Many elephant populations across Africa were or are being devastated by poaching and habitat loss, making population and range size for the species important issues of conservation concern in elephant ranging areas, including Tarangire National Park (TNP). Poaching and/or overhunting are known to have direct effects on the demography of elephant populations. The current understanding of the indirect effects of poaching on the sociality and demography of elephant populations is relatively poor, both at the group and an individual level. In this thesis the importance of genetic relatedness (as influenced by poaching) on sociality and demography of the female elephants was studied, using a combination of genetic, observation of behaviour and two decades of demography data collected from the northern subpopulation of TNP. I investigated and characterized the relatedness categories within elephant groups. Using a conceptual model for group size analysis in aggregation economies, I hypothesized that elephant populations subject to social disruptions due to poaching would exhibit characteristics of the free entry model, whereas more stable, closed populations would better fit the group-controlled model. I present a rare quantitative analysis of genetic relatedness and group size patterns among groups of adult female elephants in two wild populations: one in Tarangire National Park (TNP), Tanzania, and another in Addo Elephant National Park (AENP), South Africa. I demonstrate that the group size in African elephant populations is governed by genetic relatedness, and that poaching/overhunting has a significant influence upon the apparent group formation and size in elephants. I then focused on the effect of relatedness on agonistic interactions between adult females. I hypothesized that individual-based aggressive interactions among adult female African elephants would vary according to degree of kinship, with closely related dyads showing less aggression towards each other in resource-limited environments, thereby leading to indirect fitness benefits for individuals. As predicted, females did not show agonistic interactions to their close kin most often, and the frequency and intensity of aggressive interactions was inversely related to the degree of relatedness of the interactants. The effect of group relatedness and structure on reproductive success of individual female African elephant in TNP was also investigated. Adult female reproductive success was significantly influenced by within- group relatedness and structure. Higher reproductive success (with higher frequency of calf production and survival and more female calves produced) were more evident in the closely related groups than groups with low relatedness, suggesting that females from genetically disrupted groups are less likely to be reproductive than those in closely related groups. The possibility of negative effects of poaching on the subsequent generation of poached adult females and the alternative of a positive demographic response through reduced density was assessed by analysing the demographic patterns of the first generation (F1) females of prime-aged adult female African elephants in TNP using within- group relatedness and size. I also compared vital rate (age of first birth and interbirth interval) responses of first generation (F1) cows from Tarangire (poached) elephants with other females from poached (Northern Luangwa National Park, Zambia) and unpoached (Amboseli National Park, Kenya and Addo Elephant National Park, South Africa) populations. Group relatedness had no significant effect on sex ratio of the F1 cows‟ calves. There was a significant difference between the mean age of first birth and interbirth interval of F1 cows from the two (poached, Tarangire and unpoached, Amboseli National Park) elephant populations, suggesting that elephant populations reduced by poaching to low levels show an increase vigour through release from density constraints. Based on these results, the broader implications of secondary effects of poaching on elephant populations are critically evaluated. Also the importance of understanding the consequences of these effects is highlighted in light of other elephant conservation and management approaches. This understanding is useful in making conservation and management decisions for elephants and other biodiversity.

The continual increase in elephant numbers across southern Africa raises concern, though the interventions to manage these populations are more contentious. Within small, enclosed reserves active management is necessary to limit elephant impact. Various management tools exist with which to control fluctuations in elephant population numbers or density and simulate natural large stochastic events to control population growth. During this study, undertaken in the Munyawana Conservancy, KwaZulu-Natal, and Pilanesberg National Park, North West Province, South Africa, several management options were implemented. In order to lower the population numbers, family groups were translocated from the Munyawana Conservancy to other reserves, while to reduce population growth rate an immunocontraception was implemented. Both conservation areas introduced older bulls to normalise the bull population age structure, and expanded the conservation area by inclusion of new land to reduce population density. The influence of these management interventions on the elephant population were measured by their social, behavioural, spatial and movement responses. The older bull introduction was successful as bulls set up exclusive bull areas. There was a quick, subtle affect on the bull groups' size immediately after the older bull introduction, while there was no immediate change within the resident bulls' musth behaviour or duration. During area expansion, elephants appeared to perceive the new unexplored area as a threat although this threat became reduced through time as they became more familiar with it. The spatial scale of response was relatively small, while the temporal scale of response was relatively large. Rotational immunocontraception was shown to be a successful tool to alter herd structure by aging the population and maintaining a low population growth rate. The process of immunocontraception darting had no significant effect on herd associations and movement rates, accordingly the duration of the disruption effects were short lived. During multi-management interventions, no differences were found within the elephant social grouping. Management interventions may pose unforeseen social risks and different populations may respond differently to management induced stress. Therefore, interventions need to be considered for each elephant population which will achieve the conservation area's objectives with the most effective outcome, but with lowest holistic impact. / Thesis (M.Sc.)-University of KwaZulu-Natal, Westville, 2012.