Controls on oxygen depletion under lake ice

Even low productive, high-altitude lakes experience deep water hypoxia under ice-cover. While the changing ice phenology is expected to ripple on the magnitude of under-ice hypoxia, the lack of a mechanistic framework linking the physical impact of ice loss to biogeochemical properties has led to seemingly contradictory conclusions.  

Biogeochemical and physical processes constrain the Dissolved Oxygen (DO) dynamic at the sediment-water interface under lake ice. On the one hand, the biogeochemical hypothesis envisions a primary control of DO decay under the ice by sediment oxygen uptake, which arises from benthic microbial respiration and the release of reduced compounds. On the other hand, the physical hypothesis assumes a greater DO decay when sediment heat release reinforces the inverse stratification; the stronger is the sediment heat release, the more the bottom layer, from which oxygen is consumed, gets isolated from potential diffusive resupply from the upper layers. The outcome of a shorter ice-cover on the under-ice DO dynamics depends on the dominance of either biogeochemical or physical processes.

Based on in-situ observations of DO and temperature, we assessed the relative share of biogeochemical and physical processes on decay under the ice of 14 high-altitude lakes in the French Alps. We found highly variable DO decay rates across the different lakes and years, with exponential coefficients ranging from 1.10^-3 to 6.10^-2 d^-1.  The under-ice DO decay rates increased, within years and lakes, with sediment heat release, while biogeochemical factors played only a marginal role. We tested through a reaction-diffusion model on an archetypal, testbed lake the individual effects of biogeochemical versus physical processes on DO decay. We confirmed that the sediment heat flux at ice-on is a major driver of DO decay under the ice, explaining one mechanism by which shallower or more transparent lakes experience greater DO decay under the ice^.