PARENT SESSION
Poster Session 20: Invasive Species
Wednesday, August 10, 5:00 PM - 6:30 PM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal

Nearest Neighbor Analysis of the Effects of the Rust Fungus Uromyces scutellatus on Euphorbia spp. in Europe.

Caesar, Anthony^*,1, ¹ USDA/ARS Pest Management Research Unit, Sidney, MT, U.S.A.

ABSTRACT- Biological control of exotic invasive plants is based on the Enemy Release Hypothesis. In practice, this entails searches within the native range of the invasive for insects or plant pathogens that can damage or cause disease on the target plant species. It is increasingly understood in the field of invasive weed biocontrol that a new departure away from a strict reliance on narrow host range agents (chiefly insects) is needed for control strategies. The prevailing method has come to be known as the lottery approach and such an approach is seen as increasing the risk to nontarget species and increasing the hazards of other indirect effects. A consensus is emerging that recommends a focus on fewer, more effective agents. Thus, following tests that show a narrow host range, the magnitude of potential impacts should be assessed prior to field release. In addition to such prerelease studies there should be an assessment of impact in the native range. This would be most valuable even prior to prerelease studies. This approach should apply to candidate species that are plant pathogens as well as insects and the method would be useful where stands are irregular in size and occur in small patches. I assessed the effects of the autoecious, microcyclic rust fungus Uromyces scutellatus on stand density of Euphorbia esula/virgata by comparing population densities among stands with and without symptoms using nearest neighbor spatial analysis. At three of four sites assessed in Hungary and Austria in 2004, symptomless E. esula virgata plants were significantly more closely spaced as determined with ANOVA. The plant densities of rusted stands estimated from the distance data were 48-73% of those of neighboring symptomless stands. Thus, it appears that the candidate rust U. scutellatus reduces stand density where it occurs at sites in Europe compared to adjacent or nearby stands without rust disease. This data indicates the potential value of this fungus for effective biocontrol, especially if shown to be highly specific to E. esula/virgata. Additional data will be collected in 2005 at the same sites as in 2004 and at additional sites. Such data could be a valuable addition to current protocols to improve the effectiveness and the ecological safety of weed biocontrol.

Key words: nearest neighbor spatial analysis, biological control, enemy release hypothesis, prerelease studies