Revisiting lake nutrient budgets to predict in‐lake nutrient concentrations for research, monitoring, and management

Abstract Nutrient budgets can be used to predict in‐lake nutrient concentrations. The primary challenge in doing so has been estimating permanent burial, resulting in the development of mechanistic, semi‐mechanistic, and empirical models to quantify this loss process. Data requirements or underlying assumptions of steady state can limit the applicability of these models. Here we provide a framework for deriving and synchronizing influx and efflux estimates for nutrient budgets using limited empirical data, publicly available data, and simple statistical models. We use paleolimnological methods to estimate permanent burial. We generated nutrient budgets for Lake McDonald, Montana, USA, for water years 2008–2023. We compared our nutrient influx and efflux estimates and the predictive ability of our nutrient budgets to those generated for Lake McDonald as part of the 1975 National Eutrophication Survey, which used coarser methods to estimate nutrient influx from tributaries and atmospheric deposition and did not incorporate estimates of permanent burial. Compared to 1975, we estimated lower nutrient loading from Lake McDonald's tributaries, greater phosphorus but lower nitrogen loading from the atmosphere, and the greatest efflux of phosphorus occurred via permanent burial. Our budgets generated more accurate predictions of nutrient concentrations in Lake McDonald (RMSE: , ) than those generated for the 1975 survey (RMSE: , ). Our predictive budgets provide a plausible explanation for the nutrient enrichment event detected in 2018 through simulated wildfire‐enhanced nutrient loading. This method can supplement lake sampling and enable exploration of lake responses to environmental change.