Identifying risks to migratory birds and bats from wind development

Principal Investigator: Jim Perry.

Co-Principal Investigator: Doug Johnson.

Student: Kevin Heist.

Status: Research Work Order established and Ph.D. graduate student selected. The study was designed over the winter of 2009-2010. Data collection was completed over six field seasons (spring 2010, fall 2010, spring 2011, fall 2011, spring 2012, and fall 2012). Data analysis (phases 1 and 2), statistical analyses, and dissertation drafting were completed during 2013 and early 2014. Dissertation was successfully defended June 2014, submission of final dissertation and Kevin’s graduation occurred August 2014. One manuscript (focused on bat activity in the Great Lakes) has been completed and is currently being submitted to journals for publication, two others (focused on the relationship between acoustic bat activity and fatality rates, and automated species identification) are being written. 

Justification: Wind energy development is occurring at a rapid pace and is expected to increase dramatically under the U.S. objective of producing 20% of the Nation’s energy from wind by 2030. Although wind provides a renewable source of energy, concerns exist about the effects on wildlife, particularly migratory birds and bats. Migratory birds and any endangered bats are trust species of the federal government, and any “take” of such animals are of concern. The federal government has also made extensive investments in refuges, waterfowl productions areas, and wetland and grassland easements, primarily for the protection and production of migratory birds. It is important to understand the extent to which wildlife values associated with these investments may be compromised by wind energy development.

Objectives: The goal of this study was to explore methods to assess risks to migratory birds and bats posed by wind energy development at a local level. Specific project objectives included:

  1. Evaluate the ability of dual acoustic-ultrasonic recorder to capture nocturnal calls of birds and bats at current and potential wind power sites.
  2. Evaluate the ability of an automated species identification program to accurately identify recorded bat calls.
  3. Relate nocturnal call activity for birds and bats to results of fatality searches at operating wind farms. Our hypothesis is that the number of fatalities found per search interval will be correlated with the number of calls recorded during that search interval.
  4. Evaluate whether call activity varies in relation to prominent landscape features. We will test the hypothesis that nightly call counts vary in predictable relations to specified physiographic and landscape features.
  5. Measure changes in bird and bat activity before and then after a wind farm is constructed. We will test the hypothesis that the development of a wind farm does not affect activity patterns at the site by recording call rates in the same locations before and after a wind farm is constructed.
  6. Examine whether bird and/or bat activity varies with distance from turbines. We will test the hypothesis that bird and bat activity does not depend on proximity to a turbine by recording call rates at predetermined distances from the base of turbines.

Progress: Field sites were selected in particular locations to address these objectives. To address Objective 3, recorders were placed at wind farms where fatality searches were being conducted (independently of this study). To address Objective 4, recorders were set up at various distances from physiographic and landscape features that may influence bird and bat abundance during migration, including riparian corridors and grasslands under federal or state protection. To address Objective 5, recorders were located at sites where wind farm development was anticipated within the following two years, and are returned to the same locations after the wind farm is built. To address Objective 6, recorders were placed at specified distances from the base of wind turbines. Data from all sites were used to address Objectives 1 and 3. Most field sites were located in the Upper Midwest and Great Lakes regions, but wind farms in Pennsylvania and Texas also were included.

Six seasons of data collection have been completed, and over 2.8 million bird and bat passes have been counted. Data processing and analysis, including species identification and statistical analyses have been completed. Dissertation writing and defense have been completed. Current activities include manuscript writing and presentations of final results.

Results Overview: Recorders worked well in the field, and the automated bat call identification had mixed results, both among species and geographic regions, with low call detection rates and highly varied accuracy rates. Data did not show a strong relationship between call rates and fatality rates for either birds or bats. We found a consistent landscape-dependent pattern at recording sites near river corridors, with high bat activity near forested edges (within 500 meters) and decreasing activity with distance from edge. Call data also indicate large differences in bird and bat activity among sites at broad regional scales, with high activity levels near Great Lakes shorelines and less activity at interior sites. Recordings at one newly-constructed wind farm indicated lower levels of bird activity after construction, compared to levels prior to construction; however only one year of recordings at a single post-construction site was available for this objective. Recordings at varying distances from turbines have not indicated any consistent pattern of elevated bat activity near turbines.