Nutrient loading and nonstationarity: The importance of differentiating the independent effects of tributary flow and nutrient concentration

The “phosphorus loading concept,” or more generally the “nutrient loading concept,” arose from Richard Vollenweider’s work in the 1960–1970s that showed correlations between phosphorus loads and various eutrophication symptoms. The initial success of target loads developed for the Great Lakes solidified the concept that nutrient loading causes eutrophication, and load targets have become common tools to reduce eutrophication. Using concepts from the field of causality, we offer additional context to the nutrient loading concept to show that the correlation between nutrient load and eutrophication is spurious; load and eutrophication have common drivers, tributary flow and tributary nutrient concentration, but load itself is not causal. Consequently, in-lake conditions are not invariant to the same load delivered at differing flow-concentration combinations. We then use a simulation model to evaluate the consequences of delivering the same load at various flow-concentration combinations from the Maumee River into Lake Erie. We show that load reductions under increased tributary flows may cause in-lake phosphorus concentration increases, potentially offsetting the anticipated effect of the load reduction. Thus, particularly under a scenario where climate change may cause systematic flow changes, it will be important to expand the nutrient loading concept to consider the independent effects of tributary flow and nutrient concentrations, to assess the effectiveness of nutrient reduction strategies.