Spatial versus temporal variability in surface oxygen and its metabolic consequences in temperate and subtropical...

Shallow lakes are biogeochemical hotspots that play a disproportionate role in global carbon cycles, yet surface oxygen (O₂) dynamics-critical for estimating metabolic activity and CO₂ fluxes-remain poorly characterized across spatial and temporal scales. While most studies emphasize temporal and vertical variations, horizontal heterogeneity across lake surfaces is often overlooked. To address this gap, we compiled high-frequency, multi-site surface O₂ monitoring data throughout 2021 in two large shallow lakes-the temperate Lake Hulun and the subtropical Lake Taihu. By partitioning horizontal spatial and temporal O₂ variance, we found that single-site sampling misestimated lake-wide O₂ concentrations by up to 29% and metabolic rates by up to 147%. Seasonal changes were the dominant driver of O₂ dynamics, introducing even greater errors (34-35%), while daily variability contributed 13-20% uncertainty, highlighting the limitations of infrequent sampling. Environmental controls operated at distinct scales: spatial O₂ heterogeneity was shaped by ecosystem respiration and algal gradients; seasonal variations were driven primarily by temperature and metabolic cycles; and daily (diel) fluctuations were governed by the interplay between production and respiration, moderated by carbonate buffering. Horizontal heterogeneity consistently distorted lake-wide metabolic estimates, underscoring the need to integrate spatial O₂ data into ecosystem models to reduce errors in lake metabolism estimates. High-resolution spatiotemporal monitoring and targeted seasonal sampling are imperative to constrain uncertainties in lake metabolism and carbon flux estimates under climate change.