Assessing the non-target impacts of classical biological control agents: is host-testing always necessary?

Abstract

Release of a biocontrol agent in New Zealand is typically preceded by non-target testing of native or valued species. Nevertheless, if both the target pest and the natural enemy are very different from any native fauna, then there may be no scientific justification for host testing. Gonatocerus ashmeadi (Girault) (Hymenoptera: Mymaridae) is being considered as a biocontrol agent for glassy winged sharpshooter, Homalodisca vitripennis (Germar) (Hemiptera: Cicadellidae), should the pest arrive. An assessment of the potential impact of G. ashmeadi on New Zealand’s Cicadellidae and Membracidae, from published literature data, indicates that none of these insects is at risk, as their eggs will not be recognised by the parasitoid because either their size or location places them outside the parasitoid’s search pattern. Consequently, there is no scientific case for any non-target host-testing to be carried out in containment.