Determinants of adult odonate community structure at several spatial scales: effects of habitat type and landscape context

Abstract

Dragonflies (Insecta: Odonata) use both aquatic and terrestrial ecosystems, and the abundance and diversity of odonates should be good indicators of habitat integrity. To determine which environmental variables affect odonates, we sampled adult dragonflies three times at 12 sites in Pickens and Greenville Counties, SC, USA, in different habitats, at different spatial scales, across a landscape gradient from intact forest to urban locations. At each site, we established two 2 m × 20 m plots along the shoreline of each aquatic habitat. We sampled dragonflies in ten 2 m × 2 m subplots/plot, described the vegetation and substrate in these subplots and adjacent aquatic subplots, and measured the percent cover of different landforms within 500 m of each plot center. Using nested ANOVA and Akaike information criteria models, habitat type and correlating environmental variables (substrate type and bank vegetation) were the best predictors of community structure at all spatial scales. Streams and rivers had fewer individuals and species than lakes, and had a nested subset of species found in lake communities. Landscape elements were also important, with indices declining as barren land and grasslands increased. At the largest scale, anthropogenic changes to the landscape had mixed effects. Small habitats isolated in urban areas had a significantly depauperate, nested subset of species found in communities inhabiting larger natural areas. However, odonate abundance and diversity was highest at human-made lakes and ponds, suggesting that these anthropogenic features help maintain odonate communities.

Keywords:

• dragonfly
• community ecology
• nestedness
• anthropogenic effects
• Odonata

Acknowledgements

We thank Dr John Quinn for his statistical help with Akaike modeling. We thank the South Carolina Department of Parks Recreation and Tourism, the South Carolina Department of Natural Resources Heritage Trust Division, Greenville County Parks and Recreation, Lake Conestee Nature Preserve, The Nature Conservancy, and the City of Greenville for allowing us to sample at these sites. This research was supported by a grant from The Furman Advantage to MGC and a Furman Standard grant to WBW.