Short-term effects of canopy openness on insect herbivores in a rain forest in Guyana

1. A before/after-control/impact experiment (BACI) was used to examine the shortterm effects of canopy openness on insect herbivores foraging on the seedlings of five rain forest species in an unlogged forest in Guyana. During the first and second study years, insects were collected monthly from 250 sites, representing 9750 seedlings spread over a plot of 1 km(2). At the onset of the second collecting year, half of the parent trees at the sites were felled, creating an average canopy openness of 26%. 2. Comparisons of the control and impact sites before and after the felling showed that overall insect abundance, particularly of psyllids, and species richness increased at the impact sites, whereas evenness decreased markedly. These responses were consistent with the prediction that intermediate disturbance promotes high diversity in tropical systems. 3. Twenty-nine per cent of insect species analysed showed a significant response to felling, with some populations doubling or halving at the impact sites after felling. The abundance of most species increased, although responses varied even between congeners. 4. A canonical correspondence analysis at the impact sites after felling indicated that most species responded to an increase in leaf production by seedlings, not to canopy openness per se. Species declining at the impact sites were more likely to feed indiscriminately on young and mature foliage and probably reacted to changes in microclimate. 5. The study illustrates the difficulty in identifying higher taxa, umbrella or indicator species that adequately integrate the responses of insect herbivores to forest disturbance. Because the integrity of insect communities in the understorey of this forest depends more on minimizing damage to plant resources than on canopy openness, we found some support for the conservation value of reduced impact logging in tropical wet forests.