that swims through a kelp forest leaves something behind: skin cells, mucus, scales, waste — all of it carrying DNA. Filter a few litres of seawater, extract that genetic trace, and you can produce a species list for a reef without seeing a single animal. This is environmental DNA, or eDNA, and over the past decade it has moved from a laboratory curiosity to one of the most practical tools in marine monitoring. For anyone trying to measure whether a kelp forest is truly recovering — not just whether kelp is standing, but whether the whole community is coming back — eDNA has become almost indispensable.
Organisms constantly shed DNA into their environment. In the ocean, that DNA floats suspended in the water column or settles into sediment, where it persists for hours to days before sunlight, microbes and chemistry break it down. An eDNA survey is, at heart, a three-step argument:
The elegance is in step one: the ocean does the sampling. One dive team can collect water at ten sites in a morning that a visual survey crew would need weeks to census.
Two lab approaches dominate, and knowing the difference matters for reading any eDNA study.
Metabarcoding uses PCR primers that amplify a short "barcode" gene region shared across a broad group — the 12S gene for vertebrates, COI for many invertebrates — then sequences everything that amplified, in bulk, from one sample. The result is a community-scale species list. This is the tool for biodiversity questions: how many fish species use this reef? Did the invertebrate community shift after the heatwave?
Quantitative PCR (qPCR) and its digital cousin (ddPCR) look for one species at a time, using primers designed to match only that species' DNA. They are exquisitely sensitive — the method of choice for detecting rare, cryptic or invasive species. Searching for the first returning juvenile sunflower sea stars on a recovering reef, or watching for an invasive alga's arrival, is qPCR territory.
In practice, serious programmes run both: metabarcoding for the community picture, targeted assays for the species that matter most.
Honesty about limits is what separates eDNA as a science from eDNA as a sales pitch.
Detection is not abundance. A strong sequence signal might mean many animals, one large animal nearby, or DNA washed in from elsewhere. Water moves; DNA moves with it. Researchers are making progress on quantitative interpretation, but "how many" remains far harder than "whether".
DNA degrades and drifts. A signal reflects the last hours-to-days of life in roughly the local water mass — usually a strength, sometimes a blur, especially on current-swept headlands.
The reference library is the ceiling. Sequences are identified by matching to databases of known DNA barcodes. If a species has never been barcoded — and plenty of invertebrates haven't — its DNA is invisible or misassigned. Library gaps skew results toward well-studied groups like fish.
It can't see size, age or health. eDNA won't tell you whether the kelp forest holds big old breeding fish or only juveniles, whether an abalone bed is reproductive, or whether a black abalone population is breeding or merely persisting. For that you still need eyes in the water.
The mature position, and the one we take in our own monitoring, is that eDNA complements rather than replaces visual surveys, catch data and remote sensing. Each method sees a different slice of the truth.
Kelp restoration succeeds or fails at the community level. Standing kelp is the beginning; the endpoint is a functioning food web — grazers in balance, predators present, juvenile fish recruiting. That is precisely the kind of multi-species, repeating measurement eDNA is built for, and it slots neatly into the modern restoration toolkit:
At Tidewright, water samples from our restoration and reference sites feed into the same monitoring platform as our satellite canopy maps and drone surveys — the biodiversity layer of Canopy Ledger, sitting alongside the carbon layer. A restored forest that only regrew kelp, and not its community, would show up in that data, and it would not count as success.
Three frontiers are worth watching:
Typically hours to a few days, depending on temperature, sunlight and microbial activity. Warm, sunlit surface water degrades DNA fastest. This short persistence is often useful: what you detect is what was recently, and roughly locally, alive.
No — and it shouldn't try to. eDNA sees presence far more sensitively than divers can, but it cannot see size structure, behaviour or habitat condition. The strongest monitoring programmes, including ours, pair eDNA with visual surveys and remote sensing rather than substituting one for the others.
Surprisingly little — commonly one to five litres per site per replicate, though protocols vary with turbidity and target. The discipline is in replication and contamination control, not volume.
Per-sample lab costs have fallen dramatically over the past decade, and the real economy is in the field: replacing weeks of diver time with a morning of water collection changes what questions are affordable to ask at all.
Yes — filtering seawater through a field kit is well within trained volunteer capability, and community-collected eDNA already contributes to genuine monitoring programmes. Our careers and education coverage tracks citizen-science openings, and the explore section covers how divers can fold data collection into ordinary diving.
Marine heatwaves are weeks of anomalously hot ocean that can deforest entire coastlines. Here's the science behind them and what restoration can do.
Dr. Lena Fischer ·
Giant kelp is the fastest-growing organism on Earth and the architect of undersea forests. Learn its biology, life cycle, range and why it matters.
Dr. Theo Nakamura ·
The sunflower sea star was the kelp forest's great urchin hunter. Wasting disease erased billions of them — and the consequences reshaped coastlines.
Dr. Mara Ellison ·