A study part-funded by NatureScot models substantial long-term effects from wind-turbine collisions on hen harriers and golden eagles in the Flow Country. The agency disputes the scale of the predicted losses, while acknowledging that Scotland does not hold a comprehensive national record of birds killed by turbines.
NatureScot has challenged the findings of a scientific study it helped fund which models potentially substantial long-term effects from wind farms on protected bird populations in northern Scotland.
The research examines the cumulative effects of wind-turbine collisions and displacement on birds associated with the Caithness and Sutherland Peatlands Special Protection Area and the wider Natural Heritage Zone 5, which includes much of the Flow Country.
Its most striking projection concerns the hen harrier.
Under the study’s collision assumptions, the hen harrier population associated with the Special Protection Area would, after 30 years, be only 44% of the size of an equivalent population modelled without wind-turbine collision mortality. Across the wider Natural Heritage Zone, the equivalent figure was 74%.
NatureScot says those projections overstate the actual risk.
The disagreement comes as NatureScot’s own records show at least 71 birds of prey were reported killed by collisions with onshore wind turbines in Scotland between 2014 and 2024.
Those records include four hen harriers, seven golden eagles and nine white-tailed eagles.
But NatureScot says the 71 deaths are not a complete Scottish total. Reports are submitted voluntarily and the agency explicitly states that it does not hold a comprehensive record of bird collisions with wind turbines. Offshore deaths are even more difficult to establish because birds that collide with turbines at sea may never be recovered.
The gap between modelled risk and recorded mortality is now at the centre of a considerably larger question about how Scotland measures the environmental consequences of expanding wind generation.
What the Flow Country study found
NatureScot Research Report 1350 was produced by Lucy Mitchell and Elizabeth Masden of the Environmental Research Institute and part-funded by NatureScot, Highland Council and the RSPB.
The researchers examined four species considered vulnerable to turbine collisions: hen harrier, golden eagle, red-throated diver and black-throated diver.
They also considered displacement effects on golden eagle, golden plover, dunlin and greenshank.
Rather than attempting simply to count dead birds, the researchers constructed population models to estimate what repeated additional mortality could mean over the expected life of wind developments.
For hen harriers in the Special Protection Area, the model reduced the population growth rate from 1.11 to 0.94 once estimated cumulative turbine-collision mortality was added.
After 30 years, the modelled affected population was 44% of the equivalent population without that collision effect.
The researchers also considered what could happen if a breeding male were killed. Female hen harriers undertake incubation and depend heavily upon males to provide food during incubation and the early nestling period. The loss of a male can therefore affect more than the individual bird if a nest subsequently fails.
Under that additional productivity scenario, the modelled population was 65% of the equivalent unaffected population after 30 years.
Golden eagles also showed a long-term difference.
For the Special Protection Area population, the modelled population exposed to cumulative collision mortality was 88% of the equivalent unimpacted population after 30 years. Across Natural Heritage Zone 5, it was 75%.
These are projections, not counts of birds which have already been killed.
The calculations depend on assumptions about population size, survival, productivity and expected collision rates. The researchers themselves identify uncertainty in some of those inputs.
NatureScot disputes the scale of the collision risk
NatureScot took the unusual step of publishing its own position statement alongside the research.
For hen harriers, the agency says there is “very little evidence gathered to date” of birds colliding with turbines.
It also says the collision estimates used by the researchers did not account for mitigation already required at some individual wind farms.
At Hollandmey, Slickly, Strathy North and Strathy South, for example, measures have been attached to developments with the intention of reducing habitat suitability or otherwise limiting predicted effects on hen harriers.
NatureScot also points to limited demographic information for the species, which reduces confidence in predicted survival rates.
Its conclusion is that actual residual collision risk is below the level which would cause the hen harrier population to decline at either Special Protection Area or Natural Heritage Zone scale.
The agency also believes the model overstates collision risk for golden eagles.
In their case, NatureScot argues that displacement is the greater concern because satellite-tracking evidence indicates that eagles tend to avoid wind farms.
That means a turbine does not have to kill an eagle to affect it. A development can alter the amount or quality of habitat available to a territory if birds avoid areas they previously used.
Scotland has recorded 71 raptor deaths, but says the figure is incomplete
The assertion that collision evidence is limited becomes more complicated when placed alongside NatureScot’s own mortality records.
In response to a freedom of information request, the agency disclosed 71 reported raptor deaths caused by collisions with onshore wind turbines over the decade to 2024.
The records include hen harriers, golden and white-tailed eagles, ospreys, red kites, buzzards, kestrels, peregrines, sparrowhawks, barn owls, a goshawk and a merlin.
Four of the recorded fatalities were hen harriers.
The number supports NatureScot’s position that reported hen harrier collisions have been uncommon. It does not establish that four were the only hen harriers killed.
NatureScot’s collision reporting system relies upon information sent voluntarily to an email address established in 2013.
The agency said directly in its response that it could not provide an accurate national figure because its records were not comprehensive.
That limitation is particularly relevant when assessing cumulative risk.
Scotland is attempting to estimate what repeated mortality may mean for bird populations over several decades without having a complete retrospective dataset showing how many birds have already collided with turbines.
There is no reliable worldwide death total
International research confirms that turbine collisions kill birds, but there is no scientifically reliable global total.
Carcass surveys require researchers to correct for bodies removed by scavengers, birds missed by search teams and casualties falling outside the area being searched. Monitoring standards vary between wind farms and countries, while obtaining comparable data from offshore developments is considerably more difficult.
Studies therefore report rates and modelled estimates rather than a worldwide count.
One large Canadian study examined carcass-search data from 43 wind farms and corrected the findings for scavenger removal, searcher efficiency and carcasses falling outside the search area.
It estimated an average of 8.2 birds killed per turbine per year.
The figure varied sharply between individual developments, from zero to 26.9 birds per turbine annually.
Applied to Canada’s 2,955 installed turbines at the end of 2011, the researchers estimated approximately 23,300 bird deaths a year, with a 95% confidence interval of 20,000 to 28,300.
The variation between wind farms is as significant as the average. Geography, habitat, migration routes, the species present and turbine design can all influence collision rates.
It would therefore be misleading simply to multiply Canada’s historic average by the number of turbines in Scotland and present the result as a Scottish mortality estimate.
North American research estimated hundreds of thousands of deaths
A separate peer-reviewed analysis examined 116 studies at 71 wind-energy facilities in the United States and Canada.
Researchers recorded 4,975 bird fatalities in the studies they examined. Small passerines accounted for 62.5% of the recovered birds, while diurnal raptors accounted for 7.8%.
After adjusting and extrapolating the results to installed wind capacity at the time, the researchers estimated approximately 214,000 to 368,000 bird deaths annually from turbine collisions across the United States and Canada.
Another major US analysis cited within that work produced a central estimate of about 234,000 deaths a year, within a range of 140,000 to 328,000.
Those figures date from an earlier period of North American wind deployment and should not be treated as a current 2026 mortality estimate.
They do, however, show why a simple headline total can obscure the conservation issue.
In the North American analysis, the estimated annual collision loss for individual small-passerine species was less than one-tenth of 1% of each species’ continental population.
A similar number of deaths among a small population of long-lived, slowly reproducing raptors can have a very different effect.
That is precisely why the Flow Country report concentrates on population consequences rather than simply asking how many birds might collide with turbines.
Some collision risks can be reduced
Research also suggests that turbine design and mitigation can alter mortality.
A study at the Smøla wind farm in Norway tested whether making turbine blades more visible could reduce collisions.
Researchers painted one of the three rotor blades black on four turbines and compared mortality with neighbouring turbines. They reported a reduction of more than 70% in annual bird fatalities at the treated turbines.
The experiment involved only four treated turbines and its authors called for replication elsewhere before the finding could be applied broadly.
It nevertheless demonstrated that turbine-related mortality is not necessarily fixed.
NatureScot is now proposing further work of its own, including reviewing collision-avoidance assumptions, using post-construction monitoring and exploring collaborative monitoring between wind farms in northern Scotland.
Wind development around the Flow Country
The research was commissioned because individual planning applications do not necessarily reveal what happens when many developments occupy the same landscape.
The turbine dataset used by the researchers was last updated in December 2022.
At that point, Natural Heritage Zone 5 contained 256 constructed turbines, with another 126 approved, 82 under construction and 123 in planning.
Of the 256 already constructed, 205 were within five kilometres of what was then the proposed World Heritage Site.
Those numbers are historical inputs to the study and are not a count of the current number of turbines in the area in 2026.
The Flow Country was subsequently inscribed as a UNESCO World Heritage Site in 2024. It covers 187,026 hectares of Caithness and Sutherland and is recognised for its blanket-bog ecosystem.
UNESCO specifically identifies wind turbines as a potential threat, both because of their supporting infrastructure and because of possible effects on birds which form part of the ecosystem for which the site was designated.
Its list of potential pressures also includes forestry, drainage, intensive agriculture, burning, deer management and climate change.
The new research does not conclude that wind farms cannot be developed around the Flow Country.
Nor does it demonstrate that hen harriers or golden eagles will decline by the percentages produced by its models.
It reveals a more difficult problem.
The study produces potentially serious population effects from cumulative turbine collisions. NatureScot believes those estimates are too high. Its own records confirm that protected raptors have been killed by turbines, while simultaneously showing that Scotland has no complete national record of how often that happens.
International studies show why neither side of that argument can be reduced to a single number. Collision rates can vary from virtually none at one wind farm to many times the average at another, and the ecological consequence depends heavily upon which species are being lost.
For the Flow Country, that leaves the central question unresolved.
Scotland knows turbines have killed hen harriers and eagles. It knows the number recorded is incomplete. It also has a scientific model suggesting cumulative losses could eventually alter some populations… and a national nature agency which believes that model overstates the danger.
The next test will be whether Scotland can gather enough real-world monitoring data to determine which is closer to the truth.
Sources
NatureScot — Research Report 1350: Cumulative collision impact modelling on bird species of Caithness and Sutherland Peatlands Special Protection Area
https://www.nature.scot/doc/naturescot-research-report-1350-cumulative-collision-impact-modelling-bird-species-caithness-and
NatureScot — Freedom of Information Request: Deaths of Birds of Prey in Scotland from windfarms
https://www.nature.scot/doc/freedom-information-request-deaths-birds-prey-scotland-windfarms
UNESCO World Heritage Centre — The Flow Country
https://whc.unesco.org/en/list/1722/
Zimmerling, Pomeroy, d’Entremont and Francis — Canadian Estimate of Bird Mortality Due to Collisions and Direct Habitat Loss Associated with Wind Turbine Developments, Avian Conservation and Ecology
https://doi.org/10.5751/ACE-00609-080210
US Department of Energy / Pacific Northwest National Laboratory Tethys — Canadian Estimate of Bird Mortality Due to Collisions and Direct Habitat Loss Associated with Wind Turbine Developments
https://tethys.pnnl.gov/publications/canadian-estimate-bird-mortality-due-collisions-direct-habitat-loss-associated-wind
Erickson et al. — A Comprehensive Analysis of Small-Passerine Fatalities from Collision with Turbines at Wind Energy Facilities, PLOS ONE
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0107491
Loss, Will and Marra — Estimates of bird collision mortality at wind facilities in the contiguous United States, Biological Conservation
https://doi.org/10.1016/j.biocon.2013.10.007
May et al. — Paint it black: Efficacy of increased wind turbine rotor blade visibility to reduce avian fatalities, Ecology and Evolution
https://doi.org/10.1002/ece3.6592
