Since 2003, when it was discovered that large numbers of bats were being killed at wind turbines in the eastern United States, our understanding of the impact of wind energy development on bats has increased and consistent patterns of fatality, including seasonality and species composition have become evident. Yet, many questions remain despite the wealth of data collected across numerous post-construction monitoring studies. We synthesized the recent literature to provide an overview of our current understanding of patterns of bat fatalities at wind energy facilities in the United States and Canada. Our understanding of the impact of wind energy development on bats continues to be hindered by inconsistencies among studies and lack of publicly available data. It will be difficult to fully address this complex issue and develop sustainable strategies to reduce the impact of wind turbines on bats and generate wind energy without standardized protocols for field methods, estimation of fatality, and greater cooperation among stakeholders.
Estimates of impacts
With only a fraction of individual studies available, it is difficult to place impact of wind energy development on bats into context. Several attempts to develop cumulative estimates for a given region or year have been made, each using different assumptions and datasets. Kunz et al. (2007)  estimated cumulative bat fatalities for the Mid-Atlantic Highlands would range from 33,000 to 62,000 or 59,000 to 111,000, depending on the projected installed capacity in the region by the year 2020. Cryan (2011)  used the average 11.6 bats/megawatt, based on data provided in Arnett et al. (2008)  and multiplied it by the total installed capacity in the United States, which at the time was approximately 40,000 megawatts, to estimate >450,000 bat fatalities each year in North America.
Two recent attempts were made to estimate bat fatality in the United States for 2012. Hayes (2013)  followed a similar approach to Cryan (2011) and based his analysis primarily on the limited dataset from Arnett et al. (2008). Hayes (2013) indicated that >600,000 bats were killed at wind energy facilities in 2012 and suggested that this was a conservative estimate. Smallwood (2013)  estimated up to 888,000 bats were killed in the United States in 2012. He used a larger dataset than Hayes (2013) and applied a common estimator and 3 adjustment factors to improve comparability among sites.
Arnett and Baerwald (2013)  estimated cumulative bat fatalities in the United States and Canada using data from 122 post-construction fatality studies from 73 facilities. They calculated a weighted mean based on regional means and weighted by installed capacity for each year from 1999 to 2010, then calculated and multiplied by total installed megawatts for each year. Arnett and Baerwald (2013) estimated [that] cumulative bat fatalities in the United States and Canada ranged from 0.8 to 1.7 million between 2000 and 2011. Based on their assumptions and installed wind power capacity, this estimate was projected to increase by 200,000 to 400,000 bats in 2012.
We recommend caution when using any of these estimates and to articulate the assumptions and limitations when citing these publications. Huso and Dalthorp (2013)  provided a critique of the methodology used by Hayes (2013), but many of the key issues could be applied to each of these cumulative estimates. In addition, Smallwood (2013) details numerous biases associated with individual studies that make comparing and combining data problematic. Each of these studies suffers from limited datasets that were based on public availability of studies and were not a representative sample of fatality across the region of inference. Although Arnett and Baerwald (2013) had by far the most data for their analysis, this collection of studies likely was still not representative of the entire United States and Canada. Yet, theirs is the only study to weight their estimates by both region and installed wind power capacity, which may provide a more conservative and accurate estimate (Arnett et al. 2016 ).
Even if more data were publicly available for use, another major challenge in estimating cumulative fatalities is lack of consistency in study design among sites. This, in part, is due to changes in turbine size and advances in study protocols and fatality estimation (Huso and Dalthorp 2013, Smallwood 2013). Nevertheless, varying levels of effort (e.g., temporally and spatially), differing methods for adjusting for imperfect detection, and different estimators used among sites are so large that estimates cannot be compared or combined (Huso 2011 ). Piorkowski et al. (2012) , recognizing the impossibility of obtaining reliable estimates of fatality from currently available data, identified development of a standardized experimental design and generalized fatality estimator as the number one issue in addressing impact of wind energy development on bats. Until this is realized, any attempt to develop cumulative estimates or project estimates of bat fatalities into the future is problematic.
We suggest that each of these be considered an order of magnitude estimate; taken together, they highlight the almost certain large number of bats being killed (i.e., on the order of hundreds of thousands per year) in the United States and Canada. Given that bats have a low reproductive rate—typically only having 1 or 2 pups/year—and require high adult survivorship to avoid population declines (Barclay and Harder 2003 ), this level of impact presumably puts bat populations at risk. Moreover, many species were thought to be declining prior to the onset and expansion of wind energy development, including species impacted by white-nose syndrome (Winhold et al. 2008 , Frick et al. 2010 ). Although population data are sparse or lacking for many bat species, current and presumed future level of fatality is considered to be unsustainable, and actions to reduce impact of wind turbines on bats should be implemented immediately.
Cris Hein, Michael Schirmacher
Bat Conservation International
Human-Wildlife Interactions. 10. 19-27.
Download original document: “Impact of wind energy on bats: A summary of our current knowledge ”