The biodiversity of arthropods is often linked to the availability of different niches and their ability to disperse. Spiders are known to be long-distance dispersers that discriminate based on microsite variation provided by the surrounding vegetative structure. We wanted to know at what spatial scale spider community composition and diversity differs in an eastern deciduous forest. Furthermore, we wanted to determine if these trends varied based on tree species and habitat type (xeric and mesic). A four-level hierarchical design (individual tree, forest stand, site, and ecoregion) was used to sample canopy spiders in southern Ohio and Indiana, USA. Spiders were collected via a combination of insecticidal canopy fogging and ground-based funnels during June and August 2000. All spiders were identified to family, while all adults were identified to species. To analyze diversity across hierarchical levels, we used additive and multiplicative partitioning methods based on the PARTITION software developed by Crist et al. 2003 via an under development R package: PARTITIONR. We tested the significance of each level’s alpha and beta against a null, random distribution. Community overlap for each scale and for tree species and habitat was determined using PERMANOVA and NMDS.
Results/Conclusions
Collected spiders belonged to 94 species in 23 families. Spider family diversity was greater than expected at the forest stand level using both additive and multiplicative methods. Species diversity was also greater than expected for site and ecoregion using multiplicative partitioning while only ecoregion was greater than expected using additive. Individual tree and forest stand levels had fewer spider species than expected. As indicated by ordination results suggest that forest stands drive spider species diversity. Ordination results revealed that most tree species overlapped in the spiders observed with select tree species having more spider species; PERMANOVA on tree species with spider species also indicated significant differences. When habitats were compared, we saw a similar trend with all spider species observed in mesic habitats being nested within the xeric habitat. We suspect that the differences in spider communities among tree species and habitats are directed by changes in spider abundance in addition to species richness. As hypothesized, spider species are more associated with certain spatial scales possibly due to dispersal ability of various spider functional guilds and resources provided by unique microsites of various tree species. For future spider collection and conservation, we recommend that multiple sites across ecoregions are considered.