Here, we provide a worked example of network analysis in a wild animal population using data from Weber and colleagues (2013). The data in this study were collected using proximity loggers deployed on 51 individuals in a UK population of European badgers (Meles meles) naturally infected with bovine tuberculosis (for more details on the methods, we refer readers to the original study).

We provide R code demonstrating how to calculate the individual-level and population-level network metrics discussed in this article (see table 2), plot the network, and calculate its community structure and modularity (see supplemental material). The badger population has a social network with high modularity and six cliques or communities detected (Q = 0.75 for this subdivision). Modularity structure is driven principally by association with a main sett (the communal burrows used by territorial social groups) and is illustrated by node color in figure 3. There is also considerable individual variation in centrality in this network (table S1), and this is demonstrated by the size of the nodes in figure 3.

Figure 3.

A contact network of badger social interactions. Nodes are colored by social community membership as was determined by the fastgreedy algorithm applied in igraph. Node size is related to the flow betweenness of the individual; those with greater flow betweenness are represented by larger nodes. Badger image obtained under a Creative Commons license from http://animalsclipart.com/stencil-badger-clipart-design.

Open in new tabDownload slide

The relationship between network position and infection is complex in this population, with infected individuals tending to form more out-of-group contacts and fewer within-group contacts at particular times of year (Weber et al. 2013). Nevertheless, there is an overall trend for bTB-infected individuals to have higher-degree (i.e., more contacts) and lower P_i (i.e., more out-of-group contacts), as would be expected from these results (figure 4).

Figure 4.

The degree (number of connections) and participation coefficient (Pi; proportion of connection within their own network community) for European badgers in relation to bovine tuberculosis infection status as defined by Weber and colleagues (2013). The solid line represents the median, the box the interquartile range and the whiskers extend 1.5 times above and below the upper and lower quartiles respectively.

Open in new tabDownload slide

By using the information in figure 3 and supplemental table S2, it would be theoretically possible to identify the individuals most likely to spread infection through the network, raising the possibility that such individuals could be targeted by management interventions. For example, individual 29 has the highest values of flow betweenness and degree and the lowest value of P_i and so may be pivotal in facilitating disease transmission through the network.