Does fundamental host range testing reflect ecological host range: Larval substrate considerations informing Drosophila suzukii classical biological control

Host specificity testing of candidate importation biological control agents is crucial for determining their suitability for release into the invaded range of a target pest species. Ecological risks associated with such releases are often characterized using physiological host range testing. Laboratory experiments oversimplify natural processes that influence non-target host availability and physiological host range testing can overestimate non-target attack in nature. In the Drosophila suzukii Matsumura (Diptera: Drosophilidae) biocontrol system, two parasitoids – Leptopilina japonica (Novković & Kimura) and Ganaspis brasiliensis Ihering (both Hymenoptera: Figitidae) – have been considered for the control of the small fruit pest. Past physiological host range work for these parasitoids assessed non-target attack in cornmeal Drosophila rearing substrate; an unnatural microhabitat for non-target larvae. We conduct similar laboratory experiments to assess non-target parasitism by L. japonica both in cornmeal diet and spoiled cherries – a more likely natural substrate for non target larvae – and compare L. japonica parasitism between substrates and non-target Drosophila species. We found no difference in parasitoid attack between rearing substrates but found differences in parasitism rate between Drosophila species. Our findings improve the ecological relevance of past physiological host range testing for L. japonica as non-target attack in cornmeal adequately describes non-target attack in spoiled fruits in the laboratory.