Predictive Classification Modeling of Freshwaters

Project Lead: Patricia A. Soranno
Collaborators: Mary T. Bremigan, Kendra Spence Cheruvelil, Craig A. Stow, Tyler Wagner, Katherine E. Webster
Graduate students and post-docs: C. Emi Fergus, Brett Alger

Governmental entities are responsible for managing and conserving large numbers of lake, river, and wetland ecosystems that only rarely can be addressed on a case-by-case basis. We are developing a predictive classification modeling approach that is grounded in the theoretical foundation of landscape limnology and that creates a tractable number of ecosystem classes to tailor management actions to. We have demonstrated our system by applying two types of predictive classification modeling approaches to developing nutrient criteria for eutrophication management in 1,998 North Temperate lakes as described in Soranno et al. (2010). Our predictive classification system promotes the effective management of multiple ecosystems across broad geographic scales by explicitly connecting management and conservation goals to the classification modeling approach, considering multiple spatial scales as drivers of ecosystem dynamics, and acknowledging the hierarchical structure of freshwater ecosystems. Such a system is critical for adaptive management of complex mosaics of freshwater ecosystems and for balancing competing needs for ecosystem services in a changing world.

Funding sources: US Environmental Protection Agency; Michigan Department of Natural Resources; MSU-Center for Water Sciences; US Geological Survey



Michigan State University