Abstract
WEST, Inc. (2006) reported on the first year of fatality monitoring in the Diablo Winds Energy Project, which replaced 169 vertical axis wind turbines with 31 larger horizontal axis wind turbines in the Altamont Pass Wind Resource Area (APWRA). WEST, Inc. also compared bird utilization and mortality in the project area to utilization and mortality reported by Smallwood and Thelander (2004) across the APWRA. West, Inc. concluded mortality declined after Diablo Winds even though raptor utilization of the area increased. I reviewed the WEST, Inc. report, other reports of bird utilization and mortality at other wind farms, and I analyzed the Smallwood and Thelander data collected at the 169 wind turbines that were replaced.
I found WEST, Inc. inappropriately compared bird observations to 800 m to the data Smallwood and Thelander collected to 300 m. Recording birds seen to 800 m misses many of the birds that should have occurred between 300 and 800 m, but even so this practice artificially inflates the number of birds seen per unit time because the total area within an 800-m radius is seven times the area within a 300-m radius. I quantified the bias and adjusted the WEST, Inc. utilization estimates, revealing raptor activity in the study area decreased rather than increased after Diablo Winds.
The WEST, Inc. mortality estimator also is statistically biased because its adjustments for searcher detection error and scavenger removal rates were based on incorrect assumptions that inflated mortality estimates in the case of searcher detection error, and lowered estimates in the case of scavenger removal error. It was inappropriate to use rock dove as a surrogate for large birds and large raptors in searcher detection trials, and the same was true of scavenger removal trials. The mean number of days to carcass removal, which is the scavenger removal adjustment term used by WEST, Inc., turned out to increase systematically with the duration of the trial and with increasing sample size of carcasses used (until about 50 were used). Except for the case of large raptors, most carcasses deposited under wind turbines are quickly carried off by vertebrate scavengers, but when many carcasses are placed in the field at once vertebrate scavengers simply cannot carry them all away before some decompose to unattractive levels The remaining carcasses "mummify," and will usually remain throughout the trial, so the more days used in the trial, the more these mummified carcasses will factor into the calculation of the mean days to carcass removal. I quantified these biases, and developed an alternative scavenger removal term.
Relying on all suitable reports I obtained of scavenger removal trials, I developed a simple model to predict the percentage of carcasses remaining each day following the commencement of a removal trial or of the next fatality search interval, assuming a steady state of carcass deposition by wind turbines. Adjusted mortality estimates caused by the new and replaced wind turbines indicated overall bird mortality was reduced 70% by the Diablo Winds Energy Project, and raptor mortality was reduced 62%. Burrowing owl mortality was reduced 85%, and most of the total bird mortality reduction appeared to be among song birds. On the other hand, red-tailed hawk mortality increased nearly three-fold, and some species were killed by Diablo Winds that were not reported killed by the replaced turbines during Smallwood and Thelander's study, including golden eagle and bats. Differences in mortality were likely due to the reduced number of wind turbines, turbine siting, and the increased height above ground of the turbines, but additional research will be needed. Also, the repowering did not change the risk of collision for all raptors or all birds, perhaps because avian utilization of the Diablo Winds project site declined along with mortality between studies,1 or perhaps because the utilization data are not reliably comparable.
My analysis of bird utilization and mortality data among the new and replaced wind turbines indicates a decline occurred in bird utilization over the last 8 years, but the Diablo Winds Energy Project is killing smaller numbers of birds than did the Flowind vertical axis turbines. I believe bird mortality could have been reduced further had the wind turbine siting and operations recommendations of Smallwood and Thelander (2004, 2005) been implemented, and I believe future repowering efforts in the APWRA can also reduce mortality more substantially by adopting the recommendations of Smallwood and Thelander (2004, 2005), Smallwood and Neher (2004), and Smallwood and Spiegel (2005a,b,c).
The mortality adjustments in this report include multiple uncertainties and potential statistical biases yet to be completely characterized. Directed research is needed to address these and other uncertainties. Furthermore, the first annual report of Diablo Winds fatalities is preliminary because the time period was short and the sample size of fatalities small. Several years of monitoring will be needed to make more robust comparisons of mortality before and after the project.