sacs as food-storage devices (similar to the cheek pouches of Old World
monkeys, which play no acoustic role; Schön-Ybarra 1995). However, Fay
(1960) found this last suggestion unlikely, suggesting that Sleptsov’s specimens had regurgitated stomach contents into the sacs in their death throes.
In any case, the “bell production” and “life preserver” hypotheses are both
well-supported, indicating at least two functions for walrus pharyngeal sacs
and suggesting that air sacs may serve multiple functions in other species
as well.
The most commonly cited possible function for vocal air sacs is that they
play a role in impedance matching from the vocal tract to the atmosphere
(see below for more details). Such a role has been suggested for most
anurans as well as for the inflatable (soft-walled) air sacs of cercopithecid
monkeys such as guenons (Cercopithecus spp., Gautier 1971) and siamangs
(Hylobates syndactylus, Napier and Napier 1985). Impedance matching has
also been suggested as a function of the inflatable esophagus of male bitterns (Botaurus lentiginosus), which during the breeding season produce a
loud, low-pitched booming sound (Chapin 1922). Similar observations have
been made concerning the role of the swim bladder in sound-producing
teleost fishes (Demski et al. 1973) as well as other birds that inflate the crop
or a gular air sac during vocalization (e.g., doves and pigeons, grouse,
ostriches, bustards, and other species; Ziswiler and Farner 1972). There are
many other species with air sacs attaching to the vocal tract that have not
been studied but where a similar impedance-matching function seems plausible, for example in the “drumming” of emus (Dromaius novaehollandiae)
(McLelland 1989), in some baleen whales (Quayle 1991), or in the pharyngeal sacs of the walrus described above. In several species, an impedancematching function has been experimentally verified simply by puncturing
the air sacs and observing that normal-sounding vocalization continues but
at a much reduced amplitude (Gautier 1971; Gans 1973).
Another hypothesized role of air sacs is to allow air recycling, where the
same volume of air is used repeatedly to excite the voice source. Air
expelled through the lungs passes through the larynx into the elastic sac,
which then deflates, returning the air to the lungs. Such a role is clear for
anurans, where the recycling of air probably allows much higher rates of
vocalization than would be possible given the relatively inefficient mechanisms available to anurans to inflate the lungs; this mechanism may also
allow some conservation of mechanical energy (Dudley and Rand 1991).
Although most anurans appear to vocalize upon expiration, members of the
relatively primitive genera Discoglossus and Bombina vocalize upon inspiration (Schneider 1988). An air-recycling function also seems very likely for
the large laryngeal air sacs seen in Mysticete (baleen) whales (Hosokawa
1950; Quayle 1991), which can vocalize for long periods under water
without releasing air. Although no experimental data are available,
Mysticete air sacs are heavily invested with muscle, which would aid in
returning the expired air to the lungs.
94
W.T. Fitch and M.D. Hauser
monkeys, which play no acoustic role; Schön-Ybarra 1995). However, Fay
(1960) found this last suggestion unlikely, suggesting that Sleptsov’s specimens had regurgitated stomach contents into the sacs in their death throes.
In any case, the “bell production” and “life preserver” hypotheses are both
well-supported, indicating at least two functions for walrus pharyngeal sacs
and suggesting that air sacs may serve multiple functions in other species
as well.
The most commonly cited possible function for vocal air sacs is that they
play a role in impedance matching from the vocal tract to the atmosphere
(see below for more details). Such a role has been suggested for most
anurans as well as for the inflatable (soft-walled) air sacs of cercopithecid
monkeys such as guenons (Cercopithecus spp., Gautier 1971) and siamangs
(Hylobates syndactylus, Napier and Napier 1985). Impedance matching has
also been suggested as a function of the inflatable esophagus of male bitterns (Botaurus lentiginosus), which during the breeding season produce a
loud, low-pitched booming sound (Chapin 1922). Similar observations have
been made concerning the role of the swim bladder in sound-producing
teleost fishes (Demski et al. 1973) as well as other birds that inflate the crop
or a gular air sac during vocalization (e.g., doves and pigeons, grouse,
ostriches, bustards, and other species; Ziswiler and Farner 1972). There are
many other species with air sacs attaching to the vocal tract that have not
been studied but where a similar impedance-matching function seems plausible, for example in the “drumming” of emus (Dromaius novaehollandiae)
(McLelland 1989), in some baleen whales (Quayle 1991), or in the pharyngeal sacs of the walrus described above. In several species, an impedancematching function has been experimentally verified simply by puncturing
the air sacs and observing that normal-sounding vocalization continues but
at a much reduced amplitude (Gautier 1971; Gans 1973).
Another hypothesized role of air sacs is to allow air recycling, where the
same volume of air is used repeatedly to excite the voice source. Air
expelled through the lungs passes through the larynx into the elastic sac,
which then deflates, returning the air to the lungs. Such a role is clear for
anurans, where the recycling of air probably allows much higher rates of
vocalization than would be possible given the relatively inefficient mechanisms available to anurans to inflate the lungs; this mechanism may also
allow some conservation of mechanical energy (Dudley and Rand 1991).
Although most anurans appear to vocalize upon expiration, members of the
relatively primitive genera Discoglossus and Bombina vocalize upon inspiration (Schneider 1988). An air-recycling function also seems very likely for
the large laryngeal air sacs seen in Mysticete (baleen) whales (Hosokawa
1950; Quayle 1991), which can vocalize for long periods under water
without releasing air. Although no experimental data are available,
Mysticete air sacs are heavily invested with muscle, which would aid in
returning the expired air to the lungs.
94
W.T. Fitch and M.D. Hauser