RESUMO
Herbivore-induced plant volatiles (HIPVs) are important cues for natural enemies to find their hosts. HIPVs are usually present as blends and the effects of combinations of individual components are less studied. Here, we investigated plant volatiles in a tritrophic system, comprising the parasitoid wasp Lytopylus rufipes Nees (Hymenoptera: Braconidae), the Oriental fruit moth Grapholita molesta (Busck) (Lepidoptera: Tortricidae), and Japanese pear, Pyrus pyrifolia 'Kosui', so as to elucidate the effects of single components and blends on wasp behaviors. Bioassays in a four-arm olfactometer, using either shoots or their isolated volatiles collected on adsorbent, revealed that female wasps preferred volatiles from host-infested shoots over those from intact shoots. Analyses identified (Z)-3-hexenyl acetate (H), linalool (L), (E)-ß-ocimene (O), (E)-3,8-dimethyl-1,4,7-nonatriene (D), and (E,E)-α-farnesene (F). Among them, only F was induced by infestation with G. molesta. When tested singly, only O and D elicited positive responses by L. rufipes. Binary blends of HO and DF elicited a positive response, but that of HD elicited a negative one, even though D alone elicited a positive response. Remarkably, wasps did not prefer either the ODF or HL blends, but showed a highest positive response to a quinary blend (HLODF). These results show that synergism among volatiles released from host-infested plants is necessary for eliciting high behavioral responses in L. rufipes, enabling L. rufipes to find its host efficiently.