Evaluation of Probabilistic and Targeted Sampling Designs Used to Assess Wadeable Streams in Wisconsin
Michael A. Miller 1, Alison C. C. Colby 1, Paul D. Kanehl 2 and Karen Blocksom 3
1 Wisconsin Department of Natural Resources, Bureau of Fisheries Management, Madison, Wisconsin
2 Wisconsin Department of Natural Resources, Bureau of Science Services, Madison , Wisconsin
3 U.S. Environmental Protection Agency, ORD, Cincinnati Ohio
The Wisconsin Department of Natural Resources (WDNR), with support from the U.S. EPA’s Regional Environmental Monitoring and Assessment Program conducted an assessment of the physical, chemical, and biological conditions of wadeable streams in the Driftless Area ecoregion in western Wisconsin, using a probabilistic sampling design. For each of the study streams (n = 60), watershed land use, riparian and in-stream habitat, water chemistry, macroinvertebrate, and fish assemblage data were collected at a randomly-selected stream site and an associated “modified-random” sampling site, accessed via a road crossing nearest the random site. Least-disturbed stream reference sites (n = 23) were also sampled to develop reference conditions. Study results show no statistically significant differences between the random and modified-random assessment sites for any of the 9 physical habitat, 11 water chemistry, 7 macroinvertebrate, or 7 fish metrics analyzed. We also provide evidence that targeted sampling data routinely gathered by the WDNR indicate that overall streams are in better condition than indicated by probabilistic sampling. There were significant differences between the random sample population (and by inference the population of all streams in the Driftless Area) and the reference conditions for a number of physical, chemical, and biological measures, suggesting significant environmental degradation of the Driftless Area stream population. While further evaluation of the statistical rigor of using a modified-random sampling is warranted, sampling randomly-selected stream sites accessed via the nearest road crossing may provide a more economical way to apply probabilistic sampling designs in stream sampling programs.
Keywords: streams, probabilistic sampling, reference conditions, bioassessment