Response of a complex ecosystem model of the northern Adriatic Sea to a regional climate change scenario

This paper investigates the impact of a regional climate change scenario on an ecosystem model of the northern Adriatic Sea. The study was performed by applying a biogeochemical biomass-based ecosystem model (ERSEM: European Regional Seas Ecosystem Model) coupled with a 1-dimensional version of a commonly used hydrodynamical model (POM: Princeton Ocean Model). The model response was studied by performing 2 simulations: one with atmospheric forcing functions for present-day climate conditions ('Control', 1970-1999), and one for the future climate after a doubling of the CO2 concentration ('Scenario', 2060-2089). Time-series of meteorological forcing functions were obtained from a greenhouse gas experiment with the high-resolution version of the ECHAM4 atmospheric general circulation model with a horizontal resolution of about 100 km. Present-day simulation results were compared to the available historical observations and showed a satisfactory agreement with the main seasonal cycles of hydrodynamical variables and nutrients in the limits of the 1-dimensional representation. Under the Scenario conditions, the model predicted an overall enhancement of the water-column stratification on an annual basis, with stronger intensification during the summer periods. The diffusion of oxygen and nutrients between surface and bottom layers is reduced, and the transfer of organic matter through the food web is shifted towards the smaller components of the microbial web. The model gives indications of an increase in the specific phytoplankton uptake rates of inorganic carbon in the Scenario, but the photosynthesized carbon ends up in the DOC pool because of the reduced supply of nutrients. The mean annual concentration of DOC is thus higher in the Scenario, with possible negative consequences on the water quality. Additional experiments show that the response of primary producers is not directly linked to the increase in ambient temperature. The structure of the food web modulates the interaction of phytoplankton with the abiotic conditions. The results give indications of a possible non-linear response of the biogeochemical cycles to the climate change warming scenario and confirm the feasibility of using the downscaling technique to couple large-scale climate models with comprehensive regional ecosystem models.