Bat detectors are an important tool for ecological studies of bats. However
, quality and quantity of data may be affected by the devices used to recor
d the output from the detector. We compared recordings of bat activity from
audiocassette recorders and computers. Numbers of calls/hour, passes/hour,
identifiable passes/hour, and feeding buzzes/hour were similar (all Ps >0.
1) among recording devices. All call characteristics, except for the minimu
m frequency and characteristic frequency, differed (P <0.05) between tapes
and computers. Species identification using discriminate function analysis
was less reliable with tape data than with computer data, particularly when
the model built with computer-recorded reference calls was tested with tap
e-recorded calls. Therefore, we suggest that when tape recorders are used f
or field recording, they also are used to record reference calls.