Abstract
The marine biosphere plays a fundamental role in the earth system dynamics (carbon cycle, atmosphere composition, etc.) and provides numerous essential services to humanity (fisheries, carbon sequestration). Its vulnerability to global change (climate change, growing exploitation of natural resources, pollution) makes the study of its evolutions imperative. In this framework, the aim of this thesis is the study of the structuring of the marine biodiversity by hydro-climatic variability of the global ocean, as well as how this structuring impacts on the ecosystems functioning. First, a modelling framework is developed to account for the impact of environment from individuals to populations to communities. It is based on a biomass size spectrum model which represents fish communities with individuals size and species maximum length as only variables. Detailed in Maury O. & Poggiale J. C. (2013) the model is implemented to represent a generic fish community. Indicators of structure, diversity and metabolism are developed to study so represented ecosystems. Once the methodogical framework defined the model is used for the idealized study of the biomass size spectrum properties when constrained with different environmental conditions. The impact of distinct constant primary production and temperature levels are investigated. First the static impact, forcing ecosystems with constant levels. Communities present similar properties for increasing primary production or decreasing water temperature. A succession of four domains characterized by similar fish community features are observed with increasing primary production, decreasing temperature. These distinct charateristics will induce distinct sensitivities of ecosystems function of the level of primary production or temperature. In order to link the results of these numerical experiments to reality the properties of ecosystems along latitudes are also computed. The community biomass spectrum is simulated along stations at different latitudes from pole to equator forcing with mean primary productions and temperatures. The properties of so represented communities enlight the observation of increasing species length with latitude, the so called Bergmann's rule. After the results for static spectra, the dynamic properties of fish community spectra are analyzed through the seasonality. Forcing communities at different latitudes with a seasonal primary production and temperature cycle a more or less strong species sucession is observed. The succession impacts the capacity of communities to maintain larger species during the bad season of poor conditions between two peaks of primary production. Once the properties of the biomass community spectrum investigated in an idealized manner, the model is implemented in the ecosystem model APECOSM in order to account for the spatial interactions and the link with habitat. Applied in the global ocean and forced with physical and biogeochemical NEMO-PISCES domains the model allows the modelling of ecosystems characteristics. Different known characteristics of marine ecosystems are reproduced, especially Bergmann's rule. The current models describing marine ecosystems provide a simplified representation of biodiversity (e.g. NPZD type biochemical models, Ewe or Atlantis type box models, OSMOSE and APECOSM type models). It induces a limitation of their use in the study of the impact of climate change on biodiversity and reversely; for the study of the impact of biodiversity changes on the functioning of ecosystems. The model we describe mechanistically allows the representation of the dynamic of ecosystems from individual bioenergetic and predation interactions while keeping diversity.