Shallow lakes are known to transition between two alternative stable states: 1) a clear state that typically supports abundant submerged aquatic vegetation (SAV) and provides high quality wildlife habitat and (2) a turbid state with frequent algal blooms supporting little to no SAV. Using yearly observations of 123 shallow lakes in Minnesota over three years, we developed a hidden Markov model to both classify lake states and to examine the influence of nutrients (a proxy for resilience) and fish (major disruptors) on the stability of each state.
We used chlorophyll a, SAV abundance, depth, and total phosphorus (TP) to classify lake states. To assess the stability of the clear and turbid states, we modeled state transition probabilities as a function of TP levels and fish populations (i.e., presence and abundance of planktivores, benthivores, and/or piscivores).
Results/Conclusions
Both states were highly stable in accordance with theory. Clear lakes were more likely to transition to the turbid state as TP levels increased and when benthivores were present. Clear lakes were also more likely to transition to the turbid state as planktivore and/or benthivore abundance increased between years. In the other direction, turbid lakes were more likely to transition to the clear state as planktivore abundance decreased between years, but changes in benthivore abundance and TP were not predictive of transitions from the turbid to clear state. Piscivore presence and abundance were not informative for predicting transitions in either direction. We conclude that TP, benthivore presence, and changes in planktivore and benthivore abundance can be used to help predict shallow lake state transitions. Our results can be applied in a management context to assess the vulnerability of lakes to undesirable state shifts and prioritize lakes for management actions.