Male bottlenose dolphins (Tursiops sp.) in Shark Bay, WA, form several levels of alliances. Determining the relationship between paternity, relatedness and alliance membership is crucial in seeking evolutionary adaptive explanations for alliance formation. Previous behavioural data have revealed a social system whose complexity is unparalleled outside humans. Pairs or trios of male dolphins cooperate as stable first-order (1? alliances to sequester and control reproductive females. Two 1?alliances sometimes cooperate as small second-order (2? alliances to attack other 1?alliances or defend against attacks. Some males choose a different strategy by forming large superalliances of approximately fourteen individuals to attack 1?and 2?alliances. Kinship appears to play a role in the structuring of male alliances, but its importance differs with the alliance type. Relatedness analyses showed that on average, males in 1?and 2?alliances are strongly related, while members of the superalliance are not. Further, the strength of the association of partners within the superalliance was not correlated with their genetic relatedness. Thus, within one sex, it appears that there may be more than one simultaneous mode of group formation and its evolution. There was also an association between alliance behaviour and reproductive success as predicted by some theories of group formation. I assigned nine paternities to six out of 107 mature males. Males with alliance partners were significantly more successful in fathering offspring than males without partners. Compared to non-allied males, the chance of obtaining a paternity was significantly higher for members of 1?alliances, and reproductive success was significantly skewed among 1?alliance members. Assessment of kinship and relatedness required a number of technical developments and some preliminary data. I first refined a biopsy system in order to obtain tissue samples from free-ranging dolphins with minimum behavioural effects irrespective of age-group or gender. Then I carried out population genetics analyses, which demonstrated that there was only weak population structure within Shark Bay. Microsatellites showed a weak pattern of isolation by distance, and eight haplotypes of the mitochondrial DNA control region suggested weak female philopatry. The high number of migrants between locations allowed all samples from East Shark Bay to be pooled for subsequent analyses.

Male bottlenose dolphins (Tursiops sp.) in Shark Bay, WA, form several levels of alliances. Determining the relationship between paternity, relatedness and alliance membership is crucial in seeking evolutionary adaptive explanations for alliance formation. Previous behavioural data have revealed a social system whose complexity is unparalleled outside humans. Pairs or trios of male dolphins cooperate as stable first-order (1? alliances to sequester and control reproductive females. Two 1?alliances sometimes cooperate as small second-order (2? alliances to attack other 1?alliances or defend against attacks. Some males choose a different strategy by forming large superalliances of approximately fourteen individuals to attack 1?and 2?alliances. Kinship appears to play a role in the structuring of male alliances, but its importance differs with the alliance type. Relatedness analyses showed that on average, males in 1?and 2?alliances are strongly related, while members of the superalliance are not. Further, the strength of the association of partners within the superalliance was not correlated with their genetic relatedness. Thus, within one sex, it appears that there may be more than one simultaneous mode of group formation and its evolution. There was also an association between alliance behaviour and reproductive success as predicted by some theories of group formation. I assigned nine paternities to six out of 107 mature males. Males with alliance partners were significantly more successful in fathering offspring than males without partners. Compared to non-allied males, the chance of obtaining a paternity was significantly higher for members of 1?alliances, and reproductive success was significantly skewed among 1?alliance members. Assessment of kinship and relatedness required a number of technical developments and some preliminary data. I first refined a biopsy system in order to obtain tissue samples from free-ranging dolphins with minimum behavioural effects irrespective of age-group or gender. Then I carried out population genetics analyses, which demonstrated that there was only weak population structure within Shark Bay. Microsatellites showed a weak pattern of isolation by distance, and eight haplotypes of the mitochondrial DNA control region suggested weak female philopatry. The high number of migrants between locations allowed all samples from East Shark Bay to be pooled for subsequent analyses.

The biology of bottlenose dolphins from the east coast of southern Africa is investigated. Births occur predominantly in summer, after a one year gestation period. Calves are born at a length of approximately 103 cm and a mass of nearly 14 kg. Neonates are closely attended by the mother, she determines the calf's respiration rate and behaviour as it swims in a negative pressure vortex high on her flank. The initial weeks are marked by the calf's acquisition of swimming skills and its ability to regulate its own respiration rate. Although the calf takes solid food between six months and one year, suckling may continue for as much as three years. Initial growth is rapid but slows with the approach to puberty. Females reach sexual maturity at about ten years of age and may undergo a series of rapid ovulations before fertilisation occurs. Subsequently, the mean ovulation rate indicates a three year calving interval. Males show a mass growth spurt at the onset of puberty, between 10 and 12 years, and reach sexual maturity between 12 and 15 years of age. In both sexes, physical maturity and asymptotic size are reached at about 15 years and maximum life-span is in excess of 40 years. Although a wide variety of fish and cephalopods are taken, the fishes Pomadasys olivaceum, Scomber japonicus, Pagellus bellotti, Trachurus delagoae, and the cephalopods Sepia officinalis and Loligo sp. contribute some 60 % by mass of all prey taken. Different sex and maturity classes of dolphins consume differing sizes and species of prey, implying a partitioning of food resources within the group. Comparisons of stomach volumes with estimates of prey mass and nutritional requirements suggest that calves and lactating females may need to feed more often than other sex and maturity classes. Examination of shark stomachs indicates that shark predation may be an important component of dolphin natural mortality. Four species of shark, the Zambesi (Carcharhinus leucas), the tiger (Galeocerdo cuvien), the great white (Carcharodon carcharias) and the dusky shark (Carcharhinus obscurus) are implicated as dolphin predators. Estimates from the number of these four species caught annually and the frequency of occurrence of dolphin flukes and vertebrae in their stomachs suggest that a minimumof 20 bottle nosed dolphins or 2.2 % of the estimated population in southern Natal coastal waters are killed each year by sharks. Some 32 bottlenose dolphins, about 4 % of the estimated population, are captured in shark nets annually. Calves of two years or less constitute nearly 45 % of the catch, while lactating females make up a further 15 %. An analysis of biological, environmental and physiographic factors associated with each capture, suggest that feeding and probably prey movement and distribution are directly associated with capture. PCB, t-DDT and dieldrin concentrations in the blubber of male bottlenose dolphins increase with age and reach levels that may impair testosterone production. In females, there is an 80 % decline in residue concentrations after the first or second ovulation. Evidence presented suggests that first born calves receive possibly lethal doses of residues within two months of birth. Residue concentrations in dolphins from different geographical areas varied significantly, implying a degree of isolation of sections of the population. Investigation of reproductive parameters indicates that mortality of bottlenose dolphins off Natal probably equals or exceeds the replacement rate. The necessity for and requirements of further research on the natural history of bottlenose dolphins off the east coast of southern Africa are summarised. / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1989.