1.Parasite virulence varies greatly. Theory predicts that this arises from parasites optimising a trade-off between the mortality they inflict on current hosts, and their transmission to future hosts. The effect of the environment on this coevolution is rarely considered. 2.Geographic mosaics are fertile systems for studying coevolution, but again, the diversity of outcomes is often assumed to result from co-evolutionary dynamism, rather than being moulded by the environment. 3.Here we quantify variation in virulence among lakes in a geographic mosaic of coevolution between a trematode ectoparasite (Gyrodactylus arcuatus) and its three-spined stickleback (Gasterosteus aculeatus) host. 4.Virulence varies greatly in this system, and parasites are generally locally adapted to their hosts. 5.Parasites are also locally adapted to the water in their own lake, and virulence is strongly related to lake pH, the dominant axis of abiotic environmental variation in this system. 6.These results suggest that the evolution of virulence can be substantially affected by the abiotic environment, which has important implications for understanding coevolution. There are also implications for the evolutionary management of disease e.g. ectoparasites in aquaculture, the impacts of which might be expected to reduce given ongoing acidification of aquatic ecosystems.

Gasterosteus and Gyrodactylus: results of artificial infection experiments showing local adaptationThe file contains the results from five experiments. Experiments 1 to 4 show the outcome of artificial infections of stickleback hosts from different populations (lakes) with Gyrodactylus parasites from different lakes. Data shown are the populations of origin of the host and parasite, the sex, length (mm) and family (where applicable) of hosts, and the total number of parasites counted on the host over the course of the experiment. Different experiments are for different combinations of host and parasite populations, and had slightly different experimental conditions (e.g. number of parasites used to start an infection, and frequency of counts). Experiment 5 shows the results (in hours until death) of an experiment in which parasites from different lakes (and detached from hosts), were reciprocally exposed to water from the same lakes.Func Ecol DRYAD data.xlsx