COS 100-4
Habitat selection as a mechanism structuring larval odonate communities across a canopy cover gradient

Thursday, August 13, 2015: 9:00 AM
320, Baltimore Convention Center
Sarah K. French, Biology, University of Toronto Mississauga, Mississauga, ON, Canada
Shannon J. McCauley, Biology, University of Toronto Mississauga, Mississauga, ON, Canada

Habitat selection can shape species’ distributions, acting as a filter on the colonization of potential habitats before many other mechanisms of community assembly can act (e.g., differential performance in a habitat). In heterogeneous habitats where species can choose between habitats of differing quality, habitat selection is difficult to distinguish from species-sorting. The level of canopy cover (i.e., pond shading) can shape community structure in aquatic habitats, and habitat selection may drive this pattern. For example, adult dragonflies detect ponds based on polarized light, which decreases with increased canopy cover, potentially reducing their colonization of suitable habitat. We assessed the relative contributions of habitat selection and species-sorting in determining species’ distributions, by testing whether canopy cover affects adult dragonfly habitat selection and consequently larval distributions. Three field experiments tested whether canopy cover reduced the abundance of adult and/or larval dragonflies at artificial water bodies. Firstly, we manipulated temperature and artificial canopy cover in a full factorial design. Secondly, mesocosms were placed under a gradient of natural canopy cover. Mesocosms were observed for adult dragonfly visits and sampled for larvae. Finally, larvae were caged in mesocosms that varied in canopy cover, and mortality was measured to test for species-sorting.


In all experiments, canopy cover affected the number of visits to mesocosms by adult dragonflies. Artificial canopy cover with 50% shade cloth marginally reduced adult dragonfly visitation rates (p = 0.07). Under a gradient of natural canopy cover, the number of adult visits to mesocosms decreased significantly with an increase in canopy cover (p < 0.001). No relationship was found between larval abundance and artificial (p = 0.8) or natural (p = 0.6) canopy cover, and there was no differential mortality of larvae across a natural canopy cover gradient. In addition, temperature had no effect on the number of adult visits (p > 0.8) or larval abundances (p > 0.1) in either experiment. Adult dragonflies showed strong habitat selection responses to canopy cover, however canopy cover did not affect larval abundance or survival in our experimental tanks. At present, widespread forest regrowth in parts of eastern North America is modifying terrestrial and aquatic environments. Dragonflies play an important role as predators in aquatic systems and can shape aquatic communities. Thus, determining their response to the presence of canopy cover can be broadly applied to understanding the impacts of forest regrowth on aquatic community structure.