PARENT SESSION
Oral Session #19: Animal Community Ecology: Predation, community structure.
Presiding: G. Cumming
Monday, August 5. 1:00 PM to 3:45 PM. Cochise Meeting Room, TCC.

Thresholds in community composition: avoiding the presumptions of a priori classification.

Bossenbroek, Jonathan^*,1, Wagner, Helene², Wiens, John³, ¹ Colorado State University, Fort Collins, CO² WSL - Swiss Federal Institute for Forest, Snow and Landscape Research, Birmensdorf, Switzerland³ The Nature Conservancy, Arlington, VA

ABSTRACT- Traditionally, animal community composition is analyzed within the context of a classification approach, comparing populations within different vegetation classes, disturbance regimes, or some other factorial design. This approach assumes that community patterns coincide with predefined classifications, and violations of this assumption may lead to erroneous patterns or failure to detect pattern in community composition. To avoid these problems, we suggest an analysis based on gradient analysis that does not presuppose categories but emphasizes differences in the rates of species turnover for different groups across geographic space. To examine turnover rates of beetles and vascular plants, we sampled a 4 km transect in eastern Colorado. Beetle pit-fall traps and 1-m² vegetation quadrats were placed every 50 m along the transect, as well as within two intensely sampled zones. Each of these zones was 250 m long, consisting of two transects that paralleled the main transect that were sampled every 10 m. A Detrended Correspondence Analysis (DCA) was conducted separately for the beetle and plant data, and the scores of the first DCA axis were plotted against sample location along the transect. Turnover rates were measured as the first derivative of the DCA axis profile. The highest rates of turnover for beetles and for plants were typically in different locations along the transect, indicating that vegetation and beetle communities are responding to different aspects of the landscape. These regions of high turnover rates identify potential boundaries that help to define classifications based on either the vegetation or the beetle communities. If the analysis of beetle communities had been constrained by a vegetation classification, true boundaries in beetle composition would have remained undetected. We conclude that studies on animal community composition should not be based on an a priori mapping of vegetation unless the mapped boundaries in plant species composition are known to coincide with thresholds in animal community composition.

KEY WORDS: Thresholds, Gradients, Community composition, Classification