Oxygen (O2) is a fundamental parameter for life. It not only profoundly influences the biogeochemical cycling on a global scale, but also deals with the regulation of metabolic processes at microbial level, in particular the transition between aerobic and anaerobic metabolisms.However, until recently, due to the lack of high-resolution methods for O2 concentration determination, several oxygen-related processes, such as aerobic respiration in pelagic aquatic ecosystems and in naturally oxygen poor waters (e.g. Oxygen Minimum Zones, OMZs), or the oxygen regulation of nitrification and denitrification, were only partially described. In spite of the importance of aerobic respiration as a key process in the global carbon cycle, the available data are still few, and highly biased with respect to season, latitude and depth.The main aims of this Ph.D were to: i) develop and test a highly-sensitive method to assess O2 respiration rates, using of the ultrasensitive Switchable Trace amount Oxygen (STOX) sensor, ii) apply the method to investigate the O2 uptake kinetics of natural planktonic communities from OMZs, iii) investigate the influence of low O2 concentration on nitrification, in particular on nitrite-oxidizing bacteria (NOB) and the expression of their terminal oxidases genes.The results from this project show that it is indeed possible, by combining high resolution sensor technology (the STOX sensor) and all-glass bottle incubation (Appendix A) to achieve an extreme high sensitivity in O2 detection (1-10 nM, and ~ 2 nM O2 /h, for respiration rates) (Manuscript I, II and IV). Thus this new method allows for precise determination of respiration rates also in low activity waters, within few hours of incubation time (Manuscript II). This low-O2 STOX based method has been compared with other methods suitable for measurements at low O2 (Manuscript IV) showing comparable results, together with an overall higher precision and less susceptibility to bias from light/temperature effects. The resolution of the measurements with the STOX sensor is inversely proportional to the O2 concentrations within the optimal working range of the STOX sensors. Therefore, in order to reach the highest resolution, the initial O2 concentration in the samples needs to be artificially lowered. Nevertheless, similar results were obtained when comparing the respiration rates measured by our method with the ones, from full O2 saturation, measured by traditional Winkler titration (Manuscript I). It was thus possible to measure respiration rates from planktonic communities in both coastal Danish waters, with relatively high activity (600-800 nM O2/h, in summer), and in waters from four stations in the Eastern Tropical North and South Pacific OMZs, with low activity (200- 4 nM O2/h) (Manuscript II). Our investigations suggest that diverse natural planktonic communities have very low apparent Km values for oxygen (< 200 nM) and that aerobic respiration is likely to proceed efficiently down to the O2 nanomolar range. Communities with such characteristics were found in both fully oxygenated and deoxygenated aquatic environments, implying that the transition from aerobic to anaerobic metabolisms may occur at very low O2 concentrations and that some aerobic and anaerobic processes may co-exist in the water column at these micro-oxic conditions. Moreover, Km values in the O2 nanomolar range have also been found for other important aerobic processes in the ocean, such as ammonia oxidation and nitrite oxidation (Manuscript V).The application of the STOX sensor for in situ high-resolution profiling on oceanic cruises allowed us to support distribution studies of iron species in the Eastern Tropical North Pacific (ETNP) OMZ water column, and to gain evidences that this OMZ is characterized by a functionally anoxic core (Anoxic Marine Zone, AMZ) (Manuscript II, VI).Subsequently we applied our low-O2 method to pure cultures (Manuscript III), in order to assess the response of three species of NOB (Nitrospira defluvvi, N. moscoviensis and Nitrospina gracilis) to low O2 concentrations, and the oxygen regulation on the expression of the terminal oxidases genes in N.moscoviensis. The oxygen affinities of these pure cultures were lower than found for natural communities of NOB (apparent Km values~ 1- 4 µM), but higher than the ones from the well-studied opportunistic NOB Nitrobacter. The expression of high-affinity terminal oxidases in these NOB could, however, not be confirmed. Overall the results of this Ph.D. constitute important advances in our ability of assessing respiration rates in discrete samples, and in our understanding of the O2 levels involved in supporting aerobic metabolisms and the shift to anaerobic metabolic pathways in marine environments. Aerobic processes seem to proceed efficiently down to vanishingly low O2 concentrations, and they are likely to co-occur, in some environments, with anaerobic ones. These findings call for a need to change and reshape our definitions of oxic and anoxic environments, as well as the microbial metabolic pathways involved; thus to expand our knowledge on this topic through our newly acquired technical abilities to investigate lower O2 concentrations.