Search for: All records

Abstract Identifying the scaling rules describing ecological patterns across time and space is a central challenge in ecology. Taylor's law of fluctuation scaling, which states that the variance of a population's size or density is proportional to a positive power of the mean size or density, has been widely observed in population dynamics and characterizes variability in multiple scientific domains. However, it is unclear if this phenomenon accurately describes ecological patterns across many orders of magnitude in time, and therefore links otherwise disparate observations. Here, we use water clarity observations from 10,531 days of high‐frequency measurements in 35 globally distributed lakes, and lower‐frequency measurements over multiple decades from 6342 lakes to test this unknown. We focus on water clarity as an integrative ecological characteristic that responds to both biotic and abiotic drivers. We provide the first documentation that variations in ecological measurements across diverse sites and temporal scales exhibit variance patterns consistent with Taylor's law, and that model coefficients increase in a predictable yet non‐linear manner with decreasing observation frequency. This discovery effectively links high‐frequency sensor network observations with long‐term historical monitoring records, thereby affording new opportunities to understand and predict ecological dynamics on time scales from days to decades. 

Abstract Near‐term ecological forecasting can be used to improve operational resource management in freshwater ecosystems. Here, we developed a framework that uses water temperature forecasting as a tool to predict the migrations of Atlantic salmon (Salmo salar) and European eel (Anguilla anguilla) between freshwater and the sea. We used historical observations of lake water temperature and fish migrations from an internationally important long‐term monitoring site (the Burrishoole catchment, Ireland) to generate daily probabilistic predictions (0%–100%) of when relatively large numbers of fish migrate. For this, we produced daily lake water temperature forecasts that extended up to 34 days into the future using Forecasting Lake and Reservoir Ecosystems (FLARE), an open‐source ensemble‐based forecasting system. We used this system to forecast lake water temperature conditions associated with percentile‐based fish migrations. Two metrics, P66 and P95, were used to indicate days with migrations in excess of 66% and 95%, respectively, of the historical daily fish counts. The results were first validated against water temperature observations, with an overall root mean squared error (RMSE) of 0.97°C. Our forecasts outperformed two other possible water temperature forecasting approaches, using site climatology (1.36°C) and site persistence (1.19°C). The predictions for fish migrations performed better for the P66 metric than for the more extreme P95 metric based on the continuous ranked probability score (CRPS), and the best results were obtained for the salmon downstream migration. This forecasting approach with quantified uncertainty levels has the potential to assist decision making, especially in the face of increased risks for these species. We conclude by discussing the scalability of the framework to other settings as a tool aimed at supporting management practices in real time. 

Understanding controls on primary productivity is essential for describing ecosystems and their responses to environmental change. Lake primary production is strongly controlled by inputs of nutrients and colored dissolved organic matter. While past studies have developed mathematical models of this nutrient-color paradigm, broad empirical tests of these models are scarce. We compiled data from 58 diverse and globally distributed and mostly temperate lakes to test such a model and improve understanding and prediction of the controls on lake primary production. These lakes varied widely in size (0.02-2300 km2), pelagic gross primary production (20-8000 mg C m-2 d-1), and other characteristics. The data package includes high-frequency dissolved oxygen, water temperature, wind speed, and solar radiation data as well as daily estimates of GPP and ER derived from those data. In addition, the data package includes median in-lake and stream concentrations of dissolved organic carbon and total phosphorus for a subset of 18 of those lakes. 

Abstract Understanding controls on primary productivity is essential for describing ecosystems and their responses to environmental change. In lakes, pelagic gross primary productivity (GPP) is strongly controlled by inputs of nutrients and dissolved organic matter. Although past studies have developed process models of this nutrient‐color paradigm (NCP), broad empirical tests of these models are scarce. We used data from 58 globally distributed, mostly temperate lakes to test such a model and improve understanding and prediction of the controls on lake primary production. The model includes three state variables–dissolved phosphorus, terrestrial dissolved organic carbon (DOC), and phytoplankton biomass–and generates realistic predictions for equilibrium rates of pelagic GPP. We calibrated our model using a Bayesian data assimilation technique on a subset of lakes where DOC and total phosphorus (TP) loads were known. We then asked how well the calibrated model performed with a larger set of lakes. Revised parameter estimates from the updated model aligned well with existing literature values. Observed GPP varied nonlinearly with both inflow DOC and TP concentrations in a manner consistent with increasing light limitation as DOC inputs increased and decreasing nutrient limitation as TP inputs increased. Furthermore, across these diverse lake ecosystems, model predictions of GPP were highly correlated with observed values derived from high‐frequency sensor data. The GPP predictions using the updated parameters improved upon previous estimates, expanding the utility of a process model with simplified assumptions for water column mixing. Our analysis provides a model structure that may be broadly useful for understanding current and future patterns in lake primary production. 

Rivers and streams contribute to global carbon cycling by decomposing immense quantities of terrestrial plant matter. However, decomposition rates are highly variable and large-scale patterns and drivers of this process remain poorly understood. Using a cellulose-based assay to reflect the primary constituent of plant detritus, we generated a predictive model (81% variance explained) for cellulose decomposition rates across 514 globally distributed streams. A large number of variables were important for predicting decomposition, highlighting the complexity of this process at the global scale. Predicted cellulose decomposition rates, when combined with genus-level litter quality attributes, explain published leaf litter decomposition rates with high accuracy (70% variance explained). Our global map provides estimates of rates across vast understudied areas of Earth and reveals rapid decomposition across continental-scale areas dominated by human activities. 

Abstract In lakes, the rates of gross primary production (GPP), ecosystem respiration (R), and net ecosystem production (NEP) are often controlled by resource availability. Herein, we explore how catchment vs. within lake predictors of metabolism compare using data from 16 lakes spanning 39°N to 64°N, a range of inflowing streams, and trophic status. For each lake, we combined stream loads of dissolved organic carbon (DOC), total nitrogen (TN), and total phosphorus (TP) with lake DOC, TN, and TP concentrations and high frequencyin situmonitoring of dissolved oxygen. We found that stream load stoichiometry indicated lake stoichiometry for C : N and C : P (r² = 0.74 andr² = 0.84, respectively), but not for N : P (r² = 0.04). As we found a strong positive correlation between TN and TP, we only used TP in our statistical models. For the catchment model, GPP and R were best predicted by DOC load, TP load, and load N : P (R² = 0.85 andR² = 0.82, respectively). For the lake model, GPP and R were best predicted by TP concentrations (R² = 0.86 andR² = 0.67, respectively). The inclusion of N : P in the catchment model, but not the lake model, suggests that both N and P regulate metabolism and that organisms may be responding more strongly to catchment inputs than lake resources. Our models predicted NEP poorly, though it is unclear why. Overall, our work stresses the importance of characterizing lake catchment loads to predict metabolic rates, a result that may be particularly important in catchments experiencing changing hydrologic regimes related to global environmental change.