Table 4-16. Use of aquatic ecoregions to evaluate regional consistency in interpretation of 

 indicator 



Aquatic ecoregions (e.g., Omernik 1987, Omernik and Griffith 1991, Omernik 1995) can serve 

 as a regional framework to classify stream ecosystems in a target resource population 



• Based on overall similarity in several natural features (e.g., climate, soils, 

 vegetation, physiography, land use). 



Ecoregions correspond to spatial patterns in fish assemblages and abiotic characteristics of 

 streams (e.g., Pflieger 1975, Larsen etal. 1986, Rohm etal. 1987. Whittier etal. 1988, Hughes 

 and Larsen 1988, Jenkins and Burkhead 1993). 



Ecoregions have been shown to be useful in improving the consistency of interpretation of other 

 multimetric indicators applied over large geographic scales (e.g., Yoderand Rankin 1995, 

 Barbour ef a/. 1996). 



Ecoregions serve as a basis to account for natural differences in potential biotic integrity under 

 minimal human disturbance. 



• Can be used to define different expectations for individual metrics, or different 

 thresholds for indicator value (e.g., Yoder and Rankin 1995). 



• Metric-based adjustment is more suitable for EMAP indicators because of focus 

 on regional resource population estimates. 



Two examples (Fig. 4-7) are provided to demonstrate an evaluation of differences in fish assemblage 

 characteristics across the region of interest. The distributions of metric response variables across two levels 

 of aquatic ecoregion aggregations are examined using box-and-whisker plots. Regions showing restricted or 

 expanded distributions in comparison to others should be considered for possible adjustment in metric 

 expectations. For both examples (Figure 4-7 (A), number of water column species; and (B) proportion of 

 individuals of tolerant species), examination of the boxplots suggests that no substantial differences exist in 

 the range or general distribution of response values across the two levels of ecoregion aggregation. For 

 these two metrics, adjustments of expectations do not appear to be necessary. Similar analyses applied to 

 other candidate metrics have provided similar results, and at present, the indicator is being developed without 

 normalization of component metrics. 



Summary 



Aquatic ecoregions, evaluated in the context of historical zoogeography affecting fish distributions, can be 

 used to assess the natural variation in metric responses. These results can be used to adjust the expectations 

 for individual metrics. Preliminary examination suggests that normalization is not necessary for the component 

 metrics or the indicator, but additional analyses, performed in conjunction with assessing the responsiveness 

 of the indicator (Guideline 12), are needed. 



4-31 



