Qualitative Identification of Free-flying Bats Using the Anabat Detector

A variety of ultrasonic bat detectors have been used over the past 3 decades to identify freeflying bats. Analyses of recorded echolocation calls were slow and typically restricted to few calls and at a resolution obscuring details of call structure. The Anabat II detector and associated zerocrossings analysis system allows an immediate examination, via a laptop computer, of the timefrequency structure of calls as they are detected. These calls can be stored on the hard drive for later examination, editing, and measurement. Many North American bats can be identified to species by qualitatively using certain structural characteristics of calls, primarily approximate maximum and minimum frequencies and morphological aspects of calls e.g., linearity and changes in slope). To identify calls precisely, it is important to use a continuous sequence of calls from an individual in normal flight rather than from single isolated calls. All calls are not equally useful, and many fragmentary calls must be discarded before making a determination. Each sequence of calls must be examined to ensure that multiple bats have not been simultaneously recorded, which confounds correct identification. We found the percentage of nonusable calls within usable vocal sequences to be highest in vespertilionids 20-40%), whereas for other families this was frequently <10%. Active rather than passive collection of data maximizes quality and quantity of diagnostic calls and provides a contextual base for the investigator.