It has been shown that volatile organic compounds (VOCs) emitted from plant tissue can be induced in the presence of herbivory and serve as a defense mechanism by attracting the parasitoid enemies of herbivores.  Sometimes these volatiles are specific, meaning they are produced only in response to an attack and may vary in intensity, while others are released constitutively.  The objectives of this study were to 1) collect and identify the VOCs of the understory shrub, Lindera benzoin, 2) investigate the specificity of L. benzoin‘s VOCs by analyzing scent bouquets produced in response to different damage treatments (herbivory, mechanical wounding, and no damage), and 3) test the hypothesis that parasitoids are attracted to short range cues from plants induced by the presence of a specialist caterpillar, Epimecis hortaria, feeding on L. benzoin.  Volatiles were collected using field solid-phase microextraction (SPME) and analyzed using gas chromatography-mass spectrometry.  In a second experiment, insect traps were constructed around L. benzoin leaves with each of the three damage treatments, in order to determine if parasitoids were preferentially attracted to specific cues induced by the different types of damage. 

Results/Conclusions

Seventeen VOCs were positively identified (library match of 90% or better, and verification with known standards or KI), using the SPME / GC-MS method.  Trans-caryophyllene, alpha humulene, and (E)-4,8-dimethyl-1,3,7-nonatriene (DMNT), were present as some percentage of the scent blend in every collection.  The green leaf volatile cis-3-hexen-1-ol, known to be a parasitoid attractant in other systems, was found in several collections, however it did not seem to be specific to the herbivory damage treatment.  The remaining 14 known compounds varied in presence/absence and in relative percentages.  While no obvious correlation between VOC emissions and damage treatments was seen, there was a large amount of variation among individuals.  There were two clearly distinct types of scent blends observed; some individuals produced four to five compounds, while others contained seventeen to twenty five compounds.  The variation was consistent over time, leading to the hypothesis that perhaps there are different “phenotypes” of scent blends causing the individual variation. No significant differences were seen in the number of insects attracted to different damage types in the insect trapping experiment, although a power analysis suggests that if the sample size was doubled, a difference would be detectible.