Insect diversity: facts, fiction and speculation*

Abstract

Biologists are still trying to grasp the global dimensions of the phylum Arthropoda and its major class the Insecta, in spite of the fact that over a million species of arthropods have been described. The canopy of rain forest trees is believed by many to hold the key to the immense diversity of insects. In recent years the use of knock-down insecticides to sample insects from rain forest canopy has revealed information on the canopy's arthropod inhabitants and community structure. The sampling techniques involved are outlined and data reviewed on taxonomic and guild structure, species abundance, body size and biomass of insects, and the faunal similarity of trees. Calculations by Erwin (1982), based on knock-down insecticide studies of the beetle fauna of one species of Central American tree, suggest there may be 30 million species of tropical forest arthropods. Reanalysis of these calculations, using additional data, produces a range of possible estimates from about 10 to 80 million. The unknown range of plant host-specificities of tropical insects is the main weakness of this method of calculation. Assessment of the faunal importance of the canopy in relation to that of other rain forest biotopes requires comparative quantitative studies. The preliminary results of one such simple study suggest that over 42 million arthropods may be found in a hectare of Seram rain forest (at the time of study), and that 70% occur in the soil and leaf litter and 14% in the canopy. They also suggest that Collembola and Acarina are the dominant groups in this hectare, and that there are as many ants as all the other insects (excluding Collembola).