Re-test of Rhinocyllus conicus host specificity, and the prediction of ecological risk in biological control

Biological control is proposed as an ecological strategy to manage the threat of invasive plants, especially in natural areas. To pursue this strategy, we need to know that the host specificity criteria used to evaluate ecological risk with deliberate introduction of an exotic insect for biocontrol are sufficient to predict potential impact on native species. Host specificity is defined by adult feeding and oviposition preferences and larval development. One way to evaluate the criteria is to re-examine case histories where ecological effects are recorded, such as that of Rhinocyllus conicus Frölich. This flower head weevil, released in North America in 1968 to control exotic thistles like Musk thistle (Carduus nutans L), is now reducing seed production by multiple native North American thistle species (Cirsium spp.), and local population density of Platte thistle (Cirsium canescens Nutt.). We hypothesized that host specificity of R. conicus has changed since pre-release testing, providing an explanation for the unexpected magnitude of the documented ecological effects. Instead, when we re-tested host specificity of weevils naturalized over 28 generations, we found that host specificity has not changed. Naturalized adults of R. conicus showed strong feeding and oviposition preference for Musk thistle over Platte thistle. In addition, larval development by these weevils was faster and more successful (to larger size) on Musk thistle than on Platte thistle. Thus, our results indicate that a change in host specificity cannot explain the unexpectedly large build-up of R. conicus and significant ecological effect on Platte thistle. We conclude that accurate prediction of the potential level of impact on native host plants in the field requires further ecological information in addition to host specificity.