The ability to maintain important social relationships is believed to have direct fitness benefits for individuals, allowing them to maximise survival and reproductive success in constantly changing socio-ecological environments [1, 2, 30]. This is particularly apparent in large-brained, social species where information is accumulated over long life spans [1, 3–5, 27, 31]. However, extremely disruptive events, including culling, poaching and translocation to new areas or capture for captivity can ultimately lead to serious disruption of the intricate social networks that underpin social structure in these species, with severe impacts on each individual’s close social bonds and opportunities for learning from older group members [9, 11, 16, 19]. Furthermore, such disruption appears capable of driving aberrant behaviours in social animals that are akin to the post-traumatic stress disorder experienced by humans following extremely traumatic events [16, 19]. While elephants in the wild can appear to exhibit short-term resilience following social disruption, apparently forming stable and reproductively active family groups (but see 9), the results presented here suggest that important decision-making abilities that are likely to impact on fundamental aspects of the elephant’s complex social behaviour may be significantly altered in the long-term.
Our work provides an unusual opportunity to examine directly links between social structure and inherent social skills that are at the basis of individual and group-level interactions in cognitively advanced mammals [1, 2]. Cognition encompasses the mechanisms by which animals acquire, process, store and act on information from the environment, including perception, learning, memory and decision-making . Responses in our two playback experiments suggest that functionally important decision-making abilities may be significantly altered by disruption of the natural structure of kin-based social relationships. Contrasting patterns of responses to socially unfamiliar elephants in our initial tests of social knowledge could conceivably be driven by differences in social attitudes, if lack of opportunities to bond with kin in the original Pilanesberg population resulted in greater acceptance of unknown individuals [11, 29]. However, it is important to note that the Pilanesberg elephants did not show lower levels of defensive bunching overall - instead they simply failed to focus their defensive bunching on the most socially threatening individuals. Moreover, our main series of experiments subsequently tested for a social skill with direct functional relevance in both populations, the ability to assess age-related social dominance [26–28]. Here again, Pilanesberg elephants were apparently unable to distinguish between the level of social threat presented by older versus younger callers.
Previous studies have documented that a single traumatic event is sufficient to impact the neurological development of the mammalian brain [17, 18, 33, 34], and the large hippocampus of the African elephant, which mediates social memory, is thought to be particularly susceptible during growth to adolescence . The relative importance that such neurological changes might have in generating impaired decision-making versus the consequences of lack of exposure to older more experienced group members in the years following the traumatic event is hard to assess, but both may be important in driving our results. Exposure to older more experienced individuals has been shown to facilitate the development of functionally important skills in a range of mammals see [23, 24] for reviews, and non-human primates deprived of appropriate role models acquire a smaller set of learned skills [23, 35]. Although social learning has not been definitively demonstrated in wild African elephants, there is evidence that knowledge transfer does occur between experienced and naïve individuals  in common with many other large brained, socially complex species [23, 24, 37]. Further studies are now required to partition out these potential effects, and to assess their generality across populations that have experienced differing levels of disturbance.
Understanding the impacts of disrupting social bonds can both provide crucial insights into processes central to social evolution and also throw light on the functioning of advanced mammal societies that have been radically impacted by human disturbance. Our findings suggest that the health and social functioning of wild populations of long-lived and highly social species could be significantly impacted in the long-term by elevated levels of anthropogenic disturbance, which may compromise the ability of surviving individuals to respond appropriately to their conspecifics. Impairments to decision-making processes about threat may also contribute to the development of abnormally aggressive behaviour in response to other species, such as the killing of humans by female elephants in five populations established from translocated individuals that were the survivors of culls .
Although recent empirical evidence has highlighted the value of conserving functioning kin-based family groups, this remains an important issue that is often overlooked by wildlife practitioners in favour of population level management approaches that focus primarily on abundance . In particular, while the recovery of populations from human-induced depletion is often assessed on the basis of numbers, it is now becoming clear that abnormal social structure may be a more persistent effect with very significant consequences [9, 11, 13, 40, 41]. These issues are currently very relevant, as translocation of mammal groups to new areas is becoming an increasingly common response in dealing with situations of animal-human conflict , whilst the escalation of poaching is having a dramatic effect on the structure of many populations . Furthermore, in future years increasing demands on natural resources and ecosystem services from human societies is likely to intensify social disruption and conflict [14, 43, 44]. There is an assumption that wildlife responds to such pressures only in terms of demography, however our study demonstrates that cognitively advanced species such as elephants that live in complex societies may suffer more profound effects.