• Sinclair, James S.
  • Arnott, Shelley E.


1. Salt, primarily NaCl, is commonly used to de‐ice winter roads. The usage of road salt is steadily increasing as the area of urbanised land expands, and chloride concentrations in many lakes and flowing waters near roads have been steadily increasing through time. In some waterbodies, these concentrations are reaching levels that are toxic to freshwater organisms, such as zooplankton which are integral to the structure and function of freshwater communities.

2. We currently have a poor understanding of the impact of road salt contamination on zooplankton communities, and particularly the variation in response among lakes. Zooplankton communities differ in factors that could influence the impacts of elevated chloride, such as local species composition, exposure history and regional connectivity. Assessing and predicting how increasing road salt usage affects natural, freshwater zooplankton communities requires studies that investigate how communities vary in the impacts of elevated chloride, and in their response to dispersal.

3. We conducted a field mesocosm experiment that examined the effects of salt (NaCl) addition on two different zooplankton communities, which were also subjected to repeated introductions of zooplankton dispersers from the regional species pool. These treatments allowed us to determine whether communities differed in the impacts of salt, and how the effects of salt differed between communities that did and did not receive zooplankton dispersers.

4. We found variation in the effects of salt and dispersal between our experimental communities. For one lake community, salt drove zooplankton species composition towards dominance by littoral cladocerans. Zooplankton dispersal into this community ameliorated these impacts of salt by returning species composition to a state more similar to our “Control” communities. Conversely, for the other lake community, salt alone had minimal effects on zooplankton, while the combination of salt and dispersal led to declines in adult copepods.

5. Our experiment found that the impacts of salt varied between two zooplankton communities, causing a range of negative, positive and neutral community effects, and that dispersal from the regional species pool could serve as an ameliorating or exacerbating influence. Efforts to understand and predict the impacts of road salt on freshwater communities would therefore benefit from an evaluation of the potential role played by local (e.g., exposure history and biotic interactions) and regional (e.g., connectivity to possible sources of immigrants) context in the response of communities to elevated chloride.


The experiment was conducted 200 m from the shore of Opinicon Lake at the Queen’s University Biology Station, ON, Canada (44.566728 latitude, -76.325006 longitude), from 30 June 2015 until 31 August 2015. The experiment employed an unbalanced factorial design where we created two lake treatments (termed “Long” and “Loughborough”), for which we collected zooplankton from our two source lakes, and these experimental communities were crossed with a salt treatment (“Salt”: a single addition of NaCl; or “Control”: no addition of NaCl). The mesocosms that received salt additions were also crossed with a dispersal treatment (“Dispersal”: twicemonthly additions of zooplankton from the regional pool; or “No dispersal”: no twice-monthly zooplankton additions). This created three treatments for each of the “Long” and “Loughborough” experimental communities: “Control/No dispersal” (three replicates per lake), “Salt/ No dispersal” (five replicates per lake), and “Salt/Dispersal” (five replicates per lake), totalling 26 mesocosms. Higher replication was employed for the salt-treated mesocosms to better quantify the greater variability in community response that can occur in these treatments (e.g., Hintz et al., 2017; Symons & Arnott, 2013). We assessed variation in community response to elevated chloride and dispersal between the two lake communities by comparing phytoplankton and zooplankton among our “Control/No dispersal,” “Salt/ No dispersal” and “Salt/Dispersal” treatments for each lake treatment. A “Control/Dispersal” treatment for each lake was not included because the effects of simulated dispersal on mesocosm zooplankton communities have been well documented (e.g., Forbes & Chase, 2002; Howeth & Leibold, 2010; Shurin, 2000; Strecker & Arnott, 2010; Symons & Arnott, 2013; Thompson & Shurin, 2012). Additionally, our objective was not to compare the effects of dispersal between undisturbed and disturbed communities, but rather to examine variation in the effects of dispersal between two different communities experiencing the same disturbance.