E-mail : lrendell@is2.dal.ca; hwhitehe@is.dal.ca

WWW : http://is.dal.ca/~whitelab

Keywords : animal culture, cetaceans, cognition, co-evolution, cultural transmission, dolphins, evolution of culture, imitation, teaching, whales

Studies of animal culture have not normally included a consideration of cetaceans. However, with several long-term field studies now maturing, this situation should change. Animal culture is generally studied by either investigating transmission mechanisms experimentally, or observing patterns of behavioural variation in wild populations which cannot be explained by either genetic or environmental factors. Taking this second, ethnographic, approach, there is good evidence for cultural transmission in several cetacean species. However, only the bottlenose dolphin (Tursiops spp.) has been shown experimentally to possess sophisticated social learning abilities, including vocal and motor imitation; other species have not been studied. There is observational evidence for imitation and teaching in killer whales. For cetaceans, and other large wide-ranging animals, excessive reliance on experimental data for evidence of culture is not productive, we favour the ethnographic approach. The complex and stable vocal and behavioural cultures of sympatric groups of killer whales (Orcinus orca) appear to have no parallel outside humans and represent an independent evolution of cultural faculties. The wide movements of cetaceans, the greater variability of the marine environment over large temporal scales relative to that on land, and the stable matrilineal social groups of some species are potentially important factors in the evolution of cetacean culture. There have been suggestions of gene-culture coevolution in cetaceans, and culture may be implicated in some unusual behavioural and life-history traits of the whales and dolphins. We hope to stimulate both discussion and research on culture in these animals.

Introduction 

The presence of cultural processes in non-human animals is an area of some controversy (Galef 1992; de Waal 1999). In this target article we attempt to fuel the debate by reviewing the evidence for cultural transmission in whales and dolphins (order Cetacea), a group which has so far received almost no attention from students of animal culture. Studies of cetaceans have uncovered a number of patterns of behaviour and vocalizations which some cetologists have ascribed to cultural processes. Here we review these results from the perspectives used in research on cultural transmission in other animals.

Theoretical investigations suggest that cultural transmission of information should be adaptive in a broad range of environments (Boyd & Richerson 1985), but it is quite rarely documented outside humans (but see Slater 1986; Whiten et al. 1999). This discrepancy has yet to be explained (Laland et al. 1996). When stable over generations, culture can strongly affect biological evolution, in both theory (e.g. Findlay 1991) and practice - much of human behaviour is determined by a broad range of cultural processes and there is good evidence for gene-culture coevolution in our species (Feldman & Laland 1996). In contrast, among non-human animals culture is much simpler, rarer, and, except possibly in the case of bird song (Grant & Grant 1996), thought not to have the stability necessary to make a substantial impact on genetic evolution (Laland 1992; Feldman & Laland 1996).  

The logistical difficulties of studying wild cetaceans make the study of culture difficult, and often give rise to information which is incomplete and poor in detail. Nonetheless, we feel it is timely to introduce cetaceans into the wider debate surrounding animal culture for a number of reasons. Firstly, there is growing evidence of cultural transmission and cultural evolution in the cetaceans, some of which is strong, some of which is weaker, but which when taken as a whole make a compelling case for the detailed study of cultural phenomena in this group. While culture and cultural transmission have been briefly discussed in the context of cetaceans by a number of authors (Norris & Dohl 1980; Osborne 1986; Shane et al. 1986; Norris & Schilt 1988; Felleman et al. 1991; Ford 1991; Norris et al. 1994; Silber & Fertl 1995) no synthesis has been attempted. Secondly, the evidence now available describes some interesting and rare (in some cases unique outside humans) patterns of behavioural variation in the wild, likely maintained by cultural transmission processes. Thirdly, there is growing evidence that in the complexity of their social systems - the only non-human example of second-order alliances (Connor et al. 1998) - and their cognition - data suggest that dolphins can use abstract representations of objects, actions and concepts to guide their behaviour (Herman et al. 1993; 1994) - some cetaceans match or exceed all other non-human animals. Since complex social systems and advanced cognitive abilities have been suggested as good predictors of animal culture (Roper 1986), it is pertinent to ask whether these factors are reflected in the cultural faculties of cetaceans. Finally, cetaceans provide an interesting contrast to the study of culture in humans and other terrestrial animals, since they inhabit a radically different environment and perhaps represent an independent evolution of social learning and cultural transmission. 

The study of animal culture is heavily influenced by perspective. Hence, before we review culture in cetaceans, we discuss the differing approaches that have been taken to the study of non-human culture. We then review evidence for culture in cetaceans from the two principal perspectives, ethnographic patterns and the experimental study of imitation and teaching, comparing the results with the most similar phenomena described in other groups of animals. After trying to reconcile the evidence from these two approaches, we consider the evolution of cultural transmission in cetaceans, gene-culture coevolution, and the possibility that cultural processes may explain some unusual behavioural and life-history patterns of whales and dolphins.  

Clearly our review of culture in cetaceans will depend heavily upon our idea of what culture is. There is little consensus on this issue; the term 'culture' is defined in an array of subtly different ways within the literature, some of which we have listed in Table 1. We have not included definitions which make culture a trait only humans show. These were considered by Mundinger (1980) in his review of cultural theory. We agree with him both in that there is "no empirical evidence" supporting such a cultural dichotomy between humans and other animals, and that more general concepts of culture are more likely to advance understanding. The work of Boyd and Richerson (1985; 1996) has been crucially important in giving the study of cultural transmission and cultural evolution a sound theoretical basis (Bettinger 1991, p.182). Thus the definitions of cultural that they found useful are particularly important, and have heavily influenced our decision on which definition to adopt: 

Whiten and Ham (1992) list a range of "social processes" as supporting cultural transmission, and in the definition of culture we use, the term "some form of social learning" refers to these processes. These comprise exposure, social support, matched dependent learning, stimulus enhancement, observational conditioning, imitation, and goal emulation as listed and defined in Whiten and Ham (1992). 

In contrast to this broad definition, some scientists have insisted that cultural transmission only takes place under two types of social learning: phylogenetically homologous (to humans) imitation and teaching (Galef 1992; Tomasello 1994). This restriction is justified on the grounds that these processes seem to be vital elements of human culture (Galef 1992), and also that, unlike other forms of social learning, they allow complex cultures to be constructed by successive modification (Boyd & Richerson 1985). While we reject this narrower view of culture for reasons discussed below, we recognize that teaching and imitation are particularly important forms of social learning when considering cultural transmission. 

The empirical study of cultural processes in animals is generally approached in two major ways : controlled laboratory experiments on social learning mechanisms and field descriptions of behavioural variation (Lefebvre & Palameta 1988). The first follows from the restriction of culture to imitation and teaching and emphasises process - is imitation and/or teaching taking place - and the second, espoused by those accepting a generally broader definition of culture, emphasises product - cultural patterns. Both make important contributions to our understanding of culture.  

The first approach focuses on experimental study of the cognitive processes underlying cultural transmission. In general, controlled laboratory experimentation is the preferred methodological tool; this gives the approach the advantage of controlled conditions and hence less chance of ambiguity in the interpretation of data. However, the studies do not necessarily relate to what occurs in the wild, and care must be taken to establish that such studies are not simply measuring what McGrew (1992, p.21) calls the "socio-ecological validity of the captive environment" rather than the true abilities of the animals under scrutiny.  

The second approach is field-based, involving the systematic assimilation of data on the behaviour of individuals and groups often over large temporal and spatial scales. Here culture is deduced from patterns of behavioural variation in time and space, which cannot be explained by environmental or genetic factors (Nishida 1987; Boesch et al. 1994; Boesch 1996; Whiten et al. 1999). This approach has been likened to ethnography in the social sciences (Wrangham et al. 1994). The strength of this approach is that it is firmly rooted in what the animals actually do in the wild, with the unavoidable weakness that results can be more ambiguous than those derived from controlled experiments - such studies cannot usually tell us much about which specific social learning processes are involved in producing the observed behavioural variation.

These two approaches have interacted in different ways in the study of culture and cultural transmission in different taxonomic groups. Culture in humans is studied largely from an ethnographic perspective, although some experimental work has been done (e.g. Tomasello et al. 1993; Meltzoff 1996). In the study of the cultural evolution of bird-song, the two approaches have generally integrated cooperatively with laboratory and field studies complementing each other in a stimulating and progressive way (see Baker & Cunningham 1985). In non-primate mammals there exists an impressive body of work concerning the social transmission of feeding behaviour, based mainly on an experimental approach (see Galef 1996), with little reference to variation in the wild (for a notable exception, see Terkel 1996). It is in the discussion surrounding culture in non-human primates that the most severe dichotomy between these two perspectives is apparent. A lack of laboratory evidence for imitation has led to the persistent denial of culture in chimpanzees (Pan troglodytes) from some (Galef 1992; Tomasello 1994), while others, drawing on field evidence of variation in behaviour such as nut-cracking which cannot be explained by ecological or genetic factors, maintain that wild chimpanzees do have distinctive and complex cultures (Boesch et al. 1994; McGrew 1994; Boesch 1996; Whiten et al. 1999).  

We strongly believe that research on cultural processes is best served by an approach which integrates the sometimes opposing process and product oriented perspectives, as well as the laboratory and field approaches, taking good data from each. This cannot be achieved unless both perspectives are understood, and so we shall approach cetacean culture from both in turn. Following this we will bring our own perspective, as field biologists heavily influenced by evolutionary ecology, to an attempted integration.  

 The ethnographic evidence for cetacean culture is remarkably strong given the substantial difficulties of studying whales and dolphins in the wild. In only four (of ~80) species of Cetacea have more than a handful of papers on behaviour been published (Mann 1999): the bottlenose dolphin (Tursiops spp.), the killer whale (Orcinus orca), the sperm whale (Physeter macrocephalus) and the humpback whale (Megaptera novaeangliae). However, studies of each of these species have been carried out in different ocean basins and over time periods of ten years and more. In many attributes, these four species span a wide range. For instance, their sizes range from 2m (bottlenose dolphins) to 16m (sperm whales), their habitat from protected coastal lagoons (bottlenose dolphins) to deep oceanic waters (sperm whales), and trophic levels from partial planktivores (humpback whales) to top predators (killer whales). The four species have diverse social systems: humpback whales live in loose fission-fusion societies (Clapham 1993); both sexes of killer whale generally remain within their natal matrilineal group (Baird 1999); female sperm whales live in largely matrilineal groups from which males disperse to lead quite solitary adult lives (Whitehead & Weilgart 1999); while in bottlenose dolphins, males can form stable alliances while females possess a network of more labile relationships (Connor et al. 1998). Although the four well-studied cetacean species are socially diverse, they are likely unrepresentative of all cetaceans. For instance, the pelagic dolphins, beaked whales, and river dolphins may have quite different social systems (Connor et al. 1998), and cultural faculties.  

From the ethnographic perspective cultural transmission is deduced from spatial, temporal or social patterns of variation in behaviour which are not consistent with genetic or environmental determination, or individual learning. It should be noted that from this perspective, no attempt is made to deduce what particular form of social learning underlies the observed patterns. We will consider three types of pattern:  

1. rapid spread of a novel and complex form of behaviour through a segment of the population, indicating a largely horizontal - within-generation (Cavalli-Sforza & Feldman 1981) - cultural process; 
2. mother-offspring similarity in a complex form of behaviour, indicating vertical - parent-offspring (Cavalli-Sforza & Feldman 1981) - cultural transmission ; 
3. differences in complex behaviour between stable groups of animals which are hard to explain by genetic differences, shared environments, or the sizes or demographic structure of the groups. Such patterns could arise through vertical or oblique - learning from a non-parental model of the previous generation (Cavalli-Sforza & Feldman 1981) - transmission within strictly matrilineal groups, or through a combination of vertical, oblique and horizontal within-group transmission in a system with conformist traditions - individuals aligning their behaviour with that of other group members (Boyd & Richerson 1985) - within more labile groups.

 We will refer to these as 'rapid-spread', 'mother-offspring' and 'group-specific' behavioural patterns, respectively. Our categories are not discrete, as the same cultural phenomenon (such as behaviour learned primarily from the mother) could be inferred in more than one way ('mother-offspring' or 'group-specific' if groups are matrilineal). However, there is a distinction between 'rapid-spread' patterns, which are likely primarily due to within-generation transmission, and the other categories, which likely incorporate a significant between-generation transmission component. This distinction is important from an evolutionary perspective since it is between-generation transmission that has the most profound evolutionary effects (Russell & Russell 1990; Laland 1992; Feldman & Laland 1996). Here, we review examples of each pattern in cetaceans, and compare what has been found with results from other animals. We consider cases where environmental and genetic causation can be ruled out, and also those where such causes are theoretically feasible but practically unlikely.

 When new behavioural variants spread through much of a population over time scales of less than a generation, then genetic causation can be excluded, but environmental change plus individual learning must be considered as an alternative to social learning. If behavioural change is continuous, then an environmental causative factor should vary over a similar temporal scale. The spread of a single novel behaviour through a population over a short period could be caused either by environmental change and then individuals learning the appropriate behaviour independently, or by social learning (culture). Distinguishing between these alternatives requires either observation of individual or social learning (which is very hard for cetaceans-see below), or a consideration of the likelihood that a new environmental factor could have triggered a bout of independent individual innovations.  

On their winter breeding grounds male humpback whales produce songs, structured sequences of vocalizations cycling with a period of about 5-25min (Payne & McVay 1971). At any time, all males in a breeding population sing nearly the same song, but the song evolves structurally over time, changing noticeably over a breeding season, substantially over periods of several years, but remaining stable over the largely non-singing summer months (Payne, K. & Payne 1985). Males sing virtually identical songs on breeding grounds thousands of kilometres apart, and the songs on these different grounds evolve as one. For instance, songs from Maui, Hawaii and Islas Revillagigedo, Mexico (4,500km apart) are similar at any time but change in the same way over a two year period (Payne & Guinee 1983). While the mechanisms underlying this process are not fully understood, horizontal cultural transmission almost certainly plays an important role in maintaining song homogeneity as there is no conceivable environmental trigger for such a pattern of variation (Payne & Guinee 1983; Cerchio 1993) - it may be that the oceanic deep sound channel (Payne & Webb 1971) plays a role in facilitating this transmission.  

Superficially, this pattern of rapid change is similar to the cultural evolution of song in yellow-rumped caciques (Cacius cela) (Feekes 1982; Trainer 1989) and village indigobirds (Vidua chalybeata) (Payne 1985), but there are important differences. Humpback song is homogenous over entire ocean basins compared to the sharp variation over short distances in both bird species, and thousands of individual humpbacks share the same song compared to the colony- or locale-specific bird songs (Cerchio 1993). Thus it is hard to see how the changes in humpback song could be driven by imitating a few dominant males as has been suggested by Trainer (1989) for birds showing similarly rapid change - some other cultural process must be acting. Cerchio (1993) suggests that evolving humpback song may constitute dialects in the time domain, and that conformity to the current dialect may be socially significant in the same way that conformity to the local dialect is in birds (we take as our definition of dialect that of Connor (1982) - variation in the vocal behaviour of different but potentially interbreeding groups). Nevertheless, the differences in scale make humpback songs a so far unique instance among non-humans of a continuously evolving conformist culture in a large and dispersed population.

Bowhead whales (Balaena mysticetus) of the Bering Sea stock also sing during their peak-mating season. While their songs are simpler than the humpbacks', bowhead songs also change from year to year and all males on a given migration sing the same general song (Clark 1990; Würsig & Clark 1993). Bowheads have been observed apparently imitating conspecific calls (Clark 1990), hence horizontal cultural transmission also likely maintains song homogeneity in this species. However, as much less data are available for comparisons between areas and over time than with the humpbacks, the characteristics of this system await further recordings from different populations.  

Cultural innovations can also spread quite rapidly on humpback feeding grounds where they spend the summer months. In the southern Gulf of Maine, a novel complex feeding technique, 'lobtail feeding', was first observed in 1981, and by 1989 had been adopted by nearly 50% of the population (Weinrich et al. 1992). This feeding method is apparently a modification of 'bubble-cloud' feeding, a complex but common form of feeding in humpbacks in which prey schools are enveloped in clouds of bubbles formed by exhaling underwater (for the diversity of humpback feeding techniques, see Hain et al. 1982); the behaviour is modified by slamming the tail-flukes onto the water (termed lobtailing) prior to diving. The spread of the behaviour is known in some detail since it was recorded over a nine year period in individuals known from photo-identification, and in these details are clues to the transmission process (Weinrich et al. 1992). The increase in the numbers of animals showing lobtail feeding was due primarily to animals born into the study population using the technique (many of these had mothers which did not show the technique, thus genetic determination is unlikely), although some adults also adopted the method. Figure 1 shows that this pattern of spread strongly suggests social learning, given that it is best fit by an accelerating function as is expected under social learning (Lefebvre 1995), although a series of independent individual learning events cannot be entirely ruled out (Weinrich et al. 1992). Although the innovation of lobtail feeding followed a shift in diet accompanying a change in the distribution of prey species (Weinrich et al. 1992), the change was in proportional use of different food sources, not the introduction of a novel environmental element. Given that the technique is a modification of pre-existing methods, these observations suggest the potential for a 'ratchet effect' (Tomasello 1994) in culturally transmitted feeding behaviour.  

The spread of novel feeding methods through a population has been documented for a number of terrestrial and avian species (Roper 1986). Two of the most famous cases are milk-bottle top opening by birds in Britain (Fisher & Hinde 1949), and washing sweet potatoes by Japanese macaques (Macaca fuscata) (Kawai 1965). In both cases, the spread was thought to be due to imitation, but more recent work has cast doubt on this (Sherry & Galef 1984; Whiten 1989). Rates of spread of the innovations were similar to those observed for the lobtail-feeding humpbacks: milk-bottle opening took 20 years to spread across London and potato washing spread through almost all the band of macaques in 9 years.

When mother and offspring have similar, but characteristic, patterns of complex behaviour, this suggests vertical cultural transmission through imitation, teaching, or other forms of social learning. However, genetic determination or shared environments leading to parallel individual learning are also potential explanations in some cases. The limitations of current field studies on cetaceans mean that only seldom are mother-offspring relationships known among adults.  

Genetic and photo-identification studies have shown that young beluga (Delphinapterus leucas) and humpback whales follow their mothers on initial migrations between breeding and feeding grounds, and then repeat them faithfully throughout their lives (Katona & Beard 1990; O'Corry-Crowe et al. 1997). In both species the segregation of mitochondrial DNA haplotypes (a haplotype being any given DNA sequence - individuals with the same mitochondrial haplotypes have the same DNA sequence in their mitochondrial genome; mitochondrial DNA is found not in the nucleus but in the mitochondria, and is inherited through the maternal germ line) by migration routes suggest strong maternal migratory traditions (Baker et al. 1990; O'Corry-Crowe et al. 1997); in humpbacks this hypothesis is supported by photo-identification data showing calves returning to their mother's feeding grounds, even though several feeding stocks apparently intermingle on breeding grounds - making genetic inheritance unlikely (Clapham & Mayo 1987). These cases have obvious parallels in the migratory behaviour of some birds (e.g. Healey et al. 1980).  

Bottlenose dolphins in Shark Bay, Australia, carry sponges on their rostra (Smolker et al. 1997). The exact function of 'sponging' is not known; it is thought to be a foraging specialisation (Smolker et al. 1997), but could carry little obvious adaptive significance, making it directly comparable to stone-handling in Japanese macaques (Huffman 1996). What is interesting from a cultural perspective is that in a population of over 60 individuals known from the 150km² study site, only five sponge regularly; it was been observed only four times in other individuals over a 6 year study period (Smolker et al. 1997). The sponging dolphins are also unusual in their social behaviour; all female, they are markedly more solitary in their habits than the other dolphins in the population and were not observed in any large social groups during the study period. While sponging is only seen in sandy parts of the mixed sand/sea grass habitat, all members of the population experience the same mixed habitat, such that ecological explanations for this behavioural variation can be discounted. Other members of the population are seemingly aware of the technique, as evidenced by the occasional observations of sponging in other individuals, but they do not fully adopt it; thus variation is unlikely due to any genetic ability. However, the calf of one of the regular spongers itself sponges, suggesting vertical cultural transmission. Dolphins in Shark Bay also show another 'foraging specialisation' - feeding by humans at Monkey Mia beach - in which not all of the population takes part. This variation also appears to be maintained by vertical cultural transmission, since most of the dolphins taking advantage of the feeding are offspring of females which were themselves fed (Smolker et al. 1997); hence the 'specialisation' is likely learned while swimming with the mother.  

The transmission of feeding specialisations from parent to offspring is fairly common in other animals. For example, it has been documented in oystercatchers, Haematopus ostralegus, (Sutherland et al. 1996), rats, Rattus rattus, (e.g. Terkel 1996), and chimpanzees (Nishida 1987; Whiten et al. 1999), although the remarkable diversity of feeding specialisations in bottlenose dolphins both within and between areas (Shane et al. 1986; Würsig 1986) is comparable only with chimpanzees.  

Mother-offspring similarity in feeding behaviour is also known from killer whales, particularly the dramatic case of intentional stranding on beaches to catch pinnipeds (Baird 1999). This behaviour has been sufficiently well studied that it provides good evidence for teaching, and so is discussed below under transmission mechanisms.

 The principal ethnographic approach to the study of culture has been the contrasting of behavioural patterns between stable social groups of animals (e.g. Whiten et al. 1999). However social groups, especially in primates, usually occupy distinct habitats and are genetically related, leading to criticism that such ethnographic patterns may be the results of individual learning in different environments, or may have been caused by genetic differences (Galef 1992). Primatologists address such arguments by examining correlations between presumed cultural variants and environmental features or phylogenetic relatedness (Whiten et al. 1999). In the two species of cetacean with matrilineal social systems where group-specific behavioural patterns have been explicitly studied, killer and sperm whales, environmental causation is easily dismissed as groups showing distinctive behavioural patterns are often sympatric, sharing the same habitat and frequently interacting (Baird 1999; Whitehead & Weilgart 1999). We use the term 'group' here sensu Connor et al. (1998) - a set of animals with consistently stronger associations with each other than with other members of the population over periods of months to decades.   

Addressing potential genetic causation of behavioural differences between groups is more complex, as some presumed cultural traits in the matrilineal groups of these species seem to be sufficiently stable that they can show occurrence patterns very similar to parts of the maternally-inherited mitochondrial genome (Whitehead et al. 1998, Figure 2). However, except in the case of the different forms of killer whale (see below), mating appears to occur across behavioural variation 'boundaries', (Ohsumi 1966; Baird 1999). Thus, stable group-specific behavioural traits, if genetically determined, would have to lack paternal inheritance, as conventional biparental genetic transmission via the nuclear genome would lead to hybrid behaviour and scramble group-specific patterns (Whitehead 1999a). It is unlikely in the extreme that mitochondrial DNA would code for behaviour. It is theoretically possible for behavioural variants to be encoded in the nuclear genome, and give rise to the observed patterns, if the encoding genes were subject to some form of genomic imprinting, where only the alleles from one parent, in this case the mother, were expressed (see Barlow 1995; Spencer et al. 1999). However, such genomic imprinting systems are typically involved in embryonic development, and are thought to be the result of genetic conflict for developmental resources (Spencer et al. 1999); it is hard to see how such a system could have evolved in the case of group specific behaviour. Vertical (or oblique within matrilines) cultural transmission is an obviously analogous process to the maternal inheritance of mtDNA which could easily lead to strong mtDNA-behaviour correlations.  

Several species of cetacean live in stable social groups (Connor et al. 1998); of these the best known is the killer whale, particularly those that live around Vancouver Island. There are at least two different forms of killer whale in this area, which are sympatric but can be distinguished by diet, morphology, behaviour, social structure and genetics (Baird 1999). Although they are known as 'residents' and 'transients', this terminology does not really reflect the habits of the two forms (Baird & Dill 1995). Best known is the fish-feeding, 'resident', form. 'Residents' live in highly stable matrilineal 'pods' averaging 12 animals (Bigg et al. 1990); there is no known case of individuals changing pods in over 21 years of study (Baird 1999). In contrast 'transients' live in smaller pods, averaging 3 animals (Baird 1999), which appears to be the more typical case for killer whales world-wide (Heimlich-Boran & Heimlich-Boran in press). 'Transient' killer whales do occasionally leave their natal pods and travel temporarily with other 'transient' groups. The study of this species off Vancouver Island and in other areas has produced evidence for considerable behavioural variation between social groups.

The strongest evidence lies in the vocal dialects of 'resident' pods; each pod has a distinctive set of 7-17 'discrete' calls (Ford 1991; Strager 1995). These dialects are maintained despite extensive associations between pods. Some pods share up to 10 calls (Ford 1991) and pods which share calls can be grouped together in acoustic 'clans' (Ford 1991), suggesting another level of population structure. Ford (1991) found four distinct clans within two 'resident' communities, and suggested that the observed pattern of call variation is a result of dialects being passed down through vocal learning, and being modified over time. Thus, given the lack of dispersal, acoustic clans may reflect common matrilineal ancestry, and the number of calls any two pods share may reflect their relatedness (Ford 1991). So how reasonable is the assumption of vocal learning in the face of the alternative hypothesis that dialects are genetically based? Although the question has not been directly addressed, there is some evidence that killer whales are capable of vocal learning, (Janik & Slater 1997 and see section 3.1). In a fine scale analysis of call variation over time within pods, Deecke (1998) showed that individual call types accumulated modifications over a 12 year period within two pods, and that these modifications did not result in a divergence between the two pods, implying that some mechanism is preventing divergence while modification takes place (the most likely is horizontal cultural transmission); both of these findings are evidence against genetic determination of dialect in killer whales. Finally, for genetic determination of pod-specific dialects when most mating appears to take place between pods (Baird 1999), some highly unusual genetic system without paternal inheritance would be needed (as discussed above).

Between-pod variation is also evident in other aspects of killer whale behaviour, particularly foraging. Baird and Dill (1995) found strong variation in the use of their study area by 'transient' pods - some pods were seen primarily during the harbour seal (Phoca vitulina) pupping period, apparently specialising on foraging around haul-out areas during the pupping season, while others were seen foraging away from haul-outs all year round. There are strong indications that different sympatric 'resident' pods specialize on different salmon species (Oncorhynchus spp.), evidenced by correlations in the abundance of different salmon species and killer whale pods at various locations; it has been suggested that accumulated knowledge of salmon distribution results in the traditional use of specific areas by different pods (Nichol & Shackleton 1996). 'Resident' predation on marine mammals is extremely rare compared to 'transients', but interestingly, of the handful of observations of 'resident' killer whales 'harassing' marine mammals, all but one (10/11 combined observations from Osborne (1986) and Ford et al. (1998)) were by a single pod, L01. In a study of Norwegian 'residents' and their interaction with herring (Clupea harengus), Similä et al. (1996) reported pod-specific variation in migration patterns as indicated by area use, while Similä and Ugarte (1993) describe a cooperative hunting technique (carousel feeding) not seen in any other killer whale population. The feeding techniques of killer whales are as variable and adaptable as those of the bottlenose dolphin; in addition to the techniques we describe here, they also take a wide variety of other cetaceans (Jefferson et al. 1991), pinnipeds (e.g. Smith et al. 1981) and elasmobranchs (Fertl et al. 1996; Visser 1999), using a range of often complex and cooperative hunting techniques. This variability and adaptability in feeding techniques has also allowed killer whales to take advantage of new anthropogenic food sources as they become available - for example the discards of trawlers (Couperus 1994). In the Bering Sea killer whales take fish from long-lines (Yano & Dahlheim 1995); of 19 known pods in the Prince William Sound area, only 2 are known to take fish in this way (Yano & Dahlheim 1995), another example of sympatric behavioural variation.

Other behavioural patterns vary between higher-level groups of killer whales. Off Vancouver Island, there are community-specific 'greeting ceremonies' observed when 'resident' pods of one community meet (Osborne 1986); the two pods line up facing each other and stop in formation for 10-30 seconds before approaching and mingling. Some pods of another community engage in 'beach-rubbing', and again there is variability between pods in the preferred locations for rubbing (Hoyt 1990). All of this variation should be considered in the context of the markedly low genetic variability within the 'resident' and 'transient' communities (compared to other cetaceans) found by Hoelzel et al. (1998).  

Sperm whales make distinctive, stereotyped patterns of 3->12 clicks termed 'codas', which are thought to function in communication (Watkins & Schevill 1977). Distinctive coda dialects (consisting of very different proportional use of about 30 different types of coda) are a feature of partially matrilineal, but interacting, groups of about 20 female sperm whales (Weilgart & Whitehead 1997). Given the wide ranging movements of these animals - of the order of 1,000km (Dufault & Whitehead 1995) - these dialects are effectively sympatric. Among 6 sperm whale groups, there was a strong and significant correlation between inter-group dialect similarity and the similarity of their mitochondrial DNA (mtDNA) haplotypes - groups with similar coda dialects also had similar mtDNA (Whitehead et al. 1998, Figure 2). The existence of this correlation implies that mitochondrial haplotype and coda dialect are transmitted by analogous processes through the female line and show a similar order of stability. However, it also presents a conundrum as sperm whale groups are not themselves particularly stable, often consisting of two or more largely matrilineal units which swim together for periods of days (Whitehead et al. 1992; Richard et al. 1996; Christal 1998). These social units may themselves split or merge (Christal et al. 1998). How then can the groups possess highly stable dialects? Possible resolutions (Christal 1998; Whitehead 1999a) include the possibility that the coda repertoire of a group is largely determined by its numerically dominant social unit; the fact that the results on non-matrilineality of sperm whale units are based on studies of just a few units in Galápagos/Ecuadorean waters which may have been fragmented by intense whaling from Peru; the transfer of individuals between units may occur within larger, currently unrecognized, cultural trait-groups (such as the acoustic clans of killer whales, Ford 1991); that transferring individuals may have low reproductive success; and that conformist traditions maintain cultural stability within groups in situations when groups frequently interact (see below).  

There are also indications of non-vocal group-specific behaviour in sperm whales. Group identity accounted for an estimated 40% or more of the variance in 6/18 measures of the visually observable behaviour of sperm whales off the Galapagos Islands when environmental and temporal variation had been considered (Whitehead 1999b). These measures were mostly (5 out of 6) concerned with how the groups used space: heading consistency, inter-animal distance, and straight-line distance moved in daylight. The results of this study should be interpreted cautiously for a number of reasons (Whitehead 1999b). For instance, only one measure (straight-line distance moved in 12 hours) was the group-specific effect statistically significant at P<0.05 following corrections for multiple comparisons, although statistical tests for group-specific effects had little power (group identity had to account for >~65% of the variance in a measure to produce a significant effect at P<0.05).  

In a study spanning the South Pacific there were significant differences in the number of predator-inflicted marks on the tails of sperm whales between matrilineal groups (Dufault & Whitehead 1998). As with coda repertoire, the marks possessed by a group were correlated with the group's predominant mitochondrial haplotype (Whitehead et al. 1998). One explanation for this surprising result is that, like coda repertoire, methods of communal defence against predators are passed down culturally in parallel with the mitochondrial genome (Whitehead et al. 1998): sperm whale groups have been observed to defend themselves against killer whale attack in some instances by forming a tight rank and keeping their heads and jaws towards the killers (e.g. Arnbom et al. 1987), and in others by putting their heads together and allowing their bodies to radiate outwards in a 'wagon-wheel' formation so that their tails face the predators (e.g. Pitman & Chivers 1999) - although there is no data available on whether groups consistently use one or other strategy. Groups adopting the wagon-wheel formation would tend to accumulate tail markings much more readily than those that defended with their jaws.

In addition to these group-specific patterns of killer and sperm whales, there are some local behavioural patterns of cetaceans which do not live in such stable groups. Bottlenose dolphins at Laguna off the coast of Brazil have an unusual group-specific feeding technique which seems to date from 1847, and to have been transmitted within a matrilineal community since at least three generations of dolphin are involved (Pryor et al. 1990). The 25-30 dolphins and local fishers follow a strict protocol - involving no training nor commands from the fishermen - that allows the humans and dolphins to coordinate their actions. The dolphins drive fish into the nets of human fishermen, indicating as they do so when the humans should cast their nets (by performing a distinctive 'rolling' dive; the humans can also pick up from how much of the body comes out of the water on this roll an idea of how many fish are present - it is entirely unclear whether this cue is given intentionally or not) - and then feed off the fish that are stunned or missed by the net (Pryor et al. 1990). There are other bottlenose dolphins in the area which do not participate in the cooperative fishing, and sometimes try to disrupt it (Pryor et al. 1990) - hence again behavioural variation is sympatric. Only young adults whose mothers took part in the fishing later adopted it themselves, although not all the offspring of fishing mothers did so (Pryor et al. 1990). There are several other accounts of such cooperative fishing on different continents (Pryor et al. 1990). For example Irrawaddy dolphins (Orcaella brevirostris) in the Ayeyarwady River, Myanmar, have a similar, generations-old, cooperative relationship with local fishers (Anderson 1879; Smith et al. 1997).  

Group-specific culturally-transmitted behavioural patterns have parallels in other animal taxa, but in some respects these cetacean cultures, in particular those of the killer whale, appear unique outside humans. Within a number of primate species, including macaques and chimpanzees, bands have characteristic cultures, which include diet, familiarity with a home range, social signals, relationships, and, in the case of chimpanzees, tool use (Russell & Russell 1990; Whiten et al. 1999). Similarly populations of birds of a single species sing different dialects of the species-specific song (Baker & Cunningham 1985). However, these dialects and cultures are geographically based: animals in one place behave in one way, those in another behave differently. In contrast, different cultural variants of killer and sperm whales, as well as the cooperative fishing traditions of bottlenose dolphins, are sympatric and animals with different cultures often interact. Thus, members of these species are repeatedly exposed to a wide range of cultural variations, but maintain their own group-specific culture. Somewhat similar phenomena have been observed in the flock-specific calls of some birds (e.g Mammen & Nowicki 1981; Feekes 1982); however these flocks are not stable for more than a few months and hence do not support persistent cultures. Greater spear-nosed bats (Phyllostomus hastatus) modify their 'screech' calls to establish and maintain differences between stable social groups which share caves (Boughman & Wilkinson 1998) - hence demonstrating sympatric cultural variation; however, this variation is based only on modifications of a single call, and so does not approach the complexity of group-specific behaviour seen in killer whales. Another contrast is that the behavioural complexes seen in killer whales appear to encompass both vocal and physical behaviours; such complex multicultural societies where culture encompasses both the vocal and motor domains are otherwise only known from humans. However, it should be noted that in no cetacean example is there evidence of such broad suites of cultural behaviours as have been found in chimpanzees, where 39 behaviour patterns have been shown to vary culturally (Whiten et al. 1999).  

A second remarkable attribute of some of the group-specific cultural traits of cetaceans is in their stability. Killer whale dialects are highly stable, known to persist for at least six generations and it has been suggested much longer (Ford 1991). To give rise to the strong dialect-mtDNA correlations which seem to be present in sperm whales, vocal culture must be stable over many generations (Whitehead 1998). Such stability is not a feature, as far as is known, of comparable cultures outside humans (Feldman & Laland 1996). Some song-bird dialects last over 10 years, but these are apparently not related to stable social groups (e.g. Trainer 1985).  

 What kinds of social learning are cetaceans capable of? Some scientists (e.g. Galef 1992; Tomasello 1994) will only admit culture when it can be shown that behavioural patterns are being transmitted between animals by either imitation or teaching, and not by other types of social learning such as stimulus enhancement (in which individual learning is enhanced when one animal directs the attention of another towards a stimulus). While we do not subscribe to this particular view, we do think that understanding process (cultural transmission) is crucial to our understanding of the product (culture) - for example some forms of cultural transmission may be more likely to produce cultures capable of feeding on themselves, producing the "ratchet effect" (Tomasello 1994) - and so it is appropriate to ask whether cetaceans are capable of higher-order social learning such as imitation or teaching.  

In the wild, cetaceans are mostly underwater and out of view, while only the smaller species, such as bottlenose dolphins, can be kept in captivity for more than a short period. Even then, the social and ecological environment of the captives is far from natural. Therefore, there is little concrete evidence for imitation or teaching by cetaceans. For instance, it has not been experimentally proven that maturing killer whales learn their group-specific call dialects (Tyack & Sayigh 1997); only cross-fostering experiments could establish this, and given the expense and difficulty of raising killer whales in captivity even without cross-fostering, these experiments are never likely to be performed regardless of the ethical issues involved. We are also limited by the fact that the vast majority of experimental work has been performed on a single species - the bottlenose dolphin. Despite these substantial difficulties, there is some good evidence that cetaceans can both imitate and teach.  

 Bain (1986) describes a captive killer whale from Iceland learning, over a three year period, the vocal repertoire of its tank-mate from British Columbia, while Ford (1991) presents evidence for inter-pod call mimicry in the wild - both suggesting that killer whales are capable of vocal learning. Bowles et al. (1988) followed the ontogeny of vocal behaviour in a single captive killer whale. During the study, the calf was housed with its mother, another female companion, and a young male; the calf never had any contact with its father. By 398 days, the calf's output was dominated (90% of output) by the one call type which distinguished its mother's repertoire from that of the female companion (the two adult females shared other call types) even though 82% of all the calls recorded from the tank were not of that type; Bowles et al. (1988) suggest that this is due to selective learning by the calf. However, they recognise that their study cannot exclude genetic effects, although there was no trace of the father's dialect in the calf's repertoire, hence no evidence for the hybrid dialect expected under genetic determination. When viewed as a whole, the combined evidence points clearly to imitation as the transmission mechanism of vocal repertoire. Hence, it seems very likely that killer whale call dialects are, as Ford (1991) suggests, cultural institutions, even from the perspective of those who believe cultural transmission should be restricted to teaching and imitation.  

There is less direct evidence for social learning in the other good example of group-specific cetacean dialects, the sperm whale coda repertoire. However, in the remarkable 'echocodas', two animals precisely interleave their click patterns, giving rise to two overlapping codas, identical in temporal pattern to within a few milliseconds, offset by about 50-100 milliseconds (Weilgart 1990). This 'duetting' suggests that sperm whales may be matching codas in a similar way to bottlenose dolphins matching signature whistles (see Tyack 1986b); such matching would require imitative learning in some form.  

Experimental work on the learning abilities of cetaceans has been largely confined to a single species, the bottlenose dolphin, mainly because this species makes up the vast majority of the captive population. A number of studies have shown that this species is clearly capable of vocal and motor imitation (Richards et al. 1984; Richards 1986; Bauer & Johnson 1994; Kuczaj et al. 1998). Anecdotal examples include the imitation of the movements and postures of a pinniped by a dolphin sharing the same tank described by Tayler and Saayman (1973). Imitative abilities include 'simple' imitation - 'action level imitation' in the terminology of Byrne and Russon (1998) or 'copying' in Heyes (1994) - and 'true' imitation - 'goal emulation' in Whiten and Ham (1992) or 'program level imitation' in Byrne and Russon (1998). An example of the former is the imitation of motor patterns (for example shaking the head from side to side) described by Bauer and Johnson (1994); of the latter, the imitation of functional tool use described in Kuczaj et al. (1998). Such examples are deserving of more recognition than they have previously been given in discussions of animal imitation; evidence for imitation in dolphins is a consistent 'thorn in the side' (Byrne & Russon 1998) of those who deny imitation in non-humans (e.g. Galef 1992; Tomasello 1994). The grey parrot (Psittacus erithacus) rivals the bottlenose dolphin in social learning ability, being capable of both vocal and movement imitation (see Moore 1992; 1996); however, it has yet to be shown whether parrots are capable of program level imitation. Thus, to our knowledge, bottlenose dolphins are the only nonhuman animal for which both vocal imitation, and motor imitation at both action and program level, have so far been demonstrated (Herman 1986; Kuczaj et al. 1998).  

3.2 Observations of teaching in killer whales

Killer whales in the Crozet Islands and off Punta Norte, Argentina, swim ashore to capture pinnipeds (Lopez & Lopez 1985; Guinet 1991; Hoelzel 1991; Guinet & Bouvier 1995). In the clearest descriptions of the social learning process in wild cetaceans, Guinet and Bouvier (1995) describe young killer whales learning from their mothers, and sometimes other animals, the feeding technique of intentional stranding on pinniped breeding beaches. This method of feeding is profitable but risky; one of the calves in Guinet and Bouvier's (1995) study was found permanently stranded and facing death until observers returned it to the water. As the behaviour of adult killer whales towards juveniles during intentional stranding appears to (unusually for non-humans) fit definitions of teaching (Baird 1999; Heimlich-Boran & Heimlich-Boran in press), the details bear repeating here. We take Caro and Hauser's (1992) definition of teaching as modifying behaviour, at some cost or lack of benefit, only in the presence of a naïve observer such as to encourage, punish, provide experience or set an example, such that the observer acquires a skill more rapidly than it might do otherwise, or may not ever learn. While Caro and Hauser (1992) considered the evidence for teaching in killer whales 'weak', this was before the publication of Guinet and Bouvier's (1995) study. The evidence now is considerably stronger, although it is based on observations of only two calves, so we are unaware of how wide-spread the process may be.  

Adult killer whales have been observed pushing their young up the beach, then back down the beach, directing them towards prey, helping them out when they become stuck by creating wash, helping them back to deep water after a successful capture (Guinet & Bouvier 1995), and throwing prey at juveniles (Lopez & Lopez 1985) - hence they modify their behaviour in the presence of naïve observers. Adults are more successful at hunting in the absence of juveniles (Hoelzel 1991); at the extreme they throw away already captured prey (Lopez & Lopez 1985) - hence there is a demonstrable cost. Pushing juveniles onto beaches and pushing them towards prey is clearly encouragement. As to whether observers learn better with or without instruction, consider Guinet and Bouvier's (1991; 1995) study in the Crozet Islands. They followed the development of hunting by intentional stranding in two killer whale calves, A4 and A5. At the start of the study, the calves estimated ages were 4 and 3 respectively, and they were observed taking part in 'beaching play' (stranding, with their mothers and/or other adults, on beaches devoid of elephant seal, Mirounga leonina, prey); occasionally they would also strand during predation attempts by adults. Both calves were observed to strand alone for the first time at age 5. Near the end of the three year study calf A5, then aged 6, was observed successfully catching a seal pup. A4, although a year older, had not been observed hunting successfully by the end of the study. This difference between the calves is very interesting; during beaching play, A5 stranded exclusively with its mother, while A4 stranded only twice in 35 observations with its mother (Table 2). A4's mother rarely took part in beaching play, and was not observed hunting in this way. In contrast, A5's mother closely supervised its strandings. The mother was observed pushing the calf up the beach and stranding onshore in order to push the calf back into the water, accompanying the calf on unsuccessful hunting attempts and finally assisting in the first successful capture by pushing the calf towards the prey and helping the calf to return to the water following capture. Hence the behaviour of A5's mother seems to have enabled her calf to learn the hunting technique at least one year earlier (aged 6) than A4, which received very little 'instruction' - so the behaviour apparently results in a skill being learned more rapidly than it otherwise would. It is not known whether A4 ever learnt to hunt successfully this way; interestingly, A4 was the previously mentioned calf found permanently stranded and facing death, suggesting a severe fitness cost for mothers which do not give their calves much attention. Clearly, according to accepted definitions, killer whales teach. Caro and Hauser (1992) point out the rarity of such overt encouragement in other animals.  

 Does the evidence we present here legitimately allow us to attribute culture to cetaceans? We recognise that how one defines culture will inevitably affect how one attributes it, and we also recognise that we have chosen a broad definition of culture. However, we think there are good reasons for choosing such a definition, which we will attempt to explain here.  

From our evolutionary perspective, the important questions surrounding culture in humans and animals concern how cultural faculties and the behavioural complexes to which they give rise (which we would call cultures by our definition) vary in extent and form within and across species, and how this may be related to evolutionary ecology. Along with Mundinger (1980), we find that there is no "empirical evidence for any qualitative difference that would support a basic human/non-human dichotomy" - many social learning processes apparently play an important role in supporting human culture (Midford 1993; Olsen & Astington 1993; Rogoff et al. 1993; Boesch 1996; Plotkin 1996), but there are also animals, including bottlenose dolphins, that are capable of sophisticated social learning - in particular imitation. Hence we cannot support any definition which renders culture by definition something only humans can have. Surveying cultural transmission in nature, one finds what approaches a continuum between animals which acquire only a single behavioural pattern culturally (e.g. bluehead wrasse, Thalassoma bifasciatum, mating sites, Warner 1988), through animals which acquire suites of behaviours through cultural processes (including chimpanzees and, data strongly suggest, killer whales) to humans where culture has enabled us to radically alter our own environment. Given this continuum, it seems to us that the question of culture is more likely a question of extent - just how much of the behavioural repertoire must be culturally determined before a population can be said to show 'culture'? We would maintain that drawing a line on this continuum and labeling one side 'culture' and the other 'not culture' is essentially an arbitrary exercise, leading to the current variability in attributions of culture to non-humans. Instead we adopted a definition which has allowed significant progress to be made in developing a theoretical basis for understanding culture, and which is not tied to any particular species nor any particular form of culture. Such a broad definition allows us to concentrate on comparing cultures across species, and relating these comparisons to ecology.  

We have rejected an exclusively process-centred definition of culture. Such a stance contains a number of serious weaknesses. Culture has been incorporated into theoretical models in a way which is essentially process independent; as long as information is transferred between individuals extra-genetically, then, while the transmission process will affect the parameters of cultural evolution (for example the speed of acquisition of a novel trait, or the stability of resultant traditions), it does not affect the basic co-evolutionary process. To define culture in terms of the transmission mechanisms upon which human cultures depend (see Tomasello et al. 1993) is from an evolutionary perspective counter-productive and anthropocentric (de Waal 1999). We concur with Plotkin's (1996) assertion that 'dual inheritance can be, and almost certainly is, served by more than one form of social learning'; Whiten and Ham (1992) explicitly list a range of social learning processes from 'exposure' to 'goal emulation' as supporting cultural transmission. Human cultures depend on human cultural learning; teaching and imitation may be unique to humans (although there is good and growing evidence that it is not, and cetaceans are a group which has given rise to some of the strongest evidence, as presented above) and hence culture as a human trait has unique properties, the material evidence of which surrounds us. However, to then define culture as a general trait in terms of human transmission processes is, in our view, a mistake; it is akin to defining locomotion as a general trait in terms of walking on two legs; this is how humans move, but other animals achieve the same effect (moving from A to B) with a huge variety of different locomotion processes, some of which are energetically more efficient, or quieter, or faster than others (de Waal (1999) independently gives a similar argument). Clearly it is wrong to say that animals do not show locomotion if they do not move using a bipedal gait; in our view it is equally wrong to say that animals showing stable behavioural variation independent of ecology and genetics and transmitted through social learning do not show culture if that behavioural variation is not transmitted using teaching or imitation. We cannot agree with Tomasello's (1994) argument that since 'behavioral traditions of humans provide the prototypical case of cultural transmission' then human culture should be the benchmark against which all else is compared. Instead, we concur with Boesch (1996) that 'it seems rather arbitrary to single out one process of information transmission as the only one able to produce culture'.

We see other weaknesses in the process-centred definition. Concentrating on process reduces the issue of culture to a question of whether or not a species can imitate or teach in an experimental setting, as opposed to other social learning mechanisms, such as stimulus enhancement. However, there is much conceptual confusion surrounding imitative and non-imitative social learning; it is not clear how the bewildering taxonomy of terms (e.g. Galef 1988) for various social learning mechanisms relate to each other, nor that the underlying conceptual approach is really satisfactory - many of the categories are based on unobservable and ill-defined mechanisms, are not mutually exclusive and give little or no information regarding conditions for occurrence or functional significance (see Heyes 1994). The 'cross-talk and confusion' (Heyes 1996) surrounding the taxonomy of social learning processes, up to and including various forms of imitation, weakens the process-centred approach to culture; the different approaches of Heyes (1994), Zentall (1996), Tomasello et al. (1993) and Byrne and Russon (1998) have yet to be resolved into a coherent understanding of social learning. This becomes serious if, as tends to occur, observational learning (and hence culture) is rejected as an explanation for field observations until and unless all other social learning mechanisms are experimentally excluded (Boyd & Richerson 1996); we do not accept this to really be a sound approach given the lack of consensus surrounding social learning. Moreover the implicit assumption that behaviour is culturally acquired in humans, and that it is not in non-humans (McGrew 1992, p.14, p.197, p.217) leads to different null hypotheses for considering culture in humans and animals; the apparent lack of culture in animals may be due more to the placement of the burden of proof than anything else, leading to the criticism that 'a double-standard is being applied' in this human-centred approach (Boyd & Richerson 1996), and the heuristic weakness which has led this approach 'down paths of steadily decreasing interest to the rest of the community of life scientists' (Galef 1992).  

Until and unless cetacean traditions are proven experimentally to rely on teaching or imitation, those who restrict culture to imitation and teaching will deny culture to these animals. The final weakness of such an approach is that the necessary experiments will likely never be performed given the expense and difficulty of keeping, let alone raising, most cetaceans in conditions which are both sufficiently controlled for valid experiments and sufficiently naturalistic so that the animals may show realistic social behaviour. For instance, social groups of sperm whales will never be kept in captivity; the logistics are simply not feasible. Ethnographic data and field observations are all that is, and likely ever will be, available for such species. Here is a central problem with an experimental approach to cetacean culture: it can freeze the question by demanding that which will never occur (i.e. experimental studies). In our view, it is not reasonable to postpone discussion on this semi-permanent basis, when other rigorous and conceptually sound approaches are available. This is not to say that the study of social learning mechanisms is not important - as we have already stated. Human culture shows extraordinary characteristics when compared with animal culture, in particular concerning its linguistic, material and symbolic extent, and it is likely that the mechanisms by which it is propagated contribute to this uniqueness. For example, accumulating modifications over time, the so-called 'ratchet effect' (Tomasello 1994), may be greatly facilitated by enhanced observational learning abilities in humans (Boyd & Richerson 1995; 1996; Henrich & Boyd 1998).  

The field-based approach to culture is exemplified by Boesch et al.'s (1994) and Whiten et al.'s (1999) work on chimpanzee culture, Grant and Grant's (1996) work on Darwin's finches and Warner's (1988) work on bluehead wrasse. The approach is clear; systematic field observation (and manipulation of natural populations in Warner (1988)) enables the elimination of ecological and genetic factors potentially causing behavioural variation; what is left must be cultural. The resulting conclusions are weak in that the transmission process remains unproven but strong in that they are firmly rooted in how the animals actually behave in the wild. Since cultural learning is social learning, we can only fully appreciate its complexity and functional usage in animals when it is studied in a naturalistic social setting. The cultural hypothesis is strengthened if the behaviour under scrutiny varies non-adaptively or arbitrarily (Boesch 1996) since both ecological and genetic factors are more likely to produce adaptive variation, whereas culture can produce maladaptive behaviours (Boyd & Richerson 1985) or influence otherwise selectively neutral variation in behaviour (e.g. bird song, Baker & Cunningham 1985).  

This ethnographic perspective heavily influences our approach to culture. Our review suggests that at least some cetaceans are adept social learners (see also Heimlich-Boran & Heimlich-Boran in press). It seems to us most likely that these abilities, and not genetic or environmental causation, have given rise to the conspicuous patterns of rapid-spread, mother-offspring similarity, and group-specific behaviour listed in Table 3. In a few cases, such as sponge feeding and the use of human provisioning by bottlenose dolphins, it is possible to envisage scenarios of environmental change and individual learning giving rise to the observed patterns. However, the continuously evolving songs of humpback and bowhead whales have no conceivable environmental or genetic cause, and if the characteristic dialects and behaviour of the matrilineal groups of killer and sperm whales were genetically determined, there would have to be little or no paternal inheritance, a highly unusual process. Thus, from the ethnographic perspective, we believe that most, if not all, of the patterns listed in Table 2 can be ascribed to cultural transmission.  

In the case of the killer whale there are strong indications that groups possess suites of distinctive, interlocking cultural characteristics, as so far described only for humans and chimpanzees (Whiten et al. 1999). However, one obvious difference between these cultures is that there is no evidence at all for preservable material culture in cetaceans compared to both humans and chimpanzees (c.f. McGrew 1992). Human culture is intimately linked to both language and symbolism, but there is currently no empirical basis for discussing the role or non-role of language and symbolism in cetacean culture - bottlenose dolphins have been taught artificial 'languages' (e.g. Herman et al. 1993), but such work tells us little about the role of communication in the natural situation (Tyack 1993). Cetacean cultures do appear to possess other attributes that have otherwise been restricted to humans. In particular, we are aware of no phenomena outside humans comparable to the distinctive, stable and sympatric vocal and behavioural cultures which appear to exist at several levels of killer whale society.  

Some cetaceans then, seem to have evolved cultures that closely parallel those found in both chimpanzees and humans. What is perhaps surprising is that all four of the best studied cetacean species show strong evidence, from either the experimental or ethnographic approach, for social learning. Why? It is true that they possess those biological attributes which Roper (1986) suggests favour social learning: long lifetimes (~20-90yr), advanced cognitive abilities, and prolonged parental care (Tyack 1986a; Herman et al. 1994; Marten & Psarakos 1995). However, while there probably is a minimum cognitive capability required for social learning, the relative success of those individuals within a given species which are better than average at social learning likely depends ultimately on the ecological situation in which that individual must make a living - thus it is more important to look to ecology when attempting to explain species differences in social learning (Lefebvre & Palameta 1988). We think that ecological factors may have a strong role to play in explaining the social learning abilities and culture to which they give rise in cetaceans. Whitehead (1998) suggests that the structure of the marine environment may have favoured the evolution of cultural transmission in cetaceans. Here we explore this suggestion in more detail.  

Compared with life on land, marine ecosystems are more likely to switch into alternate states over time scales of months or longer (Steele 1985). This increased low-frequency temporal variability of marine systems may significantly increase the adaptiveness of culture to cetaceans, as the benefits of cultural transmission, relative to individual learning or genetic determination, are thought to be strongly related to environmental variability (Boyd & Richerson 1985; 1988; Laland et al. 1996). The scale of spatial variation may also be important; spatial autocorrelation in oceanic ecosystems weakens at ranges of about 500km (Myers et al. 1997), so that one way to deal with radical changes in the environment in any place is to move a few hundred kilometres. Many marine organisms are adapted to particular environments in which they can flourish, but also have long-range dispersal of large numbers of eggs, larvae or juveniles which allow them to colonize suitable distant ocean areas, and so to persist when conditions deteriorate in any place (Steele 1985). Long-lived marine animals with low reproductive rates similarly can use migration to avoid unfavourable conditions (Whitehead 1996). Compared to terrestrial mammals, but not birds, cetaceans have the advantage of much lower travel costs (Williams et al. 1992), and few substantial barriers. Many oceanic cetaceans do appear to use movements over hundreds of kilometres to improve environmental conditions (e.g. Whitehead 1996). For instance, the mean monthly displacement of a female South Pacific sperm whale - ~350km - is roughly ten times that of members of a particularly mobile population of a particularly mobile terrestrial mammal, the African elephant, Loxodonta africana, (Thouless 1995; Whitehead in preperation). The efficiency of these movements could be greatly enhanced by cultural transmission of desirable movement strategies vertically from mother to offspring, horizontally among animals in the same region, or, perhaps especially, between generations within stable groups (Whitehead 1996). Moreover, if the primary benefits of culture accrue from accelerated adaptation to changing circumstances or more rapid expansion into new niches relative to individual learning or genetic change (Boesch 1996; Boyd & Richerson 1996) then these benefits will be accentuated in environments which are more variable, and also in which movement into new habitats is likely or easy, conditions which are both true for cetaceans. For example, chimpanzees live in quite stable ecological situations and have a limited migratory potential and hence the adaptive advantage of culture may not have been as strong as in nomadic hominids (Boesch 1996); perhaps also this is why the evolution of culture seems to have progressed further in some directions among cetaceans than in non-human primates.  

Extensive mobility, while often primarily a function of the need to reduce variation in one key environmental variable (usually food intake, availability of water, or temperature), tends to increase variance in other aspects of an animal's environment, including its social environment. Tyack and Sayigh (1997) argue that the relatively greater mobility of cetaceans may be one reason why they show extensive capabilities for vocal flexibility and vocal learning, while terrestrial mammals do not. Consider group-living species; there are substantial advantages for individual cetaceans living in groups, be it through cooperative foraging (Similä & Ugarte 1993), food sharing (Hoelzel 1991) or communal defence (Arnbom et al. 1987), but there is also the risk of sharing food with, or being injured in defending, individuals who are not members of the same group and hence are unlikely to reciprocate. Group signatures are one way to minimize this risk. However, as Tyack and Sayigh (1997) point out, when highly mobile animals regularly interact with conspecifics of different groups, signature systems need to be flexible and sophisticated - a demand which culturally transmitted dialects meet.  

We can envision an evolutionary trajectory for cetacean cultural learning abilities similar to that proposed for psittacine birds by Moore (1992; 1996), of call learning being generalised to vocal mimicry through to more generalised imitative capabilities - although it must always be recognised that we know virtually nothing about the actual learning mechanisms cetaceans employ. Three of the four species we primarily discuss here are known or very likely vocal learners: the bottlenose dolphin, killer whale and humpback whale. For sperm whales the learning of codas, since it does not involve learning a new sound, only a pattern of known sounds (clicks), may involve contextual rather than strictly vocal learning (see Janik & Slater 1997). In addition, the link from vocal to motor mimicry through percussive behaviour proposed by Moore (1992; 1996) for birds may also be present in cetaceans - almost all species perform behaviour which involve striking the water surface (lob-tailing, flipper slapping) and it is thought that these may sometimes function as acoustic signals (e.g. Norris et al. 1994). So the mobility of cetaceans may have created selection for vocal learning, providing the roots of sophisticated social learning, while the spatial and temporal variability of the marine environment made social learning highly adaptive as a cost-reducing adjunct to individual learning about new niches (see Boyd & Richerson 1995). In long-lived animals which form stable social groups, the opportunities for cultural transmission are greatly increased, and if most other group members are kin, such information exchange would also accrue inclusive fitness benefits - leading, perhaps, to the remarkable cultures of killer whales.  

For the dialect-gene and dialect-ancestry correlations which seem to be present in sperm and killer whales, cultural transmission must be very stable, with cultural traits being passed consistently within matrilines, but very rarely between them (Whitehead 1998). How can this occur when cetacean matrilineal groups frequently meet, interact, and, in the case of sperm and 'transient' killer whales, occasionally receive new members (Christal et al. 1998; Connor et al. 1998; Baird 1999)? Conformist traditions within groups seem to be a vital element of human cultural evolution (Boyd & Richerson 1985); we actively adopt the prevalent cultures of the groups we are members of. Conformity is clearly advantageous to the group as a whole, and thus its members, when the culture refers to coordinated behaviour, such as communal foraging or within-group communication, and can lead to highly stable cultures (Cavalli-Sforza & Feldman 1981). One of the mechanisms with which cultural information could be secured is through 'social norms' (Heyes 1993; Boesch 1996); this idea equates to Boyd and Richerson's (1985) 'conformist transmission'. When group-specific behaviour is generally favoured, then conformist cultural markers of group membership may evolve, reinforcing the conformist transmission of other cultural elements (Richerson & Boyd 2000). Recent theoretical work points to the widespread conditions favouring conformist transmission and suggests a synergistic relationship between the evolution of imitation and conformism (Richerson & Boyd 2000); such an interaction could well have occurred, or be occurring, in killer and sperm whales.  

The two features in which killer and probably sperm whale cultures seem to differ from those of virtually all other non-human animals, stability and multiculturalism, are prerequisites for cultural processes to have much effect on genetic evolution. To affect genetic evolution cultures must usually be stable over many generations (Laland 1992), and if cultural variants rarely interact they will generally have only local effects (Whitehead 1998). There have been two suggestions that substantial gene-culture coevolution has occurred in whales and dolphins; since both involve historical explanation, neither can be empirically proven. However, this is no different from posited cases of gene-culture co-evolution in humans (Feldman & Laland 1996). Both Baird (1999) and Heimlich-Boran and Heimlich-Boran (in press) propose that culturally-transmitted group-specific foraging techniques initiated the divergence of the forms of killer whale, which now show genetic and morphological differences, and may well be in the process of speciation given the apparent reproductive isolation of the two forms (Baird et al. 1992). This is a plausible explanation for the ongoing sympatric speciation; however, since the genetic differences between the two forms are now so evident (Hoelzel et al. 1998), it cannot be proven that culture was responsible for the divergence.  

Mitochondrial DNA diversity in four matrilineal whale species (killer whales, sperm whales and the two pilot whale species, Globicephala spp.) is about fivefold lower than it is in most other cetacean species (Whitehead 1998). Whitehead (1998) suggests that this may have occurred by means of 'cultural hitchhiking' in which selectively-advantageous and matrilineally-transmitted cultural variants sweep through a population, incidentally reducing the diversity of analogously transmitted mitochondrial DNA. Such a process is theoretically analagous to molecular hitchhiking in which diversity in a neutral locus is reduced by selection at a linked, non-neutral locus (Kaplan et al. 1989). In the cultural hitchhiking proposed by Whitehead (1998), the non-neutral locus is a cultural trait, transmitted matrilineally between generations; selection is in the form of greater reproduction or survival for animals with certain cultural variants. Since mtDNA is also transmitted matrilineally between generations, alleles at neutral mtDNA loci will track the spread of ('hitchhike on') successful cultural traits - as successful traits spread in the population, the mtDNA alleles associated with that matriline will also spread, giving rise to the reduced mtDNA diversity now observed in the matrilineal odontocetes. Other theoretically tenable explanations for the low mtDNA diversity of the matrilineal odontocetes are population bottlenecks (Lyrholm & Gyllensten 1998), group-specific population dynamics (Siemann 1994; Amos 1999), or group-specific environments (Tiedemann & Milinkovitch 1999). However, all of these alternative explanations make assumptions or predictions that do not seem to be consistent with what we know of the biology of the matrilineal whales (Whitehead 1998; 1999a).  

Conformist traditions can lead to cultural group selection (Boyd & Richerson 1985). Group conformity increases both homogeneity within groups and heterogeneity among groups and thus elevates variation in behavioural phenotype to the group level (Boyd & Richerson 1985; Richerson & Boyd 2000); hence we would expect selection on behavioural phenotype to act at this level. For species that forage cooperatively, particularly within kin-based groups (e.g. killer whales), competition for resources may occur largely between rather than within groups, which would significantly increase the adaptive value of conformist traditions, reinforcing the whole system. Similarly, predator-prey arms races can be a potent driver of both genetic evolution (Dawkins & Krebs 1979) and, as is very apparent in human history, cultural evolution. For most whales and dolphins, the most formidable and important natural predator is another cetacean, the killer whale, (Jefferson et al. 1991) and the predatory techniques of killer whales appear to be largely determined by cultural processes. Thus, it is possible to envisage cultural arms races between killer whales and their cetacean prey.  

Theoretical studies also suggest that, during the evolution of group-specific cultures, behaviour which is not adaptive can easily arise (Boyd & Richerson 1985). There is one behavioural pattern seen in group-living cetaceans which is individually maladaptive but which could have arisen within a system of conformist traditions: mass stranding. Cetaceans of several species fatally strand en masse. In contrast to individual strandings, most of the animals involved in these mass strandings appear healthy, but when individually pulled back to sea, turn around and re-strand (Sergeant 1982) - a simple, genetically mediated, aggregation response is unlikely to produce such behaviour as it is so individually maladaptive. This phenomenon is seen as indicative of extreme social cohesion in the species which mass strand (Norris & Schilt 1988), with the usually adaptive strategy of remaining with the group proving fatal when one member makes a mistake or becomes debilitated through disease. There is evidence from pilot whale strandings that larger (presumably older) animals have a strong influence on the behaviour of the group (Fehring & Wells 1976). We suggest that cultural group conformity in movement strategies may play an important role in mass strandings; such phenomena might then be an example of the maladaptive effects of conformist cultures.  

Culture may also have had effects on the evolution of life history. Menopause is known in killer and short-finned pilot whales (Globicephala macrorhynchus), and there are indications of its occurrence in other cetacean species (Marsh & Kasuya 1986; Olesiuk et al. 1990). Like humans, and unlike any other mammal, female killer and short-finned pilot whales may live decades after the birth of their last offspring (Table 4). Within-group cultural processes may have played a part in this phenomenon, if, for instance, the role of older females in cultural transmission is very important. Menopause could be highly adaptive if the role of older females as a source of information significantly increases the fitness of her descendants, and reproduction towards the end of her life decreases survival (Heimlich-Boran & Heimlich-Boran in press; see also Norris & Pryor 1991). Guinet and Bouvier (1995) note that the juvenile killer whales they observed learning the difficult and dangerous technique of self-stranding in order to catch pinnipeds spent at least six years closely associated with their mothers; one calf was not observed to capture prey itself until it was six years old, and even then required assistance in handling the prey. They contrast this with observations near Vancouver Island where juvenile 'resident' killer whales rarely spend more than three years in such close association with their mother, and feed on salmon which they learn to catch within a year of birth (Haenel 1986). We suggest that the long time required to learn the culturally transmitted (and highly adaptive) self-stranding technique may be driving an incipient divergence in life histories, with the by-product of extended parent-offspring contact providing the opportunity for more cultural transmission. Such interactions between culture and development may closely parallel early human evolution.

Russell and Russell (1990) point out the link between maternal care and cultural transmission in early humans and other primates, and it is interesting that the cetacean species for which gene-culture coevolution has been suggested are also those with matrilineally based societies. This potential link produces a testable hypothesis: other, less studied, matrilineal cetacean species should show group-specific traditions. Preliminary evidence suggests that short-finned pilot whales, almost certainly a matrilineal species and also a species showing menopause (Kasuya & Marsh 1984), do indeed have group-specific dialects (Scheer et al. 1998). If our ideas on gene-culture co-evolution in cetaceans and cultural influence in the evolution of menopause are correct, then further investigation of this species and its congener, the long-finned pilot whale (Globicephala melas) - also matrilineal (Amos et al. 1991) - would strengthen this link.  

Although it has not been experimentally demonstrated in any case, observations of cetaceans in the wild strongly suggest that cultural transmission is important in some species. The lack of evidence in other species could well be due simply to lack of study. In the case of killer whales and possibly other matrilineal species, this transmission gives rise to stable cultures, which are in some respects unique outside humans. Our ethnological perspective, and hence our conclusions, are unlikely to be shared by all. However, our approach is internally consistent, and its conclusion is that culture should be attributed to cetaceans. We hope to stimulate a research effort which, even if it should disconfirm some aspects of our assertion, will give us a much better insight into the role of cultural transmission in the behavioural development of cetaceans. Given that there are over 80 species in the group, the possibility for comparative work is exciting.  

We have suggested several aspects of the natural history of whales and dolphins that may have promoted the evolution of these complex cultures. Of these, mobility may largely account for the apparently greater complexity of some cetacean cultures than those found in some non-human primates, whereas greater group stability and cognitive ability may be important in the differences between the cultures of cetaceans and birds. These ideas have relevance to our understanding of human pre-history. Theoretical work indicating the widespread adaptiveness of culture coupled with a dearth of empirical examples suggest there are important obstacles to the evolution of cultural transmission, obstacles which both humans and some cetaceans appear to have overcome. What ecological and social factors were common in the histories of both groups to enable this evolutionary leap? Our review suggests stable matrilineal groups as an important social factor, and environmental variability and mobility (c.f. Boesch 1996) as important ecological factors. While cetaceans are intrinsically more mobile than humans, humans have been able to use cultural innovations to become progressively more proficient travelers, overtaking first cetaceans, then birds, and so accelerating the spread and evolution of our other cultural forms.  

None of the observations of cetacean culture summarized in this paper come from research directly on cultural transmission - they are by-products of observational studies of behaviour, vocalizations or populations. Yet, together, they constitute strong evidence that, from the ethnographic perspective, these animals do have culture. Thus, there is a clear case for studying the cultural transmission of information directly as parts of the research agendas of the long-term field studies of whales and dolphins.  

 Many thanks to Robin Baird, Jim Heimlich-Boran, Kevin Laland, Peter Richerson, Brian Smith and Andrew Whiten for ideas and access to unpublished documents, and to Robin Baird, Richard Connor, John Dupre, Christophe Guinet, Andy Horn, Kevin Laland, Thierry Ripoll, Peter Tyack, Andrew Whiten and five anonymous reviewers for helpful comments on various versions of this manuscript. This work was funded by the Natural Sciences and Engineering Research Council of Canada. Luke Rendell was supported by a Canadian Commonwealth Scholarship and Izaak Walton Killam Memorial Scholarship.

Amos, W., Barrett, J. & Dover, G. A. (1991) Breeding behaviour of pilot whales revealed by DNA fingerprinting. Heredity 67: 49-55.  

Amos, W. (1999) Culture and genetic evolution in whales. Science 284:2055a.  

Anderson, J. (1879) Anatomical and zoological researches: comprising accounts of the zoological results of the two expeditions to western Yunnan in 1868 and 1875; and a monograph of the two cetacean genera, Platanista and Orcaella. Bernard Quaritich.  

Aoki, K. (1991) Some theoretical aspects of the origin of cultural transmission. In: Evolution of life: Fossils, molecules and culture, eds. S. Osaura & T. Honjo, Springer-Verlag.  

Arnbom, T., Papastavrou, V., Weilgart, L. S. & Whitehead, H. (1987) Sperm whales react to an attack by killer whales. Journal of Mammalogy 68: 450-453.  

Bain, D. E. (1986) Acoustic behavior of Orcinus: Sequences, periodicity, behavioral correlates and an automated technique for call classification. In: Behavioral biology of killer whales, eds. B. Kirkevold & J. Lockard, Alan R. Liss, Inc.  

Baird, R. W. (1999) The killer whale - foraging specializations and group hunting. In: Cetacean societies, eds. J. Mann, R. C. Connor, P. Tyack & H. Whitehead, University of Chicago Press.  

Baird, R. W. & Dill, L. M. (1995) Occurrence and behavior of transient killer whales: seasonal and pod-specific variability, foraging behavior and prey handling. Canadian Journal of Zoology 73: 1300-1311.  

Baird, R. W., Abrams, P. A. & Dill, L. M. (1992) Possible indirect interactions between transient and resident killer whales: implications for the evolution of foraging specializations in the genus Orcinus. Oecologia 89: 125-132.  

Baker, C. S., Palumbi, S. R., Lambertsen, R. H., Weinrich, M. T., Calambokidis, J. & O'Brien, S. J. (1990) Influence of seasonal migration on geographic distribution of mitochondrial DNA haplotypes in humpback whales. Nature 344: 238-240.  

Baker, M. C. & Cunningham, M. A. (1985) The biology of bird-song dialects. The Behavioral and Brain Sciences 8: 85-133.  

Barlow, D. P. (1995) Gametic imprinting in mammals. Science 270: 1610-1613.  

Bauer, G. & Johnson, C. M. (1994) Trained motor imitation by bottlenose dolphins (Tursiops truncatus). Perceptual and Motor Skills 79: 1307-1315.  

Bettinger, R. L. (1991) Hunter-gatherers: Archaeological and evolutionary theory. Plenum Press.  

Bigg, M. A., Olesiuk, P. F., Ellis, G. M., Ford, J. K. B. & Balcomb, K. C. (1990) Social organization and genealogy of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Reports of the International Whaling Commission (Special Issue) 12: 383-405.  

Boesch, C. (1996) The emergence of cultures among wild chimpanzees. Proceedings of the British Academy 88: 251-268.  

Boesch, C., Marchesi, P., Marchesi, N., Fruth, B. & Joulian, F. (1994) Is nut-cracking in wild chimpanzees a cultural behaviour? Journal of Human Evolution 26: 325-338.  

Bonner, J. T. (1980) The evolution of culture in animals. Princeton University Press.  

Boughman, J. W. & Wilkinson, G. S. (1998) Greater spear-nosed bats discriminate group mates by vocalizations. Animal Behaviour 55: 1717-1732.  

Bowles, A. E., Glenn Young, W. & Asper, E. D. (1988) Ontogeny of sterotyped calling of a killer whale calf, Orcinus orca, during her first year. Rit Fiskideildar 11: 251-275.  

Boyd, R. & Richerson, P. J. (1985) Culture and the evolutionary process. Chicago University Press.

Boyd, R. & Richerson, P. J. (1988) An evolutionary model of social learning: The effects of spatial and temporal variation. In: Social learning: Psychological and biological perspectives, eds. T. Zentall & B. G. Galef Jr., Lawrence Erlbaum Assoc.  

Boyd, R. & Richerson, P. J. (1995) Why does culture increase human adaptability? Ethology and Sociobiology 16: 123-143.  

Boyd, R. & Richerson, P. J. (1996) Why culture is common, but cultural evolution is rare. Proceedings of the British Academy 88: 77-93.  

Byrne, R. W. & Russon, A. E. (1998) Learning by imitation: A hierarchical approach. The Behavioral and Brain Sciences 21: 667-684.  

Caro, T. M. & Hauser, M. D. (1992) Is there teaching in nonhuman animals? Quarterly Review of Biology 67: 151-174.  

Cavalli-Sforza, L. L. & Feldman, M. W. (1981) Cultural transmission and evolution: A quantitative approach. Princeton University Press.  

Cerchio, S. (1993) Geographic variation and cultural evolution in songs of humpback whales (Megaptera novaeangliae) in the Eastern North Pacific. M.Sc. Thesis, Moss Landing Marine Laboratories, San Jose State University.  

Christal, J. (1998) An analysis of sperm whale social structure: patterns of association and genetic relatedness. Ph.D. Dissertation, Dalhousie University, Halifax, Nova Scotia.  

Christal, J., Whitehead, H. & Lettevall, E. (1998) Sperm whale social units: variation and change. Canadian Journal of Zoology 76: 1431-1440.  

Clapham, P. J. (1993) Social organization of humpback whales on a North Atlantic feeding ground. Symposia of the Zoological Society, London 66: 131-145.  

Clapham, P. & Mayo, C. A. (1987) Reproduction and recruitment of individually identified humpback whales (Megaptera novaeangliae) observed in Massachusetts Bay 1979-1985. Canadian Journal of Zoology 65: 2853-2863.  

Clark, C. W. (1990) Acoustic behaviour of mysticete whales. In: Sensory abilities of cetaceans, eds. J. Thomas & R. Kastelein, Plenum Press.  

Connor, D. A. (1982) Dialects versus geographic variation in mammalian vocalizations. Animal Behaviour 30: 297-298. 

Connor, R. C., Mann, J., Tyack, P. & Whitehead, H. (1998) Social evolution in toothed whales. Trends in Ecology and Evolution 13: 228-231.  

Couperus, A. S. (1994) Killer whales (Orcinus orca) scavenging on discards of freezer trawlers north-east of the Shetland Islands. Aquatic Mammals 20: 47-51.  

Dawkins, R. & Krebs, J. R. (1979) Arms races between and within species. Proceedings of the Royal Society of London B, Biological Sciences 205: 489-511.  

Deecke, V. B. (1998) Stability and change of killer whale (Orcinus orca) dialects. M.Sc. Thesis, University of British Columbia, Vancouver.  

de Waal, F. B. M. (1999) Cultural primatology comes of age. Nature 399: 635-636.  

Dufault, S. & Whitehead, H. (1995) The geographic stock structure of female and immature sperm whales in the South Pacific. Reports of the International Whaling Commission 45: 401-405.  

Dufault, S. & Whitehead, H. (1998) Regional and group-level difference in fluke markings and notches of sperm whales. Journal of Mammalogy 79: 514-520.  

Feekes, F. (1982) Song mimesis within colonies of Cacicus c. cela (Icteridae, Aves). A colonial password? Zeitschrift fur Tierpsychologie 58: 119-152.  

Fehring, W. K. & Wells, R. S. (1976) A series of strandings by a single herd of pilot whales on the west coast of Florida. Journal of Mammalogy 57: 191-194.  

Feldman, M. W. & Laland, K. N. (1996) Gene-culture coevolutionary theory. Trends in Ecology and Evolution 11: 453-457.  

Felleman, F. L., Heimlich-Boran, J. R. & Osborne, R. W. (1991) The feeding ecology of killer whales. In: Dolphin societies: Discoveries and puzzles, eds. K. Pryor & K. S. Norris, University of California Press.  

Fertl, D., Acevedo-Gutierrez, A. & Darby, F. L. (1996) A report of killer whales (Orcinus orca) feeding on a carcharhinid shark in Costa Rica. Marine Mammal Science 12: 606-611.  

Findlay, C. S. (1991) Fundamental theorem of natural selection under gene-culture transmission. Proceedings of the National Academy of Sciences of the United States of America 88: 4874-4876.  

Fisher, J. & Hinde, R. A. (1949) The opening of milk bottles by birds. British Birds 42: 347-357.  

Ford, J. K. B. (1991) Vocal traditions among resident killer whales (Orcinus orca) in coastal waters of British Columbia. Canadian Journal of Zoology 69: 1454-1483.  

Ford, J. K. B., Ellis, G. M., Barrett-Lennard, L. G., Morton, A. B., Palm, R. S. & Balcomb, K. C. (1998) Dietary specialization in two sympatric populations of killer whales (Orcinus orca) in coastal British Columbia and adjacent waters. Canadian Journal of Zoology 76: 1456-1471.  

Galef, B. G. (1988) Imitation in animals. In: Social learning: Psychological and biological perspectives, eds. T. R. Zentall & B. G. Galef, Lawrence Erlbaum.  

Galef, B. G. (1992) The question of animal culture. Human Nature 3: 157-178.  

Galef, B. G. (1996) Social enhancement of food preferences in Norway rats: A brief review. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Grant, B. R. & Grant, P. R. (1996) Cultural inheritance of song and its role in the evolution of Darwin's finches. Evolution 50: 2471-2487.  

Guinet, C. (1991) Intentional stranding apprenticeship and social play in killer whales (Orcinus orca). Canadian Journal of Zoology 69: 2712-2716. 

Guinet, C. & Bouvier, J. (1995) Development of intentional stranding hunting techniques in killer whale (Orcinus orca) calves at Crozet Archipelago. Canadian Journal of Zoology 73: 27-33.  

Haenel, N. J. (1986) General notes on the behavioral ontogeny of Puget Sound killer whales and the occurence of allomaternal behavior. In: Behavioral biology of killer whales, eds. B. C. Kirkevold & J. S. Lockard, Alan R. Liss, Inc. 

Hain, J. H. W., Carter, G. R., Kraus, S. D., Mayo, C. A. & Winn, H. E. (1982) Feeding behavior of the humpback whale, Megaptera novaeangliae, in the Western North Atlantic. Fishery Bulletin US 80: 259-268.  

Healey, R. F., Cooke, F. & Colgan, P. W. (1980) Demographic consequences of snow goose brood-rearing traditions. Journal of Wildlife Management 44: 900-905.  

Heimlich-Boran, J. R. & Heimlich-Boran, S. L. (in press) Social learning in cetaceans: hunting, hearing and hierarchies. Symposia of the Zoological Society, London 73.  

Henrich, J. & Boyd, R. (1998) The evolution of conformist transmission and the emergence of between group differences. Evolution and Human Behavior 19: 215-241.  

Herman, L. M. (1986) Cognition and language competencies of bottlenosed dolphins. In: Dolphin cognition and behavior: A comparative approach, eds. R. J. Schusterman, J. A. Thomas & F. G. Wood, Lawrence Erlbaum.  

Herman, L. M., Pack, A. A. & Morrel-Samuels, P. (1993) Representational and conceptual skills of dolphins. In: Language and communication: Comparative perspectives, eds. H. L. Roitblat, L. M. Herman & P. E. Nachtigall, Lawrence Erlbaum.  

Herman, L. M., Pack, A. A. & Wood, A. M. (1994) Bottlenose dolphins can generalize rules and develop abstract concepts. Marine Mammal Science 10: 70-80.  

Heyes, C. M. (1993) Imitation, culture and cognition. Animal Behaviour 46: 999-1010.  

Heyes, C. M. (1994) Social learning in animals: Categories and mechanisms. Biological Review 69: 207-231.  

Heyes, C. M. (1996) Introduction: Identifying and defining imitation. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Hoelzel, A. R. (1991) Killer whale predation on marine mammals at Punta Norte, Argentina; food sharing, provisioning and foraging strategy. Behavioral Ecology and Sociobiology 29: 197-204.  

Hoelzel, A. R., Dahlheim, M. & Stern, S. J. (1998) Low genetic variation among killer whales (Orcinus orca) in the eastern North Pacific, and genetic differentiation between foraging specialists. Journal of Heredity 89: 121-128.  

Hoyt, E. (1990) Orca: The whale called killer. 2nd Edition. Camden House.  

Huffman, M. A. (1996) Acquisition of innovative cultural behaviours in non-human primates: A case study of stone handling, a socially transmitted behavior in Japanese macaques. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Janik, V. M. & Slater, P. J. B. (1997) Vocal learning in mammals. Advances in the Study of Behaviour 26: 59-99.  

Jefferson, T. A., Stacey, P. J. & Baird, R. W. (1991) A review of killer whale interactions with other marine mammals: predation to co-existence. Mammal Review 4: 151-180.  

Kaplan, N. L., Hudson, R. R. & Langley, C. H. (1989) The "hitchhiking effect" revisited. Genetics 123: 887-899.  

Kasuya, T. & Marsh, H. (1984) Life history and reproductive biology of the short-finned pilot whale, Globicephala macrorhynchus, off the Pacific coast of Japan. Reports of the International Whaling Commission (Special Issue) 6: 259-310.  

Katona, S. K. & Beard, J. A. (1990) Population size, migrations and feeding aggregations of the humpback whale (Megaptera novaeangliae) in the western North Atlantic Ocean. Reports of the International Whaling Commission (Special Issue) 12: 295-305.  

Kawai, M. (1965) Newly acquired pre-cultural behavior of natural troop of Japanese monkeys on Koshima Islet. Primates 1: 1-30.  

Kuczaj, S. A., Gory, J. D. & Xitco, M. J. (1998) Using programs to solve problems: Imitation versus insight. The Behavioral and Brain Sciences 21: 695-696.  

Kummer, H. (1971) Primate societies: Group techniques of ecological adaptation. Aldine-Atherton Inc.  

Laland, K. N. (1992) A theoretical investigation of the role of social transmission in evolution. Ethology and Sociobiology 13: 87-113.  

Laland, K. N., Richerson, P. J. & Boyd, R. (1996) Developing a theory of animal social learning. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef Jr., Academic Press.  

Lefebvre, L. (1995) Culturally-transmitted feeding behaviour in primates: Evidence for accelerated learning rates. Primates 36: 227-239.  

Lefebvre, L. & Palameta, B. (1988) Mechanisms, ecology, and population diffusion of socially-learned food-finding behavior in feral pigeons. In: Social learning: Psychological and biological perspectives, eds. T. R. Zentall & B. G. Galef, Lawrence Erlbaum.  

Lopez, J. C. & Lopez, D. (1985) Killer whales of Patagonia and their behavior of intentional stranding while hunting nearshore. Journal of Mammalogy 66: 181-183.  

Lyrholm, T. & Gyllensten, U. (1998) Global matrilineal population structure in sperm whales as indicated by mitochondrial DNA sequences. Proceedings of the Royal Society of London B, Biological Sciences 265: 1679-1684.  

Mammen, D. L. & Nowicki, S. (1981) Individual differences and within-flock convergence in chickadee calls. Behavioral Ecology and Sociobiology 9: 179-186.  

Mann, J. (1999) Behavioral sampling methods for cetaceans: A review and critique. Marine Mammal Science 15: 102-122.  

Marsh, H. & Kasuya, T. (1986) Evidence for reproductive senescence in female cetaceans. Reports of the International Whaling Commission (Special Issue) 8: 57-74.  

Marten, K. & Psarakos, S. (1995) Using self-view television to distinguish between self-examination and social behavior in the bottlenose dolphin (Tursiops truncatus). Consciousness and Cognition 4: 205-224.  

McGrew, W. C. (1992) Chimpanzee material culture: Implications for human evolution. Cambridge University Press.  

McGrew, W. C. (1994) Tools compared: The material of culture. In: Chimpanzee cultures, eds. R. W. Wrangham, W. C. McGrew, F. B. M. de Waal & P. G. Heltne, Harvard University Press.  

Meltzoff, A. N. (1996) The human infant as imitative generalist: A 20 year progress report on infant imitation with implications for comparative psychology. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Midford, P. (1993) Cultural transmission is more than cultural learning. The Behavioral and Brain Sciences 16: 529-530.  

Moore, B. R. (1992) Avian movement imitation and a new form of mimicry: Tracing the evolution of a complex form of learning. Behaviour 122: 231-263.  

Moore, B. R. (1996) The evolution of imitative learning. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Mundinger, P. C. (1980) Animal cultures and a general theory of cultural evolution. Ethology and Sociobiology 1: 183-223.  

Myers, R. A., Mertz, G. & Bridon, J. (1997) Spatial scales of interannual recruitment variations of marine, anadromous and freshwater fish. Canadian Journal of Fisheries and Aquatic Sciences 54: 1400-1407.  

Nichol, L. M. & Shackleton, D. M. (1996) Seasonal movements and foraging behavior of northern resident killer whales (Orcinus orca) in relation to the inshore distribution of salmon (Oncorhynchus spp.) in British Columbia. Canadian Journal of Zoology 74: 983-991.  

Nishida, T. (1987) Local traditions and cultural transmission. In: Primate societies, eds. B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham & T. T. Struhsaker, University of Chicago Press.

Norris, K. S. & Dohl, T. P. (1980) The structure and functions of cetacean schools. In: Cetacean behavior: Mechanisms and functions, ed. L. M. Herman, Wiley-Interscience.  

Norris, K. S. & Pryor, K. (1991) Some thoughts on grandmothers. In: Dolphin Societies: Discoveries and Puzzles, eds. K. Pryor & K. S. Norris, University of California Press.  

Norris, K. S. & Schilt, C. R. (1988) Cooperative societies in three-dimensional space: On the origins of aggregations, flocks and schools, with special reference to dolphins and fish. Ethology and Sociobiology 9: 149-179.  

Norris, K. S., Würsig, B., Wells, R. S. & Würsig, M. (1994) The Hawaiian spinner dolphin. University of California Press.  

O'Corry-Crowe, G. M., Suydam, R. S., Rosenberg, A., Frost, K. J. & Dizon, A. E. (1997) Phylogeography, population structure and dispersal patterns of the beluga whale Delphinapterus leucas in the western Nearctic revealed by mitochondrial DNA. Molecular Ecology 6: 955-970.  

Ohsumi, S. (1966) Sexual segregation of the sperm whale in the North Pacific. Scientific Reports of the Whales Research Institute, Tokyo 23: 1-25.  

Olesiuk, P., Bigg, M. A. & Ellis, G. M. (1990) Life history and population dynamics of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Reports of the International Whaling Commission (Special Issue) 12: 209-243.  

Olsen, D. R. & Astington, J. W. (1993) Cultural learning and the educational process. The Behavioral and Brain Sciences 16: 531-532.  

Osborne, R. W. (1986) A behavioral budget of Puget Sound killer whales. In: Behavioral biology of killer whales, eds. B. C. Kirkevold & J. S. Lockard, Alan R. Liss, Inc.  

Payne, K. & Payne, R. S. (1985) Large-scale changes over 17 years in songs of humpback whales in Bermuda. Zeitschrift fur Tierpsychologie 68: 89-114.  

Payne, R. B. (1985) Behavioral continuity and change in local song populations of village indigobirds Vidua chalybeata. Zeitschrift fur Tierpsychologie 70: 1-44.  

Payne, R. & Guinee, L. N. (1983) Humpback whale, Megaptera novaeangliae, songs as an indicator of "stocks". In: Communication and behavior of whales, ed. R. Payne, Westview Press.  

Payne, R. & McVay, S. (1971) Songs of humpback whales. Science 173: 587-597.  

Payne, R. S. & Webb, D. (1971) Orientation by means of long-range acoustic signaling in baleen whales. Annals of the New York Academy of Sciences 188: 110-142.  

Pitman, R. L. & Chivers, S. J. (1999) Terror in black and white. Natural History 107: 26-29.

Plotkin, H. C. (1996) Non-genetic transmission of information: Candidate cognitive processes and the evolution of culture. Behavioural Processes 35: 207-213.  

Pryor, K., Lindbergh, J., Lindbergh, S. & Milano, R. (1990) A dolphin-human fishing cooperative in Brazil. Marine Mammal Science 6: 77-82.  

Richard, K. R., Dillon, M. C., Whitehead, H. & Wright, J. M. (1996) Patterns of kinship in groups of free-living sperm whales (Physeter macrocephalus) revealed by multiple molecular genetic analyses. Proceedings of the National Academy of Sciences of the United States of America 93: 8792-8795.  

Richards, D. G. (1986) Dolphin vocal mimicry and vocal object labelling. In: Dolphin cognition and behavior: A comparative approach, eds. R. J. Schusterman, J. A. Thomas & F. G. Wood, Lawrence Erlbaum.  

Richards, D. G., Wolz, J. P. & Herman, L. M. (1984) Vocal mimicry of computer-generated sounds and vocal labelling of objects by bottlenose dolphins (Tursiops truncatus): Evidence for vocal learning. Journal of Comparative Psychology 98: 10-28.  

Richerson, P. J. & Boyd, R. (2000) The evolution of human ultrasociality. In: Indoctrinability, ideology and warfare, eds. I. Eibl-Eibesfeldt & F.K. Salter, Berghahn Books.  

Rogoff, B., Chavajay, P. & Matusov, E. (1993) Questioning assumptions about culture and individuals. The Behavioral and Brain Sciences 16: 533-534.  

Roper, T. J. (1986) Cultural evolution of feeding behaviour in animals. Science Progress 70: 571-583.  

Russell, C. & Russell, W. M. S. (1990) Cultural evolution of behaviour. Netherlands Journal of Zoology 40: 745-762.  

Scheer, M., Hoffman, B. & Behr, P. I. 1998. Discrete pod-specific call repertoires among short-finned pilot whales (Globicephala macrorhynchus) off the SW coast of Tenerife, Canary Islands. Abstract in World Marine Mammal Science Conference, Monaco.  

Sergeant, D. E. (1982) Mass strandings of toothed whales (Odontoceti) as a population phenomenon. Scientific Reports of the Whales Research Institute 34: 1-47.  

Shane, S. H., Wells, R. S. & Würsig, B. (1986) Ecology, behavior and social organisation of the bottlenose dolphin: A review. Marine Mammal Science 2: 24-63.  

Sherry, D. F. & Galef, B. G. (1984) Cultural transmission without imitation: Milk bottle opening by birds. Animal Behaviour 32: 937-938.  

Siemann, L. A. (1994) Mitochondrial DNA sequence variation in North Atlantic long-finned pilot whales, Globicephala melas. Ph.D. Dissertation, Massachusetts Institute of Technology, Cambridge, Massachusetts.  

Silber, G. K. & Fertl, D. (1995) Intentional beaching by bottlenose dolphins (Tursiops truncatus) in the Colorado River Delta, Mexico. Aquatic Mammals 21: 183-186.  

Similä, T. & Ugarte, F. (1993) Surface and underwater observations of cooperatively feeding killer whales in northern Norway. Canadian Journal of Zoology 71: 1494-1499.  

Similä, T., Holst, J. C. & Christensen, I. (1996) Occurence and diet of killer whales in northern Norway: Seasonal patterns relative to the distribution and abundance of Norwegian spring-spawning herring. Canadian Journal of Fisheries and Aquatic Sciences 53: 769-779.  

Slater, P. J. B. (1986) The cultural transmission of bird song. Trends in Ecology and Evolution 1: 94-97.  

Smith, B. D., Thant, U. H., Lwin, J. M. & Shaw, C. D. (1997) Investigation of cetaceans in the Ayayarwady River and northern coastal waters of Myanmar. Asian Marine Biology 14: 173-194.  

Smith, T. G., Siniff, D. B., Reichle, R. & Stone, S. (1981) Coordinated behavior of killer whales, Orcinus orca, hunting a crabeater seal, Lobodon carcinohagus. Canadian Journal of Zoology 59: 1185-1189.  

Smolker, R. A., Richards, A. F., Connor, R. C., Mann, J. & Berggren, P. (1997) Sponge-carrying by Indian Ocean bottlenose dolphins: possible tool-use by a delphinid. Ethology 103: 454-465.  

Spencer, H. G., Clark, A. G. & Feldman, M. W. (1999) Genetic conflicts and the evolutionary origin of genomic imprinting. Trends in Ecology and Evolution 14: 197-201.  

Steele, J. H. (1985) A comparison of terrestrial and marine ecological systems. Nature 313: 355-358.  

Strager, H. (1995) Pod-specific call repertoires and compound calls of killer whales (Orcinus orca, Linnaeus 1758), in the waters off Northern Norway. Canadian Journal of Zoology 73: 1037-1047.  

Sutherland, W. J., Ens, B. J., Goss-Custard, J. D. & Hulscher, J. B. (1996) Specialization. In: The oystercatcher: From individuals to populations, ed. J. D. Goss-Custard, Oxford University Press.  

Tayler, C. K. & Saayman, G. S. (1973) Imitative behaviour by Indian Ocean bottlenose dolphins (Tursiops aduncus) in captivity. Behaviour 44: 286-298.  

Terkel, J. (1996) Cultural transmission of feeding behavior in the black rat (Rattus rattus). In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.  

Thouless, C. R. (1995) Long distance movements of elephants in nothern Kenya. African Journal of Ecology 33: 321-334.  

Tiedemann, R. & Milinkovitch, M. C. (1999) Culture and genetic evolution in whales. Science 284: 2055a.  

Tomasello, M. (1994) The question of chimpanzee culture. In: Chimpanzee cultures, eds. R. W. Wrangham, W. C. McGrew, F. B. M. de Waal & P. G. Heltne, Harvard University Press.  

Tomasello, M., Kruger, A. C. & Ratner, H. H. (1993) Cultural learning. The Behavioral and Brain Sciences 16: 495-511.  

Trainer, J. M. (1985) Changes in song dialect distributions and microgeographic variation in the song of white-crowned sparrows (Zonotrichia leucophrys nutalli). Auk 100: 568-582.  

Trainer, J. M. (1989) Cultural evolution in song dialects of yellow-rumped caciques in Panama. Ethology 80: 190-204.  

Tyack, P. (1986a) Population biology, social behavior and communication in whales and dolphins. Trends in Ecology and Evolution 1: 144-150.  

Tyack, P. (1986b) Whistle repertoires of two bottlenosed dolphins, Tursiops truncatus: mimicry of signature whistles? Behavioral Ecology and Sociobiology 18: 251-257.  

Tyack, P. L. (1993) Animal language research needs a broader comparative and evolutionary framework. In: Language and communication: Comparative perspectives, eds. H. L. Roitblat, L. M. Herman & P. E. Natchigall, Lawrence Erlbaum.  

Tyack, P. L. & Sayigh, L. S. (1997) Vocal learning in cetaceans. In: Social influences on vocal development, eds. C. T. Snowdon & M. Hausberger, Cambridge University Press.  

Visser, I. (1999) Benthic foraging on stingrays by killer whales (Orcinus orca) in New Zealand waters. Marine Mammal Science 15: 220-227.  

Warner, R. R. (1988) Traditionality of mating-site preferences in a coral reef fish. Nature 335: 719-721.  

Watkins, W. A. & Schevill, W. E. (1977) Sperm whale codas. Journal of the Acoustical Society of America 62: 1486-1490.  

Weilgart, L. S. (1990) Vocalizations of the sperm whale (Physeter macrocephalus) off the Galápagos Islands as related to behavioral and circumstantial variables. Ph.D. Dissertation, Dalhousie University, Halifax, Nova Scotia, Canada.  

Weilgart, L. & Whitehead, H. (1997) Group-specific dialects and geographical variation in coda repertoire in South Pacific sperm whales. Behavioral Ecology and Sociobiology 40: 277-285.  

Weinrich, M. T., Schilling, M. R. & Belt, C. R. (1992) Evidence for acquisition of a novel feeding behaviour: lobtail feeding in humpback whales, Megaptera novaeangliae. Animal Behaviour 44: 1059-1072.  

Whitehead, H. (1996) Variation in the feeding success of sperm whales: temporal scale, spatial scale and relationship to migrations. Journal of Animal Ecology 65: 429-438.  

Whitehead, H. (1998) Cultural selection and genetic diversity in matrilineal whales. Science 282: 1708-1711.  

Whitehead, H. (1999a) Culture and genetic evolution in whales. Science. 284: 2055a.  

Whitehead, H. (1999b) Variation in the visually observable behavior of groups of Galápagos sperm whales. Marine Mammal Science 15: 1181-1197.  

Whitehead, H. (in preparation) Analysis of animal movement using opportunistic individual identifications: application to sperm whales.  

Whitehead, H. & Mann, J. (1999) Female reproductive strategies of cetaceans. In: Cetacean societies, eds. J. Mann, R. Connor, P. L. Tyack & H. Whitehead, University of Chicago Press.  

Whitehead, H. & Weilgart, L. (1999) The sperm whale: social females and roving males. In: Cetacean societies, eds. J. Mann, R. C. Connor, P. Tyack & H. Whitehead, University of Chicago Press.  

Whitehead, H., Waters, S. & Lyrholm, T. (1992) Population structure of female and immature sperm whales (Physeter macrocephalus) off the Galápagos Islands. Canadian Journal of Fisheries and Aquatic Sciences 49: 78-84.  

Whitehead, H., Dillon, M., Dufault, S., Weilgart, L. & Wright, J. (1998) Non-geographically based population structure of South Pacific sperm whales: dialects, fluke-markings and genetics. Journal of Animal Ecology 67: 253-262. 

Whiten, A. (1989) Transmission mechanisms in primate cultural evolution. Trends in Ecology and Evolution 4: 61-62.  

Whiten, A. & Ham, R. (1992) On the nature and evolution of imitation in the animal kingdom: Reappraisal of a century of research. Advances in the Study of Behaviour 21: 239-283.  

Whiten, A., Goodall, J., McGrew, W. C., Nishida, T., Reynolds, V., Sugiyama, Y., Tutin, C. E., Wrangham, R. W. & Boesch, C. (1999) Chimpanzee cultures. Nature 399: 682-685.  

Williams, T. M., Friedl, W. A., Fong, M. L., Yamada, R. M., Dedivy, P. & Haun, J. E. (1992) Travel at low energetic cost by swimming and wave-riding bottlenose dolphins. Nature 355: 821-823.  

Wrangham, R. W., de Waal, F. B. M. & McGrew, W. C. (1994) The challenge of behavioral diversity. In: Chimpanzee cultures, eds. R. W. Wrangham, F. B. M. de Waal, W. C. McGrew & P. G. Heltne, Harvard University Press.  

Würsig, B. (1986) Delphinid foraging strategies. In: Dolphin cognition and behavior: A comparative approach, eds. R. J. Schusterman, J. A. Thomas & F. G. Wood, Lawrence Erlbaum.  

Würsig, B. & Clark, C. W. (1993) Behavior. In: The bowhead whale. Special Publication No. 2, The Society for Marine Mammalogy, eds. J. J. Burns, J. J. Montague & C. J. Cowles, Allen Press, Inc.  

Yano, K. & Dahlheim, M. E. (1995) Killer whale, Orcinus orca, depredation on long-line catches of bottomfish in the southeastern Bering Sea and adjacent waters. Fishery Bulletin US 93: 355-372.  

Zentall, T. R. (1996) An analysis of imitative learning in animals. In: Social learning in animals: The roots of culture, eds. C. M. Heyes & B. G. Galef, Academic Press.

Table 1: Some definitions of culture.

Table 2 : Numbers of self-strandings observed and the percentage in which the mother was present for two killer whale calves (data from Guinet and Bouvier (1995).