212
1997a), and that those afferents could function as simple sound detectors.
Alternatively, those cells may reflect a step in the maturation of the synaptic
connections on the sensory epithelium. New hair cells and new connections develop
as the endorgan grows throughout the life of the fish; some of those connections
may be temporary, particularly if correlated activity is favored for maintaining synaptic contacts between an afferent and multiple hair cells.
Thresholds among saccular afferents varied between 300 and 0.1 nm rms, similar
to saccular afferents of the goldfish (Fay and Ream 1986; Fay and Edds-Walton
2000). These data were important because they showed that the otophysic connection (mechanically connecting the ear and gas bladder) in goldfish does not provide
substantial improvement in sensitivity to particle motion at lower frequencies
(below 200 Hz) when compared to the toadfish, which lacks the otophysic connection. The most sensitive afferents (sensitivity is the inverse of threshold) rival mammalian cochlear afferents. Also, similarly to mammalian cochlear afferents, toadfish
afferents with low spontaneous (or background) activity tended to have higher
thresholds than those with higher spontaneous activity, though there was a continuum (not a dichotomy) of responses in all three locations investigated along the
saccule (Fig. 5 in Fay and Edds-Walton 1997a).
Additionally, a subset of afferents located all along the saccule, with 0–2 spikes/s
spontaneous activity, exhibited consistent level-dependent phase shifts with increasing stimulus levels (mean slope ± s.e.m.: 3.7°/dB ± 0.16°). Although the phase shift
may seem inconsequential, for a 5dB difference in stimulus level, there could be a
20 deg shift in the occurrence of a phase-locked spike (Fig. 4). For a biologically
relevant frequency such as 100 Hz, a 5 dB difference would translate to a 55 ms
temporal difference in the spiking activity of an afferent that exhibits phase-advance
(Fig. 4d) versus a phase-locked afferent that does not exhibit phase-advance (Fig.
4b). If we look at an octaval cell that receives input from both of the above cells, the
phase-advanced afferent representing the louder sound would provide its “information” earlier to a cell that receives both inputs. Taken together, the directional
response data and phase-advance data provide evidence that computations of interaural level differences are possible, and a binaural comparison of the activity from
the nonparallel saccules could contribute to sound source localization. These results
led to the investigation of response characteristics within the medullary nucleus of
the octaval column (Fig. 2b) that receives the majority of input from the saccule: the
dorsal division of the descending octaval nucleus (dDON).
3.4 Physiology: What the Brain “Knows”
A key question to address is what the brain does with the information from the ear.
Our studies of the dDON in the medulla and its midbrain target, nucleus centralis
(NC), revealed that the auditory circuits perform various computations that
“improve” the information about frequency, temporal pattern, and the axis of particle motion of a sound source from the information provided by the auditory endorgans of the ear. At this point, we cannot state which endorgans of the ear contribute
P.L. Edds-Walton
1997a), and that those afferents could function as simple sound detectors.
Alternatively, those cells may reflect a step in the maturation of the synaptic
connections on the sensory epithelium. New hair cells and new connections develop
as the endorgan grows throughout the life of the fish; some of those connections
may be temporary, particularly if correlated activity is favored for maintaining synaptic contacts between an afferent and multiple hair cells.
Thresholds among saccular afferents varied between 300 and 0.1 nm rms, similar
to saccular afferents of the goldfish (Fay and Ream 1986; Fay and Edds-Walton
2000). These data were important because they showed that the otophysic connection (mechanically connecting the ear and gas bladder) in goldfish does not provide
substantial improvement in sensitivity to particle motion at lower frequencies
(below 200 Hz) when compared to the toadfish, which lacks the otophysic connection. The most sensitive afferents (sensitivity is the inverse of threshold) rival mammalian cochlear afferents. Also, similarly to mammalian cochlear afferents, toadfish
afferents with low spontaneous (or background) activity tended to have higher
thresholds than those with higher spontaneous activity, though there was a continuum (not a dichotomy) of responses in all three locations investigated along the
saccule (Fig. 5 in Fay and Edds-Walton 1997a).
Additionally, a subset of afferents located all along the saccule, with 0–2 spikes/s
spontaneous activity, exhibited consistent level-dependent phase shifts with increasing stimulus levels (mean slope ± s.e.m.: 3.7°/dB ± 0.16°). Although the phase shift
may seem inconsequential, for a 5dB difference in stimulus level, there could be a
20 deg shift in the occurrence of a phase-locked spike (Fig. 4). For a biologically
relevant frequency such as 100 Hz, a 5 dB difference would translate to a 55 ms
temporal difference in the spiking activity of an afferent that exhibits phase-advance
(Fig. 4d) versus a phase-locked afferent that does not exhibit phase-advance (Fig.
4b). If we look at an octaval cell that receives input from both of the above cells, the
phase-advanced afferent representing the louder sound would provide its “information” earlier to a cell that receives both inputs. Taken together, the directional
response data and phase-advance data provide evidence that computations of interaural level differences are possible, and a binaural comparison of the activity from
the nonparallel saccules could contribute to sound source localization. These results
led to the investigation of response characteristics within the medullary nucleus of
the octaval column (Fig. 2b) that receives the majority of input from the saccule: the
dorsal division of the descending octaval nucleus (dDON).
3.4 Physiology: What the Brain “Knows”
A key question to address is what the brain does with the information from the ear.
Our studies of the dDON in the medulla and its midbrain target, nucleus centralis
(NC), revealed that the auditory circuits perform various computations that
“improve” the information about frequency, temporal pattern, and the axis of particle motion of a sound source from the information provided by the auditory endorgans of the ear. At this point, we cannot state which endorgans of the ear contribute
P.L. Edds-Walton