Short Answer: Airborne eDNA signals are primarily local, but DNA can travel considerably further than most practitioners initially assumed. The current best estimate from a national-scale study is that the median transport distance at near-ground level is around 18.6 km, with detections possible but rare beyond 75 km. In contrast, zoo studies show strong, reliable signals from within a few hundred metres. The apparent contradiction resolves when you consider particle size, atmospheric conditions, and height above ground: large particles settle quickly, small particles travel further. The honest answer is that transport distance depends critically on what organism you are looking for, what particles carry its DNA, what the weather is doing, and the landscape topology.
1. Why transport distance matters
The question of how far airborne eDNA travels is the single most important practical question for anyone designing a monitoring programme or interpreting its results. It determines:
These questions cannot be answered by a single number. Transport distance is not a fixed property of airborne eDNA — it is a variable that depends on the physics of the carrier particle, atmospheric dynamics, and local geography.
2. The zoo studies: strong signal at hundreds of metres
The earliest evidence for airborne eDNA transport distances came from zoo studies, where the known positions of animals allowed source-detector distances to be precisely measured.
Clare et al. (2022) at Hamerton Zoo found that animal DNA was detectable at distances of several hundred metres from known source locations, with no significant drop in concentration over this range. Critically, DNA from enclosed indoor animals escaped sealed buildings and was detectable in external air samples. Lynggaard et al. (2024) at a Danish mixed forest found that detection probability decreased with distance from known animal positions, but detections were still made at the limits of the experimental design (up to the scale of the study area).
Jager et al. (2025) at Rotterdam Zoo found their passive DNAir sampler detecting signal at up to 515 metres from source animals, with detection probability correlating positively with total species biomass. Kroos et al. (2026), working in open farmland with targeted qPCR for invasive wallabies, detected signal at up to 1,000 metres using a high-volume active sampler — but at that distance, detection was inconsistent and not guaranteed across replicates. This is an important qualification: 1,000 m may represent the outer edge of what is physically possible under favourable conditions, not a distance at which reliable detection can be planned for. The field has not yet reached consensus on what a "reliable" detection distance looks like in practice. Zoo studies suggest a few hundred metres; targeted qPCR in farmland can reach 1 km; community metabarcoding in natural landscapes has not yet established a clear distance limit, partly because the answer depends heavily on sampler volume, target taxon, and atmospheric conditions at the time of sampling.
Newton et al. (2026) found that airborne eDNA and spider webs together outperformed other substrates for vertebrate detection at the study site.
Key zoo finding: At near-ground level in enclosed or semi-enclosed environments, airborne eDNA provides a reliable, strong signal from within a radius of a few hundred metres. This is adequate for monitoring specific sites, nature reserves, or enclosed habitats.
3. Natural landscape studies: a broader but fuzzier catchment
Moving from controlled zoo settings to open natural landscapes introduces complexity. Wind patterns, topography, vegetation structure, and atmospheric stability all modulate how far particles travel and where they deposit.
Lynggaard et al. (2024) provided among the first rigorous evidence from a multi-day, multi-height natural forest study for airborne eDNA transport in a natural landscape, sampling at multiple heights and distances in a Danish mixed forest. They found clear evidence that weather conditions — particularly wind speed and precipitation — influenced which vertebrate taxa were detected and in what abundance. Rain washed particles out of the air, reducing detection probability. Higher wind speeds increased detection of taxa further from source.
Polling et al. (2024) compared airborne eDNA against camera trap records at three Dutch habitats and found that eDNA consistently detected species beyond the spatial footprint covered by the cameras — including species detected in the broader landscape but not immediately adjacent to the sampler.
The largest scale natural landscape data come from Tournayre et al. (2025), who used citizen science (iNaturalist) occurrence records to estimate the catchment radius of individual air quality monitoring stations used in their national UK survey. They found:
- Median estimated transport distance: 18.6 km(estimated as the radius within which iNaturalist detections matched eDNA detections for most tested taxonomic groups)
- Most signal is local:detections within 5 km were far more consistent than at 18.6 km
- Long transport is possible but not predominant:eDNA sampling near ground level (~1.5–2 m, the height of UK air quality samplers, in the human breathing zone) is constrained by local topography and landscape features
This 18.6 km figure should be interpreted carefully. It represents a maximum catchment estimate derived by an indirect method, not a direct physical measurement. Tournayre et al. used iNaturalist citizen science records as a proxy for species presence at different radii around the samplers. iNaturalist records have known spatial biases — they cluster around roadsides, accessible habitats, and areas of high observer effort — and the radius at which iNaturalist detections converge with eDNA detections may partly reflect those recording biases rather than true eDNA transport distances. Independent direct-measurement estimates of transport distance from precisely known source populations in open natural landscapes are still lacking. The 18.6 km figure is the best available estimate, but should be treated as an order-of-magnitude guide rather than a precise physical constant. Many detections will come from much closer; the estimate also represents landscape-level averages and will vary substantially with terrain, vegetation, and atmospheric conditions.
4. The archive study: regional and continental transport
Sullivan et al. (2025) took a different approach to understanding transport distances, using atmospheric back-trajectory modelling (HYSPLIT) to estimate the geographic origins of specific eDNA detections from their Swedish radionuclide archive.
For 22 µm particles (typical of many biological particles), their modelling showed:
- The primary catchment area centred on the immediate regional landscape within ~50 km
- Seasonal variation was substantial — summer back-trajectories were dominated by local sources; winter back-trajectories showed contributions from regions hundreds of kilometres away
- Detection of cod (Gadus) DNA at an inland Swedish site was explicable by back-trajectories showing marine-origin air masses from the Baltic and Norwegian seas, consistent with sea-spray aerosol transport of marine organisms
This is a profound result: not only can airborne eDNA be transported hundreds of kilometres under some atmospheric conditions, but the geographic origin of a signal can be traced using existing atmospheric modelling tools from meteorology. This is not a new technique — HYSPLIT back-trajectory analysis has been used for decades to source atmospheric pollutants, pollen, and volcanic ash (Stein et al. 2015; Sofiev et al. 2006) — but its application to biodiversity data is genuinely novel.
5. What aerosol physics tells us about particle-specific transport
Understanding why transport distance varies requires understanding aerosol physics. Biological particles in the atmosphere do not all behave the same way — their transport distance is governed primarily by their aerodynamic diameter and density, which determine their terminal settling velocity.
The fundamental relationship is described by Stokes' law for particle settling:
Settling velocity ∝ particle diameter² × particle density
Consequently:
| Particle | Diameter | Settling velocity | Typical atmospheric residence |
|---|---|---|---|
| Birch pollen | 20–25 µm | ~0.01–0.03 m/s | Hours to ~1 day |
| Grass pollen | 25–35 µm | ~0.02–0.05 m/s | Hours |
| Pine pollen (saccate) | 50–80 µm | ~0.05–0.15 m/s | Less than 1 day |
| Sub-pollen particles | 0.5–4 µm | <0.001 m/s | Days to weeks |
| Animal cell fragments | 1–20 µm | 0.0001–0.01 m/s | Hours to days |
| Bacterial cells | 0.5–5 µm | <0.001 m/s | Days to weeks |
Two important caveats apply to the table above. First, the values assume spherical particles of unit density (~1.0 g/cm³). Real biological particles deviate from both assumptions: biological particle densities range from ~0.9 g/cm³ for pollen grains with air sacs to ~1.4 g/cm³ for dense fungal spores, and irregular shapes (spines, lobes, cell fragments) cause particles to settle 1.5–3× more slowly than smooth spheres of equivalent volume, due to the shape correction factor in Stokes' Law (Hinds 1999, Aerosol Technology). Second, for particles below ~5 µm, gravitational settling becomes a secondary removal mechanism. In this size range, dry deposition is dominated by turbulent impaction (for particles 1–5 µm) and Brownian diffusion (for particles below ~1 µm). Particles in the accumulation mode (~0.1–1 µm) have a deposition velocity minimum — they are efficiently neither settled by gravity nor captured by diffusion — and can remain suspended for days to weeks. This actually strengthens the case for long-range airborne eDNA transport: the smallest DNA-carrying particles are the hardest to remove from the atmosphere.
Small particles remain aloft far longer and travel much further than large ones. Sullivan et al. (2025) modelled particles of 5, 22 and 60 µm diameter and found the catchment area differed substantially depending on particle size. The 60 µm particles (comparable to large pollen grains) showed a tight local catchment; the 5 µm particles showed regional catchments extending hundreds of kilometres.
This means that the transport distance of airborne eDNA depends on which particle carries the DNA. Vertebrate skin cell fragments and sub-pollen particles — which are small — can travel further than intact pollen grains. Bacterial and fungal DNA on very fine particles can travel intercontinental distances under favourable conditions (Prenni et al. 2009; Desprès et al. 2012).
6. How atmospheric conditions modulate transport
Transport distance is not solely a function of particle size. Atmospheric dynamics — wind speed and direction, boundary layer height, precipitation, and local turbulence — can dramatically amplify or constrain how far particles travel.
Wind acts as the primary transport vector: moderate winds extend the catchment; very high winds also dilute the local signal by mixing DNA into a larger air volume. Boundary layer dynamics determine whether particles are lifted high for long-range transport or confined near the ground for local accumulation. Rain is the primary removal mechanism — it scavenges particles from the air column and resets the atmospheric pool; post-rain detections are typically depleted until resuspension occurs. Topography constrains near-ground transport substantially, which is why fixed-point samplers at low height tend to capture locally representative signals rather than regional ones.
7. Transport distance versus detection distance
A distinction that is frequently blurred in the literature — and that matters considerably for monitoring design — is the difference between physical transport distance (how far DNA particles actually travel before depositing) and operational detection distance (the furthest distance at which a given sampler and assay can reliably detect DNA from a known source).
These are not the same. At 1 km from a source, DNA from that source is almost certainly present in the air — but in concentrations that may fall below the PCR detection threshold of a particular sampler-volume-primer combination. The Kroos et al. (2026) 1,000 m detection limit does not mean DNA does not travel beyond 1,000 m; it means the specific experimental setup could not reliably detect it at that range. A system collecting higher total air volumes per sample, or using a more sensitive targeted assay, would extend the operational detection distance without changing the underlying physics of transport.
This means that statements about detection distance are always statements about a specific methodology. Improvements in sampling volume, filter efficiency, DNA extraction, or assay sensitivity will extend detection distance. The physics-based transport limits from atmospheric modelling represent the theoretical ceiling; operational detection limits depend on the method.
Mobile sampling fundamentally changes the catchment concept. A fixed-point sampler has a catchment defined by wind direction — essentially a cone upwind of the inlet. A vehicle-mounted mobile sampler moving along a 25 km route integrates a route-length catchment: instead of sampling a single point's upwind sector, it samples a strip of landscape continuously. Martino et al. (2025) demonstrated this concretely: five taxa were recovered exclusively from mobile vehicle-mounted samplers, from habitats the stationary positions never faced downwind of. The catchment of a mobile system is therefore not a circle or cone, but a corridor shaped by the route, which can be designed to traverse specific habitat types or follow ecological gradients.
8. The concept of an eDNA catchment area
By analogy with hydrology (where a watershed defines the land area that drains to a given point), the concept of an eDNA catchment area or source catchment describes the geographic region from which organisms contribute detectable DNA to a given sampler at a given time.
This catchment is not a fixed circle — it is a dynamic, wind-shaped zone that changes with meteorological conditions. Sullivan et al. (2025) illustrated this elegantly with HYSPLIT back-trajectory maps showing that the catchment for their Swedish station was elongated along prevailing wind directions and shifted seasonally with large-scale atmospheric circulation patterns.
For practical monitoring design, the concept of a catchment area raises an important question: is this a feature or a bug? On one hand, a large catchment means a single sampler represents a large area — making monitoring cost-effective. On the other hand, it means a detection may not confirm local species presence, complicating interpretation. Tournayre et al. (2025) noted that the 18.6 km estimated median catchment provides a "sweet spot for national level biomonitoring" — large enough to provide useful regional data, small enough to be ecologically interpretable.
9. Practical implications for monitoring design
Based on current evidence, the following practical conclusions can be drawn for monitoring programme design:
For site-level monitoring (e.g. a specific nature reserve, Natura 2000 site, or biosecurity perimeter): A near-ground active or passive sampler will primarily reflect the local biological community within a few hundred metres to a few kilometres. This is sufficient for detecting species presence within or immediately adjacent to a site.
Sampler placement and orientation matter for active samplers. Active samplers draw air through an inlet, creating a directional catchment shaped by wind — the sampler primarily captures organisms upwind of the inlet. Sampling downwind of a target habitat is substantially more effective than sampling crosswind or in still air conditions. For monitoring a specific habitat (e.g. a fen, a woodland edge), placing the sampler on the downwind side relative to the dominant wind direction during the sampling window maximises detection probability. Monitoring protocols should record sampler orientation and wind direction during sampling as standard metadata.
For landscape-level monitoring (e.g. a 20 km² farmland biodiversity survey): Air quality network stations at 1.5–2 m height will likely integrate signal over a broadly regional catchment under typical atmospheric conditions — but what this means for species-level detection is not yet clear. No large multi-site, long-term study has yet demonstrated that airborne eDNA can serve as a reliable species inventory tool at landscape scale. The evidence base is promising but still early-stage: the method is plausible as a landscape-level biodiversity indicator, but the specific sampling designs, replication requirements, and detection thresholds needed to make this operational remain to be established.
Seasonal replication is essential: A single sampling event captures the community at one moment in one season. Multi-season or continuous sampling (as provided by repurposed air quality networks) captures a much more complete community picture.
10. Caveats and contested claims
The 18.6 km figure is an estimate, not a measurement. It was derived by comparing eDNA detections with iNaturalist occurrence records at different radii — an indirect approach with substantial uncertainty. Direct measurements of eDNA transport distance in open landscapes, using known source locations, remain rare.
Zoo studies likely overestimate field performance. The dense animal populations and semi-enclosed conditions of zoos produce elevated eDNA concentrations not representative of wild populations. As Clare herself acknowledged, detection at several hundred metres in a zoo does not guarantee equivalent detection in open countryside where the same species might be represented by a handful of individuals per square kilometre.
Conflicting results exist between studies. Different sampler types, heights, weather conditions, and sequencing approaches make direct comparison between studies difficult. This is part of why the field will need standardised protocols (Tulloch et al. 2025).
References
- Tournayre O et al. (2025). First national survey of terrestrial biodiversity using airborne eDNA. Scientific Reports 15:19247. https://doi.org/10.1038/s41598-025-03650-z
- Sullivan AR et al. (2025). Airborne eDNA captures three decades of ecosystem biodiversity. Nature Communications 16:11281. https://doi.org/10.1038/s41467-025-67676-7
- Lynggaard C et al. (2024). Airborne environmental DNA captures terrestrial vertebrate diversity in nature. Molecular Ecology Resources 24:e13840. https://doi.org/10.1111/1755-0998.13840
- Polling M et al. (2024). Continuous daily sampling of airborne eDNA detects all vertebrate species. Environmental DNA 6:e591. https://doi.org/10.1002/edn3.591
- Clare EL et al. (2022). Measuring biodiversity from DNA in the air. Current Biology 32(3):693–700. https://doi.org/10.1016/j.cub.2021.11.064
- Garrett NR et al. (2023). Out of thin air: surveying tropical bat roosts through air sampling of eDNA. PeerJ 11:e14772. https://doi.org/10.7717/peerj.14772
- Jager H et al. (2025). A breath of fresh air: passive vs active airborne eDNA sampling. bioRxiv. https://doi.org/10.1101/2025.03.26.645491
- Newton JP et al. (2026). Airborne DNA and spider webs outperform other eDNA sources. Molecular Ecology Resources. https://doi.org/10.1111/1755-0998.70067
- Stein AF et al. (2015). NOAA’s HYSPLIT atmospheric transport and dispersion modeling system. BAMS 96:2059–2077. https://doi.org/10.1175/BAMS-D-14-00110.1
- Sofiev M et al. (2006). Towards numerical forecasting of long-range air transport of birch pollen. International Journal of Biometeorology 50:392–402. https://doi.org/10.1007/s00484-006-0027-x
- Desprès VR et al. (2012). Primary biological aerosol particles in the atmosphere: A review. Tellus B 64:15598. https://doi.org/10.3402/tellusb.v64i0.15598
- Prenni AJ et al. (2009). Relative roles of biogenic emissions and Saharan dust as ice nuclei in the Amazon basin. Nature Geoscience 2:402–405. https://doi.org/10.1038/ngeo517
- Tulloch RL et al. (2025). Winds of Change: Charting a Pathway to Ecosystem Monitoring Using Airborne Environmental DNA. Environmental DNA 7:e70134. https://doi.org/10.1002/edn3.70134
- Hinds WC & Zhu Y (2022). Aerosol Technology: Properties, Behavior, and Measurement of Airborne Particles (3rd ed.). John Wiley & Sons.
- Johnson M & Barnes MA (2024). Macrobial airborne environmental DNA analysis: A review. Molecular Ecology Resources 24:e13998. https://doi.org/10.1111/1755-0998.13998
- Martino J et al. (2025). A direct comparison of fixed and vehicle-mounted airborne eDNA sampling methods for terrestrial vertebrate species detection at large spatial scales. bioRxiv preprint. https://doi.org/10.1101/2025.09.02.673667
- Skjøth CA et al. (2007). The long-range transport of birch pollen causes significant pre-season concentrations in Denmark. Clinical & Experimental Allergy 37:1204–1212. https://doi.org/10.1111/j.1365-2222.2007.02771.x