Consistent variations in the underwater vocalizations of marine mammals can exist among individuals, groups, or ” «. populations. Differences in vocal patterns between geographically isolated populations have been described in several species of pinnipeds and cetaceans. Good examples are regional variations in the vocal repertoire of die Weddell seal (Leptomj-chotes weddellii) (Thomas and Stirling, 1983) or the major differences in the songs of humpback whales (Megaptera novaeangliae), in different ocean areas (Payne and Guinee, 1983). These types of variations have occasionally been referred to as dialects, but are more appropriately termed geographic variations (Conner, 1982). Such differences likely result from long-term geographic and, therefore, reproductive isolation, leading to genetically determined distinctiveness, although the potential for cultural variation cannot be ruled out. Pinnipeds and cetaceans are among the few mammalian orders in which vocal mimicry and learning have been documented (Janik and Slater, 1997). Transmission of vocal patterns across generations may thus depend more on cultural than on genetic mechanisms. Copying errors and other forms of cultural mutation and drift: may be responsible for at least a portion of the vocal differences between distant populations. Geographic variation in vocal patterns most likely represents epiphenomena, or by-products, of social and genetic isolation.
Vocal learning is most probably the primary mechanism involved in the evolution of dialects among local, interacting populations or groups of marine mammals. Dialects are well known in birds, but appear to be quite rare among marine mammals or, for that matter, mammals generally. They have only been described in two marine mammal species: killer whales (Orcinus orca) and sperm whales (Physeter macrocephalus). In British Columbia, matrilineal kinship groups, or pods, of “resident” killer whales have repertoires of 7-17 call types that vary among pods (Ford, 1989, 1991). All pods have distinctive features in their call repertoires, and thus each has a unique dialect. Certain pods share a portion of their call repertoires with others, and these are considered to belong to the same acoustic clan, each of which is acoustically distinct. Pods belonging to different clans have overlapping ranges and interact frequently, despite having very different call repertoires. New pods form by gradual fission of older, larger pods, along maternal lines. This process appears to be accompanied by divergence of common dialects, thus dialects reflect the historical matrilineal genealogy of pods within clans. Evidence indicates that minor though significant repertoire divergence exists among closely related matrilines within pods (Deecke et al, 2000; Miller and Bain, 2000). Dialects likely serve as acoustical “badges” that help maintain cohesion and integrity of matrilineal groups. Genetic studies by Barrett-Lennard (2000) suggest that dialects in British Columbian killer whales also serve as a mechanism promoting outbreeding. Group-specific dialects have also been found in Norwegian killer whales (Strager, 1995) and are to be expected in other populations of the species that are strongly matrilineally organized.
The other cetacean species with group-specific dialects, the sperm whale, tends also to live in matrilineal groups, although these lack the long-term stability seen in “resident” killer whales. Sperm whale groups were found to have repertoires that consistently varied in the proportional usage of different coda types and classes (Weilgart and Whitehead, 1997). No coda type was unique to any particular group. As in killer whales, groups of sperm whales with distinct dialects interact regularly. Geographic variations in coda repertoires were also noted in different oceans and in different areas within oceans, but such variations were weaker than those observed in group-specific dialects within local regions.
