Filtering is one of the most important design choices in aquatic environmental DNA surveys. Water contains DNA as dissolved fragments, organelles, cells, mucus and material attached to particles. A filter with smaller pores may retain more of the fine fraction, but it can clog quickly in turbid or organic-rich water. A larger pore may allow more water to be processed while letting some small DNA-bearing material pass. Filter material also affects binding, extraction and inhibition. Consequently, “we filtered one litre” is not enough information to compare studies. Researchers must report pore size, membrane material, housing, volume, preservation, water conditions and whether multiple filters were combined. Pilot tests are essential because the most effective protocol balances capture efficiency, workable volume, contamination control and reliable laboratory recovery for the target organism and habitat.
Key takeaways
- Environmental DNA occurs in several particle-size fractions.
- Smaller pores can improve retention but increase clogging.
- Larger pores may permit a greater sample volume.
- Filter material influences binding, strength and extraction.
- Turbidity and organic matter can increase inhibition.
- Pilot studies should optimise the whole workflow, not pore size alone.
What is being captured
Organisms shed skin cells, mucus, faeces, gametes and other biological material. DNA may remain inside cells, attach to mineral or organic particles, or occur as small extracellular fragments. These fractions behave differently in water and across a membrane. An assay detects target sequences after extraction; it does not reveal which physical form carried each copy unless the study separates fractions.
Pore size is not a simple sieve
A nominal pore size describes the membrane but does not create an absolute boundary. Irregular particles, adsorption and the growing filter cake affect retention. Material smaller than the stated pore may stick, while deformation or channeling may alter passage. Comparisons should use empirical recovery tests rather than assuming every DNA fragment above one size is captured and everything below is lost.
The capture–volume trade-off
Small pores can trap fine material efficiently yet stop flow after a few hundred millilitres in muddy water. A larger pore may capture a lower fraction per millilitre but process many litres, potentially collecting more total target DNA. The optimum depends on concentration, particle distribution, pump pressure, labour and the point at which clogging makes additional volume impractical.
Filter materials
Common membranes include cellulose nitrate, cellulose acetate, glass fibre, polyethersulfone and enclosed cartridges. They differ in protein and DNA binding, chemical compatibility, mechanical strength and extraction method. A material successful in a clear stream may perform poorly in seawater or a tannin-rich wetland. Method changes need validation rather than being treated as interchangeable.
Open membranes and enclosed capsules
Open filter holders allow visual inspection and flexible membrane choice but increase handling. Enclosed capsules reduce direct contact and can process larger volumes, although they cost more and require an extraction method suited to the housing. Field logistics, clean handling and waste disposal can be as important as theoretical capture efficiency.
Turbidity and filter cake
Suspended sediment and algae accumulate on the membrane, raising pressure and slowing flow. The deposited layer can trap additional DNA, effectively changing filtration during the sample. It can also concentrate inhibitors. Recording turbidity, final volume, filtration time and pressure helps explain why two nominally identical filters produced different recoveries.
Sample volume
More water usually increases the opportunity to capture rare DNA, but returns can diminish and contamination or inhibition risk can rise. Several replicate filters may represent spatial variation better than one enormous volume. USGS guidance emphasises testing the combination of filter, pore size and volume under local water-quality conditions rather than adopting a universal litre target.
Pressure and pumping
Vacuum, peristaltic pumps, syringes and gravity systems impose different pressures and contact surfaces. Excess pressure can damage a membrane or force material through. Tubing can retain DNA between samples. Flow paths should be cleaned or single-use, and blanks should pass through the same equipment so contamination introduced by the apparatus is detectable.
Preserving the filter
DNA begins degrading after collection. Filters may be frozen, dried or stored in preservation buffer, depending on method and field conditions. Preservation must reach all trapped material, particularly a thick filter cake. Time, temperature and transport should be recorded. A highly efficient capture step is wasted if DNA deteriorates before extraction.
Extraction efficiency
The laboratory must release DNA from the membrane and remove substances that interfere with amplification. A filter that captures abundant material may yield little amplifiable DNA if extraction is inefficient. Spike controls and internal positive controls help distinguish low target abundance from inhibition. Comparing protocols requires measuring the entire capture-to-result chain.
Inhibition
Humic acids, metals, sediment and other compounds can reduce polymerase-chain-reaction performance. Diluting the extract may reduce inhibition while also diluting target DNA. Cleanup steps can lose DNA. A non-detection from a clogged, inhibitor-rich sample should not be interpreted like a clean sample with a strong internal control.
Replicates and spatial variability
DNA can be patchy across a river, lake or coastal site. Multiple independent water samples reveal sampling variation; repeated PCR wells reveal analytical variation. Filtering one combined volume answers a different question from filtering several replicates. The design should preserve the level at which detection probability and uncertainty will be estimated.
Contamination controls
Field blanks, equipment blanks, filtration blanks and laboratory extraction controls identify contamination at different stages. A blank should experience the same handling, filter and preservation as a sample. Positive detections in controls require investigation under pre-set rules. Smaller expected target concentrations make even tiny contamination important.
Comparing studies
Two surveys using different pore sizes, volumes, extraction kits and replicate rules do not have identical detection probabilities. A higher positivity rate may reflect biology or method. Meta-analysis and long-term monitoring should retain protocol metadata and, when methods change, run an overlap comparison. Standardisation helps, but local optimisation may still justify documented differences.
Choosing a protocol
Start with the management question and target concentration, then test several filter–volume combinations using real site water and known additions where appropriate. Measure processing time, clogging, DNA recovery, inhibition and variation. Select a method that field teams can perform consistently. The best laboratory result is not useful if the protocol routinely fails under field conditions.
A classroom filtration experiment
Students can pass equal volumes of clear and muddy water containing harmless coloured particles through coarse and fine filters. They record time, retained colour and final volume, then repeat with a larger volume through the coarse filter. The model does not reproduce molecular DNA, but it demonstrates why retention, clogging and processed volume must be considered together.
Size fractionation experiments
Researchers can pass the same water sequentially through filters of decreasing pore size or analyse material retained in several fractions. This estimates where target DNA resides physically and whether that distribution changes with species, season or degradation. Sequential filtration also changes the sample as each stage removes particles, so order matters. Fractionation studies inform protocol choice but should not be confused with ordinary monitoring, where extra handling can raise contamination and DNA-loss risk.
Flow cytometry and microscopy as supporting tools
Particle counts, microscopy and turbidity measurements describe the material challenging a filter. They cannot identify target DNA by themselves, but they help explain clogging and differences between sites. A sudden shift towards fine particles may reduce capture by a coarse membrane, while an algal bloom can increase filter cake and inhibitors. Recording these supporting variables turns an unexplained change in detection into a testable change in sample matrix.
Field standardisation and operator effects
Different technicians may stop filtration at different pressures, judge clogging differently or handle membranes for unequal times. Written stopping rules, flow meters, training and duplicate samples reduce these operator effects. A protocol should explain whether the target is a fixed volume, fixed time or pressure limit and what happens when it cannot be achieved. Consistency supports trend detection, but safety and contamination control take priority over forcing every sample to the nominal volume.
Method changes in a long monitoring programme
A better filter can improve detection while making the time series appear to show ecological increase. Before switching, teams can run old and new methods side by side across representative conditions. The overlap estimates a conversion or at least documents the discontinuity. Archived extracts or filters may allow retrospective comparison if preservation is sound. Reports should mark the change clearly rather than merging results into one uninterrupted trend line.
Cost, waste and survey scale
A highly sensitive filter may be expensive, slow or require cold transport, limiting the number of sites and replicates. A slightly less efficient method applied consistently across many locations can answer a distribution question better. Single-use plastics and chemical preservatives also create waste and transport burdens. Protocol choice should compare detection performance per site with total survey coverage, staff time, safety and quality control. Cost cutting becomes false economy if insufficient blanks or replication make positive and negative results uninterpretable.
Reporting a filtration method
A reproducible report states water source, collection depth, container, time to filtration, membrane material, nominal pore size, housing, pump, pressure or flow, attempted and achieved volume, clogging rule, preservation, storage temperature and extraction method. It also lists field and laboratory controls, replicates and inhibition tests. These details may seem mundane beside the species result, but they define detection probability. Without them, another team cannot determine whether a different result reflects ecology, hydrology or filtration.
Questions to ask before accepting the conclusion
For any claim about eDNA filter pore size, identify what was measured directly, which steps relied on models or judgement, and whether the sampled places and times represent the wider question. Check the units, comparison group, missing observations and uncertainty range. Look for an independent method with different weaknesses and ask whether it supports the same central result. Finally, separate a useful screening signal from evidence strong enough for an expensive or irreversible decision. This checklist does not demand perfect certainty; it ensures that the confidence of the conclusion matches the quality and scope of the evidence.
Related Earth Site reading
Useful background is available in eDNA transport and decay, environmental DNA monitoring, eDNA detection limits and false positives in eDNA surveys. These public Earth Site pages were checked as internal destinations; the independent sources listed in the companion notes support the research in this article.
Conclusion
Filter choice shapes which eDNA fractions are collected, how much water can be processed and how many inhibitors reach the laboratory. A smaller pore is not automatically more sensitive, and a larger volume is not automatically more representative. The useful unit is the complete, validated protocol: site selection, filter material, pore size, volume, preservation, extraction, controls and replicate design. Reporting those details allows results to be compared honestly and helps managers distinguish ecological change from a change in sampling method.
Frequently asked questions
Does a smaller pore always capture more eDNA?
It may retain finer material, but clogging can reduce the total volume and increase inhibition.
What pore size should every study use?
There is no universal size; pilot testing should match the target and water conditions.
Why record filtered volume?
Detection opportunity depends partly on how much water actually passed through the filter.
Can muddy water cause false negatives?
Yes. Clogging limits volume and co-collected substances can inhibit extraction or amplification.
Are different filter materials interchangeable?
No. Binding, strength and extraction efficiency differ and should be validated.
