The impact of heterogeneous contact structure on the evolution of pathogen virulence

Abstract Classical theory on pathogen evolution traditionally assumed homogeneous host populations with random interactions for simplicity. However, in most biological systems, individuals interact locally and exhibit significant heterogeneity, where a minority of ‘superspreaders’ accounts for the majority of transmission. While local interactions are known to select for ‘prudent’ pathogens, those that transmit more slowly and cause less host damage, previous research suggested that superspreading had no long-term evolutionary effect unless linked to traits impacting host survival. However, since all infectious disease systems are likely to show a degree of both local and heterogeneous interactions, a realistic model should include both these sources of population structure. In contrast to results assuming random transmission, we demonstrate that when local population structure is accounted for, heterogeneity can act as a key driver of virulence evolution. Specifically, superspreading selects for slower-transmitting, less virulent pathogens. Social network analyses suggest that many wildlife and human diseases exhibit this type of transmission heterogeneity, which directly shapes pathogen development. Our findings reveal that the evolutionary role of superspreading may have been overlooked, compared to its well-known impacts on epidemiology. These results provide a more realistic framework for predicting disease evolution in structured, real-world populations.