Session Lead: Rosemary Fanelli (U.S. Geological Survey)
Co-Lead(s): Kelly Maloney
Session Format: Oral presentations
Session Description:
Maintaining and improving the health of freshwater stream ecosystems remains a major goal of the Chesapeake Bay restoration effort. Landscape and regional drivers, such as urban growth, increased agricultural intensification, and shifting precipitation patterns, have altered the health and biological integrity of streams and rivers in Chesapeake Bay watershed. These landscape drivers often trigger changes in abiotic conditions in and around stream channels, including modified streamflow, degraded habitat and water quality, and increased occurrence of contaminants. These stressors can reduce the abundance, diversity, and functioning of living resources within freshwater stream ecosystems.
Local and regional managers have been tasked with implementing restoration and conservation practices to protect or restore stream ecosystems within the watershed. Knowing where stressors are absent/minimal (for conservation) or elevated (for restoration) would help guide the placement of these practices. Local and regional monitoring programs have been established to collect information about stream stressors and stream health, but spatial and temporal gaps in the networks often exist, which limits their use. However, when harmonized across multiple jurisdictions, these data can be coupled with emerging AI tools, like machine learning, to provide a more holistic picture of stream ecosystem conditions and responses to management practices. A key aspect to these models is their ability to predict conditions in unmonitored areas, which can help guide watershed-wide conservation and restoration implementation efforts needed to improve future stream health conditions.
This session will showcase studies that leverage local and regional monitoring data to better characterize, track, and predict changes in freshwater stream ecosystems. For example, the USGS has been leveraging such monitoring networks to develop predictive models for estimating benthic macroinvertebrate and fish community composition, salinity, and physical habitat across all stream reaches within the watershed. New projects are focused on estimating stream temperature and visualizing these predictions so that managers can access the results and dynamically incorporate them into their workflows. We welcome additional examples of predictive modeling, data visualization, and data analysis that leverages local and regional data collection efforts to better understand freshwater stream ecosystems and the effects of management practices.