researchers have been able to reliably observe and record individual
animals over long periods of time (e.g., Clark 1982; Tyack 1983; Dahlheim
et al. 1984; Clark et al. 1986; Silber 1986). There have been a few observations of blue and fin whales in surface-active groups with associated recordings of transient sounds (Watkins 1981; Watkins et al. 1987; Edds 1988).
Transient signals from coastal species are highly variable in their acoustic
structures. In contrast, transient signals from at least three of the pelagic
species are much simpler and share common acoustic attributes. Contact
calls from right and bowhead whales are simple, frequency-modulated
sweeps, whereas feeding calls from humpbacks, which serve to attract
distant whales to the aggregation, are long, constant-frequency sirens. At a
qualitative comparative level, the most variable transient signals from blue
and fin whales are equal to the simplest transient signals from right and
humpback whales. Values for pelagic species are either from recordings
during late winter or late summer when groups of animals were observed
and feeding was the primary activity. Figure 2.6 presents spectrographic
representations of calls for blue, fin, humpback, right, and bowhead whales.
There is enough evidence to make comparisons between the sounds of
coastal and pelagic mysticetes. Here, we will not rely on statistical analyses
(e.g., discriminant function, principle components; see Cortopassi and
Bradbury 2000) to reveal acoustic groupings. Instead, we will plot several
basic acoustic attributes by species and label them according to the preferred environment in which those sounds are typically produced. We
proceed with this mechanism under the assumption that environmental
influence on acoustic features at the evolutionary level should be relatively
obvious, especially when the differences in the acoustic environments are
so drastic. More subtle differences resulting from immediate behavioral
influences (e.g., motivation) are expected to be graded (see Clark 1983) and,
given the small species sample size, not so easily revealed.
In Figures 2.7 and 2.8, we present several graphical representations of
acoustic features for the five representative species. Figure 2.7 emphasizes
the importance of inter–note interval, frequency bandwidth, and duration
for songs. It shows that song notes for coastal species have greater timebandwidth products with shorter inter–rate intervals than songs for pelagic
species and that, for these two features, the humpback and bowhead whales
are remarkably similar while dissimilar from the blue and fin whales. Humpbacks and bowheads are rarely found in the same regions, whereas blue and
fin whales are sympatric. In Figure 2.8, species-specific spectral profiles
(solid lines) are contrasted with ambient noise spectra (hatched lines) from
the deep and shallow water environments. The left-hand column illustrates
these spectra for songs, whereas the right-hand column illustrates these
spectra for calls. This figure indicates that the overall frequency compositions of the different species’ song energies are well-matched to the ambient
noise properties of the environment. One immediate advantage of producing song in a frequency band of low ambient noise is a greater range of
40
A.H. Bass and C.W. Clark
animals over long periods of time (e.g., Clark 1982; Tyack 1983; Dahlheim
et al. 1984; Clark et al. 1986; Silber 1986). There have been a few observations of blue and fin whales in surface-active groups with associated recordings of transient sounds (Watkins 1981; Watkins et al. 1987; Edds 1988).
Transient signals from coastal species are highly variable in their acoustic
structures. In contrast, transient signals from at least three of the pelagic
species are much simpler and share common acoustic attributes. Contact
calls from right and bowhead whales are simple, frequency-modulated
sweeps, whereas feeding calls from humpbacks, which serve to attract
distant whales to the aggregation, are long, constant-frequency sirens. At a
qualitative comparative level, the most variable transient signals from blue
and fin whales are equal to the simplest transient signals from right and
humpback whales. Values for pelagic species are either from recordings
during late winter or late summer when groups of animals were observed
and feeding was the primary activity. Figure 2.6 presents spectrographic
representations of calls for blue, fin, humpback, right, and bowhead whales.
There is enough evidence to make comparisons between the sounds of
coastal and pelagic mysticetes. Here, we will not rely on statistical analyses
(e.g., discriminant function, principle components; see Cortopassi and
Bradbury 2000) to reveal acoustic groupings. Instead, we will plot several
basic acoustic attributes by species and label them according to the preferred environment in which those sounds are typically produced. We
proceed with this mechanism under the assumption that environmental
influence on acoustic features at the evolutionary level should be relatively
obvious, especially when the differences in the acoustic environments are
so drastic. More subtle differences resulting from immediate behavioral
influences (e.g., motivation) are expected to be graded (see Clark 1983) and,
given the small species sample size, not so easily revealed.
In Figures 2.7 and 2.8, we present several graphical representations of
acoustic features for the five representative species. Figure 2.7 emphasizes
the importance of inter–note interval, frequency bandwidth, and duration
for songs. It shows that song notes for coastal species have greater timebandwidth products with shorter inter–rate intervals than songs for pelagic
species and that, for these two features, the humpback and bowhead whales
are remarkably similar while dissimilar from the blue and fin whales. Humpbacks and bowheads are rarely found in the same regions, whereas blue and
fin whales are sympatric. In Figure 2.8, species-specific spectral profiles
(solid lines) are contrasted with ambient noise spectra (hatched lines) from
the deep and shallow water environments. The left-hand column illustrates
these spectra for songs, whereas the right-hand column illustrates these
spectra for calls. This figure indicates that the overall frequency compositions of the different species’ song energies are well-matched to the ambient
noise properties of the environment. One immediate advantage of producing song in a frequency band of low ambient noise is a greater range of
40
A.H. Bass and C.W. Clark