Is insect vertical distribution in rainforests better explained by distance from the canopy top or distance from the ground

The way arthropods are distributed vertically in tropical forests has been of great interest with diversity often greatest at or near the canopy top. Typically, stratification is measured up from the ground but, since the height of trees reaching the canopy top can vary, we hypothesise that distance down from the canopy top, might better explain arthropod distributions. To test this samples were collected from Australian tropical rainforest trees in both dry and wet seasons by beating foliage from five trees for each of 11 tree species at set intervals down each tree. A total of 2628 arthropods were collected. Abundant groups were Araneae, Coleoptera, Formicidae, Blattodea and Homoptera. Coleoptera were sorted to species. Since the forest was naturally disturbed by storms, height of trees reaching the canopy top ranged 10–40 m. Our results suggested that the best fit for vertical stratification, either distance from ground or distance down from the canopy, were taxon specific. For ordinal richness and abundance of arthropods the best model was distance from the ground with decreasing trends from the ground. Similarly, distance from the ground best fitted abundances of spiders, cockroaches and Homoptera. In contrast, declination from the canopy top best fitted beetle species richness and abundance, and ant abundance. The effect of vertical stratification was, however, significant only for ants in dry season: abundance of ants decreased with increasing distance down from the canopy top. We were surprised to have found taxon-specific patterns, which may be explained by highly variable canopy tree height, creating vertically heterogeneous micro-habitat conditions in this forest system.