Invasive Species Compendium

Detailed coverage of invasive species threatening livelihoods and the environment worldwide

Abstract

Non-target impacts of Phytomyza vitalbae a biological control agent of the European weed Clematis vitalba in New Zealand.

Abstract

The agromyzid leaf-mining fly Phytomyza vitalbae, which was introduced into New Zealand as a biological control agent of the invasive deciduous European vine Clematis vitalba L. (old man's beard; Ranunculaceae) in 1996, was recorded attacking native non-target Clematis forsteri and Clematis foetida in New Zealand, with C. foetida being most commonly attacked. Both the incidence and levels of P. vitalbae attack were significantly lower on the non-target species, compared to on the target plant and populations of non-target plants that were growing within 4 km of the nearest known patch of C. vitalba were most commonly attacked. No-choice starvation tests indicated that survival of P. vitalbae was low and oviposition did not occur on C. foetida unless flies had previously fed on C. vitalba until they began ovipositing, indicating that non-target attack was a "spillover" effect that is unlikely to have a major detrimental impact on the non-target plants. Parasitoid rearing indicated that P. vitalbae shares parasitoids with a closely related native-leaf-miner Phytomyza clematadi. Therefore, P. vitalbae and P. clematadi potentially compete with each other, mediated through shared parasitoids. However, we found no evidence for increasing levels of parasitism of P. clematadi mines, or for a reduction in P. clematadi abundance on C. foetida with increasing proximity to C. vitalba infestations. Both the direct and indirect non-target impacts of P. vitalbae attack, therefore, appear to be minor. Host-range testing performed prior to the release of P. vitalbae in New Zealand predicted the plant species that were most at risk of non-target attack. Our results show that the prevalence of spillover onto non-target species was underestimated and we discuss how host-range testing might be improved in the light of these findings.