• bioassessment;
  • biological traits;
  • indicators;
  • predictive models;
  • stream ecosystems


1. The use of species traits to interpret biological changes in invertebrate assemblages across environmental gradients has been suggested as a method to improve discrimination over existing species composition methods. One reason for greater potential discrimination and predictive ability is the assumed universality of traits across a range of spatial and temporal scales. We explore this assumption by comparing the consistency of stressor–response relationships of a trait characteristic (percent clinging taxa) and a common taxonomy-based metric [percent Ephemeroptera, Plecoptera and Trichoptera (EPT)] over a stressor gradient of increasing benthic fine sediment.

2. We use invertebrate assemblage and environmental data from three large-scale surveys that cover the western United States, eastern United States and the Mid-Atlantic Highlands of the US. These three datasets allow us to compare stressor–response relationships in terms of geographic position (west versus east) and spatial scale (entire east versus a sub-region of the east). We compare the slopes and intercepts of the two measures of assemblage response.

3. Trait characteristics exhibited more consistent stressor–response relationships than identity characteristics. Stressor–response relationships generated for clinging invertebrate richness had statistically similar slopes over sediment gradients, regardless of spatial scale or geographic location. In contrast, slopes were significantly different for relationships generated with EPT richness over sediment gradients.

4. Results of this study support the hypothesis that trait-based measures have a more consistent response to a stressor gradient than identity-based assemblage measures. Choosing consistent measures of community response will facilitate comparisons among assemblages across large spatial scales.