USING THE THEORY OF ALTERNATE STATES IN ECOSYSTEMS TO PREDICT HARMFUL ALGAL BLOOMS IN THREE OHIO GLACIAL LAKES

The theory of alternate states of ecosystem predicts that ecosystems can exist in multiple configurations of biotic and abiotic conditions. Perturbations can cause ecosystems to shift from one state to another: in our example, no bloom versus algal bloom. This theory also indicates that statistical properties of state variables will change prior to a change in state (e.g. variance or autocorrelation of the time series). The Quickest Detection (QD) method can be used to detect this change in variance and provide an alarm to indicate that a state change may soon occur. We explored whether the QD method could be applied in three lakes to predict the onset of harmful algal blooms in any of those lakes. The state variables we examined were Chl. a and phycocyanin which were measured daily in East Twin, West Twin and Sandy Lakes (Ohio) from late May to mid-September, 2023. The QD method requires a baseline condition, and we examined each lake’s feasibility to serve as a baseline compared to the other two. East Twin Lake had the lowest concentration and variance of Chl. a and phycocyanin of the three lakes and did not exhibit an algal bloom. West Twin Lake developed two algal blooms and the quickest detection method generated alarms before the peak of each bloom when East Twin Lake was used as a baseline. Sandy Lake had higher concentrations and variance of Chl. a and phycocyanin than either of the Twin Lakes. Sandy Lake had numerous alarms throughout the summer that did not seem to correspond to changes in the algal biomass.

Primary Presenter: Tamar Atwell, Kent State University (tatwell1@kent.edu)

Authors:

Tamar Atwell, Kent State University  (tatwell1@kent.edu)

Darren Bade, Kent State University  (dbade@kent.edu)