Model assumptions limit implications for nitrogen and phosphorus management

A recent Science paper concluding “phosphorus load reduction will make Erie more toxic” is misleading because it ignores historical lake behavior, and is based on a cellular-level model that excludes lake-level processes which influence phytoplankton community composition and toxicity. The inferences made from models are inherently limited by the assumptions underlying those models. We identify major assumptions, several of which are implicit, that were involved in extrapolating this cellular-level model to make lake-level inferences. An active adaptive management program will evaluate future nutrient load modifications if emerging evidence from well-vetted data and models indicate updates are appropriate.