Résumé : Given its crucial role in atmospheric chemistry and its global warming potential, methane(CH4) deserves to be accurately budgeted. However, the recent renewed rise in atmosphericCH4 growth rates from 2007 on, after a few years of slow-down, attests that sources are notcompensated anymore by sinks, and calls for a better assessment of the processes contributingto the global CH4 budget. Among natural sources, oceanic emissions are still subject tomany uncertainties, due to the lack of sampling. This is particularly relevant in polar regions,where the role of sea ice on CH4 sea-air fluxes is largely unknown.In an effort to contribute to a better characterisation of CH4 dynamics in oceanic environments,we investigated very contrasted settings during a journey from temperate to polarregions and applied the concentration and stable isotope approach.We start by evaluating the performance of a commercially available in situ CH4 sensor(CONTROS HydroC® CH4 from Kongsberg Contros) in controlled and natural environments,with the hope of using it in the framework of our various field campaigns. Although thissensor has the potential to significantly increase the spatial and temporal resolution comparedto discrete sampling, the long response time prevents from using its measurements as absolutevalues in dynamic natural environments and calls for progress in the field of technologies forcontinuous in situ dissolved CH4 measurements. However, the sensor turns out to be veryuseful during cruises to observe relative changes in dissolved CH4 concentrations and guidethe discrete sampling episodes.Our journey starts in the Scheldt estuary, at the transition between land and sea. Stableisotope analyses reveal that the unusual enrichment of dissolved CH4 in 13C and D in theupper estuary could result from intense microbial oxidation or an unknown source upstream.In the lower part of the estuary, this enriched CH4 mixes with depleted CH4 produced bymethanogenesis in the sediments, before entering the North Sea.In the shallow coastal Wadden Sea, we highlight the dominant contribution of coastal areasto oceanic CH4 emissions. The progressive increase in dissolved CH4 concentrations coincidedwith a 2°C warming of seawater. Submarine groundwater discharge, controlled by thespring-neap tide cycle, and tidal pumping might also have contributed to temporal variationsin dissolved CH4 concentrations and isotopic composition.In the Barents Sea, sailing towards polar latitudes, we find that the fractional sea-ice coverdid not induce a significant change in CH4 concentration nor isotopic composition at theocean-atmosphere interface. Local CH4 seepages at the seafloor might be a relict of gashydrate dissociation after the retreat of the Scandinavian Ice Sheet from the continental shelfafter the Last Glacial Maximum.Trapped in landfast sea ice at Barrow (Arctic) and Cape Evans (Antarctic), we find thatthe processes governing CH4 dynamics in sea ice happen to be dependent on the season andthe regional setting, and can be unravelled thanks to stable isotope analyses. At Barrow,the range of delta-13C values points towards in-ice microbial oxidation of CH4 produced bymethanogenesis in the underlying sediments. At Cape Evans, the much higher delta-13C valuessuggest a hydrothermal origin of CH4 trapped in sea ice and/or aerobic production withinsea ice.The journey ends in the Ross Sea, where the high variability and supersaturation observed indissolved CH4 concentrations, as well as carbon isotope signatures typical of a thermogenicorigin, suggest that gas seepages on the continental shelf might be the main source of CH4 tothe water column.This unique dataset of CH4 concentration and stable isotope composition in seawater, in seaice and in the atmosphere, highlights the spatial and temporal variability of the processesgoverning CH4 dynamics across the various oceanic environments investigated. This thesisprovides an example of how the isotopic approach can be successfully applied to disentanglethe biogeochemical cycle of CH4. To better constrain oceanic emissions, we recommend theimplementation of an extensive monitoring network to measure dissolved CH4 continuously,particularly in shallow coastal regions, which contribute the most. Eventually, further studiesshould focus on the Southern Ocean, which has yet to reveal its secrets with regard to CH4dynamics.