Abstract
Wind power has grown rapidly as an alternative energy source over the last decade. Although overall environmental impacts are relatively low, impacts to bats have yet to be fully assessed. Recent studies at other windfarms suggest regional variation in mortality rates and species affected. This study took place at Tennessee Valley Authority’s Buffalo Mountain Windfarm (BMW) in eastern Tennessee, currently the only commercial windfarm in the southeastern United States. Study objectives were to establish patterns of bat mortality and activity at BMW, and determine if and how they were related.
Mortality monitoring consisted of regular carcass searches conducted since BMW began operation in fall of 2000 and continued through fall of 2003. The adjusted bat mortality rate of 20.82 bats/turbine/year for the three-year monitoring period was greater than the average adjusted bat mortality rate at eight other windfarms (1.7 bats/turbine/year) by over an order of magnitude, but less than half the preliminary mortality rate (47.5 bats/turbine/year) reported at the Mountaineer Wind Energy Center, WV for 2003. The BMW mortality rate included adjustments determined by search bias trials (n = 6), which measured searcher efficiency (37.1%) and length of time before carcasses were removed by scavengers (average 6.3 days). Average distance of fatalities from turbines was 19.9 ± 2.3 m and distances of bat fatalities decreased logistically from turbines (y = 0.1223 – 0.1345 ln (x), r2 = 0.84, P = 0.0002), with zero fatalities predicted at 40.6 m; therefore, plot size of 50 m radius was deemed sufficient. Bat mortality showed a strong seasonal peak during late summer / early fall, with 70% of all bat fatalities occurring between 1 August and 15 September 2001 - 2003. A total of 119 bat fatalities were comprised of six species. Red bat (Lasiurus borealis) was the most common (63.1%), followed by eastern pipistrelle (Pipistrellus subflavus, 24.4%) and hoary bat (L. cinereus, 10.1%). The remaining 4.2% of bat fatalities consisted of three species: big brown bat (Eptesicus fuscus, n = 2), silver-haired bat (Lasionycteris noctivagans, n = 2), and Seminole bat (L. seminolus, n = 1). Adults were more common then juveniles (63.5% and 36.5%), and males more common than females (71.9% and 28.1%).
Bat activity was monitored with Anabat bat detectors and quantified as activity indices (AI), the proportion of one-minute increments in a night that contained one or more recorded bat calls. AI’s were compared across several time periods, locations, and altitudes, and compared with bat mortalities. Bat activity exhibited a seasonal peak in late summer / early fall during all three years, generally coinciding with the peak in mortality. Some of the variance associated with the likelihood of mortality was correlated with bat activity for 2002 and 2003 combined (r = 0.47, P < 0.0001). Considering only fresh fatalities, bat activity levels were greater during nights containing fatalities than nights without fatalities (t = 2.54, P = 0.0067). However, bat activity was not related to fatalities when the time interval was expanded from individual nights to search intervals and all fatalities were used (t = -1.05, P = 0.15). This suggests great variation in bat activity between nights, making temporal resolution an important factor when correlating levels of bat mortality and activity. Bat activity levels were influenced by presence, size and habitat configuration surrounding water features, as well as by altitude. Six species were acoustically identified as present at the BMW site, five of which were found among the turbine fatalities. Three species were found proportionally less as turbine fatalities than were acoustically recorded (eastern pipistrelle, big brown bat, and silver-haired bat) and two species were found proportionally more (red bat and hoary bat), indicating a greater collision risk for the latter species.
Overall numbers of bat fatalities at BMW, along with the lack of endangered species fatalities, indicate that population effects appear not to be significant. However, as the generation of electricity from wind increases in the eastern United States, it becomes critical to understand patterns in turbine-related mortality, and for future windfarm projects to determine population level effects of bat mortality.