Arthropod diversity is often linked to habitat and resource availability. Ants play a range of functional roles in forests (e.g. seed dispersers, omnivores, and predators), while spiders are exclusively predators. Both taxa are influenced by climate, vegetation structure, and landscape patch configuration. We tested how ant and spider diversity and community composition vary across spatial scales in eastern deciduous forests. We also examined the relative importance of vegetation, climate, landscape, and space in explaining these patterns. A four-level hierarchical design (individual tree, forest stand, site, and ecoregion) was used to sample canopy ants and spiders in southern Ohio and Indiana, USA. Ants and Spiders were collected via insecticidal canopy fogging and ground-based funnels during summer 2000. Adult spiders and ant workers were identified to species, while juvenile spiders were identified to family. To analyze diversity across hierarchical levels, we used multiplicative partitioning of species diversity, and tested the significance of each diversity component against a null distribution using an under development R package: PARTITIONR. Species composition of ants and spiders was examined across spatial scales using nested PERMANOVA. Relationships between environmental factors and arthropod diversity and community composition at the stand level were tested using mixed models and dbRDA.
Results/Conclusions
A total of 23 species of ants and 94 species in 23 families of spiders were collected during the study. Ant richness was lower than expected at the tree level, but ant beta diversity was greater than expected at the tree and forest stand levels. Similarly, spider richness was less than expected at the tree level, but spider beta diversity was greater than expected at the tree, forest stand, and site levels. PERMANOVA indicated ant species composition divergence among stands, sites, and ecoregions, but the highest dissimilarity occurred at the tree and stand levels. Conversely, spider species composition was only different at the stand and site levels, but with similarly high dissimilarity at the tree and stand levels. Mixed models and dbRDA indicated climate and forest patch configuration were important predictors of ant diversity, while climatic variables solely explained patterns of spider diversity. Local scales were the strongest drivers of ant and spider diversity and species composition, indicating dispersal ability and local habitat characteristics (e.g. resource and microsite availability) may limit canopy arthropod diversity. Relationships with environmental factors indicate differing effects of habitat configuration on diversity of ants and spiders, but similar climatic influence on diversity of both taxa.