114
T.W. Cranford
The pulse repetition rate or its inverse, the interpulse interval, is rarely
constant and sometimes changes cyclically with trained dolphins during
echolocation tasks (Au et al. 1974; Thomas and Turl 1990). These animals
exhibit considerable control over the repetition rate. Some studies have
observed that dolphins operate in a pulse-echo mode, where the animal
produces one pulse and waits for the returning echo before producing
another (Evans and Powell 1967; Morozov et al. 1972; Au et al. 1974), suggesting that repetition rate is a function of target distance (interpulse interval greater than the two-way acoustic transit time). However, work with
D. leucas has demonstrated greater flexibility in the pulse-echo pattern
and repetition rate (Au et al. 1987; Turl and Penner 1989). We may find this
flexibility is a response to a particular problem-solving strategy and may
be more widespread than we currently have evidence to suspect.
High repetition rate pulsed sounds have been variously named and frequently proposed to have social functions (Herman and Tavolga 1980;
Norris et al. 1994). When pulses are produced in high repetition rate bursts,
they take on an audible harmonic structure that is partially related to the
time separation between pulse events; the so-called "time separation pitch."
These high repetition rate pulse sequences have commonly been referred
to as "burst pulses" but many other descriptive terms can be found, especially in the early literature (ct. Lilly and Miller 1961). Some examples of
terms are raspberries, blats, bleats, squawks, moans, barks, and squeaks. The
terms are rarely explicitly defined, so the choice of terminology is likely
related to the subjective experience of the reporter.
It should be noted that the pulses in these high repetition rate bursts are
produced so rapidly that they may run together or overlap in time so that
it becomes difficult to subject individual pulses to electronic analysis or
central nervous system processing. It is possible that more than one pulse
generator is required to attain these rapid repetition rates. For example,
since some of the highest click repetition rates are more rapid than the
fastest nerve firing rates, it seems unlikely that there is a single sound generator, where each click generation event is being controlled by a corresponding nervous system depolarization event. Consequently repetition
rate may provide an important clue for understanding odontocete pulse
generation.
By comparison, the spectral frequency composition of a single click might
reveal aspects of anatomic structure and the mechanism(s) odontocetes use
to generate pulses. As mentioned earlier, frequency composition involves
subtleties that may hold the potential for confusion or a key to understanding. Unless otherwise stated, the following remarks apply only to the
spectral or "internal" frequency structure of individual pulses, rather than
the repetition frequency or relationships between pulsed events (i.e., the
time separation pitch).
It is useful to differentiate between two distinct pulse types because the
signals are fundamentally different, although they may be generated at the
T.W. Cranford
The pulse repetition rate or its inverse, the interpulse interval, is rarely
constant and sometimes changes cyclically with trained dolphins during
echolocation tasks (Au et al. 1974; Thomas and Turl 1990). These animals
exhibit considerable control over the repetition rate. Some studies have
observed that dolphins operate in a pulse-echo mode, where the animal
produces one pulse and waits for the returning echo before producing
another (Evans and Powell 1967; Morozov et al. 1972; Au et al. 1974), suggesting that repetition rate is a function of target distance (interpulse interval greater than the two-way acoustic transit time). However, work with
D. leucas has demonstrated greater flexibility in the pulse-echo pattern
and repetition rate (Au et al. 1987; Turl and Penner 1989). We may find this
flexibility is a response to a particular problem-solving strategy and may
be more widespread than we currently have evidence to suspect.
High repetition rate pulsed sounds have been variously named and frequently proposed to have social functions (Herman and Tavolga 1980;
Norris et al. 1994). When pulses are produced in high repetition rate bursts,
they take on an audible harmonic structure that is partially related to the
time separation between pulse events; the so-called "time separation pitch."
These high repetition rate pulse sequences have commonly been referred
to as "burst pulses" but many other descriptive terms can be found, especially in the early literature (ct. Lilly and Miller 1961). Some examples of
terms are raspberries, blats, bleats, squawks, moans, barks, and squeaks. The
terms are rarely explicitly defined, so the choice of terminology is likely
related to the subjective experience of the reporter.
It should be noted that the pulses in these high repetition rate bursts are
produced so rapidly that they may run together or overlap in time so that
it becomes difficult to subject individual pulses to electronic analysis or
central nervous system processing. It is possible that more than one pulse
generator is required to attain these rapid repetition rates. For example,
since some of the highest click repetition rates are more rapid than the
fastest nerve firing rates, it seems unlikely that there is a single sound generator, where each click generation event is being controlled by a corresponding nervous system depolarization event. Consequently repetition
rate may provide an important clue for understanding odontocete pulse
generation.
By comparison, the spectral frequency composition of a single click might
reveal aspects of anatomic structure and the mechanism(s) odontocetes use
to generate pulses. As mentioned earlier, frequency composition involves
subtleties that may hold the potential for confusion or a key to understanding. Unless otherwise stated, the following remarks apply only to the
spectral or "internal" frequency structure of individual pulses, rather than
the repetition frequency or relationships between pulsed events (i.e., the
time separation pitch).
It is useful to differentiate between two distinct pulse types because the
signals are fundamentally different, although they may be generated at the