A related point is that coevolutionary models of host-pathogen interactions imply coevolution at the level of local host/pathogen populations whereas the sampled fungus strains and host plants tested here are not supported by an ecological/biological rationale. In fact, besides the California grape isolates, the pathogen samples are from a global population and there is no evidence as far as I know that for example Arabidopsis is naturally infected by B. cinerea. Thus I do question whether evolutionary statements can be drawn from the studied pathogen/host collection in the absence of ecological or biological evidence of coevolution.