This action funds an NSF Postdoctoral Research Fellowship in Biology for FY 2011, Broadening Participation. The fellowship supports a research and training plan in a host laboratory for the Fellow who also presents a plan to broaden participation in biology. The title of the research and training plan for this fellowship to Shermin DeSilva is "Social stratification and leadership in elephants." The host institution for this fellowship is the Colorado State University, and the sponsoring scientist is Dr. George Wittemyer.
Socioecology explores how social and ecological factors structure animal societies. This research builds much-needed bridges between theory and empirical observation in the study of social evolution. It poses three inter-related questions: 1) How does resource availability influence social cohesion? 2) What effect does this have on dominance and leadership behavior? 3) How do individuals adjust to social disruption? Theoretical predictions are tested by comparing the social behavior of wild Asian and African elephants, sister species that experience different ecological pressures. It examines how the spatial arrangement and abundance of resources shape social relationships via individual movement decisions, whether this in turn drives the emergence of leadership and dominance hierarchies, and ultimately how individuals cope with changes such as the loss of a companion.
Training objectives include acquiring experience working with a diverse array of data ranging from animal movements, to satellite imagery, and social networks. It also offers a comparative perspective on research and conservation in the African context relative to the Asian, with direct implications for wildlife management. Both field sites host local and foreign students at the undergraduate and graduate level, including those from the U.S. and Europe, encouraging collaboration and exchange of expertise. Researchers also oversee community-based education and sustainable livelihood programs seeking to reduce conflict between people and elephants. Long-term goals are not only to further the study of behavior, but also to address the challenges people and wildlife face in living together through participatory science-based conservation.
Intellectual merit What does it mean to say an animal is ‘social’? Humans and other animals congregate in a number of different ways. Socioecological theory contends that the structure and function of social groups is a result of their social and ecological context, namely factors such as resource distribution and abundance, predation, and competition or conflict with those in one’s own group or others. However before an gathering of individuals can be termed ‘social’ it is necessary to established that these individuals are indeed together because they are expressing a social preference for one another, rather than a preference from something else – such as a resource in the environment. In this work we critically examine some of the statistical methods used to establish this basic starting point as well as provide improved approaches to this problem. We focus on elephants, both Asian and African, because these sister species exhibit both express complex social behavior as well as occupy different habitats on the two continents. They therefore offer the opportunity to compare and contrast the relationship between social behavior and ecology, while keeping constant basic similarities due to their shared evolutionary history. Broader impacts The methods we have developed can potentially be applied to other contexts – other species, or other questions. For instance, our approach offers a way to distinguish the social vs. spatial drivers of contact networks, which may have applications to the study of how diseases could spread. With respect to elephants specifically, both the behavioral studies as well as the demographic studies that were concurrently conducted, have implications for managing wild elephant populations. In particular it is important to consider social consequences of management practices such as translocation, which bring to contact individuals who would not otherwise occupy the same area, which can create competition, conflict, and ultimately have adverse consequences for well-being. This is especially relevant for Asian elephants, which are globally endangered. We also published a study of the basic demographics of this population. During this work, we have also in public talks, video presentations, a TV documentary film, social and blogging media, and other ways to communicate our research findings and raise the visibility of conservation issues concerning Asian elephants. Summary Asian (Elephas maximus) and African savannah elephants (Loxodonta africana) are sister taxa that have diverged by approximately 6 million years but exhibit gross physical and ecological similarities as climax herbivores. Conditions at the two field sites in Kenya and Sri Lanka differ, driven by the difference in absolute rainfall each location receives. The African savannah environment is drier than the Sri Lankan. We found that Asian elephants exhibit more dynamic contact patterns than African elephants, typically with smaller day-to-day group sizes which fluctuate more than the associations. As a result, they do not exhibit the clear social stratification exibited by the African sister species. We showed that African elephants display seasonal (wet vs. dry) association patterns with multiple 'tiers' that are absent in the Asian population, although both species display extended networks of associations. Correspondingly, female Asian elephants also do not exhibit clear dominance hierarchies, unlike female African elephants, which have hierarchies ordered by age (which also correlates with size). In this study we also examined the statistical basis for evaluating what constitutes a ‘social’ relationship, and whether an individual’s position in a network is meaningful or random. We were concerned with how tests performed when datasets were impoverished or when social preferences were relatively weak. We found that one version of a standard test performs reasonably well so long as certain constraints are observed, but that it has to be applied with caution when examining network properties. Secondly we develop a new methodology that uses where and when individuals are seen to derive their relationship to one another and also group individuals together in terms of their social affiliations. This method rests on the intuition that individuals who truly have social preferences for one another are more likely to appear at the same place at the same time than those that do not, even if particular locations contain features (e.g. resources) that are attractive to them. We thereby distinguish these ‘social’ preferences, from encounters due to external factors.
