Abstract Themango gall fly (Procontarinia) is a growing concern in many countries that importmango trees.Mangoes flush (produce opening buds) twice a year: once in spring, and the main flush, which occurs in early autumn flush and corresponds to the photoperiod necessary for the gall fly to emerge. Gall flies attack only very young mango flush leaves. Three cultivars displaying varying susceptibility were used in this study: Heidi, which typically is heavily infected; Keitt, which displays antixenosis properties; and Sensation, which is resistant to gall fly attack. Headspace profiles of the three cultivars were obtained over two full growing seasons. It was found that flush leaves emit higher levels of volatile organic compounds than mature leaves. Volatile emissions peaked during the main flush, after which they declined as the tree went into the winter dormant period. The levels of a- and b-pinene increased in gall fly-susceptible cultivars as the season progressed, while very low levels of these terpenes were emitted by resistant cultivars. The changes in the profiles of these emissions suggest that aand b-pinene may be linked to gall fly susceptibility. In addition, chemometric analysis confirmed the correlation between volatile emissions and gall fly susceptibility.

Abstract Mango gall fly (Procontarinia matteiana) is an orchard pest that parasitises flush leaves of mango and serious outbreaks may result in reduced fruit yield. The trigger for infestation is unknown, but terpenes emitted by the leaves appear to play a role in attraction. Metabolic profiles of three mango cultivars of varying susceptibility to mango gall fly attack were obtained by headspace profiling using GC-FID and GC-MS analysis. Chemometric models constructed from the data revealed that three terpenes, α- and β-pinene and camphene could be useful as biomarkers for susceptibility. Headspace profiles of twenty other cultivars, naturally exposed to gall fly, were obtained in the same way. Susceptibility or resistance of these cultivars was predicted using the developed orthogonal partial least squares model. Predictive outcomes were thereafter verified by visual examination of the leaves to detect gall formation, an indication of gall fly infestation. The model was found to predict the susceptibility or resistance of 90% of the cultivars accurately. This finding indicates the contributory role of the three terpene biomarkers in mango gall fly interaction and may direct future studies to determine their inter-relationship.