road salt sels voirie-eng - canada.ca · untitled document this content was archived on june 24,...

Untitled Document

This content was archived on June 24, 2013.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards. As per the Communications Policy of the Government of Canada, you can request alternate formats on the "Contact Us" page.

file:///C|/Documents%20and%20Settings/ASLOBODY/Desktop/Working%20folder/archive-eng.html2013-06-20 12:59:43 PM

http://www.tbs-sct.gc.ca/pol/doc-eng.aspx?id=12316&section=text

http://www.tbs-sct.gc.ca/pol/doc-eng.aspx?id=12316&section=text

http://www.hc-sc.gc.ca/contact/index-eng.php

Environment EnvironnementCanada Canada

Health SantéCanada Canada

PRIORITY SUBSTANCES LIST ASSESSMENT REPORT

Canadian EnvironmentalProtection Act, 1999

Road Salts

Canadian Environmental Protection Act, 1999

PRIORITY SUBSTANCES LIST ASSESSMENT REPORT

Road Salts

Environment CanadaHealth Canada

2001

TABLE OF CONTENTS

SYNOPSIS ..................................................................................................................1

1.0 INTRODUCTION ..............................................................................................5

2.0 SUMMARY OF BACKGROUND INFORMATION FOR THE ASSESSMENT

OF ROAD SALTS ..............................................................................................9

2.1 Identity, properties, production and sources ......................................92.1.1 Substance characterization ................................................................92.1.2 Physical and chemical properties ......................................................92.1.3 Production and use............................................................................11

2.1.3.1 Sodium chloride ............................................................112.1.3.2 Calcium chloride............................................................112.1.3.3 Mixture of sodium chloride and calcium chloride........122.1.3.4 Magnesium chloride ......................................................122.1.3.5 Mixture of sodium chloride and magnesium chloride ..122.1.3.6 Potassium chloride ........................................................122.1.3.7 Brines ............................................................................122.1.3.8 Ferrocyanide ..................................................................12

2.1.4 Sources and releases ........................................................................132.1.4.1 Natural sources ..............................................................132.1.4.2 Anthropogenic sources ..................................................13

2.2 Road salt loadings ................................................................................142.2.1 Surveys ......................................................................................14

2.2.1.1 Provinces ......................................................................142.2.1.2 Municipalities and regional governments ....................142.2.1.3 Salt industry data ..........................................................142.2.1.4 Municipal estimates ......................................................142.2.1.5 Provincial and municipal use of dust

suppressant salts ............................................................152.2.2 Results ..............................................................................................15

2.2.2.1 Sodium chloride ............................................................162.2.2.1.1 Recommended application rates ..................162.2.2.1.2 Loadings ......................................................16

2.2.2.2 Calcium chloride ..........................................................182.2.2.3 Total chloride loadings ..................................................182.2.2.4 Historic trends ..............................................................18

2.2.3 Reliability of loadings data ..............................................................202.2.4 Summary............................................................................................21

2.3 Snow disposal ........................................................................................212.3.1 Characterization of snow from roadways ........................................21

PSL ASSESSMENT REPORT — ROAD SALTS iii

2.3.2 Volumes of snow in Canadian cities ................................................222.3.3 Snow clearing and snow disposal methods ......................................22

2.3.3.1 Methods not involving snow removal ..........................222.3.3.2 Methods involving release into waterways

after processing or treating meltwater ..........................232.3.3.3 Methods involving release to surface water

without water treatment ................................................242.3.4 Use of disposal methods by Canadian municipalities ......................242.3.5 Summary............................................................................................25

2.4 Patrol yards ..........................................................................................252.4.1 Number of patrol yards and quantity of materials stored ................252.4.2 Patrol yard design and management ................................................262.4.3 Estimation of releases of salts from patrol yards ............................272.4.4 Salt concentrations at different areas around patrol yards ..............28

2.4.4.1 Concentrations at surface points around patrol yards....282.4.4.2 Concentrations in patrol yard well water ......................292.4.4.3 Concentrations in shallow groundwater........................292.4.4.4 Concentrations in patrol yard soils................................30

2.4.5 Off-site concentrations at patrol yards ............................................30

2.5 Environmental fate and pathways ......................................................312.5.1 Fate/transport ..................................................................................31

2.5.1.1 Chloride ........................................................................322.5.1.2 Sodium and potassium ..................................................322.5.1.3 Calcium..........................................................................322.5.1.4 Magnesium ....................................................................332.5.1.5 Sodium ferrocyanide ....................................................33

2.5.2 Environmental pathways ..................................................................332.5.2.1 Roadway applications....................................................332.5.2.2 Snow disposal sites........................................................332.5.2.3 Storage sites/patrol yards ..............................................35

2.6 Chloride concentrations in surface waters across Canada ..............352.6.1 Mapping of chloride concentrations in watersheds across

Canada ..............................................................................................352.6.2 Environmental concentrations: Lakes and rivers ............................36

2.6.2.1 Atlantic region (Newfoundland, Nova Scotia, New Brunswick and Prince Edward Island) ................36

2.6.2.2 Central region (Ontario and Quebec)............................382.6.2.3 Prairie region (Manitoba, Saskatchewan and Alberta)..412.6.2.4 Pacific region (British Columbia) ................................412.6.2.5 Yukon, Northwest Territories and Nunavut...................42

2.6.3 Environmental concentrations: Benthic sediments...........................422.6.4 Environmental concentrations: Wetlands .........................................422.6.5 Environmental concentrations: Urban lakes and ponds ..................432.6.6 Chloride concentrations based on mass balance calculations.........44

2.6.6.1 Calculation of watershed chloride concentrations resulting from the use of road salts ...............................44

PSL ASSESSMENT REPORT — ROAD SALTSiv

2.6.6.2 Calculation of chloride concentrations in roadway runoff ..............................................................45

2.6.7 Summary............................................................................................46

3.0 SUMMARY OF CRITICAL INFORMATION AND ASSESSMENT OF

“TOXIC” UNDER CEPA 1999 ....................................................................49

3.1 CEPA 1999 64(a): Environment and CEPA 1999 64(b): Environment on which life depends....................49

3.2 Groundwater ........................................................................................493.2.1 Introduction ......................................................................................493.2.2 Groundwater principles ....................................................................50

3.2.2.1 Hydrological cycle ........................................................503.2.2.2 Unsaturated flow and groundwater recharge ................503.2.2.3 Saturated flow and contaminant migration ..................50

3.2.3 Impacts of road salts on groundwater quality ..................................503.2.3.1 Factors controlling the migration of road salts ............503.2.3.2 Predicted road salt concentrations in groundwater ......513.2.3.3 Plume migration and impact assessment ......................54

3.2.4 Case studies ......................................................................................553.2.5 Conclusions ......................................................................................57

3.3 Aquatic ecosystems ..............................................................................573.3.1 Scoping and assessment approach....................................................573.3.2 Laboratory studies ............................................................................59

3.3.2.1 Short-term or acute toxicity studies ..............................593.3.2.2 Chronic toxicity ............................................................63

3.3.3 Assessment of toxicity to aquatic ecosystems ..................................643.3.3.1 Quotient-based risk characterization ............................643.3.3.2 Tier 1 and Tier 2 assessments........................................65

3.3.3.2.1 Estimated Exposure Values (EEVs) ............653.3.3.2.2 Critical Toxicity Values (CTVs) ..................663.3.3.2.3 Conclusions for Tier 1 and

Tier 2 assessments........................................663.3.3.3 Tier 3 assessments with field validation ......................67

3.3.3.3.1 Case 1: Road salt runoff and urban creeks, streams and small rivers in densely populated areas ..........................68

3.3.3.3.2 Case 2: Road salt runoff and creeks, streams and rivers in less densely populated areas ..............................70

3.3.3.3.3 Case 3: Road salt runoff and urban ponds and lakes ................................72

3.3.3.3.4 Case 4: Highway runoff and rural ponds and lakes ..................................74

3.3.3.3.5 Case 5: Salt storage depots and aquatic ecosystems ......................................76

PSL ASSESSMENT REPORT — ROAD SALTS v

PSL ASSESSMENT REPORT — ROAD SALTSvi

3.3.3.3.6 Case 6: Snow dumps ....................................793.3.3.4 Other effects ..................................................................793.3.3.5 Conclusions for aquatic ecosystems..............................81

3.4 Soils ........................................................................................................843.4.1 Soil salinization resulting from road salt application ......................84

3.4.1.1 Definitions and equations..............................................853.4.1.2 Case studies ..................................................................873.4.1.3 Discussion......................................................................893.4.1.4 Conclusions ..................................................................94

3.4.2 Biological effects of road salts on soils ............................................943.4.2.1 Ecological role of soil biotic communities ..................943.4.2.2 Effects on soil biota ......................................................953.4.2.3 Exposure of soil organisms to road salt ........................963.4.2.4 Environmental assessment ............................................963.4.2.5 Conclusions ..................................................................96

3.5 Terrestrial vegetation............................................................................963.5.1 Conservative environmental assessment ..........................................98

3.5.1.1 Effects threshold values ..............................................1013.5.1.2 Estimated Exposure Values (EEVs) ............................1013.5.1.3 Quotient estimation ....................................................103

3.5.2 Detailed environmental assessment — threshold-based analysis ..1033.5.2.1 Response to sodium in soil..........................................1033.5.2.2 Response to chloride in soil ........................................1033.5.2.3 Response to sodium chloride in soil ...........................1043.5.2.4 Response to sodium and chloride by

aerial dispersion...........................................................1043.5.2.5 Detailed environmental assessment —

summary of threshold-based analysis .........................1043.5.3 Detailed environmental assessment — reference site analysis ......105

3.5.3.1 Spread of halophytic plant species..............................1053.5.3.2 Community changes in a bog impacted by road salts...1053.5.3.3 Effects of salt spray from a four-lane highway on

peach and plum orchards in Grimsby, Ontario ...........1063.5.3.4 Effects of salt spray from a four-lane highway on

peach and plum orchards near St. Catharines, Ontario ...1063.5.3.5 Effects of salt spray from a two-lane highway

on blueberries in western Nova Scotia........................1073.5.3.6 Concentrations in soil and woody plants

adjacent to highways in interior British Columbia .....1073.5.3.7 Concentrations in soil and woody foliage adjacent

to a two-lane highway in British Columbia ................1083.5.3.8 Concentrations in woody foliage directly adjacent

to a two-lane highway in Connecticut.........................1093.5.3.9 Concentrations in soil and woody and herbaceous

foliage along roadsides in Massachusetts....................110

3.5.3.10 Concentrations in soil and woody plants downslope from a four-lane highway in British Columbia ..........110

3.5.3.11 Concentrations in soil and woody foliage adjacent to a patrol yard ............................................................112

3.5.4 Changes in application rates and current injury in Canada..........1123.5.5 Estimate of area of potential injury ................................................1143.5.6 Risk assessment summary and conclusions ....................................114

3.6 Terrestrial wildlife ..............................................................................1153.6.1 Exposure characterization ..............................................................115

3.6.1.1 Mammals ....................................................................1153.6.1.2 Birds ............................................................................116

3.6.2 Effects characterization ..................................................................1173.6.2.1 Acute toxicity of sodium chloride ..............................1173.6.2.2 Sublethal effects of excess salt ingestion in birds

and mammals ..............................................................1183.6.3 Risk characterization ......................................................................118

3.6.3.1 Estimating the contribution of salt toxicity to vehicle strikes ..............................................................118

3.6.3.2 Conclusions and discussion of uncertainty ................119

3.7 Ferrocyanides ......................................................................................1213.7.1 Exposure characterization ..............................................................121

3.7.1.1 Soil ..............................................................................1213.7.1.1.1 Non-illuminated ferrocyanide ....................1213.7.1.1.2 Illuminated ferrocyanide............................122

3.7.1.2 Water............................................................................1223.7.1.3 Air................................................................................1223.7.1.4 Biomagnification ........................................................123

3.7.2 Measured and estimated environmental concentrations ................1233.7.3 Effects characterization ..................................................................123

3.7.3.1 Aquatic biota ..............................................................1233.7.3.1.1 Microorganisms..........................................1233.7.3.1.2 Plants ........................................................1233.7.3.1.3 Invertebrates ..............................................1243.7.3.1.4 Vertebrates ................................................124

3.7.3.2 Terrestrial biota............................................................1243.7.3.2.1 Microorganisms..........................................1243.7.3.2.2 Plants ........................................................1243.7.3.2.3 Invertebrates ..............................................1243.7.3.2.4 Vertebrates ................................................124

3.7.4 Risk characterization ......................................................................1253.7.4.1 Tier 1 risk characterization..........................................125

3.7.4.1.1 Cyanide concentration in water ................1253.7.4.1.2 Cyanide concentration in soil: Case I ......1253.7.4.1.3 Cyanide concentration in soil: Case II......1253.7.4.1.4 Cyanide concentration in air ....................125

PSL ASSESSMENT REPORT — ROAD SALTS vii

3.7.4.1.5 Quotient calculations ................................1253.7.4.1.6 Tier 1 conclusions ......................................126

3.7.4.2 Tier 2 risk characterization..........................................1263.7.4.2.1 Cyanide concentration in water ................1273.7.4.2.2 Cyanide concentration in soil: Case I ......1273.7.4.2.3 Cyanide concentration in soil: Case II......1273.7.4.2.4 Cyanide concentration in air ....................1273.7.4.2.5 Quotient calculations ................................1273.7.4.2.6 Tier 2 conclusions ......................................127

3.7.4.3 Tier 3 risk characterization..........................................1273.7.4.3.1 Assumptions ..............................................1283.7.4.3.2 Estimation of the frequency of

occurrence of toxic events..........................1283.7.4.3.3 Tier 3 results and conclusions ..................129

3.7.5 Conclusion for ferrocyanides..........................................................130

3.8 Overall conclusions ............................................................................1303.8.1 Considerations ................................................................................130

3.8.1.1 Environment ................................................................1303.8.1.2 Harmful effects ............................................................1313.8.1.3 Weight of evidence ......................................................132

3.8.2 Summary of assessment ..................................................................1323.8.3 Uncertainties ..................................................................................1343.8.4 Conclusions ....................................................................................137

3.9 Considerations for follow-up (further action)..................................138

4.0 REFERENCES ..............................................................................................141

APPENDIX A SEARCH STRATEGIES EMPLOYED FOR IDENTIFICATION

OF RELEVANT DATA ......................................................................167

APPENDIX B SUPPORTING DOCUMENTS PREPARED FOR THE

PSL ASSESSMENT OF ROAD SALTS ..............................................169

PSL ASSESSMENT REPORT — ROAD SALTSviii

LIST OF TABLES

TABLE 1 Physical-chemical properties of road salts ..............................................................10

TABLE 2 Total loading of sodium chloride road salt, winter 1997–98 (from Morin and Perchanok, 2000) ..............................................................................................18

TABLE 3 Estimated quantities of calcium chloride used on roadways in a typical year (from Morin and Perchanok, 2000) .........................................................................19

TABLE 4 Total loadings of chloride (from Morin and Perchanok, 2000)...............................19

TABLE 5 Total volume of waste snow and quantity of salt used in certain Canadian cities, winter 1997–98 (from Delisle and Dériger, 2000)........................23

TABLE 6 Snow disposal methods used in certain Canadian cities, winter 1997–98 (from Delisle and Dériger, 2000).............................................................................25

TABLE 7 Number of provincial patrol yards and sodium chloride and abrasive use, by province (from Snodgrass and Morin, 2000) .....................................................26

TABLE 8 Estimate of magnitude of salt loss at patrol yards (from Snodgrass and Morin, 2000) ............................................................................................................29

TABLE 9 Chloride concentrations in various streams in the Toronto Remedial Action Plan watershed for 1990–96 (from Toronto and Region Conservation Authority, 1998) ....................................................................40

TABLE 10 Toxicity responses of organisms to sodium chloride for exposures less than 1 day (from Evans and Frick, 2001)................................................................60

TABLE 11 Toxicity responses of organisms to sodium chloride for exposures of 1 day (from Evans and Frick, 2001)..................................................................................61

TABLE 12 Four-day LC50s of various taxa exposed to sodium chloride (from Evans and Frick, 2001)..................................................................................61

TABLE 13 Seven- to 10-day LC50s and EC50s of various taxa exposed to sodium chloride (from Evans and Frick, 2001).......................................................62

TABLE 14 Summary of Tier 2 calculations (from Evans and Frick, 2001) ..............................67

TABLE 15 pH, total phosphorus and chloride optima for selected diatom species in the northeastern United States (from Dixit et al., 1999) .........................................76

TABLE 16 Sodium chloride tolerance of selected plant species in the salt-impacted mat zone of Pinhook Bog, Indiana (from Wilcox, 1982) .................78

TABLE 17 Predicted cumulative percentage of species affected by chronic exposures to chloride (from Evans and Frick, 2001) ...............................................................82

PSL ASSESSMENT REPORT — ROAD SALTS ix

TABLE 18 Modelled electrical conductivity values and chloride concentrations for Corbett Brook, downstream from Fredericton’s salt depot .....................................90

TABLE 19 Species list of roadside trees and shrubs rated for their resistance to airborne highway salt spray (from Lumis et al., 1983) ...........................................99

TABLE 20 Species list of native forest tree species rated for their resistance to highway salt spray .................................................................................................100

TABLE 21 Range of threshold values estimated for soil and water for various forms of plants (from Cain et al., 2001)...............................................................................101

TABLE 22 Range of threshold values estimated for herbaceous tissue concentrations following soil or water solution exposure (from Cain et al., 2001)......................102

TABLE 23 Range of threshold values estimated for woody tissue concentrations following aerial exposure (from Cain et al., 2001) ...............................................102

TABLE 24 Calculation of the number of particles of salt that need to be ingested in order to reach CTVs assuming a model 28-g house sparrow consuming particles at the upper end of its known preference range.........................................................119

TABLE 25 Calculation of the number of particles of salt that need to be ingested in order to reach CTVs assuming a model 28-g house sparrow consuming particles of average size ............................................................................................................120

TABLE 26 Quotient calculation results for aquatic biota — Tier 1 (from Letts, 2000a)........126

TABLE 27 Quotient calculation results for terrestrial biota — Tier 1 (from Letts, 2000a) ....126

TABLE 28 Quotient calculation results for aquatic biota — Tier 2 (from Letts, 2000a)........128

TABLE 29 Quotient calculation results for terrestrial biota — Tier 2 (from Letts, 2000a) ....128

PSL ASSESSMENT REPORT — ROAD SALTSx

LIST OF FIGURES

FIGURE 1 Provincial road network area fraction (from Morin and Perchanok, 2000) ............15

FIGURE 2 Recommended provincial application rates for sodium chloride road salts, 1998 (from Morin and Perchanok, 2000)...............................................16

FIGURE 3 Average recommended municipal application rates for sodium chloride road salts, 1998 (from Morin and Perchanok, 2000)...............................................17

FIGURE 4 Total amount of sodium chloride used, by maintenance district (from Morin and Perchanok, 2000) ..............................................................................................17

FIGURE 5 Historical salt use by provincial agencies (based on surveys conducted by the Salt Institute, 1964–1983, and Morin and Perchanok, 2000)..................................20

FIGURE 6 Average daily concentration of leachate from salt-treated sand pile, 1975–76 (from NB DOE and DOT, 1978) .............................................................................28

FIGURE 7 Chloride concentrations in Heffley Creek municipal wells (from AGRA, 1999)...31

FIGURE 8 Road salts transport pathway ...................................................................................34

FIGURE 9 Observed chloride concentrations in Canadian watersheds (from Mayer et al., 1999) ........................................................................................37

FIGURE 10 Estimated road salt chloride concentrations by watershed, calculated from average annual road salt loadings and average annual runoff (from Mayer et al., 1999) ........................................................................................45

FIGURE 11 Chloride concentrations in runoff for provincial roads in Canada..........................46

FIGURE 12 Estimated chloride concentrations in groundwater for various chloride application rates and a groundwater recharge rate of 10 cm/year (from Johnston et al., 2000).....................................................................................52

FIGURE 13 Estimated chloride concentrations in groundwater for various chloride application rates and a groundwater recharge rate of 20 cm/year (from Johnston et al., 2000).....................................................................................52

FIGURE 14 Estimated chloride concentrations in groundwater for various chloride application rates and a groundwater recharge rate of 40 cm/year (from Johnston et al., 2000)..................................................................53

FIGURE 15 Chloride concentrations in groundwater from municipal production wells in southern Ontario (from Johnston et al., 2000) ....................................................56

FIGURE 16 Species diversity across a salinity and chloride gradient (from Wetzel, 1983).......65

FIGURE 17 Experimental acute toxicity and predicted chronic toxicity for aquatic taxa (from Evans and Frick, 2001)..................................................................................83

PSL ASSESSMENT REPORT — ROAD SALTS xi

FIGURE 18 Representative short-term chloride concentrations in the Canadian aquatic environment associated with contamination by road salts and concentrations causing adverse biological effects following brief exposures .................................83

FIGURE 19 Representative long-term chloride concentrations in the Canadian aquatic environment associated with contamination by road salts and concentrations causing adverse biological effects following prolonged exposures ........................84

FIGURE 20 Areas with low to medium to high road salt application hazards, based on product of soil clay content with exchangeable sodium ratio (from Morin et al., 2000).........................................................................................88

FIGURE 21 Box plot showing 10th, 25th, 75th, and 90th percentile plus individual data above and below the 10th and 90th percentile for ditch water electrical conductivity and chloride concentrations along two highways near Fredericton’s salt depot, outside the depot’s catchment area for representative days in June, July, October 1999, and January 2000. “Well-drained”, “cross-flow”, and “stagnant” refer to three conditions: ridges from which water drains, stream/road crossings, and ditch depressions with no visible outflow, respectively (from Arp, 2001)..................................................................................89

FIGURE 22 Estimates for total dissolved solids (TDS) and electrical conductivity (EC) of average soil solution and surface waters, by level 2 watersheds and by municipal boundary, Ontario and Quebec (from Morin et al., 2000) ....................92

FIGURE 23 Exchangeable sodium percentage (ESP) along roadsides, by provincial road maintenance district (from Morin et al., 2000) ...............................................93

FIGURE 24 Salt tolerance and crop yield relative to soil salinity (electrical conductivity, EC) and the estimated percentage of roadsides with corresponding levels of electricalconductivity (from Bresler et al., 1982) ..................................................................95

PSL ASSESSMENT REPORT — ROAD SALTSxii

LIST OF ACRONYMS AND ABBREVIATIONS

CAS Chemical Abstracts ServiceCEPA Canadian Environmental Protection ActCEPA 1999 Canadian Environmental Protection Act, 1999CTV Critical Toxicity ValueEC electrical conductivityEC50 median effective concentrationEEV Estimated Exposure ValueENEV Estimated No-Effects ValueESA environmental site assessmentkg-bw kilogram body weightLC50 median lethal concentrationLD50 median lethal doseLOEC Lowest-Observed-Effect ConcentrationLOEL Lowest-Observed-Effect LevelNOEC No-Observed-Effect ConcentrationNOEL No-Observed-Effect LevelPSL Priority Substances List

PSL ASSESSMENT REPORT — ROAD SALTS xiii

Road salts are used as de-icing and anti-icingchemicals for winter road maintenance, withsome use as summer dust suppressants. Inorganicchloride salts considered in this assessmentinclude sodium chloride, calcium chloride,potassium chloride and magnesium chloride. Inthe environment, these compounds dissociate intothe chloride anion and the corresponding cation.In addition, ferrocyanide salts, which are added asanti-caking agents to some road salt formulations,were assessed. It is estimated that approximately4.75 million tonnes of sodium chloride were usedas road salts in the winter of 1997–98 and that110 000 tonnes of calcium chloride are used onroadways in a typical year. Very small amountsof other salts are used. Based on these estimates,about 4.9 million tonnes of road salts can bereleased to the environment in Canada everyyear, accounting for about 3.0 million tonnes ofchloride. The highest annual loadings of road saltson a road-length basis are in Ontario and Quebec,with intermediate loadings in the Atlanticprovinces and lowest loadings in the westernprovinces.

Road salts enter the Canadianenvironment through their storage and use andthrough disposal of snow cleared from roadways.Road salts enter surface water, soil andgroundwater after snowmelt and are dispersedthrough the air by splashing and spray fromvehicles and as windborne powder. Chloride ionsare conservative, moving with water withoutbeing retarded or lost. Accordingly, all chlorideions that enter the soil and groundwater canultimately be expected to reach surface water; itmay take from a few years to several decades ormore for steady-state groundwater concentrationsto be reached. Because of the widespreaddispersal of road salts through the environment,environmental concerns can be associated withmost environmental compartments.

In water, natural backgroundconcentrations of chloride are generally no morethan a few milligrams per litre, with some localor regional instances of higher natural salinity,notably in some areas of the Prairies and BritishColumbia. High concentrations of chloride relatedto the use of road salts on roadways or releasesfrom patrol yards or snow dumps have beenmeasured. For example, concentrations ofchloride over 18 000 mg/L were observed inrunoff from roadways. Chloride concentrations upto 82 000 mg/L were also observed in runoff fromuncovered blended abrasive/salt piles in a patrolyard. Chloride concentrations in snow clearedfrom city streets can be quite variable. Forexample, the average chloride concentrations insnow cleared from streets in Montréal were3000 mg/L for secondary streets and 5000 mg/Lfor primary streets. Waters from roadways, patrolyards or snow dumps can be diluted to variousdegrees when entering the environment. In theenvironment, resulting chloride concentrationshave been measured as high as 2800 mg/L ingroundwater in areas adjacent to storage yards,4000 mg/L in ponds and wetlands, 4300 mg/Lin watercourses, 2000–5000 mg/L in urbanimpoundment lakes and 150–300 mg/L in rurallakes. While highest concentrations are usuallyassociated with winter or spring thaws, highconcentrations can also be measured in thesummer, as a result of the travel time of theions to surface waters and the reduced waterflows in the summer. Water bodies most subjectto the impacts of road salts are small ponds andwatercourses draining large urbanized areas, aswell as streams, wetlands or lakes draining majorroadways. Field measurements have shown thatroadway applications in rural areas can result inincreased chloride concentrations in lakes locateda few hundred metres from roadways.

The potential for impacts on regionalgroundwater systems was evaluated using a mass

PSL ASSESSMENT REPORT — ROAD SALTS 1

SYNOPSIS

PSL ASSESSMENT REPORT — ROAD SALTS2

balance technique that provides an indication ofpotential chloride concentrations downgradientfrom saltable road networks. The mass balancemodelling and field measurements indicatedthat regional-scale groundwater chlorideconcentrations greater than 250 mg/L will likelyresult under high-density road networks subject toannual loadings above 20 tonnes sodium chlorideper two-lane-kilometre. Considering data on roadsalt loadings, urban areas in southern Ontario,southern Quebec and the Atlantic provinces facethe greatest risk of regional groundwater impacts.Groundwater will eventually well up into thesurface water or emerge as seeps and springs.Research has shown that 10–60% of the saltapplied enters shallow subsurface waters andaccumulates until steady-state concentrationsare attained. Elevated concentrations of chlorideshave been detected in groundwater springsemerging to the surface.

Acute toxic effects of chloride on aquaticorganisms are usually observed at relativelyelevated concentrations. For example, the 4-daymedian lethal concentration (LC50) for thecladoceran Ceriodaphnia dubia is 1400 mg/L.Exposure to such concentrations may occur insmall streams located in heavily populated urbanareas with dense road networks and elevated roadsalt loadings, in ponds and wetlands adjacent toroadways, near poorly managed salt storagedepots and at certain snow disposal sites.

Chronic toxicity occurs at lowerconcentrations. Toxic effects on aquatic biotaare associated with exposures to chlorideconcentrations as low as 870, 990 and 1070 mg/Lfor median lethal effects (fathead minnowembryos, rainbow trout eggs/embryos anddaphnids, respectively). The No-Observed-EffectConcentration (NOEC) for the 33-day early lifestage test for survival of fathead minnow was252 mg chloride/L. Furthermore, it is estimatedthat 5% of aquatic species would be affected(median lethal concentration) at chlorideconcentrations of about 210 mg/L, and 10%of species would be affected at chloride

concentrations of about 240 mg/L. Changesin populations or community structure can occurat lower concentrations. Because of differences inthe optimal chloride concentrations for the growthand reproduction of different species of algae,shifts in populations in lakes were associated withconcentrations of 12–235 mg/L. Increased saltconcentrations in lakes can lead to stratification,which retards or prevents the seasonal mixingof waters, thereby affecting the distribution ofoxygen and nutrients. Chloride concentrationsbetween 100 and 1000 mg/L or more havebeen observed in a variety of urban watercoursesand lakes. For example, maximum chlorideconcentrations in water samples from fourToronto-area creeks ranged from 1390 to4310 mg/L. Chloride concentrations greater thanabout 230 mg/L, corresponding to those havingchronic effects on sensitive organisms, have beenreported from these four watercourses throughmuch of the year. In areas of heavy use ofroad salts, especially southern Ontario, Quebecand the Maritimes, chloride concentrations ingroundwater and surface water are frequently atlevels likely to affect biota, as demonstrated bylaboratory and field studies.

Application of road salts can also resultin deleterious effects on the physical and chemicalproperties of soils, especially in areas that sufferfrom poor salt, soil and vegetation management.Effects are associated with areas adjacent to saltdepots and roadsides, especially in poorly draineddepressions. Effects include impacts on soilstructure, soil dispersion, soil permeability, soilswelling and crusting, soil electrical conductivityand soil osmotic potential. These can have, inturn, abiotic and biotic impacts on the localenvironment. The primary abiotic impact is theloss of soil stability during drying and wettingcycles and during periods of high surface runoffand wind. Biological impacts relate primarily toosmotic stress on soil macro- and microflora andmacro- and microfauna, as well as salt-inducedmobilization of macro- and micronutrients thataffect flora and fauna.

A number of field studies havedocumented damage to vegetation and shifts inplant community structure in areas impacted byroad salt runoff and aerial dispersion. Halophyticspecies, such as cattails and common reed-grass,readily invade areas impacted by salt, leadingto changes in occurrence and diversity of salt-sensitive species. Elevated soil levels of sodiumand chloride or aerial exposures to sodium andchloride result in reductions in flowering andfruiting of sensitive plant species; foliar, shootand root injury; growth reductions; and reductionsin seedling establishment. Sensitive terrestrialplants may be affected by soil concentrationsgreater than about 68 mg sodium/kg and 215 mgchloride/kg. Areas with such soil concentrationsextend linearly along roads and highways or otherareas where road salts are applied for de-icing ordust control. The impact of aerial dispersionextends up to 200 m from the edge of multi-lanehighways and 35 m from two-lane highwayswhere de-icing salts are used. Salt injury tovegetation also occurs along watercourses thatdrain roadways and salt handling facilities.

Behavioural and toxicologicalimpacts have been associated with exposureof mammalian and avian wildlife to road salts.Ingestion of road salts increases the vulnerabilityof birds to car strikes. Furthermore, intakecalculations suggest that road salts may poisonsome birds, especially when water is not freelyavailable during severe winters. Road salts mayalso affect wildlife habitat, with reduction in plantcover or shifts in communities that could affectwildlife dependent on these plants for food orshelter. Available data suggest that the severityof road kills of federally protected migratorybird species (e.g., cardueline finches) and thecontribution of road salts to this mortality havebeen underestimated.

Ferrocyanides are very persistent but areof low toxicity. However, in solution and in thepresence of light, they can dissociate to formcyanide. In turn, the cyanide ion may volatilizeand dissipate fairly quickly. The ultimate effectsof ferrocyanides therefore depend on the complexbalance between photolysis and volatilization,which in turn depends on environmental factors.Modelling studies undertaken in support of thisassessment indicate that there is a potential forcertain aquatic organisms to be adversely affectedby cyanide in areas of high use of road salts.

Based on the available data, it isconsidered that road salts that containinorganic chloride salts with or withoutferrocyanide salts are entering theenvironment in a quantity or concentration orunder conditions that have or may have animmediate or long-term harmful effect on theenvironment or its biological diversity or thatconstitute or may constitute a danger to theenvironment on which life depends. Therefore,it is concluded that road salts that containinorganic chloride salts with or withoutferrocyanide salts are “toxic” as defined inSection 64 of the Canadian EnvironmentalProtection Act, 1999 (CEPA 1999).

The use of de-icing agents is an importantcomponent of strategies to keep roadways openand safe during the winter and minimize trafficcrashes, injuries and mortality under icy andsnowy conditions. These benefits were recognizedby the Ministers’ Expert Advisory Panel on theSecond Priority Substances List, even as theyrecommended that this assessment of potentialimpacts on the environment be conducted. Anymeasures developed as a result of this assessmentmust never compromise human safety; selectionof options must be based on optimization ofwinter road maintenance practices so as not tojeopardize road safety, while minimizing thepotential for harm to the environment. Any actiontaken to reduce impacts on the environment isalso likely to reduce potential for contaminationof groundwater-based drinking water supplies,which is clearly desirable.


Future management should focus onkey sources in areas where the assessment hasindicated concerns. These relate to the following:

• Patrol yards: Key concerns relate to thecontamination of groundwater at patrolyards and the discharge to surface water. Inaddition, overland flow of salty snowmeltwaters can result in direct impacts to surfacewater and near-field vegetation. Based onsurveys and reviews, salt losses from patrolyards are associated with loss at storage piles(which include salt piles as well as piles ofsand and gravel to which salts have beenadded) and during the handling of salts,relating to both storage and loading andunloading of trucks. The discharge of patrolyard washwater is also a potential sourceof release of salts. Measures and practicesshould therefore be considered to ensurestorage of salts and abrasives to reduce lossesthrough weathering, to reduce losses duringtransfers and to minimize releases ofstormwater and equipment washwater.

• Roadway application: Key environmentalconcerns have been associated with areasof high salt use and high road density.Regions of southern Ontario and Quebecand the Atlantic provinces have the highestrate of salt use on an area basis and as suchhave the highest potential for contaminationof soils, groundwater and surface water byroad salts as a result of roadway applications.In addition, urban areas in other parts ofthe country where large amounts of salts areapplied are of potential concern, especiallyfor streams and aquifers that are whollysurrounded by urban areas. In rural areas,surface waters receiving drainage fromroadways may also be susceptible tocontamination. Areas where splash or sprayfrom salted roads can be transported through

air to sensitive vegetation are a potentialconcern. Wetlands that directly adjoinroadway ditches and that receive runoff inthe form of salty snowmelt waters are alsopotential management concerns. Therefore,measures should be considered to reduce theoverall use of chloride salts in such areas.The selection of alternative products or ofappropriate practices or technology to reducesalt use should be considered while ensuringmaintenance of roadway safety.

• Snow disposal: Key environmental concernsrelate to eventual loss of meltwater intosurface water and into soil and groundwaterat snow disposal sites. Measures to minimizepercolation of salty snowmelt waters into soiland groundwater at snow disposal sites shouldbe considered. Practices to direct the releaseof salty snowmelt waters into surface watersthat have minimal environmental sensitivityor into storm sewers could be considered.Measures should also be considered to ensuresufficient dilution before release.

• Ferrocyanides: This assessment indicatesthat there is a possible adverse exposurefor the more sensitive aquatic vertebrates inareas of very high use of road salts. Riskscould be reduced by reducing total salt useor reducing content of ferrocyanides in roadsalt formulations. To reduce the possibilityof exposure, producers of road salts couldconsider reducing the addition rate offerrocyanide to road salts. Any reduction intotal salt use would be expected to resultin an equivalent reduction in release offerrocyanides.


The Canadian Environmental Protection Act,1999 (CEPA 1999) requires the federal Ministersof the Environment and of Health to prepareand publish a Priority Substances List (PSL)that identifies substances, including chemicals,groups of chemicals, effluents and wastes, thatmay be harmful to the environment or constitutea danger to human health. The Act also requiresboth Ministers to assess these substances anddetermine whether they are “toxic” or capable ofbecoming “toxic” as defined in Section 64 of theAct, which states:

...a substance is toxic if it is entering or may enterthe environment in a quantity or concentrationor under conditions that

(a) have or may have an immediate or long-termharmful effect on the environment or itsbiological diversity;

(b) constitute or may constitute a danger to theenvironment on which life depends; or

(c) constitute or may constitute a danger inCanada to human life or health.

Substances that are assessed as “toxic”as defined in Section 64 may be placed on theList of Toxic Substances in Schedule I of theAct and considered for possible risk managementmeasures, such as regulations, guidelines,pollution prevention plans or codes of practiceto control any aspect of their life cycle, fromthe research and development stage throughmanufacture, use, storage, transport and ultimatedisposal.

Based on initial screening of readilyaccessible information, the rationale for assessingroad salts provided by the Ministers’ ExpertAdvisory Panel on the Second Priority SubstancesList (Ministers’ Expert Advisory Panel, 1995) wasas follows:

The Panel recognized the benefits associated withthe use of road salts. However, these substanceshave negative effects on the environment. Largevolumes are released through road salting,

particularly in Ontario, Quebec and the Atlanticprovinces. There is evidence of adverse localenvironmental effects to groundwater and to plantand animal life following exposure. Algae andbenthic fauna have been shown to be particularlysensitive to changes in chloride ion concentrations,resulting in a reduction of fish populations.The Panel recognizes that there has beenconsiderable progress in upgrading storagefacilities. However, given the widespread exposureto these substances, and their release in largevolumes into the Canadian environment, the Panelbelieves that an assessment is needed to determinetheir ecological effects.

The basis for inclusion of road saltsby the Ministers’ Expert Advisory Panel on theSecond Priority Substances List was limitedto environmental effects and did not identifyconcerns with respect to human health. Humansare exposed to road salts principally through thecontamination of roadside well waters, wherechloride and sodium levels can be increased andtaste adversely affected. These elements are notconsidered toxic; in fact, the Canadian drinkingwater guidelines for chloride and sodium arebased on taste, which is affected at levels wellbelow those that might be of concern for toxicity(Health Canada, 1996). Other substancescontained in road salts, including ferrocyanidecompounds and certain metals, are present onlyat trace levels.

With respect to the potential for healtheffects of sodium chloride road salts, an extensiveliterature search failed to identify any studies thatwere adequate to serve as the basis for a healthrisk assessment. In one study dermal exposureto concentrated solutions of road salt causeddermal irritation, but no sensitization (Cushmanet al., 1991). There is a single correlational studyin which a statistical association between road saltuse and mortality from a number of types ofcancer in the United States was identified (Foster,1993). However, in this study, road salt use wouldhave varied in parallel with a number of other


1.0 INTRODUCTION


factors; for example, larger amounts of salt areused in urban centres, where there is also moreexposure to air pollution. There would alsohave been differences among states in smokingpatterns, diet and a host of other factors. Thedesign of this exploratory study does not permitadjustment for these other factors, many ofwhich are more plausible causes of cancer thanis road salt.

In view of the focus of this assessment onenvironmental effects defined by the Ministers’Expert Advisory Panel and the extremely limitedand inadequate data available related to thepotential impacts of road salts on human health,this assessment solely addresses effects on theenvironment (i.e., the determination of whetherroad salts are “toxic” under Paragraphs 64(a) and64(b) of CEPA 1999).

Descriptions of the approaches toassessment of the effects of Priority Substanceson the environment are available in a publishedcompanion document. The document entitled“Environmental Assessments of PrioritySubstances under the Canadian EnvironmentalProtection Act. Guidance Manual Version 1.0 —March 1997” (Environment Canada, 1997a)provides guidance for conducting environmentalassessments of Priority Substances in Canada.This document may be purchased from:

Environmental Protection PublicationsEnvironmental Technology Advancement

DirectorateEnvironment CanadaOttawa, OntarioK1A 0H3

It is also available on the Existing SubstancesBranch web site at www.ec.gc.ca/cceb1/eng/psap.htm under the heading “Technical GuidanceManual.” It should be noted that the approachoutlined therein has evolved to reflect changesin the Canadian Environmental Protection Actand to incorporate recent developments in riskassessment methodology, which will be addressedin future releases of the guidance manual for the

assessment of effects of Priority Substances onthe environment.

The search strategies employed in theidentification of data relevant to the assessmentof potential effects on the environment (priorto May 2001) are presented in Appendix A.Review articles were consulted where appropriate.However, all original studies that form the basisfor determining whether road salts are “toxic”under CEPA 1999 have been critically evaluated.

Preparation of the environmentalcomponents of the assessment was led byB. Elliott under the direction of R. Chénier.Extensive supporting documentation (AppendixB) related to the environmental assessment ofroad salts was prepared and reviewed by theEnvironmental Resource Group established inJune 1997 by Environment Canada to supportthe environmental assessment:

P. Arp, University of New BrunswickM. Barre, Environment Canada

(until December 1999)Y. Bourassa, Environment Canada

(since December 1999)L. Brownlee, Environment Canada

(until April 2000)B. Butler, University of WaterlooN. Cain, Cain Vegetation Inc.R. Chénier, Environment CanadaC. Delisle, École Polytechnique

de MontréalM. Eggleton, Environment Canada

(until February 2000)B. Elliott, Environment CanadaM. Evans, Environment CanadaK. Hansen, Ontario Ministry of the

Environment J. Haskill, Environment CanadaK. Howard, University of Toronto

(until January 2000)A. Letts, Morton International Inc.B. Mander, Environment CanadaT. Mayer, Environment Canada

P. Mineau, Environment Canada(since April 2000)

D. Morin, Environment CanadaM. Perchanok, Ontario Ministry of

TransportationR. Smith, Transportation Association

of CanadaW. Snodgrass, Snodgrass ConsultantsK. Taylor, Environment CanadaM. Weese, Ontario Ministry of

Transportation (until January 1999)

Environmental Resource Group membersalso reviewed the Assessment Report preparedby Environment Canada. Conclusions andinterpretations in this report are those ofEnvironment Canada and do not necessarilyrepresent those of all members of theEnvironmental Resource Group.

Environmental supporting documentationwas also reviewed by:

K. Adare, Environment CanadaJ. Addison, Royal Roads UniversityJ. Van Barneveld, British Columbia Ministry

of Environment, Lands and ParksY. Bédard, Ministère des Transports

du QuébecD. Belluck, Minnesota Department of

TransportY. Blouin, Ville de Cap-de-la-MadeleineR. Brecher, Global ToxC. Chong, University of GuelphA. Decréon, Ministère des Transports

du QuébecR. Delisle, Ministère des Transports

du QuébecG. Duval, Ministère des Transports

du QuébecA. El-Shaarawi, Environment CanadaA. Fraser, Environment CanadaM. Frénette, Ville de MontréalD. Gutzman, Environment CanadaG. Haines, New Brunswick Department

of Transportation

R. Hodgins, Ecoplans LimitedM. Kent, British Columbia Ministry

of Transportation and HighwaysC. MacQuarrie, New Brunswick

Department of TransportationB. Mason, City of TorontoL. McCarty, L.S. McCarty ConsultingG. McRae, Ontario Ministry of

TransportationT. Pollock, Environment CanadaM. Roberts, University of New BrunswickD. Rushton, Nova Scotia Department

of Transportation and Public WorksK. Solomon, University of GuelphR. Stemberger, Dartmouth College,

New HampshireT. Young, Clarkson University, New York

The Assessment Report was reviewedand approved by the Environment Canada/HealthCanada CEPA Management Committee.

The Assessment Report for Road Saltswas released in August 2000 for a 60-day publiccomment period. Following consideration ofcomments received, the Assessment Report wasrevised. A summary of the public comments andresponses from Environment Canada and HealthCanada is available by contacting the ExistingSubstances Branch (see address below) or on theInternet at:

www.ec.gc.ca/cceb1/ese/eng/psap.htm

Copies of this Assessment Report areavailable upon request from:

Inquiry CentreEnvironment CanadaMain Floor, Place Vincent Massey351 St. Joseph Blvd.Hull, QuebecK1A 0H3

or on the Internet at:

www.ec.gc.ca/cceb1/ese/eng/psap.htm



Unpublished supporting documentation,which presents additional information, is availableupon request from:

Existing Substances BranchEnvironment Canada14th Floor, Place Vincent Massey351 St. Joseph Blvd.Hull, QuebecK1A 0H3e-mail: [email protected]

2.1 Identity, properties, productionand sources

2.1.1 Substance characterization

Road salts can refer to any salt applied toroadways for roadway maintenance. As outlinedbelow, road salts are mainly used in Canada asde-icing and anti-icing agents during winter roadmaintenance, but smaller quantities are also usedas dust suppressants. “Salt” can refer to anycompound consisting of the cation from a baseand the anion from an acid and which is readilydissociated in water. While sodium chloride(NaCl) is by far the most frequently used road saltin Canada, other inorganic salts used in Canadainclude calcium chloride (CaCl2), magnesiumchloride (MgCl2) and potassium chloride (KCl).Few additives are used in Canada as part ofroad salt formulations; sodium ferrocyanide(Na4Fe(CN)6·10H2O) is the only product regularlyadded as an anti-caking agent in Canada. Organicsalts and a few other products are essentially usedin Canada for maintenance of airports and forplane de-icing or limited roadway trials.Abrasives such as sand are used for winterroadway maintenance; while these are not salts,chloride salts are frequently mixed with abrasives.Accordingly, abrasives blended with salts can bea source of chloride salts to the environment.

This assessment focuses on the inorganicchloride salts used for roadway maintenance.This is based on the following considerations:

• The original recommendations of theMinisters’ Expert Advisory Panel called foran ecological assessment of the mostcommonly used de-icing chloride salts.

• Sodium chloride and, to a lesser extent,calcium chloride are by far the mostcommonly used salts on roadways in Canada.

• All inorganic chloride salts have broadlysimilar behaviour and effects in theenvironment. Notably, effects related to thetoxicity of chloride can be considered todepend on the cumulative input of all chloridesalts.

In addition, public concerns have beenexpressed regarding the use of ferrocyanide saltsin formulations of road salts, notably given that,in solution, they can photolyse to yield freecyanide ions, which are highly toxic to aquaticorganisms. Since ferrocyanides are commonlyused in road salt formulations, their entry,exposure and effects are also considered as partof the assessment of road salts.

Organic salts are not used in Canada orare used in specific circumstances, such as atairports (rather than on roadways), and they arenot assessed in this report. While abrasives areused in large quantities in Canada, the natureand potential for environmental effects of thesecompounds are distinct from those of road saltsand are not considered in this assessment.

Limited data were available for parkinglots, industrial, commercial and other privateproperties. However, the assessment was basedessentially on uses and releases from publicroadways, snow disposal sites and patrol yards.

2.1.2 Physical and chemical properties

The Chemical Abstracts Service (CAS) registrynumbers and physical-chemical properties offour inorganic salts (sodium chloride, calciumchloride, magnesium chloride and potassiumchloride) and ferrocyanides used as road salts aregiven in Table 1.


2.0 SUMMARY OF BACKGROUND INFORMATION

FOR THE ASSESSMENT OF ROAD SALTS

TABLE 1 Physical-chemical properties of road salts1

Substance CAS No. Specific Molecular Eutectic2 Working3 Waterapplication weight temperature temperature solubility

(°C) (°C) (g/100mL)(°C)


Sodiumchloride,NaCl

Calciumchloride,CaCl2

Mixture ofsodium/calciumchloride (80/20 mix)

Magnesiumchloride,MgCl2

Mixture ofsodium/magnesiumchloride(80/20 mix)

Potassiumchloride, KCl

Sodiumferrocyanide,Na4Fe(CN)6·10H2O

Ferricferrocyanide,Fe4[Fe(CN)6]3

1 Physical-chemical properties are from Weast et al. (1989) and Chang et al. (1994).2 Eutectic temperature refers to the lowest temperature at which the substance will melt ice.3 Working temperature refers to the lowest effective de-icing temperature.

7647-14-5

10043-52-4

7786-30-3

7447-40-7

13601-19-9

14038-43-8

Road de-icer and anti-icer, de-icing additive for sand

Road de-icer, de-icing additive, anti-icer, prewetter, dust suppressant, road construction

Road de-icer, road anti-icer

Road de-icer, de-icing additive, road anti-icer, dust suppressant

Road de-icer

Alternative road de-icer

Anti-caking additive

Anti-caking additive

58.44

110.99

95.21

74.55

484.07

859.25

–21

–51.1

n.a.

–33.3

n.a.

–10.5

0 to –15

<–23

–12

–15

<–15

–3.89

35.7 (0)39.12 (100)

37.1 (0)42.5 (20)

54.25 (20)72.7 (100)

56.7 (100)

31.85 (20) 156.5 (98)

insoluble

The eutectic temperature is the lowestfreezing temperature that can be achieved forwater by adding a given salt to it. The greaterthe difference between the ambient temperatureand eutectic temperature, the higher the rate ofmelting (OECD Scientific Expert Group, 1989).Thus, from the salts presented in Table 1, calciumchloride would produce the highest meltingrate. The rate of reaction is observed to beapproximately the same for sodium chloride andcalcium chloride at temperatures between -1 and -4°C; from -5°C downwards, however, sodiumchloride acts more slowly than calcium chloridewhen equal quantities are applied (OECDScientific Expert Group, 1989).

Salinity is defined as the total dissolvedsolids in water after all carbonates have beenconverted to oxides, all bromides and iodideshave been replaced by chlorides and all organicmatter has been oxidized (Stumm and Morgan,1981). Since road salts in this assessment arechloride salts, chloride (Cl–) is the principalcontributing anion to salinity resulting from theapplication of these salts, but other contributinganions in the environment include bicarbonate(HCO3

–), carbonate (CO32–) and sulphate (SO4

2–).Cations that contribute significantly to salinityinclude calcium (Ca2+), magnesium (Mg2+),sodium (Na+) and potassium (K+). Salinity isclosely related to the total halide concentration,which is often called chlorinity. The relationbetween the two is described by the empiricallyderived Knudson equation (Mayer et al., 1999).

Ferrocyanide is a very complex anionof limited solubility composed of a central ironatom surrounded by an octahedral configurationof cyanide ligands (Letts, 2000a).

2.1.3 Production and use

Road salts (mostly sodium chloride) have beenused as ice-disbonding and ice-melting agents inCanada since the 1940s (Perchanok et al., 1991).

2.1.3.1 Sodium chloride

The predominant chloride salt used as a de-icerin North America is sodium chloride, which iscomposed of about 40% sodium and 60%chloride by weight. Trace elements, includingtrace metals, may represent up to 5% of thetotal salt weight. Substances potentially presentinclude phosphorus (14–26 mg/kg), sulphur(6.78–4200 mg/kg), nitrogen (6.78–4200 mg/kg),copper (0–14 mg/kg) and zinc (0.02–0.68 mg/kg)(MDOT, 1993).

World production of sodium chloridetotalled 189 000 kilotonnes in 1995. The largestproducers are the United States (21%), China(15%), Germany (8%) and Canada (7%) (NaturalResources Canada, 1998). The largest globalmarket for sodium chloride is the chemicalindustry (60%), followed by table salt (20%) androad de-icing (10%) (CIS, 1994).

As of 1993, there were 12 manufacturersof sodium chloride in Canada, with 24 plants inseven provinces. Total nameplate capacity was13 645 kilotonnes per year, and total domesticproduction was 10 895 kilotonnes. A further1010 kilotonnes were imported from the UnitedStates and Mexico, making a total supply of11 905 kilotonnes. Of this, Canada exported3106 kilotonnes, mainly to the United States.Canadian domestic demand was 8799 kilotonnes(CIS, 1994). The largest market for sodiumchloride in Canada is snow and ice control,which accounts for about half the domesticdemand (4240 kilotonnes in 1993). Total amountsused for de-icing fluctuate from year to year,depending on weather conditions (CIS, 1994).

2.1.3.2 Calcium chloride

Calcium chloride is the second most commonlyused road salt in North America. Calcium chlorideis the leading chemical used for dust suppressionin Canada. Liquid calcium chloride is appliedprimarily to gravel roads to consolidateaggregates and control dust. Calcium chlorideis also used to pre-wet salt or sand in winter



highway maintenance and to stabilize road basemixtures after pulverization. Of these three uses,dust control accounts for approximately 97%of total use (Morin and Perchanok, 2000). Thequantity of calcium chloride used for winter roadmaintenance may increase as different agenciesstart experimenting with pre-wetting techniques.

In 1995, there were three calciumchloride producers in Canada, with one plantin Ontario and four brine wells in Alberta.Total nameplate capacity for calcium chloride inCanada for that year was 629 kilotonnes. At thattime, domestic production totalled 399 kilotonnesand 25 kilotonnes were imported, creating a totalsupply of 424 kilotonnes. Of this, 156 kilotonneswere exported, while Canadian domestic demandwas 268 kilotonnes. The total amount used in1995 in road dust suppression and in roadconstruction was reported at 201 kilotonnes(CIS, 1996).

2.1.3.3 Mixture of sodium chloride andcalcium chloride

Sodium chloride pre-wetted with calcium chloridebrine has been recommended for reducing totalsalt applications (Gooding and Bodnarchuk,1994). The total amount used for roads in Canadais included in the above estimates for sodiumchloride and calcium chloride.

2.1.3.4 Magnesium chloride

Although not recommended for dust controlon roads, magnesium chloride is an alternativedust suppressant for use on material piles, roadshoulders or material transfer ponds (OMEE,1993). Using data provided by industry,Morin and Perchanok (2000) estimated thatapproximately 25 000–35 000 tonnes ofmagnesium chloride are used annually forroadway de-icing activities in Canada. Noinformation was found concerning magnesiumchloride production volumes in Canada, butvolumes are likely quite low.

2.1.3.5 Mixture of sodium chloride andmagnesium chloride

While sodium chloride–based de-icing productscan contain magnesium chloride as well as somecorrosion-inhibiting additives (MDOT, 1993),there is no evidence of their current use inCanada.

2.1.3.6 Potassium chloride

There is limited use of potassium chloride in roadde-icing chemicals; however, potash mine tailingscontaining 0.5–2% potassium chloride (the rest isprimarily sodium chloride) are applied to someCanadian roads for de-icing (CFI, 1997). The totalamount of potassium chloride used annually forroadway de-icing in Canada has been estimated tobe 2000 tonnes (Morin and Perchanok, 2000) and3300 tonnes (CFI, 1997). It is estimated that totalCanadian production of potassium chloride in1999 was 13.6 million tonnes, and the potashproducing industry operated at 61% of productioncapacity (Prud’homme, 2000). Most of the potashproduced was used for agricultural purposes.

2.1.3.7 Brines

Oil field brine is a traditional dust suppressantcontaining calcium, magnesium, sodium andchloride. Oil field brine is obtained as formationwater generated from oil exploration operations atseveral locations in southwestern Ontario (OMEE,1993).

2.1.3.8 Ferrocyanide

To prevent the clumping of chloride saltsduring storage and de-icing operations, sodiumferrocyanide and ferric ferrocyanide can beadded to these salts. In Canada, sodiumferrocyanide is added to sodium chloride inamounts of 30–124 mg/kg (Letts, 2000a).

Sodium ferrocyanide is not currentlyproduced in Canada. However, approximately300–350 tonnes per year are imported fromEuropean and Asian manufacturers for use as ananti-caking agent. Road salts containing sodium

ferrocyanide are used in Ontario, Quebec and theAtlantic provinces. From Manitoba to the interiorof British Columbia, the by-product salt fromSaskatchewan potash mines that is used as roadsalt is not treated with sodium ferrocyanide. Onthe west coast, rock salt imported from Chile istreated with sodium ferrocyanide at a rate of60 mg/kg. The quantity of ferrocyanide used invarious jurisdictions can be estimated from theamount of salt used and its average ferrocyanidecontent.

2.1.4 Sources and releases

2.1.4.1 Natural sources

Natural sources of sodium chloride, calciumchloride, magnesium chloride and potassiumchloride salts in the environment includeweathering and erosion of rocks and soils,atmospheric precipitation and groundwaterdischarge. The composition of the rock material,relief and climate are important factors thatlargely control the rates of weathering anddissolution. In aquatic ecosystems, cations suchas calcium, magnesium and potassium are notuniquely associated with chlorides but can bederived from natural sources such as carbonatesfrom soils and bedrock in the watershed (Mayeret al., 1999). Igneous rocks and associated soilshave generally lower salt contents than shales andlimestones (Pringle et al., 1981). The contributionof salinity from atmospheric sources isparticularly important in the coastal maritimeregions. Areas that are dry with low rates ofprecipitation, such as the Canadian Prairies, havelandscapes that tend to concentrate salts in surfacesoils because of high net evapotranspiration fromthe land. Natural sources of salts in aquaticecosystems are reviewed by Evans and Frick(2001).

Sodium and chloride are stored invegetation, but they are also easily lost by ionuptake, leaf washing and litter decomposition.Chloride is an important inorganic solute ofosmotic regulation. The chloride concentration inthe environment is not affected by chemical orbiological reactions. Chloride does not adsorb

onto particulate matter. Potassium ion tends tobe the most constant quantity in animal cells.Calcium is the most reactive ion of all the majorcations contributing to salinity; it is required as anutrient in higher plants and is one of the basicinorganic elements present in algae. Magnesiumis important in enzymatic transformations and isrequired by chlorophyll-bearing plants.

While ferrocyanides do not occurnaturally in the environment, cyanide compoundsare naturally occurring substances reported fromover 2000 plant species (Conn, 1980). Sodiumferrocyanide and ferric ferrocyanide are highlystable and relatively immobile in the environment.When they are exposed to sunlight, theferrocyanide complex decomposes and releasescyanide ions (CN–) (Meeussen et al., 1992a).Under natural conditions, the cyanide ionhydrolyses to the volatile hydrogen cyanide(HCN) molecule (Shifrin et al., 1996).

2.1.4.2 Anthropogenic sources

Anthropogenic sources of inorganic chloride saltsinclude domestic sewage and industrial processes(Sonzogni et al., 1983), such as effluent from thechemical and petrochemical industry (Johnsonand Kauss, 1991), gas manufacturing and acidmine drainage (U.S. EPA, 1973; Pringle et al.,1981). Once these substances enter theenvironment, they remain in aqueous phaseuntil their concentrations exceed their solubilityproducts, when crystallization and subsequentsedimentation of mineral salts might occur.

Salts may be released to the environmentduring their production, mining, mixing, and bulktransportation and storage. Since these industrialsources are not all limited to ultimate use as roadsalts, they are not considered further in thisassessment.

Ultimately, all road salts enter theenvironment as a result of:

• storage at patrol yards (including losses fromstorage piles and during handling);



• roadway application (at the time ofapplication as well as subsequent movementof the salts off the roadways); and

• disposal of waste snow.

Releases are therefore associated with both pointsources (storage and snow disposal) and linesources (roadway application). The followingsections review material pertinent to entry into theenvironment, notably loadings (amounts appliedto roadways), disposal of waste snow and patrolyards.

2.2 Road salt loadings

This section summarizes findings in a reportprepared by Morin and Perchanok (2000), whichcharacterizes the use of road salts in Canada.Information describing application rates andannual mass loadings per highway length andper geographic district area is provided. Thesedetailed, geographically based data were usedto calculate exposure and to support subsequentsections of the assessment. Data in this sectionfocus on the use of sodium chloride and calciumchloride salts. Magnesium chloride and potassiumchloride salts are used to a lesser extent. Industryrepresentatives estimated that approximately25 000–35 000 tonnes of magnesium chlorideand 2000 tonnes of potassium chloride are usedfor roadway de-icing in Canada annually.

2.2.1 Surveys

Information on the use of road salts in Canadawas obtained from surveys conducted by theTransportation Association of Canada, the OntarioMinistry of Transportation and EnvironmentCanada. The Environment Canada surveyincluded data from provinces, municipalities andthe private sector. Details on the methodologyused in the surveys are presented in Morin andPerchanok (2000).

2.2.1.1 Provinces

Information on salt loadings and road networklength was obtained from all provincial andterritorial departments of transportation. The1997–98 winter had the most complete datanationwide; these data were used to estimate totalsalt use. Comparisons between data from saltsuppliers and data from provincial agenciessuggest that the quantities reported are reliable(Morin and Perchanok, 2000).

2.2.1.2 Municipalities and regionalgovernments

Information on salt use and road network lengthwas obtained from 104 municipalities acrossCanada. Data on salt use were also obtained fromall regional municipalities and counties inOntario. Some respondents provided informationon salt use and road network length for each ofthe past 5 years; others provided information onsalt use for the 1997–98 winter only.

2.2.1.3 Salt industry data

Salt suppliers provided purchase and tender datafor 500 municipalities in Canada. A comparisonbetween tender data and purchase data indicates astrong concordance between the two.

2.2.1.4 Municipal estimates

Information on salt use by road type is availableonly for 104 municipalities. Even by combiningdata obtained from municipalities (104) andsalt suppliers (500 municipalities), this does notrepresent total municipal salt use. Estimates weremade to account for salt use by municipalitieswhere no data were submitted.

Population was chosen as the variableto predict municipal salt use. This was done byderiving an average annual loading per personfor municipalities with survey or purchase/tenderdata and multiplying this by the population of all

other municipalities in a maintenance district. Bysumming known municipal salt use and estimatedmunicipal salt use, it was possible to estimatetotal municipal salt use by maintenance district.After estimates were made, salt use was mappedby maintenance district.

2.2.1.5 Provincial and municipal use of dustsuppressant salts

The major producer of calcium chloride providedmarket estimates of calcium chloride use onroadways by province. Recommended applicationrates for calcium chloride were obtained fromindustry. Rates vary depending on the type ofgravel road, traffic and policies regardingroadway maintenance. Minimum and maximumloadings for a gravel road 7.4 m wide werecalculated.

2.2.2 Results

Detailed results are presented in Morin andPerchanok (2000).

Using survey information, maps depictingthe distribution of provincial roads were created.A provincial road network area fraction wascalculated and mapped (Figure 1 indicates thearea of saltable roads maintained by provincialagencies per district area). This map can be usedto estimate salt loadings to surface water indrainage areas of a similar size and shape.Districts with the highest proportion of saltableroad to district area are in parts of the Atlanticprovinces, southern Quebec, southern Ontario andparts of British Columbia.


FIGURE 1 Provincial road network area fraction (from Morin and Perchanok, 2000)

2.2.2.1 Sodium chloride

2.2.2.1.1 Recommended application rates

Information on recommended provincialapplication rates is depicted in Figure 2; rates forselect municipalities are presented in Figure 3.Application rates cannot be used to quantifyannual loadings; they provide information on thequantity of salt applied with each application, notthe number of applications and resulting totalmass applied. Loadings refer to average totalannual mass, not application rates.

2.2.2.1.2 Loadings

Detailed data were analysed separately forprovinces, municipalities and regionalgovernments, including calculations of mass ofsalts per length and surface area of two-lanesaltable roads and per area of maintenance district(Morin and Perchanok, 2000). Areas with highestloadings were in central and southern Ontarioand Quebec, followed by the Atlantic provinces;lowest loadings were in the Prairies.

Municipal loadings were combined withprovincial and territorial loadings to estimate totalsalt use by maintenance district for the 1997–98winter season (Figure 4); it is estimated that4 418 462 tonnes of sodium chloride were usedas de-icers on roadways in Canada.

The above estimates of total loadinginclude data on the use of road salts by municipal,provincial, territorial and regional governmentsonly. Private contractors, industry and agenciessuch as port and transit authorities also use roadsalts. Cheminfo (1999) estimated that the total useby commercial and industrial customers is 5–10%of the road salt market. A midpoint value of 7.5%was used to estimate the quantity of road saltsused by commercial and industrial sources inCanada (Table 2). If these quantities are included,it is estimated that 4 749 847 tonnes of sodiumchloride road salts were used during the 1997–98winter season.


FIGURE 2 Recommended provincial application rates for sodium chloride road salts, 1998(from Morin and Perchanok, 2000)


FIGURE 3 Average recommended municipal application rates for sodium chloride road salts,1998 (from Morin and Perchanok, 2000)

FIGURE 4 Total amount of sodium chloride used, by maintenance district (from Morin andPerchanok, 2000)

2.2.2.2 Calcium chloride

The quantity of calcium chloride used by provinceand territory is listed in Table 3. While Ontario isthe province where the largest quantity of calciumchloride is used on roadways, analyses by Morinand Perchanok (2000) indicate that Nova Scotia isthe province with the highest usage per unit areaof land.

2.2.2.3 Total chloride loadings

Estimates of the quantity of sodium chloride andcalcium chloride used on roadways in Canada arecombined to estimate total chloride use onroadways by province. Table 4 presents totalchloride use by province, based on total sodiumchloride loadings for the 1997–98 winter seasonand the estimated use of calcium chloride in atypical year. Mass of chloride used per area of

province was calculated to provide a basis onwhich to compare loadings. Thus, while Ontarioand Quebec are the provinces where the mostchloride is used on roadways, Nova Scotia has thehighest loading per unit area of province (Morinand Perchanok, 2000).

2.2.2.4 Historic trends

Changes in road salt loadings over time wereconsidered. A few agencies that maintainroadways provided long-term information ontotal salt use. While these data suggest that totalroad salt loadings have generally increased overtime, they also indicate that there are annualfluctuations. Furthermore, it is difficult toconclude that loadings have increased withoutinformation on the length of the road networkthat is maintained.


TABLE 2 Total loading of sodium chloride road salt, winter 1997–98 (from Morin and Perchanok, 2000)

Province Loading of sodium chloride road salt (tonnes)Provincial Estimated Total Estimated Total

county + provincial+ commercial + municipal county + industrial1

municipal British Columbia 83 458 48 199 131 657 9 874 141 531Alberta 101 063 67 870 168 933 12 670 181 603Saskatchewan 44 001 4 844 48 845 3 663 52 508Manitoba 36 780 28 256 65 036 4 878 69 914Ontario 592 932 1 123 653 1 716 585 128 744 1 845 329Quebec 609 550 827 205 1 436 755 107 757 1 544 512New Brunswick 189 093 75 826 264 919 19 869 284 788Prince Edward Island 23 051 4 300 27 351 2 051 29 402Nova Scotia 270 105 77 761 347 866 26 090 373 956Newfoundland 159 200 47 558 206 758 15 507 222 265Yukon 1 791 120 1 911 143 2 054Northwest Territories2 1 846 0 1 846 138 1 984

Total 2 112 870 2 305 592 4 418 462 331 385 4 749 847

1 Commercial and industrial road salt use is assumed to be at 7.5% of provincial and municipal use (7.5% based on estimateby Cheminfo, 1999).

2 Including Nunavut.

One way to roughly ascertain if annualroad salt loadings have changed is by comparingsurvey data compiled by the Salt Institute fromthe mid 1960s to early 1980s (Salt Institute,1964–1983) with data collected for the 5-yearsurvey period (1993–94 to 1997–98). Thiscomparison suggests that road salt loadingsper two-lane-kilometre have not decreased.

While some agencies used similar quantitiesof salt per two-lane-kilometre over both timeperiods, road salt loadings by other agenciesincreased slightly from the early 1980s to mid1990s. Figure 5 shows average provincialloadings per two-lane-kilometre of provincialroad for Ontario, Quebec, New Brunswick andNova Scotia.


TABLE 3 Estimated quantities of calcium chloride used on roadways in a typical year (from Morinand Perchanok, 2000)

Province Amount of calcium chloride (tonnes)British Columbia 12 573Alberta 6 985Saskatchewan 2 794Manitoba 6 985Ontario 45 405Quebec 20 956New Brunswick 1 746Prince Edward Island 349Nova Scotia 5 169Newfoundland 839Yukon 1 397Northwest Territories1 2 794Total 107 992


TABLE 4 Total loadings of chloride (from Morin and Perchanok, 2000)

Province Chloride loading (tonnes)British Columbia 93 900Alberta 114 641Saskatchewan 33 642Manitoba 46 880Ontario 1 148 570Quebec 950 444New Brunswick 173 896Prince Edward Island 18 061Nova Scotia 230 182Newfoundland 135 384Yukon Territory 2 139Northwest Territories1 2 989Total 2 950 728



The comparison and merger of data fromdifferent sources for different time periods wereassessed by comparing the correspondence ofdifferent data sets for provincial agencies inOntario and Nova Scotia. This type of comparisoncould not be done for the other provinces, becauseoverlapping data were not available. When datacould be compared, however, there was aconcurrence between the quantities reported.For example, data obtained from the SaltInstitute indicate that the Ontario Ministry ofTransportation used 418 997, 486 648 and402 346 tonnes of sodium chloride for the wintersof 1978–79, 1981–82 and 1982–83, respectively(Salt Institute, 1964–1983). Data obtained fromthe Ontario Ministry of Transportation indicatethat it used 415 381, 486 648 and 402 346 tonnesof sodium chloride for the same years,respectively. Data obtained from the Salt Instituteand Nova Scotia Department of Transportationand Public Works both indicate that 201 989 and129 453 tonnes of sodium chloride were usedduring the winters of 1981–82 and 1982–83,respectively. While this comparison is limited

in scope, it suggests that data obtained from theSalt Institute and provincial agencies concur.

2.2.3 Reliability of loadings data

Although a concerted effort was made to obtaindata from the most accurate sources, errors indata and calculations are possible. Quantitiesof materials applied may have been inaccuratelyrecorded by agencies, or there may be non-reported salt usage. There could also beinconsistencies in the data that were provided.For example, some provinces purchase salts foruse by municipalities or maintain municipalroads. Hence, municipal data may be includedin the provincial loadings for certain districts.Another problem could be that survey datagenerally refer to salt purchased and not salt used.While all the salts purchased will eventually beused, there is no way of determining the quantityof salt that is applied each year. Potential errorswith data from salt suppliers were assessed bycorrelating provincial bids and usage. Overall,both data sets are highly correlated.

FIGURE 5 Historical salt use by provincial agencies (based on surveys conducted by the Salt Institute,1964–1983, and Morin and Perchanok, 2000)

One possible source of error could bethe use of 1997–98 data to characterize saltloadings. While these quantities are a bestestimate, available provincial data indicate thatloadings varied by year. A comparison of thetotal salt loadings per unit length of road for allagencies that provided 5 years of data suggeststhat the 1997–98 winter season is not atypical(Figure 5). To validate assumptions that the1997–98 winter season was representative, twostatistical analyses were done (Collins, 2000).Data were analysed using an analysis of variance(ANOVA), and, since data were not normallydistributed, they were also analysed using thenon-parametric Friedman’s two-way analysis byranks. Whenever there was a significant differenceamong years, Ryan’s step-down procedure testedfor differences among years. Results of theseanalyses indicate that the loadings were somewhatlower for the 1997–98 period (Collins, 2000).

A potential limitation of analyses maybe the use of a “saltable” category to characterizeroad salt loadings. Since highways generallyreceive higher loadings, it could be argued thatthe saltable category underestimates salt loadingson high-volume roads. Another limitation may bethe use of population to estimate salt loadings formunicipalities with no data. While there is a goodcorrelation between municipal population and saltuse at the provincial level, the strength of thisrelation varies by maintenance district.

Despite these potential limitations,data presented are good indicators of thequantities of salt used on roadways in Canada.This is particularly so when considering that thereis a difference of 9.4% (417 501 tonnes) betweenknown and estimated usage of sodium chloride bymunicipal, county and provincial agencies for the1997–98 winter.

2.2.4 Summary

The estimated annual use of sodium chlorideand calcium chloride road salts in Canada wasdetermined from surveys of provincial, territorialand municipal agencies, industry data and

estimates based on population. It is estimatedthat approximately 4.75 million tonnes of sodiumchloride de-icers were used in the 1997–98winter and that 110 000 tonnes of calciumchloride are used on roadways in a typical year.When combined, it is estimated that 2.95 milliontonnes of chloride were used on roads in Canadaduring the 1997–98 year. It is recognized thatamounts used will vary on a yearly basis, notablybecause of variation in climatic conditions.

Annual loadings vary geographically,with the highest loadings on a length of road basisoccurring in Ontario and Quebec, intermediateloadings in the Atlantic provinces and lowestloadings in the western provinces. Overall, datapresented in this section indicate that road saltloadings have not decreased over the past 20years. Furthermore, data for some provincessuggest that loadings have increased since thelate 1970s (Figure 5).

2.3 Snow disposal

Delisle and Dériger (2000) reviewed thephysicochemical and ecotoxicologicalcharacteristics of roadside snow and the differentmethods used for removing snow from roads andsidewalks.

2.3.1 Characterization of snowfrom roadways

Road salts contribute to the presence of sodium,chloride and cyanide ions in snow. Other majorcontaminants also found in urban snow includedebris, suspended solids, oil and grease, andmetals (lead, manganese, iron, chromium)(Gouvernement du Québec, 1997). Averageconcentrations of chlorides in roadside snow or itsmeltwater for certain municipalities in Quebecranged from 3.8 to 5689 mg/L (see Delisle andDériger, 2000).

Chenevier (1997) analysed thephysicochemical parameters of roadside snowfrom Montréal’s primary and secondary streets



in 1997. Concentrations for all parameterswere higher for the primary streets. Averageconcentrations of chloride were 3115 mg/L forsecondary streets and 5066 mg/L for primarystreets (Delisle and Dériger, 2000). Thisdifference is probably due to the higher frequencyor rates of application of road salts. A similarstudy by Delisle et al. (1997) monitored thephysicochemical characteristics of roadsidesnow for two snowstorms in January 1997.Mean chloride concentrations for samples takenduring the first and second storms were 7716 and3663 mg/L, respectively. Chloride concentrationsfor the 24 samples taken during both stormsranged between 1366 and 18 230 mg/L.

The characteristics of roadside snowdepend on a wide range of factors, includingcommunity size, the land occupancy factor, trafficdensity, quantities of de-icing salt and abrasivesused, and duration of snow clearing (Malmqvist,1985; Delisle et al., 1995). Contaminant contentgenerally changes with time. For example,concentrations of chloride were 530 mg/L after72 hours for snow from residential city streetsand 7496 mg/L after 72 hours for snow fromcommercial city streets (Delisle and Leduc, 1987).Concentrations subsequently diminished to590 mg/L after more than 1 week, presumablydue to loss of ions though meltwater.

2.3.2 Volumes of snow in Canadian cities

Removal and disposal of snow may be requiredwhen the accumulation of snow on or alongroadways may hamper traffic or safety. As such,the quantity of snow to be removed and disposedof depends on the volume of snow and the extentof urban development (Table 5). Given the largeamount of snowfall in Montréal (annual snowfallin southern Quebec is between 200 and 350 cm)and the size and density of the city, about 11.258million cubic metres of snow were brought tosnow disposal sites during the winter of 1997–98.This quantity can be compared with 1.5 millioncubic metres for Toronto (before municipalamalgamation) and a total of about 30 million

cubic metres of snow for all municipalitiesin Quebec.

2.3.3 Snow clearing and snow disposalmethods

Clearing of snow generally involves plowingsnow to the side of the roadway. In cities, snowclearing can begin as snowfall reaches 2.5 cm,with the snow being pushed to the sides ofstreets and onto sidewalks. However, whenthere is considerable accumulation (more than10 cm) or when necessary, snow is cleared andtransported to various disposal sites (City ofMontréal, 1998). The various snow disposalmethods have been reviewed by Delisle andDériger (2000) and can be grouped into threecategories, as described below.

2.3.3.1 Methods not involving snow removal

Methods for clearing snow from the roadwaywithout transporting it to snow disposal sitestypically involve plowing snow to the sideof the road or blowing it onto land adjacent tothe roadway. These methods are generally noteffective in areas with a high land occupancyfactor (Delisle, 1994). These methods are,however, the most prevalent approach to theclearing of snow from roadways in non-urbanareas. This type of approach contributes topotential impacts of salts on roadside soils,vegetation, surface water and groundwater. Datacollected by Watson (2000) characterized chlorideconcentrations in ponds and wetlands adjacent toroadways in southern Ontario. Results of thisstudy indicate that the chloride concentrationsare variable, but concentrations greater than4000 mg/L were observed. This study indicatesthat runoff from roadways can adversely affectaquatic environments adjacent to roadways.

A study by Delisle (1999), whichassociated elevated chloride concentrations inmunicipal wells with the use of road salts, alsoindicated that high chloride concentrations wereobserved in an area where ditch waters were left

to percolate through the soils. Concentrations ofchloride at three municipal production wells in thestudy area increased by 102, 116 and 145 mg/Lbetween 1983 and 1994.

2.3.3.2 Methods involving release intowaterways after processing or treatingmeltwater

Roadway snow can be transported to snowdisposal sites where snow melts and the meltwateris treated. This typically involves dumping snowat surface sites or in quarries where runoff ischannelled to treatment facilities. Generally, snowdisposal sites are located on impermeable orslightly permeable ground or must be equippedwith a geotextile membrane. Some sites are alsoequipped with sedimentation facilities or aredesigned to direct the meltwater towards awastewater treatment system (Gouvernement duQuébec, 1991). These sites should not be locatednext to watercourses that could be affectedby runoff.

Pinard et al. (1989) characterizedchloride concentrations in runoff from snowdisposal sites. This study indicated that only 2%of the salt spread on city streets was present inmeltwater from snow disposal sites, with mostof the salt likely released to the environmentfrom the roadway or roadside. This concurs withDelisle and Leduc (1987), who indicated thatchloride concentrations in roadside snow initiallyincrease then decrease with increasing time. Theconcentration of chloride in snow removed fromroadways will, however, vary by street type;concentrations in snow collected from primarystreets can be an order of magnitude higher thanconcentrations in snow from secondary streets.

While the percentage of salt presentin snow transferred to snow depots may be low(e.g., 2%), the concentration of chloride in runoffis still elevated. A study by Péloquin (1993)indicated that the average chloride concentrationin meltwater from a snow disposal site was414 mg/L. Pinard et al. (1989) monitored chloride


TABLE 5 Total volume of waste snow and quantity of salt used in certain Canadian cities, winter1997–98 (from Delisle and Dériger, 2000)

City Total snow volume Quantity of salt used(m3)1 (tonnes)

Vancouver <100 000 2 254Calgary 320 000 20 428Regina 150 000 2 067Winnipeg 175 000 24 525Toronto (before amalgamation) 1 500 000 17 884Ottawa-Carleton 1 220 000 68 000Montréal 11 258 000 60 000Québec City 3 000 000 35 000Moncton 159 000 12 695Halifax N/A 41 679Charlottetown N/A 2 300St. John’s N/A 21 530

1 These quantities are approximate and vary yearly with snowfall amount. N/A = not available.2 Plus 100 000 tonnes of a sand/salt mix containing 5% NaCl.3 Plus 39 000 tonnes of a sand/gravel mix (1:9 for sidewalks and 5:5 for streets) during freezing rain.

2

3


concentrations in runoff from a Québec Citysnow disposal site from April 18 to the end ofJune 1988. Concentrations in runoff ranged fromapproximately 100 to 1100 mg/L. Concentrationswere highest in the early sampling and graduallydecreased throughout the spring. It is not knownif more elevated chloride concentrations werepresent in runoff prior to the monitoring programor if the concentrations increased throughout thesummer due to a decrease in the volume ofrunoff water.

The potential impact of snowmelt ongroundwater quality will vary by disposal site.A study described in Morin (2000) indicatesthat chloride concentrations between 233 and1820 mg/L were measured at monitoring wellsinstalled to assess the impact of a snow disposalsite on shallow groundwater quality.

Other agencies (e.g., City of Montréal)dump snow through chutes linked to municipalsewer networks (Godbout, 1996; Couture, 1997).Certain agencies use snow melters. Theunderlying principle of melters is to dump waterrather than snow into the sewer system. Melterscan be stationary or mobile. Snow is dumped bytrucks into preheated water tanks equipped withoil or gas burners. The City of Toronto uses thistechnique occasionally. So-called “geothermal”snow melting, using geological formations as anatural storage reservoir for water during thesummer and keeping the water sufficiently hot tomelt the snow during the winter, has been usedsince January 1998 in Cap-Rouge, Quebec(Bilodeau, 1999).

2.3.3.3 Methods involving release to surfacewater without water treatment

This method of disposing of snow involvesdumping snow directly into a waterway or ontoits banks. Snow can also be dumped down sewerchutes that are not linked to treatment plants.This results in the release of waste snow and anycontaminants to surface water directly (dumpinginto rivers or into the ocean), with some removal

of large debris (dumping onto banks) or withpossible dilution by stormwater (dumping intosewer chutes not linked to treatment plants).

2.3.4 Use of disposal methods by Canadianmunicipalities

The use of snow disposal approaches by selectedCanadian municipalities is shown in Table 6.The City of Montréal uses the greatest varietyof disposal methods. In 1997–98, the City ofMontréal used surface sites (7 sites), sewer chutes(11 sites), dumping into the St. Lawrence River(3 sites) and quarry dumps (2 sites) (City ofMontréal, 1998).

Surface sites are clearly the mostcommon method of disposal in Canada. Forexample, Regina uses between two and fivesurface disposal sites, depending on the year(City of Regina, 1997). In the case of theRegional Municipality of Ottawa-Carleton,several studies have been conducted todetermine the number of surface sites neededto accommodate snow removal in that region(McNeely Engineering, 1990; McNeely-TunnockLtd., 1995). A few municipalities (e.g., St. John’s,Halifax and Vancouver) dump snow into the seaat harbours. The direct dumping of snow tofresh surface water is restricted and will not bepermitted in Quebec as of 2002 (Gouvernementdu Québec, 2001).

In the past, municipalities such asMontréal blew snow onto private property andoccasionally used snow melters. Blowing ontoprivate land was virtually abandoned by Montréalbecause of social and political pressure, and snowmelters proved to be too expensive to operatebecause of high fuel costs (City of Montréal,1998). In certain cities like Halifax, largequantities of de-icing salt are used, but little snowis removed; roadway meltwater is channelled tostorm sewer systems (Delisle and Dériger, 2000).

2.3.5 Summary

In addition to inorganic ions from road salts,waste snow can contain a broad range of physicaland chemical contaminants. Chloride ions aredissolved and ultimately transported in meltwaterand are largely not affected by physical orbiological water treatment. Impacts depend ontotal amounts and concentrations of the ions atthe point of release of meltwater into surfacewater or into soil and groundwater. Whilechloride concentrations in waste snow and itsmeltwater can be quite variable, they can be ashigh as 18 000 mg/L. Average chlorideconcentrations in snow from streets in Montréalwere approximately 3000 and 5000 mg/L forsecondary and primary streets, respectively.Canadian municipalities use a variety of snowdisposal techniques, but surface sites are the mostcommon. While there is uncertainty regardingthe portion of salt that may be transported tosnow disposal sites, it is clear that runoff fromthese sites has elevated chloride concentrations.

2.4 Patrol yards

Patrol yards (also referred to as storage yardsor maintenance yards) are used to store roadmaintenance materials before their application toroadways. The following sections describe andcharacterize these facilities, outline environmentalexposure pathways and present measuredconcentrations from specific points around patrolyards. Most of the data are based on pre-1998standards for patrol yard design and the storageof road salts. While more effective designs arecurrently being promoted to reduce salt loss(TAC, 1999), current design and managementare largely similar to those considered in thisassessment. Data in the following sections weresummarized from a report prepared by Snodgrassand Morin (2000).

2.4.1 Number of patrol yards and quantityof materials stored

Data compiled by Morin and Perchanok (2000)were used to estimate the number of provincial


TABLE 6 Snow disposal methods used in certain Canadian cities, winter 1997–98 (from Delisle andDériger, 2000)

City Waste snow disposal (% of total snow volume)Surface Quarries Sewer Melters Waterways dumps chutes or sea

Vancouver 10 – – – 90Calgary 100 – – – –Regina 100 – – – –Winnipeg 100 – – – –Toronto 90 – – 10 –Ottawa-Carleton 100 – – – –Montréal 25 37.7 18 – 19.3Québec City 100 – – – –Moncton 100 – – – –Halifax – – – – 100St. John’s – – – – 100

and territorial patrol yards and the quantity ofroad salts and abrasives stored at these yards.There is no definite information on the numbersof patrol yards in Canada, but it is estimated thatthere are 1300 provincial patrol yards. Assumingthat total salt and abrasive use by jurisdictionsis distributed equally among patrol yards, theaverage quantities used at patrol yards wereestimated (Table 7). This table does notinclude municipal and county yards and storagefacilities used by private contractors and hencesignificantly underestimates the number ofpatrol yards in Canada. The average amount ofsalts and abrasives stored at patrol yards variesconsiderably across Canada. Table 7 suggests thatthe average amount of salts stored at patrol yardsin Ontario, Quebec and the Maritime provincesranges between 1300 and 3800 tonnes. Theaverage amount of abrasives for these provincesranges between 1000 and 8600 tonnes. Thenumber of patrol yards by district and theaverage tonnage of sodium chloride and abrasivesthat would be used by patrol yards for selected

provinces are presented in Snodgrass andMorin (2000).

Data for the winter of 1997–98 from 117patrol yards administered by the Ontario Ministryof Transportation indicate that the quantity ofsodium chloride stored at these patrol yardsranges between 45 and 21 400 tonnes per yard;the quantity of abrasives stored ranges between10 and 13 200 tonnes per yard.

2.4.2 Patrol yard design and management

Patrol yards can be located in a variety ofsettings, and patrol yard design and standardsvary substantially across Canada. Accordingly,the degree of protection against weathering variesconsiderably. Covered facilities used to store roadsalts can include domes/igloos, sheds and lean-tos. Doors or walls may or may not be present,and storage under an overpass can be consideredas covered by some agencies. Salt may be storedon asphalt or concrete pads or outside on a thick


TABLE 7 Number of provincial patrol yards and sodium chloride and abrasive use, by province(from Snodgrass and Morin, 2000)

Province No. of Total Total Tonnes Tonnes of patrol NaCl abrasives of NaCl/ abrasives/yards used used patrol yard patrol yard

British Columbia 166 83 458 720 717 503 4342Alberta 71 101 063 300 464 1423 4232Saskatchewan 86 44 001 30 939 512 360Manitoba 67 36 780 37 800 549 564Ontario 158 592 932 555 859 3753 3518Quebec 478 609 550 548 996 1275 1149New Brunswick 95 189 093 376 705 1990 3965Prince Edward Island 10 23 051 85 735 2305 8574Nova Scotia 82 270 105 81 943 3294 999Newfoundland 44 159 200 269 460 3618 6124Yukon 23 1 791 38 000 78 1652Northwest Territories 1 10 1 846 11 400 185 1140Total 1 290 2 112 870 3 058 017


plastic tarp and covered by another tarp. Whilestorage in covered facilities can protect againstprecipitation, which leaches the salt, there is nouniversal standard design for storage facilities.Use of best management practices that promotegood housekeeping at patrol yards is equallyimportant to reduce salt loss. The effectivenessof best practices, however, depends on the rigourwith which they are implemented.

Salts can be released from mixedabrasive/salt piles as well as from salt piles.A mixed abrasive/salt pile consists of either sandor gravel mixed with salt to prevent the abrasivefrom freezing and to keep the abrasive in aflowing state (NB DOE and DOT, 1978). Thepercentage of salt varies — most agencies have5% mixtures, but this can range between 2.5 and15% salt. A New Brunswick study (NB DOE andDOT, 1978) monitored the quantity and quality ofleachate from a 2000-tonne abrasive pile with a2.5% salt content. During the first year, 420 m3

of leachate passed through the monitoring system.Sodium and chloride concentrations in theleachate are depicted in Figure 6. The maximumconcentrations for sodium and chloride were37 000 and 66 000 mg/L, respectively. It wasestimated that total salt loss through leachingduring the first year was 18.2 tonnes, or over athird of total salt added to the pile, even though80% of the abrasive pile had already beenremoved for sanding operations by the end ofJanuary. Results of the second year also indicatedhigh concentrations in leachate from the pile.A total of 255 m3 of leachate was measured. Themaximum concentrations for sodium and chloridewere 49 000 and 82 000 mg/L, respectively.Considering these concentrations and the resultsof a survey suggesting that few provincialagencies cover mixed abrasive/salt piles(Snodgrass and Morin, 2000), the potential forsalt loss from mixed abrasive/salt piles is high.

Spillage during stockpiling andloading of spreaders is a major source of salt loss(TAC, 1999), as confirmed by a review ofelectromagnetic surveys done at patrol yards(Snodgrass and Morin, 2000). These surveys,which typically identify subsurface materials and

groundwater (within 6 m of the surface) that arehighly conductive (e.g., soils and groundwaterwith elevated levels of salts), indicated highconductivity in areas adjacent to driveways andaround storage facilities where salts are handled.

Another potential source ofcontamination is from washwater used to washwinter road maintenance vehicles. At patrol yardsin urban areas, the washwater is likely releasedto a municipal sewage treatment system; inrural areas, discharge is probably to a dry welland left to percolate through the soil. Washwatersamples from seven patrol yards had chlorideconcentrations between 3500 and 37 000 mg/L;the average concentration was 16 000 mg/L(Beck et al., 1994). This study also estimated thatbetween one-third and two-thirds of patrol yardsin Ontario discharge washwater to dry wells.Chloride concentrations in washwater can alsobe estimated by analysing chloride concentrationsin oil/water interceptors. These devices havebeen installed at some patrol yard garages toremove oil, grease and similar compoundsfrom washwater waste streams. Since oil/waterinterceptors do not remove salt, concentrationsin discharges will be the same as in the influent.A study by INTERA Consultants (1996)indicated that chloride concentrations in oil/waterinterceptors at three patrol yards ranged between1100 and 35 400 mg/L. These concentrationsroughly concur with those monitored by Becket al. (1994).

2.4.3 Estimation of releases of salts frompatrol yards

Estimates of releases of salts from patrol yardsare addressed in this section; details of theassumptions used to calculate these releasesare in Snodgrass and Morin (2000). Estimatesof releases to the environment were made byconsidering a hypothetical scenario of releasesfrom 190 active yards. Total estimated releasesfrom patrol yards were then compared with atotal roadway application of approximately590 000 tonnes annually. This hypotheticalscenario corresponds roughly to a major provinceor several smaller provinces as the aggregate basis


for the calculations.

Table 8 provides three estimates basedon the following scenarios: (1) a patrol yard withbest management practices where salt and mixedabrasive piles are stored indoors; (2) a patrol yardwhere salt piles are stored indoors and abrasivepiles are stored outdoors; and (3) a patrol yardwhere neither abrasive piles nor salt piles arestored indoors. Of the three scenarios, options 1and 2 are probably most representative of patrolyards. The last column of Table 8 estimates thepercentage of total road salt use that may be lostat patrol yards. Depending on the facility and saltmanagement options, 0.2–20% of total salt usecan be lost at patrol yards.

2.4.4 Salt concentrations at different areasaround patrol yards

The following sections provide an overview ofconcentrations observed at surface and subsurfaceendpoints at patrol yards.

2.4.4.1 Concentrations at surface points aroundpatrol yards

A limited database for surface drainagesystems around patrol yards is available fromenvironmental site assessments (ESAs) doneat some Ontario Ministry of Transportationpatrol yards (MTO, 1993–1999). Chlorideconcentrations at 29 yards where samples weretaken ranged between 4 and 4880 mg/L; the meanchloride concentration was 659 mg/L.

A similar type of study was done fora yard in Alberta. A sample taken from theevaporation pond located on this property had achloride concentration of 32 100 mg/L (AlbertaInfrastructure, 2000). A study by the Nova ScotiaDepartment of Transportation and Public Worksindicated that runoff and shallow groundwaterflowing from a patrol yard to a small brookincreased chloride concentrations in a pondlocated approximately 250 m downgradient fromless than 75 mg/L to approximately 1000 mg/L(Rushton, 1999).


FIGURE 6 Average daily concentration of leachate from salt-treated sand pile, 1975–76 (from NB DOEand DOT, 1978)

2.4.4.2 Concentrations in patrol yard well water

A study by the Ontario Ministry ofTransportation (MTO, 1997) graphed chlorideconcentrations from over 300 patrol yards inOntario over a 20- to 30-year period ending in1988. A 5-year period, 1983–1988, was selectedto characterize the range of concentrations thatcould be observed in patrol yard wells. Of theyards in the analysis, most (229 yards, 74%) hadchloride concentrations below 200 mg/L. Asmaller number of yards (82 yards, 26%) hadconcentrations above 200 mg/L, including30 yards (9%) with concentrations greater than500 mg/L and 14 yards (4%) with concentrationsgreater than 1000 mg/L.

ESAs done for patrol yards in theOntario Ministry of Transportation’s EasternRegion tested the quality of water in potablewater wells (MTO, 1993–1999). Concentrationsof chloride in samples from 36 wells rangedbetween 1.5 and 5050 mg/L; the meanconcentration was 722 mg/L. Sixty-four percentof the samples had chloride concentrations greaterthan 100 mg/L, and 50% of the samples hadchloride concentrations that exceeded 250 mg/L.Thirty-five percent of the samples hadconcentrations greater than 500 mg/L. The depthof the wells sampled ranged from 12 to 113 m,with an average depth of 39.2 m. Elevated

chloride concentrations at wells do not necessarilyindicate that the entire aquifer is contaminated.For example, faulty well construction can causesalts to flow down around the well casing to thegroundwater.

2.4.4.3 Concentrations in shallow groundwater

The same set of ESAs (MTO, 1993–1999)was also used to assess the quality of shallowgroundwater. At several patrol yards, monitoringwells were installed to depths between 0.26and 5.25 m and groundwater was sampled.Chloride concentrations ranged between 1 and24 000 mg/L; the mean chloride concentration forthe 102 samples was 2600 mg/L. A summary ofthe data indicated that 75% of the samples hadchloride concentrations greater than 100 mg/L and69% of the samples had chloride concentrationsthat exceeded 250 mg/L.

In similar types of analyses, chlorideconcentrations in groundwater at three Albertapatrol yards ranged between 26 and 26 400 mg/L(Alberta Infrastructure, 2000). Chlorideconcentrations in shallow groundwater at threepatrol yards managed by the Nova ScotiaDepartment of Transportation and Public Worksranged between 254 and 38 600 mg/L (Rushton,1999).


TABLE 8 Estimate of magnitude of salt loss at patrol yards (from Snodgrass and Morin, 2000)

Patrol yard type Salt loss from management Percentage of total area of 190 patrol yards road salt use

(tonnes/year) (590 000 tonnes)Option 1 — Best management practices 1 290 0.2patrol yards — indoor storage for mixed abrasive and salt pilesOption 2 — Patrol yards where salt piles 9 900 2are stored indoors and abrasive piles are stored outdoorsOption 3 — Patrol yards where neither 128 000 20abrasive nor salt piles are stored indoors

2.4.4.4 Concentrations in patrol yard soils

At several patrol yards subjected to ESAs, testpits and bore holes were dug and soil sampleswere analysed (MTO, 1993–1999). Chlorideconcentrations from 53 test pits ranged between 5and 14 500 µg/g soil; sodium concentrations for46 samples ranged between 39 and 13 100 µg/gsoil. The average concentrations for chlorideand sodium were 2100 and 2600 µg/g soil,respectively.

Chloride concentrations in soilsamples from 46 bore holes ranged between2 and 13 300 µg/g soil. The average chlorideconcentration was 1500 µg/g soil. Sodiumconcentrations for 98 samples ranged between86 and 6720 µg/g soil; the average sodiumconcentration was 870 µg/g. Bore holes werebetween 0.1 and 5 m deep, with an average depthof 1.6 m.

2.4.5 Off-site concentrations at patrol yards

A review of select case studies by Morin(2000) indicated that there have been high off-siteconcentrations of salts at several locationswhere road salts are stored. A few examplesare presented in the following paragraphs.

Arp (2001) examined how the presenceof a salt depot can be associated with elevatedlevels of salinity in surface waters. Sodiumconcentrations in areas surrounding the depotvaried by type (pond and stream water), locationand season. Sodium concentrations were highestat and near the salt depot, ranging from 3000to more than 10 000 mg/L. Runoff andseepage from the depot were quickly dilutedby flowing surface water next to the depot,with sodium concentrations decreasing toabout 100–300 mg/L. There were also seasonalvariations: salt concentrations near the saltdepot were lowest during snowmelt and the rainyfall season. During summer, salt concentrationsincreased with reduced precipitation andincreasing air temperature. Highest concentrationsoccurred during summer months. At this time, thehigh salt concentration envelope that surrounds

the salt depot spread into the adjacent pond andcaused elevated salt concentrations in the maindrainage stream.

Exposed abrasive piles and salt reservesstored in sheds that offered poor protection fromthe elements resulted in the contamination of themunicipal water supply for Heffley Creek, BritishColumbia (BC MoTH, 1993). Road salts andunprotected abrasive piles with a 3% salt mixturehad been stored at this site since 1973 (AGRA,1995). (Prior to 1975, concentrations of chloridehad been less than 5 mg/L.) In 1993, the B.C.Ministry of Transportation and Highwaysreceived an official complaint from theMunicipality of Heffley Creek that groundwaterfor one of its two wells had been contaminated byroad salts stored at the Heffley Creek patrol yard.Water quality tests indicated that chlorideconcentrations at this well were approximately1000 mg/L. By 1994, there were elevated chlorideconcentrations at both municipal wells locatedbetween 175 and 200 m from the patrol yard(Figure 7). Chloride concentrations eventuallypeaked at approximately 3200 mg/L in 1995.While chloride concentrations have decreasedsince 1996, well #1 still had chlorideconcentrations higher than 500 mg/L in 1998(AGRA, 1999).

In 1996, complaints were registered withthe Nova Scotia Departments of Transportationand Public Works and the Environment regardingsalt contamination of residential wells in BibleHill, Nova Scotia. Tests undertaken at three wellsthat supplied six duplexes indicated that chlorideand sodium concentrations ranged from 130 to640 mg/L (Rushton, 1999). The Bible Hill patrolyard, which is located relatively close to theseresidences, was determined to be the source ofsalt contamination. While road salts are stored ina salt dome, abrasive mixtures have traditionallybeen stored outside. The Nova Scotia Departmentof Transportation and Public Works indicatedthat groundwater contamination was typical oflocations where mixed abrasive/salt piles are leftuncovered (Rushton, 1999).


A study by Ohno (1990) conducted onwetlands near sand/salt storage sites in Maineshowed that concentrations of chloride andsodium at the affected sites were 2–3 ordersof magnitude higher than those of controlsites. The concentrations at the control siteswere 6 mg sodium/L and 8 mg chloride/L, whilethe concentrations at the affected sites rangedfrom 16 to 8663 mg sodium/L and from 14 to12 463 mg chloride/L.

Pinhook Bog, LaPorte County, Indiana,was adversely impacted by contamination from asalt storage pile from the late 1960s to the early1980s (Wilcox, 1982). Chloride concentrations atcontrol sites in 1980 and 1981 were 5–6 mg/L.This compares with a maximum single dailyreading for salt-impacted locations of 1468 mgchloride/L in 1979, 982 mg chloride/L in 1980and 570 mg chloride/L in 1981. The total chlorideinputs to the bog over the 10-year period whensalt storage occurred were estimated as follows:2.3 million kilograms from the salt pile,0.4 million kilograms from road salting and0.012 million kilograms from direct precipitation.

2.5 Environmental fate andpathways

Road salts can enter and move through theenvironment in their salt form or as theirdissociated ions. In aquatic systems, chloride saltsare present in dissociated form as chloride ionsand corresponding cations (sodium, potassium,calcium, magnesium). The circulation of chloridethrough the hydrological cycle is due mostly tophysical rather than chemical processes. Thechloride ions pass readily though soil, entergroundwater and eventually drain into surfacewaters. Because chloride ions are persistent andare entrained in the hydrological cycle, allchloride ions applied to roadways as road salts orreleased from patrol yards or disposal sites can beexpected to be ultimately found in surface water.This section presents information on the fate andtransportation of road salts in the environment andtheir potential pathways.

2.5.1 Fate/transport

Once road salts enter surface waters, theyremain in solution unless their concentrationsare very high and exceed their solubility, when


FIGURE 7 Chloride concentrations in Heffley Creek municipal wells (from AGRA, 1999)


crystallization and subsequent sedimentation ofmineral salts could occur. When water containingroad salts percolates through soil, positivelycharged ions (i.e., sodium, calcium, magnesium,potassium) are attracted to and bond with theinherently negatively charged soil surfaces.The extent of bonding depends on the cationexchange capacity and the number of negativelycharged sites in the soil. The sodium ion has ahigh solubility and may, once dissolved, remainin solution; however, since it is readily adsorbedonto soil particles, it is less probable that itwill reach groundwater and surface waters.Nevertheless, in cases with limited adsorptionor where extreme leaching from the soil occurs,the sodium ion will follow water pathways andeventually find its way to groundwater andsurface waters (MDOT, 1993).

There are no major removalmechanisms, such as volatilization, degradation(photodegradation, biodegradation), sorption (toparticulates) or oxidation, that would remove thesalts from surface waters. Dilution resulting frommixing with low-salinity water would decreasethe salt concentrations in the aqueous phase.Hence, the salts will always be present in theaqueous phase rather than in the particulate phase(suspended or bottom sediments). In the caseof benthic sediments, salts may accumulate insediment interstitial water (pore water).

Dissolved salts may alter the physicalproperties of water by changing its density.Density increases linearly with increasing salinity(Ruttner, 1963). The increases in the densityresulting from salinity changes are large incomparison with density changes associatedwith temperature and have important implicationsfor lakes, as inorganic salts can accumulatetemporarily or permanently in deeper strata andprevent lake waters from mixing (Section 3.3.3).

The chemical and biochemical behaviourof major ions from four inorganic salts and theferrocyanide anti-caking agent considered in thisreport are discussed below.

2.5.1.1 Chloride

Chloride is the principal contributing anion tosalinity from application of road salts and isinfluential in general osmotic salinity balance(Wetzel, 1975; Hammer, 1977). Chloride is ahighly soluble and mobile ion that does notvolatilize or easily precipitate or adsorb ontosurfaces of particulates. In freshwater ecosystems,chloride behaves conservatively (i.e., itsconcentration in water is not affected by chemicalor biological reactions) (Wetzel, 1975; Pringleet al., 1981). Chloride levels in surface waterwill largely change only in response to theaddition of chloride, dilution by precipitationor inflow, or concentration by evaporation ofwater. In the majority of inland surface waters,the concentrations of chloride and the respectivecations rarely reach the solubility products ofthe respective chloride salts, and so precipitationrarely occurs.

2.5.1.2 Sodium and potassium

In surface waters, the monovalent cations sodiumand potassium are relatively conservative in theirchemical reactivity and low biotic requirements.Because of this, the spatial and temporaldistributions of sodium and potassium inunimpacted freshwater systems are uniform,with little seasonal variation.

2.5.1.3 Calcium

Calcium is the most reactive ion of all the majorcations contributing to salinity. Its concentrationsin surface waters are affected by chemical andbiological processes within the aquatic systems.It is required as a micronutrient for higher plantsand is one of the basic inorganic elements ofalgae. Calcium concentrations in lakes with hardwater undergo seasonal changes, with a decreasein calcium concentrations and total alkalinityin the summer, when biogenically induceddecalcification removes calcium from the watercolumn. Decreased concentrations of inorganiccarbon in the epilimnion, resulting from increasedrates of photosynthesis, are responsible for


the precipitation of calcium with bicarbonate.Some of the precipitated calcium carbonate isresolubilized in the hypolimnion, and some isentrained in sediments. Decalcification of thetrophogenic zone changes monovalent:divalentcation ratios, which has an effect on thedistribution and dynamics of algae and largeraquatic plants in freshwater ecosystems (Wetzel,1975).

2.5.1.4 Magnesium

In general, magnesium compounds are moresoluble than other compounds. In hardwatersystems, calcium carbonates precipitate beforemore soluble magnesium carbonates. Significantprecipitation of magnesium carbonate andhydroxide occurs only at very high pH (>10).Magnesium is a micronutrient in enzymatictransformations of organisms. It is required bychlorophyll-bearing plants as the magnesiumporphyrin of the chlorophyll molecules (Wetzel,1975). However, the metabolic requirementsfor magnesium are small in comparison with itsavailability in freshwater systems.

2.5.1.5 Sodium ferrocyanide

Sodium ferrocyanide dissociates to yield sodiumand ferrocyanide ions. In the ferrocyanide ion,the chemical bond between the cyanide groupand the iron is very strong; in water, however,it can be photolysed to release cyanide ions(Hsu, 1984). Laboratory tests determined thata 15.5 mg/L solution of ferrocyanide wouldproduce 3.8 mg cyanide/L upon exposure tosunlight for 30 minutes (U.S. EPA, 1971).

2.5.2 Environmental pathways

The transport pathway for road salts applied toroadways is illustrated in Figure 8. The followingsections explain the different environmentalcompartments that can be affected by release ofroad salts via roadway applications, snow disposalsites and salt storage sites (patrol yards).

2.5.2.1 Roadway applications

Road salts applied to roadways can enter surfacewater from:

• salt-laden runoff from roadways into drainageditches, watercourses or wetlands soon afterapplication;

• salt-contaminated meltwater from roadwaysurfaces or from snow plowed to roadsidesafter application of salts; and

• the release of salts stored in surface soils(Scott, 1980).

Factors affecting the degree of roadwayrunoff into aquatic ecosystems include a) thelength of major road treated and drained, b) theamount of salts applied prior to the thaw period,c) road drainage pattern and topography, d) levelof discharge of the receiving stream, e) degreeof urbanization, f) rate of rise and duration oftemperatures above freezing and g) precipitation(Scott, 1981).

Road salts applied to roadways canenter soils and groundwater from meltwaterrunoff from roadways and wet and dry depositionof airborne salt.

Road salts applied to roadways can enterair from windborne vehicle spray or splash andwindborne dry powder from residue on roads.Windborne vehicle spray or splash can be affectednotably by vehicle speed, wind, road gradientand geometrical features of the road (McBean andAl-Nassri, 1987).

2.5.2.2 Snow disposal sites

Road salts released through snow disposal willenter surface water from:

• direct release when dumping into surfacewaters, releasing to untreated storm sewers ormelting following disposal on banks of waterbodies;

• indirect releases via wastewater treatmentplants in the case of dumping to treated storm


FIG

UR

E8

Roa

d sa

lts tr

ansp

ort p

athw

ay

sewers or of surface disposal sites draining toa treated water system;

• overland runoff of stormwater and snowmeltwater; and

• stormwater pipage, which drains part of thedisposal area to a surface water body.

Road salts released through snow disposalwill enter soil and groundwater from directpercolation of precipitation and meltwater fromsnow disposal sites.

2.5.2.3 Storage sites/patrol yards

Road salts released from storage sites or patrolyards will enter surface water from:

• salt-laden leachate from unprotected materialpiles;

• overland runoff of stormwater that hasdissolved salts and of snowmelt water; and

• vehicle washwater.

Road salts released from storage sites orpatrol yards will enter groundwater from:

• groundwater recharge from salt-laden soil andsurface water;

• releases from vehicle washwater where theeffluent is discharged to a dry well; and

• entry through wells or around well casings.

Road salts released from storage sites orpatrol yards will enter soil from:

• spillage of salts at storage sheds duringstockpiling or transferring of salts or blendedabrasives/salts;

• spillage of salts or blended abrasives/saltsduring loading and unloading of maintenancevehicles;

• spillage of salts in the remainder of the yardas maintenance vehicles drive through;

• wind-driven erosion of salts from piles andfrom uncovered winter maintenance vehicles;

• salt-laden leachate from unprotected salt andblended abrasive/salt piles;

• precipitation or snowmelt waters, whichdissolve salts and infiltrate into the soil; and

• release from vehicle washwater dischargingonto the soil.

2.6 Chloride concentrations insurface waters across Canada

This section presents information on theenvironmental concentrations of constituents ofroad salts in Canadian surface waters and presentsestimates of chloride concentrations in surfacewaters across Canada based on mass balancemodelling methods. Information in this sectionis summarized in Mayer et al. (1999) and Evansand Frick (2001) and is based on a review ofnumerous agency reports, scientific publicationsand monitoring surveys.

The surface water ecosystems coveredinclude running (lotic) waters, such as riversand creeks, and standing (lentic) waters, whichinclude lakes, ponds and wetlands. Both thepelagic and the benthic compartments of theseecosystems are discussed. Also presented isa review of background concentrations andconcentrations for water bodies that have beenimpacted by the release of road salts. Examplesof water bodies that have been impacted by thestorage of road salts were included in the sectionon patrol yards (see Section 2.4). The data suggestthat the surface waters most sensitive to road saltimpacts are low-dilution environments such aswetlands, small urban lakes and ponds with longresidence times and small streams draining largeurbanized areas.

2.6.1 Mapping of chloride concentrations inwatersheds across Canada

Concentrations of chloride in water across Canadawere mapped by watershed. Observed chlorideconcentrations in Canadian watersheds areprovided in Figure 9. The concentrations ofchloride in surface water are based on meanvalues obtained from federal and provincial waterquality monitoring stations in Canada. Themethods and database used in generating the mapare provided in Mayer et al. (1999). Basically,


monitoring data from near the mouth of eachwatershed are combined to provide estimates ofthe concentration for the entire watershed. Theamount of data are varied, ranging from manystations and a monthly sampling interval tointermittent spot samples. Accordingly, data inFigure 9 should be used as an indication of therelative variations in the concentrations ofchloride across Canada.

Watersheds containing major urban areasand areas of intense road salt loadings in southernQuebec, Ontario and the Maritime provinceshave elevated chloride concentrations. Note thatthese concentrations represent averages over awatershed and, as such, do not indicate actualexposure concentrations of a specific watercourse.The map also indicates that some watershedsin the Prairie provinces have elevated ambientconcentrations. These concentrations arerelated to the arid climate of the Prairies and areassociated with the saline lakes and rivers situatedin this region.

Road salts are not the only input thatinfluences the concentrations in Figure 9.Effluents from various industries (e.g., pulpand paper mills) and releases from private septicsystems and municipal wastewater treatmentplants can influence chloride concentrations.It has been estimated for the lower Great Lakesthat road salts contribute approximately 20%of the chloride load, with natural and otheranthropogenic sources contributing the balance(Jones et al., 1986).

2.6.2 Environmental concentrations:Lakes and rivers

In the absence of anthropogenic sources, surfacewaters acquire their characteristics by dissolutionand chemical reactions with solids, liquids andgases with which they come into contact duringvarious phases of the hydrological cycle (Stummand Morgan, 1981). For instance, stream waterchemistry can be related to the chemistry of thesource rock and to equilibria controlling the

formation of solid phases (Feth et al., 1964;Garrels and Mackenzie, 1967). Watershedbedrock and surficial geology play a dominantrole in determining the salt concentrations insurface waters. In addition, the climate and theproximity of the surface waters to the sea will beimportant in regulating the background saltconcentrations. Thus, the background saltconcentrations (salinity) of surface waters vary.

Surface waters range from fresh waters,with salinities lower than 500 mg/L, to salinewaters, with salinities equal to or greater than3000 mg/L (Hammer, 1986). While enclosedlakes and ponds may become saline, streams andrivers will rarely have natural salinity highenough to be classified as saline.

For the purpose of this AssessmentReport, surface water quality is discussed bygeographic region rather than by hydrogeologicalregion in order to better address the effects ofhuman factors on surface water quality. Bothnaturally occurring concentrations and thoseimpacted by human activity are presented. Whileanthropogenic sources such as road salts, sewageand industrial effluents can all increase salinity,the main emphasis for data on impacted areasrelates to the releases of road salts.

2.6.2.1 Atlantic region (Newfoundland, NovaScotia, New Brunswick and PrinceEdward Island)

The median concentrations of chloride andcalcium from recent surveys (1985 and later)were compiled by Jeffries (1997). The data setincludes 38 lakes from Labrador, 63 lakes fromNewfoundland, 150 lakes from Nova Scotia and166 lakes from New Brunswick. The medianchloride and calcium concentrations in lakes inlargely unimpacted areas ranged between 0.3and 4.5 mg/L and between 0.82 and 1.10 mg/L,respectively.

Comprehensive studies by Kerekeset al. (1989), Kerekes and Freedman (1989) and


Freedman et al. (1989) at Kejimkujik NationalPark, Nova Scotia, investigated the water qualityof several lakes and rivers in the park. Thesestudies concluded that sea salt influence issignificant in Nova Scotia. The concentrationranges were 3.6–5.4 mg/L for chloride,2.3–3.4 mg/L for sodium and 0.36–0.96 mg/L forcalcium. The levels of these ions werecomparable with those measured in other surfacewaters in the Atlantic region (Jeffries, 1997).

Studies on the Chain Lakes nearHalifax indicate that winter de-icing activitiescan have an impact on water quality. Thirmurthiand Tan (1978) explained that, from June toSeptember 1975, the chloride concentration wasapproximately 19 mg/L. Concentrations began toincrease in November and eventually peaked inMay 1976, at 43 mg chloride/L. The chlorideconcentration then dropped rapidly through the

rest of May and June, decreasing to 19 mg/L inJuly. The increase in chloride concentrations forboth lakes represents an approximate addition of19 440 kg of chloride to First Chain Lake and7920 kg to Second Chain Lake. The Chain Lakescontinued to be studied through the early 1980s(Hart, 1985, 1988). Chloride concentrations inboth lakes increased gradually from 1976 toreach a maximum in the mid-1980s ofapproximately 120 mg/L in First Chain Lake andabout 170 mg/L in Second Chain Lake. Highestchloride concentrations were observed duringspring runoff.

Road salts have impacted ChocolateLake in Nova Scotia. The lake is small (82.7 ha)and shallow (mean depth 3.9 m, maximum depth12.2 m), with an estimated volume of 350 000 m3

(Kelly et al., 1976). The watershed is large(900 ha), approximately 11 times as great as


FIGURE 9 Observed chloride concentrations in Canadian watersheds (from Mayer et al., 1999)


the lake surface area. Chloride concentrationswere highly elevated in the lake between Apriland August 1975. Average summer chlorideconcentrations were 207.5 mg/L compared withan estimated background value of 15–20 mg/Lfor non-impacted lakes. Chloride concentrationsvaried with depth, ranging from 199 to 224 mg/Lat the surface, from 189 to 217 mg/L at 6.1 m,and from 225 to 330 mg/L at 12.1 m. Sodiumconcentrations ranged from 108 to 125 mg/L,from 102 to 125 mg/L and from 116 to 183 mg/Lat the same depths. The gradient in saltconcentrations and the absolute concentration ofsalt in deep waters were large enough to preventcomplete vernal vertical mixing of the watercolumn; deep waters became anoxic in summer.The primary source of chloride was attributed tohighway runoff containing road salts. Kelly et al.(1976) estimated that 58.4 tonnes of road salt(representing 35 409 kg of chloride and 22 737 kgof sodium) were applied to the Chocolate Lakedrainage basin during the winter of 1974–75.

A study on the water quality of 50 lakesin the Halifax/Dartmouth metropolitan area notedthat most lakes experienced a very pronouncedincrease in the concentrations of sodium andchloride between 1980 and 1991 (Keizer et al.,1993). Of the lakes sampled in 1980, 12 hadchloride concentrations greater than 50 mg/Land 3 had concentrations greater than 100 mg/L(the maximum concentration was 125 mg/L). In1991, 22 of the 50 lakes sampled had chlorideconcentrations greater than 50 mg/L, 12 hadconcentrations greater than 100 mg/L and 4had concentrations greater than 150 mg/L (themaximum concentration was 197 mg/L). Theconcentrations of sodium in these lakes showedsimilar trends. Of the lakes sampled in 1980,9 lakes had sodium concentrations greater than40 mg/L and 3 had concentrations greater than60 mg/L (the maximum concentration of sodiumwas 73 mg/L). In 1991, 21 lakes had sodiumconcentrations greater than 40 mg/L, 12 hadconcentrations greater than 60 mg/L and 7had concentrations greater than 80 mg/L (themaximum concentration was 114 mg/L). It isnoteworthy that sampling was at the surface in

the middle of the lakes and may not becharacteristic of bays, where higherconcentrations may occur. Furthermore, sincedenser, salt-laden water generally sinks to thebottom, these samples may be a low estimateof the overall lake concentrations.

Keizer et al. (1993) found that lakes inmore developed watersheds had higher chlorideand sodium concentrations than lakes in ruralareas. In addition, salt concentrations from lakesin less developed areas did not increase markedlybetween the two sampling periods, unlike lakesin urban areas. For example, 19 of the rural lakessampled had chloride concentrations below25 mg/L in both 1980 and 1991; 23 of the lakessampled had sodium concentrations below20 mg/L in 1980 and 1991. Three of the lakeswith the lowest concentrations had chlorideconcentrations of 5.1, 3.9 and 5 mg/L in 1980and 4.5, 3.4 and 5 mg/L in 1991. Sodiumconcentrations were 3.4, 2.5 and 3.2 mg/L in1980 and 2.9, 2.2 and 2.9 in 1991.

In general, the chloride concentrationsin unimpacted water bodies in the Atlanticregion do not exceed a concentration of 20 mg/L.Lower concentrations (<10 mg/L) are generallyobserved in inland water bodies, while the higherconcentrations occur in coastal areas.

2.6.2.2 Central region (Ontario and Quebec)

The concentrations of chloride, sodium andcalcium in lakes and rivers in the CanadianShield, which covers a large area of Quebec andOntario, including the upper Great Lakes, are verylow (e.g., 1–10 mg chloride/L) (Jeffries, 1997).According to Weiler and Chawla (1969), theaverage annual concentrations of chloride in lakesErie and Ontario were 24.6 and 27.5 mg/L,respectively, while the corresponding chlorideconcentration in Lake Superior was only1.3 mg/L. Earlier studies, which investigatedwater quality trends in the Great Lakes, reported arapid increase in chloride concentrations related tohuman activities (Kramer, 1964; Beeton, 1965;Dobson, 1967). The largest increase was observed


for Lake Ontario. Later data (Williams et al.,1998), however, showed a decline in chlorideconcentrations from about 22 mg/L in 1977 toabout 16 mg/L in recent years (1985–1993). Thedecline in the lower Great Lakes can probably beascribed to lower loadings from industrial anddomestic sources, resulting from improvedcontrol/treatment of industrial and domesticeffluents. The tributaries to Lake St. Clair(e.g., Sydenham and Thames rivers) havesomewhat higher chloride concentrations (about30–40 mg/L) than other southern Ontario riversbecause they drain the Paleozoic sedimentary rockadjacent to Michigan Basin, which contains salineformations such as salt (sodium chloride) andanhydrite (calcium sulphate) bearing strata. Thesestrata also contribute to salt (sodium chloride andcalcium sulphate) inputs to Lake St. Clair andwestern Lake Erie.

In urbanized areas, application of de-icing salts contributes significantly to chlorideinputs. Long-term monitoring locations in theGreater Toronto Area show a gradual increasein chloride concentrations. In Highland Creek,a highly degraded urban watercourse, thestatistical trend analysis shows an increase inchloride concentrations from 150 mg/L in 1972(presumably already impacted) to over 250 mg/Lby 1995 (Bowen and Hinton, 1998). In DuffinCreek, the median chloride concentrationsincreased from 10–20 mg/L in the 1960s to30–40 mg/L in the early 1990s (Bowen andHinton, 1998). In urban streams and rivers, suchas the Don River in Toronto, concentrations ofchloride increased to more than 1000 mg/L inwinter (Scott, 1980; Schroeder and Solomon,1998), with increases appearing to coincidewith thaws. The autumn baseline values inthe Don River were 100–150 mg chloride/L.Similar baseline concentrations of chloride(50–100 mg/L) were reported in another Torontowatercourse, Black Creek (Scott, 1980). Chlorideconcentrations 50 times higher than baselinevalues were measured during the thaw periods inthis creek. Crowther and Hynes (1977) reportedhigh winter concentrations of chloride(1770 mg/L), sodium (9550 mg/L) and calcium

(4890 mg/L) in Laurel Creek, which passesthrough urban Waterloo. Rodgers (1999) observedhigh concentrations of chloride (4355 mg/L) inRed Hill Creek in Hamilton, Ontario.

The Ontario Ministry of Transportationhas monitored water quality in a number ofstreams in Ontario, including Toronto-areawatersheds, between 1990 and 1996. Results forfour watercourses in urban areas are shown inTable 9. Highest concentrations were recordedduring the winter months, indicating peak inputsof road salts. Chloride concentrations frequentlyexceeded 250 mg/L and often exceeded 500 mg/L(Table 9).

Little Round Lake in central Ontariowas affected by cultural disturbances thatincluded urbanization, septic tanks, highway roadsalt runoff and seepage from a salt storage depot(Smol et al., 1983). This small lake (7.4 ha,maximum depth 16.8 m) became meromictic.Meromictic conditions were attributed by Smolet al. (1983) to input of road salts from runoff andstorage. Salt concentrations in the monimolimnionor deep layer were 58.4 mg sodium/L and103.7 mg chloride/L, well in excess of thatexplainable by the natural geology of the region.Road salt additions apparently had resulted in theformation of meromixis in the 30 years prior tothe study in 1981.

The Ministère des Transports duQuébec (1980,1999) investigated the impacts ofroad salt on Lac à la Truite, near Sainte-Agathe-des-Monts. The lake drainage area, estimatedat 728 ha, was affected by a 7-km stretch ofhighway (approximately 30 lane-kilometres).At certain places, the highway was located asclose as 250 m from the lake. The lake has asurface area of 48.6 ha, a mean depth of 21.5 mand an estimated volume of 486 000 m3. In 1972,the average chloride concentration for the lakewas 12 mg/L. This increased through the 1970sto reach a maximum concentration of 150 mg/Lin 1979. This corresponds to an addition of67 068 kg of chloride to the lake. To preventfurther increases, abrasives (containing some salt


to prevent caking) were used instead of road salts.Chloride concentrations fell through the 1980s toreach 45 mg/L in 1990 and have remained at thatlevel since.

In the United States, significant increasesin chloride concentrations observed in fourAdirondack streams were attributed to winter saltapplication (Demers and Sage, 1990). There wasa significant difference in chloride concentrationsbetween samples upstream and downstream fromroadways. Demers (1992) reported that chlorideconcentrations were as much as 66 times higher indownstream samples. The overall mean chloridelevel was 0.61 mg/L upstream and 5.23 mg/Ldownstream. The terrain characteristics ofAdirondack streams (New Hampshire) are similarto those of streams in the Eastern Townships ofQuebec. Chloride concentrations of 11 000 mg/L(Hawkins and Judd, 1972) were measured inwinter of 1969 in Meadowbrook, which drainsabout 10 km2 of suburbs in Syracuse, New York

(Hawkins and Judd, 1972). Also in the UnitedStates, chloride concentrations of about 400 mg/Lwere measured in the bottom waters ofIrondequoit Bay, a small bay on Lake Ontarionear Rochester, New York (Bubeck et al.,1971, 1995).

In summary, the chloride concentrationsof unimpacted water bodies in the Shield regionof Central Canada are among the lowest in thecountry and vary within the range <1–5 mg/L.Higher background concentrations of up to10–30 mg/L are observed in water bodies situatedwithin the St. Lawrence Lowlands region, whichincludes the basin of the lower Great Lakes andthe St. Lawrence River. A marked increase inchloride concentrations above background levelsis particularly noticeable in many small urbanlakes and watercourses in this region, as it isthe most densely populated area of Canada.Seasonal and spatial trends indicate the importantcontribution of road salts to these increases.

TABLE 9 Chloride concentrations in various streams in the Toronto Remedial Action Plan watershedfor 1990–96 (from Toronto and Region Conservation Authority, 1998)

Station location No. of Minimum Maximum Mean Median % frequency observations (mg/L) (mg/L) (mg/L) (mg/L) exceedence

250 mg/L 500 mg/LEtobicoke Creek• at Derry Road 38 10 3780 278 135 18 8• at Burnhamthorpe 40 0 2670 392 206 40 20

Rd.• at Highway #2 74 43 2140 351 208 38 19

Mimico Creek• at Highway #2 37 51 3470 553 276 62 22

Black Creek• at Scarlett Rd. 38 20 4310 495 248 47 21

Highland Creek• at Highway 55 22 1390 310 220 42 13

Creek Park

2.6.2.3 Prairie region (Manitoba, Saskatchewanand Alberta)

Most naturally saline lakes in Canada arelocated within two regions. The first regionencompasses several endorheic drainage basinson the Canadian Prairies (Hammer, 1984).The area where saline lakes are most numerousstretches over southern Alberta and Saskatchewan.The Alberta–Saskatchewan saline lake regionis largely underlain by Cretaceous bedrock,composed mainly of shales, silts and sandstones(Hammer, 1984). Of the Prairie provinces,Saskatchewan has by far the greatest number andvolume of saline lakes (Hammer, 1986). Therelative proportions of cations in most Prairielakes are as follows: Na > Mg > Ca. Of the lakesinvestigated, there are only two knownmeromictic lakes on the Canadian Prairies,Waldsea Lake and Deadmoose Lake (Hammer,1984). Numerous potholes (shallow lakes orponds), which range in salinity from fresh tosaline, can be found in the southern Prairies.

There are fewer naturally saline lakes(about 30) in Alberta (Hammer, 1984). Most ofthem are concentrated in the Provost and Hannaregions, although they occur as far north asEdmonton. Even fewer saline lakes are inManitoba. Barica (1978) investigated a group ofover 100 small potholes, which varied in salinityfrom fresh to moderately saline, near Erickson insouthwestern Manitoba. The concentration rangeswere 1–448 mg/L for chlorine, 0.8–1075 mg/L forsodium and 27–380 mg/L for calcium. Twenty-three sodium chloride–dominated sites along thewestern shore of Lake Winnipegosis were studiedby McKillop et al. (1992). The concentrationranges for chloride, sodium and calcium were861–33 750 mg/L, 587–21 313 mg/L and59–1400 mg/L, respectively.

Much lower concentrations of chlorideand calcium are reported for 27 and 193 lakesin northern portions of Saskatchewan andAlberta, respectively (Jeffries, 1997). The medianconcentrations of chloride were 0.2 and 0.5 mg/L,respectively. The calcium concentrations of

Saskatchewan and Alberta lakes were 2.0 and14.0 mg/L, respectively. The chloride and calciumconcentrations of 26 lakes in Manitoba were 2.0and 7.5 mg/L, respectively (Jeffries, 1997).

In summary, many of the saline waterbodies in Canada are located within the InteriorPlains, which cover the southern portions of thePrairie provinces. In general, dissolved saltconcentrations, including chloride, are higher inthis geographical region because of the underlyinggeology. However, low concentrations of chloride(<5 mg/L) are reported for unimpacted waterbodies situated in the northern region of theseprovinces, outside of the Interior Plains region.

2.6.2.4 Pacific region (British Columbia)

The second region in Canada where a largenumber of naturally saline lakes are located isthe Southern Interior Plateau (Fraser Plateau) ofBritish Columbia, in the rain shadow of the CoastMountains. Saline lakes in British Columbia aresmall and shallow (Topping and Scudder, 1977).The limnology of four of these lakes wasdescribed by Northcote and Halsey (1969), whoshowed not only that these lakes differed in totaldissolved solids concentrations among themselves,but also that there were differences within thelakes (surface to near-bottom). Concentrationsranging from 5.1 to 800 mg/L were reported. Themeromixis in three of these lakes was maintainedby chemical density gradients, while in the fourth,the mixing was inhibited by morphometricfeatures. In non-saline regions, the reportedmedian chloride and calcium concentrations of sixacid-sensitive lakes located in the southwest of theprovince were 2.5 and 3.7 mg/L, respectively(Phippen et al., 1996; Jeffries, 1997).

The impact of road salts on water qualitywas monitored for the Serpentine River in BritishColumbia’s Lower Fraser Valley. Electronic datarecorders monitored water temperature, pH,water level and conductivity every 15 minutes.Three-fold increases in conductivity wereobserved over 10- to 20-hour periods during thawperiods following a cold period when roads were



sanded and salted (Whitfield and Wade, 1992).Furthermore, periods of elevated conductivitywere followed by increased water levels, whichare characteristic of snowmelt. Results of thisstudy indicate that organisms living in certainstreams during winter months may be subject tofluctuations in salt concentrations.

In summary, there are several lakeslocated in the Southern Interior Plateau of BritishColumbia whose chloride concentrations arehigher than 100 mg/L, and several of these lakesare naturally saline. The chloride concentrationsof unimpacted water bodies in the remainingparts of British Columbia are generally less than5 mg/L.

2.6.2.5 Yukon, Northwest Territoriesand Nunavut

Few data were identified from the Yukon,Northwest Territories and Nunavut.

2.6.3 Environmental concentrations:Benthic sediments

Because inorganic chloride salts used for roadmaintenance are highly soluble and interactionswith sediment particulates are minimal, inorganicsalts are expected to accumulate in sediment porewater rather than in the solid phase. The study byMayer et al. (1999) of an urban pond (RougeRiver Pond) shows that the sediment pore water,which is in equilibrium with the overlying water,enriched in inorganic salts, may attain high saltconcentrations (see Section 2.6.5). Highconcentrations of salts in sediment pore watercan result not only in osmotic stress and directtoxic effects on benthic biota but also incomplexation of metals with chloride, augmentingthe concentrations of dissolved heavy metals (e.g.,cadmium), which are toxic to benthic organisms.

2.6.4 Environmental concentrations:Wetlands

Wetlands are an important landscape feature inCanada. Their most distinguishing features are

presence of standing water, unique wetland soilsand vegetation adapted to or tolerant of saturatedsoils (Mitsch and Gosselink, 1986). Canada has agreat variety of wetlands; detailed descriptions ofvarious types of wetlands can be found in Mitschand Gosselink (1986).

There is only limited information onthe chemical characteristics of coastal and inlandfreshwater wetlands with respect to chloride, formuch of the work on wetlands is concerned withnutrient cycling, acidification and geochemistryof metals. A 1987 study of wetlands withinKejimkujik National Park (Wood and Rubec,1989) assessed the chemical characteristicsof wetlands. The authors reported significantdifferences in peat chemistry between the bogsand fens, in particular in sodium and calciumconcentrations. The concentrations of calcium,sodium and chloride in peat from bogs were0.17, 0.8 and 0.15 mg/L, respectively. Theconcentrations of calcium, sodium and chloridein peat from fens were 0.32, 1.5 and 0.12 mg/L.There was no difference in the surfacewater chemistry between bog and fen. Theconcentrations of calcium, sodium and chloridein water of bogs were 6 × 10–4, 3.0 × 10–3

and 3.5 × 10–3 mg/L, respectively, and theconcentrations of calcium, sodium and chloridein water of fens were 6 × 10–4, 2.8 × 10–3 and3.9 × 10–3 mg/L, respectively.

Bourbonniere (1985, 1998) investigatedthe geochemistry of wetlands in Nova Scotia andOntario. He measured the major ion chemistry inpore water of the ombrotrophic bog in BarringtonCounty in Nova Scotia. The measuredconcentrations were 9.6–24 mg/L for chlorine,5.2–12.8 mg/L for sodium and 0.47–1.5 mg/Lfor calcium. The concentrations of calcium,sodium and chlorine in Beverly Swamp,Ontario, were 58.1–95.3 mg/L, 5.1–20.0 mg/Land 8.7–41.2 mg/L, respectively. The watercomposition of potholes described in Section2.6.3.3 may be considered to be characteristicof the prairie wetlands. No Canadian studieswere found that address the impact of road saltapplication on the chemistry of wetlands.


Two studies were presented in the patrol yardsection (Section 2.4) that focus on the impact ofsalt depots on wetlands.

2.6.5 Environmental concentrations: Urbanlakes and ponds

Natural small urban lakes and ponds that receiveroad runoff are susceptible to changes in ionconcentrations. There are only limited data onthe chloride concentrations in such bodies. In theabsence of such data, ponds that are part of anurban stormwater management system can beused as a reasonable surrogate. Numerous pondshave been constructed and remain a widely usedform of watershed management in urban areas.Although these ponds are engineered waterbodies designed primarily for hydraulic flowmanagement and stormwater treatment, they havedeveloped aquatic ecology similar to natural smallurban ponds. Like natural urban aquatic systems,these ponds are wetland habitats for many speciesof flora and fauna. They are likely candidates forshowing maximum chloride concentrations, sincethey represent the type of water body that receivesthe maximum exposure to chloride loadings fromwinter maintenance activities.

A study of Lake Wabukayne, a smallhuman-made lake in Mississauga, Ontario,reported chloride concentrations between 200and 2000 mg/L and temporary meromixisduring the winter and early spring (Free andMulamoottil, 1983). Chloride concentrations of1100–2000 mg/L (Vickers, 2000) were measuredin the Harding Pond in Richmond Hill in theGreater Toronto Area. Seasonal surveys of thestormwater retention ponds in the Greater TorontoArea (Mayer et al., 1996; Mayer, 2000) showedhigher concentrations of chloride during thewinter months. Chloride concentrations of 380and 800 mg/L, respectively, were measured in theHeritage and Unionville ponds, which are situatedin residential settings (Mayer, 2000), whilechloride concentrations as high as 5910 mg/Lwere measured in the Col. S. Smith Reservoir,receiving runoff from a multi-lane highway(Queen Elizabeth Way) in February of 1999.

Both a mass balance study and a sedimentstudy were performed on the Rouge River Pond,situated in the Rouge River Valley, Scarborough,Ontario, near Highway 401 and Port Union Road.This pond is an urban stormwater managementfacility designed to treat highway stormwater byremoving 70% of suspended solids. The massbalance study was conducted to measure thepollutant removal and the dynamics of severalwater quality constituents between 1995 and 1998(Liang, 1998). The initial studies during 1995,1996 and 1997 focused on balances during theice-free season. A chloride imbalance related topermanent meromictic conditions in the pondwas shown, with meromixis likely due to acombination of the pond design and high road saltloadings. The salt concentrations calculated assodium chloride in the bottom layer of thepond were of the order of 5000 mg/L. Elevatedconcentrations of metals and ammonia were alsomeasured in this bottom layer. The sediment study(Mayer et al., 1999) investigated the chemistryand toxicity of sediment pore water and assessedthe solid-phase metal chemistry. A geochemicalmodel was formulated to assess the impact ofsalts on metal speciation. The investigationrevealed that sediment pore water, which is inequilibrium with the overlying water, enrichedin inorganic salts, may itself attain high saltconcentrations. At the sediment–water interface,concentrations of chloride and sodium greaterthan 3000 mg/L and 2000 mg/L, respectively,were measured in the pore water. A gradualdecrease with depth in sediment was observedfor both ions; however, even at a depth of 40 cmbelow the sediment–water interface, the chlorideconcentrations were about 1500 mg/L.

A study by Bishop et al. (2000) focusedon the presence of contaminants in six stormwaterretention ponds in the Greater Toronto Areaand nine stormwater retention ponds in Guelph,Ontario. Of the 15 ponds in this study, 4 ponds inGuelph were sampled between 19 and 21 timeseach to determine the concentration of chloride instormwater. These samples were taken at differenttimes during the months of September, Octoberand November 1997 and April, May, June, July,


August, September and November 1998. Sampleswere also taken as close to the ponds’ outfalls aspossible (Struger, 2000). Mean concentrations ofchloride in stormwater samples from these fourponds were between 120 and 282 mg/L; themaximum concentrations observed at the fourponds were between 416 and 1230 mg/L (Struger,2000). While samples were not taken fromDecember to March, the average chlorideconcentrations of samples for November, Apriland May were higher than the average chlorideconcentrations of samples for the other months ateach of the ponds (336 vs. 153 mg/L; 183 vs.79 mg/L; 469 vs. 82 mg/L; 444 vs. 188 mg/L).

The samples were taken close to theponds’ outfalls because these concentrations areprobably indicative of the concentrations enteringreceiving environments. The stormwater retentionponds treat suspended solids and other attachedconstituents, but do not remove dissolved saltsfrom runoff (OME, 1991; Bishop et al., 2000).Depending on the volume of runoff entering theseponds, they are typically designed to retain runofffor a period ranging between 24 and 72 hours;retention time may be as low as 10 hours duringpeak periods of large events.

Watson (2000) sampled the quality ofwater in 89 ponds located near roadways insouthern Ontario. Multiple samples were takenduring spring and summer of 2000, and the pondswere categorized according to the number of lanesof the nearest road. The mean concentration ofchloride in ponds located near two-lane roadswas 95 mg/L (range 0–368 mg/L). The meanconcentration of chloride in ponds located nearroads with more than two lanes but fewer thansix lanes was 124 mg/L (range 0–620 mg/L).The mean concentration of chloride in pondslocated near roads with six lanes or more was952 mg/L (range 49–3950 mg/L). Watson (2000)does not provide an indication of backgroundconcentrations, but indicates that the main sourceof chlorides would be from road salts.

2.6.6 Chloride concentrations based onmass balance calculations

2.6.6.1 Calculation of watershed chlorideconcentrations resulting from the useof road salts

A mass balance model was used to estimatechloride concentrations in Canadian watershedsfrom the use of road salts. The calculated chlorideconcentrations were compared with the observedconcentrations presented in Section 2.6.1. Moredetails of this comparison and the analysis arepresented in Mayer et al. (1999).

While there are some simplificationsinherent to the model used (e.g., assumptionof similar road salt application and volume ofrunoff from year to year), the model providesa reasonable estimate of the potential chlorideconcentrations resulting from road salt use. Themodel is also useful for identifying the relativeconcentrations of chlorides in the differentwatersheds. Sub-basin watersheds are used inthe calculations; a more accurate comparison ofcalculated chloride concentrations would havebeen obtained if finer-scale resolution watershedmaps had been used.

The results are presented in Figure 10.The Island of Montréal has the highest modelledconcentration of chloride in runoff. Figure 10 alsoindicates that watersheds on the north shore ofLake Ontario, notably those largely influencedby municipalities in the Toronto area, also havehigh chloride concentrations related to road salts.The watersheds in urban areas, particularlythose in southern Ontario, southern Quebec, theMaritime provinces and larger urban centres inthe Prairie provinces, have the highest modelledconcentrations of chloride in runoff.

It is noteworthy that the concentrationspresented in Figure 10 are modelled averages forwatersheds. Accordingly, the concentrations donot indicate actual exposure concentrations forspecific watercourses or water bodies. Data inFigure 10 can be used to identify areas that face

the greatest risk of adverse effects from the use ofroad salts.

2.6.6.2 Calculation of chloride concentrationsin roadway runoff

The previous section assessed chlorideconcentrations caused by road salts for entirewatersheds. At a watershed scale, modelledconcentrations are diluted by runoff from areasother than just the roadway. This section estimatesthe concentration of chlorides at a roadway scale,using a mass balance approach.

For these estimates, it was assumed thatthe mass of salt used per two-lane-kilometre ofprovincially maintained road was dissolved anddiluted by the volume of average annual runoff

that would accumulate on the roadway surface(i.e., in 7400 m2, based on an assumed width ofa two-lane kilometre of 7.4 m).

Figure 11 presents estimatedconcentrations of chloride and the lengthof provincial roads associated with theseconcentrations. Concentrations of chloride rangebetween 30 and 31 000 mg/L. Based on theseestimates, the majority of the concentrations(85%) range between 1000 and 10 000 mg/L.These estimates are of the concentrations ofchloride in runoff from roadways; variousapplication factors can be used to account fordilution in receiving waters. The accuracy ofresults depends on the quality of data used tocalculate concentrations. One limitation is thatestimates of average annual runoff were derived


FIGURE 10 Estimated road salt chloride concentrations by watershed, calculated from average annualroad salt loadings and average annual runoff (from Mayer et al., 1999)


using a map at a scale of 1:7 500 000. Moredetailed maps (i.e., 1:50 000) might result inslightly different results. Furthermore, estimatesof average annual runoff were used. Differentvalues for runoff may have been calculated ifrunoff associated with winter months or springthaw had been used.

The concentrations of chloride observedin a recent study (Mayer et al., 1998) of highwayrunoff from three sites in southern Ontario areconsistent with the above estimated chlorideconcentrations. The study reported chlorideconcentrations in runoff from roadways at threesites: a) the Burlington Skyway Bridge (a four-lane road with 92 000 cars per day), b) Highway2 east of Brantford (a two-lane road with 31 100cars per day) and c) Plains Road in Burlington(a two-lane roadway with 15 460 cars per day).Highest chloride concentrations (up to19 135 mg/L) were observed in winter runofffrom the Skyway Bridge, although highconcentrations were also observed at Plains Road.The study, which investigated runoff toxicity,showed that salt-laden runoff has the potential toadversely affect aquatic organisms in low-dilutionaquatic systems. The high chloride concentrationsreported in this study are representative of

concentrations at point of discharge and representthe worst-case scenario, which may be foundalong roadside ditches or small wetlands adjacentto major roadways.

2.6.7 Summary

Data show a broad range of background chlorideconcentrations in Canadian surface waters.Differences stem from the variance in bedrockand surficial geology, climate and proximity tothe sea. Impacts of high use of road salts areevident from the data, particularly in highlyurbanized areas. Water bodies most sensitive tothe releases of road salts are low-dilutionenvironments, such as small urban lakes andponds with long residence times, urbanstormwater ponds with short residence times,streams draining large urbanized areas andwetlands adjacent to major roadways.

In terms of natural concentrations acrossCanada, the following regional trends haveevolved:

• In general, the chloride concentrations inunimpacted water bodies in the Atlanticprovinces do not exceed about 20 mg/L.

FIGURE 11 Chloride concentrations in runoff for provincial roads in Canada

0

2 000

4 000

6 000

8 000

10 000

12 000

14 0000

-0.5

>0

.5-1

.0

>1

.0-1

.5

>1

.5-2

.0

>2

.0-2

.5

>2

.5-3

.0

>3

.0-3

.5

>3

.5-4

.0

>4

.0-4

.5

>4

.5-5

.0

>5

.0-5

.5

>5

.5-6

.0

>6

.0-6

.5

>6

.5-7

.0

>7

.0-8

.0

>8

.0-9

.0

>9

.0-1

0.0

>1

0.0

-15.0

>1

5.0

-25.0

>2

5.0

Chloride concentration (g/L)

Le

ng

th o

f p

rim

ary

ro

ad

(km

)

0

2

4

6

8

10

12

14

16

18

% o

f to

tal

km % of total

Lower concentrations (<10 mg/L) aregenerally observed in inland water bodies.

• Chloride concentrations of unimpacted waterbodies in the Canadian Shield region ofCentral Canada are among the lowest in thecountry and vary within the range <1–5 mg/L.A marked increase in chloride concentrationsis noticeable in many small urban lakesand streams.

• Many of the naturally saline water bodies inCanada are in the southern portions of thePrairie provinces. However, naturally lowconcentrations of chloride (<5 mg/L) arereported for unimpacted water bodiesin the northern region of these provinces.

• There are several naturally saline lakeslocated in the Southern Interior Plateau ofBritish Columbia (chloride concentrationshigher than 100 mg/L). The chlorideconcentrations of unimpacted water bodiesare generally less than 5 mg/L.

A spatial analysis of chlorideconcentrations was done based on twoapproaches. The first aggregated monitoringdata to a watershed scale; the second used a mass

balance analysis to model chloride concentrationsat a watershed level. Mass balance techniqueswere also used to compute average annualconcentrations in runoff from specific highwaysacross Canada. The calculated concentrationsrange up to 31 000 mg/L, but mainly lie between1000 and 10 000 mg/L. These estimates areconsistent with measured undiluted runofffrom roadways.

Elevated chloride concentrations werealso measured in ponds and wetlands adjacentto roadways. Concentrations up to 4300 mg/Lwere observed in watercourses, and concentrationsbetween 150 and 300 mg/L were observed in rurallakes that had been impacted by the use of roadsalts. While highest concentrations are usuallyassociated with winter or spring thaws, highconcentrations can also be measured in thesummer, as a result of the travel time of the ionsto surface waters and the reduced water flows inthe summer.


3.1 CEPA 1999 64(a): Environmentand CEPA 1999 64(b):Environment on which lifedepends

The approach for the environmental assessmentof PSL substances involves characterizing theentry of, exposure to and effects of a substance,which ultimately permits the characterization ofthe risk posed by the substance (EnvironmentCanada, 1997a). This may involve the use of bothlaboratory and field-derived data. Empirical fieldevidence linking environmental changes to entryof the substance can provide strong indications ofactual risk.

It can be assumed that all road saltseventually enter the environment, whether fromlosses from storage or patrol yards, from roadwayapplication or from disposal of snow. Althoughall chloride ions will ultimately be found insurface waters, ions from inorganic chloride roadsalts and ferrocyanides will be released to ormove through soil, groundwater, surface waterand air, and organisms can be exposed to roadsalts through all these media. Therefore, allenvironmental compartments are of concernfor the assessment of road salts. Environmentaleffects can be either biotic or abiotic. Thus,endpoints of concern can include effectson individual organisms, populations orcommunities, as well as abiotic endpoints, such asdeterioration of soil structure or lake stratification,which disrupts seasonal mixing of waters.

While there has not been systematicassessment of exposure to road salts for anygiven compartment across Canada that wouldpermit detailed quantitative risk characterizationat the national scale, many individual studiesare available that permit the establishment of

relationships between the use of road salts andtheir impacts on biotic and abiotic systems.

The following sections characterizethe risks associated with releases of road saltsinto the environment, by reviewing the entry,exposure and effects related to five environmentalcompartments or biotic groups with regardto inorganic chloride salts. Data pertinent toferrocyanides and their environmental risks arereviewed in a separate section (Section 3.7).

3.2 Groundwater

3.2.1 Introduction

This section summarizes the fate and impactof road salts on groundwater quality. Discussionin this section is based on a report prepared byJohnston et al. (2000). A summary of case studies(Morin, 2000) is also presented.

Data on concentrations of salts ingroundwater are frequently discussed inpublished and unpublished literature in termsof exceedences of drinking water qualityguidelines. These guidelines are set on the basisof taste thresholds at 200 mg sodium/L and250 mg chloride/L (CCME, 1998). Although theuse of groundwater as a drinking water resourceis not considered in this assessment, a review ofreported exceedences of a chloride concentrationof 250 mg/L is pertinent to the assessment ofpotential impacts on aquatic biota. As reviewedin Section 3.3, the No-Observed-EffectConcentration (NOEC) for the 33-day survivalof early life stage fathead minnow was 252 mgchloride/L. Furthermore, lethality data modelledfor chronic exposure indicated that 10% ofaquatic species would be affected at chlorideconcentrations of about 240 mg/L.


3.0 SUMMARY OF CRITICAL INFORMATION AND

ASSESSMENT OF “TOXIC” UNDER CEPA 1999

3.2.2 Groundwater principles

3.2.2.1 Hydrological cycle

The degree to which road salts may have animpact on the groundwater environment will varysignificantly, depending on the road salt loading,climatic conditions, surface and subsurface soilconditions and location of the site within theoverall hydrogeological environment.

Precipitation that falls across an areaas rain or snow contributes to a number ofhydrological processes. A portion of theprecipitation will run off as overland flow fromimpervious surfaces, such as roadways andparking lots, and is eventually directed by openditches or subsurface piped systems that dischargeto streams, wetlands, lakes or other surface waterfeatures. Overland flow will result in the directmigration of surficial contaminants, like roadsalts, to surface water features, typically in thetime frame of hours.

In certain settings, a portion ofprecipitation will infiltrate the shallow soil zoneand move laterally through the unsaturated zoneabove the water table as interflow to the localdischarge area. The transport of contaminants tosurface water features as interflow typicallyoccurs in the time frame of hours to days.

The balance of precipitation is subjectto evapotranspiration or infiltrates as groundwaterrecharge to the water table. Contaminantsintroduced to the local, intermediate or regionalgroundwater flow systems may take several yearsto several hundred years before discharging asbaseflow to surface water features.

3.2.2.2 Unsaturated flow and groundwaterrecharge

The amount of precipitation that infiltrates tothe water table depends on a number of factors,including the duration and intensity of the rainfallevent, the initial soil moisture content and the soilmoisture characteristics. Based on the general

temperature and precipitation pattern in Canada,the majority of groundwater recharge in a yearis expected to occur in the late winter andearly spring, as a result of winter snowmelt andspring rains (Gerber and Howard, 1997; Singeret al., 1997).

Vertical travel times through theunsaturated zone may be expected to range fromless than 0.5 m/year to more than 20 m/year. Inareas with thick unsaturated zones (i.e., 20–30 m),travel times for infiltrating precipitation to reachthe water table may be in the range of 5–20 yearsin sandy soils. These travel times are importantwhen considering the impacts to downgradientreceptors and the time required before remedialmeasures implemented at the source would havenoticeable effects.

3.2.2.3 Saturated flow and contaminantmigration

Once below the water table, the movement ofwater is controlled by the fluid potential and thepermeability of the geological material. Whenevaluating the migration of solutes, such assodium, calcium and chloride ions, withingroundwater environments, several other physicaland chemical processes must be considered.Mechanical dispersion, molecular diffusion anddensity effects represent the primary physicaltransport mechanisms that affect the migrationof sodium, calcium and chloride ions.

3.2.3 Impacts of road salts on groundwaterquality

3.2.3.1 Factors controlling the migration ofroad salts

One of the primary concerns associated withinorganic sodium chloride road salts is theirhigh solubility in water. The dissociated ions willmigrate with infiltrating precipitation to the watertable. The migration rate of chloride ion will bethe same as that of the water. This is not the casefor cations such as the sodium ion, which can beaffected by cation exchange reactions in certain


geological environments. The following equationpresents the ion exchange reaction for sodium ion:

2Na+ + Ca(adsorbed) � Ca2+ + 2Na(adsorbed)

In the equation, sodium ion is exchangedwith calcium adsorbed to clay-rich materials.The net effect of this reaction is a decrease insodium ion concentrations within the infiltratinggroundwater and a Na:Cl ratio of less than 1.Reduced sodium ion concentrations in groundwaterimpacted by road salts have been documented byHoward and Beck (1993) and Pilon and Howard(1987).

3.2.3.2 Predicted road salt concentrations ingroundwater

To provide an understanding of the impact thatroad salting may have on groundwater quality, anestimate of the amount of road salt available fordissolution in infiltrating precipitation is required.A simple spreadsheet mass balance modelwas developed to provide an indication of theequilibrium chloride ion concentration in shallowgroundwater (Johnston et al., 2000). The model isbased on the dissolution of the total chloride ionretained within the soil by recharging precipitationand does not consider any additional chlorideion from other sources. The calculated chlorideion concentrations can therefore be consideredrepresentative of regional-scale equilibriumchloride ion concentrations downgradient of asaltable road network. It should be noted thatthe model does not provide an indication ofpotential chloride ion concentrations immediatelydowngradient of a source area, such as anindividual roadway, where higher chloride ionconcentrations may be expected.

Input to the model includes groundwaterrecharge rates, road salt loadings and road densitiestypical of southern Ontario. Recharge rates rangingfrom 10 to 40 cm/year were selected based on arange of soil types. The chloride ion loadingswere based on typical road salt application rates forboth provincial and municipal road networks, asdiscussed in Section 2.2. The road salt available forinfiltration to the water table was estimated based

on the ranges reported by Wulkowicz and Saleem(1974), Scott (1980) and Howard and Haynes(1993). Road densities in the range of 1–15 two-lane-kilometres per square kilometre were selectedbased on the data presented in Morin andPerchanok (2000) and Stantec Consulting Ltd.(2000). Road densities of 0.01–0.1 two-lane-kilometres per square kilometre were assumed tobe characteristic of low-density rural land use, roaddensities of 1–5 two-lane-kilometres per squarekilometre were assumed to be characteristic ofmedium-density rural to semi-rural land use, androad densities of 10–15 two-lane-kilometres persquare kilometre were assumed to be characteristicof urban land use.

Predicted chloride ion concentrations for agiven recharge rate increase with increased loading,road network density and percentage chloride ionretained. Chloride concentrations decrease withincreasing recharge rates due to the dilution effectsof increased recharge. The highest chloride ionconcentration, 3600 mg/L, was predicted for soilwith a recharge rate of 10 cm/year, a loading of40 tonnes chloride ion per two-lane-kilometreand 60% of the chloride ion applied availablefor dilution in recharging precipitation. Forcomparison, the chloride ion concentration for thesame loading and percentage retained is 900 mg/Lfor soils with a recharge rate of 40 cm/year. It isexpected that the percentage of chloride ionretained will vary significantly between a site witha recharge rate of 10 cm/year versus 40 cm/year.This is due to the fact that runoff would be higherfor the site with a recharge rate of 10 cm/year, andtherefore less chloride ion should be retained fordissolution in infiltrating precipitation.

Figures 12–14 present regional-scalegroundwater chloride ion concentrations thatwere predicted for various road salt loadings,road densities and recharge rates. In each figure,an envelope corresponding to the maximum andminimum chloride ion concentrations is presented.Average chloride ion concentration envelopesfor medium- and high-density road networks arealso presented, assuming the average percentagechloride ion retained of 30% and 40%.



FIGURE 12 Estimated chloride concentrations in groundwater for various chloride application ratesand a groundwater recharge rate of 10 cm/year (from Johnston et al., 2000)


For soils with recharge rates of10 cm/year, typical of clay-rich soils, chloride ionconcentrations may exceed 250 mg/L for chlorideion loadings of approximately 5 tonnes chlorideper two-lane-kilometre under high-density roadnetworks (Figure 12). Under medium roaddensities, chloride ion concentrations arepredicted to remain below 250 mg/L for road saltloadings of approximately 10–15 tonnes chlorideper two-lane-kilometre. For low-density roadnetworks, typical of a rural setting, regional-scalechloride concentrations are predicted to remainwell below 250 mg/L (Johnston et al., 2000).

Figure 13 presents the estimatedgroundwater concentrations for a soil with arecharge rate of 20 cm/year, typical of sand. Due todilution from the increased recharge, slightly higherchloride ion loadings can be supported before aconcentration of 250 mg/L is exceeded. Under high-and medium-density road networks, chloride ionloadings above 11 tonnes chloride per two-lane-kilometre and 25–34 tonnes chloride per two-lane-kilometre, respectively, may result in regional-scalechloride ion concentrations above 250 mg/L.

At recharge rates typical of verypermeable soils, such as gravels or fractured rock,groundwater chloride ion concentrations exceed250 mg/L only for high-density road networks.In this case, chloride ion loadings above about16–22 tonnes chloride per two-lane-kilometreare predicted to result in concentrations above250 mg/L (Figure 14).

The mass balance modelling completedfor this section indicates that road salt loadingsabove approximately 15 tonnes chloride per two-lane-kilometre, or 24 tonnes sodium chloride pertwo-lane-kilometre, will result in regional-scalegroundwater concentrations greater than250 mg/L in any soil condition under high-densityland uses. Considering that road salt loadingsin southern Ontario, southern Quebec andthe Maritime provinces are in the range of20–50 tonnes sodium chloride per two-lane-kilometre, highly developed urban areas in theseprovinces are at the greatest risk of wide-scalegroundwater impacts associated with road salting.In addition to groundwater impacts, surfacewater quality within these areas of Canada will



be subject to the discharge of groundwater withelevated sodium and chloride ion concentrationsas baseflow, particularly during the late summerand fall.

The mass balance modelling indicatesthat under low-density land use, chloride ionconcentrations will remain well below 250 mg/L,reaching a maximum concentration of 24 mg/Lfor road salt loadings of 40 tonnes sodiumchloride per two-lane-kilometre. It should benoted, however, that this concentration representsan average regional concentration, and higherconcentrations may be expected immediatelydowngradient from individual roadways.

3.2.3.3 Plume migration and impact assessment

The above section provides an overview of theregional impact associated with non-point sourceloading from road salting in rural and urban areas.The impact on groundwater quality and plumedevelopment from point source loading associatedwith road salting along an individual roadway isdiscussed below.

To provide an indication of maximumchloride ion concentrations and the effects ofdispersion on plume concentrations, a numericalsolute transport model was developed for a simpleunconfined sand aquifer system using the finitedifference modelling programs MODFLOW(McDonald and Harbaugh, 1988) and MT3D(Zheng, 1990). The simplified aquifer wasassigned a lateral hydraulic conductivity of1 × 10–4 m/s, vertical hydraulic conductivity of1 × 10–5 m/s, specific yield of 0.25, effectiveporosity of 0.30, longitudinal dispersivity of10 m, and lateral and vertical transversedispersivities of 1 m and 0.1 m, respectively.Linear constant-head boundaries were assignedat the upgradient (17.5 m) and downgradient(15 m) extents of the model to simulate alateral hydraulic gradient of 0.005 m/m acrossthe 500-m model length. Recharge was assignedat 30 cm/year to the entire model domain.

Within the simulated steady-stateflow field, a square constant concentration sourceterm of 1000 mg/L was assigned at the watertable to represent the chloride ion loading from asection of roadway. This concentration is withinthe range (323–2930 mg/L, average 923 mg/L)for shallow groundwater beneath a four-lanearterial road in Kitchener, Ontario (StantecConsulting Ltd., 2000). The formation andmigration of an unretarded, dissolved-phasechloride ion plume was simulated for 5000 days,followed by the dissipation of the plume afterremoving the concentration source term.

At a simulated time of 100 days,typical of early summer following rechargeof groundwater with elevated chloride ionconcentrations from winter road salting,groundwater concentrations parallel to theplume showed the formation of a typical “cigar-shaped” plume as defined by the 10 mg/L contour.At 100 days, the 10 mg/L contour has migratedapproximately 40 m downgradient from thesource area, with the 250 mg/L contour restrictedto within approximately 10 m of the source area.After a period of approximately 5 years, theplume has reached a steady-state condition asdefined by the 10 mg/L contour in which the fluxfrom the source area is balanced by the fluxacross the plume boundary. The 10 mg/L contourunder steady-state conditions is locatedapproximately 200 m downgradient from thesource, while the 250 mg/L contour moved 20 mdowngradient from the source area.

After a simulation period of approximately5 years, the chloride ion source was removed toprovide an indication of the time required forconcentrations to dissipate. Chloride concentrationsdecreased rapidly after 50 days, with no simulatedconcentrations above 250 mg/L.


The above simulations indicate thatimpacts associated with road salting alongindividual highways may be limited in extentdue to groundwater flow and transport processes,depending on the aquifer properties. Based onthe modelling results, shallow wells and surfacewater features located within 20 m of saltedroads are at the greatest risk of chloride ionimpacts. The modelling does suggest that if roadsalting activities are discontinued, chloride ionconcentrations within the shallow aquifer mayimprove significantly over time frames of monthsto years.

3.2.4 Case studies

This section presents modelling and fieldmeasurements for specific areas in Canada.Additional case studies involving impacts ongroundwater quality associated with roadwaysand maintenance salt storage yards are presentedin Morin (2000).

One method for assessing the regional-scale impacts associated with road salting wasthat applied by Howard and Haynes (1993) in theHighland Creek catchment located near Toronto,Ontario. Through detailed measurements ofstream flow and chloride ion concentrations overa 3-year period, it was estimated that only 45%of the chloride ion applied as road salt withinthe basin was removed by surface water runoff.The remaining 55% was interpreted to be storedwithin the groundwater system. At current saltapplication rates, Howard and Haynes (1993)estimated an average chloride ion concentrationof 426 ± 50 mg/L in groundwater dischargingas baseflow to Highland Creek. While road saltapplication rates were not presented, the totalloading over the basin was approximately 200 gsodium chloride/m2. Assuming a road densitytypical of an urban environment, the road saltapplication rate over the study area is estimated tobe in the range of 20–30 tonnes sodium chlorideper two-lane-kilometre, which is similar to thatpredicted based on the mass balance modellingcompleted as part of this report (Section 3.2.3.2).

Further evidence of the impact ongroundwater quality and discharge to surfacewater features due to road salting in urban areas ispresented in Howard and Beck (1993) and Eylesand Howard (1988). Chloride concentrations insprings discharging to Lake Ontario, along theScarborough Bluffs near Toronto, were measuredand found to be approximately 400 mg/L, similarto those estimated by Howard and Haynes (1993)for the Highland Creek watershed.

Long-term chloride ion concentrationsfrom municipal supply wells provide furtherevidence of impacts on groundwater qualityassociated with winter road salting (Johnstonet al., 2000). Figure 15 presents chloride ionconcentrations from municipal water supply wellsin an urban centre of southern Ontario. Data oncurrent chloride ion application rates in thestudy area near the wells indicate that applicationrates have ranged from 18 to 36 tonnes sodiumchloride per two-lane-kilometre, with an averageapplication rate of 28 tonnes sodium chloride pertwo-lane-kilometre. Chloride concentrationsbeginning in the early 1960s were less than50 mg/L and increased steadily to 250–300 mg/Lin the late 1990s (Figure 15) (data after 1998 notplotted). One of the key issues with respect to thetrends indicated in Figure 15 is that chloride ionconcentrations are continuing to rise and have notreached steady state. Since 1998, chlorideconcentrations at these wells have increased at arate similar to that seen since 1993 on Figure 15,with Well 3 now over 350 mg/L. However, it isnoted that the groundwater being pumped bythese wells represents precipitation that rechargedthe aquifer in the mid to late 1970s, with traveltimes from ground to surface on the order of30 years. Hence, increased chlorideconcentrations observed at these productionwells are the result of road salts that were spread30 years ago. Given that these wells pump onaverage 3.3 million cubic metres per year andhave 20-year capture zones that extend out over12 km2, the concentrations predicted at these wellsare considered to reflect regional-scale chlorideconcentrations. A comparison of the currentconcentrations of approximately 350 mg/L withthe predicted concentrations of 300–450 mg/L


indicates that the mass balance modellingapproach presented in the previous sectionsprovides reasonable estimates of chlorideconcentrations (Johnston et al., 2000).

The time to steady state (the time whensalt inputs are balanced by salt output) dependson local hydrogeological conditions and thesize of the catchment area (Howard et al., 1993).Howard et al. (1993) used two numerical modelsto demonstrate temporal and spatial changes inwater quality. The model FLOWPATH showedthat chemically conservative contaminants such aschloride released within a few kilometres of riversand Lake Ontario will discharge in about 5 years.Contaminants that are released in more centralareas will take more than 100 years to be flushedfrom the groundwater system. When road salt isevenly distributed in a representative 460-km2

region of the Greater Toronto Area, Howard et al.(1993) demonstrated, using the FLOWPATHmodel, that the average chloride concentrationswill reach steady-state values within 200 yearsof initial salt application. Using another model,AQUA, in a smaller area near Toronto, steadystate was reached in just 30 years. At steady state,the levels of chloride at a distance of 200 m fromthe salted highways are 2–3 times the levels ofchloride observed in the discharging baseflow.

Williams et al. (1997) investigatedchloride concentrations in 20 springs insoutheastern Ontario. Chloride concentrationsranged from 8.1 to 1149 mg/L. The higherchloride concentrations originated fromgroundwater that may be contaminated byroad salt. Williams et al. (1999) continuedthis research, noting that the mean chlorideconcentrations were 2.1 mg/L at the Glen MajorsConservation Area and about 100 mg/L inrural areas. Chloride concentrations in springsnear urban areas were higher (>200 mg/L),with the maximum concentration 1345 mg/Land mean 1092 mg/L. The spring with highestconcentrations was adjacent to a highway andbridge. Chloride concentrations at these urbansites also increased between November 1996and November 1997. Salinity contamination wasrelated to road salt. Williams et al. (1999) notedthat spatial patterns in road salt contaminationwere more readily detected by sampling springsthan by sampling creeks, because chlorideconcentrations in spring waters (i.e., discharginggroundwater) exhibited relatively little seasonalvariability. In contrast, chloride concentrations increeks were highly variable seasonally.


FIGURE 15 Chloride concentrations in groundwater from municipal production wells in southernOntario (from Johnston et al., 2000)

Howard and Taylor (1998) reported highconcentrations of chloride in springs dischargingfrom aquifers along the Scarborough Bluffs, astretch of Lake Ontario shoreline east of Toronto,and along an urban–rural transect from theOak Ridges Moraine to Metropolitan Toronto.Concentrations exceeding 2800 mg chloride/Lwere detected in springs issuing from shallowaquifers in the Oak Ridges Moraine–Toronto area.The primary source of the chloride contaminationwas road de-icing chemicals (Howard and Beck,1993).

In 1970, an expressway was constructedat a distance of 50–200 m from wells used by themunicipality of Trois-Rivières-Ouest to obtaingroundwater. By 1976, an increase in salinitywas noted (Gélinas and Locat, 1988). In theearly 1980s, a plume of salty groundwaterwas identified. Chloride concentrations insamples from piezometers ranged between 3and 1500 mg/L. Data collected in 1985 revealedthat the shape and extent of the plume varieddepending on the time of year and the quantityof water pumped from each of the six municipalpumping stations. Data for 1985–86 indicate thatconcentrations gradually increase throughout thespring (during snowmelt) to peak in August andSeptember. Higher concentrations were observedfor a municipal pumping station locateddowngradient from the highway. Concentrationsat most wells upgradient from the highway wereless than 5 mg/L. Mean chloride concentrationsin one station downgradient (80 mg/L) were about6 times higher than the average concentration in anearby station upgradient (14 mg/L).

The municipality of Cap-de-la-Madeleinepumps water from 27 wells. While the quality ofthis water was considered to be extremely good,certain wells have experienced a gradual increasein the concentration of chloride since the 1970sthat is attributed to the use of de-icing road salts.In areas where ditches were not connected toa drainage system, runoff is left to percolatethrough the soils (Delisle, 1999).

3.2.5 Conclusions

The potential for impacts on groundwaterand surface water quality was evaluated usinga mass balance technique that provides anindication of potential regional-scale chloride ionconcentrations downgradient of saltable roadnetworks. The mass balance modelling indicatesthat for road salt application rates above 20 tonnessodium chloride per two-lane-kilometre, regional-scale groundwater chloride ion concentrationsgreater than 250 mg/L will likely result underhigh-density road networks, typical of urbanareas. Considering road salt application ratesthroughout Canada, urban areas in southernOntario, southern Quebec and the Maritimeprovinces are at the greatest risk of wide-scalegroundwater impacts associated with road salting.The mass balance modelling completed as part ofthis report indicates that rural and semi-rural areasare not as likely to be impacted by chloride ionconcentrations above 250 mg/L on a regionalscale. However, local impacts along individualroadways have been documented.

Groundwater containing road saltswill eventually upwell into the surface wateror emerge as springs. Research has shown thatonly a portion of the road salts that are appliedto roads are removed by surface water runoff,and a significant proportion of the salt may bestored within the groundwater system. Elevatedconcentrations of chlorides have been detectedin both groundwater and groundwater that hasemerged to the surface as springs. Predictedand measured concentrations in groundwatersand springs exceed those identified as lethalitythresholds for many organisms, as identified inSection 3.3 (e.g., 10% of species can be expectedto be affected by concentrations greater than240 mg/L).

3.3 Aquatic ecosystems

3.3.1 Scoping and assessment approach

Aquatic ecosystems are vulnerable to variousimpacts from road salts. Evans and Frick (2001)assessed such impacts, focusing on a) a literature


review of the effects of road salt applications onaquatic ecosystems and b) the lethal and sublethaleffect levels of sodium chloride and other chloridesalts as determined from laboratory studies. Thesestudies were then used in the characterization ofthe risks of road salts. Highlights of this reportare presented in this section of the AssessmentReport.

The aquatic ecosystems that wereconsidered in this assessment included a widevariety of habitats, including streams, rivers,wetlands, ponds and lakes. These ecosystemswere located across Canada and in a widevariety of regions, some experiencing minimaland others intense anthropogenic stress. Manyaquatic ecosystems are located in relativelypristine areas of Canada, where anthropogenicimpacts are minimal. Nevertheless, such impactsare of concern, because pristine habitats arevulnerable to relatively minor anthropogenicstresses. Small shifts in species composition,particularly for phytoplankton, are highlyindicative of such impacts. This is because thespecies assemblage of pristine environments is, inlarge measure, composed of taxa that are adaptedto unperturbed environments; pollution-tolerantspecies generally are minor constituents of suchassemblages. Therefore, there is the concern thatsmall increases in chloride concentrations fromhighway runoff or loss from salt storage depotsmay have measurable impacts on suchcommunities.

Other aquatic ecosystems are locatedin rural and agricultural areas. Taxa inhabitingthese environments are experiencing a variety ofstresses as a result of a variety of anthropogenicactivities, including habitat loss and increasedchemical inputs. With increased developmentin the Canadian countryside, including the lossof naturally occurring ponds and wetlands toagricultural usage, the narrow tens of metresof space on the sides of roadways may provideimportant habitat for displaced plant and wildlifecommunities. Moreover, this land can serve asmigration corridors between larger, relativelyundisturbed areas (Nadec, 2001).

Aquatic ecosystems are also found inurban areas, with many such aquatic ecosystemsgradually being incorporated into the growingurban landscape. In such areas, provision ismade to protect these ecosystems by limitingdevelopment around their immediate shorelineand by minimizing anthropogenic impacts intheir immediate watershed. Well-known examplesof such incorporated parkland areas include thoseestablished around the many creeks runningthrough the Metropolitan Toronto area and parkssuch as Stanley Park in Vancouver. These aquaticsystems, while no longer pristine, neverthelessmerit protection against further environmentaldegradation. Increasingly, the design ofsubdivisions includes the incorporation of ruralfeatures such as lakes and creeks. Many of theseengineered ecosystems are built within thenatural landscape over existing aquatic habitats(e.g., ponds, wetlands, sloughs, creeks), whichare rapidly recolonized by temporarily displacedaquatic organisms.

Stormwater detention ponds, which areoften constructed in a natural aquatic ecosystem,e.g., a creek or wetland, are engineered habitats.Nevertheless, these ponds do provide habitat fora wide variety of organisms (Bishop et al., 2000).While clearly not ideal because of the varietyof contaminants associated with them and alteredhydrological regime, they provide aquatic habitatand resources for both the organism and thepopulation (which requires a critical area for self-maintenance) in an increasingly fragmented urbanand rural landscape.

Aquatic organisms considered in theassessment included all components of theaquatic ecosystem, i.e., bacteria, fungi,protozoans, zooplankton, benthic invertebrates,fish and amphibians. However, not all groupswere equally well represented in the literature.For example, laboratory studies investigatingchloride toxicity were heavily weighted towardsfish, zooplankton and benthic invertebratesand for short-term exposures. Such studies, bytheir nature, focus on the capacity of species toendure short-term osmotic stress due to elevated


salinities. Few studies investigated chronictoxicity; chronic toxicity was estimated usingapproaches followed by the U.S. EPA (1988) forsetting water quality guidelines for chloride.

Relatively few field studies have beenconducted specifically investigating the impactsof road salts on aquatic ecosystems. The majorityof these studies were conducted in the 1970s and1980s. Many of the studies that were found wereconducted in the United States but have generalapplication to the Canadian environment.

As provided for in CEPA 1999 and asnoted in Environment Canada (1997a), a weightof evidence approach involving consideration ofseveral lines of evidence is used to strengthen theconfidence in assessment conclusions. These linesof evidence include consideration of short-termtoxicity studies, estimates of acute toxicity, fieldstudies, emerging information and the generalknowledge of the use patterns of road salts inthe Canadian environment. In addition, a tieredapproach, using quotient-based assessments, isused to determine the likelihood of adverse effectsoccurring in aquatic ecosystems as a result ofrelease of road salts into the environment.

3.3.2 Laboratory studies

The literature review of laboratory studies ofchloride toxicity focused on obtaining informationon sodium chloride toxicity, but also collecteddata for calcium, potassium and magnesiumchloride toxicity, since these chlorides are alsopresent in various road salt formulations. Toxicitydata are essential for assessing the environmentalconcentrations and exposure durations that maybe harmful. Both short-term (hours to days)toxicity and long-term (weeks to months) toxicityare of concern. In the following paragraphs, thehighlights of these determinations are presented.

3.3.2.1 Short-term or acute toxicity studies

One exposure scenario that is of concern is short-term exposures (hours to days) to elevatedchloride concentrations, particularly when thedilution capacity of the receiving water body is

weak. Direct highway runoff, particularly intocreeks, streams and other roadside waterways,is of particular concern. Organisms inhabitinglarger water bodies such as rivers and lakes, withtheir greater dilution capacity, are less likely to beexposed to conditions in which acute toxicity isof concern. Organisms inhabiting wetlands, whichare relatively small, stationary water bodies withslow water exchange, may experience significantacute and chronic toxicity.

An extensive review of the literaturerevealed several dozen studies that investigatedthe short-term or acute toxicity of sodium chlorideto aquatic organisms; fewer studies were locatedfor magnesium, calcium and potassium salts(Evans and Frick, 2001). These acute studies weresubsequently grouped into four time intervals toestimate toxicity following various exposures toelevated chloride concentrations:

• <24 hours: For exposures of less than 1 day,four studies were located for fish (3) and abenthic organism (1) (Table 10). The LC50sranged from 6063 to 30 330 mg chloride/L,with a geometric mean of 12 826 mgchloride/L. Natural salinities in this rangeare associated with estuaries, tidal marshes,oceans and inland saline lakes. Salinities inthis range have also been associated withhighway runoff (and spray) from multiple-lane highways, waste snow from urban areas,leachate and groundwater from patrol yards.Elevated chloride concentrations at the lowerend of this range have also been observed inwell water, a stream and a wetland near a roadsalt depot (see Sections 2.3, 2.4, 2.6 and 3.4).

• 24 hours: For exposures of 24 hours, sevenstudies were located testing fish (4) andcladocerans (3) (Table 11). The LC50s rangedfrom 1652 to 8553 mg chloride/L, with ageometric mean of 3746 mg chloride/L.Salinities in this range are associated withestuaries, tidal marshes and inland salinelakes. Salinities in this range have also beenassociated with highway runoff (and spray)from multiple-lane highways, waste snowfrom urban areas and leachate from patrol


yards. Elevated chloride concentrations atthe lower end of this range have also beenobserved in well water, a stream and awetland near a road salt depot (see Sections2.3, 2.4, 2.6 and 3.4).

• 3 or 4 days: Several studies were foundinvestigating the 4-day toxicity of sodiumchloride to a variety of organisms, and asmaller set of 3-day exposures was alsofound. The 3-day exposure data wereconverted into 4-day estimates using aconversion factor as described in Evans andFrick (2001). This resulted in 28 observations,including fish (13), cladocerans (7) and otherinvertebrates (8) (Table 12). Some specieswere the subject of several studies, e.g.,Daphnia magna and fathead minnow(Pimephales promelas). There weredifferences in some LC50s, probably as a resultof test conditions (e.g., use of reconstitutedwater in Birge et al., 1985) and/or differencesin the physiological tolerances of the testorganisms. In general, fish had greatertolerances than invertebrates. The LC50sranged from 1400 to 13 085 mg chloride/L,with a geometric mean of 4033 mgchloride/L. The higher 4-day than 1-daytolerances is due, in large measure, to theinclusion of mosquito fish (Gambusia affinis,a hardy fish that is used worldwide formosquito control) and American eel (Anguilla

rostrata, a fish that spends its adult life in theocean) in the data set. Salinities in this rangeare associated with estuaries, tidal marshesand inland saline lakes. Salinities in thisrange have also been associated withhighway runoff (and spray) from multiple-lane highways, waste snow from urban areasand leachate from patrol yards. Elevatedchloride concentrations at the lower end ofthis range have also been observed in wellwater, a stream and a wetland near a road saltdepot (see Sections 2.3, 2.4, 2.6 and 3.4).Moreover, salinities in the 1000–4000 mg/Lrange have been reported for several creeks inthe Metropolitan Toronto area.

• 7–10 days: Seventeen studies were foundestimating chloride toxicity at exposuretimes of 7–10 days (Table 13). Exposuresof this duration are at the upper end ofacute exposure times but are not sufficientlylong to be considered chronic for long-livedorganisms such as amphibians and fish.These studies examined responses such asmortality, reduced growth, fecundity andfailure to complete development, e.g., fromegg to embryo. There were five fish tests, twoamphibian tests, nine zooplankton tests andone algal test. The LC50s and EC50s rangedfrom 874 to 3660 mg chloride/L, with ageometric mean of 1840 mg chloride/L.


TABLE 10 Toxicity responses of organisms to sodium chloride for exposures less than 1 day(from Evans and Frick, 2001)

Species Common NaCl Cl (mg/L) Response Time Referencename/taxon (mg/L) (hour)

Salvelinus Brook trout 50 000 30 330 LC50 0.25 Phillips, fontinalis 1944

Lepomis Bluegill 20 000 12 132 LC47 6 Waller et macrochirus al., 1996

Oncorhynchus Rainbow trout 20 000 12 132 LC40 6 Waller et mykiss al., 1996

Chironomus Chironomid 9 995 6 063 LC50 12 Thorton attenatus and Sauer,

1972

61

TABLE 11 Toxicity responses of organisms to sodium chloride for exposures of 1 day (from Evans andFrick, 2001)

Species Common NaCl Cl Response Referencename/taxon (mg/L) (mg/L)

Lepomis macrochirus Bluegill 14 100 8 553 LC50 Doudoroff and Katz, 1953Daphnia magna Cladoceran 7 754 4 704 LC50 Cowgill and Milazzo, 1990Cirrhinius mrigalo Indian carp fry 7 500 4 550 LC50 Gosh and Pal, 1969Labeo rohoto Indian carp fry 7 500 4 550 LC50 Gosh and Pal, 1969Catla catla Indian carp fry 7 500 4 550 LC50 Gosh and Pal, 1969Daphnia pulex Cladoceran 2 724 1 652 LC50 Cowgill and Milazzo, 1990Ceriodaphnia dubia Cladoceran 2 724 1 652 LC50 Cowgill and Milazzo, 1990

TABLE 12 Four-day LC50s of various taxa exposed to sodium chloride (from Evans and Frick, 2001)

Species Common NaCl Cl Referencesname/taxon (mg/L) (mg/L)

Anguilla rostrata American eel, 21 571 13 085 Hinton and Eversole, 1978black eel stage

Anguilla rostrata American eel, 17 969 10 900 Hinton and Eversole, 1978black eel stage

Gambusia affinis Mosquito fish 17 500 10 616 Wallen et al., 1957Lepomis macrochirus Bluegill 12 964 7 864 Trama, 1954Oncorhynchus mykiss Rainbow trout 11 112 6 743 Spehar, 1987Pimephales promelas Fathead minnow 10 831 6 570 Birge et al., 1985Culex sp. Mosquito 10 254 6 222 Dowden and Bennett, 1965Lepomis macrochirus Bluegill 9 627 5 840 Birge et al., 1985Pimephales promelas Fathead minnow 7 681 4 600 WI SLOH, 1995Pimephales promelas Fathead minnow 7 650 4 640 Adelman et al., 1976Carassius auratus Goldfish 7 341 4 453 Adelman et al., 1976Anaobolia nervosa Caddisfly 7 014 4 255 Sutcliffe, 1961Limnephilus stigma Caddisfly 7 014 4 255 Sutcliffe, 1961Daphnia magna Cladoceran 6 709 4 071 WI SLOH, 1995Chironomus attenatus Chironomid 6 637 4 026 Thorton and Sauer, 1972Daphnia magna Cladoceran 6 031 3 658 Cowgill and Milazzo, 1990Hydroptila angusta Caddisfly 5 526 4 039 Hamilton et al., 1975Cricotopus trifascia Chironomid 5 192 3 795 Hamilton et al., 1975Catla catla Indian carp fry 4 980 3 021 Gosh and Pal, 1969Labeo rohoto Indian carp fry 4 980 3 021 Gosh and Pal, 1969Cirrhinius mrigalo Indian carp fry 4 980 3 021 Gosh and Pal, 1969Lirceus fontinalis Isopod 4 896 2 970 Birge et al., 1985Physa gyrina Snail 4 088 2 480 Birge et al., 1985Daphnia magna Cladoceran 3 939 2 390 Arambasic et al., 1995Daphnia magna Cladoceran 3 054 1 853 Anderson, 1948Ceriodaphnia dubia Cladoceran 2 630 1 596 WI SLOH, 1995Daphnia pulex Cladoceran 2 422 1 470 Birge et al., 1985Ceriodaphnia dubia Cladoceran 2 308 1 400 Cowgill and Milazzo, 1990

PSL ASSESSMENT REPORT — ROAD SALTS

This salinity range is defined as subsaline(Hammer, 1986). Such salinities are foundat the interface between the marine andfreshwater realms and in inland subsalinelakes. Species diversity declines rapidlywith increasing salinity in this range (Wetzel,

1983). Salinities in this range have beenassociated with highway runoff (and spray),waste snow, leachate from patrol yards,well water, a stream, a wetland near a roadsalt depot, and creeks and rivers in theMetropolitan Toronto area.


TABLE 13 Seven- to 10-day LC50s and EC50s of various taxa exposed to sodium chloride 1 (from Evansand Frick, 2001)

Species Common NaCl Cl Response Referencename/taxon (mg/L) (mg/L)

Daphnia magna Cladoceran 6 031 3 660 LC50 Cowgill and Milazzo, 1990Daphnia magna Cladoceran 5 777 3 506 EC50 (mean Cowgill and Milazzo, 1990

number of broods)

Pimephales promelas Fathead minnow 5 490 3 330 LC50 Beak, 1999larvae

Pimephales promelas Fathead minnow 4 990 3 029 EC50 Beak, 1999larvae (growth)

Daphnia magna Cladoceran 4 310 2 616 EC50 Cowgill and Milazzo, 1990(dry weight)

Daphnia magna Cladoceran 4 282 2 599 EC50 (total Cowgill and Milazzo, 1990progeny size)

Daphnia magna Cladoceran 4 040 2 451 EC50 (mean Cowgill and Milazzo, 1990brood size)

Xenopus laevis Frog embryo 2 940 1 784 LC50 Beak, 1999Oncorhynchus mykiss Rainbow trout 2 630 1 595 LC50 (survival) Beak, 1999

embryo/alvinXenopus laevis Frog embryo 2 510 1 524 EC50 (survival) Beak, 1999Nitzschia linearais Diatom 2 430 1 474 EC50 (cell Gonzalez-Moreno et al.,

numbers) 1997Oncorhynchus mykiss Rainbow trout 2 400 1 456 LC50 (survival) Beak, 1999

egg embryoCeriodaphnia dubia Cladoceran 2 077 1 261 LC50 Cowgill and Milazzo, 1990Ceriodaphnia dubia Cladoceran 1 991 1 208 EC50 (mean Cowgill and Milazzo, 1990

number of broods)

Ceriodaphnia dubia Cladoceran 1 761 1 068 EC50 (mean Cowgill and Milazzo, 1990brood size)

Ceriodaphnia dubia Cladoceran 1 761 1 088 EC50 Cowgill and Milazzo, 1990(total progeny)

Pimephales promelas Fathead minnow 1 440 874 LC50 (survival) Beak, 1999embryos

1 LC50/EC50 ratio calculated for studies where both parameters were estimated (i.e., Cowgill and Milazzo, 1990; Beak, 1999).

3.3.2.2 Chronic toxicity

Long-term or chronic toxicity to sodium chloride,which may be expected to occur over weeks tomonths, has seldom been investigated in thelaboratory. Only two studies were located in theliterature review.

Birge et al. (1985) estimated the 4-dayLC50 toxicity of chloride to fathead minnow as6570 mg/L. The NOEC for the 33-day early lifestage test was 252 mg chloride/L and the Lowest-Observed-Effect Concentration (LOEC) was352 mg/L, giving a geometric mean of 298 mgchloride/L as the estimated chronic value and anacute to chronic ratio (ACR) of 22.1. Birge et al.(1985) also tested Daphnia pulex. The 4-day LC50

was 1470 mg chloride/L. For the 21-day test,the chronic toxicity, calculated as the geometricmean of the NOEC (314 mg/L) and the LOEC(441 mg/L), was 372 mg chloride/L. This valueis similar to that for fathead minnow. However,because of the lower acute toxicity value forDaphnia, the ACR was 3.95. Birge et al. (1985)conducted a second 4-day toxicity test with D.pulex using natural stream water and reportedan EC50 of 3050 mg chloride/L, or twice thetolerance of that using reconstituted water. (If thisvalue were used to estimate the ACR, the valuewould become 8.20.) Birge et al. (1985)calculated the geometric mean of the Daphniaand fathead minnow chronic toxicity values toestimate a chronic chloride toxicity concentrationof 333 mg/L, although they also indicated someuncertainty in the D. pulex values.

In a later study, the U.S. EPA (1988)developed water quality criteria for chloride,relying largely on the results of the Birge et al.(1985) study for acute toxicity and ACRestimates. They used the same toxicity dataand ACR (3.95) for Daphnia pulex. However,for fathead minnow, they estimated the chronictoxicity as 433.1 mg/L (based on the geometricmean of 9% impaired survival at 352 mg/L and15% impaired survival at 533 mg/L), giving anACR of 15.17. They also cited a rainbow trout(Salmo gairdneri, now Oncorhynchus mykiss)

study conducted by Spehar (1987), who reportedan acute toxicity of 6743 mg chloride/L, a chronictoxicity of 922.7 mg chloride/L and an ACR of7.31. The geometric mean of these three ACRswas used to develop a final ACR of 7.59. A finalacute genus toxicity value of 1720 mg/L, basedon the most sensitive genus (Daphnia), wasdivided by this ACR to give a final chronictoxicity value of 226.5 mg/L.

A modification of the ACR approachwas used to estimate the chronic toxicity of theorganisms for which there were 4-day acutetoxicity data. Two approaches could have beenused. The first was to apply the Daphnia ACR(3.95) as developed by Birge et al. (1985) andlater used by the U.S. Environmental ProtectionAgency (U.S. EPA, 1988) to the invertebrate dataand the geometric mean (10.53) of the fish ACRdata reported by the U.S. EPA. However, Birgeet al. (1985) considered the acute toxicity data(and ACR) to be too low. Application of twodifferent ratios also changes the general orderingof species sensitivities based on the 4-day acutetoxicity data, so that fish now appear moresensitive and aquatic invertebrates more tolerantto long-term elevated chloride exposures.This seems highly unlikely, with aquaticorganisms such as cladocerans and insectlarvae being poorly represented in the marine andestuarine environment, while fish taxa such as thesalmonids have strong marine affinities. A moresuitable approach, then, is to follow the U.S. EPAapproach and to use the mean ACR of 7.59 basedon the geometric mean of the three ACR studies.

Using a mean ACR of 7.59, the acutetoxicity of chloride was estimated as ranging from184.5 mg/L for Ceriodaphnia dubia to 1724 mg/Lfor the black eel stage of the American eel, witha geometric mean value of 512.6 mg/L chloride(Evans and Frick, 2001). The lower value isapproximately 24 times greater than the averagechloride concentration (8 mg/L) of the world’sriver waters (Wetzel, 1983). The geometric meanof the chronic toxicities, 551.9 mg chloride/L, isat the lower end of the 500–1000 mg chloride/Lrange, where many researchers have reported


losses in freshwater species with increasingsalinity (Williams, 1987; Hart et al., 1990;Leland and Fend, 1998; Evans and Frick, 2001).A marked drop in the species diversity offreshwater ecosystems occurs as salinity increasesfrom about 2000 mg/L to about 5000 mg/L(Figure 16; see also Wetzel, 1983). In general,for freshwater species, the number of speciesdecreases as salinity increases, with the greatestand most rapid decrease in species numbers atchloride ion concentrations of 1000–3000 mg/L.Salinities in this 185–1724 mg/L range havebeen associated with highway runoff (and spray),waste snow, leachate from patrol yards and creeksin the Metropolitan Toronto area. The lowersalinity values have been associated withchloride-impacted lakes.

3.3.3 Assessment of toxicity to aquaticecosystems

Three approaches were used to assess the toxicityof road salts in the aquatic environment. Theseare a) toxicity thresholds and quotient-based riskcharacterization, as described in EnvironmentCanada (1997a); b) field evidence of populationor ecosystem effects; and c) consideration ofabiotic effects. Briefy, these approaches can bedescribed as follows:

• Toxicity thresholds and quotient-based riskcharacterization: This is a quotient-basedapproach, which is used to determine thelikelihood of adverse effects occurring inaquatic ecosystems as a result of release ofroad salts into the environment. This approachis based on comparisons of concentrationsof chloride in the aquatic environmentresulting from the release of road salts(based on literature and analyses done forthis assessment) and toxicity thresholdsfor chloride salts as identified in laboratorystudies.

• Field evidence of population or ecosystemeffects: The quotient-based approach generallyfocuses on the consideration of toxicitythresholds. However, community structure

can be affected by environmental conditionsthat, while not necessarily exposingorganisms to concentrations of contaminantsthat are themselves toxic, can result inselective advantage for certain populations orspecies and thereby result in shifts inpopulations and changes in communitystructure. Thus, field studies are particularlywell suited for investigating the chroniceffects of elevated chloride concentrations inthe environment. However, because not allvariables are controlled in such studies, it ispossible that other, covarying variables affectsome observed responses.

• Abiotic effects: Through their interactionswith physical, chemical or biologicalcomponents of the environment, substancesmay alter the abiotic environment and therebyaffect ecosystems. Road salts may operate intwo important ways. The first is by affectingthe density of water and hence its mixingproperties. This is especially important inrelatively stationary water bodies such aslakes, ponds and wetlands. Second, chloridemay increase the solubility of a variety ofcompounds, including metals.

3.3.3.1 Quotient-based risk characterization

The quotient-based environmental riskassessment is based on the procedures outlined inEnvironment Canada (1997a). Based on possibleexposure scenarios, a hyperconservative orconservative Estimated Exposure Value (EEV)is selected. Next, an Estimated-No-EffectsValue (ENEV) is determined by dividing anexperimental Critical Toxicity Value (CTV) by anapplication factor to account for the uncertaintysurrounding the extrapolation to chronic or pulsedexposure, the extrapolation from laboratory tofield conditions, and interspecies and intraspeciesvariations in sensitivity. The CTV can be basedon a wide range of toxicity responses, includingmortality, growth, reproduction, fecundity,longevity, productivity, community structure anddiversity.


After calculating ENEVs, ahyperconservative (Tier 1) quotient (EEV/ENEV) is calculated for pertinent sensitive taxa in order to determine whether there ispotential environmental risk from the release ofthat substance. If these quotients are less than 1,it can be concluded that the substance poses littlerisk to the environment, and the risk assessmentis completed. If, however, the quotient is greaterthan 1 for a particular assessment endpoint, thenthe risk assessment for that endpoint proceedsto a Tier 2 analysis, where a more realistic butconservative quotient is calculated for eachpertinent taxon. If these quotients are greater than1, the assessment proceeds to a Tier 3 level. Here,the broader likelihood and magnitude of effectsare considered. This latter approach involvesa more thorough consideration of sources ofvariability and uncertainty in the risk analysis.

3.3.3.2 Tier 1 and Tier 2 assessments

3.3.3.2.1 Estimated Exposure Values (EEVs)

A Tier 1 assessment is based on ahyperconservative exposure scenario, generally

assuming that the EEV is the maximumconcentration measured or likely to beencountered in the Canadian environment. Thehighest reported sodium chloride concentrationsin the aquatic environment have been associatedwith highway runoff, salt-contaminated snow andcontaminated waters near salt storage depots.

In Ontario, chloride concentrationsas high as 19 135 mg/L were reported inhighway runoff from the Skyway Bridge(Mayer et al., 1998). Delisle et al. (1995) assessedthe concentration of chloride in snow clearedfrom city streets in Montréal. The averageconcentration of chloride was 3851 mg/L, with amaximum reported concentration of 10 000 mg/L.Concentrations of chloride in waste snow inMontréal averaged 3115 mg/L for secondarystreets and 5066 mg/L for primary streets(Delisle and Dériger, 2000; Section 2.3.1).Chloride concentrations in leachate frompoorly maintained salt storage depots may reach66 000 mg/L (Section 2.4.2.1). While noCanadian wetland studies have been identified, inMaine, chloride concentrations in surface watersof a bog located near a salt storage depot reached


FIGURE 16 Species diversity across a salinity and chloride gradient (from Wetzel, 1983)

13 500 mg/L; concentrations remained elevatedthroughout the March to October sampling period(Ohno, 1990). Therefore, for the purposes ofTier 1 assessments, an EEV of 10 000 mg/Lwas used to represent estimated maximumchloride concentrations. Such a concentrationis representative of snowmelt draining fromhighways into small receiving water bodies suchas roadside creeks and wetlands and approximatesthat observed in wetlands near poorly designedroad salt storage depots.

Tier 2 assessments involve a furtheranalysis of exposure and/or effects to calculatea quotient that is still conservative, but is morerealistic than the hyperconservative quotientcalculated in Tier 1 (Environment Canada,1997a). In Canada, the best-documented impactsof road salt on chloride levels in groundwaterand other aquatic systems have been for theMetropolitan Toronto area. Williams et al. (1997,1999) demonstrated the impact of road salts ongroundwater and the invertebrate communitiesinhabiting groundwater-fed springs. There isalso a comprehensive water quality data setfor the numerous creeks flowing through theMetropolitan Toronto area. The southernOntario region is also an area of high roadsalt application rates. Hence, this data set wasexamined to estimate a Tier 2 EEV.

Chloride concentrations were measuredin several streams in the Toronto RemedialAction Plan watershed over 1990–1996 (seeEvans and Frick, 2001; Table 9). Data werecollected seasonally. Maximum chlorideconcentrations reported at individual stationsincluded 2140–3780 mg/L for three stations onEtobicoke Creek, 3470 mg/L at Mimico Creek,4310 mg/L at Black Creek, 96–4310 mg/L at fivestations on the Humber River, 960–2610 mg/Lat three stations on the Don River and 1390 mg/Lat Highland Creek. Seasonal plots of chlorideconcentration variations in Highland Creekover 1990–1993 show several winter samplingperiods in which chloride concentrationsexceeded 1000 mg/L for what appear to beseveral periods of 1 week. Mean chloride

concentrations at these sites ranged from 278 to553 mg/L, with winter concentrationsapproximately twice those of summer. Williams etal. (1999) reported that an Ontario spring locatednear a highway and bridge had a mean chlorideconcentration of 1092 mg/L as a probable resultof road salt contamination of groundwater. Whilethe maximum chloride value observed in creeksand rivers in the Toronto area that have beencontaminated by road salts is in the range of2000–4000 mg/L, the frequency and duration ofsuch occurrences are likely variable. Therefore, aless conservative estimate of 1000 mg/L was usedas the EEV in the Tier 2 assessments. Chlorideconcentrations in this range have been commonlyobserved in Toronto-area creeks and rivers, in acontaminated spring in the Toronto area and in abog contaminated by a salt storage depot (Wilcox,1982). It is assumed that, for these scenarios,exposure times would be in the order of days.

3.3.3.2.2 Critical Toxicity Values (CTVs)

Data used for generating the CTVs used in Tier 1and 2 assessments were obtained during theliterature review as described above and presentedin detail in Evans and Frick (2001). Suitable datawere not found for all trophic groups. The mostcomprehensive data were from the 2- to 4-daystudies. The endpoint most typically investigatedwas mortality (zooplankton, benthos), unsuccess-ful development of eggs to larval stages (EC50

data) and reduced growth (phytoplankton,macrophytes).

To calculate ENEVs, an application factorof 100 was used for toxicity data based on LC50

and EC50 data as recommended in the guidancemanual (Environment Canada, 1997a) and 10 fortoxicity data based on LC25 and EC25 data.

3.3.3.2.3 Conclusions for Tier 1 and Tier 2assessments

The Tier 1 assessments indicated that riskquotients were greater than 1 for all taxaconsidered. Accordingly, a Tier 2 assessment wasconducted for all taxa. Tier 2 assessments were


conducted by reducing the EEV from 10 000 mgchloride/L to 1000 mg chloride/L. Results arepresented in Table 14 (given the assumeddifferences in exposure scenarios between Tier 1and Tier 2, all Tier 1 quotients are 10 timesgreater than the Tier 2 quotients). Even if theTier 2 application factors were reduced by afactor of 10, most quotients would still exceed 1.Tier 3 assessments were therefore done for allgroups of organisms considered in Tiers 1 and 2.

3.3.3.3 Tier 3 assessments with field validation

Tier 3 assessments include considerations of thedistributions of exposures and/or effects in theenvironment (Environment Canada, 1997a). Thisis a challenging task, given the wide variety ofroad salt entries into the Canadian environment,e.g., leakage from salt storage depots, road andhighway runoff (and spray) and transport viacreeks into larger aquatic ecosystems, such aswetlands, rivers, ponds and lakes. It is also a


TABLE 14 Summary of Tier 2 calculations (from Evans and Frick, 2001)

Organism Exposure time/ CTV Reference Application EEV/ENEVendpoint NaCl factor

(and Cl)(mgL)

1. Fungi 48-hour, increased 659 Sridhar and 100 250Unknown aquatic fungi sporulation (400) Barlocher, 1997

2. Protozoans 17% reduction of cells 577 Cronkite 1 2.9Paramecium cultured in light, (350) et al., 1997tetrourelia 57-day test

3. Phytoplankton 120-hour, 50% 2430 Patrick et al., 100 67.8Nitzschia linearais reduction in number (1475) 1968

of cells

4. Macrophytes 45-day, 43% reduction 2471 Wilcox, 1984 100 66.7Sphagnum fimbriatum in growth (1500)

5. Zooplankton 7-day, 50% mortality 2019 Cowgill and 100 81.2Ceriodaphnia dubia (1260) Milazzo, 1990

6. Benthic invertebrates 48-hour LC25 2000 Hamilton 10 8.2Nais variabilis (1214) et al., 1975

7. Amphibians 7-day EC50 2510 Beak, 1999 100 65.6Xenopus laevis (1524)

8. Fish 7-day EC25, 1630 Beak, 1999 10 10.1Oncorhynchus mykiss egg/embryo (989)

complex task, as watercourses vary tremendouslyin their size, location in relation to a potentialsource of road salt source, flow regimes, etc.

While data on chloride concentrationshave been collected as part of routine waterquality monitoring programs, most data havenot been subjected to rigorous analysis for spatialand temporal distributions and for causal factors.Thus, while Evans and Frick (2001) conductedan extensive review of the literature, a relativelysmall number of case history studies werelocated, especially for the Canadian environment.This situation may begin to change as researchersand agencies devote more effort to betterunderstanding the impacts of road salts onchloride concentrations in the environment.

In order to conduct the Tier 3 assessment,a scenario approach is employed to describethe different settings in which road salts haveor may have impacted the Canadian aquaticenvironment. This approach is based oninformation on road salt usage patterns acrossCanada, known information on chloride levelsin the aquatic environment, toxicity data asdescribed previously and reported environmentalimpacts or case studies. Many of the case studiesare drawn from literature in the United States,where a greater number of studies have beenconducted, but are representative of use inCanada.

3.3.3.3.1 Case 1: Road salt runoff and urbancreeks, streams and small rivers indensely populated areas

As cities expand into the rural landscape, theyenclose an increasingly large network of creeks,streams and rivers. Moreover, two-lane roadsincrease in density and, for larger cities, four-laneand larger highways are constructed, resultingin a highly altered watershed. Unlike the rurallandscape, where a significant fraction of therunoff may be retained in relatively closeproximity to the roadway, in the highly pavedurban setting, much of the runoff is delivered inpulses into creeks, streams and rivers. Moreover,the absolute volume of road salt used is higher.

The Metropolitan Toronto area representsthe clearest example of the impacts of a densenetwork of highways on the numerous creeks,streams and, eventually, rivers flowing throughthis densely populated urban area. As they flowsouth towards Lake Ontario, these creeks pick upmore and more of the urban runoff. As previouslynoted, chloride concentrations can exceed severalhundred milligrams per litre for much of thewinter, with peak concentrations reaching1000–4000 mg/L. Watercourses with highchloride concentrations include Etobicoke,Mimico, Black and Highland. The Don Rivercan also have high chloride concentrations duringmuch of the winter.

Schroeder and Solomon (1998) conducteda formal investigation of chloride concentrationsand toxicity at three stations on the Don River.Chloride concentrations from November 1995 toApril 1996 ranged up to about 950 mg/L at site 1,up to about 2600 mg/L at site 2 and up to about1150 mg/L at site 3. These researchers alsocollected acute lethal toxicity data (LC50 and EC50)for 13 fish and 7 invertebrate taxa; values rangedfrom about 1000 to 30 000 mg/L. Next, theresearchers superimposed the cumulative toxicitydata on the cumulative chloride distribution data.They then estimated that 10% of the specieswould be expected to experience acute effects ofchloride toxicity at least 90% of the time at twoof the sites and at least 85% of the time at thethird site.

Several field studies were located thatprovided corroborating evidence that chlorideconcentrations in the 500–1000 mg/L rangecan impact stream and creek ecosystems. Thefollowing paragraphs summarize some of thesestudies.

Williams et al. (1997) detected significantvariations in macroinvertebrate communitystructure related to different concentrationsof chloride in 20 groundwater-fed springsin southeastern Ontario in the MetropolitanToronto area. The chloride content of thesesprings ranged from 8.1 to 1149 mg/L. Tipulidaeand Ceratopogonidae were associated with


springs containing higher chloride concentrations,whereas taxa such as Gammarus pseudolimnaeus(an amphipod) and Tubellaria (a flatworm)were found only in springs with low chlorideconcentrations. Crowther and Hynes (1977)reported that 20% of a test population ofG. pseudolimnaeus dies with a 1-day exposureto 4121 mg sodium chloride/L (2500 mgchloride/L). The higher chloride concentrationsin some of the springs originated fromgroundwater that apparently had beencontaminated by road salt.

Crowther and Hynes (1977)experimentally investigated the effect of chlorideconcentrations on the drift of benthic invertebratesin Lutteral Creek, a trout stream located insouthern Ontario. This creek is a small, spring-fedtributary of the Speed River. They alsoinvestigated chloride concentrations in LaurelCreek (annual discharge 19 × 106 m3), a tributaryto the Grand River, which passes through urbanWaterloo. At that time, Waterloo had a populationof 32 000. Therefore, the general results of thestudy by Crowther and Hynes (1977) may applyto the creeks flowing through the numerous smalltowns in such regions as southern Ontario,Quebec and the Maritimes, where road salt isheavily used. Chloride levels in Laurel Creekvaried seasonally, with peak concentrations ashigh as 1770 mg/L being observed for a 2-weekperiod between January 5 and February 6, 1975.

Considering the elevated concentrationsin Laurel Creek, a series of experiments wasconducted in Lutteral Creek to assess theimpacts of pulses of sodium chloride–ladenwater on benthic drift. Fifteen metres of LutteralCreek were divided longitudinally by sheets ofcorrugated steel and benthic drift assessed on thecontrol side and experiment side of the partition.Highlights are presented in Evans and Frick(2001).

Crowther and Hynes (1977) concludedtheir study by noting that, as chloride levelsin streams approach 1000 mg/L, impacts onbenthic drift may begin to appear in streams.

During their study, they noted that most of theseeffects would be expected to occur during winterand early spring. However, they warned that, asgroundwater continued to become contaminatedwith chloride salts, stream flows might becontaminated by chloride-rich groundwaterinflows during summer low-flow periods.

Dickman and Gochnauer (1978) studiedthe effects of exposure to sodium chloride on thedensity of bacteria and algae in Heyworth Stream,near Heyworth, Quebec, between July 24 andOctober 1, 1973. Sodium chloride was added atone location along the stream to maintain chlorideconcentrations of 1000 mg/L (1653 mg sodiumchloride/L). This concentration was selectedto simulate the impact of road salt fromstorm sewers on a receiving stream microfloraecosystem. It would also simulate the impacts ofelevated chloride concentrations in groundwaterof spring ecosystems. Sodium chlorideconcentrations at the control site were 2–3 mg/L.The stream was described as shallow, fast flowingand densely shaded. Artificial substrates (tiles)were placed in the creek at the experimental siteand at an upstream control and recovered atweekly intervals. The treated site had loweralgal diversity and lower standing crops ofphotosynthetic periphyton or algae, which wasrelated to osmotic stress. Auxospore (a restingstage) formation of the diatom Cocconeisplacentula was noted at the treated site onday 28 but not at the control site; auxosporeformation is often triggered by environmentalstress. Density of bacteria was greater at thetreated site, which was attributed to a reductionin the grazing pressure on the bacterial populationbecause of the reduced number of grazers, such asflagellates, ciliates and rhizopods. The incidenceof diatom parasitism was lower at the treated sitethan at the control site, possibly because thesodium chloride inhibited fungal growth, asindicated by other researchers (Kszos et al., 1990;Rantamaki et al., 1992).

Other studies investigated the effects ofsalinity of benthic and algal community structure.Although the salinity was not due to road salt,


these studies provide further support for thehypothesis that stream community compositionis sensitive to salinity, with species compositionchanging as salinity increases towards 1000 mgchloride/L.

Leland and Fend (1998) investigatedbenthic distributions in the San Joaquin RiverValley, California. Total dissolved solidsconcentrations ranged from 55 to 1700 mg/L.Distinct assemblages were observed at the highand low salinities, and the distribution of manytaxa indicated salinity optima. Ephemeroptera(mayfly larvae) rarely occurred at salinitiesabove 1000 mg/L (600 mg chloride/L). Whileother, covarying factors may have affected theserelationships, the authors noted that patternswere not related to pesticide distributions or towater discharge and irrigation regime.

Short et al. (1991) investigated benthicinvertebrates in a Kentucky stream subject tochloride seepage from nearby oil field operations.Ephemeropterans were the group least tolerantof elevated sodium chloride levels and wereabsent in regions where salinity exceeded2000 mg chloride/L. Fish appeared to be moretolerant of these elevated salinities.

Blinn et al. (1981) investigated theseasonal dynamics of phytoplankton at threelocations along the Chevelon Creek system inArizona. The chloride concentration at site 1was 88 mg/L at baseflow compared with900–1100 mg/L at site 3, located 10.8 kmdownstream of site 1 and 3.5 km downstreamof site 2 (chloride concentration at site 2 wasnot given). The elevated chloride concentrationat site 3 was due to numerous seeps and springsfrom the canyon wall. There were markeddifferences in phytoplankton between the threesites, with species associated with differences insalinities.

Overall, these studies suggest thatcontinuous exposure to chloride concentrationsas low as 1000 mg/L (1653 mg sodiumchloride/L) for time periods as short as 1 weekcan result in changes in stream periphyton and

benthic communities. Furthermore, if theseexposures persist, community compositionremains different from lower-salinity upstreamand downstream sites. Several severely chloride-contaminated streams and ponds adjacentto roadways have been identified in theMetropolitan Toronto area and surroundingmunicipalities. The number of affected streamsand creeks is likely to increase with increasedpopulation growth in southern Ontario. Chloride-contaminated streams may also occur in othersimilar areas of Quebec and elsewhere wherethere is a dense network of highways and heavyroad salt use. Areas in which continuous (orlong-term) exposures are most likely to occurare seepages from road salt storage yards andcontaminated groundwaters that later emergeas springs. Data suggest that pulses of sodiumchloride–laden water during spring melt willhave pronounced effects on stream communities,as will the continued release of highconcentrations of sodium chloride from streambanks during summer and autumn months.Periphyton form the base of the food webin many creeks, being grazed upon byinvertebrates, which in turn serve as foragefor fish. Reduced fungal biomass resultingfrom increased sodium chloride concentrationsmay also impact invertebrates, forage fish and,ultimately, predatory fish communities.

3.3.3.3.2 Case 2: Road salt runoff and creeks,streams and rivers in less denselypopulated areas

Substantially lower chloride concentrations havebeen observed in creeks flowing through lessdensely populated areas with a smaller highwaynetwork. Moreover, at these low levels, whichmay be only 10–100 mg/L above backgroundconcentrations, many anthropogenic activitiesmay be associated with this increase. Forexample, within the Waterford River Basin(Newfoundland), chloride concentrationsaveraged 70 mg/L at the Donovans Station(located in an industrial park), comparedwith 17 mg/L in an area that had no knownindustrial activities (Arsenault et al., 1985).High concentrations at the Donovans Station were


related to industrial activities in addition to roadsalting and contamination from two road saltdepots. Unfortunately, no study was specificallydesigned to investigate chloride concentrationsimmediately following road salt application andthe following snowmelt.

Several studies conducted in theUnited States have shown similarly low chlorideconcentrations. One study shows how, in theselow-impact streams, relatively high chloridepulses may occur. Boucher (1982) conducteda study investigating the impacts of road saltrunoff on Penjajawoc Stream in Maine.Chloride concentrations upstream of acommercial development averaged 5–15 mg/Lover 1979–1982; within the development,concentrations averaged 10–50 mg/L. However,chloride levels rose during runoff events to reacha maximum concentration of 621 mg/L. Otherstudies have also associated the use of road saltswith increased chloride concentrations. Theseconcentrations have ranged between 10 and150 mg/L in small urban and rural streams(Cherkauer and Ostenso, 1976; Gosz, 1977;Hoffman et al., 1981; Smith and Kaster, 1983;Prowse, 1987; August and Graupensperger, 1989;Demers and Sage, 1990; Mattson and Godfrey,1994; Herlighy et al., 1998). Other factors mayhave affected some of the increases in averagechloride concentrations in these streams.However, where studies have been conductedto assess the effects of snowmelt on chloridelevels, a clear pulse has been observed(Prowse, 1987). Prowse (1987) and Mattson andGodfrey (1994) also noted a relationship betweenthe concentration of chloride in streams andhighway density and length of the roadway thatdrains directly into the watercourse (e.g., wherethe road is adjacent to the watercourse).

The impacts on stream and creekcommunities of relatively small increases inchloride concentrations or larger increases overa short duration (hours to days) are uncertain.

Demers (1992) investigated the effectsof elevated chloride levels on the aquaticmacroinvertebrates inhabiting four streams near

the town of Newcomb in the Adirondack regionof northern New York. All four streams werelocated along a 2-km stretch of a state highway.Chloride concentrations in the creeks weremeasured weekly, yielding an overall meanconcentration of chloride in upstream locationsof 0.61 mg/L compared with 5.23 mg/L indownstream areas. Artificial substrates wereplaced in riffle or fast-flowing sections of thestream upstream and downstream of the highway.Samplers were within 50–100 m of the road andwere left in place between April 22 and June 3,1988, at which time they were recovered tocharacterize the colonizing invertebrates.Benthic diversity was lower in downstreamthan in upstream sites. There was an increaseof Chironomidae in downstream sites, whilePerlodidae (a stonefly family) and Ephemerellidae(a mayfly family) declined downstream of thehighway. A similar increase in chironomid anddecrease in mayfly and stonefly dominance wereobserved in the Humber River (Ontario) betweenNovember, when road salt had not yet beenapplied, and the following March, after road salthad been applied (Kersey and Mackay, 1981).

Other studies have reported differencesin benthic communities upstream and downstreamof highways, but these differences were relatedto the larger differences in flow regime, sedimentsuspended load, habitat and inorganic and organiccontaminants (Molles and Gosz, 1980; Kersey andMackay, 1981; Smith and Kaster, 1983; Maltbyet al., 1995). In the absence of carefully designedfield studies, the impacts of small increases inchloride concentrations on stream, creek andriver ecosystems remain unknown. An increase inmean chloride concentration from about 3 mg/L(world average from rivers) to 30–300 mg/L mayaffect succession of species with different salinityoptima. Recent paleolimnological studies areshowing that algal species, in fact, have distinctsalinity optima, and changes in lake salinity arebeing inferred based on long-term changes inalgal communities (Dixit et al., 1999).

Some researchers are beginning toinvestigate the role that water chemistry playsin determining community structure in rivers.


For example, Rott et al. (1998) investigatedperiphyton assemblages in the Grand River,Ontario, where chloride concentrations rangedfrom 7.7 to 85.0 mg/L and conductivity from 180to 540 µS/cm. Canonical correspondence analysisshowed that the largest portion of variability inspecies composition in the river over the studyperiod could be explained by a seasonal gradientrelated to temperature and latitudinal gradientsof nitrite-nitrate, conductivity and chloride.Phosphorus, ammonia, pH, turbidity andoxygen were of lesser importance.

Chetelat et al. (1999) investigatedperiphyton biomass and community compositionin 13 rivers in southern Ontario that differed intheir nutrient concentrations and conductivity.Biomass was strongly correlated with totalphosphorus (r 2 = 0.56) and even more stronglycorrelated with conductivity (r 2 = 0.71). Theysuggested that the predominance of Cladophoraat high-conductivity sites was related to the highercalcium concentrations at this site. Conductivityranged from 65 to 190 µS/cm. Major differencesin species composition were observed betweenlow- and high-conductivity sites.

It is not clear whether periphytoncommunities were responding directly tochloride or whether chloride is an indicator ofother perturbations in the watershed. However,given newer studies involving algal communitycomposition in lakes (e.g., Dixit et al., 1999),it is probable that at least some of the responsesare directly due to small differences in saltconcentrations.

3.3.3.3.3 Case 3: Road salt runoff and urbanponds and lakes

Ponds and lakes differ from creeks, streams andrivers in three important ways. First, ponds andlake systems do not have the pronounced andseasonally varying flow velocities characteristicof streams, creeks and rivers. Second, waterresidence time in ponds and lakes is longer,ranging from weeks to years. Third, ponds andlakes receive water from a variety of sources,and so their dilution capacity for contaminants

can be higher than that for streams and creekslocated near similar contaminant sources. Thus,chloride levels tend to be substantially lowerin road salt–contaminated ponds, although theexposure time is generally much longer. Thedegree of contamination also depends on thedegree of dilution that can be afforded. Generally,larger water bodies have a greater capacity fordilution.

There is a tremendous amount ofvariation in the degree to which road salt runoffaffects the chloride concentration of urban pondsand lakes. However, in general, impacts on theCanadian environment have been modest, withchloride increases limited to the tens to hundredsof milligrams per litre rather than the thousands ofmilligrams per litre as observed for creeks in theMetropolitan Toronto area.

In Newfoundland, Arsenault et al.(1985) investigated chloride concentrationsin five ponds in the Waterford River basin.Average concentrations ranged from 6 to 24 mgchloride/L, with the higher values associated withhighway runoff and possibly runoff from a saltstorage depot.

In Halifax, the First and Second Chainlakes have shown seasonal increases in chlorideconcentration, apparently related to road salting;chloride concentrations increased through theearly 1980s to peak at 120–170 mg/L in the mid1980s. Both lakes are shallow (mean depths of 4.1and 3.2 m, respectively) and with a large urbanwatershed. More recently, Keizer et al. (1993)investigated the water quality of 51 lakes in theHalifax/Dartmouth Metro Area, comparing datacollected in April 1991 with those collectedin April 1980. The most pronounced change inthese lakes was a near doubling in conductivity,with the majority of the increase due toincreased concentrations of sodium and chloride.Whereas only 3 lakes had chloride concentrationsgreater than 100 mg/L in 1980, 12 had chlorideconcentrations exceeding 100 mg/L in 1991,with one lake reaching 197 mg/L.


In southeastern Ontario, Little RoundLake apparently became meromictic (or morestrongly meromictic) in the 1950s as a resultof road salt entering the lake from two highwaysand seepage from a salt storage shed (Smol et al.,1983). Chloride concentrations in themonimolimnion were 104 mg/L in about the early1970s. This meromixis resulted in a reduction inseasonal nutrient exchange from deep to surfacewaters, a reduction in lake productivity and achange in algal community structure.

In Ontario, an ongoing study isproviding evidence of elevated chlorideconcentrations in ponds next to highways.Chloride concentrations in ponds within 10 mof the highway average 88 mg/L (5–150 mg/L)for two-lane highways and 970 mg/L(84–2630 mg/L) for four-lane highways (Watson,2000). Chloride concentrations remained elevated50 m from the highway, with concentrationsaveraging 115 mg/L (35–368 mg/L) for two-laneand 1219 mg/L (50–3950 mg/L) for four-lanehighways. Free and Mulamoottil (1983) reportedchloride concentrations of up to 282 mg/L inLake Wabekayne, a stormwater impoundment inMississauga.

The Ontario Ministry of Transportationhas unpublished data on the mean (May–October1995) chloride concentrations in eight waterbodies in the Humber River watershed (Scanton,1999). Chloride concentrations averaged10.6 mg/L at Lake St. George, 26.8 mg/L atPreston Lake, 47.0 mg/L at Wilcox Lake,60.8 mg/L at Heart Lake, 107.9 mg/L atClaireville Reservoir, 110.3 mg/L atG. Ross Lord Dam, 174.0 mg/L at LakeAquitanne and 408.9 mg/L at Grenedier Pond.The Humber River, like Highland Creek, isa major urban catchment basin. Road saltapplications may be impacting the Humber Riverbasin, as has been observed for the HighlandCreek basin (Howard and Haynes, 1997).

More extreme examples of increasedchloride concentrations in ponds and creeksin urban areas are to be found in U.S. studies.

Cherkauer and Ostenso (1976) reported thatNorthridge Lakes in Milwaukee, Wisconsin,were affected by elevated chloride concentrationsduring winter ice cover. These lakes are shallow(9.2 m), artificial and interconnected. Chlorideconcentrations reached 2500 mg/L. Anotherextreme example is Irondequoit Bay, a smallembayment located on the southern shore ofLake Ontario. This is a moderately deep bay(maximum depth 23.8 m, mean depth 6.8 m)surrounded by a dense network of highways.Chloride concentrations began increasing in thisbay in the 1940s and continued increasing intothe early 1970s. Chloride concentrations in deepwaters reached more than 400 mg/L; verticalmixing of the water column was prevented(Bubeck et al., 1971). Remedial actions weredeveloped; by the early 1980s, chlorideconcentrations in Ides Cove, within the bay,declined to a maximum of 140–150 mg/L,and vertical mixing was improved (Bubeck et al.,1995).

In Ann Arbor, Michigan, Fonda Lakewas adversely affected by highway runoff thatentered the lake via two drainage pipes. Chlorideconcentrations reached 177 mg/L in the mid1960s, and vertical mixing was disrupted (Judd,1969). After this was learned, the city reducedits salt usage in the subdivision and redirectedrunoff. However, chloride concentrationscontinued to increase because of increased inputsfrom the greater number of highways. Chlorideconcentrations reached a maximum of 720 mg/Lin deep waters (Judd and Stegall, 1982). However,because the salt entered the lake in diffuse sourcesrather than through two drainage pipes, this saltmixed into the lake more effectively, and verticalmixing was not disrupted.

As already noted, few biological studieshave accompanied these urban pond and lakestudies. In addition to the Smol et al. (1983)study, Free and Mulamoottil (1983) noted adecrease in benthic invertebrate densities thatoccurred when bottom water became anoxic.Based on estimates of chronic toxicity, directchloride effects on the individual species may be


expected to occur at concentrations as low as200 mg/L for individual species. However, recentstudies involving phytoplankton suggest that foralgae, at least, impacts may occur at even lowerlevels. This is discussed in the next section.

3.3.3.3.4 Case 4: Highway runoff and ruralponds and lakes

Chloride increases due to road salt have also beenreported in rural ponds and lakes. As in the urbansetting, the increase is related to pond/lake sizeand the degree of contact with the roadway.

In Terra Nova National Park inNewfoundland, chloride concentrations in PineHill Pond increased nearly 7-fold in spring froman average of 14 mg/L over April–December 1969to 94 mg/L in April (Kerekes, 1974). Thisincrease was related to road salt usage.

Underwood et al. (1986) surveyed 234Nova Scotia lakes. The background chlorideconcentration was 8.1 mg/L, while the maximumconcentration was 28.1 mg/L. The Nova ScotiaDepartment of the Environment (Briggins et al.,1989) indicated that waters with chlorideconcentrations above 25 mg/L must have beenreceiving chloride inputs from anthropogenicsources. In watersheds with numerous highways,road salt would have been the probable source.

The Ministère des Transports du Québec(1980, 1999) investigated the impacts of road salton Lac à la Truite, by Highway 15 and nearSainte-Agathe-des-Monts. The lake’s drainagearea, estimated at 728 ha, was affected by a 7-kmstretch of highway, with a maximum slope of 5%.The lake had a surface area of 48.6 ha, a meandepth of 21.5 m and an estimated volume of486 000 m3. In 1972, the average chlorideconcentration for the lake was 12 mg/L. Thisincreased through the 1970s to reach a maximumconcentration of 150 mg/L in 1979. Thiscorresponds to an addition of approximately67 068 kg of salt to the lake. The amount of saltapplied to the roads was reduced, and chlorideconcentrations fell through the 1980s to reach

45 mg/L in 1990. While concentrations declined,they remained elevated at 42 mg/L at 3 m and49 mg/L at 10 m.

In the United States, as in Canada, thereis growing evidence of increasing chloride levelsin lakes, with these increases being related, atleast in part, to road salt application.

Several studies conducted in theUnited States provide further information on theprobable impact of road salts on rural lakes andponds that may be applicable to the Canadianenvironment. In Maine, Hanes et al. (1970)discussed the effects of road salt on a variety ofaquatic ecosystems. They noted that farm pondslocated near highways had sodium concentrationsranging from 1.4 to 115 mg/L and chlorideconcentrations ranging from <1 to 210 mg/L.Salt concentrations were higher in April 1966 thanin July 1965. In 1967, chloride concentrationsranged from 1.4 to 221 mg/L, suggesting thatchloride levels were increasing.

Sparkling Lake in the northernWisconsin Lake District experienced an increasein chloride concentration due to contaminationfrom road salt–laden groundwater (Bowser,1992). The morphometry of Sparkling Lake wasnot reported. Road salt was initially applied toroads above the lake, and the salt apparentlyleached into the groundwater prior to entering thelake. Chloride concentrations in unaffected lakesand groundwater in the area were 0.3–0.5 mg/Lcompared with 2.61 mg/L in 1982 and 3.68 mg/Lin 1991 in Sparkling Lake. The load of chloriderequired to produce such an increase in chlorideconcentration between 1982 and 1991 wascalculated by Bowser (1992) to be 1200 kg/year.

Eilers and Selle (1991) comparedconductivity, alkalinity, calcium and pH datacollected at 149 northern Wisconsin lakes over1925–1931 with data collected over 1973–1983.All parameters increased between the two timeperiods, with the greatest increases associatedwith increased land development on lakeshorelines. Mean conductivity increased from


31.5 µS/cm in 1925 to 44.3 µS/cm in 1983 indeveloped watersheds and from 14.3 µS/cm in1925 to 14.7 µS/cm in 1983 in undevelopedwatersheds. Increased conductivity appearedto be associated with a combination of factors,including road salt, cultural eutrophication andchanges in hydrology. The strongest increaseswere associated with lakes located near highwaysor paved roads.

The impacts of small increases in chlorideconcentration from road salt on lakes and pondswere not investigated in these studies. However,increased chloride (and other salts) concentrationstraditionally have been viewed with concern(Beeton, 1969; Pringle et al., 1981) and actionstaken to protect aquatic environments from theseincreased salt levels.

Small increases in chloride concentrationare most likely to impact species compositionand productivity. There is emerging evidencethat microorganisms such as fungi, planktonand macroinvertebrates have salinity optima.Changes in salinity that alter the competitivebalance between species should result incompositional changes at the base of the foodweb. Productivity may also increase. Evans andFrick (2001) addressed this issue by investigatingwhat is known about the effects of salinity andconductivity on freshwater systems in general.

The addition of road salts to aquaticenvironments may enhance productivity througha variety of means. Road salt contains tracenutrients, including phosphorus (14–26 mg/kg)and nitrogen (7–4200 mg/kg). The addition of100 mg chloride salt/L could result in an increasein phosphorus concentration of 1.4–2.6 µg/Land an increase in nitrogen concentration of0.7–420 µg/L. Metals such as copper and zincare also essential elements, and the additionof these metals to aquatic ecosystems may alsoenhance productivity. The systems that wouldbe most vulnerable to a road salt–stimulatedeutrophication would be low-productivityecosystems. Paleolimnological studies in theGreat Lakes have demonstrated enhanced

productivity in the Bay of Quinte as early asthe late 1600s, which was associated with landclearing and the increased inputs of nutrients (andtrace metals) from the watershed (Schelske et al.,1983, 1985; Stoermer et al., 1985).

More recent limnological studies areinvestigating community structure in a widevariety of lakes and as a function of a numberof limnological variables, such as conductivity,nutrient concentrations and major ions. Thisliterature was briefly examined to determinewhether additional information could be foundon the responses of aquatic organisms to smallgradients in salinity in addition to the expectedresponses to phosphorus concentrations. Sincethese studies were conducted in freshwater lakes,conductivity was dominated primarily by calciumcarbonates. The following are noted.

Dixit et al. (1999) assessed changesin water quality in the northeastern United Statesusing diatoms that had become depositedin lake sediments as indicators of change.The composition of diatoms in the surfacesediments was related to current water qualityparameters, including pH, chloride concentrationand total phosphorus concentration. Totalphosphorus and chloride optima were developedfor 235 of the common species (Table 15).Mathematical models were then developed topredict the pH, chloride concentration or totalphosphorus concentration of the study lakes basedon the diatom assemblage (i.e., not individualtaxa) found on surface sediments. The diatomassemblage was then examined at the 30-cmsediment depth in core samples to determinewhether or not the water quality had changedsince before the 1850s. Marked deteriorationin water quality was noted in hundreds oflakes. In these lakes, phosphorus and chlorideconcentrations clearly were higher in modernthan in preindustrial times. Moreover, therehas been a marked increase in the number ofeutrophic lakes in the Coastal Lowlands/Plateau.This increase in the trophic status of theselakes was associated not only with increasingphosphorus concentrations but also with


increasing chloride ion concentrations.Sodium and chloride concentrations werehighly correlated in these lakes, suggestingthat increased chloride concentrations wereassociated with road salting. However, it wasalso recognized that agriculture, silvicultureand urbanization can contribute to increases inchloride concentration. Future research, basedon more detailed paleolimnological studies,will resolve these issues. However, it is alreadyapparent that diatoms have optimal chlorideconcentrations, and deviations in theseconcentrations are associated with changes inspecies composition. This also must occur forother organisms, such as other algal groups,plants, benthic invertebrates and zooplankton.

Studies, while few in number,strongly suggest that small increases in chlorideconcentration will result in shifts in phytoplanktoncommunity composition in ponds and lakes.Furthermore, it is highly likely that standingstocks of plants and animals will also be affected.Productivity is likely to be enhanced throughincreased phosphorus, nitrogen and trace elementinputs.

3.3.3.3.5 Case 5: Salt storage depots andaquatic ecosystems

The actual number of road salt storage depots inCanada is unknown. Morin and Perchanok (2000)estimated that there were more than 1300 patrolyards at the provincial level and an unknownnumber of municipal and privately maintainedyards. Snodgrass and Morin (2000) also notedthat the design of these yards varies considerablyacross Canada. Some are well designed, withminimal salt leaching through the winter. Poorlydesigned yards may lose 22% or more of their saltthrough leaching. Salt concentrations in suchleachate can exceed 80 000 mg/L (see Section2.4). Ontario Ministry of Transportation (MTO,1997) studies conducted in Ontario determinedthat 26% of patrol yards had well waterconcentrations that exceeded 500 mg/L. Inanother study, 69% of shallow groundwatersamples had chloride concentrations thatexceeded 250 mg/L. While leachate eventuallywill be diluted on entering a receiving water body,considerable damage may be inflicted on theaquatic environment on the dilution path. Thisis of particular concern for wetlands and small


TABLE 15 pH, total phosphorus and chloride optima for selected diatom species in the northeasternUnited States (from Dixit et al., 1999)

Species pH Total phosphorus (µg/L) Chloride (mg/L)Achnanthes altaica 6.8 7 0.7Achnanthes clevei 8.1 7 8.7Amphora ovalis 8.0 22 4.2Amphora perpusilla 8.3 25 21.0Cyclotella meneghiniana 8.3 66 39.5Cymbella cesatii 7.8 10 0.7Fragilaria crotonensis 8.0 14 6.9Navicula bremensis 6.2 7 0.5Nitzschia linearais 7.4 8 0.8Stephanodiscus niagrae 8.1 16 10.5Synedra ulna 7.9 15 4.5Tabellaria fenestrata 7.5 13 1.8Tabellaria quadriseptata 5.5 11 1.1

streams such as investigated by Dickman andGochnauer (1978). Moreover, even after dilution,chloride concentrations may remain high enoughto create meromixis in small lakes and/or tosignificantly raise their salinity.

Poorly designed or maintained saltstorage depots clearly present a threat to theCanadian environment. Brief (hours to a fewdays) exposures to undiluted leachate are wellwithin the toxic range for a variety of aquaticorganisms. Where saline water collects forextended periods of time, as in a bog, sensitivemarsh vegetation species may be killed andreplaced by more salt-tolerant forms (Wilcox,1982). Where saline water flows into lakes, thenormal circulation patterns of the water columnmay be disrupted, and the lake becomes morestrongly stratified. Relatively small increases insalinity to as little as 200–300 mg chloride/Lfall within the chronic toxicity range for certainaquatic organisms.

Some studies have implicated road saltstorage depots as having adversely affectedchloride levels in the Canadian environment.Arsenault et al. (1985) related small increasesin chloride concentrations at the DonovansStation on the Waterford River, Newfoundland, tosalt storage depots in addition to potential impactsfrom the highway. In addition, Arp (2001)(Section 3.4.1.2) reported elevated sodiumand chloride concentrations in creek and ditchwater located downstream of a salt storage depot.Highest chloride levels occurred in summer.In southeastern Ontario, Smol et al. (1983)conducted a paleolimnological study on LittleRound Lake, noting that the lake was oligotrophicprior to European settlement in the mid 1800s.With increased logging in the watershed, thelake became eutrophic, presumably becauseof increased nutrient and mineral inputsassociated with land disturbance. A railroad andtwo highways were later built adjacent to thelake. The lake returned to oligotrophic conditionsin the late 1960s as an apparent result of the lakebecoming meromictic. The deep-water chlorideconcentration in the early 1970s was 104 mg/L,

well above background levels. This meromixisprevented nutrients regenerated in the deep layerof the lake from being mixed back into surfacewaters. The elevated chloride concentrations(and meromixis) were related to highwayrunoff and seepage from a salt storage shed.Unfortunately, very few studies have beenconducted around salt storage depots in Canada,despite the fact that there are more than athousand such facilities in Canada and the factthat many clearly are contaminating groundwaterand well water with chloride.

More limnological studies have beenconducted around road salt storage depots inthe United States. Such studies have showna variety of impacts, which indicates that theproper management of such facilities is ofenvironmental concern. Thus, the results of thesestudies have broad application to the Canadianenvironment, particularly for storage depots thatare not properly maintained (i.e., covered and onan asphalt pad) and where best practices are notfollowed with respect to patrol yard washwaterwaste.

Wilcox (1982) investigated the effectsof sodium chloride contamination by a road saltstorage pile at Pinhook Bog in Indiana. In 1963,an uncovered sodium chloride storage depot wasestablished overlooking the bog. Salt-laden runofffrom the storage pile resulted in major alterationsin the bog vegetation within a 2-ha area, asdid runoff from the highway. The salt pilewas covered in 1972; after winter 1980–81,road salt ceased to be stored at this site. Impactswere studied from 1979 to 1983. Sodiumconcentrations as high as 468 mg/L and chlorideconcentrations as high as 1215 mg/L wererecorded in interstitial waters of the bog mat inareas of the strongest road salt impact (Wilcox,1982). These readings were made in 1979, sohigher concentrations may have occurred inearlier years. Chloride concentrations at controlsites in 1980 and 1981 were 5–6 mg/L. Themaximum single daily readings for salt-impactedlocations were 1468 mg chloride/L in 1979,982 mg chloride/L in 1980, 570 mg chloride/L


in 1981 and 610 mg chloride/L in 1983. Thisindicates that decommissioning of the storagefacility resulted in reductions in loadings andresulting concentrations in the bog, but that thechloride was subsequently largely retained in thewetland.

Native species, such as Sphagnum spp.and Larix laricina, were absent from the impactedareas of the bog, while salt-tolerant species, suchas Typha angustifolia, invaded the bog. Salinitytolerances of various plant species were defined(Wilcox, 1982), and taxa were shown to besensitive to sodium chloride concentrations in bogwater as low as 280 mg/L (170 mg chloride/L)(Table 16). As salt concentrations decreasedsome 50% over 1980–1983, many endemic bogplants, including Sphagnum, recolonized the bog(Wilcox, 1982). Sphagnum began growing onlow hummocks in areas where interstitial chlorideconcentrations had dropped to approximately300 mg/L.

Tuchman et al. (1984) used core samplesfrom lake sediments to investigate the impacts ofhistoric salinization on diatoms of Fonda Lake,Michigan. A salt storage facility has been locatedadjacent to Fonda Lake since 1953; an asphalt padwas added at the salt storage facility in the early

1970s to reduce loss of salt. A reduction indiversity of algal species began in 1960. Diatomdiversity reached a minimum in 1968, when avariety of salt-tolerant (or halophilic) taxaattained their highest relative abundance. Inlater years, diversity increased slightly and somehalophilic taxa decreased in relative abundance,suggesting a decrease in salt loading to the lake.Unfortunately, lake salinity was not determined aspart of this study. Zeeb and Smol (1991) extendedthe study of Fonda Lake to scaled chrysophytes.Mallamonas caudata, a chloride-indifferent taxon,dominated at all depths in the core. Mallamonaselongata, a widely distributed taxon found morecommonly in eutrophic and alkaline waters,and Mallamonas pseudocoronata, a chloride-intolerant taxon, declined during the period whenchloride concentrations apparently were highest.Mallamonas tonsurrata, which occurs mainlyin eutrophic lakes, became more abundantduring this period. Major shifts in diatomand chrysophyte assemblages were associatedwith relatively small changes in salinity (i.e.,from about 12 to 235 mg chloride/L, or about20–387 mg sodium chloride/L).

Overall, these studies show that poorlydesigned or managed road salt storage depots canhave a number of adverse impacts on the aquatic


TABLE 16 Sodium chloride tolerance of selected plant species in the salt-impacted mat zone ofPinhook Bog, Indiana (from Wilcox, 1982)

Scientific name Common name Tolerance (mg/L)

NaCl Na ClBidens connata Purple-stemmed tickseed 1030 405 625Pyrus floribunda Purple chokeberry 1070 420 649Hypericum virginicum Marsh St. John’s wort 1070 420 649Sphagnum Bog moss 770 303 467Solidago graminifolia Grass-leafed goldenrod 760 299 461Vaccinium corymbosum Highbush blueberry 580 228 352Vaccinium atrococcum Black highbush blueberry 400 157 243Drosera intermedia Oblong leafed sundew 360 142 218Nemopanthus mucronata Mountain holly 280 110 170Larix laricina Tamarack 280 110 170

environment. This is of particular concernbecause biological monitoring programs have notbeen established around these yards to ensure thatthe local aquatic and terrestrial environments arenot being damaged.

3.3.3.3.6 Case 6: Snow dumps

Snow collected from roadways to whichroad salts have been applied has been reported tocontain chloride at concentrations reaching up to2000–10 000 mg/L. Chloride concentrations inthis range can be acutely toxic to a variety oforganisms for exposures of 1 day or less. Themajority of municipalities that remove roadsidesnow use surface dumps (Delisle and Dériger,2000). The actual location of snow dumps issubject to significant variation and variableenvironmental regulation. The number of snowdumps is also unknown.

Snow deposited on low-lying landscan result in the increased chloride content ofunderlying soils and wetlands when the snowmelts. Similarly, snow deposited on elevatedareas eventually melts, transporting salt into thegroundwater and into aquatic ecosystems, such asstreams, wetlands and ponds.

Few studies have been conductedaround snow dumps. However, elevated chlorideconcentrations (233–994 mg/L) have beenreported in monitoring wells established toassess the impact of snow disposal on shallowgroundwater (Morin, 2000). This contaminatedgroundwater could, in turn, contaminategroundwater-fed springs, as observed by Williamset al. (1999).

In British Columbia, Warrington andPhelan (1998) reported that the ionic compositionof two lakes was shifted from carbonatedominated to chloride dominated as a result ofsalt-laden snow being pushed into the lakes fromthe highway. Salt-laden highway runoff may alsohave been a contributing factor. The biologicaleffects were not investigated, but some could beinferred from recent paleolimnological studies of

the effects of increased chloride concentrations onphytoplankton communities.

No other studies were located on theadverse impacts of surface snow dumps onaquatic ecosystems. However, given the probablethousands of snow dumps across Canada andthe high chloride concentration in meltwaters,the potential harm to the Canadian aquaticenvironment of poorly located dumps is ofconcern. This is particularly so because there arenot uniform regulations regarding snow dumppractices and very few monitoring programs inplace to ensure that the environment is not beingharmed by these chloride releases. Unsuitablelocations for snow dumps include wetlands, smallponds and lakes, and elevated areas that drain intosmall creeks. Groundwater contamination alsoneeds to be considered.

3.3.3.4 Other effects

Road salts can have adverse effects on the aquaticenvironment in addition to their apparent directeffects on species composition and abundance.Chloride salts can affect aquatic systems throughinteractions with abiotic components of theenvironment. Notably, chloride salts tend to bemore soluble than carbonate salts and can thus,through various reactions, enhance the mobility oftrace metals in aquatic ecosystems. Increased saltloadings can affect water density, therebyaffecting mixing processes in lakes and thusdisrupting the ecological functioning of thatecosystem.

Increased concentrations of chloridesalts in surface water systems can lead to therelease of metals from sediments and suspendedparticulate matter. By competing for particulatebinding sites, sodium chloride acts as an enhancerof dissolved and potentially bioavailable tracemetals, such as cadmium, copper and zinc(Warren and Zimmerman, 1994). Cadmium,copper and zinc are acutely toxic to aquaticorganisms at concentrations as low as 1 µgcadmium/L (rainbow trout), 6.5 µg copper/L(Daphnia magna) and 90 µg zinc/L (rainbow


trout) (CCME, 1991). In addition, since road saltcontains a variety of contaminants, includingmetals such as copper (0–14 µg/g), zinc(0.02–0.68 µg/g) and cadmium (MDOT, 1993;Maltby et al., 1995), snowmelt containing roadsalt is a potential source of metal contaminationto receiving water bodies. Furthermore, becausethe highway surface contains a variety of metalcontaminants from the automobile itself(lead, copper, zinc and cadmium), road salt canfacilitate the mobilization and transport of thesecontaminants into the aquatic ecosystem (Maltbyet al., 1995). These metals may accumulatein various reaches of streams on sedimentaryparticles (Maltby et al., 1995) or in stationarywaters such as ponds. There, they have thepotential to exert toxic effects, especially throughprocesses involving sediments.

High concentrations of salts insediment pore water were shown to augmentthe concentrations of dissolved heavy metals(e.g., cadmium) and resulting toxicity to a benthicinvertebrate, Hyalella azteca (Mayer et al., 1999).Wang et al. (1991) found that 709 mg chloride/L(or 0.02 mol/L) substantially enhanced the releaseof mercury from freshwater sediments. Mercurycan be acutely toxic to invertebrate species andfish in concentrations as low as 0.002 mg/L and0.02 mg/L, respectively (CCME, 1991) and, whenpresent in its methylated form, accumulates inthe tissues of fish. Sodium chloride also enhancesmercury mobilization from soils (MacLeod et al.,1996).

Road salt can also affect ponds and lakesby reducing their capacity for vertical mixing.Most lakes undergo vertical mixing, resulting inan exchange of deep and surface waters. Suchexchanges are important for transferring oxygen-rich surface water to the deeper regions of thelake. In the absence of such exchanges, deepwaters can become anoxic. Vertical exchangesare also important in transferring nutrientsregenerated in the deeper regions of the lake tothe surface, where they become available to thephytoplankton community. Vertical mixing isdriven by the winds, which mix surface waters

down to depths dependent on lake fetch,surrounding topography and water temperature.Vertical mixing is also driven by changes in watertemperature that affect the density of water. Sincewater has a maximum density at 4°C, springwarming from freezing to 4°C is accompaniedby an increase in water density, promoting thevertical exchange of warmer waters with colderand less dense deep waters. Lake cooling inautumn and lake warming to 4°C in spring areaccompanied by a change in water density,which promotes the vertical exchange of waters.

Meromixis occurs in lakes whenconditions develop that prevent the verticalexchange of surface and deep waters (Wetzel,1983). This occurs when a sufficient gradientexists in salinity to override the effects ofseasonal variations in water temperature ondensity and lake mixing. Road salts, by enteringlakes through surface flow (overland runoff,ditches, streams) or groundwater discharge (seeps,springs), have the potential to impair the normalcirculation of lakes. Small, moderately deep lakeswill be the most vulnerable to such meromixis,especially in areas of heavy road salt application.Larger lakes are less vulnerable, because theintruding saltwater plumes experience greaterdilution as the denser salt-laden water flows alongthe lake floor towards the deeper regions of thelake. In addition, larger lakes have greater fetchesand hence more powerful wind-driven currentsand other water exchanges.

The formation of meromixis can havea number of impacts on lakes. The low oxygenconditions that develop below the chemocline canresult in the loss of all but the most resilient ofdeep-water benthic species. Zooplankton may beexcluded from their deep-water daytime refuges,being forced to live in the brighter surface layers,where they may be more vulnerable to size-selective fish predation. Hypolimnetic fish suchas lake trout may also be adversely affected.The onset of lake meromixis will affectsediment–water exchanges. Phosphorus andvarious metals are more readily released frompoorly oxygenated than from well-oxygenated


sediments (Wetzel, 1983). This increase inphosphorus release from the sediments mayenhance the productivity and eutrophication of thelake, particularly if there is sufficient exchange atthe chemocline. Alternatively, if there is limitedexchange at the chemocline, production maybe reduced, and a eutrophic lake may becomeoligotrophic (Smol et al., 1983).

3.3.3.5 Conclusions for aquatic ecosystems

Short-term acute toxic effects have beenassociated in the laboratory with high chlorideconcentrations. However, as exposure timeincreases, sensitivity to chloride increases.For example, the 4-day LC50 is 1400 mg/Lfor Ceriodaphnia dubia (a cladoceran) and1261 mg/L at 9 days (Cowgill and Milazzo,1990). Exposure to such high lethalconcentrations would most likely be associatedwith poorly managed salt storage depots,inappropriately placed snow dumps on wetlandareas, roadside ditches in areas of high use,contaminated groundwater springs and smallwatercourses in heavily populated urban areas(e.g., Metropolitan Toronto) that have a densenetwork of highways and road salt application.

Longer-term toxicity occurs atsubstantially lower levels, estimated to beginat concentrations as low as 210 mg/L. Chlorideconcentrations at these levels have been observedin a variety of urban creeks, streams and lakes,although primarily in highly populated areas andsmall water bodies near highways with high use.Such levels are also expected to occur in thevicinity of poorly managed road salt storagedepots and in inappropriately placed snow dumps,e.g., on wetlands, in small ponds, near headwatercreeks and near contaminated groundwater-fedsprings. Aquatic ecosystems experiencing suchchloride levels are expected to be impaired.

A family of curves was prepared basedon chloride toxicity data following 4-day, 7-dayand predicted long-term (chronic) exposures(Figure 17). The toxicity data used to developFigure 17 can be found in Evans and Frick(2001). One-day and less than 1-day data are

excluded because studies were few in number.Data are plotted on a linear-log scale with 95%confidence intervals around each curve. A log-logistic model was used in the curve fitting(Environment Canada, 1997b). Based on thepredicted data, it can be expected that 5% ofspecies would be affected (median lethalconcentration) at chloride concentrations of about210 mg/L (Table 17), while 10% of species wouldbe affected at chloride concentrations of about240 mg/L.

Figures 18 and 19 show variouschloride concentrations in the Canadian aquaticenvironment juxtaposed with concentrationscausing adverse biological effects. Theenvironmental chloride concentrations specifiedare assumed to arise from road salt contamination.Unless otherwise specified, all values for effectsrelate to lethality.

Figure 18 presents data on short-termexposure pertinent to a period of 5 days or less.The concentrations given are for rivers and creekswhere chloride peaks associated with events suchas freeze/thaw occurrences would not be expectedto last for more than a few days at any givensampling location. The biological effects specifiedare for toxicity tests with an endpoint of 5 days orless (e.g., 4-day LC50).

Figure 19 presents data pertinent toexposure for a period of more than 5 days.The environmental concentrations given aremostly for lakes and ponds where chloride levelsare generally expected to remain stable over alonger period. The biological effects specified arefor toxicity tests with an endpoint of more than5 days (e.g., 33-day LC50).

As previously noted, there is growingevidence of a widespread increase in chloridelevels in rivers and lakes in the northeastern andmidwestern United States, including lakes in anon-urban setting. For example, Dixit et al.(1999) investigated 257 lakes and reservoirs in thenortheastern United States and noted that while34% of the reservoirs and 6% of the lakes hadbackground chloride concentrations of >8 mg/L


prior to the 1850s, this had increased to 46%and 18%, respectively, by present times. Suchincreases are related, in part, to the number ofhighways in the area, land disturbance and roadsalt use. Furthermore, researchers are beginningto note changes in algal communities as thesechloride levels increase from background levelsof 1.8–3.6 mg/L to over 7 mg/L (Dixit et al.,1999). Studies conducted in Canada, althoughless detailed, are also pointing towards increasedchloride levels, in the range at which changesin algal composition may be expected, in smallponds and lakes located near certain highwaysand in a variety of streams and lakes in an urbansetting. However, it is important to note that thevast majority of these studies are limited to areasof heavy road salt usage, primarily southernOntario, the Maritimes and Quebec.

The U.S. EPA (1988) developed waterquality guidelines for chloride and concluded that,except possibly where a locally important speciesis very sensitive, freshwater organisms and theiruses should not be affected unacceptably if:

• the 4-day average concentration of chloride,when associated with sodium, does notexceed 230 mg/L more than once every 3years on average;

• the 1-hour average chloride concentrationdoes not exceed 860 mg/L more than onceevery 3 years on average.

They noted that these criteria will not beadequately protective when the chlorideis associated with potassium, calcium ormagnesium, and that because animals have anarrow range of acute sensitivities to chloride,excursions above this range might affect asubstantial number of species. Evans and Frick(2001) also found evidence that potassiumchloride and magnesium chloride salts were moretoxic than sodium chloride. In addition, fishappeared to be less sensitive to calcium chloridethan to sodium chloride, and the converse wastrue for invertebrates, although the data sets weresmall. Although the focus of the assessment wasprimarily on sodium chloride, they concluded thatpotassium chloride and magnesium chloride saltsappear to be as toxic as or even more toxic thansodium chloride.

To conclude, it is considered that highconcentrations of chloride associated with roadsalts may have immediate or long-term harmfuleffects on surface water systems, based onseveral exceedences of effect thresholds in theenvironment and on field evidence of ecosystem


TABLE 17 Predicted cumulative percentage of species affected by chronic exposures to chloride(from Evans and Frick, 2001) 1

Cumulative % of Mean chloride Lower confidence Upper confidence species affected concentration (mg/L) limit (mg/L) limit (mg/L)

5 212.6 135.9 289.510 237.9 162.3 313.625 328.7 260.2 397.250 563.2 504.8 621.775 963.7 882.3 1045.190 1341.1 1253.8 1428.4

1 Data are from Figure 17.


FIGURE 17 Experimental acute toxicity and predicted chronic toxicity for aquatic taxa (from Evansand Frick, 2001)

FIGURE 18 Representative short-term chloride concentrations in the Canadian aquatic environmentassociated with contamination by road salts and concentrations causing adverse biologicaleffects following brief exposures

population effects. These effects are most likelyto occur at improperly managed road salt storagedepots, at inappropriately placed snow dumps andin small watercourses along a dense network ofhighways. At lower concentrations, increasedchloride concentrations may affect communitystructure, diversity and productivity. Elevatedchloride levels have been shown to increasemetal bioavailability; by affecting densitygradients in lakes, road salts may have a majorimpact on lake ecosystems, notably in terms ofdepth-dependent availability of oxygen andnutrients. Most instances in which this occursare in areas of high road salt usage, primarilysouthern Ontario, Quebec and the Maritimes.

3.4 Soils

3.4.1 Soil salinization resulting from roadsalt application

This section summarizes the informationpresented in the report prepared by Morin et al.(2000) on the extent of soil salinization due toroad salt application. The report presented soilmaps for all provinces in Canada. This mappingeffort is complemented with a few case studiesdesigned to show how the salinity of soilsolutions changes dynamically in salt-affectedareas, including short distances from point andline salt sources, and with season. The salt ionsexamined are sodium, calcium, sulphate andchloride. Contributions due to other ions such as


FIGURE 19 Representative long-term chloride concentrations in the Canadian aquatic environmentassociated with contamination by road salts and concentrations causing adverse biologicaleffects following prolonged exposures

potassium and magnesium are neglected, becausethese generally contribute little to overall soilsalinity in most locations.

Mapping of road salt inputs (annualinputs) and salt concentrations in soils was doneas spatial averages compiled for Level 2watersheds, as spatial averages at the municipalitylevel and as spatial averages at the road level(mainly principal roads and expressways). Thesethree levels of analysis were done because of thegeneral connection between soil salinity andsurface water salinity, i.e., most surface watersreceive direct input from soils as a result ofrunoff, interflow, soil percolation and baseflow.Similarly, downslope soils often receive seepagewater from higher surface water locations.Gradual dissolution of easily weathered soiland rock minerals further contributes todownslope soil salinization, as do agricultural andhorticultural fertilizer applications and seepagefrom domestic and industrial sewage. Evaporationof water from dry soils in arid regions such as thePrairie provinces also contributes to soil salinity,which, in the extreme, leads to above- and below-ground salt accumulations, especially in landscapedepressions and on hillslope seeps. Industrialactivities, especially those associated with saltmining and with the sodium hydroxide–basedextraction of fossil fuels from tar and oil sands,add further to environmental salt loading in areasspecific to these activities. This section, however,focuses on the extent to which road saltapplications in combination with atmosphericdeposition contribute to the overall salinity insoils next to roads, and how these applicationsmay affect the general background salinity ofwater as it reemerges from the soil in variousdownslope locations. In general, calculatedaverage Level 2 soil salt concentrations fromroad salt loadings are expected to be lower thanarea-wide municipal averages, and area-widemunicipal averages are expected to be lowerthan roadside soil averages. Hence, a three-waymapping effort was done to estimate:

• expected average salt concentrations in soilsolutions based on total road and atmosphericsalt inputs per Level 2 watershed;

• expected average salt concentrations in soilsolutions based on total road and atmosphericsalt inputs per municipal area; and

• expected average salt concentrations in soilsolutions based on total road and atmosphericsalt inputs in immediate roadside vicinities.

To quantify these averages, informationcontained in various national, provincial andmunicipal databases was compiled to determine:

• local average annual road salt applications,by municipality and county;

• local average annual atmospheric depositionregarding precipitation and wet sodium,calcium, chloride and sulphate ion deposition(a dry deposition database does not yet exist;including dry deposition further adds tooverall salt loadings, especially in the Atlanticregion on account of sea spray);

• average annual runoff; and• local soil depth, clay content and organic

matter content.

For the mapping effort, seasonalaspects are not considered, but soil salinitieswould vary depending on the volume of road saltloadings, weather and topography. Some of theseasonal aspects are addressed in this section byway of special case studies that deal with soil andsurface water salinity in downslope locationsfrom a salt depot.

3.4.1.1 Definitions and equations

Compiled data on atmospheric sodium, calcium,chloride and sulphate ion loadings refer tovolume-weighted ion concentrations (wetdeposition only), either in mg/L or in meq/L.Salt loadings, in contrast, are generally expressedin terms of g/m2 per year or kg/ha per year. Inthis report, atmospheric and road salt loadings areconverted into total dissolved salt concentrationsin soil solutions by dividing the calculated


combined annual salt loadings by the expectedannual soil percolation rate, where:

For simplicity, annual soil percolation rates(mm/year) are estimated to be equivalent tostream discharge rates (mm/year), where streamdischarge rates are obtained from watershed-specific hydrometric discharge observations. Forthe mapping effort, expected soil percolation ratesare therefore set equal to stream discharge rates.Maps located in Morin et al. (2000) depict:

• millimetres of stream water discharge at anational scale, as obtained from NaturalResources Canada; and

• cubic metres of stream water discharge perLevel 2 watershed (as obtained from thenational stream discharge map).

Watersheds with highest runoffcoefficients occur along the Pacific and theAtlantic coasts. Watersheds located in thesouthern part of the Prairie provinces have thelowest quantity of runoff to dissolve and dilutethe salts. Areas with low rates of runoff areimportant to identify because these areas havehigh road salt concentrations in runoff.

For the mapping of these parameters, itwas also essential to obtain national soil surveyinformation on:

• soil clay content (as a % of mineral portionof soil) (particulate matter <2 mm);

• soil organic matter (in %); and• cation exchange capacity (in meq/100 g oven-

dry soil).

The soil solution estimates for totaldissolved salt loadings and related ionconcentrations were used to calculate:

• total cations and total anions (in meq/L);• soil electrical conductivity (in mmho/cm or

mS/cm);

• osmotic potential (in bar);• sodium absorption ratio (dimensionless); • exchangeable sodium ratio (dimensionless); • exchangeable sodium percentage (in %);• exchangeable sodium content (in meq/100 g

oven-dry soil);• soil clay and silt dispersion (as % of soil clay

content); and • relative changes in soil hydraulic conductivity

(dimensionless).

Likely salt-induced effects on soilsinclude substantial lowering of soil osmoticpotentials, increased soil swelling, reduced soilstability (loss of soil structure), decreased soilpermeability, increased potential for soil erosion,increased clay and silt dispersion, and increasedturbidity in surface waters. All of these impactsget worse with increasing sodium content in thesoil. This is because increased sodium ioncontents decrease the affinity that soil particleshave for each other on account of the strongaffinity between sodium ions and watermolecules. To some extent, salt-induced effectsdepend strongly on soil clay mineralogy (e.g., soilswelling and shrinking in soils containing largeportions of montmorillonite). In some otherrespects, salt-induced effects are independentof soil mineralogy and are simply affected byelectrolyte-mediated affinity among individualparticles in the soil or in suspension (e.g., affinityis regulated by interparticulate electrolyte forcesthat determine the degree of particle coagulationand dispersion; see Ali et al., 1987). Withincreasing soil dispersion (e.g., when salinizedsoils receive low electrolyte irrigation water andare then subjected to water and wind erosion),not only does dispersion include fine soilparticles, but these particles may also be carriersfor environmental contaminants such as nutrients,heavy metals and microbiota by way of water-and wind-induced soil erosion.


soil percolation = precipitation – actual evapotranspiration – surface runoff rate (mm/year) rate (mm/year) rate (mm/year) (mm/year)

Soils that have a relatively high claycontent and continuously or periodically receivehigh sodium inputs will likely show the strongestimpacts in the long run (i.e., should receive a highsalt hazard rating). To map areas with a high salthazard rating, a simple salt hazard index wasobtained by setting:

salt hazard index = exchangeable sodium ratio ×soil clay content (in %)

In this way, areas with highest hazardratings were located for southern Ontario nearthe Greater Toronto Area and in southern Quebecaround the Island of Montréal (Figure 20).Other urban areas in southern Ontario andsouthern Quebec also have high ratings.Additional areas are found in the southernportions of the Prairie provinces, where soilsalinization on account of natural processes is amajor regional concern (see Eilers et al., 1995).

3.4.1.2 Case studies

Three case studies were conducted at KejimkujikNational Park, Nova Scotia, along select streetsin Fredericton, New Brunswick, and in a forestedarea next to Fredericton’s municipal salt depot(University of New Brunswick woodlot). Thesestudies focused on:

• providing numbers for the general turnoverrate of sodium ions in the soils of headwaterbasins at Kejimkujik National Park (anenvironment that is essentially free of roadsalts but is influenced by sea spray);

• assessing the status of soil salinity underneathlawns next to the curb of select roads inFredericton during the fall season;

• examining how the salinity of surface waterchanges with distance from Fredericton’s saltdepot, by season; and

• examining the salinity of ditch water alongthe Fredericton–New Maryland Highway.

The results of these case studies indicated that theability of soils and watersheds in the KejimkujikNational Park to retain incoming sodium andchloride ions is very low, i.e., sodium and

chloride ions are readily flushed through the soilsand through the basins of the park. Ionconcentrations are highest in soils and streamsduring early fall, and lowest concentrations occurat the time of snowmelt.

Along the city streets of Fredericton,generally little sodium remains on the ionexchange sites in soils next to the road, asobserved in early fall. The data indicate that ionexchangeable sodium is particularly low alongthe most frequently travelled roads at this timeof year. Furthermore, exchangeable sodiumconcentrations are lower next to curbs than 2–3 maway from curbs. This difference indicates thatrepeated and substantial traffic splashing and roadrunoff during spring, summer and fall are flushingsodium out of the roadside soils.

Compared with Kejimkujik NationalPark, background sodium and chloride ionconcentrations in soil solutions and in surfacewaters are somewhat smaller at the Universityof New Brunswick’s woodlot (2 mg/L versus3 mg/L, respectively). Chloride concentrationsin ditch and stream water around the Universityof New Brunswick’s woodlot, however, varyby location. Arp (2001) measured theelectrical conductivity (later related to chlorideconcentrations) of ditch water along theFredericton–New Maryland Highway and BishopDrive for representative days in June, July,October (1999) and January (2000). Results aresummarized in Figure 21 by way of box plots,according to the hydrological condition at eachpoint of measurement. Ditches in this analysiswere categorized according to the following threelocation characteristics:

• well-drained locations where the ditches tendto run dry (crest locations);

• poorly drained areas where water accumulatesand becomes stagnant; and

• areas of continuous water flow, i.e., brookroad crossings (culverts).

Figure 21 shows that electricalconductivities of ditch water peak in July,especially in areas where water is not stagnant.


At stagnant locations, electrical conductivityreadings remain fairly high regardless of seasonand tend to be highest in June and July, but arelowest and generally least variable in the fall.Where streams, brooks and/or ditch water crossunderneath the highways (by way of culverts),conductivities are lowest, at or near backgroundlevel, due to continuous flushing — i.e., ditchwater, as it enters streams and brooks, is quicklydiluted, on a year-round basis. However, whereditch water drains into poorly drained, water-accumulating depressions and where the waterturnover rate is low, dilution does not occur,and electrical conductivity levels (and, therefore,salinity levels) remain elevated year-round.

Chloride concentrations are highest atand near the salt depot, ranging from 300 to morethan 1000 mg/L. Listed in Table 18 are chlorideconcentrations for stream and pond water withincreasing stream catchment area of the CorbettBrook downstream from Fredericton’s salt depotat various times of the year (May, June, July,October, January). At first, considerable dilutionoccurs in the nearby “Larch Swale Pond”(0 km in Table 18), which receives water fromthe 172-ha headwater system. Below this area,dilution continues either gradually or in jumps,with each jump being specific to the catchmentarea of each of the adjoining brooks. For example,at about 2 km away from the depot, CorbettBrook is joined by the O’Leary Brook. At that


FIGURE 20 Areas with low to medium to high road salt application hazards, based on product of soilclay content with exchangeable sodium ratio (from Morin et al., 2000)

point, dilution increases sharply, but in proportionwith the combined catchment area for O’LearyBrook and Corbett Brook.

Season has a strong effect on electricalconductivity measurements and the related streamdilution factor. In principle, conductivitymeasurements are related to stream dischargerates: high discharge rates imply increaseddilution, and therefore decreased conductivityvalues; low discharge rates imply less dilution,and therefore increased conductivity values.For intermittent stream flow conditions (as mayoccur in summer and early fall), the overallinfluence of the salt depot appears to be confinedto the vicinity of the salt depot, while areasfurther removed from the salt depot receive lesssalt-enriched water. This “apparent” dilutionwith increasing catchment area at lower streamlocations is greater during times of intermittentflow (such as July) than during times ofcontinuous flow. In July, concentrations greaterthan 250 mg/L were observed up to 500 m fromthe depot and were still 100 mg/L at 1250 mfrom the depot.

3.4.1.3 Discussion

Road salt application effects on physical andchemical soil parameters such as osmoticpotential, electrical conductivity, soilpermeability, soil dispersion and exchangeablesodium ratio (see Morin et al., 2000, for detailedinformation on other physical and chemicalparameters) are all calculated to be highest inareas next to salt depots and roads receiving saltinputs. At the municipal level, impacts of soilwaters on downslope positions can be expected tovary greatly depending on the density of the roadnetwork, on local conditions, on the actuallocation of road salt applications and on thephysical condition of salt storage facilities. Forexample, in highly paved urban areas, road saltapplications are expected to produce higheroverall salt concentrations in soils and surfacewaters than in less developed areas. Williams etal. (1999), for example, showed that saltconcentrations in springs within the MetropolitanToronto area increase as predominant land usechanges from rural to urban. These authors alsoshowed that:


FIGURE 21 Box plot showing 10th, 25th, 75th, and 90th percentile plus individual data above andbelow the 10th and 90th percentile for ditch water electrical conductivity and chlorideconcentrations along two highways near Fredericton’s salt depot, outside the depot’scatchment area for representative days in June, July, October 1999, and January 2000.“Well-drained”, “cross-flow”, and “stagnant” refer to three conditions: ridges fromwhich water drains, stream/road crossings, and ditch depressions with no visible outflow,respectively (from Arp, 2001).

• these salt concentrations peak during the timeof salt application;

• some of these peaks have chloride ionconcentrations in excess of 1000 mg/L; and

• secondary salt concentration peaks occur inmidsummer (thereby supporting the results ofone of the above case studies), with highestchloride ion concentrations in excess of200 mg/L.

Salt concentrations in surface water,groundwater and soil solutions, therefore, maybe high at the time of road salt application, butelevated concentrations may recur later in summer

during periods of drought when water evaporatesand when salty water from, for example, subsoilsseeps back into downslope locations. Duringtimes of high precipitation, low electricalconductivity water will help to flush brinywater from these areas. Along roadsides,frequent splashing during rain events mayfurther accelerate the displacement of briny waterand the return to low soil exchangeable sodiumpercentage values, but this can occur effectivelyonly in soils of high permeability in areas wherewater does not accumulate (i.e., does not stagnatefor significant parts of the year).


TABLE 18 Modelled electrical conductivity values and chloride concentrations for Corbett Brook,downstream from Fredericton’s salt depot

Catchment area (ha) Distance Electrical conductivity (mS) Chloride concentrations (mg/L)from May June July Oct. Jan. May June July Oct. Jan.depot (km)

172 0.00 489 580 2109 191 629 150 179 667 58 195213 0.25 395 459 1270 168 514 121 141 398 51 158255 0.50 332 378 837 152 435 101 116 260 46 134296 0.75 286 321 587 139 378 87 98 181 42 116337 1.00 251 279 431 128 334 76 85 133 39 102378 1.25 224 246 328 120 300 68 75 100 36 92420 1.50 202 219 257 113 272 61 67 78 34 83461 1.75 184 198 206 107 249 56 60 62 32 76502 2.00 169 180 168 102 230 51 55 51 30 70543 2.25 157 166 140 97 213 47 50 42 29 65585 2.50 146 153 118 93 199 44 46 35 28 60626 2.75 136 142 100 89 187 41 43 30 27 57667 3.00 128 132 86 86 176 38 40 26 26 53708 3.25 121 124 75 83 166 36 37 22 25 50750 3.50 114 117 65 80 158 34 35 19 24 48791 3.75 108 110 58 78 150 32 33 17 23 45832 4.00 103 104 51 75 143 31 31 15 22 43873 4.25 98 99 46 73 137 29 30 13 22 41915 4.50 94 94 41 71 131 28 28 12 21 39956 4.75 90 89 37 69 125 27 27 11 21 38997 5.00 86 85 33 68 121 26 25 10 20 36

1038 5.25 83 82 30 66 116 25 24 9 20 351080 5.50 79 78 28 65 112 24 23 8 19 341121 5.75 77 75 25 63 108 23 22 7 19 321162 6.00 74 72 23 62 104 22 21 7 18 31

Underneath paved surfaces (as in urbanareas), soils may not benefit from seasonal saltflushing. Salt in these locations likely may remainconcentrated or may accumulate even morewith additional salt applications, because theselocations are, at least in part, outside the paths ofpercolating soil water. These locations, therefore,become a source of soil and surface water salinity,by way of either slow seepage or high-salinityflushes during times of pavement repair orunusually wet weather.

The overall salt impacts on soils needto be evaluated in terms of landscape-levelsource/sink positioning. For example, the locationof the roadside in relation to the road (above road,below road, same level as road, upwind fromroad, downwind from road) and terrain conditionsalong roads (e.g., open versus forested, well-drained versus poorly drained) have to beconsidered. Typically, salts are spread from roadto roadside by vehicle spray, dust and runoff.Accordingly, soil salinity is expected to increasefrom ridge tops (recharge areas) to poorly draineddepressions.

The process of salt dilution throughsubsequent precipitation and traffic splashingis important in terms of allowing much of theCanadian soil roadscape to recover partially orcompletely from seasonal road salt applications.Apparently, splashing from roadsides acceleratesthe leaching loss of winter-accumulated saltfrom the roadsides, as noted in one of the abovecase studies. However, during times of highexchangeable sodium percentage values and lowelectrical conductivity splash water, this effectcould have deleterious effects on soil aggregation,e.g., soil clay and silt dispersion, which, in turn,can lead to reduced soil infiltration rates andthereby enhanced soil erosion by water andenhanced dispersion of soil particles by wind.This effect would be highest where roadsides arenot covered by vegetation.

Figure 22 shows expected average valuesfor total dissolved solids, osmotic potential andelectrical conductivity of soil solutions andsurface water at the Level 2 watershed level and

at the municipal area level. An inspection of thetotal dissolved solids results indicates that thesevalues are all higher than the “background” levelsthat would be expected due to local atmosphericdeposition and subsequent evapotranspirationalone. These area-averaged concentrations,however, are far from being of concern, exceptfor highly urbanized areas. As expected,calculated values for total dissolved solids,osmotic potential and electrical conductivityincrease with decreasing area averaging.

Values for total dissolved solids, osmoticpotential and electrical conductivity for soilsolutions and surface water for roadway rights-of-way were calculated. These values are based onprovincial road salt loading data, by maintenancedistrict. Overall, these calculations indicate thatareas immediately surrounding the right-of-waytend to have high annual averages for totaldissolved solids, osmotic potential and electricalconductivity. Seasonally, these values would behigher through periods of salt application andsubstantially lower during periods of highflushing. High peaks, however, can be expectedto return during midsummer in areas of wateraccumulation and may remain high in stagnantwater pools or depressions with limited surfacerunoff, as discussed above.

The map in Figure 23 suggests thatcertain areas in Ontario and Quebec may haveroads with elevated exchangeable sodiumpercentage values. These values are primarilyassociated with areas that have high road saltloadings. Figure 23 suggests that certain areasin the Prairie provinces may also have highexchangeable sodium percentage values. Whileroads in the Prairie provinces have low road saltloadings compared with most municipal areasin Ontario and Quebec, there is less precipitationto dissolve and dilute the salts. The situation issuch that many soils in southern Alberta andsouthern Saskatchewan and some soils in southernManitoba are already saline due to natural causes,especially in depressions and near wetlands anddrainage courses. Here, road salts would add tothe total electrolyte loads of soils, as these wouldsubsequently accumulate in the depressed areas


next to the roads. In general, the spatial patternsof the background levels of electrolyte ions inthe soil solutions based on atmospheric depositionalone match the spatial patterns of soil salinityas displayed by Eilers et al. (1995) fairly well,thereby confirming the simple mass balanceprotocol of the model that produced the mapsof this report.

Increased soil salinization throughroad salt applications may affect soil biota(macro- and microflora and macro- andmicrofauna). For a simplified summary ofexpected plant sensitivities and related species-specific threshold tolerances to falling soil

osmotic potentials, refer to Figure 24. Figure 24associates the sensitivity of crops to soil salinityby demonstrating how crop yield decreases withincreasing levels of electrical conductivity. Thedashed boxes in this figure show the percentageof provincial roadways with certain levels ofestimated soil salinity. Data in this graph suggestthat salinity in soil solution alongside the majorityof provincial roads may adversely affect manyshrub and tree species. This graph also suggeststhat certain roadsides may have levels of soilsalinity that can adversely affect salt-tolerantplants like wheatgrass. Impacts on terrestrialvegetation are discussed more comprehensivelyin Section 3.5.


FIGURE 22 Estimates for total dissolved solids (TDS) and electrical conductivity (EC) of average soil solutionand surface waters, by level 2 watersheds and by municipal boundary, Ontario and Quebec (fromMorin et al., 2000)

Road salt effects on soils are expected tobe highly dependent on the extent of vegetativecover. Soils that are bare likely suffer the worstimpacts in terms of surface crusting, surfacerunoff, and clay and silt dispersal by water andwind. Soils that are originally covered byvegetation will also suffer impacts once thesurface vegetation reacts negatively to highnegative osmotic potentials in the soil.Here, gradual loss of surface vegetation will leadto increased mineral soil exposure; this leads togradually decreasing levels of soil organic matterand, therefore, increasing soil dispersion, and thusdecreasing soil permeability and increasing soilerosion. Proper vegetation management, salt

application and road maintenance practices wouldobviously go a long way to reducing any sucheffects. Additional details on soil salinity andmany other aspects related to soil managementand soil conservation can be found in, forexample, Anon. (1992), Holms and Henry (1982),Steppuhn and Curtin (1992) and Acton andGregorich (1995).

Other impacts of road salts not coveredin this section deal with salt-induced release ofheavy metals (e.g., zinc, copper, cobalt, mercury,cadmium) and base cations (calcium, magnesium,potassium). All of these may have positive ornegative effects on soil flora and fauna, including:


FIGURE 23 Exchangeable sodium percentage (ESP) along roadsides, by provincial road maintenancedistrict (from Morin et al., 2000)

• encouraging the growth of certain species inotherwise nutrient-poor ecosystems;

• enhancing the loss of plant nutrients such ascalcium, magnesium and potassium due to ionexchange with sodium ions; and

• heavy metal mobilization in soil and aquatichabitats, depending on local circumstances.

Heavy metal mobilization occurs throughformation of water-soluble chloride–metal cationcomplexes, thereby rendering many heavy metalcations more phyto-available. This would increasethe entrance of heavy metals into the local foodchain by way of plant uptake. The plantsthemselves could experience heavy metal toxicitysymptoms in the worst-case scenarios. For anexample of the processes involved, see Smoldersand McLaughlin (1996) regarding the case ofchloride-enhanced cadmium mobilization andsubsequent phyto-uptake. For additional detailsregarding soil salinization effects, see, forexample, Bresler et al. (1982), Eilers et al.(1995), Butler and Addison (2000) and Cain et al.(2001).

3.4.1.4 Conclusions

Application of road salts can result in deleteriouseffects on physical and chemical parameters ofsoils, especially in salt-affected areas that alsosuffer from general salt, soil and vegetationmanagement neglect. Highest effects arecalculated in areas that would be directlyimpacted by salt depots and along roadsides,especially in poorly drained depressions.Electrolyte-related effects impact on soilstructural stability, soil dispersion, soilpermeability, soil swelling and crusting, soilelectrical conductivity and soil osmotic potential.Surface waters downslope from salt-affected soilsare also affected by briny seepage water or brinyrunoff with salt dispersed sediments. All of thishas, in turn, abiotic and biotic impacts on thelocal environment. Abiotic impacts primarilydeal with loss of soil stability during drying andwetting cycles and during periods of high surfacerunoff and wind; biological impacts primarilydeal with osmotic stressors on soil macro-and microflora and macro- and microfauna

and salt-induced mobilization of macro- andmicronutrients, which, in turn, affect flora andfauna in situ and in downslope seepage areas.

3.4.2 Biological effects of road salts on soils

A comprehensive review of the effects of salt onsoil microflora and microfauna was compiled byButler and Addison (2000). There appear to bevery few data dealing with road salt exposureand soil microbial communities and/or theiractivities. Thus, available literature on themetabolic adaptations required for microbialsalt tolerance and on the effects of high soilsalinity on agriculturally important processeswas reviewed and addressed for this assessment,although the originators of the quoted datahad purposes very different from those of thisdocument. An additional difficulty is the variedways of reporting salt exposures in the literature.Soil-to-soil comparisons are typically based onelectrical conductivity, water potential or osmoticpotential, not added salt concentration, becausethe ionic composition of different soil solutionsvaries. However, salt concentrations (mass/liquidvolume) are more usually reported in themicrobiology literature. Studies involving soilfauna typically report salt inputs as amount perunit area.

3.4.2.1 Ecological role of soil bioticcommunities

The biological communities of soils andsediments are responsible for the degradationand mineralization of organic matter and cyclingof elements (e.g., carbon, nitrogen, phosphorus,sulphur) in these environments. In terrestrialsystems, there is only a minor direct contributionto primary production by photo- andchemolithotrophic microbes. However,experiments have indicated that the primaryproductivity of higher plants is greatly enhancedby the activities of the below-ground microbialand faunal biota (Faber and Verhoef, 1991;Setälä and Huhta, 1991; Setälä et al., 1998).Mycorrhizae are features of nearly all terrestrialplants, with the fungal partner of this symbioticplant–fungus relationship contributing to nutrient


(especially phosphorus) uptake and translocationand potential protection against plant pathogens,nematodes and toxicants such as heavy metalsin the rooting zone (Paul and Clark, 1996).The enzymatic capabilities of the soil microfauna(e.g., protozoa, nematodes) and mesofauna(e.g., mites, wingless insects) are more restrictedthan those of the prokaryotes and microfungi,so that most of the chemical transformationsinvolved in organic matter decomposition andnutrient cycling are mediated by soil microbes. Thecontributions of the soil fauna to these processesinclude more indirect effects such as physicalbreakdown of litter, dispersal of microbialpropagules (including mycorrhizae and pathogens),selective grazing on microbial species and, in thecase of macrofauna, modification of soil structure.For example, earthworms, where present, can exerta tremendous impact on soil aeration, aggregatestructure, water-holding capacity and bulkdensity of soils, as well as influence nitrogenmineralization rates (Lee, 1985).

3.4.2.2 Effects on soil biota

Data on effects of sodium chloride on soilorganisms are presented by Butler and Addison(2000). Soil bacteria can be quite sensitiveto sodium chloride. Hattori & Hattori (1980)studied the growth of 40 oligotrophic soilbacteria isolated from various Japanese soils.Two of the 40 bacteria were moderately inhibitedby 0.15 g sodium chloride/L (correspondingto approximately 60 mg sodium/L and 90 mgchloride/L). One bacterium was strongly inhibitedby 1 g sodium chloride/L (approximately 400 mgsodium/L and 600 mg chloride/L), while 17 werestrongly inhibited by 5 g sodium chloride/L(approximately 2000 mg sodium/L and 3000 mgchloride/L). Eleven of the 40 bacteria werestrongly inhibited by 10 g sodium chloride/L,while the remaining 9 were moderately inhibitedat 10 g/L. McCormick and Wolf (1980) found


FIGURE 24 Salt tolerance and crop yield relative to soil salinity (electrical conductivity, EC) and theestimated percentage of roadsides with corresponding levels of electrical conductivity(from Bresler et al., 1982)

that soil nitrification was significantly reducedwhen 0.25 g/kg (approximately 100 mg sodium/kgand 150 mg chloride/kg) or more sodium chloridewas present. Sodium chloride at �10 g/kgcompletely inhibited nitrification. Some rhizobia,soil bacteria that form nitrogen-fixing nodules withleguminous plants (including important crops likesoybean, pea and clover), tend to be salt-sensitive.The generation times of salt-sensitiveBradyrhizobia japonicum strains were increasedwith exposure to 2.6 g sodium chloride/L, whilethere was little effect on salt-insensitive variants(Upchurch and Elkan, 1977). Earthworms can alsobe quite sensitive to salt. Earthworms of thespecies Eisenia fetida died within 2 weeks whenexposed to �5 g sodium chloride/kg (Kaplan et al.,1980). McCormick and Wolf (1980) reported afairly sharp decline in carbon dioxide evolutionfrom a sandy loam treated for 14 weeks withsodium chloride when sodium chloride exceeded10 g/kg (corresponding to about 4000 mgsodium/kg and 6000 mg chloride/kg). Soilreceiving 100 g sodium chloride/kg did not releasedetectable carbon dioxide.

3.4.2.3 Exposure of soil organisms to road salt

Soil concentrations of sodium and chloride arepresented in Cain et al. (2001). The highest soilchloride concentration reported was 1050 mg/kgin a sample taken in the highway median, and thehighest soil sodium concentration was 890 mg/kgin a sample taken 10 m from the highway, bothfrom a study along a four-lane highway in Ontario(Hofstra and Smith, 1984). Concentrations ofsodium and chloride decreased rapidly withdistance from the road. Concentrations of sodiumwere below 100 mg/kg at distances greater thanabout 20 m from the road at all reported Canadiansites, while concentrations of chloride were below300 mg/kg at distances greater than about 30 mfrom the road.

Sodium concentrations of up to13 100 mg/kg and chloride concentrations up to14 500 mg/kg have been reported in soils at patrolyards (Section 2.4.4.2).

3.4.2.4 Environmental assessment

Sensitive soil bacteria are moderately inhibitedby sodium chloride at concentrations as low as150 mg sodium chloride/L (corresponding toapproximately 60 mg sodium/L and 90 mgchloride/L). Soil nitrification can be significantlyreduced at a sodium chloride concentration of250 mg/kg (approximately 100 mg sodium/kgand 150 mg chloride/kg). Soil concentrations ofsodium exceeding 60 mg/kg have been reportedwithin about 30 m from the edges of roads inCanada. Soil concentrations of chloride exceeding200 mg/kg have been reported within about200 m from the edges of roads in Canada.

Concentrations high enough to adverselyaffect microorganisms and other soil organismshave also been reported for soils at patrol yards.

3.4.2.5 Conclusions

Concentrations of sodium and chloride in soilnear the edges of roads in Canada and in patrolyards are high enough to have adverse effectson sensitive soil organisms. How serious theseeffects might be to soil ecosystem function isvery difficult to gauge. Soil microorganisms mayhave considerable ability to adapt to increasedsalinity or to reestablish from unaffected areas.The dispersal capabilities of soil fauna such asearthworms are low, as most species are adaptedto the relatively high humidity of the soilenvironment. Where affected areas are smallenough — for example, strips of soil adjacent toroads — populations of impacted soil organismsmay be able to be reestablished quite quickly ifthe unfavourable salt conditions are abated.

3.5 Terrestrial vegetation

Information presented in this risk assessmentsection is taken from Cain et al. (2001).

Elevated levels of sodium and chloride inthe substrate or soil affect plants in various ways:


• inhibition of water and nutrient absorption byplants due to osmotic imbalances, resulting inreduced shoot and root growth and drought-like symptoms;

• nutritional imbalances for some species bydisrupting uptake of other nutrients;

• long-term growth inhibition and, at higherlevels, direct toxicity to the plant cells,manifested in leaf burn symptoms and tissuedeath; and

• deterioration of soil structure.

Excess soil salinity leads to deteriorationof soil structure due to soil crusting and cloggingof soil pores by entrapment of dispersed soil clayand silt particles (Morin et al., 2000). Soilcrusting reduces shoot emergence of subsurfaceseeds and root penetration of both surface andsubsurface seeds, resulting in reduced plantestablishment. Clogging of soil pores reducesa) soil space available for air and water retentionand b) air and water penetration and permeation.Reduced soil aeration is a concern, since reducedoxygen supply to plant roots affects root growth.

An additional impact of elevated chloridein roadside soils is to increase the availabilityof heavy metals. Chloride forms complexes withheavy metals, rendering many of them more watersoluble and therefore more available for uptakeby plant roots. This would increase plant uptakeof such metals, possibly resulting in plant toxicity,depending on the type and amount of metalspresent in the soil. Smolders and McLaughlin(1996), for example, reported on chloride-relatedenhancement of cadmium mobilization throughthe soil and increased plant uptake.

Laboratory and experimental fieldstudies (done away from roadside locations)have demonstrated that soil and spray applicationsof sodium and calcium chloride severelyinjure plants and have characterized the injurysymptoms (Hall et al., 1972; Cordukes andMaclean, 1973; Dirr, 1975, 1978; Hofstra andLumis, 1975; Paul et al., 1984; Headley andBassuk, 1991; Dkhili and Anderson, 1992;Harrington and Meikle, 1994).

Plant symptoms following exposureto elevated soil levels of sodium and chlorideinclude general plant decline, reduction in leafsize and plant growth, leaf chlorosis, leaf burnand tissue death. Seed germination can be reducedor delayed as well.

Injury of woody plants along roadsidesdue to salt spray is the result of tissue drought ordesiccation and is related to the penetration ofphytotoxic ions of sodium and chloride throughthe stem, bud and leaf tissues (Barrick andDavidson, 1980; Chong and Lumis, 1990).Death of stems and death of leaf and flower buds,usually on first-year shoots, are the predominantsymptoms on deciduous woody plants. Flowerbuds are more sensitive to salt injury than leafbuds. Conifers exhibit premature leaf drop andbrowning of needles.

Salt levels in soil or plant tissue aregenerally highest in the spring and declinewith rainfall (leaching of soil) or plant growth.Some species may be able to recover or regrowfollowing injury early in the growing season.However, many herbaceous plant speciesgerminate or regrow from dormant propagules inthe spring due to specific day length, temperatureand moisture requirements. Once a bank of seedsor propagules on a site is killed by elevated saltlevels, the affected species may not regrow,resulting in a change in the plant community.

With woody species, even though plantinjury is often worst in the spring and some plantsrecover later in the growing season, repeatedinjury, year after year, reduces the vigour andgrowth of the plants. If the soil or tissue levelsare high enough, a plant may be killed outright ormay have extensive dieback, which results inplant weakening and death over a number of years(Sucoff, 1975; Lumis et al., 1976). In addition,woody plants that are weakened by the repeatedstress of road salt are more sensitive to disease,insects or abiotic stresses, such as winter injury(Sucoff and Hong, 1976) or drought.

For woody species that flower in thespring, flower buds lost due to salt damage will


not be recovered later in the season. If thisdamage occurs year after year, there will be anegative impact on the reproductive capabilityof spring flowering species.

Dirr (1978) pointed out that the patternof plant injury and elevated soil or tissue levelsshould be used to confirm that the cause of thedamage is de-icing salts. The following injurypatterns are associated with road salt injury toplants (Lumis et al., 1973; Dirr, 1976):

• injury occurs in a linear pattern along roads orin areas where runoff from roads collects;

• injury is more severe on the side of the plantfacing the road;

• injury decreases with distance from the road;• injury is worse on the downwind side of the

road; • parts of woody plants that are covered by

snow or were sheltered from spray lack injurysymptoms;

• parts of trees that are above the salt sprayzone are not injured or are injured less;

• salt spray injury extends only a short distanceinto dense plants;

• injury in coniferous trees becomes evident inlate winter and continues into the growingseason; and

• injury in deciduous trees becomes evident inspring when growth resumes and continuesinto the growing season.

Additional information about injuryto plants characteristic of road salt exposure ispresented in Cain et al. (2001).

There have been extensive compilationsof ratings of sensitivity of woody plants fromvarious observational and experimental studies.A survey of these woody plant ratings(ornamental and forest species) in the literatureindicated that 29.6–50.8% of the plants listed ineach report were deemed sensitive to salt in atleast one rating (Sucoff, 1975; Dirr, 1976; Lumiset al., 1977; Thuet, 1977; Beckerson et al., 1980;Dobson, 1991).

Lumis et al. (1977), making fieldobservations of plants on Ontario roadsides, rated30% of 73 species as sensitive to road salt; 7%had severe injury symptoms that could result inplant death. A review of sensitivity ratings ofplants commonly used in Canadian landscapeplantings (Beckerson et al., 1980) found that40% of 89 species had been rated as sensitive toroad salt.

A survey of the salt tolerance ratingsof 32 cultivars or varieties of fruit trees, vinesor shrubs, representing commercial crop andlandscape plants (Appendix 2 in Cain et al.,2001), indicated that 65% were consideredsensitive to salt in at least one rating.

There are 15 principal genera of trees(over 25 species) within Canadian forestedregions, according to Rowe (1972), and 11 ofthese have been rated as sensitive to road salt(Section 4.5 in Cain et al., 2001). Looking atindividual species listed by Rowe (1972), 8 of 15(53%) have been rated as sensitive to road salt inat least one rating.

A list of roadside trees and shrubs ratedfor their resistance to airborne highway salt spray,based on Lumis et al. (1983), is presented inTable 19. A list of native forest trees rated fortheir resistance to airborne highway salt spray ispresented in Table 20.

Additional information about thesensitivity of various types of plants is included inCain et al. (2001).

3.5.1 Conservative environmentalassessment

EC25 and CTV threshold values were estimatedfrom the scientific literature for chloride, sodiumand sodium chloride in growing media (soil,soil water or water); for sodium chloride in foliarspray solution; and for chloride and sodium inplant tissue (Cain et al., 2001). Lowest-Observed-Effect Level (LOEL) threshold values wereidentified for chloride and sodium chloride in



TABLE 19 Species list of roadside trees and shrubs rated for their resistance to airborne highway saltspray (from Lumis et al., 1983)

Tree/shrub type Injury Tree/shrub type Injuryrating 1 rating 1

Deciduous trees Deciduous shrubsNorway maple Acer platanoides 1 Siberian peashrub Caragana arborescens 1Horse-chestnut Aesculus hippocastanum 1 Sea-buckhorn Hippophae rhamnoides 1Tree of Heaven Ailanthus altissima 1 Staghorn sumac Rhus typhina 2Honeylocust Gleditsia triacanthos inermis 1 Burningbush Euonymus alatus 2Cottonwood Populus deltoides 1 Honeysuckle Lonicera spp. 2Black locust Robinia pseudoacacia 1 Japanese tree lilac Syringa amurensis japonica 2Shagbark hickory Carya ovata 2 Common lilac Syringa vulgaris 3Russian-olive Elaeagnus angustifolia 2 Speckled alder Alnus rugosa 3White ash Fraxinus americana 2 Border forsythia Forsythia x intermedia 3Largetooth aspen Populus grandidentata 2 Privet Ligustrum spp. 3Lombardy poplar Populus nigra ‘Italica’ 2 Mockorange Philadelphus spp. 4Trembling aspen Populus tremuloides 2 Flowering-quince (Sweet) Chaenomeles speciosa 4Choke cherry Prunus virginiana 2 Beautybush Kolkwitzia amabilis 4Pear Pyrus sp. 2 Bumalda spirea Spiraea x bumalda 4Red oak Quercus rubra 2 European cranberry-bush Viburnum opulus 5Mountain-ash Sorbus aucuparia 2 Gray dogwood Cornus racemosa 5Amur maple Acer ginnala 3 Red-osier dogwood Cornus stolonifera Red maple Acer rubrum 3Silver maple Acer saccharinum 3Sugar maple Acer saccharum 3 ConifersPaper birch Betula papyrifera 3 Blue spruce Picea pungens Englem. glauca 1Gray birch Betula populifolia 3 Jack pine Pinus banksiana 1Northern catalpa Catalpa speciosa 3 Mugo pine Pinus mugo 1Quince Cydonia oblonga 3 Austrian pine Pinus nigra 1Green ash Fraxinus pennsylvanica 3 Red cedar Juniperus virginiana 1

lanceolata Juniper Juniperus spp. 2Black walnut Juglans nigra 3 Norway spruce Picea abies 2English walnut Juglans regia 3 Yew Taxus spp. 3Black willow Salix nigra 3 White spruce Picea glauca 3Basswood Tilia americana 3 Red pine Pinus resinosa 4White elm Ulmus americana 3 Scots pine Pinus sylvestris 4Siberian elm Ulmus pumila 3 White cedar Thuja occidentalis 4Manitoba maple Acer negundo 4 White pine Pinus strobus 4Allegheny serviceberry Amelanchier laevis 4 Hemlock Tsuga canadensis 5Hawthorn Crataegus spp. 4Apple Malus sp. 4Crabapple Malus spp. 4Mulberry Morus sp. 4Peach Prunus persica 4Weeping golden willow Salix alba ‘Tristis’ 4American beech Fagus grandifolia 5

1 A rating of 1 indicates no twig dieback or needle browning of conifers and no dieback, tufting or inhibitions of flowering ofdeciduous trees and shrubs. Ratings of 5 represent complete branch dieback and needle browning of conifers, and completedieback, evidence of previous tufting and lack of flowering of deciduous trees and shrubs. Under severe conditions, plantsrated 5 will eventually die. Ratings of 2, 3 and 4 encompass slight, moderate and extensive gradations of the above injurysymptoms.


TABLE 20 Species list of native forest tree species rated for their resistance to highway salt spray

Scientific name Common name Salt Referencessensitivity 1

Fagus grandifolia American beech S Beckerson et al., 1980; Dobson, 1991Quercus rubra Red oak S-M Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991; Barker, 2000 Quercus velutina Black oak S-M Barker, 2000Quercus palustris Pin oak S-M Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991; Barker, 2000 Quercus alba White oak S-M Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991; Barker, 2000 Quercus macrocarpa Bur oak S-M Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Carya ovata Shagbark hickory S-M Beckerson et al., 1980; Dobson, 1991Abies balsamea Balsam fir S-MT Beckerson et al., 1980; Dobson, 1991;

Barker, 2000Acer saccharum Sugar maple M-MT Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Acer rubrum Red maple M-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Acer saccharinum Silver maple M-T Beckerson et al., 1980; Dobson, 1991Acer negundo Manitoba maple, box-elder M-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Salix nigra Black willow M-T Beckerson et al., 1980; Dobson, 1991Betula papyrifera White birch S-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991; Barker, 2000 Pinus banksiana Jack pine S-T Beckerson et al., 1980; Dobson, 1991Populus tremuloides Trembling aspen S-T Beckerson et al., 1980; Dobson, 1991;

Barker, 2000Picea glauca White spruce S-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Salix sp. Willow S-T Dobson, 1991Pinus ponderosa Ponderosa pine S-T Sucoff, 1975; Dobson, 1991 Pseudotsuga menziesii Douglas-fir S-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Thuja plicata Western red cedar S-T Dobson, 1991Juglans nigra Black walnut S-T Beckerson et al., 1980; Dobson, 1991Pinus resinosa Red pine S-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991Pinus strobus Eastern white pine S-T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991; Barker, 2000 Tsuga canadensis Eastern hemlock T Sucoff, 1975; Beckerson et al., 1980;

Dobson, 1991 Betula alleghaniensis Yellow birch T Dobson, 1991; Barker, 2000

1 S – sensitive, MS – moderately sensitive, M – intermediate, MT – moderately tolerant, T – tolerant.

growing media, and No-Observed-Effect Level(NOEL) threshold values were identified forsodium chloride in growing media.

3.5.1.1 Effects threshold values

Threshold values, derived from experimental dataor roadside studies, have been tabulated in Tables21, 22 and 23 and in Section 3.6 in Cain et al.(2001). The plant species represented includewetland species (e.g., rushes, sphagnum moss,bulrush, cattails and duckweed), herbaceousspecies (e.g., sedges, grasses and beets),coniferous species (e.g., pine and spruce) anddeciduous species (e.g., maple, peach and plum).The threshold values ranged from 67.5 to300 ppm sodium, 215 to 1500 ppm chloride and280 to 66 600 ppm sodium chloride in growingmedia (applied soil solution) (Table 21); from<280 to 17 800 ppm sodium and <4100 ppmchloride in herbaceous tissue following soilor water solution exposure (Table 22); andfrom 200 to 16 100 ppm sodium and 800 to70 700 ppm chloride in woody tissue followingaerial exposure and 7140 ppm sodium chloridefoliar spray concentration (Table 23).

For the conservative assessment, thevalues used were the lowest concentrations thatwere identified in the threshold value analysis(Section 3.6 in Cain et al., 2001) for each typeof threshold described above from experimentalevaluations or from sampling done beyond theroad or highway right-of-way limits.

3.5.1.2 Estimated Exposure Values (EEVs)

The EEVs used in the conservative assessmentwere the highest values extracted from theliterature dealing with environmentalconcentrations in Canada of chloride or sodiumin growing media (soil or aqueous) in roadsideconditions from sampling done beyond thehighway right-of-way or the highest valuesextracted from literature dealing withconcentrations of chloride or sodium found invegetation growing along salted roadsides inCanada beyond the highway right-of-way.

The exposure values used for theassessment were 240 ppm soil chloride and45 ppm soil sodium in samples taken outsideof the highway right-of-way, 40 m from thehighway, both from a study along a four-lanehighway in Ontario (Hofstra and Smith, 1984).


TABLE 21 Range of threshold values estimated for soil and water for various forms of plants(from Cain et al., 2001)

Pathway Form Plant form Threshold type Threshold range (ppm)Soil Na Herbaceous EC25 202–270Soil Na Woody EC25 67.5–300Root uptake Na All species EC25 67.5–300

Water solution Cl Wetland LOEL 300–1500Soil Cl Woody EC25 215–500Root uptake Cl All species EC25, LOEL 215–1500

Soil NaCl Woody EC25 600–5500Solution culture NaCl Herbaceous EC25 <2500–10 000Solution culture NaCl Wetland NOEL, LOEL 280–66 600Solution culture NaCl Woody EC25, CTV 836–25 000Root uptake NaCl All species EC25, CTV, NOEL, LOEL 280–66 600

The highest soil chloride concentration reportedalong a highway was 1050 ppm in a sample takenin the highway median, and the highest soilsodium concentration was 890 ppm in a sampletaken 10 m from the highway, both from the samestudy.

No Canadian references reported soilconcentrations of sodium chloride. A value ofsoil sodium chloride was estimated based on theconcentration of chloride found in the soil 40 mfrom the highway, 240 ppm, which was multipliedby a factor of 1.6485, resulting in a value of396 ppm sodium chloride. A second value ofsoil sodium chloride was estimated based on theconcentration of sodium found in the soil 40 mfrom the highway, 45 ppm, which was multipliedby a factor of 2.542, resulting in a value of114 ppm sodium chloride. The average of thesetwo values, 255 ppm sodium chloride, was used inthe conservative estimation.

In order to estimate the EEVs for aerialexposure of plants to sodium chloride followingroad de-icing, references were reviewed thatreported concentrations of chloride or sodium inplant tissues at given distances from the road or

highway, following exposure to road salt aerialdispersion (refer to Section 2.3 of Cain et al.,2001). Three references were reviewed for tissueconcentrations of chloride and sodium. Thesedata included plant tissue samples taken from10 to 200 m from the road or highway. Thesereferences reported studies from Ontario on Prunus(peach and plum) and unspecified vegetation.

The tissue concentrations used for theassessment were 4600 ppm chloride and 3500ppm sodium reported in peach (Prunus persica)growing 40 m from a four-lane highway, outsidethe highway right-of-way, in southern Ontario(Northover, 1987). By comparison, the highesttissue chloride concentration was 11 000 ppm,reported in unspecified vegetation growing 60 mfrom a four-lane highway in Ontario (Hofstra andSmith, 1984). Northover (1987) also reported9000 ppm chloride in peach growing 20 m from afour-lane highway in southern Ontario (Northover,1987). The highest tissue sodium concentrationwas 6900 ppm, reported in twig tissue of peachgrowing 20 m from a four-lane highway insouthern Ontario (Northover, 1987).


TABLE 22 Range of threshold values estimated for herbaceous tissue concentrations following soil orwater solution exposure (from Cain et al., 2001)

Pathway Form Plant form Threshold type Threshold range (ppm)Soil Na Herbaceous EC25 <280–17 800Soil Cl Herbaceous EC25 <4100

TABLE 23 Range of threshold values estimated for woody tissue concentrations following aerialexposure (from Cain et al., 2001)

Pathway Form Plant form Threshold type Threshold range (ppm)Vegetation Na Woody EC25 200–16 100concentration

Vegetation Cl Woody EC25, CTV 800–70 700concentration

Foliar spray NaCl Woody CTV 7140concentration

3.5.1.3 Quotient estimation

From Section 3.5.1.2, it can be seen that chlorideand sodium have been reported in Canadian soilsand in plant tissues outside of the highway right-of-way at concentrations higher than or, in onecase, in the range of the effects threshold valuespresented in Section 3.5.1.1 and in Section 3.6of Cain et al. (2001). Therefore, a detailedassessment of the effects of road salts onterrestrial vegetation is warranted.

3.5.2 Detailed environmental assessment —threshold-based analysis

This risk characterization reviews concentrationsof chloride and sodium in soils or tissue thatproduce biologically relevant effects onvegetation and relates these to concentrationsfound in the Canadian environment. The datapresented represent primarily Canadian literatureand incorporate threshold values that werepresented in Section 3.5.1.1 above and in Section3.6 of Cain et al. (2001).

3.5.2.1 Response to sodium in soil

The range of threshold values for sodium chlorideapplied to roots was 67.5–300 ppm sodium(Table 21 above and Section 3.6 of Cain et al.,2001). The most sensitive plants reported were 2-year-old ponderosa pine (Pinus ponderosa)seedlings, with a threshold value of 67.5 ppm,based on foliar injury following soil application ofa sodium chloride solution. Mortality occurred ata concentration of 140 ppm sodium (Bedunah andTrlica, 1979). Seed germination rates of prairiegrass and broad-leaved wildflower species werereduced by about 60–70% and root growth wasreduced by up to 73% when plants wereexposed to 100–400 ppm sodium chloride(about 40–160 ppm sodium) (Harrington andMeikle, 1994). The EC25 threshold values thatwere estimated from this study for sodiumconcentration in soil were 202–270 ppm.

The concentration of sodium in soilexceeded the estimated threshold range of67.5–300 ppm in a number of soil samplescollected in Canada at a distance of about 30 mor less from the roads (Figure 2.17 in Cain et al.,2001). A concentration of 890 ppm sodium wasfound in soil 10 m from a four-lane highway insouthern Ontario (Hofstra and Smith, 1984).A concentration of 45 ppm sodium was found40 m from the highway, outside the right-of-way,in the same study.

3.5.2.2 Response to chloride in soil

The range of threshold values for sodium chlorideapplied to roots was 215–1500 ppm chloride(Table 21 above and Section 3.6 of Cain et al.,2001). The most sensitive plants reported were 2-year-old ponderosa pine seedlings, with athreshold value of 215 ppm chloride, basedon foliar injury following soil application of asodium chloride solution. Mortality occurred at aconcentration of 350 ppm chloride (Bedunah andTrlica, 1979).

The estimated LOEL for reductionin mean length growth of sphagnum moss,Sphagnum recurvum, was 300 ppm chloride in ahydroponic solution (Wilcox, 1984). Wilcox andAndrus (1987) reported reproductive effects,reduced gametophore production, in Sphagnumfimbriatum at 300 ppm chloride.

The concentration of chloride in soilsamples collected in Canada at a distance of upto 200 m from the roads exceeded the estimatedlowest threshold of 215–500 ppm for a numberof samples (Figure 2.16 in Cain et al., 2001).Concentrations of chloride in soils exceeded500 ppm only at the edge of the roads. A soilconcentration of 240 ppm chloride was foundin samples taken 40 m from a four-lane highway,beyond the right-of-way, in southern Ontario(Hofstra and Smith, 1984). A soil concentrationof 1200 ppm chloride was found in the soilbehind a patrol yard located along a highway innorthwestern Ontario (Racette and Griffin, 1989).


3.5.2.3 Response to sodium chloride in soil

The range of threshold values for sodium chlorideapplied to roots was 280–66 600 ppm sodiumchloride (Table 21 above and Section 3.6 of Cainet al., 2001). The most sensitive reported specieswere two wetland species, three-fruited sedge(Carex trisperma) and common mountain holly(Nemopanthus mucronata [mucronatus]), whichhad salt tolerance limits of 280 ppm based onsodium chloride concentration gradients in bogwater (Wilcox, 1986a). Survival of Coloradoblue spruce (Picea pungens) and growth of Scotspine (Pinus sylvestris) were affected by sodiumchloride concentrations of 600 ppm (Werkhovenet al., 1966).

Schauffler (1993) reported sublethaleffects, biomass increase and tiller production ofcottongrass (Eriophorum vaginatum var. spissum)with a concentration of 400 ppm sodium chloridein a hydroponic culture, and reproductive effects,reduced flower production, with a concentrationof 1600 ppm.

No Canadian references reported soilconcentrations of sodium chloride; however,values have been estimated using reportedconcentrations of sodium and chloride. A valueof soil sodium chloride of 1978 ppm wasestimated by multiplying the highest Canadianconcentration of chloride found in the soiladjacent to a patrol yard, 1200 ppm, by a factorof 1.6485. A second value of soil sodium chlorideof 2262 ppm was estimated by multiplying thehighest concentration of sodium found in the soil10 m from the highway, 890 ppm, by a factorof 2.542. The average of these two values, 2120,was about 10 times the lowest threshold value of280 ppm.

Using the same method of calculation, avalue of 255 ppm sodium chloride outside of thehighway right-of-way was calculated from soilvalues of 240 ppm chloride and 40 ppm sodium,found 40 m from a four-lane highway in southernOntario (Hofstra and Smith, 1984). This value isin the range of the threshold value of 280 ppm.

3.5.2.4 Response to sodium and chloride byaerial dispersion

The range of threshold values for aerial exposureto sodium chloride was 800–70 700 ppm chlorideand 200–16 100 ppm sodium (Table 23 aboveand Section 3.6 of Cain et al., 2001). The mostsensitive species reported in a Canadianenvironment were peach (Prunus persica) andplum (Prunus domestica), with a threshold valueof 800 ppm chloride and 650 ppm sodium, basedon twig dieback (a sublethal effect) followingaerial exposure (McLaughlin and Pearson, 1981).Mortality of multiple pine species (Pinus spp.)had threshold values of 13 000 ppm chloride and8000 ppm sodium (Townsend and Kwolek, 1987).Reproductive impacts, evaluated as peach fruityield, had threshold values of 1000 ppm chlorideand 1700 ppm sodium (Northover, 1987).

Tissue concentrations exceeding thelowest threshold values of 800 ppm chloride and650 ppm sodium estimated for sensitive plantspecies (Section 3.6.2 of Cain et al., 2001) havebeen reported at distances up to about 120 mfrom highways for sodium and up to 200 m forchloride. Tissue concentrations of 4600 ppmchloride and 3500 ppm sodium were reported inpeach growing 40 m from a four-lane highway,outside the highway right-of-way, in southernOntario (Northover, 1987). The highestconcentrations reported in woody plants growingalong a road were 9000 ppm chloride and6900 ppm sodium in peach twig tissue growing20 m from a four-lane highway in southernOntario (Northover, 1987).

3.5.2.5 Detailed environmental assessment —summary of threshold-based analysis

Elevated soil levels of sodium and chloride havebeen recorded up to 30 m from four-lanehighways (Hofstra and Smith, 1984; Davis et al.,1992) and up to 15 m from two-lane highways(Davis et al., 1992; Soilcon Laboratories Ltd.,1995). The threshold range of 67.5–300 ppmsodium and 215–500 ppm chloride in soil wouldbe exceeded in a zone up to about 30 m fromhighways (Figures 2.7 and 2.8 in Cain et al.,


2001), depending on the class of road, applicationrates, local topography and climate. Thresholdvalues have also been exceeded on sites wheresalty runoff has collected (Racette and Griffin,1989).

Similarly, elevated tissue concentrationsof sodium and chloride have been reported forvegetation growing adjacent to highways insouthern Ontario. Tissue concentrations exceedingthe lowest threshold values of 650 ppm sodiumand 800 ppm chloride estimated for sensitiveplant species (Section 3.6.2 of Cain et al., 2001)have been reported at distances up to about 120 mfrom highways for sodium and up to 200 m forchloride.

3.5.3 Detailed environmental assessment —reference site analysis

Reference sites are presented that documenteffects on vegetation that can be traced toapplication of road salts. A number ofCanadian studies have evaluated environmentalconcentrations of sodium and chloride and plantimpacts as a result of de-icing salt application toroads or runoff as a result of handling activitiesin highway maintenance patrol yards. Some U.S.studies that are relevant to Canada have also beenincluded in this section. Details about thesestudies can be found in Section 4.4 of Cain et al.(2001).

3.5.3.1 Spread of halophytic plant species

The spread of halophytic species, characteristicof sea coasts and salt marshes, has been recordedalong roadsides in Ontario. Catling and McKay(1980) reported the occurrence of 20 salt-tolerantplant species along roadsides in Ontario, of which4 species were reported for the first time in theprovince. The range of an additional four specieswas extended significantly. The species discussedby Catling and McKay (1980) were naturallyoccurring plants, not ones planted by the roadauthorities for roadside stabilization orlandscaping. These species were spreadingnaturally along highways in response to the use

of de-icing salts and the decline of non-adaptedplanted vegetation. The sites studied had sodiumconcentrations in the range of 1000 ppm andgrowing conditions that ranged from aquatic tomoist to periodically moist to dry. These sitesincluded roadside ditches, road and highwaymedians, areas adjacent to highways that receivesalty runoff, and areas adjacent to salt storagedepots and snow dumps. These included areaswithin the engineered right-of-way, within theplanted right-of-way and outside the right-of-way.

Species that were tolerant of soilsodium levels of 500–1000 ppm and were oftenassociated with halophytic species includednarrow-leaved cattail (Typha angustifolia),common reed-grass (Phragmites australis) andweedy species such as annual and perennial sow-thistle (Sonchus oleraceus, Sonchus arvensis,Sonchus uliginosus), common ragweed (Ambrosiaartemisiifolia), wild carrot (Daucus carota) andkochia (Kochia scoparia).

Reznicek et al. (1976) reported theincrease in distribution of Carex praegracilis,an alkali-tolerant western sedge species, sinceits first observation in Ontario in 1973. The plantwas observed in 28 highway locations in 14counties, generally within highway rights-of-way.Additional reports of roadside and highwaymedian occurrences of this species were reportedby Catling and McKay (1980).

Similar observations of the spread of salt-tolerant species have been made in Michigan,which has a climate and flora similar to those ofsouthern Ontario (Reznicek, 1980).

3.5.3.2 Community changes in a bog impactedby road salts

The effects of sodium and chloride contaminationwere studied in Pinhook Bog, located innorthwestern Indiana, southeast of Michigan City,which had been impacted by road salt runoff(Wilcox, 1982, 1986a,b). This study providesevidence of plant community impacts by elevatedsalt levels. Many of the predominant plant species


in this bog are native wetland species that arefound in Canada. This part of Indiana is part ofthe Great Lakes basin and has the same climaticzone as parts of southern Ontario, the Montréalregion, the Atlantic provinces and BritishColumbia.

An uncovered road salt storage pile hadbeen located on an embankment adjacent to thebog from 1963 to 1972. A protective dome wasplaced over the salt pile in 1972 to limit runoff,and salt storage was discontinued on this site inmid 1981. The bog also received runoff from a 1-km stretch of a four-lane highway. The annualroad salt application for the adjacent highway was11 300 kg/km (5650 kg per two-lane-kilometre).

The most severely impacted area hadmean salt concentrations as high as 486 ppmsodium and 1215 ppm chloride in the interstitialwaters of the peat mat. The pattern of elevatedsodium and chloride concentrations in the bogwas correlated with zones of affected vegetation.The salt concentrations in the root zone decreasedeach year from 1979 to 1981.

Vegetation evaluations were carried outon the unimpacted and impacted portions of thebog from 1980 to 1983. The salt contaminationhad resulted in a shift in vegetation species in theimpacted areas. Nearly all of the endemic plantspecies were absent from the portion of the bogwith elevated salt levels. They had been replacedwith non-bog species dominated by narrow-leaved cattail.

As salt levels declined by 50% over the 4-year study period, many endemic speciesreturned to the impacted bog. Some of the speciesthat had invaded during the high salt conditionsdeclined, while others, such as narrow-leavedcattail and purple chokeberry (Pyrus floribunda),remained dominant in parts of the impacted bog.This study provides evidence that wetlandcommunities can recover once a source of salt isremoved, although certain invasive species mayremain, resulting in a permanent change incommunity composition.

3.5.3.3 Effects of salt spray from a four-lanehighway on peach and plum orchards inGrimsby, Ontario

The effect of distance on the impact of salt spraywas studied in a number of peach and plumplantings along the Queen Elizabeth Way, a four-lane, controlled-access highway in the Grimsbyand Beamsville area of Ontario (McLaughlin andPearson, 1981). These plantings were on growerproperties outside of the highway right-of-way.Peach and plum trees were evaluated for terminaltwig dieback in April 1980 and 1981. Twigsamples were evaluated for chloride and sodiumcontent as well.

The chloride and sodium contents of thetwigs were highest closest to the highway andgenerally decreased with increasing distance fromthe highway (Figures 2.11 and 2.12 in Cain et al.,2001). Twig tissue sodium concentration was2030 ppm at 45 m from the highway, and chlorideconcentration was 2300 ppm at 40 m from thehighway.

Twig dieback was greatest in the plantsclosest to the highway and decreased withincreasing distance from the highway (Figure 4.5in Cain et al., 2001); however, the degree ofdieback differed between years and species. Injurywas greater on the south side of the highway inApril 1981; peach twig dieback was 288 mm at22 m from the highway, decreasing to 97 mm at82 m from the highway. In April 1980, peach twigdieback on the south side of the highway fell from91 mm at 36 m from the highway to 8 mm at82 m from the highway. Twig dieback wascorrelated with tissue concentrations of sodiumand chloride. Twig dieback increased withincreasing concentrations of sodium and chloridein the combined peach and plum data (thresholdevaluation, Appendix 8 in Cain et al., 2001).

3.5.3.4 Effects of salt spray from a four-lanehighway on peach and plum orchardsnear St. Catharines, Ontario

The effect of salt spray was evaluated onLoring peach trees during the 1973–74 winter


(Northover, 1987). The trees were located on thesouth side of an east–west section of the QueenElizabeth Way, a busy, four-lane, controlled-access highway near St. Catharines, Ontario.The orchard was located at the same elevation asthe highway, on a grower property outside of thehighway right-of-way.

Sodium and chloride tissue content ofshoot tissue decreased with increasing distancefrom the highway. The chloride content of thepeach twig tissue, from trees located 20 m fromthe highway, was greater than 4 times the levelsfound in trees 120 m from the highway(9000 ppm compared with 1900 ppm chloride).The sodium content was elevated more than 7times the levels detected 120 m from the highway(6900 ppm compared with 900 ppm sodium).

Fruit yield was reduced at distances lessthan 80 m from the highway, and peach fruitnumber per shoot increased with increasingdistance from the highway. Shoots in the treecanopy were 97% dead 20 m from the highway.Shoot death decreased with increasing distancefrom the highway, falling to 8% at 80 m fromthe highway and decreasing to a background levelof less than 1% at 120 m from the highway.The sodium and chloride contents of the shootswere positively correlated with the percentage ofdead wood in the canopy of the trees.

The soil levels of sodium and chloridewere not elevated on this site, indicating thatthe elevated tissue levels and plant effectswere primarily due to aerial salt spray fromthe highway.

3.5.3.5 Effects of salt spray from a two-lanehighway on blueberries in westernNova Scotia

Eaton et al. (no date) evaluated the effects ofroad salt from a two-lane highway in the FollyMountain area, west of Truro, over the wintersfrom 1993 to 1995 on naturally occurringmanaged lowbush blueberry stands on adjacentlands. Salt application in this area was 33 and

40 tonnes of sodium chloride per kilometre ofhighway during the winters of 1993–94 and1994–95, respectively.

Blueberry blossom number and yieldwere reduced within 35 m of the two-lanehighway and approached background levelsbeyond 50 m from the highway. Blossom numberwas reduced by 60–90% within 10 m of thehighway, and yield was reduced by 50% or morecompared with samples taken at 50 m from thehighway. Yield under protective structures ofnursery film was 4224 kg/ha of fresh fruit in 1994compared with less than 2000 kg/ha at 30 m andless than 200 kg/ha at 5 m from the highway. In1995, yield under the protective structures was8092 kg/ha compared with less than 3000 kg/haat 35 m and less than 1500 kg/ha at 5 m from thehighway.

Stem tissue salt levels taken under theprotective structures were 0.53 and 0.3 g sodiumchloride per 20 stems in 1993–94 and 1994–95,respectively. This compared with levels from 6 toover 10 g at 10 m from the road in spring 1994and 3 to over 8 g in spring 1995. Tissue levelswere elevated up to 45 or 50 m from the highway.Reductions in both blossom number and yieldwere negatively correlated with elevated saltcontent of the blueberry stems taken in winter andearly spring.

3.5.3.6 Concentrations in soil and woody plantsadjacent to highways in interior BritishColumbia

Davis et al. (1992) reported on the causes ofobserved roadside tree injury and decline on 17forested sites in south-central and eastern BritishColumbia, located on two-, three- or four-lanehighways.

At 15 of the sites, sodium chloride wasused for winter maintenance; application ratesvaried between 60 and 130 kg per lane-kilometreas conditions dictated. On two of the sites thathad gravel roads, sodium chloride was not usedfor de-icing, although sand with 3–5% sodium


chloride was applied in the winter. In the summer,calcium chloride or magnesium chloride was usedfor dust suppression, initially at 2 L/m2, but at1.6 L/m2 at the time of the study.

A detailed field evaluation wasundertaken on the 17 sites. Three sample plotswere established, including two injury plots withmoderate to severe injury symptoms and onecontrol plot with no visible injury symptoms.A detailed assessment was made of plant saltinjury symptoms and injury due to other causes,such as abiotic stress, disease, insects orbreakage. Soil and vegetation samples wereanalysed for sodium and chloride content andother parameters, and a detailed history wascompleted for each site.

The pattern of injury to conifer trees on14 of the sites was similar. At these sites, injurydecreased with increasing distance from thehighway and was typical of road salt injury inwoody plants, including leaf scorch, leaf loss,thinning of crowns and severe defoliation anddieback in severely injured trees. In most of thesites, the injury occurred on a narrow strip, within5–15 m downslope of the highway. At other sites,injury occurred as close as 2 m from the road,since the forest stand was close to the road.At one site, injury was evident for 40 m into astand that received road runoff from a culvert.At another site, minor injury was evident 50 mfrom the highway and was attributed to salt sprayfrom the road. Injury occurred within the highwayright-of-way of many of the sites, since thetypical vegetated right-of-way in BritishColumbia is 4.5–5 m along a two-lane highway orup to 10 m on some highways. At other sites,injury was outside the highway right-of-way,more than 5 or 10 m from the highway.

Soil chloride concentrations at 0- to 10-cm depth in the control plots ranged from <1to 12 ppm. Soil concentrations of chloride on theplots with plant injury ranged from 0.9 to230 ppm chloride and were greater than 25 ppmon six of the sites. Two of the sites that wereadjacent to a four-lane highway had soil chloridelevels ranging from 3.0 to 198 ppm.

Soil sodium concentrations at 0- to 10-cmdepth in the control plots ranged from <1 to8.5 ppm. Soil concentrations of sodium rangedfrom 0.02 to 50.2 ppm on the plots with plantinjury. At 13 sites, soil sodium concentrationswere greater than the control on one or both of theinjury plots and were greater than 17 ppm on 7 ofthe sites. Two of the sites that were adjacent to afour-lane highway had soil sodium levels as highas 43 ppm.

The chloride content of the needles fromthe injury plots was elevated up to 3.4 times thatfound in the control needle samples. The sodiumcontent of needles from the injury plots waselevated up to 206 times that found in the controlneedles.

Sodium tissue levels were not elevated onthe two sites where sodium chloride was not usedfor winter de-icing, but tissue chloride levels were2–2.4 times greater than in the control plots on atleast one of the injury plots. Plant injury occurredless than 10 m from the roadside on these sites,both inside and outside the highway right-of-way.On one site, soil chloride was elevated to225 ppm, while soil sodium was not elevated.The use of calcium chloride or magnesiumchloride for dust suppression as part of summermaintenance programs was implicated in theinjury of trees on these two sites.

In this study, sodium chloride used for de-icing and calcium chloride or magnesiumchloride used for summer dust control werestrongly implicated as major contributors to theobserved roadside tree injury and decline on 16 ofthe sites. This conclusion is based on plant injurysymptoms, the pattern of injury and elevatedsodium and chloride concentrations in soil andvegetation.

3.5.3.7 Concentrations in soil and woodyfoliage adjacent to a two-lane highwayin British Columbia

Soilcon Laboratories Ltd. (1995) investigated thelevels of salt in soil and plant tissue with respectto plant injury levels along a 2-km section of


Loon Lake Road, a two-lane highway northwestof Kamloops in British Columbia. The dominantforest species on the site were ponderosa pine andDouglas-fir (Pseudotsuga menziesii).

The individual rock salt application rateswere 22, 88 or 153 kg per two-lane-kilometre,depending on whether low, average or heavyapplications were applied. The total salt loadingover the winter was 1561 kg per two-lane-kilometre, consisting of 1031 kg applied as rocksalt and 530 kg applied as sand with 5% salt.Salt or sand applications were made fromNovember 15, 1993, to February 23, 1994.

Nine sites were evaluated within a 2-kmsection along Loon Lake Road, where treesdisplayed low, moderate or severe salt injurysymptoms. Trees that fell in the moderate andsevere injury classes were located from 3 to 8.4 mfrom the road, both within and outside thehighway right-of-way. In comparison, trees thatfell in the low injury class were located 10–20 mfrom the road outside of the highway right-of-way.

Soil concentrations of chloride andsodium were sampled on the nine sites inDecember 1993 and April 1994 and comparedwith control values sampled in April 1991. Onsites with moderate and severe tree injury, soilconcentrations of chloride measured in the 0- to10-cm layer were 31.3 ppm chloride (range84.5–4.1 ppm) in December. By April, theconcentrations fell to 5.12 ppm chloride (range12.6–1.6 ppm). On sites with low tree injury, soilconcentrations of chloride measured in the 0- to10-cm layer were 7.2 ppm chloride (range23.5–1.4 ppm) in December. By April, theconcentrations fell to 2.9 ppm chloride (range4.5–1.3 ppm). This compared with backgroundlevels of 3.8 ppm chloride (range 7.5–0.2 ppm).

On sites with moderate and severe treeinjury, soil concentrations of sodium measuredin the 0- to 10-cm layer were 15.9 ppm (range28.2–1.9 ppm) in December. By April, theconcentrations fell to 9.1 ppm sodium (range30.7–1.0 ppm). On sites with low tree injury, soil

concentrations of sodium measured in the 0- to10-cm layer were 4.7 ppm (range 11.7–5.6 ppm)in December. By April, the concentrations fellto 1.2 ppm sodium (range 1.8–0.5 ppm). Thiscompared with background levels of 2.3 ppmsodium (range 2.7–2.0 ppm).

Soil concentrations of chloride andsodium were higher at sites with plants in themoderate- and severe-injury class than at siteswith plants in the low-injury class in bothDecember and April. Soil concentrations wereapproaching background levels by the Aprilsampling period. It was likely that sodium andchloride had leached from the soil by April byrainfall, since the last salt application was inFebruary.

Over the December to April samplingperiod, foliar chloride concentrations ranged from110 to 560 ppm in trees in the low-injury class,from 110 to 3180 ppm in trees in the moderate-injury class and from 870 to 5590 ppm in trees inthe severe-injury class.

During the same period, foliar sodiumconcentrations ranged from 0.5 to 70 ppm in treesin the low-injury class, from 0.5 to 190 ppm intrees in the moderate-injury class and from 30 to920 ppm in trees in the severe-injury class.

3.5.3.8 Concentrations in woody foliagedirectly adjacent to a two-lane highwayin Connecticut

Button and Peaslee (1966) reported foliarconcentrations of chloride and sodium inmature (about 70-year-old) sugar maple (Acersaccharum) leaves. The trees were growing close(from 1 to 4 m) to Route 17 between Middletownand Durham, a two-lane highway in Connecticut,likely within the highway right-of-way. Theresearch is relevant to Canadian conditions sincesugar maple is a native forest species, usedfor maple syrup production and planted as alandscape tree. This study also provides anexample of injury due to soil exposure to roadsalt via runoff.


Although salt load is not provided in thepaper, the authors indicate that 63 000 tonnesof rock salt and 2250 tonnes of calcium chloridewere used annually on 7180 km of highwayin Connecticut. This converts to an annualapplication of 9850 kg rock salt/km and 350 kgcalcium chloride/km.

The trees growing on the west side ofthe highway had road surface drainage flowingdirectly towards the trees. The trees growing onthe east side of the highway had drainage divertedby a curb that had been installed during thesummer prior to the study. Foliar concentrationsof chloride at the different sampling periodsaveraged 5200–6700 ppm on the west side of theroad. These were significantly higher than theconcentrations found on the east side of the road,averaging 1600–5400 ppm. This compared withlevels ranging from 200 to 1400 ppm chloride incontrol trees of the same species and age, locatedaway from a road. Foliar sodium levels rangedfrom an average of 1100–1600 ppm on the westside of the highway compared with non-detectable to 200 ppm on the east side of thehighway and 0 ppm in the control trees.

The higher concentrations of chlorideand sodium found in the foliage of the treeson the west side of the highway with overlandsurface flow were correlated with moderate tosevere injury symptoms in the trees on the westside of the road. This compared with no or lightinjury symptoms in the trees on the east side ofthe road and no injury in the control plants.

3.5.3.9 Concentrations in soil and woody andherbaceous foliage along roadsides inMassachusetts

Soil and vegetation samples were taken atdifferent distances along numerous highways ininland (central and western) Massachusetts duringJune 2000 (Barker, 2000; Mills and Barker, 2001).The research is relevant to Canadian conditions,since most of the sampled species are native ornaturalized species in Canada and since theapplication rates used in Massachusetts in 1999were comparable to application rates in central

and eastern Canada. Application rates inMassachusetts were 240 lb./lane-mile (136 kg pertwo-lane-kilometre), compared with rates inOntario ranging from 130 to 170 kg per two-lane-kilometre.

Samples of eight of the woody plantspecies and a range of herbaceous species weretaken at 1.5–12 m from different highways andwere analysed for sodium concentrations. Plantinjury was observed within 4–9 m from thehighways and was correlated with increasedsodium concentrations in the plant tissues inmultiple pine species (Pinus spp.) and sumac(Rhus sp.). In pine, the mean foliar concentrationof sodium was 2139 ppm in damaged foliagecompared with 28 ppm in a sample collected froman off-roadside site. In sumac, the mean foliarconcentration of sodium was 209 ppm indamaged foliage compared with 177 ppm inhealthy foliage. There was not a great differencefound in sodium concentrations in a single sampleof poplar (Populus sp.). The sodium concentrationwas 310 ppm in damaged foliage compared with338 ppm in healthy foliage.

The foliar sodium concentrations inplants were elevated between 3 and 7.6 m fromthe road and continued to fall off up to 12 mfrom the road. Extractable sodium measured inthe roadside soils was elevated within 6.1 mfrom the road, as high as 154 ppm, falling offto 26 ppm at 12 m from the road. Soil pH waselevated within 3 m from the road, above 7,compared with less than 6.5 at greater than 6 mfrom the road. There was not a consistent patternto electrical conductivity, which ranged from0.12 dS/m at 9.1 m to 0.30 dS/m at 12 m fromthe road.

3.5.3.10 Concentrations in soil and woodyplants downslope from a four-lanehighway in British Columbia

A site investigation conducted in 1989 in BoitanioPark in British Columbia investigated the sourcesof tree injury and decline that were evident ontrees in the east section of the park near Highway97 (Van Barneveld and Louie, 1990). Boitanio


Park is located in Williams Lake, southwest of afour-lane section of Highway 97. The dominanttree species in the park was rocky mountainDouglas-fir (Pseudotsuga menziesii var. glauca).

The area of concern was near the highwayand sloped towards the southwest with a gradeof 5–15%. Two culverts were located alongthe highway above the park, so the water waschannelled through the park. The highway had atraffic volume of approximately 12 000 vehiclesper day in this area, and sodium chloride de-icingsalt was applied to the highway at a rate of60–130 kg salt per lane-kilometre.

Trees along the drainage channels andadjacent to the highway displayed symptoms ofsevere injury and rapid decline that were reportedin the spring of 1989. Preliminary soil samplesfound high levels of sodium in the soil, which ledto a comprehensive sampling and field analysis ofvegetation and soil within the park in July 1989.Samples were taken using an incomplete gridpattern running away from the highway, withsampling locations every 50 m along the grid.Vegetation was evaluated for plant injurysymptoms and insect and disease status. Foliartissue was sampled and analysed for chloride andsodium content. Soil samples were taken fromsuccessive soil layers of 25 cm down to 150 cm.

Plant injury was evaluated using a ratingscale of low, moderate and severe. Trees in thelow-injury class had less than 10% discolorationand dead branches with healthy leaders.Trees in the moderate-injury class had 10–50%discoloration, defoliation and some dead brancheswith leaders alive. Trees in the severe-injuryclass were dead or had greater than 50% foliardiscoloration, defoliation and dead tops.

The pattern of tree injury was associatedwith roadside locations and with the drainagepatterns from the two culverts along drainagechannels within the park. The injury was mostsevere along these drainage pathways.

Soil concentrations of sodium andchloride were elevated on sites with severe or

moderate injury compared with those with lowinjury. The average soil concentration of sodiumwas 108.7 ppm on the severe-injury sites,37.9 ppm on the moderate-injury sites and29.0 ppm on the low-injury sites. The average soilconcentration of chloride was 52.1 ppm on thesevere-injury sites, 50.0 ppm on the moderate-injury sites and 20.3 ppm on the low-injury sites.Severe-injury sites had soil sodium concentrationsthat were 3.7 times the concentration in low-injurysites and soil chloride concentrations that were2.6 times the concentration in low-injury sites.

Foliar concentrations of sodium andchloride were higher in the plants with a severeor moderate injury rating compared with theplants with a low injury rating. The averageconcentration of sodium in the 0- to 3-yearneedles was 113 ppm on the severe-injury sites,54.8 ppm on the moderate-injury sites and59.7 ppm on the low-injury sites. This comparedwith average concentrations of sodium in the 4-to 6-year needles of 208 ppm on the severe-injurysites, 72.7 ppm on the moderate-injury sites and47.9 ppm on the low-injury sites.

The average concentration of chloridein the 0- to 3-year needles was 3560 ppm on thesevere-injury sites, 2630 ppm on the moderate-injury sites and 572 ppm on the low-injury sites.A similar pattern was observed in the averageconcentrations of chloride in the 4- to 6-yearneedles: 4180 ppm on the severe-injury sites,2667 ppm on the moderate-injury sites and1060 ppm on the low-injury sites.

Other factors such as recent drought,presence of other contaminants in the runoff,insects and recent brush removal may haveaggravated the decline of the trees. However, itwas determined that these factors were not theprimary cause of the decline in the park. This sitehas a dry climate, with annual May to Augustprecipitation of 158 mm of rain. This low rainfallamount and recent drought would have aggravatedthe problem, in that elevated levels of sodium andchloride in the soil would not be leached out.


The injury symptoms on the Douglas-firin this site were consistent with salt injury andsupported by the pattern of more severe plantinjury associated with roadside locations and withthe drainage patterns from the two culverts alongdrainage channels within the park. In these areas,elevated soil and tissue concentrations of sodiumand chloride were correlated with more severeinjury symptoms in the trees.

Injury due to highway runoff followinga drainage course into forest stands has beenreported by Kliejunas et al. (1989). Theyrecorded conifer needle tip dieback typical of saltabsorption from the soil in forest stands 91–122 mfrom Highway 89 at Emerald Bay and fromInterstate 80 at Donner Summit in the Lake Tahoearea of Nevada. High levels of sodium andchloride (5800 ppm sodium and 13 700 ppmchloride) were found in injured white fir (Abiesconcolor) and Jeffrey pine (Pinus jeffreyi) foliageat Donner Summit located within a drainagechannel 76 m downslope from Interstate 80.

3.5.3.11 Concentrations in soil and woodyfoliage adjacent to a patrol yard

Racette and Griffin (1989) reported on theinvestigation of the cause of severe woody plantinjury in the vicinity of a highway maintenancepatrol yard east of Kenora, Ontario. Trees inthe area behind the sand dome and salt storagesheds and along a drainage channel that receiveddrainage from the highway and patrol yard areahad severe injury symptoms.

In a nearby sand and gravel storagearea, salt-treated sand had been stored on atleast one occasion. In this area, jack pine (Pinusbanksiana), spruce (Picea sp.), balsam fir (Abiesbalsamea) and poplar (Populus sp.) trees haddied a number of years before. Existingvegetation was damaged along a drainage coursein this area. Grasses had been killed, andtrembling aspen (Populus tremuloides) hadsymptoms of severe marginal necrosis.

Plant tissue levels of sodium andchloride were sampled in September 1988 in

the areas where the injury was evident. Foliarconcentrations of chloride ranged from 1800 ppm(serviceberry [Amelanchier sp.] foliage) to22 000 ppm (white birch [Betula papyrifera]foliage). Foliar concentrations of sodium rangedfrom 10 ppm (black spruce [Picea mariana]foliage) to 86 ppm (serviceberry foliage).Background foliar concentrations of chlorideranged from less than 20 ppm (serviceberryand white birch) to 1900 ppm (trembling aspen).Background foliar concentrations of sodiumranged from 7 ppm (black spruce) to 19 ppm(serviceberry).

Twig concentrations of sodium andchloride were also elevated in these areas.Twig concentrations of chloride ranged from900 ppm (trembling aspen) to 9100 ppm (blackspruce). Twig concentrations of sodium rangedfrom 9 ppm (trembling aspen) to 78 ppm (blackspruce). Background twig concentrations ofchloride ranged from less than 20 ppm (tremblingaspen) to 200 ppm (black spruce). Backgroundtwig concentrations of sodium ranged from 7 to17 ppm (both in black spruce).

Surface soil concentrations of sodium andchloride at these two sites, taken at 0- to 5-cmdepth, were elevated compared with a control siteremote from the area. The soil concentration ofchloride ranged from 780 to 1200 ppm, and thesoil concentration of sodium ranged from 370to 680 ppm. Background concentrations were15 ppm chloride and 91 ppm sodium.

Changes were initiated in the winter of1988–89 to reduce losses of salt from the patrolyard operations. The area and severity of thevegetation injury had decreased by August 1989.The concentrations of sodium and chloride in thefoliage also declined, although levels were stillelevated in comparison with the control.

3.5.4 Changes in application rates andcurrent injury in Canada

The Canadian field studies discussed in Section3.5.3 and the environmental concentrations


discussed in Section 3.5.1 use data on plant orsoil chloride and sodium concentrations and plantinjury data from field sampling done between1980 and 1995. Application rates of sodiumchloride have decreased by 20–30% in somejurisdictions since the 1980s; however, totalannual loadings have not declined (Morin andPerchanok, 2000).

The total amount applied during a winter(annual loading) and the total amount of soil oraerial deposition are the critical values for plants(Hofstra and Lumis, 1975). The total amount ofsalt applied during a season will be higher inwinters with more snow events, with more dayswith snowfall, with higher snowfall amounts perevent or with more freezing rain events. As well,plants are exposed to more aerial salt spray inyears with more salt applications, more early-and late-season storms, and freezing conditionsrequiring salt application that begin earlier thanusual in the fall and extend later into spring.

Current data from two municipalities inOntario and Alberta demonstrate the variation intotal application of road salts per winter season.In the North District of the Region of Peel inOntario, total application amounts per year perlane-kilometre varied by 22% over 1995–2000,from 22 to 27 tonnes per lane-kilometre (Zidar,2000). Over the years of 1996–97 to 1998–99,the City of Edmonton practised “sensible salting”and “best management practices” to minimize theeffect of road salt on the environment. They foundthat total road salt usage during this period rangedfrom 11 000 tonnes sodium chloride plus69 400 L calcium chloride to 21 300 tonnessodium chloride plus 40 900 L calcium chloride.This represents a variation of more than 90% forsodium chloride and 70% for calcium chloride(Neehall, 2000).

In other scenarios, with early- or late-season storms that occur when the frost is outof the ground, more salty runoff would penetratethe soil around plant roots rather than run offover frozen ground (Headley and Bassuk, 1991).Spring salt applications would also result in aerialsalt deposition on expanding woody buds and

succulent young shoots as well as soil exposure ofgerminating seedlings. In years with dry springs,salt deposition on woody plant stems and leavesmay not be washed off by rainfall (Hofstra et al.,1979; Simini and Leone, 1986), and salt presentin soil may not be leached as much as in yearswith wetter springs. All these variables mean thatsalt exposure of plants via soil or aerial depositiongoes up and down depending on climaticconditions each year (Hofstra et al., 1979).

Salt injury on roadside plants remainsevident along rural and urban roadside conditionsin Canada. The City of Edmonton has observednegative impacts attributable to salt on roadsidegrass, shrub and tree vegetation, but notes thatthere are few alternatives to road salt to achievesafe winter driving conditions (Neehall, 2000).Although city staff practise “sensible salting” and“best management practices” to minimize theeffect of road salt on the environment, they havefound that soil chloride concentrations areexacerbated in years with drought conditions thatreduce soil moisture and flushing of soluble saltsfrom the soil. Negative impacts on roadside treesin Edmonton varied in severity from year to yearand were more severe in drought years.

Salt injury to trees has also been observedin rural areas of Ontario during the 2000 growingseason — for example, along County Road 124in Dufferin County (Figure 3.4 in Cain et al.,2001, and accompanying discussion) and alongHighway 401. The Region of Peel in Ontarioevaluated salt damage to trees along roads in thenorthern part of their district on roads maintainedby the Region of Peel and the Ontario Ministryof Transportation (Thompson, 2000; Zidar, 2000).Injury seemed to be more severe than usualfollowing the 1999–2000 winter. Salt injury wasobserved most often on sensitive tree species,such as eastern white cedar (Thuja occidentalis),red pine (Pinus resinosa) and Scots pine (Pinussylvestris). Other contributing factors includedsummer drought conditions during the previoustwo summers, which contributed to tree stress,and the occurrence of mild winter days over the1999–2000 winter, which possibly resulted inwet road conditions, contributing to salt spray.


The occurrence of injury also depended on thethickness of forest along the roads, tree distancefrom the road, soil type and slope, speed of roadand volume of traffic.

3.5.5 Estimate of area of potential injury

The zone of impact of road salts on susceptibleplants extends generally up to 80 m from the edgeof multi-lane highways or freeways and up to35 m from two-lane highways. Using highwaylength data, one could estimate the area of landin Canada where sensitive vegetation could beaffected by road salts. Using data published byTransport Canada (2001), the length of freewaysin Canada in 1995 was 16 571 two-lane-equivalent kilometres. This number could bedivided by 2 to approximate the linear lengthof highway. The potential area of land wheresensitive plants would be affected up to 80 mfrom the edge of the highway would be132 568 ha. In the same year, the length of urbanand rural paved roads in Canada was 301 348two-lane-equivalent kilometres, providing a linearlength of the same amount. Along these roads,the potential area of land where sensitive plantswould be affected up to 35 m from the edge ofthe roads would be 2 109 436 ha.

In total, sensitive vegetation couldbe affected by road salts on an estimated2 242 000 ha of land along Canadian roadsides,using road length data for 1995. Not all of theseroads and highways are salted, however, so theactual amount of potentially affected land wouldbe substantially less.

3.5.6 Risk assessment summary andconclusions

Sodium chloride applied as a de-icer can have aharmful effect on plants growing along roads orin areas that have localized contamination dueto runoff with elevated salt levels. Elevated soillevels of sodium and chloride or aerial dispersionof sodium and chloride results in severereductions in flowering and fruiting of sensitiveplant species; severe foliar, shoot and root injury;

growth reductions; and reductions in germinationand seedling establishment. The species affectedinclude native grass species; wetland species,such as sphagnum moss and sedges; and woodyspecies, such as maple, pine, Douglas-fir,dogwood, peach and plum.

Severe effects may result in plantmortality, while repeated (chronic) exposure ofperennial plants to elevated salt levels, year afteryear, results in reduced plant growth and death ofshoot and leaf tissue. Repeated stress on plantsreduces plant vigour and the ability to deal withinsects, disease and abiotic stresses. These effectscan lead to loss of plants within a community aswell as impacts on other organisms that rely onplants for food and cover.

Changes in plant communities in responseto elevated levels of de-icing salts have beenrecorded in areas impacted by road salt runoffand aerial dispersion. Halophytic species, such ascattails and common reed-grass, readily invadeareas impacted by salt, leading to the changes inoccurrence and diversity of salt-sensitive species.The composition of non-salt-tolerant communitieschanges due to the loss of sensitive species andreplacement by salt-tolerant species.

The zone of impact is usually linear,along roads and highways or other areas wheresodium chloride or other chloride salts are appliedfor de-icing or dust control. Salt injury alsooccurs along drainage courses that direct saltyrunoff from roads and salt handling facilities.Aerial dispersal of sodium and chloride occursalong roads during the winter months when de-icing is required and impacts mainly woodyplants. The impact of aerial dispersion of sodiumand chloride extends up to about 80 m from theedge of multi-lane highways and up to about 35 mof two-lane highways where de-icing salts areused. The degree and distance of dispersiondepend on the salt load, traffic volume andvelocity, local topography and prevailing winds.

Impacts of road salts on forest speciesadjacent to roadsides have been documented incentral and southern British Columbia. In these


areas, the zone of impact was in the first 10–15 mfrom two-lane highways. Adjacent to a four-lanehighway in British Columbia, the zone of severeinjury of forest trees covered the first 50 m fromthe highway. Areas of high forest density inCanada that experience medium to high saltloadings (based on average annual loadings, bymaintenance district) include Vancouver Island,southwestern and central Alberta, central Ontarioand parts of southern Ontario, southern andcentral Quebec, and areas in New Brunswick andNova Scotia. Fifty-three percent of the primaryspecies of forest trees in Canada have been ratedas sensitive to road salt.

Fruit crops experience severe plant injuryand reduced fruit production in a zone up to50–80 m from four-lane highways in southernOntario and about 35 m from two-lane highwaysin Nova Scotia. Sixty-five percent of 32 cultivarsor varieties of fruit trees, vines or shrubs havebeen rated as sensitive to salt. The sensitivity ofthese species provides an indication of sensitivityof native fruiting plants that occur in naturalcommunities along roadsides.

Sensitive landscape plantings and nurserycrops suffer severe plant injury and plant deathdue to elevated soil levels of road salt and saltspray. From 30 to 40% of common landscapespecies used in Ontario and throughout Canadahave been rated as sensitive to road salts.

3.6 Terrestrial wildlife

The effects of laying down a road network onvertebrate wildlife, both positive and negative,have been the subject of numerous scientificreviews (e.g., Bennett, 1991), and these aspectswill not be covered here. Impacts specific tothe use of road salts may be mediated through areduction in plant cover or community shifts asdescribed in Section 3.5. Under some conditions,especially where habitat is limiting, road vergesare valuable, both as breeding habitat for wildlifeand as a corridor between suitable habitat patches(e.g., Oetting and Cassell, 1971; Drake and

Kirchner, 1987; Bennett, 1991; Camp and Best,1993a,b). Efforts are currently under wayin several jurisdictions to improve roadsidemanagement in order to increase the value ofthis habitat. Road salts can have an adverse effecton these management efforts. Also, concernsabout the impact of road salts on species withan aquatic life stage such as frogs and otheramphibians were covered in Section 3.3. Thissection addresses the direct toxicological effectsof road salts on terrestrial wildlife, namely birdsand mammals.

3.6.1 Exposure characterization

Sodium is an indispensable component of thephysiological processes of all vertebrates, but itis toxic when taken in excess. Typically, surplussodium is removed by an increase in glomerularfiltration rate and a decrease in the percentage ofsodium reabsorbed by the kidneys. Terrestrialbirds, especially herbivorous and granivorousspecies, are more likely to be salt deficient andare poorly equipped to deal with excess sodium.Any animal attempting to satisfy a “salt hunger”generally overshoots its deficit by a substantialamount, and the increase persists for some timeafter the deficit is replaced (Schulkin, 1991).Overconsumption of salt is in turn regulated viaa feedback mechanism: the thirst response. Theabsence of a source of water therefore increasesthe toxicity of ingested salt (see Section 3.6.2.1).Sodium-deficient wildlife are known to travelgreat distances to obtain and ingest salt and tovisit roads where salt has been applied (Schulkin,1991). One concern raised by this assessment isthat the “footprint” of a salted road may be quitelarge when highly mobile wildlife species areattracted from great distances.

3.6.1.1 Mammals

The most visible interaction between road saltand wildlife concerns large ungulates, such asmoose (Alces alces), mule deer (Odocoileushemionus), white-tailed deer (Odocoileusvirginianus), elk (Cervus canadensis) and bighornsheep (Ovis canadensis). Although there may


be other reasons for mammalian wildlife to visitroadsides, the use of road salt has repeatedlybeen identified as a major factor contributing tocar strikes. In Ontario, the peak of moose–vehicleaccidents corresponds to the period of highestsodium hunger, not the period of highest vehiculartraffic (Fraser and Thomas, 1982). Most sightingsof moose and about half of all moose–vehicleaccidents occurred at or very near salt pools(Fraser and Thomas, 1982). A technical reportpublished by the Ontario Ministry ofTransportation states that deer and moosedrinking salty water tend to lose their fear ofhumans and vehicles and are prone to bolt,sometimes in the path of a vehicle, rather thanmove away as they normally would (Jones et al.,1986). In one Quebec study, moose visitationto roadside ponds increased with sodium andcalcium levels, and the number of car strikeswas higher near ponds receiving higher moosevisitation (Grenier, 1973). Overall, there weretwice as many moose killed per kilometre of roadwhen roadside pools were present than wherethere were no pools. Miller and Litvaitis (1992)found that radio-collared moose in northern NewHampshire extended their home ranges in orderto include roadside pools heavily contaminatedby salt. Road salt attraction has been identifiedas a main reason for kills of bighorn sheep and aminor reason for kills of elk in Jasper NationalPark (Bradford, 1988). The role that road saltcould play in the mortality of other wildlifespecies such as small mammals commonly killedby traffic has not, to our knowledge, been thesubject of formal study, but another OntarioMinistry of Transportation report states thatwoodchucks (Marmota monax), snowshoe hares(Lepus americanus) and porcupines (Erethizondorsatum) are frequently observed feeding onroadside salt (Hubbs and Boonstra, 1995). Trainerand Karstad (1960) reported a case of poisoningby road salt in eastern cottontails (Sylvilagusfloridanus).

3.6.1.2 Birds

Twelve published reports of bird kills associatedwith salted roads were identified (Brownlee et al.,2000). Two incidents were formally diagnosedas salt poisoning (Trainer and Karstad, 1960;Martineau and Lair, 1995), and observations ofaberrant behaviour suggestive of toxicosis weremade in several other cases. Sodium chloride wasusually identified as the substance to which birdswere exposed, although in one case the road saltapplied was calcium chloride (Meade, 1942).Geographically, these kills occurred at several siteswithin the Canadian and U.S. snow belt from 1942to 1999. One report originated from Germany. Atleast two reported kills were large, involving morethan a thousand birds each. Given the difficulty offinding avian carcasses (Mineau and Collins,1988), the high rate of scavenging reported fromsites where birds are routinely killed at roadside(Benkman, 1998; Woods, 1998) and the low rateof reporting for wildlife mortality events ingeneral, the number of reports of bird mortalityassociated with the use of road salt is significantand suggests that kills are more widespread andfrequent than indicated by documented reportsalone. Supporting this view is the fact that otherresearchers and biologists, when contacted, werevery familiar with kills of birds on saltedhighways, even though their observations had notbeen published. In one location (Mount RevelstokePark), mortality of siskins and other winter finches(see below) has been seen frequently enough overthe last 25 years for the birds to be called “grillbirds” by the local inhabitants, in reference to theirpropensity to be collected by the front end ofmoving vehicles (Woods, 1998). The kills oftenoccur in sections of the highway with curves orwhere there are many cracks and crevasses in thepavement where salt can accumulate, althoughthey may occur anywhere the roads have beensanded and salted. Baker (1965) stated that theinhabitants of one Maine community “lookedupon this bird slaughter as a natural and everydayoccurrence.” Another researcher reported seeing“hundreds of dead crossbills that were hit byvehicles while on the road eating salt” (Benkman,1998).


Cardueline finches (family Fringillidae,subfamily Carduelinae: crossbills, grosbeaksand siskins) were involved in 11 out of 12 publishedincidents. All cardueline finches are protectedspecies listed under the Migratory Birds ConventionAct. This group of seed-eating birds rangesthroughout the boreal forest, moving to moresouthern latitudes when seed yield is low. The extentof these “invasions” is best visualized throughanimated computer-generated maps of yearlyabundance (see Cornell Laboratory of Ornithology,2000a,b). Although finches will eat deciduous seed,insects and berries, their primary diet, particularly inthe winter, consists of coniferous seeds. Theirattraction to salt is well known. Red crossbills(Loxia curvirostra) have been trapped using salt asbait (Dawson et al., 1965), and cardueline fincheswere found to prefer sodium over other mineralsregardless of grit size (Bennetts and Hutto, 1985).Fraser and Thomas (1982) identified three finchspecies — evening grosbeaks (Hesperiphonavespertina), purple finches (Carpodacus purpureus)and pine siskins (Carduelis pinus) — as the mainvisitors to roadside salt pools in northern Ontario.

The information from reported kills and thebehavioural/social characteristics of the carduelinefinches suggest that this is the group most at risk ofacute salt toxicity from ingesting road salt.However, because the birds tend to travel and foragein large flocks in the winter, there is undoubtedly abias for a higher detectability and reporting of kills.Trainer and Karstad (1960) reported bobwhite quail(Colinus virginianus), ring-necked pheasant(Phasianus colchicus) and feral dove (Columbalivia) as well as eastern cottontails being poisoned.

It is often assumed that salt is ingestedin order to fulfil a physiological need associatedwith a largely vegetarian diet. However, the takingof salt crystals as grit cannot be ruled out. Grit isused to aid in the grinding of food and to providesupplementary minerals to bird species with low-calcium diets (Gionfriddo and Best, 1995). Acrossspecies, grit size preferences are linearly related tobird size. The size of road salt particles from a saltmine in Ontario ranged from 0.6 to 9.5 mm(Canadian Salt Company Limited, 1991). Becausesalt particles applied to roads undergo a gradual size

reduction as they melt (Letts, 1999), salt particlesoverlap broadly with the preferred grit size for anysize of bird from the time of application to someundetermined time after application. The largevariation in the starting size of road salt particlesensures that a suitable size of grit will be availablefor an extended period.

3.6.2 Effects characterization

3.6.2.1 Acute toxicity of sodium chloride

Incidents of vertebrate salt toxicity in domestic orcaptive stock fall under several main categories:accidental overdose in feed beyond a level that canbe compensated for by drinking more water(Sandals, 1978; Swayne et al., 1986; Howelland Gumbrell, 1992; Wages et al., 1995; Khannaet al., 1997), exposure to saline drinking water(Franson et al., 1981), exposure to saline lakes(Windingstad et al., 1987; Meteyer et al., 1997) orthe provision of salt supplements with deprivationof water (Trueman and Clague, 1978; Scarratt et al.,1985). Reports of salt toxicosis in free-rangingvertebrate wildlife have been the result of exposureto road salt, drought, ice, hypersaline lakes anddischarge ponds from various industries (Baeten andDein, 1996).

Sodium chloride has an oral LD50 of3000 mg/kg-bw in rats and 4000 mg/kg-bw in mice(Bertram, 1997). Sodium chloride administered as asaturated water solution (approximately2100 mg/kg-bw) usually resulted in death within 24hours (Trainer and Karstad, 1960). Repeated smalldoses of salt produced no illness if water was notrestricted.

The Canadian Cooperative WildlifeHealth Centre in Saskatoon and the NationalWildlife Research Centre of Environment Canadarecently conducted a joint study on the acuteeffects of salt on wild-caught house sparrows(Passer domesticus) (Wickstrom et al., 2001).The “up-and-down” method was used in a pilotstudy to estimate the approximate lethal oral doseof granular sodium chloride in birds that were non-fasted but without access to water for 6 hourspost-dose. Results indicated an approximate


LD50 of 3000–3500 mg/kg-bw, which is similarto the rodent values. A tentative no-effect level(mortality) was estimated to be 2000 mg/kg-bwin this pilot study, although mortality was seenat 1500 mg/kg-bw in a subsequent phase ofthe study.

3.6.2.2 Sublethal effects of excess saltingestion in birds and mammals

Several authors reporting wildlife incidentsassociated with road salt made note ofbehavioural deficits in exposed birds. The mostcommon observation is that the birds appearedunusually fearless and could be approached easily(Meade, 1942; Trainer and Karstad, 1960; Baker,1965; Thiel, 1980; Smith, 1981; Woods, 1999).Others described the birds as “weak and slow”(Martineau and Lair, 1995) or appearing sick(Meade, 1942). Trainer and Karstad (1960)reported “depression, tremors, torticollis,retropulsion, and partial paralysis” for birds andmammals affected in one incident.

These signs are closely matchedby observations made in the course of laboratorystudies. Pheasants fed increased amountsof sodium in their mash exhibited signsof intoxication when the drinking waterwas restricted. These included depressionfollowed by excitement, tremors, torticollis,opisthotonos, retropulsion, completeincoordination and coma (Trainer and Karstad,1960). Not all signs were seen in the same animal.Frequent signs reported in poultry and in swineinclude salivation, diarrhea, a staggering gait,muscular spasms, twitching and prostration(Humphreys, 1978). In Wickstrom et al. (2001),following the pilot study reported above, housesparrows were dosed orally with granular sodiumchloride at 0, 500, 1500, 2500 or 3500 mg/kg-bw.Overt clinical signs were observed at 1500 mg/kg-bw or higher and included rapid onset(<30 minutes) of depression, ataxia and inabilityto fly or perch, with death in as little as 45minutes. Birds that survived for 6 hours usuallyrecovered. Plasma sodium concentrations above200 mmol/L were consistently associated withovert clinical signs. Lesions in the form of gizzard

edema were observed after 1 hour in most birdsdosed with 500 mg/kg-bw or higher, and brainstem vacuolation, also consistent with edema, wasseen in some birds showing clinical signs.

There is no record of moose or deersuffering from salt toxicosis, although thereare anecdotal observations that moose drinkingsalty water tend to lose their fear of humans andvehicles (Jones et al., 1992), in contrast to moosenot drinking salty water, which move awayfrom approaching humans. The only documentedcase of small mammals exhibiting signs ofsalt toxicosis is eastern cottontails, reportedduring a severe winter in Wisconsin (Trainerand Karstad, 1960).

3.6.3 Risk characterization

Mortality has been documented and is clearlymuch more frequent and widespread than isreflected by the published record. The risk tocardueline finches at least is significant. Locally,the risk to large mammals is also high. It is clearthat the mere presence of road salt contributes tovehicle collisions by increasing the attractivenessof roadsides. The main uncertainty in the riskcharacterization is whether salt toxicity increasesthe vulnerability of some species to vehiclecollisions. Also, if salt toxicity is occurring, itis relevant to ask whether wildlife, carduelinefinches in particular, could be killed by saltingestion, even in the absence of a vehicle strike.

3.6.3.1 Estimating the contribution of salttoxicity to vehicle strikes

The severe behavioural impairments (depression,tremors, torticollis, retropulsion and partialparalysis) observed at one site, as well as thefinding of elevated brain sodium levels at another,argue strongly for a contributory role of salttoxicity to roadside bird kills. The more commonobservation of unusual fearlessness documentedin both birds and mammals could be an earlysymptom of intoxication or the manifestationof animals under extreme salt hunger modifyingtheir usual behaviour towards potential danger.


Although overt signs of intoxication wereseen in the house sparrow at 1500 mg/kg-bw andhigher, the level of dosing at which a more subtlebehavioural impairment could be manifest is notknown. A lower limit for potential impairment,using the house sparrow as a model, wasestimated as follows.

The distribution of normal sodiumvalues in the plasma of the sparrows isvery narrow (mean 163.2 mmol/L, range158–168 mmol/L, n = 12). The assumption is thatadverse effects (such as reduced coordination andweakness) may begin when sodium homeostasisis breached and elevated sodium levels are beingdelivered to target organs. This level in the housesparrow model was determined to be 266 mg/kg-bw (Brownlee et al., 2000).

To estimate exposure, we assume thatsalt crystals are most likely to be picked upwhole, as are grit particles. A finch trying tosatisfy a salt hunger might pick up a granule of asize that it is “comfortable” with. The maximumrecorded size for grit in the house sparrow is2.4 mm (Gionfriddo and Best, 1995). Table 24shows the number of particles that need to beingested by birds in order to attain specific CTVs.This calculation suggests that birds could besublethally impaired following ingestion of asingle large salt particle and could be killed byas few as two large particles.

On the other hand, the average size ofgrit ingested by the house sparrow was 0.5 mm(Gionfriddo and Best, 1995). Assuming birdswere intent on picking up salt particles ofthis “preferred” size for digestion, the abovecalculations were repeated for average-sizedgrit. In fact, this average grit size is likely anunderestimate, because it is based on “worn” gritextracted from sparrow gizzards. The average alsoencompasses very fine mineral particles, whichwere probably ingested accidentally along withfood. These results are given in Table 25.Although the number of particles that need to beingested to reach CTVs is much higher than thatcalculated in Table 24, this number of particlescan easily be ingested based on a comparison withaverage grit counts. Therefore, birds in need ofreplenishing their grit supply may easily achievedoses that would be toxicologically relevant.

3.6.3.2 Conclusions and discussion ofuncertainty

Road salt is widely acknowledged to be animportant factor in attracting large mammalsto roads, and it does appear to increase thefrequency of vehicular collisions.

It is clear that, even without road salt,some birds would be coming to highways for grit,clay, road-killed prey, etc. However, observationsmade at roadside and the biology of carduelinefinches especially suggest that road salt is an


TABLE 24 Calculation of the number of particles of salt that need to be ingested in order to reach CTVsassuming a model 28-g house sparrow consuming particles at the upper end of its knownpreference range

CTV Rationale No. of 2.4-mm-diameter No. of 2.4-mm-diameter(mg/kg-bw) spherical salt particles to cubic salt particles to

attain CTV in 28-g bird attain CTV in 28-g bird266 Breach of homeostasis 0.47 0.25500 Edematous lesions in the gizzard 0.88 0.47

1 500 Overt signs of toxicity; mortality first recorded 2.6 1.4

3 000 Approximate median lethal dose 5.2 2.8

important attractant. The evidence to date suggeststhat road salt increases the vulnerability of birds tocar strikes by causing impairment. It is commonsense that birds at the roadside that are “weak andslow” (Martineau and Lair, 1995), that appear sick(Meade, 1942) or that show “depression, tremors,torticollis, retropulsion, and partial paralysis”(Trainer and Karstad, 1960) have a diminishedcapacity to avoid moving vehicles. The biologicaleffects of road salts differentiate them from, forexample, sand, which is known to attract birds insearch of grit to highways but which does notcause impairment. Furthermore, intakecalculations suggest that road salt may poisonsome birds directly, especially when water is notfreely available during severe winters. Birds havebeen observed eating snow, catching snowflakesduring a storm and chipping ice to obtain water(Wolfe, 1996). Oeser (1977) reported thatcrossbills were eating snow in the reportedGerman kill incident. There could be negativeenergetic consequences of snow ingestion inresponse to salt overconsumption. Furthermore,melted snow on the roads may be the mostobtainable source of water, but, depending onthe concentration of salt in the snowmelt,consumption of such water may increase saltingestion rather than alleviate salt toxicosis.

A remaining uncertainty is whether thehouse sparrow is a suitable toxicological modelfor cardueline finches and other native birds.The house sparrow has Middle Eastern originsand may therefore be genetically predisposed tobe more salt tolerant than cardueline finches.On the other hand, consumption rates of gritmay be high in house sparrows relative to otherbird species (Gionfriddo and Best, 1995).The scenarios presented above assumed aconsumption of salt based on grit need (Table25) as well as on optimizing salt intake byingesting larger particles at the limit of toleratedgrit size (Table 24). Regardless of the exactparticle size taken, salt-deprived birds couldingest toxicologically relevant doses beforefeedback mechanisms reduce the desire for salt.

The overlap between the distributionof winter finches and the Canadian road systemreceiving road salt was determined to be extensivewhen mapped by Brownlee et al. (2000). This isespecially true during “invasion” years whenthe finches descend out of the boreal forest. It isreasonable to assume that finches will be attractedmore to some types of roads than to others. Forexample, a busy multi-lane highway servicinga large metropolitan area may not offer many


TABLE 25 Calculation of the number of particles of salt that need to be ingested in order to reachCTVs assuming a model 28-g house sparrow consuming particles of average size

CTV Rationale No. of 0.5-mm-diameter No. of salt particles as (mg/kg-bw) salt particles to attain CTV proportion (%) of average

gizzard grit count (580 particles as in Gionfriddo and Best, 1995)

Spherical Cubic Spherical Cubic particles particles particles particles

266 Breach of 52 27 9.0 4.7homeostasis

500 Edematous lesions 98 51 17 8.8in the gizzard

1500 Overt signs of 294 154 51 27toxicity; mortality first recorded

3000 Approximate median 587 307 101 53lethal dose

opportunities for a flock of birds to alight,regardless of how much salt is available. However,the data needed to assess the probability that a flockwill land on any given roadway or the proportion ofthose visits that result in kills (whether fatalattraction, toxicosis-induced car strikes or lethalingestion) are not available. The length of time theapplied salt is retained on or near the road surface(e.g., in the soil), where it could create an artificial“lick” attractive to both birds and mammals evenafter granules have disappeared, is highly stochasticand difficult to assess, but undoubtedly important.

Despite these uncertainties, the scientificdata available to date suggest that the number ofroad kills of federally protected migratory birdspecies (especially the cardueline finches) andthe contribution of road salt to this mortalityhave been underestimated by wildlife managersand transport officials. On the other hand, theimportance of road salt in increasing thenumber of collisions between large mammalsand motor vehicles has long been recognizedand acknowledged by transport officials.

3.7 Ferrocyanides

This section is based on a detailed review offerrocyanides presented by Letts (2000a).

3.7.1 Exposure characterization

Sodium ferrocyanide dissolves in water torelease the stable ferrocyanide anion (Fe(CN)6

4–),which is not volatile and resists further breakupunless illuminated. The anion is fairly mobile ingroundwater but reacts readily with iron to form theprecipitate ferric ferrocyanide, a very stable, non-toxic, relatively immobile compound, which is amajor removal pathway of ferrocyanide fromgroundwater. Under favourable conditions,including non-illumination, the half-life offerrocyanide is approximately 2.5 years (Higgs,1992), while that of ferric ferrocyanide can extendto a thousand years (Meeussen et al., 1992a).

Two characteristics of the ferrocyanideanion are paramount. It is of extremely lowtoxicity in the complexed form (Rader et al.,1993), but in solution under direct light itphotodecomposes, liberating free cyanide, whichrapidly reacts to produce the highly toxic andvolatile hydrogen cyanide (HCN). Outside ofsolution, photolysis of ferrocyanide is slight(American Cyanamid Company, 1953).

Photolysis of the ferrocyanide anion isa well-studied phenomenon (Broderius, 1973;Broderius and Smith, 1980; Otake et al., 1982;Clark et al., 1984; Meeussen et al., 1992a).It can liberate as many as five of the six cyanideions associated with the parent compound.The reaction is a function of the initialconcentration, intensity of illumination, pHand temperature, among other factors.

Free cyanide in the environment issubject to a number of attenuation processes.It is readily chelated with numerous transitionmetals to form more stable, less toxic materials(Chatwin, 1990). The presence of clay ororganic matter removes free cyanide by surfaceadsorption (Theis and West, 1986). Additionally,it is oxidized to the relatively innocuouscyanates and thiocyanates in the presence ofsulphur.

3.7.1.1 Soil

3.7.1.1.1 Non-illuminated ferrocyanide

The largest portion of research covering thebehaviour of non-illuminated ferrocyanides insoil has focused on waste compounds fromformer manufactured gas plant sites. Studiesat such sites indicate that the presence offerrocyanide is tightly restricted around the sourcesite (Effenberger, 1964; Parker and Mather, 1979).This low mobility is apparently due to thepresence of sulphur and transition metals inthe soil. Iron under pH conditions of about4–7 combines with the ferrocyanide anion to


produce the sparingly soluble ferricferrocyanides known as prussian blue (Meeussenet al., 1992b). This compound is evident in theblue coloration of the soil at many abandonedmanufactured gas plant sites. In alkaline soils (pHabove 7), ferrocyanide in solution retains theanion form and is fairly mobile. It has been notedthat under strongly acidic conditions, ferrocyanideanions dissociate to release cyanide ion (Belluck,2000). It has been suggested that the appropriateconditions may occur in the laboratory but seldomor never in the environment (Chatwin, 1990).

Retention and immobilization in soilsin low-toxicity forms are the major attenuationpathways limiting the potential toxic effects offerrocyanide. Particle adsorption in clay-bearingmoderate-pH soils may effectively containferrocyanide anions. Goethite (�-FeOOH), acommon surface coating of soil particles, removesas much as 95% of the ferrocyanide anion throughadsorption at pH 4 (Theis and West, 1986). Theseremoval methods are abetted by moderate pHlevels. Soils in Canada are acidic in forestedregions, and especially so in Eastern Canada,where salting activities are heavy. In contrast,soils are neutral to alkaline in the Prairies,where soils are naturally salt enriched(Acton and Gregorich, 1995; see also CANSISat http://sis.agr.gc.ca/cansis/intro.html).Immobilized ferrocyanide is also subject tochemical conversion to compounds of lowtoxicity in the presence of sulphur, andmicroorganisms that degrade metal-complexedcyanides have been identified (Mudder, 1991).

3.7.1.1.2 Illuminated ferrocyanide

Photolysis of ferrocyanide under illumination is acomplex, intensively studied reaction (Broderius,1973; Broderius and Smith, 1980; Otake et al.,1982; Clark et al., 1984; Meeussen et al., 1992a).Factors affecting the process include pH,temperature, initial concentration, season and timeof day. In their study of photolytic reactions,Miller et al. (1989) reported that free cyaniderelease is restricted to a soil depth of 1 mm.This suggests that only ferrocyanide anions withinthe top 1 mm of soil are subject to the process.

Hydrogen cyanide produced within thesoil is mildly persistent. It has an estimated half-life of 4 weeks to 6 months (Howard et al., 1991).It is adsorbed at the rates of 0.05 mg/g by claysand 0.5 mg/g by organic matter (Chatwin, 1989).It is also subject to hydrolysis to formic acid upto a rate of 4% in saturated soils (Chatwin, 1990).Volatilization of hydrogen cyanide from soilsaccounts for as much as 10% of losses, butsaturation of the soil retards this process bya factor of 104 (Chatwin, 1990). Biologicaldegradation of various cyanides by soil microbes(Towill et al., 1978; Dubey and Holmes, 1995) orassimilation by plants (Fuller, 1984; Knowles andBunch, 1986) is an efficient removal system thatis strongly diminished under saturated anaerobicconditions (Strobel, 1967).

3.7.1.2 Water

Ferrocyanide dissociates very slowly under non-illumination. Under this condition, the amount offree cyanide released from concentrations offerrocyanide in the range of 1–100 000 mg/Lvaries from 1.2 to 6.1 µg/L. However, underenvironmental illuminated conditions, rapidconversion of ferrocyanide to free cyanide hasbeen reported (Ferguson, 1985; Singleton, 1986).

In water, free cyanide is converted tohydrogen cyanide, a weak acid, through theequilibrium reaction H+ + CN–

� HCN. Thehydrogen cyanide state is highly favouredbelow pH 6. Hydrogen cyanide is highly volatileand, under environmental conditions, quicklydissipates from water bodies, its volatilization ratebeing a function of concentration, temperature,pH and water flow rate (Schmidt et al., 1981;Melis et al., 1987). Secondary cyanide removalprocesses include complexation and precipitation,biological degradation, adsorption by organicmaterial and uptake by plants. A number ofnaturally occurring organisms are known toproduce cyanides (Knowles, 1976).

3.7.1.3 Air

Ferrocyanide itself does not volatilize. Hydrogencyanide is highly volatile and is expected to


rapidly dissipate in the atmosphere, with anestimated half-life of 89 days to 2.4 years(Howard et al., 1991).

3.7.1.4 Biomagnification

Towill et al. (1978) found no studies reportingbiomagnification of cyanide.

3.7.2 Measured and estimatedenvironmental concentrations

Reported cyanide concentrations measured inrunoff waters range from 2.3 to 22 µg/L for freecyanide and from 3 to 270 µg/L for total cyanide(Hsu, 1984; Ohno, 1989; Novotny et al., 1997;Mayer et al., 1998). Since actual data are limited,estimated concentrations were also derived.

Morin and Perchanok (2000) documentedroad salting activity in Canada. The highestlevel of use in Canada is 10 150 g salt/m2, andthe highest addition rate of sodium ferrocyanideto road salt is reported to be 124 mg/kg(Letts, 2000b). This corresponds to 86 mgferrocyanide/kg, or a stoichiometric cyanidecontent of 63.7 mg/kg. From these values, thedistribution rate is calculated as being 877.8 mgferrocyanide/m2 or 646.6 mg cyanide ion/m2.Using annual precipitation data, the concentrationrange of ferrocyanide anion in runoff water canbe determined to be 0.000 142–1.42 mg/L, whichcorresponds to a stoichiometric content of0.000 104–1.04 mg cyanide ion/L.

In soils, since local conditions willdictate the mobility of ferrocyanide, two setsof assumptions are proposed. Under Case I,ferrocyanide is precipitated out of solution and theannual application is retained in the upper 2 cmof soil. Under Case II, ferrocyanide remains insolution and 20% of the annual application isretained in the first 20 cm of soil. For Case I, theresultant soil concentration ranges from 0.0043 to43.8 mg ferrocyanide/kg soil, which correspondsto a stoichiometric cyanide ion content of0.0032–32.3 mg/kg soil. Under Case II, it rangesfrom 0.000 086 to 0.88 mg ferrocyanide/kg soil,

which corresponds to a stoichiometric cyanide ioncontent of 0.000 064–0.64 mg/kg soil.

3.7.3 Effects characterization

Initial assessment suggests that the greatestinherent danger in the use of sodium ferrocyanidelies with the ability of the ferrocyanide anionto photodecompose, resulting in free cyanide.Therefore, the following sections summarizethe known effects of both these compounds onterrestrial and aquatic biota. The limitations ofdetermining toxicity under artificial laboratoryconditions must be taken into account.

3.7.3.1 Aquatic biota

The main factors affecting toxicity in water areconcentration and water temperature. Toxicity isalso modified by the level of dissolved oxygenand salinity. The literature contains a number ofcomprehensive reviews on the toxicity of cyanideto freshwater organisms (Doudoroff, 1966, 1976,1980; Leduc et al., 1982; U.S. EPA, 1985). Someorganisms display reversible acclimatization tothe toxic effects.

3.7.3.1.1 Microorganisms

The effects of cyanide are quite diverse amongvarious microbial species. Many species exhibitremarkable resistance to noxious effects; indeed,some species are employed in cyanidedetoxification processes (Chatwin, 1990; Fallonet al., 1991). On the other hand, adverse effectson Pseudomonas putida have been reported at the1 µg/L level (U.S. EPA, 1985). This is the lowestconcentration at which adverse effects on aquaticmicroorganisms have been reported (Letts,2000a).

3.7.3.1.2 Plants

Data on toxicity to plants are scarce. The mostsensitive species, Scenedesmus quadricauda,exhibits incipient cell inhibition under a 96-hourexposure to 31 µg free cyanide/L (U.S. EPA,1985).


3.7.3.1.3 Invertebrates

In invertebrates, onset of ferrocyanide toxicitygenerally exceeds 1000 µg/L. On the other hand,the lowest free cyanide 96-hour LC50 reported inthe literature is 83 µg/L for Daphnia magna (U.S.EPA, 1985; Eisler, 1991). Divergent sensitivity inmixed subspecies groups may cause a populationstructure shift following exposure in somecommunities (Smith et al., 1979). For additionaldata, see Letts (2000a).

3.7.3.1.4 Vertebrates

Toxicity to vertebrates has been intensivelystudied — for example, by Leduc et al. (1982).Vertebrates are less sensitive to ferrocyanide(Schraufnagel, 1965) than to free cyanide(Leduc et al., 1982). Fish as a group appearmore sensitive than invertebrates to free cyanide.The 96-hour LC50 estimates lie in the range of40–200 µg hydrogen cyanide/L (U.S. EPA, 1985).A wide variety of physiological and behaviouraleffects has been reported. Life stage investigationsuggests that the juvenile phase is most sensitive,with a negative impact on growth commonlyreported. Adverse effects on predation andavoidance behaviours and swimming ability havebeen reported. Adverse reproductive effects havebeen selected as a toxicity target indicator.Schraufnagel (1965) and Smith et al. (1979)found spawning to be inhibited at concentrationsof 5.0 and 5.2 µg/L, respectively.

3.7.3.2 Terrestrial biota

The primary effect of cyanide is an interferencewith respiration as a result of disruption of amitochondrial enzyme (Towill et al., 1978; Wayet al., 1988). If the effect is not rapidly fatal,recovery is generally quick.

3.7.3.2.1 Microorganisms

Microorganisms show few adverse effectsas a result of cyanide exposure. Some showadaptations, such as alternative respiratorypathways. Evidence suggests the existence of a“cyanide microcycle” enabling microorganisms as

well as plants to use hydrogen cyanide as a sourceof nitrogen and carbon (Allen and Strobel, 1966).

3.7.3.2.2 Plants

As a group, terrestrial plants are fairly insensitiveto ferrocyanide and free cyanide. As much as10 g ferrocyanide/kg soil mix is needed to inhibitgrowth. Like microorganisms, some plantspossess alternative respiratory pathways and/orthe ability to metabolize hydrogen cyanide.However, some plants do exhibit adverse effects.A 25% inhibition of seedling emergence wasobserved in radish plants at concentrations as lowas 1.2 mg/kg (Environment Canada, 1995). Thisrepresents the lowest concentration at whichadverse effects on terrestrial plants have beenreported.

3.7.3.2.3 Invertebrates

While some invertebrates are resistant to cyanideor produce cyanides as a defensive mechanism,most demonstrate some level of sensitivity.The lowest free cyanide concentration inducingan adverse effect in terrestrial invertebrates is4 mg/kg, which caused mortality in 25% ofearthworms (Eisenia fetida) exposed for 14 days(Environment Canada, 1995).

3.7.3.2.4 Vertebrates

Ferrocyanides have been approved as foodadditives and used as therapeutics. Rats exposedto 3200 mg/kg-bw per day were unaffected(Dvorak et al., 1971). For hydrogen cyanide,existing standards with a 7- to 8-fold safetymargin limit exposure to 5 mg/m3 (Klaassen,1966). The U.S. EPA developed a subchronicreference dose of 0.08 mg/kg-bw per day forfree cyanide from available research. From theavailable literature, the mammal most sensitiveto adverse effects is the rabbit. Ballantyne (1987)observed an LD50 value of 0.6 mg/kg-bw. Amongbirds, mallards (Anas platyrhynchos) appear tobe the most sensitive, exhibiting an LD50 for oralingestion of sodium cyanide of 2.7 mg/kg-bw(Henny et al., 1994).


3.7.4 Risk characterization

As previously stated, the greatest environmentaldanger from the use of sodium ferrocyanide isthe ability of the ferrocyanide anion in solutionto undergo photodecomposition, resulting inthe release of cyanide ion and its subsequenthydrolysis to hydrogen cyanide. The process hasbeen extensively examined by Broderius andSmith (1980), who reported the rate of reactionto be generally a direct function of light intensity(which depends on time of year, time of day, skyconditions), temperature and concentration and aninverse function of pH and latitude. A maximumof five out of six of the available molecularcyanides can be released.

For assessment under Tiers 1 and 2assumptions, quotient values were calculated forbiota by dividing the highest measured and/orestimated free cyanide concentrations in water,air or soil by the individual ENEVs. The ENEVsused correspond to the CTVs, which are thelowest effect concentrations found in toxicitystudies from the available literature, divided byappropriate application factors.

3.7.4.1 Tier 1 risk characterization

The Tier 1 assessment applied worst-casecondition correction factors in conjunctionwith typical Canadian location values to computethe photolysis rate. The reaction is proposed tooccur at latitude 45°N at noon on a cloudless dayof 4°C. The rate under these conditions is28.5% per hour. The maximum number ofcyanide anions that can be released from theferrocyanide molecule through the photolysisprocess is five out of six. The cyanide soproduced is considered to be conservedentirely within the considered environmentalcompartment.

3.7.4.1.1 Cyanide concentration in water

Calculating from the range of annual mean peakcyanide concentrations expected in runoff waterthat was determined above, it can be seen that

the maximum potential cyanide concentrationafter total potential cyanide anion release throughphotolysis ranges from 0.087 to 870 µg cyanideion/L. Given a 1-hour reaction window, thecyanide ion released to water through photolysisgives a concentration ranging from 0.025 to250 µg/L.

3.7.4.1.2 Cyanide concentration in soil: Case I

Using the maximum annual salt application rate(i.e., 10 150 g/m2) and Tier 1 and soil Case Iassumptions, the maximum potential cyanideconcentration after total potential cyanide anionrelease through photolysis would be 1.35 mgcyanide ion/kg soil. After a 1-hour reactionwindow, the cyanide ion released throughphotolysis would give a concentration of0.39 mg/kg.

3.7.4.1.3 Cyanide concentration in soil:Case II

Using the maximum annual salt application rate(i.e., 10 150 g/m2) and Tier 1 and soil Case IIassumptions, the maximum potential cyanideconcentration after total potential cyanide anionrelease through photolysis would be 0.027 mgcyanide ion/kg soil. After a 1-hour reactionwindow, the cyanide ion released throughphotolysis would give a concentration of0.0078 mg/kg.

3.7.4.1.4 Cyanide concentration in air

The volatilization rate of cyanide atconcentrations of 0.1–0.5 mg/L is 0.22 mgcyanide ion/m2 per hour (Higgs, 1992). For Tier 1purposes, a closed-box model corresponding to a10-cm layer of air above the roadway was used.After 1 hour, the maximum cyanide concentrationin this layer would be 2.2 mg cyanide ion/m3.

3.7.4.1.5 Quotient calculations

Quotient calculations were based on the highestestimated free cyanide and measured totalcyanide (assumed to be free cyanide for the Tier 1


analysis) concentrations in water (i.e., 250 and270 µg/L) (Novotny et al., 1997) and werecalculated as follows: Quotient = EEV/ENEV.

For terrestrial organisms, quotientcalculations were undertaken using the highestestimated free cyanide concentration in soil underTier 1 assumptions (i.e., 0.39 mg/kg soil). Nomeasured free cyanide soil concentrations wereavailable in the literature. For birds and mammals,EEVs were calculated using a multipath modelincorporating intake from soil, air and water.

3.7.4.1.6 Tier 1 conclusions

With the exception of terrestrial vertebrates andinvertebrates, all quotients exceed 1 (Tables 26and 27). Aquatic organisms appear particularlysusceptible. A Tier 2 assessment is warranted.


The Tier 2 assessment, while still conservative,is based on more representative scenarios andassumptions. Typical and measured variables aregenerally employed.

The rate of addition of sodiumferrocyanide has decreased over the past years.The 1997 addition rate of 80 mg sodiumferrocyanide/kg salt corresponding to 55.8 mgferrocyanide ion/kg and 41.1 mg cyanide ion/kgroad salt is more typical of present-day use andwill be used in Tier 2.

Labadia and Buttle (1996) measured thepattern of road salt dispersal along a section ofsouthern Ontario highway. Their observationsshow that slightly more than 50% of applicationsare redistributed to the adjacent snowpack withina 3.7-m splash zone. This will also be taken intoconsideration in Tier 2.


TABLE 26 Quotient calculation results for aquatic biota — Tier 1 (from Letts, 2000a)

Group Name Application factor Quotient 1

Microorganisms Pseudomonas putida 5 1250 (e)1250 (m)

Plants Scenedesmus quadricauda 5 40.3 (e)43.5 (m)

Invertebrates Daphnia magna 10 30.12 (e)32.5 (m)

Vertebrates Lepomis macrochirus 5 250 (e)270 (m)

1 (m) = Using measured water concentration; (e) = Using estimated water concentration.

TABLE 27 Quotient calculation results for terrestrial biota — Tier 1 (from Letts, 2000a)

Group Name Application factor QuotientPlants Raphanus sativa 10 3.25Invertebrates Eisenia fetida 10 1Vertebrates (mammals) Sylvilagus floridanus none 1Vertebrates (birds) Anas platyrhynchos 10 1.22

The ferrocyanide photolysis ratecalculation was corrected to reflect an averagedaily light intensity and a condition of 70% clearsky effect. Corrections were also made to accountfor changes in initial ferrocyanide concentration.Overall Tier 2 assumptions produced an adjustedphotolysis rate of 13.8% per hour.

3.7.4.2.1 Cyanide concentration in water

Based on Tier 2 assumptions and the annual meanpeak chloride ion concentrations compiled byMorin and Perchanok (2000), the cyanide ionconcentration range in water after 1 hour wascalculated to be 0.0039–39 µg/L.

3.7.4.2.2 Cyanide concentration in soil: Case I

Using the maximum annual salt application rate(i.e., 10 150 g/m2) and Tier 2 and soil Case Iassumptions, the soil concentration followingthe photolytic release of cyanide ion from theavailable ferrocyanide would be 0.12 mg cyanideion/kg soil after 1 hour.

3.7.4.2.3 Cyanide concentration in soil:Case II

Using the maximum annual salt application rate(i.e., 10 150 g/m2) and Tier 2 and soil Case IIassumptions, the soil concentration followingthe photolytic release of cyanide ion from theavailable ferrocyanide would be 0.0024 mgcyanide ion/kg soil after l hour.

3.7.4.2.4 Cyanide concentration in air

The volatilization rate of cyanide atconcentrations of 0.1–0.5 mg/L is 0.22 mgcyanide ion/m2 per hour (Higgs, 1992). For thisassessment, it is assumed the available cyanide iscompletely transferred to the air above theroadway without dispersal and that the gaseousdiffusion rate is the same as the interfacialdiffusion rate. Under these conditions, after 1hour, the maximum cyanide concentration in thislayer would be 0.22 mg/m3.

3.7.4.2.5 Quotient calculations

Quotient calculations were undertaken usingthe highest estimated free cyanide concentrationin runoff water under Tier 2 assumptions (i.e.,39 µg/L) and the highest measured concentrationof 41 µg/L derived as the geometric mean ofmeasured values reported in the literature. TheCTVs used in the Tier 1 assessment are usedagain in Tier 2, but some application factors weremodified to reflect a more realistic analysis.

Assessment of terrestrial organismswas conducted using the highest estimatedfree cyanide concentration in soil under Tier 2assumptions (i.e., 0.12 mg/kg soil). No measuredfree cyanide soil concentrations were available inthe literature. Calculation of EEVs for birds andmammals was the same as in Tier 1.

3.7.4.2.6 Tier 2 conclusions

Based on the calculated quotients (Tables 28 and29), a Tier 3 assessment should be conducted foraquatic organisms.


Tier 1 and Tier 2 analyses have shown that,through the formation of cyanide, sodiumferrocyanide can have the potential to negativelyimpact the aquatic environment. The Tier 3assessment quantifies the probability of adverseeffect within the aquatic compartment.

As already stated, photolytic releaseof free cyanide from ferrocyanide anion remainsthe most serious concern from ferrocyanidereleases to the environment. Total photolysisacross the relevant wavelengths and losses due tovolatilization can be mathematically modelled;from these models, the concentration of freecyanide in solution can be calculated for anygiven time after the salt application event. Thesecalculations were part of the Tier 3 assessment.


3.7.4.3.1 Assumptions

For purposes of Tier 3, the following assumptionsand correction factors were used in estimatingferrocyanide photolysis and volatilization rates.Because photolysis is a function of water depthand clarity, ferrocyanide was considered to becontained within the first 20 cm, and claritymeasurements from a range of water bodies wereused. Likewise, the reaction rate is a function ofdate, time of day, sky conditions and temperature.Calculations were therefore made for every hourof the day for January 1, April 1, June 21 andSeptember 21 under conditions of 70% clear skyeffect with Henry’s law constant corrected for anaverage temperature of 4°C. The observations ofLabadia and Buttle (1996), who reported thatapproximately 50% of an initial salt application isredistributed to the adjacent snowpack, were, aspreviously, taken into consideration. Both streamand wind velocities were estimated from directCanadian source measurements. Mean stream

current was set at 0.52 m/s, and wind velocity wasset to 5.0, 4.7, 3.8 and 3.8 m/s for the months ofJanuary, April, June and September, respectively.The chloride ion runoff water concentrations wereselected from the range of values collected byMorin and Perchanok (2000). The probability ofselection of a given concentration for inclusionin a calculation was a function of the size ofthe area reporting that concentration. Sodiumferrocyanide addition rates were selected forinclusion in calculations on a probability basisreflecting the percentage of the market served bythe corresponding source mine.

3.7.4.3.2 Estimation of the frequency ofoccurrence of toxic events

Calculations were conducted by randomlyselecting each variable from the appropriatefrequency distribution. Variables included chlorideconcentration in runoff water, concentration offerrocyanide in salt, current speed, wind speed


TABLE 28 Quotient calculation results for aquatic biota — Tier 2 (from Letts, 2000a)

Group Name Application factor Quotient 1

Microorganisms Pseudomonas putida none 39 (e)41.5 (m)

Plants Scenedesmus quadricauda none 1.26 (e)1.3 (m)

Invertebrates Daphnia magna 10 4.7 (e)5.0 (m)

Vertebrates Lepomis macrochirus none 7.8 (e)8.3 (m)

1 (m) = Using measured water concentration; (e) = Using estimated water concentration.

TABLE 29 Quotient calculation results for terrestrial biota — Tier 2 (from Letts, 2000a)

Group Name Application factor QuotientPlants Raphanus sativa 5 0.2Invertebrates Eisenia fetida 10 0.15Vertebrates (mammals) Sylvilagus floridanus none 0.118Vertebrates (birds) Anas platyrhynchos 10 0.126

and attenuation coefficient. Photodecomposition–volatilization curves were generated for each date.The maximum free cyanide value so producedwas then recorded. This procedure was repeated10 000 times, and the resulting “maximum freecyanide observed” values (EEVmax) were sortedinto ascending order. The CTVs for each aquaticorganism were then compared with theserandomly generated distributions, and theprobability that an exposure event would equal orexceed the CTV was determined. The procedurewas run in duplicate using data derived fromacross Canada and from the heaviest use region,the Québec City to Windsor corridor.

As extensive data on salt use by manyCanadian municipalities were available, theabove procedure was repeated using the saltingdata from three municipalities selected asrepresentative of the highest, average andlowest Canadian municipal use. The selectedmunicipalities were La Tuque, Quebec;Newmarket, Ontario; and White River, Ontario.Results are briefly summarized below. Full resultsare presented in the supporting document by Letts(2000a).

3.7.4.3.3 Tier 3 results and conclusions

1) Aquatic microorganisms

Analyses indicate that Pseudomonas putidawould be adversely affected in 100% of modelledexposures. In the municipality selected torepresent a case of highest use, Microregmaheterostoma would be expected to be affectedadversely in as many as 60% of cases. However,the use of road salts in that municipality is twiceor greater than the level in all but the two nexthighest users in 90 Canadian cities. For othercities, effects would be expected in 10 to 20% ofscenarios. No adverse effects are indicated for anyof the other five aquatic microorganismsconsidered. Thus, while ferrocyanide from roadsalts may be viewed as posing little threat to fiveof seven microorganisms for which data areavailable, the most sensitive species aresusceptible to harm.

2) Aquatic plants

Results demonstrate that Scenedesmusquadricauda, the most sensitive alga identified,could be adversely affected in 16–28% of possibleexposure scenarios, with a value as high as 82%in the municipality with highest use. Valuestherefore indicate that there is a potential formoderate effects at certain times for certainsensitive species, with lower levels of risk formost aquatic plants at most sites and times.

3) Aquatic invertebrates

Adverse effects on the doughboy scallop(Chlamys asperrimus) may occur on as many as34% of exposures during April in the Québec Cityto Windsor corridor. The likelihood of negativeeffects on this organism falls to less than 25%of cases for other locations and dates. For themunicipality with highest use, adverse effectswere predicted for 88% of occurrences. No otherorganism is shown to risk harm at greater than3.5% of exposure levels. Generally, membersof this group are not considered at serious risk,but significant exposure of sensitive speciesmay occur.

4) Aquatic vertebrates

The highest potential for significant exposureswas indicated for aquatic vertebrates. Among thestudies reviewed by Letts (2000a), the 96-hourLC50 for rainbow trout was reported to be 17 µg/L(Higgs, 1992), 27 µg/L (Smith et al., 1979) and28 µg/L (Kovacs, 1979; Kovacs and Leduc,1982). The probability of occurrence ofconcentrations equal to or greater than these is,respectively, 52.6%, 26.8% and 25.2% forJanuary and 67.2%, 36.2% and 33.8% for Aprilfor locations within the Québec City to Windsorcorridor. This is the area of higher concentrationsof the two broad-based sources of data.Concentrations at these levels are most likely tooccur in areas where annual mean peak road saltschloride concentrations in runoff are 1000 mg/Lor more (modelled as 53 out of 90 cities). Thus, inthese areas, ferrocyanide in undiluted runoff couldaffect aquatic invertebrates.


3.7.5 Conclusion for ferrocyanides

The assessment indicates that, at present levels ofroad salt use, potential adverse effects related tosodium ferrocyanide are limited to aquatic biota,with possible negative consequences to sensitivespecies of aquatic microorganisms, aquatic plantsand aquatic invertebrates. The more sensitiveaquatic vertebrates in roadside ditches and watercourses in areas of high use would be mostsusceptible to harmful exposures. The modellingof likelihood of effects was based on exposurespredicted in roadside ditches, and did not accountfor further dilution by receiving waters, whichwould tend to overestimate exposure. At thesame time, modelling was based on dilutionof salts by total yearly precipitation, whichwould underestimate exposure at any given time.The ultimate impact of ferrocyanides on fishpopulations depends on a wide range ofparameters, including those influencing thequantities of ferrocyanides entering water courses(e.g., mass and timing of road salts application,runoff), the photolysis of the ferrocyanidesto cyanide (e.g., insolation, water turbidity,temperature), volatilization of cyanide (e.g., waterflow, temperature), status of fish or amphibianpopulations in receiving waters, extent of dilutionin receiving waters, etc. Accordingly, whileavailable data indicate a potential for adverseeffects, it is not possible to predict reliably theimpact of ferrocyanide on vertebrate populationsin water bodies adjacent to or near roadways.

3.8 Overall conclusions

3.8.1 Considerations

As provided by Paragraph 76(1) of CEPA 1999,PSL substances are to be assessed to determinewhether they are “toxic or capable of becomingtoxic.” In this assessment, road salts are assessedas to whether they are “toxic or capable ofbecoming toxic” as defined in Paragraphs 64(a)and 64(b) of CEPA 1999, namely:

... a substance is toxic if it is entering or may enterthe environment in a quantity or concentrationor under conditions that

(a) have or may have an immediate or long-termharmful effect on the environment or itsbiological diversity;

(b) constitute or may constitute a danger to theenvironment on which life depends.

3.8.1.1 Environment

CEPA 1999 defines “environment” quite broadly,to include all biotic and abiotic components of thebiosphere (CEPA 1999, Section 3):

“environment’’ means the components of the Earthand includes

(a) air, land and water;(b) all layers of the atmosphere;(c) all organic and inorganic matter and living

organisms; and(d) the interacting natural systems that include

components referred to in paragraphs (a)to (c).

This definition does not distinguish between“natural” and “non-natural” environments. Indeed,all ecosystems can be considered to have beensubjected to stress or modifications by humanactivity, be it at a local (e.g., habitat destruction,emission of non-persistent pollutants in liquideffluent), regional (e.g., dispersion of atmosphericpollutants) or global scale (e.g., persistent organicpollutants, global atmospheric effects).

Road salts are stored or used and snowis disposed of in engineered environments —namely, storage or patrol yards, roadways and snowdumps. The degree to which these are in fact“engineered” is quite variable: salts can be stored inspecially designed facilities at patrol yards or intemporary piles at the margins of roadways or patrolyards; roadways can be part of complex corridorswith rights-of-ways with lined ditches or run parallelto and drain directly into streams, lakes or wetlands;snow dumps can range from secure facilities withmeltwater retention and treatment to dumps onavailable vacant lots. Thus, the degree to which roadsalts are released to well-circumscribed engineeredsystems is quite variable, as is the level of concernor need for protection at the periphery of theseengineered systems.


In some cases, elements of theenvironment that are part of or immediatelyadjacent to the engineered systems can beecologically significant (Nadec, 2001). Roadsideshave been shown to be extensively used as habitatby a considerable diversity of species (Perringet al., 1964; Oetting and Cassel, 1971; Way, 1977;Laursen, 1981; Adams and Geiss, 1983; Auestadet al., 1999; Bellamy et al., 2000). Accordingly,roadsides play two ecologically significant roles:

• Roadsides contribute to maintainingbiodiversity by providing habitat for somespecies. For example, 35 of the 257 nationallyrare plant species in Great Britain are found inroadsides, including some for which roadsidesare the main remaining habitat (Way, 1977). InQuebec, 75 species of birds were observed inroad rights-of-way, among which 29 werelong-range migratory birds and 10 were year-round local residents (Lacroix and Bélanger,2000). Migratory birds are protected species,subject to the provisions of the MigratoryBirds Convention Act. In parts of Quebec,roadside ditches were reported to be the onlysite where the frog Pseudacris triseriata,identified as vulnerable on the provincialendangered species list, was observed to breed(Saumure, 2001). In southwestern Ontario,only 0.5–3.8% of the original tallgrass prairieor oak savanna ecosystems remain that existedat the time of land surveys during the 1700sand 1800s (Bakowsky and Riley, 1994).At least 35 native tallgrass prairie plantspecies persist in prairie-like conditions foundalong roadsides in southern and centralOntario (Woodliffe, 2001). Thirteen of thesespecies are on the provincial “rare, threatened,and endangered” list (Argus et al., 1987;Woodliffe, 2001).

• Roadsides contribute in reducing theprobability of local species extirpationdue to habitat fragmentation. While habitatfragmentation has been shown to lead tospecies extinction (Wilcox and Murphy, 1985),research has also shown that, given afragmentation event, the linking of fragmentsby strips of hospitable habitat allows the

maintenance of biodiversity within thefragments (Noss, 1983; Lefkovitch and Fahrig,1985; Bennett, 1990; Bélanger, 2000).Roadsides have been shown to act as stripsof hospitable habitat linking fragments andensuring habitat continuity (Huey, 1941;Getz et al., 1978; Lewis, 1991; Andreassenet al., 1996).

Changes to roadside habitats that could result fromcontamination by road salts could thereforecompromise their ecological value.

3.8.1.2 Harmful effects

It is recognized that any environmental change isnot necessarily adverse, nor does it necessarilyresult in harmful effects. Changes within the rangeof normal variation may not necessarily constituteor result in harm, but more substantive ones maybe harmful. The potential for harm resulting fromincreases in salt concentrations in the environmentwould depend, among other factors, on themagnitude, duration and frequency of changes.

Effects observed at the population,community or ecosystem level are generallyconsidered more environmentally harmful andare of more concern than those observed only atlower levels of organization. Few studies havedirectly tested Priority Substances for effects atthe population, community or ecosystem level oforganization. Most toxicity studies are conducted inthe laboratory using relatively small sample sizesrelative to population sizes in natural communities.However, many of the effects measured inlaboratory and field studies have implicationsfor populations, communities and ecosystems.Effects on processes such as growth, reproductionor survival are closely related to the viability ofnatural populations and are used as an indicationof the potential for environmental harm. Whilequantification of likely ecosystem impacts is oftennot feasible, it can be assumed that where ambientconcentrations approach those associated withserious effects on individual organisms (suchas median lethal concentrations), harmfulenvironmental impact may also be expected.


3.8.1.3 Weight of evidence

CEPA 1999, Section 76.1, provides that, whenconducting and interpreting the results of anassessment to determine whether a PSL substanceis “toxic or capable of becoming toxic,” “theMinisters shall apply a weight of evidenceapproach and the precautionary principle.”

It is impossible to specify a rigid cut-offpoint where effects are considered sufficient inextent or harm to declare a substance “toxic orcapable of becoming toxic.” In using a weightof evidence approach, several lines of evidenceand endpoints are considered. This includesconsideration of several lines of evidence for agiven environmental compartment (e.g., surfacewater assessment considering toxic thresholdsfor a range of aquatic biota, potential for shifts infreshwater algal populations, changes in physicalstructure of lakes, exposure resulting fromreleases of contaminated groundwater throughsprings, etc.). Ultimate overall conclusions mustbe based on consideration of the full array of linesof evidence. While some lines of evidence mayprovide for only tenuous conclusions, others willprovide stronger evidence. By using a weight ofevidence approach, risk assessment can reduce thebiases and uncertainties associated with usingonly one approach to estimate risk.

3.8.2 Summary of assessment

Road salts are used as de-icing and anti-icingchemicals for winter road maintenance, withsome use as summer dust suppressants. Inorganicchloride salts considered in this assessmentinclude sodium chloride, calcium chloride,potassium chloride and magnesium chloride.In the environment, these compounds dissociateinto the chloride anion and the correspondingcation. In addition, ferrocyanide salts, which areadded as anti-caking agents to some road saltformulations, were assessed. It is estimated thatapproximately 4.75 million tonnes of sodiumchloride were used as road salts in the winter of1997–98 and that 110 000 tonnes of calciumchloride are used on roadways in a typical year.Very small amounts of other salts are used. Based

on these estimates, about 4.9 million tonnes ofroad salts can be released to the environmentin Canada every year, accounting for about3.0 million tonnes of chloride. The highest annualloadings of road salts on a road-length basisare in Ontario and Quebec, with intermediateloadings in the Atlantic provinces and lowestloadings in the western provinces.

Road salts enter the Canadianenvironment through their storage and useand through disposal of snow cleared fromroadways. Road salts enter surface water, soiland groundwater after snowmelt and are dispersedthrough the air by splashing and spray fromvehicles and as windborne powder. Chlorideions are conservative, moving with water withoutbeing retarded or lost. Accordingly, all chlorideions that enter the soil and groundwater canultimately be expected to reach surface water;it may take from a few years to several decades ormore for steady-state groundwater concentrationsto be reached. Because of the widespreaddispersal of road salts through the environment,environmental concerns can be associated withmost environmental compartments.

In water, natural backgroundconcentrations of chloride are generally no morethan a few milligrams per litre, with some localor regional instances of higher natural salinity,notably in some areas of the Prairies and BritishColumbia. High concentrations of chloride relatedto the use of road salts on roadways or releasesfrom patrol yards or snow dumps have beenmeasured. For example, concentrations ofchloride over 18 000 mg/L were observed inrunoff from roadways. Chloride concentrations upto 82 000 mg/L were also observed in runoff fromuncovered blended abrasive/salt piles in a patrolyard. Chloride concentrations in snow clearedfrom city streets can be quite variable. Forexample, the average chloride concentrationsin snow cleared from streets in Montréal were3000 mg/L (secondary streets) and 5000 mg/L(primary streets). Waters from patrol yards,roadways or snow dumps can be diluted tovarious degrees when entering the environment.In the environment, resulting chloride


concentrations have been measured as high as2800 mg/L in groundwater in areas adjacent tostorage yards, 4000 mg/L in ponds and wetlands,4300 mg/L in watercourses, 2000–5000 mg/Lin urban impoundment lakes and 150–300 mg/Lin rural lakes. While highest concentrations areusually associated with winter or spring thaws,high concentrations can also be measured in thesummer, as a result of the travel time of the ionsto surface waters and the reduced water flowsin the summer. Water bodies most subject tothe impacts of road salts are small ponds andwatercourses draining large urbanized areas, aswell as streams, wetlands or lakes draining majorroadways. Field measurements have shown thatroadway applications in rural areas can result inincreased chloride concentrations in lakes locateda few hundred metres from roadways.

The potential for impacts to regionalgroundwater systems was evaluated using amass balance technique that provides anindication of potential chloride ion concentrationsdowngradient from saltable road networks. Themass balance modelling and field measurementsindicated that regional-scale groundwater chlorideconcentrations greater than 250 mg/L will likelyresult under high-density road networks subject toannual loadings above 20 tonnes sodium chlorideper two-lane-kilometre. Considering data onloadings of road salts, urban areas in southernOntario, southern Quebec and the Atlanticprovinces face the greatest risk of regionalgroundwater impacts. Groundwater willeventually well up into the surface water oremerge as seeps and springs. Research has shownthat 10–60% of the salt applied enters shallowsubsurface waters and accumulates until steady-state concentrations are attained. Elevatedconcentrations of chlorides have been detected ingroundwater springs emerging to the surface.

Acute toxic effects of chloride on aquaticorganisms are usually observed at relativelyelevated concentrations. For example, the 4-daymedian lethal concentration (LC50) for thecladoceran Ceriodaphnia dubia is 1400 mg/L.Exposure to such concentrations may occur insmall streams located in heavily populated urban

areas with dense road networks and elevated roadsalt loadings, in ponds and wetlands adjacent toroadways, near poorly managed salt storagedepots and at certain snow disposal sites.

Chronic toxicity occurs at lowerconcentrations. Toxic effects on aquatic biotaare associated with exposures to chlorideconcentrations as low as 870, 990 and 1070 mg/Lfor median lethal effects (fathead minnowembryos, rainbow trout eggs/embryos anddaphnids, respectively). The NOEC for the 33-day early life stage test for survival offathead minnow was 252 mg chloride/L.Furthermore, it is estimated that 5% of aquaticspecies would be affected (median lethalconcentration) at chloride concentrations of about210 mg/L, and 10% of species would be affectedat chloride concentrations of about 240 mg/L.Changes in populations or community structurecan occur at lower concentrations. Because ofdifferences in the optimal chloride concentrationsfor the growth and reproduction of differentspecies of algae, shifts in populations in lakeswere associated with concentrations of12–235 mg/L. Increased salt concentrations inlakes can lead to stratification, which retards orprevents the seasonal mixing of waters, therebyaffecting the distribution of oxygen and nutrients.Chloride concentrations between 100 and1000 mg/L or more have been observed in avariety of urban watercourses and lakes. Forexample, maximum chloride concentrations inwater samples from four Toronto-area creeksranged from 1390 to 4310 mg/L. Chlorideconcentrations greater than about 230 mg/L,corresponding to those having chronic effectson sensitive organisms, have been reported fromthese four watercourses through much of the year.Available studies indicate that, in areas ofheavy use of road salts, especially southernOntario, Quebec and the Maritimes, chlorideconcentrations in groundwater and surface waterare frequently at levels likely to affect biota, asdemonstrated by laboratory and field studies.

Application of road salts can also result indeleterious effects on the physical and chemicalproperties of soils, especially in areas that suffer


from poor salt, soil and vegetation management.Effects are associated with areas adjacent to saltdepots and roadsides, especially in poorly draineddepressions. Effects include impacts on soilstructure, soil dispersion, soil permeability, soilswelling and crusting, soil electrical conductivityand soil osmotic potential. These can have, inturn, abiotic and biotic impacts on the localenvironment. The primary abiotic impact is theloss of soil stability during drying and wettingcycles and during periods of high surface runoffand wind. Biological impacts relate primarily toosmotic stress on soil macro- and microflora andmacro- and microfauna, as well as salt-inducedmobilization of macro- and micronutrients thataffect flora and fauna.

A number of field studies havedocumented damage to vegetation and shiftsin plant community structure in areas impactedby road salt runoff and aerial dispersion.Halophytic species, such as cattails and commonreed-grass, readily invade areas impacted by salt,leading to changes in occurrence and diversity ofsalt-sensitive species. Elevated soil levels ofsodium and chloride or aerial exposure to sodiumand chloride results in reductions in flowering andfruiting of sensitive plant species; foliar, shootand root injury; growth reductions; and reductionsin seedling establishment. Sensitive terrestrialplants may be affected by soil concentrationsgreater than about 68 mg sodium/kg and 215 mgchloride/kg. Areas with such soil concentrationsextend linearly along roads and highways or otherareas where road salts are applied for de-icing ordust control. The impact of aerial dispersionextends up to 200 m from the edge of multi-lanehighways and 35 m from two-lane highwayswhere de-icing salts are used. Salt injury tovegetation also occurs along watercourses thatdrain roadways and salt handling facilities.

Behavioural and toxicological impactshave been associated with exposure ofmammalian and avian wildlife to road salts.Ingestion of road salts increases the vulnerabilityof birds to car strikes. Furthermore, intakecalculations suggest that road salts may poisonsome birds, especially when water is not freely

available during severe winters. Road salts mayalso affect wildlife habitat, with reduction in plantcover or shifts in communities that could affectwildlife dependent on these plants for food orshelter. Available data suggest that the severityof road kills of federally protected migratorybird species (e.g., cardueline finches) and thecontribution of road salts to this mortality havebeen underestimated.

Ferrocyanides are very persistent but areof low toxicity. However, in solution and in thepresence of light, they can dissociate and formcyanide. In turn, the cyanide ion may volatilizeand dissipate fairly quickly. The ultimate effectsof ferrocyanides therefore depend on the complexbalance between photolysis and volatilization,which in turn depends on environmental factors.Modelling studies undertaken in support of thisassessment indicate that there is a potential forcertain aquatic organisms to be adversely affectedby cyanide in areas of high use of road salts.

3.8.3 Uncertainties

There are a number of uncertainties andassumptions inherent in the environmental riskcharacterization of road salts presented in thisreport. These uncertainties can typically beassociated with knowledge and data gaps inlaboratory and field data that were used tocharacterize the use of road salts in Canadaand assess the impact of road salts on differentenvironmental endpoints. In addition todiscussions of uncertainties in previous sections,the following paragraphs present furtherdiscussions of uncertainties pertinent to theconclusions for this assessment.

A concerted effort was made to obtaindata on road salt loadings from the most accuratesources. Data were also analysed to determineif the base year used to characterize road saltloadings was abnormal. Results of this analysisindicate that the 1997–98 winter season was notcharacterized by elevated loadings (incomparison with the 5-year window for whichdata were collected). One limitation, however, isthat data are characterized by provincial


maintenance district. While these estimates arereliable, it is uncertain how the use of road saltsis distributed within a maintenance district. Itwas assumed that road salts were used equallythroughout the maintenance districts. Certainroads, however, are subject to higher loadings,and the density of roads will vary within amaintenance district. Estimates were made toaccount for municipalities where data were notavailable and for the amount of salt that would beused for commercial purposes (e.g., in parkinglots). Estimates were not made to account for thevolume of salt that would be used by individualsor other agencies, such as transit or portauthorities.

Most of the data presented in theAssessment Report relate to the use of road saltson roadways. Road salts are also used inparking lots and other commercial and industrialproperties. Data on chloride concentrations inrunoff from such properties and in receivingwaters were not presented in this report;however, there is a potential for receiving watersto be impacted by such releases.

The number of active patrol yards inCanada is uncertain. Information from provincialand territorial agencies indicates that they manageapproximately 1300 patrol yards. Considering thatother levels of government (e.g., counties andmunicipalities) and private sector agencies alsomaintain roadways, it is probable that thisnumber considerably underestimates the numberof patrol yards in Canada. Another uncertainty isthe amount of salt that may be stored at patrolyards. Data and estimates of the amount of saltthat may be stored at patrol yards were presented,but these are rough estimates. For example, it wasestimated that the amount of salt stored at patrolyards ranges between 45 and 21 000 tonnes.Other uncertainties relate to the type of structuresused to store salts, the number of agencies thatstore blended abrasive/salt piles outside, thepercentage of salt in blended piles, the dischargeof washwater and the range of practices used toreduce the impact of road salt storage on theenvironment.

Another uncertainty relates to therelevance of the use of exposure data obtainedfrom engineered environments such as ditches,rights-of-ways and stormwater retention pondsto estimate possible environmental effects.Modified environments can be ecologicallyimportant. In areas heavily impacted byurbanization, industrialization or agriculture,modified landscapes such as rights-of-ways orimpoundments can in fact offer the only habitatsavailable for many species locally or regionally,including sensitive species as well as protectedorganisms such as migratory birds. As such,consideration of potential impacts in thesemodified environments may be justifiable.In some cases, concentrations in engineeredenvironments can be used as surrogates forconcentrations that may be found in adjacentareas. Thus, concentrations in ditches can berepresentative of environmental concentrationslikely to occur in streams or wetlandsimmediately adjacent to roadways, andconcentrations in groundwater below storagefacilities may be indicative of groundwaterconcentrations immediately off-site. In yet otherinstances, it may be more appropriate to considerdata from engineered environments as an estimateof concentrations that may be released to theenvironment following little or extensive dilution.

There is also a certain degree ofuncertainty regarding the models used to estimatethe impact of road salts on soils, surface waterand groundwater resources. At the outset, anyassumptions made when characterizing road saltloadings (e.g., equal use over a maintenancedistrict) were carried forward to these models.Furthermore, these models were designed tobe as simple as possible to derive reasonableapproximate estimates of the spatial effects of roadsalts on the endpoints at a national scale. Becauseof the numerous and often confounding factors atplay in the environment, it was not possible todevelop one model that would provide site-specific results. The calculations and proceduresused in these models are based on regional-scaleassumptions. If the different parameters includedin these models were measured on a site-specificbasis, more accurate estimates of the impacts of


road salts on environmental endpoints could bemodelled. Accordingly, the estimates generatedfrom these models should be viewed ascharacteristic of the range of impacts that can beexpected at a national scale.

Some of the exposure data presented inthe Assessment Report were collected as part ofroutine water quality monitoring; samples werenot necessarily collected within a frameworkdesigned to assess the impact of road salts onwater quality. Studies like the one by Whitfieldand Wade (1992) demonstrate the importanceof continuous monitoring to characterize thetemporal impact of road salts on watercourses.The temporal aspect of these concentrations isimportant to determine if organisms are exposedto concentrations that cause acute or chroniceffects. Continuous monitoring data were seldomavailable; however, several years of monitoringdata were used to characterize concentrations thatwould be typical for certain locations at differenttimes. A few studies with continuous or extendedsampling were available that permittedcorroboration of other studies.

Without constant monitoring data, itis difficult to determine if the grab samples arerepresentative of peak concentrations or aremore characteristic of ambient concentrationsthat are typical of certain time periods.Recent unpublished monitoring data for sixwatercourses in the Toronto area from September2000 to April 2001 suggest that elevated chlorideconcentrations can be observed throughout wintermonths. For example, samples taken fromOctober 30 to the end of April for EtobicokeCreek, Mimico Creek, Humber River, Don River,Highland Creek and Rouge River had chlorideconcentrations that ranged between 448 and 3771,419 and 5376, 264 and 2110, 87 and 2247, 51 and1755 and 101 and 434 mg/L, respectively (City ofToronto, 2001). These data generally concur withother data that were used to characterize chlorideconcentrations in watercourses in the GreaterToronto Area. Furthermore, this information(and monitoring data presented previously)seems to suggest that aquatic organisms in these

watercourses would be exposed to elevatedchloride concentrations throughout winter andspring. Other work in New Brunswick indicatedmaximum concentrations in summer.

There is also a certain degree ofuncertainty regarding the species that may beimpacted by the use of road salts. The sectionson aquatic ecosystems, vegetation, wildlife andferrocyanides compared toxicity values withexposure concentrations that are characteristicof the same region. Within a region, however,the distribution of these species and thecorresponding exposure concentrations may vary.While there is a certain degree of uncertainty,case studies were used to reduce the degree ofuncertainty and highlight the type of impact thatcan be expected.

Case studies, which document theimpact of road salts on environmental endpoints,are presented in this Assessment Report andsupporting documentation. The review of casestudies should not be viewed as exhaustive.The number of case studies presented likelyunderestimates the number of instances whereconcentrations of concern for salts have beenexceeded. Case studies in this report havegenerally been brought to the forefrontbecause the taste of groundwater-based drinkingwater supplies has been affected as a resultof an elevation in salt concentrations fromcontamination by road salts. It is expected thatcontamination in areas away from drinking watersources would frequently not be detected orreported.

It is not possible to predict the impact ofpotential climate change on the net use of roadsalts in the future, notably as this relates to thefrequency and extent of precipitation events atair temperatures of about –10°C to 0°C. However,it can be expected that any yearly increase inevapotranspiration or decrease in precipitationwould result in an increase in concentrations ofchloride and other solutes in surface water andshallow groundwater. Field evidence in theExperimental Lakes Area in western Ontario


indicated that an increase in average airtemperature from 14°C to 16°C caused anincrease in evaporation of 30% (Schindler et al.,1996). Despite predicted increases in precipitationin parts of the Great Lakes basin, greatlyincreased evapotranspiration in the catchmentof the Great Lakes is expected, resulting inreductions in wetlands (Mortsch and Quinn, 1996)and a corresponding increase in concentrations ofsolutes.

3.8.4 Conclusions

Almost 5 million tonnes of road salts thatcontain inorganic chloride salts with or withoutferrocyanides are used every year in Canada.All of these salts are ultimately found in theenvironment, whether as a result of dispersiveuse on roadways or through losses from patrolyards and snow dumps. The report has identifiedsituations where the resulting environmentalconcentrations approach or exceed thoseassociated with harmful effects on physicalproperties of soils or water bodies or onorganisms associated with freshwater andterrestrial habitats at numerous sites acrossCanada.

Local and regional contamination ofgroundwater can lead, sometimes after severalyears or decades, to high concentrations ofchloride, including concentrations that exceedacute and chronic toxicity values for organismsthat may occur in springs.

For surface waters associated withroadways or storage facilities, episodes of salinityhave been reported during the winter and springin some urban watercourses in the rangeassociated with acute toxicity in laboratoryexperiments. Reports of lower levels of exposureranging from near those associated with chronictoxicity in laboratory experiments to thosemodestly above background salinity levels aremore frequent and widespread. Some measurablechanges in aquatic biota and/or communities havebeen found in laboratory experiments and in field

studies at these lower concentrations, includingpossible impacts on lake stratification.

Increased salt loadings into soils canlead to modification of soil properties criticalfor soil health. A number of field studies havedocumented damage to vegetation and shifts inplant community structure in areas impacted byroad salt runoff and aerial dispersion. Elevatedsoil levels of sodium and chloride or aerialexposure to sodium and chloride have beenrecorded frequently at levels impacting growth,reproduction and survival of sensitive plantspecies along roadways and watercourses thatdrain roadways and salt handling facilities.

Behavioural and toxicological impactshave been associated with exposure ofmammalian and avian wildlife to road salts.Road salts may also affect wildlife habitat, withreduction in plant cover or shifts in communitiesthat could affect wildlife dependent on theseplants for food or shelter.

Since ferrocyanides can dissociate in theenvironment to form cyanide, there is a potentialfor certain aquatic organisms to be adverselyaffected by cyanide in areas of high use of roadsalts.

While the ecological significance of manyof the potential impacts outlined above cannot bequantified, it is clear from available informationthat there is a reasonable probability that roadsalts that contain inorganic chloride salts withor without ferrocyanide salts may be having animmediate or long-term harmful effect on someCanadian surface water organisms, terrestrialvegetation and wildlife and may also constitute adanger to the environment on which life dependsthrough its impacts on aquatic systems and soilsand terrestrial habitats. Thus, road salts thatcontain inorganic chloride salts with or withoutferrocyanide salts should be considered “toxic”under CEPA 1999 because of tangible threats ofserious or irreversible environmental damage.


CEPA 1999 64(a): Based on the available data, itis considered that road saltsthat contain inorganic chloridesalts with or withoutferrocyanide salts are enteringthe environment in a quantityor concentration or underconditions that have or mayhave an immediate or long-term harmful effect on theenvironment or its biologicaldiversity. Therefore, it isconcluded that road salts thatcontain inorganic chloridesalts with or withoutferrocyanide salts are “toxic”as defined in Paragraph 64(a)of CEPA 1999.

CEPA 1999 64(b): Based on the available data,it is considered that road saltsthat contain inorganicchloride salts with or withoutferrocyanide salts are enteringthe environment in a quantityor concentration or underconditions that constitute ormay constitute a danger to theenvironment on which lifedepends. Therefore, it isconcluded that road salts thatcontain inorganic chloridesalts with or withoutferrocyanide salts are “toxic”as defined in Paragraph 64(b)of CEPA 1999.

Overall conclusion: Based on critical assessment

of relevant information, roadsalts that contain inorganicchloride salts with or withoutferrocyanide salts areconsidered to be “toxic” asdefined in Section 64 ofCEPA 1999.

3.9 Considerations for follow-up(further action)

The use of de-icing agents is an importantcomponent of strategies to keep roadways openand safe during the winter and minimize trafficcrashes, injuries and mortality under icy andsnowy conditions. These benefits were recognizedby the Ministers’ Expert Advisory Panel on theSecond Priority Substances List, even as theyrecommended that this assessment of potentialimpacts on the environment be conducted. Anymeasures developed as a result of this assessmentmust never compromise human safety; selectionof options must be based on optimization ofwinter road maintenance practices so as to notjeopardize road safety, while minimizing thepotential for harm to the environment. Any actiontaken to reduce impacts on the environment islikely also to reduce potential for contaminationof groundwater-based drinking water supplies,which is clearly desirable.

Future management should focus onkey sources in areas where the assessment hasindicated concerns. These relate to the following:

• Patrol yards: Key concerns relate tocontamination of groundwater at patrolyards and the discharge to surface water.In addition, overland flow of salty snowmeltwaters can result in direct impacts on surfacewater and near-field vegetation. Based onsurveys and reviews, salt losses from patrolyards are associated with loss at storage piles(which include salt piles as well as piles ofsand and gravel to which salts have beenadded) and during the handling of salts,relating to both storage and loading andunloading of trucks. The discharge of patrolyard washwater is also a potential sourceof release of salts. Measures and practicesshould therefore be considered to ensurestorage of salts and abrasives to reducelosses through weathering, to reduce lossesduring transfers and to minimize releases ofstormwater and equipment washwater.


• Roadway application: Key environmentalconcerns have been associated with areasof high salt use and high road density.Regions of southern Ontario and Quebec andthe Atlantic provinces have the highest rateof salt use on an area basis and as such havethe highest potential for contamination ofsoils, groundwater and surface water byroad salts as a result of roadway applications.In addition, urban areas in other parts of thecountry where large amounts of salts areapplied are of potential concern, especiallyfor streams and aquifers that are whollysurrounded by urban areas. In rural areas,surface waters receiving drainage fromroadways may also be susceptible tocontamination. Areas where splash or sprayfrom salted roads can be transported throughair to sensitive vegetation are a potentialconcern. Wetlands that directly adjoinroadway ditches and that receive runoff inthe form of salty snowmelt waters are alsopotential management concerns. Therefore,measures should be considered to reduce theoverall use of chloride salts in such areas.The selection of alternative products or ofappropriate practices of technology to reducesalt use should be considered while ensuringmaintenance of roadway safety.

• Snow disposal: Key environmental concernsrelate to eventual loss of meltwater intosurface water and into soil and groundwaterat snow disposal sites. Measures to minimizepercolation of salty snowmelt waters into soiland groundwater at snow disposal sites shouldbe considered. Practices to direct the releaseof salty snowmelt waters into surface waterswith minimal environmental sensitivity orinto storm sewers could be considered.Measures should also be considered to ensuresufficient dilution before release.

• Ferrocyanides: This assessment indicatesthat there is a possible adverse exposure forthe more sensitive aquatic vertebrates in areasof very high use of road salts. Risks couldbe reduced by reducing total salt use orreducing content of ferrocyanides in road saltformulations. To reduce the possibility ofexposure, producers of road salts couldconsider reducing the addition rate offerrocyanide to road salt. Any reductionin total salt use would be expected to resultin an equivalent reduction in release offerrocyanides.


Acton D.F. and L.J. Gregorich (eds). 1995.The Health of Our Soils. Toward sustainableagriculture in Canada. Centre for Land andBiological Resources Research. ResearchBranch, Agriculture and Agri-Food Canada.Publication 1906/E.

Adams, L.W. and A.D. Geiss. 1983. Effectsof roads on small mammals. J. Appl. Ecol.20: 403–415.

Adelman, I.R., L.L.J. Smith and G.D. Siesennop.1976. Acute toxicity of sodium chloride,pentachlorophenol, guthion and hexavalentchromium to fathead minnows (Pimephalespromelas) and goldfish (Carassius auratus).J. Fish. Res. Board Can. 33: 203–208.

AGRA. 1995. Phase 1 of groundwater flow andcontaminant assessment, Heffley Creek,British Columbia. Report prepared by AGRAEarth and Environmental for the BritishColumbia Ministry of Transportation andHighways, Kamloops, British Columbia(Project No. KX11322).

AGRA. 1999. Chloride plume monitoring —October 1998 and February 1999, HeffleyCreek, British Columbia. Report preparedby AGRA Earth and Environmental for theBritish Columbia Ministry of Transportationand Highways, Kamloops, British Columbia(Project No. KX11538).

Alberta Infrastructure. 2000. Draft Phase 3environmental site assessments for theWorsley Satellite Yard, Obed Satellite Yardand 25th Baseline Satellite Yard. Reportprepared by Green Plan Limited for AlbertaInfrastructure, Civil Projects Branch,Edmonton, Alberta.

Ali, O.M., M. Yousaf and J.D. Rhoades. 1987.Effect of exchangeable cation and electrolyteconcentration on mineralogy of clay dispersedfrom aggregates. Soil Sci. Soc. Am. J.51: 896–900.

Allen, J. and G.A. Strobel. 1966. The assimilationof H14CN by a variety of fungi. Can. J.Microbiol. 12: 414–416 [cited in Towill et al.,1978].

American Cyanamid Company. 1953. Thechemistry of the ferrocyanides. The BeaconPress, Inc., New York, N.Y.

Anderson, B.G. 1948. The apparent thresholdsof toxicity to Daphnia magna for chloridesof various metals when added to Lake Eriewater. Am. Fish. Soc. 78: 96–113.

Andreassen, H.P., R.A. Ims and O.K. Steinset.1996. Discontinuous habitat corridors: effectson male root vole movements. J. Appl. Ecol. 33: 555–560.

Anon. 1992. Dryland salinity investigationsprocedures manual. Conservation andDevelopment Branch, Alberta Agriculture,and Prairie Farm RehabilitationAdministration, Agriculture and Agri-FoodCanada. 110 pp.

Arambasic, M.B., S. Bjelic and G. Subakov. 1995.Acute toxicity of heavy metals (copper, lead,zinc), phenol and sodium on Allium cepa L.,Lepidium sativum L. and Daphnia magna St.:comparative investigations and the practicalapplications. Water Res. 29: 497–503.


4.0 REFERENCES

Argus, G.W., K.M. Pryer, D.J. White andC.J. Keddy (eds.). 1987. Atlas of the rarevascular plants of Ontario, 1982–1987.National Museum of Natural Sciences,Ottawa, Ontario.

Arp, P.A. 2001. Electrical conductivity andchloride contents of stream, pond andditchwater locations near a salt depot: a casestudy. Faculty of Forestry and EnvironmentalManagement, University of New Brunswick,Fredericton, New Brunswick. 14 pp.

Arsenault, R.A., T.L. Pollock, S. Ratnam andD. Hansen. 1985. Surface water qualitystudy in the Waterford River Basin. UrbanHydrology Study of the Waterford RiverBasin. 91 pp. (Technical Report No. UHS-WRB1.4).

Auestad, I., A. Norderhaug and I. Auestad. 1999.Road verges — Species rich habitats. AspectsAppl. Biol. 54: 269–275.

August, L.L. and T.A. Graupensperger. 1989.Impacts of highway deicing programs ongroundwater and surface water in Maryland.In: Groundwater issues and solutions in thePotomac River Basin. Chesapeake BayRegion. National Water Well AssociationConference, May 14–16, 1989, Washington,D.C. pp. 37–394.

Baeten, L.A. and F.J. Dein. 1996. Salt toxicosis inwild birds. Am. Assoc. Zoo Vet. Annu. Proc.1996: 296–297.

Baker, K.D. 1965. An observation of birdmortality on highways. Blue Jay 23(2): 79–80.

Bakowsky, W. and J.L. Riley. 1994. A survey ofthe prairies and savannas of southern Ontario.In: R.G. Wickett, P.D. Lewis, A. Woodliffeand P. Pratt (eds.), Proceedings of the 13thNorth American Prairie Conference, 1992.Department of Parks and Recreation,Windsor, Ontario.

Ballantyne, B. 1987. Clinical and experimentaltoxicology of cyanides. In: B. Ballantyne andT.C. Marrs (eds.), Clinical and experimentaltoxicology of cyanides. John Wright, Bristol,U.K. pp. 41–126.

Barica, J. 1978. Variability in ionic compositionand phytoplankton biomass of salineeutrophic lakes within a small geographicarea. Arch. Hydrobiol. 81(3): 304–326.

Barker, A.V. 2000. Search of literature andplanning of research on salt damage alonghighways in Massachusetts. MassachusettsHighway Department and University ofMassachusetts Cooperative TransportationResearch Program. 44 pp. (ResearchQuarterly Report).

Barrick, W.E. and H. Davidson. 1980. Deicingsalt spray injury in Norway maple asinfluenced by temperature and humiditytreatments. HortScience 15(2): 203–205.

BC MoTH (British Columbia Ministry ofTransportation and Highways). 1993.Groundwater contamination study: HeffleyCreek Pit, Thompson-Okanagan Region.Report prepared by R. Buchanan,Geotechnical and Materials Engineering,Victoria, British Columbia.

Beak. 1999. Ecotoxicity test results. Unpublisheddata for M.S. Evans. Beak International Inc.,Brampton, Ontario.

Beck, P., S. Kok and W.J. Snodgrass. 1994.Technology assessment — removal of totaldissolved solids from salty washwatersgenerated from vehicle washing operations.Report prepared for Ontario Ministry ofTransportation, Toronto, Ontario.

Beckerson, D.W., N.P. Cain, G. Hofstra,D.P. Ormrod and P.A. Campbell. 1980.A guide to: plant sensitivity to environmentalstress. Landscape Archit. May: 299–303.


Bedunah, D. and M.J. Trlica. 1979. Sodiumchloride effects on carbon dioxide exchangerates and other plant and soil variablesof ponderosa pine. Can. J. For. Res. 9(3): 349–353.

Beeton, A.M. 1965. Eutrophication of theSt. Lawrence Great Lakes. Limnol. Oceanogr. 10: 240–254.

Beeton, A.M. 1969. Changes in the environmentand the biota of the Great Lakes. In:Eutrophication: Causes, consequences,correctives. Printing and Publishing Office,National Academy of Sciences, Washington,D.C. pp. 150–187.

Bélanger, L. 2000. Personal communication.Section Faune et Habitats, EnvironnementCanada, Sainte-Foy, Québec.

Bellamy, P.E., R.F. Shore, D. Ardeshir,J.R. Treweek and T.H. Sparks. 2000. Roadverges as habitat for small mammals inBritain. Mammal Rev. 30(2): 131–139.

Belluck, D.A. 2000. Personal communication.Minnesota Department of Transport, St. Paul,Minnesota.

Benkman, C.W.1998. [Crossbills hit by vehicleswhile eating salt]. Personal communicationfrom C.W. Benkman (Dept. of Biology.New Mexico State University, Las Cruces,New Mexico) to Lorna Brownlee (CanadianWildlife Service, National Wildlife ResearchCentre, Ottawa). Email dated 18 February1998.

Benkman, C.W. 2000. [The scavenging of fincheskilled at roadside in Algonquin Park] Personalcommunication from Craig W. Benkman(Dept. of Biology. New Mexico StateUniversity, Las Cruces, New Mexico) toP. Mineau (Canadian Wildlife Service,National Wildlife Research Centre, Ottawa).Email dated 11 May 2000.

Bennett, A.F. 1990. Habitat corridors andthe conservation of small mammals in afragmented forest environment. LandscapeEcol. 4: 109–122.

Bennett, A.F. 1991. Roads, roadsides and wildlifeconservation: a review. In: D.A. Saunders andR.J. Hobbs (eds.), Nature conservation 2:The role of corridors. Surrey Beatty and Sons,Victoria, Australia. pp. 99–118.

Bennetts, R.E. and R.L. Hutto. 1985. Attractionof social fringillids to mineral salts:An experimental study. J. Field Ornithol. 56(2): 187–189.

Bertram, B.M. 1997. Sodium chloride. In:J.I. Kroschwitz (exec. ed.), M. Howe-Grant(ed.), Kirk-Othmer encyclopedia of chemicaltechnology. 4th edition. Vol. 22. John Wiley& Sons, New York, N.Y. p. 371.

Bilodeau, S. 1999. La fondeuse à neige de Cap-Rouge, un an après son démarrage.Vecteur Environnement 32(2): 70–73.

Birge, W.J., J.A. Black, A.G. Westerman,T.M. Short, S.B. Taylor, D.M. Bruser andE.D. Wallingford. 1985. Recommendationson numerical values for regulating iron andchloride concentrations for the purpose ofprotecting warmwater species of aquatic lifein the Commonwealth of Kentucky. KentuckyNatural Resources and EnvironmentalProtection Cabinet, Lexington, Kentucky(Memorandum of Agreement No. 5429).

Bishop, C.A., J. Struger, L.J. Shirose, L. Dunnand G.D. Campbell. 2000. Contaminationand wildlife communities in stormwaterdetention ponds in Guelph and the GreaterToronto Area, Ontario, 1997 and 1998. Part 2— Contamination and biological effects ofcontamination. Water Qual. Res. J. Can.35(3): 437–474.


Blinn, D.W., M. Hurley and L. Brokaw. 1981.The effects of saline seeps and restricted lightupon the seasonal dynamics of phytoplanktoncommunities within a southwestern (USA)desert canyon stream. Archiv. Hydrobiol. 92: 287–305.

Boucher, R.J. 1982. Penjajawoc Stream SaltStudy. Materials and Research Division,Maine Department of Transportation. 15 pp.(Technical Paper 82-85).

Bourbonniere, R.A. 1985. Unpublished data.National Water Research Institute, Burlington,Ontario.

Bourbonniere, R.A. 1998. Unpublished data.National Water Research Institute, Burlington,Ontario.

Bowen, G.S. and M.J. Hinton. 1998. Thetemporal and spatial impacts of road salt onstreams draining the Greater Toronto Area. In:Groundwater in a watershed context:proceedings. Ontario Ministry of theEnvironment and Geological Survey ofCanada, Burlington, Ontario. pp. 303–309.

Bowser, C.J. 1992. Groundwater pathways forchloride pollution of lakes. In: F.M. D’Itri(ed.), Chemical deicers and the environment.Lewis Publishers, Boca Raton, Florida. pp. 283–301.

Bradford, W. 1988. A plan to reduce wildlifemortality on transportation corridors in JasperNational Park. Unpublished managementplan. Natural Resources Conservation, JasperNational Park, Jasper, Alberta.

Bresler, E., B.L. McNeil and D.L. Carter. 1982.Saline and sodic soils. Advanced Series inAgriculture Sciences 10. Springer-Verlag,Berlin. 236 pp.

Briggins, D., J. Walsh, C.L. Lin, K.A. Macdermidand R.H. Johnson (eds.). 1989. Theenvironmental implications of road salting in

Nova Scotia. Joint release of the Nova ScotiaDepartment of the Environment and theNova Scotia Department of Transportationand Communication. 40 pp.

Broderius, S.J. 1973. Determination of molecularhydrocyanic acid in water and studies of thechemistry and toxicity to fish of metalcyanide complexes. Ph.D. thesis, OregonState University, Corvallis, Oregon [cited inDoudoroff, 1976; U.S. EPA 1985].

Broderius, S.J. and L.L. Smith, Jr. 1980. Directphotolysis of hexacyanoferrate complexes:proposed applications to the aquaticenvironment. Environmental ResearchLaboratory, U.S. Environmental ProtectionAgency, Duluth, Minnesota (Technical ReportEPA-600/3-80-003).

Brownlee, L., P. Mineau and A. Baril. 2000. Riskcharacterization for wildlife. Report submittedto the Environment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts, May 2000. CommercialChemicals Evaluation Branch, EnvironmentCanada, Hull, Quebec.

Bubeck, R.C., W.H. Diment, B.L. Deck,A.L. Baldwin and S.D. Lipton. 1971.Runoff of deicing salt: Effect on IrondequoitBay, Rochester, New York. Science 172: 1128–1132.

Bubeck, R.C., W.W. Staubitz, A.D. Weidemannand L.P. Spittal. 1995. Physical and chemicallimnology of Ides Cove near Rochester, NewYork, 1970–1982. U.S. Geological Survey,Ithaca, New York. 44 pp. (Water-ResourcesInvestigations Report 92-4052).

Butler, B.J. and J. Addison. 2000. Biologicaleffects of road salts on soils. Report submittedto the Environment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts, March 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec.


Button, E.F. and D.E. Peaslee. 1966. The effectof rock salt upon roadside sugar maples inConnecticut. Highw. Res. Rec. 161: 121– 131.

Cain, N.P., B. Hale, E. Berkelaar and D. Morin.2001. Critical review of effects of NaCl andother road salts on terrestrial vegetation inCanada. Report submitted to the EnvironmentCanada CEPA Priority Substances ListEnvironmental Resource Group on RoadSalts. July 2001. Existing Substances Branch,Environment Canada, Hull, Quebec.

Camp, M. and L.B. Best. 1993a. Nest densityand nesting success of birds in roadsidesadjacent to rowcrop fields. Am. Midl. Nat. 131: 347–358.

Camp, M. and L.B. Best. 1993b. Bird abundanceand species richness in roadsides adjacentto Iowa rowcrop fields. Wildl. Soc. Bull. 21: 315–325.

Canadian Salt Company Limited. 1991. Windsorsalt, Safe-T-Coarse - Highway. Point Claire,Quebec (PDS VIII-3 5/91).

Catling, P.M. and S.M. McKay. 1980. Halophyticplants in southern Ontario. Can. Field-Nat. 94(3): 248–258.

CCME (Canadian Council of Ministers of theEnvironment). 1991. Canadian water qualityguidelines. Environment Canada, Ottawa,Ontario.

CCME (Canadian Council of Ministers of theEnvironment). 1998. Guidelines for Canadiandrinking water quality. Environment Canada,Ottawa, Ontario.

CFI (Canadian Fertilizer Institute). 1997.[Regarding report on potassium chloride inpotash mine tailings used for salting roads].Personal communication from J. Farrell(Manager, Technical Affairs, CFI) to R. Chenier(Chemicals Evaluation Division, EnvironmentCanada, Hull). Letter dated April 18, 1997.

Chang, N.-Y., W. Pearson, J.I.J. Chang, A. Gross,M. Meyer, M. Jolly, B. Vang and H. Samour.1994. Environmentally sensitive sanding anddeicing practices. Colorado TransportationInstitute and Colorado Department ofTransportation, Denver, Colorado. 116 pp.

Chatwin, T.D. 1989. Cyanide attenuation/degradation in soil (final report). ResourceRecovery and Conservation Company(R2C2), Salt Lake City, Utah.

Chatwin, T.D. 1990. Cyanide attenuation/degradation in soil (Appendix). ResourceRecovery and Conservation Company(R2C2), Salt Lake City, Utah.

Cheminfo. 1999. Socio-economic background andoptions study on the Canadian salt industry.Report prepared by Cheminfo, Toronto,Ontario, for Regulatory and AssessmentBranch, Environment Canada, Hull, Quebec.

Chenevier, C. 1997. Synthèse des donnéesphysico-chimiques sur les neiges usées auQuébec et leurs impacts environnementaux.M.Sc.A., Département des génies civile,géologique et des mines, École Polytechniquede Montréal, Montréal, Québec.

Cherkauer, D.S. and N.A. Ostenso. 1976. Theeffect of salt on small artificial lakes. WaterRes. Bull. 12(6): 1259–1266.

Chetelat, J., F.R. Pick, A. Morin andP.B. Hamilton. 1999. Periphyton biomass andcommunity composition in rivers of differentnutrient status. Can. J. Fish. Aquat. Sci. 56: 560–569.

Chong, C. and G.P. Lumis. 1990. Reduction ofsalt build-up and twig injury in roadsidepeach trees with film-forming sprays. Transp.Res. Rec. 1279: 45–53.

CIS (Camford Information Services). 1994.Sodium chloride. CPI Product Profiles.Don Mills, Ontario. 3 pp.


CIS (Camford Information Services). 1996.Calcium chloride. CPI Product Profiles.Don Mills, Ontario. 3 pp.

City of Montréal. 1998. Actualisation etcomplément d’information relatifs auprogramme d’assainissement des lieuxd’élimination de neige de la Ville deMontréal. Montréal, Quebec. 28 pp.

City of Regina. 1997. Soil investigation at citysnow dump sites 1996–1997. City of ReginaMunicipal Engineering Department, Regina,Saskatchewan. 11 pp. + appendices.

City of Toronto. 2001. [Unpublished monitoringdata for Toronto area watercourses]. Personalcommunication from G. Gris (Works andEmergency Services, City of Toronto) toD. Morin (Chemicals Evaluation Division,Environment Canada). April and May 2001.

Clark, M.J.R., H. Hanssen, G. Van Aggelen andS. Horvath. 1984. Acute toxicity of iron cyanidespecies to rainbow trout and to Daphnia magnaunder exposure to different light intensities. In:Proceedings of the 11th Annual Aquatic ToxicityWorkshop. British Columbia Ministry of theEnvironment, Vancouver, British Columbia [citedin Higgs, 1992].

Collins, B. 2000. Comparison of salt use amongyears. Memorandum dated November 3,2000. Migratory Bird Populations Division,National Wildlife Research Centre,Environment Canada.

Conn, E.E. 1980. Cyanogenic compounds. Annu.Rev. Plant Physiol. 31: 433 [cited in Knowlesand Bunch, 1986].

Cordukes, A.J. and W.E. Maclean. 1973.Tolerance of some turfgrass species todifferent concentrations of salt in soils. Can.J. Plant Sci. 53: 69–73.

Cornell Laboratory of Ornithology. 2000a.Biennial irruptions of pine siskins acrossNorth America. http://birdsource.cornell.edu/results/irruptions.htm

Cornell Laboratory of Ornithology. 2000b.Population trends in evening grosbeak.http://birdsource.cornell.edu/features/evegro

Couture, P. 1997. La gestion des neiges usées :Dilemmes et contradictions. VecteurEnvironnement 30(6): 18–20.

Cowgill, U.M. and D.P. Milazzo. 1990. Thesensitivity of two cladocerans to water qualityvariables: Salinity and hardness. Arch.Hydrobiol. 120(2): 185–196.

Crowther, R.A. and H.B.N. Hynes. 1977.The effect of road deicing salt on the drift ofstream benthos. Environ. Pollut. 14: 113–126.

Cushman, J.R., V.A. Duff, G.H. Buteau,L.B. Aust, N. Caldwell and W. Lazer. 1991.Evaluation of calcium magnesium acetate androad salt for contact hypersensitivity potentialand dermal irritancy in humans. ContactDermatitis 24: 289–292.

Davis, G., S. Krannitz and M. Goldstein. 1992.Phase II: A reconnaissance study of roadsidetree injury and decline at 17 sites in interiorBritish Columbia for the Roadside Tree InjuryCommittee. British Columbia Ministry ofEnvironment, Lands and Parks, Victoria,British Columbia. 98 pp. plus appendices.

Dawson, W.R., V.H. Shoemaker, H.B. Tordoff andA. Borut. 1965. Observations on metabolismof sodium chloride in the red crossbill. Auk 82: 606–623.

Delisle, C.E. 1994. La gestion des neiges usées.Routes et transports 24(4): 20–27.


Delisle, C.E. and L. Dériger. 2000.Caracterisation et élimination des neigesusées : impacts sur l’environnement. Reportsubmitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts. CommercialChemicals Evaluation Branch, EnvironmentCanada, Hull, Quebec.

Delisle, C.E. and A. Leduc. 1987. Évolution dansle temps et dans l’espace de la qualité de laneige usée et de l’eau de ruissellement depluie du territoire de la Ville de Montréal.Centre de développement technologique del’École Polytechnique de Montréal. Finalreport. 155 pp. + annexes.

Delisle, C.E., J.-F. Paradis and P. André. 1995.The Montréal, Quebec experience in usedsnow disposal and treatment. In: Ecologicalreclamation in Canada at century’s turn.Proceedings of the 35th Annual Meetingof the Canadian Society of EnvironmentalBiologists. Canadian Plains Research Center,Regina, Saskatchewan. pp. 10–18.

Delisle, C.E., C. Chenevier and P. André. 1997.Synthèse des données physico-chimiquessur les neiges usées au Québec et leursrépercussions environnementales. VecteurEnvironnement 30(6): 33–44.

Delisle, R. 1999. Autoroute 40 — Cap-de-la-Madeleine : Étude de la contamination dela nappe aquifère par les sels déglacants.Ministère des transports du Québec (Projet # 20-6373-9501).

Demers, C.L. 1992. Effects of road deicing salton aquatic invertebrates in four Adirondackstreams. In: F.M. D’Itri (ed.), Chemicaldeicers and the environment. LewisPublishers, Boca Raton, Florida. pp. 245–252.

Demers, C.L. and R.W. Sage, Jr. 1990. Effectsof road deicing salt on chloride levels in fourAdirondack streams. Water Air Soil Pollut. 49: 369–373.

Dickman, M.D. and M.B. Gochnauer. 1978.Impact of sodium chloride on the microbiotaof a small stream. Environ. Pollut. 17: 109–126.

Dirr, M.A. 1975. Effects of salts and applicationmethods on English ivy. HortScience 10(2): 182–184.

Dirr, M.A. 1976. Selection of trees for toleranceto salt injury. J. Arboric. 2(11): 209–216.

Dirr, M.A. 1978. Tolerance of seven woodyornamentals to soil-applied sodium chloride.J. Arboric. 4(7): 162–165.

Dixit, S.S., J.P. Smol, D.F. Charles, R.M. Hughes,S.G. Paulsen and G.B. Collins. 1999.Assessing water quality changes in the lakesof the northeastern United States usingsediment diatoms. Can. J. Fish. Aquat. Sci. 56: 131–152.

Dkhili, M. and B. Anderson. 1992. Salt effectson seedling growth of switchgrass and bigbluestem. In: D. Smith and C.C. Jacobs (eds.),Proceedings of the 12th North AmericanPrairie Conference 1990. University ofNorthern Iowa, Cedar Falls, Iowa. pp. 13–16.

Dobson, H.H. 1967. Principal ions and dissolvedoxygen in Lake Ontario. In: Proceedingsof the 10th Conference on Great LakesResearch. International Association forGreat Lakes Research, Ann Arbor, Michigan.Braun-Brumfield, Inc., Ann Arbor, Michigan. pp. 337–356.

Dobson, M.C. 1991. De-icing salt damage to treesand shrubs. Forestry Commission, Surrey,British Columbia. 64 pp. (Bulletin 101).

Doudoroff, P. 1966. Some experiments on thetoxicity of complex cyanides to fish. SewageInd. Wastes 28(8): 1020.


Doudoroff, P. 1976. Toxicity to fish of cyanidesand related compounds. A review.Environmental Research Laboratory,Office of Research and Development, U.S.Environmental Protection Agency, Duluth,Minnesota (EPA-600/3-76-038).

Doudoroff, P. 1980. A critical review of recentliterature on the toxicity of cyanides to fish.American Petroleum Institute, Washington,D.C.

Doudoroff, P. and M. Katz. 1953. Critical reviewof literature on the toxicity of industrialwastes and their components to fish. SewageInd. Wastes 25: 802–839 [cited in Evans andFrick, 2001].

Dowden, B.F. and H.J. Bennett. 1965. Toxicity ofselected chemicals to certain animals. J. WaterPollut. Control Fed. 37: 1308–1316 [cited inEvans and Frick, 2001].

Drake, L. and B. Kirchner. 1987. Protectingremnant natural communities along ruralroadsides. Nat. Areas J. 7(2): 79–80.

Dubey, S.K. and D.S. Holmes. 1995. Biologicalcyanide destruction mediated bymicroorganisms World J. Microbiol.Biotechnol. 11: 257.

Dvorak, P., M. Gunther, U. Zorn and A. Catsch.1971. Metabolisches verhalten vonkolloidalem ferrihexacyanoferrat (II).Naunyn-Schmiedebergs Arch. Pharmakol.269: 48–56 [cited in Shifrin et al., 1996].

Eaton, L.J., J. Hoyle and A. King. No date.The impact of road salt on lowbushblueberries. Nova Scotia Agricultural Collegeand Nova Scotia Department of Agricultureand Marketing, Truro, Nova Scotia.

Effenberger, E. 1964. Verunreinigunen eivesgrund-wasser durch cyanide. Arch. Hyg.Bakteriol. 148(4–5): 271–278 [cited inChatwin, 1990].

Eilers, J.M. and A.R. Selle. 1991. Changes inthe conductivity, alkalinity, calcium, andpH during a 50-year period in selectedWisconsin lakes. Can. J. Fish. Aquat. Sci. 46: 1929–1944.

Eilers, R.G., W.D. Eilers, W.W. Pettapiece andG. Lelyk. 1995. Salinization of soils. In:D.F. Acton and L.J. Gregorich (eds.), Thehealth of our soils. Agriculture and Agri-FoodCanada. pp. 77–86 (Publication 1906/E).

Eisler, R. 1991. Cyanide hazards to fish, wildlife,and invertebrates: A synoptic review. U.S.Fish and Wildlife Service (FWS Biol. Rep.85(1.23); NTIS No. PB94-120144).

Environment Canada. 1995. Toxicity testing ofNational Contaminated Sites RemediationProgram priority substances for thedevelopment of soil quality criteria forcontaminated sites. Retesting VolatileSubstances Guidelines Division, Evaluationand Interpretation Branch, EcosystemConservation Service.

Environment Canada. 1997a. Environmentalassessments of Priority Substances underthe Canadian Environment Protection Act.Guidance manual Version 1.0 — March 1997.Environmental Protection Series. CommercialChemicals Evaluation Branch, EnvironmentCanada (EPS/2/CC/3E).

Environment Canada. 1997b. CanadianEnvironmental Protection Act —Problem formulation for the environmentalassessment of the Priority Substance roadsalts. Government of Canada.

Evans, M. and C. Frick. 2001. The effects of roadsalts in stream, lake and wetland ecosystems.National Water Research Institute, Saskatoon,Saskatchewan.


Eyles, N. and K.W.F. Howard. 1988. Urbanlandsliding caused by heavy rains:Geochemical identification of recharge watersalong the Scarborough Bluffs, Toronto,Ontario. Can. Geotech. J. 25: 455–466.

Faber, J.H. and H.A. Verhoef. 1991. Functionaldifferences between closely-related soilarthropods with respect to decompositionprocesses in the presence or absence of pineroots. Soil Biol. Biochem. 23(1): 15–23.

Fallon, R.D., D.A. Cooper, R. Speece andM. Henson. 1991. Anaerobic biodegradationof cyanide under methanogenic conditions.Appl. Environ. Microbiol. 57(6): 1656–1662.

Ferguson, K.D. 1985. The photolysis of iron-cyanide in a stream receiving effluent froma gold cyanidation mill. EnvironmentalProtection Service, Pacific Region,Environment Canada.

Feth, J.H., C.E. Roberson and W.L. Polzer. 1964.Sources of mineral constituents in water fromgranitic rocks, Sierra Nevada, California andNevada. U.S. Geological Survey, Washington,D.C. (Water Supply Papers No. 1535-I).

Foster, H.D. 1993. Road de-icing salt and cancer:the need for further study. J. Natl. CancerInst. 85(19): 1603–1605.

Franson, J.C., L. Sileo and W.J. Flemming. 1981.Iatrogenic salt poisoning in captive sandhillcranes. J. Am. Vet. Med. Assoc. 179(11): 1211–1213.

Fraser, D. and E.R. Thomas. 1982.Moose–vehicle accidents in Ontario: Relationto highway salt. Wildl. Soc. Bull. 10(3):261–265.

Free, B.M. and G.G. Mulamoottil. 1983. Thelimnology of Lake Wabukayne, a storm-waterimpoundment. Water Res. Bull. 19(5): 821–827.

Freedman, B., J. Kerekes and G. Howell. 1989.Patterns of water chemistry among twenty-seven oligotrophic lakes in KejimkujikNational Park, Nova Scotia. Water Air SoilPollut. 46: 119–130.

Fuller, W.H. 1984. Cyanides in the environmentwith particular attention to soil. In:D. Van Zyl (ed.), Cyanide and theenvironment. Vol. 1. GeochemicalEngineering Program, Colorado StateUniversity, Fort Collins, Colorado. pp. 19–46.

Garrels, R.M. and F.T. Mackenzie. 1967. Originof the chemical compositions of some springsand lakes. In: Equilibrium concepts in naturalwater systems. American Chemical Society,Washington, D.C. pp. 222–242 (Advances inChemistry Series No. 67).

Gélinas, P. and J. Locat. 1988. Effets des selsdéglaçants sur la qualité de l’eau del’aquifère de Trois-Rivières Ouest.Ministère des transports du Québec (Report No. RTQ-87-05).

Gerber, R.E. and K.W.F. Howard. 1997. Ground-water to the Oak Ridges Moraine. In:N. Eyles (ed.), Environmental geologyof urban areas. Geological Associationof Canada, St. John’s, Newfoundland. pp. 173–192.

Getz, L.L., F.R. Cole and D.L. Gates. 1978.Interstate roadsides as dispersal routes forMicrotus pennsylvannicus. J. Mammal. 59: 208–212.

Gionfriddo, J.P. and L.B. Best. 1995. Grit use byhouse sparrows: Effects of diet and grit size.The Condor 97: 57–67.

Godbout, P. 1996. Chute à neige à l’égoutmotorisée et asservie — Une innovationquébécoise unique en Amérique du Nord.Les consultants BGH - Planning, Inc.Symposium, Hôtel Le Méridien, November19–21, 1996. Montréal, Quebec. 27 pp.


Gonzalez-Moreno, S., J. Gomez-Barrera,H. Perales and R. Moreno-Sanchez. 1997.Multiple effects of salinity on photosynthesisof the protist Euglena gracilis. Physiol. Plant. 101: 777–786.

Gooding, D. and A.J. Bodnarchuk. 1994. Fieldtrials of prewetting salt and sand with MgCl2

and CaCl2 brines: Efficiency and effects.Highway Environment Branch, BritishColumbia Ministry of Transportation andHighways. 7 pp. (MoTH-ENV-94-01).

Gosh, A.K. and R.N. Pal. 1969. Toxicity of fourtherapeutic compounds to fry of Indian majorcarps. Fish. Technol. 6: 120–123 [cited inHammer, 1977].

Gosz, J.R. 1977. Effects of ski area developmentand use in stream water quality on theSanta Fe Basin, New Mexico. For. Sci. 23(2): 167–179.

Gouvernement du Québec. 1991. Guide pourl’aménagement des lieux d’élimination desneiges usées. Ministère de l’Environnementdu Québec. 95 pp.

Gouvernement du Québec. 1997. Gestion desneiges usées : Guide pour l’aménagement deslieux d’élimination de neige et mise en oeuvredu règlement sur les lieux d’élimination deneige. Ministère de l’Environnement duQuébec. 56 pp.

Gouvernement du Québec. 2001. Guided’aménagement des lieux d’élimination deneige et mise en oeuvre du Règlement dur leslieux d’élimination de neige. Ministère del’environnement du Québec. Downloaded on June 13, 2001, from:www.menv.gouv.qc.ca/matieres/neiges_usees/index.htm

Grenier, P. 1973. Moose killed on the highway inthe Laurentides Park, Quebec, 1962 to 1972.Proc. N. Am. Moose Conf. Workshop 9: 155–193.

Hall, R., G. Hofstra and G.P. Lumis. 1972. Effectsof deicing salt on eastern white pine: foliarinjury, growth suppression, and seasonalchanges in foliar concentrations of sodiumand chloride. Can. J. For. Res. 2: 224–249.

Hamilton, R.W., J.K. Buttner and R.G. Brunetti.1975. Lethal levels of sodium chloride andpotassium chloride for an oligochaete, achironomid midge, and a caddisfly of LakeMichigan. Environ. Entomol. 4: 1003–1006.

Hammer, U.T. 1977. The effects of alkali halideson fish and other aquatic fauna. In: Theeffects of alkali halides in the Canadianenvironment. Associate Committee onScientific Criteria for Environmental Quality,National Research Council of Canada,Ottawa, Ontario. pp. 99–115 (NRCCPublication No. 15019).

Hammer, U.T. 1984. The saline lakes of Canada.In: F.B. Taub (ed.), Lakes and reservoirs.Elsevier Scientific Publishing Company,Amsterdam. pp. 521–540.

Hammer, U.T. 1986. Saline lake resources ofthe Canadian Prairies. Revue canadienne desressources en eau 11(1): 43–57.

Hanes, R.E., L.W. Zelazny and R.E. Blaser. 1970.Effects of deicing salts on water quality andbiota: Literature review and recommendedresearch. Highway Research Board, NationalAcademy of Sciences/National Academy ofEngineering, Washington, D.C. 70 pp.

Harrington, J.A. and T. Meikle. 1994. Road salteffects on the germination of eight selectprairie species. In: R.G. Wickett, P.D. Lewis,A. Woodliffe and P. Pratt (eds.), Proceedingsof the 13th North American PrairieConference, 1992. Department of Parks andRecreation, Windsor, Ontario. pp. 183–192.


Hart, W.C. 1985. Effects on the chemistry ofwater supply lakes due to chloride inputs.Centre for Water Resources Research,Technical University of Nova Scotia,Halifax, Nova Scotia. 16 pp. [cited inBriggins et al., 1989].

Hart, W.C. 1988. Chloride measured in thevicinity of Second Chain Lake, 1987–1988.Centre for Water Resources Research,Technical University of Nova Scotia, Halifax,Nova Scotia [cited in Briggins et al., 1989].

Hart, B.T., P. Bailey, R. Edwards, K. Hortle,K. James, A. McMahon, C. Meredith andK. Swadling. 1990. Effects of salinity onriver, stream, and wetland ecosystems inVictoria, Australia. Water Res. 24: 1103–1107.

Hattori, R. and T. Hattori. 1980. Sensitivity tosalts and organic compounds of soil bacteriaisolated on diluted media. J. Gen. Microbiol. 26: 1–14.

Hawkins, R.H. and J.H. Judd. 1972. Waterpollution as affected by street salting. WaterResour. Bull. 8: 1246–1252.

Headley, D. B. and N. Bassuk. 1991. Effect oftime and application of sodium chloride in thedormant season on selected tree seedlings.J. Environ. Hortic. 9(3): 130–136.

Health Canada. 1996. Guidelines for Canadiandrinking water quality. 6th edition. Ottawa,Ontario.

Henny, C.J., R.J. Hallock and E.F. Hill. 1994.Cyanide and migratory birds at gold mines inNevada, USA. Ecotoxicology 3: 45–58.

Herlihy, A.T., J.L. Stoddard and C.B. Johnson.1998. The relationship between streamchemistry and watershed cover data in themid-Atlantic region, U.S. Water Air SoilPollut. 105: 377–386.

Higgs, T.W. 1992. Technical guide for theenvironmental management of cyanide inmining. Prepared for Cyanide Subcommittee,British Columbia Technical and ResearchCommittee on Reclamation, by T.W. HiggsAssociates Ltd., Vancouver, British Columbia.

Hinton, M.J. and A.G. Eversole. 1978. Toxicityof ten commonly used chemicals to Americaneels. Agricultural Experiment Station,U.S. Department of Agriculture (TechnicalContribution 1595).

Hoffman, R.W., C.R. Goldman, S. Paulson andG.R. Winters. 1981. Aquatic impacts ofdeicing salts in the central Sierra NevadaMountains, California. Water Res. Bull. 17: 280–285.

Hofstra, G. and G.P. Lumis. 1975. Levels ofdeicing salt producing injury on apple trees.Can. J. Plant Sci. 55(1): 113–115.

Hofstra, G. and D.W. Smith. 1984. The effectsof road deicing salt on the levels of ions inroadside soils in southern Ontario. J. Environ.Manage. 19: 261–271.

Hofstra, G., R. Hall and G.P. Lumis. 1979.Studies of salt-induced damage to roadsideplants in Ontario. J. Arboric. 5: 25–31.

Holms, H.M. and J.L. Henry. 1982.Understanding salt-affected soils. PlantIndustry Branch, Saskatchewan Agriculture.14 pp.

Howard, K.W.F. and P.J. Beck. 1993.Hydrogeochemical implications ofgroundwater contamination by road de-icingchemicals. J. Contam. Hydrol. 12: 245–268.

Howard, K.W.F. and J. Haynes. 1993.Groundwater contamination due to road de-icing chemicals — salt balanceimplications. Geosci. Can. 20(1): 1–8.


Howard, K.W.F. and J. Haynes. 1997.Contamination of urban groundwater byroad de-icing chemicals. In: N. Eyles (ed.),Environmental geology of urban areas.Geological Association of Canada, St. John’s,Newfoundland.

Howard, K.W.F. and L.C. Taylor. 1998.Hydrogeochemistry of springs in urbanToronto, Canada. In: J. Van Brahana,Y. Eckstein, L.K. Ongley, R. Scheinderand J.E. Moore (eds.), Gambling withgroundwater — physical, chemical, andbiological aspects of aquifer-stream relations.American Institute of Hydrology, St. Paul,Minnesota. pp. 639–644.

Howard, K.W.F., J.I. Boyce, S.J. Livingstone andS.L. Salvatori. 1993. Road salt impacts onground-water quality — the worst is still tocome. GSA Today 3(12): 301–321.

Howard, P.H., R.S. Boethling, W.F. Jarvis,W.M. Meylan and E.M. Michalenko. 1991.Handbook of environmental degradationrates. Lewis Publishers, New York, N.Y. pp. 132–133.

Howell, J. and R.C. Gumbrell. 1992.Salt poisoning in broiler chickens. N. Z. Vet. J. 40: 85.

Hsu, M. 1984. Anticaking agents in deicing salt.Materials and Research Division, MaineDepartment of Transportation. 18 pp.(Technical Paper No. 84-4).

Hubbs, A.H. and R. Boonstra. 1995. Study designto assess the effects of highway medianbarriers on wildlife. Research andDevelopment Branch, Ontario Ministryof Transportation (MAT-94-03).

Huey, L.M. 1941. Mammalian invasion via thehighway. J. Mammal. 22: 383–385.

Humphreys, D.J. 1978. A review of recent trendsin animal poisoning. Br. Vet. J. 134: 128–141.

INTERA Consultants. 1996. Applicationguidelines for installation of oil/waterseparators at existing MTO patrol yards.Report submitted to Environmental Office,Ontario Ministry of Transportation, Toronto,Ontario.

Jeffries, D.S. 1997. 1997 Canadian acid rainassessment: The effects on Canada’s lakes,rivers and wetlands. Vol. 3. Aquatic effects.Environment Canada, Ottawa, Ontario.

Johnson, G.D. and P.B. Kauss. 1991. Tributarycontaminant inputs on the St. Clair andDetroit rivers and Lake St. Clair 1984–1985.Water Resources Branch, Ontario Ministry ofthe Environment, Toronto, Ontario.

Johnston, C.T., R.C. Sydor and C.L.S. Bourne.2000. Impact of winter road salting on thehydrogeologic environment — an overview.Stantec Consulting Ltd., Kitchener, Ontario.

Jones, P.H., B.A. Jeffrey, P.K. Watler andH. Hutchon. 1986. Environmental impactof road salting – State of the art. OntarioMinistry of Transport and Communications(MTC No. RR237).

Jones, P.H., B.A. Jeffery, P.M. Watler andH. Hutchon. 1992. Environmental impactof road salting. In: F.M. D’Itri (ed.),Chemical deicers and the environment. LewisPublishers, Chelsea, Michigan.

Judd, J.H. 1969. Lake stratification caused byrunoff from street deicing, Milwaukee,Wisconsin. Center for Great Lakes Studies.25 pp.

Judd, J.H. and J.W. Stegall. 1982. Reevaluation ofthe effects of deicing salt runoff on a smalllake. National Technical Information Service,Springfield, Virginia. 28 pp. (A-114-MICH).


Kaplan, D.L., R. Hartenstein, E.F. Neuhauserand M.R. Malecki 1980. Physicochemicalrequirements in the environment of theearthworm (Eisenia foetida). Soil Biol.Biochem. 12: 347–352.

Keizer, P.D., D.C. Gordon, J.W. Rowell,R. McCurdy, D. Borgal, T.A. Clair, D. Taylor,J.D. Ogden and G.E.M. Hall. 1993. Synopticwater quality survey of Halifax/DartmouthMetro Area lakes on April 16, 1991.Biological Services Branch, Scotia-FundyRegion, Department of Fisheries and Oceans.

Kelly, D., J. Underwood and J. Hansen. 1976.Influence of highway deicing salts upon waterquality in Chocolate Lake, Halifax, NovaScotia. Nova Scotia Department of theEnvironment, Halifax, Nova Scotia. 118 pp.

Kerekes, J.J. 1974. Limnological conditions infive small oligotrophic lakes in Terra NovaNational Park, Newfoundland. J. Fish. Res.Board Can. 31: 555–583.

Kerekes, J. and B. Freedman. 1989. Seasonalvariation of water chemistry in oligotrophicstreams and rivers in Kejimkujik NationalPark, NS. Water Air Soil Pollut. 46(1–4): 131–144.

Kerekes, J., B. Freedman, S. Beauchamp andR. Tordon. 1989. Physical and chemicalcharacteristics of three acidic oligotrophiclakes and their watersheds in KejimkujikPark, Nova Scotia. Water Air Soil Pollut. 46: 99–117.

Kersey, K. and R.J. Mackay. 1981. Laboratoryand field studies on the effects of road de-icing salt on stream invertebrates SIC-9.Snow and Ice Control Working Group,Institute for Environmental Studies,University of Toronto, Toronto, Ontario.

Khanna, C., H.J. Boermans and B. Wilcock. 1997.Fatal hypernatremia in a dog from saltingestion. J. Am. Anim. Hosp. Assoc. 33: 113–117.

Klaassen, C.D. (ed.). 1996. Casarett and Doull’stoxicology. The basic science of poisons.5th edition. McGraw-Hill, New York, N.Y.

Kliejunas, J., M. Marosy and J. Pronos. 1989.Conifer damage and mortality associated withhighway de-icing and snow removal in theLake Tahoe area. Forest Pest Management,U.S. Department of Agriculture ForestService. 18 pp. (Report No: 89-11).

Knowles, C.J. 1976. Microorganisms and cyanide.Bacteriol. Rev. 40: 652–680.

Knowles, C.J. and A.W. Bunch. 1986. Microbialcyanide metabolism. Adv. Microbial Physiol. 27: 73–111.

Kovacs, T.G. 1979. The effect of temperature onthe cyanide toxicity to rainbow trout: I, Acuteeffects. II, Chronic effects. M.Sc. thesis,Concordia University, Montréal, Quebec.69 pp. [cited in Leduc et al., 1982].

Kovacs, T.G. and G. Leduc. 1982. Sublethaltoxicity of cyanide to rainbow trout (Salmogairdneri) at different temperatures. Can. J.Fish. Aquat. Sci. 39: 1389–1395.

Kramer, J.R. 1964. Theoretical model for thechemical composition of fresh water withapplication to the Great Lakes. In:Proceedings of the 7th Conference on GreatLakes Research. Great Lakes ResearchDivision, Institute of Science and Technology,University of Michigan, Ann Arbor,Michigan. pp. 147–160 (Publication No. 11).

Kszos, L.A., J.D. Winter and T.A. Storch. 1990.Toxicity of Chautauqua Lake bridge runoffto young-of-the-year sunfish (Lepomismacrochirus). Bull. Environ. Contam. Toxicol. 45: 923–930.

Labadia, C.F. and J.M. Buttle. 1996. Road saltaccumulation in highway snow banks andtransport through the unsaturated zone of theOak Ridges Moraine, southern Ontario.Hydrol. Processes 10: 1575–1589.


Lacroix, G. and L. Bélanger. 2000. Suiviécologique du projet-pilote de gestiondifférenciée des emprises autoroutières: voletavifaune et mortalité par collision. Preparedfor the Department of Biological Chemistryof the Université du Québec at Trois-Rivièresand the Quebec Ministry of Transportation bythe Canadian Wildlife Service, EnvironmentCanada, Sainte-Foy, Quebec.

Laursen, K. 1981. Birds on roadside verges andthe effect of mowing on frequency anddistribution. Biol. Conserv. 20: 59–68.

Leduc, G., R.C. Pierce and I.R. McCracken. 1982.The effects of cyanides on aquatic organismswith emphasis on freshwater fishes. AssociateCommittee on Scientific Criteria forEnvironmental Quality, National ResearchCouncil of Canada, Ottawa, Ontario (NRCCNo. 19246).

Lee, K. 1985. Earthworms: Their ecology andrelationships with soils and land use.Academic Press, New York, N.Y.

Lefkovitch, L.P. and I. Fahrig. 1985. Spatialcharacteristics of habitat patches andpopulation survival. Ecol. Model. 30: 297–308.

Leland, H.V. and S.V. Fend. 1998. Benthicinvertebrate distributions in the San JoaquinRiver, California, in relation to physical andchemical factors. Can. J. Fish. Aquat. Sci. 55:1051–1067.

Letts, A. 1999. [The melting of salt applied toroads]. Personal communication from A. Letts(Technical Manager, Morton InternationalInc., Chicago, Illinois) in conversation withL. Brownlee (Canadian Wildlife Service,National Wildlife Research Centre, Ottawa).

Letts, A.D. 2000a. Effects of sodium ferrocyanidederived from road salting on the ecosystem.Morton Salt, Morton International Inc.,Chicago, Illinois.

Letts, A.D. 2000b. A model to predict chlorideconcentrations in meltwater runoff from roadsalt. Morton Salt, Morton International Inc.,Chicago, Illinois.

Lewis, S.A. 1991. The conservation andmanagement of roadside vegetation in SouthAustralia. In: D.A. Saunders and R.J. Hobbs(eds.), Nature conservation 2: The role ofcorridors. Beatty & Sons, Surrey. pp. 314–318.

Liang, W. 1998. Performance monitoring of theRouge River stormwater management facility.Draft report. Stormwater AssessmentMonitoring and Performance Program,Ontario Ministry of the Environment,Etobicoke, Ontario.

Lumis, G.P., G. Hofstra and R. Hall. 1973.Sensitivity of roadside trees and shrubs toaerial drift of deicing salt. Can. J. Plant Sci. 8(6): 475–477.

Lumis, G.P., G. Hofstra and R. Hall. 1976.Roadside woody plant susceptibility tosodium and chloride accumulation duringwinter and spring. Can. J. Plant Sci. 56: 853–859.

Lumis, G.P., G. Hofstra and R. Hall. 1977. Saltdamage to roadside plants. Ontario Ministryof Agriculture and Food Factsheet. 3 pp.(Order No. 77-063 Dec. 1977).

Lumis, G.P., G. Hofstra and R. Hall. 1983. Saltdamage to roadside plants. Ontario Ministryof Agriculture and Food Factsheet. 3 pp.

MacLeod, C.L., M.P. Borcsik and P.R. Jaffe.1996. Effect of infiltrating solutions onthe desorption of mercury from aquifersediments. Environ. Technol. 17: 465–475.

Malmqvist, P.-A. 1985. Environmental effects ofsnow disposal in Sweden. Molndalsvagen,Göteborg, Sweden. 10 pp.


Maltby, L., D.M. Farrow, A.B.A. Boxall, P. Calowand C.I. Betton. 1995. The effects ofmotorway runoff on freshwater ecosystems: 1. Field study. Environ. Toxicol. Chem. 14: 1079–1092.

Martineau, D. and S. Lair. 1995. Nécropsie.Canadian Cooperative Wildlife Health Centre,Saint-Hyacinthe, Quebec (unpublishedpathology report 95-1893).

Mattson, M.D. and P.J. Godfrey. 1994.Identification of road salts contaminationusing multiple regression and GIS. J. Environ.Manage. 18: 767–773.

Mayer, T. 2000. Unpublished data. National WaterResearch Institute, Burlington, Ontario.

Mayer, T., J. Marsalek and E. Delos Reyes. 1996.Nutrients and metal contaminants status ofurban stormwater detention ponds. LakeReserv. Manage. 12(3): 348–368.

Mayer, T., Q. Rochfort, J. Marsalek, M. Servos,A. Jurkovic and R. McKinnis. 1998. Effectsof de-icing salts on the chemistry andtoxicology of highway runoff. National WaterResearch Institute, Burlington, Ontario.

Mayer, T., W.J. Snodgrass and D. Morin. 1999.Spatial characterization of the occurrence ofroad salts and their environmentalconcentrations as chlorides in Canadiansurface waters and benthic sediments. WaterQual. Res. J. Can. 34(4): 545-574. See also:Mayer, T., W.J. Snodgrass and D. Morin.1999. Spatial characterization of theoccurrence of road salts and theirenvironmental concentrations as chlorides inCanadian surface waters and benthicsediments. Environment Canada. NWRIcontribution No. 99-097 (revised 2001).44 pp.

McBean, E. and S. Al-Nassri. 1987. Migrationpattern of de-icing salts from roads.J. Environ. Manage. 25: 231–238.

McCormick, R.W. and D.C. Wolf. 1980. Effectsof sodium chloride on CO2 evolution,ammonification and nitrification in a sassafrassandy loam. Soil Biol. Biochem. 12: 153–157.

McDonald, M.G. and A.W. Harbaugh. 1988.A modular three-dimensional finite-differenceground-water flow model. U.S. GeologicalSurvey, Washington, D.C. 586 pp.(Techniques of Water ResourcesInvestigations, Book 6).

McKee, J.E. and H.W. Wolf. 1963. Water qualitycriteria. 2nd edition. California State WaterResources Control Board. 542 pp.

McKillop, W.B., R.T. Patterson, L.D. Delormeand T. Nogrady. 1992. The origin, physico-chemistry and biotics of sodium chloridedominated saline waters on the western shoreof Lake Winnipegosis, Man. Can. Field-Nat.106(4): 454–473.

McLaughlin, D.L. and R.G. Pearson. 1981.Effectiveness of snow fence barriers inreducing salt spray injury to fruit orchardsadjacent to the Queen Elizabeth Highway:1978–1981. Phytotoxicology Section, AirResources Branch, Ontario Ministry of theEnvironment, Brampton, Ontario.

McNeely Engineering. 1990. Snow disposal studyfor the designated growth centres. Reportpresented to the Transportation Department,Regional Municipality of Ottawa-Carleton,Ottawa, Ontario.

McNeely-Tunnock Ltd. 1995. Strategic plan forsnow disposal inside the Greenbelt. Finalreport presented to the Regional Municipalityof Ottawa-Carleton, Ottawa, Ontario.

MDOT (Michigan Department of Transportation).1993. The use of selected de-icing materialson Michigan roads: Environmental andeconomic impacts. 105 pp.


Meade, G.M. 1942. Calcium chloride — a deathlure for crossbills. Auk 59: 439–440.

Meeussen, J.C.L., M.G. Keizer and F.A.M.De Haan. 1992a. Chemical stability anddecomposition rate of iron cyanide complexesin soil solutions. Environ. Sci. Technol. 26(3): 511–516.

Meeussen, J.C.L., M.G. Keizer, W.H. vanRiemsdijk and F.A.M. de Haan. 1992b.Dissolution behavior of iron cyanide (prussianblue) in contaminated soils. Environ. Sci.Technol. 26(9): 1832–1838.

Melis, L.A., P.I. Tones and S.M. Swanson. 1987.Identification and evaluation of impactsresulting from the discharge of gold milleffluents. Report prepared for EnvironmentCanada. 121 pp. (Project No. 52SS.KE145-6-095) [cited in Higgs, 1992].

Meteyer, C.U., R.R. Dubielzig, F.J. Dein,L.A. Baeten, M.K. Moore, J.R. Jehl, Jr. andK. Wesenberg. 1997. Sodium toxicity andpathology associated with exposure ofwaterfowl to hypersaline playa lakes ofsoutheast New Mexico. J. Vet. Diagn. Invest. 9: 269–280.

Miller, B.K. and J.A. Litvaitis. 1992. Use ofroadside salt licks by moose, Alces alces, innorthern New Hampshire. Can. Field-Nat. 106: 112–117.

Miller, G.C., V.R. Hebert and W.W. Miller. 1989.Effect of sunlight on organic contaminants atthe atmosphere–soil interface. Soil ScienceSociety of America, Madison, Wisconsin. pp. 99–110. (SSSA Special Publication No. 22).

Mills, G.E. and A.V. Barker. 2001. Sodiumaccumulation in soils and plants alongMassachusetts roadsides. In: Proceedingsof the Annual Meeting of Northeast RegionAmerican Society for Horticultural Science(Abstract).

Mineau, P. and B.T. Collins. 1988. Avianmortality in agro-ecosystems 2. Methods ofdetection. In: M.P. Greaves, B.D. Smith andP.W. Greig-Smith (eds.), Field methods forthe study of environmental effects ofpesticides. British Crop Protection Council,Croydon, U.K. pp. 13–27.

Ministère des Transports du Québec. 1980.Prévention contre la pollution du Lac àla Truite en hiver. Document prepared byFrançois Berard, Office des autoroutes duQuébec. Document obtained from the QuebecMinistry of Transportation’s document centre.

Ministère des Transports du Québec. 1999.Personal communication from G. D’Astous(Direction des Laurentides-Lanaudière)and A. Décréon (Service de l’aménagementdes infrastructures et de l’environnement) toD. Morin (Environment Canada) regardingthe impact of road salt loadings on lac à laTruite.

Ministers’ Expert Advisory Panel. 1995. Reportof the Ministers’ Expert Advisory Panel onthe second Priority Substances List underthe Canadian Environmental Protection Act(CEPA). Government of Canada, Ottawa,Ontario. 26 pp.

Mitsch, W.J. and J.G. Gosselink. 1986. Wetlands.Van Nostrand Reinhold Company, New York,N.Y. 539 pp.

Molles, M.C. and J.R. Gosz. 1980. Effects of skiarea on the water quality and invertebrates ofa mountain stream. Water Air Soil Pollut. 14: 187–205.

Morin, D. 2000. Environmental impacts fromthe spreading and storage of road salts: Casestudies. Supporting document for the roadsalts PSL assessment. Report submitted tothe Environmental Resource Group for RoadSalts, Commercial Chemicals EvaluationBranch, Environment Canada. May 2000.84 pp.


Morin, D. and M. Perchanok. 2000. Road saltloadings in Canada. Supporting documentfor the road salts PSL assessment. Reportsubmitted to the Environment Canada CEPAPriority Substances List EnvironmentalResource Group on Road Salts, May 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec. 85 pp.

Morin, D., W. Snodgrass, J. Brown and P.A. Arp.2000. Impacts evaluation of road salt loadingson soils and surface waters. Supportingdocument for the road salts PSL assessment.Report submitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts, June 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec. 76 pp.

Mortsch, L.D. and F.H. Quinn. 1996. Climatechange scenarios for Great Lakes Basinecosystem studies. Limnol. Oceanogr. 41: 903–911.

MTO (Ontario Ministry of Transportation).1993–1999. Phase 1 and/or Phase 2environmental site assessments done byDST Consulting Engineers, Oliver MangioneMcCalla & Associates, NAR EnvironmentalConsultants, Conestoga-Rovers & Associatesor Ecoplans for the following patrol yardsowned and operated by the Ontario Ministryof Transportation: Alexandria, Almonte,Apsley, Barriefield, Belleville Sub PY,Bloomfield, Bobcaygeon, Brighton,Calabogie, Camden East, Carlsbad Springs,Cloyne, Coboconk, Crystal Beach, Denbigh,Eganville, Erinsville, Grafton, Greely,Hawkesbury, Johnstown, Kemptville,Lansdowne, Lorneville, Madawaska, Madoc,Marysville, Morrisburg, Navan, Norwood,Pembroke, Roseneath, Sharbot Lake,Summerstown, Toledo, Tory Hill, Trenton SubPY, Vankleek Hill, Verona, Winchester, WolfIsland, Youngs Point.

MTO (Ontario Ministry of Transportation). 1997.Graphical compilation of chlorideconcentrations in patrol yards. Managedby the Ontario Ministry of Transportation,Toronto, Ontario.

Mudder, T. 1991. Personal communication.Steffen Robertson and Kristen, Seattle,Washington [cited in Higgs, 1992].

Nadec (Nadec Experts & Consultants Inc.). 2001.The ecological significance of roadsidehabitat — Revised final draft. Presented toCommercial Chemicals Branch, EnvironmentCanada, Hull, Quebec. March.

Natural Resources Canada. 1998. Mineraland metal commodity reviews — salt. In:Canadian minerals yearbook, 1998. Mineralsand Metals Sector, Natural Resources Canada.

NB DOE and DOT (New Brunswick Departmentof Environment and Department ofTransportation). 1978. An investigationof the quantity and quality of leachate from ahighway salt treated sand pile under normaloperating conditions. Prepared by theWater Resources Branch, New BrunswickDepartment of Environment, for and incooperation with the New BrunswickDepartment of Transportation.

Neehall, R. 2000. Comments on the draftassessment report on salts, City of Edmonton,Office of the Environment, dated October 11,2000.

Northcote, T.G. and T.G. Halsey. 1969. Seasonalchanges in the limnology of some meromicticlakes in southern British Columbia. J. Fish.Res. Board Can. 26: 1763–1787.

Northover, J. 1987. NaCl injury to dormantroadside peach trees and its effect on theincidence of infections by Leucostoma spp.Phytopathology 77(6): 835–840.


Noss, R.F. 1983. A regional landscape approach tomaintain diversity. Bioscience 33: 700–706.

Novotny, V., D. Muehring, D.W. Smith,D.H. Zitomer and R. Facey. 1997. Impactof nonpoint source snowmelt on receivingwaters. In: Proceedings of the WaterEnvironmental Federation 70th AnnualConference, October 19, Chicago, Illinois(Workshop #118).

OECD Scientific Expert Group. 1989. Curtailingusage of de-icing agents in wintermaintenance. Organisation for Economic Co-operation and Development, Paris, France.

Oeser, R., Jr. 1977. The crossbill (Loxiacurvirostra) — Highway mortalities inthe Fichtelbers area. Beitr. Vogelkd. 23(5): 278–280.

Oetting, R.B. and J.F. Cassel. 1971. Waterfowlnesting on interstate highway right-of-wayin North Dakota. J. Wildl. Manage. 35(4): 774–781.

Ohno, T. 1989. Determination of levels of freecyanide in surface and in ground watersaffected by MDOT salt storage facilities.Technical Services Division, MaineDepartment of Transportation, Orono, Maine(Technical Report 86C:16-17).

Ohno, T. 1990. Levels of total cyanide and NaClin surface waters adjacent to road salt storagefacilities. Environ. Pollut. 67: 123–132.

OME (Ontario Ministry of the Environment).1991. Stormwater quality best managementpractices. Environmental Sciences &Standards/Water Resources, OntarioMinistry of the Environment. June(ISBN 0-7729-8478-6).

OMEE (Ontario Ministry of Environment andEnergy). 1993. A study of dust suppressantsin Ontario. Final report. Prepared by CH2MHill Engineering Ltd. for Waste ManagementBranch, Ontario Ministry of Environment andEnergy, Toronto, Ontario.

Otake, T., S. Tone, I. Komasawa and M.Kawashima. 1982. Mechanism and kinetics ofdecomposition of water soluble cyanoferratesby ultraviolet radiation. Osaka University,Osaka, Japan [translation].

Parker, A. and J.D. Mather. 1979. Aninvestigation into pollution from a disusedgasworks site near Ladybank, Fife. AtomicEnergy Research Establishment, Harwell,Berkshire. 19 pp. (AERE-R 9213) [cited inChatwin, 1990].

Patrick, R., J.J. Cairns and A. Scheier. 1968. Therelative sensitivity of diatoms, snails and fishto twenty common constituents of industrialwastes. Prog. Fish-Cult. 30: 137–140.

Paul, E.A. and F.E. Clark. 1996. Soilmicrobiology and biochemistry. 2nd edition.Academic Press, San Diego, California.

Paul, R., M. Rocher and R. Impens. 1984.[Influence of winter deicing with CaCl2 onmountain ash, maple, lime and plane.] Bull.Soc. R. Bot. Belg. 117: 277–284 (in French).

Péloquin, Y. 1993. Qualité des eaux de fonteprovenant d’un site de surface pourl’élimination des neiges usées à la Ville deLaval. Université du Québec à Montréal,Montréal, Québec. 91 pp.

Perchanok, M.S., D.G. Manning andJ.J. Armstrong. 1991. Highway deicers:Standards, practice and research in theProvince of Ontario. Research andDevelopment Branch, Ontario Ministry ofTransportation, Downsview, Ontario. 53 pp.(MAT-91-13).


Perring, F.H., P.D. Sell, S.M. Walters andH.L.K. Whitehouse. 1964. A flora ofCambridgeshire. Cambridge University Press,Cambridge, U.K.

Phillips, A.M.J. 1944. The physiological effect ofsodium chloride upon brook trout. Trans. Am.Fish. Soc. 74: 297–309.

Phippen, B.W., D.C. Parks, L.G. Swain,R. Nordin, C.J.P. McKean, J.B. Holms,P.D. Warrington, R. Nijman, J. Denisegerand L. Erickson. 1996. A ten year assessmentof water quality in six acid-rain sensitiveBritish Columbia lakes (1984–1994).British Columbia Ministry of Environment,Lands and Parks. 250 pp.

Pilon, P. and K.W.F. Howard. 1987. Originsof elevated chloride concentrations ingroundwater systems. Ontario Ministry of theEnvironment Technology Transfer ConferenceNo. 7, Toronto, December; Part C — Liquidand Solid Waste Research. pp. 376–398(ISSN 0-825-4591).

Pinard, D., J.B. Sérodes and P.A. Côté. 1989.Charactérisation des eaux de fonte d’un dépotà neiges usées. Sci. Tech. Eau 22(3): 211-215.

Pringle, C.M., D.S. White, C.P. Rice andM.L. Tuchman. 1981. The biological effectsof chloride and sulfate with the specialemphasis on the Laurentian Great Lakes.University of Michigan, Ann Arbor,Michigan. pp. 1–51.

Prowse, C.W. 1987. The impact of urbanizationon major ion flux through catchments: a casestudy in southern England. Water Air SoilPollut. 32: 277–292.

Prud’homme, M. 2000. Potash. In: Canadianminerals yearbook, 1999. Mineralsand Metals Sector, Natural ResourcesCanada. Downloaded on November 10, 2000,from www.nrcan.gc.ca/mms/cmy/1999CMY_E2.html.

Racette, D.J. and H.D. Griffin. 1989. Vegetationassessment surveys near the Ministry ofTransportation’s Longbow Lake Patrol Yard1988–1989. Ontario Ministry of theEnvironment, Technical Support Section,N.W. Region. 9 pp.

Rader, W.S., L. Solujic, E.B. Milosavljevicand J.L. Hendrix. 1993. Sunlight-inducedphotochemistry of aqueous solutions ofhexaferrocyanate (II) and (III) ions. Environ.Sci. Technol. 27(9): 1875–1879.

Rantamaki, J., L. Cerenius and K. Soderhall.1992. Prevention of transmission of thecrayfish plague fungus (Aphanomyces astaci)to the freshwater crayfish (Astacus astacus)by treatment with MgCl2. Aquaculture 104: 11–18.

Reznicek, A.A. 1980. Halophytes along aMichigan roadside with comments on theoccurrence of halophytes in Michigan. Mich.Bot. 19: 23–30.

Reznicek, A.A., P.M. Catling and S.M. McKay.1976. Carex praegracilis W. Boott, recentlyadventive in southern Ontario. Can. Field-Nat. 90: 180–183.

Rodgers, G.K. 1998. Winter stratification inHamilton Harbour. National Water ResearchInstitute, Burlington, Ontario (NWRIContribution 98-231).

Rodgers, G.K. 1999. Personal communicationregarding data in NWRI Contribution 98-231[see Rodgers, 1998]. National Water ResearchInstitute, Burlington, Ontario.

Rott, E., H.C. Duthie and E. Pipp. 1998.Monitoring organic pollution andeutrophication in the Grand River, Ontario,by means of diatoms. Can. J. Fish. Aquat. Sci. 55: 1443–1453.

Rowe, J.S. 1972. Forest regions of Canada.Canadian Forestry Service, Department ofthe Environment, Ottawa, Ontario. 172 pp.


Rushton, D. 1999. [Data regarding Bible Hill,Saulnierville and Oxford patrol yards].Personal communication from D. Rushton(Environmental Services, Nova ScotiaDepartment of Transportation and PublicWorks) to D. Morin (Chemicals EvaluationDivision, Environment Canada). October andNovember, 1999.

Rushton, D. 2000. [Data regarding Bible Hill,Saulnierville and Oxford patrol yards].Personal communication from D. Rushton(Environmental Services, Nova ScotiaDepartment of Transportation and PublicWorks) to D. Morin (Chemicals EvaluationDivision, Environment Canada). Winter,2000.

Ruttner, F. 1963. Fundamentals of limnology.Translated by D.G. Frey and F.E.J. Fry.University of Toronto Press, Toronto, Ontario. 295 pp.

Salt Institute. 1964–1983. Survey of salt, calciumchloride, and abrasives use in the UnitedStates and Canada for 1964–65, 1966–67,1969–70, 1973–74, 1978–79, 1982–83.Alexandria, Virginia.

Sandals, W.C.D. 1978. Acute salt poisoning incattle. Can. Vet. J. 19: 136–137.

Saumure, R.A. 2001. Personal communication.Shark Reef at Mandalay Bay, Las Vegas,Nevada.

Scanton, D. 1999. Personal communication(letter dated May 19, 1999). Ontario Ministryof Transportation, Toronto, Ontario.

Scarratt, W.K., T.J. Collins and D.P. Sponenberg.1985. Water deprivation–sodium chlorideintoxication in a group of feeder lambs.J. Am. Vet. Med. Assoc. 186(9): 977–978.

Schauffler, M. 1993. Capture of road-salt aerosolsin an acidic peatland in central Maine.MA thesis, University of Maine. 41 pp.

Schelske, C. L., E. F. Stoermer, D. J. Conley,J. A. Robbins, and R. M. Glover. 1983. Earlyeutrophication in the lower Great Lakes: newevidence from biogenic silica in sediments.Science (Wash., DC) 222:320-322.

Schelske, C. L., D. J. Conley, and W. F. Warwick.1985. Historical relationships betweenphosphorus loading and biogenic silicaaccumulation in Bay of Quinte sediments.Can. J. Fish. Aquat. Sci. 42:1401-1409.

Schindler, D.W., S.E. Bayley, B.R. Parker,K.G. Beaty, D.R. Cruikshank, E.J. Fee,E.U. Schindler and M.P. Stainton. 1996. Theeffects of climatic warming on the propertiesof boreal lakes and streams at theExperimental Lakes Area, northwesternOntario. Limnol. Oceanogr. 41: 1004–1017.

Schmidt, J.W., L. Simovic and E. Shannon. 1981.Natural degradation of cyanide in gold milleffluents. Presented at Cyanide in GoldMining Seminar, January 1981, Ottawa,Ontario [cited in Higgs, 1992].

Schraufnagel, F.H. 1965. Chlorides. Committeeon Water Pollution, Wisconsin Public HealthService, Madison, Wisconsin. pp. 1–13 plusappendix.

Schroeder, J.E. and K.R. Solomon. 1998. Impactof road salts on an urban watershed. Posterpresented at 19th Annual Meeting of theSociety of Environmental Toxicology andChemistry (SETAC), Charlotte, NorthCarolina.

Schulkin, J. 1991. Sodium hunger: The searchfor a salty taste. Cambridge University Press,Cambridge, U.K. pp. 458–499.

Scott, W.S. 1980. Road salt movement into twoToronto streams. J. Environ. Eng. Div. EE3: 547–560.


Scott, W.S. 1981. An analysis of factorsinfluencing de-icing salt levels in streams.J. Environ. Manage. 13: 269–287.

Setälä, H. and V. Huhta. 1991. Soil faunaincrease Betula pendula growth — Laboratoryexperiments with coniferous forest floor.Ecology 72(2): 665–671.

Setälä, H., J. Laakso, J. Mikola and V. Huhta.1998. Functional diversity of decomposerorganisms in relation to primary production.Appl. Soil Ecol. 9: 25–31.

Shifrin, N.S., B.D. Beck, T.D. Gauthier,S.D. Chapnick and G. Goodman. 1996.Chemistry, toxicology, and human healthrisk of cyanide compounds in soil at formermanufactured gas plant sites. Regul. Toxicol.Pharmacol. 23: 106–116.

Short, T.M., J.A. Black and W.J. Birge. 1991.Ecology of a saline stream: communityresponses to spatial gradients ofenvironmental conditions. Hydrobiologia 226: 167–178.

Simini, M. and I.A. Leone. 1986. The responseof dormant Norway and sugar maples tosimulated de-icing salt spray. J. Arboric.12(1): 1–5.

Singer, S.N., C.K. Cheng and M.G. Scafe. 1997.The hydrogeology of southern Ontario. Vol. 1.Ontario Ministry of Environment and Energy,Toronto, Ontario. 80 pp.

Singleton, H.J. 1986. Water quality criteriafor cyanide. Report and technical appendix.Water Management Division, BritishColumbia Ministry of Environment, Victoria,British Columbia. 84 pp.

Smith, L.L., Jr., S.J. Broderius, D.M. Oseid,G.L. Kimball, W.M. Koenst and D.T. Lind. 1979. Acute and chronic toxicityof HCN to fish and invertebrates. U.S.Environmental Protection Agency. 129 pp. (EPA-600/3-79-009).

Smith, M.E. and J.L. Kaster. 1983. Effect ofrural highway runoff on stream benthicmacroinvertebrates. Environ. Pollut. Ser. A32: 157–170.

Smith, W.G. 1981. Observations on a largehighway kill of Evening Grosbeaks in BritishColumbia. Syesis 14: 163.

Smol, J. P., S.R. Brown and R.N. McNeely. 1983.Cultural disturbances and trophic history of asmall meromictic lake from central Canada.Hydrobiologia 103: 125–130.

Smolders, E. and M.J. McLaughlin. 1996.Chloride increases cadmium uptake in Swisschard in a resin-buffered nutrient solution.Soil Sci. Soc. Am. J. 60: 1443–1447.

Snodgrass, W.J. and D. Morin. 2000. Patrol(maintenance/works) yards. Supportingdocument for the road salts PSL assessment.Report submitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts, July 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec.

Soilcon Laboratories Ltd. 1995. Detailed study ofde-icing salt in the Loon Lake environment.Draft prepared for Highway EnvironmentalBranch, Ministry of Transportation andHighways, Victoria, British Columbia. 97 pp.

Sonzogni, W.C., W. Richardson, P. Rodgers andT.J. Monteith. 1983. Chloride pollution of theGreat Lakes. J. Water Pollut. Control Fed. 55: 513–521.

Spehar, R.L. 1987. Memorandum to C. Stephan,U.S. Environmental Protection Agency,Duluth, Minnesota, June 24 [cited in U.S.EPA, 1988].

Sridhar, K.R. and F. Barlocher. 1997. Waterchemistry and sporulation by aquatichyphomycetes. Mycol. Res. 101(5): 591–596.


Stantec Consulting Ltd. 2000. Calculation ofhistorical road densities within well fieldcapture zones. Kitchener, Ontario.

Steppuhn, H. and D. Curtin (eds.). 1992.Salinity and sustainable agriculture.Swift Current Research Station, ResearchBranch, Agriculture Canada, Swift Current,Saskatchewan. 192 pp. (Prairie SalinityPublication No. 1).

Stoermer, E. F., J. A. Wolin, C. L. Schelske,and D. J. Conley. 1985. Postsettlementdiatom succession in the Bay of Quinte,Lake Ontario. Can. J. Fish. Aquat. Sci. 42: 754-767.

Strobel, G.A. 1967. Cyanide utilization in soil.Soil Sci. 103(4): 299–302 [cited in Chatwin,1990].

Struger, J. 2000. Data on Cl concentrations in theoutfall from water detention ponds. Data sentby e-mail on December 15, 2000.

Stumm, W. and J.J. Morgan. 1981. Aquaticchemistry. John Wiley & Sons, New York,N.Y. 780 pp.

Sucoff, E. 1975. Effect of deicing salts on woodyvegetation along Minnesota Roads. Minn.Agric. Exp. Stn. Tech. Bull. 303, For. Ser. 20. 49 pp.

Sucoff, E. and S.G. Hong. 1976. Effect of NaClon cold hardiness of Malus spp. and Syringavulgaris. Can. J. Bot. 54: 2816–2819.

Sutcliffe, D.W. 1961. Studies on salt and waterbalance in caddis larvae (Trichoptera): II.Osmotic and ionic regulation of body fluidsin Limnephilus stigma Curtis and Anabolianervosa Leach. J. Exp. Biol. 38: 521–530.

Swayne, D.E., A. Shlosberg and R.B. Davis.1986. Salt poisoning in turkey poults. AvianDis. 39(4): 847–852.

TAC (Transportation Association of Canada).1999. Salt management guide. Ottawa,Ontario.

Theis, T.L. and M.J. West. 1986. Effects ofcyanide complexation on adsorption of tracemetals at the surface of goethite. Environ.Technol. Lett. 7: 309–318 [cited in Shifrinet al., 1996].

Thiel, R.P. 1980. Unusual vehicle causedmortality in red crossbills. Passenger Pigeon 42(1): 23–24.

Thirmurthi, D. and K.-H. Tan. 1978. Establishingbaseline data for management of Halifax,N.S., lake watersheds. J. Am. Water WorksAssoc. 70(8): 436–445.

Thompson, P. 2000. Salt damage to trees alongroads in the Regional Municipality of Peel.Report for Regional Municipality of Peel,July 10, 2000.

Thorton, K.W. and J.R. Sauer. 1972. Physiologicaleffects of NaCl on Chironomus attenuatus(Diptera: Chironomidae). Ann. Entomol.Soc. Am. 65: 872–875.

Thuet, J.H. 1977. Environment. In: Economicimpact of highway snow and ice control: stateof the art. Federal Highway Administration,Washington, D.C. pp. 146–168, 231–247(Report No. FHWA-RD-77-20).

Topping, M.S. and G.G.E. Scudder. 1977.Some physical and chemical features ofsaline lakes in central British Columbia.Syesis 10: 145–166.

Toronto and Region Conservation Authority.1998. 1990 to 1996 Water Quality Datafrom the Toronto RAP Watershed; basedon data from the Ontario Ministry of theEnvironment's Provincial Water QualityMonitoring Network. Toronto, Ontario.


Towill, L.E., J.S. Drury, B.L. Whitfield,E.B. Lewis, E.L. Galyan and A.S. Hammons.1978. Reviews of the environmental effects of pollutants: V. Cyanide. 190 pp.(ORNL/EIS-81; EPA-600/1-78-027).

Townsend, A.M. and W.F. Kwolek, 1987. Relativesusceptibility of thirteen pine species tosodium chloride spray. J. Arboric. 13(9): 225–228.

Trainer, D.O. and L. Karstad. 1960. Saltpoisoning in Wisconsin wildlife. J. Am. Vet.Med. Assoc. 13: 614–617.

Trama, F.B. 1954. The acute toxicity of somecommon salts of sodium, potassium, andcalcium to the common bluegill (Lepomismacrochirus). Proc. Acad. Nat. Sci.Philadelphia 196: 1885 [cited in McKee andWolf, 1963].

Transport Canada. 2001. Length of public roads.Table presented on website www.tc.gc.ca/pol/en/ExcelSpreadsheets2/main.asp on May 28, 2001.

Trueman K.F. and D.C. Clague. 1978. Sodiumchloride poisoning in cattle. Aust. Vet. J. 54: 89–91.

Tuchman, M.L., E.F. Stoermer and H.J. Carney.1984. Effects of increased salinity on thediatom assemblage in Fonda Lake, Michigan.Hydrobiologia 109: 179–188.

Underwood, J.K., J.G. Ogden and D.L. Smith.1986. Contemporary chemistry of NovaScotian lakes. Water Air Soil Pollut. 30: 991–999.

Upchurch, R.G. and G.H. Elkan. 1977.Comparison of colony morphology, salttolerance, and effectiveness in Rhizobiumjaponicum. Can. J. Microbiol. 23: 1118–1122.

U.S. EPA (United States EnvironmentalProtection Agency). 1971. Water qualityresearch. Environmental impact of highwaydeicing. Edison Water Quality Laboratory,Edison, New Jersey (NJ 11040 GKK).

U.S. EPA (United States EnvironmentalProtection Agency). 1973. Water qualitycriteria. 1972. National Academy of Scienceand National Academy of Engineering,Washington, D.C. (EPA-R3-73-03) [cited inSingleton, 1986].

U.S. EPA (United States EnvironmentalProtection Agency). 1985. Ambient waterquality criteria for cyanide, 1984. Criteriaand Standards Division, U.S. EnvironmentalProtection Agency, Washington, D.C.(EPA/440/5-84-028).

U.S. EPA (United States EnvironmentalProtection Agency). 1988. Ambientwater quality criteria for chloride.U.S. Environmental Protection Agency,Washington, D.C.

Van Barneveld, J.W. and R.H. Louie. 1990.Boitanio Park, causes of tree decline, soil andvegetation analysis. Working report. BritishColumbia Ministry of Environment, Victoria,British Columbia. 34 pp. plus appendices.

Vickers, S. 2000. Metal accumulation andsediment–water exchange in an urban wetdetention pond. M.Sc. thesis, GeographyDepartment, University of Toronto, Toronto,Ontario.

Wages, D.P., M.D. Ficken, M.E. Cook andJ. Mitchell. 1995. Salt toxicosis incommercial turkeys. Avian Dis. 39: 158–161.

Wallen, I.E., W.C. Greer and R. Lasater. 1957.Toxicity to Gambusia affinis of certain purechemicals in turbid waters. Environ. Future29: 695 [cited in McKee and Wolf, 1963].


Waller, D.L., S.W. Fisher and H. Dabrowska.1996. Prevention of zebra mussel infestationand dispersal during aquaculture operations.Prog. Fish Cult. 58(2): 77–84.

Wang, J.S., P.M. Huang, W.K. Liaw andU.T. Hammer. 1991. Kinetics of thedesorption of mercury from selectedfreshwater sediments as influenced bychloride. Water Air Soil Pollut. 56: 533–542.

Warren, L.A. and A.P. Zimmerman. 1994.The influence of temperature and NaCl oncadmium, copper and zinc partitioning amongsuspended particulate and dissolved phasesin an urban river. Water Res. 28(9): 1921.

Warrington, P. and C. Phelan. 1998. Road saltsand winter maintenance — a bestmanagement practices guide to protectingwater quality. Water Management Section,British Columbia Ministry of Environment,Lands and Parks, Victoria, British Columbia.

Watson T. 2000. [Data on chloride concentrationsin ponds adjacent to roadways in SouthernOntario]. Personal communication fromT. Watson (University of Toronto, Dept.of Biology) to D. Morin (ChemicalsEvaluation Division, Environment Canada).Communication between October, 2000,and January, 2001.

Way, J.L., P. Leung, E. Cannon, R. Morgan,C. Tamulinas, J. Leong-Way, L. Baxter,A. Nagi and C. Chui. 1988. The mechanismof cyanide intoxication and its antagonism. In:Proceedings of the Symposium on CyanideCompounds in Biology, March 15–17. CibaFoundation, London, U.K. pp. 233–243.

Way, J.M. 1977. Roadside verges andconservation in Britain: a review. Biol.Conserv. 12: 65–74.

Weast, R.C., D.R. Lide, M.J. Astle and W. Beyer(eds.). 1989. CRC handbook of chemistry andphysics, 1989–1990. 70th ed. CRC Press,Boca Raton, Florida.

Weiler, R.R. and V.K. Chawla. 1969. Dissolvedmineral quality of Great Lakes waters.In: Proceedings of the 12th Conference onGreat Lakes Research. Vol. 1. InternationalAssociation for Great Lakes Research, AnnArbor, Michigan. Braun-Brumfield, Inc., AnnArbor, Michigan. pp. 801–818.

Werkhoven, C.H.E., P.J. Salisbury, andW.H. Cram. 1966. Germination and survivalof Colorado spruce, Scots pine, caragana, andSiberian elm at four salinity and two moisturelevels. Can. J. Plant Sci. 46(1): 1–7.

Wetzel, R.G. 1975. Limnology. W.B. SaundersCo., Philadelphia, Pennsylvania. 743 pp.

Wetzel, R.G. 1983. Limnology. 2nd edition.Saunders College Publishing, Toronto,Ontario.

Whitfield, P.H. and N.L. Wade. 1992. Monitoringtransient water quality events electronically.Water Resour. Bull. 28(4): 703–711.

Wickstrom, M.L., P. Mineau, L.J. Brownlee andT.K. Bollinger. 2001. Toxicity of road salt tohouse sparrows. Draft report. University ofSaskatchewan, Saskatoon, Saskatchewan.

Wilcox, B.A. and D.D. Murphy. 1985Conservation strategy: the effects offragmentation on extinction. Am. Nat. 125: 879–887.

Wilcox, D.A. 1982. The effects of deicing salts onwater chemistry and vegetation in PinhookBog, Indiana Dunes National Lakeshore.Ph.D. dissertation, Purdue University, WestLafayette, Indiana. 134 pp.

Wilcox, D.A. 1984. The effects of NaCl deicingsalt on Sphagnum recurvum P. Beauv.Environ. Exp. Bot. 24(4): 295–304.

Wilcox, D.A. 1986a. The effects of deicing saltson vegetation in Pinhook Bog, Indiana. Can.J. Bot. 64(4): 865–874.


Wilcox, D.A. 1986b. The effects of deicing salton water chemistry in Pinhook Bog, Indiana.Water Resour. Bull. 22(1): 57–64.

Wilcox, D.A. and R.E. Andrus. 1987. The roleof Sphagnum fimbriatum in secondarysuccession in a road salt impacted bog. Can.J. Bot. 65: 2270–2275.

Williams, D.D., N.E. Williams and Y. Cao. 1997.Spatial differences in macroinvertebratecommunity structure in springs insoutheastern Ontario in relation to theirchemical and physical environment. Can. J.Zool. 75: 1404–1414.

Williams, D.D., N.E. Williams and Y. Cao. 1999.Road salt contamination of groundwater in amajor metropolitan area and development of abiological index to monitor its impact. WaterRes. 34(1): 127–138.

Williams, D.J., K.W. Kuntz, S. L’Italien andV. Richardson. 1998. Lake OntarioSurveillance Program. Spatial and temporaltrends of selected parameters, with emphasison 1992–93 results. EnvironmentalConservation Branch/Ontario Region,Ecosystem Health Division, EnvironmentCanada. 24 pp. (Report No. EHD/ECB-OR/98-04/l).

Williams, W.D. 1987. Salinity as a determinantof salt lake fauna: a question of scale.Hydrobiologia 210: 151–160.

Windingstad, R.M., F.X. Kartch, R.K. Stroudand M.R. Smith. 1987. Salt toxicosis inwaterfowl in North Dakota. J. Wildl. Dis. 23(3): 443–446.

WI SLOH (Wisconsin State Laboratory ofHygiene). 1995. Unpublished data on chloridetoxicity to aquatic species. From A. Letts(Technical Manager, Morton International Inc.Chicago, Illinois) to M.S. Evans (NationalHydrology Research Institue, EnvironmentCanada), August 11, 1998.

Wolfe, D.F.G. 1996. Opportunistic winter wateracquisition by pine grosbeaks. Wilson Bull.108(1): 186–187.

Wood, J.A. and C.D.A. Rubec. 1989. Chemicalcharacterization of several wetlands inKejimkujik National Park, Nova Scotia.Water Air Soil Pollut. 46: 177–186.

Woodliffe, P.A. 2001. Personal communication.Ontario Ministry of Natural Resources–Aylmer District, Chatham, Ontario.

Woods, J. 1998. [The recurrent kills of pinesiskins witnessed by local inhabitants inRevelstoke National Park]. Personalcommunication from John Woods (ParksCanada, Mount Revelstoke & GlacierNational Parks, Revelstoke, British Columbia)with P. Mineau and G. Fox (CanadianWildlife Service, National Wildlife ResearchCentre, Ottawa). Meeting, December 4, 1998.

Wulkowicz, G.M. and Z.A. Saleem. 1974.Chloride balance of an urban basin inthe Chicago area. Water Resour. Res. 10(5): 974–982.

Zeeb, B.A. and J.P. Smol. 1991.Paleolimnological investigation of the effectsof road salt seepage on scaled chrysophytesin Fonda Lake, Michigan. J. Paleolimnol. 5: 263–266.

Zheng, C. 1990. MT3D: A modular three-dimensional transport model for simulation ofadvection, dispersion and chemical reactionsof contaminant in groundwater systems.Report to the U.S. Environmental ProtectionAgency, Ada, Oklahoma. 170 pp.

Zidar, L. 2000. Public comment. Salt damage totrees in North District of the Region of Peel.Letter to J. Buccini, dated October 11, 2000.System Operations, Region of Peel,Brampton, Ontario.


Data relevant to the assessment of whether roadsalts are “toxic” to the environment underCEPA 1999 were identified from existing reviewdocuments, published reference texts and on-linesearches of the following databases, conductedbetween January and May 1996: Aqualine(1990–1996), ASFA (Aquatic Sciences andFisheries Abstracts, Cambridge ScientificAbstracts; 1996), BIOSIS (BiosciencesInformation Services; 1990–1996), CAB(Commonwealth Agriculture Bureaux;1990–1996), CESARS (Chemical EvaluationSearch and Retrieval System, Ontario Ministryof the Environment and Michigan Departmentof Natural Resources; 1996), Chemical Abstracts(Chemical Abstracts Service, Columbus, Ohio;1990–1996), CHRIS (Chemical Hazard ReleaseInformation System; 1964–1985), CurrentContents (Institute for Scientific Information;1990–1992, 1996), ELIAS (EnvironmentalLibrary Integrated Automated System,Environment Canada library; January 1996),Enviroline (R.R. Bowker Publishing Co.;November 1995 – June 1996), EnvironmentalAbstracts (1975 – February 1996), EnvironmentalBibliography (Environmental Studies Institute,International Academy at Santa Barbara;1990–1996), GEOREF (Geo ReferenceInformation System, American GeologicalInstitute; 1990–1996), HSDB (HazardousSubstances Data Bank, U.S. National Library ofMedicine; 1990–1996), Life Sciences (CambridgeScientific Abstracts; 1990–1996), NTIS (NationalTechnical Information Service, U.S. Departmentof Commerce; 1990–1996), Pollution Abstracts(Cambridge Scientific Abstracts, U.S. NationalLibrary of Medicine; 1990–1996), POLTOX

(Cambridge Scientific Abstracts, U.S. NationalLibrary of Medicine; 1990–1995), RTECS(Registry of Toxic Effects of ChemicalSubstances, U.S. National Institute forOccupational Safety and Health; 1996), Toxline(U.S. National Library of Medicine; 1990–1996),TRI93 (Toxic Chemical Release Inventory, U.S.Environmental Protection Agency, Office of ToxicSubstances; 1993), U.S. EPA-ASTER(Assessment Tools for the Evaluation of Risk,U.S. Environmental Protection Agency; up toDecember 21, 1994), WASTEINFO (WasteManagement Information Bureau of the AmericanEnergy Agency; 1973 – September 1995) andWater Resources Abstracts (Geological Survey,U.S. Department of the Interior; 1990–1996).Reveal Alert was used to maintain an ongoingrecord of the current scientific literaturepertaining to the potential environmental effectsof road salts.

In addition, the TransportationAssociation of Canada has reviewed informationon sodium/calcium road salt as part of its RoadSalt Management Initiative (TAC, 1999).

Data searches were done by individualworking groups preparing separate components ofthe supporting documentation. Search strategiesor data sources are provided in individual reports.

Data obtained after May 2001 were notconsidered in this assessment.


APPENDIX A SEARCH STRATEGIES EMPLOYED FOR

IDENTIFICATION OF RELEVANT DATA

Brownlee, L., P. Mineau and A. Baril. 2000. Riskcharacterization for wildlife. Report submittedto the Environment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts, May 2000. CommercialChemicals Evaluation Branch, EnvironmentCanada, Hull, Quebec.

Butler, B.J. and J. Addison. 2000. Biologicaleffects of road salts on soils. Report submittedto the Environment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts, March 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec.

Cain, N.P., B. Hale, E. Berkelaar and D. Morin.2001. Critical review of effects of NaCl andother road salts on terrestrial vegetation inCanada. Report submitted to the EnvironmentCanada CEPA Priority Substances ListEnvironmental Resource Group on RoadSalts. July 2001. Existing Substances Branch,Environment Canada, Hull, Quebec.

Delisle, C.E. and L. Dériger. 2000.Caracterisation et élimination des neigesusées: impacts sur l’environnement. Reportsubmitted to the Environment Canada CEPAPriority Substances List EnvironmentalResource Group on Road Salts. CommercialChemicals Evaluation Branch, EnvironmentCanada, Hull, Quebec.Environment Canada.1997. Canadian Environmental Protection Act— Problem formulation for the environmentalassessment of the Priority Substance roadsalts. Government of Canada.

Evans, M. and C. Frick. 2001. The effects of roadsalts aquatic ecosystems. National WaterResearch Institute, Saskatoon,Saskatchewan.Report submitted to the

Environment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts. August 2001. ExistingSubstances Branch, Environment Canada,Hull, Quebec.

Johnston, C.T., R.C. Sydor and C.L.S. Bourne.2000. Impact of winter road salting on thehydrogeologic environment — an overview.Stantec Consulting Ltd., Kitchener, Ontario.Report submitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts, May 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec.

Letts, A.D. 2000. Effects of sodium ferrocyanidederived from road salting on the ecosystem.Morton Salt, Morton International Inc.,Chicago, Illinois. Report submitted to theEnvironment Canada CEPA PrioritySubstances List Environmental ResourceGroup on Road Salts, May 2000. CommercialChemicals Evaluation Branch, EnvironmentCanada, Hull, Quebec.

Mayer, T., W.J. Snodgrass and D. Morin. 1999b.Spatial characterization of the occurrenceof road salts and their environmentalconcentrations as chlorides in Canadiansurface waters and benthic sediments.Environment Canada. NWRI contributionNo. 99-097 (revised 2001). 44 pp.

Morin, D. 2000. Environmental impacts fromthe spreading and storage of road salts: Casestudies. Supporting document for the roadsalts PSL assessment. Report submitted tothe Environmental Resource Group for RoadSalts, Commercial Chemicals EvaluationBranch, Environment Canada. May 2000. 84 pp.


APPENDIX B SUPPORTING DOCUMENTS PREPARED FOR THE

PSL ASSESSMENT OF ROAD SALTS

Morin, D. and M. Perchanok. 2000. Road saltloadings in Canada. Supporting documentfor the road salts PSL assessment. Reportsubmitted to the Environment Canada CEPAPriority Substances List EnvironmentalResource Group on Road Salts, May 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec. 85 pp.

Morin, D., W. Snodgrass, J. Brown and P.A. Arp.2000. Impacts evaluation of road salt loadingson soils and surface waters. Supportingdocument for the road salts PSL assessment.Report submitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts, June 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec. 76 pp.

Snodgrass, W.J. and D. Morin. 2000. Patrol(maintenance/works) yards. Supportingdocument for the road salts PSL assessment.Report submitted to the Environment CanadaCEPA Priority Substances List EnvironmentalResource Group on Road Salts, July 2000.Commercial Chemicals Evaluation Branch,Environment Canada, Hull, Quebec.


© Minister of Public Works and Government Services 2001

Canadian Cataloguing in Publication Data

Additional information can be obtained at Environment Canada’s Web site atwww.ec.gc.ca or at the Inquiry Centre at 1-800-668-6767.