Bat population censusing with passive acoustics

Abstract

Passive acoustic monitoring is a widely used method to identify bat species and determine spatial and temporal activity patterns. One area where acoustic methods have not yet been successfully applied, however, is in determining population counts, especially from roosts. Typically, most roost counts are obtained with thermal imagery that may be prohibitively expensive for many natural resource managers or require complex computer programming. Here, we demonstrate a new acoustic technique to estimate population size of Brazilian free-tailed bats (Tadarida brasiliensis) emerging from large cave colonies. Data were acquired across multiple nights and at 9 cave locations with different roost structures and flight behavior profiles. We used a single microphone to monitor echolocation activity and simultaneously recorded the emerging bats with thermal video. Bat abundance counts were determined from a single video frame analysis (every 10 s) and were compared to different acoustic energy measures of a 1-s long acoustic sequence recorded at the time of the analyzed video frame. For most cave locations, linear regression models successfully predicted bat emergence count based on acoustic intensity of the emerging stream. Here, we describe our method and report on its application for counting bats from different roost locations.