Estimating nutrient thresholds for eutrophication management: Novel insights from understudied lake types

A continuous classification of the 476,697 lakes of the conterminous US based on geographic archetypes

“…The design of monitoring programs to collect in situ ecosystem data is notoriously challenging because of the need to capture the full range of ecologically relevant heterogeneity that exists at both local and regional scales (Janousek et al 2019;Soranno et al 2020). Our results show that a continuous approach based on lake archetype classification can facilitate the identification of lakes with similar limnological states and responses, two key management targets that can sometimes form the foundation for such monitoring programs (Soranno et al 2010;Yuan et al 2014;Poikane et al 2022). For example, because all 476,697 conterminous US lakes ≥ 1 ha are assigned a weight for each archetype, it is possible to use the archetype model output to estimate potential ranges of water quality related in-lake measurements (nutrients, carbon, clarity), even for lakes that have not been sampled.…”

“…Classification systems are diverse in their structure and aims, with some of them focusing on in situ observations of temperature, water chemistry, or biodiversity of the ecosystem studied, while others have focused on the physical or geographic properties of the aquatic ecosystem itself or its surrounding terrestrial landscape. Using these approaches, researchers have identified regions or types of lakes with (1) similar ecosystem properties such as biodiversity, nutrient concentrations, or alkalinity (Emmons et al 1999; Phillips et al 2008; Dodds et al 2019; Lemm et al 2021), or (2) similar driver‐response relationships, as exemplified by stronger limnological relationships between concentrations of total Nitrogen (TN) and total Phosphorus (TP), or between algal biomass and nutrients within rather than among regions or types (Yuan et al 2014; Poikane et al 2022). Classification approaches have thus been successful in categorizing lakes into meaningful and useful groups, providing a predictive framework for understanding limnological properties and responses among broadly distributed ecosystems.…”

“…While all classification approaches essentially achieve this variable reduction, a lack of consensus on methods and approaches, perhaps explained by a combination of variability in spatial and temporal scales, project objectives, and data availability, has complicated efforts to make and compare inferences across different classification systems. For example, European countries have developed tens of lake classification systems, often at the national level, resulting in hundreds of differently defined lake types under the European Water Framework Directive (Poikane et al 2022).…”

“…Furthermore, classification systems can be discrete and rely on a smaller number of in situ variables available for already sampled lakes such as trophic status (Carlson 1977; Wetzel 2001), stratification (Lewis Jr 1983), or alkalinity (Kelly et al 2012; Solheim et al 2019; Poikane et al 2022), but cannot be readily applied to lakes for which the classifying variables are not available. Other approaches have used a larger number of ex situ variables for all lakes within a region (e.g., topographic or geological variables available from geospatial datasets; Hill et al 2018; Solheim et al 2019), but empirical demonstrations of the ability of these classifications to describe lake functioning remain rare.…”

Response of lake metabolism to catchment inputs inferred using high‐frequency lake and stream data from across the northern hemisphere

“…Hence, catchment‐lake management may be more successful when the models used to make predictions about lake conditions include stream load information. Future field and modeling studies related to how changes in catchment loading, hydrology, or other factors not considered here (i.e., temperature) affect metabolism at multiple timescales, as well as studies on lakes outside of the north‐temperate zone (e.g., Brighenti et al 2018; Poikane et al 2022), will help us better predict changes in lake metabolism related to global change. Furthermore, as much as stream C, N, and P loadings can affect lake metabolism, lake metabolism can also imprint in downstream exports of C, N, and P. Therefore, understanding how metabolism is influenced by multiple catchment forcings will not only lead to better predictions of how lake metabolism may change in the future, but also how this affects landscape‐scale carbon and nutrient cycling along the aquatic continuum.…”