Sources of variability in plant resistance against insects: free caterpillars show strongest effects

Tests with herbivorous animals have shown great variability in foliage quality within individual plants. This may be a product of plant architecture and microclimate as well as a specific response to herbivory. We tested the level of constitutive and herbivore‐induced resistance within the canopy of mountain birch ( Betula pubescens ssp. czerepanovii ) stems of different age, using shoots with different architectural position, function and age. We tested foliage both with uncaged larvae of Epirrita autumnata feeding freely on foliage and with growth trials in the laboratory. The only indication of systematic within‐tree variability was low consumption by third instar larvae growing on long shoot leaves. This may protect long shoots that would suffer from reduced growth if their first leaves were consumed at an early stage of development. Stem age and architecture did not contribute to within‐tree variation and the effects induced by herbivory were similar throughout the tree. A small number of caterpillars was sufficient to elicit responses in larval growth or consumption. This suggests that the presence of caterpillars may act as a cue of increased risk of future herbivory for mountain birch attacked by E. autumnata with poor mobility and regular outbreaks. Previous defoliation and young age of stems increased the disappearance rate of free larvae. The effect of defoliation suggests that induced resistance may act to prevent future buildup of E. autumnata populations. The small size and simple canopy architecture of young stems may in turn increase the disappearance rate by increasing the probability of predators finding prey. This may partly explain the low vulnerability of young mountain birch forests to E. autumnata outbreaks. Overall, changes in the disappearance rate were greater than changes in larval growth and consumption. This suggests that tests in the laboratory or inside cages may underestimate the efficacy of induced resistance against herbivores.