Field Censuses.

Sites containing at least one of the three focal morning glory species in North Carolina were censused to determine the natural distribution of infection during the years 2006–2008 (Fig. 1). Each year, we visited these communities during the early summer after plant germination had occurred and recorded which Ipomoea species were present. During repeated visits over the summer, we assessed the proportion of plants of each species harboring infections. By late August, these proportions fell into two discrete categories: >90% of plants infected or <10% of plants infected.

Map of census site locations and sites used in the cross-inoculation experiment. The presence of host species and their infection status, if present, is indicated by colored portions of divided circles. Each circle includes three portions, one for each host species, as indicated in Inset. White portions indicate the absence of the given host species, green portions indicate presence without infection, and orange portions indicate presence and infection. Approximate locations are indicated by black dots or stars; stars indicate the five censused locations chosen for cross-inoculation experimentation.