Temperate reservoirs and lakes worldwide are experiencing decreases in ice cover, which will likely alter the net balance of gross primary production (GPP) and respiration (R) in these ecosystems. However, most metabolism studies to date have focused on summer dynamics, thereby excluding winter dynamics from annual metabolism budgets. To address this gap, we analyzed six years of year-round high-frequency dissolved oxygen data to estimate daily rates of net ecosystem production (NEP), GPP, and R in a eutrophic, dimictic reservoir that has intermittent ice cover. Over six years, the reservoir exhibited slight heterotrophy during both summer and winter. We found winter and summer metabolism rates to be similar: summer NEP had a median rate of -0.06 mg O2 L-1 day-1 (range: -15.86 to 3.20 mg O2 L-1 day-1), while median winter NEP was -0.02 mg O2 L-1 day-1 (range: -8.19 to 0.53 mg O2 L-1 day-1). Despite large differences in the duration of ice cover among years, there were minimal differences in NEP among winters. Overall, the inclusion of winter data had a limited effect on annual metabolism estimates, likely due to short winter periods in this reservoir (ice durations 0–35 days) relative to higher-latitude lakes. Our work reveals a smaller difference between winter and summer NEP than in lakes with continuous ice cover. Ultimately, our work underscores the importance of studying full-year metabolism dynamics in a range of aquatic ecosystems to help anticipate the effects of declining ice cover across lakes worldwide.

Near-term, iterative ecological forecasts can be used to help understand and proactively manage ecosystems. To date, more forecasts have been developed for aquatic ecosystems than other ecosystems worldwide, likely motivated by the pressing need to conserve these essential and threatened ecosystems and increasing availability of high-frequency data. Forecasters have implemented many different modelling approaches to forecast freshwater variables, which have demonstrated promise at individual sites. However, a comprehensive analysis of the performance of varying forecast models across multiple sites is needed to understand broader controls on forecast performance. Forecasting challenges (i.e., community-scale efforts to generate forecasts while also developing shared software, training materials, and best practices) present a useful platform for bridging this gap to evaluate how a range of modelling methods perform across axes of space, time, and ecological systems. Here, we analysed forecasts from the aquatics theme of the National Ecological Observatory Network (NEON) Forecasting Challenge hosted by the Ecological Forecasting Initiative. Over 100,000 probabilistic forecasts of water temperature and dissolved oxygen concentration for 1-30 days ahead across seven NEON-monitored lakes were submitted in 2023. We . For water temperature, we found that forecast skill degraded with increases in forecast horizons, process-based models and models that included air temperature as a covariate generally exhibited the highest forecast performance, and that the most skillful forecasts often Overall, the NEON Forecasting Challenge provides an exciting opportunity for a model inter-comparison to learn about the relative strengths of a diverse suite of models and advance our understanding of freshwater ecosystem predictability.

Water level drawdowns are increasingly common in lakes and reservoirs worldwide as a result of both climate change and water management. Drawdowns can have direct effects on physical properties of a waterbody (e.g., by altering stratification and light dynamics), which can interact to modify the waterbody’s biology and chemistry. However, the ecosystem-level effects of drawdown remain poorly characterized in small, thermally-stratified reservoirs, which are common in many regions of the world. Here, we intensively monitored a small eutrophic reservoir for two years, including before, during, and after a month-long drawdown that reduced total reservoir volume by 36%. During drawdown, stratification strength (maximum buoyancy frequency) and surface phosphate concentrations both increased, contributing to a substantial surface phytoplankton bloom. The peak in phytoplankton biomass was followed by cascading changes in surface water chemistry associated with bloom degradation, with sequential peaks in dissolved organic carbon, dissolved carbon dioxide, and ammonium concentrations that were up to an order of magnitude higher than the previous year. Dissolved oxygen concentrations substantially decreased in the surface waters during drawdown (to 41% saturation), which was associated with increased total iron and manganese concentrations. Combined, our results illustrate how changes in water level can have cascading effects on coupled physical, chemical, and biological processes. As climate change and water management continue to increase the frequency of drawdowns in lakes worldwide, our results highlight the importance of characterizing how water level variability can alter complex in-lake ecosystem processes, thereby affecting water quality.

Water temperature forecasting in lakes and reservoirs is a valuable tool to manage crucial freshwater resources in a changing and more variable climate, but previous efforts have yet to identify an optimal modelling approach. Here, we demonstrate the first multi-model ensemble (MME) reservoir water temperature forecast, a forecasting method that combines individual model strengths in a single forecasting framework. We developed two MMEs: a three-model process-based MME and a five-model MME that includes process-based and empirical models to forecast water temperature profiles at a temperate drinking water reservoir. Our results showed that the five-model MME improved forecast performance by 8-30% relative to individual models and the process-based MME, as quantified using an aggregated probabilistic skill score. This increase in performance was due to large improvements in forecast bias in the five-model MME, despite increases in forecast uncertainty. High correlation among the process-based models resulted in little improvement in forecast performance in the process-based MME relative to the individual process-based models. The utility of MMEs is highlighted by two results: 1) no individual model performed best at every depth and horizon (days in the future), and 2) MMEs avoided poor performances by rarely producing the worst forecast for any single forecasted period (<6% of the worst ranked forecasts over time). This work presents an example of how existing models can be combined to improve water temperature forecasting in lakes and reservoirs and discusses the value of utilising MMEs, rather than individual models, in operational forecasts.