AbstractSeveral findings suggest that CO2 emissions in lakes are not always directly linked to changes in metabolism but can be associated with interactions with the dissolved inorganic carbon equilibrium. Alkalinity has been described as a determining factor in regulating the relative contributions of biological and inorganic processes to carbon dynamics in lakes. Here we analyzed the relationship between metabolic changes in dissolved oxygen (DO) and dissolved inorganic carbon (DIC) at different timescales in eight lakes covering a wide range in alkalinity. We used high-frequency data from automatic monitoring stations to explore the sensitivity of DIC to metabolic changes inferred from oxygen. To overcome the problem of noisy data, commonly found in high-frequency measurements datasets, we used Singular Spectrum Analysis to enhance the diel signal-to-noise ratio. Our results suggest that in most of the studied lakes, a large part of the measured variability in DO and DIC reflects non-metabolic processes. Furthermore, at low alkalinity, DIC dynamics appear to be mostly driven by aquatic metabolism, but this relationship weakens with increasing alkalinity. The observed deviations from the metabolic 1:1 stoichiometry between DO and DIC were strongly correlated with the deviations expected to occur from calcite precipitation, with a stronger correlation when accounting also for the benthic contribution of calcite precipitation. This highlights the role of calcite precipitation as an important driver of CO2 supersaturation in lakes with alkalinity above 1 meq L−1, which represents 57% of the global area of lakes and reservoirs around the world.