The Gold Standard of Ecology | Sediment Resuspension, Legacy eDNA and False Recent Presence

1. Define what recent presence means for the decision

Back to contents

Recent can mean hours for a rapid-response operation, days for a spawning survey or months for seasonal occupancy. State the organism, life stage, water body, spatial claim and decision window before interpreting the sequence. The same detection may be adequate for broad catchment surveillance and inadequate for declaring a live local population.

Separate three claims: target DNA occurred in the sample, target material entered the water system, and living organisms occupied the site during the stated window. Molecular replication strengthens the first claim. Sediment and transport evidence are needed to move toward the third.

2. Treat sediment as both sink and delayed source

Back to contents

Cells, tissue fragments and extracellular DNA can adsorb to minerals or organic particles and settle. Burial may reduce exposure to ultraviolet light and change microbial degradation, allowing some fragments to persist longer than in the water column. The amount retained depends on grain size, organic matter, chemistry, temperature and the physical state of the DNA.

Sediment storage is not a perfect archive with a known date. Layers can mix through floods, burrowing animals, propeller wash or sampling. A later water detection can arise when stored particles are reintroduced, even if no new shedding occurred nearby.

3. Measure the processes that resuspend material

Back to contents

Record discharge, shear stress, waves, wind, tides, rainfall, dam operations, vessel activity, dredging and benthic disturbance as relevant. A turbidity spike can mark particle mobilisation, but turbidity alone does not identify which sediment layer or DNA age entered the water.

Sample around events rather than only after them. Pair eDNA with suspended solids, particle size, tracers or hydrodynamic observations. If the detection appears only during disturbance, resuspension becomes a plausible source that must be compared with fresh biological shedding.

4. Distinguish particle states and analytical pathways

Back to contents

Different filters capture different size fractions, and centrifugation, prefiltration or extraction changes which DNA states are measured. A particle-rich storm sample may yield more sediment-bound DNA and more inhibitors at the same time. Report filtration, mass or volume, inhibition controls and recovery so concentration changes are not mistaken for ecology.

Consider size-fractionated sampling or paired water and suspended-sediment extracts when legacy signal is central. The aim is not to assign an exact age from particle size, but to test whether detections track the fraction most likely to have been stored and remobilised.

5. Design water, sediment and timing controls together

Back to contents

Collect upstream or external controls, local water replicates, surficial sediment and, where appropriate, deeper sediment or settling material. Repeat before, during and after a resuspension event. Field and laboratory blanks still matter because legacy DNA is not the only false-recent mechanism.

A time series can distinguish a brief particle pulse from sustained fresh input. If DNA, turbidity and suspended sediment peak together then fall while a live-source indicator remains absent, the evidence supports remobilisation more than ongoing local shedding.

6. Worked example: invasive fish DNA after dredging

Back to contents

A harbour survey reports a strong invasive-fish signal one day after maintenance dredging. No fish are captured, earlier water samples were negative and the signal is concentrated in the larger particle fraction. Archived sediment cores contain the same target at low levels, and turbidity remains elevated near the dredged berth.

The team does not announce a newly established population. It repeats water sampling after particles settle, samples adjacent sediment and uses an independent assay. Detections disappear within four days. The management report states that target DNA was confirmed but recent live presence was not established; surveillance continues at likely biological entry points.

7. Use multiple lines of evidence for recency

Back to contents

Shorter DNA fragments, RNA, environmental context, repeated concentration trajectories or organism-specific behaviour may sometimes inform recency, but none is a universal clock. Degradation rates vary, and laboratory persistence may not transfer to protected sediment. Capture, visual, acoustic or telemetry evidence can provide a more direct check when feasible.

Use a decision matrix that combines molecular repeatability, disturbance timing, particle association, sediment occurrence and independent organism evidence. Predefine what combination triggers containment, further sampling or a public claim.

8. Changed-context transfer check: a reservoir drawdown

Back to contents

Move from a dredged harbour to a reservoir whose water level drops and exposes old littoral sediment. Wind then refloods and resuspends the margin. The transferable logic is to compare current detections with storage, disturbance and repeated post-event sampling. The changed system adds shoreline erosion, stratification and long retention times.

Stations should follow circulation and depth, not just distance from shore. A positive surface sample may represent mixed legacy material, while a persistent depth-specific plume near suitable habitat may support a different interpretation.

9. Model alternative sources instead of choosing one story

Back to contents

Build competing explanations: fresh local shedding, upstream transport, wastewater or bait input, sediment resuspension and contamination. Link each to expected timing, particle fraction, spatial pattern and control results. A Bayesian or ensemble model can quantify uncertainty, but even a transparent qualitative evidence table is better than a single untested narrative.

Do not infer age directly from concentration. A high legacy pulse can exceed weak fresh shedding, and dilution can hide a nearby population. The inference should report which sources remain compatible with the evidence and what observation would separate them.

10. Know the limits and common misconceptions

Back to contents

A legacy detection is not necessarily a laboratory false positive; the target DNA can be genuinely present while the ecological timing claim is wrong. Sediment does not always preserve DNA longer, and resuspension does not explain every storm-associated signal. Living organisms may also become active or move during high flow.

Negative follow-up samples do not prove no organism was present. They can reflect settling, dilution, inhibition or imperfect detection. The correct conclusion is bounded by the sampling design, hydrodynamics and assay performance.

11. Audit the field-to-recency inference chain

Back to contents

Publish sampling times, coordinates, depths, recent disturbances, hydrology, turbidity, suspended solids, sediment methods, filtration, controls, inhibition tests, replicate results, thresholds and assay sequences. Preserve raw amplification and sequencing records.

Show the timeline linking environmental events to detections and non-detections. Reanalyse under alternative decay, transport and resuspension assumptions. A future reader should be able to see why the study supported recent presence, legacy signal or continued uncertainty.

12. The gold standard separates detection from biological time

Back to contents

Environmental DNA becomes more useful when its physical history is treated as evidence rather than nuisance. Sediment can store, protect, move and release DNA, so a sample date is not automatically the date of biological presence.

The gold standard defines the management window, samples around disturbance, measures particles and controls, tests alternative sources and seeks independent evidence where the decision is costly. It reports a positive molecular result confidently while refusing to assign false recency that the water and sediment cannot support.