Lake management: are tipping points real?

Posted on December 14, 2021


Freshwater biologists often seek to restore turbid shallow lakes dominated by phytoplankton to clear-water systems with a high abundance of submerged plants.  The usual plan of operation, which is often guided by lake ecosystem modelling, is to reduce external inputs of nutrients (e.g., from agricultural runoff or urban waste streams) to a threshold level at which the lake should quickly flip from a turbid to a clear-water state.  However, recent studies and reviews suggest that gradual changes are actually more common than sudden shifts, which makes it desirable to review relevant assumptions and models.  Because conventional modelling treats a lake as a simplified, fully mixed water column of uniform depth, a team of environmental scientists assessed the performance of an ecosystem model with a more realistic representation of depth-dependent variations in temperature, light levels and nutrient concentrations in the water column and sediment layers.  The model was tested using data from Lake Hinge, a small Danish lake with a maximum depth of 2.6 metres, and the simulations involved dividing the water column into 16 layers, each with a sediment component.  In the model, submerged plants are assumed to promote the settlement of suspended solids, suppress phytoplankton by competing for light and nutrients, and encourage the growth of piscivorous fish so as to put top-down pressure on phytoplankton via a food chain cascade.  The modelling results agreed well with observed data for water temperature, nitrogen, phosphorus and chlorophyll concentrations.  Rather than there being a sudden “all or nothing” effect at a critical turbidity, light penetration through the water column helped plants to re-establish in the shallow zones, encouraging gradual colonisation of the deeper areas.  These findings show that conventional modelling exaggerates the suddenness of the turbid – clear transition, and that in order to accurately estimate the nutrient load at which lakes change to a clear-water condition, it’s vital to include depth-dependency in the model.  This conclusion is supported by the fact that in the case of Lake Hinge, conventional modelling indicated that a 50% reduction in phosphate loading would be required, but the new model predicted that a reduction of only 20-30% would be needed – quite a significant difference in terms of economic cost.

Reference:  Andersen, T.K. et al.  2020. Predicting ecosystem state changes in shallow lakes using an aquatic ecosystem model: Lake Hinge, Denmark, an example.  Ecological Applications 30(7),  e02160.  https://doi.org/10.1002/eap.2160