An onboard acomtic data logger to record biosonar of free-ranging bottlenose dolphins. Douglas P. Nowacek*, Peter L. Tyack*, Randall S. Wells~, and Mark P. Johnson~ *Department of B iology and ~Departmenl of Applied Ocean Physics and Engineering, Woods Hole Oceanographic Institution, Woods Hole, MA 02543; tChicago biological Society, do Mole Marine Laboratory, Sarasota, FL 34236 Abstract: me ecology of the odontocete echolocation system is not well understood despite a solid understanding of the system’s operation. To gain insight into the functional uses of dolphin biosorrsr we have developed an acoustic data logger which utilizes a miniature DAT recorder and two suction-cup hydrophores. The first hydrophore is located 10 cm posterior of the blowhole, and (he second 20 cm below the lateral base of the dorssf fin. The anterior ‘high-frequency’ hydrophore, designed specifically to record echolocation signals, has unity gain and a one-pole 10kHz high pass filter. The ‘ambient’ hydrophore located at the base of the dorsal fin has +18 dB gain and has a one-pole I kHz high pass filter. To obtain echolocation recordings the ‘high-frequency’ hydrophore was filtered through a simple demodulator in one of the deployments, The package was attached to temporarily restrained animals which, after release, were followed to record behavioral data. During (he two successful deployments to date the logger recorded animal vocalizations, surfacing events, the sounds of passing boats, and hydrodynamic sounds produced by the animal’s fluke strokes. Odontocete cetaceans have been known for 37 years to use echolocation (Norris et af. 1961). The characteristics of the sonar system of the bottlenose dolphin, Tursiops truncatus, have been elucidated through intensive study of captive animals. Au (1993) reviews this research describing the dolphin transmission and receiving systems and documenting the chstracterislic acoustic features of the echolocalion signals. We know also from these studies lhat the dolphins’ sonar syslcm is excellent for target detection, discrimination, and classification and for range discrimination (Au 1993). While the performance of the dolphin echolocation system is well characterized, its functional uses by wild animals are not well understood, Recent studies have begun to elucidate some details of odontocete echolocation use in the conlext of foraging (Verfuss and Schni(zler 1995; Miller et al. 1995). These studies document changes in echolocation signals and use patterns as the animals move through a predation sequence, a phenomenon also seen in foraging microchiropteran bats (Schnitzler & Henson 1980; Kick & Simmons 1984). Bat research has successfully elucidated many of the operational and functional details of echolocalion. In addition to a good understanding of the performance (Schnitzler & Henson 1980) and neural processing (Dear& Suga 1995), the ecology of the bal echolocation system is much more fully understood than is the odontocete system (Neuweiler 1983; Surlykke 1988). In fact, few data exist which can address even basic questions: how do odontocetes use echolocation for navigation an~or foraging? Do patlerns of use change diurnally? DATA LOGGER DESCRIPTION AND R~ULTS One reason that odont(xete echoiocation research has not progressed as quickly as bat research is the difficulty in obtaining individually identified recordings of animals echolocating on biologically relevant targets. To procure such recordings we have developed an onboard acoustic data logger utilizing a two-channel DAT recorder housed in aluminum and attached to the dorsal fin with a Track Pack o (Figure 1). The recorder has a flat frequency response from 10 Hz-14 kHz, and each [ape can store 120 stereo-minutes. The first suction-cup hydrophore (sensitivity -205 dB re 1 @a) is located 10 cm posterior of the blowhole, and the second 20 cm blow the lateral base of the dorsal fin (Figure 1). The anterior ‘high-frequency’ hydrophore, designed specifically to record echolocation signals, has unity gain and a one-pole 10 kHz high pass filter. The ‘ambient’ hydrophore has +18 dB gain and has a one-pole 1 kHz high pass filter. To oblain echolocation recordings the ‘high-frequency’ hydrophore was frllered through a simple demodulator in one of the deployments. This frequency shift circuit is similar to a single-side-band demodulator, consisting of a high-pass filter (HP~ with passband edge at 70 kHz and a multiplier (implemented by an analog switch), modulating the filtered signal with a 70 kHz square wave. The result is that the 70-85 kHz band is shifted to the O-15 kHz band of the DAT recorder. The purpose of the HPF is to minimize distortion in the recorded signal due to aliasing of the 55-70 kHz band by the multiplier and the quality of the recording depends upon the stopband attenuation of the HPF. As the primary application of tie frequency-shifted recordings WaSto be in estimating click rates, a high degree of alias rejwtion was not required, Given this and the small volume available in the tag for circuitry, we found that a straightforward combination of tunable notch and 3-pole active high-pass filters was satisfactory. VHF radio transmitters were mounled 1409