Abstract
The Tasmanian wedge-tailed eagle (WTE, Aquila audax fleayi) and the white-bellied sea-eagle (WBSE, Haliaeetus leucogaster) are present on the Bluff Point (37 Vestas V66 turbines) and Studland Bay (25 Vestas V90 turbines) Wind Farms in north-west Tasmania, Australia. These species have been intensively studied since the commencement of operations in 2002 and 2007, respectively, as part of compliance monitoring. Monitoring has included documenting collisions with turbines, breeding success surveys, and movement and behaviour studies. Additional investigations (outside regulatory requirements) have also been conducted, including targeted studies and trials of collision mitigation techniques. Both species of eagle have continued to use the sites during construction and operation of the wind farms. The average collision rates for WTE were 1.54 and 0.95 per year, and for WBSE 0.36 and 0 per year at Bluff Point and Studland Bay, respectively (calculated up to October 2012). These are below maximum rates estimated in collision risk modeling which formed part of the information for the assessment of the wind farms. The collision rate for WTE was constant across years, although there was some evidence the rate could be declining at Studland Bay. Analyses could not be conducted on WBSE due to small sample sizes. Seasonal and other temporal patterns were tested for in the collision data, but all evidence supported the theory that the strikes were independent and random in time, with no support found for some proposed theories about why eagles collide with turbines. A spatial analysis of collisions was not possible, again due to small sample sizes. Eagles continued to breed at the sites, with at least the same level of success as nests outside the wind farms. The observational studies provided useful data about how eagles interacted with turbines at these sites. These data were used to calculate turbine avoidance rates and to assess how rates changed with development of the wind farm and when turbines were operational or not.