The Results and Discussion will be considered in three sections including: (1) Physicochemical parameters; (2) Toxicity assessments; and (3) Land-use management. The physicochemical analyses section will include a discussion of results in three subsections covering the pH, TDS, and EC (Section 3.1.1); DOC, COD, and TSS (Section 3.1.2); and metals and PAHs (Section 3.1.3). The toxicity section presents the two toxicity assays R. subcapitata (Section 3.2.1) and V. fischeri (Section 3.2.2). The final land-use section will be used to discuss the impacts of land-use on the individual catchment area estimated pollutant loadings into the SSR. Figure 1 presents a map of the CoS and a summary of all physicochemical, metals, and PAH data in the form of box and whisker plots. Further details will be discussed in each of the following subsections. Results by event are included in Table 1 (see Table S5 for results by site; Tables S6-S7 for full SP and SM results).
Previous studies have collected snow by sampling SPs at different depths (Calonne et al. 2011; McLaren 1982), by coring, or by simple grab sampling of the top 20 cm of the SP (McLaren 1982) and SM runoff by scooping samples from puddles (Kuchta and Cessna 2009). However, studies sampling SPs are limited and there is no standard procedure for sample collection available. Due to the mixing of various roadside snows in removal trucks, their deposition on top of old packs, and potential-freeze thaw events causing partial or preferential elution of contaminants, it was assumed pollutants would not be uniformly distributed in the pile. Previous research on pollutant loading in snow packs found grid sampling to best minimize variations in mass load estimates (Vijayan et al. 2021), though this team used drills to minimize depth uncertainty, which was not performed in the current study.
3.1 Physicochemical Parameters
3.1.1 pH, Total Dissolved Solids (TDS), and Electrical Conductivity (EC)
The average pH for SP samples was 9.00 ± 0.54 and 8.07 ± 0.31 for SM samples (Table 1). No variation was observed across either SP or SM by site or event for pH with one exception (p ≤ 0.05): the March 7th, 2020 SP samples were significantly different from all 2019 samples. Generally, pH was significantly higher in SP samples than in SM samples (p ≤ 0.05). Curiously, USask SP samples possessed the lowest pH of all SP samples on April 2nd, 2019 (pH = 8.76) and March 7th, 2020 (pH = 7.56) (Table S6), with pH increasing over the 2019 season (which was not observed at other sites). Due to the low SP sampling size, more field seasons are needed to confirm if lower pH and melt trends are characteristic to the site. Compared to other sites sampled on the same dates, Central Ave. SM typically had lower pH values, while Wanuskewin Rd. SM had higher pH values (Table S7). For SP samples, pH had strong inverse correlation with TDS (R = -0.938) and EC (R = -0.937), though this was driven by TDS content in March 7th, 2020 SP samples. This correlation was not observed in SM samples with R = -0.243 and R = -0.247, respectively).
The EC and TDS are related parameters so will be discussed together herein. The average EC for SP and SM were 124 ± 131 µS/cm and 3,561 ± 4,017 µS/cm, respectively, while the average TDS were 59.4 ± 64.1 mg/L and 1,900 ± 2,290 mg/L respectively (Table 1). No site variation was observed for EC nor TDS for SP or SM (p ≤ 0.05). Event-wise, with respect to SP and similarly to pH, the March 7th, 2020, samples differed significantly from 2019 samples for both TDS and EC (p ≤ 0.05). With respect to SM, variation was observed between March 26th, 2020 and April 2nd, 2019 (TDS only), April 17th, 2020, April 23rd, 2020 (TDS only), and all samples after April 28th, 2020 (inclusive) (p ≤ 0.05). No significant differences were observed between any other events. Except the March 7th, 2020 SP Valley Rd. and USask samples, EC and TDS results for SP samples were 1-2 orders of magnitude lower than those of SM samples. Most SM samples display elevated EC and TDS relative to summer SW, with the March 26th, 2020, Valley Rd. sample having a TDS 7-fold and an EC 6-fold greater than those of 2019 SW (collected as part of a parallel study). Though no seasonal trends were observed during the 2019 season due to elution of TDS from the SPs, the TDS in 2020 SM samples remained elevated throughout the sampling season (Table 1), driven by individual site-specific contaminant spikes (Table S7), with a pulse of 1,721 mg/L TDS and 3,350 µS/cm observed at Valley Rd. as late as May 1st, 2020.
Measurements of chloride were performed on select samples (Table S8). Generally, chloride comprised 39-67% of TDS concentrations, except the April 28th, 2020, USask sample, where it only comprised 11% of TDS (the value returned to a set-typical value on May 5th ). Compared to summer SW chloride concentrations (141 ± 56 mg/L), concentrations on April 2nd, 2019 SM were similar (150 mg/L) but chloride was more abundant in less-melted March 7th, 2020 samples. Mirroring TDS and EC concentrations, USask SP contained more slightly more chloride than SM, while at Valley Rd. significantly more chloride was observed in SM. Making this observation within chloride as opposed to TDS is interesting: 62% of the TDS burden in the Valley Rd. SM sample is chloride (46% in SP), while this proportion is only 41% in USask SM (47% in SP). The sites are evidently within different stages of preferential elution early into the sampling season, likely due to the unique impervious surface at the Valley Rd. site. Throughout the 2020 spring melt the largest three pulses of chloride are observed at this site (2,600-5,800 mg/L), with the next largest three pulses observed at Central Ave (1,100-1,700 mg/L). As chloride encourages dissolved partitioning of metals, correlations between chloride and metals concentrations were examined. There is some correlation between the presence of chloride and dissolved concentrations of arsenic (R = 0.594), cadmium (R = 0.489), manganese (R = 0.684), nickel (R = 0.658), and strontium (R = 0.482) in tested samples, the implications of which are further discussed in Section 3.2.
A two-week melt period was observed prior to sampling in April 2019, which was not observed in March 2020. Additionally, the seasonal sampling events on April 2nd, 2019, and March 7th, 2020, took place on cooler days relative to other events. Both sampling conditions may explain the relative similarity in pH (as temperature governs the chemical composition of the SW matrix), but differences in EC and TDS. As dissolved contaminants in snow are preferentially eluted with meltwater out of SPs (Viklander 1996), it is expected that EC and TDS would be elevated in SM relative to SPs. In cold climates, these parameters are a major concern for toxic impact due to the practice of road salting (Exall et al. 2011b; Mayer et al. 2011a; Murphy et al. 2014). Dissolved salts, especially chloride, affect the partitioning of metals, increasing those in the dissolved phase (Bäckström et al. 2004; Galfi et al. 2017; Mayer et al. 2011b; Reinosdotter and Viklander 2007). A previous study conducted over a period of 2 years found significantly elevated EC, TSS, TOC, copper, and zinc in melt season samples compared to rain-season samples (Helmreich et al. 2010). However, the chemodynamics of snow are complex. For example, when sand is used as an anti-slip agent (often in combination with NaCl or MgCl2, CaCl2 salt), SM pH and TSS are elevated. This may counteract the dissolving effect of chloride by providing more particulate surfaces for contaminants to bind (e.g., metals, PAHs). Some of the highest SM pH values in this study occurred in Valley Rd. samples, the impermeable surface of which would retain TDS and TSS in SM runoff.
3.1.2 Dissolved Organic Carbon (DOC), Chemical Oxygen Demand (COD), and Total Suspended Solids (TSS)
Despite the presence of geese contributing organic waste onsite throughout the melt season (personal observation of H.P.) at the USask and Central Ave. sites, no site-wise significant variation in DOC nor COD was observed for either SP or SM (p ≤ 0.05). The average DOC of SM (6.93 ± 6.25 mg/L) approximately double than that of SP (3.12 ± 2.73 mg/L). Average COD did not follow a similar trend, with SP COD (460 mg/L) slightly higher than SM COD (353 mg/L). There was significant variation in COD obtained from both pile (SD ± 230 mg/L) and melt (SD ± 293 mg/L) samples. Unsurprisingly, there was little correlation between COD and DOC concentrations in SP (R = -0.068); conversely, COD and DOC were correlated in SM (R = 0.668). The highest DOC values in SP occurred in USask samples (4.45 ± 1.75 mg/L; Table S5) for all 2019 events. In SM, the highest DOC values occurred in Valley Rd. samples on March 7th, 2020 (20.4 mg/L) and March 26th, 2020 (27.4 mg/L). The highest COD values were measured in April 13th, 2019, pile samples (684 ± 7.4 mg/L) and March 26th, 2020, Valley Rd. sample (645 mg/L ± 301 mg/L). For SP, event-wise differences in DOC were observed between March 7th, 2020 and April 13th, 18th, and 24th 2019 (p ≤ 0.05); the sole differences in event-wise COD were observed between April 13th, 2019 and all other SP samples (p ≤ 0.05). For SM, one event-specific variation in DOC occurred between March 7th, 2020 and May 12th, 2020 (p ≤ 0.05); only the March 26th, 2020 event otherwise differed from April 2nd, 2019, April 10th, April 17th, April 23rd, April 28th, May 5th, and May 12th, 2020 samples (p ≤ 0.05). Only the March 26th, 2020 and May 12th, 2020 samples had significantly different COD from one another. The concentrations of both DOC and COD fluctuated between sites and sampling events for both SP and SM with no clear seasonal trend, though no COD values exceeding 300 mg/L were measured in SM samples after April 23rd, 2020.
The overall range of TSS varied from 31 to 2,982 mg/L in SM and 506 to 3,513 mg/L in SP. Generally, SP samples contained more TSS than SM samples, with an average TSS (1,662 ± 750 mg/L) four times that of SM (400 ± 387 mg/L) (Table 1). It is not unexpected that coarse TSS remain within the SP as opposed to becoming entrained in SM runoff (Westerlund and Viklander 2011). No significant difference related to site nor event was observed for any samples. The highest TSS was observed on April 13th, 2019 SP. Elevated TSS values in SM were observed on April 17th, 2020 (USask, 1,800 mg/L), and April 28th, 2020 (Central Ave, 2,982 mg/L), coinciding with warm days in which higher flow more capable of transporting more, and potentially larger, particles was expected. Interestingly, the SM trend in TSS concentration did not decrease as the spring season proceeded. Though COD and DOC were correlated in SM and not SP, this observation was reversed with respect to TSS, with COD and TSS correlated in SP (R = 0.593) but not SM (R = 0.143).
Larger, less chemically reactive particles remained within the SP, which is consistent with previous findings that large particles are deposited onsite and not transported with meltwater (Viklander 1999). The TSS likely experienced delayed releases from the SP after extended periods of warming melted the surface of the SP and transported those particles through the pack. This would also explain the relationship between TDS and TSS values at this site, where peaks in TSS concentration were observed in the days following TDS peaks (March 26th -March 29th, April 17th -April 20th, April 23rd -April 28th, all 2020 season). Relative to the preceding three sampling dates, elevated TSS was observed at Valley Rd. through April 28th, May 1st, and May 5th, 2020, though the concentration had dropped almost half by May 12th, 2020. Similar TSS behaviour was observed by Westerlund and Viklander (2011), who noted that over 90% of the metals burden in SPs was in particulate form and followed the same trends as TSS. It is possible the high TSS contained particulate COD, which has been previously observed in snow samples (Westerlund and Viklander 2011). Moreover, particulate COD tends to dissolve in meltwater (Viklander 1996), and some partial dissolution may explain the slightly higher variability in SM (SD ± 101 mg/L) relative to snow (SD ± 79 mg/L).
3.1.3 Metals and Polycyclic Aromatic Hydrocarbons (PAHs)
The average sum of measured metal species was 559 ± 187 µg/L in SP and 725 ± 412 µg/L in SM (Figure 2 with concentrations of individual elements included in Tables S9-S10). Concentrations of dissolved aluminum, iron, lead, and zinc were generally higher in SP samples, though SP concentrations were comparable to those in summer SW obtained as part of a parallel study (Popick et al. 2021). The largest overall contaminant peaks occurred in Valley Rd. SM on March 26th, 2020 (1,455 µg/L), April 17th, 2020 (1,988 µg/L) and April 23rd, 2020 (1,654 µg/L). All three of these contaminant peaks were accompanied by discharges of manganese measuring 575-919 µg/L. Other species-specific pulses in SP included aluminum at 401 µg/L (Valley Rd., April 13th, 2019;), arsenic at 26.2 µg/L (Wanuskewin Rd., April 18th, 2019), copper at 135 µg/L (USask, April 2nd, 2019), lead at 7.50 µg/L (USask, April 24th, 2019), and selenium at 15.8 µg/L (USask, April 2nd, 2019). In SM, species-specific pulses include mercury at 0.18 µg/L (Central Ave, May 5th, 2020) and zinc at 581 µg/L (Valley Rd, March 29th, 2020). In SP and SM samples from April 2nd, 2019 and March 7th, 2020, dissolved metal concentrations in SM were similar, but March 7th, 2020 SP concentrations were approximately 67% of those observed in April 2nd, 2019 SP.
While metals are naturally occurring geogenic elements, with their presence in sediments and soils predating industrialization (Owca et al. 2020), their mobilization in the urban environment can be markedly increased by human activities (Soto et al. 2011; Galfi et al. 2017; Sakson, et al. 2018). Metals may be present in SM in either dissolved or particulate forms with the dissolved metals being considered as bioavailable, and potentially toxic, to aquatic organisms. Chloride from winter road maintenance is a significant contributor to dissolved inorganic ligands in solution, thus enhancing the complexation of metals and increasing their dissolved concentration (Mayer et al. 2011a; Valtanen et al. 2014). Metals of particular concern due to their toxicity in SW, especially in cold climates, include lead, manganese, chromium, zinc, nickel, copper, and cadmium. At lower tropic levels, these metals can cause DNA damage, growth and biomass reduction, cellular respiration and enzyme dysfunction, inhibition of photosynthesis, and nervous system damage, among other adverse effects (Jaishankar et al. 2014; Zubala et al. 2017). In humans, arsenic, copper, lead, and zinc are associated with health effects including cancer, bone disease, hypertension, DNA and enzymatic dysfunction, nervous system damage, infertility, and organ failure (Chung et al. 2014; Ma et al. 2016; Zubala et al. 2017).
The average of measured dissolved ΣPAHs were 77 ± 157 µg/L in SP and 0.587 ± 0.962 µg/L in SM (Figure 3 with individual species in Table S11). Generally, dissolved PAHs in SP were 2 orders of magnitude greater than those detected in SM. With the exception of the Valley Rd. April 24th, 2019 sample, at least 20 µg/L ΣPAHs were detected in all SP samples. SM samples typically contained < 2 µg/L ΣPAHs, with the exception of USask April 23rd, 2020 at 4.82 µg/L and Valley Rd. March 29th, 2020 at 1.14 µg/L. The most notable PAH discharge of 565 µg/L occurred in SP samples from Wanuskewin Rd. on April 13th, 2019, including notable discharges of pyrene, phenanthrene, and anthracene, while the largest pulses in SM occurred in Valley Rd. March 26th, 2020 (1.14 µg/L) and USask April 23rd, 2020 (4.82 µg/L) samples. Nearly all SP samples exceeded the CCME guidelines for fluorene, benzo[a]pyrene, pyrene, phenanthrene, and anthracene, while 50% of SM samples exceeded guideline thresholds for benzo[a]pyrene, pyrene and all except one exceeded the threshold for anthracene.
High concentrations of PAHs in SW are often correlated with high concentrations of heavy metals, indicating that roadways are also a major source of PAHs, mainly as vehicle exhaust, diesel soot, or from vehicle and pavement degradation (Aryal et al. 2010; Legret and Pagotto 1999; Moghadas et al. 2015). Considering that the Wanuskewin Rd. April 13th 2019 sample contained comparable DOC and COD to other April 13th, 2019 samples, its PAHs are likely traffic-derived. Sampling at this site did not disproportionately target contaminated snow relative to other sites or events; it is possible that oil-enriched particles were entrained in the sample and repartitioned into the dissolved phase when melted; however, previous observations of the same suite of PAHs noted 88-95% of species remained particle-bound in SW samples (Mayer et al. 2011b; Vijayan et al. 2019). This would be consistent with observations of preferential elution, in which the heavier and less polar compounds are the last to leave the SP. The April 13th, 2019 Wanuskewin Rd. sample may be the result of selecting a localized patch of pollution within the SP; even if this magnitude of pollution only occurred at one location onsite, the effluents may still be highly concentrated even after mixing. A pollution peak may also have been occurring at this site alone on the sampling day. The phenanthrene, pyrene, and anthracene found in this sample are likely compounds derived from coal tar, roadway degradation, and vehicle emissions which contribute the bulk of pollution in roadside snow banks (Kuoppamäki et al. 2014; Moghadas et al. 2015; Vijayan et al. 2019; Viklander 1999).
3.2 Toxicology Assessments
3.2.1 R. subcapitata Bioassay
Significant toxicity was generally not observed in either SP or SM samples, with the calculation of an EC10 necessary to quantify toxic response for many samples (Table 2). It is assumed that chloride might have played a significant role in the osmotic stress experienced by R. subcapitata. While all samples inhibited growth at 100% concentration, few inhibited growth rates more than 50%. Furthermore, all serial dilutions below and including 50% often displayed stimulation instead of inhibition relative to the negative control. Toxicity did not seem related to any site-specific sample characteristics, though toxic responses were observed for all tested April 17th, 2020 samples. The most toxic response for which a 95% CI could be calculated was the USask SM sample on this date (EC50 at 43.7 percent sample concentration; 95% CI 39.5-48.2). This sample most remarkably contains the highest dissolved copper discharge of all SM samples (Table S10). The April 17th, 2020 Central Ave. sample generated the second-lowest EC50 threshold (54.3 percent sample concentration) but no CI could be calculated. This sample is not particularly remarkable compared to inter-event samples (the Valley Rd. site has higher TDS and chloride readings), inter-site samples (the April 23rd, 2020 Central Ave. sample has similar quality measurements but no detected EC50), or the entire dataset (no contaminant spikes were measured in this sample). However, it does possess the greatest dissolved zinc concentration of any April 17th, 2020 sample (119 µg/L). These toxicity results seem to agree with previous observations that idiosyncratic sample mixtures are likely providing additive, synergistic, and antagonistic effects which cannot be attributed to a single species (Fulladosa et al. 2005; Tsiridis et al. 2006).
Table 2
72-hour growth inhibition EC50 and EC10 for R. subcapitata and 15-minute luminescence inhibition for V. fischeri (Microtox). Results are presented as percent of total sample concentration required to read the endpoint. EC50 and EC10 values were generated using GraphPad Prism as a dose-response curve (see Methods) of, in the case of R. subcapitata, the initial (t = 0) fluorescence observed after 72 hours, and in the case of V. fischeri, the initial (t = 0) luminescence observed after 15 minutes. EC10 values were interpolated from relating 10% luminescence inhibition to sample concentration on a standardized curve (see Methods).
Year
|
Sample Date
|
Site Name
|
Algae
|
MicroTox
|
EC50 (CI 95%)
|
EC10 (CI 95%)
|
EC50−15 min (CI 95%)
|
EC10−15 min (CI 95%)
|
2019
|
April 2nd
|
USask (Snow)
|
103.4 (96.9-111.6)
|
28.1 (31.9-24.6)
|
54.0 (27.3-99.2)
|
53.4 (30.3-97.0)
|
USask (SM)
|
ND
|
95.6 (50.3-NC)
|
54.4 (NC)
|
52.1 (29.0-96.7)
|
April 18th
|
Valley Rd.
|
ND
|
36.8 (29.8-44.2)
|
56.5 (NC-68.3)
|
45.4 (33.7-49.6)
|
Central Ave.
|
ND
|
62.0 (51.4-73.5)
|
43.6 (34.2-54.3)
|
36.0 (26.3-45.3)
|
Wanuskewin Rd.
|
99.2 (94.8-104.1)
|
94.0 (66.8-NC)
|
55.1 (NC)
|
50.8 (36.1-79.7)
|
April 24th
|
Valley Rd.
|
ND
|
NC
|
52.6 (NC)
|
49.9 (27.6-95.4)
|
USask
|
-
|
-
|
55.2 (42.4-72.0)
|
41.7 (28.4-70.5)
|
2020
|
March 7th
|
Valley Rd. (Snow)
|
-
|
-
|
56.9 (NC)
|
54.9 (50.0-87.0)
|
Valley Rd. (SM)
|
ND
|
33.5 (23.6-44.3)
|
59.5 (NC)
|
57.4 (44.3-91.8)
|
USask (Snow)
|
ND
|
32.8 (24.1-42.2)
|
70.7 (NC)
|
71.1 (50.2-NC)
|
USask (SM)
|
43.7 (39.5-48.2)
|
18.4 (14.9-22.7)
|
54.2 (NC)
|
52.0 (33.4-86.9)
|
March 26th
|
Valley Rd.
|
ND
|
50.9 (35.7-65.3)
|
45.8 (35.8-56.8)
|
NC
|
April 17th
|
Valley Rd.
|
123.2 (99.6-177.3)
|
38.7 (24.4-55.8)
|
91.2 (7.28-NC)
|
NC
|
USask
|
96.4 (NC)
|
84.1 (49.9-158.3)
|
58.4 (NC)
|
55.4 (39.7-89.4)
|
Central Ave.
|
54.3 (NC)
|
49.4 (28.9-93.2)
|
52.6 (49.8-NC)
|
50.8 (27.8-96.7)
|
Wanuskewin Rd.
|
-
|
-
|
31.3 (21.7-44.2)
|
34.0 (23.9-50.8)
|
April 23rd
|
Valley Rd.
|
-
|
-
|
ND
|
NC
|
Central Ave.
|
ND
|
NC
|
1.19 (NC)
|
NC
|
April 28th
|
Valley Rd.
|
-
|
-
|
0.062 (NC)
|
NC
|
USask
|
-
|
-
|
NC
|
NC
|
May 1st
|
Valley Rd.
|
-
|
-
|
58.4 (NC)
|
60.5 (50.2-96.9)
|
May 5th
|
Central Ave.
|
-
|
-
|
99.9 (18.5-NC)
|
NC
|
May 12th
|
Valley Rd.
|
ND
|
90.7 (69.0-99.8)
|
56.8 (39.1-77.8)
|
45.9 (32.0-81.6)
|
ND – not observable toxicity detected at the respective threshold; NC – not calculated. |
Organic matter and TSS have been linked to acute toxicity as bound PAHs and metals cause mortality at acutely toxic concentrations and can cause adverse, potentially cumulative, effects at chronic concentrations (Barbosa et al. 2012; Glen and Sansalone 2002; Rossi et al. 2013). Concentrations of PAHs exceeding CCME guideline thresholds (see above) did not contribute to observable toxicity in SP over SM samples. This observation is not entirely unexpected, as Bragin et al. (2016) found PAHs did not contribute significant growth inhibition in R. subcapitata. Despite copper and zinc commonly causing toxicity in aquatic organisms (Babich and Stotzky 1978; Mayer et al. 2011b) alongside notable concentrations in the most toxic samples, no correlation between copper or zinc concentrations and algae toxicity was observed in this study (R = 0.090 and R = 0.263, respectively). As R. subcapitata can acclimate to aquatic concentrations between 0.5-100 µg Cu/L (Bossuyt and Janssen 2005), local algae species may accommodate higher geogenic background levels. Brix et al. (2010) identified relatively low zinc toxicity risk for brief, one-hour events and relatively significant toxicity for chronic exposures, which could have toxicity implications for spring melt runoff from galvanized SW pipes. In cold climates, road salts can contribute significantly to toxicity and affect metal toxicity in winter and spring runoff. In a study examining the toxicity of winter highway runoff, undiluted samples containing high road salt observed no survival in D. magna, V. fischeri, Ceriodaphnia dubia, or Oncorhynchus mykiss (Mayer et al. 2011a). When testing the impact of the three road salts on rainbow trout, Hintz and Relyea (2017) found CaCl2 was the most harmful to fish growth, followed by NaCl and then MgCl2. Conversely, emphasizing the intraspecies variability of toxicity, Hopkins et al. (2013) found MgCl2 and NaCl caused comparable developmental deformities in rough-skinned newts.
3.3.2 V. fischeri Bioassay
Similar to R. subcapitata, displayed some degree of luminescence inhibition in samples at 100% concentration (Table 2). However, inhibition was not detected in all samples and stimulation relative to the negative control was often observed in serial dilutions. No remarkable site- or event-related trends in toxicity were observed, though the survival rates in diluted sample wells created difficulties in calculating IC50 with 95% CI for many samples. The April 23rd, 2020 Valley Rd. sample displayed the highest toxic potential (15-min IC50 = 8.3% sample concentration); this pulse notably had the highest concentration of dissolved manganese (919 µg/L) and the second-highest concentrations of dissolved mercury (0.15 µg/L) and chloride (2,900 mg/L).
Fulladosa et al. (2005) found metals to be toxic to V. fischeri in the order of mercury > silver > copper > zinc, while cobalt, cadmium, chromium(VI), arsenic(V), and arsenic(III) showed no significant toxicity. However, in this study, as with algae, no significant correlations were noted between dissolved metals and toxic response. Moderate correlations were found between toxicity and chromium (R = 0.453), manganese (R = 0.480), nickel (R = 0.479), and strontium (R = 0.354). In testing binary mixtures of metals, Fulladosa et al. (2005) found copper-zinc mixtures to be additive, while Tsiridis et al. (2006) observed a synergistic effect. Interestingly, the latter comments that all copper-zinc IC50 results differed significantly from theoretical predictions; when combined with mixtures of humic acids, the toxicity of the solution decreased relative to the metals-only mixture. Interestingly, most of the aforementioned metal species were found to be significantly correlated with chloride concentrations (see Section 3.1). The complexity of sample mixture is potentially reflected in these results. Chloride concentrations have been implicated as the dominant driver in SM toxicity relative to metals and PAHs (Mayer et al. 2011a; Prosser et al. 2017), they appear to have increased the influence of select metals on toxicity, as previously observed (Bäckström et al. 2004; Reinosdotter and Viklander 2007). However, despite the impact which would be expected from, for example, the Valley Rd. March 26th, 2020 (5,800 mg/L chloride) or the April 23rd, 2020 sample (2,900 mg/L chloride, 919 µg/L manganese; 0.15 µg/L mercury), none was observed. That antagonistic mechanisms against chloride or metal toxicity seem apparent in samples is most easily indicated by comparing R. subcapitata and V. fischeri data: EC50 thresholds for algae often occurred at greater sample concentrations, though algae should be more susceptible to salt toxicity relative to V. fischeri (Cook et al. 2000). Previous studies which found significant salt toxicity sampled winter road runoff in the Ontario (Mayer et al. 2011a; Prosser et al. 2017). Though carbonate-rich soils are local to the Saskatoon area, with the CoS reporting a river water hardness of 191 mg CaCO3/L on the municipal website, Prosser et al. found 3,159 mg/L of chloride impacted 50% mortality in freshwater mussels (Lampsilis siliquoidea) at a water hardness of 237 mg CaCO3/L (Prosser et al. 2017). As local water hardness therefore would explain the lack of toxic chloride impact in this study, it is probable that the aging of SPs and its onsite SM puddles alter their toxic potential relative to fresh road runoff.
3.4 Land use management
To estimate the quantity of snow at CoS facilities prior to the melt season, UAS-LIDAR imagery of the Valley Rd. facility was obtained on March 13th, 2020 with a snow estimate of 151,910 m3. Assuming the Valley Rd. facility receives half of the CoS’s plowed snow (pers. comm.), a city-wide facility volume of approximately 300,000 m3 was obtained. Using a sample taken from Valley Rd., a SP density of approximately 520 kg/m3 was determined which was then used in conjunction with a water density of 1,000 kg/m3 to produce an estimate of 79,000 m3 in SWE for the Valley Rd. site and 34,500 m3 at each of CoS other facilities. Similarly, the USask site snow volume was estimated and the SWE determined to be 10,000 m3. Assuming these volumes were discharged to the environment through the melt season, the 2020 seasonal loading estimate was calculated using the average measured concentration of TDS, TSS, COD, DOC, copper, chromium, manganese, nickel, lead, selenium, zinc, and ΣPAHs (Table 3). It should be noted that this loading is representative of CoS traffic-related surfaces, however, SM runoff within catchments is likely to follow the same transport paths as SW. However, contaminant concentrations in this non-road snow were not determined in the current study. As 35-85% of SM has been observed to infiltrate on the Canadian Prairies, even over frozen soils (Mohammed et al. 2019), it is not unreasonable to assume roadside SPs (which would discharge directly to storm sewers if not removed) account for a significant portion of urban contamination in snow-derived runoff.
Given it is the largest snow storage facility, as expected the Valley Rd. site contributed the highest calculated loadings for all parameters shown in Table 3. This includes exceptionally high values for TDS (169,231 kg), COD (24,524 kg), and DOC (611 kg). Interestingly, the TSS loadings for the three CoS sites were all in a similar range (30,000 to 40,000 kg) despite having significant snow volume differences. However, this could be attributed to soil erosion on the smaller sites given the Valley Rd. site is the only facility contained on a concrete surface. This is also likely the reason whey the TDS loading was very high at the Valley Rd. facility as there is no opportunity for infiltration as available at other sites. The TDS and COD were strongly correlated (R = 0.962) as were the COD and DOC (R = 0.966). Organics were not significantly present in snow with DOC comprising 0.36-1.03% of TDS loading. Two sites contributed 89% of DOC loading: Valley Rd. (611 kg) and Central Ave. (199 kg). Strong correlations were also observed between COD loading and copper (R = 0.918), chromium (R = 0.980), nickel (R = 0.973), zinc (R = 0.958), and ΣPAHs (R = 0.953) along with a moderate correlation in lead (R = 0.948). The largest metal loading estimates were of manganese (25 kg) and zinc (12 kg) driven by single-event pulses observed at Valley Rd. The next highest loading estimates were copper (2.25 kg with 52% comprising Valley Rd.) and nickel (1.38 kg with 64% comprising Valley Rd.). Loading estimates for copper, manganese, selenium, and ΣPAHs were not found to be significantly correlated to snow volume, though correlations were observed for chromium (R = 0.974), nickel (R = 0.945), lead (R = 0.921), and zinc (R = 0.950). However, each site’s characteristics are quite variable making direct comparisons difficult and potentially erroneous. For example, the TDS, COD, and DOC values at the Central Ave. site were markedly higher than the Wanuskewin Rd. despite each have the same estimated SWE volume.
During the winter in cold-region Canadian cities, precipitation falls as snow and runoff events do not occur for several months, though winter activities continue in the catchment. After a snowfall event, municipalities will plow snow from the roads into roadside banks and apply grit material (road salt or sand). Within hours of deposition, these snowbanks become contaminant sinks for heavy metals and particulates from traffic and winter maintenance activities, as well as transformation products of primarily organic contaminants (Glen and Sansalone 2002; Sillanpää and Koivusalo 2013; Westerlund and Viklander 2011). In this study, while sample DOC contributed to oxygen demand, the majority of COD is likely inorganic: COD has been observed as a major source of contamination in roadway runoff (Huang et al. 2007) and it is not unreasonable that COD in snow would be mainly traffic-derived. Strong COD correlations with reactive metals species Cr, Ni, Pb, and Zn indicate probable road and vehicle origins. However, despite all snow facilities receiving a mixture of roadside snow, the estimated contaminant loads for each facility varied. Contaminant deposition decreases drastically with increased distance from the roadway, with the bulk of pollutants deposited immediately beside the road (Hautala et al. 1995; Kuoppamäki et al. 2014). Hautala et al. (1995) observed chloride concentrations were significantly greater at 10 m than 30 m from the road edge while the opposite was true for low-molecular-weight (LMW) polyaromatic hydrocarbons (PAHs). Industrial emissions have additionally been implicated in non-roadside SP contamination (Li et al. 2015). The comparatively low PAH (0.07 kg) to DOC loads (907 kg) are likely due to high particle affinity, but deposition distribution may play some role in inter-site variation. Facilities likely received volumes of non-roadside snow which contributed different contaminant species and concentrations. The Valley Rd. site possessed the greatest contamination load as expected due to lack of on-site infiltration. However, its TSS values were comparable to those observed at facilities with half the estimated SM volume at Wanuskewin Rd. and Central Ave. The Central Ave. site possessed 89-fold higher TDS and 70-fold higher COD than the Wanuskewin Rd. site despite seemingly similar environmental characteristics. The infiltration capacity at the Central Ave site may have been reduced by high-TDS SM runoff over time, resulting in overall poorer SM quality leaving the site and impacting surrounding environments.