COS 87-2
Non-host neighbors provide associational susceptibility of a native thistle to an invasive weevil

Wednesday, August 12, 2015: 1:50 PM
326, Baltimore Convention Center
Alyssa S. Hakes, Biology, Lawrence University, Appleton, WI
Background/Question/Methods: The non-native weevil, Larinus planus, is a biocontrol agent for the noxious weed, Canada thistle, but also attacks native thistles including the federally threatened dune thistle, Cirsium pitcheri. Seed predation from L. planus has the potential to cut C. pitcheri’s extinction time by half. To help prevent further damage to this vulnerable species, it is important to understand the factors that influence weevil damage. An observational study conducted to document the spatial variability in weevil damage revealed that C. pitcheri located near Ammophila grass and at lower dune elevations were associated with more Larinus damage, although these two variables (proximity to grass neighbors and elevation) were confounded (grass density decreased as elevation increased). To test the hypothesis that non-host grass neighbors increase susceptibility of C. pitcheri to L. planus independent of elevation, I manipulated the presence of neighbors surrounding thistles at low and high elevations. Ninety-nine reproductive thistles were tagged; one-third located at high elevations that had grass neighbors, one-third located at low elevations that had grass neighbors, and the remaining third located at low elevations that had grass neighbors within a 1 m radius clipped. Weevil oviposition, proportion of flowers damaged, and seed predation was monitored.

Results/Conclusions: There was a significant difference in the frequency of oviposition holes among treatments; 44% of C. pitcheri located at low elevations with grass neighbors had oviposition holes present, whereas the frequency of oviposition holes was lower for plants at the same elevation with grass neighbors (9%) and for plants at higher elevations (15%). This difference in damage among treatments was also significant at the season’s end. C. pitcheri in the low-elevation grass treatment had on average twice as many flower heads damaged than those in the low-elevation clipped treatment. There was not a significant difference in seed predation between the low-elevation clipped treatment and high elevation grass treatment. Together these results suggest that L. planus weevils prefer to disperse at low dune elevations and through the dominant Ammophila grass matrix to locate and climb onto its host. C. pitcheri that germinate further away from beach grass may have a better chance of successfully reproducing. I discuss the mechanisms that may explain why a non-host grass provides associational susceptibility to an invasive herbivore’s unintended host.