Élaboration d'un diagnostic robuste de mise en place d'aires marines protégées, dans un contexte d'incertitudes

The Eastern Channel is a strategic area that supports many human activities, in particular fishing, freight, aggregates extraction and tourism. Many zones have also been identified that are suitable to host windfarms. A good understanding of interactions between these activities is necessary in order to estimate, forestall and manage risks of major spatial conflicts. Besides, their potential interactions with marine populations have not yet been well established. Several Marine Protected Areas (MPAs) have been created in the Channel in the past few years, to protect key habitats or populations, or are still being discussed. It is necessary to predict to what extent these MPAs can be an asset for a sustainable exploitation of marine resources, and what their effects will be on the organisation of sectors that depend on these resources. Thus, modelling becomes a relevant approach. One major issue when trying to model marine ecosystems is that they have high uncertainties: marine populations and fishing intensity are very variable, and their state is often poorly known. These uncertainties increase risks that ecosystem management goals may not be reached. Therefore, management measures that are robust to uncertainties and minimise these risks have to be found. Methods have been developed within Decision Theory that aim at allowing decision making under severe uncertainty. One of these methods is the information-gap decision theory. The info-gap method has started to permeate ecological modelling, with recent applications to conservation. However, these practical applications have so far been restricted to rather simple models with analytical solutions. Here we set up a nonprobabilistic management measures robustness assessment approach based on decision theory for a complex, non-analytical model. This approach is applied to the management of the Eastern Channel flatfish fishery. Using the ISIS-Fish modelling platform, we model populations of sole and plaice in this area. A wide range of values of management parameters is tested, so that we can determine which values allow reaching maximum sustainable yield (MSY) with the highest robustness to uncertainties. This model is then further enhanced in order to evaluate the impacts of human activities on various elements of the Eastern Channel ecosystem, in a context of developing spatial management. Effects of MPAs on fishing activities, fish and benthic populations are modelled according to information found in the literature. Impacts of sediment extraction on benthic communities can be determined by means of data obtained on experimental dredging sites located in the Bay of Seine. This study took place from 2007 to 2012, the analysis being performed at the level of the trophic guilds that appear in the diets of sole and plaice. Based on these data, we determine the dynamics of benthic populations at the trophic guild level on dredged sites and on reference sites. These dynamics are then included in the ISIS-Fish model of the area. Finally, we test various management scenarios corresponding to different levels of access limitations to Marine Protected Areas. Among all scenarios tested, we search for those that allow reaching management goals, despite potentially large uncertainties on several parameters linked to the environment, human behavior, or fish biology.