The effect of network structure on pathogen prevalence and levels of immunity through time. In the top row, we depict a representative network from each of the five ensembles. The second row shows the distributions of each of three response variables for prevalence of the pathogen and specific immunity over the course of the simulation, with the units for the horizontal axes given by the panel headings. We depict one point for each randomized network structure and box-plots indicating the median and inter-quartile range of each network-type's distribution. Network generating algorithms were tuned to produce networks of the same size and approximate connectance and model parameters were either the same for all populations and across simulations (σ = 8, μ = 0.1, δ = 0.01, and γ = 0.66) or randomized for each population in each simulation (initial densities of infectious and immune individuals [0, 1] and β value corresponding to a \(\tilde{R_0}\) within [1, 6] for each population).