Virulence is expected to be linked to parasite fitness via transmission. However, it is not clear whether this relationship is genetically determined, nor if it differs when transmission occurs continuously during, or only at the end of, the infection period. Here, we used inbred lines of the macro-parasitic spider mite Tetranychus urticae to disentangle genetic vs non-genetic correlations among traits, while varying parasite density and opportunities for transmission. A positive genetic correlation between virulence and the number of transmitting stages produced was found under continuous transmission. However, if transmission occurred only at the end of the infection period, this genetic correlation disappeared. Instead, we observed a negative relationship between virulence and the number of transmitting stages, driven by density dependence. Thus, within-host density dependence caused by reduced opportunities for transmission may hamper selection for higher virulence, providing a novel explanation as to why limited host availability leads to lower virulence.

This dataset was collected by infecting 4 cm² leaf patches of bean plants (Phaseoulus vulgaris, variety Pongo) with different inbred lines of the macroparasite Tetranychus urticae (spider mites) at different densities (5, 10 or 20) for 4 days. After this period,  spider mites were removed, the number of eggs was measured and a photograph was taken to determine the percentage of host tissue damaged by spider mites. The offspring emerging from the eggs laid by the parasite was allowed to devolp until adulthood for 14 days. Transmission to uninfected hosts was allowed during the infection period (experiment 1 - Continuous transmission) or not (experiment 2 - Transmission at the end of infection period). After 14 days, the total number of adult daughters (transmiting stage) produced was determined, as well as the number of daughters infection new hosts (for experiment 1).