Abstract
Criterion Power Partners, LCC, completed construction and initiated operation of the Criterion Wind Project in Garrett County, Maryland in 2010. The project includes 28, 2.5 megawatt wind turbine generators for a total generating capacity of 70 MW. Beginning in April 2011, Criterion initiated the first year of post-construction monitoring surveys to estimate the impacts of project operations on bird and bat species. The following report contains results of the fatality and associated field surveys conducted from April to November 2011.
The primary objective of the fatality monitoring study was to determine the level of bird and bat mortality attributable to collisions with wind turbines for the entire facility for the study period. The monitoring study consisted of four components: 1) standardized carcass surveys of the project turbines; 2) searcher efficiency trials to estimate the percentage of carcasses found by searchers; 3) carcass removal trials to estimate the length of time that a carcass remained in the field for possible detection; and 4) adjusted fatality estimates for bird and bat species calculated by correcting survey results for potential biases (e.g., area searched, searcher efficiency, carcass removal).
The monitoring study period was from April 5 to November 15, 2011. Search plots were established around all 28 turbines in the project and the carcass search schedule was for daily searches at all turbines (weather and safety permitting). Search plot size varied in shape and size, due to habitat constraints, but in most cases areas up to approximately 40-50 m (~130-165 ft) from the turbines were cleared of vegetation for access and construction purposes and this area was used as the search plot. Parallel transects were spaced and delineated approximately 5 m (~16 ft) apart within the search plot and surveyors systematically walked the transects while scanning the ground for fatalities or injured birds or bats.
During the study, 262 birds representing 46 species and 706 bats representing eight species were found either during standardized carcass searches or incidentally during the study period. The most commonly found bird species were red-eyed vireo and blackpoll warbler, while eastern red bat and hoary bat accounted for the majority of the bat fatalities found. Bird and bat fatalities were spread throughout entire survey period and throughout the entire project with the number of fatalities peaking for both birds and bats in the fall.
A total of 78 bird carcasses and 72 bat carcasses were placed for searcher efficiency trials. Searcher efficiency was estimated for the first half of the study period (April 5 to July 15 – labeled the spring) and the second half (July 16-November 15 – labeled the fall) to investigate potential changes over time. For small birds searcher efficiency was 0.70 in spring and 0.32 in fall, for large birds was 0.87 in spring and fall, and for bats was 0.72 in spring and 0.38 in fall. A total of 168 carcasses were placed for carcass removal trials. The mean carcass removal rate was similar between spring and fall for small and large birds and over the entire study period was 11.57 days for small birds and 10.94 for large birds. Mean carcass removal for bats varied by season and was 19.50 days for the spring and 13.33 days for the fall.
The probability that a carcass would remain in a search plot and be found by a searcher was 0.91 in the spring and 0.79 in fall for small birds and was 0.93 in spring and fall for large birds. The probability that a bat carcass was available and detected was 0.95 in spring and 0.84 in the fall.
Fatality estimates were adjusted based on the corrections for carcass removal, observer detection bias, and the area searched to account for carcasses potentially falling outside the plot. Combining both spring and fall estimates, the overall adjusted estimate for small birds was 15.25 small birds per turbine for the study period or 6.10 small birds per MW. For large birds the overall adjusted fatality estimate was 0.76 large birds per turbine for the study period or 0.30 large birds per MW. The fatality estimate for all birds combined was 16.01 birds per turbine or 6.40 birds per MW for the study period. For bats the overall estimate fatality estimate was 39.03 bats per turbine for the study period or 15.61 bats per MW.
Because the study utilized daily carcass searches the time of death for most casualties recovered was known, and an analysis to investigate the influence of nightly weather on fatality rate was conducted. The analysis utilized weather data collected at the turbines to investigate the correlation between weather variables and bird and bat mortality as well as an analysis to model what combination of weather variables was best predictive of mortality. Visibility data collected from a Maryland State Highway weather monitoring station was used to investigate the correlation between nightly visibility and mortality. Bat mortality was negatively correlated with wind speed and positively correlated with temperature; that is, as average nightly wind speed increased or the proportion of nights with wind speed greater than 6 m/s increased bat mortality decreased, and as average nightly temperature increased bat mortality increased. Mean nightly temperature, nightly wind speed, and the interaction of temperature and wind speed were the best variables for predicting nightly bat mortality. Bird mortality was negatively correlated with wind speed but was not correlated with nightly temperature. Bird mortality was negatively correlated with visibility. That is, as visibility increased bird mortality decreased.