Leaf compositional differences predict variation in Hypsipyla robusta damage to Toona ciliata in field trials

Abstract

Hypsipyla robusta Moore is a shoot-boring moth that feeds on species in the Swietenioideae subfamily of Meliaceae, including the rain forest tree Toona ciliata M. Roemer. Damage from Hypsipyla has been a major barrier to growing these species in plantations. Although there has been speculation regarding the role of plant chemistry in determining host selection by Hypsipyla, there is no substantial evidence to support a role for any particular class of compounds. In this study, we used near-infrared spectroscopy (NIRS) to quantify variation in tissue composition to determine whether compositional variation could be linked with differences in H. robusta damage in a sample of 153 T. ciliata tree stems. We found that a discriminant analysis using NIRS data successfully classified most leaflets into high- and low-damage classes. Regression models based on NIRS data were also able to predict variation in leaflet nitrogen and tree height. Taller specimens of T. ciliata were more frequently damaged. Leaf nitrogen varied only a little, making it a weak explanatory variable for insect attack. The capacity of NIRS to predict variation in H. robusta attack suggests a link between T. ciliata leaf chemistry and H. robusta behaviour.