Volume 10, Issue 2 e1629
Open Access

The ecosystem implications of road salt as a pollutant of freshwaters

Hilary A. Dugan

Corresponding Author

Hilary A. Dugan

Center for Limnology, University of Wisconsin-Madison, Madison, Wisconsin, USA


Hilary A. Dugan, Center for Limnology, University of Wisconsin-Madison, 680 N. Park St., Madison, WI 53706, USA.

Email: [email protected]

Contribution: Conceptualization (equal), Data curation (lead), Formal analysis (lead), Visualization (lead), Writing - original draft (equal)

Search for more papers by this author
Shelley E. Arnott

Shelley E. Arnott

Department of Biology, Queen's University, Kingston, Ontario, Canada

Contribution: Conceptualization (equal), Writing - original draft (equal)

Search for more papers by this author
First published: 19 December 2022
Citations: 3
Edited by: Wendy Jepson, Co-Editor-in-Chief

Funding information: National Science Foundation, Grant/Award Number: DEB-2144750; Natural Sciences and Engineering Research Council of Canada, Grant/Award Number: RGPIN-2019-04315


Salt pollution is a threat to freshwater ecosystems. Anthropogenic salt inputs increase lake and stream salinity, and consequently change aquatic ecosystem structure and function. Elevated salt concentrations impact species directly not only through osmoregulatory stress, but also through community-level feedbacks that change the flow of energy and materials through food webs. Here, we discuss the implications of road salt pollution on freshwater rivers and lakes and how “one size fits all” ecotoxicity thresholds may not adequately protect aquatic organisms.

This article is categorized under:

  • Science of Water > Water Quality
  • Water and Life > Nature of Freshwater Ecosystems
  • Water and Life > Stresses and Pressures on Ecosystems

Graphical Abstract

Salt pollution is a threat to freshwater ecosystem. Research and education are needed to advance our approaches to studying freshwater salinization, and outreach and education are needed to reform our practices and protect our freshwater and human health.


Humans have a long and storied history with salt, a mineral so common to our everyday lives that we rarely eat a meal without it, and yet so consequential in the history of humanity that this simple mineral has led to historic events like Gandhi's Salt March, an act of historic nonviolent civil protest against colonial oppression. Human's demand for salt does not stop at food; its presence is found in industries, agriculture, and transportation, and yet our extensive use of salt is at odds with humanity's need for freshwater resources.

Freshwater can be contaminated by many pollutants, but only the addition of salts will strip the designation of “fresh.” Given the susceptibility of freshwater to salinization, the omnipresence of salt pollution from human activities leaves many freshwater resources vulnerable (Hintz, Fay, et al., 2022). In the United States, upwards of 50–60 million tons (45–55 million metric tons) of NaCl salt is consumed annually, the majority of which is used in either the chemical industry or as road deicer (Figure 1, US Geological Survey [USGS], 2022). Salt (hereafter NaCl) is applied to roads and other paved surfaces to reduce or prevent the build up of ice, thereby creating safer conditions for vehicles and pedestrians. Salt works by lowering the freezing point of water to −21°C. It works best in the temperature range between 0 and −10°C and therefore it is sometimes used in association with abrasives and other salts (e.g., CaCl2 and MgCl2, which have lower eutectic points [lowest possible freezing temperature]) in regions with cold winters. De-icing salts are commonly applied in solid form (i.e., rock salt), but application as liquid brine (23% salt) is becoming more common. Liquid brine has been found to be more effective per pound of salt, which reduces overall salt use (Claros et al., 2021). The use of road salt in North America started in the 1940s and has steadily increased through the decades as the road network has grown (Figure 1; Corsi et al., 2010).

Details are in the caption following the image
Total annual consumption of salt in the United States (light blue bars). Over the last decade, roughly 40% of domestically produced and important salt is used as road deicer (dark blue bars). All data from US Geological Survey (USGS) mineral commodity summaries (USGS, 2022). For context, 20 million tons of salt would be required to shift 20,735 Olympic-size swimming pools from pure water to ocean salinity.

The entirety of road deicing salt is released into the environment, moving directly through storm sewers or gradually percolating through soils, and is known to be the leading cause of widespread salinization of freshwater lakes, rivers, wetlands, and groundwater in regions with heavy road salt use (Dugan et al., 2017a; Herbert et al., 2015; Kaushal et al., 2018; Stets et al., 2017). In many cities, salinization is exacerbated by point source pollution from water softener and industrial effluent discharged directly to rivers. Here, we discuss the ecosystem implications of salt as a pollutant of freshwaters, with a focus on road deicers given the magnitude and breadth of their use. Although much of the research cited here comes from north-central and north-eastern North America, a region known for both plentiful freshwater and heavy road salt use, our discussion of the ecosystem implications of salinization encompasses all agents of freshwater salinization. Globally, these include climate change (e.g., increasing aridity), freshwater diversions, seawater intrusion, mining (e.g., potash), and agriculture (e.g., intensive irrigation) (Cunillera-Montcusí et al., 2022).


Monitoring of salt pollution is conducted via measurements of salinity and electrical conductivity (EC), which are integrative measurements of all dissolved ions, as well as chloride concentrations because it is a common anion found in most road salts (Figure 2). Salinity is a chemical term that refers to the mass of dissolved inorganic solids found in water, and for most limnological purposes, can be considered the sum of the mass fraction of the major cations (Na+, Mg2+, Ca2+, K+), anions (Cl, SO 4 2 ), and carbonate species ( HCO 3 , CO 3 2 ). Because salinity is dominated by charged ionic constituents, salinity can be approximated by EC; a measure of water's ability to conduct electrical flow, expressed in μS cm−1. For ease of comparison, EC should be temperature corrected to 25°C and reported as specific conductance (SpC).

Details are in the caption following the image
Global surface waters span a range of salinities. Freshwater lakes are generally considered to be <1–3 g L 1 . Salinization is monitored through a variety of water quality measurements. Salinity, total dissolved solids (TDS), specific conductance (SpC), electrical conductivity (EC), and chloride concentrations, are all indicators of changing salinity.

The delineation between fresh and saline water is not fixed, but is often marked at a salinity of 3 g L−1. However, in countries replete with freshwater (e.g., Canada), it is not uncommon to see freshwater categorized as a salinity <1 g L−1, and drinking water regulated to be <0.5 g L−1. For reference, a teaspoon of salt weighs 5–6 g. Therefore, a teaspoon of salt poured into 2 L or a half-gallon of pure water would render that water saline, and six times over the drinking water limit.

Typically salt (singular) refers to NaCl, the most commonly used form of salt, and salts (plural) refers to any noncharged chemical compound made up of cations and anions (e.g., saltpeter KNO3). Anthropogenic salt pollution is often in the form of chloride salts (e.g. NaCl, CaCl2, MgCl2) or chloride brines, and therefore changes in salinity and EC are often proportional to changes in chloride (Cl). In biological applications, it is often more appropriate to monitor individual ions over salinity, as toxicity and water quality criteria are evaluated at the level of individual chemicals (i.e., Na+, Cl) rather than bulk salinity.

To monitor chloride concentrations, manual water samples are analyzed in a laboratory, often by using ion chromatography. Discrete sampling can be supplemented with high-frequency measurements of EC, where sondes are deployed in lakes or streams. Care must be taken when correlating EC to Cl, since biologically driven changes in inorganic carbon concentrations ( HCO 3 , CO 3 2 ) unrelated to anthropogenic salinization will change EC.

Where high temporal sampling is not feasible, known patterns in salt dynamics can inform initial monitoring strategies. For instance, numerous studies have concluded that impervious surfaces (i.e., roads) are a strong predictor of chloride pollution (Oswald et al., 2019; Dugan et al., 2020). Therefore, priority should be given to lakes and streams with urban landcover or major highways. Secondarily, freshwaters downstream of wastewater discharges or land with intensive agriculture should be monitored. In rivers, temporal sampling is critical. Often chloride concentrations peak in late winter/early spring due to road salt runoff. Chloride concentrations can be orders of magnitude higher than summer baseflow concentrations (Figure 3a,b). In lakes, longer water residence times diminish seasonal peaks (Figure 3c). When the water column is mixed, a surface sample may be enough to ascertain total lake concentration. In deep lakes, dense saline inflows may pool at depth, which is easily revealed by a water column profile.

Details are in the caption following the image
(a) Annual dynamics and (b) monthly ranges in chloride concentrations in the urban Kinnickinnic River in Milwaukee, WI, USA. (c) Long-term increases in chloride in three lakes in Madison, WI, USA. Data from the US Water Quality Portal (Read et al., 2017), and the North Temperate Lakes Long-Term Ecological Research Program (NTL-LTER, 2018, 2020).

Quantifying the magnitude of chloride pollution requires a knowledge of “background or natural” concentrations. For instance, in Lake Mendota, WI, chloride concentrations in the 1940s were often <3 mg L−1 (Figure 3c). In this case, long-term data enables us to draw conclusions about the rate and magnitude of anthropogenic salinization. Unfortunately, few datasets extend back past the 1950s, prior to the advent and explosion of road salt use in North America. However, using long-term data and current data of “pristine” systems, we can draw the conclusion that the background chloride concentrations in most freshwater lakes, rivers, and wetlands should be <10–20 mg L−1, and in many cases <5 mg L−1 (Hintz & Relyea, 2019) in central and eastern North America.


In evaluating salt as a freshwater pollutant it is important to understand the impacts on the physical dynamics of aquatic systems. In rivers, turbulent mixing is driven by flow, whereas in lakes, water column mixing is induced by both turbulent (wind) and buoyant (density) forces. Mixing in aquatic systems is an important control on the thermal and biogeochemical (oxygen, nutrients) habitat. In many lakes, density differences in the water column will cause the lake to stratify for extended periods of time. This density-imposed stratification in freshwater lakes is driven by temperature differences, and is a barrier to full lake mixing (Boehrer & Schultze, 2008).

High salt concentrations will also stabilize lake water columns because saline water is denser than freshwater. The accumulation of anthropogenic salts in a lake can increase density gradients in the water column and lead to delayed, diminished, or disrupted lake mixing. This phenomenon has been documented in 12-m deep Woods Lake in Michigan (Koretsky et al., 2012; Sibert et al., 2015), and 23-m deep Irondequoit Bay in Lake Ontario (Bubeck & Burton, 1989) at hypolimnetic chloride concentrations of 290 and 360–410 mg L−1, respectively.

Changing the density of lake water due to anthropogenic salt inputs can shift a normally dimictic lake (mixes twice a year) to monomictic (mixes once a year) or meromictic (never mixes). A unique example of disrupted mixing, took place in 191-m deep Lake Traunsee, Austria. When industrial saline discharge into the lake was stopped, the surface freshened while the bottom remained at ~ 170 mg-Cl L−1, and the lake became meromictic (Ficker et al., 2019). Reduced mixing can have serious consequences on lake biogeochemistry and lake habitat, as bottom waters become anoxic due to lack of oxygenation from surface water renewal (Ladwig, Hanson, et al., 2021; Straile et al., 2003).

Lake mixing is most likely to be impacted in the spring following high-salinity late-winter inflows from road salts flushing off the landscape. Formation of a halocline (salinity gradient) in the water column due to anthropogenic sources can be intensified by natural processes such as ion rejection from ice formation (Dugan et al., 2017b) and redox reactions at the sediment water interface (Cortés et al., 2017). It is unknown how pervasive of a problem delayed mixing is across lakes regionally, due to the difficulty in obtaining observational data during the narrow window of time between ice-off and spring mixing. In dimictic lakes, spring mixing (“overturn”) is the transition period from winter inverse stratification to summer stratification when the water column is experiencing full mixing. The start of this period is typically defined as when the water column becomes isothermal, that is, constant temperature with depth (Yang et al., 2020). This method is based on the assumption that density stratification in freshwater lakes is driven by temperature. However, if the lake water column has a salinity gradient, temperature measurements alone may be a misleading indicator of mixing dynamics. Take for example, high-frequency data from the epilimnion and hypolimnion of 20-m deep Lake Monona, Wisconsin, which show evidence for salinity-induced delayed spring mixing (Ladwig, Rock, et al., 2021). Following ice-off on March 20, 2020, the water column became isothermal at roughly 2.5°C. This temperature data taken alone would, erroneously, indicate the start of the spring overturn period. In reality, on March 20, the SpC of the hypolimnion was >1300 μS cm−1 compared to epilimnion at 600 μS cm−1, indicating stable stratification. For 20 days following ice off, the epilimnion warmed to 7°C, while the hypolimnion remained cold at 2.5°C. The density gradient induced by salinity and temperature, was not broken down until 20 days post ice-off on April 9, when it is clear from both the temperature (isothermal) and SpC (isohaline) data that the water column fully mixed (Figure 4). In this eutrophic lake, spring mixing is associated with reoxygenation of bottom waters, and a large pulse of inorganic nitrogen and phosphorus to the surface, which fuels phytoplankton production (Matsuzaki et al., 2020). Salinity-induced changes to spring mixing likely have cascading effects on the seasonal timing of trophic interactions (Sommer et al., 2012) in the epilimnion, beyond the direct effects of salinity on species composition, as well as on redox-reactions in the hypolimnion associated with prolonged anoxia.

Details are in the caption following the image
High frequency loggers beneath the lake ice in Lake Monona, Wisconsin, record surface (2 m) and bottom (20 m) water temperature and the gradient in specific conductance (SpC) between 2 and 20 m. The recorded data reveal a 20-day delay in water column mixing following ice-off. The period of lake ice is delineated by the vertical dashed lines. Data from Ladwig, Rock, et al. (2021).

Rivers and lakes also experience direct effects of high salt concentrations on geochemistry. Salinization has been found to impact trace metal concentrations, ion mobilization/retention and exchange rates, pH, and organic matter and nutrient cycling (see Berger et al., 2019; Kaushal et al., 2022). This complex set of geochemical interactions and their effects on freshwater ecosystems has been coined the freshwater salinization syndrome, and an in-depth review is provided by Kaushal et al. (2021). Riverine geochemical interactions have been better studied than lakes, likely due to higher salinities (Figure 3). Overall, there is a critical need to study the interactive geochemical responses to salinization in freshwater ecosystems.


In excess, salt is toxic to freshwater organisms. Freshwater organisms are osmoregulators, who maintain concentrations of organic and inorganic molecules within their bodies at higher concentrations than the external environment. This is achieved through a variety of mechanisms, such as actively transporting ions against concentration gradients or regulating water loss through osmoregulatory organs such as kidneys that produce dilute urine (Griffith, 2017). If environmental salinity increases, organisms must use energy to maintain osmotic balance, which can result in less energy for reproduction and growth. As salinity rises, osmotic stress can impair physiological function and result in death, even when environmental salinity remains below organismal solute concentrations (Kefford, 2019). Concentrations at which salinity becomes toxic to freshwater plants, animals, and microbes is species specific. Generally, larger vertebrates (i.e., fish) are more tolerant than invertebrates (Castillo et al., 2018).

Most biological studies on the impacts of road salt focus on chloride toxicity because this is the principal anion associated with de-icing salts. However, the toxicity of salt depends on the cation component as well, with NaCl and CaCl2 being less toxic than MgCl2, which is less toxic than KCl (Erickson et al., 2022; Mount et al., 1997). Because over 90% of road de-icers are NaCl, most studies have focused on Cl from NaCl. Here, we focus on NaCl, but also acknowledge the complexity of ecotoxicology and the limited scope of this review.

4.1 Impact of salinization on food webs

Chloride concentrations have been measured in excess of 10,000 mg L−1 in urban runoff and stormwater ponds, and over 1000 mg L−1 in urban rivers (Szklarek et al., 2022). Even in lakes, where chloride concentrations are still relatively low, present day concentrations can be 50–100x higher than historical concentrations (Figure 3c).

Relatively small increases in chloride in freshwater ecosystems (sometimes much less than 500 mg L−1) have been shown to cause decreases in population growth rate, reduced species richness, lower total community abundance, and impaired ecosystem functioning (e.g., grazing rates) (Gillis, 2011; Hébert et al., 2022; Hintz & Relyea, 2019; Szklarek et al., 2022). Consequently, salinization can impact the flow of energy and materials through food webs, resulting in varied effects. For example, a reduction in crustacean zooplankton abundance at 645 mg Cl L−1 compared to ambient concentrations, reduced competition, resulting in shorter time to and greater mass at metamorphosis in gray treefrog larvae (Van Meter et al., 2011). In contrast, the loss of zooplankton prey at high salinities resulted in decreased growth of spotted salamander larvae predators (Petranka & Francis, 2013). Reductions in abundance of herbivorous species can also cause increases in algae abundance because of lower community grazing rates (Hintz et al., 2017; Hintz, Arnott, et al., 2022). Salinization can lead to species shifts that impact ecosystem services such as drinking water quality and recreation. For example, algal communities can shift toward cyanobacterial dominance as salinity increases (Fournier et al., 2022; McClymont et al., 2022; McGowan et al., 2020; Pilkaitytë et al., 2004), likely because of variation in sensitivity among species (Tonk et al., 2007) and/or a loss of grazers (Tõnno et al., 2016). Although fish generally withstand higher concentrations of chloride and salt (>1000 mg L−1) before exhibiting behavioral and fitness changes (Leite et al., 2022; Zhou et al., 2022), there have been few investigations into how fish may be impacted by salinization-induced food web changes (Hintz et al., 2017).

4.2 Factors that modify impacts

Organism's sensitivity to salt is influenced by the local environment. For example, in lakes, zooplankton are among the most sensitive species and within this group, rotifers tend to be more tolerant than cladocerans and copepods, but relative sensitivities vary among studies (Hintz, Arnott, et al., 2022). A number of factors can influence sensitivity, including the quantity and quality of food (e.g., how much algae and what kind of algae). As the concentration of available food decreases (measured as carbon per liter), Daphnia experience mortality and reduced reproduction at lower concentrations of salt (both NaCl and CaCl2; Brown & Yan, 2015). The fatty acid composition of food can also influence toxicity (Isanta-Navarro et al., 2021). Water hardness (the amount of Ca and Mg in the water, expressed as CaCO3 concentration) can modify salt toxicity such that organisms in softwater are more sensitive to chloride than in hardwater (Elphick et al., 2011; Soucek et al., 2011). This is consistent with results that found that Daphnia and other cladocerans in Precambrian Shield lakes (that typically have low ion concentrations and low algae concentrations) are sensitive to chloride at concentrations below current Canadian water quality guidelines (Arnott et al., 2020). There is also some evidence that the local community composition, that is, species interactions, can influence macroinvertebrate and zooplankton species responses to increased salinity (Arnott et al., 2022; Bray et al., 2019).

4.3 Evolutionary responses to salinization

Lastly, most research has focused on the impact of increased salinity on freshwater organisms, and usually has ignored the role evolution might play in enabling organisms to adapt to rising salinity in freshwater environments. Will species evolve osmoregulatory strategies, such as decreasing the permeability of membranes to water, changing the expression and activity of ion transporters, or synthesizing osmolytes to maintain osmotic pressures (Lee et al., 2022)? History of exposure is a factor that may contribute to variation in salt toxicity. For example, Daphnia from waterbodies with higher salt concentrations tend to have higher tolerances to salt (Latta et al., 2012; Teschner, 1995). Evolutionary shifts in salt tolerance through time have been detected in populations of Ceriodaphnia dubia that were resurrected from Moon Lake sediments that were associated with different periods of drought and salinization (Elmarsafy et al., 2021), and from populations of the copepod Eurytemora affinis after they invaded freshwater lakes from saline estuaries (Lee, 2016). Some shifts in salt tolerance may be attributed to genomic changes related to osmoregulation. Studying a population of Daphnia pulicaria through a 25-year period of increasing salinity, Wersebe and Weider (2022) found the population evolved higher salinity tolerance over time. Genome sequencing revealed signatures of natural selection that were associated with genes related to osmoregulation, and found mutations in a gene that encodes a key chloride channel protein. Continued research is needed to understand the contributions of the identified mutations to widespread salinity tolerance.

Field experiments have suggested that adaption to salt can happen over a few generations, likely due to strong selection. Daphnia pulex that had been exposed to experimental salt treatments for 2.5 months, had higher survival (Coldsnow et al., 2017) and abundance (Hintz et al., 2019) when subsequently exposed to intermediate levels of salinity, compared to naive Daphnia with no previous elevated salt exposure. Although there is a massive need for research into eco-evolutionary dynamics, it is clear that organisms will vary widely in their ability to tolerate and adapt to changing salinity. However, many questions remain. It is unclear how selection for salt tolerant individuals will impact population responses to other stressors or how it will influence the population's trait distributions (Ribeiro & Lopes, 2013). For example, there is some evidence that Daphnia that have evolved increased tolerance to NaCl also have increased tolerance to CaCl2 and moderate levels of MgCl2. However, increased salt tolerance may have a trade-off; Daphnia that have adapted to high salinity have lower abundance under low salt conditions compared to Daphnia with no history of salt exposure (Hintz et al., 2019).


Recognizing the toxicity of salt, governments have set water quality guidelines for chloride; however, many of these are science-based water management criteria that reflect the maximum desirable surface water concentrations rather than enforceable regulations (Figure 5). In some countries, thresholds or guidelines exist to protect drinking water for humans. For instance, in North America and Europe, the drinking water quality guideline for chloride is 250 mg L−1, the concentration at which water starts to taste salty. In Canada, aquatic life is protected by acute exposure (short-term pulses of chloride) and chronic exposure (long-term, sustained exposure) water quality guidelines that were developed based on laboratory toxicity tests conducted on a variety of species including plants, invertebrates, amphibians, and fish. Effects endpoints are usually growth, reproduction, and survival, but can also include ecologically relevant behavioral and physiological traits. Because organisms vary in sensitivity, Canadian guidelines are derived from the 5th percentile of a species sensitivity distribution (SSD), which should in theory protect the majority of species (Figure 6). Current Canadian aquatic life guidelines are 120 mg Cl L−1 for chronic toxicity and 640 mg Cl L−1 for acute toxicity. The US Environmental Protection Agency (US EPA) water quality chronic criteria of 230 mg Cl L−1 is higher than Canada's guideline because it was based on the geometric mean of only three species: Daphnia pulex, rainbow trout, and the fathead minnow, which varied widely in their chloride tolerance (Elphick et al., 2011; US EPA, 1988).

Details are in the caption following the image
Chloride toxicity guidelines for Canada and the United States are all less than 1000 mg L 1 . For reference, the concentration of chloride in the global ocean is 19,400 mg L 1 .
Details are in the caption following the image
(a) Species sensitivity distribution (SSD) endpoints (including 24-h EC 50 , 48-h EC 50 , and 96-h LC 50 ) used to determine the freshwater short-term Canadian Water Quality Guideline for chloride. Species names associated with all points are given on the right axis. The fifth percentile of a log-normal model fit to the endpoints is 640 mg L 1 , as shown by the red dashed lines. (b) Endpoints (including 24-h EC 10 , 48-h IC 10 , 7- and 28-day EC 25 , 8- and 60-day No Observed Effect Concentration, 33-day LC 10 , and 8- to 28-day Maximum Acceptable Toxicant Concentration) used to determine the freshwater long-term Canadian Water Quality Guideline for chloride. The fifth percentile of a log-normal model fit to the endpoints is 120 mg L 1 , as shown by the red dashed lines. Reproduced from CCME (2011).

Toxicity tests used to inform water quality guidelines should meet a rigorous set of criteria (Canadian Council of Ministers of the Environment [CCME], 2011) that includes testing of media conditions to ensure adequate oxygen concentrations, and high performance of control animals. Despite this, there are limitations associated with these studies for setting guidelines. For example, the chronic guideline for chloride in Canada is based on 28 species, but this is a small subset of the species that live in freshwater habitats. Within a species, few populations are tested, which fails to capture geographical variation in sensitivity (Arnott et al., 2022), that could arise from acclimation, plasticity, evolutionary change, or environmental context. Toxicity tests are typically conducted under ideal temperature and food conditions, and often do not consider the influence of environmental context, such as the influence of potentially modifying ions, food quantity and quality, or the presence of other stressors, including predators (Hintz & Relyea, 2017; Liu & Steiner, 2017). Using chronic laboratory exposures, Arnott et al. (2020) found that six species of Daphnia experienced decreased reproduction and survival at chloride concentrations between 5 and 40 mg L−1; well below Canadian water quality guidelines for chronic exposure (120 mg L−1). The high sensitivity to chloride was likely associated with the softwater medium that was used for the bioassays, whereas most published toxicity tests for chloride have been conducted in medium to hardwater (see Elphick et al., 2011; Soucek et al., 2011). Furthermore, laboratory tests typically only assess direct effects of salt, despite evidence that species interactions can influence sensitivity to pollutants (Zhao et al., 2020). However, a recent review provided some evidence that an SSD based on individuals and populations offers as much protection as an SSD based on community-level experiments (Venâncio et al., 2022), indicating that laboratory-based toxicity tests may adequately inform water quality guidelines. This review, however, did not capture recent community-level field experiments that provide compelling evidence that current water quality guidelines do not adequately protect organisms in all lakes (Astorg et al., 2022; Greco et al., 2021; McClymont et al., 2022). A globally coordinated field experiment highlighted variation in community responses across a wide geographic region, with many communities being sensitive to salt at or below current water quality guidelines (Hébert et al., 2022; Hintz et al., 2022b). Water chemistry (ions and nutrients) explained some of the variation in chloride sensitivity for rotifers, but not for cladocerans or copepods. More research is needed to understand how environmental context influences chloride toxicity so we can ensure lakes in all regions are protected.


It is difficult to imagine a salt-free future for global freshwaters. It is easy for us to dissociate the millions of tons of salt dumped into the environment annually from the equivalent mass of salt rivers and lakes must bear as a result. To turn off the tap of salt requires a societal shift toward recognizing and prioritizing freshwater quality and eliminating our dilution mindset (Liboiron, 2021). We must act to lessen the salt load through targeted management approaches and reassessing the suitability of “one-size-fits-all” water quality criteria. Furthermore, road salt is retained in soils (as high as 40%–90%) and gradually released to streams and lakes over time (Oswald et al., 2019), which implies continued water quality impairment for future generations (Kelly et al., 2008). Only when we reduce or eliminate salt inputs (Schuler et al., 2019) will we be able to confront the legacy of salt pollution that has accumulated in the terrestrial landscape and be able to predict and protect the future of freshwater.


Hilary Dugan: Conceptualization (equal); data curation (lead); formal analysis (lead); visualization (lead); writing – original draft (equal). Shelley Arnott: Conceptualization (equal); writing – original draft (equal).


Thank you to our students and collaborators for their dedication to investigating freshwater salinization, and to two reviewers whose comments improved this manuscript. Cartoons in the graphical abstract were designed by Caitlin Bourbeau from the Wisconsin Applied Population Lab.


    This work was support by the US National Science Foundation (DEB-2144750) and the National Sciences and Engineering Research Council of Canada (RGPIN-2019-04315).


    The authors have declared no conflicts of interest for this article.


    The ecology and biodiversity of urban ponds

    River ecosystem conceptual models and non-perennial rivers: A critical review


    All data used in this manuscript include DOIs are from publicly available resources.