Reservoirs in the Great Plains are ecologically and economically vital for supporting regional fishery and waterfowl production, yet their ecological dynamics remain underexplored. In Kansas, most reservoirs are highly productive (eutrophic to hypereutrophic), but the temporal dynamics of this productivity are uncertain. Moreover, complex bathymetry in reservoirs creates potential for significant spatial variation in productivity. To investigate these patterns, we deployed an array of buoy-mounted sondes in Clinton Lake, a large reservoir in eastern Kansas, to measure diel dissolved oxygen levels throughout the 2024 growing season (spring–fall). These measurements were used to estimate gross primary production, ecosystem respiration, and net ecosystem production, providing a detailed characterization of lake metabolism across spatial and temporal scales. We found extreme variation in metabolism metrics through time, with NEP shifting from positive to negative on an almost daily basis, implying high-frequency changes from autochthony to heterotrophy. Metabolic variation in the reservoir is associated with mixing events, driven by air/water temperatures, wind speed, and precipitation/inflow events. Calm wind and thermally stratified periods encourage cyanobacteria dominance, which are then pushed deeper in the water column and decay during mixing events. Metabolism metrics often followed similar trends across locations in both main arms of the reservoir and main basin, implying climatic control of lake metabolism. This was not always the case, however, as there were periods with variation between reservoir arms. This suggests that ecosystem metabolism in reservoirs is generally controlled by climate, but site-level differences can alter function as well. Our findings offer new insights into the ecological function of different zones within Clinton Lake, enhancing our understanding of Great Plains reservoirs as ecosystems and fisheries.