PS 46-58 - Canonical correspondence analysis as a tool for integrating species abundance and stable isotope data

Wednesday, August 5, 2009
Exhibit Hall NE & SE, Albuquerque Convention Center
Jonathan A. Freedman, Pa Coop Fish & Wildlife Research Unit, Penn State University, University Park, PA, R. Allen Curry, Biology Department, University of New Brunswick, Canada, Kelly R. Munkittrick, Department of Biology, University of New Brunswick, Saint John, NB, Canada, Robert F. Carline, PA Coop Fish & Wildlife Research Unit, Penn State University, University Park, PA and Jay R. Stauffer Jr., School of Forest Resources, Penn State University, University Park, PA
Background/Question/Methods

The structure of animal communities can be assessed using a variety of ecological indices, such as measures of diversity and evenness. Meanwhile, stable isotope analysis is increasingly being used to provide information about the environment and to answer ecological questions. One of the more common ecological applications of stable isotope analysis is to delineate food-web patterns and trophic interactions within animal communities. These studies are often used to compare different sites which may vary along environmental gradients (such as primary productivity in aquatic systems). Many studies of this type can and do benefit from analysis of both species abundance and stable isotopes, but to-date there has not been an adequate method which can incorporate both types of data into one quantitative analysis to directly assess the interactions between the data sets. Canonical correspondence analysis (CCA) is a multivariate technique which integrates species abundance data with environmental variables by combining the two data matrices along a common site axis. We used CCA to analyze species abundance data and stable isotope data of carbon and nitrogen (δ13C and δ15N as the “environmental” gradients) for two data sets. The first data set consisted of ten sites in a temperate reservoir, located upstream, downstream, and on the far-shore from a pulp mill’s effluent discharge.  The second data set consists of six sites in a large river, three of which have been altered by gravel dredging and three which acted as references. For both data sets, multidimensional scaling was used to separately analyze similarities in fish species abundance and stable isotope signatures.

Results/Conclusions

We found that while multidimensional scaling was able to adequately differentiate sites, CCA provided the ability to observe the relationships among sites, species abundance, and stable isotope data. For the first data set, this allowed us to determine the effects which differences in stable isotope signatures (indicative of exposure to pulp mill effluent and the resulting eutrophication) have on fish assemblages among sites. CCA revealed that the both the downstream and far-shore sites show clear relationships between species abundance and stable isotope signatures indicative of exposure to pulp mill effluent, which otherwise could only have been loosely inferred from the multidimensional scaling. Likewise for the second data set, CCA showed the relationship between lower species abundance and a reduced dependence on benthic nutrients at dredged versus undredged sites. CCA is therefore a valuable tool for integrating community and stable isotope data.

Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.