Optimizing selective withdrawal strategies to mitigate hypoxia under water-level reduction in Germany's largest drinking water reservoir.

Water-level reduction frequently occurs in deep reservoirs, but its effect on dissolved oxygen concentration is not well understood. In this study we used a well-established water quality model to illustrate effects of water level dynamics on oxygen concentration in Rappbode Reservoir, Germany. We then systematically elucidated the potential of selective withdrawal to control hypoxia under changing water levels. Our results documented a gradual decrease of hypolimnetic oxygen concentration under decreasing water level, and hypoxia occurred when the initial level was lower than 410 m a.s.l (71 m relative to the reservoir bottom). We also suggested that changes of hypoxic region, under increasing hypolimnetic withdrawal discharge, followed a unimodal trajectory with the maximum hypoxic area projected under the discharge between 3 m3/sec and 4 m3/sec. Besides, our results illustrated the extent of hypoxia was most effectively inhibited if the withdrawal strategy was applied at the end of stratification with the outlet elevation at the deepest part of the reservoir. Moreover, hypoxia can be totally avoided under a hybrid elevation withdrawal strategy using surface withdrawal during early and mid stratification, and deep withdrawal at the end of stratification. We further confirmed the decisive role of thermal structure in the formation of hypoxia under water-level reduction and withdrawal strategies. We believe the conclusions from this study can be applied to many deep waters in the temperate zone, and the results should guide stakeholders to mitigate negative impacts of hypoxia on aquatic ecosystems.

Consistent stoichiometric long-term relationships between nutrients and chlorophyll-a across shallow lakes.

Aquatic ecosystems are threatened by eutrophication from nutrient pollution. In lakes, eutrophication causes a plethora of deleterious effects, such as harmful algal blooms, fish kills and increased methane emissions. However, lake-specific responses to nutrient changes are highly variable, complicating eutrophication management. These lake-specific responses could result from short-term stochastic drivers overshadowing lake-independent, long-term relationships between phytoplankton and nutrients. Here, we show that strong stoichiometric long-term relationships exist between nutrients and chlorophyll a (Chla) for 5-year simple moving averages (SMA, median R² = 0.87) along a gradient of total nitrogen to total phosphorus (TN:TP) ratios. These stoichiometric relationships are consistent across 159 shallow lakes (defined as average depth < 6 m) from a cross-continental, open-access database. We calculate 5-year SMA residuals to assess short-term variability and find substantial short-term Chla variation which is weakly related to nutrient concentrations (median R² = 0.12). With shallow lakes representing 89% of the world's lakes, the identified stoichiometric long-term relationships can globally improve quantitative nutrient management in both lakes and their catchments through a nutrient-ratio-based strategy.