Conservation et planification spatiale des Ecosystèmes Marins Vulnérables en Méditerranée

Bottom trawling represents the main threat to the integrity of continental shelves and benthic communities. Sharing their habitat with commercially valuable fish and crustacean species, Vulnerable Marine Ecosystems (VMEs) are subject to the collateral impacts of this fishing activity. The cold-water corals, sea pens, sponges, and crinoids that constitute these ecosystems are sessile and form erect structures that are easily destroyed by trawling. Their life-history traits (e.g., slow-growth, late-maturity, and longevity) further limit their capacity for recolonization after disturbance. To date, the conservation of Mediterranean VMEs remains constrained by a complex geopolitical context, fragile legislative framework, and a lack of spatial data, highlighting the need for effective planning, even more urgent in the face of the ongoing climate change. This thesis first aimed to produce new spatial information on the structure and composition of megabenthic communities, the location of areas most sensitive to trawling, as well as the current suitable habitat and future climate refugia of VME indicator species. These data were then integrated into a systematic conservation planning exercise to identify priority networks that are representative, protective of sensitive species, and climate-smart, while minimizing socio-economic impacts. A biogeographic approach confirmed the bathymetric structuring of megabenthic communities. While their low resistance to bottom trawling was demonstrated, variable resilience capacities among taxa were also revealed. Finally, significant habitat losses were predicted for VME indicator species under climate change. The continental shelves and slopes of the Gulf of Lion, the Catalan Sea, and the Strait of Sicily were identified as strategic areas for the establishment of effective conservation measures. Multiple scenarios were proposed to offer flexible solutions for decision-makers. These findings highlight the importance of an adaptive and integrated approach to improve current conservation networks and achieve the ambitious environmental targets set for 2030. The reliability of these results could be enhanced through access to new observations, particularly in the eastern and southern parts of the basin. Integrating the connectivity of marine communities, the consequences of fishing effort redistribution, and improving transparency in communicating uncertainties are also promising perspectives for future conservation planning. Finally, the development of a deep learning approach could revolutionize our ability to predict the location of deep-sea habitats, which remain largely unknown.