199
and others examined. Popper (1981) proposed a broad classification system for
saccular hair cell orientation “patterns,” and Coombs and Popper (1979) suggested
a functional hypothesis for some of the diversity, based on their studies of Hawaiian
squirrelfish species with different audiograms and different hair cell orientation patterns. Although that hypothesis has not proven to be a valid generalization for all
fishes, those studies led to major questions about the functional significance of the
variations observed among the saccules of teleosts (Popper and Coombs 1982).
Work by Flock (1971) and Hudspeth and Corey (1977) on the physiological polarity
of vertebrate hair cells provided the basis for a variety of hypotheses on the potential
role that hair cell orientations might play in encoding the direction of a sound source
(e.g., Moulton and Dixon 1967; Schuijf 1975, 1976; Saidel and Popper 1983;
Schellart and deMunck 1987; Rogers and Zeddies 2008). Documented variations in
the apical structures (a single kinocilium and a stair-step array of stereovillae) on
individual hair cells also stimulated research on frequency response and the potential for regional response differences along the saccular epithelium (e.g., Furukawa
and Ishii 1967; Sugihara and Furukawa 1989).
Around that time, Fay (1984) conducted a landmark study that introduced the
field of fish bioacoustics to his three-dimensional particle motion stimulus system,
which allowed him to collect the data that proved all three otolithic endorgans of the
goldfish (Carassius auratus) had overlapping frequency responses. In addition,
each endorgan encoded the axis of particle motion as predicted, based on the orientation of the endorgan and the orientations of the hair cells on each sensory epithelium. Fay went on to study many aspects of the sense of hearing in goldfish by
classical conditioning (respiration or heart rate) and various psychophysical procedures (see Fay 1988 for details), first obtaining a response to a stimulus the fish
could easily detect and then determining whether the fish could detect another stimulus with a difference in frequency content, intensity, or temporal parameters, or a
stimulus in the presence of various types of “noise.” The reader is referred to Fay
(2014), wherein he summarized his contributions to our understanding of the sense
of hearing in goldfish, especially as it compares to the sense of hearing in humans.
It is important to note here that the goldfish is a nonvocal teleost fish that nonetheless has specializations of the auditory pathway (Weberian ossicles) that provide
increased sensitivity to higher frequencies (e.g., 500 Hz to about 5000 Hz) than the
majority of fish species investigated to date, which lack similar specializations and
hear only lower frequencies (e.g., below 500 Hz; Popper and Fay 1999). The importance of Fay’s work with goldfish was in revealing what the goldfish “knows” about
the sounds (or noise) around it, including the concept of “auditory scene” analyses
(introduced by Bregman 1990) or the “soundscape,” which Fay has often addressed
in his papers and presentations (Fay 2009, 2014).
As a beginning graduate student in the Popper lab, I was struck by the volume of
information known about auditory processing in the non-vocal goldfish and the relative lack of information about auditory processing in vocal fish, particularly with
regard to sound source localization. The oyster toadfish, Opsanus tau (Linnaeus
1766) was my first marine research subject as an undergraduate working in the lab
of the famous “shark lady,” Eugenie Clark, where I studied their shelter-seeking
What the Toadfish Ear Tells the Toadfish Brain About Sound
and others examined. Popper (1981) proposed a broad classification system for
saccular hair cell orientation “patterns,” and Coombs and Popper (1979) suggested
a functional hypothesis for some of the diversity, based on their studies of Hawaiian
squirrelfish species with different audiograms and different hair cell orientation patterns. Although that hypothesis has not proven to be a valid generalization for all
fishes, those studies led to major questions about the functional significance of the
variations observed among the saccules of teleosts (Popper and Coombs 1982).
Work by Flock (1971) and Hudspeth and Corey (1977) on the physiological polarity
of vertebrate hair cells provided the basis for a variety of hypotheses on the potential
role that hair cell orientations might play in encoding the direction of a sound source
(e.g., Moulton and Dixon 1967; Schuijf 1975, 1976; Saidel and Popper 1983;
Schellart and deMunck 1987; Rogers and Zeddies 2008). Documented variations in
the apical structures (a single kinocilium and a stair-step array of stereovillae) on
individual hair cells also stimulated research on frequency response and the potential for regional response differences along the saccular epithelium (e.g., Furukawa
and Ishii 1967; Sugihara and Furukawa 1989).
Around that time, Fay (1984) conducted a landmark study that introduced the
field of fish bioacoustics to his three-dimensional particle motion stimulus system,
which allowed him to collect the data that proved all three otolithic endorgans of the
goldfish (Carassius auratus) had overlapping frequency responses. In addition,
each endorgan encoded the axis of particle motion as predicted, based on the orientation of the endorgan and the orientations of the hair cells on each sensory epithelium. Fay went on to study many aspects of the sense of hearing in goldfish by
classical conditioning (respiration or heart rate) and various psychophysical procedures (see Fay 1988 for details), first obtaining a response to a stimulus the fish
could easily detect and then determining whether the fish could detect another stimulus with a difference in frequency content, intensity, or temporal parameters, or a
stimulus in the presence of various types of “noise.” The reader is referred to Fay
(2014), wherein he summarized his contributions to our understanding of the sense
of hearing in goldfish, especially as it compares to the sense of hearing in humans.
It is important to note here that the goldfish is a nonvocal teleost fish that nonetheless has specializations of the auditory pathway (Weberian ossicles) that provide
increased sensitivity to higher frequencies (e.g., 500 Hz to about 5000 Hz) than the
majority of fish species investigated to date, which lack similar specializations and
hear only lower frequencies (e.g., below 500 Hz; Popper and Fay 1999). The importance of Fay’s work with goldfish was in revealing what the goldfish “knows” about
the sounds (or noise) around it, including the concept of “auditory scene” analyses
(introduced by Bregman 1990) or the “soundscape,” which Fay has often addressed
in his papers and presentations (Fay 2009, 2014).
As a beginning graduate student in the Popper lab, I was struck by the volume of
information known about auditory processing in the non-vocal goldfish and the relative lack of information about auditory processing in vocal fish, particularly with
regard to sound source localization. The oyster toadfish, Opsanus tau (Linnaeus
1766) was my first marine research subject as an undergraduate working in the lab
of the famous “shark lady,” Eugenie Clark, where I studied their shelter-seeking
What the Toadfish Ear Tells the Toadfish Brain About Sound