Oxythermal Habitat Characterization and Predictive Modeling of Upper Midwestern Glacial Lakes

Cold-water fish species in the Upper Midwest are culturally, economically, and ecologically important; however, they are especially sensitive to climate warming and variability. Threats to habitat conditions and availability for cold-water fish intensify as glacial lakes become warmer and nutrient loading increases. Summertime warm water temperatures and hypolimnion hypoxia cause fish to be “squeezed” into a thin layer of oxythermal habitat during lake stratification, causing physiological stress and possible death_. Considerable variation in sufficient oxythermal habitat availability both between and within summers has been observed in Northern Wisconsin lakes, making it well-suited for predictive modeling and management. We leverage data from Wisconsin’s North Temperate Lakes- Long Term Ecological Research program (NTL-LTER) and the NOAA Physical Sciences Laboratory to create statistical models predicting various summertime average oxythermal stress metrics for cisco (Coregonus artedi) at a season-ahead time scale. Predictions are issued pre-summer, utilizing machine learning models conditioned on local hydrology, lake characteristics, and large-scale climate features as predictors. Results indicate skillful prediction of vertical habitat thickness; notably global climate features represent strong predictors. Responding to requests from natural resource managers, we expanded the analysis to thousands of glacial lakes across Upper Midwestern states. Our study highlights how season-ahead predictions can inform actionable resource decisions and compliment short-term and long-term adaptation strategies for cold-water fish in glacial lakes. Management strategies by the Department of Natural Resources and other lake organizations are often reacting to adverse conditions, whereas pre-summer season prediction models may uniquely offer prospects for proactive management.

Presentation Preference: Oral

Primary Presenter: Emma Blackford, University of Wisconsin- Madison (eblackford@wisc.edu)

Authors:

Emma Blackford, University of Wisconsin-Madison  (eblackford@wisc.edu)

Paul Block, University of Wisconsin-Madison  (paul.block@wisc.edu)