237
for each test intensity and frequency to facilitate quicker generation of the entire
audiogram data while the single neural recording was stable. Sound pressure levels
were calibrated with a B&K hydrophone as described above for AEPs.
Thresholds were determined for each test frequency by beginning with a suprathreshold intensity followed by decreasing intensities in 5 dB increments until the
neuron no longer responded to the stimulus. Threshold was defi ned as the lowest
intensity to produce a Rayleigh statistic, or Z value, of ≥4.5 (Lu and Fay 1993 ;
Batschelet 1981 ). The Z value measures the signifi cance of phase-locking and is
defi ned as R
2 × N , where N is the total number of action potentials sampled, and R is
the synchronization coeffi cient, or vector strength calculated according to (Goldberg
and Brown 1969 ). The degree of phase-locking is generally a good predictor of
auditory frequency encoding among vertebrates for low frequency systems (≤1
kHz) (Fay 1978b ; Javel and Mott 1988 ; Sisneros and Bass 2003 ).
The four different recording locations compared in this study are depicted in
Fig. 2b (saccular potential, AEP, hindbrain and midbrain single neurons), and all
experiments used the same underwater speaker positioned beneath the fi sh as a
stimulus. While the Hawaiian sergeant fi sh is likely most sensitive to particle motion
rather than sound pressure, due to technical limitations and for comparisons to other
studies, we only characterized the stimulus for all recordings in terms of sound pressure levels (dB re: 1 μPa) measured and calibrated in the experimental tanks with a
hydrophone. We agree, however, that future studies on fi sh hearing should attempt
to measure both sound pressure and particle motion in their experimental setups
whenever possible as recently suggested by Popper and Fay ( 2011 ). This information would allow for better interpretation of auditory capabilities in biologically
relevant contexts, as recent work shows differences in threshold curves expressed
in terms of pressure versus particle motion primarily for species with special adaptations to transfer pressure fl uctuations from the swim bladder to the inner ear
(Horodysky et al. 2008 ; Wysocki et al. 2009 ; Radford et al. 2012 ).
3 Results
3.1 Saccular Potential Recordings
Similar to previous studies (Fay 1974 ; Fay and Popper 1974 ; Sisneros 2007 ), saccular potentials from the Hawaiian sergeant fi sh were evoked maximally at twice
the stimulus frequency rather than at the same stimulus frequency (Fig. 3 ). This
double frequency effect is due to hair cell populations with opposite orientations
and is also dependent on the nonlinearity of the saccular potential such that the
cancellation of two sinusoidal waveforms 180° out of phase with each other is
avoided (Fay 1974 ). Best frequency was defi ned as the frequency that evoked the
saccular potential with the lowest threshold and ranged from 109 to 124 dB re: 1
μPa at 75 Hz (the lowest frequency tested) for all individuals tested. The majority of
saccular potential tuning curves showed lowest thresholds at this best frequency of
Comparison of Electrophysiological Auditory Measures in Fishes
for each test intensity and frequency to facilitate quicker generation of the entire
audiogram data while the single neural recording was stable. Sound pressure levels
were calibrated with a B&K hydrophone as described above for AEPs.
Thresholds were determined for each test frequency by beginning with a suprathreshold intensity followed by decreasing intensities in 5 dB increments until the
neuron no longer responded to the stimulus. Threshold was defi ned as the lowest
intensity to produce a Rayleigh statistic, or Z value, of ≥4.5 (Lu and Fay 1993 ;
Batschelet 1981 ). The Z value measures the signifi cance of phase-locking and is
defi ned as R
2 × N , where N is the total number of action potentials sampled, and R is
the synchronization coeffi cient, or vector strength calculated according to (Goldberg
and Brown 1969 ). The degree of phase-locking is generally a good predictor of
auditory frequency encoding among vertebrates for low frequency systems (≤1
kHz) (Fay 1978b ; Javel and Mott 1988 ; Sisneros and Bass 2003 ).
The four different recording locations compared in this study are depicted in
Fig. 2b (saccular potential, AEP, hindbrain and midbrain single neurons), and all
experiments used the same underwater speaker positioned beneath the fi sh as a
stimulus. While the Hawaiian sergeant fi sh is likely most sensitive to particle motion
rather than sound pressure, due to technical limitations and for comparisons to other
studies, we only characterized the stimulus for all recordings in terms of sound pressure levels (dB re: 1 μPa) measured and calibrated in the experimental tanks with a
hydrophone. We agree, however, that future studies on fi sh hearing should attempt
to measure both sound pressure and particle motion in their experimental setups
whenever possible as recently suggested by Popper and Fay ( 2011 ). This information would allow for better interpretation of auditory capabilities in biologically
relevant contexts, as recent work shows differences in threshold curves expressed
in terms of pressure versus particle motion primarily for species with special adaptations to transfer pressure fl uctuations from the swim bladder to the inner ear
(Horodysky et al. 2008 ; Wysocki et al. 2009 ; Radford et al. 2012 ).
3 Results
3.1 Saccular Potential Recordings
Similar to previous studies (Fay 1974 ; Fay and Popper 1974 ; Sisneros 2007 ), saccular potentials from the Hawaiian sergeant fi sh were evoked maximally at twice
the stimulus frequency rather than at the same stimulus frequency (Fig. 3 ). This
double frequency effect is due to hair cell populations with opposite orientations
and is also dependent on the nonlinearity of the saccular potential such that the
cancellation of two sinusoidal waveforms 180° out of phase with each other is
avoided (Fay 1974 ). Best frequency was defi ned as the frequency that evoked the
saccular potential with the lowest threshold and ranged from 109 to 124 dB re: 1
μPa at 75 Hz (the lowest frequency tested) for all individuals tested. The majority of
saccular potential tuning curves showed lowest thresholds at this best frequency of
Comparison of Electrophysiological Auditory Measures in Fishes