Predicting ecosystem state changes in shallow lakes using an aquatic ecosystem model: Lake Hinge, Denmark, an example

Area of Science:

• Ecological modeling
• Limnology
• Environmental science

Background:

• Shallow lakes often exist in turbid, phytoplankton-dominated or clear-water, macrophyte-dominated states.
• Dynamic lake models like PCLake predict nutrient thresholds for regime shifts, but may oversimplify vertical gradients.
• Recent studies suggest lake ecosystem changes are often gradual, challenging the traditional regime shift concept.

Purpose of the Study:

• To investigate if accounting for depth-dependent heterogeneity influences the gradualness of lake ecosystem regime shifts.
• To assess the impact of vertical gradients in water and sediment layers on lake model predictions.
• To refine predictions of external nutrient load thresholds for restoring clear-water conditions.

Main Methods:

• Coupling the 1D hydrodynamic model GOTM with the aquatic ecosystem model PCLake within the FABM framework.
• Implementing vertical gradients in water column and sediment layers for a comprehensive depth-dependent analysis.
• Calibrating and validating the GOTM-FABM-PCLake model using a 14-year dataset from Lake Hinge, Denmark, with Auto-Calibration Python (ACPy).

Main Results:

• The GOTM-FABM-PCLake model demonstrated strong agreement with observed data for temperature, nitrogen, and phosphorus species, and adequate simulation of chlorophyll a, zooplankton, and macrophyte coverage.
• Bifurcation analysis revealed that incorporating depth heterogeneity led to more gradual responses in macrophyte extent and phytoplankton chlorophyll-a to reduced phosphorus loads.
• Hysteresis effects were still observed, indicating complex dynamics in lake ecosystem state changes.

Conclusions:

• Including depth heterogeneity in lake models provides a more realistic representation of ecosystem dynamics, showing gradual rather than abrupt regime shifts.
• Accurate determination of external nutrient load thresholds for achieving clear-water states necessitates the inclusion of depth-dependent processes.
• This approach enhances the predictive power of lake management models for restoring aquatic ecosystems.