Main content area

Probabilistically assessing the role of nutrient loading in harmful algal bloom formation in western Lake Erie

Bertani, Isabella, Obenour, Daniel R., Steger, Cara E., Stow, Craig A., Gronewold, Andrew D., Scavia, Donald
Journal of Great Lakes research 2016 v.42 no.6 pp. 1184-1192
algal blooms, lakes, models, phosphorus, pollution load, prediction, rivers, spring, uncertainty, water quality, Lake Erie
Harmful algal blooms (HABs) have increased in frequency and magnitude in western Lake Erie and spring phosphorus (P) load was shown to be a key driver of bloom intensity. A recently developed Bayesian hierarchical model that predicts peak bloom size as a function of Maumee River phosphorus load suggested an apparent increased susceptibility of the lake to HABs. We applied that model to develop load–response curves to inform revision of Lake Erie phosphorus load targets under the 2012 Great Lakes Water Quality Agreement. In this application, the model was modified to estimate the fraction of the particulate P (PP) load that becomes bioavailable, and it was recalibrated with additional bloom observations. Although the uncertainty surrounding the estimate of the bioavailable PP fraction is large, inclusion in the model improves prediction of bloom variability compared to dissolved reactive P (DRP) alone. The ability to characterize model and measurement uncertainty through hierarchical modeling allowed us to show that inconsistencies in bloom measurement represent a considerable portion of the overall uncertainty associated with load–response curves. The updated calibration also lends support to the system's apparent enhanced susceptibility to blooms. The temporal trend estimated by the model results in an upward shift of the load–response curve over time such that a larger load reduction is required to achieve a target bloom size today compared to earlier years. More research is needed to further test the hypothesis of a shift in the lake's response to stressors over time and, if confirmed, to explore underlying mechanisms.