Rapidly Induced Defenses and Talking Trees: The Devil's Advocate Position

We review the evidence that rapidly induced chemical changes in damaged leaves, for example, after attack by insects, (1) significantly influence the population dynamics of herbivorous insects and mites in the field, and (2) significantly reduce damage to plants by such herbivores. We also review the evidence for "talking trees"; that is, following experimental damage to tree foliage, chemical changes are induced in adjacent undamaged trees that mirror those in damaged trees. Most of the evidence for the effects of rapidly induced changes in damaged plants on herbivorous insects and mites is laboratory based, and often reveals only small effects (< 10%) on such things as larval development time or pupal weights. Many of the reported larger effects are statistically flawed. We argue that most induced responses may have negligible consequences for field populations. There is only one field study demonstrating significant effects of rapidly induced defenses on insect population dynamics. Evidence that foliage damage is overdispersed on tree leaves, interpreted as possibly caused by induced defenses, is equivocal and subject to several plausible alternative explanations. Two experiments (Baldwin and Schultz 1983; Rhoades 1983) purport to show intertree communication after experimental damage. One of these studies is statistically flawed (Baldwin and Schultz 1983); results from Rhoades' study may be due to an infectious disease transmitted between caterpillars and not to intertree communication. Finally, we report on a field experiment in which we look for reductions in herbivore damage levels in birch (Betula pubescens) saplings following experimental defoliations of 5% and 25%. We also look for evidence of intertree communication. We found no evidence either that experimental defoliation reduces subsequent levels of herbivore attack or that trees communicate.