Chapter 2: What Do We Know about Impacts & Risk Factors to Wildlife and Habitat?

How do we measure collision impacts risk for birds and bats at wind energy facilities?

Conducting structured searches for bird and bat carcasses beneath operational wind turbines is our main source of information about collision impacts. These post-construction monitoring, or “PCM,” studies are designed to estimate bird and bat fatality rates for a given wind project; collectively, PCM study results have provided a basis for estimating collision fatalities across all wind energy projects and have helped to shed light on patterns of collision risk and risk factors for bird and bat species.

PCM study protocols

In PCM studies, investigators search for bird and bat carcasses at a sample of wind turbines, in order to estimate collision rates for a project. The protocol – that is, objectives, design, and methodology – used for PCM studies includes the number of turbines to monitor, the size and shape of plots to be searched and how searchers cover them, the interval of time between searches, and type of searcher—canine vs human. Topography and types of land cover that may affect searcher access and the visibility of carcasses, seasonal activity, and landcover found within a project area should all be considered when designing PCM protocols. For each carcass observed, searchers typically record the date, species, turbine being searched and the distance of the carcass to that turbine, GPS coordinates, vegetation cover, and condition of the carcass. Some PCM protocols include searches of roads and pads around wind turbines or incorporate carcasses found by project operations staff outside of scheduled searches.

Studies also include searcher efficiency and carcass persistence trials to estimate the number of fatalities that may not be detected during scheduled fatality searches.

See the U.S. Fish & Wildlife Service’s Land-based Wind Energy Guidelines, pp. 34-39, for a more detailed description of the questions to be answered and the methods and metrics to be used in PCM (Tier 4a) studies.

Searcher efficiency and carcass persistence trials

When we survey for plants and animals in the field, there is always a certain number that will be missed, either because they are hidden or because humans are naturally imperfect. In PCM studies, searchers may miss a carcass within the search plot or a carcass may have been removed from a search plot from a scavenger or due to decay before it can be discovered. The rate at which a carcass remains on the landscape is known as “carcass persistence”; the success rate by which searchers are able to find carcasses that remain on the landscape is known as “searcher efficiency”. Estimating these rates is an important part of estimating collision rates from fatality searches. Investigators often estimate these rates by conducting trials in which decoys or surrogate carcasses are placed in known locations. Because eagle and raptor carcasses are difficult to obtain, investigators often use surrogates, such as pheasants or other game birds for carcass persistence trials. However, studies are showing that these surrogates often have shorter persistence times than raptor carcasses, which may lead to study fatality estimates that are higher than the true fatality rate occurring at a project. Some studies propose that median carcass persistence rates from a region can be used in place of conducting a site-specific carcass persistence estimate.

Technologies that may supplement or replace fatality searches

Various technologies are being developed to automate fatality monitoring and wind energy projects using cameras or strike detectors. These are intended to supplement – or in some cases replace – fatality searches. A study is currently underway to validate a system that uses infrared cameras to detect bat fatalities. Microphones and accelerometers mounted on turbine blades may be used as strike detectors that capture collision events in real time. Sensor readings would need to be paired with carcass searches to identify the species and any demographic data associated with the fatality. Systems that pair cameras with sensor readings can identify species of larger bodied birds, in some cases, but are unlikely to be able to provide information on species or age for bats and smaller birds. Using technologies as they are improved and verified may increase our understanding of the factors that influence collision risk for birds and bats and may additionally lower cost and effort associated with traditional PCM. Strike detection technologies have the added potential to expand fatality monitoring in the offshore environment, where traditional PCM methods are not possible.

Reporting PCM results

PCM results are typically reported as adjusted fatality rates per project and per megawatt (MW) of project nameplate capacity (maximum rated electricity output) per year. Raw counts from carcass searches must be adjusted to account for fatalities that occur outside the study period and carcasses that land outside the search area, as well as for two other sources of imperfect detection: scavenger removal of carcasses (carcass persistence) and carcasses missed by searchers (searcher efficiency) (see above). Learn more about how raw carcass counts are adjusted to estimate actual fatality rates in Allison et al. 2019, p. 6.

Are PCM study results available to the public?

PCM study reports from over 100 wind energy projects are publicly available, but many more PCM results are confidential and have been unavailable for analysis. The American Wind Wildlife Information Center (AWWIC) encourages voluntary data contributions from wind energy projects across the U.S. by maintaining data confidentially, thus making more data available for analysis. AWWIC reports summarize bird and bat fatality rates (bird or bat fatalities per MW per year) and fatality incident data (individual fatalities) from wind energy facilities, including data from projects in regions with few published PCM studies.

AWWIC reports are updated regularly to provide the most up-to-date summary of fatality rates at hundreds of projects, and of species composition and timing of collisions to support the development of hypotheses that can be tested with additional analysis. It is important to note that these data do not represent a comprehensive or randomized dataset, and conclusions or extrapolations made from these data may change as additional data are added.

What do we know about collision impacts to birds and bats?

For birds, 75% of studies in AWWIC reported 2.3 or fewer fatalities per MW per year, with a median fatality estimate of 1.3 birds per MW per year. (The median is reported instead of the mean because of the right-skewed distribution of fatality estimates.) Overall, bird fatality rates ranged from 0 to 19 fatalities per MW per year.

Adjusted bat fatality rates tend to be higher and more variable than bird fatality rates. 75% of studies in AWWIC reported estimates of fewer than 7.7 bat fatalities per MW per year, with a median of 3 bats per MW per year. Some projects along the forested ridgelines of the central Appalachians report rates close to 50 bats per MW per year.

Which species are most impacted by collision with turbines?

• The majority of bird fatalities are small passerines. Studies in AWWIC reported 307 species of birds discovered during PCM studies and an additional 13 that were found incidentally. Raw counts of small passerines (songbirds, which make up nearly 90% of all land birds) account for approximately 57% of fatalities reported in both publicly available and private studies conducted at U.S. wind facilities. However, current turbine-related fatalities constitute a very small percentage (typically <0.02%) of their total population size; the current level of mortality is unlikely to contribute to population declines in most bird species.
• Fatalities of diurnal raptors are reported more often than expected given the relatively low abundance of these species. Diurnal raptors account for approximately 7% of reported fatalities. Red-tailed hawk and American kestrel are the most commonly reported fatalities; they are also the two most abundant diurnal raptors in the U.S. and have carcasses that tend to persist longer than other species and therefore may be more likely to be detected in searches.
• Reported fatalities of other large bird species are very low. Collision fatalities of prairie grouse and waterbirds and waterfowl (such as ducks, gulls and terns, shorebirds, loons and grebes) are reported relatively infrequently at land-based wind facilities.
• The majority of bat fatalities are migratory tree-roosting bat species. At least 25 species of bats have been recorded as collision fatalities in North America, but approximately 70% of the reported fatalities at wind facilities for all North American regions combined are from three migratory tree-roosting species: hoary bat, eastern red bat, and silver-haired bat.
• Mexican free-tailed bat account for a significant percentage of bat fatalities over most of the southern U.S. One of the most abundant bat species in the U.S., Mexican free-tailed bats make up 41-86% of the estimated number of bats killed at wind facilities within this species’ range.

Do collision rates vary by region and season?

• Bat fatality rates appear to vary substantially among regions in the U.S. Adjusted fatality rates of bats are highest at wind energy facilities in the upper Midwest and eastern forests and tend be much lower throughout the Great Plains and western U.S.
• Passerine fatalities are more common during migration seasons. Modest peaks in fatalities of small passerines occur during spring and fall at most wind facilities, presumably reflecting the passage of migrating species during these times.
• Bat fatalities peak in the northern U.S. during late summer and early fall. Several studies in the northern U.S. have shown a peak in the incidence of bat fatalities in late summer and early fall, coinciding with the migration season of tree bats.

What risk factors affect collision risk for birds and bats?

• Landscape and topographic features may influence the abundance and risk-related behavior of large raptors around turbines. Large raptors are known to take advantage of wind currents created by ridge tops, upwind sides of slopes, and canyons that are favorable for local and migratory movements.
• It is unknown if collision risk increases with turbine height and rotor-swept area. The height and rotor-swept area of turbines has been increasing. It has been hypothesized that collision fatalities will also increase due to the greater overlap of taller turbines with flight heights of migrating songbirds and bats. Further, a larger rotor-swept area could increase the collision risk zone. A study using PCM data from two sites in northern California from 1998-2007 found that replacing older, smaller turbines, with fewer, larger turbines decreased mean annual fatality rates . Conversely, another study in southern California, which conducted high effort PCM from 2018-2019, found fatality rates to increase proportionally with the rated capacity of turbines (in MW). A study is underway to explore the relationship between turbine height, rotor-swept area, and collision risk for both birds and bats.
• Turbine lighting plays an important role in managing risk to birds but the effect on bats is less understood. Nocturnally migrating birds are attracted to some types of artificial lighting, which can alter their flight behavior and put them at higher risk of collision with nearby structures. The Federal Aviation Administration (FAA) regulates the lighting required on structures taller than 199 feet in height above ground level to ensure air traffic safety. The FAA currently recommends strobe or strobe-like lighting that produces momentary flashes, interspersed with 3-second dark periods. Wind facilities must only light one in five turbines, firing all lights synchronously. The number of bat and songbird fatalities at turbines using FAA-approved lighting is not greater than that recorded at unlit turbines. FAA-approved, approach lighting systems (ALS) can further minimize risk from artificial lights—these systems only light up when they detect an approaching plane. A study published in 2014 found more red bat fatalities at unlit turbines than at turbines lit with red aviation lights. Conversely, a study published in 2022 suggested that bats were attracted to turbines that were lit up by ultraviolet light.

What causes bat fatalities at wind energy facilities?

• Low wind speeds and warmer temperatures appear to be risk factors for bats. Bat activity is influenced by nightly wind speed and temperature, and some studies indicate that bat fatalities occur primarily on nights with low wind speeds. Other weather-related variables such as temperature, wind direction, or changing barometric pressure may also be important. Additional research on weather as a predictor of bat activity and fatalities could support targeted mitigation efforts to reduce bat fatalities
• Some bat species may be attracted to wind turbines. It has been hypothesized that the relatively high number of bat fatalities that have been observed for some species and locations may be explained by attraction to wind turbines or wind facilities. Several factors that might attract bats have been proposed, including the sounds produced by turbines, a concentration of insects near turbines, and bat mating behavior. However, definitive tests of these hypotheses are still needed.
• Bat abundance during pre-construction surveys remains a poor predictor of collision fatalities during operation for these animals. The use of radar and bat acoustic detectors is a common feature of pre-construction risk assessments for siting wind energy facilities. To date, however, studies have not found a relationship between pre-construction activity surveys and post-construction collision risk, potentially because many bat species may be attracted to the area after construction and operation begins because of the presence of turbines.
• Landscape features may predict collision risk. In a study analyzing AWWIC bat fatality data from wind projects in the Midwest and Northeast, fatalities were higher at facilities that had a higher percentage of developed land within 25 km. Fatalities also appeared to be associated with wetland configuration, with increased risk in areas having multiple small or a combination of small and large wetland patches, suggesting that broadscale landscape features may be better predictors of fatality risk for migratory tree-roosting bats.
• Barotrauma does not appear to be an important source of bat mortality at wind energy facilities. Forensic examination of bat carcasses found at wind energy facilities suggests that the importance of barotrauma, that is, injury resulting from rapidly altered air pressure caused by fast-moving wind turbine blades, is substantially less than originally suggested. A 2020 study comparing the pressure changes experienced by bats near a turbine blade with the pressure changes that cause barotrauma mortality in mammals similar in size to bats (rats and mice) concluded that it is unlikely that barotrauma is responsible for a significant number of turbine-related fatalities.

For more detailed information about collision impacts to wildlife from wind energy, view AWWI’s summary of wind turbine interactions with wildlife and their habitats.