Emergent trade-offs and selection for outbreak frequency in spatial epidemics

van Ballegooijen, W.M. and Boerlijst, M.C., 2004. Emergent trade-offs and selection for outbreak frequency in spatial epidemics. Proceedings of the National Academy of Sciences, 101(52), pp.18246-18250.

Most studies that consider the evolution of pathogens assume inherent life history tradeoffs with infectiousness, mode of transmission, or virulence. This paper described a different evolutionary mechanism that can produce a tradeoff between infection rate and period: spatial selection processes drive pathogen evolution. To study the impact of spatial processes on pathogen evolution, they developed a spatial SIR model (on a 120×120 cell lattice) with infection occurring between neighboring cells. Their general framework allows each cell to be representative of communities, schools, or individuals. The parameters for transmission rate and period could change over time due to random mutations following pathogen spread. But first they studied the behavior of the model without letting transmission rate or infectious period vary. For different values of infection rates and periods, the model produced a variety of self-organized patterns: small localized clusters, regularly spaced circular waves, and turbulent waves.
When the infectious rate and period could evolve, the trajectories of pathogen traits proceeded towards a trade-off between infection rate and period (regardless of initial starting point). The ridge which all trajectories converged to corresponded to parameters which optimized outbreak frequency (average frequency that hosts were infected). They also investigated how their spatial model differed from non-spatial theory in predictions of maximal infection rate and period. Non-spatial theory predicts “runaway” selection. In contrast, their model predicts an intermediate level of infection rate and period. Their results are interesting because instead of just assuming the classical trade-off between infectious rate and period, their model shows how these different trade-offs emerge by explicitly considering multi-scale spatial processes.