during intense socializing (Ford 1989). These calls do not differ among pods
in a consistent way (Ford and Fisher 1983).
The exact function of discrete calls is unclear. They are given in various
social situations, and Ford (1989) hypothesized that they function as contact
calls, enabling pod mates to stay in acoustic contact while traveling and foraging. Ford speculated that calls might be important in coordinating foraging efforts during group hunts. Presumably, whales can use call differences
among pods to determine the pod to which the caller belongs, although call
repertoires overlap substantially (Fig. 4.7) and playback studies to test the
group signature and contact-call hypotheses have not been conducted. The
high degree of repertoire congruence among pods suggests that group identification might be based on structural variation between pods. Ford (1991)
hypothesized that shared calls indicate the level of relatedness between
pods or maternal lineage and might facilitate outbreeding (Ford and Fisher
1983). Unfortunately, there are no data to test most of these hypotheses.
Testing hypotheses for call function with playback experiments is critical
for establishing function and could also illuminate the selective pressures
that might favor a learned or genetic basis to calls. A single call type is sufficient to identify group membership; the occurrence and function of repertoires remain unexplained, but the pattern suggests additional functions
for repertoires. Genetic data indicate that mating usually occurs outside the
pod and clan, so mates are likely to differ acoustically (Barrett-Lennard
unpublished). This provides indirect support for the hypothesis that whales
base mate choice on the degree of acoustic similarity.
No direct published evidence that calls are learned exists. One observation of captive whales is consistent with the learning hypothesis. In this case,
a single Icelandic female produced calls similar to the BC whale with which
she was housed for many years (Bain 1988).
Despite this lack of direct evidence. Ford (1989, 1991) proposed that calls
are learned and that call sharing arises because pods that come into frequent contact imitate each others’ calls. Deecke (1998) investigated this
latter hypothesis with respect to structural variation, but not repertoire variation, by comparing the association patterns and vocal similarities of nine
matrilineal units in three pods. He tested four call types, two of which
showed some support for the hypothesis; however, the pattern was not
particularly strong. Even though a significant correlation existed between
acoustic similarity and association, dendrograms for acoustic similarity for
4. Comparative Vocal Learning
179
Figure 4.7. Killer whale social calls. (A) Sonograms of four call types. (B) Call types
in the repertoire of the pods studied. Three closely related pods have almost identical calls (1, A4, A5) but produce calls that other more distantly related pods do
not (B through Ill). Pods B through H have very similar repertoires. All pods except
Ill produce call N3, N5, and a variant of N7 and N8. (From Ford and Fisher 1982.
Reprinted with permission from the International Whaling Commission.)
᭣
in a consistent way (Ford and Fisher 1983).
The exact function of discrete calls is unclear. They are given in various
social situations, and Ford (1989) hypothesized that they function as contact
calls, enabling pod mates to stay in acoustic contact while traveling and foraging. Ford speculated that calls might be important in coordinating foraging efforts during group hunts. Presumably, whales can use call differences
among pods to determine the pod to which the caller belongs, although call
repertoires overlap substantially (Fig. 4.7) and playback studies to test the
group signature and contact-call hypotheses have not been conducted. The
high degree of repertoire congruence among pods suggests that group identification might be based on structural variation between pods. Ford (1991)
hypothesized that shared calls indicate the level of relatedness between
pods or maternal lineage and might facilitate outbreeding (Ford and Fisher
1983). Unfortunately, there are no data to test most of these hypotheses.
Testing hypotheses for call function with playback experiments is critical
for establishing function and could also illuminate the selective pressures
that might favor a learned or genetic basis to calls. A single call type is sufficient to identify group membership; the occurrence and function of repertoires remain unexplained, but the pattern suggests additional functions
for repertoires. Genetic data indicate that mating usually occurs outside the
pod and clan, so mates are likely to differ acoustically (Barrett-Lennard
unpublished). This provides indirect support for the hypothesis that whales
base mate choice on the degree of acoustic similarity.
No direct published evidence that calls are learned exists. One observation of captive whales is consistent with the learning hypothesis. In this case,
a single Icelandic female produced calls similar to the BC whale with which
she was housed for many years (Bain 1988).
Despite this lack of direct evidence. Ford (1989, 1991) proposed that calls
are learned and that call sharing arises because pods that come into frequent contact imitate each others’ calls. Deecke (1998) investigated this
latter hypothesis with respect to structural variation, but not repertoire variation, by comparing the association patterns and vocal similarities of nine
matrilineal units in three pods. He tested four call types, two of which
showed some support for the hypothesis; however, the pattern was not
particularly strong. Even though a significant correlation existed between
acoustic similarity and association, dendrograms for acoustic similarity for
4. Comparative Vocal Learning
179
Figure 4.7. Killer whale social calls. (A) Sonograms of four call types. (B) Call types
in the repertoire of the pods studied. Three closely related pods have almost identical calls (1, A4, A5) but produce calls that other more distantly related pods do
not (B through Ill). Pods B through H have very similar repertoires. All pods except
Ill produce call N3, N5, and a variant of N7 and N8. (From Ford and Fisher 1982.
Reprinted with permission from the International Whaling Commission.)
᭣