This community of Atlantic spotted dolphins exhibits fission/fusion dynamics very similar to that of the association patterns of coastal bottlenose dolphins and chimpanzees (see Connor et al. 2000). Strongest associations were between the same sex and age classes, though some variations were found. Males formed strong associations within and between male pairs/trios that remained evident over all time lags, while females had preferred casual acquaintances that disassociate over time and were affected by reproductive status and social familiarity. Mixed sex pairs showed similar patterns to female-female associations, though they were weaker and less stable over time even though mixed sex groups were common. Rapid disassociation (within one day) was a common theme in all LAR indicating a level of daily fluidity within the more stable associations of this community. The overall LAR reported here supports the community structure documented previously of three interacting social clusters (Elliser and Herzing 2012) because this type of LAR can be produced by a social system of permanent social units that associate temporarily (Whitehead 2008a). Combined with the fact that all sex class associations leveled out above the null association rate, this indicates a community with distinct interacting social clusters along with differential association patterns due to sex. The detail of this study reveals how sex and age class interact in their influence on associations and social structure.
The pattern of male associations was consistent with the rapid disassociation and constant companion model, where although there will be some rapid disassociation on a daily basis, males remained with their preferred companions consistently over all time lags. Socio-ecological factors determine female grouping and association patterns that in turn determine the options (regarding socio-sexual strategies, male associations/relationships and dispersal) left for males because they compete primarily for access to fertile females (Hill and Van Hooff 1994, Van Hooff and van Shaik 1994). The male spotted dolphins in this study show long-term strong associations between individuals and pair/trios of males, but are these male coalitions and/or alliances? de Waal and Harcourt (1992) define a coalition as a joining of forces by two or more parties during a conflict of interests with other parties, and an alliance as an enduring cooperative relationship in which repeated coalitions are formed. Male alliances in primates, lions and dolphins are primarily attributed to increased access (directly or indirectly) to females (e.g., Packer et al. 1991, Watts 1998, Connor et al. 2000).
Herzing (1996) described male coalitions (as defined above) of spotted dolphins consisting of three to four dolphins that chased and surrounded a female and eventually mated with her. This monopolization involves tending/following a female in apparent estrus, surrounding her, escorting her to the bottom during feeding bouts and fending off other male groups (Herzing and Johnson 1997; Herzing and Elliser, in press). The absolute duration of these behaviors is unknown, but females have been documented with the same male pair/trio during encounters (minutes to hours), multiple encounters in one day and in some cases across multiple days (DLH, unpublished data). Although this monopolizing behavior is not as overt as the herding by Shark Bay dolphins (Connor et al. 2000), or mate guarding in chimpanzees (Watts 1998), it seems to serve the same purpose: males cooperating to gain and maintain access to females. The importance of defense of females by first and second order alliances (discouraging attacks from other alliances) in these instances may have contributed to the selection pressure for use of coercion in consortships (Connor and Vollmer 2009). Cooperative behaviors within and between spotted dolphin male coalitions were also documented during intra and interspecific aggressive interactions, including synchronized swimming behavior, postures, and vocalizations (Herzing 1996, Cusick 2012). Synchrony has been shown to be an important component of bottlenose dolphin male alliances, and may be a useful measure of alliance unity (Connor et al. 2006). The behavioral evidence combined with the long-term association analysis presented here indicates that these strong male associations are alliances because they represent enduring cooperative relationships, whose function is, at least partially, to gain access to females.
Spotted dolphin male alliances share with Sarasota and Shark Bay bottlenose dolphins the characteristic of long-term stability, lasting at least 12 yr. These were strong, long-term stable core pairs/trios with CoAs ≥ 0.70, similar to first order alliances in Shark Bay (Connor et al. 1992) and pair alliances in Sarasota (Wells et al. 1987). The structure of spotted dolphin male alliances, however, more closely resembles that of the Shark Bay bottlenose dolphin alliances with at least two levels of alliance formation (Connor and Mann 2006, Connor et al. 2011).
There were groupings of two to three alliances, with lower CoAs ranging from twice the community average to 0.69 in a given pooled period and changing membership across pooled years. These groupings have been observed during courtship activities (Herzing 1996; Herzing and Johnson 1997; Herzing and Elliser, in press) as well as in both intra- and interspecies aggressive encounters (Herzing 1996; Cusick 2012; Herzing and Elliser, in press), suggesting enduring cooperative relationships indicative of alliance formation. These are similar in structure to the second order alliances of Shark Bay where there are strong associations between members of different alliances; however, the temporal stability of the second order alliances varies between the two species. In Shark Bay there can be shifts in membership over time, although many second order alliances in Shark Bay are very stable over many years (Connor et al. 1992; Connor 2007, 2010). The spotted dolphins seem to have less stability in second order alliances over longer periods (>3 yr). It is important to note that the pooling of data may have diluted the length of the associations we observed (they could be 4 to 5 yr long possibly); however, this is still shorter than the majority of stable second order alliances in Shark Bay. This difference in temporal stability between these two species may be due to varying social and ecological pressures, such that selection for long-term second order alliances is favored in Shark Bay, but not in the Bahamas.
While the majority of spotted dolphin second order alliances were between two stable core pairs, one larger group of males had more labile association patterns, similar to the 14 male bottlenose dolphin superalliance of Shark Bay (Connor et al. 1999). Although these male spotted dolphins (up to seven individuals) had varying associates within the group over the years, distinct partner preferences and avoidances were documented, similar to the superalliance (Connor 2007). However there were no associations between clear stable first order alliances as seen in the smaller second order alliances described above. The varied associations may also be influenced by competition for females and/or between other individuals/alliances. The superalliance members in Shark Bay joined forces and competed directly with smaller teams of stable alliances (Connor et al. 1999). It may be that these varied associations within this larger group are a result of dolphins associating with certain individuals during particular behavioral events (Gero et al. 2005). It is unclear what the purpose and significance are for this larger grouping of males in the Bahamas. Further behavioral research is needed to determine the function of this large grouping of males and how they interact with first and second order alliances.
Age class seems to be an important determinant in alliance formation as male association patterns were influenced heavily by the age of their associate. Alliance members were weakly associated during juvenile years when they were speckled, and the majority of spotted dolphin males that were not part of any alliance were speckled. The bonds between males apparently grow from relationships developed in subadult groups or earlier (Wells 1991), where more affiliative associations between juveniles may indicate the early stages of alliance formation (Gero et al. 2005). Spotted dolphin CoAs strengthened as they became mottled, starting at 10 yr of age and older. The majority of alliance pairs involved mottled and fused males of the same age class; with the strongest CoAs of first order alliances between fused males aged ≥16 yr. This structure is similar to that seen in Sarasota, where the minimum age for pair formation was 7 yr old, and most male pair bonds formed in their early teens. As males increase in age (15–20 or more years), so does the probability that the male was currently paired, or has had a partner in the past (Owen et al. 2002). Males became more restricted in their associations with other males of the same age class after the onset of sexual maturity (Wells et al. 1987).
There is preliminary evidence that these alliances of older, sexually mature males are important to successful reproduction in this population. In a recent genetic study, seven males were assigned paternity (for 10 calves). BigGash and Romeo (a long-term first order alliance), each had two calves and two other males (siring three calves) were in first order alliances. The final three males were within the larger more labile alliance. All paternities were assigned to fused males (≥16 yr old) (Green et al. 2011). Although limited sample size precludes absolute support, these results are similar to Shark Bay, where reproductive success was significantly skewed toward members of first order alliances (Krützen et al. 2004).
Access to females may not be the only reason for spotted dolphin male alliances. Regularly occurring interspecies encounters reveal more complex spotted dolphin male interactions. Spotted dolphins spend 15% of their time together with sympatric bottlenose dolphins. In many instances during aggressive interspecies interactions, coalitional male behavior by the spotted dolphins was critical in determining the outcome, needing six spotted dolphins to chase away one bottlenose dolphin (Herzing and Johnson 1997). First order and second order alliances were prevalent during these interspecies encounters, however, their associations were more complex; males had strong associates other than their first and second order alliance members (Elliser 2010). This has some similarities (though for a different purpose) to a possible third order alliance structure seen in Shark Bay where second order alliances have been seen to associate in amicable, regular low-level associations with other male groups in contests for females (Connor 2007, Connor et al. 2011). The need for allies to defend females may explain the formation of third order alliances if former allies were no longer present (Connor et al. 2011). For spotted dolphins the need for allies may be for defense against bottlenose dolphins. The function of these interspecies interactions is not fully understood, but alliance behavior has been seen to ward off and intercept matings between the species and defend individual males against the physical dominant behavior of the bottlenose dolphins (Herzing and Johnson 1997). Detailed behavioral analysis has documented focused, synchronized behavior within and between alliances during aggressive encounters towards bottlenose dolphins (Herzing and Johnson 1997, Cusick 2012), indicating cooperation for a common goal and some level of relationship between the individuals (at least during these encounters). Although de Waal and Harcourt (1992) define alliances/coalitions as interactions between members of the same species, the context of these interspecies encounters often mirrors that of intraspecies interactions (access to females, directly or indirectly). This is a unique case where the terms alliance and coalition can be used when describing interactions within and between species. Interestingly the spotted and bottlenose dolphins have also formed temporary interspecies associations while defending against third party intruders (such as a shark or offshore ecotype bottlenose dolphin [nonresident]), indicating cooperation even beyond intraspecies alliances/coalitions.
The question becomes, are these relationships during these encounters alliances or coalitions? Shorter-term coalitions often form during agonistic interactions in many populations (e.g., Tursiops sp.: Connor et al. 1999, Lusseau 2007; primates: Goodall 1986, Noë 1994, Silk 1999). It is evident that first and second order alliances for spotted dolphins, at the very least, form temporary coalitions during aggressive interactions with the larger, more dominant bottlenose dolphins. It is unclear whether these coalitions are indeed temporary and inconsistent between encounters, or are enduring cooperative relationships that constitute a third level of alliance formation. Future behavioral and association pattern research will help to illuminate the complex male relationships in this population and their regular interactions with the sympatric bottlenose dolphins.
Spotted dolphin females also showed rapid disassociation on a daily basis but contrary to the males, females had preferred casual acquaintances that disassociated and then may have re-associated again over time. These associations leveled out above the null association rate (and mixed sex LAR), most likely due to their consistent associations with other females in their social cluster (from 5 to 25 females). They had low-level associations in a network of females, but with no long-term consistent subsets of individuals. Generally bottlenose dolphin females have this type of “network,” rather than the specific subgroups of two to three individuals seen in male-male associations (Wells et al. 1987, Smolker et al. 1992, Connor et al. 2000, Rogers et al. 2004). There is evidence for increased relatedness of spotted dolphin females within clusters (genetic differentiation, Green 2008). This has also been documented in bottlenose dolphins (e.g., Wells 1991). Increased relatedness between females may reduce the fitness cost of competing/sharing resources, while also gaining the benefits of receiving aid in rearing young (Sterck and Watts 1997). This may encourage females to remain in their natal cluster, as the potential costs of emigration/immigration (such as increased aggression, decreased foraging and energetic travel costs) may be high, as seen in chimpanzees (Kahlenberg et al. 2008).
Strong associations between females were correlated with reproductive status and past social familiarity, supporting previous work on female associations (Herzing and Brunnick 1997). Female fitness and reproductive success are dependent on the successful rearing of young, and females will use social relationships to achieve their reproductive goals, as described in primates (Sterck and Watts 1997). Benefits to female grouping may be ecological in nature, such as increased predator protection and food distribution (Sterck and Watts 1997), or social, including calf care and social learning (Miles and Herzing 2003, Bender et al. 2008, Gibson and Mann 2008). Results indicate that familiarity and reproduction are strong influences in female sociality. Adaptive value of sociality is described for female bottlenose dolphins in a unique approach by Frère et al. (2010), showing that sociality influences the fitness trait in a wild population, consistent with the results of many social analyses (like this study) that show strong associations between females of the same reproductive status, and between nonrelated individuals. Thus genetic and social effects on fitness are intertwined, both important in determining female success (Frère et al. 2010).
Contrary to male-male associations, age was not a significant factor in female only associations. Female-female CoAs within and between age class were not significantly different and the majority of associations were between age classes. Spotted dolphin females had strong associations across age classes within their cluster because they associate highly with their older speckled and even mottled offspring. They also associate with other females and their older offspring, with whom they have had previous associations. It is obvious that females would have strong associations across classes between adults and calves, due to dependency during the first few years of the calves' life. Subsequently mother/calf associations tend to drop significantly between calf years three and four (spotted dolphins: Herzing and Brunnick 1997; bottlenose dolphins: Wells et al. 1987, Smolker et al. 1992), however, this study shows that some strong associations can remain, through adulthood of the offspring.
Consistent mother-offspring associations up to 11 yr were documented in both this study and previously (Herzing and Brunnick 1997), indicating strong relationships through at least three age classes of the offspring (up to mottled). While the mother and offspring are closely associated, the offspring will be exposed to and have relationships with their mother's associates and their offspring. Female associates may be daughters of their mother's close associates, with whom they spent part of their infancy or juvenile period (Wells et al. 1987, Möller and Harcourt 2008). The sociability of Shark Bay bottlenose dolphin female calves has been shown to mirror that of their mothers (Gibson and Mann 2008). This parity may translate into adulthood, continuing on the “network” of female relationships. The formation of the Northern, Central, and Southern clusters may be influenced by both kinship and social familiarity between females, while reproduction and social familiarity affect the patterns of within-cluster associations.