4. Acoustic Communication in Whales and Dolphins
175
principle is that if the frequency of a sound is influenced by a physical structure such as a resonant cavity, then the frequency of the sound will be lower
for an animal with a larger resonator. If females choose larger males by
selecting for those with sounds of the lowest peak frequency, then this will
create a selection pressure for males to make the lowest frequencies they
can. However, if the minimum peak frequency a male can produce is constrained by the size of his sound production apparatus and if the size of this
apparatus correlates with body size (e.g., Myrberg et al. 1993 for bicolor
damselfish), then males may be constrained to produce an honest advertisement of their body size. This correlation between the lowest peak frequency and body size does not hold for most terrestrial mammals (Fitch
1997). Fitch (1997) argues that a different acoustic cue, called formant frequency dispersion, is a better predictor of body size in mammals, but neither
feature has been tested with marine mammals.
Ryan (1994) argues that preexisting biases in the sensory systems of
females may be an important factor in the evolution of displays by sexual
selection. Most communication systems show a match between the speciesspecific frequency ranges of the signals produced and the sensory perception system. However, in two species of frogs studied by Ryan and
colleagues (1990,1992), the typical frequency of the male advertisement call
was higher in frequency than the frequency at which females were most
sensitive. The larger the male, the lower the frequency of its advertisement,
and the closer it was to the frequencies at which females are more sensitive. Ryan argues that this sensory bias can form the basis of female choice
of male displays.
Sexual selection may also drive female preferences for complexity in
advertisements. Humans find many advertisement displays, from the tail of
the peacock to the song of the nightingale, to be complex and beautiful.
Fisher (1958) explained the evolution of complex and beautiful secondary
sexual characters in terms of a runaway process of sexual selection. The
"runaway process" emphasizes positive feedback between the female preference and the elaboration of a male display. Let us start by assuming that
females have developed a preference for an acoustic display with a particular feature. This preference could arise because the display was correlated
with some valuable male trait, the display could make the male more easy
to find, or females could simply have some perceptual bias to respond preferentially to a particular stimulus (Lande and Arnold 1981). Whatever the
origin of the preference, the tendency for females to prefer males with the
feature means that genes for the preference will covary with genes for
the feature. If females on average select males with extreme development
of the feature, then the next generation will have more sons with the exaggerated feature and more females with the preference. This creates a positive feedback loop, potentially producing a runaway process leading to
extreme and exaggerated development of secondary sexual characters in
males and preferences in females. The most obvious consequence of this
175
principle is that if the frequency of a sound is influenced by a physical structure such as a resonant cavity, then the frequency of the sound will be lower
for an animal with a larger resonator. If females choose larger males by
selecting for those with sounds of the lowest peak frequency, then this will
create a selection pressure for males to make the lowest frequencies they
can. However, if the minimum peak frequency a male can produce is constrained by the size of his sound production apparatus and if the size of this
apparatus correlates with body size (e.g., Myrberg et al. 1993 for bicolor
damselfish), then males may be constrained to produce an honest advertisement of their body size. This correlation between the lowest peak frequency and body size does not hold for most terrestrial mammals (Fitch
1997). Fitch (1997) argues that a different acoustic cue, called formant frequency dispersion, is a better predictor of body size in mammals, but neither
feature has been tested with marine mammals.
Ryan (1994) argues that preexisting biases in the sensory systems of
females may be an important factor in the evolution of displays by sexual
selection. Most communication systems show a match between the speciesspecific frequency ranges of the signals produced and the sensory perception system. However, in two species of frogs studied by Ryan and
colleagues (1990,1992), the typical frequency of the male advertisement call
was higher in frequency than the frequency at which females were most
sensitive. The larger the male, the lower the frequency of its advertisement,
and the closer it was to the frequencies at which females are more sensitive. Ryan argues that this sensory bias can form the basis of female choice
of male displays.
Sexual selection may also drive female preferences for complexity in
advertisements. Humans find many advertisement displays, from the tail of
the peacock to the song of the nightingale, to be complex and beautiful.
Fisher (1958) explained the evolution of complex and beautiful secondary
sexual characters in terms of a runaway process of sexual selection. The
"runaway process" emphasizes positive feedback between the female preference and the elaboration of a male display. Let us start by assuming that
females have developed a preference for an acoustic display with a particular feature. This preference could arise because the display was correlated
with some valuable male trait, the display could make the male more easy
to find, or females could simply have some perceptual bias to respond preferentially to a particular stimulus (Lande and Arnold 1981). Whatever the
origin of the preference, the tendency for females to prefer males with the
feature means that genes for the preference will covary with genes for
the feature. If females on average select males with extreme development
of the feature, then the next generation will have more sons with the exaggerated feature and more females with the preference. This creates a positive feedback loop, potentially producing a runaway process leading to
extreme and exaggerated development of secondary sexual characters in
males and preferences in females. The most obvious consequence of this