Exploring relationships between in-stream conditions and ecological health while assessing landuse and climate scenarios

Human disturbances can have significant impacts on physicochemical and biological conditions of streams. A good understanding of the relationships among these factors will help decision makers in sustainable management of the ecosystems. To address these issues, the following research objectives were developed: 1) bridge the gap between hydrologic models and ecological conditions using the Soil and Water Assessment Tool, 2) identify influential in-stream variables to explain fish and macroinvertebrate measures, 3) compare fuzzy logic techniques with statistical approaches to describe and model ecological health, 4) use in-stream variables obtained from SWAT to predict the impacts of different landuse and climate scenarios, and evaluate the effectiveness of best management practices, in regards to aquatic health. A high resolution SWAT model was built for the Saginaw River basin of Michigan, and flow and water quality outputs were linked with measured biological data. Results indicate that SWAT models can be an effective tool to produce in-stream variables, explaining 21% to 57% of variation (R2) in ecological measures. Fuzzy logic methods are effective approach to model ecological health and outperformed other statistical methods tested here. Average annual flow rate had the strongest correlation with IBI, whereas nutrient concentrations showed the largest influence on all other ecological measures. Results suggest that efforts to model historic baseline conditions and to provide context for stream health assessments should include both pre-settlement land use and climate conditions. Meanwhile, the conservation practice, native grass, showed the most improvement to stream health, followed by residue management and no-tillage.