Spatiotemporal variation of ecosystem metabolism in a eutrophic, polymictic reservoir: Water column stability begets metabolism stability

Reservoirs are globally important aquatic ecosystems that provide critical anthropogenic functions including water storage, flood control, and fisheries, yet their ecosystem metabolism remains understudied compared to natural lakes. We estimated daily lake metabolism (net ecosystem production [NEP], gross primary production [GPP], and ecosystem respiration [R]) at nine locations throughout Clinton Lake, a temperate, hypereutrophic, polymictic reservoir in Kansas, USA, during the 2024 growing season. Using dissolved oxygen monitoring and Bayesian hierarchical generalized additive mixed models, we assessed spatiotemporal variation in metabolism and relationships with environmental predictors including thermal stratification, temperature, wind, and phytoplankton pigment data. Clinton Lake exhibited dynamic metabolism with frequent daily alternations between autotrophy and heterotrophy, resulting in near-zero mean NEP (0.01 mg O₂ L⁻¹ d⁻¹) across the study period. All metabolism components showed strong seasonal patterns, with GPP and R magnitudes increasing throughout the growing season. Temporal variation dominated over spatial variation in all metabolism metrics, but modest site-level effects were present in both GPP and R models. Thermal stratification emerged as a main driver of metabolism variability. Stronger stratification was associated with higher GPP and R rates, while changes in stratification strength corresponded to directional changes in NEP: stratification events promoted autotrophy, while destratification promoted heterotrophy. We suggest the coupling between metabolism and stratification is driven by the dominance of buoyant cyanobacteria controlling oxygen dynamics near the surface. Our results suggest that Great Plains reservoirs can function as highly dynamic ecosystems where wind-driven mixing regimes precede nutrients or light for the primary control of metabolic processes. The frequent stratification-destratification cycles characteristic of polymictic systems create spatiotemporally variable ecosystem function. These findings provide critical baseline understanding for monitoring future changes in reservoir ecosystems and highlight the importance of physical forcing in structuring metabolism in engineered aquatic systems.