Application of spatial food web simulations at a marine Natura 2000 site: analysis of vulnerability and management actions

Marine Natura 2000 sites play an important role in nature conservation, while representing an opportunity to promote sustainable natural resources exploitation. To be effective, Natura 2000 sites should be managed considering spatial and temporal ecosystem dynamics. In this study, conducted within the Interreg project CASCADE, we applied a spatially explicit food web model in Ecopath with Ecosim to the Tegnùe di Chioggia SCI, a Natura 2000 site characterized by biogenic rocky outcrops and still lacking a proper management plan. Following the principles of the Ecosystem Approach, biological and economic compartments were included in the model. Species groups distribution and spatial dynamics of main fishing fleets were considered. Through temporal simulations, we first assessed the vulnerability of the socio-ecological system to climate change and fishing effort scenarios. Spatial simulations and multi-criteria analysis were combined, to compare the effects of alternative management actions on the ecosystem (artisanal fishing in the SCI, SCI area extension and mussel farms expansion). Ecosystem indicators were used to summarize simulations outputs. Results confirm the expected vulnerability to climate change: increased water temperature induces a reduction in biomass diversity (lower Kempton index) and commercial species abundance, while species associated to rocky outcrops (e.g., Diplodus annularis) are predicted to decrease substantially. Results indicated that bycatch threats trophic groups when fishing effort is increased over time. Spatial food web simulations also reveal how ecosystem complexity can lead to unexpected results when different management options are compared. Indeed, based on the multi-criteria analysis, the expansion of mussel farming areas was the best management scenario, partly explained by the collateral no-take zone effect for trawlers. Constraints on model performances imposed by data availability, and future monitoring efforts to fill these gaps, are discussed.