The secret lives of African forest elephants: Using genetics, networks, and telemetry to understand sociality

Identifying the social structure of a species is critical for understanding the overall evolution and behavioral ecology of that species. An individual’s social relationships have consequences that impact the movements, habitat use, and mate choice of other individuals, and collectively influence spatial patterns and gene flow in the population or species. My dissertation research focuses on the social structure of African forest elephants (Loxodonta cyclotis) and tests for fission-fusion sociality, which has been detected in the other extant elephant species, the African savanna (L. africana) and Asian (Elephas maximus ) elephants. Observational studies at forest clearings have shown that African forest elephants have the smallest group sizes of the extant elephants, and are typically composed of an adult female and her dependent calves. However, associations may be more extensive because it is difficult to detect individuals that may be obscured by the dense forest vegetation. I used satellite telemetry, behavioral observations, non-invasive genetic sampling, and social networks to look for evidence of multi-tiered fission fusion sociality, in which an adult female and her dependent calves (tier one) form a “family group” with related adult females (tier two), that fuse into larger “bond groups” (tier three), and larger “clans” (tier four) of up to hundreds of elephants.;First, I describe satellite telemetry results from six adult female forest elephants in Loango National Park, Gabon. As home range overlap can provide indirect information about the social interactions between individuals, this study found small home ranges with low volume of intersection indices between individuals, indicating that the probability of co-occurrence between dyads of individuals in the same area is also low. This suggests spatial avoidance among these adult females, which contrasts with the patterns of family groups overlapping in home ranges within sub-populations in African savanna elephants.;Second, I used non-invasive genetic approaches to infer sociality from African forest elephants in Lopé National Park (LNP), Gabon. I found evidence of fine-scale genetic structure with individuals being more closely related to each other than expected by chance at distances of five kilometers or less. Through network models created by genotyping dung samples collected together at the same time, location, and of the same freshness, I found larger group sizes of forest elephants compared to those from observations alone and that groups largely consisted of individuals of the same mitochondrial DNA matriline. These results support evidence of higher order social structure including family groups (representing second tier relationships), and possibly even bond groups (third tier).;Finally, I tracked the relationships of known individuals and created social networks of adult females within LNP. Social networks revealed evidence of kin-based fission-fusion sociality with one large component of twenty-two adult females, followed by smaller ones of four. I observed many solitary females (females observed alone or only with dependent calves) throughout the study and few individuals had preferred associations. Although these results reveal that forest elephant females associate with other females in fission-fusion patterns, they frequently separate from them, and when preferred associations do form, it is typically with only one other individual.;When comparing these results with knowledge about the sociality of the other extant elephant species, they suggest that although there is evidence of kin-based fission-fusion sociality, African forest elephants differ from African and Asian elephants. Social networks from both African savanna and Asian elephants have more connected networks, and larger component sizes in networks. In contrast, forest elephant social networks and networks created from noninvasive genetic sampling had many small components and solitary individuals, few large components, and were disconnected.;Several ecological factors may contribute to the more limited sociality observed in forest elephants. Forest elephants have the most “closed” habitats of the three extant species, which may physically prevent larger aggregations. There also may be costs associated with foraging in larger groups, as patchily distributed resources such as fruits, on which forest elephants forage heavily, will quickly be depleted, resulting in increased travel for forage. Finally, large cooperatively hunting predators are currently absent in most locations throughout forest elephant range, negating the need for groups to form for defense against predators. Therefore, forest elephants may lack strong benefits associated with group-living and suffer from costs which may drive the type of fission-fusion patterns observed.