In the context of climate change, Arctic marine ecosystems are affected by rapid environmental modifications, whose effects on biotic communities are still debated. The sea-ice decline and the increase in freshwater inputs and turbidity are likely to impact Arctic primary producers, with cascade effects on a key-process in those ecosystems: the trophic relationship between primary producers and benthic consumers (generally referred as “pelagic-benthic coupling”). The direct study of such complex interaction is not straightforward in the Arctic. The biological model of filter-feeding bivalves offers the possibility to get around these problems, allowing to study those ecological processes indirectly. Among the advantages of this model, there is first of all the fact that these organisms record in their shell, in the carbonate layers, some dynamics of their environments. The information recorded in such “bioarchives” are interpreted through the methods of sclerochronology and sclerochemistry and relate to a time window corresponding to the organism lifespan (from some years to more than 500 years). Given that these organisms are primary consumers, another advantage of this biological model is that the study of their diet can provide information about the trophic relationship with primary producers. With the methods of trophic ecology, especially fatty acids and stable isotopes, the study of the tissues allows the investigation of sources assimilated at a timescale of weeks/months.The main objective of this thesis is to test the potential of bivalves Astarte moerchi (borealis complex) as a biological model for the study of marine Arctic ecosystems. A coupled approach is used to combine shell analysis by the methods of sclerochronology and sclerochemistry (elemental ratios) and tissue analysis by the methods of trophic ecology (fatty acids, carbon and nitrogen stable isotopes, compound-specific carbon stable isotopes on individual fatty acids). Two living A. moerchi populations have been studied in two fjords ...