Use of single versus multiple biotic communities as indicators of biological integrity in northern prairie wetlands
Matthew J. Wilson and Suzanne E. Bailey
Ecological Indicators 20:187-195; 2012
As much as 70% of prairie wetlands in Canada have been lost. Although further degradation of natural wetlands is considered to be somewhat offset by wetland construction and restoration, Canada lacks bioassessment tools that can track ecosystem health in prairie wetlands. Indices of biological integrity (IBIs) use one or more biotic communities to compare the biological condition of a particular site to conditions found in least-impacted reference sites. Using the IBI approach, we evaluated the potential of 5 biotic communities to assess wetland health in northern prairie wetlands in Canada. Vegetation in the wet meadow, emergent and open-water zones as well as wetland-dependent songbirds and waterbirds were sampled at 81 semi-permanent/permanent natural and compensation wetlands spanning an environmental stress gradient. Metrics with strong linear relationships to the stress gradient (R2 > 0.2) were combined into an IBI for each biotic community and were subsequently validated at a suite of test sites. After validation, the entire data set was combined and each IBI was evaluated based on its linear relationship to environmental stress. Wet meadow zone vegetation was a strong indicator of environmental stress (R2 = 0.68, p < 0.001), as was the wetland-dependent songbird community (R2 = 0.59, p < 0.001). The emergent zone vegetation community was a relatively weak and inconsistent indicator of environmental stress, while the open-water zone vegetation and waterbird communities were poor indicators. To evaluate whether monitoring more than one biotic community provided additional information about a site's biological and environmental condition, we produced a two-taxon IBI that combined wet meadow zone vegetation and wetland-dependent songbird metrics. The two-taxon IBI had a marginally stronger linear relationship to the stress gradient (R2 = 0.72, p < 0.001) than any single biotic community alone, although we argue that this added information would not warrant the extra cost, effort, and logistical barriers of sampling both plants and birds. The wet meadow zone vegetation and wetland-dependent songbird IBIs were strong surrogates of one another (R2 = 0.57), suggesting that wet meadow zone vegetation can be used to predict the health of wetland-dependent songbirds, and visa versa. Our results suggest that habitat for healthy wet meadow zone vegetation and wetland-dependent songbird communities is being degraded as compensation sites are replacing their natural analogs.