Modeling the impact of flushing on a lake meta-ecosystem using PCLakeS+

Designing effective management strategies for lake networks is challenging as nutrient dynamics within a lake can be complex and nonlinear, and potentially cause spatial cascades. The recently developed lake meta-ecosystem model PCLakeS+ can aid in finding such management strategies. Here, we present the first real-world application of PCLakeS+ on the Frisian Lake Network, a lake meta-ecosystem located in the North of the Netherlands. We aim to 1) evaluate the performance of the PCLakeS+ model against empirical data, and 2) explore the long-term impacts of an experimental flushing scheme on the lake ecosystem status and network wide nutrient retention of the Frisian Lakes. We showed that model performance fitted reasonably well with measured water quality indicators in the lake, especially given uncertainties in water flows. We found that deriving the required water flows to simulate a meta-ecosystem network was not easily accomplished in relatively complex and heavily managed lake networks, even though a detailed hydrodynamic model was available. In terms of management, we found that the long-term application of the flushing scheme alone would not suffice to initiate an ecosystem state change to a clear, macrophyte dominated system in all lakes but one. Rather, we found that the flushing scheme would increase the nutrient transport towards the Northern regions of the system, as nutrient retention capacity of the lakes in the network was limited. Our lake network model application revealed the importance of additional management measures to achieve desirable environmental outcomes in the Frisian Lake Network.