Predicting anoxia in low‐nutrient temperate lakes

Abstract

Absence of dissolved oxygen (anoxia) in the hypolimnion of lakes can eliminate habitat for sensitive species and may induce the release of sediment‐bound phosphorus. Lake anoxia generally results from decomposition of organic matter, which is exacerbated by high nutrient loads. Total phosphorus (TP) in lakes is regulated by static aspects of the lake’s watershed, but lake TP can be readily increased by human activities. In some low‐nutrient lakes, basin morphometry may induce naturally occurring anoxia. The occurrence of natural anoxia is especially important to consider in lake water quality assessments that compare observed conditions to expected reference conditions. To investigate the occurrence of natural vs. anthropogenically influenced anoxia, we constructed a logistic regression model to calculate the probability of low‐nutrient lakes (TP < 15 µg/L) developing aerial anoxic extent ≥10% by testing the predictive potential of variables related to basin morphometry, depths of lake thermal strata, epilimnetic TP, and dissolved organic carbon (DOC). Maximum lake depth and the proportion of lake area under the top of the metalimnion were the most important variables to predict the likelihood of hypolimnetic anoxia, which correctly predicted anoxic condition in 84% of lakes (Model 1). Adding TP as a third variable to Model 1 produced a significantly improved model (Model 2) but the prediction success rate was comparable (86%). We also present a model for lakes with limited bathymetric data, which predicts anoxia with 81% accuracy based on maximum lake depth and mean thermocline depth at peak stratification. DOC was relatively low (4.3 ± 1.5 mg/L [mean ± SD]) in the study lakes and its inclusion did not improve model performance. In Model 1, lakes with an anoxic extent ≥10% of lake area had significantly higher epilimnetic TP than lakes with oxic hypolimnia, regardless of prediction category or success. Our results indicate that including TP as a variable helps refine models based on morphometry alone, but lake morphometry and stratification dynamics are the most important factors in the development of anoxic extent in low‐nutrient temperate lakes. Our approach informs studies concerned with identifying key factors that influence regime shifts in a variety of ecosystems.

Predicting the presence of hypoxic hypolimnia in lakes at large spatial scales

“…Morphometric variables, however, could not predict the presence of hypoxia. This finding is in contrast with a previous study that found that morphometry successfully predicted hypoxia in low nutrient lakes (Deeds et al 2021). In our study, morphometry emerged in predicting hypoxic hypolimnia in the regression tree as the secondary variable and in the random forest focused on dystrophic lakes.…”

“…The influence of heavy precipitation on sudden allochthonous DOC inputs has been shown to cause and maintain hypoxic conditions in a shallow lake (Brothers et al 2014), but the role of browning in driving lakes to hypoxia and anoxia remains under explored. Some studies found no relationships specifically with DOC (Nürnberg 1995; Deeds et al 2021; Senar et al 2021), whereas another site‐specific study did, but only in oligotrophic lakes (Knoll et al 2018). This could potentially be due to a differential influence of browning depending on the nutrient‐color status of the lake.…”

“…For example, in a warmer climate, temperate lakes will have a later ice‐in and earlier ice‐out, where the reduction in period of ice‐cover would effectively increase the duration of thermal stratification, thus promoting, and prolonging widespread hypolimnetic hypoxic conditions (Bartosiewicz et al 2019; Woolway et al 2021; Jane et al 2022). Factors influencing the presence of hypoxic bottom waters appear to vary across lakes (e.g., Nürnberg 1995; Brothers et al 2014; Deeds et al 2021), but include nutrients, color, and lake shape. As such, and in the context of human activities, determining what controls the presence of a physically isolated hypolimnion and promotes its hypoxia at large spatial scales would be useful to predict, in order to understand how lakes process material more broadly across different landscapes or may change and influence the ability to deliver certain ecosystem services.…”

Predicting lake bathymetry from the topography of the surrounding terrain using deep learning

“…Poorer performance for lake morphometry variables was expected, as the model is trained to minimize pixel-wise error. While modeling of lake-level morphometric variables is sufficient for some applications, others require bathymetric maps, for example, for analyses of the distribution and biomass of submerged macrophytes (Duarte and Kalff 1986;Lehmann 1998), greenhouse gas emissions (Li et al 2020), and extent of anoxia (Deeds et al 2021). For such applications, having bathymetric maps available enables improved quantification of such processes on regional to global scales.…”

“…Morphometric variables such as lake volume, mean ( z mean ), and maximum ( z max ) depths are important predictors of many processes, including water retention time, nutrient loading, and cycling, as well as productivity of phytoplankton, zooplankton, and fish (Håkanson 2005). Even more usefully, hypsographic relationships and bathymetric maps communicate the extent of littoral zones colonized by submerged plants (Seekell et al 2021) and hypolimnion zones that might be subject to seasonal oxygen depletion (Deeds et al 2021), and also can be incorporated into physical models (Fricker and Nepf 2000). Lake bathymetry and derived morphometric variables are thus used extensively to investigate lake functioning.…”