UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The bioavailability of trace metal contaminants in the Brunette River watershed Yuan, Yi 2000

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

Item Metadata

Download

Media
831-ubc_2000-0622.pdf [ 7.75MB ]
Metadata
JSON: 831-1.0089693.json
JSON-LD: 831-1.0089693-ld.json
RDF/XML (Pretty): 831-1.0089693-rdf.xml
RDF/JSON: 831-1.0089693-rdf.json
Turtle: 831-1.0089693-turtle.txt
N-Triples: 831-1.0089693-rdf-ntriples.txt
Original Record: 831-1.0089693-source.json
Full Text
831-1.0089693-fulltext.txt
Citation
831-1.0089693.ris

Full Text

THE BIOAVAILABILITY OF TRACE METAL CONTAMINANTS IN THE BRUNETTE RIVER WATERSHED by Yi Yuan B.E., Beijing University of Chemical Technology, 1982 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T OF T H E R E Q U I R E M E N T S F O R THE DEGREE OF M A S T E R OF SCIENCE  in THE FACULTY OF GRADUATE STUDIES (Resource Management and Environmental studies) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH C O L U M B I A August, 2000 © Y i Y u a n , 2000  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  Department of Resource Management and Environmental Studies The University of British Columbia Vancouver, Canada Date September 1. 2000  Abstract Urban expansion with the development of our transportation network and its associated impervious surfaces, that readily transport surface runoff to our urban streams, has become the major non-point source of pollution in our urban watersheds. Trace metals and organic contaminants are of particular importance due to their potential toxic effects on aquatic organisms. The association of these contaminants with suspended solids and bed sediments, and the dynamic nature of the stream environment, has generated an interest in determining if these contaminants are available to aquatic organisms.  The objective of this study was to investigate the potential bioavailability of the trace metals, copper and zinc, in stormwater runoff and the streams of an urban watershed. Two techniques, namely metal speciation and dialysis with resin receiver were employed in the Brunette River watershed located mainly in the metropolitan area of Great Vancouver, British Columbia. Stormwater and baseflow samples were collected along a hydrologic gradient and the trace metals partitioned into chelex resin extractable, dissolved and particulate components. The suspended solids, filtered from samples, were subjected to geochemical sequential partitioning into the exchangeable, easily reducible and residual trace metal phases. Chelex resin placed in dialysis tubes was deployed in streams and sediment traps for periods of a week to complex "available metals" during dry summer baseflow and wet winter stormrunoff periods. Manganese and iron were also investigated since they could affect the geochemical partitioning of the more toxic trace metals. Recent studies have demonstrated that manganese is accumulating in the Brunette watershed probably as a result of the gasoline additive MMT.  The results of this study indicate that both Cu and Zn are at relatively high concentrations in this watershed, compared to the relevant water quality and toxicity criteria. The concentrations of these two metals decreased along the urban hydrological gradient, with highest level in the street surface runoff, decreasing in Still Creek and still further in the Brunette River as a result of sedimentation in Burnaby Lake. There was a high correlation of suspended sediment (SS) and the concentration of the four metals Cu, Zn, Mn and Fe in ii  surface runoff, which indicated that the trace metals were mostly associated with the particles flushed from impervious surfaces and transported in the urban streams.  Analysis of trace metal speciation demonstrates that only about 10 % of Cu was found in the immediate bioavailable form (resin extractable) at all sites. The particulate bioavailable fraction of Cu (exchangeable and reducible components) only accounted for 30% in street runoff and less than 10% in streams. Fifty percent of the Zn was observed in the immediate bioavailable form (resin extractable) in the watershed. The particulate bioavailable fraction of Zn accounted for 11 -24% along the watershed. Both these trace metals showed potential toxic effects to organisms in the Brunette River watershed, especially in streetrunoff and upstream in Still Creek when compared to the relevant criteria. Copper contributed a higher toxic effect to the aquatic life than zinc in this watershed since it was proportionally higher compared to the criteria.  There was no clear relationship between Mn and the bioavailable fractions of Cu and Zn within the watershed observed in this study. However, there was an increase in particulate easily reducible Cu and Zn (manganese bound metal fraction) during stormflow conditions when compared to baseflow.  In stream resin uptake experiments clearly demonstrated that the trace metal uptake rates were high during storm flow, especially when there was a long dry period before the rainfall. However, there was still considerable uptake of trace metals under baseflow conditions. The good correlation between bioavailable metal species in stream and metal uptake by the receiver resin suggested that both approaches are useful in evaluating the bioavailability of trace metals in the urban aquatic environment and can be used as tools to assess the impact of trace metal on the aquatic organisms in the urban receiving water.  iii  TABLE OF CONTENTS Abstract  ii  List of Tables  ix  List of Figures  xi  Acknowledgements  xiii  C H A P T E R 1 INTRODUCTION  1  C H A P T E R 2 L I T E R A T U R E REVIEW  7  2.1 Trace metals in the urban water system  8  2.2 Toxicity of trace metals to aquatic organisms  9  2.2.1 Bioaccumulation 2.2.2 Bioavailability of trace metals 2.2.3 Trace metal toxicity studies 2.2.3.1 Trace metal forms 2.2.3.2 Aquatic organisms 2.2.3.3 Environmental control factors 2.2.3.4 Manganese and iron oxide control the trace metal bioavailability 2.3 Techniques used to determine bioavailable metal species 2.3.1 Separation of metal species in dissolved phase 2.3.2 Speciation of metal species in the particulate phase 2.3.3 Dialysis with resin receiver - simulation of metal uptake by the organisms  10 10 12 12 14 16 17 18 19 20 22  2.4 Trace metal studies in brunette river watershed  24  2.5 Relevant environmental criteria for assessing trace metals  29  CHAPTER 3 M E T H O D O L O G Y  32  3.1 Sampling sites  32  3.2 Trace metal speciation  34  3.2.1 Water sampling 3.2.2 Trace metal speciation 3.2.2.1 Separation of the water samples into dissolved and particulate phases 3.2.2.2 Separation of trace metals in the dissolved phase 3.2.2.3 Speciation of suspended sediment 3.2.2.4 Total metal concentration (TM) 3.2.3 Other water quality data 3.2.3.1 Suspended sediment (SS) 3.2.3.2 pH iv  34 37 38 38 39 40 40 40 40  3.2.3.3 Precipitation data and flow data  40  3.3 Dialysis with resin receiver at field sites  41  3.3.1 Field receiver experiment procedure 3.3.2 Receiver materials 3.3.3 Laboratory analysis 3.3.3.1 Analysis of resin 3.3.3.2 Analysis of trace metal in membrane 3.3.3.3 Exchange capacity and interference test for chelex resin  41 41 43 43 43 43  3.4 QA/QC and tatistics  44  C H A P T E R 4 RESULTS A N D DISCUSSION  45  4.1 Bioavailable metal study by trace metal partitioning 4.1.1 Total water fractionation 4.1.1.1 Grab water samples 4.1.1.2 Intensive stormwater samples 4.1.1.3 Discussion on dissolved metal components 4.1.2 Metal partitioning in suspended sediment 4.1.2.1 Partitioning results 4.1.2.2 Discussion on particulate metal components 4.1.3 Bioavailable copper and zinc in the watershed 4.1.3.1 Fractionation of Cu and Zn 4.1.3.2 Concentrations of bioavailable cu and zn in the watershed 4.1.4 Regulating effects of M n on Cu and Zn 4.1.4.1 Regulating effect on Cu 4.1.4.2 Regulating effect on Zn  46 46 47 52 60 68 68 72 75 76 78 80 80 80  4.2 Bioavailable metal study by dialysis with resin receiver in situ 4.2.1 Trace metal uptake under baseflow condition 4.2.1.1 Copper uptake 4.2.1.2 Zinc uptake 4.2.1.3 Manganese uptake 4.2.1.4 Iron uptake 4.2.2 Uptake of trace metal in stormflow and snowmelt in the 4.2.2.1 Copper uptake 4.2.2.2 Zinc uptake 4.2.2.3 Manganese uptake 4.2.2.4 Ion uptake 4.2.2.5 Trace metal uptake over one rain event 4.2.3 Resin uptake of trace metals in watershed 4.2.3.1 Seasonal variation of resin uptake 4.2.3.2 Spatial distribution of resin uptake 4.2.4 Environmental factors affecting the metal uptake 4.2.4.1 In stormflow condition 4.2.4.2 In baseflow condition  82 82 82 84 85 86 87 87 88 88 90 92 93 95 97 98 98 99  v  field  4.3 Suspended sediment and trace metal transportation 4.3.1 Suspended sediment 4.3.2 Trace metal transport with ss  99 99 102  4.4 Other relevant data  104  4.4.1 pH  104  4.4.2 Flow data and precipitation data  104  4.5 Assessment of trace metal toxicity in the watershed  105  4.5.1 Assessment with relevant water quality criteria 4.5.2 Assessment with bioassy uptake 4.5.2.1 Uptake in street runoff 4.5.2.2 Uptake in stream 4.5.2.3 Uptake in baseflow  105 107 108 108 109  4.6 Discussion on the two approaches  110  4.6.1 Relationship between the two approaches 4.6.1.1 Cu 4.6.1.2 Zn 4.6.2 Method suitability and experimental design 4.6.2.1 Partitioning method 4.6.2.2 Receiver experiment  110 110 Ill 112 112 113  CHAPTER 5 S U M M A R Y  114  5.1 Potentially bioavailable trace metals in Brunette watershed  115  5.2 Methods suitability  117  5.2.1 Water sampling 5.2.2 Receiver technique  117 118  5.3 Recommendation and further work  118  REFERENCES  120  APPENDIX  128  Appendix A Trace metal speciation in the water samples Appendix B Metal speciation in the suspended sediment (ss) Appendix C Stormwater quality (sq) at intervals over the storm event (Nov. 19, 1999) Appendix D Chelex receiver uptake results Appendix E Resin evaluation for exchange capacity and ca & mg interference E.l Exchange capacity test E.2 Interference test for chelex resin  vi  128 130 132 133 140 140 144  List of Tables Table 2- 1 Possible forms of trace metal in water  11  Table 2- 2 Uptake of metals by Chelex-100 in a 3500 M W C O dialysis bag  23  Table 2- 3 Water Quality Objectives in the Brunette Watershed (pg/L total metal)  29  Table 2- 4 E P A Water Target Concentrations (pg/L total metal) (EPA 1983)  30  Table 2- 5 E E C proposed EQO Values for Protecting Freshwater Life against Heavy Metals  31  Table 3- 1 Description of sampling sites  33  Table 3- 2 Water sampling during study period  36  Table 3-3 Field receiver experiments during the study period  42  Table 4- 1 Climatic and hydrological conditions during the sampling and receiver  46  Table 4- 2 Resin exchangeable metal (RE) (ug/L) in baseflow  51  Table 4-3 R E / TD for Cu, Zn, Mn and Fe (Averaged over three stormwater samples)  65  Table 4- 4 R E / T M for Cu, Zn, Mn and Fe  66  Table 4- 5 Summary of resin uptake over four rain periods  98  Table 4- 6 Comparison of total Cu and Zn with Water Quality Objectives (BC Canada) ..106 Table 4- 7 USEPA Water Target Concentrations for total Cu and total Zn  106  Table 4- 8 Comparison of total dissolved Cu and Zn with E E C water quality criteria  107  Table 4- 9 Trace metal uptake rates in street runoff (pg/mm .h )  108  Table 4- 10 Trace metal uptake rate in stream (pg/mm .h)  108  2  2  Table 4- 11 Trace metal uptake rate under baseflow (pg/mm .h )  109  Table A - 1 Speciation for baseflow water  128  Table A - 2 Speciation for stormwater (1)  128  Table A - 3 Speciation for stormwater (2)  129  Table A - 4 Speciation for stormwater (3)  129  Table B- 1 Cu Speciation in SS  130  Table B- 2 Zn Speciation in SS  130  Table B- 3 M n Speciation in SS  131  2  vii  Table B- 4 Fe Speciation in SS  131  Table C- 1 Stormwater quality data at Site #1  132  Table C- 2 Stormwater quality data at Site #2  132  Table C- 3 Stormwater quality data at Site #3  132  Table D- 1 Results in July (7.23 - 7. 30), 1998  133  Table D- 2. Results in August (8.17 - 9.2), 1998  134  Table D-3. Results in Febrary (2.11 -2.18), 1999  135  Table D- 4. Results in October (10.27 - 11.3), 1999  136  Table D-5. Results in November (11.19- 11.25), 1999  137  Table D-6. Results in November (11.19), 1999 (One storm)  138  Table D- 7. Results in January (1.3- 1.10), 2000  139  Table E- 1 Chelex resin exchange capacity  142  Table E- 2 Resin exchange capacity test for Cu solution in 2.3, 1999  143  Table E- 3 Resin exchange capacity test for Zn solution in 2.3, 1999  143  Table E- 4 Formation constants of metal complexes of Chelex resin functional group (EDTA)  145  Table E- 5 Solution of Cu+Ca+Mg (mg/L)  147  Table E- 6 Solution of Cu-Ca (mg/L)  148  viii  List of Figures Figure 2- 1 Brunette River Watershed  25  Figure 2- 2 Trace metal source, sink and transport in the urban area  27  Figure 3- 1 Locations of the sampling sites in Brunette River watershed  33  Figure 3- 2 Field sampling scheme  34  Figure 3- 3 Field samplings and in situ experiments  35  Figure 3-4 The speciation scheme for the water sampling  37  Figure 4- 1 Partitioning of Cu in stormwater sampled at three sites  48  Figure 4- 2 Partitioning of Zn in stomwater sampled at three sites  49  Figure 4- 3 Partitioning of M n in stormwater sampled at three sites  50  Figure 4- 4 Partitioning of Fe in stormwater sampled at three sites  51  Figure 4- 5 Copper partitioning over one storm event (Nov. 19, 1999)  54  Figure 4- 6 Zinc partitioning over one storm event (Nov. 19, 1999)  55  Figure 4- 7 Manganese partitioning over one storm event (Nov. 19, 1999)  57  Figure 4- 8 Iron partitioning over one storm event (Nov. 19, 1999)  59  Figure 4- 9 Ratio of TD/ T M for Cu, Zn, M n and Fe  61  Figure 4-10 Distribution of copper in three particulate fractions  69  Figure 4-11 Partitioning of zinc in three particulate fractions at three sites  70  Figure 4- 12 Distribution of manganese in three particulate fractions  71  Figure 4- 13 Distribution of iron in three particulate fractions at three sites  72  Figure 4- 14 Fractions of particulate Cu in SS at three sites  73  Figure 4- 15 Fractions of particulate Zn in SS at three sites  74  Figure 4-16 Fractions of particulate M n in SS at three sites  75  Figure 4- 17 Copper and Zinc partitioning in five fractions among the sampling sites  77  Figure 4- 18 Bioavailable Cu and Zn levels in the watershed  79  Figure 4- 19 Effect of M n on Zn bioavailability  81  Figure 4- 20 Copper taken up in baseflow by (a) resin and (b) membrane  83  Figure 4- 21 Zinc taken up in baseflow by (a) resin and (b) membrane  84  ix  Figure 4- 22 Manganese taken up in baseflow by (a) resin and (b) membrane  85  Figure 4- 23 Iron taken up in baseflow by (a) resin and (b) membrane  86  Figure 4- 24 Copper taken up in stormwater by (a) resin and (b) membrane  87  Figure 4- 25 Zinc taken up in stormwater by (a) resin and (b) membrane  89  Figure 4- 26 Manganese taken up in stormwater by (a) resin and (b) membrane  90  Figure 4- 27 Iron taken up in stormwater by (a) resin and (b) membrane  91  Figure 4- 28 Metal uptake by resin - over one storm  93  Figure 4- 29 Metal uptake by membrane - over one storm  93  Figure 4-30 Spatial and seasonal variation of Cu, Zn, M n and Fe uptake by resin  95  Figure 4-31 Variation of TSS concentration over one storm at three sites  100  Figure 4- 32 Corresponding rainfall intensity and stream discharge  100  Figure 4-33 Trace metal concentration change with TSS over a storm event  103  Figure 4- 34 Rainfall and flow data over the study periods in the Brunette River watershed 105 Figure 4-35 Correlation of Cu uptake rate to TD-Cu and RE-Cu  111  Figure 4-36 Correlation of Zn uptake rate to TM-Zn, TD-Zn, and RE-Zn  112  Figure E- 1 Resin exchange capacity test for Cu (a) and Zn (b)  141  Figure E- 2 The effect of calcium and magnisium on the complaxition  146  Figure E- 3 The effect of calcium on the complaxition of copper on Chelex resin  146  x  Acknowledgement  M y heartfelt thanks go to Professor Ken Hall, my supervisor, for providing me the research opportunity, for helping me in field samplings and laboratory experiments, and for financially supporting me to attend an international conference. His countless help, valuable comments and instructions, and constant encouragement made this thesis possible.  M y heartfelt thanks also go to Professor Les Lavculich and Professor Hans Schreier. They provided me not only with valuable comments and suggestions, but also with the research facilities. Professor Lavkulich provided me with laborotary facilities in the Department of Soil Sciences to enable me to analyze the large amount of samples.  Thanks further go to Mr. Gevan Mattu, my fellow postgraduate student, for many valuable discussion and helpful comments. Thanks are also due to all my fellow postgraduate students in IRE for their efforts to make a friendly and supportive environment.  Thanks also go to Paula D. Parkinson, Susan Harper and Keren Ferguson in both the Environmental Engineering Laboratory and the Soil Science Laboratory for their guidance and assistance of all my analysis work.  Last, but not least, my thanks go to my husband Jingnan Guo and my son Jerry for all their supports and encouragement during my study.  xi  Chapter 1 Introduction  The growth of population and intensive human activities are a typical feature of the urban watershed. Following the rapid industrial expansion since World War II, industries have been the major contaminant sources, referred to as point sources, and responsible for much of the pollution in the urban watershed. Over the last two decades, pollution abatement efforts have identified, regulated and mitigated many of these point sources to our waterways (McCallum 1995). However, the increase of non-point source pollution, in the form of urban runoff, has been recognized as a major continuing source of contaminants to urban streams (Gibb et al. 1991). As many investigators have pointed out, pollution from non-point sources in an urban area can account for over half of the contaminant load entering the water system (Whipple et al. 1974, Wanielista et al. 1977). Colston (1974) also found significant contribution of urban non-point sources to receiving creeks. He showed that even if 100% of the pollutants from raw municipal wastewater was removed, the total reduction of contaminants discharged to the receiving creek would only be 52% for chemical oxygen demand (COD), 59% for biochemical oxygen demand (BOD ), 5% for suspended solids, and 9-43% for toxic trace 5  metals such as Cr, Cu, Pb, and Zn (Colston 1974).  Numerous contaminants exist in urban runoff, they include hydrocarbons, nutrients, pesticides and toxic trace metals (Hunter et al. 1979, Hoffman et al. 1982, Bedient et al. 1980, Murphy and Carleo 1978, Wilber and Hunter 1979). Contaminants, such as trace metals, are deposited on urban impervious surfaces and are transported to the receiving water during stormwater runoff periods (MacDonald et al. 1997, IRE 1997). The data from the US Environmental Protection Agency's (EPA) nationwide urban stormwater runoff monitoring  1  program indicated that the most prominent toxic trace metals, namely copper, zinc and lead, exceeded the freshwater acute toxicity criteria 50, 12, and 27% of the time, respectively (Cole et al. 1984). Similar levels of toxicity criteria exceedance have been reported in various urban watershed studies (Morrison 1985, Jarvis 1992, MacDonald et al. 1997). In British Columbia the monitoring data from the Brunette River watershed indicated that among 13 stations within the watershed, copper concentration exceeded the 2 ug/L water quality objective (WQO) in 60 of 103 samples (58%) under baseflow conditions. The average zinc concentration exceeded 30 ug/L WQO in the Still Creek, a tributary to the Brunette, and an average copper concentration of 50 pg/L was observed during stormflow (MacDonald et al. 1997). The sediments in streams and lakes in the watershed also showed high level of trace metals. McCallum (1995) reported that the contamination of Burnaby Lake surface sediments with Pb, Cu, Zn, and Cd is indicated by their respective enrichment factors of 10, 6, 3.8, and 1.5, respectively, while stream contamination is indicated by the failure of all 33 stream stations to meet watershed sediment criteria for Pb, Cu, Zn.  The ultimate concern of trace metal contaminants in the receiving water is their toxic impact on the aquatic organisms and fish species (Morrison et al. 1987, Revitt et al. 1987, Kushner 1993, Bindra and Hall 1977). The toxic impact of the trace metals is caused by the accumulation of metals in the organisms. This process is referred to as bioaccumulation. Traditionally, total trace metal concentrations in water were used as environmental criteria for controlling metal pollution (Morrison 1985). However, researchers have demonstrated that the total trace metal concentration provides little information on metal toxicity, since only metal species termed "bioavailable metal form" play a key role for the toxicity estimation of metals to organisms in the water system (Bindra and Hall 1977, Morrison et al. 1987).  Trace metals in water generally exist in two phases. In the dissolved phase the metal is dissolved in the water column. In the particulate phase, the metals are either bound by the suspended particles, or deposited in the bed sediment. It has been recognized that the dissolved metal fractions are considered potentially more hazardous because they can be taken up directly by the aquatic organisms (Bindra 1983, Morrison 1985). Although the  2  sediment-bound metals appear to be of little concern upon preliminary considerations, later investigations have shown that particulate metal might prove to be very important. This is due to the fact that sediments often contain 100 times higher concentrations of trace metals than that in the water column (Bindra 1983), and the release of trace metals from sediment into the water may happen under certain environmental conditions. Also ingestion of sediment particles by filter feeding organisms can provide another route for bioaccumulation.  Morrison (1985) suggested that storm runoff over the urban impervious surface brought large amount of trace metals into the receiving water, of which metals in the dissolved phase might cause an acute toxic impact on the organisms. On the other hand, the metals in particulate phase led to the accumulation of these metals in the stream sediment. Aquatic organisms can either directly ingest these sediments or assimilate the dissolved metals released from sediment to the water column.  The existing trace metal forms and their concentrations in the water and sediment can be affected by various physical-chemical factors. In an urban watershed, these factors include the metal sources discharged from different land use areas, such as the industrial, commercial and residential lands; the transport pathways by the urban hydrological characteristics, such as creeks, drainage channel, and the lakes; and the water chemistry, such as pH, hardness, dissolved oxygen; as well the effects of inorganic and organic substances and other metal species. A l l of these are considered as regulators of trace metal species and their toxic effect in the urban aquatic environment (Morrison 1985, MaCdonald et al. 1997, IRE 1997, SoloGabriele and Perkins 1997, Yuan et al. 1999)  Water chemistry plays a key role in metal bioavailability in the water system. It is generally accepted that trace metals will be more toxic at lower pH values, where most of the trace metals will exist in a free ionic form. Increases in pH may lead to the formation of insoluble hydroxides and oxides of the metal which prevents the metal of being assimilated by the aquatic organisms, thus reducing the toxicity of the trace metals (Gadd and Griffiths 1978). A laboratory bioassay study supported this conclusion by finding an increase of metal toxicity with a decrease of pH from 8.0 to 5.0 (Hall and Anderson 1987). The effects of water  3  hardness on toxicity of trace metals to fish and invertebrates have also been noted. Moore and Ramamoorthy (1984) reported that the increasing hardness could reduce the toxicity of C u , P b , N i , and Z n . Collins and Stotzky (1989) also suggested that metallic cations are 2+  2+  2 +  2+  more toxic to many microorganisms in soft than in hard water, because of the higher bicarbonate / carbonate contents of the latter. The redox condition of the sediment has been shown to be an important factor in controlling the chemical exchange reactions, which affect the transport of trace metals across the sediment water interface (Bindra 1983). Gibbs (1973) found that Fe and M n existed in sediments as hydrous oxide coatings on particles. One typical example is that the release of manganese (Mn) and iron (Fe) from the sediment to the water column was accompanied by the depletion of dissolved oxygen in the water (Chen et al. 1976). Morgan and Stumm (1964) reported that under anoxic conditions, Fe and M n in sediment become unstable and the mobilization of associated trace metals occurs. Singh and Subramanian (1984) concluded that manganese and iron play a significant role in influencing the distribution and transport of a variety of metal ions in the water system, since manganese oxides, along with iron oxides, are well known as the scavengers of the trace metals.  Trace metal pollution has been studied in the Brunette River watershed in the metropolitan region of Vancouver, British Columbia, since 1970's. Considerable data of water quality, urban hydrology, and trace metals both in water and in sediment have been collected over a twenty-year period (McCallum 1995). High concentrations of manganese have been observed in the stream sediment within the Brunette River watershed in recent years. The methylcyclopentadienyl manganese tricarbonyl (MMT), a manganese based gasoline additive used in the automobile since 1970's, may account for this finding (Mattu unpublished data). The high level of manganese in the Brunette water system may be the consequences of combustion products from M M T , possibly in the forms of Mn-phosphates, Mn-oxides, and Mn-sulphates (Ethyl 1997). These are deposited on the urban impervious surface, and discharged into the water by surface runoff. Since manganese is one of the regulators for the trace metal toxicity, the examination of metal manganese in the water system is also an important part of trace metal bioavailable studies.  4  This thesis focuses mainly on the examination o f bioavailable trace metal species in the waterway and their distribution under different flow and chemical conditions in the Brunette River watershed. T w o approaches were employed in the study. The first is the speciation of trace metals in the water samples. The information on possible forms o f bioavailable trace metals existing in urban storm runoff and in streams can be obtained with this physicochemical analytical technique. The second is dialysis with receiver resin technique. Through this technique, the uptake o f trace metals by aquatic organisms is simulated. The results o f this approach are to provide more meaningful expression o f the potential availability o f trace metal contaminants in an urban watershed, especially during intermittent stormflow conditions.  This study is expected to provide more information on trace metal partitioning in the typical urban water system, and gain insight into the spatial (along the metal transport route from street runoff to stream) and temporal (stormwater, snowmelt, as well as baseflow) distribution o f the potential toxic effects o f the trace metals to the aquatic organisms. The specific objectives o f this study are:  (1) To characterize the bioavailability o f trace metal contaminants (Cu, Z n , M n , and Fe) in the Brunette River watershed by following the hydrological gradient through the watershed. (2) To collect information on the potential metal uptake by the organisms during baseflow, stormwater runoff, and snowmelt conditions using dialysis with the receiver resin technique. (3) To investigate the influence o f manganese on the bioavailability o f trace metal copper and zinc in the water stream.  The thesis is organized in the following sequence. In Chapter 1, the general introduction o f trace metal bioavailability in the aquatic system and the objectives o f this study are presented. Chapter 2 reviews the relevant literature on research on metal toxicity, the techniques used to measure bioavailable trace metals, and mechanisms o f trace metal transportation in urban water system. The background information o f metal contamination in  5  the Brunette River watershed and relevant water quality are also reviewed in this chapter. Chapter 3 describes the methodology of this study, which includes sampling programs, in situ  dialysis with resin receiver experiment, and the speciation techniques. Chapter 4 presents  the results of the trace metal speciation and the in situ resin receiver experiments. The bioavailability of trace metals Cu and Zn in Brunette River watershed, the effects of metal manganese in regulating the partition of trace metals Cu and Zn in the water system, the control factors affecting the trace metal uptake and partitioning, and the metal transport in the surface runoff are also discussed in this chapter. Chapter 5 draws the conclusions from this study and makes recommendations for further monitoring and research.  6  Chapter  2  Literature Review  Stormwater runoff in the urban watershed contributes significant amounts of trace metals to the receiving waters by washing off from impermeable surfaces. The monitoring data in various urban watersheds demonstrate that these intermittent trace metal discharges may exceed defined chronic and acute criteria for protection of aquatic life when compared to relevant water quality standards (Morrison 1985). Florence and Batley (1980) pointed out that in the case of metal toxicity, the determination of the total concentration of trace metal in a water sample provided very little information about the toxicity of the water. It is generally accepted that only the metals in specific forms can be assimilated by the organisms, which are termed bioavailable metal forms.  In natural surface water, trace metals present in a wide range of chemical forms, which are generally grouped into dissolved and particulate phases (Morrison 1989). The dissolved phase consists of hydrated ions, inorganic and organic complexes, together with species associated with heterogeneous colloidal dispersions and organometallic compounds. On the other hand, the particulate phase contains elements in various chemical associations, ranging from weak adsorption to binding in the detritus mineral matrix. The definition for these two phases is based on the conventional concept that the "dissolved" metals are all metal species that pass through 0.45-pm filter, and the remaining metals on the filter are called "particulate" metals.  7  When relating metal levels to toxic effects in the aquatic system, it is important to recognize the phase of the metals. It has been found that the dissolved metal can be taken up by the aquatic organisms more directly and easily, which may cause acute toxicity if there are  sufficient dissolved metal species in the water column (Morrison 1989). However, it has been noted that high levels of trace metals are in the particulate phase either suspended in the water column or deposited into the bottom as the bed sediment depending on the particle size. In the urban watershed, a considerable loading of dissolved metals in storm discharges has been observed (Morrison 1985). The particulate metal may also be mobilized into the dissolved phase under certain hydrological and chemical conditions, or ingested directly by the organisms. The accumulation of trace metals in the organisms through uptake of both dissolved and particulate metal species cause the toxic impact on the aquatic ecological system.  2.1 Trace metals in the urban water system Trace metal contaminants have been regarded as the typical non-point source pollution to the receiving water from urban watershed (Macdonald et al. 1997, IRE 1997>. The impacts of trace metal pollution on the urban water system is determined by the trace metal load from the urban watershed area and the metal concentration in the urban stream. Considerable research has been conducted on the heavy metal pollution in urban watersheds (Sartor and Boyd 1972, Hall et al. 1976, Harrison et al. 1981, Morrison 1985, Morrison and Revitt 1987, Jarvis 1992, Solo-Gabriele and Perkins 1997). These studies demonstrated that trace metal characteristics, the availability of the metals from the urban contributing area, and the urban hydrological conditions are the major factors which control the cycling and transportation of the heavy metals in the urban environment.  Urban impervious surfaces such as highway, street and parking lot are recognized as the major metal contributing area, where the heavy metal emissions from urban transportation are deposited on these surfaces and transported to the receiving waters through surface runoff  8  (Morrison and Revitt 1987). Sartor and Boyd (1972) found that highway surface sediments often contain elevated levels of trace metals. Similar observations were made by Harrison et al. (1981), namely that heavy metals on the road surface, derived mainly from vehicular emissions, are largely adsorbed onto or associated with highway particulate. A monitoring program conducted in 1997 (Hall unpublished data) in the Brunette River watershed, a highly urbanized area in B.C. Canada, demonstrated that the discharges of metal zinc and the total suspended sediment in storm runoff were highly correlated. These findings were also supported by other studies (Morrison 1985, Jarvis 1992, Solo-Gabriele and Perkins 1997). It is expected that the suspended sediments play the key role on the metal transport during storm events.  The rainfall volume, rainfall intensity, watershed size, and slope as well as the antecedent dry period have all been identified as controlling factors in regulating the loading of heavy metals from an urban area (Sonzogni et al. 1980, Jarvis 1992, McDonalds et al. 1997). Sartor and Boyd (1972) concluded that the quantity of materials existing on the urban street was dependent on the length of time elapsed since the last rain. Jarvis (1992) suggested that washoff of the highway pollutants is generally assumed to be a function of the amount of pollution on the highway surface at the time of the storm, and the rainfall intensity.  Considerable research and modeling have also been done to predict the trace metal loading in the urban watershed, and the transportation of trace metals in the urban watershed (SoloGabriele and Perkins 1997, Yuan et al. 1999).  2.2 Toxicity of trace metals to aquatic organisms  Bioavailable trace metals refer to the fraction of the trace metals in their surrounding environment that are able to be taken up by organisms. The surrounding environment may include water, sediment, suspended particles, and food items (Spacie and Hamelink 1984). Toxicity occurs when the organisms are unable to cope with the trace metals that are taken up and accumulated in their body. The toxicity of the trace metal to the aquatic organism is  9  determined by the bioavailability of the trace metal species in the water environment, and the uptake of the metal by the aquatic organisms. The definitions of toxicity and bioavailability of the trace metals and their relations are reviewed as follows.  2.2.1 Bioaccumulation The toxic impact of trace metals on the aquatic organisms is due to the accumulation of trace metals in the organisms, which is referred to as bioaccumulation. Bioaccumulation occurs when the uptake of a chemical by an organism exceeds its elimination (Spacie and Hamelink 1984).  Two mechanisms are involved  in the bioaccumulation  process, namely  bioconcentration and biomagnification. The first mechanism refers to the process that, through direct adsorption and/or partitioning from the aquatic environment, the contaminants dissolved in water or bound to the sediment are taken up directly by the organisms and accumulated in their body (Tinsley 1979). The second mechanism is due to the pollutant distribution in different levels of the food chain, and is referred to as food chain biomagnification. Tinsley (1979) stated that accumulation in a food chain depends on the movement of mass through the food chain, and on the fact that less than 50% of this mass is converted into tissue of organisms in the next higher level. The uptake of contaminants in each level in the food chain leads to the accumulation of the contaminants in the higher level of predator, and finally the toxic effect to the organisms. A typical example of trace metal bioaccumulation was found for mercury (Hg), or more accurately methylmercury, which is accumulated through the aquatic food chain. Even in an extremely low level in natural water, bioaccumulation of mercury occurs. This is because mercury, when methylated, is very effectively absorbed by a variety of aquatic organisms. Wren et al. (1995) reported that aquatic organisms accumulated mercury from food, water and sediments. They found that even though most of the mercury in water is in the divalent inorganic form, the mercury in the methylated forms predominates in tissues of aquatic biota. Cox et al. (1975) measured concentrations of total mercury and methylmercury in aquatic biota collected from a stream receiving a continuous input of 0.8 ug of mercuric chloride per liter. They found that the bioconcentration factors of mercury ranged from 75 for water boatmen to 29,000 for  damselfly nymphs.  10  2.2.2 Bioavailability of trace metals Observations of trace metal toxicity impact demonstrated that the accumulation of metals in organisms was correlated to the metals in some specific forms rather than the total metal concentration. The term "bioavailability" of trace metals existing in the aquatic environment may represent those metal species that can be taken up and accumulated by the aquatic organism. Further partitioning of metals in dissolved and particulate phases shows the possible forms of trace metal in surface water, which are presented in Table 2-1 (Florence and Batley 1980, Morrison 1989).  Table 2- 1 Possible Forms of Trace Metal in Water Approximate Phase  Chemical form  Example  diameter (nm) (Molecular weight)  Simple ionic species  Zn(H 0) 2  0.8  2+ 6  (0-150)  Dissolved  Simple inorganic or  Zn(H 0) C l  organic complexes  Cu-glycinate  (0-150)  Weak complexes  Cu-Fulvic acid  2-4  Lipid-soluble complexes  CH HgCl  (0-150)  Adsorbed on inorganic and  Cu-Fe(OH) - humic  10-500  organic colloidals  2  +  5  3  3  acid  1-2  (150-10 ) 7  Pb -humic acid 2+  Particulate  Metals adsorbed onto or  Retained by  >450  contained within clay  0.45 um filter  (>10 ) 7  and/or organic particles  Three groups of metal species in dissolved phase have been identified as having contributions to the bioavailability of the metals, and are listed in Table 2-1 (Florence and Batley 1980, Morrison 1989). They are metals in simple ionic form, metals in weak complexes, and metals in lipid-soluble complexes. The reason to classify these three metal forms as "bioavailable" is based on the transport mechanisms of trace metal through bio-  11  membrane, e.g. transport by passive diffusion; transport by carrier-mediated pathways; and active transport process. Morrison (1987) suggested that the passive diffusion might dominate metal transport across the membrane of a living cell. In the passive diffusion process, the rate of diffusion is determined by the molecular size. As such, free metal ions can diffuse through a membrane fast, while the colloidal species cannot move easily across the membrane because of their larger size. For lipid-soluble complexes, a rapid penetration pathway was provided by the passive diffusion process so that the highly toxic effects of these species to the organisms were observed. The other two transport routes such as carriermediated pathways and the active transport process are also favored for the transport of free and weakly complexed metal forms. More transport mechanisms were described in detail by Luoma (1983), Viarengo (1985), Spacie and Hamelink (1984) and Morrison (1989).  As mentioned above, the soluble form of the trace metal accounts for only a small portion in the water system, most of the metals actually exist in the particulate phases that are bound to the sediment. Therefore the ingestion of particulate trace metal by the aquatic organisms is also a very important part for the studies of trace metal bioavailability.  2.2.3 Trace metal toxicity studies Trace metal toxicity studies have found that various factors can affect trace metal accumulation and its toxicity to the aquatic organism. They include (1) existing trace metal forms in the water environment, (2) the type of the aquatic organisms exposed to the trace metals, and (3) the physical-chemical conditions of the water. Numerous studies have already been conducted to investigate the effects of these factors on trace metal toxicity.  2.2.3.1 Trace metal forms Trace metals in different forms have different accumulation rates and thus have different toxicity to organisms. Bindra and Hall (1977) monitored Pb levels in chironomids sampled from two watersheds - one in urban and one in rural areas of Lower Mainland of British Columbia. They found that Pb concentration ratio of animals to sediments was 2.76:1 in the urban watershed, while it was relatively low in rural watershed with the range of 0.27:1 to  12  0.34:1. They concluded that the large difference in the accumulation of Pb in those two watersheds was probably the result of the different forms of Pb in the sediments.  a) Ionic metal The ionic form of dissolved metal is often considered the most readily available and toxic form. In the early toxicity studies, Petersen (1982) found that free Cu ion was toxic in the range of 10" to 10" mole to the freshwater alga, Scendesmus quadricauda. Zevenhuisen et 10  12  al. (1979) reported that the concentrations of 10" to 10" mole free Cu ion were found to 6  10  inhibit bacterium, Klebsiella aerogenes. Freedman et al. (1980) demonstrated that the mortality of the amphipod Hyallela azteca strongly correlates to ionic Pb concentration.  b) Weakly bound metal forms It has been found that the weakly complexed metal species can also be assimilated by aquatic organisms. Theis and Dodge (1979) observed the uptake of CuOH and Cu  by the midge  larva, Chironomous tentans. Guy and Kean (1980) showed that five organic ligands with relatively high Cu stability constants were toxic at a concentration of 10 ~ M Cu. Luoma 8  (1983) found that the toxicity was higher for citric acid ( 1 0  10  M Cu) and ethylenediamine  (10~ M Cu) which have low stability constants with Cu. He suggested that these small 9  organic ligands may transport Cu into the cell via the biological membrane. By using E D T A as the complexing ligand to simulate the uptake by organisms, Allen et al. (1980) found that ZnOH  +  and Z n  2 +  were the toxic species to the alga Microcystis aeruginosa. Both the  carbonate and free ionic forms of Pb have been reported as toxic to rainbow trout (Davies et al. 1976)  Humic substances act as the important complexation ligands for trace metal ions in natural water (Morrison 1989). It is believed that complexation of metals by humic substances reduces the metal bioavailbility, and thus the potential toxicity. Winner (1985) observed that a linear relationship between the LC50 value with increasing humic substance concentration, which reduced both the acute (3-day) and chronic (42-day) toxicity of Cu to the crustacean  D. pulex.  13  C) Metal bound in Sediment Metals bound in sediment are classified as the particulate form. It has been found to be an important source for the metal accumulation in aquatic organisms. Although there was no straightforward relationship between trace metal levels in organisms and total levels of metals in the sediment (Thomas 1997), the high metal levels both in the tissue of organisms and in the sediment in which they were exposed, suggested the uptake of trace metals by the organisms (Stecko and Bendell-Young 1999, Thomas 1997, Bindra 1983). The mechanisms of metal accumulation in organisms through sediment ingestion are complicated and this process is controlled by various factors. One example of sediment ingestion by worms was shown by Luoma and Bryan (1979). They found that Cu released from ingested sediments by stomach acids is readily absorbed by the tissue of the worms. Bindra (1983) reported that some organisms like chironomids, which constituted a major food source for freshwater and estuarine fish, could accumulate toxic trace metals from contaminated sediment substrates. In the process, toxic elements may be mobilized directly from sediments to food chain which is linked to the human aquatic food source.  It has been recognized that there is a clear relationship between the trace metal level in the organisms and the specific metal-binding forms, called 'biological relevant' components, in the sediment (Thomas and Bendell -Young 1998, Bindra and Hall 1977, Bindra 1983, Stecko and Bendell-Young 1999). Thomas and Bendell-Young (1998) found that the copper concentration in the tissue of bivalves correlated positively with ER-Cu (Mn-oxide bound) and RED-Cu (Fe-oxide bound) copper in the sediment. The amount of copper recovered in the R E D phase (RED-Cu) accounted for 38% of the variability in tissue levels. The major interests of the bioavailability studies on the particulate metals have now focused on what kind of metal-binding forms in the sediment, and how easy they can be released to the water column under various physical, chemical and biological conditions, or to the sediment geochemistry.  2.2.3.2 Aquatic organisms Bioavailable trace metal species in the water may not be taken up by all of the organisms living in the water environment. Rand and Petrocelli (1984) reported that substantial  14  assimilation differences might exist among individual organisms. For example, not all organisms respond in a quantitatively identical manner to the same concentration of a toxicant. The effects of such an exposure might vary from very intense in some organisms to minimal or none in other organisms. That is, some organisms may die and others survive with apparently minimal adverse effects. The observation by Bindra (1983) in his microcosm studies indicated that contaminated sediments were most toxic to the opossum shrimp, but were much less toxic to the oligochaetes worm, since the oligochaetes worm appeared to have the higher capacity to mobilize and excrete trace metals from its tissue. Due to this characteristic, some organisms, such as worms, molluscs and aquatic insects, are usually used as indicators to determine the trace metal availability in the aquatic system.  Bryan (1971) proposed that the polychaete worm, Nereis diversicolor, seemed to reflect the extent of Cu contamination of its environment and can be used as one of the suitable indicator organism species for metal contamination of sediments. In his study, polychaetes, sampled from two estuarine sediments had Cu concentration of 407 and 49 p.g/L, had Cu contents of 22 and 9 M-g/L, respectively.  The results of studies on Zn accumulation in oligochaetes are often contradictory. Ray et al. (1979) reported that Zn in Nereis virens remained constant when exposed to water containing 1.0 mg/L of Zn. Guthrie et al. (1979) attributed the high concentration of Zn in polychaetes from a marine microcosm to its ability to accumulate Zn from both sediment and water.  Molluscs are filter feeders or deposit feeders. They depend upon the ingestion of sediment or suspended particulate matter for their food supply. It has been reported that mussels from the waters close to major population and industrial centers often contain high concentrations of Cu (Fowler et al. 1974, Anderson 1977, Popham et al. 1980). Fowler and Oregioni (1976) attributed the accumulation of Cu by mussels to their feeding habit and their slow uptake and excretion mechanisms. Oysters are known to accumulate high concentrations of Cu and appear to reflect the levels in the environment. In 22 weeks of exposure of oysters to contaminated seawater and sediment, the concentration of Cu in the oysters ranged from < 40 p.g/L to 600 p.g/L (Watling and Watling 1976).  15  Freshwater clams and mussels are known to have a high affinity for the metal Zn (Forester 1980). Anderson (1977b) reported Zn concentration in six species of freshwater clam and sediments collected from the Fox River near Chicago. Zinc levels in clams were greater than that in the sediments. Anderson (1977b) found that the distribution of Zn in various organs of two species of clams (Anodonta marginata and Lasmigona complanatd) indicated maximum concentrations in gills and viscera, which suggested that Zn uptake was mainly through food intake.  Chironomids, a kind of aquatic insect, was also found to accumulate the trace metal Cu. Bindra and Hall (1977) found considerable accumulation of Cu by these organisms, created bioaccumulation ratios ranging from 1.1 to 4.5 under field conditions. The accumulation of zinc in aquatic insects was also reported by Namminga and Wilhm (1977). In their study, the zinc levels were found in chironomides and sediments collected from Skeleton Creek, Oklahoma. The mean concentration of Zn in the organisms was 57 pg/L, 3.6 times higher than that in the sediments. Bindra and Hall (1979) studied the distribution of trace metals in various geochemical phases of sediments and in the burrowing oligochaete worms, collected from various watersheds of the Lower Mainland of British Columbia. Sediments were characterized for a number of chemical and physical parameters such as percent organic and inorganic carbon, pH, redox potential, and particulate size distribution. No single parameter appeared to control Zn in the worms.  2.2.3.3 Environmental control factors Numerous factors such as temperature, pH, salinity, hardness, dissolved oxygen etc. in the aquatic environment influence the toxicity of pollutants to organisms. Two selected examples of these effects are illustrated below.  a) Water pH Water pH is an important factor in regulating bioavailable metal forms and their toxicity. Allard et.al. (1987) reported that in normal freshwater pH condition, Cu mostly existed in the  16  particulate phase, and jumped from 60% to 90% when the pH increased from 5.5 to 7. The particulate Zn, on the other hand, increased from 10% to 70% with pH change in the same range. Morrison (1989) proposed that hydrogen ion competes with free metal ion for cellular binding sites at low pH, thereby reducing metal toxicity; whereas above pH 6.5, the free metal concentration decreases due to the presence of Cu hydroxide complexes which may not be toxic. Fleming and Trevor (1989) found that water pH greatly affects copper speciation. Petersen et al. (1984) reported that Cu toxicity to the green alga S. quadricauda increases from pH 5.0 to pH 6.5 and is constant above pH 6.5. Harrison et al. (1986) found that a decrease of pH from 7 to 5 led to a 50 % to 60 % reduction of Zn flux into the alga  Chlamdomonas variabilis. b) Water hardness Water hardness, usually expressed as mg/L CaCC>3, substantially affects metal toxicity. The almost universal observation on the relation of water hardness to metal toxic effects is that metals are less toxic in hard water, provided the pH is kept constant. The effect of hardness varies with the metal, being most dramatic with cadmium, because of its precipitation in hard water (Rattner and Heath 1995). Copper and zinc are intermediate in their interaction with hardness (Brown 1968, Miller and Mackay 1980, Bratley and Sprague 1985a). In the case of zinc, the effect of hardness seems to rest with changes in the fish gills rather that with metal speciation. Increasing hardness reduces the uptake rate of zinc by gill tissue, the prime target organ for zinc becoming lethal (Bratley and Sprague 1985b). Playle et al. (1992) found that calcium (but not CaC03) reduced gill accumulation of copper at pH 4.8, but not at pH 6.3. Apparently, both calcium and hydrogen ions compete with copper for binding sites on gills.  2.2.3.4 Manganese and iron oxide control the trace metal bioavailability Manganese-oxide, as well as iron-oxide have been recognized as important controlling factors in trace metal bioavailability and toxicity. Singh and Subramanian (1984) pointed out that the incorporation of trace metals into hydrous manganese or iron oxide was due to adsorption (specific or electrostatic), ion-exchange or co-precipitation. The low solubility of hydrous M n and Fe oxides together with their large surface area approximately 300 m /g, 2  17  i.e., small grain size (of the order of 100 A ) and favorable surface change in the pH range generally found in natural aquatic systems make these hydrous oxides efficient scavengers of trace metals.  Thomas and Bendell-Young (1998) investigated the trace metal partitioned in the surficial sediment and linking these metal forms to the deposit feeder Macoma balthica at 26 locations in the intertidal region of the Fraser River estuary in B.C. Canada. Three biologically relevant sediment components, named easily reducible M n (ER-Mn), reducible Fe (RED-Fe) and organic matter, were characterized with chemical extraction procedure. The trace metal Cd, Cu, Pb, N i and Zn were separated and measured in these three fractions in the sediment. The analysis clearly indicated that metal levels in the tissue and shell of M. balthica were best related to the concentration of metal associated with the E R - M n components of the sediment.  Studies suggested that trace metals associated with the iron and manganese oxides in the sediment are most available to deposit feeding organisms (Thomas 1997). Bendell-Young and Harvey (1994) found that zinc and copper concentrations in chironomids correlated with zinc and copper concentration associated with the manganese oxide portion as modified by amounts of organic matter. Luoma et al. (1995) reported that the concentration of silver associated with the iron and manganese oxides was a reasonable good predictor of silver bioavailability to Macoma balthica.  2.3 Techniques used to determine bioavailable metal species  With the realization that bioavailable metal species depend on their chemical form, various physio-chemical separation methods and associated analytical techniques have been developed (Florence and Batley 1980, Morrison 1985, Bendell-Young et al. 1992). Numerous studies on trace metal toxicity and bioavailability have been based on these techniques (Bindra and Hall 1977, Florence and Batley 1980, Harrison et al. 1981, Morrison 1985, 1987, 1989, Bendell-Young et al. 1992, Thomas and Bendell-Young 1998, Stecko and  18  Bendell-Young 1998). These techniques are still widely used in the study of metal bioavailability.  2.3.1 Separation of metal species in dissolved phase Florence and Batley (1980) summarized commonly used separation techniques in speciation of metals in the dissolved phase, such as liquid-liquid extraction, dialysis, ultrafiltration, centrifugation, and ion-exchange. Of all these physio-chemical separation techniques, ionexchange using Chelex resin to separate bioavailable metals from other forms was recommended because of its selectivity.  Chelex chelating resin is classed as a weakly acidic cation exchange resin by virtue of its carboxylic acid groups. The resin is made by copolymers of styrene divinylbenzene. The paired iminodiacetate ions contained in the resin act as chelating groups in binding polyvalent metal ions (BIO-RAD), which gives it a higher binding strength to the metal. The mechanism of resin selectivity for trace metals is due to its structure with homogeneous pores down to 1.5-3.0 nm (Morrison 1985). The stability of the iminodiacetate groups in the resin (log k « 13 for Cu) is such that weakly complexed metals may be dissociated and preferentially associate with the Chelex-100 (Morrison 1985). When polluted water contacts the resin, the metal ion may adsorb either to the resin surface or to the resin pore surface by diffusion. The transport of metal ions into the resin pore is driven by a strong diffusion gradient and then undergoes rapid chelation. The metal in colloidal forms is prevented from adsorption by the resin because of their large molecular size. The efficient separation of metal species in seawater by Chelex resin was found with pore size structure of resin about 1.5 -3.0 nm to bind metal in the ionic form (small size) and prevent binding of colloidal metal in the large molecular size (Florence and Batley 1976).  Morrison (1985) employed Chelex -100 resin in the metal partitioning of stormwater. In his study, three dissolved metal fractions were separated, which were termed electro-chemically available metal fraction, representing the metal in ionic form; chelex removable fraction, representing ionic and weakly bound inorganic and organic low molecular weight form; and  19  colloids (strongly bound fraction). The determination of chelex removable fraction is by passing stormwater through a resin column, trace metals in the water are exchanged with the cation such as Ca  or Na on the resin and become complexed with the resin. The trace  metal in the resin can then be eluted with acidic solution. The metals in acidic elutant are referred to as the resin removable or resin extractable fraction, which implies that they are the potentially bioavailable metal species in the stormwater. By using this separation scheme, the metal in the resin removable fraction was found to represent a significant proportion of total metal for both Zn (48%) and Cd (57%) in the urban stormwater, which showed the potential toxic impact on receiving water (Morrison 1985, Morrison and Revitt 1987).  The high selectivity of Chelex resin to the freshwater trace metals, such as Cd, Cu and Zn, was described in the manufacturer's instruction manual (BIO-RAD). The recovery test of metal copper and zinc from Chelex resin in stormwater suggested that the resin can be used as an efficient tool for the determination of bioavailable metal species in the urban freshwater environment (Morrison 1985, Morrison and Revitt 1987).  2.3.2 Speciation of metal species in the particulate phase Speciation techniques account for the most sediment bioavailable trace metal studies since anthropogenic metals introduced into the aquatic environment occur either in particulate form or are rapidly sorbed to particles (Regnier and Wollast 1993). Stecko and BendellYoung (1999) pointed out that once the trace metals enter water, they are adsorbed relatively fast and contribute to the pool of metals associated with either suspended particulate matter (SPM) within the water column or are ultimately incorporated into deposited sediments (DS). The ingestion of SPM or DS by organisms as their food source will lead to the trace metal accumulation in the organism and possible transfer via the food chain to a higher trophic level. On the other hand, the geochemically unstable metal fraction in S P M or DS may release trace metals to the dissolved bioavailable forms under certain conditions (Morrison and Revitt 1987).  20  The  sequential  selective  extraction  procedure  (SEQ)  with  atomic  absorption  spectrophotometry (AAS) analysis for the geochemical speciation of metal species in the particulate phase has been developed and widely applied (Tessier et al. 1979, Harrison 1981, Bendell-Young 1992). In general, six geochemical forms have been classified by extraction with selected reagents for the metal associates, which are termed exchangeable form, carbonate form, easily reducible form (bound to Mn-oxide), reducible form (bound to Feoxide), organic form and residual form. The modifications of this analytical method have been conducted by various researchers based on their particular interest on the sediment metal geochemistry and bioavailability (Harrison et al. 1981, Morrison 1985, Bendell- Young et al. 1992, Thomas and Bendell-Young 1998, Stecko and Bendell-Young 1999). BendellYoung et al. (1992) proposed the simultaneous sediment extraction procedure (SIM). Comparing it to SEQ, they suggested that for sediment with low carbonate content, the SIM is a more efficient partitioning method because the loss of metal due to the "in between reagent rinses" in each fraction by SEQ can be overcome, and also SIM needs less time for the sample processing. However, a larger sediment sample is needed when using the SIM method.  The sediment speciation technique has been widely used to define geochemical components related to various soils, deposited sediment and suspended sediment in the water system. Harrison (1981) investigated the chemical association of lead, copper, zinc and cadmium in the street dusts and roadside soils by using sequential extraction procedure. He found that, by partitioning metals into five fractions, lead, cadmium and zinc were predominantly associated with carbonates and Fe-Mn oxides, whereas copper was largely in organic association.  Stecko and Bendell-Young (1999) studied the geochemistry of suspended particulate matter (SPM) contrasting to the deposited sediment (DS) in the Fraser River estuary, B C , Canada. A distinct difference was found in the geochemistry between these two particulate phases. Concentrations of organic matter and reducible Fe (R-Fe) in SPM were significant higher in the winter months than in the rest of the year, while the organic matter has no seasonal change in the DS. Concentrations of Cu, Zn and Pb in SPM were up to 17 times greater than  21  that in DS, with a higher proportion of these metals associated with the easily reducible component (oxides of M n and amorphous forms of Fe oxides) during winter as compared to summer months. The findings of this study suggested that the seasonal change in the partitioning of metals, in addition to greater proportion of the metal occurring in an easily reducible form in S P M relative to DS, has potentially important implications for sediment ingesting organisms capable of filter-feeding on both S P M and DS (Stecko and BendellYoung 1999).  Morrison (1987) partitioned the suspended sediment from urban stormwater into three fractions, namely exchangeable, hydrous oxide, and organic fractions, by using a selective sequential extraction technique. He suggested that the exchangeable fraction may represent the bioavailable metal source because metals in this fraction may exchange or complex with anions, so that they might be released into the soluble phase under normal pH conditions. The hydrous oxide (Mn-oxide or Fe-oxide) and organic fractions contain more strongly bound metals and are unlikely to have any immediate biological impact, but are potentially released into the water column when there is a significant drop in pH or redox potential in the water or sediment environment.  2.3.3 Dialysis with resin receiver - simulation of metal uptake by the organisms Bioassays are considered an accurate method for studying trace metal bioavailability, though they  are time  consuming, and sometimes are species-specific that lead to  the  misunderstanding of potential trace metal impacts in the water environment (Florence and Batley 1980). A dialysis with resin receiver approach has been developed to simulate the uptake of metal species by the aquatic organisms (Morrison 1987). It is generally accepted that metal species either in ionic form or bound to small inorganic or organic molecules, such as weak complexes or lipid soluble forms, with molecular weighs less than 1000, can be assimilated by a living cell. The dialysis membrane bag (MWCO (molecule weight cut off value) = 1000) can be used since it provides a fairly efficient separation between colloidal and ionic species (Morrison 1985). The Chelex resin is encapsulated in the bag to make it the  22  receiver. The consideration for the receiver structure is that the dialysis membrane has been designed to simulate the membrane of the living cell, and the Chelex resin within the bag as the constituent of the cell. When exposed in situ for a certain time, the trace metal species will penetrate the dialysis membrane and become chelated by the resin inside the bag. The concentration gradient of the membrane bag causes metal species to traverse the membrane from the polluted water since the resin chelates the metal continuously inside the bag creating a diffusion gradient. The trace metal assimilated by the receiver represent the metal species that is bioavailable in the aquatic environment, and can be quantitatively expressed as the metal uptake rate by the organisms (pg metal mm" .hr"') based on the bag area and the 2  exposure time. Morrison (1987) suggested that the dialysis with receiver resin technique should be particularly suitable for stormwater studies over a long time sampling period where intermittent discharges of metal are experienced. These may normally be missed by a grab sample during stormflow. In addition to the resin uptake of trace metals, the membrane uptake rate can also be examined and expressed in pg metal mm~ .hr . The uptake of metal 2  _1  by membrane surface may represent those metal species that are available to the surface of the organism but may not readily penetrate a membrane (Morrison 1987).  By applying dialysis with resin receiver technique in the urban stormwater, Morrison (1987) reported that an increase of bioavailable metal uptake rate was observed during experiment periods containing rainfall events, while the low and fairly constant uptake rates were found during baseflow. The sensitivity of metal uptake by the receiver with the metal level in the water was also tested in the field under stormflow conditions. The results are given in Table 2-2.  Table 2- 2 Uptake of metals by Chelex-100 in a 3500 M W C O dialysis bag (Morrison 1985) Time period  Runoff  *  volume (m )  Metal uptake rate, pg / mm . hr Zn Cd Pb  Nov.22-26, 1984  650.4  73.1  0.12  0.51  19.3  Nov.26-29, 1984  169.2  26.5  0.05  0.25  2.03  3  23  Cu  The data in Table 2-2 indicate an increase in metal uptake during the period o f higher storm runoff. It may therefore be concluded that the dialysis with resin receiver is sensitive to the increased intermittent stormwater discharge (Morrison 1985).  2.4 Trace metal studies in Brunette River watershed  The Brunette River watershed is located in the geographic center o f the Great Vancouver Regional District ( G V R D ) . It is a small tributary system to the lower Fraser River with an area o f 98 k m (Fig. 2-1). The watershed is highly urbanized, and is within the municipalities 2  of Burnaby, Vancouver and N e w Westminster. Four predominant  land uses within the  watershed include residential (42%), industrial (5.5%), commercial (15%), and open space and forested  (31%) (MacDonald et al. 1997). Highway 401 and Lougheed Highway  following the lower elevation contours around both sides o f Burnaby Lake, as well as the streets and alleys, occupy approximately 15% o f the total land within the watershed, and make up significant impervious surfaces for stormwater runoff (Hall and Anderson 1988). Two shallow lakes, namely Burnaby Lake (120 ha) and Deer Lake (80 ha) are the prominent features o f this drainage basin (Bindra and Hall 1977). The main tributary up-stream of Burnaby Lake is called Still Creek. This creek collects the major stormwater from watershed upstream and discharges it into the west-end o f the Burnaby Lake, while the Brunette River at the downstream o f the Burnaby Lake enters the Fraser River with all the flow from the watershed. The streams and lakes within the watershed have been classified as the storm sewer by the Great Vancouver Regional District ( G V R D ) .  Stream flow characteristics are highly dependent on the region's topography and climatic conditions. Within the watershed, while the steep gradients o f the land in the upper reaches of both Still Creek and the Brunette River cause high stream flow, the lower reaches o f Still Creek (from Gilmore Avenue to Burnaby Lake) experience near-quiescent conditions during dry periods leading to a variation in stream velocities ( M c C a l l u m 1995). The maximum discharges occur during the high precipitation winter months, and the minimum discharges are in the summer.  24  Figure 2- 1 Brunette River Watershed  25  Stormwater contaminants in Still Creek have been an issue since the beginning of urbanization in the region. The first zoning of industrial land in the watershed (Still Creek area) in 1947 marked the beginning of a prolonged commercial and industrial expansion (Dawson et al. 1985). The dramatic increase of the population in the Lower Mainland caused the rapid increase of automobile traffic. The associated problems of roadway congestion and air and water pollution are a major concern in the Brunette watershed as a result of large traffic volumes throughout (McCallum 1995).  There were significant changes in wildlife populations following the clearing and settlement of the region. Water pollution and habitat destruction caused the complete elimination of the salmon population in Still Creek and in Brunette River during 1960-1970's. After the efforts of Sapperton Fish and Game Club to clear river debris, eliminate point sources of contamination, and re-stock fish from hatcheries, a few salmon have returned to the Brunette River recently (Rudolph 1995).  The examination of water quality and trace metal contamination in the Brunette River watershed started in 1970's (Hall et al. 1976). Based on the contaminant transport characteristics in the urban area (see Fig. 2-2), contaminants have been examined in various components such as baseflow, stormflow, street sediment, stream sediment, lake sediment, as well as fish tissues and benthic invertebrate. The trace metals Pb, Cu, Zn, Cd, N i and Hg have been studied since they are the predominant contaminants in the urban watershed.  In order to examine water quality and the influence of non-point source pollution from urban stormwater runoff on the aquatic ecosystem, a comprehensive monitoring program was conducted by Fraser River Action Plan and Fraser Pollution Abatement Office of Environmental Canada from 1994 to 1995 (MacDonald et al. 1997). The report of this study indicated that under baseflow condition, dissolved oxygen in the lower reaches of Still Creek and in the upper reaches of Brunette River threatens the safe passage of fish during the summer. The levels of oxygen in the surface waters of lower Still Creek drop to <5 mg/L (50% saturation), while the bottom waters are often depleted. Fish kills have been observed in this reach of the river. The low dissolved oxygen (2.6 mg/L, 30% saturation) in the  26  Brunette River was also observed in summer, which was due to the highly colored water (avg. 70 mg/L) from Burnaby Lake that prevented light from penetrating into the water column and limited photosynthesis. The monitoring data showed that copper at all three stormwater sampling stations exceeded the water quality criterion o f 2 ug/L for the protection o f aquatic life.  During rainfall, the high concentrations o f trace metals were found in the stormwater. The average total zinc concentration was found to exceed the safe criterion (30 pg/L) in both Still Creek and Eagle Creek during storm runoff period. The average copper concentration in Still Creek exceeded 50 p g / L , which is far higher than the water quality criterion for copper. Even though lead additives have been removed from gasoline, the average concentration o f 10 ug/L lead in Still Creek was still detected, which exceeded the water quality criterion o f 4 pg/L for the protection o f aquatic life (Macdonald et al. 1997).  Rainfall (M  eta!  Source  A tm o s p h e r i c  I)  (M etal S o u r c e II) E x h a u s t of transportation  particles  *  Street deposit ( I m p e r v i o u s surface)  Street  O t h e r area ( G r e e n spaces)  runoff  G r o u n d w a ter  S t o r m w a t e r in stream  S tre am  S tre am  sediment  Lake  Baseflow  sediments  Figure 2- 1 Trace metal source, sink and transport in the urban area.  27  In earlier studies (Hall et al. 1976), a series of stream sediments and street surface materials were collected for one year period from July 1973 to June 1974, and 11 trace metals were analyzed. In this research, the relatively high levels of trace metals, especially copper, lead and zinc were found in the Still Creek stream sediments that were related to industrial activity and heavy traffic flow in this area. The low flow areas of both Still Creek and Brunette River were areas of trace metal sedimentation in the stream bed. The study indicated that the street surface materials appear to be one of the major sources of trace metal contamination.  Bioassay tests (Hall and Anderson 1987) using the cladoevran, Daphnia and stormwater indicated that the suspended sediments were toxic to aquatic organisms, as most contaminants in the stormwater were associated with these particles. Macdonald et al. (1997) reported that the suspended sediment at the Gilmore stream station (Still Creek) had an EC50 of 1-2%, which was characterized as a slightly toxic response. The suspended sediments sampled in runoff from a highly traveled roadway (Willingdon Ave.) were considered moderately toxic ( E C  50  0.1-0.9%).  Manganese in the watershed water system is another contaminant of concern. The replacement of M M T (Mn based fuel additive) with lead since 1970's in Canada leads to the increase of manganese in the Brunetter River watershed. Data from McCallum (1995) and Schreier (1991) have reported an increase in M n concentration in the stream sediments. A 2600% increase in the acid extractable M n between 1973 and 1993 was reported by McCallum (1995). The observations from a recent manganese study within the watershed suggested that the lightly bound M n in the acid extractable fraction may have an effect on trace metal dynamics in the system since Mn-oxides are scavengers of other trace metals (Mattu and Hall 1999).  28  2.5 Relevant environmental criteria for assessing trace metals The establishment of criteria for the assessment of metal bioavailability must consider the effect of toxic metals on the organisms, and the interaction of the metal with surrounding organic and inorganic substances (Morrison 1985). The problems encountered in the assessment of metal toxicity may be caused by (1) chemicals in the water environment affecting the trace metal toxicity, and (2) trace metals exhibiting an extremely variable toxicity between species (Borgmann 1983). One example is that the humic acid causes a significant reduction of Cu toxicity to Daphnia magna. It is generally accepted that metal toxicity decreases with increasing water hardness. The relevant water quality criteria based on the water hardness were used for the freshwater toxicity assessment and are listed in Table 2-3 (Swain 1989), Table 2-4, and Table 2-5 (Morrison 1985).  Table 2-3 provides the specific trace metal objectives for the Brunette River system, which was established by Swain (1989) and reiterated by the Fraser River Estuary Management Program (FREMP) in 1991. The values listed in the table are for protection of freshwater aquatic life and wildlife, and for the secondary contact recreation (MacDonald et al. 1997).  Table 2- 3 Water Quality Objectives in the Brunette Water System (total metal) Hardness  Cu  (mg/L as  (Ug/L)  CaCO ) 3  Zn  Pb  Cr  (lig/L)  (lig/L)  Long term  Long term  Maximum  30 day  Maximum  Maximum  Ave.  Maximum  Value  Ave.  Value  Value  30  4  18  20  2 >20  4  >30  5  The US E P A trace metal criteria under different water hardness is presented in Table 2-4. Since these criteria take into account the effect of intermittent total metal discharges, it is considered to be more realistic for the assessment of stormwater impact.  29  Table 2- 4 E P A Water Target Concentrations (total metal) (EPA 1983)  Heavy  Total  metal  hardness  Freshwater toxic criteria  Intermittent discharge criteria  (Mg/L)  (ug/L)  24 hour  Maximum  Threshold  (mg/L)  Copper  Zinc  Cadmium  Lead  Significant mortality  50  5.6  12  20  50-90  100  5.6  22  35  90-150  300  5.6  62  115  265-500  50  47  180  380  870-3200  100  47  321  680  1500-4500  300  47  800  1700  3850-11000  50  0.01  1.5  3  7-160  100  0.02  3.0  6.6  15-350  300  0.08  9.6  20  45-1070  50  0.75  74  150  350-3200  100  3.8  172  360  820-7500  300  50.0  660  1400  3100-29000  The proposed value of environmental quality objective by European Economic Council (EEC) is given in Table 2-5. Under different total water hardness, the dissolved rather than total metal concentration was set up to protect freshwater organisms against toxic trace metals. However, the intermittent metal discharge was not considered in the criteria. The relevant freshwater toxicity criteria for trace metals have been set for some particular trace metals.  30  Table 2- 5 EEC proposed E Q O * Values for Protecting Freshwater Life against Heavy Metals (Mance and O'Donnell 1983) Total Hardness  Dissolved Pb  Dissolved Zn  Dissolved Cu  (mg/L)  (Ug/L)  (Mg/L)  (Ug/L)  0-50  50  70  1  50-100  125  170  6  100-150  125  250  10  150-200  250  250  10  200-250  250  250  10  250+  250  500  28  EQO - Environmenta quality objective  The  previous studies in the Brunette River watershed demonstrated trace metal  contamination in the water system (MacDonald et al. 1997). It was found that under the same water hardness (0-100 mg/L as CaCOs), both the total and dissolved concentration of Cu and Zn exceeded the B C criteria listed in Table 2-3 by several times among street runoff and in stream. The dissolved Cu also exceeded the corresponding EEC limitation listed in Table 25. The total Cu at the two stream sites also exceeded EPA's criteria for 24-hour average for freshwater toxicity. The total Cu in the runoff even significantly exceeds the threshold of the intermittent discharge criteria.  The concentration of total metal Zn, on the other hand, exceeded E P A ' s criteria for both the 24-hour and maximum values for freshwater toxicity in street runoff. The concentration in Still Creek exceeded the 24-hour limitation for freshwater toxicity, but lower than the maximum limitation of the same criteria. Total Zn in downstream in Brunette was safely lower than all the criteria listed in Table 2-5.  31  Chapter  3  Methodology Two methods, namely metal speciation and dialysis with resin receiver, were used in this metal bioavailability study. While the metal speciation technique was used to characterize bioavailable trace metal species in the urban water; the resin receiver technique was used to simulate the uptake of these metals by the aquatic organisms. The potential toxic impacts of trace metals in the urban aquatic system were assessed with these two approaches.  Three sampling sites were selected within the watershed for this study. In order to characterize the bioavailable metal species in the Brunette River system, the sampling was conducted under baseflow, stormflow, and snowmelt conditions in the past one and half years. Each water sample was analyzed through the speciation technique with atomic absorption spectrophotometry(AAS) to identify the bioavailable species under different flow conditions. The in-situ experiments of dialysis with resin receiver were also conducted in all three sites under the same set of flow conditions.  3.1 Sampling sites  The locations of the three sampling sites are shown in Figure 3 - 1 . The site selection is based on the transport route of trace metals from urban non-point sources to the receiving water following the urban hydrologic gradient. Site #1 is a sediment trap of a road in a commercial land use area. This site was chosen to evaluate the trace metal contaminants from urban impervious surface through surface runoff. Site #2 is located in Still Creek - upstream of the lakes in the watershed. It was used to investigate the trace metal species from a highly  32  impervious sub-watershed. Located in Brunette River, Site #3 was selected to study the metal species in the downstream section o f the Burnaby Lake before entering the Fraser River.  Figure 3- 1 Locations o f the sampling sites in Brunette River watershed  Table 3- 1 Description o f sampling sites Station  Station  Number  Name  #1  Ventura  Location  Description  Ventura Drive and Lake City  •  Light industrial area.  way SE. corner o f intersection  •  Representing the street runoff.  •  Industrial area.  •  Representing the watershed  (On south slope of Burnaby Mountain)  #2  Douglas  Still Creek at Douglas A v e .  Ave.  #3  North  upstream o f the lake.  Brunette River at North R d .  Road  •  Sampled within Hume Park.  •  Representing downstream o f Burnaby Lake  33  The detailed description for each site is given in Table 3-1. A l l the water sampling and receiver experiments of this study were conducted at these three sites. Another important consideration in selecting these three sampling sites is that considerable monitoring data are already available for comparison and analysis (MacDonald et al. 1997).  3.2 Trace metal speciation  The whole process of trace metal speciation consisted of three parts: water sampling, sample speciation and sample analysis.  3.2.1 Water sampling Water was sampled at all the three sampling sites described in 3.1. Two sampling programs were conducted - grab sampling and intensive sampling. While the grab samples were collected in baseflow and stormwater conditions, the intensive water sampling was conducted over a typical storm event. The water samples were collected in a rotating sequence from site to site for one typical storm event. At each site, the water was sampled at intervals of approximately 40 min. The sampling procedure is shown in Figure 3-2.  Water sampling  Storm flow  Base flow  I  Grab sampling  r Grab sampling  Intensive sampling over one storm  Figure 3-2 Field sampling scheme  34  Both the water sampling and in situ receiver experiments were conducted from July 1998 to January 2000. They were designed to follow the precipitation patterns of the watershed, as shown in Fig. 3-3. Of the six grab-samplings, two were sampled under baseflow conditions, and four were collected in the early rainfall under stormflow conditions. Field receiver experiments were also conducted six times during the period. They were almost all started at the same time as the grab sampling, except for one in February 1999. The only intensive water sampling was carried out in November 1999, during a typical rain event of the wet season. The detailed sampling records are shown in Table 3-2.  Jun-98  Jul-98  Sep-98  Nov-98  Dec-98  — Receiver experiments;  Feb-99  ij.  Mar-99  May-99  Jul-99  Aug-99  ~ Grab water sampling;  Oct-99  Q-  Dec-99  Mar-00  intensive water sampling  Figure 3-3 Field samplings and in situ experiments  35  Jan-00  Table 3 - 2 Water sampling during study period  Sampling  Sampling  Sample  date  station  characteristics  #1  Baseflow  July 23, 1998  quantity (L)  Nov. 19,  In the trap  1.0  #2  Baseflow  In stream  4.0  #3  Baseflow  In stream  4.0  #1  Baseflow  In the trap  12.0  Trap water #2  Baseflow  In stream  12.0  #3  Baseflow  In stream  12.0  #1  Street runoff  Free falling from  10.0  road surface  Oct. 23, 1999  Sample  Trap water  A u g . 17, 1998  Sampling point  #2  Stormwater  In stream  10.0  #3  Stormwater  In stream  10.0  #1  Street runoff  Free falling from  4x4 +2x2*  1999  road surface  (intensive  #2  Stormwater  In stream  4x4 +2x2  sampling)  #3  Stormwater  In stream  4x4 +2x2  Jan. 3,  #1  Street runoff  Free falling from  3.0  road surface  2000 #2  Stormwater  In stream  3.0  #3  Stormwater  In stream  3.0  Jan. 3,  #1  Street runoff  In the trap  2.0  2000  #2  Stormwater  In stream  2.0  #3  Stormwater  In stream  2.0  *Six samplings in total. 4 L each in the first four and 2 L each in the rest two samplings.  36  3.2.2 Trace metal speciation The major trace metals studied in this project were copper and zinc, because the high concentrations o f these metals were observed within the watershed during previous studies (MacDonald et al. 1997). Manganese and iron were also analyzed to examine their effects o f regulating the more toxic trace metals such as C u and Z n in the aquatic environment.  The speciation procedure followed by this study is based on the scheme developed by Morrison (1985). Metals in the soluble phase were partitioned into resin extractable (or resin removable) fraction and residual dissolved fraction to distinguish bioavailable and nonbioavailable metal species. For particulate metals, three fractions, namely exchangeable, M n oxide bound and residual trace metals, were determined to evaluate the potentially bioavailable metals. The reason for studying the Mn-oxides bound fraction in the particulate phase was to evaluate the effect o f manganese in regulating trace metals in this urban aquatic system. Figure 3-4 shows the flow chart o f the speciation scheme. The detailed procedure for each analysis step is described below.  Water Samples (2-1 OL)  Filtration (GFC 1.2 mm)  I Dissolved Phase  Suspended Solid Phase  Exchange  Digestion  with Resin  with Concentrated HNOo  1  Extraction  Extraction  with 1M  with 0.01 M  with Concentrated  Magnesium Chloride  Hydroxylamine  H N 0 , Centrifuge  Extraction  3  Hydrochloride  Resin Removable Fraction  Dissolved Residual Fraction  Exchangeable Fraction  Mn-Oxide Binding Fraction  Particulate Residual Fraction  Figure 3-4 The speciation scheme for the water sampling  37  3.2.2.1 Separation of the water samples into dissolved and particulate phases Two to ten-liter water samples (baseflow or stormwater) were filtered through 1.2 Lim glass fiber filters ( G F C ) . The reason for using G F C filter rather than 0.47 Lim polymer filter was to provide a large amount o f filtrate for further bioavailable metal speciation in the dissolved phase.  3.2.2.2 Separation of trace metals in the dissolved phase a) Chelex Resin removable fraction (RE) 1 - 3 L filtrates o f stormwater (or 10 L o f baseflow water) were exchanged with Chelex-100 resin (6 g) in a glass column o f 25 cm in length and 1.5 cm in diameter. The flow rate was 5 -  10 m L / m i n . The resin after exchange with sample water was transferred into a 50 m L  polyethylene tube. Twenty m L o f 2 M HNO3 was added to the resin and shaken for 3 hours on a lateral action shaker (Eberbach Cop.) to elute the trace metals. The sample was centrifuged at 3000 rpm ( B E C K M A N Centrifuge J2-21M/E) for 20-30 m i n and the elutants were filtered through Whatman #4 filter paper into a 50 m L volumetric flask. The resin was extracted with 15 m L de-ionized water, shaken for half an hour and centrifuged for 20 min. The water extract was added to the acid in the volumetric flask and the sample made to volume. The Chelex resin extractable  fraction was analyzed by atomic absorption  spectrometer (SpectrAA-220FS) for Cu, Z n , Fe, and M n to give bioavailable metal forms.  b) Residual dissolved fraction Residual dissolved fraction indicates the metal species bound to the colloidal forms in the dissolved phase. It can be determined by the difference between total dissolved metal (TD) and resin removable metal (RE). Total dissolved metal was measured by the following steps. The filtrate (250-500 m L ) was digested with 5 m L concentrated H N O 3 , and evaporated to dryness on a hot plate. The residue in the digestion flask was dissolved in 10 m L o f I M HNO3, rinsed with 10 m L o f deionized water into a 25 m L volumetric flask, and then filtered  38  through a Whatman #4 filter paper. This sample was analyzed by A A S for C u , Z n , M n and Fe.  3.2.2.3 Speciation of suspended sediment  a) Exchangeable fraction Trace metals in the exchangeable fraction were analyzed with the following procedure. First the filter containing the particulate residue was cut into small pieces and then placed into a polyethylene tube. Twenty m L o f I M M g C b at pH=7 was added to the tube, the tube was then shaken at room temperature for 1 hour. The sample was then centrifuged at 10000 rpm for 30 min, and the extract was filtered through Whatman #4 filter paper into a 50 m L volumetric flask. The filter in the polyethylene tube was extracted with 20 m L o f deionized water on the shaker for another 20 min and added into the volumetric flask. This extract was finally analyzed by A A S for C u , Z n , M n and Fe. The results represent the exchangeable fraction. The remaining filter residue was then used in the following analysis.  b) Hydrous manganese oxide fraction The remaining filter in the polyethylene tube was used for further extraction. It was extracted with 20 m L o f 0.01 M acidic hydroxylamine hydrochloride, N H 2 O H . H C I (dissolved in 0.01M  H N O 3 , pH=2) by shaking at room temperature for 0.5 hour. The sample was  centrifuged at 10000 rpm for 30 min, and the supernatant was transferred to a 50 m L volumetric flask through Whatman #4 filter paper. The filter in the polyethylene tube was extracted with 20 m l o f deionized water on the shaker for another 20 m i n and added into the volumetric flask. The sample was analyzed by A A S for C u , Z n , M n and Fe. The remaining filter residue is used in the following procedure.  c) Particulate residual fraction The remaining filter was transferred to a porcelain crucible and digested with a 20 m L mixture o f concentrated HNO3 and HC1 (4:1) on a hot plate till the solution evaporated. The  39  residue was dissolved in 15-20 m L o f 0.05 M H N O 3 , and diluted to 25 m L . The sample was analyzed by A A S as the particulate residual fraction for the metals C u , Z n , Fe and M n .  3.2.2.4 Total metal concentration The total metal concentration ™ was determined in the water samples, and T M concentration is used to calculate the total particulate metal concentration (TP) in stormwater expressed as TP=TM-TD.  A non-fdtered water sample (250 m L ) in an Erlenmeyer flask was digested  with 5 m l concentrated HNO3 and evaporated to dryness on a hot plate. The flask was rinsed with 10 m L 1 M HNO3 and 10 m L o f deionized water and transferred to a 25 m L volumetric flask through a paper filter (Whatman #4, washed by de-ionized water). C u , Z n , Fe, and M n were analyzed with A A S .  3.2.3 Other water quality data 3.2.3.1 Suspended sediment (SS) The suspended sediment was analyzed by following a standard analysis procedure ( A P H A 1981). The samples were filtered through G F / F glass fiber filter (0.7 pm). Total suspended sediments were measured for stormwater over a rain event..  3.2.3.2 pH The p H o f the water samples o f the intensive sampling was measured in the laboratory with p H meter ( O R I O N 420A), which was standardized with p H 4 and p H 7 buffers.  3.2.3.3 Precipitation data and flow data Precipitation data and flow data from January 1998 to January 2000 were obtained from G V R D . The flow data o f two stations, Station SC4 in Still Creek and Station B T 5 in Brunette River, were collected. While Station S C 4 was at exactly the same location as sampling site #2, Station B T 5 was just upstream o f sampling site #3. They were then used for the  40  stormflow and baseflow analysis. The rain gauge number BU7 was located in the center of the watershed, and provided the precipitation data.  3.3 Dialysis with resin receiver at field sites The dialysis with resin receiver technique was used to measure the uptake rate of the trace metals copper, zinc, manganese, and iron in the urban water system. This technique is based on the fact that the chelating resins are able to reasonably simulate the cell uptake. It was conducted in two steps: the in situ receiver experiment and the laboratory analysis. The insitu experiment was to expose the receiver to the water for a period of time, to simulate the metal uptake by aquatic organisms.  3.3.1 Field receiver experiment procedure The in situ receiver experiments were conducted in the same three sites used for the water sampling. A total of six in situ receiver experiments were conducted under baseflow, stormflow and snowmelt conditions. The cellulose dialysis tubing bags within which one gram of Chelex resin (Chelex -100 in Na form) was encapsulated were emerged in the sediment trap and streams. After the exposure of a period of time (one storm, one week and two weeks), the receivers were returned to the laboratory for analysis. The description of all the in-situ receiver experiment is given in Table 3-3.  3.3.2 Receiver materials The resin receiver was made from a Spectra/Pro wet dialysis membrane tubing, which has a diameter of 24 mm and the molecular weight cut-off value of 1000. The membrane tubing was treated with 0.01 E D T A overnight and washed with deionized water before use. The suitable tubing length of 60-80 mm (Morrison 1985) was accurately measured. The Chelex100 chelating resin in Na form (1.0 g) was slurried into the bag with deionized water (approximately 20 ml). The tubing bag was tightly sealed at both ends and suspended within  41  a plastic tube. The plastic tubes were tied to concrete blocks and placed in the stream sediment trap.  Table 3-3 Field receiver experiments during the study period Experiment  Sampling  Flow  Experiment  Exposure  No. of  date  station  characteristics  place  time  dialysis  (days)  receivers  July 23 - 30,  #1  Baseflow  In the trap  7  4  1998  #2  Baseflow  In stream  7  4  #3  Baseflow  In stream  7  4  #1  Baseflow  In the trap  15  4  Sept. 1,  #2  Baseflow  In stream  15  4  1998  #3  Baseflow  In stream  15  4  Feb. 1 1 - 1 8  #1  Street runoff  In the trap  7  4  1999  #2  Stormwater  In stream  7  8  #3  Stormwater  In stream  7  6  #1  Street runoff  In the trap  7  6  #2  Stormwater  In stream  7  0  1999  #3  Stormwater  In stream  7  6  N o v . 19, 1999  #1  Street runoff  In the trap  7  4  (intensive  #2  Stormwater  In stream  7  4  sampling  #3  Stormwater  In stream  7  4  Nov. 1 9 - 2 5 ,  #1  Street runoff  In the trap  7  4  1999  #2  Stormwater  In stream  7  4  #3  Stormwater  In stream  7  4  Jan. 3 - 1 0 ,  #1  Street runoff  In the trap  7  4  2000  #2  Stormwater  In stream  7  4  #3  Stormwater  In stream  7  4  A u g . 17 -  Oct. 27 -  Nov. 3  42  3.3.3 Laboratory analysis The trace metals were analyzed for both the receiver resin and membrane tubing. The sample preparations were conducted in a Soil Science laboratory and analyzed in a Environmental Engineering laboratory by atomic absorption spectrophotometry ( A A S ) . The procedure of laboratory analysis is described below.  3.3.3.1 Analysis of resin The resin was carefully released from the tubing bag into a beaker, and the bag was washed with 10 m L deionized water to transfer all residue resin into the beaker. The resin was transferred with water into a 25 m L plastic centrifuge tube, and the supernatant was removed with pipettes. Ten m l o f I M HNO3 was added into the tube and the metals was eluted from resin by shaking 3- 4 hours. The sample was centrifuged at 3000 rpm for 20-30 m i n and the elutants were filtered through Whatman #4 filter paper into a 25 m L volumetric flask. Ten m L o f deionized water was added to the tube, and shaken for another half an hour. The elutant was centrifuged and filtered into the acid extract. The sample was diluted to 25 m L , and analyzed by A A S for C u , Z n , M n , and Fe.  3.3.3.2 Analysis of trace metal in membrane After the resin was released from the membrane bag, the bag was put into a 25 m l plastic centrifuge tube. The membrane bag was extracted with 10 m L o f 2 M H N O 3 by shaking for 34 hours. The elutant was filtered with the Whatman #4 filter paper into a 25 m L volumetric flask. Five m L o f deionized water was added into the tube to rinse the bag by shaking for another half an hour. The elutants was filtered into the volumetric flask. The filtrate was diluted to 25 m L , and the sample was then analyzed by A A S for C u , Z n , M n , and Fe.  3.3.3.3 Exchange capacity and interference test for Chelex resin The resin exchange capacity was tested to ensure enough Chelex resin was available in the dialysis bag for the targeted trace metal complexation. W i t h the consideration that the  43  common divalent metals such as calcium and magnesium are dominant elements in the urban water, the interference o f these metal species were tested in the laboratory to ensure that these metals had little interference on the trace metal copper and zinc complexation with Chelex resin. The results o f these two experiments are given in the Appendix E .  3.4 Q A / Q C and statistics  A l l glassware was acid-washed before use. The blanks used in the speciation and elution processes went through the complete analytical procedure. Reproducibility o f the analysis was measured on duplicate samples for the trace metals in the water sample. The receiver membrane was treated with 0.02M E D T A and washed with deionized water to eliminate metal contaminants adsorbed in the membrane surface before use. The replicates o f receiver were used in each in situ experiment and the analysis.  The data obtained through the trace metal speciation o f the water samples and the in situ receiver experiments, along with the precipitation and flow data, were analyzed statistically to obtain the following information on bioavailable trace metals in the watershed: 1. Correlation between metals M n and Fe with trace metals C u and Z n . It is to see how M n and Fe regulate the toxicity o f C u and Z n . 2. Correlation between trace metals with suspended sediment transport over surface runoff. 3. Correlation o f the two bioavailable trace metal study approaches, i.e. the relationship between the bioavailable trace metal speciation and the trace metal uptake.  44  Chapter  4  Results and D i s c u s s i o n  The results obtained from the two technical approaches - namely metal speciation and dialysis with resin receiver - are presented in this chapter. The speciation o f trace metal in the water samples provides the information on the concentration o f bioavailable trace metal species, in both dissolved form and the particulate form associated with suspended sediment. These forms were studied under baseflow, stormwater, and snowmelt conditions. The uptake rate of bioavailable metals measured in the field experiments using dialysis with resin receiver helps to evaluate the potential accumulation o f trace metal species by the aquatic organisms throughout the drainage system o f this highly urbanized watershed.  The behavior o f trace metals and their potential bioavailability in the watershed are discussed in this chapter. This discussion includes their spatial distribution and seasonal variation, and their potential toxicity either through immediate cell uptake (dissolved metal) or through particulate metal ingestion. The focus o f the discussion is on: 1) the bioavailable forms o f trace metal in relation to their total concentration; 2) the discharge pattern o f the trace metals through storm runoff; and 3) the potential toxic effects o f the trace metals on the receiving water organisms. The possible control factors in regulating the bioavailable metal species and the uptake o f these species by the receiver resin have also been analyzed. The correlation between dissolved bioavailable metal species and their uptake by receiver resin are also discussed.  45  Table 4-1 summarized the details o f climatic and hydrological conditions during the water sampling and resin receiver experiments. The baseflow sampling were carried out in two ocassions under non rainfall period. There was a short rainfall during the August field experiment. Stormwater sampling was try to catch each storm at the starting o f the rainfall, but due to the transportation delay, the sampling can only be conducted in the early storm. A l l storms were in the medium to large, except the storm in Feberary, 1999, which is relatively light comparing to the others.  Table 4- 1 Climatic and hydrological conditions during the sampling and receiver experiment (Need to put some data—check for data) Sampling & R e s i n experime-ntal time (7 days exposure)  Dry period before rain (day)  Total rainfall over exp. period (mm)  Total flow/ exp. period @ Site #2 (m )  Total flow/ exp. period @ Site #3 (m )  7 30  0 16.8  102826 379296  1091059 2232576  0 12 5 0  56.2 85.2 90.4 83.2  1731024 2086128 2181600 2373840  3462048 4172256 4363200 4747680  3  3  Baseflow July, 98 A u g . , 98  Stormflow Feb., 99 Oct., 99 Nov., 99 Jan., 00  4.1 Bioavailable metal study by trace metal partitioning  4.1.1 Total water fractionation The series physical-chemical partitioning procedures were employed for all the water samples collected during baseflow and stormflow, which included grab water samples and the intensive stormwater samples. The results o f the street runoff (Site #1) demonstrate the bioavailable trace metals discharged directly from the urban impervious surface; while the results obtained at Site #2 and Site #3 indicate the spatial variation o f the bioavailable trace metals when they are transported  through Brunette River water  46  system. The metal  partitioning o f the intensive stormwater samples provides insight into the variation of bioavailable trace metal species in both surface runoff and in stream water over a rain event.  4.1.1.1 Grab water samples Grab samples were collected from stormwater and baseflow. The total ( T M ) , total dissolved (TD), and resin extractable (RE) fractions for the trace metals copper, zinc, manganese and iron were measured in the water samples. The R E fraction measured the dissolved phase o f each metal and was regarded as the most possible bioavailable trace-metal species that can be directly taken up through the cell o f the organisms. The suspended sediments collected in each grab samples were analyzed with the speciation technique to understand the different geochemical binding forms o f the particulate metals. The complete data set o f these metalpartitioning results is given in Appendix A .  a. Metal partitioning in stormwater Stormwater sampling was conducted four times through the whole study period (Tab. 4-1). The detailed sampling procedure was described in chapter 3 (Table 3-2). The partitioning o f the four trace metals - C u , Z n , M n , and Fe - in stormwater was conducted on the grab samples collected on Jan. 3 and Jan. 10, 2000, respectively. The grab samples collected in October 1999 were used for the preliminary investigation o f the study methods, therefore their results are not included in this chapter. The results o f the intensive stormwater sampling program (November 1999), on the other hand, w i l l be presented separately in section 4.1.1.2. The results o f the mean concentrations o f C u , Z n , M n , and Fe in their T M , T D and R E forms are presented from F i g . 4-1 to Fig. 4-4, with their 90% confident interval (n - 4).  Copper A s shown in Figure 4-1, it can be seen a significantly higher total copper concentration ( T M Cu) was observed at Site #1 than at the other two sampling sites. O f the two stream sites, a relatively higher level o f total copper was found in Still Creek, upstream o f Burnaby Lake (Site # 2), than that observed at the down stream site in the Brunette River (Site #3).  47  Site 1  Site 2  Site 3  Figure 4- 1 Partitioning o f C u in stormwater sampled at three sites There was no obvious difference o f dissolved copper (TD-Cu) among the three sites. A quite similar mean T D - C u concentration (24.5 ug/L) was detected at Site #1 and Site #2, but a slightly lower level o f 17.4 p g / L was found at Site #3. The resin removable copper (RE-Cu) fraction, on the other hand, was at a very low level (below 10 pg/1) at all three sites.  Zinc A s shown in Figure 4-2, Z n behaved similarly to C u . H i g h T M concentration o f zinc was found in the street runoff (Site #1) with the mean concentration o f 369 p g / L . In the stream, the total zinc ( T M - Z n ) concentration was measured in the range o f 80-188 p g / L in Still Creek (Site #2), and decreased approximately 40% to 37-59 p g / L downstream in Brunette River (Site #3). It can be seen in Fig. 4-2 that the total dissolved zinc (TD-Zn) was highly corelated to T M - Z n at all three sampling sites, which accounted for approximately 50-70% o f the T M - Z n concentration. Zinc in the resin removable fraction (RE-Zn) was also closely corelated to the dissolved zinc, accounting for about 60-70% o f total dissolved zinc (TD-Zn) at all three sites. The higher R E - Z n fraction in dissolved phase indicates that a large amount o f zinc exists as ionic and weakly complexed forms in the stormwater.  48  ffiBTM  90%  D)  ^ c  i n  c 0)  o o  D T D  ED R E  c o n f i d e n c e i n t e r v a l (n = 4)  o o  Site 1  Site 2  Site  3  Figure 4- 2 Partitioning o f Z n in stomwater sampled at three sites  Manganese The partitioning o f manganese among T M , T D and R E fractions in the stormwater, shown in Fig. 4-3, demonstrates that the total manganese ( T M - M n ) concentration was quite high in the street runoff (Site #1), with the average value o f 214.3 ug/L. A t the stream sites, the mean T M - M n concentration was 82.4 pg/L i n the upstream (Site #2), and 61.8 p.g/L i n the downstream location (Site 3). The total dissolved manganese ( T D - M n ) i n the street runoff (Site #1) was not significantly higher than that at the stream sites, which shows a similar pattern to that o f copper. The resin removable manganese ( R E - M n ) fraction also accounted for a large proportion o f the total dissolved manganese, which was found between 65-80% of the T D - M n at all three sites.  49  EfflTM  c o B  '.5  k_  c u o C  0  o  9 0 %  o m  D T D  c o n f i d e n c e  LURE  i n t e r v a l  (n =  4)  T—  o o  ^  Site 1  Site 2  Site 3  Figure 4- 3 Partitioning o f M n in stormwater sampled at three sites  Iron The partitioning o f metal iron between T M , T D and R E fractions is shown in F i g . 4-4. It clearly indicates that iron exists mainly in the particulate forms in the stormwater throughout the watershed. Although the mean T M - F e concentration in street runoff (Site #1) was as high as 11.0 mg/L, only 1.1 m g / L o f dissolved iron was detected. It was ten times less than the total concentration (TM-Fe) at the same site. T M - F e concentrations at the two stream sites were significantly lower than that from the street samples. The mean concentration was measured 1.1 m g / L in Still Creek and 0.9 m g / L in Brunette River, respectively. The corresponding T D Fe concentration was also very low in the stream, less than 0.3 m g / L at the two stream sites was observed. The extremely low concentrations o f R E - F e fraction were found at all the three sites, with the highest concentration o f 0.5 mg/L at Site #1, and less than 0.02 m g / L at the two stream sites.  50  ITM  •  TD  O R E  90% c o n f i d e n c e i n t e r v a l (n = 4)  Site 1  Site 3  Site 2  Figure 4- 4 Partitioning o f Fe in stormwater sampled at three sites  b. Resin extractable metals in baseflow The grab sampling was conducted only once under baseflow condition at three sites in August 1998. Only the metals in bioavailable fraction - R E form were measured. The total and total dissolved metals were not analyzed because o f their very low concentrations in the baseflow. The R E fractions o f C u , Z n , M n , and Fe in the baseflow samples are listed in Table 4-1.  Table 4- 2 Resin exchangeable metal (RE) (ug/L) in baseflow  Cu  Zn  Mn  Fe  (Mg/L)  (Mg/L)  (Mg/L)  Site #1  3.9  44  50  3.77  Site #2  0.8  79  53  0.29  Site #3  0.2  16  10  0.28  51  A similar distribution o f the R E fraction o f C u and Fe was found at all three sites, where the concentrations were very similar at each site. Although the R E concentrations o f these two metals were at very low levels at all three sites, considerably higher concentrations o f C u and Fe were found in the sediment trap (Site #1) than at the stream sites (#2 and #3).  Regarding to the R E fractions o f Z n and M n , R E fraction o f Z n was found higher in streams, but lower in street runoff (site #1). However, the maximum concentrations o f these two metals were not found at the street site, but in Still Creek. M u c h lower R E levels o f these two metals were found in the Brunette River - downstream o f the lake.  From the above results it is obvious that while the total concentration o f Fe is always much higher than that o f the other three metals in the water system, its R E fraction is at a very low level in both stormwater and baseflow samples.  4.1.1.2 Intensive stormwater samples The partitioning o f trace metals in stormwater samples was carried out over a storm event. It covered each o f the four metals at each sampling site. The results o f this analysis demonstrate the variations o f the trace metal components over a rain event. This indicates the transport characteristics o f trace metal contaminants through urban waterways, as well as the distribution o f bioavailable trace metal species over a storm period. The data on the partitioning o f C u , Z n , M n and Fe, the suspended sediment, and p H o f the target storm event are given in Appendix C .  Copper The partitioning o f copper in samples at the three sites is shown in F i g . 4-5. The corresponding rainfall intensity (l6o) is provided with the street runoff at Site #1 in F i g . 45(a). For the two stream sites, the corresponding flow rates over the same period in Still Creek and Brunette River are also included in Figs. 4-5(b) and 4-5(c).  52  Figure 4-5(a) demonstrates that the T M - C u increased from 33 u g / L in the early stage o f the storm to 113 ug/L in the later period o f the rain event in the street runoff. T D - C u concentration, on the contrary, decreased from 20 p g / L to 5 u g / L during the same period. The R E - C u concentration o f 9 ug/L was only observed in the early storm, and decreased to the undetectable level later. This clearly indicates that the dissolved copper was washed out early in the rainfall event. The increase of T M - C u with the storm development implies the increase o f particulate copper (expressed as T P = T M - T D ) discharge with the progress o f the rain event. This might be due to the discharge characteristics o f metals associated with suspended sediment (SS) from the impervious surfaces (Solo and Perkins 1997, Jarvis 1992, Yuan et al. 1999), and w i l l be presented section 4.3.  Totally different C u partitioning patterns were observed at the two stream sites (Site #2 and Site #3). A t Site #2 (Fig.4-9(b)), T M - C u , T D - C u and R E - C u were all at low levels o f 10-17 ug/L through the whole rain event, except one sample at 11:30. The reason o f the big change at this time is unknown. Even lower copper levels were measured at Site #3 (Fig.4-9(C)). The concentrations o f T M - C u and T D - C u were in the range o f 3-10 ug/L, and R E - C u only ranged from 0.6-3.0 p g / L in this site. The low C u in the down stream site was expected to be due to the sedimentation o f C u in Burnaby Lake. There were no obvious relationships between the copper concentrations and its partitioning and the flow rates o f the two stream sites.  Zinc Zinc partitioning over one storm at the three sampling sites is shown in Figure 4-6. It shows very similar pattern to that o f the copper at street runoff (Fig. 4-6(a)). A t this site, T M - Z n concentration increased from 151 ug/L to 613 ug/L during the rain event. T D - Z n , on the other hand, kept relatively constant concentration through the sampling period (78-135 ug/L). R E - Z n was closely related to the T D - Z n and showed a similar trend to that o f T D - Z n over the whole storm event. The early wash-off o f the dissolved phase was not so obvious for zinc. The dissolved phase accounted for a large part o f the zinc in the street runoff during most o f the rainfall event.  53  C u - S ite  #1  3  Time  (a) Site #1.  Time  7 5  (b) Site #2. C u  tP  <?  - Site  #3  .t£  >>  5  /S>  .<?  .<£  &  Time  (c) Site #3. Figure 4- 5 Copper partitioning over one storm event (Nov. 19, 1999)  54  (a) Site #1.  &  .(£>&  >•  ft> K V& .&  KV  C>-  Time  (b) Site #2.  -i  o  ro s  .«$>  »•  J> .?>•  .•$>  .?>•  #  ^-  <$> K  v  <S>  ^-  #  ^^Time  jy «5>  _.<$»  <v-  J&  <u-  _.•$> <v-  ..^ <s-  „.#  ^f>  (c) Site #3 Figure 4- 6 Zinc partitioning over one storm event (Nov. 19, 1999)  55  A t the stream sites, unlike copper, the increasing T M - Z n was observed with the progress o f the storm, which increased from 66 ug/L to 126 p g / L at Still Creek (Fig. 4-6(b)) and from 34 pg/L to 65 p g / L in the Brunette River (Fig. 4-6(c)). The first flush effect for T D - Z i n c seemed also to occur at Site #2, where T D - Z n decreased to its minimum at the end o f the rain event even though the flow was still increasing. A t site #3, lower T D - Z n and R E - Z n were observed early in the rain period. Later increase o f T D - Z n and R E - Z n might be due to the time-delay of the flow at downstream sites. H i g h ratio o f R E - Z n over T D - Z n was also observed at all three sites for most o f the samples. This observation agrees with the grab sample results, and demonstrates that the dissolved zinc in the stormwater is mostly in the bioavailable form (42 - 94 %).  Manganese The total manganese ( T M - M n ) concentration in street runoff (shown in Fig. 4-7(a)) increased from 62 ug/L to 345 ug/L during the rain event. Similar to the grab samples, a large portion (>70%) o f manganese in street runoff was found to be associated with the particulate form. The maximum T M - M n concentration shown at the end o f the rain might be also due to the discharge characteristics o f metals associated with SS. The dissolved and resin removable manganese fractions over the sampling period remained at low and consistent levels (14-25 ug/L for T D - M n and 9-25 p g / L for R E - M n ) . The total manganese ( T M ) concentration was found to be higher at Site #2 in Still Creek (Figure 4-7(b)), with a range o f 116-155 p g / L throughout the sampling period. A high ratio o f T D - M n was observed in most o f the samples at Site #2, which was similar to that found in the grab samples. It might be concluded that a large proportion o f manganese was in the soluble form in the upper region o f the watershed. A t Site #3 (Figure 4-7(c)) in the Brunette River, T M - M n concentration was between 57-98 ug/L. The low T D - M n and associated R E - M n at Site #3 indicate that manganese exists largely in the particulate form in the downstream region o f the Brunette River system. This is supported by the observation o f Mattu and Hall in this watershed (1999).  56  T i m e  (b) Site #2. M  n  -  Site  #3  (c) Site #3. Figure 4- 7 Manganese partitioning over one storm event (Nov. 19, 1999)  57  Iron A clear partitioning pattern was observed at all three sampling sites (Figure 4-8) during the storm event - most o f the iron was in the particulate phase. The highest T M - F e concentration was found in the street runoff (Site #1), which increased from 13 m g / L i n the early stage to 70 mg/L at the end o f the rain event. T M - F e in Still Creek (Site #2) also showed the increasing trend, starting from 7 mg/L to the end o f 10 mg/L. L o w T M - F e concentration was found in Brunette River (Site #3), with the concentration o f 0.7-1.0 m g / L . The possible reason for this big concentration difference at the two stream sites is the regulation o f the Burnaby Lake. Lower T D - F e was observed at all three sites. The maximum T D - F e was found at Site #2 with the range o f 0.4-0.9 mg/L, which was less than ten times that o f T M - F e at the same site. Very low T D - F e level o f 0.2-0.5 mg/L was at Site #1 and 0.3-0.4 mg/L at Site #3. R E - F e levels were also very low at all three sites, with the maximum value o f 3-25 ug/L.  58  (a) Site #1.  (b) Site #2. F e  - S ite  #3  rm,  I T M ] R  E  T  D  Flow  rate  (c) Site #3. Figure 4- 8 Iron partitioning over one storm event (Nov. 19, 1999)  59  4.1.1.3 Discussion on the dissolved metal components The toxicity o f trace metal pollutants on aquatic organisms was found to be directly related to the dissolved trace metal species in the water system. Metal in the ionic form or weakly complexed forms with small particle size (RE-fraction) was recognized as the most direct contributor o f this toxic effects (Macdonald et. al. 1997, Morrison 1985, Mance and O'Donnell 1983). In the following discussion, we w i l l look at how much the dissolved phase accounts for in the total metals ( T D / T M ) , and then the percentage o f dissolved bioavailable metals in the total dissolved phase ( R E / T D ) . Lastly, the fraction o f dissolved bioavailable metals in the total trace metals can be summarized ( R E / T M ) .  a.  TD/TM  The T D / T M ratio o f four metals, C u , Z n , M n and Fe, at three sites along the urban hydrological gradient in the Brunette watershed is presented in F i g . 4-9. The results o f this study (designated as 99') are the average o f three grab stormwater samples collected in N o v . 1999, Jan. 3 and Jan. 10, 2000. The previous monitoring results (Macdonald et al. 1997) at these three sites are also plotted for comparison, which are the average o f 18 storms for street runoff at Site #1 and the average o f 12 storms in Still Creek (Site #2), respectively. N o data for the Brunette River (Site #3) and no data about Fe were collected in the 1997 monitoring program.  C u - T D / T P  EDTD  Sampling  (a)  60  s ite  H T P  Zn-TD/TP 1 00%  8 0%  6 0%  40% 2 0%  0 % S Its  #2  S a m p l i n g  s ite  (b) EE T D  M n -T D /T P  IT P  100 % 80 % 60 % 40 % 20% 0% S Ite  #2  S a m p l i n g  s ite  (c) F e - T D IT  P  GET D  8 0 %  40 % 2 0 %  S a m p l i n g  site  (d) Figure 4- 9 Ratio of T D / T M for C u , Z n , M n and Fe  61  SIT P  TD/TM ratio in the street surface runoff Figure 4-9(a) indicates that the dissolved copper (TD-Cu) in the street runoff (Site #1) accounts for approximately 37% o f the total C u concentration in both studies. It means that more than 60% o f copper was in the particulate form when it was washed off from urban impervious surfaces. It has been reported that the high copper concentration in street runoff was due to the high traffic densities - as the results o f wear o f break linings ( M c C a l l u m 1995). However, the potential sources o f soluble copper may be also from house and garden pesticides containing copper sulfate and copper chromate, and from household plumbing that is commonly solubilised by the relatively corrosive water supply. The metals from all these sources could be deposited on the lawn and roadway, and then washed off during a storm (Macdonald etal. 1997).  M u c h higher ratio o f dissolved Z n was found in this study (approximately 60%) than in the 1997 study (about 40%) in street runoff (Site #1) (Fig.4-9(b)). The high soluble zinc in urban runoff might be contributed by the specific uses of Z n including corrosion o f galvanized parts, zinc oxides in paints, zinc chloride in antifreeze as a rust inhibitor, tire wear, and even lubricating oils containing zinc dialkyl (Atlins and Hawley 1978).  Relatively low ratio o f dissolved M n (approximately 30%) was found in street runoff samples in both studies (Fig. 4-9(c)). However, as the concentration o f total manganese discharged from surface runoff was very high (as high as 241 pg/L), the dissolved manganese component was also at a relatively high concentration. Macdonald et al (1997) reported that, in addition to natural manganese sources such as erosion o f mineral soils in the watershed, the major suspected anthropogenic source was the gasoline additive M M T (methylcyclopentadienyl  manganese  tricabonyl)  used  in  Canada  since  1970s.  The  combustion products o f M M T , thought to be Mn-phosphates, Mn-oxides and Mn-sulfates (Ethyl 1997), may be the major sources o f dissolved M n in stormwater (Mattu and H a l l 1999).  62  Figure 4-9(d) shows that dissolved Fe observed in street runoff had the lowest ratio o f 7%, which implied that iron washed out from urban impervious area was mostly bound to the suspended sediment. Shuman (1977) noted that hydrous oxides o f Fe are found to some extent in all soils, and this hydrous ferric hydroxide is strongly attached to clay minerals as part o f the particulate phase.  TD/TM ratio in streams  A t stream sites, a large portion o f C u was found in the dissolved forms i n the early stage o f the storm. Over 80% o f dissolved C u was found in Still Creek (Site #2) and over 90% was found in Brunette river (Site #3), respectively. A relatively high ratio o f dissolved Z n was also observed in the stream sites. The dissolved Z n accounted for approximately 70% at Still Creek and nearly 92% at Brunette River, respectively. The total dissolved ratios o f these two metals agreed well in the two studies, and both had larger or narrower ratios o f the dissolved fraction in downstream than in upstream portion o f the waterway.  The T D / T M ratio o f M n i n streams showed an opposite pattern to those o f copper and zinc, where the higher percentage o f dissolved M n was observed in the upstream portion. About 82%) o f dissolved M n was measured in Still Creek (Site #2), and around 54% was measured in the Brunette River (Site #3). A similar pattern was also reported by Mattu and H a l l (1999) in early storm samples, in which they measured T D / T M ratio o f 90% in the upstream and 2040%o in the downstream sections. The T D / T M ratio o f iron maintained a low level in both upstream and downstream - less than 15%. The low percentage o f dissolved iron implied that iron was mainly associated with the particulate phase in streams.  The T D / T M ratio is regulated by many factors such as the type o f metal compounds accumulated in the urban surface, the aquatic physical-chemical environment, and the urban hydrological flow regime. It was reported that the high ratio o f dissolved metals in the upstream can be either due to the direct response o f first flush effect between the street runoff and the connected receiving water (Jarvis 1992, Solo and Perkins 1997), or due to the flow regime i n the watershed. The flow rates at two stream sites and the corresponding rainfall intensity recorded over a rainfall event were presented in Figure 4-32. It can be seen that the  63  flow rate in Still Creek (Site #2) is relatively low. During rainfall, the particulate metals transported through this section may deposit into the streambed (Hall et. al. 1976), which leads to the high ratio o f the metal in the dissolved phase. However, the flow rate at Site #3, a downstream site in the Brunette River, was relatively higher. The high proportion of dissolved metals at this site, on the other hand, is attributed to the sedimentation function of the Burnaby Lake located upstream. The possible mechanism o f sedimentation is that when the metals are washed into the Burnaby Lake, large amount o f particulate associated metals from upper stream (Still Creek) was deposited in the lake due to the slower flow in the lake area. This sedimentation process leads to the decrease o f both particulate and total metal concentrations downstream. A s a result, a higher ratio o f dissolved metal in the down stream is observed.  The higher T D / T M ratio o f manganese upstream can be explained as follows: The high soluble manganese found in the Still Creek was related to the M M T combustion products directly flushed from street runoff. This dissolved M n is apparently oxidized as it passes through Burnaby Lake, resulting in the lower soluble manganese downstream o f the lake Brunette River (Mattu and H a l l 1999).  L o w ratio o f soluble Fe in streams was mostly due to the fact that Fe that originated from the urban storm runoff was largely in the particulate associated form (clay mineral associated Fehydrous oxide or Fe phosphate). Another possible explanation is that when the ionic Fe enters the receiving water, the oxidation process o f F e  2 +  to F e  3 +  takes place in a few minutes  under freshwater p H conditions, and then the great affinity between hydroxide ions and F e  3 +  quickly forms the insoluble ferric hydroxide (Stumm and Morgan 1970). Aquatic p H seems to be the most important influencing factor in the precipitation (oxidation) o f iron and manganese. Stumm and Lee (1960) reported that iron is readily oxidized by dissolved oxygen to the ferric form in the alkaline-neutral to slightly acid p H range; while manganese requires a much higher p H for equivalent rates o f oxidation.  b. RE/TD  64  The resin extractable fraction o f dissolved trace metals has been found to be more toxic as it can be assimilated directly by the aquatic organisms, which is also regarded as a bioavailable fraction o f the dissolved trace metals (Florence and Batley 1980, Morrison 1985 and 1989). Table 4-3 lists the percentage o f resin extractable fraction over the total dissolved copper, zinc, manganese and iron.  Table 4- 3 R E / T D for C u , Z n , M n and Fe (Averaged over three stormwater samples) Metal Site #1 Site #2 Site #3 (%)  RE/TD  Cu  31  13  7  Zn  75  60  56  Mn  70  73  Fe  44  1  .  75 5  Although the dissolved C u accounted for quite a large proportion o f the total metal C u in urban receiving water, the resin removable fraction o f C u , on the contrary, was only a smaller percentage. It was reported that C u in the dissolved phase was largely in the inorganic or organic colloidal forms, such as C u - F e ( O H ) - h u m i c acid, which were regarded as non3  bioavailable forms due to their larger molecular size (Florence and Batley 1980, Morrison 1989). A s suggested in the literature, C u is more likely associated with organic matter in the aquatic system (Allard et al. 1987). In the Brunette watershed, high concentration o f organic matter and the initial high color (1500-2000 mg Pt/L) in the storm water indicate that the runoff could contain high concentration o f humic substances and/or lignins and tannins - the main characteristics o f the leaf degradation products from the watershed area (Macdonald et al. 1997).  The resin removable fraction o f metal Z n (RE-Zn), on the other hand, accounted for a larger proportion o f the dissolved phase in stormwater. The ratio o f R E / T D was around 60% at the two stream sites and over 70% in the street runoff. H i g h portion o f R E - Z n was also observed in the urban stormwater by Morrison (1985). He suggested that these soluble forms were free and weakly complexed Z n coming from the soluble chloride complexes o f Z n in the urban contributing area.  65  Manganese also showed a high percentage o f bioavailable fraction in dissolved form at all the sites, where over 70% o f the dissolved manganese was potentially bioavailable - leaving only 20 - 30% in the colloidal fraction. H i g h ratio o f manganese in resin removable form indicates the divalent state o f M n in the water system, which is extremely soluble and was also the common manganese form observed during stormflow in this aquatic environment ( M c C a l l u m 1995)  H i g h ratio o f R E fraction o f iron in street runoff implies that large amount o f dissolved iron may be in the ionic forms when it was released form trap under reducing conditions. However ionic form o f iron was rarely found in the streams, where the R E / T D ratio was only 3-6%. It has been found that metal iron in the stream is mainly in the particulate phase such as Fe oxides or bound to the organic matters and clay minerals (Shuman 1977, Stecko and Bendell-Young 1999). It can be concluded that very little iron in streams was bioavailable.  c. R E / T M The proportion o f the dissolved bioavailable metals (RE) over the total trace metals is summarized in Table 4-4. It gives the percentage o f trace metals from the watershed that w i l l potentially contribute to the toxicity of organisms through cell uptake. A l s o shown in the Table 4-4 is the R E / T M ratio obtained from Morrison's study in Bergsjon - a small urban watershed in Sweden (Morrison 1985).  Table 4 - 4 Ratio (%)  R E / T M for C u , Z n , M n and Fe Metal  Site #2  Site #1  Site #3  Morrison (1985)  RE/TM  Cu  11  11  7  18  Zn  45  42  51  19  Mn  21  60  41  ~  Fe  3  0.2  1  66  L o w ratio o f R E - C u (resin removable fraction o f Cu) over its total concentration was measured in a range o f 7 % - l 1% at three sites, which was even lower than that observed in storm runoff in Bergsjon watershed. L o w ratio o f R E - C u indicates the low ionic form o f copper in the water, which is largely affected by p H o f the water environment. Allard et al. (1987) pointed out that the partitioning o f hydrolysable transition elements such as C u , Z n , C d and Pb between dissolved and particulate phase is very strongly related to p H o f the aqueous environment. It was suggested that the p H in the stream greater than 5.5 led to the trend o f C u adsorption into the solid phase (Morrison 1987). A n adsorption curve obtained from laboratory experiment showed that C u mostly existed in the particulate phase in normal freshwater p H condition, which jumped from 60% to 90% when the p H increased from 5.5 to 7. The particulate Z n , on the other hand, increased from 10% to 70% with p H change in the same range (Allard et.al. 1987).  The p H range o f Brunette watershed was measured as 6.2-6.8, a range when low ionic C u and high ionic zinc was expected. However, the formation o f the particulate Z n was also expected in this p H range. A l l o w a y and Ayres (1993) pointed out that in general, C u tended to be adsorbed by sediments very strongly, while Z n was usually held weakly. This implies that Z n is likely to be more labile and bioavailable. The transfer coefficients o f trace metals in the soil-plant system also supported this conclusion (Kloke et al. 1984). The transfer coefficient - the concentration o f metal in the aerial portion o f the plant relative to total concentration in the soil - was 0.01-0.1 for C u and 1-10 for Z n , which indicates that C u is more strongly bound to the soil colloids and less available to the plant than Z n . Resin extractable zinc (RE-Zn), as expected, accounted for much higher portion, which made up approximately 50% o f the total Z n ( T M - Z n ) in the urban water.  The R E / T M ratio o f manganese indicated that approximately 2 1 % o f M n was in the dissolved bioavailable form when washed off from the urban street. The higher ratio o f 60% at up stream and 4 1 % at down stream sites showed a similar pattern to that o f the T D / T M . A s described before, the elevated manganese was from storm runoff, o f which a considerable amount o f M n was in the soluble form (due to its sources possibly the combustion products of M M T ) . Therefore, the high ratio o f R E / T M indicated the existence o f ionic manganese in  67  the street (Site #1) and in Still Creek (Site #2). The lower R E / T M ratio at the downstream site might be attributable to the oxidation process when manganese is transported through Burnaby Lake (Mattu and Hall 1999). It was reported that the high concentration o f organic matter or bacteria in the lake could play a role in this oxidation, where M n electrons and being oxidized to the more insoluble M n  4 +  2 +  was loosing  form (Chapnick et al. 1982, Sunda  and Kleber 1994).  Very low R E / T M ratio o f 3% for iron in the street runoff demonstrated that only a small portion o f iron was in the ionic form from surface runoff. Even lower ratio o f R E / T M in streams indicated that most of iron was in the particulate forms when entering the urban receiving water.  4.1.2 Metal partitioning in suspended sediment 4.1.2.1  Partitioning results  Three fractions o f particulate metals in suspended sediments - exchangeable fraction ( E X ) , easily reducible Mn-oxide bound fraction (ER), and residue fraction ( R E S ) - were measured for metals C u , Z n , M n , and Fe. The detailed results o f the speciation are provided in Appendix B .  Figure 4-10 shows the copper partitioning in the particulate phase in both baseflow (BF) (August 1998) and stormwater (SF) samples (November 1999 and January 2000). It can be seen that under baseflow conditions, quite similar distributions o f copper in three fractions were observed at all three sites, with the percentage o f each fraction o f 28-36% in E X , 2737% in E R , and 21-43% in R E S .  The variation o f copper partitioning in stormwater was also observed in F i g . 4-10. Three particulate C u fractions in street runoff (Site #1)  showed a consistent ratio among  E X / E R / R E S as 1/1.9/4 (averaged over all stormwater samples). A t stream sites, relatively higher E R - C u was observed in the November samples in both Still Creek (Site #2) and Brunette River (Site #3). The ratio o f E R - C u over R E S - C u as 1.5-2.7/1 was obtained. In  68  January samples, however, E R - C u and R E S - C u were only recorded at Site #2, and the average ratio o f E R - C u / R E S - C u was found to be 1/1.3. It should be noted that, due to low SS collection, exchangeable fraction ( E X fraction) was not measured in the suspended sediment of stormwater at these two stream sites. A s such, the E X fraction was calculated as a component o f E R - C u fraction.  Figure 4- 10 Distribution o f copper in three particulate fractions (BF-- baseflow, SF— stormflow)  Zinc partitioning in the particulate phase is given in F i g . 4-11. The residual zinc fraction (RES) was found prominent in the baseflow samples, which ranges from 56% - 80% in the SS collected both in the sediment trap and in streams. For the stormwater, the partitioning of E X - Z n , E R - Z n and R E S - Z n in street runoff (Site #1) was quite similar over four samples, with the average ratio among these three fractions o f 1/1.57/2. A t the two stream sites, a similar partitioning pattern was observed, o f which the E R - Z n fraction (averaged over two samples) was 58% at Site #2 and 53% at Site #3. This implied the existence of potentially bioavailable zinc associated with the suspended sediment in the water streams.  69  IEX D E R  Zn-speciation  E2  RES  SF  Figure 4- 11 Partitioning o f zinc in three particulate fractions at three sites (BF— baseflow, SF— stormflow) Figure 4-12 shows the results o f manganese partitioning in the matrix o f the suspended sediment (SS) at the three sampling sites. A consistent high E R - M n fraction was observed at the two stream sites, in which higher concentration o f E R - M n was found in the Brunette River (Site #3) than that in Still Creek (Site #2) under all flow conditions. The highest concentration o f E R - M n was found at Site #3 (stormwater in November), which was as high as 7508 mg/kg SS. In the sediment trap (Site #1), ratios o f the three particulate fractions E X / E R / R E S was 1/0.5/1.03 in baseflow and 1/1.1/2.7 during stormflow (averaged over four stormwater samples).  70  o  EX D E R ^ R E S  Mn-speciation  o  SF  1397  00  BF  I  SF  SF SF SF  o<3>  ^  <^  Site 1  ^  *° •  Site 2  *  ^  ^  ^ Site 3  Figure 4 - 1 2 Distribution o f manganese in three particulate fractions (BF-- baseflow, SF— stormflow) The predominant component o f particulate iron was found in the residue fraction in both baseflow (80-90%) and stormwater (70-80%) except for November samples, as shown in Fig. 4-13. In the stormwater, relatively higher concentrations o f iron in E R - M n fraction were observed at the two stream sites, with the value o f 4.5-116 g/kg SS. In baseflow samples, iron in E R - M n fraction among three sampling sites ranges from 0.6-1.3 g/kg SS. The iron in exchangeable fraction (EX-Fe) was lower in both baseflow (0.2-0.4 g/kg SS) and stormwater (0.06-0.6 g/kg SS).  71  Fe-speciation  SEX D E R 0 RES 210  SF  |  CO CO O) c o  '  ©  re c  0) O  o  c » o o  M •  1  ^  SF SF  SF  ^  BF  SF  • I  SF  BF  1 1  ^  Site 1  Site 2  Site 3  Figure 4- 13 Distribution o f iron in three particulate fractions at three sites. (BF— baseflow, SF— stormflow)  4.1.2.2 Discussion on particulate metal components O f the three metal particulate forms in suspended sediments, exchangeable ( E X ) and easily reduceable (ER) metal froms are regarded as potentially bioavailable. The bioavailable components o f the particulate trace metals at the three sites, as well as their seasonal variations, w i l l be discussed below. Due to the fact that most o f the particulate iron is in R E S form, and also it is a less toxic metal in the water system, the particulate bioavailable iron is not discussed here.  a. Bioavailable Cu in suspended sediment The partitioning o f particulate C u in suspended sediment (SS) in four stormflow and one baseflow conditions is shown in F i g . 4-14. It indicates that SS generally contains a large portion o f the potentially bioavailable copper, namely E X - C u and E R - C u (bound to M n oxides). E X - C u and E R - C u accounted for approximately 60% o f the total particulate C u in baseflow samples at all three sampling sites. O f which E X - C u along accounted for about  72  38% o f the total particulate copper, which was similar to the result o f E X - C u = 3 5 . 2 % from a Swedish catchment study (Morrison 1985). E R - C u contributed a relatively small portion (20%>) during baseflow, which was probably due to the manganese oxide formation in water column.  Cu  IEX  BF  o o  SF  SF  SF  SF  I 1 IT I1 I  i  1  T  III UA  o  00 o co  BF  SF  SF  SF  DER BF  E3  RES  SF  1 | II | IT iI! I  f-p-p-f I ' l l  o  TT  o  CN  ^ Y*  A  \^ N  ^  J>  \f>  4>  v  ^  Site 1  Site 2  Site 3  Figure 4 - 1 4 Fractions o f particulate C u in SS at three sites (BF - baseflow; SF - stormflow)  During the rainfall event the residual fractions were >50% from the street runoff samples, while in the stream sites, a great proportions o f the C u was bound to M n in the E R - C u phase of the suspended sediments. A similar pattern was also reported by Stecko and BendellYoung (1999). Their results from SS collected in the Fraser River Estuary showed that the great proportions o f E R - C u was found during the winter months in the wet weather.  b. Bioavailable Zn in suspended sediment The percentage o f the bioavailable components o f Z n in SS (Fig. 4-15) varied significantly from baseflow to stormflow at Site #1 (sediment trap), where bioavailable fractions ( E X - Z n  73  and E R - Z n ) accounted for only 20% o f the total particulate Z n in baseflow, but increased dramatically to 60% in stormflow. A similar ratio o f 22.3% in baseflow was also reported by Morrison (1985) in his study o f two urban catchments. He reported that the E X - Z n is derived mostly from road runoff solids and gully pot basal sediments.  EX D E R  Zn 5?  BF  SF  SF  SF  SF  BF  SF  1  o  00  (0  - l - l - l -  o  _1_  o  TJ-  ri  -JJJ-IJ  SF  H  U J  BF  RES  SF  SF  l-J-jj-  o  CM  Aug  Nov.(a)  N o v . (b)  Jan (a)  Jan (b)  Aug  Site 1  Nov.  Jan. (b)  Aug  Site 2  Nov.  Jan. (b)  Site 3  Figure 4 - 1 5 Fractions o f particulate Z n in SS at three sites (BF - baseflow; SF - stormflow)  The bioavailable components o f Z n at two stream sites, on the other hand, accounted for a stable but relatively low portion in both baseflow and stormflow conditions, ranging from 3040%). This may be attributed to the binding o f Z n to Fe-oxides and other organic and inorganic materials existing in the water column, as a high level o f particulate Fe and color was observed in the watershed streams. However, E X + E R fractions still provide an important potentially bioavailable Z n pool within suspended sediment, which may be released into the soluble phase under normal p H conditions or into the water column with the significant drop in p H or redox conditions (Morrison 1985).  c. Bioavailable Mn in suspended sediment  74  Figure 4-16 demonstrates three component o f particulate manganese in SS among the three sampling sites under different flow conditions. It can be seen that the partitioning pattern o f M n varied significantly from site to site and under different flow conditions. H i g h variability of E R - M n proportion in SS was also observed in previous studies (Stecko and BendellYoung 1999). They suggested that the high variability in concentrations o f E R - M n in the suspended sediment could be related to many interactions between biota and M n oxide.  EX D E R  Mn BF  o o  SF  SF  SF  SF  SF  TIN  o CO  o  to o TJ  o  0 SF  RES SF  m r n r1 n  'I  CM  1  <oi  ^  -fr  ^  Site 1  Site 2  Site 3  Figure 4 - 1 6 Fractions o f particulate M n in SS at three sites (BF - baseflow; SF - stormflow)  However, much higher E R - M n proportion was generally found i n the streams than in street runoff, which caused the ratio o f particulate bioavailable M n to be higher in streams than in street runoff.  4.1.3  Bioavailable Copper and Zinc in the Watershed  Copper and zinc are two o f the most important and toxic trace metals in the Brunette water system. The concentrations and distributions o f these two metals and their bioavailable components are discussed below.  75  4.1.3.1 Fractionation o f C u and Z n Five major fractions o f C u and Z n , namely R E for resin extractable and S T R - B for strongly bound (colloid form) in the dissolved phase and E X for exchangeable, E R for easily removable and R E S for strongly bound in particulate phase, in stormflow at three sites are summarized in F i g . 4-17. It should be noted that due to the limited SS samples collected at Site #2 and Site #3, the particulate metals were only partitioned into E R and R E S fraction, and no further partitioning for E X was done at these two sites. E R fraction at these two sites can then be treated as E X + E R - the total potentially bioavailable metal in particulate phase.  Figure 4-17 demonstrates that a significantly high proportion (80-90%) o f total dissolved C u was observed at stream sites. However the R E fraction was only a small portion (7-11%), indicating that copper in the dissolved phase is largely in colloidal form. The dissolved C u in street runoff (Site #1), on the other hand, was only approximately 30% o f the total C u , of which about two thirds were in the colloid form ( S T R - B ) . This indicates that most o f C u discharged from the urban impervious surface was in particulate and non-bioavailable forms.  O f the high percentage o f particulate C u in street runoff, about 25% was associated with the Mn-oxide fraction, which may imply that M n compounds on the urban street play important role in the transport o f C u . The E X + E R C u in the particulate phase accounted for about 30% of the T M - C u , indicating the potential move o f the bioavailable C u fraction from urban nonpoint source to the receiving water. The lower percentage o f particulate C u found at stream sites is possibly due to the deposition o f large amount o f metal associated particles onto the bed sediment o f stream and lake (Macdonald et al 1997).  The partitioning pattern o f Z n was quite similar at all sampling sites. Relatively higher dissolved Z n was found both in street runoff and in streams, which was similar to Morrison's study (1985). The R E - Z n fraction made up 42 - 51 % o f T M - Z n at all three sites, indicating a  76  high level o f Z n in ionic and weakly complexed forms. It was believed that the high level of •KB  cu-#i 100  80  80  60  60  Zn-m  40  o  I SIRS •adved  100  20  I  20  RE RriaJ*>  Cu-ffi  100  80  SIRS Dssdved  ER  RES  PartiaJate  2h-#2  80  c o u  eo  60  (0  o  •o2 40  0)4,  20  20  STRB  Rrtaiie  100  EX  OJ-«3  100  EX+t±< Ftriaiao  FES  Zn-#3  80  80 o  S  60  1  o  20  M  i  SIRS •ssctad  o (0  » 40 o  •  20  STRB  EX-tS!  EX+€R Faticidate  RES  Figure 4- 17 Copper and Zinc partitioning in five fractions among the sampling sites  77  E R - Z n in stormwater was attributed by the soluble Z n compounds from urban impervious surface (Macdonald et al. 1997) and the behavior o f Z n in the freshwater system (Florence an Batley 1980). Particulate Z n accounted for approximately 30-40% o f total zinc at all sites, while one o f its bioavailable fractions, E X - Z n , made up only 8% o f the total Z n in the surface runoff - much less than that o f 22.3% in Morrison's study (1985). The other bioavailable particulate fraction E R - Z n was in the range o f 11%-16% at all three sites.  4.1.3.2 Concentrations of bioavailable Cu and Zn in the watershed The concentrations o f bioavailable C u and Z n , i.e. the resin extractable fraction (RE) in dissolved phase and the exchangeable and easily removable fractions ( E X and E R ) in particulate phase, are shown in F i g . 4-18. The results reflect three flow conditions at three sites. It can be seen that the concentration o f Z n is much higher than that o f C u , in every fraction o f the three bioavailable forms, at all three sites.  Fig. 4-18 illustrates that the concentrations o f bioavailable C u in both dissolved and particulate phases were mostly in the range o f N D - 1 0 p g / L ( N D - non-detectable). The maximum concentration o f the bioavailable C u was always found at Site #1 - the street runoff site. This indicates that the street runoff is one o f the major sources o f copper to the watershed streams. The concentrations o f the bioavailable C u at the two stream sites were significantly lower, and the concentration at downstream site (Site #3) was further lower. The regulating and sedimentation function o f the Burnaby Lake lying between Site #2 and Site #3 is quite obvious.  Figure 4-18 also indicates that the concentration o f the bioavailable C u is higher in stormflow than in baseflow, for all the three sites. The concentration o f dissolved C u (RE) was always lower than that o f the particulate C u ( E X + E R ) in street runoff at Site #1, especially in stormflow. However, this difference was not so obvious in streams, where the sedimentation of the particulate metals occurred.  78  B a s efIo w  Cu  Cu #2  #1  Cu  Z n  Z n  Z n  #3  #1  #2  #3  S a m p l i n g  s i t e s  (b) Storm  C u #  1  -January  C u  C u  Z n  Z n  Z n  #2  #3  #  #2  #3  Sampling  1  sites  (C)  Figure 4- 18 Bioavailable C u and Z n levels in the watershed under baseflow (a) and under stormflow (b) and (c)  79  M u c h higher concentrations o f both dissolved and particulate Z n were also found in street runoff than in streams during rainfall - they were about two times higher than those in upstream (Site #2) and six times higher than those in downstream site (Site #3). This implies that street runoff was also the major source o f Z n to the water system. The concentrations o f both dissolved and particulate Z n were also significantly higher in stormflow than in baseflow in street runoff. However, this difference is not so obvious in streams, where most components  o f the bioavailable Z n showed comparable levels in both baseflow and  stormflow.  4.1.4 Regulating Effects of Mn on Cu and Zn It has been recognized that the manganese bound C u or Z n is produced when the dissolved C u or Z n interacts with particulate M n . If sedimentation occurs during the interaction, this decreases the concentration of C u and Z n in the dissolved phase and consequently reduces the bioavailable C u and Z n in the water system. It can be expected that the amount o f M n bound trace metals is determined by both the concentration o f particulate M n and the concentration o f dissolved C u or Z n .  4.1.4.1  Regulating effect on C u  A s presented in this chapter, Mn-bound C u (ER-Cu) accounted for a large proportion of the particulate C u - over 25% in most o f the samples. The relationship between total particulate M n and the M n bound C u , or specifically the relationship between total particulate M n and the ratio o f M n bound C u over all C u contributed to E R - C u ( T D - C u + E R - C u ) were analyzed. It was found that no clear trends o f manganese regulating copper in the stormwater was observed. The variation between E R - M n and T D - C u could be caused by various physical and chemical conditions in the water system.  4.1.4.2  Regulating effect on Zn  The same relationship was analyzed for Z n . A positive relationship between E R - M n and the ratio o f E R - Z n / T D - Z n + E R - Z n is evident in F i g . 4-19 (R=0.56). However, considerable  80  variability in this relationship makes it difficult to draw a firm conclusion o f the regulating effect o f manganese on the bioavailable Z n in the stormwater.  A s mentioned before, the regulating effect o f M n on C u and Z n in the water column is determined by the particulate fraction o f the metal M n . However, it has been found that the partitioning o f M n changes significantly in the Brunette water system. A monitoring program over a rain event in Brunette watershed showed that in Still Creek, M n in the soluble form was approximately 90% at the beginning o f the storm, and decreased to less than 20% with increasing flow (Mattu and Hall 1999).  Z n -M n 0.3  c Oi UJ  l tvi  a.  -i  .  0.25  •  0.2  ^  0.15  0.1  .  UJ  •  •  0.05  0  -I 0  1 5  1  1  1  1  10  15  20  25  ER-Mn  concentration  30  (ug/L)  Figure 4 - 1 9 Effect o f M n on Z n bioavailability A t the Brunette River site, the dissolved M n fraction made 20-40%) o f the total M n . However, this suggests that the M n seemed to be oxidized as it traveled down into the lake, since M n was mostly in the soluble form when entering Burnaby Lake, and mostly in the particulate form when leaving the lake. Therefore, it can be expected that the change o f particulate M n in the watershed waterways is likely to help regulate particle associated C u and Z n in the water system, which w i l l reduce the dissolved C u and Z n phases especially downstream o f Burnaby Lake. This is confirmed by the higher levels o f M n and associated Z n and C u that precipitate on the resin membranes during storm events at the Brunette River site.  81  4.2 Bioavailable metal study by dialysis with resin receiver in situ  The field experiments using dialysis with resin receiver were conducted with both long time (1-2 weeks) and short time (24 hours) exposures to simulate the uptake o f trace metal species by the aquatic organisms in the urban stream. The uptake rates for all targeted metals were measured in the sediment trap and at the two stream sites. The experiments were conducted seasonally based on the flow pattern in the urban streams (two in baseflow and four in stormflow and snowmelt conditions). A complete data set for the metal uptake is listed in Appendix D .  4.2.1 Trace metal uptake under baseflow condition Trace metals taken up by receiver resin and membrane in baseflow condition were examined in two exposure periods in 1998. One was conducted for 7 days in July, and the other was for 15 days in August.  4.2.1.1 Copper uptake The copper taken up by receiver resin and membrane at the three sites is shown in Fig. 4-20. The lowest resin uptake rate was found in the sediment trap at Site #1, with the range o f 1.95 -2.89 pg/mm .h. The uptake rates at Site #2 and Site #3 were found higher, ranging from 5.4 2  to 5.9 pg/mm .h at Site #2 and 4.9 to 5.4 pg/mm .h at Site #3, respectively. It was noted that 2  2  although the resin uptake rates of the copper varied from site to site, they were relatively consistent in July and August during baseflow at all three sites. This may imply that the bioavailable copper species in the watershed during the drier summer period is low but at consistent levels.  82  tM  E E  a> Q. 3  S ite 2  S ite 3  July  (a) Resin uptake  CM  E E  "5 > a.  3  S ite 3  July  (b) Membrane uptake Figure 4- 20 Copper taken up in baseflow by (a) resin and (b) membrane The membrane uptake rate for copper, on the other hand, varied not only from site to site, but also from time to time (Fig. 4-20 (b)). Relatively higher uptake was found in the sediment trap than at the stream sites. The range o f uptake rate o f 5.9 -6.7 pg/mm .h was found at Site #1, whereas the range of uptake rate o f 2.4-5.8 pg/mm .h and 2.7-4.9 pg/mm .h was observed 2  at Site #2 and Site #3, respectively.  83  2  4.2.1.2 Zinc uptake The receiver uptake for zinc was quite variable between the two experiments (Figure 4-21).  (a) Resin uptake  (b) Membrane uptake Figure 4- 21 Zinc taken up in baseflow by (a) resin and (b) membrane In July, the uptake rates by resin at all three sites were quite similar (with the range between 19-25 pg/mm .h), and so were the uptake rates by membrane (13.5-15 pg/mm .h). However, the differences between the uptake rates by resin and by membrane were quite significant at the sampling sites. In the two experiments, the maximum resin uptake rate was found at Site #2, while the highest membrane uptake was found at Site #1 and Site #3, respectively. These results might be due to the different physical and chemical conditions that affect the uptake of zinc at these three sites. For instance, a water condition appropriate for membrane adsorption or precipitation may prevent the transmission o f metal species inside the dialysis bag, and allow uptake by the resin.  84  4.2.1.3 Manganese uptake Manganese taken up by the receiver is shown in Fig. 4-22. The high resin uptake rates were found in the sediment trap (Site #1) in both experiments, which was 265 pg/mm h in July and 2  124 pg/mm h in August. Significantly low resin uptake rates were measured at the two stream sites - 24-35 pg/mm h in Still Creek (Site #2) and 15-16 pg/mm h at Brunette River 2  2  (Site #3). A totally opposite pattern o f the membrane uptake was observed at these three experimental locations, where the lowest uptake rate o f 3-7 pg/mm h was at Site #1, and 2  highest rate o f 344-486 pg/mm h was at Site #3. However, membrane uptake was also low at 2  Site #2 with the range o f 35-58 pg/mm h. 2  S  ite  3  (a) Resin uptake  July  (b) Membrane uptake Figure 4- 22 Manganese taken up in baseflow by (a) resin and (b) membrane  85  4.2.1.4 Iron uptake The distribution o f iron uptake at the three experimental sites was quite similar to that o f the manganese, as shown in Fig. 4-23. Very high resin uptake rate (6962 pg/rnm h) was found at 2  Site #1 in July's experiment. The rates at the two stream sites were also significantly low - 98 2  2  pg/mm h at Site #2 and 65 pg/mm h at Site #3. Similarly, the highest iron uptake rate by the membrane was also found in Brunette River (Site #3) - with the value o f 1234 pg/mm h in 2  July. The rates o f iron taken up by the membrane at the two other sites were similar, which 2  2  was 398 pg/mm h at Site #1 and 355 pg/mm h at Site #2 in July's experiment.  S  Ite  3  July  (a) Resin uptake  0)  in  July  (b) Membrane uptake Figure 4- 23 Iron taken up in baseflow by (a) resin and (b) membrane  86  4.2.2 Uptake of trace metal in stormflow and snowmelt in the field Four one-week-exposure  experiments were conducted during stormflow and snowmelt  conditions. They were snowmelt condition in February 1999, stormwater conditions in October and November 1999, and January 2000, respectively. It should be noted that the uptake data are not available for the experiment in October 1999 at Site #2, as the receiver was lost during the experiment. The results o f trace metal uptake study over one rain event are also presented in this section.  4.2.2.1 Copper uptake Although the copper uptake rates by resin were quite variable in the four experiments, they showed little difference at the three experimental sites, as shown in Fig. 4-24(a).  (a) Resin uptake  (b) Membrane uptake Figure 4- 24 Copper taken up in stormwater by (a) resin and (b) membrane  87  The maximum uptake was recorded in a stormflow in October 1999, which was at least three times higher than those in other three stormflow or snowmelt. The membrane uptake o f C u also varied greatly in the four experiments, as shown in Fig. 424(b). However, very low rates were recorded in October 1999 at all sites, which corresponded to the maximum resin uptake rate o f the four experiments. Higher membrane uptakes were observed in February and November 1999. The highest rate was in February at Site #2 o f 33.3 pg/mm .h. 2  4.2.2.2 Zinc uptake The maximum resin uptake of Z n was also found in October 1999, especially in the sediment trap (115 pg/mm -h) (Fig. 4-25). The uptake rates from the other experimental periods were 2  significantly lower - ranging from 20 to 40 pg/mm -h. Membrane uptake of Z n , on the 2  contrary, was also lower in October's experiment. H i g h membrane uptake was detected in February and November at both Site #1 and Site #2. Also like the case for copper, the zinc uptake rates by membrane were considerable low for all the experiments at Site #3.  4.2.2.3 Manganese uptake Manganese uptake by resin was found much higher at the two stream sites than that at the sediment trap, as shown in Figure 4-26(a). It ranged from 52-88 pg/mm .h at Site #2, and 602  82 pg/mm .h at Site #3. While at Site #1, the manganese uptake rate was only 8-35 pg/mm h of the four experiments.  Manganese uptake by membrane showed the opposite pattern to the uptake by resin at site #1 and site #3 (Fig. 4-26(b)), i.e., the higher resin uptake corresponded to the lower membrane uptake. However, high membrane uptake o f manganese was also found at Site #2, with the rates ranging from 45-147 pg/mm h. It may imply the high manganese concentration in the Still Creek. Figure 4-26 also demonstrates that the relatively low manganese uptake rates were found at Site #1, the sediment trap, both by resin and by membrane.  88  Feb.99  Oct. 99  Nov.  99  Jan.00  (a) Resin uptake  Feb.99  O c t . 99  N o v . 99  Jan.00  (b) Membrane uptake Figure 4- 25 Zinc taken up in stormwater by (a) resin and (b) membrane  89  (a) Resin uptake  (b) Membrane uptake Figure 4- 26 Manganese taken up in stormwater by (a) resin and (b) membrane  4.2.2.4 Iron uptake Iron taken up by resin was at low levels compared to their precipitation onto the membrane, as shown in F i g 4-27(a) and (b).  90  • #1 m*2  H#3  (a) Resin uptake  o o o m  (b) Membrane uptake Figure 4- 27 Iron taken up in stormwater by (a) resin and (b) membrane In the sediment trap, the uptake rate by resin was quiet variable, ranging from 16-250 pg/mm -h in stormflow and snowmelt. The relatively constant high uptake was measured in 2  Still Creek at Site #2, where the rates were from 105-141 pg/mm .h. Lower uptake was 2  recorded at Site #3 at 14-83 pg/mm -h. The membrane uptake rate also varied significantly 2  from one experiment to another, with the highest rate o f 3188 pg/mm .h in November and lower than 100 pg/mm .h in January. It can be seen clearly that the higher membrane uptake 2  91  rate corresponded to lower resin uptake rate at this site. H i g h rates were also occurred at Site #2 in all experiments, ranging from 1020-4112 pg/mm .h. Membrane uptake o f iron at Site 2  #3 was at the rate o f 125-785 pg/mm .h, where the relatively higher rates were found in 2  October and November.  4.2.2.5 Trace metal uptake over one rain event One storm metal uptake experiment  was conducted in November  1999. During this  experiment, the receivers were installed before the rainfall at the three sites, and collected after the rainfall. The rain lasted for about four hours in this experiment (See Fig. 4-32 for the detailed rain and flow). The total exposure time of the receiver was about 24 hours. The resin uptake results o f the four metals are shown in Figure 4-28, and the membrane uptake results are shown in Figure 4-29.  It is evident from Fig.4-28 that the bioavailable copper and zinc in the stormwater quickly penetrated the dialysis membrane and complexed with resin inside the bag. The uptake rates were as high as 6.8-11.4 pg/mm .h for copper, and 13-25 pg/mm .h for zinc at all three sites. 2  2  The high uptake rate indicates that the receiver is highly sensitive to the available metal species in the stormwater. M u c h higher uptake of manganese by resin was found at the two stream sites. It was at 194 pg/mm .h at Site #2 and 174 pg/mm .h at Site #3. This finding may imply that with the effects o f the storm runoff, manganese is largely in the bioavilable form (ionic or weakly complexed form) in transport through Brunette River waterways.  For the membrane uptake (Figure 4-29), only iron demonstrated a high uptake rate in the stormwater. The rate o f iron taken up by membrane was 128 pg/mm .h at Site #1, 646 2  pg/mm .h at Site #2, and 160 pg/mm .h at Site #3. Copper taken up by membrane was not 2  2  detected at all three sites over this four-hour rainfall exposure period. Zinc taken up by the membrane was measured as 7-12 pg/mm .h. The rate o f membrane uptake for manganese was 2 pg/mm .h at Site #1, 12 pg/mm .h at Site #2, and 28 pg/mm .h at Site #3. 2  2  2  92  Resin  uptake  #1  #2  fflCu  MZn  #3  EDI M n E l F e  Figure 4- 28 Metal uptake by resin - over one storm  M em b r a n e u p t a k e  0Cu  I Z n  I M n E3Fe  Figure 4- 29 Metal uptake by membrane - over one storm  4.2.3 Resin uptake of trace metals in the watershed Figure 4-30 presents the receiver resin uptake rates o f copper(a), zinc(b), manganese(c) and iron(d) from the experiments at three sites at difference flow conditions over the study periods. Their distributions and variation w i l l be discussed.  93  (b)  94  CN  E E  "3>  a. o CO i_ O  ro Q.  Jul. 98 (BF)  A u g . 98 [BF)  N o v . 99 (SF)  J an. 00 (S F )  #1  (C)  Fe - R e s i n  uptake  C  *-  o  CQ  a  J u l . 98 (BF)  A u g . 98 (BF)  Feb.99 (SM)  O c t . 99 (SF)  N o v . 99 (SF)  Jan.00 (SF)  (d) Figure 4- 30 Spatial and seasonal variation o f C u , Z n , M n and Fe uptake by resin ( B F - baseflow, S M - snowmelt, and SF - stormflow)  4.2.3.1 Seasonal variation of resin uptake A similar pattern o f resin uptake was observed for both C u and Z n (Fig. 4-30 (a) and (b)). During summer baseflow (July and August 1998), the uptake o f copper and zinc by receiver was relatively consistent. However, the uptake rates o f these two metals varied considerably from storm to storm in the wet season. A much higher uptake rates for both C u and Z n were  95  observed in stormwater in October - the beginning o f the wet season o f the year. The significant variation through different storms was also observed in Morrison's study (1985). The reason for this uptake variation is expected to be due to the variation of bioavailable trace metal levels in the stormwater and the flow regime in the urban waterway, possibly the dilution factor. There is no significant difference o f C u and Z n uptakes between stormflow in January and snowmelt in February at all three sampling sites, which differs from Morrison's conclusion that higher uptake was detected during a snowmelt period. This may be due to the climate difference o f the two watersheds. In the Brunette watershed, rainfall events usually follow soon after a snow, which makes the snow melt very fast. The accumulation o f metal pollutants in the snow does not usually have time to occur.  The manganese uptake was totally different from those o f C u and Z n as shown in F i g . 430(c). The maximum uptake rate was measured in baseflow in the sediment trap, which was much higher than the others. This is possibly due to the release o f manganese from the reducing environment o f the bottom sediment caused by the organic matter degradation under quiescent conditions in the trap, since the manganese is a redox sensitive species (Stecko and Bendell-Young 1999). In stormflow conditions, higher uptake rates occurred mostly in streams.  It can be seen in Fig. 4-30(d) that, except for the sediment trap under baseflow, low uptake rates o f iron were found in all receiver experiments. This phenomenon clearly demonstrated that iron in the urban water system exists mostly in the non-bioavailable fractions. The extremely high uptake o f iron in the sediment trap during baseflow (July 1998) seems due to the release o f ionic iron from the iron binding fraction o f the sediment under stronger reducing conditions in the trap. There was a very long dry period before and during the experiment, and which leads to the dramatic decrease o f dissolved oxygen (DO) (60-80% decrease o f the D O saturation level caused by the bacteria activities) in the sediment (Morrison 1985).  96  4.2.3.2 Spatial distribution of resin uptake The resin uptake rates o f C u and Z n in baseflow were relatively lower in the sediment trap (Site #1) than in streams (Sites #2 and #3). However, according to the discussions in Section 4.1.3.2 o f this chapter, the concentrations o f R E - C u and R E - Z n were found mostly lower in streams than in the trap. One possible explanation o f the high metal uptake versus low bioavailable metal concentration in baseflow is the difference o f trace metal pools in both sediment trap and streams environment. While the receiver was exposed to a low but consistent level o f bioavailable trace metal in streams, limited amount o f R E - C u and R E - Z n were formed in the trap during the dry period. Another possible explanation is the concentration gradient layer that encloses the resin receiver. The uptake rate may be determined by the thickness o f this layer, which is inversely proportional to the flow rate. Since there is no flow in the sediment trap during the dry season, the thicker gradient layer slows down the uptake rate at this site. However, this possible impact needs to be investigated further.  During the storm event, on the other hand, the higher uptake amounts o f C u and Z n were generally found in the street runoff than in the streams. The resin uptake rates o f C u and Z n at both street and stream sites were well correlated to the bioavailable copper and zinc measured in stormwater, which w i l l be discussed further in the following section.  Quite high uptake rates o f manganese were found in both upstream (Site #2) and downstream site (Site #3) o f the water system in the wet season. This is because a large amount of manganese, as the combustion products o f M M T , was discharged through urban impervious surfaces and contributed considerable soluble manganese which is potentially available to the organisms in the urban water stream (Mattu and H a l l 1999). The reason that less uptake o f M n was measured in the sediment trap during storms is unknown, since the R E - M n level at this site was found similar to those in streams. N o significant difference o f iron uptake was observed in the watershed. The possible explanation is that iron exists mostly in the nonbioavailable forms.  97  4.2.4 Environmental factors affecting the metal uptake Several factors that may affect the C u and Z n uptake were examined. They are found functioning in different flow conditions.  4.2.4.1 In stormflow condition In storm condition, two major factors have been regarded as the contributors for increasing the bioavailable metal uptake rate, namely the rainfall intensity and the antecedent dry period (Morrison 1987, H a l l and Anderson 1988). Previous work found that the dry period before the rainfall plays a very important role - the longer dry period causes more trace metal accumulation on the urban impervious surfaces, and leads to more trace metals flushed into the water system during the rain (Yuan et al. 1999). The metal uptake in four rain periods is summarized both for C u and Z n , and presented in Table 4-5. The analysis between Table 4-1 and Table 4-5 indicates that a 12-day dry period before the rain causes the C u uptake rate at Site #3 approximately four times as high as that o f a 5-day dry period before the rain, and five times as high as that o f 0-day dry period. Higher uptake o f Z n after a longer dry period was observed in the same experiment, but it was not as significant as that for C u .  Table 4- 5 Summary o f the resin uptake in four rain periods *  Resin uptake rate (pg/mm .h) Zn Cu  Resin experimental (7 days exposure)  Feb. 1999 Oct. 1999  Site #2  Site #3  Site #2  Site #3  3.6  2.7  38  23  15  —  —  49  N o v . 999  4.8  3.8  21  13  Jan. 2000  3.6  3.0  23  21  The other factor affecting the metal uptake rate, the rainfall intensity, is also listed in Table 41. However, this factor is not as significant as that o f antecedent dry days before the rainfall in affecting the metal uptake.  98  4.2.4.2 In baseflow condition H i g h uptake o f trace metals in baseflow conditions was expected to be due to the low level of dissolved oxygen (DO) in the water system. It was found that the D O level in Still Creek at Douglas station is consistently less than 5 mg/L during the summer (Macdonald et al. 1997), because Still Creek at this section is deep and slow moving, leading to the minimum reaeration. The typically low D O level has also been found in the upper reach o f the Brunette River, (Cariboo Road Station) with the measured value o f 2.6 m g / L in the summer time. The ecological effects in Burnaby Lake might also affect the D O level in the lower reach o f Brunette, since decomposition o f algae and macrophytes could result in decreased D O (Macdonald et al. 1997). L o w level o f D O may also lead to the release o f sediment bound metals ( E R - M n bound or R E - F e bound) into the water column as the dissolved phase that can be assimilated by the Chelex resin.  The uptake o f metal in the sediment trap during the dry season might be also affected by the same factor, where the reducing conditions in the trap cause the release o f particulate M n and Fe to the dissolved phase and further liberate C u and Z n bound to them. The release o f M n and Fe was indicated by the higher uptake of M n and Fe by the receiver resin, as described above.  F l o w rate might be another factor that influences the metal uptake rate, as stated in 4.2.3.2 in this chapter. However, this needs to be studied in future research.  4.3 Suspended sediment and trace metal transportation  4.3.1 Suspended sediment Suspended sediments were measured for total suspended sediment (TSS) over a rain event (Nov. 19, 1999), and the results are given in Appendix C . The changes o f the T S S concentration over a rain event at three sampling sites was shown in Figure 4-31. The  99  corresponding rainfall intensity in the region and the flow rates at the two stream sites during the rain event are shown in Fig. 4-32.  0 : 0  0  0 : 2  8  0 : 5 7  1:2  1:5  6  5  2 : 2 4  2 : 5 2  3 : 2  1  Time Figure 4-31 Variation o f TSS concentration over one storm at three sites (Nov. 19, 1999, sampling time refer to Appendix C )  0:00  0:28  0:57  1:26  1:55  2:24  2:52  3:2 1  Time  Figure 4 - 3 2 Corresponding rainfall intensity and stream discharge at sites #2 and #3 (Nov. 19, 1999) The significant increase o f T S S concentration in the surface runoff with the rainfall development is indicated at Site #1 (Fig. 4-31). This phenomenon agrees with the previous studies on the SS discharge pattern under stormflow that, the T S S increases with the storm  100  development (Jarvis 1992, Solo and Perkins 1997, H a l l , unpublished data). The slight increase o f T S S concentration at Site #2 and Site #3 are also shown in Fig. 4-31. A smaller change at Site #2 in the stream might be due to dilution and the effect o f flow regime such as low flow (Fig. 4-32) in Still Creek. However, a smaller change in T S S during the rainfall event at Site #3 was controlled by sedimentation in Burnaby Lake, even though a relatively high flow was observed in this site (2-6 m /s). The different T S S concentration trends among 3  these three sites indicate that the direct response o f TSS to rainfall intensity was significant in the street runoff from urban impervious surface (Site #1).  4.3.2 Trace metal transport with S S Trace metal contaminants discharged from urban impervious surface through storm runoff have been regarded as the typical non-point source pollution to the receiving water. Sampling o f street runoff (Site #1) is to sample closer to the potential pollutant sources as urban activity generates many o f the pollutants in the stormwater (MacDonald et al. 1997). Many previous studies found that trace metals transported through urban impervious surface is largely associated with total suspended sediment (TSS) (Hall and Anderson 1988, Jarvis 1992, Solo-Gabriele and Perkins 1997). In this study, an intensive sampling activity was conducted for street runoff. The T S S and metal C u , Z n , M n and Fe were measured through six sequential samples over a rainfall period and the results are presented in F i g . 4-33.  Very good correlation between the total metal ( T M ) and the total suspended sediment (TSS) has been found for each o f the four metals, with the correlation coefficient o f Rc =0.98 for u  C u , R z n - 0.92 for Z n , Rivin 0.94 for M n , and Rp = 0.92 for Fe. This conclusion further =  e  demonstrates that trace metals transported through urban impervious surfaces is largely associated with total suspended sediment.  101  TSS - Cu  o o oo  CT  E c cu o c o u 00 00  9:10  o o o  ^ — UI  E, C O  CO l_  9:55  10:40  11:25  12:05  10:40  11:25  12:05  TSS - Zn  o o 00  o O  <°  a!  5 c  o u 00 00  o o o o  CM  9:10  9:55  102  12:45  TSS-Mn  o Ui  c o I °  9:10  9:55  1 0:40  11:25  12:05  2  c « u c o o  1 2:45  TSS-Fe  o 00  o E  c g  _  9:10  9:55  10:40  11:25  12:05  c V o c o o  0)  12:45  Figure 4 - 3 3 Concentration change o f C u , Z n , M n and Fe with T S S over a storm event (Nov. 19, 1999 at street runoff (Site #1))  Figure 4-33 also indicates that T S S has no relationship with either dissolved metals (TD), or their bioavailable forms (RE) in storm runoff. This pattern was also observed by many other  103  studies, and the conclusion was drawn that metals in dissolved form are washed off during the early part o f a storm event (Jarvis 1992, Solo-Gabriele and Perkins 1997, McDonalds et al. 1997). It can then be concluded that the bioavailable trace metal species that can be directly taken up by the aquatic organisms are mostly washed off at the early stage o f the storm event. The storm intensity w i l l also be important regarding the mobilization o f street dirt and movement o f stream sediments and their associated particulate metals by the higher flow.  4.4 Other relevant data  Listed in the following were the relevant data that were also measured or collected in this study. They correspond to each of the sampling under baseflow and stormflow conditions.  4.4.1 pH The p H values were only measured in the water samples from the intensive sampling scheme. The results showed that stormwater p H was relatively constant with the value o f 6.4 - 6.7 at all three sampling sites, and the data were given in Appendix C in the intensive sampling data set.  4.4.2 Flow data and precipitation data Precipitation data during the study period, from January 1998 to January 2000, and the flow data at two sampling sites in the watershed streams were collected from G V R D , and are shown in Fig. 4-34.  It is obvious that the rainfall occurs mainly in the winter - from November to February. The rainfall in this wet season is several times more than that in the dry season - from June to October, as are the flow rates in the stream (shown in Fig. 4-34).  104  Aug-98  Nov-98  Feb-99  Date  May-99  Sep-99  Dec-99  Figure 4- 34 Rainfall and flow data over the study periods in the Brunette River watershed  4.5 Assessment of trace metal toxicity in the watershed  4.5.1 A s s e s s m e n t with relevant water quality criteria The potential toxicity o f trace metals C u and Z n in the Brunette River water system was assessed according to the existing water quality objectives (Swain 1989, E P A 1983, and E E C 1983). Table 4-6 compares the average total C u and total Z n in street runoff and streams with the Water Quality Objectives ( W Q O ) for the Brunette Watershed set up by the Ministry o f Environment (Swain 1989). This is a long term objective for protection o f aquatic life and wildlife.  It can be seen that both total C u and total Z n are over 10 times higher than those allowed by the Objective in the street runoff. The trace metal concentrations in streams, though are significantly lower than those in street runoff, but are still much higher than those recommended by the Objective. C u is 5.8 times higher and Z n is 4.4 times higher in Still  105  Creek (Site #2), and C u is 3.6 times higher and Z n is 1.3 times higher i n the Brunette River (Site #3) than the W Q O s .  Table 4- 6 Comparison o f total C u and Z n with Water Quality Objectives ( B C Canada) Sampling Sites  W Q O Standard ( B C ) C u (pg/L)  Cu  (pg/L)  Site 2  Site 1  Site 3  Hardness >30 m g / L as CaC0  Zn  3  76  29  18  5  Site 1  Site 2  Site 3  Z n (pg/L)  370  132  39  30  (pg/L)  When one compares the concentrations o f C u and Z n at the three sampling sites with the U S E P A criteria based on the total metal concentration ( E P A 1983), it can be seen that total C u of this study exceeds the 24-hour average at all three sites, and the maximum for freshwater toxicity at both Sites #1 and #2. The total C u in street runoff significantly exceeds the threshold o f the intermittent discharge criteria, and may create significant mortality. The concentration o f total Z n , on the other hand, exceeds both the 24-hour and maximum values for freshwater toxicity only in surface runoff (Site #1). Z n in the Still Creek (Site #2) exceeds the 24-hour criteria for freshwater toxicity, but is lower than the maximum limit. Total Z n at Site #3 is below the criteria listed in Table 4-7.  Table 4- 7 U S E P A Water Target Concentrations for total C u and total Z n Total  Freshwater toxic criteria  Intermittent discharge criteria  Trace  hardness  (ug/L)  (ug/L)  metal  (mg/1)  24 hour  Maximum  Threshold  Significant mortality  Copper  50  5.6  12  20  50-90  100  5.6  22  35  90-150  106  Zinc  50  47  180  380  870-3200  100  47  321  680  1500-4500  B y comparison to E E C ' s Environmental Quality Objective (EQO), a criterion based on the total dissolved metals (TD), copper does show toxic effects in the Brunette River water system since the dissolved C u concentrations are much higher than the E Q O criteria at all three sites (Table 4-8). For zinc, while the dissolved Z n exceeds the E Q O criteria at Site #1, it may also have possible toxic effect at Site #2 depending on water hardness. A g a i n at Site #3, the concentration o f dissolved Z n is far below toxic level.  Table 4- 8 Comparison o f total dissolved C u and Z n with E E C water quality criteria Cu Site #1  Site #2  17-32  24-25  (ug/L) E Q O standard ( E E C )  Site #3  1-6  15-20  Hardness = 0-100 m g / L as C a C 0 Zn  3  (ug/L)  Site #1  Site #2  Site #3  E Q O standard ( E E C )  179-266  77-92  30-41  70-170 Hardness = 0-100 m g / L as C a C 0  3  4.5.2 Assessment with bioassy uptake There is still no criterion for the trace metal uptake rate by organisms. The trace metal uptake rates measured in the Brunette River watershed are compared with previous work in the Bergsjon catchment (Morrison 1987) - a residential area (15.6 ha) located 10 k m northeast o f central  Goteborg,  Sweden.  The uptake experiment  in that work was  conducted  in  Backebolsmotel o f Goteborg - a drainage area o f a 600 m section o f the heavily used motorway.  107  4.5.2.1 Uptake in street runoff Table 4-9 shows that higher C u and Z n uptake rates both by resin and dialysis membrane were found in Brunette River watershed compared to that o f the Backebolsmotel. Both the resin and membrane uptake rates varied considerably in the experiments in both catchments. It can be concluded from the comparison that the bioavailable C u and Z n from street runoff are significantly higher in the Brunette River watershed than in Backebolmotel.  Table 4- 9 Trace metal uptake rates in street runoff (pg/mm . h ) 2  Brunette River Watershed Street Runoff  Backebolsmotel  (Site #1) Resin  Membrane  Resin  Membrane  7.9  10.3  1.6  2.6  (0.8-25.0)  (0.0-39.8)  (0.6-4.4)  (0.7-7.5)  50.2  47.9  9.5  2.9  (16.9-164.0)  (1.3-194.2)  (2.5-21.3)  (0.8-8.3)  Cu  Zn  4.5.2.2 Uptake in stream The trace metal uptake in stormwater, on the other hand, is at about the same level in the two studies, as indicated in Table 4-10. Except for a much lower membrane uptake rate o f Z n , the Bergsjon Catchment shows quite similar C u and Z n uptake rates by resin and membrane in the stormwater to those detected in Brunette River Watershed.  Table 4 - 1 0 Trace metal uptake rate in stream (pg/mm .h) Stormwater  Cu  Zn  Brunette River watershed  Bergsjon catchment  Resin  Membrane  Resin  Membrane  5.3  10.0  13.4  22.4  (0.0-16.0)  (0.0-37.4)  (0.6-41.3)  (4.4-78.7)  26.8  31.0  25.6  2.4  (10.6-64.0)  (0.0-124.5)  (4.0-67.1)  (0.9-4.5)  108  4.5.2.3 Uptake in baseflow Table 4-11 indicates that both studies detected relatively low C u and Z n uptake rates in baseflow. The uptake rates were quite similar in these two urban watersheds under baseflow conditions. The variable uptake rates in baseflow conditions were explained as the contamination o f in-pipe sediment remaining after storm events, which resulted in metals being leached by ground water entering the pipe system (Morrison 1987).  Table 4- 11 Trace metal uptake rate under baseflow (pg/mm . h ) 2  Stormwater  Brunette River Watershed Resin  Cu  Membrane  Bergsjon Catchment Resin  5.6  4.3  (4.9-6.4)  (2.8-5.9)  17.3  11.2  (5-27)  (4.0-14.4)  Zn  Membrane  A maximum biological uptake flux was used to assess Z n , i.e. 2.35 pg/mm .h - a maximum 2  rate o f Z n taken up by the seawater alga  Phaeodactylum cornutum (Turner 1983). It can be  seen that Z n uptake rates in the Brunette River watershed exceed this maximum biological uptake rate even in the dry season. The exceedance is much more serious in the wet season, in which the highest Z n uptake was found to be as high as 177.9 pg/mm .h in street runoff 2  and 62.3 pg/mm .h in streams. From this comparison it might be concluded that Z n is toxic in the Brunette River watershed, i f the Chelex resin can be considered analogous to an aquatic organism, such as algae.  4.6  Discussion on the two approaches  109  4.6.1 Relationship between the two approaches Trace metal partitioning and trace metal uptake experiments were the two  different  approaches to measure metal bioavailability. It is reasonable to believe that these two approaches are correlated. The correlation between the metal uptake and the trace metal partitioning was analyzed in this study. The receiver experiments and the water samplings at three sites i n two storm periods, seven days each in Jan.3 and Jan. 10 2000, were chosen for this analysis. The results demonstrated a fairly good correlation between the metal uptake and the resin removable fraction ( R E fraction) o f C u and Z n .  4.6.1.1 C u The correlation between C u uptake rate and C u concentration i n total dissolve (TD-Cu) and resin removal (RE-Cu) fractions in the water is shown in F i g . 4-35. It can be seen that the uptake rate o f copper seems well correlated with resin removable C u concentration in the water. The correlation coefficient o f C u uptake to the resin removable fraction o f C u ( R E Cu) in stormwater was R = 0.77. T D - C u and copper uptake showed less linear correlation compared to the R E - C u , with the correlation coefficient o f R = 0.59. N o clear relationship 2  between T M - C u and C u uptake was found. It can then be concluded that although the resin removal fraction o f C u accounts for only a small fraction o f dissolved C u , as well as in total C u , it is the bioavailable portion that determines C u uptake by the receivers.  110  Figure 4-35 Correlation o f C u uptake rate to T D - C u and R E - C u  4.6.1.2 Zn Positive correlation between receiver uptake rate for Z n and concentrations o f three Z n fractions, T M - Z n , T D - Z n , and R E - Z n , is shown in F i g . 4-36, i.e. the higher concentration o f the metal in the stormwater, the higher the metal taken up by receiving resin in the dialysis bag. The curve fittings indicate that the best correlation o f Z n uptake is to the bioavailable zinc (RE-Zn), with the coefficient o f R = 0.97. A significant correlation o f resin uptake rate with T D - Z n ( R = 0.94) and T M - Z n (R = 0.79) was also observed. However, it should be 2  2  noted that this is based on a very limited data set and additional texts are needed to confirm these results.  ill  •  TM  •  TD  A  RE R' = 0.7903  P o l y . (RE) - - - P o l y . (TD)  R' = 0.9372 ..£3  P o w e r (TM)  R = 0.9747  ,.-.<«3""  10  15  20  25 Uptake  30  35  40  45  rate ( p g / m m 2.h)  Figure 4 - 3 6 Correlation o f Z n uptake rate to T M - Z n , T D - Z n , and R E - Z n  4.6.2 Method suitability and experimental design  4.6.2.1 Partitioning M e t h o d In this study, the grab sample was used to measure bioavailable metal species i n the urban runoff. Even though the water was taken in the early rainfall period, the grab samples only represented a snap shot o f that rainfall event but might not represent the distribution o f bioavailable metal species in the runoff over time. The intensive sampling over a whole rain event is necessary i f a more meaningful mean value is desired, especially in the case o f heavy storms with highly variable runoff.  Filtration to collect suspended sediment can cause another problem that needs to be addressed. The filter itself may bring contamination to the analysis, especially when several filters are used for one sample. Using a continuous centrifuge i n the field may avoid this problem.  112  4.6.2.2 Receiver experiment The study o f trace metal bioavailability by using "Dialysis with Resin Receiver" technique is thought to be a more direct and more meaningful method. However, the variability o f resin and membrane uptake rates among the experiments was observed. This phenomenon reflects the intermittent and variable nature o f bioavailable metal discharge in the stormwater (Morrison 1987).  It should be noted that trace metals assimilated by organisms are expressed as an uptake rate with the unit o f pg / mm .h, which indicates that the uptake rate or flux o f the bioavailable 2  metal species through the dialysis membrane is closely correlated to the available metal species in the water, the dialysis membrane exposure area and the exposure time. The uptake rate therefore is a time-averaged value used to describe the available metal species that can penetrate a cell under a certain time period. It was found in this study that the exposure time is a very important factor that influences the uptake rate. This is because the exposure o f the receiver in the stream w i l l lead to the adsorption o f particulate metal and other substance on the membrane surface (especially during storm period). This may clog the pathway o f the metal species to the cell surface and lead to the non-linear uptake. The longer a receiver is exposed in the stream, the greater potential there is for adsorption to occur on the dialysis membrane. A s such, the exposure time design for this experiment must be considered, which should be correlated to the rainfall and the flow pattern in the urban receiving water. A comparative experiment with the Chelex resins in dialysis tubes and live organisms held in flow through enclosures needs to be conducted to confirm the use o f these resins as surrogate organisms for metal bioavailability studies.  113  Chapter 5  Summary Trace metals, their bioavailability, their spatial distributions, and their seasonal changes in the water system, were studied with two techniques, namely metal speciation and dialysis with resin receiver. Both techniques were used to investigate the trace metal bioavailability in the Brunette River watershed from July 1998 to January 2000.  The concentrations o f trace metals and their bioavailable forms both in dissolved (total water partitioning) and particulate (sediment geochemistry) phases in both street runoff and streams were measured. Control factors regulating the metal forms in the urban receiving water were analyzed. The potential assimilation of bioavailable trace metals by the aquatic organisms was simulated through dialysis with resin receiver technique. The uptake rate o f resin inside the tubing can provide an indication o f the bioavailable metal species in the urban receiving water and their potential accumulation and toxic effects on the aquatic organisms.  The spatial distribution o f bioavailable trace metals in the Brunette water system was analyzed along a hydrological gradient from street runoff to streams above and below Burnaby Lake. The temporal change o f the bioavailable metals was examined with the sampling and analysis under baseflow and stormflow conditions.  114  5.1 Potentially bioavailable trace metals in Brunette Watershed  The following conclusions can be made o f the potential bioavailable trace metals in Brunette water system.  >  Copper in three bioavailable components, R E - C u in dissolved phase and E X - C u and E R C u in particulate phase, accounted for 4 1 % in street runoff, 2 1 % in Still Creek and 9% in Brunette River o f the total copper concentrations over the period o f investigation. Although the dissolved C u contributed a large portion in the stormwater samples, it was mostly in the non-bioavailable forms. This study found that copper in the dissolved phase was largely in the colloidal fraction - 70% in the street runoff and 85-92% in stream sites. The high colloid C u in the dissolved phase may be attributable to organic matter from urban runoff that can contain humic-like substances and/or lignins and tannins - the leaf degradation products in the watershed (Macdonald et al. 1997). Thus, the immediate bioavailable C u (RE-Cu) concentration was relatively low - 5-6 ug/1 at street site, 3-5 ug/1 in Still Creek and less than 2 pg/1 in Brunette River. This represents approximately 10 % o f the total C u concentration ( T M - C u ) . C u in the particulate bioavailable fraction ( E X + E R ) also showed a low proportion o f the total C u in both street runoff (30%) and in streams (<10%). The particulate fraction o f C u ( E X + E R ) clearly declined from street runoff to Still Creek and then to the Brunette River under all flow conditions. The seasonal change o f these two bioavailable particulate fractions was observed only at the street runoff site, where a higher concentration o f particulate bioavailable C u ( E X + E R ) was found in surface runoff-fall and winter period (10-22 pg/1) than in baseflow (6 pg/1) drier summer period.  >  Zinc was found largely in bioavailable forms in the Brunette watershed.  Three  bioavailable Z n fractions accounted for 50-70%) o f the total Z n concentration at all three sampling sites. The R E - Z n dissolved form was predominant, and accounted for 40-50% of the total Z n . The particulate Z n in the bioavailable fractions ( E X + E R ) , on the other hand, contributed a relatively low percentage (11-24%). This might imply that Z n was mainly in the ionic and weakly complexed bioavailable fractions. A decrease o f R E - Z n  115  concentration following the hydrological gradient was observed during storm runoff, in which the R E - Z n was 119 ug/1 in the street runoff, decreased to 46 pg/1 in Still Creek and to 17 pg/1 in the Brunette River. The concentration o f bioavailable Z n in baseflow was relatively low, and there was no significant difference between up and down stream reaches. However, a high level o f R E - Z n (79 ug/1) was still measured in Still Creek during the dry season.  >  In stream trace metal uptake by the receiver resin was highly variable during storm flow. The higher uptake rates were observed in the early part o f the rainfall season (October). The analysis showed that major control factors seemed to be the rainfall intensity and the antecedent dry period that determined the accumulation o f C u and Z n and their potential bioavailability to the receiving water organisms. N o significant difference o f resin uptake rate was found under stormflow and snowmelt conditions, which is attributable to the rapid snow melt that is characteristic o f our mild, marine coastal climate. Considerable in situ resin uptake o f trace metals was also found under baseflow conditions. L o w dissolved oxygen in the sediment trap and in the deeper water o f Still Creek probably caused the release o f ionic metals from sediments to the water column, which then increased the bioavailable trace metals species to the water under baseflow conditions. Membrane uptake rate showed an opposite pattern to the resin uptake rate, i.e., the higher uptake rate by membrane, the lower uptake rate by resin. The oxidation o f iron and manganese on the dialysis tubing probably facilitated adsorption o f other metals to the tubing.  >  Trace metal concentration obviously decreased along the urban hydrological gradient. The highest concentrations o f C u and Z n were found in the street runoff - the major nonpoint source o f the trace metals. The concentrations o f trace metals were found to be highly correlated to the suspended sediment in surface runoff, indicating the particleassociated characteristic o f trace metals from the wash-off. C u and Z n i n the upstream section o f the watershed (Still Creek) showed significantly higher concentration than downstream (Brunette River). This difference seemed to be caused by the hydrologic  116  feature o f the watershed, such as the low flow in Still Creek and sedimentation in Burnaby Lake in the middle o f the watershed between the two streams.  >  The effect o f manganese on regulating C u and Z n in the Brunette River water system was not clearly demonstrated in this study, as the concentration o f M n was not well correlated to those o f the Mn-bound C u and Z n . However, considerable particulate E R - C u and E R Z n and high concentration o f manganese  were observed in urban runoff. Since  manganese is well recognized as the trace metal scavenger, it is reasonable to believe that the regulating effect o f M n on C u and Z n does exist in the Brunette River water system, but a more detailed, focused study is needed to elucidate this relationship.  >  Assessment o f C u and Z n measured in the watershed with the relevant water quality and toxicity criteria indicates that both these trace metals can have potential toxic effects on the organisms in the Brunette River watershed, especially in street runoff and upstream in Still Creek. Comparisons to W Q criteria also demonstrated that copper seemed to have a higher toxic effect to the aquatic life than zinc in this watershed since the criteria for copper toxicity is much lower.  5.2 Methods suitability  5.2.1 Water sampling In this study, the grab sample was used to measure bioavailable metal species in the urban watershed. The grab samples were normally taken in the early stage o f a rainfall event, which only represents a snap shot o f that rain period but not really the bioavailable trace metal species in the whole rainfall period. Intensive sampling needs to be done over the whole rain event to measure the mean value and provide more meaningful data.  Filtration to collect suspended sediment in this study might cause problems. Since many filters were used to provide sufficient particulate material for the geochemical fraction, this might lead to contamination o f samples for analysis. Continuous flow through centrifugation  117  should be used in the future work for suspended sediment collection. This would provide enough particulate metal sample for the analysis, and filter blank contamination during the filtration and digestion processes can be avoided.  5.2.2 Receiver technique The study o f trace metals using Dialysis with Resin Receiver technique is thought to be a more direct and more meaningful method of measuring bioavailability. However, the variability o f resin and membrane uptake rates during the experiments was observed. This phenomenon reflects the intermittent and variable nature o f bioavailable metal discharges in the stormwater (Morrison 1987). It should be noted that trace metals assimilated by organisms is expressed as uptake rate in the unit o f pg / mm .h. The uptake rate or flux o f the 2  bioavailable metal species is closely correlated to the available metal species in the water, the dialysis membrane exposure area and the exposure time in water. The uptake rate therefore is an averaged value used to describe the available metal species that can penetrate cell under certain time period. It was found in this study that the exposure time is a very important factor that influences the uptake rate. A too long exposure time w i l l lead to the adsorption o f particulate metal and other substance on the membrane surface (especially during storm period), which may clog the pathway o f the metal species into the cell surface and lead to the non-linear uptake. Thus the optimal design o f the exposure time must be taken into account for the future receiver experiment, which should be correlated to the rainfall and the flow pattern in the urban receiving water.  5.3 Recommendation and further work  Field work (water sampling and receiver experiment) and sample analysis (water and suspended sediment speciation and receiver resin analysis) in this study provided meaningful data on the bioavailability o f copper and zinc and their distribution in the Brunette River watershed. This information, in conjunction with the water quality data, precipitation data and flow data in the watershed, provided a more detailed understanding o f trace metal contamination and their toxic effects in this typical urban watershed.  118  Future investigations o f the bioavailable trace metal should focus on a more detailed analysis of suspended sediment both spatially and temporally, to provide a better understanding o f the bioavailable trace metal in the particulate phases. More baseflow water analysis needs to be done to verify the bioavailable trace metals in watershed in non-rainfall period, since the limited data in this study only provides preliminary conclusions. The water quality measurements need to be conducted at the same time as the water sampling and receiver resin experiments, so that a more accurate explanation o f the existing bioavailable trace metal species in the Brunette River watershed can be obtained. Comparative studies with in situ receiver resins and bioassay organisms need to be conducted to firm up the use o f this technique as a bioavailability assessment tool.  119  References Allard, B., Hakansson, K . , and Karlsson, S. (1987). The importance o f sorption phenomena in relation to trace element speciation and mobility. In: Speciation of Metals in Water, Sediment and Soil System, Landner, L., (ed.), Springer-Verlag, 99. Allen, H . E . and Hall, R . H . , and Brissbin, T.D. (1980). Metal speciation effects on aquatic toxicity. Environ. Sci. Technol, 14, 441. Alloway, B.J., and Ayres, D . C . (1993). Chemical principles o f environmental pollution. Published by Blackie Academic and professional, an imprint of Chapman & Hall. 131 pp A P H A (American Public Health Association). (1981). examination o f water and wastewater. 15 edition. 90.  Standard  methods  for  the  th  Atlins, E . D . , and Hawley, J.R. (1978). Sources o f metals and metals levels in municipal wastewater. Pollution Control Branch, Ontario Ministry of Environment, Toronto, Project #75-1-43. Anderson, R . V . (1977). Concentration o f cadmium, copper, lead and zinc in six species of freshwater clams. Bull. Envrion. Contam. Toxicol, 18, 492. Bedient, P.B., Lambert, J . L . , and Springer, N. K . (1980). Stormwater pollution load runoff relationships. Journal of the Water Pollution Control Federation, 52, 2396. Bendell -Young, L.I., Dutton, M . , and Pick, F . R. (1992). Contrasting two methods for determining trace metal partitioning in oxidized lake sediments. Biogeochemistry, 17, 205. Bendell -Young, L.I., and Harvey H . H . (1994). Geochemistry o f M n and Fe in lake sediments in relation to lake acidity. Limnol. Oceanogr., 37, 603. Bindra, K.S. (1983). Mobilization o f selected trace metals i n the aquatic environment (Sediment to water column and benthic invertebrates). Ph. D Thesis, Dept. of Civil Eng., University of British Columbia, Canada. Bindra, K.S., and Hall, K . (1977). Geochemical partitioning o f trace metals in sediments and factors affecting bioaccumulation in bethic organisms. Draft research paper, Dept. of Civil Eng., University of British Columbia, Canada. Borgmann , U . (1983). Metal speciation and toxicity o f free metal ions to aquatic biota. Adv. Environ. Sci. and Tech., 13, 47. Bratley, R.W., and Sprague, J.B. (1985a). The influence o f p H , water hardness, and alkalinity on the acute lethality o f zinc to rainbow trout (Salmo gairdneri). Can. J. Fish, Aquat. Sci., 42, 731.  120  Bratley, R.W., and Sprague, J.B. (1985b). Accumulation o f zinc by rainbow trout as influenced by p H . Can. J. Fish, Aquat. Sci., 42, 825. Brown, V . M . (1968). The calculation o f the acute toxicity o f mixture o f poisons to rainbow trout. Water Res. 2, 723. Bryan, G . W . (1971). The effects o f heavy metals (other than mercury) on marine and estuarine organisms. Proceedings of the Royal Society, London (Series B) 177, 389. Chapnick, S.D., Willard, S.M., and Kenneth, H . N . (1982). Microbially manganese oxidation i n a freshwater lake. Limnol. Oceanography, 27, 1004-14.  mediated  Chen, K . Y . , Gupta, S.K., Sycip, A . Z . , L u , J.C.S, Knezevic, M . , and Choi, W. (1976). Research study on the effects o f dispersion, settling and resedimentation on migration of chemical constituents during open-water disposal o f dredged materials. Dredged material research program contract report D-76-1, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Miss., 211 pp. Cole, R . H . , Frederick, R . E . , Healy, R. P., and Rolan, R . G . (1984). Preliminary findings o f the priority pollutant monitoring project o f the nationwide urban runoff program. Journal of the Water Pollution Control Federation. 56, 898. Collins, Y., and Stotzky, G . (1989). Factors affecting the toxicity o f heavy metals to microbes. In: Metal Ions and Bacteria, T.J. Beveridge and R.J. Doyle (eds.), New York, John Wiley & Sons, 31. Colston, N.V. (1974). Characterization and treatment o f urban land runoff. USA, EPA, Cincinnati, OH, EPA-670/2-74-096. Cox, M . F . , Holm, H . M . , Kania, H . J . , and Knight, R . L . (1975). Methylmercury and total mercury concentrations in selected stream biota. In: Trace Substances in Environmental Health, Hempill, D.D., (ed.), 9, 1975, 151. Davies, P.H., Goettl, J.P. Sinley, J.R., and Smith, N.F. (1976). Acute and chronic toxicity of lead to rainbow trout, Salmo gairdneri in hard and soft water. Water Res., 10, 199. Dawson, L . , Falherty, M . , and Gang, M . (1985). Still Creek Inventory. Part I RupertCassiar Neighbourhood Association. Ethyl. (1997). Charaterization o f manganese particulates from vehicles using M M T fuel - A compilation o f results from studies. Ethyl Corp., Richmond Va., USA. Flemming, C . A . , and Trevor, J . T . (1989). Copper toxicity environment: a review. Water Air Soil Pollu., 44, 143.  121  and chemistry in the  Florence, T.M., and Batley, G.E. (1980). C R C Critical reviews. In: Analytical 219.  Chemistry  Florence, T.M., and Batley, G.E. (1976). Trace metal species in seawater-I, Removal of trace metals from seawater by a chelating resin. Talanta, 23, 179. Forester, A.J. (1980). Monitoring the bioavailability o f toxic metals in acid stressed shield lakes using pelecypod molluscs (clams, mussels). In: Trace Substances in Environmental Health, Vol. XIV, D.D. Hemphill (ed.). Proceedings University of Missori's 14 annual conference on Trace Substances in Environmental Health, June 1980, 142. th  Fowler, S.W., Oregioni, B., La Rosa, J., and Renfro, W.C. (1974). Trace element concentrations in mussels o f the Ligurian Sea. Report IAEA-163, International Atomic Energy Agency, Vienna, 119 pp. Fowler, S.W., and Oregioni, B. (1976). Trace metals Mediterranean. Marine Pollution Bulletin, 7, 26.  in mussels  from the N W  Freedman, M.E., Cunningham, P.M., Schindler, J.E., and Zimmerman, M.J. (1980). Effect o f lead speciation on toxicity. Bull. Environ. Contam. Toxicol, 25, 389. Freiser, H. and Fernando, Q. (1963). Ionic equilibria in analytical chemistry. John Wiley & Sons, INC., 334pp Gadd, G.M., and Griffiths, A.J. (1978). Microorganisms and heavy metal toxicity. Microbial Ecology, 4, 303. Gibb, A., Bennet, B., and Birkbeck, A. (1991). Urban runoff quality and treatment: A comprehensive review. British Columbia Research Coporation, Vancouver, B.C. Canada. Gibbs, R.J. (1973). Mechanisms o f trace metal transport in rivers. Science, 180, 73. Guy, R.D., and Kean, A.R. (1980). Algae as a chemical speciation monitor I - A comparison o f algal growth and computer calculated speciation. Water Res., 14, 891. Guthrie, R.K., Davis, E.M., Cherry, D.S., and Murray, H.E. (1979). Biomagnification o f heavy metals by organisms in a marine microcosm. Bull. Envrion. Contam. Toxicol, 21, 53. Hall, K., and Anderson, B.C. (1987). The toxicity and chemical composition o f urban stormwater runoff. Can. J. Eng., 15, 98. Hall, K.J., Koch, F.A., and Yesaki, I. (1974). Further investigations into water quality conditions in the Lower Fraser River system. Technical report No. 4, Westwater Research Center, University of British Columbia, B.C., Canada.  122  Hall, J.K., Tesaki, I., and Chan, J. (1976). Trace metals and chlorinated hydrocarbons in the sediments o f a metropolitan watershed. Tech. Report No. 10, Westwater Research University of British Columbia, B.C., Canada.  Center,  Harrison, R.M., Laxen, D.P.H., and Wilson, S.J. (1981). Chemical associations o f lead cadmium, copper and zinc in street dusts and roadside soils. Environ. Sci. Tech., 15, 1378.  Harrison, R.M., and Wilson, S.J. (1983). " 2 International conference on heavy metals in nd  the environment". CEP Consultants Ltd, Edinburgh, 1, 684.  Harrison, G.I., Compbell, P.G.C., and Tessier, A. (1986). Effects o f p H changes on Z n uptake by Chlamydomonas variabilis grown in batch culture. Can. J. Fish. Aquat. Sci., 43, 687.  Hoffman,E.J., Latimer, J.S., Mills, G.L., and Quinn, J.G. (1982). Petroleum hydrocarbons in urban runoff from a commercial land use area. Journal of the Water Pollution Federation, 4, 1517.  Control  Hunter, J.V., Sabation, T., Gomperts, R., and Mckenzie, M.J. (1979). Contribution of urban runoff to hydrocarbon pollution. Journal of the Water Pollution Control Federation, 51, 2129. IRE. (1997). Intergrated watershed management. CD used for course in IRE, University of British Columbia, B.C., Canada.  watershed  management  Jarvis, B. (1992). Investigation o f stormwater runoff from highway catchment. Thesis, (Honour). Dept. Environ. Eng., University of Western Australia, Nedlands, W.A., Australia.  Kloke, A., Sauerbeck, D. R., and Vetter, H. (1984). Trace elements in soils and plants. CRC press, Boca Raton, FL.  Kushner, D.J. (1993). Effects o f speciation o f toxic metals on their biological activity. Water Poll. Res. J. Canada, 28, 111.  Luoma, S.N. (1983). Bioavailability of trace metals to aquatic organisms - a review. Sci. Total Environ., 28, 1.  Luoma, S.N., and Bryan, G.W. (1979). Trace metal bioavailability: modelling chemical and biologicl interactions o f sediment-bound zinc. In: Chemical Modeling in Aqueous Systems, Jenne, E.A. (ed.), A.C.S. Symp. Series 93, American Chemical Society, Washington, D.C., 577.  Luoma, S.N., Ho, Y.B., and Bryan, G.W. (1995). Fate, bioavailability and toxicity of silver in estuarine environments. Marine Pollution Bulletin, 31, 44.  123  Macdonald, R., Hall, K . J . , and Schreier, H . (1997). Water quality and stormwater contaminants in the Brunette River Watershed, British Columbia, 1994/1995. Research report, Westewater Research Center, University of British Columbia, Canada. Mance, G . , and O'Donnell, A.R. (1983). Application o f the European communities directive dangerous substances to list II: Substances in the U . K . Water Sci. Tech., 16, 159. Mattu S. G . , and Hall, J . K . (1999). The dynamics o f manganese from the gasoline additive M M T in the Brunette River watershed. CWRA conference in Richmond, B.C. Canada. Mattu G . (unpublished). The transport, geochemistry, and environmental fate o f manganese from M M T in the Brunette River Watershed. M.Sc. thesis, IRE, University of British Columbia, Canada. McCallum, D.W. (1995). A n examination of trace metal contamination and land use in an urban watershed. M. Sc. thesis, Dept. of Civil Eng., University of British Columbia, Canada. Miller, T . G . , and Mackay, W . C . (1980). The effects o f hardness, alkalinity and p H o f test water on the toxicity o f copper to rainbow trout (Salmo gairdneri). Water Res., 14, 129. Moore, J.W., and Ramamoorthy, S. (1984). Heavy metals in natural waters: Applied monitoring and impact assessment. New York: Springer-Verlag, 268pp. Morrison, G.M.P. (1985). Metal speciation in urban runoff. Ph.D University of Technology, U.K.  Thesis,  Chalmers  Morrison, G.M.P., and Revitt, D . M . (1987). Assessment o f metal species bioavailability and geochemical mobility in polluted waters. Environ. Tech. Letters, 8, 361. Morrison, G.M.P. (1987). Bioavailable metal uptake rate determination in polluted waters by dialysis with receiving resins. Environ. Tech. Letters, 8, 393. Morrison, G.M.P. (1989). Trace element speciation and its relationship to bioavailability and toxicity in natural waters. In: Trace Element Speciation, Chapter 2, Analytical Methods and Problems. CRC, Graeme E. Batley (ed.), 350pp Morrison, G.M.P., Revitt, D . M . , and Ellis, J.B. (1990). Metal speciation i n separate stormwater systems. Wat. Sci. Tech., 22, 53. Morgan, J . , and Stumm, W . (1964). Colloid chemical properties o f manganese dioxide. J. Colloid Interface Sci., 19, 347. Murphy, C.B., and Carleo, P.J. (1978). The contribution o f mercury and chlorinated organics from urban runoff. Water Research, 12, 531.  124  Namminga, H . , and Wilhm, J . (1977). Heavy metals in water, sediments and chironomids. Journal of the Water Polluion Control Federation, 53, 993. Petersen, R. (1982). Influence o f copper and zinc on the growth o f a freshwater algae, Sccendesmus quadricauda: the significance o f chemical speciation. Environ. Sci. Technol, 16, 443. Petersen, H . G . , Heakey, F.P., and Wagemann, R. (1984). Metal toxicity to algae: a highly p H dependent phenomenon. Can, J. Fish. Aquat. Sci., 41, 974. Playle, R . C . , Gensemer, R.W., and Dixon, D . G . (1992). Copper accumulation on gills o f fathead minnows: influence o f water hardness, complexation and p H o f the gill microenvironment. Environ. Toxicol. Chem., 11, 381. Popham, J.D., Johnson, D.C., and D'Auria, J . M . (1980). Mussels (Mytilus edulis) as 'point source' indicators o f trace metal pollution. Marine Pollution Bulletin, 11, 261. Rand, G . M . and Petrocelli, SJ*. (1984). Chapter 1 in Fundamentals o f Aquatic Toxicology, Hemisphere Publishing Corp. Rattner, B.A., and Heath, A . G . (1995). Environmental factors affecting contaminant toxicity in aquatic and terrestrial vertebrates. Chapter 22 in Handbook of Ecotoxicology. Hoffman, D.J. et al. (ed.), CRC Press, Inc., Lewis publishers. Ray, S., McLeese, D.W., and Pezzack, D. (1979). Chelation and inter-elemental effects on the bioaccumulation o f heavy metals by marine invertebrates. International Confrence Proceedings on Management and Control of Heavy Metals in the Environment, Imperial College, London, 35. Revitt, D . M . , and Morrison, G.M.P. (1987). Metal speciation variations within separate stormwater systems. Environ. Tech. Letters, 8, 373. Regnier, P., and Wollast, R. (1993). Distribution o f trace metals in suspended matter of the Scheldt estuary. Mar. Chem., 43, 3. Rudolph, E . (1995). Personal communication. Sapperton Westminster, B.C., Canada.  Fish and Game Club,  New  Sartor, J.D., and Boyd, G.B. (1972). Water pollution aspects o f street surface contaminants. U.S. Environmental Protection Agency Report, EPA-R2-72-081, U.S. Government Printing Office, Washington, D.C. Schreier, H . , Brown S., and Hall K . (1991). The land-water interface in the Fraser River Basin. In: Water in Sustainable Development, Dorcey, A. and J. Griggs (eds.), Westwater Research Center, University of British Columbia, B.C., Canada.  125  Singh, S., and Subramanian, V. (1984). Hydrous Fe and M n oxides - Scavengers o f heavy metals in the aquatic environment. Critical Review in Environmental Control, 14, 33. Shuman, L.M. (1977). Adsorpton o f Z n by Fe and A l hydrous oxides as influenced by aging and p H . Soil Sci. Am. J., 41, 703.  Sonzogni, W.C., Chesters, G., Coote, D.R., and Jeffs, D.N. (1980). "Pollution from land r u n o f f . Environ. Sci. Technol,  14, 148.  Spacie, A., and Hamelink, J.L. (1984). Bioaccumulation. Chapter 17 in Fundamentals Aquatic Toxicology, Rand, G.M. andPetrocelli, S.R.(ed.), Hemisphere Publishing Corp.  of  Solo-Gabriele, H.M., and Perkins, F.E. (1997). Watershed-specific model for streamflow, sediment, and metal transport. J. Environ. Eng., 123, 61.  Stecko J.R.P., and Bendell-Young, L.I. (1999). Contrasting the geochemistry o f suspended particulate matter and deposited sediments within an estuary. Applied Geochemistry, 1.  Stumm, W., and Lee, G.F. (1960). The chemistry o f aquatic iron. Schweiz. Z. Hydro!., 21, 295.  Stumm, W., and Morgan, J.J. (1970). Aquatic chemistry. Wiley Inter-Science, N.Y., 583pp. Sunda, W.G., and Kleber, D.J. (1994). Oxidation o f humic substances by manganese oxides yields low- molecular weight organic substances. Nature, 367, 62. Swain, L.G. (1989). Water quality assessment and objectives. Technical Appendix, Coquitlam-Pitt River area, tributaries to the lower Fraser river along the north shore, BC Ministry of Environment, Province of British Columbia.  Tessier, A., Campbell, P. G., and Bisson, M. (1979). Sequential extraction procedure for the speciation o f particulate trace metals. Analytical Chem., 51, 844.  Theis, T. L., and Dodge, E.E. (1979). Effect o f chemical speciation on the uptake o f copper by Chironomous tentans. Environ. Sci. Technol,  13, 1287.  Thomas, C A . (1997). Linking sediment geochemistry in the Fraser River intertidal region to metal accumulation i n Macima Balthica. M.Sc Thesis, Dept. Biological Sciences, Simon Fraser University, B.C., Canada.  Thomas, C.A., and Bendell-Young, L.I. (1998). Linking the sediment geochemistry o f an intertidal region to metal bioavailability i n the deposit feeder Macoma Ecology Progress Series, 173, 197.  balthica.  Tinsley, I.J. (1979). Chemical concepts  Wiley-Interscience  Publication,  John Wiley & Sons, 96.  126  i n pollutant  behavior.  Marine  Turner, D.R. (1983). Relationships between biological availability and chemical measurement. In: Metal ions in biological systems. Sigel, H (ed.), 18, 137, Marcel Dekker, New York. Viarengo, A . (1985). Biochemical effect o f trace metals. Mar. Pollut. Bull, 16, 153. Wanielista, M.P., Yousef, Y . A . , and Mclellon, W . L . (1977). Nonpoint source effects o f water quality. Journal of the Water Pollution Control Federation, 49, 441. Watling. H.R., and Watling, R.J. (1976). Trace metals in oysters from Knysna Estuary. Marine Pollution Bulletin, 7, 45. Wren, C D . , Harris, S., and Harttrup, N.A. (1995). Ecotoxicology o f mercury and cadmium. Section II, 17, in Handbook of Ecotoxicology, Hoffman, D.J. et al. (eds.), 32 pp Whipple, W., Hunter, J.V., and Y u , S.L. (1974). Unrecorded pollution from urban runoff. Journal of the Water pollution Control Federation, 46, 873. Wilber, W . G . , and Hunter, J.V. (1979). Aquatic transport o f heavy metals in the urban environment. Water Resources Bulletin, 13, 721. Winner, R.W. (1985). Bioaccumulation and toxicity o f copper as affected by interactions between humic acid and water hardness. Water Research. 19, 449. Yuan, Y . , Hall K . J . , and Oldham, C . (1999). A preliminary model for predicting heavy metal contaminant loading from an urban catchment. Eighth International Symposium: The Interactions between Sediment and Water. Sept. 13-17, 1999 Beijing, China. Zevenhuisen ,L.P., Dolling, J . , Eshuis, E . J . , and Scholten-Koersehman, I.J. (1979). Inhibitory effects o f copper on bacteria related to the free ion concentration. Microbiol. Ecol, 5, 139.  127  Appendix Appendix A  Trace metal speciation in the water samples  Table A - 1 Speciation for baseflow water Sampli ng data  Metal (ug/L)  Cu  Zn Aug. 17, 1998  Mn  Fe  Samplin g site  Total Metal (TM)  Total D i ssolved (Tl 3) RE STR-B  #1 #2 #3 #1 #2 #3 #1 #2 #3 #1 #2 #3  3.9 0.8 0.2 44.0 79.0 16.0 50.0 53.0 10.0 3.8 0.3 0.3  Total Particulate (TP) EX ER RES 3.9 0.8 0.2 1.1 2.5 4.0 3.0 4.0 21.0 8.0 8.0 7.0  1.9 1.7 1.1 1.2 3.6 3.0 2.0 4.0 57.0 20.0 18.0 26.0  2.4 2.3 2.1 9.6 11.0 13.0 3.0 4.0 29.0 255.0 208.0 332.0  Table A - 2 Speciation for stormwater (1) Sampli ng data  Metal (ug/L)  Samplin g site  Total Metal (TM)  Cu  #1 #2  35.0 28.0 7.0 176.0 71.0 44.0 78.0 141.0 64.0 10563.0 7880.0 777.0  Zn Nov. 19, 1999  Mn  Fe  #3 #1 #2 #3 #1 #2 #3 #1 #2 #3  Total D issoved (Ti D) RE STR-B 5.0 5.0 2.0 76.0 40.0 12.0 16.0 43.0 9.0 10.0 13.0 11.0  128  10.0 18.0 4.0 32.0 27.0 8.0 3.0 55.0 8.0 297.0 889.0 297.0  EX  Total Particulate (TP) ER RES  2.7 0.0 0.0 18.0 0.0 0.0 12.0 0.0 0.0 45.0 0.0 0.0  7.5 1.9 1.0 24.0 8.0 10.0 15.0 31.0 43.0 3700.0 890.0 134.0  8.9 0.7 0.7 30.0 15.0 23.0 28.0 4.0 8.0 7500.0 1610.0 268.0  Table A - 3 Speciation for stormwater (2) Sampling data  Metal  (Mg/L)  Cu  Zn  Jan. 3, 2000  Mn  Fe  Total D issoved (Tl RE STR-B  Total Particulate (TP) EX ER RES 6.8  Samplin g site  Total Metal (TM)  #1 #2 #3 #1  67.2 29.8 15.5 373.8  4.4 3.1 0.6 219.6  12.8 21.9 14.8 46.7  #2 #3 #1 #2 #3 #1 #2 #3  132.8 39.1 187.2 76.8 78.4 9980.0 1320.0 1330.0  45.5 22.3 51.4 33.0 27.0 920.0 5.0 10.0  31.4 7.2 15.7 7.2 12.7 810.0 75.0 290.0  25.0 —  13.0  27.0  10.5 3.1 0.0 66.0  27.3 4.0 1.9 73.0  --  —  —  —  5.0 15.0 19.0 79.0 95.0 75.0  30.0 7.0 5.0 2731.0 395.0 320.0  Table A - 4 Speciation for stormwater (3) Sampling data  Metal  (Mg/L)  Cu  Zn Jan, 10 2000  Mn  Fe  Sampling Site  Total Metal (TM)  #1 #2 #3 #1 #2 #3 #1 #2 #3 #1 #2 #3  85.1 28.8 19.8 364.7 130.3 38.4 241.4 88.0 45.0 11950.0 960.0 380.0  Total D issoved (Tl D) RE STR-B  Total Particulate (TP) EX ER RES  7.8 2.5 0.0 104.4 57.8 21.6 41.0 59.5 30.2 130.0 10.0 20.0  12.0  129  20.0 21.4 19.6 74.5 34.7 19.1 23.4 18.1 10.7 350.0 280.0 200.0  —  —  37.0 0.0 0.0 22.0 —  ~  47.0 —  —  15.0 4.0 0.0 65.0 15.0 5.0 21.0 5.0 4.0 274.0 176.0 69.0  33.0 5.5 2.0 88.0 21.0 7.0 83.0 10.0 3.0 7727.0 691.0 268.0  Appendix B  Metal speciation in the suspended sediment (SS)  Table B - 1 C u Speciation in SS  Aug#l Aug#2 Aug#3 Nov#la Nov#lb Nov#2 Nov#3 3,Jan#l 3,Jan#2 3,Jan#3 10,Jan#l 10,Jan#2 10,Jan#3  EX 72.3 107.1 89.3 29.2 17.0 0.0 0.0 45.0  15.8  mg / K g S S ER 57.4 78.0 52.2 79.7 29.5 249.8 182.3 69.3 145.1 0.0 19.8 140.5 0.0  RES 72.6 104.1 104.6 94.7 107.7 91.6 117.4 180.5 188.1 161.6 43.5 193.1 196.3  EX 2.4 2.4 1.8 2.7 6.5  6.8  12.0  ug/L ER 1.9 1.7 1.1 7.5 11.3 1.9 1.0 10.5 3.1 0.0 15.0 4.0 0.0  RES 2.4 2.3 2.1 8.9 41.2 0.7 0.7 27.3 4.0 1.9 33.0 5.5 2.0  Table B - 2 Z n Speciation in SS  Aug#l Aug#2 Aug#3 Nov#la Nov#lb Nov#2 Nov#3 3,Jan#l 10,Jan#l 10,Jan#2 10,Jan#3  EX 31.0 117.0 199.0 245.0 199.0 0.0 0.0 164.0 52.0 0.0 0.0  mg / K g S S ER 34.0 168.0 149.0 327.0 178.0 1043.0 1746.0 436.0 92.0 562.0 482.0  RES 273.0 514.0 647.0 409.0 314.0 1955.0 4016.0 480.0 124.0 800.0 770.0  EX 1.1 2.5 4.0 18.0 76.0  25.0 37.0 0.0 0.0  130  ug/L ER 1.2 3.6 3.0 24.0 68.0 8.0 10.0 66.0 65.0 15.0 5.0  RES 9.6 11.0 13.0 30.0 120.0 15.0 23.0 73.0 88.0 21.0 7.0  Table B - 3 M n Speciation in SS  Aug#l Aug#2 Aug#3 Nov#la Nov#lb Nov#2 Nov#3 3,Jan#l 3,Jan#2 3,Jan#3 10,Jan#l 10,Jan#2 10,Jan#3  EX 99.0 160.0 1053.0 163.0 73.0 0.0 0.0 84.0  30.0  mg/KgSS ER 47.0 182.0 2856.0 204.0 118.0 4040.0 7508.0 34.0 686.0 1575.0 28.0 174.0 437.0  RES 102.0 175.0 1434.0 381.0 248.0 521.0 1397.0 197.0 328.0 426.0 109.0 341.0 309.0  EX 3.0 4.0 21.0 12.0 28.0 0.0  RES 7.6 9.3 16.5 102.2 94.4 209.8 46.8 18.1 18.6 27.0 10.2 24.5 26.3  EX 8.0 8.0 7.0 45.0 230.0 0.0  13.0  22.0  ug/L ER 2.0 4.0 57.0 15.0 45.0 31.0 43.0 5.0 15.0 19.0 21.0 5.0 4.0  RES 3.0 4.0 29.0 28.0 95.0 4.0 8.0 30.0 7.0 5.0 83.0 10.0 3.0  ug/L ER 20.0 18.0 26.0 3700.0 1200.0 890.0 134.0 79.0 95.0 75.0 274.0 176.0 69.0  RES 255.0 208.0 332.0 7500.0 36140.0 1610.0 268.0 2731.0 395.0 320.0 7727.0 691.0 268.0  Table B - 4 Fe Speciation in SS  Aug#l Aug#2 Aug#3 Nov#la Nov#lb Nov#2 Nov#3 3,Jan#l 3,Jan#2 3,Jan#3 10,Jan#l 10,Jan#2 10,Jan#3  EX 0.2 0.3 0.4 0.6 0.6  0.2  0.1  g/KgSS ER 0.6 0.8 1.3 50.4 3.1 116.0 23.4 0.5 4.5 6.3 0.4 6.2 6.8  27.0  47.0  131  Appendix C Stormwater Quality (SQ) at Intervals over the Storm Event (Nov. 19, 1999) Table C - 1 Stormwater quality data at Site #1 SS  Mn (ug/L)  Zn (pg/L)  Cu (ug/L)  Fe (ug/L)  Time  pH  (mg/L)  TM  TD  RE  TM  TD  RE  TM  TD  RE  TM  TD  RE  9:10  6.5  89  33  20  9  151  89  75  62  18  12  13154  179  3  9:55  6.4  31  40  12  6  234  108  64  106  25  25  25523  465  5  10:40  6.5  35  31  10  2  161  135  91  71  20  15  17449  401  15  11:25  6.5  73  36  17  3  158  100  74  74  14  13  17528  181  15  12:05  6.4  223  60  14  1  104  98  51  47  14  9  10991  272  12  12:45  6.6  948  113  5  0  613  78  57  345  19  13  65952  163  8  Table C - 2 Stormwater quality data at Site #2 SS  Cu  (Ug/L)  Mn (pg/L)  Zn (ug/L)  Fe (ug/L)  Time  pH  (mg/L)  TM  TD  RE  TM  TD  RE  TM  TD  RE  TM  TD  RE  9:25  6.4  3  13  11  2  66  62  34  145  136  92  7830  949  6  10:05  6.5  1  14  11  2  65  55  48  155  93  45  8468  972  7  10:50  6.6  2  10  8  5  60  60  57  116  106  15  7629  882  25  11:30  6.6  5  75  60  9  94  92  22  149  58  2  7591  804  14  12:15  6.7  20  12  11  2  121  87  33  138  91  68  11464  598  17  13:00  6.6  18  17  14  3  126  47  48  119  67  44  10319  444  9  Table C - 3 Stormwater quality data at Site #3 SS  Cu  Zn (pg/L)  (pg/L)  Mn (ug/L)  Fe (pg/L)  Time  pH  (mg/L)  TM  TD  RE  TM  TD  RE  TM  TD  RE  TM  TD  RE  9:40  6.8  5  9  8  2  34  31  13  57  18  17  702  273  6  10:25  6.7  4  3  3  1  41  19  15  65  13  9  812  359  8  11:05  6.7  5  7  6  3  48  18  10  68  11  2  810  278  9  11:50  6.6  7  10  9  3  53  13  11  64  22  6  783  321  20  12:30  6.6  7  8  7  3  55  22  21  72  24  21  875  322  16  13:20  6.7  10  7  7  2  65  53  50  98  18  16  998  376  25  132  Appendix D  Chelex receiver uptake results  Table D - 1 Results in July (7.23 - 7. 30V 1998 D-1A  Resin uptake  Metal  Site No.  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  2  2  Cu  #1 #2 #3  2.89 5.85 5.36  1.41-1.38 5.36-7.33 3.38-7.33  4 4 4  Zn  #1 #2 #3  19 25 20  13-27 22-27 13-17  4 4 4  Mn  #1 #2 #3  265 24 16  186-349 17-33 7-26  4 4 4  Fe  #1 #2 #3  6962 98 65  5590-8506 14-289 19-108  4 4 4  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  #1 #2 #3  6.67 5.85 4.86  5.36-7.33 3.38-9.30 3.38-7.33  4 4 4  #1 #2  14 14  #3  15  11-16 7-27 9-25  4 4 4  Mn  #1 #2 #3  3 58 486  1-6 39-82 259-814  4 4 4  Fe  #1 #2 #3  398 355 1234  308-512 174-617 607-2123  4 4 4  D-1B  Membrane uptake  Metal  Site No.  2  Cu  Zn  133  2  Table D - 2. Results i n August (8.17 - 9.2). 1998 D-2A  Resin uptake  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  Cu  #1 #2 #3  1.95 5.41 4.85  0.62-3.28 5.05-5.77 4.78-4.91  2 2 2  Zn  #1 #2 #3  9 16 5  5-13 10-22 4-5  2 2 2  Mn  #1 #2 #3  124 35 15  79-168 13-58 12-18  2 2 2  Fe  #1 #2 #3  521 56 20  379-664 31-80 14-27  2 2 2  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  Metal  Site No.  2  D-2B  2  Membrane uptake  Metal  Site No.  2  2  Cu  #1 #2 #3  5.90 2.36 2.73  4.98-6.81 2.33-2.40 2.27-3.19  2 2 2  Zn  #1 #2 #3  30 6 9  23-37 6-7 6-12  2 2 2  Mn  #1 #2 #3  7 29 344  7-8 25-34 304-385  2 2 2  Fe  #1 #2 #3  300 155 584  255-344 145-166 473-695  2 2 2  134  Table D - 3. Results i n Febrary (2.11 - 2.18). 1999  D-3A  Resin uptake  Metal  Site No.  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  2  2  Cu  #1 #2 #3  4.73 3.57 2.66  0.79-8.75 0.74-7.65 0.00-7.25  4 8 6  Zn  #1 #2 #3  21 38 23  18-25 17-62 20-26  4 8 6  Mn  #1 #2 #3  17 88 73  15-18 68-109 65-80  4 8 6  Fe  #1 #2 #3  36 105 14  25-50 59-174 9-32  4 8 6  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  Cu  #1 #2 #3  8.98 33.31 0.00  8.98-8.98 16.98-22.23 0  4 8 6  Zn  #1 #2 #3  47 94 2  41-52 70-125 0-6  4 8 6  Mn  #1 #2 #3  26 147 10  24-28 111-200 4-23  4 8 6  Fe  #1 #2 #3  870 1893 294  820-919 1893-2493 148-573  4 8 6  D-3B  Membrane uptake  Metal  Site No.  2  135  2  Table D - 4. Results i n October (T0.27 - 11.3V 1999 D-4A  Resin uptake  Metal  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  #1 #2 #3  18.00  9.00-25.00  15.00  9.00-16.00  6 6 6  #1 #2 #3  115  89-164  49  29-64  #1 #2 #3  35  29-40  60  45-71  #1 #2 #3  250  226-290  Site No.  2  Cu  Zn  Mn  Fe  D-4B  —  —  —  —  —  —  Mn  Fe  6 6 6  19-194  6 6 6  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  #1 #2  0.45  0.00-1.37  #3  1.25  0.00-2.42  6 6 6  #1 #2 #3  7  3-9  13  8-20  1  0-1  86  60-108  157  71-249  785  548-1064  —  —  83  Site No.  2  Zn  6 6 6  Membrane uptake  Metal  Cu  2  #1 #2 #3 #1 #2 #3  136  2  6 6 6 6 6 6 6 6 6  Table D - 5 . Results in November T i l . 1 9 - 11.25V 1999 D-5A  Resin uptake  Metal  Site No.  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  2  2  Cu  #1 #2 #3  4.80 3.80 3.00  4.20-5.30 1. 70-6.20 2.60-3.80  4 4 4  Zn  #1 #2 #3  22 21 13  17-29 17-27 1-19  4 4 4  Mn  #1 #2 #3  8 51 73  6-11 43-66 60-96  4 4 4  Fe  #1 #2 #3  16 141 45  14-19 92-184 22-73  4 4 4  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  D - 5 B Membrane uptake  Metal  Site No.  2  2  Cu  #1 #2 #3  30.00 25.00 1.00  18.00-40.00 12.00-37.00 0.00-1.00  4 4 4  Zn  #1 #2 #3  134 78 8  93-194 38-118 7-8  4 4 4  Mn  #1 #2 #3  69 142 65  55-95 92-191 53-76  4 4 4  Fe  #1 #2 #3  3233 4158 784  2544-4096 2214-5841 740-884  4 4 4  137  Table D - 6. Results i n November (11.19V 1999 (One storm) D-6A  Resin uptake  Metal  Site No.  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  2  2  Cu  #1 #2 #3  6.80 9.60 11.40  2.70-10.50 7.40-12.60 9.80-13.60  4 4 4  Zn  #1 #2 #3  13 25 21  8-16 14-37 11-33  4 4 4  Mn  #1 #2 #3  24 194 174  22-27 176-213 133-213  4 4 4  Fe  #1 #2 #3  7 154 21  0-15 134-175 14-29  4 4 4  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  D-6B  Membrane uptake  Metal  Site No.  2  2  Cu  #1 #2 #3  0 0 0  0 0 0  4 4 4  Zn  #1 #2 #3  7 12 11  0-15 9-16 9-13  4 4 4  Mn  #1 #2 #3  2 12 28  0-6 6-19 25-33  4 4 4  Fe  #1 #2 #3  128 646 160  74-211 499-907 70-262  4 4 4  138  Table D - 7. Results in January (1.3- 1.10). 2000 D-7A  Resin uptake  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  Cu  #1 #2 #3  3.98 4.65 4.04  3.26-4.47 3.66-5.02 2.91-6.90  4 4 4  Zn  #1 #2 #3  42 23 21  40-44 22-24 20-24  4 4 4  Mn  #1 #2 #3  17 66 82  15-19 56-72 77-92  4 4 4  Fe  #1 #2 #3  51 107 16  41-72 84-142 12-21  4 4 4  Metal  Site No.  2  D-7B  2  Membrane uptake  Metal  Site No.  Mean (pg/mm .hr)  Range (pg/mm .hr)  No. of samples  2  2  Cu  #1 #2 #3  1.90 8.30 1.30  0.63-3.63 5.49-11.79 0.39-3.23  4 4 4  Zn  #1 #2 #3  4 20 2  1-9 15-23 2-3  4 4 4  Mn  #1 #2 #3  2 45 7  0-4 28-58 5-9  4 4 4  Fe  #1 #2 #3  80 1020 125  28-182 790-1217 107-133  4 4 4  139  Appendix E Resin evaluation for exchange capacity and Ca & Mg interference E.1  E x c h a n g e capacity test.  Chelex-100 chelating resin ( B I O - R A D ) was used in the Dialysis with Resin Receiver in the field experiments to determine trace metal avilability. When the receiver was contacted with the water, the bioavailable trace metal species in the water w i l l penetrate the dialysis membrane and complex with the Chelex resin. The exchange capacity test conducted herein is to ensure that the resin used inside the dialysis bag is suffucient for the available trace metals it w i l l be complexed when the receiver was kept in contact with trace metal polluted water for a certain time period (exposure time). The test procedure, the results and the conclusion were described below.  A . Procedure: Two grams o f chelex resin were used for the exchange capacity test for trace metal C u and Zn. The solution concentration for C u and Z n were prepared as below:  Metal solution  (1) C u ^ = 4 meq in 100 m l o f water, C c = 1270.8 mg C u ^ / L u  (2) Z n ^ = 4meq in 100 m l o f water,  C z = 1307.4 mg Zn++ / L n  The experiment was conducted in the following steps. Ten m l o f metal solution were exchanged with Chelex-100 resin (2 g) in a separate column o f 25 cm in length and 1.5 cm in diameter, and the solution after exchange was collected in a 50 m l volumetric flask. After each 10 m l solution exchange, 30 m l deionized water were added to the exchange column, and the water was continuous received into the flask. The flow rate was 0.5-1 ml/min. The metal solution in the volumetric flask was diluted to 50 m l and A A S analysis for C u or Z n concentration. The above procedure was repeated for each 10 m l o f trace metal solution to  140  get the exchange capacity curve. Each metal was tested twice to ensure the results were consistent.  B . Results The resin capacity curve for trace metal C u and Z n showing in Figure E - l below indicated that the exchange capacity o f the Chelex-100 resin reaches the saturation at approximately 70 ml o f metal solution passing through the column.  Resin  o O  e x c h a n g e  capacity  test  (Cu)  I  CN  0  20  40  60  Solution  volume  80  100  (mL)  (a) R e s i n e x c h a n g e c a p a c i t y test O  (Zn)  -i  CM  0  20  40  60  Solution volume  80  (mL)  (b) Figure E - 1 Resin exchange capacity test for C u (a) and Z n (b)  141  1 00  The calculation gives that the Chelex-100 resin exchange capacity for metal C u and Z n in Table E - l .  Table E - 1 Chelex resin exchange capacity  In the  Exchange capacity  Cu  Zn  mg(metal) / g(resin)  44.48  45.76  field  experiment,  the highest  metal uptake by Chelex resin was  7.9  mg/g  (Fe+Mn+Cu+Zn in July, 1998 experiment), which is approximately 6 times lower than the resin exchange capacity.  C. Conclusion The use o f 1 gram o f resin in the receiver is sufficient for the field experiment to test trace metal bioavailability over the exposure periods studied in the Brunette River watershed.  142  D . Experiment data  Table E - 2 Resin exchange capacity test for C u solution in 2.3, 1999  Vol. Solution (ml) 10 m l 20ml 30ml 40ml 50ml 60ml 70ml 80ml 90ml 100ml  Cu concentration for resin excha nge capacity test at 2.3,1999 (solution: Cu only, res n are all old resin) abs(nm) C (mg/1) abs(nm) C (mg/1)  nl n2 n3 n4 n5 n6  n7 n8 n9  0.00 0.00 0.00 0.03 0.08 . 0.52 0.73 0.73 0.73  -0.13 -0.13 -0.13 1.71 5.40 57.30 173.81 173.81 173.81  ol o2 o3 o4 o5 06  o7 08  o9 olO  0.00 0.00 0.00 0.01 0.05 0.57 0.73 0.73 0.74 0.74  -0.13 -0.13 -0.06 0.38 3.70 65.05 159.80 176.19 176.19 176.19  Table E - 3 Resin exchange capacity test for Z n solution in 2.3, 1999 Zn concentration for resin exchange capacity test at 2.3,1999 (solution: Zn only, resin are all old resin) Vol.  abs(nm)  C (mg/t)  0.03 0.00 0.01 0.04 0.56 1.03 1.10 1.11 1.11  0.25 -0.12 -0.05 0.42 22.73 121.66 168.03 168.03 168.03  abs(nm)  C (mg/1)  0.00 0.00 0.01 0.02 0.49 1.05 1.10 1.11 1.11  -0.16 -0.16 0.00 0.12 18.86 133.26 171.60 171.60 171.60  Solution (ml) 10 m l 20ml 30ml 40ml 50ml 60ml 70ml 80ml 90ml  nl n2 n3 n4 n5 n6 n7 n8 n9  ol o2 o3 o4 o5 06  o7 08  o9  143  E.2 Interference test for Chelex resin With the consideration that calcium (Ca) and magnesium (Mg) are the most common ionic metals in the urban water, interference o f these metal species may affect the complexation o f resin with the target trace metals in the water. The interference o f C a and M g were tested in the laboratory to see that i f there w i l l be the interface o f these metals on the trace metal copper and zinc complexation with Chelex resin.  A . Procedure: Two grams o f chelex resin were used for the interference test for trace metal C u . Three solutions were prepared in this test. The first is prepared by C u , M g and C a , second consists of C a and C u only and the third is only made by C u for comparison. The constituents o f each solution were described below.  Metal solution  Solution 1: C u Ca  4 meq in 100 m l o f water, C c = 127.08 mg Cu* 1100 m l u  + i  Mg  5 meq in 100 m l of water, C c = 100.2 mg Ca** I100 m l a  +  2 meq in 100 m l o f water, C  M g  = 24.3 l m g Mg* I 100 m l  Solution 2: C u ^ = 4 meq in 100 m l o f water, C c = 127.08 mg C u / 1 0 0 m l + +  u  C a " ^ 10 meq in 100 m l o f water, C c = 200.4 mg Cz** I100 m l a  Solution 3: C u ^ = 4 meq in 100 m l o f water, C c = 127.08 mg Cu* 1100 m l u  The interference experiment was conducted in the following steps. Ten m l o f metal solution were eluted through a separate column (25 cm in length and 1.5 cm in diameter) containing Chelex-100 resin (2 g) with flow rate o f 0.5-1.0 ml/min. The eluant was collected in a 50 m l volumetric flask. After each 10 m l metal solution , 30 m l deionized water were added to the  144  exchange column, and the water was collected in the same volumetric flask. The metal solution in the volumetric flask then was diluted to 50 m L and A A S analysis for the concentration o f each element sample. Repeat above procedure for each 10 m l o f trace metal solution to get the curves for C a , C u and M g . The elution profiles were conducted in duplicate.  B.  Results  Figure E-2 and E-3 showed the interference o f C a and M g on the trace metal C u complexation with Chelex resin.  For the comparison purpose, the measurement o f C u  solution was also drawn in these two figures. It can be seen in the figures that the presence of Ca  and M g  + 2  + 2  affect the complexation between the chelex resin and metal C u . The elution  profiles o f the C u in the presence of C a and M g (Fig. E-2) or only C a (Fig. E-3) demonstrates different  complexation ability when compared  to the pure C u solution (The  distinct  breakthrough curve). In the early stage, C a and M g in the water may compete with C u for the resin complexation sites, showing in Figure E-2 that C a and M g curve increased dramatically after elution o f 20 m l solution through the exchange column. Based on this competition, the complexation o f C u in the resin w i l l be delayed. However, it is clear that the priority o f resin complexation was with C u , since C a or M g reached saturation on the resin quickly, while C u kept binding to the Chelex resin. Table E-4 (Freiser, H . and Fernando, Q. 1963) lists the relative stability constant o f C a , M g , C u + 2  + 2  + 2  and Z n  + 2  for the Chelex functional group,  which ensure that the priority o f the C u and Z n complexed with the Chelex resin than C a , + 2  Mg  + 2  in the water system.  Table E - 4 Formation constants of metal complexes o f Chelex resin functional group (EDTA)  Metal Ion  logk  Cu  + 2  18.8  Zn  + 2  16.5  Ca  + 2  10.7  Mg  8.7  + 2  145  Figure E - 2 The effect o f calcium and magnesium on the complexation o f copper on Chelex resin  Figure E - 3 The effect o f calcium on the complexation o f copper on Chelex resin  146  C. Conclusion C a and M g in the water column affect the initial complexation profile o f C u with the chelex resin. However, the experimental curve indicates that the chelex resin can continue to bind C u even when saturated with C a and M g . So interference from divalent cations such as C a and mg  + 2  should be small considering their relative concentrations in the field water samples  compared to the levels used in the laboratory experiment.  D . Experimental data (Average value o f two parallel experiments)  Table E - 5 Solution o f Cu+Ca+Mg (mg/L) Vol.of Cu+Ca+Mg solution (mL) 10  Cu(alone)  Cu(mg/L;  Ca(mg/L)  Mg(mg/L)  run-1  run-2  run-1  run-2  run-1  run-2  2  1  10  5  3  1  1  20  1  17  18  19  21  6  9  30  5  24  36  234  165  98  94  40  6  65  87  286  177  71  71  50  170  105  116  272  171  58  61  60  185  136  160  236  144  51  53  70  194  175  198  205  119  48  49  80  228  193  187  195  119  49  51  90  225  198  209  183  105  48  51  100  241  232  232  163  93  49  50  (mg/L)  147  Table E - 6 Solution of C u - C a (mg/L) Vol.ofCu+Ca solution (mL) 10  Cu(alone)  20  Cu(mg/LJ  Ca(mg/L)  run-1  run-2  run-1  run-2  2  1  7  13  1  30  51  163  316  30  5  43  99  214  503  40  6  70  124  513  505  50  170  98  160  505  491  60  185  144  182  456  462  70  194  169  204  434  442  80  228  179  205  379  445  90  225  193  222  393  428  100  241  210  238  397  444  (mg/L) 2  C u (only) solution for the comparison  148  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089693/manifest

Comment

Related Items