COS 32-4
Is cryptic diversity in the Pilosella leaf gall wasp associated with geography, host plant, or Wolbachia?

Tuesday, August 6, 2013: 9:00 AM
M100HC, Minneapolis Convention Center
Chandra E. Moffat, Department of Biology, University of New Brunswick, Fredericton, NB, Canada
Robert G. Lalonde, Department of Biology, University of British Columbia Okanagan, Kelowna, BC, Canada
Jason Pither, Biology, University of British Columbia, Okanagan Campus, Kelowna, BC, Canada
Gitta Grosskopf-Lachat, CABI Europe-Switzerland, Delémont, Switzerland
Kevin D. Floate, Lethbridge Research Centre, Agriculture and Agri-Food Canada, Lethbridge, AB, Canada
Background/Question/Methods

Biological control programs require accurate assessments of the host-range of agents prior to introduction.  It is increasingly clear that cryptic genetic variation is common in specialist herbivorous insects.  For weed biocontrol agents, genetic structure has been reported only for a few post-introduction systems.  We demonstrate, for the first time, cryptic genetic differentiation in a candidate weed biocontrol agent prior to its release.

The gall wasp Aulacidea pilosellae (Cynipidae) is a candidate biocontrol agent for multiple species of European hawkweeds (Pilosella, syn. Hieracium) that are invasive in North America.  Preliminary surveys in the wasp’s native range in Europe suggest that it has both Northern and Southern biotypes, which appear to differ in voltinism, host range, and reproductive mode. We surveyed wasps on multiple host species in four geographic areas and sequenced three gene regions (CO1, 28S, ITSII) to determine whether (i) there was genetic evidence for multiple bio-types, and (ii) genetic variation was associated with geographic separation (e.g local adaptation, as suggested by the difference in voltinism), host-plant association (differences in host range) and/or infections with the bacterial endosymbiont Wolbachia (known to alter reproductive mode in Hymenoptera). 

Results/Conclusions

We found considerable genetic divergence among populations of A. pilosellae, supporting the hypothesis that this species has multiple biotypes. We found this variation only in CO1, not in the nuclear regions. A 50% majority-rule consensus phylogeny suggested three distinct lineages, which primarily corresponded to host association, not geography. Wolbachia infection was found in multiple populations, but all of these were in one host-associated lineage. Our results have important implications for hawkweed biocontrol, because they suggest that different wasp accessions will better target different members of the invasive Pilosella complex. Our results demonstrate the value of genetic typing in source populations of biocontrol agents, in order to define host associations and reduce the risk of non-target attack.