UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Metal distribution, speciation, and bioavailability in stormwater management oil/grit chamber systems… Cohen, Tamira 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-104783.pdf [ 26.23MB ]
Metadata
JSON: 831-1.0092327.json
JSON-LD: 831-1.0092327-ld.json
RDF/XML (Pretty): 831-1.0092327-rdf.xml
RDF/JSON: 831-1.0092327-rdf.json
Turtle: 831-1.0092327-turtle.txt
N-Triples: 831-1.0092327-rdf-ntriples.txt
Original Record: 831-1.0092327-source.json
Full Text
831-1.0092327-fulltext.txt
Citation
831-1.0092327.ris

Full Text

METAL DISTRIBUTION, SPECIATION, AND BIOAVAILABILITY IN STORMWATER MANAGEMENT OIL/GRIT CHAMBER SYSTEMS AND MARINE RECEIVING WATERS by TAMIRA COHEN B.Sc, University of Toronto, 1990 B.Sc, Hon. University of Toronto, 1993 M.Sc, University of California, Los Angeles, 1995 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Zoology) THE UNIVERSITY OF BRITISH COLUMBIA June 2005 © Tamira Cohen, 2005 ABSTRACT The stormwater management practice of oil/grit chamber systems is utilized to reduce pollutants from urban/industrial runoff to natural waters. Metal determinations (Cu, Zn, Cd, N i , Pb) were used to test chamber effectiveness, elucidate distribution patterns of total and labile (Cu, Cd, N i only) metals in water and sediments, and examine influences on metal bioavailability. Water, sediment, and biota analyses revealed chambers do not function adequately in reducing dissolved, fine particulate-bound, colloidal, and DGT-labile metals. Runoff from industrial/commercial areas resulted in elevated bioavailabilities of Cu, Zn, Cd, and Pb. Cu and Zn were of significant concern, especially during the onset of the rainy season. Chamber systems appear to provide some reduction of metal-bound coarse particles; however, significant portions of total metals easily flow into receiving water bodies. Furthermore, re-suspension of previously deposited particulate metals within chambers may provide additional opportunities for increased metal dissolution and lability. The diffusive gradient in thin film technique (DGT) used in water provided an indication of high metal discharge and proved to be a good measure of time-integrated average of elevated levels. Consequently, this technique can reflect effective metal means for whole storm event periods. DGT in water also effectively reflected bioavailabilities for the three elements assessed in the order (Cu>Cd»>Ni) . For N i , DGT may be more sensitive in discriminating between sites than bioaccumulations. Comparatively, the sediment DGT technique was not very effective due to large spatial and temporal variabilities; it is not recommended for environmental monitoring work, where these conditions may apply. The barnacle, Balanus glandula, was useful as a biomonitoring species and shows promise for future metal bioavailability work within its geographical range. Variability in metal bioaccumulations was noted as a result of substrata and vertical positioning and should be considered in any future studies using B. glandula as a metal biomonitor. Ideally, contaminant free substrata should be utilized for the prevention of potential influences from naturally occurring surfaces. Altogether, these studies indicate that oil/grit chambers function as poor primary treatments. Better stormwater detention capacity and metal removal processes are necessary to facilitate the goal of metal reduction in stormwater runoff to waterways. 11 TABLE OF CONTENTS A B S T R A C T ii T A B L E OF CONTENTS i i i LIST OF TABLES vii LIST OF FIGURES x LIST OF ABBREVIATIONS xv A C K N O W L E D G E M E N T S xvi CHAPTER 1 INTRODUCTION A N D B A C K G R O U N D INFORMATION 1 1.1. Introduction 2 1.1.1. Motivation 2 1.1.2. Objectives 3 1.1.3. Approach 6 1.1.4. Thesis Organization 6 1.2. Background Information 6 1.2.1. Study Area 6 1.2.2. Urban Stormwater Runoff and Management Practices 7 1.2.3. Trace Elements Investigated (Cu, Zn, Cd, Pb, Ni) 11 1.2.3.1. Natural and Anthropogenic Sources 11 1.2.3.2. Physical, Chemical, and Biological Properties 11 1.2.4. Metal Bioavailability 14 1.3. References 21 CHAPTER 2 DISTRIBUTION OF M E T A L S IN WATERS OF STORMWATER M A N A G E M E N T OIL/GRIT C H A M B E R S A N D A D J A C E N T M A R I N E RECEIVING A R E A S 25 2.1. Introduction 26 2.2. Materials and Methods 28 2.2.1. Field Program 28 2.2.2. Analytical Program 30 2.2.2.1. Reagents 30 2.2.2.2. Instrumental Metal Analysis 30 2.2.3. Diffusion Gradient in Thin Film (DGT) 31 2.2.3.1. Background 31 2.2.3.2. Preparation and Performance Testing 33 in 2.2.3.2.1. Gels 33 2.2.3.2.2. Resin in Gels 33 2.2.3.2.3. DGT Performance 34 2.2.3.2.4. Time Dependence 35 2.2.3.2.5. Ionic Strength and Temperature 35 2.2.4. Determination of Dissolved Metals 37 2.2.5. Determination of Particulate Metals 40 2.3. Results 40 2.3.1. Temperature and Salinity. 41 2.3.2. pH 42 2.3.3. Dissolved and DGT-Labile Metals 43 2.3.3.1. Copper 43 2.3.3.2. Nickel 45 2.3.3.3. Cadmium 47 2.3.3.4. Zinc and Lead 48 2.3.3.5. Suspended Particulate Matter 49 2.3.4. Chamber Impact on Stormwater Metals 52 2.3.5. DGT-Lability 54 2.4. Discussion 54 2.4.1. Temperature and Salinity 54 2.4.2. Stormwater: Dissolved metals and Regional Context 55 2.4.3. Chamber Effectiveness 56 2.4.4. DGT-Lability 61 2.4.5. Salinity and Dispersion 64 2.5. Conclusions 68 Table 70 2.6. References 120 CHAPTER 3 ASSESSMENT OF M E T A L DISTRIBUTION IN SEDIMENTS OF STORMWATER RUNOFF RECEIVING A R E A S 125 3.1. Introduction 126 3.2. Materials and Methods 130 3.2.1. Field Program 130 3.2.2. Sediments 132 3.2.3. Trace metals 132 3.2.3.1. Total metals 132 3.2.3.2. Acid extractable metals 133 3.2.4. Carbon 133 3.2.5. Porewaters 134 3.2.6. DGT 135 3.2.6.1. Theory 135 3.2.6.2. Field Application 136 iv 3.2.7. Grain size 137 3.2.8. Analytical Program 138 3.2.8.1. Reagents 138 3.2.8.2. Instrumental Metal Analysis 138 3.3. Results 139 3.3.1. Grain Size 139 3.3.2. Organic Carbon 139 3.3.3. Total and Acid-Extractable Metals 140 3.3.4. Porewater Metals 143 3.3.5. DGT Metals 145 3.4. Discussion 146 3.5. Conclusions 151 3.6. References 204 CHAPTER 4 BIOMONITOR ASSESSMENT OF M E T A L DISTRIBUTION IN M A R I N E A R E A S RECEIVING STORMWATER RUNOFF 209 4.1. Introduction 210 4.2. Materials and Methods 211 4.2.1. Field Program 211 4.2.2. Analytical Program 212 4.2.2.1. Reagents 212 4.2.2.2. Sample Preparation 213 4.2.2.3. Instrumental Metal Analysis 214 4.2.2.4. Analysis of Reference Materials 214 4.2.3. Statistical Analyses 214 4.3. Results 215 4.3.1. Macroalgae 215 4.3.2. Mytilus trossulus 216 4.3.3. Balanus glandula 216 4.4. Discussion 217 4.4.1. Comparative Bioaccumulation 220 4.4.2. Environmental Metals and Bioaccumulation 222 4.5. Conclusions 227 4.6. References 262 CHAPTER 5 THE B A R N A C L E , BALANUS GLANDULA, AS A BIOMONITOR OF T R A C E M E T A L S 268 5.1. Introduction 269 5.2. Materials and Methods 270 5.2.1. Species Identification 270 5.2.2. Field Program 270 5.2.3. Analytical Program 272 5.2.4. Statistical Analyses 272 5.3. Results 273 5.4. Discussion 274 5.5. Conclusions 277 Station 279 5.6. References 286 CHAPTER 6 6.1. Conclusions and Recommendations 290 6.2. References 294 APPENDIXA D A T A S U M M A R Y 295 Table A l . Water Sampling Data 296 Table A2. Sediment Data 304 Table A3. Water Supplementary Data 349 Table A4. Diffusion Coefficients of Metals in Distilled Water at 25°C 356 Supplementary Precipitation Figures Figure A l . Total precipitation (mm) during field biota metal exposures,2001.... 357 Figure A.2. Seasonal precipitation patterns during 1998-2001 sampling periods.358 APPENDIX B INSTRUMENTAL OPERATING CONDITIONS 359 Table B l . PlasmaQuad™ ICP-MS Operating Conditions 360 Table B2. Element2™ HP-ICP-MS Operating Conditions 360 LIST OF TABLES Table 1.1. Typical Pollutant Concentrations Found in Urban Stormwater Runoff. 8 Table 1.2. Median Pollutant Removal of Stormwater Treatment Practices 10 Table 1.3. Major Inorganic Species in Natural Waters 12 Table 2.1. Monthly climate and precipitation summaries 70 Table 2.2. Frequency of sampling during different weather conditions and corresponding climate data 71 Table 2.3. Analyses of control DGT field units for blanks and method detection limits. 72 Table 2.4. Concentrations of dissolved metals used in deionized water DGT calibration experiment 72 Table 2.5. Target Spike Ratios 72 Table 2.6. Concentrations used in Isotope Dilution Mass Spectrometry Trials 73 Table 2.7. Analysis efficiency of original and spiked standard estuarine reference material SLEW-2 during laboratory trials 74 Table 2.8. Analyses of original and spiked standard estuarine reference material SLEW-2 and blanks using isotope dilution mass spectrometry method 75 Table 2.9. Analyses of marine sediment reference material PACS-2 (Esquimalt Harbour, BC) and filter blanks using microwave digestion method 76 Table 2.10. Metal concentrations in water from study sites 77 Table 2.11. Suspended particulate matter and particulate metal concentrations from study sites 78 Table 2.12. Ancillary data from study sites 79 Table 2.13. Fall 2000 sampling dates, times and water velocity at Philip site 80 Table 2.14. Stormwater concentrations from side drainage location at Philip Site 80 Table 2.15. Stormwater metal concentrations at Philip Site 81 Table 2.16. Stormwater metal concentrations of particulate matter on September 29, 2000 at Philip site 82 Table 2.17. Comparision of stormwater metal concentrations total and (dissolved) within the Greater Vancouver Regional District (GVRD) 83 Table 2.18. Characteristics of Catchment Areas 84 Table 2.19. Settling properties of particles within the Philip chamber 84 Table 2.20. Stormwater residence time 85 Table 2.21. Particulate-bound metal concentrations within the Greater Vancouver Regional District (GVRD) 85 Table 2.22. Percentage reductions of stormwater metals through chamber at the Philip site 86 Table 2.23. Stormwater flow at the Philip site 86 Table 3.1. Metal analyses of marine sediment reference material PACS-2 (Esquimals Harbour, B.C.) and blanks using microwave digestion method 153 Table 3.2. Total carbon analyses of marine sediment reference material 153 Table 3.3. Sediment grain size of surface and core samples from Philip and Blueridge sites 154 Table 3.4. Sediment characteristics and total mean metal concentrations from selected areas of Burrard Inlet 156 vii Table 4.1. Analyses of marine lobster hepatopancreas reference material TORT-2 (Prince Edward Island) and blanks using microwave digestion method 229 Table 4.2. Metal concentrations of macroalgal species Fucus gardneri from Philip, Blueridge, and Lighthouse Park sites 229 Table 4.3. Metal concentrations of macroalgal species Enteromorpha sp. from Philip, Blueridge, and Lighthouse Park sampling sites 230 Table 4.4. Weight-adjusted metal concentrations in soft tissues of Mytilus trossulus from Philip, Blueridge, and Lighthouse Park sampling sites 231 Table 4.5. Trace metal concentrations in soft tissues of Balanus glandula found on rocks from Philip, Blueridge, and Lighthouse sites 232 Table 4.6. Ranges of metal bioaccumulation for Philip, Blueridge and Lighthouse sites. 233 Table 4.7. Metal concentrations in bodies of selected Mytilus species from various locations and metal-contamination levels 234 Table 4.8. Trace metal Median International Standards (MIS) for shellfish and indicators of high concentrations (85th percentile concentrations) in the United States (NS&T) and California, U.S.A. (SWRCB) Mussel Watch Programs 236 Table 4.9. Metal concentrations in bodies of selected Fucus species from various locations and metal-contamination levels 237 Table 4.10. Metal concentrations in bodies of selected Enteromorpha species from various locations and metal-contamination levels 238 Table 4.11. Metal concentrations in bodies of selected barnacle species from various locations and metal-contamination levels 239 Table 4.12. Water salinities, temperatures, and pH from the Philip, Blueridge, and Lighthouse (reference) sites during spring sampling periods of 1999 and biota collection period, May 21-22, 2001 241 Table 4.13. DGT-labile Cu, Cd, and N i concentrations in water from all sampling sites during the 2001 winter/spring deployment periods from Feb. 6 to May 20 using seven 10-19 day deployment periods 242 Table 4.14. Dissolved and (DGT-labile) N i concentrations in water during spring sampling periods of 1999 and biota collection period, May 21-22, 2001 243 Table 4.15. Dissolved and (DGT-labile) Cu concentrations in water during spring sampling periods of 1999 and biota collection period, May 21-22, 2001 244 Table 4.16. Dissolved Zn concentrations in water during fall and summer sampling periods of 1999-2000 and biota collection period, May 21-22, 2001 245 Table 4.17. Dissolved and (DGT-labile) Cd concentrations in water during spring sampling periods of 1999 and biota collection period, May 21-22, 2001 246 Table 4.18. Dissolved Pb concentrations in water during spring sampling periods of 1999 and biota collection period, May 21-22, 2001 247 Table 4.19. Sediment characteristics: percentage organic carbon content, percentage silt content, extractable Fe, and extractable Mn from the Philip, Blueridge, and Lighthouse (reference) sites during biota collection period, May 21-22, 2001 248 Table 4.20. N i concentrations in suspended particulate matter, total sediments, extractable from sediments, and fluxes from sediments as measured by DGT from the Philip, Blueridge and Lighthouse (reference) sites during biota collection period, May 21-22,2001 249 v i n Table 4.21. Cu concentrations in suspended particulate matter, total sediments, extractable from sediments, and fluxes from sediments as measured by DGT from the Philip, Blueridge and Lighthouse (reference) sites during biota collection period, May 21-22, 2001 250 Table 4.22. Zn concentrations in suspended particulate matter, total sediments, and extractable from sediments from the Philip, Blueridge and Lighthouse (reference) sites during biota collection period, May 21-22, 2001 251 Table 4.23. Cd concentrations in suspended particulate matter, total sediments, extractable from sediments, and fluxes from sediments as measured by DGT from the Philip, Blueridge and Lighthouse (reference) sites during biota collection period, May 21-22, 2001 252 Table 4.24. Pb concentrations in suspended particulate matter, total sediments, and extractable from sediments from the Philip, Blueridge and Lighthouse (reference) sites during biota collection period, May 21-22, 2001 253 Table 5.1. Trace metal concentrations in soft tissues of Balanus glandula grown on plexiglass frames from Philip site 279 Table 5.2. Trace metal concentrations in soft tissues of Balanus glandula found on rocks from Philip, Blueridge, and Lighthouse sites 280 Table 5.3. Regression coefficients of metal concentrations (dependent variables) vs. distance (independent variable) (on log-log data transformations) 281 i x LIST OF FIGURES Figure 1.1. Map of North Vancouver and Vancouver Harbour 17 Figure 1.2. Philip site 18 Figure 1.3. Blueridge site 19 Figure 1.4. Schematic of Oil/Grit Chamber 20 Figure 2.1. Schematic representation of the free concentration of ionic species in a gel assembly in contact with natural water where the concentration is Cb (DBL, diffusive boundary layer) 87 Figure 2.2. Measured metal mass in resin layer for gel assemblies immersed in a stirred mixed metal (25 ug/L Cu, 1 ug/L Cd, 10 ug/L Ni) N a N 0 3 (0.01 M) solution at 20°C for different times 88 Figure 2.3. Measured metal mass in resin layer for gel assemblies immersed in stirred solutions of mixed metal standards and 0.01 M NaN03 at temperatures of A) 12°C and B) 20°C 89 Figure 2.4. Effect of total metal concentrations on the effective diffusion coefficient Deff for Cu, N i , and Cd 90 Figure 2.5. Measured metal mass in resin layer for gel assemblies immersed in stirred solutions of mixed metal standards in deionized water 91 Figure 2.6. August 25-26, 1998: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 92 Figure 2.7. September 30-October 4, 1998: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site 93 Figure 2.8. November 14-15, 1998: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip and Blueridge Sites 94 Figure 2.9. March 23-26, 1999: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 95 Figure 2.10. April 17-18, 1999: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 96 Figure 2.11. June 3-4, 1999: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 97 Figure 2.12.November 12-13, 1999: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 98 Figure 2.13. November 21-22, 1999: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site 99 Figure 2.14. September 7-8, 2000: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip Site 100 Figure 2.15. September 18-19, 2000: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, N i , and Pb levels in water at Philip Site 101 Figure 2.16. September 27-30, 2000: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip and Blueridge Sites 102 Figure 2.17. May 21-22, 2001: A) Daily precipitation record; B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip, Blueridge and Lighthouse Reference (LHR)Site 103 Figure 2.18. Total suspended particulate matter (SPM) at Philip, Blueridge and Lighthouse (Reference) Sites 104 x Figure 2.19. A) Particulate Cu in SPM and B) Particulate Cu in waters at Philip, Blueridge and Lighthouse (Reference) Sites 105 Figure 2.20. A) Particulate N i in SPM and B) Particulate N i in waters at Philip, Blueridge and Lighthouse (Reference) Sites 106 Figure 2.21. A) Particulate Cd in SPM and B) Particulate Cd in waters at Philip, Blueridge and Lighthouse (Reference) Sites 107 Figure 2.22. A) Particulate Zn in SPM and B) Particulate Zn in waters at Philip, Blueridge and Lighthouse (Reference) Sites 108 Figure 2.23. A) Particulate Pb in SPM and B) Particulate Pb in waters at Philip, Blueridge and Lighthouse (Reference) Sites 109 Figure 2.24. Concentrations of dissolved and DGT-labile Cu, N i , and Cd produced upon mixing unfiltered stormwater with unfdtered seawater from the Philip site on September 29, 2000 110 Figure 2.25. DGT-labile Cu, N i , and Cd as percentage of total dissolved in mixing experiment of September 29, 2000 waters 111 Figure 2.26. Characteristics of moderate and severe storms: A) Storm properties including number of antecedent dry days, total precipitation, and rainfall intensity; and B) DGT-labile concentrations for Cu, N i , and Cd 112 Figure 2.27. Dissolved and DGT-labile metals measured in stormwaters during a light storm A) November 13, 1999 and moderate storm B) November 22, 1999 within Philip chamber inlet, Philip chamber and Philip chamber outlet waters 113 Figure 2.28. Dissolved, DGT-labile, particulate, and SPM-bound metals measured in A) September 29, 2000 stormwater (severe storm) and B) September 30, 2000 receiving basin water within Philip chamber inlet, Philip chamber and Philip chamber outlet waters 114 Figure 2.29. Speciation of Cu, N i , Cd, Pb, and Zn in mixing experiments calculated with the Visual MINTEQ computer program and using estimated values of dissolved organic matter of A) 2 mg/L and B) 20 mg/L 115 Figure 2.30. Correlations of measured DGT-labile metal versus calculated free metal ion using estimated D O M values of A) 2 mg/L and B) 20 mg/L for Cu, N i , and Cd 116 Figure 2.31. Correlations of measured DGT-labile metal versus calculated total inorganic metals using estimated D O M values of A) 2 mg/L and B) 20 mg/L for Cu, N i , andCd 117 Figure 2.32. Proportion of dissolved metals at Philip stations relative to stormwater chamber levels as a function of A) distance, B) salinity, and C) a salinity based mixing index for three moderate/severe storms (November 22, 1999, September 7, 2000, and September 29, 2000) 118 Figure 2.33. Metal, salinity, pH and SPM summary for sampling period September 28-30, 2000 at Philip site 119 Figure 3.1. DGT Sediment Probe 157 Figure 3.2. Sediment grain size distributions at the Philip site from 1998 surface (0-2 cm) sediment sampling during A . August, and B. March 158 Figure 3.3. Sediment grain size distributions from September 2000 core sampling at Philip and Blueridge sites from A.Sediment core depths and B. Surficial sediments.... 159 Figure 3.4. Organic carbon in sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 1999, C. 2000) and Blueridge (D. 2000) sites 160 x i Figure 3.5. Organic carbon of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 161 Figure 3.6. Moisture content in sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 1999, C. 2000) and Blueridge (D. 2000) sites 162 Figure 3.7. Total and acid extractable Cu in sediment cores sampled during A. 1998 and B. 1990 at the Philip site 163 Figure 3.8. Total and acid extractable Cu in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites 164 Figure 3.9. Total and acid extractable Zn in sediment cores sampled during A. 1998 and B. 1990 at the Philip site 165 Figure 3.10. Total and acid extractable Zn in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites 166 Figure 3.11. Total and acid extractable Cd in sediment cores sampled during A. 1998 and B. 1990 at the Philip site 167 Figure 3.12. Total and acid extractable Cd in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites 168 Figure 3.13. Total and acid extractable Pb in sediment cores sampled during A. 1998 and B. 1990 at the Philip site 169 Figure 3.14. Total and acid extractable Pb in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites 170 Figure 3.15. Total and acid extractable N i in sediment cores sampled during A. 1998 and B. 1990 at the Philip site 171 Figure 3.16. Total and acid extractable N i in sediment cores sampled in 2000 at A . Philip and B. Blueridge sites 172 Figure 3.17. Total Cu of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 173 Figure 3.18. Total Zn of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 174 Figure 3.19. Total Cd of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 175 Figure 3.20. Total Pb of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 176 Figure 3.21. Total N i of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites 177 Figure 3.22. Total metal-metal correlations of all samples obtained 1998-2001 178 Figure 3.23. Acid extractable metal-metal correlations of all samples obtained 1998-2001 179 Figure 3.24. Mean total metal concentrations in Philip site surface sediments obtained during 1998-2001 180 Figure 3.25. Mean total metal concentrations in Blueridge site surface sediments obtained during 1998-2001 181 Figure 3.26. Porewater concentrations of dissolved A . Mn and B. Fe sampled at the Philip site in 1998 182 Figure 3.27. Porewater concentrations of dissolved A . Mn and B. Fe sampled at the Philip site in 1999 183 x i i Figure 3.28. Porewater concentrations of dissolved Cu sampled in A. 1998 and B. 1999 at the Philip site 184 Figure 3.29. Porewater concentrations of dissolved Zn sampled in A . 1998 and B. 1999 at the Philip site 185 Figure 3.30. Porewater concentrations of dissolved Cd sampled in A. 1998 and B. 1999 at the Philip site 186 Figure 3.31. Porewater concentrations of dissolved Pb sampled in A . 1998 and B. 1999 at the Philip site 187 Figure 3.32. Porewater concentrations of dissolved N i sampled in A . 1998 and B. 1999 at the Philip site 188 Figure 3.33. Porewater concentrations and porewater to DGT fluxes of dissolved Cu sampled in 1998 at stations A l , A2, and A3 at the Philip site 189 Figure 3.34. Porewater concentrations and porewater to DGT fluxes of dissolved Cd sampled in 1998 at stations A l , A2, and A3 at the Philip site 190 Figure 3.35. Porewater concentrations and porewater to DGT fluxes of dissolved N i sampled in 1998 at stations A l , A2, and A3 at the Philip site 191 Figure 3.36. Porewater concentrations and porewater to DGT fluxes of dissolved Cu sampled in 1999 at stations A l , A2, and A3 at the Philip site 192 Figure 3.37. Porewater concentrations and porewater to DGT fluxes of dissolved Cd sampled in 1999 at stations A l , A2, and A3 at the Philip site 193 Figure 3.38. Porewater concentrations and porewater to DGT fluxes of dissolved N i sampled in 1999 at stations A l , A2, and A3 at the Philip site 194 Figure 3.39. Mean porewater to DGT fluxes of dissolved N i , Cu, and Cd sampled in September 2000 at the Philip site (stations A l , A2, A3, A4) and at the Blueridge site (stations C4, C5) 195 Figure 3.40. Mean porewater to DGT fluxes of dissolved N i , Cu, and Cd sampled in May 2001 at the Philip site (stations AO to A6) and at the Blueridge site (stations C3 to C6) 196 Figure 3.41. Acid extractable Cu associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A. Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001 197 Figure 3.42. Acid extractable Zn associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A. Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001 198 Figure 3.43. Acid extractable Cd associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A. Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001 199 Figure 3.44. Acid extractable Pb associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A. Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001 200 Figure 3.45. Acid extractable N i associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001 201 Figure 3.46. Acid extractable metal associations with percentage silt in surface sediments (0-2 cm) at the Philip and Blueridge sites in all samples 202 Xll l Figure 3.47. Acid extractable metal associations with percentage sand in surface sediments (0-2 cm) at the Philip and Blueridge sites in all samples 203 Figure 4.1. Metal concentrations of invertebrate species at Philip, Blueridge, and Reference sites 254 Figure 4.2. Correlation analysis between log-log transformed mean concentrations of N i , Cu, Zn, Cd and Pb in Balanus glandula and weight-adjusted Mytilus trossulus 255 Figure 4.3. Metal concentrations of macroalgae species at Philip, Blueridge, and Reference sites 256 Figure 4.4. Correlation analysis between log-log transformed mean concentrations of N i , Cu, Zn, Cd and Pb in Enteromorpha sp. and Fucus gardneri 257 Figure 4.5. Species-species correlation analysis between log-log transformed mean concentrations of N i , Cu, Zn, Cd and Pb in invertebrates and macroalgae 258 Figure 4.6. Associations of water metal data with metal bioaccumulations in Enteromorpha sp., Fucus gardneri, Mytilus trossulus, and Balanus glandula 259 Figure 4.7. Associations of sediment metal data with metal bioaccumulations in Enteromorpha sp., Fucus gardneri, Mytilus trossulus, and Balanus glandula 260 Figure 4.8. Associations of sediment characteristics with metal bioaccumulations in Enteromorpha sp., Fucus gardneri, Mytilus trossulus, and Balanus glandula 261 Figure 5.1. Schematic of plexiglass frame with barnacle settling plates 282 Figure 5.2. Metal concentrations of Balanus glandula grown at different plate (tidal) heights at the Philip site 283 Figure 5.3. Metal concentrations of Balanus glandula grown on plexiglass (bottom plate) or rock substrata at the Philip site 284 Figure 5.4. Average DGT-labile N i , Cu, and Cd concentrations at the Philip, Blueridge, and Lighthouse sites for seven deployment periods during February 6 to May 20, 2001. 285 xiv LIST OF ABBREVIATIONS A V S Acid volatile sulfides A N C O V A Analysis of covariance A N O V A Analysis of variance BD Below detection cm Centimeter C V Coefficient of variation DGT Diffusive gradient in thin film (Diffusive gel technique) DOC Dissolved organic carbon D O M Dissolved organic matter EG Environmental grade g Gram hr Hour ICP-MS Inductively coupled plasma mass spectrometry (spectrometer) kg Kilogram L Litre MIS Median International Standard for edible shellfish pg Microgram um Micrometer mg Milligram mL Millilitre mm Millimeter min Minute M Mol/litre nm Nanometer N Normal (1 gram equivalent/liter) ppt (%o) parts per thousand RSD Relative standard deviation s Second S E M Simultaneously extracted metals sp. Species (single) spp. Species (plural) STD Standard deviation SPM Suspended particulate matter v/v Volume ratio W Watt wt. Weight w/w Weight ratio xv ACKNOWLEDGEMENTS I am truly indebted to the many people that contributed in so many ways for the completion of this work. My journey was long and arduous and greatly eased by your skills, labour, encouragement and support. I am grateful to my supervisor, Dr. A l Lewis, for his support of the project from its inception to the many final administrative steps necessary for its completion; to Sharon DeWreede for her warmth in welcoming me to Vancouver and always being helpful; to Richard Boase and the District of North Vancouver for all of their assistance; to Allison Fedrigo along with Dr. Ken Hall and Richard Boase for the introduction to this project; and to those who served on my supervisory committee, Dr. Kristin Orians, Dr. Leah Bendell-Young, Dr. Ken Hall, Dr. Tom Pedersen, for their availability to answer questions, provide guidance and feedback, and of course equipment. In particular, I want to thank Dr. Orians for the generous provision of lab space where I could perform so much of my work and to Dr. Tom Pedersen and Dr. Steve Calvert for the use of their trace metal spaces in the Oceanography department. I am very thankful to Bert Mueller for the training, technical assistance and invaluable guidance with many aspects of trace metal chemistry and ICP-MS protocol development; and Maureen Soon for teaching and guiding me through microwave digestions, coulometry, and C N sample preparations. I also wish to acknowledge and thank Jay McNee and Lorax Environmental Services Ltd. for their contributions of time, space, and materials in assisting me with the preparation of DGT gels. Much gratitude is owed to Helen Drost, April Truchan, Saskia Arneson, for their tireless efforts in the field at all hours and in all weather conditions, for their much needed assistance with the endless lab work (you too Mohammed), and for being great students willing to learn and share insights. I also wish to acknowledge here appreciation for the G.R.E.A.T. Scholarship received from the Science Council of B.C. Finally, I must acknowledge and thank my family for their relentless and loving encouragement and support throughout my post-graduate experience; and my close friends for providing all the necessary reminders and cheers throughout the last few years. And of course a heartfelt thanks to Finnegan for temporarily sacrificing his svelte physique as short walks replaced his usual romps while I wrote and wrote and wrote. xvi CHAPTER 1 INTRODUCTION AND BACKGROUND INFORMATION l 1.1. Introduction 1.1.1. Motivation The expanding urbanization of North America with population growth and increased industrialization has dramatically modified the natural landscape and waterways. Where watersheds once provided natural containment and storage for stormwater, urban development has led to increased paved surfaces. These impervious streets, parking lots, driveways, rooftops etc. rather than absorb stormwater produce large volumes of runoff. Over the past thirty years, industries have faced increasing pollution abatement regulations. As a result, significant reductions in municipal and industrial wastewater discharges have followed. At the same time, non-point pollution sources, such as urban runoff or stormwater, have grown to be significant contributors in the contamination of natural waters (Burton & Pitt, 2002; Gibb et al, 1991). In the last two decades, multiple studies have demonstrated the significant contribution of trace metals from stormwater runoff occurring from various land use areas (Butler et al, 1995; Fedrigo & Boase, 1997; McCallum, 1995; Morrison etal, 1988; Schueler & Shepp, 1993). These metals were identified by the Nationwide Urban Runoff Program in the United States as the pollutants having the greatest risk to aquatic life (McCallum, 1995; Whipple Jr., 1987). In recognition of the risks imposed by elevated concentrations of metals and other chemicals in stormwater runoff, some municipalities have adopted strategies to control the release of these pollutants into receiving waters. Oil/grit chamber systems are one such management practice that was adopted in the early 1990s by the Environmental Protection Office of the District of North Vancouver. In a collaborative effort between researchers at the University of British Columbia and the District of North Vancouver, a preliminary study conducted a few years after installation of oil/grit chamber systems along the North Shore demonstrated that total Cu was accumulated within and then discharged out of these chambers during storm events (Fedrigo & Boase, 1997). This trend was demonstrated for Cu and other metals (Zn, Cd, Pb) in similar stormwater interceptor chambers (Schueler & Shepp, 1993) as well as in comparable roadside gullypot systems (Butler et al, 1995; Lee et al, 1997; Morrison et al, 1988; Wei & Morrison, 1993) in the United States and Europe. Altogether those results 2 presented an opportunity to investigate metal contamination from stormwater runoff within a marine setting and place it within the context of potential ecological risks. A detailed evaluation of the effectiveness of the chamber system as a management technique for urban stormwater runoff was sought with specific emphasis on the management of biologically available metals. Metal bioavailability, which refers to the uptake of metals by organisms, and field estimation of metal lability are, to variable degrees, better indicators of potential ecological/biological impacts due to metal pollution than measures of total metals. Consequently, metal fate and distribution were addressed with the emphasis on bioavailability within marine chamber receiving systems. In addition, suitability of certain monitoring tools was evaluated for their assessments of ecological risks imposed by metal pollution in aquatic systems. The selection of metals for this study included Cd and N i in addition to Cu, Zn, and Pb, which were examined within these chamber systems previously by Fedrigo and Boase (1997). A l l of these trace metals are of interest due to their anthropogenic sources and common contamination problems in estuarine and coastal environments as well as their potential for toxicity to aquatic organisms. 1.1.2. Objectives On the basis of these primary interests, six main objectives were defined for the evaluation of metal fate, distribution, and associated potential risks within two North Vancouver stormwater chamber systems and adjacent marine receiving waters. 1. Examination of the influence of chamber system on metal lability in addition to total metals in stormwater runoff during different storm conditions, seasons, and years. Fedrigo and Boase (1997) determined total and total dissolved (<0.45 pm pore size) Cu, Zn and Pb in water from the inlet and outlet of an industrial site chamber in North Vancouver during two fall storms in 1995. Further elucidation was required and considered advisable and beneficial for the general assessment of these stormwater management systems in their local geographic setting. Furthermore, metal lability was not previously addressed nor was any attempt made to relate the findings to potential environmental risks. Within this study, the changes in distribution and 3 speciation of metals from stormwater runoff to marine receiving waters were examined with the understanding that the risks imposed by metal contamination are reliant on metal fate in the environment (metal physical and biogeochemical associations). 2. Determination of salinity influences on stormwater metal lability and/or fate. The coastal location of the chamber systems provided the opportunity to examine salinity impacts on metal concentrations and distributions. Changes in salinity have been shown to influence metal concentration and speciation through flocculation, precipitation and aggregation (Sholkovitz, 1976, 1978), scavenging by iron and manganese oxyhydroxides (Johnson, 1986; Warren & Zimmerman, 1993), ion-exchange mechanisms that release metals from suspended sediments (Duinker, 1980), particulate organic matter (Paulson et al, 1991) or release from and exchange between solid sorbent phases (Brennan, 1991). Since environmental fate of metals within marine systems is inherently dependent on salinity influences, some form of assessment was warranted. 3. Determination of metal distribution and speciation within sediments of the two oil/grit chambers and receiving basins during different years for determination of large scale changes and/or improvements in total and labile metal concentrations. Total and labile metals were determined and used to assess differences in sediment metal lability over space and time attributable to stormwater runoff and influence of the oil/grit chamber systems. 4. Assessing monitoring techniques for the determination of metal lability in sediments and water. The diffusive gradient in thin film or DGT technique has shown promise in field measurements of labile metals by the research work of H. Zhang and W. Davison (Davison et al, 1994; Zhang & Davison, 1995; Zhang et al, 1998; Zhang et al, 1995). The technique's potential for quick and ecologically relevant metal determinations offered an attractive tool for monitoring metals in stormwater runoff 4 and receiving waters. DGT was used alongside other field monitoring analyses in water (filtration: total dissolved, suspended particulate matter) and sediment (labile: acid-extractable, total: digestion, porewater: peepers) for assessment of metal distributions and lability. 5. Assessing metal bioavailability through metal bioaccumulation in a suite of littoral species to determine the usefulness of metal biomonitoring as it relates to urban stormwater runoff to marine systems. As well, these metal bioaccumulations are related to environmental measurements in the assessment of suitable estimates of metal bioavailability. The determination of metal loads in biota tissue, in particular from several species with variable metal exposures (i.e. solute, suspended particulate) provides an assessment of metal distribution that is biologically relevant in terms of potential for exerting toxic effects. Furthermore, the ability of certain aquatic species to bioaccumulate metals gives an integrated measure of metal contamination that can be assessed over spatial and temporal scales. 6. Evaluation of a common North American barnacle species as metal biomonitor with cosmopolitan potential. Much focus has been applied to metal bioaccumulation in barnacles, and these sessile crustaceans have been demonstrated to be effective bioaccumulators of trace metals (Phillips & Rainbow, 1988; Powell & White, 1990; Rainbow et al, 2004; Rainbow et al, 2002; Rainbow et al, 2000; Walker & Foster, 1979; Weeks et al, 1995). The common North American species, Balanus glandula, has not been used previously in metal biomonitoring work. This thesis provides some relevant information on the metal bioaccumulation characteristics of this species and baseline metal data for future work within this region and for comparisons across regions of this barnacle's geographical distribution. 5 1.1.3. Approach Over a three year period, sampling of water and sediments within oil/grit chambers and their marine receiving waters was performed at the only two suitable sites in North Vancouver (Fig. 1.1). One site within an industrial/commercial setting was investigated intensely over different seasons and climate conditions, while the second site was added to the sampling regiment on a few occasions for comparative purposes. Biota were sampled once at both sites and at an additional coastal reference site in North Vancouver. Total metal loads were determined and compared across sites, stations, and species. 1.1.4. Thesis Organization This thesis consists of six chapters. The motivation, objectives, and approach have been discussed; the following provides a description of the study area, some general aspects of stormwater runoff and management and the physical, chemical, and biological properties of the trace metals investigated as well as an introduction to trace metal bioavailability. The results and findings of this study are presented in the next four chapters. Metals in water, sediment, and biota are presented in three separate chapters (Chapters 2 to 4) while Chapter 5 focuses entirely on metals in the resident barnacle species, Balanus glandula. The final chapter (Chapter 6) presents summary conclusions for the entire thesis and recommendations for future research. 1.2. Background Information 1.2.1. Study Area Two sites with installed oil/grit chambers were selected for study within the District of North Vancouver, which lies within the Greater Vancouver Regional District (GVRD) in British Columbia, Canada (Fig. 1.1). The chambers at both sites lie at the foot of small marine embayments that lead into Vancouver Harbour which forms the major portion of Burrard Inlet, situated between First and Second Narrows. The Harbour, which is bordered 6 by North Vancouver on its north shore and the city of Vancouver on the south shore, is an average 2.5 kilometers wide and eight kilometers long with a maximum depth of 66 m. Vancouver harbour is the largest port on the west coast of Canada and has undergone significant shoreline modification for dock construction. The experimental site, Philip (Fig. 1.2) is located within a highly industrialized and commercialized area at the foot of Philip Ave. and is close enough to receiving waters for frequent saltwater intrusions into the chamber. The receiving basin is partially enclosed at the seaward end where a small culvert permits passage between this inner basin and the outer marine embayment. The control site, Blueridge (Fig. 1.3) is located downstream from an oil/grit chamber situated at an elevation of approximately 45 m. The chamber drains a fairly new residential area and discharges directly into a small freshwater stream (Blueridge Creek) that, within 300 m, flows into a small marine embayment. Water and biota samples were also obtained for selected analyses from a reference site, Lighthouse Park (Fig. 1.1), in a West Vancouver recreational area. 1.2.2. Urban Stormwater Runoff and Management Practices Hydrology within a given area can be significantly altered due to urban development. Natural porous materials (i.e. soil, vegetation) permit infiltration and absorption of rain and snowmelt, while impervious paved surfaces (i.e. streets, parking lots, roofs) can lead to massive, fast-moving lateral stormwater flows transported to our waterways. Pollutants and debris that accumulate on surfaces between rain events are carried with these flows and commonly include organic contaminants, trace metals, nutrients, sediments, pathogens, and other toxins. Contributions can occur from various activities including gardening (pesticides, fertilizers), industrial and commercial businesses (chemical leaks, spills), and highway traffic (metals, particulates, petroleum compounds), and can be directly deposited onto surfaces, eventually settle after atmospheric release (i.e. vehicle exhaust), or are wet deposited with precipitation scavenging. Typical pollutants and their average concentrations in urban stormwater are listed in Table 1.1. The trace metals that commonly occur in high levels are cadmium, copper, lead and zinc. 7 Table 1.1. Typical Pollutant Concentrations Found in Urban Stormwater Runoff* Pollutants Units Average Concentration*'* Total Suspended Solids mg/L 80 Total Phosphorus mg/L 0.30 Total Nitrogen mg/L 2.0 Total Organic Carbon mg/L 12.7 Fecal Coliform Bacteria MPN & /100 mL 3600 E. coli Bacteria MPN/100 mL 1450 Petroleum Hydrocarbons mg/L 3.5 Cadmium ug/L 2 Copper pg/L 10 Lead pg/L 18 Zinc Hg/L 140 Chlorides (winter only) mg/L 230 Insecticides pg/L 0.1 to 2.0 Herbicides pg/L 1 to 5.0 ( 'F rom: (Maryland Department of Environment, 2000) ( & ) M P N : Most probable number ( } Mean or median concentrations at typical sites (may be greater during individual storms and can be 2 to 10 times higher from stormwater hotspots). The volume of stormwater runoff and related urban pollutant loads are directly related to the degree of imperviousness within a given watershed and some level of impact is inevitable within any urban setting. The degree of impact can usually be minimized greatly through integrated watershed management approaches at the urban design stage. Often, however, the practical approach involves a retrofit of stormwater management options within already developed areas. The best management practices for urban stormwater runoff comprise two approaches: source control and treatment. Source control focuses on the reduction of pollutants through pollution prevention type strategies, while treatment options involve removal of pollutants from stormwater runoff through delay, capture, storage, treatment or infiltration options prior to discharge into natural waterbodies. The broad 8 categories include ponds, wetlands, infiltration, filtering systems, and open channels. A review of 139 stormwater management performance studies in the U.S. from 1977 to 2000 by the Center for Watershed Protection, Maryland (Winer, 2000) produced mean pollutant removal efficiencies for the different treatment groups which are summarized in Table 1.2. Although few studies with considerable variability could be included within each group, the general conclusions were that infiltration and filtration practices had the highest removal rates while dry ponds and the related oil/grit systems and stormceptors had the lowest. 9 Table 1.2. Median Pollutant Removal (%) of Stormwater Treatment Practices1 T S S ( i ) Sol P ( 3 ) T N ( 4 ) NOx<5> Cu Zn N ( 6 ) Stormwater Dry Ponds Quality Control Ponds 3 19 0 5 9 10 5 3 Dry Extended DetentionPond 61 20 -11 31 -2 29 29 6 Stormwater Wet Ponds Wet Extended DetentionPond 80 55 67 35 63 44 69 14 Multiple Pond System 91 76 69 N / A 87 N / A N / A 1 Wet Pond 79 49 62 32 36 58 65 29 Stormwater Wetlands Shallow Marsh 83 43 29 26 73 33 42 23 Extended Detention Wetland 69 39 32 56 35 N / A -74 4 Pond/Wetland System 71 56 43 19 40 58 56 10 Submerged Gravel Wetland 83 64 -10 19 81 21 55 2 Filtering Practices Organic Filter 88 61 30 41 -15 66 89 7 Perimeter Sand Filter 79 41 68 47 -53 25 69 3 Surface Sand Filter 87 59 -17 32 -13 49 80 8 Vertical Sand Filter 58 45 21 5 -87 32 56 2 Bioretention N / A 65 N / A 49 16 97 95 1 Infiltration Practices N / A 100 100 42 82 N / A N / A 3 Porous Pavement 95 65 10 83 N / A N / A 99 3 Open Channels Ditches 31 -16 -25 -9 24 14 0 9 Grass Channel 68 29 40 N / A -25 42 45 3 Dry Swale 93 83 70 92 90 70 86 4 Wet Swale 74 28 -31 40 31 11 33 2 Other Oil-Gr i t Separator -8 -41 40 N / A 47 -11 17 1 Stormceptor® 25 19 21 N / A 6 30 21 1 ( 'F rom: (Winer, 2000) ( I ) TSS: Total Suspended Solids; ( 2 ) TP: Total Phosporus; ( 3 ) Sol P: Soluble Phosphorus; ( 4 ) T N : Total Nitrogen; ( 5 ) N O x : Nitrate and Nitrite Nitrogen; ( 6 ) N : Number of studies incorporated in review. 10 The oil/grit chamber systems installed at various locations within the District of North Vancouver by the Environmental Protection Office in an effort to reduce release of stormwater pollutants and wastes into receiving waters are essentially based on a delay-capture design similar to stormwater ponds. These underground concrete sediment chambers, however, are generally used for smaller drainage areas of high imperviousness. A schematic of a typical chamber system is presented in Fig. 1.4. As stormwater is channeled through the chambers, baffles fitted within the chamber function to reduce turbulence and facilitate settling of particles and associated contaminants, while the inverted elbow pipe allows for oil and grease separation via flotation (BC Research Corporation, 1992). 1.2.3. Trace Elements Investigated (Cu, Zn, Cd, Pb, Ni) 1.2.3.1. Natural and Anthropogenic Sources The metals investigated in this study, Cu, Zn, Cd, Pb, and N i , are considered trace metals due to their natural occurrence in low concentrations generally <5 pg/L dissolved in seawater and <100 mg/kg in crustal rock (Bruland, 1983; Cox, 1989; Sadiq, 1992), although all can be found in higher concentrations in natural ores. They can all be introduced to the oceans through the atmosphere (wind-borne soil, volcanic activity, forest fires) and river systems (anthropogenic, fluvial, naturally dissolved) and can also be present associated with marine and terrestrial biogenic particles. Anthropogenic sources can exceed natural inputs, with the most significant sources arising from industrial, domestic, and agricultural wastes; automobile wastes and emissions; and metal smelting, refining and manufacturing industries (Kayhanian et al, 2003; McCallum, 1995; Nriagu, 1989; Nriagu & Pacyna, 1988). 1.2.3.2. Physical, Chemical, and Biological Properties Cu, Zn, Cd, and N i are transition elements with one to nine outer shell (d) electrons, stable ionic forms, relatively low ionization potentials and the ability to form organic and inorganic complexes. The bivalent oxidation state is most common, although Cu can form 11 stable +1 complexes. Pb is a /?-block element with 4 valence electrons of which two are readily ionizable producing the +11 oxidation state typical of its predominant inorganic forms in natural aquatic systems. Metal cations are commonly classified into three groups (Class A, Class B, and transition cations) according to their metal-ligand complex stabilities and electron configuration (Stumm & Morgan, 1996). Class A metal ions (Ca 2 + , M g 2 + , Na + , K + , A l 3 + ) exhibit the inert gas type (d°) electron configuration which is assumed to be of spherical symmetry. The electron sheaths are described as hard spheres since they are not readily deformed by surrounding charged ions (low polarizability). In contrast, class B metal ions (Cu +, Z n 2 + , C d 2 + , Pb 2 +) exhibit nd 1 0 and nd1 0(n+l)s2 configurations which form easily deformable electron sheaths (i.e. soft spheres with higher polarizabilities; Stumm & Morgan, 1996). Transition metal cations (Ni 2 + , C u 2 + , Z n 2 + , Mn 2 + ) , with their partially filled d orbitals, tend to display complex stability that increases as ionic potential (valence/radius) and electronegativity increase in the generally established Irving-Williams order, M n 2 + (d5) < Fe 2 + (d6) < C o 2 + (d8) < N i 2 + (d8) < C u 2 + (d9) > Z n 2 + (d1 0) (McNee, 1997; Stumm & Morgan, 1996). These classifications help explain the complexes most often seen in natural waters (Table 1.3). The selected trace elements are often complexed with inorganic ligands such as C ( V " , Cl" and OFT. Additionally, organic complexation is common for Cu but also observed with Zn, Cd, and N i . In anoxic environments, the formation of insoluble metal sulfides is the predominant complexation reaction. Table 1.3. Major Inorganic Species in Natural Waters* Element Major Species'1' Cu(II) CuC0 3 , Cu(OH) 2° Zn(II) Zn 2 + , ZnC0 3 ° (Zn 2 + , ZnCf) Cd(II) C d 2 + , CdC0 3 ° (CdCl 2) Ni(II) N i 2 + , NiC0 3 ° (N i 2 + , NiCl + ) Pb(II) PbC0 3 °, (PbCl + , PbC0 3 ) 'Adapted from Stumm & Morgan (1996). ( l )Where major species in seawater deviate from those in fresh waters, major seawater species are given in parentheses. 12 Particles in aquatic systems play a leading role in the transport and cycling of trace elements in aquatic systems. Despite their great compositional diversity (clays, minerals, organic particles, humus, biological debris, organic coatings), they display a strong capacity to bind trace metals via ion exchange or adsorption (Stumm & Morgan, 1996). Autochthonous, atmospheric, and most importantly riverine sources provide metal-laden particulates to oceans that are subject to biological processes (uptake by organisms and subsequent release through association with organic debris) and physico-chemical mechanisms (adsorption onto particle surfaces, particle aggregation and particle-facilitated metal precipitation). Trace metals are then transported to sediments via sinking particles. In the case of Pb, particle scavenging is quite efficient resulting in immediate declines within the water column both with water depth and with distance from anthropogenic sources. Many trace metals (as in this case for Cd, N i , Zn) follow nutrient-type profiles in the ocean where removal from surface waters by particles is followed by remineralization at depth which is controlled by their specific biogeochemical cycles, the characteristics of the surrounding medium, circulation and physical mixing. Cu oceanic profiles reflect a combination of both nutrient and scavenged type distributions with depth (Li, 1991). The free ions are the most toxic form of these metals in aquatic systems and for Zn and N i are the predominant forms found in seawater (Birge & Black, 1980; Sadiq, 1992; Stumm & Morgan, 1996; Turner et al, 1981). Metal toxicity is generally reduced when salinity increases as a result of the increased complexing capacity of seawater (most notable in chlorocomplexation of Cd) or elimination from the solute phase via adsorption and precipitation, and/or flocculation, aggregation, and coagulation processes involving both inorganic Fe/Mn oxides/hydroxides and organic material (Santschi et al, 1997; Sholkovitz, 1978; Stumm & Morgan, 1996). The trace metals investigated consist of both essential (Cu, Zn, Ni) and non-essential (Cd, Pb) elements. Cu, Zn, and N i are required by most aquatic organisms in order to maintain normal metabolic function while Cd and Pb can be detrimental at even low concentrations (Campbell et al, 1988). In reality, irrespective of nutritional necessity, all trace metals can exert toxic effects when above a threshold bioavailability. Of the five elements studied, Cd is the most toxic and generally one of the most biotoxic elements within the aquatic environment (Sadiq, 1992). Potential toxicities of trace elements are carefully 13 considered in the development of environmental quality guidelines reported alongside the presentation of data within the next chapters. 1.2.4. Metal Bioavailability Knowledge of the biological availability of a metal is central to the understanding of its biotic impact. The concentration of a metal available for uptake by aquatic organisms or the bioavailable fraction is not necessarily the concentration within the surrounding medium (Bjerregaard & Depledge, 1994; Langston, 1990; Luoma & Carter, 1991; Sadiq, 1992). Both abiotic and biological factors are involved in determining the portion of metal that is accessible to living organisms. Metal uptake can occur through direct environmental exposure or ingestion and three strategies are generally described for the resulting metal fate: (1) participation in metabolic processes with nutritional or toxic role; (2) elimination from the organism back to the environment; and (3) sequestration within the organism with reduction in toxic potential. Different organisms vary in their metal uptake pathways which can depend on the organism's physiology and relative cellular processes, ecology, and relative ability to regulate metals. What remains key for ecological considerations within broad pollution monitoring work is the potential for detrimental effects if biologically available metals occur in excess. The chemical form of a metal in the aquatic environment plays a major role in its biological availability. Low concentrations of metals normally occur in the water column as a result of the tendency of metals to complex and precipitate out of solution. For some metals, an even smaller portion (usually the free ionic form) of the total dissolved metal is correlated with biotic uptake (Bjerregaard & Depledge, 1994; Campbell, 1995; Langston, 1990; Sadiq, 1992). This is based on the free-ion activity model formulated by Morel (1983) which assumes that metals in the free ionic form pass unhindered through cell membranes where they may then exert an effect (Campbell, 1995; Campbell et al., 1988). A considerable body of experimental work that has been compiled and reviewed by Campbell (1995) has supported the model's prediction (with some exceptions and qualifications). Metal speciation which subsequently influences bioavailability (Barron, 1995) within the 14 water column can be affected by water quality parameters such as pH, redox state, temperature, salinity, carbonate buffer system, dissolved organic matter content and total hardness (expressed as the equivalent quantity of calcium carbonate and primarily related to the presence of calcium and magnesium). Other crucial considerations are temporal and spatial variations including episodic, seasonal, and annual changes in metal concentrations as well as transport and depositional processes, which can redistribute metals (Frazier, 1979; Luoma & Carter, 1991). Although certain chemical species represent a small portion of the total metal concentration available for biotic absorption within the water column, a substantial portion can be available from sediments. Concentrations of metals within sediments usually exceed concentrations found within the water column and can be greater by a factor of 3 or more (Barron, 1995; Bryan & Langston, 1992). The physical proximity to or nutritional dependence on sediments creates a precarious situation for benthic or epibenthic invertebrates, benthic algae, and other benthic biota in polluted systems. These organisms tend to have higher tissue-metal concentrations than planktonic or pelagic organisms (Campbell et al, 1988; Campbell & Tessier, 1991). Most studies investigating the relationship between concentrations of metals in sediment and bioavailability measure and use only those metal concentrations released from certain sediment fractions (exchangeable fraction, carbonates and hydrous metal oxide fractions, organic fraction, total mobile or labile fraction, which includes all but lattice-bound metals). For organisms living in close proximity to sediments, the contribution from interstitial water can be significant. While sediment solids may serve as a source of metal input, the interstitial water is in direct contact with organisms. Metals can occur as dissolved or colloidal forms within sediment interstitial water or can be adsorbed at particle surfaces, associated with organic matter, or present in the lattice structure of minerals. Only a fraction of these forms is relevant in the study of biological impacts. The ionic and weakly complexed forms of most metals are considered to be the bioavailable and, in excess, toxic to most organisms (Barron, 1995; Bryan & Langston, 1992; Tessier et al, 1984). Benthic detritivores and suspension feeders can ingest metals associated with sediment particles. Depending on the metals' binding capacity to the sediment, they can be released by digestive processes and absorbed into biological tissues. Competition between different binding sites can also prove to be critical. The presence and 15 influence of these metal-binding sites are in turn associated with sediment composition, grain size, dissolved organic matter content (within porewater) and concentrations of various ligands. Smaller grain sizes such as clay or silt, iron or manganese oxides (which scavenge trace metals), and increased organic carbon content have all been related to low bioavailability of trace metals (Sadiq, 1992; Timmermans, 1993). While the determination of metal bioavailability is of prime interest, the approach often utilized for its practicality is its estimation by direct measurements of metal lability within physical environmental media. The labile fraction of a metal is operationally defined as the fraction of metal able to dissociate to the free metal ion within the time frame and conditions of the analytical technique employed. Aside from the direct metal bioavailabilities obtained using biomonitors, the analytical techniques employed for the assessment of metal lability in this study were DGT in water and sediments, porewater peepers in sediments, and cold acid extractions in sediments. 16 Figure 1.1. Map of North Vancouver and Vancouver Harbour. Location of sampling sites, Philip (PH) and Blueridge (BR), and reference site, Lighthouse (LH) are indicated. 17 Stormwater Runoff Burrard Inlet Figure 1.2. Philip site. Sampling stations AO to A6 are indicated. 18 Figure 1.3. Blueridge site. Sampling stations C3 to C6 are indicated. Figure 1.4. Schematic of Oil/Grit Chamber. 20 1.3. References Barron, M.G. (1995). Bioaccumulation and bioconcentration in aquatic organisms. In Handbook of Ecotoxicology. (ed. D.J. Hoffman, B.A. Rattner, G.A. Burton Jr. & J. Cairns Jr.), pp. 652-666. CRC Press, Inc., Boca Raton. BC Research Corporation (1992). Urban Runoff Quality Control Guidelines for British Columbia. Prepared for: Municipal Waste Reduction Branch, Environmental Protection Division, BC Environment, Victoria, BC. Birge, W.J. & Black, J.A. (1980). Aquatic toxicology of nickel. In Nickel in the Environment. (ed J.O. Nriagu), pp. 349-366. JW and Sons, New York. Bjerregaard, P. & Depledge, M . H . (1994). Cadmium accumulation in Littorina littorea, Mytilus edulis and Carcinus maenas: the influence of salinity and calcium ion concentrations. Mar. Biol, 119, 385-395. Brennan, B. (1991). Chemical partitioning and remobilization of heavy metals from sewage sludge dumped in Dublin Bay. Water Res., 25, 1193-1198. Bruland, K.W. (1983). Trace elements in seawater. In Chemical Oceanography, (ed. J.P. Riley & R. Chester). Academic Press, New York. Bryan, G.W. & Langston, W.J. (1992). Bioavailability, accumulation and effects of heavy metals in sediments with special reference to United Kingdom estuaries: a review. Environ. Pollut., 76, 89-113. Burton, G.A. & Pitt, R. (2002). Stormwater effects handbook: a toolbox for watershed managers, scientists, and engineers Lewis Publishers, New York. pp. Butler, D., Xiao, Y. , Karunaratne, S.H.P.G., & Thedchanamoorthy, S. (1995). The gully pot as a physical, chemical and biological reactor. Water. Sci. Technol, 31, 219-228. Campbell, P.G.C., Lewis, A .G. , Chapman, P.M., Crowder, A . A . , Fletcher, W.K., Imber, B., Luoma, S.N., Stokes, P.M., & Winfrey, M . (1988). Biologically available metals in sediments, Rep. No. No. 27694. National Research Council of Canada, Ottawa. Campbell, P.G.C. & Tessier, A . (1991). Biological availability of metals in sediments: analytical approaches. In Heavy Metals in the Environment, (ed J.P. Vernet), pp. 161-173. Elsevier Sci. Publ. Co. Inc., Amsterdam. Campbell, P.G.C. (1995). Interactions between Trace Metals and Aquatic Organisms: A Critique of the Free-ion Activity Model. In Metal Speciation and Bioavailability in Aquatic Systems, (ed. A. Tessier & D.R. Turner), pp. 45-102. John Wiley & Sons Ltd., New York. Cox, P. A . (1989). The Elements: Their origin, Abundance and Distribution Oxford University Press, Oxford, U.K. pp. Davison, W., Zhang, H. , & Grime, G.W. (1994). Performance characteristics of gel probes used for measuring the chemistry of pore waters. Environ. Sci. Technol, 28, 1623-1632. Duinker, J.C. (1980). Suspended matter in estuaries: Adsorption and desorption processes. In Chemistry and Biochemistry of Estuaries, (ed. E. Olausson & I. Cato). Wiley, New York. Fedrigo, A . M . & Boase, R. (1997). Sediment and Stormwater Chemistry of Oil-Grit Chambers. Department of Civil Engineering, University of British Columbia, Vancouver and District of North Vancouver, B.C. 21 Frazier, J .M. (1979). Bioaccumulation of cadmium in marine organisms. Environ. Health Persp.,2%, 75-79. Gibb, A. , Bennet, B., & Birkbeck, A. (1991). Urban runoff quality and treatment: a comprehensive review. Prepared for the City of Burnaby, B.C. Johnson, C A . (1986). The regulation of trace element concentrations in river and estuarine waters contaminated with acid mine drainage: The adsorption of Cu and Zn on amorphous Fe oxyhydroxides. Geochim. Cosmochim. Acta, 50, 2433-2438. Kayhanian, M . , Singh, A. , Suverkropp, C , & Borroum, S. (2003). Impact of annual average daily traffic on highway runoff pollutant concentrations. J. Environ. Eng., 129, 975-990. Langston, W.J. (1990). Toxic effects of metals and the incidence of metal pollution in marine ecosystems. In Heavy Metals in the Marine Environment, (ed. R. W. Furness & P.S. Rainbow), pp. 101-122. CRC Press, Boca Raton. Lee, P.K., Baillif, P., & Touray, J.C. (1997). Geochemical behaviour and relative mobility of metals (Mn, Cd, Zd and Pb) in recent sediments of a retention pond along the A-71 motorway in Sologne, France. Environ. Geo!., 32, 142-152. L i , Y . - H . (1991). Distribution patterns of the elements in the ocean: A synthesis. Geochim. Cosmochim. Acta, 55, 3223-3240. Luoma, S.N. & Carter, J.L. (1991). Effects of trace metals on aquatic benthos. In Metal Ecotoxicology: Concepts and Applications, (ed. M.C. Newman & A.W. Mcintosh), pp. 261-300. Lewis Publications, Chelsea. Maryland Department of Environment (2000). Maryland Stormwater Manual, Baltimore. 212 pp. McCallum, D. (1995). An Examination of Trace Metal Contamination and Land Use in an Urban Watershed. M . A . S c , University of British Columbia, Vancouver, B.C. McNee, J.J. (1997). The Post-Depositional Cycling of Cd, Cu, Mo and Zn in Several Hydrographically Distinct B.C. Fjords. Ph.D., University of British Columbia, Vancouver. Morel, F . M . M . (1983). Principles of Aquatic Chemistry Wiley-Interscience, New YOrk. 301 pp. Morrison, G.M.P., Revitt, D .M. , Ellis, J.B., Svensson, G., & Balmer, P. (1988). Transport mechanisms and processes for metal species in a gullypot system. Water Res., 22, 1417-1427. Nriagu, J.O. & Pacyna, J.M. (1988). Quantitative assessment of worldwide contamination of air, water and soils with trace metals. Nature, 333, 134-139. Nriagu, J.O. (1989). A global assessment of natural sources of atmospheric trace metals. Nature, 338, 47-49. Paulson, A.J. , Curl, H.C., & Cokelet, E.D. (1991). Remobilization of Cu from marine particulate organic matter and from sewage. Mar. Chem., 33, 41-60. Phillips, D.J.H. & Rainbow, P.S. (1988). Barnacles and mussels as biomonitors of trace elements: a comparative study. Mar. Ecol-Prog. Ser., 49, 83-93. Powell, K.J. & White, S.L. (1990). Heavy metal accumulation by barnacles and its implications for their use as biological monitors. Mar. Environ. Res., 30, 91-118. Rainbow, P.S., Wolowicz, M . , Fialkowski, W., Smith, B.D., & Sokolowski, A . (2000). Biomonitoring of trace metals in the Gulf of Gdansk, using mussels (Mytilus trossulus) and barnacles (Balanus improvisus). Water Res., 34, 1823-1829. 22 Rainbow, P.S., Smith, B.D., & Lau, S.S.S. (2002). Biomonitoring of trace metal availability in the Thames estuary using a suite of littoral biomonitors. J. Mar. Biol. Ass. U.K., 82, 793-799. Rainbow, P.S., Fialkowski, W., Sokolowski, A. , Smith, B.D., & Wolowicz, M . (2004). Geographical and seasonal variation of trace metal bioavailabilities in the Gulf of Gdansk, Baltic Sea using mussels (Mytilus trossulus) and barnacles (Balanus improvisus) as biomonitors. Mar. Biol, 144, 271-286. Sadiq, M . (1992). Toxic Metal Chemistry in Marine Environments. Marcel Dekker, Inc., New York. pp. Santschi, P.H., Lenhart, J.J., & Honeyman, B.D. (1997). Heterogeneous processes affecting trace contaminant distribution in estuaries: The role of natural organic matter. Mar. Chem., 58, 99-125. Schueler, T. & Shepp, D. (1993). The Quality of Trapped Sediments and Pool Water Within Oil Grit Separators in Suburban MD. , Metro Washingto COG. Sholkovitz, E.R. (1976). Flocculation of dissolved organic and inorganic matter during the mixing of river water and seawater. Geochim. Cosmochim. Acta, 40, 831-845. Sholkovitz, E.R. (1978). The flocculation of dissolved Fe, Mn, A l , Cu, N i , Co and Cd during estuarine mixing. Earth and Planetary Science Letters, 41, 77-86. Stumm, W. & Morgan, J.J. (1996). Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, pp. 1022. John Wiley & Sons, Inc., New York. Tessier, A. , Campbell, P .G .C , Auclair, J .C , & Bisson, M . (1984). Relationships between the partitioning of trace metals in sediments and their accumulation in the tissues of the freshwater mollusc Elliptio complanata in a mining area. Can. J. Fish. Aquat. Sci., 41, 1463-1472. Timmermans, K.R. (1993). Accumulation and effects of trace metals in freshwater invertebrates. In Ecotoxicology of Metals in Invertebrates, (ed. R. Dallinger & P.S. Rainbow), pp. 143-175. CRC Press, Boca Raton. Turner, D.R., Whitfield, M . , & Dickson, A .G . (1981). The equilibrium speciation of dissolved components in freshwater and seawater at 25°C and 1 atm pressure. Geochim. Cosmochim. Acta, 45, 855-882. Walker, G. & Foster, p. (1979). Seasonal variation of zinc in the barnacle, Balanus balanoides (L.) maintained ona raft in the Menai Strait. Mar. Environ. Res., 2, 209-221. Warren, L .A . & Zimmerman, A.P. (1993). Trace metal-suspended particulate matter associations in a fluvial system: physical and chemical influences. In Particulate Matter and Aquatic Contaminants, (ed S.S. Rao). Lewis Publishers. Weeks, J .M., Rainbow, P.S., & Depledge, M . H . (1995). Barnacles (Chthamalus stellatus) as biomonitors of trace metal bioavailability in the waters of Sao Miguel (Azores). Agoreana, Supplement, 103-111. Wei, C. & Morrison, G . M . (1993). Effect of stormwater runoff on metal distribution in the sediment and interstitial waters of an urban river. Environ. Technol, 14, 1057-1064. Whipple Jr., W. (1987). Implementing dual-purpose stormwater detention programs. Res. Plann. Manage. Div., (Am. Soc. Civ. Eng.), 113, 779-792. Winer, R. (2000). National Pollutant Removal Performance Database for Stormwater Treatment Practices. 2nd Edition. Prepared for EPA Office of Science and 23 Technology In association with TetraTech, Inc. Center for Watershed Protection, Ellicott City, Maryland. Zhang, H. & Davison, W. (1995). Performance characteristics of diffusion gradients in thin films for the insitu measurement of trace metals in aqueous solution. Anal. Chem., 67, 3391-3400. Zhang, H., Davison, W., Miller, S., & Tych, W. (1995). In situ high resolution measurements of fluxes of N i , Cu, Fe, and M n and concentrations of Zn and Cd in porewaters by DGT. Geochim. Cosmochim. Acta, 5 9 , 4181-4192. Zhang, H., Davison, W., Knight, B., & McGrath, S. (1998). In situ measurements of solution concentrations and fluxes of trace metals in soils using DGT. Environ. Sci. Technol., 32, 704-710. 24 CHAPTER 2 DISTRIBUTION OF METALS IN WATERS OF STORMWATER MANAGEMENT OIL/GRIT CHAMBERS AND ADJACENT MARINE RECEIVING AREAS 25 2.1. Introduction The wet climate, vast waterways, coastal location, and urban/industrial development of southwestern British Columbia make stormwater pollution a significant concern in the region. The mild marine climate of the Greater Vancouver area produces a rainy season that lasts from fall to spring with an annual average rainfall of 1,155 mm (Environment Canada). The greatest rainfall and most severe storms occur in fall and winter (typically 70% of annual precipitation), while the lighter rainfall period from late winter to early spring is typically characterized by regular predictable storms and evenly distributed precipitation (20% of annual precipitation). Late spring and summer tend to be drier with occasional localized rainfall events. The annual rainfall for the period of study (1998-2001) and general study area ranged between 1,443 and 1,984 mm with winter and fall precipitation averaging 181 and 244 mm, spring 100 mm and summer 54 mm (Table 2.1). Oil/grit chambers installed in an attempt to reduce stormwater pollution through retention of storm-water sediments have functioned poorly as substantial evidence has indicated (Schueler & Shepp, 1993; Shepp, 1995b, 1995a). In North Vancouver, the preliminary work conducted in a collaborative effort of the University of British Columbia and the District of North Vancouver by Fedrigo and Boase (1997) had indicated that total metals were able to flow unhindered into receiving waters. Under the premise that determining metal lability is critical to the understanding of environmental impacts due to metal pollution, a more thorough investigation was prompted into the metal-related effectiveness of this stormwater management practice, the influence on metal lability, and the ease and usefulness of the DGT technique as a monitoring tool. The selection of an industrial and residential/parkland site provided the opportunity to examine potentially large variations in metal concentrations and provide relevant information pertinent in the increasing urban landscape of modern times. Burrard Inlet has long been identified as a depository of various pollutants with sediments from various locations exceptionally enriched and often exceeding regulatory guidelines (including those for Cu, Zn, Cd, and Pb) (BIEP, 1992). While the Vancouver Wharves area was identified as a major metals contaminant source in the Inlet, other areas 26 have also been found to be contaminated. Recent measurements from a number of sites within the Inlet found Cu, Zn, Cd, Pb, and Ni below provincial objectives set for maximum levels; however, detection limits for Cu, Pb, and N i exceeded objectives for average water quality (MWLAP, 2001). Reported sediment quality was similar to previous findings. The occurrence of stormwater metal contamination is largely documented (in particular for Cu, Zn and Pb) and related to the process of urbanization with its related reduction of permeable surfaces (Cohn-Lee & Cameron, 1992; Davis et al, 2001; Field & Turkeltaub, 1981; McCallum, 1995; Sansalone et al, 1998). Nonetheless, complexities arise in evaluating stormwater contributions to receiving waters and providing useful management tools. Stormwater discharges are periodic and pollutant loads can vary both in type and quantity depending on several factors such as previous meteorological conditions, land use patterns, storm intensity and duration, and watershed characteristics. The need for regional characterization of urban stormwater for its effective management has been recognized and implemented in the Greater Vancouver Regional District (MacDonald, 2003). While the District of North Vancouver (location of study sites) is equipped with separate pipes for stormwaters, stormwater runoff in addition to combined sewer overflows (CSOs) and marine activity related pollution are all well known contributors of non-point sources of pollution to Burrard Inlet. CSO and total stormwater loadings to the Inlet Inner Harbour have been estimated respectively at 1300 kg and 1100 kg of Cu per year, 1200 kg and 4600 kg of Zn per year, 270 kg and 740 kg Pb per year, and 9 kg of Cd in total stormwater per year (Hall et al, 1998; MacDonald, 2003). While the issue of metal contamination has been well addressed, the evaluation of biological impacts or potential for biological impacts of said pollution is at an early stage of development. Limited toxicity testing has been performed on stormwater effluent as a whole (GVRD, 1998; MacDonald, 2003) but does not have the capacity to discriminate between constituents. The more recent efforts by the G V R D are expanding toward an ecosystem approach. The G V R D has outlined an initial extensive program for the evaluation of potential biological effects using benthic community structure as an indicator of impairment due to stormwater discharges in streams (EVS, 2003). 27 A complement to this approach with the ability to identify potential impacts of specific pollutants or category of pollutants would also be useful in this regard. The determination of metal lability, i f easily conducted, can be performed regularly to elucidate specific risks. The diffusion gradient in thin-film hydrogel (DGT) technique has shown promise in this area with its ability to integrate trace metal concentrations over time and the potential to separate species based on the principles of size exclusion and metal lability (Zhang & Davison, 1995). While temporal and spatial variability cannot be discreetly determined, a time-averaged approach may be more than adequate for the evaluation of potential biological impacts of stormwater discharges. Recent investigations of the technique have identified certain limitations of the technique, but still point to it as an effective tool for monitoring purposes (Twiss & Moffett, 2002). The purpose of this study was to examine the influence of the oil/grit chamber system on labile metals (using DGT) in addition to total metals in stormwater runoff discharged to marine waters during different storm conditions, seasons, and years. Salinity effects on stormwater metal lability and fate of discharged trace metals was also of interest and an attempt was made to characterize these influences. 2.2. Materials and Methods 2.2.1. Field Program Field sampling was carried out at the two North Vancouver sites (Figs. 1.1 to 1.3) during low tides over a three year period, from 1998 to 2001. Water samples were obtained from a total of fifteen sampling periods during three seasons (summer, fall, spring) and different weather conditions. Tides in the general area are mixed semi-diurnal with two low and two high water periods during each 25 hour diurnal cycle, one fortnightly spring-neap cycle, and small monthly and annual variations. The tidal range is fairly consistent throughout Burrard Inlet with a mean tidal range of 3.3 m and a tidal range of 5 m for large tides (Thomson, 1981). Magnitude and direction of currents are spatially variable and generated by the tidally-induced fluctuations in water level. Net water movement, however, is very small due to the oscillatory nature of these currents (Thomson, 1981). The two small 28 sites studied were largely influenced by local episodic stormwater discharges in addition to tidally forced currents. Table 2.2 details the number of sampling occasions during different weather conditions as well as corresponding climate data. Weather conditions were categorized according to cumulative precipitation levels and estimated storm intensities occurring within the 24 hour period prior to sampling. Intensity within a given period was quantified by dividing the total rainfall volume of a continuous rain event by its duration as noted in field observations. From 1998-2001, DGT-labile metals were measured in water (using 24-hour deployment periods) frequently at the Philip site and on selected dates for the Blueridge site. From 1999 to 2001, water samples were collected for dissolved metal analyses. Metals associated with suspended particulate matter or particulate metals were analyzed at both sites on four occasions reflecting three weather conditions. Stormwater concentrations entering the chamber were determined for two storms in November 1999 and one severe storm on September 29, 2000. Water temperature, conductivity, pH, flow, depth, and in some instances dissolved oxygen were also measured. Conductivity was measured using a Hanna Instruments model 9033 multi-range conductivity meter. Flow was measured with a Flow-Mate™ model 2000 portable flowmeter (Marsh-McBirnay, Inc.) and dissolved oxygen was measured with a MultiLine P4™ probe (Wissenschaftlich-Technische Werkstatten, Gimblt Weilheimy Germany). Precipitation levels and climate data were obtained from climatological stations within the vicinity of the sample sites (Environment Canada Climate Data Services, Vancouver, BC). Chamber and outer basin water used for a simulated mixing experiment was collected from the Philip site on September 29. Dissolved and DGT-labile metals were determined for a range of salinities using different ratios of unfiltered stormwater and receiving basin waters. The mixtures were continuously stirred and kept at 15°C overnight prior to a 24 hr DGT deployment period. 29 2.2.2. Analytical Program 2.2.2.1. Reagents Stock and working solutions of trace metals and blanks were prepared using nanopure water and ultrapure grade concentrated nitric acid (Seastar Chemicals Inc., Sidney, BC). Acetic acid and ammonia for buffering and microwave digestions were ultrapure grade (Seastar Chemicals Inc., Sidney, BC). Concentrated nitric, hydrochloric, and hydrofluoric acids used for digestions were environmental grade (EG; Anachemia Science, Richmond BC). A l l plasticware was acid cleaned by soaking in 10% ACS (American Chemical Society) reagent grade or environmental grade (Anachemia Science, Richmond BC) HNO3 for 24 h, rinsing and then soaking in nanopure water for 24 h. Final nanopure rinses occurred before allowing plasticware to air dry in a laminar flow hood. 2.2.2.2. Instrumental Metal Analysis Analyses of metals were performed on two inductively coupled plasma mass spectrometers (ICP-MS). DGT-labile metals from water samples were analyzed on the PlasmaQuad™ ICP-MS (VG PlasmaQuad™ 2+ turbo, V G Elemental, Winsford, Cheshire, U.K.) fitted with a 1 mL flow injection loop for sample introduction into a pneumatic nebulizer. A l l metals were measured in pulse counting mode (with Co and In as internal standards for low and high masses respectively). A l l other metal analyses were performed on the Element2™ HP-ICP-MS (ThermoFirmigan Elements™ High Performance ICP-MS , Finnigan-MAT, Mississauga, Canada) fitted with a 24-position autosampler. A l l metals were analyzed in medium resolution with Rh as the internal standard for particulate metal samples. Isotope dilution was used for the analyses of dissolved metals in water. The enriched isotope of each analyte served as the ideal internal standard. Operating parameters for both instruments are given in Appendix B. Calibrations were performed internally using certified single-element standards (Can Lab VWR, Vancouver, BC). Calibration standards and corresponding samples were diluted in either 1 N or 1% Seastar H N O 3 . Sn as well as C d " 1 113 to Cd ratios were continuously monitored in samples so that potential isobaric 30 interferences with Cd could be detected and corrected for through the acquisition of interference free isotopic intensities. Certified reference materials (purchased from the National Research Council of Canada or NRC) were used to verify accuracy of sample preparation and analyses. 10% of all samples were run in duplicate with percentage relative standard deviation (%RSD) generally <5%. 2.2.3. Diffusion Gradient in Thin Film (DGT) 2.2.3.1. Background Davison and Zhang (1994) developed a technique that simulates diffusion of the labile metal fraction from an aqueous solution through biological membranes. It uses an ion-exchange resin backed to an ion-permeable gel membrane. It is represented as a diffusive gradient in thin films (DGT) that is described using the properties of the thin film model (Fig. 2.1). Transport through the gel membrane is strictly controlled by molecular diffusion and restricted to small ions or molecules, which can pass through the >5 nm pores (Zhang & Davison, 1999) of the polyacrylamide gel. The resin backing is composed of a chelating agent designed to bind free metal ions and weakly complexed metals able to diffuse through the gel layer. The concentration of the labile metal can be calculated depending on the time that the device is in contact with the bulk solution. In addition to providing a measure of metal lability, the nature of the chelating resin provides a concentrating effect and separates the metal ions of interest from the interfering salt water matrix, eliminating the need for an additional procedure. Zhang and Davison (1995) calculated a concentration enhancement of 864 (resin concentration/bulk concentration) after 24 hours using a typical diffusion coefficient for water of 10"5 cm2/s, a gel thickness of 1 mm and resin thickness of 0.1 mm. Ion diffusion across the gel is governed by Fick's first law: Ag + S where F = flux of metal ion, D = diffusion coefficient of metal ion, Q, = bulk concentration of free metal ion, Cr - concentration of free metal ion in resin gel layer, Ag = thickness of 31 diffusive gel layer, and 8 = diffusion boundary layer thickness. Zhang et al. (1995) have demonstrated that 6 can be neglected for natural waters where water movement tends to be above a minimum threshold velocity, and that Cr is effectively zero below the resin's saturation level. Eq. (1) can be simplified to: F=°± (2, and since flux through the diffusive gel can be calculated as the mass of the metal ion (M) that diffuses through the gel exposure area (A), after a given deployment time (t) or: Mean be calculated by rearranging Eqs (2) and (3): , . DCJA M = B— (4) The mass of the ion diffusing during period t (MM) can then be determined by analyzing a known volume of acid used to extract the metal ions from the resin gel: M M = C A V " ^ + V - L ) (5) where Ce = measured concentration in acid extract solution, VGE/ = volume of gel, VHNO3 = volume of added acid, and^ = extraction efficiency of acid for the metal ion which is typically 0.8 (Alfaro-Del la Torre et al, 2000; Warnken et al, 2004; Zhang & Davison, 1995). The measured mass (MM) can then be used to calculate the concentration of the metal ion in the original solution (CDGT)- Rearranging Eq. (4) and substituting MM (measured metal ion mass) for M (mass of metal ion) and CDGT (calculated concentration of metal ion in bulk solution through DGT) for Cb (concentration of metal ion in bulk solution) gives: CDGT = (6) DGI DtA Subsequent calculations using Eqns. 4-6 use the diffusion coefficient for each metal at 25 °C as given by L i and Gregory (1974) and included in Appendix A (Table A4) after correction for temperature (described later). 32 2.2.3.2. Preparation and Performance Testing 2.2.3.2.1. Gels Gel preparation followed a slightly modified version of the method described in Zhang and Davison (1995) and was conducted under a laminar flow hood within the cleanroom facilities of Lorax Environmental Services Ltd. in Vancouver, BC. Gels were prepared by initially mixing 57.5 mL 1.5 M Tris (tris(hydroxylemthyl)aminomethane), 85 mL 40% acrylamide/BIS (N,N'-methylene-bis-acrylamide, crosslinker, mixed in a ratio of 37.5:1) and 82.5 mL nanopure water. In the following order, 2.25 mL 10% sodium dodecylsulphate, 1.13 mL 10% ammoniumpersulphate (initiator) and 112.5 pL T E M E D (N,N,N',N'-tetramethylethylenediamine, catalyst) were added and the solution was well mixed. The gels were cast by pipetting the gel solution between glass plates separated by acid-cleaned clear plastic spacers in a Protean II™ xi multi-gel casting chamber (BioRad, Mississauga, ON). The gels were allowed to set for a minimum of 2-4 hours after which the glass plates were separated and the gels cut using a plastic disc cutter. The gel discs were gently removed from the plates and soaked in nanopure water for at least 24 hours with periodic water changes. The gels were then cleaned by soaking in 1 M ultrapure H N 0 3 (Seastar) overnight, rinsing several times with nanopure water, buffered to a pH of 6.5 using ultrapure ammonia and a solution of 2 M ammonium acetate buffer (using ultrapure ammonia and acetic acid, Seastar), and rinsed again several times with nanopure water. The gels were stored in nanopure water. The fully hydrated diffusive gel thicknesses were consistent between batches and measured to be 0.65±0.02 cm using a digitizing microscope. The gel expansion factor (volume of hydrated gel: volume of initially prepared gel) was estimated to be between 1.7 and 2.1 which indicates adequate proportion of free water for unrestricted diffusion of metal ions and small complexes (~5 nm) (Zhang & Davison, 1999). 2.2.3.2.2. Resin in Gels The resin layers were prepared as above with the addition of 89 g of Chelex-100 (100-200 mesh, BioRad) to 137 mL of gel solution prior to gel casting. Cutting, cleaning, 33 and buffering were performed as described for the diffusive gels above. The resin layers were stored in nanopure water until use. The piston-style plastic DGT deployment units with diffusion window of 2 cm were purchased from DGT Research Ltd. (Lancaster, U.K.) and have been described elsewhere (Zhang and Davison, 1995). Assembly and disassembly of the DGT units was performed under a laminar flow hood situated within a dust-free enclosure of a Chemistry Department laboratory at the University of British Columbia (Vancouver, BC). The resin layer was placed on the moulding base followed by the diffusive gel layer. An acid-cleaned 0.45 urn polyethersulfone membrane was placed over the diffusive gel to prevent fouling before the unit cap was snapped into place. 2.2.3.2.3. DGT Performance The performance of the DGT units was assessed with controlled laboratory tests. Acid-clean plastic containers were used to hold the test solutions. For each test, five or six containers with varying concentrations of a mixed metal standard in 2 L of 0.01 M N a N 0 3 or nanopure water were used. The solutions were stirred gently and continuously using Teflon coated magnetic stirrers. Sets of five DGT units were immersed and suspended vertically within each solution for a period of 24 hours. This exposure time, which was also selected for the field units, allows for the accumulation of detectable metal concentrations within the resin while preventing saturation. The DGT units were removed from the solutions, rinsed with nanopure water, and disassembled. The resin layer of each unit was transferred to a pre-weighed acid-clean polypropylene microtube, and soaked for a minimum of 12 hours in 1 mL of 1 M high purity nitric acid (Seastar). The metals in the eluent acid were then analyzed by ICP-MS, and these concentrations were converted into labile metal concentrations using equations (5) and (6) given above. For Cu, N i , and Cd, there was good agreement with expected concentrations, with average recoveries of 89%, 102%, and 92%, respectively, between DGT-labile and dissolved metal concentrations obtained through direct ICP-MS analyses of the prepared water solutions. Poor recovery was obtained for Pb during calibration experiments (52%) possibly due to colloid formation and container wall 34 adsorption (Warnken et al, 2004). Problems were also encountered with Zn, with inconsistent results during calibration experiments and high field variation (5-175%). Consequently, both DGT-labile Pb and Zn were eliminated from any further reporting or evaluation of results. Blank concentrations of Cu, N i , and Cd from field control DGT units (processed as the field units) and detection limits are reported in Table 2.3. 2.2.3.2.4. Time Dependence DGT units were deployed in a solution of 24 ug/L Cu, 1 ug/L Cd, and 12 ug/L N i for different time periods up to 48 hours. Similar to results presented by Zhang and Davison (1995), measured metal mass in the resin layer increased linearly with time (Fig. 2.2). The experimental data also agreed with the theoretical prediction of metal mass (M) using Eq. (4). 2.2.3.2.5. Ionic Strength and Temperature Zhang and Davison (1995) demonstrated that concentrations measured by their DGT devices lead to accurately predicted bulk concentrations of Cd when ionic strength (I) was in the range of 10 nM-1 M NaN03. This concurs with earlier findings by L i and Gregory (1974) that the diffusion coefficient decreases by at most 8% in seawater (I = 0.7) compared to distilled water (I = 0). Alfaro-De la Torre et al. (2000) reported significant increases in the diffusivity coefficient for Cd and N i below ionic strengths of 5 x 10"4 M due to electrostatic interactions and co-diffusion resulting from a negative N a + gradient from the chelex resin across the diffusional gel at these low ionic strengths. Other researchers using bis-acrylamide and agarose cross-linked gels (Na + and N H 4 + form of chelex) demonstrated the occurrence of non-selective weak binding of C d 2 + and C u 2 + within the diffusional gel at low ionic strengths that, at higher ionic strengths (>1 mM), becomes negligible. This is due to the availability of excess cations in the bulk solution which can then dominate binding (Peters et al, 2003; Sangi et al., 2002). To assess the effect of ionic strength, DGT units were deployed in solutions with varying concentrations of a mixed metal standard (Cu, N i , Cd, Zn, Pb, Mn) in 0.01 M NaN03 or deionized water. Two sets of N a N 0 3 solutions were 35 used to test for temperature effects (12 and 20 °C). Diffusivity coefficients for metal ions in water at different temperatures were calculated as previously reported (Zhang & Davison, 1995) by combining the Stoke-Einstein equation (Simpson & Carr, 1958) which uses the temperature dependence of diffusion coefficients (D) to the viscosity of water (rj), and the absolute temperature (7): Dr] JT\ Dr] \ T JT2 (7) and the well established temperature (t) dependence of water viscosity (Atkins, 1982; Dorsey, 1940): , ^25 1.37023(^-25) + 8.36xl0- 4 (^-25) 2 l o g - — = — (8) nt 109 + t Combining Eqs. (7) and (8) and knowing D25, the diffusion coefficient of ions in water at 25 °C (Li & Gregory, 1974) allows calculation of Dt at any temperature : , _ 1.37023(^-25) + 8.36xl0- 4 (^-25) 2 , ZX, (273 + 0 ,™ logD, = —+ l o g ^ - ! : (9) 109 + r 298 The temperature corrected diffusion coefficients for each metal were used in subsequent calculations of metal ion concentrations in solution using DGT. Fig. 2.3 reveals good predictability of Eq. (4) for solutions with ionic strength of 0.01 M at two different temperatures (12°C and 20°C). In deionized water, diffusion through the diffusional gel was enhanced at low bulk metal concentrations but approached theoretical diffusion coefficients at higher bulk solution metal concentrations (Fig. 2.4 and Table 2.4). This agrees with other reported findings (Alfaro-Del la Torre et al, 2000; Sangi et al, 2002) where single metal testing occurred. Precision of measured metal mass in the resin layer was poor at low bulk solution concentrations but greatly improved at higher concentrations (Fig. 2.5). While this general trend has been reported elsewhere (Peters et al, 2003), in these experiments improved predictability accompanied increased total trace metal concentration in the bulk solution (0-3 x 10"3 M) in addition to increased ionic strength (0 and 0.01 M ) 36 using NaN0 3 . Theoretical diffusivity coefficients for water were used to calculate C D GT for all DGT-labile samples given that the mixed metal concentrations used in the tests adequately approximated minimal stormwater trace metal concentrations within this study. Although diffusion coefficients can be influenced by increasing ionic strength, adjustments for increased salinity were not made. Adjustments would not have produced major changes with the potential of only small differences in the diffusion coefficients between deionized water and seawater (up to 8%) as reported by L i and Gregory and the large field variability in salinity (sometimes over small time scales and within the 24 hour sampling periods). Blueridge chamber/outfall samples, while consisting of fresh water and low trace metal concentrations, were not adjusted since concentrations were consistent with expectations. 2.2.4. Determination of Dissolved Metals Water samples were filtered within a few hours of collection using acid-clean polypropylene/polyethylene syringes and syringe filters of mixed cellulose esters (0.45 um pore size). Samples were immediately put on ice until filtration could be performed. Syringe filters were acid cleaned by slow passage of 20 mL of 2 M Seastar HNO3 followed by 50 mL nanopure water and then conditioned with 15 mL of sample before the filtrate was collected for metal analyses. Once filtered, samples were acidified with ultrapure nitric acid to a pH < 2 (0.2% v/v 70% nitric acid) and stored at 4°C. Metals were determined using isotope dilution mass spectrometry which required the quantitative addition of an enriched isotope for each element of interest. A chelex batch technique (involving direct addition of chelex beads to samples) was used to separate metals of interest from the salt-water matrix and interfering cations. Isotopic spikes were added to each water sample in order to produce the desired target ratios of an element's isotopes (Table 2.5). Target ratios were selected by balancing the need to minimize uncertainty introduced by propagation of error through the isotope dilution equation (described below) and the need to minimize influence of isobaric interferences that can occur at the enriched isotope mass. Since a ratio close to the geometric mean of the IS1/IS2 ratios of the spike and sample (where IS 1 and IS2 refer to the two element isotopes and the spike is enriched in ISi) 37 minimizes the former, while increasing the IS1/IS2 spike ratio decreases the correction-related error for the latter (Wu & Boyle, 1997), target ratios were selected to be just higher than the geometric means of IS1/IS2. In order to obtain target ratios of an element's isotopes, the exact metal concentrations of samples must be known before the isotope dilution technique can be successfully utilized. Since sample concentrations, as indicated by DGT data, exhibited a large range of concentrations for all metals of interest, laboratory trials were used to assess if one enriched isotope spike addition per analyte could be used to adequately quantify all samples. Two different mixed metal concentrations of samples were assumed for the preparation of two stock spike standards (IDMS1 and IDMS2) using the previously determined target ratios for calculations (Table 2.6). The exact spike standard concentrations were determined using a reverse isotope dilution technique with standard mixed metal solutions of natural isotope abundances. Each spike standard was added to a set of 1 acidified blank (nanopure water, 0.2% Seastar HNO3, v/v) and 4 seawater reference samples (SLEW-2, NRC) with varying additions of stock metal standards representing the anticipated concentration range of field samples (all in triplicate) (Table 2.6). Acidified samples (3 mL) in 4 mL conical-bottom polypropylene tubes were spiked with enriched isotope N i , Cu, 6 7 Zn , 1 u C d , and 2 0 6 Pb spikes (100 pL) and allowed to equilibrate for 24 hours. Samples were adjusted to pH 5.7 (low chelation efficiency for Ca and Mg) using aqueous N H 3 (10-25 pL) and N H 4 O A c (60 uL). The required volumes of NH3 were previously empirically determined on sacrificial samples. The tubes were capped after the addition of 50 pi of an acid-clean chelex slurry at pH 5.7 to each sample. The samples were shaken occasionally throughout a 24-hr period, after which the supernatant was discarded. The remaining chelex beads were rinsed twice with nanopure water and then allowed to sit in 1 mL of 1 M Seastar HNO3 overnight for metal extraction. Appropriate volumes of the acid extract were then diluted into 1% Seastar H N 0 3 for ICP-MS analysis. Analyte concentrations of each sample were calculated by the Element2 HP-ICP-MS Software and based on the following expression: 38 where C is the analyte concentration (ug/g), Wis the sample weight (g), Mspike is the mass of the stable isotope spike, A is the spike or natural (nat) abundance of either the spike (IS1) or reference (IS2) isotope, K is the ratio of natural and spike atomic weights, and R is the measured isotope ratio of reference isotope to spike isotope after spike addition. The measured ratio (R) was corrected for instrumental mass bias by comparing to a solution of well-known isotopic composition (natural element standard) prior to each session of the ICP-MS. Any mass offsets were then incorporated into all instrument ratio determinations. Typical mass offsets were < 0.001% for Cu, N i , Cd and Zn, and < 0.005% for Pb. Analysis efficiency was often poor and inconsistent for IDMS1; but all elements were fairly accurately determined for the IDMS2 trial (89-109%) at almost all sample concentrations tested (Table 2.7). The only exceptions were for very low concentrations of Zn, Cd and Pb. The poor precision and relative accuracy of very low metal concentrations was considered acceptable given the relatively high metal concentrations within the field site waters and the nature of this study. Consequently, the IDMS2 enriched mixed spike standard was used as the basis for all field sample isotope spiking. The general procedure used was as described above with 2 blanks, 1 estuarine water reference material (SLEW-2), and 2 spiked estuarine water reference material samples (SLEW-SP1 and SLEW-SP4) added to each set of approximately 20 samples. Additionally, 10% of samples were performed in replicate. Recoveries for standard reference material (original and spiked) were similar to trial findings with low precision and decreased accuracy associated with very low concentrations of Zn, Cd, and Pb (Table 2.8). This was deemed acceptable for the purposes of this study. Method replicate analysis revealed 50-77% of samples below a RSD of 3%; and 80-90% of samples below a RSD of 10% for N i , Cu, Zn and Cd. Higher RSDs were associated with low concentrations near detection. Pb had 42% of samples below RSDs of 10% and 50% of samples with RSDs above 20% (also primarily associated with concentrations near the detection limit). 39 2.2.5. Determination of Particulate Metals Water samples taken during September 28-30, 2000 and May 22, 2001, were analyzed for particulate metals. Particulate matter was collected by filtration of 250 mL water samples through acid-cleaned polycarbonate membranes (Poretics Corp., Livermore, CA, diameter 47 mm, pore size 0.4 urn). The filters were individually placed in polyethylene microtubes, and dried within a wrapped plastic container in a convection oven at 60°C. The plastic enclosure was open to the atmosphere through plastic tubing fitted with a high density polyethylene frit for air exchange to minimize contamination by airborne particulates. Once dry, the filters were placed in 10 mL Teflon tubes and dissolved in 1 mL Seastar ammonia overnight. The ammonia was evaporated to dryness on a hotplate within a laminar flow fumehood prior to microwave digestion (Milestone Microwave M L S 1200 Mega, Milestone, U.S.A.) with 0.8 mL H N 0 3 , 0.1 mL HF, and 0.2 mL HC1. Digestions were performed in a carousel of twelve PTFE pressure bombs accommodating 3 vials per bomb. Two standard reference material standards and two blanks were included with each microwave carousel of 32 samples. A timed microwave program of 250W for 12 min, 450W for 12 min, 650W for 24 min, and 250W for 12 min was utilized. Samples were allowed to vent and cool before removing vials and evaporating to dryness. The samples were then reconstituted with 1 mL of 1 N Seastar HNO3 before preparation for ICP-MS analyses. Teflon tubes were cleaned between runs by soaking in soap solution overnight, rinsing twice with distilled water and then soaked for 24 hours in a 4 N HNO3 bath. Tubes were rinsed twice with distilled water and then twice with distilled, deionized water before microwaving with 1 mL EG HNO3 through a cleaning program (250W for 10 min, 650W for 10 min, and 250W for 10 min). The acid was discarded and the tubes were allowed to dry overnight under a trace-clean laminar flow fumehood. Analysis of the marine sediment reference material (PACS-2) agreed well with certified values (Table 2.9). 2.3. Results Weather condition categories are defined in Table 2.2 and described above. Water results are summarized in Tables 2.10 and 2.11 for the Philip (chamber, inner basin, outer 40 basin) and Blueridge (chamber, receiving basin) sites according to weather condition categories. Metals were measured at the Lighthouse reference site during May 21-22, 2001 and in general, dissolved and DGT metal levels were comparable to dry-moderate storm levels within the Blueridge receiving basin and lower than levels at the Philip site. Chamber inlet concentrations were obtained from two consecutive storms at Philip in November 1999 and during the storm period of September 29-30, 2000. Chamber inlet samples were obtained through the manhole immediately preceding the chamber. On September 29, 2000, stormwater samples were also taken from a local drainage pipe on the east bank of the receiving creek (seaward of station A l , Fig. 1.2). This sampling point is directly adjacent to an autobody shop, a cement/tile facility and heavily used railroads, and reflects local metal contamination sources. The land uses in the immediate vicinity, however, are not unique within this catchment area. Fig. 2.6 to Fig. 2.17 represent dissolved and DGT-labile metals, as well as salinity and precipitation data for each of the sampling periods. 2.3.1. Temperature and Salinity Water temperatures observed within this study, which represent shallow landward conditions, were highest in summer and early fall (14-19°C), lowest in late fall and winter (6-11°C) and then increased in spring (9-14°C) (Table 2.12). Storm-induced temperature drops were observed across seasons. Salinity within the Philip inner basin ranged from 3.6-16%o during dry periods of sampled seasons with a fall/summer range of 9.8-16.1 %o and a lower spring range of 3.6-11.3%o with probable influence from regular rainfall occurring during this season. For example, from May 21 s t to May 22 n d in 2001, salinity increased at each of the two sampling sites in addition to the coastal reference site (LH) where the preceding rain event of 74 mm total had occurred during March 13-16th (Fig. 2.9). Chamber salinities were comparable to inner basin levels across seasons. Light storms produced salinities as low as 2.5%o during spring in chamber and inner basin, a range of 2.5-6.2%o during ongoing storms and 5.5-9%o (Oct 4/98, Jun 4/99, Sep 8/00, Fig. 2.7B, Fig. 2.1 IB and Fig. 2.14B) during sampling dates 41 where rain had ended several hours prior to measurement. Moderate and severe storms resulted in salinity levels ranging from 0.03-3.2%o during storms and 9.3 and 13.4%o for the two post-storm measurements (Nov 15/98 and Sep 30/00, Fig. 2.8B and Fig. 2.16B). Philip outer basin salinities ranged from 17.3-28.2%o with the exception of a few low salinities occurring in spring (9.7%o, May 21/01, Fig. 2.17B) and severe storms in fall (10.6%o, Nov 14/98 and ll . l-1596o, Oct 2-3/98, Fig. 2.8B and Fig. 2.7B). Blueridge receiving basin salinities were similar to Philip outer basin over the two dry weather sampled periods in spring and early fall. The salinity range showed slight variation during a moderate storm event in early fall 2000 (17-24.8%o, Fig. 2.14B) but was notably reduced in the middle of the most severe storm sampled in mid fall 1998 (0.03-3. l%o, Fig. 2.8B). 2.3.2. pH pH in sampled waters was generally decreased at low salinities as in stormwaters and rose with increases in salinity. pH tended to be lower in chambers (5.2-6.5) than inner basin (Philip, 5.3-8.0) or receiving basin (Blueridge, 7.0-7.6) at both sites and was similar within the two site chambers (Table 2.12 and Appendix A , Table A3). Slightly acidic conditions occurred with stormwater at both sites (pH of 5.2-6.5). These low pH levels were detected within the Philip chamber and inner basin during storms but recovered to higher levels (6-8) with increases in salinity during dry weather. Blueridge receiving basin pH levels were lowest on November 14 th, 1998 (6.5) but recovered to pH 7 on the next day with the end of the storm. The Philip outer basin pH ranges were higher than inner basin levels but slightly lower than Blueridge receiving basin levels. Both sites showed some degree of seasonal and storm activity influence on pH levels. Within the Philip outer basin and Blueridge receiving basin stations, storms were often accompanied by slight pH drops and spring pH levels (6.9-7.1) were lower than summer/fall levels (7.1-7.8) during dry conditions. 42 2.3.3. Dissolved and DGT-Labile Metals 2.3.3.1. Copper Dry weather sampling across stations and three seasons revealed copper concentrations that were relatively low for both dissolved (0.42-12.9 pg/L) and DGT-labile (BD-7.4 pg/L) at both sites (Table 2.10). Consistently lower levels were found in the outer basin relative to the inner basin and chamber at Philip (approximately 50% decrease). Relative to dry weather sampling, on most occasions light storm sampling revealed slightly elevated dissolved and DGT-labile levels (50-400%) increase) within the Philip chamber and inner basin. A l l three moderate storms, sampled at the Philip site chamber and inner basin, produced dissolved concentrations that were 10-20 fold increases relative to dry weather concentrations and 10-100 fold increases for DGT-labile concentrations. The highest dissolved copper level of 364 and 315 pg/L for chamber and inner basin respectively were measured during one severe storm (September 29, 2000, Fig. 2.16C) at Philip and was 20-50 times greater than previous dry weather concentrations. The subsequent moderate storm was accompanied by declines in dissolved levels (191 and 116 pg/L) within the chamber and inner basin. DGT water sampling during two severe storms enabled the determination of storm DGT-labile metal concentrations. The November 14-15th, 1998 event (Fig. 2.8C) resulted in concentrations of DGT-Cu that increased by over 3000% in both chamber and inner basin while the October 2-3 r d, 1998 event (Fig. 2.7C) resulted in a 200% increase in the inner basin and a 1000% increase in the chamber. The light storm immediately following (October 3-4, 1998) resulted in 25-40 fold elevations over previous dry weather concentrations. The highest DGT-labile Cu concentration levels were observed for the moderate storm sampled from September 29-30, 2000 at 579 and 338 pg/L within the chamber and inner basin respectively (Fig. 2.16C). Differences between the Philip chamber and adjacent inner basin station (station A l , Fig. 1.2) were small in dry weather sampling. The trend was inconsistent and a pattern was difficult to discern based on the water data collected within the present study. Differences in dissolved copper became more distinct with storm sampling. During storm events and with the onset of stormwater flow that resulted in lower salinities, dissolved concentrations were higher in the chamber relative to the inner basin (18-69%). Once salinities rose following 43 rainfall cessation, chamber concentration levels fell relative to the inner basin (e.g. Nov. 12-13 and Nov. 21-22 1999, Sep. 7-8 and Sep. 29-30, 2000 storms, Figs. 2.12-2.16). The same trend was reflected in the DGT concentrations of October 1998 (Fig. 2.7). In general, DGT-labile concentration changes were less consistent relative to dissolved concentrations which reflects the dynamic condition of the chamber and inner basin waters as captured by the time-averaged nature of DGT measurements. The Philip outer basin dissolved and DGT-labile Cu concentrations remained relatively unchanged between dry weather and light storms (BD-7.72 ug/L), however modest elevations up to 55.8. ug/L dissolved copper and 15.4 ug/L DGT-labile copper were observed with moderate and severe storm sampling (Fig. 2.16 and Appendix A). Three out of four moderate and severe storms had 10-20 fold increases for dissolved copper while DGT-labile copper was only elevated by a possible 2-fold increase for the severe storm in November 1998 (Fig. 2.8). Overall, similar to dry weather sampling, storm sampling revealed lower concentrations in the outer basin relative to the inner basin and chamber (2-12 fold drops for dissolved concentrations and 10-40 fold drops for DGT-labile concentrations). The Blueridge site chamber outfall and receiving basin generally had lower concentrations relative to Philip. Dry weather dissolved Cu concentrations (measured for five sampling dates consisting of two sampling periods) in the chamber were 50% lower than Philip chamber concentrations (Fig. 2.16 and Fig. 2.17). DGT-labile concentrations (measured for four sampling dates) were similar between site chambers for September 27-28, 2000 (Fig. 2.16) and 30% lower for the Blueridge chamber outfall for May 21-22, 2001 (Fig. 2.17). Blueridge stations were not sampled during light storms but were sampled during one moderate storm in 2000 and one severe storm (DGT only) in 1998. The receiving basin concentrations for the moderate storm were approximately 60-75% lower than chamber outfall concentrations during dry weather. The 1998 severe storm produced an order of magnitude elevation in DGT-labile Cu concentrations (Fig. 2.8). Dissolved Cu concentrations were 87% and 75% lower during the moderate and severe storms of 2000 respectively. DGT-labile concentrations were 1.0-5.2 ug/L and BD- 1.3 ug/L for the chamber outfall and receiving basin during dry weather and moderate storm sampling but reached 42.2 ug/L within the receiving basin during the severe storm in November 1998. 44 In general, replicate field samples had coefficients of variation (CVs) less than 14% for dissolved copper except for two concentrations above 150 pg/L with CVs of 40% and 20%. DGT-labile concentrations above 5 pg/L had CVs typically below 15%. As expected, decreasing concentrations coincided with increasing CVs. Sampling during dry periods revealed DGT to dissolved concentration ratios in the range of 0.2-0.8 in the Philip chamber, 0.2-0.6 in the inner basin station A l , 0.2-0.7 for the inner basin station average, and 0.3-0.4 in the outer basin. Within both basins, this ratio is greater in spring when compared to fall and summer and lowest in the dry summer of August 1998 within the chamber. Blueridge ratios were 0.9-1.1 in the chamber, and BD and 1.1 in the receiving basin. It should be noted that the receiving basin DGT-labile concentrations for fall of 2000 were below detection and that this chamber consists of freshwater with no tidal influence. During storm events, the dissolved copper measurements reflect one time point whereas the DGT-labile copper concentrations provide a time-averaged measure of copper concentration fluctuations during the DGT deployment period. Direct comparison of these two measurement techniques is thus difficult especially since one presumes water concentrations remain constant during the sampling periods. 2.3.3.2. Nickel Relative to Cu within the Philip chamber and inner basin, nickel displayed a narrower range of dissolved (0.31-8.60 pg/L) and DGT-labile (BD-10.6 pg/L) concentrations across weather categories (Table 2.10). Concentration ranges overlap between different weather categories and seem to obscure trends that may be present for both dissolved and DGT-labile concentrations. By examining sampling periods individually, clearer patterns emerge. During the September 18-19, 2000 sampling period, Philip chamber concentrations were elevated and comparable to storm levels (7.21 and 8.60 pg/L dissolved; 8.2 pg/L DGT-labile). Inner basin concentrations were similar to other dry sampling dates. Aside from September 18-19, 2000, all dry period concentrations within the Philip chamber fell below 3 pg/L (dissolved) and 4 pg/L (DGT-labile). Elevations in dissolved concentrations with storm activity were observed during one of six light storm days (November 21, 1999, 4.23 and 4.19 45 pg/L) and for all four moderate/severe storm days within the chamber and inner basin. DGT-labile concentrations were elevated for one of four light storms (November 12-13, 1999; 8.2 and 4.4 pg/L) and for all three moderate storms measured (November 21-22, 1999, September 7-8 and 29-30, 2000). The two severe storms sampled by DGT revealed concentrations within the chamber and inner basin that were similar to or lower than concentrations occurring during dry periods. An examination of the two multi-day sampling periods demonstrated similar yet slight increases for the DGT-labile storm concentration levels relative to the preceding dry period concentrations for the October 1998 sampling; and similar dissolved concentrations for the two storm days (September 29 and 30 t h, 2000) within the chamber and inner basin station A l but slightly elevated concentrations on the second lighter storm day within the remaining length of the inner basin. Similar to Cu, N i differences between the Philip chamber and station A l were not consistent during dry weather sampling. In contrast to Cu patterns, changes in dissolved N i did not correlate with salinity levels. Within the spring light storm salinity range of 2.5 to 9%o, all four dates produced dissolved concentrations greater within the chamber than at station A l . Of the two moderate and one severe storms with extremely low salinities, chamber dissolved concentrations were greater than at station A l during the severe storm on September 29, 2000 only. The other two storms produced no difference in concentrations between these stations on September 7, 2000 and November 22, 1999. Philip outer basin dissolved (0.38-3.12 pg/L) and DGT-labile (BD-3.2 pg/L) concentrations were generally lower than inner basin concentration levels across seasons and weather categories. Blueridge chamber outfall concentrations (0.50-1.78 pg/L, dissolved; BD-1.2 pg/L, DGT-labile) were similar to receiving basin concentrations (0.36-0.74 pg/L, dissolved; BD-3.4 pg/L, DGT-labile) and both were consistently lower than concentrations at the Philip site. Replicate field samples had CVs less than about 20% for dissolved and DGT-labile concentrations above 4 pg/L and 3 pg/L respectively and below the upper ranges of the sample concentrations. Sampling during dry periods produced DGT:dissolved concentration ratios of 0.8-2.5 in the Philip chamber, 0.8-1.4 and 15 (May 21-22, 2001) for station A l , 0.7-46 1.2 and 11 (May 21-22, 2001) for inner basin averages, and BD, 0.8-1 and 6.4 (May 21-22, 2001) for outer basin averages. 2.3.3.3. Cadmium Dissolved (0.266-2.30 ug/L) and DGT-labile (0.35-1.85 ug/L) cadmium concentration ranges within the Philip chamber and inner basin were similar between dry and light storm periods (Table 2.10). Elevated concentrations within the chamber and inner basin during these relatively calm periods suggest that Cd may be released from sediments into overlying waters. Increased dissolved levels (5.12-7.36 ug/L) were seen with two of the three moderate storms (November 22, 1999 and September 30, 2000) and the preceding fall severe storm (September 29, 2000). Of the five DGT measured moderate/severe storms, only the two moderate storms resulted in elevated concentrations (2.57 and 7.10 ug/L, November 22, 1999 and September 30, 2000). Of the two severe storms, DGT concentrations were only slightly elevated for the October 3, 1998 event (1.69 and 2.43 ug/L, chamber and inner basin) relative to the preceding dry period (0.46 and 0.49 ug/L). Unlike the Cu trend for this sampling period, Cd concentration levels declined with the subsequent light storm (0.97 and 0.99 ug/L) but were still elevated (2-fold increase) with respect to the initial dry period. The September 28-30, 2000 sampling period revealed dissolved levels that were highest on the first severe storm day within the chamber and inner basin station A l but higher on the following moderate storm day within the remaining length of the inner basin. Chamber concentration levels were often higher than the adjacent inner basin station (Al ) during dry weather periods. A similar trend was apparent in samples during the moderate/severe storm periods. Sampling during light storms revealed relatively higher DGT concentrations within the chamber station for three of five storm periods where DGT sampling occurred. Dissolved Cd was higher in the chamber for two of five light storms and higher at station A l for one of five light storms. Philip outer basin dissolved and DGT concentrations were generally lower than chamber and inner basin concentrations and comparable to inner basin dry period concentrations. Blueridge chamber outfall concentrations (BD-0.138 ug/L, dissolved; 0.09-47 0.30 ug/L, DGT-labile) were slightly elevated relative to the downstream receiving basin concentrations (BD-0.059 ug/L, dissolved; BD-0.21 ug/L, DGT-labile) (September 28-30, 2000 and May 2001). However, concentrations at Blueridge were always lower than at Philip. CVs were generally below 15% for both dissolved and DGT-labile concentrations but increased up to 35% for DGT levels below 0.5 ug/L and dissolved levels greater than 4 ug/L. DGT:dissolved concentration ratios for dry weather sampling were 0.5-0.9 in the Philip chamber, 0.4-0.7 at inner basin station A l , 0.4-1.4 for the inner basin average, BD-0.6 for the outer basin, and 1:1 for the two sampling dates of the Blueridge chamber. Ratios could not be calculated for the Blueridge receiving basin where concentrations were below detection. Philip inner basin concentration ratios were highest in spring and lowest in fall and summer. 2.3.3.4. Zinc and Lead Due to contamination and detection issues for zinc and lead, only certain periods are reported for these two elements. Zinc is included where field blanks for a given sampling day produced consistent low metal concentrations and lead is included when at least one field station produced concentrations above detection. Dissolved zinc and lead are reported for November 1999, September 2000 and May 2001. Lead was additionally measured during the September 18-19 dry period. During three dry dates, Zn concentrations ranged between 66.4 and 222 ug/L within the Philip chamber and inner basin. The range of dissolved Zn concentrations for the four light storm periods (192-246 ug/L) was similar to the dry periods within the chamber and inner basin. The three moderate storms and one severe storm showed elevated concentrations of up to one order of magnitude relative to dry periods within the chamber (506-1260 ug/L) and inner basin (460-902 ug/L). Except for the September 7-8 storm period where concentrations were similar, concentrations within the chamber were higher than at the inner basin station A l (spring dry and fall storm). The Philip outer basin dry and storm period concentrations were comparable to those measured within the inner basin during dry periods. During September 29 and 30 t h severe and moderate storm days, the outer basin A4 station Zn concentrations were elevated at 405 and 48 365 \xgfL relative to dry periods but concentrations declined at the adjacent seaward A5 station (178 and 163 pg/L). Blueridge Zn concentration levels were higher within chamber outfall waters relative to the receiving basin (an order of magnitude) but both were considerably lower than Philip concentrations (1-2 orders of magnitude). Other than concentrations near the method detection limit, CVs were generally below 17%. Lead concentrations were below 2 pg/L within the Philip chamber and inner basin for dry days with the exception of the inner basin station A3 on September 28 t h . One light storm day (Nov. 13/99) produced Pb concentrations slightly higher within the chamber (2.7 pg/L) and at station A3 (3.2 pg/L). The moderate storms on September 7 and November 22 produced the highest dissolved lead concentrations within chamber and inner basin (6.3 and 8.0 pg/L). The September 29 t h severe storm resulted in a modest concentration elevation within the chamber (3.3 pg/L) but not throughout the inner basin. Similar to Pb, Zn concentrations at station A3 were slightly elevated (3.6 pg/L). The subsequent moderate September 30 t h storm revealed reduced Zn concentrations that were comparable to those measured during dry periods (<2 pg/L). Chamber concentrations were higher than inner basin station A l concentrations for all four moderate/severe storm days. Dry and light storm days revealed no detectable difference in Zn concentrations between these two stations. Philip outer basin concentration levels were below 2 pg/L for all dates and Blueridge chamber and receiving basin concentrations were close to or below detection. Similar to Zn, Pb CVs were generally below 20%. 2.3.3.5. Suspended Particulate Matter Suspended particulate matter (SPM) was sampled during the dry spring date of May 22, 2001 and prior to and during a major storm period from September 28-30 th, 2000. The suspended particulate matter concentrations, SPM-metal content, and water particulate metal concentrations are presented in Figs 2.18 to 2.23. Total SPM concentrations between sites showed small variation for the storm and post-storm (Sep. 29 and Sep. 30) samples in 2000. From the antecedent dry period (Sep. 28) to the period after the onset of the moderate/severe fall 2000 storm, total SPM within the Philip chamber increased by 27% and then dropped to 49 pre-storm levels after storm cessation. Three sampling points provide a coarse look at the fall 2000 storm-related changes in SPM and particulate copper concentrations. SPM concentrations are expected to vary over the course of a storm due to the delivery of the largest loads within the early runoff period (Morrison et al, 1988; Sansalone & Buchberger, 1997; Schueler, 1987). The first stormwater sampling was 17 hours after the storm onset coincident with the end of the severe rainfall intensity period and likely followed any peak SPM discharges. Within sites, concentrations at the outfall stations were lower within the chamber for all dates and similar between dates with the possible exception of Sep. 29 which showed an elevation of 16% compared to the preceding dry date. Concentrations at the Blueridge site were higher than at the Philip chamber and inner basin in May 2001; however, the highest concentrations were at the Philip outer basin which were also comparable to the Lighthouse Reference site (Fig. 2.18). Philip inner basin concentration averages and standard deviations were highest on Sep. 28 at the seaward station A3 (Fig. 1.2) where water level and measured flow were lowest (Table 2.13). The trend of total SPM concentrations for the fall storm sampling within the Philip inner basin was Sep.28 > Sep.30 > Sep.29 or pre-storm>post-storm>storm and the opposite or Sep.29 > Sep.30 > Sep.28 (storm>post-storm>pre-storm) for water flow velocity. While sampling was always conducted within about 2 hours of low tide, direction of tide flow differed for the consecutive September sampling dates as shown in Table 2.13. Only the pre-storm sampling occurred at the beginning of the incoming tide cycle. Flow, salinity, and tide direction (i.e. source input) may have influenced the observed SPM concentrations on this date. Particulate metal concentrations within Philip waters including the outer basin were elevated on September 29 t h with flow of storm runoff (one order of magnitude for all elements except Zn, Figs. 2.19 to 2.23). Post storm (September 30) SPM-Cu and Zn concentrations (dry weight) decreased but were still elevated relative to pre-storm concentrations (most markedly within the chamber). Post storm, the outer basin revealed a quicker return to pre-storm concentrations for both elements. SPM-metal concentrations showed a similar decline to particulate metal in water for Cu and Zn. Post storm SPM-metals were comparable (only slightly lower) than storm metal conentrations with the exception of the chamber and station A3 where post-storm > storm. Pb, Cd and N i particulate concentrations at Philip were similarly elevated on September 29 t h but reduced drastically by 50 September 30 . Concentrations at station A3 were elevated for SPM-Cu, Zn, Pb, and Cd relative to the other two inner basin stations, although pre- and post-storm Zn and Cd concentrations were similar and Pb concentrations were higher during the pre-storm period than during the post-storm period. Station A3 differs from the other two inner basin stations with respect to SPM-metal and particulate metals in water. A slight elevation in dissolved Pb was also noted during the light storm of Nov. 13/99. This station was the shallowest of the inner stations at the Philip site with water depth ranges of 5 to 12 cm during sampling periods. Aside from the possibility of sampling disturbances to the sediment surface, water samples from the shallow water depth at this station may provide a close reflection of sediment/water interface processes. Such processes include sediment resuspension and early diagenesis with resulting metal dissolution from both. A l l storm and post-storm particulate metals were higher within the Philip chamber and at the Blueridge chamber outfall compared to their respective receiving basin stations. The east bank drainage pipe which was discharging a small volume of stormwater directly into the creek also contributed elevated concentrations of total SPM and particulate metals relative to station A l of the inner basin (Table 2.14). These concentrations were comparable to those within chamber inlet stormwater which are reported in the next section. Particulate metal concentrations at Station A2 were consequently elevated relative to A l on September 29, indicating measurable influence from this direct source input. May 22, 2001 Philip metal concentrations were similar to September 28 t h 2000 with the exception of the outer basin (May>September) and station A3 (September>May). Blueridge receiving basin particulate-metal levels do not show much variation between sampling dates and are similar to Philip dry weather sampling dates and the May Lighthouse reference site. The highest Blueridge chamber outfall particulate-metal levels were only slightly (10-20%) lower than Philip chamber levels during the September 29 t h storm day, even when Philip site elevations reached almost 2 orders of magnitude. The post-storm (September 30) particulate metal levels for the Blueridge chamber outfall decreased but remained elevated relative to pre-storm levels for Cu, Pb, Cd, and Zn. 51 2.3.4. Chamber Impact on Stormwater Metals Stormwater metal concentrations at the Philip site were assessed with stormwater passage through the chamber system. In order to capture runoff before and after chamber exposure, chamber water was compared to chamber inlet and outlet waters. Three storms were evaluated for dissolved and DGT-labile metals (Table 2.15). The dissolved metal samples provide a short-term (1 hour) average of stormwater concentrations (1 sample every 20 minutes) whereas DGT metals provide an approximate 24 hour time averaged concentration. DGT units were deployed throughout the three storm events and all three had some extent of saline freshwater mixing. The November 12-13 light storm differed from the other two in that intrusion of saline receiving basin waters was detected at the chamber and outlet at the start and finish of the sampling period (salinity ~ 5%o) (Fig. 2.12B). Field observations and meteorological data indicate that this light storm was fairly consistent with respect to rainfall intensity. Dissolved metals from November 13 th represent the final stages of this storm while DGT metals provide a time average for the whole storm. The November 21 -22 DGT units were deployed in pre-storm saline conditions (20% of deployment time for chamber and outlet only), however, 80% of the DGT deployment period was within stormwaters. Similar to the previous storm, dissolved metals from November 22 samples represent the end of this storm and DGT samples represent the bulk of this moderate storm. The September 29-30 DGT deployment period reflected post-storm influences from the 14 hr event that ended at the time of sampling on the 29 t h, moderate storm conditions that occurred for 12 hours of the deployment period, and post-storm conditions that followed for the remaining DGT deployment period. Given the elevated DGT and dissolved metal concentrations within the saline waters on September 30 (Fig. 2.16), it seems unlikely that DGT metal concentrations were significantly underestimated for this combined storm event. For the purposes herein, they are assumed to provide an adequate representation of these combined events. The dissolved metals from September 29 stormwater samples represent the end of the initial severe event. A l l three storms revealed reductions in DGT-labile Cu after passage through the chamber. DGT-labile Cd was also reduced at the outlet for the moderate (Nov.22/99) and moderate/severe (Sep.29/00) storms. Although DGT measurements could be 52 underestimating storm metal concentrations during the November 22 event, given the pre-storm dissolved concentrations (Fig. 2.13C), this would at most lead to a decrease of ~11% in reported concentrations. Chamber processes are therefore likely influencing DGT-labile stormwater metal reduction for these two elements (Cu and Cd). DGT-labile N i was significantly elevated at the inlet during the light November 13 storm (41.3±13.6 ug/L) and largely reduced (by 84%) at the outlet. While the September 29 and November 13 storms both revealed elevated chamber inlet DGT-Ni concentrations relative to typical dry weather concentrations, only the noticeable elevation of the light storm samples was decreased at the outlet by 91%. Stormwater velocity at the time of sampling indicated that the fast movement of stormwaters on November 22 (20.5 cm/s) coincided with non-discernable changes in dissolved metal concentrations after passage through the chamber system. The slower movement of stormwaters on November 13 t h (2.8 cm/s) and September 29 t h (12.3 cm/s) were accompanied by changes in outlet concentrations for some metals but showed differences between these two sampling days. Dissolved N i was generally elevated in stormwater of the moderate and severe storms and did not display any reduction at the outlets. Concentrations were slightly elevated within the chambers for all three storms. Dissolved Cd remained low during the light storm, and was only reduced for the severe storm sampling. Only a slight reduction for dissolved Cu at the outlet occurred during the light storm. Changes for dissolved Zn within the chamber or at the outlet are not discernable within the range of sampling variation. Dissolved Pb was reduced at the chamber outlet for the severe storm and slightly enriched within the chamber for the slower moving stormwaters of the light and severe storms. Dissolved metal concentrations reflect the conditions at the time of sampling and, at least for Pb, may be directly related to stormwater velocity. In addition to dissolved and DGT-labile measurements reported in the previous section, SPM analyses was performed during the severe/moderate storms of September 28-30, 2000. Stormwater particulate metal concentrations are reported in Table 2.16. SPM collected from the chamber inlet was higher in Zn, Cd, and especially Pb relative to the chamber and chamber outlet at the Philip site. N i and Cu SPM content (mg metal/g dry SPM weight) was higher within the chamber than at the outlet. Total SPM concentrations were reduced from the inlet to the outlet (86.8 to 13.3 ug/L) with consequent reductions in all 53 particulate metal contributions to total stormwater metals (pg/L). Of course influence from reductions in SPM metal content for Cd, Zn, and Pb would have also occurred. 2.3.5. DGT-Lability The results of the mixing experiments, performed for the elucidation of dissolved and DGT metal behaviour (using stormwater and outer basin saline water samples from September 29 t h), are presented in Fig. 2.24. Dissolved and DGT-labile Cu, Cd, and N i are plotted against salinity. Dissolved and DGT-labile Cu both deviated from the ideal mixing line demonstrating non-conservative behaviour. Dissolved and DGT-labile N i behaved conservatively during mixing (i.e. adhered to the straight ideal mixing line). DGT-labile Cd, similar to DGT-Cu, appeared non-conservative; however, dissolved Cd appeared above the mixing line at least for salinities of 10 and 17 %o. Each of the three metals assessed demonstrated a slightly different distribution between dissolved and DGT-labile fractions (Fig. 2.25). The percentage of DGT-labile Cu quickly decreased to approximately 60% of total dissolved Cu by a salinity of 5 %o and then appeared to stabilize to 40% at higher salinities. DGT-labile and dissolved N i agreed well at lower salinities; however, the DGT to dissolved proportion progressively decreased at higher salinities (88 to 58%). DGT-Cd was similar to Cu, decreasing at low salinities (68% to 75%) but to a lesser degree. At higher salinities, DGT-Cd represented 36-42% of total dissolved Cd. 2.4. Discussion 2.4.1. Temperature and Salinity Extensive description of the region's oceanography was reviewed by Thomson (1981) and provides the basis for assessing salinity and temperature trends. In general, Burrard Inlet temperature and salinity vary seasonally and most noticeably within the top 5 m of the water column. Freshwater contributions come from various rivers, land seepage and sewer discharges and are closely tied to precipitation with greatest influence during the autumn to 54 winter rainy season. Temperatures of near surface waters are highest in late July to early August and can approach air temperatures when winds are light over shallow regions despite general cold surface temperatures due to intense tidal mixing within First and Second Narrows. Cold water temperatures prevail in winter with depth uniformity. Salinities are higher in winter with a small decrease in surface waters (25-30%o), and lower in summer (10-28 %o) with a more apparent depth gradient. During the summer high discharge period the freshwater plume of the Fraser River may also penetrate through First Narrows into the inner harbour (Thomson, 1981). The shallow landward properties of the studied area would lead to temperatures expected from previously reported patterns for Burrard Inlet; however, salinity patterns are confounded by proximity to direct storm-related freshwater inputs. 2.4.2. Stormwater: Dissolved metals and Regional Context The relative range of element concentrations during storm conditions (Zn>Cu»Ni>Pb>Cd) shows a slightly different trend than during dry weather periods (Zn»Cu>Ni>Cd>Pb) within the urban/industrial setting of the Philip Street site. Large contributions of Zn and Cu from stormwater discharges were associated with moderate and severe storms. N i , Cd, and Pb stormwater concentrations did not reach the same magnitude during this study although moderate increases were observed for some storms. While stormwater contaminant contributions within a given location are expected to vary with any number of factors, including land usage and intensity, storm intensity, duration, stormwater volume, and antecedent dry days, it was interesting to note that the two largest DGT-labile Cu levels in the immediate receiving waters at the Philip site were measured during two storms (Nov. 14-15, 1998 and Sep. 29-30, 2000) with the largest total storm volumes (Fig. 2.26). Average and initial flash storm intensities in addition to antecedent dry days did not vary consistently between these two storms; however, the two longest dry periods (40 and 18 days) preceeded the events on Sept. 8 and Sept. 30, 2000 with the largest dissolved and DGT-labile N i discharges respectively. The frequency and efficiency of chamber maintenance (sediment removal) will also affect potential discharge into receiving waters. District maintenance standards for the period prior to 1998 indicate chambers were checked and serviced every three months (Fedrigo and Boase, 1997). Maintenance records, while not 55 available for 1998-2001, indicated a more frequent cleaning regime (at least in fall and winter) commencing in late 2002 (Hudda-Musani, 2004). A perusal of the records indicated that despite the increased diligence and possible reduction of chamber pollutant load immediately after chamber cleaning, several significant storm events occurred periodically between maintenance activities and could have discharged previously collected chamber contaminants into receiving waters in spring, fall and winter. Cu and Zn, in addition to Pb have a history of significantly elevated storm-related concentration levels within the broader Greater Vancouver Regional District (GVRD) and elsewhere (Table 2.17). Dissolved stormwater Cu, Zn, and Cd concentrations were higher at Philip than total metal concentrations previously reported in other G V R D catchment areas but still within the concentration ranges in typical stormwater (BC Research Corporation, 1992). As in recent G V R D stormwater monitoring studies, Cu, Zn, and Cd regularly exceeded water quality guidelines at Philip. Land use (including proximal traffic density), total drainage area, and total impervious area can all contribute to metal loadings in stormwater runoff, and most noticeably do so in concert (Kayhanian et al, 2003; McCallum, 1995). The Philip site drains a reasonably sized and largely impervious industrial/commercial area. The result is large metal inputs which not surprisingly sets this location apart from the other catchment areas in the region (Table 2.18). Total metal concentrations for Philip and Blueridge for one early fall storm are also both significantly higher, but these concentrations are not flow-weighted and represent only one severe event which among the storms sampled here, had the highest dissolved and DGT-labile Cu, Zn, and Cd concentrations (September 29-30, 2000, Fig. 2.16C). 2.4.3. Chamber Effectiveness The analysis of the three storm periods with available chamber inlet concentrations reveal that reductions in stormwater metal concentrations may be occurring to some extent for Cu, Cd and Pb (Fig. 2.27 and Fig. 2.28). Elevated concentrations of dissolved N i and Cd within the chamber can possibly be explained by resuspension of finer material including colloidal matter within the chamber and/or metal leaching from associated particulates. It is 56 expected that increased suspended colloid and resulting colloid-metal levels would be measured in dissolved samples but not with the DGT technique. However, very few of the storms sampled during this study have the sufficiently elevated metal concentrations or DGT deployment coverage to adequately elucidate this trend. Metals associated with suspended particulate matter provide insight to the processes occurring within this system (Fig. 2.28). Reduction of SPM and SPM-Zn, Cd and Pb content within the chamber and at the chamber outlet suggests that heavier enriched particles are settling out within the chamber and resulting in a reduced discharge of particulate metals and consequently total metals. The available literature suggest that particle size distributions in stormwater runoff are quite variable, likely site-specific, and related to land use patterns, storm volume and intensity. Most previous studies reported that stormwater consists primarily of fine particles with 60-100% of particles <100 pm diameter (Kobriger & Geinopolos, 1984) and 80% < 25 pm (Randall, 1982) and that metals were predominantly associated with this fine fraction in detained highway runoff solids (Price & Yonge, 1995; Xanthopoulos & Hahn, 1990). In contrast, Sansalone et al. (1998) have shown, through the examination of highway runoff from 13 storms over 2 years, that although most particles were <25 pm in diameter, only 10% of the total solid particulate mass accounted for diameters <100 pm, 25-60-% between 100 and 400 pm, and 40-70% of the mass was >400 pm. These authors also reported that particle surface areas were approximately 3 orders of magnitude greater than is usually estimated based on the assumption of spherical particles (due to stormwater particle surface character which consisted of folds, pores, notches, pits, and roughness). Over eight storms and four seasons, most of the total surface area was provided by particles with diameters between 425 and 850 pm with only a minor contribution from particles <100 pm. Further work by this group and others has demonstrated that a large proportion of the total metal load is associated with the coarser particles of urban runoff and street sediments (Sansalone, 2003; Stone & Marsalek, 1995). Given the mixed urban/industrial/roadway nature of the stormwater particles from this study, similar metal enrichment in the coarser particle fraction is not surprising. Changes in element adsorption efficiencies can be expected to occur with the slight pH decline from inlet (5.9) to chamber (6.4). These pH levels lay within the adsorption edge (the narrow pH range where rapid adsorption/desorption occurs, usually between pH 5 and 7) 57 for the elements of interest (Stumm & Morgan, 1996). Pb and Cu are expected to be primarily associated with the bound phase of manganese and iron oxides at pH > 6; however, Zn would be expected to primarily reside in the dissolved phase (Campbell & LaZerte, 1988; Dzombak & Morel, 1990; Liu et al, 2004; Stumm & Morgan, 1996). Both Cd and Zn adsorption declines with decreasing pH (< 6.5) yet both of these elements are seen at higher SPM-bound concentrations at the lower pH of the inlet station. Mn is present at slightly higher concentrations within SPM of the inlet compared to the chamber outlet (618 and 293 mg/kg Mn). The adsorption changes for specific storm particulates within the chamber require a given residence time not likely encountered here. Liu et al, (2004) observed <30 minutes for 50% adsorption removal by manganese oxides in experimental work with storm runoff and that major cations such as Ca(II) can compete for adsorpition sites (Liu et al, 2004; Zasoski & Burau, 1988). Given these reported observations, it is not surprising that the slight rise in pH did not result in observable Cd and Zn-SPM increases from stormwater inlet to chamber. The above observations and the magnitude of the metal SPM content suggests that the chambers are successful in retaining at least some metal-laden particles (at least for Pb, Zn, Cd). The reduction in total Cu concentration on September 29, 2000 appears to be primarily due to the decrease in SPM load (Fig. 2.28). The small increase in dissolved N i within the chamber relative to the inlet may represent an increase of colloidal metal (able to pass through the 0.45 pm filter). Given the likelihood that resuspension of fine and colloidal matter occurred at some point during this storm (with characteristic long settling times), the slight rise in stormwater pH within the chamber and outlet may have led to increased Ni adsorption and subsequent colloidal N i concentration. The higher Ni -SPM content within the chamber supports this argument and can be related to this element's potential for significant increase in iron and manganese oxide association (2-40% between pH 6 and 6.5) and its ability to form strong Ni-fulvic complexes at this pH. Other researchers have demonstrated inert and very slowly-dissociating Ni-fulvic acid species in freshwaters at pH 6.4 and 8.1 (Lam et al, 1999; Lavigne et al, 1987). Cu-SPM enrichment within the chamber may reflect scavenging by newly introduced particulate and colloidal material or release of previously collected Cu-laden particulates. This explanation is supported by the study of Fedrigo and Boase (1997) who reported that 58 high flow storm events resulted in resuspension and discharge of previously trapped particulates and associated Cu from a similar oil/grit chamber in the vicinity. Evidence from settled sediment data within the chamber and receiving creek (presented in Chapter 3) indicates that on average, Cu, Cd and Zn are not effectively retained within the chambers. The two factors that determine which particles will settle out within a sedimentation chamber are the water flow rate through the chamber and the settling velocity of the suspended particles in the water (Salvia, 2000). As an illustration, the terminal settling velocities of three particle sizes, calculated based on Stokes' law, are given in Table 2.19 along with the estimated time for settling in the chamber. The convective residence time in the chamber (Table 2.20) was estimated from the measured stormwater velocity measured at the chamber outlet. Although this estimate neglects non-ideal flow due to baffles in the chamber, it provides a lower limit of the convective residence time. It becomes apparent that all but the heaviest particles (> 100-400 urn) are expected to be transported through the chamber to the receiving waters. Resuspension of settled particles can also be calculated using the following equation which represents the scour velocity for a particle: S/3(s-\)gdnU2 f (11) where: v h = horizontal velocity that will cause scour, s = specific gravity of particle (2.65 g/mL for quartz/sand), (3 = dimensionless constant ranging from 0.04 to 0.06 (0.05 used),/= Darcy-Weisbach friction factor (usually 0.02-0.03, 0.025 used), d- diameter of particle, and g = gravitational acceleration (9.806 cm/s ). Rearranging and inputting the stormwater velocity of September 29 gives a particle diameter of ~60 urn. In other words, not only are all particles <100-400 um expected to flow freely through the chamber, but also, all previously settled particles <60 um are likely resuspended and transported out of the chamber and into receiving waters. Ultimately, the partitioning of metals between solution and solid phases is controlled by stormwater properties such as pH, ligand (including adsorption site) concentrations, concentrations of competing metal ions (for binding sites), and binding strength to adsorption 59 sites. The nature of particles and presence of other contaminants such as oil and grease (field observations from this study and previously reported (Fedrigo & Boase, 1997)) will also influence metal content. The large accumulation of street sediments within the chamber presents a large pool of particulates and organic matter, which also create a spatial context for particle metal reactivity that may be difficult to detect from total water metal concentrations, especially if stormwater dilution has occurred. The residence time within the chamber will also provide the time scale for potential reactivity. The residence time of stormwater within the chamber at the time of sampling on Sept. 29 was calculated to be ~ 16s (Table 2.19). Whether suspended particles settle out or are transported out of the chamber with flowing stormwater, the time available for reactions within the chamber (length ~ 200 cm) while remaining in suspension is estimated by tcu = Z,/vCh = 24 s (^Ch is residence time within chamber, L is length of outer chamber detention area -300 cm, and vCh is the horizontal velocity of stormwater within the chamber = 12.4 cm/s). Adsorption and precipitation reactions are generally in the order of minutes and would presumably require maximum stormwater velocities of 0.2-5 cm/s for adequate chamber residence times. Stormwater particulate metal concentrations were similar for both the Blueridge and Philip chamber outlets (Fig. 2.28); however, dissolved metals were much lower at Blueridge. Both chambers may receive particles from similar sources or different sources with similar contamination. Both locations drain catchments with high traffic volumes from nearby highways and are characterized by a high degree of land impermeability. Additionally, ongoing construction adjacent to the Blueridge site may have contributed metal sources. Traffic density has been related to Pb, Cu, and Zn in street sediment (McCallum, 1995), while high traffic, building siding and roofs have also been related to high Cu, Cd, Zn and Pb in stormwater runoff (Davis et al., 2001; El Samrani et al., 2004). N i , Cu, Zn, Cd and Pb all exceed stormwater SPM levels found in some previous studies within the G V R D however, estimates using data from the early 1980s indicate metals fall within a broad range for both a residential and an industrial catchment (Table 2.21). The prolific use of metals and presence of on site exposed metals within industrial and commercial areas increase the potential for leaching and entrainment of particle-metal associations within stormwater. N i , Pb and Cu are associated with metallic alloys and Zn with galvanized metals, all of which have to some extent contributed to stormwater discharges within this study. While the G V R D studies 60 report flow-weighted composite averages for a particular event (usually 6-10 hr), the data reported here represent the tail end of a significant storm lasting in excess of 12 hours. Presumably, the highest concentrations would be found at the start of a storm or first flush when particulates and contaminants are initially washed away by stormwater and diminish with storm duration and increased flow (MacDonald et al, 1997; Sekela et al, 1998; Swain, 1983). Although grab samples can misrepresent an event, the sample timing, the finding of similar results for both sites and all inner basin stations within the Philip site, and corroboration with DGT data indicate that this event is likely well represented by the available samples. The stormwater reductions through the chamber system during the severe September 29 t h storm for Cu, N i , Cd, Zn and Pb were 70%, 58%, 94%, 96%, and 94% for particulate-metal load and 39%, 40%, 52%, 42%, and 93% for total metal loads (Table 2.22). Relative to the Blueridge chamber and dry weather conditions at Philip, the Philip outlet concentrations still represent significant increases for Zn, Cd and Cu. The dissolved and particulate data on September 30 t h (post-storm saline waters) indicate, not surprisingly, that most metals are still elevated prior to inflow of tide waters (Fig. 2.28). The data presented here seem to indicate that the chambers did retain particulates and associated metals (Pb, Zn, and Cd) during the events assessed; however, other than for metals that are predominantly particulate (Pb in this case), this does not appear to be sufficiently effective for overall stormwater discharge (in particular for Zn, Cd, and Cu) to the receiving waterbodies. 2.4.4. DGT-Lability Given the duration of the DGT deployment periods and the lack of similar time-averaged dissolved metals, a relationship between the two metal fractions is difficult to define. Some generalities can be gleaned from the simulated mixing results using Philip waters from September 29, 2000. For stormwaters, DGT measurements coincide with dissolved levels for Cu, N i , and Cd. As mixing occurs with saline receiving basin waters, DGT-labile Cu and Cd both decrease proportionately from dissolved levels. DGT and dissolved N i agree at low salinities with only a small decrease in DGT levels at the higher salinities of the receiving basin waters. 61 The DGT units operate on the principles of size exclusion and metal lability. Chelex binding of both free metal ions and labile metal complexes can only occur for those metal species able to pass through the -5 nm pore size of the diffusive gel. The size restriction effectively excludes most colloidal metals (1 nm - 1 pm). In general it can be expected that DGT metal measurements will not include large dissolved organic metal complexes, colloidal metals, inert metals, or slowly dissociating dissolved metal complexes that are not labile within the time frame of DGT (-minutes; (Zhang & Davison, 1995). Recent studies have been able to demonstrate that metal organic complexes in natural waters can be sequestered by the DGT probes; however, increasing molecular sizes lead to slower diffusion through the gels relative to the free metal ion (Luider et al, 2004; Twiss & Moffett, 2002; Zhang & Davison, 1999, 2000). Downward et al (2003) demonstrated that Cu and A l associated with fulvic acids diffuse through the diffusive gel at the same rate as the free metal ions and that differences in measured metal lability between DGT and a flow injection analysis technique using a chelating ion exchanger (FIA) were related to the experimental time scales of the techniques (1-3 s for FIA and mins. for DGT). Their data imply that moderately labile and possibly some inert metal complexes as operationally defined by FIA are included in DGT-labile measurements. Dissolved organic carbon (DOC) levels present in estuaries can be significant. DOC concentrations of 2 mg/L in Burrard Inlet in the 1980s (BIEP, 1992) is within the range of 1-11 mg/L measured in polluted estuaries in the United States (Avery Jr. et al, 2003; Moss, 2001) and the U K (Martino et al, 2002). Larger DOC concentrations of 7-12 mg/L in G V R D stormwaters in the 1990s (Sekela et al., 1998) are on the low side of the 4-202 mg/L range measured in stormwaters of four urban areas in the U.S. (Sansalone, 2003). These higher stormwater levels indicate at least the potential for high episodic discharges into natural receiving waters. Given the much lower DOC ranges measured in most polluted estuaries, these levels likely rapidly decrease due to physical and chemical processes including among others dispersion, flocculation, and degradation. It is generally accepted that organic complexation influences the speciation of many metals at least to some extent (Buffle, 1988; Santschi et al, 1997). Cu speciation in particular is known to be dominated by organic complexation. Twiss and Moffet (2002) have shown in a recent study of both pristine and contaminated U.S. coastal waters that while DGT probes provide consistent reproducible data, they will include to a substantial but variable degree, 62 organic complexed Cu. This was particularly true for natural waters with greater proportions of organically complexed Cu. The heterogeneous suite of organic species and associated metal affinities along with the temporal variability of relative composition and concentration are considerable in natural waters let alone stormwaters. The organic pool is not yet well characterized in most natural waters and hardly addressed in stormwaters. With the high metal levels associated with stormwater discharges into receiving waters, it is probable that the strongest and most metal-specific organic ligands (generally occurring in low concentrations) will tend to be saturated first thus leaving an assortment of more labile and potentially free metal concentrations within the system. The exact nature of these more labile species is not easy to predict and thus great difficulty accompany any attempts to convert DGT measurements to non-operationally defined labile data. Another added complication is that DGT to dissolved ratios are possibly underestimated since dissolved metals in this study will include the metal colloidal fraction. Cd dissolved in stormwater and receiving basin saline water is similar to DGT measured Cd, while the DGT fraction is slightly decreased relative to total dissolved Cd at all of the intermediate salinity treatments. At higher salinities, Cd forms stable complexes with CI" which is measured by DGT. Colloidal Cd complexes are known to occur in seawater (Dai et al, 1995) and may explain the positive deviation from the mixing line (Fig. 2.24). The Visual MINTEQ program was used to predict metal species distribution (Fig. 2.29) and in particular to provide an estimate of the free metal ion. Two simulations were performed using dissolved organic matter (DOM) values of 2 and 20 mg/L respectively in order to predict possible conditions given the ranges found in other regions. The correlations between the calculated free metal ion and DGT measured metal were quite good for Cd and N i , but not for Cu (Fig. 2.30). This is consistent with the fact that the association with D O M in saline waters is significant for Cu and not for Cd or N i (always <5% metal-DOM complexes except in pure stormwater) and that metal-DOM associations are measured to some extent by DGT. Given that DGT measures labile as well as free metals and that dissolved inorganic complexes are small enough to pass through the gel pores, an attempt to correlate DGT metal with inorganic metal was made. While chelex can bind most inorganic metal complexes; the DGT units limit measured species to those that can react with the resin within the time frame of metal-resin exposure (minutes). Most inorganic complexes would be expected to be 63 acquired by the DGT technique with the exception of possibly Ni-carbonates (Leinonen & Lehto, 2000). These species, however, are only expected in minor concentrations in saline waters (Fig. 2.29). Correlations of inorganic metal and DGT metals (Fig. 2.31) are similar to those seen for the free metal ions with the exception of N i and Cd at higher D O M . Metal-D O M complexes increase within stormwater for these two metals with increased D O M which in turn reduces the free metal levels. At elevated salinities, the higher concentrations of seawater cations (e.g. C a 2 + and Mg 2 + ) effectively compete for organic compound binding sites and form stable complexes with metals (Stumm & Morgan, 1996) which is accounted for in the program. Both N i and Cd generally form weaker associations with organic matter than Cu or Pb (Stumm & Morgan, 1996). With both D O M scenarios it is interesting to note the relationship of measured DGT levels with predicted inorganic and free species. DGT N i appears to better represent inorganic N i than free N i . DGT Cd appears somewhere between free and inorganic Cd, while DGT-Cu is greater than either free or inorganic Cu. This suggests a significant role of colloidal matter for Cd since the predicted levels are based on a dissolved + colloidal fraction, and further support the claim that some Cu-DOM is sequestered by DGT. Without an accurate knowledge of the different metal species and any potential metal binding ligands within the waters being analyzed, the task of quantifying metal speciation is not possible; however, agreement with dissolved metals for Cd and N i make DGT useful for measuring acute metal pulses as in this study. Further modifications, as recently suggested (Twiss & Moffett, 2002; Zhang & Davison, 1999, 2000) to utilize a combination of DGT units with different pore sizes will enable a better assessment of inorganic Cu and other metals which under freshwater conditions may complex more readily with D O M . Such measurements will at least be able to provide important monitoring ability which is critical to any biological impact considerations. 2.4.5. Salinity and Dispersion The wide range of salinities, sampling conditions, and stormwater metal concentrations in this study present a complex system. In particular, the common practice of establishing conservative or non-conservative behaviour of metals during estuarine mixing using dissolved concentration vs. salinity plots is not appropriate for this data set. For storm 64 events, when a discharge of freshwater is occurring into the receiving basin, a salinity gradient does not occur within the small inner receiving basin. The saline waters are either displaced by stormwaters during moderate and severe events or mix to produce salinity decreases that are constant within the shallow waters and small reach of the inner basin. As the stormwaters penetrate to the outer receiving basin, dispersion and mixing occur simultaneously and almost instantaneously due to the abrupt increase in basin width and depth, and subsequent mixing with inlet waters. In order to elucidate distance versus salinity influences on stormwater metals, the ratio of metal concentration to chamber concentrations were plotted against distance, salinity and a mixing index for the three storms where stormwater concentrations (albeit from chamber) were available (Fig. 2.32). A salinity-based mixing index was devised in order to represent the proportion of mixing with inlet waters since these salinities were not consistent between storms. Because of the limited applicability of this mixing index resulting from the small area covered by the outer basin stations, caution should be used in generalizing any observed trends to other situations. For the purposes of this study, the derived mixing index was considered adequate in providing a method of discriminating between salinity and dispersion influences at the Philip site. Since the primary source of freshwater entering the saline receiving basin was stormwater, a salinity-based mixing index was calculated as follows: where Si is the salinity of the field sample, S c h is the salinity of the chamber water, and Ssw is the salinity of the seawater end member. Figs. 2.32B and C reveal a bimodal distribution with high metal concentrations in stormwater, relatively lower concentrations in receiving basin waters, and the lack of a mixing gradient. Cu levels for moderate and severe storms appear independent of distance for the periods studied while for the remaining metals distance may at least on occasion influence reductions in dissolved concentrations. The three storms that are represented differ with respect to stormwater pH, water velocity and flow at the time of sampling (Table 2.23). The storm with slowest moving waters and highest pH (Sep. 29/00) coincides with metal reductions for all four elements. These reductions could be explained by particle reactivity (sorption, aggregation) and settling in slow moving waters. 65 The storm with low pH, intermediate velocity and highest flow (Nov. 22) coincides with N i and Cd reduction at only the farthest inner basin station. The low pH which does not change within the inner basin and the high stormwater velocity make the possibilies of increasing metal adsorption and colloidal deposition unlikely. However, during these storms there is a reduction in water velocity within the inner basin which coincides with its width expansion compared to the chamber system. It may be that N i and Cd transport occurred via larger colloids with a specific urban source or after enrichment within the chamber. As discussed earlier, N i particulates are enriched within the chamber and Cd particulates are enriched in initial stormwaters. While three storms are insufficient to draw general conclusions, there is an indication that at least for some storms, dissolved and/or colloidal metals can flow unhindered into natural water bodies. The two metals with the greatest concentrations overall (Cu and Zn) follow this trend. The particulate analyses reveal the increase in particulate metals in stormwater as discussed. A graphical summary of dissolved, DGT-labile, particulate, salinity, pH and total SPM at the Philip site during the September 28-30 sampling period is given in Fig. 2.33. One interesting observation at the Philip site during the storm on September 29 t h is the increase in SPM-Cu and Zn within the outer basin (stations A4-A6) relative to the inner basin (A1-A3) and the chamber (AO). SPM-Pb also shows an increasing trend with distance within the inner basin which drops off at the outer basin stations. At the time of sampling, the inner basin consisted of undiluted stormwater (salinity = 0) whereas mixing had occurred in the outer basin (salinity >20%o). SPM-Mn concentrations were slightly lower at the outer stations compared to the inner stations but showed a roughly 5-fold elevation over the previous dry period and subsequent dry spring period. With the increase in salinity and pH (Fig. 2.33), some degree of colloidal aggregation, flocculation and associated increases in metal adsorption would be expected to occur, especially in the presence of the large storm metal contributions. Increases in salinity can enhance metal removal from solution through aggregation, coagulation, and flocculation of metal-bearing colloids (organic and inorganic) and subsequent adsorption onto newly formed flocculants (Santschi et al, 1997; Sholkovitz, 1978; Stumm & Morgan, 1996). Increases in pH (e.g. 5 to 8) also lead to metal removal via precipitation of insoluble metal oxides, oxidation of iron and precipitation of iron and aluminum oxyhydroxides with subsequent adsorbtion of other metals (Stumm & Morgan, 66 1996). Resuspension of fines is also possible; however, increases in total SPM measurements did not occur at these outer stations and the increase in depth from station A4 to A5 is accompanied by slightly elevated metal-SPM concentrations. Pb's strong particle affinity, pH dependence, and association with Fe-oxides (Santschi et al, 1997) may result in slightly quicker deposition and removal within the outer basin stations. The other interesting observation is that the post storm sampling of September 30 t h, when salinities and pH had both risen (>10 %o and >6.5 respectively), indicating mixing with saline waters, SPM-Cu and Zn are still elevated within the chamber and inner basin, but reduced at the outer basin (S=27%o and pH=7). By September 30 t h, there is a slight elevation in total SPM; however, SPM-Ni, Cd, Pb and Mn are significantly reduced by this time. This, along with the well established high degree of association with organic matter (in particular colloidal) (Dai et al, 1995; Santschi et al, 1997), may indicate that at least for Cu, the elevated levels both on the 29 t h and the 30 t h are due to its involvement with colloidal organic material. Without further data, however, it is difficult to ascertain the relative contributions of all possible geochemical or physical processes to the observed elevated metal levels. With the small outward water flow still occurring on the 30 t h, fine particulates could be discharging into the inner basin from the metal-enriched chamber station or resuspension of previously deposited particulates within the inner basin itself could be occurring. Metal desorption from enriched particulates (urban runoff sources or existing) and other surfaces within the chamber and inner basin could be continuously occurring throughout the storm event with passage of overlying low pH waters. Even in dry periods, lower chamber and inner basin pH relative to the outer basin suggests degradation of organic matter is occurring along with remineralization processes which would contribute metals to overlying waters. Additionally, the increased turbidity at the storm onset could have introduced a different set of particulates with higher metal affinities (clay, Fe or other metal oxides). Of course the diverse nature of inorganic and organic pollutants found in urban stormwater runoff further complicates the situation and obscures precise interpretation. DGT-labile Cu, N i , and Cd indicate elevated average concentrations from the 29 t h to the 30 t h for non-colloidal forms which to a great extent reflect storm conditions (as mentioned previously). The other element of concern is Cd. Similar to other metals, dissolved Cd levels are exceptionally high on the 29 t h and 30 t h and reflected in the elevated DGT-labile concentrations. Similar to Zn, 67 and as seen for some of the other storms in this study (e.g. Nov.22, 1999), while SPM-Cd is elevated on the 29 t h, particulate Cd is only a small proportion of total Cd throughout the system. Elevated dissolved Cd concentrations in the shallow inner basin waters during most dry days are suggestive of some remineralization activity within surficial sediments. A general Cd flux to overlying waters likely leads to transport of dissolved Cd out of the inner basin under most storm conditions. 2.5. Conclusions The evaluation of trace metal data in dissolved, particulate, and DGT-labile forms from two catchments of different land use patterns has provided useful information in the operation of oil/grit chamber systems in urban settings. The industrial area revealed elevated stormwater metal discharges characteristic of this type of setting (high industrial/commercial activity and high impervious ground cover). The residential area may also contribute elevated metal levels, although only periodically, which can be related to new developments, construction and housing materials (siding and rooftops) and similar high ground imperviousness. Of the studied metals, all but N i are of significant concern due to their magnitude, persistence, and potential toxicity. Of particular note in this study are Cu in all metal fractions, Zn in dissolved and particulate forms, Pb in particulate forms, and possibly dissolved and DGT-labile Cd. Early fall rainfall events with characteristic sudden onsets of high intensity storms and prolonged antecedent dry days can contribute significant elevations of dissolved, particulate and DGT-labile Cu, Zn, and to a lesser extent Cd, Pb, and N i within industrial catchments. Oil/grit chamber systems regularly installed in such settings provide inadequate reductions for dissolved, fine particulate-bound, colloidal, and DGT-labile metals. Some reduction of particulate metals, especially those bound to coarse particles do occur; however, significant portions of total metals easily flow into receiving water bodies and resuspension of previously deposited particulate metals may provide additional opportunities for increased metal dissolution and lability. Stormwater flow through the chamber system and the shallow receiving area at the industrial site, in addition to high metal runoff sources, may also be 68 contributing metals from diagenetic remobilization within bottom sediments, particulate metals from sediment resuspension and associated metal dissolution. The volume and intensity of most storms in this region coupled with its urban character indicate the need for a more effective stormwater management technique. Oil/grit chambers may function well as primary treatments. However, better stormwater retention capacity and metal removal processes are necessary to facilitate the goal of metal reduction. Salinity and dispersion both act to reduce dissolved, particulate, and DGT-labile metals in stormwater discharges. Salinity increases induce metal removal from the dissolved phase as coagulation, aggregation, and flocculation processes combine to precipitate and adsorb metals. Dilution of large metal inputs occurs after stormwaters flow to the much larger water basin of Burrard Inlet. Salinity alone may function to reduce Cu and Cd; however, dissolved and DGT-labile N i does not appear to be affected. This has severe implications for stormwaters discharged to small streams and lakes. DGT does provide a fairly easy method of measuring non-colloidal metals. This may be adequate for Cd and N i ; however, an expansion of the technique to include restricted pore-sized gels may help garner more information with respect to metals such as Cu whose speciation is dominated by organics. Despite the difficulties in metal speciation interpretations, DGT does provide a clear indication of high metal discharges, a good time-integrated average of elevated levels, and can consequently reflect effective metal means for whole storm event periods. It can certainly accompany current monitoring practices of measuring total metals. With minimal relative effort and especially if two gel pore sizes are incorporated into the technique, it can provide more ecologically meaningful data for stormwater metal impacts. 69 Table 2.1. Monthly climate and precipitation summaries. Precipitation data from North Vancouver Wharves climate station and air temperatures from Vancouver Harbour climate station (provided by Environment Canada Climate Services). Year Month Mean Mean Mean Temp Total Precip. Min Temp (°C) Max Temp (°C) (°C) (mm) 1998 January 3.4 7.8 5.6 252.9 February 5.5 10.0 7.8 157.5 March 5.7 11.1 8.4 159.8 April 7.0 14.8 11.3 36.9 May 10.9 17.5 14.2 131.1 June 13.6 20.6 17.1 39.4 July 16.1 23.6 19.8 53.9 August d 15.5 24.6 20.0 18.3 September 12.7 21.4 17.1 11.9 October d s 9.1 14.5 11.8 187.0 November s 7.0 10.5 8.8 426.7 December 2.7 6.5 4.6 370.9 1999 January 4.0 7.6 5.8 313.2 February 4.1 8.7 6.4 301.3 March s 4.2 9.9 7.1 165.3 A p r i l d 5.6 14.0 9.8 80.0 May 8.0 15.4 11.7 93.7 June d s 11.4 18.3 14.8 138.3 July 13.5 21.8 17.7 44.8 August 15.3 22.7 19.0 55.6 September 11.2 19.5 15.3 26.3 October 7.6 13.8 10.7 182.2 November s 6.3 10.6 8.5 293.2 December s 3.9 7.2 5.6 290.2 2000 January s 2.1 6.7 4.5 190.1 February 3.4 9.0 6.2 95.2 March 5.0 10.8 7.9 158.2 April 7.2 14.7 11.0 66.6 May 8.9 15.9 12.4 154.1 June 12.3 20.6 16.5 116.9 July 14.2 22.1 18.2 87.6 August 14.1 22.3 18.2 8.0 Septemberds 12.1 19.5 15.8 100.2 October 8.5 14.4 11.5 183 November 4.3 9.2 6.8 111.6 December 2.9 7.3 5.1 171.9 2001 January 4.2 9.1 6.7 178.9 February 2.8 9.0 5.9 43.4 March 5.1 11.5 8.3 160.9 April 7.3 14.9 10.9 151.9 May d 9.5 16.7 13.1 95.0 June 11.8 18.8 15.3 95.3 July 14.2 21.3 17.8 50.7 August 14.6 21.5 18.0 139.2 September 12.2 18.6 15.4 53.9 October 7.8 12.9 10.4 236.6 November 6.5 10.7 8.7 195.9 December 2.1 6.9 4.5 282.3 Superscripts indicate the occurrence of dry weather ( ) and/or storm (s) sampling during the month. Table 2.2. Frequency of sampling during different weather conditions and corresponding climate data. Climate data was recorded at North Vancouver Wharves Climate Station and provided by Environment Canada Climate Services. Dry Weather Light Storms Moderate Storms Severe Storms Seasons in Spring, summer, fall Spring, summer, fall Summer, fall Fall Years 1998, 1999, 2000, 2001 1998, 1999, 2000 1999, 2000 1998 Precipitation 24-hour cumulative means (mm) 0 1-20 6-36 38-70 Storm intensity* (mm/hr) 0 0.1-1 1.2-2 2.7-3 Air temperature (°C) Range Daily mean 8.7-23.9 13.1-18.8 3.2-18.1 7.2-15.2 3.9-20.1 6.0-15.1 9.3-15.9 11.1-14.5 No. of sampling occasions: Philip site Dissolved metals DGT-labile metals Particulate metals 10 6 2 7 4 0 3 3 1 1 2 1 Blueridge site Dissolved metals DGT-labile metals Particulate metals 3 2 2 - 1 1 1 1 1 1 Lighthouse Park (Reference site) Dissolved metals DGT-labile metals Particulate metals 1 1 1 - --Antecedent dry days <lmm (Mean) 1-18 (7.1) 0-2 (0.3) 0,8 7 *Storm intensity was estimated using Environment Canada precipitation levels and recorded field observations. 71 Table 2.3. Analyses of control DGT field units for blanks and method detection limits. Concentrations (Mean ± SD) in pg/L. Ni Cu Cd Blanks N=12 0.696 ± 0.272 1.11 ± 0 . 1 4 2 0.080 ± 0 . 0 1 2 Detection Limi t ' (3a) 0.816 0.426 0.035 'Detection Limit calculated as three times the standard deviation of the blanks. Table 2.4. Concentrations (pg/L) of dissolved metals used in deionized water DGT calibration experiment. Solution Cu Ni Cd Zn Pb M n 1.1 0.09 0.24 0.078 7.0 0.3 0.24 1.2 0.28 0.82 0.080 31.5 0.9 1.20 1.3 0.87 2.79 0.158 22.5 2.5 2.86 1.4 2.77 3.45 0.279 24.2 2.5 4.80 1.5 5.49 8.22 0.729 40.7 2.5 10.7 1.6 20.7 17.8 1.54 83.1 10.0 20.9 Table 2.5. Target Spike Ratios. N i Cu Zn C d Pb Mass isotope 1 62 65 67 111 206 Mass isotope 2 60 63 66 114 208 Sample IS i : % abundance of isotope 1 in Sample (natural) 3.59 30.8 4.1 12.8 24.1 Sample IS2: % abundance of isotope 2 in Sample (natural) 26.1 69.2 27.9 28.7 52.4 Spike ISi: % abundance of isotope 1 in Spike 98.7 99.7 94.6 96.4 99.9 Spike IS2: % abundance of isotope 2 in Spike 0.63 0.30 1.95 0.59 0.02 Sample Ratio (RS=IS,/IS2) 0.138 0.446 0.147 0.446 0.460 Spike Ratio (R t=ISi/IS2) 157 332 48.5 164 4346 Geometric Mean (R s *R t ) 1 / 2 4.64 12.2 2.67 8.54 44.7 Target Ratio (IS1/IS2) 5 15 5 10 50 72 Table 2.6. Concentrations used in Isotope Dilution Mass Spectrometry Trials (pg/L). N i Cu Zn Cd Pb Test Samples: SLEW 0.709 1.62 1.10 0.019 0.027 SLEWSP1 5.71 21.6 21.1 1.02 5.03 SLEWSP2 50.7 102 101 5.02 50.0 SLEWSP3 301 602 601 30.0 300 Target Concentrations for Spike Standard 1 and IDMS 1 Trial 0.7 1.6 1.1 0.02 0.03 Target Concentrations for Spike Standard 2 and IDMS 2 Trial 5.7 22 21 1 5 73 Table 2.7. Analysis efficiency of original and spiked standard estuarine reference material SLEW-2 during laboratory trials. Ni Cu Zn Cd Pb IDMS1 Target Ratio calculated using expected sample concentration 0.7 1.6 1.1 0.02 0.03 (Ug/L) Analysis Efficiency or AE and (%RSD) for : SLEW2 87 84 188 139 68 (3) (2) (124) (8) (676) SLEW2-SP1 97 99 102 139 100 (1) (4) (9) (34) (9) SLEW2-SP2 133 167 134 20 85 (8) (4) ( ID (2000) (42) SLEW2-SP3 92 106 142 31 87 (35) (28) (93) (204) (143) IDMS2 Target Ratio calculated using expected sample concentration 5.7 22 21 1 5 (ug/L) Analysis Efficiency or AE and (%RSD) for: SLEW2 91 94 793 144 -180 (2) (1) (40) (53) (30) SLEW2-SP1 94 100 97 94 104 (3) (1) (5) (2) (1) SLEW2-SP2 97 109 96 93 104 (2) (4) (4) (0.2) (4) SLEW2-SP3 89 109 95 100 104 (1) (2) (2) (1) (2) ' ^ = i o o x [ M e t o / ] — [Metal\.xpected 74 Table 2.8. Analyses of original and spiked standard estuarine reference material SLEW-2 and blanks using isotope dilution mass spectrometry method. Concentrations in pg/L. Ni Cu Zn Cd Pb SLEW2 Certified 0.709 ± 0.054 1.62 ± 0.11 1.1 ± 0 . 1 4 0.019 ± 0 . 0 0 2 0.027 ± 0 . 0 0 5 Analyzed (Mean ± SD) N=12 0.67 ± 0.03 1.50 ± 0 . 0 8 1.2 ± 0 . 3 0.028 ± 0.005 0.0 ± 0.008 SLEW2-SP1.1 Expected 5.53 20.9 20.4 0.986 4.87 Analyzed (Mean ± SD) N=12 5.66 ± 0 . 1 7 20.8 ± 0 . 4 21.9 ± 0.9 0.981 ± 0 . 0 3 9 5.10 ± 0.23 SLEW2-SP4 Expected 3.16 11.4 10.9 0.102 0.497 Analyzed (Mean ± SD) N=12 3.19 ± 0 . 0 6 11.1 ± 0 . 2 11.9 ± 1.3 0.104 ± 0 . 0 0 4 0.520 ± 0 . 0 5 2 Blanks N=24 0.03 ± 0.03 0.21 ± 0 . 0 2 1.21 ± 0 . 4 1 0.006 ± 0.004 0.174 ± 0.162 Detection Limit1 (3a) 0.09 0.05 1.2 0.012 0.485 'Detection Limit calculated as three times the standard deviation of the blanks. 75 Table 2.9. Analyses of marine sediment reference material PACS-2 (Esquimalt Harbour, BC) and filter blanks using microwave digestion method. Concentrations (Mean ± SD) in mg/kg. Ni C u Zn Cd Pb PACS-2 Certified 39.5 ±2 .3 310 ± 12 364 ± 23 2.11 ±0.15 183 ± 8 Analyzed N=10 38.7 ±5.2 332 ± 42 389 ± 2 8 2.05 ±0.37 177 ± 3 1 Blank filters' N=12 3.40 ± 1.10 0.453 ±0.197 6.48 ±2.14 0.0139 ±0.0122 0.065 ± 0.027 Detection Limit 1 ' 2 (3a) 3.29 0.590 6.41 0.0365 0.081 'Blanks and detection limits are reported for a 10 mg sample of sediment. 2Detection Limit calculated as three times the standard deviation of the blanks. 76 Table 2.10. Dissolved and DGT-labile metal concentrations in water& from study sites. Water Properties Weather Philip Site Blueridge Site Lighthouse Park preference Site) Chamber Inner Basin Outer Basin Chamber Receiving Basin Dissolved Copper (ug/L) D 1 7.16-12.9 4.65-11.2 1.35-7.72 4.56-6.11 0.42-2.95 1.36,2.63 L 2 15.3-21.1 8.35-22.5 1.52-6.00 - - -M 3 90.1-191 80.6-116 7.10-43.4 6.58 0.90-2.28 -S 4 364 315 55.8 9.24 1.72-3.24 -DGT-labile Copper (Hg/L) D 2.0-7.4 1.6-5.0 BD*-2.7 4.1,5.2 BD, 1.3 0.9 L 9.8-14.1 6.9-16.6 BD-3.1 - - -M 29.6-579 26.8-338 1.4-8.6 1.0 BD -S 32.9, 132 6.3, 153 0.7, 8.6 - 4.9, 42.2 -Dissolved Zinc (Hg/L) D 130-192 66.4-222 39.8-94.5 18.7-56.1 BD-4.6 5.8,6.1 L 126-240 122-246 50.0-136 - - -M 506-991 460-767 129-267 16.0 2.3 -S 1260 902 284 33.7 5.8 -Dissolved Cadmium (Hg/L) D 0.465-1.83 0.344-2.30 0.181-2.35 BD-0.196 BD-0.076 0.045, 0.049 L 0.266-1.83 0.52-1.70 0.229-1.40 - - -M 1.05-6.65 1.13-4.75 0.537-1.40 0.034 0.042 -S 7.36 4.27 1.43 0.138 0.059 -DGT-labile Cadmium (ug/L) D 0.35-1.57 0.42-1.85 BD-1.51 0.09, 0.20 BD, 0.12 0.08 L 0.68-1.13 0.66-0.99 0.10-0.37 - - -M 0.84-7.10 1.15-2.68 BD-0.91 0.07 BD -S 0.90, 1.69 0.74,2.14 BD, 0.40 0.30 0.21 -Dissolved Nickel (Hg/L) D 0.45-7.21 0.31-4.03 0.38-2.85 0.50-1.11 0.31-0.89 0.13,0.43 L 1.54-10.1 1.20-7.53 0.53-2.07 - - -M 5.89-8.60 5.48-6.65 2.07-3.12 1.12 0.38 -S 7.43 4.38 1.92 1.78 0.54 -DGT-labile Nickel (WJ/L) D 0.9-8.2 BD-3.3 BD-2.8 BD, 0.8 B D , 3.5 1.3 L 1.4-8.2 BD-4.4 BD-2.2 - - -M 7.4-13.7 4.8-10.6 1.0-3.2 1.2 BD -S 1.2, 1.7 2.1,2.4 BD, 1.1 B D BD -Dissolved Lead (Mg/L) D BD-1.5 BD-3.6 BD-1.2 BD-0.9 BD B D L BD-2.7 BD-3.6 BD-1.6 - - -M 1.4-8.3 BD-8.0 BD-1.4 B D BD -S 3.3 1.9 1.1 B D BD -Weather categories include: D 1 : dry weather, Ll station replicates. BD: Below detection. light or weak rain event, M : moderate storm, S : severe storm. Means of basin stations or chamber Table 2.11. Suspended particulate matter and particulate metal concentrations from study sites. Water Properties Weather Philip Site Blueridge Site Lighthouse Park (Reference site) Chamber Inner Basin Outer Basin Chamber/ Outfall Receiving Basin Suspended Particulate Matter (mg/L) D 1 15.1, 19.9 10.3-47.8 10.8-71.8 14.7, 28.5 14.1-46.1 78.2 L 2 - - - - - -M 3 16.5 11.0-25.3 14.4, 26.3 36.4 12.4-39.3 -S4 19.2 10.5-21.3 7.36, 11.1 16.1 13.3-29.1 -Particulate Copper (ug/L) D 4.13-10.5 16.3,39.6 11.6,28.5 0.08, 1.6 0.17-2.1 2.0 L - - - - - -M 230 153 48.7 4.25 0.87 -S 187 154 102 126 0.89 -Particulate Zinc (ug/L) D 4.5-22.0 12.3-40.6 15.2-21.1 BD, 14.6 BD,4.0 3.5 L - - - - - -M 48.1 62.1 24.1 23.0 4.27 -S 27.2 46.6 38.0 37.1 1.78 -Particulate Cadmium (ug/L) D 0.013-0.019 0.045-0.098 0.028-0.073 0.001, 0.003 BD BD L _ - - - - -M 0.030 0.082 0.020 0.034 0.005 _ S 0.098 0.233 0.062 0.152 0.004 -Particlulate Nickel (ug/L) D BD BD BD, 0.67 0.15,0.66 BD,0.21 1.26 L - - - - - -M 1.08 2.39 1.33 0.84 0.40 -S 9.26 5.81 2.79 6.84 BD -Particulate Lead (ug/L) D 1.75-1.79 2.82-7.84 1.18-4.41 0.21,0.65 0.26-0.55 0.24 L - - - - - -M 4.80 5.71 1.63 3.24 0.73 -S 7.83 14.7 5.12 11.3 0.84 -Weather categories include: D 1 : dry weather, L 2 : light or weak rain event, M 3 : moderate storm, S 4: severe storm. &Means of basin stations or chamber station replicates. BD: Below detection. Table 2.12. Ancillary data from study sites. Water Properties Weather Philip Site Blueru ge Site Lighthouse Park (Reference Site) Chamber Inner Basin Outer Basin Chamber Receiving Basin No. Stations A l l 1 3 1-3 1 1-4 1 Length (m) A l l 2 91 50 2 80 -Width (m) A l l 1 4.3-6.5 4-10 1 -Depth (m) A l l 0.1-1.3 0.05-0.3 0.5-3 3-5 0.03-0.7 >10 Temperature (°C) D ' 9-18 9-19 9-19 12-13 13-18 15 L2 8-13 6-15 6-15 - - -Mi 5-14 5-14 7-13 13 15 -S 4 11-16 11-16 11-13 10-14 10-14 -pH D 6.0-6.9 6.0-8 6.5-7.5 6.5 7.5-7.6 7 L 5.8-6.5 5.8-6.5 6.0-7.0 - - -M 5.2-6.7 5.3-6.8 6.1-7.0 5.5 7.0 -S 5.5-6.5 5.6-6.5 6.0-7.8 5.5-6.5 6.2, 7.0 -Salinity (%o) D 5.0-15.8 4.2-16.1 9.7-28.2 0 12.5-22.2 20 .0 ,21 .5 L 2.5-9.9 2.5-11.5 16.9-26.3 - - -M 0-13.4 0-16.8 21-27.2 0 18.7-24.8 -S 0.2-9.3 0.1-10.8 10.6-26.2 0 15.6-22.3 -Weather categories include: D 1 : dry weather, L 2 : light or weak rain event, M 3 : moderate storm, S 4: severe storm. &Means of basin stations or chamber station replicates. Table 2.13. Fall 2000 sampling dates, times and water velocity at Philip site. Sampling Date Weather Category1 Salinity (%o) Time of Low Tide (PST)2 Time of Philip Site Sampling (PST) Water velocity (cm/s) Chamber Outfall3 (AO - Al) Inner Basin (A3)4 September 28,2000 D 14.5-15.4 12:52 14:00-15:00 1.0 6.0 September 29,2000 S 0.7-0.9 13:34 10:45-11:15 12.3 2.1 September 30, 2000 M 13.4-16.8 14:17 12:00-13:00 5.2 0.7 Weather categories include: D: dry weather, M : moderate storm, S: severe storm. 2 PST: Pacific Standard Time 3Chamber Outfall is the location within the inner basin immediately adjacent to the chamber outfall pipe, (between stations AO and A l , Fig. 1.2). 4 A3 is the farthest seaward station within the inner basin, Fig. 1.2) Table 2.14. Stormwater concentrations (Mean ± SD) from side drainage location at Philip Site. Dissolved1 (ug/L) Particulate metal (ug/L) SPM metal1 (mg/kg) Cu 257±126 518 5720 ± 3590 N i 4.24 ±0.77 8.81 99.8 ± 66.2 Cd 5.78 ±0.69 4.50 32.1 ±4 .0 Zn 1050 ± 2 0 1160 8290±1030 Pb 1.59 ±0.36 385 2550 ±740 Concentration means and standard deviations are derived from three samples taken 20 minutes apart. Total suspended particulate matter was 132±93 mg/L. 80 Table 2.15. Stormwater metal concentrations (Mean ± SD) at Philip Site. Dissolved1 (ug/L) DGl r-Labile2 (ug/L) Chamber Inlet Chamber Chamber Outlet Chamber Inlet Chamber Chamber Outlet Nov. 13, 1999 C u 34.4 ± 2 . 7 16.8 ± 2 . 0 25.6 ± 2 . 9 58.1 ± 3 . 2 9 . 8 ± 1.1 9.2 ± 1.0 N i 0.83 ± 0.09 2.04 ± 0.06 2.26 ± 0 . 5 5 41.3 ± 13.6 8.2 ± 0 . 9 3.8 ± 0 . 3 C d 0.213±0.011 0.402 ± 0 . 0 1 7 0.871±0.027 0.51 ± 0 . 1 9 0.68 ± 0.02 0.41 ± 0 . 1 2 Z n 6 0 9 ± 1 5 7 751 ± 2 8 9 7 6 8 ± 1 8 9 - - -Pb 0.5 ± 0 . 1 2.7 ± 0 . 1 1.4 ± 0.1 - - -Nov. 22, 1999 C u 92.0 ± 16.4 95.7 ± 7 . 8 90.6 ± 8.4 45.9 ± 5 . 4 29.6 ± 6 . 4 37.3 ± 10.1 N i 5.55 ± 0 . 6 9 7.45 ± 0 . 8 1 6.58 ± 0 . 8 3 3.7 ± 0 . 5 3.6 ± 0 . 4 3.1 ± 1.2 C d 4.09 ± 1.06 5.12 ± 0 . 9 4 4.26 ± 0 . 1 8 4 . 0 2 ± 1.13 2.57 ± 0.15 2.41 ± 0 . 6 3 Zn 250 ± 3 3 304 ± 5 2 216 ± 26 - - -Pb 7.2 ± 2 . 6 6.8 ± 1.6 6.3 ± 1.5 - - -Sept. 29, 2000 C u 289 ± 4 1 364 ± 2 8 . 8 306 ± 6 0 603 ± 89 5 7 9 ± 1 2 6 345 ± 75 N i 5.03 ± 0.99 7.43 ± 1.97 5.24 ± 1.83 8.9 ± 1.3 10.3 ± 2 . 1 1 1 . 0 ± 1.9 C d 8.73 ± 1.59 7.36 ± 2 . 0 6 5.05 ± 1.75 6.71 ± 0 . 8 7 7 . 1 0 ± 0 . 8 6 2.94 ± 0 . 3 5 Z n 1120 ± 2 5 0 1260 ± 3 0 7 1000 ± 2 0 0 - - -Pb 2.0 ± 0 . 6 3.3 ± 2 . 7 1.1 ± 0 . 4 - - -'Concentration means and standard deviations are derived from three samples taken 20 minutes apart. 2 D G T deployment was approximately 24 hours. 3Chamber and chamber outlet waters reflect tidal influence (S~5%o) during this light storm. 81 Table 2.16. Stormwater metal concentrations of particulate matter (Mean ± SD) on September 29, 2000 at Philip site. Particulate metal (pg/L) SPM metal1 (mg/kg) Chamber Chamber Chamber Chamber Chamber Chamber Inlet Outlet Inlet Outlet Cu 421 187 125 4840±1080 10200±1380 9420 ±400 N i 12.2 18.1 5.12 139 ± 21 979 ± 65 386 ± 14 Cd 2.12 0.189 0.131 23.2 ± 1.8 10.3 ± 1.9 9.95 ± 1.14 Zn 668 49.0 28.4 7750 ± 700 2670 ± 520 2160 ±550 Pb 180 14.4 11.5 2170 ±530 780 ± 92 871± 720 'Concentration means and standard deviations are derived from three samples taken 20 minutes apart. Total suspended particulate matter was 86.8±52.1 mg/L at inlet, 19.2±2.7 mg/L within chamber, and 13.3±1.0 mg/L at outlet. 82 Table 2.17. Comparision of stormwater metal concentrations total and (dissolved), in pg/L, within the Greater Vancouver Regional District (GVRD). Ni Cu Zn Cd Pb Aquatic Life Criteria Total' (Dissolved)2 25 (16) 2-4 (2.9-3.8) 30 (37) 0.017 (0.83-0.95) 1-7 (0.54-14) Typical Stormwater' Range Mean <2-126 12 4-560 34 10-5750 160 0.7-30 0.7 3-280 140 GVRD Summary4 Range Mean -5.6-232 32 13-237 73 <0.1-0.6 0.24 <l-84 11.7 GVRD Catchments: Wagg Creek, North Vancouver5 <l-3 (1) 5-19 (1.7) 12-56 (1) <0.2-0.2 (<0.2) 2.2-49 (<0.1) Still Creek, Burnaby5 <l-4 (4) 8-57 (6.1) 34-110 (36) <0.1-0.6 (<0.1) <l-84 (0.9) Serpentine River, Surrey5 <2-5 (<1) 7-37 (1.9) 22-167 (1) <0.2-0.3 (<0.2) <1-21 (<0.1) Crown Street, North Vancouver6 -35.7-43.9 (19.4-25.9) 68.8-127 (8.02-45.6) -_ This Study: Philip Street, North Vancouver7 10.3 (5.24-6.88) 431 (84.9-306) 1030 (456, 1000) 5.18 (1.04-5.05) 12.6 (1.1-7.3) Blueridge Creek, North Vancouver7 8.62 (1.12, 1.78) 135 (6.58, 9.24) 70.8 (16.0,33.7) 0.290 (0.034, 0.138) 11.3 (<0.5) ' (CCME, 2003) 2(Buchman, 1999) 3 (BC Research Corporation, 1992): Compilation of various studies; Cu, Zn, and Pb means from Metro Seattle; N i and Cd means from U.S. Nationwide Urban Runoff data base. 4(MacDonald, 2003): Compilation of several storm sampling reports during 1993-1997 within the G V R D ; estimated mean represents calculated 'typical'urban runoff character. 5 { (GVRD, 1998): Total (5 storms) and dissolved (1 storm) metals are composite, flow-weighted means, 1996-1997. 6(Fedrigo & Boase, 1997): Chamber outlet flow-weighted means, 3 storms sampled (Nov., Feb., Mar., 1997). 7This Study: Chamber outfall station (stormwater only); dissolved metal represents 3 moderate/severe (Philip) and 2 (Blueridge) moderate/severe storms from 1999-2000; total metal concentration represents 1 storm, September 29, 2000. 83 Table 2.18. Characteristics of Catchment Areas. Catchment Area Major Land Uses Drainage Area (ha) Total Impervious Area (%TIA) Average Slope (%) Wagg Creek Catchment, North Vancouver' Residential, Traffic 60 37 5 Still Creek Catchment, Burnaby' Indus/Comm, Residential, Traffic 152 60 0.14 Serpentine River Catchment, Surrey' Residential, Parkland 141 50 0.55 Crown Street Catchment, North Vancouver 2 Industrial 5 99 0.59 Philip Street Catchment, North Vancouver 2 Industrial/Comm, Traffic 78 71 0.71 Blueridge Creek Watershed, North Vancouver 3 Res idential/Parkland, Traffic 4 119 56 -' (GVRD, 1998) 2(Fedrigo & Boase, 1997) 3 ( G V R D , 1999): 1996 estimates. 4Land use estimates from this study. This area is bounded by major highway routes (Fig. 1.3). Table 2.19. Settling properties of particles within the Philip chamber. Particle diameter: 25 um 100 um 400 um Terminal settling velocity*: 18/7 (from Stokes' law) 0.056 cm/s 0.89 cm/s 14 cm/s Chamber residence time required for particle settling& U = D / V s 30 min 2 min 7s *g = gravitational acceleration, 9.806 m/s2, n, = viscosity of the fluid (0.001005 kg/m/s for water at 20°C), p s = density of particle (2.65 g/mL for quartz/sand), p = density of fluid (1 g/mL), d= diameter of particle, (Stumm & Morgan, 1996). t c h = chamber residence time, D = water depth in chamber, V s = particle settling velocity 84 Table 2.20. Stormwater residence time. Storm: 1999 2000 November 13 November 22 September 29 Water velocity at chamber outlet 8.16 cm/s 20.5 cm/s 12.3 cm/s Chamber residence time 25 s 10 s 16 s Chamber system residence time 74 s 29 s 49 s Table 2.21. Particulate-bound metal concentrations (mg/kg) within the Greater Vancouver Regional District (GVRD). N i C u Zn C d Pb Suspended Solids1: Still Creek Catchment, Burnaby 35.9 422 1800 5.6 306 Serpentine River Catchment, Surrey 35.3 199 904 2.5 230 Industrial Catchment, Burnaby2 - 102 103 - 103 Residential Catchment, Burnaby2 - 103 103-104 - 103 Street Sediments3: Brunette River Watershed, Burnaby 11-76 30-268 112-1383 <3-4.7 66-475 ' (GVRD, 1998): One storm event represented for each location during 1996-1997. Estimated from (EVS et al, 1996). 3(McCallum, 1995): Various locations and land uses, September 1993. 85 Table 2.22. Percentage reductions of stormwater metals through chamber at the Philip site. Event Element Dissolved Particulate Total DGT-Labile Nov. 13,1999 Cu 26 - - 84 N i -170 - - 91 Cd -310 - - 19 Zn -26 - - -Pb -160 - - -Nov. 22,1999 Cu 2 - - 19 N i -19 - - 18 Cd -4 - - 40 Zn 14 - - -Pb 12 - - -Sept. 29, 2000 Cu -6 70 39 43 N i -4 58 40 -23 Cd 42 94 52 56 Zn 11 96 42 -Pb 44 94 93 -Table 2.23. Stormwater flow at the Philip site. Parameters/Storm November 22,1999 September 7, 2000 September 29,2000 pH 5.3 6.5 5.3 Water velocity (cm/s) 20.5 27.5 12.3 Flow (L/s) 148 51.2 84.2 86 «— Ag • «-5+ Distance Figure 2.1. Schematic representation of the free concentration of ionic species in a gel assembly in contact with natural water where the concentration is Cb (DBL, diffusive boundary layer). From Zhang & Davison (1995). 87 5.0e-2 Time (hrs) Figure 2 . 2 . Measured metal mass in resin layer for gel assemblies immersed in a stirred mixed metal (25 pg/L Cu, 1 pg/L Cd, 10 pg/L Ni) N a N 0 3 (0.01 M) solution at 20°C for different times. The solid lines are predicted using Eqs. 4 and 5 and diffusion coefficients of distilled water. Error bars are ±1 standard deviation. 88 A. 12°C B. 20°C Figure 2.3. Measured metal mass in resin layer for gel assemblies immersed in stirred solutions of mixed metal standards and 0.01 M NaN03 at temperatures of A) 12°C and B) 20°C. The solid line represents predicted metal mass (M) in resin using Eq. 4. Error bars are ±1 standard deviation. 89 1.0e-5 • • Cu • • 1 1 • 0.0 0.5 1.0 1.5 2.0 2.5 Bulk Solution [Total Metals] (10"6 M) 4.0e-5 3.5e-5 -\ 3.0e-5 \ 2.5e-5 -|_ 2.0e-5 -It: <d 1.5e-5 Q 1.0e-5 -| 5.0e-6 0.0 0.0 Ni 0.5 1.0 1.5 2.0 Bulk Solution [Total Metals] (1CT6 M) 2.5 3.0e-5 0.5 1.0 1.5 2.0 Bulk Solution [Total Metals] (10 6 M) 2.5 Figure 2.4. Effect of total metal concentrations on the effective diffusion coefficient Deff for Cu (top), N i (centre), and Cd (bottom). The solid lines are the diffusion coefficients for the specified metal ion in distilled water (Li and Gregory, 1974). The bulk solutions include Cu, N i , Cd, Zn, Pb, and Mn in the concentrations listed in Table 2.4. Deff is calculated by rearranging Eq. 4: Deff = MAg CbtA 90 0.4 0.3 H CD O 0.2 H </> m 0.1 Cu 0.0 0.4 0.3 0.2 0.1 0.0 3.5e-2 3.0e-2 — 2.5e-2 -o 2.0e-2 O w 1.5e-2 ro 2 1.0e-2 -5.0e-3 -0.0 5 10 15 [Cu] in Bulk Solution (ug/L) 20 25 Ni \ -< • 0 2 4 6 8 10 12 14 16 18 20 [Ni] in Bulk Solution (ug/L) Cd o.o 0.5 1.0 1.5 [Cd] in Bulk Solution (ug/L) 2.0 Figure 2.5. Measured metal mass for Cu (top), N i (centre), and Cd (bottom) in resin layer for gel assemblies immersed in stirred solutions of mixed metal standards in deionized water. Error bars are ±1 standard deviation. 91 Chamber Inner Basin Distance from Chamber (m), Philips Site Outer Basin o [D] AUG25/98 (3PM) Dissolved o [D] AUG26/98 (1 PM) Dissolved • [D] AUG25-26/98 DGT-Labile AUG25/98 (3PM) Salinity AUG26/98 (1PM) Salinity Figure 2.6. August 25-26, 1998: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site. E E 0J Q_ ro o h-Q. Q. >, E 80 60 40 20 0 CD O 50 -1 40 30 20 10 0 cn O 10 E E_ 80 sT 60 B 4 0 f 20 o 0 p 0CT2 A 0CT1 0CT3 0 C T 4 6AM 6PM 6AM 6PM 9/7/98 9/14/98 9/21/98 9/28/98 10/5/98 0. 6AM 6PM 6AM 6PM Date 30 - —S V-20 10 -; . . . . A . . _ _ i , — - — 0 A A : A • . . . A —\ S-• • o 0 Chamber O o 10 15 Inner Basin Distance from Chamber (m), Philip Site Outer Basin B • * ^ • o o i * , S 80 [ 90 100 110 O [D] SEP30-OCT1/98 DGT-Labile • [S] OCT2-3/98 DGT-Labile • [L] OCT3-4/98 DGT-Labile • a • SEP30 & OCT1/98 Salinity » OCT2/98 (12PM) Salinity A OCT3/98 (1PM) Salinity * OCT4/98 (1PM) Salinity Figure 2.7. September 30-October 4, 1998: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site. 93 E E g 80 '•So 60 ? 2 0 £ 0 (0 o •i n n B -fin - „ „ „ n IL „ 200 O) 150 ~T 100 •a O 50 0 2 1 0 10 E .§,80 c 60 o • 1 2 0 '8 0 NOV14 NOV15 10/19/98 10/26/98 11/2/98 11/9/98 11/16/980-Date 6AM 6PM 6AM 6PM 10 15 80 90 100 110 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philip and Blueridge Sites A 20 Blueridge Receiving Basin [S] N0V14-15/98 DGT-Labile NOV14/98 (12PM) Salinity NOV15/98 (3PM) Salinity B Figure 2.8. November 14-15,1998: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip and Blueridge Sites. 94 D. O a. 80 60 ^ 40 20 0 innfll 5 2/22J'99 3/1/ 30 20 ^ 10 0 50 40 30 20 10 0 a. S 1 0 5- 10 =1 i" 5 0 3/8/99 3/15/99 3/22/99 3/29/99 Date -k k-B a % % 1 1 k k 1 $ , 1 • i 1 * § 1 1 > 1 1 > > t • i t 1 T r 1 1 I I 0 5 10 15 80 90 100 110 Chamber Inner Basin Distance from Chamber (m), Philip Site Outer Basin o [L] MAR23/99 (4PM) Dissolved • [L] MAR26/99 (5PM) Dissolved • [L] MAR23-26/99 DGT-Labile MAR23/99 (4PM) Salinity MAR26/99 (5PM) Salinity Figure 2.9. March 23-26, 1999: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site. 95 80 60 40 4 "8 2 0 , n n n r 3/22/99 3/29/99 4/5/99 Date 4/12/99 4/19/99 B Chamber Inner Basin Distance from Chamber (m), Philip Site Outer Basin o [D] APR17/99 (4PM) Dissolved o [D]APR18/99 (1PM) Dissolved • [D] APR17-18/99 DGT-Labile APR17/99 (4PM) Salinity APR18/99 (1PM) Salinity Figure 2.10. April 17-18, 1999: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site. 96 o 5/3/99 30 —j— 20 -10 -0 — 5/10/99 5/17/99 5/24/99 5/31/99 6/7/99 Q_ Date 50 -40 30 -i 20 -10 i 0 J -a 1 O § 0 • JUN4* (Misty Rain) JUN3 1 6AM 6PM 6AM 6PM B 4 £ 10 5 0 Chamber 10 15 80 90 Inner Basin Distance from Chamber (m), Philip Site 100 110 Outer Basin" o [D] JUN3/99 (3PM) Dissolved © [L] JUN4/99 (1PM) Dissolved • [L] JUN3-4/99 DGT-Labile • • JUN3/99 (3PM) Salinity •* JUN4/99 (1PM) Salinity Figure 2.11. June 3-4, 1999: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site. June 4 water sampling occurred under misty rain conditions prior to significant rainfall. 97 10/11/99 10/18/99 10/25/99 11/1/99 11/8/99 11/15/99 Date E, 80 c 60 I 4 0 CO « 20 o 0 £ NOV13 NOV12 6AM 6PM 6AM 6PM B Inner Basin Outer Basin Distance from Chamber (m), Philips Site o [L]NOV12/99 (5PM) Dissolved © [L] NOV13/99 (3PM) Dissolved • [L] NOV12-13/99 DGT-Labile NOV12/99 (5PM) Salinity NOV13/99 (3PM) Salinity Figure 2.12.November 12-13, 1999: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and N i levels in water at Philip Site. 98 80 c o 75 60 5 40 20 0 E •E 80 § 60 | 4 0 , ij i,ii,iniiiMii?[l, ^[U, i,Lci^l ,1 o o 10/25/99 11/1/99 11/8/99 11/15/99 11/22/99°-Date N0V21 NOV22 6AM 6PM 6AM 6PM B Inner Basin Outer Basin Distance from Chamber (m), Philips Site o [L]NOV21/99 (4PM) Dissolved o [M] NOV22/99 (3PM) Dissolved • [M] NOV21-22/99 DGT-Labile NOV21/99 (4PM) Salinity NOV22/99 (3PM) Salinity Figure 2.13. November 21-22, 1999 A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, and Ni levels in water at Philip Site. 99 80 60 40 20 0 E E, 80 C 60 O •5.20 o o P SEP7 SEP8 8/7/00 8/14/00 8/21/00 8/28/00 9/4/00 9/11/00 Q-Date 6AM 6PM 6AM 6PM A B Inner Basin Distance from Chamber (m), Philips Site Outer Basin o [M] SEP7/00 (1PM) Dissolved o [L] SEP8/00 (8AM) Dissolved • [M] SEP7-8/00 DGT-Labile • • SEP7/00 (1PM) Salinity * SEP8/00 (8AM) Salinity Figure 2.14. September 7-8, 2000: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip Site. 100 'is 60 -a. 40 i £ o i ^ <S 8/21/00 8/28/00 9/4/00 Date 'i i 9/11/00 9/18/00 ~ 30 CL Q L — 20 >. C 11 To W 0 _ 5 0 =d 40 3 30 3 20 O 10 0 D) 2 "O 1 o d 10 H a, _ 5 0 T 4 cn Q. ] 1 rr~+- ——» — A — I A 1 1 ( r 1 1 i ) 1 — 1 1 8 — • — # 1 $ • t ? 6 6 • • i i I I I § o 1 i i '4 • i i , i i 0 Q - T • .—9 i — • • i 0 5 10 15 80 90 100 110 Chamber Inner Basin Distance from Chamber (m), Philips Site Outer Basin ~~ B o [D] SEP18/00 (3PM) Dissolved o [D] SEP19/00 (2PM) Dissolved • [D]SEP18-19/00 DGT-Labile SEP18/00 (3PM) Salinity SEP19/00 (2PM) Salinity Figure 2.15. September 18-19, 2000: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, Ni , and Pb levels in water at Philip Site. 101 E £ o 80 j B 60 i cip 40 i |U 20 \ ft 0 15 o (-30 Q. Q. 20 ->, E 10 -"to W n 9/4/00 „ 600 _ l 5 400 <3 2 0 0 0 8 -I 6 4 2 0 15 co -a O co 10 -5 5 1500 a, 1000 c 500 N 0 10 8 6 4 2 0 ! a. 9/11/00 9/18/00 Date 9/25/00 10/2/00 -• • • : s 0 ] * « o • ) * 0 1 V c 0 o 10 15 1 c 80 90 100 110 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philips Site — 80 g 60 ra 40 9-20 '8 0 SEP29 SEP28 SEP30 "Mi a 3 0 S 20 ii 1 0 C/5 0 40 co 3 o 20 0 --1.0 a. '0.5 •a O 0 0 15 a. i 5 0 ,-.100 _J , co -3=50 c N 0 10 =d 8 3 6 iT 4 0. 2 6AM 6PM 6AM 6PM 6AM 6PM B ty I I I « I . I I —' 1-0 10 20 30 40 50 60 70 80 CH/OF Receiving Basin Distance from Chamber Outfall (m), Blueridge Site o [D] SEP28/00 (2PM) Dissolved e [S]SEP29/00 (11AM) Dissolved • [M] SEP30/00 (1 PM) Dissolved • [D] SEP27-28/00 DGT-Labile • [M] SEP29-30/00 DGT-Labile — - SEP28/00 (2PM) Salinity • • * • SE29/00 (11AM) Salinity * SE30/00 (1PM) Salinity Figure 2.16. September 27-30, 2000: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip and Blueridge Sites. 102 E E Q_ 75 o h-Q. Q. >. 30 -20 -"E 10 Sal 0 30 _J i 20 Cu 10 0 _ i 3 Z Cd 1 -0 _ i 10 -3 5 z 0 ^200 D) c 100 N 0 8 _ i "3i 3 6 4 Pb 2 n 80 60 40 -j 20 | 0 3 4/23/01 n„n,-, , 'i 4/30/01 5/7/01 Date • c 10 15 5/14/01 5/21/01 t 3 0 • J ^ 80 90 100 1.10 Outer Basin Chamber Inner Basin Distance from Chamber (m), Philip Site 10 30 13> 20 n 10 O 0 3 2 ~&> 3 T> 1 -O 0 A 1 L_^ , , A — ^ 1 # 5" 10 \ •3" 5 =d200 CO 3 ^100 N 0 8 6 4 2 0 B • _0 10 20 30 40 50 60 70 8Q CH/OF Receiving Basin LHR Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site o [D] MAY21/01 (12PM) Dissolved • [D] MAY22/01 (1PM) Dissolved • [D] MAY21-22/01 DGT-Labile • - A • MAY21/01 (12PM) Salinity * MAY22/01 (1PM) Salinity Figure 2.17. May 21-22, 2001: A) Daily precipitation record (recorded at nearby North Vancouver Wharves Climate Station, provided by Environment Canada Climate Services); B) Salinity levels; and C) Cu, Cd, N i , Zn and Pb levels in water at Philip, Blueridge and Lighthouse Reference (LHR) Site. 103 • O • MAY22/01 —• SEP28/00 - A - SEP29/00 - V - SEP30/00 _0 20 40 60 80 100 J20 0 10 20 30 40 50 60 70 80 LHR Chamber Inner Basin Outer Basin CH/OF Receiving Basin *" Distance from Chamber (m), Philip Site Distance from Chamber Outfall (m), Blueridge and Lighthouse Reference (LHR) Sites Figure 2.18. Total suspended particulate matter (SPM) at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 104 • o • MAY22/01 - o • SEP28/00 SEP29/00 v SEP30/00 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philip Site 1_\ 1 * i—y/ 1 • % *^\>- -i- o ^ - i . -a— 0 10 20 30 40 50 60 70 80 LHR CH/OF Receiving Basin Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site Figure 2.19. A,B) Particulate Cu in SPM and C,D) Particulate Cu in waters at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 105 • o - May 22/01 - a • SEP28/00 - A - SEP29/00 v SEP30/00 J20 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philip Site t 1 4 Q_ CO 1.2 cn •) 1.0 E, ~ 0.8 06 C £ 0.4 I (11 o 0.2 ™ 0.0 C ZJ o CD • o t: ra 0. 15 10 j 25 -1 — 20 1 CH/0F Receiving Basin Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site Figure 2.20. A,B) Particulate N i in SPM and C,D) Particulate N i in waters at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 106 • -o • MAY22/01 - o SEP28/00 — S E P 2 9 / 0 0 - - T - SEP30/00 5 25 0. w O) 20 ai 3 •o 15 O •a c =J 10 o CD A) 5 O CO CL 0 06 "3i 3 0.5 "D O 04 -o <z 03 o i <u 0.2 o arti 0 1 0.0 120 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philip Site 5 25 o_ CO CO 20 3 T3 15 O T J C 10 D o £• d> 5 o t Pa 0 0 6 _ i O) 3 0 5 -a O 04 "o =! 0 3 O 00 1 cu 02 o t 0.1 CO 0 0 n —s/-A. 5 " T ^ I S 0 .1' -0 10 20 30 40 50 60 70 80 LHR CH/OF Receiving Basin Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site Figure 2.21. A,B) Particulate Cd in SPM and C,D) Particulate Cd in waters at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 107 • o - MAY22/01 SEP28/00 SEP29/00 -v- SEP30/00 CD 3 c N X} c ZS o CD i <v o r ra Q. 150 100 •S 50 c \ I * ' l i _0 Chamber 20 40 60 80 100 120 Inner Basin Outer Basin Distance from Chamber (m), Philip Site D_ CO -S" Ol c N -a c o CO 12 10 «2 2 150 en 3 tz N •D sz Z> o CQ i _a> o t C L 100 50 [B * i z ^ J b -7, » • =n • —^  •— — . u A. . 0 10 20 30 40 50 60 70 80 LHR CH/OF Receiving Basin Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site Figure 2.22. A,B) Particulate Zn in SPM and C,D) Particulate Zn in waters at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 108 • -o • MAY22/01 — • SEP28/00 A — SEP29/00 V SEP30/00 120 Chamber Inner Basin Outer Basin Distance from Chamber (m), Philip Site 2 0 . 00 oi CO £ -Q Q_ ~o c o CO cb o to Q-co - O Q_ x> =J o 00 I o •e ro 0. 1 4 1 2 1.0 0.8 0.6 0.4 0.2 0.0 50 40 30 20 10 [B 0 10 20 30 40 50 60 70 80 LHR CH/OF Receiving Basin Distance from Chamber Outfall (m), Blueridge Site, and Lighthouse Reference (LHR) Site Figure 2.23. A,B) Particulate Pb in SPM and C,D) Particulate Pb in waters at Philip, Blueridge and Lighthouse (Reference) Sites. Error bars represent ±1 standard deviation. 109 400 300 O Dissolved • DGT-Labile co 3 200 o 100 10 6 H 0 5 o 10 15 Salinity (ppt) 20 0 • 25 o Dissolved • DGT-Labile 30 2 0 10 8 6 co a . T> 4 O 0 5 QO 9 Q 0 5 10 15 20 Salinity (ppt) 25 O Dissolved • DGT-Labile 0 10 15 Salinity (ppt) 20 25 30 30 Figure 2.24. Concentrations of dissolved and DGT-labile Cu (top), N i (centre), and Cd (bottom) produced upon mixing unfiltered stormwater with unfiltered seawater from the Philip site on September 29, 2000. 110 Figure 2.25. DGT-labile Cu (top), N i (centre), and Cd (bottom) as percentage dissolved in mixing experiment of September 29, 2000 waters. A. Storm Properties B. DGT-Labile Metals Q. O £ 0. o I-200 150 100 50 24-hour Rain Volume. Total Storm Volume III lu rioui 6 E E 4 >. :ens 2 c 0 Average Storm Intensity Initial intensity I III 3 O O Q 800 -] _J 600 -O 0) 400 -15 ro h— 200 -O Q 0 -5" DI 3 15 -z 0J 10 -5 03 _l h- 5 -O Q 0 -• Chamber A Inner Basin • Outer Basin Storm Dates Storm Dates Figure 2.26. Characteristics of moderate and severe storms: A) Storm properties including number of antecedent dry days (top), total precipitation (centre), and rainfall intensity (bottom); and B) DGT-labile concentrations (pg/L) for Cu (top), N i (centre), and Cd (bottom). Storm parameters are estimated from Environment Canada data and field observations. 112 Dissolved DGT-Labile A. November 13,1999 B. November 22,1999 3-o 100 80 60 40 20 0 Cu PCI PC P C O PCI PC PCO Figure 2.27. Dissolved and DGT-labile metals (pg/L) measured in stormwaters during a light storm A) November 13, 1999 and moderate storm B) November 22, 1999 within Philip chamber inlet (PCI), Philip chamber (PC) and Philip chamber outlet (PCO) waters. 113 A. September 29, 2000 • D isso lved (ug/L) 1 Part iculate (ug/L) M e t a l - S P M (mg/kg) - 1.5e+4 - 1.0e+4 Q_ 00 - 5.0e+3 ZJ o 0.0 • r 1.5e+3 - 1.0e+3 CL °? - 5.0e+2 z 0.0 • "3 r 40 E - 30 —• - 20 a_ °? •6 - 10 o 0 • p 1e+4 C D - 8e+3 E - 6e+3 0-- 4e+3 °? c - 2e+3 N - 0 • r - 4.0e+3 - 3.0e+3 £ is - 2.0e+3 o_ C O - 1.0e+3 o_ 0.0 • 1000 800 600 400 200 0 E 0. co September 30, 2000 I D i sso l ved (ug/L) I I Par t icu late (ug/L) D G T - L a b i l e (ug/L) , (Sept . 29 -30 dep loyment ) M e t a l - S P M (mg/kg) «1000 00 800 PCO BCO PCO BCO Figure 2.28. Dissolved, DGT-labile, particulate, and SPM-bound metals measured in A) September 29, 2000 stormwater (severe storm) and B) September 30, 2000 receiving basin water within Philip chamber inlet (PCI), Philip chamber (PC), Philip chamber outlet (PCO), and Blueridge chamber outlet (BCO) waters. pH and total SPM (bottom) are also provided for both dates. DGT-labile metals are only provided for September 30, 2000 and represent the 24-hr deployment period from September 29-30, 2000. Error bars are ±1 standard deviation. 114 - 100 CO O 75 H (-— 50 o v? 25 -100H O 75 t 50 H o 25 A. DOM=2 mg/L Copper 3 3 4 2 5 9 2 5 5 2 1 1 1 5 7 7 5 . 2 2 8 . 9 -100H O 75 ! l 50 H o vP 25 0 2 3 5 10 17 26 Salinity (ppt) Nickel 7 . 2 6 . 9 6 . 3 5 . 8 5 . 5 2 . 9 1,5 p5 m 0 2 3 5 10 17 26 Salinity (ppt) Cadmium 7 . 2 6 . 4 5 . 6 5 4 3.1 0 . 5 0 2 3 5 10 17 26 Salinity (ppt) 0 2 3 5 10 17 26 Salinity (ppt) Zinc 1 1 2 1 1 0 0 0 9 7 9 9 5 2 7 1 3 4 8 7 1 7 9 w pa yg jm g • 0 2 3 5 10 17 26 Salinity (ppt) c m C u + 2 (777! Cu DOM c m Cu0H + 'otal 100 -CuCI+ 'otal 75 -irmm CuS0 4 (aq) h- 50 -szsa CuC0 3 (aq) o 25 -CuHC0 3 + 0 ™ Cu(C03)2"2 c m Ni ' r777i Ni DOM m n n N i c i + rxxa NiS04 (aq) n a N1CO3 (aq) NiHC0,+ C m Cd~ CdCI+ • M CdCI2 (aq) rrrrm CdS0 4 (aq) CZZ2 Cd DOM C Z Z 3 Pb DOM « = m Pb0H+ • • • Pbcr M B PbCI2 (aq) irrmn PbS04(aq) H i PbC0 3 (aq) ESSKRKS PbHC0 3 + Pb(C03)2"2 1 = 1 Zn^ C Z Z 2 Zn DOM ZnCI+ • • • ZnCI2 (aq) rmm ZnS0 4 (aq) EHM ZnC0 3 laq) B. DOM=20mg/L Copper 3 3 4 2 5 9 2 5 5 2 1 1 1 5 7 7 5 . 2 2 8 i 100 75 50 25 0 _ 100 ra -,c Z 75 t 50 = 25 * 0 0 2 3 5 10 17 26 Salinity (ppt) Nickel 7 . 2 6 . 9 6 . 3 5 8 5 . 5 2 . 9 1.5 ^ ^ 0 2 3 5 10 17 26 Salinity (ppt) Cadmium 7 . 2 6.4 5 . 6 5.4 3.1 0 . 5 0 2 3 5 10 17 26 Salinity (ppt) Lead 2 . 3 1.7 1.6 1.4 1 .3 1.2 1.1 0 2 3 5 10 17 26 Salinity (ppt) Zinc 1 1 2 1 1 0 0 0 9 7 9 9 5 2 7 1 3 4 8 7 1 7 9 2 n 7T\ l l i 0 2 3 5 10 17 26 Salinity (ppt) Figure 2.29. Speciation of Cu, N i , Cd, Pb, and Zn in mixing experiments calculated with the Visual MINTEQ computer program (Gustafsson, 2004) and using estimated values of dissolved organic matter (DOM) of A) 2 mg/L and B) 20 mg/L. Species with abundance less than 1% are not shown and metal organic complexes are shown as metal-DOM. Measured dissolved metal concentrations corresponding to each individual bar are given on the top axis of each subplot. 115 A. DOM = 2 mg/L B. DOM = 20 mg/L 0 20 40 60 80 100 120 140 160 Free C u 2 t (ug/L) 8 10 Free N i 2 + (ug/L) 0 1 2 3 4 5 6 7 Free C d 2 + (ug/L) T3 O CD Q 0 20 40 60 80 100 120 140 160 Free C u 2 + (ug/L) Free N i 2 + (ug/L) 2 3 4 5 6 Free C d 2 + (ug/L) Figure 2.30. Correlations of measured DGT-labile metal versus calculated free metal ion using estimated D O M values of A) 2 mg/L and B) 20 mg/L for Cu (top), N i (centre), and Cd (bottom). 116 A. DOM = 2 mg/L B. DOM = 20 mg/L Figure 2.31. Correlations of measured DGT-labile metal versus calculated total inorganic metals using estimated D O M values of A) 2 mg/L and B) 20 mg/L for Cu (top), N i (centre), and Cd (bottom). 11 3 o 3 o 1.2 1.0 0.8 -0.6 -04 0 2 0.0 20 40 60 80 100120 Distance (m) 5 10 15 20 25 30 Salinity (ppt) 1 2 1 0 0 8 0.6 0 4 0.2 0 0 20 40 60 80 100120 Distance (m) "O o o 1.2 0.6 0.4 0.2 0.0 1.0 0.8 N c N Q . .O 0-0.2 0.0 1.2 1.0 0.8 0.6 0.4 0.2 0 0 0 «-! 1.2 o m x 1 0 o © 0 g 0.6 • 0.2 o 0.0 0 20 40 60 80 100120 Distance (m) 1.2 8 - o g o 0 x 1.0 o c 0.8 N S 0.6 c N 0.4 0.2 0.0 20 40 60 80 100120 Distance (m) b° 0 9 o o 20 40 60 80 100120 Distance (m) 0.0 0.2 0.4 0.6 0.8 1.0 Salinity Based Mixing Index 1.2 o 1-0 o I r u 0.8 -I f 0 z 0.6 I ° O c P z 0.4 -o 0.2 -0.0 --5 10 15 20 25 30 Salinity (ppt) ° o 5 10 15 20 25 30 Salinity (ppt) 5 10 15 20 25 30 Salinity (ppt) 0.0 0.2 0.4 0.6 0.8 1.0 Salinity Based Mixing Index 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 0 2 0.4 0 6 0 8 1.0 Salinity Based Mixing Index 5 10 15 20 25 30 Salinity (ppt) 0.0 0.2 0.4 0.6 0.8 1.0 Salinity Based Mixing Index Figure 2.32. Proportion of dissolved metals at Philip stations relative to stormwater chamber levels as a function of A) distance, B) salinity, and C) a salinity based mixing index for three moderate/severe storms (November 22, 1999, September 7, 2000, and September 29, 2000). Solid symbols represent chamber concentrations, shaded symbols represent inner basin concentrations and open symbols represent outer basin concentrations. 118 I I Dissolved (ug/L) I 1 Particulate (ug/L) DGT-Labile (ug/L) Sept. 28, 2000 Sept. 29, 2000 Sept. 30, 2000 May 22, 2001 u> 600 £ 20 01 °? 15 ? 10 ° 0 2 3 0 | 2 0 ra <" 1 0 O 0 150 t t T T t f A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 i ' r T - T r T - i 4 h ^ r t r t 1.8e+4 | 1.5e+4 E 1.2e+4 <; 9.0e+3 fe 6.0e+3 i 3.0e+3 ° 0.0 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 r * ! l » l l " l T T i 1.2e+3 1.0e+3 8.0e+2 6.0e+2 4.0e+2 2.0e+2 0.0 oi 40 =* oi 30 i. 20 10 Q_ 0 0 •o O A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 T I I I I I I I I I I I — | 1 T 1 — I 1 1 I 1 1 I 1 1 I ' A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 1.2e+4 1 1.0e + 4 E. 8.0e + 3 2 6.0e + 3 fe 4.0e+3 2.0e+3 0.0 N n • I*! -r 1 — i — i — i — r T i A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 o 5 o o ® ® O O O O ^ » A A. A. O o o o ° • • • • • • " T r T T T 2.5e + 3 5 1 2.0e + 3 I5 1.5e+3 ^ 1.0e+3 fe 5.0e+2 £ 0.0 8 i 7 tt 6 " 5 _ 1000 | 800 g 600 ^ 400 fe 200 c 0 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 A O A 1 A 2 A 3 A 4 A 5 A 6 Figure 2.33. Metal, salinity, pH and SPM summary for sampling period September 28-30, 2000 at Philip site. Error bars represent ±1 standard deviation. 119 2.6. References Alfaro-Del la Torre, M.C. , Beaulieu, P.-Y., & Tessier, A . (2000). In situ measurement of trace metals in lakewater using the dialysis and DGT techniques. Anal. Chim. Acta, 418,53-68. Atkins, P.W. (1982). Physical Chemistry. 2nd ed. edn. Oxford University Press, Oxford, U K . pp. Avery Jr., G.B., Willey, J.D., & Kieber, R.J. (2003). Flux and bioavailability of Cape Fear River and rainwater dissolved organic carbon to Long Bay, southeastern United States. Global Biogeochem. Cy., 17, 1042. BC Research Corporation (1992). Urban Runoff Quality Control Guidelines for British Columbia. Prepared for: Municipal Waste Reduction Branch, Environmental Protection Division, BC Environment, Victoria, BC. BIEP (1992). Summary Report of Existing Environmental Quality Information. Prepared by the Environmental Quality Objectives/Monitoring Action Team(s). Burrard Inlet Environmental Action Program, Vancouver, BC. Buchman, M.F. (1999). N O A A Screeningick Reference Tables, Rep. No. N O A A H A Z M A T Report 99-1. Coastal Protection and Restoration Division, National Oceanic and Atmospheric Administration, Seattle WA. Buffle, J. (1988). Complexation Reactions in Aquatic Systems: An Analytical Approach Ellis Horwood Limited, Chichester, pp. Campbell, P.G.C. & LaZerte, B. (1988). Environmental Chemistry - General Overview. In Manganese in the Canadian Environment, (ed P .M. Stokes). National Research Council of Canada, Ottawa, Canada. C C M E (2003). Canadian Environment Quality Guidelines: Summary of Existing Canadian Environmental Quality Guidelines. Canadian Council of Ministers of the Environment. Cohn-Lee, R.G. & Cameron, D .M. (1992). Urban stormwater runoff contamination of the Chesapeake Bay: Sources and mitigation. Environ. Prof, 14, 10-27. Dai, M . , Martin, J.-M., & Cauwet, G. (1995). The significant role of colloids in the transport and transformation of organic carbon and associated trace metals (Cd, Cu and Ni) in the Rhone delta (France). Mar. Chem., 51, 159-175. Davis, A.P., Shokouhian, M . , & S., N . (2001). Loading estimates of lead, copper, cadmium, and zinc in urban runoff from specific sources. Chemosphere, 44, 997-1009. Davison, W., Zhang, H. , & Grime, G.W. (1994). Performance characteristics of gel probes used for measuring the chemistry of pore waters. Environ. Sci. Technol, 28, 1623-1632. Dorsey, N.E. (1940). Properties of ordinary water-substance Reinhold, New York. 673 pp. Downward, A.J. , Panther, J., Kim, Y . - C , & Powell, K.J. (2003). Lability of metal ion-fulvic acid complexes as probed by FIA and DGT: a comparative study. Anal. Chim. Acta, 499, 17-28. Dzombak, D. & Morel, F . M . M . (1990). Surface Complexation Modeling: Hydrous Ferric Oxide. Wiley-Interscience, New York. pp. El Samrani, A .G . , Lartiges, B.S., Ghanbaja, J., Yvon, J., & Kohler, A . (2004). Trace element carriers in combined sewer during dry and wet weather: an electron microscope investigation. Water Res., 38, 2063-2076. 120 EVS, Sea Science, & Frank Gobas Environmental Consulting (1996). Modelling the fate of contaminant discharges in Burrard Inlet. Burrard Inlet Environmental Action Program, North Vancouver, B.C., Vancouver, B.C. EVS (2003). Receiving Environment Monitoring Program for Stormwater Discharges within the Greater Vancouver Regional District. Prepared for the Greater Vancouver Regional District (GVRD), Burnaby, B.C. by EVS Environmental Consultants, North Vancouver, B.C. Fedrigo, A . M . & Boase, R. (1997). Sediment and Stormwater Chemistry of Oil-Grit Chambers. Department of Civil Engineering, University of British Columbia, Vancouver and District of North Vancouver, B.C. Field, R. & Turkeltaub, R. (1981). Urban runoff receiving water impacts: program overview. J. Environ. Eng. Div., 107, 83-100. Gustafsson, J.P. (2004). Visual MINTEQ. K T H , Dept. of Land and Water Resources Engineering, Stockholm, Sweden. Based on MINTEQ.A2 Ver. 4.0 and originally released in 1999 by C E A M , EPA, U.S.A. G V R D (1998). Characterization of stormwater runoff in three catchments. Prepared for Greater Vancouver Regional District by Morrow Environmental Consultants, Burnaby, B.C. G V R D (1999). Watershed Flows and Stormwater Runoff: Estimates for 1996 & 2036. Final Report. Greater Vancouver Regional District, Burnaby, B.C. Hall, K.J. , McCallum, D.W., Lee, K. , & MacDonald, R. (1998). Characterization and aquatic impacts of combined sewer overflows in Greater Vancouver, British Columbia. Water. Sci. Technol., 38, 9-14. Hudda-Musani, F. (2004). District of North Vancouver, B.C., North Vancouver, B.C. Kayhanian, M . , Singh, A. , Suverkropp, C , & Borroum, S. (2003). Impact of annual average daily traffic on highway runoff pollutant concentrations. J. Environ. Eng., 129, 975-990. Kobriger, N.P. & Geinopolos, A. (1984). Sources and migration of highway runoff pollutants, Rep. No. Rep. FHWA/RD-84/059 (PB86-227915). U.S. Dept. of Transp., Fed. Hwy. Admin., Washington, D.C. Lam, M.T., Murimboh, J., Hassan, N . M . , & Chakrabarti, C.L. (1999). Competitive ligand exchange/adsorptive cathodic stripping voltammetry (CLE/AdCSV) for kinetic studies of nickel speciation in aqueous environmental samples containing heterogeneous, macromolecular, organic complexants. Anal. Chim. Acta, 402, 195-209. Lavigne, J.A., Langford, C.H., & Mak, M.K.S . (1987). Kinetic study of speciation of nickel(II) bound to a fulvic acid. Anal. Chem., 59, 2616-2620. Leinonen, H. & Lehto, J. (2000). Ion-exhange of nickel by iminodiacetic acid chelating resin Chelex 100. React. Fund. Polym., 43, 1-6. L i , Y . - H . & Gregory, S. (1974). Diffusion of ions in sea water and in deep-sea sediments. Geochim. Cosmochim. Acta, 38, 703-714. Liu, D., Sansalone, J.J., & Cartledge, F.K. (2004). Adsorption characteristics of oxide oated buoyant media (ps<1.0) for storm water treatment. I: Batch equilibria and kinetics. J. Environ. Eng., 130, 374-382. 121 Luider, C D . , Crusius, J., Playle, R.C., & Curtis, P.J. (2004). Influence of natural organic matter source on copper speciation as demonstrated by Cu binding to fish gills, by ion selective electrode, and by DGT gel sampler. Environ. Sci. TechnoL, 38, 2865-2872. MacDonald, R.H., Kail , H. , & Schreier, H. (1997). Water Quality and Stormwater Contaminants in the Brunette River Watershed, British Columbia 1994/5. Prepared for the Fraser River Action Plan and the Fraser Pollution Abatement Office (FPAO), by the Westwater Research Unit, Institute of Resources and Environment, University of British Columbia. MacDonald, R.H. (2003). Estimated Urban Runoff Character and Contaminant Loadings in the GVRD. Prepared for the Regional Utilities Planning Division, Policy and Planning Department of the Greater Vancouver Regional District (GVRD), Burnaby, B.C. by Ron Macdonald, M A S c , P. Eng., Vancouver, B.C. Martino, M . , Turner, A. , Nimmo, M . , & Millard, G.E. (2002). Resuspension, reactivity and recycling of trace metals in the Mersey Estuary, U K . Mar. Chem., 77, 171-186. McCallum, D. (1995). An Examination of Trace Metal Contamination and Land Use in an Urban Watershed. M . A . S c , University of British Columbia, Vancouver, B.C. Morrison, G.M.P., Revitt, D .M. , Ellis, J.B., Svensson, G., & Balmer, P. (1988). Transport mechanisms and processes for metal species in a gullypot system. Water Res., 22, 1417-1427. Moss, M . (2001). Coastal Ocean Monitoring Program. Final Progress Report for N O A A Award #NA96RP0259 1 September 1999 to 31 May 2001. University of North Carolina at Wilmington. M W L A P (2001). Assessment of Burrard Inlet Water and Sediment Quality 2000. Prepared for the Ministry of Water, Land and Air Protection by Burke Phippen, B WP Consulting, Victoria, BC. Peters, A.J. , Zhang, H. , & Davison, W. (2003). Performance of the diffusive gradients in thin films technique for measurement of trace metals in low ionic strength freshwaters. Anal. Chim. Acta, 478, 237-244. Price, F. & Yonge, D. (1995). Enhancing contaminant removal in stormwater detention basins by coagulation. Transport. Res. Rec, 1483, 105-111. Randall, C.W. (1982). Stormwater Detention Ponds for Water Quality Control. In Stormwater Detention Facilities Planning. Design, Operation and Maintenance, (ed W. DeGroot). Engineering Foundation, Urban Water Resources Research, Council of the American Society of Civil Engineers, New York. Salvia, S. (2000). Application of Water-Quality Engineering Fundamentals to the Assessment of Stormwater Treatment Devices, pp. Project SCI 8-02, Final Technical Memorandum, Oakland, CA. Sangi, M.R., Halstead, M.J. , & Hunter, K . A . (2002). Use of the diffusion gradient thin film method to measure trace metals in fresh waters at low ionic strength. Anal. Chim. Acta, 456, 241-251. Sansalone, J., Koran, J., Smithson, J., & Buchberger, G. (1998). Physical characteristics of highway solids transported during rainfall. J. Environ. Eng., 124, 427-440. Sansalone, J. (2003). Physical and chemical characteristics of transportation land use runoff, Columbus, Ohio. Sansalone, J.J. & Buchberger, S.G. (1997). Partitioning and first flush of metals in urban roadway storm water. J. Environ. Eng., 123, 134-143. 122 Santschi, P.H., Lenhart, J.J., & Honeyman, B.D. (1997). Heterogeneous processes affecting trace contaminant distribution in estuaries: The role of natural organic matter. Mar. Chem., 58, 99-125. Schueler, T. (1987). Controlling Urban Runoff: A Practical Manual for Planning and Designing Urban BMPs., Metro Washington COG. Schueler, T. & Shepp, D. (1993). The Quality of Trapped Sediments and Pool Water Within Oil Grit Separators in Suburban MD. , Metro Washingto COG. Sekela, M . , Brewer, R , Tuominen, T., Sylvestre, S., & Moyle, G. (1998). Effect of a Rainfall Event on Contaminant Levels in the Brunette River Watershed (Data Report). Rep. No. DOE FRAP 1997-36. Aquatic and Atmospheric Sciences Division, Environmental Conservation Branch, Environment Canada, Pacific and Yokon Region, Vancouver, B.C. Shepp, D. (1995a). Petroleum Hydrocarbons in Urban Runoff: A Study Overview. Prepared for Maryland Dept. of Environment. Department of Environmental Programs. Metropolitan Washington Council of Governments., Washington, D.C. Shepp, D. (1995b). A Performance Assessmnt of an Oil-Grit Separator in Suburban Maryland. Final Report prepared for Maryland Department of Environment. Metropolitan Washinton Council of Governments., Washington, D.C. Sholkovitz, E.R. (1978). The flocculation of dissolved Fe, Mn, A l , Cu, N i , Co and Cd during estuarine mixing. Earth and Planetary Science Letters, 41, 77-86. Simpson, J.H. & Carr, H.J. (1958). Diffusion and nuclear spin relaxation in water. Phys. Rev., 2nd. Series, 1201-1202. Stone, M . & Marsalek, J. (1995). Trace metal composition and speciation in street sediment: Sault Ste. Marie, Canada. Water Air Soil. Poll, 83, 1-21. Stumm, W. & Morgan, J.J. (1996). Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, pp. 1022. John Wiley & Sons, Inc., New York. Swain, L .G. (1983). Stormwater Monitoring of a Residential Catchment Area. Resource Quality Section, Water Management Branch, B.C. Ministry of Environment, Vancouver, B.C. Thomson, R.E. (1981). Oceanography of the British Columbia coast, Ottawa. 291 pp. Twiss, R.R. & Moffett, J.W. (2002). Comparison of copper speciation in coastal marine waters measured using analytical voltammetry and diffusion gradient in thin-film techniques. Environ. Sci. Technol, 36, 1061-1068. Warnken, K.W., Zhang, H., & Davison, W. (2004). Performance characteristics of suspended particulate reagent-iminodiacetate as a binding agent for diffusive gradients in thin films. Anal. Chim. Acta, 508, 41-51. Wu, J. & Boyle, E.A. (1997). Low blank preconcentration technique for the determination of lead, copper, and cadmium in small-volume seawater samples by isotope dilution ICPMS. Anal. Chem., 69, 2464-2470. Xanthopoulos, C. & Hahn, H.H. (1990). Pollutants attached to particles from drainage areas. Sci. Total Environ., 93, 441-448. Zasoski, R.J. & Burau, R.G. (1988). Sorption and sorptive interaction of cadmium and zinc on hydrous manganese oxide. Soil Sci. Soc. Am. J., 52, 81-87. Zhang, H. & Davison, W. (1995). Performance characteristics of diffusion gradients in thin films for the insitu measurement of trace metals in aqueous solution. Anal. Chem., 67, 3391-3400. 123 Zhang, H. & Davison, W. (1999). Diffusional characteristics of hydrogels used in DGT DET techniques. Anal. Chim. Acta, 398, 329-340. Zhang, H. & Davison, W. (2000). Direct in situ measurements of labile inorganic and organically bound metal species in synthetic solutions and natural waters using diffusive gradients in thin films. Anal. Chem., 72, 4447-4457. CHAPTER 3 ASSESSMENT OF METAL DISTRIBUTION IN SEDIMENTS OF STORMWATER RUNOFF RECEIVING AREAS 125 3.1. Introduction Sediments can sequester large quantities of trace metals over long periods of time, and can act as metal sources rather than sinks to benthic and resident pelagic species when sediments are disturbed (e.g., storms, bioirrigation, dredging). Although chemical speciation provides only a small portion of free ions available for biotic absorption in the water column, quite a substantial portion can be available from sediments. Concentrations of metals within sediments usually exceed concentrations found within the water column and can be greater by a factor of 103 or more (Barron, 1995; 1992). The physical proximity to or nutritional dependence on sediments creates a precarious situation for benthic invertebrates, benthic algae, and other benthic biota in polluted systems. These organisms tend to reveal higher tissue-metal concentrations than planktonic or pelagic organisms (Campbell et al, 1988; Campbell & Tessier, 1991). Most studies investigating the relationship between concentrations of metals in sediments and bioavailability measure and use only those metal concentrations released from certain sediment fractions (exchangeable fraction, carbonates and hydrous metal oxide fractions, organic fraction). For organisms living in close proximity to sediments, the contribution from interstitial water is significant. While sediments may serve as a source of metal input, the interstitial water is in direct contact with organisms and partitioning into porewaters can play an important role in controlling the bioavailability of the free metal ion or other labile aqueous metal species. Metals can occur as dissolved or colloidal forms within sediment interstitial water or can be adsorbed at particle surfaces, associated with organic matter, or present in the lattice structure of minerals. Only a fraction of these forms is relevant in the study of biological impacts. The ionic and weakly complexed forms of most metals are considered to be the bioavailable and, in excess, toxic form to most organisms (Barron, 1995; Bryan & Langston, 1992; Tessier et al., 1984). Benthic detritivores and suspension feeders can ingest metals associated with sediment particles. Depending on the metals' binding capacity to the sediment, they can be released by the digestive process and absorbed into biological tissues. Competition between different binding sites can also prove to be critical. The presence and influence of these metal-binding sites are in turn associated with sediment composition, grain size, dissolved organic matter content (within porewater) and concentrations of various 126 ligands. Smaller grain sizes such as clay or silt matter, iron or manganese oxides (which scavenge trace metals), and increased organic carbon content have all been related to low bioavailability of trace metals (Sadiq, 1992; Timmermans, 1993). Bioavailability within sediments is also influenced by transformation reactions such as methylation, synergistic effects with other metals, and sediment chemistry, including salinity changes, redox state, and pH (Barron, 1995). While metal uptake depends on the organism's physiology, ecology, relative exposure, metal regulation in some instances and other relevant cellular processes, the potential exists for detrimental effects if biologically available metal is in excess. Sediments consist of several geochemical compartments with varying strengths of metal association. The strengths of these associations are determined by the metals involved, the inherent nature and relative size of the different sediment fractions, and other physicochemical features such as pH, ionic strength, redox potential and temperature (Campbell et al., 1988). The partitioning of metals between these different fractions can be used as a measure of the metal risk to biological organisms. The sediment compartments of an oxic environment consist of 1) the exchangeable fraction, consisting of weakly sorbed metals including metals adsorbed at specific sites and those associated by true ion-exchange with various solid substrates; 2) the carbonate-bound fraction, consisting of metals associated with sediment carbonates; 3) the iron and manganese oxide-bound fraction, consisting of metals associated with coatings, grains, precipitates or primary Fe/Mn oxide minerals; 4) the organic bound fraction, consisting of metals bound to various forms of organic matter including living organisms, detritus, humic and fulvic acids, coatings on mineral particles or other substrates; and 5) the residual fraction, consisting of silicate-bound metals (Campbell et al, 1988; Jenne & Luoma, 1977; Tessier et al, 1979). The residual fraction usually contains the largest proportion of metals; however, these metals are effectively unreactive and thus considered unavailable for biological uptake (Campbell et al, 1988). Fe and Mn oxides are major scavengers of trace metals and have been shown in many studies to reduce the bioavailable fraction of As, Cu, Pb, Zn, and Ag through a possible competitive effect (Bendell-Young et al, 1994; Bendell-Young & Harvey, 1992; Langston, 1980; Luoma & Bryan, 1979, 1982; Tessier et al, 1984). Similar results have been found for the organic sediment compartment for Pb, Zn, Co, and possibly Cu (Bendell-Young et al., 1994; Bendell-Young & Harvey, 1992; Luoma & Bryan, 1978, 1979; Tessier et al, 1984). 127 The bioaccumulation of cadmium has been correlated with the easily exchangeable and the easily reducible geochemical sediment fractions for Mytilus edulis (Rule & Alden III, 1990), but a much more complex model involving total sedimentary Cd concentrations, Fe oxyhydroxides, organic carbon concentrations and pH were required to describe bioaccumulation in the freshwater bivalve, Anodonta grandis (Tessier et al, 1993). The carbonate fraction may or may not be relevant to metal bioavailability depending on the abundance of fine grained carbonate particles and the particular chemistry of the sediment environment (Campbell et al, 1988). Sediment extraction techniques have been developed to help distinguish between these different sediment compartments. Non-selective, sequential and simultaneous extraction techniques have all been tried. In some instances, single leaching has provided better correlations with bioavailability than any other one extractant of a multi-extractant procedure (Campbell et al, 1988; Luoma & Bryan, 1981, 1982). Single leaching procedures use a single extractant such as dilute HC1 (0.5 or 1 N) to release most of the easily extractable metals (includes exchangeable, carbonate, Fe/Mn oxide and organic matter fractions) while leaving lattice-bound metals intact (Campbell et al, 1988). Numerous studies have demonstrated a relationship between biological uptake and one or more sediment fractions using a series of extraction steps (Bendell-Young et al, 1994; Bendell-Young & Harvey, 1992; Campbell & Tessier, 1991; Fan etal, 2002; Rule & Alden III, 1990; Tessier et al, 1984; Tessier et al, 1993; Thomas & Bendell-Young, 1998). Previous research has demonstrated that different sediment compartments may have more influence than others on bioavailability depending on the metal studied and the specific characteristics of the sediment involved (Bendell-Young & Harvey, 1992; Forstner, 1984; Tessier et al, 1984). The multi-extractant schemes are only applicable to oxic sediments since different geochemical controls exist within anoxic zones. These analytical procedures also operationally define rather than specifically determine the metal compartment and the reflected partitioning behaviour can be affected by the extraction sequence, the contact time between sediment and reagent, and the ratio of sediment to reagent, etc. (Pickering, 1981). Additionally, the available reagents are not as selective as desired, and overlap between identified phases is inevitable (Campbell & Tessier, 1991). 128 Monovalent labile sulfides in sediment or as they are operationally referred to, acid volatile sulfides (AVS), have the potential to influence metal bioavailability and toxicity due to metal binding within anoxic sulfide-bearing sediments and the subsequent unavailability to porewaters where biota exposures could occur (Chapman et al, 1998; Di Toro et al, 1992). Based on this principle, the molar ratio of the sum of simultaneously extracted metals (SEM) to A V S has been proposed (Di Toro et al, 1992) and demonstrated (Ankley et al, 1994; Ankley et al, 1991; Berry et al, 1996; Carlson et al, 1991; Cassas & Crecelius, 1994; DeWitt et al, 1996; Hansen et al, 1996; Liber et al, 1996; Pesch et al, 1995; Sibley et al, 1996) as indicative of potential toxicity when SEM: A V S >1 and generally lacking in toxicity when <1. However, recent studies using environmentally realistic metal concentrations have demonstrated a lack of correlation of metal bioaccumulation and/or toxicity with A V S and presented evidence that other geochemical considerations are relevant such as the availability of other metal ligands, biological exposure pathways such as ingestion, and/or sediment equilibration times (Lee et al, 2004; Lee et al, 2001; Simpson et al, 2000; Sundelin & Eriksson, 2001; Wiklund & Sundelin, 2002). Furthermore, Simpson et al. (2001), and Brumbaugh & Arms (1996) have demonstrated that S E M Cu is subject to erroneous and ambiguous interpretations due to artifacts of the A V S / S E M analytical procedure and in the case of the former, authors have suggested that A V S / S E M theory should only be applied to Cd, Pb, and Zn-contaminated sediments. At best, the SEM-AVS approach is limited by the complexities of ecotoxicological considerations, but may have limited usefulness under certain stable field conditions and for rapid screening of sediment quality (Lee et al, 2001; Simpson, 2001). A V S analyses in September 2000 of the present field program using the methods described in Allen et al. (1991; 1993) indicated only bottom sections of sediment cores (below 5-6 cm depth) at Philip stations A2 and A3 and Blueridge station C6 contained relatively low A V S concentrations of 1.3-3.6 umol /g and 0.07-0.53 umol/g at Philip and Blueridge respectively. Considering the availability of vast quantities of sulphates in the region (BIEP, 1992), these data may not have been representative of true conditions but reflective of losses due to sample storage time. Considering the above mentioned points and the dynamic nature of the basins as indicated by frequent resuspension events (Chapter 2), observed grain size distributions and variable surficial sediment redox conditions (this 129 Chapter), and the consequential alteration of relevant significant geochemical indicators of potential risk to biota (i.e. aerobic-Fe/Mn oxides, organic carbon vs. anaerobic-FeS, AVS) , A V S measurements were not pursued further nor reported in this study. Lability, bioavailability, and interstitial concentrations of trace metals in sediments and soils have also been assessed by the DGT technique (Hooda et al, 1999; Zhang et al, 1998; Zhang et al, 1995). This technique offers a potentially attractive monitoring tool due to its generally good performance in natural waters (Chapter 2) and relative ease of use. The nature of the chelating resin and the dilute concentrations normally found in interstitial waters provide a concentrating effect that enhances the normally difficult task of detecting and measuring porewater concentrations. As discussed in Chapter 2, oil/grit chambers can also serve as a source of metal enriched particulates to receiving waters. The characterization of sediment metals at the Philip (industrial/commercial) and Blueridge (residential/parkland) sites was sought using sediment extractions, porewater profiles and DGT. The intention was to provide a more complete assessment of metal contamination within the two sampling sites situated in areas of contrasting land-use patterns, and to further assist in the understanding of stormwater runoff and chamber control system influences on metal distributions within this study. Since surface sediments at the sample sites were often anoxic and any oxic layers when present were difficult to separate (coarse sediment textures and black oily colouration), only single extractions were performed on sediment samples using a weak acid ( I N HC1). This was in line with the goals of the study, which were to assess spatial and temporal differences in metal lability attributable to stormwater runoff and the oil/grit chamber system rather than the direct linkage of bioavailability to sediment geochemistry. 3.2. Materials and Methods 3.2.1. Field Program Sediment samples were collected in triplicate for trace metal analysis on several separate occasions from the Philip and Blueridge sites during 1998 to 2001. Sampling at 130 both sites occurred in January 1998 and May 2001 (surficial sediment) and in September 2000 (sediment cores). Sediment cores were also obtained from the inner basin stations at the Philip site alone in August 1998 and August 1999. Surficial sediments were collected from areas with higher deposition rates as determined by grain size analysis (elevated silt content). In different years, sediment core collections occurred in different sampling locations within each station. 1998 cores were sampled from central locations of the inner basin, 1999 cores were closer to the side banks, while 2000 cores were sampled across the width with one central core and two cores adjacent to either bank. Sediment cores (10 cm depth) were retrieved using butyrate core tubes (7-8 cm diameter), capped and transported to the laboratory in nitrogen filled double bags. Surficial sediments (0-2 cm depth) were scooped using polyethylene containers and stored therein. A l l sediments were transported on ice, and stored frozen until sample processing could begin (up to 6 months). Two of the three replicates collected in 1999 were lost which left only one set of cores per station for analysis. In January 1998, surface sediments were sampled at an additional oil/grit chamber receiving area close to the Philip site. This site was excluded from further evaluations due to the anomalous nature of sediments within the chamber and receiving area caused by the overwhelming presence of wood chips from an adjacent wood processing facility. During the summers of 1998 and 1999, porewater depth profiles were obtained using plexiglass porewater peepers prior to core collections and within 2 m of where sediment cores were obtained. The porewater peepers were constructed at Simon Fraser University, Burnaby and based on the design of Carignan (1985). DGT sediment probes were simultaneously utilized alongside the porewater peepers during these two occasions to allow comparison of the two techniques with respect to trace metals. Prior to porewater peeper retrieval from the field, DGT units were deployed (in triplicate) for 24 hours within sediments at corresponding stations. DGT units were inserted within 0.5 m of peepers at each station. Similar DGT sediment probes were also subsequently deployed in September 2000 for depth metal profiles. Piston style disc DGT units (same as those used for water sampling) were used for sediment surface metal measurements in May 2001. Ancillary measurements during field sampling included temperature, pH and salinity. Grain size analysis was conducted on separate sediment cores (0-10 cm depth) obtained during the 2000 sediment collection and surface sediments (0-5 cm) in March and August 1998. Due to the 131 unconsolidated nature of the sediments within the oil/grit chamber, core sampling was not possible and sediments from this station in 2000 were obtained using a plastic scoop. 3.2.2. Sediments Sediment core processing was performed within a nitrogen-filled glove-bag. Once cores were thawed, they were separated into six layers by careful extrusion. A plexiglass disc fitted within the bottom of each core tube was designed to slide through each core tube at specified intervals through the use of sectioned plexiglass rods (each cut to extrude the six desired sediment layers). Cores were sectioned from top to bottom in increasing thicknesses (0.5 cm, 1 cm, 1.5 cm, 1.5 cm 2.5 cm, and 3 cm) with the expectation that subsequent decreasing resolution coincides with decreasing porosity and change in metal concentrations. The perimeter of each sediment layer, which can be distorted during coring, was discarded. The remaining sediment was mixed, divided into separate aliquots for each different analysis and, with the exception of acid extractable metals and acid volatile sulfides, stored frozen in acid clean vials for later analysis. Only plastic or Teflon utensils were used for sediment handling. A l l plasticware and Teflon used for sediment sampling and processing were acid washed using the protocol described in Chapter 2. Surface sediments and cores were analyzed for total metals, cold acid extractable metals, and total and inorganic carbon. 3.2.3. Trace metals 3.2.3.1. Total metals Total metals were determined by the complete digestion of sediment samples. Frozen subsamples, in polyethylene vials with caps ajar, were freeze-dried in a high vacuum Edwards 4K Modulyo freeze dryer. Dried samples were weighed for moisture content and then ground to 200 mesh with a Herzog HSM-100 tungsten-carbide mill and stored sealed in new vials. Accurately weighed aliquots of powdered samples (10-20 mg) were transferred to 10 mL Teflon tubes and placed in Teflon vessels (3 tubes per vessel) for microwave 132 digestion prior to analysis by ICP-MS. The digestion and metal analysis protocol is described in Chapter 2. Replicate analysis of within core samples generally produced relative standard deviations (%RSDs) of <25% except at the low and high ends of the concentration ranges where %RSDs were elevated and as high as 80%. The precision of sediment microwave digestion was assessed to be quite good with %RSDs (n=10) of 5.6%, 2.6%, 4.2%, 4.2%, and 3.6% for N i , Cu, Zn, Cd, and Pb respectively. To test for possible contamination, the marine sediment reference material (PACS-2) samples were run through the grinding mill in the same manner as field sediments. Analysis of these PACS-2 samples, as well as those microwave-digested directly, agreed well with certified values (Table 3.1). Analysis of blanks and detections are also given. 3.2.3.2. Acid extractable metals Acid extractable metals were determined with a cold weak acid (1 N hydrochloric acid) known to extract all but lattice bound metals in sediments (Campbell et al, 1988). Wet sediment samples were weighed (~2 g) and placed in 40 mL polyethylene vials. 20 mL of 1 N HC1 was added to samples which were then capped and allowed to sit overnight with occasional shaking. The extract solution from each sample was decanted into individual 50 mL centrifuge tubes. Samples were centrifuged and 1 mL aliquots of particulate free solution were transferred to 10 mL Teflon tubes. Samples were evaporated to dryness under a laminar flow fumehood and then reconstituted with 1.5 mL Seastar HNO3. Samples were transferred to 1.5 mL microtubes and stored until analyzed for metals by ICP-MS. 3.2.4. Carbon Organic carbon was derived indirectly as the difference between total and carbonate carbon in each sediment sample. Total carbon was measured gas-chromatographically on a Carlo-Erba CNS analyzer Model N A 1500 and a carbonate coulometer UIC model CM5014. Ground dried sediment samples (20-30 mg) were accurately weighed into small tin cups which were then crimped shut before being loaded into a sample carousel. Samples were 133 combusted at 1000°C and the resulting gases separated chromatographically and detected by thermal conductivity. Instrument calibration was performed using NRC marine sediment reference materials MESS-1 and BCSS-1 and sulfaniliamide standards. Each carousel consisted of 39 samples, 2 blanks, 4 marine sediment reference material samples (MESS-1 and BCSS-1), and 5 sulfanilamide standards. Recoveries were over 99% for sediment reference materials (Table 3.2) and precision (n=15) was estimated in %RSD as 3.16%. Carbonate carbon was determined by coulometry on a Coulometrics 5010 coulometer and organic carbon was calculated by subtraction from total-carbon as measured above. Coulometry involves the conversion of carbonates into C 0 2 with the addition of 10% HC1. 50-100 mg dried ground samples were accurately weighed and placed into clean glass reaction tubes. Each tube was flushed with C02-free air for two minutes for the removal of residual atmospheric C O 2 . Tubes were lowered into the instrument heating block and a few mL of 10% HC1 was added. CCVfree air was bubbled through the heated mixture which liberated and carried the evolved C O 2 through a blue ethanolamine solution, causing a rapid reaction and change in colour of the solution from blue to clear. A photo-detector measures the change in transmittance which leads to generation of OH" ions after reduction of water at a silver electrode. The proportionality of carbonate-carbon to the change in current required for the return to the original transmittance of the ethanolamine solution is used to determine the quantity of carbonate-carbon in each sample. Instrument calibration was achieved using high purity CaC0 3 . Replicate analysis of the C a C 0 3 standard (12.000% carbonate carbon) achieved a measured value of 11.95 ±0.05% carbonate carbon (n=14). The detection limit was estimated to be 5.86 pg carbonate carbon (5a of the blank, n=7). Method and instrument precision were measured by replicate analysis of sediment samples and CaC03 and determined to be 5.33% and 0.41% respectively. 3.2.5. Porewaters Porewater metals were obtained by using the "in situ" dialysis technique with plexiglass peepers modeled after the methods of Carignan et al. (1985). Each porewater peeper was 60 cm long and 8 cm wide with separate 4 mL wells spaced 1 cm apart 134 lengthwise. The wells were filled with nanopure water, covered with an acid-clean 0.2 pm pore size polysulfone filter membrane (Gelman HT-200) and secured with a plexiglass cover whose apertures matched each well opening. The peepers were then deoxygenated for a minimum of 24 hours within an anaerobic chamber consisting of nanopure water continuously bubbled with nitrogen. Peepers were transported to the field in nitrogen filled double bags. At each station, triplicate peepers were inserted vertically into the sediment (within 1 m of each other) for the maximum depth possible (20-40 cm) and left to equilibrate for approximately four weeks in 1998 and two and a half weeks in 1999. Upon retrieval, peepers were rinsed with nanopure water to remove sediment and debris, kept horizontal and cool, placed in nitrogen filled double bags, and carefully transported to the lab for porewater removal (within 45 minutes). Using acid clean polyethylene transfer pipettes (one for each sample), the filter membrane was punctured and well waters were removed and immediately pipetted into acid-clean polyethylene tubes containing 20 pL Seastar 70% H N O 3 . Samples were stored in capped tubes at 4°C until analysis by ICP-MS. 3.2.6. DGT 3.2.6.1. Theory DGT was developed initially for use in natural waters and the basic theory developed by Zhang & Davison (1995) as described in Chapter 2 remains essentially the same; however, the constraints of the sediment environment (e.g. lack of turbulence) introduces a few different aspects that are considered herein. Whether the DGT device is measuring porewater concentrations or some measure of metal flux depends on the particular metal, its interstitial concentration, and the binding capacity of the substrates within the sediment (Zhang et al, 1995). An important consideration to note is that within sediments, metal transport will generally be governed by molecular diffusion which can become a limiting factor. The DGT unit, by its nature, perturbs the sediment environment by introducing a metal sink i.e. chelex resin, within which metals swiftly accumulate and are in turn depleted from interstitial waters. There are three different representations described by Zhang et al. (1998) in which the metal flux to DGT relates to metals in soil or sediments: (1) Fully 135 sustained: Metal ion removal from interstitial waters by the DGT device are rapidly resupplied through diffusion or desorption processes from solid phases creating a quasi 'steady-state' condition in which concentrations and fluxes to DGT can be calculated using Chapter 2 Eqns. 6 and 3 respectively. In this case, the DGT flux is lower than the maximum potential flux of solid to solution within porewaters. (2) Unsustained: Metals are not resupplied from solids to solution. Transport is governed by molecular diffusion alone and porewater metals decline over the time of DGT deployment. (3) Partially sustained: Some resupply from solid to solution phase occurs but is inadequate for the maintenance of initial porewater concentrations. Comparison of DGT with other porewater measurement techniques can be performed in order to distinguish between the three cases. Additionally, the hydrodynamics of the bulk solution can be controlled by selection of variable diffusive gel thicknesses, which changes the degree of perturbation to the sediment environment or DGT demand and effectively controls the overall rate of mass transport (Zhang et al, 1995). Thinner gels lead to an increased demand and quicker porewater metal depletion. 3.2.6.2. Field Application The DGT sediment probes (15 x 5 cm dimensions) were constructed out of plexiglass and consisted of a back plate and cover with a 13 x 1 cm central window open for diffusion (Fig. 3.1). Similar to the water units described in the preceding chapter, a chelating resin gel layer was topped with a diffusive gel, followed by a 0.45 pm polyethersulfone membrane, the open top plate, and secured with plastic screws. The units were placed in polyethylene containers filled with nanopure water and deoxygenated with continuously bubbled nitrogen for 24 hours. They were sealed in these deaerated containers, transported to the field and vertically inserted into sediments lengthwise. They were removed from sediments after 24 hours, rinsed with nanopure water, and transported to the laboratory in acid clean plastic containers. Two DGT probes used as blanks were similarly prepared and transported between field and laboratory. Under a laminar flow hood, all DGT units were disassembled. The exposed section of the resin layer sitting directly beneath the diffusion window was sliced into 1 cm squares with an acid-clean plastic knife and placed individually into acid-clean microtubes. Metals were eluted with 1 M HNO3 and analyzed by ICP-MS as detailed 136 in Chapter 2. Calculations using Eqns.2 through 6 of Chapter 2 were used to determine mean metal fluxes to the DGT resin over the sediment deployment time alongside potential concentration interpretations which could be compared to peeper porewater concentrations. 3.2.7. Grain size Grain size distribution analysis of sediments was conducted for the purpose of quantifying the relative proportions of each of the main sand, silt and (for representative samples) clay fractions. A dry sieve-based grain size analysis on surface sediments (0-5 cm) was conducted in 1998 in order to establish sediment deposition patterns for future sampling. In September 2000, the grain size protocol for sediment cores was refined and extended in order to improve accuracy for assessment purposes. The new protocol was based on the method of Folk (1968) and consisted of wet sieving for coarse fractions followed by the pipette method for the separation of silt and clay. The pipette method uses the settling property of particles based on Stokes' Law, and measures the percent by weight of each particle size class per sample. While assumptions of this method (equal density and perfect spherosity for all particles) cannot be completely satisfied, it was assumed adequate for the determination of the approximate distribution of grain sizes. Cores sampled for grain size analysis were sectioned as described in Section 3.2.2 for the same depth layers used for metal and carbon analysis. Approximately 15 g of each sediment layer was placed in a beaker with 100 mL deionized water where samples were disaggregated with a glass stirring rod. As recommended by Menking et al. (1993), for the removal of carbonate and organic material, 150 mL of Morgan's solution (27 mL glacial acetic acid buffered to pH 5 with 82 g sodium acetate) and 20 mL 30% H2O2 was added to each sample and allowed to stand overnight, stirring once every 8-12 hours. Samples were then heated for 2 to 4 hours at 150 °C to remove excess H2O2. Samples were decanted and wet-sieved to separate gravel (>2 mm diameter), sand (2-0.064 mm), and silt and clay (<63pm). Gravel and sand fractions were dried at 60°C and separated using a nested sieve mechanical shaker. The silt and clay fractions were poured into 1 L cylinders. A dispersant was added (5 mL 5% sodium hexa-metaphosphate) to prevent flocculation and solutions were made up to 1000 mL. Each solution was vigorously stirred for 2 minutes, then allowed to settle for 20 seconds at which 137 time 20 mL was removed by pipette at a 20 cm depth from the water surface. The 20 mL aliquots were placed in preweighed aluminum dishes, and allowed to dry in a heated oven at 60°C. The dried mass remaining was accurately weighed, corrected for added salt, and multiplied by 50 to obtain the total mass of silt and clay within the original cylinder. An additional aliquot was pipetted after 3 hours and 33 minutes at a 20 cm depth to obtain the clay fraction (<0.002 pm). 3.2.8. Analytical Program 3.2.8.1. Reagents ICP-MS metal stock and working solutions were prepared using nanopure water and ultrapure grade concentrated nitric acid (Seastar Chemicals Inc., Sidney, BC). Concentrated nitric, hydrochloric, and hydrofluoric acids used for digestions were environmental grade (Anachemia Science, Richmond BC). A l l plasticware was acid cleaned as described in Chapter 2. 3.2.8.2. Instrumental Metal Analysis Analyses of total, acid extractable, and DGT-labile metals in sediments was performed on the Element2 HP-ICP-MS (ThermoFinnigan Element2 High Performance ICP-MS , Finnigan-MAT, Mississauga, Canada) as previously described in Chapter 2. Metals in porewaters were analyzed on the same instrument using a standard additions method for calibration in order to minimize matrix interferences from dissolved salts. A minimum 15-fold dilution was determined through calculation and instrument response (depression of internal standard and argon peak intensities) to be adequate for the reduction of dissolved solids to the maximum capacity of the HP ICP-MS (<0.2%). Assuming the salinity along a peeper would not be extremely variable along its 20-40 cm depth and metals would follow a simple curvilinear profile, selected samples from along the depth of each porewater peeper were chosen to assess which samples had the lowest barely detectable metal concentrations. 138 One sample per peeper was selected as the matrix-matched blank to which standards were added for calibration. Each subsequent porewater sample from that peeper was analyzed in relation to its specific peeper blank. Dilution factors were kept constant within each calibration set and instrument rinses were increased to prevent salt build-up within instrument components. 3.3. Results 3.3.1. Grain Size Grain size analysis conducted on sediment samples across the width of the Philip inner basin from the east to west banks generally revealed sandy sediments (>50% sand, mean grain diameter range 0.190-2.24 mm) with increased silt content at the centre of the inner basin in 1998 and outer banks in 2000 (Table 3.3, Figs. 3.2 and 3.3). Sediment texture was finer in the outer basin in 2000 (9.73-16.1% silt, 0.61-1.18 mm) but not 1998 (0.45-6.51% silt, 0.53-2.45 mm). Sediments were coarser within the Philip chamber (2.37-3.37 mm) than within the inner or outer basins (0.19-3.45 mm) during each of the three sampling occasions. The Blueridge site consisted of higher silt content at the inner station (59-89% silt) but exhibited grain size distributions that were similar to the Philip site at its two more seaward stations (<6% silt). 3.3.2. Organic Carbon Organic carbon varied between 0.201 and 13.8% at Philip and Blueridge across sediment depth and years (Fig. 3.4). Depth profiles showed different patterns between years and along the length of the inner basin at the Philip site. Differences between the two earlier years were attributed to spatial differences between the central line of the inner basin (higher silt content) sampled in 1999 and the coarser sediment texture of the outer edges sampled in 1998. September 2000 cores, however, may have been influenced by the first season's severe storm that occurred two weeks prior to sampling. The basin is small enough to have 139 been scoured of fine particles and associated organic material with this storm and indications of such activity were revealed in Chapter 2. In August 1999, the two stations closest to the oil/grit chamber had the highest subsurface maximum (1-2.5 cm) of between 6 and 8% organic carbon and a sharp decline to 2% with depth. In August 1998, station A l organic carbon declined slightly from - 3 % at 0-2 cm depth to - 1 % by 4 cm depth while station A2 showed a reverse trend with organic carbon at ~1% at the surface and increasing to just over 2% at depth with the exception of the deepest sediment layer. The most seaward inner basin station (A3) core profile in 1999 was lower than in 1998 (-2-9%). Station replicates showed large but inconsistent variability at station A3 in both 1998 and 2000 which was not tested in 1999 with the analysis of only one sample per station (Fig. 3.5). A l l three inner basin stations (A1-A3) in 2000 revealed low organic carbon content (<2%) with the absence of any distinct peaks. Organic carbon from the outer basin station A4 (sampled in 2000) was relatively high compared to the inner basin stations but showed greater variability with no distinct peaks or patterns between station replicates. The Blueridge site (also sampled in 2000) showed a distinction between the two mid stations (C4 and C5) with low uniform organic carbon content with depth, and the two stations lying at either end of the receiving basin (C3-closest to the chamber outlet and C6-the most seaward station) with high organic carbon (—3-14%), large variability and slight subsurface peaks. As expected, moisture content profiles (Fig. 3.6) reveal the same patterns seen for organic carbon and silt content. General field observations of higher organic debris at Blueridge station C3 and the Philip station A3 seem to be reflected for the most part in the organic carbon determinations. 3.3.3. Total and Acid-Extractable Metals Total Cu, Zn, Cd, Pb and N i in sediments followed similar patterns to each other over years, stations, and sediment depth (Figs. 3.7-3.16). Some spatial variability was noted within station locations (Figs. 3.17-3.21) however correlations were apparent between all elements (less for Ni) (Fig. 3.22 and Fig. 3.23) indicating a possible common source and/or similar geochemical controls. 140 Concentrations of total metals in surface sediments at the Philip site, sampled over all years and excluding chamber sediments, ranged over one order of magnitude for Cu (307-2260 mg/kg) and Pb (78.5-612 mg/kg), and two orders of magnitude for Zn (248-18,300 mg/kg) and Cd (0.47-48.2 mg/kg). N i concentrations varied over a narrow 2-fold range (15.7-39.9 mg/kg). For the 2000/01 sampling periods, lower mean maximum concentrations for Cu and Cd produced narrower but still substantial concentration ranges of 307-1370 and 0.47-15.5 mg/kg respectively. Zn concentrations were substantially reduced with a range of 248-938 mg/kg. The maximum mean Pb concentration occurred in 2001 and represented a 2-fold increase over the maximum concentration of 381 mg/kg for 1998-99. Surface sediments sampled within the chamber in 2001 contained the lowest total Cu (74 mg/kg), Zn (136 mg/kg) and Cd (0.31 mg/kg) concentrations and represent a 5-fold decrease for Cu and 2-fold decreases for Zn and Cd over lower range concentrations measured within Philip receiving basins. Sediments at depths of 1.5-10 cm produced similarly broad ranges for Cu (202-3,030 mg/kg), Pb (87.0-612 mg/kg), Zn (281-23,300 mg/kg) and Cd (0.82-57.5 mg/kg). N i concentrations reflected similarly narrow surface range of 16.4-45.9 mg/kg with variable metal depth profiles between stations and years. In 1998, a general decrease of metal concentrations at sediment surfaces is seen with increasing sediment depth at station A l whereas at stations A2 and A3 metal concentrations tend to increase with maximums at bottom sediment depths. In 1999, absolute concentrations were greater at both stations A l and A2 compared to 1998; however profiles were similar to station A l in 1998 with decreasing metal concentrations from surface maximums. At station A3 in 1999, metal concentrations peak at depth for Cu, Pb, and Cd, while both a subsurface peak (2 cm) and lower core peak (8cm) was observed for Zn. It should be recalled that only one core was available for analysis at each station in 1999 so any spatial variability was undetermined. In 2000, sediment concentration changes with depth are reduced compared to the previous two years. Similar to 1998 however, is the greater spatial variability and observance of subsurface peaks at station A3 seen with the core replicates presented in Figs. 3.17-3.21. This is also observed within the outer basin station A4 sampled in 2000. Otherwise, surface sediment concentration ranges are somewhat similar between stations for all elements. 141 Blueridge site surface sediments (Fig. 3.25) sampled in 2000 and 2001 were markedly lower than Philip sediments for total Cu (19.6-108 mg/kg), Zn 71-423 mg/kg), Cd 0.09-0.92 mg/kg), and Pb (16.2-66.6 mg/kg). Similar low concentrations were determined for 2001 deeper core sediments with concentration ranges of 15.3-134 mg/kg (Cu), 116-493 mg/kg (Zn), 0.02-0.81 mg/kg (Cd), and 12.5-92.1 mg/kg (Pb). Total N i within both surface (11.4-27.7 mg/kg) and deeper core sediments (8.0-39.4 mg/kg) at Blueridge were only slightly reduced compared to the Philip site. Some degree of spatial variability was seen for all elements in at least one of the four stations with the largest concentration range observed at stations C4 and C6 for Cu, Zn, Cu and N i ; at station C6 for Pb, and stations C5 and C6 for Cd. Unlike Philip chamber sediments, total Cu, Cd, and Pb were all moderately elevated relative to receiving basin sediments. Acid extractable metal concentrations showed more variation between elements with a much larger range for Cu, spanning 4 orders of magnitude. The general metal patterns, however, were similar to those observed for total metals (Figs. 3.7-3.16). Metal concentration ranges of sediment surface and core depths were similar. Overall ranges at the Philip site were 0.02-264 mg/kg (Cu), 85-15,200 mg/kg (Zn), 0.09-49.3 (Cd), 1.5-9.3 (Ni) and 46.0-569 mg/kg (Pb). Chamber sediments at Philip were again somewhat reduced for Cu (11.7 mg/kg) and Zn (38.5 mg/kg) and within the low range for Cd (0.06 mg/kg), N i (2.4 mg/kg) and Pb (48.4 mg/kg). A certain amount of spatial variation is seen for all stations in all years; however, the same slight enrichment of Cu, Zn, Cd and Pb at stations A3 and A4 is seen in 2000. Blueridge concentrations were decreased with ranges of 3.6-48.0 mg/kg (Cu), 26-204 (Zn), 0.02-0.36 mg/kg (Cd), 2.1-8.8 mg/kg (Ni), and 4.1-62.8 mg/kg (Pb). The highest metal concentrations primarily occurred in surficial sediments. Concentrations of total and acid extractable metals within surface oxic sediments (i.e. excluding 1999) showed no significant decreases from 1998 to 2001 for Cu, Cd, and N i (Figs. 3.24, 3.25, and 3.7-16). The increase in Pb in 2001 (3-fold) and decreases in Zn from 1998 to 2001 (3-fold) at Philip showed an inverse proportionality and may have been related to higher Fe/Mn oxides (20% and 3-fold decreases respectively), spatial/temporal variability in sediments or runoff sources, or other geochemical considerations rather than as a result of chamber influences. This is highlighted further by the lack of element inter-correspondence in this regard. 142 Considerable variability across stations, sites, years and between elements was seen in the proportion of trace metals extracted from sediments (Figs 3.7-16). Extractable Cu ranged between near 0-33% in sediments of 1998/99, 12-46%) in sediments at Philip and 4-100%o at Blueridge in 2000/01. Extractable Cu in 1999 was lower in core depths and generally stayed below 25%. Similarly, extractable N i ranged between 14% and 37% across depths at Philip in 1999-2001 with a slightly lower percentage range in 1998 (7.5-17%). Blueridge percentages were similar at 14-57% in 2000/01. Comparatively, Zn, Cd, and Pb all exhibited elevated ranges at Philip with 17-100+%, 14-97%, and 18-100+% respectively. Blueridge percentages were similar at 11-95%, 30-100+% (with one sample at 3% within C5 core), and 27-100%) respectively. 3.3.4. Porewater Metals Porewater profiles of dissolved Mn and Fe obtained in 1998 and 1999 are presented in Figs. 3.26 and 3.27 respectively. The porewater profiles of M n 2 + indicate low concentrations with depth at all three stations within the Philip inner basin in 1998 (<0.2 mg/L). A sediment surface enrichment of both Fe 2 + and M n 2 + in one of the replicates at station A l , and Fe 2 + alone in two replicates, indicates a reducing environment for Fe and Mn and suboxic conditions. At station A3, suboxic conditions may be occurring at or just below the sediment-water interface as indicated by subsurface peaks in two of three replicates. At station A2, suboxic conditions may be occurring at about 5 cm with the increase in dissolved M n indicative of Mn reduction. The dissolved Fe profile appears somewhat scattered at this station and may reflect sampling artifacts. M n reduction and production of dissolved M n 2 + usually precedes Fe reduction and resulting increases in Fe 2 + . The apparent simultaneous processes of Fe and M n reduction may be attributable to the combination of anoxic conditions, low availability of both Mn and Fe oxides (~2-5 mg/g, as estimated by cold acid extraction, Appendix A , Table A2), and large availability of sulphates (typical of marine systems) which could conceivably result in Fe 2 + consumption with Fe-sulfide precipitation within shallow depths and result in overlapping Fe and Mn peaks. 143 Trace metal porewater profiles for 1998 and 1999 are reported in Figs 3.28 to 3.32. In 1998, concentrations of all five trace metals in porewaters were enriched (increases <l-2 orders of magnitude) at subsurface sediment depths compared to water concentrations reported in Chapter 2. Spatial variability within station locations indicate comparisons between stations should be treated cautiously. Nonetheless, the metal enrichment within surficial sediments is suggestive of metal release with remineralization of organic carbon which is in line with the dissolved Fe and Mn profiles and anoxic conditions discussed above. The trace metal porewater profiles for Cu, Zn, and Cd are all somewhat similar in 1998 with decreasing concentrations from bottom waters to surficial sediment (at approximately 2 cm depth). The downward trace metal decline reflected in the Fe 2 + decline supports the suggestion of not only increased Fe-sulfide precipitation but also the co-occurrence of trace metal sulfide precipitation. A l l four elements show a subsurface maximum below 10 cm depth at station A2. The Fe 2 + and M n 2 + concentrations at station A2 also appear elevated at around this depth which may indicate a pocket of organic matter oxidation. Dissolved N i is somewhat anomalous with an unusually high (-60 ug/L) subsurface peak at station A l at approximately 15 cm for one station replicate. Dissolved Pb shows a similar subsurface maximum (~30 ug/L) to the other trace metals at about 2-5 cm for Station A l for two of the three replicates. Only stations A l and A3 were examined for porewaters in 1999. M n 2 + profiles at station A l showed only the slightest increase at just below a 5 cm sediment depth which was quickly followed by an increase in Fe 2 + at around 8 cm. Spatial variability was higher at station A3 than A l , with one replicate showing Mn increases by about 5 cm and the other below 10 cm. Once again, Fe 2 + at this station followed closely with maximums at 8 and 12 cm respectively. 1999 porewater profiles of dissolved Cu, Zn, and Cd at both stations all displayed similar patterns of declining concentrations from bottom waters to 8 cm sediment depth at station A l and 2-5 cm at station A3. These profiles are consistent with a progressive reducing environment and anoxia with resulting dissolved metal depletion and increased precipitation of metal sulfides. This same subsurface decline is not seen for Pb or N i at either station. Pb shows a subsurface maximum at about 10 cm (-20 ug/L). Dissolved N i primarily varies between below detection to 10 ug/L and shows variability and marked 144 scatter between depths at station A l . Station A3 displays the same deep sediment concentration peak for all five trace metals at approximately 25 cm depth. 3.3.5. DGT Metals In order to determine exactly how the DGT trace metal data should be interpreted (i.e. flux or concentrations), DGT fluxes for N i , Cu, and Cd are presented together with mean porewater concentrations for 1998 (Figs. 3.33-3.35) and 1999 (Figs. 3.36-3.38). The graphical correspondence between the dissolved and DGT-flux metal scales is arranged so that the dissolved metal scales also function as the interpreted DGT-labile metal scale i f porewater concentrations are not depleted or diminished during the DGT deployment period. DGT Zn and Pb concentrations are not reported due to the poor recoveries obtained in calibration experiments (discussed previously in Chapter 2). Initially in 1998, only the larger gel thickness was used (G2=0.08 cm) since it provides a longer diffusion distance to the DGT resin and consequently slows the potential metal depletion from porewaters. DGT-Cu fluxes remained fairly uniform within the measured 12 cm sediment depth profile. Surface DGT concentration estimates were in close agreement with porewater concentrations at stations A l and A2 but the DGT replicate mean reflected a 50% underestimation of porewater Cu at station A3. DGT overestimated porewater concentrations below 2-4 cm depth with the exception of station A2 where the porewater concentrations intersected with DGT estimates at the 12 cm sediment depth. DGT N i variability within surface sediments (stations A l and A3), underestimation of porewater concentrations at station A2, and overestimation at various points along the sediment depth profiles at all three stations may be a result of both low N i concentrations and spatial heterogeneity. DGT-Cd shows large variability between DGT replicates, and while concentration estimates overlap porewater concentrations at many points, a general conclusion cannot be drawn. In 1999, when two diffusive gel thicknesses (Gl=0.062 cm and G2=0.080 cm) were utilized, DGT fluxes provided closer estimates for interpreted concentrations for Cu than Cd and both were better than N i . DGT-Cu concentration estimates of both gel sizes were equal 145 to each other and with porewater concentrations within surficial sediments (0-2 cm) for station A l and within 30% of each other and porewater concentrations at station A3. At lower depths, increased variability was seen between the two gel thicknesses along with general lack of correspondence between DGT estimates and porewater concentrations. Large variances which appear between same size gel replicates likely indicate the high level of lateral spatial variability for these proximal locations. In September 2000, DGT measurements revealed fluxes for N i and Cd that were very similar across sampled stations at the Philip site (station AO to A4, Fig. 3.39). Replicate variability was high and trends were difficult to discern other than the higher mean flux of DGT-Cu at station A l relative to the other three stations. Blueridge, in relation to Philip DGT metal fluxes in surface sediments, were lower for Cu and Cd and within the same range for N i . The May 2001 DGT surface sediment metal fluxes were similar to the 2000 fluxes and with the exception of possibly Cd, similar between sites (Fig. 3.40). Cu fluxes in 2001 were elevated at station A l but otherwise revealed similar values to Blueridge. The Philip outer basin and two most seaward stations at Blueridge were similar and slightly lower than other stations within the respective sites. Cd fluxes were low at Philip and even lower at Blueridge. 3.4. Discussion Grain size distributions which show coarse grain sizes only within the chamber, reveal the lack of particle retention within the Philip chamber system and the heterogeneity of sediments across sites, stations, and years reflecting the dynamic nature of water flow and sedimentation processes resulting from storm-related activities within this restricted geographic area. This heterogeneity extends to sediment metal concentrations and geochemical properties and further suggests that the constant supply of storm particulate depositions and storm related sediment resuspension are key factors in the distribution of both total and labile metals within Philip sediments. Diagenetic processes are also important considerations during dry weather periods when sediment metal fluxes have the potential to supply remineralized metals to overlying waters and during resuspension events when 146 previously sequestered metals (i.e. sulfide bound within anoxic zones) have the potential for release to the aqueous phase with sulfide oxidation. This process, however, may only lead to small increases in dissolved metals possibly due to immediate adsorption to available oxic substrates (Chapman et al, 1998; Forstner, 1995; Zhuang et al, 1994). Concentrations of porewater trace metals showed both spatial and temporal fluctuations. Grain sizes and organic matter showed similar trends which are not uncommon and indicative of a strong involvement of these two parameters in the trace metal distributions observed over time and space. Diagenesis primarily drives vertical gradients which are redox potential driven and involve dissolution of Fe and Mn oxides, while lateral gradients can be due to differences in grain sizes, organic matter content, and bioturbation and/or bioirrigation (Aller, 1982; Wang et al, 2001). Climatic conditions may have been influential as observed in September 2000 and May 2001. Metal concentrations within Philip site sediments appear to be within the same concentration range reported for some of the more polluted regions of Vancouver Harbour in 2000 and similarly exceed Canadian environmental quality guidelines and sediment quality objectives for Cu, Zn, Cd, and Pb (Table 3.4). Reported metal concentrations in surface sediments are lower for Pb, Cu, and Zn in 2000 compared to 1985-87 for the Vancouver Wharves and areas within the vicinity of the Philip site. The degree of variability for these elements as noted in this study and these previous studies, however, make it difficult to judge if the observed reductions are indicative of temporal declines or spatial heterogeneity due to either diagenetic processes or localized metal "hotspots" or both. In particular, the Wharves area is expected to have certain elevated regions of metal contamination related to material loading and unloading activities. Blueridge metals are well within the range of concentrations previously reported within the area in 1985-87 and 2000. A l l elements at this site within the current study's 2000/01 sediment collections, except Zn (in 6 of 28 samples across stations and depths), were within or close to guidelines and below probable effects levels set for sediments. Philip chamber sediments were generally lower in metal content than inner or outer basin stations; however, samples from only two occasions were available and both were during (January 1998) or following (September 2000) active storm periods. The coarser grain sizes found within the chamber sediments and previously determined high metal concentrations in suspended particulate matter (Chapter 2) suggest that sampling 147 followed periods of chamber particulate discharge to the receiving basin. The large metal variability reported previously for Philip chamber sediments from June 1995 to March 1997 by Fedrigo & Boase (1997) (total Cu: 50-750 mg/kg, Pb: 25-500 mg/kg, Zn: 100-700 mg/kg) indicate that sampling can easily miss the high metal concentrations whenever metal enriched sediments move out of oil/grit chambers and into receiving basins. Blueridge chamber sediments did not reveal the same relative depleted total metal concentrations for Cu, Cd, and Pb; however, the much lower metal content in receiving basin sediment in this area may act to dilute metal contributions from discharged particulates. In addition, relative to Philip, the Blueridge chamber sits deeper in the ground relative to the receiving creek and thus much greater storm activity would be required, i f at all possible, to discharge coarser particles of runoff contributions from within the chamber. One notable observation of the 2000/01 collections was the elevated Cu concentrations within sediments of the Philip outer basin station relative to the inner basin sediments for both years. Pb and Cd were similarly elevated in the outer basin in 2000 but showed the opposite trend with elevated levels in the inner basin in 2001. Zn followed with outer basin elevations in 2000 but large spatial variability in 2001 precluded a definite assessment. At least with respect to Cd and Pb in 2001, inner basin concentrations represent 2-3 fold enrichments over outer basin sediments and values reported within the vicinity by others (Table 3.4). While this is not a drastic enrichment, it may indicate contributions from storm related anthropogenic influences which settle out quickly within the inner basin. The investigation of the exact nature and mineralogy of sediments was beyond the scope of this study, but it was noted earlier that stromwater passing through the chamber system decreased in suspended particulate-Zn, Cd, and Pb, which led to the suggestion that these elements were enriched in the larger size particle fractions (Chapter 2). Furthermore, these particles originating from urban environments are thought to consist of much larger surface areas than theoretically predicted (3 orders of magnitude higher) due to characteristic rough and uneven particle surfaces (Sansalone et al, 1998). By extension, the mild storm activity of May 2001 could have presented a similar pattern of metal particulate distributions with more Cd, Pb, and possibly Zn settling out with the coarser particle fractions within the inner basin while the severe storms occurring just prior to the September 2000 sediment sampling could have carried both fine and coarse particles from the inner to the outer basin, resulting in the 148 detected higher element load there. Grain size distributions seem to corroborate this with the higher percentage of fines found at the outer basin in 2000 but not in 1998. Cu concentrations are elevated within outer basin sediments for both 2000 and 2001 which is consistent with earlier findings for Cu content of suspended particulate matter that suggests a strong association with finer stormwater runoff particles (Chapter 2). The proportion of metals within each geochemical fraction was not determined in this study and thus the relative importance of each is unknown. However, the primary interest was in metal lability as it extends to bioavailability for which the relative importance of each fraction in terms of biological uptake (bioaccumulation and potential toxicity) is still at an early stage of discovery and not yet well defined. An attempt was made, however, to elucidate the predominant metal associations within surface sediments through examination of graphical relationships between acid extractable metals and geochemical variables (Fe and Mn oxides, organic carbon, and carbonate carbon) and grain size distribution (Figs. 3.41 to 3.47). No correlations were apparent with grain size distributions (both %fines and %sand, Figs. 3.46 and 3.47) and this was attributable to the sand-skewed distributions in the primarily sandy textured sediments and nature of storm particulates with large metal complexation potential as discussed above. Acid extractable Fe and Mn were used as estimates of Fe and Mn oxides respectively (Luoma & Bryan, 1981). At Philip, Cu, Zn, Cd, Pb and N i all demonstrated a certain degree of association with Fe and Mn; however, only Cu showed a positive correlation with organic carbon. Trends could be discerned in samples from the 2000 and 2001 collections for Philip but not Blueridge sediments (Figs 3.41 to 3.47). The sharp decline of acid extractable metal concentrations with sediment depth generally coincides with declines in Fe and Mn-oxides, indicating trace metal associations that release these trace metals to the solute phase with diagenetic oxide reductions during burial. The lack of correlation of acid extractable trace metals with Fe/Mn oxides in surface sediments suggests that other factors are overriding with respect to trace element associations at this site. Both extractable Fe and Mn are higher within surface sediments at Blueridge than at Philip; however, two of the four Blueridge stations have higher organic carbon content. Higher extractable Fe and Mn would indicate elevated levels of Fe/Mn oxides in an oxic environment; however, organic matter could compete for trace metals, in particular Cu. The absence of any correlations with Fe, Mn, or 149 organic carbon at Blueridge suggests by elimination that sulfides dominate trace metal associations. Some of the cores are likely within anoxic sulfidic zones which would reduce Fe/Mn oxide correlations and be detected by poor trace metal recoveries with the acid extractable technique (Allen et al, 1993; Luoma & Bryan, 1981; Tessier et al, 1979). While different subsamples were used for total and acid extractable metal determinations, some general patterns emerged. The occurrence at the Philip site of significantly high percentages of acid extractable to total or labile metals for Cu (12-46%), Zn (17-100%), Cd (14-97%) and Pb (18-100%) coupled with total concentrations for these elements that frequently exceed environmental quality guidelines highlight these metals as ones that are potentially problematic with respect to bioavailability and associated environmental risks. Additionally, the observation that these elements produced strong correlations between acid extractable and total metals indicates that extractability and potential lability is to some extent reliant on total concentrations. Metals were generally more labile at the Blueridge site, but absolute concentrations were reduced and likely not of significant environmental concern. Percentages of labile Cu and N i were both decreased in 1998 compared to 1999-2001. Cu was also decreased in 1999. This finding would be expected if the trace metal forms within sediments were primarily as metal sulfide precipitates or metals adsorbed to other solid metal sulfides. These were likely the predominant metal forms given the prevalent anoxic conditions in 1998 as evidenced by the porewater profiles. Both Cu and N i are much less soluble in weak acids when complexed to or precipitated as sulfides (Allen et al, 1993; Brumbaugh & Arms, 1996; Cooper & Morse, 1998; Simpson etal, 1998). As a final note, DGT was useful for assessing metal lability in surface sediments but would not be reliable as a sole analytical technique. In particular, sediments that exhibit great spatial variability in terms of trace metal concentrations and adsorption substrates can be difficult to define through DGT. The question of flux versus concentration for trace metals will always need to be addressed and necessitates using alternate techniques especially where spatial and/or temporal variability is known or suspected. Although pure research aspects and very specific uses for sediment DGT are attractive, the benefit of DGT utilization in sediments is not readily apparent and not recommended for most environmental monitoring programs where techniques should be fast, practical and consistently reliable. 150 3.5. Conclusions The evaluation of metal distributions within sediments of two local areas with contrasting land-use patterns revealed significant differences in both total and labile metals with the highest concentrations of Cu, Zn, Cd, and Pb occurring at the industrial/commercial site (Philip). Sediments in these two areas were fairly coarse-grained with primarily sandy textures which are typical of the region. Relationships of metals to sediment texture could not be found and of the geochemical fractions examined, labile metals were most strongly associated with Fe and Mn oxides. Total metal concentrations within both sites as compared to historical data from their respective vicinities were not substantially changed (with the possible exception of Zn in Philip vicinity) since the mid 1980s. Large variability was seen in very small localized spaces (cm to m) and over time revealing the extreme spatial and temporal patchiness which is not uncommon for estuaries in general but would in particular be expected in restricted areas subject to frequent sediment resuspension from considerable stormwater inputs, irregular introduction (in terms of both quality and quantity) of runoff debris and organic material and resulting burial and diagenetic activity. Cu, Zn, Cd and Pb concentrations all exceeded Canadian environmental quality and probable effects guidelines at the Philip site and indicate that these elements could also be of concern within other areas of coastal North Vancouver receiving industrial/commercial stormwater runoff. The evidence from grain size distributions and metal content across stations suggest that the oil/grit chamber at least within the polluted Philip site is not effectively retaining particulates and associated total or labile metals. Evidence from this chapter coupled with results of suspended particulate matter analyses discussed in Chapter 2 indicate that substantial resuspension of particulates both within the chamber and previously settled within the immediate receiving areas may be occurring and contributing to metal mobility (especially Cu and likely Pb, Cd, Zn as well) and potential bioavailability within the area studied. The single extraction technique using 1 N HC1 was useful and provided an efficient method of separating labile metal fractions of potential ecological risk from total metal concentrations that include the often substantial inert residual fractions. The sediment DGT technique, due to its variable nature in restricted environments (i.e. interstitial water), is not recommended for this type of environmental monitoring work where both spatial and temporal variability is 151 likely in terms of metal porewater concentrations and abundances and nature of sediment substrates. 152 Table 3.1. Metal analyses of marine sediment reference material PACS-2 (Esquimals Harbour, BC) and blanks using microwave digestion method. Concentrations (Mean ± SD) in mg/kg. Ni Cu Zn Cd Pb PACS-2 Certified 39.5 ± 2 . 3 310 ± 12 364 ± 23 2.11 ± 0 . 1 5 183 ± 8 Analyzed N=33 40.9 ± 4 . 6 9 323 ± 27 419 ± 25 2.07 ± 0 . 1 8 166 ± 2 3 Grinding Test N=6 42.9 ± 2 . 3 3 2 6 ± 19 423 ± 19 2.09 ± 0.08 165 ± 9 Blank ( l ) N=34 0.263 ± 0 . 3 0 1 0.734 ± 1.89 5.83 ± 4 . 1 3 0.0038 ± 0 . 0 1 0 2 0.281 ± 0584 Detection Limit™ 2 ' (3a) 0.903 5.67 12.4 0.031 1.75 'Blanks and Detection Limits are reported for a 10 mg sample of sediment. 'Detection Limits calculated as three times the standard deviation of the blanks. Table 3.2. Total carbon analyses of marine sediment reference material. %Carbon MESS-1 BCSS-1 Reported 2.99 2.19 Measured (Mean ± SD) 2.98 ± 0.03 2.19 ±0.02 153 Table 3.3. Sediment grain size (diameter Mean ± SD, mm) of surface and core samples from Philip and Blueridge sites. A . March 1998, Philip Site (surficial sediments, 0-5 cm) Philip Site Sample Mean Grain Diameter (mm) Chamber Station AO - 2.44 ± 1.49 1 2.18 ± 1.53 Inner Basin Station A1 2 2.31 ± 1.43 3 3.24 ± 1.54 B. August 1998, Philip Site (surficial sediments, 0-5 cm) Philip Site Sample Mean Grain Diameter (mm) Chamber Station AO - 2.37 ± 1.50 1 1.20 ± 1.64 Station A1 2 1.27 ± 1.69 3 1.47 ± 1.69 1 2.19 ± 1.89 Inner Basin Station A2 2 0.23 ±2.19 3 1.93 ± 1.54 1 2.24 ± 1.54 Station A3 2 0.19 ±2.22 3 0.42 ± 2.09 1 2.45 ± 1.65 Outer Basin Station A4 2 3.45 ± 1.57 3 0.53 ± 1.97 (Continued on next page) 154 Table 3.3 (continued) C. September 2000, Philip and Blueridge Sites (sediment cores 0-10 cm) Site and Station Sample Mean Grain Diameter (mm) Surface (0-1.5 cm) Depth (1.5-10 cm) Philip Site Chamber Station AO - 3.37 ±2.90* -1 0.91 ± 1.09 0.91 ± 1.25 Station A1 2 1.62 ± 1.37 2.23 ± 1.48 3 1.49 ± 1.53 1.36 ± 1.35 1 0.64 ± 1.23 0.73 ± 1.27 Inner Basin Station A2 2 1.07 ± 1.58 0.84 ± 1.52 3 0.58 ± 1.31 0.45 ± 1.71 1 0.53 ± 1.57 0.48 ± 1.47 Station A3 2 0.62 ± 1.37 0.59 ± 1.47 3 0.33 ± 1.66 0.96 ±2.54 1 1.18 ± 2.13 1.59 ±2.45 Outer Basin Station A4 2 0.830 ±2.07 1.10 ±2.06 3 0.610 ± 1.90 1.50 ±2.20 Blueridge Site Station C3 1 0.080 ±0.960 0.120 ± 1.30 2 0.140 ± 1.43 0.160 ± 1.48 3 0.240 ± 1.83 0.270 ± 1.88 Station C4 1 1.85 ± 1.43 1.24 ± 1.79 Receiving Basin 2 0.970 ± 1.48 1.12 ± 1.71 3 0.510 ± 1.37 0.950 ± 1.50 Station C5 1 2.36 ± 1.94 2.09 ± 1.43 2 2.67 ± 1.23 1.96 ± 1.29 3 2.72 ± 1.77 2.22 ± 1.45 *Chamber sediments (-0-5 cm) were obtained with plastic scoop. 155 Table 3.4. Sediment characteristics and total mean metal concentrations from selected areas of Burrard Inlet. Depth (cm) Texture Organic Carbon (mg/g) Ni (mg/kg) Cu (mg/kg) Zn (mg/kg) Cd (mg/kg) Pb (mg/kg) ISQG ( # 1 18.7 124 0.700 30.2 P E L ( $ ) 30 ( 4 ) 108 271 4.2 112 SQO < & ) 100 150 1.0 30 Average upper continental crust (%) 20 25 71 0.1 20 Port Moody ( l ), 2000; East end of Burrard Inlet 0-15 & surf grabs Silt-clay 36-41 33-38 129-140 204-242 1.5-2.3 82-88 Off South Shore(2), 1995 0-2 2-28 -2.84-3.06 1.06-3.89 39.5-57.5 30.0-88.8 260-453 65-470 214-262 128-406 0.50-1.04 0.476-1.51 81.4-125 41.1-861 Vancouver Harbour, North Shore Wharves'n, 2000 Surf grabs Sand 23 44 985 575 2.5 126 Wharves(3), 1985-87 0-2 Sand -40 88, 296 4083, 4353 654,2267 2.4,5.2 585,15420 Philip vicinity (3), 1985-87 0-2 Sand -30-80 19- 44 542-1967 828-1372 1.5-7.4 93-262 Mosquito Creek 0 ), 2000 0-15 Sand - 33,40 239, 633 99, 100 1.6, 2.1 90, 168 **Philip Site, 2000-01 Outer Basin (Inner Basin) 0-2 2-10 (0-2) (2-10) Sand Sand (Sand) (Sand) 32-46 26-46 (6.6-15) (5.7-22) 18.9-27.5 22.4-39.0 (15.7-32.8) (16.4-45.9) 687-1367 718-1580 (307-726) (241-1520) 93.8-289 608-1240 (248-763) (281-1300) 0.47-3.50 1.95-4.71 (0.95-15.5) (0.82-4.63) 78.5-189 115-253 (82.6-612) (87.8-213) Blueridge vicinity(3); 1985-87 0-2 - - 13-22 115-307 146-646 0.4-0.7 39-99 Sites east of Blueridge, 2000 ( l ) 0-15, surf Sand 14, 19.4 17-23 25.9, 56 54.9, 120 <0.8 30, 101 **Blueridge Site, 2000-01 0-2 2-10 Sand 4.7- 100 2.8- 100 11.4-27.7 2.1-17.0 19.6-108 15.3-135 71-423 116-493 0.10-0.92 0.03-0.81 16.2-66.6 12.5-92.1 ( ' I S Q G : Interim sediment quality guideline, Canadian environmental quality guideline for marine sediments; ( 'PEL: Probable Effects Level, Canadian environmental quality guideline for marine sediments; < & ) SQO: Sediment quality objectives for long-term total concentrations within Burrard Inlet (CCME, 2003). ^Lowest effects limit guideline for N i in marine sediments as proposed by Nagpal et al., (2001); ( % )Average concentrations of upper continental crust (Taylor & McClennan, 1985) **This Study References cited: ( 1 )Phippen, (2001); , 2 ) McNee, (1997); (3 )Goyette & Boyd, (1989). C -5 a a o H C B o 3 O —J on o ^-;-:-.:.:v:Resin GellayfeT/-:"^ Diffusive Gel Layer Membrane Layer Front Window Plate V Q s AUGUST MARCH o 100 £ 80 co b cu 60 N CO c 40 ro 1— CD AO A1 A2 A3 Stations A4 AO A1 Stations GRAVEL SAND SILT AND CLAY Figure 3.2. Sediment grain size distributions at the Philip site from 1998 surface (0-2 cm sediment sampling during A . August, and B. March. A. SEDIMENT CORE SUBSURFACE DEPTHS 1.5-10 CM B. SURFICIAL SEDIMENTS 0-1.5 CM Philip Blueridge K. A1 A2 A3 A4 C3 C4 C5 Stations Stations • — G R A V E L • SAND — SILT AND/OR CLAY TZZZZZZA CLAY Figure 3.3. Sediment grain size distributions from September 2000 core sampling at Philip and Blueridge sites from A . Sediment core depths 1.5-10 cm and B. Surficial sediments 0-1.5 cm. 159 A. Philip Site, August 1998 B. Philip Site, August 1999 C. Philip Site, September 2000 D. Blueridge Site, September 2000 o 2 4 H 10 —•— A1 o A2 — A - A3 \ i—h—i — D - A4 4—-p—1 y 0 2 4 6 8 10 12 14 Organic Carbon (wt.%) • C3 o C4 — A — - C5 — D - C6 0 2 4 6 8 10 12 14 Organic Carbon (wt.%) Figure 3.4. Organic carbon (% by weight) in sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 1999, C. 2000) and Blueridge (D. 2000) sites. Averages for sampling stations at Philip (stations A l to A4) and Blueridge (stations C3 to C6) are presented with error bars representing ±1 standard deviation for within station replicates. 160 A. Philip, 1998 Station A1 ••• REP 1 REP 2 Station A2 REP 3 Station A3 —1 1 1 1 T~ 0 2 4 6 8 10121416 0 2 4 6 8 10121416 0 2 4 6 8 10121416 B. Philip, 2000 Station A1 Station A2 Station A3 Station A4 0 -% ? ? 2-o £ 4 " Q . CD Q 6 -r «i . \ • / 1 t 8 -1 1 r~ 0 2 4 6 8 101214 0 2 4 6 8 101214 0 2 4 6 8 101214 0 2 4 6 8 10121416 C. Blueridge, 2000 Station C3 Station C4 Station C5 Station C6 0 -\ E 2 £ 4 (D Q 6 -1 1 1 1 1 1 "I 1 1 1 1 1 1 0 2 4 6 8 101214 0 2 4 6 8 101214 0 2 4 6 8 101214 0 2 4 6 8 10121416 Organic Carbon (%) Figure 3.5. Organic carbon (% by weight) of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. 161 A. Philip Site, August 1998 B. Philip Site, August 1999 Figure 3.6. Moisture content (% by weight) in sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 1999, C. 2000) and Blueridge (D. 2000) sites. Averages for sampling stations at Philip (stations A l to A4) and Blueridge (stations C3 to C6) are presented with error bars representing ±1 standard deviation for within station replicates. 162 A. September 1998, Philip Site Station A1 Station A2 Station A3 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 1.0 2.0 3.0 4.0 Cu (mg/g) Cu (mg/g) Cu (mg/g) B. August 1999, Philip Site Station A1 Station A2 Station A3 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Cu (mg/g) Cu (mg/g) Cu (mg/g) Figure 3.7. Total (solid bars) and acid extractable (shaded bars) Cu (mg/g) in sediment cores sampled during A. 1998 and B. 1990 at the Philip site. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 163 STATION A1 September 2000 A. PHILIP SITE STATION A2 STATION A3 STAITON A4 ~ 2 sz 4 *-> Q. CD O R 0.0 1.0 2.0 3.0 0.0 1.0 2.0 3.0 0.0 1.0 2.0 3.0 0.0 1.0 2.0 3.0 Cu (mg/g) Cu (mg/g) Cu (mg/g) Cu (mg/g) STATION C 3 o 1 — 2 E £ 4 Q. CU Q 6 B. B L U E R I D G E SITE STATION C4 STATION C 5 STATION C6 I o 2 o 2 A r 6 0.0 1.0 2.0 3.0 0.0 Cu (mg/g) 0 0 0.2 1.0 2.0 3.0 0.0 1.0 Cu (mg/g) 2.0 3.0 0.0 1.0 2.0 3.0 Cu (mg/g) Cu (mg/g) Figure 3.8. Total (solid bars) and acid extractable (shaded bars) Cu (mg/g) in sediment cores sampled in 2000 at A . Philip and B. Blueridge sites. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 164 E o CU Q 1 Station A1 2 r 4 r 0 2 4 A. September 1998, Philip Site Station A2 5 10 15 20 25 0 Zn (mg/g) Station A3 5 10 15 20 25 0 Zn (mg/g) 5 10 15 20 25 Zn (mg/g) B. August 1999, Philip Site Station A1 Station A2 Station A3 0 5 10 15 20 25 0 5 10 15 20 25 0 5 10 15 20 25 Zn (mg/g) Zn (mg/g) Zn (mg/g) Figure 3.9. Total (solid bars) and acid extractable (shaded bars) Zn (mg/g) in sediment cores sampled during A. 1998 and B. 1990 at the Philip site. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 165 September 2000 STATION A1 A. PHILIP SITE STATION A2 STATION A3 STAITON A4 0 1 2 3 0 1 2 3 0 1 2 Zn (mg/g) Zn (mg/g) Zn (mg/g) 3 o 1 2 3 Zn (mg/g) STATION C3 B. BLUERIDGE SITE STATION C4 STATION C5 STATION C6 0 1 2 3 0 1 2 3 0 1 2 Zn (mg/g) Zn (mg/g) Zn (mg/g) 3 o 1 2 3 Zn (mg/g) Figure 3.10. Total (solid bars) and acid extractable (shaded bars) Zn (mg/g) in sediment cores sampled in 2000 at A . Philip and B. Blueridge sites. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 166 A. September 1998, Philip Site Station A1 T £ 4 -4 CL 6 Station A2 Station A3 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) B. August 1999, Philip Site Station A1 Station A2 Station A3 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) Figure 3.11. Total (solid bars) and acid extractable (shaded bars) Cd (mg/kg) in sediment cores sampled during A . 1998 and B. 1990 at the Philip site. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 167 STATION A1 September 2000 A. PHILIP SITE STATION A2 STATION A3 STAITON A4 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) STATION C3 B. BLUERIDGE SITE STATION C4 STATION C5 STATION C6 0 1 2 3 0 1 2 3 0 1 2 3 0 1 2 3 Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) Cd (mg/kg) Figure 3.12. Total (solid bars) and acid extractable (shaded bars) Cd (mg/kg) in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 168 A. September 1998, Philip Site Station A1 Station A2 Station A3 0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000 Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) B. August 1999, Philip Site Station A1 Station A2 Station A3 0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000 Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) Figure 3.13. Total (solid bars) and acid extractable (shaded bars) Pb (mg/kg) in sediment cores sampled during A. 1998 and B. 1990 at the Philip site. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 169 STATION A1 September 2000 A. PHILIP SITE STATION A2 STATION A3 STAITON A4 0 250 500 750 1000 0 250 500 750 1000 0 250 500 750 1000 0 250 500 750 1000 Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) STATION C3 B. BLUERIDGE SITE STATION C4 STATION C5 STATION C6 0 250 500 750 1000 0 250 500 750 1000 0 250 500 750 1000 0 250 500 750 1000 Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) Pb (mg/kg) Figure 3.14. Total (solid bars) and acid extractable (shaded bars) Pb (mg/kg) in sediment cores sampled in 2000 at A . Philip and B. Blueridge sites. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 170 E o CL Q Station A1 A. September 1998, Philip Site Station A2 Station A3 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) B. August 1999, Philip Site Station A1 Station A2 Station A3 Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) Figure 3.15. Total (solid bars) and acid extractable (shaded bars) N i (mg/kg) in sediment cores sampled during A . 1998 and B. 1990 at the Philip site. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 171 September 2000 STATION A1 A. PHILIP SITE STATION A2 STATION A3 STAITON A4 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) STATION C3 B. BLUERIDGE SITE STATION C4 STATION C5 STATION C6 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) Ni (mg/kg) Figure 3.16. Total (solid bars) and acid extractable (shaded bars) N i (mg/kg) in sediment cores sampled in 2000 at A. Philip and B. Blueridge sites. Averages for sampling stations are presented with error bars representing ±1 standard deviation for within station replicates. 172 R E P 1 A. Philip, 1998 Station A1 R E P 2 - o - R E P 3 Station A2 R E P 4 Station A3 0 - 0 -E 2 ^ o ( 2 -£ 4 i 4 -"5. <u Q 6 - 6 -8 - i 8 -B. Philip, 2000 Station A1 o 1 Station A2 E 2 o Q . Q 6 • > : / <s i \ C. Blueridge, 2000 Station C3 Station C4 0 E 2 £ 4 Q 6 t»<5 A i A \ I I A \ \ \ A • b 0 1 2 3 4 5 Station A3 Station A4 0 1 2 0 1 2 3 Station C5 Station C6 0.0 0.1 0.0 0.1 0.0 0.1 Total Cu (mg/g) o.o 0.1 0.2 Figure 3.17. Total Cu (mg/g) of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. 173 • •• • REP 1 - A - REP 2 A. Philip, 1998 Station A1 REP 4 Station A3 0 2 4 6 8 10 0 2 4 6 8 10 B. Philip, 2000 Q . CD Q Station A1 Station A2 •> 4 < •\ \\ •.\ Station A3 0 2 4 6 8 10 Station A4 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 2.5 C. Blueridge, 2000 Station C3 Station C4 Station C5 Station C6 0 E 2 _CJ £ 4 CU Q 6 8 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 1.0 Total Zn (mg/g) Figure 3.18. Total Zn (mg/g) of station replicate sediment cores sampled during 1998-at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. • •• • REP 1 - A - - REP 2 A. Philip, 1998 Station A1 R E P 4 Station A3 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 B. Philip, 2000 Station A1 Station A2 Station A3 Station A4 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 0 2 4 6 8 C. Blueridge, 2000 Station C3 Station C4 Station C5 Station C6 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 2.0 Total Cd (mg/kg) Figure 3.19. Total Cd (mg/kg) of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. 175 A. Philip, 1998 Station A1 R E P 1 REP 2 - o - REP 3 Station A2 REP 4 Station A3 A \ \ • \ \ ^ \ \ ' • \ \ • 0 200 400 600 8001000 0 200 400 600 8001000 0 200 400 600 8001000 B. Philip, 2000 Station A1 Station A2 Station A3 Station A4 E o Q . CU 0 - 5P 2 -I 4 - •> •/ l* 6 - f ' I " t 8 -\ . \- IL 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 C. Blueridge, 2000 Station C3 Station C4 Station C5 Station C6 CU Q 6 0 40 80 120 0 40 80 120 0 40 80 120 0 40 80 120 160 Total Pb (mg/kg) Figure 3.20. Total Pb (mg/kg) of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. 176 • • • • REP 1 A. Philip, 1998 Station A1 REP 2 REP 4 Station A3 0 20 40 60 0 20 40 60 0 20 40 60 B. Philip, 2000 Station A1 Station A2 Station A3 Station A4 Ah* • 0 20 40 60 0 20 40 60 0 20 40 60 120 0 20 40 60 C. Blueridge, 2000 Station C3 Station C4 Station C5 Station C6 E 2 £ 4 CU Q 6 0 20 40 60 0 20 40 60 0 20 40 60 0 Total Ni (mg/kg) 20 40 60 Figure 3.21. Total N i (mg/kg) of station replicate sediment cores sampled during 1998-2001 at the Philip (A. 1998, B. 2000) and Blueridge (C. 2000) sites. 177 Zn Cd Pb 1500 3000 4500 Cd 80 60 40 20 0 1500 3000 4500 0 1000 5000 0 1500 3000 4500 0 120 5000 10000 0 20 40 60 80 Ni 120 i 100 -80 -60 -A 40 - * A A 20 -0 -0 1500 3000 4500 0 Cu 5000 10000 0 20 40 60 80 0 200 400 600 800 1000 Zn Cd Pb Figure 3.22. Total metal-metal correlations (concentrations in mg/kg) of all samples obtained 1998-2001. C u Zn C d Pb 0 100 200 300 400 0 1500 3000 4500 0 10 20 30 40 50 60 70 Figure 3.23. Acid extractable metal-metal correlations (concentrations in mg/kg) of all samples obtained 1998-2001. 179 - • — JAN 1998 • SEPT 1998 - A — AUG 1999 0 SEPT 2000 ••• MAY 2001 AO A1 A2 A3 A4 A5 A6 AO A1 A2 A3 A4 A5 A6 AO A1 A2 A3 A4 A5 A6 Stations 50 j CT) 40 -£ 30 -"a O 20 --2 10 --Q 0 -800 j CT) 7h 600 -g 400 -Q-~2 200 -o I— 0 --AO A1 A2 A3 A4 A5 A6 AO A1 A2 A3 A4 A5 A6 CT) E ro o AO A1 A2 A3 A4 A5 A6 Stations Figure 3.24. Mean total metal concentrations (mg/kg) in Philip site surface sediments obtained during 1998-2001. 180 ^1000 1 800 + £ 600 4-JAN 1998 0 SEPT 2000 • • • • MAY 2001 0 \ Mn CO C1 C3 C4 C5 C6 O) E, _Q a TO O cn E o 80 60 40 20 0 25 20 15 10 I I 1 I CO C1 C3 C4 C5 C6 ° Pb 0 "•5 V 1 I 1 I CO C1 C3 C4 1 C5 C6 V J • ° Ni u \ . / 0 • • I I I 1 CO C1 C3 C4 C5 C6 Sta t ions S ta t i ons Figure 3.25. Mean total metal concentrations (mg/kg) in Blueridge site surface sediments obtained during 1998-2001. 181 A. Mn STATION A1 August 1998 STATION A2 STATION A3 E o Q. OJ Q o H 10 20 30 H 40 0.0 0.1 0.2 0.0 0.1 0.2 0.0 0.1 0.2 Dissolved Mn (mg/L) Dissolved Mn (mg/L) Dissolved Mn (mg/L) B. Fe STATION A1 STATION A2 STATION A3 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Dissolved Fe (mg/L) Dissolved Fe (mg/L) Dissolved Fe (mg/L) Figure 3.26. Porewater concentrations (mg/L) of dissolved A . M n and B. Fe sampled at the Philip site in 1998. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 182 A. Mn August 1999 STATION A1 STATION A3 0.0 0.1 0.2 0.0 0.1 0.2 Dissolved Mn (mg/L) Dissolved Mn (mg/L) B. Fe STATION A1 STATION A3 Dissolved Fe (mg/L) Dissolved Fe (mg/L) Figure 3.27. Porewater concentrations (mg/L) of dissolved A. Mn and B. Fe sampled at the Philip site in 1999. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 183 A. August 1998 STATION A1 STATION A2 STATION A3 0 200 400 600 800 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Dissolved Cu (ug/L) Dissolved Cu ( L i g / L ) Dissolved Cu (ug/L) B. August 1999 STATION A1 STATION A3 1' " ' I " " I ' ' 1 1 I 1 1 1 1 I I " 1 1 I 1 " 1 I 1 1 1 1 I " 1 1 I 0 40 80 120 160 0 40 80 120 160 Dissolved Cu (ug/L) Dissolved Cu (ug/L) Figure 3.28. Porewater concentrations (ug/L) of dissolved Cu sampled in A. 1998 and B. 1999 at the Philip site. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 184 A. August 1998 STATION A1 STATION A2 STATION A3 30 A 40 A 0.0 0.5 1.0 1.5 2.00.0 0.5 1.0 1.5 2.00.0 0.5 1.0 1.5 2.0 Dissolved Zn (mg/L) Dissolved Zn (mg/L) Dissolved Zn (mg/L) B. August 1999 STATION A1 STATION A3 | - " " | " " | " " | " " | " " | | - l l , l l l " l " " l " " l " " l 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5 2.0 2.5 Dissolved Zn (mg/L) Dissolved Zn (mg/L) Figure 3.29. Porewater concentrations (pg/L) of dissolved Zn sampled in A. 1998 and B. 1999 at the Philip site. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 185 A. August 1998 STATION A1 STATION A2 STATION A3 40 A 0.0 0.2 0.4 0.6 0.8 1.0 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Dissolved Cd (ug/L) Dissolved Cd (ug/L) Dissolved Cd (ug/L) B. August 1999 STATION A1 STATION A3 I"""!""!""!""!""! r " , , l " " l " " l " " l " " l 0 1 2 3 4 5 0 1 2 3 4 5 Dissolved Cd (ug/L) Dissolved Cd (ug/L) Figure 3.30. Porewater concentrations (ug/L) of dissolved Cd sampled in A . 1998 and B. 1999 at the Philip site. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 186 STATION A1 A. August 1998 STATION A2 STATION A3 '|MM|. 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Dissolved Pb (ug/L) Dissolved Pb (ug/L) Dissolved Pb (ug/L) B. August 1999 STATION A1 STATION A3 | " " | " " | " M | l 0 20 40 60 80 100 0 20 40 60 80 100 Dissolved Pb (|ig/L) Dissolved Pb (ug/L) Figure 3.31. Porewater concentrations (pg/L) of dissolved Pb sampled in A . 1998 and B. 1999 at the Philip site. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 187 A. August 1998 STATION A1 STATION A2 STATION A3 40 H 0 20 40 60 80 100 0 5 10 15 20 25 30 0 5 10 15 20 25 30 Dissolved Ni (ug/L) Dissolved Ni (ug/L) Dissolved Ni (ug/L) B. August 1999 STATION A1 STATION A3 0 5 10 15 20 0 5 10 15 20 Dissolved Ni (ug/L) Dissolved Ni (ug/L) Figure 3.32. Porewater concentrations (ug/L) of dissolved N i sampled in A . 1998 and B. 1999 at the Philip site. Replicate peepers are presented: replicate 1 (solid symbols), replicate 2 (open symbols), and replicate 3 (shaded symbols). 188 - • - Cu Flux (ng/cm Is) C u F | u x (^g / c r n 2/ S ) - • - Cu Flux (ng/cm2/s) 0 1e-6 2e-6 3e-6 0 1e-6 2e-6 3e-6 0 1e-6 2e-6 3e-6 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Dissolved Cu (ug/L) Dissolved Cu (ug/L) Dissolved Cu (ug/L) STATION A1 STATION A2 STATION A3 Figure 3.33. Porewater concentrations (solid line, ug/L) and porewater to DGT fluxes (filled symbols, pg/cm2/s) of dissolved Cu sampled in 1998 at stations A l , A2, and A3 at the Philip site. Fluxes are presented as means ±1 standard deviation. 189 - • - Cd Flux (ug/cm Is) C d F , u x ( u g/ C m 2 /s) - • - Cd Flux (ug/cm2/s) 0 5e-8 1e-7 0 5e-8 1e-7 0 5e-8 1e-7 h~i—i—i—i—I—i—i—i—i—I—i—i—i \r~i—i—i—i—I—i—i—i—i—I—i—i—i b~i—i—i—i—I—i—i—i—i—I—i—r— o i-0 0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 Dissolved Cd (ug/L) Dissolved Cd (ug/L) Dissolved Cd (ug/L) STATION A1 STATION A2 STATION A3 Figure 3.34. Porewater concentrations (solid line, ug/L) and porewater to DGT fluxes (solid symbols, ug/cm2/s) of dissolved Cd sampled in 1998 at stations A l , A2, and A3 at the Philip site. Fluxes are presented as means ±1 standard deviation. 190 Ni Flux ((ig/cm Is) E o Q . CU Q 2e-7 4e-7 6e-7 2 4 6 8 10 — Dissolved Ni (ug/L) STATION A1 —•— Ni Flux (ng/cm/s) —•— Ni Flux (ng/cm Is) 0 2e-7 4e-7 6e-7 0 5e-7 1e-6 —1—I—I—|—I—I—I—I—|—I—I—I—I—\- —l 1 1 - + ~1 I 1 0 2 4 6 8 10 0 2 4 6 8 10 12 14 Dissolved Ni (ug/L) Dissolved Ni (ug/L) STATION A2 STATION A3 Figure 3.35. Porewater concentrations (solid line, pg/L) and porewater to DGT fluxes (filled symbols, pg/cm2/s) of dissolved N i sampled in 1998 at stations A l , A2, and A3 at the Philip site. Fluxes concentrations are presented as means ±1 standard deviation. 191 - • - G2 Cu Flux (ug/crrr/s) 5e-6 1e-5 i i . I i . i . I • -o- G1 Cu Flux (ug/cm2/s) 5.0e-6 1 .Oe-5 1,5e-5 —•— G2 Cu Flux (ug/cm2/s) 0 2e-6 4e-6 6e-6 I—i—i—i—i—|—i—i—i—i—j—i—i—\—i—|—i—i—i— — ° ~ G1 Cu Flux (ug/cm2/s) 0 3e-6 6e-6 9e-6 50 100 150 - Dissolved Cu (ug/L) STATION A1 20 40 60 80 100 Dissolved Cu (ug/L) STATION A3 Figure 3.36. Porewater concentrations (solid line, ug/L) and porewater to DGT fluxes (symbols, ug/cm2/s) of dissolved Cu sampled in 1999 at stations A l , A2, and A3 at the Philip site. Filled and open symbols represent DGT fluxes using diffusive gel thickness' of 0.062 cm (Gl) and 0.080 cm (G2) respectively. Fluxes are presented as means + or - 1 standard deviation. 192 - G2 Cd Flux (|.ig/cm2/s) 2e-7 4e-7 6e-7 >- G1 Cd Flux ((ig/cm2/s) 5e-7 1e-6 + 4 6 8 10 Dissolved Cd (ug/L) E o Q . CU Q —•— G2 Cd Flux (ng/crrr/s) 0 1e-7 2e-7 1 i • i i I i i • i I i —c— G1 Cd Flux (|ig/cm2/s) 0 1e-7 2e-7 3e-7 1 1 1 1 1 1 1 1 1 1 1 1 1 14 + __ __r-=o + + 1 2 3 Dissolved Cd (u.g/L) STATION A1 STATION A3 Figure 3.37. Porewater concentrations (solid line, pg/L) and porewater to DGT fluxes (filled symbols, pg/cm2/s) of dissolved Cd sampled in 1999 at stations A l , A2, and A3 at the Philip site. Filled and open symbols represent DGT fluxes using diffusive gel thickness' of 0.062 cm (GI) and 0.080 cm (G2) respectively. Fluxes are presented as means + or - 1 standard deviation. 193 —•— G2 Ni Flux (ug/cm7s) 0 2e-7 4e-7 6e-7 8e-7 |—I—I—I—I—|—I—I—I—I—|—I—I—I—I—|—I—I—I—I—|—r - o - G1 Ni Flux (ug/cm2/s) 0 3e-7 6e-7 9e-7 I . . . . I 0 2 4 6 8 10 12 Dissolved Ni (ug/L) STATION A1 — G 2 Ni Flux (ug/cm7s) 0 2e-7 4e-7 6e-7 1 . i i . I i . i . I i . i . I i - o - G1 Ni Flux (ug/cm2/s) 0 2e-7 4e-7 6e-7 8e-7 T—.—i—i—|—.—i—.—i—|—.—i—i—i—|—i—.—i—i—|—r G2 Ni Flux (ng/cm2/s) 2e-6 — G1 Ni Flux (Hg/cm2/s) 0 le-6 2e-6 3e-6 M I i i i i I i i i i E o Q. CD Q 2 4 6 8 10 Dissolved Ni (ug/L) STATION A3 Figure 3.38. Porewater concentrations (solid line, ug/L) and porewater to DGT fluxes (filled symbols, ug/cm2/s) of dissolved N i sampled in 1999 at stations A l , A2, and A3 at the Philip site. Filled and open symbols represent DGT fluxes using diffusive gel thickness' of 0.062 cm (Gl) and 0.080 cm (G2) respectively. Fluxes are presented as means + or - 1 standard deviation. 194 Ni Cu Cd Figure 3.39. Mean porewater to DGT fluxes (pg/cm2/sec) of dissolved N i , Cu, and Cd sampled in September 2000 at the Philip site (stations A l , A2, A3, A4) and at the Blueridge site (stations C4, C5). Error bars represent ±1 standard deviation. 195 10 10 o 4.0e-7 CT) 3 . X 3 _ 2.0e-7 CD Q 0.0 3.0e-6 f sec] E .y 2.0e-6 -)^ 3. X 3 Cu FI 1.0e-6 -DGT 0.0 3.0e-7 u •y 2.0e-7 CT) x 3 3 1.0e-7 CD Q 0.0 — i — i — i — i — i — i — r ~ AO A1 A2 A3 A4 A5 A6 u CD "Bi E 4.0e-7 . _ CT) Z =1 0) </> x 3 ~ 2.0e-7 CD Q — I 1 1 1 1 I 1 — AO A1 A2 A3 A4 A5 A6 Philip Stations C3 C4 C5 C6 Blueridge Stations Figure 3.40. Mean porewater to DGT fluxes (ug/cm2/sec) of dissolved N i (top), Cu (centre), and Cd (bottom) sampled in May 2001 at the Philip site (stations AO to A6) and at the Blueridge site (stations C3 to C6). Error bars represent ±1 standard deviation. 196 A. Philip Site, 1998 & 1999 C u - F e 1998 • A1 • A2 • A3 1999 o A1 v A2 • A3 Cu - Mn Cu - Organic Carbon 0 1 2 3 4 5 6 Acid Extractable Fe (mg/g) 0 10 20 30 Acid Extractable Mn (mg/kg) 2 4 6 8 Organic Carbon (%) Cu - Carbonate Carbon 140 - 0 120 - o O 100 -80 -60 -40 -D • T •• 20 -0 — M 10 0 100 200 300 400 500 Carbonate Carbon (mg/kg) B. Philip & Blueridge Sites, 2000 & 2001 Cu - Fe SEPTEMBER 2000: Philip • A1 v A2 • A3 • A4 Blueridge o C3 v C4 • C5 o C6 MAY 2001 Philip *A0 - * A 6 Blueridge *C3 - * C 6 300 -250 - o 0 0 200 - 0 150 -100 i 50 -0 *A0 OTD P Cu - Mn 'A3 QO *C3 Cu - Organic Carbon Cu - Carbonate Carbon 0 2 4 6 8 10 12 14 16 18 Acid Extractable Fe (mg/g) 0 20 40 60 80 100 120 Acid Extractable Mn (mg/kg) 2 4 6 8 10 12 Organic Carbon (%) 0 500 1000 1500 2000 Carbonate Carbon (mg/kg) Figure 3.41. Acid extractable Cu associations with Fe oxides, M n oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001. 197 A. Philip Site, 1998 & 1999 03 CT E_ C N V .Q X 111 •g o < Zn • Fe V V 9 o 0 1 w 0 1 2 3 4 5 6 Acid Extractable Fe (mg/g) 1998 • A1 T A2 • A3 1999 o A1 v A2 • A3 Zn - Mn Zn - Organic Carbon 10 15 20 25 30 35 40 Acid Extractable Mn (mg/kg) • • o o Zn - Carbonate Carbon 0 2 4 6 8 10 0 100 200 300 400 500 Organic Carbon (%) Carbonate Carbon (mg/kg) B. Philip & Blueridge Sites, 2000 & 2001 SEPTEMBER 2000: Philip • A1 v A2 • A3 O A4 Blueridge o C3 v C4 • C5 O C6 Zn-Fe Zn - Mn Zn - Organic Carbon 0 2 4 6 8 10 12 14 16 18 Acid Extractable Fe (mg/g) 0 20 40 60 80 100 120 Acid Extractable Mn (mg/kg) Organic Carbon (%) MAY 2001 Philip *A0- *A6 Blueridge *C3- *C6 Zn - Carbonate Carbon • o T • A 1 'A5 o o m4° o o 0 Bra • • 6 8 10 12 0 500 1000 1500 2000 Carbonate Carbon (mg/kg) Figure 3.42. Acid extractable Zn associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999' and B. Philip and Blueridge sites in 2000 and 2001. 198 A. Philip Site, 1998 & 1999 C d - F e 0 1 2 3 4 5 6 Acid Extractable Fe (mg/g) 1998 • A1 • A2 • A3 1999 o A1 v A2 • A3 Cd - Mn • V • £ ° Cd - Organic Carbon Cd - Carbonate Carbon 0 10 20 30 40 Acid Extractable Mn (mg/kg) 0 2 4 6 8 10 Organic Carbon (%) 0 100 200 300 400 500 Carbonate Carbon (mg/kg) B. Philip & Blueridge Sites, 2000 & 2001 10 O LU •g o < 4 2 H 0 Cd - Fe *A1 *A2 0 5 10 15 20 0 2 4 6 8 10 12 14 16 18 Acid Extractable Fe (mg/g) SEPTEMBER 2000: Philip © A1 v A2 • Blueridge o C3 v CA A3 • A4 • C5 O C6 MAY 2001 Philip *A0 -*A6 Blueridge *C3 -*C6 Cd - Mn Cd - Organic Carbon Cd • Carbonate Carbon 20 40 60 100 120 Acid Extractable Mn (mg/kg) 0 2 4 6 8 10 12 Organic Carbon (%) 0 500 1000 1500 2000 Carbonate Carbon (mg/kg) Figure 3.43. Acid extractable Cd associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001. 199 A. Philip Site, 1998 & 1999 600 1998 • A1 T A2 • A3 1999 o A1 v A2 n A3 Pb - Fe Pb-Mn Pb - Organic Carbon • Qp 8 v Pb - Carbonate Carbon 0 1 2 3 4 5 6 Acid Extractable Fe (mg/g) 0 10 20 30 40 Acid Extractable Mn (mg/kg) 0 2 4 6 8 10 0 100 200 300 400 500 Organic Carbon (%) Carbonate Carbon (mg/kg) B. Philip & Blueridge Sites, 2000 & 2001 CT 600 -CO E 500 -.O Q_ 400 CO -Q 300 -ro o ra 200 -•s LU 100 -•o o < 0 Pb - Fe *A1 *A2 SEPTEMBER 2000: Philip o A1 v A2 • A3 O A4 Blueridge o C3 v C4 • C5 o C6 MAY 2001 Philip *A0- *A6 Blueridge *C3- *C6 Pb - Mn Pb - Organic Carbon Pb - Carbonate Carbon 0 2 4 6 8 10 12 14 16 18 Acid Extractable Fe (mg/g) 0 20 40 60 80 100 120 Acid Extractable Mn (mg/kg) 2 4 6 8 10 Organic Carbon (%) 500 1000 1500 2000 Carbonate Carbon (mg/kg) Figure 3.44. Acid extractable Pb associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001. 200 A. Philip Site, 1998 & 1999 Ni-Fe 1998 • A1 • A2 • A3 1999 o A1 v A2 • A3 Ni - Mn Ni - Organic Carbon Ni - Carbonate Carbon 9 V o 0 V • • • T » • • 0 1 2 3 4 5 6 Acid Extractable Fe (mg/g) 0 10 20 30 40 Acid Extractable Mn (mg/kg) 0 2 4 6 8 10 0 100 200 300 400 500 Organic Carbon (%) Carbonate Carbon (mg/kg) B. Philip & Blueridge Sites, 2000 & 2001 Ni-Fe SEPTEMBER 2000: Philip • A1 T A2 • A3 • A4 Blueridge o C3 v C4 • C5 O C6 MAY 2001 Philip *A0- *A6 Blueridge *C3- *C6 Ni - Mn Ni - Organic Carbon Ni - Carbonate Carbon 0 2 4 6 8 10 12 14 16 18 Acid Extractable Fe (mg/g) 0 20 40 60 80 100 120 Acid Extractable Mn (mg/kg) 0 2 4 6 8 10 12 Organic Carbon (%) 0 500 1000 1500 2000 Carbonate Carbon (mg/kg) Figure 3.45. Acid extractable N i associations with Fe oxides, Mn oxides, organic carbon, and carbonate carbon in surface sediments (0-2 cm) at the A . Philip site in 1998 and 1999 and B. Philip and Blueridge sites in 2000 and 2001. 201 O < 0 20 40 60 80 100 Percentage Silt 0 20 40 60 80 100 Percentage Silt 0 20 40 60 80 100 Percentage Silt E TO E a) XI (TJ O CO s CD o < Ni 25 -20 - A 15 -A A 10 -5 -0 -8 A T A A r-0 20 40 60 80 100 Percentage Silt 0 20 40 60 80 100 Percentage Silt Figure 3.46. Acid extractable metal associations with percentage silt in surface sediments (0-2 cm) at the Philip and Blueridge sites in all samples. 202 Cu Zn Cd 80 100 Percentage sand Percentage sand Percentage sand •g o < 0 20 40 60 80 100120 Percentage sand 0 20 40 60 80 100120 Percentage sand Figure 3.47. Acid extractable metal associations with percentage sand in surface sediments (0-2 cm) at the Philip and Blueridge sites in all samples. 203 3.6. References Allen, H.E., Fu, G., Boothman, W.S., D i Toro, D .M. , & Mahony, J.D. (1991). Determination of acid volatile sulfide and selected simultaneously extractable metals in sediment. Office of Water Regulations and Standards, U.S. Environmental Protection Agency, Washington, D.C. Allen, H.E., Fu, G., & Deng, B. (1993). Analysis of acid-volatile sulfide (AVS) and simultaneously extracted metals (SEM) for the estimation of potential toxicity in aquatic sediments. Environ. Toxicol. Chem., 12, 1441-1453. Aller, R.C. (1982). The effects of macrobenthos on chemical properties of marine sediments and overlying water. In Animal-Sediment Relations, (ed. P.L. McCall & M.J.S. Tevesz), pp. 53-102. Plenum Press, New York. Ankley, G.T., Phipps, G.L., Leonard, E.N., Benoit, D.A., Mattson, V.R., Kosian, P.A., Cotter, A . M . , Dierkes, J.R., Hansen, D.J., & Mahony, J.D. (1991). Acid-volatile sulfide as a factor mediating cadmium and nickel bioavailability in contaminated sediments. Environ. Toxicol. Chem., 10, 1299-1307. Ankley, G.T., Leonard, E.N., & Mattson, V.R. (1994). Prediction of bioaccumulation of metals from contaminated sediments by the oligochaete, Lumbriculus variegatus. Water Res., 28, 1071-1076. Barron, M.G. (1995). Bioaccumulation and bioconcentration in aquatic organisms. In Handbook of Ecotoxicology. (ed. D.J. Hoffman, B.A. Rattner, G.A. Burton Jr. & J. Cairns Jr.), pp. 652-666. CRC Press, Inc., Boca Raton. Bendell-Young, L.I. & Harvey, H.H. (1992). Metal concentrations in chironomid larvae in relation to the geochemical characteristics of surficial sediments. Arch. Environ. Contam. Toxicol, 21,202-211. Bendell-Young, L.I., Chouinard, J., & Pick, F.R. (1994). Metal concentrations in chironomids in relation to peatland geochemistry. Arch. Environ. Contam. Toxicol, 27, 186-194. Berry, W.J., Hansen, D.J., Mahony, J.D., Robson, D.L., Di Toro, D .M. , Shipley, B.P., Rogers, B., Corbin, J.M., & Boothman, W.S. (1996). Predicting the toxicity of metal-spiked laboratory sediments using acid-volatile sulfide and interstitial water normalizations. Environ. Toxicol Chem., 15, 2067-2079. BIEP (1992). Summary Report of Existing Environmental Quality Information. Prepared by the Environmental Quality Objectives/Monitoring Action Team(s). Burrard Inlet Environmental Action Program, Vancouver, BC. Brumbaugh, W.G. & Arms, J.W. (1996). Quality control considerations for the determination of acid-volatile sulfide and simultaneously extracted metals in sediments. Environ. Toxicol Chem., 15, 282-285. Bryan, G.W. & Langston, W.J. (1992). Bioavailability, accumulation and effects of heavy metals in sediments with special reference to United Kingdom estuaries: a review. Environ. Pollut., 76, 89-113. Campbell, P.G.C., Lewis, A .G. , Chapman, P.M., Crowder, A . A . , Fletcher, W.K., Imber, B., Luoma, S.N., Stokes, P.M., & Winfrey, M . (1988). Biologically available metals in sediments, Rep. No. No. 27694. National Research Council of Canada, Ottawa. 204 Campbell, P.G.C. & Tessier, A . (1991). Biological availability of metals in sediments: analytical approaches. In Heavy Metals in the Environment, (ed J.P. Vernet), pp. 161-173. Elsevier Sci. Publ. Co. Inc., Amsterdam. Carignan, R., Rapin, F., & Tessier, A . (1985). Sediment porewater sampling for metal analysis: A comparison of techniques. Geochim. Cosmochim. Acta, 49, 2493-2497. Carlson, A.R., Phipps, G.L., Mattson, V.R., Kosian, P.A., & Cotter, A . M . (1991). The role of acid-volatile sulfide in determining cadmium bioavailaility and toxicity in freshwater sediments. Environ. Toxicol. Chem., 10, 1309-1319. Cassas, A . M . & Crecelius, E.A. (1994). Relationship between acid volatile sulfide and the toxicity of zinc, lead and copper in marine sediments. Environ. Toxicol. Chem., 13, 529-536. C C M E (2003). Canadian Environment Quality Guidelines: Summary of Existing Canadian Environmental Quality Guidelines. Canadian Council of Ministers of the Environment. Chapman, P.M., Want, F., Janssen, C.R., Persoone, G., & Allen, H.E. (1998). Ecotoxicology of metals in aquatic sediments: binding and release, bioavailability, risk assessment, and remediation. Can. J. Fish. Aquat. Sci., 55, 2221-2243. Cooper, D.C. & Morse, J.W. (1998). Extractability of metal sulfide minerals in acidic solutions: Application to environmental studies of trace metal contamination within anoxic sediments. Environ. Sci. Technol, 32, 1076-1078. DeWitt, T.H., Swartz, R.C., Hansen, D.J., McGovern, D., & Berry, W.J. (1996). Bioavailability and chronic toxicity of cadmium in sediment to the estuarine amphipod Leptocheirusplumulosus. Environ. Toxicol. Chem., 15, 2095-2101. Di Toro, D.M. , Mahony, J.D., Hansen, D.J., Scott, K.J. , Carlson, A.R., & Ankley, G.T. (1992). Acid volatile sulfide predicts the acute toxicity of cadmium and nickel in sediments. Environ. Sci. Technol., 26, 96-101. Fan, W., Wang, W.-X., & Chen, J. (2002). Geochemistry of Cd, Cr, and Zn in highly contaminated sediments and its influences on assimilation by marine bivalves. Environ. Sci. Technol., 36, 5164-5171. Fedrigo, A . M . & Boase, R. (1997). Sediment and Stormwater Chemistry of Oil-Grit Chambers. Department of Civil Engineering, University of British Columbia, Vancouver and District of North Vancouver, B.C. Folk, R.L. (1968). Petrology of Sedimentary Rocks Hemphill's, Austin, Texas, pp. Forstner, U . (1984). Chemical forms and reactivities of metals in sediments. In Chemical Methods for Assessing Bio-Available Metals in Sludges and Soils, (ed. R. Leschber, R.D. Davis & P. L'Hermite). Elsevier Applied Science Publishers, New York. Forstner, U . (1995). Non-linear release of metals from aquatic sediments. In Biogeodynamics of pollutants in soils and sediments: risk assessment of delayed and non-linear responses, (ed. W. Salomons & W . M . Stigliani), pp. 247-307. Springer-Verlag, Berlin. Goyette, D. & Boyd, J. (1989). Distribution and Environmental Impact of Selected Benthic Contaminants in Vancouver, Harbour, British Columbia. 1985-1987. Environment Canada, Conservation and Protection, Environmental Protection, Pacific and Yukon Region, Regional Program Report 89-02. Hansen, D.J., Berry, W.J., Mahony, J.D., Boothman, W.S., Di Toro, D .M. , Robson, D.L., Ankley, G.T., Ma, D., Yan, Q., & Pesch, C.E. (1996). Predicting the toxicity of 205 metal-contaminated field sediments using interstitial concentration of metals and acid-volatile sulfide normalizations. Environ. Toxicol. Chem., 15, 2080-2094. Hooda, P.S., Zhang, H. , Davison, W., & Edwards, A .C . (1999). Measuring bioavailable trace metals by diffusive gradients in thin films (DGT): soil moisture effects on its performance in soils. Eur. J. Soil. Sci., 50, 285-294. Jenne, E.A. & Luoma, S.N. (1977). Forms of trace elements in soils, sediments, and associated waters: An overview of their determination and biological availability. E R D A Symposium Series 42, Springfield, V A . Langston, W.J. (1980). Arsenic in U.K. estuarine sediments and its availability to deposit-feeding bivalves. J. Mar. Biol. Ass. U.K., 60, 869-881. Lee, J.-S., Lee, B.-G., Yoo, H. , Koh, C.-FL, & Luoma, S.N. (2001). Influence of reactive sulfide (AVS) and supplementary food on Ag, Cd and Zn bioaccumulation in the marine polychaete Neanthes arenaceodentata. Mar. Ecol-Prog. Ser., 216, 129-140. Lee, J.-S., Lee, B.-G., Luoma, S.N., & Yoo, H. (2004). Importance of equilibration time in the partitioning and toxicity of zinc in spiked sediment bioassays. Environ. Toxicol. Chem., 23, 65-71. Liber, K., Call, D.J., Markee, T.P., Schmude, K . L . , Balcer, M.D. , Whiteman, F.W., & Ankley, G.T. (1996). Effects of acid-volatile sulfide on zinc bioavailability and toxicity to benthic macroinvertebrates: a spiked-sediment field experiment. Environ. Toxicol. Chem., 15, 2113-2125. Luoma, S.N. & Bryan, G.W. (1978). Factors controlling the availability of sediment-bound lead to the estuarine bivalve Scrobiculariaplana. J. Mar. Biol. Ass. U.K., 58, 793-802. Luoma, S.N. & Bryan, G.W. (1979). Trace metal bioavailability: modeling chemical and biological interactions of sediment-bound zinc. In Chemcal Modeling in Aqueous Systems, (ed E.A. Jenne), pp. 577-609. ACS Symp. Ser. No. 93, Am. Chem. Soc, Washington, D.C. Luoma, S.N. & Bryan, G.W. (1981). A statistical assessment of the form of trace metals in oxidized estuarine sediments employing chemical extractants. Sci. Total Environ., 17, 165-196. Luoma, S.N. & Bryan, G.W. (1982). A statistical study of environmental factors controlling concentrations of heavy metals in the burrowing bivalve Scrobicularia plana and the polychaete Nereis diversicolor. Estuar. Coast. Mar. Sci., 15, 95-108. McNee, J.J. (1997). The Post-Depositional Cycling of Cd, Cu, Mo and Zn in Several Hydrographically Distinct B.C. Fjords. Ph.D., University of British Columbia, Vancouver. Menking, K . M . , Musler, H .M. , Fitts, J.F., Bischoff, J.L., & Anderson, R.S. (1993). Sediment Size Analyses of the Owens Lake Core, Rep. No. USGS Open File Report 93-683. U.S. Geological Survey. Nagpal, N.K. , Pommen, L.W., & Swain, L .G. (2001). A Compendium of Working Water Quality Guidelines for British Columbia: Updated August 23, 2001. Water Management Branch, Environmental and Resource Management Department, Ministry of Water, Land and Air Protection. Pesch, C.E., Hansen, D.J., Boothman, W.S., Berry, W.J., & Mahony, J.D. (1995). The role of acid-volatile sulfide and interstitial water metal concentrations in determining 206 bioavailability of cadmium and nickel from contaminated sediments to the marine polychaete Neanthes arenaceodentata. Environ. Toxicol. Chem., 14, 129-141. Phippen, B. (2001). Assessment of Burrard Inlet Water and Sediment Quality 2000. Prepared by Burke Phippen, BWP Consulting for the Water Protection Branch, Ministry of Water, Land, and Air Protection. Burrard Inlet Environmental Action Program, Victoria, BC. Pickering, W.F. (1981). Selective chemical extraction of soil components and bound metal species. CRCCr. Rev. Anal. Chem., 12, 233-266. Rule, J.H. & Alden III, R.W. (1990). Cadmium bioavailability to three estuarine animals in relation to geochemical fractoins to sediments. Arch. Environ. Contam. Toxicol, 19, 878-885. Sadiq, M . (1992). Toxic Metal Chemistry in Marine Environments. Marcel Dekker, Inc., New York. pp. Sansalone, J., Koran, J., Smithson, J., & Buchberger, G. (1998). Physical characteristics of highway solids transported during rainfall. J. Environ. Eng., 124, 427-440. Sibley, P.K., Ankley, G.T., Cotter, A . M . , & Leonard, E.N. (1996). Predicting chronic toxicity of sediments spiked with zinc: an evaluation of the acid-volatile sulfide model using a life-cycle test with the midge Chironomus tentans. Environ. Toxicol. Chem., 15, 2102-2112. Simpson, J.H., Apte, S.C., & Batley, G.E. (1998). Effect of short-term resuspension events on trace metal speciation in polluted anoxic sediments. Environ. Sci. Technol., 32, 620-625. Simpson, S.L., Rosner, J., & Ellis, J. (2000). Competitive displacement reactions of cadmium, copper, and zinc added to a polluted, sulfidic estuarine sediment. Environ. Toxicol. Chem., 19, 1992-1999. Simpson, S.L. (2001). A rapid screening method for acid-volatile sulfide in sediments. Environ. Chem., 20, 2657-2661. Sundelin, B. & Eriksson, A . - K . (2001). Mobility and bioavailability of trace metals in sulfidic coastal sediments. Environ. Toxicol. Chem., 20, 748-756. Taylor, S.R. & McClennan, S.M. (1985). The Continental Crust; Its composition and evolution; an examination of the geochemical record preserved in sedimentary rocks. Blackwell Scientific Publications, Oxford. 312 pp. Tessier, A. , Campbell, P .G .C , & Bisson, M . (1979). Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem., 51, 844-851. Tessier, A. , Campbell, P .G.C, Auclair, J .C , & Bisson, M . (1984). Relationships between the partitioning of trace metals in sediments and their accumulation in the tissues of the freshwater mollusc Elliptio complanata in a mining area. Can. J. Fish. Aquat. Sci., 41, 1463-1472. Tessier, A. , Couillard, Y. , Campbell, P .G .C , & Auclair, J.C. (1993). Modeling Cd partitioning in oxic lake sediments and Cd concentrations in the freshwater bivalve Anodonta grandis. Limnol. Oceanogr., 38, 1-17. Thomas, C A . & Bendell-Young, L.I. (1998). Linking the sediment geochemistry of an intertidal region to metal bioavailability in the deposit feeder Macoma balthica. Mar. Ecol-Prog. Ser., 173, 197-213. 207 Timmermans, K.R. (1993). Accumulation and effects of trace metals in freshwater invertebrates. In Ecotoxicology of Metals in Invertebrates, (ed. R. Dallinger & P.S. Rainbow), pp. 143-175. CRC Press, Boca Raton. Wang, F., Tessier, A. , & Hare, L. (2001). Oxygen measurements in the burrows of freshwater insects. Freshwater Biol, 46. Wiklund, A . - K . E . & Sundelin, B. (2002). Bioavailability of metals to the amphipod Monoporeia affinis: interactions with authigenic sulfides in urban brackish-water and freshwater sediments. Environ. Toxicol. Chem., 21, 1219-1228. Zhang, H. & Davison, W. (1995). Performance characteristics of diffusion gradients in thin films for the insitu measurement of trace metals in aqueous solution. Anal. Chem., 67, 3391-3400. Zhang, H., Davison, W., Miller, S., & Tych, W. (1995). In situ high resolution measurements of fluxes of N i , Cu, Fe, and Mn and concentrations of Zn and Cd in porewaters by DGT. Geochim. Cosmochim. Acta, 59, 4181-4192. Zhang, H. , Davison, W., Knight, B., & McGrath, S. (1998). In situ measurements of solution concentrations and fluxes of trace metals in soils using DGT. Environ. Sci. Technol, 32, 704-710. Zhuang, Y. , Allen, H.E., & Fu, G. (1994). Effect of aeration of sediment on cadmium binding. Environ. Toxicol. Chem., 13, 717-724. 208 CHAPTER 4 BIOMONITOR ASSESSMENT OF METAL DISTRIBUTION IN MARINE AREAS RECEIVING STORMWATER RUNOFF 4.1. Introduction Pollution sources and areas with elevated metal concentrations are traditionally identified through measurements of metal concentrations in sediment and water. However, metal uptake by biota, especially sessile species, has the potential to provide environmental metal distribution patterns that are ecologically relevant. Certain coastal organisms are able to bioaccumulate metals and provide a relative measure of metal contamination over space and time. These organisms are termed 'biomonitors' and can accumulate metals to very high levels in their tissues. Tissue concentrations can and have been interpreted as time-integrated measures of the ecologically relevant bioavailable portions within the ambient environment (Rainbow & Blackmore, 2001; Rainbow et al, 2004; Rainbow & Phillips, 1993; Rainbow et al., 2002). The bioaccumulation patterns of individual biomonitor species can differ significantly due to their susceptibility to different sources of bioavailable metals and species-specific bioaccumulation strategies. For a complete portrayal of metal bioavailability in an aquatic environment, the recommended approach requires the use of a suite of biomonitors that reflect metal bioavailabilities from various sources (Rainbow, 1995; Rainbow & Phillips, 1993). Mussels (genus: Mytilus) have been recognized as useful biomonitors of environmental pollution and related changes. Mussel Watch programs (Besada et al, 2002; Cantillo, 1998; Claisse, 1989; Franco etal, 2002; O'Connor, 1996), and numerous independent research monitoring programs, have been extensively used to assess coastal environmental pollution since the mussel watch concept was first suggested in the 1970s (Goldberg, 1975; Goldberg et al, 1978; Phillips, 1976, 1977). Mytilus trossulus and M. galloprovincialis (both species common to the North American Pacific Coast), in addition to the well known M. edulis (formally all were thought to be the same species (McDonald & Koehn, 1988)) have been demonstrated as useful and sensitive biomonitors for trace metal assessments in coastal areas (Besada et al., 2002; Blackmore & Wang, 2003; Rainbow et al., 2004; Rainbow et al, 2000; Szefer et al, 2004; Ugur et al, 2002). Barnacles have also become increasingly useful as biomonitors with intensive focus in the Indo-Pacific, but also extending to some areas of the Americas and Europe (Blackmore, 1996; Rainbow, 1993; Rainbow et al., 2004; Rainbow et al, 1993; Weeks et al, 1995). In recognition of their high 210 metal accumulation capacity, seaweeds too have been used as metal biomonitors. Enteromorpha and Fucus spp. are often used in macroalgal assessments of metal contamination in the solute phase (Barnett & Ashcroft, 1985; Favero & Frigo, 2002; Fuge & James, 1973; Haritonidis & Malea, 1995; Leal et al, 1997; Muse et al, 1999; Villares et al, 2001) or together with other groups of organisms for more extensive metal monitoring studies (Larsen et al, 2001; Ostapczuk et al, 1997; Phillips, 1979; Storelli et al, 2001; Struck et al, 1997). For the best representation of metal bioavailabilities within this study, the four sessile species that occurred at each of the three sites were selected for metal analyses. These consisted of the brown alga (Fucus gardneri), the green alga (Enteromorpha sp.), the mussel (Mytilus trossulus) and the barnacle (Balanus glandula). Fucus, Enteromorpha, Mytilus, and barnacles in general, have been broadly and effectively used in bioaccumulation studies and for the assessment of spatial and temporal metal distributions. For an assessment of metal bioavailability within areas influenced by stormwater discharges, a comparison of metal bioaccumulation profiles was conducted as well as an examination of linkages with environmental metal concentrations. Due to the introduction of the DGT technique in representing bioavailable aquatic metals (Zhang & Davison, 1995, 2000), its comparability with bioaccumulations was also sought. 4.2. Materials and Methods 4.2.1. Field Program Barnacles (Balanus glandula), mussels (Mytilus trossulus), and two macroalgae (Enteromorpha sp., Fucus gardneri) were collected on May 22, 2001 from several stations at three sites (Philip, Blueridge, Lighthouse Park) along the North shore of Vancouver as described earlier (Chapters 1 and 2, Figs. 1.1 to 1.3). Briefly, each of the three marine locations is characterized by a different degree of pollution influences (stormwater, urban, industrial). The Philip site lies in an urban/industrial setting and is subject to large inputs of metal-laden stormwater discharges. The Blueridge site is within a residential/parkland area and receives small stormwater discharges with low metal inputs. The stormwater 211 management practice of oil/grit chamber system utilization is established at both sites. Stormwater is directed through the oil/grit chamber systems prior to its discharge into marine receiving waters. The Lighthouse Park location is used as a reference site and is located within a relatively clean parkland/residential area free from any direct pollutant discharges. The Philip inner basin (stations AO to A3) in addition to lacking three of the four study species is also subject to large salinity fluctuations (Chapter 2) which lessens comparability to the other two sites. It is discussed in detail in the next chapter with respect to Balanus. These stations are, however, included for the examination of associations between environmental variables and biota metal accumulations, as they encompass a complete data set for at least this one species Biota was collected within the lower eulittoral zone from existing substrata (rock) from all three sites and transported back to the laboratory in acid clean plastic containers. A l l biota samples were kept cool during transport and stored frozen at -20°C until sample processing (~2 weeks). Barnacles (basal diameters of 5-6 mm) were carefully removed from rocks using stainless steel instruments. Triplicate DGT units were deployed in the ambient water prior to biota collections, for seven 10-19 day intervals each during February 6 to May 20, 2001, and averaged to provide concurrent DGT-labile metal concentrations with biota metal exposures. Water salinity, temperature, and pH were measured at deployment times. DGT units were also deployed in sediment for approximately 24 hours (May 20-21) for metal DGT-fluxes. Water, suspended particulate matter, and sediments were collected simultaneously with biota collection. Sampling and analysis protocols for water, sediment, and DGT have previously been described in Chapters 2 and 3. A l l samples were analyzed for N i , Cu, Zn, Cd, and Pb. 4.2.2. Analytical Program 4.2.2.1. Reagents ICP-MS metal stock and working solutions were prepared using nanopure water and ultrapure grade concentrated nitric acid (Seastar Chemicals Inc., Sidney, BC). High purity 212 nitric acid (70% w/w OmniTrace, E M D Chemicals Inc., CA) and hydrogen peroxide (30% w/w Suprapur, E M D Chemicals Inc.) were used for microwave digestions (EMD Chemicals, CA.) . A l l plasticware was acid cleaned as described in Chapter 2. 4.2.2.2. Sample Preparation Upon thawing, barnacle soft tissues were removed from shells using acid clean plastic forceps, carefully rinsed with nanopure water to remove any detritus, and placed in acid-clean high density polyethylene vials. Fifty individuals were pooled for each sample to provide sufficient tissue for analysis (100-200 mg dry weight). Mussel soft tissues were separated from shells and byssal threads which were discarded. Whole soft tissue bodies were carefully rinsed with nanopure water and retained for analyses. To maintain consistency between stations, only individuals within the shell length range of 2.5-3.7 cm were used. Seven to twelve individuals from each station were pooled together to produce each of four replicates. Soft tissues were homogenized and aliquots of approximately 10 g wet weight transferred to acid clean vials. Macroalgae samples were similarly processed with fronds (Enteromorpha) or thalli (Fucus) from 5-8 similarly sized individual organisms pooled together for each replicate sample and 10 g wet weight aliquots of homogenized material used per sample. A l l samples were completely dried using a high vacuum freeze dryer (Edwards 4K Modulyo) and weighed after allowing for weight stabilization under normal laboratory atmospheric conditions (minimum of 48 hours). Dried macroalgae and mussel samples were pulverized using Teflon utensils and aliquots of powdered samples were transferred directly to Teflon vessels for digestion (200-300 mg for macroalgae and 500 mg for mussels). Dried barnacle samples were transferred whole to Teflon tubes that were inserted within the Teflon vessels for digestion (3 tubes per vessel). Concentrated nitric acid (OmniTrace, E M D Chemicals Inc., CA) was added to all prepared samples which were then allowed to sit overnight under a laminar flow fumehood with loose caps prior to microwave digestion. Hydrogen peroxide (30% w/w SUPRAPUR, E M D Chemicals Inc.) was then added to each sample for a nitric acid to hydrogen peroxide ratio of 4:1. Each species was processed in 213 separate runs using a Milestone microwave digestion system (MLS 1200 Mega, Milestone, U.S.A.) with 2 blanks and 2 lobster hepatopancrease marine reference material samples (TORT-2, NRC). A timed microwave program of 250W for 3 min, 0W for 2 min, 250W for 7 min, 450W for 12 min, 650W for 24 min, and 250W for 12 min was utilized. Samples were allowed to vent and cool before removing vials and evaporating to dryness. The samples were then reconstituted with 1 mL of 1 N Seastar HNO3 before preparation for ICP-MS analyses. Procedures for cleaning of Teflon tubes and vessels between runs and acid-cleaning of all other plasticware were described in Chapter 2. 4.2.2.3. Instrumental Metal Analysis Analyses of metals was performed on the Element2™ HP-ICP-MS (ThermoFinnigan Element2™ High Performance ICP-MS, Finnigan-MAT, Mississauga, Canada) as previously described in Chapter 2. 4.2.2.4. Analysis of Reference Materials Certified reference materials (purchased from NRC) were used to verify accuracy of sample preparation and analyses. Good agreement was achieved between certified and analyzed values (Table 4.1). 4.2.3. Statistical Analyses Two parametric tests were performed in the statistical treatment of data. Analysis of variance (ANOVA) was used for macroalgae data and A N O V A or analysis of covariance (ANCOVA) for invertebrate data. A l l distributions were tested for normality and homogeneity of variances using both graphical means (frequency/probability distributions and analysis of residuals) and using ad hoc tests (Kolmogorov-Smirnov goodness-of-fit test and Levene test of homogeneity of variance, using the p<0.05 significance level). If 214 required, data was log-transformed to remove skewness and heteroscedasticity. A l l statistical computations were performed with Statistica (Ver. 5.1, 97 ed, StatSoft, OK). With respect to invertebrate metal data, the potential exists for size effects on body metal concentrations and a generally accepted model of the relationship between metal concentration (y) and body dry weight (x) is the power function y = axb (Phillips & Rainbow, 1988; Rainbow et al., 2000). A l l data were therefore transformed logarithmically to create additive data sets and to satisfy the assumptions of normality and variance homogeneity. Invertebrate data were first tested for significant correlations (Pearsons's product-moment correlation test,/j><0.05, and examination of scatterplots) between metal concentration (ug/g) dry weight and mean body dry weight of pooled bodies. Whole data sets were examined per metal within each species and per metal within each species at each site. If a size effect was detected, an A N C O V A was performed which essentially combines regression with A N O V A . Without a size effect, an A N O V A could be carried out. The further assumption under A N C O V A of homogeneity of covariate regression coefficients (slopes of regression lines are the same for each group) was tested to determine if adjustments of metal concentrations (for organisms of an overall mean body weight) were necessary. Post-hoc analyses using the Tukey Honest Significant Difference test (Tukey HSD) were employed for both A N O V A and A N C O V A to detect where differences were located. 4.3. Results 4.3.1. Macroalgae Macroalgae metal concentrations, along with statistical results, are presented in Table 4.2 for Fucus gardneri and Table 4.3 for Enteromorpha. Analyses within each species showed significant differences for Cu between Philip and Blueridge stations with elevations at the former site while Blueridge was similar to the reference site. Both species also show similar significant differences between sites for Zn although some overlap occurs between the Philip and Blueridge sites for Fucus and between the Blueridge and reference site for Enteromorpha. Elevated Cd in Enter omorpha was significantly different (p<0.05) at the 215 Philip site compared to the other two sites and Pb in both species was lower at the reference site compared to Philip and Blueridge. 4.3.2. Mytilus trossulus Analyses of the effect of size on metal concentrations in the mussel Mytilus trossulus showed significant (p<0.05) correlations for four out of the five metals studied (Cu, Zn, Cd, Pb). For these metals, A N C O V A was therefore used to compare accumulated metal concentrations between stations after confirmation of parallel regression lines (/?<0.05). Due to the presence of a size effect and to allow meaningful comparisons, metal concentrations are presented as weight-adjusted mean concentrations (estimated by the best-fit double log regression for each site and using the mean weight of mussel soft tissue for the whole data set for each metal). Metal concentrations in mussels and results of post hoc analyses are presented in Table 4.4. Cu, Zn, Cd, and Pb all show significant elevations in mussels sampled from the Philip site compared to the Blueridge site. A further reduction is also apparent for the mussels collected from the reference site which is significant for all four of the metals when compared to the Philip site and for Cu and Pb in relation to the Blueridge site. 4.3.3. Balanus glandula Regression analyses performed on log-transformed data sets for each metal did not demonstrate a significant effect (p<0.05) of mean body weight on metal concentration. Careful selection of individual barnacles with basal diameters of 5-6 mm and pooling of 50 individuals per sample effectively removed the size bias from the study's barnacle data. The mean body weight was 3.48±0.80 mg (dry weight). Metal concentrations and statistical findings for barnacles from the Philip, Blueridge, and Lighthouse sites are reported in Table 4.5. An overall A N O V A revealed significant differences between sites for Cu, Zn, and Pb but not Cd and N i . Differences between the 216 three sites were strongly apparent for Pb and present with decreasing strength for Cu followed by Zn as revealed by non-significant station overlap. At the Philip site, Cu and Pb, and to a lesser degree Zn and Cd, in addition, revealed distinct separation between station A4 and the seaward stations A5 and A6. The reference station was distinct from both the Philip and the Blueridge sites in terms of Pb and Zn bioaccumulations while for Cu, overlap was apparent for only the most seaward station at Blueridge. Very little variation occurred between stations at Blueridge with the exception of slight elevations at C5 which was only significant for Zn. 4.4. Discussion The pathway of trace metal uptake is both dependent on the element in question and the species under investigation. The environmental partitioning of an element in addition to an individual organism's exposure routes both play a part. Metal uptake can occur from ingestion of particulates as in the case of suspension feeders or from water exposure as applies in general to aquatic organisms, or from some combination of both. In every case, the exact environmental metal source(s) is species-specific and can even be dependent on other more variable factors such as health condition, reproductive stage, season, and local habitat. While some examination of metal bioaccumulation associations with environmental variables and potential sources of contributing metals follows, the primary intent of this study was to assess the degree of metal pollution inclusive of ecotoxicological considerations. As such, the focus is on the trace metal bioaccumulations in the selected biomonitors which gives a measure of the metal bioavailabilities. Summed together, the biomonitors and their respective metal bioavailabilities across sources, enable a more generalized ecologically meaningful conclusion regarding metal distributions. Variations in bioavailability data across four species were able to distinguish between sites of different pollution levels for four of the five elements analyzed in the order Cu>Zn>Pb>Cd and demonstrate the elevated levels at the Philip site with its storm related metal discharges. A l l four species revealed significant differences for the first three elements, while Cd tissue levels in three of the four species (all but Fucus gardneri) similarly 217 distinguished between sites. Due to the occurrence of only two species (Balanus and Mytilus) at the Philip outer basin station (A4), and the potential for storm-localized influences at this station with its proximity to the small and dynamic inner basin, a summary table is presented which indicates ranges and statistical findings for the invertebrate species with this station highlighted (Table 4.6). With the exclusion of station A4, Balanus bioaccumulations of Zn and Cd are no longer distinct between the Philip and Blueridge sites although the reference site is still distinct from both of these sites. Mytilus site differences remain unchanged. The overall importance of Cu is emphasized with its consistent and significant elevation at the industrial/commercial site for all four species. Popham et al. (1980) in their study of Mytilus metal bioaccumulations from the same vicinity along the north shore coast of Burrard Inlet in 1979, reported spiked elevations of Cu (15-44 pg/g, dry body weight), Zn (250-805 pg/g) and Pb (30-470 pg/g). They were able to attribute this to runoff contributions from a stormwater outfall. With their extended coastline sampling between the Lions Gate Bridge and the Second Narrows Bridge, they found baseline concentrations of 12-15 pg/g for Cu, 150-220 pg/g for Zn and 2-8 pg/g or lower for Pb. Their maps suggest that the storm drainage outfall of this earlier work could very well be the Philip site sampled in this study. At a minimum, it is within the same local area and represents similar land-use patterns given that drastic changes have not occurred within this area during the 20 year sampling gap. Additionally, comparison of the salinity ranges and distances between sampling stations for the two studies suggests that sampling locations were similar and are comparable. The contrast of 12-15%o between the outfall stations and 21.5-21.8%o of stations starting from 30 m away in the July sampling of Popham et al. (1980) is comparable to the dry spring salinity ranges found in this study between the Philip inner basin (3-13%o) and the ~60 m seaward outer basin stations (A5, A6, 21-28%o). A temporal examination of the storm outflow area indicates that Mytilus Cu concentrations have increased by around an order of magnitude, Zn concentrations have stayed about the same or possibly increased slightly, while Pb concentrations have declined. Considering baseline levels of the earlier study are similar to the Blueridge and Lighthouse concentrations within this study and the noted differences are relatively large in magnitude, these temporal changes seem quite probable despite any potential protocol and/or natural sampling variations. The implication therefore, is that the installation of the oil/grit chambers in the 1990s for pollutant 218 mitigation of stormwater runoff has not been effective for the control of Cu or Zn and in fact Cu discharges may have increased. The decline in Pb may be attributable to the oil/grit chambers, the nationwide environmental decreases following the phase-out of leaded-gasoline in the 1980s, or some combination of both. Mussel metal concentrations from different regions in the world are presented in Table 4.7 and include both Mytilus edulis and M. galloprovincialis in addition to M trossulus. Validation for the comparison of these three species is supported by the lack of interspecific concentration differences found at a site off the Columbia River, U.S.A. (O'Connor, 2002) and further confirmed in experimental work by Blackmore and Wang (2003). In order to serve as guidelines of general elevated metal pollution levels, the Median International Standards (MIS) for edible shellfish and the 85 t h percentile concentrations from two Mussel Watch programs in the U.S. are provided in Table 4.8. The Mytilus N i range (1.3-2.8 ug/g) falls within the median range of Mytilus concentrations found in other studies and is lower than the 85 t h percentile concentrations of both California and the national U.S. programs. Cu concentrations at Blueridge (13.8-17.0 ug/g) are at about the 85 t h percentile level in the U.S. and fall within the lower range of concentrations found in several studies from Europe and the middle-east, while Philip concentrations (92-255 ug/g) represent an order of magnitude increase, exceeding the MIS, and fall in line with the highest concentrations found in polluted coastal areas of Korea and Japan (172, 385 ug/g). Zinc concentrations at Blueridge and Lighthouse are similar to concentrations found within dense urban areas (>150 ug/g, Table 4.7) and more polluted areas of the U.S. (200-290 ug/g; Table 4.8). Philip concentrations (815-1010 ug/g) are at the high end of the more polluted areas of Europe and Asia (200-1200 ug/g) and exceed the MIS. Cd levels at Philip are either at or above the pollution indicator concentrations of the U.S. mussel watch programs. The more polluted A4 and A5 stations exceed the MIS and are similar to the most polluted areas of Europe (0.1-8 ug/g) but lower than some of the industrialized or more polluted areas of Asia and Europe (12.9-100 ug/g). Similar to Cd, Pb concentrations at Blueridge and Lighthouse (0.634-3.31 ug/g) are low compared to many urban/industrial areas around the world; however, Philip concentrations (12.7-18.5 ug/g) exceed the MIS and are elevated compared to the 85 t h percentile concentrations of the U.S. programs. While these levels are relatively high, they are far below some of the more polluted areas in Europe, Asia and Australia (20-219 200 pg/g) and (100-1000 pg/g) from an industrialized region of Norway (Julshamn & Grahl-Hielsen, 1996). For both macroalgal species and Balanus, comparison with like species across different regions in the world is compromised due to either identification difficulties arising from complex taxonomies, as in the genus Enteromorpha, or the lack of studies focusing on the species that are indigenous to the North Vancouver sampling areas as for Balanus glandula and Fucus gardneri. While in general, interspecific comparisons are not valid for ascertaining absolute metal differences, a very coarse examination of the current species concentrations within a global context is attempted with the presentation of literature values for closely related species (Tables 4.9-4.11). This is provided solely for the purposes of indicating broad metal pollution levels. For Enteromorpha, Seeliger and Wallner (1988) did not find metal bioaccumulation differences in a laboratory study of four different species and so comparisons for this genus alone may be justified (Seeliger & Wallner, 1988; Villares et al, 2001). Cu and Zn in other studies reveal similar respective macroalgal concentrations and relative differences between contaminated and reference sites as in this study (Tables 4.9 and 4.10). Enteromorpha at Philip is slightly enriched relative to other studies which may be due to species differences or the morphological susceptibility to metal contamination by fine grained particulates which was a problem encountered by Villares et al. (2001). N i , Cd, and Pb in F. gardneri are all below concentrations presented in the literature. Unlike the macroalgal metal concentration ranges across related species, metal content of Balanus spp. and other eulittoral barnacle species vary broadly (up to two orders of magnitude) from species to species for each of the five elements presented within similar regions (Table 4.11). Interspecific comparisons are not attempted here other than to note that Balanus glandula metal concentration ranges from this study are all within the large ranges observed amongst other species from various locations. 4.4.1. Comparative Bioaccumulation In general, Balanus Cu and Zn bioaccumulations exceeded those of Mytilus and the two macroalgal species by a minimum of one order of magnitude and up to two orders of 220 magnitude for Zn. This finding was expected considering the availability of solution and particulate metal sources compared to a sole solution exposure route for algae and the non-regulation of these two elements in Balanus (Powell & White, 1990; Rainbow, 1985) compared to the partial regulation exhibited by Mytilus (Phillips, 1976; Phillips & Rainbow, 1988, 1989). The similarity of N i bioaccumulation ranges is likely related to the low concentrations found at all sites. The two non-essential elements, Pb and Cd, both exhibited bioaccumulation ranges that were slightly different across species. Pb bioaccumulation ranges were similar for all four species with the exception of the two fold decrease for Fucus at Philip. Conversely, Cd in Enteromorpha was about an order of magnitude lower than the other three species. Invertebrate Cd bioaccumulation ranges were somewhat elevated compared to the two macroalgae with a three-fold elevation of Balanus over Mytilus at the Blueridge and Lighthouse sites. Some form of assessment was desired in order to examine whether different species were responding to the same bioavailable sources for a particular metal and to provide information on bioaccumulation patterns and strategies. Comparing absolute metal concentrations between species is not appropriate due to the potential for even closely related species within the same locations to significantly differ in their total metal accumulations (Lobel et al, 1990; Rainbow & Phillips, 1993; Rainbow et al, 1993). Rank correlation analysis has been recommended and successfully performed as a means of interspecific comparisons between sites (Chan et al, 1990; Phillips & Rainbow, 1988); however, the availability of only three sites makes this type of analysis impractical. Instead, as in Rainbow et al (2004), a correlation analysis of log transformed metal concentrations was performed between species for each metal using species collected from the same stations. Cu and Pb bioaccumulations were significantly (p<0.05) correlated between the two invertebrate species (Fig. 4.2). Similarly, Cu and Pb, in addition to N i were significantly correlated between the two macroalgal species (Fig. 4.4). At a minimum, this suggests that the bioavailable sources of at least some elements are very similar for species within the respective species pairs. The small number of stations in the study (10) in addition to the even smaller number of stations where species collections overlapped (4-7) reduces the power of the correlation analyses to detect significant differences i f they exist. Consequently, the lack of significance may not be biologically meaningful. Understanding 221 the bioaccumulation strategies of individual species and an examination of findings from other studies provides some assistance in this regard. For example, mussels are known to partially regulate Zn and Cu, whereas barnacles continuously accumulate both of these metals from their environment (Powell & White, 1990; Rainbow et al, 1990). This in itself seems to suggest that weaker correlations would be found for these elements as compared to elements which are not regulated in both species. Of course no correlation would result between an element's bioaccumulation in two species where complete regulation occurred in one and no regulation occurred in the other. Two studies that both compared metal bioaccumulation between Mytilus trossulus and the eulittoral barnacle, Balanus improvisus in Hong Kong coastal waters (Phillips & Rainbow, 1988) and on the Baltic coast of Poland (Rainbow et al., 2004) were able to demonstrate interspecies correlations for Cu, Zn, Pb but not Cd. While care should be taken in drawing any firm conclusions, Fig. 4.2's non-significant metal data distributions between Mytilus and Balanus shows Cd with a much weaker relatedness (r=0.05) than either N i (r=0.36) or Zn (r=0.67). At least for these two species, it appears that Cu and Pb may be available for accumulation from similar sources. The further comparison of invertebrates with macroalgae (Fig. 4.5) shows Pb followed by Cu (5 and 4 incidents of significance in 6 pairwise correlation analyses) are most commonly correlated between species. Additionally Zn accumulation between Mytilus and both macroalgal species is significant. The significant correlations between Mytilus and both macroalgal species suggest the importance of both Zn and Cu uptake from solution to all three of these species. The same can be said for Pb with 3 of 4 macroalgal-invertebrate comparisons resulting in significant correlations. More specific conclusions with respect to definitive sources of metal bioavailability are not possible with this data set. 4.4.2. Environmental Metals and Bioaccumulation In order to assess the key environmental metal sources to biota, comparisons were sought of both dissolved metals in the water column and surface sediment (oxic) metals with biota metal accumulations. A l l organisms within this study spend significant portions, i f not all of their existence, bathed in the waters of their aquatic environments. In addition to 222 exposure within the solution phase, filter-feeding invertebrates such as Mytilus and Balanus in this study, are also potentially exposed to metal uptake through food particles. Metals in suspended particulate matter are included in the analysis. Additionally, sediment metal accumulation from overlying waters not only represents a time-integrated process but can also be indicative of metal exposures to organisms close to the sediment-water interface, and potential exposures to these organisms during resuspension events. The examination of the sediment metals in relation to at least the epibenthic organisms in this study is potentially useful given the proximity to stormwaters shown earlier (Chapter 2) and related contributions from dissolved and particulate-bound metals. The biota used in the sampling collections would have been most recently exposed to the light storm spring conditions of 2001. B. glandula ages were 1-4 months (Chapter 5), and estimated for M. trossulus from growth rates on the southern coast of British Columbia (Yanick et al, 2003; Zis et al, 2004) to be between 4 and 12 months. Additionally, metal accumulation in mussels can occur quickly and then plateau after a period of time as short as three months (Regoli & Orlando, 1994; Webb & Keough, 2002). Water ancillary and DGT data for the three month period prior to biota collections are reported in Tables 12 and 13. As reported in earlier chapters and Appendix A, the water (alongside spring sampling data from 1999) and sediment metal data used in the analysis are included here in Tables 4.14-4.18 and 4.19-4.23 respectively. The dissolved water data, although representing two days' sampling, was used as reflective of the average conditions for the spring period of 2001 and representative of recent exposure conditions for resident biota. The dissolved metal concentrations for these two days are mostly within the ranges of other spring dates sampled in 1999 representing dry and light storm conditions. A comparison of seasonal precipitation patterns (Appendix A , Fig. A l ) reveals similar frequencies of dry, light and medium to heavy storm days. The findings reported in Chapter 2 noted that any elevations in dissolved metal concentrations would have only occurred during rainy days (approximately 20% of all days within the spring period for both 1999 and 2001). Consequently, average dissolved concentrations for the spring period, as seen for the spring 2001 DGT measurements, would not have substantially increased over dry period concentrations. The representative nature of the May 21-22 SPM data is less certain and associations to biota metal bioaccumulations are speculative at best. The variable nature in both quantity and quality of the SPM load can 223 vary over very short periods e.g. with tidal cycle (Wang et al, 1996), or with changing storm conditions (Chapter 2, Figs. 2.6-2.17). While the influence on metal bioaccumulations can be extensive (Bendell-Young & Arifin, 2004; Pollet & Bendell-Young, 1999; Wang et al, 1996), this type of SPM investigation was beyond the scope of this work. Plots of log-log metal bioaccumulations versus metal concentrations in water are presented in Fig. 4.6. DGT and dissolved metal concentrations reveal similar trends with bioaccumulations with a somewhat better relationship for DGT than dissolved Cu. A strong association is seen for Cu between DGT and bioaccumulations for all four species. A similar but weaker trend is seen for Cd. Only Mytilus and Enteromorpha show slight elevations at the higher Cd dissolved and DGT concentrations. No pattern is apparent for N i . While DGT-Zn and Pb data are not available, dissolved concentrations reveal the same increasing trend with bioaccumulations that is clear for Pb in all four species and somewhat apparent but weaker for Zn. Positive relationships with SPM metal content were evident for Cu and Pb in both invertebrate species as well as Cd in Mytilus. Cu, Zn, Cd, and Pb-SPM contributions in water were also correlated with Mytilus. Weaker correlations for all SPM metals but Cu, were seen for both Fucus and Enteromorpha. The metal exposure to these macroalgal species is through solution but the larger range of dissolved and DGT Cu and corresponding SPM concentrations would explain this finding. In other words this is likely a reflection of simultaneous increases in water and SPM Cu concentrations rather than a relationship of SPM algal bioaccumulations or possible issues of contamination. Another consequence of field assessments where simultaneous metal elevations occur in both dissolved and particulate compartments is that relationships between solution versus digestion exposures for the invertebrate species are difficult to discern. This is more apparent for Cu than Pb since the latter is strongly particle reactive and as previously reported (Chapter 2), there are only slight elevations in dissolved metal with storm conditions. Coupled with the only slight elevations in dissolved Pb would be the short duration of this solution exposure to biota due to the strong particle scavenging effect and salinity driven flocculation processes which would further reduce colloidal matter and enhance overall deposition to sediments. With dissolved Pb exposures slight and short-lived in the more contaminated areas, the water-biota associations suggest a significant role of ingestion pathways for Pb for both Mytilus and Balanus. Overall biota-water associations 224 (for dissolved and particulate metals) are strongest for Cu and Pb, followed by Cd and then Zn. DGT measurements reveal the same patterns for Ni , Cu, and Cd that are seen in overall water conditions but not in biota. N i bioaccumulations for all species show similarities among sites, whereas differences are seen for Cu and Cd to a less degree, with elevations occurring at the strongly storm exposed stations at the Philip site. DGT was able to better discriminate between sites for both Cd and N i . This is likely due to the low average environmental levels of the latter two elements and natural biological variations. These findings are corroborated by the recent work of Webb and Keough (2002). Despite deployments of mussels and DGT units during different years and different seasons, they found a similar improved discriminating power of DGT (Cu and Cd) over mussel bioaccumulations (Cu). Sediment-biota associations (Fig. 4.7) are not particularly revealing with the small number of stations and only two sites (sediments were not obtained from the reference station). Frequent storm related disturbances and scouring of the small study sites (as revealed by low silt content in Chapter 3) actually produce surface sediment conditions that are not necessarily reflective of recent water metal conditions but dependent on the severity of the storms and duration of relatively calm conditions during post storm periods. Although infrequent, larger storms did occur during the spring 2001 period as indicated by >20 mm (24-hour) precipitation levels as recently as five days prior to sampling (Appendix A , Fig. A l ) . Despite the dynamic conditions of the sites, the conclusions that can be drawn are that biota and sediment Cu and Pb, and perhaps to some extent Cd, show similar increased levels at the more contaminated sites. Biota accumulations are highest at the Philip site for all four metals as is the case with water and sediment concentrations. Furthermore, differences are most notable for Cu in water, sediment and biota. Examination of the relationship between biota metals and sediment properties (Fig. 4.8) reveal the possible importance to Balanus and Mytilus of decreased extractable Fe or Mn (i.e. Fe or Mn-oxides) to both Pb and Cu bioaccumulation and reduced organic carbon to Pb assimilation. In other words, Fe-oxides may be acting to lessen Cu uptake, whereas Fe, Mn-oxides and organic carbon may all play a role in the reduction of Pb uptake in suspension feeders. Similar conclusions were reached by Montouris et al. (2002) through the statistical assessment of heavy metal bioavailability models linking sediment characteristics to invertebrate metal bioaccumulation. They used 225 data from various regions of the U.S. and Greece, and were able to explain most of the variation in bioconcentration factors of Pb and Zn with organic matter concentrations and for Cu with iron concentrations. In general, they found that sediment oxides were more important factors in metal bioavailtity than organic carbon. These results and the associations observed in this study are also in agreement with recent work by others (Dong et al, 2000; Tessier et al, 1996; Trivedi & Axe, 2000) . A further observation with this data set was the increased invertebrate Pb bioaccumulations seen with increased sand content. This corroborates the suggestion made above with respect to the possible importance of ingestion pathways for Pb uptake; areas with higher particulate exposures (lower overall deposition of SPM as evidence by increased sand in underlying sediments) led to increases in Pb bioaccumulations in Mytilus and Balanus. Other factors which also have the potential for influencing metal bioavailability are temperature, salinity, and pH. Temperatures between stations and sites are similar, however salinity and pH do exhibit differences between sites (Table 4.12).