Summary
We have compared the echolocation and feeding behaviours of Myotis lucifugus, M. californicus, M. volans, and M. auriculus based on observations and recordings of bats in the field. Myotis lucifugus and M. californicus appeared to detect prey at close range (≤ 1 m) and regularly made several attempts to capture insects over short distances; both used similar frequency-modulated echolocation calls. Myotis volans detected prey at greater distances (5–10 m), made only one attempt to capture insects per pass through a feeding area, and used an echolocation call with a distinct constant-frequency component. Myotis auriculus fed mainly on resting insects, mostly moths. The echolocation calls of this species were of shorter duration, lower intensity, broader frequency range with a higher frequency of maximum energy, and showed an initial upward sweep in frequency relative to the calls of the other Myotis we studied. Myotis auriculus did not increase their pulse repetition rate as they closed with stationary prey, and they appeared to fix on resting insects from about 2 m. This species rarely made more than one attempt to capture an insect per pass through a feeding area.
Methodology
We observed bats as they hunted insects in a variety of locations, sometimes with the aid of a night-vision scope or low light level television camera. Their echolocation calls were monitored using a broadband microphone as described by Simmons, Fenton, Ferguson et al. (1979), or with a QMC S100 bat detector, in either case coupled to a zero-crossing period meter which provided an oscilloscope display of the fundamental frequency-time structures of the echolocation calls (Simmons, Fenton, Ferguson et 01. 1979). At both locations some bats were light tagged using modifications of the technique described by Buchler (1976u), and (or) marked with reflective celluloid split bands (A. C. Hughes, England). Observations of the echolocation calls (via the period meter) of marked bats permitted us to recognize most species in an area by their orientation calls (Fenton and Thomas 1979). In this report we consider a 'complete' capture to represent an observation of the entire capture sequence of an insect by a bat, from before the bat had started to directly approach its target. An 'incomplete capture' observation involved our seeing only the capture of the insect by the bat. By observing 'complete' captures we were able to estimate the distance from which the bat homed in on its target. We use this distance as an indication of the range of detection of the insect by the bat and realize that it is a minimum value.
Echolocation calls were recorded using a broad-band microphone and a Lockheed Store 4D tape recorder operated at 76 cmlii. Far each species at least 20 approaches and attempts to catch insects- were recurded Tor analysis {rarely did these recnrdinpr include 'complete' capture sequences). During analysis. tapes were slowed by 8 to 16 times and played into a Princeton Applied Research model 4513 FFT (fast Fourier transform) real time spectrum analyzer (2048 lines, alias filter in, flat weighting) operated in a 'transient capture' mode, and a Kay sonograph model 7029A