PAST AND ONGOING RESEARCH
This section of the meeting consisted of six presentations, with discussion of those presentations. (1) S.A. Gauthreaux, Jr., summarized studies in the United States, excluding some ongoing work that is described in the next two presentations. (2) T.J. Cade summarized work organized by the Kenetech Avian Task Force. (3) R.W. Thresher summarized work presently being funded (or soon to be funded) by the U.S. Federal Wind Energy Program . (4) S. Byrne summarized utility experience in designing and conducting research on the bird-power line interaction problem in the U.S.A., and lessons that this may provide for research on bird-wind power interactions. (5) J.E. Winkelman summarized work on bird-wind power issues in Europe, emphasizing the extensive studies done in the Netherlands. (6) R. Martí of the Sociedad Española de Ornitologia summarized ongoing research on bird-wind plant interactions in southern Spain.
The following five subsections are based on the extended abstracts for the five prepared reviews, augmented by points made during the oral presentations and subsequent discussion. The sixth subsection summarizes the presentation of R. Martí.
The History of Wind-Related Avian Research in the U.S.A.
by
Sidney A. Gauthreaux, Jr., Clemson University
This presentation summarized the major studies already completed, and mentioned the main ongoing pre-construction studies. Ongoing work organized under the aegis of Kenetech Windpower and the Federal Wind Energy Program is not described here, but is summarized in the next two presentations (p.
*, *). Points raised in the discussion of this presentation have been incorporated into the relevant subsections rather than at the end. ERDA/NASA 100 kW Experimental Wind Turbine, OhioStudies of the potential impact of wind turbines on birds in the U.S.A. began in the middle 1970s when the potential for environmental effects from the development of electrical generating capacity using wind energy was assessed at the ERDA/NASA 100 kW Experimental Wind Turbine at NASA's Lewis Research Center Plum Brook Station near Sandusky, Ohio (Rogers et al. 1977). The avian component of this study concentrated on assessing the potential for nocturnal migrant collisions with a wind turbine, because this was the only impact considered significant enough to warrant field studies. (It was assumed that birds would see and avoid the turbine in the daytime.) Observation methods for nighttime migratory movements included (a) making vertical ceilometer beam observations with 10x50 binoculars and an image intensifier scope (5x), and (b) monitoring migration over the general area using ARSR-2 radar units. Daytime studies included grounded migrant surveys and searches for dead and wounded birds at the base of the wind turbine generator and the meteorological tower. During four migration seasons only two birds were found dead at the meteorological tower and one bird at the wind turbine generator. It was concluded that the wind turbine was not proven to be a high risk to birds, including nocturnal migrants.
Boeing/PG&E MOD-2 Wind Turbine, CaliforniaApproximately five years later, bird movements and collision mortality were studied as part of Pacific Gas and Electric Company's performance monitoring program for a 350 ft Boeing MOD-2 wind turbine and associated meteorological tower located on the western edge of the Suisun Marsh just south of Cordelia, Solano County, California (Byrne 1983, 1985). Both raptor and waterfowl movements were monitored prior to construction. Nocturnal migration over the site was monitored during the fall of 1982 and the spring of 1983 using a portable ceilometer and image intensifier system. Dead bird searches were conducted five days a week during nocturnal migration monitoring and once a week thereafter. Weather data were gathered to examine relationships between migration intensity and weather conditions. Findings indicated relatively low rates of waterfowl movements and nocturnal songbird migration over the wind turbine site. Raptor activity in the area was moderate to high. Migration rates were considerable lower than those recorded in the eastern United States. During the year of mortality monitoring (1 Sep 1982—31 Aug 1993), seven dead birds were found: one was observed to have collided with the rotor (an American Kestrel), four more were thought to have collided with the wind turbine, and two were thought to have collided with guy lines or the meteorological tower. The mortality adjusted for scavenger removal and detectability suggests an actual mortality during the study as high as 54 birds.
California Energy Commission StudiesIn 1989 the California Energy Commission Report, "Avian mortality at large wind energy facilities in California: Identification of a problem" was released (CEC 1989). The report reviewed existing data on bird injuries and mortality caused by wind turbines in 1984-1988. Nearly all incidents involved raptors, including 108 individuals of seven species (72 collisions and 36 electrocutions). Raptors are protected by both California and federal laws. These collisions were in the Altamont and Tehachapi areas. Most of the reported mortality was during winter, not during migration.
This report stimulated a two-year study conducted by BioSystems Analysis, Inc., to evaluate the extent and significance of the impact of wind turbines on birds, to identify the causes and factors contributing to bird injuries and deaths, and to recommend mitigation measures (Orloff and Flannery 1992). The study areas included an established wind turbine area at Altamont Pass, Calif., and another still under development in Solano County, Calif. Observations and dead bird searches were conducted for six seasons. Of 183 dead birds found, 119 (65%) were raptors and only 19 carcasses were fresh. Most dead raptors were Red-tailed Hawks (36%), followed by American Kestrels (13%) and Golden Eagles (11%). Fifty-five percent of the mortality was attributed to collisions with turbines, 11% to collisions with wires, 8% to electrocutions, and 26% unknown. During the study no birds were observed flying into wind turbines. The investigators concluded that, of the potential factors contributing to mortality, the following were most important: end-row turbines, turbines within 500 m of a canyon, the elevation of a turbine, and a lattice type turbine tower. The statistical robustness of these conclusions has been the subject of much subsequent discussion.
A follow-up study, also by BioSystems Analysis, Inc., is presently underway. This will include further mortality searches in the Altamont, and additional analysis of the existing Altamont mortality data in relation to turbine characteristics.
Ongoing Pre-Construction SurveysCurrently, several studies are underway or planned to assess the potential impact of wind farm development on bird injury and mortality. Nearly all of these studies involve pre-construction monitoring of bird movements, including local movements of resident species and migration traffic rates of transient species. These studies are being done in the vicinity of potential wind farm developments in order to identify the species that have the greatest risks of negative impact. Many of the studies are using direct visual observations during the day and image intensification and radar observations at night to quantify the amount of movement, height of flight, and flight behavior of local and migratory species. These studies are underway in the Kibby Range of Maine, eastern Lake Ontario/Tug Hill region of New York, and the Fort Davis area of the Trans-Pecos of Texas. Similar studies are planned for the Searsburg and Readsboro areas of Vermont, eastern Oregon, and the Norris Hill Wind Resource Area of Montana.
The opportunity to acquire valuable pre-construction data is great, but to date the methodologies applied in different studies have not been standardized. If this were done, meaningful comparisons of data would be possible. Likewise, if post-construction studies use standardized methods of monitoring bird movements and measuring mortality, a reliable data base can be accumulated in a relatively short period of time on a continental scale. These data will be very useful in determining the geographical variability in avian mortality and/or species and numbers of birds that may be at risk.
Industry Research: Kenetech Windpower
by
Tom J. Cade, Chairman
Kenetech Avian Research Task Force
With growing concern that wind power plants present certain hazards to birds and cause some injuries and fatalities, in 1992 Kenetech Windpower assembled a group of outside experts to review information about avian collisions at the Altamont and Solano wind farms, where strong concerns exist about accidents involving raptors. This group is called the Kenetech Avian Research Task Force. It consists of five academic scientists with interests in raptor ecology and behavior, aerodynamics of avian flight, flight behavior and orientation, migration, and avian sensory physiology. The functions of the Task Force are to recommend and oversee specific areas of research needed to understand the interactions between birds and wind turbines, and to use the information (a) to help design ways to reduce the frequency of bird collisions, and (b) to assess the biological significance of wind farm-related fatalities. While trying to maintain the scientific rigor of these studies, the Task Force is aware that its recommendations for research have to be implemented within a sociopolitical arena in which the wind power industry is under regulatory constraint to "solve the problem" of individual bird fatalities, regardless of "whether raptor populations are significantly affected by turbine-related mortality".
Research InitiatedThe Task Force feels that it is important to acquire precise quantitative data in sufficient quantities to allow reliable assessments and conclusions. After reviewing the existing reports on avian activities and fatalities in the Altamont and Solano plants, the Task Force was impressed by the fact that, although systematic searching turned up numbers of dead birds on the ground over the course of a year, the rate of accidents per turbine was so low that direct human observations of collisions were impractical for determining how and why birds get into trouble.
One approach to this problem has been to focus on controlled flights of homing pigeons in and amongst the turbines as a way to obtain quantifiable data on the behavior of birds during close encounters with wind turbines. Pigeons are not raptors, but the Task Force believes that much can be learned about general bird/turbine interactions from careful tests with pigeons. Pigeons released near strings of turbines during the daytime clearly recognize the turbines and, when necessary, adjust their flight to avoid them. Of about 2,270 study flights near turbines to date, three pigeons have collided with the turbines.
Another approach has been the development of automatic, machine-recording systems such as video monitoring. However, review of tapes is very time consuming, and the resolution of standard video systems is a limiting factor. "Smart-camera" equipment that records only when a bird is detected may prove valuable, but its usefulness in this application is unproven.
Because visual observation and even standard video-recording do not provide an accurate location of a flying bird relative to other objects in three-dimensional space, considerable effort has been directed to the development of a tracking system that will automatically and accurately record the position of a flying bird as it moves through space near turbines. Such a system would allow researchers to measure and understand avoidance behavior or failure to avoid under a variety of conditions, and to demonstrate behavioral changes associated with modifications of turbines. A system employing two simultaneously recording video cameras has been field-tested and shows promise for yielding the needed information.
In order to modify turbines in ways to make them more avoidable by birds, it is necessary to know how birds perceive their world. In particular, we need to know their visual and auditory capabilities. To that end, the Task Force recommended a program of research, now under way at Boise State University, on the visual and acoustic capacities of raptors. American Kestrels and Red-tailed Hawks are being trained for use in these tests. Study to date on vision indicates that contrast, color, and rotation are the three most important variables influencing a raptor's detection of turbine blades.
Contrast between the blades and their background is most important. Thus, an effective warning pattern should maximize pattern contrast against environmental background, especially since birds, including raptors, have poorer ability than humans to resolve spatial frequencies at low contrast. It is hoped to begin field tests this year.
Early observations in the Altamont indicated that raptors perch on turbine towers and even on the blades when they are not rotating. Recent study shows that this behavior is more common than previously supposed, and that there may be a tendency for raptors to perch more often on end-of-row turbines, where fatalities are also indicated to be higher (Orloff and Flannery 1992). The possibility of a significant association between perching and collisions requires detailed investigations, which are under way.
It is desirable, from the standpoint of public perception and ethical considerations, to reduce bird collisions as much as practicable, not to mention the legal requirements to do so. However, the Task Force takes the view that some level of mortality associated with wind plant operations is acceptable, so long as it does not influence the long-term population viability of any species negatively. How fatalities at wind plants fit into the overall balance between natality and mortality of avian populations is the main biological question that needs to be addressed.
In the Altamont region, the Golden Eagle is the species of first concern. Numbers in the area are unusually high, and some are killed by collisions with wind turbines. The Task Force recommended in 1993 that a long-term study of eagle population dynamics be carried out in the region to determine the impact, if any, of eagle fatalities in the wind farms on the associated breeding population of Golden Eagles. A one-year pilot study by the Santa Cruz Predatory Bird Research Group has been funded by NREL and Kenetech to obtain first estimates of relevant population parameters and to test study methods and identify needed resources for a multi-year project. Since January 1994, 31 adult and immature eagles have been trapped in the Altamont and equipped with radio-transmitters, and 25 nestling eagles from the surrounding Diablo Mountains have also been radio-tagged. Thirty-one active nests and 50 resident pairs have been located in the Diablo Mountain/Altamont region.
Recommended General ApproachAfter reviewing avian-wind power problems for two years, the Kenetech Avian Task Force feels strongly that adequate management of bird collisions at the wind plants must be approached at four levels. (1) Initial plans for siting wind farms must take into consideration the entire annual cycle and pattern of avian use of the proposed area. If the area proves to be one of high use and dense concentration for birds, then alternative sites should be sought. (2) The size and physical configuration of the wind plant, spacing of turbines, position of turbine rows, etc., need to be evaluated with respect to the kinds of birds and their activities in the area. (3) The structure of turbines and turbine towers should be designed to reduce collisions by reducing perching opportunities to a minimum. In addition, turbine blades should be patterned to maximize their visibility to birds under as wide a range of conditions as possible; exactly how to accomplish this remains to be worked out. (4) Where unpreventable fatalities may continue to occur, off-site mitigation can be helpful. For example, if Golden Eagles foraging in the Altamont continue to contribute to an increase to population-wide mortality, off-site mitigation to insure the long-term integrity of eagle nesting territories in the Diablo Mountains could do more to perpetuate population viability of eagles in the region than spending millions of dollars in efforts to eliminate all eagle fatalities on the wind farms.
DiscussionHoming Pigeon Releases.—Individual pigeons are released repeatedly, and presumably gain experience with wind turbines over time. It was noted that the relevance of the pigeon experiments has been the subject of much discussion.
Raptor Behavior and Mortality in Wind Plants.—Is there information on whether soaring birds take advantage of local updrafts caused by the turbines themselves? Answer: No, and it would be difficult to distinguish micrometeorological effects of turbines from those of the often-hilly terrain on which the turbines are situated.
What is the distribution of mortality relative to lattice vs. tubular towers? There have been suggestions that death rates of raptors in the Altamont are higher at turbines with lattice towers. However, interpretation is confounded by various factors, including the non-random distribution of turbine and tower types in relation to topography. The CEC report notes that causal links could not be isolated (Orloff and Flannery 1992, p. 3-80ff, 4-8 and 4-13). They note (p. 4-13) that "Any causal agent must be statistically associated with mortality, but the association of a variable with mortality does not necessarily imply causation. We do not imply that any of the variables we found to be associated with mortality were actually causing bird deaths." Some meeting attendees commented that observations such as this should be treated as hypotheses that could be investigated by controlled tests.
Are birds found dead in the wind plant being autopsied to determine cause of death, and the possible contribution of debilitating factors such as lead poisoning? To some degree, but there are complications associated with the treatment of bird bodies as evidence for potential prosecutions. This is an example of a situation in which policy/legal questions have direct effects on research.
Golden Eagle Population Study.—What are the objectives? See following presentation (p.
*). One important objective is to determine whether the eagles killed in the wind plant are breeding birds, and whether the mortality is affecting recruitment into the breeding population.No "control" Golden Eagle population is being studied as part of the present pilot study, but Golden Eagles have been studied elsewhere.
The Pacheco Pass area south of the Altamont has topography similar to that of the Altamont, and only a few wind turbines. No systematic study of birds has been done in that potential "control" area, but it seems similarly suitable for eagles.
Approach.—Regarding the suggestion that off-site mitigation could be more effective than costly efforts to eliminate all eagle fatalities on the wind farms, it was noted that this concept is not unanimously accepted from either an ethical or a legal perspective. Likewise, the view that some individual bird deaths are acceptable (legally or otherwise), provided that bird populations are unaffected, is controversial.
U.S. Federal Wind Energy Program Avian Research Projects
by
Robert W. Thresher, NREL
The two-year study by BioSystems Analysis Inc. to examine the impact of wind turbines on birds in the Altamont Pass and Solano County wind resource areas has focused attention on avian mortality related to wind turbines. The BioSystems study concluded that there was significant Golden Eagle mortality due to collisions with wind turbines. However, the impact on the population of Golden Eagles normally using the Altamont Wind Resource Area was not addressed. NREL identified a population study of the Golden Eagles using this wind resource area as fundamental to understanding the significance of the observed deaths.
A pilot study of the Golden Eagles in the Altamont Pass wind resource areas, and particularly on their population dynamics, has been initiated, as noted in the preceding presentation (p.
*). The University of California at Santa Cruz is carrying out this pilot study. The pilot study will develop methods to obtain preliminary estimates for the following population data:Determine first estimates of the population of Golden Eagles using the Altamont Wind Resource Area.
Develop an estimate of the transient Golden Eagle using the area.
Identify what attracts Golden Eagles to the wind plants and what activities increase the risk of collision.
Estimate the yearly mortality rates caused by wind turbines and the impact on the overall population.
The primary focus of this study is to provide estimates for population data. In addition, the pilot study is to develop a comprehensive plan for surveying the Golden Eagle population in Altamont. The plan will recommend experimental methods and procedures, and provide a schedule and budget that will be based on the experience and data from the pilot study. The pilot study is to be completed in December 1994. In this way, proposed follow-on study plans can be based on real data and field experience, and the benefits of a more comprehensive study can be judged more accurately.
NREL regularly runs a solicitation to encourage university participation in the Federal Wind Energy Program. This past spring an avian research category was included in this solicitation as a high priority research topic. Boise State University's proposed research on avian perching and related mortality has been selected for negotiation of a research subcontract. The Statement of Work for this effort has not been completed at this time. The proposal outlined a series of experiments designed to determine the influence of perching on raptor mortality and to test anti-perching devices to determine their effectiveness. The research effort is planned for three years beginning this fall, but the "perching hypothesis" will be tested first and follow-on efforts adjusted accordingly. The wind industry partner for this proposed effort is Kenetech Windpower, and field experiments will take place in Altamont Pass.
The Utility Wind Turbine Verification Program sponsored by DOE and EPRI (p.
*) provides cost sharing for the deployment of small 6 MW wind power plants using the latest technology. Under this program, pre-construction environmental impact studies are being carried out to assess the potential impact on birds. In addition, several other high potential wind farm development areas are planning pre-construction avian monitoring studies, and the Federal Wind Program is considering support of these efforts.Lessons from Utility Structure Environmental Impacts
by
Sheila Byrne, PG&E
The first published account of birds colliding with overhead wires dates from the 1870s, shortly after the first telegraph wires were stretched across the prairies (Avery et al. 1980). Incidents were reported sporadically through the next century. Reports and studies of bird collisions with powerlines have greatly increased, starting with the early 1970s and the passage of the National Environmental Policy Act and the Endangered Species Act. Other impacts of utility structures on birds have similarly been documented. Birds collide with tall stacks and cooling towers. They are electrocuted on distribution and transmission lines, and rights-of-way modify their habitats. Birds, in turn, impact utility operations. They perch and nest on utility structures, shorting lines and fouling insulators and substations. Electrocutions often result in outages, and can cause fires.
These problems have been examined extensively for nearly 25 years. Several national meetings have been held on bird/utility issues. The most recent was in December 1993, in Miami, sponsored by the Electric Power Research Institute and the Avian Powerline Interaction Committee (APLIC) (EPRI 1993).
The paper is based on utility experience in two areas: biological and procedural. Particular biological lessons may or may not be applicable to wind turbine developments. However, I think that the procedural lessons are all applicable.
Biological Generalizations |
If you put it in the air, sooner or later a bird will fly into it. |
||||
|
Collision vulnerability varies with species, age, sex, habitat use, weather, human disturbance, and especially location. Raptors and gulls seem less vulnerable than some other types of birds to collisions with power lines. |
||||
|
Not all dead birds found under a line were killed by colliding with it, and it is often difficult or impossible to determine the specific cause of death. |
||||
|
Although some problems can be solved solely by better engineering, knowledge of bird behavior is critical to finding quick and cost-effective solutions. For example,
|
||||
|
The significance of mortality depends upon the population affected. |
Procedural Generalizations
|
It is difficult to get good estimates of collision mortality. It is easier to do a "worst case analysis" whose errors and unsupported assumptions will be copied in all subsequent studies. Reasonable correction factors for scavenger removal and search bias can now be obtained. Estimates of the percentage of birds crippled or lost in inaccessible habitats are difficult to obtain, and easy to misuse. |
|
Always do necropsies. |
|
Developing solutions to impacts requires good scientific studies, which are time consuming and expensive. These studies are not easily accommodated in the permitting process, even though the lack of these studies results in even more delay and expense to projects. |
|
The perceived significance of mortality depends upon many non-biological factors. |
Avian Powerline Interaction Committee (APLIC)
Utilities have responded to the problems inherent in developing solutions to utility structure impacts on birds by forming a cooperative study group, the Avian Powerline Interaction Committee. APLIC has recently
|
tested powerline marking devices, and |
|
developed a manual for studying and mitigating collisions with powerlines,and is |
|
revising the manual on preventing raptor electrocutions. |
APLIC's recommendations are widely accepted by utilities and agencies. In the author's view, the following factors have contributed to APLIC's success:
|
it is a truly cooperative effort, involving not only agencies and utilities but also environmental groups and academic scientists, |
|
dedicated agency and utility biologists have nurtured it, often through periods of limited funding, and |
|
it has kept a limited focus, restricting its attention primarily to collisions and electrocutions. |
Other attendees who have been involved in APLIC reiterated the preceding comments on factors that have contributed to APLIC's success, and suggested several additional points:
|
APLIC has been very successful in developing methods to reduce avian mortality due to electrocutions and to some degree to powerline collisions. |
|
APLIC has sponsored conferences to disseminate these results. |
|
It has been recognized that one can reduce but not totally eliminate bird deaths caused by powerlines. Through openness and cooperation, stakeholders have accepted this as a worthwhile objective. |
|
APLIC provides interested parties with information about the available options for dealing with bird collision issues; decisions as to how to use this information are left to the responsible groups. |
|
The use of standardized study methodologies is important to allow meaningful comparisons of results from different studies. |
Bird/Wind Turbine Investigations in Europe
by
J.E. Winkelman, The Netherlands
This presentation gave an overview of research carried out in Europe, with special emphasis on the results of the two most detailed studies. (1) Oosterbierum wind park, Netherlands, with 18 middle-sized (300 kW) turbines in cluster formation on 55 ha of arable land close to the Wadden Sea (Winkelman 1992a-d). (2) Urk wind park, Netherlands, with 25 middle-sized (300 kW) turbines in line formation along a 3-km dike bordering lake IJsselmeer, a major wintering area for ducks (Winkelman 1989). Most results of other European studies are consistent with those from these two areas. The main exception is the recent work in southern Spain (p. *), where bird mortality (mainly of raptors) has been more evident than in the Netherlands.
A more detailed version of this presentation, including Tables and a Bibliography of European bird-wind power research, appears as Appendix 2B (p. *). English-language summaries of the Winkelman (1989, 1992a-d) reports on the Urk and Oosterbierum studies appear in Appendix 2C (p. *).
In Europe discussions about the possible impact of wind energy on birds started in the late seventies, when the first national wind energy strategies were formulated and the first (mostly small sized, solitary) wind turbines were erected. This was followed by a huge number of speculative articles in newspapers, magazines and popular scientific journals, nearly all of them focusing on the possible collision risks for birds. The first research results became available in Sweden, Denmark and The Netherlands in 1983 and 1984, again followed by many articles and reports on possible effects, pre-construction studies, progress reports, and overview studies.
To date, 14 studies have been finalized in Europe, covering 108 different sites with one or more wind turbines. These studies were in southern Sweden (2 studies, 2 sites), Denmark (3,18), northern Germany (1,10), Netherlands (6,85), and United Kingdom (2,3). Research is now underway or will start soon in the south of Spain, The Netherlands, and Denmark.
Most studies include small, solitary turbines (<100-150 kW). Wind parks, and especially middle-sized (250-500 kW) and large (MW) turbines, are less studied. Three general topics have been studied:
1. Collisions with rotor, tower, power lines; 2. Disturbance (usually without much habitat modification): Loss of habitat for breeding, feeding, migration; 3. Behavioral Changes: Flight behavior when approaching wind turbines.
Studies on bird collisions were mostly carried out by searches for dead birds. The proportion of birds colliding in relation to the total numbers passing the wind turbines was studied at 13 sites. Estimates of the total numbers of bird victims could only be made in three studies (3 sites), as the other studies did not take into account the search efficiency, predation pressure, number of days with searches, causes of death of the corpses found, and/or total areas searched. These factors may all have strong effects on total estimates. It was proven that, even with careful searching, the proportion of the bird bodies found could be low, especially for small birds in high vegetation.
At the 108 European study sites, a total of 303 dead birds were found, of which at least 124 (41%) were proven collision victims. In the Dutch Oosterbierum wind park, only 27% of all birds found were killed by collision. It is noteworthy that there were no nights with large kills, that very few collision victims were found near small wind turbines, and that almost all victims were of common species. Virtually none of the victims were scarce or rare species. Estimated average numbers of collision victims in the Oosterbierum and Urk wind parks, in birds per turbine/day, varied between 0.04 (Urk, autumn) and 0.09 (Oosterbierum, spring), depending on season and site. These figures were based on regular searches.
Based on nocturnal observations with a thermal image intensifier in the Oosterbierum wind park in autumn 1988, an estimated 170 birds collided with turbines during seven consecutive nights, or 0.051 dead birds/h/100 m front. This was equivalent to 2.5% of all birds passing at rotor height (20-50 m). In both the Oosterbierum and Urk studies, most bird victims were found after nights with both poor flight conditions and visibility. Mean numbers per kilometer of wind park are comparable to the numbers of birds killed by traffic per km of highway, and comparable to or somewhat lower than the numbers of victims per km of power line in risky situations. Total numbers likely to be killed per 1,000 MW of wind power capacity are low relative to other human-related causes of death.
In the Oosterbierum wind park, only a few birds were seen very close to a rotor during daylight. Of these, one (14%) was hit and killed. During the night, 20% of all birds crossing a rotor were killed. It was noteworthy that not all observed collisions were fatal, and that some "collisions" were caused by the wake behind the rotor. In the latter cases, birds that did not contact the rotor were sometimes swept down by the wake, and injured or killed as a result.
Disturbance and Habitat Loss
Several studies evaluated the effects of disturbance and habitat loss on numbers of birds present. Five of these studies concerned loss of habitat for breeding birds (mainly waders). Three studies concerned resting birds (several larger bird species), three concerned daytime migrants and two concerned nocturnal migrants—largely songbirds.
Habitat loss/disturbance effects were demonstrated at distances up to 250-500 m from the nearest turbines. The reduction in the numbers present in the disturbed zones ranged up to 95%. Some bird species were far more vulnerable than others, and vulnerability depended on site, season, tide, and whether or not the wind park was in operation. Breeding waders seemed less vulnerable than some other birds. However, those results may have been confounded by the high site fidelity and long life spans of waders, coupled with the fact that the studies were carried out for only one or a few breeding seasons. From a European nature conservation point of view, disturbance/habitat loss effects associated with wind plants are thought to be of much more importance than direct bird mortality due to collisions. However, the ongoing study in Spain (see p. *) may be an exception to this generalization.
Changes in flight behavior during migration were examined during seven studies of day-time migration involving 28 wind turbine sites, and during three studies of nocturnal migration at three sites. Aspects studied included numbers and types of reactions. These were mostly within 100 m of the nearest turbine during diurnal migration, and within 20 m of a rotor during nocturnal migration. Changes in flight paths were also studied. During diurnal migration these changes mostly occurred within 300-500 m.
During daylight, proportionally fewer of the migrating birds reacted when the turbines were not operating than when turbines were operating (2% vs. 11-18%). The frequency of reactions with turbines operating depending on the distance between the turbines, with reactions being more frequent when the turbines were 150 m apart than when they were 300 m apart. During the day-time, most reactions were calm and gradual, mainly consisting of horizontal shifts. Only a minority of the approaching birds needed more than one passing attempt before crossing the wind park.
In the Oosterbierum wind park only a very few birds were seen within 20 m of a rotor during daylight. Nocturnal migrants were more commonly seen within 20 m of the rotors. During the night, reactions of 47 birds within 20 m of a rotor were observed by means of a thermal image intensifier. Of these, 43% approached without hesitation. The proportion of the birds that reacted depending on the wind direction. Birds flying with headwinds were more likely to react, perhaps because they encounter the rotor wake before reaching the operating rotor. During the daytime, 15% of the closely-approaching birds altered their flight paths calmly to avoid crossing the rotor, whilst at night 36% did so. The other birds all crossed or tried to cross between the rotor blades. While doing so, most of them either flapped their wings powerfully or fluttered.
Migration Behavior and Collisions.—In Dr. Winkelman's opinion, direct collision mortality with wind turbines does not now constitute a significant biological problem for bird populations in most of Europe; Spain may be an exception. This conclusion was discussed at some length. It is not established whether this conclusion applies to wind plant situations in the United States, or that it would apply after larger scale wind development in either continent. Also, it was noted that bird deaths may be a legal, ethical or political problem whether or not they are significant in the context of bird population dynamics.
What will be the impact on birds of offshore wind plant developments planned for Europe? Initially, offshore wind plants will be near the shore. However, there are proposals for wind plants up to at least 10 km offshore, e.g. between Denmark and Sweden. Nothing specific is known about their effects on birds. Also, little is known about the effects of coastal wind plants on nocturnal migrants flying low along coastlines. (But see Buurma and van Gasteren 1989.)
Were field observers aware of the hypotheses being tested when they were categorizing behavioral reactions to the turbines? Yes, it was not possible to do this in a "blind" manner. In the Netherlands, it has proven to be very important to use observers who are highly experienced in visual observation of migration. Also, it should be noted that some key observations were based on videotape, where there is a permanent record of the event. The birds thrown to the ground by the turbine wake were documented on videotape.
Why is the kill rate so much higher in the Netherlands, e.g. 0.05 birds/turbine/night during migration, than in the Altamont? What are the implications of proposed wind developments in other parts of the United States with different species and numbers of xbirds, and with concentrated migration corridors? The amount of nocturnal migration at low altitude is probably considerably higher near the Netherlands coast than in the Altamont. During nights of high-density nocturnal migration, on the order of 4,000 birds/night pass over the Oosterbierum wind park at low altitudes (0-50 m above ground). Some parts of North America are also known to have higher rates of nocturnal migration than have been documented in California. However, in most areas little is known about the proportions of the nocturnal migrants that are at altitudes low enough to encounter wind turbines. In any case, the Altamont, with a very high density of Golden Eagles and other raptors, is not comparable to areas of Europe and North America with fewer raptors but concentrated nocturnal migration. Many participants believed that the types and numbers of birds that would collide with turbines in some parts of North America will differ widely from those documented in the Altamont.
In the U.S.A., collisions are so infrequent at any given turbine that it is impractical to observe them directly. In Europe, are collisions common enough to allow direct study? Even with the higher collision rate in the Netherlands, collisions are infrequent events. They can be observed directly, but these observations are very labor intensive. It took one year to analyze the thermal imaging recordings obtained on 17 nights. Automated techniques for video review were investigated in the Netherlands several years ago. At that time there did not seem to be any practical automated method for automated video review. New technology for video surveillance might now allow automated detection of birds. However, meeting participants were not aware of any case where this has been implemented for birds in a field situation.
What is the present state of knowledge about the effects of lighting on collision rates at night? Large numbers of nocturnal migrants occasionally collide with tall structures that are continuously lit, e.g. by floodlights, by interior lighting visible through windows, or by steady or slowly-flashing red lights. On these rare occasions with much mortality, there is usually fog or drizzle. On these nights, migrants are attracted to the lights; many individuals circle or hover near the lights. Many fewer nocturnal migrants collide with structures illuminated by flashing white strobe lights.
Disturbance and Habitat Loss.—What attributes of wind parks in the Netherlands cause reductions in bird densities in habitats near those wind parks? This is not known. Noise and movement are possibilities. However, the same disturbance effect has been noted at a noisy and a quieter wind park, suggesting that turbine noise may not be the main factor.
Why do disturbance effects extend several hundred meters from a wind park, when the literature suggests that, for birds, maximum disturbance distances around other "public access areas" are more typically about 50 m? This is not known. Because of the relatively large radius of effect in Europe, questions have been raised as to whether it is better Ø to have a few large wind parks rather than a large number of small ones, and Ø for turbines to be in line or cluster formation.
Was there any possibility that the disturbance effect was partly an artefact of observer bias? No, the effect was real. The same types of counts were done at varying distances from the turbines, on control as well as wind park areas, and before construction as well as during wind plant operations. The same observers were involved in all of this work.
Bird/Wind Turbine Investigations in Southern Spain
by
Ramón Martí, Spain
Ramón Marti noted that he was presenting preliminary results based on the first few months of an ongoing study. This work is being conducted by Luis Barrios and Enrique Aguilar, from Sociedad Española de Ornitologia (SEO/BirdLife Spain) under contract with the Agencia de Medio Ambiente de Andalucía (Andalusian Environmental Agency).
Wind development in Spain is now concentrated at Tarifa, at the southern tip of Spain near Gibraltar (Fig. 1). The average wind speed at Tarifa is 6.5-7.5 m/s, which is not as high as at some other locations in Spain. However, the Tarifa area is attractive for wind power development because useable winds blow on over 95% of the days of the year.
Tarifa, the Spanish edge of the Strait of Gibraltar area, forms one of the two main "bottle-necks" for concentrated bird migration in the Mediterranean basin. The other is the Bosphorus Strait around the eastern Mediterranean, between the Balkans and Turkey. A third route, less important compared with the other two, is across the Messina Strait, Italy (Fig. 2). Large numbers of migratory birds, including a high proportion of the soaring raptors and storks that nest in Western Europe and winter in Africa, migrate through the Tarifa/Gibraltar area. Finlayson (1992) provides a recent general review of bird movements and populations in the area. About 20 species of soaring birds totalling at least 300,000 individuals migrate through the area in autumn. The Honey Buzzard, Black Kite and White Stork are especially common, along with large numbers of non-soaring birds.
Because of the importance of this area for migratory birds, the area has been given international recognition as an "Important Bird Area" by the ICBP (International Council for Bird Preservation), now called BirdLife International. It has also been declared a "Special Protection Area" under European Union Directive 79/409 on conservation of wild birds. Furthermore, because of the migratory birds and other natural values, it has also been declared a Natural Park by the Andalusian Government.
The main migration axis for soaring birds in the Tarifa area during fall migration, when there are many more birds than during spring migration, is NNW-SSE under the dominant easterly-wind conditions (Fig. 3).PgNum+3 code here to allow for three diagrams. Several wind parks have been established within the area traversed by soaring birds on migration. Many of the turbine strings are aligned roughly parallel to the main NNW-SSE migration direction, but some strings cross that axis at an angle (Fig. 3). Large numbers (hundreds or thousands) of soaring birds sometimes land and roost on flat ground or promontories in the Tarifa area, including some locations with existing or proposed wind turbines, while waiting for weather conditions good for crossing the sea.
The SEO/BirdLife study now underway includes searches for dead birds near the turbines plus direct observations of the behaviour of soaring birds nearby. During the first few months of study, a number of birds killed by collisions with turbine blades have been found. These have raised concern at SEO/BirdLife Spain. The casualties include 14 protected species, but the majority were Griffon Vultures. They are large and relatively unmanoeuvrable birds that depend on slope-winds and thermals. This study, and another one on wind resources, will provide the basis for further rational development of wind energy in Tarifa area, if appropriate.
No studies of nocturnal migration in relation to wind power development are being done in the Tarifa area. In general, the region near the Strait of Gibraltar is known to be an important corridor for nocturnal migration of passerines and other species travelling between Western Europe and Africa, based on radar and (to a lesser degree) moon-watch studies in past years. This nocturnal migration, and the daytime migration of non-soaring species, occurs on a broader front than does the daytime migration of soaring raptors and storks.
