3
Ecological Effects of WindEnergy Development

CHAPTER OVERVIEW

At regional to global scales, the effects of wind energy on the environment often are considered to be positive, through the production of renewable energy and the potential displacement of mining activities, air pollution, and greenhouse gas emissions associated with nonrenewable energy sources (see Chapter 2). However, wind-energy facilities have been demonstrated to kill birds and bats and there is evidence that wind-energy development also can result in the loss of habitat for some species. To the extent that we understand how, when, and where wind-energy development most adversely affects organisms and their habitat, it will be possible to mitigate future impacts through careful siting decisions. In this chapter, we review the effects of wind-energy development on ecosystem structure and functioning, through direct effects of turbines on organisms, and on landscapes through alteration and displacement. We recommend a research and monitoring framework for reducing these impacts. Although the focus of our analysis is the Mid-Atlantic Highlands, we use all available information to assess general impacts. Although other sources of development on sites that are suitable for wind-energy development affect wildlife and their habitats (e.g., mineral extraction, cutting of timber), and there are other sources of anthropogenic mortality to animals, as stated previously, this committee was charged to focus on wind energy, and therefore did not conduct a comprehensive comparative analysis of impacts from other sources of development.

Wind turbines cause fatalities of birds and bats through collision, most



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 67
Environmental Impacts of Wind-Energy Projects 3 Ecological Effects of WindEnergy Development CHAPTER OVERVIEW At regional to global scales, the effects of wind energy on the environment often are considered to be positive, through the production of renewable energy and the potential displacement of mining activities, air pollution, and greenhouse gas emissions associated with nonrenewable energy sources (see Chapter 2). However, wind-energy facilities have been demonstrated to kill birds and bats and there is evidence that wind-energy development also can result in the loss of habitat for some species. To the extent that we understand how, when, and where wind-energy development most adversely affects organisms and their habitat, it will be possible to mitigate future impacts through careful siting decisions. In this chapter, we review the effects of wind-energy development on ecosystem structure and functioning, through direct effects of turbines on organisms, and on landscapes through alteration and displacement. We recommend a research and monitoring framework for reducing these impacts. Although the focus of our analysis is the Mid-Atlantic Highlands, we use all available information to assess general impacts. Although other sources of development on sites that are suitable for wind-energy development affect wildlife and their habitats (e.g., mineral extraction, cutting of timber), and there are other sources of anthropogenic mortality to animals, as stated previously, this committee was charged to focus on wind energy, and therefore did not conduct a comprehensive comparative analysis of impacts from other sources of development. Wind turbines cause fatalities of birds and bats through collision, most

OCR for page 67
Environmental Impacts of Wind-Energy Projects likely with the turbine blades. Species differ in their vulnerability to collision, in the likelihood that fatalities will have large-scale cumulative impacts on biotic communities, and in the extent to which their fatalities are discovered and publicized. This chapter reviews information on the probabilities of fatalities, which are affected by both abundance and behavioral characteristics of each species. Factors such as the type, location, and operational schedules of turbines that may influence bird and bat fatalities are reviewed in this chapter. The overall importance of turbine-related deaths for bird populations is unclear. Collisions with wind turbines represent one element of the cumulative anthropogenic impacts on bird populations; other impacts include collisions with tall buildings, communications towers, other structures, and vehicles, as well as other sources of mortality such as predation by house cats (Erickson et al. 2001, 2005). While estimation of avian fatalities caused by wind-power generation is possible, the data on total bird deaths caused by most anthropogenic sources, including wind turbines, are sparse and less reliable than one would wish, and therefore it is not possible to provide an accurate estimate of the incremental contribution of wind-powered generation to cumulative bird deaths in time and space at current levels of development. Data on bat fatalities are even sparser. While there have been a few reports of bat kills from other anthropogenic sources (e.g., through collisions with buildings and communications towers), the recent bat fatalities from wind turbines appear to be unprecedentedly high. More data on direct comparisons of turbine types are needed to establish whether and why migratory bats appear to be at the greatest risk of being killed. Clearly, a better understanding of the biology of the populations at risk and analysis of the cumulative effects of wind turbines and other anthropogenic sources on bird and bat mortality are needed. The construction and maintenance of wind-energy facilities alter ecosystem structure, through vegetation clearing, soil disruption, and potential for erosion, and this is particularly problematic in areas that are difficult to reclaim, such as desert, shrub-steppe, and forested areas. In the MidAtlantic Highlands forest clearing represents perhaps the most significant potential change through fragmentation and loss of habitat for forest-dependent species. Changes in forest structure and the creation of openings alter microclimate and increase the amount of forest edge. There may also be important interactions between habitat alteration and the risk of fatalities, such as bat foraging behavior near turbines. The recommendations in this chapter address the types of studies that need to be conducted prior to siting and prior to and following construction of wind-energy facilities to evaluate the potential and realized ecological impacts of wind-energy development. The recommendations also address

OCR for page 67
Environmental Impacts of Wind-Energy Projects assessing the degree to which a particular site is acceptable for wind-energy development and the types of research and monitoring needed to help inform decision makers. INTRODUCTION There are two major ways that wind-energy development may influence ecosystem structure and functioning—through direct impacts on individual organisms and through impacts on habitat structure and functioning. Environmental influences of wind-energy facilities can propagate across a wide range of spatial scales, from the location of a single turbine to landscapes, regions, and the planet, and a range of temporal scales from short-term noise to long-term influences on habitat structure and influences on presence of species. In this chapter, we review the documented and potential influences of wind-energy development on ecosystem structure and functioning, focusing on scales of relevance to siting decisions and on influences on birds, bats, and other vertebrates. Construction and operation of wind-energy facilities directly influence ecosystem structure. Site preparation activities, large machinery, transportation of turbine elements, and “feeder lines,” transmission lines that lead from the wind-energy facility to the electricity grid, all can lead to removal of vegetation, disturbance, and compaction of soil, soil erosion, and changes in hydrologic features. Although many of these activities are relatively local and short-term in practice (e.g., construction), there may be substantial effects on habitat quality for a variety of organisms. These changes will likely be detrimental to some species and beneficial to others. Wind-energy development that is focused on specific topographic features (e.g., ridgelines) that represent key habitat features for some species may have disproportionately detrimental impacts on those species that depend on or are closely associated with these habitats. Recent reviews of available literature have clearly documented direct impacts of wind turbines on birds and bats (GAO 2005; Barclay and Kurta 2007; Kunz et al. 2007), including death from colliding with turbine blades. As discussed below, little is known about the circumstances contributing to fatalities, but issues such as turbine height and design, rotor velocity, number and dispersion of turbines, location of the turbine on the landscape, and the abundance, migration, and behavioral characteristics of each species present are likely to influence fatality rates. In addition, non-flying organisms may be affected by turbine construction and operation, because of alteration of habitat and behavioral avoidance, possibly due to noise, vibration, motion of turbines, or their mere presence in the landscape. We can make three general predictions about the large-scale and longterm impacts of individual fatalities. First, life-history theory predicts that

OCR for page 67
Environmental Impacts of Wind-Energy Projects characteristics of populations of affected species determine the consequences of increased mortality: organisms whose populations are characterized by low birth rate, long life span, naturally low mortality rates, a high trophic level, and small geographic ranges are likely to be most susceptible to cumulative, long-term impacts on population size, genetic diversity, and ultimately, population viability (e.g., McKinney 1997; Purvis et al. 2000). Bats are unusual among mammals with respect to their life-histories, because they typically have small body sizes but long life spans (Barclay and Harder 2003), and the probability of extinction in bats has been linked to several of these characteristics (Jones et al. 2003). Second, the effects of a decline in one species on entire biotic communities is determined by the role of the species in the larger context: losses of keystone species, organisms that have a disproportionately high impact on ecosystem functioning (Power et al. 1996), and those that provide important ecosystem services (Daily et al. 1997) are of most concern. Species that are important predators and perform critical top-down control over communities, and species that are important prey sources can be keystone species in both natural and human-altered ecosystems (Cleveland et al. 2006). Notably, many raptors and insectivorous bats fill these roles. Finally, we do not know how the migration patterns of affected species will influence regional-scale mortality; we also do not understand the consequences of deaths of individuals of these migrating species to the local populations they originate from. Unfortunately this type of information is nearly impossible to obtain. The ecological influences of wind-energy facilities are complex, and can vary with spatial and temporal scale, location, season, weather, ecosystem type, species, and other factors. Moreover, many of the influences are likely cumulative, and ecological influences can interact in complex ways at wind-energy facilities and at other sites associated with changed land-use practices and other anthropogenic disturbances. Because of this complexity, evaluating ecological influences of wind-energy development is challenging and relies on understanding factors that are inadequately studied. Despite this, several patterns are beginning to emerge from the information currently available. Increased research using rigorous scientific methods will be critical to filling existing information gaps and improving reliability of predictions. In this chapter, we review the literature on the ecological effects of wind-energy development, focusing on wildlife and their habitats. We then provide an assessment of projected impacts of future development in the Mid-Atlantic Highland region based on the limited information currently available. Finally, we provide an overview of current methods and metrics for monitoring ecological impacts of wind-energy facilities, and propose research and monitoring priorities.

OCR for page 67
Environmental Impacts of Wind-Energy Projects BIRD DEATHS IN CONTEXT A primary question that arises from considerations of current and projected cumulative bird deaths from wind turbines is whether and to what degree they are ecologically significant. A related (but nonetheless different) question is how the number of turbine-caused bird deaths compares with the number of all anthropogenically caused bird deaths in the United States. The committee approaches the answer to the latter question with great hesitation, for four reasons. First, the accuracy and precision of data available to answer the question are poor. Although it is clear that more birds are killed by other human activities than by wind turbines, both natural mortality rates for many species and fatalities resulting from many types of human activities are poorly documented. In addition, different sources of human-caused fatalities do not affect all bird species to the same degree. Second, the demographic consequences of various mortality rates are poorly understood for most bird species, as are factors such as the timing of fatalities and sex or age bias in fatalities resulting from different anthropogenic causes, which could have a variety of demographic impacts. Moreover, the demographic and ecological importance of any given mortality rate being considered is relative to population size, which is poorly known for most species. Third, grouping all species together in any estimate provides information that is not ecologically relevant. For example, the ecological consequences and conservation implications of the deaths of 10,000 starlings (Sturnus vulgaris) are far different from those of the deaths of 10,000 bald eagles (Haliaeetus leucocephalus). Finally, consideration of aggregate bird fatalities across the United States from any cause—including those caused by wind-energy installations—is not the appropriate spatial scale to address the question of interest. Region-specific information about the demographic effects of any cause of mortality on species of interest would be much more informative. Thus, for example, it is more important to know how many raptors of a particular species are killed by turbines and other human mortality sources in a particular region than it is to know how many raptors are killed nationwide. Having said the above, we provide here estimates summarized by Erickson et al. (2005) and estimates reported by the U.S. Fish and Wildlife Service (USFWS 2002a). Those sources emphasize the uncertainty in the estimates, but the numbers are so large that they are not obscured even by the uncertainty. Collisions with buildings kill 97 to 976 million birds annually; collisions with high-tension lines kill at least 130 million birds, perhaps more than 1 billion; collisions with communications towers kill between 4 and 5 million based on “conservative estimates,” but could be as high as 50 million; cars may kill 80 million birds per year; and collisions with wind turbines killed an estimated 20,000 to 37,000 birds per year in

OCR for page 67
Environmental Impacts of Wind-Energy Projects 2003, with all but 9,200 of those deaths occurring in California. Toxic chemicals, including pesticides, kill more than 72 million birds each year, while domestic cats are estimated to kill hundreds of millions of songbirds and other species each year. Erickson et al. (2005) estimate that total cumulative bird mortality in the United States “may easily approach 1 billion birds per year.” Clearly, bird deaths caused by wind turbines are a minute fraction of the total anthropogenic bird deaths—less than 0.003% in 2003 based on the estimates of Erickson et al. (2005). However, the committee re-emphasizes the importance of local and temporal factors in evaluating the effects of wind turbines on bird populations, including a consideration of local geography, seasonal bird abundances, and the species at risk. In addition, it is necessary to consider the possible cumulative bird deaths that can be expected if the use of wind energy increases according to recent projections (see Chapter 2). TURBINES CAUSE FATALITIES TO BIRDS AND BATS Information on fatalities of birds and bats associated with wind-energy facilities in the Mid-Atlantic Highlands is limited, largely because of the relatively small amount of wind-energy development in the region to date, the modest investments in monitoring and data collection, and in some cases, restricted access to wind-energy facilities for research and monitoring. This lack of information requires the use of information from other parts of the United States (and elsewhere). The following discussion summarizes what is known regarding bird and bat fatalities caused by windenergy facilities throughout the United States. National and regional results are related to the potential for fatalities in the Mid-Atlantic Highlands where appropriate. Early industrial wind-energy facilities, most of which were developed in California in the early 1980s, were planned, permitted, constructed, and operated with little consideration for the potential impacts to birds or bats (Anderson et al. 1999). Discoveries of raptor fatalities at the Altamont Pass Wind Resource Area (APWRA) (Anderson and Estep 1988; Estep 1989; Orloff and Flannery 1992) triggered concern about possible impacts to birds from wind-energy development on the part of regulatory agencies, environmental groups, wildlife resource agencies, and wind- and electricutility industries throughout the country. Initial discoveries of bird fatalities resulted from chance encounters by industry maintenance personnel with raptor carcasses at wind-energy facilities. Although fatalities of many bird species have since been documented at wind-energy facilities, raptors have received the most attention (Anderson and Estep 1988; Estep 1989; Howell and Noone 1992; Orloff and Flannery

OCR for page 67
Environmental Impacts of Wind-Energy Projects 1992, 1996; Howell 1995; Martí 1995; Anderson et al. 1996a,b, 1997, 1999, 2000; Johnson et al. 2000a,b; Thelander and Rugge 2000; Hunt 2002; Smallwood and Thelander 2004, 2005; Hoover and Morrison 2005). This attention is likely because raptors are lower in abundance than many other bird species, have symbolic and emotional value to many Americans, and are protected by federal and state laws. Raptor carcasses also remain much longer than carcasses of small birds, making fatalities of raptors more conspicuous to observers. Raptors occur in most areas with potential for wind-facility development, although raptor species appear to differ from one another in their susceptibility to collisions. Early studies of wind-energy facility impacts on birds were based on the carcasses discovered during planned searches. However, fatality estimates did not account for potential survey biases, most importantly biases in searcher efficiency and carcass “life expectancy” or persistence. Most current estimates of fatalities include estimates for all species and are based on extrapolation of the number of observed fatalities at surveyed turbines to the entire wind-energy facility, although not all studies adequately correct for observer-detection bias and carcass persistence, the latter usually referred to as scavenger-removal bias (e.g., Erickson et al. 2004). Until relatively recently, little attention has been given to bat fatalities at wind-energy installations. This is largely because few bat fatalities have been reported at most wind-energy facilities (Johnson 2005). While some bat fatalities were reported beginning in the early 1990s, few of the earliest studies of fatalities at wind-energy facilities were designed to look for or evaluate bat fatalities, and thus did not use systematic search protocols or account for observer bias or scavenging. The scarcity of reported fatalities also may be due in part to the rarity of post-construction studies designed specifically to detect bat fatalities at wind-energy facilities. Recent surveys indicate that some wind-energy facilities have killed large numbers of bats in the United States (Arnett 2005; Johnson 2005), Europe (Dürr and Bach 2004; Hötker et al. 2004; UNEP/EUROBATS 2006), and Canada (R.M.R. Barclay, University of Calgary, personal communication 2006). BIRD AND BAT FATALITIES In the following discussion, fatality rate is presented as fatalities/ turbine/year or fatalities/MW/year. Because turbine size, and presumably risk, varies from facility to facility, we have chosen to make comparisons of fatalities among turbines using the metric fatalities/MW/year. The MW used in this metric represents the nameplate capacity for the turbines and does not represent the actual amount of MW produced by a turbine or wind-energy plant. The reader is referred to Chapter 2 for a more general discussion of nameplate capacity. A more accurate measure of MW pro-

OCR for page 67
Environmental Impacts of Wind-Energy Projects duction for individual turbines would provide a much better metric for comparison purposes. For example, two turbines with the same nameplate capacity may operate a much greater percentage of time at a Class 5 wind site than in a Class 4 wind site. Bird Species Prone to Collisions with Wind Turbines Songbirds (order Passeriformes) are by far the most abundant bird group in most terrestrial ecosystems, and also the most often reported as fatalities at wind-energy facilities. The number of fatalities reported by individual studies in the eastern United States ranges from 0 during a five-month study at the Searsburg, Vermont facility (Kerlinger 1997) to 11.7 birds per MW during a one-year study at Buffalo Mountain, Tennessee (Nicholson 2003). In a review of bird collisions reported in 31 studies at wind-energy facilities, Erickson et al. (2001) reported that 78% of the carcasses found at facilities outside of California were protected passerines (i.e., songbirds protected by the Migratory Bird Treaty Reform Act of 2005). The remainder of the fatalities included waterfowl (5.3%), waterbirds (3.3%), shorebirds (0.7%), diurnal raptors (2.7%), owls (0.5%), fowl-like (galliform) birds (4.0%), other (2.7%), and non-protected birds (e.g., starling, house sparrow, and rock dove or feral pigeon; 3.3%). Despite the relatively high proportion of passerines recorded, actual fatalities of passerines probably are underrepresented in most studies, because small birds are more difficult to detect and scavenging of small birds can be expected to be higher (e.g., Johnson et al. 2000b). Moreover, given the episodic nature of bird migration, it is possible that many previous studies with relatively long search intervals failed to detect some fatalities of small birds during the migration season, and thus existing estimates of fatalities could be underestimates. Data allowing accurate estimates of bird fatalities at wind-energy facilities in the United States are limited, particularly in the Mid-Atlantic Highlands region. Of the studies reviewed for this report, 14 were conducted using a survey protocol for all seasons of occupancy for a one-year period (Table 3-1) and incorporated scavenging and searcher-efficiency biases into estimates (Erickson et al. 2000, 2004; Young et al. 2001, 2003a,b; Howe et al. 2002; Johnson et al. 2002, 2003b; Nicholson 2003; Kerns and Kerlinger 2004; Koford et al. 2004). Protocols used in these 14 studies varied considerably, but all generally followed the guidance in Anderson et al. (1999). The wind-energy facilities included in these studies contain turbines that range in size from 600 kW to 1.8 MW. Passerines make up 75% of the fatalities at these facilities and 76% of the fatalities at the two forested facilities in the eastern United States (Table 3-2, Figure 3-1). The greatest difference between fatalities at wind-energy facilities in the eastern United States and those in other regions is the relative abundance of doves,

OCR for page 67
Environmental Impacts of Wind-Energy Projects pigeons, and “other” species (e.g., swifts and hummingbirds, cuckoos, woodpeckers) in the east. Total annual bird fatalities per turbine and per MW are similar for all regions examined in these studies, although data from the two sites evaluated in the eastern United States suggest that more birds may be killed at windenergy facilities on forested ridge tops than in other regions. It is not known whether this is due to higher risk of collisions at these sites, or higher abundance of birds in the region. Most studies report that passerine fatalities occur throughout facilities, with no identified relationship to site characteristics (e.g., vegetation, topography, turbine density). The relatively high proportion of passerines probably reflects the fact that this group is by far the most abundant of all birds at the facilities where these fatalities occurred. Relative exposure is difficult to measure and there are no data suggesting that fatalities expressed as percentages are proportional to abundance. As discussed below, behavior appears to be important in determining the risk of collision. The combined average raptor fatality rate for the 14 studies (Table 3-2) is 0.03 birds per turbine/year and 0.04 per MW/year. The regional raptor fatalities per MW/year are similar, ranging from 0.07 in the Pacific Northwest region to 0.02 in the eastern United States. With the exception of the two eastern facilities, Mountaineer and Buffalo Mountain, which are in forest (68 MW combined), the land use/land cover is similar in all regions (Table 3-1). Most of the wind-energy facilities occur in agricultural areas (333 MW combined) and agriculture/grassland/Conservation Reserve Program lands (438 MW combined), and the remainder occur in short-grass prairie (68 MW combined). Landscapes vary from mountains, plateaus, and ridges, to areas of low relief. Aside from the size of the rotor-swept area, each of these facilities used similar technologies. Bird abundance may be an important factor in fatalities (discussed in more detail below), although standard estimates of bird use are not available for all 14 studies. Interpreting fatalities of breeding and migrating passerines is challenging because of inadequate estimation of exposure of different species to risk. The most common fatalities reported at wind-energy facilities in the western and middle United States are relatively common species, such as horned lark (Eremophila alpestris), vesper sparrow (Pooecetes gramineus), and bobolink (Dolichonyx oryzivorus). These species perform aerial courtship displays that frequently take them high enough to enter the rotor-swept area of a turbine (Kerlinger and Dowdell 2003). The western meadowlark (Sturnella neglecta), on the other hand, is quite common and is frequently reported in fatality records, yet is not often seen flying as high as the rotorswept area of wind turbines. By contrast, crows, ravens, and vultures are among the most common species seen flying within the rotor-swept area of turbines (e.g., Orloff and Flannery 1992; Erickson et al. 2004; Smallwood and Thelander 2004, 2005), yet they seldom are found during carcass

OCR for page 67
Environmental Impacts of Wind-Energy Projects TABLE 3-1 Description of Wind-Energy Facilities Based on Data Collected During the Period of Bird Occupancy over a Minimum Period of One Year and Where Standardized Bird Mortality Studies Conducted, Including Scavenging and Searcher Efficiency Biases. Vegetation Categories Include Agriculture (AG), Grass Land (Grass), Conservation Reserve Program (CRP) Grasslands, Short-Grass Steppe, and Forest. Seasons Include Spring (S), Summer (Su), Fall (F), and Winter (W) Wind Facility Vegetation Dates of Study Vansycle, OR Ag/Grass/CRP 1/99-12/99 Nine Canyon, WA Ag/Grass/CRP 9/02-8/03 Stateline, OR/WA Ag/Grass/CRP 1/02-12/03 Combine Hills, OR Ag/Grass/CRP 2/04-2/05 Klondike, OR Ag/Grass/CRP 2/02-2/03 Foote Creek Rim, WY Phase I Short-grass Steppe 11/98-12/00 Foote Creek Rim, WY Phase II Short-grass Steppe 7/99-12/00 Wisconsin Agriculture Spring 98-12/00 Buffalo Ridge, MN Phase I Agriculture 4/94-12/95     3/96-11/99 Buffalo Ridge, MN Phase II Agriculture 3/98-11/99 Buffalo Ridge, MN Phase III Agriculture 3/99-11/99 Top of Iowa, IW Agriculture 4/03-12/03 Buffalo Mountain, TN Forest 10/01-9/02 Mountaineer, WV Forest 4/03-11/03 surveys. Clearly, abundance and behavior interact to influence exposure of breeding passerines and other birds to the risk of collisions. While estimated bird fatalities for these 14 wind-energy facilities are relatively low when compared to other sources of bird fatalities (Erickson et al. 2001), the lack of multiyear estimates of density and other population characteristics at most wind-energy facilities makes it difficult to draw general conclusions about their effects on populations of bird fatalities. In addition, lack of replication of studies among facilities and years makes it impossible to evaluate natural variability and the likelihood of unusual episodic events in relation to bird fatalities. Influences of Turbine Design on Bird Fatalities The structure and design of existing wind turbines vary considerably, and it is likely that additional modifications will occur over time. Changes in turbine design result from technological improvements, differences in

OCR for page 67
Environmental Impacts of Wind-Energy Projects Search Interval Number of Turbines in Facility Number of Turbines Searched Reference 28 days 38 38 Erickson et al. (2000) 14 days S, Su, F 37 37 Erickson et al. (2003b) 28 days W       14 days 454 124-153 Erickson et al. (2004) 28 41 41 Young et al. (2005) 28 days 16 16 Johnson et al. (2003b) 28 days 69 69 Young et al. (2001) 28 days 36 36 Young et al. (2003b) Daily-weekly 31 31 Howe et al. (2002) 7 days 73 50 Johnson et al. (2002) 14 days 73 21   14 days 143 40 Johnson et al. (2002) 14 days 138 30 Johnson et al. (2002) 2-3 days 89 26 Koford et al. (2004) 2/week-weekly 3 3 Nicholson (2003) S-11 days 44 44 Kerns and Kerlinger Su-28 days     (2004) F-7 days       generation capacity, and in some cases, modifications to meet site-specific needs (such as modification of height because of Federal Aviation Administration [FAA] constraints). Differences in design of turbines could affect fatality rates of birds. For example, as turbine heights increase, nocturnally migrating passerines could be increasingly affected because they tend to migrate at levels above 400 feet (see Appendix C for further discussion). Much of the early work on fatalities at wind-energy facilities occurred in California, because most wind energy was produced at three windresource areas: APWRA, San Gorgonio, and Tehachapi. Not coincidently, some of the existing concern regarding the impact of wind-energy facilities on birds is rooted in the fatalities that have occurred at the APWRA, and thus although many of the characteristics of APWRA differ from those of the Mid-Atlantic Highlands region, the history of APWRA provides important background and context. The APWRA currently has between 5,000 and 5,400 turbines of various types and sizes, with an installed capacity of approximately 550 MW

OCR for page 67
Environmental Impacts of Wind-Energy Projects vent them from coming into contact with turbines, or it is possible that their rarity has not yet led to a recorded fatality of any of those species. Ecological Implications of Projected Cumulative Impacts These projections of cumulative bat and bird fatalities for the Mid-Atlantic Highlands by the year 2020 assume that bat and bird populations living in or migrating through the region each year would be constant. The latter assumption is likely to be violated given assorted caveats about expected inter-annual variability; however, given that we have presented both worst-case (maximum number of fatalities/year) and best-case (minimum number of fatalities/year) scenarios, our projected fatality rates in the Mid-Atlantic Highlands bracket expected extremes. These projected fatalities can best be considered as hypotheses to be tested with future data on fatalities from the Mid-Atlantic Highlands and other regions where bird and bat fatalities have been reported, and by adjusting monitoring protocols to minimize potentially confounding assumptions (Kunz et al. 2007). A question that arises from these projections is whether they are of biological importance to bat and bird populations. The answer differs for birds and bats and for migratory and local populations. For birds, it is unlikely that this predicted level of fatalities would result in measurable impacts to migratory populations of most species. However, for rare species and local populations, the impacts, when combined with other sources of mortality such as large weather-related bird kills, could affect viability, and thereby affect overall risks to populations. A definitive conclusion on these predicted impacts requires more information on the demographics of rare and local populations of birds than is currently available. For bats, the question draws attention to the almost complete lack of data for population estimates of any species considered here, either on a regional or continental scale (Kunz et al. 2007). A risk assessment of biological impacts typically requires knowledge of baseline populations. Nonetheless, the numbers of fatalities projected above for bats in the MidAtlantic Highlands suggest that bat populations might be at risk, because they reflect fatality rates as high as or higher than fatality rates that have been reported for bats from other measurable anthropogenic sources (Kunz et al. 2007). CONCLUSIONS AND RECOMMENDATIONS Our understanding of the ecological effects of wind-energy development in the Mid-Atlantic Highlands region and elsewhere is limited by minimal monitoring efforts at existing wind-energy facilities and by poor understanding of key aspects of species ecology, of causal mechanisms

OCR for page 67
Environmental Impacts of Wind-Energy Projects underlying fatalities at wind-energy facilities, and of the reliability of our projections of fatalities at wind-energy facilities. This section contains the committee’s conclusions about the known and potential ecological effects of wind-energy projects, identification of information needs, and recommendations for research and monitoring. Ecological Effects of Wind-Energy Projects While research and monitoring studies admittedly are limited, a synthesis of the existing studies indicates that adverse effects of wind-energy facilities on ecosystem structure and functioning have occurred. This knowledge should be used to guide decisions on planning, siting, and operation. Wind turbines cause fatalities of birds and bats through collision, most likely with the turbine blades. Species differ in their vulnerability to collision. The probability of fatality is most likely a function of abundance, local concentrations, and the behavioral characteristics of species. Migratory tree-roosting bat species appear to be most susceptible to direct impacts. To date, the highest fatality rates have been reported in the Mid-Atlantic Highlands, although recent evidence suggests that bats from grassland and agricultural landscapes may also experience high fatality rates. Migratory tree bats constitute over 78% of all fatalities reported at wind-energy facilities, and thus appear to be killed disproportionately to highly colonial species. To date, no endangered species have been reported being killed at existing wind-energy facilities, although only a few sites have been monitored. Increased risks are expected as more wind-energy facilities are developed. Risks of fatalities to bats in the southwestern United States, especially in Texas, where large wind-energy facilities exist and have been proposed, are largely unknown because data have not been reported for most of these facilities. Abundance interacts with behavior to influence exposure of breeding passerines, raptors, and bats to the risk of collisions. Raptors appear to be the most vulnerable to collisions. On average raptors constitute 6% of the reported fatalities at wind-energy facilities, yet they are far less abundant than most other groups of birds (e.g., passerines). By contrast, crows, ravens, and vultures are among the most common species seen flying within the rotor-swept area of turbines, yet they are seldom found during carcass surveys. Nocturnally migrating passerines are the most abundant species at most wind-energy facilities and are the most commonly reported fatalities. Nonetheless, fatalities among passerines vary more than can be explained by abundance alone. Species differ in the extent to which their fatalities are discovered

OCR for page 67
Environmental Impacts of Wind-Energy Projects and publicized. Small birds and bats are more difficult to find than others during planned searches and incidentally. Large birds such as raptors are more easily seen, and are often more publicized because of their charismatic status and perceived importance in the environment. The location of wind-energy facilities on the landscape (e.g., agricultural lands, ridge tops, canyons, grasslands) influences bird and bat fatalities. Available evidence suggests that fatalities are positively correlated with bird abundance. Landscape features influence density by concentrating prey or through providing favorable conditions for other activities such as nesting, feeding, and flying (e.g., updrafts for raptor soaring and linear landscapes for bats). The characteristics (e.g., rotor-swept area, height, support structure, lighting, number of turbines) of wind-energy facilities may act synergistically to cause bird and bat fatalities. Newer, larger turbines installed on monopoles may cause fewer bird fatalities per MW than the smaller, older, lattice-style turbines, but the ability to determine the significance of these characteristics is limited by sparse data; in addition, other factors such as the local and regional abundances of birds and bats and landscape variation confound understanding of the effects of turbine characteristics noted above. The lack of estimates of population sizes and other population parameters for birds and bats and the lack of multiyear studies at most existing wind-energy facilities make it difficult to draw general conclusions about how wind turbines and population characteristics interact to influence mortality of birds and bats. In addition, lack of replication of studies among facilities and years makes it impossible to evaluate natural variability, in particular unusual episodic events, in relation to fatalities and to predict the potential for future population effects. It is essential that the potential for population effects be evaluated as wind-energy facilities become more numerous. Fatality rates of migratory tree bats appear to be high in some landscapes (e.g., forested ridge tops), although almost nothing is known about the population status of these species, and the biological significance of reported fatalities. Nonetheless, this lack of data on bat populations points to a critical need to evaluate the status of these and other species that may be at risk, especially as wind-energy facilities proliferate, and a need to evaluate where major cumulative impacts could be expected. The construction and maintenance of wind turbines and associated infrastructure (e.g., roads) alters ecosystem structure through vegetation clearing, soil disruption, and potential for erosion and noise. Based on similar types of construction and development, it is likely that wind-energy facilities will adversely alter ecosystems indirectly, especially through the following cumulative impacts:

OCR for page 67
Environmental Impacts of Wind-Energy Projects Forest clearing resulting from road construction, transmission lines leading to the grid, and turbine placements represents perhaps the most significant potential change through habitat loss and fragmentation for forest-dependent species. This impact is particularly important in the Mid-Atlantic Highlands, because wind-energy projects there all have been constructed or proposed in forested areas. Changes in forest structure and the creation of openings may alter microclimate and increase the amount of forest edge. Plants and animals throughout the ecosystem respond differently to these changes, and particular attention should be paid to species listed under the ESA and species of concern (Appendix C) that are known to have narrow habitat requirements and whose niches are disproportionately altered. Information Needs Here we identify information needs related to understanding, predicting, and managing bird and bat fatalities and landscape and habitat alterations. For each of these categories we suggest important information needs that we judge should be given the highest priority for monitoring and research based on our collective understanding of the issues, weighed by tractability and best practices. The following recommendations are not meant to apply to every situation and should be modified given the characteristics of the site being developed, the species of concern, the results of pilot studies, and the amount of information applicable to that site. If wind-energy development continues in a region, research and monitoring protocols should evolve as more becomes known. Research is needed to develop mitigation approaches for existing facilities and to aid in assessing risk at proposed facilities. The latter is particularly important in landscapes where unusually high bird and bat fatalities have already been reported and in regions where facilities are planned where little is known about migration, foraging, and fatalities associated with wind-energy facilities (e.g., the Mid-Atlantic Highlands and the south-western United States). Following accepted scientific protocols, hypotheses should be developed to help address unanswered questions. Testing hypotheses promises to provide science-based answers that will help inform developers, decision makers, policy makers, and other stakeholders concerning actual and expected impacts of wind-energy development on bat and bird population and on landscapes and habitats of other animals that might be altered by construction. Some of these information needs are beyond the scope of any individual developer (e.g., population status of affected species). Therefore, a collab-

OCR for page 67
Environmental Impacts of Wind-Energy Projects orative effort by industry and agencies to fund the necessary research to address these overarching questions should be initiated. Other information could be developed as part of the permitting process. Decision makers could require owners and developers to fund research and monitoring studies by qualified researchers at the proposed wind-energy facilities; developers and operators should provide full access (subject to safety and proprietary concerns) to researchers at existing wind-energy facilities. The research should be conducted openly and the protocols and results should be subject to peer review. Follow established scientific principles in conducting monitoring studies and experiments. Follow established research methods and metrics (summarized in Appendix C). Evaluate the efficacy of tools needed to make reliable predictions that would assess measures to reduce the risk of fatalities (e.g., evaluate potential mitigation measures). Develop new quantitative tools to predict fatalities at proposed and existing wind-energy facilities. Develop estimates of exposure for use in evaluating fatalities and for estimating risk (e.g., radar studies at existing facilities in combination with fatality data to develop stronger risk-assessment tools). Improve tools and protocols that can discriminate migrating birds from migrating bats, operate in inclement weather, and provide cost-effective estimates of numbers and movements of flying birds and bats. Develop models to predict risk based on geographic region, topography, season, weather, lunar cycles, and characteristics of different turbines. Improve methods and metrics to determine the context of the number of fatalities related to the number of birds moving through the airspace (proportionality). Identify potential biases associated with estimation of fatalities, including necessary search effort (plot size, frequency of search, methods of searching), the probability that a carcass will be detected if present, and the probability that a carcass will be removed so that its detection probability is zero. Encourage and conduct studies to support impact assessments. Assess effects of changing technologies (e.g., larger turbines) on bird and bat fatalities. Identify impacts of different types of lighting on bat and bird fatalities.

OCR for page 67
Environmental Impacts of Wind-Energy Projects Assess how different landscape features may affect bird and bat fatalities (mountain ridges, agriculture, grassland, canyons). Assess how weather fronts influence bat and bird fatalities. Identify bat and bird migratory patterns over space and time. Determine whether migratory birds and bats adjust their migratory paths or exhibit other behaviors that may cause them to avoid turbines. Determine whether fatalities from turbines reduce the breeding or stopover density and reproductive success of birds and bats. Conduct studies to identify methods of mitigating impacts of wind turbines on bats, birds, and other wildlife. Hypothesis-Based Research on Bats Knowledge about bat fatalities at wind-energy plants is very limited, mainly because the large number of bats killed has been recognized only recently. Eleven hypotheses are listed below, as examples, to help address how, when, where, and why bats are being killed at wind-energy facilities (Kunz et al. 2007). These hypotheses are not mutually exclusive, as several postulated factors might act synergistically to produce the high fatalities that have been reported. Linear-Corridor Hypothesis: Wind-energy facilities constructed along forested ridge tops create clearings with linear landscapes that are attractive to bats. Bats frequently use these linear landscapes during migration and while commuting and foraging (Limpens and Kapteyn 1991; Verboom and Spoelstra 1999; Hensen 2004; Menzel et al. 2005a), and thus may be placed at increased risk of being killed (Dürr and Bach 2004). Roost-Attraction Hypothesis: Tree-roosting bats commonly seek roosts in tall trees (Pierson 1998; Kunz and Lumsden 2003; Barclay and Kurta 2007) and thus if wind turbines are perceived as potential roosts (Ahlén 2002, 2003; Hensen 2004), their presence could contribute to increased risks of being killed when bats search for night roosts or during migratory stopovers. Landscape-Attraction Hypothesis: Modifications of landscapes needed to install wind-energy facilities, including the construction of wide power-access corridors and removal of trees to create clearings (usually 0.5-2 ha) around each turbine site, create conditions favorable for insects on which bats feed (Lewis 1970; Grindal and Brigham 1998; Hensen 2004). Thus, bats that are attracted to and feed on insects in these altered landscapes may be at an increased risk of being killed by wind turbines. Low Wind-Velocity Hypothesis: Fatalities of aerial feeding and

OCR for page 67
Environmental Impacts of Wind-Energy Projects migrating bats are highest on nights during periods of low wind velocity (Fiedler 2004; Hensen 2004; Arnett 2006), in part because flying insects are most active under these conditions (Ahlén 2002, 2003). Heat-Attraction Hypothesis: Flying insects are attracted to the heat produced by nacelles of wind turbines (Corten and Veldkamp 2001; Ahlén 2002, 2003; Hensen 2004). As bats respond to high densities of flying insects near wind turbines, they may be at increased risk of being struck by turbine blades. Acoustic-Attraction Hypothesis: Bats are attracted to audible and ultrasonic sound produced by wind turbines (Schmidt and Joermann 1986; Ahlén 2002, 2003). Sounds produced by the turbine generator and the swishing sounds of rotating turbine blades may attract bats, thus increasing risks of collision and fatality. Visual-Attraction Hypothesis: Insects flying at night are visually attracted to wind turbines (von Hensen 2004). Inasmuch as bats may feed on those insects, they become vulnerable to collisions with the turbine blades. Echolocation-Failure Hypothesis: Migrating and foraging bats fail to detect wind turbines by echolocation, or miscalculate rotor velocity (Ahlén 2002, 2003). If bats are unable to detect the moving turbine blades, they may be struck and killed directly. Electromagnetic-Field Disorientation Hypothesis: If bats have receptors sensitive to magnetic fields (Buchler and Wasilewski 1985), and wind turbines produce complex electromagnetic fields in the vicinity of the nacelle, the flight behavior of bats may be altered by these fields and thus increases their risk of being killed by rotating turbine blades. Decompression Hypothesis: Bats flying in the vicinity of turbines may experience rapid decompression (Dürr and Bach 2004; Hensen 2004). Rapid pressure change may cause internal injuries or disorientation, thus increasing risk of death. Thermal-Inversion Hypothesis: The altitude at which bats migrate and/or feed may be influenced by thermal inversions, forcing them to the altitude of rotor-swept areas (Arnett 2005). The most likely impact of thermal inversions is to create dense fog in cool valleys, possibly concentrating both bats and insects on ridges, and thus encouraging bats to feed over the ridges on those nights, if for no other reason than to avoid the cool air and fog. Research Recommendations Research should focus on two general lines of inquiry, including methodological research addressing improved tools and monitoring protocols as necessary, and hypothesis-driven research to provide information that will help inform developers, decision makers, policy makers, and other

OCR for page 67
Environmental Impacts of Wind-Energy Projects stakeholders to deal with actual and expected impacts of wind-energy development on populations and ecosystems. At a national scale, it would be appropriate to identify multiyear research goals that place the impacts of wind-energy development into a broad environmental perspective. Research initiatives should be encouraged to identify biological impacts of wind-energy development, and compare these impacts and risks with those of competing power-generating technologies. Research should focus on regions and sites where existing and new information suggest the greatest potential for biologically significant adverse impacts on birds and bats at proposed and existing wind-energy facilities. For example, while current evidence suggests that bat fatalities have been the highest at wind-energy facilities in forested mounted ridge tops in the Mid-Atlantic Highlands, recent monitoring studies in agricultural landscape in the Midwest and at wind-energy facilities in southwestern Alberta, Canada, suggest that fatality rates of migratory tree bats may be as high as those reported for the Mid-Atlantic Highlands. We also expect that high bat fatalities are occurring or will occur in the southwestern United States, where large numbers of Brazilian free-tailed bats form maternity colonies (McCracken 2003), and where there is high bat-species richness (O’Shea and Bogan 2003). However, to date, no appropriately designed fatality surveys have been reported at wind-energy facilities in this region. Given the observed geographic variation in fatality rates of both birds and bats, research is needed to evaluate where the risks or fatalities are high so that similar areas can be avoided. Improved assessments, with a focus on evaluation of causes and cumulative impacts, should be an urgent research priority. Proceeding with large-scale development of wind-energy facilities before identifying risks likely threatens both bats and the public acceptance of wind energy as an environmentally friendly form of energy (Kunz et al. 2007). Thus, the initial developments should be used as an opportunity to understand the risks before the full wind-energy potential of the Mid-Atlantic Highlands is developed. The highest priority for avian habitat is the quantification and prediction of habitat impacts, including loss because of the spatial demands of wind-energy facilities (e.g., roads and turbine pads) and displacement impacts because of behavioral response or habitat degradation, particularly on forest-dwelling and shrub-steppe and grassland birds. In addition, the role of wind in large-scale fragmentation of habitat for species dependent on forests should be evaluated. Finally, the impact of habitat loss or modification should be evaluated in terms of the potential for demographic impacts on ground-nesting birds. Clearly defined pre- and post-construction studies are needed to inform decision makers about the feasibility of constructing a new project and

OCR for page 67
Environmental Impacts of Wind-Energy Projects mitigating the adverse effects of existing facilities. The studies should be replicable and compared with other studies conducted in areas with similar topography and habitat. Where appropriate, pre- and post-construction studies should be conducted as recommended below. Pre-siting Studies Conduct pre-siting studies that allow the comparison of multiple sites when making decisions about where to develop wind energy. Identify species of special concern and their habitat needs; these include species listed under the federal ESA, such as the West Virginia northern flying squirrel, as well as species listed by the appropriate state, such as the Allegheny woodrat. Pre-construction Studies Conduct regional assessments to identify species of concern, including those vulnerable to direct impacts and those vulnerable to habitat loss. Develop pre-construction estimates of potential biological significance of fatalities based on estimated fatality rates and demographics of the species of concern. Conduct multiyear studies when appropriate to assess daily, seasonal and interannual variability of bird and bat populations. Establish species-specific abundance, periods of use (both seasonally and within a day), and behavior in relation to proposed turbines placement locally, regionally, and nationally. Identify habitat characteristics for birds, bats, and other animals, such as topography and types of vegetation at each proposed sites. Post-construction Studies Conduct full-season, multiyear, post-construction studies where appropriate to assess variability of bird and bat fatalities. Identify number, species composition, and timing of fatalities. Estimate the biological significance of bird and bat fatalities. Clarify the relationship of small-scale (e.g., habitat disturbance and species displacement) versus large-scale impacts (e.g., landscape alteration and fragmentation) of development on bird and bat populations. Conduct experiments to test alternative mitigation procedures (strategic shutdowns, feathering, blade painting and other potential deterrents, and lighting) that could avoid or reduce current fatality rates—independent of a meta-analysis to assess biological significance and adverse cumulative impacts.

OCR for page 67
Environmental Impacts of Wind-Energy Projects General Develop predictive and risk-assessment models of potential cumulative impacts of proposed wind-energy facilities, based on monitoring studies and hypothesis-based research. Summary More information is needed on the characteristics of bird and bat fatalities at wind facilities in all regions of the county, and in particular areas that are relatively unstudied such as the Mid-Atlantic Highlands, the arid southwest, and coastal areas. Turbine characteristics, turbine siting, and abundance appear to be important factors in determining the risk of raptor fatalities at wind-energy facilities. Compared to relatively high raptor fatalities at some older facilities in California, direct impacts of wind-energy development on passerines at the current level of development appear to be minimal. At current levels of development existing data suggest that new-generation turbines (e.g., fewer turbines mounted on monopoles with greater rotor-swept zones) may cause lower bird fatalities in agricultural and grassland areas than older smaller turbines have caused in California. Data on bird fatalities are absent for many existing wind-energy facilities, particularly in Texas and the southwestern United States. Additionally, new areas are being proposed for development where no previous data on bird and bat fatalities exist. It is important to assess impacts in existing and new areas to determine if trends are consistent with existing information. In particular, only two short-term post-construction studies have been conducted in the Mid-Atlantic Highlands and any new facilities should be used as learning opportunities. Additional information also is needed to characterize bat fatalities in all regions of the country where wind-energy development has occurred or where it is expected. Most wind-turbine-related bat fatalities in the United States have been of migratory species. To date, no fatalities of federally listed bat species have been documented, although as wind-energy development increases geographically, some threatened and endangered species could be at risk. Among the studies that have been conducted, the highest bat fatality rates appear to occur episodically in late summer and early autumn during periods of relatively low wind speeds (< 6 m/sec), at times when wind-energy generation is low, especially following passing weather fronts. To date, few studies have evaluated fatalities during spring migration or during the summer maternity period. Moreover, among fatality surveys that have been conducted, few have consistently corrected results for observer bias and scavenger removal, protocols that are needed to provide reliable data on fatalities. While current evidence suggests that the highest fatality rates are of migratory tree-roosting species along ridge tops in

OCR for page 67
Environmental Impacts of Wind-Energy Projects eastern deciduous forests, recent evidence suggests that similar fatality rates may occur in some agricultural and grassland regions. Bats in other regions of the country that have high wind capacity and are currently undergoing rapid wind-energy development (e.g., southwestern United States), where some of the largest bat colonies in North America are known, may be at considerable risk from wind-energy development during both migratory and maternity periods. Projected development of wind-energy facilities throughout the United States should be evaluated for cumulative impacts on different species considered at risk.