UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Toxicity of urban stormwater runoff Anderson, Bruce Campbell 1982

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1982_A7 A63.pdf [ 9.3MB ]
Metadata
JSON: 831-1.0062987.json
JSON-LD: 831-1.0062987-ld.json
RDF/XML (Pretty): 831-1.0062987-rdf.xml
RDF/JSON: 831-1.0062987-rdf.json
Turtle: 831-1.0062987-turtle.txt
N-Triples: 831-1.0062987-rdf-ntriples.txt
Original Record: 831-1.0062987-source.json
Full Text
831-1.0062987-fulltext.txt
Citation
831-1.0062987.ris

Full Text

TOXICITY OF URBAN STORMWATER RUNOFF by BRUCE CAMPBELL ANDERSON B.Sc, The University of Toronto, 1977 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE i n THE FACULTY OF GRADUATE STUDIES Department of C i v i l Engineering We accept t h i s thesis as conforming to the required standard •THE UNIVERSITY OF BRITISH COLUMBIA July 1982 © Bruce Campbell Anderson, 1982 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head o f my department or by h i s or her r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . Department Of C i v i l Engineering  The U n i v e r s i t y of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 July 14, 1982 - i i -ABSTRACT This work involves the study of the e f f e c t s of land use on the chemical composition of urban stormwater runoff, and i t s subsequent acute t o x i c i t y to the aquatic invertebrate Daphnia pulex. Samples were obtained from the Brunette drainage basin of Burnaby, B r i t i s h Columbia, from a v a r i e t y of s i t e s i n the land use c l a s s i f i c a t i o n s commercial (C), i n d u s t r i a l ( I ) , r e s i d e n t i a l (R) and open/greenspace (0). Results indicate that the t o x i c i t y to D. pulex and the chemical composition of the stormwater (measured by such parameters as COD, a l k a l i n i t y , hardness, hydrocarbons and trace metals) were influenced by land use and the i n t e r v a l between r a i n f a l l events. The i n d u s t r i a l and commercial land use s i t e s were the major source of those trace metals most often considered toxic to aquatic organisms, with runoff from the commercial s i t e s proving most toxic to the test organism ( t o x i c i t y followed the sequence C>I>R»0). Bioassays with synthetic stormwater (Cu, Fe, Pb and Zn, at concentra-tions observed from f i e l d samples) demonstrated that pH and suspended s o l i d s helped to regulate the t o x i c i t y of the trace metals, and implicated the importance of these elements i n natural stormwater t o x i c i t y . S t a t i s t i c a l comparison between synthetic and natural stormwater runoff t o x i c i t y yielded poor c o r r e l a t i o n ; however, t h i s was expected due to the inherent differences between the laboratory and f i e l d environments. A deta i l e d study of a single storm event indicated that while the " f i r s t - f l u s h " of the storm may be contributing to t o x i c i t y through the physi-c a l scouring of insoluble pollutants, the soluble pollutants proved to be more toxic and were washed out of the area over the en t i r e duration of the event. This prompted the author to propose the complete treatment of a l l stormwater runoff, and not simply the slug load of the f i r s t hour. - i i i -TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS i i i LIST OF TABLES v LIST OF FIGURES v i i ACKNOWLEDGEMENT i x 1. INTRODUCTION 1 2. LITERATURE REVIEW 2 2.1 Non-Point P o l l u t i o n Sources 2 2.2 Urban Stormwater Runoff 6 2.3 Contaminants i n Urban Runoff 7 2.4 T o x i c i t y of Trace Metals to Aquatic Invertebrates 9 2.5 F i e l d Versus Laboratory Derived Data 14 2.6 Other E f f e c t s of Urban Runoff 15 3. SAMPLING AND METHODOLOGY 19 3.1 Introduction 19 3.2 Phase One: In Situ F i e l d Sampling 19 3.2.1 Sampling Sites and Study Area 20 3.2.2 Sampling Equipment and On-Site Operation 26 3.2.3 Sampling and Lab Testing 28 3.2.4 A n a l y t i c a l Procedures 28 a) To t a l Suspended Solids 29 b) Conductivity 29 c) pH 29 d) Total A l k a l i n i t y 29 e) Total and Calcium Hardness 29 f) Chemical Oxygen Demand 29 g) Hydrocarbons 30 h) Trace Metals 30 3.2-5 Bioassay Materials and Procedures 31 a) D i l u t i o n Water 31 b) Choice of Daphnid Species 32 c) Bioassay Procedure 32 3.3 Phase Two: Laboratory Bioassay Using Synthetic Stormwater .... 35 3.3.1 Synthetic Stormwater Preparation 35 a) Trace Metals 35 b) pH E f f e c t s on Trace Metal T o x i c i t y 36 c) Sediment E f f e c t s on Trace Metal T o x i c i t y 37 d) Calcium Hardness E f f e c t s on Trace Metal T o x i c i t y .. 38 e) Hydrocarbon E f f e c t s on Trace Metal Toxicty 38 f) Combined E f f e c t s on Trace Metal T o x i c i t y 39 - i v -Page 3.4 Phase Three:"First-Flush"Analysis 40 3.4.1 Sampling Site 40 3.4.2 Sampling Equipment and Operation 40 3.4.3 Sampling and Lab Testing 41 3.5 S t a t i s t i c a l Procedures 42 3.5.1 L C 5 Q Determination 42 3.5.2 U.B.C. TRP 44 3.5.3 Marking-Dawson A d d i t i v i t y Index 45 4. RESULTS AND DISCUSSION 48 4.1 Phase One Introduction 48 4.1.1 General Parameters 49 a) COD 51 b) pH 54 c) Total Suspended Solids 54 d) A l k a l i n i t y , Hardness and Conductance 56 e) Hydrocarbons 60 4.1.2 Trace Metal Contamination i n Stormwater i n the Brunette River 61 4.1.3 Acute Bioassay Results 88 4.2 Phase Two: Bioassay Using Synthetic Stormwater 97 4.2.1 Calcium Hardness-Trace Metal and Hydrocarbon-Trace Metal Interactions 97 4.2.2 Experimental Comments 99 i ) Choice of Trace Metals 99 i i ) pH D r i f t Over Test Duration 100 4.2.3 Bioassay Results 100 i ) Trace Metal Interactions 101 i i ) Trace Metal-pH-T.S.S. Interactions 106 4.2.4 Conclusions: Laboratory Versus F i e l d Data 117 4.3 Phase Three: " F i r s t - F l u s h " Analysis 120 5. CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER RESEARCH 135 5.1 Conclusions 135 5.2 Recommendations 141 REFERENCES 143 APPENDICES Appendix A - Miscellaneous Data 149 Appendix B - Stormwater Data from the L i t e r a t u r e , as Summarized by Ferguson and H a l l (1979) 152 Appendix C - Complete Trace Metals Data for Phase One Sampling ....157 - v -LIST OF TABLES Page 2-1 Origins of Contaminants Contributing to NPS P o l l u t i o n : Sources ... 5 2-2 Origins of Contaminants Contributing to NPS P o l l u t i o n : A c t i v i t i e s 5 2-3 Cadmium and Lead T o x i c i t y to Aquatic Organisms 12 2- 4 T o t a l Trace Metals i n the Renfrew (B.C.) Stormwater Runoff 12 3- 1 Predominant Land Use at Each Sampling Site 25 3- 2 Detection Limits and Absorption Wavelengths for Trace Metals i n Stormwater Runoff 31 4- 1 Phase One Chemical Analysis Results 50 4-2 Phase One Trace Metal Analysis Results 63 4-3 L i n e a r i t y of Relationship Between Total Trace Metals and Total Suspended Sol i d s , A l l Sites 76 4-4 L i n e a r i t y of Relationship Between Total Trace Metals and Buildup Time, A l l Sites 76 4-5 E f f e c t of Land Use on Trace Metal Content i n Stormwater Runoff ... 81 4-6 Trace Metal Contamination S i g n i f i c a n t l y Different From Greenspace Areas 86 4-7 Trace Metal Data From the L i t e r a t u r e 87 4-8 Phase One Acute T o x i c i t y Bioassay Results 89 4-9 P a r t i a l Linear Regression C o e f f i c i e n t s f o r Chemical Parameters as a Function of T o x i c i t y 93 4-10 Multiple Regression Equations Relating T o x i c i t y to Independent Parameters 96 4-11 Deer Lake D i l u t i o n Water Chemical Analysis 102 4-12 Trace Metals Interaction Results 102 - v i -Page 4-13 Marking-Dawson Index Analysis Results: Trace Metals Interactions . 103 4-14 TRP Analysis on Trace Metal T o x i c i t y 105 4-15 Trace Metals - pH Interaction Results 107 4-16 Trace Metals - Total Suspended Solids Interaction Results 112 4-17 Trace Metals - pH-Total Suspended Solids Interaction Results 114 4-18 Marking-Dawson Index Analysis Results: Trace Metals - pH-T.S.S Interactions 114 4-19 TRP Analysis on pH and Total Suspended Solids E f f e c t s on Trace Metal T o x i c i t y 119 APPENDIX A - MISCELLANEOUS DATA 149 Al Sampling and R a i n f a l l Data f o r Phase One 150 A2 Deer Lake Water Chemical Analysis 151 APPENDIX B - STORMWATER DATA FROM THE LITERATURE AS SUMMARIZED BY FERGUSON AND HALL (1979) 152 Bl Seattle Urban Runoff Pollutant Concentrations Summary 153 B2 Summary Westwater Research S t i l l Creek Study Data 153 B3 GVSDD - S t i l l Creek Chemical Analyses Summary 154 B4 Results of EPS Stormwater Sampling 155 B5 Results of Stormwater Monitoring Programs Reported i n the L i t e r a t u r e 156 APPENDIX C - COMPLETE TRACE METALS DATA FOR PHASE ONE SAMPLING 157 Page - v i i -LIST OF FIGURES 2- 1 C h a r a c t e r i s t i c Dissolved Oxygen Sag Curve for Urban Runoff Input to Streams 16 3- 1 Brunette River Basin Study Area and Sampling Sites 21 3-2 Land Use Within the Brunette River Basin 22 3-3 T r a f f i c Volumes Within the Brunette River Basin 23 3-4 Typical Manhole/Stormsewer i n the Brunette River Basin 27 3- 5 Marking-Dawson A d d i t i v i t y Index 47 4- 1 COD-Buildup Time Relationship, A l l Sites 52 4-2 Conductance, Total A l k a l i n i t y and Total Hardness Vs. Buildup Time, A l l Sites 59 4-3 Hydrocarbons Vs. Buildup Time, A l l Sites 62 4-4 Cadmium D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin ... 64 4-5 Chromium D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin .. 65 4-6 Copper D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin .... 66 4-7 Iron D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin 67 4-8 Manganese D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin . 68 4-9 Nickel D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin .... 69 4-10 Lead D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin 70 4-11 Zinc D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin 71 4-12 Means, Ranges and Standard Deviations of Trace Metal Content i n Runoff From Each Land Use C l a s s i f i c a t i o n 79-80 4-13 Acute T o x i c i t y of Stormwater From the Brunette River Basin 90 - v i i i -Page 4-14 Trace Metal - pH T o x i c i t y Relationship 108 4-15 Trace Metal - T.S.S T o x i c i t y Relationship I l l 4-16 Trace Metal - T.S.S.-pH T o x i c i t y Relationship f o r Pb/Zn 115 4-17 Trace Metal - T.S.S-pH T o x i c i t y Relationship for Cu/Pb/Zn 116 4-18 Pollutograph: Total Suspended Solids and Conductivity 122 4-19 Pollutograph: Total A l k a l i n i t y and Total Hardness 123 4-20 Pollutograph: pH and COD 124 4-21 Pollutograph: Calcium and Copper 125 4-22 Pollutograph: Iron and Magnesium 126 4-23 Pollutograph: Lead and Nickel 127 4-24 Pollutograph: Zinc and L C 5 Q T o x i c i t y 128 - i x -ACKNOWLEDGEMENTS The author would l i k e to express his gratitude and thanks to h i s advisor, Dr. K.J. H a l l , for his kind assistance i n the research and prepara-t i o n of th i s t h e s i s . He also thanks Dr. D.S. Mavinic for reviewing t h i s manuscript. The author i s very g r a t e f u l to those people who as s i s t e d i n the c o l l e c t i o n and analysis of the samples used i n th i s research: Susan Liptak and Sue Jasper of the Environmental Engineering Labortory, U.B.C; and N.B. MacKinnon and R.L. S a c k v i l l e . As well, the author thanks K e l l y Lamb and Jane Gardner for t h e i r help i n the preparation of th i s manuscript. F i n a n c i a l support provided by the National Science and Engineering Research Council i s g r a t e f u l l y acknowledged. 1, CHAPTER 1 Introduction Much work has been c a r r i e d out on the i d e n t i f i c a t i o n of pollutants entering the aquatic environment, with the majority of researchers agreeing that these inputs stem from either well defined point sources, or nonpoint sources (NPS) which are much more d i f f i c u l t to i s o l a t e and contain. Of these nonpoint sources, i t i s suggested that urban runoff r e s u l t i n g from storm or other p r e c i p i t a t i o n events i s making one of the largest contributions to the o v e r a l l problem. In l i g h t of t h i s , attempts to mathematically model these runoff events, with the s p e c i f i c goal of c h a r a c t e r i z a t i o n of the runoff patterns and behaviour, have been common i n recent years and have met with mixed success; the s p a t i a l and temporal discreteness of i n d i v i d u a l sampling s i t e s within and between studies make the r e s u l t s of these analyses somewhat l i m i t e d i n a p p l i c a t i o n . While the need for these types of modelling studies seems j u s t i f i e d by the sheer scope of t h i s nonpoint p o l l u t i o n problem, l i t t l e or no work has been done on the actual e f f e c t s of the stormwater runoff on the inhabitants of the aquatic environment. The purpose of t h i s research was to address t h i s s p e c i f i c facet of urban stormwater runoff. The Brunette River basin, i n Burnaby, B r i t i s h Columbia, had been picked as the sampling l o c a t i o n due to previous studies, and a number of s i t e s had been selected from a v a r i e t y of land use c l a s s i f i c a t i o n s ; namely commercial, i n d u s t r i a l , r e s i d e n t i a l and open/greenspace/recreational. Stormwater runoff from these s i t e s was c o l l e c t e d and analyzed, and bioassays were run using Daphnia pulex i n an attempt to q u a l i f y the influence of land use on the pollutant loading of of urban runoff, as well as attempting to define and i s o l a t e any toxic components of the runoff from the d i f f e r e n t land uses. By .2. doing this the author has tried to show the importance of these kinds of studies in the overall analysis of urban stormwater runoff pollution; the author suspects that some toxic effects may be resulting from these kinds of inputs, and hopes that the results of this research w i l l point out the need for treatment of this type of water pollution. CHAPTER 2 3 . Literature Review 2.1 Nonpoint Pollution Sources (NPS) Organic and/or inorganic contaminants that are allowed to enter the aquatic environment are generally classified as coming from one of two sources: a point source, usually a single or series of individual, well defined and easily measureable points of origin, such as the outfall from a sewage treatment plant; or a nonpoint source (NPS) which i s not defined or easily identified, and i s therefore not quantitatively measureable to any degree of certainty. Contamination from point sources, by definition, i s well documented and easily amendable to control processes; however, as many studies have pointed out, this point source contamination may only be a small fraction of the total pollution problem. In an extensive study by the U.S. Council on Environmental Quality, i t was found that i n approximately 80% of the studied urban areas, downstream water quality was controlled by nonpoint sources of pollution (Wanielista, 1977). Whipple et a l . , (1974) found that in many developing urban and suburban areas, once secondary treatment had been implemented at municipal treatment plants (point sources), the nonpoint pollution sources accounted for more than half of the total contamination i n rivers and streams. For this reason they suggested that planning considera-tions in a l l urban areas must take into account these unrecorded wastes, as they w i l l be found too large to ignore. The authors estimated that for a relatively clean urban development, largely residential in nature and with associated shopping and service f a c i l i t i e s , we may expect an unrecorded B0D5 loading of approximately 0.02 to 0.03 lb/day/person (9-13.6 g/day/person), 4. with a s i g n i f i c a n t l y higher loading from areas with i n d u s t r i a l and/or heavy commercial f a c i l i t i e s . It seems clear that without planning considerations to treat t h i s unrecorded p o l l u t i o n source, very l i t t l e abatement of t o t a l p o l l u t i o n can take place, and may i n turn lead to grossly inadequate p o l l u t i o n control plans i n many cases. Colston (1974) concluded from his study on nonpoint sources that, i f the study town were to provide 100% removal of organic matter and suspended s o l i d s from the raw municipal wastewater on an annual basis, the t o t a l reduction of contaminants discharged to the receiving creek would only be 52% for COD, 59% for ultimate B O D 5 , 5% for suspended s o l i d s and 9, 6, 11, 21, 12 and 43% for the trace metals Cr, Cu, Fe, Pb, Ni and Zn, r e s p e c t i v e l y . Because the study town had separate storm sewers, from which we would expect a major discharge of NPS contaminants, the problems of NPS c o n t r o l were gr e a t l y compounded. As w e l l , a recent study by the U.S. General Accounting O f f i c e (Barton, 1978) concluded that even i f a l l point source pollutant discharges were to be c o n t r o l l e d , p o l l u t i o n from nonpoint sources would s t i l l prevent the achievement of the Clean Water Act's goal of s u f f i c i e n t l y high q u a l i t y waters for f i s h i n g and swimming by 1983. It w i l l be r e a d i l y seen that, i n order for any type of NPS co n t r o l strategy to be s u c c e s s f u l l y implemented, the general o r i g i n s of the contami-nants should be known. Whipple and Hunter (1977), a f t e r an extensive study of the problem, suggested a number of sources, which are summarized i n Table 2-1. We see from t h i s table that the sources are, for the most part, a r e s u l t of human a c t i v i t i e s and the subsequent wastes from these a c t i v i t i e s . While the l i s t i s far from complete, we do get some idea of the types of processes that produce NPS p o l l u t i o n . Further to t h i s , i n an e f f o r t to more exactly define the major a c t i v i t i e s contributing to NPS p o l l u t i o n , the U.S. 5. TABLE 2-1 ORIGINS OF CONTAMINANTS CONTRIBUTING TO NPS POLLUTION: SOURCES1 1. Drainage from animal feed l o t s 2. Commercial f e r t i l i z e r s 3. Pesticides 4. Acid P r e c i p i t a t i o n 5. P r e c i p i t a t i o n containing s u l f a t e s , mercury, cadmium, automotive o i l hydrocarbons or lead from exhaust gases 6. Saline i n t r u s i o n to groundwater 7. Wastewater and i n d u s t r i a l waste leakage 8. Septic tanks 9. Soli d waste disposal 10. Animal and human v i r a l and b a c t e r i a l contamination 11. De-icing s a l t s 12. Various wastes from recreational and i n d u s t r i a l watercraft i . e . , b i l g e water, garbage and sewage. 13. Sediment erosion 14. Waste heat 1 Whipple and Hunter, 1977 TABLE 2-2 ORIGINS OF CONTAMINANTS CONTRIBUTING TO NPS POLLUTION: ACTIVITIES 1 1. S i l v i c u l t u r e - growing and harvesting of timber f o r lumber and paper production 2. Agriculture 3. Mining 4. Construction of roads and buildings 5. Salt-water i n t r u s i o n into freshwater supplies 6. Subsurface excavations, including i n j e c t i o n wells, septic tanks and l a n d f i l l s 7. Hydrologic modifications i . e . , p o l l u t i o n r e s u l t i n g from changes i n movement, flow or c i r c u l a t i o n of surface groundwaters, including changes by dams, levees, channels or flow diver s i o n s . 8. Urban runoff 1 U.S. EPA; from Barton, 1978 6. EPA looked into the problem, and the sources they suggest are summarized i n Table 2-2 (Barton, 1978). From these tables, we see that the o r i g i n s of NPS pollutants are many and varied, and not e a s i l y defined; however, for any type of source control to be su c c e s s f u l l y implemented, these NPS o r i g i n s must be defined and q u a l i f i e d . 2.2 Urban Stormwater Runoff Included i n Table 2-2 i s urban runoff, a nonpoint source which many f e e l i s a major contribution to the o v e r a l l p o l l u t i o n problem. In f a c t , the U.S. EPA suggests that urban runoff may r i v a l a g r i c u l t u r e as the worst contributor to NPS p o l l u t i o n , and may be a far more serious p o l l u t e r i n many areas (Barton, 1978). For example, Washington, D.C.'s Occoquan watershed area development p o l i c y o r i g i n a l l y c a l l e d f o r the construction of an advanced wastewater treatment plant, on the premise that such a plant would encourage the conversion of a g r i c u l t u r e lands to suburban development, and thereby reduce o v e r a l l p o l l u t i o n by reducing a g r i c u l t u r e runoff. However, a June, 1977 study (North V i r g i n i a Planning D i s t r i c t Commission et a l . , 1977) revealed that pollutant loadings i n runoff from urbanized areas were considerably higher than those from a g r i c u l t u r e lands, and concluded that any improvement i n water qu a l i t y r e s u l t i n g from the i n s t a l l a t i o n of the proposed treatment plant would be more than o f f s e t by increases i n nonpoint source p o l l u t i o n ( i n the form of urban runoff created by the ensuing urbanization). The contaminants contained i n urban runoff are again many and varied, ranging from toxic trace metals associated with p a r t i c u l a t e matter to hydro-carbons and excess nutrients, and as such, i t would be expected that the sources of these contaminants are just as diverse. The U.S. EPA (Barton, 1978) suggest that these sources may include pesticides and f e r t i l i z e r s from 7. construction s i t e s , lawns and gardens; o i l y residue from brakes, clutches and t i r e s ; street l i t t e r ; dust; a i r p o l l u t i o n p a r t i c u l a t e s ; toxic heavy metals from the combustion of petroleum products; and sanitary and storm sewer over-flows, a l l of which may be contributing to the severe water p o l l u t i o n problem i n urban areas. Due to the great number of contributors to urban runoff, Whipple and Hunter (1977) suggested that u n t i l the stormwater s i t u a t i o n i s analyzed and co r r e c t i v e measures taken, there would be l i t t l e or no sense i n seeking higher l e v e l s of treatment e f f i c i e n c y i n e x i s t i n g secondary plants. 2.3 Contaminants i n Urban Runoff There are numerous studies d e t a i l i n g the contribution of urban runoff to the o v e r a l l p o l l u t i o n problem of the urban environment. For example, Hunter and MacKenzie (1979) have estimated that 4.2 x 10 9 L of automobile and indus-t r i a l l u b r i c a n t s are l o s t to the environment annually i n the U.S., ei t h e r d i r e c t l y by disposal to sewers and land a p p l i c a t i o n , or i n d i r e c t l y by s p i l l -age and leakage from v e h i c l e s . This, they suggest, r e s u l t s i n a high concen-t r a t i o n of these f r a c t i o n s i n runoff from t h i s environment, due mainly to the fact that most petroleum hydrocarbons are associated with the p a r t i c u l a t e matter which i s most e a s i l y transported during a storm event (as high as 86.4%, according to t h i s study). While runoff constituents such as trace organic compounds ( i e . , p e s t i -cides, p o l y c y c l i c aromatic hydrocarbons, e t c . ) , and nutrients such as n i t r o -gen and phosphorous (which contribute to problems of entrophication) are themselves important factors i n the urban runoff problem, by far the most research has been conducted into the quantities of trace metals being washed into the aquatic environment. V i t a l e and Sprey (1974) estimate that for a t y p i c a l moderate-sized c i t y (population 25,000 or more), annual loading rates 8. of lead and mercury to nearby receiving waters would be i n the order of 100,000 - 200,000 pounds (45,360 - 90,720 kg) for Pb and 6,000 - 30,000 pounds (2,270 -13,600 kg) for Hg. Getz et a l . , (1977) proposed that as much as 16,000 kg/year of lead i s released to the aquatic environment as a d i r e c t r e s u l t of automobile emissions, for the same moderate-sized urban centre. Newton et a l . , (1973) looked at t h i s same problem and suggested that lead i s the primary contaminant of automobile emissions and tends to be deposited within a short distance from emission on the road bed. The highest concen-t r a t i o n , they found, was deposited d i r e c t l y on the road bed where i t tended to remain unadsorbed. During a runoff event, t h i s Pb would be washed o f f the roadbed ( i n an e s s e n t i a l l y complete scouring process), where i t would be mixed and adsorbed onto the surrounding p a r t i c u l a t e matter and c a r r i e d away i n the runoff. They concluded that we should expect to f i n d the highest concentrations of Pb on the roadbed, with decreasing amounts with increased distance from the road, i n low runoff periods and dry weather. 0 Further, Wilber and Hunter (1979) looked at the d i s t r i b u t i o n of trace metals i n street sweepings, stormwater s o l i d s and urban aquatic sediments as a r e s u l t of stormwater runoff i n the Lodi, N.J. area, and they too found high lead concentrations as a r e s u l t of automobile emissions (they a t t r i b u t e d t h i s to the combustion of gasoline containing an a l k y l - l e a d antiknock component). They a l s o found that metal concentrations gradually increased with distance downstream through the urban area, suggesting a cumulative e f f e c t of deposi-t i o n as a d i r e c t r e s u l t of stormwater runoff (and, to a lesser degree, the other nonpoint sources previously mentioned). In connection with t h i s , the authors examined the enrichment of "a v a i l a b l e " metals (those trace metals either dissolved i n the water column, or e a s i l y exchangable and thereby read-i l y a v a i l a b l e for uptake by aquatic organisms) i n the upstream area of the 9. urban centre as compared to the downstream area, and found increases i n "available" metal concentrations i n the downstream area ranging from 610% for Zn, to 320%, 420% and 430% for Pb, Cu and Ni, respectively. This suggests a massive input of these trace metals into the stream from runoff in the urban area. Finally, studies on the amounts of toxic trace metals being carried into open water courses by urban runoff, and the general avai l a b i l i t y of these metals to aquatic organisms (Hall et a l . , 1974; Hall et a l . , 1976, Koch et a l . , 1977; Pitt and Amy, 1973) conclude that these loadings should be consi-dered a potential contributing factor to stream deterioration, as a direct result of their accumulation in the bottom sediments of quiescent stream areas. These studies found that, while the general solubilities of the trace metals are low (in the range of less than 10%), Pb, Cu, and Zn are s t i l l sufficiently soluble to create toxic conditions for aquatic organisms under select circumstances (such as soft water). It therefore appears that d i l u -tion by the watercourse i s not always sufficient to maintain these metals at non-toxic concentrations, and with the expected slug loadings from urban runoff, toxic conditions may be resulting. 2.4 Toxicity of Trace Metals to Aquatic Invertebrates It seems obvious from a review of the available literature that urban stormwater runoff makes an extremely large contribution to stream pollution. However, the actual effects of this type of pollution are d i f f i c u l t to quan-t i f y , due to the complex chemical composition of the runoff and the transient nature of the receiving body i e . , one would expect factors such as mixing and dilution, taking place i n water bodies of sufficient volume, to overcome any acutely toxic effects arising from stormwater input, and i n most cases this 10. i s probably what happens. However, under select conditions such as low flow volume, high slug load of pollutant, l i t t l e mixing or extreme concentration i n l o c a l i z e d areas, we might expect to f i n d several detrimental e f f e c t s on the aquatic b i o t a , and a number of studies have been undertaken to examine t h i s . Please note that t h i s review i s only dealing with acutely t o x i c e f f e c t s , with the r e a l i z a t i o n that other, l e s s obvious e f f e c t s i . e . , long term, sub-lethal t o x i c i t y ; biomagnification and bioconcentration; e f f e c t s on population dynamics through reproductive impairment, etc., are just as preva-l e n t and, i n some cases more important. The concentrations of trace metals at which toxic e f f e c t s have been shown to take place are well documented (Beisinger and Christensen, 1972; Bringmann and Kuhn, 1978; Applegate et a l . , 1957; Brown, 1968; Calamari and Marchetti, 1973; Hale, 1977; Davies et a l . , 1976), and expected concentra-tions of metals i n urban stormwater have been recently well studied (Koch et^ a l . , 1977; Ferguson and H a l l , 1978; Sidhu, 1975(a); Sidhu, 1975(b); M i l l s , 1977). Together, some idea can be obtained as to the possible short-term e f f e c t s of exposure of s p e c i f i c indigenous aquatic species to the trace metal concentrations associated with some urban stormwater runoff. Beisinger and Christensen (1972) estimate the "safe concentrations" of trace metals to Daphnia magna (those concentrations to which the organism can be exposed with no acutely toxic e f f e c t s ) as Cu 22pg/L; Zn 70 ug/L; and Cd 0.17 ug/L; at the same time, Koch et a l . , (1977) report metal concentrations i n the Renfrew (Vancouver, B.C.) storm sewer area to be far i n excess of these values, and a f t e r sampling i n the Burnaby (B.C.) stormsewer, Ferguson and H a l l (1979) report values for metal concentrations to be, i n some cases, as high as ten times the suggested "safe concentration". This would i n d i c a t e that, i n some cases, storm sewer e f f l u e n t may be highly toxic to some 11. aquatic organisms i n short-term exposure; however, i n t e r p r e t a t i o n and a p p l i -cation of laboratory data to a f i e l d environment i s extremely d i f f i c u l t , and any such useage should be viewed with the greatest caution. As well, the form of the trace metal may vary between studies; as such, exact comparison can be d i f f i c u l t . Spehar et a l . , (1978) investigated the t o x i c i t y of cadmium and lead to a number of freshwater species. Four i n s e c t s , one s n a i l and one amphipod were exposed to various concentrations of cadmium [as CdCl^] and lead [as FbCNO^^] for time periods ranging from 28 days to 3 years. Their r e s u l t s are summar-iz e d i n Table 2-3. They found that the 28-day LC ^ Q t o x i c i t y values for cadmium-exposed s n a i l s and lead-exposed amphipods were 11 and 14 times lower than the seven and four day (96 hour) values for these metals, respectively, i n d i c a t i n g more t o x i c i t y with longer exposure periods. As well, they observed that the lowest e f f e c t concentrations obtained a f t e r 28 days for cadmium-exposed mayflies and s n a i l s and lead-exposed amphipods were s i m i l a r to those a f f e c t i n g f i s h exposed over t h e i r e n t i r e l i f e cycle i n water of si m i l a r q u a l i t y , i n d i c a t i n g a much greater s e n s i t i v i t y by the insects compared with f i s h . There were no l e t h a l threshold concentrations observed for species exposed to either metal, i n d i c a t i n g that possible e f f e c t s could occur at lower concentrations during longer exposure periods. F i n a l l y , i t was shown that cadmium and lead concentrations i n the animals tested generally increased with increasing water concentrations, and were 9,000 -30,000 times greater than corresponding metal concentrations i n the water. By comparing these lowest e f f e c t concentrations with known values for concen-tr a t i o n s of trace metals i n stormsewers (Table 2-4), some i n d i c a t i o n of the p o t e n t i a l damage th i s runoff may cause can be obtained. In spite of the 1 2 . TABLE 2-3 CADMIUM AND LEAD TOXICITY TO AQUATIC ORGANISMS1 METAL LOWEST EFFECT CONCENTRATION (Ug/L) EXPOSURE PERIOD TEST WATER HARDNESS (as mg/LCaC03) SPECIES Cadmium [as CdCl 2] 3.0 27.5 8.1 28d 28d lOOd 45 45 45 Ephemerella sp.(Mayfly) Physa integra (Snail) Jordanella f l o r i d a e (Flagfish) 3.4 3 yrs 45 Salvelinus f o n t i n a l i s (Brook Trout) Lead fas Pb(N0o)ol 32 28d 45 Gammarus pseudolimnaeus (Amphipod) 119 3 yrs 44 Salvelinus f o n t i n a l i s (Brook Trout) 7.6 2-3 yrs 28 Salmo gairdneri (Rainbow Trout) 1 Spehar et a l . , 1978 TABLE 2-4 TOTAL TRACE METALS IN THE RENFREW (B.C.) STORMSEWER RUNOFF1 DATE TIME TOTAL TRACE METALS (Mg/L) Cu Fe Mn Ni Pb Zn 19 Nov. 0900 20 2100 90 — 155 130 1245 15 910 71 — 55 213 1603 12 670 32 <1 31 174 21 Nov. 0835 8 590 87 <1 <1 25 1200 6 610 109 <1 <1 41 1520 8 1040 123 <1 7 158 Median 10 790 88 <1 19 144 Mean 12 990 85 — — 124 Standard Deviation 5 570 32 75 Koch et a l . , 1977 1 3 . d i f f i c u l t y involved i n drawing an absolute comparison between these figures (since Table 2-3 involves soluble trace metal, while Table 2-4 deals with t o t a l trace metals), one can suggest that stormwater of t h i s q u a l i t y could conceivably be toxic to c e r t a i n aquatic organisms under the proper conditions, i e . , l i t t l e mixing, low flow, etc. Eaton, McKim and Holcombe (1978) examined the t o x i c i t y of cadmium and copper to l a r v a l and juvenile f i s h populations and found that, i n the case of embryos and larvae of seven freshwater species exposed to low concentrations of cadmium i n soft water, a l l species were k i l l e d or t h e i r growth retarded by concentrations ranging from approximately 4 to 12 pg/L Cd. From experimenta-t i o n using the same species exposed to varying concentrations of copper, i t was concluded that, for a l l species except Northern pike, the concentrations of copper that caused s i g n i f i c a n t e f f e c t s on the l a r v a l standing crop were s i m i l a r , i n the range from 31.7 to 43.5 pg/L Cu. In a d d i t i o n , the authors decided that the agreement between t h e i r r e s u l t s and those from l i f e - c y c l e chronic t o x i c i t y studies indicated that embryo and l a r v a l exposures w i l l give r e l i a b l e estimates of the chronic t o x i c i t y of cadmium to a d d i t i o n a l f i s h species. While comparison of these t o x i c i t y l e v e l s to trace metal concentra-tions i n c e r t a i n stormsewers doesn't appear to suggest any expected t o x i c i t y , further data on trace metals i n urban runoff may i n d i c a t e the opposite. While the data from Table 2-4 were obtained i n November, a time of frequent p r e c i p i t a t i o n i n the area and correspondingly low contaminant buildup, figures for the same stormsewer i n July i n d i c a t e that, i n at least one instance, copper l e v e l s were much higher than the experimentally determined toxic concentrations for these p a r t i c u l a r f i s h species. Again, one must be aware of the problems comparing t o t a l versus soluble trace metal f r a c t i o n , but there i s evidence to indicate the p o s s i b i l i t y of toxic e f f e c t s r e s u l t i n g 14. from urban runoff i n th i s p a r t i c u l a r instance. 2.5 F i e l d Versus Laboratory Derived Data A common c r i t i c i s m of laboratory-derived chronic or acute t o x i c i t y l e v e l s for trace metals i s that t h i s data cannot be applied to the f i e l d environment with any degree of c e r t a i n t y , and i n many cases these t o x i c i t y l e v e l s have no r e a l s i g n i f i c a n c e outside the lab (Winner and F a r r e l l , 1976; Stephan and Mount, 1973; Sprague, 1976). In add i t i o n , the f i e l d i n v e s t i g a -t i o n of toxic e f f e c t s r e s u l t i n g from urban stormwater runoff i s made very d i f f i c u l t by a multitude of factors i . e . , transient nature of the receiving body; d i f f i c u l t y i n d i f f e r e n t i a t i n g natural population death from runoff-induced mortality; lack of controls i n the f i e l d , etc. Marshall (1978), i n his study on the f i e l d v e r i f i c a t i o n of cadmium t o x i c i t y to laboratory daphnia populations however, concluded that by studying the e f f e c t s of toxic stress on both laboratory and natural population rates-of-increase, we could draw meaningful comparisons between the lab and the f i e l d environments, and apply lab data to the f i e l d with an increased degree of confidence. In another study, Digiano et a l . , (1975), looking at the f i e l d e f f e c t s of urban runoff p o l l u t i o n , showed a p a r t i c u l a r l y s i g n i f i c a n t r e l a t i o n s h i p between trace metal concentration i n the runoff and the condition of the benthic macroinverte-brate community, an important part of the food chain for f i s h . They found that, at the upper stations of the study r i v e r area, removed from the i n f l u -ence of urban runoff, l a r v a l forms of mayflies, s t o n e f l i e s , c a d d i s f l i e s and midges were i n abundance. However, species d i v e r s i t y was much reduced at stations downstream, p a r t i c u l a r l y by l a t e summer, with s n a i l s comprising the bulk of the macroinvertebrate population at the lower r i v e r reaches. This, they suggest, was due to some adverse influence which was acting to eliminate 15. or Impede the col o n i z a t i o n of the r i v e r bottom by various species. Corres-pondingly, i t was found that metal concentrations i n the sediments and d e t r i t u s were greater at the lower sampling stations (those more influenced by measured urban runo f f ) , and the benthic organisms contained metal concen-tr a t i o n s at s t i l l higher l e v e l s , ranging up to two orders of magnitude greater than normally accepted acute t o x i c i t y thresholds. It was concluded that, while there were no in d i c a t i o n s the urban runoff was having any e f f e c t on r e g u l a r l y monitored physical and chemical water parameters, the runoff may have been exerting a much larger than expected influence on the macroinverte-brate population as a r e s u l t of the a s s o c i a t i o n of runoff trace metals to the bottom sediments and d e t r i t u s . 2.6 Other E f f e c t s of Urban Runoff The e f f e c t s of urban stormwater runoff are not l i m i t e d to t o x i c i t y from trace metals associated with p a r t i c u l a t e matter. Numerous studies have been done on the e f f e c t s of slug loads of organic p a r t i c u l a t e matter on the dissolved oxygen l e v e l s i n the receiving r i v e r or stream, and i n many cases, i t has been shown that t h i s runoff can reduce the D.O. to dangerously low l e v e l s . In a study ca r r i e d out by Enviro Control, Inc. for the U.S. Council on Environmental Quality (Whipple and Hunter, 1977), i t was determined that, a f t e r stormwater runoff entered an estuary, the ambient D.O. l e v e l dropped 2.05 mg/L immediately a f t e r the storm, and the l e v e l continued to decrease u n t i l a minimum l e v e l was reached a f u l l s ix days a f t e r the storm. Further, i t was found that the estuary system took from two to s i x more days to f u l l y recover from t h i s slug stormwater load. Obviously t h i s would r e s u l t i n considerable shock and probable damage to the system involved. Rimer, Nissen and Reynolds (1978) also looked at t h i s problem, as a 16. 17. function of runoff from various land use types. From their findings they derived certain general characteristics about the effects of NPS runoff, as well as a characteristic D.O. sag curve for stormwater input (Figure 2-1). They propose that the low flow antecedent D.O. levels are generally low for most of the year, due to the presence of point source discharges only (A). As surface runoff from a storm event begins to enter the stream, the effects of increased flow tend to drive the D.O. level up (B). As the runoff con-tinues to enter the stream however, the effects of the oxygen-demanding materials washed into the stream exert themselves and begin to outweigh the effects of increased stream reaeration and temperature fluctuations; this i n turn causes an oxygen sag of varying magnitude and duration (E), until a low point i s reached (C). As flow continues to increase, temperature and flow effects appear to reassert themselves, driving the D.O. concentration up once more. From their results, the authors proposed a number of general character-i s t i c s about NPS runoff which they feel can be applied i n most situations. These include: 1. NPS runoff generally depresses D.O. below antecedent conditions by about 1.0 mg/L at c r i t i c a l impact locations. 2. In certain stream reaches where point source impacts are large and ante-cedent D.O. levels depressed, the net effect of the NPS pollution i s to raise the D.O.' levels above antecedent conditions. 3. In those cases where NPS pollution causes D.O. violations, they w i l l last from a few hours to ten days, for an 8-year recurrence interval storm. 4. NPS generated D.O. violations are restricted to certain selected reaches of the major receiving streams of the study area. In conclusion, the authors suggest that while the data does indicate that NPS 18. runoff pollution depresses the D.O. concentrations i n streams for almost a l l storm events, the overall impact i s small and the receiving body i s able to renew i t s e l f with l i t t l e trouble. There are, of course, other detrimental effects generated by the input of NPS runoff to a receiving body i.e., eutrophication through high nitrogen and phosphorous inputs; toxicity resulting from trace organic compounds such as hydrocarbons, f e r t i l i z e r s , pesticides, P.A.H's, etc.; suspended sediments and turbidity problems, i n the form of fish g i l l abrasion, smothering of eggs and smaller invertebrates, reduction of light penetration, etc. This review and research has, however, concentrated on those aspects of urban runoff considered most important by investigators. CHAPTER 3  Sampling and Methodology 19. 3.1 Introduction The experimental work i n t h i s research i s divided into three sections; 1) i n - s i t u f i e l d sampling at storm sewers; 2) laboratory studies i n v o l v i n g bioassays using synthetic stormwater; and 3) " f i r s t - f l u s h " a n alysis using stormwater c o l l e c t e d from the s i t e with the worst-case pollutant loading, as determined i n section one. The sampling and a n a l y t i c a l procedures for each phase are described i n t h i s chapter. 3.2 Phase One: In-Situ F i e l d Sampling Twelve sampling s i t e s were chosen i n the Brunette River drainage basin, encompassing S t i l l Creek, Burnaby Lake and Deer Lake. These s i t e s were selec-ted on the basis of the proximity of t h e i r stormsewer outlets to the above mentioned water bodies (and t r i b u t a r i e s ) . The s i t e s were taken from a range of land use c l a s s i f i c a t i o n s , i n an attempt to delineate e x i s t i n g r e l a t i o n -ships between the land use and p o l l u t i o n a l concentrations or loadings from stormwater runoff i n the p a r t i c u l a r area. The land uses studied were r e s i -d e n t i a l , i n d u s t r i a l , commercial and open/green space. At each s i t e , sampling of the runoff was undertaken during a storm event, using automated water samplers positioned at the stormsewers. The c o l l e c t e d water was analyzed i n the laboratory f o r a number of key chemical and physical properties and bioassays were run to determine the acute t o x i -c i t y of the stormwater to Daphnia pulex. 20. 3.2.1 Sampling Sites and Study Area The research was ca r r i e d out e n t i r e l y i n the Brunette River drainage basin, a highly urban/industrialized area encompassed within the municipali-t i e s of Burnaby, Vancouver and New Westminster, B.C. This area was selected with the idea of integ r a t i n g e x i s t i n g information on trace metal contamina-tio n within the basin, obtained from previous research ( H a l l et a l . , 1976; Koch et a l . , 1977; Ferguson and H a l l , 1979; and Bindra and H a l l , unpublished) with the updated r e s u l t s from t h i s study. This would enable the author to obtain more comprehensive information on the magnitude of the trace metal p o l l u t i o n problem i n t h i s p a r t i c u l a r basin, and the degree to which urban stormwater runoff i s contributing to the problem. The major water bodies located within the basin include two shallow lakes, namely Burnaby Lake (120 hectares; 300 acres) and Deer Lake (32 hectares; 80 acres); S t i l l Creek, running into Burnaby Lake from the west, and the Brunette River, draining Burnaby Lake from the east (with an average discharge of 2.7 m 3/sec; 85 c f s , as measured at the l e v e l control dam at the head of the Brunette River). S t i l l Creek, the main t r i b u t a r y entering Burnaby Lake (58 percent of the flow) i s c l a s s i f i e d as a storm sewer by the Greater Vancouver Regional D i s t r i c t , and as such some portions of the creek are contained i n covered sewers. Figure 3-1 i l l u s t r a t e s the main features of the drainage basin. Land use i n the area i s quite varied, encompassing the wide v a r i e t y of designations most commonly found i n a highly populated urban environment. Of a t o t a l area of 6,060 hectares (15,000 acres), 1972 estimates had 42% residen-t i a l , 31% open space and forested, 15% commercial and i n s t i t u t i o n a l , 5.5% i n d u s t r i a l , 5% recreational and 1% major transportation corridors ( H a l l £t a l . , 1976). F i g . 3-2 shows a more recent (1974) land use map of the area. Figure 3-1 Brunette River Basin Study Area and Sampling Sites Figure 3-2 Land Use W i t h i n the Brunette R i v e r B a s i n . N.B. Numbered S i t e s From Study by H a l l et a l . , 1976. 24. The commercial and i n s t i t u t i o n a l land uses are scattered throughout the basin, while the heavy i n d u s t r i a l areas are concentrated along the north sides of S t i l l Creek and Burnaby Lake. The most extensive area of open space i s on the steeper slopes of Burnaby Mountain, the edge of Burnaby Lake and Deer lake, and the southern, poorly drained reaches of S t i l l Creek. In a d d i t i o n to these land uses, the basin also contains the major east-west transportation corr i d o r between Vancouver and the outlaying areas of the Fraser Valley. There are two main highways following the lower elevation contours throughout the length of the basin (Highway 401 and Lougheed High-way) as well as a number of major north-south access roads to these highways. As indicated by Figure 3-3, the 1977 average weekday t r a f f i c volume for High-way 401 was 70,000 veh i c l e s , while the Lougheed Highway c a r r i e d approximately 32,000 veh i c l e s , as measured at Boundary Road. This represents a s i g n i f i c a n t increase i n t r a v e l along Highway 401 since 1972 (approximately 40%), and as such also suggests a major source of trace metal p o l l u t i o n as a r e s u l t of automotive exhaust, o i l leakage, etc. According to the Department of Muni-c i p l e A f f a i r s (1974), 15.4 percent of the t o t a l land area of Burnaby i s taken up by st r e e t s , roads and a l l e y s , i n d i c a t i n g the degree to which transporta-t i o n i s a major land use i n the area. Figure 3-1 shows the l o c a t i o n of the twelve sampling s i t e s used i n t h i s study while Table 3-1 indicates the main land use at each s i t e . As a f i r s t step, the land use i n the Brunette basin was divided into i n d u s t r i a l , commer-c i a l , r e s i d e n t i a l and open/recreational, and from each of these categories three s i t e s were chosen with the i n t e n t i o n of c o r r e l a t i n g f i n a l trace metal and acute bioassay t o x i c i t y r e s u l t s with s p e c i f i c land use. The basis for s i t e s e l e c t i o n was the l o c a t i o n of stormsewers i n the sampling area i n r e l a -t i o n to water bodies, i n l i g h t of the purpose of the study, i . e . , to examine 25. the acute toxicity of stormwater run off to aquatic organisms. Sites were chosen which were as close to these water bodies as possible, so that the data and conclusions obtained from sampling could be more readily transposed to the water body i t s e l f . It i s realized that each site encompassed a variety of land uses, and as such could not be thought of as draining any one particular type; however an exact breakdown of uses at each site proved extremely d i f f i c u l t to obtain. The author i s satisfied with the usage of the term "predominant land use" at each site, and suggests that, because that particular use encompasses at least 75% of the land use in the sampling areas, any effects seen from the runoff of the area can be ascribed to the particular land use indicated in Table 3-1. A more detailed breakdown i s beyond the scope of this research, but may be borne i n mind for future studies. TABLE 3-1 PREDOMINANT LAND USE AT EACH SAMPLING SITE SITE NO. SITE LOCATION LAND USE* 1 Brentwood Mall C 2 Beaverbrook Drive at Beaverbrook Cres R 3 Production Way at Thunderbird Cres. I 4 Lougheed Mall C 5 Wedgewood Avenue at 1st Avenue R 6 Robert Burnaby Park 0 7 Deer Lake Park 0 8 Haszard Street at Buckingham Avenue R 9 Warner Loat Park 0 10 Brunette Avenue at Capilano Street I 11 Alpha Way at Alaska Street I 12 Boundary Road at Canada Way C C = Commercial R = Residential I = Industrial 0 = Open/Greenspace/Recreational 26. \ 3.2.2 Sampling Equipment and On-Site Operation At each s i t e , samples were taken using an ISCO automated water sampler. The stormsewer cover was removed and the sampler positioned with tube extended into the sewer, below the dry weather flow water l e v e l . A t y p i c a l stormsewer of the basin i s shown i n Figure 3-4. At the onset of a r a i n f a l l event, the sampler was activated by hand as soon as there was a constant runoff flow from the street surface surrounding the stormsewer; t h i s buildup time varied between s i t e s as a r e s u l t of the differences i n i n t e n s i t y between each i n d i v i d u a l storm; as well, the length of sampling period at each s i t e was greatly dependent on the duration and i n t e n s i t y of the storm. Many studies have indicated the presence of a " f i r s t - f l u s h " e f f e c t , or a washing out of the majority of the pollutant load within the f i r s t hour of most char-a c t e r i s t i c storm events (DiGiano et a l . , 1975; Helsel et a l . , 1979) and i t i s of obvious importance that t h i s f r a c t i o n of the storm be sampled; however, i f only the f i r s t hour of the storm were sampled, the r e s u l t s would be biased toward a higher pollutant concentration than was a c t u a l l y washed out over the e n t i r e storm event. In order to avoid t h i s , the r e s u l t s between f i r s t - f l u s h concentrations and those of the remainder of the storm should be flow-weighted and averaged to give a more representative impression of the average p o l l u t i o n a l load of the p a r t i c u l a r storm i f sampling only selected portions of the storm. In t h i s study, however, i t was decided that sampling be ca r r i e d out over the entire period of the storm (where possible), to obtain average as well as peak pollutant concentrations. F i g u r e 3 - 4 T y p i c a l Manhole/Stormsewer i n the Brunette River Basin. 2 8 . 3.2.3 Sampling and Lab Testing During a r a i n f a l l event, following an antecedent dry period of at least 24 hours, the sampler was manually activated and 500 mL of stormwater was c o l l e c t e d from the base of the stormsewer at ten minute i n t e r v a l s i n t o 500 mL wide-mouth p l a s t i c sampling j a r s . The water was transported to the lab immediately upon completion of sampling, where te s t i n g was begun on i t s phy-s i c a l and chemical properties. Following i n i t i a l t e s t s , the samples were a c i d i f i e d using concentrated n i t r i c a c i d at an add i t i o n rate of 5 percent by volume, and placed i n cold storage (4°C). A l l tests were complete within 96 hours of sampling. Although the stated i n t e n t i o n was to sample over the e n t i r e duration of a storm, most of the storms sampled during t h i s phase were of an unusually short duration ( i n general, le s s than two hours), and there was consequently l i t t l e d i f f e r e n t i a t i o n between the f i r s t - f l u s h and periods of lower pollutant concentration. For t h i s reason, a composite sample was prepared from the stormwater obtained for each s i t e . A l l t e s t s , including bioassays, were performed using these composite samples, and only trace metal analysis and pH tests were c a r r i e d out on the i n d i v i d u a l ten-minute samples, for s i t e s one to eight. Due to time constraints, only si n g l e grab samples were taken at s i t e s nine to twelve, and these were treated as the composite samples from these s i t e s . S u f f i c i e n t runoff time was allowed to obtain a representative sample from these s i t e s , given the number of antecendent dry days. Table A - l (Appendix A) shows sampling dates, buildup time and r a i n f a l l data for the sampling c a r r i e d out at each s i t e . 3.2.4 A n a l y t i c a l Procedures A l l tests were performed with reference to Standard Methods for the Examination of Water and Wastewater (A.P.H.A., 1978). 29. a) T o t a l Suspended S o l i d s : A known volume of sample was vacuum f i l t e r e d through a Gooch c r u c i b l e with g l a s s - f i b r e f i l t e r . The cr u c i b l e was dried overnight at 104°C to evaporate a l l excess moisture, and weighed to give a sol i d s weight per unit volume measurement of the suspended s o l i d s . b) Conductivity: The actual conductivity and d i l u t e d conductivity (a better i n d i c a t i o n of the r e a l conductivity of the sample, due to the absence of i o n crowding and pair formation) were measured on a l l composite samples at room temperature (corrected to 25°C) using a Radiometer CDM3 Conductivity Meter with a platinum-electrode s p e c i f i c conductance immersion c e l l . c) pH: The pH was determined i n a l l samples (composite and 10-minute) as soon as possible a f t e r sampling, using a Fisher Model 210 pH meter with glass reference electrode, standardized against Fisher pH4 and pH7 buffer s o l u t i o n . d) T o t a l A l k a l i n i t y : A l l samples had i n i t i a l pH of less than 8.3 (see Chapter 4, r e s u l t s ) the phenolphthalein end point, and therefore a l k a l i n i t y i n the samples was due mainly to the bicarbonate i on HCO3 -, as measured by the methyl orange end point. A l l samples were t i t r a t e d to th i s end point using 0.02N standard a c i d , and res u l t s reported as milligrams per l i t r e a l k a -l i n i t y as calcium carbonate (CaC03). e) T o t a l and Calcium Hardness: A l l composite samples were analyzed f o r t o t a l and calcium hardness as soon as possible a f t e r sampling. The appro-priate i n d i c a t o r (Eriochrome black T or Eriochrome blue black R) was added to sample a l i q u o t s and these were t i t r a t e d to t h e i r subsequent equivalence points using 0.01M EDTA. f) Chemical Oxygen Demand: COD tests were run on the composite samples as a measure of carbon, since t h i s test i s rapid compared to the B0D5 procedure and provides a useful estimate of the t o t a l carbon-containing elements i n the sample. Twenty m i l l i l i t r e a liquots were refluxed i n the presence of excess 30. potassium dichromate and mercurous sulphate i n an a c i d environment (cone. R^SO^ + Ag 2S0 1 +), and the excess dichromate t i t r a t e d using standardized ferrous ammonium sulphate and f e r r o i n i n d i c a t o r . g) Hydrocarbons: Five hundred m i l l i l i t r e s of composite sample were a c i d i -f i e d with 2.5 mL of concentrated H^SO^, followed by the a d d i t i o n of 2.5g of NaC£ and 20 mL of carbon t e t r a c h l o r i d e , and the solutions mixed. The lower CC^ phase was drained o f f and the C-H stretching frequency of t h i s f r a c t i o n measured from 3300 - 2600 cm - 1 on a Perkin-Elmer Model 247 Grating Infrared Spectrophotometer. The i n t e n s i t y of absorption was compared to standard curves prepared with iso-octane, and the r e s u l t s reported as milligrams per l i t r e iso-octane (Gruenfeld, 1973 and Z e l l e r , 1974; as reported i n Koch et_ a l . , 1977). h) Trace Metals: Preliminary testing on Deer Lake d i l u t i o n water indicated that t o t a l trace metal concentrations i n natural water samples were so close to the lower detection l i m i t s for most metals as to be unreliable i n d i r e c t a s p i r a t i o n , and a concentration technique was used (Liptak, 1979). Detection l i m i t s for the investigated trace metals are presented i n Table 3-2. For each sample, a 250 mL aliquot was evaporated to 50 mL i n an a c i d environment (HNO3 and HC& 1), and vacuum f i l t e r e d to removed any p a r t i c u l a t e material. Concentration was c a r r i e d out within 24 hours of sample c o l l e c t i o n . 1 It was discovered that some zinc contamination was present i n the hydrochloric a c i d ; therefore this a c i d wasn't used i n the concentration technique for samples taken from s i t e four to s i t e twelve. 31. TABLE 3-2 DETECTION LIMITS AND ABSORPTION WAVELENGTHS FOR TRACE METALS IN STORMWATER RUNOFF1 Trace Metal Detection Limit (ug/L) Absorption Wavelengths (nm) Ca 500 422.7 Cd 1 228.8 Cr 50 357.9 Cu 1 324.7 Fe 50 248.3 Mg 50 285.2 Mn 10 279.8 Ni 1 232.0 Pb 1 283.3 Zn 1 213.8 1 A.P.H.A., 1974 A l l t o t a l trace metal analysis was done using the J a r r e l l Ash Model 810 Atomic Absorption Spectrophotometer, with the concentrated sample solutions aspirated d i r e c t l y and th e i r absorbances at the appropriate wavelengths recorded. Table 3-2 l i s t s these absorption wavelengths for the investigated metals. The absorption i n t e n s i t y for each metal was compared to those of the standard curves prepared using reagent grade metal solutions, and r e s u l t s reported as milligrams per l i t r e . Interferences were minimized using a non-absorbing l i n e (NAL) background c o r r e c t i o n . 3.2.5 Bioassay Materials and Procedures Procedures established i n the Environmental Engineering Lab, U.B.C, using daphnia as a bioassay organism (McDonald, unpublished) were chosen for t h i s research. a) D i l u t i o n Water: As recommended i n that report, water from Deer Lake (Burnaby, B.C.) was used as the d i l u t i o n water throughout the experimenta-t i o n , both because i t had been successfully used i n previous bioassay work, and because the background chemical composition was v i r t u a l l y unchanged over 32. the preceding two year period. The l e v e l s of the various physiochemical parameters of the water were low enough to provide a stable and, more import-antly, neutral d i l u t i o n water. Table A2 (Appendix A) shows the r e s u l t s of chemical analysis of Deer Lake water as determined i n previous studies and t h i s i n v e s t i g a t i o n (12/6/79 - 16/3/80). A l l parameters not described i n Section 3.2.4 were determined using methodology as det a i l e d by Standard Methods (A.P.H.A., 1978). Periodic spot checks throughout the experimenta-t i o n confirmed the a c c e p t a b i l i t y of the Deer Lake water for t h i s procedure. b) Choice of Daphnid Species: The two main daphnid species a v a i l a b l e for bioassay work are Daphnia magna and Daphnia pulex. Unfortunately, there i s no clear choice of one over the other for a l l bioassays, with the U.S. EPA recommending either one ( P e l t i e r , 1978) and Standard Methods recommending D. magna. For t h i s experimentation, D. pulex was chosen for a number of reasons: 1. a well established standing mass breeder set-up (with the organisms o r i g i n a l l y obtained from Deer Lake) was a v a i l a b l e i n the Environ-mental Engineering Lab, and the organisms were well acclimated to both the lab environment and the Deer Lake d i l u t i o n water over a period of many generations; 2. numerous other reports had made D. pulex the organism of choice; 3. while D. pulex i s not e n t i r e l y a freshwater r i v e r or lake organism, the r e s u l t s of bioassays using the species can be applied to the f i e l d with a high degree of confidence; 4. counting and set-up are easy with D. pulex; and 5. the organism i s r e a d i l y cultured i n the lab. c) Bioassay Procedure: Two to three weeks p r i o r to the s t a r t of bioassay work, one female was removed from the mass breeder and placed i n a 110 mL 3 3 . f l i n t - g l a s s j a r containing 100 mL f i l t e r e d Deer Lake water and s u f f i c i e n t food (100 pL d i l u t e powdered trout chow and 10 pL a l g a l nutrient s o l u t i o n ) . The f i r s t brood of 15-20 neonates was produced within three days, whereupon each neonate was transferred to i t s own prepared j a r . In t h i s manner, appro-ximately 150 i n d i v i d u a l breeding jars were established for use i n bioassay work. Upon production of a brood (approximately every t h i r d day) the neo-nates were removed, either for bioassay purposes or to be discarded; t h i s ensured that only parthenogenic production of female daphnids was taking place. A l l jars were checked d a i l y and cleared of neonates whenever neces-sary, to ensure that only those less than 24 hours old were a v a i l a b l e for bioassay work. A l l breeders were fed twice weekly. When a stormwater sample was ready for bioassay, a s u f f i c i e n t volume of Deer Lake water was vacuum f i l t e r e d through a Whatman #4 f i l t e r , to remove any suspended s o l i d s and indigenous aquatic species. From Table A2 (Appendix A), i t i s seen that there was a r e l a t i v e l y high concentration of t o t a l residue present i n the u n f i l t e r e d water; t h i s was mostly made up of suspended organic matter i n the form of duckweed and other algae. It was therefore necessary to f i l t e r the water to reduce t h i s organic load, to approximately 0 mg/L; the dissolved s o l i d s content of the water, as evidenced by the conductivity, (Table Al) was r e l a t i v e l y low before f i l t e r i n g (around 150 pS/cm), a value which compares favourably with the conductivity of most potable water (50 - 1500 pS/cm; A.P.H.A., 1974), and i t was therefore decided that the dissolved s o l i d s f r a c t i o n would be assumed to be of no s i g n i f i c a n c e to the o v e r a l l water chemistry. Only water c o l l e c t e d within the previous 36 hours was used i n t e s t i n g . The various s e r i a l d i l u t i o n s of stormwater, i n the series 1.0, 1.8, 3.2, 5.6 and 10 percent by volume (multiplied or divided by a factor of ten, as 34. necessary, dependent on the strength of the waste), and s u f f i c i e n t d i l u t i o n water to bring the t o t a l volume to 100 mL were prepared i n 110 mL wide-mouth round f l i n t - g l a s s j a r s and allowed to adjust to the ambient test temperature of 17±1°C. Due to the low t o x i c i t y of the stormwater, d i l u t i o n s of les s than 10 percent were r a r e l y necessary. Five j a r s containing 100 mL of f i l t e r e d d i l u t i o n water were used as controls for each sample bioassayed, and each d i l u t i o n was run i n t r i p l i c a t e . In a d d i t i o n , an extra j ar was prepared for the second lowest d i l u t i o n (second highest concentration) and the c o n t r o l , for each sample. These jars were used for pH and dissolved oxygen measure-ments at 48 and 96 hours, to monitor any i n t e r n a l changes i n the d i l u t i o n water. No neonates were added to these j a r s . Once the test d i l u t i o n s were prepared and allowed to adjust to ambient tes t chamber temperature, neonate daphnids les s than twenty-four hours old were removed from the i n d i v i d u a l breeding jars using a pasteur pipette with an approximately two millimeter f i r e - p o l i s h e d opening, and transferred to a beaker containing f i l t e r e d Deer Lake water, for the purpose of washing. A t o t a l of f i v e washed neonates were transferred to each test j a r . The trans-fer procedure was done as gently as possible to avoid putting undue stress on the organisms. The jars were then covered with a r i g i d p l a s t i c sheet with 1/8 inch holes d r i l l e d at 1/2 inch centres. After two hours the jars were examined and any f l o a t i n g daphnids were gently submerged using a pasteur pipette to d r i p water on top of the f l o a t e r . The organisms were not fed f o r the duration of the t e s t s . The l i g h t regime i n the test chamber was main-tained at the same 16 hour - 8 hour light-dark cycle that the i n d i v i d u a l breeders were under. The number of motile daphnids (those not immobilized, as defined by P e l t i e r [1978]) were counted and recorded at 24, 48 and 96 hours, and the 35. dissolved oxygen (using a ca l i b r a t e d YSI Model 54 Oxygen Meter with attached s t i r r i n g probe) and pH were measured and recorded on the extra jars at 48 and 96 hours. 3.3 Phase Two; Laboratory Bioassay Using Synthetic Stormwater As i t was the in t e n t i o n of t h i s research to investigate the types and eff e c t s of trace metals present i n urban stormwater runoff, and the i n t e r a c -t i v e e f f e c t s of sel e c t parameters on the t o x i c i t y of these metals, the author wanted, i n phase two, to create a "synthetic" stormwater comprising the worst-case metal concentrations observed i n phase one. This water was used i n a con t r o l l e d series of daphnia bioassays to observe the toxic e f f e c t s of the metals i n question by themselves and i n combination with each other, as well as i n combination with several other chemical parameters; namely pH, t o t a l sediment content, calcium hardness and hydrocarbon concentration. In thi s way, any sy n e r g i s t i c and/or antagonistic i n t e r a c t i o n s would be more r e a d i l y observed, and the parameters responsible f o r the t o x i c i t y found i n phase one i d e a l l y i d e n t i f i e d and explained. 3.3.1 Synthetic Stormwater Preparation a) Trace Metals: It was decided that the metals of most concern and there-fore those that would be used i n t h i s experimentation would include Cu, Fe, Pb and Zn (Ha l l et a l . , 1976). The worst-case concentration of each of these metals was determined from phase one, and standard solutions of each were made up using Fisher Atomic Absorption Stock Solutions (available i n concen-t r a t i o n s of 1 mg/mL). The stock metal solutions were d i l u t e d to the appro-pr i a t e concentrations with f i l t e r e d Deer Lake water (with minimal background trace metal content). These stock so l u t i o n s , with d i l u t e a c i d as the s o l -vent, lowered the pH of the standards, making i t necessary to bring the pH 36. back up to background d i l u t i o n water pH (7.0) by adding approximately 5N NaOH. The standard solutions were then l e f t for 24 hours at 17 °C to bring them to bioassay temperature, and tested for pH and metal content (to ensure that each had the desired concentration of trace metal and an acceptable test pH). These solutions were then used i n basic acute bioassay procedures (se c t i o n 3.2.5c), and i n d i v i d u a l LC^^'s determined for each trace metal. When the metals were to be tested i n combination ( i n p a i r s , t r i p l e t s and the one quadruplet, so that a l l combinations and permutations were tested), equal amounts of standard solutions were combined so that each of these new test solutions contained an equal concentration of trace metals ( i n a 1:1 r a t i o ) , and the i n d i v i d u a l trace metal concentration of t h i s new standard solution was at the desired l e v e l . It i s r e a l i z e d that t h i s might not be representative of the f i e l d s i t u a t i o n , i n that the metals could combine and exert influence i n r a t i o s other than a simple 1:1 (as governed by such f a c -tors as adsorption-desorption by the p a r t i c u l a t e matter, transport mechanics of the i n d i v i d u a l storm, e t c . ) ; however, as a f i r s t approximation of the i n t e r a c t i o n between d i f f e r e n t trace metals as far as t o x i c i t y i s concerned, the 1:1 r a t i o was judged to be v a l i d and u s e f u l . Research into varying con-cen t r a t i o n a l combinations of trace metals and t h e i r influence on the aquatic environment are beyond the scope of t h i s study. ^ < J Q v a l u e s were derived from the bioassays using these combined standard solutions, b) pH E f f e c t s on Trace Metal T o x i c i t y : In order to determine the e f f e c t s of pH on the t o x i c i t y of the trace metals being tested (possible increased/ decreased t o x i c i t y at higher and lower pH), the r e s u l t s from phase one were examined, and I t was decided that the trace metal combinations be tested i n the pH range 5.0 to 8.0, i n standard bioassay procedures. Due to time constraints, not a l l combinations of trace metals could be tested i n ass o c i a -37. t i o n with each pH; therefore only the combinations which showed highest t o x i -c i t y from the preliminary r e s u l t s were used. These combinations included zi n c , lead/zinc and lead/zinc/copper, and each of these were bioassayed at pH 5.0, 5.5, 6.0, 6.5, 7.5 and 8.0, at the determined concentration prepared as described i n Section 3.3.1a. The Deer Lake water used to bring the metal solutions to the desired volume was pH adjusted to test pH before stock metal solutions (1 mg/mL) were added. The e n t i r e standard s o l u t i o n was then readjusted to the desired test pH. This was done i n order to reduce the t o t a l volume of a c i d or base needed for pH adjustment and thereby reduce the chances of the acid/base exerting toxic e f f e c t s by i t s e l f . The controls consisted of daphnids i n the pre-adjusted Deer Lake d i l u t i o n water i n order to observe any e f f e c t s of the lowered/raised pH by i t s e l f , with no metal i n t e r a c t i o n s . As well, one j a r of normal Deer Lake water (pH 7.0) containing f i v e daphnids was run with each batch so that any natural population death could be observed. The pH and D.O. were measured on a l l jars at 24, 48 and 96 hours. LC,.„ values were obtained from the J ' 50 subsequent mortality curves. c) Sediment E f f e c t s on Trace Metal T o x i c i t y : The e f f e c t of suspended and se t t l e d s o l i d matter on the t o x i c i t y of trace metals was investigated i n t h i s section of experimentation. From the r e s u l t s of phase one i t was decided that the trace metal combinations zinc, lead/zinc and copper/lead/zinc would be bioassayed i n a s s o c i a t i o n with 50, 100 and 200 mg/L t o t a l s o l i d s i n the form of diatom f i l t e r powder (Vortex Products). This diatomaceous earth was found to be non-toxic by i t s e l f i n preliminary bioassays, and was used because i t approximated natural suspended s o l i d s i n i t s a b i l i t y to exchange m e t a l l i c ions. Other s o l i d s considered for use included k a o l i n i t e and montmorillonite clays, but these were judged to be too unrepresentative of 38. the s o l i d s one would expect to fi n d i n urban runoff (Wilber and Hunter, 1979), as well as having too high a cation exchange capacity for use i n these experiments (Pettigrew, personal communication). The metal solutions were made up as before (Section 3.3a) and i t was found that the diatomaceous earth was best added d i r e c t l y to the test j a r s ; therefore, s e r i a l d i l u t i o n s of the trace metal standards were added to the j a r s , and enough diatomaceous earth weighed out such that 5.0, 10.0 and 20.0 mg were added per 100 mL t o t a l s o l u t i o n . The controls contained no diatoma-ceous earth. Standard bioassay procedure was followed. d) Calcium Hardness E f f e c t s on Trace Metal T o x i c i t y : It was determined from e a r l i e r work that calcium hardness comprised 80 to 90% of the t o t a l hardness of a t y p i c a l stormwater sample, and therefore only calcium hardness e f f e c t s on trace metal t o x i c i t y were investigated. As well, the manufacture of a calcium hardness standard solu t i o n , using r e a d i l y a v a i l a b l e CaC03, could be done more e a s i l y and pr e c i s e l y than making a t o t a l hardness standard comprising both CaC03 and Mg(0H)2 i n varying proportions. Stock Fisher calcium carbonate (lmg/mL) was t o t a l l y dissolved i n f i l t e r e d Deer Lake water to make up standard hardness solutions of 300, 200, 100 and 50 mg/L hardness, as CaC03. The background Deer Lake d i l u t i o n water hardness of 40 mg/L as CaC03 was taken i n t o account i n the d i l u t i o n c a l c u l a t i o n s . Standard bioassay procedure was followed, testing these hard-ness solutions i n as s o c i a t i o n with the trace metal combinations Zn, Pb/Zn and Cu/Pb/Zn, and lC,-Q values derived. e) Hydrocarbon E f f e c t s on Trace Metal T o x i c i t y : V i s u a l observation of stormwater runoff suggested that a high volume of gasoline, and other petro-leum products were being washed off the street surface during a r a i n f a l l event; for t h i s reason the ef f e c t s of hydrocarbons i n the form of iso-octane 3 9 . (2,2,4-trimethylpentane, a major component of gasoline) on trace metal t o x i -c i t y were investigated. Using Fisher grade Iso-octane i n a solvent of carbon t e t r a c h l o r i d e , standard solutions of 2, 5 and 10 mg/L were prepared. From preliminary experimentation i t was found that the iso-octane was best added d i r e c t l y to the bioassay j a r s , which avoided the su b s t a n t i a l losses encountered i n the tr a n s f e r r i n g and d i l u t i n g of manufactured standards i n volumetric f l a s k s . The hydrocarbon was added d i r e c t l y using micropipettes of the necessary volumes to f i l t e r e d Deer Lake water. During the bioassays i t was observed that the iso-octane tended to f l o a t on the water surface and was thus e f f e c t i v e l y removed from contact with the neonate daphnids; to encourage mixing, therefore (to simulate natural turbu-lent conditions), duplicates of each test bioassay were run, with one set unmixed and the other mixed at slow speed using magnetic s t i r r i n g bars and mixing pl a t e s . Again, the trace metal combinations used i n association with the hydrocarbons were Zn, Pb/Zn and Cu/Pb/Zn. f) Combined E f f e c t s on Trace Metal T o x i c i t y : It was decided to manufacture a synthetic stormwater containing a l l of the previously described chemical properties i n an e f f o r t to duplicate, as c l o s e l y as possible, the natural stormwater samples of phase one. This synthetic stormwater was made up of the worst-case trace metals i n combination with varying pH and t o t a l s o l i d s . Due to time constraints, only the trace metal combination Pb/Zn was used. The following combinations were manufactured and bioassayed using standard procedure: Pb/Zn at pH 5.0 with 200, 100 and 50 mg/L s o l i d s ; Pb/Zn at pH 6.0 with 200, 100 and 50 mg/L s o l i d s ; and Pb/Zn at pH 8.0 with 200, 100 and 50 mg/L s o l i d s , a l l made up using the previously described procedures. Ninety-s i x hour L C ^ Q values were obtained for a l l combinations. Pb/Zn trace metal concentrations were as before. 40. 3.4 Phase Three: "First-Flush" Analysis Much of the literature reviewed suggests the presence of a " f i r s t - f l u s h " phenomenon in urban stormwater runoff, a process whereby most pollutants may be washed off the street surface in the f i r s t hour of r a i n f a l l runoff (Wilber and Hunter, 1977; Kaufman and Lai, 1981). Obviously this slug load of pollu-tants, i f present, might be expected to have adverse effects on aquatic organisms in the receiving body, but i t is extremely d i f f i c u l t to quantify these effects. The f i r s t step should be the verification of the " f i r s t -flush", and the confirmation of i t s higher contaminational threat when integrated with the remainder of the storm. It was decided that phase three of this research would entail a study of one sampling site through the duration of one storm event. This would help to qualify the phases of the storm, and confirm or deny the existance of the " f i r s t - f l u s h " and i t s potential environmental effects. 3.4.1 Sampling Site The results from phase one were reviewed and the sampling site with the highest pollutional load in terms of trace metals and the other investigated parameters was chosen. Unfortunately, the storm sewer at this site (Lougheed Mall; Station 4) had been covered with wood chips and peat moss being used in landscaping. As i t would have been impossible to obtain a representative stormwater sample from this site, i t was decided that the next worst-case site (Brentwood Mall; Station 1) would be used for this study. 3.4.2 Sampling Equipment and Operation Use of the Isco automated water samples was suggested as the sampling mechanism for this research; however, having used the device in phase one sampling, i t was seen that the sampler took a long time (up to 10 minutes) to 41. c o l l e c t a s i n g l e stormwater sample and as such was i n e f f e c t c o l l e c t i n g an integrated sample over a period of time, as opposed to the desired i n s t a n -taneous grab sample. While t h i s composite sample was s u f f i c i e n t for the preliminary analyses performed i n phase one, i t was decided that a grab sample would be more useful for our purposes i n phase three. A series of grab samples would allow rapid parameter changes during the course of the storm to be seen. Therefore, each sample was obtained using an acid-washed and rinsed p l a s t i c four gallon bucket, submerged into the stormsewer at prescribed times. The stormwater was then transferred to i n d i v i d u a l a c i d -washed 500 mL wide-mouth capped p l a s t i c b o t t l e s . The volume obtained with t h i s bucket from one grab sample was s u f f i c i e n t f o r a l l chemical tests and bioassays performed. 3.4.3 Sampling and Lab Testing It was determined that the sample storm should be four to f i v e hours i n length, and sampling be done every twenty minutes. The samples were trans-ported to the lab immediately a f t e r the storm ended, and testing procedures begun. A l l samples not analysed within twenty-four hours of c o l l e c t i o n were a c i d i f i e d with HC£ • S u f f i c i e n t sample volume was obtained to provide an "as i s " sample ( u n f i l t e r e d ) a " f i l t e r e d " sample, and, once i n each hour, an "acid" sample, for each sampling period over the duration of the storm. A l l samples were analyzed for the parameters described i n Section 3.2.4; as w e l l , bioassays were run on a l l samples. The "as i s " samples were run without being f i l t e r e d , to give some idea of t o t a l metal concentrations; the " f i l t e r e d " samples were vacuum f i l t e r e d through Whatman #4 paper f i l t e r s , to give dissolved ( n o n - f i l t r a b l e ) metal concentrations, and the "acid" samples were a c i d i f i e d to pH3 with HC£, vacuum f i l t e r e d , and readjusted to t h e i r o r i g i n a l pH with NaOH, to quantify the acid extractable metal f r a c t i o n . 42. 3.5 St a t i s t i c a l Procedures 3.5.1 LCsn Determination As stated by Sprague (1973) and others, the objective of the acute (short-term) lethal bioassay i s to estimate the incipient lethal level, the concentration at which acute toxicity ceases. This lethal threshold concen-trations (incipient LC 5 0) i s that level of contaminant beyond which 50% of the population cannot live for an indefinite time (constrained, of course, by natural population dynamics). The basic procedure i s to choose a logarithmic series of concentrations spanning the range from 0 to 100% mortality, and expose the animal for a sufficiently long period of time to determine a threshold concentration. Percent mortality within each test chamber i s noted at a logarithmic series of observation times in order to obtain a toxicity curve at each observation time for the various toxicant test concentrations. These graphs are then combined into a toxicity curve for estimation of the incipient LC 5Q. This i s the basic procedure for the quantitative or time base method of acute bioassay. The method used in this research i s the quantal method (Brown, 1973) in which mortality of the test organism i s recorded at fixed times (eg. 24, 48 and 96 hours), from a series of logarithmic serial dilutions, and the LG\-0 derived by interpolation from the resultant curve of % mortality vs. log % effluent concentration. This method has the advantage that confidence inter-vals can be calculated in terms of concentration rather than time, but has the disadvantage that i f most of the test organisms die, much precision i s lost by not measuring survival time. Numerous authors have suggested that quantal data requires many more observations than quantitative, to obtain a given level of accuracy (Gaddum, 1953; Burdick, 1960). Sprague (1969) suggests sampling more frequently i n quantal tests and recording the time to 4 3 . 50% mortality, which he claims should make the data equally useful. In this research, the value of LC 5 0 for each composite stormwater sample was obtained directly from the curve of % mortality vs. log % effluent concentration, and the toxicity curve, as recommended above, was not drawn. This was done for a number of reasons: 1) the toxicity curve i s derived from a number of replicates of an effluent test concentration at various exposure times: i n this research, sufficient stormwater volumes could not be obtained from each storm to carry out these test replicates, and therefore each sample was tested at different concentrations over a set time period (96 hours); and 2) more importantly, the toxicity curve i s very useful i n delineating modes of toxicity and different pathways of action with increasing time i.e., a toxicant may have different effects at 24 hours than at 96 hours, and the slope of the toxicity curve w i l l readily reveal this. While these modes of action are important to recognize, i t was the desired purpose of this research to look at general toxicity and the extent to which various external factors (land use, trace metal presence and concentration, e t c ) affect and influence this toxicity. Therefore, this research can be looked at as the f i r s t step in a detailed study at stormwater runoff toxicity, and as such, only general trends are dealt with. The need for more detailed toxicity work is obvious, but i s not the stated purpose of this research, and as such, the preliminary LC 5 0 values are deemed more than adequate for general comparitive purposes. Finally, Sprague (1970) states that acute toxicity seems somewhat less important for invertebrates than such subtle effects as delayed responses, inhibited growth, egg sensitivity and behavioural deviations which render the animal more sensitive to predation. He suggests that sublethal responses of invertebrate larval forms have a good potential for correlating well with 44. chronic fish bioassays. These suggestions, while valid in nature, again represent work beyond the scope of this preliminary research, and would be considered i n a more detailed examination of stormwater toxicity. The author i s less interested here i n modes and subtleties of toxicant action than the evidence of a toxic response existing in the f i r s t place. 3.5.2 U.B.C. TRP An analysis of stormwater toxicity and i t s causes was carried out using the computer program U.B.C. TRP (Triangular Regression Package; Le and Tenisci, 1977). This package i s able to provide partial correlation coeffi-cients for a l l chemical parameters tested against toxicity, through the use of the program INMSDC. In addition, the program STPREG i s used to generate multiple regression equations for toxicity as a function of a l l s t a t i s t i c a l l y significant parameters. STPREG performs stepwise forward and backward regression analysis and in the process eliminates those independent variables which make no significant contribution to the dependent variable at the pro-bability level specified (usually 95%). The output of STPREG includes R2, standard error, F-probability of the overall regression, standard error of the coefficients of the equations and F-probability of a l l independent va r i -ables. The TRP manual describes these s t a t i s t i c a l parameters as follows: R2 (Standard Multiple Regression Coefficient) i s always between zero and one; the closer i t i s to one, the better the regression equations f i t the data. R2 i s the proportion of the variance of the dependent variable which i s accounted for by the regression line; F-Probability i s the likelihood of obtaining a value of R2 greater than or equal to the one calculated, given that there i s no association between the dependent and independent variables. If this value i s less than 0.05, i t i s concluded that R2 i s significantly different from zero; and 45. Standard Error of Dependent Variables. The square of this value i s an estimate of the variance of the dependent variable about the regression hyperplane. 3.5.3 Marking-Dawson Additivity Index (Marking and Dawson, 1975) This index, used to analyze the bioassay results from phase two, allows one to evaluate the contribution each component of a chemical mixture makes to the overall toxicity of the mixture, and whether the chemicals i n combination are more or less toxic than individually. The effective contributions of each chemical (A and B) of a mixture are represented by where A and B are the chemicals in question, i and m are the toxicities (LC^Q'S) of the individual chemicals and mixtures, respectively, and S i s the sum of the biological activity of the chemical mixture, as measured by the toxicity of this mixture to the test organism. If the sum of toxicity of the chemicals i s simply additive, S = 1.0; sums that are less than 1.0 indicate greater-than-additive toxicity (synergism between the chemicals), and sums greater than 1.0 indicate less-than-additive toxicity (antagonism). This sum S could function as a quantitative indication of additive toxicity, except that values greater than 1.0 are not linear with values less than 1.0. However, l i n e r i t y can be established and a zero, positive and negative additivity index derived by the use of a modifying factor for S. This factor i s either: Am Bm Ai Bi (3.1) for S< 1.0 or (3.2) S(-l) + 1.0 for S > 1.0 (3.3) 46. A sum S of 1.0 yields an index value of zero by either modifying procedure and represents simple additive toxicity. Figure 3-5 presents these indices in a readily understood form. Please note that these indices are easily adaptable for chemical mixtures of more than two components; in this case, * TH Am Bm Cm + Am ( 3 # 4 ) AT Bi Ci M T H and the use of the additivity index i s the same. GREATER THAN ADDITIVE TOXICITY LESS THAN ADDITIVE TOXICITY 1.0 2.0 3.0 4.0 SUM OF TOXIC CONTRIBUTION 5.0 S(-1I+1 = LESS THAN ADDITIVE TOXICITY GREATER THAN ADDITIVE TOXICITY 2-0 -1.0 0 1.0 2.0 CORRECTED SUM OF TOXIC CONTRIBUTIONS 3.0 Figure 3-5 Marking-Dawson A d d i t i v i t y Index 48. CHAPTER 4  Results and Discussion 4.1 Phase One: Introduction In order to f a c i l i t a t e the presentation of r e s u l t s , research and analy-s i s data have been summarized i n several tables and figures, to which the text w i l l r e f e r . General physical-chemical properties of the composite stormwater samples obtained from the preliminary sampling s i t e s are shown i n Table 4-1, while t o t a l trace metal data for these samples i s contained i n Table 4-2. In an attempt to indicate areas of trace metal buildup i n storm-water within the Brunette River basin and the r e l a t i o n s h i p of these metals to land use, values for cadmium, chromium, copper, i r o n , manganese, n i c k e l , lead and zinc were divided into ranges of concentration and plotted on maps of the drainage basin (Figures 4-4 to 4-11). With the exception of i r o n and manga-nese, these, along with mercury are the trace metals most often considered toxic to aquatic organisms i n th e i r soluble form (F.W.P.C.F., 1968). The preliminary metals data were used to compare the r e l a t i v e degree of trace metal contamination i n the basin, as a function of the d i f f e r e n t land use c l a s s i f i c a t i o n s sampled; namely i n d u s t r i a l , commercial, r e s i d e n t i a l and open/greenspace. For purposes of comparison, the open/greenspace category was used for background trace metal values. A graphic representation of t h i s land use contamination i s presented i n Figure 4-12, with a s t a t i s t i c a l analy-s i s i n Tables 4-5 and 4-6. Bioassay data i s included i n Table 4-8 as the 96 hour LC^^ value (%), the acute t o x i c i t y of the stormwater samples to D. pulex, and Figure 4-13 shows these comparative t o x i c i t i e s plotted on a map of the drainage basin. 4 9 . As well, results of the s t a t i s t i c a l analysis of toxicity as a function of the other parameters, using computer programs from U.B.C. TRP (Le and Tenisci, 1978) are presented. 4.1.1 General Parameters1 Values for the investigated physiochemical properties of the stormwater samples taken from the 12 preliminary sampling sites are presented i n Table 4-1. These include chemical oxygen demand (COD), pH, total suspended solids, total alkalinity, total and calcium hardness, measured and diluted conduc-tance and hydrocarbons (as iso-octane). For purposes of comparison, Table A2 (Appendix A) contains the results of the chemical analysis of Deer Lake water, used as the dilution water for a l l bioassays undertaken i n this research (essentially a "clean", unpolluted water), and Appendix B contains the results of stormwater quality analyses as reported elsewhere in the literature. Comparing these values to the ones obtained i n this research gives some indication of the relative quality of stormwater i n the Brunette River basin. 1 Please note that the multiple regression component of the TRP program was used to relate these parameters to the observed toxicity values of the bioassay work (sections 4.1.3 and 4.2), i n an attempt to better understand and determine the causes of toxicity of stormwater to D. pulex. Therefore, this section w i l l only be dealing with general observations about the physiochemical properties of the stormwater. TABLE 4-1 PHASE ONE CHEMICAL ANALYSIS RESULTS N U M B E R STATION NAME SAMPLING DATE C L A S S BUILDUP TIME 2 i C O D V N A L Y T K PH 3 : A L P A R A M E I T.S.S. C E R S 1 T O T A L ALKALINITY4 T O T A L HARDNESS11 CALCIUM HARDNESS11 C O N D U C T A N C E 5 MEASURED DILLUTED C O N D . 5 H Y D R O C A R B O N S 6 1 Brentwood Mall, S.W. corner 15/AUG/79 C 33 1030.8 5.70 129.00 52.4 327.0 284.0 496.0 523.2 6.66 2 Beaverbrook Drive at Beaverbrook Cresent 26/SEPT/79 R 18 763.1 5.80 64.10 33.6 142.4 112.0 512.3 1336.2 5.80 3 Production Way at Thunderblrd Cresent 14/OCT/79 I 17 269.0 6.00 16.50 63.0 81.6 71.6 95.6 852.8 4.62 4 Lougheed Mall 24/OCT/79 C 1.5 108.0 5.82 55.50 12.0 12.2 11.8 14.6 139.4 2.03 5 Wedgewood Avenue at 1st Avenue 21/NOV/79 R 4 66.8 6.80 15.32 33.0 28.8 23.0 130.0 170.9 2.35 6 Robert Burnaby Park 21/NOV/79 0 4 69.9 6.55 21.81 12.0 14.8 14.0 83.6 102.9 1.79 7 Deer Lake Park 21/NOV/79 0 4 77.9 5.82 75.56 8.4 16.8 18.0 58.2 296.3 3.16 8 Haszard St. at Buckingham Avenue 21/NOV/79 R 4 98.8 6.28 122.16 20.0 19.2 20.0 108.8 102.4 2.57 9 Warner Loat Park 15/DEC/79 0 7 45.9 6.96 151.11 22.4 54.4 53.8 73.2 78.4 1.82 10 Brunette Avenue at Capllano Street 31/JAN/80 I 15 429.9 8.10 1867.72 122.0 84.8 66.4 469.6 1097.7 9.17 11 Alpha Way at Alaska Street 31/JAN/80 I 15 405.6 7.52 443.65 64.0 92.0 81.0 692.6 805.0 8.84 12 Boundary Road at Canada Way 31/JAN/80 C 15 168.8 7.34 32.71 154.0 142.0 114.4 587.5 662.5 4.33 1 Values i n mg/L, except where otherwise indicated 2 Values i n Days 3 Values unltless 4 Values i n mg/L as CaC03 5 Values i n us-cnT 1 6 Values i n mg/L Hydrocarbon as Iso-octane 51. a) COD: Values ranged from 46 mg/L at station nine, Warner Loat Park (which would be representative of a background value since this park i s essentially isolated from the influence of the adjacent semi-industrial buildup), to a high of more than 1000 mg/L at station one, the commercial Brentwood Mall area. This relatively high value for the commercial area might simply be a function of the unusually long buildup time that occurred before sampling was undertaken (33 days, compared to an average dry period buildup time of 11.4 days, or 9.5 days i f this outlier value i s excluded). Given this length of dry period, we could expect greater deposition of oxygen demanding carbonaceous organic and inorganic material ( i . e . , from the heavy vehicular t r a f f i c i n the area, debris and detritus deposited on the roadways and sidewalks, etc.); this in turn would lead to a higher observed COD. This i s further Indicated when we examine the COD-buildup time relationship at the other sites within the different land use classifications. Figure 4-1 suggests that there i s an approximately linear relationship between COD and buildup time, i.e., the more antecedent dry days before sampling, (allowing for greater pollutant buildup), the higher w i l l be the COD. This i s most easily observed in the commercial and residential areas, while the industrial areas suggest the possibility of an inverse relationship between the two parameters. It may be reasonably assumed that this i s primarily due to a lack of replicates within this land-use category, and that the relationship i s essentially the same in a l l cases. In further comparison between the different land use areas, the commer-c i a l sites had the highest mean COD (436 mg/L) followed by the industrial sites (368 mg/L), with the residential sites having the lowest (310 mg/L) as compared with the background open space site mean value of 65 mg/L. However, i f these COD values are compared in relation to buildup time, using an 52. Figure 4-1 C0D-3uildup Time R e l a t i o n s h i p , A l l S i t e s . 53. a r b i t r a r y " l o a d i n g " value of COD buildup per day (mg/L-day), i t i s found that the r e s i d e n t i a l areas have the highest "l o a d i n g " (36 mg/L-day) with the i n d u s t r i a l and commercial areas having e s s e n t i a l l y the same load values (23 and 26 mg/L-day r e s p e c t i v e l y ) . This would seem to suggest a higher presence of COD-contributing substances i n the r e s i d e n t i a l areas, which could be due to the greater amount of v e g e t a l and a s s o c i a t e d matter i n these areas ( i n the form of l e a v e s , grasses, p l a n t s , f e r t i l i z e r s , etc.) when compared with the i n d u s t r i a l and commercial areas, both of which are e s s e n t i a l l y c l e a r e d of t r e e s , bushes, e t c . ) , and t h e r e f o r e do not have t h i s source of COD load to as great a degree. From t h i s , a higher COD i n the runoff of r e s i d e n t i a l areas could be expected, d e s p i t e the absence of heavy v e h i c u l a r t r a f f i c . F o l lowing t h i s argument, i t might be suggested t h a t , because the open/ greenspace areas have the highest amount of v e g e t a l matter of any of the land use c l a s s e s , we should expect these areas to c o n t r i b u t e the g r e a t e s t amount of v e g e t a l a s s o c i a t e d COD to stormwater r u n o f f . However, Rimer et a l . , (1978), from t h e i r study of the c h a r a c t e r i z a t i o n and impact of stormwater runoff from d i f f e r e n t land use types, suggested that the COD l e v e l s are g e n e r a l l y a f u n c t i o n of the amount of development and the percentage of impervious surface w i t h i n the catchment area being drained. This study confirmed that i n c r e a s i n g values of COD c o n c e n t r a t i o n can be expected from land cover types which have an i n c r e a s i n g percent of impervious s u r f a c e . This i n c r e a s i n g COD c o n c e n t r a t i o n would a l s o be a f u n c t i o n of the amount of oxygen demanding source m a t e r i a l present i n the a r e a . For example, a green-space s i t e , t o t a l l y l a c k i n g i n paved a r e a , a c t s as a b u f f e r a gainst storm-water flow through p h y s i c a l r e s t r a i n t , p e r c o l a t i o n , i n f i l t r a t i o n , e t c . As such very l i t t l e of the otherwise abundance oxygen demanding v e g e t a l matter of the area would be expected to c o n t r i b u t e to COD under normal stormwater 5 4 . flow conditions. In the same respect, an almost totally paved industrial area would offer excellent flow characteristics, but because there would be l i t t l e source material present, the expected COD levels i n stormwater from these areas would again be low. Therefore, the residential areas offer the proper conditions of percentage impervious surface and amount of oxygen demanding vegetal matter, to make a high contribution to stormwater COD load. These results would seem to support this. A review of the literature (Appendix B) suggests stormwater COD values ranging from 12 mg/L (Tulsa, Oklahoma, 1968-69) to as high as 3100 mg/L (Stockholm, Sweden). It would appear that the mean COD values found i n this study are more or less average for stormwater, but the high value of 1031 mg/L i s somewhat greater than average. b) pH: Values ranged from a low of 5.70, again at Brentwood Mall (site 1) and again probably due to the extremely long buildup period before sampling, to a high of 8.10 at the industrial site 10, situated at the mouth of the Brunette River, close to a lumber processing plant and a brewery. This i s not a wide range of pH and well within the tolerance range of most freshwater aquatic organisms. The mean pH values for the four land-use classes a l l f a l l within the range 6.56 ± 0.44, again not a very large range and easily within tolerance limits for the bioassay test species (A.P.H.A., 1978). These values are quite comparable to those found i n the literature (Appendix B) and, from Table Al, the pH range of Deer Lake water over approximately a two year period i s 7.26 ± 0.46, which also compares favourably with the mean storm-water pH. c) Total Suspended Solids: Most of the solids measured as T.S.S. were medium sized sediment particles (~0.4 mm), usually in the form of fine g r i t 55. or sand, which were readily suspended under moderate mixing, (ie., the turbu-lent flow usually associated with stormwater runoff) but which settled out of the water column within approximately 12 hours under quiescent conditions. In addition to this sediment, other forms of solid materials were common to the runoff flow i n the stormsewer area; these were usually organic and/or detrital i n nature, i.e., wood chips from landscaping a c t i v i t i e s , pieces of leaves, etc.; however, nonorganic material such as cigarette butts and packs, f i l t e r s , etc. were not uncommon in some areas. Care was taken during samp-ling to exclude any such materials from the stormwater obtained. Most of the sampling areas had stormwater T.S.S. concentrations of 100 mg/L or less, with only two sites having concentrations greater than 150 mg/L. The solids content does not appear to be related to buildup time, but i n one case i s directly related to the industrial activity of the area. At site 10, near the lumber yard, the area i s mostly unpaved and lacking i n road curbs; because of this, storm runoff i s undirected, flowing over a large area of dirt roads. As such the runoff i s able to suspend a great deal of sedi-ment i n the flow. This was evidenced by the fact that the stormsewer from which samples were taken was at least two-thirds f i l l e d with sediment, which was acting as a natural dam or barrier to stormwater flow through the sewer system. The samples taken at this site were obtained by directing the sur-face runoff flow into the sampling containers, with care being taken to avoid sampling from the inside the stormsewer proper. In this way, a more repre-sentative sample was obtained from this area. The solids content also does not appear to be highly related to land use, with the commercial, residential and open areas a l l having mean T.S.S. concentrations of approximately 75 ± 8 mg/L. It might be expected that the 5 6 . open/greenspace sites have runoff with a much higher solids content, but as explained in Section 4.1.1a, the vegetal matter of the area functions as a barrier to stormwater flow, and acts as a f i l t e r for the pollutants being carried in the runoff. The residential and commmercial areas, having a high percentage of impervious land surface, would not be expected to contribute much suspended sediment to the stormwater flow due to a general lack of sedi-ment sources, and this appears to be the case. Only the industrial areas appear to be contributing a high sediment load, and this can be explained by the mixture of covered and deforested uncovered land surface in these samp-ling areas. The uncovered areas would be contributing a large amount of sediment to the stormwater flow, and the impervious roadways would act to channel this sediment-laden runoff to the stormsewers. This is only a gener-alization and might not apply in a l l cases; however, i t might be expected that with a higher percentage of uncovered, unforested (cleared) land in the area, the higher w i l l be the sediment load to the stormwater flow. A review of the literature (Appendix B) indicates that these values are more or less average and could be expected in a heavily populated city area, d) Alkalinity, Hardness and Conductance: From Table 4-1 the alkalinity of the stormwater samples is low at most sites (mean value 50 ± 46 mg/L as CaC03), with the highest value occuring at site 12, the parking lot of a small urban shopping centre, and the lowest found in Deer Lake Park, site seven. Mean values indicate the highest alkalinity occurs at the commercial and industrial sites (73 and 83 mg/L as CaC03, respectively), while the resi-dential areas have a much lower mean alkalinity (29 mg/L as CaC03). The open areas, as expected, show the lowest mean value as a group as well as individual sites. These alkalinity measurements, in general, are very low, but i t must 5 7 . be kept i n mind that they are indicative of alkalinity contributed mainly by the carbonate-bicarbonate-hydroxide system (although any buffering component within the pH ti t r a t i o n range i s actually measured i n the titration); i n this case, with the pH being between 5.7 and 8.1, the alkalinity i s assumed to be due to the bicarbonate species (HCO3 -). An examination of stormwater hardness shows that, at every site but two, the hardness of the samples far exceeded the alkalinity (maximum 327, minimum 12.2, mean 144 - 90, a l l measured as mg/L CaC03). This suggests the presence of excess non-carbonate hardness, contributed by other chemical species. A detailed analysis would have to be done to determine exactly what those species are, but they may include such anions as Cl" and S0h~2 (A.P.H.A., 1978). Thus, alk a l i n i t y measurements can only give limited information about the presence of anionic species in a sample. The presence of these charged species i s further suggested by an exami-nation of the conductance of the stormwater samples. As suggested by Standard Methods (A.P.H.A., 1978), the concentration of the dissolved ionic matter in a sample may often be estimated by multiplying the conductivity (in ymhos/cm) by an empirical factor. This factor may range from 0.55 to 0.90, depending on the soluble components of the water and the temperature of the measurement. Relatively high factors may be required for saline or boiler waters, whereas lower factors may apply where considerable hydroxide or free acid i s present. In addition to this more or less general conversion factor, an approximation of the milliequivalents per l i t r e of either anions or cations in some water samples may be obtained by multiplying the conductance (in umhos/cm) by 0.01. In either case, an estimate of the total dissolved ionic content of a water sample can be made with a relative degree of certainty, using measured conductance values. 58. The mean conductance of a l l samples i s 514 ± 314 uS/cm, a r e l a t i v e l y high value, which indicates the presence of considerable amounts of dissolved i o n i c material, depending on the empirical conversion factor used. The i n d u s t r i a l land use s i t e s had the highest mean conductance at 919 pS/cm, while the commercial and r e s i d e n t i a l areas again had comparable values, at 442 and 537 uS/cm r e s p e c t i v e l y . The open land use s i t e s had mean conductance values of 159 uS/cm, seemingly i n d i c a t i v e of a background area. While no chemical analysis was done to determine the i d e n t i t y of a l l dissolved i o n i c species, t h e i r existence i n concentrations r i v a l l i n g those of the t o t a l suspended s o l i d s must be noted, as well as th e i r possible contributions to stormwater t o x i c i t y . Figure 4-2 shows the r e l a t i o n s h i p between conductance, t o t a l a l k a l i n i t y and t o t a l harness i n terms of buildup time. It can be seen that a l i n e a r r e l a t i o n s h i p e x i s t s for a l l three parameters. There are only a few points that do not f i t the established r e l a t i o n s h i p s , most notably those of samples c o l l e c t e d from Brentwood Mall (Station one) a f t e r the 33 day buildup period. It i s evident that, because t h i s dry period was so unusually long, many of the sample parameters do not, upon a n a l y s i s , conform to deduced patterns; nevertheless, the data i s s t i l l generally valuable, i f only to convey i n f o r -mation about the e f f e c t s of very infrequent, unusually long dry periods. This l i n e a r r e l a t i o n s h i p does not take into account the s p e c i f i c land use of each sampling s i t e , and a comparison of the data on that basis reveals no obvious patterns. It would appear, therefore that for these three para-meters, as with COD, buildup time i s more i n f l u e n t i a l on stormwater q u a l i t y than i s land use. 59. 0 0 10 20 30 40 BUILDUP TIME (DAYS) Figure 4-2 Conductance, To t a l A l k a l i n i t y and Total Hardness Vs. Buildup Time, A l l S i t e s . 60. e) Hydrocarbons: Values ranged from a minimum of 1.79 mg/L at Robert Burnaby Park, to a maximum value of 9.17 mg/L at s i t e 10, the heavy indus-t r i a l s i t e at the mouth of the Fraser River. A l l hydrocarbon values are reported as mg/L as iso-octane, i n an attempt to r e l a t e the concentration of t h i s petroleum product to the type of automotive a c t i v i t y at the p a r t i c u l a r sampling s i t e s . Mean hydrocarbon values seem to r e f l e c t t h i s a c t i v i t y , i n that the i n d u s t r i a l s i t e s have the highest mean hydrocarbon concentrations (7.54 mg/L), with the commercial s i t e s second highest at 4.34 mg/L. The r e s i d e n t i a l s i t e s are lower s t i l l (3.57 mg/L) and the open s i t e s again show background "clean area" values with a mean concentration of 2.26 mg/L. The commercial s i t e mean hydrocarbon values may be r e f l e c t i v e of the pattern of t r a f f i c i n those areas, with not so much a high volume as a long residence time. For example, cars would be parked at shopping malls, or s i t t i n g i d l i n g for long periods of time, as well as s t a r t i n g up more often than i n other areas; this longer residence time would lead to a high concentration of com-bustion waste products i n the area, as well as allow increased leakage from o i l pans, gas tanks, etc. The same mechanism might apply to the r e s i d e n t i a l areas, but with a somewhat reduced t r a f f i c volume as well as reduced i d l i n g time and start-up frequency, r e s u l t i n g i n a lower o v e r a l l hydrocarbon concen-t r a t i o n . The open areas, with their low t r a f f i c volumes, would be expected to have r e l a t i v e l y low hydrocarbon concentrations, as i s the case. T r a f f i c flow alone, however, doesn't seem to explain the higher values found i n the i n d u s t r i a l areas which, due to th e i r r e l a t i v e i s o l a t i o n from the major public thoroughfares, would be expected to have even less t r a f f i c than the residen-t i a l areas. It might be assumed that the character of the i n d u s t r i a l t r a f f i c , i . e . , large trucks and heavy machinery, could be having some impact on stormwater hydrocarbon content, but further research i s indicated. Figure 4-3 shows that, with the exception of two sites (10 and 11), there i s a good linear relationship between hydrocarbon concentration and buildup time; this would follow from the argument that with longer dry periods, the t r a f f i c of a given area would contribute more to hydrocarbon buildup through deposition. As well, the different land uses can be gener-al l y characterized on the graph. The higher and middle values are from the commercial and residential sites, and the lower values (l e f t side of figure) are from the open/greenspace sampling sites. The fact that the values from the two industrial sites do not f i t the relationship seems to suggest a secondary source of hydrocarbon input to stormwater runoff i n these areas, perhaps i n the form of leakage from storage tanks, mechanical equipment, dumping of service waste products, etc. (Hunter et a l . , 1979). 4.1.2 Trace Metal Contamination i n Stormwater i n the Brunette Basin A l l composite and grab samples were analyzed for the trace metals cad-mium (Cd), chromium (Cr), copper (Cu), iron (Fe), manganese (Mn), nickel (Ni), lead (Pb) and zinc (Zn), as well as calcium (Ca) and magnesium (Mg). Results of the analyses on the composite stormwater samples obtained at each site are presented i n Table 4-2, and Appendix C contains detailed total trace metal data for a l l samples taken. In order to present this data i n an under-standable format (relating trace metal contamination to land use), total con-centration values for those metals most often considered toxic to aquatic organisms (cadmium, chromium, copper, nickel, lead and zinc; FWPCF, 1968), as well as manganese and iron, were divided into ranges of concentrations and plotted on maps of the drainage basin (Figures 4-4 to 4-11). Due to the generally low concentrations found in, stormwater samples, no obvious patterns of trace metal contamination are present. Such patterns may emerge i n the discussion of land use influences on metals concentrations. 6 2 . Figure 4 - 3 Hydrocarbons Vs. Buildup Time, A l l S i t e s . TABLE 4-2 PHASE ONE TRACE METAL ANALYSIS ON COMPOSITE SAMPLES STATION ANALYTICAL PARAMETERS 1 NUMBER NAME SAMPLING DATE CLASS BUILDUP TIME (DAYS) Ca C d 2 C r 2 Cu Fe Mg Mn N I 2 Pb Zn 1 Brentwood M a l l 1 5 / A U G / 7 9 C 33 5 2 . 6 22 . 3 0 . 0 .34 4 .93 3 . 7 8 0 .43 5 3 . 1 .75 2 . 7 0 2 B e a v e r b r o o k D r i v e a t B e a v e r b r o o k C r e s e n t 2 6 / S E P T / 7 9 R 18 4 0 . 0 7 . 1 0 . 0 . 2 9 2 .88 3 . 3 8 0 . 4 6 33 . 1 .44 2 . 5 6 3 3 P r o d u c t i o n Way a t T h u n d e r b i r d C r e s e n t 1 4 / O C T / 7 9 I 17 2 6 . 3 2 . <10. 0 .28 1 .68 1 . 8 0 0 . 1 0 3 . 0 . 3 3 0 . 6 2 3 4 Lougheed M a l l 2 4 / O C T / 7 9 C 1.5 3 . 3 <1. 16 . 0 . 0 8 1 .48 0 . 7 0 0 .05 <2. 0 .26 0 . 2 6 3 5 Wedgewood Avenue a t 1 s t Avenue 2 1 / N O V / 7 9 R 4 7 .7 <!• <10. <0.04 0 . 4 9 0 .84 0 . 0 6 2. 0 . 1 1 0 . 0 9 3 6 R o b e r t Burnaby P a r k 2 1 / N O V / 7 9 0 4 2 . 9 <1. <10. <0.04 0 . 5 8 0 .66 0 . 0 8 <2. 0 .04 0 . 0 6 3 7 Deer L a k e Park 2 1 / N O V / 7 9 0 4 4 . 7 <2. <10. <0.04 0 . 6 6 0 .75 0 . 0 3 <2. 0 .07 0 .04 3 8 H a s z a r d S t r e e t a t Buckingham Avenue 21 /NOV/79 R 4 6 .4 <2. <10. <0.04 0 .59 0 .82 0 .05 <2. 0 .19 0 . 1 1 3 9 Warner L o a t P a r k 1 5 / D E C / 7 9 0 7 4 . 7 <2. <10. <0.04 0 .73 0 . 6 3 0 .04 <2. 0 . 0 6 0 . 0 8 10 B r u n e t t e Avenue a t C a p l l a n o S t r e e t 3 1 / J A N / 8 0 I 15 2 4 . 5 24. 92 . 0 .25 2 5 . 2 0 6 .27 0 . 6 7 126 . 0 . 9 9 0 .72 11 A l p h a Way a t A l a s k a S t r e e t 3 1 / J A N / 8 0 I 15 12 .8 2 . 72 . 0 . 27 1 1 . 2 0 5 .15 0 . 3 6 3 . 1 . 08 0 .61 12 B o u n d a r y Road a t Canada Way 3 1 / J A N / 8 0 C 15 27 .2 <2. <10. <0.04 2 .80 5 .94 0 .17 3 . 0 .41 0 . 3 0 * A l l m e t a l s v a l u e s i n m g / L , except as i n d i c a t e d . 2 V a l u e s i n p g / L . P o s s i b l e z i n c c o n t a m i n a t i o n o f samples from d i s t i l l e d w a t e r . Figure 4-4 Cadmium D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin. Figure 4-5 Chromium D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin. Figure 4-6 Copper D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin. Figure 4 - 7 Iron D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin. • 0-2 • 2-50 • 50-100 • >100 Figure 4-9 N i c k e l D i s t r i b u t i o n i n Stormwater i n the Brunette R i v e r Basin. Figure 4-10 Lead D i s t r i b u t i o n i n Stormwater i n the Brunette River Basin. Figure 4-11 Zinc D i s t r i b u t i o n i n Stormwater i n the Brunette R i v e r Basin. 72. When looking at maximum total metals values, half of the metals examined had maxima i n samples taken from site one, the Brentwood Mall area, which may again reflect the effect of the length of antecedent dry period more so than the land use around the sampling site (keeping i n mind the problems inherent i n differentiating between land use effects and dry period effects on storm-water trace metals levels). Manganese (0.64 mg/L), copper (0.45 mg/L), cad-mium (50 Pg/L) and nickel (74. Pg/L) a l l had maximum values at this site, while iron (30 mg/L), chromium (0.032 mg/L) and lead (4.14 mg/L) had maximum values at site four, Lougheed Mall. Only zinc (5.40 mg/L) had a maximum value at a site other than a commercial one (site five, a residential area close to Robert Burnaby park). While i t would appear, as mentioned previously, that the presence of these comparatively high total trace metal values at site one i s more a func-tion of buildup time than anything else (from Table 4-2, composite metal values in almost a l l cases tend to increase with increasing buildup, time, regardless of the land use from which the sample i s taken), maximum values at site four are attained after only a 1.5 day buildup. This would suggest one of two things about this sampling site: either the area around the storm-sewer, consisting of approximately 90% paved parking surface and 10% vegeta-tive cover, was subjected to an inordinantly high amount of iron, chromium and lead-containing substances before sampling (an amount not explained by the usual volume of t r a f f i c and the other daily a c t i v i t i e s i n the area), or, more r e a l i s t i c a l l y , the previous storm failed to completely wash out these particular trace metals from around the catchment area. Assuming that these metals were closely associated with particulate matter (as has been suggested by a number of studies; G r i f f i n et a l . , 1980; Helsel et a l . , 1979) this author would subscribe to the theory of Wilber and Hunter (1977) who state 7 3 . that the less consolidated (less closely packed) a sediment mass i s , the more easily a " c r i t i c a l erosion velocity" may be reached, at which point the sedi-ment mass i s resuspended and washed out of the catchment area. Failure of a storm event to remove a l l the contaminants from the area via this scouring process would obviously lead to erroneous and inflated conclusions as far as contaminant concentrations per individual storms are concerned. This serves to i l l u s t r a t e the problem of deciding what factor has the most influence on stormwater metals concentrations; only when the previous storm has completely washed out trace metal and other contaminants from the area can useful estimates of individual pollutant concentrations and loadings for discrete storm events be obtained. However, with no other basis to go on (due to lack of sampling replicates at the sites through different storms) i t must be assumed that each storm scour i s essentially complete, thereby making each storm temporally discrete. An examination of the composite trace metal values i n Table 4-2 reveals no obvious relationships between trace metal presence at the sampling site and the land use at that site; the main correlation seems to be between total trace metal concentration and buildup time. There are, however, certain exceptions not explained by this trace metal-buildup time relationship, i.e. the composite sample with the highest Fe, Mg, Cr, Ni and Mn values was taken from the industrial site 10 after a buildup time of 15 days (compared with a mean buildup time for a l l sites of 11.5 days), not an excessively long dry period. This might be explained by the fact that only one grab sample, as opposed to a number of 10-minute individual samples, was taken from this site, and as such i t would be unrealistic to say that this was a true compo-site sample. However, sufficient runoff time from the start of the storm was allowed to ensure a more homogenous sample was obtained, giving, i n the 74. authors' opinion, a representative average stormwater sample from the site. Therefore, a look at the land use in the area might explain this apparent exception to an otherwise general relationship. This site has accounted for anomalies in other parameters as well, and i t might be surmised that the land use activities at the site i.e., lumber processing and brewery related work, with associated heavy t r a f f i c and waste by-products, are contributing high amounts of trace metals and other contami-nants to the stormwater runoff, which has direct access to the nearby Fraser River. Further investigation of this site seems warranted i n order to obtain more complete knowledge of the potential environmental effects resulting from stormwater runoff from this area. The existence of the trace metal-solids complex, as mentioned previously, might be expected to account for any patterns seen i n the presence (or absence) of trace metals at the various sampling sites. G r i f f i n et a l . , (1980) suggest that upwards of 90% of the total trace metal fraction present i n the water column may be associated with particulate matter; from this we would assume that a greater concentration of trace metal w i l l be present i n association with a high concentration of total solids. This would follow from the previous discussion on the scouring process of stormwater runoff (Wilber and Hunter, 1977) i n that the unassociated metal fraction would be easily washed out of the system under typical runoff flow and would therefore exert l i t t l e influence on the system. On the other hand, the metal-particulate complex would remain in the system for a longer period of time, i e . , until the c r i t i c a l erosion velocity, as defined by Wilber and Hunter, i s reached. Because such a high percentage of the total trace metals are associated with particulate matter, any relationships between these para-meters should be readily observable from the data. These results, while 75. greatly influenced by the presence of a relatively low suspended solids content (mean T.S.S. concentration for a l l sites, excluding site 10, i s 103 mg/L), and the var i a b i l i t y of the runoff flow velocity between individual sites (an assumption i s made that the c r i t i c a l erosion velocity was achieved at a l l sampling sites), tend to show this relationship i n some, but not a l l cases. Table 4-3 shows the linear correlation values (y 2) for total trace metals as a function of total suspended solids, measured at a l l sampling sites. The closer the correlation (goodness-of-fit) i s to 1.00, the better i s the f i t of a straight line drawn through the points on a graph, and the more linear i s the relationship (Dixon and Massey, 1969). For comparative purposes, Table 4-4 includes the correlation values for trace metals as a o t function of buildup time. It can be seen that, with the exception of magne-sium, cadmium and manganese, the trace metal concentrations are closely related to either total suspended solids content or buildup time, and these two parameters readily account for the trace metal presence at each sampling site. Iron, chromium, and nickel are linearly related to suspended solids content, while zinc, lead, calcium and copper show a definite correlation to buildup time. Magnesium, cadmium and manganese show an equally strong or, conversly, equally weak relationship to either suspended solids or buildup time. These relationships are, at f i r s t glance, independent of the land use at the various sampling stations i.e., from Table 4-2 there i s no obvious correlation between land use and trace metal content; i t must be remembered, however, that the land use can exert a great deal of indirect "internalized" influence on the properties of stormwater runoff. For example, Helsel e^ t a l . , (1979) suggest that parameters such as percent impervious cover, average METAL 76. TABLE 4-3 LINEARITY OF RELATIONSHIP BETWEEN TOTAL TRACE METALS AND  TOTAL SUSPENDED SOLIDS, ALL SITES Ca Cd Cr Cu Fe Mg Mn Ni Pb Zn CORRELATION .12 .69 .88 .33 .97 .59 .73 .88 .30 .03 Y 2 METAL TABLE 4-4 LINEARITY OF RELATIONSHIP BETWEEN TOTAL TRACE METALS AND  BUILDUP TIME, ALL SITES Ca Cd Cr Cu Fe Mg Mn Ni Pb Zn CORRELATION .95 .68 .31 .84 .33 .64 .68 .47 .87 .83 Y 2 land slope, dwelling unit density, curb length per unit area, drainage type, rooftop area, street area, f e r t i l i z e d lawn area and off-street parking area wi l l a l l have an influence on the trace metals found in storm drainage. Since these factors w i l l be different for each land use category, and vari-able even within a single category, the land use of any particular area w i l l definitely be exerting some influence on the properties of the stormwater runoff from that area. It would follow from this that the suspended solids concentration of the runoff, as well as other parameters, would be greatly influenced by the type of land use ac t i v i t i e s and characteristics existing i n 77. any particular sampling area, and this i n turn w i l l have an influence on the trace metals concentrations. Rimer et a l . , (1978) suggest that, for a factor like suspended solids, the imperviousness of an area w i l l have a large effect on the total solids content of runoff from that area i.e., in an open/ greenspace area, with very l i t t l e paved area and a low percentage of imper-vious surface, the suspended solids concentrations w i l l be low, contrary to what might be assumed and expected in an area influenced by s o i l erosion. Conversely, i n an urban setting (residential and commercial), the percentage of impervious surface increases dramatically and with i t the suspended solids content of runoff. From the findings i n Table 4-3, i t would thus be expected that the Cr, Fe, Mn and Ni concentrations would be equally elevated i n these urban areas, regardless of the particular land use ac t i v i t i e s ongoing in the sampling area, but subject to the "internalized " influence of the character-i s t i c factors mentioned above. While i t i s very d i f f i c u l t to delineate each of these characteristic factors, certain general observations can be made about their affects on stormwater runoff. A more detailed study would be needed to detail the exact influence of each. In general then, i t can be concluded from Tables 4-3 and 4-4 that, i f the trace metal concentrations are found to be correlated to total suspended solids, there w i l l be no great effect on these metals by the land use of the sampling area i.e., the total trace metal concentrations are more or less independent of land use, but greatly dependent on the specific land charac-teristics of Rimer et a l . , (1978). On the other hand, If i t i s found that the trace metals are correlated with buildup time, as from Table 4-4, i t may be concluded that these metals are greatly influenced by the land use, and subsequent a c t i v i t i e s , of the sampling area. With a longer buildup time, whatever a c t i v i t i e s are contributing to the trace metal content of the runoff 78. (ie., t r a f f i c flow, manufacturing, processing, etc.), and are thus character-i s t i c of that particular area, w i l l have the opportunity to influence the buildup of these metals to a greater degree than with a shorter buildup period. Further, this buildup w i l l be greater i n the industrial/commercial areas than i n the open areas. From this one could expect to see high trace metal concentrations with longer buildup time i n those areas where the a c t i -vities were contributing the most to trace metal concentrations (industrial/ commercial), and lower concentrations i n the recreational areas (for buildup time correlated metals). Conversely, i f the metals are more greatly influenced by the suspended solids we would expect more or less the same metal concentrations whether the buildup period i s short or long, and the dichotomy between industrial/commercial/residential and open/ greenspace areas would be reduced. While these relationships are very general and greatly variable between different areas, they are nonetheless valuable i n characterizing stormwater runoff. A comparison of trace metal concentrations found within each land use classification would be useful in delineating the contributions of these individual land uses to the overall trace metals pollution problem. Figure 4-12 and Table 4-5 present these in the form of mean trace metal concentra-tions, along with the standard deviation and the range (to give some idea of the variability within each land use). It should be noted that there are obvious problems involved i n drawing meaningful comparisons between para-meters sampled at different sites and i n different storms. G r i f f i n et a l . , (1980), i n recognizing this problem, listed the reasons for what they f e l t should be a cautious appraisal of any data derived from this type of study. These include (a) natural discountinuities exist i n space and time i n the 79. 1001 Ca 501 75 5d 25 25 4> 501 92. Cr' 501 25 25 20 15 25 C R 30 Fe R 3 Ui a z o o § LU Ol All values in mg/L except a jjg/L RANGE MEAN STANDARD DEVIATION LAND USE I-INDUSTRIAL C-COMMERCIAL R-RESIDENTIAL G-GREEN/OPEN SPACE Figure 4-12 Means, Ranges and Standard Deviations of Trace Metal Content in Runoff From Each Land Use Classification. 8 0 . 8| 6 100i 75 50 25 200 150 100 50 Mg 5CH 25 126, Ni 200| 150 100\ 501 I C R G r ^ 6 zn I 4* I DC 67| 6< X Mn(x10 ) R G 414J 216) 2 4 3 r ~ l 1 PbOcIo"2) RANGE STANDARD DEVIATION MEAN All values in m g / L except a ^ j g / L LAND USE I-INDUSTRIAL C-COMMERCIAL R-RESIDENTIAL G - G R E E N / O P E N S P A C E Figure 4-12 Continued. 81. TABLE 4-5 EFFECT OF LAM) USE ON TRACE METAL CONTENT IN STORMWATER RUNOFF PARAMETER TRACE METAL CONCENTRATION1 Ca Cd2 Cr Cu Fe Ma NL2 Pb Zn ARITHMETIC MEAN 23.9 5. 0.033 0.21 7.34 3.23 0.23 21. 0.68 0.78 INDUSTRIAL STANDARD DEVIATION 5.7 8. 0.034 0.05 8.63 1.76 0.22 46. 0.38 0.16 RANGE 12.8-30.7 2.-24. <0.01-0.092 0.14-0.27 1.76-25.2 1.80-6.27 0.10-0.67 3.-126. 0.35-1.19 0.61-1.10 ARITHMETIC MEAN 27.2 11. 0.02 0.21 5.37 2.63 0.25 27. 1.26 1.34 COMMERCIAL STANDARD DEVIATION 24.5 14. 0.01 0.13 8.03 1.95 0.21 29. 1.17 1.20 RANGE 2.2-54.8 <1-50. <0.01-0.032 <0.04-0.45 1.04-30.0 0.50-5.94 0.03-0.64 <2.-74. 0.15-4.14 0.20-3.00 ARITHMETIC MEAN 15.2 3. <0.01 0.10 1.36 1.41 0.16 22. 0.50 1.09 RESIDENTIAL STANDARD DEVIATION 15.6 3. 0 0.10 1.91 1.46 0.20 28. 0.72 1.67 RANGE 4.2-57.1 <1-8. — <0.04-0.30 0.01-6.94 0.13-5.33 0.03-0.51 <2.-86. 0.06-2.16 0.06-5.40 ARITHMETIC MEAN 4.7 2. <0.01 0.04 0.49 0.61 0.05 2. 0.05 0.06 OPEN/GREENSPACE STANDARD DEVIATION 2.2 1. 0 0 0.43 0.30 0.03 <1. 0.04 0.04 RANGE 1.2-9.4 <1-2. — — 0.01-1.59 0.22-1.36 0.02-0.13 <2.-3. 0.02-0.15 0.02-0.1B 1 \alues in mg/L as total trace metal expect where indicated. 2 Values in pg/L. 82. runoff procedure, such that the runoff patterns between two physically separ-ate sites are never exactly identical; (b) identical r a i n f a l l patterns usually do not occur on physically separate sites (this becomes much more important when the sampling sites are small); and (c) any measure of pollut-ant yield used i n a "wet" study (one involving sampling of stormwater run-off), i s unavoidably extrapolative in nature; that i s , for example a para-meter such as kg/hectare/cm r a i n f a l l - a pollutant load of 10 kg resulting from 0.25 cm of r a i n f a l l i s no guarantee that 20 kg of the same pollutant would result from an event of 0.50 cm, a l l other factors remaining equal. It therefore becomes very d i f f i c u l t to draw meaningful conclusions between pollutant runoff from separate storms, unless a l l factors involved i n the temporal and spatial distribution of each storm event are taken into consi-deration. Since this would involve an inordinant amount of measurement and analysis, i t would have to be assumed that, by taking composite samples from each site, an average pollutant concentration from these sites i s being obtained, and these averages are amenable to comparison (keeping i n mind that the differences between individual storm events make these comparisons some-what less than absolute). From Figure 4-12 we can see that there i s a high degree of var i a b i l i t y i n the trace metal concentration within any single land use area. This may be partly accounted for by the small number of samples taken from each land class (twelve i n total), as well as by the natural discountinuities described above. Other studies have encountered the same kinds of va r i a b i l i t y i.e., Hall et a l . , (1976) found i n their study of trace metals in street surface contaminants, that for many of the trace metals, especially i n the industrial and commercial areas, one or two large values resulted in standard deviations which exceeded the lower range values. This appears to be true for this 8 3 . study as well, most notably i n the residential land use class. However, accepting that this large variability i s an integral part of a study of this type, a comparison of mean trace metal concentrations reveals some interest-ing trends. Like Hall et a l . , i t can be seen that the highest degree of contamina-tion of those metals deemed most toxic to aquatic organisms (Cd, Cr, Cu, Ni, Pb and Zn) occurs i n the industrial and commercial land use areas. As well, the other metals studied (Ca, Fe, Mg and Mn) also show the highest contamina-tion i n the industrial and commercial areas. The metals most prevalent and, therefore, those with the highest contamination potential i n the industrial areas appear to be Cr, Fe and Mg, of which only Cr i s deemed to be environ-mentally important at the concentrations found (F.W.P.C., 1968). In the commercial areas, Ca, Cd, Mn, Ni, Pb and Zn are found i n highest concentra-tion, while Cu i s found at equally high concentrations i n both areas. It should be noted that while Cr, Fe, Pb and Zn appear to be i n highest mean concentration at one or the other of the industrial or commercial sites, these mean concentrations are a l l biased by one very high value. Nonethe-less, in almost a l l cases, the residential land use areas show much lower mean trace metal concentrations, and the open/greenspace areas a l l show the low trace metal levels indicative of background or "natural" contamination. It would appear from this that the commercial land use sites, those areas associated with large shopping plazas (Brentwood Mall and Lougheed Mall) or smaller concentrated clusters of shops along major t r a f f i c thoroughfares, are the most severely contaminated by the trace metals studied. These are followed by the industrial, residential and open/ greenspace sites, respectively. This seems a reasonable observation i f a comparison i s drawn between the general site categories i n terms of the land 84. use parameters Helsel et a l . , listed in their study of the land use influences on metals in storm drainage, and previously mentioned in this chapter. In terms of such parameters as percent impervious surface, curb length per unit area (a parameter which Hall et a l . , suggest act as flow channels, directing surface runoff to the storm sewers and creeks), street area and off-street parking area - a l l indicative of t r a f f i c volume in the respective areas - the commercial stations would be expected to have the highest contamination values of the different land use categories. For example, for a parameter like percent impervious surface, the commercial areas consist mainly of paved parking surface and pedestrian walkways which would allow very l i t t l e i n f i l t r a t i o n of precipitation, and any trace metals deposited in these areas would be readily picked up and washed out in storm-water runoff. On the other hand, the industrial areas investigated have a higher percentage of unpaved, bare-earth walkways and roads, and would be expected to contribute significantly less trace metal contamination by mass, due in part to increased precipitation i n f i l t r a t i o n and diminished stormwater volume. The same would apply to the residential areas, with their increased lawn and vegetal area. Exactly these kinds of trends are seen in the trace metal data for the other land use parameters as well, and i t must therefore be concluded that the commercial areas are contributing a high proportion of trace metal contamination in stormwater runoff. The assumption that the open/greenspace trace metal values are representative of background levels of contamination is deemed a valid one, due to the relatively isolated nature of these sites from the influence of t r a f f i c , industrial processing and manufacturing a c t i v i t i e s , and residential activities such as lawn f e r t i l i z a t i o n , etc. Using these as background values, some idea of the relative degree of contamination at the other sites 85. can be obtained. This i s accomplished through the use of s t a t i s t i c a l methods to look at degrees of significant difference between each land use c l a s s i f i -cation and the open/greenspace areas. Table 4-6 summarizes this analysis. From i t and Figure 4-12 i t can be seen that, with the exception of one metal, a l l the land use classes sampled had stormwater runoff with trace metal concentrations significantly different from the open/greenspace areas, thus suggesting a higher degree of contamination. Again i t should be noted that the data i s far from comprehensive, and other factors such as atmospheric d r i f t can transport contaminants into any of the sampling areas, thereby raising the naturally occurring activity-related trace metal levels measured; however, the data are useful for general comparative purposes. Overall, runoff from the commercial areas tends to show consistently higher levels of toxic trace metals than either of the industrial or residen-t i a l sampling sites, with Cr, Cu, Mn, Pb, Zn, Cd and Ni a l l occurring at higher levels i n these areas. More specifically, some of these metals (Cu, Mn, Pb & Zn) occur at the same level of significant difference i n both the commercial and industrial sites, while Cd and NL occur at the same level of significant different in the commercial and residential sites. Only Fe i s found at significantly higher levels at sites other than commercial ones, i n this case the industrial sites. While i t i s d i f f i c u l t , without further detailed study of each sampling site, to explain exactly why some metals are found at the same level of significant difference in two different land use classifications while others are not, the general pattern of stormwater trace metal contamination seems to be one in which the commercial sites show the highest pollutant levels, followed by the industrial sites. The residential sites show consistently lower levels of trace metal contamination. Generally this agrees with the results of other studies. Hall et a l . , (1976) summarized 8 6 . TABLE 4-6 TRACE METAL CONTAMINATION SIGNIFICANTLY DIFFERENT FROM GREENSPACE AREAS TRACE METAL SIGNIFICANT DIFFERENCE 1 INDUSTRIAL Ca Cd Cr Cu Fe Mg Mn NL Pb Zn A C B A A A A C A A COMMERCIAL A B A A B A A A A A RESIDENTIAL B B _2 B C B B A B B 1 A - s t a t i s t i c a l l y d i f f e r e n t at the 1% l e v e l of s i g n i f i c a n c e (a = 0.01) B - s t a t i s t i c a l l y d i f f e r e n t a t the 5% l e v e l of s i g n i f i c a n c e (a = 0.05) C - s t a t i s t i c a l l y d i f f e r e n t at the 10% l e v e l of s i g n i f i c a n c e (a = 0.10) 2 No s i g n i f i c a n t s t a t i s t i c a l d i f f e r e n c e . the findings of several U.S. studies which concluded that commercial and i n d u s t r i a l streets were the most heavily contaminated with trace metals; further, i t was found that Cd, Cu, Pb and Zn reached t h e i r highest concentra-tions i n these commercial areas; t h i s agrees with the present study. While i t appears that there i s s i g n i f i c a n t trace metal contamination at most of the sampling stations i n the commercial, i n d u s t r i a l and r e s i d e n t i a l areas, a comparison of act u a l trace metal l e v e l s obtained from t h i s study (Table 4-2) with those of other studies i n the l i t e r a t u r e provides some i n d i -cation of the r e l a t i v e degree to which these sampling stations are contamina-ted. Table 4-7 summarizes t h i s information. Generally these values seem to be comparable for Cd, Cr, Cu, Mn, Pb and Zn, with the ranges obtained i n t h i s study f a l l i n g e a s i l y within those of the other studies. Only Ni appears to TABLE 4-7 TRACE METAL DATA FROM THE LITERATURE TRACE METAL CONCENTRATION1 Cd Cr Cu Mn Ni Pb Zn STUDY Min Max Min Max Min Max Min Max Min Max Min Max Min Max S e a t t l e ( a ) 0.002 0.05 - - - - - - - - 0.008 3.5 0.003 1.6 S t i l l C r e e k ^ 1 0.001 0.008 - - 0.004 0.03 0.13 0.28 - - 0.001 0.12 0.033 0.10 2 0.001 0.005 - - 0.002 0.011 0.09 0.23 - - 0.001 0.031 0.002 0.06 S t i l l , . Creek>c-' 0.001 0.021 0.005 0.07 0.006 0.13 0.02 0.59 0.001 0.07 0.003 0.81 0.01 0.47 E P S ( d ) - 0.05 - - - 0.07 - - - 0.05 - 0.25 - 2.70 Prince, .2 George^ - <0.03 <0.02 0.61 <0.01 0.38 - - - <0.06 <0.02 0.60 <0.01 6.50 ( f 1 z Kamloopsv ' <0.03 <0.06 <0.03 5.0 <0.03 0.21 - - <0.06 0.07 <0.03 6.0 <0.03 0.21 East York (8) - <0.01 - 0.11 - 0.09 - 0.30 - 0.03 - - - 0.61 a) UFerguson and Hall, 1979 J Values in mg/L unless otherwise indicated b) Ferguson and Hall, 1979 2 Extractable trace metal. c) Ferguson and Hall, 1979 d) Ferguson and Hall, 1979 e) Sidhu, 1975 a f) Sidhu, 1975 b g) Mills, 1977 88. be found at a somewhat elevated level i n this study, with the maximum concen-tration of 0.126 mg/L being approximately ten times higher than any of the other maximum values. However, i t must be noted that relatively small quan-t i t i e s of the metals are being found here, with most values being measured i n the pg/L or low mg/L range, and the d i f f i c u l t y of comparing different storm-water study results on anything more than a semi-quantitative basis must be kept i n mind. Overall i t may be concluded that the stormwater runoff from the land use sites i n this study has average trace metal contamination. 4.1.3 Phase One Acute Toxicity Bioassay Results Acute toxicity bioassay results for the composite stormwater samples collected from each sampling site are summarized i n Table 4-8, and graphic-a l l y illustrated on a map of the drainage basin i n Figure 4-13. A prelimi-nary analysis of the data reveals no obvious relationships between toxicity and type of land use surrounding each sampling station i.e., six of the 12 sites produced stormwater that was toxic i n some degree to Daphnia pulex, with three of these being designated commercial land use, two industrial and one residential. The open/greenspace sites, being representative of natural background contamination, a l l produced stormwater runoff that was non-toxic to D. pulex i n the static 96-hour tests. The most toxic stormwater sample (at LC 5 0 = 10.0%, indicating that half the test organisms would die within the prescribed time period upon exposure to a bioassay solution containing one part stormwater i n ten parts dilution water) was obtained at a residential sampling site (site 2, Beaverbrook Drive and Beaverbrook Crescent), and the least toxic of those samples which showed a lethal concentration was taken from the industrial site 10 (at LC 5 0 = 60.2%). From the previous discussion of the chemical analyses of the storm-water samples obtained from the twelve preliminary sites (sections 4.1.1 and 8 9 . TABLE 4-8 PHASE ONE ACUTE TOXICITY BIOASSAY RESULTS1 DATE STATION CLASS pH TOXICITY (LC 5 Q%) INITIAL FINAL 15/AUG 1 C 6.92 7.04 11.5 27/SEPT 2 R 6.64 6.90 10.0 NT2 14/OCT 3 I 6.30 6.92 24/OCT 4 . C 7.55 7.62 39.6 22/NOV 5 R 7.22 7.40 NT 6 0 7.35 7.28 NT 7 0 7.48 7.62 NT 8 R 6.73 6.82 NT 16/DEC 9 0 6.72 6.78 NT 31/JAN 10 I . 7.38 7.24 60.2 11 I 7.42 7.30 22.3 12 C 7.62 7.70 40.03 1 A l l tests at 17°C for 96 hrs. 2 NT = Non Toxic after 96 hrs. 3 Toxicity may be resulting from entrapment of organisms i n sediments 4.1.2), these toxicity results are almost exactly opposite to what might have been expected, taken on an individual site basis, in that the residential areas i n general had the least contamination of any of the three land use classifications; yet, the most toxic stormwater came from a residential site. On the other hand, site 10 might have been expected to yield a highly toxic sample when, in fact, a much lower toxicity was found. By looking at the land use classifications as a group, however, the toxicity results seem to follow the pattern established by the other chemical parameters i.e., the commercial sites show the highest toxicity, followed by the industrial sites, with the residential sites having the least toxic stormwater runoff. The fact that the most highly toxic stormwater sample came from a residential area must be ascribed to the particular environment of this site; the area seemed to be more heavily travelled than other 100% nontoxic Figure 4-13 Acute Toxicity-'- of Stormwater From the Brunette River Basin. Measured as 96-hr. LC50: Presented As 100-LC50%. 91. residential areas, since Beaverbrook Drive connects two main thoroughfares and as such acts as a shortcut for the t r a f f i c in the area. As well, visual observation of the stormwater runoff suggested the presence of a reasonably high volume of gasoline (and possible other hydrocarbons), although the iso-octane concentration determined doesn't indicate an especially high hydro-carbon presence. In any case, i t would appear that this sampling site i s unusual i n the toxicity of the urban runoff obtained there; the reasons for this are d i f f i c u l t to ascertain and would require further study. Generally, the pattern of land use versus degree of contamination found earlier seems to apply. Examination of Figure 4-13 indicates a pattern i n the toxicity of the stormwater samples, i n that the sampling sites i n the upper and lower reaches of the basin (1, 11 and 12; 2, 4 and 10) produced runoff that was toxic to D.  pulex, while those sites i n the middle of the basin a l l had runoff which was non-toxic. In general, these middle reaches of the basin, being at the lower gently sloping elevations and close to the main water bodies, are reserved for residential and open/greenspace land uses, while the commercial and industrial areas are found towards the outer edges of the basin (see Figure 3-2). It would appear from this that the pattern of toxicity i s merely a reflection of the land use patterns i n the basin; however, drainage maps of the area (Municipality of Burnaby, Dept. of Engineering; personal communica-tion) indicate that the highest amount of drainage from the northern and southern elevations occurs at the ends of the basin, while the middle areas receive less drainage contribution from these elevated areas. This would suggest that the industrial and commercial areas at the ends of the basin should be expected to have a more highly contaminated urban runoff, as this runoff i s being contributed from a larger area than that of the middle reaches; the shear volume and loading of this drainage contribution would be expected to account in part for the observed toxicity pattern. Therefore, i t cannot be stated for certain whether the observed toxicity pattern i s a result of land use or drainage contributions, only that a definite pattern of runoff toxicity exists. Table 4-9 summarizes the output of the INMSDC program, from which par-t i a l linear regression coefficients were obtained for a l l the chemical para-meters investigated, i n relation to toxicity. The data from the sampling sites were grouped and analyzed according to land use classification, and i t can be seen from the table that there are no obvious patterns of chemical parameter relationship to toxicity within any of the land use groups. For the commercial sites, pH i s the only independent variable that shows any kind of positive correlation to toxicity, and even then the correlation doesn't appear to be very strong (at 0.57). As such i t might be suspected that pH i s having some effect on the other chemical parameters at these sites (most li k e l y affecting the forms and ava i l a b i l i t y of the trace metals), but there i s no evidence to suggest that this effect i s a strong one. As far as the industrial land use sites are concerned, buildup time and calcium content appear to be positively correlated to toxicity. Calcium, however, i s not usually associated with trace metal toxic effects to aquatic organisms; as such, the strong positive correlation shown (0.91) may be more a function of the significant difference in the Ca concentration at these sites when compared to other trace metals than to any real relationship to observed toxicity. On the other hand, the buildup time parameter i s merely an indication of the number of dry days preceding the sample storm, and i s thus indicative only of the collection at each site of trace metals and other contributors to daphnid toxicity; therefore, the fact that the buildup time TABLE 4-9 PARTIAL LINEAR REGRESSION COEFFICIENTS FOR  CHEMICAL PARAMETERS AS A FUNCTION OF TOXICITY INDEPENDENT VARIABLES DATA CONSIDERED BUILDUP COD PH T.S.S T.ALK. T.HARD. CHARD COND. H.CAR. Fe Zn Mg Cr Ni Pb Cd Ca Cu Mn COMMERCIAL -.90 -.99 .57 -.98 .25 -.91 -.92 -.25 -.86 -.92 -.99 -.09 -.96 -.99 -.99 -.99 -.87 -.99 -.95 INDUSTRIAL .88 -.80 -.72 -.24 -.04 -.97 -.62 .13 -.84 -.42 .06 -.73 -.74 -.02 -.92 -.02 .91 .34 -.47 RESIDENTIAL -1.0 -.99 .85 .05 -.53 -.99 -.99 -.99 -.99 -.99 -1.0 -1.0 -1.0 -.99 -.98 -.99 -.99 -1.0 -.99 OPEN/GREENSPACE The potei i n s i g n i f i t i a l cant Inde inde] >endent 1 endent ' arlables arlable c for deper r Is fore ident va; ed out i iable t the o x i c l t y egress lc are . m eqi i l l i t atioi islgn f icai i t . ' "he ci mstai it i s eithi r an ALL SITES THAT SHOWED TOXICITY .85 -.57 .89 -.23 -.99 -.99 -.98 -.99 -.99 -.99 .74 .72 -.44 -.99 -.77 -.02 -.99 .88 -.99 94. parameter has a strong l i n e a r r e l a t i o n s h i p to t o x i c i t y i s important, but by i t s e l f doesn't give any i n d i c a t i o n as to the exact cause of the observed t o x i c i t y . A l l that can be stated i s that, at the i n d u s t r i a l s i t e s , those parameters which contribute to daphnid t o x i c i t y are l i n e a r l y related to buildup, and the longer t h i s buildup time i s , the higher w i l l be the concen-t r a t i o n of these parameters and the more toxic w i l l be the runoff. The r e s i d e n t i a l s i t e s , those with the lowest t o x i c i t y of any of the land uses when considered as a group, show only pH to be strongly correlated (0.85) with daphnid t o x i c i t y . This parameter can be seen to be c l o s e l y r e l a -ted to the forms and a v a i l a b i l i t y of trace metals and other chemical para-meters i n the stormwater runoff, but i s again only i n d i c a t i v e of the other parameters a f f e c t i n g and contributing to stormwater t o x i c i t y . By i t s e l f , i t i s not a d i r e c t cause of t o x i c i t y . It might be concluded, therefore, that there are some i n d i c a t i o n s that land use i s related to, and in f l u e n c i n g stormwater t o x i c i t y as far as p a r t i a l l i n e a r regression i s concerned. However, t h i s approach can't t e l l us exactly what i s responsible for the observed t o x i c i t y . Another method of a n a l y s i s used involved combining the data from a l l sample s i t e s which produced t o x i c i t y values, regardless of land use at the p a r t i c u l a r s i t e s . By doing t h i s , i t i s found that buildup time, pH, Zn, Pb and Cu a l l show reasonably strong p o s i t i v e c o r r e l a t i o n to t o x i c i t y . This would suggest that t o x i c i t y i s not so much influenced by the land use i n the area as much as by the s p e c i f i c types of a c t i v i t i e s ongoing around any p a r t i -cular sampling s i t e . Conversely, some of these a c t i v i t i e s may be said to be influenced by land use i . e . , manufacturing, processing, etc., but the major-i t y of a c t i v i t i e s would be more or les s common to a l l land uses, except the open/ greenspace s i t e s i e . , t r a f f i c flow. The lack of c o r r e l a t i o n from the analysis between land use groups and their relation to toxicity would seem to support this. The use of partial linear regression coefficients to draw meaningful conclusions from this data must be viewed with a certain degree of caution due to the lack of replicates from each sampling site, as well as the natural spatial/temporal discontinuities existing between the individual sites and sampled storms (Helsel et a l . , 1979; G r i f f i n et a l . , 1980). A better indica-tion of the individual parameters' relation to toxicity may be obtained from a multivariate stepwise analysis of the data, the results of which are pre-sented in Table 4-10. Again, the data was f i r s t analyzed based on land use, and then on toxicity. The results seem to indicate even less of a relation-ship between the chemical parameters and toxicity than was found in the par-t i a l regression analysis. The commercial sites' toxicity appears to be related to the Zn concen-tration of the stormwater, while the residential sites' toxicity is related to buildup time, again more indicative of the contributions of various chemi-cal parameters to toxicity due to a long buildup period. It should be noted that the negative sign in this analysis indicates a positive relationship between the parameter and toxicity, due to the use of an inverted toxicity scale i.e., the lower the LC 5 Q value, the more toxic i s the sample. The industrial sites show no continuity among themselves, and the regression equation is simply a constant. These equations do not necessarily include a l l of the values that show a good linear correlation coefficient, since the c r i t e r i a for selection depends upon the F-probability of the independent variable rather than i t s R2 value. In general, there seems to be very l i t t l e relationship between the sampling sites within a specific land use and the toxicity of the stormwater taken from those sites. Even when only the sites TABLE 4-10 MULTIPLE REGRESSION EQUATIONS RELATING TOXICITY TO INDEPENDENT PARAMETERS DATA CONSIDERED EQUATION R2 F-PROB LEVEL F-PROB STANDARD ERROR LC 5„ COMMERCIAL TOXICITY (LCs 0 ) - 43.07 - 11.69 Zn .99 .05 .017 .61 INDUSTRIAL TOXICITY ( L C 5 0 ) - 61.17 0 .05 1.00 39.36 RESIDENTIAL TOXICITY (LCgn) - 127 - 6.50 BUILDUP 1.00 .05 .006 0.0 OPEN/GREENSPACE NO EQUATION EXISTS RELATING TOXICITY TO INDEPENDENT PARAMETERS — — — ALL SITES THAT SHOWED TOXICITY TOXICITY ( L C 5 0 ) - 83.77 - 5.29 Pb .99 .05 .012 .82 which produced a toxicity value were considered i n the regression equations, the relationship between the Independent and dependent variables i s poor, with only Pb showing any kind of strong positive correlation. In summary, there does appear to be some evidence to suggest that trace metal content makes a significant contribution to stormwater toxicity (from the commercial and overall regression equations), more as a function of the ac t i v i t i e s surrounding an individual sampling site than the specific land use designation of the site, but the mode of this contribution and the influence of other physiochemical parameters have to be investigated before any conclu-sions can be made. 4.2 Phase Two: Bioassay Using Synthetic Stormwater The LC 5 0 toxicity values derived from the bioassays performed during this phase of the research were analysed using the TRP computer package. Partial linear regression coefficients and multiple regression equations were produced to relate toxicity to the presence of the various trace metals by themselves and i n combination with varying pH and suspended solids content. These analyses were performed in an attempt to delineate the interactions and relationships between trace metals and the other parameters, with the inten-tion of explaining the observed toxicity of phase one, i f possible. In con-junction with this, LC 5 0 toxicity values were also used in an analysis of the data through the Marking-Dawson Additivity Index (Marking and Dawson, 1975; section 3.5.3). 4.2.1 Calcium Hardness-Trace Metal and Hydrocarbon-Trace Metal  Interactions Experimental d i f f i c u l t i e s prevented the obtaining of meaningful results from the bioassays using hardness and hydrocarbons i n association with trace metals. Calcium hardness concentrations of up to 300 mg/L (as CaC03) were 98. added to the Deer Lake dilution water; however, from the results of the bio-assays i t was discovered that D. pulex were too sensitive to hardness concen-trations of this magnitude, and couldn't survive very well in water with a hardness much above that of their natural habitat (approximately 40 mg/L as CaC03). Therefore the effects of hardness on trace metal toxicity could not be investigated but i t might be suggested that the calcium ion component of the hardness or the undissolved hardness particulate matter may be contribut-ing to the overall toxicity of the stormwater; while this i s not evident from the literature (most investigations involved the use of fish instead of daphnids), i t might be investigated using a less sensitive invertebrate as the test organism. A combination of mechanical and chemical d i f f i c u l t i e s prompted the ter-mination of the hydrocarbon-trace metal interaction investigations. F i r s t l y , the carbon tetrachloride solvent proved to be toxic to the daphnids at very low concentrations; therefore, the iso-octane was added directly to the bio-assay jars at the concentrations specified (2 to 10 mg/L). As mentioned previously, however, the hydrocarbon tended to float on the surface of the water, effectively removed from contact with the test organisms. Because of this, a magnetic stirring bar-mixer assembly was devised to provide enough circulation to ensure proper mixing of the iso-octane. Unfortunately, the circulation speed needed to keep the mixing bars in motion proved to be too hazardous to the daphnids which, more often than not, were crushed by the bar i t s e l f or by contact with the sides of the jar. It was thus d i f f i c u l t to differentiate between death by hydrocarbon toxicity and death by the process of mechanical mixing. There is very l i t t l e relevant data in the literature pertaining to hydrocarbon toxicity; however, we might suggest from a compar-ison of hydrocarbon concentrations (Hunter et a l . , 1979) that these samples 99. show average content and as such one would not expect to f i n d any unusual acute t o x i c i t y r e s u l t i n g from bioassays using these samples. Further inves-t i g a t i o n i s needed to confirm t h i s . 4.2.2 Experimental Comments ( i ) Choice of Trace Metals; From the r e s u l t s of the work c a r r i e d out i n the e a r l i e r phase of experimentation i t was seen that Cu, Fe, Pb and Zn were the trace metals present i n the highest concentrations and, as these (with the exception of Fe) have been designated as being most harmful to aquatic organ-isms, they were the trace metals used for the bioassay t e s t i n g i n phase two. However, a problem arose through the use of Fe which could not be solved to the author's s a t i s f a c t i o n and i t was decided a f t e r the f i r s t set of bioassays that only Cu, Pb and Zn be used for subsequent research. According to Standard Methods (1974), the f e r r i c i o n i s present and soluble i n the test s o l u t i o n only at low pH (the m e t a l l i c solutions used to make up the synthetic stormwater samples a l l had very low pH,. approximately 2.00, due to the use of an acid solvent); as the pH r i s e s , the i o n can be brought out of s o l u t i o n i n the f e r r i c form or, more l i k e l y , i t may hydrolyze to form i n s o l u b l e hydrated f e r r i c oxide, a yellowish p r e c i p i t a t e which s e t t l e s out of s o l u t i o n . Since a l l synthetic stormwater samples had to have a pH of approximately 7.00±0.05, the preliminary solutions were pH-adjusted using HCl and NaOH to t h i s ambient l e v e l for a l l tests except those dealing with pH e f f e c t s on trace metal t o x i c i t y - i n those cases, the pH was adjusted to the desired test l e v e l using HCl and feOH. As a r e s u l t of t h i s , the f e r r i c i on was p r e c i p i t a t e d out of solution and e f f e c t i v e l y removed from contact with the bioassay organism. If t h i s were to happen i n the natural aquatic environment, t h i s p r e c i p i -tate would most l i k e l y be washed out of the system and be unable to exert an 100. influence on the aquatic biota; there i s a possibility, however, that the precipitate could plug the f i l t e r i n g mechanisms of aquatic organisms and thus result i n mortality. While this possibility does exist, i t was decided that, in spite of the comparitively high concentrations of Fe found in the samples (up to 30 mg/L) the Fe metal was not greatly toxic to Dj^  pulex, a conclusion supported by the results of preliminary bioassays in which Fe alone and i n combination with other trace metals produced a very high LC 5 Q toxicity value, or no toxicity at a l l . Therefore, Fe was not used i n the subsequent examina-tions of pH and T.S.S. effects on trace metal toxicity; the author would recommend, however, a more detailed study of Fe effects on the f i l t e r i n g mechanisms of aquatic organisms. The other trace metals used showed no signs of precipitating out of solution when the synthetic samples were neutralized to pH 7, and i t was assumed that any observable effects were due to the target parameters and could be naturally occurring i n the f i e l d environment. ( i i ) pH Drift Over Test Duration: On average, the pH level was within ±6% of the i n i t i a l pH after 96 hours, the experimental duration, and i t was conclu-ded that the target pH was maintained against the weak natural buffering system of the Deer lake dilution water. 4.2.3 Bioassay Results The results of a l l bioassays performed in this research phase are pre-sented i n Tables 4-12 to 4-15, as the LC 5 Q acute toxicity value. In general, a l l bioassays produced an LC 5 Q value, indicating that a l l test solutions were acutely toxic to IK^  pulex to some degree, ranging from a high of less than 1% to a low toxicity of more than 90%. However, general trends are d i f f i c u l t to discern from this data alone; therefore, the results of TRP analysis for a l l groups, as well as the Marking-Dawson additlvity index are presented i n 101. Tables 4-16 and 4-17, respectively. These analyses were performed i n an attempt to f u l l y explain the observed toxicity and to isolate the toxic component(s) of the mixtures of chemicals and their various subtle interac-tions. In addition, the chemical composition of Deer Lake dilution water used for a l l bioassays i n this phase of experimentation i s presented i n Table 4-11. (i) Trace Metals Interactions; The results of the bioassays performed using Cu Fe, Pb and Zn, at the concentrations dictated by phase one results (0.45 mg/L, 30.0 mg/L, 4.14 mg/L and 3.20 mg/L, respectively), under constant pH (7.00±0.05) and solids (0.0 mg/L, after filtering) conditions are presented in Table 4-12. The trace metals were run both individually and i n combina-tion, and i t can be seen that the resultant LC 5 Q values encompass the above mentioned range of less than 1% to greater than 90% LC 5 0 toxicity. The most obvious trend apparent from this data i s the extremely low toxicity of Fe, as suggested previously. By i t s e l f , this metal produced an LC 5 Q value of 85.8%, less than half the toxicity exhibited by any of the other metals, and in the various trace metals combinations the toxicity i n a l l cases was substantially reduced by the presence of this ferric component. The Marking-Dawson index (Table 4-13) seems to further confirm this, i n that a l l combinations of Fe with the other trace metals produced an lowered S value (-0.38 to -0.91) when compared with the other combinations, indicating a high degree of antagonism between Fe and the other metals. One would have expected Fe to produce a very low i f not negligible toxicity, due to the precipitation of the ferric oxide at neutral test pH; however, i t appears that Fe also has a negative effect on the toxicity of the other metals. This may indicate that the ferric oxide precipitate acts as an insoluble solid i n the test solution, binding and/or dif f e r e n t i a l l y adsorbing the other trace metals i n solution, 102 TABLE 4-11 DEER LAKE DILUTION WATER CHEMICAL ANALYSIS PARAMETER COD pH T.S.S. Total Alkalinity Total Hardness Calcium Hardness Conductance Ca Cu Fe Mg Pb Zn CONCENTRATION! 19.26 7.02 0.0 80.03 60.03 44.03 343.8k 17.6 0.06 0.50 3.39 <0.01 0.02 1 A l l values i n mg/L, except as indicated Value unitless 3 Values i n mg/L as CaC03 ^ Value i n ps-crn-1 TABLE 4-12 TRACE METALS INTERACTION RESULTS RUN2 TRP DESIGNATION TRACE METAL COMBINATION LC 5 Q% AND CONCENTRATION1 Cu Fe Pb Zn — 4.14 — 35.8 0.45 9.20 30.0 — — 85.8 3.20 6.40 0.45 4.14 — <1.00 4.14 3.20 2.39 30.0 4.14 — 85.9 30.0 — 3.20 9.65 0.45 — 3.20 8.24 0.45 30.0 — — 92.0 __ 30.0 4.14 3.20 47.5 0.45 — 4.14 3.20 3.50 0.45 30.0 — 3.20 24.8 0.45 30.0 4.14 3.20 16.6 A - i B C D —1 E F G H I J METAL 1 METAL 2 K "I L METAL 3 M -J N ] METAL 4 1 A l l values i n mg/L 2 A l l runs at pH 7.0; TSS = 0 mg/L 103. TABLE 4-13 MARKING-DAWSON INDEX ANALYSIS RESULTS; TRACE METALS INTERACTIONS EXPERIMENTAL CONDITIONS Trace Metals pH T.S.S. Combinations (mg/L) S Pb-Cu 7.0 0 6.30 Pb-Zn 7.0 0 1.27 Pb-Fe 7.0 0 -0.71 Zn-Fe 7.0 0 -0.38 Zn-Cu 7.0 0 -0.27 Fe-Cu 7.0 0 -0.91 Fe-Zn-Pb 7.0 0 -0.89 Pb-Zn-Cu 7.0 0 -0.03 Fe-Zn-Cu 7.0 0 -0.85 Fe-Zn-Cu-Pb 7.0 0 -0.80 thereby removing them from contact with the test organism. In this way, the true effects of the other trace metals are not being seen, as they are being removed from solution i n some way by this ferric precipitate. For this reason and those mentioned previously (ie. the low toxicity of Fe by i t s e l f ) , Fe was not used i n the subsequent experimentation. Further research i s required to determine i f this i s what i s happening in the test solutions, and whether or not this process could occur in the natural environment. While a l l trace metal combinations run i n bioassay i n this phase produced an LC 5 0 toxicity value, an examination of the S values for these combinations reveals that, with very few exceptions, the trace metals were more toxic by themselves than they were i n combination with each other. Of the ten combinations used, eight produced a negative S value when comparing individual toxicity to combined toxicity, and only the Pb/Cu and Pb/Zn combi-nations showed any synergism (combined toxicity being greater than the sum of the individual toxicities), as evidenced by a relatively high positive S 104. value (6.30 and 1.27, respectively). The high degree of synergism between Pb and Cu, as suggested by the extremely high toxicity exhibited by the pair in bioassay (LC 5 0 = <1.0%), is at the same time both notable and unique, in that this pair is the only one to show any strong associative effect (positive or negative); i t also appears that, of the metals combinations used in bioassay, very few show evidence of having any enhancing or decreasing effect on the toxicity of the overall test mixture. In general, mean S values of -0.5 are found, indicating a trend toward slightly decreased toxicity with metals combinations as opposed to individual trace metals, and in turn perhaps indi-cating some of the reasons behind the lack of toxicity found in some of the stormwater samples of phase one. The analysis of toxicity data and any trends apparent from this data is perhaps better suited to the use of TRP regression analysis than simply the use of the Marking-Dawson index. This index, while useful in looking at relationships between individual components of a mixture, does not allow for any overall conclusions to be drawn about the mixture. Through the use of both analyses, however, one should be better able to explain the observed toxicity. Table 4-14 presents the results of TRP analysis for this data in the form of both independent variables and multiple regression equations. Again, i t i s d i f f i c u l t to draw definite conclusions from this table. The only inde-pendent variable seen to be of any significance in any of the metals groups (singles, pairs, triplets and quadruplet) is Fe, which as previously discussed, might not be considered to be a toxic trace metal at the concen-trations used here; i t is included as a variable mostly due to i t s s i g n i f i -cantly high concentration in the test solutions (in a s t a t i s t i c a l sense). A l l other trace metals seem to have minimal correlation to toxicity when TABLE 4-14 TRP ANALYSIS ON TRACE METAL TOXICITY DATA CONSIDERED1 Fe Zn Pb Cu pH REGRESSION EQUATION R2 F-PROB F-PROB LEVEL STANDARD ERROR LC 5 Q Metal 1 .93 -.51 .03 -.45 — Toxicity LC5Q=34.30 0 1.00 0.05 36.80 Metal 2 .74 -.67 -.09 .01 — Toxicity LC50=33.20 0 1.00 0.05 43.36 Metal 3 .86 — -.86 — — Toxicity LC50=25.27 0 1.00 0.05 22.00 Metal 1+2+3+4 -.11 -.23 — Toxicity LC50=30.49 + 1.41Fe - 11.48Zn 0.73 0.0007 0.05 18.91 Data defined in Table 4-12 106 . considered in groups (METAL 1,2,3 and 4, as designated i n Table 4-12). In addition, i n three of the four multiple regression analyses, the groupings of the various trace metals show no significant relationships whatsoever (with the regression equations consisting of a constant); only in the case of a l l four test metals being used (and then only i n one replicate) i s a significant regression equation produced, and again Fe i s included as a significant vari-able i n this equation. Generally, therefore, conclusions are hard to draw from these trace metal groupings, as there appears to be very poor correlation between the individual trace metals and toxicity. Perhaps i t would be best to examine the results of the individual bioassays, in an attempt to define any meaning-fu l and significant relationships. F i r s t , however, the trace metal-pH-suspended solids interactions data are presented. ( i i ) Trace Metal-pH-T.S.S Interactions: The individual bioassay results for trace metal-pH, trace metal-suspended solids, and trace metal-suspended solids-pH interactions are presented in Tables 4-15 to 4-17 respectively. Please note that only those trace metal combinations which proved to be the worst-case combinations i n previous experimentation were used here (Zn, Pb/Zn and Cu/Pb/Zn) Again, a l l combinations of trace metals and pH-solids parameters produced toxicity values, with LC 5 0's ranging from a high of <1.0% to a low of only 12.8% ( s t i l l a relatively high toxicity when compared with the samples examined in phase one). The actual trace metal concentrations corresponding to these toxicity levels would be i n the microgram per l i t r e range, suggesting that these synthetic solutions are quite highly toxic, although there i s no comparable data in the literature to confirm this. The effects of the varying levels of pH and suspended solids, both 1 0 7 . TABLE 4-15 TRACE METALS-pH INTERACTION RESULTS1 RUN2 VARIABLE PARAMETER pH TRACE METAL COMBINATIONS AND CONCENTRATION3 Cu Pb Zn LC 5 Q% A 5.0 — — 3.20 12.8 B 5.0 — 4.14 3.20 2.06 C 5.0 0.45 4.14 3.20 4.20 D 5.5 — 3.20 1.50 E 5.5 — 4.14 3.20 <1.00 F 5.5 0.45 4.14 3.20 1.18 G 6.0 — 3.20 2.80 H 6.0 — 4.14 3.20 5.71 I 6.0 0.45 4.14 3.20 2.90 J 6.5 — 3.20 5.70 K 6.5 — 4.14 3.20 9.10 L 6.5 0.45 4.14 3.20 2.81 M 7.5 — — 3.20 6.50 N 7.5 — 4.14 3.20 7.00 0 7.5 0.45 4.14 3.20 4.00 P 8.0 — 3.20 7.60 o 8.0 — 4.14 3.20 6.85 R 8.0 0.45 4.14 3.20 4.10 1 TRP Designation: pH Effects 2 A l l runs at TSS = 0 mg/L 3 A l l values i n mg/L. individually and i n combination, on trace metal toxicity are not easily discernable from the tables; therefore Figures 4-14 to 4-16 present the data in the form of graphs of the toxicity of the samples versus the pH or suspen-ded solids content. From these, the relationship between the examined para-meters becomes quite evident. Figure 4-14 shows the effects of pH on the toxicity of the trace metal combinations Zn, Pb/Zn and Cu/Pb/Zn, and in a l l cases toxicity decreases with increasing pH. This could be expected because, as suggested by Gadd and Griffiths 108. Figure 4-14 Trace Metal - pH T o x i c i t y Relationship. 109. (1978) among others, a lower pH w i l l generally increase the av a i l a b i l i t y of of metal ions, while a higher pH w i l l decrease this a v a i l a b i l i t y . The mecha-nism of decreased-increased avai l a b i l i t y differs for different metals at different pH, but, as a generalization, at an acid pH trace metals exist as free ionic cations, while at a higher alkaline pH these cations may precipi-tate out of solution as generally insoluble hydroxides or oxides, i n much the same manner as the Fe precipitation observed earlier. For example, Zn may precipitate as Zn(0H)2 at pH>5, and at pH>8.5, i t may form zincate ion which i s precipitated by calcium. Exactly at what pH this precipitation occurs i s governed by a number of other factors, i e . , temperature, redox potential, etc., and these pH's are general guidelines. As well, some metals such as Cu have more than one valence state and the oxidized form i s favoured at high pH; the hydroxides of these oxidized states are less soluble than those of reduced states and therefore precipitate at low pH. Figure 4-14 would appear to suggest that this sort of process i s taking place within these samples. At the lower pH (5.0), Zn i s the more toxic trace metal; however, as the pH rises, the Zn sample toxicity greatly decreases, suggesting that the availability of the metal i s diminished at these higher pH's; the Pb/Zn combination exhibits the same trend, i n that at low pH the toxicity of this combination i s relatively the same as Zn alone, and as pH rises, the toxicity decreases i n a manner again similar to Zn alone. This suggests that Zn i s dominant in the Pb/Zn pairing, and thus the toxicity of this pair follows the toxicity of Zn alone. On the other hand, the Cu/Pb/Zn combination exhibits a much smaller diminishment of toxicity at high pH, and i s i n turn the most toxic combination at the highest test pH (8.0). This would seem to suggest that the antagonistic nature of Zn i s overcome In the mixture of Pb and Cu, resulting i n a more generally toxic 110. mixture. Marking-Dawson S values (Table 4-18) would seem to confirm this, i n that there i s a more generally synergistic interaction at most of the pH levels tested for the Cu/Pb/Zn combination. In general, from Table 4-15 i t appears that the optimum pH at which high toxicity i s occurring for a l l metal combinations i s 5.5, at which there are very strong synergistic interactions indicated (S = 1.19 to 4.43), with LC 5 0 toxicity values for a l l combinations less than 1.5%. Table 4-16 presents the interactions of suspended solids on trace metal toxicity for a l l samples run. While the results must be viewed with caution, due to the limited number of replicates and the inherent high variability built into the experimentation through the use of an optimized "average" T.S.S. range of 50 to 200 mg/L (as dictated by phase one results), certain general observations can be made i n conjunction with the Marking-Dawson S values for these sample runs (Table 4-18). The literature suggests that one can expect to see a generally decreased toxicity with increasing solids con-tent, due to the a b i l i t y of the sediment to bind trace metals differentially (depending on the cation exchange capacity of the solid in question), making them unavailable for uptake by aquatic organisms. Some studies suggest that certain oxidized sediments can bind up to 96% of an added metal (Gadd and Griffiths, 1978), while others have found that up to 99% of trace metals may be bound by present solids (Wilber and Hunter, 1979). One could reasonably expect to see this same pattern (diminished toxicity with increased solids content) in these samples; however, from Figure 4-15 i t would appear that exactly the opposite i s happening. The slope of the straight lines suggest that there i s an increased toxicity with increased solids content. Preliminary bioassays, designed to look at the toxicity of the diatomaceous earth used as the solid i n this experimentation, indicated 111. TOXICITY C L C 5 0 % ^ 50 100 150 200 T O T A L S U S P E N D E D SOLIDS (mg/L ) Figure 4-15 Trace Metal - T.S.S. T o x i c i t y Relationship. 112. TABLE 4-16 TRACE METALS-TOTAL SUSPENDED SOLIDS INTERACTION RESULTS1 RUN2 VARIABLE PARAMETER T.S.S. TRACE METAL COMBINATION AND CONCENTRATION3 Cu Pb Zn LC 5 Q% A 200 — — 3.20 6.48 B 100 — — 3.20 6.70 C 50 — — 3.20 8.71 D 200 — 4.14 3.20 5.60 E 100 — 4.14 3.20 8.49 F 50 — 4.14 3.20 7.55 G 200 0.45 4.14 3.20 3.46 H 100 0.45 4.14 3.20 3.46 I 50 0.45 4.14 3.20 5.45 1 TRP Designation: Sediment Effects 2 A l l runs at pH 7.0 3 A l l values in mg/L that i t was completely non-toxic to JD^  pulex, and could safely be used for the desired purpose. Therefore, the reasons for this reversed trend would appear to l i e elsewhere in the data. One possible explanation for this seeming anomaly may l i e in the fact that the test organisms are essentially filter-feeders. As such they would be expected to ingest any adsorbed trace metals when they feed on the parti-culate matter present in the water column. With a higher concentration of trace metals in the system, coupled with a high particulate content, one could reasonably expect more trace metal adsorption and, upon ingestion by the organism a greater potential for toxicity. In this event i t is found that greater toxicity is occurring in the presence of high particulate matter content. Numerous studies have found much the same results with respect to 113. pesticide adsorption and bioaccumulation within aquatic organisms (summarized in Reese et a l . , 1972; and U.S. EPA, 1972), and one may therefore assume that the same kind of mechanism might be active here. This process i s indicated in Figure 4-15, with the trace metal group Cu/Pb/Zn showing the highest toxicity (lowest LC 5 0), and the groups Pb/Zn and Zn showing approximately the same toxicity. In general, while the conclu-sions must necessarily be limited to these results only, i t would appear that the Cu/Pb/Zn trace metal combination i s the most toxic to I), pulex, and this group i s less influenced by the suspended solids than are the other trace metal groups (as evidenced by the higher toxicity of Cu/Pb/Zn). A l l group toxicities are reduced to some degree by the presence of these suspended solids, however, as indicated by the fact that a l l Marking-Dawson S values for these combinations are negative (Table 4-18). A comparison of the combined effects of pH and suspended solids on trace metal toxicity was carried out and the results are presented i n Table 4-17 and Figures 4-16 and 4-17. Although necessarily brief in scope, using only the two worst-case trace metal conclusions Pb/Zn and Cu/Pb/Zn, some meaning-ful conclusions may be drawn. From the figures, i t appears that pH i s con-troll i n g the shape of the curves, with a l l showing very high correlation to straight lines; as well, again one sees the relationship of high toxicity at lower pH, with correspondingly decreased toxicity as pH increases. On the other hand, the amount of solids in the sample appears to be the determining factor i n the positioning of the individual curves, i n much the same manner as i t did i n the "solids only" effects on toxicity. One also sees that, as a general rule, with more solids present i n the sample, the higher w i l l be the overall toxicity. Again caution must be used in the application of this data, i n that i t may only be unique to this particular solid (which would 114. TABLE 4-17 TRACE METALS - pH - TOTAL SUSPENDED SOLIDS INTERACTION RESULTS1 RUN VARIABLE pH PARAMETERS TRACE METAL COMBINATION T.S.S AND CONCENTRATION2 Cu LC 5 0% Pb Zn 4.14 3.20 3.44 4.14 3.20 2.04 4.14 3.20 5.63 4.14 3.20 4.68 4.14 3.20 4.89 4.14 3.20 5.72 4.14 3.20 5.24 4.14 3.20 1.98 4.14 3.20 6.40 4.14 3.20 5.56 4.14 3.20 6.21 4.14 3.20 6.90 4.14 3.20 6.88 4.14 3.20 5.57 4.14 3.20 7.95 4.14 3.20 7.03 4.14 3.20 7.43 4.14 3.20 8.48 A 5.0 200 — B 5.0 200 0.45 C 5.0 100 — D 5.0 100 0.45 E 5.0 50 — F 5.0 50 0.45 G 6.0 200 — H 6.0 200 0.45 I 6.0 100 — J 6.0 100 0.45 K 6.0 50 — L 6.0 50 0.45 M 8.0 200 — N 8.0 200 0.45 0 8.0 100 — P 8.0 100 0.45 Q 8.0 50 — R 8.0 50 0.45 1 TRP Designation: Combined Effects 2 A l l values i n mg/L. TABLE 4-18 MARKING-DAWSON INDEX ANALYSIS RESULTS; TRACE METALS -pH-T.S.S INTERACTIONS VARIABLE PARAMETERS S Cu-Pb-Zn pH T.S.S Zn Pb-Zn 5.0 0 -0.99 1.64 -0.23 5.5 0 3.27 4.43 1.19 6.0 0 1.29 -0.05 0.17 6.5 0 0.12 -0.68 0.21 7.5 0 -0.02 -0.29 -0.18 8.0 0 -0.19 -0.26 -0.21 2 2001 -0.01 -0.03 -0.02 100 -0.05 -0.56 -0.02 50 -0.36 -0.39 -0.60 5.0 200 - 0.58 -0.02 5.0 100 - -0.04 0.03 5.0 50 - 0.11 -0.09 6.0 200 - 0.04 -0.10 6.0 100 - -0.18 -0.05 6.0 50 - -0.14 -0.01 8.0 200 - -0.27 -0.12 8.0 100 - -0.46 -0.10 8.0 50 — 0.37 -0.02 1 TRP Designation: Sediment Effects 2 Runs at pH 7.0 115. 116. 10.0 5.0 5.5 6.0 6.5 pH 7.0 7.5 8.0 &5 Figure 4-17 Trace Metal - T.S.S. For Cu/Pb/Zn. - pH T o x i c i t y R e l a t i o n s h i p 117. have to be verified by further experimentation). In general, however, i t appears that these synthetic samples are most toxic at low pH and high suspended solids concentrations. Finally, Table 4-19 presents the results of the TRP analyses run on a l l samples i n the pH and suspended solids effects investigations. While i t appears that there i s no regression equation relating pH effects to toxicity, with the regression equation consisting of a constant (indicating that, in spite of the linearity of the relationship between pH and toxicity, this relationship contains no significant independent variables that would relate pH to toxicity i n an equation form), regression equations were produced for the effect of solids alone and i n combination with pH on trace metal toxi-c i t y . In both equations, suspended solids i s included as a significant inde-pendent variable, indicating that the suspended solids have a pronounced effect on trace metal toxicity. In addition, Cu seems to be an important controlling factor where sediment effects on toxicity are concerned, suggest-ing a strong correlation between Cu, sediments and total toxicity, and pH i s found to be strongly significant when a l l effects are considered. The lack of inclusion in the regression equations of other variables doesn't necessar-i l y mean that these variables are not important; i t only indicates that they are not significantly relatable to the dependent variable i n this particular set of experimental circumstances. Most importantly therefore, i t appears that the pH and solids content, factors which control* the av a i l a b i l i t y and form of trace metals such as Pb and Cu in solution, are greatly responsible for the observed toxicity. 4.2.4 Conclusions: Laboratory Versus Field Data The laboratory derived bioassay data seems to suggest that the trace metals Cu, Pb and Zn, at the concentrations found i n the f i e l d samples and TABLE 4-19 TRP ANALYSIS ON pH AND T.S.S. EFFECT ON TRACE METAL TOXICITY DATA CONSIDERED1 Fe Zn Pb Cu PH REGRESSION EQUATION R2 F-PROB F-PROB LEVEL STANDARD ERROR LC 5 0 pH Effects — — .30 -.40 .26 Toxicity LC 5 Q= 4.88 0 1.00 0.05 3.10 Sediment Effects — — -.05 — — Toxicity LC 5 Q= 8.80 - 0.13TSS - 6.96Cu 0.87 0.002 0.05 0.81 Combined Effects r ind« 10 significant 'pendent variables Toxicity LC50=1.24 + 0.90 pH - 0.01TSS 0.87 0.002 0.05 0.58 Data defined in Tables 4-15, 4-16 and 4-17 119. moderated by such factors as pH and total solids content, are largely respon-sible for the toxicity observed within the D. pulex test populations. How-ever, this conclusion cannot be a l l inclusive, i n that other physiochemical properties of the stormwater sample, not easily amendable to the kind of bioassay testing that was carried out here (i.e., hardness, hydrocarbon con-centration, C.O.D., etc.) may also be contributing to the overall toxicity of the stormwater. Until these properties can be tested individually and i n conjunction with each other and their effects on toxicity delineated, one may make only limited conclusions based on this data. It appears, however, that the trace metal content does have a large influence on the acute toxicity of this synthetic urban stormwater runoff. Comparison between the lab-derived and f i e l d bioassay data, while l i m i -ted by the constraints mentioned previously (Winner and Farrel l , 1976; Stephan and Mount, 1973; and Sprague, 1976), seems to further confirm the assumption of trace metal toxicity. While the f i e l d data generally produced a lower acute toxicity value than the lab-derived sample data (a high of 10% and low of 60%) this would be expected for a number of reasons: 1) the lab data comes from experimentation under ideal, worst-case conditions based on the overall composite f i e l d data; however, i n the f i e l d the total overall trace metals content of the individual samples i s somewhat lower than those used i n the lab and, i f the hypothesis of trace metal toxicity i s valid, one would expect to see subsequently lower f i e l d toxicity, which i s i n fact the case; 2) there i s an increased presence of easily measured natural inter-ferences to trace metal toxicity i n the f i e l d samples and, as was shown with pH and suspended solids, these may work to lower the overall toxicity of the sample; and 3) the presence of unseen natural interferences, i . e . trace organics like PAH's, would be expected to be greater in the f i e l d samples and 120. as such may be expected to exert a potential antagonistic/synergistic i n f l u -ence on trace metal presence, thereby changing the overall toxicity of the sample; this change may be i n the form of increased/decreased acute toxicity. Generally, however, the data shows that trace metals are responsible for a high degree of toxicity i n the lab samples, and one must assume that this influence i s present i n the f i e l d samples as well, although i n some altered (enhanced/decreased) state, due to natural interferences. It has been shown that, these interferences tend to generally decrease trace metal toxicity i n the lab, and i n natural f i e l d conditions the same kind of decrease might be assumed (based on pH and suspended solids data). This i s a limited assump-tion, as "natural f i e l d conditions" encompass a wide variety of conditions and, consequently, effects, i.e. at pH 5.5 trace metal toxicity i s enhanced, but at 7.0 and above, the toxicity i s somewhat depressed. The more complex a solution i s , the more d i f f i c u l t i s i t to model; however, based on these results, the importance of trace metals to the acute toxicity of the storm-water samples i s evident. 4.3 Phase Three: "First-Flush" Analysis Much of the literature reviewed i n Chapter 2 makes reference to the "fi r s t - f l u s h " of a storm event, wherein most of the pollutant load from the storm w i l l be washed out of the system within the f i r s t hour of runoff (Wilber and Hunter, 1977; Kaufman and Lai, 1981; Digiano et a l . , 1975; Helsel et a l . , 1979; G r i f f i n et a l . , 1980; Mi l l s , 1977). If this i s i n fact what happens, one might expect this slug load to have a detrimental effect on indigenous aquatic l i f e i n the water body receiving the runoff. However, as discussed previously, the exact effects of this kind of pollution loading are extremely d i f f i c u l t to ascertain when, in,fact, the existence of the " f i r s t -1 2 1 . flush" effect i t s e l f i s s t i l l a source of debate among researchers. The purpose here i s to examine one individual storm event through i t s duration, i n an attempt to define the various stages of the storm and to confirm or deny the existence of the "f i r s t - f l u s h " effect as far as pollutant loadings are concerned. To this end, Figures 4-18 to 4-24 present the data i n the form of pollu-tographs, or the flow influenced change of an individual parameters' concen-tration i n the runoff through time. Investigated parameters include total suspended solids, conductivity, total alkalinity, total hardness, pH and COD, as well as the trace metals copper, iron, lead, nickel and zinc. In addi-tion, calcium and magnesium were measured through the storm, and plotted on pollutographs. Finally, bioassays were run on a l l samples obtained, and the resultant LC 5 0 toxicity values plotted on a pollutograph, i n order to observe changes i n stormwater toxicity over the duration of the r a i n f a l l event. Rainfall amount (as an exact measurement, courtesy of G.V.R.D., 1980) and flow level (an arbitrary high-low scale) were also plotted. Each sample was divided into an "as i s " fraction and a " f i l t r a t e " fraction, to give some idea of the relative amounts of total and dissolved material present. In addi-tion, once per hour a sample was taken and the materials i n i t acid extracted to give a total acid extractable sample. This data was, however, of limited use for a complete analysis and i s therefore not included in this discussion. According to G r i f f i n et a l . , (1980), the pollutant phase i.e., soluble or particulate, i s a major factor i n determining whether or not a " f i r s t -flush" of that pollutant occurs during a runoff event. This, they suggest, i s partially explained by the factors responsible for pollutant removal. For example, the insoluble pollutants are generally removed by physical RAINFALL (Jn/hrxK)"2) 1 A 800 700 600 500 400 120l 90 60 30 "AS IS"SAMPLE I TOTAL! 60 120 180 240 300 TIME OF STORM IMIN.I HIGH • TOTAL SUSPENDED SOLIDS (mg/C) FLOW A CONDUCTIVITYfuS/cm 1) LOW ^ r ' ' "FILTRATE"SAMPLE (DISSOLVED! 60 120 T55 235" TIME OF STORM (MIN.I 355" Figure 4-18 Pollutograph: Total Suspended Solids and Conductivity. Figure 4-19 Pollutograph: Total A l k a l i n i t y and Total Hardness. Figure 4-20 Pollutograph: pH and COD, RAINFALL (in/hrx10" 2) A 0.3o| 30| 0.20 0.10 20j 10^ "AS IS"SAMPLE I TOTAL 60 120 180 240 TIME OF STORM (MIN.] 300 HIGH FLOW LOW • CALCIUMfmg/O A COPPER(mg/Q FILTRATE"SAMPLE IDISSOLVEDI 60 120 180 "240" TIME OF STORM IMIN.] 300 Figure 4-21 Pollutograph: Calcium and Copper. ON Figure 4-22 Pollutograph: Iron and Magnesium. N3 Figure 4-23 Pollutograph: Lead and Nickel. Figure 4-24 Pollutograph: Zinc and LC 5 0 Toxicity. 129. processes, and the majority of these would probably be entrained during the rising limb of the runoff hydrograph and thus exhibit a " f i r s t - f l u s h " characteristic. On the other hand, soluble pollutants removal i s more gener-al l y governed by solubility equilibria, exchange capacity, and adsorption-desorption processes; i n these cases, the physical scouring removal mechanism would not affect their concentrations when they are already i n the runoff stream. Therefore, an Insoluble pollutant like suspended solids would exhi-bit " f i r s t - f l u s h " characteristics, while a soluble parameter like phosphate would not. This type of pattern i s definitely reflected by this data, and i s most readily observed in Figure 4-18, in which total suspended solids and conduc-t i v i t y concentrations are plotted against time. The insoluble suspended solids show a very strong " f i r s t - f l u s h " i n the total samples (and, as would be expected, the dissolved samples contain much lower suspended solids con-centrations, which have been filtered out of solution); conversely, the dissolved solids, as measured by the conductivity (discussed in Section 4.1.1(d)), show less of a " f i r s t - f l u s h " i n the total samples, although there does appear to be a peak concentration reached at approximately two hours -this would seem to be the point at which the solubility equilibria of these particular dissolved solids allow them to be washed out of the system in the stormwater runoff. By s t r i c t adherence to the definition of the " f i r s t -flush", i.e., the majority of the pollutant i s washed out of the system i n the f i r s t hour, one can say that the soluble dissolved solids do not exhibit " f i r s t - f l u s h " characteristics, as would be expected according to G r i f f i n e£ a l . A reasonably similar pattern to the total samples i s exhibited for these dissolved solids i n the dissolved samples' pollutograph, again indicative of the soluble nature of these solids. 130. Other parameters' concentrations plotted again cumulative storm runoff time include total alkalinity and hardness, and COD (Figures 4-19 and 4-20). Once again the differences between soluble and insoluble pollutants (and the subsequent differences in removal mechanisms in the stormwater runoff) can be seen from these figures. Both total alkalinity and total hardness, essenti-a l l y comprised of the measurement of the HCOg buffering system and Ca + 2 and Mg+2 ionic species, respectively, would have to be considered as dissolved soluble species at the neutral pH's found in these particular samples (Figure 4-20). As such one would not expect to find the characteristic " f i r s t - f l u s h " runoff pattern i n the pollutographs for these species. While total a l k a l i -nity seems to show the characteristic pattern, i n that the highest concentra-tions are found at the beginning of the storm, there i s no buildup to this peak concentration i.e., i t appears that this peak i s residual alkalinity from the dry weather deposition period. Since the dissolved sample pattern i s almost identical to that of the total sample for alkalinity, i t must be concluded that most of the alkalinity i s present i n the dissolved soluble state, and there i s no " f i r s t - f l u s h " runoff occurring for this parameter. The reasons for the seemingly characteristic pattern would appear to l i e elsewhere. The pollutograph of total hardness seems to show the delayed peak found with the dissolved solids, and from this and the identical pattern found i n both the total and dissolved samples, i t can be concluded that the total hardness parameter exhibits no " f i r s t - f l u s h " runoff pattern. Further, the pollutograph for COD (Figure 4-20) seems to exhibit the characteristic " f i r s t - f l u s h " pattern for the total samples, albeit in a somewhat altered form i n that there i s a bimodal peak concentrational pattern associated with the oxygen-demanding carbonaceous material. The f i r s t peak, at the 70 minute 131. mark f i t s the theory of the majority of the pollutant being washed out i n the f i r s t hour of runoff; the second peak, at the 180 minute mark, would appear to be the result of one unusually high COD measurement i n the third hour of the storm. This value, while i t does suggest the presence of a second peak i n the pollutograph, i s nonetheless an isolated case and, as such, should be overlooked i t i n favour of the more general pattern exhibited over the entire duration of the storm. Therefore, i t can be concluded that COD does exhibit "fir s t - f l u s h " tendancies, something that would be expected when one considers the nature of the COD-contributing substances, (vegetal material and other insoluble organics) and the mode by which these are removed during the storm event, i e . , by physical scouring processes usually associated with the f i r s t part of the runoff, once the " c r i t i c a l erosion velocity" i s achieved. This i s further suggested by the somewhat reduced COD concentrations found in the dissolved sample pollutograph, indicative of the insoluble nature of the pollutant. A general pattern of soluble versus insoluble pollutants emerges when the theories of G r i f f i n et a l are considered. The pollutographs examined so far suggest that COD and T.S.S. exhibit " f i r s t - f l u s h " characteristics, while alkalinity, hardness and dissolved solids (conductance) do not. This f i t s well with G r i f f i n et a l . , from which one would expect COD and T.S.S. to be measures of insoluble, physical parameters (less so i n the case of COD, but the theory i s s t i l l applicable), while the others would be soluble chemical parameters not expected to exhibit the " f i r s t - f l u s h " characteristic due to differences i n removal mechanisms. This difference i s also quite evident from the pollutographs in that, in the case of the insoluble pollutants, the total and dissolved graphs are markedly different, with the dissolved sample concentrations being much lower. However, the graphs for soluble pollutants 132. are almost identical, indicating that the dissolved fraction is the major component making up the total sample. This readily observable difference may be useful for identifying the nature and expected runoff pattern of specific pollutants. Figures 4-21 to 4-24 present pollutographs for the trace metals Ca, Cu, Mg, Pb, Ni and Zn, respectively, as well as a graphic representation of the toxicity of the stormwater (LC 5 0%) through the duration of the storm. Again the difference between soluble and insoluble species is somewhat apparent; as well, from the discussion of land use influences versus buildup time i n f l u -ences on trace metal concentrations (section 4.1.2 and Tables 4-3 and 4-4) one can see the differences between the various trace metal species. From that discussion i t was concluded that Fe, Cr and Ni were more closely i n f l u -enced by and associated with the solids content of the runoff, as a direct result of the characteristics of the sampling site but not the activities ongoing at that site; on the other hand, one found that Zn, Pb, Ca and Cu were closely associated with buildup time and, subsequently, with the speci-f i c activities ongoing at the sampling site. Mg, Cd and Mn have an equal association with both parameters. This being the case, one could expect to see these associations reflected in the pollutographs of these specific trace metals, as a function of the solubility/insolubility of the particular species. An examination of the individual total sample pollutographs con-firms this. Fe exhibits a very pronounced " f i r s t flush" runoff pattern (almost an ideal runoff), while Ni shows a very erratic runoff pattern throughout the storm, with the average concentration changing greatly between each sample. However, according to the previous discussion, one would have to say that there is a probable " f i r s t - f l u s h " because of the difference between the total and dissolved sample pollutographs for Ni, with the 133. dissolved samples showing lower concentrations, indicating that most of the Ni i s in a insoluble form (probably adsorbed to the particulate material of the runoff). The pollutographs for Zn, Pb, Ca and Cu, while erratic and far from ideal, suggest the absence of any "fi r s t - f l u s h " runoff, by definition. In most cases, the total and dissolved sample pollutographs are approximately equal i n runoff pattern, and lacking in any discernible " f i r s t - f l u s h " runoff within the f i r s t hour of the storm. As well, there appears to be a peak concentration reached for each metal anywhere from 1.5 to 3 hours after the start of runoff, and this "delayed" peak, indicative of the different removal mechanisms involved for soluble pollutants, further confirms the lack of "fi r s t - f l u s h " runoff. Therefore, i f a s t r i c t definition of the "f i r s t - f l u s h " runoff pattern i s adhered to, i t must be concluded that trace metals like Zn, Pb, Ca and Cu, as well as chemical properties like alkalinity, hardness and conductivity, do not exhibit " f i r s t - f l u s h " runoff tendancies, while the insoluble parameters like solids content, Fe and Ni do. However, one must be aware that the para-meters that don't exhibit " f i r s t - f l u s h " runoff do nonetheless reach a peak concentration i n the runoff flow, albeit on a somewhat delayed basis. This becomes very important in planning considerations for the treatment of storm-water runoff, since, i f only the "f i r s t - f l u s h " runoff were to be collected for treatment, many of the soluble, potentially toxic components of the run-off would be allowed to freely enter the aquatic environment untreated. This i s further confirmed when one examines the acute toxicity of this stormwater runoff through the duration of the storm (Figure 4-24). From this pollutograph, i t i s seen that there i s no definable pattern of acute toxicity to D. pulex, with many of the total samples proving completely non-toxic to 134. the organism. However, the most acutely toxic stormwater runoff i s coming from the dissolved samples (LC 5 0 as low as 14%) as much as 3-4 hours after the start of runoff, indicating the highly toxic nature of these soluble parameters. One might have expected this from the results of phase two experimentation (section 4.2.3), i n which i t was found that the trace metals Zn, Cu and Pb were highly toxic to D. pulex; these same soluble trace metals are evidently responsible for the toxicity observed here. It therefore becomes apparent that the most harmful stormwater runoff may not necessarily be associated with the " f i r s t - f l u s h " alone, and the complete treatment of a l l stormwater runoff would seem to be the only feas-ible alternative. While Helsel et a l . , (1979) suggest that the average i n c i -dence of " f i r s t - f l u s h " runoff increases with the increasing degree of urbani-zation (attributable to the fact that impervious subbasins produce a dwindling source of pollutants during a storm event while the forested sub-basin sources remain essentially constant through the storm), the presence of the "delayed" flush of toxic soluble pollutants, while probably constrained by the same pattern, would appear to confirm the need for complete treatment of a l l stormwater runoff i n the urban environment, i n the author's opinion. 135. CHAPTER 5 Conclusions and Recommendations for Further Research  5.1 Conclusions The following conclusions regarding the effects of land use on urban stormwater runoff and i t s toxicity to aquatic organisms may be drawn from this research: 1. From preliminary analysis of stormwater runoff i t was found that the chemical oxygen demand, total alkalinity and hardness, and conductivity (as a measure of dissolved solids) were linearly related to the dry weather buildup period preceeding a storm event. This was true for most samples taken from the commercial (C) and residential (R) land use areas as well as, to a lesser extent, those from the industrial (I) areas. The open/ greenspace areas (0) produced samples with lower, "background" values expec-ted of a clean, unpolluted area. This suggested that a more polluted runoff w i l l result from a longer dry period for these parameters, regardless of the specific land use classification of any sampling site (C,R,I or 0), but as a direct result of the specific individual ac t i v i t i e s ongoing at that site which are contributing to the pollutant load. These ac t i v i t i e s are non "class-dependent" and may be occurring equally i n any of the land use classes sampled. It was noted that while the non "class-dependent" act i v i t i e s contributing to stormwater runoff pollutant loads are free from the external influences of land use class, they are s t i l l constrained by the internallized factors defined by Rimer et a l . , (1978). 2. It was seen that hydrocarbon accumulation (measured as iso-octane) was related both to buildup time, like the previously mentioned parameters, as well as to the land use of the sampling si t e . This suggested that there 136. was some influence on the hydrocarbon content of stormwater runoff by "class-dependent" act i v i t i e s , which contributed more to the runoff with a longer • buildup period in the commercial and industrial areas, and less in the resi-dential and open areas. 3. The contributions to mean runoff COD from each land use class followed the sequence C>I>R>>0. However, a better comparison of COD loading, measured as COD buildup per day (mg/L-day COD) resulted in the sequence R>I= C>0. This sequence was expected because the residential areas offered the most favourable combination of impervious surface and source of oxygen demanding materials. These internal factors, defined by Rimer et a l . , are important in determining the potential for polluted runoff from any sampling site. 4. A l l preliminary samples had a relatively high dissolved solids content (as measured by conductivity), following the sequence I>>C = R>0. The presence of these anionic species (ie., CSL~ and SO^-2) as well as the species contributing to the non-carbonate hardness observed, were important to note due to the possible interactions between them and other parameters of the stormwater runoff. 5. From preliminary trace metals analysis, i t was seen that Cr, Fe and Ni concentrations were linearly relatable to the total suspended solids content of the stormwater runoff, while Ca, Cu, Pb and Zn were linearly rela-ted to the buildup time before sampling. Cd, Mg and Mn did not show a strong relationship to either parameter. This suggested that Cr, Fe and Ni concen-trations were controlled by non "class-dependent" activities which affected the TSS concentration, regardless of land use classificaiton. On the other hand, Ca, Cu, Pb and Zn concentrations were controlled by "class-dependent" activities occurring as a direct result of the land use classification of the 137. sampling site, indicating that these trace metal concentrations would be higher at one specific class of site i e . , commercial, rather than another. 6. The industrial and commercial land use sites were the source of the majority of those metals most often considered toxic to aquatic organisms. Cr was most prevalent in the industrial areas; Cd, Ni, Pb and Zn were most prevalent in the commercial areas, and Cu was found at equally elevated levels in both. The sequence of trace metal contamination was therefore" determined as C>I>R»0, with the commercial areas being the most contamina-ted. A l l sites in the commercial, industrial and residential sampling areas had trace metal contamination that was significantly different from the open/ greenspace areas, at a l l levels of s t a t i s t i c a l significance tested except one. It was therefore decided that the assumption of the open/ greenspace areas having background trace metal contamination was valid. Further, com-parison with data from the literature revealed that a l l sampling sites had average contamination for trace metals, as well as the other parameters investigated. 7. Results from the preliminary acute toxicity bioassays indicated that stormwater runoff from the commercial land use areas was the most toxic to D. pulex i n static 96 hour bioassays. The acute toxicity to D. pulex followed the sequence C>I>R, with a l l runoff from the open/greenspace areas proving non-toxic to the test organisms. It was not clear whether this toxi-city pattern was a reflection of the land use in the various sampling areas or simply a result of the drainage -patterns in the basin, with the outlying commercial areas receiving most of the drainage flow. However, review of the preceeding trace metal and physiochemical parameter data suggested that one could expect this sequence of toxicity. 8. The TRP (Triangular Regression Package) computer program was used 1 3 8 . in an attempt to isolate the toxic components of the runoff samples, but very l i t t l e correlation was obtained through either the partial linear regression analysis (INMSDC) or the step-wise multiple regression package (STPREG). There were indications that the toxicity of the runoff was related to buildup time in the industrial areas, and to pH in the residential and commercial sampling sites; this was not confirmed by the STPREG analysis, however. Further, the buildup time and pH variables in the toxicity equation were considered more indicators of toxicity than actual causes of i t , i e . , i f trace metals were responsible for the observed toxicity, the pH might be a controlling factor by regulating the availability and form of the metals. It was therefore not possible to isolate the toxic components of the stormwater runoff samples from preliminary sampling. However, there were indications that the toxicity occurred as a result of "class-dependent" activities (through correlation to buildup time and pH). 9. Cu, Fe, Pb and Zn were used in the manufacture of a synthetic stormwater, at the worst-case concentrations determined from preliminary sampling. These metals were bioassayed in combination with each other under constant conditions in an attempt to duplicate the runoff toxicity observed earlier. Varying conditions of hardness and hydrocarbon concentration were also used, but a combination of experimental and mechanical d i f f i c u l t i e s prevented this author from obtaining meaningful results, i t was noted, however, that the test organism (D. pulex) did not survive well at total hardness conditions much above those of i t s natural habitat (approximately 40 mg/L as CaC03). In a l l bioassays involving trace metals under constant hard-ness and pH conditions, results indicated that the metals were more toxic individually than in combination, but a l l proved acutely toxic to some degree (a high of <1.00% to a low of 92.0%, measured as LC s n%). This was confirmed 139. strongly synergistic association while Pb-Zn was less strongly synergistic; a l l other combinations showed slightly antagonistic interactions. Analysis of this data using TRP did not confirm this, and there was very l i t t l e correla-tion using either INMSDC or STPREG, although there were indications that Fe and Zn were contributing to toxicity. 10. A closer approximation of natural stormwater runoff was attempted using the trace metals Cu, Pb and Zn under varying conditions of TSS and pH. Fe was not used, due to questionable results obtained earlier. A l l trace metal combinations under a l l conditions of pH and TSS were acutely toxic to D. pulex to some degree (a high of <1.00% to a low of 12.8%, measured as LC 5 Q%). It was seen that the presence of suspended solids under different pH conditions enhanced the toxicity of the trace metals being tested, with the overall toxicity of the mixture being maximized under conditions of low pH and high suspended solids. A l l combinations were most toxic at pH 5.5, the "maximum toxicity point", and i t was suggested that pH was controlling the form and availability of the trace metals. The relationship of higher toxi-city with higher suspended solids was unexpected; however, adsorption-desorption characteristics and the fact that the test organisms are f i l t e r -feeders explained this. TRP analysis confirmed the role of pH and TSS i n controlling the form and avail a b i l i t y of the toxic trace metals. Overall correlation between synthetic and natural stormwater runoff toxicity was poor; this was expected due to the differences between the f i e l d and labora-tory environments. The comparison did suggest, however, that the trace metals Cu, Pb and Zn were responsible for much of the toxicity observed in the preliminary runoff samples, regulated to some degree by the pH and suspended solids content of the runoff. 140. 11. An analysis of the magnitude of occurrence and toxic effects of the "fi r s t - f l u s h " of stormwater runoff was carried out. According to Grif f i n et a l . , (1980), the form of the pollutant (soluble or insoluble i n the runoff flow) i s important i n determining whether or not a "f i r s t - f l u s h " w i l l occur, due to the differences i n removal mechanisms between the two. This research confirmed this, in that the insoluble pollutants TSS, Fe and Ni (at the investigated pH) didn't show "f i r s t - f l u s h " characteristics, while the expec-tedly soluble Ca, Cu, Pb, Zn, alkalinity, hardness and conductance did exhi-bit classic " f i r s t - f l u s h " trends. The differentiation between soluble and insoluble parameters was made easier through the use of pollutographs., which plot the flow induced change In parameter concentration i n the runoff through time. For a l l pollutants considered, each sample was divided into a total fraction and a dissolved fraction, and these were plotted on pollutographs. It was observed from this analysis that i f the total and f i l t r a t e curves closely approximated each other, the pollutant could be classified as soluble, and no "fi r s t - f l u s h " would be expected; i f , however, the curves were markedly different, the pollutant was insoluble and "f i r s t - f l u s h " was occur-ring. 12. While the " f i r s t - f l u s h " of insoluble pollutants can be important, i t was suggested from the acute toxicity bioassays of these samples that the soluble pollutants (ie., Cu, Pb and Zn) were the most toxic to the test organisms and, from the pollutograph of toxicity, these may be flushed out of the system anywhere to 3 hours after the start of a storm event. These same . pollutants were responsible for the toxicity observed in the other phases of this research, and point up an obvious need for treatment of this toxic stormwater runoff. 141. 5.2 Recommendations 1. This study dealt exclusively with the short-term, acute toxicity effects of urban stormwater runoff to one aquatic invertebrate. The longer term chronic effects and modes of action on reproduction and life-cycle popu-lation dynamics on other invertebrates and fish species would seem to be the next logical step in the complete characterization of the toxicity of storm-water runoff. 2. The d i f f i c u l t i e s in comparing the results of different studies, and even between different sampling sites in the same study have been pointed out. These arise mostly from the inherent spatial and temporal differences of the individual sites, and demonstrate the discreteness of every site with-in a particular study. A long term study of one "representative" site from each land use category may provide some idea of the internal changes occur-ring within these sites over a period of time, and some characteristics of the specific sites that may be applied to the entire land use class may be derived. In conjunction with this, a more detailed breakdown of each land use site, the contributions to the pollutant runoff flow of the particular characteristics of Rimer et a l . , (1978), and more detailed information on the drainage patterns at each site would be necessary. 3. An in-situ investigation of the effects of stormwater runoff seems to be called for. This could be carried out in an isolated section of a river or stream over a period of time to see i f dilution effects are s u f f i -cient in reducing the toxic nature of the runoff, or i f , in isolated circum-stances and under conditions of low flow or l i t t l e mixing, toxic conditions can result, as well as seeing what changes through time and distance are occurring in the specific pollutants in this natural environment. 142. 4. Further investigations into the effects of hardness, hydrocarbons and iron concentrations on aquatic invertebrates i s needed. This study concluded that the test organisms were greatly affected as a result of the hardness of the test solution, and more information of these effects should be collected to determine whether or not hardness can be considered a toxic component at sufficiently high concentrations in urban stormwater runoff. In addition, the effects of the ferric ion precipitation and i t s a b i l i t y to remove other metals by adsorption processes should be looked at. 5. This research concluded that there i s a definite need for some kind of treatment of stormwater runoff, as i t s toxic effects are too pronounced to ignore. The specific mode of this treatment should be investigated, i e . , complete or partial treatment, separately or i n conjunction with domestic sewage, etc. 143. REFERENCES Adema, D.M.M., 1978. "Daphnia magna as a Test Animal in Acute and Chronic Toxicity Tests." Hydrobiologia, 59(2):125-134. American Public Health Association, 1978. Standard Methods for the Examina- tion of Water and Wastewater. 14th Ed., Washington, D.C. Applegate, V.C, J.H. Howell, A.E. Hall and M.A. Smith, 1957. "Toxicity of 4,346 Chemicals to larval lampreys and Fishes." Fish & Wildlife Service, Special Science Rep. Fish., 207:1-7. Barton, K., 1978. "The Other Water Pollution." Environment, 20(5):12-20. Beisinger, K.E. and G.M. Christensen, 1972. "Effects of Various Metals on Survival, Growth, Reproduction and Metabolism of Daphnia magna." J.  Fish. Res. Board. Can., 29_:1691. Bindra, K.S. and K.J. Hall, 1977. "Geochemical Partitioning of Trace Metals in Sediments and Factors Affecting Bioaccumulation i n Benthic Organisms." Ph.D. Thesis, University of British Columbia (in preparation). Bringmann, G. and R. Kuhn, 1978. "The Toxicity of Waterborne Contaminants Towards Daphnia magna." WRC Info., (G.B.) .5, 78-0262. Brown, V.M., 1968. "The Calculation of the Acute Toxicity of Mixtures of Poi-sons to Rainbow Trout." Water Res., 2:723-733. Brown, V.M. and R.A. Dalton, 1970. "The Acute Lethal Toxicity to Rainbow trout of Mixtures of Cu, Phenol, Zn & Ni." J. Fish. Biol., 2^ :211-216. Brown, V.M., 1973. "Concepts and Outlook i n Testing the Toxicity of Substances to Fish." In Bioassay Techniques and Environmental Chemistry, Glass, G.E.(ed.), pp. 73-95. Burdick, G.E., 1960. "The Use of Bioassays by the Water Pollution Control Agency." Biol. Problems Water Pollution, Trans, of Second Seminar, 1959, U.S. Public Health Service, R.A. Taft Sanit. Engng. Center Tech. Report W60-3, pp. 145-148. Cairns, J.Jr. and A. Scheier, 1968. "A Comparison of Toxicity of Some Common Industrial Waste Components Tested Individually and Combined." Prog. Fish. Cult., 30(1):3-8. Cairns, J.Jr. and K.L. Dickson, 1971. "A Simple Method for the Biological Assessment of the Effects of Waste Discharges on Aquatic Bottom-Dwelling Organisms." J.W.P.C.F., 43:755. Calamari, D. and R. Marchetti, 1973. "The Toxicity of Mixtures of Metals and Surfactants to Rainbow Trout." Water Res., 7^:1453-1464. 144. Canton, J.H. and D.M.M. Adema, 1978. "Reproducibility of Short-Term and Reproduction Toxicity Experiments with Daphnia magna and Comparison of the Sensitivity of D. magna with D. Pulex and D. cucullata in Short Term Experiments." Hydrobiologia, 59,(2):135-140. Daniel, T.C., McGuire, P.E., Bubenzer, G.D., Madison, F.W. and J.G. Konrad, 1978. "Assessing the Pollutional Load From Nonpoint Sources: Planning Considerations and a Description of Automated Water Quality Monitoring Program." Environmental Management, 2(1):55-65. Davies, P.H., J.P. Goettl Jr., J.R. Sinley and N.F. Smith, 1976. "Acute and Chronic Toxicity of Pb (Lead) to Rainbow Trout i n Hard and Soft Water." Water Res., 10:199-206. DiGiano, F.A. et a l , 1975. "A Projection of Pollutional Effects of Urban Runoff i n the Green River, Mass." In Urbanization and Water Quality °Control, Amer. Water Res. Assoc. Dixon, W.J. and F.J. Massey Jr., 1969. Introduction to Sta t i s t i c a l Analysis, 3rd Ed., McGraw-Hill Book Company, New York. Eaton, J.G., McKim, J.M. and G.W. Holcombe, 1978. "Metal Toxicity to Embryos and Larvae of Seven Freshwater Fish Species - I. Cadmium." Bull. Env.  Con, and Tox., 19(1):95-103. Federal Water Pollution Control Federation, 1968. Water Quality Criteria, Washington, D.C. Ferguson, K. and K.J. Hall, 1979. "Stormwater Discharge to the Lower Fraser River/Estuary." A Component Report i n the Water Quality Study of the Joint Federal-Provincial Fraser Estuary Study. Fraser River Estuary Study Steering Committee, 1978. "Fraser River Estuary Study - 6. Water Quality." Joint Research by the Province of B.C. and the Government of Canada. Gadd, G.M. and A.J. Griffiths, 1978. "Microorganisms and Heavy Metal Toxicity." Microbial Ecology, h}303-317. Gaddum, J.H., 1953. "Bioassays and Mathematics." Pharmacol. Rec., _5:87-134. Getz, L. et a l . , 1977. "Transport and Distribution in a Watershed Ecosystem." In Boggess, W.R. and B.G. Wixson (eds.), Lead i n the  Environment, NSF Report NSF/RA-770214, ch.6, pp. 105. Griff i n , D.M.Jr., Grizzard, T.J., Helsel, D.R. and J.P. Hartigan, 1980. "Analysis of Non-Point Pollution Export from Small Catchments." J.W.P.C.F., 52(4):780-790. Hale, J.G., 1977. "Toxicity of Metal Mining Wastes." Bull. Env. Con, and Toxicol., 17(1):66-73. 145. Hall, K.J. and K. Fletcher, 1974. "Trace Metal Pollution from a Metropolitan Area: Sources and Accumulation in the Lower Fraser River and Estuary." Proc. Inter. Conf. Transport Persistent Chemicals Aquatic Ecosystems, Ottawa, Canada 1-83. Hall, K.J., F.A. Koch and I. Yesaki, 1974. "Further Investigations i n Water Quality Conditions i n the Lower Fraser River System." Technical Report  No. 4, Westwater Research Centre, U.B.C. Hall, K.J., Yesaki, I and J. Chan, 1976. "Trace Metals and Chlorinated Hydrocarbons i n the Sediments of a Metropolitan Watershed." Technical  Report No. 10, Westwater Research Centre, U.B.C. Heaney, J.P., Field, R. and R.H. Sullivan, 1979. "Nation Wide Cost of Wet-Weather Pollution Control." J.W.P.C.F., 51(8):2043-2053. Helsel, D.R., Kim, J.I. and C.W. Randall, 1979. "Land Use Influences on Metals i n Storm Drainage." J.W.P.C.F., 51(4):709-717. Hunter, J.V. & M.J. MacKenzie, 1979. "Contribution of Urban Runoff to Hydro-carbon Pollution." J.W.P.C.F., 51(8):2129-2138. Kaufman, H.L. and F. Lai, 1980. "Cost Effective Combined Sewer Overflow Pollution Abatement Planning." J.W.P.C.F., 52(8):2220-2229. Koch, F.A., Hall, K.J. and I. Yesaki, 1977. "Toxic Substances in the Waste-waters from a Metropolitan Area." Technical Report No. 12, Westwater Research Centre, U.B.C. Krenkel, P.A., 1979. "Problems in the Establishment of Water Quality Criteria." J.W.P.C.F., 51(8):2168-2188. Laxen, D.P.H. and R.M. Harrison, 1977. "The Highway as a Source of Water Pollution: An Appraisal with the Heavy Metal Lead." Water Res., 11(1). Le, C. and T. Tenisci, 1978. "UBC TRP-Triangular Regression Package." Computing Centre, University of British Columbia. Leeuwangh, P., 1978. "Toxicity Tests with Daphnids: Its Application in the Management of Water Quality." Hydrobiologia, 59(2):145-148. Liptak, S., 1979. "Concentration Techniques for Trace Metals Analysis." Environmental Engineering, University of British Columbia. Unpublished. Litchfield, J.T.Jr. and F. Wilcoxon, 1949. "A Simplified Method of Evaluating Dose-Effect Experiments." J. Pharm. Exp. Therap., 96:99-113. Marking, L.L. and V.K. Dawson, 1975. "Method for Assessment of Toxicity or Efficacy of Mixtures of Chemicals." Investigations In Fish Control, No. 67, U.S. Dept. of Interior, Fish & Wildlife Service, Washington, D.C. Marmorek, D.R., 1979. "A C r i t i c a l Review of Fish Bioassays and Their Use In Water Pollution Control." Unpublished. 146. Marshall, J.S., 1978. "Field Verification of Cadmium Toxicity to laboratory Djipjmia Populations ." Bull. Env. Con. and Tox., 20(3) :387-393. McDonald, E., 1978. "Report on Bioassay Procedure for Daphnia." Environ-mental Engineering, University of British Columbia. Unpublished. McKim, J.M., Eaton, J.G. and G.W. Holcombe, 1978. "Metal Toxicity of Embryos and Larvae of Eight Species of Freshwater Fish - II: Copper." Bull. Env. Con, and Tox., 19(5):608-616. Mills, W.G., 1977. "Water Quality of Urban Storm Water Runoff in the Borough of East York." Research Program for the Abatement of Municipal Pollu- tion Within the Provisions of the Canada-Ontario Agreement on Great  Lakes Water Quality: Research Report No. 66. Training and Technology Transfer Division (Water), Environmental Protection Service, Fisheries and Environment Canada, Ottawa, Ontario. Newton, CD., Shephard, W.W. and M.S. Coleman, 1974. "Street Runoff as a Source of Lead Pollution." J.W.P.C.F., 46(5):999-1000. North Virginia Planning District Commission and Dept. of C i v i l Engineering, Virginia Polytechnic Institute and State University, 1977. "Interim Report: Occoquan Four Mile Run Nonpoint Source Correlation Study." Prepared for Metropolitan Washington Water Resources Planning Board. Oberts, G.L., 1977. "Water Quality Effects of Potential Urban Best Management Practices: a Literature Review." Tech. Bull 97, Dept. of Nat. Res., Madison, WIS. Parkhurst, B.R., Gehrs, C.W. and I.B. Rubin, 1979. "Value of Chemical Frac-tionation for Identifying the Toxic Components of Complex Aqueous Efflu -ents." Aquatic Toxicology, ASTM STP 667, L.L. Marking & R.A. Kimerle, Eds., American Society for Testing and Materials, pp. 122-130. Parkhurst, B.R., Bradshaw, A.S., Forte, J.L. and G.P. Wright, 1979. "An Evaluation of the Acute Toxicity to Aquatic Biota of a Coal Conversion Effluent and Its Major Components." Bull. Env. Con, and Tox., 23:349-356. Peltier, W., 1978. "Methods for Measuring the Acute Toxicity of Effluents to Aquatic Organisms." Environmental Monitoring and Support Laboratory. Office of Research and Development, U.S. EPA EPA-600/4-78-012, Cincinnati, Ohio 4568. Pitt, R.E. and G. Amy, 1973, Toxic Materials Analysis of Street Surface Contaminants, U.S. E.P.A., E.P.A.-R2-73-283. Pratt, J.M. and R.A. Coler, 1976, A Procedure for the Routine Biological Evaluation of Urban Runoff in Small Rivers. Water Research, 10:1019-1025. Reese, CD., Dodson, I.W., Ulrich, V., Becker, D.L. and C J . Kempter, 1972. "Pesticides i n the Aquatic Environment." U.S. EPA, Washington, D.C 147. Rimer, A.E., Nissen, J.A. and D.E. Reynolds, 1978. "Characterization and Impact of Stormwater Runoff From Various Land Cover Types." J.W.P.C.F., 50(2):252-264. Randall, C.W., Garland, J.A., Grizzard. T.J. and R.C. Hoehn, 1977. "The Significance of Stormwater Runoff i n an Urbanizing Watershed." Prog. Water Tech., 9_(3): 547-562. Seymour, M.D., Schubert, S.A., Clayton, J.W.,Jr. and Q. Fernando, 1978. "Variations i n the Acid Content of Rain Water in the Course of a Single Precipitation." Water, Air and Soil Pollution, 10:147-161. Sidhu, S.T., 1975(*a^. "A Preliminary Study on Wastewater Characteristics of Prince George Stormwater and Sewage Discharges." Dept. of the Environ-ment, Environmental Protection Service, Pacific Region, Vancouver, B.C., Manuscript Report //EPS 75-2. 1 9 7 5 ( b ) , "A Preliminary Study on Wastewater Characteristics of Kamloops Stormwater Discharge." Dept. of the Environment, Environmental Protection Service, Pacific Region, Vancouver, B.C., Manuscript Report  //EPS 75-3. Slooff, W., E. Rab and T.P. Spierenburg, 1978. "Continuous Standard Water Delivery System for Bioassays With Aquatic Organisms." Prog. Fish. Cult., 40(3):112-114. Spehar, R.L., Anderson, R.L. and J.T. Fiandt, 1978. "Toxicity and Bioaccumu-lation of Cadmiu and Lead i n Aquatic Invertebrates." Environ.  Pollution, 15:195-208. Sprague, J.B., 1969. "Review Paper: Measurement of Pollutant Toxicity to Fish 1. Bioassay Methods for Acute Toxicity." Water Research, 3_:793-821. Sprague, J.B., 1970. "Review Paper: Measurement of Pollutant Toxicity to Fish II. Utilizing and Applying Bioassay Results." Water Research, -4: 3-32. Sullivan, R.H. and J.P. Heaney, 1978. "Evaluation of the Magnitude and Significance of Pollution Loadings from Urban Stormwater Runoff i n Ontario." Research Program for the Abatement of Municipal Pollution  Within the Provisions of the Canada-Ontario Agreement on Great Lakes  Water Quality: Research Report No. 81. Training and Technology Transfer Division (Water), Environmental Protection Service, Fisheries and Environment Canada, Ottawa, Ontario. Sutherland, R.C. and R.H. McCuen, 1978. "Simulation of Urban Non-Point Source Pollution." Water Resources Bulletin, 14(2):409-428. U.S. EPA, 1972. "Effects of Pesticides in Water: A Report to the States." U.S. EPA, Washington, D.C. 148. Vigers, G.A. and A.W. Maynard, 1977. "The Residual Oxygen Bioassay: A Rapid Procedure to Predict Effluent Toxicity to Rainbow Trout." Water  Research, 11:343. Vitale, A.M. and P.M. Sprey, 1974. "Total Urban Water Pollution Loads: The Impact of Stormwater." U.S. Cone on Env. Quality. Wanielista, M.P., Yousef, Y.A. and W.M. McLellon, 1977. "Nonpoint Source Effects of Water Quality." J.W.P.C.F., 49:441. Whipple, W.J., Hunter, J.V. and S.L. Yu, 1977. "Effects of Storm Frequency on Pollution from Urban Runoff." J.W.P.C.F., 49:2243. Whipple, W.Jr. and J.V. Hunter, 1977. "Nonpoint Sources and Planning for Water Pollution Control." J.W.P.C.F., 49(1):15-23. Whipple, W., Hunter, J.V. and S.L. Yu, 1974. "Unrecorded Pollution from Urban Runoff." J.W.P.C.F., 46(5):873-885. Wilber, A. and J.H. Hunter, 1975,a,b. See - Oberts, G.L., 1977. Wilber, W.G. and J.V. Hunter, 1977. "Aquatic Transport of Heavy Metals i n the Urban Environment." Water Resources Bulletin, 13(4):721-734. Wilber, W.G. and J.V. Hunter, 1979. "Distribution of Metals in Street Sweep-ings, Stormwater Solids and Urban Aquatic Sediments." J.W.P.C.F., 51(12):2810-2822. Wu, J.S. and R.C. Ahlert, 1978. "Assessment of Methods for Computing Storm Runoff Loads." Water Resources Bulletin, 14(2):429-439. 149. APPENDIX A Miscellaneous Data 150. TABLE Al STORM SAMPLING AND RAINFALL DATA1 STORM DATE BUILDUP TIME SITES SAMPLED2 HOURS OF RAIN DAILY TOTAL (DAYS) (in. x 10"2) 15 AUG 79 33 1 5 6 26 SEPT 79 18 2 2 4 14 OCT 79 17 3 3 3 24 OCT 79 1.5 4 12 38 21 NOV 79 4 5-8 22 130 15 DEC 79 7 9 2 5 31 JAN 80 15 10-12 4 5 ' Rainfall Data from Sperling Avenue Monitoring Station; courtesy of G.V.R.D. See Table 3-1 for key to sampling site locations. 151. TABLE A2 DEER LAKE WATER CHEMICAL ANALYSIS COMPONENTS SAMPLING DATE 26 JUNE 1978 10 J U L Y 1978 17 J U L Y 1978 12 JUNE 1979 16 MARCH 1980 B O D 5 <2 - - - -COD 12 16 17 2 . 6 5 1 9 . 2 6 C o l o u r - A . P . H . A . 2 20 20 - 20 -T o t a l P h o s p h o r o u s <0.05 0 .14 <0.05 <0.01 -T o t a l N i t r o g e n <1.5 <1.1 <1.1 0 .56 -T o t a l R e s i d u e 127 122 120 309 -V o l a t i l e R e s i d u e 71 78 78 272 -Ammonia N i t r o g e n <0.3 <0.3 <0.3 <0.3 -7 .8 7 . 0 7 .7 6 .8 7 . 0 A l k a l i n i t y - p H 4 . 5 3 44 50 50 45 62 C o n d u c t a n c e 4 145 151 151 156 153 O r g a n i c C a r b o n 3 2 4 9 -I n o r g a n i c C a r b o n 15 14 14 8 -T o t a l H a r d n e s s 3 - - - 4 6 . 4 6 0 . 0 C a l d i u m H a r d n e s s 3 - - - 3 7 . 6 4 4 . 0 T a n n i n - l i k e Compounds , 0 . 4 8 0 .48 0 . 4 8 - -as T a n n i c A c i d Cadmium <0.001 - - <0.001 -C a l c i u m 13 14 14 1 3 . 3 1 7 . 6 C o p p e r 0 . 0 1 0 .02 0 . 0 3 0 .016 0 . 0 6 3 5 I r o n 0 . 2 8 0 . 3 0 0 . 3 3 0 . 2 4 0 .50 L e a d - - - 0 .005 <0.005 Magnes ium 3 . 2 0 3 . 2 5 3 . 4 0 3 .37 3 . 3 9 Manganese 0 .12 - - 0 . 1 3 -P o t a s s i u m 1 .6 1 .5 1 .6 1 .85 -Sodium 8 . 9 9 . 8 9 . 5 9 .75 -Z i n c 0 .02 0 .02 0 .02 0 .013 0 .019 1 A l l V a l u e s i n m g / L e x c e p t as i n d i c a t e d 2 V a l u e s u n i t l e s s 3 V a l u e s i n m g / L as C a C 0 3 ** V a l u e s i n Ps /cm 5 V a l u e u n r e a l i a b l e due to s i g n i f i c a n t d i s t i l l e d w a t e r b l a n k APPENDIX B Stormwater Data from the Literature, as Summarized by Ferguson and Hall (1979) Please Note: A l l values are in mg/L, except as follows: Conductivity in ymhos Total Hardness and Total Alkalinity in mg/L as CaCO pH unitless 153. TABLE Bl SEATTLE URBAN RUNOFF POLLUTANT CONCENTRATIONS SUMMARY VIEWRIDGE SOUTH SEATTLE SOUTHCENTRE Mean Range Mean Range Mean Range NFR 51.1 1.01-465 130 7.6-1172 19.5 4.6-291 Conductivity 77.8 1.67-271 42.3 6.6-154 13.6 24.2-235 Cd 0.004 0.003-0.006 0.005 0.004-0.012 0.004 0.002-0.050 Pb 0.198 0.008-0.71 0.21 0.05-0.59 0.48 0.08-3.5 Zn 0.089 0.005-0.30 0.20 0.06-0.54 0.13 0.003-1.6 TABLE B2 SUMMARY WESTWATER RESEARCH STILL CREEK STUDY DATA January - June, 1973 STILL CREEK AT GILMORE BRUNETTE RIVER AT BRAID STREET Mean Range Mean Range Conductance 187 117-218 144 90-167 Total Hardness 70.2 40.2-81.5 43.5 27.2-57.4 Total Alkalinity 53.1 28.1-64.0 32.9 20.3-45.0 pH 7.2 6.8-7.5 7.1 6.8-7.4 Ca 20.8 8.8-26.0 13.4 5.9-18.9 Mg 4.4 2.5-8.9 2.8 0.5-7.2 Na 11.0 6.3-19.8 9.7 5.3-13.3 Fe 1.06 0.63-1.60 1.02 0.60-1.60 Mn 0.21 0.13-0.28 0.14 0.09-0.23 Cd 0.001 0.001-0.008 0.001 0.001-0.005 Cu 0.015 0.004-0.030 0.0048 0.002-0.011 Pb 0.016 0.001-0.120 0.0116 0.001-0.031 Zn 0.008 0.033-0.100 0.017 0.002-0.060 154. TABLE B3 GVSDD - STILL CREEK CHEMICAL ANALYSES SUMMARY STATION 3 STATION A STATION B STATION C Mean Range Mean Range Mean Range Mean Range COD 30.0 7. 9-98.4 21.0 2.0-188 31.1 5.8-165 17.2 1.9-83.7 Conductance 135 30 -180 165 45-670 130 55-200 160 55-260 Hardness 60.5 14 .5-98.0 61.5 20.7-84.8 54.5 21.0-70.0 63.5 25.5-84.0 Total 40.0 8. 0-71.5 35.0 8.0-75.0 35.3 10.8-52.5 46.0 17.6-67.0 Alkalinity pH 7.1 6. 7-7.3 7.0 6.6-7.4 7.0 6.4-7.6 7.0 6.5-7.6 NFR 14 2-53 6.0 0-208 6.5 1-2445 3.0 0-50 Na 5.3 1. 2-7.8 8.1 2.5-11.7 8.6 2.3-12.9 9.1 3.9-19.8 Mg 2.7 0. 91-4.1 2.7 1.1-4.9 2.2 1.0-3.8 2.6 1.1-5.2 Ca 15.8 3. 3-23.7 16.5 4.7-30.4 14.9 6.5-43.7 17.4 8.8-55.0 Cd - 0. 002-0.01 - 0.001-0.021 - 0.001-0.02 0.01 0.001-0.02 Cu - 0. 01-0.05 - 0.01-0.13 0.04 0.006-0.09 0.04 0.04-0.04 Fe 1.0 0. 4-2.4 0.9 0.2-9.16 0.8 0.30-2.3 0.56 0.28-1.95 Mn 0.08 0. 04-0.49 0.2 0.02-0.59 0.1 0.02-0.25 0.10 0.02-0.23 Pb - 0. 13-0.16 - 0.003-0.81 0.15 0.02-0.37 0.13 0.009-0.17 Zn 0.05 0. 02-0.31 0.04 0.01-0.47 0.05 0.01-0.29 0.03 0.01-0.12 Cr - 0. 005-0.07 - 0.005-0.07 - 0.005-0.07 - 0.005-0.07 Ni - 0. 001-0.07 - 0.002-0.07 - 0.001-0.07 - 0.001-0.07 Oi l and Grease 143 TABLE B4 RESULTS OF EPS STORMWATER SAMPLING SAMPLING LOCATION Zn Cu Cr Ni Cd Pb Fe Oils COD Total Diss. Total Diss. Total Diss. Total Diss. Total Diss. Total Diss. Total Diss. & Grease pH NRF Jericho Park 27 0.12 0.11 0.01 0.02 0.02 0 02 0 05 0.05 0 01 0.01 0.04 0.02 3.0 1.1 5 6.6 10 Grandview Hwy. 65 0.18 0.14 0.04 0.01 0.02 0.19 0.02 5.2 0.40 7.1 42 Nootka 25 0.02 0.01 0.01 0.01 0.02 0.03 0 .01 1.4 0.35 7.3 5 Boundary S. 73 0.02 0.01 0.01 0.01 0.02 0.02 0.02 5.0 2.4 7.0 10 2nd Hwy. 86 0.05 0.03 0.01 0.01 0.02 0.02 9.5 2.8 6.8 23 UEL 35 2.7 0.14 0.05 0.01 0.03 0.08 4.0 0.05 7.4 72 UEL 20 0.04 0.03 0.02 0.01 0.02 0.02 0.21 0.06 7.2 5 Kyle St. 52 0.19 0.07 0.03 0.01 0.02 0 05 0.18 0.11 0.06 7.1 239 Kyle St. 20 0.02 0.01 0.01 0.02 0.02 0.01 0.02 0.55 0.17 7.3 12 Williams St. 20 1.0 0.14 0.01 0.01 0.02 0.03 1.0 0.60 7.0 12 Williams St. 20 0.02 0.03 0.01 0.01 0.02 0.02 1.2 0.73 7.1 6 Schoolhouse Cres. 20 0.10 0.12 0.01 0.01 0.02 0.03 0.80 0.30 7.3 26 Schoolhouse Cres. 20 0.01 0.01 0.01 0.01 0.02 0.02 0.44 0.19 7.5 6 Rhodes St. 51 0.13 0.10 0.03 0.01 0.04 0.20 2.3 0.13 7.1 49 Rhodes St. 20 0.03 0.02 0.01 0.01 0.02 0.02 0.53 0.08 7.5 5 Collingwood St. 59 0.17 0.10 0.04 0.01 0.02 0.25 4.4 0.03 7.4 93 Collingwood St. 20 0.04 0.02 0.02 0.01 0.02 0.03 0.98 0.09 7.2 7 Byrne Road 20 0.06 0.02 0.01 0.01 0.02 0.02 0.55 0.12 7.5 8 Byrne Road 20 0.03 0.02 0.01 0.01 0.02 0.02 0.75 0.30 7.6 23 BCIT Pond 38 0.46 0.07 0.01 0.01 0.02 0.06 2.8 0.12 7.3 34 BCIT Pond 24 0.02 0.01 0.01 0.01 0.02 0.02 0.48 0.28 8.0 6 Springer Ave. 46 0.08 0.07 0.01 0.01 0.02 0.04 1.2 1.0 7.5 31 Springer Ave. 20 0.01 0.01 0.01 0.01 0.02 0.02 0.55 0.31 7.7 5 Kaymar Raveen 20 0.02 0.01 0.01 0.01 0.02 0.02 0.60 0.03 7.7 14 Kaymar Raveen 20 0.02 0.01 0.01 0.01 0.02 0.02 0.18 0.05 7.8 5 Buckingham 20 1.3 0.05 0.01 0.01 0.04 0.02 0.58 0.12 7.9 10 Deer Lake 20 0.01 0.01 0.01 o.oi • 0.02 0.02 f 0.16 0.09 8.1 5 23rd St. 84 0.12 0.11 0.07 0.05 0.02 0.20 0 16 1.0 0.15 7.2 24 Central Area 20 0.03 0.02 0.01 0.09 0.02 0.02 0.02 2.0 0.35 7.1 6 King George 20 0.04 0.01 0.01 0.01 0.02 0.03 1.1 0.03 7.1 5 & Fraser Hwy. 22 0.05 0.02 0.03 0.01 0.02 0.03 1.1 0.03 7.4 5 88th Ave. & 20 0.01 0.01 0.01 0.01 0.02 0.02 1.3 0.15 7.8 23 King George Hwy. 20 0.01 0.01 0.01 0.01 0.03 + ^ 0.02 y 0.40 0.07 • 8.4 7 156. TABLE B5 RESULTS OF STORMWATER MONITORING PROGRAMS REPORTED IN THE LITERATURE STATION 1 NFR (mg/L) COD (mg/L) pH Average Range Average Range Average Range 1 300 — — — — _ 2a 2080 650-11900 - - - -2c 505 95-1053 - - - -2d - - 224 - - -2e - 1013 - - - -2f 81 10-1000 - - - -2g 26 - - - - -2h 30 - - - - -2i 71 3-211 58 21-176 - -2j 247 84-2052 85 12-405 - -2k - 130-11280 - 29-1514 - -3 - 15-410 - - - -4 33 1.5-207 116 3-1000 - -5 128 2.5-635 79 7-277 - -6 322 19-1251 - - 0.47 0.07-0.95 7a 227 5-1200 I l l 20-610 - -7b 313 5-2074 79 30-159 - -7d - - 188 - - -7f 14541 - - - - -7g — 1000-3500 - - - -Data obtained from following studies: 1. Values are supposedly representative of data obtained in Canada & U.S. 2. Values are averages of data from monitoring programs in following c i t i e s : a) Ann Arbor, Michigan, 1965 c) Des Moines, Iowa, 1969 d) Durham, N.C., 1968 e) Los Angeles, 1967-1968 f) Madison, Wisconsin, 1970-1971 g) New Orleans, 1967-1969 h) Roanoke, Virginia, 1969 i) Sacramento, California, 1968-1969 j) Tulsa, Oklahoma, 1968-1969 k) Washington, D.C, 1969 3. Data from study in North York, Ontario 4. E.P.S. study of 10 stormdrains (30 samples only) in Kamloops, B.C. 5. E.P.S. study of 10 stormdrains (20 samples only) in Prince George, B.C 6. Results of stormwater monitoring program conducted in Halifax, N.S. 7. Results of stormwater monitoring studies conducted in: a) Cincinnati, Ohio b) Cashocton, Ohio d) Stockholm, Sweden f) Leningrad, USSR g) Moscow, USSR APPENDIX C Complete Trace Metals Data for Phase One Sampling 158. SITE 1: Brentwood Mall (C) Sample1 1 2 3 4 5 6 Metal 2 Ca 53.70 53.70 52.80 53.56 54.75 52.80 Cd 3 18. 19. 50. 21. 14. 17. Cr <0.03 <0.03 <0.03 <0.03 <0.03 <0.03 Cu 0.45 0.38 0.31 0.31 0.29 0.28 Fe 14.30 9.40 3.50 2.59 1.95 2.33 Mg 5.45 5.15 3.68 3.23 2.75 3.02 Mn 0.64 0.56 0.42 0.38 0.36 0.38 N i 3 68. 55. 74. 43. 64. 48. Pb 1.82 2.18 . 1.88 1.95 1.72 1.91 Zn 3.00 2.77 2.70 2.62 2.31 2.52 SITE 2: Beaverbrook Dr. at Beaverbrook Cres. (R) Sample 1 2 3 4 Metal * Ca 57.1 33.6 33.5 35.0 Cd 3 7. 8. 8. 8. Cr 0.01 0.01 0.01 0.01 Cu 0.27 0.23 0.30 0.26 Fe 6.94 3.84 3.28 2.74 Mg 5.33 3.22 3.10 3.02 Mn 0.49 0.51 0.48 0.48 Ni 3 43. 37. 41. 33. Pb 2.16 1.68 1.54 1.34 Zn 2.04 2.86 3.18 3.00 159. SITE 3: Production Way at Thunderbird Cres. (I) Sample 1 2 3 4 5 Metal Ca 22.10 24.20 24.60 30.70 28.60 Cd 3 2. 2. 3. 2. 2. Cr 0.026 <0.01 <0.01 <0.01 <0.01 Cu 0.26 0.18 0.14 0.17 0.19 Fe 6.82 1.76 1.76 2.36 2.26 Mg 2.80 1.80 1.80 2.50 2.30 Mn 0.12 0.11 0.12 0.10 0.10 N i 3 6. 5. 4. 3. 3. Pb 1.19 0.40 0.38 0.40 0.35 Zn 0.82 1.10 0.83 0.67 0.72 SITE 4: Lougheed Mall (C) Sample 1 2 3 4 5 6 7 Metal Ca 10.8 4.5 3.7 3.8 3.6 3.2 2.2 Cd 3 <1. 4. <1. <1. <1. <1. <1. Cr 0.032 0.015 <0.01 <0.01 <0.01 <0.01 <0.01 Cu 0.18 0.10 0.09 0.08 0.11 0.11 0.15 Fe 30.0 2.28 1.06 1.16 1.34 1.48 1.04 Mg 3.3 1.0 0.7 0.7 0.7 0.7 0.5 Mn 0.32 0.07 0.05 0.05 0.05 0.04 0.03 N i 3 15. <2. <2. <2. <2. <2. <2. Pb 4.14 0.51 0.23 0.22 0.22 0.26 0.15 Zn 1.00 0.40 0.26 0.26 0.23 0.21 0.20 160. SITE 5: Wedgewood Ave. at 1st St. (R) Sample 1 2 3 4 5 6 Metal Ca 6.63 7.75 8.36 6.07 4.86 4.21 Cd 3 <1. <1. <1. <1. <1. <1. Cr <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Cu <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 Fe 1.07 0.70 0.42 0.72 0.50 0.63 Mg 1.20 0.90 0.42 0.83 0.13 0.47 Mn 0.080 0.062 0.042 0.052 0.032 0.042 N i 3 25. <2. <2. <2. 2. <2. Pb 0.194 0.119 0.063 0.119 0.080 0.112 Zn 0.09 0.08 0.07 0.06 0.10 5.40 SITE 6: Robert Burnaby Park (0) Sample 1 2 3 4 5 6 Metal Ca 1.21 3.78 4.47 3.71 2.89 4.47 Cd 3 <1. <1. <1. <1. <1. <1. Cr <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Cu <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 Fe 1.59 0.50 0.30 0.33 0.47 0.24 Mg 1.36 0.64 0.57 0.62 0.64 0.51 Mn 0.125 0.083 0.075 0.077 0.067 0.078 N i 3 <2. <2. <2. <2. <2. <2. Pb 0.146 0.047 0.026 0.036 0.037 0.031 Zn 0.18 0.06 0.08 0.05 0.04 0.04 161. SITE 7: Deer Lake Park (0) Sample 1 2 3 4 5 6 Metal Ca 9.44 5.87 6.76 6.33 5.98 1.88 Cd 3 <2. <2. <2. <2. <2. <2. Cr <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Cu <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 Fe 0.14 0.58 0.22 0.10 0.11 1.09 Mg 0.54 0.72 0.31 0.27 0.22 0.93 Mn 0.028 0.032 0.031 0.020 0.017 0.033 N i 3 <2. <2. 3. <2. <2. <2. Pb 0.032 0.055 0.031 0.023 0.015 0.098 Zn 0.07 0.04 0.06 0.06 0.02 0.02 SITE 8: Haszard St. at Buckingham Ave. (R) Sample 1 2 3 4 5 6 Metal Ca 9.85 8.53 8.03 7.73 7.10 5.12 Cd 3 <2. <2. <2. <2. <2. <2. Cr <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 Cu <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 Fe 0.48 0.12 0.17 0.03 0.01 0.08 Mg 0.98 0.74 0.70 0.66 0.52 0.33 Mn 0.060 0.042 0.040 0.038 0.028 0.027 N i 3 <2. <2. 72. <2. <2. 86. Pb 0.174 0.082 0.092 0.084 0.083 0.086 Zn 0.14 0.11 0.12 0.08 0.09 0.07 Each sample is discrete and there is a 20 minute runoff time between each. A l l trace metal values in mg/L, except where indicated. Values in Pg/L. 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0062987/manifest

Comment

Related Items