Transforming a complex ecosystem model to a MICE and optimizing between complexity and performance

Ecosystem modelling constitutes a fundamental approach to gain insights into the structure, function and dynamic processes of complex marine ecosystems. In this context, species aggregation introduces a way to reduce model complexity and evaluate the effects on model output. Increasing model complexity does not necessarily improve model performance. Regarding model structure and complexity, Models of Intermediate Complexity for Ecosystem assessments (MICE) are positioned between whole-of-ecosystem models, such as Ecopath with Ecosim (EwE), and single-species models. In this study, different Ecopath models were developed, based on an existing EwE model of the Aegean Sea, in which their functional groups (FGs) were gradually aggregated from 44 to 10. A sensitivity analysis was conducted, concerning this stepwise reduction in the number of FGs, aiming in determining the optimum level or “sweet spot” of model complexity. A simplified model of 20 FGs was identified as the optimal balance point between complexity and performance, resulting in an EwE-MICE approach which can be later investigated further with different management scenarios, including spatio-temporal dynamics. Aggregation altered the values of several indicators related to the structural and ecological properties of the ecosystem, while the type of aggregation approach also influenced model output. These findings confirmed that aggregation can be a powerful tool; nevertheless, the optimum complexity can be influenced by research objectives and further research might be required, regarding the methodology for identifying this “sweet spot” in ecosystem models.