Abstract
While wind energy has been employed for electricity production since the 1880s, it wasn't until the oil crisis of the 1970s that commercial wind energy production was pursued actively in the United States. Wind energy use has grown rapidly since it began to be promoted as an alternative to fossil fuels and was accorded sponsorship by the state of California in the 1980s and by the Federal Government beginning in the late 1990s. Concerns about avian and chiropteran deaths caused by wind turbines emerged in the early 1990s [Howell and DiDonato, 1991], with widely varying estimates of the fatality rates, and studies were mounted to assess these rates as early as 1998 [Smallwood and Thelander, 2005]. Aggregate U.S. mortality estimates have been reported ranging from 20,000 to 573,000 birds annually [Erickson et al., 2001, 2005, Loss et al., 2013, Manville, 2009, Smallwood, 2013, Sovacool, 2012]. High profile lawsuits in such places as Altamont, CA (2007), Ventura, CA (2012), Nantucket Sound, MA (2012), Port Clinton, OH (2014) have brought the issue to national prominence.
The naïve approach to estimating turbine-related avian and chiropteran mortality - surveying periodically for bird and bat carcasses in designated areas near turbines at prescribed time intervals, and scaling the counts by time interval and study area - leads to grossly distorted estimates, for a variety of reasons. Some carcasses will be removed by scavengers before the survey, for example; some carcasses may be present but undetected at the time of the survey; some fatally injured birds or bats may survive long enough to alight outside the study area; and carcasses may be discovered whose death arose from other causes or during other time periods.
A number of investigators have developed modeling approaches leading to proposed adjustment formulas intended to overcome the distortions and biases of the naïve approach [Erickson et al., 1998, Johnson et al., 2003, Shoenfeld, 2004, Pollock, 2007, Huso, 2011], each embodying slightly different assumptions about the processes affecting carcass discovery. The wide variability of these estimation formulas leaves practitioners uncertain which of them (if any) to use. Here we explain the assumptions that underlie four commonly used estimation formulas, illustrate when each is appropriate and how they differ, and propose a new modelbased Avian and Chiropteran Mortality Estimator called "ACME" that extends all four of them and introduces three new features to improve the reliability of mortality estimates: the diminishment of Field Technician (FT) discovery proficiency as carcasses age; the reduced rate of scavenger removal as carcasses age; and the possibility that some but not all carcasses present but undiscovered by FTs in one search may be discovered in a later search.