Do Striped Cusk-Eels Ophidion marginatum (Ophidiidae) Produce the “Chatter” Sound Attributed to Weakfish Cynoscion regalis (Sciaenidae)?

Weakfish Cynoscion regalis and striped cusk-eels Ophidion marginatum both produce sounds, but there has been confusion in the literature on a particular sound, the “chatter.” It has been stated that this sound is produced by weakfish using their pharyngeal teeth. Striped cusk-eels make a similar sound (but not identified in the literature as a chatter) using sonic muscles associated with the swim bladder and vertebral components. The striped cusk-eel identifications were based on captive fish sound recordings, whereas the weakfish identifications were based on recordings made in situ where weakfish were visually observed but other sound-producing organisms could have been present. Based on new signal analysis of striped cusk-eel sounds made in captivity, we identify that species as the source of the chatter sound in our field recordings. The dominant frequency of the sounds increased while the pulse period decreased over the temperature range 18.0–27.5 C. The acoustic characteristics presented here will aid researchers in their identification of these sounds.

Fig. 1.

(A) Oscillogram of chatters recorded at 1810 on 25 August 1997 near Hatteras Inlet in Pamlico Sound, North Carolina (35°11′33.758″N, 75°46′42.129″W). (B) Oscillogram produced by a captive striped cusk-eel recorded in air by Mann et al. (1997). (C) A sonogram of same sound as in (A) using 1024-point Hanning windowed Fast Fourier Transforms with 7/8 overlap; power spectral densities (PSDs) on a relative scale with 0 dB set to background level. (D) Sonogram of same sound as in (B) produced using the same techniques as in (C)

Fig. 2.

Oscillograms showing single pulses. (A) Oscillogram of pulse in the field recorded chatter shown in Figure 1A. (B) Oscillogram of pulse in the captive striped cusk-eel sound shown in Figure 1B

Fig. 3.

Average power spectra of the long calls from Figure 1 produced using consecutive 1024-point Hanning windowed Fast Fourier Transforms. The power spectral densities (PSDs) are on a relative scale with 0 dB set to the background level. (A) Average power spectrum of the same field recorded chatter as in Figure 1A. The low-frequency components in this sonogram were a result of 60 Hz noise and were not part of the fish sounds. (B) Average power spectrum of the same captive striped cusk-eel sound as in Figure 1B

Fig. 4.

Regression analyses of sound characteristics of field-recorded chatter sounds as a function of temperature. n = 81 pulses. (A) Number of pulses versus temperature. (B) Pulse period versus temperature. (C) Dominant frequency versus temperature

Accepted: 05 Jan 2001