Humans, animals and plants often become co-infected with different genotypes of the same parasite species. A number of theoretical studies have predicted that co-infections select for higher levels of parasite virulence. All these models are based on the assumption that more virulent parasites are competitively superior and obtain greater transmission from mixed infections than less virulent parasites. However, few experimental studies have tested this fundamental assumption underlying theoretical models. Instead, many studies have focused on directly comparing some proxies of virulence of single-genotype and mixed-genotype infections of experimentally infected hosts. Thus, at present we are left with theoretical models predicting evolutionary end-point magnitudes of virulence at the population level and experimental studies measuring virulence at the scale of one co-infected host individual. To bridge the gap between theory and experiment, we here develop a simple mathematical model to investigate the consequences of the three most commonly assumed (theory) and tested (experiments) mixed-infection scenarios on both the level of virulence at the scale of one coinfection and the evolutionary end-point magnitudes of virulence at the population level. 

Results/Conclusions

Under the first scenario, only competition for host resources occurs. At the scale of one coinfection, this leads to a parasite density and virulence which are the same as thOSE of the most virulent parasite on its own and, on over long-term evolutionary time, this leads to an increased level of virulence. Under the second scenario, parasites in co-infected hosts respond to the presence of each other by phenotypically increasing their growth rate. This results in a higher parasite density and virulence at the scale of one coinfection but no increase of virulence over long-term evolutionary time. Under the last scenario, the host immune system response against two different parasite genotypes is less efficient than against one parasite genotype. At the scale of one coinfection, this leads to a parasite density which is the same as that of the most virulent parasite on its own but also to an increased virulence, whereas it leads to a decreased virulence over long-term evolution. By demonstrating the absence of correlation between levels of virulence that can be measured experimentally and the oneS that are reached over evolutionary time, this model stresses the need for experimentalists to track the intra-host parasite densities in order to identify the mecHanisms responsible for the measured virulence.