The average path length and clustering coefficient were 4.39 and 0.37 respectively, showing characteristics of a small-world structure. The monthly network size grew steadily over the study period, reaching its maximum value in the last study month (December 2019) with 31,176 nodes and 18,420 edges. The analysis of the global time series for network metrics using the trend test increased significantly over time (p<0.001) for betweenness, diameter, mean number of neighbours, shortest paths, and giant weakly and strongly connected components; on the other hand, we found a significant decreasing tendency (p<0.001) for closeness, clustering coefficient and edge density (Fig. 5).
Visualizing the Ecuadorian network metrics videos (Supporting information) help to understand the network dynamics for each study year (2017-2019), along with the spatial distribution of movements.
Figure 5. Temporal distribution of monthly networks metrics characterising the 2017–2019 pig movements in Ecuador.