Social Networking for Animals?

When we hear the words “social network,” most of us think about the worlds of social media on Twitter and Facebook. An ecologist or animal behaviorist may think of a different type of social network, a model of the social connections between animals, not utilizing the internet. A social network is most simply a set of social units and the relationships between them, represented by nodes and edges respectively¹. A human social network could be developed where an individual is the social unit or node and the ties between each individual or edges are determined by their association. This association between people could be defined by a number of different connections, maybe phone calls between them or just being friends on Facebook.

An example of a human social network using Facebook connections. http://www.touchgraph.com/facebook 

Let’s imagine trying to develop a social network that exists in a typical high school. As a researcher, you would observe teenagers interacting at different times of day, maybe before school begins and during lunch. The network could be determined by interactions and associations. If you were looking at interactions, you could count every high five or hug between individuals. Associations could be determined by people sitting together at lunch. Mapping these relationships would give a picture of the high school students’ cliques and which individuals bridged between these groups.  Other factors could also be analyzed to see what determined these groupings. Were the teenagers associating because they had similar hobbies? Would the stereotypical cliques emerge, with the jocks, band kids, and skaters only associating together? Or would the associations be more strongly related by race or social class? This snapshot of a young population of our society brings insight into the structure of human society.

The study of this type of network is termed Social Network Analysis (SNA). Scientists create graphical representations of the network called sociograms to visualize the connections in a social group and determine its structure. Then mathematical modeling can be used to detect patterns within the networks.² These patterns can be compared with other characteristics of the individuals to see which factors determine the social connections. For example, are the social units associating together because they are genetically related or they are a similar age? The network sociograms can convey a lot of information that would be difficult to graph in other ways, such as the strength and directionality of the social connections. The strength of the whole social network is determined by the number of connections between individuals relative to the total possible connections. A network can also be analyzed to determine the importance of an individual in the group through the number of connections they have to others.¹ By determining these key players in a network, predictions can be made about the social group. For example, social learning can be predicted using this type of analysis. If an individual who is very central to the network learns a task, it may be predicted that the others in the group will learn that task faster than if an individual on the outside of the group learns the task. Disease transmission can also be predicted by looking at social networks.  

  

A human social network within a company, demonstrating that Alan is a central individual.
http://hbr.org/2002/06/the-people-who-make-organizations-go-or-stop/ar/1

Analysis of the networks of other species can give us insight into how their societies are structured differently than ours and the factors affecting relationships. These factors may be genetic relatedness, age, or endocrine levels of the individuals. The relationships could also be controlled by outside factors, such as season or availability of food in the environment. The insights into social dynamics of an animal population gained from these analyses are useful for wildlife management strategies and conservation. It is important to maintain the strength of the social structure and protect the key individuals crucial for the cohesion of a network. ⁴ Monitoring the social network of a population can determine any effects human activities are having on the well-being of the population.

Social network analysis has helped to clarify the social organization of both African and Asian elephants and allowed for comparison between the two species. The analysis of African elephant society has demonstrated that their society is made up of hierarchical tiers. The most basic level of association is between the mother and calf, who sometimes group with other closely related females, and then sometimes with multiple groups of families, and finally with multiple herds. The elephants grouped into these different tiers sometimes depending on ecological factors such as food availability or season.⁷ In a study of Asian elephants inhabiting a national park in Sri Lanka, social network analysis helped determine that most ties between individuals are weaker than in African elephants. The Asian elephant network is less interconnected and therefore they cannot have the same hierarchical structure as African elephants.⁶ Another study demonstrated that there was greater association between Asian elephants in the dry season, which is very different from African elephants who associate more in the wet season. It is possible that Asian elephants aggregate in the dry season in order to access and protect scarce resources. ⁵

Social networks of Asian (E. maximus) and African (L. africana) elephants. The strength of the African elephant network is demonstrated by the dense connections between individuals and greater fragmentation when weak ties are removed. de Silva & Wittemyer,2012 

Using a network approach, the social interactions of male elephants have been further investigated. It was previously thought that male elephants had very weak social ties, but a recent study has shown that their sociality is affected by sexual state. When looking at a network of only African males that were not sexually active, the network appears to be much denser and the associations between males stronger. SNA allows for these patterns to emerge and the ability to visualize these male elephant relationships. It was also shown that sexually inactive males were associating with other males based on age. ⁸
Social network analysis is an exciting tool for the investigation of the social structure of many different populations, including our own. Further research on elephant societies in the wild may show that there are more factors influencing their society than we currently understand. It will also be an important tool as we continue to investigate the impact of poaching on elephant societies. Think about your own social network and the factors influencing the people who you are connected with. Although our societies are very different, we form relationships in some of the same ways. Maybe our social groups don't change based on season, but humans do tend to associate with relatives.

Two African elephant families hanging out in a river together.

¹ Wasserman S. & Faust K. 1994. Social network analysis: methods and applications. In: Granovetter M, editor. Structural analysis in the social sciences. Cambridge: Cambridge University Press. 825.

² Freeman LC. 2004. The development of social network analysis. Vancouver, BC: Empirical Press.

³ Brent L., Lehmann J., & Ramos-Fernandez G. 2011. Social network analysis in the study of non-human primates: a historical perspective. American Journal of Primatology. 73, 720-730.

⁴ Lusseau D, NewmanMEJ. 2004. Identifying the role that animals play in their social networks. Biology Letters. 271:S477–S481.

⁵ de Silva S., Ranjeewa A., & Kryazhimskiy S. 2011. The dynamics of social networks among female Asian elephants. BMC Ecology. 11:17.

⁶ de Silva S. & Wittemyer G. 2012. A comparison of social organization  in Asian elephants and African Savannah elephants. International Journal of Primatology. 

⁷ Wittemyer G., Douglas-Hamilton I., & Getz W. M. 2005. The socioecology of elephants: analysis of the processes creating multitiered social structures. Animal Behaviour. 69, 1357-1371.

⁸ Goldenberg S., de Silva S., Rasmussen H., Douglas-Hamilton I., & Wittemyer G. 2014. Controlling for behavioural state reveals social dynamics among male African elephants, Loxodonta africana. Animal Behaviour. 95, 111-119.