Andreas Floren, Karl Eduard Linsenmair, Tobias Müller

• Although much is known about the ecology and functional importance of canopy arthropods in temperate forests, few studies have tried to assess the overall diversity and investigate the composition and dynamics of tree-specific communities. This has impeded a deeper understanding of the functioning of forests, and of how to maintain system services. Here, we present the first comprehensive data of whole arthropod communities, collected by insecticidal knockdown (fogging) from 1159 trees in 18 study areas in Central Europe during the last 25Although much is known about the ecology and functional importance of canopy arthropods in temperate forests, few studies have tried to assess the overall diversity and investigate the composition and dynamics of tree-specific communities. This has impeded a deeper understanding of the functioning of forests, and of how to maintain system services. Here, we present the first comprehensive data of whole arthropod communities, collected by insecticidal knockdown (fogging) from 1159 trees in 18 study areas in Central Europe during the last 25 years. The data includes 3,253,591 arthropods from 32 taxa (order, suborder, family) collected on 24 tree species from 18 genera. Fogging collects free-living, ectophytic arthropods in approximately the same number as they occur in the trees. To our knowledge, these are the most comprehensive data available today on the taxonomic composition of arboreal fauna. Assigning all arthropods to their feeding guild provided a proxy of their functional importance. The data showed that the canopy communities were regularly structured, with a clear dominance hierarchy comprised of eight ‘major taxa’ that represented 87% of all arthropods. Despite significant differences in the proportions of taxa on deciduous and coniferous trees, the composition of the guilds was very similar. The individual tree genera, on the other hand, showed significant differences in guild composition, especially when different study areas and years were compared, whereas tree-specific traits, such as tree height, girth in breast height or leaf cover, explained little of the overall variance. On the ordinal level, guild composition also differed significantly between managed and primary forests, with a simultaneous low within-group variability, indicating that management is a key factor determining the distribution of biodiversity and guild composition.…