Quantitative Method for Selection of Identifiable Search-phase Calls Using the Anabat System.

Abstract

The authors investigated the use of a new filter in the analysis program 'Analook' to quantify selection of identifiable search-phase calls of bats in the eastern United States. This filter is a mathematical function that served to remove unwanted material from a recorded Anabat file. They tested the effectiveness of the new filter using calls of four bat species: Myotis lucifugus, M. septentrionalis, Pipistrellus subflavus, and Eptesicus fuscus. Each Anabat file was cleaned manually by two researchers and automatically by the Analook filter. For all four species, total number of pulses selected per sequence by the filter and the authors varied considerably. This variation indicates that the filter selected search-phase pulses somewhat differently than the two biologists. The relatively small amount of variation imposed by the use of the filter suggests that the filter would provide suitable cleaning for the process of acoustic identification.