Using knowledge of landscapes to guide biological monitoring

Landscape scale data were used to develop a framework to designate sampling sites for long-term biomonitoring in Wisconsin using models that classify streams based on thermal and flow regimes and models characterizing anthropogenic disturbance. Long-term data are essential to quantify temporal change; site selection is critical to understanding change. Two groups of sites were selected; one represented wadeable streams with relatively low levels of human impact; these “least-impacted” reference streams were stratified across four major ecoregions. The second group represented randomly selected sites, not screened for disturbance, distributed across 24 watershed management units; both groups were stratified across a stream temperature-flow classification or “natural community”. Approximately 36,000 stream segments were classified into eight natural communities that represent cold, cool-cold transitional, cool-warm transitional, and warm water systems in both wadable headwater and mainstem river systems. “Least-impacted” streams were identified by developing a human disturbance model using a geographic information system. Field samples were collected for fish, invertebrates, habitat, nutrients, and suspended sediment during 2007 – 2009 at “least-impacted” sites; field sampling is planned for randomly selected sites during 2010. Data from both groups will be used by managers to establish benchmarks, evaluate biotic integrity and quality, verify models, and evaluate trends.