Background/Question/Methods Rare plant species are expected to benefit from reduced pressure from host-specific pathogens that behave in a density-responsive manner. In addition, rare species may benefit from reduced competition by locally dominant plants that suffer greater disease pressure. Such variation in pressure from host-specific pathogens and pests has long been suggests as a mechanism to help maintain plant diversity. However, most plant pathogens are polyphagous rather than host specific – they are able to cause disease on a number of host species in any given community. This then raises the question, why should there be a rare-species advantage if most pathogens are not host specific?

Results/Conclusions Cross-inoculation studies and analysis of published host ranges show that which hosts are susceptible is not a random subset of the community; host range is phylogenetically structured so that closely related plant species are more likely to share a given pathogen. We applied an empirically derived relationship between shared susceptibility and evolutionary distance between plant species to estimate the likelihood that pairs of plant species should share a particular pathogen. When combined with local density estimates from temperate and tropical mapped forest plots, we found that a locally rare plant is often just one of several alternative hosts for a pathogen that together increase host abundance. This should effectively dampen the rare-species advantage expected in forest communities. For plant pathogens to shape plant community structure through rare-species advantage, more specialized pathogens must have a greater impact on hosts than do less specialized pathogens. Alternatively, host competence (the ability of a plant to support pathogen reproduction) must be more phylogenetically restricted than host susceptibility.