We initiated a multi-year pre- and post-construction study in mid-summer 2006 to determine patterns of bat activity and evaluate the use of acoustic monitoring to predict mortality of bats at a proposed wind energy facility in northwest Massachusetts. The primary objectives of this study are to 1) determine level and patterns of activity of different species groups of bats using the proposed wind facility prior to and after construction of turbines; 2) correlate bat activity with weather and other environmental variables; and 3) combine results from this study with those from similar efforts to determine if indices of pre-construction bat activity can be used to predict post-construction bat fatalities at proposed wind facilities. Here we report results from the first year of pre-construction data collection.
We recorded echolocation calls of bats with Anabat II zero-crossing ultrasonic detectors programmed to record calls from 1900 to 0700 hr each day of the study from 26 July to 20 December 2006. We used meteorological (met) towers to vertically array detectors for acoustic sampling during this study (one detector at 10, 31, and 39 m above ground level (AGL) at each tower). We recorded a total of 4,816 bat calls from all detectors at all met tower locations combined from 26 July through 11 November, an average of 8.9 calls per tower per night. No bat calls were recorded between 11 November and 20 December 2006. Bat activity generally was highest immediately after sunset and declined through the night until just before sunrise the following morning. High (>35 kHz, mostly Myotis species and red bats [Lasiurus borealis]) and low (<35 kHz, mostly hoary bats [Lasiurus cinereus] and big brown bats [Eptesicus fuscus]) frequency-emitting echolocating bats tended to fly at different heights on the study area. We estimated that activity rates for bats using high frequencies were 1.5–4.0 times higher than those using low frequencies at low altitudes (10 m). This trend was reversed at medium altitudes (31 m) where it was estimated that activity rates for bats using low frequencies were 5.75–22.9 times greater than those for high frequency users, and at 39 m AGL it was estimated that activity rates for bats using low frequencies were 11.2–38.8 times greater than those for high frequency users.
The best model of bat activity in relation to weather variables included linear effects of temperature and wind speed and their interaction, the quadratic effect of wind speed, and the interaction of temperature with high versus low frequency echolocating bat groups. Bat activity was strongly related to temperature, but the effect differed for high and low frequency groups; high frequency users responded more to temperature than did low frequency users. For every 1o C increase in temperature, bat activity increased 14–57% for high frequency users and 5–34% for those using low frequencies. Bat activity increased slightly with increasing wind speed, but then decreased at higher wind speeds. Wind speeds at which maximum bat activity was predicted changed with temperature. In general, when temperatures were warm (>20ºC) and wind speeds were moderate (about 8 m/s), the predicted number of passes on any night was low, except for high frequency using bats.
This study was conducted at one proposed wind energy facility located on a forested ridge, so statistical inferences are limited to this site. However, our findings may reflect patterns of bat activity on similar forested ridges with comparable vegetation composition and topography in Massachusetts and surrounding states. Due to logistical constraints we were unable to gather activity data for the entire period of the year when bats are known to be active in this region (generally April through November), but our sampling did encompass the fall migration period when bats have been most frequently reported killed at wind facilities. Analyses presented in this report are exploratory, in part because so little data exist upon which to develop a priori, confirmatory hypotheses and associated candidate models. The current analysis only estimates activity rates and differences in activity patterns of two species groups (high and low frequency). Development of species-specific models is anticipated to occur in future years of the study. A second year of pre- construction acoustic monitoring was initiated in June 2007 and will continue through November 2007, and a third year of pre-construction acoustic data may be collected if the project is not constructed in 2008. After turbines are constructed, we anticipate gathering two consecutive years of post-construction activity and fatality data from April through November each year.