Characterization of chemicals mediating ovipositional host-plant finding byAmyelois transitella females.

Ovipositional host-finding in the navel orangeworm,Amyelois transitella (Walker), is brought about by an in-flight response to host odors. Wind-tunnel studies of the response of gravid females to almonds showed that this response is mediated primarily by long-chain fatty acids, particularly oleic acid and linoleic acid. Evidence for the behavioral activity of fatty acids is based on the fact that: (1) behavioral activity of almond oil was concentrated in a single liquid chromatographic fraction whose composition was predominantly long-chain fatty acids, (2) behavioral activity was lost when either almond oil or the active fraction of that oil was treated with diazomethane, (3) full activity was elicited by a selective extraction of free fatty acids from crude almond oil, and (4) upwind response by females was elicited by a blend of synthetic oleic and linoleic acids, albeit at a level less than that elicited by almond oil. Five fatty acids identified from the almond oil were: myristic acid (1%), palmitic acid (16%), stearic acid (3%), oleic acid (58%), and linoleic (22%). Attraction to various combinations of synthetic acids was observed only when oleic acid was present, and oleic acid elicited upwind flights to the source when presented alone; however, short-range responses were enhanced by the addition of linoleic acid, which elicited no long-range orientation by itself. Despite significant levels of attraction to synthetic blends, the percentage of females flying to the source was lower than that flying to acidulated almond oil, the best natural attractant tested. Thus, although longrange response may be mediated primarily by a blend of oleic and linoleic acids, additional and as yet unidentified components must also play an important role. Long-range chemically modulated host finding in this and other generalist plant feeders is discussed with respect to current models of the evolution of host finding, and it is argued that suggestions that long-range host finding should be correlated with narrowness of host utilization are logically flawed and are not supported by our current understanding of specific examples of host finding.