Biology Reference
In-Depth Information
Fig. 14.4
Spectrogram of a red deer roar representing the distribution of the energy
(in grey levels) across time (x axis) and frequency (y axis). The first four formants are visible
as dark bands of energy (labelled F1 to F4), which decrease throughout the vocalization.
The spacing between the formants is shown by the arrowed lines. The overall spacing
between the formants (or 'formant dispersion') is estimated by linear regression and
changes from 339 Hz at the beginning of the roar (corresponding to a 51.7 cm vocal
tract) to 243 Hz at the end of the roar (corresponding to a 72 cm vocal tract). Red deer
stags have a descended and mobile larynx that enables them to lengthen their vocal tract
during roaring, causing the observed drop in formant frequencies. The minimum formant
frequencies attained at the end of the roar reflect the fully extended vocal tract and,
therefore, communicate information about body size. From Reby and McComb (2003).
With permission from Elsevier. Photo © David Reby.
Red Deer
We now return to the example of red deer roars, which have been studied in depth by
David Reby, Karen McComb and colleagues. As discussed above, roaring is used in the
assessment process by which stags choose whether to escalate conflicts to physical
conflict. Do stags use some acoustic property of the roar, as with pitch in toads, to assess
the size of competitors and, therefore, their likely fighting ability?
Things are slightly more complex in mammals such as red deer, because the size of
larynx can be varied independently of body size and so, in contrast to toads, the pitch
of sound produced by the vocal cords is not a reliable index of body size. However,
acoustic theory predicts that the resonant frequencies of the vocal tract (termed
formants) and the average distance between them (termed formant dispersion)
should be negatively correlated with vocal tract length (Fig. 14.4; Fitch & Reby,
2001). Consequently, if vocal tract length were constrained by body size, then
formant frequencies and dispersion would provide a reliable index of body size. Reby
and McComb (2003) tested this by comparing the roars of different stags from
an intensively studied population on the island of Rum (Inner Hebrides, UK).
They found that, as predicted, larger stags produced lower formant frequencies
(Fig. 14.5a). Furthermore, the long-term data on the population showed that stags
Formant
frequencies are a
reliable index of
body size in red
deer
