139
motion (“vector detection”) by populations of differently oriented hair cells. The
second component is a resolution of “the 180° ambiguity” problem, i.e. disambiguation of the inherent ambiguity in the directionality of an oscillating particle. Schuijf
(1975) conceived of a clever solution to this problem that entailed determining the
phase angle between particle motion and sound pressure. This “phase model”
requires that both the sound pressure and particle motion waveforms be encoded by
the inner ear, and that appropriate central computations take place, exploiting the
relative phase or timing relations to extract the true direction of incidence. While we
ourselves have provided some empirical support for a role of pressure sensitivity via
the swim bladder in midshipman localization (e.g., Coffin et al. 2014), the phase
model (and pressure sensitivity) does not sufficiently explain sound localization in
all scenarios. For example, particle motion–pressure relationships are easily predicted for monopolar sound sources (e.g., an expanding and contracting gas bubble), but many underwater sound sources are dipolar at least. Dipolar sources (e.g.,
a vibrating sphere with a constant volume) produce complex sound fields in which
the particle motion vectors do not always point toward and away from the sound
source, and the particle motion–pressure relationship does not reliably correspond
to any one direction. A more serious empirical challenge to the phase model is that
sharks and other elasmobranchs are able to locate sound sources from relatively far
distances (Nelson and Gruber 1963; Nelson and Johnson 1972) despite their probable lack of pressure sensitivity (elasmobranchs lack a swim bladder or other inner
ear-associated gas bubble). Thus, sharks are apparently able to localize sound on the
basis of acoustic particle motion alone. Another difficulty with the phase model is
that it requires the use of sinusoidal signals while broadband signals such the clicks
transmitted by cod and haddock, which are far more common than sinusoidal signals in nature, cannot be used with the phase model.
Approximately over the past three decades, models of directional hearing in fish
have proliferated, often focusing on improved solutions to the 180° ambiguity problem. Recent models have included (1) an “orbital” model by Schellart and de Munck
(1987; but also see de Munck and Schellart 1987) in which sound pressure and
particle motion together cause the otolith orbits to rotate either clockwise or counterclockwise depending on whether the source is to the left or right, (2) a computational model by Rogers et al. (1988) that also requires both pressure and particle
motion sensitivity, (3) a more algorithmic approach proposed by Kalmijn (1997) in
which fish maintain a constant angle with respect to the axis of vibration even if the
axis of vibration does not point toward the source, and (4) a “multipole” model by
Rogers and Zeddies (2008) which applies the theory of multipole sensors to the fish
ear, specifically the concept of an uncovered hair cell with no overlying otolith that
responds to sound as a lateral quadrupole capable of resolving the 180° ambiguity
when coupled to otolith-covered hair cells that act as dipole detection mechanisms.
Interestingly, the model by Rogers and Zeddies (2008) is one of the few that
addresses how the 180° ambiguity might be resolved for fish without a gas bubble,
but this may only be possible in the far field.
All of the phase models of sound localization that have been proposed (Schuijf
1975; Schellart and de Munck 1987; Rogers et al. 1988) require that a fish be able
Directional Hearing and Sound Source Localization in Fishes
motion (“vector detection”) by populations of differently oriented hair cells. The
second component is a resolution of “the 180° ambiguity” problem, i.e. disambiguation of the inherent ambiguity in the directionality of an oscillating particle. Schuijf
(1975) conceived of a clever solution to this problem that entailed determining the
phase angle between particle motion and sound pressure. This “phase model”
requires that both the sound pressure and particle motion waveforms be encoded by
the inner ear, and that appropriate central computations take place, exploiting the
relative phase or timing relations to extract the true direction of incidence. While we
ourselves have provided some empirical support for a role of pressure sensitivity via
the swim bladder in midshipman localization (e.g., Coffin et al. 2014), the phase
model (and pressure sensitivity) does not sufficiently explain sound localization in
all scenarios. For example, particle motion–pressure relationships are easily predicted for monopolar sound sources (e.g., an expanding and contracting gas bubble), but many underwater sound sources are dipolar at least. Dipolar sources (e.g.,
a vibrating sphere with a constant volume) produce complex sound fields in which
the particle motion vectors do not always point toward and away from the sound
source, and the particle motion–pressure relationship does not reliably correspond
to any one direction. A more serious empirical challenge to the phase model is that
sharks and other elasmobranchs are able to locate sound sources from relatively far
distances (Nelson and Gruber 1963; Nelson and Johnson 1972) despite their probable lack of pressure sensitivity (elasmobranchs lack a swim bladder or other inner
ear-associated gas bubble). Thus, sharks are apparently able to localize sound on the
basis of acoustic particle motion alone. Another difficulty with the phase model is
that it requires the use of sinusoidal signals while broadband signals such the clicks
transmitted by cod and haddock, which are far more common than sinusoidal signals in nature, cannot be used with the phase model.
Approximately over the past three decades, models of directional hearing in fish
have proliferated, often focusing on improved solutions to the 180° ambiguity problem. Recent models have included (1) an “orbital” model by Schellart and de Munck
(1987; but also see de Munck and Schellart 1987) in which sound pressure and
particle motion together cause the otolith orbits to rotate either clockwise or counterclockwise depending on whether the source is to the left or right, (2) a computational model by Rogers et al. (1988) that also requires both pressure and particle
motion sensitivity, (3) a more algorithmic approach proposed by Kalmijn (1997) in
which fish maintain a constant angle with respect to the axis of vibration even if the
axis of vibration does not point toward the source, and (4) a “multipole” model by
Rogers and Zeddies (2008) which applies the theory of multipole sensors to the fish
ear, specifically the concept of an uncovered hair cell with no overlying otolith that
responds to sound as a lateral quadrupole capable of resolving the 180° ambiguity
when coupled to otolith-covered hair cells that act as dipole detection mechanisms.
Interestingly, the model by Rogers and Zeddies (2008) is one of the few that
addresses how the 180° ambiguity might be resolved for fish without a gas bubble,
but this may only be possible in the far field.
All of the phase models of sound localization that have been proposed (Schuijf
1975; Schellart and de Munck 1987; Rogers et al. 1988) require that a fish be able
Directional Hearing and Sound Source Localization in Fishes