the environment and to disruption by atmospheric turbulence. Near ground
level, however, sounds with very low frequencies will also experience
increased levels of attenuation because of destructive interference between
direct waves and waves reflected from the ground (Wiley and Richards
1978). Thus, near ground level in some environments, a “sound window”
of minimal excess attenuation exists for frequencies in the range of about
1–3 kHz (Morton 1975; Marten et al. 1977; Waser and Brown 1986).
The temporal structure of a signal can also influence the amount of
degradation it experiences (Richards and Wiley 1980; Ryan and Sullivan
1989; Mathevon et al. 1996). Reverberations and air turbulence can blur
amplitude modulation contained within signals, favoring tonal signals over
amplitude-modulated signals (Richards and Wiley 1980). The frequency of
the signal also influences these effects—sounds between 2 and 8 kHz are
less subject to reverberations from the vegetation and ground and are thus
favored with respect to the maintenance of temporal fidelity (Wiley and
Richards 1978).
3.2. Testing the Acoustic Adaptation Hypothesis
Like signals, all habitats are not equal with respect to signal transmission.
Selection on call structure may be stronger, and thus elicit a stronger evolutionary response, in one type of environment versus another (Morton
1975; Ryan et al. 1990). Alternatively, the optimal call structure for longdistance communication can vary among habitats or microhabitats. For
example, it has been shown that amplitude-modulated calls are favored for
communication in open environments, whereas tonal calls are favored in
forested environments (Morton 1975; Sorjonen 1986). Similarly, the presence and shape of the frequency sound window for low excess attenuation
appears to differ among environments as well as between different heights
within the same environment (Morton 1975; Marten and Marler 1977;
Waser and Brown 1986). The frequency window may be present in some
locations and absent in others, or the range of frequencies experiencing low
excess attenuation may differ among habitats.
Given that different habitats can impose differing selection on the structure of calls used for long-distance communication, one might expect predictable divergence among the signals used by the species living in these
habitats. These predictions can be used to test the hypothesis that longdistance communication signals have evolved in response to selection
generated by habitat acoustics. Adaptation to the acoustic environment is
one possible explanation when the observed trends in call characteristics
match the predictions. Following this logic, a number of studies have used
comparative methods to test the hypothesis that the signals used for longdistance communication have evolved in response to selection for decreased
degradation and increased transmission distance within their home environment.
234
M.J. Ryan and N.M. Kime
level, however, sounds with very low frequencies will also experience
increased levels of attenuation because of destructive interference between
direct waves and waves reflected from the ground (Wiley and Richards
1978). Thus, near ground level in some environments, a “sound window”
of minimal excess attenuation exists for frequencies in the range of about
1–3 kHz (Morton 1975; Marten et al. 1977; Waser and Brown 1986).
The temporal structure of a signal can also influence the amount of
degradation it experiences (Richards and Wiley 1980; Ryan and Sullivan
1989; Mathevon et al. 1996). Reverberations and air turbulence can blur
amplitude modulation contained within signals, favoring tonal signals over
amplitude-modulated signals (Richards and Wiley 1980). The frequency of
the signal also influences these effects—sounds between 2 and 8 kHz are
less subject to reverberations from the vegetation and ground and are thus
favored with respect to the maintenance of temporal fidelity (Wiley and
Richards 1978).
3.2. Testing the Acoustic Adaptation Hypothesis
Like signals, all habitats are not equal with respect to signal transmission.
Selection on call structure may be stronger, and thus elicit a stronger evolutionary response, in one type of environment versus another (Morton
1975; Ryan et al. 1990). Alternatively, the optimal call structure for longdistance communication can vary among habitats or microhabitats. For
example, it has been shown that amplitude-modulated calls are favored for
communication in open environments, whereas tonal calls are favored in
forested environments (Morton 1975; Sorjonen 1986). Similarly, the presence and shape of the frequency sound window for low excess attenuation
appears to differ among environments as well as between different heights
within the same environment (Morton 1975; Marten and Marler 1977;
Waser and Brown 1986). The frequency window may be present in some
locations and absent in others, or the range of frequencies experiencing low
excess attenuation may differ among habitats.
Given that different habitats can impose differing selection on the structure of calls used for long-distance communication, one might expect predictable divergence among the signals used by the species living in these
habitats. These predictions can be used to test the hypothesis that longdistance communication signals have evolved in response to selection
generated by habitat acoustics. Adaptation to the acoustic environment is
one possible explanation when the observed trends in call characteristics
match the predictions. Following this logic, a number of studies have used
comparative methods to test the hypothesis that the signals used for longdistance communication have evolved in response to selection for decreased
degradation and increased transmission distance within their home environment.
234
M.J. Ryan and N.M. Kime