Assessing models' sensitivity to the effects of forest management and climate change on carbon and water fluxes in European beech forests

The consequences of climate change continue to threaten European forests, particularly for species located at the edges of their latitudinal and altitudinal ranges. While extensively studied in Central Europe, European beech forests require further investigation to understand how climate change will affect these ecosystems in Mediterranean areas. Proposed silvicultural options increasingly aim at sustainable management to reduce biotic and abiotic stresses and enhance these forest ecosystems' resistance and resilience mechanisms. Process-based models (PBMs) can help us to simulate such phenomena and capture early stress signals while considering the effect of different management approaches. In this study, we focus on estimating sensitivity of two state-of-the-art PBMs forest models by simulating carbon and water fluxes at the stand level to assess productivity changes and feedback resulting from different climatic forcings. Utilizing 3D-CMCC-FEM and MEDFATE models, we simulated and analyzed carbon (C) and water (H20) fluxes in diverse forest plots under managed vs. unmanaged scenarios and under current climate and different climatic forcings (RCP4.5 and RCP8.5), in two sites, on the Italian peninsula, Cansiglio in the north and Mongiana in the south. To ensure confidence in the models' results, we first evaluated their performance in simulating C and H2O flux in three additional beech forests along a latitudinal gradient spanning from Denmark to central Italy. The results from both models for C and H2O flux assessment showed generally good model accuracy. At the Cansiglio site in northern Italy, both models simulated a general increase in C and H2O fluxes under the RCP8.5 climate scenario compared to the current climate. Still, no benefit in managed plots compared to unmanaged ones, as the site does not have water availability limitations, and thus, competition for water is low. At the Mongiana site in southern Italy, both models simulate a decrease in C and H2O fluxes and sensitivity to the different climatic forcings compared to the current climate, with an increase in C and H2O fluxes considering specific management regimes compared to unmanaged scenarios. Conversely, in both models, under unmanaged scenarios, plots are simulated to experience first signals of mortality prematurely due to water stress (MEDFATE) and carbon starvation (3D-CMCC-FEM) scenarios. In conclusion, while management interventions may be considered a viable solution for the conservation of beech forests under future climate conditions at moister sites like Cansiglio, in drier sites like Mongiana may not lie in management interventions alone but rather in the establishment of synergistic mechanisms with other species.