Introduction
The contamination of freshwater habitats in temperate areas by road de-icing salts is now increasingly recognized as a widespread problem (reviewed by Cañedo-Argüelles et al., Reference Cañedo-Argüelles, Hawkins and Kefford2016; Hintz & Relyea, Reference Hintz and Relyea2019). Road salts applied in North America typically comprise ~98% sodium chloride (NaCl), with the remaining 2% represented by calcium chloride (CaCl2) and/or magnesium chloride (MgCl2) (Environment Canada, 2001; Sanzo & Hecnar, Reference Schell2006). Road salts usually reach aquatic environments by leaching from nearby roads during spring rains and snow melt that wash residue into ground and surface waters, as well as into soil (Lax & Peterson, Reference Lax and Peterson2009; Findlay & Kelly, Reference Findlay and Kelly2011). Because NaCl-saturated soils can exhibit slow release of ions over time, this can result in chronic salt introduction, which keeps salinity levels high through the summer months (Findlay & Kelly, Reference Findlay and Kelly2011). As road salt is applied to hard surfaces, water bodies in urbanized environments are more likely to display elevated salinity (reviewed by Tiwari & Rachlin, Reference Tiwari and Rachlin2018), and electric conductivity can be 20 times higher in water bodies near roads compared to forests (Karraker et al., Reference Karraker, Gibbs and Vonesh2008; Dananay et al., Reference Dananay, Krynak, Krynak and Benard2015). Conductivity reflects total dissolved solids, and is often used as a surrogate measure for dissolved salts (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017). The ratio of salinity to conductivity for use in conversion between these measures varies among freshwater bodies (Pawlowicz, Reference Pawlowicz2008), but conductivity values (in micromhos) were approximately double those of NaCl (in mg/l) in a field study of ponds in Ontario (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017). Thus, salinity levels in freshwater habitats vary spatially and temporally, and measured in different ways.
Chloride (Cl−) is often used as a proxy for road salt contamination in freshwater as ~60% of NaCl is chloride by weight, and it also more readily leaches (Lax & Peterson, Reference Lax and Peterson2009; Tiwari & Rachlin, Reference Tiwari and Rachlin2018). Correspondingly, Cl− concentrations in surface waters throughout the US have been steadily increasing for many years (Tiwari & Rachlin, Reference Tiwari and Rachlin2018). Most road-salt-contaminated freshwater systems surveyed to date have values of <200 mg Cl−/l; however, depending on the location, a range of 10–13,500 mg Cl−/l has been found in ponds and wetlands (Hintz & Relyea, Reference Hintz and Relyea2019). To put these values into context, Canadian guidelines suggest a limit of 250 mg Cl−/l in drinking water (Tiwari & Rachlin, Reference Tiwari and Rachlin2018), and chronic levels below 212.6 mg Cl−/l in natural habitats to avoid adverse effects on the vast majority of freshwater species (Elphick et al., Reference Elphick, Bergh and Bailey2011). In terms of NaCl as a whole, concentrations of 150 mg/l have been measured in rural lakes, with levels reaching as high as 4000 mg/l in ponds and wetlands, and 4300mg/l in streams (Sanzo & Hecnar, Reference Schell2006).
When considering exposure to environmentally realistic Cl− or NaCl concentrations, various effects on freshwater organisms have been reported. Many macroinvertebrates are relatively tolerant of moderate to high salinity levels, but taxa vary in their sensitivity (see reviews by Tiwari & Rachlin, Reference Tiwari and Rachlin2018; Hintz & Relyea, Reference Hintz and Relyea2019). For instance, stoneflies and black fly larvae can survive acute exposure to 1000–10,000 mg/l NaCl; however, this was lethal to the amphipods and caddisfly species tested (Blasius & Merritt, Reference Blasius and Merritt2002). Elevated salinity can also be detrimental to freshwater vertebrates such as larval fish and larval amphibians (Tiwari & Rachlin, Reference Tiwari and Rachlin2018; Hintz & Relyea, Reference Hintz and Relyea2019). Among these, rainbow trout fry exhibited reduced growth when exposed to 3000 mg Cl−/l (Hintz & Relyea, Reference Hintz and Relyea2017), and behavioural and developmental abnormalities occurred in wood frog tadpoles after exposure to 77.5 and 1030 mg/l NaCl, respectively (Sanzo & Hecnar, Reference Schell2006).
Beyond the direct effects of road salt exposure on freshwater animals, such as reduced growth or survival, potential indirect and sub-lethal impacts must also be taken into account. For example, elevated salinity can reduce biodiversity in freshwater habitats, but also shift ecological communities to favour salt-tolerant species (Bray et al., Reference Bray, Reich, Nichols, Kon Kam King, Mac Nally, Thompson, O'Reilly-Nugent and Kefford2018; Hintz & Relyea, Reference Hintz and Relyea2019). Another important sub-lethal effect to consider is whether road salt exposure can alter susceptibility to infection by pathogens or parasites, particularly given the significance of emerging infectious diseases for the conservation of freshwater species (Johnson & Paull, Reference Johnson and Paull2011; Thrush et al., Reference Thrush, Murray, Brun, Wallace and Peeler2011). The few studies to date that have examined this aspect generally indicate increased susceptibility to infection when potential hosts are exposed to environmentally relevant salt levels. Milotic et al. (Reference Milotic, Milotic and Koprivnikar2017) found that wood frog tadpoles exposed to 600 or 1050 mg NaCl/l not only had higher intensities of infection by trematode parasites, but also exhibited reduced anti-parasite behaviour in these conditions. Buss & Hua (Reference Buss and Hua2018) similarly reported that trematode parasitism in larval wood frogs increased by 68% at concentrations of 1000 mg/l of NaCl or more (note: NaCl concentrations throughout are converted to mg/l if originally reported in other units), and larval dragonflies exhibited a reduced immune response to simulated parasite challenge when subjected to chronic NaCl exposure at 3000 mg/l (Mangahas et al., Reference Mangahas, Murray and McCauley2019).
But in order to understand the net effects of road salt exposure on host–parasite dynamics, it is critical to consider how free-living parasite or pathogen infectious stages may be affected (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017; Buss & Hua, Reference Buss and Hua2018). Notably, aquatic infectious stages can be negatively affected by many environmental factors, from contaminants to pH, with consequences for their transmission success (reviewed by Pietrock & Marcogliese, Reference Pietrock and Marcogliese2003; Sures et al., Reference Sures, Nachev, Selbach and Marcogliese2017). For instance, Merrick & Searle (Reference Merrick and Searle2019) reported salinity-driven reductions in Daphnia fungal infections that were potentially mediated by reduced parasite infectivity, and high salt concentrations can effectively treat freshwater fish infected by the protozoan ectoparasite Ichthyophthirius multifiliis owing to the sensitivity of the latter (Miron et al., Reference Miron, Silva, Golombieski and Baldisserotto2003). However, the results of other studies regarding NaCl effects on freshwater parasites suggest wide variation. Eggs of the tapeworm Schistocephalus solidus developed normally in concentrations up to 12,500 mg NaCl/l (Simmonds & Barber, Reference Simmonds and Barber2016). In addition, cercariae of the trematode Petasiger nitidus had optimal survival at 2900 mg/l of NaCl (Shostak, Reference Shostak1993), and those of Bolbophorus confusus actually exhibited NaCl-dependent survival up to 2500 mg/l (Venable et al., Reference VanAcker, Lambert, Schmitz and Skelly2000).
Here, we examined survival and activity in cercariae of four freshwater trematodes (Ribeiroia ondatrae, Echinostoma sp., Cephalogonimus sp. and an unidentified strigeid-type) found in North America during exposure to environmentally realistic road salt concentrations. Of these trematodes, R. ondatrae, Echinostoma sp. and Cephalogonimus sp. typically use larval amphibians as second intermediate hosts (Schell, Reference Schell1985). Given the detrimental effects of many trematodes on amphibians (reviewed by Koprivnikar et al., Reference Koprivnikar, Marcogliese, Rohr, Orlofske, Raffel and Johnson2012), and that larval amphibian exposure to road salt increases their susceptibility to trematode infection (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017; Buss & Hua, Reference Buss and Hua2018), it is important to understand the extent to which this widespread contaminant also affects cercariae in order to predict net infection outcomes.
Materials and methods
Trematode acquisition
Aquatic snails (Helisoma sp.) were acquired from a pond in the Glenridge Quarry Naturalization Site in St Catherines, Ontario, where no road salt application occurs, as well as from another pond in Cambridge, Ontario, where measured NaCl concentrations did not exceed 150 mg/l in the spring or summer of 2015 (see Milotic et al., Reference Milotic, Milotic and Koprivnikar2017). Snails were then screened for trematode infection in the lab using standard methods (see Szuroczki & Richardson, Reference Szuroczki and Richardson M2009). Emerged cercariae were identified with standard keys (Schell, Reference Sanzo and Hecnar1970, Reference Schell1985) using a compound microscope. Owing to the similar morphotypes of certain cercariae, and the often-cryptic nature of closely related species, it was not always possible to identify cercariae to genus level; thus, we only do so here where appropriate. As such, four common trematode types were identified during our screening process and used in subsequent experiments: R. ondatrae, Echinostoma sp., Cephalogonimus sp. and an unidentified cercaria with a furcocercous morphotype (see Schell, Reference Sanzo and Hecnar1970, Reference Schell1985), which will subsequently be referred to as a ‘strigeid-type’ cercaria. Infected snails were kept separate in communal tubs filled with 5 l of dechlorinated water at a density of 10 snails/tub, but were also grouped by infection type. All were fed raw organic spinach ad libitum.
Experimental procedure
We used five salinity treatments with the following target concentrations: control (no salt added), 200, 400, 600 and 800 mg/l of NaCl. These reflected naturally occurring levels (50–560 mg/l NaCl) measured in 11 local ponds during the spring and summer of 2015 in which larval amphibians can be found (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017), and are consistent with mean values reported for North American ponds and wetlands (Environment Canada 2001; Daley et al., Reference Daley, Potter and McDowell2009; Todd & Kaltenecker, Reference Todd and Kaltenecker2012), although levels as high as 4300 mg/l have been reported within urban environments (e.g. Sanzo & Hecnar, Reference Schell2006). Importantly, wood frog tadpoles exposed to 600 or 1050 mg/l NaCl are more susceptible to infection by R. ondatrae (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017), and exposure to >1000 mg/l also increased infection by an echinostomatid trematode (Buss & Hua, Reference Buss and Hua2018). Treatment solutions were made by dissolving commercially available road salt (Windsor Safe-T-Salt® (Pointe Claire, Quebec, Canada)) into dechlorinated water, and the resulting salinity levels were measured with a VWR Traceable® salinity probe (VWR, Mississauga, Ontario, Canada). Thus, the actual concentrations of NaCl for each treatment were 160, 360, 560, 760 and 960 mg/ml of NaCl. As with similar studies (e.g. Hua et al., Reference Hua, Buss, Kim, Orlofske and Hoverman2016; Milotic et al., Reference Milotic, Milotic and Koprivnikar2019), we elected to use separate 96-well tissue culture plates for each of our treatments so as reduce potential cross-contamination. To avoid plate-level effects, the experiment for each cercaria species/type used 25 plates in total, with five plates allocated to each of the five treatments. Within each treatment plate, five consecutive wells were filled with the corresponding solution. Because an individual cercaria was pipetted into each of the five wells, N = 25 per treatment for each trematode species/type.
The experiments for each cercaria genus/type were carried out on separate consecutive days. Excluding R. ondatrae, cercariae were obtained in the morning of the experimental day by placing snails with the corresponding infection within Petri dishes of dechlorinated water under a lamp for 1 h. Emerged cercariae were then transferred into a common dish for mixing so as to randomize variation in their age and genotype. From this common dish, only actively swimming cercariae were transferred into single wells of the tissue culture plates that contained 1 ml of their designated treatment solution, and none obtained in this way were older than 3 h by the time the experiment commenced. As R. ondatrae cercariae emerge nocturnally (Hannon et al., Reference Hannon, Calhoun, Chadalawada and Johnson2018), snails harbouring this trematode were kept within Petri dishes overnight and cercariae were collected early the next morning (7 a.m.). However, this did introduce an element of age disparity between R. ondatrae and the other types of cercariae because the latter were no more than 1 h old before the experiment commenced while the former likely had a range in post-emergence age. This being said, many R. ondatrae cercariae were still active after 6 h of salt exposure (see Results), which suggests that they were <6 h old at the beginning of the experiment considering that they typically become inactive ~12 h post-emergence (Altman et al., Reference Altman, Paull, Johnson, Golembieski, Stephens, LaFonte and Raffel2016). For the experiment carried with each trematode genus/type, we filled plates with cercariae in random order, organized them in a corresponding manner on a lab bench and then systematically examined the plates at each time point in the same order in which they were filled. In this manner, the plate that first received cercariae was examined first, and so on, ensuring that exposure to a given treatment was the same for each cercaria, but also that this did not differ among the treatments (see Milotic et al., Reference Milotic, Milotic and Koprivnikar2019).
After the experiment commenced, the plates containing cercariae were examined every 2 h under a dissecting microscope for the first 8 h (i.e. at 2, 4, 6 and 8 h), and then at 24 h. At each of these time points, we recorded the activity (yes or no) and survival (dead or alive) of each cercaria. We considered cercariae as active if they were swimming and could potentially orient towards a possible host, and dead once they no longer responded to stimulus by a probe (see Hua et al., Reference Hua, Buss, Kim, Orlofske and Hoverman2016; Milotic et al., Reference Milotic, Milotic and Koprivnikar2019). These observation time points were selected because cercariae of most species do not live beyond 24 h (e.g. Shostak, Reference Shostak1993; McCarthy, Reference McCarthy1999), but also by considering the period during which they are infective to the next host in their life cycle – that is, the temporal period most critical and relevant for transmission. For instance, R. ondatrae cercariae live ~22 h (Lafonte et al., Reference LaFonte, Raffel, Monk and Johnson2015), but are usually not active beyond 12 h (Altman et al., Reference Altman, Paull, Johnson, Golembieski, Stephens, LaFonte and Raffel2016). For cercariae of species within the family Echinostomatidae, various studies have reported that Echinostoma trivolvis cercariae are maximally infective to hosts such as tadpoles within the first 8 h after emergence (e.g. Fried et al., Reference Fried, Pane and Reddy1997). Similarly, infectivity was found to peak at 3 h post-emergence for Hypoderaeum conoideum, and at 2.5 h for Euparyphium albuferensis, thereafter decreasing steadily with increasing cercariae age (Toledo et al., Reference Toledo, Munoz-Antoli, Pérez and Esteban1999). To the best of our knowledge, the period of maximum infectivity for Cephalogonimus sp. cercariae is unknown; however, their lifespan is generally >14 h (Calhoun et al., Reference Calhoun, Bucciarelli, Kats, Zimmer and Johnson2017). Diplostomum spathaceum and Schistosoma mansoni both belong to the order Strigeatida, as does the unidentified strigeid-type cercaria used here, showing declines in infectivity after 6 and 5 h post-emergence, respectively (Olivier, Reference Olivier1966; Karvonen et al., Reference Karvonen, Paukku, Valtonen and Hudson2003). Given that we exposed the cercariae used here to NaCl by the time they were ~3 h old, our focus on effects during the first 8 h likely corresponds to the period of maximal infectivity (3–11 h post-emergence) that is of greatest biological relevance.
Data analysis
We used a series of generalized linear mixed models with repeated measures to examine the activity and survival, respectively, of each cercaria at each time point for the four different trematode genera/types. However, as no cercariae were left alive after 24 h for R. ondatrae and Echinostoma sp., data for this time point were only retained for Cephalogonimus sp. and the strigeid-type cercaria. The dependent measures (yes or no for activity and survival, respectively) were given a binomial distribution with a logit link function. We used the fixed effects of treatment and time point (both categorical), as well as their interaction. The well and plate number for each cercaria were also entered as random categorical effects. Non-significant interactions and fixed effects in the overall model were dropped, and the analyses were re-run in a step-by-step manner until a final model was achieved, followed by Tukey's Honest Significant Difference post-hoc tests for significant main effects where applicable. All analyses were performed using SPSS 24.0(International Business Machines Corporation (IBM), New York, USA).
Results
Cercariae mortality
After model simplification, the final overall model that retained time for R. ondatrae was significant, with cercariae survival decreasing throughout the exposure period (F 3,472 = 7.647, P < 0.001; fig. 1), but exposure to salt had no significant effect (P = 0.197). The final model for Echinostoma sp. also retained time as a fixed effect, but the overall model was not significant (F 3,400 = 1.582, P = 0.193), nor was the effect of treatment (P = 0.812). Similarly, the final overall model for the strigeid-type cercariae retained time, but was not significant (F 4,550 = 0.604, P = 0.660), and salt exposure also had no effect (P = 0.786). In contrast, the survival of Cephalogonimus sp. cercariae was significantly affected by treatment (final overall model: F 4,580 = 3.432, P = 0.009), but not time (P = 0.093), with the post-hoc tests indicating specific differences in mortality between the following salt concentrations: 160 vs. 560 mg NaCl/L (P = 0.009), vs. 760 (P = 0.002) and vs. 960 (P = 0.012); and 360 vs. 760 mg/l (P = 0.025), with marginally insignificant differences vs. 560 (P = 0.071) and vs. 960 (P = 0.087). There was no significant interaction of treatment and time on survival for any of the four cercariae genera/types.
Fig. 1. Effects of NaCl exposure on the survival of cercariae representing four different species/types of freshwater trematode: (a) Ribeiroia ondatrae; (b) Echinostoma sp.; (c) strigeid-type; (d) Cephalogonimus sp. Survival is denoted by the total number of cercariae for each species/type that were alive at each time point (out of a maximum of 25) within five different NaCl treatments.
Cercariae activity
The overall model for R. ondatrae cercariae was significant (F 7,468 = 11.967, P < 0.001), including treatment (P = 0.048) and time (P < 0.001) because activity was reduced at later observation points, but also at the lowest and highest salt concentrations (fig. 2). Specifically, post-hoc tests indicated a difference in activity for R. ondatrae cercariae subjected to 160 vs. 560 mg/l of NaCl (P = 0.038), as well as between 360 and 760 mg/l (P = 0.024), and between 560 and 960 mg/l. For Echinostoma sp., the final model indicated a significant effect of time (F 3,400 = 3.334, P = 0.020), but not of salt treatment (P = 0.452). The final model for activity of the strigeid-type cercariae was significant (F 8,546 = 3.982, P < 0.001), and contained treatment (P < 0.001) and time (P = 0.011). Cercariae activity decreased over time, but was generally higher in the middle-range concentrations, with significant differences seen between the following treatments: 160 and 760 mg/l of NaCl (P = 0.005); 360 and 760 mg/l (P = 0.002); 560 and 760 mg/l (P = 0.028); and 760 and 960 mg/l (P = 0.014). The final model for Cephalogonimus sp. retained treatment as a fixed effect, but this was not significant (F 4,580 = 0.937, P = 0.442), nor was time (P = 0.653). There was no significant interaction of treatment and time on the activity of any genus/type of cercariae.
Fig. 2. Effects of NaCl exposure on the activity of cercariae representing four different species/types of freshwater trematode: (a) Ribeiroia ondatrae; (b) Echinostoma sp.; (c) strigeid-type; and (d) Cephalogonimus sp. Activity is denoted by the total number of cercariae for each species/type that were alive at each time point (out of a maximum of 25) within five different NaCl treatments.
Discussion
Road salt had very few detrimental effects on the four different genera/types of freshwater cercariae (R. ondatrae, Echinostoma sp., Cephalogonimus sp. and an unidentified strigeid-type) examined here. Not only were cercariae survival and activity relatively unharmed by environmentally realistic levels of NaCl, but cercariae of some genera/types actually performed best at the medium-range concentrations (360–760 mg/l). Importantly, some common second intermediate hosts of freshwater trematodes have been shown to exhibit increased susceptibility to infection when exposed to similar salt concentrations (e.g. Buss & Hua, Reference Buss and Hua2018; Mangahas et al., Reference Mangahas, Murray and McCauley2019), setting up a possible mismatch in host vs. parasite effects when road salts contaminate natural habitats.
This is particularly relevant for larval amphibians, with heavier infection by R. ondatrae seen in wood frogs exposed to 600 or 1050 mg/l of NaCl, and northern leopard frogs exposed to 1140 mg/l harbouring more Echinostoma sp. cysts (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017) – similar to the findings of Buss & Hua (Reference Buss and Hua2018) for wood frog tadpoles and an echinostomatid species. Given that salt exposure did not reduce the survival of R. ondatrae cercariae over a range of 160–960 mg NaCl/l, with their activity lowest at only the lowest and highest concentrations, this suggests that amphibian larvae subjected to NaCl levels ~500–700 mg/l will be disproportionately affected – that is, host susceptibility will increase while R. ondatrae cercariae are unharmed. This is even more likely for tadpole infection by echinostomatid trematodes as neither the survival nor activity of Echinostoma sp. cercariae was affected by salt treatment here. Whether road salt exposure affects susceptibility to infection by Cephalogonimus sp. in potential hosts such as larval amphibians (Schell, Reference Schell1985) is not known; however, salinity had no effect on the activity of these cercariae, and their survival was greatest in the 760 mg NaCl/l treatment. Our unidentified strigeid-type cercaria showed a similar pattern, but salinity did not affect survival, and cercariae in the 760 mg NaCl/l treatment exhibited the greatest activity. The order Strigeatida contains a diverse array of species using various hosts, including those responsible for avian and human schistosomiasis (reviewed by Colley et al., Reference Colley, Bustinduy, Secor and King2014; Horák et al., Reference Horák, Mikeš, Lichtenbergová, Skála, Soldánová and Brant2015). The results of the present study suggest that such cercariae can easily tolerate moderate salinization of freshwater habitats, and actually thrive in these conditions compared to water bodies with relatively low salt concentrations. However, it will be necessary to conduct further experiments and determine whether the infectivity of the cercariae studied here is affected by salt exposure.
Although there have been few studies examining how salinity affects the free-living infectious stages of freshwater parasites and pathogens, our findings are generally consistent with those for trematode cercariae in terms of tolerance for relatively high salt concentrations. For instance, Donnelly et al. (Reference Donnelly, Appleton and Schutte1984a) found that the survival of cercariae from three different species of Schistosoma was not affected until NaCl exceeded 5250 mg/l, but their infectivity was reduced above 1750 mg/l. Christensen et al. (Reference Christensen, Frandsen and Nansen1979) also reported that reductions in S. mansoni infectivity occurred at lower salt concentrations than those at which mortality increased. Similar to our findings of peak cercariae survival and activity at relatively moderate salinity levels for some trematode genera/types, the longevity of Schistosoma mattheei cercariae was greatest at 1750 mg NaCl/l (Donnelly et al., Reference Donnelly, Appleton and Schutte1984a). Other studies have also found that Schistosoma sp. cercariae show greater survival or activity in salt solutions compared to controls of relatively fresh water (e.g. Becker, Reference Becker1971; Asch, Reference Asch1975). Cercariae survival for B. confusus, another member of the order Strigeatida, similarly showed a positive association with salinity up to 2500 mg/l (Venable et al., Reference VanAcker, Lambert, Schmitz and Skelly2000). Experiments with another freshwater species in the family Echinostomatidae, P. nitidus, found that cercariae survived the longest at 2900 mg NaCl/l, with reductions beyond this concentration (Shostak, Reference Shostak1993).
Based on the results of our study and those of others with freshwater trematodes (e.g. Donnelly et al., Reference Donnelly, Appleton and Schutte1984a; Shostak, Reference Shostak1993; Venable et al., Reference VanAcker, Lambert, Schmitz and Skelly2000), cercariae seem to withstand NaCl concentrations (~360–2900 mg/l) that easily fall within the range reported from ponds and wetlands in North America (Sanzo & Hecnar, Reference Schell2006; Hintz & Relyea, Reference Hintz and Relyea2019). The ova and miracidia of trematodes also seem tolerant of relatively high (<2500 mg/l) NaCl levels (e.g. Donnelly et al., Reference Donnelly, Appleton and Schutte1984b), as are other parasitic flatworms. For instance, various species of freshwater monogenean show no apparent detrimental effects after exposure to concentrations as high as 6000 mg NaCl/l (e.g. Soleng & Bakke, Reference Soleng and Bakke1997). However, other parasites and pathogens found in freshwater habitats may be more sensitive. Catfish infected with the protozoan I. multifiliis showed improved survival after exposure to 2000 or 4000 mg NaCl/l as these concentrations inhibited parasite replication or survival (Miron et al., Reference Miron, Silva, Golombieski and Baldisserotto2003). Similarly, the relative sensitivity shown by the chytrid fungus Batrachochytrium dendrobatidis to conditions of 4000 mg NaCl/l compared to its amphibian host has been suggested as a means by which to mitigate this deadly infection (Stockwell et al., Reference Stockwell, Storrie, Pollard, Clulow and Mahony2015).
Considering that freshwater habitats typically have <500 mg/l of total dissolved salts (Battaglia, Reference Battaglia1959), it would seem puzzling that trematode cercariae not only tolerate higher concentrations, but often exhibit optimal survival and activity above this threshold – as shown here and by other studies (e.g. Donnelly et al., Reference Donnelly, Appleton and Schutte1984a; Venable et al., Reference VanAcker, Lambert, Schmitz and Skelly2000). A proposed explanation is that cercariae require less energy for their ionic and osmotic regulatory processes as the difference between external and internal salinity concentrations is reduced, allowing cercariae to devote more resources to movement and survival considering that they rely on endogenous energy reserves (Donnelly et al., Reference Donnelly, Appleton and Schutte1984a). However, we also note that cercariae may be expected to withstand salt concentrations that reflect those within the fluids of their hosts. For instance, the salinity of larval amphibian tissues is similar to that of typical amphibian Ringer's saline solution recipes (Funkhouser, Reference Funkhouser1977), which often call for 6600 mg/l of NaCl alone (e.g. Brandwein & Morholt, Reference Brandwein and Morholt1986).
Although road salt contamination is one factor contributing to the salinization of freshwater habitats, there are others (e.g. agriculture, resource extraction and land clearing) that make this a large-scale problem (Cañedo-Argüelles et al., Reference Cañedo-Argüelles, Hawkins and Kefford2016; Kaushal et al., Reference Kaushal, Likens, Pace, Utz, Haq, Gorman and Grese2018). The implications of this for freshwater community diversity and structure have been noted (reviewed by Tiwari & Rachlin, Reference Tiwari and Rachlin2018; Hintz & Relyea, Reference Hintz and Relyea2019), but it is also important to consider how salinization may affect host–parasite dynamics. For instance, in an earlier study (Milotic et al., Reference Milotic, Milotic and Koprivnikar2017), we found spring and summer NaCl concentrations of 50–560 mg/l in ponds known to contain larval amphibians with trematode infections, indicating that these hosts and helminths may be routinely affected by road salt contamination. Related to this, VanAcker et al. (Reference Venable, Gaudé and Klerks2019) found that suburbanization was positively related to amphibian trematode infection loads, as was site conductivity, although conductivity itself was not a significant predictor of parasitism. Given that other field studies have reported relationships between salinity and parasitism in freshwater systems (e.g. Rogowski & Stockwell, Reference Rogowski and Stockwell2006; Sanchez et al., Reference Sanchez, Coccia, Valdecasas, Boyero and Green2015), more investigations of this type are needed to extrapolate from the findings of laboratory studies. For trematode infections, it will also be necessary to consider how elevated salinity may affect the gastropod first intermediate hosts, as well as other second intermediate hosts considering the emphasis on larval amphibians to date. Given that different gastropod species vary in their responses to salinity (Hintz & Relyea, Reference Hintz and Relyea2019), this may promote the presence of some trematode species while reducing others. In addition, other road salt formulations should be studied in terms of host and parasite impacts, as NaCl, CaCl2 and MgCl2 can exert different effects (e.g. Hintz & Relyea, Reference Hintz and Relyea2017).
It is necessary to examine both hosts and parasites in order to understand the net effects of exposure to contaminants, including road salt. This is important if one party is relatively unaffected but the other is not, and may also vary across host–parasite systems. Here, we found that the cercariae of four different freshwater genera/types were highly tolerant of NaCl concentrations known to increase infection susceptibility in some of their hosts. Predicting net effects on host–parasite interactions is critical for many reasons. For instance, host pathology stemming from parasitism is often intensity-dependent, but infection may also interact negatively with the physiological stress of salinization in an additive or synergistic manner (e.g. Cespedes et al., Reference Céspedes, Valdecasas, Green and Sánchez2019). Given that the problem of freshwater habitat salinization is likely to worsen, with half of the rivers and streams in the US alone expected to have salt levels increase 50% by 2100 (Kaushal et al., Reference Kaushal, Likens, Pace, Utz, Haq, Gorman and Grese2018), understanding possible implications for infectious disease dynamics is important given that the latter in itself represents a substantial issue for the conservation of freshwater species (Johnson & Paull, Reference Johnson and Paull2011; Thrush et al., Reference Thrush, Murray, Brun, Wallace and Peeler2011).
Acknowledgements
We thank S.Y.S. Wang for her assistance in collecting snails.
Financial support
This work was supported by a NSERC Discovery grant to J.K. (grant number RGPIN-2015-05566).
Conflicts of interest
None.
Ethical Standards
This article does not contain any studies with human participants or vertebrate animals performed by any of the authors.
