Inducibility of chemical defences in young oak trees is stronger in species with high elevational ranges

Elevational gradients have been highly useful for understanding the underlying forces driving variation in plant traits and plant-insect herbivore interactions. A widely held view from these studies has been that greater herbivory under warmer and less variable climatic conditions found at low elevations has resulted in stronger herbivore selection on plant defences. However, this prediction has been called into question by conflicting empirical evidence, which could be explained by a number of causes such as an incomplete assessment of defensive strategies (ignoring other axes of defence such as defence inducibility) or unaccounted variation in abiotic factors along elevational clines. We conducted a greenhouse experiment testing for inter-specific variation in constitutive leaf chemical defences (phenolic compounds) and their inducibility in response to feeding by gypsy moth larvae (Lymantria dispar L., Lepidoptera) using saplings of 18 oak (Quercus, Fagaceae) species. These species vary in their elevational distribution and together span >2400 m in elevation, therefore allowing us to test for among-species elevational clines in defences based on the elevational range of each species. In addition, we further tested for elevational gradients in the correlated expression of constitutive defences and their inducibility and for associations between defences and climatic factors potentially underlying elevational gradients in defences. Our results showed that oak species with high elevational ranges exhibited a greater inducibility of phenolic compounds (hydrolysable tannins), but this gradient was not accounted for by climatic predictors. In contrast, constitutive defences and the correlated expression of constitutive phenolics and their inducibility did not exhibit elevational clines. Overall, this study builds towards a more robust and integrative understanding of how multivariate plant defensive phenotypes vary along ecological gradients and their underlying abiotic drivers.